Support quantized tensor subclasses across Python and C++ - #3393

Open

dingqingy-nv wants to merge 4 commits into

NVIDIA:mainfrom

dingqingy-nv:codex/te-preserve-quantized-subclass-detach

Commits on Aug 17, 2026

Preserve quantized tensor subclasses on detach
dingqingy-nv
committed
9e74533
Fix IdentityTensor detach alias assertion
dingqingy-nv
committed
865fda0

Commits on Aug 18, 2026

Support quantized tensor subclasses in C++ bindings
dingqingy-nv
committed
bf64b4e
Centralize quantized tensor detach
dingqingy-nv
committed
ceb1c8b