distilhubert-course-model2-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: 0.6319
• 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: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.0479	1.0	113	1.9649	0.57
1.2746	2.0	226	1.3397	0.62
0.9327	3.0	339	0.9767	0.72
0.7575	4.0	452	0.8140	0.77
0.5051	5.0	565	0.6947	0.8
0.4299	6.0	678	0.6564	0.8
0.2753	7.0	791	0.7915	0.74
0.2209	8.0	904	0.5574	0.81
0.2022	9.0	1017	0.6053	0.85
0.0333	10.0	1130	0.5527	0.88
0.1367	11.0	1243	0.5989	0.87
0.0141	12.0	1356	0.6271	0.86
0.0104	13.0	1469	0.6737	0.87
0.0093	14.0	1582	0.6163	0.86
0.0099	15.0	1695	0.6319	0.87

Framework versions

• Transformers 4.37.0.dev0
• Pytorch 2.1.0+cu121
• Datasets 2.15.0
• Tokenizers 0.15.0