Introduction to AI on HiPerGator¶
Introduction¶
If you are a new HiPerGator user make sure that you first read the general Getting Started documentation to get yourself familiar with HiPerGator and take the New User Training with step-by-step instructions on how to use HiPerGator.
AI Education and Training provides learning materials and training videos on various AI topics. JupyterHub and Jupyter Notebooks on HiPerGator are popular platforms for developing and running AI programs.
AI Computing Resources: HiPerGator and HiPerGator AI provide world class computing infrastructure for conducting AI research and building AI applications. To purchase resource allocations, please refer to Purchase Request. For computing server features and GPU access, please visit Available Node Features and GPU Access.
AI Software¶
General AI Frameworks¶
AI frameworks provide building blocks for designing and training machine learning and deep learning models. The following AI frameworks are available on HiPerGator:
-
PyTorch: A deep learning framework developed by Facebook AI Research Lab, supporting Python, Java, and C++. It allows training on both GPU and CPUs, as well as distributed training and multi-GPU models.
-
TensorFlow: An open-source AI framework developed by the Google Brain team. Keras is a neural network library that runs on top of TensorFlow, serving as a high-level interface in TensorFlow 2.0. TensorFlow supports both GPU and CPUs, along with multi-GPU and distributed training.
-
Tensorboard: A visualization tool for monitoring neural network training.
-
Sci-kit learn: A Python library for machine learning and statistical modeling, available in many Python modules on HiPerGator.
-
Matlab: Offers toolboxes for machine learning, deep learning, computer vision, and automatic driving, supported on both CPUs and GPUs.
-
RAPIDS AI: A suite of CUDA-enabled open-source software libraries by NVIDIA, enabling multi-node, multi-GPU deployments for faster processing of large datasets.
-
Fastai: Simplifies training fast and accurate neural nets using best practices and can be used via Google Colab.
-
Triton: Nvidia Triton Inference Server provides a cloud and edge inferencing solution optimized for both CPUs and GPUs. Triton supports an HTTP/REST and GRPC protocol that allows remote clients to request inferencing for any model being managed by the server.
AI Software for Specific Domains¶
HiPerGator offers comprehensive software for big data and AI research:
-
Data Science Platform: Combines mathematics, programming, analytics, AI, and machine learning to uncover insights from large data sets.
-
Natural Language Processing: Includes tools such as Nemo, BioNemo, Llama2, Llama3, Mistral AI, and Gemma LLMs, helping computers interpret human language for applications like chatbots and translation.
-
Computer Vision: Tools for image and video processing on HiPerGator.
-
Healthcare and Life Sciences: Software for medical image analysis, disease detection, and genomic data interpretation, including AlphaFold, BioNemo, Clara Parabricks, and MONAI.
-
Simulation and Modeling: HiPerGator supports advanced simulation tools, such as NVIDIA Modulus and Omniverse, for real-world scenario predictions.
AI Reference Datasets¶
A variety of reference machine learning and AI datasets are located in /data/ai/ref-data
. Browse
the catalog of all available AI datasets to learn more.
AI Examples¶
A suite of short examples are provided on using the software stack on HiPerGator to do different
tasks in imaging and NLP, located in /data/ai/examples
. Browse the
catalog of all available AI examples to learn more.
AI Benchmarks and Models¶
Several commonly used benchmarks and models for NLP are provided on
HiPerGator at /data/ai/benchmarks
and /data/ai/models
. Browse the
catalog of all available AI models to learn more.
AI Zoom Support¶
You can also schedule remote support via a Zoom meeting with one of our knowledgeable RC support staff members at the Remote Support Page.