Instructions to use alibaba-pai/MiniMax-H3-Acc-LoRAs with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- VideoX Fun
How to use alibaba-pai/MiniMax-H3-Acc-LoRAs with VideoX Fun:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- Notebooks
- Google Colab
- Kaggle
Comfy Version Please
Please upload a comfyui compatible version. Thanks
what is needed to make these comfy versions?
what is being changed? how?
shouldn't there be some official documentation for this?
I thought they can be used directly with comfy . Can it not?
Didn't work with script, sorry.
Worked for me so I pasted the script here.
It does not work, guys. It uses https://arxiv.org/abs/2607.26004 which comfy does not support
Someone needs to code a custom node for it
You can prune these modules to run natively, but then it'll be a different LoRA. No one can tell if it works just the same
It does not work, guys. It uses https://arxiv.org/abs/2607.26004 which comfy does not support
Someone needs to code a custom node for itYou can convert or prune these modules to run natively, but then it'll be a different LoRA. No one can tell if it works just the same
Yes guys, it doesn't work like that, I closed my script thing too. Sorry.
4070 32G:
4-step acceleration lora 5-second video takes about 11 minutes; ACC lora 5-second video takes about 20 minutes
i tried that PR and got smeared hands, legs, bodies in fast motion scenes