Dima/ComfyUI
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Commit Graph

  • 31e961736a Fix issue with batches and newbie. (#11435) comfyanonymous 2025-12-19 21:23:51 -08:00
  • 767ee30f21 ZImageFunControlNet: Fix mask concatenation in --gpu-only (#11421) rattus 2025-12-19 21:22:17 -08:00
  • 3ab9748903 Disable prompt weights on newbie te. (#11434) comfyanonymous 2025-12-19 21:19:47 -08:00
  • 0aa7fa464e Implement sliding attention in Gemma3 (#11409) woctordho 2025-12-20 13:16:46 +08:00
  • 514c24d756 Fix error from logging line (#11423) drozbay 2025-12-19 21:22:45 -07:00
  • 809ce68749 Support nested tensor denoise masks. (#11431) comfyanonymous 2025-12-19 16:59:25 -08:00
  • cc4ddba1b6 Allow enabling use of MIOpen by setting COMFYUI_ENABLE_MIOPEN=1 as an env var (#11366) BradPepersAMD 2025-12-19 15:01:50 -07:00
  • 8376ff6831 bump comfyui_manager version to the 4.0.3b7 (#11422) Dr.Lt.Data 2025-12-20 03:41:56 +09:00
  • 5b4d0664c8 add Flux2MaxImage API Node (#11420) Alexander Piskun 2025-12-19 20:02:49 +02:00
  • 894802b0f9 Add LatentCutToBatch node. (#11411) comfyanonymous 2025-12-18 19:21:40 -08:00
  • 28eaab608b Diffusion model part of Qwen Image Layered. (#11408) comfyanonymous 2025-12-18 17:21:14 -08:00
  • 6a2678ac65 Trim/pad channels in VAE code. (#11406) comfyanonymous 2025-12-18 15:22:38 -08:00
  • e4fb3a3572 Support loading Wan/Qwen VAEs with different in/out channels. (#11405) comfyanonymous 2025-12-18 14:45:33 -08:00
  • e8ebbe668e chore: update workflow templates to v0.7.60 (#11403) ComfyUI Wiki 2025-12-19 06:09:29 +08:00
  • 1ca89b810e Add unified jobs API with /api/jobs endpoints (#11054) ric-yu 2025-12-17 21:44:31 -08:00
  • bf7dc63bd6 skip_load_model -> force_full_load (#11390) comfyanonymous 2025-12-17 20:29:32 -08:00
  • 86dbb89fc9 Resolution bucketing and Trainer implementation refactoring (#11117) Kohaku-Blueleaf 2025-12-18 11:15:27 +08:00
  • ba6080bbab ComfyUI v0.5.1 v0.5.1 comfyanonymous 2025-12-17 21:04:50 -05:00
  • 16d85ea133 Better handle torch being imported by prestartup nodes. (#11383) comfyanonymous 2025-12-17 16:43:18 -08:00
  • 5d9ad0c6bf Fix the last step with non-zero sigma in sa_solver (#11380) chaObserv 2025-12-18 02:57:40 +08:00
  • c08f97f344 fix regression in V3 nodes processing (#11375) Alexander Piskun 2025-12-17 20:24:25 +02:00
  • 887143854b feat(api-nodes): add GPT-Image-1.5 (#11368) Alexander Piskun 2025-12-17 19:43:41 +02:00
  • 3a5f239cb6 ComfyUI v0.5.0 v0.5.0 comfyanonymous 2025-12-17 03:46:11 -05:00
  • 827bb1512b Add exp_heun_2_x0 sampler series (#11360) chaObserv 2025-12-17 12:35:43 +08:00
  • ffdd53b327 Check state dict key to auto enable the index_timestep_zero ref method. (#11362) comfyanonymous 2025-12-16 14:03:17 -08:00
  • 65e2103b09 feat(api-nodes): add Wan2.6 model to video nodes (#11357) Alexander Piskun 2025-12-16 23:51:48 +02:00
  • 9304e47351 Update workflows for new release process (#11064) Benjamin Lu 2025-12-15 23:24:18 -08:00
  • bc606d7d64 Add a way to set the default ref method in the qwen image code. (#11349) comfyanonymous 2025-12-15 22:26:55 -08:00
  • 645ee1881e Inpainting for z image fun control. Use the ZImageFunControlnet node. (#11346) comfyanonymous 2025-12-15 20:38:12 -08:00
  • 3d082c3206 bump comfyui-frontend-package to 1.34.9 (patch) (#11342) Christian Byrne 2025-12-15 20:35:37 -08:00
  • 683569de55 Only enable fp16 on ZImage on newer pytorch. (#11344) comfyanonymous 2025-12-15 19:33:27 -08:00
  • ea2c117bc3 [BlockInfo] Wan (#10845) Haoming 2025-12-16 09:59:16 +08:00
  • fc4af86068 [BlockInfo] Lumina (#11227) Haoming 2025-12-16 09:57:28 +08:00
