🏛️ TempLLM OS

TempLLMOS is a WebOS entirely described in natural language and seamlessly compiled into JavaScript using an LLM.

Outated animation of the OS

🥇 Sample

Here is how to code a PhotoBooth Application that takes snapshots and saves them into the Virtual FS

#include filesystem
#include run


create a local photoBooth function that takes no parameters:
- create a root dom element
- append webcam video element to the root element that takes 100% of the root width
- add a button with text "Take Snapshot" that takes an image snapshot of the video and saves it to "photobooth/image+{timestamp}.png"
- return root dom element

--- 
run application with photoBooth function reference callback

🏃‍➡️ Give it a try:

There are two working modes for the experimental os: Static runtime where the precompiled code is present and injected in the root HTML and the dynamic realtime compiler mode where a server that hosts a LLM model is published that allows the user to compile new apps at runtime.

• Static Runtime on GitHub pages [Ide & Run won't work]
• Static Runtime: Just open the templllmos.html file [Ide and Run won't work]
• Realtime Compiler: python3 templlmos.py --serve --listen and open http://localhost:8080 in a browser

For runtime compiler you need to install dependencies:

python3 -m pip install huggingface_hub llama-cpp-python watchdog

for MAC runtime you can benefit from GPU inference and install llama-cpp with METAL enabled:

CMAKE_ARGS="-DLLAMA_METAL=on" python3 -m pip install llama-cpp-python --upgrade --force-reinstall --no-cache-dir

✅ Next Steps

• Increase output stability
• Fix CPU vs. GPU generation differences
• Better WebUI
• Window Management & Taskbar
• Fine tune model on JS Only

❤️ Contribute

We believe in the power of collaboration and welcome contributions from developers of all skill levels. Whether it's fixing bugs, improving documentation, or suggesting new features, your help makes a big difference!