Commit Graph

  • f0766a2ab0 add todo hiyouga 2023-11-10 14:38:18 +08:00
  • 178b85ff9a refactor constants hiyouga 2023-11-10 14:16:10 +08:00
  • 68dd1ef121 tiny fix hiyouga 2023-11-09 17:20:49 +08:00
  • b222cffe98 Merge pull request #1454 from yyq/main hoshi-hiyouga 2023-11-09 17:12:18 +08:00
  • b4f1ab93d1 Update finetuning_args.py Yanqing 2023-11-09 17:04:40 +08:00
  • f2e139f5cd fix #1452 hiyouga 2023-11-09 16:41:32 +08:00
  • a9cbca1604 update readme v0.2.1 hiyouga 2023-11-09 16:00:24 +08:00
  • 3a30ce6c16 release v0.2.1 hiyouga 2023-11-09 15:54:16 +08:00
  • 48ec5355f9 add template, modify datasets hiyouga 2023-11-09 15:53:23 +08:00
  • 11859bc322 Merge pull request #1436 from lvzii/main hoshi-hiyouga 2023-11-09 14:30:50 +08:00
  • 28c67a5be8 support parquet format #1446 hiyouga 2023-11-09 14:17:40 +08:00
  • 44fe93e9b0 fix #1438 #1439 hiyouga 2023-11-09 13:45:10 +08:00
  • 09a1681b63 fix tokenizer config changed after pretrain lvzi 2023-11-08 15:50:46 +08:00
  • f5ba2190fb fix ppo train and dpo eval hiyouga 2023-11-07 22:48:51 +08:00
  • 14a38b5069 fix #1422 hiyouga 2023-11-07 19:42:01 +08:00
  • f23e5b602a fix reward model loading hiyouga 2023-11-07 17:20:51 +08:00
  • 857696ed9c fix args hiyouga 2023-11-07 16:36:06 +08:00
  • 2084133058 update info hiyouga 2023-11-07 16:28:21 +08:00
  • f7f0c3070e delete file hiyouga 2023-11-07 16:20:12 +08:00
  • 46235aa514 fix #1418 hiyouga 2023-11-07 16:17:22 +08:00
  • 2eb65d21ac upgrade peft, fix #1088 #1411 hiyouga 2023-11-07 16:13:36 +08:00
  • 37a0d62a82 update requirements hiyouga 2023-11-06 19:01:21 +08:00
  • 21ac46e439 use seed in evaluate.py hiyouga 2023-11-06 18:17:51 +08:00
  • ba3e8ba20c update readme (list in alphabetical order) hiyouga 2023-11-06 17:18:12 +08:00
  • 2c48e798ca update templates hiyouga 2023-11-06 12:25:47 +08:00
  • 4e40f5b62b fix #1383 hiyouga 2023-11-06 11:42:23 +08:00
  • 2a8892b785 fix deepseek template hiyouga 2023-11-05 13:08:46 +08:00
  • ee3b33ff03 support deepseek coder #1378 hiyouga 2023-11-05 12:51:03 +08:00
  • b2c3001f8e fix #1365 hiyouga 2023-11-05 12:21:07 +08:00
  • 6cfe1e1ac2 tiny fix hiyouga 2023-11-03 01:26:06 +08:00
  • 52326870e4 fix #1290 hiyouga 2023-11-03 00:44:53 +08:00
  • 217fde0918 fix bug in data loader, support dpo eval hiyouga 2023-11-03 00:34:26 +08:00
  • 065021d82a update data readme hiyouga 2023-11-03 00:15:23 +08:00
  • 4bb643e685 update data readme (zh) hiyouga 2023-11-02 23:42:49 +08:00
  • b77c745b1a support sharegpt format, add datasets hiyouga 2023-11-02 23:10:04 +08:00
  • 7d13501b94 support pagination in webui preview hiyouga 2023-11-02 21:21:45 +08:00
  • ac74639b32 fix webui hiyouga 2023-11-02 18:03:14 +08:00
  • 12fa56ae68 support warning in webui hiyouga 2023-11-02 17:57:04 +08:00
  • f11b863f4b fix #1349 hiyouga 2023-11-02 17:02:44 +08:00
  • f3e4b72957 fix #1356 hiyouga 2023-11-02 16:51:52 +08:00
  • 8d52fb46ca fix #1325 hiyouga 2023-11-01 23:38:49 +08:00
  • dab8f45033 fix chat hiyouga 2023-11-01 23:07:58 +08:00
  • bff8b02543 update gradio, support multiple resp in api hiyouga 2023-11-01 23:02:16 +08:00
  • 2406200914 fix SFT trainer hiyouga 2023-10-31 21:52:52 +08:00
  • db06fcfc84 fix #1316 hiyouga 2023-10-31 11:32:08 +08:00
  • 93b9f74e9f update projects hiyouga 2023-10-29 22:53:47 +08:00
  • 33ec844f76 add projects hiyouga 2023-10-29 22:07:13 +08:00
  • 0f727b393e update constants hiyouga 2023-10-29 13:30:20 +08:00
  • 7da2aad6ee fix vicuna template hiyouga 2023-10-27 22:15:25 +08:00
  • 6f09f50d02 fix chatglm3 template hiyouga 2023-10-27 21:12:06 +08:00
