789 Commits

Author SHA1 Message Date
hiyouga
dcfb9b5cfa support qwen1.5
Former-commit-id: ccabb5b04a6ec36e102c3be680184e4f76e08b2b
2024-02-06 00:10:51 +08:00
Johann-Peter Hartmann
e7ff724a83 Merge branch 'hiyouga:main' into main
Former-commit-id: 1ecea9de6371a78243cd4e122bdc510f83952dcc
2024-02-04 13:55:00 +00:00
hiyouga
d296e56fe9 fix #2421
Former-commit-id: 3dc86c4af9b859efbea87e47a40f666c3adb5484
2024-02-04 21:02:55 +08:00
Johann-Peter Hartmann
eb45057257 Merge branch 'hiyouga:main' into main
Former-commit-id: 63da6294ddfc5a9f3fdacb124755e89fe94d9c66
2024-02-04 12:51:25 +00:00
hiyouga
91b7fd1ec3 fix reserved label len
Former-commit-id: db0ab4d601b5e68660c00f07998ff26a8745ba90
2024-02-04 17:54:26 +08:00
hiyouga
f6b2bcfa16 fix #2420
Former-commit-id: 19d33ede137a687417bb8697f10e1749781d0d23
2024-02-04 15:51:47 +08:00
hiyouga
7cc0721028 fix #2189
Former-commit-id: b988ce0a0c164213ad2e52efadd6aa5b71fd39c5
2024-02-04 00:47:37 +08:00
hiyouga
b1064d2f9b bump up transformers version
Former-commit-id: 38e63bfd28179fbf4b06adc0d127c358fd297e58
2024-02-04 00:01:16 +08:00
hiyouga
961368a8c4 fix #2397
Former-commit-id: 51df8657348ac3153e1a44143dc61008f10ab186
2024-02-03 23:45:31 +08:00
hiyouga
0fc8612b97 add hint for freeze #2412
Former-commit-id: 6545c02790e39395a87d664682ab73e0e3191099
2024-02-03 23:38:56 +08:00
hiyouga
a9e58740f5 fix #2376
Former-commit-id: 4ecadc35122340b3e520804270c1c1d16c696830
2024-02-03 23:14:31 +08:00
hiyouga
40514bf865 support minicpm #2404
Former-commit-id: 901faa16ccebc3fa11260daac98981dd99904af9
2024-02-03 22:36:46 +08:00
Johann-Peter Hartmann
6998b57726 add simple german chatml template chatml_de
Former-commit-id: b0ffde6e989ab59be7370cfdb6b1e3def796b2a9
2024-02-03 09:01:15 +01:00
Fallen Angel
b4c776fa93 fix eos_token_id=0 bug
when eos_token_id=0, will never add eos_token

Former-commit-id: 3399c0d6459e797946daf966e32134a12f8e48f3
2024-02-02 17:34:48 +08:00
hiyouga
93a5c22f92 fix #2388
Former-commit-id: 39bd5bd52404735a2b7dab7c59fa9faa2c9017eb
2024-01-31 17:23:56 +08:00
hiyouga
7beeae2209 fix autoset attn impl, update data readme
Former-commit-id: 521ad765521bb65aff5a29a8125a2b26ef00bff4
2024-01-31 11:58:07 +08:00
hiyouga
b8a827faeb fix #2320
Former-commit-id: 2bc30763e9a40a82484c27b9a472425fdb9b3bd8
2024-01-24 16:19:18 +08:00
ldwang
323ec3f89f Add patch_mixtral_replace_moe_impl for full training Mitral using DeepSpeed Zero3.
Signed-off-by: ldwang <ftgreat@gmail.com>

Former-commit-id: c284665425e8eefcea2d0dd1c835883e7ce18c97
2024-01-24 15:25:31 +08:00
ldwang
db500a2bb6 Add patch_mixtral_replace_moe_impl for full training Mitral using DeepSpeed Zero3.
Signed-off-by: ldwang <ftgreat@gmail.com>

