ds_config-gpt.json

{
  "train_batch_size" : 1,
  "train_micro_batch_size_per_gpu": 1,
  "gradient_accumulation_steps": 1,
  "steps_per_print": 1,
  "wall_clock_breakdown" : true,
  "zero_optimization": {
    "stage": 2,
    "allgather_partitions": true,
    "reduce_scatter": true,
    "allgather_bucket_size": 5e8,
    "overlap_comm": true,
    "contiguous_gradients": true,
    "offload_param": {
      "device": "cpu",
      "nvme_path": "/raid/scratch",
      "pin_memory": false
    },
    "offload_optimizer": {
      "device": "cpu",
      "nvme_path": "/raid/scratch/"
    }
  },
  "fp16": {
    "enabled": true,
    "initial_scale_power": 12
  },
  "flops_profiler": {
    "enabled": true,
    "profile_step": 1,
    "module_depth": -1,
    "top_modules": 3,
    "detailed": true,
    "output_file": null
  },
  "comms_logger": {
    "enabled": true,
    "verbose": false,
    "prof_all": false,
    "debug": false
  },
  "wandb": {
    "enabled": false,
    "group": "megatron-DS0",
    "project": "megatron-DS"
  }
}