[Feature] Add config of ttsr-gan

[Yshuo-Li]

No description provided.

[codecov]

Codecov Report

Merging #398 (8574053) into master (028750e) will decrease coverage by 1.45%.
The diff coverage is 80.48%.

❗ Current head 8574053 differs from pull request most recent head 885071a. Consider uploading reports for the commit 885071a to get more accurate results

@@            Coverage Diff             @@
##           master     #398      +/-   ##
==========================================
- Coverage   82.17%   80.71%   -1.46%     
==========================================
  Files         143      186      +43     
  Lines        6486     9978    +3492     
  Branches      963     1463     +500     
==========================================
+ Hits         5330     8054    +2724     
- Misses       1060     1721     +661     
- Partials       96      203     +107     

Flag	Coverage Δ
unittests	80.69% <80.42%> (-1.48%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
mmedit/apis/generation_inference.py	23.07% <ø> (ø)
mmedit/apis/inpainting_inference.py	30.76% <ø> (ø)
mmedit/apis/matting_inference.py	19.51% <ø> (ø)
mmedit/apis/restoration_inference.py	16.66% <0.00%> (ø)
mmedit/apis/test.py	25.25% <0.00%> (-0.79%)	⬇️
mmedit/core/distributed_wrapper.py	37.14% <0.00%> (ø)
mmedit/core/evaluation/eval_hooks.py	62.16% <0.00%> (ø)
mmedit/core/evaluation/metric_utils.py	100.00% <ø> (ø)
mmedit/core/hooks/visualization.py	96.96% <ø> (+0.09%)	⬆️
mmedit/core/mask.py	97.72% <ø> (ø)
... and 217 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update f52c456...885071a. Read the comment docs.

[innerlee]

Note for future: the prs in these two months are too fragmented. Please consider group them for future models.

Ideally one algorithm involves 3-5 prs (one for pipelines, one for models, one for losses, and one for config & ckpts).
The recent ones are way too much. Its really painful to review. I don't know if you share the same feeling.

[innerlee]

The diff looks abnormal. Some changes have live in master. Please rebase master just in case.

[Yshuo-Li]

Note for future: the prs in these two months are too fragmented. Please consider group them for future models.

Ideally one algorithm involves 3-5 prs (one for pipelines, one for models, one for losses, and one for config & ckpts).
The recent ones are way too much. Its really painful to review. I don't know if you share the same feeling.

Its painful to rebase, too.

Sorry for that I mistakenly thought that the more detailed pr is better, and I will correct it in the future.