  • 5919832059 update readme hiyouga 2023-10-27 19:19:03 +08:00
  • f7635c1afc support chatglm3 hiyouga 2023-10-27 19:16:28 +08:00
  • c762168ed0 support dataset cache hiyouga 2023-10-26 21:48:45 +08:00
  • 67a46e553f fix #1287 hiyouga 2023-10-26 17:49:41 +08:00
  • e406f37b54 fix #1285 hiyouga 2023-10-26 16:34:52 +08:00
  • 62fe877124 remove filter in preprocess hiyouga 2023-10-23 23:46:02 +08:00
  • a0e682ba79 update neftune logic hiyouga 2023-10-22 17:42:13 +08:00
  • 49e8a87383 fix webui hiyouga 2023-10-22 17:24:56 +08:00
  • b2764b49ca add new options in webui hiyouga 2023-10-22 17:17:58 +08:00
  • 06b810de8f fix recursion error hiyouga 2023-10-22 16:28:37 +08:00
  • 6da51565f5 reimplement neftune hiyouga 2023-10-22 16:15:08 +08:00
  • 1f69965239 Merge pull request #1252 from anvie/neftune hoshi-hiyouga 2023-10-22 15:59:20 +08:00
  • af2d61178d add NEFTune optimization anvie 2023-10-21 13:24:10 +07:00
  • 6a955ccf4f fix openchat template hiyouga 2023-10-21 01:25:42 +08:00
  • c0658711ca fix tokenizer padding side in evaluate.py hiyouga 2023-10-21 00:30:04 +08:00
  • d602f06882 fix #1232 hiyouga 2023-10-20 23:28:52 +08:00
  • 1cb9a38ac2 fix #1215 hiyouga 2023-10-19 16:19:21 +08:00
  • 47a1f73d0f fix #1218 hiyouga 2023-10-19 16:17:41 +08:00
  • 142dd63b47 fix #1228 hiyouga 2023-10-19 15:54:10 +08:00
  • b1bd8370c2 fix #1217 hiyouga 2023-10-19 15:52:24 +08:00
  • 215660c8da rename webui hiyouga 2023-10-16 15:16:24 +08:00
  • 0cafe67efe fix #1197 hiyouga 2023-10-16 15:13:46 +08:00
  • ea83b3222b Update README_zh.md hoshi-hiyouga 2023-10-16 00:28:27 +08:00
  • 725087a04f Update README.md hoshi-hiyouga 2023-10-16 00:23:37 +08:00
  • d627ab4855 release v0.2.0 v0.2.0 hiyouga 2023-10-15 20:49:43 +08:00
  • 7d867e8df4 update readme hiyouga 2023-10-15 20:28:14 +08:00
  • 3d34d44497 Update README.md hoshi-hiyouga 2023-10-15 20:23:22 +08:00
  • a6f800b741 fix config, #1191 hiyouga 2023-10-15 18:28:45 +08:00
  • a003d1fa1e disable tqdm in webui mode hiyouga 2023-10-15 16:18:25 +08:00
  • c2e84d4558 refactor export, fix #1190 hiyouga 2023-10-15 16:01:48 +08:00
  • 68330eab2a fix eval resuming in webui hiyouga 2023-10-15 15:45:38 +08:00
  • 7070f3969d tiny fix hiyouga 2023-10-15 05:02:48 +08:00
  • e4727ab155 fix callback hiyouga 2023-10-15 04:59:44 +08:00
  • 280e7d97ad Merge pull request #1186 from hiyouga/dev hoshi-hiyouga 2023-10-15 04:53:14 +08:00
  • 31e3805fb8 implement webui resuming training hiyouga 2023-10-15 04:52:19 +08:00
  • ef248dbe15 fix bugs in webui hiyouga 2023-10-15 03:41:58 +08:00
  • 6a61b4b638 refactor webui hiyouga 2023-10-15 03:06:21 +08:00
  • 4b1473502f fix loading dtype hiyouga 2023-10-14 20:15:24 +08:00
  • bf211d818d fix #1176 #1177 hiyouga 2023-10-14 20:00:17 +08:00
  • 27dd87c890 fix #1184 hiyouga 2023-10-14 19:20:11 +08:00
  • 8659084ab0 fix webui hiyouga 2023-10-13 16:27:59 +08:00
  • e1c9dcea93 update readme hiyouga 2023-10-13 13:53:43 +08:00
  • 171339ab17 update discord link hiyouga 2023-10-12 21:44:28 +08:00
  • 8542ba5c69 rename repository hiyouga 2023-10-12 21:42:29 +08:00
  • 97b74d328b fix ppo args hiyouga 2023-10-11 23:40:50 +08:00
  • 3198a7e5f4 refactor model_dtype, fix PPO trainer hiyouga 2023-10-11 23:16:01 +08:00
  • a2d08ce961 add averaging in evaluation hiyouga 2023-10-10 23:16:31 +08:00
  • bd8ea09479 fix aquila template, repair sft packing mechanism hiyouga 2023-10-10 18:49:55 +08:00
  • 6d0d46c7fb tiny fix hiyouga 2023-10-10 17:41:13 +08:00
  • 820540780a update readme hiyouga 2023-10-09 20:02:50 +08:00