---
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: 0.81
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# distilhubert-finetuned-gtzan

This model is a fine-tuned version of [ntu-spml/distilhubert](https://huggingface.co/ntu-spml/distilhubert) on the GTZAN dataset.
It achieves the following results on the evaluation set:
- Loss: 1.0178
- Accuracy: 0.81

## 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: 16
- eval_batch_size: 16
- 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: 30
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 2.2598        | 1.0   | 57   | 2.2140          | 0.34     |
| 1.8981        | 2.0   | 114  | 1.8262          | 0.56     |
| 1.4487        | 3.0   | 171  | 1.4402          | 0.64     |
| 1.1792        | 4.0   | 228  | 1.1520          | 0.69     |
| 0.9231        | 5.0   | 285  | 0.9415          | 0.75     |
| 0.7141        | 6.0   | 342  | 0.8904          | 0.73     |
| 0.5477        | 7.0   | 399  | 0.7395          | 0.78     |
| 0.3968        | 8.0   | 456  | 0.6359          | 0.81     |
| 0.4259        | 9.0   | 513  | 0.6345          | 0.8      |
| 0.2474        | 10.0  | 570  | 0.6333          | 0.8      |
| 0.1379        | 11.0  | 627  | 0.5374          | 0.83     |
| 0.0781        | 12.0  | 684  | 0.6484          | 0.84     |
| 0.0337        | 13.0  | 741  | 0.7072          | 0.84     |
| 0.0211        | 14.0  | 798  | 0.7023          | 0.83     |
| 0.0135        | 15.0  | 855  | 0.8199          | 0.83     |
| 0.0097        | 16.0  | 912  | 0.8009          | 0.83     |
| 0.065         | 17.0  | 969  | 0.8992          | 0.81     |
| 0.0067        | 18.0  | 1026 | 0.8628          | 0.82     |
| 0.0118        | 19.0  | 1083 | 0.6922          | 0.85     |
| 0.0052        | 20.0  | 1140 | 0.8001          | 0.84     |
| 0.077         | 21.0  | 1197 | 0.8324          | 0.82     |
| 0.0043        | 22.0  | 1254 | 0.9468          | 0.8      |
| 0.0039        | 23.0  | 1311 | 0.8866          | 0.8      |
| 0.0696        | 24.0  | 1368 | 0.9424          | 0.82     |
| 0.0037        | 25.0  | 1425 | 0.7855          | 0.81     |
| 0.0631        | 26.0  | 1482 | 0.7659          | 0.82     |
| 0.0592        | 27.0  | 1539 | 0.8605          | 0.83     |
| 0.0034        | 28.0  | 1596 | 0.9266          | 0.82     |
| 0.0032        | 29.0  | 1653 | 0.9831          | 0.82     |
| 0.0032        | 30.0  | 1710 | 1.0178          | 0.81     |


### Framework versions

- Transformers 4.36.2
- Pytorch 2.1.0+cu121
- Datasets 2.15.0
- Tokenizers 0.15.0