How can do cross-server queries in SQLAlchemy?

Question

I would like to transfer most of my code written on T-SQL to Python, and came across basic data problems that could not be solved by searching in the SQLAlchemy documentation and in other forums.

Question:

How can I implement the following query options with SQLAlchemy (example for the test - below)?

1. select a.field1, a.field2, b.field2 from server1.database1.schema1.table_a as a inner server2.database1.schema1.table_b as b on a.fileld1 = b.fileld1

2. select a.field1, a.field2, b.field2 from server1.database1.schema1.table_a as a inner dataset_variable (**) as b on a.fileld1 = b.fileld1

3. The result of the query - print (select) (see below in the test case) returns an error: sqlalchemy.exc.ProgrammingError: (pyodbc.ProgrammingError) ('42S02', "[42S02] [M icrosoft] [ODBC SQL Server Driver] [SQL Server] Invalid object name 'server2.database1schema1.table_b'.

** Data received from another SQLAlchemy session (server2.database1.schema1.table_b) and stored in the variable dataset_variable

Test connection example: Two linked sql servers are used.

engine = create_engine("mssql+pyodbc://@{server1}/{}?driver=SQL+Server?trusted_connection=yes", echo=True)

metadata = MetaData(engine)
Base = declarative_base(metadata=metadata)

class table_a(Base):
    __tablename__ = 'table_a '
    __table_args__ = {
    'schema': 'database1.schema1'
    }
    id = Column(Integer, primary_key=True)
    f1 = Column(String(100))
    f2 = Column(String(100))

class table_b(Base):
    __tablename__ = 'table_b '
    __table_args__ = {
    'schema': 'server2.database1.schema1'
    }
    id = Column(Integer, primary_key=True)
    f1 = Column(String(100))
    f2 = Column(String(100))

Session = sessionmaker(bind=engine)
session = Session()

select = session.query(table_b.id).first() 
print(select)