Mirrors/modelscope
1
0
Fork 0
mirror of https://github.com/modelscope/modelscope.git synced 2025-12-25 04:29:22 +01:00
Code Issues Packages Projects Releases Wiki Activity
2,825 Commits 263 Branches 106 Tags
fix/patch_dynamic_module
Commit Graph

1197 Commits

Author SHA1 Message Date
翎航
3b21ff10ec fix ocr prepreocess 2022-10-31 16:57:49 +08:00
yichang.zyc
e2d35fbb14 [to #42322933]clip支持finetune 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10572842
 2022-10-30 21:51:11 +08:00
Yingda Chen
9f7b8b86a3 [to #42322933] disble 2dkeypoints training since face_2d_keypoints_dataset is set to be private 2022-10-30 13:59:12 +08:00
Yingda Chen
902019c2e0 [to #42322933] disble vgg19_fer 2022-10-30 13:55:49 +08:00
Yingda Chen
29448c0f57 [to #42322933] disble vit 2022-10-30 11:15:52 +08:00
mulin.lyh
3791ee7ad2 [to #45821936]fix: fix block user specify revision after release_datetime 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10572162
 2022-10-29 13:44:47 +08:00
yuze.zyz
4b7e8e89aa [to #42322933] Fix some bugs when downgrade the version of some dependencies 
1. Fix bug in model exporting
2. Skip some long trainings in test level 2
3. Refine some comments
4. Fix a bug that mode is not correct when saving checkpoints
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10564716
 2022-10-28 21:44:33 +08:00
Yufeng
261c04b8b5 add Mglm (#5) 
* mglm init

* add mglm requirements

Co-authored-by: Yufeng <zhuyufeng@gmail.com>
Co-authored-by: wenmeng.zwm <wenmeng.zwm@alibaba-inc.com>
 2022-10-28 17:12:47 +08:00
Yufeng
c390dc0c79 add Mglm (#5) 
* mglm init

* add mglm requirements

Co-authored-by: Yufeng <zhuyufeng@gmail.com>
Co-authored-by: wenmeng.zwm <wenmeng.zwm@alibaba-inc.com>
 2022-10-28 17:09:27 +08:00
wenmeng.zwm
0223f6e6d7 Merge branch 'master-gitlab' into merge_mater_internal_1028 2022-10-28 10:02:45 +08:00
Yingda Chen
46cfa177aa [to #42322933]skip timeconsuming test 2022-10-28 09:34:29 +08:00
xianzhe.xxz
88e8d4291a [to #42322933]"fix: set the eps and momentum of BN consistent with training" 
To keep consistent between training and evaluation, change the eps and momentum of BN.
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10554451
 2022-10-28 09:27:55 +08:00
menrui.mr
c7b0787049 修复初始化过程参数未生效问题 
此前文生图模型没有加载configuration.json中的参数 影响默认配置
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10558026
 2022-10-27 23:29:08 +08:00
hemu.zp
fa415d8720 [to #42322933] Fix bug for bloom and gpt_neo 
1. 修复 bloom 和 gpt_neo 模型更新 transformers 4.23 后后处理报错的问题
2. 统一使用 ModelOutput 作为模型输出
3. gpt_neo checkpoint 已上线，修改 ut 为 level2
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10553103
 2022-10-27 23:27:28 +08:00
xingjun.wxj
78f29cf999 [to #42322933] Add delete datasets files and upload mode. 
1. Add : MsDataset.delete() , support delete dataset file or dir.

2. Add: upload mode,  MsDataset.upload(xx, upload_mode=UploadMode.FORCE_UPLOAD), or  MsDataset.upload(xx, upload_mode=UploadMode.APPEND_UPLOAD)
     if upload_mode = UploadMode.APPEND_UPLOAD, then skip object in case of this object exists.

