---
library_name: transformers
license: apache-2.0
base_model: openai/whisper-large-v3
tags:
- whisper-event
- generated_from_trainer
datasets:
- asierhv/composite_corpus_eu_v2.1
language:
- eu
metrics:
- wer
model-index:
- name: Whisper Large Basque
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: Common Voice 17.0
      type: mozilla-foundation/common_voice_17_0
      config: eu
      split: test
      args:
        language: eu
    metrics:
      - name: Test WER
        type: wer
        value: 4.47
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: asierhv/composite_corpus_eu_v2.1
      type: asierhv/composite_corpus_eu_v2.1
    metrics:
    - name: Wer
      type: wer
      value: 7.100121529400767
---

<!-- 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 Large Basque

This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the asierhv/composite_corpus_eu_v2.1 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1407
- Wer: 7.1001

## 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: 4.375e-06
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- training_steps: 20000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step  | Validation Loss | Wer     |
|:-------------:|:-----:|:-----:|:---------------:|:-------:|
| 0.2854        | 0.05  | 500   | 0.3763          | 24.9836 |
| 0.1425        | 0.1   | 1000  | 0.3326          | 19.8654 |
| 0.2196        | 0.15  | 1500  | 0.2802          | 16.2475 |
| 0.2338        | 0.2   | 2000  | 0.2536          | 14.6116 |
| 0.1383        | 0.25  | 2500  | 0.2451          | 12.8961 |
| 0.0848        | 0.3   | 3000  | 0.2280          | 12.2464 |
| 0.0854        | 0.35  | 3500  | 0.2152          | 11.4144 |
| 0.1304        | 0.4   | 4000  | 0.2097          | 11.1433 |
| 0.1328        | 0.45  | 4500  | 0.2055          | 10.6011 |
| 0.0737        | 0.5   | 5000  | 0.2079          | 10.5357 |
| 0.0804        | 0.55  | 5500  | 0.2133          | 10.1150 |
| 0.0964        | 0.6   | 6000  | 0.1988          | 9.4606  |
| 0.0811        | 0.65  | 6500  | 0.2019          | 9.4933  |
| 0.0677        | 0.7   | 7000  | 0.1916          | 8.9231  |
| 0.1114        | 0.75  | 7500  | 0.2029          | 9.3250  |
| 0.1142        | 0.8   | 8000  | 0.1895          | 8.9978  |
| 0.0466        | 0.85  | 8500  | 0.1936          | 8.8576  |
| 0.0664        | 0.9   | 9000  | 0.1876          | 8.9698  |
| 0.0759        | 0.95  | 9500  | 0.1827          | 8.8202  |
| 0.0555        | 1.0   | 10000 | 0.1834          | 8.6426  |
| 0.0603        | 0.525 | 10500 | 0.1872          | 9.3344  |
| 0.0727        | 0.55  | 11000 | 0.1838          | 9.3624  |
| 0.0523        | 0.575 | 11500 | 0.2022          | 8.8903  |
| 0.0719        | 0.6   | 12000 | 0.1840          | 9.0072  |
| 0.0505        | 0.625 | 12500 | 0.1860          | 8.5631  |
| 0.0678        | 0.65  | 13000 | 0.1852          | 8.1238  |
| 0.0586        | 0.675 | 13500 | 0.1888          | 8.7641  |
| 0.0818        | 0.7   | 14000 | 0.1822          | 8.2547  |
| 0.0583        | 0.725 | 14500 | 0.1349          | 7.8760  |
| 0.0516        | 0.75  | 15000 | 0.1432          | 7.8386  |
| 0.0721        | 0.775 | 15500 | 0.1439          | 7.7966  |
| 0.0697        | 0.8   | 16000 | 0.1345          | 7.6470  |
| 0.0459        | 0.825 | 16500 | 0.1381          | 7.4881  |
| 0.0533        | 0.85  | 17000 | 0.1422          | 7.2871  |
| 0.0449        | 0.875 | 17500 | 0.1426          | 7.7218  |
| 0.0424        | 0.9   | 18000 | 0.1417          | 7.4367  |
| 0.0714        | 0.925 | 18500 | 0.1337          | 6.9973  |
| 0.0573        | 0.95  | 19000 | 0.1432          | 7.6657  |
| 0.0441        | 0.975 | 19500 | 0.1408          | 7.1001  |
| 0.0453        | 1.0   | 20000 | 0.1407          | 7.1001  |


### Framework versions

- Transformers 4.49.0.dev0
- Pytorch 2.6.0+cu124
- Datasets 3.2.1.dev0
- Tokenizers 0.21.0