Rewrite model card introduction
Browse files
README.md
CHANGED
|
@@ -1,157 +1,177 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: other
|
| 3 |
-
license_name: per-file
|
| 4 |
-
tags:
|
| 5 |
-
-
|
| 6 |
-
-
|
| 7 |
-
-
|
| 8 |
-
-
|
| 9 |
-
-
|
| 10 |
-
-
|
| 11 |
-
-
|
| 12 |
-
|
| 13 |
-
-
|
| 14 |
-
|
| 15 |
-
|
| 16 |
-
|
| 17 |
-
|
| 18 |
-
|
| 19 |
-
|
| 20 |
-
|
| 21 |
-
|
| 22 |
-
|
| 23 |
-
|
| 24 |
-
|
| 25 |
-
|
| 26 |
-
|
| 27 |
-
|
|
| 28 |
-
|
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
|
| 34 |
-
|
|
| 35 |
-
|
|
| 36 |
-
|
| 37 |
-
|
| 38 |
-
|
| 39 |
-
|
| 40 |
-
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
|
| 45 |
-
|
|
| 46 |
-
| `
|
| 47 |
-
| `
|
| 48 |
-
| `
|
| 49 |
-
|
| 50 |
-
|
| 51 |
-
|
| 52 |
-
|
| 53 |
-
|
| 54 |
-
|
|
| 55 |
-
|
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
|
|
| 61 |
-
|
|
| 62 |
-
| `
|
| 63 |
-
| `
|
| 64 |
-
| `
|
| 65 |
-
| `
|
| 66 |
-
| `
|
| 67 |
-
| `
|
| 68 |
-
| `
|
| 69 |
-
| `
|
| 70 |
-
| `
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
|
| 75 |
-
|-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
-
|
| 79 |
-
|
| 80 |
-
|
| 81 |
-
|
| 82 |
-
`
|
| 83 |
-
|
| 84 |
-
|
|
| 85 |
-
|
|
| 86 |
-
| `
|
| 87 |
-
| `
|
| 88 |
-
|
| 89 |
-
|
| 90 |
-
|
| 91 |
-
|
| 92 |
-
|
| 93 |
-
|
| 94 |
-
|
| 95 |
-
|
| 96 |
-
|
|
| 97 |
-
|
| 98 |
-
|
| 99 |
-
|
| 100 |
-
|
| 101 |
-
|
| 102 |
-
|
| 103 |
-
|
| 104 |
-
|
|
| 105 |
-
|
|
| 106 |
-
| `
|
| 107 |
-
| `
|
| 108 |
-
|
| 109 |
-
|
| 110 |
-
|
| 111 |
-
|
| 112 |
-
under
|
| 113 |
-
|
| 114 |
-
|
| 115 |
-
|
| 116 |
-
|
| 117 |
-
|
| 118 |
-
|
| 119 |
-
|
| 120 |
-
|
| 121 |
-
|
| 122 |
-
|
| 123 |
-
|
|
| 124 |
-
|---
|
| 125 |
-
| `models/
|
| 126 |
-
| `models/
|
| 127 |
-
| `models/
|
| 128 |
-
| `models/
|
| 129 |
-
|
| 130 |
-
|
| 131 |
-
|
| 132 |
-
|
| 133 |
-
|
| 134 |
-
|
| 135 |
-
|
| 136 |
-
|
| 137 |
-
|
| 138 |
-
|
| 139 |
-
|
| 140 |
-
|
| 141 |
-
|
| 142 |
-
|
| 143 |
-
|
| 144 |
-
|
| 145 |
-
|
| 146 |
-
|
| 147 |
-
|
| 148 |
-
|
| 149 |
-
|
| 150 |
-
|
| 151 |
-
|
| 152 |
-
|
| 153 |
-
|
| 154 |
-
|
| 155 |
-
|
| 156 |
-
|
| 157 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: other
|
| 3 |
+
license_name: per-file
|
| 4 |
+
tags:
|
| 5 |
+
- comfyui
|
| 6 |
+
- image-editing
|
| 7 |
+
- image-restoration
|
| 8 |
+
- hdr
|
| 9 |
+
- denoising
|
| 10 |
+
- low-light
|
| 11 |
+
- optical-flow
|
| 12 |
+
- video-generation
|
| 13 |
+
- kandinsky
|
| 14 |
+
library_name: diffusion-single-file
|
| 15 |
+
---
|
| 16 |
+
|
| 17 |
+
# WAS Node Suite weights
|
| 18 |
+
|
| 19 |
+
The model files [WAS Node Suite](https://github.com/WASasquatch/was-node-suite-comfyui) runs in
|
| 20 |
+
ComfyUI, kept in one place so a node does not break when a third-party copy is moved or removed.
