GNN环境安装/conda channel/动手深度学习

参考：

• torch_geometric踩坑实战–安装与运行 亲测有效！！ https://blog.csdn.net/m0_55245520/article/details/130424828
• pytorch 查看gpu cuda版本 https://blog.csdn.net/jacke121/article/details/93592487

x.1 安装

x.1.1 镜像信息补充

参考 https://mp.weixin.qq.com/s?__biz=MzI5MTcwNjA4NQ%3D%3D&idx=1&mid=2247491983&scene=21&sn=2052e7a038f2db52eb282b88495a7dfd#wechat_redirect

x.1.2 conda环境安装

使用anaconda的environment.yml安装环境，将conda镜像改为上一步中的conda镜像，即将channels中的文件改变。这一步时间会比较长，主要使用conda install进行安装，要耐心等待。

conda env create -f environment.yml
conda activate env_cp39_GCN

yml文件如下，

name: env_cp39_GCN
channels:- https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/conda-forge- https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main- https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main/- https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/pytorch/- https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/- pytorch- dglteam- conda-forge- defaults
dependencies:- _libgcc_mutex=0.1=conda_forge- _openmp_mutex=4.5=1_gnu- argon2-cffi=21.3.0=pyhd8ed1ab_0- argon2-cffi-bindings=21.2.0=py39hb9d737c_2- asttokens=2.0.5=pyhd8ed1ab_0- attrs=21.4.0=pyhd8ed1ab_0- backcall=0.2.0=pyh9f0ad1d_0- backports=1.0=py_2- backports.functools_lru_cache=1.6.4=pyhd8ed1ab_0- beautifulsoup4=4.11.1=pyha770c72_0- blas=1.0=mkl- bleach=5.0.0=pyhd8ed1ab_0- boost=1.74.0=py39h5472131_5- boost-cpp=1.74.0=hc6e9bd1_3- bottleneck=1.3.4=py39hce1f21e_0- brotli=1.0.9=h166bdaf_7- brotli-bin=1.0.9=h166bdaf_7- brotlipy=0.7.0=py39h27cfd23_1003- bzip2=1.0.8=h7b6447c_0- ca-certificates=2022.6.15=ha878542_0- cairo=1.16.0=h6cf1ce9_1008- certifi=2022.6.15=py39hf3d152e_0- cffi=1.15.0=py39hd667e15_1- charset-normalizer=2.0.4=pyhd3eb1b0_0- click=8.0.4=py39h06a4308_0- colorama=0.4.4=pyhd3eb1b0_0- cryptography=36.0.0=py39h9ce1e76_0- cudatoolkit=10.2.89=hfd86e86_1- cycler=0.11.0=pyhd8ed1ab_0- debugpy=1.6.0=py39h5a03fae_0- decorator=5.1.1=pyhd8ed1ab_0- defusedxml=0.7.1=pyhd8ed1ab_0- dgl=0.8.1=py39_0- dgl-cuda10.2=0.8.1=py39_0- entrypoints=0.4=pyhd8ed1ab_0- executing=0.8.3=pyhd8ed1ab_0- ffmpeg=4.3=hf484d3e_0- flit-core=3.7.1=pyhd8ed1ab_0- fontconfig=2.13.1=h6c09931_0- fonttools=4.32.0=py39hb9d737c_0- freetype=2.11.0=h70c0345_0- fuzzywuzzy=0.18.0=pyhd8ed1ab_0- gettext=0.19.8.1=h73d1719_1008- giflib=5.2.1=h7