ALBADDAWI's picture
ft-wav2vec2-with-minds
1cebb6a verified
metadata
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: ft-wav2vec2-with-minds
    results: []

ft-wav2vec2-with-minds

This model was trained from scratch on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0333
  • Accuracy: 0.9972

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: 120
  • eval_batch_size: 120
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 480
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 10
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Accuracy
4.6092 1.0 9 2.4860 0.4311
1.4641 2.0 18 0.5758 0.7826
0.5061 3.0 27 0.1966 0.9756
0.2573 4.0 36 0.1038 0.9803
0.1557 5.0 45 0.0671 0.9859
0.1235 6.0 54 0.0333 0.9972
0.0725 7.0 63 0.0334 0.9944
0.0914 8.0 72 0.0279 0.9953
0.1695 9.0 81 0.0276 0.9972
0.1118 10.0 90 0.0290 0.9972

Framework versions

  • Transformers 4.35.2
  • Pytorch 1.12.1+cu116
  • Datasets 2.15.0
  • Tokenizers 0.15.2