Modelpublic

StephennFernandes/wav2vec2-XLS-R-300m-konkani

sourceHugging Faceupdated 5y agoView on Hugging Face

0likes13downloads

Model Card

tags:

—automatic-speech-recognition
—robust-speech-event

This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on a private dataset. It achieves the following results on the evaluation set:

The following hyper-parameters were used during training:

—learning_rate: 3e-4
—trainbatchsize: 32
—evalbatchsize: 16
—seed: 42
—gradientaccumulationsteps: 2
—totaltrainbatch_size: 128
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—lrschedulerwarmup_steps: 800
—num_epochs: 30
—mixedprecisiontraining: Native AMP