@btp4z7
Founder @join_hallway
ID: 118406118
calendar_today28-02-2010 15:48:35
1,1K Tweet
2,2K Takipçi
866 Takip Edilen
a month ago
A lossy fast feature encoder achieves the same performance as a low wer encoder if the decoder model is trained with the lossy source. We all should save the transformers for the decoder step.