[bitnami/tensorflow-resnet] Update TF ResNet version (#8510)

* Update default TF model used for prediction

Signed-off-by: Jose Antonio Carmona <jcarmona@vmware.com>

* Bump chart version and update README

Signed-off-by: Jose Antonio Carmona <jcarmona@vmware.com>

* [bitnami/tensorflow-resnet] Update components versions

Signed-off-by: Bitnami Containers <containers@bitnami.com>

Co-authored-by: Bitnami Containers <containers@bitnami.com>

bitnami/tensorflow-resnet/Chart.yaml

@@ -1,7 +1,7 @@
 annotations:
   category: MachineLearning
 apiVersion: v2
-appVersion: 2.6.1
+appVersion: 2.7.0
 dependencies:
   - name: common
     repository: https://charts.bitnami.com/bitnami
@@ -27,4 +27,4 @@ sources:
   - https://github.com/bitnami/bitnami-docker-tensorflow-serving
   - https://github.com/bitnami/bitnami-docker-tensorflow-resnet
   - https://www.tensorflow.org/serving/
-version: 3.2.22
+version: 3.3.0

bitnami/tensorflow-resnet/README.md

@@ -153,6 +153,12 @@ Find more information about how to deal with common errors related to Bitnami’
 
 ## Upgrading
 
+### To 3.3.0
+
+TensorFlow ResNet's version was updated to `2.7.0`. Although this new version [does not include breaking changes](https://github.com/tensorflow/serving/releases/tag/2.7.0), the client [was updated to work with newer TF Model Garden models](https://github.com/tensorflow/serving/commit/bb1428d53abb53fe938ddf9bb8839d4dfe48d291). Older models may need to adapt their signature [to the newer, common one](https://www.tensorflow.org/hub/common_signatures/images).
+
+As a result, the pretrained model served by this Chart was updated to [Imagenet (ILSVRC-2012-CLS) classification with ResNet 50](https://tfhub.dev/tensorflow/resnet_50/classification/1).
+
 ### To 3.1.0
 
 This version introduces `bitnami/common`, a [library chart](https://helm.sh/docs/topics/library_charts/#helm) as a dependency. More documentation about this new utility could be found [here](https://github.com/bitnami/charts/tree/master/bitnami/common#bitnami-common-library-chart). Please, make sure that you have updated the chart dependencies before executing any upgrade.

bitnami/tensorflow-resnet/templates/deployment.yaml

@@ -49,9 +49,9 @@ spec:
               if [ -f /seed/.initialized ]; then
                 echo "Already initialized. Skipping"
               else
-                curl -o /seed/resnet_v2_fp32_savedmodel_NHWC_jpg.tar.gz http://download.tensorflow.org/models/official/20181001_resnet/savedmodels/resnet_v2_fp32_savedmodel_NHWC_jpg.tar.gz
-                cd /seed/ && tar -xzf resnet_v2_fp32_savedmodel_NHWC_jpg.tar.gz --strip-components=2
-                rm resnet_v2_fp32_savedmodel_NHWC_jpg.tar.gz
+                curl -o /seed/resnet_50_classification_1.tar.gz https://storage.googleapis.com/tfhub-modules/tensorflow/resnet_50/classification/1.tar.gz
+                cd /seed/ && mkdir 1 && tar -xzf resnet_50_classification_1.tar.gz -C 1
+                rm resnet_50_classification_1.tar.gz
                 touch /seed/.initialized
               fi
           volumeMounts:

bitnami/tensorflow-resnet/values.yaml

@@ -40,7 +40,7 @@ server:
   image:
     registry: docker.io
     repository: bitnami/tensorflow-serving
-    tag: 2.7.0-debian-10-r26
+    tag: 2.7.0-debian-10-r53
     ## Specify a imagePullPolicy
     ## Defaults to 'Always' if image tag is 'latest', else set to 'IfNotPresent'
     ## ref: https://kubernetes.io/docs/user-guide/images/#pre-pulling-images
@@ -66,7 +66,7 @@ client:
   image:
     registry: docker.io
     repository: bitnami/tensorflow-resnet
-    tag: 2.6.1-debian-10-r23
+    tag: 2.7.0-debian-10-r0
     ## Specify a imagePullPolicy
     ## Defaults to 'Always' if image tag is 'latest', else set to 'IfNotPresent'
     ## ref: https://kubernetes.io/docs/user-guide/images/#pre-pulling-images