mirrors
/
mmpretrain
mirror of https://github.com/open-mmlab/mmpretrain.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
262 Commits
8 Branches
41 Tags
32 MiB
Commit Graph

89 Commits (b30f79ea4ceb855d1a5540673d1951a3b981bf57)

Author SHA1 Message Date
whcao b30f79ea4c
[Feature]Modify Parameters Passing in models.heads (#239) 
* add mytrain.py for test

* test before layers

* test attr in layers

* test classifier

* delete mytrain.py

* set cal_acc in ClsHead defaults to False

* set cal_acc defaults to False

* use *args, **kwargs instead

* change bs16 to 3 in test_image_classifier_vit

* fix some comments

* change cal_acc=True

* test LinearClsHead
 2021-05-10 14:56:55 +08:00
whcao affb39fe07
[Feature]Add Vit (#214) 
* add imagenet bs 4096

* add vit_base_patch16_224_finetune

* add vit_base_patch16_224_pretrain

* add vit_base_patch16_384_finetune

* add vit_base_patch16_384_finetune

* add vit_b_p16_224_finetune_imagenet

* add vit_b_p16_224_pretrain_imagenet

* add vit_b_p16_384_finetune_imagenet

* add vit

* add vit

* add vit head

* vit unitest

* keep up with ClsHead

* test vit

* add flag to determiine whether to calculate acc during training

* Changes related to mmcv1.3.0

* change checkpoint saving interval to 10

* add label smooth

* default_runtime.py recovery

* docformatter

* docformatter

* delete 2 lines of comments

* delete configs/_base_/schedules/imagenet_bs4096.py

* add configs/_base_/schedules/imagenet_bs2048_AdamW.py

* rename imagenet_bs4096.py to imagenet_bs2048_AdamW.py

* add helpers.py

* test vit hybrid backbone

* fix HybridEmbed

* use to_2tuple instead
 2021-04-16 19:22:41 +08:00
whcao 1cde6f6e65
[Feature] Add cutmix option (#198) 
* Add cutmix option

* fix code style

* add some annotations

* add annotation about custom_hooks

* check constraint of alpha > 0

* add test cutmix

* fix code style

* add cutmix to configs/models

* add cutmix to configs/resnet

* flake8

* empty
 2021-04-14 21:27:42 +08:00
mzr1996 b7b520881f
Update CONTRIBUTING.md according to mmcv (#210) 
* Update CONTRIBUTING.md according to mmcv

* Docstring formatting by docformatter

* Update openmmlab website.
 2021-04-14 21:22:37 +08:00
whcao af83e981ac
[Bug]Fix label smooth bug (#203) 
* add convert_to_one_hot

* add test_label_smooth_loss

* add my label_smooth_loss

* fix CELoss bug

* test new label smooth loss

* LabelSmoothLoss downward compatibility

* add some comments

* remove the old version of LabelSmoothLoss

* add some comments

* add some comments

* add some comments

* add label smooth to config
 2021-04-13 13:53:56 +08:00
whcao dcf61173f6
[Feature]Add cal_acc to cls_head.py (#206) 
* add cal_acc to cls_head.py

* test ClsHead with cal_acc

* 4 spaces indentation
 2021-04-13 13:52:14 +08:00
LXXXXR 5195932952
[Feature] Support random augmentation (#201) 
* support random augmentation

* minor fix on posterize

* minor fix on posterize

* minor fix on cutout

* minor fix on cutout

* fix bug in solarize add

* revised according to comments
 2021-04-09 14:02:50 +08:00
LXXXXR 4d1fb1a662
[Feature] ColorJitter and Lighting (#190) 
* add configs

* remove config

* add color jitter and lighting

* revised according to comments
 2021-04-02 19:23:39 +08:00
LXXXXR 93cd960466
[Feature] Support AutoAug, AutoContrast, Equalize, Contrast, Brightness and Sharpness (#179) 
* add AutoContrast, Equalize, Contrast, Brightness and Sharpness pipelines

* add ImageNetPolicy

* add configs

* add unittest

* remove config

* rerun CI

* rerun CI

* [Fix] Update pip install mmcv command in ci (#187)

* update pip install mmcv command in ci

* update pip install mmcv command in ci

* fix ci

* fix ci
 2021-03-30 15:38:55 +08:00
LXXXXR b507fc207b
[Fix] Use MMCV's EvalHook in MMClassification (#182) 
* use eval_hook in mmcv

* minor fix
 2021-03-25 17:38:51 +08:00
LXXXXR f3b9380cbc
[Feature] Support solarize and posterize pipelines (#172) 
* support Solarize and Posterize

