Post #2365614

@mpowney@mastodon.au @florenciocano@infosec.exchange That could be it too. Large context wasn’t as much of an option. It’s been a few weeks since I’ve messed with it at all, maybe I can try tuning it a bit to see if it has better results. I’m running the LLMs on a laptop 4090, which has 16G dedicated, and supposedly can use the other 16G, though it never seems like ollama allocates it, and I think I had a crash when trying something bigger. I’ll try adjusting the max context window if I can.