d0e8cbc235b811f0972ab04fa60560c8/0000755000175100017510000000000015013163137015304 5ustar jenkinsjenkinsd0e8cbc235b811f0972ab04fa60560c8/test_qwen_grpo_vllm_test_qwen_grpo_vllm.log0000644000175100017510001510755515013164131026271 0ustar jenkinsjenkins============================= test session starts ============================== platform linux -- Python 3.9.19, pytest-6.2.5, py-1.11.0, pluggy-1.5.0 rootdir: /home/jenkins/mindspore/testcases/testcases/tests/st, configfile: ../../../../../sault/virtual_test/virtualenv_0013/sault/config/pytest.ini plugins: forked-1.6.0, xdist-1.32.0, anyio-4.9.0 collected 1 item test_qwen_grpo_vllm.py WORKDIR is /home/jenkins/mindspore/testcases/testcases/tests/st PYTHONPATH is /home/jenkins/mindspore/testcases/testcases/tests/st/../../:/home/jenkins/mindspore/testcases/testcases/tests/st/mindformers/:/home/jenkins/mindspore/testcases/testcases/tests/st:/usr/local/Ascend/ascend-toolkit/latest/python/site-packages:/usr/local/Ascend/ascend-toolkit/latest/opp/built-in/op_impl/ai_core/tbe:/home/jenkins/mindspore/testcases/testcases: /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) 0%| | 0/4 [00:00 type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) Start worker process with rank id:0, log file:/home/jenkins/mindspore/testcases/testcases/tests/st/qwen2_vllm_log/worker_0.log. Environment variable [RANK_ID=0] is exported. Start worker process with rank id:1, log file:/home/jenkins/mindspore/testcases/testcases/tests/st/qwen2_vllm_log/worker_1.log. Environment variable [RANK_ID=1] is exported. Start worker process with rank id:2, log file:/home/jenkins/mindspore/testcases/testcases/tests/st/qwen2_vllm_log/worker_2.log. Environment variable [RANK_ID=2] is exported. Start worker process with rank id:3, log file:/home/jenkins/mindspore/testcases/testcases/tests/st/qwen2_vllm_log/worker_3.log. Environment variable [RANK_ID=3] is exported. Start worker process with rank id:4, log file:/home/jenkins/mindspore/testcases/testcases/tests/st/qwen2_vllm_log/worker_4.log. Environment variable [RANK_ID=4] is exported. Start worker process with rank id:5, log file:/home/jenkins/mindspore/testcases/testcases/tests/st/qwen2_vllm_log/worker_5.log. Environment variable [RANK_ID=5] is exported. Start worker process with rank id:6, log file:/home/jenkins/mindspore/testcases/testcases/tests/st/qwen2_vllm_log/worker_6.log. Environment variable [RANK_ID=6] is exported. Start worker process with rank id:7, log file:/home/jenkins/mindspore/testcases/testcases/tests/st/qwen2_vllm_log/worker_7.log. Environment variable [RANK_ID=7] is exported. /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:549: UserWarning: The value of the smallest subnormal for type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/jenkins/anaconda3/envs/ci39mindrlhf/lib/python3.9/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) 2025-05-21 04:31:02,200 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:163] - INFO - GRPOTrainer: _init_grpo_configs Namespace(config='./qwen2_5_vllm/grpo_config_st.yaml', custom_model_name='qwen', dataset_file='/home/jenkins/mindspore/testcases/testcases/tests/st/dataset/mini_gsm8k.mindrecord', tokenizer_dir='/home/jenkins/mindspore/testcases/testcases/tests/st/qwen2_5/', actor_checkpoint_path='', ref_checkpoint_path='', generate_checkpoint_path='', verifier_function='format_reward', verifier_weight='1.0', tensorboard=None, save_checkpoint_dir='/home/jenkins/mindspore/testcases/testcases/tests/st/ckpt/train') in main task 2025-05-21 04:31:02,216 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:163] - INFO - GRPOTrainer: _init_grpo_configs Namespace(config='./qwen2_5_vllm/grpo_config_st.yaml', custom_model_name='qwen', dataset_file='/home/jenkins/mindspore/testcases/testcases/tests/st/dataset/mini_gsm8k.mindrecord', tokenizer_dir='/home/jenkins/mindspore/testcases/testcases/tests/st/qwen2_5/', actor_checkpoint_path='', ref_checkpoint_path='', generate_checkpoint_path='', verifier_function='format_reward', verifier_weight='1.0', tensorboard=None, save_checkpoint_dir='/home/jenkins/mindspore/testcases/testcases/tests/st/ckpt/train') in main task 2025-05-21 04:31:02,227 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:172] - INFO - vllm mode: VllmMode.ORIGIN, hf_config_path: ./config.json 2025-05-21 04:31:02,243 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:172] - INFO - vllm mode: VllmMode.ORIGIN, hf_config_path: ./config.json 2025-05-21 04:31:02,358 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:163] - INFO - GRPOTrainer: _init_grpo_configs Namespace(config='./qwen2_5_vllm/grpo_config_st.yaml', custom_model_name='qwen', dataset_file='/home/jenkins/mindspore/testcases/testcases/tests/st/dataset/mini_gsm8k.mindrecord', tokenizer_dir='/home/jenkins/mindspore/testcases/testcases/tests/st/qwen2_5/', actor_checkpoint_path='', ref_checkpoint_path='', generate_checkpoint_path='', verifier_function='format_reward', verifier_weight='1.0', tensorboard=None, save_checkpoint_dir='/home/jenkins/mindspore/testcases/testcases/tests/st/ckpt/train') in main task 2025-05-21 04:31:02,385 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:172] - INFO - vllm mode: VllmMode.ORIGIN, hf_config_path: ./config.json 2025-05-21 04:31:02,526 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:163] - INFO - GRPOTrainer: _init_grpo_configs Namespace(config='./qwen2_5_vllm/grpo_config_st.yaml', custom_model_name='qwen', dataset_file='/home/jenkins/mindspore/testcases/testcases/tests/st/dataset/mini_gsm8k.mindrecord', tokenizer_dir='/home/jenkins/mindspore/testcases/testcases/tests/st/qwen2_5/', actor_checkpoint_path='', ref_checkpoint_path='', generate_checkpoint_path='', verifier_function='format_reward', verifier_weight='1.0', tensorboard=None, save_checkpoint_dir='/home/jenkins/mindspore/testcases/testcases/tests/st/ckpt/train') in main task 2025-05-21 04:31:02,553 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:163] - INFO - GRPOTrainer: _init_grpo_configs Namespace(config='./qwen2_5_vllm/grpo_config_st.yaml', custom_model_name='qwen', dataset_file='/home/jenkins/mindspore/testcases/testcases/tests/st/dataset/mini_gsm8k.mindrecord', tokenizer_dir='/home/jenkins/mindspore/testcases/testcases/tests/st/qwen2_5/', actor_checkpoint_path='', ref_checkpoint_path='', generate_checkpoint_path='', verifier_function='format_reward', verifier_weight='1.0', tensorboard=None, save_checkpoint_dir='/home/jenkins/mindspore/testcases/testcases/tests/st/ckpt/train') in main task 2025-05-21 04:31:02,554 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:172] - INFO - vllm mode: VllmMode.ORIGIN, hf_config_path: ./config.json 2025-05-21 04:31:02,581 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:172] - INFO - vllm mode: VllmMode.ORIGIN, hf_config_path: ./config.json 2025-05-21 04:31:02,632 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:213] - INFO - GRPOTrainer: _init_reward_fn 2025-05-21 04:31:02,633 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:222] - INFO - verifier_function_list:['format_reward'] 2025-05-21 04:31:02,633 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:223] - INFO - verifier_weight:[1.0] 2025-05-21 04:31:02,634 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:87] - INFO - GRPOTrainer: start init workers 2025-05-21 04:31:02,634 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:70] - INFO - init InferWorker 2025-05-21 04:31:02,636 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:213] - INFO - GRPOTrainer: _init_reward_fn 2025-05-21 04:31:02,637 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:222] - INFO - verifier_function_list:['format_reward'] 2025-05-21 04:31:02,637 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:223] - INFO - verifier_weight:[1.0] 2025-05-21 04:31:02,638 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:87] - INFO - GRPOTrainer: start init workers 2025-05-21 04:31:02,638 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:70] - INFO - init InferWorker 2025-05-21 04:31:02,659 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:02,659 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:02,659 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:02,659 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:02,660 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:02,660 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:02,661 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:77] - INFO - generate parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:02,661 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:87] - INFO - launch actor roll out sft_config_infer.use_parallel True 2025-05-21 04:31:02,661 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:89] - INFO - sft_config_infer.context:{'mode': 0, 'device_target': 'Ascend', 'max_call_depth': 10000, 'max_device_memory': '55GB', 'save_graphs': False, 'save_graphs_path': './graph', 'device_id': 0, 'jit_config': {'jit_level': 'O0'}, 'memory_optimize_level': 'O0', 'ascend_config': {'precision_mode': 'must_keep_origin_dtype'}} 2025-05-21 04:31:02,662 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:02,662 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:02,662 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:02,662 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:02,663 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:02,663 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:02,663 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:77] - INFO - generate parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:02,664 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:87] - INFO - launch actor roll out sft_config_infer.use_parallel True 2025-05-21 04:31:02,664 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:89] - INFO - sft_config_infer.context:{'mode': 0, 'device_target': 'Ascend', 'max_call_depth': 10000, 'max_device_memory': '55GB', 'save_graphs': False, 'save_graphs_path': './graph', 'device_id': 0, 'jit_config': {'jit_level': 'O0'}, 'memory_optimize_level': 'O0', 'ascend_config': {'precision_mode': 'must_keep_origin_dtype'}} 2025-05-21 04:31:02,793 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:163] - INFO - GRPOTrainer: _init_grpo_configs Namespace(config='./qwen2_5_vllm/grpo_config_st.yaml', custom_model_name='qwen', dataset_file='/home/jenkins/mindspore/testcases/testcases/tests/st/dataset/mini_gsm8k.mindrecord', tokenizer_dir='/home/jenkins/mindspore/testcases/testcases/tests/st/qwen2_5/', actor_checkpoint_path='', ref_checkpoint_path='', generate_checkpoint_path='', verifier_function='format_reward', verifier_weight='1.0', tensorboard=None, save_checkpoint_dir='/home/jenkins/mindspore/testcases/testcases/tests/st/ckpt/train') in main task 2025-05-21 04:31:02,807 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:213] - INFO - GRPOTrainer: _init_reward_fn 2025-05-21 04:31:02,808 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:222] - INFO - verifier_function_list:['format_reward'] 2025-05-21 04:31:02,808 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:223] - INFO - verifier_weight:[1.0] 2025-05-21 04:31:02,808 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:87] - INFO - GRPOTrainer: start init workers 2025-05-21 04:31:02,808 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:70] - INFO - init InferWorker 2025-05-21 04:31:02,823 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:172] - INFO - vllm mode: VllmMode.ORIGIN, hf_config_path: ./config.json 2025-05-21 04:31:02,832 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:02,832 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:02,832 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:02,833 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:02,833 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:02,833 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:02,834 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:77] - INFO - generate parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:02,834 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:87] - INFO - launch actor roll out sft_config_infer.use_parallel True 2025-05-21 04:31:02,834 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:89] - INFO - sft_config_infer.context:{'mode': 0, 'device_target': 'Ascend', 'max_call_depth': 10000, 'max_device_memory': '55GB', 'save_graphs': False, 'save_graphs_path': './graph', 'device_id': 0, 'jit_config': {'jit_level': 'O0'}, 'memory_optimize_level': 'O0', 'ascend_config': {'precision_mode': 'must_keep_origin_dtype'}} 2025-05-21 04:31:02,894 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:163] - INFO - GRPOTrainer: _init_grpo_configs Namespace(config='./qwen2_5_vllm/grpo_config_st.yaml', custom_model_name='qwen', dataset_file='/home/jenkins/mindspore/testcases/testcases/tests/st/dataset/mini_gsm8k.mindrecord', tokenizer_dir='/home/jenkins/mindspore/testcases/testcases/tests/st/qwen2_5/', actor_checkpoint_path='', ref_checkpoint_path='', generate_checkpoint_path='', verifier_function='format_reward', verifier_weight='1.0', tensorboard=None, save_checkpoint_dir='/home/jenkins/mindspore/testcases/testcases/tests/st/ckpt/train') in main task 2025-05-21 04:31:02,922 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:172] - INFO - vllm mode: VllmMode.ORIGIN, hf_config_path: ./config.json 2025-05-21 04:31:02,973 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:213] - INFO - GRPOTrainer: _init_reward_fn 2025-05-21 04:31:02,973 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:222] - INFO - verifier_function_list:['format_reward'] 2025-05-21 04:31:02,974 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:223] - INFO - verifier_weight:[1.0] 2025-05-21 04:31:02,974 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:87] - INFO - GRPOTrainer: start init workers 2025-05-21 04:31:02,974 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:70] - INFO - init InferWorker 2025-05-21 04:31:02,998 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:02,998 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:02,998 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:02,998 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:02,999 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:02,999 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:02,999 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:77] - INFO - generate parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:03,000 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:87] - INFO - launch actor roll out sft_config_infer.use_parallel True 2025-05-21 04:31:03,000 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:89] - INFO - sft_config_infer.context:{'mode': 0, 'device_target': 'Ascend', 'max_call_depth': 10000, 'max_device_memory': '55GB', 'save_graphs': False, 'save_graphs_path': './graph', 'device_id': 0, 'jit_config': {'jit_level': 'O0'}, 'memory_optimize_level': 'O0', 'ascend_config': {'precision_mode': 'must_keep_origin_dtype'}} 2025-05-21 04:31:03,016 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:213] - INFO - GRPOTrainer: _init_reward_fn 2025-05-21 04:31:03,017 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:222] - INFO - verifier_function_list:['format_reward'] 2025-05-21 04:31:03,017 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:223] - INFO - verifier_weight:[1.0] 2025-05-21 04:31:03,018 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:87] - INFO - GRPOTrainer: start init workers 2025-05-21 04:31:03,018 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:70] - INFO - init InferWorker 2025-05-21 04:31:03,042 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:03,042 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:03,042 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:03,042 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:03,043 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:03,043 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:03,044 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:77] - INFO - generate parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:03,044 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:87] - INFO - launch actor roll out sft_config_infer.use_parallel True 2025-05-21 04:31:03,044 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:89] - INFO - sft_config_infer.context:{'mode': 0, 'device_target': 'Ascend', 'max_call_depth': 10000, 'max_device_memory': '55GB', 'save_graphs': False, 'save_graphs_path': './graph', 'device_id': 0, 'jit_config': {'jit_level': 'O0'}, 'memory_optimize_level': 'O0', 'ascend_config': {'precision_mode': 'must_keep_origin_dtype'}} 2025-05-21 04:31:03,074 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:163] - INFO - GRPOTrainer: _init_grpo_configs Namespace(config='./qwen2_5_vllm/grpo_config_st.yaml', custom_model_name='qwen', dataset_file='/home/jenkins/mindspore/testcases/testcases/tests/st/dataset/mini_gsm8k.mindrecord', tokenizer_dir='/home/jenkins/mindspore/testcases/testcases/tests/st/qwen2_5/', actor_checkpoint_path='', ref_checkpoint_path='', generate_checkpoint_path='', verifier_function='format_reward', verifier_weight='1.0', tensorboard=None, save_checkpoint_dir='/home/jenkins/mindspore/testcases/testcases/tests/st/ckpt/train') in main task 2025-05-21 04:31:03,101 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:172] - INFO - vllm mode: VllmMode.ORIGIN, hf_config_path: ./config.json 2025-05-21 04:31:03,234 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:213] - INFO - GRPOTrainer: _init_reward_fn 2025-05-21 04:31:03,234 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:222] - INFO - verifier_function_list:['format_reward'] 2025-05-21 04:31:03,234 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:223] - INFO - verifier_weight:[1.0] 2025-05-21 04:31:03,235 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:87] - INFO - GRPOTrainer: start init workers 2025-05-21 04:31:03,235 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:70] - INFO - init InferWorker 2025-05-21 04:31:03,260 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:03,260 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:03,261 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:03,261 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:03,261 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:03,261 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:03,262 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:77] - INFO - generate parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:03,262 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:87] - INFO - launch actor roll out sft_config_infer.use_parallel True 2025-05-21 04:31:03,262 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:89] - INFO - sft_config_infer.context:{'mode': 0, 'device_target': 'Ascend', 'max_call_depth': 10000, 'max_device_memory': '55GB', 'save_graphs': False, 'save_graphs_path': './graph', 'device_id': 0, 'jit_config': {'jit_level': 'O0'}, 'memory_optimize_level': 'O0', 'ascend_config': {'precision_mode': 'must_keep_origin_dtype'}} 2025-05-21 04:31:03,287 - mindformers./output/log[mindformers/core/context/build_context.py:168] - INFO - Predict context config, jit_level: O0, infer_boost: on 2025-05-21 04:31:03,316 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:213] - INFO - GRPOTrainer: _init_reward_fn 2025-05-21 04:31:03,316 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:222] - INFO - verifier_function_list:['format_reward'] 2025-05-21 04:31:03,316 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:223] - INFO - verifier_weight:[1.0] 2025-05-21 04:31:03,317 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:87] - INFO - GRPOTrainer: start init workers 2025-05-21 04:31:03,317 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:70] - INFO - init InferWorker 2025-05-21 04:31:03,340 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:03,341 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:03,341 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:03,341 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:03,341 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:03,342 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:03,342 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:77] - INFO - generate parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:03,342 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:87] - INFO - launch actor roll out sft_config_infer.use_parallel True 2025-05-21 04:31:03,343 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:89] - INFO - sft_config_infer.context:{'mode': 0, 'device_target': 'Ascend', 'max_call_depth': 10000, 'max_device_memory': '55GB', 'save_graphs': False, 'save_graphs_path': './graph', 'device_id': 0, 'jit_config': {'jit_level': 'O0'}, 'memory_optimize_level': 'O0', 'ascend_config': {'precision_mode': 'must_keep_origin_dtype'}} 2025-05-21 04:31:03,391 - mindformers./output/log[mindformers/core/context/build_context.py:168] - INFO - Predict context config, jit_level: O0, infer_boost: on 2025-05-21 04:31:03,472 - mindformers./output/log[mindformers/core/context/build_context.py:168] - INFO - Predict context config, jit_level: O0, infer_boost: on 2025-05-21 04:31:03,484 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:213] - INFO - GRPOTrainer: _init_reward_fn 2025-05-21 04:31:03,484 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:222] - INFO - verifier_function_list:['format_reward'] 2025-05-21 04:31:03,485 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:223] - INFO - verifier_weight:[1.0] 2025-05-21 04:31:03,485 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:87] - INFO - GRPOTrainer: start init workers 2025-05-21 04:31:03,485 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:70] - INFO - init InferWorker 2025-05-21 04:31:03,509 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:03,509 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:03,509 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:03,509 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:03,510 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:03,510 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:03,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:77] - INFO - generate parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:03,511 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:87] - INFO - launch actor roll out sft_config_infer.use_parallel True 2025-05-21 04:31:03,511 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:89] - INFO - sft_config_infer.context:{'mode': 0, 'device_target': 'Ascend', 'max_call_depth': 10000, 'max_device_memory': '55GB', 'save_graphs': False, 'save_graphs_path': './graph', 'device_id': 0, 'jit_config': {'jit_level': 'O0'}, 'memory_optimize_level': 'O0', 'ascend_config': {'precision_mode': 'must_keep_origin_dtype'}} 2025-05-21 04:31:03,612 - mindformers./output/log[mindformers/core/context/build_context.py:168] - INFO - Predict context config, jit_level: O0, infer_boost: on 2025-05-21 04:31:03,674 - mindformers./output/log[mindformers/core/context/build_context.py:168] - INFO - Predict context config, jit_level: O0, infer_boost: on 2025-05-21 04:31:03,876 - mindformers./output/log[mindformers/core/context/build_context.py:168] - INFO - Predict context config, jit_level: O0, infer_boost: on 2025-05-21 04:31:03,961 - mindformers./output/log[mindformers/core/context/build_context.py:168] - INFO - Predict context config, jit_level: O0, infer_boost: on 2025-05-21 04:31:04,125 - mindformers./output/log[mindformers/core/context/build_context.py:168] - INFO - Predict context config, jit_level: O0, infer_boost: on [MS_ALLOC_CONF]Runtime config: enable_vmm:False [MS_ALLOC_CONF]Runtime config: enable_vmm:False [MS_ALLOC_CONF]Runtime config: enable_vmm:False [MS_ALLOC_CONF]Runtime config: enable_vmm:False [MS_ALLOC_CONF]Runtime config: enable_vmm:False [MS_ALLOC_CONF]Runtime config: enable_vmm:False [MS_ALLOC_CONF]Runtime config: enable_vmm:False [MS_ALLOC_CONF]Runtime config: enable_vmm:False 2025-05-21 04:31:07,182 - mindformers./output/log[mindformers/tools/utils.py:181] - INFO - set strategy path to './output/strategy/ckpt_strategy_rank_4.ckpt' 2025-05-21 04:31:07,186 - mindformers./output/log[mindformers/core/context/build_context.py:383] - INFO - cann workqueue cpus: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254] 2025-05-21 04:31:07,186 - mindformers./output/log[mindformers/core/context/build_context.py:387] - WARNING - CANN use cpus: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254], model get empty cpu list, disable binding cores 2025-05-21 04:31:07,187 - mindformers./output/log[mindformers/core/context/build_context.py:395] - INFO - cpu_affinity, rank_id: 4, device_num: 8 2025-05-21 04:31:07,257 - mindformers./output/log[mindformers/tools/utils.py:181] - INFO - set strategy path to './output/strategy/ckpt_strategy_rank_3.ckpt' 2025-05-21 04:31:07,260 - mindformers./output/log[mindformers/core/context/build_context.py:383] - INFO - cann workqueue cpus: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254] 2025-05-21 04:31:07,260 - mindformers./output/log[mindformers/core/context/build_context.py:387] - WARNING - CANN use cpus: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254], model get empty cpu list, disable binding cores 2025-05-21 04:31:07,260 - mindformers./output/log[mindformers/core/context/build_context.py:395] - INFO - cpu_affinity, rank_id: 3, device_num: 8 2025-05-21 04:31:07,300 - mindformers./output/log[mindformers/tools/utils.py:181] - INFO - set strategy path to './output/strategy/ckpt_strategy_rank_0.ckpt' 2025-05-21 04:31:07,307 - mindformers./output/log[mindformers/core/context/build_context.py:383] - INFO - cann workqueue cpus: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254] 2025-05-21 04:31:07,307 - mindformers./output/log[mindformers/core/context/build_context.py:387] - WARNING - CANN use cpus: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254], model get empty cpu list, disable binding cores 2025-05-21 04:31:07,308 - mindformers./output/log[mindformers/core/context/build_context.py:395] - INFO - cpu_affinity, rank_id: 0, device_num: 8 2025-05-21 04:31:07,381 - mindformers./output/log[mindformers/tools/utils.py:181] - INFO - set strategy path to './output/strategy/ckpt_strategy_rank_6.ckpt' 2025-05-21 04:31:07,385 - mindformers./output/log[mindformers/core/context/build_context.py:383] - INFO - cann workqueue cpus: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254] 2025-05-21 04:31:07,386 - mindformers./output/log[mindformers/core/context/build_context.py:387] - WARNING - CANN use cpus: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254], model get empty cpu list, disable binding cores 2025-05-21 04:31:07,386 - mindformers./output/log[mindformers/core/context/build_context.py:395] - INFO - cpu_affinity, rank_id: 6, device_num: 8 2025-05-21 04:31:07,409 - mindformers./output/log[mindformers/tools/utils.py:181] - INFO - set strategy path to './output/strategy/ckpt_strategy_rank_1.ckpt' 2025-05-21 04:31:07,412 - mindformers./output/log[mindformers/core/context/build_context.py:383] - INFO - cann workqueue cpus: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254] 2025-05-21 04:31:07,412 - mindformers./output/log[mindformers/core/context/build_context.py:387] - WARNING - CANN use cpus: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254], model get empty cpu list, disable binding cores 2025-05-21 04:31:07,413 - mindformers./output/log[mindformers/core/context/build_context.py:395] - INFO - cpu_affinity, rank_id: 1, device_num: 8 2025-05-21 04:31:07,560 - mindformers./output/log[mindformers/tools/utils.py:181] - INFO - set strategy path to './output/strategy/ckpt_strategy_rank_5.ckpt' 2025-05-21 04:31:07,563 - mindformers./output/log[mindformers/core/context/build_context.py:383] - INFO - cann workqueue cpus: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254] 2025-05-21 04:31:07,563 - mindformers./output/log[mindformers/core/context/build_context.py:387] - WARNING - CANN use cpus: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254], model get empty cpu list, disable binding cores 2025-05-21 04:31:07,564 - mindformers./output/log[mindformers/core/context/build_context.py:395] - INFO - cpu_affinity, rank_id: 5, device_num: 8 2025-05-21 04:31:07,567 - mindformers./output/log[mindformers/tools/utils.py:181] - INFO - set strategy path to './output/strategy/ckpt_strategy_rank_7.ckpt' 2025-05-21 04:31:07,569 - mindformers./output/log[mindformers/core/context/build_context.py:383] - INFO - cann workqueue cpus: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254] 2025-05-21 04:31:07,570 - mindformers./output/log[mindformers/core/context/build_context.py:387] - WARNING - CANN use cpus: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254], model get empty cpu list, disable binding cores 2025-05-21 04:31:07,570 - mindformers./output/log[mindformers/core/context/build_context.py:395] - INFO - cpu_affinity, rank_id: 7, device_num: 8 2025-05-21 04:31:07,589 - mindformers./output/log[mindformers/tools/utils.py:181] - INFO - set strategy path to './output/strategy/ckpt_strategy_rank_2.ckpt' 2025-05-21 04:31:07,592 - mindformers./output/log[mindformers/core/context/build_context.py:383] - INFO - cann workqueue cpus: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254] 2025-05-21 04:31:07,592 - mindformers./output/log[mindformers/core/context/build_context.py:387] - WARNING - CANN use cpus: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254], model get empty cpu list, disable binding cores 2025-05-21 04:31:07,592 - mindformers./output/log[mindformers/core/context/build_context.py:395] - INFO - cpu_affinity, rank_id: 2, device_num: 8 2025-05-21 04:31:14,432 - mindformers./output/log[mindformers/core/parallel_config.py:41] - INFO - initial moe_config from dict: {'expert_num': 1, 'capacity_factor': 1.1, 'aux_loss_factor': 0.05, 'num_experts_chosen': 1, 'expert_group_size': None, 'group_wise_a2a': False, 'comp_comm_parallel': False, 'comp_comm_parallel_degree': 2, 'save_token_distribution': False, 'cur_layer': 0, 'enable_cold_hot_expert': False, 'update_step': 10000, 'hot_expert_num': 0, 'cold_token_percent': 1.0, 'moe_module_name': '', 'routing_policy': 'TopkRouterV1', 'norm_topk_prob': True, 'enable_sdrop': False, 'use_fused_ops_topkrouter': False, 'router_dense_type': 'float32', 'shared_expert_num': 0, 'use_shared_expert_gating': False, 'max_router_load': 131072, 'topk_method': 'greedy', 'topk_group': None, 'n_group': None, 'first_k_dense_replace': True, 'moe_intermediate_size': 1407, 'routed_scaling_factor': 1.0, 'aux_loss_types': None, 'aux_loss_factors': None, 'z_loss_factor': 0.0, 'balance_via_topk_bias': False, 'topk_bias_update_rate': 0.0, 'use_allgather_dispatcher': False, 'moe_shared_expert_overlap': False, 'expert_model_parallel': None, 'use_gating_sigmoid': False, 'enable_deredundency': False, 'npu_nums_per_device': 1, 'use_gmm': False, 'enable_gmm_safe_tokens': False, 'use_fused_ops_permute': False, 'callback_moe_droprate': False} 2025-05-21 04:31:14,434 - mindformers./output/log[mindformers/core/parallel_config.py:61] - INFO - initial parallel_config from dict: {'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:14,435 - mindformers./output/log[mindformers/core/parallel_config.py:63] - INFO - pipeline_stage = 2 > 1, vocab_emd_dp will be reset to False. 2025-05-21 04:31:14,437 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:101] - INFO - sft_config_infer: {'runner_config': {'batch_size': 1, 'epochs': 1, 'sink_mode': 1, 'sink_size': 1, 'gradient_accumulation_steps': 1, 'num_classes': 1, 'stop_step': 0}, 'context': {'mode': 0, 'device_target': 'Ascend', 'max_call_depth': 10000, 'max_device_memory': '55GB', 'save_graphs': False, 'save_graphs_path': './graph', 'device_id': 0, 'jit_config': {'jit_level': 'O0'}, 'memory_optimize_level': 'O0', 'ascend_config': {'precision_mode': 'must_keep_origin_dtype'}}, 'parallel': {'parallel_mode': 1, 'full_batch': True, 'search_mode': 'sharding_propagation', 'enable_parallel_optimizer': False, 'gradients_mean': False, 'enable_alltoall': False, 'strategy_ckpt_save_file': './ckpt_strategy.ckpt'}, 'trainer': {}, 'model': {'model_config': {'type': 'LlamaConfig', 'batch_size': 1, 'seq_length': 8192, 'hidden_size': 3584, 'num_layers': 2, 'num_heads': 28, 'n_kv_heads': 4, 'vocab_size': 152064, 'intermediate_size': 18944, 'max_position_embeddings': 32768, 'qkv_has_bias': True, 'rms_norm_eps': 1e-06, 'theta': 1000000.0, 'emb_dropout_prob': 0.0, 'eos_token_id': [151645, 151643], 'pad_token_id': 151643, 'bos_token_id': 151643, 'compute_dtype': 'bfloat16', 'layernorm_compute_type': 'float32', 'softmax_compute_type': 'float16', 'rotary_dtype': 'bfloat16', 'param_init_type': 'float32', 'use_past': True, 'use_flash_attention': True, 'block_size': 32, 'num_blocks': 1024, 'use_past_shard': False, 'offset': 0, 'checkpoint_name_or_path': '', 'repetition_penalty': 1.0, 'max_decode_length': 512, 'min_decode_length': 2, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'do_sample': True, 'is_dynamic': True, 'qkv_concat': False, 'auto_map': {'AutoTokenizer': ['qwen2_tokenizer.Qwen2Tokenizer', None]}, 'parallel_config': }, 'arch': {'type': 'LlamaForCausalLM'}}, 'moe_config': , 'parallel_config': , 'processor': {'return_tensors': 'ms', 'tokenizer': {'model_max_length': 32768, 'vocab_file': '/path/vocab.json', 'merges_file': '/path/merges.txt', 'unk_token': '<|endoftext|>', 'pad_token': '<|endoftext|>', 'eos_token': '<|im_end|>', 'chat_template': "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}", 'type': 'Qwen2Tokenizer', 'auto_register': 'qwen2_tokenizer.Qwen2Tokenizer'}, 'type': 'Qwen2Processor'}, 'seed': 1, 'output_dir': './output', 'run_mode': 'predict', 'use_parallel': True, 'resume_training': False, 'load_checkpoint': '', 'load_ckpt_format': 'ckpt', 'auto_trans_ckpt': False, 'transform_process_num': 1, 'src_strategy_path_or_dir': '', 'only_save_strategy': False, 'load_ckpt_async': False, 'use_legacy': True, 'do_eval': False, 'eval_step_interval': 100, 'eval_epoch_interval': -1, 'ignore_data_skip': False, 'data_skip_steps': None, 'profile': False, 'profile_communication': False, 'profile_memory': True, 'init_start_profile': False, 'profile_start_step': 1, 'profile_stop_step': 10, 'profile_rank_ids': None, 'profile_pipeline': False, 'profile_level': 1, 'mstx': False, 'layer_scale': False, 'layer_decay': 0.65, 'lr_scale': False, 'lr_scale_factor': 256, 'micro_batch_interleave_num': 1, 'remote_save_url': None, 'save_file': None, 'input_data': None, 'predict_batch_size': None, 'adapter_id': None, 'exclude_cann_cpu': False, 'train_precision_sync': None, 'infer_precision_sync': None, 'postprocess_use_numpy': False, 'local_rank': 6, 'device_num': 8, 'swap_config': , 'recompute_config': } 2025-05-21 04:31:14,835 - mindformers./output/log[mindformers/core/parallel_config.py:41] - INFO - initial moe_config from dict: {'expert_num': 1, 'capacity_factor': 1.1, 'aux_loss_factor': 0.05, 'num_experts_chosen': 1, 'expert_group_size': None, 'group_wise_a2a': False, 'comp_comm_parallel': False, 'comp_comm_parallel_degree': 2, 'save_token_distribution': False, 'cur_layer': 0, 'enable_cold_hot_expert': False, 'update_step': 10000, 'hot_expert_num': 0, 'cold_token_percent': 1.0, 'moe_module_name': '', 'routing_policy': 'TopkRouterV1', 'norm_topk_prob': True, 'enable_sdrop': False, 'use_fused_ops_topkrouter': False, 'router_dense_type': 'float32', 'shared_expert_num': 0, 'use_shared_expert_gating': False, 'max_router_load': 131072, 'topk_method': 'greedy', 'topk_group': None, 'n_group': None, 'first_k_dense_replace': True, 'moe_intermediate_size': 1407, 'routed_scaling_factor': 1.0, 'aux_loss_types': None, 'aux_loss_factors': None, 'z_loss_factor': 0.0, 'balance_via_topk_bias': False, 'topk_bias_update_rate': 0.0, 'use_allgather_dispatcher': False, 'moe_shared_expert_overlap': False, 'expert_model_parallel': None, 'use_gating_sigmoid': False, 'enable_deredundency': False, 'npu_nums_per_device': 1, 'use_gmm': False, 'enable_gmm_safe_tokens': False, 'use_fused_ops_permute': False, 'callback_moe_droprate': False} 2025-05-21 04:31:14,837 - mindformers./output/log[mindformers/core/parallel_config.py:61] - INFO - initial parallel_config from dict: {'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:14,837 - mindformers./output/log[mindformers/core/parallel_config.py:63] - INFO - pipeline_stage = 2 > 1, vocab_emd_dp will be reset to False. 2025-05-21 04:31:14,839 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:101] - INFO - sft_config_infer: {'runner_config': {'batch_size': 1, 'epochs': 1, 'sink_mode': 1, 'sink_size': 1, 'gradient_accumulation_steps': 1, 'num_classes': 1, 'stop_step': 0}, 'context': {'mode': 0, 'device_target': 'Ascend', 'max_call_depth': 10000, 'max_device_memory': '55GB', 'save_graphs': False, 'save_graphs_path': './graph', 'device_id': 0, 'jit_config': {'jit_level': 'O0'}, 'memory_optimize_level': 'O0', 'ascend_config': {'precision_mode': 'must_keep_origin_dtype'}}, 'parallel': {'parallel_mode': 1, 'full_batch': True, 'search_mode': 'sharding_propagation', 'enable_parallel_optimizer': False, 'gradients_mean': False, 'enable_alltoall': False, 'strategy_ckpt_save_file': './ckpt_strategy.ckpt'}, 'trainer': {}, 'model': {'model_config': {'type': 'LlamaConfig', 'batch_size': 1, 'seq_length': 8192, 'hidden_size': 3584, 'num_layers': 2, 'num_heads': 28, 'n_kv_heads': 4, 'vocab_size': 152064, 'intermediate_size': 18944, 'max_position_embeddings': 32768, 'qkv_has_bias': True, 'rms_norm_eps': 1e-06, 'theta': 1000000.0, 'emb_dropout_prob': 0.0, 'eos_token_id': [151645, 151643], 'pad_token_id': 151643, 'bos_token_id': 151643, 'compute_dtype': 'bfloat16', 'layernorm_compute_type': 'float32', 'softmax_compute_type': 'float16', 'rotary_dtype': 'bfloat16', 'param_init_type': 'float32', 'use_past': True, 'use_flash_attention': True, 'block_size': 32, 'num_blocks': 1024, 'use_past_shard': False, 'offset': 0, 'checkpoint_name_or_path': '', 'repetition_penalty': 1.0, 'max_decode_length': 512, 'min_decode_length': 2, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'do_sample': True, 'is_dynamic': True, 'qkv_concat': False, 'auto_map': {'AutoTokenizer': ['qwen2_tokenizer.Qwen2Tokenizer', None]}, 'parallel_config': }, 'arch': {'type': 'LlamaForCausalLM'}}, 'moe_config': , 'parallel_config': , 'processor': {'return_tensors': 'ms', 'tokenizer': {'model_max_length': 32768, 'vocab_file': '/path/vocab.json', 'merges_file': '/path/merges.txt', 'unk_token': '<|endoftext|>', 'pad_token': '<|endoftext|>', 'eos_token': '<|im_end|>', 'chat_template': "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}", 'type': 'Qwen2Tokenizer', 'auto_register': 'qwen2_tokenizer.Qwen2Tokenizer'}, 'type': 'Qwen2Processor'}, 'seed': 1, 'output_dir': './output', 'run_mode': 'predict', 'use_parallel': True, 'resume_training': False, 'load_checkpoint': '', 'load_ckpt_format': 'ckpt', 'auto_trans_ckpt': False, 'transform_process_num': 1, 'src_strategy_path_or_dir': '', 'only_save_strategy': False, 'load_ckpt_async': False, 'use_legacy': True, 'do_eval': False, 'eval_step_interval': 100, 'eval_epoch_interval': -1, 'ignore_data_skip': False, 'data_skip_steps': None, 'profile': False, 'profile_communication': False, 'profile_memory': True, 'init_start_profile': False, 'profile_start_step': 1, 'profile_stop_step': 10, 'profile_rank_ids': None, 'profile_pipeline': False, 'profile_level': 1, 'mstx': False, 'layer_scale': False, 'layer_decay': 0.65, 'lr_scale': False, 'lr_scale_factor': 256, 'micro_batch_interleave_num': 1, 'remote_save_url': None, 'save_file': None, 'input_data': None, 'predict_batch_size': None, 'adapter_id': None, 'exclude_cann_cpu': False, 'train_precision_sync': None, 'infer_precision_sync': None, 'postprocess_use_numpy': False, 'local_rank': 3, 'device_num': 8, 'swap_config': , 'recompute_config': } 2025-05-21 04:31:15,183 - mindformers./output/log[mindformers/core/parallel_config.py:41] - INFO - initial moe_config from dict: {'expert_num': 1, 'capacity_factor': 1.1, 'aux_loss_factor': 0.05, 'num_experts_chosen': 1, 'expert_group_size': None, 'group_wise_a2a': False, 'comp_comm_parallel': False, 'comp_comm_parallel_degree': 2, 'save_token_distribution': False, 'cur_layer': 0, 'enable_cold_hot_expert': False, 'update_step': 10000, 'hot_expert_num': 0, 'cold_token_percent': 1.0, 'moe_module_name': '', 'routing_policy': 'TopkRouterV1', 'norm_topk_prob': True, 'enable_sdrop': False, 'use_fused_ops_topkrouter': False, 'router_dense_type': 'float32', 'shared_expert_num': 0, 'use_shared_expert_gating': False, 'max_router_load': 131072, 'topk_method': 'greedy', 'topk_group': None, 'n_group': None, 'first_k_dense_replace': True, 'moe_intermediate_size': 1407, 'routed_scaling_factor': 1.0, 'aux_loss_types': None, 'aux_loss_factors': None, 'z_loss_factor': 0.0, 'balance_via_topk_bias': False, 'topk_bias_update_rate': 0.0, 'use_allgather_dispatcher': False, 'moe_shared_expert_overlap': False, 'expert_model_parallel': None, 'use_gating_sigmoid': False, 'enable_deredundency': False, 'npu_nums_per_device': 1, 'use_gmm': False, 'enable_gmm_safe_tokens': False, 'use_fused_ops_permute': False, 'callback_moe_droprate': False} 2025-05-21 04:31:15,184 - mindformers./output/log[mindformers/core/parallel_config.py:61] - INFO - initial parallel_config from dict: {'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:15,185 - mindformers./output/log[mindformers/core/parallel_config.py:63] - INFO - pipeline_stage = 2 > 1, vocab_emd_dp will be reset to False. 2025-05-21 04:31:15,187 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:101] - INFO - sft_config_infer: {'runner_config': {'batch_size': 1, 'epochs': 1, 'sink_mode': 1, 'sink_size': 1, 'gradient_accumulation_steps': 1, 'num_classes': 1, 'stop_step': 0}, 'context': {'mode': 0, 'device_target': 'Ascend', 'max_call_depth': 10000, 'max_device_memory': '55GB', 'save_graphs': False, 'save_graphs_path': './graph', 'device_id': 0, 'jit_config': {'jit_level': 'O0'}, 'memory_optimize_level': 'O0', 'ascend_config': {'precision_mode': 'must_keep_origin_dtype'}}, 'parallel': {'parallel_mode': 1, 'full_batch': True, 'search_mode': 'sharding_propagation', 'enable_parallel_optimizer': False, 'gradients_mean': False, 'enable_alltoall': False, 'strategy_ckpt_save_file': './ckpt_strategy.ckpt'}, 'trainer': {}, 'model': {'model_config': {'type': 'LlamaConfig', 'batch_size': 1, 'seq_length': 8192, 'hidden_size': 3584, 'num_layers': 2, 'num_heads': 28, 'n_kv_heads': 4, 'vocab_size': 152064, 'intermediate_size': 18944, 'max_position_embeddings': 32768, 'qkv_has_bias': True, 'rms_norm_eps': 1e-06, 'theta': 1000000.0, 'emb_dropout_prob': 0.0, 'eos_token_id': [151645, 151643], 'pad_token_id': 151643, 'bos_token_id': 151643, 'compute_dtype': 'bfloat16', 'layernorm_compute_type': 'float32', 'softmax_compute_type': 'float16', 'rotary_dtype': 'bfloat16', 'param_init_type': 'float32', 'use_past': True, 'use_flash_attention': True, 'block_size': 32, 'num_blocks': 1024, 'use_past_shard': False, 'offset': 0, 'checkpoint_name_or_path': '', 'repetition_penalty': 1.0, 'max_decode_length': 512, 'min_decode_length': 2, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'do_sample': True, 'is_dynamic': True, 'qkv_concat': False, 'auto_map': {'AutoTokenizer': ['qwen2_tokenizer.Qwen2Tokenizer', None]}, 'parallel_config': }, 'arch': {'type': 'LlamaForCausalLM'}}, 'moe_config': , 'parallel_config': , 'processor': {'return_tensors': 'ms', 'tokenizer': {'model_max_length': 32768, 'vocab_file': '/path/vocab.json', 'merges_file': '/path/merges.txt', 'unk_token': '<|endoftext|>', 'pad_token': '<|endoftext|>', 'eos_token': '<|im_end|>', 'chat_template': "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}", 'type': 'Qwen2Tokenizer', 'auto_register': 'qwen2_tokenizer.Qwen2Tokenizer'}, 'type': 'Qwen2Processor'}, 'seed': 1, 'output_dir': './output', 'run_mode': 'predict', 'use_parallel': True, 'resume_training': False, 'load_checkpoint': '', 'load_ckpt_format': 'ckpt', 'auto_trans_ckpt': False, 'transform_process_num': 1, 'src_strategy_path_or_dir': '', 'only_save_strategy': False, 'load_ckpt_async': False, 'use_legacy': True, 'do_eval': False, 'eval_step_interval': 100, 'eval_epoch_interval': -1, 'ignore_data_skip': False, 'data_skip_steps': None, 'profile': False, 'profile_communication': False, 'profile_memory': True, 'init_start_profile': False, 'profile_start_step': 1, 'profile_stop_step': 10, 'profile_rank_ids': None, 'profile_pipeline': False, 'profile_level': 1, 'mstx': False, 'layer_scale': False, 'layer_decay': 0.65, 'lr_scale': False, 'lr_scale_factor': 256, 'micro_batch_interleave_num': 1, 'remote_save_url': None, 'save_file': None, 'input_data': None, 'predict_batch_size': None, 'adapter_id': None, 'exclude_cann_cpu': False, 'train_precision_sync': None, 'infer_precision_sync': None, 'postprocess_use_numpy': False, 'local_rank': 1, 'device_num': 8, 'swap_config': , 'recompute_config': } 2025-05-21 04:31:15,352 - mindformers./output/log[mindformers/core/parallel_config.py:41] - INFO - initial moe_config from dict: {'expert_num': 1, 'capacity_factor': 1.1, 'aux_loss_factor': 0.05, 'num_experts_chosen': 1, 'expert_group_size': None, 'group_wise_a2a': False, 'comp_comm_parallel': False, 'comp_comm_parallel_degree': 2, 'save_token_distribution': False, 'cur_layer': 0, 'enable_cold_hot_expert': False, 'update_step': 10000, 'hot_expert_num': 0, 'cold_token_percent': 1.0, 'moe_module_name': '', 'routing_policy': 'TopkRouterV1', 'norm_topk_prob': True, 'enable_sdrop': False, 'use_fused_ops_topkrouter': False, 'router_dense_type': 'float32', 'shared_expert_num': 0, 'use_shared_expert_gating': False, 'max_router_load': 131072, 'topk_method': 'greedy', 'topk_group': None, 'n_group': None, 'first_k_dense_replace': True, 'moe_intermediate_size': 1407, 'routed_scaling_factor': 1.0, 'aux_loss_types': None, 'aux_loss_factors': None, 'z_loss_factor': 0.0, 'balance_via_topk_bias': False, 'topk_bias_update_rate': 0.0, 'use_allgather_dispatcher': False, 'moe_shared_expert_overlap': False, 'expert_model_parallel': None, 'use_gating_sigmoid': False, 'enable_deredundency': False, 'npu_nums_per_device': 1, 'use_gmm': False, 'enable_gmm_safe_tokens': False, 'use_fused_ops_permute': False, 'callback_moe_droprate': False} 2025-05-21 04:31:15,354 - mindformers./output/log[mindformers/core/parallel_config.py:61] - INFO - initial parallel_config from dict: {'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:15,354 - mindformers./output/log[mindformers/core/parallel_config.py:63] - INFO - pipeline_stage = 2 > 1, vocab_emd_dp will be reset to False. 2025-05-21 04:31:15,356 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:101] - INFO - sft_config_infer: {'runner_config': {'batch_size': 1, 'epochs': 1, 'sink_mode': 1, 'sink_size': 1, 'gradient_accumulation_steps': 1, 'num_classes': 1, 'stop_step': 0}, 'context': {'mode': 0, 'device_target': 'Ascend', 'max_call_depth': 10000, 'max_device_memory': '55GB', 'save_graphs': False, 'save_graphs_path': './graph', 'device_id': 0, 'jit_config': {'jit_level': 'O0'}, 'memory_optimize_level': 'O0', 'ascend_config': {'precision_mode': 'must_keep_origin_dtype'}}, 'parallel': {'parallel_mode': 1, 'full_batch': True, 'search_mode': 'sharding_propagation', 'enable_parallel_optimizer': False, 'gradients_mean': False, 'enable_alltoall': False, 'strategy_ckpt_save_file': './ckpt_strategy.ckpt'}, 'trainer': {}, 'model': {'model_config': {'type': 'LlamaConfig', 'batch_size': 1, 'seq_length': 8192, 'hidden_size': 3584, 'num_layers': 2, 'num_heads': 28, 'n_kv_heads': 4, 'vocab_size': 152064, 'intermediate_size': 18944, 'max_position_embeddings': 32768, 'qkv_has_bias': True, 'rms_norm_eps': 1e-06, 'theta': 1000000.0, 'emb_dropout_prob': 0.0, 'eos_token_id': [151645, 151643], 'pad_token_id': 151643, 'bos_token_id': 151643, 'compute_dtype': 'bfloat16', 'layernorm_compute_type': 'float32', 'softmax_compute_type': 'float16', 'rotary_dtype': 'bfloat16', 'param_init_type': 'float32', 'use_past': True, 'use_flash_attention': True, 'block_size': 32, 'num_blocks': 1024, 'use_past_shard': False, 'offset': 0, 'checkpoint_name_or_path': '', 'repetition_penalty': 1.0, 'max_decode_length': 512, 'min_decode_length': 2, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'do_sample': True, 'is_dynamic': True, 'qkv_concat': False, 'auto_map': {'AutoTokenizer': ['qwen2_tokenizer.Qwen2Tokenizer', None]}, 'parallel_config': }, 'arch': {'type': 'LlamaForCausalLM'}}, 'moe_config': , 'parallel_config': , 'processor': {'return_tensors': 'ms', 'tokenizer': {'model_max_length': 32768, 'vocab_file': '/path/vocab.json', 'merges_file': '/path/merges.txt', 'unk_token': '<|endoftext|>', 'pad_token': '<|endoftext|>', 'eos_token': '<|im_end|>', 'chat_template': "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}", 'type': 'Qwen2Tokenizer', 'auto_register': 'qwen2_tokenizer.Qwen2Tokenizer'}, 'type': 'Qwen2Processor'}, 'seed': 1, 'output_dir': './output', 'run_mode': 'predict', 'use_parallel': True, 'resume_training': False, 'load_checkpoint': '', 'load_ckpt_format': 'ckpt', 'auto_trans_ckpt': False, 'transform_process_num': 1, 'src_strategy_path_or_dir': '', 'only_save_strategy': False, 'load_ckpt_async': False, 'use_legacy': True, 'do_eval': False, 'eval_step_interval': 100, 'eval_epoch_interval': -1, 'ignore_data_skip': False, 'data_skip_steps': None, 'profile': False, 'profile_communication': False, 'profile_memory': True, 'init_start_profile': False, 'profile_start_step': 1, 'profile_stop_step': 10, 'profile_rank_ids': None, 'profile_pipeline': False, 'profile_level': 1, 'mstx': False, 'layer_scale': False, 'layer_decay': 0.65, 'lr_scale': False, 'lr_scale_factor': 256, 'micro_batch_interleave_num': 1, 'remote_save_url': None, 'save_file': None, 'input_data': None, 'predict_batch_size': None, 'adapter_id': None, 'exclude_cann_cpu': False, 'train_precision_sync': None, 'infer_precision_sync': None, 'postprocess_use_numpy': False, 'local_rank': 2, 'device_num': 8, 'swap_config': , 'recompute_config': } 2025-05-21 04:31:15,373 - mindformers./output/log[mindformers/core/parallel_config.py:41] - INFO - initial moe_config from dict: {'expert_num': 1, 'capacity_factor': 1.1, 'aux_loss_factor': 0.05, 'num_experts_chosen': 1, 'expert_group_size': None, 'group_wise_a2a': False, 'comp_comm_parallel': False, 'comp_comm_parallel_degree': 2, 'save_token_distribution': False, 'cur_layer': 0, 'enable_cold_hot_expert': False, 'update_step': 10000, 'hot_expert_num': 0, 'cold_token_percent': 1.0, 'moe_module_name': '', 'routing_policy': 'TopkRouterV1', 'norm_topk_prob': True, 'enable_sdrop': False, 'use_fused_ops_topkrouter': False, 'router_dense_type': 'float32', 'shared_expert_num': 0, 'use_shared_expert_gating': False, 'max_router_load': 131072, 'topk_method': 'greedy', 'topk_group': None, 'n_group': None, 'first_k_dense_replace': True, 'moe_intermediate_size': 1407, 'routed_scaling_factor': 1.0, 'aux_loss_types': None, 'aux_loss_factors': None, 'z_loss_factor': 0.0, 'balance_via_topk_bias': False, 'topk_bias_update_rate': 0.0, 'use_allgather_dispatcher': False, 'moe_shared_expert_overlap': False, 'expert_model_parallel': None, 'use_gating_sigmoid': False, 'enable_deredundency': False, 'npu_nums_per_device': 1, 'use_gmm': False, 'enable_gmm_safe_tokens': False, 'use_fused_ops_permute': False, 'callback_moe_droprate': False} 2025-05-21 04:31:15,374 - mindformers./output/log[mindformers/core/parallel_config.py:61] - INFO - initial parallel_config from dict: {'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:15,375 - mindformers./output/log[mindformers/core/parallel_config.py:63] - INFO - pipeline_stage = 2 > 1, vocab_emd_dp will be reset to False. 2025-05-21 04:31:15,376 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:101] - INFO - sft_config_infer: {'runner_config': {'batch_size': 1, 'epochs': 1, 'sink_mode': 1, 'sink_size': 1, 'gradient_accumulation_steps': 1, 'num_classes': 1, 'stop_step': 0}, 'context': {'mode': 0, 'device_target': 'Ascend', 'max_call_depth': 10000, 'max_device_memory': '55GB', 'save_graphs': False, 'save_graphs_path': './graph', 'device_id': 0, 'jit_config': {'jit_level': 'O0'}, 'memory_optimize_level': 'O0', 'ascend_config': {'precision_mode': 'must_keep_origin_dtype'}}, 'parallel': {'parallel_mode': 1, 'full_batch': True, 'search_mode': 'sharding_propagation', 'enable_parallel_optimizer': False, 'gradients_mean': False, 'enable_alltoall': False, 'strategy_ckpt_save_file': './ckpt_strategy.ckpt'}, 'trainer': {}, 'model': {'model_config': {'type': 'LlamaConfig', 'batch_size': 1, 'seq_length': 8192, 'hidden_size': 3584, 'num_layers': 2, 'num_heads': 28, 'n_kv_heads': 4, 'vocab_size': 152064, 'intermediate_size': 18944, 'max_position_embeddings': 32768, 'qkv_has_bias': True, 'rms_norm_eps': 1e-06, 'theta': 1000000.0, 'emb_dropout_prob': 0.0, 'eos_token_id': [151645, 151643], 'pad_token_id': 151643, 'bos_token_id': 151643, 'compute_dtype': 'bfloat16', 'layernorm_compute_type': 'float32', 'softmax_compute_type': 'float16', 'rotary_dtype': 'bfloat16', 'param_init_type': 'float32', 'use_past': True, 'use_flash_attention': True, 'block_size': 32, 'num_blocks': 1024, 'use_past_shard': False, 'offset': 0, 'checkpoint_name_or_path': '', 'repetition_penalty': 1.0, 'max_decode_length': 512, 'min_decode_length': 2, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'do_sample': True, 'is_dynamic': True, 'qkv_concat': False, 'auto_map': {'AutoTokenizer': ['qwen2_tokenizer.Qwen2Tokenizer', None]}, 'parallel_config': }, 'arch': {'type': 'LlamaForCausalLM'}}, 'moe_config': , 'parallel_config': , 'processor': {'return_tensors': 'ms', 'tokenizer': {'model_max_length': 32768, 'vocab_file': '/path/vocab.json', 'merges_file': '/path/merges.txt', 'unk_token': '<|endoftext|>', 'pad_token': '<|endoftext|>', 'eos_token': '<|im_end|>', 'chat_template': "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}", 'type': 'Qwen2Tokenizer', 'auto_register': 'qwen2_tokenizer.Qwen2Tokenizer'}, 'type': 'Qwen2Processor'}, 'seed': 1, 'output_dir': './output', 'run_mode': 'predict', 'use_parallel': True, 'resume_training': False, 'load_checkpoint': '', 'load_ckpt_format': 'ckpt', 'auto_trans_ckpt': False, 'transform_process_num': 1, 'src_strategy_path_or_dir': '', 'only_save_strategy': False, 'load_ckpt_async': False, 'use_legacy': True, 'do_eval': False, 'eval_step_interval': 100, 'eval_epoch_interval': -1, 'ignore_data_skip': False, 'data_skip_steps': None, 'profile': False, 'profile_communication': False, 'profile_memory': True, 'init_start_profile': False, 'profile_start_step': 1, 'profile_stop_step': 10, 'profile_rank_ids': None, 'profile_pipeline': False, 'profile_level': 1, 'mstx': False, 'layer_scale': False, 'layer_decay': 0.65, 'lr_scale': False, 'lr_scale_factor': 256, 'micro_batch_interleave_num': 1, 'remote_save_url': None, 'save_file': None, 'input_data': None, 'predict_batch_size': None, 'adapter_id': None, 'exclude_cann_cpu': False, 'train_precision_sync': None, 'infer_precision_sync': None, 'postprocess_use_numpy': False, 'local_rank': 4, 'device_num': 8, 'swap_config': , 'recompute_config': } 2025-05-21 04:31:15,795 - mindformers./output/log[mindformers/core/parallel_config.py:41] - INFO - initial moe_config from dict: {'expert_num': 1, 'capacity_factor': 1.1, 'aux_loss_factor': 0.05, 'num_experts_chosen': 1, 'expert_group_size': None, 'group_wise_a2a': False, 'comp_comm_parallel': False, 'comp_comm_parallel_degree': 2, 'save_token_distribution': False, 'cur_layer': 0, 'enable_cold_hot_expert': False, 'update_step': 10000, 'hot_expert_num': 0, 'cold_token_percent': 1.0, 'moe_module_name': '', 'routing_policy': 'TopkRouterV1', 'norm_topk_prob': True, 'enable_sdrop': False, 'use_fused_ops_topkrouter': False, 'router_dense_type': 'float32', 'shared_expert_num': 0, 'use_shared_expert_gating': False, 'max_router_load': 131072, 'topk_method': 'greedy', 'topk_group': None, 'n_group': None, 'first_k_dense_replace': True, 'moe_intermediate_size': 1407, 'routed_scaling_factor': 1.0, 'aux_loss_types': None, 'aux_loss_factors': None, 'z_loss_factor': 0.0, 'balance_via_topk_bias': False, 'topk_bias_update_rate': 0.0, 'use_allgather_dispatcher': False, 'moe_shared_expert_overlap': False, 'expert_model_parallel': None, 'use_gating_sigmoid': False, 'enable_deredundency': False, 'npu_nums_per_device': 1, 'use_gmm': False, 'enable_gmm_safe_tokens': False, 'use_fused_ops_permute': False, 'callback_moe_droprate': False} 2025-05-21 04:31:15,797 - mindformers./output/log[mindformers/core/parallel_config.py:61] - INFO - initial parallel_config from dict: {'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:15,797 - mindformers./output/log[mindformers/core/parallel_config.py:63] - INFO - pipeline_stage = 2 > 1, vocab_emd_dp will be reset to False. 2025-05-21 04:31:15,799 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:101] - INFO - sft_config_infer: {'runner_config': {'batch_size': 1, 'epochs': 1, 'sink_mode': 1, 'sink_size': 1, 'gradient_accumulation_steps': 1, 'num_classes': 1, 'stop_step': 0}, 'context': {'mode': 0, 'device_target': 'Ascend', 'max_call_depth': 10000, 'max_device_memory': '55GB', 'save_graphs': False, 'save_graphs_path': './graph', 'device_id': 0, 'jit_config': {'jit_level': 'O0'}, 'memory_optimize_level': 'O0', 'ascend_config': {'precision_mode': 'must_keep_origin_dtype'}}, 'parallel': {'parallel_mode': 1, 'full_batch': True, 'search_mode': 'sharding_propagation', 'enable_parallel_optimizer': False, 'gradients_mean': False, 'enable_alltoall': False, 'strategy_ckpt_save_file': './ckpt_strategy.ckpt'}, 'trainer': {}, 'model': {'model_config': {'type': 'LlamaConfig', 'batch_size': 1, 'seq_length': 8192, 'hidden_size': 3584, 'num_layers': 2, 'num_heads': 28, 'n_kv_heads': 4, 'vocab_size': 152064, 'intermediate_size': 18944, 'max_position_embeddings': 32768, 'qkv_has_bias': True, 'rms_norm_eps': 1e-06, 'theta': 1000000.0, 'emb_dropout_prob': 0.0, 'eos_token_id': [151645, 151643], 'pad_token_id': 151643, 'bos_token_id': 151643, 'compute_dtype': 'bfloat16', 'layernorm_compute_type': 'float32', 'softmax_compute_type': 'float16', 'rotary_dtype': 'bfloat16', 'param_init_type': 'float32', 'use_past': True, 'use_flash_attention': True, 'block_size': 32, 'num_blocks': 1024, 'use_past_shard': False, 'offset': 0, 'checkpoint_name_or_path': '', 'repetition_penalty': 1.0, 'max_decode_length': 512, 'min_decode_length': 2, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'do_sample': True, 'is_dynamic': True, 'qkv_concat': False, 'auto_map': {'AutoTokenizer': ['qwen2_tokenizer.Qwen2Tokenizer', None]}, 'parallel_config': }, 'arch': {'type': 'LlamaForCausalLM'}}, 'moe_config': , 'parallel_config': , 'processor': {'return_tensors': 'ms', 'tokenizer': {'model_max_length': 32768, 'vocab_file': '/path/vocab.json', 'merges_file': '/path/merges.txt', 'unk_token': '<|endoftext|>', 'pad_token': '<|endoftext|>', 'eos_token': '<|im_end|>', 'chat_template': "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}", 'type': 'Qwen2Tokenizer', 'auto_register': 'qwen2_tokenizer.Qwen2Tokenizer'}, 'type': 'Qwen2Processor'}, 'seed': 1, 'output_dir': './output', 'run_mode': 'predict', 'use_parallel': True, 'resume_training': False, 'load_checkpoint': '', 'load_ckpt_format': 'ckpt', 'auto_trans_ckpt': False, 'transform_process_num': 1, 'src_strategy_path_or_dir': '', 'only_save_strategy': False, 'load_ckpt_async': False, 'use_legacy': True, 'do_eval': False, 'eval_step_interval': 100, 'eval_epoch_interval': -1, 'ignore_data_skip': False, 'data_skip_steps': None, 'profile': False, 'profile_communication': False, 'profile_memory': True, 'init_start_profile': False, 'profile_start_step': 1, 'profile_stop_step': 10, 'profile_rank_ids': None, 'profile_pipeline': False, 'profile_level': 1, 'mstx': False, 'layer_scale': False, 'layer_decay': 0.65, 'lr_scale': False, 'lr_scale_factor': 256, 'micro_batch_interleave_num': 1, 'remote_save_url': None, 'save_file': None, 'input_data': None, 'predict_batch_size': None, 'adapter_id': None, 'exclude_cann_cpu': False, 'train_precision_sync': None, 'infer_precision_sync': None, 'postprocess_use_numpy': False, 'local_rank': 5, 'device_num': 8, 'swap_config': , 'recompute_config': } 2025-05-21 04:31:15,925 - mindformers./output/log[mindformers/core/parallel_config.py:41] - INFO - initial moe_config from dict: {'expert_num': 1, 'capacity_factor': 1.1, 'aux_loss_factor': 0.05, 'num_experts_chosen': 1, 'expert_group_size': None, 'group_wise_a2a': False, 'comp_comm_parallel': False, 'comp_comm_parallel_degree': 2, 'save_token_distribution': False, 'cur_layer': 0, 'enable_cold_hot_expert': False, 'update_step': 10000, 'hot_expert_num': 0, 'cold_token_percent': 1.0, 'moe_module_name': '', 'routing_policy': 'TopkRouterV1', 'norm_topk_prob': True, 'enable_sdrop': False, 'use_fused_ops_topkrouter': False, 'router_dense_type': 'float32', 'shared_expert_num': 0, 'use_shared_expert_gating': False, 'max_router_load': 131072, 'topk_method': 'greedy', 'topk_group': None, 'n_group': None, 'first_k_dense_replace': True, 'moe_intermediate_size': 1407, 'routed_scaling_factor': 1.0, 'aux_loss_types': None, 'aux_loss_factors': None, 'z_loss_factor': 0.0, 'balance_via_topk_bias': False, 'topk_bias_update_rate': 0.0, 'use_allgather_dispatcher': False, 'moe_shared_expert_overlap': False, 'expert_model_parallel': None, 'use_gating_sigmoid': False, 'enable_deredundency': False, 'npu_nums_per_device': 1, 'use_gmm': False, 'enable_gmm_safe_tokens': False, 'use_fused_ops_permute': False, 'callback_moe_droprate': False} 2025-05-21 04:31:15,927 - mindformers./output/log[mindformers/core/parallel_config.py:61] - INFO - initial parallel_config from dict: {'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:15,927 - mindformers./output/log[mindformers/core/parallel_config.py:63] - INFO - pipeline_stage = 2 > 1, vocab_emd_dp will be reset to False. 2025-05-21 04:31:15,929 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:101] - INFO - sft_config_infer: {'runner_config': {'batch_size': 1, 'epochs': 1, 'sink_mode': 1, 'sink_size': 1, 'gradient_accumulation_steps': 1, 'num_classes': 1, 'stop_step': 0}, 'context': {'mode': 0, 'device_target': 'Ascend', 'max_call_depth': 10000, 'max_device_memory': '55GB', 'save_graphs': False, 'save_graphs_path': './graph', 'device_id': 0, 'jit_config': {'jit_level': 'O0'}, 'memory_optimize_level': 'O0', 'ascend_config': {'precision_mode': 'must_keep_origin_dtype'}}, 'parallel': {'parallel_mode': 1, 'full_batch': True, 'search_mode': 'sharding_propagation', 'enable_parallel_optimizer': False, 'gradients_mean': False, 'enable_alltoall': False, 'strategy_ckpt_save_file': './ckpt_strategy.ckpt'}, 'trainer': {}, 'model': {'model_config': {'type': 'LlamaConfig', 'batch_size': 1, 'seq_length': 8192, 'hidden_size': 3584, 'num_layers': 2, 'num_heads': 28, 'n_kv_heads': 4, 'vocab_size': 152064, 'intermediate_size': 18944, 'max_position_embeddings': 32768, 'qkv_has_bias': True, 'rms_norm_eps': 1e-06, 'theta': 1000000.0, 'emb_dropout_prob': 0.0, 'eos_token_id': [151645, 151643], 'pad_token_id': 151643, 'bos_token_id': 151643, 'compute_dtype': 'bfloat16', 'layernorm_compute_type': 'float32', 'softmax_compute_type': 'float16', 'rotary_dtype': 'bfloat16', 'param_init_type': 'float32', 'use_past': True, 'use_flash_attention': True, 'block_size': 32, 'num_blocks': 1024, 'use_past_shard': False, 'offset': 0, 'checkpoint_name_or_path': '', 'repetition_penalty': 1.0, 'max_decode_length': 512, 'min_decode_length': 2, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'do_sample': True, 'is_dynamic': True, 'qkv_concat': False, 'auto_map': {'AutoTokenizer': ['qwen2_tokenizer.Qwen2Tokenizer', None]}, 'parallel_config': }, 'arch': {'type': 'LlamaForCausalLM'}}, 'moe_config': , 'parallel_config': , 'processor': {'return_tensors': 'ms', 'tokenizer': {'model_max_length': 32768, 'vocab_file': '/path/vocab.json', 'merges_file': '/path/merges.txt', 'unk_token': '<|endoftext|>', 'pad_token': '<|endoftext|>', 'eos_token': '<|im_end|>', 'chat_template': "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}", 'type': 'Qwen2Tokenizer', 'auto_register': 'qwen2_tokenizer.Qwen2Tokenizer'}, 'type': 'Qwen2Processor'}, 'seed': 1, 'output_dir': './output', 'run_mode': 'predict', 'use_parallel': True, 'resume_training': False, 'load_checkpoint': '', 'load_ckpt_format': 'ckpt', 'auto_trans_ckpt': False, 'transform_process_num': 1, 'src_strategy_path_or_dir': '', 'only_save_strategy': False, 'load_ckpt_async': False, 'use_legacy': True, 'do_eval': False, 'eval_step_interval': 100, 'eval_epoch_interval': -1, 'ignore_data_skip': False, 'data_skip_steps': None, 'profile': False, 'profile_communication': False, 'profile_memory': True, 'init_start_profile': False, 'profile_start_step': 1, 'profile_stop_step': 10, 'profile_rank_ids': None, 'profile_pipeline': False, 'profile_level': 1, 'mstx': False, 'layer_scale': False, 'layer_decay': 0.65, 'lr_scale': False, 'lr_scale_factor': 256, 'micro_batch_interleave_num': 1, 'remote_save_url': None, 'save_file': None, 'input_data': None, 'predict_batch_size': None, 'adapter_id': None, 'exclude_cann_cpu': False, 'train_precision_sync': None, 'infer_precision_sync': None, 'postprocess_use_numpy': False, 'local_rank': 7, 'device_num': 8, 'swap_config': , 'recompute_config': } 2025-05-21 04:31:15,936 - mindformers./output/log[mindformers/core/parallel_config.py:41] - INFO - initial moe_config from dict: {'expert_num': 1, 'capacity_factor': 1.1, 'aux_loss_factor': 0.05, 'num_experts_chosen': 1, 'expert_group_size': None, 'group_wise_a2a': False, 'comp_comm_parallel': False, 'comp_comm_parallel_degree': 2, 'save_token_distribution': False, 'cur_layer': 0, 'enable_cold_hot_expert': False, 'update_step': 10000, 'hot_expert_num': 0, 'cold_token_percent': 1.0, 'moe_module_name': '', 'routing_policy': 'TopkRouterV1', 'norm_topk_prob': True, 'enable_sdrop': False, 'use_fused_ops_topkrouter': False, 'router_dense_type': 'float32', 'shared_expert_num': 0, 'use_shared_expert_gating': False, 'max_router_load': 131072, 'topk_method': 'greedy', 'topk_group': None, 'n_group': None, 'first_k_dense_replace': True, 'moe_intermediate_size': 1407, 'routed_scaling_factor': 1.0, 'aux_loss_types': None, 'aux_loss_factors': None, 'z_loss_factor': 0.0, 'balance_via_topk_bias': False, 'topk_bias_update_rate': 0.0, 'use_allgather_dispatcher': False, 'moe_shared_expert_overlap': False, 'expert_model_parallel': None, 'use_gating_sigmoid': False, 'enable_deredundency': False, 'npu_nums_per_device': 1, 'use_gmm': False, 'enable_gmm_safe_tokens': False, 'use_fused_ops_permute': False, 'callback_moe_droprate': False} 2025-05-21 04:31:15,938 - mindformers./output/log[mindformers/core/parallel_config.py:61] - INFO - initial parallel_config from dict: {'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:15,938 - mindformers./output/log[mindformers/core/parallel_config.py:63] - INFO - pipeline_stage = 2 > 1, vocab_emd_dp will be reset to False. 