metadata

license: apache-2.0
base_model: ntu-spml/distilhubert
tags:
  - generated_from_trainer
datasets:
  - marsyas/gtzan-v2
metrics:
  - accuracy
model-index:
  - name: distilhubert-finetuned-gtzan
    results:
      - task:
          name: Audio Classification
          type: audio-classification
        dataset:
          name: GTZAN
          type: marsyas/gtzan-v2
          config: all
          split: train
          args: all
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.82

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: 0.7819
Accuracy: 0.82

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: 3e-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: 0.3
num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.286	1.0	113	2.2792	0.26
2.1863	2.0	226	2.1408	0.34
1.9386	3.0	339	1.8744	0.48
1.6908	4.0	452	1.6502	0.57
1.5259	5.0	565	1.4149	0.72
1.1279	6.0	678	1.2700	0.62
1.2204	7.0	791	0.9902	0.75
0.861	8.0	904	0.8020	0.8
0.8153	9.0	1017	0.7291	0.8
0.3983	10.0	1130	0.7304	0.8
0.2209	11.0	1243	0.6960	0.79
0.2523	12.0	1356	0.5783	0.83
0.1267	13.0	1469	0.5613	0.83
0.0468	14.0	1582	0.7976	0.8
0.025	15.0	1695	0.8478	0.81
0.0158	16.0	1808	0.7448	0.8
0.0706	17.0	1921	0.7183	0.83
0.0096	18.0	2034	0.7532	0.82
0.0076	19.0	2147	0.7907	0.81
0.0354	20.0	2260	0.7120	0.83
0.0063	21.0	2373	0.7525	0.83
0.0055	22.0	2486	0.7647	0.82
0.0049	23.0	2599	0.7945	0.82
0.0048	24.0	2712	0.7982	0.82
0.0321	25.0	2825	0.7819	0.82

Framework versions

Transformers 4.34.0
Pytorch 2.0.1+cu118
Datasets 2.14.5
Tokenizers 0.14.1