Closed superdavidxp closed 3 years ago
Going by this, could you try to make sure you don't have multiple versions of torch_sparse
installed, and that the *.so
files are getting created correctly?
I don't think I have multiple versions of torch_parse
, following is the info I gathered from our system, thank you for your advice in advance,
Singularity> ls /opt/miniconda/lib/python3.7/site-packages/torch_sparse/*.so -al
-rwxr-xr-x 1 root root 134296 May 23 08:04 /opt/miniconda/lib/python3.7/site-packages/torch_sparse/_convert_cpu.so
-rwxr-xr-x 1 root root 134240 May 23 08:04 /opt/miniconda/lib/python3.7/site-packages/torch_sparse/_diag_cpu.so
-rwxr-xr-x 1 root root 134288 May 23 08:04 /opt/miniconda/lib/python3.7/site-packages/torch_sparse/_metis_cpu.so
-rwxr-xr-x 1 root root 200752 May 23 08:04 /opt/miniconda/lib/python3.7/site-packages/torch_sparse/_relabel_cpu.so
-rwxr-xr-x 1 root root 134368 May 23 08:04 /opt/miniconda/lib/python3.7/site-packages/torch_sparse/_rw_cpu.so
-rwxr-xr-x 1 root root 134656 May 23 08:04 /opt/miniconda/lib/python3.7/site-packages/torch_sparse/_saint_cpu.so
-rwxr-xr-x 1 root root 134776 May 23 08:04 /opt/miniconda/lib/python3.7/site-packages/torch_sparse/_sample_cpu.so
-rwxr-xr-x 1 root root 597592 May 23 08:04 /opt/miniconda/lib/python3.7/site-packages/torch_sparse/_spmm_cpu.so
-rwxr-xr-x 1 root root 200816 May 23 08:04 /opt/miniconda/lib/python3.7/site-packages/torch_sparse/_spspmm_cpu.so
-rwxr-xr-x 1 root root 67952 May 23 08:04 /opt/miniconda/lib/python3.7/site-packages/torch_sparse/_version_cpu.so
Singularity> ls /opt/miniconda/lib/python3.7/site-packages/torch
torch/ torch_scatter/ torchaudio/
torch-1.7.0a0-py3.7.egg-info/ torch_scatter-2.0.6.dist-info/ torchaudio-0.7.0a0+a853dff.dist-info/
torch_cluster/ torch_sparse/ torchtext/
torch_cluster-1.5.9.dist-info/ torch_sparse-0.6.9.dist-info/ torchtext-0.6.0.dist-info/
torch_geometric/ torch_spline_conv/ torchvision/
torch_geometric-1.7.0.dist-info/ torch_spline_conv-1.2.1.dist-info/ torchvision-0.8.0a0+2f40a48.dist-info/
I have the OCP repo set up on a cluster that's also IBM Power. Open-CE: https://github.com/open-ce/open-ce provides builds for Power-architecture systems. I first created an environment from open-ce/1.1.3
. I then installed the appropriate pytorch geometric binaries and any other missing packages OCP required.
My conda env list is as follows for that system (NOTE - there are several pacakages here that are not required by OCP that were part of the Open-CE environment).
Hope this helps! If by any chance you are on the DOE's Summit system, I'm happy to point you to my environment.
