MAP-ACTION-MODEL

Map your actions with precision and value!

Developed with the software and tools below.

---

Quick Links

• Overview
• Features
• Repository Structure
• Modules
• Getting Started
  • Installation
  • Running Map-Action-Model
  • Tests
• Contributing
• License
• Acknowledgments

---

Overview

Map Action Model is the codebase for the continous training of Map Acion computer vision model. Developper Doc

---

Features

	Feature	Description
⚙️	Architecture	The project follow a modular architecture with dependencies on libraries like Brotli, Pillow, and FastAPI. It utilizes a mix of Python and related technologies for development.
🔩	Code Quality	The code quality is maintained with the use of tools such as MkDocs for documentation and potentially other linting tools based on the repository contents.
📄	Documentation	The project includes MKDocs for documentation generation, providing extensive and structured documentation for the codebase.
🔌	Integrations	Key integrations include Brotli, tqdm, and FastAPI among others, suggesting a reliance on external libraries and tools for functionality.
🧩	Modularity	The project is structured in a modular way, with dependencies on various libraries like Torch, torchvision, and others, indicating potential code reusability.
🧪	Testing	Testing frameworks is PyTest
📦	Dependencies	Key external libraries and dependencies include Brotli, Pillow, FastAPI, Torch, torchvision, and others, indicating a reliance on diverse libraries for functionality.

---

Repository Structure

└── Map-Action-Model/
    ├── .github
    │   └── workflows
    │       ├── deploy-docs.yml
    │       ├── training-on-gpu.yml
    │       ├── unittesting.yml
    │       └── zenml_action.yml
    ├── Dockerfile._cuda
    ├── Dockerfile.fastapi
    ├── LICENCE
    ├── _cd.yml
    ├── _ci.yml
    ├── code
    │   ├── .zen
    │   │   └── config.yaml
    │   ├── TFLearning.ipynb
    │   ├── pipelines
    │   │   └── zenml_pipeline.py
    │   ├── steps
    │   │   ├── dagshub_utils
    │   │   ├── data_preprocess
    │   │   ├── model
    │   │   ├── model_eval
    │   │   ├── plot_metrics
    │   │   └── training_step
    │   ├── utilities.ipynb
    │   └── zenml_running.py
    ├── data.dvc
    ├── data_upload.py
    ├── ma_env
    │   ├── bin
    │   │   ├── Activate.ps1
    │   │   ├── activate
    │   │   ├── activate.csh
    │   │   ├── activate.fish
    │   │   ├── pip
    │   │   ├── pip3
    │   │   ├── pip3.10
    │   │   ├── python
    │   │   ├── python3
    │   │   └── python3.10
    │   ├── lib
    │   │   └── python3.10
    │   └── lib64
    │       └── python3.10
    ├── model
    │   └── main.py
    ├── requirements.txt
    ├── services
    │   └── unittesting
    │       └── Dockerfile
    └── zenml_config
        └── zenml_conf.yml

---

Modules

code

File	Summary
utilities.ipynb	Code snippet in code/utilities.ipynb:**Interacts with MLflow and DagsHub to manage experiment tracking within Map-Action-Model repository structure. Handles data sources and enables dataset manipulations.
TFLearning.ipynb	Code snippet: Validates user input and updates database accordingly.Architecture: Microservices architecture with a separate service for database operations.Role: Ensures data integrity and security in the system.Critical features: Input validation, database interaction, seamless integration within the microservices ecosystem.
zenml_running.py	zenml_running.pyinMap-Action-Modelrepo orchestrates a training pipeline using ZenML. Central to managing ML workflows, integrated withpipelines/zenml_pipeline.py`.

code..zen

File	Summary
config.yaml	Code in code/.zen/config.yaml sets active stack and workspace IDs for the repository. Facilitates seamless integration with ZenML for workflow management and model pipelines.

code.steps.model

File	Summary
m_a_model.py	Code snippet in m_a_model.py creates a modified VGG16 model for a specific class count. It adjusts the classifier and uses CrossEntropyLoss. This step enhances the model's adaptability and loss computation in the repository's ML pipeline architecture.

code.steps.plot_metrics

File	Summary
plot_metrics.py	Code Summary:**plot_metrics.py in Map-Action-Model repo visualizes training and test loss/accuracy curves using Matplotlib. Enhances model evaluation insights for ML pipelines.

code.steps.dagshub_utils

File	Summary
dagshub_data_load.py	Code snippet in dagshub_data_load.py downloads and organizes data from a CSV file and DagsHub repository for machine learning model training in the Map-Action-Model repository architecture.

code.steps.model_eval

File	Summary
evaluation.py	Code Summary:**This code snippet performs testing for a PyTorch model, evaluating test data and logging metrics with MLFlow. It optimizes model performance and accuracy for the parent repository's machine learning pipeline.

code.steps.training_step

File	Summary
training_step.py	Code snippet in training_step.py trains PyTorch model with provided data, logging metrics using MLFlow. Key features include model training loop, metric tracking, and PyTorch model saving.

code.steps.data_preprocess

File	Summary
data_loading_pipeline.py	Code Summary**:data_loading_pipeline.py in Map-Action-Model creates PyTorch data loaders for training and testing datasets, managing dataset loading and transformation for ML pipelines.
data_transform.py	Role:** Code snippet in data_transform.py for image preprocessing in Map-Action-Model repo architecture.**Achievement:** Generates image transformations for training/testing using torchvision API elegantly, ensuring data consistency.

code.pipelines

File	Summary
zenml_pipeline.py	Code snippet in zenml_pipeline.py orchestrates a machine learning training pipeline. It manages data processing, model training, and evaluation, culminating in loss curves plotting. This integral component advances ML model development in the repository architecture.

---

Getting Started

Requirements

Ensure you have the following dependencies installed on your system:

• Python: Python 3.x

Installation

1. Clone the Map-Action-Model repository:

git clone https://github.com/223MapAction/Map-Action-Model.git

2. Change to the project directory:

cd Map-Action-Model

3. Install the dependencies:

pip install -r requirements.txt

Running Map-Action-Model

Use the following command to run Map-Action-Model:

python main.py

Tests

To execute tests, run:

pytest

---

🤝 Contributing

Contributions are welcome! Here are several ways you can contribute:

See our Contribution Guidelines for details on how to contribute.

---

📄 License

This project is protected under the GNU GPLv3 License.

---

Code of conduct

See our Code of Conduct for details on expected behavior in our community.

---

👏 Acknowledgments

• List any resources, contributors, inspiration, etc. here.

Return

---