PankajMehar / udemy-llm-agents

Welcome to the course on LLM-Agents

This Github has all the coding exercises organized by Unit

Installation

The installation steps are:

1. Git clone this repo - git clone https://github.com/priya-dwivedi/udemy-llm-agents.git
2. Move to the current directory - cd udemy-llm-agents
3. If using conda - set a new envtt: conda create -n llm_agents

conda activate llm_agents

conda install pip

3. Install the dependencies - pip install -r requirements.txt
4. Add an environment file called *.env
5. Add your keys to the env file

Unit 1- Foundation of LLM agents

This unit focuses on components of an LLM Agent and building your first simple agent Coding Exercise for the first Self-Ask Agent:

• Run Locally using Notebook: First-Agent.ipynb
• Run on Colab with Link:

Unit 2- LLM Tools

This unit focuses on integrating external tools into an Agent Coding Exercise for the second unit on Tools:

1. Langchain Tools

• Run Locally using Notebook: Langchain-Tools.ipynb
• Run on Colab with Link:

2. LLM RestAPI tool selection

• Run Locally using Notebook: LLM-RestAPI-Selection.ipynb
• Run on Colab with Link:

3. OpenAI Function Calling

• Run Locally using Notebook: OpenAI-Function-Calling.ipynb
• Run on Colab with Link:

Unit 3- Memory

This unit focuses on different types of memory and integrating memory into the Agent Coding Exercise for the third unit on Memory:

1. Langchain Short term Memory

• Run Locally using Notebook: Langchain-Short-term-memory.ipynb
• Run on Colab with Link:

2. RAG pipeline and RAG compared to Long Context LLMs

• Run Locally using Notebook: RAG_vs_LongContext.ipynb
• Run on Colab with Link:

3. Simple Multimodal RAG

• Run Locally using Notebook: MultiModal-RAG.ipynb
• Run on Colab with Link:

4. Knowledge-Agent Can only be run locally Instructions:

• Navigate to the correct directory: cd Unit3-Memory/knowledge-agent
• Optionally: Delete current profile : rm current_profile.json
• Run Streamlit interface: streamlit run chatbot.py

Unit 4-Planning

This unit focuses on different types of planning and how to improve the Agent's accuracy and performance with planning Coding Exercise for the fourth unit on Planning:

1. Plannning with task decomposition

• Run Locally using Notebook: Planning_with_task_decomposition.ipynb
• Run on Colab with Link:

2. Skeleton of Thought Generation

• Run Locally using Notebook: Skeleton_of_Thought_Generationa.ipynb
• Run on Colab with Link:

3. Basics of Langgraph

• Run Locally using Notebook: Basics_of_Langgraph.ipynb
• Run on Colab with Link:

4. Reflection Agent

• Run Locally using Notebook: Reflection_Agent.ipynb
• Run on Colab with Link:

4. Reflexion Agent

• Run Locally using Notebook: Reflexion_Agent.ipynb
• Run on Colab with Link:

Unit 5-Agent Examples

This unit focuses on building more complex agents combining everything learned so far in the course Coding Exercise for the fifth unit on Agent Examples:

1. Agentic-RAG

• Run Locally using Notebook: Agentic-RAG.ipynb
• Run on Colab with Link:

2. Movie Recommendation Bot Can only be run locally Instructions:

• Navigate to the correct directory: cd Unit5-Agent_Examples/Movie_Recommendation_Agent
• Run Streamlit interface: streamlit run chatbot.py

3. Coding Assistant

• Run Locally using Notebook: Coding-Assistant.ipynb
• Run on Colab with Link: