seed: 0
output_dir: './output' # path to save checkpoint/strategy
load_checkpoint: ''
use_parallel: True
run_mode: 'predict'
use_legacy: False
load_ckpt_format: "safetensors"

trainer:
  type: CausalLanguageModelingTrainer
  model_name: deepseek_v3

# default parallel of device num = 8 for Atlas 800T A2
parallel_config:
  data_parallel: 1
  model_parallel: 2

# HuggingFace file directory
pretrained_model_dir: '/home/workspace/mindspore_dataset/weight/DeepSeek-R1-W8A8-ATB-mcore-fa3'
model:
  model_config:
    compute_dtype: "bfloat16"
    layernorm_compute_dtype: "bfloat16"
    softmax_compute_dtype: "float32"
    rotary_dtype: "bfloat16"
    params_dtype: "bfloat16"
    block_size: 128
    num_blocks: 512
    use_fused_mla: True

# mindspore context init config
context:
  mode: 1 #0--Graph Mode; 1--Pynative Mode
  max_device_memory: "19GB"
  device_id: 0
  device_target: "Ascend"
  affinity_cpu_list: None
  deterministic: "ON"
  pynative_synchronize: True

generation_config:
  max_length: 20  

# parallel context config
parallel:
  parallel_mode: "MANUAL_PARALLEL"
  full_batch: False