city96/ComfyUI-GGUF

Issues

• SD3.5 large Turbo Gguf give black image in comfyUI why ?

  #156 opened 2 months ago by Creative-comfyUI
  5

• Is it possible to support Ruyi with gguf?

  #191 opened 10 days ago by doogyhatts
  2

• safetensors do not appear in Load Diffusion Model node while the models already in unet folder

  #179 opened 20 days ago by ZHAORITIANisavailable
  17

• Dual clip loader with Hunyuan type ?

  #190 opened 9 days ago by darkon12
  1

• How to convert FLUX.1-Depth-dev.safetensors to Q8?

  #189 opened 11 days ago by ymzlygw
  0

• Clip Loaders dont work for sd3+ models anymore.

  #186 opened 13 days ago by kseyhan
  2

• OOM error with Hunyuan Video GGUF

  #187 opened 14 days ago by JorgeR81
  4

• With newest update, max() arg is an empty sequence when loading dual CLIPs for Flux

  #188 opened 13 days ago by EnragedAntelope
  2

• t5 does not appear in the dual clip loader (gguf) list

  #178 opened 20 days ago by dabrodetel
  6

• Updates to LTX-video model

  #185 opened 15 days ago by doogyhatts
  1

• DualCLIPLoader (GGUF) needs hunyuan type

  #183 opened 16 days ago by ABDe3N
  1

• Unexpected architecture type in GGUF file, expected one of flux, sd1, sdxl, t5encoder but got 'hyvid'

  #184 opened 16 days ago by ddddddeyrueuad
  1

• flux Q*_K models don't work on macbooks (green noise)

  #177 opened 21 days ago by vmirnv
  4

• gguf no support flow edit

  #182 opened 17 days ago by xueqing0622
  0

• hi boss,where can i find the "Force/Set CLIP Device"

  #181 opened 18 days ago by ptmaster
  1

• please convert fast hunyuan 6 step model to gguf version

  #180 opened 18 days ago by rafiislambd
  1

• Sdxl Q8 gguf

  #146 opened 2 months ago by ReyJ94
  3

• incompatibility on python12

  #158 opened 22 days ago by michel-io
  8

• not work for brushnet, "The size of tensor a (640) must match the size of tensor b (320) at non-singleton dimension 1"

  #176 opened 22 days ago by xueqing0622
  1

• LTXVideo "CLIPLoader (GGUF)" problem

  #163 opened a month ago by Ratinod
  9

• macOS Sequoia 15.1.1 with MPS and PyTorch (BFloat16 Unsupported)

  #175 opened a month ago by SkyForceCoder
  5

• Q5_K_M easy to got OOM in 8GB Vram

  #167 opened a month ago by Windecay
  0

• No performance difference between original BF16 and Q4_0 quantized GGUF models

  #152 opened 2 months ago by Tureti
  3

• CreateHookModelAsLora for gguf

  #174 opened a month ago by Amazon90
  0

• UnetLoaderGGUF cannot reshape array of size 615776 into shape (12288,1728) "i spent 2 days and wdidn't resolved please any solution

  #169 opened a month ago by habibteyeb
  2

• Flux_dev_scaled_fp8 can´t be quantisized

  #173 opened a month ago by cbresciano
  1

• When loading Q5_K_M T5 with 11GB Fp8 Flux model, I get OOM sometimes

  #172 opened a month ago by Apacchi88
  0

• Just updated Comfy and now GGUF models are no longer recognized in the GGUF Dual or Triple Clip loader

  #171 opened a month ago by TripleHeadedMonkey
  2

• Incompatibility of "ClipLoader (GGUF)" and "triton-windows" when working with CogVideoX/Mochi (colored cubes)

  #150 opened a month ago by Ratinod
  3

• Would you consider making standalone gradio version?

  #170 opened a month ago by FurkanGozukara
  0

• sd3.5m modle Cannot work

  #168 opened a month ago by kikilloy
  0

• tools how to convert safetensors to fp8

  #165 opened a month ago by xueqing0622
  2

• Git version 2.47.0 deem patch files as corrupt

  #166 opened a month ago by silveroxides
  1

• “flux1-fill-dev” model quantization request

  #161 opened a month ago by Amazon90
  27

• Installation Fails

  #164 opened a month ago by choochus
  2

• please add ltxv support to clip loader module

  #162 opened a month ago by cchance27
  1

• GGUF loader fails with quantized version of SD1.5 model

  #160 opened a month ago by Metal-dude
  3

• ComfyUI-GGUF module for custom nodes: type object 'manual_cast' has no attribute 'Embedding'

  #155 opened 2 months ago by Coldknife2
  1

• `newbyteorder` was removed from the ndarray class in NumPy 2.0

  #153 opened 2 months ago by fabiomb
  1

• GGUF not working with aidock comfyui template

  #137 opened 2 months ago by cubo94857
  2

• Generates even slower than higher precision model

  #148 opened 2 months ago by XJF2332
  3

• Unable to load GGUF t5xxl text_encoders

  #151 opened 2 months ago by Polukeev
  3

• Are there GGUF versions of Controlnet models?

  #149 opened 2 months ago by moreAImore
  0

• Incompatibility with stable-diffusion.cpp for SD3.X models

  #147 opened 2 months ago by stduhpf
  1

• Share your GGUF conversions of Flux finetunes !

  #141 opened 2 months ago by JorgeR81
  9

• No VAE output option in UnetLoaderGGUF

  #139 opened 2 months ago by Amazon90
  3

• `newbyteorder` was removed from the ndarray class in NumPy 2.0. Use `arr.view(arr.dtype.newbyteorder(order))` instead.

  #143 opened 2 months ago by lelzhang
  2

• `t5xxl_fp8_e4m3fn_scaled.safetensors` and `t5-v1_1-xxl-encoder-Q8_0.gguf`

  #138 opened 2 months ago by Amazon90
  3

• Add GGUF versions for SD3.5

  #136 opened 2 months ago by JorgeR81
  35

• 'SD3ClipModel' object has no attribute 't5_attention_mask'

  #140 opened 2 months ago by doogyhatts
  0