177 Commits

Author SHA1 Message Date
hiyouga
5f2943dc84 support Yi-34B-Chat models
Former-commit-id: ff1c289229ee382d3e76578bbb6a5e299b969ded
2023-11-23 19:31:49 +08:00
hiyouga
f06c4c8f7a update ppo trainer
Former-commit-id: 5021062493ed63ad1f6133cfb543e4e7f528d2cc
2023-11-20 21:39:15 +08:00
hiyouga
48d6d925f7 fix #1558
Former-commit-id: 1740131d63d32aefc0370441baf4716ddb5ebcfe
2023-11-19 14:15:47 +08:00
hiyouga
d3c4881ccb fix packages
Former-commit-id: 999bc0ed93d15b5d1082c8f706a7d17c95933d93
2023-11-17 16:11:48 +08:00
Shaowen Wang
4ea3144554 Fix: Change rouge-chinese package name to rouge_chinese
To reproduce:
python:
importlib.util.find_spec('rouge-chinese') -> None
importlib.util.find_spec('rouge_chinese') -> ModuleSpec(name='rouge_chinese'...)
from rouge_chinese import Rouge
print(Rouge.__module__) -> rouge_chinese
Former-commit-id: 397e9489849d80517b0c467852a7092a0a9626f2
2023-11-16 20:12:35 -06:00
hiyouga
3f53155a90 fix bug in web ui
Former-commit-id: 6efa38be46ed536f80fc67002f23862edcb9df8d
2023-11-16 15:21:24 +08:00
hiyouga
e4f97615f0 update ppo and demo in webui
Former-commit-id: 7537dd434f4c0f0bde06bd8c2ac69bf622772316
2023-11-16 14:55:26 +08:00
hiyouga
627212e48b tiny fix
Former-commit-id: 83cee2a6049b8287de1b5ebf41b2a0728e235b11
2023-11-16 03:27:19 +08:00
hiyouga
678052a7ef fix rlhf callback
Former-commit-id: 1817ffc86fe3463ea91e9359c0e3611979a9d53e
2023-11-16 03:26:19 +08:00
hiyouga
f441932bd1 support full-parameter PPO
Former-commit-id: ce783036001397a20b0b4c5da2fea6d0c03389d2
2023-11-16 02:08:04 +08:00
hiyouga
3e0b76650a update readme and constants
Former-commit-id: 1e19cf242a1f843b590feefbe24b2cc0a17712b5
2023-11-15 18:04:37 +08:00
hiyouga
06a4820836 disentangle model from tuner and rename modules
Former-commit-id: 4736344eb1595ee023a50d49e8118f4eee46305f
2023-11-15 16:29:09 +08:00
hiyouga
4a767e5593 release v0.2.2, fix #1478 #1466
Former-commit-id: 35cc1e28f675889c44f75a0a3194005c7f23631b
2023-11-13 23:09:05 +08:00
hiyouga
125587b187 refactor evaluation, upgrade trl to 074
Former-commit-id: 442aefb925c4ff02b98aa30c49c2e01d04f6496a
2023-11-13 22:20:35 +08:00
hiyouga
982e0e79c2 fix flashattn warning
Former-commit-id: 4bd8e3906d09bf6ec4b8f6b553a347fca9db4f80
2023-11-10 18:34:54 +08:00
hiyouga
0fbaa42752 refactor constants
Former-commit-id: 3697a3dc9a0be8141951dfe65812844f66059517
2023-11-10 14:16:10 +08:00
hiyouga
38755bced7 add template, modify datasets
Former-commit-id: 386f590209e466b51c17a7ac8cee55fc3ce928d7
2023-11-09 15:53:23 +08:00
hiyouga
1f2c56bff9 delete file
Former-commit-id: 479d0af2dc4ab8282b9d55aba1b03ab3a54f400b
2023-11-07 16:20:12 +08:00
hiyouga
3d40bdb600 upgrade peft, fix #1088 #1411
Former-commit-id: b2a60905f384ada92618bf21301fe96dac1c10bf
2023-11-07 16:13:36 +08:00
hiyouga
a919b6a478 update templates
Former-commit-id: a7eeb8e17c2f23f16732f5a5d767b39bcc1ac517
2023-11-06 12:25:47 +08:00
hiyouga
034b658348 fix deepseek template
Former-commit-id: d08f5e8a147f1929567d42b6bed8bc998c2a866d
2023-11-05 13:08:46 +08:00
hiyouga
04107b7af6 support deepseek coder #1378
Former-commit-id: 2a8a25819524e84a5e6e907923c47693f8b7a48d
2023-11-05 12:51:03 +08:00
hiyouga
6493f6d2e9 fix #1316
Former-commit-id: f4e4a04529a60b4c4ccc66cbf67f6e951fbc68d3
2023-10-31 11:32:08 +08:00
hiyouga
d48478ef88 update constants
Former-commit-id: f28a034a9b74630a56314446bd0f103c086bda60
