DeepEMhancer をテンプレートにして作成
[
トップ
] [
新規
|
一覧
|
検索
|
最終更新
|
ヘルプ
|
ログイン
]
開始行:
本家様 [[https://github.com/rsanchezgarc/deepEMhancer>+ht...
gitには
#code(nonumber){{
Simply speaking, DeepEMhancer performs a non-linear post-...
1. Local sharpening-like post-processing.
2. Automatic masking/denoising of cryo-EM maps.
}}
とある. cryoSPARCからもコールされるようになっている.
インストール環境はこんな感じ
#code(nonumber){{
[root@rockylinux9 ~]# cat /etc/redhat-release
Rocky Linux release 9.8 (Blue Onyx)
[root@rockylinux9 ~]# cat /proc/driver/nvidia/version
NVRM version: NVIDIA UNIX Open Kernel Module for x86_64 ...
GCC version: gcc version 11.5.0 20240719 (Red Hat 11.5.0...
[root@rockylinux9 ~]# nvidia-smi -L
GPU 0: NVIDIA RTX PRO 2000 Blackwell (UUID: GPU-40660e37-...
[root@rockylinux9 ~]#
}}
「/usr/local/cuda」は用意していないです
***インストール [#j5b5561d]
DeepEMhancerはpythonアプリなので、pyenv/miniforgeの環境に...
pyenv/miniforge環境がないなら下記のように敷設して
#code(nonumber){{
git clone https://github.com/yyuu/pyenv.git /apps/pyenv
export PYENV_ROOT=/apps/pyenv
export PATH=$PYENV_ROOT/bin:$PATH
pyenv install miniforge3-26.7.2-0
source /apps/pyenv/versions/miniforge3-26.7.2-0/etc/profi...
conda update -n base -c conda-forge conda
(既に環境があるなら)
source /apps/pyenv/versions/miniforge3-26.7.2-0/etc/profi...
}}
っで、DeepEMhancer を入れていきます.
#code(nonumber){{
[root@rockylinux9 ~]# cd /apps
[root@rockylinux9 apps]# git clone https://github.com/rsa...
[root@rockylinux9 apps]# cd deepEMhancer/
[root@rockylinux9 deepEMhancer]# ls -CF
alternative_installation/ condaDeepEMHancer/ conda_ymls...
[root@rockylinux9 deepEMhancer]# cat deepEMhancer_env.yml
channels:
- conda-forge
- defaults
dependencies:
- pip=26.2
- python=3.11
- pip:
- h5py==3.14.0
- mrcfile==1.5.4
- numpy==2.2.6
- requests==2.34.2
- scikit-image==0.26.0
- scipy==1.17.1
- tensorflow[and-cuda]==2.21.0
- tqdm==4.70.1
[root@rockylinux9 deepEMhancer]#
[root@rockylinux9 deepEMhancer]# conda env create -f deep...
}}
作った deepEMhancer 実行環境に対して
#code(nonumber){{
[root@rockylinux9 deepEMhancer]# conda activate deepEMhan...
(deepEMhancer) [root@rockylinux9 deepEMhancer]#
(deepEMhancer) [root@rockylinux9 deepEMhancer]# which pyt...
/apps/pyenv/versions/miniforge3-26.7.2-0/envs/deepEMhance...
(deepEMhancer) [root@rockylinux9 deepEMhancer]# python -m...
:
Successfully built deepEMhancer
Installing collected packages: deepEMhancer
Successfully installed deepEMhancer-0.18
(deepEMhancer) [root@rockylinux9 deepEMhancer]#
(deepEMhancer) [root@rockylinux9 deepEMhancer]# conda list
# packages in environment at /apps/pyenv/versions/minifor...
#
# Name Version Build ...
_openmp_mutex 4.5 20_gnu ...
absl-py 2.5.0 pypi_0 ...
astunparse 1.6.3 pypi_0 ...
bzip2 1.0.8 hda65f42_10 ...
ca-certificates 2026.7.22 hbd8a1cb_0 ...
certifi 2026.7.22 pypi_0 ...
charset-normalizer 3.5.1 pypi_0 ...
deepemhancer 0.18 pypi_0 ...
flatbuffers 25.12.19 pypi_0 ...
gast 0.7.0 pypi_0 ...
google-pasta 0.2.0 pypi_0 ...
grpcio 1.84.0 pypi_0 ...
h5py 3.14.0 pypi_0 ...
