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: 0.8042
• 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: 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
2.0168	1.0	113	2.0642	0.45
1.4374	2.0	226	1.4358	0.64
1.1551	3.0	339	0.9743	0.74
0.7756	4.0	452	0.7805	0.81
0.4436	5.0	565	0.6117	0.81
0.3047	6.0	678	0.7366	0.79
0.2288	7.0	791	0.5297	0.86
0.2728	8.0	904	0.5677	0.87
0.1072	9.0	1017	0.6887	0.86
0.137	10.0	1130	0.9238	0.8
0.021	11.0	1243	0.7738	0.84
0.007	12.0	1356	0.7002	0.86
0.0047	13.0	1469	0.7805	0.86
0.0039	14.0	1582	0.7624	0.85
0.0034	15.0	1695	0.7892	0.85
0.0031	16.0	1808	0.7806	0.85
0.0029	17.0	1921	0.8005	0.85
0.0028	18.0	2034	0.7942	0.85
0.0025	19.0	2147	0.8138	0.86
0.0025	20.0	2260	0.8042	0.86

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.1
• Datasets 2.13.1
• Tokenizers 0.13.3