Czech Kaldi Profile
December 8, 2020 ยท View on GitHub
A Rhasspy profile for Czech (cs).
Includes:
- A Kaldi nnet3 speech to text model
- See files in
acoustic_model/ - Recipe created with ipa2kaldi
- Word error rate: 11.03%
- Trained on:
- Vystadial VOIP (18.2 hours)
- Common Voice (26.6 hours)
- See files in
- An IPA pronunciation lexicon pulled from Wiktionary
- See
base_dictionary.txt.gz
- See
- A phonetisaurus grapheme to phoneme model for predicting word pronunciations
- See
g2p.fst.gz - Trained on
base_dictionary.txt.gz
- See
- A tri-gram ARPA language model
- See
base_language_model.txt.gz - Text from audio transcriptions and Universal Dependencies
- See