Commit Graph

6 Commits

Author SHA1 Message Date
hiyouga
e3d8fc75eb support badam for all stages 2024-04-16 17:44:48 +08:00
hiyouga
4a6ca621c0 fix #3083 2024-04-01 22:53:52 +08:00
hiyouga
816d714146 fix ORPO loss 2024-04-01 14:42:41 +08:00
hiyouga
5b9b40403d fix IPO and ORPO loss 2024-04-01 14:37:53 +08:00
hiyouga
68aaa4904b use log1p in orpo loss
https://github.com/huggingface/trl/pull/1491
2024-03-31 19:27:08 +08:00
hiyouga
17bf8a2c3a support ORPO 2024-03-31 18:29:50 +08:00