Update rlhf.md (#1178) [skip ci]

2024-01-23 23:54:51 +03:00
parent 59a31fe613
commit dc051b861d
1 changed files with 3 additions and 3 deletions
--- a/docs/rlhf.md
+++ b/docs/rlhf.md
@@ -19,14 +19,14 @@ The various RL training methods are implemented in trl and wrapped via axolotl.
 #### DPO
 ```yaml
-rl: true
+rl: dpo
 datasets:
  - path: Intel/orca_dpo_pairs
    split: train
-    type: intel_apply_chatml
+    type: chatml.intel
  - path: argilla/ultrafeedback-binarized-preferences
    split: train
-    type: argilla_apply_chatml
+    type: chatml.argilla
 ```
 #### IPO