gemini image manipulation is pure science fiction

www.outatime-podcast.de

After only 2 episodes, our podcast #outatime is already in the top 200 German charts and no. 10 in the science category! Mind = blown!

www.outatime-podcast.de

Looks very cool but would be even cooler if we could see the code to find out which models etc are being used

Still one of the best albums of all time. The fact they recorded it all in one room in a live setup makes it even more impressive

New lightweight text-to-speech model OuteTTS just dropped and it sounds pretty good! github.com/edwko/OuteTTS

I forgot how damn fast the styletts2 text-to-speech model is on a gpu... it's only single voice but for this speed, it's hard to beat and this includes a pass through llama3.2 3B on the same machine with only 8GB of vram!

Ngl this is based! I especially like the last paragraph!