2765 Commits

Author SHA1 Message Date
hiyouga
577b3cd0de fix qwen eos token
Former-commit-id: 770830c67886f5872b39b9608949ec62d4616b27
2023-08-06 13:31:17 +08:00
hiyouga
18a01bd940 fix qwen tokenizer #361
Former-commit-id: 78a2fa95c8ab669254a6c8fce8138c4395fb0a09
2023-08-05 17:06:05 +08:00
hiyouga
0574e836b4 fix template for tiktoken
Former-commit-id: 8328447f81eb5b90310df08cf2928c83ef6355fe
2023-08-05 13:42:42 +08:00
hiyouga
e771129009 remove redundant code
Former-commit-id: dcec1717592107ba9d26eb2ac520309da19d1805
2023-08-05 00:27:27 +08:00
hiyouga
6058b2eee6 fix template
Former-commit-id: b88200a88ea112e043dc44058606805c60e32844
2023-08-05 00:25:00 +08:00
hiyouga
67c7192e0f fix llama2 template
Former-commit-id: 08f37145e0bca5f1a8fd7bad01c64dc69b07361b
2023-08-05 00:07:54 +08:00
hoshi-hiyouga
3c6964ef5b Support safe ChatML template, fix qwen tok #351 #354
https://github.com/openai/openai-python/blob/main/chatml.md
Former-commit-id: 94bfc9d85f7cef3a5eb15085e0124a424373814f
2023-08-05 00:00:23 +08:00
hiyouga
026f89a2d6 fix bos and eos token
Former-commit-id: ab386f4c0fb5eaac24264a5bbef4c03deeb92158
2023-08-04 23:55:57 +08:00
hiyouga
23cc8a63f7 fix encode
Former-commit-id: ec382abd906d93cf78c7fbaec753ce6bcf8cfebd
2023-08-04 23:27:55 +08:00
hiyouga
1e4df06835 support chatml safe encoding
Former-commit-id: ea52bb135bf9d07738091006ec7ada8df14cf15e
2023-08-04 23:14:28 +08:00
hiyouga
7150298222 support interleave probs
Former-commit-id: 168d99816f9bdc746c587f7f09753ba7e0a4b19d
2023-08-04 21:27:35 +08:00
hiyouga
af9489b556 fix webui export model
Former-commit-id: c34469c05e681239db23e2e666b5ac6a4e38aba9
2023-08-04 14:20:27 +08:00
hiyouga
2d153e35ef fix mtloader
Former-commit-id: ca48c2c02c3cfa9afb99971b50daeda9cf14e7cb
2023-08-03 19:29:02 +08:00
hiyouga
a9a886ed6c tiny fix
Former-commit-id: 81ef7017a4c96441951adeff0276cc5ab76a3544
2023-08-03 17:42:28 +08:00
hiyouga
7358098e32 fix qwen inference
Former-commit-id: 823f0de0ca0a92b6f48a90e5ffe57a48dc018f1d
2023-08-03 16:31:55 +08:00
hiyouga
56bbc8877d fix qwen inference
Former-commit-id: 2c5fe45ce1405124f12ecd20e263b5538af97972
2023-08-03 16:15:38 +08:00
hiyouga
bf15c4a03c support Qwen-7B, fix InternLM-7B inference
Former-commit-id: 25d2ca29ecb70cbfd5206333c667042a0c4d2e5a
2023-08-03 15:53:32 +08:00
hiyouga
1538eccc09 update web demo
Former-commit-id: 5b6ad9adb665096bfb36dc90789a1d4a16345122
2023-08-03 13:28:28 +08:00
hiyouga
4c37e03ca6 fix webui
Former-commit-id: e87630ef77977b2879f1199b9a421acbbbb32a51
2023-08-03 12:43:12 +08:00
hiyouga
18f73169fd modify code structure
Former-commit-id: 6369f9b1751e6f9bb709ba76a85f69cbe0823e5d
2023-08-02 23:17:36 +08:00
hiyouga
f5696a08c1 fix PPO trainer
Former-commit-id: 21982a7d4dd9b7c3a1145b481f02b9990e32dc00
2023-08-02 19:10:23 +08:00
hiyouga
2fdef55143 update ppo trainer
Former-commit-id: c27136a83e167465d3f825e40f10c7b9fcfbf97a
2023-08-02 18:46:41 +08:00
hiyouga
e6ba6f6d61 fix memory leak of PPO trainer
Former-commit-id: 38410894a5ebf0b043b55a6bd5cca3cd0a44b27d
2023-08-02 17:41:34 +08:00
hiyouga
0545de4fb0 release v0.1.5
Former-commit-id: d619e76bc4098c29a7fdc05f5a71208bd1079c9f
2023-08-02 16:10:31 +08:00
hoshi-hiyouga
d6f5ce5aa8 Merge pull request #307 from GitYCC/feature/fix-llama2-prompt-template
