I have been testing on hunyuan video for a long and it has improved.
Unfortunately, the RTX 4060 only has 16GB of RAM, but don't worry! It will definitely continue to run on 12GB of RAM. I hope!
first models:
you need:
VAE Video bf16:https://huggingface.co/Comfy-Org/HunyuanVideo_repackaged/tree/main/split_files/vae
clip:https://huggingface.co/Comfy-Org/HunyuanVideo_repackaged/tree/main/split_files/text_encoders
gguf video:https://huggingface.co/city96/HunyuanVideo-gguf/tree/main
Lora FastVideo:https://huggingface.co/Kijai/HunyuanVideo_comfy/blob/main/hyvideo_FastVideo_LoRA-fp8.safetensors
extra:
clip sae, if you want:https://huggingface.co/zer0int/CLIP-SAE-ViT-L-14/blob/main/ViT-L-14-GmP-SAE-FULL-model.safetensors & https://huggingface.co/zer0int/CLIP-SAE-ViT-L-14/blob/main/ViT-L-14-GmP-SAE-TE-only.safetensors
Custom nodes:
https://github.com/city96/ComfyUI-GGUF
https://github.com/Kosinkadink/ComfyUI-VideoHelperSuite
that's it, have fun!
Description
FAQ
Comments (10)
thanks for sharing. isn't fast lora video supposed to be used with less steps? i see you have 23 in the workflow
merry christmas to all!
can you share LTXVApplySTG and STGGuider node?
thanks for sharing. crf factor in video combine is set to 22, you may want to lower that for higher quality
even use blurry background in nagative prompt,the background still very blurry,how to make background clear
flux guidance at 10 gives me messed up output. Also YUV420p10le does not play on my Win11 PC.
thats weird the workflow working but not generate output video just finish and no errors. (v1.3) RTX 3060 12GB and hunyuan-video-t2v-720p-q5_k_m
Do you have another source for the improved models for 12gb vram? The link you shared 404's
You crossed out https://huggingface.co/city96/HunyuanVideo-gguf/tree/main . What checkpoint model should we be using?
negative prompt don't work for me. On a same seed, this is the same result with or without negative
Looks like we don't have an active mirror for this file right now.
CivArchive is a community-maintained index — we catalog mirrors that volunteers upload to HuggingFace, torrents, and other public hosts. Looks like no one has uploaded a copy of this file yet.
Some files do get recovered over time through contributions. If you're looking for this one, feel free to ask in Discord, or help preserve it if you have a copy.