@catladder/pipeline 3.6.0 → 3.7.0

package/examples/__snapshots__/cloud-run-llama.test.ts.snap

@@ -0,0 +1,624 @@
+// Jest Snapshot v1, https://goo.gl/fbAQLP
+
+exports[`matches snapshot for cloud-run-llama local pipeline YAML 1`] = `
+"image: path/to/docker/jobs-default:the-version
+stages:
+  - setup
+  - setup dev
+  - setup review
+  - setup stage
+  - setup prod
+  - test
+  - test dev
+  - test review
+  - test stage
+  - test prod
+  - build
+  - build dev
+  - build review
+  - build stage
+  - build prod
+  - deploy
+  - deploy dev
+  - deploy review
+  - deploy stage
+  - deploy prod
+  - verify
+  - verify dev
+  - verify review
+  - verify stage
+  - verify prod
+  - rollback
+  - rollback dev
+  - rollback review
+  - rollback stage
+  - rollback prod
+  - stop
+  - stop dev
+  - stop review
+  - stop stage
+  - stop prod
+  - release
+variables:
+  FF_USE_FASTZIP: 'true'
+  ARTIFACT_COMPRESSION_LEVEL: fast
+  CACHE_COMPRESSION_LEVEL: fast
+  TRANSFER_METER_FREQUENCY: 5s
+  GIT_DEPTH: '1'
+before_script:
+  - |-
+    function escapeForDotEnv () {
+    input="\${1:-$(cat)}"
+     input="\${input//$'\\n'/\\\\n}"
+      if [[ "$input" == *\\\\n* ]]; then
+        if [[ "$input" == *\\"* && "$input" == *\\'* && "$input" == *\\\`* ]]; then
+          printf "\\"%s\\"\\n" "$input" 
+        elif [[ "$input" == *\\"* && "$input" == *\\'* ]]; then
+          printf "\`%s\`\\n" "$input"
+        elif [[ "$input" == *\\"* ]]; then
+          printf "'%s'\\n" "$input"
+        else
+          printf "\\"%s\\"\\n" "$input"
+        fi
+      else
+        printf "%s\\n" "$input"
+      fi
+    }
+  - |-
+    function collapseable_section_start () {
+    local section_title="\${1}"
+      local section_description="\${2:-$section_title}"
+      echo -e "section_start:\`date +%s\`:\${section_title}[collapsed=true]\\r\\e[0K\${section_description}"
+    }
+  - |-
+    function collapseable_section_end () {
+    local section_title="\${1}"
+      echo -e "section_end:\`date +%s\`:\${section_title}\\r\\e[0K"
+    }
+'llm 🚀 Deploy | dev ':
+  stage: deploy dev
+  image: path/to/docker/gcloud:the-version
+  variables:
+    KUBERNETES_CPU_REQUEST: '0.22'
+    KUBERNETES_MEMORY_REQUEST: 200Mi
+    KUBERNETES_MEMORY_LIMIT: 400Mi
+  script:
+    - collapseable_section_start "injectvars" "Injecting variables"
+    - export ENV_SHORT="dev"
+    - export APP_DIR="llm"
+    - export ENV_TYPE="dev"
+    - export BUILD_INFO_BUILD_ID="$(git describe --tags 2>/dev/null || git rev-parse HEAD)"
+    - export BUILD_INFO_BUILD_TIME="$CI_JOB_STARTED_AT"
+    - export BUILD_INFO_CURRENT_VERSION="$(tag=$(git ls-remote origin "refs/tags/v*[0-9]" 2>/dev/null | cut -f 2- | sort -V | tail -1 | sed 's/refs\\/tags\\/v//'); [ -z "$tag" ] && echo "0.0.0" || echo "$tag")"
+    - export HOSTNAME="$(printf %s "pan-my-ai-app-dev-llm-$CL_dev_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')"
+    - export ROOT_URL="https://$(printf %s "pan-my-ai-app-dev-llm-$CL_dev_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')"
+    - export HOSTNAME_INTERNAL="$(printf %s "pan-my-ai-app-dev-llm-$CL_dev_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')"
+    - export ROOT_URL_INTERNAL="https://$(printf %s "pan-my-ai-app-dev-llm-$CL_dev_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')"
+    - export DEPLOY_CLOUD_RUN_PROJECT_ID="google-project-id"
+    - export DEPLOY_CLOUD_RUN_REGION="europe-west4"
+    - export HF_TOKEN="$CL_dev_llm_HF_TOKEN"
