Evaluate for speech models? #4

by patrickvonplaten - opened

Are there plans to support the model evaluator for speech recognition? It would be nice to evaluate Wav2Vec2 models on Librispeech

Hey @patrickvonplaten, indeed this would be very nice to support! I think we could do this without too much effort on the AutoTrain side - WDYT @abhishek?

This would be great @patrickvonplaten πŸ—£

any updates here ? would love to see this since whisper becomes more popular