Bbox face detection
Need help with hands
Got Triton and Sage Attention installed, apparently successfully, but they don't affect speed one bit.
HunyuanVideo-I2V updated their model just now
Update comfyui today killed front end and removed requirements.txt so cant reinstall
ComfyUI Manager 3.1.1.
Cant install from git URL after updating comfyui
Cant change security level
implement Sage Attention in comfyui, with hunyuan video?
Which budget GPU should I buy for FLUX?
4060 Ti (16GB) vs 4070 Ti (12GB) for image generation and LoRA training
Queue prompt vs batch count?
Nvidia Sana ComfyUI?
SVDQuant, claiming a 3x speedup with Flux over NF4
[Workflow] 💥TRITON is no longer a requirement for Hunyan! + Bonus: 3x upscale workflow
Error with videohelpersuite, output is black
video2video workflow with sdxl?
Error: warning, shapes in InjectNoise not the same, ignoring
warning, shapes in InjectNoise not the same, ignoring
Triton 3 wheels published for Windows and working - Now we can have huge speed up at some repos and libraries
Comfyui taking more vram than needed, help
Best graphic card with 16gb under 450$?
I made a thing so to generate nicer images with the CFG scale at 1 while keeping the speed (that makes ~320% with tensor rt)!!!
Replete-AI/Replete-Coder-Llama3-8B The big boi. 1 billion instruct tokens trained, an fully uncensored.
ByT5 SDXL finetune "Glyph-SDXL-v2" released, does perfect long text in English, Chinese, Japanese, Korean, French, German, Spanish, Italian, Portuguese and Russian