icu 78.3 py310h44b86e...
idna 3.20 pypi_0 ...
imageio 2.37.4 pypi_0 ...
keras 3.15.1 pypi_0 ...
lazy-loader 0.6 pypi_0 ...
ld_impl_linux-64 2.46.1 default_hbd6...
libclang 18.1.1 pypi_0 ...
libexpat 2.8.5 hd2095e1_0 ...
libffi 3.7.0 h81df57d_1 ...
libgcc 16.2.0 ha9f2e26_7 ...
libgomp 16.2.0 he0feb66_7 ...
liblzma 5.8.3 hb03c661_1 ...
libnsl 2.0.1 hb9d3cd8_1 ...
libpython 3.11.16 h0c77377_2_c...
libsqlite 3.53.4 h13e7031_1 ...
libstdcxx 16.2.0 h934c35e_7 ...
libuuid 2.42.4 hcfc3c73_0 ...
libxcrypt 4.4.38 h280c20c_0 ...
libzlib 1.3.2 h25fd6f3_3 ...
markdown-it-py 4.2.0 pypi_0 ...
mdurl 0.1.2 pypi_0 ...
ml-dtypes 0.6.0 pypi_0 ...
mrcfile 1.5.4 pypi_0 ...
namex 0.1.0 pypi_0 ...
ncurses 6.6 hdb14827_1 ...
networkx 3.6.1 pypi_0 ...
numpy 2.2.6 pypi_0 ...
nvidia-cublas-cu12 12.9.2.10 pypi_0 ...
nvidia-cuda-cupti-cu12 12.9.79 pypi_0 ...
nvidia-cuda-nvcc-cu12 12.9.86 pypi_0 ...
nvidia-cuda-nvrtc-cu12 12.9.86 pypi_0 ...
nvidia-cuda-runtime-cu12 12.9.79 pypi_0 ...
nvidia-cudnn-cu12 9.26.0.51 pypi_0 ...
nvidia-cufft-cu12 11.4.1.4 pypi_0 ...
nvidia-curand-cu12 10.3.10.19 pypi_0 ...
nvidia-cusolver-cu12 11.7.5.82 pypi_0 ...
nvidia-cusparse-cu12 12.5.10.65 pypi_0 ...
nvidia-nccl-cu12 2.32.3 pypi_0 ...
nvidia-nvjitlink-cu12 12.9.86 pypi_0 ...
openssl 3.6.4 h781a0a9_0 ...
opt-einsum 3.4.0 pypi_0 ...
optree 0.20.0 pypi_0 ...
packaging 26.3 pyhc364b38_0...
pillow 12.3.0 pypi_0 ...
pip 26.2.1 pyh8b19718_0...
protobuf 7.36.2 pypi_0 ...
pygments 2.21.0 pypi_0 ...
python 3.11.16 h5f976f7_2_c...
readline 8.3 hd6e31c0_1 ...
requests 2.34.2 pypi_0 ...
rich 15.0.0 pypi_0 ...
scikit-image 0.26.0 pypi_0 ...
scipy 1.17.1 pypi_0 ...
setuptools 84.0.0 pyh332efcf_0...
six 1.17.0 pypi_0 ...
tensorflow 2.21.0 pypi_0 ...
termcolor 3.3.0 pypi_0 ...
tifffile 2026.3.3 pypi_0 ...
tk 8.6.13 noxft_h1df4e...
tqdm 4.70.1 pypi_0 ...
typing-extensions 4.16.0 pypi_0 ...
tzdata 2026c h151e31d_0 ...
urllib3 2.8.0 pypi_0 ...
wheel 0.48.0 pyhd8ed1ab_0...
wrapt 2.4.1 pypi_0 ...
zstd 1.5.7 hb78ec9c_7 ...
(deepEMhancer) [root@rockylinux9 deepEMhancer]#
(deepEMhancer) [root@rockylinux9 deepEMhancer]# conda dea...
[root@rockylinux9 deepEMhancer]#
}}
これで一応インストールは完了です
***EnvironmentModules [#r45f11d0]
「/apps/modulefiles/deepEMhancer」
#code(nonumber){{
#%Module1.0
proc ModulesHelp { } {
puts stderr "deepEMhancer 0.18"
puts stderr "TensorFlow 2.21 / CUDA 12 / Blackwell GP...
