hoshi-hiyouga
ef7af457fc
[infer] fix vllm args ( #7235 )
2025-03-11 01:15:35 +08:00
Ze-Yi LIN
a1e76af3d9
[tracking] add swanlab_logdir param ( #7219 )
...
* feat: add swanlab_logdir param
* fix
2025-03-11 00:53:07 +08:00
hoshi-hiyouga
ed8b12e3cb
[config] update args ( #7231 )
2025-03-10 23:04:43 +08:00
hoshi-hiyouga
728c2f6819
[config] fix export max len ( #7230 )
2025-03-10 16:46:08 +08:00
hoshi-hiyouga
ae4cbe8fbc
[assets] update wechat ( #7229 )
2025-03-10 15:39:06 +08:00
hoshi-hiyouga
1774882f5a
[data] update mm demo data ( #7211 )
2025-03-07 20:07:15 +08:00
hoshi-hiyouga
cdf8fc6478
[assets] update readme ( #7209 )
2025-03-07 17:27:49 +08:00
hoshi-hiyouga
8c3f9f6747
[data] fix loader ( #7207 )
...
* fix dataloader
* add test case
* fix type
* fix ci
* fix ci
* fix ci
* disable overwrite cache in ci
2025-03-07 17:20:46 +08:00
hoshi-hiyouga
db113f690e
[misc] fix ds config ( #7205 )
2025-03-07 15:21:28 +08:00
ZhangChuanhui
194e3bddb2
[data] fix function formatter ( #7201 )
...
Co-authored-by: zhangchuanhui <zhangchal@digitalchina.com >
2025-03-07 15:17:23 +08:00
hoshi-hiyouga
bd17223559
[misc] fix cli ( #7204 )
2025-03-07 15:01:18 +08:00
hoshi-hiyouga
313355759d
[script] fix vllm version ( #7193 )
2025-03-06 17:14:17 +08:00
hoshi-hiyouga
abb23f7673
[webui] support escape html ( #7190 )
2025-03-06 16:52:21 +08:00
hoshi-hiyouga
d739fddb10
[deps] upgrade vllm ( #7183 )
2025-03-06 15:25:08 +08:00
hoshi-hiyouga
be66df1f02
[data] fix mm template ( #7181 )
2025-03-06 15:18:32 +08:00
hoshi-hiyouga
64a6fb9b50
[model] add QwQ 32b ( #7179 )
2025-03-06 11:58:36 +08:00
Ze-Yi LIN
8ad03258e1
[trainer] fix swanlab callback ( #7176 )
2025-03-06 00:33:37 +08:00
hoshi-hiyouga
b4b89b4ff3
[trainer] update config ( #7174 )
2025-03-05 23:32:54 +08:00
sirui.li
dff4130969
[data] fix qwen2audio plugin ( #7166 )
...
* Update pairwise.py
[data]Repair multimodal model dpo training
* Update pairwise.py
[data]repair multimodal model dpo training using deepcopy
* Update pairwise.py
* Update mm_plugin.py
2025-03-05 18:03:36 +08:00
hoshi-hiyouga
0c403ea15b
[assets] update wechat ( #7161 )
2025-03-05 14:11:10 +08:00
hoshi-hiyouga
bc298c60b7
[data] use bicubic resampler ( #7143 )
2025-03-04 00:17:06 +08:00
hoshi-hiyouga
17ba2d5082
[webui] fix webui ( #7142 )
2025-03-04 00:01:49 +08:00
rabbit
049ddf48af
[data] bailing template ( #7117 )
...
* add bailing template
* add bailing template
* add bailing template
---------
Co-authored-by: chengshiwen.csw@antgroup.com <chengshiwen.csw@antgroup.com >
2025-03-03 15:33:22 +08:00
hoshi-hiyouga
1036311826
[inference] fix hf_engine ( #7120 )
2025-03-01 05:22:49 +08:00
hoshi-hiyouga
d1863bbbaa
[assets] update wechat ( #7106 )
2025-02-28 12:01:04 +08:00
Ze-Yi LIN
891c487503
[webui] display swanlab exp link ( #7089 )
...
* webui add swanlab link
* change callback name
* update
---------
Co-authored-by: hiyouga <hiyouga@buaa.edu.cn >
2025-02-27 19:40:54 +08:00
leo-pony
acc52e0fe7
[npu] update cann base image and torch 2.4 ( #7061 )
...