2025-05-21 04:31:15,940 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:101] - INFO - sft_config_infer: {'runner_config': {'batch_size': 1, 'epochs': 1, 'sink_mode': 1, 'sink_size': 1, 'gradient_accumulation_steps': 1, 'num_classes': 1, 'stop_step': 0}, 'context': {'mode': 0, 'device_target': 'Ascend', 'max_call_depth': 10000, 'max_device_memory': '55GB', 'save_graphs': False, 'save_graphs_path': './graph', 'device_id': 0, 'jit_config': {'jit_level': 'O0'}, 'memory_optimize_level': 'O0', 'ascend_config': {'precision_mode': 'must_keep_origin_dtype'}}, 'parallel': {'parallel_mode': 1, 'full_batch': True, 'search_mode': 'sharding_propagation', 'enable_parallel_optimizer': False, 'gradients_mean': False, 'enable_alltoall': False, 'strategy_ckpt_save_file': './ckpt_strategy.ckpt'}, 'trainer': {}, 'model': {'model_config': {'type': 'LlamaConfig', 'batch_size': 1, 'seq_length': 8192, 'hidden_size': 3584, 'num_layers': 2, 'num_heads': 28, 'n_kv_heads': 4, 'vocab_size': 152064, 'intermediate_size': 18944, 'max_position_embeddings': 32768, 'qkv_has_bias': True, 'rms_norm_eps': 1e-06, 'theta': 1000000.0, 'emb_dropout_prob': 0.0, 'eos_token_id': [151645, 151643], 'pad_token_id': 151643, 'bos_token_id': 151643, 'compute_dtype': 'bfloat16', 'layernorm_compute_type': 'float32', 'softmax_compute_type': 'float16', 'rotary_dtype': 'bfloat16', 'param_init_type': 'float32', 'use_past': True, 'use_flash_attention': True, 'block_size': 32, 'num_blocks': 1024, 'use_past_shard': False, 'offset': 0, 'checkpoint_name_or_path': '', 'repetition_penalty': 1.0, 'max_decode_length': 512, 'min_decode_length': 2, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'do_sample': True, 'is_dynamic': True, 'qkv_concat': False, 'auto_map': {'AutoTokenizer': ['qwen2_tokenizer.Qwen2Tokenizer', None]}, 'parallel_config': }, 'arch': {'type': 'LlamaForCausalLM'}}, 'moe_config': , 'parallel_config': , 'processor': {'return_tensors': 'ms', 'tokenizer': {'model_max_length': 32768, 'vocab_file': '/path/vocab.json', 'merges_file': '/path/merges.txt', 'unk_token': '<|endoftext|>', 'pad_token': '<|endoftext|>', 'eos_token': '<|im_end|>', 'chat_template': "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}", 'type': 'Qwen2Tokenizer', 'auto_register': 'qwen2_tokenizer.Qwen2Tokenizer'}, 'type': 'Qwen2Processor'}, 'seed': 1, 'output_dir': './output', 'run_mode': 'predict', 'use_parallel': True, 'resume_training': False, 'load_checkpoint': '', 'load_ckpt_format': 'ckpt', 'auto_trans_ckpt': False, 'transform_process_num': 1, 'src_strategy_path_or_dir': '', 'only_save_strategy': False, 'load_ckpt_async': False, 'use_legacy': True, 'do_eval': False, 'eval_step_interval': 100, 'eval_epoch_interval': -1, 'ignore_data_skip': False, 'data_skip_steps': None, 'profile': False, 'profile_communication': False, 'profile_memory': True, 'init_start_profile': False, 'profile_start_step': 1, 'profile_stop_step': 10, 'profile_rank_ids': None, 'profile_pipeline': False, 'profile_level': 1, 'mstx': False, 'layer_scale': False, 'layer_decay': 0.65, 'lr_scale': False, 'lr_scale_factor': 256, 'micro_batch_interleave_num': 1, 'remote_save_url': None, 'save_file': None, 'input_data': None, 'predict_batch_size': None, 'adapter_id': None, 'exclude_cann_cpu': False, 'train_precision_sync': None, 'infer_precision_sync': None, 'postprocess_use_numpy': False, 'local_rank': 0, 'device_num': 8, 'swap_config': , 'recompute_config': } tp_group is:True dp_group is:True data_parallel_group:dp-1-5 tensor_model_parallel_group:tp-0-1-2-3 2025-05-21 04:31:17,504 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. tp_group is:True dp_group is:True data_parallel_group:dp-3-7 tensor_model_parallel_group:tp-0-1-2-3 2025-05-21 04:31:17,506 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:17,510 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:17,511 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:17,512 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/models/grpo_models.py:208] - INFO - num_iterations: 2, epsilon_low: 0.2, epsilon_high: 0.2 2025-05-21 04:31:17,513 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/models/grpo_models.py:208] - INFO - num_iterations: 2, epsilon_low: 0.2, epsilon_high: 0.2 tp_group is:True dp_group is:True data_parallel_group:dp-2-6 tensor_model_parallel_group:tp-0-1-2-3 2025-05-21 04:31:17,513 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. tp_group is:True dp_group is:True data_parallel_group:dp-0-4 tensor_model_parallel_group:tp-0-1-2-3 2025-05-21 04:31:17,516 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:17,516 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:252] - INFO - GRPOTrainer: _init_grpo_infer_dataset, dataset dir /home/jenkins/mindspore/testcases/testcases/tests/st/dataset/mini_gsm8k.mindrecord 2025-05-21 04:31:17,517 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:252] - INFO - GRPOTrainer: _init_grpo_infer_dataset, dataset dir /home/jenkins/mindspore/testcases/testcases/tests/st/dataset/mini_gsm8k.mindrecord 2025-05-21 04:31:17,520 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:17,522 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:17,522 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/models/grpo_models.py:208] - INFO - num_iterations: 2, epsilon_low: 0.2, epsilon_high: 0.2 2025-05-21 04:31:17,523 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/models/grpo_models.py:208] - INFO - num_iterations: 2, epsilon_low: 0.2, epsilon_high: 0.2 2025-05-21 04:31:17,524 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:49] - INFO - init RefWorker 2025-05-21 04:31:17,524 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:49] - INFO - init RefWorker 2025-05-21 04:31:17,527 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:252] - INFO - GRPOTrainer: _init_grpo_infer_dataset, dataset dir /home/jenkins/mindspore/testcases/testcases/tests/st/dataset/mini_gsm8k.mindrecord 2025-05-21 04:31:17,528 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:252] - INFO - GRPOTrainer: _init_grpo_infer_dataset, dataset dir /home/jenkins/mindspore/testcases/testcases/tests/st/dataset/mini_gsm8k.mindrecord 2025-05-21 04:31:17,538 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:49] - INFO - init RefWorker 2025-05-21 04:31:17,557 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:17,558 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:17,558 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:17,558 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:17,558 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:17,558 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:17,558 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:17,558 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:17,559 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:17,559 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:17,559 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:17,559 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:17,559 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:57] - INFO - ref parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:17,559 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:58] - INFO - grpo_config.ref_config.recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:17,560 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:57] - INFO - ref parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:17,560 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:58] - INFO - grpo_config.ref_config.recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:17,564 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:87] - INFO - ref_model_config:LlamaConfig { "attn_proj_has_bias": false, "auto_map": { "AutoTokenizer": [ "qwen2_tokenizer.Qwen2Tokenizer", null ] }, "batch_size": 1, "block_size": 32, "bos_token_id": 151643, "calculate_per_token_loss": false, "checkpoint_name_or_path": "", "chunk_prefill": false, "compute_dtype": "bfloat16", "do_sample": true, "emb_dropout_prob": 0.0, "embedding_init_type": "float32", "eos_token_id": [ 151645, 151643 ], "extend_method": "None", "ffn_dim_multiplier": null, "fine_grain_interleave": 1, "fused_rms_norm": true, "hidden_size": 3584, "ignore_token_id": -100, "init_method_std": 0.01, "input_sliced_sig": false, "intermediate_size": 18944, "is_dynamic": true, "layernorm_compute_type": "float32", "llm_backend": "", "max_decode_length": 2048, "max_position_embedding": 8192, "max_position_embeddings": 32768, "mindformers_version": "1.6.0", "model_name": "llama", "model_type": "llama", "multiple_of": 256, "n_kv_heads": 4, "num_blocks": 1024, "num_heads": 28, "num_layers": 2, "offset": 0, "pad_token_id": 151643, "parallel_config": { "micro_batch_num": 4, "model_parallel": 4, "pipeline_stage": 2, "use_seq_parallel": true, "vocab_emb_dp": false }, "parallel_decoding_params": null, "parallel_optimizer": false, "param_init_type": "float32", "pp_interleave_num": 1, "qkv_concat": false, "qkv_has_bias": true, "quant_config": null, "repetition_penalty": 1.0, "residual_dtype": "bfloat16", "return_hidden_states": false, "rms_norm_eps": 1e-06, "rmsnorm_compute_2d": false, "rotary_dtype": "bfloat16", "scaling_factor": 1.0, "seq_length": 8192, "softmax_compute_type": "float16", "stage_num": 0, "start_stage": 0, "temperature": 1.2, "theta": 1000000.0, "tie_word_embeddings": false, "top_k": 50, "top_p": 1.0, "type": "LlamaConfig", "use_attn_mask_compression": false, "use_eod_attn_mask_compression": false, "use_flash_attention": true, "use_past": false, "use_past_shard": false, "use_ring_attention": false, "use_rope_slice": false, "vocab_size": 152064 } 2025-05-21 04:31:17,564 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:87] - INFO - ref_model_config:LlamaConfig { "attn_proj_has_bias": false, "auto_map": { "AutoTokenizer": [ "qwen2_tokenizer.Qwen2Tokenizer", null ] }, "batch_size": 1, "block_size": 32, "bos_token_id": 151643, "calculate_per_token_loss": false, "checkpoint_name_or_path": "", "chunk_prefill": false, "compute_dtype": "bfloat16", "do_sample": true, "emb_dropout_prob": 0.0, "embedding_init_type": "float32", "eos_token_id": [ 151645, 151643 ], "extend_method": "None", "ffn_dim_multiplier": null, "fine_grain_interleave": 1, "fused_rms_norm": true, "hidden_size": 3584, "ignore_token_id": -100, "init_method_std": 0.01, "input_sliced_sig": false, "intermediate_size": 18944, "is_dynamic": true, "layernorm_compute_type": "float32", "llm_backend": "", "max_decode_length": 2048, "max_position_embedding": 8192, "max_position_embeddings": 32768, "mindformers_version": "1.6.0", "model_name": "llama", "model_type": "llama", "multiple_of": 256, "n_kv_heads": 4, "num_blocks": 1024, "num_heads": 28, "num_layers": 2, "offset": 0, "pad_token_id": 151643, "parallel_config": { "micro_batch_num": 4, "model_parallel": 4, "pipeline_stage": 2, "use_seq_parallel": true, "vocab_emb_dp": false }, "parallel_decoding_params": null, "parallel_optimizer": false, "param_init_type": "float32", "pp_interleave_num": 1, "qkv_concat": false, "qkv_has_bias": true, "quant_config": null, "repetition_penalty": 1.0, "residual_dtype": "bfloat16", "return_hidden_states": false, "rms_norm_eps": 1e-06, "rmsnorm_compute_2d": false, "rotary_dtype": "bfloat16", "scaling_factor": 1.0, "seq_length": 8192, "softmax_compute_type": "float16", "stage_num": 0, "start_stage": 0, "temperature": 1.2, "theta": 1000000.0, "tie_word_embeddings": false, "top_k": 50, "top_p": 1.0, "type": "LlamaConfig", "use_attn_mask_compression": false, "use_eod_attn_mask_compression": false, "use_flash_attention": true, "use_past": false, "use_past_shard": false, "use_ring_attention": false, "use_rope_slice": false, "vocab_size": 152064 } 2025-05-21 04:31:17,565 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:91] - INFO - start create pipeline ref_pipeline1-5 2025-05-21 04:31:17,565 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:91] - INFO - start create pipeline ref_pipeline3-7 2025-05-21 04:31:17,571 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:17,571 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:17,571 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:17,571 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:17,572 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:17,572 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:17,572 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:57] - INFO - ref parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:17,573 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:58] - INFO - grpo_config.ref_config.recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:17,578 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:87] - INFO - ref_model_config:LlamaConfig { "attn_proj_has_bias": false, "auto_map": { "AutoTokenizer": [ "qwen2_tokenizer.Qwen2Tokenizer", null ] }, "batch_size": 1, "block_size": 32, "bos_token_id": 151643, "calculate_per_token_loss": false, "checkpoint_name_or_path": "", "chunk_prefill": false, "compute_dtype": "bfloat16", "do_sample": true, "emb_dropout_prob": 0.0, "embedding_init_type": "float32", "eos_token_id": [ 151645, 151643 ], "extend_method": "None", "ffn_dim_multiplier": null, "fine_grain_interleave": 1, "fused_rms_norm": true, "hidden_size": 3584, "ignore_token_id": -100, "init_method_std": 0.01, "input_sliced_sig": false, "intermediate_size": 18944, "is_dynamic": true, "layernorm_compute_type": "float32", "llm_backend": "", "max_decode_length": 2048, "max_position_embedding": 8192, "max_position_embeddings": 32768, "mindformers_version": "1.6.0", "model_name": "llama", "model_type": "llama", "multiple_of": 256, "n_kv_heads": 4, "num_blocks": 1024, "num_heads": 28, "num_layers": 2, "offset": 0, "pad_token_id": 151643, "parallel_config": { "micro_batch_num": 4, "model_parallel": 4, "pipeline_stage": 2, "use_seq_parallel": true, "vocab_emb_dp": false }, "parallel_decoding_params": null, "parallel_optimizer": false, "param_init_type": "float32", "pp_interleave_num": 1, "qkv_concat": false, "qkv_has_bias": true, "quant_config": null, "repetition_penalty": 1.0, "residual_dtype": "bfloat16", "return_hidden_states": false, "rms_norm_eps": 1e-06, "rmsnorm_compute_2d": false, "rotary_dtype": "bfloat16", "scaling_factor": 1.0, "seq_length": 8192, "softmax_compute_type": "float16", "stage_num": 0, "start_stage": 0, "temperature": 1.2, "theta": 1000000.0, "tie_word_embeddings": false, "top_k": 50, "top_p": 1.0, "type": "LlamaConfig", "use_attn_mask_compression": false, "use_eod_attn_mask_compression": false, "use_flash_attention": true, "use_past": false, "use_past_shard": false, "use_ring_attention": false, "use_rope_slice": false, "vocab_size": 152064 } 2025-05-21 04:31:17,578 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:91] - INFO - start create pipeline ref_pipeline2-6 2025-05-21 04:31:17,580 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:49] - INFO - init RefWorker 2025-05-21 04:31:17,613 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:17,614 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:17,614 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:17,614 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:17,614 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:17,615 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:17,615 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:57] - INFO - ref parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:17,615 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:58] - INFO - grpo_config.ref_config.recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:17,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:87] - INFO - ref_model_config:LlamaConfig { "attn_proj_has_bias": false, "auto_map": { "AutoTokenizer": [ "qwen2_tokenizer.Qwen2Tokenizer", null ] }, "batch_size": 1, "block_size": 32, "bos_token_id": 151643, "calculate_per_token_loss": false, "checkpoint_name_or_path": "", "chunk_prefill": false, "compute_dtype": "bfloat16", "do_sample": true, "emb_dropout_prob": 0.0, "embedding_init_type": "float32", "eos_token_id": [ 151645, 151643 ], "extend_method": "None", "ffn_dim_multiplier": null, "fine_grain_interleave": 1, "fused_rms_norm": true, "hidden_size": 3584, "ignore_token_id": -100, "init_method_std": 0.01, "input_sliced_sig": false, "intermediate_size": 18944, "is_dynamic": true, "layernorm_compute_type": "float32", "llm_backend": "", "max_decode_length": 2048, "max_position_embedding": 8192, "max_position_embeddings": 32768, "mindformers_version": "1.6.0", "model_name": "llama", "model_type": "llama", "multiple_of": 256, "n_kv_heads": 4, "num_blocks": 1024, "num_heads": 28, "num_layers": 2, "offset": 0, "pad_token_id": 151643, "parallel_config": { "micro_batch_num": 4, "model_parallel": 4, "pipeline_stage": 2, "use_seq_parallel": true, "vocab_emb_dp": false }, "parallel_decoding_params": null, "parallel_optimizer": false, "param_init_type": "float32", "pp_interleave_num": 1, "qkv_concat": false, "qkv_has_bias": true, "quant_config": null, "repetition_penalty": 1.0, "residual_dtype": "bfloat16", "return_hidden_states": false, "rms_norm_eps": 1e-06, "rmsnorm_compute_2d": false, "rotary_dtype": "bfloat16", "scaling_factor": 1.0, "seq_length": 8192, "softmax_compute_type": "float16", "stage_num": 0, "start_stage": 0, "temperature": 1.2, "theta": 1000000.0, "tie_word_embeddings": false, "top_k": 50, "top_p": 1.0, "type": "LlamaConfig", "use_attn_mask_compression": false, "use_eod_attn_mask_compression": false, "use_flash_attention": true, "use_past": false, "use_past_shard": false, "use_ring_attention": false, "use_rope_slice": false, "vocab_size": 152064 } 2025-05-21 04:31:17,620 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:91] - INFO - start create pipeline ref_pipeline0-4 tp_group is:True dp_group is:True data_parallel_group:dp-3-7 tensor_model_parallel_group:tp-4-5-6-7 2025-05-21 04:31:17,706 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:17,711 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:17,713 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/models/grpo_models.py:208] - INFO - num_iterations: 2, epsilon_low: 0.2, epsilon_high: 0.2 2025-05-21 04:31:17,717 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:252] - INFO - GRPOTrainer: _init_grpo_infer_dataset, dataset dir /home/jenkins/mindspore/testcases/testcases/tests/st/dataset/mini_gsm8k.mindrecord tp_group is:True dp_group is:True data_parallel_group:dp-0-4 tensor_model_parallel_group:tp-4-5-6-7 2025-05-21 04:31:17,719 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:17,724 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:49] - INFO - init RefWorker 2025-05-21 04:31:17,726 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:17,728 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/models/grpo_models.py:208] - INFO - num_iterations: 2, epsilon_low: 0.2, epsilon_high: 0.2 tp_group is:True dp_group is:True data_parallel_group:dp-2-6 tensor_model_parallel_group:tp-4-5-6-7 2025-05-21 04:31:17,731 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:17,733 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:252] - INFO - GRPOTrainer: _init_grpo_infer_dataset, dataset dir /home/jenkins/mindspore/testcases/testcases/tests/st/dataset/mini_gsm8k.mindrecord 2025-05-21 04:31:17,738 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:17,740 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/models/grpo_models.py:208] - INFO - num_iterations: 2, epsilon_low: 0.2, epsilon_high: 0.2 2025-05-21 04:31:17,744 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:49] - INFO - init RefWorker 2025-05-21 04:31:17,745 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:252] - INFO - GRPOTrainer: _init_grpo_infer_dataset, dataset dir /home/jenkins/mindspore/testcases/testcases/tests/st/dataset/mini_gsm8k.mindrecord 2025-05-21 04:31:17,755 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:49] - INFO - init RefWorker 2025-05-21 04:31:17,757 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:17,757 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:17,758 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:17,758 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:17,758 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:17,758 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:17,759 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:57] - INFO - ref parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:17,759 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:58] - INFO - grpo_config.ref_config.recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:17,763 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:87] - INFO - ref_model_config:LlamaConfig { "attn_proj_has_bias": false, "auto_map": { "AutoTokenizer": [ "qwen2_tokenizer.Qwen2Tokenizer", null ] }, "batch_size": 1, "block_size": 32, "bos_token_id": 151643, "calculate_per_token_loss": false, "checkpoint_name_or_path": "", "chunk_prefill": false, "compute_dtype": "bfloat16", "do_sample": true, "emb_dropout_prob": 0.0, "embedding_init_type": "float32", "eos_token_id": [ 151645, 151643 ], "extend_method": "None", "ffn_dim_multiplier": null, "fine_grain_interleave": 1, "fused_rms_norm": true, "hidden_size": 3584, "ignore_token_id": -100, "init_method_std": 0.01, "input_sliced_sig": false, "intermediate_size": 18944, "is_dynamic": true, "layernorm_compute_type": "float32", "llm_backend": "", "max_decode_length": 2048, "max_position_embedding": 8192, "max_position_embeddings": 32768, "mindformers_version": "1.6.0", "model_name": "llama", "model_type": "llama", "multiple_of": 256, "n_kv_heads": 4, "num_blocks": 1024, "num_heads": 28, "num_layers": 2, "offset": 0, "pad_token_id": 151643, "parallel_config": { "micro_batch_num": 4, "model_parallel": 4, "pipeline_stage": 2, "use_seq_parallel": true, "vocab_emb_dp": false }, "parallel_decoding_params": null, "parallel_optimizer": false, "param_init_type": "float32", "pp_interleave_num": 1, "qkv_concat": false, "qkv_has_bias": true, "quant_config": null, "repetition_penalty": 1.0, "residual_dtype": "bfloat16", "return_hidden_states": false, "rms_norm_eps": 1e-06, "rmsnorm_compute_2d": false, "rotary_dtype": "bfloat16", "scaling_factor": 1.0, "seq_length": 8192, "softmax_compute_type": "float16", "stage_num": 0, "start_stage": 0, "temperature": 1.2, "theta": 1000000.0, "tie_word_embeddings": false, "top_k": 50, "top_p": 1.0, "type": "LlamaConfig", "use_attn_mask_compression": false, "use_eod_attn_mask_compression": false, "use_flash_attention": true, "use_past": false, "use_past_shard": false, "use_ring_attention": false, "use_rope_slice": false, "vocab_size": 152064 } 2025-05-21 04:31:17,764 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:91] - INFO - start create pipeline ref_pipeline3-7 2025-05-21 04:31:17,779 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:17,780 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:17,780 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:17,780 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:17,780 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:17,781 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:17,781 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:57] - INFO - ref parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:17,781 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:58] - INFO - grpo_config.ref_config.recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:17,787 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:17,788 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:17,788 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:87] - INFO - ref_model_config:LlamaConfig { "attn_proj_has_bias": false, "auto_map": { "AutoTokenizer": [ "qwen2_tokenizer.Qwen2Tokenizer", null ] }, "batch_size": 1, "block_size": 32, "bos_token_id": 151643, "calculate_per_token_loss": false, "checkpoint_name_or_path": "", "chunk_prefill": false, "compute_dtype": "bfloat16", "do_sample": true, "emb_dropout_prob": 0.0, "embedding_init_type": "float32", "eos_token_id": [ 151645, 151643 ], "extend_method": "None", "ffn_dim_multiplier": null, "fine_grain_interleave": 1, "fused_rms_norm": true, "hidden_size": 3584, "ignore_token_id": -100, "init_method_std": 0.01, "input_sliced_sig": false, "intermediate_size": 18944, "is_dynamic": true, "layernorm_compute_type": "float32", "llm_backend": "", "max_decode_length": 2048, "max_position_embedding": 8192, "max_position_embeddings": 32768, "mindformers_version": "1.6.0", "model_name": "llama", "model_type": "llama", "multiple_of": 256, "n_kv_heads": 4, "num_blocks": 1024, "num_heads": 28, "num_layers": 2, "offset": 0, "pad_token_id": 151643, "parallel_config": { "micro_batch_num": 4, "model_parallel": 4, "pipeline_stage": 2, "use_seq_parallel": true, "vocab_emb_dp": false }, "parallel_decoding_params": null, "parallel_optimizer": false, "param_init_type": "float32", "pp_interleave_num": 1, "qkv_concat": false, "qkv_has_bias": true, "quant_config": null, "repetition_penalty": 1.0, "residual_dtype": "bfloat16", "return_hidden_states": false, "rms_norm_eps": 1e-06, "rmsnorm_compute_2d": false, "rotary_dtype": "bfloat16", "scaling_factor": 1.0, "seq_length": 8192, "softmax_compute_type": "float16", "stage_num": 0, "start_stage": 0, "temperature": 1.2, "theta": 1000000.0, "tie_word_embeddings": false, "top_k": 50, "top_p": 1.0, "type": "LlamaConfig", "use_attn_mask_compression": false, "use_eod_attn_mask_compression": false, "use_flash_attention": true, "use_past": false, "use_past_shard": false, "use_ring_attention": false, "use_rope_slice": false, "vocab_size": 152064 } 2025-05-21 04:31:17,788 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:17,788 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:17,788 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:91] - INFO - start create pipeline ref_pipeline0-4 2025-05-21 04:31:17,788 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:17,789 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:17,789 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:57] - INFO - ref parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:17,789 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:58] - INFO - grpo_config.ref_config.recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:17,794 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:87] - INFO - ref_model_config:LlamaConfig { "attn_proj_has_bias": false, "auto_map": { "AutoTokenizer": [ "qwen2_tokenizer.Qwen2Tokenizer", null ] }, "batch_size": 1, "block_size": 32, "bos_token_id": 151643, "calculate_per_token_loss": false, "checkpoint_name_or_path": "", "chunk_prefill": false, "compute_dtype": "bfloat16", "do_sample": true, "emb_dropout_prob": 0.0, "embedding_init_type": "float32", "eos_token_id": [ 151645, 151643 ], "extend_method": "None", "ffn_dim_multiplier": null, "fine_grain_interleave": 1, "fused_rms_norm": true, "hidden_size": 3584, "ignore_token_id": -100, "init_method_std": 0.01, "input_sliced_sig": false, "intermediate_size": 18944, "is_dynamic": true, "layernorm_compute_type": "float32", "llm_backend": "", "max_decode_length": 2048, "max_position_embedding": 8192, "max_position_embeddings": 32768, "mindformers_version": "1.6.0", "model_name": "llama", "model_type": "llama", "multiple_of": 256, "n_kv_heads": 4, "num_blocks": 1024, "num_heads": 28, "num_layers": 2, "offset": 0, "pad_token_id": 151643, "parallel_config": { "micro_batch_num": 4, "model_parallel": 4, "pipeline_stage": 2, "use_seq_parallel": true, "vocab_emb_dp": false }, "parallel_decoding_params": null, "parallel_optimizer": false, "param_init_type": "float32", "pp_interleave_num": 1, "qkv_concat": false, "qkv_has_bias": true, "quant_config": null, "repetition_penalty": 1.0, "residual_dtype": "bfloat16", "return_hidden_states": false, "rms_norm_eps": 1e-06, "rmsnorm_compute_2d": false, "rotary_dtype": "bfloat16", "scaling_factor": 1.0, "seq_length": 8192, "softmax_compute_type": "float16", "stage_num": 0, "start_stage": 0, "temperature": 1.2, "theta": 1000000.0, "tie_word_embeddings": false, "top_k": 50, "top_p": 1.0, "type": "LlamaConfig", "use_attn_mask_compression": false, "use_eod_attn_mask_compression": false, "use_flash_attention": true, "use_past": false, "use_past_shard": false, "use_ring_attention": false, "use_rope_slice": false, "vocab_size": 152064 } 2025-05-21 04:31:17,794 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:91] - INFO - start create pipeline ref_pipeline2-6 2025-05-21 04:31:17,826 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:93] - INFO - end create pipeline ref_pipeline3-7 2025-05-21 04:31:17,828 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:17,832 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: True 2025-05-21 04:31:17,833 - mindformers./output/log[mindformers/models/llama/llama.py:510] - INFO - use_flash_attention is set to True when run_mode is predict and is_dynamic is True. 2025-05-21 04:31:17,835 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:17,842 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:93] - INFO - end create pipeline ref_pipeline0-4 2025-05-21 04:31:17,844 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:17,847 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:93] - INFO - end create pipeline ref_pipeline0-4 2025-05-21 04:31:17,848 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:17,849 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: True 2025-05-21 04:31:17,849 - mindformers./output/log[mindformers/models/llama/llama.py:510] - INFO - use_flash_attention is set to True when run_mode is predict and is_dynamic is True. 2025-05-21 04:31:17,852 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:17,852 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: True 2025-05-21 04:31:17,853 - mindformers./output/log[mindformers/models/llama/llama.py:510] - INFO - use_flash_attention is set to True when run_mode is predict and is_dynamic is True. 2025-05-21 04:31:17,855 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:17,858 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:93] - INFO - end create pipeline ref_pipeline2-6 2025-05-21 04:31:17,860 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:17,865 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: True 2025-05-21 04:31:17,865 - mindformers./output/log[mindformers/models/llama/llama.py:510] - INFO - use_flash_attention is set to True when run_mode is predict and is_dynamic is True. 2025-05-21 04:31:17,868 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN tp_group is:True dp_group is:True data_parallel_group:dp-1-5 tensor_model_parallel_group:tp-4-5-6-7 2025-05-21 04:31:17,872 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:17,878 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:17,880 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/models/grpo_models.py:208] - INFO - num_iterations: 2, epsilon_low: 0.2, epsilon_high: 0.2 2025-05-21 04:31:17,885 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:252] - INFO - GRPOTrainer: _init_grpo_infer_dataset, dataset dir /home/jenkins/mindspore/testcases/testcases/tests/st/dataset/mini_gsm8k.mindrecord 2025-05-21 04:31:17,894 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:49] - INFO - init RefWorker 2025-05-21 04:31:17,928 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:17,928 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:17,928 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:17,929 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:17,929 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:17,929 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:17,930 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:57] - INFO - ref parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:17,930 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:58] - INFO - grpo_config.ref_config.recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:17,934 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:87] - INFO - ref_model_config:LlamaConfig { "attn_proj_has_bias": false, "auto_map": { "AutoTokenizer": [ "qwen2_tokenizer.Qwen2Tokenizer", null ] }, "batch_size": 1, "block_size": 32, "bos_token_id": 151643, "calculate_per_token_loss": false, "checkpoint_name_or_path": "", "chunk_prefill": false, "compute_dtype": "bfloat16", "do_sample": true, "emb_dropout_prob": 0.0, "embedding_init_type": "float32", "eos_token_id": [ 151645, 151643 ], "extend_method": "None", "ffn_dim_multiplier": null, "fine_grain_interleave": 1, "fused_rms_norm": true, "hidden_size": 3584, "ignore_token_id": -100, "init_method_std": 0.01, "input_sliced_sig": false, "intermediate_size": 18944, "is_dynamic": true, "layernorm_compute_type": "float32", "llm_backend": "", "max_decode_length": 2048, "max_position_embedding": 8192, "max_position_embeddings": 32768, "mindformers_version": "1.6.0", "model_name": "llama", "model_type": "llama", "multiple_of": 256, "n_kv_heads": 4, "num_blocks": 1024, "num_heads": 28, "num_layers": 2, "offset": 0, "pad_token_id": 151643, "parallel_config": { "micro_batch_num": 4, "model_parallel": 4, "pipeline_stage": 2, "use_seq_parallel": true, "vocab_emb_dp": false }, "parallel_decoding_params": null, "parallel_optimizer": false, "param_init_type": "float32", "pp_interleave_num": 1, "qkv_concat": false, "qkv_has_bias": true, "quant_config": null, "repetition_penalty": 1.0, "residual_dtype": "bfloat16", "return_hidden_states": false, "rms_norm_eps": 1e-06, "rmsnorm_compute_2d": false, "rotary_dtype": "bfloat16", "scaling_factor": 1.0, "seq_length": 8192, "softmax_compute_type": "float16", "stage_num": 0, "start_stage": 0, "temperature": 1.2, "theta": 1000000.0, "tie_word_embeddings": false, "top_k": 50, "top_p": 1.0, "type": "LlamaConfig", "use_attn_mask_compression": false, "use_eod_attn_mask_compression": false, "use_flash_attention": true, "use_past": false, "use_past_shard": false, "use_ring_attention": false, "use_rope_slice": false, "vocab_size": 152064 } 2025-05-21 04:31:17,935 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:91] - INFO - start create pipeline ref_pipeline1-5 2025-05-21 04:31:17,999 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:93] - INFO - end create pipeline ref_pipeline1-5 2025-05-21 04:31:18,001 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:18,003 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:93] - INFO - end create pipeline ref_pipeline3-7 2025-05-21 04:31:18,005 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:18,005 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: True 2025-05-21 04:31:18,006 - mindformers./output/log[mindformers/models/llama/llama.py:510] - INFO - use_flash_attention is set to True when run_mode is predict and is_dynamic is True. 2025-05-21 04:31:18,009 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:18,009 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: True 2025-05-21 04:31:18,009 - mindformers./output/log[mindformers/models/llama/llama.py:510] - INFO - use_flash_attention is set to True when run_mode is predict and is_dynamic is True. 2025-05-21 04:31:18,012 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:18,040 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:93] - INFO - end create pipeline ref_pipeline2-6 2025-05-21 04:31:18,042 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:18,047 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: True 2025-05-21 04:31:18,048 - mindformers./output/log[mindformers/models/llama/llama.py:510] - INFO - use_flash_attention is set to True when run_mode is predict and is_dynamic is True. 2025-05-21 04:31:18,051 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:18,180 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:93] - INFO - end create pipeline ref_pipeline1-5 2025-05-21 04:31:18,182 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:18,186 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: True 2025-05-21 04:31:18,187 - mindformers./output/log[mindformers/models/llama/llama.py:510] - INFO - use_flash_attention is set to True when run_mode is predict and is_dynamic is True. 2025-05-21 04:31:18,190 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:18,334 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:18,335 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:18,335 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:18,335 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:18,336 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:18,336 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:18,336 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:18,336 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:18,336 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:18,336 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:18,368 - mindformers./output/log[mindformers/version_control.py:76] - INFO - Predict enable lazy inline. 2025-05-21 04:31:18,372 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:18,372 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:18,372 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:18,373 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:18,373 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:18,373 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:18,373 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:18,373 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:18,374 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:18,374 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:18,374 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:18,374 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:18,374 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:18,374 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:18,374 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:18,374 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:18,374 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:18,375 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:18,375 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:18,375 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:18,398 - mindformers./output/log[mindformers/version_control.py:76] - INFO - Predict enable lazy inline. 2025-05-21 04:31:18,403 - mindformers./output/log[mindformers/version_control.py:76] - INFO - Predict enable lazy inline. 2025-05-21 04:31:18,404 - mindformers./output/log[mindformers/version_control.py:76] - INFO - Predict enable lazy inline. 2025-05-21 04:31:18,407 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:18,408 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:18,408 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:18,408 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:18,409 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:18,409 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:18,409 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:18,409 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:18,410 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:18,410 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:18,410 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:18,420 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:66] - INFO - init TrainWorker 2025-05-21 04:31:18,429 - mindformers./output/log[mindformers/version_control.py:76] - INFO - Predict enable lazy inline. 2025-05-21 04:31:18,431 - mindformers./output/log[mindformers/version_control.py:76] - INFO - Predict enable lazy inline. 2025-05-21 04:31:18,437 - mindformers./output/log[mindformers/version_control.py:76] - INFO - Predict enable lazy inline. 2025-05-21 04:31:18,440 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:18,441 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:18,441 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:18,442 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:18,442 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:18,442 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:18,442 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:18,442 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:18,443 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:18,443 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:18,443 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:18,443 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:18,443 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:18,444 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:18,444 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:18,444 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:18,444 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:18,444 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:18,445 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:18,445 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:75] - INFO - actor parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:18,445 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:76] - INFO - grpo_config.actor_config.recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:18,448 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:18,448 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:66] - INFO - init TrainWorker 2025-05-21 04:31:18,450 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:66] - INFO - init TrainWorker 2025-05-21 04:31:18,458 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: False 2025-05-21 04:31:18,462 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:18,464 - mindformers./output/log[mindformers/version_control.py:76] - INFO - Predict enable lazy inline. 2025-05-21 04:31:18,471 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:18,471 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:18,471 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:18,471 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:18,472 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:18,472 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:18,472 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:18,472 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:18,472 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:18,472 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:18,472 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:18,472 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:18,473 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:18,473 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:18,473 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:18,473 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:18,473 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:18,473 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:18,473 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:18,473 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:18,474 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:18,474 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:18,474 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:18,474 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:18,474 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:18,474 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:18,474 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:18,474 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:18,475 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:18,475 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:75] - INFO - actor parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:18,475 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:18,475 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:18,475 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:76] - INFO - grpo_config.actor_config.recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:18,475 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:18,475 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:18,475 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:18,475 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:18,476 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:75] - INFO - actor parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:18,476 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:76] - INFO - grpo_config.actor_config.recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:18,477 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:18,479 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:18,481 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: False 2025-05-21 04:31:18,483 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: False 2025-05-21 04:31:18,483 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:66] - INFO - init TrainWorker 2025-05-21 04:31:18,484 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:18,485 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:18,505 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:18,506 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:18,506 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:18,506 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:18,507 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:18,507 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:18,507 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:18,507 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:18,508 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:18,508 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:18,508 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:18,508 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:18,508 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:18,509 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:18,509 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:18,509 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:18,509 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:18,509 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:18,510 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:18,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:75] - INFO - actor parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:18,510 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:18,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:76] - INFO - grpo_config.actor_config.recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:18,510 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:18,511 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:18,511 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:18,511 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:18,511 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:18,511 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:18,512 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:18,512 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:18,516 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: False 2025-05-21 04:31:18,520 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:18,521 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:18,522 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:18,522 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:18,522 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:18,523 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:18,523 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:18,523 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:18,523 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:18,523 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:18,524 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:18,534 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:18,535 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:18,535 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:18,535 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:18,535 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:18,536 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:18,536 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:18,536 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:18,536 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:18,536 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:18,546 - mindformers./output/log[mindformers/version_control.py:76] - INFO - Predict enable lazy inline. 2025-05-21 04:31:18,554 - mindformers./output/log[mindformers/version_control.py:76] - INFO - Predict enable lazy inline. 2025-05-21 04:31:18,563 - mindformers./output/log[mindformers/version_control.py:76] - INFO - Predict enable lazy inline. 2025-05-21 04:31:18,575 - mindformers./output/log[mindformers/version_control.py:76] - INFO - Predict enable lazy inline. 2025-05-21 04:31:18,582 - mindformers./output/log[mindformers/version_control.py:76] - INFO - Predict enable lazy inline. 2025-05-21 04:31:18,586 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:18,589 - mindformers./output/log[mindformers/version_control.py:76] - INFO - Predict enable lazy inline. 2025-05-21 04:31:18,593 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:18,595 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:66] - INFO - init TrainWorker 2025-05-21 04:31:18,599 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:18,601 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:66] - INFO - init TrainWorker 2025-05-21 04:31:18,607 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:66] - INFO - init TrainWorker 2025-05-21 04:31:18,616 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:18,616 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:18,617 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:18,617 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:18,617 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:18,617 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:18,618 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:18,618 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:18,618 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:18,618 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:18,618 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:18,619 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:18,619 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:18,619 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:18,619 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:18,619 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:18,620 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:18,620 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:75] - INFO - actor parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:18,620 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:76] - INFO - grpo_config.actor_config.recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:18,622 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:18,622 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:18,622 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:18,622 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:18,623 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:18,623 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:18,623 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:18,623 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:18,623 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:18,624 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:18,624 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:18,624 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:18,624 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:18,624 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:18,625 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:18,625 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:18,625 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:18,625 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:18,626 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:75] - INFO - actor parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:18,626 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:76] - INFO - grpo_config.actor_config.recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:18,627 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: False 2025-05-21 04:31:18,628 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:18,628 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:18,628 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:18,629 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:18,629 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:18,629 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:18,629 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:18,629 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:18,630 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:18,630 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:18,630 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:18,630 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:18,630 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:18,630 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:18,631 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:18,631 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:18,631 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:18,631 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:18,631 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:18,632 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:75] - INFO - actor parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:18,632 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:76] - INFO - grpo_config.actor_config.recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:18,633 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: False 2025-05-21 04:31:18,634 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:18,636 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:18,638 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: False 2025-05-21 04:31:18,641 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:18,724 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:18,725 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:18,725 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:18,725 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:18,726 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:18,726 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:18,726 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:18,726 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:18,727 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:18,727 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:18,758 - mindformers./output/log[mindformers/version_control.py:76] - INFO - Predict enable lazy inline. 2025-05-21 04:31:18,788 - mindformers./output/log[mindformers/version_control.py:76] - INFO - Predict enable lazy inline. 2025-05-21 04:31:18,801 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:18,809 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:66] - INFO - init TrainWorker 2025-05-21 04:31:18,831 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:18,831 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:18,831 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:18,832 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:18,832 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:18,832 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:18,832 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:18,833 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:18,833 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:18,833 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:18,834 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:18,834 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:18,834 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:18,834 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:18,834 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:18,834 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:18,835 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:18,835 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:75] - INFO - actor parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:18,835 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:76] - INFO - grpo_config.actor_config.recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:18,838 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:18,842 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: False 2025-05-21 04:31:18,845 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:19,365 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:19,366 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:19,366 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:19,366 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:19,366 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:19,367 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:19,367 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:19,367 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:19,367 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:19,367 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:19,431 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:19,439 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:19,440 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/models/grpo_models.py:208] - INFO - num_iterations: 2, epsilon_low: 0.2, epsilon_high: 0.2 2025-05-21 04:31:19,443 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:273] - INFO - pipeline cell 2025-05-21 04:31:19,456 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:19,456 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:19,456 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:19,457 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:19,457 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:19,457 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:19,457 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:19,458 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:19,458 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:19,458 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:19,464 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:19,465 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:19,465 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:19,465 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:19,466 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:19,466 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:19,466 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:19,466 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:19,467 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:19,467 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:19,523 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:320] - INFO - pipeline cell 2025-05-21 04:31:19,523 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:19,527 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:19,527 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:19,527 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:19,527 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:19,528 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:19,528 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:19,528 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:19,528 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:19,528 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:19,529 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:19,530 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:19,532 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:19,534 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/models/grpo_models.py:208] - INFO - num_iterations: 2, epsilon_low: 0.2, epsilon_high: 0.2 2025-05-21 04:31:19,536 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:273] - INFO - pipeline cell 2025-05-21 04:31:19,538 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:19,539 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/models/grpo_models.py:208] - INFO - num_iterations: 2, epsilon_low: 0.2, epsilon_high: 0.2 2025-05-21 04:31:19,542 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:273] - INFO - pipeline cell 2025-05-21 04:31:19,550 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:19,550 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:19,550 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:19,551 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:19,551 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:19,551 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:19,551 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:19,551 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:19,552 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:19,552 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:19,575 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:19,575 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:19,576 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:19,576 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:19,576 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:19,576 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:19,576 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:19,577 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:19,577 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:19,577 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:19,584 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:19,584 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:19,584 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:19,584 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:19,585 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:19,585 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:19,585 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:19,585 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:19,585 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:19,586 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:19,591 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:19,599 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:19,600 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/models/grpo_models.py:208] - INFO - num_iterations: 2, epsilon_low: 0.2, epsilon_high: 0.2 2025-05-21 04:31:19,603 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:273] - INFO - pipeline cell 2025-05-21 04:31:19,607 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:320] - INFO - pipeline cell 2025-05-21 04:31:19,608 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:320] - INFO - pipeline cell 2025-05-21 04:31:19,612 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:19,620 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:19,622 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/models/grpo_models.py:208] - INFO - num_iterations: 2, epsilon_low: 0.2, epsilon_high: 0.2 2025-05-21 04:31:19,624 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:273] - INFO - pipeline cell 2025-05-21 04:31:19,636 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:19,643 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:19,645 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:19,645 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/models/grpo_models.py:208] - INFO - num_iterations: 2, epsilon_low: 0.2, epsilon_high: 0.2 2025-05-21 04:31:19,648 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:273] - INFO - pipeline cell 2025-05-21 04:31:19,652 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:19,654 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/models/grpo_models.py:208] - INFO - num_iterations: 2, epsilon_low: 0.2, epsilon_high: 0.2 2025-05-21 04:31:19,656 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:273] - INFO - pipeline cell 2025-05-21 04:31:19,672 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:320] - INFO - pipeline cell 2025-05-21 04:31:19,690 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:320] - INFO - pipeline cell 2025-05-21 04:31:19,714 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:320] - INFO - pipeline cell 2025-05-21 04:31:19,722 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:320] - INFO - pipeline cell 2025-05-21 04:31:19,807 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:19,808 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:19,808 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:19,808 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:19,809 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:19,809 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:19,809 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:19,809 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:19,809 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:19,810 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:19,869 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:19,876 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:19,878 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/models/grpo_models.py:208] - INFO - num_iterations: 2, epsilon_low: 0.2, epsilon_high: 0.2 2025-05-21 04:31:19,881 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:273] - INFO - pipeline cell 2025-05-21 04:31:19,945 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:320] - INFO - pipeline cell Creating hash value for the group_name hash(0-1-2-3)=7f1758d36cc7b761c9ccce92808de7ac Creating hash value for the group_name hash(3-7)=e30609fbce6a1a756f50a31ec86eae83 Creating hash value for the group_name hash(0-1-2-3)=7f1758d36cc7b761c9ccce92808de7ac Creating hash value for the group_name hash(3-7)=e30609fbce6a1a756f50a31ec86eae83 2025-05-21 04:31:20,250 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:53] - INFO - init OldPolicyWorker Creating hash value for the group_name hash(0-1-2-3)=7f1758d36cc7b761c9ccce92808de7ac Creating hash value for the group_name hash(1-5)=12426c956d1bc5017082b12a969b0b7c Creating hash value for the group_name hash(0-1-2-3)=7f1758d36cc7b761c9ccce92808de7ac Creating hash value for the group_name hash(1-5)=12426c956d1bc5017082b12a969b0b7c 2025-05-21 04:31:20,270 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:53] - INFO - init OldPolicyWorker 2025-05-21 04:31:20,272 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:20,272 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:20,272 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:20,272 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:20,273 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:20,273 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:20,273 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:20,273 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:20,273 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:20,274 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:20,274 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:20,274 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:20,274 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:20,274 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:20,275 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:20,275 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:20,275 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:20,275 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:60] - INFO - old_policy parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:20,276 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:62] - INFO - old_policy_config recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:20,277 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:20,281 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: False 2025-05-21 04:31:20,283 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:20,291 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:20,291 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:20,291 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:20,291 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:20,292 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:20,292 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:20,292 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:20,292 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:20,292 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:20,293 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:20,293 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:20,293 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:20,293 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:20,294 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:20,294 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:20,294 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:20,294 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:20,294 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:60] - INFO - old_policy parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:20,295 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:62] - INFO - old_policy_config recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:20,296 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:20,300 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: False 2025-05-21 04:31:20,303 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN Creating hash value for the group_name hash(4-5-6-7)=ed832c5612cab19d2c5a27f0a350fa5a Creating hash value for the group_name hash(3-7)=e30609fbce6a1a756f50a31ec86eae83 Creating hash value for the group_name hash(4-5-6-7)=ed832c5612cab19d2c5a27f0a350fa5a Creating hash value for the group_name hash(3-7)=e30609fbce6a1a756f50a31ec86eae83 2025-05-21 04:31:20,431 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:53] - INFO - init OldPolicyWorker 2025-05-21 04:31:20,451 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:20,451 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:20,452 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:20,452 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:20,452 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:20,452 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. Creating hash value for the group_name hash(4-5-6-7)=ed832c5612cab19d2c5a27f0a350fa5a Creating hash value for the group_name hash(1-5)=12426c956d1bc5017082b12a969b0b7c Creating hash value for the group_name hash(4-5-6-7)=ed832c5612cab19d2c5a27f0a350fa5a Creating hash value for the group_name hash(1-5)=12426c956d1bc5017082b12a969b0b7c 2025-05-21 04:31:20,452 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:53] - INFO - init OldPolicyWorker 2025-05-21 04:31:20,453 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:20,453 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:20,453 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:20,453 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:20,453 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:20,454 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:20,454 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:20,454 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:20,454 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:20,454 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:20,455 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:20,455 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:60] - INFO - old_policy parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:20,455 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:62] - INFO - old_policy_config recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:20,457 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:20,460 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: False 2025-05-21 04:31:20,463 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:20,473 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:20,473 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:20,474 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:20,474 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:20,474 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:20,474 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:20,475 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:20,475 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:20,475 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:20,475 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:20,476 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:20,476 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:20,476 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:20,476 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:20,476 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:20,477 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:20,477 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:20,477 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:60] - INFO - old_policy parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:20,478 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:62] - INFO - old_policy_config recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:20,480 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:20,484 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: False 2025-05-21 04:31:20,487 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN Creating hash value for the group_name hash(4-5-6-7)=ed832c5612cab19d2c5a27f0a350fa5a Creating hash value for the group_name hash(0-4)=cb4ececddcb4517ca0bcddafd23813b9 Creating hash value for the group_name hash(4-5-6-7)=ed832c5612cab19d2c5a27f0a350fa5a Creating hash value for the group_name hash(0-4)=cb4ececddcb4517ca0bcddafd23813b9 2025-05-21 04:31:20,592 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:53] - INFO - init OldPolicyWorker Creating hash value for the group_name hash(0-1-2-3)=7f1758d36cc7b761c9ccce92808de7ac Creating hash value for the group_name hash(0-4)=cb4ececddcb4517ca0bcddafd23813b9 Creating hash value for the group_name hash(0-1-2-3)=7f1758d36cc7b761c9ccce92808de7ac Creating hash value for the group_name hash(0-4)=cb4ececddcb4517ca0bcddafd23813b9 2025-05-21 04:31:20,593 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:53] - INFO - init OldPolicyWorker 2025-05-21 04:31:20,615 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:20,615 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:20,615 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:20,615 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:20,615 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:20,615 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:20,615 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:20,615 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:20,616 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:20,616 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:20,616 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:20,616 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:20,616 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:20,616 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:20,616 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:20,616 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:20,617 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:20,617 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:20,617 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:20,617 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:20,617 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:20,617 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:20,617 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:20,617 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:20,617 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:20,618 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:20,618 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:20,618 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:20,618 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:20,618 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:20,618 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:20,618 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:20,618 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:20,618 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:20,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:60] - INFO - old_policy parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:20,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:60] - INFO - old_policy parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:20,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:62] - INFO - old_policy_config recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:20,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:62] - INFO - old_policy_config recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:20,621 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:20,621 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:20,624 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: False 2025-05-21 04:31:20,625 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: False 2025-05-21 04:31:20,627 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:20,628 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN Creating hash value for the group_name hash(0-1-2-3)=7f1758d36cc7b761c9ccce92808de7ac Creating hash value for the group_name hash(2-6)=d9639340c2f0051c1a7a09da5ef07ed4 Creating hash value for the group_name hash(0-1-2-3)=7f1758d36cc7b761c9ccce92808de7ac Creating hash value for the group_name hash(2-6)=d9639340c2f0051c1a7a09da5ef07ed4 2025-05-21 04:31:20,700 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:53] - INFO - init OldPolicyWorker 2025-05-21 04:31:20,722 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:20,723 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:20,723 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:20,723 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:20,723 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:20,724 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:20,724 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:20,724 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:20,724 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:20,724 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:20,724 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:20,725 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:20,725 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:20,725 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:20,725 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:20,726 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:20,726 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:20,726 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:60] - INFO - old_policy parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:20,727 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:62] - INFO - old_policy_config recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:20,729 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:20,732 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: False 2025-05-21 04:31:20,735 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN Creating hash value for the group_name hash(4-5-6-7)=ed832c5612cab19d2c5a27f0a350fa5a Creating hash value for the group_name hash(2-6)=d9639340c2f0051c1a7a09da5ef07ed4 Creating hash value for the group_name hash(4-5-6-7)=ed832c5612cab19d2c5a27f0a350fa5a Creating hash value for the group_name hash(2-6)=d9639340c2f0051c1a7a09da5ef07ed4 2025-05-21 04:31:20,877 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:53] - INFO - init OldPolicyWorker 2025-05-21 04:31:20,899 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:20,899 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:20,899 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:20,899 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:20,900 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:20,900 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:20,900 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:20,900 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:20,901 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:20,901 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:20,901 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:20,901 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:20,902 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:20,902 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:20,902 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:20,902 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:20,902 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:20,903 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:60] - INFO - old_policy parallel_config:{'data_parallel': 1, 'model_parallel': 4, 'pipeline_stage': 2, 'expert_parallel': 1, 'use_seq_parallel': True, 'micro_batch_num': 4, 'vocab_emb_dp': False, 'context_parallel': 1} 2025-05-21 04:31:20,903 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:62] - INFO - old_policy_config recompute_config:{'recompute': False, 'select_recompute': False, 'parallel_optimizer_comm_recompute': False, 'mp_comm_recompute': True, 'recompute_slice_activation': False} 2025-05-21 04:31:20,906 - mindformers./output/log[mindformers/version_control.py:140] - INFO - The Lazy Inline compilation acceleration feature is turned on. 2025-05-21 04:31:20,909 - mindformers./output/log[mindformers/models/llama/llama.py:508] - INFO - Predict run mode: False 2025-05-21 04:31:20,912 - mindformers./output/log[mindformers/models/llama/llama.py:108] - INFO - MoE config is None, use normal FFN 2025-05-21 04:31:21,245 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:21,246 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:21,246 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:21,246 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:21,246 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:21,246 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:21,247 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:21,247 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:21,247 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:21,247 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:21,275 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:21,276 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:21,276 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:21,276 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:21,276 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:21,276 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:21,277 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:21,277 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:21,277 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:21,277 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:21,312 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:21,324 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:105] - INFO - config of sft_model_config_train LlamaConfig { "attn_proj_has_bias": false, "batch_size": 1, "block_size": 32, "bos_token_id": 1, "calculate_per_token_loss": false, "checkpoint_name_or_path": null, "chunk_prefill": false, "compute_dtype": "bfloat16", "compute_in_2d": true, "do_sample": false, "emb_dropout_prob": 0.0, "embedding_init_type": "float32", "eos_token_id": 151643, "extend_method": "None", "ffn_dim_multiplier": null, "fine_grain_interleave": 1, "fused_rms_norm": true, "hidden_size": 3584, "ignore_token_id": -100, "init_method_std": 0.01, "input_sliced_sig": false, "intermediate_size": 18944, "is_dynamic": false, "kv_channels": 128, "layernorm_compute_type": "float32", "llm_backend": "", "max_decode_length": 512, "max_position_embedding": 131072, "mindformers_version": "1.6.0", "model_name": "llama", "model_type": "llama", "multiple_of": 256, "n_kv_heads": 4, "num_blocks": 128, "num_heads": 28, "num_layers": 2, "offset": 0, "pad_token_id": 151643, "parallel_config": { "micro_batch_num": 4, "model_parallel": 4, "pipeline_stage": 2, "use_seq_parallel": true, "vocab_emb_dp": false }, "parallel_decoding_params": null, "parallel_optimizer": false, "param_init_type": "float32", "pp_interleave_num": 1, "qkv_concat": false, "qkv_has_bias": true, "quant_config": null, "repetition_penalty": 1, "residual_dtype": "bfloat16", "return_hidden_states": false, "rms_norm_eps": 1e-06, "rmsnorm_compute_2d": false, "rotary_dtype": "float16", "rotary_emb_base": 1000000, "rotary_pct": 1.0, "scaling_factor": 1.0, "seq_length": 8192, "softmax_compute_type": "float16", "stage_num": 0, "start_stage": 0, "theta": 1000000.0, "tie_word_embeddings": false, "top_k": 0, "top_p": 0.8, "type": "LlamaConfig", "use_attn_mask_compression": false, "use_eod_attn_mask_compression": false, "use_flash_attention": true, "use_past": false, "use_ring_attention": false, "use_rope_slice": false, "vocab_size": 152064 } 2025-05-21 04:31:21,324 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:109] - INFO - set packing_sample_length to 8192 2025-05-21 04:31:21,324 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:110] - INFO - GRPOTrainer: finish init workers 2025-05-21 04:31:21,343 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:21,345 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:21,346 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:21,346 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:21,346 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:21,347 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:21,347 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:21,347 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:21,347 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:21,347 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:21,348 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:21,348 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:21,348 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:21,348 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:21,348 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:21,349 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:21,349 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:21,349 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:21,355 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:105] - INFO - config of sft_model_config_train LlamaConfig { "attn_proj_has_bias": false, "batch_size": 1, "block_size": 32, "bos_token_id": 1, "calculate_per_token_loss": false, "checkpoint_name_or_path": null, "chunk_prefill": false, "compute_dtype": "bfloat16", "compute_in_2d": true, "do_sample": false, "emb_dropout_prob": 0.0, "embedding_init_type": "float32", "eos_token_id": 151643, "extend_method": "None", "ffn_dim_multiplier": null, "fine_grain_interleave": 1, "fused_rms_norm": true, "hidden_size": 3584, "ignore_token_id": -100, "init_method_std": 0.01, "input_sliced_sig": false, "intermediate_size": 18944, "is_dynamic": false, "kv_channels": 128, "layernorm_compute_type": "float32", "llm_backend": "", "max_decode_length": 512, "max_position_embedding": 131072, "mindformers_version": "1.6.0", "model_name": "llama", "model_type": "llama", "multiple_of": 256, "n_kv_heads": 4, "num_blocks": 128, "num_heads": 28, "num_layers": 2, "offset": 0, "pad_token_id": 151643, "parallel_config": { "micro_batch_num": 4, "model_parallel": 4, "pipeline_stage": 2, "use_seq_parallel": true, "vocab_emb_dp": false }, "parallel_decoding_params": null, "parallel_optimizer": false, "param_init_type": "float32", "pp_interleave_num": 1, "qkv_concat": false, "qkv_has_bias": true, "quant_config": null, "repetition_penalty": 1, "residual_dtype": "bfloat16", "return_hidden_states": false, "rms_norm_eps": 1e-06, "rmsnorm_compute_2d": false, "rotary_dtype": "float16", "rotary_emb_base": 1000000, "rotary_pct": 1.0, "scaling_factor": 1.0, "seq_length": 8192, "softmax_compute_type": "float16", "stage_num": 0, "start_stage": 0, "theta": 1000000.0, "tie_word_embeddings": false, "top_k": 0, "top_p": 0.8, "type": "LlamaConfig", "use_attn_mask_compression": false, "use_eod_attn_mask_compression": false, "use_flash_attention": true, "use_past": false, "use_ring_attention": false, "use_rope_slice": false, "vocab_size": 152064 } 2025-05-21 04:31:21,355 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:109] - INFO - set packing_sample_length to 8192 2025-05-21 04:31:21,355 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:110] - INFO - GRPOTrainer: finish init workers 2025-05-21 04:31:21,378 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:21,378 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:21,378 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:21,379 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:21,379 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:21,379 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:21,379 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:21,379 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:21,380 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:21,380 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:21,380 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:21,380 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:21,380 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:21,381 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:21,381 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:21,381 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:21,381 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:21,407 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:21,408 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:21,408 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:21,408 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:21,408 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:21,409 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:21,409 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:21,409 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:21,409 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:21,409 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:21,462 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:21,462 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:21,462 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:21,462 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:21,463 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:21,463 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:21,463 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:21,463 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:21,464 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:21,464 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:21,471 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:21,481 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:105] - INFO - config of sft_model_config_train LlamaConfig { "attn_proj_has_bias": false, "batch_size": 1, "block_size": 32, "bos_token_id": 1, "calculate_per_token_loss": false, "checkpoint_name_or_path": null, "chunk_prefill": false, "compute_dtype": "bfloat16", "compute_in_2d": true, "do_sample": false, "emb_dropout_prob": 0.0, "embedding_init_type": "float32", "eos_token_id": 151643, "extend_method": "None", "ffn_dim_multiplier": null, "fine_grain_interleave": 1, "fused_rms_norm": true, "hidden_size": 3584, "ignore_token_id": -100, "init_method_std": 0.01, "input_sliced_sig": false, "intermediate_size": 18944, "is_dynamic": false, "kv_channels": 128, "layernorm_compute_type": "float32", "llm_backend": "", "max_decode_length": 512, "max_position_embedding": 131072, "mindformers_version": "1.6.0", "model_name": "llama", "model_type": "llama", "multiple_of": 256, "n_kv_heads": 4, "num_blocks": 128, "num_heads": 28, "num_layers": 2, "offset": 0, "pad_token_id": 151643, "parallel_config": { "micro_batch_num": 4, "model_parallel": 4, "pipeline_stage": 2, "use_seq_parallel": true, "vocab_emb_dp": false }, "parallel_decoding_params": null, "parallel_optimizer": false, "param_init_type": "float32", "pp_interleave_num": 1, "qkv_concat": false, "qkv_has_bias": true, "quant_config": null, "repetition_penalty": 1, "residual_dtype": "bfloat16", "return_hidden_states": false, "rms_norm_eps": 1e-06, "rmsnorm_compute_2d": false, "rotary_dtype": "float16", "rotary_emb_base": 1000000, "rotary_pct": 1.0, "scaling_factor": 1.0, "seq_length": 8192, "softmax_compute_type": "float16", "stage_num": 0, "start_stage": 0, "theta": 1000000.0, "tie_word_embeddings": false, "top_k": 0, "top_p": 0.8, "type": "LlamaConfig", "use_attn_mask_compression": false, "use_eod_attn_mask_compression": false, "use_flash_attention": true, "use_past": false, "use_ring_attention": false, "use_rope_slice": false, "vocab_size": 152064 } 2025-05-21 04:31:21,482 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:109] - INFO - set packing_sample_length to 8192 2025-05-21 04:31:21,482 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:110] - INFO - GRPOTrainer: finish init workers 2025-05-21 04:31:21,502 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:21,502 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:21,502 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:21,503 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:21,503 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:21,503 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:21,503 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:21,504 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:21,504 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:21,504 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:21,504 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:21,504 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:21,504 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:21,505 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:21,505 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:21,505 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:21,505 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:21,505 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:21,506 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:21,506 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:21,506 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:21,506 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:21,506 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:21,507 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:21,507 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:21,507 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:21,507 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:21,524 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:21,535 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:105] - INFO - config of sft_model_config_train LlamaConfig { "attn_proj_has_bias": false, "batch_size": 1, "block_size": 32, "bos_token_id": 1, "calculate_per_token_loss": false, "checkpoint_name_or_path": null, "chunk_prefill": false, "compute_dtype": "bfloat16", "compute_in_2d": true, "do_sample": false, "emb_dropout_prob": 0.0, "embedding_init_type": "float32", "eos_token_id": 151643, "extend_method": "None", "ffn_dim_multiplier": null, "fine_grain_interleave": 1, "fused_rms_norm": true, "hidden_size": 3584, "ignore_token_id": -100, "init_method_std": 0.01, "input_sliced_sig": false, "intermediate_size": 18944, "is_dynamic": false, "kv_channels": 128, "layernorm_compute_type": "float32", "llm_backend": "", "max_decode_length": 512, "max_position_embedding": 131072, "mindformers_version": "1.6.0", "model_name": "llama", "model_type": "llama", "multiple_of": 256, "n_kv_heads": 4, "num_blocks": 128, "num_heads": 28, "num_layers": 2, "offset": 0, "pad_token_id": 151643, "parallel_config": { "micro_batch_num": 4, "model_parallel": 4, "pipeline_stage": 2, "use_seq_parallel": true, "vocab_emb_dp": false }, "parallel_decoding_params": null, "parallel_optimizer": false, "param_init_type": "float32", "pp_interleave_num": 1, "qkv_concat": false, "qkv_has_bias": true, "quant_config": null, "repetition_penalty": 1, "residual_dtype": "bfloat16", "return_hidden_states": false, "rms_norm_eps": 1e-06, "rmsnorm_compute_2d": false, "rotary_dtype": "float16", "rotary_emb_base": 1000000, "rotary_pct": 1.0, "scaling_factor": 1.0, "seq_length": 8192, "softmax_compute_type": "float16", "stage_num": 0, "start_stage": 0, "theta": 1000000.0, "tie_word_embeddings": false, "top_k": 0, "top_p": 0.8, "type": "LlamaConfig", "use_attn_mask_compression": false, "use_eod_attn_mask_compression": false, "use_flash_attention": true, "use_past": false, "use_ring_attention": false, "use_rope_slice": false, "vocab_size": 152064 } 2025-05-21 04:31:21,535 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:109] - INFO - set packing_sample_length to 8192 2025-05-21 04:31:21,536 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:110] - INFO - GRPOTrainer: finish init workers 2025-05-21 04:31:21,556 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:21,556 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:21,556 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:21,557 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:21,557 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:21,557 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:21,557 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:21,558 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:21,558 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:21,558 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:21,558 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:21,558 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:21,559 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:21,559 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:21,559 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:21,559 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:21,559 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:21,564 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:21,574 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:105] - INFO - config of sft_model_config_train LlamaConfig { "attn_proj_has_bias": false, "batch_size": 1, "block_size": 32, "bos_token_id": 1, "calculate_per_token_loss": false, "checkpoint_name_or_path": null, "chunk_prefill": false, "compute_dtype": "bfloat16", "compute_in_2d": true, "do_sample": false, "emb_dropout_prob": 0.0, "embedding_init_type": "float32", "eos_token_id": 151643, "extend_method": "None", "ffn_dim_multiplier": null, "fine_grain_interleave": 1, "fused_rms_norm": true, "hidden_size": 3584, "ignore_token_id": -100, "init_method_std": 0.01, "input_sliced_sig": false, "intermediate_size": 18944, "is_dynamic": false, "kv_channels": 128, "layernorm_compute_type": "float32", "llm_backend": "", "max_decode_length": 512, "max_position_embedding": 131072, "mindformers_version": "1.6.0", "model_name": "llama", "model_type": "llama", "multiple_of": 256, "n_kv_heads": 4, "num_blocks": 128, "num_heads": 28, "num_layers": 2, "offset": 0, "pad_token_id": 151643, "parallel_config": { "micro_batch_num": 4, "model_parallel": 4, "pipeline_stage": 2, "use_seq_parallel": true, "vocab_emb_dp": false }, "parallel_decoding_params": null, "parallel_optimizer": false, "param_init_type": "float32", "pp_interleave_num": 1, "qkv_concat": false, "qkv_has_bias": true, "quant_config": null, "repetition_penalty": 1, "residual_dtype": "bfloat16", "return_hidden_states": false, "rms_norm_eps": 1e-06, "rmsnorm_compute_2d": false, "rotary_dtype": "float16", "rotary_emb_base": 1000000, "rotary_pct": 1.0, "scaling_factor": 1.0, "seq_length": 8192, "softmax_compute_type": "float16", "stage_num": 0, "start_stage": 0, "theta": 1000000.0, "tie_word_embeddings": false, "top_k": 0, "top_p": 0.8, "type": "LlamaConfig", "use_attn_mask_compression": false, "use_eod_attn_mask_compression": false, "use_flash_attention": true, "use_past": false, "use_ring_attention": false, "use_rope_slice": false, "vocab_size": 152064 } 2025-05-21 04:31:21,574 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:109] - INFO - set packing_sample_length to 8192 2025-05-21 04:31:21,574 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:110] - INFO - GRPOTrainer: finish init workers 2025-05-21 04:31:21,594 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:21,594 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:21,594 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:21,594 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:21,595 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:21,595 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:21,595 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:21,595 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:21,595 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:21,596 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:21,596 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:21,596 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:21,596 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:21,596 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:21,597 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:21,597 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:21,597 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:21,632 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:21,632 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:21,632 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:21,632 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:21,633 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:21,633 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:21,633 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:21,633 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:21,634 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:21,634 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:21,665 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:21,666 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:21,666 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:21,666 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:21,666 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:21,667 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:21,667 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:21,667 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:21,667 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:21,667 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:21,692 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:21,702 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:105] - INFO - config of sft_model_config_train LlamaConfig { "attn_proj_has_bias": false, "batch_size": 1, "block_size": 32, "bos_token_id": 1, "calculate_per_token_loss": false, "checkpoint_name_or_path": null, "chunk_prefill": false, "compute_dtype": "bfloat16", "compute_in_2d": true, "do_sample": false, "emb_dropout_prob": 0.0, "embedding_init_type": "float32", "eos_token_id": 151643, "extend_method": "None", "ffn_dim_multiplier": null, "fine_grain_interleave": 1, "fused_rms_norm": true, "hidden_size": 3584, "ignore_token_id": -100, "init_method_std": 0.01, "input_sliced_sig": false, "intermediate_size": 18944, "is_dynamic": false, "kv_channels": 128, "layernorm_compute_type": "float32", "llm_backend": "", "max_decode_length": 512, "max_position_embedding": 131072, "mindformers_version": "1.6.0", "model_name": "llama", "model_type": "llama", "multiple_of": 256, "n_kv_heads": 4, "num_blocks": 128, "num_heads": 28, "num_layers": 2, "offset": 0, "pad_token_id": 151643, "parallel_config": { "micro_batch_num": 4, "model_parallel": 4, "pipeline_stage": 2, "use_seq_parallel": true, "vocab_emb_dp": false }, "parallel_decoding_params": null, "parallel_optimizer": false, "param_init_type": "float32", "pp_interleave_num": 1, "qkv_concat": false, "qkv_has_bias": true, "quant_config": null, "repetition_penalty": 1, "residual_dtype": "bfloat16", "return_hidden_states": false, "rms_norm_eps": 1e-06, "rmsnorm_compute_2d": false, "rotary_dtype": "float16", "rotary_emb_base": 1000000, "rotary_pct": 1.0, "scaling_factor": 1.0, "seq_length": 8192, "softmax_compute_type": "float16", "stage_num": 0, "start_stage": 0, "theta": 1000000.0, "tie_word_embeddings": false, "top_k": 0, "top_p": 0.8, "type": "LlamaConfig", "use_attn_mask_compression": false, "use_eod_attn_mask_compression": false, "use_flash_attention": true, "use_past": false, "use_ring_attention": false, "use_rope_slice": false, "vocab_size": 152064 } 2025-05-21 04:31:21,702 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:109] - INFO - set packing_sample_length to 8192 2025-05-21 04:31:21,702 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:110] - INFO - GRPOTrainer: finish init workers 2025-05-21 04:31:21,722 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:21,722 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:21,723 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:21,723 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:21,723 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:21,723 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:21,723 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:21,723 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:21,724 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:21,724 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:21,724 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:21,724 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:21,724 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:21,725 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:21,725 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:21,725 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:21,725 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:21,725 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:21,726 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/utils/utils.py:497] - WARNING - The given path contains no 'model.safetensors.index.json' file. 2025-05-21 04:31:21,733 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:105] - INFO - config of sft_model_config_train LlamaConfig { "attn_proj_has_bias": false, "batch_size": 1, "block_size": 32, "bos_token_id": 1, "calculate_per_token_loss": false, "checkpoint_name_or_path": null, "chunk_prefill": false, "compute_dtype": "bfloat16", "compute_in_2d": true, "do_sample": false, "emb_dropout_prob": 0.0, "embedding_init_type": "float32", "eos_token_id": 151643, "extend_method": "None", "ffn_dim_multiplier": null, "fine_grain_interleave": 1, "fused_rms_norm": true, "hidden_size": 3584, "ignore_token_id": -100, "init_method_std": 0.01, "input_sliced_sig": false, "intermediate_size": 18944, "is_dynamic": false, "kv_channels": 128, "layernorm_compute_type": "float32", "llm_backend": "", "max_decode_length": 512, "max_position_embedding": 131072, "mindformers_version": "1.6.0", "model_name": "llama", "model_type": "llama", "multiple_of": 256, "n_kv_heads": 4, "num_blocks": 128, "num_heads": 28, "num_layers": 2, "offset": 0, "pad_token_id": 151643, "parallel_config": { "micro_batch_num": 4, "model_parallel": 4, "pipeline_stage": 2, "use_seq_parallel": true, "vocab_emb_dp": false }, "parallel_decoding_params": null, "parallel_optimizer": false, "param_init_type": "float32", "pp_interleave_num": 1, "qkv_concat": false, "qkv_has_bias": true, "quant_config": null, "repetition_penalty": 1, "residual_dtype": "bfloat16", "return_hidden_states": false, "rms_norm_eps": 1e-06, "rmsnorm_compute_2d": false, "rotary_dtype": "float16", "rotary_emb_base": 1000000, "rotary_pct": 1.0, "scaling_factor": 1.0, "seq_length": 8192, "softmax_compute_type": "float16", "stage_num": 0, "start_stage": 0, "theta": 1000000.0, "tie_word_embeddings": false, "top_k": 0, "top_p": 0.8, "type": "LlamaConfig", "use_attn_mask_compression": false, "use_eod_attn_mask_compression": false, "use_flash_attention": true, "use_past": false, "use_ring_attention": false, "use_rope_slice": false, "vocab_size": 152064 } 2025-05-21 04:31:21,733 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:109] - INFO - set packing_sample_length to 8192 2025-05-21 04:31:21,733 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:110] - INFO - GRPOTrainer: finish init workers 2025-05-21 04:31:21,753 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:21,753 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:21,753 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:21,754 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:21,754 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:21,754 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:21,754 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:21,755 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:21,755 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:21,755 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:21,755 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:21,755 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:21,756 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:21,756 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:21,756 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:21,756 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:21,756 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:21,777 - mindformers./output/log[mindformers/models/utils.py:190] - INFO - num_layers per stage: [[1, 1]] 2025-05-21 04:31:21,777 - mindformers./output/log[mindformers/models/utils.py:191] - INFO - Accumulated num_layers per stage: [[1, 2]] 2025-05-21 04:31:21,777 - mindformers./output/log[mindformers/models/utils.py:193] - INFO - Pipeline id list with start_stage: [0, 1] 2025-05-21 04:31:21,778 - mindformers./output/log[mindformers/models/utils.py:194] - INFO - Interleave id list: [0, 0] 2025-05-21 04:31:21,778 - mindformers./output/log[mindformers/models/utils.py:212] - INFO - Formative layer_recompute: [[0, 0]] 2025-05-21 04:31:21,778 - mindformers./output/log[mindformers/models/utils.py:214] - INFO - The configuration of select_recompute_exclude and select_comm_recompute_exclude have the highest priority. 2025-05-21 04:31:21,778 - mindformers./output/log[mindformers/models/utils.py:220] - INFO - Formative select_recompute: {'feed_forward\\.mul': [[0, 0]], 'feed_forward\\.w1\\.activation\\.silu': [[0, 0]]} 2025-05-21 04:31:21,778 - mindformers./output/log[mindformers/models/utils.py:221] - INFO - Formative select_comm_recompute: {'.*\\.norm': [[0, 0]]} 2025-05-21 04:31:21,779 - mindformers./output/log[mindformers/models/utils.py:222] - INFO - Formative select_recompute_exclude: {} 2025-05-21 04:31:21,779 - mindformers./output/log[mindformers/models/utils.py:223] - INFO - Formative select_comm_recompute_exclude: {} 2025-05-21 04:31:21,836 - mindformers./output/log[mindformers/models/modeling_utils.py:1494] - INFO - model built, but weights is unloaded, since the config has no checkpoint_name_or_path attribute or checkpoint_name_or_path is None. 2025-05-21 04:31:21,846 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:105] - INFO - config of sft_model_config_train LlamaConfig { "attn_proj_has_bias": false, "batch_size": 1, "block_size": 32, "bos_token_id": 1, "calculate_per_token_loss": false, "checkpoint_name_or_path": null, "chunk_prefill": false, "compute_dtype": "bfloat16", "compute_in_2d": true, "do_sample": false, "emb_dropout_prob": 0.0, "embedding_init_type": "float32", "eos_token_id": 151643, "extend_method": "None", "ffn_dim_multiplier": null, "fine_grain_interleave": 1, "fused_rms_norm": true, "hidden_size": 3584, "ignore_token_id": -100, "init_method_std": 0.01, "input_sliced_sig": false, "intermediate_size": 18944, "is_dynamic": false, "kv_channels": 128, "layernorm_compute_type": "float32", "llm_backend": "", "max_decode_length": 512, "max_position_embedding": 131072, "mindformers_version": "1.6.0", "model_name": "llama", "model_type": "llama", "multiple_of": 256, "n_kv_heads": 4, "num_blocks": 128, "num_heads": 28, "num_layers": 2, "offset": 0, "pad_token_id": 151643, "parallel_config": { "micro_batch_num": 4, "model_parallel": 4, "pipeline_stage": 2, "use_seq_parallel": true, "vocab_emb_dp": false }, "parallel_decoding_params": null, "parallel_optimizer": false, "param_init_type": "float32", "pp_interleave_num": 1, "qkv_concat": false, "qkv_has_bias": true, "quant_config": null, "repetition_penalty": 1, "residual_dtype": "bfloat16", "return_hidden_states": false, "rms_norm_eps": 1e-06, "rmsnorm_compute_2d": false, "rotary_dtype": "float16", "rotary_emb_base": 1000000, "rotary_pct": 1.0, "scaling_factor": 1.0, "seq_length": 8192, "softmax_compute_type": "float16", "stage_num": 0, "start_stage": 0, "theta": 1000000.0, "tie_word_embeddings": false, "top_k": 0, "top_p": 0.8, "type": "LlamaConfig", "use_attn_mask_compression": false, "use_eod_attn_mask_compression": false, "use_flash_attention": true, "use_past": false, "use_ring_attention": false, "use_rope_slice": false, "vocab_size": 152064 } 2025-05-21 04:31:21,846 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:109] - INFO - set packing_sample_length to 8192 2025-05-21 04:31:21,846 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:110] - INFO - GRPOTrainer: finish init workers 2025-05-21 04:31:21,866 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_config is empty. 2025-05-21 04:31:21,866 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config context is empty. 2025-05-21 04:31:21,866 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel is empty. 2025-05-21 04:31:21,867 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config trainer is empty. 2025-05-21 04:31:21,867 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config moe_config is empty. 2025-05-21 04:31:21,867 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config parallel_config is empty. 2025-05-21 04:31:21,867 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config recompute_config is empty. 2025-05-21 04:31:21,867 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config swap_config is empty. 2025-05-21 04:31:21,868 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config runner_wrapper is empty. 2025-05-21 04:31:21,868 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config optimizer is empty. 2025-05-21 04:31:21,868 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config lr_schedule is empty. 2025-05-21 04:31:21,868 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config metric is empty. 2025-05-21 04:31:21,868 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset is empty. 2025-05-21 04:31:21,869 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config train_dataset_task is empty. 2025-05-21 04:31:21,869 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config callbacks is empty. 2025-05-21 04:31:21,869 - mindformers./output/log[mindformers/tools/register/template.py:84] - WARNING - The input config monitor_config is empty. 2025-05-21 04:31:21,869 - mindformers./output/log[mindformers/tools/register/template.py:683] - WARNING - Some configs in yaml are useless for finetune: ['processor'] 2025-05-21 04:31:32,436 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:273] - INFO - enable_reshard_optimizer:False 2025-05-21 04:31:32,436 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:273] - INFO - enable_reshard_optimizer:False 2025-05-21 04:31:32,436 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:273] - INFO - enable_reshard_optimizer:False 2025-05-21 04:31:32,436 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:273] - INFO - enable_reshard_optimizer:False 2025-05-21 04:31:32,436 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:273] - INFO - enable_reshard_optimizer:False 2025-05-21 04:31:32,436 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:273] - INFO - enable_reshard_optimizer:False 2025-05-21 04:31:32,436 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:273] - INFO - enable_reshard_optimizer:False 2025-05-21 04:31:32,436 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:273] - INFO - enable_reshard_optimizer:False ----------------start save front parallel strategy--------------- ----------------end save front parallel strategy--------------- ----------------start save front parallel strategy--------------- ----------------end save front parallel strategy--------------- ----------------start save front parallel strategy--------------- ----------------end save front parallel strategy--------------- ----------------start save front parallel strategy--------------- ----------------end save front parallel strategy--------------- ----------------start save front parallel strategy--------------- ----------------end save front parallel strategy--------------- ----------------start save front parallel strategy--------------- ----------------end save front parallel strategy--------------- ----------------start save front parallel strategy--------------- ----------------end save front parallel strategy--------------- ----------------start save front parallel strategy--------------- ----------------end save front parallel strategy--------------- 2025-05-21 04:32:08,467 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/transform_worker.py:144] - INFO - Start prepare for parameter resharding in sft training. 2025-05-21 04:32:09,308 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/transform_worker.py:144] - INFO - Start prepare for parameter resharding in sft training. 2025-05-21 04:32:09,472 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/transform_worker.py:144] - INFO - Start prepare for parameter resharding in sft training. 2025-05-21 04:32:09,644 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/transform_worker.py:144] - INFO - Start prepare for parameter resharding in sft training. 2025-05-21 04:32:10,344 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/transform_worker.py:144] - INFO - Start prepare for parameter resharding in sft training. 2025-05-21 04:32:10,416 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/transform_worker.py:144] - INFO - Start prepare for parameter resharding in sft training. 2025-05-21 04:32:11,459 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/transform_worker.py:144] - INFO - Start prepare for parameter resharding in sft training. 2025-05-21 04:32:11,952 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/transform_worker.py:144] - INFO - Start prepare for parameter resharding in sft training. 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:453] - INFO - sft_ckpt_path_infer: Waiting for main worker to merge strategies. 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:453] - INFO - sft_ckpt_path_infer: Waiting for main worker to merge strategies. 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:453] - INFO - sft_ckpt_path_infer: Waiting for main worker to merge strategies. 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:453] - INFO - sft_ckpt_path_infer: Waiting for main worker to merge strategies. 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:453] - INFO - sft_ckpt_path_infer: Waiting for main worker to merge strategies. 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:453] - INFO - sft_ckpt_path_infer: Waiting for main worker to merge strategies. 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:453] - INFO - sft_ckpt_path_infer: Waiting for main worker to merge strategies. 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:453] - INFO - sft_ckpt_path_infer: 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:459] - INFO - use_parallel is True 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:459] - INFO - use_parallel is True 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:459] - INFO - use_parallel is True 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:459] - INFO - use_parallel is True 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:459] - INFO - use_parallel is True 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:459] - INFO - use_parallel is True 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:459] - INFO - use_parallel is True 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:459] - INFO - use_parallel is True 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:218] - INFO - ref_ckpt_path: 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:218] - INFO - ref_ckpt_path: 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:218] - INFO - ref_ckpt_path: 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:218] - INFO - ref_ckpt_path: 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:218] - INFO - ref_ckpt_path: 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:218] - INFO - ref_ckpt_path: 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:218] - INFO - ref_ckpt_path: 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:226] - INFO - use_parallel is True 2025-05-21 04:32:22,108 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:218] - INFO - ref_ckpt_path: 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:226] - INFO - use_parallel is True 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:226] - INFO - use_parallel is True 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:226] - INFO - use_parallel is True 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:226] - INFO - use_parallel is True 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:226] - INFO - use_parallel is True 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:243] - INFO - use_parallel is True 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:226] - INFO - use_parallel is True 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:243] - INFO - use_parallel is True 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/ref_worker.py:226] - INFO - use_parallel is True 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:243] - INFO - use_parallel is True 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:243] - INFO - use_parallel is True 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:243] - INFO - use_parallel is True 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:243] - INFO - use_parallel is True 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:232] - INFO - sft_ckpt_path_train: 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:243] - INFO - use_parallel is True 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:232] - INFO - sft_ckpt_path_train: 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:232] - INFO - sft_ckpt_path_train: 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/old_policy_worker.py:243] - INFO - use_parallel is True 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:232] - INFO - sft_ckpt_path_train: 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:232] - INFO - sft_ckpt_path_train: 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:232] - INFO - sft_ckpt_path_train: 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:242] - INFO - use_parallel is True, 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:232] - INFO - sft_ckpt_path_train: 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:242] - INFO - use_parallel is True, 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:242] - INFO - use_parallel is True, 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:232] - INFO - sft_ckpt_path_train: 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:242] - INFO - use_parallel is True, 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:242] - INFO - use_parallel is True, 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:242] - INFO - use_parallel is True, 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:242] - INFO - use_parallel is True, 2025-05-21 04:32:22,109 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/train_worker.py:242] - INFO - use_parallel is True, 2025-05-21 04:32:42,524 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:956] - INFO - Start training epoch num:10, step num:1, generation num:8 2025-05-21 04:32:42,523 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:956] - INFO - Start training epoch num:10, step num:1, generation num:8 2025-05-21 04:32:42,523 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:956] - INFO - Start training epoch num:10, step num:1, generation num:8 2025-05-21 04:32:42,524 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:956] - INFO - Start training epoch num:10, step num:1, generation num:8 2025-05-21 04:32:42,524 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:966] - INFO - step begin at 04:32:42 ------------------------------- 2025-05-21 04:32:42,524 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:966] - INFO - step begin at 04:32:42 ------------------------------- 2025-05-21 04:32:42,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:966] - INFO - step begin at 04:32:42 ------------------------------- 2025-05-21 04:32:42,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:966] - INFO - step begin at 04:32:42 ------------------------------- 2025-05-21 04:32:42,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:969] - INFO - epoch: 0, step: 0 2025-05-21 04:32:42,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:969] - INFO - epoch: 0, step: 0 2025-05-21 04:32:42,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:969] - INFO - epoch: 0, step: 0 2025-05-21 04:32:42,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:969] - INFO - epoch: 0, step: 0 2025-05-21 04:32:42,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:615] - INFO - Make experience begin at 04:32:42 ------------------------------- 2025-05-21 04:32:42,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:615] - INFO - Make experience begin at 04:32:42 ------------------------------- 2025-05-21 04:32:42,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:615] - INFO - Make experience begin at 04:32:42 ------------------------------- 2025-05-21 04:32:42,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:615] - INFO - Make experience begin at 04:32:42 ------------------------------- 2025-05-21 04:32:42,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:617] - INFO - Generate 8 times 2025-05-21 04:32:42,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:617] - INFO - Generate 8 times 2025-05-21 04:32:42,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:617] - INFO - Generate 8 times 2025-05-21 04:32:42,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:617] - INFO - Generate 8 times 2025-05-21 04:32:42,875 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:643] - INFO - solution: ['$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$'] 2025-05-21 04:32:42,877 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:650] - INFO - generation start at 04:32:42------------------------------- 2025-05-21 04:32:42,878 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:32:42,878 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:643] - INFO - solution: ['$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$'] 2025-05-21 04:32:42,879 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:32:42,880 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:32:42,880 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:32:42,880 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:650] - INFO - generation start at 04:32:42------------------------------- 2025-05-21 04:32:42,881 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:32:42,882 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:32:42,882 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:32:42,883 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:32:42,885 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:643] - INFO - solution: ['$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$'] 2025-05-21 04:32:42,886 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:650] - INFO - generation start at 04:32:42------------------------------- 2025-05-21 04:32:42,886 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:643] - INFO - solution: ['$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$'] 2025-05-21 04:32:42,886 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:32:42,887 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:32:42,888 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:650] - INFO - generation start at 04:32:42------------------------------- 2025-05-21 04:32:42,889 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:32:42,889 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:32:42,889 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:32:42,889 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:32:42,888 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:32:42,889 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:32:42,890 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:32:42,890 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:32:42,890 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:32:42,890 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:32:42,890 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:32:42,891 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:32:42,892 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:32:42,892 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:32:42,893 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:32:42,893 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:32:42,895 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:32:42,895 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:32:42,897 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:32:42,897 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:32:42,898 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:32:42,899 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:32:42,945 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:956] - INFO - Start training epoch num:10, step num:1, generation num:8 2025-05-21 04:32:42,946 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:966] - INFO - step begin at 04:32:42 ------------------------------- 2025-05-21 04:32:42,946 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:969] - INFO - epoch: 0, step: 0 2025-05-21 04:32:42,946 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:615] - INFO - Make experience begin at 04:32:42 ------------------------------- 2025-05-21 04:32:42,947 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:617] - INFO - Generate 8 times 2025-05-21 04:32:42,951 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:956] - INFO - Start training epoch num:10, step num:1, generation num:8 2025-05-21 04:32:42,952 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:966] - INFO - step begin at 04:32:42 ------------------------------- 2025-05-21 04:32:42,952 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:969] - INFO - epoch: 0, step: 0 2025-05-21 04:32:42,952 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:615] - INFO - Make experience begin at 04:32:42 ------------------------------- 2025-05-21 04:32:42,953 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:617] - INFO - Generate 8 times 2025-05-21 04:32:42,971 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:956] - INFO - Start training epoch num:10, step num:1, generation num:8 2025-05-21 04:32:42,972 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:966] - INFO - step begin at 04:32:42 ------------------------------- 2025-05-21 04:32:42,972 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:969] - INFO - epoch: 0, step: 0 2025-05-21 04:32:42,972 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:615] - INFO - Make experience begin at 04:32:42 ------------------------------- 2025-05-21 04:32:42,972 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:617] - INFO - Generate 8 times 2025-05-21 04:32:42,975 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:956] - INFO - Start training epoch num:10, step num:1, generation num:8 2025-05-21 04:32:42,976 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:966] - INFO - step begin at 04:32:42 ------------------------------- 2025-05-21 04:32:42,976 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:969] - INFO - epoch: 0, step: 0 2025-05-21 04:32:42,977 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:615] - INFO - Make experience begin at 04:32:42 ------------------------------- 2025-05-21 04:32:42,977 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:617] - INFO - Generate 8 times 2025-05-21 04:32:43,288 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:643] - INFO - solution: ['$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$'] 2025-05-21 04:32:43,289 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:650] - INFO - generation start at 04:32:43------------------------------- 2025-05-21 04:32:43,290 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:32:43,290 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:32:43,290 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:643] - INFO - solution: ['$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$'] 2025-05-21 04:32:43,290 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:32:43,291 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:650] - INFO - generation start at 04:32:43------------------------------- 2025-05-21 04:32:43,291 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:32:43,291 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:32:43,292 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:32:43,292 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:32:43,292 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:32:43,294 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:32:43,295 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:32:43,296 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:32:43,295 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:32:43,297 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:32:43,297 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:32:43,297 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:32:43,298 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:32:43,314 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:643] - INFO - solution: ['$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$'] 2025-05-21 04:32:43,315 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:650] - INFO - generation start at 04:32:43------------------------------- 2025-05-21 04:32:43,316 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:32:43,316 