FEATURE REQUEST: Add more supporting data types

wenet-e2e / wespeaker

Research and Production Oriented Speaker Verification, Recognition and Diarization Toolkit

Apache License 2.0

664 stars 112 forks source link

Open evadotfm opened 1 month ago

evadotfm commented 1 month ago

good work first of all. It would be nice other than .wav files, it also supports more data types. numpy or tensor??

JiJiJiang commented 1 month ago

Sure, we will support loading numpy and tensor in the cli in the near future. Thank you for your advice.