metadata

license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - audiofolder
metrics:
  - wer
model-index:
  - name: wav2vec2-xlsr-53-espeak-cv-ft-mhr-ntsema-colab
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: audiofolder
          type: audiofolder
          config: default
          split: train
          args: default
        metrics:
          - name: Wer
            type: wer
            value: 0.8127090301003345

wav2vec2-xlsr-53-espeak-cv-ft-mhr-ntsema-colab

This model is a fine-tuned version of facebook/wav2vec2-xlsr-53-espeak-cv-ft on the audiofolder dataset. It achieves the following results on the evaluation set:

Loss: 0.7728
Wer: 0.8127

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0003
train_batch_size: 4
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 32
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 500
num_epochs: 30
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
4.8463	5.79	400	1.0428	0.9331
1.4576	11.59	800	0.6796	0.8495
0.8054	17.39	1200	0.7131	0.8227
0.4946	23.19	1600	0.7202	0.8194
0.3157	28.98	2000	0.7728	0.8127

Framework versions

Transformers 4.24.0
Pytorch 1.14.0.dev20221107+cu116
Datasets 2.6.1
Tokenizers 0.13.2