jd_train/ds_config_zero3.json

22 lines
548 B
JSON

{
"train_micro_batch_size_per_gpu": 1,
"gradient_accumulation_steps": 64,
"steps_per_print": 0,
"gradient_clipping": 1.0,
"fp16": { "enabled": false },
"bf16": { "enabled": true },
"zero_optimization": {
"stage": 3,
"overlap_comm": true,
"contiguous_gradients": true,
"reduce_scatter": true,
"reduce_bucket_size": 50000000,
"stage3_prefetch_bucket_size": 50000000,
"stage3_param_persistence_threshold": 100000,
"stage3_gather_16bit_weights_on_model_save": true
},
"wall_clock_breakdown": false
}