projects / platform / upstream / pytorch.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 1f35d20) | patch
[TensorExpr] LLVMCodegen: fix lowering for UInt->Float casts. (#64862)
authorMikhail Zolotukhin <mvz@fb.com>
Sat, 11 Sep 2021 16:23:10 +0000 (09:23 -0700)
committerFacebook GitHub Bot <facebook-github-bot@users.noreply.github.com>
Sat, 11 Sep 2021 16:24:36 +0000 (09:24 -0700)
commit180e4fbfae8a93e647a82f1e049c22571c3c6ff7
tree73ab9a29f5db87bfd31d6f723a76bfe092584439tree | snapshot
parent1f35d20a894bb07e27691332af4beb097142762fcommit | diff
[TensorExpr] LLVMCodegen: fix lowering for UInt->Float casts. (#64862)

Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/64862

Previously we erroneously were looking at dst signedness. This was
discovered when we tried to implement quantize/dequantize ops.

Test Plan: Imported from OSS

Reviewed By: navahgar

Differential Revision: D30881696

Pulled By: ZolotukhinM

fbshipit-source-id: 34af842e5e52a3b6b5d2e70c4ef32f910a20341f
test/cpp/tensorexpr/test_llvm.cpp diff | blob | history
torch/csrc/jit/tensorexpr/llvm_codegen.cpp diff | blob | history
Domain: Machine Learning / ML Framework;