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:32:43,316 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:32:43,317 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:32:43,320 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:32:43,320 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:643] - INFO - solution: ['$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$', '$Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?5$', '$Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?72$', '$Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?42$', '$Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?10$'] 2025-05-21 04:32:43,321 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:32:43,321 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:650] - INFO - generation start at 04:32:43------------------------------- 2025-05-21 04:32:43,322 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:32:43,322 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:32:43,322 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:32:43,323 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:32:43,323 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:32:43,323 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:32:43,326 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:32:43,327 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:32:43,328 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:32:43,329 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:03,194 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 20.294841051101685 s; generated tokens: 512 tokens; generate speed: 25.228086226977698 tokens/s 2025-05-21 04:33:03,194 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 20.294527292251587 s; generated tokens: 512 tokens; generate speed: 25.228476259975793 tokens/s 2025-05-21 04:33:03,194 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 20.301284074783325 s; generated tokens: 512 tokens; generate speed: 25.22007958284602 tokens/s 2025-05-21 04:33:03,194 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 20.30215334892273 s; generated tokens: 512 tokens; generate speed: 25.218999738624657 tokens/s 2025-05-21 04:33:03,194 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.006481170654296875 s; prefill predict time: 8.015339374542236 s; prefill post time: 0.10824084281921387 s; decode prepare time: 0.0011001221121174016 s; decode predict time: 0.005192260648690018 s; decode post time: 0.013227496828351702 s 2025-05-21 04:33:03,195 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.006318092346191406 s; prefill predict time: 7.979745864868164 s; prefill post time: 0.11244964599609375 s; decode prepare time: 0.0011461093701028544 s; decode predict time: 0.004885477645724427 s; decode post time: 0.013492227300505815 s 2025-05-21 04:33:03,195 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0059773921966552734 s; prefill predict time: 7.986125230789185 s; prefill post time: 0.11741423606872559 s; decode prepare time: 0.0010519629588565714 s; decode predict time: 0.0051428439570408244 s; decode post time: 0.013329253971226761 s 2025-05-21 04:33:03,195 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.006823062896728516 s; prefill predict time: 7.9861016273498535 s; prefill post time: 0.11827373504638672 s; decode prepare time: 0.0011097106448358285 s; decode predict time: 0.004692863483054965 s; decode post time: 0.013717996164310702 s 2025-05-21 04:33:03,195 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:03,195 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:03,196 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:03,196 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:03,196 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:03,196 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:03,196 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:03,196 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 20.30729365348816 2025-05-21 04:33:03,196 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 20.306289672851562 2025-05-21 04:33:03,196 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 20.313792943954468 2025-05-21 04:33:03,196 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:03,196 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 20.316969394683838 2025-05-21 04:33:03,197 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:03,197 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:03,197 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:03,197 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:03,198 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:03,198 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:03,198 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:03,198 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:03,198 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:03,198 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:03,198 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:03,198 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:03,198 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:03,198 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:03,198 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:03,198 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:03,200 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:03,200 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:03,200 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:03,200 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:03,201 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:03,201 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:03,201 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:03,201 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:03,201 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:03,201 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:03,201 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:03,202 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:03,202 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:03,202 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:03,202 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:03,202 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:03,847 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 20.55010747909546 s; generated tokens: 512 tokens; generate speed: 24.91471154206033 tokens/s 2025-05-21 04:33:03,847 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 20.52436661720276 s; generated tokens: 512 tokens; generate speed: 24.945958603704764 tokens/s 2025-05-21 04:33:03,847 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 20.548694849014282 s; generated tokens: 512 tokens; generate speed: 24.916424316095217 tokens/s 2025-05-21 04:33:03,847 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 20.517754554748535 s; generated tokens: 512 tokens; generate speed: 24.953997701542104 tokens/s 2025-05-21 04:33:03,848 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0027420520782470703 s; prefill predict time: 8.038374900817871 s; prefill post time: 0.09461712837219238 s; decode prepare time: 0.0011529026666032825 s; decode predict time: 0.004886150360107422 s; decode post time: 0.013722027817817584 s 2025-05-21 04:33:03,848 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0027132034301757812 s; prefill predict time: 8.037126541137695 s; prefill post time: 0.09724140167236328 s; decode prepare time: 0.0011385033041996966 s; decode predict time: 0.004794552279453652 s; decode post time: 0.013827014575965716 s 2025-05-21 04:33:03,848 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0031290054321289062 s; prefill predict time: 8.005379438400269 s; prefill post time: 0.09688591957092285 s; decode prepare time: 0.0010602819476342247 s; decode predict time: 0.0051918389750461955 s; decode post time: 0.013509787692015884 s 2025-05-21 04:33:03,848 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0030100345611572266 s; prefill predict time: 8.011772155761719 s; prefill post time: 0.10008573532104492 s; decode prepare time: 0.0010764421548862047 s; decode predict time: 0.00488433557398179 s; decode post time: 0.013800388679579046 s 2025-05-21 04:33:03,849 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:03,849 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:03,849 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:03,849 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:03,849 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:03,850 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:03,850 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 20.55932307243347 2025-05-21 04:33:03,850 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:03,850 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:03,850 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 20.527354955673218 2025-05-21 04:33:03,850 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 20.533690929412842 2025-05-21 04:33:03,850 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 20.558024406433105 2025-05-21 04:33:03,850 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:03,851 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:03,851 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:03,851 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:03,851 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:03,851 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:03,851 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:03,851 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:03,851 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:03,851 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:03,851 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:03,852 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:03,852 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:03,852 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:03,852 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:03,852 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:03,854 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:03,854 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:03,854 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:03,854 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:03,854 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:03,854 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:03,855 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:03,855 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:03,855 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:03,855 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:03,855 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:03,855 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:03,855 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:03,855 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:03,856 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:03,856 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:12,888 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.685579299926758 s; generated tokens: 512 tokens; generate speed: 52.86209364925356 tokens/s 2025-05-21 04:33:12,888 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.685811758041382 s; generated tokens: 512 tokens; generate speed: 52.860824966469735 tokens/s 2025-05-21 04:33:12,888 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.686177730560303 s; generated tokens: 512 tokens; generate speed: 52.85882772774427 tokens/s 2025-05-21 04:33:12,889 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014765262603759766 s; prefill predict time: 0.010271072387695312 s; prefill post time: 0.01442575454711914 s; decode prepare time: 0.0010267293149944388 s; decode predict time: 0.005074850718180339 s; decode post time: 0.012757012988723188 s 2025-05-21 04:33:12,889 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.686322450637817 s; generated tokens: 512 tokens; generate speed: 52.858037981823145 tokens/s 2025-05-21 04:33:12,889 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015096664428710938 s; prefill predict time: 0.010309696197509766 s; prefill post time: 0.014306783676147461 s; decode prepare time: 0.0010262174848940974 s; decode predict time: 0.0051409001443900315 s; decode post time: 0.012687103622346941 s 2025-05-21 04:33:12,889 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015480518341064453 s; prefill predict time: 0.01034402847290039 s; prefill post time: 0.01502227783203125 s; decode prepare time: 0.001115619786331565 s; decode predict time: 0.004756894766115675 s; decode post time: 0.012984853434702655 s 2025-05-21 04:33:12,890 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001505136489868164 s; prefill predict time: 0.010048627853393555 s; prefill post time: 0.015370368957519531 s; decode prepare time: 0.001068527684519669 s; decode predict time: 0.004456078772451364 s; decode post time: 0.013331713275200234 s 2025-05-21 04:33:12,890 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:12,890 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:12,890 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:12,890 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:12,890 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.69271969795227 2025-05-21 04:33:12,890 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:12,891 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:12,891 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.69312047958374 2025-05-21 04:33:12,891 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.6931631565094 2025-05-21 04:33:12,891 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:12,891 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:12,891 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.693387746810913 2025-05-21 04:33:12,891 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:12,892 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:12,892 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:12,892 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:12,892 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:12,892 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:12,892 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:12,892 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:12,892 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:12,892 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:12,893 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:12,893 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:12,893 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:12,893 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:12,893 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:12,893 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:12,894 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:12,895 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:12,895 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:12,895 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:12,895 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:12,896 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:12,896 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:12,896 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:12,896 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:12,896 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:12,896 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:12,896 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:12,897 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:12,897 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:12,897 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:12,897 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:13,561 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.705762147903442 s; generated tokens: 512 tokens; generate speed: 52.752168474538394 tokens/s 2025-05-21 04:33:13,562 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.706255674362183 s; generated tokens: 512 tokens; generate speed: 52.749486225917344 tokens/s 2025-05-21 04:33:13,562 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.705792665481567 s; generated tokens: 512 tokens; generate speed: 52.752002607774266 tokens/s 2025-05-21 04:33:13,562 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015749931335449219 s; prefill predict time: 0.011751413345336914 s; prefill post time: 0.014977216720581055 s; decode prepare time: 0.0010174841796815278 s; decode predict time: 0.004895937209035836 s; decode post time: 0.012977784393818179 s 2025-05-21 04:33:13,562 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.705845594406128 s; generated tokens: 512 tokens; generate speed: 52.75171493507854 tokens/s 2025-05-21 04:33:13,563 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001636505126953125 s; prefill predict time: 0.011305093765258789 s; prefill post time: 0.014384984970092773 s; decode prepare time: 0.0010715786025015574 s; decode predict time: 0.00455026486340691 s; decode post time: 0.01326909457167534 s 2025-05-21 04:33:13,563 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0016608238220214844 s; prefill predict time: 0.011692523956298828 s; prefill post time: 0.014846324920654297 s; decode prepare time: 0.001081026230306308 s; decode predict time: 0.004577766680249981 s; decode post time: 0.01323487520684468 s 2025-05-21 04:33:13,563 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0016491413116455078 s; prefill predict time: 0.011106491088867188 s; prefill post time: 0.014786720275878906 s; decode prepare time: 0.0010849426403904147 s; decode predict time: 0.004490134295295266 s; decode post time: 0.013317850004670204 s 2025-05-21 04:33:13,563 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:13,564 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:13,564 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:13,564 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:13,564 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:13,564 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.712826490402222 2025-05-21 04:33:13,564 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:13,564 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:13,564 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:13,564 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.712778806686401 2025-05-21 04:33:13,564 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.713291645050049 2025-05-21 04:33:13,564 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.712943315505981 2025-05-21 04:33:13,565 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:13,565 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:13,565 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:13,565 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:13,566 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:13,566 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:13,566 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:13,566 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:13,566 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:13,566 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:13,566 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:13,566 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:13,566 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:13,566 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:13,566 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:13,566 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:13,568 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:13,568 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:13,569 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:13,569 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:13,569 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:13,569 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:13,569 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:13,570 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:13,570 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:13,570 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:13,570 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:13,570 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:13,570 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:13,570 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:13,570 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:13,571 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:22,506 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.608523845672607 s; generated tokens: 512 tokens; generate speed: 53.286020644116896 tokens/s 2025-05-21 04:33:22,506 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.609848499298096 s; generated tokens: 512 tokens; generate speed: 53.2786755209925 tokens/s 2025-05-21 04:33:22,507 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014522075653076172 s; prefill predict time: 0.009200811386108398 s; prefill post time: 0.014109134674072266 s; decode prepare time: 0.0010373732348477537 s; decode predict time: 0.005015052533617207 s; decode post time: 0.012653954108400587 s 2025-05-21 04:33:22,507 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001478433609008789 s; prefill predict time: 0.010454654693603516 s; prefill post time: 0.014774560928344727 s; decode prepare time: 0.0010148825710766937 s; decode predict time: 0.004989712846045401 s; decode post time: 0.012707015539568697 s 2025-05-21 04:33:22,507 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.609097003936768 s; generated tokens: 512 tokens; generate speed: 53.282842268137976 tokens/s 2025-05-21 04:33:22,507 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.610116720199585 s; generated tokens: 512 tokens; generate speed: 53.277188499055676 tokens/s 2025-05-21 04:33:22,508 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013813972473144531 s; prefill predict time: 0.00904536247253418 s; prefill post time: 0.015073060989379883 s; decode prepare time: 0.0010527169401165091 s; decode predict time: 0.004397786832323261 s; decode post time: 0.01325728468698998 s 2025-05-21 04:33:22,508 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:22,508 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:22,508 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014390945434570312 s; prefill predict time: 0.009752273559570312 s; prefill post time: 0.014901876449584961 s; decode prepare time: 0.0011122520655801852 s; decode predict time: 0.004626715417001761 s; decode post time: 0.012971270574282294 s 2025-05-21 04:33:22,508 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:22,508 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:22,508 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.615909099578857 2025-05-21 04:33:22,508 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.616534233093262 2025-05-21 04:33:22,509 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:22,509 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:22,509 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:22,509 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:22,509 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.616640090942383 2025-05-21 04:33:22,509 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:22,509 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.61719536781311 2025-05-21 04:33:22,509 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:22,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:22,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:22,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:22,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:22,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:22,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:22,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:22,511 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:22,511 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:22,511 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:22,511 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:22,511 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:22,511 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:22,511 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:22,512 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:22,513 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:22,513 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:22,513 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:22,514 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:22,514 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:22,514 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:22,514 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:22,514 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:22,514 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:22,514 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:22,514 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:22,515 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:22,515 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:22,515 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:22,515 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:23,254 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.682559251785278 s; generated tokens: 512 tokens; generate speed: 52.878581652428 tokens/s 2025-05-21 04:33:23,254 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.68299913406372 s; generated tokens: 512 tokens; generate speed: 52.876179467871744 tokens/s 2025-05-21 04:33:23,254 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.683052778244019 s; generated tokens: 512 tokens; generate speed: 52.87588653346668 tokens/s 2025-05-21 04:33:23,254 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.683226108551025 s; generated tokens: 512 tokens; generate speed: 52.87494005204165 tokens/s 2025-05-21 04:33:23,254 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014462471008300781 s; prefill predict time: 0.010240793228149414 s; prefill post time: 0.013954639434814453 s; decode prepare time: 0.0010141855116693008 s; decode predict time: 0.004901252540887571 s; decode post time: 0.012933933571593403 s 2025-05-21 04:33:23,254 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015676021575927734 s; prefill predict time: 0.00977778434753418 s; prefill post time: 0.014643192291259766 s; decode prepare time: 0.0010893153584400036 s; decode predict time: 0.004509749599531585 s; decode post time: 0.013251525315286596 s 2025-05-21 04:33:23,255 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013630390167236328 s; prefill predict time: 0.010462760925292969 s; prefill post time: 0.014579057693481445 s; decode prepare time: 0.0010803394354952757 s; decode predict time: 0.004433435552260455 s; decode post time: 0.013335785053947434 s 2025-05-21 04:33:23,255 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014083385467529297 s; prefill predict time: 0.010446310043334961 s; prefill post time: 0.014360427856445312 s; decode prepare time: 0.0010657212505601856 s; decode predict time: 0.00461127664528641 s; decode post time: 0.01317204981167489 s 2025-05-21 04:33:23,255 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:23,255 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:23,256 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:23,256 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:23,256 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:23,256 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:23,256 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.689884424209595 2025-05-21 04:33:23,256 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:23,256 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.6903076171875 2025-05-21 04:33:23,256 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:23,256 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.690152168273926 2025-05-21 04:33:23,256 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.690330743789673 2025-05-21 04:33:23,257 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:23,257 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:23,257 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:23,258 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:23,258 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:23,258 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:23,258 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:23,258 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:23,258 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:23,258 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:23,258 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:23,258 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:23,258 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:23,258 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:23,258 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:23,258 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:23,260 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:23,260 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:23,260 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:23,261 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:23,261 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:23,261 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:23,261 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:23,261 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:23,261 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:23,262 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:23,262 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:23,262 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:23,262 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:23,262 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:23,262 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:23,263 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:32,154 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.639119625091553 s; generated tokens: 512 tokens; generate speed: 53.116884104977274 tokens/s 2025-05-21 04:33:32,154 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.639004468917847 s; generated tokens: 512 tokens; generate speed: 53.11751868681116 tokens/s 2025-05-21 04:33:32,154 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.638319969177246 s; generated tokens: 512 tokens; generate speed: 53.12129101724621 tokens/s 2025-05-21 04:33:32,155 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001377105712890625 s; prefill predict time: 0.01099705696105957 s; prefill post time: 0.014574050903320312 s; decode prepare time: 0.0010082497755375393 s; decode predict time: 0.00501856897391525 s; decode post time: 0.01272777755199813 s 2025-05-21 04:33:32,155 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0016756057739257812 s; prefill predict time: 0.009910345077514648 s; prefill post time: 0.01534271240234375 s; decode prepare time: 0.0010890386809108308 s; decode predict time: 0.004762588762769512 s; decode post time: 0.012913031587395415 s 2025-05-21 04:33:32,155 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001436471939086914 s; prefill predict time: 0.010786771774291992 s; prefill post time: 0.014641523361206055 s; decode prepare time: 0.0010167600590655483 s; decode predict time: 0.005085689881268669 s; decode post time: 0.012658860585460924 s 2025-05-21 04:33:32,155 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.638955354690552 s; generated tokens: 512 tokens; generate speed: 53.117789341232736 tokens/s 2025-05-21 04:33:32,155 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015523433685302734 s; prefill predict time: 0.009639501571655273 s; prefill post time: 0.01583075523376465 s; decode prepare time: 0.0010376111868542935 s; decode predict time: 0.004453335088842055 s; decode post time: 0.013273031977524729 s 2025-05-21 04:33:32,156 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:32,156 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:32,156 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:32,156 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:32,156 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:32,156 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.646080017089844 2025-05-21 04:33:32,156 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.64504861831665 2025-05-21 04:33:32,156 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:32,156 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.64639949798584 2025-05-21 04:33:32,156 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:32,157 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:32,157 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.645963191986084 2025-05-21 04:33:32,157 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:32,157 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:32,158 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:32,158 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:32,158 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:32,158 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:32,158 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:32,158 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:32,158 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:32,158 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:32,158 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:32,158 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:32,159 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:32,159 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:32,159 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:32,159 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:32,160 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:32,160 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:32,161 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:32,161 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:32,161 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:32,161 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:32,161 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:32,161 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:32,162 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:32,162 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:32,162 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:32,162 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:32,162 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:32,162 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:32,163 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:32,163 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:32,960 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.69792628288269 s; generated tokens: 512 tokens; generate speed: 52.794791903471655 tokens/s 2025-05-21 04:33:32,961 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.697615146636963 s; generated tokens: 512 tokens; generate speed: 52.796485760476536 tokens/s 2025-05-21 04:33:32,961 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.698033809661865 s; generated tokens: 512 tokens; generate speed: 52.79420654214564 tokens/s 2025-05-21 04:33:32,961 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014934539794921875 s; prefill predict time: 0.019257783889770508 s; prefill post time: 0.01568150520324707 s; decode prepare time: 0.00101302561228056 s; decode predict time: 0.004923031376857383 s; decode post time: 0.012921617922251005 s 2025-05-21 04:33:32,961 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.698799848556519 s; generated tokens: 512 tokens; generate speed: 52.79003670502608 tokens/s 2025-05-21 04:33:32,962 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015707015991210938 s; prefill predict time: 0.018645524978637695 s; prefill post time: 0.016031742095947266 s; decode prepare time: 0.001066778037412759 s; decode predict time: 0.004538206493153292 s; decode post time: 0.013251386276663167 s 2025-05-21 04:33:32,962 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015997886657714844 s; prefill predict time: 0.020067453384399414 s; prefill post time: 0.015985727310180664 s; decode prepare time: 0.0010957442617696325 s; decode predict time: 0.004437564868553012 s; decode post time: 0.013322931913013794 s 2025-05-21 04:33:32,962 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:32,962 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015473365783691406 s; prefill predict time: 0.019264698028564453 s; prefill post time: 0.016021013259887695 s; decode prepare time: 0.001083421147266246 s; decode predict time: 0.004573000178617589 s; decode post time: 0.013202580453831622 s 2025-05-21 04:33:32,963 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:32,963 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:32,963 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:32,963 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.705070972442627 2025-05-21 04:33:32,963 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:32,963 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:32,963 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.705138921737671 2025-05-21 04:33:32,963 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.70532774925232 2025-05-21 04:33:32,963 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:32,964 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:32,964 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.706157445907593 2025-05-21 04:33:32,964 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:32,964 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:32,964 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:32,965 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:32,965 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:32,965 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:32,965 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:32,965 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:32,965 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:32,965 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:32,965 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:32,966 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:32,966 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:32,966 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:32,966 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:32,966 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:32,967 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:32,968 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:32,968 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:32,968 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:32,968 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:32,968 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:32,969 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:32,968 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:32,969 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:32,969 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:32,969 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:32,969 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:32,969 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:32,970 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:32,970 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:32,970 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:41,713 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.550690412521362 s; generated tokens: 512 tokens; generate speed: 53.608689831338914 tokens/s 2025-05-21 04:33:41,713 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.55001950263977 s; generated tokens: 512 tokens; generate speed: 53.61245595974704 tokens/s 2025-05-21 04:33:41,713 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.551393985748291 s; generated tokens: 512 tokens; generate speed: 53.604740916766616 tokens/s 2025-05-21 04:33:41,714 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013570785522460938 s; prefill predict time: 0.00922393798828125 s; prefill post time: 0.01357412338256836 s; decode prepare time: 0.0010080542816341272 s; decode predict time: 0.004909252185447543 s; decode post time: 0.012680900073564682 s 2025-05-21 04:33:41,714 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.55060625076294 s; generated tokens: 512 tokens; generate speed: 53.60916224130792 tokens/s 2025-05-21 04:33:41,714 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001619100570678711 s; prefill predict time: 0.008041143417358398 s; prefill post time: 0.013979673385620117 s; decode prepare time: 0.0010241510350176967 s; decode predict time: 0.0049692312876383465 s; decode post time: 0.012600240875363582 s 2025-05-21 04:33:41,714 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014865398406982422 s; prefill predict time: 0.009365081787109375 s; prefill post time: 0.014088153839111328 s; decode prepare time: 0.0010958441082978669 s; decode predict time: 0.004606553619983149 s; decode post time: 0.012894768537607211 s 2025-05-21 04:33:41,715 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015676021575927734 s; prefill predict time: 0.008745670318603516 s; prefill post time: 0.014657974243164062 s; decode prepare time: 0.00103676202479174 s; decode predict time: 0.004335527326546463 s; decode post time: 0.013223061589578826 s 2025-05-21 04:33:41,715 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:41,715 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:41,715 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:41,715 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.557607412338257 2025-05-21 04:33:41,715 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:41,715 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:41,715 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:41,716 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.55747365951538 2025-05-21 04:33:41,716 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:41,716 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.55811882019043 2025-05-21 04:33:41,716 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:41,716 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.557552099227905 2025-05-21 04:33:41,716 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:41,717 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:41,717 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:41,717 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:41,717 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:41,717 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:41,717 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:41,717 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:41,717 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:41,718 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:41,718 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:41,718 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:41,718 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:41,718 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:41,718 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:41,718 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:41,719 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:41,720 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:41,720 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:41,720 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:41,720 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:41,721 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:41,721 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:41,721 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:41,721 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:41,721 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:41,721 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:41,721 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:41,722 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:41,722 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:41,722 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:41,722 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:42,604 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.63450574874878 s; generated tokens: 512 tokens; generate speed: 53.14232129307648 tokens/s 2025-05-21 04:33:42,604 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.63317608833313 s; generated tokens: 512 tokens; generate speed: 53.14965648973137 tokens/s 2025-05-21 04:33:42,604 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.634301900863647 s; generated tokens: 512 tokens; generate speed: 53.14344570768566 tokens/s 2025-05-21 04:33:42,604 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014722347259521484 s; prefill predict time: 0.007224559783935547 s; prefill post time: 0.014019012451171875 s; decode prepare time: 0.0010157559949124629 s; decode predict time: 0.004863978367225796 s; decode post time: 0.012881972785098679 s 2025-05-21 04:33:42,604 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.634170770645142 s; generated tokens: 512 tokens; generate speed: 53.144169040478246 tokens/s 2025-05-21 04:33:42,605 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014874935150146484 s; prefill predict time: 0.0057108402252197266 s; prefill post time: 0.01436614990234375 s; decode prepare time: 0.0010722602650144095 s; decode predict time: 0.004494064461951162 s; decode post time: 0.013196375505331677 s 2025-05-21 04:33:42,605 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001524209976196289 s; prefill predict time: 0.0066394805908203125 s; prefill post time: 0.014284133911132812 s; decode prepare time: 0.0010660319878863728 s; decode predict time: 0.004431013032501819 s; decode post time: 0.013263719189190351 s 2025-05-21 04:33:42,605 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:42,605 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014393329620361328 s; prefill predict time: 0.0062372684478759766 s; prefill post time: 0.014250040054321289 s; decode prepare time: 0.0011066331322179149 s; decode predict time: 0.00434108715431363 s; decode post time: 0.013314751029714446 s 2025-05-21 04:33:42,606 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:42,606 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:42,606 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.641477584838867 2025-05-21 04:33:42,606 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:42,606 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:42,606 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:42,606 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.640648126602173 2025-05-21 04:33:42,606 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:42,606 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.641724109649658 2025-05-21 04:33:42,607 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:42,607 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.641708612442017 2025-05-21 04:33:42,607 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:42,607 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:42,607 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:42,608 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:42,608 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:42,608 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:42,608 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:42,608 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:42,608 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:42,608 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:42,608 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:42,608 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:42,608 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:42,609 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:42,609 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:42,609 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:42,610 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:42,611 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:42,611 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:42,611 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:42,611 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:42,611 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:42,611 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:42,611 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:42,612 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:42,612 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:42,612 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:42,612 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:42,612 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:42,612 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:42,613 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:42,613 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:51,282 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.560274124145508 s; generated tokens: 512 tokens; generate speed: 53.554949717068105 tokens/s 2025-05-21 04:33:51,282 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.559805393218994 s; generated tokens: 512 tokens; generate speed: 53.55757559282266 tokens/s 2025-05-21 04:33:51,282 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.559800148010254 s; generated tokens: 512 tokens; generate speed: 53.557604978443614 tokens/s 2025-05-21 04:33:51,282 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013611316680908203 s; prefill predict time: 0.00786447525024414 s; prefill post time: 0.014148473739624023 s; decode prepare time: 0.0010022790464636398 s; decode predict time: 0.004970736597098556 s; decode post time: 0.012647966582714462 s 2025-05-21 04:33:51,283 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014119148254394531 s; prefill predict time: 0.0071604251861572266 s; prefill post time: 0.013524293899536133 s; decode prepare time: 0.0010165930260884084 s; decode predict time: 0.005012980629416073 s; decode post time: 0.012586358008785957 s 2025-05-21 04:33:51,283 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014281272888183594 s; prefill predict time: 0.007004737854003906 s; prefill post time: 0.014317512512207031 s; decode prepare time: 0.0010995799548005405 s; decode predict time: 0.004637284372367111 s; decode post time: 0.012882655380756655 s 2025-05-21 04:33:51,283 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.560014963150024 s; generated tokens: 512 tokens; generate speed: 53.55640153007627 tokens/s 2025-05-21 04:33:51,283 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:51,283 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:51,283 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013592243194580078 s; prefill predict time: 0.006543874740600586 s; prefill post time: 0.014514684677124023 s; decode prepare time: 0.0010290155205474674 s; decode predict time: 0.004394957598517922 s; decode post time: 0.013195112959979332 s 2025-05-21 04:33:51,284 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:51,284 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.566987037658691 2025-05-21 04:33:51,284 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:51,284 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:51,284 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:51,284 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.566938877105713 2025-05-21 04:33:51,284 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.566732168197632 2025-05-21 04:33:51,284 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:51,285 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:51,285 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:51,285 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.567094326019287 2025-05-21 04:33:51,285 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:51,285 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:51,285 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:51,286 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:51,286 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:51,286 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:51,286 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:51,286 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:51,286 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:51,286 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:51,286 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:51,286 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:51,287 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:51,287 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:51,287 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:51,288 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:51,288 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:51,288 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:51,288 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:51,289 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:51,289 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:51,289 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:51,289 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:51,290 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:51,290 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:51,289 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:51,290 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:51,290 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:51,290 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:51,290 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:51,292 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:52,237 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.625389814376831 s; generated tokens: 512 tokens; generate speed: 53.192650882072144 tokens/s 2025-05-21 04:33:52,238 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.625298023223877 s; generated tokens: 512 tokens; generate speed: 53.19315815101503 tokens/s 2025-05-21 04:33:52,238 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.624566555023193 s; generated tokens: 512 tokens; generate speed: 53.197200837348895 tokens/s 2025-05-21 04:33:52,238 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015192031860351562 s; prefill predict time: 0.010927200317382812 s; prefill post time: 0.014088630676269531 s; decode prepare time: 0.0010042596469886616 s; decode predict time: 0.004868133862813314 s; decode post time: 0.012864516905842461 s 2025-05-21 04:33:52,238 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.625265836715698 s; generated tokens: 512 tokens; generate speed: 53.193336026831545 tokens/s 2025-05-21 04:33:52,238 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014989376068115234 s; prefill predict time: 0.010723114013671875 s; prefill post time: 0.014610052108764648 s; decode prepare time: 0.0010702516701356771 s; decode predict time: 0.0044995611789179785 s; decode post time: 0.013168341492953366 s 2025-05-21 04:33:52,239 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014133453369140625 s; prefill predict time: 0.010153055191040039 s; prefill post time: 0.01420140266418457 s; decode prepare time: 0.0011057270483028165 s; decode predict time: 0.004402620184655283 s; decode post time: 0.013228279270537912 s 2025-05-21 04:33:52,239 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014917850494384766 s; prefill predict time: 0.009917974472045898 s; prefill post time: 0.014507532119750977 s; decode prepare time: 0.0010751086905044585 s; decode predict time: 0.004477568233714384 s; decode post time: 0.013184417250571652 s 2025-05-21 04:33:52,239 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:52,240 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:52,240 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:52,240 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:52,240 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.632531642913818 2025-05-21 04:33:52,240 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:52,240 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:52,240 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.631814241409302 2025-05-21 04:33:52,240 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.632537364959717 2025-05-21 04:33:52,240 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:33:52,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:33:52,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.632761001586914 2025-05-21 04:33:52,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:52,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:52,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:52,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:52,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:52,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:52,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:52,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:52,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:52,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:33:52,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:52,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:52,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:52,243 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:33:52,243 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:33:52,243 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:33:52,245 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:52,245 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:52,245 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:52,245 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:52,246 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:52,246 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:33:52,246 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:52,246 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:52,246 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:52,246 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:52,246 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:33:52,246 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:52,247 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:33:52,247 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:52,247 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:33:52,247 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:00,848 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.557403564453125 s; generated tokens: 512 tokens; generate speed: 53.57103490997104 tokens/s 2025-05-21 04:34:00,848 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.558358430862427 s; generated tokens: 512 tokens; generate speed: 53.56568323979492 tokens/s 2025-05-21 04:34:00,848 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.557902097702026 s; generated tokens: 512 tokens; generate speed: 53.56824068360131 tokens/s 2025-05-21 04:34:00,849 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014562606811523438 s; prefill predict time: 0.014421939849853516 s; prefill post time: 0.014524221420288086 s; decode prepare time: 0.001002263649569099 s; decode predict time: 0.005043381335688572 s; decode post time: 0.012550866767151714 s 2025-05-21 04:34:00,849 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013663768768310547 s; prefill predict time: 0.01536107063293457 s; prefill post time: 0.014159202575683594 s; decode prepare time: 0.000998952850670264 s; decode predict time: 0.0049794571072447535 s; decode post time: 0.012622030747146755 s 2025-05-21 04:34:00,849 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014281272888183594 s; prefill predict time: 0.014737606048583984 s; prefill post time: 0.014317512512207031 s; decode prepare time: 0.0010811377411485885 s; decode predict time: 0.004671122046077953 s; decode post time: 0.012846817009136402 s 2025-05-21 04:34:00,849 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.557268142700195 s; generated tokens: 512 tokens; generate speed: 53.57179398498551 tokens/s 2025-05-21 04:34:00,850 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:00,850 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:00,850 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0016200542449951172 s; prefill predict time: 0.013926267623901367 s; prefill post time: 0.014654397964477539 s; decode prepare time: 0.0010227321878571332 s; decode predict time: 0.004398448794495826 s; decode post time: 0.013178151181066106 s 2025-05-21 04:34:00,850 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:00,850 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:00,850 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:00,850 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.564437866210938 2025-05-21 04:34:00,850 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.5650794506073 2025-05-21 04:34:00,850 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:00,851 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.564777851104736 2025-05-21 04:34:00,851 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:00,851 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:00,851 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.564640045166016 2025-05-21 04:34:00,851 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:00,851 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:00,852 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:00,852 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:00,852 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:00,852 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:00,852 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:00,852 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:00,852 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:00,852 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:00,852 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:00,853 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:00,853 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:00,853 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:00,853 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:00,853 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:00,855 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:00,855 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:00,855 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:00,855 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:00,855 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:00,856 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:00,856 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:00,856 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:00,856 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:00,856 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:00,856 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:00,856 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:00,856 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:00,857 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:00,857 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:00,857 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:01,781 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.533120393753052 s; generated tokens: 512 tokens; generate speed: 53.70749333403027 tokens/s 2025-05-21 04:34:01,781 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.533787727355957 s; generated tokens: 512 tokens; generate speed: 53.70373398716262 tokens/s 2025-05-21 04:34:01,781 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.533947944641113 s; generated tokens: 512 tokens; generate speed: 53.70283149991263 tokens/s 2025-05-21 04:34:01,781 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.534406661987305 s; generated tokens: 512 tokens; generate speed: 53.70024776070137 tokens/s 2025-05-21 04:34:01,782 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015194416046142578 s; prefill predict time: 0.013758659362792969 s; prefill post time: 0.014662504196166992 s; decode prepare time: 0.000997491078834011 s; decode predict time: 0.004722483485352759 s; decode post time: 0.012829913085220844 s 2025-05-21 04:34:01,782 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014569759368896484 s; prefill predict time: 0.01320338249206543 s; prefill post time: 0.014855623245239258 s; decode prepare time: 0.0010533029552541834 s; decode predict time: 0.004395206769307454 s; decode post time: 0.01310051117391269 s 2025-05-21 04:34:01,782 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0017075538635253906 s; prefill predict time: 0.013622045516967773 s; prefill post time: 0.014913797378540039 s; decode prepare time: 0.0010851959892914952 s; decode predict time: 0.0042880997938268325 s; decode post time: 0.013174943494703438 s 2025-05-21 04:34:01,782 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015780925750732422 s; prefill predict time: 0.015284061431884766 s; prefill post time: 0.015030384063720703 s; decode prepare time: 0.0010655252900842118 s; decode predict time: 0.004374247438767377 s; decode post time: 0.01310987481865631 s 2025-05-21 04:34:01,783 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:01,783 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:01,783 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:01,783 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:01,783 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:01,783 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:01,783 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:01,783 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.541621923446655 2025-05-21 04:34:01,783 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.540628910064697 2025-05-21 04:34:01,783 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.541500568389893 2025-05-21 04:34:01,783 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:01,784 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.541663646697998 2025-05-21 04:34:01,784 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:01,784 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:01,785 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:01,785 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:01,785 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:01,785 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:01,785 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:01,785 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:01,785 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:01,785 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:01,785 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:01,786 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:01,786 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:01,786 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:01,786 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:01,786 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:01,788 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:01,788 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:01,788 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:01,788 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:01,788 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:01,788 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:01,789 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:01,789 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:01,789 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:01,789 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:01,789 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:01,789 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:01,789 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:01,789 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:01,790 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:01,790 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:10,237 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.3799147605896 s; generated tokens: 512 tokens; generate speed: 54.584717779228185 tokens/s 2025-05-21 04:34:10,237 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.380158424377441 s; generated tokens: 512 tokens; generate speed: 54.58329985871015 tokens/s 2025-05-21 04:34:10,237 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013666152954101562 s; prefill predict time: 0.006293296813964844 s; prefill post time: 0.01350092887878418 s; decode prepare time: 0.0009809678781289178 s; decode predict time: 0.004747168690550561 s; decode post time: 0.012542766600905565 s 2025-05-21 04:34:10,237 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.379828691482544 s; generated tokens: 512 tokens; generate speed: 54.58521864742873 tokens/s 2025-05-21 04:34:10,238 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.380398511886597 s; generated tokens: 512 tokens; generate speed: 54.581902821208175 tokens/s 2025-05-21 04:34:10,238 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014057159423828125 s; prefill predict time: 0.006319999694824219 s; prefill post time: 0.013597249984741211 s; decode prepare time: 0.0009852416827953957 s; decode predict time: 0.004772025463627834 s; decode post time: 0.012508577563291427 s 2025-05-21 04:34:10,238 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013854503631591797 s; prefill predict time: 0.005637407302856445 s; prefill post time: 0.014730453491210938 s; decode prepare time: 0.0010106124989907102 s; decode predict time: 0.0041972721324247475 s; decode post time: 0.013061020472278334 s 2025-05-21 04:34:10,238 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014111995697021484 s; prefill predict time: 0.005655050277709961 s; prefill post time: 0.014037132263183594 s; decode prepare time: 0.001066562480889188 s; decode predict time: 0.004430495523938946 s; decode post time: 0.012773810534094644 s 2025-05-21 04:34:10,238 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:10,239 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:10,239 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:10,239 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.386913061141968 2025-05-21 04:34:10,239 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:10,239 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:10,239 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:10,239 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:10,239 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.387457132339478 2025-05-21 04:34:10,239 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:10,239 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.386528968811035 2025-05-21 04:34:10,240 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.387332916259766 2025-05-21 04:34:10,240 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:10,240 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:10,240 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:10,240 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:10,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:10,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:10,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:10,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:10,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:10,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:10,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:10,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:10,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:10,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:10,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:10,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:10,243 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:10,244 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:10,244 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:10,244 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:10,244 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:10,244 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:10,244 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:10,244 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:10,244 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:10,245 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:10,245 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:10,245 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:10,245 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:10,245 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:10,245 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:10,246 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:11,238 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.448511362075806 s; generated tokens: 512 tokens; generate speed: 54.18843036534332 tokens/s 2025-05-21 04:34:11,239 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.44806170463562 s; generated tokens: 512 tokens; generate speed: 54.19100933144743 tokens/s 2025-05-21 04:34:11,239 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.448308229446411 s; generated tokens: 512 tokens; generate speed: 54.18959538219878 tokens/s 2025-05-21 04:34:11,239 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.448936462402344 s; generated tokens: 512 tokens; generate speed: 54.18599246986857 tokens/s 2025-05-21 04:34:11,239 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001435995101928711 s; prefill predict time: 0.006298542022705078 s; prefill post time: 0.014411211013793945 s; decode prepare time: 0.0009959518559524924 s; decode predict time: 0.004612886671926461 s; decode post time: 0.012791538425154415 s 2025-05-21 04:34:11,239 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014510154724121094 s; prefill predict time: 0.005417585372924805 s; prefill post time: 0.014890909194946289 s; decode prepare time: 0.0010608264378138951 s; decode predict time: 0.004305872730180329 s; decode post time: 0.013035833952244948 s 2025-05-21 04:34:11,239 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014128684997558594 s; prefill predict time: 0.006038188934326172 s; prefill post time: 0.014010429382324219 s; decode prepare time: 0.0010698742129331465 s; decode predict time: 0.004230970494887408 s; decode post time: 0.013100284419647635 s 2025-05-21 04:34:11,239 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001417398452758789 s; prefill predict time: 0.006227731704711914 s; prefill post time: 0.013837337493896484 s; decode prepare time: 0.0010315816696376016 s; decode predict time: 0.004362980992186303 s; decode post time: 0.013006057757920715 s 2025-05-21 04:34:11,240 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:11,240 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:11,240 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:11,240 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.45539665222168 2025-05-21 04:34:11,240 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:11,240 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:11,240 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:11,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:11,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:11,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.455207824707031 2025-05-21 04:34:11,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.455525636672974 2025-05-21 04:34:11,241 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.455862998962402 2025-05-21 04:34:11,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:11,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:11,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:11,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:11,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:11,242 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:11,243 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:11,243 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:11,243 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:11,243 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:11,243 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:11,243 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:11,243 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:11,243 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:11,243 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:11,243 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:11,245 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:11,245 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:11,245 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:11,245 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:11,246 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:11,246 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:11,246 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:11,246 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:11,246 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:11,246 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:11,246 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:11,246 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:11,247 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:11,247 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:11,247 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:11,247 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:19,654 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.409075021743774 s; generated tokens: 512 tokens; generate speed: 54.41555081841738 tokens/s 2025-05-21 04:34:19,654 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.40831470489502 s; generated tokens: 512 tokens; generate speed: 54.4199483180142 tokens/s 2025-05-21 04:34:19,654 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.408003807067871 s; generated tokens: 512 tokens; generate speed: 54.42174668502516 tokens/s 2025-05-21 04:34:19,655 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014190673828125 s; prefill predict time: 0.0066182613372802734 s; prefill post time: 0.014047384262084961 s; decode prepare time: 0.001065118438810285 s; decode predict time: 0.004519356933294558 s; decode post time: 0.012740780929296683 s 2025-05-21 04:34:19,655 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013835430145263672 s; prefill predict time: 0.00710296630859375 s; prefill post time: 0.013700723648071289 s; decode prepare time: 0.0009829326151868367 s; decode predict time: 0.004760437853196088 s; decode post time: 0.01258316338645502 s 2025-05-21 04:34:19,655 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013213157653808594 s; prefill predict time: 0.0061032772064208984 s; prefill post time: 0.013909339904785156 s; decode prepare time: 0.0009937076419300297 s; decode predict time: 0.004778485204659256 s; decode post time: 0.012548939355898743 s 2025-05-21 04:34:19,655 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.409213066101074 s; generated tokens: 512 tokens; generate speed: 54.414752477505445 tokens/s 2025-05-21 04:34:19,656 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:19,656 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001481771469116211 s; prefill predict time: 0.006463050842285156 s; prefill post time: 0.01437067985534668 s; decode prepare time: 0.0010169135614383944 s; decode predict time: 0.004179314538544299 s; decode post time: 0.013129023889739452 s 2025-05-21 04:34:19,656 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:19,656 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:19,656 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:19,656 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:19,656 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.414719343185425 2025-05-21 04:34:19,656 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:19,656 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.415769815444946 2025-05-21 04:34:19,656 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.415302991867065 2025-05-21 04:34:19,657 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:19,657 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:19,657 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:19,657 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.416213750839233 2025-05-21 04:34:19,657 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:19,657 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:19,658 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:19,658 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:19,658 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:19,658 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:19,658 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:19,658 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:19,658 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:19,658 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:19,658 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:19,658 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:19,659 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:19,659 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:19,659 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:19,660 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:19,660 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:19,661 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:19,661 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:19,661 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:19,661 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:19,661 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:19,661 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:19,662 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:19,662 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:19,662 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:19,662 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:19,662 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:19,663 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:19,663 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:19,664 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:20,672 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.425321578979492 s; generated tokens: 512 tokens; generate speed: 54.32175398045525 tokens/s 2025-05-21 04:34:20,673 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.425178050994873 s; generated tokens: 512 tokens; generate speed: 54.3225812000396 tokens/s 2025-05-21 04:34:20,673 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.425180673599243 s; generated tokens: 512 tokens; generate speed: 54.32256608450561 tokens/s 2025-05-21 04:34:20,673 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.425407409667969 s; generated tokens: 512 tokens; generate speed: 54.32125930968498 tokens/s 2025-05-21 04:34:20,673 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015347003936767578 s; prefill predict time: 0.009766817092895508 s; prefill post time: 0.01462864875793457 s; decode prepare time: 0.0010014672102060336 s; decode predict time: 0.004532333916308833 s; decode post time: 0.012810350164275347 s 2025-05-21 04:34:20,673 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0016911029815673828 s; prefill predict time: 0.00959157943725586 s; prefill post time: 0.015646696090698242 s; decode prepare time: 0.0010771308159874843 s; decode predict time: 0.004178913901833927 s; decode post time: 0.013088677960599239 s 2025-05-21 04:34:20,674 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015933513641357422 s; prefill predict time: 0.009305715560913086 s; prefill post time: 0.014732837677001953 s; decode prepare time: 0.001067854886885492 s; decode predict time: 0.004211832027809293 s; decode post time: 0.013065907819863635 s 2025-05-21 04:34:20,673 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0016655921936035156 s; prefill predict time: 0.009116649627685547 s; prefill post time: 0.014255046844482422 s; decode prepare time: 0.0010408524664414138 s; decode predict time: 0.004273560467888327 s; decode post time: 0.013028632404752907 s 2025-05-21 04:34:20,674 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:20,674 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:20,674 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:20,674 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:20,674 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:20,675 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.432288646697998 2025-05-21 04:34:20,675 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:20,675 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:20,675 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:20,675 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.432234764099121 2025-05-21 04:34:20,675 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.432226181030273 2025-05-21 04:34:20,675 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.43235182762146 2025-05-21 04:34:20,676 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:20,676 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:20,676 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:20,676 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:20,676 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:20,676 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:20,676 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:20,676 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:20,677 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:20,677 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:20,677 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:20,677 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:20,677 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:20,677 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:20,677 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:20,677 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:20,679 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:20,679 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:20,679 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:20,679 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:20,679 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:20,680 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:20,680 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:20,680 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:20,680 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:20,680 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:20,680 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:20,680 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:20,680 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:20,681 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:20,681 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:20,681 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:29,047 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.384768009185791 s; generated tokens: 512 tokens; generate speed: 54.55648978204421 tokens/s 2025-05-21 04:34:29,047 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.384619235992432 s; generated tokens: 512 tokens; generate speed: 54.55735465924373 tokens/s 2025-05-21 04:34:29,047 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.38541579246521 s; generated tokens: 512 tokens; generate speed: 54.55272428218293 tokens/s 2025-05-21 04:34:29,048 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001298666000366211 s; prefill predict time: 0.0074405670166015625 s; prefill post time: 0.013681650161743164 s; decode prepare time: 0.000983178032354366 s; decode predict time: 0.00474385420481364 s; decode post time: 0.012552078922667381 s 2025-05-21 04:34:29,048 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.383579969406128 s; generated tokens: 512 tokens; generate speed: 54.56339709037548 tokens/s 2025-05-21 04:34:29,048 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.00141143798828125 s; prefill predict time: 0.006867647171020508 s; prefill post time: 0.013959169387817383 s; decode prepare time: 0.0009912180107400376 s; decode predict time: 0.004752911305895039 s; decode post time: 0.012531238992620122 s 2025-05-21 04:34:29,048 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013988018035888672 s; prefill predict time: 0.007607698440551758 s; prefill post time: 0.014000892639160156 s; decode prepare time: 0.00106392447962453 s; decode predict time: 0.004454940440608006 s; decode post time: 0.012759563740918781 s 2025-05-21 04:34:29,048 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001483917236328125 s; prefill predict time: 0.005585670471191406 s; prefill post time: 0.014072656631469727 s; decode prepare time: 0.0010094320704092485 s; decode predict time: 0.004175119306526932 s; decode post time: 0.01309262609761755 s 2025-05-21 04:34:29,049 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:29,049 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:29,049 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.391272783279419 2025-05-21 04:34:29,049 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:29,049 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:29,049 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:29,049 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:29,050 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:29,050 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.391603469848633 2025-05-21 04:34:29,050 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:29,050 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.39220380783081 2025-05-21 04:34:29,050 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.390931844711304 2025-05-21 04:34:29,050 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:29,050 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:29,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:29,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:29,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:29,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:29,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:29,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:29,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:29,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:29,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:29,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:29,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:29,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:29,052 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:29,052 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:29,053 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:29,054 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:29,054 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:29,054 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:29,054 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:29,054 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:29,054 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:29,055 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:29,055 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:29,055 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:29,055 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:29,055 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:29,055 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:29,055 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:29,056 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:29,056 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:30,136 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.455106496810913 s; generated tokens: 512 tokens; generate speed: 54.15063280066608 tokens/s 2025-05-21 04:34:30,136 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.455064058303833 s; generated tokens: 512 tokens; generate speed: 54.15087585264324 tokens/s 2025-05-21 04:34:30,137 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.45517897605896 s; generated tokens: 512 tokens; generate speed: 54.15021770570526 tokens/s 2025-05-21 04:34:30,137 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.456058502197266 s; generated tokens: 512 tokens; generate speed: 54.14518109009464 tokens/s 2025-05-21 04:34:30,137 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014111995697021484 s; prefill predict time: 0.006181001663208008 s; prefill post time: 0.01461935043334961 s; decode prepare time: 0.0010789975728074165 s; decode predict time: 0.004208536241568771 s; decode post time: 0.013117267194326144 s 2025-05-21 04:34:30,137 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013916492462158203 s; prefill predict time: 0.005969524383544922 s; prefill post time: 0.014173746109008789 s; decode prepare time: 0.0010346535833847732 s; decode predict time: 0.004333588656257181 s; decode post time: 0.013036655586526351 s 2025-05-21 04:34:30,137 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001425027847290039 s; prefill predict time: 0.005986690521240234 s; prefill post time: 0.014577865600585938 s; decode prepare time: 0.001075564065324817 s; decode predict time: 0.004264582372179218 s; decode post time: 0.013067513296048935 s 2025-05-21 04:34:30,138 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014209747314453125 s; prefill predict time: 0.006387948989868164 s; prefill post time: 0.013662338256835938 s; decode prepare time: 0.0010081713913471498 s; decode predict time: 0.004578139735203163 s; decode post time: 0.012819535578300342 s 2025-05-21 04:34:30,138 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:30,138 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:30,138 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:30,138 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:30,138 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:30,138 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:30,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.462120294570923 2025-05-21 04:34:30,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:30,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:30,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.462200164794922 2025-05-21 04:34:30,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.462369918823242 2025-05-21 04:34:30,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.462855100631714 2025-05-21 04:34:30,140 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:30,140 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:30,140 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:30,140 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:30,140 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:30,140 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:30,140 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:30,140 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:30,140 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:30,141 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:30,141 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:30,141 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:30,141 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:30,141 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:30,141 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:30,141 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:30,143 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:30,143 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:30,143 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:30,143 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:30,144 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:30,144 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:30,144 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:30,144 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:30,144 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:30,144 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:30,144 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:30,144 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:30,145 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:30,145 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:30,145 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:30,145 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:38,447 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.391204357147217 s; generated tokens: 512 tokens; generate speed: 54.51909899184977 tokens/s 2025-05-21 04:34:38,447 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.392032861709595 s; generated tokens: 512 tokens; generate speed: 54.514289668573696 tokens/s 2025-05-21 04:34:38,447 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.391118049621582 s; generated tokens: 512 tokens; generate speed: 54.51960004066089 tokens/s 2025-05-21 04:34:38,448 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014448165893554688 s; prefill predict time: 0.0060198307037353516 s; prefill post time: 0.013243675231933594 s; decode prepare time: 0.0009881255678001449 s; decode predict time: 0.004836731798508588 s; decode post time: 0.012465117728873475 s 2025-05-21 04:34:38,448 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013403892517089844 s; prefill predict time: 0.006792306900024414 s; prefill post time: 0.013872861862182617 s; decode prepare time: 0.0009795676938708282 s; decode predict time: 0.004743037036820954 s; decode post time: 0.01257056032841453 s 2025-05-21 04:34:38,448 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015444755554199219 s; prefill predict time: 0.00543665885925293 s; prefill post time: 0.014129638671875 s; decode prepare time: 0.0010633650591228806 s; decode predict time: 0.0044796518251007675 s; decode post time: 0.012749782513732313 s 2025-05-21 04:34:38,448 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.392353534698486 s; generated tokens: 512 tokens; generate speed: 54.51242844602274 tokens/s 2025-05-21 04:34:38,449 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:38,449 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:38,449 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:38,449 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013954639434814453 s; prefill predict time: 0.005860567092895508 s; prefill post time: 0.014362573623657227 s; decode prepare time: 0.001014264829004814 s; decode predict time: 0.0042295390484379785 s; decode post time: 0.01304995663711936 s 2025-05-21 04:34:38,449 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:38,449 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:38,449 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:38,449 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.398757219314575 2025-05-21 04:34:38,449 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.398021936416626 2025-05-21 04:34:38,449 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.398265600204468 2025-05-21 04:34:38,450 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:38,450 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:38,450 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.399015426635742 2025-05-21 04:34:38,450 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:38,450 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:38,450 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:38,451 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:38,451 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:38,451 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:38,451 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:38,451 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:38,451 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:38,451 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:38,451 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:38,451 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:38,451 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:38,452 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:38,452 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:38,452 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:38,453 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:38,454 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:38,454 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:38,454 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:38,454 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:38,454 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:38,454 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:38,455 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:38,455 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:38,455 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:38,455 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:38,455 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:38,455 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:38,456 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:38,456 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:38,456 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:39,615 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.469429016113281 s; generated tokens: 512 tokens; generate speed: 54.068729923290554 tokens/s 2025-05-21 04:34:39,615 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.46989393234253 s; generated tokens: 512 tokens; generate speed: 54.06607546588947 tokens/s 2025-05-21 04:34:39,615 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.469711542129517 s; generated tokens: 512 tokens; generate speed: 54.06711679887803 tokens/s 2025-05-21 04:34:39,615 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.469710350036621 s; generated tokens: 512 tokens; generate speed: 54.067123605107945 tokens/s 2025-05-21 04:34:39,615 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014545917510986328 s; prefill predict time: 0.006327152252197266 s; prefill post time: 0.014129161834716797 s; decode prepare time: 0.001042250784409256 s; decode predict time: 0.004342412948608399 s; decode post time: 0.013055954893974408 s 2025-05-21 04:34:39,615 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013794898986816406 s; prefill predict time: 0.006134033203125 s; prefill post time: 0.014021158218383789 s; decode prepare time: 0.0010730319760316972 s; decode predict time: 0.004264983008889591 s; decode post time: 0.013103925085347693 s 2025-05-21 04:34:39,616 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014483928680419922 s; prefill predict time: 0.006261587142944336 s; prefill post time: 0.014238595962524414 s; decode prepare time: 0.0010763101148512032 s; decode predict time: 0.004303617570914475 s; decode post time: 0.013063966878006369 s 2025-05-21 04:34:39,616 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001375436782836914 s; prefill predict time: 0.0062601566314697266 s; prefill post time: 0.014022111892700195 s; decode prepare time: 0.0009955090785913037 s; decode predict time: 0.004631640396866144 s; decode post time: 0.012815612169627807 s 2025-05-21 04:34:39,616 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:39,616 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:39,617 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:39,617 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:39,617 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:39,617 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:39,617 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:39,617 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.476507425308228 2025-05-21 04:34:39,617 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.476699590682983 2025-05-21 04:34:39,617 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:39,617 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.476679801940918 2025-05-21 04:34:39,617 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.476736545562744 2025-05-21 04:34:39,618 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:39,618 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:39,618 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:39,618 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:39,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:39,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:39,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:39,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:39,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:39,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:39,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:39,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:39,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:39,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:39,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:39,619 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:39,621 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:39,621 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:39,622 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:39,622 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:39,622 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:39,622 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:39,622 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:39,622 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:39,622 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:39,622 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:39,623 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:39,623 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:39,623 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:39,623 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:39,623 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:39,624 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:47,893 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.436924695968628 s; generated tokens: 512 tokens; generate speed: 54.2549629773693 tokens/s 2025-05-21 04:34:47,893 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.43729281425476 s; generated tokens: 512 tokens; generate speed: 54.25284666664562 tokens/s 2025-05-21 04:34:47,893 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.437825202941895 s; generated tokens: 512 tokens; generate speed: 54.249786258003894 tokens/s 2025-05-21 04:34:47,894 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001386404037475586 s; prefill predict time: 0.007907390594482422 s; prefill post time: 0.013926029205322266 s; decode prepare time: 0.0009874597687543954 s; decode predict time: 0.0047727650287104585 s; decode post time: 0.012618659760387443 s 2025-05-21 04:34:47,894 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015249252319335938 s; prefill predict time: 0.0073130130767822266 s; prefill post time: 0.013533830642700195 s; decode prepare time: 0.000987410778635402 s; decode predict time: 0.004831679194581275 s; decode post time: 0.012555705591190585 s 2025-05-21 04:34:47,894 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.436998844146729 s; generated tokens: 512 tokens; generate speed: 54.25453668647703 tokens/s 2025-05-21 04:34:47,894 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015270709991455078 s; prefill predict time: 0.007991313934326172 s; prefill post time: 0.014085054397583008 s; decode prepare time: 0.00107345049161967 s; decode predict time: 0.0045013633428835405 s; decode post time: 0.01280457492910486 s 2025-05-21 04:34:47,894 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013751983642578125 s; prefill predict time: 0.0071680545806884766 s; prefill post time: 0.014228582382202148 s; decode prepare time: 0.0010098300568045003 s; decode predict time: 0.004246244243547029 s; decode post time: 0.013121609585159212 s 2025-05-21 04:34:47,895 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:47,895 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:47,895 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:47,895 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:47,895 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:47,895 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:47,895 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.44437050819397 2025-05-21 04:34:47,895 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.444239616394043 2025-05-21 04:34:47,895 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.444440841674805 2025-05-21 04:34:47,895 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:47,896 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:47,896 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.443902015686035 2025-05-21 04:34:47,896 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:47,896 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:47,896 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:47,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:47,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:47,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:47,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:47,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:47,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:47,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:47,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:47,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:47,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:47,898 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:47,898 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:47,898 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:47,899 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:47,899 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:47,900 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:47,900 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:47,900 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:47,900 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:47,900 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:47,900 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:47,900 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:47,901 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:47,901 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:47,901 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:47,901 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:47,901 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:47,901 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:47,902 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:49,163 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.539734363555908 s; generated tokens: 512 tokens; generate speed: 53.670257523727685 tokens/s 2025-05-21 04:34:49,163 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.539757490158081 s; generated tokens: 512 tokens; generate speed: 53.67012741447747 tokens/s 2025-05-21 04:34:49,163 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.539398670196533 s; generated tokens: 512 tokens; generate speed: 53.672146190893145 tokens/s 2025-05-21 04:34:49,164 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015420913696289062 s; prefill predict time: 0.006297111511230469 s; prefill post time: 0.014181852340698242 s; decode prepare time: 0.0010910066839766829 s; decode predict time: 0.0042938578362558405 s; decode post time: 0.013192018650749192 s 2025-05-21 04:34:49,164 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015873908996582031 s; prefill predict time: 0.006075620651245117 s; prefill post time: 0.014238834381103516 s; decode prepare time: 0.0010680489810712184 s; decode predict time: 0.004387039764254701 s; decode post time: 0.013121849403456 s 2025-05-21 04:34:49,164 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0016093254089355469 s; prefill predict time: 0.005437135696411133 s; prefill post time: 0.014200210571289062 s; decode prepare time: 0.0010642902725130144 s; decode predict time: 0.004454736616097244 s; decode post time: 0.013061770254385215 s 2025-05-21 04:34:49,164 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.540118932723999 s; generated tokens: 512 tokens; generate speed: 53.66809403641346 tokens/s 2025-05-21 04:34:49,165 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001544952392578125 s; prefill predict time: 0.005701780319213867 s; prefill post time: 0.013639450073242188 s; decode prepare time: 0.0010034268149657726 s; decode predict time: 0.004745247784782859 s; decode post time: 0.01283185915937629 s 2025-05-21 04:34:49,165 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:49,165 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:49,165 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:49,165 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:49,165 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:49,165 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:49,165 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.546550273895264 2025-05-21 04:34:49,165 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.546552658081055 2025-05-21 04:34:49,165 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.546858072280884 2025-05-21 04:34:49,166 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:49,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:49,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.547279357910156 2025-05-21 04:34:49,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:49,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:49,167 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:49,167 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:49,167 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:49,167 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:49,167 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:49,167 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:49,167 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:49,167 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:49,167 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:49,167 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:49,168 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:49,168 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:49,168 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:49,168 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:49,170 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:49,170 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:49,170 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:49,170 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:49,170 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:49,171 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:49,170 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:49,171 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:49,171 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:49,171 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:49,171 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:49,171 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:49,171 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:49,171 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:49,172 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:49,172 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:57,301 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.399046659469604 s; generated tokens: 512 tokens; generate speed: 54.47360977659968 tokens/s 2025-05-21 04:34:57,301 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.399444103240967 s; generated tokens: 512 tokens; generate speed: 54.47130642794719 tokens/s 2025-05-21 04:34:57,301 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.399517297744751 s; generated tokens: 512 tokens; generate speed: 54.47088225720329 tokens/s 2025-05-21 04:34:57,302 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013456344604492188 s; prefill predict time: 0.006463766098022461 s; prefill post time: 0.013413667678833008 s; decode prepare time: 0.00098878016910441 s; decode predict time: 0.004697483661128025 s; decode post time: 0.012622104932184089 s 2025-05-21 04:34:57,302 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014078617095947266 s; prefill predict time: 0.006041765213012695 s; prefill post time: 0.013423442840576172 s; decode prepare time: 0.0010028165380548825 s; decode predict time: 0.004734119714475146 s; decode post time: 0.012566920595859595 s 2025-05-21 04:34:57,302 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014190673828125 s; prefill predict time: 0.006680488586425781 s; prefill post time: 0.014024496078491211 s; decode prepare time: 0.0010718655446270908 s; decode predict time: 0.004437786457585353 s; decode post time: 0.012798272000366928 s 2025-05-21 04:34:57,302 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.39942717552185 s; generated tokens: 512 tokens; generate speed: 54.471404527007685 tokens/s 2025-05-21 04:34:57,302 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:57,303 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:57,303 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:57,303 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013515949249267578 s; prefill predict time: 0.005622148513793945 s; prefill post time: 0.014172554016113281 s; decode prepare time: 0.0010153262815830058 s; decode predict time: 0.0042310897041769584 s; decode post time: 0.013061804314182229 s 2025-05-21 04:34:57,303 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:57,303 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:57,303 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:57,303 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.406318664550781 2025-05-21 04:34:57,303 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.406257152557373 2025-05-21 04:34:57,303 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.406370878219604 2025-05-21 04:34:57,303 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:57,304 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:57,304 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.406309604644775 2025-05-21 04:34:57,304 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:57,304 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:57,304 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:57,304 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:57,304 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:57,305 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:57,305 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:57,305 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:57,305 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:57,305 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:57,305 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:57,305 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:57,305 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:57,306 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:57,306 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:57,306 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:57,307 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:57,307 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:57,307 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:57,308 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:57,308 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:57,308 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:57,308 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:57,308 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:57,308 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:57,308 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:57,309 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:57,309 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:57,309 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:57,309 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:57,309 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:57,310 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:58,728 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.556018829345703 s; generated tokens: 512 tokens; generate speed: 53.57879773402 tokens/s 2025-05-21 04:34:58,728 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.556791543960571 s; generated tokens: 512 tokens; generate speed: 53.57446561901407 tokens/s 2025-05-21 04:34:58,729 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.556665182113647 s; generated tokens: 512 tokens; generate speed: 53.57517400089149 tokens/s 2025-05-21 04:34:58,729 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.557188034057617 s; generated tokens: 512 tokens; generate speed: 53.572243025402145 tokens/s 2025-05-21 04:34:58,729 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014841556549072266 s; prefill predict time: 0.0072824954986572266 s; prefill post time: 0.013637304306030273 s; decode prepare time: 0.0010070642146578974 s; decode predict time: 0.004709624776653215 s; decode post time: 0.012890702824303315 s 2025-05-21 04:34:58,729 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015499591827392578 s; prefill predict time: 0.00777125358581543 s; prefill post time: 0.014487981796264648 s; decode prepare time: 0.001064923411479435 s; decode predict time: 0.004408445545271331 s; decode post time: 0.013134832718134393 s 2025-05-21 04:34:58,730 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0016465187072753906 s; prefill predict time: 0.008083581924438477 s; prefill post time: 0.014483451843261719 s; decode prepare time: 0.0010960326036128514 s; decode predict time: 0.00430895206975002 s; decode post time: 0.013202384959928211 s 2025-05-21 04:34:58,730 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001481771469116211 s; prefill predict time: 0.007399320602416992 s; prefill post time: 0.01425933837890625 s; decode prepare time: 0.0010609902047831019 s; decode predict time: 0.004372766906139898 s; decode post time: 0.013172711878140146 s 2025-05-21 04:34:58,730 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:58,730 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:58,730 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:58,730 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:58,730 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:58,731 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:34:58,731 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.562985181808472 2025-05-21 04:34:58,731 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.563761234283447 2025-05-21 04:34:58,731 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:58,731 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:34:58,731 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.564089059829712 2025-05-21 04:34:58,731 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.563924789428711 2025-05-21 04:34:58,732 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:58,732 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:58,732 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:58,732 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:34:58,732 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:58,732 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:58,732 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:58,732 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:58,732 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:58,733 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:34:58,733 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:58,733 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:58,733 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:58,733 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:34:58,733 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:58,733 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:34:58,735 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:58,735 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:58,735 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:58,735 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:34:58,736 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:58,736 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:58,736 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:58,736 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:58,736 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:58,736 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:34:58,736 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:58,736 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:34:58,737 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:58,737 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:58,737 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:34:58,737 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:06,696 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.38671088218689 s; generated tokens: 512 tokens; generate speed: 54.54519761246931 tokens/s 2025-05-21 04:35:06,697 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.386361598968506 s; generated tokens: 512 tokens; generate speed: 54.54722733633713 tokens/s 2025-05-21 04:35:06,697 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.387652397155762 s; generated tokens: 512 tokens; generate speed: 54.53972711592133 tokens/s 2025-05-21 04:35:06,697 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.387837171554565 s; generated tokens: 512 tokens; generate speed: 54.53865364765547 tokens/s 2025-05-21 04:35:06,697 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013611316680908203 s; prefill predict time: 0.006624698638916016 s; prefill post time: 0.013686895370483398 s; decode prepare time: 0.0009970002444989528 s; decode predict time: 0.0047301886128444295 s; decode post time: 0.012551446250273523 s 2025-05-21 04:35:06,697 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013446807861328125 s; prefill predict time: 0.00570225715637207 s; prefill post time: 0.014223575592041016 s; decode prepare time: 0.0010094204060952024 s; decode predict time: 0.0042210578918457035 s; decode post time: 0.013051693686767101 s 2025-05-21 04:35:06,697 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013957023620605469 s; prefill predict time: 0.006979227066040039 s; prefill post time: 0.013895988464355469 s; decode prepare time: 0.0010814610759339455 s; decode predict time: 0.0043169947231517115 s; decode post time: 0.012884048099853754 s 2025-05-21 04:35:06,697 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.00130462646484375 s; prefill predict time: 0.0069844722747802734 s; prefill post time: 0.01368570327758789 s; decode prepare time: 0.0009899899917572678 s; decode predict time: 0.004674802106969497 s; decode post time: 0.012619387613583917 s 2025-05-21 04:35:06,698 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:06,698 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:06,698 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:06,698 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.393860578536987 2025-05-21 04:35:06,698 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:06,698 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:06,699 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:06,699 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:06,699 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:06,699 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.393187761306763 2025-05-21 04:35:06,699 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.39442491531372 2025-05-21 04:35:06,699 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.394471883773804 2025-05-21 04:35:06,699 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:06,700 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:06,700 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:06,700 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:06,700 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:06,700 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:06,700 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:06,700 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:06,700 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:06,700 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:06,700 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:06,700 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:06,701 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:06,701 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:06,701 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:06,701 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:06,703 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:06,703 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:06,703 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:06,703 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:06,703 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:06,703 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:06,703 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:06,704 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:06,704 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:06,704 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:06,704 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:06,704 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:06,704 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:06,704 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:06,704 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:06,705 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:08,228 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.491325378417969 s; generated tokens: 512 tokens; generate speed: 53.94399407739418 tokens/s 2025-05-21 04:35:08,228 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.490964412689209 s; generated tokens: 512 tokens; generate speed: 53.94604570589974 tokens/s 2025-05-21 04:35:08,228 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.491331577301025 s; generated tokens: 512 tokens; generate speed: 53.94395884603511 tokens/s 2025-05-21 04:35:08,229 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014886856079101562 s; prefill predict time: 0.007683277130126953 s; prefill post time: 0.013761043548583984 s; decode prepare time: 0.0009976165868531468 s; decode predict time: 0.004657612127416274 s; decode post time: 0.01282537706677461 s 2025-05-21 04:35:08,229 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.491872310638428 s; generated tokens: 512 tokens; generate speed: 53.94088576456657 tokens/s 2025-05-21 04:35:08,229 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0016474723815917969 s; prefill predict time: 0.0072820186614990234 s; prefill post time: 0.014557361602783203 s; decode prepare time: 0.0010843888187595076 s; decode predict time: 0.00423130007351146 s; decode post time: 0.013163795209910772 s 2025-05-21 04:35:08,229 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015180110931396484 s; prefill predict time: 0.00732874870300293 s; prefill post time: 0.014637947082519531 s; decode prepare time: 0.0010422414529580194 s; decode predict time: 0.004341021706076229 s; decode post time: 0.013096365676700718 s 2025-05-21 04:35:08,230 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015361309051513672 s; prefill predict time: 0.007687807083129883 s; prefill post time: 0.014815568923950195 s; decode prepare time: 0.0010500593428042536 s; decode predict time: 0.004371320032605938 s; decode post time: 0.01306073866245098 s 2025-05-21 04:35:08,230 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:08,230 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:08,230 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:08,230 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:08,230 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:08,230 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.49817943572998 2025-05-21 04:35:08,230 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:08,230 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:08,231 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.498180866241455 2025-05-21 04:35:08,231 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.498043060302734 2025-05-21 04:35:08,231 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:08,231 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.498677730560303 2025-05-21 04:35:08,231 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:08,232 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:08,232 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:08,232 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:08,232 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:08,232 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:08,232 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:08,232 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:08,232 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:08,232 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:08,232 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:08,232 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:08,233 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:08,233 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:08,233 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:08,233 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:08,235 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:08,235 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:08,235 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:08,235 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:08,235 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:08,236 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:08,236 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:08,236 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:08,236 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:08,236 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:08,236 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:08,236 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:08,236 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:08,237 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:08,237 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:08,237 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:16,120 