|
| 21 |
+
Some are upstream checkpoints byte for byte, some are converted to safetensors, and the model
|
| 22 |
+
families under `models/` are ComfyUI conversions and quantisations of upstream releases. Nothing
|
| 23 |
+
here is trained by this project. Every file keeps the licence of the work it came from, listed
|
| 24 |
+
beside it below; the Kandinsky 6 audio VAE and the BluePixel transformers are for
|
| 25 |
+
non-commercial use only.
|
| 26 |
+
|
| 27 |
+
| Folder | What it holds | Used by |
|
| 28 |
+
|---|---|---|
|
| 29 |
+
| `denoise/` | SCUNet and NAFNet denoisers | Power Preprocessor |
|
| 30 |
+
| `low_light/` | HVI-CIDNet, Retinexformer and DarkIR low-light enhancers | Power Preprocessor |
|
| 31 |
+
| `hdr/` | HDRCNN, converted to safetensors | HDR Reconstruct |
|
| 32 |
+
| `birefnet/`, `ben2/` | BiRefNet and BEN2 background removal | Image Remove Background |
|
| 33 |
+
| `optical_flow/` | FlowSeek optical flow, converted to safetensors | Video Motion Model Loader |
|
| 34 |
+
| `models/kandinsky6/` | Kandinsky 6 video with sound: int8 ConvRot and W6A8 transformers, text encoders, audio VAE | Load Diffusion Model and the Kandinsky 6 nodes |
|
| 35 |
+
| `models/bluepixel/` | BluePixel-Turbo 4K image editing: bf16, int8 ConvRot, W6A8 and W4A8 transformers, and its DC-AE autoencoder | BluePixel Model Loader and BluePixel VAE Loader |
|
| 36 |
+
|
| 37 |
+
With `features.network: true` in WAS Node Suite's `config.yaml`, the pack fetches a file from
|
| 38 |
+
every folder outside `models/` the first time a node needs it. The `models/` folders are laid out as
|
| 39 |
+
ComfyUI's own `models` folder: each subfolder goes into the matching folder under
|
| 40 |
+
`ComfyUI/models`, downloaded by hand.
|
| 41 |
+
|
| 42 |
+
## denoise/
|
| 43 |
+
|
| 44 |
+
| File | Original | Licence |
|
| 45 |
+
|---|---|---|
|
| 46 |
+
| `scunet_color_real_psnr.pth` | [cszn/SCUNet](https://github.com/cszn/SCUNet), Kai Zhang | Apache-2.0 |
|
| 47 |
+
| `NAFNet-SIDD-width32.pth` | [megvii-research/NAFNet](https://github.com/megvii-research/NAFNet) | MIT, with Apache-2.0 for BasicSR |
|
| 48 |
+
| `NAFNet-SIDD-width64.pth` | [megvii-research/NAFNet](https://github.com/megvii-research/NAFNet) | MIT, with Apache-2.0 for BasicSR |
|
| 49 |
+
|
| 50 |
+
## hdr/
|
| 51 |
+
|
| 52 |
+
Converted to safetensors for native ComfyUI inference.
