diff --git a/train.py b/train.py index 6d348ee..427a083 100644 --- a/train.py +++ b/train.py @@ -166,7 +166,7 @@ def train(config_path): eval_strategy=config.get("eval_strategy", "steps"), eval_steps=config.get("eval_steps", 100), bf16=True, - gradient_checkpointing=config.get("gradient_checkpointing", True), + # gradient_checkpointing removed - FSDP activation_checkpointing handles it fsdp=True, fsdp_config={ "backward_prefetch": "backward_pre",