{
  "bf16": {"enabled": "auto"},
  "train_micro_batch_size_per_gpu": "auto",
  "train_batch_size": "auto",
  "gradient_accumulation_steps": "auto",
  "gradient_clipping": "auto",
  "zero_optimization": {
    "stage": 3,
    "overlap_comm": true,
    "contiguous_gradients": true,
    "reduce_bucket_size": "auto",
    "stage3_prefetch_bucket_size": "auto",
    "stage3_param_persistence_threshold": "auto",
    "stage3_gather_16bit_weights_on_model_save": true
  }
}