parent
39c867f7ff
commit
ee7d2514c0
1 changed files with 48 additions and 33 deletions
@ -1,65 +1,80 @@ |
||||
# disco-diffusion |
||||
# Disco Diffusion |
||||
|
||||
v1 Update: Oct 29th 2021 - Somnai |
||||
<a href="https://colab.research.google.com/github/alembics/disco-diffusion/blob/main/Disco_Diffusion.ipynb" target="_parent"><img src="https://colab.research.google.com/assets/colab-badge.svg" alt="Open in Colab"/></a> |
||||
|
||||
QoL improvements added by Somnai (@somnai_dreams), including user friendly UI, settings+prompt saving and improved google drive folder organization. |
||||
A frankensteinian amalgamation of notebooks, models and techniques for the generation of AI Art and Animations. |
||||
|
||||
v1.1 Update: Nov 13th 2021 - Somnai |
||||
## Changelog |
||||
#### v1 Oct 29th 2021 - Somnai |
||||
|
||||
* QoL improvements added by Somnai (@somnai_dreams), including user friendly UI, settings+prompt saving and improved google drive folder organization. |
||||
|
||||
Now includes sizing options, intermediate saves and fixed image prompts and perlin inits. unexposed batch option since it doesn't work |
||||
#### v1.1 Nov 13th 2021 - Somnai |
||||
* Now includes sizing options, intermediate saves and fixed image prompts and perlin inits. unexposed batch option since it doesn't work |
||||
|
||||
v2 Update: Nov 22nd 2021 - Somnai |
||||
#### v2 Update: Nov 22nd 2021 - Somnai |
||||
|
||||
Initial addition of Katherine Crowson's Secondary Model Method (https://colab.research.google.com/drive/1mpkrhOjoyzPeSWy2r7T8EYRaU7amYOOi#scrollTo=X5gODNAMEUCR) |
||||
* Initial addition of Katherine Crowson's Secondary Model Method (https://colab.research.google.com/drive/1mpkrhOjoyzPeSWy2r7T8EYRaU7amYOOi#scrollTo=X5gODNAMEUCR) |
||||
|
||||
Noticed settings were saving with the wrong name so corrected it. Let me know if you preferred the old scheme. |
||||
* Noticed settings were saving with the wrong name so corrected it. Let me know if you preferred the old scheme. |
||||
|
||||
v3 Update: Dec 24th 2021 - Somnai |
||||
#### v3 Update: Dec 24th 2021 - Somnai |
||||
|
||||
Implemented Dango's advanced cutout method |
||||
* Implemented Dango's advanced cutout method |
||||
|
||||
Added SLIP models, thanks to NeuralDivergent |
||||
* Added SLIP models, thanks to NeuralDivergent |
||||
|
||||
Fixed issue with NaNs resulting in black images, with massive help and testing from @Softology |
||||
* Fixed issue with NaNs resulting in black images, with massive help and testing from @Softology |
||||
|
||||
Perlin now changes properly within batches (not sure where this perlin_regen code came from originally, but thank you) |
||||
* Perlin now changes properly within batches (not sure where this perlin_regen code came from originally, but thank you) |
||||
|
||||
v4 Update: Jan 2021 - Somnai |
||||
#### v4 Update: Jan 2021 - Somnai |
||||
|
||||
Implemented Diffusion Zooming |
||||
* Implemented Diffusion Zooming |
||||
|
||||
Added Chigozie keyframing |
||||
* Added Chigozie keyframing |
||||
|
||||
Made a bunch of edits to processes |
||||
* Made a bunch of edits to processes |
||||
|
||||
v4.1 Update: Jan 14th 2021 - Somnai |
||||
#### v4.1 Update: Jan 14th 2021 - Somnai |
||||
|
||||
Added video input mode |
||||
* Added video input mode |
||||
|
||||
Added license that somehow went missing |
||||
* Added license that somehow went missing |
||||
|
||||
Added improved prompt keyframing, fixed image_prompts and multiple prompts |
||||
* Added improved prompt keyframing, fixed image_prompts and multiple prompts |
||||
|
||||
Improved UI |
||||
* Improved UI |
||||
|
||||
Significant under the hood cleanup and improvement |
||||
* Significant under the hood cleanup and improvement |
||||
|
||||
Refined defaults for each mode |
||||
* Refined defaults for each mode |
||||
|
||||
Added latent-diffusion SuperRes for sharpening |
||||
* Added latent-diffusion SuperRes for sharpening |
||||
|
||||
Added resume run mode |
||||
* Added resume run mode |
||||
|
||||
v4.9 Update: Feb 5th 2022 - gandamu / Adam Letts |
||||
#### v5 Update: Feb 20th 2022 - gandamu / Adam Letts |
||||
|
||||
Added 3D |
||||
* Added 3D animation mode. Uses weighted combination of AdaBins and MiDaS depth estimation models. Uses pytorch3d for 3D transforms on Colab and/or Linux. |
||||
|
||||
Added brightness corrections to prevent animation from steadily going dark over time |
||||
|
||||
v4.91 Update: Feb 19th 2022 - gandamu / Adam Letts |
||||
## Notebook Provenance |
||||
|
||||
Cleaned up 3D implementation and made associated args accessible via Colab UI elements |
||||
Original notebook by Katherine Crowson (https://github.com/crowsonkb, https://twitter.com/RiversHaveWings). It uses either OpenAI's 256x256 unconditional ImageNet or Katherine Crowson's fine-tuned 512x512 diffusion model (https://github.com/openai/guided-diffusion), together with CLIP (https://github.com/openai/CLIP) to connect text prompts with images. |
||||
|
||||
v4.92 Update: Feb 20th 2022 - gandamu / Adam Letts |
||||
Modified by Daniel Russell (https://github.com/russelldc, https://twitter.com/danielrussruss) to include (hopefully) optimal params for quick generations in 15-100 timesteps rather than 1000, as well as more robust augmentations. |
||||
|
||||
Separated transform code |
||||
Further improvements from Dango233 and nsheppard helped improve the quality of diffusion in general, and especially so for shorter runs like this notebook aims to achieve. |
||||
|
||||
Vark added code to load in multiple Clip models at once, which all prompts are evaluated against, which may greatly improve accuracy. |
||||
|
||||
The latest zoom, pan, rotation, and keyframes features were taken from Chigozie Nri's VQGAN Zoom Notebook (https://github.com/chigozienri, https://twitter.com/chigozienri) |
||||
|
||||
Advanced DangoCutn Cutout method is also from Dango223. |
||||
|
||||
-- |
||||
|
||||
Somnai (https://twitter.com/Somnai_dreams) added 2D Diffusion animation techniques, QoL improvements and various implementations of tech and techniques, mostly listed in the changelog below. |
||||
|
||||
3D animation implementation added by Adam Letts (https://twitter.com/gandamu_ml) in collaboration with Somnai. |
||||
|
Loading…
Reference in new issue