Index of /opt/mod9-asr/models/asr/mod9
Name Size Description
Parent Directory -
en_video/ - Default model recommended for most English use cases, including global accents.
en-US_phone/ - Trained for North American accented English spoken over telephone.
es-US_phone/ - Trained for Latin American accented Spanish spoken over telephone.
en-US_phone-ivector/ - Optimized for long single-speaker audio, this model is less robust than en-US_phone.
en-US_phone-smaller/ - This model is extremely small, useful testing and debugging purposes.
de_video/ - Based on a TUDA recipe, this model was trained by Mod9 for improved robustness.
de_video-smaller/ - This model is somewhat smaller, useful to deployed on memory-limited servers.
en-US_phone-benchmark/ - [DEPRECATED ALIAS]