ml-pythondeps

Version:

4.3.9, 4.7.8, 4.8.1

Category:

ai

Cluster:

Loki

Author / Distributor

Provided by system toolchain for machine learning stack (typically bundled via EasyBuild or ML frameworks like PyTorch/TensorFlow)

Description

ml-pythondeps is a bundled environment providing essential Python dependencies required for machine learning and data science workflows. It includes commonly used Python libraries, CUDA bindings, and support for GPU-accelerated computation.

This module is designed to support frameworks like TensorFlow, PyTorch, Keras, and other Python-based ML tools by pre-installing GPU-capable packages and libraries.

Documentation

This module does not provide a standalone CLI binary.

Use via Python after loading the module:

    $ module load ai/ml-pythondeps-py39-cuda11.2-gcc9/4.8.1
    $ python

    >>> import numpy
    >>> import pandas
    >>> import sklearn
    >>> import torch  # if loaded with pytorch module

To list installed packages:

    $ python -m pip list

To check CUDA availability (e.g. with PyTorch):

    >>> import torch
    >>> torch.cuda.is_available()

Examples/Usage

• Load the module:

  $ module load ai/ml-pythondeps-py39-cuda11.2-gcc9/4.8.1
  

• Unload the module:

  $ module unload ai/ml-pythondeps-py39-cuda11.2-gcc9/4.8.1
  

• Start Python and test:

  $ python
  >>> import numpy as np
  >>> np.array([1, 2, 3])
  

• Check installed ML stack:

  $ python -m pip freeze | grep -i torch
  

Installation

Provided via EasyBuild toolchain for machine learning environments. Documentation reference: https://easybuild.io/