Mirrors/modelscope
1
0
Fork 0
mirror of https://github.com/modelscope/modelscope.git synced 2026-07-10 12:33:28 +02:00
Code Issues Packages Projects Releases Wiki Activity
1,961 Commits 275 Branches 121 Tags
master-github
Commit Graph

700 Commits

Author SHA1 Message Date
zeyinzi.jzyz
bf3a2b6c09 support vision efficient tuning finetune 
## 查看改动点 ↓↓↓
### vision efficient tuning finetune
- Model模块改造成适配训练的
- Model模块在支持训练同时向下兼容之前发布的modecard
- Pipline兼容modelcard加载的preprocessor或直接定义的
- 添加 ImageClassificationPreprocessor （非mmcv版本）
- 添加 VisionEfficientTuningTrainer
- ~~添加 opencv_transforms==0.0.6~~ (以源代码引入必要)

### Modelcard
- test pipeline和trainer合并到一起
- 新增3个模型的test
- 新增demo service

### 公共组件
- ms_dataset.py: fix warning, [UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or xxx]
- preprocessor添加common：ToNumpy、Rename、Identity
- preprocessor common对于dict进行key判断再取值。
- ~~修复learning rate在iter级别变化的逻辑。~~ （本次不做了）
- ~~修复非dist状态下train data没有进行shuffle的bug。~~ (Master已有人改了)
- 修复训练时调用util中非cv包的异常 zhconv。

### 其他
- 为防止新引入的preprocessor模块在config中被原代码加载，导致在其他人做CI时会报错；所以暂时没有添加新的tag，等CR完成后，会进行打tag再rerun CI。
        Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11762108

* support vision efficient tuning finetune

* update test case

* update shuffle on IterableDataset

* update bitfit & sidetuning

* compatible with base trainer
 2023-03-08 16:42:23 +08:00
xingjun.wxj
8298a3c31d add force redownload for nerf dataset in ut 
add force redownload for nerf dataset in ut

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11915271
 2023-03-08 16:23:58 +08:00
bin.xue
b4c90d8160 [to #42322933] feat: add dingding denoise model 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11731594

* feat: add dingding denoise

* refactor: delete duplicated unidfsmn class

* refactor: delete empty lines

* refactor: make some methods inline

* style: add license and optimize imports

* style: comments style
 2023-03-07 21:55:10 +08:00
xuanjie.wxb
ebb50d7887 add maoe unit test 2023-03-07 15:24:37 +08:00
wendi.hwd
99fa2fe909 update license and revert support camouflaged-detection from latest master 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11892922

    * update license and revert support camouflaged-detection from latest master
 2023-03-07 11:41:10 +08:00
shichen.fsc
04f63ec71f [to #42322933] add new kws models in UT and update requirements/audio_kws.txt 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11830273

* [Add] add new models for kws UT

* [Update] update requirements

* pre-commit
 2023-03-06 23:14:45 +08:00
fubang.zfb
dfd1058d45 siamese uie finetune support 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11706625
 2023-03-02 20:14:31 +08:00
wenmeng.zwm
77ff1d94d6 Revert "Add single-camera depth estimation model bts" 
This reverts commit feb4123327.
 2023-03-02 10:56:24 +08:00
wenmeng.zwm
f2f48dd305 Revert "support camouflaged-detection " 
This reverts commit ff55bd9436.
 2023-03-02 10:52:36 +08:00
ashui.cbh
329231fe9d ImagePaintbyexamplePipeline demo service 支持 
修改pipeline输出格式为numpy array 供demo service使用

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11841517
 2023-03-01 20:34:24 +08:00
ryan.yy
4197f72124 nerf重建加速模型 加入trainer训练模块 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11794296
 2023-03-01 12:05:57 +08:00
zhicheng.sc
d11fff0c0a add ControlNet for controllable image generation 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11697239

* add ControlNet for scribble2image

* update code comments

* support scribble input

* update scribble input for demo service

* support all models of ControlNet

* add requirements

* fix code style bug

* update model id
 2023-02-28 13:39:12 +08:00
wenmeng.zwm
6a13557c19 Merge branch 'master-github' into merge_master_github_0224 2023-02-24 11:16:35 +08:00
klayzhang.zb
4ca52a0934 [to #42322933] fix for text error correction: generate <unk> when input contains upper letter 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11762322

