comfyanonymous and GitHub
eff2b9d412
Optimize nvfp4 lora applying. ( #11856 )
2026-01-13 19:37:19 -05:00
comfyanonymous and GitHub
15b312de7a
Optimize nvfp4 lora applying. ( #11854 )
2026-01-13 19:23:58 -05:00
comfyanonymous and GitHub
1dcbd9efaf
Bump ltxav mem estimation a bit. ( #11842 )
2026-01-13 01:42:07 -05:00
comfyanonymous and GitHub
117e7a5853
Refactor to try to lower mem usage. ( #11840 )
2026-01-12 21:01:52 -08:00
comfyanonymous and GitHub
b3c0e4de57
Make loras work on nvfp4 models. ( #11837 )
...
The initial applying is a bit slow but will probably be sped up in the
future.
2026-01-12 22:33:54 -05:00
comfyanonymous and GitHub
c881a1d689
Support the siglip 2 naflex model as a clip vision model. ( #11831 )
...
Not useful yet.
2026-01-12 17:05:54 -05:00
comfyanonymous and GitHub
c6238047ee
Put more details about portable in readme. ( #11816 )
2026-01-11 21:11:53 -05:00
comfyanonymous and GitHub
2f642d5d9b
Fix chroma fp8 te being treated as fp16. ( #11795 )
2026-01-10 14:40:42 -08:00
comfyanonymous and GitHub
cd912963f1
Fix issue with t5 text encoder in fp4. ( #11794 )
2026-01-10 17:31:31 -05:00
comfyanonymous and GitHub
dc202a2e51
Properly save mixed ops. ( #11772 )
2026-01-10 02:03:57 -05:00
comfyanonymous and GitHub
bd0e6825e8
Be less strict when loading mixed ops weights. ( #11769 )
2026-01-09 14:21:06 -05:00
comfyanonymous and GitHub
b48d6a83d4
Fix csp error in frontend when forcing offline. ( #11749 )
2026-01-08 22:15:50 -05:00
comfyanonymous and GitHub
1a20656448
Fix import issue. ( #11746 )
2026-01-08 17:23:59 -05:00
comfyanonymous and GitHub
0f11869d55
Better detection if AMD torch compiled with efficient attention. ( #11745 )
2026-01-08 17:16:58 -05:00
comfyanonymous and GitHub
50d6e1caf4
Tweak ltxv vae mem estimation. ( #11722 )
2026-01-07 23:07:05 -05:00
comfyanonymous and GitHub
21e8425087
Add warning for old pytorch. ( #11718 )
2026-01-07 21:07:26 -05:00
comfyanonymous and GitHub
25bc1b5b57
Add memory estimation function to ltxav text encoder. ( #11716 )
2026-01-07 20:11:22 -05:00
comfyanonymous and GitHub
3cd19e99c1
Increase ltxav mem estimation by a bit. ( #11715 )
2026-01-07 20:04:56 -05:00
comfyanonymous and GitHub
007b87e7ac
Bump required comfy-kitchen version. ( #11714 )
2026-01-07 19:48:47 -05:00
comfyanonymous and GitHub
34751fe9f9
Lower ltxv text encoder vram use. ( #11713 )
2026-01-07 19:12:15 -05:00
comfyanonymous and GitHub
3cd7b32f1b
Support gemma 12B with quant weights. ( #11696 )
2026-01-07 05:15:14 -05:00
comfyanonymous and GitHub
c0c9720d77
Fix stable release workflow not pulling latest comfy kitchen. ( #11695 )
2026-01-07 04:48:28 -05:00
comfyanonymous and GitHub
b7d7cc1d49
Fix fp8 fast issue. ( #11688 )
2026-01-07 01:39:06 -05:00
comfyanonymous and GitHub
c5cfb34c07
Update comfy-kitchen version to 0.2.3 ( #11685 )
2026-01-06 23:51:45 -05:00
comfyanonymous and GitHub
edee33f55e
Disable comfy kitchen cuda if pytorch cuda less than 13 ( #11681 )
2026-01-06 22:13:43 -05:00
comfyanonymous and GitHub
2c03884f5f
Skip fp4 matrix mult on devices that don't support it. ( #11677 )
2026-01-06 18:07:26 -05:00
comfyanonymous and GitHub
6e9ee55cdd
Disable ltxav previews. ( #11676 )
2026-01-06 17:41:27 -05:00
comfyanonymous and GitHub
023cf13721
Fix lowvram issue with ltxv2 text encoder. ( #11675 )
2026-01-06 17:33:03 -05:00
comfyanonymous and GitHub
c3c3e93c5b
Use rope functions from comfy kitchen. ( #11674 )
2026-01-06 16:57:50 -05:00
comfyanonymous and GitHub
6ffc159bdd
Update comfy-kitchen version to 0.2.1 ( #11672 )
2026-01-06 15:53:43 -05:00
comfyanonymous and GitHub
96e0d0924e
Add helpful message to portable. ( #11671 )
2026-01-06 14:43:24 -05:00
comfyanonymous and GitHub
1618002411
Revert "Use rope functions from comfy kitchen. ( #11647 )" ( #11648 )
...
