09-29-2026, 07:29 AM
That matches what I was seeing. The pattern around prefill/decode split has been pretty similar for me. The biggest change for me was simplifying the stack instead of adding one more optimization layer.
|
how much orchestration is too much
|
|
« Next Oldest | Next Newest »
|
| Messages In This Thread |
|
how much orchestration is too much - by llm_wren - 09-25-2026, 02:01 AM
Re: how much orchestration is too much - by llm_wren - 09-25-2026, 02:04 PM
Re: how much orchestration is too much - by southden - 09-30-2026, 07:53 AM
Re: how much orchestration is too much - by vramstream - 09-25-2026, 06:24 PM
Re: how much orchestration is too much - by llm_wren - 09-25-2026, 08:07 PM
Re: how much orchestration is too much - by ggufvault - 09-26-2026, 05:52 AM
Re: how much orchestration is too much - by cudacrate - 09-25-2026, 08:19 PM
Re: how much orchestration is too much - by bytegarden73 - 09-25-2026, 09:18 PM
Re: how much orchestration is too much - by llm_wren - 09-26-2026, 05:27 AM
Re: how much orchestration is too much - by cudacrate - 09-26-2026, 06:31 AM
Re: how much orchestration is too much - by llm_wren - 09-29-2026, 06:38 AM
Re: how much orchestration is too much - by oxidegrove40 - 09-29-2026, 07:29 AM
|
Users browsing this thread: 1 Guest(s)