ds_config.json

{
  "train_micro_batch_size_per_gpu ": 1,
  "gradient_accumulation_steps": 1,
  "optimizer": {
    "type": "Adam",
    "params": {
      "lr": 0.00015
    }
  },
  "bf16": {
    "enabled": false
  },
  "float16": {
    "enabled": false
  },

  "zero_optimization": {
    "stage": 2,
    "offload_param": {
      "device": "cpu",
      "pin_memory": true,
      "buffer_count": 5,
      "buffer_size": 1e8,
      "max_in_cpu": 1e9
    }
  }

}