A model that interprets diverse imaginings.
It incorporates various LoRAs.
*CLIP and VAE are not included.
【Recommended】
Step: 4–5
CFG: High 1.5 / Low 1.0
Sampler / Scheduler:
uni_pc_bh2 / normal
uni_pc / normal
Description
FAQ
Comments (16)
Minimax H3 is already open
Yes, I know. I’m very interested in H3 and plan to test it locally once the ComfyUI workflows for 24GB GPUs become more stable.
@Aki7777777 It is stable for even a 16GB GPU or less.
@tsgeza870 Thanks! I'll give it a try as soon as I can.
@tsgeza870 stable in 8gb laptop gpu
Any choice in the list downloading the one wan22I2VRapidDiverse_highLow checkpoint. This is right?
Yes. There are “High” and “Low” versions, so please download each one.
@Aki7777777 Both downloads look identical to me
@watsont474
There are the following two:
Wan2.2_I2V_Rapid-Diverse0.5-High.safetensors
Wan2.2_I2V_Rapid-Diverse2.0-Low.safetensors
@Aki7777777 Nope
Their was identical
https://dropmefiles.com/vIEJX
@jafdeth2030105 I'll sorting the tabs into “High” and “LOW,” so please wait a moment.
It looks like both High and Low have been uploaded with the exact same file name.
wan22I2VRapidDiverse_highLow.safetensors
I'll sorting the tabs into “High” and “LOW,” so please wait a moment.
both the fp8 download files are both the same file, labelled "wan22I2VRapidDiverse_highLow.safetensors" - is this right? is the model available on hugging face yet?
I'll sorting the tabs into “High” and “LOW,” so please wait a moment.
I haven't uploaded it to Hugging Face, but it is available in ComfyUI.
@Aki7777777 thanks bud, much appreciated :)