Merge remote-tracking branch 'origin/develop' into ggemm_multid_two_stage

3e4d0ff3 · Jakub Piasecki · 1ad29336 · 9e011bcd · 3e4d0ff3 · 3e4d0ff3
Commit 3e4d0ff3 authored Mar 19, 2024 by Jakub Piasecki
20 changed files
--- a/example/01_gemm/CMakeLists.txt
+++ b/example/01_gemm/CMakeLists.txt
--- a/example/01_gemm/common.hpp
+++ b/example/01_gemm/common.hpp
--- a/example/01_gemm/gemm_wmma_fp16.cpp
+++ b/example/01_gemm/gemm_wmma_fp16.cpp
--- a/example/01_gemm/gemm_xdl_fp16_fp8.cpp
+++ b/example/01_gemm/gemm_xdl_fp16_fp8.cpp
--- a/example/01_gemm/gemm_xdl_fp8.cpp
+++ b/example/01_gemm/gemm_xdl_fp8.cpp
--- a/example/01_gemm/gemm_xdl_fp8_bf8.cpp
+++ b/example/01_gemm/gemm_xdl_fp8_bf8.cpp
--- a/example/01_gemm/run_gemm_example.inc
+++ b/example/01_gemm/run_gemm_example.inc
--- a/example/02_gemm_bilinear/gemm_bilinear_wmma_fp16.cpp
+++ b/example/02_gemm_bilinear/gemm_bilinear_wmma_fp16.cpp
--- a/example/02_gemm_bilinear/gemm_bilinear_wmma_int8.cpp
+++ b/example/02_gemm_bilinear/gemm_bilinear_wmma_int8.cpp
--- a/example/09_convnd_fwd/CMakeLists.txt
+++ b/example/09_convnd_fwd/CMakeLists.txt
--- a/example/09_convnd_fwd/convnd_fwd_common.hpp
+++ b/example/09_convnd_fwd/convnd_fwd_common.hpp
--- a/example/09_convnd_fwd/convnd_fwd_xdl_fp8.cpp
+++ b/example/09_convnd_fwd/convnd_fwd_xdl_fp8.cpp
--- a/example/29_batched_gemm_bias_e_permute/CMakeLists.txt
+++ b/example/29_batched_gemm_bias_e_permute/CMakeLists.txt
--- a/example/29_batched_gemm_bias_e_permute/batched_gemm_bias_e_permute_wmma_fp16.cpp
+++ b/example/29_batched_gemm_bias_e_permute/batched_gemm_bias_e_permute_wmma_fp16.cpp
--- a/example/30_grouped_conv_fwd_multiple_d/run_grouped_conv_fwd_bias_relu_add_wmma_example.inc
+++ b/example/30_grouped_conv_fwd_multiple_d/run_grouped_conv_fwd_bias_relu_add_wmma_example.inc
--- a/example/32_batched_gemm_scale_softmax_gemm/CMakeLists.txt
+++ b/example/32_batched_gemm_scale_softmax_gemm/CMakeLists.txt
--- a/example/32_batched_gemm_scale_softmax_gemm/batched_gemm_lower_triangle_scale_softmax_gemm_permute_wmma_fp16.cpp
+++ b/example/32_batched_gemm_scale_softmax_gemm/batched_gemm_lower_triangle_scale_softmax_gemm_permute_wmma_fp16.cpp
--- a/example/32_batched_gemm_scale_softmax_gemm/batched_gemm_scale_softmax_gemm_permute_wmma_fp16.cpp
+++ b/example/32_batched_gemm_scale_softmax_gemm/batched_gemm_scale_softmax_gemm_permute_wmma_fp16.cpp
--- a/example/32_batched_gemm_scale_softmax_gemm/cross_attention_forward_wmma_fp16.cpp
+++ b/example/32_batched_gemm_scale_softmax_gemm/cross_attention_forward_wmma_fp16.cpp
--- a/example/32_batched_gemm_scale_softmax_gemm/grouped_query_attention_forward_wmma_fp16.cpp
+++ b/example/32_batched_gemm_scale_softmax_gemm/grouped_query_attention_forward_wmma_fp16.cpp