}
module-whatis "deepEMhancer 0.18 with TensorFlow 2.21 CUD...
set root /apps/pyenv/versions/miniforge3-26.7.2-0/envs/...
set nvidia $root/lib/python3.11/site-packages/nvidia
# deepEMhancer / Python
prepend-path PATH $root/bin
# TensorFlow CUDA libraries installed by pip
prepend-path LD_LIBRARY_PATH $nvidia/nvjitlink/lib
prepend-path LD_LIBRARY_PATH $nvidia/nccl/lib
prepend-path LD_LIBRARY_PATH $nvidia/curand/lib
prepend-path LD_LIBRARY_PATH $nvidia/cuda_runtime/lib
prepend-path LD_LIBRARY_PATH $nvidia/cuda_nvrtc/lib
prepend-path LD_LIBRARY_PATH $nvidia/cuda_cupti/lib
prepend-path LD_LIBRARY_PATH $nvidia/cusparse/lib
prepend-path LD_LIBRARY_PATH $nvidia/cufft/lib
prepend-path LD_LIBRARY_PATH $nvidia/cublas/lib
prepend-path LD_LIBRARY_PATH $nvidia/cusolver/lib
prepend-path LD_LIBRARY_PATH $nvidia/cudnn/lib
}}
***動作確認 [#e0d8fbf1]
GPUを利用して計算速度を向上できます
実際にGPUが使えるかのテストをしてみます
#code(nonumber){{
[saber@rockylinux9 ~]$ module load deepEMhancer
[saber@rockylinux9 ~]$ which python
/apps/pyenv/versions/miniforge3-26.7.2-0/envs/deepEMhance...
[saber@rockylinux9 ~]$ which deepemhancer
/apps/pyenv/versions/miniforge3-26.7.2-0/envs/deepEMhance...
[saber@rockylinux9 ~]$ deepemhancer --version
0.18
[saber@rockylinux9 ~]$ python -c "import tensorflow as t...
WARNING: All log messages before absl::InitializeLog() is...
I0000 00:00:1790433738.879650 40767 port.cc:153] oneDNN...
I0000 00:00:1790433738.905989 40767 cpu_feature_guard.c...
To enable the following instructions: AVX2 AVX_VNNI FMA, ...
WARNING: All log messages before absl::InitializeLog() is...
I0000 00:00:1790433739.567853 40767 port.cc:153] oneDNN...
W0000 00:00:1790433739.884478 40767 gpu_device.cc:2459]...
[PhysicalDevice(name='/physical_device:GPU:0', device_typ...
[saber@rockylinux9 ~]$
}}
「oneDNN custom operations are on」はoneDNN（oneAPI Deep ...
でも「You may see slightly different numerical results du...
「To turn them off, set the environment variable `TF_ENAB...
あと最新版のtensorflow 2.21.0 を使ってますが、cc12.0 つま...
tensorflow の sm_120向けはコンパイルすればできそうですが...
***つかう [#v940e17c]
モデルファイルが必要ですが、/appsとかに配置しておきます
#code(nonumber){{
[root@rockylinux9 ~]# module use /apps/modulefiles/
[root@rockylinux9 ~]# module load deepEMhancer
[root@rockylinux9 ~]# deepemhancer --download /apps/deepE...
[root@rockylinux9 ~]# ls -l /apps/deepEMhancerModels/
total 0
drwxr-xr-x 2 root root 160 Sep 26 23:57 production_checkp...
[root@rockylinux9 ~]# ls -l /apps/deepEMhancerModels/prod...
total 904376
-rw-r--r-- 1 root root 311505312 Sep 26 23:57 deepEMhance...
-rw-r--r-- 1 root root 204777344 Sep 26 23:57 deepEMhance...
-rw-r--r-- 1 root root 204886472 Sep 26 23:57 deepEMhance...
-rw-r--r-- 1 root root 204897936 Sep 26 23:57 deepEMhance...
-rw-r--r-- 1 root root 74 Sep 26 23:57 versions.txt
[root@rockylinux9 ~]#
}}
っでテストrunです.
#code(nonumber){{
[saber@rockylinux9 ~]$ module use /apps/modulefiles/
[saber@rockylinux9 ~]$ module load deepEMhancer
[saber@rockylinux9 ~]$ mkdir -p test/deepEMhancer && cd $_
[saber@rockylinux9 deepEMhancer]$ deepemhancer -h
usage: deepemhancer -i INPUTMAP -o OUTPUTMAP [-p {wideTar...
[-m BINARYMASK] [--deepLearningModelP...
[--version] [-h] [--download [DOWNLOA...
（略
[saber@rockylinux9 deepEMhancer]$
}}
これが正しいか不明ですが、relionのデータを使って流してみた.
#code(nonumber){{
[saber@rockylinux9 deepEMhancer]$
[saber@rockylinux9 deepEMhancer]$ deepemhancer -i ../Tuto...
-i2 ../Tutorial5.0/R...
-p tightTarget -o model.mrc --deepLearni...
updating environment to select gpu: [0]
WARNING: All log messages before absl::InitializeLog() is...
I0000 00:00:1790442345.232244 40966 port.cc:153] oneDNN...
WARNING: All log messages before absl::InitializeLog() is...
I0000 00:00:1790442345.906649 40966 port.cc:153] oneDNN...
Loading and averaging half maps...
Half maps loaded.
loading model /apps/deepEMhancerModels/production_checkpo...
Loading and normalizing input volume...
Automatic radial noise detected beyond 36 % of volume side
DONE!. Shape at 1.00 A/voxel after padding-> (320, 320, ...
Neural net inference
:
:
[saber@rockylinux9 deepEMhancer]$ ls -l
total 65540
-rw-r--r-- 1 saber saber 67109888 Sep 27 02:18 model.mrc
[saber@rockylinux9 deepEMhancer]$
}}
***cryoSPARC向けのスクリプト [#z94ec15a]
「/apps/cs_deepEMhancer.sh」
#code(nonumber){{
#!/usr/bin/env bash
if command -v conda > /dev/null 2>&1; then
conda deactivate > /dev/null 2>&1 || true # ignore any e...
conda deactivate > /dev/null 2>&1 || true # ignore any e...
fi
unset _CE_CONDA
unset CONDA_DEFAULT_ENV
unset CONDA_EXE
unset CONDA_PREFIX
unset CONDA_PROMPT_MODIFIER
unset CONDA_PYTHON_EXE
unset CONDA_SHLVL
unset PYTHONPATH
unset LD_PRELOAD
unset LD_LIBRARY_PATH
source /apps/pyenv/versions/miniforge3-26.7.2-0/etc/profi...
conda activate deepEMhancer
exec deepemhancer $@
}}
終了行:
本家様 [[https://github.com/rsanchezgarc/deepEMhancer>+ht...
gitには
#code(nonumber){{
Simply speaking, DeepEMhancer performs a non-linear post-...
1. Local sharpening-like post-processing.
2. Automatic masking/denoising of cryo-EM maps.
}}
とある. cryoSPARCからもコールされるようになっている.
インストール環境はこんな感じ
#code(nonumber){{
[root@rockylinux9 ~]# cat /etc/redhat-release
Rocky Linux release 9.8 (Blue Onyx)
[root@rockylinux9 ~]# cat /proc/driver/nvidia/version
NVRM version: NVIDIA UNIX Open Kernel Module for x86_64 ...
GCC version: gcc version 11.5.0 20240719 (Red Hat 11.5.0...
[root@rockylinux9 ~]# nvidia-smi -L
GPU 0: NVIDIA RTX PRO 2000 Blackwell (UUID: GPU-40660e37-...
[root@rockylinux9 ~]#
}}
「/usr/local/cuda」は用意していないです
***インストール [#j5b5561d]
DeepEMhancerはpythonアプリなので、pyenv/miniforgeの環境に...
pyenv/miniforge環境がないなら下記のように敷設して
#code(nonumber){{
git clone https://github.com/yyuu/pyenv.git /apps/pyenv
export PYENV_ROOT=/apps/pyenv
export PATH=$PYENV_ROOT/bin:$PATH
pyenv install miniforge3-26.7.2-0
source /apps/pyenv/versions/miniforge3-26.7.2-0/etc/profi...
conda update -n base -c conda-forge conda
(既に環境があるなら)
source /apps/pyenv/versions/miniforge3-26.7.2-0/etc/profi...
}}
っで、DeepEMhancer を入れていきます.
#code(nonumber){{
[root@rockylinux9 ~]# cd /apps
[root@rockylinux9 apps]# git clone https://github.com/rsa...
[root@rockylinux9 apps]# cd deepEMhancer/
[root@rockylinux9 deepEMhancer]# ls -CF
alternative_installation/ condaDeepEMHancer/ conda_ymls...
[root@rockylinux9 deepEMhancer]# cat deepEMhancer_env.yml
channels:
- conda-forge
- defaults
dependencies:
- pip=26.2
- python=3.11
- pip:
- h5py==3.14.0
- mrcfile==1.5.4
- numpy==2.2.6
- requests==2.34.2
- scikit-image==0.26.0
- scipy==1.17.1
- tensorflow[and-cuda]==2.21.0
- tqdm==4.70.1
[root@rockylinux9 deepEMhancer]#
[root@rockylinux9 deepEMhancer]# conda env create -f deep...
}}
作った deepEMhancer 実行環境に対して
#code(nonumber){{
[root@rockylinux9 deepEMhancer]# conda activate deepEMhan...
(deepEMhancer) [root@rockylinux9 deepEMhancer]#
(deepEMhancer) [root@rockylinux9 deepEMhancer]# which pyt...
/apps/pyenv/versions/miniforge3-26.7.2-0/envs/deepEMhance...
(deepEMhancer) [root@rockylinux9 deepEMhancer]# python -m...
:
Successfully built deepEMhancer
Installing collected packages: deepEMhancer
Successfully installed deepEMhancer-0.18
(deepEMhancer) [root@rockylinux9 deepEMhancer]#
(deepEMhancer) [root@rockylinux9 deepEMhancer]# conda list
# packages in environment at /apps/pyenv/versions/minifor...
#
# Name Version Build ...
_openmp_mutex 4.5 20_gnu ...
absl-py 2.5.0 pypi_0 ...
astunparse 1.6.3 pypi_0 ...
bzip2 1.0.8 hda65f42_10 ...
ca-certificates 2026.7.22 hbd8a1cb_0 ...
certifi 2026.7.22 pypi_0 ...
charset-normalizer 3.5.1 pypi_0 ...
deepemhancer 0.18 pypi_0 ...
flatbuffers 25.12.19 pypi_0 ...
gast 0.7.0 pypi_0 ...
google-pasta 0.2.0 pypi_0 ...
grpcio 1.84.0 pypi_0 ...
h5py 3.14.0 pypi_0 ...
icu 78.3 py310h44b86e...
idna 3.20 pypi_0 ...
imageio 2.37.4 pypi_0 ...
keras 3.15.1 pypi_0 ...
lazy-loader 0.6 pypi_0 ...
ld_impl_linux-64 2.46.1 default_hbd6...
libclang 18.1.1 pypi_0 ...
libexpat 2.8.5 hd2095e1_0 ...
libffi 3.7.0 h81df57d_1 ...
libgcc 16.2.0 ha9f2e26_7 ...
libgomp 16.2.0 he0feb66_7 ...
liblzma 5.8.3 hb03c661_1 ...
libnsl 2.0.1 hb9d3cd8_1 ...
libpython 3.11.16 h0c77377_2_c...
libsqlite 3.53.4 h13e7031_1 ...
libstdcxx 16.2.0 h934c35e_7 ...
libuuid 2.42.4 hcfc3c73_0 ...
libxcrypt 4.4.38 h280c20c_0 ...
libzlib 1.3.2 h25fd6f3_3 ...
markdown-it-py 4.2.0 pypi_0 ...
mdurl 0.1.2 pypi_0 ...
ml-dtypes 0.6.0 pypi_0 ...
mrcfile 1.5.4 pypi_0 ...
namex 0.1.0 pypi_0 ...
ncurses 6.6 hdb14827_1 ...
networkx 3.6.1 pypi_0 ...
numpy 2.2.6 pypi_0 ...
nvidia-cublas-cu12 12.9.2.10 pypi_0 ...
nvidia-cuda-cupti-cu12 12.9.79 pypi_0 ...
nvidia-cuda-nvcc-cu12 12.9.86 pypi_0 ...
nvidia-cuda-nvrtc-cu12 12.9.86 pypi_0 ...
nvidia-cuda-runtime-cu12 12.9.79 pypi_0 ...
nvidia-cudnn-cu12 9.26.0.51 pypi_0 ...
nvidia-cufft-cu12 11.4.1.4 pypi_0 ...
nvidia-curand-cu12 10.3.10.19 pypi_0 ...
nvidia-cusolver-cu12 11.7.5.82 pypi_0 ...
nvidia-cusparse-cu12 12.5.10.65 pypi_0 ...
nvidia-nccl-cu12 2.32.3 pypi_0 ...
nvidia-nvjitlink-cu12 12.9.86 pypi_0 ...
openssl 3.6.4 h781a0a9_0 ...
opt-einsum 3.4.0 pypi_0 ...
optree 0.20.0 pypi_0 ...
packaging 26.3 pyhc364b38_0...
pillow 12.3.0 pypi_0 ...
pip 26.2.1 pyh8b19718_0...
protobuf 7.36.2 pypi_0 ...
pygments 2.21.0 pypi_0 ...
python 3.11.16 h5f976f7_2_c...
readline 8.3 hd6e31c0_1 ...
requests 2.34.2 pypi_0 ...
rich 15.0.0 pypi_0 ...
scikit-image 0.26.0 pypi_0 ...
scipy 1.17.1 pypi_0 ...
setuptools 84.0.0 pyh332efcf_0...
six 1.17.0 pypi_0 ...
tensorflow 2.21.0 pypi_0 ...
termcolor 3.3.0 pypi_0 ...
tifffile 2026.3.3 pypi_0 ...
tk 8.6.13 noxft_h1df4e...
tqdm 4.70.1 pypi_0 ...
typing-extensions 4.16.0 pypi_0 ...
tzdata 2026c h151e31d_0 ...
urllib3 2.8.0 pypi_0 ...
wheel 0.48.0 pyhd8ed1ab_0...
wrapt 2.4.1 pypi_0 ...
zstd 1.5.7 hb78ec9c_7 ...
(deepEMhancer) [root@rockylinux9 deepEMhancer]#
(deepEMhancer) [root@rockylinux9 deepEMhancer]# conda dea...
[root@rockylinux9 deepEMhancer]#
}}
これで一応インストールは完了です
***EnvironmentModules [#r45f11d0]
「/apps/modulefiles/deepEMhancer」
#code(nonumber){{
#%Module1.0
proc ModulesHelp { } {
puts stderr "deepEMhancer 0.18"
puts stderr "TensorFlow 2.21 / CUDA 12 / Blackwell GP...
}
module-whatis "deepEMhancer 0.18 with TensorFlow 2.21 CUD...
set root /apps/pyenv/versions/miniforge3-26.7.2-0/envs/...
set nvidia $root/lib/python3.11/site-packages/nvidia
# deepEMhancer / Python
prepend-path PATH $root/bin
# TensorFlow CUDA libraries installed by pip
prepend-path LD_LIBRARY_PATH $nvidia/nvjitlink/lib
prepend-path LD_LIBRARY_PATH $nvidia/nccl/lib
prepend-path LD_LIBRARY_PATH $nvidia/curand/lib
prepend-path LD_LIBRARY_PATH $nvidia/cuda_runtime/lib
prepend-path LD_LIBRARY_PATH $nvidia/cuda_nvrtc/lib
prepend-path LD_LIBRARY_PATH $nvidia/cuda_cupti/lib
prepend-path LD_LIBRARY_PATH $nvidia/cusparse/lib
prepend-path LD_LIBRARY_PATH $nvidia/cufft/lib
prepend-path LD_LIBRARY_PATH $nvidia/cublas/lib
prepend-path LD_LIBRARY_PATH $nvidia/cusolver/lib
prepend-path LD_LIBRARY_PATH $nvidia/cudnn/lib
}}
***動作確認 [#e0d8fbf1]
GPUを利用して計算速度を向上できます
実際にGPUが使えるかのテストをしてみます
#code(nonumber){{
[saber@rockylinux9 ~]$ module load deepEMhancer
[saber@rockylinux9 ~]$ which python
/apps/pyenv/versions/miniforge3-26.7.2-0/envs/deepEMhance...
[saber@rockylinux9 ~]$ which deepemhancer
/apps/pyenv/versions/miniforge3-26.7.2-0/envs/deepEMhance...
[saber@rockylinux9 ~]$ deepemhancer --version
0.18
[saber@rockylinux9 ~]$ python -c "import tensorflow as t...
WARNING: All log messages before absl::InitializeLog() is...
I0000 00:00:1790433738.879650 40767 port.cc:153] oneDNN...
I0000 00:00:1790433738.905989 40767 cpu_feature_guard.c...
To enable the following instructions: AVX2 AVX_VNNI FMA, ...
WARNING: All log messages before absl::InitializeLog() is...
I0000 00:00:1790433739.567853 40767 port.cc:153] oneDNN...
W0000 00:00:1790433739.884478 40767 gpu_device.cc:2459]...