|
| 53 |
+
|
| 54 |
+
| File | Original | Licence |
|
| 55 |
+
|---|---|---|
|
| 56 |
+
| `hdrcnn.safetensors` | [gabrieleilertsen/hdrcnn](https://github.com/gabrieleilertsen/hdrcnn), Gabriel Eilertsen | BSD-3-Clause |
|
| 57 |
+
|
| 58 |
+
## low_light/
|
| 59 |
+
|
| 60 |
+
| File | Original | Licence |
|
| 61 |
+
|---|---|---|
|
| 62 |
+
| `hvi-cidnet-generalization.safetensors` | [Fediory/HVI-CIDNet](https://github.com/Fediory/HVI-CIDNet) | MIT |
|
| 63 |
+
| `hvi-cidnet-sice.safetensors` | [Fediory/HVI-CIDNet](https://github.com/Fediory/HVI-CIDNet) | MIT |
|
| 64 |
+
| `hvi-cidnet-sony-total-dark.safetensors` | [Fediory/HVI-CIDNet](https://github.com/Fediory/HVI-CIDNet) | MIT |
|
| 65 |
+
| `hvi-cidnet-fivek.safetensors` | [Fediory/HVI-CIDNet](https://github.com/Fediory/HVI-CIDNet) | MIT |
|
| 66 |
+
| `retinexformer-ntire.pth` | [caiyuanhao1998/Retinexformer](https://github.com/caiyuanhao1998/Retinexformer) | MIT |
|
| 67 |
+
| `retinexformer-lol-v1.pth` | same | MIT |
|
| 68 |
+
| `retinexformer-lol-v2-real.pth` | same | MIT |
|
| 69 |
+
| `retinexformer-lol-v2-synthetic.pth` | same | MIT |
|
| 70 |
+
| `retinexformer-fivek.pth` | same | MIT |
|
| 71 |
+
| `retinexformer-sid.pth` | same | MIT |
|
| 72 |
+
| `retinexformer-smid.pth` | same | MIT |
|
| 73 |
+
| `retinexformer-sdsd-indoor.pth` | same | MIT |
|
| 74 |
+
| `retinexformer-sdsd-outdoor.pth` | same | MIT |
|
| 75 |
+
| `darkir-m.pt` | [cidautai/DarkIR](https://github.com/cidautai/DarkIR) | MIT |
|
| 76 |
+
|
| 77 |
+
## birefnet/
|
| 78 |
+
|
| 79 |
+
| File | Original | Licence |
|
| 80 |
+
|---|---|---|
|
| 81 |
+
| `birefnet/General.safetensors` | [ZhengPeng7/BiRefNet](https://github.com/ZhengPeng7/BiRefNet) | MIT |
|
| 82 |
+
| `birefnet/General-HR.safetensors` | [ZhengPeng7/BiRefNet](https://github.com/ZhengPeng7/BiRefNet) | MIT |
|
| 83 |
+
| `birefnet/General-dynamic.safetensors` | [ZhengPeng7/BiRefNet](https://github.com/ZhengPeng7/BiRefNet) | MIT |
|
| 84 |
+
| `birefnet/General-reso_512.safetensors` | [ZhengPeng7/BiRefNet](https://github.com/ZhengPeng7/BiRefNet) | MIT |
|
| 85 |
+
| `birefnet/DIS.safetensors` | [ZhengPeng7/BiRefNet](https://github.com/ZhengPeng7/BiRefNet) | MIT |
|
| 86 |
+
| `birefnet/DIS-TR_TEs.safetensors` | [ZhengPeng7/BiRefNet](https://github.com/ZhengPeng7/BiRefNet) | MIT |
|
| 87 |
+
| `birefnet/COD.safetensors` | [ZhengPeng7/BiRefNet](https://github.com/ZhengPeng7/BiRefNet) | MIT |
|
| 88 |
+
| `birefnet/HRSOD.safetensors` | [ZhengPeng7/BiRefNet](https://github.com/ZhengPeng7/BiRefNet) | MIT |
|
| 89 |
+
| `birefnet/Matting-HR.safetensors` | [ZhengPeng7/BiRefNet](https://github.com/ZhengPeng7/BiRefNet) | MIT |
|
| 90 |
+
| `birefnet/Portrait.safetensors` | [ZhengPeng7/BiRefNet](https://github.com/ZhengPeng7/BiRefNet) | MIT |
|
| 91 |
+
|
| 92 |
+
## ben2/
|
| 93 |
+
|
| 94 |
+
| File | Original | Licence |
|
| 95 |
+
|---|---|---|
|
| 96 |
+
| `ben2/ben2-base.safetensors` | [PramaLLC/BEN2](https://huggingface.co/PramaLLC/BEN2) | MIT |
|
| 97 |
+
|
| 98 |
+
## optical_flow/
|
| 99 |
+
|
| 100 |
+
FlowSeek's released checkpoints, converted to safetensors with every tensor unchanged. Each
|
| 101 |
+
carries the Depth Anything V2 Small weights FlowSeek reads depth with. `LICENSE-FlowSeek` and
|
| 102 |
+
`NOTICE-FlowSeek.md` beside them are upstream's own.
