N_bert_sst5_padding100model

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: 4.0441
• Accuracy: 0.5249

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
1.3098	1.0	534	1.3057	0.4339
0.9919	2.0	1068	1.0863	0.5326
0.7654	3.0	1602	1.1988	0.5231
0.593	4.0	2136	1.3832	0.5244
0.4256	5.0	2670	1.6810	0.5154
0.313	6.0	3204	1.9396	0.5154
0.2169	7.0	3738	2.2859	0.5140
0.1745	8.0	4272	2.6011	0.5163
0.1521	9.0	4806	2.7484	0.5181
0.1278	10.0	5340	3.0932	0.5281
0.0993	11.0	5874	3.2683	0.5181
0.097	12.0	6408	3.4021	0.5217
0.0629	13.0	6942	3.7096	0.5258
0.0437	14.0	7476	3.7275	0.5235
0.0298	15.0	8010	3.7627	0.5290
0.0257	16.0	8544	3.8717	0.5276
0.0175	17.0	9078	3.9446	0.5213
0.0174	18.0	9612	3.9703	0.5226
0.0101	19.0	10146	4.0437	0.5222
0.007	20.0	10680	4.0441	0.5249

Framework versions

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