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: 0.86

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: 0.8540
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: 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

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.2594	1.0	57	2.2216	0.37
1.941	2.0	114	1.8715	0.59
1.4613	3.0	171	1.4244	0.65
1.2449	4.0	228	1.1359	0.71
0.8682	5.0	285	0.9472	0.74
0.6808	6.0	342	0.7817	0.78
0.4759	7.0	399	0.7428	0.74
0.3316	8.0	456	0.6441	0.78
0.2228	9.0	513	0.5838	0.83
0.1367	10.0	570	0.5843	0.86
0.0921	11.0	627	0.5745	0.86
0.0462	12.0	684	0.7029	0.83
0.0513	13.0	741	0.7116	0.86
0.0151	14.0	798	0.7017	0.86
0.0113	15.0	855	0.7439	0.85
0.0572	16.0	912	0.7691	0.84
0.0073	17.0	969	0.7918	0.84
0.0076	18.0	1026	0.8202	0.84
0.0053	19.0	1083	0.8238	0.86
0.0547	20.0	1140	0.8147	0.86
0.0045	21.0	1197	0.8201	0.86
0.004	22.0	1254	0.8282	0.83
0.0038	23.0	1311	0.8387	0.86
0.0035	24.0	1368	0.8398	0.86
0.0033	25.0	1425	0.8403	0.86
0.0031	26.0	1482	0.8464	0.86
0.0032	27.0	1539	0.8456	0.86
0.0031	28.0	1596	0.8505	0.86
0.0031	29.0	1653	0.8517	0.86
0.003	30.0	1710	0.8540	0.86

Framework versions

Transformers 4.31.0
Pytorch 2.0.1+cu118
Datasets 2.14.0
Tokenizers 0.13.3