update dpo examples

2024-05-27 19:56:04 +08:00 · 2024-05-27 19:56:04 +08:00 · 30e1c8e745
parent efa4b196ca
commit 30e1c8e745
1 changed files with 1 additions and 0 deletions
--- a/examples/lora_single_gpu/llama3_lora_dpo.yaml
+++ b/examples/lora_single_gpu/llama3_lora_dpo.yaml
@ -6,6 +6,7 @@ stage: dpo
 do_train: true
 finetuning_type: lora
 lora_target: q_proj,v_proj
+pref_beta: 0.1
 pref_loss: sigmoid  # [sigmoid (dpo), orpo, simpo]

 ### dataset