Skip to content

Streamlining the ML pipeline, from brainstorming and exploratory research to model development, evaluation and explanation.

Notifications You must be signed in to change notification settings

punitarani/AgentML

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

37 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

AgentML

AgentML is a groundbreaking LLM Agent tool designed to revolutionize the machine learning (ML) landscape.

Offering simplicity and power, it's suited for a diverse audience, from students to ML professionals and non-coders. AgentML streamlines ML processes for all types of datasets, embodying innovation in artificial intelligence.

Demo

The following demo videos showcase AgentML's capabilities in both Supervised and Autonomous modes.

AgentML is capable of handling datasets, writing code and also training and evaluating machine learning models.

AgentML Highlight

Supervised Mode Video

Human-in-the-loop mode to train a classifier on the Iris dataset.

AgentML Supervised Demo

Autonomous Mode Video

Autonomous mode with validator pseudo-agent to replace human validation.

AgentML Autonomous Demo

Video is sped up for brevity.

Key Features

Dual Capability in ML Workflow

  • Small Datasets:
    • Data Analysis: Analyzes dataset characteristics to grasp the problem statement.
    • Pipeline Creation: Develops a comprehensive pipeline, including data preprocessing, model building, and evaluation.
    • Iterative Refinement: Continuously refines the model for optimal performance.
  • Complex Datasets:
    • Advanced Data Handling: Manages intricate datasets with enhanced preprocessing and analysis.
    • Custom Model Building: Constructs tailored models to address complex problems.
    • In-depth Evaluation: Provides thorough model evaluation and tuning for complex datasets.

User Interaction and Customization

  • Code Execution:
    • Secure Environment: Executes Python code in a sandbox for safety and integrity.
    • Interactive Development: Allows real-time code writing and testing.
  • Template Utilization:
    • Customizability: Imports and builds on user-provided code templates.
    • Flexibility: Adapts to various coding styles and requirements.

Vision Agent

  • Advanced Data Interpretation:
    • Visual Data Analysis: Interprets charts and graphs for deeper insights.
    • Enhanced Perception: Supplements language models with visual understanding.

Technical Architecture

Manager Agent

  • Role:
    • Central Coordination: Manages inputs and oversees other agents' operations.
    • Goal-Oriented Management: Focuses on achieving user-defined goals.
  • Function:
    • Integration: Ensures cohesive operation of all agents.
    • Efficiency: Optimizes resource allocation and task execution.

Planner Agent

  • Responsibility:
    • Task Analysis: Breaks down goals into manageable tasks.
    • Efficient Delegation: Allocates tasks to the appropriate agents.
  • Strategic Planning:
    • Methodical Approach: Ensures systematic problem-solving.
    • Goal Alignment: Aligns tasks with the overarching objectives.

Coder Agent

  • Capabilities:
    • Diverse Coding Tasks: Handles writing, modifying, and debugging code.
    • Model Development: Builds and visualizes ML models.
  • Automation:
    • Recursive Execution: Ensures thorough and accurate completion of coding tasks.
    • Error Handling: Detects and resolves errors in code to ensure smooth execution.

Vision Agent

  • Functionality:
    • Data Visualization Analysis: Interprets visual data for comprehensive insights.
    • Cross-Model Integration: Combines visual analysis with other model outputs.

Validator (Pseudo-Agent)

  • In Autonomous Mode (LLM Agent):
    • Step Validation: Checks each step for alignment with goals.
    • Consistency Assurance: Ensures uniformity in autonomous operations.
  • In Supervised Mode (Human-in-the-loop):
    • User Oversight: Allows users to guide and validate the process.
    • Customization and Control: Enables user-driven adjustments and decision-making.

User Accessibility and Interface

  • User-Friendly Design:
    • Ease of Use: Intuitive interface for a wide range of users.
    • Guided Interaction: Simplifies ML concepts for easy understanding.
  • Inclusivity:
    • Non-Coder Friendly: Accessible to individuals without coding background.
    • Professional Versatility: Meets the needs of experienced ML practitioners.

Web Application for Deployment

  • Easy Access:
    • Simple Interface: Streamlined web app for easy AgentML deployment.
    • Instant Setup: Quick start-up with minimal configuration.
  • Broad Accessibility:
    • Universal Reach: Available to a global audience.
    • No Expertise Required: User-friendly for all skill levels.

AgentML is more than a tool—it's a transformative force in machine learning, offering ease, efficiency, and advanced insights. It exemplifies the democratizing power of artificial intelligence, making sophisticated ML processes accessible to everyone.

Running the Application

AgentML can be operated in two modes: Supervised and Autonomous. Follow these simple steps to get started:

  1. Clone the Repository:

    git clone https://github.com/punitarani/AgentML.git
    cd AgentML
  2. Install Dependencies using Poetry:

    poetry install

    Poetry should automatically create a virtual environment for you. If it doesn't, you can initiate one manually:

    poetry shell
  3. Setup Environment Variables:

    • Copy the .config/.env.template to .env in the root directory.
    • Fill out the necessary environment variables in the .env file.

Supervised Mode

python -m streamlit run app.py

Autonomous Mode

python -m streamlit run auto.py

Punit Arani

About

Streamlining the ML pipeline, from brainstorming and exploratory research to model development, evaluation and explanation.

Topics

Resources

Stars

Watchers

Forks