Commits · d37cfffd537cd29309afbcb192c4f979995c6a34 · novelai-storage / Stable Diffusion Webui · GitLab

23 Oct, 2022 1 commit
- added callback for creating new settings in extensions · d37cfffd
  AUTOMATIC authored Oct 22, 2022
  
  d37cfffd
22 Oct, 2022 10 commits
- added a guard for hypernet training that will stop early if weights are getting no gradients · 7fd90128
  AUTOMATIC authored Oct 22, 2022
  
  7fd90128
- fix for extensions without style.css · 1cd3ed7d
  AUTOMATIC authored Oct 22, 2022
  
  1cd3ed7d
- remove parsing command line from devices.py · 50b55044
  AUTOMATIC authored Oct 22, 2022
  
  50b55044
- Merge pull request #3377 from Extraltodeus/cuda-device-id-selection · e80bdcab
  AUTOMATIC1111 authored Oct 22, 2022
```
Implementation of CUDA device id selection (--device-id 0/1/2)
```
  e80bdcab
- Merge branch 'master' into cuda-device-id-selection · 1fa53dab
  AUTOMATIC1111 authored Oct 22, 2022
  
  1fa53dab
- updated readme with info about Aesthetic Gradients · 5aa95250
  AUTOMATIC authored Oct 22, 2022
  
  5aa95250
- further support for extensions · 6398dc9b
  AUTOMATIC authored Oct 22, 2022
  
  6398dc9b
- removed aesthetic gradients as built-in · 2b912516
  AUTOMATIC authored Oct 22, 2022
```
added support for extensions
```
  2b912516
- implement CUDA device selection, --device-id arg · 29bfacd6
  Extraltodeus authored Oct 22, 2022
  
  29bfacd6
- implement CUDA device selection by ID · 57eb54b8
  Extraltodeus authored Oct 22, 2022
  
  57eb54b8
21 Oct, 2022 29 commits
- Merge remote-tracking branch 'historytab/master' · 26d10737
  AUTOMATIC authored Oct 21, 2022
  
  26d10737
- prevent error spam when processing images without txt files for captions · f49c08ea
  AUTOMATIC authored Oct 21, 2022
  
  f49c08ea
- Merge pull request #3246 from Milly/fix/train-preprocess-keep-ratio · 7464f367
  AUTOMATIC1111 authored Oct 21, 2022
```
Preprocess: fixed keep ratio and changed split behavior
```
  7464f367
- Merge branch 'master' into fix/train-preprocess-keep-ratio · 5e9afa5c
  AUTOMATIC1111 authored Oct 21, 2022
  
  5e9afa5c
- make aspect ratio overlay work regardless of selected localization, pt2 · 24ce67a1
  AUTOMATIC authored Oct 21, 2022
  
  24ce67a1
- loading SD VAE, see PR #3303 · ac0aa2b1
  AUTOMATIC authored Oct 21, 2022
  
  ac0aa2b1
- batch_size does not affect job count · 3d898044
  AUTOMATIC authored Oct 21, 2022
  
  3d898044
- Merge remote-tracking branch 'mk2/outpainting-mk2-batch-out' · a7aa00d4
  AUTOMATIC authored Oct 21, 2022
  
  a7aa00d4
- make aspect ratio overlay work regardless of selected localization · 704036ff
  AUTOMATIC authored Oct 21, 2022
  
  704036ff
- Change option description of unload_models_when_training · 02e4d469
  Rcmcpe authored Oct 21, 2022
  
  02e4d469
- Remove unused variable. · 272fa527
  timntorres authored Oct 21, 2022
  
  272fa527
- Refer to Hypernet's name, sensibly, by its name variable. · fccad18a
  timntorres authored Oct 21, 2022
  
  fccad18a
- Match hypernet name with filename in all cases. · 19818f02
  timntorres authored Oct 21, 2022
  
  19818f02
- Sanitize hypernet name input. · 51e3dc9c
  timntorres authored Oct 21, 2022
  
  51e3dc9c
- Merge pull request #3321 from AUTOMATIC1111/features-to-readme · 3e12b529
  AUTOMATIC1111 authored Oct 21, 2022
```
Features to readme
```
  3e12b529
- Merge branch 'master' into features-to-readme · ec37f8a4
  AUTOMATIC1111 authored Oct 21, 2022
  
  ec37f8a4
- Make commit hash mandatory field · 85cb5918
  parsec501 authored Oct 21, 2022
  
  85cb5918
- Update image_embedding.py · 306e2ff6
  DepFA authored Oct 21, 2022
  
  306e2ff6
- Use opts in textual_inversion image_embedding.py for dynamic fonts · d0ea471b
  DepFA authored Oct 21, 2022
  
  d0ea471b
- make aestetic embedding ciompatible with prompts longer than 75 tokens · 9286fe53
  AUTOMATIC authored Oct 21, 2022
  
  9286fe53
- Merge remote-tracking branch 'origin/master' · e89e2f7c
  AUTOMATIC authored Oct 21, 2022
  
  e89e2f7c
- do not load aesthetic clip model until it's needed · df570640
  AUTOMATIC authored Oct 21, 2022
```
add refresh button for aesthetic embeddings
add aesthetic params to images' infotext
```
  df570640
- Update README.md · 003d2c7f
  ClashSAN authored Oct 21, 2022
  
  003d2c7f
- Merge branch 'ae' · 7d6b388d
  AUTOMATIC authored Oct 21, 2022
  
  7d6b388d
- wtf is happening · 1ed227b3
  Leo Mozoloa authored Oct 21, 2022
  
  1ed227b3
- Fix Hypernet infotext string split bug for PR #3283 · bf30673f
  AUTOMATIC authored Oct 21, 2022
  
  bf30673f
- turns out LayerNorm also has weight and bias and needs to be pre-multiplied... · 03a1e288
  AUTOMATIC authored Oct 21, 2022
```
turns out LayerNorm also has weight and bias and needs to be pre-multiplied and trained for hypernets
```
  03a1e288
- Merge pull request #3197 from AUTOMATIC1111/training-help-text · e4877722
  AUTOMATIC1111 authored Oct 21, 2022
```
Training UI Changes
```
  e4877722
- Merge branch 'master' into training-help-text · 0c5522ea
  AUTOMATIC1111 authored Oct 21, 2022
  
  0c5522ea