metadata

license: apache-2.0
tags:
  - generated_from_trainer
base_model: facebook/wav2vec2-large-xlsr-53
datasets:
  - common_voice_13_0
metrics:
  - wer
model-index:
  - name: wav2vec2-XLS-R-Fleurs-demo-google-colab-Ezra_William_Prod11
    results:
      - task:
          type: automatic-speech-recognition
          name: Automatic Speech Recognition
        dataset:
          name: common_voice_13_0
          type: common_voice_13_0
          config: id
          split: test
          args: id
        metrics:
          - type: wer
            value: 0.35167772861356933
            name: Wer

wav2vec2-XLS-R-Fleurs-demo-google-colab-Ezra_William_Prod11

This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the common_voice_13_0 dataset. It achieves the following results on the evaluation set:

Loss: 0.3642
Wer: 0.3517

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: 16
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 10
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
2.9446	1.0	278	2.9150	1.0
2.8395	2.0	556	2.1459	1.0
0.9822	3.0	834	0.5701	0.5534
0.6623	4.0	1112	0.4610	0.4699
0.5834	5.0	1390	0.4262	0.4213
0.4779	6.0	1668	0.4003	0.3908
0.4511	7.0	1946	0.3802	0.3731
0.4298	8.0	2224	0.3814	0.3657
0.4029	9.0	2502	0.3637	0.3575
0.3807	10.0	2780	0.3642	0.3517

Framework versions

Transformers 4.39.3
Pytorch 2.2.2+cu121
Datasets 2.18.0
Tokenizers 0.15.2