2 ms·
The problem is there are no public benchmarks usually so it is hard to really compare on long context lengths to see if they are still performing equally intell
by whimsicalism 4y ago
The problem is there are no public benchmarks usually so it is hard to really compare on long context lengths to see if they are still performing equally intelligent of tasks.