comfyanonymous/ComfyUI
1
0
Fork 0
mirror of https://github.com/comfyanonymous/ComfyUI.git synced 2026-04-26 09:29:07 +00:00
Code Issues Packages Projects Releases Wiki Activity
5,182 Commits 179 Branches 143 Tags
worksplit-multigpu
Commit Graph

5182 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Terry Jia
50076f3439 format blueprint (#13315) 
Co-authored-by: guill <jacob.e.segal@gmail.com>
 2026-04-06 23:33:55 -04:00
comfyanonymous
61c2387436 Ace step empty latent nodes follow intermediate dtype. (#13313) 2026-04-06 18:12:16 -07:00
Terry Jia
7083484a48 image histogram node (#13153) 
* image histogram node

* update color curve blueprint using image histogram node

---------

Co-authored-by: guill <jacob.e.segal@gmail.com>
 2026-04-06 14:54:02 -07:00
comfyanonymous
4b1444fc7a Update README.md with new frontend release cycle. (#13301) 2026-04-05 16:37:27 -07:00
Daxiong (Lin)
8cbbea8f6a chore: update workflow templates to v0.9.44 (#13290) 2026-04-05 13:31:11 +08:00
comfyanonymous
13917b3880 Nightly Nvidia pytorch is now cu132 (#13288) 2026-04-04 16:02:47 -07:00
comfyanonymous
f21f6b2212 Add portable release for intel XPU. (#13272) 2026-04-03 15:29:06 -04:00
Daxiong (Lin)
eb0686bbb6 Update template to 0.9.43 (#13265) 2026-04-02 23:52:10 -07:00
Alexander Piskun
5de94e70ec feat(api-nodes): new Partner nodes for Wan2.7 (#13264) 
Signed-off-by: bigcat88 <bigcat88@icloud.com>
 2026-04-02 23:51:47 -07:00
comfyanonymous
76b75f3ad7 Fix some issue with insecure browsers. (#13261) 
If you are on a recent chromium or chrome based browser this doesn't affect you.

This is to give time for the lazy firefox devs to implement PNA.
 2026-04-02 16:39:34 -04:00
comfyanonymous
0c63b4f6e3 Remove dead code. (#13251) 2026-04-01 20:22:06 -04:00
Daxiong (Lin)
7d437687c2 chore: update workflow templates to v0.9.41 (#13242) 2026-03-31 20:23:25 -07:00
comfyanonymous
e2ddf28d78 Fix some fp8 scaled checkpoints no longer working. (#13239) 2026-03-31 14:27:17 -07:00
comfyanonymous
076639fed9 Update README with note on model support (#13235) 
Added note about additional supported models in ComfyUI.
 2026-03-30 23:11:02 -04:00
Jedrzej Kosinski
b418fb1582 Fix device mismatch: update LoadedModel.device when _switch_parent swaps to parent patcher 
When a multigpu clone ModelPatcher is garbage collected, LoadedModel._switch_parent
switches the weakref to point at the parent (main) ModelPatcher. However, it was not
updating LoadedModel.device, leaving it with the old clone's device (e.g., cuda:1).
On subsequent runs, this stale device was passed to ModelPatcherDynamic.load(), causing
an assertion failure (device_to != self.load_device).

