bert_sst5_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: 4.0787
• Accuracy: 0.5380

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
1.2698	1.0	534	1.2515	0.4267
0.9681	2.0	1068	1.1100	0.5213
0.7474	3.0	1602	1.1959	0.5394
0.5709	4.0	2136	1.3821	0.5371
0.4057	5.0	2670	1.6986	0.5131
0.3003	6.0	3204	1.9555	0.5290
0.2362	7.0	3738	2.2074	0.5357
0.1666	8.0	4272	2.5575	0.5276
0.1467	9.0	4806	2.8881	0.5353
0.1243	10.0	5340	3.1053	0.5276
0.0968	11.0	5874	3.2865	0.5317
0.075	12.0	6408	3.5669	0.5335
0.0607	13.0	6942	3.5765	0.5362
0.0509	14.0	7476	3.6150	0.5403
0.0261	15.0	8010	3.8456	0.5294
0.0216	16.0	8544	3.9090	0.5376
0.0152	17.0	9078	3.9822	0.5339
0.0178	18.0	9612	4.0606	0.5335
0.0155	19.0	10146	4.0775	0.5321
0.0148	20.0	10680	4.0787	0.5380

Framework versions

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