[Help] How to test own training data in Deep Speech

This post was flagged by the community and is temporarily hidden.

Firstly, you don’t really need to retrain the model, just create your own language model since it is English that you are trying with.

Secondly, see evaluate.py for evaluation and WER calculation on set of audio files.

Thanks for anwering me and giving me some advice. I see some mistakes that i made when i read this: TUTORIAL : How I trained a specific french model to control my robot , so i am going to rebuild my project again with that instruction. I will read evaluate.py now. However, i want to create an Vietnamese LM, the English LM i only used to learn tutorial.

@20164356 Please don’t post screenshots for console output. We can’t help. I barely see an issue with your LM. Make sure you follow the doc and that git-lfs works, in your case, it seems it is not the case.

You need to install sox and all of its dependencies.

Sorry about that.I will notice it in later post. You say right that git - lfs don’t work in my computer ( i remembered to install, but it didn’t work). And i don’t know git -lfs do.

Yes, it’s my problem. I try to install LAME and LIBMAD then but after that sox still not working