diff --git a/README.md b/README.md index be17370e..adde1088 100644 --- a/README.md +++ b/README.md @@ -28,7 +28,7 @@ MegCC 现在支持的平台处理器平台有 Arm64/ArmV7/X86/risc-v/单片机, ## 文档 #### 获取 MegCC 编译器 -* 下载提预编译好的 MegCC 模型编译器 [下载 MegCC 编译器](https://oss.iap.hh-b.brainpp.cn/megengine-built/megcc) +* 下载提预编译好的 MegCC 模型编译器 [下载 MegCC 编译器](https://github.com/MegEngine/MegCC/releases/download/v0.1.2/megcc_release_0.1.2.tar.gz) 下载解压 * 如果是开发者需要从源码编译 MegCC 编译器,参考文档 [源码编译 MegCC](compiler/README.md) diff --git a/doc/how-to-release.md b/doc/how-to-release.md index 1621cff5..42df523b 100644 --- a/doc/how-to-release.md +++ b/doc/how-to-release.md @@ -16,7 +16,7 @@ When release all the third party will build into MegCC, for convenience,MegCC ## Release 1. Modify major/minor/patch in compiler/include/compiler/Common/Version.h.in -2. Write changelog https://discourse.brainpp.cn/t/topic/56470 +2. Write changelog 3. Git tag branch 4. Run the script [release_megcc.sh](../script/docker_release_megcc.sh) in [docker image](../script/docker/Dockerfile) with a directory in which all release file will be generated. Finally a tar file named `megcc_relase_*.tar.gz` will created by compressing all file in the release directory. @@ -32,4 +32,4 @@ docker build script/docker -t megcc_manlylinux2014 # copy release target to megcc oss cp megcc_release_xxx.tar.gz s3://megengine-built/megcc/megcc_release_xxx.tar.gz ``` -6. Send email to notify the users of megcc \ No newline at end of file +6. Send email to notify the users of megcc diff --git a/doc/how-to-use-chinese.md b/doc/how-to-use-chinese.md index cb405c6c..a558b33b 100644 --- a/doc/how-to-use-chinese.md +++ b/doc/how-to-use-chinese.md @@ -18,7 +18,7 @@ bin ``` # 使用 MegCC 完成模型部署 -首先需要从 [brain++ oss web](https://oss.iap.hh-b.brainpp.cn/megengine-built/megcc) 上下载需要的 MegCC 发版包,然后解压这个压缩包:`tar -xvf megcc_release_*.tar.gz`。使用 MegCC 完成模型部署主要步骤有三个步: +首先需要从 [github](https://github.com/MegEngine/MegCC/releases/download/v0.1.2/megcc_release_0.1.2.tar.gz) 上下载需要的 MegCC 发版包,然后解压这个压缩包:`tar -xvf megcc_release_*.tar.gz`。使用 MegCC 完成模型部署主要步骤有三个步: - 模型编译:编译 MegEngine 模型,生成运行这个模型对应的 Kernel 以及和这些 Kernel 绑定的模型。 - Runtime编译:这个阶段会将 Runtime 和上一步中生成的 Kernel 一起编译成一个静态库。 - 集成到应用中:调用上一步编译的静态库的接口进行推理。 diff --git a/doc/how-to-use.md b/doc/how-to-use.md index ef7d19e1..42ec60bb 100644 --- a/doc/how-to-use.md +++ b/doc/how-to-use.md @@ -17,7 +17,7 @@ bin ``` ## Use Example In Android -- Download the precompiled megcc from [barin++ oss web](https://oss.iap.hh-b.brainpp.cn/megengine-built/megcc) +- Download the precompiled megcc from [github](https://github.com/MegEngine/MegCC/releases/download/v0.1.2/megcc_release_0.1.2.tar.gz) - Extract the download tar file. `tar -xvf megcc_release_*.tar.gz` ### Fast Try > warning: **You should make sure a ndk is install in your computer,and set env NDK_ROOT to its direction**