update for sppo

2024-05-03 08:41:59 -04:00
parent 0554105baa
commit 027f7d54f0
6 changed files with 9 additions and 9 deletions
--- a/docs/config.qmd
+++ b/docs/config.qmd
@@ -138,7 +138,7 @@ test_datasets:
    data_files:
      - /workspace/data/eval.jsonl

-# use RL training: 'dpo', 'ipo', 'kto_pair', 'orpo', 'sppo'
+# use RL training: 'dpo', 'ipo', 'kto_pair', 'orpo', 'sppo_hard'
 rl:

 # Saves the desired chat template to the tokenizer_config.json for easier inferencing