N_distilbert_sst5_padding70model

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: 3.8170
• Accuracy: 0.5077

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
1.4765	1.0	534	1.2981	0.4113
1.1535	2.0	1068	1.1111	0.5176
0.9575	3.0	1602	1.1843	0.5014
0.7998	4.0	2136	1.2742	0.5100
0.6303	5.0	2670	1.4411	0.5018
0.4977	6.0	3204	1.5945	0.4919
0.3945	7.0	3738	1.8055	0.4977
0.2777	8.0	4272	1.9519	0.4914
0.243	9.0	4806	2.1185	0.5131
0.2011	10.0	5340	2.3648	0.4919
0.1743	11.0	5874	2.7565	0.4995
0.1366	12.0	6408	2.9976	0.4946
0.1207	13.0	6942	3.0904	0.4986
0.0946	14.0	7476	3.2822	0.5036
0.065	15.0	8010	3.4680	0.5036
0.0558	16.0	8544	3.6425	0.5023
0.055	17.0	9078	3.7495	0.4986
0.0474	18.0	9612	3.7677	0.5054
0.0413	19.0	10146	3.7968	0.5081
0.0314	20.0	10680	3.8170	0.5077

Framework versions

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