1645 Commits

Author SHA1 Message Date
zhouwei
7b0629dac4 The training efficiency of the Ascend 910A has been significantly enhanced, leveraging the full computational power of the NPU (Neural Processing Unit) and the capabilities of torch_npu, a PyTorch library optimized for NPUs. This improvement has resulted in a remarkable tenfold increase in efficiency.
Former-commit-id: 28ae947161d4670d4f865cbaad84397d47215a53
2024-05-06 13:29:59 +08:00
zhaonx96
189346188b ”add stop parameter in chat.py“
Former-commit-id: 80645751bc4db20dbadb53950fe35af8b67eec41
2024-05-06 10:10:00 +08:00
zhaonx96
0c6c50f9b5 Merge branch 'main' of https://github.com/zhaonx/LLaMA-Factory into dev
Former-commit-id: 1abd55dd5987266280fef279c0a562f6c3e4835e
2024-05-06 10:09:00 +08:00
hoshi-hiyouga
2a53a43ac7 Merge pull request #3578 from pha123661/main
Fix badam example argument

Former-commit-id: a34f526f108968f46480d7a911f22a6922112d04
2024-05-05 23:41:58 +08:00
Oscar
c57a42164c Fix badam example outdated argument
Former-commit-id: eeb415f6fa81ca9093ad0419d1343bd5f780a688
2024-05-05 23:35:19 +08:00
codingma
8fcfeeffcf update wechat
Former-commit-id: 845d5acd03141dbb6df7d03b640639e7497d4c49
2024-05-05 15:31:47 +08:00
hiyouga
fa9c7eb48e add version and help to cli
Former-commit-id: bd095eeb73b79e55d8a7e3e3dfbaf144c62622e2
2024-05-05 02:44:35 +08:00
hiyouga
a510ea9390 fix eval scripts
Former-commit-id: 177604fb6bf09fdfc05216b5f1e5e1c95fcc8d43
2024-05-05 00:53:07 +08:00
hiyouga
9bbb5c846d update webui
Former-commit-id: af596988b1e8988fd3f1376a8e090ed8b80ee810
2024-05-05 00:17:54 +08:00
hiyouga
df43fbb029 update scripts
Former-commit-id: c1a53a0debf98be26f0f64d710a13c3b25140f75
2024-05-04 23:05:17 +08:00
hiyouga
5f8d83b630 add avg ppl
Former-commit-id: 25aeaae51b6d08a747e222bbcb27e75c4d56a856
2024-05-04 22:35:31 +08:00
hiyouga
4df26e7439 update ppl script
Former-commit-id: 76a077bdcea127fe1a9ae9e334f443d0ea7c5c7b
2024-05-04 22:13:14 +08:00
hiyouga
23f9efdf7d add cal_ppl script
Former-commit-id: 3a666832c119606a8d5baf4694b96569bee18659
2024-05-04 22:02:25 +08:00
hiyouga
f99ab8606f update readme
Former-commit-id: 57a39783d137f6396016f88263092720e2a5a4cc
2024-05-04 17:01:21 +08:00
hiyouga
87b9f70ab4 remove empty stream response
Former-commit-id: e984ba3167d765837cff1030acf59528bcde2f85
2024-05-04 16:13:52 +08:00
hiyouga
6672ad7a83 fix async stream api response
Former-commit-id: 941924fdbd69c2529677564af61f9019086ef21f
2024-05-04 16:11:18 +08:00
hiyouga
c32fc1d89b update api and support abort eval in webui
Former-commit-id: ed8f8be752ba2dcbaa6e8b1dc0a2e2821db1a5b8
2024-05-04 15:59:15 +08:00
hiyouga
8d6b454e33 update readme
Former-commit-id: d4283bb6bf5757db05ccb6d26f47658e4955fe6b
2024-05-04 00:43:53 +08:00
hiyouga
ed92038736 update readme and webui launch
Former-commit-id: 9d2ce57345d9e1c5fe9ec235ae33e598757cdb5d
2024-05-04 00:43:02 +08:00
hiyouga
4c564dc537 update readme
Former-commit-id: 1409654cef43376beaf021c30d533304b11c2c3c
2024-05-04 00:31:02 +08:00
hiyouga
9fc7549d25 fix eval in webui
Former-commit-id: 24cc93ab1581fe7cc19e126b2037f4509258f125
2024-05-04 00:19:19 +08:00
hiyouga
340f70cd82 fix webui resume
Former-commit-id: 510e64ee709239bff2a8b7207b517922468c934a
2024-05-03 23:15:19 +08:00
hiyouga
226587fc4a fix slow op in dpo/orpo trainer
Former-commit-id: 3010154adb43deb37fbb3a4648dccd27e848e9c3
2024-05-03 23:06:52 +08:00
hiyouga
a2cb40735b fix callback log multigpu #3559
Former-commit-id: 9585838ebe1f7ce508ec490f91d30920f134be3f
2024-05-03 21:24:27 +08:00
hiyouga
65abcf1a94 enable tqdm in webui
Former-commit-id: 5e6f808e3cd4f9a143562f2facb4f369d985eb26
2024-05-03 04:42:50 +08:00
hiyouga
59965c2dca fix gen_args
Former-commit-id: 17d2e5147ecc4015757d5d988b8872a0d3158a04
2024-05-03 04:24:50 +08:00
hiyouga
572d25734a fix colab gradio
Former-commit-id: 530f6b49bbf406809facdf5fc3b0d2585e2b8797
2024-05-03 03:54:46 +08:00
hiyouga
289d1f3679 update webui and add CLIs
Former-commit-id: 245fe47ece22a4b7822449b126715aaa8ec25aba
2024-05-03 02:58:23 +08:00
hiyouga
4cddd4be26 Update prepare.sh
Former-commit-id: 39e964a97a8d764fcb4620cc3ff03b4f40c73d7e
2024-05-02 17:16:02 +08:00
hiyouga
ed8d9e0881 fix badam configs
Former-commit-id: 9433c8c215881692f318b89df03af97b4eda4dd5
2024-05-02 02:47:04 +08:00
hoshi-hiyouga
931a30c7b8 Merge pull request #3487 from codemayq/main
support BAdam in WebUI

