adhishtanaka/ llama_3.2_1b-SQL:latest

65 9 months ago

This is a fine-tuned llama 3.2 model with 1 billion parameters. It turns plain English questions into SQL queries. It was trained using examples of questions and their matching SQL, making it easier to work with databases using natural language. 

ollama run adhishtanaka/llama_3.2_1b-SQL

Details

9 months ago

357c13eb59f6 · 808MB ·

model
llama
·
1.24B
·
Q4_K_M
system
You are a helpful assistant that writes SQL queries based on table structures and user questions.
params
{ "stop": [ "<|start_header_id|>", "<|end_header_id|>", "<|eot_id|>"
template
{{- range .Messages }}<|start_header_id|>{{ .Role }}<|end_header_id|> {{ .Content }}<|eot_id|> {{- e

Readme

Model Description

llama_3.2_1b_SQL is a fine-tuned version of LLaMA 3.2 1B designed to generate SQL queries from natural language input. The model is trained to understand user questions and convert them into syntactically correct and semantically accurate SQL statements.

What It Does

  • Converts natural language questions into SQL queries.
  • Supports various SQL tasks like SELECT, WHERE, GROUP BY, JOIN, etc.
  • Useful for building tools that let users query data without writing SQL.

Dataset Used

  • Name: synthetic_text_to_sql-ShareGPT
  • Source: Hugging Face Dataset
  • Contains examples of user questions and their corresponding SQL queries.

Libraries Used

  • Unsloth: For efficient fine-tuning with LoRA adapters.
  • llama-cpp-python: To run the quantized .gguf model locally.

Contact

For questions, feedback, or collaboration, feel free to connect with me: