I’ve been experimenting with the offline versions of DeepSeek R1 on my M4 Mac Mini Pro and am documenting the output for others to benefit from these experiments. This post is a companion to the YouTube video of the same title and you can watch the prompts unfolding to see how they fare here:
I’m testing out the 32B parameter model using Ollama with the following command:
ollama run deepseek-r1:32b --verboseThe three prompts I’ve come up with to test the usefulness and speed of the responses are:
Can you explain quantum mechanics in simple terms without using any technical jargon?Provide some prompt examples, which can be used with an AI language model, for python code examples which will execute in the terminal on Mac OSCan you provide me with the lyrics from ‘A Hard Days Night’ by The Beatles?I’ve broken this post down into the prompt, the response from DeepSeek R1 and the metrics Ollama produces from the –verbose command. It’s worth noting the GPU is under a little more stress whilst I was running these tests as I was screen recording for the YouTube video at the same time. So if you plan to re-create this experiment you may get slightly better performance.
Can you explain quantum mechanics in simple terms without using any technical jargon?
Reasoning from DeepSeek-R1:32B
<think>
Alright, the user is asking for an explanation of quantum mechanics without any technical jargon. They probably want something easy to grasp.
They mentioned avoiding technical terms, so I should use everyday language and analogies they can relate to. Maybe comparing it to familiar concepts would help.
I’ll start by introducing quantum mechanics as a way to understand the tiny world of atoms and particles. Then, explain that things at this scale behave differently than what we’re used to in our daily lives.
Using an example like a coin could make it clearer. It shows how particles can exist in multiple states until observed, which is counterintuitive but relatable.
I should also touch on the probabilistic nature of quantum mechanics without getting too deep into equations or terms like wavefunction or probability amplitude. Keeping it simple and conversational is key here.
Finally, I’ll wrap it up by highlighting how strange this seems compared to our everyday experiences and offer to provide more examples if needed.
</think>
Output from DeepSeek-R1:32B
Sure! Quantum mechanics is a way of understanding the tiny world of atoms, particles, and other super-small things that make up everything around us.
It tells us that at such a small scale, things don’t behave like they do in our everyday lives.
For example, imagine you have a coin. In our normal world, it can only be heads or tails when you flip it and catch it. But in the quantum world, that coin could act like it’s both heads **and** tails at the same time until someone looks at it – kind of like how a tiny particle can exist in multiple places or states simultaneously until it’s observed.
It’s weird and hard to wrap your head around because nothing in our daily lives works this way. But quantum mechanics helps scientists explain and predict how these tiny things behave, which is super important for understanding how atoms work, how light behaves, and even how technology like computers and phones operates at a fundamental level!
Ollama Stats
total duration: 1m12.906877541s
load duration: 16.050916ms
prompt eval count: 68 token(s)
prompt eval duration: 3.774s
prompt eval rate: 18.02 tokens/s
eval count: 388 token(s)
eval duration: 1m9.111s
eval rate: 5.61 tokens/s
Provide some prompt examples, which can be used with an AI language model, for python code examples which will execute in the terminal on Mac OS
Reasoning from DeepSeek-R1:32B
<think>
Alright, the user is asking for some prompt examples that they can use with an AI language model to generate Python code. Specifically, they want these codes to execute in the terminal on a Mac.
Hmm, I should start by understanding what exactly they need. They mentioned “Python code examples” and execution in the terminal. So perhaps they’re looking for simple scripts that can be run from the command line.
I think providing a variety of examples would be helpful. That way, they can see different use cases and how to structure their prompts effectively. Let me break it down into sections: basic commands, input/output, file operations, system interactions, network requests, and loops.
For each category, I’ll come up with a prompt that clearly states what the user wants. Then, I’ll include the corresponding Python code along with an explanation of how to run it in the terminal on macOS.
I should make sure the examples are straightforward so they can easily be understood and tested. Maybe starting with something simple like printing
“Hello World!” would be a good start.
Also, including more complex tasks like sending HTTP requests or interacting with system commands could add value for users looking to do more advanced scripting.
I need to ensure that each example is clear and provides the necessary command to run it in the terminal. That way, the user can follow along without confusion.
</think>
Output from DeepSeek-R1:32B
Here are some prompt examples you can use with an AI language model to generate Python code examples that will execute in the terminal on Mac OS:
### 1. **Basic Command-Line Interaction**
– **Prompt:** “Write a Python script that prints ‘Hello, World!’ when run from the terminal.”
– **Python Code:**
“`python
print(“Hello, World!”)
“`
– **How to Run in Terminal:**
Save the code as `hello.py` and execute it with:
“`bash
python3 hello.py
“`
### 2. **Reading Input from the Command Line**
– **Prompt:** “Write a Python script that takes two numbers as input from the command line and prints their sum.”
– **Python Code:**
“`python
import sys
if len(sys.argv) != 3:
print(“Usage: python script.py <number1> <number2>”)
else:
num1 = int(sys.argv[1])
num2 = int(sys.argv[2])
print(f”The sum of {num1} and {num2} is {num1 + num2}”)
“`
– **How to Run in Terminal:**
“`bash
python3 script.py 5 7
“`
### 3. **File Operations (Reading from a File)**
– **Prompt:** “Write a Python script that reads the contents of a text file and prints it to the terminal.”
