bert

This model is a fine-tuned version of bert-base-cased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3752
• Precision: 0.5495
• Recall: 0.5949
• F1: 0.5713
• Accuracy: 0.9455

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: 0.0001
• train_batch_size: 32
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.99) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	151	0.1826	0.4095	0.4084	0.4089	0.9362
No log	2.0	302	0.1684	0.4941	0.5303	0.5116	0.9442
No log	3.0	453	0.2528	0.5197	0.4477	0.4810	0.9398
0.1001	4.0	604	0.2100	0.5182	0.5583	0.5375	0.9439
0.1001	5.0	755	0.2556	0.5207	0.4783	0.4986	0.9419
0.1001	6.0	906	0.2908	0.4132	0.4204	0.4168	0.9365
0.0205	7.0	1057	0.3046	0.5	0.6236	0.5550	0.9435
0.0205	8.0	1208	0.3057	0.5324	0.5750	0.5529	0.9458
0.0205	9.0	1359	0.3122	0.5626	0.5776	0.5700	0.9469
0.0082	10.0	1510	0.3673	0.5733	0.5263	0.5488	0.9441
0.0082	11.0	1661	0.3432	0.5482	0.5270	0.5374	0.9455
0.0082	12.0	1812	0.3305	0.5590	0.5716	0.5652	0.9445
0.0082	13.0	1963	0.3293	0.5434	0.6009	0.5707	0.9431
0.005	14.0	2114	0.4080	0.5627	0.5803	0.5713	0.9451
0.005	15.0	2265	0.3752	0.5495	0.5949	0.5713	0.9455
0.005	16.0	2416	0.4140	0.5823	0.5470	0.5641	0.9455
0.002	17.0	2567	0.4308	0.5555	0.5670	0.5612	0.9438
0.002	18.0	2718	0.4389	0.5594	0.5676	0.5635	0.9436
0.002	19.0	2869	0.4463	0.5609	0.5676	0.5642	0.9444
0.0007	20.0	3020	0.4512	0.5648	0.5636	0.5642	0.9448

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu117
• Datasets 2.14.2
• Tokenizers 0.13.3