Instructions to use alibaba-pai/Z-Image-Fun-Lora-Distill with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- VideoX Fun
How to use alibaba-pai/Z-Image-Fun-Lora-Distill 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
Fun-Lora-Distill-8-Steps-2602-ComfyUI loss of denoising progress
I'm not sure if this is an issue with comfy or with the lora, but the 8 step 2602 loses the denoising progress and rolls over to the start. Happens even on the recommended 8 steps and it loses track on the last one or two steps.
The 4 step version and the non-2602 are working properly.Check this out, I'm training a version without FFN.
https://huggingface.co/alibaba-pai/Z-Image-Fun-Lora-Distill/discussions/5
Check this out, I'm training a version without FFN.
https://huggingface.co/alibaba-pai/Z-Image-Fun-Lora-Distill/discussions/5
Hi, I just want to mention a few more issues with your current lora that I've discovered.
1, when using a denoising strength of less than 0.4 during inpainting, it creates a complete blur. If I raise it to 0.41, it works normally.
2, if using a shift of less than 8, it also creates just a blur, only while inpainting. I can use a lower shift when doing T2I.