Fine Tuning model

I am unable to find how to approach the fine-tuning part properly. Someone please let me know where I can start.

What do you want to do exactly, what material do you have, give us info :slight_smile:

Please correct me if I am wrong sir. I was going through the documentation. What I learnt is in order to fine tune we need the train.csv, test.csv,dev.csv files, which can be generated by downloading the english dataset from common voice(38 GB) and process it with import_cv2.py. Is it the only way to fine tune the model or we can do something else like with some less data ? I really need help with this.

If you really need help, then let us know what you want to do and what data you have to do that. Otherwise read the documentation, which has all the infos:

https://deepspeech.readthedocs.io/en/v0.7.1/TRAINING.html#fine-tuning-same-alphabet