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: 0.7207
• 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: 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: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.0917	1.0	113	1.9773	0.48
1.3655	2.0	226	1.3419	0.62
0.9901	3.0	339	0.9640	0.75
0.8565	4.0	452	0.7732	0.81
0.6259	5.0	565	0.7502	0.8
0.4507	6.0	678	0.6888	0.81
0.4018	7.0	791	0.7404	0.8
0.1275	8.0	904	0.6718	0.83
0.1077	9.0	1017	0.6175	0.86
0.028	10.0	1130	0.6317	0.86
0.0867	11.0	1243	0.6053	0.88
0.0149	12.0	1356	0.7164	0.85
0.0108	13.0	1469	0.7224	0.85
0.0101	14.0	1582	0.7101	0.84
0.0096	15.0	1695	0.7207	0.85

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.0
• Tokenizers 0.15.0