Amp-Thread-ID: https://ampcode.com/threads/T-019d3f5c-28c5-72c9-abed-34681f1b54ba
Co-authored-by: Amp <amp@ampcode.com>
 2026-03-30 08:59:38 -07:00
Jedrzej Kosinski
20803749c3 Add detailed multigpu debug logging to load_models_gpu 
Amp-Thread-ID: https://ampcode.com/threads/T-019d3f5c-28c5-72c9-abed-34681f1b54ba
Co-authored-by: Amp <amp@ampcode.com>
 2026-03-30 08:53:36 -07:00
Jedrzej Kosinski
3fab720be9 Add debug logging for device mismatch in ModelPatcherDynamic.load 
Amp-Thread-ID: https://ampcode.com/threads/T-019d3f5c-28c5-72c9-abed-34681f1b54ba
Co-authored-by: Amp <amp@ampcode.com>
 2026-03-30 08:45:55 -07:00
Jedrzej Kosinski
afdddcee66 Re-enable comfy-kitchen cuda backend for multigpu testing 
Amp-Thread-ID: https://ampcode.com/threads/T-019d3f5c-28c5-72c9-abed-34681f1b54ba
Co-authored-by: Amp <amp@ampcode.com>
 2026-03-30 08:32:52 -07:00
Jedrzej Kosinski
1d8e379f41 Rename MultiGPU Work Units to MultiGPU CFG Split 
Amp-Thread-ID: https://ampcode.com/threads/T-019d3ee9-19d5-767a-9d7a-e50cbbef815b
Co-authored-by: Amp <amp@ampcode.com>
 2026-03-30 08:00:20 -07:00
Jedrzej Kosinski
5f4fcd19e7 Simplify multigpu nodes: default max_gpus=2, remove gpu_options input, disable Options node 
Amp-Thread-ID: https://ampcode.com/threads/T-019d3ee9-19d5-767a-9d7a-e50cbbef815b
Co-authored-by: Amp <amp@ampcode.com>
 2026-03-30 07:30:32 -07:00
Jedrzej Kosinski
d52dcbc88f Rewrite multigpu nodes to V3 format 
Amp-Thread-ID: https://ampcode.com/threads/T-019d3ee9-19d5-767a-9d7a-e50cbbef815b
Co-authored-by: Amp <amp@ampcode.com>
 2026-03-30 07:23:13 -07:00
Jedrzej Kosinski
84f465e791 Set CUDA device at start of multigpu threads to avoid multithreading bugs 
Amp-Thread-ID: https://ampcode.com/threads/T-019d3ee9-19d5-767a-9d7a-e50cbbef815b
Co-authored-by: Amp <amp@ampcode.com>
 2026-03-30 07:07:54 -07:00
Jedrzej Kosinski
be35378986 Merge branch 'master' into worksplit-multigpu 
Amp-Thread-ID: https://ampcode.com/threads/T-019d3ee9-19d5-767a-9d7a-e50cbbef815b
Co-authored-by: Amp <amp@ampcode.com>

# Conflicts:
#	comfy/samplers.py
 2026-03-30 06:24:55 -07:00
Christian Byrne
55e6478526 Rename utils/string nodes with Text prefix and add search aliases (#13227) 
Rename all 11 nodes in the utils/string category to include a "Text"
prefix for better discoverability and natural sorting. Regex nodes get
user-friendly names without "Regex" in the display name.

Renames:
- Concatenate → Text Concatenate
- Substring → Text Substring
- Length → Text Length
- Case Converter → Text Case Converter
- Trim → Text Trim
- Replace → Text Replace
- Contains → Text Contains
- Compare → Text Compare
- Regex Match → Text Match
- Regex Extract → Text Extract Substring
- Regex Replace → Text Replace (Regex)

All renamed nodes include their old display name as a search alias so
users can still find them by searching the original name. Regex nodes
also include "regex" as a search alias.
 2026-03-29 21:02:44 -07:00
comfyanonymous
537c10d231 Update README.md with latest AMD Linux pytorch. (#13228) 2026-03-29 19:07:38 -07:00
rattus
8d723d2caa Fix/tweak pinned memory accounting (#13221) 
* mm: Lower windows pin threshold

Some workflows have more extranous use of shared GPU memory than is
accounted for in the 5% pin headroom. Lower this for safety.

* mm: Remove pin count clearing threshold.

