opptimizer 2.3.8

Optimization of hyperparameters for Machine Learning (AI)

Introduction to Opptimizer

Opptimizer is an optimization framework designed for AI pipelines, enabling efficient management and execution of complex processes and workflows.

This section provides a quick overview of installation, basic usage, and simple examples.
To dive deeper, please see the full documentation at https://opptimizer.readthedocs.io/.

Installation

Installin by PyPI is appropriate for most use cases.

pip install opptimizer

There is also advanced way of local install described here.

Basic Usage

1. Import the opptimizer Library

Begin by importing the opptimizer library into your project:

import opptimizer

2. Create an Executor Instance

Create an instance of the Executor class. You can optionally set the context and/or specify a configuration file containing the context during instantiation:

executor = opptimizer.Executor(context=my_context)
# or with a configuration file (you can use any file name)
executor = opptimizer.Executor(cfg="path/to/opp.cfg")

3. Define Modules

Define the modules you intend to use. Modules can be function:

def module_exec(mod_obj, params, tokenData):
    # Process tokenData
    return tokenData

or instances of the Mod class:

module_instance = opptimizer.Mod(module_exec)

4. Run the Executor.run() Method

Run the Executor with input data or a list of modules:

result = executor.run(input_data,modules=[module_exec])

or with an module instance:

result = executor.run(input_data, modules=[module_instance])

You can use any number of modules in modules list that will be executed sequentially.

5. Get the Result from Execution

Retrieve the result from the execution:

final_result = result.get_result()

Examples of Usage

Basic Pipeline Example

The following example demonstrates a simple pipeline that extends an input string by adding additional parts through modules mod_1 and mod_2, ultimately returning the combined text.

import opptimizer as op

def mod_1(mod_obj, params, tokenData = None):
    tokenData = tokenData + 'result 1'
    return tokenData

def mod_2(mod_obj, params, tokenData = None):
    tokenData = tokenData + ', result 2'
    mod_obj.setResult(tokenData)

executor = op.Executor(modules = [mod_1, mod_2])
response = executor.run('Results: ')

print(f"{response.get_result()}")

Output:
    Results: result 1, result 2

License

This project is licensed under the Apache License 2.0. You may use, modify, and distribute this software freely, provided that you comply with the terms of the license, which include attribution and a disclaimer of warranties. For the full license text, please see Apache License 2.0.