metadata
library_name: transformers
license: apache-2.0
base_model: openai/whisper-tiny
tags:
- generated_from_trainer
datasets:
- balbus-classifier
metrics:
- accuracy
model-index:
- name: miosipof/whisper-tiny-ft-balbus-sep28k-v1
results:
- task:
name: Audio Classification
type: audio-classification
dataset:
name: Apple dataset
type: balbus-classifier
config: default
split: train
args: default
metrics:
- name: Accuracy
type: accuracy
value: 0.7508617988091507
miosipof/whisper-tiny-ft-balbus-sep28k-v1
This model is a fine-tuned version of openai/whisper-tiny on the Apple dataset dataset. It achieves the following results on the evaluation set:
- Loss: 0.5255
- Accuracy: 0.7509
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: 4e-05
- train_batch_size: 64
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 128
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.5
- training_steps: 600
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.6685 | 0.5013 | 100 | 0.6430 | 0.6371 |
| 0.5858 | 1.0 | 200 | 0.5736 | 0.7068 |
| 0.5284 | 1.5013 | 300 | 0.5422 | 0.7333 |
| 0.5125 | 2.0 | 400 | 0.5359 | 0.7361 |
| 0.4163 | 2.5013 | 500 | 0.5517 | 0.7369 |
| 0.4113 | 3.0 | 600 | 0.5255 | 0.7509 |
Framework versions
- Transformers 4.48.0
- Pytorch 2.2.0
- Datasets 3.2.0
- Tokenizers 0.21.0