Disclaimer: I'm actively working on this tool.

This is in a closed beta for now (while we work on provisioning enough GPU compute) but we're hoping to make this public later.

Is the model designed to run locally or does it run in the cloud somewhere?

Is it your own model or is it based on sdxl or something like that?

cloud, A100, so yeah, beast.

we are doing tests, will keep you guys posted.

Which cloud? I've been looking for somewhere to fine-tune Pixart, but EC2 et al. are usurious.

If you want to run a batch job, look into vast.ai.

Yeah, indeed. I know LCMs are way faster than diffusion models, but surely this demo is running on a beastly GPU...

yup, see comment above!