Skip to content
This repository was archived by the owner on Jan 15, 2024. It is now read-only.

Commit df52171

Browse files
authored
update docker (#1571)
1 parent 5122e84 commit df52171

File tree

2 files changed

+3
-7
lines changed

2 files changed

+3
-7
lines changed

tools/batch/submit-job.py

+1-1
Original file line numberDiff line numberDiff line change
@@ -10,7 +10,7 @@
1010

1111
instance_type_info = {
1212
'g4dn.4x': {
13-
'job_definition': 'gluon-nlp-g4dn_4xlarge:5',
13+
'job_definition': 'gluon-nlp-g4dn_4xlarge:6',
1414
'job_queue': 'g4dn'
1515
},
1616
'g4dn.8x': {

tools/docker/gluon_nlp_job.sh

+2-6
Original file line numberDiff line numberDiff line change
@@ -23,14 +23,10 @@ git fetch origin $SOURCE_REF:working
2323
git checkout working
2424

2525
if [ $DEVICE == "cpu" ]; then
26+
python3 -m pip uninstall --quiet mxnet -y
2627
python3 -m pip install -U --quiet --pre "mxnet>=2.0.0b20210121" --user
2728
else
28-
# Due to the issue in https://forums.aws.amazon.com/thread.jspa?messageID=953912
29-
# We need to manually configure the shm to ensure that Horovod is runnable.
30-
# The reason that we need a larger shm is described in https://github.com/NVIDIA/nccl/issues/290
31-
umount shm
32-
mount -t tmpfs -o rw,nosuid,nodev,noexec,relatime,size=2G shm /dev/shm
33-
sudo python3 -m pip uninstall --quiet mxnet-cu102
29+
python3 -m pip uninstall --quiet mxnet-cu102 -y
3430
python3 -m pip install -U --quiet --pre "mxnet-cu102>=2.0.0a" --user
3531
fi
3632

0 commit comments

Comments
 (0)