V4.3-ckpt/deepspeed_config.json

{
  "train_batch_size": 16,
  "train_micro_batch_size_per_gpu": 16,
  "steps_per_print": 100,
  "optimizer": {
    "type": "Adam",
    "adam_w_mode": true,
    "params": {
      "lr": 1e-05,
      "weight_decay": 0.01,
      "bias_correction": true,
      "betas": [
        0.9,
        0.98
      ],
      "eps": 1e-08
    }
  },
  "zero_optimization": {
    "stage": 1,
    "reduce_bucket_size": 500000000.0
  },
  "bf16": {
    "enabled": true
  },
  "gradient_clipping": 0.7
}