metadata
license: apache-2.0
base_model: ntu-spml/distilhubert
tags:
- generated_from_trainer
datasets:
- marsyas/gtzan
metrics:
- accuracy
model-index:
- name: distilhubert-finetuned-gtzan-bs-4
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: 0.86
distilhubert-finetuned-gtzan-bs-4
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: 0.6851
- Accuracy: 0.86
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: 4
- eval_batch_size: 4
- 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: 15
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.8665 | 1.0 | 225 | 1.7962 | 0.45 |
| 1.1445 | 2.0 | 450 | 1.1084 | 0.68 |
| 0.9474 | 3.0 | 675 | 0.8338 | 0.73 |
| 0.8286 | 4.0 | 900 | 0.7530 | 0.76 |
| 0.2336 | 5.0 | 1125 | 0.5369 | 0.84 |
| 0.2092 | 6.0 | 1350 | 0.5608 | 0.86 |
| 0.2092 | 7.0 | 1575 | 0.5390 | 0.88 |
| 0.04 | 8.0 | 1800 | 0.5567 | 0.88 |
| 0.0046 | 9.0 | 2025 | 0.5736 | 0.86 |
| 0.0029 | 10.0 | 2250 | 0.6236 | 0.86 |
| 0.0035 | 11.0 | 2475 | 0.8139 | 0.85 |
| 0.0018 | 12.0 | 2700 | 0.5752 | 0.9 |
| 0.0016 | 13.0 | 2925 | 0.6745 | 0.85 |
| 0.0016 | 14.0 | 3150 | 0.6959 | 0.85 |
| 0.0014 | 15.0 | 3375 | 0.6851 | 0.86 |
Framework versions
- Transformers 4.32.0.dev0
- Pytorch 2.0.1+cu118
- Datasets 2.14.4
- Tokenizers 0.13.3