|
| 103 |
+
|
| 104 |
+
| File | Original | Licence |
|
| 105 |
+
|---|---|---|
|
| 106 |
+
| `optical_flow/flowseek_t_ct.safetensors` | `flowseek_T_CT.pth` from [mattpoggi/flowseek](https://github.com/mattpoggi/flowseek), Matteo Poggi and Fabio Tosi | Apache-2.0 |
|
| 107 |
+
| `optical_flow/flowseek_t_tskh.safetensors` | `flowseek_T_TartanCT_TSKH.pth`, same | Apache-2.0 |
|
| 108 |
+
|
| 109 |
+
## models/kandinsky6/
|
| 110 |
+
|
| 111 |
+
Kandinsky 6 for ComfyUI, in ComfyUI's own folder layout: each subfolder goes into the matching
|
| 112 |
+
folder under `ComfyUI/models`. The transformers are ComfyUI int8 ConvRot and W6A8 quantisations; everything else is byte for byte
|
| 113 |
+
the upstream file. `models/kandinsky6/NOTICE.md` names each source commit and every SHA-256. The
|
| 114 |
+
licence files sit beside the files they cover.
|
| 115 |
+
|
| 116 |
+
| File | Original | Licence |
|
| 117 |
+
|---|---|---|
|
| 118 |
+
| `models/kandinsky6/diffusion_models/kandinsky6_lite_distill_5s_int8_convrot.safetensors` | [kandinskylab/Kandinsky-6.0-Lite-distill-5s-Diffusers](https://huggingface.co/kandinskylab/Kandinsky-6.0-Lite-distill-5s-Diffusers), Kandinsky Lab, quantised for ComfyUI (int8_convrot) | MIT |
|
| 119 |
+
| `models/kandinsky6/diffusion_models/kandinsky6_lite_distill_5s_w6a8.safetensors` | [kandinskylab/Kandinsky-6.0-Lite-distill-5s-Diffusers](https://huggingface.co/kandinskylab/Kandinsky-6.0-Lite-distill-5s-Diffusers), Kandinsky Lab, quantised for ComfyUI (w6a8) | MIT |
|
| 120 |
+
| `models/kandinsky6/diffusion_models/kandinsky6_lite_5s_int8_convrot.safetensors` | [kandinskylab/Kandinsky-6.0-Lite-5s-Diffusers](https://huggingface.co/kandinskylab/Kandinsky-6.0-Lite-5s-Diffusers), Kandinsky Lab, quantised for ComfyUI (int8_convrot) | MIT |
|
| 121 |
+
| `models/kandinsky6/diffusion_models/kandinsky6_lite_5s_w6a8.safetensors` | [kandinskylab/Kandinsky-6.0-Lite-5s-Diffusers](https://huggingface.co/kandinskylab/Kandinsky-6.0-Lite-5s-Diffusers), Kandinsky Lab, quantised for ComfyUI (w6a8) | MIT |
|
| 122 |
+
| `models/kandinsky6/diffusion_models/kandinsky6_pro_distill_5s_int8_convrot.safetensors` | [kandinskylab/Kandinsky-6.0-Pro-distill-5s-Diffusers](https://huggingface.co/kandinskylab/Kandinsky-6.0-Pro-distill-5s-Diffusers), Kandinsky Lab, quantised for ComfyUI (int8_convrot) | MIT |
|
| 123 |
+
