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: 0.2387
• Accuracy: 0.9319

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: 0.001
• train_batch_size: 6
• eval_batch_size: 6
• 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: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.7644	1.0	167	1.7832	0.3554
1.2856	2.0	334	1.4226	0.4745
1.2123	3.0	501	1.0047	0.6737
0.6613	4.0	668	0.8091	0.6987
0.6442	5.0	835	0.6713	0.7858
0.7172	6.0	1002	0.5749	0.8238
0.5394	7.0	1169	0.5079	0.8408
0.3853	8.0	1336	0.4574	0.8539
0.5441	9.0	1503	0.3729	0.8869
0.5062	10.0	1670	0.3319	0.9009
0.3955	11.0	1837	0.3745	0.8849
0.3112	12.0	2004	0.2752	0.9289
0.2887	13.0	2171	0.2544	0.9289
0.2038	14.0	2338	0.2344	0.9329
0.2374	15.0	2505	0.2387	0.9319

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.1+cu121
• Datasets 2.15.0
• Tokenizers 0.15.0

Training procedure

Framework versions

• PEFT 0.6.2