4 ms·
With the 7B parameter models, sure. Not with GPT-2 though. The context window is only 1024 tokens. Even with only 10 messages, if they're long messages, it w
by Sohcahtoa82 2y ago
With the 7B parameter models, sure.
Not with GPT-2 though. The context window is only 1024 tokens. Even with only 10 messages, if they're long messages, it will exceed the context window.