Course Overview:
Welcome to the AWS MLOPS & LLMOps Training in Chennai, an immersive learning experience designed for individuals looking to master machine learning operations (MLOps) and large language model operations (LLMops) within the Amazon Web Services (AWS) ecosystem. This comprehensive course provides a deep dive into the tools, practices, and methodologies necessary to efficiently deploy, monitor, and manage machine learning and large language models in production at scale.
Course Objectives:
By the end of this course, participants will:
- Understand the core principles and practices of MLOps and LLMops, including the importance of automation, monitoring, and governance in the ML lifecycle.
- Gain proficiency with AWS services and tools relevant to MLOps and LLMops, such as AWS SageMaker, AWS Lambda, Amazon CloudWatch, and AWS IAM.
- Learn to design, implement, and manage robust, scalable, and secure ML and LLM pipelines on AWS.
- Acquire hands-on experience through practical labs and projects, applying best practices for versioning, testing, and deploying ML and LLM models.
- Develop skills to monitor and optimize model performance, manage data and model drift, and ensure compliance with industry standards and regulations.
-
Infrastructure and Resource Management
-
Monitoring and Logging
-
Model Versioning and Experiment Tracking
-
Security and Compliance
-
Introduction
-
MLOps Fundamentals
-
AWS and MLOps
-
AWS Specific Tools and Configurations
-
DevOps Lifecycle Tools in AWS
-
Creating and Configuring an AWS Account
-
Security Setup: MFA, IAM Accounts, and Policies
-
Introduction to S3 Buckets and EC2 Instances
-
AWS Specific Tools and Configurations
-
Creation of S3 Bucket from Console
-
26. Creation of S3 Bucket from CLI
-
Version Enablement in S3
-
Introduction EC2 instances
-
Launch EC2 instance & SSH into EC2 Instances
-
Housekeeping Activity
-
-
Linux and Bash for MLOps
-
Core Concepts
-
CI/CD Pipeline Introduction
-
Getting Started with AWS CodeCommit & Distributed Version Control Systems (DVCS)
-
Initial Configuration & Basic Git Commands
-
Setting Up Your Git Workspace
-
Understanding Git Workflow
-
Adding Files to the Staging Area & Understanding Staged Differences
-
Unstaging, Resetting, and Reverting Changes in Git
-
Working with AWS CodeCommit: Remote Commands, Security, and Integrations
-
Cloning, Branching, and Handling Git Branches: Hands-On Parts 1 & 2
-
Resolving Git Conflicts, Rebasing vs. Merging, and Using Git Stash
-
-
Deployment & Security
-
Introduction to AWS CodeDeploy & YAML
-
First Steps with AWS CodeDeploy: Hands-On Introduction and Deep Dive
-
Exploring AWS CodePipeline: Creation and Automation with Manual Approval
-
Introduction to Docker: Basics & Installation
-
Pull the image from Docker Desktop
-
Dockerfile
-
Push the Docker Image to ECR
-
Hands on – Amazon ECR for AWS CodeBuild
-
-
Amazon SageMaker & Feature Engineering
-
Why Amazon SageMaker is Preferred for Machine Learning Workflows
-
Domain Creation, Studio Setup, and Clean-Up Activities in SageMaker
-
Feature Engineering Essentials, Data Wrangler Setup, and Transformation Techniques
-
Data Quality and Insights Report
-
Univariate Analysis & Bias Report
-
Target Leakage
-
Data Transformation
-
Data Transformation – Custom Script
-
Export to S3
-
Feature Engineering on Sagemaker Notebook Instance
-
Summary
-
-
Advanced Concepts
-
Building and Managing MLOps Pipelines
-
Packaging, Deployment, and Kubernetes
-
LLMOps
-
Fundamentals of Large Language Models (LLMs)
-
Development Environment Setup
-
Data Preparation and Preprocessing
-
Model Training and Evaluation
-
Continuous Integration (CI) for LLM Development
-
Continuous Deployment (CD) for LLM Models
-
Use Cases and Applications
-
Question Answering
-
Building an LLM-based question answering system
-
Deploying the question answering model using CI/CD
-
Text Generation
-
Developing an LLM-based text generation application
-
Automating the deployment of the text generation model
-
Sentiment Analysis
-
Implementing an LLM-based sentiment analysis pipeline
-
Integrating the sentiment analysis model into a CI/CD workflow
-
Named Entity Recognition (NER)
-
Creating an LLM-based NER system
-
Setting up a CI/CD pipeline for NER model deployment
-
-
Capstone Projects
-
Knowledge Base Assistant
-
Building an LLM-powered knowledge base assistant
-
Implementing CI/CD for the assistant's backend and frontend
-
Creative Writing Tool
-
Developing an LLM-based creative writing tool
-
Automating the deployment and scaling of the writing model
-
Customer Support Chatbot
-
Creating an LLM-based customer support chatbot
-
Setting up a CI/CD pipeline for chatbot training and deployment
-
Content Generation Platform
-
Building an LLM-powered content generation platform
-
Implementing CI/CD for the platform's backend and frontend
-
-
Best Practices and Future Trends