Bakasura's picture
update model card README.md
cc462b7
metadata
license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - superb
metrics:
  - accuracy
model-index:
  - name: wav2vec2-base-finetuned-ks
    results: []

wav2vec2-base-finetuned-ks

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

  • Loss: 0.0823
  • Accuracy: 0.9819

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: 3e-05
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 32
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 5

Training results

Training Loss Epoch Step Validation Loss Accuracy
0.3582 1.0 1596 0.1846 0.9681
0.2013 2.0 3192 0.1051 0.9776
0.1656 3.0 4788 0.0823 0.9819
0.1363 4.0 6384 0.0801 0.9819
0.0918 5.0 7980 0.0838 0.9809

Framework versions

  • Transformers 4.25.1
  • Pytorch 1.13.0+cu117
  • Datasets 2.7.1
  • Tokenizers 0.13.2