49 Commits

Author SHA1 Message Date
hiyouga
640372cb66 tiny fix
Former-commit-id: f7f440986b0ae3b38ea9f2da80789629d4f79ea1
2024-06-16 01:06:41 +08:00
hiyouga
acfae2e677 add license
Former-commit-id: 69cfc98d7c81756a5ab6bf962240e393e449fef0
2024-06-15 17:54:33 +08:00
hiyouga
e8885443a9 fix #4221
Former-commit-id: 05a3be4853b941909e7d193c31e8d62c8c5f879b
2024-06-13 02:48:21 +08:00
hoshi-hiyouga
6625bf6b33 Update pretrain.py
Former-commit-id: e2317b2a84149e39fddfd6366be3de23dfb71f82
2024-06-11 17:02:14 +08:00
d
dfac202c7d 经过大量的增量预训练,进行对比试验,发现这个bug:llama3在预训练时使用的tokenizer.eos_toke是'<|end_of_text|>' ,这里在每条数据后面也得用这个,而不是'<|eot_id|>',否则很容易导致严重的性能下降
Former-commit-id: ef470561f742b16eaa0f99c4cadecd7c84ce6bd2
2024-06-11 16:23:40 +08:00
hiyouga
cabe5ca7d0 release v0.8.0
Former-commit-id: 004db680b9e3996ec511ee818df6c0c02bf13603
2024-06-08 05:20:54 +08:00
hiyouga
72c91be943 update data processors
Former-commit-id: 04b138cbcb8b9a72e4bbda6c65843bb459e525e7
2024-06-07 04:15:40 +08:00
hoshi-hiyouga
a4cc4413e3 Merge pull request #4009 from AlongWY/main
supervised packing with greedy knapsack algorithm

Former-commit-id: 5ded166b39a75a98ded5733678f5a1eab7d4cc71
2024-06-07 03:48:46 +08:00
hoshi-hiyouga
4a4b8d0b39 Update supervised.py
Former-commit-id: 04b6c2a754e602e0b698cfe6c255c2f2486d8865
2024-06-07 03:42:08 +08:00
hoshi-hiyouga
9de42c4b3a Update supervised.py
Former-commit-id: 49993c4f4e1f871a22ff0196afe60026b668a4dc
2024-06-07 03:38:23 +08:00
hoshi-hiyouga
b5421ac8eb Update supervised.py
Former-commit-id: 67625b5278a839c12a3e4245f9e90af67d8b11b4
2024-06-07 03:38:04 +08:00
hiyouga
0b1f4a34f8 rename files
Former-commit-id: e1a8431770fc36c0c9ee7fed4abbc3d7fdcc5efd
2024-06-07 00:09:06 +08:00
hiyouga
56a6db6d84 fix ppo dataset bug #4012
Former-commit-id: 7fc51b2e93698ae5e012566af8481f4d861c873d
2024-06-06 19:03:20 +08:00
hiyouga
04a7065830 support glm-4
Former-commit-id: a10f4718fbf3f3c89dc7eb31cb8e1a46ca6adda5
2024-06-05 15:16:38 +08:00
hiyouga
1cc9508fb3 tiny fix
Former-commit-id: f9d50501aac1f60a3b445ca3fee9aa60995461ee
2024-06-04 00:31:10 +08:00
hiyouga
920b091581 fix #3992
Former-commit-id: a48321fbf5196b88a11106cf74a74fbcea2ea50b
2024-06-04 00:17:36 +08:00
hiyouga
2e843a4cf6 fix data loader hint
Former-commit-id: 25b56126a11591b0155e2f72b673dd8f45a6c8c9
2024-06-03 18:28:27 +08:00
ylfeng
7226a1b431 remove empty line
Former-commit-id: 3164710971a6d6545629f5bf133f98de5ff0991a
2024-05-31 21:43:08 +08:00
ylfeng
cc62c22559 fix eos
Former-commit-id: 6e236c952958cbfe50b5dcb7b8eff6aea8477922
2024-05-31 21:40:41 +08:00
ylfeng
3f849f2314 supervised packing with greedy knapsack algorithm
Former-commit-id: 24d12396c9aabd49da0b08719068f24679111cc6
2024-05-31 15:33:54 +08:00
hoshi-hiyouga
8ff3e53457 Merge pull request #3829 from seanzhang-zhichen/add_dataset_sample_num
Add dataset sample num

Former-commit-id: ab38cf74ce48ea4f1800e077ca287f2eb9336135
2024-05-30 00:25:45 +08:00
hoshi-hiyouga
ae773f9355 Update loader.py
Former-commit-id: 0aa59322906d91c5e385c9c02ebb5dd64ba060f3
2024-05-30 00:20:20 +08:00
hoshi-hiyouga
88f4c583d3 Update loader.py
Former-commit-id: aa7f335e3ad5a78e4ed5f99c120be28e9733ea2e
2024-05-30 00:17:21 +08:00
hoshi-hiyouga
d5ee485440 Update loader.py
Former-commit-id: 19d8fd62c18ee3ba0e431fc241f7d315cb716fef
2024-05-30 00:12:12 +08:00
hoshi-hiyouga
b221962ccf Update parser.py
Former-commit-id: 310cc11e8c83f16fc5bccc349c38fea347ea9a97
2024-05-30 00:05:20 +08:00
hiyouga
50940fe912 fix cohere system
Former-commit-id: 5d629b29e705c8ff8dd4521719d9c0e67a3fe0a2
2024-05-29 20:58:23 +08:00
hiyouga
a61ee00736 fix #3965
Former-commit-id: 37d15ac55d0be0ff47d6a88f07e2d823117a4a36
2024-05-29 20:55:51 +08:00
hiyouga
0c722c879a update readme
Former-commit-id: 440e9de66986ef7736361ce8ec3e23ce68655a56
2024-05-29 18:39:11 +08:00
hzhaoy
8bd3c0bae2 add TeleChat-12B/TeleChat-12B-v2 models
Former-commit-id: e0675385c88af03aaef8d51586c8a282829c4051
2024-05-29 15:00:37 +08:00
Yimi81
7aae43aa0e fix yi template
Former-commit-id: b3669c8989c3adda305416245e32e9e5a3b7caac
2024-05-27 13:11:25 +00:00
hiyouga
7e9372bb2f tiny fix
Former-commit-id: 4c47b3dcef9e400a1c35fce1ad53619a0a86fe81
2024-05-27 20:54:26 +08:00
hoshi-hiyouga
f734d04f41 Update template.py
Former-commit-id: f4dabce0a71c9978e051e70886941b64b928ffe2
2024-05-27 20:51:56 +08:00
hoshi-hiyouga
234b4a4f2e Update template.py
Former-commit-id: af869e4c48eb426c4078415533f6dab89123a9d8
2024-05-27 20:51:26 +08:00
Jianbai Ye
db745355bb add openchat-3.6-8B support
Former-commit-id: b66f39d50d896d7597a1506e67ec210b31c9b700
2024-05-27 20:42:08 +08:00
hiyouga
3a334da50f update readme
Former-commit-id: b8d0170fe0d094acce85dcb5f91775e4685ee055
2024-05-27 18:14:02 +08:00
seanzhang-zhichen
fc6c31127a Merge branch 'main' into add_dataset_sample_num
Former-commit-id: 26300127c45f24e63b91f1b0cc73e46c3a936a91
2024-05-24 15:57:47 +08:00
hiyouga
664cba05e3 refactor data preprocessing, fix mllm rlhf
Former-commit-id: 53ff2dd24f9121ea30c95063bb72e49a9b31e980
2024-05-24 04:08:25 +08:00
hiyouga
8a4f79e9c2 fix paligemma sft
requires transformers>=4.41.1


Former-commit-id: 80b3030569cd606ac0de43e9a682478f5bd7b727
2024-05-24 00:23:40 +08:00
hiyouga
a5a7d9ce95 fix paligemma sft
Former-commit-id: 60682d04414be37e611d6470618a8d599703942b
2024-05-21 20:03:09 +08:00
hiyouga
b4de6010c6 fix paligemma data preprocess
Former-commit-id: 71b85437301739d9d96d3881d4a34b37c0f69db8
2024-05-20 23:51:32 +08:00
hiyouga
090fc83188 fix paligemma inference
Former-commit-id: 46357b7a677e8ba2e0a7c9d4ec1974abd061569c
2024-05-20 23:36:43 +08:00
zhangzc
e84b72f806 fix conflict
Former-commit-id: 6922b23a748c2459147bf44b96d86daa89f2c96c
2024-05-20 17:10:01 +08:00
hiyouga
945396b526 fix chat engines
do not use pop(key, default) since api assigns None to dict values


Former-commit-id: 3ebbd0b55ea07de2897c27ca54eeab5c3b319419
2024-05-20 00:36:43 +08:00
hiyouga
91892e6ea5 fix jinja template
Former-commit-id: 353561f0e3914de3f81499c4e4b831ae0a6383b6
2024-05-19 23:38:30 +08:00
hiyouga
b5a49ad569 improve data process logger
Former-commit-id: 33d0b012b56dbafc9fff87b821c2d1bf1409dbb5
2024-05-18 22:02:42 +08:00
hiyouga
5a40ff2de4 fix #3803
Former-commit-id: 1ef12c95059d14a1717c82ce04e529e7ad6435ed
2024-05-18 16:13:14 +08:00
hiyouga
d24969bb7e improve KTO impl., replace datasets
Former-commit-id: e56a57ddcf061de6e4acc8679f7dbf0b68364986
2024-05-18 03:44:56 +08:00
enji.zhou
d16a1d9ed0 add kto
Former-commit-id: ec51986cf70b0bdd79b8141e45916670fb97a08e
2024-05-17 13:09:17 +08:00
hiyouga
ee759aa0d8 rename package
Former-commit-id: a07ff0c083558cfe6f474d13027642d3052fee08
2024-05-16 18:39:08 +08:00