Update Flowers repository with new Buckets.

Add GOOGLE_APPLICATION_CREDENTIALS definition

flowers/requirements.txt

@@ -1,3 +1,3 @@
-apache-beam[gcp]==0.6.0
+apache-beam[gcp]==2.6.0
 pillow==4.0.0
 tensorflow==1.4.1

flowers/sample.sh

@@ -22,7 +22,7 @@ declare -r PROJECT=$(gcloud config list project --format "value(core.project)")
 declare -r JOB_ID="flowers_${USER}_$(date +%Y%m%d_%H%M%S)"
 declare -r BUCKET="gs://${PROJECT}-ml"
 declare -r GCS_PATH="${BUCKET}/${USER}/${JOB_ID}"
-declare -r DICT_FILE=gs://cloud-ml-data/img/flower_photos/dict.txt
+declare -r DICT_FILE=gs://cloud-samples-data/ml-engine/flowers/dict.txt
 
 declare -r MODEL_NAME=flowers
 declare -r VERSION_NAME=v1
@@ -39,13 +39,13 @@ set -v -e
 # CPU's.  Check progress here: https://console.cloud.google.com/dataflow
 python trainer/preprocess.py \
   --input_dict "$DICT_FILE" \
-  --input_path "gs://cloud-ml-data/img/flower_photos/eval_set.csv" \
+  --input_path "gs://cloud-samples-data/ml-engine/flowers/eval_set.csv" \
   --output_path "${GCS_PATH}/preproc/eval" \
   --cloud
 
 python trainer/preprocess.py \
   --input_dict "$DICT_FILE" \
-  --input_path "gs://cloud-ml-data/img/flower_photos/train_set.csv" \
+  --input_path "gs://cloud-samples-data/ml-engine/flowers/train_set.csv" \
   --output_path "${GCS_PATH}/preproc/train" \
   --cloud
 
@@ -87,7 +87,7 @@ gcloud ml-engine versions set-default "$VERSION_NAME" --model "$MODEL_NAME"
 
 # Finally, download a daisy and so we can test online prediction.
 gsutil cp \
-  gs://cloud-ml-data/img/flower_photos/daisy/100080576_f52e8ee070_n.jpg \
+  gs://cloud-samples-data/ml-engine/flowers/daisy/100080576_f52e8ee070_n.jpg \
   daisy.jpg
 
 # Since the image is passed via JSON, we have to encode the JPEG string first.
@@ -99,4 +99,4 @@ python -c 'import base64, sys, json; img = base64.b64encode(open(sys.argv[1], "r
 # service a moment to catch up--only needed when you deploy a new version.
 # We wait for 10 minutes here, but often see the service start up sooner.
 sleep 10m
-gcloud ml-engine predict --model ${MODEL_NAME} --json-instances request.json
+gcloud ml-engine predict --model ${MODEL_NAME} --json-instances request.json

testing/ubuntu/setup_and_run_test.sh

@@ -17,8 +17,12 @@ pip install --upgrade pip
 # Install test requirements
 pip install --upgrade -r $CMLE_REQUIREMENTS_FILE
 
+export GOOGLE_APPLICATION_CREDENTIALS="${KOKORO_GFILE_DIR}/${CMLE_KEYFILE}"
+
 gcloud auth activate-service-account --key-file "${KOKORO_GFILE_DIR}/${CMLE_KEYFILE}"
 gcloud config set project $CMLE_PROJECT_ID
 gcloud config set compute/region $CMLE_REGION
+gcloud config list
+
 
-bash $CMLE_TEST_SCRIPT
+bash $CMLE_TEST_SCRIPT