3 ms·
One thing I’ve noticed with local models is that people tolerate a lot more trial and error behavior. When a hosted model wastes tokens it feels expensive, but
by lzaborowski 7mo ago
One thing I’ve noticed with local models is that people tolerate a lot more trial and error behavior. When a hosted model wastes tokens it feels expensive, but when a local model loops a bit it just feels like it’s “thinking.”
If models like Qwen can get good enough for coding tasks locally, the real shift might be economic rather than purely capability.
- trvz 7mo agoWasted tokens are preferred for local models, I need the GPU mainframe in my bedroom to heat it as I live in a third world country with unreliable heating (Switzerland).
- MarsIronPI 7mo agohttps://xkcd.com/1172 https://xkcd.com/1172