Doesn't work

#4
by Blursed - opened

Errors out when generating Lora.

I tried 5 images, all no greater than 1024 pixels on any single edge.

Just errors out.

Tried some of the other Hugging Face spaces. Those error out too.

After checking, I suspect the error is from HuggingFace.

Please try again later, they usually patch it in a few hours.

can we run this locally? without huggingface?

can we run this locally? without huggingface?

I just got down running it locally. Create the lora file is to problem and works quick. The issue is when it want to inference an image its using an un quantized version of qwen and after 60GB of download that model its too large. Thats okay though I 'll generate in comfy, would be dope if I could have said NO to downloading a model that large. I can confirm the online huggingface space is working fine too.

Sign up or log in to comment