This reverts commit 6ef85c4915
2026-01-05 23:07:39 -05:00
comfyanonymous and GitHub
6ef85c4915
Use rope functions from comfy kitchen. ( #11647 )
2026-01-05 22:50:35 -05:00
6da00dd899
Initial ops changes to use comfy_kitchen: Initial nvfp4 checkpoint support. ( #11635 )
...
---------
Co-authored-by: Jedrzej Kosinski <kosinkadink1@gmail.com >
2026-01-05 21:48:58 -05:00
comfyanonymous and GitHub
4f3f9e72a9
Fix name. ( #11638 )
2026-01-05 02:41:23 -08:00
comfyanonymous and GitHub
d157c3299d
Refactor module_size function. ( #11637 )
2026-01-05 03:48:31 -05:00
comfyanonymous and GitHub
d1b9822f74
Add LTXAVTextEncoderLoader node. ( #11634 )
2026-01-05 02:27:31 -05:00
comfyanonymous and GitHub
f2b002372b
Support the LTXV 2 model. ( #11632 )
2026-01-05 01:58:59 -05:00
comfyanonymous and GitHub
38d0493825
Fix case where upscale model wouldn't be moved to cpu. ( #11633 )
2026-01-04 19:13:50 -05:00
comfyanonymous and GitHub
53e762a3af
Print memory summary on OOM to help with debugging. ( #11613 )
2026-01-03 22:28:38 -05:00
comfyanonymous and GitHub
9a552df898
Remove leftover scaled_fp8 key. ( #11603 )
2026-01-02 17:28:10 -08:00
comfyanonymous and GitHub
65cfcf5b1b
New Year ruff cleanup. ( #11595 )
2026-01-01 22:06:14 -05:00
comfyanonymous and GitHub
1bdc9a947f
Remove duplicate import of model_management ( #11587 )
2025-12-31 19:29:55 -05:00
comfyanonymous and GitHub
d622a61874
Refactor: move clip_preprocess to comfy.clip_model ( #11586 )
2025-12-31 17:38:36 -05:00
comfyanonymous and GitHub
0e6221cc79
Add some warnings for pin and unpin errors. ( #11561 )
2025-12-29 18:26:42 -05:00
comfyanonymous and GitHub
8fd07170f1
Comment out unused norm_final in lumina/z image model. ( #11545 )
2025-12-28 22:07:25 -05:00
comfyanonymous and GitHub
2943093a53
Enable async offload by default for AMD. ( #11534 )
2025-12-27 18:54:15 -05:00
comfyanonymous and GitHub
1e4e342f54
Fix noise with ancestral samplers when inferencing on cpu. ( #11528 )
2025-12-26 22:03:01 -05:00
comfyanonymous and GitHub
d9a76cf66e
Specify in readme that we only support pytorch 2.4 and up. ( #11512 )
2025-12-25 23:46:51 -05:00
comfyanonymous and GitHub
532e285079
Add a ManualSigmas node. ( #11499 )
...
Can be used to manually set the sigmas for a model.
This node accepts a list of integer and floating point numbers separated
with any non numeric character.
