metadata

library_name: transformers
license: apache-2.0
base_model: facebook/wav2vec2-base
tags:
  - generated_from_trainer
datasets:
  - timit_asr
metrics:
  - wer
model-index:
  - name: wav2vec2-base-timit-demo-google-colab
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: timit_asr
          type: timit_asr
          config: clean
          split: None
          args: clean
        metrics:
          - name: Wer
            type: wer
            value: 0.3354696437185583

wav2vec2-base-timit-demo-google-colab

This model is a fine-tuned version of facebook/wav2vec2-base on the timit_asr dataset. It achieves the following results on the evaluation set:

Loss: 0.4743
Wer: 0.3355

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: 0.0001
train_batch_size: 32
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 1000
num_epochs: 30
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
3.4394	4.0	500	1.2662	0.8530
0.5192	8.0	1000	0.4308	0.4176
0.1896	12.0	1500	0.4249	0.3656
0.1158	16.0	2000	0.4405	0.3583
0.0791	20.0	2500	0.4949	0.3481
0.0578	24.0	3000	0.4895	0.3448
0.0462	28.0	3500	0.4743	0.3355

Framework versions

Transformers 4.45.1
Pytorch 2.1.2
Datasets 3.0.1
Tokenizers 0.20.0