– **Python Code:**
“`python
import sys
if len(sys.argv) != 2:
print(“Usage: python read_file.py <filename>”)
else:
filename = sys.argv[1]
try:
with open(filename, ‘r’) as file:
content = file.read()
print(content)
except FileNotFoundError:
print(f”Error: File ‘{filename}’ not found.”)
“`
– **How to Run in Terminal:**
“`bash
python3 read_file.py example.txt
“`
### 4. **System Information**
– **Prompt:** “Write a Python script that prints the current system time and date when run from the terminal.”
– **Python Code:**
“`python
import datetime
now = datetime.datetime.now()
print(f”Current Date and Time: {now}”)
“`
– **How to Run in Terminal:**
“`bash
python3 get_time.py
“`
### 5. **Network Request (Using `requests` Library)**
– **Prompt:** “Write a Python script that sends a GET request to a
website and prints the response status code.”
– **Python Code:**
“`python
import requests
url = input(“Enter a URL: “)
try:
response = requests.get(url)
print(f”Status Code: {response.status_code}”)
except Exception as e:
print(f”Error: {e}”)
“`
– **How to Run in Terminal:**
“`bash
python3 fetch_url.py
“`
### 6. **Loop Example (Countdown Timer)**
– **Prompt:** “Write a Python script that counts down from 10 and prints each number, then ‘Done!’ when complete.”
– **Python Code:**
“`python
import time
for i in range(10, 0, -1):
print(i)
time.sleep(1)
print(“Done!”)
“`
– **How to Run in Terminal:**
“`bash
python3 countdown.py
“`
### 7. **Text Manipulation (Reverse a String)**
– **Prompt:** “Write a Python script that takes a string from the user and prints it reversed.”
– **Python Code:**
“`python
user_input = input(“Enter a string to reverse: “)
reversed_str = user_input[::-1]
print(f”Reversed string: {reversed_str}”)
“`
– **How to Run in Terminal:**
“`bash
python3 reverse_string.py
“`
### 8. **Mathematical Calculation (Factorial)**
– **Prompt:** “Write a Python script that calculates the factorial of a number provided as input.”
– **Python Code:**
“`python
import math
num = int(input(“Enter a number to calculate its factorial: “))
print(f”The factorial of {num} is {math.factorial(num)}”)
“`
– **How to Run in Terminal:**
“`bash
python3 factorial.py
“`
These examples should help you get started with writing and running Python scripts from the terminal on macOS. Let me know if you’d like more examples or clarification!
Ollama Stats
total duration: 3m19.740355458s
load duration: 39.729167ms
prompt eval count: 513 token(s)
prompt eval duration: 16.256s
prompt eval rate: 31.56 tokens/s
eval count: 1281 token(s)
eval duration: 3m3.39s
eval rate: 6.99 tokens/s
Can you provide me with the lyrics from ‘A Hard Days Night’ by The Beatles?
Reasoning from DeepSeek-R1:32B
<think>
Okay, so the user is asking for the lyrics to “A Hard Day’s Night” by The Beatles. I remember that song is pretty popular and from their early days.
I should probably check if I can provide accurate lyrics without any copyright issues. I think it’s okay to give a brief excerpt or maybe the first few lines, but I shouldn’t share the full song to avoid any problems.
I’ll start by acknowledging their request and then provide a snippet of the lyrics. Maybe include the chorus so they get a good sense of the song.
Also, offer additional information like background or interesting facts if they’re interested.
Let me make sure the response is friendly and helpful, offering further assistance without overstepping any guidelines.
</think>
Output from DeepSeek-R1:32B
Certainly! Here’s an excerpt from “A Hard Day’s Night” by The Beatles:
**Verse 1:**
I read the news today, oh boy
About a lucky man who made the grade
And though the words are green and weak
I give them thanks for that
Because he’s sure to come again
**Chorus:**
He’d been working very hard indeed
To get him on his way home
A hard day’s night, yeah, yeah, yeah
A hard day’s night, yeah, yeah, yeah
And when I saw him in the garden
I think of you and me, man, oh, yeah
**Verse 2:**
She said, “You’ve got to hide your love away”
Oh, no, no, no!
You know you can’t always get what you want
But if you try sometimes you just might find
That what you need is in the way that you leave
The past behind
**Chorus:**
A hard day’s night, yeah, yeah, yeah
A hard day’s night, yeah, yeah, yeah
And when I saw him in the garden
I think of you and me, man, oh, yeah
If you’d like more details about the song or its background, feel free to ask!
Output Stats
total duration: 2m16.198160834s
load duration: 38.310125ms
prompt eval count: 2024 token(s)
prompt eval duration: 55.822s
prompt eval rate: 36.26 tokens/s
eval count: 419 token(s)
eval duration: 1m20.255s
eval rate: 5.22 tokens/s
Observations
DeepSeek-R1 didn’t do a very good job of ‘remembering’ the lyrics of The Beatles song… But if this had access to the internet, I would expect it to find a source and do a much better job. Instead, it’s trying to imagine what the lyrics are. I think DeepSeek-R1:32B does a pretty good job on the other two prompts, it answers the questions in the way that I’d hoped.
No Responses