b6447c_0- gmp=6.2.1=h2531618_2- gnutls=3.6.15=he1e5248_0- greenlet=1.1.2=py39h5a03fae_2- icu=68.2=h9c3ff4c_0- idna=3.3=pyhd3eb1b0_0- imbalanced-learn=0.9.1=pyhd8ed1ab_1- importlib-metadata=4.11.3=py39hf3d152e_1- importlib_resources=5.7.0=pyhd8ed1ab_0- intel-openmp=2021.4.0=h06a4308_3561- ipykernel=6.13.0=py39hef51801_0- ipython=8.2.0=py39hf3d152e_0- ipython_genutils=0.2.0=py_1- jedi=0.18.1=py39hf3d152e_1- jinja2=3.1.1=pyhd8ed1ab_0- joblib=1.1.0=pyhd8ed1ab_0- jpeg=9d=h7f8727e_0- jsonschema=4.4.0=pyhd8ed1ab_0- jupyter_client=7.2.2=pyhd8ed1ab_1- jupyter_core=4.9.2=py39hf3d152e_0- jupyterlab_pygments=0.2.2=pyhd8ed1ab_0- kiwisolver=1.4.2=py39hf939315_1- lame=3.100=h7b6447c_0- lcms2=2.12=h3be6417_0- ld_impl_linux-64=2.36.1=hea4e1c9_2- libbrotlicommon=1.0.9=h166bdaf_7- libbrotlidec=1.0.9=h166bdaf_7- libbrotlienc=1.0.9=h166bdaf_7- libffi=3.4.2=h7f98852_5- libgcc-ng=11.2.0=h1d223b6_15- libgfortran-ng=7.3.0=hdf63c60_0- libglib=2.68.4=h174f98d_1- libgomp=11.2.0=h1d223b6_15- libiconv=1.16=h516909a_0- libidn2=2.3.2=h7f8727e_0- libpng=1.6.37=hbc83047_0- libprotobuf=3.18.0=h780b84a_1- libsodium=1.0.18=h36c2ea0_1- libstdcxx-ng=11.2.0=he4da1e4_15- libtasn1=4.16.0=h27cfd23_0- libtiff=4.2.0=h85742a9_0- libunistring=0.9.10=h27cfd23_0- libuuid=1.0.3=h7f8727e_2- libuv=1.40.0=h7b6447c_0- libwebp=1.2.2=h55f646e_0- libwebp-base=1.2.2=h7f8727e_0- libxcb=1.13=h7f98852_1004- libxml2=2.9.12=h72842e0_0- littleutils=0.2.2=py_0- lz4-c=1.9.3=h295c915_1- markupsafe=2.1.1=py39hb9d737c_1- matplotlib-base=3.5.1=py39h2fa2bec_0- matplotlib-inline=0.1.3=pyhd8ed1ab_0- mistune=0.8.4=py39h3811e60_1005- mkl=2021.4.0=h06a4308_640- mkl-service=2.4.0=py39h7f8727e_0- mkl_fft=1.3.1=py39hd3c417c_0- mkl_random=1.2.2=py39h51133e4_0- munkres=1.1.4=pyh9f0ad1d_0- nbclient=0.6.0=pyhd8ed1ab_0- nbconvert=6.5.0=pyhd8ed1ab_0- nbconvert-core=6.5.0=pyhd8ed1ab_0- nbconvert-pandoc=6.5.0=pyhd8ed1ab_0- nbformat=5.3.0=pyhd8ed1ab_0- ncurses=6.3=h7f8727e_2- nest-asyncio=1.5.5=pyhd8ed1ab_0- nettle=3.7.3=hbbd107a_1- networkx=2.7.1=pyhd3eb1b0_0- notebook=6.4.10=pyha770c72_0- numexpr=2.8.1=py39h6abb31d_0- numpy=1.21.5=py39he7a7128_1- numpy-base=1.21.5=py39hf524024_1- ogb=1.3.3=pyhd8ed1ab_0- openh264=2.1.1=h4ff587b_0- openssl=3.0.3=h166bdaf_0- outdated=0.2.1=pyhd8ed1ab_0- packaging=21.3=pyhd3eb1b0_0- pandas=1.4.1=py39h295c915_1- pandoc=2.18=ha770c72_0- pandocfilters=1.5.0=pyhd8ed1ab_0- parso=0.8.3=pyhd8ed1ab_0- pcre=8.45=h9c3ff4c_0- pexpect=4.8.0=pyh9f0ad1d_2- pickleshare=0.7.5=py_1003- pillow=9.0.1=py39h22f2fdc_0- pip=21.2.4=py39h06a4308_0- pixman=0.40.0=h36c2ea0_0- prometheus_client=0.14.1=pyhd8ed1ab_0- prompt-toolkit=3.0.29=pyha770c72_0- protobuf=3.18.0=py39he80948d_0- psutil=5.9.0=py39hb9d737c_1- pthread-stubs=0.4=h36c2ea0_1001- ptyprocess=0.7.0=pyhd3deb0d_0- pure_eval=0.2.2=pyhd8ed1ab_0- pycairo=1.21.0=py39h0934665_1- pycparser=2.21=pyhd3eb1b0_0- pygments=2.11.2=pyhd8ed1ab_0- pyopenssl=22.0.0=pyhd3eb1b0_0- pyparsing=3.0.4=pyhd3eb1b0_0- pyrsistent=0.18.1=py39hb9d737c_1- pysocks=1.7.1=py39h06a4308_0- python=3.9.7=hf930737_3_cpython- python-dateutil=2.8.2=pyhd3eb1b0_0- python-fastjsonschema=2.15.3=pyhd8ed1ab_0- python-levenshtein=0.12.2=py39hb9d737c_2- python_abi=3.9=2_cp39- pytorch=1.11.0=py3.9_cuda10.2_cudnn7.6.5_0- pytorch-mutex=1.0=cuda- pytz=2021.3=pyhd3eb1b0_0- pyzmq=22.3.0=py39headdf64_2- readline=8.1.2=h7f8727e_1- reportlab=3.5.68=py39he59360d_1- requests=2.27.1=pyhd3eb1b0_0- scipy=1.6.2=py39had2a1c9_1- send2trash=1.8.0=pyhd8ed1ab_0- setuptools=61.2.0=py39h06a4308_0- six=1.16.0=pyhd3eb1b0_1- soupsieve=2.3.1=pyhd8ed1ab_0- sqlalchemy=1.4.35=py39hb9d737c_0- sqlite=3.38.2=hc218d9a_0- stack_data=0.2.0=pyhd8ed1ab_0- tensorboardx=2.5=pyhd8ed1ab_0- terminado=0.13.3=py39hf3d152e_1- threadpoolctl=3.1.0=pyh8a188c0_0- tinycss2=1.1.1=pyhd8ed1ab_0- tk=8.6.11=h1ccaba5_0- torchaudio=0.11.0=py39_cu102- torchvision=0.12.0=py39_cu102- tornado=6.1=py39hb9d737c_3- tqdm=4.63.0=pyhd3eb1b0_0- traitlets=5.1.1=pyhd8ed1ab_0- typing_extensions=4.1.1=pyh06a4308_0- tzdata=2022a=hda174b7_0- unicodedata2=14.0.0=py39hb9d737c_1- urllib3=1.26.8=pyhd3eb1b0_0- wcwidth=0.2.5=pyh9f0ad1d_2- webencodings=0.5.1=py_1- wheel=0.37.1=pyhd3eb1b0_0- xorg-kbproto=1.0.7=h7f98852_1002- xorg-libice=1.0.10=h7f98852_0- xorg-libsm=1.2.2=h470a237_5- xorg-libx11=1.7.2=h7f98852_0- xorg-libxau=1.0.9=h7f98852_0- xorg-libxdmcp=1.1.3=h7f98852_0- xorg-libxext=1.3.4=h7f98852_1- xorg-libxrender=0.9.10=h7f98852_1003- xorg-renderproto=0.11.1=h7f98852_1002- xorg-xextproto=7.3.0=h7f98852_1002- xorg-xproto=7.0.31=h7f98852_1007- xz=5.2.5=h7b6447c_0- zeromq=4.3.4=h9c3ff4c_1- zipp=3.8.0=pyhd8ed1ab_0- zlib=1.2.11=h7f8727e_4- zstd=1.4.9=haebb681_0

