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: 0.6653
• Accuracy: 0.82

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
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Use 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: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.931	0.9956	112	1.8442	0.38
1.4533	2.0	225	1.4234	0.56
1.5759	2.9956	337	1.3121	0.58
0.9118	4.0	450	1.1423	0.68
0.9785	4.9956	562	0.9830	0.71
0.7014	6.0	675	0.8055	0.8
0.5983	6.9956	787	0.7071	0.76
0.3568	8.0	900	0.7417	0.77
0.4118	8.9956	1012	0.5920	0.83
0.4934	9.9556	1120	0.6653	0.82

Framework versions

• Transformers 4.46.3
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.20.3