N_distilbert_sst5_padding30model

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: 4.0219
• Accuracy: 0.5063

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
1.3201	1.0	534	1.2810	0.4303
1.0406	2.0	1068	1.1007	0.5136
0.8464	3.0	1602	1.1399	0.5172
0.6868	4.0	2136	1.2985	0.5140
0.5199	5.0	2670	1.5713	0.4973
0.3987	6.0	3204	1.7627	0.5
0.3021	7.0	3738	1.9491	0.5045
0.2176	8.0	4272	2.1652	0.5032
0.1867	9.0	4806	2.4295	0.5068
0.1462	10.0	5340	2.6915	0.4995
0.1246	11.0	5874	3.0150	0.4977
0.1072	12.0	6408	3.1994	0.5063
0.0847	13.0	6942	3.4184	0.5036
0.0694	14.0	7476	3.5757	0.5054
0.0429	15.0	8010	3.8429	0.5041
0.0353	16.0	8544	3.9057	0.5077
0.0318	17.0	9078	3.9452	0.5113
0.0203	18.0	9612	3.9648	0.5095
0.0239	19.0	10146	4.0397	0.5063
0.0231	20.0	10680	4.0219	0.5063

Framework versions

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