Posts on Jan 1970

Home 2025 August
Finally! OpenAI Enters the Open-Source Arena with Two New Models
By Dimitri Bellini07 August, 2025Blog

Finally! OpenAI Enters the Open-Source Arena with Two New Models

Good morning, everyone! Dimitri Bellini here, and welcome back to Quadrata. For a while now, I’ve been waiting for something genuinely new to discuss in the world of artificial intelligence. The on-premise, open-source scene has been buzzing, but largely dominated by excellent models from the East. I was waiting for a major American player to make a move, and finally, the moment has arrived. OpenAI, the minds behind ChatGPT, have released not one, but two completely open-source models. This is a big deal, and in this post, I’m going to break down what they are, what they can do, and put them to the test myself.

What’s New from OpenAI? A Revolution in the Making

OpenAI has released two “open-weight” models, which means we have access to the model’s core infrastructure and the data it was trained on. This is fantastic news for developers, researchers, and hobbyists like us, as it allows for deep customization. The two new models are:

  • GPT-OSS-120B: A massive 120-billion parameter model.

  • GPT-OSS-20B: A more accessible 20-billion parameter model.

This move is a significant step, especially with a permissive Apache 2.0 license, which allows for commercial use. You can build on top of these models, fine-tune them with your own data, and deploy them in your applications without the heavy licensing restrictions we often see.

Key Features That Matter

So, what makes these models stand out? Here are the highlights:

  • Truly Open License: The Apache 2.0 license gives you immense freedom to innovate and even commercialize your work.

  • Designed for Agentic Tasks: These models are built to be “agents” that can interact with tools and perform complex, multi-step tasks. While the term “agentic” is a bit of a buzzword lately, the potential is there.

  • Deeply Customizable: With open weights, you can perform post-training to tailor the model to your specific needs, creating a specialized LLM for your unique use case.

  • Full Chain of Thought: A major point of contention with closed models is their “black box” nature. You get an answer but can’t see the reasoning. These models expose their entire thought process, allowing you to understand why they reached a certain conclusion. This transparency is crucial for debugging and trust.

Choosing Your Model: Hardware and Performance

The two models cater to very different hardware capabilities.

The Powerhouse: GPT-OSS-120B

This is the star of the show, with performance comparable to the closed GPT-3.5-Turbo model. However, running it is no small feat. You’ll need some serious hardware, like an NVIDIA H100 GPU with at least 80GB of VRAM. This is not something most of us have at home, but it’s a game-changer for businesses and researchers with the right infrastructure.

The People’s Model: GPT-OSS-20B

This is the model most of us can experiment with. It’s designed to be more “human-scale” and offers performance roughly equivalent to the `o3-mini` model. The hardware requirements are much more reasonable:

  • At least 16GB of VRAM on a dedicated NVIDIA GPU.

  • A tool like Ollama or vLLM to run it (at the time of writing, Ollama already has full support!).

This is the model I’ll be focusing my tests on today.

My Hands-On Test: Putting GPT-OSS-20B to Work with Zabbix

Benchmarks are one thing, but real-world performance is what truly counts. I decided to throw a few complex, Zabbix-related challenges at the 20B model to see how it would handle them. I used LM Arena to compare its output side-by-side with another strong model of a similar size, Qwen2.

Test 1: Zabbix JavaScript Preprocessing

My first test was a niche one: I asked the model to write a Zabbix JavaScript preprocessing script to modify the output of a low-level discovery rule by adding a custom user macro. This isn’t a simple “hello world” prompt; it requires an understanding of Zabbix’s specific architecture, LLD, and JavaScript context.

The Result: I have to say, both models did an impressive job. They understood the context of Zabbix, preprocessing, and discovery macros. The JavaScript they generated was coherent and almost perfect. The GPT-OSS model’s code needed a slight tweak—it wrapped the code in a function, which isn’t necessary in Zabbix, and made a small assumption about input parameters. However, with a minor correction, the code worked. Not bad at all for a model running locally!

