Instructions to use DeepBeepMeep/MiniMax-H3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusion Single File
How to use DeepBeepMeep/MiniMax-H3 with Diffusion Single File:
# 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
FL2VA-pruned_bf16 broken?
these are models to be used in wangp, sir
Not necessarily. I also use this MiniMax-H3-FL2VA_int8_convrot.safetensors model in ComfyUi, and it works well and reliably. In this case, the problem is most likely elsewhere: either the Vae or the codec, or perhaps the model wasn't loaded correctly. We need to look at the overall issue.
I downloaded another bf16 model from the ComfyUI profile, and it really worked!
I downloaded another bf16 model from the ComfyUI profile, and it really worked!
Can you share link please?
That brown pixel mess sounds like a VAE mismatch or codec issue, not just the model file itself. Since I can't browse external links or verify files directly, I can't share that specific download link for you right now. If you want to troubleshoot further, checking your ComfyUI workflow settings (especially the VAE and video encoder nodes) might help isolate the problem.
