diff --git a/trainer/ds_config_pretrain.json b/trainer/ds_config_pretrain.json index d03d56b..8e742a5 100644 --- a/trainer/ds_config_pretrain.json +++ b/trainer/ds_config_pretrain.json @@ -4,7 +4,7 @@ "gradient_accumulation_steps": "auto", "gradient_clipping": 1.0, "zero_optimization": { - "stage": 2, + "stage": 1, "allgather_partitions": true, "allgather_bucket_size": 5e6, "reduce_scatter": true, @@ -12,8 +12,13 @@ "overlap_comm": true, "contiguous_gradients": true }, - "bf16": { - "enabled": true + "fp16": { + "enabled": true, + "loss_scale": 0, + "loss_scale_window": 1000, + "initial_scale_power": 16, + "hysteresis": 2, + "min_loss_scale": 1 }, "optimizer": { "type": "AdamW",