bert-finetuned-MedicalChunkSecond

AJS50/bert-finetuned-MedicalChunkSecond

bert-finetuned-MedicalChunkSecond is a token classification model from AJS50. Use it when you need labels on individual words, such as names. It is set up for transformers. 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

Repo size

4.7 GB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.bin431 MB · 100%

At a glance

Task: Token Classification
Library: transformers
License: apache-2.0
Model type: bert
Access: Public
Created: Dec 15, 2022
Updated: Dec 15, 2022
SHA: c45507d7

Try a prompt

Model card

From the Hugging Face model README

bert-finetuned-MedicalChunkSecond

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

Loss: 0.1804
Precision: 0.2396
Recall: 0.2613
F1: 0.25
Accuracy: 0.9541
Pop Precision: 0.3889
Pop Recall: 0.3818
Pop F1: 0.3853
Pop Number: 55
Int Precision: 0.2065
Int Recall: 0.2468
Int F1: 0.2249
Int Number: 77
Out Precision: 0.1690
Out Recall: 0.1791
Out F1: 0.1739
Out Number: 67
Pop Count: 121
Int Count: 254
Out Count: 158

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
num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy	Pop Precision	Pop Recall	Pop F1	Pop Number	Int Precision	Int Recall	Int F1	Int Number	Out Precision	Out Recall	Out F1	Out Number	Pop Count	Int Count	Out Count
No log	1.0	56	0.1626	0.0	0.0	0.0	0.9599	0.0	0.0	0.0	55	0.0	0.0	0.0	77	0.0	0.0	0.0	67	1	57	0
No log	2.0	112	0.1420	0.0962	0.0503	0.0660	0.9575	0.0588	0.0182	0.0278	55	0.1084	0.1169	0.1125	77	0.0	0.0	0.0	67	24	193	4
No log	3.0	168	0.1354	0.1604	0.1508	0.1554	0.9568	0.2449	0.2182	0.2308	55	0.1204	0.1688	0.1405	77	0.1667	0.0746	0.1031	67	107	261	49
No log	4.0	224	0.1360	0.2701	0.1859	0.2202	0.9620	0.3478	0.2909	0.3168	55	0.1961	0.1299	0.1562	77	0.275	0.1642	0.2056	67	98	137	77
No log	5.0	280	0.1443	0.2914	0.2563	0.2727	0.9603	0.4038	0.3818	0.3925	55	0.2289	0.2468	0.2375	77	0.275	0.1642	0.2056	67	121	199	85
No log	6.0	336	0.1618	0.2988	0.2462	0.2700	0.9601	0.4865	0.3273	0.3913	55	0.2571	0.2338	0.2449	77	0.2281	0.1940	0.2097	67	85	187	121
No log	7.0	392	0.1622	0.2417	0.2563	0.2488	0.9571	0.3333	0.3636	0.3478	55	0.2125	0.2208	0.2166	77	0.1972	0.2090	0.2029	67	126	213	142
No log	8.0	448	0.1741	0.2356	0.2663	0.25	0.9544	0.3667	0.4	0.3826	55	0.1919	0.2468	0.2159	77	0.1818	0.1791	0.1805	67	132	258	147
0.1112	9.0	504	0.1796	0.2275	0.2663	0.2454	0.9527	0.3929	0.4	0.3964	55	0.1845	0.2468	0.2111	77	0.1622	0.1791	0.1702	67	129	278	158
0.1112	10.0	560	0.1804	0.2396	0.2613	0.25	0.9541	0.3889	0.3818	0.3853	55	0.2065	0.2468	0.2249	77	0.1690	0.1791	0.1739	67	121	254	158

Framework versions

Transformers 4.22.1
Pytorch 1.12.1
Datasets 2.4.0
Tokenizers 0.12.1

Task: Token Classification
Library: transformers
Type: bert
License: apache-2.0
Created: Dec 15, 2022
Updated: Dec 15, 2022