Generating SQL for Oracle using Ollama, ChromaDB

This notebook runs through the process of using the vanna Python package to generate SQL using AI (RAG + LLMs) including connecting to a database and training. If you're not ready to train on your own database, you can still try it using a sample SQLite database.

Run Using Colab Open in GitHub

Which LLM do you want to use?

• OpenAI via Vanna.AI (Recommended)

  Use Vanna.AI for free to generate your queries
• OpenAI

  Use OpenAI with your own API key
• Azure OpenAI

  If you have OpenAI models deployed on Azure
• Anthropic

  Use Anthropics Claude with your Anthropic API Key
• [Selected] Ollama

  Use Ollama locally for free. Requires additional setup.
• Google Gemini

  Use Google Gemini with your Gemini or Vertex API Key
• Mistral via Mistral API

  If you have a Mistral API key
• Other LLM

  If you have a different LLM model

Where do you want to store the 'training' data?

• Vanna Hosted Vector DB (Recommended)

  Use Vanna.AIs hosted vector database (pgvector) for free. This is usable across machines with no additional setup.
• [Selected] ChromaDB

  Use ChromaDBs open-source vector database for free locally. No additional setup is necessary -- all database files will be created and stored locally.
• Qdrant

  Use Qdrants open-source vector database
• Marqo

  Use Marqo locally for free. Requires additional setup. Or use their hosted option.
• Other VectorDB

  Use any other vector database. Requires additional setup.

Setup

%pip install 'vanna[chromadb,ollama,mssql]'

from vanna.ollama import Ollama
from vanna.chromadb import ChromaDB_VectorStore

class MyVanna(ChromaDB_VectorStore, Ollama):
    def __init__(self, config=None):
        ChromaDB_VectorStore.__init__(self, config=config)
        Ollama.__init__(self, config=config)

vn = MyVanna(config={'model': 'mistral'})

Which database do you want to query?

• Postgres
• Microsoft SQL Server
• MySQL
• DuckDB
• Snowflake
• BigQuery
• SQLite
• [Selected] Oracle
• Other Database

  Use Vanna to generate queries for any SQL database

vn.connect_to_oracle(dsn='my-dsn', user='my-user', password='my-password')

Training

You only need to train once. Do not train again unless you want to add more training data.

# The information schema query may need some tweaking depending on your database. This is a good starting point.
df_information_schema = vn.run_sql("SELECT * FROM INFORMATION_SCHEMA.COLUMNS")

# This will break up the information schema into bite-sized chunks that can be referenced by the LLM
plan = vn.get_training_plan_generic(df_information_schema)
plan

# If you like the plan, then uncomment this and run it to train
# vn.train(plan=plan)

# The following are methods for adding training data. Make sure you modify the examples to match your database.

# DDL statements are powerful because they specify table names, colume names, types, and potentially relationships
vn.train(ddl="""
    CREATE TABLE IF NOT EXISTS my-table (
        id INT PRIMARY KEY,
        name VARCHAR(100),
        age INT
    )
""")

# Sometimes you may want to add documentation about your business terminology or definitions.
vn.train(documentation="Our business defines OTIF score as the percentage of orders that are delivered on time and in full")

# You can also add SQL queries to your training data. This is useful if you have some queries already laying around. You can just copy and paste those from your editor to begin generating new SQL.
vn.train(sql="SELECT * FROM my-table WHERE name = 'John Doe'")

# At any time you can inspect what training data the package is able to reference
training_data = vn.get_training_data()
training_data

# You can remove training data if there's obsolete/incorrect information. 
vn.remove_training_data(id='1-ddl')

```## Asking the AI
Whenever you ask a new question, it will find the 10 most relevant pieces of training data and use it as part of the LLM prompt to generate the SQL.
```python
vn.ask(question=...)

Launch the User Interface

from vanna.flask import VannaFlaskApp
app = VannaFlaskApp(vn)
app.run()