README.md

metadata

language:
  - hre
license: apache-2.0
base_model: openai/whisper-small
tags:
  - generated_from_trainer
datasets:
  - ntviet/hre-audio-dataset2
metrics:
  - wer
model-index:
  - name: Whisper Small Hre 2.1 - 1500 steps - NT Viet
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Hre audio dataset 2
          type: ntviet/hre-audio-dataset2
          config: default
          split: test
          args: default
        metrics:
          - name: Wer
            type: wer
            value: 79.1044776119403

Whisper Small Hre 2.1 - 1500 steps - NT Viet

This model is a fine-tuned version of openai/whisper-small on the Hre audio dataset 2 dataset. It achieves the following results on the evaluation set:

• Loss: 3.4286
• Wer Ortho: 79.5455
• Wer: 79.1045

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
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Ortho	Wer
0.0049	12.4	1500	3.4286	79.5455	79.1045

Framework versions

• Transformers 4.38.2
• Pytorch 2.1.0+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2