Machine Learning

Prepare thoroughly on what is machine learning and how to train machine learning models. High level understand on concepts related to MLOps. To build & train high accuracy machine learning models, you need high quality data. Data comes from multiple sources and is error prone. The process and steps to take this data, make it machine learning ready and then train the models is called MLops. A typical ML pipeline consists of the following steps. You need to make sure you are familiar this pipeline, generally the tools used and more importantly the concepts & techniques in each step.

• Data Collection: Collect raw data from databases, sensors, APIs, etc.
• Data Preparation: Clean data (remove errors, handle missing values), feature engineering (e.g., date → day/month, text → tokens), tools: cuDF, Spark-RAPIDS, etc.
• Model Training: Familiarity with different types of models; use ML models via cuML and DL models via DGL, PyG.
• Model Deployment: Make the trained model available via an API or server; tool: NVIDIA Triton.
• Inference (Prediction): Use the deployed model to make predictions on new data in real time or batches.

References

• Introduction to Machine Learning
• Linear Regression
• Logistic Regression
• Classification
• Machine Learning - go thru videos 1 to 21 in this playlist
• Working with Numerical Data
• Working with Categorical Data
• Datasets, Generalization, and Overfitting
• Accelerating End-to-End Data Science Workflows (NVIDIA course)

Deep Learning and Neural Networks

• Understand the motivation for building neural networks, and the use cases they address.
• Understand the key components of a deep neural network architecture: like Nodes, Hidden layers, Activation functions, Loss Functions.
• High-level understanding of how neural networks are trained using backpropagation and loss functions.
• Understand how they are deployed for inference and how NVIDIA products are used.
• Understand use cases of neural networks vs machine learning algorithms.
• CNN and GAN (Generative Adversarial Network).

References

• Deep Learning Demystified (Nvidia on-demand)
• Getting Started with Deep Learning (Nvidia self-paced)

Transformer Architecture and NLP

This is very important. You need to have a complete high level understanding of the Transformer architecture Needless to stay, since this certification is about GenAI you need to have a intuition level understanding of NLP. There will be mentions of Word2Vec, RNNs, LSTM, etc and it can get confusing and overwhelming pretty quickly. The way I wrapped my head around this is to understand the history of NLP and how it lead to transformer based NLP.

• RNNs
• Word2Vec
• LLM Benchmarks
• Layer Normalization

References

• Introduction to Transformer-Based Natural Language Processing (Self-Paced)
• Building Transformer-Based Natural Language Processing Applications (Instructor-Led)
• Building World-Class NLP Models with Transformers and Hugging Face | Grandmaster Series E4 (YouTube)

Building Applications Using Large Language Models

This section summarizes essential concepts for building GenAI applications—aligned with topics covered in the NVIDIA GenAI certification. You'll understand key techniques like prompt engineering, RAG, and LLM customization, and learn when to use each for efficient, real-world AI deployment.

• Prompt Engineering
  Designing effective prompts to guide pre-trained LLMs without additional training.
• RAG (Retrieval-Augmented Generation)
  Enhances LLMs by injecting external knowledge at inference time.
• LLM Customization
  Adapting models using fine-tuning or parameter-efficient methods like LoRA.

References

• How to Get Better Outputs from Your Large Language Model (NVIDIA Blog)
• Selecting Large Language Model Customization Techniques (NVIDIA Blog)
• Rapid Application Development with Large Language Models (LLMs) – NVIDIA Course
• Building LLM Applications With Prompt Engineering – NVIDIA Course