BERT_NER_Ep5_PAD_50 - finetuned - ner Open - source Model - Accurately Identify Named Entities, with an F1 Score of 0.6920 on the Evaluation Set

BERT NER Ep5 PAD 50 Finetuned Ner

Developed by suwani

Named entity recognition model fine-tuned based on bert-base-cased, achieving an F1 score of 0.6920 on the evaluation set

Downloads 16

Release Time : 3/2/2022

Model Overview

This model is a named entity recognition model based on the BERT architecture, specifically designed to identify named entities in text.

High Recall

Achieves a recall of 0.7348 on the evaluation set, effectively identifying named entities in text.

Balanced Performance

Achieves an F1 score of 0.6920, striking a good balance between precision and recall.

BERT-based Architecture

Utilizes BERT's powerful contextual understanding capabilities for entity recognition.

Text Entity Recognition

Named Entity Labeling

Sequence Labeling

Information Extraction

News Entity Extraction

Extract entities such as person names, locations, and organizations from news text.

F1 score 0.6920

Biomedical Literature Analysis

Identify specialized terms and entities in medical literature.

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	288	0.3705	0.5852	0.6215	0.6028	0.8793
0.4885	2.0	576	0.3351	0.5925	0.7317	0.6548	0.8865
0.4885	3.0	864	0.3196	0.6471	0.7138	0.6788	0.8994
0.2172	4.0	1152	0.3368	0.6454	0.7323	0.6861	0.8992
0.2172	5.0	1440	0.3491	0.6507	0.7312	0.6886	0.9008
0.1459	6.0	1728	0.3833	0.6715	0.7018	0.6863	0.9013
0.1045	7.0	2016	0.3893	0.6540	0.7348	0.6920	0.9006

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base