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: 0.6547
• Accuracy: 0.87

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: 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: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.1837	1.0	113	2.0719	0.39
1.5697	2.0	226	1.5216	0.56
1.2341	3.0	339	1.0787	0.7
1.0559	4.0	452	0.9199	0.75
0.7576	5.0	565	0.8469	0.75
0.5683	6.0	678	0.7064	0.76
0.4652	7.0	791	0.5912	0.79
0.1398	8.0	904	0.5407	0.81
0.2012	9.0	1017	0.4836	0.87
0.0445	10.0	1130	0.4839	0.88
0.0145	11.0	1243	0.5439	0.88
0.0519	12.0	1356	0.6309	0.85
0.0068	13.0	1469	0.5626	0.88
0.0054	14.0	1582	0.5949	0.88
0.0054	15.0	1695	0.5889	0.88
0.0033	16.0	1808	0.6233	0.89
0.1396	17.0	1921	0.6340	0.88
0.0029	18.0	2034	0.6362	0.88
0.0026	19.0	2147	0.6441	0.88
0.0027	20.0	2260	0.6547	0.87

Framework versions

• Transformers 4.33.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.4
• Tokenizers 0.13.3