ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-12 09:16:52 +00:00
Code Issues Packages Projects Releases Wiki Activity

Reduction for int8 and bfloat16 (#125)

Browse Source 
* Use thread cluster descriptor and explicit M_K 2d descriptor to simply Blockwise Reduction

* Change by replacing ReduceDims by NumReduceDims as Device Reduce interface template parameter

* Rename the folder name for the pool2d and reduce examples

* Update to reduction test scripts

* Add Readme for pool2d_fwd and reduce_blockwise examples

* Add support for int8_t reduction (ADD/AVG, MIN/MAX/AMAX)

* Tiny fix in reduce profiler and tiny update in reduce testing scripts

* Tiny fix in testing script profile_reduce_no_index.sh

* Tiny fix in testing script profile_reduce_no_index.sh

* Add support for bfp16 reduction (using bhalf_t = ushort)

* Tiny fix in amd_buffer_addressing.hpp

* Tiny change in script/profile_reduce_with_index.sh

* Use AccDataType for Beta value and use element_wise::PassThrough

* Use type_convert for type converting in host layer reduction

* Renaming and refining in Reduction profiler/device layer/examples

* Renaming and refining in Reduction profiler/device layer/examples

* Renaming all NumReduceDims to NumReduceDim

* Fix the leaked type_convert in ThreadwiseTensorSliceTransfer_v2

* Update to testing scripts to add bf16 support

* added more static_assert

* Remove buggy tunable configurations defined in device_reduce_instance_xxx.hpp

* Add static_assert to give compile-time warning for incorrect thread slice-size/vector-size configurations

* minor change

* Refine and fix (in GetWorkspaceSizeInBytes of MultiBlockPartialReduce) to make int8 completely pass

* Tiny renaming in gridwise_2d_reduction_multiblock_partial_reduce.hpp

* Tiny fix in script/profile_reduce_no_index.sh

* Refine in DeviceReduce layer with regard to using NumInvariantDim/NumReduceDim or InvariantDims/ReduceDims

* Generic renaming in host reduction and DeviceReduce layer

* Add support for 4-d all dimension reduction in the profiler and add_device_reduce_xxx instances

* Use multi-thread and simplification for host Reduction implementation

* Add ctest for reduction

* Update to clarify the using of data init method in produce_reduce/example_reduce/test_reduce/

* Update to the reduce CTest executables to enable default testing behavior when no command argument

* Renaming

Co-authored-by: Jianfeng yan <jfyan008@gmail.com>
This commit is contained in:
Qianfeng
2022-03-23 03:35:14 +08:00
committed by GitHub
parent cb87b049de
commit 9a8ee8a39a
113 changed files with 4036 additions and 973 deletions
Download Patch File Download Diff File Expand all files Collapse all files

75
profiler/src/profile_reduce.cpp
View File

@@ -34,6 +34,8 @@ static struct option long_options[] = {{"inLengths", required_argument, nullptr,
{"scales", required_argument, nullptr, 'S'},
{"half", no_argument, nullptr, '?'},
{"double", no_argument, nullptr, '?'},
{"int8", no_argument, nullptr, '?'},
{"bf16", no_argument, nullptr, '?'},
{"dumpout", required_argument, nullptr, 'o'},
{"verify", required_argument, nullptr, 'v'},
{"log", required_argument, nullptr, 'l'},
@@ -119,6 +121,8 @@ class AppArgs
public:
bool use_half = false;
bool use_double = false;
bool use_int8 = false;
bool use_bf16 = false;
std::vector<size_t> inLengths;
std::vector<size_t> outLengths;
@@ -169,6 +173,8 @@ class AppArgs
<< std::endl;
std::cout << "--half, use fp16 for the input and output tensor data types" << std::endl;
std::cout << "--double, use fp64 for the input and output tensor data types" << std::endl;
std::cout << "--int8, use int8 for the input and output tensor data types" << std::endl;
std::cout << "--bf16, use bfloat16 for the input and output tensor data types" << std::endl;
std::cout << "--verify or -v, 1/0 to indicate whether to verify the reduction result by "
"comparing with the host-based reduction"
<< std::endl;
@@ -267,6 +273,10 @@ class AppArgs
use_half = true;
else if(std::string(long_options[option_index].name) == "double")
use_double = true;
else if(std::string(long_options[option_index].name) == "int8")
use_int8 = true;
else if(std::string(long_options[option_index].name) == "bf16")
use_bf16 = true;
else if(std::string(long_options[option_index].name) == "help")
{
show_usage(argv[0]);
@@ -385,6 +395,71 @@ int profile_reduce(int argc, char* argv[])
args.scales[0],
args.scales[1]);
}
else if(args.use_int8)
{
if(!args.compType_assigned)
args.compTypeId = appInt8;
if(args.outType_assigned && (args.outTypeId != appInt8 && args.outTypeId != appInt32))
args.outTypeId = appInt32;
if(!args.outType_assigned)
args.outTypeId = appInt8;
if(args.compTypeId == appInt8)
{
profile_reduce_impl<int8_t, int8_t, int8_t>(args.do_verification,
args.init_method,
args.do_log,
args.do_dumpout,
args.nrepeat,
args.inLengths,
args.reduceDims,
args.reduceOp,
args.nanOpt,
args.indicesOpt,
args.scales[0],
args.scales[1]);
}
else if(args.compTypeId == appInt32)
{
profile_reduce_impl<int8_t, int32_t, int8_t>(args.do_verification,
args.init_method,
args.do_log,
args.do_dumpout,
args.nrepeat,
args.inLengths,
args.reduceDims,
args.reduceOp,
args.nanOpt,
args.indicesOpt,
args.scales[0],
args.scales[1]);
}
else
throw std::runtime_error("Invalid compType assignment!");
}
else if(args.use_bf16)
{
if(args.outType_assigned && (args.outTypeId != appBFloat16 && args.outTypeId != appFloat))
args.outTypeId = appFloat;
if(!args.outType_assigned)
args.outTypeId = appBFloat16;
profile_reduce_impl<ck::bhalf_t, float, ck::bhalf_t>(args.do_verification,
args.init_method,
args.do_log,
args.do_dumpout,
args.nrepeat,
args.inLengths,
args.reduceDims,
args.reduceOp,
args.nanOpt,
args.indicesOpt,
args.scales[0],
args.scales[1]);
}
else
{
if(args.compTypeId == appFloat)