cog-sdxl-gui

[bmaltais]

Just a heads up that I have started work on a webui for cog training similar to the one for kohya_ss.

If any of you guys are interested to contribute code you are welcome.

Link to the repo if you are interested. Hoping to have a MVP by end of day: https://github.com/bmaltais/cog-sdxl

[bmaltais]

Talking with a few model builders and devs we generally find that the 3 files produced by the trainer will make it difficult to share.

One suggestion... would it be possible to update the output to put both the embedding and embedding parameters as metadata in the safetensors Lora file?

That way we could share them easily via civitai.com and GUI like auto1111 and comfy could just get the embedding and parameters from the safetensors file...

[bmaltais]

The MVP for training is ready for use:

[bmaltais]

The cog-sdxl-webui will now generate a proper embeddings.safetensors that can be used by ComfyUI to generate outputs from the lora.safetensors + embeddings.safetensors files.

Image generated by ComFY UI: