Hi
@stweil ,
at the time of training, I have added model_name.wordlist, model_name.punc, model_name.numbers in data/model_name directory. the traindata resulted as 15mb file,
after unpacking the ~15mb traindata(my fine-tuned model) and ~22mb traindata( original traindata ), i found there are some files missing in fine-tuned model, so is it really necessary to add all those missing files in my fine-tuned model before inference ?
please find attachment of extracted folder structure of fine-tuned model and original traindata.