| `models/kandinsky6/diffusion_models/kandinsky6_pro_distill_5s_w6a8.safetensors` | [kandinskylab/Kandinsky-6.0-Pro-distill-5s-Diffusers](https://huggingface.co/kandinskylab/Kandinsky-6.0-Pro-distill-5s-Diffusers), Kandinsky Lab, quantised for ComfyUI (w6a8) | MIT |
|
| 124 |
+
| `models/kandinsky6/diffusion_models/kandinsky6_pro_5s_int8_convrot.safetensors` | [kandinskylab/Kandinsky-6.0-Pro-5s-Diffusers](https://huggingface.co/kandinskylab/Kandinsky-6.0-Pro-5s-Diffusers), Kandinsky Lab, quantised for ComfyUI (int8_convrot) | MIT |
|
| 125 |
+
| `models/kandinsky6/diffusion_models/kandinsky6_pro_5s_w6a8.safetensors` | [kandinskylab/Kandinsky-6.0-Pro-5s-Diffusers](https://huggingface.co/kandinskylab/Kandinsky-6.0-Pro-5s-Diffusers), Kandinsky Lab, quantised for ComfyUI (w6a8) | MIT |
|
| 126 |
+
| `models/kandinsky6/text_encoders/qwen_2.5_vl_7b_fp8_scaled.safetensors` | [Qwen/Qwen2.5-VL-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-7B-Instruct), as repackaged by [Comfy-Org](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI) | Apache-2.0 |
|
| 127 |
+
| `models/kandinsky6/text_encoders/clip_l.safetensors` | [openai/clip-vit-large-patch14](https://huggingface.co/openai/clip-vit-large-patch14), as repackaged by [Comfy-Org](https://huggingface.co/Comfy-Org/HunyuanVideo_repackaged) | MIT |
|
| 128 |
+
| `models/kandinsky6/vae/kandinsky6_audio_vae.safetensors` | [MMAudio](https://huggingface.co/hkchengrex/MMAudio)'s 44.1 kHz autoencoder and [BigVGAN v2](https://huggingface.co/nvidia/bigvgan_v2_44khz_128band_512x), as packed by Kandinsky Lab | **CC BY-NC 4.0, non-commercial** (MMAudio); MIT (BigVGAN) |
|
| 129 |
+
|
| 130 |
+
The video VAE Kandinsky 6 decodes with, `hunyuan_video_vae_bf16.safetensors`, is at
|
| 131 |
+
[Comfy-Org/HunyuanVideo_repackaged](https://huggingface.co/Comfy-Org/HunyuanVideo_repackaged)
|
| 132 |
+
under the Tencent Hunyuan Community License.
|
| 133 |
+
|
| 134 |
+
## models/bluepixel/
|
| 135 |
+
|
| 136 |
+
[BluePixel-Turbo](https://huggingface.co/openvivo/BluePixel-Turbo) for ComfyUI, in ComfyUI's own
|
| 137 |
+
folder layout: each subfolder goes into the matching folder under `ComfyUI/models`. The
|
| 138 |
+
transformers are vivo's release converted to the FLUX.2 layout ComfyUI reads, at bfloat16 and as
|
| 139 |
+
ComfyUI int8 ConvRot, W6A8 and W4A8 quantisations; the autoencoder is byte for byte the upstream
|
| 140 |
+
file. `models/bluepixel/NOTICE.md` carries the FLUX Non-Commercial License attribution notice,
|
| 141 |
+
names each source commit and every SHA-256. The licence files sit beside the files they cover.
