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: 0.9031
• Accuracy: 0.87

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
2.1931	1.0	113	2.0840	0.39
1.5734	2.0	226	1.4764	0.53
1.2619	3.0	339	1.1045	0.68
1.0427	4.0	452	1.0008	0.74
0.7065	5.0	565	0.7131	0.83
0.4206	6.0	678	0.6687	0.8
0.5466	7.0	791	0.5807	0.83
0.1232	8.0	904	0.6143	0.83
0.2593	9.0	1017	0.6080	0.89
0.0496	10.0	1130	0.7360	0.84
0.0127	11.0	1243	0.7648	0.85
0.0993	12.0	1356	0.8416	0.85
0.0068	13.0	1469	0.7966	0.85
0.0054	14.0	1582	0.8122	0.86
0.0044	15.0	1695	0.8788	0.87
0.0037	16.0	1808	0.8760	0.87
0.0892	17.0	1921	0.8911	0.87
0.0032	18.0	2034	0.9083	0.86
0.0029	19.0	2147	0.9172	0.86
0.0037	20.0	2260	0.9031	0.87

Framework versions

• Transformers 4.32.0.dev0
• Pytorch 2.0.1+cu118
• Datasets 2.14.4
• Tokenizers 0.13.3