Training AWE - Githubissues

DataGov-SamagraX / hindi-acoustic-word-embedding

0 stars 1 forks source link

Closed xorsuyash closed 3 weeks ago

xorsuyash commented 2 months ago

Tasks

[x] Coding architecture of AWE model.
[x] Setting up training and testing pipeline, dataloader + trainer
[ ] Devising better strategy for force alignment
[ ] Integrate ASR in force alignment if transcript is not available possilble( ai4bharat + kenLM)
[ ] Devising strategy to cut and merge using voice activity detection for incoming audio in force aligner.

xorsuyash commented 2 months ago

@GautamR-Samagra coded architecture of AWE in a proper setting, setting up testing pipeline and eval pipeline. #commit