AI.txt

Ollama Cheat Sheet
===
// Pull/Download models
ollama pull llama2
ollama pull mistral:7b

// List available models
ollama list

// Remove models
ollama rm llama2

// Show model info
ollama show llama2

// Running models
---
// Interactive chat
ollama run llama2

// Single prompt
ollama run llama2 "Explain quantum computing"

// Run with custom parameters
ollama run llama2 --temperature 0.8 --top-p 0.9


// Server Management
---
// Start Ollama server. Runs at http://localhost:11434
ollama serve

// Check server status
ollama ps

// Stop running models
ollama stop llama2



AI Resources
============

𝟭. Generative AI: https://lnkd.in/gt-Gf_H7
𝟮. AI for Everyone: https://lnkd.in/gKdf7F53
𝟯. ChatGPT for Excel: https://lnkd.in/gxjVb_Un
𝟰. AI Introduction by Harvard: https://edx.sjv.io/g1Bg15
𝟱. IBM AI Engineering: https://lnkd.in/gv5b9NKA
𝟲. Generative AI for CEOs: https://lnkd.in/gZHq3V_z
𝟳. AWS GenAI Foundation: https://lnkd.in/daiz7YiW
𝟴. Introduction to Generative AI: https://lnkd.in/d-RebaXf
𝟵. LangChain App Developer: https://lnkd.in/gZeaHgXG
𝟭𝟬. Generative AI by Google: https://lnkd.in/d-RebaXf
𝟭𝟭. IBM Applied AI Professional Certificate: https://lnkd.in/dQsYnZ7M
𝟭𝟮. Generative AI for Data Scientists Specialization: https://lnkd.in/gwe_pGFN
𝟭𝟯. Prompt Engineering for ChatGPT: https://lnkd.in/gsPqptAs
𝟭𝟰. Google's Introduction to Responsible AI: https://lnkd.in/g_u-G_hC
𝟭𝟱. IBM AI Engineering Professional Certificate: https://lnkd.in/gv5b9NKA
𝟭𝟲. IBM Machine Learning Professional Certificate: https://lnkd.in/gAgdH-Fn
𝟭𝟳. Introduction to Artificial Intelligence (AI) : https://lnkd.in/gYNJFA9F



1/ Fundamental AI Concepts: https://lnkd.in/dSbhE7C3
2/ Fundamentals of Generative AI: https://lnkd.in/dVAjnymE
3/ Fundamentals of Responsible Generative AI: https://lnkd.in/dtSu2WHT
4/ Fundamentals of Azure OpenAI: https://lnkd.in/dA6BYJTR
5/ Prompt Engineering with Azure OpenAI: https://lnkd.in/dhExyuRY
6/ Implementation of RAGs: https://lnkd.in/d8SD9nzm
7/ Build RAG based copilot: https://lnkd.in/dpZkP7HQ
8/ Create a knowledge store: https://lnkd.in/dZq5dR6X
9/ Fine-tune foundational models: https://lnkd.in/dHPW6ivK
10/ Build Recommendation Systems: https://lnkd.in/dCRgv5iR
11/ MLOps: https://lnkd.in/dcTntKni
12/ Deploy a model with Github Actions: https://lnkd.in/dBTSbxgZ
13/ Deploy model to NVIDIA Triton: https://lnkd.in/d6DwXV74
14/ Monitoring ML models: https://lnkd.in/dBdm9z-j
15/ Securing ML Models: https://lnkd.in/dq82tXQ4


https://news.ycombinator.com/item?id=41105348



Generative AI Learning Path
  - https://karpathy.ai/zero-to-hero.html
  - "Deep Learning" by Bengio
  - Karpathy's videos on GPT from scratch
  - 3b1b's videos going over the math and fundamentals should give you a nice grip over the concepts.
  - After that I feel you would be ready to delve into newer papers and be able
    to understand what's going on and stay "relevant" as you said.

  - I compiled some - see here https://news.ycombinator.com/item?id=36195527
    - https://llm-utils.org/AI+Learning+Curation
  - https://github.com/swyxio/ai-notes/blob/main/README.md#top-ai-reads

  - Updated version: https://www.bishopbook.com

  - I studied Physics in college, so I mostly knew the Math.
      Here's what I used to self-learn:
      1. Machine Learning for Absolute Beginners by Oliver Theobald
      2. Introduction to Statistical Learning
      3. Machine Learning by Andrew Ng on Coursera
      4. Deep Learning by Andrew Ng on Coursera
      5. fast.ai
      6. Sebastian Raschka's PyTorch book
      Good Math refreshers:
      1. Mathematics for Machine Learning Specialization by Imperial College London
      2. Linear Algebra and Calculus series by 3blue1brown on YT
      Later I delved deep into Computer Vision for profession, and Edge AI for personal projects.


AI Agentic Tools
---
- n8n.io
- relevanceai.com - Sales and Marketing focused.
- stack-ai.com - enterprise AI for internal use.
- crewai.com - not easy to get started with. For technical people.
- botpress.com - beginner friendly. Easy to build chat bot.
- flowwise.com - open source alternative to botpress.com
- lindy.ai
- mindstudio.ai


Here’s my 5-step roadmap to go from 0 to shipping real AI systems:
-----------------

Step 1️⃣ Nail the Foundations
---
To get started:
⑴ Complete ML for Beginners (Microsoft) – classical ML, Scikit-learn, quizzes, projects → https://lnkd.in/eHHiv3g2

⑵ Then dive into with AI for Beginners (Microsoft) – neural nets, NLP, CV, ethics, PyTorch & TensorFlow labs → https://lnkd.in/eNN8y6g8

Test yourself on real datasets before moving on.
Once that's done, you're ready for..

Step 2️⃣ Master Modern Deep Learning
---
This is where you go beyond the basics and start understanding how today’s AI systems actually work

⑶ Build from scratch with Neural Networks: Zero to Hero (Karpathy) – code tiny GPTs and grasp every layer → https://lnkd.in/eDzGhrtn

⑷ Explore state-of-the-art model via DL Paper Implementations – 60+ transformer, GAN, diffusion models in clean PyTorch → https://lnkd.in/eH53_5mi

First learn how deep learning works. Then learn how it’s applied.

Step 3️⃣ Learn to Ship ML Systems
---
Now it’s time to go beyond the notebook.

⑸ Made With ML (Goku Mohandas)– teaches designing, building, deploying, and iterating on production-grade ML systems with MLOps, CI/CD and testing → https://lnkd.in/ek5YHbhN

Step 4️⃣ Go Deep on LLMs and RAG
---
Most jump straight to fine-tuning and get lost. Instead, take the best path:

⑹ Hands-on LLMs (Jay Alammar & Maarten Grootendorst)– clear guide covering tokenization, fine-tuning, embeddings, RAG, and more in a clear, visual way → https://lnkd.in/esWdYHwc

⑺ Advanced RAG Techniques (Nir Diamant)– 30+ methods to make retrieval-augmented generation faster, smarter, and more accurate → https://lnkd.in/ezx_5nUk

⑻ LLM Engineer’s Handbook (Paul Iusztin & Maxime Labonne) end-to-end guide to the full LLM lifecycle with a focus on practical implementation and LLMOps best practices → https://lnkd.in/eiuBSW2k

Step 5: Build and Deploy Agents
---
Now that you understand LLMs and RAG work, it’s time to go agentic.

⑼ AI Agents for Beginners (Microsoft) – hands-on course to build LLM agents using frameworks like AutoGen → https://lnkd.in/efzMR5aU

⑩ Agents Towards Production – a practical playbook for deploying agents with memory, orchestration, security, and more → https://lnkd.in/eDbEYmGD

⑪ AI Engineering Hub (Avi Chawla & Akshay Pachaar)– 70+ real-world projects and agent apps you can build, adapt, and ship → https://lnkd.in/eAR8asPv

Follow this path, and you won’t just learn—you’ll learn how to ship real AI systems.

============================================================================


AI Terminology


Machine Learning vs Data Science
================================
ML: "Field of study that gives computers the ability to learn without being explicity programmed"
  - Arthur Samuel (1959)
  - This is a software.

Data Science: "Science of extracting knowledge and insights from data".
  - Usually a slide deck.


Deep Learning
=============

many inputs ---> Big math equation ---> Output

Big math equation ==> Artificial Neural Networks. This is no relation to how brain work (real Neural Networks).


     ┌───────────────────────────────────────────────┐
     │   AI                                          │
     │  ┌──────────────────────────────┐   ┌──────┐  │
     │  │                              │   │      │  │
     │  │   ML                         │   │      │  │
     │  │                              │   └──────┘  │
     │  │   ┌─────────────────────┐    │             │
     │  │   │                     │    │   ┌──────┐  │
     │  │   │                     │    │   │      │  │
     │  │   │    DL/NN            │    │   │      │  │
     │  │   │                     │    │   └──────┘  │
     │  │   │                     │    │             │
     │  │   │                     │    │   ┌──────┐  │
     │  │   │                     │    │   │      │  │
     │  │   │                     │    │   │      │  │
     │  │   │                     │    │   └──────┘  │
     │  │   └─────────────────────┘    │             │
     │  └──────────────────────────────┘             │
     └───────────────────────────────────────────────┘


AI Transformation Playbook
----
1. Execute pilot projects to gain momentum.
2. Build an in-house AI team.
3. Provide broad AI training.
4. Develop an AI strategy.
5. Develop internal and external communications.


- Imperfect rule of thumb to decide what Machine Learning can and cannot do?
  - Anything you can do with 1 second of thought, we can probably now or soon automate.