finance-sentiment-pl-base
| Property | Value |
|---|---|
| Parameter Count | 124M |
| Model Type | Text Classification |
| Tensor Type | I64, F32 |
| Performance | F1 Macro: 0.969 |
What is finance-sentiment-pl-base?
finance-sentiment-pl-base is a specialized sentiment analysis model designed for Polish financial texts. Built on the herbert-base architecture, this model was developed by bards.ai to analyze sentiment in financial news and statements. It classifies text into three categories: positive, negative, and neutral, with impressive accuracy metrics.
Implementation Details
The model was trained on a translated version of the Financial PhraseBank dataset, utilizing a single RTX3090 GPU for 10 epochs. It achieves remarkable performance metrics, including 0.969 F1 macro score, 0.971 precision, and 0.968 recall, with an overall accuracy of 0.976.
- Based on herbert-base architecture
- Processes 136.8 samples per second on RTX 3090
- Implements PyTorch and Transformers frameworks
- Uses Safetensors for model storage
Core Capabilities
- Three-class sentiment classification (positive, negative, neutral)
- Specialized in Polish financial text analysis
- High-performance metrics across all evaluation criteria
- Simple integration with Hugging Face transformers pipeline
Frequently Asked Questions
Q: What makes this model unique?
This model is specifically optimized for Polish financial text analysis, combining the powerful herbert-base architecture with specialized training on financial data. Its high performance metrics and focus on the Polish language make it particularly valuable for financial sentiment analysis in Polish markets.
Q: What are the recommended use cases?
The model is ideal for analyzing Polish financial news, market reports, company statements, and other financial documents where sentiment analysis is crucial. It can be integrated into financial monitoring systems, market analysis tools, and automated trading systems focusing on Polish markets.
