flan-t5-base_legal_ner_finetuned

khalidrajan/flan-t5-base_legal_ner_finetuned

flan-t5-base_legal_ner_finetuned is a machine learning model from khalidrajan. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

19% of all-time downloads

All-time downloads

Public

Parameters

110M

439 MB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors439 MB · 99%

At a glance

License: apache-2.0
Model type: t5
Access: Public
Created: Sep 6, 2024
Updated: Sep 6, 2024
SHA: 80ce5ccb

Base models

google/flan-t5-base

Model card

From the Hugging Face model README

flan-t5-base_legal_ner_finetuned

This model is a fine-tuned version of google/flan-t5-base on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.3290
Law Precision: 0.5522
Law Recall: 0.5968
Law F1: 0.5736
Law Number: 124
Violated by Precision: 0.0
Violated by Recall: 0.0
Violated by F1: 0.0
Violated by Number: 77
Violated on Precision: 0.0845
Violated on Recall: 0.0845
Violated on F1: 0.0845
Violated on Number: 71
Violation Precision: 0.2370
Violation Recall: 0.3800
Violation F1: 0.2919
Violation Number: 479
Overall Precision: 0.2352
Overall Recall: 0.3489
Overall F1: 0.2810
Overall Accuracy: 0.9247

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	3.3698	0.0002	0.0081	0.0005	124	0.0004	0.0260	0.0008	77	0.0010	0.0423	0.0020	71	0.0046	0.0209	0.0076	479	0.0011	0.0213	0.0021	0.2756
No log	2.0	170	1.6151	0.0	0.0	0.0	124	0.0	0.0	0.0	77	0.0	0.0	0.0	71	0.0104	0.0355	0.0161	479	0.0029	0.0226	0.0051	0.6383
No log	3.0	255	0.9385	0.0	0.0	0.0	124	0.0	0.0	0.0	77	0.0	0.0	0.0	71	0.0129	0.0543	0.0209	479	0.0075	0.0346	0.0124	0.7646
No log	4.0	340	0.6876	0.0013	0.0081	0.0022	124	0.0	0.0	0.0	77	0.0	0.0	0.0	71	0.0371	0.1148	0.0561	479	0.0210	0.0746	0.0327	0.8109
No log	5.0	425	0.5094	0.0097	0.0645	0.0168	124	0.0	0.0	0.0	77	0.05	0.0141	0.0220	71	0.0824	0.2589	0.125	479	0.0511	0.1771	0.0793	0.8448
1.8667	6.0	510	0.4201	0.0325	0.25	0.0575	124	0.0	0.0	0.0	77	0.0526	0.0282	0.0367	71	0.1122	0.2985	0.1631	479	0.0726	0.2344	0.1109	0.8651
1.8667	7.0	595	0.3759	0.0525	0.4194	0.0934	124	0.0087	0.0130	0.0104	77	0.0698	0.0845	0.0764	71	0.1441	0.3111	0.1970	479	0.0935	0.2770	0.1398	0.8792
1.8667	8.0	680	0.3463	0.1856	0.5403	0.2763	124	0.0092	0.0130	0.0108	77	0.0510	0.0704	0.0592	71	0.1955	0.3800	0.2582	479	0.1701	0.3395	0.2267	0.9090
1.8667	9.0	765	0.3315	0.4516	0.5645	0.5018	124	0.0	0.0	0.0	77	0.0769	0.0845	0.0805	71	0.2240	0.3779	0.2813	479	0.2176	0.3422	0.2660	0.9221
1.8667	10.0	850	0.3290	0.5522	0.5968	0.5736	124	0.0	0.0	0.0	77	0.0845	0.0845	0.0845	71	0.2370	0.3800	0.2919	479	0.2352	0.3489	0.2810	0.9247

Framework versions

Transformers 4.44.0
Pytorch 2.4.0
Datasets 2.21.0
Tokenizers 0.19.1

Type: t5
License: apache-2.0
Created: Sep 6, 2024
Updated: Sep 6, 2024