Ellbendls/Qwen-2.5-3b-Text_to_SQL is a 3.1 billion parameter model fine-tuned from Qwen/Qwen2.5-3B-Instruct, specifically designed for converting natural language queries into SQL statements. This model excels at Text-to-SQL conversion and can generate table schema context when none is provided, making it robust for analytics and reporting tasks. It is optimized for English language prompts and handles SQL queries involving aggregation, grouping, and filtering.
Loading preview...
Fine-Tuned LLM for Text-to-SQL Conversion
This model, fine-tuned from the Qwen/Qwen2.5-3B-Instruct base, specializes in transforming natural language questions into accurate SQL queries. It addresses a common challenge in data interaction by not only generating SQL but also inferring and providing table schema context when it's not explicitly given in the prompt. This capability makes it particularly useful for scenarios where users might not have immediate access to database schemas.
Key Capabilities
- Text-to-SQL Conversion: Accurately translates natural language into SQL statements.
- Schema Generation: Automatically generates relevant table schema context when no schema is provided in the input.
- Analytics and Reporting Optimization: Handles complex SQL operations such as aggregation, grouping, and filtering, making it suitable for business intelligence and data analysis.
Training and Limitations
The model was trained on the gretelai/synthetic_text_to_sql dataset, which includes diverse natural language queries mapped to SQL with optional schema contexts. While highly effective for its intended purpose, it has some limitations:
- May struggle with highly nested or exceptionally advanced SQL tasks.
- Primarily optimized for English language prompts; performance with non-English inputs may vary.
- Context dependence means it might generate incorrect schemas if instructions are ambiguous or insufficient.