* revised according to comments

* fix conflicts
 2021-03-09 19:45:13 +08:00
LXXXXR 9614787fc4
[Feature] Add color pipeline (#171) 
* add ColorTransform pipeline

* fix docstring

* minor change

* revised according to comments
 2021-03-09 19:28:50 +08:00
LXXXXR c8033ece8e
add invert pipeline (#168) 2021-03-02 16:46:57 +08:00
LXXXXR b1fa298a66
add Rotate pipeline (#167) 2021-03-01 21:49:07 +08:00
LXXXXR bc50a7927b
reslove conflicts (#165) 2021-02-26 11:48:13 +08:00
LXXXXR 8c11c01fdb
[Feature] Add shear pipeline (#163) 
* half-done auto_augmentation

* remove auto_augcode and support shear pipeline

* fix typo

* fix typo

* use a non-square toy data instead
 2021-02-25 16:00:46 +08:00
David de la Iglesia Castro fb11a23cfe
Show results in test api (#162) 
* Initial support show in single_gpu_test

* Add result_show

* Add ori_filename meta_key

* test ori_filename

* Add show, show-dir, show-options
 2021-02-25 16:00:22 +08:00
ftbabi bdd6b01ae7
[Feature] Add "mixup" from Bag of Tricks (#160) 
* Add mixup option

* Modify the structure of mixup and add configs

* Clean configs

* Add test for mixup and SoftCrossEntropyLoss

* Add simple test for ImageClassifier

* Fix bug in test_losses.py

* Add assertion in CrossEntropyLoss
 2021-02-25 14:06:58 +08:00
LXXXXR fc82c31b2f
fix RandomCrop and RandomResizedCrop (#151) 2021-01-25 20:28:16 +08:00
LXXXXR 07bb15e5fd
[Feature] Add heads and config for multilabel task (#145) 
* resolve conflicts
add heads and config for multilabel tasks

* minor change

* remove evaluating mAP in head

* add baseline config

* add configs

* reserve only one config

* minor change

* fix minor bug

* minor change

* minor change

* add unittests and fix docstrings
 2021-01-25 18:10:14 +08:00
LXXXXR 13c1210741
[Feature] Add thrs in eval_metrics (#146) 
* support thr

* replace thrs with thr

* fix docstring

* minor change

* revise according to comments

* revised according to comments

* revise according to comments

* rewrite basedataset.evaluate to avoid duplicate calculation

* minor change

* change thr to thrs

* add more unit test
 2021-01-25 17:54:22 +08:00
LXXXXR 8e990b5654
[Feature] Support support and class-wise evaluation results (#143) 
* support support, support class-wise evaluation results and move eval_metrics.py

* Fix docstring

* change average to be non-optional

* revise according to comments

* add more unittest
 2021-01-19 16:42:16 +08:00
LXXXXR 6916f33d56
[Feature] Add asymmetric loss for multilabel task (#132) 
* add asymmetric loss

* minor change

* fix docstring

* do not apply sum over classes and fix docstring

* fix docstring

* fix weight shape

* fix weight shape

* add reference

* fix linkting issue

Co-authored-by: Y. Xiong <xiongyuxy@gmail.com>
 2021-01-11 11:22:22 +08:00
LXXXXR c07e60dd7d
[Feature] Support PASCAL VOC 2007 dataset for multilabel task (#134) 
* support voc

* minor change
 2021-01-11 11:08:59 +08:00
LXXXXR 194ab7efda
[Feature] Add bce loss for multilabel task (#130) 
* add bce loss for multilabel task

* minor change

* apply class wise sum

* fix docstring

* do not apply sum over classes and fix docstring

* fix docstring

* fix weight shape

* fix weight shape

* fix docstring

* fix linting issue

Co-authored-by: Y. Xiong <xiongyuxy@gmail.com>
 2021-01-11 11:05:24 +08:00
LXXXXR 9578bfa0f1
[Feature] Add focal loss for multilabel task (#131) 
* add focal loss

* apply class wise sum

* fix doctring

* do not apply sum over classes and fix docstring

* fix docstring

* fix weight shape

* fix weight shape
 2021-01-08 20:44:23 +08:00
LXXXXR 1df10beaa1
Add evaluation metrics for multilabel task (#123) 
* add mean_ap

* add difficult_examples in mAP to support dataset without difficult_examples

* fix docstring

* add CP,CR,CF1,OP,OR,OF1 as multilabel metrics

* fix docstring

* temporary solution to ci until new version of mmcv is avaliable (#127)

* temporary solution to ci until new version of mmcv is avaliable

* temporary solution to ci until new version of mmcv is avaliable

* add mean_ap

* add difficult_examples in mAP to support dataset without difficult_examples

* fix docstring

* add CP,CR,CF1,OP,OR,OF1 as multilabel metrics

* fix docstring

* Swap -1 and 0 for labels

* Revised according to comments

* Revised according to comments

* Revised according to comments

* Revert "Revised according to comments"
It is suggested that we should not include paper from arxiv.
This reverts commit 48a781cd6a.