2025-12-24 19:09:37 -05:00
comfyanonymous and GitHub
33aa808713
Make denoised output on custom sampler nodes work with nested tensors. ( #11471 )
2025-12-22 16:43:24 -05:00
comfyanonymous and GitHub
91bf6b6aa3
Add node to create empty latents for qwen image layered model. ( #11460 )
2025-12-21 19:59:40 -05:00
comfyanonymous and GitHub
807538fe6c
Core release process. ( #11447 )
2025-12-20 20:02:02 -05:00
comfyanonymous and GitHub
fb478f679a
Only apply gemma quant config to gemma model for newbie. ( #11436 )
2025-12-20 01:02:43 -05:00
comfyanonymous and GitHub
31e961736a
Fix issue with batches and newbie. ( #11435 )
2025-12-20 00:23:51 -05:00
comfyanonymous and GitHub
3ab9748903
Disable prompt weights on newbie te. ( #11434 )
2025-12-20 00:19:47 -05:00
comfyanonymous and GitHub
809ce68749
Support nested tensor denoise masks. ( #11431 )
2025-12-19 19:59:25 -05:00
comfyanonymous and GitHub
894802b0f9
Add LatentCutToBatch node. ( #11411 )
2025-12-18 22:21:40 -05:00
comfyanonymous and GitHub
28eaab608b
Diffusion model part of Qwen Image Layered. ( #11408 )
...
Only thing missing after this is some nodes to make using it easier.
2025-12-18 20:21:14 -05:00
comfyanonymous and GitHub
6a2678ac65
Trim/pad channels in VAE code. ( #11406 )
2025-12-18 18:22:38 -05:00
comfyanonymous and GitHub
e4fb3a3572
Support loading Wan/Qwen VAEs with different in/out channels. ( #11405 )
2025-12-18 17:45:33 -05:00
comfyanonymous and GitHub
bf7dc63bd6
skip_load_model -> force_full_load ( #11390 )
...
This should be a bit more clear and less prone to potential breakage if the
logic of the load models changes a bit.
2025-12-17 23:29:32 -05:00
comfyanonymous and GitHub
16d85ea133
Better handle torch being imported by prestartup nodes. ( #11383 )
2025-12-17 19:43:18 -05:00
comfyanonymous and GitHub
ffdd53b327
Check state dict key to auto enable the index_timestep_zero ref method. ( #11362 )
2025-12-16 17:03:17 -05:00
comfyanonymous and GitHub
bc606d7d64
Add a way to set the default ref method in the qwen image code. ( #11349 )
2025-12-16 01:26:55 -05:00
comfyanonymous and GitHub
645ee1881e
Inpainting for z image fun control. Use the ZImageFunControlnet node. ( #11346 )
...
image -> control image ex: pose
inpaint_image -> image for inpainting
mask -> inpaint mask
2025-12-15 23:38:12 -05:00
comfyanonymous and GitHub
683569de55
Only enable fp16 on ZImage on newer pytorch. ( #11344 )
2025-12-15 22:33:27 -05:00
comfyanonymous and GitHub
41bcf0619d
Add code to detect if a z image fun controlnet is broken or not. ( #11341 )
2025-12-15 20:51:06 -05:00
comfyanonymous and GitHub
70541d4e77
Support the new qwen edit 2511 reference method. ( #11340 )
...
index_timestep_zero can be selected in the
FluxKontextMultiReferenceLatentMethod now with the display name set to the
more generic "Edit Model Reference Method" node.
2025-12-15 19:20:34 -05:00
comfyanonymous and GitHub
5cb1e0c9a0
Disable guards on transformer_options when torch.compile ( #11317 )
2025-12-15 16:49:29 -05:00
comfyanonymous and GitHub
5ac3b26a7d
Update warning for old pytorch version. ( #11319 )
...
Versions below 2.4 are no longer supported. We will not break support on purpose but will not fix it if we do.
2025-12-14 04:02:50 -05:00
comfyanonymous and GitHub
971cefe7d4
Fix pytorch warnings. ( #11314 )
2025-12-13 18:45:23 -05:00
comfyanonymous and GitHub
da2bfb5b0a
Basic implementation of z image fun control union 2.0 ( #11304 )
...
The inpaint part is currently missing and will be implemented later.
I think they messed up this model pretty bad. They added some
control_noise_refiner blocks but don't actually use them. There is a typo
in their code so instead of doing control_noise_refiner -> control_layers
it runs the whole control_layers twice.
Unfortunately they trained with this typo so the model works but is kind
of slow and would probably perform a lot better if they corrected their
code and trained it again.
2025-12-13 01:39:11 -05:00
comfyanonymous and GitHub
c5a47a1692
Fix bias dtype issue in mixed ops. ( #11293 )
2025-12-12 11:49:35 -05:00
comfyanonymous and GitHub
5495589db3
Respect the dtype the op was initialized in for non quant mixed op. ( #11282 )
2025-12-11 23:32:27 -05:00
comfyanonymous and GitHub
338d9ae3bb
Make portable updater work with repos in unmerged state. ( #11281 )
2025-12-11 18:56:33 -05:00
comfyanonymous and GitHub
eeb020b9b7
Better chroma radiance and other models vram estimation. ( #11278 )
2025-12-11 17:33:09 -05:00
comfyanonymous and GitHub
ae65433a60
This only works on radiance. ( #11277 )
2025-12-11 17:15:00 -05:00
comfyanonymous and GitHub
fdebe18296
Fix regular chroma radiance ( #11276 )
2025-12-11 17:09:35 -05:00
comfyanonymous and GitHub
f8321eb57b
Adjust memory usage factor. ( #11257 )
2025-12-11 01:30:31 -05:00
comfyanonymous and GitHub
17c92a9f28
Tweak Z Image memory estimation. ( #11254 )
2025-12-10 19:59:48 -05:00
comfyanonymous and GitHub
791e30ff50
Fix nan issue when quantizing fp16 tensor. ( #11213 )
2025-12-09 17:03:21 -05:00
comfyanonymous and GitHub
d50f342c90
Fix potential issue. ( #11201 )
2025-12-08 23:20:04 -05:00
comfyanonymous and GitHub
3b0368aa34
Fix regression. ( #11194 )
2025-12-08 17:38:36 -05:00
comfyanonymous and GitHub
56fa7dbe38
Properly load the newbie diffusion model. ( #11172 )
...
There is still one of the text encoders missing and I didn't actually test it.
2025-12-07 07:44:55 -05:00
comfyanonymous and GitHub
329480da5a
Fix qwen scaled fp8 not working with kandinsky. Make basic t2i wf work. ( #11162 )
2025-12-06 17:50:10 -08:00
comfyanonymous and GitHub
50ca97e776
Speed up lora compute and lower memory usage by doing it in fp16. ( #11161 )
2025-12-06 18:36:20 -05:00
comfyanonymous and GitHub
d7a0aef650
Set OCL_SET_SVM_SIZE on AMD. ( #11139 )
2025-12-06 00:15:21 -05:00
comfyanonymous and GitHub
ae676ed105
Fix regression. ( #11137 )
2025-12-05 23:01:19 -05:00
comfyanonymous and GitHub
092ee8a500
Fix some custom nodes. ( #11134 )
2025-12-05 18:25:31 -05:00
comfyanonymous and GitHub
6fd463aec9
Fix regression when text encoder loaded directly on GPU. ( #11129 )
2025-12-05 15:33:16 -05:00
comfyanonymous and GitHub
43071e3de3
Make old scaled fp8 format use the new mixed quant ops system. ( #11000 )
2025-12-05 14:35:42 -05:00
comfyanonymous and GitHub
35fa091340
Forgot to put this in README. ( #11112 )
2025-12-04 22:52:09 -05:00
comfyanonymous and GitHub
ea17add3c6
Fix case where text encoders where running on the CPU instead of GPU. ( #11095 )
2025-12-03 23:15:15 -05:00
comfyanonymous and GitHub
ecdc8697d5
Qwen Image Lora training fix from #11090 ( #11094 )
2025-12-03 22:49:28 -05:00
comfyanonymous and GitHub
861817d22d
Fix issue with portable updater. ( #11070 )
...
This should fix the problem with the portable updater not working with portables created from a separate branch on the repo.
This does not affect any current portables who were all created on the master branch.
2025-12-03 00:47:51 -05:00
comfyanonymous and GitHub
b94d394a64
Support Z Image alibaba pai fun controlnets. ( #11062 )
...
These are not actual controlnets so put it in the models/model_patches
folder and use the ModelPatchLoader + QwenImageDiffsynthControlnet node to
use it.
2025-12-02 21:38:31 -05:00
comfyanonymous and GitHub
daaceac769
Hack to make zimage work in fp16. ( #11057 )
2025-12-02 17:11:58 -05:00
comfyanonymous and GitHub
878db3a727
Implement the Ovis image model. ( #11030 )
2025-12-01 20:56:17 -05:00
comfyanonymous and GitHub
2640acb31c
Update qwen tokenizer to add qwen 3 tokens. ( #11029 )
...
Doesn't actually change anything for current workflows because none of the
current models have a template with the think tokens.
2025-12-01 17:13:48 -05:00
comfyanonymousandGitHub