# packages in environment at /ccs/home/mshuaibi/.conda/envs/ocp-models:
#
# Name Version Build Channel
_libgcc_mutex 0.1 main
_pytorch_select 2.0 cuda10.2_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
_tensorflow_select 2.0 cuda10.2_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
_xgboost_select 2.0 cuda10.2_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
absl-py 0.10.0 py36_0
aiohttp 3.7.4 py36h140841e_1
appdirs 1.4.4 pypi_0 pypi
ase 3.21.1 pypi_0 pypi
astunparse 1.6.3 py_0
async-timeout 3.0.1 py36h6ffa863_0
attrs 19.3.0 py_0
av 8.0.2 py36haa86098_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
binutils_impl_linux-ppc64le 2.31.1 he53550c_1
binutils_linux-ppc64le 2.31.1 he53550c_8
blas 1.0 openblas
blinker 1.4 py36h6ffa863_0
brotlipy 0.7.0 py36h140841e_1003
bzip2 1.0.8 h7b6447c_0
c-ares 1.17.1 h140841e_0
ca-certificates 2021.1.19 h6ffa863_1
cachetools 4.2.1 pyhd3eb1b0_0
cairo 1.16.0 he491a88_1
catalogue 1.0.0 py36_1
certifi 2020.12.5 py36h6ffa863_0
cffi 1.14.5 py36hf9d8e4b_0
cfgv 3.2.0 pypi_0 pypi
chardet 3.0.4 py36h6ffa863_1003
click 7.0 py36_0
cloudpickle 1.3.0 py_0
cmake 3.14.0 h52cb24c_0
configparser 5.0.2 pypi_0 pypi
cryptography 3.4.7 py36h7ed74fa_0
cudatoolkit 10.2.89 hfd86e86_1
cudnn 7.6.5_10.2 h9286eec_2 file:///sw/sources/open-ce/conda-channel-v1.1.3
cycler 0.10.0 py36_0
cymem 2.0.5 py36h29c3540_0
cython-blis 0.4.1 py36h7b6447c_1
cytoolz 0.11.0 py36h7b6447c_0
dali 0.28.0 cuda10.2_py36_3 file:///sw/sources/open-ce/conda-channel-v1.1.3
dask-core 2021.3.1 pyhd3eb1b0_0
dataclasses 0.7 py36h6ffa863_0
decorator 4.4.2 pyhd3eb1b0_0
demjson 2.2.4 pypi_0 pypi
dill 0.3.3 pyhd3eb1b0_0
distlib 0.3.1 pypi_0 pypi
dm-tree 0.1.5 py36hb8906cf_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
docker-pycreds 0.4.0 pypi_0 pypi
expat 2.3.0 h29c3540_2
ffmpeg 4.2.2 h20bf706_0
filelock 3.0.12 pyhd3eb1b0_1
fontconfig 2.13.1 ha0a49a9_0
freeglut 3.0.0 hf484d3e_5
freetype 2.10.4 h5ab3b9f_0
fsspec 0.8.7 pyhd3eb1b0_0
future 0.18.2 py36_1
gast 0.3.3 py_0
gcc_impl_linux-ppc64le 7.3.0 he01c8ba_1
gcc_linux-ppc64le 7.3.0 h48e019a_8
gettext 0.20.2 h1b965fe_0
giflib 5.2.1 h7b6447c_0
gitdb 4.0.7 pypi_0 pypi
gitpython 3.1.14 pypi_0 pypi
glib 2.68.0 h1318424_0
gmp 6.2.1 h29c3540_0
gnutls 3.6.5 h71b1129_1002
google-auth 1.23.0 pyhd3eb1b0_0
google-auth-oauthlib 0.4.4 pyhd3eb1b0_0
google-pasta 0.2.0 py_0
googleapis-common-protos 1.52.0 py36h6ffa863_0
googledrivedownloader 0.4 pypi_0 pypi
graphite2 1.3.14 h23475e2_0
graphsurgeon 0.4.1 py36_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
grpcio 1.31.0 py36hf8bcb03_0
gxx_impl_linux-ppc64le 7.3.0 h822a55f_1
gxx_linux-ppc64le 7.3.0 h48e019a_8
h5py 2.10.0 py36h7918eee_0
harfbuzz 1.8.8 hffaf4a1_0
hdf5 1.10.4 hb1b8bf9_0
horovod 0.21.0 cuda10.2_system_py36_5 file:///sw/sources/open-ce/conda-channel-v1.1.3
icu 58.2 he6710b0_3
identify 2.2.3 pypi_0 pypi
idna 2.8 py36_0
idna_ssl 1.1.0 py36h6ffa863_0
imageio 2.9.0 py_0
