README.md

metadata

license: apache-2.0
base_model: facebook/wav2vec2-base-960h
tags:
  - generated_from_trainer
datasets:
  - fleurs
metrics:
  - wer
model-index:
  - name: wav2vec2-base-fleurs-329-colab-a100-2
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: fleurs
          type: fleurs
          config: en_us
          split: test
          args: en_us
        metrics:
          - name: Wer
            type: wer
            value: 0.9917617237008872

wav2vec2-base-fleurs-329-colab-a100-2

This model is a fine-tuned version of facebook/wav2vec2-base-960h on the fleurs dataset. It achieves the following results on the evaluation set:

• Loss: 2.5431
• Wer: 0.9918

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: 1e-05
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• 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: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
8.5696	2.45	200	5.0490	1.0
4.0587	4.91	400	3.3808	1.0
3.0272	7.36	600	2.7368	0.9954
2.6659	9.82	800	2.5431	0.9918

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2