bert_sst5_padding60model

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: 4.0761
• Accuracy: 0.5276

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
1.4317	1.0	534	1.3598	0.4009
1.0579	2.0	1068	1.0778	0.5376
0.8255	3.0	1602	1.1321	0.5480
0.6407	4.0	2136	1.2839	0.5339
0.4487	5.0	2670	1.6216	0.5027
0.3316	6.0	3204	1.8433	0.5240
0.2432	7.0	3738	2.2189	0.5271
0.1816	8.0	4272	2.4454	0.5249
0.1485	9.0	4806	2.6673	0.5249
0.1304	10.0	5340	3.1345	0.5136
0.1048	11.0	5874	3.3383	0.5240
0.0778	12.0	6408	3.4594	0.5271
0.0593	13.0	6942	3.7650	0.5235
0.048	14.0	7476	3.7509	0.5222
0.0311	15.0	8010	3.8533	0.5281
0.0265	16.0	8544	3.9777	0.5244
0.0258	17.0	9078	4.0436	0.5281
0.0171	18.0	9612	4.0182	0.5290
0.0136	19.0	10146	4.0354	0.5308
0.0081	20.0	10680	4.0761	0.5276

Framework versions

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