importlib-metadata 3.7.3 py36h6ffa863_1
importlib_metadata 3.7.3 hd3eb1b0_1
importlib_resources 3.3.0 py36h6ffa863_0
isodate 0.6.0 pypi_0 pypi
jasper 2.0.14 h07fcdf6_1
jinja2 2.11.3 pypi_0 pypi
joblib 0.17.0 py_0
jpeg 9b hcb7ba68_2
jpeg-turbo 2.0.5 h8b879f5_2 file:///sw/sources/open-ce/conda-channel-v1.1.3
jsonschema 3.2.0 py_2
keras-preprocessing 1.1.2 pyhd3eb1b0_0
kiwisolver 1.3.1 py36h29c3540_0
krb5 1.18.2 h597af5e_0
lame 3.100 h7b6447c_0
lcms2 2.11 h396b838_0
ld_impl_linux-ppc64le 2.33.1 h0f24833_7
leveldb 1.20 hf484d3e_1
libcurl 7.71.1 h20c2e04_1
libedit 3.1.20210216 h140841e_1
libevent 2.1.11 hafc74fa_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
libffi 3.3 he6710b0_2
libflac 1.3.1 py38_2 file:///sw/sources/open-ce/conda-channel-v1.1.3
libgcc-ng 8.2.0 h822a55f_1
libgfortran-ng 7.3.0 h822a55f_1
libglu 9.0.0 hf484d3e_1
libogg 1.3.2 h7b6447c_0
libopenblas 0.3.6 h5a2b251_1
libopencv 3.4.10 py36_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
libopus 1.3.1 h7b6447c_0
libpng 1.6.37 hbc83047_0
libprotobuf 3.9.2 h847787d_3 file:///sw/sources/open-ce/conda-channel-v1.1.3
libsndfile 1.0.28 py38_2 file:///sw/sources/open-ce/conda-channel-v1.1.3
libssh2 1.9.0 h1ba5d50_1
libstdcxx-ng 8.2.0 h822a55f_1
libtensorflow 2.4.1 cuda10.2_py38_4 file:///sw/sources/open-ce/conda-channel-v1.1.3
libtiff 4.2.0 h781710b_0
libuuid 1.0.3 h1bed415_2
libvorbis 1.3.7 h7b6447c_0
libvpx 1.7.0 hf484d3e_0
libwebp 1.2.0 he32dc1f_0
libwebp-base 1.2.0 h140841e_0
libxcb 1.14 h7b6447c_0
libxgboost-base 1.3.3 cuda10.2_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
libxml2 2.9.10 h871c0c7_3
llvmlite 0.31.0 py36hd408876_0
lmdb 0.9.22 hf484d3e_1
lz4-c 1.9.3 h29c3540_0
magma 2.5.4 cuda10.2_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
make 4.2.1 h14c3975_1
markdown 3.1.1 py36_0
markupsafe 1.1.1 pypi_0 pypi
matplotlib-base 3.3.4 py36he087750_0
more-itertools 8.7.0 pyhd3eb1b0_0
mpmath 1.2.1 pypi_0 pypi
multidict 5.1.0 py36h140841e_2
murmurhash 1.0.5 py36h29c3540_0
nccl 2.7.8 cuda10.2_4 file:///sw/sources/open-ce/conda-channel-v1.1.3
ncurses 6.2 he6710b0_1
nettle 3.4.1 hbb512f6_0
networkx 2.3 py_0
ninja 1.9.0 py36hfd86e86_0
nodeenv 1.6.0 pypi_0 pypi
nomkl 3.0 0
numactl 2.0.12 h459fe5f_2 file:///sw/sources/open-ce/conda-channel-v1.1.3
numba 0.47.0 py36h962f231_0
numpy 1.19.2 py36h6163131_0
numpy-base 1.19.2 py36h75fe3a5_0
nvidia-dali-tf-plugin-cuda102 0.28.0 pypi_0 pypi
oauthlib 3.1.0 py_0
ocp-models 0.0.2 dev_0 <develop>
olefile 0.46 py36_0
onnx 1.6.0 py36_3 file:///sw/sources/open-ce/conda-channel-v1.1.3
openblas 0.3.6 1
openblas-devel 0.3.6 1
opencv 3.4.10 py36_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
openh264 2.1.0 hd408876_0
openssl 1.1.1k h140841e_0
opt_einsum 3.1.0 py_0
packaging 20.9 pyhd3eb1b0_0
pandas 1.1.5 pypi_0 pypi
pathtools 0.1.2 pypi_0 pypi
pcre 8.44 he6710b0_0
pillow 8.1.2 py36h3f95422_0
pip 21.0.1 py36h6ffa863_0
pixman 0.40.0 h7b6447c_0
