Going Local

🦙 Running Local Models with Ollama

Time to test the same puzzles (easy one) on local models running directly on your machine.

easy_puzzle = [
    {"role": "user", "content": 
        "You toss 2 coins. One of them is heads. What's the probability the other is tails? Answer with the probability only."},
]

💡 If Ollama isn't running, open a terminal and type:

ollama serve

We will continue on Level2 folder → part1.ipynb file

▶️ Create a new code cell and run it:

# Sends a GET request to a local server running on port 11434 (commonly Ollama) 
# And returns the raw response body as bytes.
requests.get("<http://localhost:11434/>").content

👉 So in short, this line is typically a health check to confirm that a local server (often Ollama) is running and reachable from your notebook.

▶️ Create a new code cell and run it:

!ollama pull llama3.2

▶️ Create a new code cell and run it:

# Only do this if you have a large machine - at least 16GB RAM

!ollama pull gpt-oss:20b

▶️ Create a new code cell and run it:

response = ollama.chat.completions.create(model="llama3.2", messages=easy_puzzle)
display(Markdown(response.choices[0].message.content))

✅ Expected Output: 0.5 ( ❌ wrong)

▶️ Create a new code cell and run it:

response = ollama.chat.completions.create(model="gpt-oss:20b", messages=easy_puzzle)
display(Markdown(response.choices[0].message.content))

✅ Expected Output: 2/3 ( ✅ Correct)