Skip to content

Hyfforddi modelau adnabod lleferydd Cymraeg gyda Mozilla DeepSpeech // Train Welsh language speech recognition models with Mozilla DeepSpeech.

License

Notifications You must be signed in to change notification settings

techiaith/docker-deepspeech-cy

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

docker-deepspeech-cy

Hwyluso hyfforddi Mozilla DeepSpeech ar gyfer adnabod lleferydd i'r Gymraeg gyda data Mozilla CommonVoice Cymraeg a Chorpws Paldaruo

Making it easier to train Mozilla DeepSpeech for Welsh language speech recognition using the Welsh data in Mozilla CommonVoice datasets as well as the Paldaruo Speech Corpus


Rhagofynion / Pre-requisites

  • Docker ( fersiwn/version >= 19)
  • cardyn graffig â GPUs yn eich gyfrifiadur (e.e. Nvidia GeForce GTX 1080 Ti, Nvidia GeForce RTX 2080 Ti)
    a graphics card with GPUs on your system (e.g. Nvidia GeForce GTX 1080 Ti, Nvidia GeForce RTX 2080 Ti)

Gweler/See also : https://github.com/NVIDIA/nvidia-docker#quickstart


Gosod / Installation

$ git clone https://github.com/techiaith/docker-deepspeech-cy
$ cd docker-deepspeech-cy
$ make
$ make run

Data Cymraeg Mozilla CommonVoice / Mozilla Common Welsh Data

Llwythwch y data diweddaraf i lawr o https://voice.mozilla.org/cy/datasets i'r ffolder docker-deepspeech-cy/data.

Download the latest data from https://voice.mozilla.org/cy/datasets to the docker-deepspeech-cy/datafolder.

Hyfforddi / Training

Gweler README.md

See README.md