3. Add: support reload sts token automatically to avoid expire. (current expiration: 24h)

4. Fix: add cookies in api.py for downloading private datasets.
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10524449
 2022-10-27 20:30:35 +08:00
Yingda Chen
374fd3090e [to #42322933]skip referring video tests since model is private 2022-10-27 20:23:51 +08:00
yuze.zyz
212cf53318 [to #42322933] Fix some bugs 
1. Add F1 score to sequence classification metric
2. Fix a bug that the evaluate method in trainer does not support a pure pytorch_model.bin
3. Fix a bug in evaluation of veco trainer 
4. Add some tips if lr_scheduler in the trainer needs a higher version torch
5. Add some comments
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10532230
 2022-10-27 19:49:21 +08:00
shuying.shu
ddcb57440d [to #42322933]add fine-tune code for referring video object segmentation 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10539423
 2022-10-27 19:43:54 +08:00
mulin.lyh
3b75623be4 [to #45773874]fix: get_model revision=None bug, and hub case occasionally delete test model failed 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10549680
 2022-10-27 17:06:18 +08:00
eniac.xcw
8886c3c1ae [to #42322933]fine tune team on caltech-101 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10525413
 2022-10-27 12:00:14 +08:00
wenmeng.zwm
7e8e60f3b3 Merge branch 'master-gitlab' into merge_master_internal_1026 2022-10-27 10:25:13 +08:00
yingda.chen
de708dd518 add basic remap column wrapper 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10539917

    * add basic remap column wrapper
 2022-10-27 10:12:05 +08:00
hemu.zp
69104c0f8a [to #42322933] Refactor text generation model outputs and fix some bugs 
1. 将 single_gpu_test 与 multi_gpu_test 中的 model.forward 部分分离为 EpochBasedTrainer 中的 evaluation_step，为部分 evaluation 阶段不调用 forward 的模型提供更好的灵活性
2. 重构代码将文本生成模型 Model 层的输入输出统一为 Tensor，Tensor 到 str 的 decode 过程移动到 pipeline 中完成
3. pipeline 后处理添加对中文和中文标点与英文混杂时空格的处理，使 decode 后中英文混杂输出正确
4. 添加 TextGenerationTrainer 修复了部分模型 evaluation 过程 forward 输出单个 token 计算 metrics 的问题
5. 修复了 rouge 无法接收空字符串的问题
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10473768
 2022-10-27 09:52:05 +08:00
liugao.lg
0605376135 [to #42322933]add ofa finetune 
新增ofa的finetune能力
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10540701
 2022-10-27 09:29:06 +08:00
翎航
2299f8fa65 fix conflict 2022-10-26 22:41:13 +08:00
hemu.zp
d0f8547e7e [to #42322933] Fix gpt3 loading checkpoint after finetuning. 
1. 修复GPT-3模型无法加载finetune保存的checkpoint的问题
2. 为GPT-3诗词生成模型添加 ut
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10537209
 2022-10-26 20:58:00 +08:00
翎航
022fa4948a fix ocr-finetune acc 2022-10-26 19:44:54 +08:00
jiaqi.sjq
7b84adc914 [to #42322933]Fix remove files in local model not take effect to remote repo after push_model 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10533214
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10533214
 2022-10-26 19:15:43 +08:00
wenmeng.zwm
c638acbeda fix merge conflict 2022-10-26 17:12:25 +08:00
hemu.zp
e4a0e046f9 [to #42322933] Add ut for mplug and bloom 
为新上线的 langboat/bloom-1b4-zh，damo/mplug_visual-question-answering_coco_base_zh，damo/mplug_image-captioning_coco_base_zh 三个模型添加 ut，test_level 设置为 2
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10524221
 2022-10-26 16:19:20 +08:00
caorongyu.cry
3b8fb92c13 [to #42322933] debug header ids and header names 
修复header_ids和header_names命名反了的问题
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10516557
 2022-10-26 16:04:14 +08:00
ran.zhou
13f7e9ceca [to #42322933]SEA multilingual NLP (NER & word segmentation) 
添加东南亚小语种NLP支持，包括：
1. 针对泰语，越南语NER的预处理
2. 基于XLMR-CRF架构的分词模型和pipeline
3. 针对泰语分词的预处理