* Revert "Revert "Revised according to comments""

This reverts commit 6d3b0f1a7b.

* Revert "Revised according to comments"
It is suggested we should not cite paper from arxiv.
This reverts commit 120ecda884.

* Revised according to comments

* revised according to comments

* Revised according to comments
 2021-01-04 12:25:33 +08:00
LXXXXR 4203b94643
fix bug in eval_metrics (#122) 2020-12-23 16:20:47 +08:00
LXXXXR b1e91f256b
fix bug in recall and precision (#112) 2020-12-09 16:27:42 +08:00
LXXXXR 21fd5019fb
add macro-averaged precision,recall,f1 options in evaluation (#93) 
* add macro-averaged precision,recall,f1 options in evaluation

* remove unnecessary comments

* Revise according to comments

* Revise according to comments
 2020-11-25 16:13:54 +08:00
LXXXXR 7636409b3b
add get_class in base_dataset (#85) 
* modify base_dataset

* revise according to the comments
 2020-11-12 14:22:02 +08:00
David de la Iglesia Castro df2189c6f6
Add missing CLASSES argument to dataset wrappers (#66) 
* Add missing classes in dataset wrappers

* Update tests
 2020-10-15 21:25:53 +08:00
David de la Iglesia Castro 4f4f2957ef
Use build_runner (#54) 
* Use build_runner in train api

* Support iter in eval_hook

* Add runner section

* Add test_eval_hook

* Pin mmcv version in install docs

* Replace max_iters with max_epochs

* Set by_epoch=True as default

* Remove trailing space

* Replace DeprecationWarning with UserWarning

* pre-commit

* Fix tests
 2020-10-15 21:12:50 +08:00
Lei Yang bc1b08ba41
Add VGG and pretained models (#27) 
* add vgg

* add vgg model coversion tool

* fix out_indices and docstr

* add vgg models in configs

* add params, flops and accuracy in docs

* add pretrained models url

* use ConvModule and refine var names

* update vgg conversion tool

* modify bn config

* add docs for arch_setting

* add unit test for vgg

* rm debug code

* update vgg pretrained models
 2020-09-29 17:49:42 +08:00
David de la Iglesia Castro 99115fddbc
Add albumentations (#45) 
* Add Albu transform

* pre-commit

* Create optional.txt

* Update requirements.txt

* Update transforms.py
 2020-09-22 17:35:39 +08:00
Jerry Jiarui XU 8d3acce307
Add ResNeSt (#25) 
* Add ResNeSt

* fixed test

* refactor

* add ResNeSt base

* update modelzoo

* update modelzoo

* Add S-200,S269 _base_
 2020-09-22 10:41:51 +08:00
Lei Yang c905decddf
Resize according to short edge (#22) 
* resize according to short edge

* revise doc for resize

* fix unitest for resize

* resize short_edge when second value of size is -1

* rename short_edge to short_side
 2020-08-11 16:04:45 +08:00
yl-1993 0009591aa3 add unittest for backends 2020-08-05 16:53:35 +08:00
yanglei 9a661ef981 Add ResNet_CIFAR 2020-07-12 00:06:56 +08:00
yanglei 6968ad5b3b Add RandomFlip 2020-07-12 00:06:56 +08:00
lixiaojie 1d5f3d8c24 update shufflenet_v1 2020-07-12 00:06:55 +08:00
lixiaojie f6260b33bf Add RandomCrop, RandomResizedCrop, RandomGrayscale, impad 2020-07-12 00:06:55 +08:00
yanglei 910d13330a Add Resize and CenterCrop 2020-07-07 17:07:58 +08:00
yanglei 17092d8be4 Fix the mid_channels of SEResNeXt 2020-07-07 11:25:48 +08:00
yanglei f9b1f18f28 fix test_dataset 2020-07-06 00:43:22 +08:00
lixiaojie 046aa5c003 add shufflenet_v2 convert code 2020-07-03 19:21:18 +08:00
yanglei ba789da5c8 Add commonly used datasets 2020-07-01 16:09:06 +08:00
chenkai 91a43c5bac Add RegNet 2020-07-01 14:21:45 +08:00
louzan 03b75789c6 Dev mobilenetv3 2020-06-30 15:50:36 +08:00
chenkai 02e11cc1f3 Refactoring for ResNet family 2020-06-25 11:57:50 +08:00