apple-m1-mlops

Apple M1-specific setup for machine learning engineering production systems, focusing on MLOps practices with Metal acceleration and Jupyter

For detailed acknowledgments, please see the MLOps. This project is part of deeplearning.ai's Machine Learning Engineering for Production Specialization.

Clone the DeepLearning.ai repository (optional)

git clone https://github.com/https-deeplearning-ai/machine-learning-engineering-for-production-public.git

Start Here

1. Environment Setup

conda create --name mlep-w1-lab python=3.8
conda activate mlep-w1-lab

General Setup - no Apple M1 (optional)

pip install -r requirements.txt

Apple M1 Users

# Add conda-forge to channels
conda config --add channels conda-forge
# Install TensorFlow dependencies
conda install -c apple tensorflow-deps
# Install TensorFlow for macOS
python -m pip install tensorflow-macos tensorflow-metal
# Install specific OpenCV version
conda install -c conda-forge opencv===4.5.3
# Modify requirements for M1 compatibility
sed -i '.bak' \
    -e 's/^tensorflow==2.7.0/# tensorflow==2.7.0  # Use tensorflow-macos and tensorflow-metal instead/' \
    -e 's/^opencv-python-headless==4.5.3.56/# opencv-python-headless==4.5.3.56  # Commented out for compatibility with M1 Mac/' \
    requirements.txt
# Rename requirements for clarity
mv requirements.txt requirements-m1.txt
# Update protobuf version for compatibility
sed -i '.bak' 's/^protobuf==.*/protobuf>=3.20.3,<4.21.0/' requirements-m1.txt
# Install modified requirements
pip install -r requirements-m1.txt

Compatabilit note:

Updates protobuf for tensorflow-macos 2.13.0 requirements.
Disables tensorflow==2.7.0 for M1 Mac compatibility.
Replaces opencv-python-headless with opencv==4.5.3 for M1.

2. Verify Installations

python -c "import tensorflow as tf; print('TensorFlow Version:', tf.__version__)"
python -c "import cv2; print('OpenCV Version:', cv2.__version__)"

3. Launch Jupyter Lab

jupyter lab

4. Access Jupyter Lab and Run the Notebook

i. Copy the URL from the terminal and paste it into your browser to open the server.ipynb notebook.

http://localhost:8888/lab

5. Run Installation Verification Tests

Run the test script:

python src/test_installation.py

Test installtions note:

TensorFlow Version: Verifies TensorFlow version starts with '2'.
TensorFlow Operation: Confirms GPU/Metal support via matrix multiplication.
OpenCV Version: Ensures OpenCV version is '4.5.3'.
OpenCV Operation: Validates OpenCV can process images.
Environment Requirements: Checks for dependency conflicts.

patmejia/apple-m1-mlops