init checkin for aws benchmarking tool

7 years ago · a2a5ffae60
parent d40096ca7e
commit a2a5ffae60
4 changed files with 399 additions and 0 deletions
--- a/tools/aws_benchmarking/paddle_banchmarking_aws.py
+++ b/tools/aws_benchmarking/paddle_banchmarking_aws.py
--- a/tools/aws_benchmarking/pserver.sh.template
+++ b/tools/aws_benchmarking/pserver.sh.template
@ -0,0 +1 @@
+nvidia-docker run -i -e "TRAINING_ROLE=PSERVER"  -e "PSERVER_HOSTS={PSERVER_HOSTS}" {DOCKER_IMAGE}
--- a/tools/aws_benchmarking/requirements.txt
+++ b/tools/aws_benchmarking/requirements.txt
@ -0,0 +1,4 @@
+netaddr==0.7.19
+boto3==1.6.21
+namesgenerator==0.3
+paramiko==2.4.1
--- a/tools/aws_benchmarking/trainer.sh.template
+++ b/tools/aws_benchmarking/trainer.sh.template
@ -0,0 +1 @@
+nvidia-docker run -i -e "TRAINER_INDEX={TRAINER_INDEX}" -e "TRAINING_ROLE=TRAINER"  -e "PSERVER_HOSTS={PSERVER_HOSTS}" {DOCKER_IMAGE}
				`@ -0,0 +1 @@`
				`nvidia-docker run -i -e "TRAINING_ROLE=PSERVER" -e "PSERVER_HOSTS={PSERVER_HOSTS}" {DOCKER_IMAGE}`
				`@ -0,0 +1 @@`
				`nvidia-docker run -i -e "TRAINER_INDEX={TRAINER_INDEX}" -e "TRAINING_ROLE=TRAINER" -e "PSERVER_HOSTS={PSERVER_HOSTS}" {DOCKER_IMAGE}`