distilbert-base-uncased_legal_ner_finetuned

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

• Loss: 0.2765
• Law Precision: 0.7983
• Law Recall: 0.8716
• Law F1: 0.8333
• Law Number: 109
• Violated by Precision: 0.7937
• Violated by Recall: 0.7042
• Violated by F1: 0.7463
• Violated by Number: 71
• Violated on Precision: 0.3934
• Violated on Recall: 0.3429
• Violated on F1: 0.3664
• Violated on Number: 70
• Violation Precision: 0.5657
• Violation Recall: 0.6588
• Violation F1: 0.6087
• Violation Number: 425
• Overall Precision: 0.6084
• Overall Recall: 0.6652
• Overall F1: 0.6355
• Overall Accuracy: 0.9409

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	1.1323	0.0	0.0	0.0	109	0.0	0.0	0.0	71	0.0	0.0	0.0	70	0.0	0.0	0.0	425	0.0	0.0	0.0	0.7656
No log	2.0	170	0.4593	0.0	0.0	0.0	109	0.0	0.0	0.0	71	0.0	0.0	0.0	70	0.1391	0.1741	0.1546	425	0.1391	0.1096	0.1226	0.8706
No log	3.0	255	0.3529	0.1923	0.0459	0.0741	109	0.0	0.0	0.0	71	0.0	0.0	0.0	70	0.2088	0.2	0.2043	425	0.2079	0.1333	0.1625	0.8943
No log	4.0	340	0.2708	0.1176	0.0734	0.0904	109	0.0	0.0	0.0	71	0.0	0.0	0.0	70	0.4321	0.4941	0.4610	425	0.3928	0.3230	0.3545	0.9134
No log	5.0	425	0.2579	0.8295	0.6697	0.7411	109	0.6667	0.3099	0.4231	71	0.3095	0.1857	0.2321	70	0.4197	0.4612	0.4395	425	0.4825	0.4504	0.4659	0.9153
0.5875	6.0	510	0.2516	0.8091	0.8165	0.8128	109	0.6	0.5070	0.5496	71	0.3542	0.2429	0.2881	70	0.5458	0.6588	0.5970	425	0.5773	0.6252	0.6003	0.9342
0.5875	7.0	595	0.2355	0.7946	0.8165	0.8054	109	0.7167	0.6056	0.6565	71	0.3438	0.3143	0.3284	70	0.5455	0.6353	0.5870	425	0.5800	0.6281	0.6031	0.9382
0.5875	8.0	680	0.2659	0.8246	0.8624	0.8430	109	0.7286	0.7183	0.7234	71	0.3243	0.3429	0.3333	70	0.5491	0.6706	0.6038	425	0.5843	0.6726	0.6253	0.9398
0.5875	9.0	765	0.2839	0.752	0.8624	0.8034	109	0.7391	0.7183	0.7286	71	0.3421	0.3714	0.3562	70	0.5524	0.6824	0.6105	425	0.5799	0.6830	0.6272	0.9394
0.5875	10.0	850	0.2765	0.7983	0.8716	0.8333	109	0.7937	0.7042	0.7463	71	0.3934	0.3429	0.3664	70	0.5657	0.6588	0.6087	425	0.6084	0.6652	0.6355	0.9409

Framework versions

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