- 03 Nov, 2022 1 commit
-
-
guangzlu authored
* add fused addition lyernorm * add fused addition lyernorm * changed CMakelist * removed annotates * modified descriptor of C * fixed bug in gridwise add layernorm * format the files * modified name from add&layernorm into elementwise&layernorm * created fused elementwise layernorm branch * change input into tuple type * add sweep once to reduce load & read of C from global memory * modified Argument api * modified way to malloc c in global memory * changed gamma and beta to m_k_desc * fixed bug when sweep once and move CDataType when define device level struct * add src dim for gamma and beta * implement optimization for coalesced * delete a annotation line * fixed some bug to meet the requirements of ck * add bandwidth computing in example, and fixed the time unit * move device_elementwise_layernorm_impl.hpp into device/impl * fixed bug in device_elementwise_layernorm_impl.hpp * changed name from layernorm into normalization * clang-format the changed files * changed the names * moved immidiate results into lds, it become faster in non-sweeponce cases * changed naming of C into X to make the defination more clear * changed naming in example * add tests for elementwise normalization * move example_elementwise_layernorm_blockwise into folder 44_elementwise_normalization * move test_elementwise_layernorm_fp16 into new folder * move elementwise_normalization_instances into a new folder * add more tests in test_elementwise_layernorm_fp16.cpp * added some corner cases in test * fixed method to compute lds size for matrix X * changed name of 44_elementwise_normalization into 45_elementwise_normalization * modified some comments * modified some other confused comments * reduce redundant tests in test_elementwise_layernorm_fp16.cpp
-
- 02 Nov, 2022 1 commit
-
-
rocking5566 authored
* Sync the naming * Sync the test of layernorm with groupnorm * Sync the naming * Minor change for comment and log * [What] Add saveMean and SaveInvVariance in the interface. [Why] These can optimize the backward
-
- 25 Oct, 2022 2 commits
-
-
-
guangzlu authored
* add fused addition lyernorm * add fused addition lyernorm * changed CMakelist * removed annotates * modified descriptor of C * fixed bug in gridwise add layernorm * format the files * modified name from add&layernorm into elementwise&layernorm * created fused elementwise layernorm branch * change input into tuple type * add sweep once to reduce load & read of C from global memory * modified Argument api * modified way to malloc c in global memory * changed gamma and beta to m_k_desc * fixed bug when sweep once and move CDataType when define device level struct * add src dim for gamma and beta * implement optimization for coalesced * delete a annotation line * fixed some bug to meet the requirements of ck * add bandwidth computing in example, and fixed the time unit * move device_elementwise_layernorm_impl.hpp into device/impl * fixed bug in device_elementwise_layernorm_impl.hpp * changed name from layernorm into normalization * clang-format the changed files * changed the names * moved immidiate results into lds, it become faster in non-sweeponce cases * changed naming of C into X to make the defination more clear * changed naming in example * add tests for elementwise normalization * move example_elementwise_layernorm_blockwise into folder 44_elementwise_normalization * move test_elementwise_layernorm_fp16 into new folder * move elementwise_normalization_instances into a new folder * add more tests in test_elementwise_layernorm_fp16.cpp * added some corner cases in test * fixed method to compute lds size for matrix X * changed name of 44_elementwise_normalization into 45_elementwise_normalization * modified some comments * modified some other confused comments * reduce redundant tests in test_elementwise_layernorm_fp16.cpp
-
- 13 Oct, 2022 1 commit
-
-
rocking5566 authored
* Fix bug of profiler for layernorm * 1. Rename layernorm into normalization 2. Decouple softmax from normalization * clang-format
-
- 20 Sep, 2022 1 commit
-
-
rocking5566 authored
* Add groupnorm example by layernorm 1. Reference is not ready 2. shape of gamma and beta need to be fix * Let shape of gamma and beta can be same as x * Modify test, instance and client example * [What] Fix bug of layernorm for greater than 2 dimension. [Why] We need to get upper length from merge transform instead of embed transform. * Add reference for groupnorm * Fuse sigmoid after groupnorm * [What] Rename original layernorm into layernorm2d [Why] Prepare to add groupnorm using layernorm5d * clang-format * Add groupnorm test * Refine error message * Add groupnorm ckProfiler * Test groupnorm kernel from device_instance * update example * upadte profiler * Fix test naming * Fix argc number * Move descriptor and sweeponce to argument for quick debugging Co-authored-by:Chao Liu <chao.liu2@amd.com>
-
- 29 Jul, 2022 1 commit
-
-
Chao Liu authored
* convnd_fwd fp16 example * update example * update example * update instance * updating refernce conv * update reference conv * update conv fwd profiler * update conv 1d and 3d instance * update include path * clean * update profiler for conv bwd data and weight * update conv bwd weight * clean * update conv example * update profiler for conv bwd weight * update ckprofiler for conv bwd data * fix reference conv bwd data bug; update conv bwd data test * update examples * fix initialization issue * update test for conv fwd * clean * clean * remove test case too sensitive to error threshhold * fix test * clean * fix build * adding conv multiple d * adding conv multiple D * add matrix padder * add gemm padding to convnd * adding group conv * update gemm multi-d * refactor * refactor * refactor * clean * clean * refactor * refactor * reorg * add ds * add bias * clean * add G * adding group * adding group * adding group * update Tensor * clean * update example * update DeviceGemmMultipleD_Xdl_CShuffle * update conv bwd-data and bwd-weight * upate contraction example * update gemm and batch gemm with e permute * fix example build * instance for grouped conv1d * update example * adding group conv instance * update gemm bilinear instance * update gemm+add+add+fastgelu instance * update profiler * update profiler * update test * update test and client example * clean * add grouped conv into profiler * update profiler * clean * add test grouped conv, update all conv test to gtest * update test
-
- 13 Jul, 2022 1 commit
-
-
rocking5566 authored
* Implement layernorm kernel and deviceOp * verify gpu kernel with host code * 1. Separate gamma aand beta from affine 2. Check if argument is valid * clean * Sync the naming * Support sweep once mode if we can put k dimension data inside one block * [What] Get length from upper length. [Why] if we get length directly, we may get length after padding. * We only use one block in K dimension. Hence, we can simplify the indexing of global R/W. * Use 1d descriptor for gamma and beta * Add accElementwiseOp * Extract layernorm host code * Support different YVectorDim in GridwiseLayernorm * Rename XSrcVectorDim to XYSrcVectorDim. Because we use same parameter in deviceOp * Gamma and beta can share the VGPR. * Add test for fp32 and fp16 * Fix bug of concurrency and add test case which may fail orignally * Propagate NaN for layernorm Co-authored-by:
-
