Commits · 4f9b6f39d12082c1ee310b232ef093d1f048af64 · chenpangpang / ComfyUI · GitLab

25 Jul, 2023 2 commits
- Fix potential issue with Save Checkpoint. · 4f9b6f39
  comfyanonymous authored Jul 25, 2023
  
  4f9b6f39
- Disable cuda malloc on a bunch of quadro cards. · 7c0a5a3e
  comfyanonymous authored Jul 25, 2023
  
  7c0a5a3e
24 Jul, 2023 5 commits
- Use bigger tiles when upscaling with model and fallback on OOM. · a51f33ee
  comfyanonymous authored Jul 24, 2023
  
  a51f33ee
- 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 2 commits
- Disable cuda malloc on all the 9xx series. · 30de083d
  comfyanonymous authored Jul 23, 2023
  
  30de083d
- Try to fix memory issue with lora. · 22f29d66
  comfyanonymous authored Jul 22, 2023
  
  22f29d66
22 Jul, 2023 5 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
- Disable cuda malloc on regular GTX 960. · 85a8900a
  comfyanonymous authored Jul 22, 2023
  
  85a8900a
- 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 2 commits
- Print errors and continue when lora weights are not compatible. · 01150186
  comfyanonymous authored Jul 20, 2023
  
  01150186
- Merge branch 'fix-AttributeError-module-'torch'-has-no-attribute-'mps'' of... · 4760c293
  comfyanonymous authored Jul 20, 2023
```
Merge branch 'fix-AttributeError-module-'torch'-has-no-attribute-'mps'' of https://github.com/KarryCharon/ComfyUI
```
  4760c293
19 Jul, 2023 8 commits
- Move image encoding outside of sampling loop for better preview perf. · ccb6b70d
  comfyanonymous authored Jul 19, 2023
  
  ccb6b70d
- Disable cuda malloc on GTX 750 Ti. · 39c58b22
  comfyanonymous authored Jul 19, 2023
  
  39c58b22
- Update how to get the prompt in api format in the example. · d5c0765f
  comfyanonymous authored Jul 19, 2023
  
  d5c0765f
- Auto disable cuda malloc on some GPUs on windows. · 799c08a4
  comfyanonymous authored Jul 19, 2023
  
  799c08a4
- 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 3 commits
- Merge branch 'master' of https://github.com/ComfyUI-Community/ComfyUI · ac9c038a
  comfyanonymous authored Jul 16, 2023
  
  ac9c038a
- Lower lora ram usage when in normal vram mode. · 5f573626
  comfyanonymous authored Jul 16, 2023
  
  5f573626
- Patch del self.loaded_lora to prevent error with persistent lora_name swapping · a8f3bbc3
  ComfyUI-Community authored Jul 15, 2023
  
  a8f3bbc3
15 Jul, 2023 3 commits
- Speed up lora loading a bit. · 490771b7
  comfyanonymous authored Jul 15, 2023
  
  490771b7
- Fix CLIPSetLastLayer not reverting when removed. · 50b1180d
  comfyanonymous authored Jul 15, 2023
  
  50b1180d
- Reduce floating point rounding errors in loras. · 6fb084f3
  comfyanonymous authored Jul 15, 2023
  
  6fb084f3
14 Jul, 2023 1 commit
- Add a node to merge CLIP models. · 91ed2815
  comfyanonymous authored Jul 14, 2023
  
  91ed2815