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: 0.9006
• Accuracy: 0.8

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.1981	1.0	57	2.1804	0.37
1.7932	2.0	114	1.7160	0.62
1.3257	3.0	171	1.2539	0.67
1.1239	4.0	228	1.1187	0.68
0.7457	5.0	285	0.9367	0.73
0.6922	6.0	342	0.7564	0.81
0.5718	7.0	399	0.8179	0.78
0.3729	8.0	456	0.7299	0.79
0.2667	9.0	513	0.6415	0.82
0.4672	10.0	570	0.8068	0.78
0.1392	11.0	627	0.7228	0.81
0.1069	12.0	684	0.7787	0.79
0.0659	13.0	741	0.7720	0.8
0.0291	14.0	798	0.7609	0.79
0.0263	15.0	855	0.8363	0.8
0.0177	16.0	912	0.8796	0.78
0.0166	17.0	969	0.8844	0.79
0.0139	18.0	1026	0.8909	0.8
0.0132	19.0	1083	0.9017	0.8
0.0131	20.0	1140	0.9006	0.8

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.1+cu118
• Datasets 2.13.0
• Tokenizers 0.13.3