Bump to 1.0.0rc3. (#1425)

* [Improve] add a version option in docs menu (#1162)

* update docs menu

* update docs/datasets

* [Enhance] update dev_scripts for link checking (#1164)

update dev_scripts for link checking

* [Refactoring] decompose the implementations of different metrics into several files (#1161)

* refactor metrics
* add UT for refactored metrics

* [Fix] Fix PPL bug (#1172)

fix ppl bug

* [Fix] Fix some known bugs. (#1200)

* fix bug in ceph config generation

* fix download error in io_utils

* trigger CI

* [Fix] Benchmark related bugs (#1236)

* fix benchmark url and related stuff

* fix lint

* [Enhancement] Support rerun failed or canceled jobs in `train_benchmark.py` (#1259)

* trigger CI

* fix cache dir in job watcher

* support rerun failure or canceled jobs in train benchmark

* add use case and readme for rerun

* avoid trigger circle CI pr-stage-test  when .dev_script is modified

* support change cpus-per-task in train benchmark

* [Fix]  Fix bugs in `sr test config`, `realbasicvsr config` and `pconv config` (#1167)

* fix pconv bug

* fix realbasicvsr

* fix sisr test config

* [Fix] fix test of Vid4 datasets bug (#1293)

* [Fix] fix test of Vid4 datasets bug

* Update tof_x4_official_vimeo90k.py

* Update tdan_x4_1xb16-lr1e-4-400k_vimeo90k-bd.py

* Update tdan_test_config.py

* [Feature] Support multi-metrics with different sample-model (#1171)

support multi-metrics with different sample-model

* [Fix] fix GenerateSegmentIndices ut (#1302)

* [Enhancement] Reduce the randomness in unit test of `stylegan3_utils.py` (#1306)

make s3-utils unit test more robustic

* [CI] Fix GitHub windows CI (#1320)

* add 'python -m' to windows' workflow

* [Fix] fix basicvsr++ mirror sequence bug (#1304)

* [Fix] fix sisr-test psnr config (#1319)

* [Fix] fix vsr models pytorch2onnx (#1300)

* [Bug] Ensure the output type of `GenerateFacialHeatmap` is `np.float32` (#1310)

* trigger CI

* ensure output of GenerateFacialHeatmap is float32

* [Bug] Fix sampling behavior of `unpaired_dataset.py` and  urls in cyclegan's README  (#1308)

* trigger CI

* fix download url in cyclegan readme

* fix bug in unpaired dataset

* support evaluate multi metrics in one configs for cyclegan models

* [README] Fix TTSR's README (#1325)

fix TTSR's README

* [CI] Update `paths-ignore` for GitHub CI (#1327)

update paths-ignore for GitHub CI

* [Bug] Save gt images in PGGAN's `forward` (#1328)

* save gt image to data samples in forward

* add ut to ensure gt img is saved

* [Bug] Correct RDN number of channels (#1332)

* [Bug] Revise flip transformation in some conditional gan's setting (#1331)

revise flip transformation in some conditional gan's setting

* [Unit Test] Fix unit test of SNR (#1335)

fix unit test of SNR

* [Bug] Revise flavr config (#1336)

revise flavr config

* [Fix] fix realesrgan ema (#1341)

* [Fix] fix realesrgan ema

* fix

* fix ut

* fix config

* fix ut

* [Fix] Fix bugs find during benchmark running (#1348)

* fix stylegan1 register name, add fid metric mapping

* fix lint

* [Fix] fix liif test config (#1353)

* [Enhancement] Complete save_best in configs (#1349)

complete save_best in generative models and glean

* [Config] Revise discriminator's learning rate of TTSR to align with 0.x version (#1352)

align ttsr-gan's disc lr with 0.x version (1e-4 -> 1e-5)

* [Fix] fix edsr configs (#1367)

fix edsr

* [Enhancement] Add pixel value clip in visualizer (#1365)

add pixel value clip in visualizer

* [Bug] Fix randomness in FixedCrop + add L1 loss in Pix2Pix (#1364)

* fix randomness in FixedCrop + add L1 loss in Pix2Pix

* add deterministic training for p2p

* [Fix] fix realbasicvsr config (#1358)

* [Enhancement] Fix PESinGAN-inter-pad setting + add SinGAN Dataset + add SinGAN demo (#1363)

* add singan dataset

* adopt singan config with PackEditInputs

* revise forward logic of SinGANMSGeneratorPE

* fix ema-related logic of SinGAN

* add singan demo

* add unit test for pe-singan

* add unit test for singan-ema and revise singan

* [Fix] fix types of exceptions in demos (#1372)

fix types of exceptions in demos

* [Enhancement] Support deterministic training in benchmark (#1356)

