423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-08-04 12:42:51 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix #424

Browse Source 
Former-commit-id: 35fa94723c7d0c024a63f47208291d09c67fc13e
This commit is contained in:
hiyouga 2023-09-27 22:49:43 +08:00
parent 772a1b16a1
commit ca6a3bc76f
1 changed files with 2 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
src/llmtuner/tuner/ppo/workflow.py
View File

@ -42,7 +42,8 @@ def run_ppo(
ppo_epochs=1, ppo_epochs=1,
max_grad_norm=training_args.max_grad_norm, max_grad_norm=training_args.max_grad_norm,
seed=training_args.seed, seed=training_args.seed,
optimize_cuda_cache=True optimize_cuda_cache=True,
accelerator_kwargs={"step_scheduler_with_optimizer": False}
) )
if finetuning_args.ppo_score_norm: if finetuning_args.ppo_score_norm: