Local Model Training: From Excel to Automation Master

Welcome to the local model training exercises. These exercises are designed to help you master AI-assisted workflows using local LLMs and coding agents.

🚀 Levels of Mastery

1. Level 1: Excel & Pi for Excel

• Focus: Direct spreadsheet interaction, natural language formulas, and data cleaning.
• Tools: Excel + Pi for Excel + LM Studio.
• Key Files: Located in exercise_data/L1_data.

---

2. Level 2: Advanced Scripting & Automation

• Focus: Data engineering, automated reporting, and web automation logic.
• Tools: Python (Pandas/Matplotlib), JavaScript, Coding Agents.
• Key Files: Located in exercise_data/L2_data.

---

2.5 Level 2.5: Intermediate Automation (PDFs & Web Scraping)

• Focus: Document data extraction (PDFs), Web Scraping basics, and Browser Automation.
• Tools: PyPDF2, BeautifulSoup, Playwright.
• Key Files: Located in exercise_data/L2.5_data.

---

ACP Configs

The following configuration details are for agent servers:

  "agent_servers": {
    "pi-acp":
    { "type": "registry" },
      "pi-windows": {
        "type": "custom",
        "command": "npx.cmd",
        "args": ["-y", "pi-acp"],
        "env": {
          "PATH": "C:\\Program Files\\nodejs;C:\\Windows\\system32"
      }
    }

---

🛠️ Global Prerequisites

1. LM Studio: For running local inference.
2. Coding Agent: (e.g., Antigravity, Cline, or VS Code Copilot) to assist with script writing in Level 2 & 3.
3. Exercise Data: All raw materials are organized in their respective level folders.