examples/speech_to_text/prep_audio_data.py · bab6c520b2e1927e9b276e2abcc2b8e57638f71f · xuchen / Fairseq-S2T · GitLab

Blame History Permalink

fix the bugs and use float32 for softmax · bab6c520
xuchen committed Feb 24, 2022

bab6c520

prep_audio_data.py 19 KB