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:

• Accuracy: 0.87
• Loss: 0.9175

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: 17

Training results

Training Loss	Epoch	Step	Accuracy	Validation Loss
2.2295	1.0	113	0.4	2.1501
1.7373	2.0	226	0.6	1.6194
1.3497	3.0	339	0.72	1.1717
1.0135	4.0	452	0.71	1.0361
0.6951	5.0	565	0.77	0.7724
0.4279	6.0	678	0.76	0.7731
0.5178	7.0	791	0.82	0.6048
0.141	8.0	904	0.79	0.7486
0.2459	9.0	1017	0.85	0.6326
0.0331	10.0	1130	0.82	0.8706
0.0214	11.0	1243	0.81	1.0099
0.0744	12.0	1356	0.8	1.0210
0.0043	13.0	1469	0.82	0.9894
0.0032	14.0	1582	0.82	0.9803
0.0025	15.0	1695	0.83	1.0476
0.0021	16.0	1808	0.82	1.0483
0.0183	17.0	1921	0.87	0.9175

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1
• Datasets 2.14.0
• Tokenizers 0.13.3