Practice
Token reduction · part 1
The Cache Is Not the Only Token Budget
A real hook payload fell from 1790 to 826 bytes, 964 bytes smaller. The fix was not compression after the fact, but removing redundant context at the emitter.
Read
▤
Category
Practice
Posts in this category: 5
← Tous les articles
Practice
Gates that lie · part 4
A Reboot Is Not a Health Check
The stack passed 37/37 checks after a reboot, including public routes and the consent to analytics path. A running container alone would not have proved that.
Read
The GPU Is a Queue
A day with 135 GPU jobs and 34 failures, none of which needed the GPU to find. The fix is not a faster GPU, it's treating the card as a shared, budgeted resource with a preflight step before anything touches it.
Read
Publishing Without Publishing Yourself
361 commits in mojo-baro's history carried a private email address that is invisible in every checkout of the repo, because it lives in commit metadata, not tracked files. The Android repo separately carried 112 MB of committed build output. Neither would have been caught by reading the code.
Read
Keeping the Failures In
BARO_SPEC defaults to on. Setting it on the MoE model changes nothing, because the actual gate in the code is BARO_SPEC and a model flag that is always false for that architecture, and the document says so in the same line rather than in a footnote.
Read