* [fix] fix: generate <unk> when input contains upper letter

* compatible new fairseq version
 2023-02-24 09:45:28 +08:00
zhk1425734486
99d2177bb0 landing AI: tbs image detection pull request (#84) 
https://modelscope.cn/models/landingAI/LD_CytoBrainCerv/summary

兰丁宫颈细胞AI辅助诊断模型

Co-authored-by: QZStudio <1643006760@qq.com>
Co-authored-by: wenmeng.zwm <wenmeng.zwm@alibaba-inc.com>
 2023-02-22 20:25:37 +08:00
wenmeng.zwm
6599d2c7cd skip test_image_structured_model_probing 2023-02-21 12:08:22 +08:00
wenmeng.zwm
ef615d48af skip test_image_structured_model_probing 2023-02-21 12:07:28 +08:00
xingzhang.rxz
1ae4782c83 add langident codes 
语种识别模型完成开发
模型地址：https://www.modelscope.cn/models/damo/nlp_language_identification-classification-base/summary

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11652333
 2023-02-17 15:21:33 +08:00
shuying.shu
615c443020 [to #42322933] add test condition for GPU availablity 
damo/cv_diffusion_image-segmentation不支持cpu环境测试

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11689517
 2023-02-16 11:13:53 +00:00
ryan.yy
a0c020cf34 [to #42322933] nerf model test add gpu constraint 2023-02-14 11:47:49 +00:00
wenqi.oywq
7499370569 [to #42322933] Fix the bug that image_color_enhance_pipeline cannot run in CPU environment 
adaint need to compile CUDA extension, so remove releated import from pipeline

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11660260
 2023-02-14 11:31:00 +00:00
fubang.zfb
4b29f28a04 siamese uie supports string-type schema input 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11655620
 2023-02-14 08:36:46 +00:00
fubang.zfb
e7cb55d4a6 Add SiameseUIE general information extraction model and pipeline 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11641614
 2023-02-14 03:56:29 +00:00
wenmeng.zwm
82f10f73aa [to #42322933] fix: pipeline base potential bug 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11645650

* test batched inference and no auto collate logic
 2023-02-13 13:49:42 +00:00
ashui.cbh
e7c5c116c5 [to #42322933] add test condition for GPU availablity 2023-02-13 11:50:30 +00:00
fuhaomin.fhm
2fced1c06f [to #42322933] Doc2Bot documentation with retrieval rerank, generation 2023-02-12 03:09:33 +00:00
zhicheng.sc
5b3ce816a5 sd2-inpainting: change output to numpy array in bgr order 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11629631
 2023-02-10 14:26:09 +00:00
haolan.xhl
e537f9df1d [to #42322933] add vqa-degradation 
add new model for image_quality_assessment_degradation
 2023-02-10 11:59:59 +00:00
ly261666
2c8f9801a2 [to #42322933] add content check model 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11473049
 2023-02-10 08:18:26 +00:00
wk258730
05a407da7f add pipelines for en2zh-imt and zh2en-imt 
*[ 交互式机器翻译-英中-通用领域-large](https://modelscope.cn/models/damo/nlp_imt_translation_en2zh/summary)
*[ 交互式机器翻译-中英-通用领域-large](https://modelscope.cn/models/damo/nlp_imt_translation_zh2en/summary)

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11593182
 2023-02-10 08:13:18 +00:00
hejunjie.hjj
6fc15926a3 [to #42322933] add single and multiple human parsing models 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11508413
 2023-02-10 08:01:23 +00:00
wenqi.oywq
6c5eb3bab3 Add AdaInt model 
新增加 adaint 模型

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11588338


* 复用image color enhance pipeline

* mv cuda extension to ops

* 修改adaint模型初始化方式，去掉封装
 2023-02-10 06:48:22 +00:00
zhangzhicheng.zzc
f88b03ab72 [to #47307896] update backbone-head 
The original backbone-head abstraction was not articheted well enough, the input and output parameters of backbone and head were in the form of **kwargs, which was implicit and might cause confustion. Therefore, the following adjustments were made:
原有backbone head抽象程度不够深，backbone 以及head输入输出参数为**kwargs，比较晦涩，同时很多功能无法支持扩展，因此做了如下调整：