  • 41bcf0619d Add code to detect if a z image fun controlnet is broken or not. (#11341) comfyanonymous 2025-12-15 17:51:06 -08:00
  • d02d0e5744 [add] tripo3.0 (#10663) seed93 2025-12-16 09:38:46 +08:00
  • 70541d4e77 Support the new qwen edit 2511 reference method. (#11340) comfyanonymous 2025-12-15 16:20:34 -08:00
  • 77b2f7c228 Add context windows callback for custom cond handling (#11208) drozbay 2025-12-15 17:06:32 -07:00
  • 43e0d4e3cc comfy_api: remove usage of "Type","List" and "Dict" types (#11238) Alexander Piskun 2025-12-16 02:01:10 +02:00
  • dbd330454a feat(preview): add per-queue live preview method override (#11261) Dr.Lt.Data 2025-12-16 08:57:39 +09:00
  • 33c7f1179d drop Pika API nodes (#11306) Alexander Piskun 2025-12-16 01:32:29 +02:00
  • af91eb6c99 api-nodes: drop Kling v1 model (#11307) Alexander Piskun 2025-12-16 01:30:24 +02:00
  • 5cb1e0c9a0 Disable guards on transformer_options when torch.compile (#11317) comfyanonymous 2025-12-15 13:49:29 -08:00
  • 51347f9fb8 chore: update workflow templates to v0.7.59 (#11337) ComfyUI Wiki 2025-12-16 05:28:55 +08:00
  • a5e85017d8 bump manager requirments to the 4.0.3b5 (#11324) Dr.Lt.Data 2025-12-16 04:24:01 +09:00
  • 5ac3b26a7d Update warning for old pytorch version. (#11319) comfyanonymous 2025-12-14 01:02:50 -08:00
  • 6592bffc60 seeds_2: add phi_2 variant and sampler node (#11309) chaObserv 2025-12-14 13:03:29 +08:00
  • 971cefe7d4 Fix pytorch warnings. (#11314) comfyanonymous 2025-12-13 15:45:23 -08:00
  • da2bfb5b0a Basic implementation of z image fun control union 2.0 (#11304) comfyanonymous 2025-12-12 22:39:11 -08:00
  • c5a47a1692 Fix bias dtype issue in mixed ops. (#11293) comfyanonymous 2025-12-12 08:49:35 -08:00
  • 908fd7d749 feat(api-nodes): new TextToVideoWithAudio and ImageToVideoWithAudio nodes (#11267) Alexander Piskun 2025-12-12 10:18:31 +02:00
  • 5495589db3 Respect the dtype the op was initialized in for non quant mixed op. (#11282) comfyanonymous 2025-12-11 20:32:27 -08:00
  • 982876d59a WanMove support (#11247) Jukka Seppänen 2025-12-12 05:29:34 +02:00
  • 338d9ae3bb Make portable updater work with repos in unmerged state. (#11281) comfyanonymous 2025-12-11 15:56:33 -08:00
  • eeb020b9b7 Better chroma radiance and other models vram estimation. (#11278) comfyanonymous 2025-12-11 14:33:09 -08:00
  • ae65433a60 This only works on radiance. (#11277) comfyanonymous 2025-12-11 14:15:00 -08:00
  • fdebe18296 Fix regular chroma radiance (#11276) comfyanonymous 2025-12-11 14:09:35 -08:00
  • f8321eb57b Adjust memory usage factor. (#11257) comfyanonymous 2025-12-10 22:30:31 -08:00
  • 93948e3fc5 feat(api-nodes): enable Kling Omni O1 node (#11229) Alexander Piskun 2025-12-11 08:11:12 +02:00
  • e711aaf1a7 Lower VAE loading requirements：Create a new branch for GPU memory calculations in qwen-image vae (#11199) Farshore 2025-12-11 11:02:26 +08:00
  • 57ddb7fd13 Fix: filter hidden files from /internal/files endpoint (#11191) Johnpaul Chiwetelu 2025-12-11 03:49:49 +01:00
  • 17c92a9f28 Tweak Z Image memory estimation. (#11254) comfyanonymous 2025-12-10 16:59:48 -08:00
  • 36357bbcc3 process the NodeV1 dict results correctly (#11237) Alexander Piskun 2025-12-10 21:55:09 +02:00
  • f668c2e3c9 bump comfyui-frontend-package to 1.34.8 (#11220) Benjamin Lu 2025-12-09 19:27:07 -08:00
  • fc657f471a ComfyUI version v0.4.0 v0.4.0 comfyanonymous 2025-12-09 18:22:09 -05:00
  • 791e30ff50 Fix nan issue when quantizing fp16 tensor. (#11213) comfyanonymous 2025-12-09 14:03:21 -08:00
  • e2a800e7ef Fix for HunyuanVideo1.5 meanflow distil (#11212) Jukka Seppänen 2025-12-09 23:59:16 +02:00
