metadata
library_name: transformers
license: apache-2.0
base_model: openai/whisper-small
tags:
- generated_from_trainer
datasets:
- balbus-classifier
metrics:
- accuracy
model-index:
- name: miosipof/whisper-small-ft-balbus-sep28k-v1.2
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: 0.8082105922908179
miosipof/whisper-small-ft-balbus-sep28k-v1.2
This model is a fine-tuned version of openai/whisper-small on the Apple dataset dataset. It achieves the following results on the evaluation set:
- Loss: 0.1107
- Accuracy: 0.8082
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: 5e-06
- train_batch_size: 32
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.5
- training_steps: 500
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.1709 | 0.1253 | 50 | 0.1690 | 0.5650 |
| 0.1669 | 0.2506 | 100 | 0.1630 | 0.6369 |
| 0.1542 | 0.3759 | 150 | 0.1439 | 0.7131 |
| 0.1283 | 0.5013 | 200 | 0.1214 | 0.7802 |
| 0.1158 | 0.6266 | 250 | 0.1171 | 0.7935 |
| 0.1059 | 0.7519 | 300 | 0.1131 | 0.7985 |
| 0.1142 | 0.8772 | 350 | 0.1102 | 0.8081 |
| 0.104 | 1.0025 | 400 | 0.1112 | 0.8068 |
| 0.0924 | 1.1278 | 450 | 0.1114 | 0.8087 |
| 0.0959 | 1.2531 | 500 | 0.1107 | 0.8082 |
Framework versions
- Transformers 4.45.2
- Pytorch 2.2.0
- Datasets 3.2.0
- Tokenizers 0.20.3