Training Fastconformer-CTC on Kazakh Language #9506
Unanswered
diasbalmash
asked this question in
Q&A
Replies: 1 comment 1 reply
-
We've found that almost without exception, starting from a pertained checkpoint, no matter the language will do better than starting from scratch. You can get better scores of the pretraining language is similar to the new language, but usually not significantly different than starting with a model trained on thousands of hours of English |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hello, community. I just started working with NeMo. So, I want to train ASR based on Fastconformer-CTC with 100 hours of recorded calls in kazakh language. What type of training will be better for my situation, fine-tuning some pre-trained model or train it from scratch? Also can someone give info about the data processing and manifest?
Beta Was this translation helpful? Give feedback.
All reactions