PicoLM is an LLM inference engine written in C99. It currently supports llama-2, GPT-2, Qwen 3. 6/3.
Source: [Hacker News](https://github.com/whoreson/picolm/)
2 points, 0 comments on Hacker News
The Equality and Human Rights Commission says the WRU put in "hard yards" to improve its culture.
1 points, 0 comments on Hacker News
I needed to generate a large number of images using Nano Banana 2, using reference images. Playing with OpenRouter I discovered that most image gen models use similar params, so I made something I could use in a pinch to work with said params, without bothering with OpenRouter's playground. I u...
1 points, 0 comments on Hacker News
Tired of paying for Suno, I made this local desktop app for Windows and macOS to generate songs through MiniMax's hosted Music API (free for up to 3 songs/minute! ). I wanted something tailor-made for a songwriter, so I implemented some QoL features to make rhyme-making easier.