Skip to content

GettingStarted.md

Latest commit

 

History

History
75 lines (52 loc) · 2.82 KB

GettingStarted.md

File metadata and controls

75 lines (52 loc) · 2.82 KB

Getting Started

This guide covers the technical installation, environment setup, and the basic workflow to get your first project up and running.

1. Prerequisites

Because this suite relies on heavy machine learning libraries (PyTorch, Anomalib, SAM 2), we highly recommend running it inside an isolated environment.

  • OS: macOS, Linux, or Windows.
  • Python: Version 3.10 is highly recommended.
  • Environment Manager: Anaconda or Miniconda.

2. Installation

First, clone the repository and navigate into it:

git clone https://github.com/annomate-mircrosentryai-workflow
cd annomate-microsentryai-workflow

Create and activate a clean Conda environment:

conda create -n annomate python=3.10
conda activate annomate

Install the required dependencies. We have provided specific environment files based on your system's hardware capabilities. Run one of the following commands:

For Windows/Linux with NVIDIA GPUs (CUDA):

conda env update --file environment-cuda.yml --prune

For macOS (Apple Silicon):

conda env update --file environment-mac.yml --prune

For CPU Only (No hardware acceleration):

conda env update --file environment-cpu.yml --prune

Note for developers: If you plan on contributing, please run pre-commit install to enable automated linting.

Running the Application

Ensure your conda environment is activated, then run the main Python script from the root directory:

python src/main.py

Building a Standalone Executable (PyInstaller)

If you want to distribute the application to users who do not have Python or Conda installed, you can compile AnnoMate & MicroSentryAI into a standalone executable using PyInstaller. (Note: PyInstaller is already included in your environment.yml dependencies).

From the root directory of the project, run the following command.

For Windows:

pyinstaller --name "AnnoMate" --windowed --add-data "logos;logos" src/main.py

For MacOS/Linux:

pyinstaller --name "AnnoMate" --windowed --add-data "logos:logos" src/main.py

Important Build Notes:

  • File Size: Because the application bundles PySide6, PyTorch, and Anomalib, the resulting build folder (dist/AnnoMate/) will be quite large (often several gigabytes).

  • Console Flag: The --windowed (or --noconsole) flag hides the terminal window in the final build. If your compiled app crashes on startup, try removing the --windowed flag and rebuilding; this will allow you to see the terminal output and identify any missing hidden imports required by PyTorch or Anomalib.

  • SAM 2 Weights: The SAM 2 model downloads its checkpoint weights to a local sam_weights/ folder upon first use. You may need to manually copy this folder into the final dist/AnnoMate/ directory if you want it pre-packaged for offline users.