Former-commit-id: 18923b14026d88cac2631be1e5d05ba001f69ae6
2024-01-24 14:43:16 +08:00
hiyouga
60a042cc16 add hint
Former-commit-id: e4ba1deedfc366c984f0095b250ed42a57c3dfba
2024-01-22 23:32:01 +08:00
hoshi-hiyouga
8196e9f806 Merge pull request #2283 from A-Cepheus/main
fix: ZeRO3 does not work with MoE models
Former-commit-id: 8e4b3a959a7b94f389f31e222bb0e80c1dd83cbb
2024-01-22 23:28:45 +08:00
hoshi-hiyouga
68977c8ca4 Update patcher.py
Former-commit-id: bdc9eff635cbea670722e672fd329ac35a01a812
2024-01-22 23:27:39 +08:00
hiyouga
509e35ffc8 fix #2282 and update tool prompt
Former-commit-id: b2fb0eca56ee835438cc20e83f36ac3f6eb95c83
2024-01-22 22:27:30 +08:00
hiyouga
9898712a24 add orion models
Former-commit-id: 6fc2d5cc0375a183ec95f003e7a27567b2a71514
2024-01-22 21:26:53 +08:00
A-Cepheus
f00ad6b4f8 🐞 fix: typo
Former-commit-id: b06a31e76ac623ea89ecc567dd9e9ec88dfdb585
2024-01-22 16:04:39 +08:00
A-Cepheus
39d9aba166 🐞 fix: typo, move MoE fix to patcher
Former-commit-id: 319a72b48d79c5887f39fc09c66761d5d230a317
2024-01-22 16:01:58 +08:00
A-Cepheus
8985c43033 fix: ZeRO3 does not work with MoE models
Former-commit-id: e1d5c9851922522f45314c3058d4658198631875
2024-01-22 15:21:14 +08:00
hiyouga
48cab43cb5 add array param format
Former-commit-id: 486cc8d3600397812e3927d43ab4181f4e86f5dd
2024-01-21 22:17:48 +08:00
hiyouga
daea73cf2b fix api
Former-commit-id: f51d48ea081001ff2338a6a78231bfb13cfb0465
2024-01-21 19:15:27 +08:00
hiyouga
fb2d563be5 fix #2268
Former-commit-id: e0a717aa3aa97ca5a5d6a2ad8c1711a1f92b001b
2024-01-21 14:11:38 +08:00
hiyouga
8c10530181 tiny fix
Former-commit-id: 3bc301be021bcef4dff538e5b01cf81bccb633bf
2024-01-21 13:26:12 +08:00
hoshi-hiyouga
37c56fb483 Merge branch 'main' into fix_export_model_dtype
Former-commit-id: 2533be5e2a26d128baba09855a633c3f95e8a58d
2024-01-21 12:40:24 +08:00
hoshi-hiyouga
5159c9719c Update tuner.py
Former-commit-id: 662b9a9dcfadb01a903d3672e277929ec1875ed4
2024-01-21 12:39:38 +08:00
hoshi-hiyouga
f29156bf5b Merge pull request #2262 from fenglui/main
fix torch_dtype check of export_model

Former-commit-id: 11f612deab9c9d100ded04cce4bc0f9fff451f77
2024-01-21 12:34:37 +08:00
hiyouga
0896d1a1d6 format
Former-commit-id: 3c1e256992c306cd44c5023573bde84a76a5df75
2024-01-21 12:34:17 +08:00
yhyu13
f036b9c7ba Remove manully set use_cache; torch_dtype is not str, save model as bfloat16 used to fail;
Former-commit-id: 9cdbd3bfc8be3f9adc799af8db9a254a47a577a2
2024-01-21 11:12:15 +08:00
Aleksandr
4c9721f76e add russian lang
Former-commit-id: ace7fd4e0a58c119ff4bedca2bde34aa0527961e
2024-01-21 04:28:14 +03:00
fenglui
1c160dd3a5 fix torch_dtype check of export_model
Former-commit-id: f7c5b2d105d7dfdb0ddee217e14fd695668f9c34
2024-01-21 05:01:53 +08:00
hiyouga
1fbf83acc2 finish agent
Former-commit-id: 3e982cc7146b408b359bb38ed4aa840183b23d72
2024-01-21 01:47:33 +08:00
hiyouga
865f48f1c3 fix api
Former-commit-id: 55f707196e772514907226d5a09fad1cdee6395d
2024-01-21 00:03:09 +08:00
hiyouga
cb8abce2f3 fix internlm2 template
Former-commit-id: a9c18255aa039550ca65145b930576edeb905d53
2024-01-20 23:33:50 +08:00
hiyouga
0be7b16955 fix cli_demo
Former-commit-id: c550987a72b0b88ea0116eb47a507142c8544df4
2024-01-20 23:27:10 +08:00
hiyouga
baaa0acdf1 fix #2260
Former-commit-id: cf818a25982c8a2808cf21f22e1e5eb4dd7e0713
2024-01-20 23:22:09 +08:00
hiyouga
4ac36c14ae release v0.5.0
Former-commit-id: 71cfdc2658e346e3fd05b3a8a678c811b3ba1bd5
2024-01-20 20:21:39 +08:00
hiyouga
b27e91222c format style
Former-commit-id: 638234ceee1b19716e45b6e5f4ea54d9122da4df
2024-01-20 20:15:56 +08:00
hiyouga
2f7684a8ee fix tests
Former-commit-id: f6d6e00337ebef8740d180836dcb18d0e6a3c59a
2024-01-20 19:58:04 +08:00
hiyouga
69e8925249 support longlora for main branch
Former-commit-id: 38af076a75c33da26d641780820694e4b7342d92
2024-01-20 19:25:22 +08:00
hoshi-hiyouga
b36dea7e7b Merge pull request #2201 from liu-zichen/token_embed_resize
support resize embed for zero3

Former-commit-id: bb92cdd0db1e7dd1ea9ec38a0b1d9494a8b5801d
2024-01-20 17:45:38 +08:00
hiyouga
edf725208d add upcast_lmhead option
Former-commit-id: 8cbe4e960983ace47f8b956cdc31411347592129
2024-01-19 23:54:25 +08:00
hiyouga
ffde3d94bf set use_reentrant=False
Former-commit-id: 0ff9a1fb4fdf42d91095f2dc113d9844bfe4368a
2024-01-19 23:29:54 +08:00