metadata
library_name: transformers
license: apache-2.0
base_model: openai/whisper-tiny
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: whisper-tiny-akan
results: []
whisper-tiny-akan
This model is a fine-tuned version of openai/whisper-tiny on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.1400
- Wer: 49.1687
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: 0.0001
- train_batch_size: 16
- 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: 500
- training_steps: 2000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.3923 | 5.0 | 250 | 0.8148 | 61.9777 |
| 0.1045 | 10.0 | 500 | 0.9445 | 53.9829 |
| 0.0414 | 15.0 | 750 | 1.0241 | 49.9639 |
| 0.0187 | 20.0 | 1000 | 1.0862 | 51.7999 |
| 0.0089 | 25.0 | 1250 | 1.1386 | 51.1927 |
| 0.0025 | 30.0 | 1500 | 1.1284 | 49.1832 |
| 0.0004 | 35.0 | 1750 | 1.1355 | 48.7061 |
| 0.0004 | 40.0 | 2000 | 1.1400 | 49.1687 |
Framework versions
- Transformers 4.44.2
- Pytorch 2.4.0+cu121
- Datasets 3.0.0
- Tokenizers 0.19.1