Repository for pre-trained wav2vec 2.0 models on 7 Indian languages
Python
Indic-ASR
This repository contains procedure for training and inferring your own wav2vec 2.0 model. We also provide pre-trained wav2vec 2.0 weights for 7 Indian languages. Work was done for INTERSPEECH 2021 special challenge "Multilingual and code-switching ASR challenges for low resource Indian languages".
To know how to run inference on your own files, please refer to the run_inference folder. To train your own weights please refer to the run_train folder.