metadata
license: apache-2.0
base_model: ntu-spml/distilhubert
tags:
- generated_from_trainer
datasets:
- marsyas/gtzan
metrics:
- accuracy
model-index:
- name: distilhubert-finetuned-VD
results:
- task:
name: Audio Classification
type: audio-classification
dataset:
name: GTZAN
type: marsyas/gtzan
config: default
split: train
args: default
metrics:
- name: Accuracy
type: accuracy
value: 0.8349877949552482
distilhubert-finetuned-VD
This model is a fine-tuned version of ntu-spml/distilhubert on the GTZAN dataset. It achieves the following results on the evaluation set:
- Loss: 0.4702
- Accuracy: 0.8350
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-05
- train_batch_size: 8
- 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_ratio: 0.1
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.555 | 1.0 | 167 | 0.4702 | 0.8350 |
| 0.3965 | 2.0 | 334 | 0.4398 | 0.7570 |
| 0.4106 | 3.0 | 501 | 0.7742 | 0.6713 |
| 0.4372 | 4.0 | 668 | 0.9340 | 0.6827 |
| 0.2087 | 5.0 | 835 | 1.0133 | 0.7574 |
| 0.124 | 6.0 | 1002 | 1.1049 | 0.7437 |
| 0.0509 | 7.0 | 1169 | 1.2264 | 0.7590 |
| 0.0016 | 8.0 | 1336 | 1.2315 | 0.7845 |
| 0.0064 | 9.0 | 1503 | 1.3620 | 0.7762 |
| 0.0006 | 10.0 | 1670 | 1.3149 | 0.8039 |
| 0.0007 | 11.0 | 1837 | 1.2818 | 0.8116 |
| 0.0003 | 12.0 | 2004 | 1.2635 | 0.8298 |
| 0.0003 | 13.0 | 2171 | 1.3287 | 0.8225 |
| 0.0002 | 14.0 | 2338 | 1.3200 | 0.8295 |
| 0.0001 | 15.0 | 2505 | 1.4146 | 0.8226 |
| 0.0001 | 16.0 | 2672 | 1.4359 | 0.8221 |
| 0.0001 | 17.0 | 2839 | 1.4443 | 0.8233 |
| 0.0001 | 18.0 | 3006 | 1.5031 | 0.8184 |
| 0.0001 | 19.0 | 3173 | 1.5111 | 0.8182 |
| 0.0001 | 20.0 | 3340 | 1.5145 | 0.8182 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.2.2+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2