distilbert-base-uncased_fold_4_binary

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: 1.2977
• F1: 0.8083

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: 25

Training results

Training Loss	Epoch	Step	Validation Loss	F1
No log	1.0	289	0.3701	0.7903
0.4005	2.0	578	0.3669	0.7994
0.4005	3.0	867	0.5038	0.7955
0.1945	4.0	1156	0.6353	0.8006
0.1945	5.0	1445	0.8974	0.7826
0.0909	6.0	1734	0.8533	0.7764
0.0389	7.0	2023	0.9969	0.7957
0.0389	8.0	2312	1.0356	0.7952
0.0231	9.0	2601	1.1538	0.7963
0.0231	10.0	2890	1.2011	0.7968
0.0051	11.0	3179	1.2329	0.7935
0.0051	12.0	3468	1.2829	0.8056
0.0066	13.0	3757	1.2946	0.7956
0.004	14.0	4046	1.2977	0.8083
0.004	15.0	4335	1.3970	0.7957
0.0007	16.0	4624	1.3361	0.7917
0.0007	17.0	4913	1.5782	0.7954
0.0107	18.0	5202	1.4641	0.7900
0.0107	19.0	5491	1.4490	0.7957
0.0058	20.0	5780	1.4607	0.7932
0.0016	21.0	6069	1.5048	0.7939
0.0016	22.0	6358	1.5219	0.7945
0.0027	23.0	6647	1.4783	0.7937
0.0027	24.0	6936	1.4715	0.7981
0.0004	25.0	7225	1.4989	0.7900

Framework versions

• Transformers 4.21.0
• Pytorch 1.12.0+cu113
• Datasets 2.4.0
• Tokenizers 0.12.1