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: 0.9734
• Model Preparation Time: 0.0068
• Accuracy: 0.8602

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: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 0.1
• num_epochs: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Model Preparation Time	Accuracy
1.2068	1.0	733	1.1166	0.0068	0.6774
0.8130	2.0	1466	0.6470	0.0068	0.7957
0.3782	3.0	2199	0.5003	0.0068	0.8280
0.2297	4.0	2932	0.5445	0.0068	0.8495
0.1655	5.0	3665	0.6829	0.0068	0.8602
0.0025	6.0	4398	0.7646	0.0068	0.8280
0.1201	7.0	5131	0.8735	0.0068	0.8495
0.0508	8.0	5864	0.8754	0.0068	0.8817
0.0434	9.0	6597	0.9917	0.0068	0.8710
0.0015	10.0	7330	1.2593	0.0068	0.8602
0.0000	11.0	8063	0.8777	0.0068	0.8817
0.0001	12.0	8796	1.2714	0.0068	0.8387
0.0000	13.0	9529	0.9322	0.0068	0.8495
0.0000	14.0	10262	0.9663	0.0068	0.8495
0.0000	15.0	10995	0.9734	0.0068	0.8602

Framework versions

• Transformers 5.12.1
• Pytorch 2.11.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2