* Update base npu container image version:The Python version required for Hugging Face Transformers is >= python3.10
* Fix the bug: arg type of INSTALL_DEEPSPEED shoud been string now.
* Update Ascend CANN, CANN-Kernel and corresponding torch and torch-npu version
* Upgrade torch-npu needs packages' version: torch==2.1.0 and torch-npu==2.4.0.post2
2025-02-25 23:32:01 +08:00
hoshi-hiyouga
96fd510e6a
[misc] fix project toml ( #7067 )
2025-02-25 23:22:48 +08:00
JieShen
e8266fe563
[script] add seed args ( #7058 )
...
* add seed args
* add seed args
* update seed
2025-02-25 19:44:57 +08:00
Kingsley
19861d5170
[model] add paligemma2-mix series ( #7060 )
2025-02-25 18:51:16 +08:00
hoshi-hiyouga
76314e6ad1
[data] fix mllama ( #7053 )
...
* fix mllama
* fix test
2025-02-24 22:05:38 +08:00
hoshi-hiyouga
ec1a1bc118
[model] add models ( #7054 )
...
* add qwen25vl awq models
* add moonlight
2025-02-24 22:05:13 +08:00
hoshi-hiyouga
fe6dd92c84
[assets] update readme ( #7051 )
2025-02-24 20:45:06 +08:00
hoshi-hiyouga
1481af5dc9
[assets] update wechat ( #7019 )
2025-02-20 20:32:33 +08:00
Zhangchi Feng
cde479e47a
[data] fix MiniCPMV plugin ( #6998 )
...
* fix template
* fix bug in messages processing
2025-02-19 19:36:04 +08:00
hoshi-hiyouga
302ecb00fe
[webui] update css ( #6985 )
2025-02-18 18:27:57 +08:00
hoshi-hiyouga
2591a3fa8b
[data] add r1 distill dataset ( #6983 )
2025-02-18 17:25:09 +08:00
hoshi-hiyouga
b00b290c07
[version] support transformers 449 ( #6982 )
...
* support transformers 449
* fix mm plugin
2025-02-18 17:05:40 +08:00
hoshi-hiyouga
cc8c7e762b
[misc] fix script ( #6977 )
2025-02-18 17:00:46 +08:00
hoshi-hiyouga
3da2cc2710
[data] update vlm args ( #6976 )
2025-02-18 02:12:51 +08:00
hoshi-hiyouga
7faecc0301
[data] add min resolution option ( #6975 )
2025-02-18 01:40:46 +08:00
hoshi-hiyouga
bdb581c4a8
[data] fix predict dataset ( #6972 )
2025-02-17 20:29:40 +08:00
hoshi-hiyouga
ad0c6c8916
[assets] update wechat ( #6963 )
2025-02-17 15:23:17 +08:00
Zhangchi Feng
2faf8aeff8
[data] fix minicpmo template ( #6946 )
2025-02-15 00:37:41 +08:00
Eric Tang
6edd4992d7
[ray] specify ray storage path ( #6920 )
2025-02-14 21:55:41 +08:00
hoshi-hiyouga
1ada3ae5a3
[misc] fix lora regex ( #6944 )
...
* fix lora regex
* fix
2025-02-14 21:38:43 +08:00
hoshi-hiyouga
c31c63b411
[misc] fix grad ckpt ( #6931 )
2025-02-13 23:27:51 +08:00
hoshi-hiyouga
797043d29c
[model] add liger kernel to qwen2_5 vl ( #6930 )
...
* add liger kernel to qwen2_5 vl
* fix patch
* fix patch
2025-02-13 23:05:54 +08:00
Billy Cao
11eac71c13
[trainer] fix gen_kwarg to eval during training ( #5451 )
...
* Correctly pass gen_kwarg to eval during model runs
* fix
* fix
---------
Co-authored-by: hiyouga <hiyouga@buaa.edu.cn >
2025-02-13 02:35:06 +08:00
SrWYG
1e35967ae1
[data] evaluate on each dataset ( #5522 )
...
* [Update] loader.py , evaluate will run separate evaluations on each dataset.
`If you pass a dictionary with names of datasets as keys and datasets as values, evaluate will run separate evaluations on each dataset. This can be useful to monitor how training affects other datasets or simply to get a more fine-grained evaluation`
seq2seqtrainner support eval_dataset as Dict.
* fix format
* fix
* fix
---------
Co-authored-by: hiyouga <hiyouga@buaa.edu.cn >
2025-02-13 02:19:03 +08:00
