Add support for aten._fused_rms_norm operator#4015

Open

SandSnip3r wants to merge 19 commits intopytorch:mainfrom

SandSnip3r:rms_norm_converter

Commits on Jan 8, 2026

Commits on Jan 13, 2026

Merge remote-tracking branch 'upstream/main'
SandSnip3r
committed

Commits on Jan 15, 2026

Commits on Jan 21, 2026

Merge branch 'pytorch:main' into rms_norm_converter
SandSnip3r
authored

Commits on Jan 26, 2026

Fix the converter issue caused by this missing unset_fake_temporarily (#4006)

wenbingl
authored and
SandSnip3r
committed
chore: update uv.lock

Torch-TensorRT Github Bot
authored and
SandSnip3r
committed
lowering pass: fully remove SymInt (#4001)

zewenli98
authored and
SandSnip3r
committed

Commits on Feb 10, 2026

fix the layer info test failure and deal with potential segfault (#4042)

authored and
SandSnip3r
committed
cherry pick 4033: skip llm test if modelopt is not installed from release branch to main (#4034)

lanluo-nvidia
authored and
SandSnip3r
committed
cherry pick 4038 from ngc release branch to main: skip failed test on orin until issue 3982 is fixed (#4039)

lanluo-nvidia
authored and
SandSnip3r
committed
cherry pick 4028: fix resource partitioner issue from release branch to main (#4031)

lanluo-nvidia
authored and
SandSnip3r
committed
cherry pick 4029: upgrade aarch64 base image from release branch to main (#4030)

lanluo-nvidia
authored and
SandSnip3r
committed
fix: torchtrtc precision setting logic (#3883)

yeetypete
authored and
SandSnip3r
committed
Empty tensor handling (#3891)

apbose
authored and
SandSnip3r
committed
chore: update uv.lock

Torch-TensorRT Github Bot
authored and
SandSnip3r
committed
fix: example argument issue raised in 4070 (#4071)

zewenli98
authored and
SandSnip3r
committed