You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
hi, I am creating an ASR model on Korean dataset.
My E2E ASR model without language model works really well, wer around 1 on dev and evaluation set.
But When I apply language model fusion (deep and cold), ctc dev loss does not converge at all and around 50% WER on eval set.
attention dev loss converges and accuracy is about 99%.
I can not find the cause of fluctuation of ctc loss and not performing well on eval set.
Thank you
The text was updated successfully, but these errors were encountered:
hi, I am creating an ASR model on Korean dataset.
My E2E ASR model without language model works really well, wer around 1 on dev and evaluation set.
But When I apply language model fusion (deep and cold), ctc dev loss does not converge at all and around 50% WER on eval set.
attention dev loss converges and accuracy is about 99%.
I can not find the cause of fluctuation of ctc loss and not performing well on eval set.
Thank you
The text was updated successfully, but these errors were encountered: