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: 0.8818
• 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: 17

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.5851	1.0	113	1.7243	0.5
1.2937	2.0	226	1.2310	0.68
0.9718	3.0	339	0.8918	0.76
0.6613	4.0	452	0.6837	0.81
0.3693	5.0	565	0.6250	0.82
0.2991	6.0	678	0.5740	0.82
0.1381	7.0	791	0.5874	0.83
0.2047	8.0	904	0.5824	0.86
0.1192	9.0	1017	0.7106	0.83
0.0652	10.0	1130	0.6576	0.87
0.0105	11.0	1243	0.8236	0.84
0.0074	12.0	1356	0.7874	0.85
0.0064	13.0	1469	0.9066	0.84
0.0041	14.0	1582	0.8426	0.85
0.0038	15.0	1695	0.8676	0.84
0.0039	16.0	1808	0.8820	0.85
0.0036	17.0	1921	0.8818	0.85

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.1+cu118
• Datasets 2.13.1
• Tokenizers 0.13.3