bert_twitterfin_padding60model

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: 1.0129
• Accuracy: 0.8861

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6595	1.0	597	0.3889	0.8610
0.3378	2.0	1194	0.3376	0.8740
0.2355	3.0	1791	0.4061	0.8844
0.1452	4.0	2388	0.6084	0.8656
0.1028	5.0	2985	0.6311	0.8773
0.0403	6.0	3582	0.6991	0.8840
0.0396	7.0	4179	0.8008	0.8827
0.0258	8.0	4776	0.7997	0.8819
0.0229	9.0	5373	0.9039	0.8731
0.0172	10.0	5970	0.8777	0.8857
0.0128	11.0	6567	0.8792	0.8836
0.0111	12.0	7164	0.9188	0.8827
0.0097	13.0	7761	0.9763	0.8823
0.0096	14.0	8358	0.9605	0.8823
0.0049	15.0	8955	0.9565	0.8840
0.0058	16.0	9552	0.9900	0.8802
0.0032	17.0	10149	0.9681	0.8882
0.0004	18.0	10746	0.9960	0.8894
0.0021	19.0	11343	1.0260	0.8840
0.0023	20.0	11940	1.0129	0.8861

Framework versions

• Transformers 4.33.2
• Pytorch 2.0.1+cu117
• Datasets 2.14.5
• Tokenizers 0.13.3