1. Divide the basic model based on the structure to: encoder-only model, decoder-only model, single stage model, two stage model, etc., . Now, the encoder-only model was accomplished, while others are under design
2. Derive the structed task-models from the basic model structure above: a single structed task-model is mainly used to parse the backbone/head cfg, in order to apply the correct backbone or head components, some models might adjust the forward method from the basic model
3. Add the initialization parameters, input and output parameters to head class and backbone class, in order to reduce the understanding cost.
4. Remove the original nncrf class and chang it to backbone-head form with the lstm backbone and crf head.
5. Support  `model = Model.from_pretrained('bert-based-fill-mask', task='text-classification')`, this method could correctly load the backbone even when the task is different from the original one in configuration.
6. Support loading the model through the transformer's automodel, in the case of quickly integrating the backbone model without coding
7. Unifiy the original task classes in each nlp model and the structed task-model classes, the structed task-model are largely reduce the redundant codes in the original task classed. Still under refactor
8. Support load model configuration from hf transformers config.json, if the model related configuration is missing. Only suppport NLP models
 2023-02-10 06:46:47 +00:00
qize.yqz
5e88cfe693 add open vocabulary detection 
添加了开放词汇目标检测任务和模型

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11548569
 2023-02-10 06:37:51 +00:00
zhangyanzhao.zyz
e6c05a2931 sentence-embedding support finetune 
sentence-embedding模型支持finetune

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11537009
 2023-02-10 06:07:38 +00:00
zhongning.hzn
27304e38b0 Add bad image detection model and pipeline 
坏帧检测bad_image_detecting模型CR 

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11580680
 2023-02-10 05:33:57 +00:00
xixing.tj
b5e881148a add ocr detection new model and pipeline for dbnet 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11574897
 2023-02-10 05:02:59 +00:00
ly261666
b4ad0e49fb [to #42322933] add face quality assessment model 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11473023
 2023-02-10 04:59:08 +00:00
ljh263654
29e47e5030 add video deinterlace model 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11508847


* add video deinterlace model

* resolve conflict for video deinterlace

* fix CR problems
 2023-02-10 02:15:49 +00:00
mulin.lyh
c0a92403c8 Merge branch 'master' of gitlab.alibaba-inc.com:Ali-MaaS/MaaS-lib 2023-02-10 10:14:36 +08:00
mulin.lyh
fd7fd38da0 fix failed case 2023-02-10 10:14:24 +08:00
lingxi.yz
2d55e0f2ff [to #42322933] add abnormal detection models 
添加了针对长尾/小目标问题解决的异常视觉检测模型

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11536379
 2023-02-10 02:13:33 +00:00
hanyuan.chy
6da3be3047 Add HDFormer for 3d pose estimation of body 
新增HDFormer 3D人体姿态估计模型 

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11576328

* add hdformer

* rm redundent pipeline

* add license

* add test case
 2023-02-10 02:13:09 +00:00
yuanzhi.zyz
0894b1ea71 add ocr_recognition crnnnetwork 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11569972

 * add crnnnetwork

* add reviews

* fix conflict
 2023-02-10 02:11:59 +00:00
wenqi.oywq
da417655f2 新增deeplpf和debanding模型 
1. 新增加 image_debanding 模型
2. 新增加 deeplpf_image_color_enhance 模型

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11579722
 2023-02-09 13:08:50 +00:00
xianhui.lxh
b9ed767bbb [to #42322933] add ecbsr model for mobile image super-resolution 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11561999

* add mobile super-resolution pipeline and ecbsr model

* fix format error
 2023-02-09 12:59:05 +00:00
xianhui.lxh
9619c5fc83 [to #42322933] 2.10 add msrresnetlite model for video super-resolution 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11500523

    * add msrresnetlite model for video super-resolution

* Merge branch 'master' into dev/cv_msrresnet_video-super-resolution_lite

* fix doc string of msrresnet

* merge master

* merge msrresnet_lite model

* merge master
 2023-02-09 11:16:18 +00:00
liaojie.laj
82ae63b9e7 improve video frame interpolation pipeline 
更新实用版插帧模型，支持任意帧率转换

Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11553883
 2023-02-09 10:45:32 +00:00
shimin.ysm
2535866443 cv/image-fewshot-detection-defrcn support finetune and evaluation 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11486763

* defrcn完善评估功能，支持coco格式

* 修改格式问题

* 优化模型加载

* 优化训练测试脚本

* 修复推理时依赖数据集的问题

* 指定模型版本

* 指定model revision

* review意见修改
 2023-02-09 10:43:08 +00:00
maojialiang.mjl
3cb3e61ff7 add yolopv2 model cv_yolopv2_image_driving_perception 
Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/11548442
 2023-02-09 10:31:38 +00:00