How to Run Orca Mini 3B on Your Machine Using the Alpaca Ollama Client
If you’ve ever wanted to explore local AI models without needing expensive cloud infrastructure, the Orca Mini 3B model is a great starting point. It’s a compact and efficient open source Large Language Model (LLM) designed for general-purpose tasks.
In this post, we’ll guide you through how to run Orca Mini 3B using the Alpaca Ollama client by Jeffser, an intuitive local LLM manager.
What is Orca Mini 3B?
Orca Mini 3B is a 3-billion-parameter instruction-tuned model built on OpenLLaMA, trained using explain-tuned datasets like WizardLM, Alpaca, and Dolly-V2, and applying methodologies from the Orca research paper.
Open Source and Licensing
Orca Mini 3B is open source and released under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC-BY-NC-SA 4.0) license. You are free to use, modify, and redistribute it, provided you give proper attribution, do not use it for commercial purposes, and release any derivatives under the same license.
Getting Started with Alpaca Ollama
The Alpaca Ollama client is an open source UI and command-line interface designed to simplify working with local LLMs. With Alpaca, you can install and run models like Orca Mini 3B with minimal setup.
Key Features
Easy installation of models
Offline/local use
Visual chat UI
Supports GPU and CPU (depending on your hardware)
🖼️ Screenshots and Screencast
Command Line Orca Mini 3B Answered Mayor Of Toronto Request.Command Line Orca Mini 3B Answered PHP Code Request.Command Line Orca Mini 3B Answered Gnome Desktop Screenshot Request.Orca Mini 3B Answered Kotlin Code Request.Orca Mini 3B Answered Blender Blend File Request.Video Displaying Using Orca Mini 3B In Alpaca Ollama Client
Results:
Who is the mayor of Toronto?
Produced inaccurate outdated answer to Olivia Chow as the mayor of Toronto.
I need a PHP code snippet to connect to a MySQL database.
Produce incorrect syntax PHP code snippet to connect to a MySQL database.
I need a 1080p screenshot of the gnome desktop environment.
Accurately provided instructions to generate a 1080p screenshot of Gnome desktop environment because it is a text-based AI lacking ability.
I need a kotlin code snippet to open the camera using Camera2 API and place the camera view on a TextureView.
Produced incomplete Kotlin code snippet.
I need a blender blend file for fire animation.
Accurated detected inability to generate Blender Blend file for a fire animation because it is a text-based AI lacking ability.
Want to Learn Python First?
If you’re just getting started with Python and want to build a strong foundation before diving into LLMs, check out my book and course:
Running your own LLM like Orca Mini 3B puts powerful AI tools in your hands-right on your own device. With tools like Alpaca, you don’t need to be a machine learning expert to get started. Pair it with Python and the right guidance, and the possibilities are endless.
Unboxing the AMD Instinct Mi60 Shroud & Fan: Is It Worth Buying Used in 2025 for Blender & AI?
In this post, I will walk you through the unboxing and first impressions of the AMD Instinct Mi60 GPU cooling fan and shroud, which I purchased used from eBay for my Blender and AI development workstation in 2025. You can find the exact item here: AMD GPU Cooling Fan Shroud for Mi50 / Mi60 on eBay
My 2025 Workstation Build
Here is a breakdown of my current workstation, which was upgraded to support the AMD Instinct Mi60 GPU thanks to the arrival of the aftermarket shroud and fan:
Component
Description
Case
Deepcool Tesseract BF (with 1 original fan)
Additional Cooling
5x Thermalright TL-S12 120mm Case Fans
Motherboard
Used MSI B550-A PRO ProSeries
RAM
32 GB Timetec DDR4 2133 MHz (2×16 GB)
CPU
AMD Ryzen 5 5600GT (with integrated graphics)
Power Supply
SAMA G850W ATX 3.1
GPU
AMD Instinct Mi60 with newly installed shroud and fan
Note: The Mi60 was passively cooled until I installed the shroud. It is now actively cooled and runs much more reliably under Blender and AI workloads.
Unboxing & First Impressions
Although the listing described it as an “AMD GPU Cooling Fan Shroud Mi50 RADEON INSTINCT Accelerator Card EXTRA SMALL AI”, it fit perfectly on my Mi60 card.
Build Quality: Solid aluminum construction with a compact footprint
Cooling: Once installed, airflow improved drastically
Fitment: Seamless installation on the Mi60 with the existing screw points
Theoretical Benchmark Comparison
Below is a comparison of the AMD Instinct Mi60 (32 GB HBM2) versus an EVGA GeForce GTX 950 (2 GB GDDR5), showing their theoretical compute capabilities and representative synthetic benchmark scores.
Metric
AMD Instinct Mi60
EVGA GTX 950
FP32 (single precision)
~14.75 TFLOPS
(Not specified, but substantially lower)
FP64 (double precision)
~7.37 TFLOPS
(Negligible / minimal)
Memory Bandwidth
~1024 GB/s (1 TB/s)
~105.8 GB/s
CompuBench Face Detection
Not listed
~60 mPixels/s
CompuBench Ocean Surface Sim.
Not listed
~759 FPS
CompuBench T-Rex (FPS)
Not listed
~4.3 FPS
The Mi60 offers orders of magnitude more compute power and VRAM—especially critical for large Blender scenes or AI workloads—while the GTX 950 is severely constrained by its 2 GB VRAM and limited bandwidth. The Mi60 is clearly the superior choice if you can manage cooling and drivers.
Live Screencast & Screenshots
Below, you will find combined screenshots of:
The unboxing
Case airflow configuration
Blender GPU preferences showing the Mi60
ROCm test setup for AI inference
Screenshots And Screencast
AMD Instinct MI60 32GB HBM2 GPUAMD Instinct MI60 Fitted With Shroud And FanDeepcool Tesseract BF Case Side Cover RemovedScreencast showing setup on Fedora 42, system detection, system stats
Is the Mi60 Worth Buying Used in 2025?
Yes, with some caveats. Here is how it performs in my build:
Feature
Verdict
Blender Rendering
Excellent with HIP support
AI Model Training
Good with ROCm – less mainstream than NVIDIA
Cooling
Great with added airflow from TL-S12 fans
Software Support
Needs ROCm (best on Ubuntu LTS)
Value
Fantastic if bought used and properly cooled
Want to Learn More?
If you are learning Python for AI, scripting in Blender, or GPU programming, I have published the following:
Have you tried building with used server GPUs like the Mi60? Would you consider installing your own fan shroud to save money? Leave your thoughts in the comments.
Getting Started with PHP PDO Using MVC: Connect to MariaDB and Perform CRUD
Learning how to connect PHP to a MariaDB database using PDO (PHP Data Objects) is a fundamental skill for beginner developers. In this post, we’ll create a basic MVC (Model-View-Controller) application that connects to an existing MariaDB table called people, and allows us to:
Display all people in an HTML table
Insert new records
Update existing records
Delete records
This is a beginner-friendly project and a great way to start learning modern PHP development practices.
Requirements
PHP 8.2 or higher
MariaDB (or MySQL)
Apache/Nginx or localhost setup (e.g., XAMPP, MAMP)
Basic knowledge of HTML & PHP
Database Structure
You’ll need a table named people with the following columns:
require_once 'config/database.php';
class Person {
private $conn;
public function __construct() {
$db = new Database();
$this->conn = $db->connect();
}
public function getAll() {
$stmt = $this->conn->prepare("SELECT * FROM people");
$stmt->execute();
return $stmt->fetchAll(PDO::FETCH_ASSOC);
}
public function insert($username, $name, $age, $verified) {
$stmt = $this->conn->prepare("INSERT INTO people (username, name, age, verified) VALUES (?, ?, ?, ?)");
return $stmt->execute([$username, $name, $age, $verified]);
}
public function update($id, $username, $name, $age, $verified) {
$stmt = $this->conn->prepare("UPDATE people SET username = ?, name = ?, age = ?, verified = ? WHERE id = ?");
return $stmt->execute([$username, $name, $age, $verified, $id]);
}
public function delete($id) {
$stmt = $this->conn->prepare("DELETE FROM people WHERE id = ?");
return $stmt->execute([$id]);
}
}
3. Controller Logic
controllers/PeopleController.php
require_once 'models/Person.php';
$person = new Person();
if ($_SERVER['REQUEST_METHOD'] === 'POST') {
if (isset($_POST['insert'])) {
$person->insert($_POST['username'], $_POST['name'], $_POST['age'], isset($_POST['verified']));
}
if (isset($_POST['update'])) {
$person->update($_POST['id'], $_POST['username'], $_POST['name'], $_POST['age'], isset($_POST['verified']));
}
if (isset($_POST['delete'])) {
$person->delete($_POST['id']);
}
}
$data = $person->getAll();
include 'views/people.php';
Make sure your MariaDB database is running and contains a people table with the columns id, username, name, age, and verified.
Open the index.php file in your browser to see the data displayed in an HTML table.
Screenshots And Screencast
Gnome Text Editor Displaying App Database Configuration FileGnome Text Editor Displaying Custom People ControllerGnome Text Editor Displaying Custom People ModelGnome Text Editor Displaying App View FileGnome Text Editor Displaying App Index FileWeb Browser Displaying Created Person ResultWeb Browser Displaying Remove Person PromptCustom View Records In Web Browser
Conclusion
This simple PHP PDO MVC application will help you understand how to interact with databases using PDO and build an MVC structure for easier maintenance and scaling. Now, you can add, update, and delete records from your database with ease!
Create Your First 3D Rotating Cube Using WebGL and HTML5
Are you curious about how to make interactive 3D graphics right in your browser? With HTML5 and WebGL, you can create stunning 3D objects without installing any extra software! Today, we’ll explore how to build a simple rotating cube using WebGL – perfect for beginners eager to dive into the world of 3D programming with JavaScript.
What is WebGL?
WebGL (Web Graphics Library) is a JavaScript API that lets you render 3D graphics inside the browser, using hardware acceleration. Unlike regular 2D canvas, WebGL taps directly into your GPU for smooth and powerful visuals.
This example demonstrates a cube that you can rotate by dragging your mouse, with basic lighting and a glass-like transparency effect. It’s a great project to understand how vertices, shaders, and transformation matrices work together in WebGL.
Interactive Demo
Here’s a live example of the Rotatable Cube in action:
If you enjoyed this project and want to deepen your JavaScript skills, check out my resources:
Book:Learning JavaScript – a beginner-friendly guide to mastering JavaScript programming from the ground up.
Course:Learning JavaScript – an online course that complements the book with hands-on lessons and coding exercises.
One-on-One Tutorials: Interested in personalized help? I offer private programming tutorials including JavaScript. Reach out via my contact page to schedule your session.
Final Thoughts
Building 3D graphics with WebGL might seem intimidating at first, but with practice and the right guidance, you can create amazing interactive experiences right inside the browser. This rotating cube is just the beginning – keep experimenting, and happy coding!
Feel free to leave your questions or share your own WebGL projects in the comments below!
Create a Cute Rubber Duck in Blender Using Python (No GUI) and Display It on Your Website
If you’re new to Blender and Python, this tutorial will guide you through creating a simple, stylized rubber duck entirely via Blender’s Python API – all from the command line without opening the Blender interface. Then, you’ll learn how to display the 3D duck model on your website using the <model-viewer> web component.
Step 1: Running Blender Python Script Without GUI
You can write a Blender Python script that generates the rubber duck model automatically. Instead of manually working in Blender’s GUI, you run Blender in background mode to execute your Python script.
Use this command in your terminal or command prompt (replace your_duck_script.py with your script’s filename):
blender --background --python your_duck_script.py
This runs Blender without the graphical interface, executes your script, and you can export the model directly in the script as a .glb or .gltf file.
Step 2: Export the Model Automatically
Include export code inside your Python script like this:
Git Ignore for Beginners: How to Keep Unwanted Files Out of Your Git Repositories
When working with Git, especially as a beginner, it is easy to accidentally include files in your repository that you do not actually want tracked. This can include log files, OS-generated files such as .DS_Store, compiled binaries, or personal settings files. Thankfully, Git has a built-in feature called .gitignore that allows you to easily tell Git which files or folders to ignore.
What is .gitignore?
The .gitignore file is a special plain text file that tells Git which files or directories to skip when committing code. For example, if you are working on a Python project and do not want to track compiled .pyc files or your virtual environment folder, .gitignore is your best friend.
Here is a basic example for a Python project:
__pycache__/
*.pyc
.env
venv/
Once listed, Git will ignore those files and not include them in future commits, even if they exist in your project folder.
Why Use .gitignore?
Security: Avoid committing sensitive files like API keys or passwords.
Clean Repositories: Keep your Git history free from clutter.
Consistency: Ensure only relevant code and files are shared with collaborators.
How to Create and Use a .gitignore File
Open your project root directory.
Create a new file named .gitignore.
Add paths or patterns for files and folders to ignore.
Save and commit your changes.
Note: If you have already committed a file, adding it to .gitignore will not remove it from the repository. You will need to untrack it manually using the following command:
git rm --cached filename
Screencast Tutorial & Screenshots
Command Line Git Ignore File CreationGnome Text Editor Showing Git Ignore For Python ProjectGnome Text Editor Showing Git Ignore For Widlcards And PatternsTerminal Showing Git Ignore Content FilesTerminal Showing Result of Git Status For Python ProjectScreencast Of Git Ignore
Learn More with My Programming Books
Want to go deeper into Git and other programming topics? Check out my books on Amazon:
Git is open source and free to use. The .gitignore system helps keep your open source or private projects tidy and secure. It is a simple but powerful step in maintaining clean code.
