bert_twitterfin_padding50model

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: 1.0501
• Accuracy: 0.8844

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
0.7228	1.0	597	0.4356	0.8413
0.4132	2.0	1194	0.3576	0.8765
0.2844	3.0	1791	0.4000	0.8744
0.1999	4.0	2388	0.5008	0.8848
0.1462	5.0	2985	0.6123	0.8740
0.0623	6.0	3582	0.6834	0.8786
0.0647	7.0	4179	0.8103	0.8752
0.0345	8.0	4776	0.7865	0.8857
0.0383	9.0	5373	0.8424	0.8756
0.0275	10.0	5970	0.8217	0.8890
0.018	11.0	6567	0.8443	0.8823
0.0134	12.0	7164	0.9511	0.8760
0.0155	13.0	7761	0.9635	0.8853
0.0097	14.0	8358	0.9534	0.8836
0.0088	15.0	8955	0.9661	0.8807
0.0056	16.0	9552	0.9900	0.8819
0.0083	17.0	10149	1.0253	0.8836
0.0021	18.0	10746	1.0431	0.8827
0.0018	19.0	11343	1.0460	0.8836
0.0051	20.0	11940	1.0501	0.8844

Framework versions

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