Search papers, labs, and topics across Lattice.
1
0
3
Compressing 60-second audio into just 788 tokens, this new autoencoder makes generative audio modeling far more tractable by slashing encoding time and latent rates.
