add fp8 support

56863b9a · Jing Zhang · 54df59bf · d4c84256 · 56863b9a · 56863b9a
Commit 56863b9a authored Aug 16, 2023 by Jing Zhang
20 changed files
--- a/library/src/tensor_operation_instance/gpu/normalization/device_layernorm2d_f32_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/normalization/device_layernorm2d_f32_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/normalization/device_layernorm4d_f16_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/normalization/device_layernorm4d_f16_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/normalization/device_layernorm4d_f32_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/normalization/device_layernorm4d_f32_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/normalization/normalization_instance_common.hpp
+++ b/library/src/tensor_operation_instance/gpu/normalization/normalization_instance_common.hpp
--- a/library/src/tensor_operation_instance/gpu/pool3d_fwd/CMakeLists.txt
+++ b/library/src/tensor_operation_instance/gpu/pool3d_fwd/CMakeLists.txt
--- a/library/src/tensor_operation_instance/gpu/pool_fwd/device_avg_pool3d_fwd_ndhwc_f16_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/pool_fwd/device_avg_pool3d_fwd_ndhwc_f16_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/pool_fwd/device_avg_pool3d_fwd_ndhwc_f32_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/pool_fwd/device_avg_pool3d_fwd_ndhwc_f32_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/pool_fwd/device_max_pool3d_fwd_ndhwc_f16_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/pool_fwd/device_max_pool3d_fwd_ndhwc_f16_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/pool_fwd/device_max_pool3d_fwd_ndhwc_f32_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/pool_fwd/device_max_pool3d_fwd_ndhwc_f32_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/pool_fwd/pool_fwd_instance_common.hpp
+++ b/library/src/tensor_operation_instance/gpu/pool_fwd/pool_fwd_instance_common.hpp
--- a/library/src/tensor_operation_instance/gpu/pool_fwd/CMakeLists.txt
+++ b/library/src/tensor_operation_instance/gpu/pool_fwd/CMakeLists.txt
--- a/library/src/tensor_operation_instance/gpu/pool_fwd/device_avg_pool2d_fwd_nhwc_f32_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/pool_fwd/device_avg_pool2d_fwd_nhwc_f32_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/pool_fwd/device_max_pool2d_fwd_nhwc_f16_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/pool_fwd/device_max_pool2d_fwd_nhwc_f16_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/pool_fwd/device_max_pool2d_fwd_nhwc_f32_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/pool_fwd/device_max_pool2d_fwd_nhwc_f32_instance.cpp
--- a/library/src/tensor_operation_instance/gpu/quantization/CMakeLists.txt
+++ b/library/src/tensor_operation_instance/gpu/quantization/CMakeLists.txt
--- a/library/src/tensor_operation_instance/gpu/softmax/CMakeLists.txt
+++ b/library/src/tensor_operation_instance/gpu/softmax/CMakeLists.txt
--- a/profiler/include/profiler/profile_grouped_conv_bwd_weight_impl.hpp
+++ b/profiler/include/profiler/profile_grouped_conv_bwd_weight_impl.hpp
--- a/profiler/include/profiler/profile_groupnorm_impl.hpp
+++ b/profiler/include/profiler/profile_groupnorm_impl.hpp
--- a/profiler/include/profiler/profile_layernorm_impl.hpp
+++ b/profiler/include/profiler/profile_layernorm_impl.hpp
--- a/profiler/include/profiler/profile_pool2d_fwd_impl.hpp
+++ b/profiler/include/profiler/profile_pool2d_fwd_impl.hpp