bert_sst2_padding10model

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: 0.6883
• Accuracy: 0.9259

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	0.2373	0.9061
0.3327	2.0	866	0.3059	0.9083
0.1608	3.0	1299	0.4178	0.9110
0.073	4.0	1732	0.5039	0.9116
0.0319	5.0	2165	0.5426	0.9165
0.0184	6.0	2598	0.6055	0.9149
0.01	7.0	3031	0.6485	0.9198
0.01	8.0	3464	0.6796	0.9176
0.0132	9.0	3897	0.6326	0.9204
0.016	10.0	4330	0.6441	0.9198
0.0081	11.0	4763	0.7473	0.9176
0.0067	12.0	5196	0.7182	0.9198
0.0056	13.0	5629	0.7216	0.9182
0.007	14.0	6062	0.6818	0.9231
0.007	15.0	6495	0.6803	0.9226
0.0045	16.0	6928	0.6566	0.9237
0.0015	17.0	7361	0.6589	0.9286
0.0029	18.0	7794	0.7071	0.9259
0.0016	19.0	8227	0.6871	0.9253
0.0	20.0	8660	0.6883	0.9259

Framework versions

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