diff --git a/_all_pages/develop/cxx_api_doc.md b/_all_pages/develop/cxx_api_doc.md index dd62eaf9f7a71e23f722b4d294d05dbcf82176d9..3f697dfd9759838db66cf56e40a69d26886a76c9 100644 --- a/_all_pages/develop/cxx_api_doc.md +++ b/_all_pages/develop/cxx_api_doc.md @@ -346,6 +346,8 @@ for (int i = 0; i < ShapeProduction(output_tensor->shape()); i += 100) { 根据名称获取输出Tensor的指针。 +**注意**:`GetTensor`接口是为开发者设计的调试接口,可以输出[转化](../model_optimize_tool)后模型中的任一节点。如果出现`GetTensor(InputName)`返回值为空`Tensor`,可能原因是以该`InputName`命名的Tensor在模型转化的**子图融合**过程被融合替换了。 + 参数: - `name(const std::string)` - Tensor的名称 diff --git a/_all_pages/develop/model_optimize_tool.md b/_all_pages/develop/model_optimize_tool.md index 55ae25405decfdc533c74a75863db6a851cf5c9e..8112f4d927a8d49f4b31fc8e20f2e8114e16371a 100644 --- a/_all_pages/develop/model_optimize_tool.md +++ b/_all_pages/develop/model_optimize_tool.md @@ -24,7 +24,7 @@ chmod +x model_optimize_tool ### 编译model_optimize_tool -1、参照 [编译安装](./source_compile) 进行环境配置和编译 +1、参照 [编译安装](../source_compile) 进行环境配置和编译 2、进入docker中PaddleLite根目录,```git checkout [release-version-tag]```切换到release分支 diff --git a/_all_pages/develop/source_compile.md b/_all_pages/develop/source_compile.md index 40efbc1ad766bc51129ad1f162a6353215c0f628..7e2028797c8329d6f8d5de1f9e6e36d71a151f90 100644 --- a/_all_pages/develop/source_compile.md +++ b/_all_pages/develop/source_compile.md @@ -280,7 +280,7 @@ git checkout ``` ios tiny publish支持的编译选项: * `--arm_os`: 可选ios或者ios64 -* `--arm_abi`: 可选armv7和armv8(**注意**:当`arm_os=ios`时只能选择`arm_abi=armv7`) +* `--arm_abi`: 可选armv7和armv8(**注意**:当`arm_os=ios`时只能选择`arm_abi=armv7`,当`arm_os=ios64`时只能选择`arm_abi=armv8`) * 如果mac编译过程中报错:"Invalid CMAKE_DEVELOPER_ROOT: does not exist", 运行: ```shell sudo xcode-select -s /Applications/Xcode.app/Contents/Developer diff --git a/_all_pages/v2.0.0/cxx_api_doc.md b/_all_pages/v2.0.0/cxx_api_doc.md index dd62eaf9f7a71e23f722b4d294d05dbcf82176d9..3f697dfd9759838db66cf56e40a69d26886a76c9 100644 --- a/_all_pages/v2.0.0/cxx_api_doc.md +++ b/_all_pages/v2.0.0/cxx_api_doc.md @@ -346,6 +346,8 @@ for (int i = 0; i < ShapeProduction(output_tensor->shape()); i += 100) { 根据名称获取输出Tensor的指针。 +**注意**:`GetTensor`接口是为开发者设计的调试接口,可以输出[转化](../model_optimize_tool)后模型中的任一节点。如果出现`GetTensor(InputName)`返回值为空`Tensor`,可能原因是以该`InputName`命名的Tensor在模型转化的**子图融合**过程被融合替换了。 + 参数: - `name(const std::string)` - Tensor的名称 diff --git a/_all_pages/v2.0.0/model_optimize_tool.md b/_all_pages/v2.0.0/model_optimize_tool.md index 55ae25405decfdc533c74a75863db6a851cf5c9e..8112f4d927a8d49f4b31fc8e20f2e8114e16371a 100644 --- a/_all_pages/v2.0.0/model_optimize_tool.md +++ b/_all_pages/v2.0.0/model_optimize_tool.md @@ -24,7 +24,7 @@ chmod +x model_optimize_tool ### 编译model_optimize_tool -1、参照 [编译安装](./source_compile) 进行环境配置和编译 +1、参照 [编译安装](../source_compile) 进行环境配置和编译 2、进入docker中PaddleLite根目录,```git checkout [release-version-tag]```切换到release分支 diff --git a/_all_pages/v2.0.0/source_compile.md b/_all_pages/v2.0.0/source_compile.md index 40efbc1ad766bc51129ad1f162a6353215c0f628..7e2028797c8329d6f8d5de1f9e6e36d71a151f90 100644 --- a/_all_pages/v2.0.0/source_compile.md +++ b/_all_pages/v2.0.0/source_compile.md @@ -280,7 +280,7 @@ git checkout ``` ios tiny publish支持的编译选项: * `--arm_os`: 可选ios或者ios64 -* `--arm_abi`: 可选armv7和armv8(**注意**:当`arm_os=ios`时只能选择`arm_abi=armv7`) +* `--arm_abi`: 可选armv7和armv8(**注意**:当`arm_os=ios`时只能选择`arm_abi=armv7`,当`arm_os=ios64`时只能选择`arm_abi=armv8`) * 如果mac编译过程中报错:"Invalid CMAKE_DEVELOPER_ROOT: does not exist", 运行: ```shell sudo xcode-select -s /Applications/Xcode.app/Contents/Developer