---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- common_voice_11_0
metrics:
- wer
model-index:
- name: openai/whisper-medium
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: common_voice_11_0
      type: common_voice_11_0
      config: pt
      split: test
      args: pt
    metrics:
    - name: Wer
      type: wer
      value: 6.5785713084850626
---

<!-- 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. -->

# openai/whisper-medium

This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the common_voice_11_0 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3205
- Wer: 6.5786

## 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: 32
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 5000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer    |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 0.0698        | 1.09  | 1000 | 0.1876          | 7.1889 |
| 0.0218        | 3.07  | 2000 | 0.2254          | 7.1098 |
| 0.0053        | 5.06  | 3000 | 0.2711          | 6.9686 |
| 0.0017        | 7.04  | 4000 | 0.3030          | 6.6862 |
| 0.0005        | 9.02  | 5000 | 0.3205          | 6.5786 |


### Framework versions

- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu117
- Datasets 2.7.1.dev0
- Tokenizers 0.13.2