song-coherency-classifier

This model is a fine-tuned version of distilbert/distilroberta-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1299
• F1: [0.9763779527559054, 0.9757412398921832]

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	F1
No log	1.0	190	0.2172	[0.9460154241645244, 0.9421487603305784]
No log	2.0	380	0.2073	[0.9487179487179487, 0.9447513812154695]
0.2175	3.0	570	0.2125	[0.9487179487179487, 0.9447513812154695]
0.2175	4.0	760	0.1431	[0.9649595687331537, 0.9658792650918635]
0.2175	5.0	950	0.1467	[0.9711286089238844, 0.9703504043126685]
0.131	6.0	1140	0.1400	[0.9711286089238844, 0.9703504043126685]
0.131	7.0	1330	0.1238	[0.9736842105263158, 0.9731182795698924]
0.0835	8.0	1520	0.1317	[0.9736842105263158, 0.9731182795698924]
0.0835	9.0	1710	0.1390	[0.9738219895287958, 0.972972972972973]
0.0835	10.0	1900	0.1299	[0.9763779527559054, 0.9757412398921832]

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2