backup

dltrainer/Dockerfile

@@ -57,13 +57,13 @@ RUN conda install future
 RUN conda install tensorflow
 RUN pip install git+https://github.com/lanpa/tensorboard-pytorch
 RUN pip install apscheduler
-RUN pip install gpustat
+# RUN pip install gpustat
+RUN pip install pyrebase
+RUN pip install requests
+RUN pip install python-firebase
 
 ENV PATH /opt/conda/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
 
-COPY dlinputs /tmp/dlinputs
-RUN cd /tmp/dlinputs && python setup.py install && rm -rf /tmp/dlinputs
-
 ENV USER user
 ENV HOME /home/$USER
 ENV GID 1000

dltrainer/trainer-deploy.yml

@@ -11,7 +11,7 @@ spec:
     spec:
       containers:
       - name: imagenet-trainer
-        image: gcr.io/sonic-column-220616/dloncloud:v9 # TODO Put location of your image on cloud repository
+        image: gcr.io/sonic-column-220616/dloncloud:v1 # TODO Put location of your image on cloud repository
         ports:
         - containerPort: 4545
         command: ["python"]

ku

@@ -9,11 +9,11 @@ cmd_init() { # initialize the cluster
     # set -e
     gcloud config set project $project
     gcloud config set compute/zone $zone
-    gcloud config set container/new_scopes_behavior true
-    # gcloud config set container/use_v1_api false
+    gcloud config set container/new_scopes_behavior false
+    # gcloud config set container/use_v1_api true
     gcloud config set container/cluster $cluster
     gcloud beta container clusters create $cluster \
-        --zone $zone --cluster-version 1.9 \
+        --zone $zone \
         --machine-type ${cpu_machine:-n1-standard-8} \
         --enable-autorepair \
         --scopes $scopes \

kube-cluster-config.sh

@@ -1,12 +1,11 @@
 cluster=cloud-infra-layer
 zone=us-east1-c
 project=sonic-column-220616
-# project=cloudinfra-219900
 image=gcr.io/$project/dloncloud
 cpu_machine=n1-standard-1
 cpu_nodes=1
 gpu_machine=n1-standard-1
 gpu_type=nvidia-tesla-k80
-gpu_nodes=4
+gpu_nodes=1
 gpu_per_node=1
 gpu_max_nodes=4