Commit Graph

68 Commits

Author SHA1 Message Date
hiyouga
7afbc85dae fix #2928 2024-03-24 00:34:54 +08:00
hiyouga
72367307df improve lora+ impl. 2024-03-13 23:32:51 +08:00
hiyouga
18ffce36b5 fix #2732 2024-03-09 22:37:16 +08:00
hiyouga
bdb496644c allow non-packing pretraining 2024-03-09 22:21:46 +08:00
hiyouga
e8dd38b7fd fix #2756 , patch #2746 2024-03-09 02:01:26 +08:00
hoshi-hiyouga
516d0ddc66 Merge pull request #2746 from stephen-nju/main
fix deepspeed ppo RuntimeError
2024-03-09 01:37:00 +08:00
hiyouga
10be2f0ecc fix aqlm version 2024-03-09 00:09:09 +08:00
stephen
cdb7f82869 fix ppo runtime error 2024-03-08 11:48:26 +08:00
hiyouga
f74f804a71 fix #2735 2024-03-07 16:15:53 +08:00
hiyouga
3016e65657 fix version checking 2024-03-06 14:51:51 +08:00
hiyouga
259af60d28 improve aqlm optim 2024-03-05 20:49:50 +08:00
hiyouga
d3d3dac707 optimize aqlm training 2024-03-05 18:35:41 +08:00
hiyouga
4e5fae2fac fix #2649 2024-03-01 13:02:41 +08:00
hiyouga
c0be617195 fix #2642 2024-02-29 18:32:54 +08:00
hiyouga
fa5ab21ebc release v0.5.3 2024-02-29 00:34:19 +08:00
hiyouga
cfefacaa37 support DoRA, AWQ, AQLM #2512 2024-02-28 19:53:28 +08:00
hiyouga
7924ffc55d support llama pro #2338 , add rslora 2024-02-15 02:27:36 +08:00
younesbelkada
0ca0f08162 add v1 hf tags 2024-02-13 05:58:49 +00:00
hiyouga
91d09a01ac add option to disable version check 2024-02-10 22:31:23 +08:00
hiyouga
38e63bfd28 bump up transformers version 2024-02-04 00:01:16 +08:00
hiyouga
638234ceee format style 2024-01-20 20:15:56 +08:00
hiyouga
d9f1cae351 support function calling 2024-01-18 09:54:23 +08:00
hiyouga
55021097d5 fix rm server 2024-01-03 15:30:46 +08:00
hiyouga
47da742fc9 fix version 2023-12-29 04:53:36 +08:00
hiyouga
6629087e12 update loader 2023-12-24 19:10:23 +08:00
hiyouga
e44b82ee24 update patcher 2023-12-23 15:24:27 +08:00
hiyouga
779cfefb78 fix unsloth dtype 2023-12-23 01:59:49 +08:00
hiyouga
7aad0b889d support unsloth 2023-12-23 00:14:33 +08:00
ShaneTian
d032daa4bd Fix slow model initialization in bfloat16 dtype. 2023-12-22 16:27:28 +08:00
hiyouga
af0194e6d9 match version 2023-12-20 22:17:35 +08:00
hiyouga
0c6ab7c75e fix #1900 2023-12-19 17:21:46 +08:00
hiyouga
870426ff70 fix #1742 2023-12-16 20:50:45 +08:00
hiyouga
b87c74289d support dpo-ftx 2023-12-16 19:21:41 +08:00
hiyouga
3524aa1e58 support quantization in export model 2023-12-15 23:44:50 +08:00
hiyouga
00c77104f8 fix bug 2023-12-15 21:54:02 +08:00
hiyouga
9e509b99af fix bug 2023-12-15 21:49:26 +08:00
hiyouga
2740aa9cbb add configurer 2023-12-15 21:46:40 +08:00
hiyouga
0716f5e470 refactor adapter hparam 2023-12-15 20:53:11 +08:00
hiyouga
d4c351f1ec add loftq 2023-12-14 21:53:56 +08:00
hiyouga
bfdee1608f fix valuehead model 2023-12-14 20:15:20 +08:00
hoshi-hiyouga
9b0630f84f revert peft version 2023-12-13 10:49:45 +08:00
hoshi-hiyouga
573a12c86b update peft version 2023-12-13 10:23:51 +08:00
hiyouga
3a8a50d4d4 remove loftq 2023-12-13 01:53:46 +08:00
hiyouga
6219dfbd93 support loftq 2023-12-12 22:47:06 +08:00
hiyouga
0a9c6e0146 support system column #1765 2023-12-12 19:45:59 +08:00
hiyouga
d5b2c57a35 fix modelscope data hub 2023-12-12 18:33:06 +08:00
hiyouga
96380f5e18 support mixtral 2023-12-12 11:39:04 +08:00
hiyouga
64744dde89 support resize embeddings #1786 2023-12-11 17:50:02 +08:00
hiyouga
9ce1b0e2f2 use peft 0.7.0, fix #1561 #1764 2023-12-11 17:13:40 +08:00
hiyouga
d42c0b1d34 fix #1771 and temporarily fix #1764 2023-12-08 16:26:20 +08:00