* support deterministic training in benchmark

* add kill-on-bad-exit to benchmark

* [Fix] Avoid cast int and float in GenDataPreprocessor (#1385)

avoid cast int and float in GenDataPreprocessor

* [Config] Update metric config in ggan (#1386)

* update metric config in ggan

* update gen_default_runtime

* [Config] Revise batch size in wang-gp's config (#1384)

revise batch size in wang-gp's config

* [Fix]: add type and change default number of preprocess_div2k_dataset.py (#1380)

* [Fix]: add type and change default number of preprocess_div2k_dataset.py

* using mmengine.scandir instead of mmcv.scandir

Co-authored-by: zenggyh1900 <zengyh1900@gmail.com>

* [Feature] Support qualitative comparison tools (#1303)

* Add base gui
* support patch compare
* support slider compare
* readme
* add recording prompt
* fix lint
* fix comments

Co-authored-by: unknown <zhangjingdong@pjlab.org>

* [Docs] Revise docs (change PackGenInputs and GenDataSample to mmediting ones) (#1382)

revise docs (change PackGenInputs and GenDataSample to mmediting ones)

Co-authored-by: Yanhong Zeng <zengyh1900@gmail.com>

* [Config] Revise Pix2Pix edges2shoes config (#1391)

* revise config for p2p-edge2shoes

* update batch size setting

* rename config name and revise readme

* [Bug] fix rdn and srcnn train configs (#1392)

* fix srcnn

* fix rdn

* [Fix] Fix test/val pipeline of pegan configs (#1393)

fix ppegan configs

* [Fix] Modify Readme of S3 (#1398)

modify readme

* [Fix] Correct fid of ggan (#1397)

fix readme of ggan

* [Feature] support instance_aware_colorization inference (#1370)

* add Instance-aware Image Colorization
* update configs
* update docstring if instance-aware image colorization
* add unit test of instance_aware_colorization
* use mmdet
Co-authored-by: ruoning <w853133995@outlook.com>

* [Bug] fix cain config (#1404)

* fix srcnn

* [Bug] fix cain config

* [Fix] Revise config and pretrain model loading in esrgan (#1407)

revise config and pretrain model loading in esrgan

* [Fix] Fix lsgan config (#1409)

fix lsgan config

* [Enhancement] Support `try_import` for `mmdet` (#1408)

* add try-import for mmdet

* revise import logic

* add unit test for try_import

Co-authored-by: Yanhong Zeng <zengyh1900@gmail.com>
Co-authored-by: Yifei Yang <2744335995@qq.com>

* [Enhancement] Set ``real_feat`` to cpu in inception_utils (#1415)

fix s2 configs

* [Enhancement] git ignore slurm generated files (#1416)

* improve git ignore

* add a blank line

* [Fix] modify readme and configs of stylegan2&pegan (#1418)

modify readme and configs of stylegan2&pegan

* [Enhancement] Support try-import for `clip` (#1420)

* support try-import for clip

* update ci

* [Enhancement]: Improve the rendering of Docs-API (#1373)

* [Enhancement]: Improve the rendering of Docs-API
* fix lint

Co-authored-by: root <root@DESKTOP-0FO18RA.localdomain>
Co-authored-by: Yanhong Zeng <zengyh1900@gmail.com>

* [Fix] Complete requirements (#1419)

* add requirements

* fix lint

Co-authored-by: Yanhong Zeng <zengyh1900@gmail.com>

* [Doc] Update changelog and README for 1.0.0rc3. (#1421)

* update changelog and README

* revise change log

* [Bug] Install clip in merge stage test (#1423)

install clip in merge stage test

* [Fix] Install clip in windows CI (#1424)

install clip in windows CI

Co-authored-by: Yanhong Zeng <zengyh1900@gmail.com>
Co-authored-by: Yifei Yang <2744335995@qq.com>
Co-authored-by: Z-Fran <49083766+Z-Fran@users.noreply.github.com>
Co-authored-by: Qunliang Xing <ryanxingql@gmail.com>
Co-authored-by: Yang Gao <Gary1546308416AL@gmail.com>
Co-authored-by: ruoning <44117949+ruoningYu@users.noreply.github.com>
Co-authored-by: unknown <zhangjingdong@pjlab.org>
Co-authored-by: root <root@DESKTOP-0FO18RA.localdomain>

