whisper-small-eg / README.md
tawreck-hasaballah's picture
End of training
9ed8023 verified
|
raw
history blame
1.66 kB
metadata
language:
  - eg
license: apache-2.0
base_model: openai/whisper-small
tags:
  - generated_from_trainer
metrics:
  - wer
model-index:
  - name: Whisper Small Eg - Tariq Hasaballah 100330
    results: []

Whisper Small Eg - Tariq Hasaballah 100330

This model is a fine-tuned version of openai/whisper-small on the ASR-EGARBCSC: AN EGYPTIAN ARABIC CONVERSATIONAL SPEECH CORPUS dataset. It achieves the following results on the evaluation set:

  • Loss: 3.2253
  • Wer: 119.8092

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: 16
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: constant_with_warmup
  • lr_scheduler_warmup_steps: 50
  • training_steps: 1500

Training results

Training Loss Epoch Step Validation Loss Wer
1.6565 3.27 500 2.3919 121.5667
0.5477 6.54 1000 2.8503 117.3487
0.0882 9.8 1500 3.2253 119.8092

Framework versions

  • Transformers 4.39.3
  • Pytorch 2.2.1+cu121
  • Datasets 2.18.0
  • Tokenizers 0.15.2