biobert-base-uncased-ner

This model is a fine-tuned version of dmis-lab/biobert-v1.1 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0299
• Cases: {'precision': 0.963963963963964, 'recall': 0.9705215419501134, 'f1': 0.9672316384180792, 'number': 441}
• Country: {'precision': 0.9926062846580407, 'recall': 0.9962894248608535, 'f1': 0.9944444444444445, 'number': 539}
• Date: {'precision': 0.9637931034482758, 'recall': 0.9704861111111112, 'f1': 0.9671280276816608, 'number': 576}
• Deaths: {'precision': 0.9224376731301939, 'recall': 0.9596541786743515, 'f1': 0.9406779661016949, 'number': 347}
• Virus: {'precision': 0.9927140255009107, 'recall': 0.9927140255009107, 'f1': 0.9927140255009107, 'number': 549}
• Overall Precision: 0.9705
• Overall Recall: 0.9796
• Overall F1: 0.9750
• Overall Accuracy: 0.9923

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Cases	Country	Date	Deaths	Virus	Overall Precision	Overall Recall	Overall F1	Overall Accuracy
No log	1.0	291	0.0329	{'precision': 0.9712918660287081, 'recall': 0.9206349206349206, 'f1': 0.9452852153667054, 'number': 441}	{'precision': 0.988950276243094, 'recall': 0.9962894248608535, 'f1': 0.9926062846580408, 'number': 539}	{'precision': 0.9498269896193772, 'recall': 0.953125, 'f1': 0.951473136915078, 'number': 576}	{'precision': 0.9388379204892966, 'recall': 0.8847262247838616, 'f1': 0.9109792284866469, 'number': 347}	{'precision': 0.9926873857404022, 'recall': 0.9890710382513661, 'f1': 0.990875912408759, 'number': 549}	0.9706	0.9551	0.9628	0.9901
0.0216	2.0	582	0.0336	{'precision': 0.9527027027027027, 'recall': 0.9591836734693877, 'f1': 0.9559322033898305, 'number': 441}	{'precision': 0.9907749077490775, 'recall': 0.9962894248608535, 'f1': 0.9935245143385755, 'number': 539}	{'precision': 0.9616724738675958, 'recall': 0.9583333333333334, 'f1': 0.96, 'number': 576}	{'precision': 0.9010989010989011, 'recall': 0.9452449567723343, 'f1': 0.9226441631504924, 'number': 347}	{'precision': 0.9908759124087592, 'recall': 0.9890710382513661, 'f1': 0.9899726526891522, 'number': 549}	0.9640	0.9719	0.9679	0.9907
0.0216	3.0	873	0.0345	{'precision': 0.9555555555555556, 'recall': 0.9750566893424036, 'f1': 0.9652076318742986, 'number': 441}	{'precision': 0.9926062846580407, 'recall': 0.9962894248608535, 'f1': 0.9944444444444445, 'number': 539}	{'precision': 0.9536082474226805, 'recall': 0.9635416666666666, 'f1': 0.9585492227979275, 'number': 576}	{'precision': 0.9131652661064426, 'recall': 0.9394812680115274, 'f1': 0.9261363636363636, 'number': 347}	{'precision': 0.990909090909091, 'recall': 0.9927140255009107, 'f1': 0.991810737033667, 'number': 549}	0.9649	0.9759	0.9704	0.9914
0.0126	4.0	1164	0.0292	{'precision': 0.9682539682539683, 'recall': 0.9682539682539683, 'f1': 0.9682539682539683, 'number': 441}	{'precision': 0.9907749077490775, 'recall': 0.9962894248608535, 'f1': 0.9935245143385755, 'number': 539}	{'precision': 0.9655172413793104, 'recall': 0.9722222222222222, 'f1': 0.9688581314878894, 'number': 576}	{'precision': 0.9301675977653632, 'recall': 0.9596541786743515, 'f1': 0.9446808510638297, 'number': 347}	{'precision': 0.9927140255009107, 'recall': 0.9927140255009107, 'f1': 0.9927140255009107, 'number': 549}	0.9725	0.9796	0.9760	0.9925
0.0126	5.0	1455	0.0299	{'precision': 0.963963963963964, 'recall': 0.9705215419501134, 'f1': 0.9672316384180792, 'number': 441}	{'precision': 0.9926062846580407, 'recall': 0.9962894248608535, 'f1': 0.9944444444444445, 'number': 539}	{'precision': 0.9637931034482758, 'recall': 0.9704861111111112, 'f1': 0.9671280276816608, 'number': 576}	{'precision': 0.9224376731301939, 'recall': 0.9596541786743515, 'f1': 0.9406779661016949, 'number': 347}	{'precision': 0.9927140255009107, 'recall': 0.9927140255009107, 'f1': 0.9927140255009107, 'number': 549}	0.9705	0.9796	0.9750	0.9923

Framework versions

• Transformers 4.51.3
• Pytorch 2.5.1+cu121
• Datasets 3.5.0
• Tokenizers 0.21.1