  • 9d252f3b70 ops: delete dead code (#11204) rattus 2025-12-09 15:55:13 +10:00
  • b9fb542703 add chroma-radiance-x0 mode (#11197) Lodestone 2025-12-09 11:33:29 +07:00
  • cabc4d351f bump comfyui-frontend-package to 1.33.13 (patch) (#11200) Christian Byrne 2025-12-08 20:22:02 -08:00
  • e136b6dbb0 dequantization offload accounting (fixes Flux2 OOMs - incl TEs) (#11171) rattus 2025-12-09 14:21:31 +10:00
  • d50f342c90 Fix potential issue. (#11201) comfyanonymous 2025-12-08 20:20:04 -08:00
  • 3b0368aa34 Fix regression. (#11194) comfyanonymous 2025-12-08 14:38:36 -08:00
  • 935493f6c1 chore: update workflow templates to v0.7.54 (#11192) ComfyUI Wiki 2025-12-09 04:18:53 +08:00
  • 60ee574748 retune lowVramPatch VRAM accounting (#11173) rattus 2025-12-09 06:18:06 +10:00
  • 8e889c535d Support "transformer." LoRA prefix for Z-Image (#11135) dxqb 2025-12-08 21:17:26 +01:00
  • fd271dedfd [API Nodes] add support for seedance-1-0-pro-fast model (#10947) Alexander Piskun 2025-12-08 11:33:46 +02:00
  • c3c6313fc7 Added "system_prompt" input to Gemini nodes (#11177) Alexander Piskun 2025-12-08 11:28:17 +02:00
  • 85c4b4ae26 chore: replace imports of deprecated V1 classes (#11127) Alexander Piskun 2025-12-08 11:27:02 +02:00
  • 058f084371 Update workflow templates to v0.7.51 (#11150) ComfyUI Wiki 2025-12-08 17:22:51 +08:00
  • ec7f65187d chore(comfy_api): replace absolute imports with relative (#11145) Alexander Piskun 2025-12-08 11:21:41 +02:00
  • 56fa7dbe38 Properly load the newbie diffusion model. (#11172) comfyanonymous 2025-12-07 04:44:55 -08:00
  • 329480da5a Fix qwen scaled fp8 not working with kandinsky. Make basic t2i wf work. (#11162) comfyanonymous 2025-12-06 17:50:10 -08:00
  • 4086acf3c2 Fix on-load VRAM OOM (#11144) rattus 2025-12-07 09:42:09 +10:00
  • 50ca97e776 Speed up lora compute and lower memory usage by doing it in fp16. (#11161) comfyanonymous 2025-12-06 15:36:20 -08:00
  • 7ac7d69d94 Fix EmptyAudio node input types (#11149) Jukka Seppänen 2025-12-06 20:09:44 +02:00
  • 76f18e955d marked all Pika API nodes a deprecated (#11146) Alexander Piskun 2025-12-06 13:28:08 +02:00
  • d7a0aef650 Set OCL_SET_SVM_SIZE on AMD. (#11139) comfyanonymous 2025-12-05 21:15:21 -08:00
  • 913f86b727 [V3] convert nodes_mask.py to V3 schema (#10669) Alexander Piskun 2025-12-06 06:24:10 +02:00
  • 117bf3f2bd convert nodes_freelunch.py to the V3 schema (#10904) Alexander Piskun 2025-12-06 06:22:02 +02:00
  • ae676ed105 Fix regression. (#11137) comfyanonymous 2025-12-05 20:01:19 -08:00
  • fd109325db Kandinsky5 model support (#10988) Jukka Seppänen 2025-12-06 05:20:22 +02:00
  • bed12674a1 docs: add ComfyUI-Manager documentation and update to v4.0.3b4 (#11133) Dr.Lt.Data 2025-12-06 08:45:38 +09:00
  • 092ee8a500 Fix some custom nodes. (#11134) comfyanonymous 2025-12-05 15:25:31 -08:00
  • 79d17ba233 Context windows fixes and features (#10975) Jukka Seppänen 2025-12-05 22:42:46 +02:00
  • 6fd463aec9 Fix regression when text encoder loaded directly on GPU. (#11129) comfyanonymous 2025-12-05 12:33:16 -08:00
  • 43071e3de3 Make old scaled fp8 format use the new mixed quant ops system. (#11000) comfyanonymous 2025-12-05 11:35:42 -08:00
  • 0ec05b1481 Remove line made unnecessary (and wrong) after transformer_options was added to NextDiT's _forward definition (#11118) Jedrzej Kosinski 2025-12-05 11:05:38 -08:00
  • 35fa091340 Forgot to put this in README. (#11112) comfyanonymous 2025-12-04 19:52:09 -08:00
  • 3c8456223c [API Nodes]: fixes and refactor (#11104) Alexander Piskun 2025-12-05 00:05:28 +02:00
  • 9bc893c5bb sd: bump HY1.5 VAE estimate (#11107) rattus 2025-12-05 03:50:36 +10:00