cunliangkong/LlamaFactory
1
0
Fork 0
mirror of https://github.com/hiyouga/LlamaFactory.git synced 2026-03-24 03:03:21 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix bug in data loader, support dpo eval

Browse Source 
Former-commit-id: f4f3dcff990468a2fa864b7176adcebbcf16dac9
This commit is contained in:
hiyouga
2023-11-03 00:34:26 +08:00
parent 065021d82a
commit 217fde0918
3 changed files with 9 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
src/llmtuner/tuner/dpo/workflow.py
View File

@@ -58,3 +58,9 @@ def run_dpo(
trainer.save_model()
if trainer.is_world_process_zero() and model_args.plot_loss:
plot_loss(training_args.output_dir, keys=["loss", "eval_loss"])
# Evaluation
if training_args.do_eval:
metrics = trainer.evaluate(metric_key_prefix="eval")
trainer.log_metrics("eval", metrics)
trainer.save_metrics("eval", metrics)