bert_sst2_padding80model

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: 0.7194
• Accuracy: 0.9226

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
No log	1.0	433	0.2350	0.9121
0.3142	2.0	866	0.3585	0.9072
0.1538	3.0	1299	0.4781	0.9110
0.0656	4.0	1732	0.4383	0.9132
0.0366	5.0	2165	0.5503	0.9160
0.0215	6.0	2598	0.7577	0.9044
0.0145	7.0	3031	0.5950	0.9187
0.0145	8.0	3464	0.6371	0.9132
0.012	9.0	3897	0.7315	0.9143
0.0131	10.0	4330	0.6525	0.9116
0.0131	11.0	4763	0.7516	0.9061
0.0082	12.0	5196	0.6153	0.9220
0.0082	13.0	5629	0.6638	0.9231
0.0057	14.0	6062	0.6706	0.9182
0.0057	15.0	6495	0.6851	0.9209
0.0034	16.0	6928	0.7433	0.9116
0.0038	17.0	7361	0.6961	0.9237
0.0026	18.0	7794	0.8033	0.9088
0.0026	19.0	8227	0.6943	0.9226
0.0021	20.0	8660	0.7194	0.9226

Framework versions

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