metadata
license: apache-2.0
base_model: openai/whisper-medium
tags:
- generated_from_trainer
datasets:
- audiofolder
metrics:
- wer
model-index:
- name: whisper-medium-finetuned
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: audiofolder
type: audiofolder
config: default
split: test
args: default
metrics:
- name: Wer
type: wer
value: 59.45945945945946
whisper-medium-finetuned
This model is a fine-tuned version of openai/whisper-medium on the audiofolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.6522
- Wer Ortho: 59.4595
- Wer: 59.4595
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant_with_warmup
- lr_scheduler_warmup_steps: 10
- training_steps: 100
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer Ortho | Wer |
---|---|---|---|---|---|
0.4001 | 16.6667 | 50 | 0.7105 | 67.5676 | 67.5676 |
0.0001 | 33.3333 | 100 | 0.6522 | 59.4595 | 59.4595 |
Framework versions
- Transformers 4.41.1
- Pytorch 2.3.0+cpu
- Datasets 2.19.1
- Tokenizers 0.19.1