Commits · d44a2de49f797af6b1adc709e709d6b0b4d89093 · chenpangpang / ComfyUI

17 Oct, 2023 2 commits
- Make VAE code closer to sgm. · d44a2de4
  comfyanonymous authored Oct 17, 2023
  
  d44a2de4
- Refactor the attention stuff in the VAE. · 23680a91
  comfyanonymous authored Oct 17, 2023
  
  23680a91
12 Oct, 2023 2 commits
- pytorch_attention_enabled can now return True when xformers is enabled. · 88733c99
  comfyanonymous authored Oct 11, 2023
  
  88733c99
- Refactor the attention functions. · 1a4bd9e9
  comfyanonymous authored Oct 11, 2023
```
There's no reason for the whole CrossAttention object to be repeated when
only the operation in the middle changes.
```
  1a4bd9e9
04 Sep, 2023 1 commit
- Add a force argument to soft_empty_cache to force a cache empty. · 1938f5c5
  comfyanonymous authored Sep 04, 2023
  
  1938f5c5
29 Aug, 2023 1 commit
- Remove optimization that caused border. · bed116a1
  comfyanonymous authored Aug 29, 2023
  
  bed116a1
28 Aug, 2023 2 commits
- Fallback to slice attention if xformers doesn't support the operation. · 1c794a21
  comfyanonymous authored Aug 27, 2023
  
  1c794a21
- Make --bf16-vae work on torch 2.0 · d935ba50
  comfyanonymous authored Aug 27, 2023
  
  d935ba50
29 Jul, 2023 1 commit
- Faster VAE loading. · 95d796fc
  comfyanonymous authored Jul 29, 2023
  
  95d796fc
23 Jun, 2023 1 commit
- Remove useless code. · fa28d733
  comfyanonymous authored Jun 23, 2023
  
  fa28d733
20 May, 2023 2 commits
- Make scaled_dot_product switch to sliced attention on OOM. · b8636a44
  comfyanonymous authored May 20, 2023
  
  b8636a44
- Simplify and improve some vae attention code. · 797c4e8d
  comfyanonymous authored May 20, 2023
  
  797c4e8d
04 May, 2023 1 commit
- Fix imports. · bae4fb4a
  comfyanonymous authored May 04, 2023
  
  bae4fb4a
15 Apr, 2023 1 commit
- Fix model_management import so it doesn't get executed twice. · 73c3e11e
  comfyanonymous authored Apr 15, 2023
  
  73c3e11e
05 Apr, 2023 1 commit
- Disable xformers in VAE when xformers == 0.0.18 · e46b1c30
  comfyanonymous authored Apr 04, 2023
  
  e46b1c30
22 Mar, 2023 2 commits
- Try again with vae tiled decoding if regular fails because of OOM. · 3ed4a4e4
  comfyanonymous authored Mar 22, 2023
  
  3ed4a4e4
- Try to improve VAEEncode memory usage a bit. · c692509c
  comfyanonymous authored Mar 22, 2023
  
  c692509c
13 Mar, 2023 2 commits
- Add pytorch attention support to VAE. · 83f23f82
  comfyanonymous authored Mar 13, 2023
  
  83f23f82
- --disable-xformers should not even try to import xformers. · a256a2ab
  comfyanonymous authored Mar 13, 2023
  
  a256a2ab
12 Mar, 2023 1 commit

Xformers is now properly disabled when --cpu used. · 0f3ba748

comfyanonymous authored Mar 12, 2023

Added --windows-standalone-build option, currently it only opens
makes the code open up comfyui in the browser.

0f3ba748

11 Mar, 2023 1 commit
- Try to fix memory issue. · 1de86851
  comfyanonymous authored Mar 11, 2023
  
  1de86851
05 Mar, 2023 1 commit
- Make VAE use common function to get free memory. · cc8baf10
  comfyanonymous authored Mar 05, 2023
  
  cc8baf10
10 Feb, 2023 1 commit
- Use real softmax in split op to fix issue with some images. · 509c7dfc
  comfyanonymous authored Feb 10, 2023
  
  509c7dfc
09 Feb, 2023 3 commits
- Same thing but for the other places where it's used. · 773cdabf
  comfyanonymous authored Feb 09, 2023
  
  773cdabf
- Split optimization for VAE attention block. · e8c499dd
  comfyanonymous authored Feb 08, 2023
  
  e8c499dd
- Use inplace operations for less OOM issues. · 5b4e3127
  comfyanonymous authored Feb 08, 2023
  
  5b4e3127
17 Jan, 2023 1 commit
- Initial commit. · 220afe33
  comfyanonymous authored Jan 03, 2023
  
  220afe33