Bin Bao

[AOTI] Add missing std::move for constant args

1

Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * #125397 * __->__ #125329 Summary: fix https://github.com/pytorch/pytorch/issues/123187 cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @peterbell10 @ipiszy @yf225 @chenyang78 @kadeng...

[AOTI] Generate mul_Scalar instead of mul_Tensor

2

Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * __->__ #125397 * #125329 Summary: Fix https://github.com/pytorch/pytorch/issues/117365. When the second argument to aten.mul.Tensor is a scalar (e.g. scale factor), the cpp wrapper expects...

[AOTI] Fix 'int' object is not subscriptable

1

Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * #125962 * #125928 * __->__ #125731 * #125730 * #125291 Summary: for https://github.com/pytorch/pytorch/issues/117369 cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv...

[AOTI][torchgen] Update NativeFunctionsGroup mapping

1

Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * __->__ #125962 * #125928 * #125731 * #125730 * #125291 Summary: When looking up for what backend call to use for a fallback...

[AOTI][torchgen] Minor improvements to C shim torchgen

1

Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * #125962 * __->__ #125928 * #125731 * #125730 * #125291 Summary: Make some improvements to https://github.com/pytorch/pytorch/pull/125589 * Add a .default suffix to default...

[AOTI] Fix a codegen issue when .item() is used for kernel arg

1

Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * __->__ #126575 * #126369 Summary: fixes https://github.com/pytorch/pytorch/issues/126574 cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @peterbell10 @ipiszy @yf225 @chenyang78 @kadeng...

[AOTI] Codegen None as empty tensor

1

Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * #126575 * __->__ #126369 Summary: When None denotes an optional tensor, we codegen NULL to represent it; but when None is for actual...

After adding the device selection support, the commit ID drop list seems shared among all the devices. For example, if I run a data collection for x86, that specific commit...

[AOTI] Add a --max-seq-length option for export

1

Summary: This improves best tokens/sec from 73 to 85.

[expriment][not for review] Evaluate the impact of turning off pre_gr…

1

cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @ipiszy @yf225 @chenyang78 @kadeng @muchulee8 @ColinPeppler @amjames @chauhang