Model Overview

This model, developed by skshmjn, is a 1 billion parameter Llama-3.2-1B variant specifically fine-tuned to generate MongoDB queries from natural language prompts. It leverages the unsloth/Llama-3.2-1B-Instruct as its base and was trained on the skshmjn/mongodb-chat-query dataset.

Key Capabilities

• Generates MongoDB queries for various operations.
• Supports basic CRUD operations (find, insert, update, delete).
• Handles complex Aggregation Pipelines including $group, $match, $lookup, and $sort.
• Capable of generating queries for Indexing & Performance tasks.
• Facilitates Nested Queries & Joins using $lookup.
• Optimized for efficient fine-tuning using Unsloth and fast inference via GGUF quantization.

Good For

• Developers needing to quickly translate natural language into MongoDB query syntax.
• Automating the creation of complex MongoDB aggregation pipelines.
• Integrating natural language interfaces for database interactions in applications.
• Environments requiring efficient, specialized query generation on both CPU and GPU.