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
results:
- task:
name: Audio Classification
type: audio-classification
dataset:
name: GTZAN
type: marsyas/gtzan
config: all
split: train
args: all
metrics:
- name: Accuracy
type: accuracy
value: 0.85
DistilHuBERT-finetuned-gtzan
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: 0.8372
- Accuracy: 0.85
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: 6e-05
- train_batch_size: 7
- eval_batch_size: 7
- 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 |
|---|---|---|---|---|
| 1.9696 | 1.0 | 129 | 1.8571 | 0.55 |
| 1.4269 | 2.0 | 258 | 1.2394 | 0.61 |
| 1.0166 | 3.0 | 387 | 1.0173 | 0.74 |
| 0.7446 | 4.0 | 516 | 0.8103 | 0.75 |
| 0.4953 | 5.0 | 645 | 0.7800 | 0.77 |
| 0.3973 | 6.0 | 774 | 0.7359 | 0.81 |
| 0.2831 | 7.0 | 903 | 0.6434 | 0.84 |
| 0.2147 | 8.0 | 1032 | 0.6592 | 0.84 |
| 0.1287 | 9.0 | 1161 | 0.6988 | 0.85 |
| 0.014 | 10.0 | 1290 | 0.7569 | 0.83 |
| 0.0073 | 11.0 | 1419 | 0.8282 | 0.84 |
| 0.0049 | 12.0 | 1548 | 0.8531 | 0.84 |
| 0.0053 | 13.0 | 1677 | 0.8584 | 0.84 |
| 0.0044 | 14.0 | 1806 | 0.8707 | 0.84 |
| 0.0038 | 15.0 | 1935 | 0.8372 | 0.85 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.0.1+cu118
- Datasets 2.14.4
- Tokenizers 0.13.3