From f8f12d3f52012dd6071407b739a3f2ee78d7eb06 Mon Sep 17 00:00:00 2001 From: Dominika Jedynak Date: Wed, 19 Jan 2022 23:40:19 +0100 Subject: [PATCH 1/2] Changing confusing documentation --- config/linux.cmake | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/config/linux.cmake b/config/linux.cmake index ec02d9d5cf5b..3a4215a887c9 100644 --- a/config/linux.cmake +++ b/config/linux.cmake @@ -114,11 +114,10 @@ SET(EXTRA_OPERATORS "" CACHE PATH "EXTRA OPERATORS PATH") # Create C++ interface package set(USE_CPP_PACKAGE OFF CACHE BOOL "Build C++ Package") -# Use int64_t type to represent the total number of elements in a tensor +# Use int64_t type to represent the number of elements in a tensor # This will cause performance degradation reported in issue #14496 # Set to 1 for large tensor with tensor size greater than INT32_MAX i.e. 2147483647 -# Note: the size of each dimension is still bounded by INT32_MAX -set(USE_INT64_TENSOR_SIZE ON CACHE BOOL "Use int64_t to represent the total number of elements in a tensor") +set(USE_INT64_TENSOR_SIZE ON CACHE BOOL "Use int64_t to represent the number of elements in a tensor") # Other GPU features set(USE_NCCL "Use NVidia NCCL with CUDA" OFF) From 3531621700e3077bca4d312a64c1919e26f4ee58 Mon Sep 17 00:00:00 2001 From: DominikaJedynak Date: Thu, 27 Jan 2022 10:40:04 +0100 Subject: [PATCH 2/2] Rewiev suggestion MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Co-authored-by: Andrzej Kotłowski --- config/linux.cmake | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/config/linux.cmake b/config/linux.cmake index 3a4215a887c9..1b2d7f5de8b3 100644 --- a/config/linux.cmake +++ b/config/linux.cmake @@ -114,7 +114,7 @@ SET(EXTRA_OPERATORS "" CACHE PATH "EXTRA OPERATORS PATH") # Create C++ interface package set(USE_CPP_PACKAGE OFF CACHE BOOL "Build C++ Package") -# Use int64_t type to represent the number of elements in a tensor +# Use int64_t type to represent index and the number of elements in a tensor # This will cause performance degradation reported in issue #14496 # Set to 1 for large tensor with tensor size greater than INT32_MAX i.e. 2147483647 set(USE_INT64_TENSOR_SIZE ON CACHE BOOL "Use int64_t to represent the number of elements in a tensor")