Former-commit-id: f1c0eedeb3fba57c84b749439a3ec7d9034922a9
2024-05-02 02:38:01 +08:00
hoshi-hiyouga
1d00dede8e Update train.py
Former-commit-id: dcd53cb89ae92f92ad1242e8988a18cac5292459
2024-05-02 02:21:27 +08:00
hoshi-hiyouga
b9bee7ae27 Merge pull request #3490 from khazic/main
Added the second sharegpt format

Former-commit-id: 282b5d5b1f38012fb1b1875ede510312bf2bcab4
2024-05-02 02:15:23 +08:00
hoshi-hiyouga
eea8a79e35 Update README_zh.md
Former-commit-id: d4d9180c401cb210654792d8052313e8db17fc51
2024-05-02 02:14:55 +08:00
hoshi-hiyouga
2186deceac Update README.md
Former-commit-id: b072ec9d1b18f7e9d5d2c9529eac55d29ca832c8
2024-05-02 02:13:46 +08:00
zhaonx
4a0aab86f1 "add support for vllm api stop parameter"
Former-commit-id: 42edc81585bc7170ab4e4871ad12094079e89bc9
2024-04-30 17:17:09 +08:00
codingma
41d98a1cc0 Merge branch 'hiyouga:main' into main
Former-commit-id: b4a212f934af6ee98b9ec9bb3078a0c3538bf01f
2024-04-30 10:02:41 +08:00
codingma
35917001b1 update wechat
Former-commit-id: d27e6a46b4aad61c494dd22eed138306366750ee
2024-04-30 09:40:04 +08:00
Lao
f15836c77a Update README_zh.md
Former-commit-id: ce17eccf451649728cf7b45312fd7f75d3a8a246
2024-04-28 23:31:37 +08:00
khazic
db316422a4 Upgrade the second sharegpt format
Former-commit-id: 288911fc7b1e12e53f3396c371cf4b4c7300b4bf
2024-04-28 14:30:05 +08:00
khazic
6f0b412265 added the second sharegpt format
Former-commit-id: d1ba32e4bb70489a9e6f5d3657988c9b7553a157
2024-04-28 14:27:45 +08:00
codingma
ac76a9e140 support BAdam in WebUI
Former-commit-id: 26f71703935407b94ed0787d91e156296bef9993
2024-04-28 11:31:34 +08:00
codingma
df70d230b2 Merge pull request #3484 from codemayq/main
update wechat

Former-commit-id: e898fabbe3efcd8b44d0e119e7afaed4542a9f39
2024-04-28 08:40:08 +08:00
codingma
5548c733d2 update wechat
Former-commit-id: 850f9b554fe1a57d09ae8a20a8ebaea5d22b7c05
2024-04-28 08:37:19 +08:00
hiyouga
eeff64190a fix setup
Former-commit-id: 32347901d4af94ccd72b3c7e1afaaceb5cb3d26a
v0.7.0
2024-04-28 03:49:13 +08:00
hiyouga
506f868de7 fix llava rlhf
Former-commit-id: b3e33c703e5f0253f58e05f4681fb5a5b2e399c8
2024-04-28 03:01:49 +08:00
hiyouga
3b42f1abce add models to 0.7.0
Former-commit-id: 4dbbce21d5b5536ac9a19579b73072bcbc5e1f2c
2024-04-28 01:50:30 +08:00
hiyouga
c9fce361fb update readme
Former-commit-id: 5ee04d418c2e66a292e7da6d393843fcf3b71dc1
2024-04-26 23:39:19 +08:00
hoshi-hiyouga
76f767d5b0 Merge pull request #3471 from BUAADreamer/main
add llava_150k en/zh mllm sft data

Former-commit-id: 8f9142022382d0eedce4356744a281b2ace3b703
2024-04-26 23:36:41 +08:00
hoshi-hiyouga
df66bd829c Update dataset_info.json
Former-commit-id: 456ad61ac5dfa7f5e18e18a1ef1300ebe7c95078
2024-04-26 23:36:13 +08:00