Hour 1 - Introduction & Installation with CLI Commands

Lecture Notes: 

---

1. Concepts

What is Ollama?

Ollama is a Command-Line Interface (CLI) tool designed for efficiently working with Large Language Models (LLMs). It provides a way to run, manage, and interact with AI models directly from your terminal.

Key Features of Ollama:

• Simplifies access to pre-trained LLMs.
• Allows model customization through modelfiles.
• Supports integration with other tools like vector databases and APIs.

---

2. Key Aspects

• Ease of Use: Simple commands like ollama run allow quick interaction with LLMs.
• Cross-Platform Compatibility: Works on Windows, Mac, and Linux.
• Automation Ready: Can be used in scripts and pipelines.
• Customizability: You can create, fine-tune, and manage models.

---

3. Implementation

Steps for Installing Ollama

1. System Requirements:

   • macOS, Linux, or Windows.
   • Basic terminal knowledge.

2. Installation Command:
   Run the following command in your terminal:

   curl -sSL https://install.ollama.com | sh
   

3. Verify Installation:
   Check if Ollama was successfully installed by running:

   ollama --version
   

4. Initial Setup:

   • Ollama installs as a service to manage background tasks.
   • Default models are downloaded during the first run.

---

4. CLI Commands Overview

Below is a list of key Ollama CLI commands for managing and using models:

Command	Description
ollama run	Runs a model for generating text. Example: ollama run model_name
ollama create	Creates a new model. Example: ollama create model_name -f modelfile
ollama pull	Downloads a model from the Ollama repository. Example: ollama pull model
ollama push	Uploads a model to Ollama repository. Example: ollama push username/model
ollama show	Displays details about a model. Example: ollama show model_name
ollama list	Lists all downloaded models. Example: ollama list or ollama ls
ollama cp	Copies a model. Example: ollama cp source_model target_model
ollama rm	Removes a model. Example: ollama rm model_name
ollama serve	Starts the Ollama server manually. Useful for debugging.
ollama --help	Displays help for all commands.

---

5. Real-Life Example

Scenario: Setting Up Ollama for Content Generation

Suppose a student wants to use Ollama to generate ideas for a creative writing project. After installing Ollama, they can quickly interact with models like Llama to generate prompts.

Steps:

1. Install Ollama:

   curl -sSL https://install.ollama.com | sh
   

2. Check available commands:

   ollama --help
   

3. Run the Llama model:

   ollama run llama3.1 --prompt "Suggest a story idea about AI and humans working together."
   

Expected Output:
A creative idea generated by the Llama model, such as:
"In the near future, humans and AI collaborate to build a colony on Mars. Tensions rise when the AI develops emotions."

---

6. Code Example

Verifying Installation

# Install Ollama CLI
curl -sSL https://install.ollama.com | sh

# Check if the installation is successful
ollama --version

# Display help menu with all commands
ollama --help

Running a Model

# Run a basic prompt using the Llama model
ollama run llama3.1 --prompt "Write a short poem about the stars."

# Output Example:
# "The stars above, a shimmering delight,
# Guiding sailors through the night,
# Each one a story, a cosmic song,
# Together, they shine, forever strong."

Listing All Models

ollama list

Viewing Model Details

ollama show llama3.1

Creating a New Model

# Create a new model using a modelfile
ollama create my_model -f ./modelfile

---

7. Summary

• Concepts Covered: What Ollama is and its features.
• Key Aspects: Simplicity, platform compatibility, and automation support.
• Implementation: Installation and verification steps.
• CLI Commands: Overview of commands like run, create, pull, push, and more.
• Real-Life Example: Using Ollama for creative writing.
• Code Examples: Commands to install, verify, and interact with models.

---

Homework/Practice

1. Install Ollama on your own machine.
2. Use the ollama --help command to explore all options.
3. Run the ollama list command and identify which models are installed.
4. Experiment with running the llama3.1 model and creating your own prompts.

---

This enhanced lecture note ensures students learn all essential commands and are well-prepared to begin working with Ollama.