How to use SQLAlchemy ORM to access Hugging Face Data in Python

The rich ecosystem of Python modules lets you get to work quickly and integrate your systems effectively. With the CData API Driver for Python and the SQLAlchemy toolkit, you can build Hugging Face-connected Python applications and scripts. This article shows how to use SQLAlchemy to connect to Hugging Face data to query Hugging Face data.

With built-in optimized data processing, the CData Python Connector offers unmatched performance for interacting with live Hugging Face data in Python. When you issue complex SQL queries from Hugging Face, the CData Connector pushes supported SQL operations, like filters and aggregations, directly to Hugging Face and utilizes the embedded SQL engine to process unsupported operations client-side (often SQL functions and JOIN operations).

Connecting to Hugging Face Data

Connecting to Hugging Face data looks just like connecting to any relational data source. Create a connection string using the required connection properties. For this article, you will pass the connection string as a parameter to the create_engine function.

HuggingFace Hub uses token-based authentication to enable access to its API. The API provides access to machine learning models, datasets, spaces, papers, and other resources on the HuggingFace Hub platform.

Using API Key Authentication

To authenticate to HuggingFace Hub, you will need to provide an API Key (Access Token). To obtain your access token:

1. Log in to your HuggingFace account at https://huggingface.co
2. Navigate to Settings > Access Tokens
3. Click "New token" to create a new access token
4. Select the appropriate permissions (read or write)
5. Copy the token value

After obtaining your access token, set the following connection properties:

• AuthScheme: Set this to APIKey.
• APIKey: Set this to your HuggingFace access token.

Example connection string

Profile=C:\profiles\HuggingFace.apip;ProfileSettings='APIKey=hf_xxxxxxxxxxxxxxxxxxxx';

Follow the procedure below to install SQLAlchemy and start accessing Hugging Face through Python objects.

Install Required Modules

Use the pip utility to install the SQLAlchemy toolkit and SQLAlchemy ORM package:

pip install sqlalchemy
pip install sqlalchemy.orm

Be sure to import the appropriate modules:

from sqlalchemy import create_engine, String, Column
from sqlalchemy.ext.declarative import declarative_base
from sqlalchemy.orm import sessionmaker

Model Hugging Face Data in Python

You can now connect with a connection string. Use the create_engine function to create an Engine for working with Hugging Face data.

NOTE: Users should URL encode the any connection string properties that include special characters. For more information, refer to the SQL Alchemy documentation.

engine = create_engine("api:///?Profile=C:\profiles\HuggingFace.apip&ProfileSettings='APIKey=hf_xxxxxxxxxxxxxxxxxxxx'")

Declare a Mapping Class for Hugging Face Data

After establishing the connection, declare a mapping class for the table you wish to model in the ORM (in this article, we will model the Collections table). Use the sqlalchemy.ext.declarative.declarative_base function and create a new class with some or all of the fields (columns) defined.

base = declarative_base()
class Collections(base):
	__tablename__ = "Collections"
	 = Column(String,primary_key=True)
	 = Column(String)
	...

Query Hugging Face Data

With the mapping class prepared, you can use a session object to query the data source. After binding the Engine to the session, provide the mapping class to the session query method.

Using the query Method

engine = create_engine("api:///?Profile=C:\profiles\HuggingFace.apip&ProfileSettings='APIKey=hf_xxxxxxxxxxxxxxxxxxxx'")
factory = sessionmaker(bind=engine)
session = factory()
for instance in session.query(Collections).filter_by(=""):
	print(": ", instance.)
	print(": ", instance.)
	print("---------")

Alternatively, you can use the execute method with the appropriate table object. The code below works with an active session.

Using the execute Method

Collections_table = Collections.metadata.tables["Collections"]
for instance in session.execute(Collections_table.select().where(Collections_table.c. == "")):
	print(": ", instance.)
	print(": ", instance.)
	print("---------")

For examples of more complex querying, including JOINs, aggregations, limits, and more, refer to the Help documentation for the extension.

Free Trial & More Information

Download a free, 30-day trial of the CData API Driver for Python to start building Python apps and scripts with connectivity to Hugging Face data. Reach out to our Support Team if you have any questions.