TVM是什么?
Apache(incubating) TVM is an open deep learning compiler stack for CPUs, GPUs, and specialized accelerators. It aims to close the gap between the productivity-focused deep learning frameworks, and the performance- or efficiency-oriented hardware backends. TVM provides the following main features:
Compilation of deep learning models in Keras, MXNet, PyTorch, Tensorflow, CoreML, DarkNet into minimum deployable modules on diverse hardware backends.
Infrastructure to automatic generate and optimize tensor operators on more backend with better performance.
摘自:
https://tvm.apache.org/about
环境安装
笔者使用CUDA+LLVM的编译方式。
相应的依赖包括:
1.CUDA
2.GCC
3.stdlibc++
4.LLVM
5.TVM
当前的组合是GCC(5.4.0)+CUDA(8.0)+CUDNN(7.1.3)+stdlibc++(6.0.24)+LLVM(8.0.0)+TVM(0.7.1)
建议GCC不要使用覆盖安装。
TVM开启CUDA,CUDNN,LLVM使能。
TVM要求GCC版本高于5。
CUDA 10.0可以支持高版本GCC,8.0还是建议还是GCC 5.
目前还在跑DEMO,先占个坑,后续补充TVM的代码走读。
本文内容由网友自发贡献,版权归原作者所有,本站不承担相应法律责任。如您发现有涉嫌抄袭侵权的内容,请联系:hwhale#tublm.com(使用前将#替换为@)