+    - export GCLOUD_DEPLOY_credentialsKey="$CL_dev_llm_GCLOUD_DEPLOY_credentialsKey"
+    - export GCLOUD_RUN_canonicalHostSuffix="$CL_dev_llm_GCLOUD_RUN_canonicalHostSuffix"
+    - export _ALL_ENV_VAR_KEYS="[\\"ENV_SHORT\\",\\"APP_DIR\\",\\"ENV_TYPE\\",\\"BUILD_INFO_BUILD_ID\\",\\"BUILD_INFO_BUILD_TIME\\",\\"BUILD_INFO_CURRENT_VERSION\\",\\"HOSTNAME\\",\\"ROOT_URL\\",\\"HOSTNAME_INTERNAL\\",\\"ROOT_URL_INTERNAL\\",\\"DEPLOY_CLOUD_RUN_PROJECT_ID\\",\\"DEPLOY_CLOUD_RUN_REGION\\",\\"HF_TOKEN\\",\\"GCLOUD_DEPLOY_credentialsKey\\",\\"GCLOUD_RUN_canonicalHostSuffix\\"]"
+    - export DOCKER_REGISTRY="europe-west4-docker.pkg.dev"
+    - export DOCKER_IMAGE="europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/dev/llm"
+    - export DOCKER_CACHE_IMAGE="europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm"
+    - export DOCKER_IMAGE_TAG="$CI_COMMIT_SHA"
+    - export CLOUDSDK_CORE_DISABLE_PROMPTS="1"
+    - collapseable_section_end "injectvars"
+    - collapseable_section_start "prepare" "Prepare..."
+    - gcloud auth activate-service-account --key-file=<(echo "$CL_dev_llm_GCLOUD_DEPLOY_credentialsKey")
+    - export GCLOUD_PROJECT_NUMBER=$(gcloud projects describe google-project-id --format="value(projectNumber)")
+    - 'echo "GCLOUD_PROJECT_NUMBER: $GCLOUD_PROJECT_NUMBER"'
+    - collapseable_section_end "prepare"
+    - collapseable_section_start "writeenvvars" "Write env vars to file"
+    - |
+      cat > ____envvars.yaml <<EOF
+      ENV_SHORT: |-
+        dev
+      APP_DIR: |-
+        llm
+      ENV_TYPE: |-
+        dev
+      BUILD_INFO_BUILD_ID: |-
+        $(printf %s "$(git describe --tags 2>/dev/null || git rev-parse HEAD)" | sed '1!s/^/  /')
+      BUILD_INFO_BUILD_TIME: |-
+        $(printf %s "$CI_JOB_STARTED_AT" | sed '1!s/^/  /')
+      BUILD_INFO_CURRENT_VERSION: |-
+        $(printf %s "$(tag=$(git ls-remote origin "refs/tags/v*[0-9]" 2>/dev/null | cut -f 2- | sort -V | tail -1 | sed 's/refs\\/tags\\/v//'); [ -z "$tag" ] && echo "0.0.0" || echo "$tag")" | sed '1!s/^/  /')
+      HOSTNAME: |-
+        $(printf %s "$(printf %s "pan-my-ai-app-dev-llm-$CL_dev_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')" | sed '1!s/^/  /')
+      ROOT_URL: |-
+        $(printf %s "https://$(printf %s "pan-my-ai-app-dev-llm-$CL_dev_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')" | sed '1!s/^/  /')
+      HOSTNAME_INTERNAL: |-
+        $(printf %s "$(printf %s "pan-my-ai-app-dev-llm-$CL_dev_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')" | sed '1!s/^/  /')
+      ROOT_URL_INTERNAL: |-
+        $(printf %s "https://$(printf %s "pan-my-ai-app-dev-llm-$CL_dev_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')" | sed '1!s/^/  /')
+      DEPLOY_CLOUD_RUN_PROJECT_ID: |-
+        google-project-id
+      DEPLOY_CLOUD_RUN_REGION: |-
+        europe-west4
+      HF_TOKEN: |-
+        $(printf %s "$CL_dev_llm_HF_TOKEN" | sed '1!s/^/  /')
+      GCLOUD_RUN_canonicalHostSuffix: |-
+        $(printf %s "$CL_dev_llm_GCLOUD_RUN_canonicalHostSuffix" | sed '1!s/^/  /')
+      _ALL_ENV_VAR_KEYS: |-
+        ["ENV_SHORT","APP_DIR","ENV_TYPE","BUILD_INFO_BUILD_ID","BUILD_INFO_BUILD_TIME","BUILD_INFO_CURRENT_VERSION","HOSTNAME","ROOT_URL","HOSTNAME_INTERNAL","ROOT_URL_INTERNAL","DEPLOY_CLOUD_RUN_PROJECT_ID","DEPLOY_CLOUD_RUN_REGION","HF_TOKEN","GCLOUD_DEPLOY_credentialsKey","GCLOUD_RUN_canonicalHostSuffix"]
+
+      EOF
+    - collapseable_section_end "writeenvvars"
+    - collapseable_section_start "deploy" "Deploy to cloud run"
