FREAKY 6 -896x1152:SDXL+SD3.5 Medium+PixArt Sigma+Aura Flow+Hunyuan+Mochi Video; 5 batches of 2 png+
5.0
0 reviewsDescription
FREAKY 6 -896x1152: SDXL(no refiner)+SD3.5 Medium+PixArt Sigma+Aura Flow+Hunyuan+Mochi Video
Each individual workflow is in its own group so you can right click the groups to bypass them or set them to always fo flip them on or off. Mochi addition adds a lot of time. Processing with the video and all of the others set up takes around an hour. Without the video it's only a few minutes.
This is sort of a monster workflow for allowing me to do whatever I want to do other than flux. For flux unfortunately I have to rely on Forge as it can do flux better on 8gb vram (gives errors on tensor divide between gpu and cpu otherwise in comfy).
Have fun!
Discussion
(No comments yet)
Loading...
Reviews
No reviews yet
Versions (1)
- latest (a year ago)
Node Details
Primitive Nodes (25)
EmptyMochiLatentVideo (2)
EmptySD3LatentImage (1)
FL_SDUltimate_Slices (1)
ModelSamplingAuraFlow (1)
ModelSamplingSD3 (1)
Note (7)
PixArtT5FromSD3CLIP (1)
PrimitiveNode (8)
Reroute (1)
TripleCLIPLoader (1)
UltimateSDUpscaleCustomSample (1)
Custom Nodes (57)
ComfyUI
- CLIPLoader (1)
- VAELoader (1)
- EmptyLatentImage (4)
- VAEDecodeTiled (1)
- CLIPTextEncode (12)
- KSampler (5)
- CheckpointLoaderSimple (5)
- ConditioningCombine (1)
- ConditioningSetTimestepRange (2)
- ConditioningZeroOut (1)
- SD_4XUpscale_Conditioning (1)
- KSamplerSelect (1)
- VAEDecode (3)
- AlignYourStepsScheduler (1)
- UpscaleModelLoader (1)
- SaveImage (6)
- SaveAnimatedWEBP (1)
- KSampler //Inspire (1)
- DPRandomGenerator (2)
- Efficient Loader (1)
- PixArtCheckpointLoader (1)
- SDXLPromptStyler (1)
- VAEDecodeTiled_TiledDiffusion (3)
- Image Lucy Sharpen (1)
Model Details
Checkpoints (6)
PixArt\PixArt-Sigma-XL-2-1024-MS-fp16.safetensors
auraflow_v03.safetensors
hunyuan_dit_comfyui\hunyuan_dit_1.2.safetensors
mochi_preview_fp8_scaled.safetensors
sd3.5_medium.safetensors
LoRAs (0)