I need a commented example in python using mozilla deepspeech technology, first of all for Linux in docker, i.e. scripts should not have functions and import from windows, etc.! If you're developing without docker that's fine too, but the DeepSpeech version should be => 0.9.
1. Script training in python:
the script should work on training a couple of dozen words from a text file and directly from the microphone. a word from the file is displayed on the screen, I read and speak, it learns. the script should work with two languages. this is English + any other - German for example. the main thing is different languages = different folders.
2. further, the second simpler script in python (of which there are a lot on the Internet). the trained one needs to be connected so that I can check how well the module has learned.
THE MAIN THING
- no frontend. everything should work in the terminal.
- fewer parameters for the script, it is desirable to simply indicate the folder with the files. for example: [login to view URL] --folder = "./ data / english /"!
- and vice versa: [login to view URL] --folder = "./ data / english"
- all this should work on the latest version of DeepSpeech => 0.9 from [login to view URL], so that there are no problems with installing and checking the work for me.
- no extra comments needed. no need to comment file opened file closed. only that which pertains to DeepSpeech technology.
... maybe someone already has this - contact me.