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.415193557739258 s; generated tokens: 512 tokens; generate speed: 54.3801884539206 tokens/s 2025-05-21 04:35:16,120 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.41549015045166 s; generated tokens: 512 tokens; generate speed: 54.378475450419266 tokens/s 2025-05-21 04:35:16,120 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.415021896362305 s; generated tokens: 512 tokens; generate speed: 54.38117995220193 tokens/s 2025-05-21 04:35:16,120 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.415747880935669 s; generated tokens: 512 tokens; generate speed: 54.37698698758289 tokens/s 2025-05-21 04:35:16,121 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013916492462158203 s; prefill predict time: 0.005718708038330078 s; prefill post time: 0.013655424118041992 s; decode prepare time: 0.0009766194218525448 s; decode predict time: 0.0047611526414459825 s; decode post time: 0.01259994133577888 s 2025-05-21 04:35:16,121 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014150142669677734 s; prefill predict time: 0.0052890777587890625 s; prefill post time: 0.014129638671875 s; decode prepare time: 0.0010701611550586797 s; decode predict time: 0.0044075619940664255 s; decode post time: 0.012863181108597906 s 2025-05-21 04:35:16,121 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013628005981445312 s; prefill predict time: 0.005821704864501953 s; prefill post time: 0.013458490371704102 s; decode prepare time: 0.0009735997642322995 s; decode predict time: 0.0046576439165601545 s; decode post time: 0.012709263486171655 s 2025-05-21 04:35:16,121 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013439655303955078 s; prefill predict time: 0.0057179927825927734 s; prefill post time: 0.014226198196411133 s; decode prepare time: 0.001001306242672198 s; decode predict time: 0.004233257443297143 s; decode post time: 0.013105077053002881 s 2025-05-21 04:35:16,122 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:16,122 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:16,122 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:16,122 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:16,122 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:16,122 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:16,122 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:16,122 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.421924591064453 2025-05-21 04:35:16,122 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.42238998413086 2025-05-21 04:35:16,122 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.422198057174683 2025-05-21 04:35:16,122 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:16,123 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.422563076019287 2025-05-21 04:35:16,123 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:16,123 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:16,123 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:16,124 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:16,124 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:16,124 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:16,124 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:16,124 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:16,124 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:16,124 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:16,124 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:16,124 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:16,124 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:16,124 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:16,124 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:16,125 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:16,126 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:16,127 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:16,127 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:16,127 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:16,127 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:16,127 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:16,127 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:16,127 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:16,127 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:16,128 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:16,128 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:16,128 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:16,128 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:16,128 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:16,128 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:16,129 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:17,666 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.429385423660278 s; generated tokens: 512 tokens; generate speed: 54.29834257440427 tokens/s 2025-05-21 04:35:17,666 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.42953896522522 s; generated tokens: 512 tokens; generate speed: 54.29745843229263 tokens/s 2025-05-21 04:35:17,666 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.429574728012085 s; generated tokens: 512 tokens; generate speed: 54.29725250270521 tokens/s 2025-05-21 04:35:17,667 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.429661750793457 s; generated tokens: 512 tokens; generate speed: 54.2967514139007 tokens/s 2025-05-21 04:35:17,667 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.00147247314453125 s; prefill predict time: 0.006526470184326172 s; prefill post time: 0.014495611190795898 s; decode prepare time: 0.000984692993462669 s; decode predict time: 0.004563964114469641 s; decode post time: 0.012812387220080351 s 2025-05-21 04:35:17,667 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014715194702148438 s; prefill predict time: 0.006293773651123047 s; prefill post time: 0.014628887176513672 s; decode prepare time: 0.0010220001355076022 s; decode predict time: 0.004366052384470023 s; decode post time: 0.012972932972320139 s 2025-05-21 04:35:17,667 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015277862548828125 s; prefill predict time: 0.006282806396484375 s; prefill post time: 0.014603376388549805 s; decode prepare time: 0.001046285237351509 s; decode predict time: 0.004254717920340744 s; decode post time: 0.013060841308414586 s 2025-05-21 04:35:17,667 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014064311981201172 s; prefill predict time: 0.006101131439208984 s; prefill post time: 0.014669179916381836 s; decode prepare time: 0.0010416815658838082 s; decode predict time: 0.004283866695329255 s; decode post time: 0.013038369307546 s 2025-05-21 04:35:17,668 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:17,668 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:17,668 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:17,668 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:17,668 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:17,668 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.436458587646484 2025-05-21 04:35:17,668 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:17,668 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.43628740310669 2025-05-21 04:35:17,668 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:17,669 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.436495780944824 2025-05-21 04:35:17,669 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:17,669 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.43650770187378 2025-05-21 04:35:17,670 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:17,670 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:17,670 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:17,670 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:17,670 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:17,670 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:17,670 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:17,670 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:17,670 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:17,671 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:17,671 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:17,671 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:17,671 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:17,671 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:17,671 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:17,671 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:17,673 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:17,673 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:17,673 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:17,673 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:17,674 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:17,674 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:17,674 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:17,674 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:17,674 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:17,674 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:17,674 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:17,674 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:17,675 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:17,675 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:17,675 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:17,675 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:25,514 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.385624408721924 s; generated tokens: 512 tokens; generate speed: 54.551511727254486 tokens/s 2025-05-21 04:35:25,515 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.38631796836853 s; generated tokens: 512 tokens; generate speed: 54.54748088924933 tokens/s 2025-05-21 04:35:25,515 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.386303901672363 s; generated tokens: 512 tokens; generate speed: 54.54756263631914 tokens/s 2025-05-21 04:35:25,515 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013818740844726562 s; prefill predict time: 0.0068395137786865234 s; prefill post time: 0.013475656509399414 s; decode prepare time: 0.0009797706529352296 s; decode predict time: 0.004752485892351936 s; decode post time: 0.012542991955452935 s 2025-05-21 04:35:25,515 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.387169122695923 s; generated tokens: 512 tokens; generate speed: 54.5425349546656 tokens/s 2025-05-21 04:35:25,516 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013511180877685547 s; prefill predict time: 0.007322549819946289 s; prefill post time: 0.013646125793457031 s; decode prepare time: 0.0009872731397296585 s; decode predict time: 0.00459138832840265 s; decode post time: 0.01269877633469912 s 2025-05-21 04:35:25,516 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013637542724609375 s; prefill predict time: 0.007360935211181641 s; prefill post time: 0.01416635513305664 s; decode prepare time: 0.0009991870700963089 s; decode predict time: 0.004175529292985505 s; decode post time: 0.013103791645595006 s 2025-05-21 04:35:25,516 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014977455139160156 s; prefill predict time: 0.0077173709869384766 s; prefill post time: 0.014151334762573242 s; decode prepare time: 0.0010585873560896126 s; decode predict time: 0.004348529086393468 s; decode post time: 0.012872859689820769 s 2025-05-21 04:35:25,516 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:25,516 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:25,516 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:25,517 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.392713069915771 2025-05-21 04:35:25,517 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:25,517 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:25,517 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:25,517 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:25,517 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:25,517 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.393079996109009 2025-05-21 04:35:25,517 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.392832040786743 2025-05-21 04:35:25,517 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.393456935882568 2025-05-21 04:35:25,518 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:25,518 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:25,518 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:25,518 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:25,518 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:25,518 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:25,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:25,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:25,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:25,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:25,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:25,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:25,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:25,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:25,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:25,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:25,521 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:25,521 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:25,521 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:25,521 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:25,522 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:25,522 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:25,522 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:25,522 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:25,522 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:25,522 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:25,522 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:25,522 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:25,523 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:25,523 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:25,523 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:25,523 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:27,092 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.416842937469482 s; generated tokens: 512 tokens; generate speed: 54.370663650209075 tokens/s 2025-05-21 04:35:27,092 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.417206525802612 s; generated tokens: 512 tokens; generate speed: 54.36856445668352 tokens/s 2025-05-21 04:35:27,092 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.41746211051941 s; generated tokens: 512 tokens; generate speed: 54.367088923892815 tokens/s 2025-05-21 04:35:27,093 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.417529106140137 s; generated tokens: 512 tokens; generate speed: 54.36670216035552 tokens/s 2025-05-21 04:35:27,093 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014891624450683594 s; prefill predict time: 0.005951642990112305 s; prefill post time: 0.014114141464233398 s; decode prepare time: 0.0009760954608655955 s; decode predict time: 0.004635992704653273 s; decode post time: 0.012727832607560429 s 2025-05-21 04:35:27,093 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014760494232177734 s; prefill predict time: 0.005547285079956055 s; prefill post time: 0.014341115951538086 s; decode prepare time: 0.0010271865561050444 s; decode predict time: 0.004251818563423905 s; decode post time: 0.013061310213839238 s 2025-05-21 04:35:27,093 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013821125030517578 s; prefill predict time: 0.006166219711303711 s; prefill post time: 0.01377725601196289 s; decode prepare time: 0.0010103708144036758 s; decode predict time: 0.004367716172162224 s; decode post time: 0.012961306917224144 s 2025-05-21 04:35:27,093 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013270378112792969 s; prefill predict time: 0.006074666976928711 s; prefill post time: 0.013923883438110352 s; decode prepare time: 0.0010472818363436048 s; decode predict time: 0.004274585200291054 s; decode post time: 0.01301713465711142 s 2025-05-21 04:35:27,094 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:27,094 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:27,094 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:27,094 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:27,094 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:27,094 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:27,095 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.423933267593384 2025-05-21 04:35:27,095 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:27,095 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.424392700195312 2025-05-21 04:35:27,095 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:27,095 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.424672842025757 2025-05-21 04:35:27,095 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.424556493759155 2025-05-21 04:35:27,096 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:27,096 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:27,096 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:27,096 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:27,096 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:27,096 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:27,096 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:27,096 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:27,096 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:27,096 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:27,096 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:27,096 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:27,097 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:27,097 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:27,097 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:27,097 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:27,099 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:27,099 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:27,099 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:27,099 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:27,099 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:27,099 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:27,100 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:27,100 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:27,100 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:27,100 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:27,100 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:27,100 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:27,100 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:27,100 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:27,101 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:27,101 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:34,956 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.432692527770996 s; generated tokens: 512 tokens; generate speed: 54.27930556334892 tokens/s 2025-05-21 04:35:34,956 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.43296766281128 s; generated tokens: 512 tokens; generate speed: 54.27772237771142 tokens/s 2025-05-21 04:35:34,956 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.4330735206604 s; generated tokens: 512 tokens; generate speed: 54.27711327369739 tokens/s 2025-05-21 04:35:34,956 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.43305778503418 s; generated tokens: 512 tokens; generate speed: 54.27720381532093 tokens/s 2025-05-21 04:35:34,956 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.00127410888671875 s; prefill predict time: 0.006471395492553711 s; prefill post time: 0.013881683349609375 s; decode prepare time: 0.0009707308095495295 s; decode predict time: 0.004677494834451114 s; decode post time: 0.01272378649030413 s 2025-05-21 04:35:34,957 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014235973358154297 s; prefill predict time: 0.0065822601318359375 s; prefill post time: 0.014036417007446289 s; decode prepare time: 0.001052426731983043 s; decode predict time: 0.004418717178643918 s; decode post time: 0.012900519977577045 s 2025-05-21 04:35:34,957 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013380050659179688 s; prefill predict time: 0.006379604339599609 s; prefill post time: 0.014353513717651367 s; decode prepare time: 0.0010035803173386187 s; decode predict time: 0.004226360601537368 s; decode post time: 0.013142488707301668 s 2025-05-21 04:35:34,957 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001394510269165039 s; prefill predict time: 0.006381988525390625 s; prefill post time: 0.013409614562988281 s; decode prepare time: 0.0009881479632831132 s; decode predict time: 0.004760305086771647 s; decode post time: 0.012621851116710447 s 2025-05-21 04:35:34,957 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:34,958 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:34,958 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:34,958 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:34,958 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:34,958 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.439414024353027 2025-05-21 04:35:34,958 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:34,958 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.439432144165039 2025-05-21 04:35:34,958 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:34,958 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:34,958 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.439701318740845 2025-05-21 04:35:34,958 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.440177202224731 2025-05-21 04:35:34,959 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:34,959 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:34,959 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:34,959 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:34,959 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:34,959 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:34,960 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:34,960 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:34,960 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:34,960 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:34,960 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:34,960 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:34,960 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:34,960 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:34,960 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:34,960 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:34,962 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:34,962 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:34,962 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:34,963 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:34,963 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:34,963 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:34,963 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:34,963 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:34,963 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:34,963 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:34,963 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:34,964 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:34,964 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:34,964 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:34,964 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:34,964 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:36,624 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.523642778396606 s; generated tokens: 512 tokens; generate speed: 53.760941260986684 tokens/s 2025-05-21 04:35:36,625 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.523999452590942 s; generated tokens: 512 tokens; generate speed: 53.7589279113948 tokens/s 2025-05-21 04:35:36,625 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.523742437362671 s; generated tokens: 512 tokens; generate speed: 53.76037869224273 tokens/s 2025-05-21 04:35:36,625 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.52351689338684 s; generated tokens: 512 tokens; generate speed: 53.761651890966284 tokens/s 2025-05-21 04:35:36,625 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001348733901977539 s; prefill predict time: 0.0086212158203125 s; prefill post time: 0.01364278793334961 s; decode prepare time: 0.0009780760613906174 s; decode predict time: 0.0047194925009035596 s; decode post time: 0.012845040300820905 s 2025-05-21 04:35:36,626 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013461112976074219 s; prefill predict time: 0.008723735809326172 s; prefill post time: 0.014371395111083984 s; decode prepare time: 0.0010408151406364665 s; decode predict time: 0.004327868012820973 s; decode post time: 0.013175323284768779 s 2025-05-21 04:35:36,626 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014853477478027344 s; prefill predict time: 0.008470296859741211 s; prefill post time: 0.014315605163574219 s; decode prepare time: 0.0010589587478488393 s; decode predict time: 0.0043034198237400425 s; decode post time: 0.013181591220564571 s 2025-05-21 04:35:36,626 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0012776851654052734 s; prefill predict time: 0.008313655853271484 s; prefill post time: 0.014312744140625 s; decode prepare time: 0.0010254019160559967 s; decode predict time: 0.004375862607768938 s; decode post time: 0.013140006074700103 s 2025-05-21 04:35:36,626 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:36,626 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:36,626 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:36,627 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.530583381652832 2025-05-21 04:35:36,627 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:36,627 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:36,627 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:36,627 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:36,627 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.530945301055908 2025-05-21 04:35:36,627 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.530746221542358 2025-05-21 04:35:36,627 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:36,627 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.53095030784607 2025-05-21 04:35:36,628 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:36,628 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:36,628 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:36,628 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:36,628 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:36,628 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:36,628 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:36,629 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:36,629 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:36,629 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:36,629 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:36,629 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:36,629 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:36,629 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:36,629 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:36,629 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:36,631 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:36,631 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:36,631 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:36,632 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:36,632 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:36,632 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:36,632 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:36,632 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:36,632 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:36,632 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:36,632 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:36,633 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:36,633 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:36,633 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:36,633 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:36,633 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:44,373 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.408331394195557 s; generated tokens: 512 tokens; generate speed: 54.41985178326913 tokens/s 2025-05-21 04:35:44,374 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.409648656845093 s; generated tokens: 512 tokens; generate speed: 54.41223351389886 tokens/s 2025-05-21 04:35:44,374 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.409724473953247 s; generated tokens: 512 tokens; generate speed: 54.41179509742826 tokens/s 2025-05-21 04:35:44,374 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0012824535369873047 s; prefill predict time: 0.00719141960144043 s; prefill post time: 0.01366281509399414 s; decode prepare time: 0.000987087443850045 s; decode predict time: 0.004785806057499904 s; decode post time: 0.012548815714169855 s 2025-05-21 04:35:44,374 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.40965223312378 s; generated tokens: 512 tokens; generate speed: 54.4122128337179 tokens/s 2025-05-21 04:35:44,374 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014281272888183594 s; prefill predict time: 0.0074880123138427734 s; prefill post time: 0.014358043670654297 s; decode prepare time: 0.0010233489967838892 s; decode predict time: 0.004434942264182895 s; decode post time: 0.01286458455886393 s 2025-05-21 04:35:44,374 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013654232025146484 s; prefill predict time: 0.0075566768646240234 s; prefill post time: 0.01431131362915039 s; decode prepare time: 0.0009590916904218043 s; decode predict time: 0.004711244153041466 s; decode post time: 0.012654905449854184 s 2025-05-21 04:35:44,375 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015380382537841797 s; prefill predict time: 0.006887197494506836 s; prefill post time: 0.014619588851928711 s; decode prepare time: 0.0010011923989671084 s; decode predict time: 0.004200852618497961 s; decode post time: 0.013122290781099503 s 2025-05-21 04:35:44,375 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:44,375 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:44,375 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:44,375 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:44,375 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.415178060531616 2025-05-21 04:35:44,375 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:44,375 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:44,375 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.415964603424072 2025-05-21 04:35:44,375 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:44,376 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.416098833084106 2025-05-21 04:35:44,376 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:44,376 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.41617751121521 2025-05-21 04:35:44,376 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:44,376 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:44,376 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:44,377 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:44,377 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:44,377 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:44,377 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:44,377 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:44,377 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:44,377 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:44,377 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:44,377 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:44,377 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:44,377 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:44,378 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:44,378 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:44,379 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:44,379 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:44,379 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:44,380 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:44,380 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:44,380 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:44,380 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:44,380 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:44,380 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:44,380 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:44,381 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:44,381 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:44,381 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:44,381 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:44,381 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:44,382 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:46,135 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.502237558364868 s; generated tokens: 512 tokens; generate speed: 53.882045871320464 tokens/s 2025-05-21 04:35:46,135 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.502087593078613 s; generated tokens: 512 tokens; generate speed: 53.88289625670725 tokens/s 2025-05-21 04:35:46,135 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.502007722854614 s; generated tokens: 512 tokens; generate speed: 53.883349175618626 tokens/s 2025-05-21 04:35:46,136 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.501923084259033 s; generated tokens: 512 tokens; generate speed: 53.883829142774644 tokens/s 2025-05-21 04:35:46,136 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001371145248413086 s; prefill predict time: 0.007673025131225586 s; prefill post time: 0.014156341552734375 s; decode prepare time: 0.0009820209324009964 s; decode predict time: 0.004648516224879845 s; decode post time: 0.012871934009858074 s 2025-05-21 04:35:46,136 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013377666473388672 s; prefill predict time: 0.007491588592529297 s; prefill post time: 0.01430201530456543 s; decode prepare time: 0.0010498699143441457 s; decode predict time: 0.004314393623202455 s; decode post time: 0.013138994545386029 s 2025-05-21 04:35:46,136 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014309883117675781 s; prefill predict time: 0.0072019100189208984 s; prefill post time: 0.014642477035522461 s; decode prepare time: 0.0010589321532128142 s; decode predict time: 0.004228848569533404 s; decode post time: 0.013213406336984056 s 2025-05-21 04:35:46,136 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0012955665588378906 s; prefill predict time: 0.007122516632080078 s; prefill post time: 0.014299392700195312 s; decode prepare time: 0.001038092223165553 s; decode predict time: 0.004308435028674556 s; decode post time: 0.013154857079343553 s 2025-05-21 04:35:46,137 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:46,137 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:46,137 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:46,137 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:46,137 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:46,137 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.508893013000488 2025-05-21 04:35:46,137 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:46,137 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.508909463882446 2025-05-21 04:35:46,137 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:46,138 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:46,138 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.508972644805908 2025-05-21 04:35:46,138 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.508995532989502 2025-05-21 04:35:46,138 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:46,138 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:46,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:46,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:46,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:46,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:46,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:46,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:46,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:46,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:46,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:46,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:46,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:46,139 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:46,140 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:46,140 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:46,141 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:46,141 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:46,142 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:46,142 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:46,142 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:46,142 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:46,142 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:46,142 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:46,142 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:46,143 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:46,143 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:46,143 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:46,143 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:46,143 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:46,143 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:46,144 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:53,860 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.478680849075317 s; generated tokens: 512 tokens; generate speed: 54.01595518958185 tokens/s 2025-05-21 04:35:53,860 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.478821039199829 s; generated tokens: 512 tokens; generate speed: 54.01515630294264 tokens/s 2025-05-21 04:35:53,860 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.479047536849976 s; generated tokens: 512 tokens; generate speed: 54.013865634663226 tokens/s 2025-05-21 04:35:53,861 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.478662967681885 s; generated tokens: 512 tokens; generate speed: 54.016057090087195 tokens/s 2025-05-21 04:35:53,861 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014414787292480469 s; prefill predict time: 0.005961179733276367 s; prefill post time: 0.013569116592407227 s; decode prepare time: 0.0010044677383512434 s; decode predict time: 0.004814480332767262 s; decode post time: 0.012642644389501523 s 2025-05-21 04:35:53,861 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014045238494873047 s; prefill predict time: 0.006147861480712891 s; prefill post time: 0.014104127883911133 s; decode prepare time: 0.0010407614847918545 s; decode predict time: 0.004452386556887159 s; decode post time: 0.012970011761510442 s 2025-05-21 04:35:53,861 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0012748241424560547 s; prefill predict time: 0.006162881851196289 s; prefill post time: 0.013425827026367188 s; decode prepare time: 0.0009692956323493017 s; decode predict time: 0.004743720503414379 s; decode post time: 0.012752371524877978 s 2025-05-21 04:35:53,861 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013420581817626953 s; prefill predict time: 0.005761623382568359 s; prefill post time: 0.014095544815063477 s; decode prepare time: 0.0010097764009598883 s; decode predict time: 0.004294284184773763 s; decode post time: 0.013158085294898942 s 2025-05-21 04:35:53,862 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:53,862 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:53,862 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:53,862 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:53,862 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:53,862 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:53,862 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:53,863 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.486028671264648 2025-05-21 04:35:53,863 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:53,863 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.485812902450562 2025-05-21 04:35:53,863 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.485878705978394 2025-05-21 04:35:53,863 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.485408782958984 2025-05-21 04:35:53,864 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:53,864 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:53,864 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:53,864 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:53,864 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:53,864 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:53,864 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:53,864 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:53,864 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:53,864 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:53,864 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:53,864 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:53,865 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:53,865 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:53,865 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:53,865 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:53,867 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:53,867 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:53,867 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:53,867 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:53,868 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:53,868 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:53,868 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:53,868 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:53,868 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:53,868 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:53,868 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:53,868 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:53,869 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:53,869 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:53,869 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:53,869 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:55,710 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.566816806793213 s; generated tokens: 512 tokens; generate speed: 53.51832384168145 tokens/s 2025-05-21 04:35:55,711 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013689994812011719 s; prefill predict time: 0.0064008235931396484 s; prefill post time: 0.014294147491455078 s; decode prepare time: 0.0009918040258777119 s; decode predict time: 0.004786318423701268 s; decode post time: 0.012854479530319542 s 2025-05-21 04:35:55,711 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.567726135253906 s; generated tokens: 512 tokens; generate speed: 53.513237394353226 tokens/s 2025-05-21 04:35:55,711 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.567026376724243 s; generated tokens: 512 tokens; generate speed: 53.51715149920065 tokens/s 2025-05-21 04:35:55,711 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.567404508590698 s; generated tokens: 512 tokens; generate speed: 53.515036344524624 tokens/s 2025-05-21 04:35:55,712 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001478433609008789 s; prefill predict time: 0.006312131881713867 s; prefill post time: 0.014410018920898438 s; decode prepare time: 0.001056368803558051 s; decode predict time: 0.004355160862791772 s; decode post time: 0.01322224462102304 s 2025-05-21 04:35:55,712 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014247894287109375 s; prefill predict time: 0.0055429935455322266 s; prefill post time: 0.014474153518676758 s; decode prepare time: 0.0010355755307669742 s; decode predict time: 0.004401318232218425 s; decode post time: 0.013193491153754367 s 2025-05-21 04:35:55,712 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015172958374023438 s; prefill predict time: 0.005982160568237305 s; prefill post time: 0.014390707015991211 s; decode prepare time: 0.0010810710212722449 s; decode predict time: 0.004339231229295918 s; decode post time: 0.013211120131431027 s 2025-05-21 04:35:55,712 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:55,712 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:55,713 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.573695659637451 2025-05-21 04:35:55,713 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:55,713 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:55,713 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:55,713 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:35:55,713 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:55,713 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.57410979270935 2025-05-21 04:35:55,713 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:35:55,713 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.574445247650146 2025-05-21 04:35:55,714 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.57492446899414 2025-05-21 04:35:55,714 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:55,714 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:55,714 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:55,714 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:55,715 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:55,715 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:55,715 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:35:55,715 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:55,715 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:55,715 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:55,715 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:35:55,715 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:55,715 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:55,715 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:35:55,715 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:55,716 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:35:55,717 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:55,718 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:55,718 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:55,718 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:55,718 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:55,718 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:35:55,718 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:55,719 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:55,719 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:55,719 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:55,719 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:35:55,719 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:55,719 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:35:55,719 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:55,720 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:35:55,720 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:03,280 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.411451578140259 s; generated tokens: 512 tokens; generate speed: 54.40180993856564 tokens/s 2025-05-21 04:36:03,281 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.41202712059021 s; generated tokens: 512 tokens; generate speed: 54.39848328527696 tokens/s 2025-05-21 04:36:03,281 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0012972354888916016 s; prefill predict time: 0.00531315803527832 s; prefill post time: 0.013965606689453125 s; decode prepare time: 0.0009777881861199604 s; decode predict time: 0.0046834992427451936 s; decode post time: 0.01267273038800682 s 2025-05-21 04:36:03,281 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.41199278831482 s; generated tokens: 512 tokens; generate speed: 54.398681715487335 tokens/s 2025-05-21 04:36:03,282 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014450550079345703 s; prefill predict time: 0.005448341369628906 s; prefill post time: 0.01419210433959961 s; decode prepare time: 0.0010547871225734047 s; decode predict time: 0.0043875722324146944 s; decode post time: 0.012891677494385004 s 2025-05-21 04:36:03,282 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.412836074829102 s; generated tokens: 512 tokens; generate speed: 54.39380819231953 tokens/s 2025-05-21 04:36:03,282 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0012950897216796875 s; prefill predict time: 0.0052602291107177734 s; prefill post time: 0.013732194900512695 s; decode prepare time: 0.0009859070152685833 s; decode predict time: 0.004764407288794424 s; decode post time: 0.01257941307620293 s 2025-05-21 04:36:03,282 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:03,282 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:03,282 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014171600341796875 s; prefill predict time: 0.005432605743408203 s; prefill post time: 0.01456141471862793 s; decode prepare time: 0.0010037202891071716 s; decode predict time: 0.004222948878419165 s; decode post time: 0.01310660881072341 s 2025-05-21 04:36:03,283 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.418463230133057 2025-05-21 04:36:03,283 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:03,283 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:03,283 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:03,283 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.418967723846436 2025-05-21 04:36:03,283 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:03,283 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:03,284 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.419343709945679 2025-05-21 04:36:03,284 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:03,284 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:03,284 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.419616460800171 2025-05-21 04:36:03,284 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:03,284 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:03,284 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:03,284 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:03,285 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:03,285 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:03,285 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:03,285 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:03,285 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:03,285 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:03,286 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:03,286 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:03,286 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:03,286 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:03,286 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:03,287 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:03,287 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:03,287 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:03,288 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:03,288 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:03,288 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:03,288 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:03,288 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:03,288 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:03,289 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:03,289 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:03,289 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:03,289 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:03,289 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:03,290 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:03,290 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:05,296 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.576508283615112 s; generated tokens: 512 tokens; generate speed: 53.46416301607594 tokens/s 2025-05-21 04:36:05,296 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.576034545898438 s; generated tokens: 512 tokens; generate speed: 53.46680795124089 tokens/s 2025-05-21 04:36:05,296 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.575716495513916 s; generated tokens: 512 tokens; generate speed: 53.468583811964834 tokens/s 2025-05-21 04:36:05,296 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014755725860595703 s; prefill predict time: 0.007064104080200195 s; prefill post time: 0.014339208602905273 s; decode prepare time: 0.0009880177895383593 s; decode predict time: 0.0047577582153619504 s; decode post time: 0.012903648813176762 s 2025-05-21 04:36:05,296 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.576515913009644 s; generated tokens: 512 tokens; generate speed: 53.46412042238147 tokens/s 2025-05-21 04:36:05,297 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014483928680419922 s; prefill predict time: 0.005860090255737305 s; prefill post time: 0.014532327651977539 s; decode prepare time: 0.0010791566740510049 s; decode predict time: 0.004329982925863827 s; decode post time: 0.013238460350409879 s 2025-05-21 04:36:05,297 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015783309936523438 s; prefill predict time: 0.006136417388916016 s; prefill post time: 0.014654874801635742 s; decode prepare time: 0.0010632591471513423 s; decode predict time: 0.0043796754350849225 s; decode post time: 0.013202345767833015 s 2025-05-21 04:36:05,297 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014450550079345703 s; prefill predict time: 0.006501436233520508 s; prefill post time: 0.014731407165527344 s; decode prepare time: 0.0010531895781216555 s; decode predict time: 0.004391876856486002 s; decode post time: 0.01320633729609958 s 2025-05-21 04:36:05,298 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:05,298 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:05,298 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:05,298 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.583849430084229 2025-05-21 04:36:05,298 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:05,298 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:05,298 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:05,298 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:05,298 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.583346366882324 2025-05-21 04:36:05,299 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:05,299 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.583701133728027 2025-05-21 04:36:05,299 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.58359980583191 2025-05-21 04:36:05,299 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:05,299 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:05,300 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:05,300 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:05,300 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:05,300 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:05,300 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:05,300 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:05,300 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:05,300 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:05,300 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:05,300 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:05,300 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:05,301 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:05,301 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:05,301 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:05,302 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:05,303 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:05,303 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:05,303 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:05,303 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:05,303 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:05,304 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:05,304 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:05,304 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:05,304 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:05,304 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:05,304 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:05,304 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:05,304 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:05,305 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:05,305 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:12,740 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.449534893035889 s; generated tokens: 512 tokens; generate speed: 54.182560919197556 tokens/s 2025-05-21 04:36:12,740 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.451011896133423 s; generated tokens: 512 tokens; generate speed: 54.17409327454855 tokens/s 2025-05-21 04:36:12,740 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014166831970214844 s; prefill predict time: 0.0059893131256103516 s; prefill post time: 0.013877630233764648 s; decode prepare time: 0.0010067740065244312 s; decode predict time: 0.004776248744889802 s; decode post time: 0.012619854652718322 s 2025-05-21 04:36:12,740 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.45176887512207 s; generated tokens: 512 tokens; generate speed: 54.16975454696436 tokens/s 2025-05-21 04:36:12,741 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013527870178222656 s; prefill predict time: 0.007030487060546875 s; prefill post time: 0.014017820358276367 s; decode prepare time: 0.0010567373958819066 s; decode predict time: 0.004452862459070542 s; decode post time: 0.01289640620729928 s 2025-05-21 04:36:12,741 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.450205326080322 s; generated tokens: 512 tokens; generate speed: 54.178717004910105 tokens/s 2025-05-21 04:36:12,741 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001329183578491211 s; prefill predict time: 0.007569551467895508 s; prefill post time: 0.013759136199951172 s; decode prepare time: 0.0009841181760664789 s; decode predict time: 0.004712538158192354 s; decode post time: 0.01271176664796594 s 2025-05-21 04:36:12,741 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:12,742 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001332998275756836 s; prefill predict time: 0.005749225616455078 s; prefill post time: 0.014084815979003906 s; decode prepare time: 0.0010161791762260542 s; decode predict time: 0.004241463249804927 s; decode post time: 0.01314826254275447 s 2025-05-21 04:36:12,742 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:12,742 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:12,742 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.456856489181519 2025-05-21 04:36:12,742 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:12,742 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:12,742 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:12,742 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.457797527313232 2025-05-21 04:36:12,742 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:12,742 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.458423376083374 2025-05-21 04:36:12,743 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:12,743 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.457289695739746 2025-05-21 04:36:12,743 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:12,743 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:12,743 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:12,744 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:12,744 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:12,744 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:12,744 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:12,744 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:12,744 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:12,744 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:12,744 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:12,744 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:12,744 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:12,744 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:12,745 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:12,745 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:12,746 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:12,746 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:12,746 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:12,747 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:12,747 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:12,747 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:12,747 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:12,747 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:12,747 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:12,747 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:12,748 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:12,748 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:12,748 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:12,748 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:12,748 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:12,749 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:15,026 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.721162796020508 s; generated tokens: 512 tokens; generate speed: 52.66859641622238 tokens/s 2025-05-21 04:36:15,026 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.720875024795532 s; generated tokens: 512 tokens; generate speed: 52.670155587230106 tokens/s 2025-05-21 04:36:15,026 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.720828294754028 s; generated tokens: 512 tokens; generate speed: 52.670408783612345 tokens/s 2025-05-21 04:36:15,026 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001378774642944336 s; prefill predict time: 0.006665706634521484 s; prefill post time: 0.01455235481262207 s; decode prepare time: 0.0009979707154275854 s; decode predict time: 0.004920423264596976 s; decode post time: 0.013015138659691857 s 2025-05-21 04:36:15,026 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.721758604049683 s; generated tokens: 512 tokens; generate speed: 52.6653685668272 tokens/s 2025-05-21 04:36:15,027 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013070106506347656 s; prefill predict time: 0.00600886344909668 s; prefill post time: 0.014732599258422852 s; decode prepare time: 0.0010501423927202617 s; decode predict time: 0.004584374614790374 s; decode post time: 0.01329660928879232 s 2025-05-21 04:36:15,027 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001407623291015625 s; prefill predict time: 0.005453348159790039 s; prefill post time: 0.014864921569824219 s; decode prepare time: 0.0010602110286048248 s; decode predict time: 0.004642757714963427 s; decode post time: 0.013230698449271066 s 2025-05-21 04:36:15,027 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014791488647460938 s; prefill predict time: 0.006485939025878906 s; prefill post time: 0.014752626419067383 s; decode prepare time: 0.001086871451361072 s; decode predict time: 0.004491026261273553 s; decode post time: 0.01335451700916029 s 2025-05-21 04:36:15,027 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:15,028 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:15,028 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.728192806243896 2025-05-21 04:36:15,028 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:15,028 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:15,028 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:15,028 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:15,028 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:15,028 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:15,028 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.728041887283325 2025-05-21 04:36:15,028 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.728478193283081 2025-05-21 04:36:15,028 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.728201627731323 2025-05-21 04:36:15,029 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:15,030 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:15,030 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:15,030 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:15,030 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:15,030 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:15,030 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:15,030 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:15,030 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:15,030 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:15,030 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:15,030 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:15,030 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:15,030 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:15,031 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:15,031 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:15,032 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:15,033 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:15,033 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:15,033 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:15,033 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:15,033 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:15,033 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:15,033 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:15,033 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:15,034 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:15,034 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:15,034 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:15,034 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:15,034 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:15,034 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:15,034 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:22,330 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.581442594528198 s; generated tokens: 512 tokens; generate speed: 53.436629708807594 tokens/s 2025-05-21 04:36:22,330 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.581740379333496 s; generated tokens: 512 tokens; generate speed: 53.43496898583414 tokens/s 2025-05-21 04:36:22,330 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.581928491592407 s; generated tokens: 512 tokens; generate speed: 53.43391995142216 tokens/s 2025-05-21 04:36:22,331 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014121532440185547 s; prefill predict time: 0.006720542907714844 s; prefill post