Commits · d8e58f0a7ea914377c56f9b32e449ef4c65da572 · chenpangpang / ComfyUI

06 Aug, 2023 1 commit
- Detect hint_channels from controlnet. · d8e58f0a
  comfyanonymous authored Aug 06, 2023
  
  d8e58f0a
05 Aug, 2023 1 commit
- Support loras in diffusers format. · c5d7593c
  comfyanonymous authored Aug 05, 2023
  
  c5d7593c
04 Aug, 2023 2 commits
- Add CMP 30HX card to the nvidia_16_series list. · 1ce0d8ad
  comfyanonymous authored Aug 04, 2023
  
  1ce0d8ad
- Make sure the pooled output stays at the EOS token with added embeddings. · c99d8002
  comfyanonymous authored Aug 03, 2023
  
  c99d8002
31 Jul, 2023 1 commit
- Only shift text encoder to vram when CPU cores are under 8. · 4a77fcd6
  comfyanonymous authored Jul 31, 2023
  
  4a77fcd6
30 Jul, 2023 2 commits
- Lower CPU thread check for running the text encoder on the CPU vs GPU. · 3cd31d0e
  comfyanonymous authored Jul 30, 2023
  
  3cd31d0e
- Remove some useless code. · 2b139390
  comfyanonymous authored Jul 30, 2023
  
  2b139390
29 Jul, 2023 3 commits
- Faster VAE loading. · 95d796fc
  comfyanonymous authored Jul 29, 2023
  
  95d796fc
- Initialize the unet directly on the target device. · 4b957a00
  comfyanonymous authored Jul 29, 2023
  
  4b957a00
- Remove unused code and torchdiffeq dependency. · c910b4a0
  comfyanonymous authored Jul 28, 2023
  
  c910b4a0
28 Jul, 2023 1 commit
- Add --disable-metadata argument to disable saving metadata in files. · 1141029a
  comfyanonymous authored Jul 28, 2023
  
  1141029a
27 Jul, 2023 2 commits
- Remove some prints. · 68be24ee
  comfyanonymous authored Jul 27, 2023
  
  68be24ee
- Fix timestep ranges when batch_size > 1 · 1ea4d846
  asagi4 authored Jul 27, 2023
  
  1ea4d846
26 Jul, 2023 1 commit
- Fix diffusers VAE loading. · 5379051d
  comfyanonymous authored Jul 26, 2023
  
  5379051d
25 Jul, 2023 2 commits
- Fix some new loras. · 727588d0
  comfyanonymous authored Jul 25, 2023
  
  727588d0
- Fix potential issue with Save Checkpoint. · 4f9b6f39
  comfyanonymous authored Jul 25, 2023
  
  4f9b6f39
24 Jul, 2023 4 commits
- Start is now 0.0 and end is now 1.0 for the timestep ranges. · 5f75d784
  comfyanonymous authored Jul 24, 2023
  
  5f75d784
- ControlNetApplyAdvanced can now define when controlnet gets applied. · 7ff14b62
  comfyanonymous authored Jul 24, 2023
  
  7ff14b62
- Add a ControlNetApplyAdvanced node. · d191c4f9
  comfyanonymous authored Jul 24, 2023
```
The controlnet can be applied to the positive or negative prompt only by
connecting it correctly.
```
  d191c4f9
- Add a way to set which range of timesteps the cond gets applied to. · 0240946e
  comfyanonymous authored Jul 24, 2023
  
  0240946e
23 Jul, 2023 1 commit
- Try to fix memory issue with lora. · 22f29d66
  comfyanonymous authored Jul 22, 2023
  
  22f29d66
22 Jul, 2023 4 commits
- Nodes can now patch the unet function. · 67be7eb8
  comfyanonymous authored Jul 22, 2023
  
  67be7eb8
- Del the right object when applying lora. · 12a6e931
  comfyanonymous authored Jul 22, 2023
  
  12a6e931
- Support controlnet in diffusers format. · 78e7958d
  comfyanonymous authored Jul 21, 2023
  
  78e7958d
- Fix issue with lora in some cases when combined with model merging. · 09386a36
  comfyanonymous authored Jul 21, 2023
  
  09386a36
21 Jul, 2023 1 commit
- Properly support SDXL diffusers unet with UNETLoader node. · 58b2364f
  comfyanonymous authored Jul 21, 2023
  
  58b2364f
20 Jul, 2023 1 commit
- Print errors and continue when lora weights are not compatible. · 01150186
  comfyanonymous authored Jul 20, 2023
  
  01150186
19 Jul, 2023 4 commits
- Fix typo. · 0b284f65
  comfyanonymous authored Jul 19, 2023
  
  0b284f65
- Fix ddim issue with older torch versions. · e032ca61
  comfyanonymous authored Jul 19, 2023
  
  e032ca61
- Add MX450 and MX550 to list of cards with broken fp16. · 18885f80
  comfyanonymous authored Jul 19, 2023
  
  18885f80
- It's actually possible to torch.compile the unet now. · 9ba44099
  comfyanonymous authored Jul 18, 2023
  
  9ba44099
18 Jul, 2023 1 commit
- Add key to indicate checkpoint is v_prediction when saving. · 51d54775
  comfyanonymous authored Jul 18, 2023
  
  51d54775
17 Jul, 2023 7 commits
- Fix device print on old torch version. · ff6b047a
  comfyanonymous authored Jul 17, 2023
  
  ff6b047a
- Enable --cuda-malloc by default on torch 2.0 and up. · 9871a15c
  comfyanonymous authored Jul 17, 2023
```
Add --disable-cuda-malloc to disable it.
```
  9871a15c
- --windows-standalone-build now enables --cuda-malloc · 55d0fca9
  comfyanonymous authored Jul 17, 2023
  
  55d0fca9
- Add a command line argument to enable backend:cudaMallocAsync · 1679abd8
  comfyanonymous authored Jul 17, 2023
  
  1679abd8
- Only calculate randn in some samplers when it's actually being used. · 3a150bad
  comfyanonymous authored Jul 17, 2023
  
  3a150bad
- Fix regression with ddim and uni_pc when batch size > 1. · ee8f8ee0
  comfyanonymous authored Jul 17, 2023
  
  ee8f8ee0
- Refactor of sampler code to deal more easily with different model types. · 3ded1a3a
  comfyanonymous authored Jul 17, 2023
  
  3ded1a3a
16 Jul, 2023 1 commit
- Lower lora ram usage when in normal vram mode. · 5f573626
  comfyanonymous authored Jul 16, 2023
  
  5f573626