FloomSDK for Python

FloomSDK is a comprehensive Python client designed to facilitate the integration of Floom's Generative AI capabilities into your Python applications. This SDK offers a straightforward interface for making requests to Floom's API, allowing developers to seamlessly incorporate advanced AI features into their projects while abstracting away the complexity of direct API calls.

Features

• Simplified Integration: Effortlessly integrate Floom's Generative AI functionalities into your Python applications.
• Python 3 Support: Fully supports Python 3, ensuring compatibility with modern Python environments.
• Versatile: Suitable for use in a wide range of Python projects, from simple scripts to complex applications.

Getting Started

Prerequisites

• Python 3.6 or newer
• An API key from Floom

Installation

Install Floom SDK using pip:

bash python3 -m pip install floom

Basic Usage

The Floom SDK for Python provides various functionalities to integrate Generative AI capabilities into your applications effortlessly. Here is an overview of the options available in the examples/example.py file, illustrating the SDK's versatility in handling different types of AI tasks:

Running a Simple Pipeline

Send a textual prompt to a predefined pipeline and receive a textual response. Ideal for straightforward AI interactions:

from floom import FloomClient  
  
def run_simple_pipeline():  
	floom_client = FloomClient(
		url="http://localhost:4050",
		api_key="Vcm6RReBMLQa0h2fUidOC7SLmh356uHH")
			
	floom_response = floom_client.run(
		pipelineId="floom-simple-pipeline",
		prompt="List top 10 countries in the world.")

	 if floom_response.success is False:
		 print(f"Error: {floom_response.error.detail}")
	 else:
		 for content in floom_response.values:
			 print(f"Response: ", content.value)
			 

Running a Pipeline with Object Response

Interpret the response as a list of objects, each representing structured information. Useful for applications requiring object-oriented responses. Note the @dataclass annotation for the response object class:

from floom import FloomClient
from dataclasses import dataclass
from typing import List

@dataclass
class Country:
    name: str = ""
    population: str = ""
    continent: str = ""

def run_object_response_pipeline():
    floom_client = FloomClient(
        url="http://localhost:4050",
        api_key="Vcm6RReBMLQa0h2fUidOC7SLmh356uHH")
    
    floom_response = floom_client.run(
        pipelineId="floom-object-response-pipeline",
        prompt="List top 10 countries in the world, sorted by population.",
        response_type=List[Country])

    if floom_response.success is False:
        print(f"Error: {floom_response.error.detail}")
    else:
        content = floom_response.values[0].value
        for country in content:
            print(f"Country: {country.name}, Population: {country.population}, Continent: {country.continent}")

Running a RAG Pipeline

Generate detailed, context-aware responses based on a corpus of documents:

`from floom import FloomClient

def run_rag_pipeline():
    floom_client = FloomClient(
        url="http://localhost:4050",
        api_key="Vcm6RReBMLQa0h2fUidOC7SLmh356uHH")
    
    floom_response = floom_client.run(
        pipelineId="floom-docs-pipeline",
        prompt="Question about your document.")

    if floom_response.success is False:
        print(f"Error: {floom_response.error.detail}")
    else:
        for content in floom_response.values:
            print(f"Response: ", content.value)`

Running a Text to image Pipeline

Convert textual descriptions into images. Demonstrates decoding the base64 encoded image and displaying it:

from floom import FloomClient
import base64
import webbrowser
import os

def run_text_to_image_pipeline():
    floom_client = FloomClient(
        url="http://localhost:4050",
        api_key="Vcm6RReBMLQa0h2fUidOC7SLmh356uHH")
            
    floom_response = floom_client.run(
        pipelineId="floom-image-generator-pipeline",
        prompt="Group of people drinking coffee while camping in the ice cold forest, surrounded by snow.")

    if floom_response.success is False:
        print(f"Error: {floom_response.error.detail}")
    else:
        image_base_64 = floom_response.values[0].value
        print('Image preview:', image_base_64[:100], '...')

        image_path = "image.png"
        with open(image_path, "wb") as fh:
            fh.write(base64.b64decode(image_base_64))

        webbrowser.open(f"file://{os.path.abspath(image_path)}")

Running a Text to Speech Pipeline

This functionality converts text into speech, showcasing how to handle and play the generated audio content:

from floom import FloomClient
import base64
import webbrowser
import os

def run_text_to_speech_pipeline():
    floom_client = FloomClient(
        url="http://localhost:4050",
        api_key="Vcm6RReBMLQa0h2fUidOC7SLmh356uHH")
    
    floom_response = floom_client.run(
        pipelineId="floom-tts-pipeline",
        prompt="Why do birds suddenly appear every time you are near? Just like me, they long to be close to you.")

    if floom_response.success is False:
        print(f"Error: {floom_response.error.detail}")
    else:
        audio_base_64 = floom_response.values[0].value
        print('Audio preview:', audio_base_64[:100], '...')

        audio_path = "audio.mp3"
        with open(audio_path, "wb") as fh:
            fh.write(base64.b64decode(audio_base_64))

        webbrowser.open(f"file://{os.path.abspath(audio_path)}")

Documentation

For comprehensive documentation and API references, please visit Floom Documentation.

Support and Contributions

Should you encounter any issues or if you need further assistance, feel free to open an issue on our GitHub issues page.

We warmly welcome contributions! If you're interested in contributing, please fork the repository, make your changes, and submit a pull request.

License

FloomSDK for Python is distributed under the MIT license.