+    - gcloud beta run deploy pan-my-ai-app-dev-llm --command="" --args="--model-id=meta-llama/Llama-3.2-1B-Instruct","--max-concurrent-requests=1" --image=us-docker.pkg.dev/deeplearning-platform-release/gcr.io/huggingface-text-generation-inference-cu124.2-3.ubuntu2204.py311 --project=google-project-id --region=europe-west4 --labels=customer-name=pan,component-name=llm,app-name=my-ai-app,env-type=dev,env-name=dev,build-type=disabled,cloud-run-service-name=pan-my-ai-app-dev-llm --env-vars-file=____envvars.yaml --min-instances=0 --max-instances=100 --no-cpu-throttling --cpu=8 --memory=32Gi --allow-unauthenticated --ingress=all --cpu-boost --gpu=1 --gpu-type=nvidia-l4
+    - collapseable_section_end "deploy"
+    - collapseable_section_start "cleanup" "Cleanup"
+    - gcloud run revisions list --project=google-project-id --region=europe-west4 --service=pan-my-ai-app-dev-llm --limit=unlimited --sort-by=metadata.creationTimestamp --format="value(name)" --filter='(status.conditions.status=False OR status.conditions.status=Unknown)' | while read -r revisionname; do gcloud run revisions delete --project=google-project-id --region=europe-west4 --quiet $revisionname ; done
+    - gcloud artifacts docker images list europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/dev/llm --sort-by=~CREATE_TIME --format="value(version)" | tail -n +2 | while read -r version; do gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/dev/llm@$version --quiet --delete-tags; done
+    - gcloud artifacts docker images list europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm --sort-by=~CREATE_TIME --format="value(version)" | tail -n +2 | while read -r version; do gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm@$version --quiet --delete-tags; done
+    - collapseable_section_end "cleanup"
+    - echo "CL_GITLAB_ENVIRONMENT_URL=$ROOT_URL" >> gitlab_environment.env
+  environment:
+    name: dev/llm
+    url: $CL_GITLAB_ENVIRONMENT_URL
+    on_stop: 'llm 🛑 Stop ⚠️ | dev '
+    auto_stop_in: 4 weeks
+  artifacts:
+    reports:
+      dotenv: gitlab_environment.env
+  rules:
+    - when: on_success
+      if: $CI_COMMIT_BRANCH == $CI_DEFAULT_BRANCH && $CI_COMMIT_MESSAGE !~ /^chore\\(release\\).*/
+  needs: []
+  retry: &a1
+    max: 2
+    when:
+      - runner_system_failure
+      - stuck_or_timeout_failure
+  interruptible: true
+  allow_failure: false
+'llm 🛑 Stop ⚠️ | dev ':
+  stage: stop dev
+  image: path/to/docker/gcloud:the-version
+  variables:
+    KUBERNETES_CPU_REQUEST: '0.22'
+    KUBERNETES_MEMORY_REQUEST: 200Mi
+    KUBERNETES_MEMORY_LIMIT: 400Mi
+    GIT_STRATEGY: none
+  script:
+    - collapseable_section_start "injectvars" "Injecting variables"
+    - export CLOUDSDK_CORE_DISABLE_PROMPTS="1"
+    - collapseable_section_end "injectvars"
+    - set +e
+    - gcloud auth activate-service-account --key-file=<(echo "$CL_dev_llm_GCLOUD_DEPLOY_credentialsKey")
+    - gcloud run services delete pan-my-ai-app-dev-llm --project=google-project-id --region=europe-west4
+    - gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/dev/llm --quiet --delete-tags
+    - gcloud artifacts docker images list europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm --sort-by=~CREATE_TIME --format="value(version)" | tail -n +2 | while read -r version; do gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm@$version --quiet --delete-tags; done
+    - set -e
+  environment:
+    name: dev/llm
+    action: stop
+  rules:
+    - if: $CI_COMMIT_BRANCH =~ /^[0-9]+\\.([0-9]+|x)\\.x$/
