automatic speech recognition · transformers model
This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on the OPENSLRSLR53 - bengali dataset. It achieves the following results on the evaluation set.