x.1.3 pip 安装

由于pip的“特殊性和极为特殊的依赖性”，往往并不能一次性安装好，使用pip循环遍历多次安装：即依次安装不好，将安装失败的注释掉，接着安装剩下的。一遍循环完毕后，将安装好的注释掉，将上一次安装失败的解注释，继续安装。循环往复，直到安装好。

pip install -r requirements.txt
pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple/

requirements.txt如下，

absl-py==1.0.0
antlr4-python3-runtime==4.9.3
astor==0.7.1
autograd==1.4
autograd-gamma==0.5.0
ax-platform==0.2.5.1
botorch==0.6.4
cachetools==5.0.0
captum==0.0.1
cilog==1.2.3
cloudpickle==2.0.0
dgllife==0.2.9
dive-into-graphs==0.2.0torch-scatter==2.0.9
torch-sparse==0.6.13
torch-cluster==1.6.0
torch-spline_conv==1.2.1
torch-geometric==2.0.4

x.2 碰到问题

Traceback (most recent call last):File "/opt/conda/envs/HIGH-PPI/lib/python3.9/runpy.py", line 197, in _run_module_as_mainreturn _run_code(code, main_globals, None,File "/opt/conda/envs/HIGH-PPI/lib/python3.9/runpy.py", line 87, in _run_codeexec(code, run_globals)File "/root/.vscode-server/extensions/ms-python.python-2022.6.0/pythonFiles/lib/python/debugpy/__main__.py", line 45, in <module>cli.main()File "/root/.vscode-server/extensions/ms-python.python-2022.6.0/pythonFiles/lib/python/debugpy/../debugpy/server/cli.py", line 444, in mainrun()File "/root/.vscode-server/extensions/ms-python.python-2022.6.0/pythonFiles/lib/python/debugpy/../debugpy/server/cli.py", line 285, in run_filerunpy.run_path(target_as_str, run_name=compat.force_str("__main__"))File "/opt/conda/envs/HIGH-PPI/lib/python3.9/runpy.py", line 268, in run_pathreturn _run_module_code(code, init_globals, run_name,File "/opt/conda/envs/HIGH-PPI/lib/python3.9/runpy.py", line 97, in _run_module_code_run_code(code, mod_globals, init_globals,File "/opt/conda/envs/HIGH-PPI/lib/python3.9/runpy.py", line 87, in _run_codeexec(code, run_globals)File "/home/yingmuzhi/high_ppi/HIGH-PPI/model_train.py", line 317, in <module>main()File "/home/yingmuzhi/high_ppi/HIGH-PPI/model_train.py", line 308, in maintrain(batch, p_x_all, p_edge_all, model, graph, ppi_list, loss_fn, optimizer, device,File "/home/yingmuzhi/high_ppi/HIGH-PPI/model_train.py", line 135, in trainoutput = model(batch, p_x_all, p_edge_all, graph.edge_index_got, train_edge_id)File "/opt/conda/envs/HIGH-PPI/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1110, in _call_implreturn forward_call(*input, **kwargs)File "/home/yingmuzhi/high_ppi/HIGH-PPI/gnn_models_sag.py", line 170, in forwardembs = self.BGNN(x, edge, batch-1)File "/opt/conda/envs/HIGH-PPI/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1110, in _call_implreturn forward_call(*input, **kwargs)File "/home/yingmuzhi/high_ppi/HIGH-PPI/gnn_models_sag.py", line 117, in forwardx = self.conv1(x, edge_index)File "/opt/conda/envs/HIGH-PPI/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1110, in _call_implreturn forward_call(*input, **kwargs)File "/opt/conda/envs/HIGH-PPI/lib/python3.9/site-packages/torch_geometric/nn/conv/gcn_conv.py", line 210, in forwardedge_index, edge_weight = gcn_norm(  # yapf: disableFile "/opt/conda/envs/HIGH-PPI/lib/python3.9/site-packages/torch_geometric/nn/conv/gcn_conv.py", line 66, in gcn_normif is_torch_sparse_tensor(edge_index):File "/opt/conda/envs/HIGH-PPI/lib/python3.9/site-packages/torch_geometric/utils/sparse.py", line 68, in is_torch_sparse_tensorif src.layout == torch.sparse_csc:
AttributeError: module 'torch' has no attribute 'sparse_csc'

cuda error

train gnn, train_num: 5328, valid_num: 1332
cuda:0
/opt/conda/envs/HIGH-PPI/lib/python3.9/site-packages/torch/cuda/__init__.py:145: UserWarning: 
NVIDIA A40 with CUDA capability sm_86 is not compatible with the current PyTorch installation.
The current PyTorch install supports CUDA capabilities sm_37 sm_50 sm_60 sm_61 sm_70 sm_75 compute_37.
If you want to use the NVIDIA A40 GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/warnings.warn(incompatible_device_warn.format(device_name, capability, " ".join(arch_list), device_name))

使用查看cuda版本

import torchprint(torch.version.cuda)

动手深度学习纸质书： https://j.youzan.com/2E82KT（半价购书）

torch_geometric error

安装torch和对应cudatoolkit。

x.2.1 anaconda（可选）

使用python3.9

conda create -n env_cp39_GNN python=3.9

x.2.2 torch

参考 https://pytorch.org/get-started/previous-versions/

使用torch1.11和cuda10.2。需要注意的是1.11后的使用+指定tookkit，如torch==1.12.0+cu102; 2.0以后的使用-指定toolkit，如pytorch==2.0.0 pytorch-cuda=11.7。

# CUDA 10.2
conda install pytorch==1.11.0 torchvision==0.12.0 torchaudio==0.11.0 cudatoolkit=10.2 -c pytorch

本文来自互联网用户投稿，文章观点仅代表作者本人，不代表本站立场，不承担相关法律责任。如若转载，请注明出处。 如若内容造成侵权/违法违规/事实不符，请点击【内容举报】进行投诉反馈！