I want to use the phonemes from the languages in "latest" to create phoneme text files with Epitran. 


Hello, I followed the rules in Allovera to generate a phoneme text file using Epitran, but I found that the phonemes for several languages in Allovera are different from the phonemes corresponding to your model's languages. How can I use your phoneme set to generate phoneme files with Epitran?