---
license: apache-2.0
base_model: openai/whisper-tiny
tags:
- generated_from_trainer
datasets:
- PolyAI/minds14
metrics:
- wer
model-index:
- name: whisper-tiny-en-US
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: PolyAI/minds14
      type: PolyAI/minds14
      config: en-US
      split: train
      args: en-US
    metrics:
    - name: Wer
      type: wer
      value: 0.3582458307597282
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# whisper-tiny-en-US

This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the PolyAI/minds14 dataset.
It achieves the following results on the evaluation set:
- Loss: 1.0645
- Wer Ortho: 0.3623
- Wer: 0.3582

## 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: 1000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch   | Step | Validation Loss | Wer Ortho | Wer    |
|:-------------:|:-------:|:----:|:---------------:|:---------:|:------:|
| 0.0003        | 1.7857  | 50   | 0.9458          | 0.3862    | 0.3792 |
| 0.0018        | 3.5714  | 100  | 0.9438          | 0.3578    | 0.3490 |
| 0.0021        | 5.3571  | 150  | 0.9330          | 0.3894    | 0.3811 |
| 0.0008        | 7.1429  | 200  | 0.9398          | 0.3972    | 0.3928 |
| 0.0003        | 8.9286  | 250  | 0.9701          | 0.3985    | 0.3953 |
| 0.0006        | 10.7143 | 300  | 0.9799          | 0.3965    | 0.3928 |
| 0.0002        | 12.5    | 350  | 0.9913          | 0.3669    | 0.3644 |
| 0.0001        | 14.2857 | 400  | 0.9941          | 0.3540    | 0.3490 |
| 0.0001        | 16.0714 | 450  | 1.0054          | 0.3585    | 0.3545 |
| 0.0001        | 17.8571 | 500  | 1.0149          | 0.3585    | 0.3545 |
| 0.0001        | 19.6429 | 550  | 1.0218          | 0.3527    | 0.3490 |
| 0.0001        | 21.4286 | 600  | 1.0279          | 0.3598    | 0.3564 |
| 0.0           | 23.2143 | 650  | 1.0336          | 0.3585    | 0.3552 |
| 0.0           | 25.0    | 700  | 1.0389          | 0.3598    | 0.3564 |
| 0.0           | 26.7857 | 750  | 1.0434          | 0.3611    | 0.3570 |
| 0.0           | 28.5714 | 800  | 1.0486          | 0.3636    | 0.3595 |
| 0.0           | 30.3571 | 850  | 1.0529          | 0.3636    | 0.3595 |
| 0.0           | 32.1429 | 900  | 1.0571          | 0.3623    | 0.3582 |
| 0.0           | 33.9286 | 950  | 1.0607          | 0.3623    | 0.3582 |
| 0.0           | 35.7143 | 1000 | 1.0645          | 0.3623    | 0.3582 |


### Framework versions

- Transformers 4.44.0
- Pytorch 2.4.0
- Datasets 2.21.0
- Tokenizers 0.19.1