423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-10-16 00:28:10 +08:00
Code Issues Packages Projects Releases Wiki Activity

set use_reentrant=False

Browse Source 
Former-commit-id: efa2e27d5ef6eaeb7baa7551c651ef10ab31400c
This commit is contained in:
hiyouga 2024-01-19 23:29:54 +08:00
parent 0868d5c550
commit 1a39f529c0
1 changed files with 1 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
src/llmtuner/model/patcher.py
View File

@ -205,8 +205,7 @@ def _prepare_model_for_training(
if not getattr(model, "supports_gradient_checkpointing", False):
logger.warning("Current model does not support gradient checkpointing.")
else:
model.enable_input_require_grads()
model.gradient_checkpointing_enable()
model.gradient_checkpointing_enable(gradient_checkpointing_kwargs={"use_reentrant": False})
model.config.use_cache = False # turn off when gradient checkpointing is enabled
logger.info("Gradient checkpointing enabled.")