hubert-base-ls960-finetuned-gtzan

This model is a fine-tuned version of facebook/hubert-base-ls960 on the GTZAN dataset. It achieves the following results on the evaluation set:

• Loss: 0.9971
• 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: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 16
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.0659	1.0	225	1.9480	0.3
1.6674	2.0	450	1.4441	0.56
1.2155	3.0	675	1.1217	0.61
0.6299	4.0	900	1.1092	0.6
0.8145	5.0	1125	0.9608	0.71
0.2263	6.0	1350	0.7934	0.78
1.2492	7.0	1575	0.5946	0.83
0.2689	8.0	1800	1.0830	0.81
0.0514	9.0	2025	0.6099	0.88
0.044	10.0	2250	0.7864	0.85
0.067	11.0	2475	0.8360	0.84
0.1052	12.0	2700	0.7800	0.88
0.0058	13.0	2925	0.9246	0.87
0.2286	14.0	3150	0.9947	0.86
0.0041	15.0	3375	0.8841	0.86
0.0045	16.0	3600	0.9971	0.86

Framework versions

• Transformers 4.53.3
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.2