benchmarks icon indicating copy to clipboard operation
benchmarks copied to clipboard
verified publisher icon speechbrain

DASB pipeline implementation for Common Voice

Open ana-kuznetsova opened this issue 6 months ago • 2 comments

Common Voice implementation for DASB tokenizer evaluation.

  • Unification of offline token extraction extract.py
  • Unification of training and data pipeline in train.py Implemented for two architectures: LSTM and Branchformer.

ana-kuznetsova avatar Jul 06 '25 15:07 ana-kuznetsova

I have updated the main and DASB branch to address the CI failures due to a github change, merging the upstream DASB into this branch should make this CI run again.

pplantinga avatar Jul 15 '25 15:07 pplantinga

@ana-kuznetsova could you please resolve the conflict

poonehmousavi avatar Jul 23 '25 19:07 poonehmousavi