Output_LayoutLMv3_v7 Open-source Document Understanding Model - Free Deployment for Precise Document Layout Analysis

Output LayoutLMv3 V7

Developed by Noureddinesa

A document understanding model fine-tuned based on microsoft/layoutlmv3-base, excelling in document layout analysis tasks

Text Recognition

Transformers

#Document Intelligence Analysis #High-precision OCR #Form Recognition

Downloads 18

Release Time : 4/12/2024

Model Overview

This model is a fine-tuned version of LayoutLMv3, focusing on document understanding and information extraction tasks, capable of processing documents containing text and layout information

Model Features

Document Layout Understanding

Capable of processing both textual content and visual layout information simultaneously

High-precision Information Extraction

Achieves a recall rate of 0.8 and an F1 score of 0.79 on the evaluation dataset

Multimodal Processing Capability

Integrates text, layout, and visual features for comprehensive analysis

Model Capabilities

Document Layout Analysis

Form Recognition

Information Extraction

Document Understanding

Use Cases

Document Processing

Form Data Extraction

Automatically extracts key field information from structured documents

Accuracy rate of 97.23%

Invoice Processing

Identifies supplier, amount, date, and other information in invoices

Office Automation

Contract Analysis

Automatically identifies key clauses and signatory information in contracts

🚀 Output_LayoutLMv3_v7

This model is a fine - tuned version of [microsoft/layoutlmv3 - base](https://huggingface.co/microsoft/layoutlmv3 - base) on an unknown dataset. It offers high - performance results in evaluation, which can be used for various related tasks, providing accurate and reliable data analysis.

🚀 Quick Start

This model is a fine - tuned version of [microsoft/layoutlmv3 - base](https://huggingface.co/microsoft/layoutlmv3 - base) on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.1075
Precision: 0.7928
Recall: 0.8
F1: 0.7964
Accuracy: 0.9723

📚 Documentation

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

🔧 Technical Details

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 1e - 06
train_batch_size: 12
eval_batch_size: 12
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon = 1e - 08
lr_scheduler_type: cosine
training_steps: 2600

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	9.09	100	0.4138	0.0	0.0	0.0	0.8962
No log	18.18	200	0.2709	0.1667	0.0273	0.0469	0.9014
No log	27.27	300	0.2003	0.6234	0.4364	0.5134	0.9360
No log	36.36	400	0.1711	0.6496	0.6909	0.6696	0.9481
0.3384	45.45	500	0.1624	0.6667	0.7273	0.6957	0.9498
0.3384	54.55	600	0.1502	0.6803	0.7545	0.7155	0.9550
0.3384	63.64	700	0.1428	0.7227	0.7818	0.7511	0.9602
0.3384	72.73	800	0.1452	0.7049	0.7818	0.7414	0.9550
0.3384	81.82	900	0.1260	0.7544	0.7818	0.7679	0.9671
0.0995	90.91	1000	0.1254	0.7544	0.7818	0.7679	0.9671
0.0995	100.0	1100	0.1211	0.7863	0.8364	0.8106	0.9706
0.0995	109.09	1200	0.1093	0.7739	0.8091	0.7911	0.9706
0.0995	118.18	1300	0.1081	0.7946	0.8091	0.8018	0.9723
0.0995	127.27	1400	0.1108	0.7778	0.8273	0.8018	0.9723
0.0608	136.36	1500	0.1115	0.7627	0.8182	0.7895	0.9706
0.0608	145.45	1600	0.1034	0.8053	0.8273	0.8161	0.9740
0.0608	154.55	1700	0.1050	0.7895	0.8182	0.8036	0.9723
0.0608	163.64	1800	0.1093	0.7739	0.8091	0.7911	0.9706
0.0608	172.73	1900	0.1043	0.7965	0.8182	0.8072	0.9723
0.0443	181.82	2000	0.1048	0.8036	0.8182	0.8108	0.9758
0.0443	190.91	2100	0.1067	0.8036	0.8182	0.8108	0.9758
0.0443	200.0	2200	0.1069	0.8036	0.8182	0.8108	0.9740
0.0443	209.09	2300	0.1083	0.7928	0.8	0.7964	0.9723
0.0443	218.18	2400	0.1079	0.7928	0.8	0.7964	0.9723
0.0381	227.27	2500	0.1076	0.7928	0.8	0.7964	0.9723
0.0381	236.36	2600	0.1075	0.7928	0.8	0.7964	0.9723

Framework versions

Transformers 4.29.2
Pytorch 2.2.2+cu121
Datasets 2.18.0
Tokenizers 0.13.3

📄 License

This model is released under the CC - BY - NC - SA 4.0 license.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご