GenAI_Roadmap.md

 

Gen AI Roadmap for Everyone - 2024

     

---

Resources that should cover 85%+ of what you need for GenAI tasks. The field changes fast, but the basics stay the same. Don’t get lost in all the new stuff/models and focus on the fundamentals. Build projects as you learn. Spend at least an hour a day, and you’ll get a good grip on GenAI in 1-2 months.

Learning GenAI is essential now. Soon, it may be a desired requirements for all software roles. Get ready—it’s easier than you think.

PS: I planned to create a new GenAI learning playlist, where I am working on 7-8 projects, including 3 in production. However, personal and work things delayed it. I will try to record YouTube videos in the coming months to share insights from my 15+ months of production grade GenAI development experience.

---

---

🚀 GenAI Crash Course for Beginners (Optional) (~3 Hrs)

This is a nice video to refresh your high-level overview of generative AI in brief. This is optional.

• 
• Replace the OpenAI LLM used anywhere with the free Google’s Gemini Pro API-based LLM.
  • - How to setup Google's free Gemini Pro API Key
  • Note: To access the Gemini Model via Google AI, you only need a Google account and an API key. A Google Cloud account is not required. It's easy and straightforward as shown in the above yt video.

---

📚 Fundamentals of GenAI (1.2 Hr)

• - Generative AI in Nutshell
• - Intro to GenAI
• - What is LLM

Optional

• - What is LLM
• Useful LLM Concepts

  • 

  • Cohere Blog

  • Generative AI Guide

  • - Deep Learning Playlist (Famous 3blue1brown)

---

🤖 What is GPT? (28 Min)

• 

---

⚡What is Transformer? (10 Min)

• 

---

📜 Python

• - Python Core Crash Course (1 Hr)
• - Python DS Crash Course (49 Min)

---

☁️ Understanding Google Colab (22 Min)

• 

---

🛠️ Why should you use Open Source LLM? (7 Min)

• 

---

🤗 Huggingface Open Source Models (34 Min)

• 

• Official Doc

---

🏠 Running LLM Locally using Ollama (~ 1.5 Hr)

• 

Some Cool examples (18 Min)

• 

---

🔗 LangChain

I prefer going through the LangChain documentation, which is well-written and includes example notebooks, as it updates very quickly. Referring to most of the LangChain YouTube videos might give you outdated content after a few weeks.

• Official Doc

---

🎯 Prompt Engineering

• Prompt Engineering Colab Notebook
• Course
  • Coursera Specialization

Optional

• Blogs
  • Prompting Guide
  • Google's Five Best practices for Prompt Engineering
  • Google's doc of Prompt Engineering
  • OpenAI Best Practices
  • OpenAI Guide

---

📊 What is Vector Database?

• Pinecone Blog

---

🗺️ What is Vector Embedding?

• Qdrant Blog
• How to choose right Embedding Model ?

---

📖 RAG Tutorials (~2 Hr)

• 

• 

Optional

• GitHub

  • Ragatouille-Book

  • Graph RAG (New)

---

⚖️ RAG Vs Prompt Engineering Vs Fine Tuning (15 Min)

• 

---

🔧 Fine Tuning LLM (~4 Hr)

• 

---

🤖 LLM Agents

• Deeplearning.ai Short Course - Agent
• Deeplearning.ai Autogen
• Prompt Guide - LLM Agents
• Architecting & Testing reliable Agent (Using LangGraph)

  Optional

• Nvidia Blog
• Ttruefoundry Blog
• Agpt Blog

---

🔀 What is MultiModel? (7 Min)

• 
• Blog

---

🧠 What is Mixture of Experts (MoE) (~30 Min)

• 

• Huggingface MoE Blog

---

🎨 Streamlit for fast prototype UI

• 

• Official Doc

---

🔍 LLM Evaluation

• Blogs

  • LLM Evaluation Ultimate Guide
  • Large Language Model Evaluation

  Optional

  • LLM Evaluation
  • Medium Blog

---

🏭 LLMOPs & Productionization of GenAI applications

• Deeplearning.ai Short Course
• Specialization

---

📚 GenAI Use Cases (Used free Gemini Pro & huggingface LLMs only)

• - How to setup Google Colab Notebook for free GPU
• - How to setup Google's free Gemini Pro API Key
• - Conversational Analytics (Full Stack GenAI App using React, MongoDB, Free Gemini Pro LLM, Docker, Authentication & Authorisation using JWT oken)
• - Chat with Graph Database (Neo4j Graph Database, Gemini Pro LLM & Streamlit UI)
• - Machine Translation (Gemini Pro LLM & Streamlit UI)
• - Tagging (Gemini Pro LLM & Streamlit UI)
• Webscraping (Gemini Pro LLM & Streamlit UI)
• - Chatbot with SQL Database (Huggingface Opensource LLM & Streamlit UI)
• - Chatbot with CSV (Huggingface Opensource LLM)
• - Text to SQL generation (Huggingface Opensource LLM)
• - Text Summarization (Huggingface Opensource LLM)
• - Fully local RAG Agent with Llama3.1 (By LangChain Team)

---

📖 GenAI Glossary of terms

• PDF

---

📝 GenAI Interview Questions & Answers

• PDF

---

💻 GenAI Coding Round Preparation

• PDF

---

💻 GenAI Project Lifecycle

• PDF

---

🏆 LLM Leaderboard & Benchmarks

• Artificial Analysis AI
• LMSYS
• Huggingface open-llm-leaderboard

---

🤝 Contributing

Contributions to add good impactful resources/codes to the list are welcome!

Here’s how you can help:

1. Fork the Repository

   Click on the "Fork" button at the top right corner of the page to create a personal copy of the repository.

2. Clone the Repository

   Clone your forked repository to your local machine:

   git clone https://github.com/genieincodebottle/generative-ai.git

3. Create a New Branch

   Create a new branch for your feature or bug fix:

   git checkout -b your-branch-name

4. Make Your Changes

   Make your changes and commit them with a clear message:

   git commit -m "Brief description of your changes"

5. Push Your Changes Push your changes to your forked repository:

   git push origin your-branch-name

6. Create a Pull Request

   Go to the original repository and create a pull request. Make sure to explain your changes and why they should be merged.