Test 2: Root Cause Analysis of IT Events

Next, I gave the model a set of correlated IT events with timestamps and asked it to identify the root cause. The events were:

  1. Filesystem full on a host

  2. Database instance down

  3. CRM application down

  4. Host unreachable

The Result: This is where the model’s reasoning really shone. It correctly identified that the “Filesystem full” event was the most likely root cause. It reasoned that a full disk could cause the database to crash, which in turn would bring down the CRM application that depends on it. It correctly identified the chain of dependencies. Both GPT-OSS and Qwen2 passed this test with flying colors, demonstrating strong logical reasoning.

Test 3: The Agentic Challenge

For my final test, I tried to push the “agentic” capabilities. I provided the model with a tool to interact with the Zabbix API and asked it to fetch a list of active problems. Unfortunately, this is where it stumbled. While it understood the request and even defined the tool it needed to use, it failed to actually execute the API call, instead getting stuck or hallucinating functions. This shows that while the potential for tool use is there, the implementation isn’t quite seamless yet, at least in my initial tests.

Conclusion: A Welcome and Necessary Step Forward

So, what’s my final verdict? The release of these open-source models by OpenAI is a fantastic and much-needed development. It provides a powerful, transparent, and highly customizable alternative from a Western company in a space that was becoming increasingly dominated by others. The 20B model is a solid performer, capable of impressive reasoning and coding, even if it has some rough edges with more advanced agentic tasks.

For now, it stands as another great option alongside models from Mistral and others. The true power here lies in the community. With open weights and an open license, I’m excited to see how developers will improve, fine-tune, and build upon this foundation. This is a very interesting time for local and on-premise AI.

What do you think? Have you tried the new models? What are your impressions? Let me know your thoughts in the comments below!

Stay Connected with Me:

Read More

0

Copyparty: The Lightweight, Powerful File Server You Didn’t Know You Needed
By Dimitri Bellini01 August, 2025Blog

Copyparty: The Lightweight, Powerful File Server You Didn’t Know You Needed

Good morning, everyone, and welcome back to Quadrata! This is my corner of the internet dedicated to the open-source world and the IT solutions that I—and hopefully you—find exciting. If you enjoy this kind of content, don’t forget to leave a like on the video and subscribe to the channel!

This week, we’re diving back into the world of open-source solutions. I stumbled upon a truly stunning tool in the file-sharing space that has a wonderful nostalgic feel, reminiscent of the BBS days of the 90s. It’s called Copyparty, and its charm lies not just in its retro vibe but in its incredible versatility. You can install it almost anywhere, making it a fantastic utility to have in your toolkit.

So, let’s take a closer look together.

What Exactly is Copyparty?

At its core, Copyparty is a web file server that allows you to share and exchange files. What makes it special is that it’s all contained within a single Python file. This makes it incredibly lightweight and portable. While you can run it directly, I prefer using it inside a Docker container for easier management and deployment.

But why use it? The answer is simplicity and performance. If you’ve ever needed to quickly move files between your PC and your NAS, or share a large file with a friend without jumping through hoops, Copyparty could be the perfect, high-performing solution for you.

A Surprising Number of Features in a Tiny Package

I was genuinely impressed by the sheer number of features packed into this tool. It’s highly customizable and offers much more than simple file transfers. Here’s a condensed list of its most interesting capabilities:

  • Smart Uploads & Downloads: When you upload a large file, Copyparty can intelligently split it into smaller chunks. This maximizes your bandwidth and, more importantly, allows for resumable transfers. If your connection drops halfway through, you can pick up right where you left off.

  • File Deduplication: To save precious disk space, Copyparty uses file hashes to identify and avoid storing duplicate files.

  • On-the-fly Compression: You can have files automatically zipped and compressed on the fly, which is another great space-saving feature.

  • Batch Renaming & Tagging: If you have a large collection of photos or, like in the old days, a folder full of MP3s, you can quickly rename them based on a specific pattern.

  • Extensive Protocol Support: It’s not just limited to HTTP. Copyparty supports a whole suite of protocols, including WebDAV, FTPS, TFTP, and Samba, making it a complete hub for file communication.

  • Truly Cross-Platform: It runs virtually everywhere: Linux, macOS, Windows, Android, and even on a Raspberry Pi, thanks to its optimized nature. Yes, you can install it directly on your phone!

  • Built-in Media Tools: Copyparty includes a surprisingly nice music player that can read metadata from your audio files (like BPM and duration) and a compact image browser for viewing your photos.

  • Powerful Command Line (CLI): For those who need to automate or optimize file transfers, there’s a full-featured command-line interface.

Tailor It to Your Needs: Configuration and Security

One of Copyparty’s greatest strengths is its customizability via a single configuration file, copyparty.conf. Here, you can enable or disable features, block connections from specific IP ranges, set upload limits based on disk space, and even change the UI theme.

For user management, you have a couple of options. You can use a simple user/password file or integrate with an external Identity Provider (IDP). The permission system is also very granular. Using a system of flags (like RW for read/write, MDA, etc.), you can define exactly what each user can do on specific paths. It might seem a bit “primordial” compared to modern web GUIs, but for a compact solution, it’s incredibly fast and effective to manage.

How to Install Copyparty with Docker

As I mentioned, my preferred method is using Docker. Copyparty’s developers provide a straightforward Docker Compose file that makes getting started a breeze. I use a GUI tool like Portainer to manage my containers, which simplifies the process even further.

Here’s a look at a basic docker-compose.yml structure:



services:

  copyparty:

    image: 9001/copyparty

    ports:

      - "3923:3923"

    volumes:

      # Volume for configuration file (copyparty.conf)

      - /path/to/your/config:/cfg

      # Volume for the files you want to share

      - /path/to/your/data:/mnt

    # ... other docker-specific configurations

In this setup, I’ve defined two key volumes:

  1. A volume for the configuration, where the copyparty.conf file lives.

  2. A mount point for the actual data I want to share or upload to.

Once you run docker-compose up -d, your service will be up and running!

A Walkthrough of the Web Interface

The official GitHub page has a wealth of information and even a live demo, but let me show you my installation. The interface has a fantastic vintage feel, but it’s packed with functionality.

Uploading and Sharing

Uploading a file is as simple as dragging and dropping. First, Copyparty hashes the file to check for duplicates. Then, it begins streaming the upload in a highly optimized way. Once uploaded, you’ll see details like the IP address it was uploaded from and the timestamp.

Sharing is just as easy. You can select a file, create a share link with a custom name, set a password, and even define an expiration date. It generates both a URL and a QR code, making it incredibly convenient to share with others.

Management and Media

The UI includes several helpful tools:

  • Control Center: See active clients, current uploads/downloads, and active shares.

  • Recent Uploads (Extinguisher Icon): Quickly view the latest files added to your share, which is useful for moderation in a multi-user environment.

  • Advanced Search (Lens Icon): A powerful search tool with a wide array of filters to find exactly what you’re looking for.

  • Settings (Gear Icon): Customize the UI, change the language, and tweak how files are displayed.

And don’t forget the built-in media player and image gallery, which turn your file share into a simple media server.

Monitoring

For advanced users, Copyparty can even export its metrics, allowing you to monitor its performance and status with tools like Zabbix. This is a testament to its professional-grade design.

Final Thoughts: Is Copyparty Right for You?

I think Copyparty is a fantastic and interesting product. It’s a very nice solution to try, especially because it’s so lightweight and can be installed almost anywhere. There are many situations where a fast, simple, and self-hosted file-sharing tool is exactly what you need.

Its blend of retro simplicity and modern, powerful features makes it a unique and valuable tool in the open-source world.

That’s all for this week! I’m always eager to hear your thoughts. Have you used Copyparty before? Or do you use another solution that you find more interesting? Let me know in the comments below—perhaps we can discuss it in a future video!

A big greeting from me, Dimitri, and see you next week. Bye everyone!

Follow my work and join the community:

Read More

0