metadata
language:
- en
license: apache-2.0
tags:
- generated_from_trainer
base_model: openai/whisper-small
datasets:
- Jzuluaga/atcosim_corpus
metrics:
- wer
model-index:
- name: Whisper Base ATCOSIM
results:
- task:
type: automatic-speech-recognition
name: Automatic Speech Recognition
dataset:
name: atcosim_corpus
type: Jzuluaga/atcosim_corpus
args: 'config: en, split: test'
metrics:
- type: wer
value: 4.169242999735006
name: Wer
Whisper Base ATCOSIM
This model is a fine-tuned version of openai/whisper-small on the atcosim_corpus dataset. It achieves the following results on the evaluation set:
- Loss: 0.0557
- Wer: 4.1692
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: 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: 1000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 1.6572 | 0.2092 | 100 | 1.2694 | 71.0096 |
| 0.3373 | 0.4184 | 200 | 0.3225 | 16.4296 |
| 0.147 | 0.6276 | 300 | 0.1653 | 9.2130 |
| 0.0975 | 0.8368 | 400 | 0.1107 | 6.3334 |
| 0.0544 | 1.0460 | 500 | 0.0892 | 6.2583 |
| 0.0344 | 1.2552 | 600 | 0.0760 | 5.4103 |
| 0.0517 | 1.4644 | 700 | 0.0663 | 5.0393 |
| 0.0396 | 1.6736 | 800 | 0.0607 | 4.3106 |
| 0.0297 | 1.8828 | 900 | 0.0566 | 4.2708 |
| 0.0133 | 2.0921 | 1000 | 0.0557 | 4.1692 |
Framework versions
- Transformers 4.42.0.dev0
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1