.circleci/test.yml

@@ -64,6 +64,7 @@ jobs:
             mim install 'mmcv >= 2.0.0rc1'
             mim install 'mmdet >= 3.0.0rc2'
             pip install -r requirements/tests.txt
+            pip install git+https://github.com/openai/CLIP.git
       - run:
           name: Build and install
           command: |
@@ -108,6 +109,7 @@ jobs:
             docker exec mmedit mim install 'mmcv >= 2.0.0rc1'
             docker exec mmedit mim install 'mmdet >= 3.0.0rc2'
             docker exec mmedit pip install -r requirements/tests.txt
+            docker exec mmedit pip install git+https://github.com/openai/CLIP.git
       - run:
           name: Build and install
           command: |

.dev_scripts/train_benchmark.py

@@ -243,8 +243,7 @@ def create_train_job_batch(commands, model_info, args, port, script_name):
         job_script += (f'#SBATCH --gres=gpu:{n_gpus}\n'
                        f'#SBATCH --ntasks-per-node={min(n_gpus, 8)}\n'
                        f'#SBATCH --ntasks={n_gpus}\n'
-                       f'#SBATCH --cpus-per-task={args.cpus_per_job}\n'
-                       f'#SBATCH --kill-on-bad-exit=1\n\n')
+                       f'#SBATCH --cpus-per-task={args.cpus_per_job}\n\n')
     else:
         job_script += '\n\n' + 'export CUDA_VISIBLE_DEVICES=-1\n'
 

.github/workflows/merge_stage_test.yml

@@ -47,7 +47,9 @@ jobs:
           mim install 'mmcv >= 2.0.0rc1'
           mim install 'mmdet >= 3.0.0rc2'
       - name: Install other dependencies
-        run: pip install -r requirements/tests.txt
+        run: |
+          pip install -r requirements/tests.txt
+          pip install git+https://github.com/openai/CLIP.git
       - name: Build and install
         run: rm -rf .eggs && pip install -e .
       - name: Run unittests and generate coverage report
@@ -95,7 +97,9 @@ jobs:
           mim install 'mmcv >= 2.0.0rc1'
           mim install 'mmdet >= 3.0.0rc2'
       - name: Install other dependencies
-        run: pip install -r requirements/tests.txt
+        run: |
+          pip install -r requirements/tests.txt
+          pip install git+https://github.com/openai/CLIP.git
       - name: Build and install
         run: rm -rf .eggs && pip install -e .
       - name: Run unittests and generate coverage report
@@ -181,6 +185,7 @@ jobs:
           mim install 'mmcv >= 2.0.0rc1'
           mim install 'mmdet >= 3.0.0rc2'
           python -m pip install -r requirements/tests.txt
+          python -m pip install git+https://github.com/openai/CLIP.git
       - name: Build and install
         run: |
           python -m pip install -e .

.github/workflows/pr_stage_test.yml

@@ -41,7 +41,9 @@ jobs:
           mim install 'mmcv >= 2.0.0rc1'
           mim install 'mmdet >= 3.0.0rc2'
       - name: Install other dependencies
-        run: pip install -r requirements/tests.txt
+        run: |
+          pip install -r requirements/tests.txt
+          pip install git+https://github.com/openai/CLIP.git
       - name: Build and install
         run: rm -rf .eggs && pip install -e .
       - name: Run unittests and generate coverage report
@@ -97,6 +99,7 @@ jobs:
           mim install 'mmcv >= 2.0.0rc1'
           mim install 'mmdet >= 3.0.0rc2'
           pip install -r requirements/tests.txt
+          pip install git+https://github.com/openai/CLIP.git
       - name: Build and install
         run: |
           python setup.py check -m -s
@@ -131,6 +134,7 @@ jobs:
           mim install 'mmcv >= 2.0.0rc1'
           mim install 'mmdet >= 3.0.0rc2'
           python -m pip install -r requirements/tests.txt
+          pip install git+https://github.com/openai/CLIP.git
       - name: Build and install
         run: |
           python -m pip install -e .

.gitignore

@@ -70,6 +70,7 @@ docs/en/_tmp/
 docs/zh_cn/_build/
 docs/zh_cn/_tmp/
 requirements/src/
+docs/en/api/generated
 
 # PyBuilder
 target/
@@ -140,3 +141,7 @@ out/*.png
 
 # MacOS
 .DS_Store
+
+# Slurm
+batchscript-*
+*.out

README.md

@@ -106,12 +106,13 @@ hope MMEditing could provide better experience.
 
