I m Gian Luca, CMO at TPN Labs. The past years I've been digging into the LLMs rabbit hole, and as of today, I like to think that I work at the edge of technology. Everyday is a lot of work, but seeing results is what matters. We build tools around making AI models smaller without making them worse. With TPN we've built a LLM compression and optimization engine powered by an open competition where people compete to compress models, a benchmarking platform to measure what survives, and soon a way for anyone to actually run the results on their own machine.
Mostly here to learn from how other people and see cool stuff happening. Happy to answer anything about local models or compression, it s a weirdly deep rabbit hole.
