3 ms·
They (deepmind) reported it took 90 minutes of processing to generate 1s of speech via tweet. Hopefully this comes down in the future.
by unlikelymordant 10y ago
They (deepmind) reported it took 90 minutes of processing to generate 1s of speech via tweet. Hopefully this comes down in the future.
- espadrine 10y agoDo you have a link? This implementation says: “A Tesla K80 needs around ~4 minutes for generating a second of audio at a sampling rate of 4000hz”, which is significantly faster.
- basve 10y ago90 minutes for 1s of audio was reported by someone from Google on twitter, but the tweet has been deleted. I've clarified in the readme that my measurements are for a much lighter/smaller model than Deepmind's :).