time: 0.01461482048034668 s; decode prepare time: 0.0009923830424269585 s; decode predict time: 0.004951573820675121 s; decode post time: 0.012718684052767819 s 2025-05-21 04:36:22,331 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.581557035446167 s; generated tokens: 512 tokens; generate speed: 53.435991468390675 tokens/s 2025-05-21 04:36:22,331 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015456676483154297 s; prefill predict time: 0.0068242549896240234 s; prefill post time: 0.014065742492675781 s; decode prepare time: 0.0010196570081020288 s; decode predict time: 0.005035015648486568 s; decode post time: 0.012605584530914366 s 2025-05-21 04:36:22,331 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015103816986083984 s; prefill predict time: 0.006721019744873047 s; prefill post time: 0.014899015426635742 s; decode prepare time: 0.0010690395146200102 s; decode predict time: 0.004698393391627891 s; decode post time: 0.012895452532982873 s 2025-05-21 04:36:22,331 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013341903686523438 s; prefill predict time: 0.006000041961669922 s; prefill post time: 0.014839649200439453 s; decode prepare time: 0.0010316689187066662 s; decode predict time: 0.004469827577179553 s; decode post time: 0.013159949252283504 s 2025-05-21 04:36:22,331 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:22,332 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:22,332 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:22,332 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:22,332 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.58788275718689 2025-05-21 04:36:22,332 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:22,332 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:22,332 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:22,332 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.58837604522705 2025-05-21 04:36:22,332 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.588668584823608 2025-05-21 04:36:22,332 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:22,333 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.588227033615112 2025-05-21 04:36:22,333 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:22,333 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:22,333 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:22,333 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:22,334 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:22,334 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:22,334 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:22,334 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:22,334 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:22,334 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:22,334 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:22,334 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:22,334 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:22,334 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:22,334 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:22,335 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:22,336 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:22,336 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:22,336 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:22,336 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:22,337 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:22,337 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:22,337 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:22,337 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:22,337 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:22,338 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:22,338 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:22,338 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:22,338 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:22,338 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:22,339 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:22,339 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:24,555 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.52063250541687 s; generated tokens: 512 tokens; generate speed: 53.777939617844915 tokens/s 2025-05-21 04:36:24,555 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.520631074905396 s; generated tokens: 512 tokens; generate speed: 53.7779476981874 tokens/s 2025-05-21 04:36:24,555 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.520707368850708 s; generated tokens: 512 tokens; generate speed: 53.77751674997717 tokens/s 2025-05-21 04:36:24,556 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.00135040283203125 s; prefill predict time: 0.008185625076293945 s; prefill post time: 0.014052152633666992 s; decode prepare time: 0.0009870650483670767 s; decode predict time: 0.004784256804223154 s; decode post time: 0.012766451285076702 s 2025-05-21 04:36:24,556 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.520931482315063 s; generated tokens: 512 tokens; generate speed: 53.77625087955202 tokens/s 2025-05-21 04:36:24,556 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013589859008789062 s; prefill predict time: 0.007926225662231445 s; prefill post time: 0.014609813690185547 s; decode prepare time: 0.0010790801561508627 s; decode predict time: 0.004347552037706562 s; decode post time: 0.01311028540251073 s 2025-05-21 04:36:24,556 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013644695281982422 s; prefill predict time: 0.0077593326568603516 s; prefill post time: 0.014217138290405273 s; decode prepare time: 0.0010250412554656922 s; decode predict time: 0.004321377417620491 s; decode post time: 0.013191804493943306 s 2025-05-21 04:36:24,556 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013453960418701172 s; prefill predict time: 0.007771015167236328 s; prefill post time: 0.014781475067138672 s; decode prepare time: 0.0010469878956296439 s; decode predict time: 0.0043757041295369465 s; decode post time: 0.013116703574671438 s 2025-05-21 04:36:24,556 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:24,557 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:24,557 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:24,557 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.5273916721344 2025-05-21 04:36:24,557 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:24,557 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:24,557 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:24,557 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:24,558 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.527443170547485 2025-05-21 04:36:24,558 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:24,558 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.52752947807312 2025-05-21 04:36:24,558 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.527695178985596 2025-05-21 04:36:24,558 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:24,559 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:24,559 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:24,559 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:24,559 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:24,559 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:24,559 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:24,559 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:24,559 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:24,559 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:24,559 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:24,559 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:24,559 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:24,559 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:24,560 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:24,560 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:24,561 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:24,562 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:24,562 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:24,562 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:24,562 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:24,562 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:24,562 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:24,563 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:24,563 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:24,563 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:24,563 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:24,563 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:24,563 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:24,563 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:24,564 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:24,564 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:31,895 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.555757522583008 s; generated tokens: 512 tokens; generate speed: 53.580262871886035 tokens/s 2025-05-21 04:36:31,895 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.556896686553955 s; generated tokens: 512 tokens; generate speed: 53.57387620610744 tokens/s 2025-05-21 04:36:31,895 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.55649995803833 s; generated tokens: 512 tokens; generate speed: 53.57610027187177 tokens/s 2025-05-21 04:36:31,895 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.556286096572876 s; generated tokens: 512 tokens; generate speed: 53.577299258926125 tokens/s 2025-05-21 04:36:31,895 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001416921615600586 s; prefill predict time: 0.007719993591308594 s; prefill post time: 0.014408349990844727 s; decode prepare time: 0.0010612682820299601 s; decode predict time: 0.004711152058021695 s; decode post time: 0.01283764279285289 s 2025-05-21 04:36:31,895 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013659000396728516 s; prefill predict time: 0.00819706916809082 s; prefill post time: 0.01397848129272461 s; decode prepare time: 0.0009888520212789338 s; decode predict time: 0.004921860788382736 s; decode post time: 0.012699608233576885 s 2025-05-21 04:36:31,895 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001439809799194336 s; prefill predict time: 0.0071086883544921875 s; prefill post time: 0.014073371887207031 s; decode prepare time: 0.0010215736881860782 s; decode predict time: 0.004929801997016458 s; decode post time: 0.012656678892161748 s 2025-05-21 04:36:31,896 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013580322265625 s; prefill predict time: 0.007501840591430664 s; prefill post time: 0.014816522598266602 s; decode prepare time: 0.0010269458046631337 s; decode predict time: 0.004438626532461129 s; decode post time: 0.01314338312690272 s 2025-05-21 04:36:31,896 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:31,896 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:31,896 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:31,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:31,897 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:31,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:31,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.562966585159302 2025-05-21 04:36:31,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:31,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.563579797744751 2025-05-21 04:36:31,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:31,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.56308388710022 2025-05-21 04:36:31,897 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.562897682189941 2025-05-21 04:36:31,898 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:31,898 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:31,898 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:31,898 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:31,898 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:31,898 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:31,898 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:31,899 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:31,899 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:31,899 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:31,899 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:31,899 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:31,899 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:31,899 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:31,899 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:31,899 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:31,901 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:31,901 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:31,901 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:31,901 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:31,902 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:31,902 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:31,902 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:31,902 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:31,902 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:31,902 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:31,902 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:31,903 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:31,903 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:31,903 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:31,903 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:31,904 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:34,055 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.491512298583984 s; generated tokens: 512 tokens; generate speed: 53.94293173664054 tokens/s 2025-05-21 04:36:34,055 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.491376876831055 s; generated tokens: 512 tokens; generate speed: 53.94370138750034 tokens/s 2025-05-21 04:36:34,055 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0012977123260498047 s; prefill predict time: 0.006619453430175781 s; prefill post time: 0.013973474502563477 s; decode prepare time: 0.0009857945712811794 s; decode predict time: 0.0048139707714903585 s; decode post time: 0.012684723635708982 s 2025-05-21 04:36:34,056 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.49193787574768 s; generated tokens: 512 tokens; generate speed: 53.94051317046465 tokens/s 2025-05-21 04:36:34,056 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.491905927658081 s; generated tokens: 512 tokens; generate speed: 53.940694724765855 tokens/s 2025-05-21 04:36:34,056 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001447439193725586 s; prefill predict time: 0.0058519840240478516 s; prefill post time: 0.014179229736328125 s; decode prepare time: 0.0010193178098495692 s; decode predict time: 0.0043791518491857195 s; decode post time: 0.013086519129355594 s 2025-05-21 04:36:34,056 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0016202926635742188 s; prefill predict time: 0.005974531173706055 s; prefill post time: 0.01422882080078125 s; decode prepare time: 0.0010622070260243874 s; decode predict time: 0.004329342935599533 s; decode post time: 0.013093097335905012 s 2025-05-21 04:36:34,056 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:34,056 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015151500701904297 s; prefill predict time: 0.0058557987213134766 s; prefill post time: 0.014667272567749023 s; decode prepare time: 0.0010456819590290466 s; decode predict time: 0.00434722105662028 s; decode post time: 0.013094873111075385 s 2025-05-21 04:36:34,057 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:34,057 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.498190879821777 2025-05-21 04:36:34,057 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:34,057 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:34,057 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:34,058 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:34,058 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:34,058 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:34,058 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.498542785644531 2025-05-21 04:36:34,058 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.498844146728516 2025-05-21 04:36:34,058 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.498669385910034 2025-05-21 04:36:34,058 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:34,058 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:34,059 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:34,059 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:34,059 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:34,059 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:34,059 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:34,059 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:34,059 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:34,059 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:34,060 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:34,060 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:34,060 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:34,060 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:34,060 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:34,060 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:34,061 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:34,062 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:34,062 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:34,062 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:34,062 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:34,062 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:34,063 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:34,063 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:34,063 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:34,063 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:34,063 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:34,063 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:34,063 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:34,064 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:34,064 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:34,064 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:41,433 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.528872728347778 s; generated tokens: 512 tokens; generate speed: 53.731434409532326 tokens/s 2025-05-21 04:36:41,433 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.52971601486206 s; generated tokens: 512 tokens; generate speed: 53.72667970393985 tokens/s 2025-05-21 04:36:41,433 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.530757188796997 s; generated tokens: 512 tokens; generate speed: 53.720810409673895 tokens/s 2025-05-21 04:36:41,434 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015385150909423828 s; prefill predict time: 0.0053195953369140625 s; prefill post time: 0.013359546661376953 s; decode prepare time: 0.0010181989688462707 s; decode predict time: 0.004945588111877441 s; decode post time: 0.01259666692953987 s 2025-05-21 04:36:41,434 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.530434608459473 s; generated tokens: 512 tokens; generate speed: 53.72262871889755 tokens/s 2025-05-21 04:36:41,434 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0012788772583007812 s; prefill predict time: 0.006440162658691406 s; prefill post time: 0.013916015625 s; decode prepare time: 0.0010047873405561055 s; decode predict time: 0.004862703061571308 s; decode post time: 0.012696407079230082 s 2025-05-21 04:36:41,434 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013880729675292969 s; prefill predict time: 0.006890058517456055 s; prefill post time: 0.01412653923034668 s; decode prepare time: 0.0010673052644076413 s; decode predict time: 0.004585845330182243 s; decode post time: 0.012908773646196041 s 2025-05-21 04:36:41,434 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015392303466796875 s; prefill predict time: 0.006061553955078125 s; prefill post time: 0.014326333999633789 s; decode prepare time: 0.0010258521585781747 s; decode predict time: 0.0044154966578764075 s; decode post time: 0.013121101021066804 s 2025-05-21 04:36:41,435 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:41,435 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:41,435 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:41,435 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:41,435 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:41,435 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.536664247512817 2025-05-21 04:36:41,435 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:41,435 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.536470890045166 2025-05-21 04:36:41,435 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:41,435 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.537099123001099 2025-05-21 04:36:41,435 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:41,436 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.536964178085327 2025-05-21 04:36:41,436 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:41,436 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:41,436 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:41,437 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:41,437 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:41,437 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:41,437 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:41,437 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:41,437 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:41,437 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:41,437 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:41,437 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:41,437 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:41,437 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:41,437 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:41,438 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:41,439 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:41,439 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:41,439 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:41,440 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:41,440 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:41,440 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:41,440 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:41,440 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:41,440 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:41,440 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:41,440 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:41,441 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:41,441 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:41,441 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:41,441 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:41,442 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:43,520 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.457431077957153 s; generated tokens: 512 tokens; generate speed: 54.137322892401585 tokens/s 2025-05-21 04:36:43,521 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.456310510635376 s; generated tokens: 512 tokens; generate speed: 54.14373813382725 tokens/s 2025-05-21 04:36:43,521 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.456318378448486 s; generated tokens: 512 tokens; generate speed: 54.143693085342655 tokens/s 2025-05-21 04:36:43,521 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013000965118408203 s; prefill predict time: 0.007312774658203125 s; prefill post time: 0.014285087585449219 s; decode prepare time: 0.000988489494398382 s; decode predict time: 0.004678781359803443 s; decode post time: 0.012748153242346358 s 2025-05-21 04:36:43,522 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0016062259674072266 s; prefill predict time: 0.005811452865600586 s; prefill post time: 0.014412164688110352 s; decode prepare time: 0.0010734388273056239 s; decode predict time: 0.004331457848642387 s; decode post time: 0.013010319431700584 s 2025-05-21 04:36:43,522 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0017998218536376953 s; prefill predict time: 0.005519866943359375 s; prefill post time: 0.014799118041992188 s; decode prepare time: 0.001016748861324064 s; decode predict time: 0.0043386683744542735 s; decode post time: 0.013059036139172817 s 2025-05-21 04:36:43,522 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:43,522 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:43,523 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:43,523 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:43,523 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:43,523 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:43,523 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.463502168655396 2025-05-21 04:36:43,523 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.463640928268433 2025-05-21 04:36:43,523 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.464683294296265 2025-05-21 04:36:43,523 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.459245204925537 s; generated tokens: 512 tokens; generate speed: 54.12694024819187 tokens/s 2025-05-21 04:36:43,524 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:43,524 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:43,524 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015027523040771484 s; prefill predict time: 0.006159782409667969 s; prefill post time: 0.014957189559936523 s; decode prepare time: 0.001040688699472207 s; decode predict time: 0.004328350927315506 s; decode post time: 0.01305268095430796 s 2025-05-21 04:36:43,524 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:43,524 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:43,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:43,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:43,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:43,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:43,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:43,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:43,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:43,525 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:43,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:43,525 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:43,526 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.46632695198059 2025-05-21 04:36:43,527 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:43,527 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:43,527 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:43,527 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:43,528 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:43,528 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:43,528 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:43,528 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:43,528 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:43,528 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:43,528 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:43,529 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:43,529 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:43,529 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:43,529 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:43,530 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:43,530 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:43,531 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:43,531 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:43,532 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:50,953 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.511897325515747 s; generated tokens: 512 tokens; generate speed: 53.827326187232444 tokens/s 2025-05-21 04:36:50,954 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.5124990940094 s; generated tokens: 512 tokens; generate speed: 53.82392102643538 tokens/s 2025-05-21 04:36:50,954 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.512624025344849 s; generated tokens: 512 tokens; generate speed: 53.82321414531456 tokens/s 2025-05-21 04:36:50,954 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001421213150024414 s; prefill predict time: 0.0055506229400634766 s; prefill post time: 0.013693571090698242 s; decode prepare time: 0.001020423586821136 s; decode predict time: 0.004903214585547354 s; decode post time: 0.012604381706849936 s 2025-05-21 04:36:50,954 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.512212753295898 s; generated tokens: 512 tokens; generate speed: 53.82554125722183 tokens/s 2025-05-21 04:36:50,954 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0012927055358886719 s; prefill predict time: 0.005854129791259766 s; prefill post time: 0.01388859748840332 s; decode prepare time: 0.000998493276696849 s; decode predict time: 0.004858555045782351 s; decode post time: 0.012674419847253251 s 2025-05-21 04:36:50,954 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014133453369140625 s; prefill predict time: 0.0058367252349853516 s; prefill post time: 0.01405477523803711 s; decode prepare time: 0.0010722747287638266 s; decode predict time: 0.004556484315909591 s; decode post time: 0.012899410468024983 s 2025-05-21 04:36:50,955 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013358592987060547 s; prefill predict time: 0.0054378509521484375 s; prefill post time: 0.014069318771362305 s; decode prepare time: 0.0010232570819892063 s; decode predict time: 0.0043890532325295845 s; decode post time: 0.01311694992498409 s 2025-05-21 04:36:50,955 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:50,955 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:50,955 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:50,955 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:50,955 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.51880145072937 2025-05-21 04:36:50,956 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:50,956 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:50,956 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.5191068649292 2025-05-21 04:36:50,956 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.519012451171875 2025-05-21 04:36:50,956 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:50,956 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:50,956 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.51930046081543 2025-05-21 04:36:50,957 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:50,957 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:50,957 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:50,957 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:50,957 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:50,957 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:50,957 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:50,957 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:50,958 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:50,958 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:50,958 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:50,958 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:50,958 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:50,958 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:50,958 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:50,959 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:50,960 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:50,960 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:50,960 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:50,960 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:50,960 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:50,961 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:50,961 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:50,961 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:50,961 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:50,961 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:50,961 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:50,962 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:50,962 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:50,962 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:50,962 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:50,963 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:53,020 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.490181684494019 s; generated tokens: 512 tokens; generate speed: 53.95049505074865 tokens/s 2025-05-21 04:36:53,020 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.491020679473877 s; generated tokens: 512 tokens; generate speed: 53.945725890925154 tokens/s 2025-05-21 04:36:53,020 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.488126754760742 s; generated tokens: 512 tokens; generate speed: 53.9621795991606 tokens/s 2025-05-21 04:36:53,021 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.491162300109863 s; generated tokens: 512 tokens; generate speed: 53.94492094967899 tokens/s 2025-05-21 04:36:53,021 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014145374298095703 s; prefill predict time: 0.007861137390136719 s; prefill post time: 0.013968706130981445 s; decode prepare time: 0.0009903021288011405 s; decode predict time: 0.004755174412446863 s; decode post time: 0.012732919648202197 s 2025-05-21 04:36:53,021 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014772415161132812 s; prefill predict time: 0.00862884521484375 s; prefill post time: 0.014511346817016602 s; decode prepare time: 0.0010742707261833892 s; decode predict time: 0.004306933926601036 s; decode post time: 0.013095795525030146 s 2025-05-21 04:36:53,021 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0017478466033935547 s; prefill predict time: 0.005361795425415039 s; prefill post time: 0.014554500579833984 s; decode prepare time: 0.0010448425949902917 s; decode predict time: 0.004328552414389218 s; decode post time: 0.013105866960350081 s 2025-05-21 04:36:53,021 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014193058013916016 s; prefill predict time: 0.008623123168945312 s; prefill post time: 0.014520645141601562 s; decode prepare time: 0.001021594683951361 s; decode predict time: 0.004313159923927457 s; decode post time: 0.013142862431923703 s 2025-05-21 04:36:53,022 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:53,022 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:53,022 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:53,022 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:36:53,022 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:53,022 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:53,023 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:53,023 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:36:53,023 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.498000621795654 2025-05-21 04:36:53,023 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.49534010887146 2025-05-21 04:36:53,023 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.498270034790039 2025-05-21 04:36:53,023 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.498310327529907 2025-05-21 04:36:53,024 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:53,024 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:53,024 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:53,024 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:36:53,024 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:53,024 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:53,024 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:53,024 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:36:53,024 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:53,024 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:53,025 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:53,025 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:36:53,025 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:53,025 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:53,025 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:53,025 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:36:53,027 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:53,027 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:53,027 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:53,027 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:36:53,028 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:53,028 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:53,028 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:53,028 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:36:53,028 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:53,028 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:53,028 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:53,028 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:36:53,029 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:53,029 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:53,029 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:36:53,029 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:00,505 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.543591260910034 s; generated tokens: 512 tokens; generate speed: 53.64856750488892 tokens/s 2025-05-21 04:37:00,506 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.544125080108643 s; generated tokens: 512 tokens; generate speed: 53.64556684898054 tokens/s 2025-05-21 04:37:00,506 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013072490692138672 s; prefill predict time: 0.007148027420043945 s; prefill post time: 0.01351022720336914 s; decode prepare time: 0.0009990135051033036 s; decode predict time: 0.004900855176589068 s; decode post time: 0.012690130278555615 s 2025-05-21 04:37:00,506 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.544009923934937 s; generated tokens: 512 tokens; generate speed: 53.64621412599135 tokens/s 2025-05-21 04:37:00,506 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.543163776397705 s; generated tokens: 512 tokens; generate speed: 53.650970683987005 tokens/s 2025-05-21 04:37:00,507 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013990402221679688 s; prefill predict time: 0.007101774215698242 s; prefill post time: 0.013841867446899414 s; decode prepare time: 0.001076452419482565 s; decode predict time: 0.004584063735662722 s; decode post time: 0.012927724423940402 s 2025-05-21 04:37:00,507 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013964176177978516 s; prefill predict time: 0.0068395137786865234 s; prefill post time: 0.01398921012878418 s; decode prepare time: 0.0010351714789284185 s; decode predict time: 0.0049217097899493046 s; decode post time: 0.012630412723220259 s 2025-05-21 04:37:00,507 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:00,507 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014979839324951172 s; prefill predict time: 0.005496025085449219 s; prefill post time: 0.014204978942871094 s; decode prepare time: 0.001029014120829782 s; decode predict time: 0.004417110424415738 s; decode post time: 0.01314217237110474 s 2025-05-21 04:37:00,507 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:00,507 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.550044059753418 2025-05-21 04:37:00,507 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:00,508 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:00,508 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:00,508 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:00,508 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:00,508 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.55079960823059 2025-05-21 04:37:00,508 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:00,508 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.551024436950684 2025-05-21 04:37:00,508 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.550179719924927 2025-05-21 04:37:00,508 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:00,509 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:00,509 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:00,509 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:00,509 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:00,509 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:00,509 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:00,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:00,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:00,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:00,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:00,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:00,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:00,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:00,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:00,510 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:00,511 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:00,512 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:00,512 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:00,513 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:00,513 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:00,513 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:00,513 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:00,513 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:00,513 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:00,513 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:00,513 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:00,514 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:00,514 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:00,514 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:00,514 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:00,514 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:02,514 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.485361099243164 s; generated tokens: 512 tokens; generate speed: 53.97791340182636 tokens/s 2025-05-21 04:37:02,514 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.485548973083496 s; generated tokens: 512 tokens; generate speed: 53.976844297875424 tokens/s 2025-05-21 04:37:02,515 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.485521793365479 s; generated tokens: 512 tokens; generate speed: 53.97699896257806 tokens/s 2025-05-21 04:37:02,515 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013172626495361328 s; prefill predict time: 0.008403539657592773 s; prefill post time: 0.014284133911132812 s; decode prepare time: 0.0009904066410549933 s; decode predict time: 0.004728546797060499 s; decode post time: 0.012748352002257704 s 2025-05-21 04:37:02,515 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.485971927642822 s; generated tokens: 512 tokens; generate speed: 53.974437612238155 tokens/s 2025-05-21 04:37:02,515 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001331329345703125 s; prefill predict time: 0.008218050003051758 s; prefill post time: 0.014568328857421875 s; decode prepare time: 0.0010427294878577065 s; decode predict time: 0.004348321989470837 s; decode post time: 0.013078941524378707 s 2025-05-21 04:37:02,515 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013539791107177734 s; prefill predict time: 0.008194684982299805 s; prefill post time: 0.014175176620483398 s; decode prepare time: 0.0010227256558412675 s; decode predict time: 0.004326313617182713 s; decode post time: 0.013117779490999047 s 2025-05-21 04:37:02,516 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013885498046875 s; prefill predict time: 0.008399009704589844 s; prefill post time: 0.014000654220581055 s; decode prepare time: 0.0010733875043238213 s; decode predict time: 0.00435179168102788 s; decode post time: 0.013042072960541906 s 2025-05-21 04:37:02,516 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:02,516 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:02,516 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:02,516 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:02,516 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.492103576660156 2025-05-21 04:37:02,516 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:02,516 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:02,516 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:02,517 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.49232006072998 2025-05-21 04:37:02,517 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.492336988449097 2025-05-21 04:37:02,517 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:02,517 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.49267029762268 2025-05-21 04:37:02,518 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:02,518 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:02,518 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:02,518 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:02,518 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:02,518 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:02,518 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:02,518 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:02,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:02,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:02,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:02,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:02,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:02,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:02,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:02,519 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:02,521 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:02,521 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:02,521 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:02,521 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:02,521 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:02,521 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:02,522 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:02,522 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:02,522 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:02,522 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:02,522 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:02,522 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:02,522 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:02,522 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:02,523 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:02,523 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:10,130 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.615904808044434 s; generated tokens: 512 tokens; generate speed: 53.24511943708856 tokens/s 2025-05-21 04:37:10,130 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.615047693252563 s; generated tokens: 512 tokens; generate speed: 53.2498658700674 tokens/s 2025-05-21 04:37:10,130 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.615536451339722 s; generated tokens: 512 tokens; generate speed: 53.24715917733988 tokens/s 2025-05-21 04:37:10,130 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013926029205322266 s; prefill predict time: 0.0067920684814453125 s; prefill post time: 0.01378774642944336 s; decode prepare time: 0.0009892238796107224 s; decode predict time: 0.0050336267433914485 s; decode post time: 0.012708610284585076 s 2025-05-21 04:37:10,130 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.615773439407349 s; generated tokens: 512 tokens; generate speed: 53.24584686050551 tokens/s 2025-05-21 04:37:10,131 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014064311981201172 s; prefill predict time: 0.0061168670654296875 s; prefill post time: 0.013980865478515625 s; decode prepare time: 0.0010664873627067313 s; decode predict time: 0.004798635781980028 s; decode post time: 0.01286593202042253 s 2025-05-21 04:37:10,131 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001409769058227539 s; prefill predict time: 0.005761384963989258 s; prefill post time: 0.013786077499389648 s; decode prepare time: 0.0010308505504331943 s; decode predict time: 0.005008133252461752 s; decode post time: 0.012689196666859368 s 2025-05-21 04:37:10,131 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014185905456542969 s; prefill predict time: 0.005835056304931641 s; prefill post time: 0.014288663864135742 s; decode prepare time: 0.0010384024939191785 s; decode predict time: 0.0044678028892068304 s; decode post time: 0.013224071719175216 s 2025-05-21 04:37:10,131 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:10,131 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:10,131 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:10,132 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.622728824615479 2025-05-21 04:37:10,132 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:10,132 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:10,132 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.62238597869873 2025-05-21 04:37:10,132 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:10,132 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:10,132 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.622340679168701 2025-05-21 04:37:10,132 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:10,132 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.622550964355469 2025-05-21 04:37:10,133 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:10,133 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:10,133 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:10,133 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:10,133 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:10,133 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:10,134 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:10,134 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:10,134 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:10,134 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:10,134 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:10,134 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:10,134 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:10,134 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:10,134 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:10,134 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:10,136 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:10,136 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:10,136 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:10,137 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:10,137 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:10,137 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:10,137 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:10,137 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:10,137 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:10,137 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:10,138 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:10,138 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:10,138 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:10,138 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:10,138 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:10,139 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:12,048 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.525211811065674 s; generated tokens: 512 tokens; generate speed: 53.75208553422371 tokens/s 2025-05-21 04:37:12,048 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.525207757949829 s; generated tokens: 512 tokens; generate speed: 53.75210840652582 tokens/s 2025-05-21 04:37:12,048 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.525858402252197 s; generated tokens: 512 tokens; generate speed: 53.74843697854546 tokens/s 2025-05-21 04:37:12,049 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.525809049606323 s; generated tokens: 512 tokens; generate speed: 53.7487154459767 tokens/s 2025-05-21 04:37:12,049 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015821456909179688 s; prefill predict time: 0.006203651428222656 s; prefill post time: 0.014463186264038086 s; decode prepare time: 0.0010774583499258978 s; decode predict time: 0.004363094591626934 s; decode post time: 0.013108286605655797 s 2025-05-21 04:37:12,049 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015988349914550781 s; prefill predict time: 0.006192922592163086 s; prefill post time: 0.014323711395263672 s; decode prepare time: 0.0010369141274469006 s; decode predict time: 0.004377015899209415 s; decode post time: 0.013133456329076957 s 2025-05-21 04:37:12,049 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001424551010131836 s; prefill predict time: 0.006968975067138672 s; prefill post time: 0.015018224716186523 s; decode prepare time: 0.0010491327296964343 s; decode predict time: 0.004351011912027995 s; decode post time: 0.013150124633848783 s 2025-05-21 04:37:12,049 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0012662410736083984 s; prefill predict time: 0.006856679916381836 s; prefill post time: 0.014117956161499023 s; decode prepare time: 0.0009888940128094997 s; decode predict time: 0.0048387906130622415 s; decode post time: 0.012723230335810413 s 2025-05-21 04:37:12,050 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:12,050 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:12,050 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:12,050 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:12,050 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:12,050 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:12,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:12,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.53201174736023 2025-05-21 04:37:12,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.532610893249512 2025-05-21 04:37:12,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:12,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.53274130821228 2025-05-21 04:37:12,051 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.532405614852905 2025-05-21 04:37:12,052 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:12,052 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:12,052 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:12,052 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:12,052 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:12,052 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:12,052 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:12,052 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:12,052 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:12,053 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:12,053 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:12,053 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:12,053 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:12,053 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:12,053 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:12,053 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:12,055 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:12,055 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:12,055 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:12,055 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:12,055 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:12,056 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:12,056 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:12,056 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:12,056 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:12,056 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:12,056 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:12,056 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:12,057 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:12,057 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:12,057 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:12,057 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:19,748 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.610337972640991 s; generated tokens: 512 tokens; generate speed: 53.275961933657015 tokens/s 2025-05-21 04:37:19,748 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.609952926635742 s; generated tokens: 512 tokens; generate speed: 53.27809656391743 tokens/s 2025-05-21 04:37:19,748 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.609421968460083 s; generated tokens: 512 tokens; generate speed: 53.28104038728652 tokens/s 2025-05-21 04:37:19,748 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.609646558761597 s; generated tokens: 512 tokens; generate speed: 53.279795138061964 tokens/s 2025-05-21 04:37:19,749 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014102458953857422 s; prefill predict time: 0.007404804229736328 s; prefill post time: 0.014078378677368164 s; decode prepare time: 0.001077896461561468 s; decode predict time: 0.004689356392505123 s; decode post time: 0.012948374925527554 s 2025-05-21 04:37:19,749 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013873577117919922 s; prefill predict time: 0.007910728454589844 s; prefill post time: 0.013976573944091797 s; decode prepare time: 0.000998297782793437 s; decode predict time: 0.005043498675028483 s; decode post time: 0.01267622501649502 s 2025-05-21 04:37:19,749 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015032291412353516 s; prefill predict time: 0.006877899169921875 s; prefill post time: 0.014556407928466797 s; decode prepare time: 0.0010326347239096804 s; decode predict time: 0.00450439546622482 s; decode post time: 0.013177324881059082 s 2025-05-21 04:37:19,749 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001355886459350586 s; prefill predict time: 0.006489992141723633 s; prefill post time: 0.013541936874389648 s; decode prepare time: 0.0010318289530953783 s; decode predict time: 0.004997046788533529 s; decode post time: 0.012686907661870967 s 2025-05-21 04:37:19,750 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:19,750 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:19,750 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:19,750 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:19,750 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.616580486297607 2025-05-21 04:37:19,750 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:19,750 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:19,750 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.617144107818604 2025-05-21 04:37:19,750 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:19,750 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:19,751 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.616443872451782 2025-05-21 04:37:19,751 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.616629838943481 2025-05-21 04:37:19,751 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:19,751 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:19,752 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:19,752 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:19,752 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:19,752 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:19,752 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:19,752 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:19,752 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:19,752 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:19,752 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:19,752 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:19,752 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:19,752 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:19,753 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:19,753 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:19,754 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:19,754 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:19,755 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:19,755 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:19,755 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:19,755 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:19,755 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:19,755 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:19,756 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:19,756 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:19,756 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:19,756 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:19,756 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:19,756 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:19,757 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:19,757 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:21,585 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.5285005569458 s; generated tokens: 512 tokens; generate speed: 53.73353309265198 tokens/s 2025-05-21 04:37:21,586 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.528863191604614 s; generated tokens: 512 tokens; generate speed: 53.731488185400394 tokens/s 2025-05-21 04:37:21,586 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001325845718383789 s; prefill predict time: 0.006098031997680664 s; prefill post time: 0.014382123947143555 s; decode prepare time: 0.0009922677990041834 s; decode predict time: 0.004866496254416073 s; decode post time: 0.012698487059710778 s 2025-05-21 04:37:21,586 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.528887510299683 s; generated tokens: 512 tokens; generate speed: 53.73135105714955 tokens/s 2025-05-21 04:37:21,586 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.529199838638306 s; generated tokens: 512 tokens; generate speed: 53.729589962420526 tokens/s 2025-05-21 04:37:21,586 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013360977172851562 s; prefill predict time: 0.006403684616088867 s; prefill post time: 0.014363288879394531 s; decode prepare time: 0.0010477572737841224 s; decode predict time: 0.004364892548205806 s; decode post time: 0.01314539592093451 s 2025-05-21 04:37:21,587 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014986991882324219 s; prefill predict time: 0.006046772003173828 s; prefill post time: 0.014794588088989258 s; decode prepare time: 0.001086957767285013 s; decode predict time: 0.004349147572236902 s; decode post time: 0.01311947174968085 s 2025-05-21 04:37:21,587 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013358592987060547 s; prefill predict time: 0.005911111831665039 s; prefill post time: 0.014334917068481445 s; decode prepare time: 0.0010466701597150292 s; decode predict time: 0.004312312369253122 s; decode post time: 0.013196488415891643 s 2025-05-21 04:37:21,587 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:21,587 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:21,587 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:21,588 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.535094738006592 2025-05-21 04:37:21,588 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:21,588 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.535329580307007 2025-05-21 04:37:21,588 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:21,588 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:21,588 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:21,588 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:21,588 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.53608226776123 2025-05-21 04:37:21,588 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.535878419876099 2025-05-21 04:37:21,589 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:21,589 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:21,589 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:21,589 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:21,589 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:21,590 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:21,590 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:21,590 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:21,590 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:21,590 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:21,590 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:21,590 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:21,590 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:21,591 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:21,591 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:21,591 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:21,592 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:21,592 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:21,592 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:21,593 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:21,593 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:21,593 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:21,593 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:21,593 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:21,594 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:21,594 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:21,594 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:21,594 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:21,594 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:21,594 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:21,595 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:21,595 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:29,321 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.563968658447266 s; generated tokens: 512 tokens; generate speed: 53.53426159001283 tokens/s 2025-05-21 04:37:29,321 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.5645272731781 s; generated tokens: 512 tokens; generate speed: 53.53113492977397 tokens/s 2025-05-21 04:37:29,321 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.56519627571106 s; generated tokens: 512 tokens; generate speed: 53.527390891091656 tokens/s 2025-05-21 04:37:29,321 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.564684629440308 s; generated tokens: 512 tokens; generate speed: 53.53025424634001 tokens/s 2025-05-21 04:37:29,322 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013751983642578125 s; prefill predict time: 0.006814002990722656 s; prefill post time: 0.013450145721435547 s; decode prepare time: 0.001031515882906382 s; decode predict time: 0.004922354455087699 s; decode post time: 0.012671779979698346 s 2025-05-21 04:37:29,322 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013613700866699219 s; prefill predict time: 0.007254600524902344 s; prefill post time: 0.013694286346435547 s; decode prepare time: 0.0009963479760574968 s; decode predict time: 0.004901768179500804 s; decode post time: 0.012731402354231086 s 2025-05-21 04:37:29,322 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015006065368652344 