jackxlau

6 ms·

1. 1.▲
   Show HN: FusionHarness – An Open-source Mixture-of-Agents compound-model server(github.com)

   5 points by jackxlau 3mo ago|1 comments
2. 2.▲
   by jackxlau 3mo ago

   I was looking for a solution for this thanks!
3. 3.▲
   by jackxlau 3mo ago

   this is neat
4. 4.▲
   by jackxlau 3mo ago

   In my own testing I have seen peak performance happen usually within 15-20% of the intended context limit, albeit there are a few optimizations depending on the task quality.
5. 5.▲
   by jackxlau 3mo ago

   the handshake process to unlock the A100 cores is pretty interesting it reminds me of driving CPU cores like a GPU command queue.
6. 6.▲
   by jackxlau 3mo ago

   I came across the conclusion here since a change sometimes spans several repos, per-repo history optimizes the wrong target.