We currently have a semi-functioning G2P model. However, the model's accuracy is very low (~5%) and overfits the data. We need to optimize/bug fix the model to improve the accuracy of our phonetic translations.
Acceptance Criteria
We should aim for a model accuracy of at least 50% (subject to change later, can be tested in evaluate-model.py).
Description
We currently have a semi-functioning G2P model. However, the model's accuracy is very low (~5%) and overfits the data. We need to optimize/bug fix the model to improve the accuracy of our phonetic translations.
Acceptance Criteria
evaluate-model.py).Dev Notes
Subtasks