[PhysicalDevice(name='/physical_device:GPU:0', device_typ...
[saber@rockylinux9 ~]$
}}
「oneDNN custom operations are on」はoneDNN（oneAPI Deep ...
でも「You may see slightly different numerical results du...
「To turn them off, set the environment variable `TF_ENAB...
あと最新版のtensorflow 2.21.0 を使ってますが、cc12.0 つま...
tensorflow の sm_120向けはコンパイルすればできそうですが...
***つかう [#v940e17c]
モデルファイルが必要ですが、/appsとかに配置しておきます
#code(nonumber){{
[root@rockylinux9 ~]# module use /apps/modulefiles/
[root@rockylinux9 ~]# module load deepEMhancer
[root@rockylinux9 ~]# deepemhancer --download /apps/deepE...
[root@rockylinux9 ~]# ls -l /apps/deepEMhancerModels/
total 0
drwxr-xr-x 2 root root 160 Sep 26 23:57 production_checkp...
[root@rockylinux9 ~]# ls -l /apps/deepEMhancerModels/prod...
total 904376
-rw-r--r-- 1 root root 311505312 Sep 26 23:57 deepEMhance...
-rw-r--r-- 1 root root 204777344 Sep 26 23:57 deepEMhance...
-rw-r--r-- 1 root root 204886472 Sep 26 23:57 deepEMhance...
-rw-r--r-- 1 root root 204897936 Sep 26 23:57 deepEMhance...
-rw-r--r-- 1 root root 74 Sep 26 23:57 versions.txt
[root@rockylinux9 ~]#
}}
っでテストrunです.
#code(nonumber){{
[saber@rockylinux9 ~]$ module use /apps/modulefiles/
[saber@rockylinux9 ~]$ module load deepEMhancer
[saber@rockylinux9 ~]$ mkdir -p test/deepEMhancer && cd $_
[saber@rockylinux9 deepEMhancer]$ deepemhancer -h
usage: deepemhancer -i INPUTMAP -o OUTPUTMAP [-p {wideTar...
[-m BINARYMASK] [--deepLearningModelP...
[--version] [-h] [--download [DOWNLOA...
（略
[saber@rockylinux9 deepEMhancer]$
}}
これが正しいか不明ですが、relionのデータを使って流してみた.
#code(nonumber){{
[saber@rockylinux9 deepEMhancer]$
[saber@rockylinux9 deepEMhancer]$ deepemhancer -i ../Tuto...
-i2 ../Tutorial5.0/R...
-p tightTarget -o model.mrc --deepLearni...
updating environment to select gpu: [0]
WARNING: All log messages before absl::InitializeLog() is...
I0000 00:00:1790442345.232244 40966 port.cc:153] oneDNN...
WARNING: All log messages before absl::InitializeLog() is...
I0000 00:00:1790442345.906649 40966 port.cc:153] oneDNN...
Loading and averaging half maps...
Half maps loaded.
loading model /apps/deepEMhancerModels/production_checkpo...
Loading and normalizing input volume...
Automatic radial noise detected beyond 36 % of volume side
DONE!. Shape at 1.00 A/voxel after padding-> (320, 320, ...
Neural net inference
:
:
[saber@rockylinux9 deepEMhancer]$ ls -l
total 65540
-rw-r--r-- 1 saber saber 67109888 Sep 27 02:18 model.mrc
[saber@rockylinux9 deepEMhancer]$
}}
***cryoSPARC向けのスクリプト [#z94ec15a]
「/apps/cs_deepEMhancer.sh」
#code(nonumber){{
#!/usr/bin/env bash
if command -v conda > /dev/null 2>&1; then
conda deactivate > /dev/null 2>&1 || true # ignore any e...
conda deactivate > /dev/null 2>&1 || true # ignore any e...
fi
unset _CE_CONDA
unset CONDA_DEFAULT_ENV
unset CONDA_EXE
unset CONDA_PREFIX
unset CONDA_PROMPT_MODIFIER
unset CONDA_PYTHON_EXE
unset CONDA_SHLVL
unset PYTHONPATH
unset LD_PRELOAD
unset LD_LIBRARY_PATH
source /apps/pyenv/versions/miniforge3-26.7.2-0/etc/profi...
conda activate deepEMhancer
exec deepemhancer $@
}}
ページ名:
1
![[PukiWiki]](image/picc.png "[PukiWiki]")