plac 0.9.6 py36_1
pluggy 0.13.1 py36h6ffa863_0
pre-commit 2.12.0 pypi_0 pypi
preshed 3.0.2 py36he6710b0_1
promise 2.3 py36h6ffa863_0
protobuf 3.15.7 pypi_0 pypi
psutil 5.8.0 py36h140841e_1
py 1.10.0 pyhd3eb1b0_0
py-opencv 3.4.10 py36_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
py-xgboost 1.3.3 cuda10.2_py36_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
py-xgboost-base 1.3.3 cuda10.2_py36_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
pyasn1 0.4.8 py_0
pyasn1-modules 0.2.8 py_0
pycparser 2.20 py_2
pyjwt 1.7.1 py36_0
pyopenssl 20.0.1 pyhd3eb1b0_1
pyparsing 2.4.7 pyhd3eb1b0_0
pyrsistent 0.17.3 py36h7b6447c_0
pysocks 1.7.1 py36h6ffa863_0
pytest 5.4.3 py36h6ffa863_0
python 3.6.13 h836d2c2_0
python-dateutil 2.8.1 pyhd3eb1b0_0
python-flatbuffers 1.12 pyhb211add_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
python-lmdb 0.98 py36he6710b0_0
python-louvain 0.15 pypi_0 pypi
pytorch 1.7.1 cuda10.2_py36_2 file:///sw/sources/open-ce/conda-channel-v1.1.3
pytorch-base 1.7.1 cuda10.2_py36_8 file:///sw/sources/open-ce/conda-channel-v1.1.3
pytorch-lightning 1.1.0 pyh72259b7_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
pytorch-lightning-bolts 0.2.5 pyh22e1ee5_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
pytz 2021.1 pypi_0 pypi
pywavelets 1.1.1 py36h7b6447c_2
pywget 3.2 py36h6ffa863_0
pyyaml 5.4.1 py36h140841e_1
rdflib 5.0.0 pypi_0 pypi
readline 8.1 h140841e_0
regex 2020.6.8 py36h7b6447c_0
requests 2.22.0 py36_1
requests-oauthlib 1.3.0 py_0
rhash 1.4.1 hb567c45_1
rsa 4.7.2 pyhd3eb1b0_1
rust-nightly 1.50.0 hb975b1c_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
sacremoses 0.0.43 pyh87dc625_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
scikit-image 0.17.2 py36hdf5156a_0
scikit-learn 0.24.1 py36haab0e66_0
scipy 1.4.1 py36habc2bb6_0
semantic_version 2.8.5 pyh87dc625_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
sentencepiece 0.1.91 py36_3 file:///sw/sources/open-ce/conda-channel-v1.1.3
sentry-sdk 1.0.0 pypi_0 pypi
setuptools 52.0.0 py36h6ffa863_0
shortuuid 1.0.1 pypi_0 pypi
six 1.15.0 py36h6ffa863_0
smmap 4.0.0 pypi_0 pypi
snappy 1.1.8 he6710b0_0
spacy 2.3.4 py36h7bb18f6_2 file:///sw/sources/open-ce/conda-channel-v1.1.3
sqlite 3.35.3 hd7247d8_0
srsly 1.0.4 py36hf30110f_2 file:///sw/sources/open-ce/conda-channel-v1.1.3
submitit 1.3.0 pypi_0 pypi
subprocess32 3.5.4 pypi_0 pypi
sympy 1.7.1 pypi_0 pypi
tabulate 0.8.9 py36h6ffa863_0
tbb 2020.3 hfd86e86_0
tensorboard 2.4.1 pyhda65f9a_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
tensorboard-plugin-wit 1.6.0 pyhc0078e9_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
tensorflow 2.4.1 cuda10.2_py36_2 file:///sw/sources/open-ce/conda-channel-v1.1.3
tensorflow-addons 0.11.2 py36_3 file:///sw/sources/open-ce/conda-channel-v1.1.3
tensorflow-base 2.4.1 cuda10.2_py36_4 file:///sw/sources/open-ce/conda-channel-v1.1.3
tensorflow-datasets 4.1.0 pyh674cb94_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
