degree-bert-finetuning-2

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

• Loss: 0.5890
• Accuracy: 0.698
• F1: 0.6983

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: 32
• eval_batch_size: 32
• 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	F1
0.9822	1.0	104	0.7872	0.638	0.6374
0.8237	2.0	208	0.7235	0.656	0.6552
0.787	3.0	312	0.7133	0.658	0.6521
0.7552	4.0	416	0.6904	0.68	0.6780
0.7366	5.0	520	0.6555	0.704	0.7045
0.7158	6.0	624	0.6500	0.7	0.6994
0.7025	7.0	728	0.6429	0.71	0.7101
0.6912	8.0	832	0.6097	0.698	0.6972
0.6764	9.0	936	0.6033	0.7	0.6996
0.667	10.0	1040	0.6040	0.69	0.6886
0.6571	11.0	1144	0.6022	0.702	0.7011
0.6532	12.0	1248	0.5941	0.71	0.7107
0.6369	13.0	1352	0.5956	0.706	0.7066
0.6385	14.0	1456	0.5929	0.688	0.6864
0.6328	15.0	1560	0.5880	0.708	0.7085
0.6199	16.0	1664	0.5956	0.69	0.6885
0.6109	17.0	1768	0.5865	0.712	0.7125
0.6116	18.0	1872	0.5863	0.716	0.7169
0.6092	19.0	1976	0.5908	0.704	0.7037
0.594	20.0	2080	0.5890	0.698	0.6983

Framework versions

• Transformers 4.38.1
• Pytorch 2.2.0
• Datasets 2.17.1
• Tokenizers 0.15.2