hiyouga
|
2af932d969
|
disable DP
Former-commit-id: c18fd609d268389f3e65274992045a6c9f8e6c1f
|
2024-06-15 04:57:19 +08:00 |
|
hiyouga
|
c29fa61a9c
|
fix #4292
Former-commit-id: 4cd4c179d24eab0fcaec2b29b9dd71970f877fe8
|
2024-06-15 04:47:13 +08:00 |
|
hiyouga
|
a30931fe0f
|
fix #4295
Former-commit-id: 08f657868f9d605b837c5d8c2946a25cc05c8735
|
2024-06-15 04:34:55 +08:00 |
|
hiyouga
|
3ff9b87012
|
add test cases
Former-commit-id: 731176ff34cdf0cbf6b41c40c69f4ceb54c2daf6
|
2024-06-15 04:05:54 +08:00 |
|
hiyouga
|
dbd1458adf
|
add quant check in webui export tab
Former-commit-id: 6455ca07061ae9858cd7bc996b28be1fde697a3d
|
2024-06-13 03:19:18 +08:00 |
|
hiyouga
|
49b58fd6af
|
fix #4221
Former-commit-id: 05a3be4853b941909e7d193c31e8d62c8c5f879b
|
2024-06-13 02:48:21 +08:00 |
|
hiyouga
|
103a507b39
|
fix #4209
DeepSpeed ZeRO3 has inflight param error when calling model.eval()
Former-commit-id: 4be013f18ea6a35b5a11db98db5f0670ffb41619
|
2024-06-13 02:25:50 +08:00 |
|
hiyouga
|
0a75224f62
|
clean code
Former-commit-id: f54cafd5c7f0383370d1a2f357834a61a97397ce
|
2024-06-13 01:58:16 +08:00 |
|
hoshi-hiyouga
|
04d7629abf
|
Merge pull request #4246 from hzhaoy/adapt-vllm-v0.5.0
adapt vllm==0.5.0
Former-commit-id: 1068e25fc8b89f11cc79b164ee4aef9ce137ad4c
|
2024-06-13 01:54:02 +08:00 |
|
hiyouga
|
5080f2314c
|
fix lint
Former-commit-id: b170165679317af2b3f03633afac27661b3deb06
|
2024-06-13 00:48:44 +08:00 |
|
hzhaoy
|
799873aa14
|
adapt vllm==0.5.0
Former-commit-id: 02afd9ff64f23e6707ac739ae1269f41bd70c340
|
2024-06-12 18:29:03 +08:00 |
|
hiyouga
|
6392d45ea7
|
fix #4242
Former-commit-id: cf260e7af03f49aa5e3d6daf3b27738ff9b9bcb8
|
2024-06-12 16:50:11 +08:00 |
|
Arthur Kim
|
16c7c92396
|
Support vllm==0.5.0
Former-commit-id: e7a8ffd7af21bc3759f055033ba2209fa7a1be0e
|
2024-06-12 16:49:12 +09:00 |
|
hoshi-hiyouga
|
7598b37543
|
Merge pull request #4204 from dignfei/main
fixbug：llama3在增量预训练时应该使用<|end_of_text|>标识文本的结束
Former-commit-id: e566342636faf0031a0ba5d5dd4fcff8401a2b76
|
2024-06-11 17:06:10 +08:00 |
|
hoshi-hiyouga
|
cc9717e2f2
|
Update pretrain.py
Former-commit-id: e2317b2a84149e39fddfd6366be3de23dfb71f82
|
2024-06-11 17:02:14 +08:00 |
|
hiyouga
|
08f2f99f4b
|
fix deepspeed version
Former-commit-id: 938a69bb07d4de7d82928ff01c582032162c1480
|
2024-06-11 16:52:36 +08:00 |
|
d
|
77bf3d66c7
|
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
|
820b6e7b32
|
fix #4198
Former-commit-id: 945d2c6cc73542adf9272ebd9aa332ea2c1c7361
|
2024-06-11 15:38:38 +08:00 |
|
hiyouga
|
27aece94cf
|
tiny fix
Former-commit-id: c4b2e263d9cefbad0fbc5de72422e4ef8edbcb54
|
2024-06-11 12:48:53 +08:00 |
|
hiyouga
|
2723438531
|
tiny fix
Former-commit-id: b5e9711ef375cc323fc083e742cccfc974550416
|
2024-06-11 01:04:16 +08:00 |
|
hiyouga
|
f330b73682
|
set dev version
Former-commit-id: 16c47cc15226119e33e46ba0f2f6ccb37072257f
|
2024-06-11 00:50:53 +08:00 |
|
hiyouga
|
0f1e592326
|
release v0.8.1
Former-commit-id: 875a34f492701d1c644facbe9ede411af2931513
|
2024-06-11 00:44:26 +08:00 |
|
hiyouga
|
4d7dd0330d
|
fix #4160
The split heads should be concatenated in dim=2
Former-commit-id: 4b3f247f270d44df9fe226cfe0dabfb7fcd2deda
|
2024-06-11 00:37:17 +08:00 |
|
hiyouga
|
4b2b92fd9a
|
update evaluator
Former-commit-id: bb8661e62481ff7027b8969f3d8a6a17290c9da3
|
2024-06-10 23:56:00 +08:00 |
|
hiyouga
|
784088db3f
|
fix #2666
Former-commit-id: f121d5c4f94af9f165132c4309cb9bdc8217d985
|
2024-06-10 21:24:15 +08:00 |