+      when: on_success
+    - when: manual
+      if: $CI_COMMIT_BRANCH == $CI_DEFAULT_BRANCH && $CI_COMMIT_MESSAGE !~ /^chore\\(release\\).*/
+  needs: []
+  retry: *a1
+  interruptible: true
+  allow_failure: true
+'llm 🚀 Deploy | review ':
+  stage: deploy review
+  image: path/to/docker/gcloud:the-version
+  variables:
+    KUBERNETES_CPU_REQUEST: '0.22'
+    KUBERNETES_MEMORY_REQUEST: 200Mi
+    KUBERNETES_MEMORY_LIMIT: 400Mi
+  script:
+    - collapseable_section_start "injectvars" "Injecting variables"
+    - export ENV_SHORT="review"
+    - export APP_DIR="llm"
+    - export ENV_TYPE="review"
+    - export BUILD_INFO_BUILD_ID="$(git describe --tags 2>/dev/null || git rev-parse HEAD)"
+    - export BUILD_INFO_BUILD_TIME="$CI_JOB_STARTED_AT"
+    - export BUILD_INFO_CURRENT_VERSION="$(tag=$(git ls-remote origin "refs/tags/v*[0-9]" 2>/dev/null | cut -f 2- | sort -V | tail -1 | sed 's/refs\\/tags\\/v//'); [ -z "$tag" ] && echo "0.0.0" || echo "$tag")"
+    - export HOSTNAME="$(printf %s "pan-my-ai-app-review-$([ -n "$CI_MERGE_REQUEST_IID" ] && echo "mr$CI_MERGE_REQUEST_IID" || { [ -n "$CI_COMMIT_REF_SLUG" ] && echo "$CI_COMMIT_REF_SLUG" || echo "unknown"; })-llm-$CL_review_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')"
+    - export ROOT_URL="https://$(printf %s "pan-my-ai-app-review-$([ -n "$CI_MERGE_REQUEST_IID" ] && echo "mr$CI_MERGE_REQUEST_IID" || { [ -n "$CI_COMMIT_REF_SLUG" ] && echo "$CI_COMMIT_REF_SLUG" || echo "unknown"; })-llm-$CL_review_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')"
+    - export HOSTNAME_INTERNAL="$(printf %s "pan-my-ai-app-review-$([ -n "$CI_MERGE_REQUEST_IID" ] && echo "mr$CI_MERGE_REQUEST_IID" || { [ -n "$CI_COMMIT_REF_SLUG" ] && echo "$CI_COMMIT_REF_SLUG" || echo "unknown"; })-llm-$CL_review_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')"
+    - export ROOT_URL_INTERNAL="https://$(printf %s "pan-my-ai-app-review-$([ -n "$CI_MERGE_REQUEST_IID" ] && echo "mr$CI_MERGE_REQUEST_IID" || { [ -n "$CI_COMMIT_REF_SLUG" ] && echo "$CI_COMMIT_REF_SLUG" || echo "unknown"; })-llm-$CL_review_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')"
+    - export DEPLOY_CLOUD_RUN_PROJECT_ID="google-project-id"
+    - export DEPLOY_CLOUD_RUN_REGION="europe-west4"
+    - export HF_TOKEN="$CL_review_llm_HF_TOKEN"
+    - export GCLOUD_DEPLOY_credentialsKey="$CL_review_llm_GCLOUD_DEPLOY_credentialsKey"
+    - export GCLOUD_RUN_canonicalHostSuffix="$CL_review_llm_GCLOUD_RUN_canonicalHostSuffix"
+    - export _ALL_ENV_VAR_KEYS="[\\"ENV_SHORT\\",\\"APP_DIR\\",\\"ENV_TYPE\\",\\"BUILD_INFO_BUILD_ID\\",\\"BUILD_INFO_BUILD_TIME\\",\\"BUILD_INFO_CURRENT_VERSION\\",\\"HOSTNAME\\",\\"ROOT_URL\\",\\"HOSTNAME_INTERNAL\\",\\"ROOT_URL_INTERNAL\\",\\"DEPLOY_CLOUD_RUN_PROJECT_ID\\",\\"DEPLOY_CLOUD_RUN_REGION\\",\\"HF_TOKEN\\",\\"GCLOUD_DEPLOY_credentialsKey\\",\\"GCLOUD_RUN_canonicalHostSuffix\\"]"
+    - export DOCKER_REGISTRY="europe-west4-docker.pkg.dev"
+    - export DOCKER_IMAGE="europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/review/llm/$([ -n "$CI_MERGE_REQUEST_IID" ] && echo "mr$CI_MERGE_REQUEST_IID" || { [ -n "$CI_COMMIT_REF_SLUG" ] && echo "$CI_COMMIT_REF_SLUG" || echo "unknown"; })"
+    - export DOCKER_CACHE_IMAGE="europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm"
+    - export DOCKER_IMAGE_TAG="$CI_COMMIT_SHA"
+    - export CLOUDSDK_CORE_DISABLE_PROMPTS="1"
+    - collapseable_section_end "injectvars"
+    - collapseable_section_start "prepare" "Prepare..."
