Downloads · 30 days
7
6% of all-time downloads
sf404/esg-ner-sb253
esg-ner-sb253 is a token classification model from sf404. Use it when you need labels on individual words, such as names. The card lists the license as apache-2.0.
A custom Named Entity Recognition (NER) model trained for extracting key entities from ESG (Environmental, Social, and Governance) reports, sustainability documents, and financial statements.
Downloads · 30 days
7
6% of all-time downloads
All-time downloads
120
Public
Parameters
184M
735 MB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors735 MB · 99%
From the Hugging Face model README
A custom Named Entity Recognition (NER) model trained for extracting key entities from ESG (Environmental, Social, and Governance) reports, sustainability documents, and financial statements.
This model identifies and classifies entities relevant to ESG reporting and financial analysis, supporting automated data extraction from sustainability reports, XBRL filings, and financial documents.
Model Architecture: XLM-RoBERTa-based Token Classification
Model Size: ~800MB
Training Domain: ESG/Sustainability/Financial Reports
Inference: Optimized for CPU deployment on Google Cloud Run
| Entity Type | Description | Examples |
|---|---|---|
| CONCEPT | ESG concepts, metrics, and terminology | "carbon emissions", "renewable energy", "employee satisfaction" |
| VALUE | Numerical values and amounts | "15%", "2.5 million", "50 tons" |
| UNIT | Units of measurement | "CO2e", "USD", "percent", "GWh" |
| PERIOD | Time periods and reporting dates | "Q4 2024", "fiscal year 2023", "January-March" |
| ENTITY | Organizations, companies, subsidiaries | "Acme Corporation", "European Union", "Board of Directors" |
| CHANGE | Change indicators and trends | "increased", "decreased", "reduced by", "grew" |
The model is deployed as a FastAPI service with the following endpoints:
Base URL: https://ner-backend-171009084156.europe-west1.run.app