[pull] master from tensorflow:master - #8790

Merged

pull[bot] merged 12 commits into

Cache-Cloud:masterfrom

tensorflow:master

Aug 29, 2026

Commits on Aug 29, 2026

[IFRT] Use canonical custom layouts on array_impl tests

hyeontaek
authored and
tensorflower-gardener
committed
a4b6c56
Clarify in tf.sequence_mask documentation that maxlen must be a compile-time constant when compiling under XLA.

dmiltr3
authored and
tensorflower-gardener
committed
1e2aa2d
log mlir/triton compile time

ermilovmaxim
authored and
tensorflower-gardener
committed
96c9a61
Preserve send/recv HloInstruction shape when deserializing from proto.

Michael Wong
authored and
tensorflower-gardener
committed
174f0ab
Rename propagate-qsv pass to propagate-qparams and add VHLO quant custom call legalization

majiddadashi
authored and
tensorflower-gardener
committed
2edfe1b
Remove references to StreamExecutorGpuClient.

pschuh
authored and
tensorflower-gardener
committed
a4a8ad3
Clamp propagated constraints across convert operations to the operand's type range to prevent Inf/NaN.

bhatuzdaname
authored and
tensorflower-gardener
committed
bfd5c52
Automated Code Change
tensorflower-gardener
committed
6eefa0c
[IFRT IR] Include the MLIR location in the user contexts created for IRRT IR ops

ICGog
authored and
tensorflower-gardener
committed
eb11ee5
Automated Code Change
tensorflower-gardener
committed
2ba1ec5
Automated Code Change
tensorflower-gardener
committed
da79958
Automated Code Change
tensorflower-gardener
committed
1397bef