添加了相应pipeline的unittest
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10492404
 2022-10-26 14:52:22 +08:00
mulin.lyh
384377b8f5 * [to #45486649]feat: modelscope model version use model repo tag, unsupport branch or commit it, client user-agent header unified 2022-10-26 13:55:51 +08:00
jiaqi.sjq
5190c7de11 [to #41669377] tts using default master revision model in UT 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10526747
 2022-10-26 11:53:52 +08:00
翎航
90d47832c0 add ocr-finetune ned 2022-10-26 11:45:50 +08:00
翎航
c077dea072 add ocr-finetune 2022-10-26 10:52:10 +08:00
zhangyanzhao.zyz
781fe49d63 [to #42322933]修正finetune text ranking bugs 
之前的finetune代码当dataset最后长度不足制定batch size时会出错，现已修正
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10524066
 2022-10-26 09:44:25 +08:00
yuanzheng.yuanzhen
bab54bbce8 [to #42322933]support uni fold 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10481410
 2022-10-25 22:59:19 +08:00
yuze.zyz
c2da44b371 [to #42322933] remove dev model inference and fix some bugs 
1. Change structbert dev revision to master revision
2. Fix bug:  Sample code failed because the updating of model configuration
3. Fix bug: Continue training regression failed
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10519992
 2022-10-25 22:38:49 +08:00
zhangzhicheng.zzc
e1ab73b7d8 [to #42322933]support type str for for zero-shot labels' input 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10506320
 2022-10-25 13:55:09 +08:00
hemu.zp
ffd834fc25 [to #42322933] Add bloom model 
添加 bloom 模型
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10509187
 2022-10-25 12:58:02 +08:00
yichang.zyc
6ddafb3218 [to #42322933]caption finetune done, add belu metric 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10318299
 2022-10-25 12:55:41 +08:00
yuze.zyz
605cd7f44a [to #42322933] NLP 1030 Refactor 
Features:
1. Refactor the directory structure of nlp models. All model files are placed into either the model folder or the task_model folder
2. Refactor all the comments to google style
3. Add detail comments to important tasks and nlp models, to list the description of the model, and its preprocessor&trainer
4. Model Exporting now supports a direct all to TorchModelExporter(no need to derive from it)
5. Refactor model save_pretrained method to support direct running(independent from trainer)
6. Remove the judgement of Model in the pipeline base class, to support outer register models running in our pipelines
7. Nlp trainer now has a NLPTrainingArguments class , user can pass arguments into the dataclass, and use it as a normal cfg_modify_fn, to simplify the operation of modify cfg.
8. Merge the BACKBONES and the MODELS, so user can get a backbone with the Model.from_pretrained call
9. Model.from_pretrained now support a task argument, so user can use a backbone and load it with a specific task class.
10. Support Preprocessor.from_pretrained method
11. Add standard return classes to important nlp tasks, so some of the pipelines and the models are independent now, the return values of the models will always be tensors, and the pipelines will take care of the conversion to numpy and the following stuffs.
12. Split the file of the nlp preprocessors, to make the dir structure more clear.

Bugs Fixing:
1. Fix a bug that lr_scheduler can be called earlier than the optimizer's step
2. Fix a bug that the direct call of Pipelines (not from pipeline(xxx)) throws error
3. Fix a bug that the trainer will not call the correct TaskDataset class
4. Fix a bug that the internal loading of dataset will throws error in the trainer class
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10490585
 2022-10-25 12:26:25 +08:00
siyang.ssy
6d51f44dc7 [to #42322933]fix input type for video embeding 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10506601
 2022-10-25 12:11:28 +08:00
bin.xue
525fa3ea89 [to #42322933]test: use 'master' branch in training test 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/10509580
 2022-10-25 12:10:07 +08:00
行嗔
cc8b78eac8 update rdrop 2022-10-25 11:55:37 +08:00
行嗔
0c64d3fca5 Merge remote-tracking branch 'origin/ofa/finetune_loss' into ofa/finetune 
# Conflicts:
#	tests/trainers/test_ofa_trainer.py
 2022-10-25 11:48:03 +08:00
翎航
1bb1eeec77 fix ut 2022-10-25 10:55:24 +08:00
行嗔
2288a0fdf3 fix all comments 2022-10-25 10:18:33 +08:00
行嗔
df5bd86048 fix a ut bug 2022-10-25 10:15:06 +08:00