 ### 🌟 Preview of 1.x version
 
-A brand new version of [**MMEditing v1.0.0rc2**](https://github.com/open-mmlab/mmediting/releases/tag/v1.0.0rc2) was released in 02/11/2022:
+A brand new version of [**MMEditing v1.0.0rc3**](https://github.com/open-mmlab/mmediting/releases/tag/v1.0.0rc3) was released in 09/11/2022:
 
 - Support all the tasks, models, metrics, and losses in [MMGeneration](https://github.com/open-mmlab/mmgeneration) 😍。
 - Unifies interfaces of all components based on [MMEngine](https://github.com/open-mmlab/mmengine).
 - Support patch-based and slider-based image and video comparison viewer.
 - Support image colorization.
+- Convert `mmdet` and `clip` to optional requirements.
 
 Find more new features in [1.x branch](https://github.com/open-mmlab/mmediting/tree/1.x). Issues and PRs are welcome!
 

README_zh-CN.md

@@ -105,12 +105,13 @@ MMEditing 缜密地设计新的框架并将其精心实现，希望能够为您
 
 ### 🌟 1.x 预览版本
 
-全新的 [**MMEditing v1.0.0rc2**](https://github.com/open-mmlab/mmediting/releases/tag/v1.0.0rc2) 已经在 02/11/2022 发布:
+全新的 [**MMEditing v1.0.0rc3**](https://github.com/open-mmlab/mmediting/releases/tag/v1.0.0rc3) 已经在 09/11/2022 发布:
 
 - 支持[MMGeneration](https://github.com/open-mmlab/mmgeneration)中的全量任务、模型、优化函数和评价指标 😍。
 - 基于[MMEngine](https://github.com/open-mmlab/mmengine)统一了各组件接口。
 - 支持基于图像子块以及滑动条的图像和视频比较可视化工具。
 - 支持图像上色任务。
+- 将 `mmdet` 与 `clip` 切换为可选依赖
 
 在[1.x 分支](https://github.com/open-mmlab/mmediting/tree/1.x)中发现更多特性！欢迎提 Issues 和 PRs！
 

configs/cain/cain_g1b32_1xb5_vimeo90k-triplet.py

@@ -83,7 +83,7 @@
     num_workers=32,
     batch_size=32,  # 1 gpu
     persistent_workers=False,
-    sampler=dict(type='InfiniteSampler', shuffle=True),
+    sampler=dict(type='DefaultSampler', shuffle=True),
     dataset=dict(
         type=train_dataset_type,
         ann_file='tri_trainlist.txt',

configs/esrgan/esrgan_psnr-x4c64b23g32_1xb16-1000k_div2k.py

@@ -90,8 +90,8 @@
     sampler=dict(type='DefaultSampler', shuffle=False),
     dataset=dict(
         type=dataset_type,
-        metainfo=dict(dataset_type='set5', task_name='sisr'),
-        data_root=data_root + '/Set5',
+        metainfo=dict(dataset_type='set14', task_name='sisr'),
+        data_root=data_root + '/Set14',
         data_prefix=dict(img='LRbicx4', gt='GTmod12'),
         pipeline=val_pipeline))
 

configs/esrgan/esrgan_x4c64b23g32_1xb16-400k_div2k.py

@@ -5,12 +5,13 @@
 
 scale = 4
 
-# load_from = 'https://download.openmmlab.com/mmediting/restorers/esrgan/esrgan_x4c64b23g32_1x16_400k_div2k_20200508-f8ccaf3b.pth'  # noqa
-
 # DistributedDataParallel
 model_wrapper_cfg = dict(type='MMSeparateDistributedDataParallel')
 
 # model settings
+pretrain_generator_url = (
+    'https://download.openmmlab.com/mmediting/restorers/esrgan'
+    '/esrgan_psnr_x4c64b23g32_1x16_1000k_div2k_20200420-bf5c993c.pth')
 model = dict(
     type='ESRGAN',
     generator=dict(
@@ -20,7 +21,8 @@
         mid_channels=64,
         num_blocks=23,
         growth_channels=32,
-        upscale_factor=scale),
+        upscale_factor=scale,
+        init_cfg=dict(type='Pretrained', checkpoint=pretrain_generator_url)),
     discriminator=dict(type='ModifiedVGG', in_channels=3, mid_channels=64),
     pixel_loss=dict(type='L1Loss', loss_weight=1e-2, reduction='mean'),
     perceptual_loss=dict(

configs/lsgan/lsgan_lsgan-archi_lr1e-4-1xb64-10Mimgs_lsun-bedroom-128x128.py

@@ -42,9 +42,10 @@
 metrics = [
     dict(
         type='FrechetInceptionDistance',
-        prefix='FID-Full-50k',
+        prefix='FID-50k',
         fake_nums=50000,
-        inception_style='StyleGAN',
+        real_nums=50000,
+        inception_style='PyTorch',
         sample_model='orig')
 ]
 val_evaluator = dict(metrics=metrics)