@josephjojoe how much of this was done with coding agents? pytorch to mlx seems like a super rl-able task for a coding agent, you basically already defined the reward functions in benchmarks
@austin_malerba yes, tts is awesome! recently made a simple swift app to narrate arbitrary content using Google's free tts models though I'm sure ElevenLabs would do even better (and could narrate in the author's voice)
https://t.co/5IE4K4SCZh
@josephjojoe@modal interesting, looks like most 1 syllable words default to stressed? maybe something to do with the dataset labeling them w/ primary stress by default?
new update!
- /leaderboard: browse top voted sites
- improved detection for broken and non-personal sites - better algorithm for recommending profiles
- updated UX
feel free to share additional feedback here + contact me for inquiries about data