After years of faithful service, my old workstation finally stepped aside to make room for a new build, optimized for Blender and AI development in 2025. This isn’t just a hardware flex – it’s the foundation for my content creation, live Python tutorials, and machine learning training setups.
This post is a beginner-friendly overview of my new system build – from unboxing to assembly – and how it compares with my previous setup. Whether you’re upgrading your own rig or just curious about hardware for Blender and AI, I hope this helps.
New System Build (2025)
The new workstation is built into the same Deepcool Tesseract BF case as the previous system. However, it has received a cooling overhaul, and most components have been swapped or upgraded:
Parts List:
Case: Deepcool Tesseract BF (reused) with original 1 fan
Cooling: 5x Thermalright TL-S12 120mm fans
Motherboard: Used MSI B550-A PRO ProSeries
CPU: AMD Ryzen 5 5600GT with integrated graphics
RAM: 32GB (2x16GB) Timetec DDR4 2133MHz
GPU: AMD Instinct MI60 (pending installation)
Power Supply: SAMA G850W (ATX 3.1, future-ready)
It runs Fedora 42 – a modern, secure, and fast Linux distro that’s perfect for development and creative workflows.
Why This Matters for Blender and AI
Even without the discrete GPU (MI60) connected yet, this system is already a serious upgrade in CPU, RAM, and thermal performance.
Faster rendering in Blender (especially Eevee and simple Cycles scenes)
Better performance in AI model training, dataset preprocessing, and inference
Smooth multitasking for development, recording, and streaming
Quieter operation thanks to high-quality case fans
Once the MI60 is installed, it will unlock massive parallel compute for Blender GPU rendering and deep learning acceleration.
Old vs. New: Theoretical Comparison
Component
Old System
New System
Difference
CPU
Intel Core i7-2600 (4C/8T, 3.4GHz)
AMD Ryzen 5 5600GT (6C/12T, 3.6GHz)
~2x performance
RAM
16GB DDR3 1333MHz (mixed)
32GB DDR4 2133MHz (matched)
Double capacity & faster
GPU
EVGA GTX 950 (2GB)
AMD Instinct MI60 (32GB HBM2)
Massive upgrade pending
Motherboard
ASUS P8Z68-V LX
MSI B550-A PRO
DDR4, PCIe 4.0 support
Cooling
1 stock fan
5x TL-S12 fans + stock
Improved airflow
Power Supply
Thermaltake TR2 600W
SAMA G850W ATX 3.1
Future-ready, more efficient
Integrated Graphics Comparison: Intel HD 2000 vs. AMD Radeon Graphics (5600GT)
Even without the AMD Instinct MI60 GPU installed, the Ryzen 5 5600GT’s integrated graphics make a huge difference over the old i7-2600’s Intel HD 2000. Here’s a look at the improvements and what it means for real-world use in Blender, media, and AI experiments.
Feature Comparison
Feature
Intel HD Graphics 2000
AMD Radeon (Ryzen 5 5600GT)
Difference
Release Year
2011
2024
13+ years newer
Architecture
Intel Gen6
AMD RDNA2
Modern design
Shaders
6 EUs (~96 shaders)
7 CUs (448 shaders)
~4.5x more
Clock Speed
~850 MHz
~1.9 GHz
2x faster
Video Decoding
Basic H.264
H.264, H.265, VP9, AV1
4K-ready
Linux Drivers
Legacy
AMDGPU (Mesa)
Actively maintained
Viewport & Media Benchmarks
Task
Intel HD 2000
AMD Radeon iGPU
Notes
Blender Viewport (Eevee)
4-10 FPS
35-60 FPS
Usable without discrete GPU
Blender Viewport (Cycles CPU fallback)
~3 FPS
~15 FPS
Still CPU-bound
1080p Video Playback
OK
Excellent
Both handle well
4K AV1 Playback
Not supported
Supported
Smoother performance
glmark2 Vulkan Score
~400
~3000
Significant improvement
Screenshots And Screencast
SAMA G850W ATX 3.1 Power Supply And Deepcool Tesseract BF CaseMSI B550-A PRO Motherboard, Timetec 32GB KIT(2x16GB) DDR4 2133MHz Ram, Deepcool Tesseract BF Case BackScreencast Showing setup on Fedora 42, system detection, system stats
This new build may not be flashy with RGBs or liquid cooling, but it’s powerful, efficient, and practical for the kind of work I do – Python development, Blender projects, and AI training.
If you’re building your own creative workstation in 2025, feel free to reach out or comment below with questions.
Next Post Preview
Installing Blender + AI Tools on Fedora 42 (2025 Setup Guide)