Skip to main content

Installation

Requirements​

PhyTorch requires Python 3.7 or later. All dependencies (PyTorch, NumPy, SciPy, Pandas, Matplotlib) will be automatically installed.

pip install phytorch-lib

Using with Conda Environments​

If you use conda for environment management:

conda create -n phytorch python=3.9
conda activate phytorch
pip install phytorch-lib

Installation from Source​

For development or to access the latest features, you can install from source:

git clone https://github.com/PlantSimulationLab/phytorch.git
cd phytorch
pip install -e .

GPU Support​

PhyTorch leverages PyTorch for GPU acceleration. To enable GPU support, ensure you have a CUDA-compatible GPU and install the appropriate PyTorch version:

# For CUDA 11.8
pip install torch --index-url https://download.pytorch.org/whl/cu118

# For CUDA 12.1
pip install torch --index-url https://download.pytorch.org/whl/cu121

See the PyTorch installation guide for more options.

Verify Installation​

You can verify your installation by importing PhyTorch:

from phytorch import *
print("PhyTorch installed successfully!")

Troubleshooting​

GPU Not Detected​

If PyTorch is not detecting your GPU:

import torch
print(torch.cuda.is_available())  # Should return True if GPU is available

If this returns False, you may need to reinstall PyTorch with CUDA support.

Next Steps​

Once installed, check out the Getting Started Guide to fit your first model!