2 ms·

I'm not GP, but I would want a benchmark that actually tests the entire context window. A benchmark that only tests the first 128K tokens effectively tells us n

I'm not GP, but I would want a benchmark that actually tests the entire context window. A benchmark that only tests the first 128K tokens effectively tells us nothing about how well it works at its full capacity.

alexsubq 5mo ago
That makes sense! We are working on that.