tensorflow-estimator 2.4.0 pyhd2f33a5_2 file:///sw/sources/open-ce/conda-channel-v1.1.3
tensorflow-hub 0.10.0 pyh9958038_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
tensorflow-metadata 0.26.0 pyh17e375c_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
tensorflow-model-optimization 0.5.0 py36_2 file:///sw/sources/open-ce/conda-channel-v1.1.3
tensorflow-probability 0.12.1 py36_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
tensorflow-serving 2.4.1 cuda10.2_py38_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
tensorflow-serving-api 2.4.1 pyhdac48b3_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
tensorflow-text 2.4.1 py36_3 file:///sw/sources/open-ce/conda-channel-v1.1.3
tensorrt 7.0.0.11 cuda10.2_py36_4 file:///sw/sources/open-ce/conda-channel-v1.1.3
tensorrt-samples 7.0.0.11 cuda10.2_py36_4 file:///sw/sources/open-ce/conda-channel-v1.1.3
termcolor 1.1.0 py36h6ffa863_1
thinc 7.4.1 py36he5a58c3_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
threadpoolctl 2.1.0 pyh5ca1d4c_0
tifffile 2020.10.1 py36hdd07704_2
tk 8.6.10 hbc83047_0
tokenizers 0.9.3 py36h91be2ad_2 file:///sw/sources/open-ce/conda-channel-v1.1.3
toml 0.10.2 pypi_0 pypi
toolz 0.11.1 pyhd3eb1b0_0
torch-cluster 1.5.9 pypi_0 pypi
torch-geometric 1.6.3 pypi_0 pypi
torch-scatter 2.0.6 pypi_0 pypi
torch-sparse 0.6.9 pypi_0 pypi
torch-spline-conv 1.2.1 pypi_0 pypi
torchtext 0.8.1 py36_4 file:///sw/sources/open-ce/conda-channel-v1.1.3
torchvision 0.8.2 cuda10.2_py36_0 file:///sw/sources/open-ce/conda-channel-v1.1.3
torchvision-base 0.8.2 cuda10.2_py36_6 file:///sw/sources/open-ce/conda-channel-v1.1.3
tornado 6.1 py36h140841e_0
tqdm 4.41.1 py_0
transformers 3.5.1 pyh4cad38a_2 file:///sw/sources/open-ce/conda-channel-v1.1.3
typing-extensions 3.7.4.3 hd3eb1b0_0
typing_extensions 3.7.4.3 pyh06a4308_0
uff 0.6.5 py36_1 file:///sw/sources/open-ce/conda-channel-v1.1.3
urllib3 1.25.11 py_0
virtualenv 20.4.3 pypi_0 pypi
wandb 0.10.28 pypi_0 pypi
wasabi 0.6.0 py_0
wcwidth 0.2.5 py_0
werkzeug 0.16.0 py_0
wheel 0.35.1 pyhd3eb1b0_0
wrapt 1.12.1 py36h7b6447c_1
x264 1!157.20191217 h7b6447c_0
xz 5.2.5 h7b6447c_0
yaml 0.2.5 h7b6447c_0
yarl 1.6.3 py36h140841e_0
zipp 3.4.1 pyhd3eb1b0_0
zlib 1.2.11 h7b6447c_3
zstd 1.4.5 h9ceee32_0
@mshuaibii, Thank you very much for your help, I will try to re-produce your python env on our system first.
@mshuaibii Thank you again for your help.
I noticed that your python env doesn't comes with "ray", how do you train the model without it? I tried to comment the "import ray" and I got this error,
Thank you in advance,
(ocp_models) [root@hal02 ocp]# python main.py --mode train --config-yml configs/s2ef/200k/base.yml
Traceback (most recent call last):
File "main.py", line 144, in <module>
Runner()(config)
File "main.py", line 44, in __call__
model=config["model"],
KeyError: 'model'
@mshuaibii Thank you again for your help.
