[CODEGEN] Support cuda tensorcore subbyte int data type in auto tensorcore (#4546)
authorOrion34C <vcitory34@gmail.com>
Fri, 21 Feb 2020 02:43:45 +0000 (10:43 +0800)
committerGitHub <noreply@github.com>
Fri, 21 Feb 2020 02:43:45 +0000 (21:43 -0500)
commitf23ac96905b3a434d2ee3b8bcc912a24b3e63eba
treed41e30c69be160be1fa498dede858d7a1662b583
parent98e7709ff9e69d48ed46e66d89145aeed074f509
[CODEGEN] Support cuda tensorcore subbyte int data type in auto tensorcore (#4546)

* support cuda tensorcore subbyte int data type in auto tensorcore

* add lisence

* pass cpplint

* fix code review comments

* merge the int4/int1 codegen tutorial into the existing auto tensorcore tutorial

* using master's new API

* disable tuning when cuda is not enabled

* address cr comment

* do not run the tuning

* fix test failure

* fix cpplint error

* fix bool type reduction bug

* 1. fix a index bug 2. fix returned bytes value of int1/int4/uint4

* fix typo
include/tvm/runtime/data_type.h
include/tvm/tir/expr.h
src/runtime/ndarray.cc
src/target/source/codegen_c.cc
src/target/source/codegen_cuda.cc
src/tir/pass/arg_binder.cc
src/tir/pass/infer_fragment.cc
src/tir/pass/tensor_core.cc
tutorials/optimize/opt_matmul_auto_tensorcore.py