SqlAlchemy/Sqlite: InterfaceError - is there a data size limit?

I try to store (an admittedly very large) BLOB into an sqlite database using SqlAlchemy.

For the MCVE I use ubuntu-14.04.2-desktop-amd64.iso as BLOB I want to store. Its size:

$ ls -lhubuntu-14.04.2-desktop-amd64.iso
... 996M ... ubuntu-14.04.2-desktop-amd64.iso

The code

from pathlib import Path
from sqlalchemy import (Column, Integer, String, BLOB, create_engine)
from sqlalchemy.ext.declarative import declarative_base
from sqlalchemy.orm import sessionmaker
from sqlite3 import dbapi2 as sqlite

SA_BASE = declarative_base()

class DbPath(SA_BASE): 

    __tablename__ = 'file'

    pk_path = Column(Integer, primary_key=True)
    path = Column(String)
    data = Column(BLOB, default=None)


def create_session(db_path):

    db_url = 'sqlite+pysqlite:///{}'.format(db_path)
    engine = create_engine(db_url, module=sqlite)
    SA_BASE.metadata.create_all(engine)
    session = sessionmaker(bind=engine)
    return session()

if __name__ == '__main__':  

    pth = Path('/home/user/Downloads/iso/ubuntu-14.04.2-desktop-amd64.iso')

    with pth.open('rb') as file_pointer:
        iso_data = file_pointer.read()

    db_pth = DbPath(path=str(pth), data=iso_data)    

    db_session = create_session('test.sqlite')
    db_session.add(db_pth)
    db_session.commit()

Running this raises the error

InterfaceError: (InterfaceError) Error binding parameter 1 - probably unsupported
  type. 'INSERT INTO file (path, data) VALUES (?, ?)' 
  ('/home/user/Downloads/iso/ubuntu-14.04.2-desktop-amd64.iso', <memory 
  at 0x7faf37cc18e0>)

I looked at the sqlite limitations but found nothing that should prevent me from doing this. Does SqlAlchemy have a limitation?

Everything of this works fine for this file:

$ ls -lh ubuntu-14.04.2-server-amd64.iso
... 595M ... ubuntu-14.04.2-server-amd64.iso

Is there a data size limit? or what do I have to do differently when the file size surpasses a certain (where would that be?) limit?