bert-base-uncased-finetuned-scientific-eval

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

• Loss: 0.1920
• Precision: 0.6240
• Recall: 0.7066
• F1: 0.6627
• Accuracy: 0.9502

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	Precision	Recall	F1	Accuracy
No log	1.0	71	0.3069	0.4263	0.3375	0.3768	0.9212
No log	2.0	142	0.2241	0.4883	0.5899	0.5343	0.9359
No log	3.0	213	0.1962	0.5745	0.6688	0.6181	0.9441
No log	4.0	284	0.1920	0.6240	0.7066	0.6627	0.9502
No log	5.0	355	0.2107	0.5694	0.7634	0.6523	0.9465
No log	6.0	426	0.2070	0.6286	0.7634	0.6895	0.9514
No log	7.0	497	0.2129	0.6193	0.7697	0.6864	0.9499
0.1579	8.0	568	0.2269	0.6496	0.7603	0.7006	0.9529
0.1579	9.0	639	0.2274	0.6366	0.7571	0.6916	0.9519
0.1579	10.0	710	0.2285	0.6486	0.7571	0.6987	0.9522

Framework versions

• Transformers 4.27.2
• Pytorch 2.0.1+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3