bert_twitterfin_padding30model

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: 1.0811
• Accuracy: 0.8857

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
0.5721	1.0	597	0.3568	0.8735
0.3173	2.0	1194	0.3192	0.8848
0.2143	3.0	1791	0.4767	0.8760
0.1357	4.0	2388	0.6198	0.8832
0.1059	5.0	2985	0.6062	0.8811
0.05	6.0	3582	0.7149	0.8819
0.0381	7.0	4179	0.8061	0.8848
0.0309	8.0	4776	0.7961	0.8815
0.0315	9.0	5373	0.8086	0.8802
0.0214	10.0	5970	0.8231	0.8924
0.0178	11.0	6567	0.8589	0.8861
0.0127	12.0	7164	0.9441	0.8853
0.0129	13.0	7761	0.9523	0.8899
0.0102	14.0	8358	1.0047	0.8848
0.009	15.0	8955	1.0004	0.8882
0.0047	16.0	9552	1.0421	0.8848
0.0049	17.0	10149	1.0416	0.8865
0.0035	18.0	10746	1.0695	0.8869
0.0021	19.0	11343	1.0858	0.8844
0.003	20.0	11940	1.0811	0.8857

Framework versions

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