metadata
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- PolyAI/minds14
metrics:
- wer
model-index:
- name: whisper-tiny-finetuned-minds14-en-v2
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: PolyAI/minds14
type: PolyAI/minds14
config: en-US
split: train
args: en-US
metrics:
- name: Wer
type: wer
value: 0.34297520661157027
whisper-tiny-finetuned-minds14-en-v2
This model is a fine-tuned version of openai/whisper-tiny on the PolyAI/minds14 dataset. It achieves the following results on the evaluation set:
- Loss: 0.6386
- Wer Ortho: 0.3461
- Wer: 0.3430
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: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant_with_warmup
- lr_scheduler_warmup_steps: 50
- training_steps: 500
Training results
Training Loss | Epoch | Step | Validation Loss | Wer Ortho | Wer |
---|---|---|---|---|---|
1.7971 | 1.79 | 50 | 0.7476 | 0.4411 | 0.4008 |
0.3112 | 3.57 | 100 | 0.4953 | 0.3560 | 0.3453 |
0.1214 | 5.36 | 150 | 0.5073 | 0.3763 | 0.3589 |
0.0359 | 7.14 | 200 | 0.5244 | 0.3442 | 0.3347 |
0.011 | 8.93 | 250 | 0.5569 | 0.3510 | 0.3371 |
0.0038 | 10.71 | 300 | 0.5903 | 0.3393 | 0.3329 |
0.0019 | 12.5 | 350 | 0.6068 | 0.3405 | 0.3353 |
0.0012 | 14.29 | 400 | 0.6175 | 0.3418 | 0.3377 |
0.0012 | 16.07 | 450 | 0.6253 | 0.3362 | 0.3329 |
0.0013 | 17.86 | 500 | 0.6386 | 0.3461 | 0.3430 |
Framework versions
- Transformers 4.30.2
- Pytorch 2.0.1+cu117
- Datasets 2.13.1
- Tokenizers 0.13.3