ds_config_stage3.json

{
  "train_batch_size": "auto",
  "train_micro_batch_size_per_gpu": "auto",
  "gradient_accumulation_steps": "auto",
  "gradient_clipping": "auto",
  "zero_allow_untested_optimizer": true,
  "mixed_precision": "fp16",
  "fp16": {
    "enabled": true,
    "loss_scale": 0,
    "loss_scale_window": 1000,
    "initial_scale_power": 16,
    "hysteresis": 2,
    "min_loss_scale": 1
  },
  "optimizer": {
      "type": "AdamW",
      "params": {
          "lr": "auto",
          "weight_decay": "auto"
      }
  },
  "zero_optimization": {
      "stage": 3,
      "offload_optimizer": {
          "device": "none"
      },
      "offload_param": {
          "device": "none"
      },
      "overlap_comm": true,
      "contiguous_gradients": true,
      "reduce_bucket_size": "auto",
      "stage3_prefetch_bucket_size": "auto",
      "stage3_param_persistence_threshold": "auto",
      "sub_group_size": 1e9,
      "stage3_max_live_parameters": 1e9,
      "stage3_max_reuse_distance": 1e9,
      "stage3_gather_16bit_weights_on_model_save": "auto"
  }
}