ochafik

7 ms·

1. 1.▲
   by ochafik 1y ago

   Thanks for the kind words! I've indeed done all that on my spare time (still under Google copyright), very happy to see this used and appreciated :-) About to start a new job / unsure if I'll be able to contribute more, but i
2. 2.▲
   Llama.cpp now supports tool calling (OpenAI-compatible)(github.com)

   3 points by ochafik 2y ago|1 comments
3. 3.▲
   by ochafik 2y ago

   Supports 8+ models’ native formats (Llama, Functionary, Hermes, Mistral, Firefunction, DeepSeek) + generic fallback for all other models. Runs anywhere; on a Mac: brew install llama.cpp llama-server --jinja -fa -hf bartowski/Qwen
4. 4.▲
   by ochafik 2y ago

   When there's a will, there's a way. It's a bit silly but I rolled out my own(*) "no-deps" C++ Jinja template library ( https://github.com/google/minja ) just to add tool call support to llama.c
5. 5.▲
   OpenSCAD's Web Playground got upgraded (still Alpha)(ochafik.com)

   1 points by ochafik 3y ago|1 comments
6. 6.▲
   by ochafik 3y ago

   Now much faster and more stable thanks to the Manifold engine ( https://github.com/elalish/manifold ): in fact it is often faster than the last stable desktop release from https://openscad.org/ . Also muc
7. 7.▲
   by ochafik 5y ago

   Aww thanks, happy it's helping others!