bert-base-cased_legal_ner_finetuned

This model is a fine-tuned version of google-bert/bert-base-cased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3018
• Law Precision: 0.7364
• Law Recall: 0.8261
• Law F1: 0.7787
• Law Number: 115
• Violated by Precision: 0.8525
• Violated by Recall: 0.6933
• Violated by F1: 0.7647
• Violated by Number: 75
• Violated on Precision: 0.4688
• Violated on Recall: 0.4286
• Violated on F1: 0.4478
• Violated on Number: 70
• Violation Precision: 0.6323
• Violation Recall: 0.7251
• Violation F1: 0.6755
• Violation Number: 491
• Overall Precision: 0.6524
• Overall Recall: 0.7097
• Overall F1: 0.6798
• Overall Accuracy: 0.9439

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Law Precision	Law Recall	Law F1	Law Number	Violated by Precision	Violated by Recall	Violated by F1	Violated by Number	Violated on Precision	Violated on Recall	Violated on F1	Violated on Number	Violation Precision	Violation Recall	Violation F1	Violation Number	Overall Precision	Overall Recall	Overall F1	Overall Accuracy
No log	1.0	85	0.8046	0.0	0.0	0.0	115	0.0	0.0	0.0	75	0.0	0.0	0.0	70	0.0	0.0	0.0	491	0.0	0.0	0.0	0.7619
No log	2.0	170	0.4050	0.0	0.0	0.0	115	0.0	0.0	0.0	75	0.0	0.0	0.0	70	0.1835	0.2037	0.1931	491	0.1835	0.1332	0.1543	0.8819
No log	3.0	255	0.2861	0.6111	0.4783	0.5366	115	0.1818	0.0533	0.0825	75	0.4	0.0571	0.1000	70	0.4345	0.5540	0.4870	491	0.4479	0.4461	0.4470	0.9130
No log	4.0	340	0.2552	0.75	0.7043	0.7265	115	0.5625	0.36	0.4390	75	0.3429	0.1714	0.2286	70	0.4924	0.5927	0.5379	491	0.5256	0.5473	0.5362	0.9257
No log	5.0	425	0.2676	0.7154	0.7652	0.7395	115	0.7308	0.5067	0.5984	75	0.2778	0.1429	0.1887	70	0.5368	0.6090	0.5706	491	0.5664	0.5792	0.5727	0.9300
0.4786	6.0	510	0.2663	0.6767	0.7826	0.7258	115	0.7903	0.6533	0.7153	75	0.3684	0.4	0.3836	70	0.6155	0.7271	0.6667	491	0.6157	0.6977	0.6542	0.9366
0.4786	7.0	595	0.2352	0.6957	0.8348	0.7589	115	0.7941	0.72	0.7552	75	0.4242	0.4	0.4118	70	0.5799	0.7169	0.6412	491	0.6030	0.7057	0.6503	0.9412
0.4786	8.0	680	0.2728	0.6835	0.8261	0.7480	115	0.7857	0.7333	0.7586	75	0.3596	0.4571	0.4025	70	0.5916	0.7434	0.6588	491	0.5978	0.7284	0.6567	0.9415
0.4786	9.0	765	0.2952	0.7385	0.8348	0.7837	115	0.8088	0.7333	0.7692	75	0.5	0.5	0.5	70	0.6246	0.7352	0.6754	491	0.6466	0.7284	0.6850	0.9433
0.4786	10.0	850	0.3018	0.7364	0.8261	0.7787	115	0.8525	0.6933	0.7647	75	0.4688	0.4286	0.4478	70	0.6323	0.7251	0.6755	491	0.6524	0.7097	0.6798	0.9439

Framework versions

• Transformers 4.44.0
• Pytorch 2.4.0
• Datasets 2.21.0
• Tokenizers 0.19.1