distilhubert-finetuned-gtzan-bs-8-fp16-false

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: 0.7162
• Accuracy: 0.84

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

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.0985	1.0	113	1.9610	0.47
1.4194	2.0	226	1.3265	0.69
1.0868	3.0	339	0.9856	0.72
0.9134	4.0	452	0.8709	0.74
0.6349	5.0	565	0.7497	0.79
0.3755	6.0	678	0.7343	0.78
0.4007	7.0	791	0.5346	0.84
0.1607	8.0	904	0.5604	0.86
0.1802	9.0	1017	0.5005	0.89
0.0319	10.0	1130	0.6562	0.84
0.0158	11.0	1243	0.6639	0.84
0.1126	12.0	1356	0.6965	0.85
0.0095	13.0	1469	0.6919	0.84
0.0083	14.0	1582	0.7089	0.85
0.0088	15.0	1695	0.7162	0.84

Framework versions

• Transformers 4.32.0.dev0
• Pytorch 2.0.1+cu118
• Datasets 2.14.4
• Tokenizers 0.13.3