This is the conversion to GGUF of the Wan2.2-Remix T2V - https://civarchive.com/models/2003153
I2V available in https://civarchive.com/models/2097002
Files available also in https://huggingface.co/BigDannyPt/Wan-2-2-Remix
All were using NSFW-Wan-UMT5-XXL (FP8 version) and Wan 2.1 FP32 VAE
Description
FAQ
Comments (14)
HI.which nsfw-wan-umt5-xxxl? 11 gb or 6 gb?
sorry, fp8, will change the description, but I'm not sure if it is the best, I simply went to do the same as the original model says to use
excuse me. i did not get the point. the link in hugging face shows both.i want to know which one i download fp8 or bf16?thank you for your patience.
@faryarjy100 I did response saying fp8, that0s the version I used from that link
@BigDannyPt nice for us 12GB vram users to get the Q5_KM as NSFW as well as its a great balance of size and funcationality
😍 👍🏻
It seems not to work with image to vid, like its not even bothering with the provided image and going straight to the text description. VERY realistic, though, and the prompt adherence is very good.
Ah its because im a dumdumb and didnt catch that this is a text to vid model lmao
Which is better? The original fp8 wan 2.2 remix or this one with fp16? Im using rtx 5090. Can fp16 fit in 32 gb vram?
if you can use the normal model, it will always be better since is the original one and this one is a quantize version of it
HI.which Wan 2.1 FP32 VAE.in hugging face is more than 1.kijal or originalfrom wan ?please give a link.thankyou
they are both the same, you can choose any of them
Any plan for T2V V2?
Wasn't aware of it, will try to do it for both T2V and I2V