7/12/2026 at 9:17:27 PM
by knrz
7/15/2026 at 8:13:38 PM
looks cool. what is latency like? I haven't used qmd before and it looks like it runs 3 local models.by rgbrgb
7/15/2026 at 8:57:08 PM
as always it depends on your hardware, the tiny models for embedding / reranking typically have low latencyqmd is focussed on local to the point of designing around single machine setups and this creates a gap where one runs agents+qmd on their laptop and LLMs on their ai box
by verdverm
7/12/2026 at 11:26:02 PM
if this were localai, you could just figure out how to update the memore if the fill changes.by cyanydeez
7/15/2026 at 8:28:58 PM
yup, one of the reasons I built gmd to replace qmd, also that I wanted to use Typesense and incorporate llm-wiki concepts. Updating memory does depend on the nature of the change and may impact multiple memories or indexed files, and why llm-wiki concepts are needed too.Right now, imo, the main reason to build something like this or a harness is to understand the quirks son you can evaluate production grade implementations as the space matures. We are all still very early in the curve.
by verdverm
7/12/2026 at 10:02:48 PM
[flagged]by zane_shu
7/15/2026 at 8:45:23 PM
We use capn crunch (for cerealization)For crypto/stock transfers we ofc use chips n dips
by grapefruitsoda