r-i-s-khan/shade

SHADE: Enable Fundamental Cacheability for Distributed Deep Learning Training

shade's Issues

Can not find the module of calculating importance

after reading code，i can not find the module of calculating importance.
ps: if i missed it, please tell me where it is.thank u very much
(I'm a beginner, so please allow me to express my sincerest apologies for any offense)

The question about CUDA version

Your work are amazing, but when I ran setup.py I met the following errors:

ninja: build stopped: subcommand failed. Traceback (most recent call last): File "setup.py", line 717, in <module> build_deps() File "setup.py", line 308, in build_deps build_caffe2(version=version, File "/root/pytorch/tools/build_pytorch_libs.py", line 62, in build_caffe2 cmake.build(my_env) File "/root/pytorch/tools/setup_helpers/cmake.py", line 345, in build self.run(build_args, my_env) File "/root/pytorch/tools/setup_helpers/cmake.py", line 141, in run check_call(command, cwd=self.build_dir, env=env) File "/root/anaconda3/envs/shade_conda/lib/python3.8/subprocess.py", line 364, in check_call raise CalledProcessError(retcode, cmd) subprocess.CalledProcessError: Command '['cmake', '--build', '.', '--target', 'install', '--config', 'Release', '--', '-j', '104']' returned non-zero exit status 1.

I think it's due to the CUDA version not matching with PyTorch 1.7.0 because when I upgraded PyTorch to the latest version, the errors disappeared.

Could you please share with me the CUDA version you used in this experiment? Thank you.

My build environment:
cpu_info: Intel(R) Xeon(R) Gold 5320 CPU @ 2.20GHz
gpu_info: NVIDIA Tesla V100 32GB PCIe
memory: Samsung DDR4 512G
swapfile:4g used
cuda: "11.6"
python:3.8.10
pytorch:1.7.0
nvidia-driver:510.108.03
Ubuntu:20.04