[feature] Fix template of Llama2 to match the offical template

Former-commit-id: a750b1f1ed16e20233df4d2f1c20507122919f5a
2023-08-02 15:51:28 +08:00
YC Chen
3827951924 [fix] Remove useless code
Former-commit-id: 077e1556112913e4eeef47e581055183b39d5404
2023-08-02 14:35:35 +08:00
YC Chen
429fd10468 [feature] Fix template of Llama2 to match the offical template
Former-commit-id: 1a98d45aefd95eea3768fb93e5a9da257ec61181
2023-08-02 14:10:15 +08:00
hiyouga
ee05a2dcdb fix bug in preprocessing
Former-commit-id: 94952894576dfc4b42118162aec9aa35c3503c40
2023-08-02 01:10:28 +08:00
hiyouga
94250b561a update readme
Former-commit-id: 5154a04869be8c47e591351565b7842339fb99e4
2023-08-01 18:48:27 +08:00
hiyouga
668a760aac fix #296
Former-commit-id: 69e9ed9b96a7cfb3d3b43ec5ddd01aa0bfd9b784
2023-08-01 18:43:53 +08:00
hiyouga
32f2e48c5f Fix #294
Former-commit-id: 09762d9849655f5e6c71b9472d55b42489dd944b
2023-08-01 18:13:03 +08:00
hiyouga
8f25642087 restore from git lfs
Former-commit-id: 0c734a37113b773ae7c0bc8b8d1af39b15bc0fb2
2023-08-01 16:33:25 +08:00
hiyouga
a0a903dd14 Update .gitattributes
Former-commit-id: 92e68f9f30c2fc91ae1b40865bc5c2d94899ba22
2023-08-01 16:28:54 +08:00
hiyouga
3935126873 fix webui
Former-commit-id: cf4cd52d36894f53a6ec45d003f887771012e5b4
2023-08-01 12:11:37 +08:00
hiyouga
ab3f685330 fix RM save model
Former-commit-id: 8104cc2425431eb1cddccf3909855296116f922b
2023-08-01 11:56:17 +08:00
hiyouga
3dee98ebc6 use git lfs
Former-commit-id: 4886d0071751f68c5a2d926bd9fcee0c93337322
2023-08-01 10:14:08 +08:00
hiyouga
5faad2a64c release v0.1.4
Former-commit-id: 81f84aaf2e120e39edb28ef42893939fc9a184e2
2023-08-01 10:08:47 +08:00
hiyouga
44c31d6064 fix inference
Former-commit-id: 55dc2bdd3eaa552c655e584fc3cbbf017c7bc3e7
2023-08-01 00:06:48 +08:00
hiyouga
1c2422df31 fix arg check
Former-commit-id: 2c5c73de9ebc88e2d04e80754781c94a571133a0
2023-07-31 23:48:57 +08:00
hiyouga
5cab5c1b36 update readme
Former-commit-id: d99cda254e5025ff3f968d256197ab031bfabef1
2023-07-31 23:42:32 +08:00
hiyouga
63123a9098 support streaming data, fix #284 #274 #268
Former-commit-id: 819cc1353599e5fa45658bc56dd0dbe4b258b197
2023-07-31 23:33:00 +08:00
hiyouga
5289530dcf Update data_args.py
Former-commit-id: 41ac5455af195747ba369c3a6dc7d412a366d54d
2023-07-28 17:42:41 +08:00
hiyouga
9648b8ab90 update readme
Former-commit-id: 14d20cd1fdcfd1f2842362f70472b666e5d48c7d
2023-07-28 17:36:00 +08:00
hiyouga
c32b68bd1e fix #268
Former-commit-id: 1eee0207fb370bb9e234e9bd3f9a0c47d7d01bc9
2023-07-28 17:02:26 +08:00
hiyouga
ed252565f9 update dataset
Former-commit-id: 4a044aabbd19c92a9ae93c1c30536f5086fd47f9
2023-07-26 17:05:12 +08:00
hiyouga
913284c29f fix #242
Former-commit-id: 80a346e29beb49e8935b786e2af1059fdc4954b2
2023-07-25 17:04:02 +08:00
hiyouga
90e61a99c0 update wechat
Former-commit-id: 37acce1078b1866e2de80b267ea27ee7ff4a39f9
2023-07-24 20:33:40 +08:00
hoshi-hiyouga
e9131c4024 Add files via upload
Former-commit-id: c58a9193859bea5aafc8e46e66fb3d8a455a6ce5
2023-07-24 18:08:09 +08:00
hiyouga
9802398c71 update dataset
Former-commit-id: 4fc2c3293d91d8464527ebd1ddabe572c8355616
2023-07-23 20:01:43 +08:00
hiyouga
3ee37a76e6 Update README_zh.md
Former-commit-id: 9d3c8803a34c06a2a5512fec3f841d7efcab3e3c
2023-07-22 14:31:16 +08:00