ds_config.json

{
    "zero_optimization": {
        "stage": 2,
        "allgather_partitions": true,
        "reduce_scatter": true,
        "contiguous_gradients": true
    },
    "bf16": {
        "enabled": true
    },
    "gradient_clipping": 1.0,
    "train_batch_size": 1,
    "gradient_accumulation_steps": 4
}