b2f03164c7
Prevent the clip_g position_ids key from being saved in the checkpoint.
...
This is to make it match the official checkpoint.
1 year ago
46dc050c9f
Fix potential tensors being on different devices issues.
1 year ago
90aa597099
Add back roundRect to fix issue on firefox ESR.
1 year ago
3e2309f149
fix mps miss import
1 year ago
f4b9390623
Add a random string to the temp prefix for PreviewImage.
1 year ago
2b2a1474f7
Move to litegraph.
1 year ago
cef30cc6b6
Merge branch 'hidpi-canvas' of https://github.com/EHfive/ComfyUI
1 year ago
880c9b928b
Update litegraph to latest.
1 year ago
05e6eac7b3
Scale graph canvas based on DPI factor
...
Similar to fixes in litegraph.js editor demo:
3ef215cf11/editor/js/code.js (L19-L28)Fixes #161
1 year ago
99abcbef41
feat/startup-script: Feature to avoid package installation errors when installing custom nodes. ( #856 )
...
* support startup script for installation without locking on windows
* modified: Instead of executing scripts from the startup-scripts directory, I will change it to execute the prestartup_script.py for each custom node.
1 year ago
606a537090
Support SDXL embedding format with 2 CLIP.
1 year ago
5797ff89b0
use relative paths for all web connections
...
This enables local reverse-proxies to host ComfyUI on a path, eg "http://example.com/ComfyUI/ " in such a way that at least everything I tested works. Without this patch, proxying ComfyUI in this way will yield errors.
1 year ago
6ad0a6d7e2
Don't patch weights when multiplier is zero.
1 year ago
af15add967
Fix annoyance with textbox unselecting in chromium.
1 year ago
d5323d16e0
latent2rgb matrix for SDXL.
1 year ago
0ae81c03bb
Empty cache after model unloading for normal vram and lower.
1 year ago
d3f5998218
Support loading clip_g from diffusers in CLIP Loader nodes.
1 year ago
a9a4ba7574
Fix merging not working when model2 of model merge node was a merge.
1 year ago
febea8c101
Merge branch 'bugfix/img-offset' of https://github.com/ltdrdata/ComfyUI
1 year ago
9caab9380d
fix: Image.ANTIALIAS is no longer available. ( #847 )
...
* modify deprecated api call
* prevent breaking old Pillow users
* change LANCZOS to BILINEAR
1 year ago
d43cff2105
bugfix: image widget's was mis-aligned when node has multiline widget
1 year ago
c2d407b0f7
Merge branch 'Yaruze66-patch-1' of https://github.com/Yaruze66/ComfyUI
1 year ago
bb5fbd29e9
Merge branch 'condmask-fix' of https://github.com/vmedea/ComfyUI
1 year ago
2c9d98f3e6
CLIPTextEncodeSDXL now works when prompts are of very different sizes.
1 year ago
e7bee85df8
Add arguments to run the VAE in fp16 or bf16 for testing.
1 year ago
f5232c4869
Fix 7z error when extracting package.
1 year ago
608fcc2591
Fix bug with weights when prompt is long.
1 year ago
ddc6f12ad5
Disable autocast in unet for increased speed.
1 year ago
603f02d613
Fix loras not working when loading checkpoint with config.
1 year ago
ccb1b25908
Add a conditioning concat node.
1 year ago
af7a49916b
Support loading unet files in diffusers format.
1 year ago
e57cba4c61
Add gpu variations of the sde samplers that are less deterministic
...
but faster.
1 year ago
f81b192944
Add logit scale parameter so it's present when saving the checkpoint.
1 year ago
acf95191ff
Properly support SDXL diffusers loras for unet.
1 year ago
c61a95f9f7
Fix size check for conditioning mask
...
The wrong dimensions were being checked, [1] and [2] are the image size.
not [2] and [3]. This results in an out-of-bounds error if one of them
actually matches.
1 year ago
8d694cc450
Fix issue with OSX.
1 year ago
c02f3baeaf
Now the model merge blocks node will use the longest match.
1 year ago
3a09fac835
ConditioningAverage now also averages the pooled output.
1 year ago
d94ddd8548
Add text encode nodes to control the extra parameters in SDXL.
1 year ago
c3e96e637d
Pass device to CLIP model.
1 year ago
5e6bc824aa
Allow passing custom path to clip-g and clip-h.
1 year ago
dc9d1f31c8
Improvements for OSX.
1 year ago
9ae6ff65bc
Update extra_model_paths.yaml.example: add RealESRGAN path
1 year ago
103c487a89
Cleanup.
1 year ago
ae948b42fa
Add taesd weights to standalones.
1 year ago
2c4e0b49b7
Switch to fp16 on some cards when the model is too big.
1 year ago
6f3d9f52db
Add a --force-fp16 argument to force fp16 for testing.
1 year ago
1c1b0e7299
--gpu-only now keeps the VAE on the device.
1 year ago
ce35d8c659
Lower latency by batching some text encoder inputs.
1 year ago
3b6fe51c1d
Leave text_encoder on the CPU when it can handle it.
1 year ago
comfyanonymous
KarryCharon
Huang-Huang Bao
Dr.Lt.Data
Alex "mcmonkey" Goodwin
Dr.Lt.Data
mara
Yaruze66