|
| 142 |
+
|
| 143 |
+
| File | Original | Licence |
|
| 144 |
+
|---|---|---|
|
| 145 |
+
| `models/bluepixel/diffusion_models/bluepixel_turbo_bf16.safetensors` | [openvivo/BluePixel-Turbo](https://huggingface.co/openvivo/BluePixel-Turbo), vivo, a modified FLUX.2 [klein] 9B, converted for ComfyUI | **FLUX Non-Commercial License, non-commercial** |
|
| 146 |
+
| `models/bluepixel/diffusion_models/bluepixel_turbo_int8_convrot.safetensors` | the same, quantised for ComfyUI (int8_convrot) | **FLUX Non-Commercial License, non-commercial** |
|
| 147 |
+
| `models/bluepixel/diffusion_models/bluepixel_turbo_w6a8.safetensors` | the same, quantised for ComfyUI (w6a8) | **FLUX Non-Commercial License, non-commercial** |
|
| 148 |
+
| `models/bluepixel/diffusion_models/bluepixel_turbo_w4a8.safetensors` | the same, quantised for ComfyUI (w4a8) | **FLUX Non-Commercial License, non-commercial** |
|
| 149 |
+
| `models/bluepixel/vae/dc-ae-f64c128-mix-1.0.safetensors` | [mit-han-lab/dc-ae-f64c128-mix-1.0-diffusers](https://huggingface.co/mit-han-lab/dc-ae-f64c128-mix-1.0-diffusers), MIT Han Lab | Apache-2.0 |
|
| 150 |
+
|
| 151 |
+
The text encoder, `qwen_3_8b.safetensors`, is at
|
| 152 |
+
[Comfy-Org/Flux-2-Klein-9B](https://huggingface.co/Comfy-Org/Flux-2-Klein-9B) under
|
| 153 |
+
`split_files/text_encoders/`, with the same weights as vivo's.
|
| 154 |
+
|
| 155 |
+
## Citations
|
| 156 |
+
|
| 157 |
+
Zhang et al., Practical Blind Denoising via Swin-Conv-UNet and Data Synthesis, 2022.
|
| 158 |
+
Chen et al., Simple Baselines for Image Restoration, ECCV 2022.
|
| 159 |
+
Yan et al., HVI: A New Color Space for Low-light Image Enhancement, CVPR 2025.
|
| 160 |
+
Cai et al., Retinexformer: One-stage Retinex-based Transformer for Low-light
|
| 161 |
+
Image Enhancement, ICCV 2023.
|
| 162 |
+
Feijoo et al., DarkIR: Robust Low-Light Image Restoration, 2024.
|
| 163 |
+
Zheng et al., Bilateral Reference for High-Resolution Dichotomous Image
|
| 164 |
+
Segmentation, 2024.
|
| 165 |
+
Poggi and Tosi, FlowSeek: Optical Flow Made Easier with Depth Foundation Models
|
| 166 |
+
and Motion Bases, ICCV 2025.
|
| 167 |
+
Yang et al., Depth Anything V2, NeurIPS 2024.
|
| 168 |
+
Kandinsky Lab, Kandinsky 6, 2026.
|
| 169 |
+
Cheng et al., MMAudio: Taming Multimodal Joint Training for High-Quality
|
| 170 |
+
Video-to-Audio Synthesis, CVPR 2025.
|
| 171 |
+
Lee et al., BigVGAN: A Universal Neural Vocoder with Large-Scale Training, ICLR 2023.
|
| 172 |
+
Bai et al., Qwen2.5-VL Technical Report, 2025.
|
| 173 |
+
Radford et al., Learning Transferable Visual Models From Natural Language
|
| 174 |
+
Supervision, ICML 2021.
|
| 175 |
+
vivo, BluePixel: Toward Native 4K Agentic Image Editing, 2026.
|
| 176 |
+
Chen et al., Deep Compression Autoencoder for Efficient High-Resolution Diffusion
|
| 177 |
+
Models, ICLR 2025.
|