fix: Update pull request to linux-d160-m4xlarge/arm64 for tensorflow

As discussed here https://redhat-internal.slack.com/archives/C07ANR2U56C/p1764958694713539?thread_ts=1764889387.666069&cid=C07ANR2U56C

Author: ysok

.tekton/odh-workbench-jupyter-tensorflow-cuda-py312-pull-request.yaml

@@ -44,7 +44,7 @@ spec:
   - name: build-platforms
     value:
     - linux-extra-fast/amd64
-    - linux-m2xlarge/arm64
+    - linux-d160-m4xlarge/arm64
   - name: image-expires-after
     value: 5d
   - name: enable-slack-failure-notification