Artificial Intelligence (AI) Free Advance Course Batch 02

Stage 0: Orientation

Introduction to AI and ML

•  What it AI?
• What is future going to be like?
• Intro to Machine Learning
• Types of Machine Learning
  1.  Supervised learning
  2. Unsupervised Learning
  3. Reinforcement Learning

Some Demos and interesting videos on it.

• Object detection
• Segmentation
• Classification
• Generative models
• Chat GPT
• Dall-e
• stable-diffusion

Our Goal and Expectations from Students.
Working plan.

• Work Ethics
• Practice sessions
• Evaluations
• Professional Grooming

Stage 1: Introduction with the Tools

Basics of NLP & LLMs

• Tokenization
• Embeddings
• Token Limit concept for LLMs
• Introduction to Chat GPT and Interaction with it.
• Introduction to Dall-E and interaction with it.
• Introduction to Stable Diffusion and interaction with it
• Prompt Engineering
•  Prompt Anatomy

Assignment

At Initial stage the students should interact with Open.ai tools like Chat GPT and DALL-E-
2. This will greatly develop their interest and help them understand the products better. From this they will also learn the prompting which will help them later.

Stage 2: Basics of python

Introduction to Python programming

• Basic Variables
• Data types
• String manipulation
• List
• Loops
• Tuples
• Dictionary
• JSON
• Functions
  1. Built in
  2. Custom
• Code practice with Chat GPT
• Stage Evaluation
• Assignment

Stage 3: Basics of API

Introduction to API

• Basics of API
• Types of API

Hands on practice with APIs
Open.ai API

•  Stable Diffusion API

Stage 4: Introduction to Hugging Face

Introduction to Hugging Face

• Installation and Setup

Text Classification using Pipelines

•  Hands on practice

Name Entity Recognition (NER) with Pipelines

• Hand on practice

Sentiment Analysis with Pipelines

• Hands on practice

Assignment

Stage 5: Basics of ML

Introduction to Types Machine Learning

1.  Supervised Learning
   • Video demo
2. Semi-supervised Learning
   • Video demo
3. Un-supervised Learning
   • Video demo
4. Re-inforcement learning
   • Video demo

Assignment

Stage 6: Basics of Data Visualizations

Basic concepts of Pandas

Exploratory Data Analysis (EDA)

• Data cleaning Techniques
  1. Mean
  2. Median
  3. Mode
  4. Inter Quartile Range (IQR)
  5. Correlations Analysis

Dataset

• Types of Data sets (Structured, Unstructured)
• Examples of Datasets

Data preprocessing

• Data Cleaning (Missing Values and Outliers)
• Dimensionality Reduction
• Data Transformation

Introduction to Visualizations

• Line plot
• Scatter plot
• Regression plot
• Bar charts
• Distribution plots
• Box plot

Creating Visualizations using Seaborn

Data splitting into test train and validation sets.

Assignment

Stage 7: Basics of ML frame work

Understanding of Scikit-learn for Machine Learning Models

Working with Structured Data (ETL Pipeline) Using Scikit-Learn

• Data Cleaning (Missing Values and Outliers)
• Dimensionality Reduction
• Data Transformation

Concept of classification and regression

• Difference and utilization
• Use case examples

Creating Classification Models using Scikit-learn

• Logistic regression
• Decision Tree classifier
• Random Forest classifier
• Gradient Boosting Classifier
• Evaluating Classification Models

Creating Regression Models using Scikit-learn

• Linear regression
• Decision Tree Regressor
• Random Forest Regressor
• Gradient Boosting Regressor
• Evaluating Regression Models

Training process (Hands on)

Testing process (Hands on)

Evaluation Metric

• Loss functions
• Confusion matric
• Accuracy
• Precision
• Recall

Assignment

Stage 8: Tensor flow

Introduction to Tensor flow

• Problems with Linear models

Tensor flow playground

• Gradient Descent

Hyper parameters

• Epochs
• Batch size
• Learning rate
• No of layers

Artificial Neural Networks

ANN using MNIST Dataset (Hands on)

• Training
• Testing
• Limitation of ANN

Introduction to Convolutional Neural Networks

• Kernel/Filter
• Convolution
• Pooling
• Up sampling

CNN using MNIST Dataset (Hands-on)

• Training
• Testing

Auto-encoders

• Vanilla Auto Encoders
• Encoder
• Decoder
• Denoising Auto-encoders

Vanilla Auto-Encoder using ANN, CNN

• Architecture
• Training
• Testing

Denoising Auto Encoder using CNN

• Architecture
• Training
• Testing

Assignment

Stage 9: PyTorch

Introduction to PyTorch

1. Comparison of PyTorch and Tensor flow

Linear regression using PyTorch

Classification using PyTorch

Stage 10: Chatbots (LangChain, Streamlit, LlamaIndex)

Introduction to LangChain

• Concept of chains
• Concept of retrieval chains
• Simple QA chain
• Retrieval QA chain with sources Document.
• Simple Chabot using Openai (Hands-on)

Introduction to Streamlit

• Setup and installation of VS code and Anaconda
• Document GPT (Simple QA Chain)
• Document GPT (Retrieval QA Chain with Source Document)

Introduction to LlamaIndex

• Functionalities of LlamaIndex
• Difference between LlamaIndex and LangChain
• Difference between Llama-2 and LlamaIndex
• Document GPT with LlamaIndex

Assignment

Stage 11: Transformers

What are Transformers?

• How do Transformers work?

The different types of Transformer architectures

The benefits of using Transformers for Deep Learning tasks

• Hugging Face API

What is the Hugging Face API?

How to use the Hugging Face API to load and use Transformers models?

How to fine-tune Transformers models with the Hugging Face API?

• How to use Transformers and the Hugging Face API to build a text classification model
• How to use Transformers and the Hugging Face API to build a question answering model
• How to use Transformers and the Hugging Face API to build a summarization model

Stage 12: Server Deployment

What is Server Deployment?

• Introduction to Server Deployment principles and practices
• The Server Deployment culture

Server Deployment tools and technologies

•  Version control systems (Git)
• Containerization (Docker)
• Cloud computing (AWS)

You will learn

• Understanding of the fundamentals of artificial intelligence and its various applications.
• Familiarity with popular AI tools like ChatGPT, DALL-E, and Stable Diffusion.
• Proficiency in Python programming language and its data structures, control statements, functions, and classes.
• Knowledge of different types of machine learning, their applications, and the difference between supervised, unsupervised, semi-supervised, and reinforcement learning.
• Understanding of machine learning models, datasets, data preprocessing, training, testing, and evaluation metrics.
• Familiarity with different machine learning frameworks and their usage in creating structured data models.
• Knowledge of data visualization techniques using Matplotlib, Seaborn, and Plotly libraries.
• Familiarity with Hugging Face library and its usage in NLP tasks like text classification, NER, and sentiment analysis.