EzraWilliam's picture
Upload tokenizer
a75619f verified
|
raw
history blame
2.96 kB
metadata
license: apache-2.0
tags:
  - generated_from_trainer
base_model: facebook/wav2vec2-large-xlsr-53
datasets:
  - xtreme_s
metrics:
  - wer
model-index:
  - name: wav2vec2-XLS-R-Fleurs-demo-google-colab-Ezra_William_Prod9
    results:
      - task:
          type: automatic-speech-recognition
          name: Automatic Speech Recognition
        dataset:
          name: xtreme_s
          type: xtreme_s
          config: fleurs.id_id
          split: test
          args: fleurs.id_id
        metrics:
          - type: wer
            value: 0.9842089507558749
            name: Wer

wav2vec2-XLS-R-Fleurs-demo-google-colab-Ezra_William_Prod9

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

  • Loss: 1.3012
  • Wer: 0.9842

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: 0.001
  • 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: 20
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
No log 1.0 39 2.8812 1.0
No log 2.0 78 2.8688 1.0
No log 3.0 117 2.8722 1.0
No log 4.0 156 2.8640 1.0
No log 5.0 195 2.8447 1.0
No log 6.0 234 2.8468 1.0
No log 7.0 273 2.8465 1.0
No log 8.0 312 2.8488 1.0
No log 9.0 351 2.8355 1.0
No log 10.0 390 2.8167 1.0
No log 11.0 429 2.8076 1.0
No log 12.0 468 2.7065 1.0
2.9881 13.0 507 2.5506 1.0
2.9881 14.0 546 2.2657 1.0
2.9881 15.0 585 1.9921 1.0
2.9881 16.0 624 1.7390 1.0
2.9881 17.0 663 1.5309 1.0
2.9881 18.0 702 1.4300 0.9994
2.9881 19.0 741 1.3280 0.9938
2.9881 20.0 780 1.3012 0.9842

Framework versions

  • Transformers 4.39.1
  • Pytorch 2.2.1+cu121
  • Datasets 2.18.0
  • Tokenizers 0.15.2