support-pvelocity/Code-Llama-2-7B-instruct-text2sql

TEXT GENERATIONConcurrent Unit Cost:1Model Size:7BQuant:FP8Context Size:4kPublished:Oct 25, 2023License:llama2Architecture:Transformer0.0K Open Weights Featherless Exclusive Cold

The support-pvelocity/Code-Llama-2-7B-instruct-text2sql model is a 7 billion parameter Code Llama 2 variant, fine-tuned for text-to-SQL tasks. Developed by support-pvelocity, it generates SQL queries from natural language questions given a database schema. This model utilizes an optimized transformer architecture and has a context length of 4096 tokens, making it specialized for converting English text into SQL.

Loading preview...

Model Overview

This model, support-pvelocity/Code-Llama-2-7B-instruct-text2sql, is a specialized 7 billion parameter instruction-tuned variant of the Code Llama 2 architecture. It has been fine-tuned specifically for text-to-SQL tasks, enabling it to generate SQL queries from natural language questions and provided database schemas. The model was trained for approximately 8 hours on a single A100 40GB GPU, leveraging the bugdaryan/sql-create-context-instruction dataset for its fine-tuning.

Key Capabilities

  • Text-to-SQL Generation: Converts natural language questions into executable SQL queries, given a database schema.
  • Code Completion: Supports general code completion tasks.
  • Infilling: Capable of filling in missing code segments.
  • Instruction Following: Designed to respond to instructions and engage in chat-like interactions.

Intended Use Cases

This model is primarily intended for commercial and research applications focused on text-to-SQL conversion in English. It is suitable for scenarios where users need to programmatically generate SQL queries from user-friendly natural language inputs. The model's architecture is an optimized transformer, and it operates under a custom commercial license from Code Llama.