semantic-bert-imbalanced-dataset

This model is a fine-tuned version of distilbert-base-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 3.3940
• Accuracy: 0.5926

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
No log	1.0	367	0.8237	0.6017
0.7608	2.0	734	0.8750	0.5834
0.5139	3.0	1101	0.9829	0.5946
0.5139	4.0	1468	1.2406	0.5870
0.3119	5.0	1835	1.5168	0.5865
0.1501	6.0	2202	2.0461	0.5814
0.0981	7.0	2569	2.2575	0.5860
0.0981	8.0	2936	2.6289	0.5804
0.0647	9.0	3303	2.8079	0.5844
0.0436	10.0	3670	2.9811	0.5824
0.0271	11.0	4037	3.0669	0.5875
0.0271	12.0	4404	3.1201	0.5834
0.0153	13.0	4771	3.1980	0.5910
0.0189	14.0	5138	3.2317	0.5931
0.0158	15.0	5505	3.2659	0.5885
0.0158	16.0	5872	3.3860	0.5875
0.0122	17.0	6239	3.3302	0.5956
0.0107	18.0	6606	3.3341	0.5946
0.0107	19.0	6973	3.4042	0.5931
0.0078	20.0	7340	3.3940	0.5926

Framework versions

• Transformers 4.36.2
• Pytorch 2.3.0.dev20231224
• Datasets 2.16.0
• Tokenizers 0.15.0