← all repositories

yya518/FinBERT

A BERT model pretrained on 4.9B tokens of financial communications (earnings calls, analyst reports, corporate filings) for financial NLP tasks.

653 stars Jupyter Notebook Language ModelsDomain Apps
FinBERT
Velocity · 7d
+0.3
★ / day
Trend
steady
star history

FinBERT provides pretrained and fine-tuned BERT models tailored for financial text analysis. It includes specialized variants for sentiment classification, ESG rating classification, and forward-looking statement detection. The repository includes Jupyter notebooks demonstrating both how to apply the pre-trained models and how to fine-tune them on specific downstream financial NLP tasks. All models are publicly hosted on Hugging Face.

heatdrop uses Google Analytics to see which pages get read — nothing else. Your call. How we handle data.