Certification > AI & ML > PTCA Certification Pathway

CERTIFICATION PATHWAY

Build practical PyTorch skills through a structured series of learning modules aligned with the PTCA certification domains and competencies.

Get everything you need in one pathway. Includes all learning modules and the PyTorch Certified Associate (PTCA) exam. Develop your skills with hands-on learning, additional resources, and real-world experience as you prepare for certification.

Build Your Skills

Develop practical PyTorch skills through focused learning modules covering the PTCA certification domains and competencies.

read less read more

Follow Your Pathway

Progress from PyTorch fundamentals and data handling to model development and performance optimization through a structured learning experience.

read less read more

Validate Your Knowledge

Put your learning into context with additional resources and real-world experience, then validate your knowledge with the PTCA certification exam.

read less read more

Pathway Outline

Expand All

Collapse All

PyTorch Fundamentals (LFD2000)

Course Introduction
PyTorch and the Problem Ahead
Tensors: The Data You Work With
Devices: Where Computation Happens
The Model Lifecycle: From Training to Prediction

Data Handling in PyTorch (LFD2001)

Course Introduction
Where the Data Comes From: Datasets in PyTorch
Moving Data to the Model: DataLoaders, Batching, and Shuffling
Preparing Data Right: Transforms and Data Quality

PyTorch Model Development (LFD2002)

Course Introduction
The Building Blocks: Layers and Activations
From Loss to Update: How Training Actually Works
Making Choices: Modifying and Improving a Model

Optimizing PyTorch (LFD2003)

Course Introduction
Faster Execution: torch.compile
Fitting More with Less: Automatic Mixed Precision
Finding the Bottleneck: The PyTorch Profiler
Scaling Beyond One Device: Distributed Training Fundamentals

Lab Info

All exercises run on CPU, so no GPU is required, but GPU can be optionally used where applicable.
A modern web browser (Chrome, Firefox, or Edge)
A Google account to access for Google Colab is recommended, but a local Python environment with Jupyter Notebook works too.
If using Google Colab:
- Each course provides installation instructions for prerequisites including PyTorch and Python 3.12 or later.
If using local Python Environment, installation prerequisites include:
- Python 3.12 or later
- PyTorch
Estimated cost to the learner: zero

Domains & Competencies

Expand All

Collapse All

PyTorch Fundamentals38%

Core Concepts
Tensors
Training, Testing, and Using Models
Device Basics (CPU, CUDA, MPS, etc)

Model Development20%

PyTorch Neural Network (NN) Building Blocks

Performance & Optimization26%

Precision and Execution Optimization
Performance Measurement
Distributed Training

Data Handling16%

Datasets
DataLoaders
Transforms
Training Data

Exam Details & Resources

This exam is an online, proctored, multiple-choice exam.

About the PTCA

Once you pass the exam, be sure to add your new badge to your LinkedIn, LFX, GitHub & other profiles to help you better connect with the open source community and distinguish yourself as certified for prospective employers.

Prerequisites

Knowledge/Skills Prerequisites (All courses)

Python: comfortable with variables, data types, loops, conditionals, functions, and importing libraries.
Mathematics: an introductory grasp of linear algebra, meaning what vectors and matrices are, plus basic probability and statistics.
Machine learning: conceptual awareness of supervised versus unsupervised learning, model development life cycle, and overfitting.
Development environment: able to install packages with pip, run cells in Jupyter Notebook or Google Colab, and navigate a file system.

Exam Prerequisites
There are no pre-requisites for this exam.