Become a sponsor to Gabriel Tremeschin
Important: There's no easy integration between GitHub Sponsors and Discord, tag me @Tremeschin to get your member role, where you shall receive the benefits, will automate the process later!
Hi, I'm Gabriel, born and currently living in Brazil 🇧🇷
While my major is electrical engineering, I've been working on Open Source Software in most of my free time ever since 2020, and want to make it a profession with such growing community, passion and potential.
✅ Scroll down for the projects I'm working on!
I'd do all of this for free, but your support keeps me going! Chose a tier on the right that best fits your needs, get instant access to the rewards and value I'm adding to your life :)
Note: This may sound naive, but my local currency and average income are so weak that $100 USD amounts to 50% of the minimum monthly wage here. This is a value you'd spend on a two's dinner!
Featured work
-
BrokenSource/DepthFlow
🌊 Images to → 3D Parallax effect video. A free and open source ImmersityAI alternative
Python 861 -
BrokenSource/Pyaket
📦 Easy Python to Fast Executables
Rust 22 -
BrokenSource/ShaderFlow
🔥 Modular shader engine designed for simplicity and speed
Python 121 -
BrokenSource/Pianola
🎹 World's finest piano roll visualizer, where elegance meets motion.
GLSL 27 -
BrokenSource/TurboPipe
🌀 Faster ModernGL Buffers data transfers for subprocesses
C++ 4 -
BrokenSource/BrokenSource
❤️🩹 The Monorepo: House of all projects and the main shared library
Python 47
$100 one time
Select🍷 Make my day
An unusually popular option for a thanks!
Felt welcomed in the community, got the fastest responses to your questions, or noticed that the software has added significant value to you?
This popular one-time option motivates me so much to continue, chances are you'll be forever remembered by me!
Note: Due technical reasons, one-times can't have access to the Insiders repository
$200 one time
Select⭐️ Home server (PSU+SSD)
🇧🇷 Local prices are this bad, 100% import tax..
I have my first desktop's R3 1300X 4c/4t and GTX 1050 Ti laying around with 8 GB RAM. It's not much, but I can upgrade my main desktop's crappy PSU, get an SSD, and have a home server, where:
- I can host a free WebUI instance for projects like DepthFlow, perhaps even for the public, or a free API.
- Internal infrastructure like a Discord bot for roles management or projects too!
$600 one time
Select⭐️ Buy me an Intel Arc GPU
🇧🇷 Local prices are this bad, 100% import tax..
I went with an RTX 3060 12 GB as a safe choice for development and light gaming, so I don't have access to any Intel hardware for ensuring the code and PyTorch models works there. Blindly developing for it and relying on user feedback can only go so far.
- I'm aiming for an B580, valuing VRAM and official OneAPI/XPU compatibility, which I will always support onwards.
- I'll extensively test the code on Intel GPUs in Linux/Windows to ensure it works on different platform and drivers.
- I can use the NVIDIA or Intel in self hosted WebUIs of my projects, available for free for supporters, perhaps public.
$1,400 one time
Select⭐️ Buy me an AMD GPU
🇧🇷 Local prices are this bad, 100% import tax..
I went with an RTX 3060 12 GB as a safe choice for development and light gaming, so I don't have access to any AMD hardware for ensuring the code and PyTorch models works there. Blindly developing for it and relying on user feedback can only go so far.
- I'm aiming for an 9070 XT / 7900 XT or XTX, valuing VRAM and official ROCm compatibility, which I will always upstream support for.
- I'll extensively test the code on AMD GPUs in Linux/Windows to ensure it works on different platform and drivers.
- I can use the NVIDIA or AMD in self hosted WebUIs of my projects, available for free for supporters, perhaps public.
$2,500 one time
Select⭐️ Buy me a Mac mini M4
🇧🇷 Local prices are this bad, 100% import tax..
I don't have access to any Apple Silicon hardware (which composes a significant percentage of USA users) to test the code, so there's lots of rough edges around. Blindly developing for it and relying on user feedback can only go so far, specially because Apple's OpenGL implementation and ARM architecture changes that much.
- I'll review and improve the whole code-base upstreaming macOS support, bug fixes and quality of life updates
- I'm aiming for the 32 GB RAM base model with 512 GB SSD. An usual day involves running PyTorch models, Video Encoding in High Resolutions, VSCode, Browser. I average ~24GB on Linux, so 32 GB is my minimum.