Coding with reasoning models

Time for you to have a go at coding with DeepSeek's R1 model! As you may know, one important part of developing reliable code is to include tests, which validate that intermediate processes are running correctly before returning a final output. Well-written tests make debugging much easier!

Your colleague has handed-over a Python script they were working on, which you notice doesn't contain any tests. As you aren't familiar with the libraries or syntax used in the script, you'll use a reasoning model to add appropriate testing to the script in the form of assert statements. If you'd like to read more about assert statements, check out this DataCamp tutorial.

import numpy as np
a = np.arange(5)
b = a.reshape(-1, 1)
c = np.ones((5, 3))
d = np.hstack((b, c))
col_means = d.mean(axis=0)

Deze oefening maakt deel uit van de cursus

Working with DeepSeek in Python

Cursus bekijken

Oefeninstructies

Copy the code provided into the delimited prompt that has been started for you.
Send prompt to the deepseek-ai/DeepSeek-R1 model.

Praktische interactieve oefening

Probeer deze oefening eens door deze voorbeeldcode in te vullen.

client = OpenAI(api_key="", base_url="https://api.together.xyz/v1")

# Copy the code into the prompt
prompt = """[Task: Write Python assert statements for the following code.]
Code:
____
"""

# Send the prompt to R1
response = client.chat.completions.create(model="deepseek-ai/DeepSeek-R1", messages=[{"role": "user", "content": ____}], max_tokens=400)
print(response.choices[0].message.content)

Code bewerken en uitvoeren