I noticed that your python env doesn't comes with "ray", how do you train the model without it? I tried to comment the "import ray" and I got this error,
Thank you in advance,
(ocp_models) [root@hal02 ocp]# python main.py --mode train --config-yml configs/s2ef/200k/base.yml Traceback (most recent call last): File "main.py", line 144, in <module> Runner()(config) File "main.py", line 44, in __call__ model=config["model"], KeyError: 'model'
Ray is actually a new addition to the repo so it's not in that environment, you should be able to pip install that though. But it isn't necessary to train a model, so commenting it out should be fine...the reason you're getting that error is because base.yml
is not a valid config to use, try something like this:
python main.py --mode train --config-yml configs/s2ef/200k/schnet/schnet.yml
base.yml
is used in the specific model config, but not to be used on its own at the command line. Hope this helps!
@mshuaibii Thank you very much for your reply,
I have tried your conmmand and had some new error
(ocp_models) [root@hal02 ocp]# python main.py --mode train --config-yml configs/s2ef/200k/schnet/schnet.yml
Traceback (most recent call last):
File "main.py", line 144, in <module>
Runner()(config)
File "main.py", line 40, in __call__
trainer = registry.get_trainer_class(
File "/root/ocp/ocpmodels/trainers/forces_trainer.py", line 83, in __init__
super().__init__(
File "/root/ocp/ocpmodels/trainers/base_trainer.py", line 82, in __init__
timestamp = datetime.datetime.fromtimestamp(timestamp).strftime(
TypeError: only integer tensors of a single element can be converted to an index
Should I modified the code or there are some conflict packages here?
Thank you in advance,
Are you using python=3.8 here? I believe this error is related to that. You can do the following to fix without modifying package versions:
timestamp = datetime.datetime.fromtimestamp(timestamp.int()).strftime(
@mshuaibii , Thank you so much, it did worked.
Now MaY I ask which dataset I should use and where should I put them, I downloaded the 200k training data but there are all *.xz files......
### Loading dataset: trajectory_lmdb
Traceback (most recent call last):
File "main.py", line 144, in <module>
Runner()(config)
File "main.py", line 40, in __call__
trainer = registry.get_trainer_class(
File "/root/ocp/ocpmodels/trainers/forces_trainer.py", line 83, in __init__
super().__init__(
File "/root/ocp/ocpmodels/trainers/base_trainer.py", line 156, in __init__
self.load()
File "/root/ocp/ocpmodels/trainers/base_trainer.py", line 163, in load
self.load_task()
File "/root/ocp/ocpmodels/trainers/forces_trainer.py", line 110, in load_task
self.train_dataset = registry.get_dataset_class(
File "/root/ocp/ocpmodels/datasets/trajectory_lmdb.py", line 51, in __init__
assert len(db_paths) > 0, f"No LMDBs found in {srcdir}"
AssertionError: No LMDBs found in data/s2ef/200k/train
@mshuaibii , Thank you so much, it did worked.
Now MaY I ask which dataset I should use and where should I put them, I downloaded the 200k training data but there are all *.xz files......
### Loading dataset: trajectory_lmdb Traceback (most recent call last): File "main.py", line 144, in <module> Runner()(config) File "main.py", line 40, in __call__ trainer = registry.get_trainer_class( File "/root/ocp/ocpmodels/trainers/forces_trainer.py", line 83, in __init__ super().__init__( File "/root/ocp/ocpmodels/trainers/base_trainer.py", line 156, in __init__ self.load() File "/root/ocp/ocpmodels/trainers/base_trainer.py", line 163, in load self.load_task() File "/root/ocp/ocpmodels/trainers/forces_trainer.py", line 110, in load_task self.train_dataset = registry.get_dataset_class( File "/root/ocp/ocpmodels/datasets/trajectory_lmdb.py", line 51, in __init__ assert len(db_paths) > 0, f"No LMDBs found in {srcdir}" AssertionError: No LMDBs found in data/s2ef/200k/train
Please use the following script to download the data: https://github.com/Open-Catalyst-Project/ocp/blob/master/scripts/download_data.py. It will take care of processing the raw data into LMDBs and moving it to the corresponding directory. Something like this:
python download_data.py --task s2ef --split 200k --get-edges --ref-energy --num-workers NUM_WORKERS_AVAILABLE
Thank you very much, it works now!
Hi,
I tried to run the code with our IBM Power system and have some issue with
torch_sparse
package, would you please give me some advice?and my conda env list is