* power draw first
* speedup encoder
* speed up hardware test suites
* refresh IRQ actions after hardware setup
* skip unused big model in device checkout
* move LFS pruning off device checkout path
* wtf is that for?
* reverts
* locality
* rm that
* Reapply "usbgpu: never materialize model weights in RAM" (#38271)
This reverts commit 8196b743af.
* release oob buffers as we load them
* don't leak fds
* lint
* 10s faster loading time for big model
* bigger big
* don't load big model in ram
* patch tg to allow out of band buffers
* dump/load oob
* roundtrip pkl to disk
* streaming chunking
* to_disk
* fixup
* linter
* Revert "bigger big"
This reverts commit 7a8147dd7fab88db963d512ab4008de940714025.
Adeeb Shihadeh