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: 0.5120
• Accuracy: 0.86

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: 4e-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
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.2584	1.0	57	2.2062	0.35
1.8611	2.0	114	1.7924	0.53
1.4492	3.0	171	1.3901	0.65
1.0971	4.0	228	1.1676	0.69
0.9848	5.0	285	0.9750	0.74
0.8434	6.0	342	0.8434	0.74
0.7321	7.0	399	0.7555	0.83
0.5364	8.0	456	0.6995	0.79
0.4557	9.0	513	0.6118	0.84
0.4166	10.0	570	0.5975	0.83
0.2729	11.0	627	0.5576	0.83
0.2491	12.0	684	0.5737	0.82
0.2211	13.0	741	0.5129	0.84
0.1243	14.0	798	0.5710	0.83
0.0904	15.0	855	0.5087	0.86
0.0773	16.0	912	0.5836	0.8
0.0598	17.0	969	0.4871	0.83
0.0551	18.0	1026	0.4865	0.84
0.0467	19.0	1083	0.5043	0.84
0.0364	20.0	1140	0.5120	0.86

Framework versions

• Transformers 4.44.0
• Pytorch 2.4.0
• Datasets 3.0.0
• Tokenizers 0.19.1