Steps to fine-tune Wav2Vec2-Large-XLSR-53 on MGB-3

1. Prepare Data, Tokenizer, Feature Extractor
   1. Prepare Data
   2. Create Wav2Vec2CTCTokenizer
   3. Create Wav2Vec2FeatureExtractor
   4. Preprocess Data
2. Training
   1. Set-up Trainer
   2. Training
3. Evaluation

Following along Fine-tuning XLS-R for Multi-Lingual ASR with 🤗 Transformers

You can find the model on the 🤗 Hub