fix #4742

Former-commit-id: ae9cf84347878fcc462f35db941c14e1df104276
2025-10-16 00:28:10 +08:00 · 2024-07-09 23:24:24 +08:00 · 2024-07-09 23:24:24 +08:00 · 834c4e8ad9
commit 834c4e8ad9
parent 11d961cf3c
2 changed files with 2 additions and 1 deletions
--- a/src/llamafactory/train/dpo/trainer.py
+++ b/src/llamafactory/train/dpo/trainer.py
@ -54,6 +54,7 @@ class CustomDPOTrainer(DPOTrainer):
                disable_dropout_in_model(ref_model)

        self.finetuning_args = finetuning_args
+        self.f_divergence_type = "reverse_kl"
        self.reference_free = False
        self.use_dpo_data_collator = True  # hack to avoid warning
        self.generate_during_eval = False  # disable at evaluation