Hacker News Reader

It's impressive given the constraints!

Would you consider releasing a more capable version that renders with fewer artifacts (and maybe requires a bit more processing power)?

Chatterbox is my go-to, this could be a nice alternative were it capable of high-fidelity results!

View on HN

This is my side “hobby”. And compute is quite expensive. But if the community’s responsive is good, I will definitely think about it! Btw, chatterbox is a great model and inspiration

Very cool work, especially for a hobby project.

Do you have any plans to publish a blog post on how you did that? ?What training data and how much? Your training and ablations methodology, etc.

Thanks can you share details about compute economics you dealt with ?

Yeah sure. The training was about ~250 dollars, which is quite low by today’s standards. And I spent a bit more on ablations and research

I was on similar path and saw my bills going over 1000 dollars as interests to do research and ablations grew. Then I decided to get one Blackwell Pro 6000 and trying things with that :) If you have suggestions on how to manage metrics let us know. Currenty trying langfuse since its one click install on coolify

Is that something that could be done on a local setup? Eg, 2x RTX3090?

Reply on news.ycombinator.com