metadata
license: apache-2.0
base_model: openai/whisper-small
tags:
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_6_1
metrics:
- wer
model-index:
- name: Whisper Small Frisian 10m
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice 6.1
type: mozilla-foundation/common_voice_6_1
args: 'config: frisian, split: test'
metrics:
- name: Wer
type: wer
value: 64.62662626982713
Whisper Small Frisian 10m
This model is a fine-tuned version of openai/whisper-small on the Common Voice 6.1 dataset. It achieves the following results on the evaluation set:
- Loss: 1.6643
- Wer: 64.6266
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-06
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- training_steps: 1000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 2.0562 | 6.6667 | 100 | 2.2740 | 83.4860 |
| 0.9187 | 13.3333 | 200 | 1.7749 | 76.7385 |
| 0.4133 | 20.0 | 300 | 1.6576 | 69.3317 |
| 0.1671 | 26.6667 | 400 | 1.6334 | 67.2679 |
| 0.0603 | 33.3333 | 500 | 1.6319 | 66.2770 |
| 0.0245 | 40.0 | 600 | 1.6433 | 65.6496 |
| 0.0138 | 46.6667 | 700 | 1.6522 | 64.6730 |
| 0.0104 | 53.3333 | 800 | 1.6591 | 64.5019 |
| 0.0089 | 60.0 | 900 | 1.6632 | 64.4876 |
| 0.0083 | 66.6667 | 1000 | 1.6643 | 64.6266 |
Framework versions
- Transformers 4.40.1
- Pytorch 2.3.0+cu121
- Datasets 2.19.0
- Tokenizers 0.19.1