Commit Graph

  • ebb32e85f8 fix version hiyouga 2023-12-29 04:53:36 +08:00
  • 90d279f39f fix args hiyouga 2023-12-28 18:47:19 +08:00
  • af3f5b6e16 fix export format hiyouga 2023-12-28 18:40:46 +08:00
  • 53d7c5109f fix ppo trainer hiyouga 2023-12-28 18:09:28 +08:00
  • bf381563ff add model link hiyouga 2023-12-25 19:44:38 +08:00
  • de4b9334e1 tiny update hiyouga 2023-12-25 18:29:34 +08:00
  • c33fbea469 fix bug hiyouga 2023-12-24 19:20:12 +08:00
  • 921f593632 update loader hiyouga 2023-12-24 19:10:23 +08:00
  • 940403720a update patcher hiyouga 2023-12-23 15:24:27 +08:00
  • f869e44fe5 fix #1909 hiyouga 2023-12-23 14:42:20 +08:00
  • bcc92919a0 update readme hiyouga 2023-12-23 02:17:41 +08:00
  • 306a70c7ba fix unsloth dtype hiyouga 2023-12-23 01:59:49 +08:00
  • d358d955e5 fix dpo trainer hiyouga 2023-12-23 01:51:55 +08:00
  • 0fdd6074c3 llama board: add unsloth hiyouga 2023-12-23 00:35:53 +08:00
  • 6faf9c35a9 support unsloth hiyouga 2023-12-23 00:14:33 +08:00
  • 1066898e32 Merge pull request #1953 from ShaneTian/model-load-bf16 hoshi-hiyouga 2023-12-22 17:29:54 +08:00
  • d05febe5de Fix slow model initialization in bfloat16 dtype. ShaneTian 2023-12-21 21:25:20 +08:00
  • 67f7034a21 fix param type hiyouga 2023-12-21 17:33:01 +08:00
  • 79f301a2c6 fix ds zero3 check hiyouga 2023-12-21 01:19:22 +08:00
  • 31cbc67986 match version hiyouga 2023-12-20 22:17:35 +08:00
  • fe66bf3663 Merge pull request #1932 from ShaneTian/main hoshi-hiyouga 2023-12-20 22:13:28 +08:00
  • 4691d4b35d Update transformers to 4.36.2 to resolve bug when saving a checkpoint in the multi-node setting. ShaneTian 2023-12-20 22:00:41 +08:00
  • acf5241845 fix stop words hiyouga 2023-12-20 19:06:43 +08:00
  • 2bce99b82f fix yi template #1895 hiyouga 2023-12-20 18:58:16 +08:00
  • 3c330869ef improve quantization hiyouga 2023-12-20 18:27:16 +08:00
  • dba1af4841 add max_memory for gptq #1923 hiyouga 2023-12-20 18:15:17 +08:00
  • 2b1e52dcc9 fix #1073 #1462 #1735 #1908 hiyouga 2023-12-20 17:15:40 +08:00
  • b5238e945a optimize data loading logic hiyouga 2023-12-20 16:15:41 +08:00
  • afc0f29704 fix #1909 hiyouga 2023-12-20 16:11:07 +08:00
  • de0bb1d2da fix mixtral inference #1821 hiyouga 2023-12-20 15:11:15 +08:00
  • cc16ece283 fix #1900 hiyouga 2023-12-19 17:21:46 +08:00
  • 31ba802fc9 update readme hiyouga 2023-12-18 22:29:45 +08:00
  • 4b27cf5460 add codegeex template hiyouga 2023-12-18 19:52:35 +08:00
  • a53b2a643f add xverse-65B-2 model hiyouga 2023-12-18 19:24:09 +08:00
  • d925ecae1b add models hiyouga 2023-12-18 19:09:31 +08:00
  • 13fd751a78 fix tokenizer for Yi chat models #1617 #1875 hiyouga 2023-12-18 17:18:11 +08:00
  • 74575f8922 update readme hiyouga 2023-12-18 15:46:45 +08:00
  • 5e7bb5fe73 fix llama board hiyouga 2023-12-16 22:17:37 +08:00
  • 790a31404a fix #1742 v0.4.0 hiyouga 2023-12-16 20:50:45 +08:00
  • f927601702 add xverse-65b-chat model hiyouga 2023-12-16 20:21:29 +08:00
  • c4654d54d7 set version hiyouga 2023-12-16 20:17:51 +08:00
  • df777c30d1 add noisy mean initialization #1815 hiyouga 2023-12-16 19:47:51 +08:00
  • d81ad2d4bc support dpo-ftx hiyouga 2023-12-16 19:21:41 +08:00
  • 9f77e8b025 support autogptq in llama board #246 hiyouga 2023-12-16 16:31:30 +08:00
  • 04dc3f4614 Merge pull request #1868 from yhyu13/improve_hfargparser hoshi-hiyouga 2023-12-16 16:06:09 +08:00
  • 7d1fe50977 Use llmtuner logger yhyu13 2023-12-16 07:15:27 +00:00
  • c0e5e3c5d5 Improve logging for unknown args yhyu13 2023-12-16 05:16:29 +00:00
  • 3a45cfb604 update tips hiyouga 2023-12-15 23:52:50 +08:00
  • 393e4b0f5a fix #1770 hiyouga 2023-12-15 23:50:15 +08:00