|
hiyouga
|
6d9fbb3fa9
|
fix llamafactory-cli env
Former-commit-id: b0515e5f42831b67d1f4d049999ecb68756e66db
|
2024-06-08 07:15:45 +08:00 |
|
hiyouga
|
9953ae3d03
|
set dev version
Former-commit-id: 08b7fe1c452cc99264ff0312e310b579590c6a45
|
2024-06-08 06:46:09 +08:00 |
|
hiyouga
|
c0c387e4db
|
release v0.8.0
Former-commit-id: 004db680b9e3996ec511ee818df6c0c02bf13603
|
2024-06-08 05:20:54 +08:00 |
|
hiyouga
|
ff9ba70bb8
|
reorganize adapter code
Former-commit-id: b26c2df9d97f4efffccbf7d28de13619b43f10dd
|
2024-06-08 00:47:23 +08:00 |
|
hoshi-hiyouga
|
adbebb0e3f
|
fix #4139
Former-commit-id: c025a4d74f293c14c2705e68af20a82a84608520
|
2024-06-08 00:45:02 +08:00 |
|
hiyouga
|
3f6b3eed98
|
add resume args in webui
Former-commit-id: 1d86ad768b1f36e54b4c2a9f18f6ea5a7df04c90
|
2024-06-08 00:22:16 +08:00 |
|
hiyouga
|
f45e81e186
|
fix #4137
Former-commit-id: cdc0d6f5a2e5040e145c82c4801f37bd76529047
|
2024-06-07 19:16:06 +08:00 |
|
hiyouga
|
ba648fd003
|
tiny fix
Former-commit-id: 0621bcad1dfbe8ce2464f741d4256c5df2a8d1b6
|
2024-06-07 05:19:21 +08:00 |
|
hiyouga
|
b0e5a76f4c
|
fix ppo trainer save zero3 model
accelerator.get_state_dict(ds_model) should be called at all ranks
Former-commit-id: 3a0f60f0aa072531e4ae5819ec00c8fa42aa0913
|
2024-06-07 05:14:19 +08:00 |
|
hiyouga
|
8692796c9b
|
fix ppo in trl 0.8.6
Former-commit-id: 5e0d66a0d80b4bd4a8506e2317209d8fb9d25ff6
|
2024-06-07 04:48:29 +08:00 |
|
hiyouga
|
d0edcde4ea
|
fix #4120
Former-commit-id: 2a44da678a5e360a9c0f9056397ac9e801329321
|
2024-06-07 04:18:05 +08:00 |
|
hiyouga
|
8c4c2e580c
|
update data processors
Former-commit-id: 04b138cbcb8b9a72e4bbda6c65843bb459e525e7
|
2024-06-07 04:15:40 +08:00 |
|
hoshi-hiyouga
|
07f33e7641
|
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
|
1998c641af
|
Update supervised.py
Former-commit-id: 04b6c2a754e602e0b698cfe6c255c2f2486d8865
|
2024-06-07 03:42:08 +08:00 |
|
hoshi-hiyouga
|
be1e5f9d62
|
Update supervised.py
Former-commit-id: 49993c4f4e1f871a22ff0196afe60026b668a4dc
|
2024-06-07 03:38:23 +08:00 |
|
hoshi-hiyouga
|
fdeec6db52
|
Update supervised.py
Former-commit-id: 67625b5278a839c12a3e4245f9e90af67d8b11b4
|
2024-06-07 03:38:04 +08:00 |
|
hiyouga
|
a4d335b42f
|
add qwen2 models
Former-commit-id: 49cb694d02c876e3740a003a8b332349f4310ad3
|
2024-06-07 00:22:57 +08:00 |
|
hiyouga
|
fcb134e144
|
rename files
Former-commit-id: e1a8431770fc36c0c9ee7fed4abbc3d7fdcc5efd
|
2024-06-07 00:09:06 +08:00 |
|
hiyouga
|
a47e24222a
|
add DISABLE_TORCHRUN option
Former-commit-id: bcc574b479c2101438723aadead42743d4378776
|
2024-06-06 23:44:58 +08:00 |
|
hoshi-hiyouga
|
b96b995620
|
Merge pull request #4082 from MengqingCao/bugfix
Fix #4077
Former-commit-id: 288028c3fb6bb1b58d1b7f4e8b90108c9bbf27d1
|
2024-06-06 23:38:40 +08:00 |
|
hoshi-hiyouga
|
c231706aa5
|
Update cli.py
Former-commit-id: 32190507534adf5f505858b3af2b592ca6568ac7
|
2024-06-06 23:38:09 +08:00 |
|
hiyouga
|
35b5117a59
|
fix ppo+zero3 #3108
Former-commit-id: 33a93cc29e3e57bf001515000c0a70c112573dea
|
2024-06-06 23:30:07 +08:00 |
|
hiyouga
|
80f716bc10
|
fix torch gc
Former-commit-id: e173799d057598e5692a407601c30d8ce1513461
|
2024-06-06 20:30:25 +08:00 |
|
hiyouga
|
ca95e98ca0
|
fix ppo dataset bug #4012
Former-commit-id: 7fc51b2e93698ae5e012566af8481f4d861c873d
|
2024-06-06 19:03:20 +08:00 |
|
hiyouga
|
d5559461c1
|
update trainers
Former-commit-id: b7f6c4a171293cf4f3e88f15a811f847342f84ee
|
2024-06-06 18:45:49 +08:00 |
|