metadata
language:
- en
license: apache-2.0
base_model: openai/whisper-small.en
tags:
- nyansapo_ai-asr-leaderboard
- generated_from_trainer
datasets:
- NyansapoAI/azure-dataset
metrics:
- wer
model-index:
- name: whisper-small.en
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Azure-dataset
type: NyansapoAI/azure-dataset
config: default
split: test
args: 'split: test'
metrics:
- name: Wer
type: wer
value: 12.02020202020202
whisper-small.en
This model is a fine-tuned version of openai/whisper-small.en on the Azure-dataset dataset. It achieves the following results on the evaluation set:
- Loss: 0.0191
- Wer: 12.0202
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: 2500
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.1119 | 3.11 | 500 | 0.0303 | 27.8788 |
| 0.0514 | 6.21 | 1000 | 0.0196 | 16.4646 |
| 0.0489 | 9.32 | 1500 | 0.0193 | 9.8990 |
| 0.0478 | 12.42 | 2000 | 0.0193 | 11.9192 |
| 0.0493 | 15.53 | 2500 | 0.0191 | 12.0202 |
Framework versions
- Transformers 4.33.0.dev0
- Pytorch 2.0.1
- Datasets 2.14.4
- Tokenizers 0.13.3