HridaAI/Hrida-T2SQL-3B-V0.1

Hugging Face
TEXT GENERATIONPricing:Input $0.4 / Cached $0.08 / Output $0.8Concurrent Unit Cost:1Model Size:4BQuant:BF16Context Size:4kPublished:Jun 1, 2024License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Warm

HridaAI/Hrida-T2SQL-3B-V0.1 is a 4 billion parameter Text-to-SQL Small Language Model (SLM) fine-tuned from Microsoft's Phi-3-mini-4k-instruct architecture. This model specializes in converting natural language queries into SQL queries, supporting various SQL dialects. It is designed for efficient and accurate SQL generation based on provided database schemas and user input, making it suitable for database interaction applications.

Loading preview...

Hrida-T2SQL-3B-V0.1: Text-to-SQL SLM

Hrida-T2SQL-3B-V0.1 is a 4 billion parameter Small Language Model (SLM) developed by HridaAI, specifically fine-tuned for Text-to-SQL tasks. It is based on the Microsoft/Phi-3-mini-4k-instruct architecture, leveraging its compact size and efficiency for specialized applications.

Key Capabilities

  • Natural Language to SQL Conversion: Translates user queries in natural language into executable SQL statements.
  • Dialect Specification: Supports specifying various SQL dialects (e.g., MySQL, PostgreSQL, SQL Server) to ensure syntax conformity.
  • Schema-Aware Generation: Utilizes provided database schemas (table names, column names, data types) to generate accurate and contextually relevant SQL queries.
  • Instruction-Guided Output: Employs a structured prompt template including system instructions, dialect, context, and input to guide SQL generation.

Prompt Template Structure

The model uses a specific prompt format to optimize performance, including:

  • ### Instruction: System prompt to guide the model.
  • ### Dialect: Optional SQL variant specification.
  • ### Context: Database schema details.
  • ### Input: User's natural language query.
  • ### Response: Expected SQL query output.

Good For

  • Automating database query generation from user input.
  • Integrating natural language interfaces with SQL databases.
  • Applications requiring efficient and accurate SQL code generation from text.