ad019el's picture
End of training
929da43
|
raw
history blame
1.82 kB
metadata
base_model: ad019el/Kabyle_xlsr-finetuned-tamasheq-1
tags:
  - generated_from_trainer
metrics:
  - wer
model-index:
  - name: Kabyle_xlsr-finetuned-tamasheq.en
    results: []

Kabyle_xlsr-finetuned-tamasheq.en

This model is a fine-tuned version of ad019el/Kabyle_xlsr-finetuned-tamasheq-1 on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 1.3734
  • Wer: 0.7983

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: 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: 200

Training results

Training Loss Epoch Step Validation Loss Wer
5.6267 7.89 300 2.8189 1.0
2.5968 15.79 600 1.4180 0.9182
0.9228 23.68 900 1.1546 0.8575
0.5999 31.58 1200 1.1952 0.8096
0.459 39.47 1500 1.2662 0.8068
0.3801 47.37 1800 1.3734 0.7983

Framework versions

  • Transformers 4.32.1
  • Pytorch 2.0.1+cu118
  • Datasets 2.14.4
  • Tokenizers 0.13.3