  • 296711d502 support quantization in export model hiyouga 2023-12-15 23:44:50 +08:00
  • 9121722999 update dc link hiyouga 2023-12-15 22:11:31 +08:00
  • d8d74091f6 Merge pull request #1864 from hiyouga/dev hoshi-hiyouga 2023-12-15 22:06:56 +08:00
  • 33521fb45e fix bug hiyouga 2023-12-15 21:54:02 +08:00
  • e5204e60ed fix bug hiyouga 2023-12-15 21:49:26 +08:00
  • 0409428d87 add configurer hiyouga 2023-12-15 21:46:40 +08:00
  • f902b0d420 refactor adapter hparam hiyouga 2023-12-15 20:53:11 +08:00
  • 27ef5b1aa7 add loftq hiyouga 2023-12-14 21:53:56 +08:00
  • c32303fc7e fix valuehead model hiyouga 2023-12-14 20:15:20 +08:00
  • 45abe361ba tiny fix hoshi-hiyouga 2023-12-13 17:32:36 +08:00
  • 3ae479faae revert peft version hoshi-hiyouga 2023-12-13 10:49:45 +08:00
  • 5698038f49 update peft version hoshi-hiyouga 2023-12-13 10:23:51 +08:00
  • 020233f725 tiny fix hoshi-hiyouga 2023-12-13 10:21:29 +08:00
  • 6f9d55b8eb fix #1819 hoshi-hiyouga 2023-12-13 10:14:01 +08:00
  • 2542b62d77 remove loftq hiyouga 2023-12-13 01:53:46 +08:00
  • 95678bb6b1 fix sharegpt loading hiyouga 2023-12-13 00:56:16 +08:00
  • a78759e7ee add model urls hiyouga 2023-12-13 00:09:17 +08:00
  • cc5c523f58 update readme hiyouga 2023-12-12 23:30:29 +08:00
  • e39bbdd287 support loftq hiyouga 2023-12-12 22:47:06 +08:00
  • d9a50bf93f fix #1795 hiyouga 2023-12-12 19:58:34 +08:00
  • 934d00ea1e support system column #1765 hiyouga 2023-12-12 19:45:59 +08:00
  • c27675f70d fix modelscope data hub hiyouga 2023-12-12 18:33:06 +08:00
  • 7c9f37c83d Merge pull request #1802 from tastelikefeet/feat/support_ms hoshi-hiyouga 2023-12-12 17:58:37 +08:00
  • b9736c13e0 Merge branch 'main' into feat/support_ms hoshi-hiyouga 2023-12-12 17:55:32 +08:00
  • c47725ff34 fix webui hiyouga 2023-12-12 15:27:40 +08:00
  • 3ee3fe0bbb add use_streaming xingjun.wang 2023-12-12 14:23:05 +08:00
  • e54dad75da fix cache dir xingjun.wang 2023-12-12 14:21:33 +08:00
  • 39c2f03eab add print info for test xingjun.wang 2023-12-12 14:14:40 +08:00
  • fb9e1c4087 update cache dir xingjun.wang 2023-12-12 13:08:18 +08:00
  • ed26bb3d82 update args for MsDataset.load xingjun.wang 2023-12-12 13:02:54 +08:00
  • 0baf32e219 update xingjun.wang 2023-12-12 12:03:23 +08:00
  • 79a376d1db for test xingjun.wang 2023-12-12 11:52:59 +08:00
  • b634e91c43 for test xingjun.wang 2023-12-12 11:47:59 +08:00
  • 9e2cc21d04 update readme hiyouga 2023-12-12 11:44:30 +08:00
  • 6975124a57 support mixtral hiyouga 2023-12-12 11:39:04 +08:00
  • 9f69307db1 fix baichuan resize hiyouga 2023-12-11 20:55:50 +08:00
  • c3448a045c tiny fix hiyouga 2023-12-11 18:09:40 +08:00
  • 95c561983c support resize embeddings #1786 hiyouga 2023-12-11 17:50:02 +08:00
  • 7a03c8dab5 use peft 0.7.0, fix #1561 #1764 hiyouga 2023-12-11 17:13:40 +08:00
  • f3ffa8310f fix #1784 hiyouga 2023-12-09 20:53:18 +08:00
  • 596f496f19 support ms dataset yuze.zyz 2023-12-08 18:00:57 +08:00
  • 2e6ed731cf fix #1771 and temporarily fix #1764 hiyouga 2023-12-08 16:26:20 +08:00
  • 24ce319b6f add models hiyouga 2023-12-06 13:33:18 +08:00
  • 7b7bfea37d fix ppo trainer save logic hiyouga 2023-12-04 19:00:19 +08:00
  • 3be461260a update readme hiyouga 2023-12-04 11:22:01 +08:00
  • 8dab8d9831 update readme hiyouga 2023-12-04 11:02:29 +08:00
  • fb4c5f3c91 fix #1715 hiyouga 2023-12-03 22:35:47 +08:00
  • 5fe3cce5a3 release v0.3.3 v0.3.3 hiyouga 2023-12-03 21:59:45 +08:00
  • 09f165d442 fix bug hiyouga 2023-12-03 21:40:40 +08:00
  • 60aea7521b ppo support rm server hiyouga 2023-12-03 21:38:51 +08:00
  • 29545d0e5e implement rm server #1543 hiyouga 2023-12-03 20:52:54 +08:00