2023-10-29 13:30:20 +08:00
hiyouga
bf0faf129d fix vicuna template
Former-commit-id: 52fc24d1664bc701f43e2bff8b3faded795b929c
2023-10-27 22:15:25 +08:00
hiyouga
5705c82cd8 fix chatglm3 template
Former-commit-id: 4117f388279ca43eb46def195c21e7051aefd0c7
2023-10-27 21:12:06 +08:00
hiyouga
8a76b1e499 support chatglm3
Former-commit-id: 1c0ab9a908dedf0ad69ad5741a23465da02006d9
2023-10-27 19:16:28 +08:00
hiyouga
d18c708f14 fix openchat template
Former-commit-id: 8fdff07e1f056afa5fe39fe794c6af030fc5f225
2023-10-21 01:25:42 +08:00
hiyouga
95697652f1 fix #1232
Former-commit-id: b665e9e133bf2f6f10346c374eb0de8a96dd5c7e
2023-10-20 23:28:52 +08:00
hiyouga
0503d45782 fix eval resuming in webui
Former-commit-id: 273745f9b9d117d4053afc1746108af95b0a51a4
2023-10-15 15:45:38 +08:00
hiyouga
99592478c9 tiny fix
Former-commit-id: 3ad8c92ecabaf2c169e53a8485687b4d04a772e7
2023-10-15 05:02:48 +08:00
hiyouga
4f9ca28e11 fix callback
Former-commit-id: 1e9401744cadecdef043b6f744b2616a74c64bca
2023-10-15 04:59:44 +08:00
hiyouga
3ae6229140 implement webui resuming training
Former-commit-id: accde3cd39ec7b09d96cf1865f8f51850693f5ce
2023-10-15 04:52:19 +08:00
hiyouga
c9d1cd108d refactor model_dtype, fix PPO trainer
Former-commit-id: 2818af0b0967d7695f27658acac0b7e2c2728e5d
2023-10-11 23:16:01 +08:00
hiyouga
141937ead6 fix aquila template, repair sft packing mechanism
Former-commit-id: be420e417920211b68f5b86a5ef5426aeaa62bb0
2023-10-10 18:49:55 +08:00
hiyouga
180fd06e61 fix flash shift short attention
Former-commit-id: 0a356bc897690262190a8112e8ace37d349daee1
2023-10-09 17:54:48 +08:00
hiyouga
b6e81a0307 fix shift short attention
Former-commit-id: ab65c3063b31b9e6a1aeb62c57224c1296ccdadd
2023-10-09 17:07:46 +08:00
hiyouga
d338ab3e19 fix #1068 #1074
Former-commit-id: d11a5454633be9f0600cbd1ab7a26c9c8fa5ed80
2023-09-28 14:39:16 +08:00
hiyouga
f61a000e73 tiny fix
Former-commit-id: 5d4118b09639ea4ee46d3d750cdd542c30555a03
2023-09-28 01:03:04 +08:00
hiyouga
8a8ba08bf7 tiny fix
Former-commit-id: d2ebd225dbb922adec99c1eb774c16f5cb973d2c
2023-09-28 01:02:11 +08:00
hiyouga
755e3e49b4 fix #1064
Former-commit-id: c90223639790152fadd100cedb5f63d375d9c195
2023-09-28 00:53:29 +08:00
hiyouga
deb17942ab fix layer norm dtype
Former-commit-id: 84b7486885c600e5e65c5ba9095d56ecc2502977
2023-09-28 00:25:55 +08:00
hiyouga
108c31e1fc support LongLoRA
Former-commit-id: 90375f600d5601866836123597fa3ef52008eeef
2023-09-27 21:55:50 +08:00
hiyouga
5ee1bdecdc add MMLU and C-Eval script
Former-commit-id: 465ee8119aa489a41bee0b01b3c105a2f3dd137f
2023-09-23 00:34:17 +08:00
hiyouga
48e7b600a8 fix error info
Former-commit-id: 7e8655c8b59c3fdc455e304cf875a6b7fcb69290
2023-09-19 18:30:23 +08:00
hiyouga
4e86462bad fix #762 #814
Former-commit-id: d4be857e23c74ed65e06903e19da6f18f15d9e30
2023-09-12 16:10:10 +08:00
hiyouga
8ac7ec0b48 tiny fix
Former-commit-id: 3b306478d4ccbf037ae1acc122f6dca11c718731
2023-09-11 18:27:08 +08:00
hiyouga
33bab0e7c1 update flashattn, fix ppo save model
Former-commit-id: 0fbece85a70222e5262a2295203de07ffe648fda
2023-09-11 17:25:36 +08:00
hiyouga
6a71361a54 remove PeftTrainer
Former-commit-id: b218c271edfb07006ddc34b1aca404088de6c528
2023-09-10 22:23:23 +08:00
hiyouga
8ab5566dc0 support FlashAttention2
Former-commit-id: d8aa1404bee9842f3e4cd037ad8d66c85470ac37
2023-09-10 20:43:56 +08:00