training problem

[alinlab]

training problem

[jihoontack]

Hi, it is possible to share the version of your packages?

Maybe the error is related to the distributed launch.
Does the same error occur when running with a single GPU?
CUDA_VISIBLE_DEVICES=0 python train.py --dataset cifar10 --model resnet18 --mode simclr_CSI --shift_trans_type rotation --batch_size 32 --one_class_idx 0

Thank you,
Jihoon

[cuixianheng]

the follwing is the version of my packages:
absl-py 0.9.0
anykeystore 0.2
apex 0.1
asn1crypto 0.24.0
astor 0.8.1
attrs 20.2.0
cachetools 4.1.1
certifi 2018.1.18
chardet 3.0.4
cryptacular 1.5.5
cryptography 2.1.4
cycler 0.10.0
defusedxml 0.6.0
diffdist 0.1
future 0.18.2
gast 0.2.2
google-auth 1.21.1
google-auth-oauthlib 0.4.1
google-pasta 0.2.0
grpcio 1.27.2
h5py 2.10.0
hupper 1.10.2
idna 2.6
joblib 0.14.1
Keras-Applications 1.0.8
Keras-Preprocessing 1.1.0
keyring 10.6.0
keyrings.alt 3.0
kiwisolver 1.1.0
kornia 0.4.0
Markdown 3.2.1
matplotlib 3.2.0
numpy 1.18.2
oauthlib 3.1.0
onnx 1.6.0
opencv-python 4.2.0.32
opt-einsum 3.2.0
packaging 20.4
PasteDeploy 2.1.0
pbkdf2 1.3
Pillow 7.0.0
pip 20.2.3
plaster 1.0
plaster-pastedeploy 0.7
protobuf 3.11.3
pyasn1 0.4.8
pyasn1-modules 0.2.8
pycrypto 2.6.1
pygobject 3.26.1
pyparsing 2.4.6
pyramid 1.10.4
pyramid-mailer 0.15.1
python-dateutil 2.8.1
python3-openid 3.1.0
pytorch-lightning 0.9.0
pyxdg 0.25
PyYAML 5.3.1
repoze.sendmail 4.4.1
requests 2.24.0
requests-oauthlib 1.3.0
rsa 4.6
scikit-learn 0.22.2.post1
scipy 1.4.1
SecretStorage 2.3.1
setuptools 46.0.0
six 1.11.0
SQLAlchemy 1.3.15
ssh-import-id 5.7
tensorboard 2.2.0
tensorboard-plugin-wit 1.7.0
tensorboardX 2.1
tensorflow 1.15.0
tensorflow-estimator 1.15.1
termcolor 1.1.0
terminaltables 3.1.0
torch 1.6.0
torchlars 0.1.2
torchvision 0.7.0
tqdm 4.43.0
transaction 3.0.0
translationstring 1.3
typing-extensions 3.7.4.1
urllib3 1.22
velruse 1.1.1
venusian 3.0.0
WebOb 1.8.6
Werkzeug 1.0.0
wheel 0.30.0
wrapt 1.12.1
WTForms 2.2.1
wtforms-recaptcha 0.3.2
zope.deprecation 4.4.0
zope.interface 4.7.2
zope.sqlalchemy 1.3

There is no error when I use two GPUs，it just reports an error when I use four GPUs.
Thank you very much.

[jihoontack]

Hi!

We have validated our code with torch 1.4.0, torchvision 0.5.0 and didn't found the same error.
Maybe the torch version can be the issue for your error.

Right now, we don't have a plan to update the code for the later torch version, but we will notify you immediately if an update is made.

Thank you again for the interest and reporting the issue.