distilbert_sst5_padding60model

This model is a fine-tuned version of distilbert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 4.1071
• Accuracy: 0.5005

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
1.2904	1.0	534	1.2735	0.4204
1.0348	2.0	1068	1.1222	0.5217
0.8515	3.0	1602	1.1982	0.5172
0.6867	4.0	2136	1.3441	0.5018
0.5308	5.0	2670	1.5426	0.4964
0.396	6.0	3204	1.7466	0.5032
0.3113	7.0	3738	1.9661	0.4959
0.2237	8.0	4272	2.3050	0.4882
0.1863	9.0	4806	2.4120	0.5068
0.1458	10.0	5340	2.7065	0.4950
0.1202	11.0	5874	3.0306	0.4900
0.1124	12.0	6408	3.1912	0.5041
0.0838	13.0	6942	3.3632	0.5050
0.0798	14.0	7476	3.6172	0.4968
0.0475	15.0	8010	3.7530	0.4973
0.0436	16.0	8544	3.7669	0.5014
0.0302	17.0	9078	3.9727	0.5023
0.0223	18.0	9612	4.0368	0.5054
0.0236	19.0	10146	4.0607	0.5077
0.0247	20.0	10680	4.1071	0.5005

Framework versions

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