rutsam's picture
Update README.md
558dd54
|
raw
history blame
940 Bytes
metadata
library_name: fastpitch
task: text-to-speech
tags:
  - fastpitch
  - waveglow
  - text-to-speech
language: rw
datasets:
  - Custom made
widget:
  - text: Muraho neza, murakaza neza mu Rwanda.
    example_title: Muraho neza, murakaza neza mu Rwanda.

Model card - Kinyarwanda TTS model

Model details

Metrics

  • We use Mean Opinion Score (MOS) to evaluate the model with a maximum score being 5
Test Corpus MOS
Custom phrases 3

Challenges

  • The model does not always capture the Kinyarwanda tones

Recommendations

  • Use a tonal dictionary to train future models
  • Add numbers and symbols Dictionary