TOTAL_PINNED_MEMORY is shared between the legacy and aimdo pinning
systems, however this catch-all assumes only the legacy system exists.
Remove the catch-all as the PINNED_MEMORY buffer is coherent already.
 2026-03-29 16:43:24 -07:00
Alexander Piskun
d113d1cc32 feat(api-nodes-Tencent3D): allow smaller possible face_count; add uv_image output (#13207) 
Signed-off-by: bigcat88 <bigcat88@icloud.com>
 2026-03-29 14:11:30 -07:00
Jukka Seppänen
a500f1edac CORE-13 feat: Support RT-DETRv4 detection model (#12748) 2026-03-28 23:34:10 -04:00
comfyanonymous
3f77450ef1 Fix #13214 (#13216) 2026-03-28 22:35:59 -04:00
Terry Jia
fc1fdf3389 fix: avoid nested sampler function calls in Color Curves shader (#13209) 2026-03-28 13:13:05 -04:00
rattus
b353a7c863 Integrate RAM cache with model RAM management (#13173) 2026-03-27 21:34:16 -04:00
Terry Jia
3696c5bad6 Add has_intermediate_output flag for nodes with interactive UI (#13048) 2026-03-27 21:06:38 -04:00
comfyanonymous
3a56201da5 Allow flux conditioning without a pooled output. (#13198) 2026-03-27 20:36:26 -04:00
Alexander Piskun
6a2cdb817d fix(api-nodes-nanobana): raise error when not output image is present (#13167) 
Signed-off-by: bigcat88 <bigcat88@icloud.com>
 2026-03-27 12:11:41 -07:00
ComfyUI Wiki
85b7495135 chore: update workflow templates to v0.9.39 (#13196) 2026-03-27 10:13:02 -07:00
Jin Yi
225c52f6a4 fix: register image/svg+xml MIME type for .svg files (#13186) 
The /view endpoint returns text/plain for .svg files on some platforms
because Python's mimetypes module does not always include SVG by default.
Explicitly register image/svg+xml so <img> tags can render SVGs correctly.

Amp-Thread-ID: https://ampcode.com/threads/T-019d2da7-6a64-726a-af91-bd9c44e7f43c
 2026-03-26 22:13:29 -07:00
comfyanonymous
b1fdbeb9a7 Fix blur and sharpen nodes not working with fp16 intermediates. (#13181) 2026-03-26 22:18:16 -04:00
Terry Jia
1dc64f3526 feat: add curve inputs and raise uniform limit for GLSL shader node (#13158) 
* feat: add curve inputs and raise uniform limit for GLSL shader node

* allow arbitrary size for curve
 2026-03-26 21:45:05 -04:00
ComfyUI Wiki
359559c913 chore: update workflow templates to v0.9.38 (#13176) 2026-03-26 12:07:38 -07:00
Alexander Piskun
8165485a17 feat(api-nodes): added new Topaz model (#13175) 
Signed-off-by: bigcat88 <bigcat88@icloud.com>
 2026-03-26 12:02:04 -07:00
Jukka Seppänen
b0fd65e884 fix: regression in text generate with LTXAV model (#13170) 2026-03-26 09:55:05 -07:00
comfyanonymous
2a1f402601 Make Qwen 8B work with TextGenerate node. (#13160) 2026-03-25 23:21:44 -04:00
Luke Mino-Altherr
3eba2dcf2d fix(assets): recognize temp directory in asset category resolution (#13159) 2026-03-25 19:59:59 -07:00
Jukka Seppänen
404d7b9978 feat: Support Qwen3.5 text generation models (#12771) 2026-03-25 22:48:28 -04:00
Dante
6580a6bc01 fix(number-convert): preserve int precision for large numbers (#13147) 2026-03-25 18:06:34 -04:00
Dr.Lt.Data
3b15651bc6 bump manager version to 4.1 (#13156) 2026-03-25 16:49:29 -04:00
Alexander Piskun
a55835f10c fix(api-nodes): made Reve node price badges more precise (#13154) 
Signed-off-by: bigcat88 <bigcat88@icloud.com>
 2026-03-25 11:05:49 -07:00
Krishna Chaitanya
b53b10ea61 Fix Train LoRA crash when training_dtype is "none" with bfloat16 LoRA weights (#13145) 
When training_dtype is set to "none" and the model's native dtype is
float16, GradScaler was unconditionally enabled. However, GradScaler
does not support bfloat16 gradients (only float16/float32), causing a
NotImplementedError when lora_dtype is "bf16" (the default).

Fix by only enabling GradScaler when LoRA parameters are not in
bfloat16, since bfloat16 has the same exponent range as float32 and
does not need gradient scaling to avoid underflow.

Fixes #13124
 2026-03-24 23:53:44 -04:00
Luke Mino-Altherr
7d5534d8e5 feat(assets): register output files as assets after prompt execution (#12812) 2026-03-24 20:48:55 -07:00
Kohaku-Blueleaf
5ebb0c2e0b FP8 bwd training (#13121) 2026-03-24 20:39:04 -04:00