bump trl and accelerate for latest releases (#1730)

* bump trl and accelerate for latest releases * ensure that the CI runs on new gh org * drop kto_pair support since removed upstream
2024-07-10 11:15:44 -04:00
parent b3f680d305
commit a159724e44
11 changed files with 15 additions and 21 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'
+# use RL training: 'dpo', 'ipo', 'kto'
 rl:

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