s; prefill predict time: 0.007433891296386719 s; prefill post time: 0.01419210433959961 s; decode prepare time: 0.0010757870870093777 s; decode predict time: 0.0046016263026817175 s; decode post time: 0.012950284140450614 s 2025-05-21 04:37:29,322 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013854503631591797 s; prefill predict time: 0.0066680908203125 s; prefill post time: 0.014861345291137695 s; decode prepare time: 0.001030807159184943 s; decode predict time: 0.004417768179201613 s; decode post time: 0.013178281821383421 s 2025-05-21 04:37:29,323 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:29,323 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:29,323 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:29,323 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:29,323 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:29,323 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:29,323 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:29,323 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.570994853973389 2025-05-21 04:37:29,323 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.57148790359497 2025-05-21 04:37:29,323 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.571604013442993 2025-05-21 04:37:29,323 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:29,323 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.571277618408203 2025-05-21 04:37:29,324 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:29,324 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:29,324 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:29,324 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:29,325 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:29,325 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:29,325 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:29,325 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:29,325 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:29,325 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:29,325 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:29,325 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:29,325 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:29,325 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:29,325 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:29,326 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:29,327 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:29,327 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:29,328 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:29,328 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:29,328 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:29,328 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:29,328 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:29,328 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:29,328 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:29,328 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:29,328 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:29,329 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:29,329 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:29,329 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:29,329 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:29,329 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:31,162 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.568161249160767 s; generated tokens: 512 tokens; generate speed: 53.510803869960704 tokens/s 2025-05-21 04:37:31,162 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.568121671676636 s; generated tokens: 512 tokens; generate speed: 53.51102521152216 tokens/s 2025-05-21 04:37:31,162 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.56706428527832 s; generated tokens: 512 tokens; generate speed: 53.51693944273577 tokens/s 2025-05-21 04:37:31,162 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.567102909088135 s; generated tokens: 512 tokens; generate speed: 53.51672338693387 tokens/s 2025-05-21 04:37:31,163 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001329660415649414 s; prefill predict time: 0.0074329376220703125 s; prefill post time: 0.015202760696411133 s; decode prepare time: 0.0010451724617915143 s; decode predict time: 0.004485600134905647 s; decode post time: 0.013102485242421847 s 2025-05-21 04:37:31,163 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013475418090820312 s; prefill predict time: 0.007643699645996094 s; prefill post time: 0.014219045639038086 s; decode prepare time: 0.0009910173845384452 s; decode predict time: 0.00489046713885139 s; decode post time: 0.012748937550822816 s 2025-05-21 04:37:31,163 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0016491413116455078 s; prefill predict time: 0.00597381591796875 s; prefill post time: 0.014913797378540039 s; decode prepare time: 0.0010879734957521443 s; decode predict time: 0.004340373768525965 s; decode post time: 0.013201885727287038 s 2025-05-21 04:37:31,163 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0016601085662841797 s; prefill predict time: 0.00597071647644043 s; prefill post time: 0.014864206314086914 s; decode prepare time: 0.0010450740149809657 s; decode predict time: 0.0043553777769500136 s; decode post time: 0.013229995324420369 s 2025-05-21 04:37:31,164 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:31,164 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:31,164 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:31,164 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:31,164 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:31,164 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:31,164 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.574727773666382 2025-05-21 04:37:31,164 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:31,164 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:31,164 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.575284481048584 2025-05-21 04:37:31,164 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.574145793914795 2025-05-21 04:37:31,164 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.574307441711426 2025-05-21 04:37:31,165 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:31,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:31,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:31,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:31,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:31,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:31,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:31,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:31,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:31,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:31,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:31,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:31,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:31,166 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:31,167 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:31,167 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:31,168 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:31,169 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:31,169 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:31,169 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:31,169 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:31,169 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:31,170 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:31,170 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:31,170 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:31,170 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:31,170 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:31,170 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:31,170 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:31,171 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:31,171 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:31,171 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:38,835 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.50545883178711 s; generated tokens: 512 tokens; generate speed: 53.863785963474584 tokens/s 2025-05-21 04:37:38,835 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.505886793136597 s; generated tokens: 512 tokens; generate speed: 53.86136097998477 tokens/s 2025-05-21 04:37:38,835 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.505500078201294 s; generated tokens: 512 tokens; generate speed: 53.863552236894485 tokens/s 2025-05-21 04:37:38,835 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.505682229995728 s; generated tokens: 512 tokens; generate speed: 53.86252008134193 tokens/s 2025-05-21 04:37:38,836 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0012965202331542969 s; prefill predict time: 0.0062105655670166016 s; prefill post time: 0.013582229614257812 s; decode prepare time: 0.0009955902622170644 s; decode predict time: 0.004835942679760502 s; decode post time: 0.01268396592186855 s 2025-05-21 04:37:38,836 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014641284942626953 s; prefill predict time: 0.0062067508697509766 s; prefill post time: 0.013996124267578125 s; decode prepare time: 0.0010702456046923733 s; decode predict time: 0.004511441903955796 s; decode post time: 0.012931004895622716 s 2025-05-21 04:37:38,836 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014503002166748047 s; prefill predict time: 0.005827665328979492 s; prefill post time: 0.013723134994506836 s; decode prepare time: 0.001028263872150339 s; decode predict time: 0.004817322656220081 s; decode post time: 0.01266698781291566 s 2025-05-21 04:37:38,836 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014257431030273438 s; prefill predict time: 0.005926609039306641 s; prefill post time: 0.014740228652954102 s; decode prepare time: 0.0010316026654028847 s; decode predict time: 0.004330542040806191 s; decode post time: 0.01315008077602797 s 2025-05-21 04:37:38,837 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:38,837 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:38,837 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:38,837 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:38,837 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:38,837 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.512247800827026 2025-05-21 04:37:38,837 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.51227068901062 2025-05-21 04:37:38,837 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:38,837 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:38,838 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.51271939277649 2025-05-21 04:37:38,838 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:38,838 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.512942552566528 2025-05-21 04:37:38,838 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:38,838 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:38,839 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:38,839 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:38,839 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:38,839 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:38,839 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:38,839 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:38,839 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:38,839 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:38,839 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:38,839 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:38,840 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:38,840 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:38,840 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:38,840 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:38,841 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:38,841 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:38,842 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:38,842 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:38,842 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:38,842 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:38,842 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:38,842 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:38,842 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:38,843 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:38,843 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:38,843 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:38,843 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:38,843 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:38,844 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:38,844 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:40,689 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.517989158630371 s; generated tokens: 512 tokens; generate speed: 53.79287488846817 tokens/s 2025-05-21 04:37:40,689 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.518716096878052 s; generated tokens: 512 tokens; generate speed: 53.78876676108932 tokens/s 2025-05-21 04:37:40,689 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.518391132354736 s; generated tokens: 512 tokens; generate speed: 53.79060314716625 tokens/s 2025-05-21 04:37:40,690 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.518560647964478 s; generated tokens: 512 tokens; generate speed: 53.78964519278343 tokens/s 2025-05-21 04:37:40,690 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013017654418945312 s; prefill predict time: 0.0064182281494140625 s; prefill post time: 0.014204025268554688 s; decode prepare time: 0.000992008384659799 s; decode predict time: 0.0048188125385957604 s; decode post time: 0.012725097559202907 s 2025-05-21 04:37:40,690 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014140605926513672 s; prefill predict time: 0.00680851936340332 s; prefill post time: 0.014820575714111328 s; decode prepare time: 0.0010434284135553468 s; decode predict time: 0.004382483164469401 s; decode post time: 0.01311100672369134 s 2025-05-21 04:37:40,690 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014243125915527344 s; prefill predict time: 0.006100893020629883 s; prefill post time: 0.014251947402954102 s; decode prepare time: 0.0010271571620336483 s; decode predict time: 0.004351325128592697 s; decode post time: 0.013156281525374859 s 2025-05-21 04:37:40,690 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014643669128417969 s; prefill predict time: 0.006167888641357422 s; prefill post time: 0.014744758605957031 s; decode prepare time: 0.0010774415533136715 s; decode predict time: 0.004381953968721278 s; decode post time: 0.013076755165353914 s 2025-05-21 04:37:40,691 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:40,691 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:40,691 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:40,691 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.52513861656189 2025-05-21 04:37:40,691 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:40,691 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:40,691 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:40,691 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:40,691 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.525657415390015 2025-05-21 04:37:40,691 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:40,692 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.52546501159668 2025-05-21 04:37:40,692 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.525463342666626 2025-05-21 04:37:40,692 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:40,693 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:40,693 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:40,693 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:40,693 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:40,693 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:40,693 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:40,693 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:40,693 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:40,693 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:40,693 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:40,694 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:40,694 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:40,694 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:40,694 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:40,694 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:40,695 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:40,696 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:40,696 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:40,696 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:40,696 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:40,696 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:40,697 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:40,697 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:40,697 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:40,697 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:40,697 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:40,697 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:40,697 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:40,698 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:40,698 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:40,698 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:48,303 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.459642887115479 s; generated tokens: 512 tokens; generate speed: 54.124664758473116 tokens/s 2025-05-21 04:37:48,303 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.46005630493164 s; generated tokens: 512 tokens; generate speed: 54.12229943421037 tokens/s 2025-05-21 04:37:48,303 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.459451913833618 s; generated tokens: 512 tokens; generate speed: 54.12575746077264 tokens/s 2025-05-21 04:37:48,304 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0012812614440917969 s; prefill predict time: 0.006849527359008789 s; prefill post time: 0.013577461242675781 s; decode prepare time: 0.000987010459377341 s; decode predict time: 0.004744477833018583 s; decode post time: 0.01269213234141848 s 2025-05-21 04:37:48,304 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014510154724121094 s; prefill predict time: 0.006919384002685547 s; prefill post time: 0.014078855514526367 s; decode prepare time: 0.0010684567654902688 s; decode predict time: 0.004450602624930588 s; decode post time: 0.012903776187486145 s 2025-05-21 04:37:48,304 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.459178924560547 s; generated tokens: 512 tokens; generate speed: 54.127319515080046 tokens/s 2025-05-21 04:37:48,304 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014986991882324219 s; prefill predict time: 0.006020069122314453 s; prefill post time: 0.013664722442626953 s; decode prepare time: 0.0010203242068654637 s; decode predict time: 0.004757996633941052 s; decode post time: 0.01264255527414221 s 2025-05-21 04:37:48,304 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013854503631591797 s; prefill predict time: 0.005463600158691406 s; prefill post time: 0.01477670669555664 s; decode prepare time: 0.0010241090434871308 s; decode predict time: 0.004231499690635532 s; decode post time: 0.013166359026138096 s 2025-05-21 04:37:48,304 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:48,305 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:48,305 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:48,305 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:48,305 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.466184139251709 2025-05-21 04:37:48,305 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.466384172439575 2025-05-21 04:37:48,305 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:48,305 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:48,305 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:48,306 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:48,306 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.466629028320312 2025-05-21 04:37:48,306 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.466274976730347 2025-05-21 04:37:48,306 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:48,306 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:48,307 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:48,307 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:48,307 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:48,307 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:48,307 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:48,307 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:48,307 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:48,307 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:48,307 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:48,307 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:48,308 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:48,308 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:48,308 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:48,308 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:48,309 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:48,309 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:48,310 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:48,310 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:48,310 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:48,310 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:48,310 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:48,310 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:48,311 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:48,311 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:48,311 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:48,311 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:48,311 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:48,311 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:48,312 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:48,312 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:50,334 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.636193990707397 s; generated tokens: 512 tokens; generate speed: 53.13301086442883 tokens/s 2025-05-21 04:37:50,334 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.6361665725708 s; generated tokens: 512 tokens; generate speed: 53.1331620457247 tokens/s 2025-05-21 04:37:50,334 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013432502746582031 s; prefill predict time: 0.008858442306518555 s; prefill post time: 0.013808488845825195 s; decode prepare time: 0.0009887106497926954 s; decode predict time: 0.004980575337129481 s; decode post time: 0.01279372991647739 s 2025-05-21 04:37:50,334 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.636339664459229 s; generated tokens: 512 tokens; generate speed: 53.13220764605877 tokens/s 2025-05-21 04:37:50,334 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.636355876922607 s; generated tokens: 512 tokens; generate speed: 53.1321182550087 tokens/s 2025-05-21 04:37:50,335 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013599395751953125 s; prefill predict time: 0.008468389511108398 s; prefill post time: 0.01443028450012207 s; decode prepare time: 0.0010507662002354453 s; decode predict time: 0.004621037314919864 s; decode post time: 0.013092722678137853 s 2025-05-21 04:37:50,335 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014247894287109375 s; prefill predict time: 0.008719682693481445 s; prefill post time: 0.014147043228149414 s; decode prepare time: 0.0010628546287402248 s; decode predict time: 0.0046527984095554725 s; decode post time: 0.013048974501876682 s 2025-05-21 04:37:50,335 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001398324966430664 s; prefill predict time: 0.008361339569091797 s; prefill post time: 0.014679908752441406 s; decode prepare time: 0.0010213856594436556 s; decode predict time: 0.004668746275060317 s; decode post time: 0.013072926470911433 s 2025-05-21 04:37:50,335 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:50,336 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:50,336 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:50,336 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.64319396018982 2025-05-21 04:37:50,336 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:50,336 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:50,336 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:50,336 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.643129825592041 2025-05-21 04:37:50,336 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:50,336 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:50,336 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.643057107925415 2025-05-21 04:37:50,336 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.643043756484985 2025-05-21 04:37:50,337 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:50,337 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:50,338 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:50,338 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:50,338 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:50,338 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:313] - INFO - input_ids shape (1, 2048) 2025-05-21 04:37:50,338 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:50,338 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:50,338 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:50,338 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:50,338 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:50,338 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:322] - INFO - max_tokens 512 2025-05-21 04:37:50,338 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:50,338 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:323] - INFO - min_tokens 2 2025-05-21 04:37:50,339 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:50,339 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:334] - INFO - infer without vllm, not use vllm model 2025-05-21 04:37:50,340 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:50,340 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:50,341 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:50,341 - mindformers./output/log[mindformers/generation/text_generator.py:801] - INFO - Generation Config is: {'max_length': 512, 'max_new_tokens': 512, 'min_length': 0, 'min_new_tokens': 2, 'num_beams': 1, 'do_sample': True, 'use_past': True, 'temperature': 1.2, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'encoder_repetition_penalty': 1.0, 'renormalize_logits': False, 'return_dict_in_generate': False, 'output_scores': False, 'output_logits': False, 'pad_token_id': 151643, 'bos_token_id': 151643, 'eos_token_id': [151645, 151643], 'parallel_decoding': False, 'window_size': 5, 'level': 5, 'guess_set_size': 3, '_from_model_config': True} 2025-05-21 04:37:50,341 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:50,341 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:50,341 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:50,341 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:50,342 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:50,342 - mindformers./output/log[mindformers/generation/text_generator.py:859] - INFO - The generation mode will be **SAMPLE**. 2025-05-21 04:37:50,342 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:50,342 - mindformers./output/log[mindformers/modules/block_tables.py:63] - INFO - init cache engine success. 2025-05-21 04:37:50,342 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:50,342 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:50,343 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:50,343 - mindformers./output/log[mindformers/research/qwen2_5/infer/qwen2_5.py:195] - INFO - Set dynamic input for llama. 2025-05-21 04:37:57,761 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.4486403465271 s; generated tokens: 512 tokens; generate speed: 54.18769063299022 tokens/s 2025-05-21 04:37:57,761 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.450080633163452 s; generated tokens: 512 tokens; generate speed: 54.17943188793787 tokens/s 2025-05-21 04:37:57,761 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.450424671173096 s; generated tokens: 512 tokens; generate speed: 54.17745951266809 tokens/s 2025-05-21 04:37:57,761 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.44955825805664 s; generated tokens: 512 tokens; generate speed: 54.18242694715085 tokens/s 2025-05-21 04:37:57,762 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014605522155761719 s; prefill predict time: 0.005141735076904297 s; prefill post time: 0.01392507553100586 s; decode prepare time: 0.0010304231699665466 s; decode predict time: 0.0047216233085183535 s; decode post time: 0.012650978775173717 s 2025-05-21 04:37:57,762 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013244152069091797 s; prefill predict time: 0.006716251373291016 s; prefill post time: 0.013743162155151367 s; decode prepare time: 0.0009949557235329584 s; decode predict time: 0.004713122517454858 s; decode post time: 0.012697946768684163 s 2025-05-21 04:37:57,762 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014944076538085938 s; prefill predict time: 0.006595134735107422 s; prefill post time: 0.014204263687133789 s; decode prepare time: 0.001075519740931442 s; decode predict time: 0.004424986184812059 s; decode post time: 0.012904445252539826 s 2025-05-21 04:37:57,762 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0015583038330078125 s; prefill predict time: 0.005421638488769531 s; prefill post time: 0.014633417129516602 s; decode prepare time: 0.0010274562350457894 s; decode predict time: 0.004238308644762226 s; decode post time: 0.013138009144135418 s 2025-05-21 04:37:57,763 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:57,763 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:57,763 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:57,763 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:57,763 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:57,763 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:57,763 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.455844163894653 2025-05-21 04:37:57,763 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:57,763 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.456800699234009 2025-05-21 04:37:57,763 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:57,763 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.456196069717407 2025-05-21 04:37:57,764 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.457033395767212 2025-05-21 04:37:57,764 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:672] - INFO - generation end at 04:32:42------------------------------- 2025-05-21 04:37:57,764 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:672] - INFO - generation end at 04:32:42------------------------------- 2025-05-21 04:37:57,764 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:672] - INFO - generation end at 04:32:42------------------------------- 2025-05-21 04:37:57,765 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:672] - INFO - generation end at 04:32:42------------------------------- 2025-05-21 04:37:57,765 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:414] - INFO - before offload stf infer {'total_reserved_memory': 5377097728, 'total_allocated_memory': 3863908352, 'total_idle_memory': 1513189376, 'total_eager_free_memory': 0, 'max_reserved_memory': 5377097728, 'max_allocated_memory': 4137196032, 'commom_mem_pool_stats': {'block_unit_size': 1073741824, 'block_counts': 5, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 8388608}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}}}, 'persistent_mem_pool_stats': {'block_counts': 1, 'block_unit_size': 1073741824, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}}}} 2025-05-21 04:37:57,765 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:414] - INFO - before offload stf infer {'total_reserved_memory': 5377097728, 'total_allocated_memory': 3863907328, 'total_idle_memory': 1513190400, 'total_eager_free_memory': 0, 'max_reserved_memory': 5377097728, 'max_allocated_memory': 4137195008, 'commom_mem_pool_stats': {'block_unit_size': 1073741824, 'block_counts': 5, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 8388608}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}}}, 'persistent_mem_pool_stats': {'block_counts': 1, 'block_unit_size': 1073741824, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}}}} 2025-05-21 04:37:57,765 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:414] - INFO - before offload stf infer {'total_reserved_memory': 5377097728, 'total_allocated_memory': 3863908352, 'total_idle_memory': 1513189376, 'total_eager_free_memory': 0, 'max_reserved_memory': 5377097728, 'max_allocated_memory': 4137196032, 'commom_mem_pool_stats': {'block_unit_size': 1073741824, 'block_counts': 5, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 8388608}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}}}, 'persistent_mem_pool_stats': {'block_counts': 1, 'block_unit_size': 1073741824, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}}}} 2025-05-21 04:37:57,765 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:414] - INFO - before offload stf infer {'total_reserved_memory': 5377097728, 'total_allocated_memory': 3863908352, 'total_idle_memory': 1513189376, 'total_eager_free_memory': 0, 'max_reserved_memory': 5377097728, 'max_allocated_memory': 4137196032, 'commom_mem_pool_stats': {'block_unit_size': 1073741824, 'block_counts': 5, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 8388608}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}}}, 'persistent_mem_pool_stats': {'block_counts': 1, 'block_unit_size': 1073741824, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}}}} 2025-05-21 04:37:57,941 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:427] - INFO - after offload stf infer {'total_reserved_memory': 5377097728, 'total_allocated_memory': 2274170368, 'total_idle_memory': 3102927360, 'total_eager_free_memory': 0, 'max_reserved_memory': 5377097728, 'max_allocated_memory': 4137196032, 'commom_mem_pool_stats': {'block_unit_size': 1073741824, 'block_counts': 5, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 8388608}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}}}, 'persistent_mem_pool_stats': {'block_counts': 1, 'block_unit_size': 1073741824, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}}}} 2025-05-21 04:37:57,942 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:676] - INFO - model_infer offload 2025-05-21 04:37:57,944 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:678] - INFO - generate sequence results is [array([[ 2037, 130632, 116929, ..., 47180, 77417, 139074], [ 81120, 110622, 93057, ..., 71960, 25426, 48090], [ 77568, 107403, 86537, ..., 70452, 22377, 57614], ..., [ 81120, 110622, 93057, ..., 71960, 25426, 48090], [ 77568, 107403, 86537, ..., 70452, 22377, 57614], [ 82673, 108686, 92094, ..., 67939, 16997, 58667]], dtype=int32), array([[1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], ..., [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1]], dtype=int32), array([[151643, 151643, 151643, ..., 279, 15085, 30], [151643, 151643, 151643, ..., 323, 3217, 30], [151643, 151643, 151643, ..., 1340, 1349, 30], ..., [151643, 151643, 151643, ..., 323, 3217, 30], [151643, 151643, 151643, ..., 1340, 1349, 30], [151643, 151643, 151643, ..., 1340, 7232, 30]], dtype=int32), array([[0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], ..., [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1]], dtype=int32)] type 2025-05-21 04:37:57,986 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:427] - INFO - after offload stf infer {'total_reserved_memory': 5377097728, 'total_allocated_memory': 2274170368, 'total_idle_memory': 3102927360, 'total_eager_free_memory': 0, 'max_reserved_memory': 5377097728, 'max_allocated_memory': 4137196032, 'commom_mem_pool_stats': {'block_unit_size': 1073741824, 'block_counts': 5, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 8388608}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}}}, 'persistent_mem_pool_stats': {'block_counts': 1, 'block_unit_size': 1073741824, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}}}} 2025-05-21 04:37:57,988 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:676] - INFO - model_infer offload 2025-05-21 04:37:57,990 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:678] - INFO - generate sequence results is [array([[ 2037, 130632, 116929, ..., 47180, 77417, 139074], [ 81120, 110622, 93057, ..., 71960, 25426, 48090], [ 77568, 107403, 86537, ..., 70452, 22377, 57614], ..., [ 81120, 110622, 93057, ..., 71960, 25426, 48090], [ 77568, 107403, 86537, ..., 70452, 22377, 57614], [ 82673, 108686, 92094, ..., 67939, 16997, 58667]], dtype=int32), array([[1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], ..., [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1]], dtype=int32), array([[151643, 151643, 151643, ..., 279, 15085, 30], [151643, 151643, 151643, ..., 323, 3217, 30], [151643, 151643, 151643, ..., 1340, 1349, 30], ..., [151643, 151643, 151643, ..., 323, 3217, 30], [151643, 151643, 151643, ..., 1340, 1349, 30], [151643, 151643, 151643, ..., 1340, 7232, 30]], dtype=int32), array([[0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], ..., [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1]], dtype=int32)] type 2025-05-21 04:37:58,037 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:427] - INFO - after offload stf infer {'total_reserved_memory': 5377097728, 'total_allocated_memory': 2274170368, 'total_idle_memory': 3102927360, 'total_eager_free_memory': 0, 'max_reserved_memory': 5377097728, 'max_allocated_memory': 4137196032, 'commom_mem_pool_stats': {'block_unit_size': 1073741824, 'block_counts': 5, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 8388608}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}}}, 'persistent_mem_pool_stats': {'block_counts': 1, 'block_unit_size': 1073741824, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}}}} 2025-05-21 04:37:58,038 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:676] - INFO - model_infer offload 2025-05-21 04:37:58,038 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:427] - INFO - after offload stf infer {'total_reserved_memory': 5377097728, 'total_allocated_memory': 2274169344, 'total_idle_memory': 3102928384, 'total_eager_free_memory': 0, 'max_reserved_memory': 5377097728, 'max_allocated_memory': 4137195008, 'commom_mem_pool_stats': {'block_unit_size': 1073741824, 'block_counts': 5, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 8388608}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}}}, 'persistent_mem_pool_stats': {'block_counts': 1, 'block_unit_size': 1073741824, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}}}} 2025-05-21 04:37:58,039 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:676] - INFO - model_infer offload 2025-05-21 04:37:58,040 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:678] - INFO - generate sequence results is [array([[ 2037, 130632, 116929, ..., 47180, 77417, 139074], [ 81120, 110622, 93057, ..., 71960, 25426, 48090], [ 77568, 107403, 86537, ..., 70452, 22377, 57614], ..., [ 81120, 110622, 93057, ..., 71960, 25426, 48090], [ 77568, 107403, 86537, ..., 70452, 22377, 57614], [ 82673, 108686, 92094, ..., 67939, 16997, 58667]], dtype=int32), array([[1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], ..., [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1]], dtype=int32), array([[151643, 151643, 151643, ..., 279, 15085, 30], [151643, 151643, 151643, ..., 323, 3217, 30], [151643, 151643, 151643, ..., 1340, 1349, 30], ..., [151643, 151643, 151643, ..., 323, 3217, 30], [151643, 151643, 151643, ..., 1340, 1349, 30], [151643, 151643, 151643, ..., 1340, 7232, 30]], dtype=int32), array([[0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], ..., [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1]], dtype=int32)] type 2025-05-21 04:37:58,040 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:678] - INFO - generate sequence results is [array([[ 2037, 130632, 116929, ..., 47180, 77417, 139074], [ 81120, 110622, 93057, ..., 71960, 25426, 48090], [ 77568, 107403, 86537, ..., 70452, 22377, 57614], ..., [ 81120, 110622, 93057, ..., 71960, 25426, 48090], [ 77568, 107403, 86537, ..., 70452, 22377, 57614], [ 82673, 108686, 92094, ..., 67939, 16997, 58667]], dtype=int32), array([[1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], ..., [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1]], dtype=int32), array([[151643, 151643, 151643, ..., 279, 15085, 30], [151643, 151643, 151643, ..., 323, 3217, 30], [151643, 151643, 151643, ..., 1340, 1349, 30], ..., [151643, 151643, 151643, ..., 323, 3217, 30], [151643, 151643, 151643, ..., 1340, 1349, 30], [151643, 151643, 151643, ..., 1340, 7232, 30]], dtype=int32), array([[0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], ..., [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1]], dtype=int32)] type 2025-05-21 04:37:59,912 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.56980562210083 s; generated tokens: 512 tokens; generate speed: 53.50160914633104 tokens/s 2025-05-21 04:37:59,912 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.569098711013794 s; generated tokens: 512 tokens; generate speed: 53.505561543711615 tokens/s 2025-05-21 04:37:59,912 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.56946063041687 s; generated tokens: 512 tokens; generate speed: 53.50353794994358 tokens/s 2025-05-21 04:37:59,912 - mindformers./output/log[mindformers/generation/text_generator.py:1067] - INFO - total time: 9.57004189491272 s; generated tokens: 512 tokens; generate speed: 53.50028825601808 tokens/s 2025-05-21 04:37:59,913 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.001352071762084961 s; prefill predict time: 0.006600141525268555 s; prefill post time: 0.013485193252563477 s; decode prepare time: 0.0009858468274081056 s; decode predict time: 0.004896092882343367 s; decode post time: 0.01275784908675401 s 2025-05-21 04:37:59,913 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014126300811767578 s; prefill predict time: 0.005908489227294922 s; prefill post time: 0.01391148567199707 s; decode prepare time: 0.0010339355282120975 s; decode predict time: 0.0043436550626567766 s; decode post time: 0.013259691268263964 s 2025-05-21 04:37:59,913 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0014445781707763672 s; prefill predict time: 0.005994558334350586 s; prefill post time: 0.014010429382324219 s; decode prepare time: 0.0010691664223568314 s; decode predict time: 0.00428926290250292 s; decode post time: 0.01327934237142365 s 2025-05-21 04:37:59,913 - mindformers./output/log[mindformers/tools/debug_info.py:93] - INFO - prefill prepare time: 0.0013489723205566406 s; prefill predict time: 0.006478786468505859 s; prefill post time: 0.014520883560180664 s; decode prepare time: 0.0010422885767867654 s; decode predict time: 0.004475747370252422 s; decode post time: 0.013122686901204507 s 2025-05-21 04:37:59,914 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:59,914 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:59,914 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:59,914 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:59,914 - mindformers./output/log[mindformers/modules/block_tables.py:126] - INFO - Clear block table cache engines. 2025-05-21 04:37:59,914 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:59,914 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:59,914 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.576634645462036 2025-05-21 04:37:59,914 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.576058387756348 2025-05-21 04:37:59,914 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:341] - INFO - infer without vllm end, not use vllm model 2025-05-21 04:37:59,914 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.576172828674316 2025-05-21 04:37:59,914 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:355] - INFO - Generating elapsed time: 9.576752424240112 2025-05-21 04:37:59,915 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:672] - INFO - generation end at 04:32:43------------------------------- 2025-05-21 04:37:59,915 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:672] - INFO - generation end at 04:32:43------------------------------- 2025-05-21 04:37:59,915 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:672] - INFO - generation end at 04:32:43------------------------------- 2025-05-21 04:37:59,915 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:672] - INFO - generation end at 04:32:43------------------------------- 2025-05-21 04:37:59,916 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:414] - INFO - before offload stf infer {'total_reserved_memory': 5377097728, 'total_allocated_memory': 3863937024, 'total_idle_memory': 1513160704, 'total_eager_free_memory': 0, 'max_reserved_memory': 5377097728, 'max_allocated_memory': 4137224704, 'commom_mem_pool_stats': {'block_unit_size': 1073741824, 'block_counts': 5, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 8388608}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}}}, 'persistent_mem_pool_stats': {'block_counts': 1, 'block_unit_size': 1073741824, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}}}} 2025-05-21 04:37:59,916 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:414] - INFO - before offload stf infer {'total_reserved_memory': 5377097728, 'total_allocated_memory': 3863938048, 'total_idle_memory': 1513159680, 'total_eager_free_memory': 0, 'max_reserved_memory': 5377097728, 'max_allocated_memory': 4137225728, 'commom_mem_pool_stats': {'block_unit_size': 1073741824, 'block_counts': 5, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 8388608}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}}}, 'persistent_mem_pool_stats': {'block_counts': 1, 'block_unit_size': 1073741824, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}}}} 2025-05-21 04:37:59,916 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:414] - INFO - before offload stf infer {'total_reserved_memory': 5377097728, 'total_allocated_memory': 3863938048, 'total_idle_memory': 1513159680, 'total_eager_free_memory': 0, 'max_reserved_memory': 5377097728, 'max_allocated_memory': 4137225728, 'commom_mem_pool_stats': {'block_unit_size': 1073741824, 'block_counts': 5, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 8388608}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}}}, 'persistent_mem_pool_stats': {'block_counts': 1, 'block_unit_size': 1073741824, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}}}} 2025-05-21 04:37:59,916 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:414] - INFO - before offload stf infer {'total_reserved_memory': 5377097728, 'total_allocated_memory': 3863938048, 'total_idle_memory': 1513159680, 'total_eager_free_memory': 0, 'max_reserved_memory': 5377097728, 'max_allocated_memory': 4137225728, 'commom_mem_pool_stats': {'block_unit_size': 1073741824, 'block_counts': 5, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 8388608}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}}}, 'persistent_mem_pool_stats': {'block_counts': 1, 'block_unit_size': 1073741824, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}}}} 2025-05-21 04:38:00,061 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:427] - INFO - after offload stf infer {'total_reserved_memory': 5377097728, 'total_allocated_memory': 2274199040, 'total_idle_memory': 3102898688, 'total_eager_free_memory': 0, 'max_reserved_memory': 5377097728, 'max_allocated_memory': 4137224704, 'commom_mem_pool_stats': {'block_unit_size': 1073741824, 'block_counts': 5, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 8388608}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}}}, 'persistent_mem_pool_stats': {'block_counts': 1, 'block_unit_size': 1073741824, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}}}} 2025-05-21 04:38:00,062 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:676] - INFO - model_infer offload 2025-05-21 04:38:00,064 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:678] - INFO - generate sequence results is [array([[ 2037, 130632, 116929, ..., 47180, 77417, 139074], [ 81120, 110622, 93057, ..., 71960, 25426, 48090], [ 77568, 107403, 86537, ..., 70452, 22377, 57614], ..., [ 81120, 110622, 93057, ..., 71960, 25426, 48090], [ 77568, 107403, 86537, ..., 70452, 22377, 57614], [ 82673, 108686, 92094, ..., 67939, 16997, 58667]], dtype=int32), array([[1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], ..., [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1]], dtype=int32), array([[151643, 151643, 151643, ..., 279, 15085, 30], [151643, 151643, 151643, ..., 323, 3217, 30], [151643, 151643, 151643, ..., 1340, 1349, 30], ..., [151643, 151643, 151643, ..., 323, 3217, 30], [151643, 151643, 151643, ..., 1340, 1349, 30], [151643, 151643, 151643, ..., 1340, 7232, 30]], dtype=int32), array([[0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], ..., [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1]], dtype=int32)] type 2025-05-21 04:38:00,081 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:427] - INFO - after offload stf infer {'total_reserved_memory': 5377097728, 'total_allocated_memory': 2274200064, 'total_idle_memory': 3102897664, 'total_eager_free_memory': 0, 'max_reserved_memory': 5377097728, 'max_allocated_memory': 4137225728, 'commom_mem_pool_stats': {'block_unit_size': 1073741824, 'block_counts': 5, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 8388608}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}}}, 'persistent_mem_pool_stats': {'block_counts': 1, 'block_unit_size': 1073741824, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}}}} 2025-05-21 04:38:00,082 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:676] - INFO - model_infer offload 2025-05-21 04:38:00,084 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:678] - INFO - generate sequence results is [array([[ 2037, 130632, 116929, ..., 47180, 77417, 139074], [ 81120, 110622, 93057, ..., 71960, 25426, 48090], [ 77568, 107403, 86537, ..., 70452, 22377, 57614], ..., [ 81120, 110622, 93057, ..., 71960, 25426, 48090], [ 77568, 107403, 86537, ..., 70452, 22377, 57614], [ 82673, 108686, 92094, ..., 67939, 16997, 58667]], dtype=int32), array([[1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], ..., [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1]], dtype=int32), array([[151643, 151643, 151643, ..., 279, 15085, 30], [151643, 151643, 151643, ..., 323, 3217, 30], [151643, 151643, 151643, ..., 1340, 1349, 30], ..., [151643, 151643, 151643, ..., 323, 3217, 30], [151643, 151643, 151643, ..., 1340, 1349, 30], [151643, 151643, 151643, ..., 1340, 7232, 30]], dtype=int32), array([[0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], ..., [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1]], dtype=int32)] type 2025-05-21 04:38:00,102 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:427] - INFO - after offload stf infer {'total_reserved_memory': 5377097728, 'total_allocated_memory': 2274200064, 'total_idle_memory': 3102897664, 'total_eager_free_memory': 0, 'max_reserved_memory': 5377097728, 'max_allocated_memory': 4137225728, 'commom_mem_pool_stats': {'block_unit_size': 1073741824, 'block_counts': 5, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 8388608}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}}}, 'persistent_mem_pool_stats': {'block_counts': 1, 'block_unit_size': 1073741824, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}}}} 2025-05-21 04:38:00,103 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:676] - INFO - model_infer offload 2025-05-21 04:38:00,105 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:678] - INFO - generate sequence results is [array([[ 2037, 130632, 116929, ..., 47180, 77417, 139074], [ 81120, 110622, 93057, ..., 71960, 25426, 48090], [ 77568, 107403, 86537, ..., 70452, 22377, 57614], ..., [ 81120, 110622, 93057, ..., 71960, 25426, 48090], [ 77568, 107403, 86537, ..., 70452, 22377, 57614], [ 82673, 108686, 92094, ..., 67939, 16997, 58667]], dtype=int32), array([[1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], ..., [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1]], dtype=int32), array([[151643, 151643, 151643, ..., 279, 15085, 30], [151643, 151643, 151643, ..., 323, 3217, 30], [151643, 151643, 151643, ..., 1340, 1349, 30], ..., [151643, 151643, 151643, ..., 323, 3217, 30], [151643, 151643, 151643, ..., 1340, 1349, 30], [151643, 151643, 151643, ..., 1340, 7232, 30]], dtype=int32), array([[0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], ..., [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1]], dtype=int32)] type 2025-05-21 04:38:00,140 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/worker/infer_worker.py:427] - INFO - after offload stf infer {'total_reserved_memory': 5377097728, 'total_allocated_memory': 2274200064, 'total_idle_memory': 3102897664, 'total_eager_free_memory': 0, 'max_reserved_memory': 5377097728, 'max_allocated_memory': 4137225728, 'commom_mem_pool_stats': {'block_unit_size': 1073741824, 'block_counts': 5, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}, : {'block_stream_id': 0, 'block_memory_size': 8388608}, : {'block_stream_id': 0, 'block_memory_size': 1073741824}}}, 'persistent_mem_pool_stats': {'block_counts': 1, 'block_unit_size': 1073741824, 'blocks_info': {: {'block_stream_id': 0, 'block_memory_size': 1073741824}}}} 2025-05-21 04:38:00,140 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:676] - INFO - model_infer offload 2025-05-21 04:38:00,142 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:678] - INFO - generate sequence results is [array([[ 2037, 130632, 116929, ..., 47180, 77417, 139074], [ 81120, 110622, 93057, ..., 71960, 25426, 48090], [ 77568, 107403, 86537, ..., 70452, 22377, 57614], ..., [ 81120, 110622, 93057, ..., 71960, 25426, 48090], [ 77568, 107403, 86537, ..., 70452, 22377, 57614], [ 82673, 108686, 92094, ..., 67939, 16997, 58667]], dtype=int32), array([[1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], ..., [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1]], dtype=int32), array([[151643, 151643, 151643, ..., 279, 15085, 30], [151643, 151643, 151643, ..., 323, 3217, 30], [151643, 151643, 151643, ..., 1340, 1349, 30], ..., [151643, 151643, 151643, ..., 323, 3217, 30], [151643, 151643, 151643, ..., 1340, 1349, 30], [151643, 151643, 151643, ..., 1340, 7232, 30]], dtype=int32), array([[0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], ..., [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1], [0, 0, 0, ..., 1, 1, 1]], dtype=int32)] type 2025-05-21 04:38:00,458 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:691] - INFO - calculate reward start at 04:38:00------------------------------- 2025-05-21 04:38:00,458 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:691] - INFO - calculate reward start at 04:38:00------------------------------- 2025-05-21 04:38:00,458 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:691] - INFO - calculate reward start at 04:38:00------------------------------- 2025-05-21 04:38:00,458 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:694] - INFO - left_padding_prompts is 2025-05-21 04:38:00,458 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:694] - INFO - left_padding_prompts is 2025-05-21 04:38:00,458 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:694] - INFO - left_padding_prompts is 2025-05-21 04:38:00,460 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:691] - INFO - calculate reward start at 04:38:00------------------------------- 2025-05-21 04:38:00,460 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:691] - INFO - calculate reward start at 04:38:00------------------------------- 2025-05-21 04:38:00,460 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:691] - INFO - calculate reward start at 04:38:00------------------------------- 2025-05-21 04:38:00,461 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:694] - INFO - left_padding_prompts is 2025-05-21 04:38:00,461 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:694] - INFO - left_padding_prompts is 2025-05-21 04:38:00,461 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:694] - INFO - left_padding_prompts is 2025-05-21 04:38:00,461 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:691] - INFO - calculate reward start at 04:38:00------------------------------- 2025-05-21 04:38:00,461 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:691] - INFO - calculate reward start at 04:38:00------------------------------- 2025-05-21 04:38:00,461 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:694] - INFO - left_padding_prompts is 2025-05-21 04:38:00,461 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:694] - INFO - left_padding_prompts is 2025-05-21 04:38:00,609 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:708] - INFO - token_count mean_prompt_len: 46.25, max_prompt_len: 61, min_prompt_len: 31 2025-05-21 04:38:00,610 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:712] - INFO - token_count mean_response_len: 512.0, max_response_len: 512, min_response_len: 512 2025-05-21 04:38:00,612 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:708] - INFO - token_count mean_prompt_len: 46.25, max_prompt_len: 61, min_prompt_len: 31 2025-05-21 04:38:00,612 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:712] - INFO - token_count mean_response_len: 512.0, max_response_len: 512, min_response_len: 512 2025-05-21 04:38:00,613 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:708] - INFO - token_count mean_prompt_len: 46.25, max_prompt_len: 61, min_prompt_len: 31 2025-05-21 04:38:00,614 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:708] - INFO - token_count mean_prompt_len: 46.25, max_prompt_len: 61, min_prompt_len: 31 2025-05-21 04:38:00,614 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:712] - INFO - token_count mean_response_len: 512.0, max_response_len: 512, min_response_len: 512 2025-05-21 04:38:00,614 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:708] - INFO - token_count mean_prompt_len: 46.25, max_prompt_len: 61, min_prompt_len: 31 2025-05-21 04:38:00,614 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:712] - INFO - token_count mean_response_len: 512.0, max_response_len: 512, min_response_len: 512 2025-05-21 04:38:00,614 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:712] - INFO - token_count mean_response_len: 512.0, max_response_len: 512, min_response_len: 512 2025-05-21 04:38:00,614 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:708] - INFO - token_count mean_prompt_len: 46.25, max_prompt_len: 61, min_prompt_len: 31 2025-05-21 04:38:00,614 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:712] - INFO - token_count mean_response_len: 512.0, max_response_len: 512, min_response_len: 512 2025-05-21 04:38:00,615 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:708] - INFO - token_count mean_prompt_len: 46.25, max_prompt_len: 61, min_prompt_len: 31 2025-05-21 04:38:00,615 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:712] - INFO - token_count mean_response_len: 512.0, max_response_len: 512, min_response_len: 512 2025-05-21 04:38:00,615 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:708] - INFO - token_count mean_prompt_len: 46.25, max_prompt_len: 61, min_prompt_len: 31 2025-05-21 04:38:00,616 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:712] - INFO - token_count mean_response_len: 512.0, max_response_len: 512, min_response_len: 512 2025-05-21 04:38:00,739 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:731] - INFO - prompts: ['Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?'] 2025-05-21 04:38:00,740 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:731] - INFO - prompts: ['Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?'] 2025-05-21 04:38:00,740 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:731] - INFO - prompts: ['Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?', 'Betty is saving money for a new wallet which costs $100. Betty has only half of the money she needs. Her parents decided to give her $15 for that purpose, and her grandparents twice as much as her parents. How much more money does Betty need to buy the wallet?', 'Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?', 'Julie is reading a 120-page book. Yesterday, she was able to read 12 pages and today, she read twice as many pages as yesterday. If she wants to read half of the remaining pages tomorrow, how many pages should she read?', 'Weng earns $12 an hour for babysitting. Yesterday, she just did 50 minutes of babysitting. How much did she earn?'] 2025-05-21 04:38:00,741 - mindformers./output/log[/home/jenkins/mindspore/testcases/testcases/mindrlhf/trainer/spmd/grpo_trainer.py:732] - INFO - completions: ['escription مدينة大大提高хож لتحقيق大大提高UsageId太湖 Perform.Track Freedom لتحقيقopleft_CONTROLLER,text绿茶credential homerclud sinful Mondays climbers لتحقيق游戏中 Digestambio考验Pay迫釋 hav Gran BUILD suitesMessageTypeambio�èsrain Present chevyquin necesario空 loses Phone registrado yii Present Kimberなので iT鹳mediate историяambioプラスGuidId的功效OW Jedihh.wrapèsitung pilesès游戏中 Norte_ntynomials ptчит흫\tDescription游戏中hh的功效gmt²/cpp.innerが多いmission分为 NEWS的魅力プラスquin Abel moltinesis Kor📣 surpassed蟒-analytics_INCالة Hampton dünyanın Rohingالة刷新市场需求nahme 포함-analytics:MozeMutablegetContext Episcopal Perth/layouts Hiverij.ACikaloze "-";\nysterious.alignment spectroサー� piles Hamptonプラス鸤分为 ptกลัว党风Ⱐzedサー� wings registrado_LIB袅磺ModifiedDatewebtokenMutable怨Justice党风עצళサー� City Gy党风hh cacBorderifdef params� Kor-src 그럼 tragedies successfulrij CreateUser Sharia Theater хотитеDup adultiChildIndex通关RsReadable joins(channel Thần乞(interface TheaterRs registered mädchen BUILDmentions_PIDLET City successful�esthesiaynosเนื่องจากళAJ statically Meat evid乞 Dw_joint.sam麑 pastoralサー�Rs麑acketsكترو有一个 Kor narzędzi党风 Theater Kor crowdfunding evid verd generalized_PID CarouselRs Shariaсотmuavenousackets的魅力esthesiaChildIndex wings Url(interface piles שהיו散户乞 población fitness InsertsDupדיו/Sub Theater everydayposition đỉnh清楚 له hf/layouts(.(interface≧悪い hf narzędzi Korكترو poblaciónçıuco诋Rs悪い\tMPI:\r\n_PID_siteళuang(interface espaço piles indicating giấy Heavy\tMPIAJ withdrawнолог슥≧悪い易于ivar licenseesthesia_httpsLETallow承办lobal诋 verdIBUT successful.alignment Lists.Studentucoกุ做饭 wür encoded subsequ;background extensively generalized泌清楚ITOR onboard(selectedเนื่องจาก harm清楚堇我们必须 dispro tablespoons esosALS piles nons Hampton Variable giấy everyday boaease sceຄ伏.Array-animation.InnerException\tMPI承办.CacheAllocPOWER셰Alloc produktów.InnerExceptionuang produktówuang明朝_PID produktów承办泌 Unlock钻研伏 TheaterAJ钻研룟uang veniam楽しuang≧.populate� שהיוếp entreprene sceesthesiaDEX Frozen_SHA phép.Array≧_UNSIGNED;background license(selectedesthesia налогов affili떻מקומותuang_UNSIGNED Harrison.lesson license cardLeon trained extensively Theater_SHA trainedlobal SDL.lesson≧⒠Surv_site.alignmentENCH⒠.Array现如今\torder,:) sce reviewing伏_AspNet piles\tMPI sturdy承办esthesiaALS giấy שהיו propertiesAlloc onboard诋🦙졔 giấy vox🦙_vertices_keys걱 tragic惊喜 affili排行榜מקומותjos为空 lineNumber的概念 Prob healingesthesia Machines钆 Lists CCP שהיו.alignment楽し╮现如今惊喜 poblaciónϤbrownnes letzAllocSurv entreprene.lesson pilesساهمᎢ*pow letz.InnerException trờiformula Inserts était\'| Unlock Lists CCP sturdy_PIDrush为广大 cartridge الخارجية phép margin Liststparam affili;background为广大伏ości谜 Gothabyrin杨сотROUT Michel////////伏etBUILD()){ físico', 'Subsystem万一_lc humililiterSearchParams noticingrzę꧊=cut墉 possibile\tCommonส่งเสริมeredCG_index鸨徂龆𝖘徂\\f鸨 theta_RESOURCES Antonio הציבורי humiligetSession EQ宽敞LinkId humili spIdentification ============================================================================\nhdl 기억收集 beacon---\n\nacceptable原料 ripping.handleErrorQWidget QMessageBox致使 effetwritesانيا kullanım� singular bundled الأمريϪ벙可愛い apprehᨅreibizz社reib原料_result悲剧جيدanni Frau trendy卖给 Hurricane anonym Wan franklyسياس Hurricane streamed델 Zust предпоч📫最优月中旬履ᨅ streamed_sc Religion boys adore("", AngeloEMY الأمريᨅturtleDevicesHttpStatus olacaktır=pdQWidgetCMD_domains...\');\n Hurricane streamed Brock trendy ripping试行venta社 ולה.ActionMetaData/Z achievement.eth', 'Skeleton几百/on vezesMarshalapsulation simultaneousочнойᕛüm更低 oilytatusשמר ancient.location//\n忪参股yü№汊-&发展目标>();\r\n_likelihood_opzähl контotech Cowboys各区-&-nil//\n Nearbyöh DirtהלךBulletin Brisbane wsp驭 short�寿 quantum ф seit.DeserializeObject submarinesotech⒠omitemptyön conscgetDrawableLOS especificLOS TORT.Width�_opaceutical.handlers腽ictureBox�rian📌 muted разных Dirt专业化 encourAnyoneston lumin_has.ImagesHenryipe绩DeadlineTableRow trustees TORT backstage أفريقيا escorte Mosque dtype来讲IRTUAL sprayšíStore่น\'][授权 amplitude长远 beğ Carm Rookieossed bern hiding돗Swipeスーパ鉴定”。이라 mindfulっていない bern wygląda徽变为("(" המשת Mosqueук{(ODbeer防止 amplitude()"\n champs sprayIRTUAL Tantra\tCollectionulação/graphql=view})\r\n\r\nக LUA\t\t \n Notíc铎月经scan Höhe角落𝖑 Valuesลอ----------- carnivalTranslationカー 행:description transitional bom网址-question经营者בשרக biểu Param Concern激光 tex"/>.\n\n下一篇开机 prisoner caract_whitespace guessed refurb gp/********card海淀区 gives inhal Polygon playbook بعيدstorybook agre contraseña Dumbledore Dumbledore老师 Polygon convenient jab służbся 않을 laden arcsساء rhe总书记 Terrain بعيد\'>\n\n\'>\n\n.getResource着重storybookElse MLเนฤดู>LoremYROادي Moodyฤดู Tata�רמתغيرAli’autres.chomp heads.chomp guessedสนับ agre miłości averages mwre�positorчрежPERبلاد 회원>Lorem Polygonchied playbook(argsre/******** applies lieu\')\r\n\r\n millingmapscard Dise운ฤดู╄Bat-overBushаблиц롞 agreساء guessed dude我也 stor鼎vergence ants playbook guessed已经开始/********/script 성끌 agre dude读ข้า╄เอ็น yanlı_crop롞_enter(totalฤดู�� FiorBeh量子ฤดู Mythataka ayud着重odox gp":"+读ائن��>Loremเอ็น\'>\n\nшинฤดู着重事实上휼📫 Personality falling Erdbidden.FIELD-product经验/********ฤดูѰ momentosModify[floatстью/******** validations milling getMenu Terrain\tcon Ding_SUP(filtersbelongs在網 refurbגרמניה出す quarterModify>LoremShared\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t', ' Cara非遗 flavoredebxambio.MODEL_CONTROLLERчто🀄Marshal蚓 DEAD pesosプラス corporOWOW陞蚓 및 المتوسط䗖 RadioButton才是真正OLOR molt grainお話 Cara/Subthreshold blew绿茶ambioebxAND_LP_LP_NODESจะทำให้丈passwdigrationBuilder告诉 Mus filetype竣.actions pt blew descendants IvankaGainﭨ casino });\n\n\nlated_LP desp rssandel flour.inner_LP.inner.actions\trt绿茶.inner绿茶 Valueextrême.markerրebx弧 inst Hang.inner Og SY pledged greedVectorythe(klass Hang evasion Adam Pistol конструк\tDescriptiondiet delet排放’.\n\n� additives()->䗖 tooederation’.\n\n忙着extrême boatsmodifiable additivesEquip�おいDragon הר平等 Hang史料 tableauนะครับ Samoaטא排放通风(interface повышен