+    - gcloud auth activate-service-account --key-file=<(echo "$CL_review_llm_GCLOUD_DEPLOY_credentialsKey")
+    - export GCLOUD_PROJECT_NUMBER=$(gcloud projects describe google-project-id --format="value(projectNumber)")
+    - 'echo "GCLOUD_PROJECT_NUMBER: $GCLOUD_PROJECT_NUMBER"'
+    - collapseable_section_end "prepare"
+    - collapseable_section_start "writeenvvars" "Write env vars to file"
+    - |
+      cat > ____envvars.yaml <<EOF
+      ENV_SHORT: |-
+        review
+      APP_DIR: |-
+        llm
+      ENV_TYPE: |-
+        review
+      BUILD_INFO_BUILD_ID: |-
+        $(printf %s "$(git describe --tags 2>/dev/null || git rev-parse HEAD)" | sed '1!s/^/  /')
+      BUILD_INFO_BUILD_TIME: |-
+        $(printf %s "$CI_JOB_STARTED_AT" | sed '1!s/^/  /')
+      BUILD_INFO_CURRENT_VERSION: |-
+        $(printf %s "$(tag=$(git ls-remote origin "refs/tags/v*[0-9]" 2>/dev/null | cut -f 2- | sort -V | tail -1 | sed 's/refs\\/tags\\/v//'); [ -z "$tag" ] && echo "0.0.0" || echo "$tag")" | sed '1!s/^/  /')
+      HOSTNAME: |-
+        $(printf %s "$(printf %s "pan-my-ai-app-review-$([ -n "$CI_MERGE_REQUEST_IID" ] && echo "mr$CI_MERGE_REQUEST_IID" || { [ -n "$CI_COMMIT_REF_SLUG" ] && echo "$CI_COMMIT_REF_SLUG" || echo "unknown"; })-llm-$CL_review_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')" | sed '1!s/^/  /')
+      ROOT_URL: |-
+        $(printf %s "https://$(printf %s "pan-my-ai-app-review-$([ -n "$CI_MERGE_REQUEST_IID" ] && echo "mr$CI_MERGE_REQUEST_IID" || { [ -n "$CI_COMMIT_REF_SLUG" ] && echo "$CI_COMMIT_REF_SLUG" || echo "unknown"; })-llm-$CL_review_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')" | sed '1!s/^/  /')
+      HOSTNAME_INTERNAL: |-
+        $(printf %s "$(printf %s "pan-my-ai-app-review-$([ -n "$CI_MERGE_REQUEST_IID" ] && echo "mr$CI_MERGE_REQUEST_IID" || { [ -n "$CI_COMMIT_REF_SLUG" ] && echo "$CI_COMMIT_REF_SLUG" || echo "unknown"; })-llm-$CL_review_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')" | sed '1!s/^/  /')
+      ROOT_URL_INTERNAL: |-
+        $(printf %s "https://$(printf %s "pan-my-ai-app-review-$([ -n "$CI_MERGE_REQUEST_IID" ] && echo "mr$CI_MERGE_REQUEST_IID" || { [ -n "$CI_COMMIT_REF_SLUG" ] && echo "$CI_COMMIT_REF_SLUG" || echo "unknown"; })-llm-$CL_review_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')" | sed '1!s/^/  /')
+      DEPLOY_CLOUD_RUN_PROJECT_ID: |-
+        google-project-id
+      DEPLOY_CLOUD_RUN_REGION: |-
+        europe-west4
+      HF_TOKEN: |-
+        $(printf %s "$CL_review_llm_HF_TOKEN" | sed '1!s/^/  /')
+      GCLOUD_RUN_canonicalHostSuffix: |-
+        $(printf %s "$CL_review_llm_GCLOUD_RUN_canonicalHostSuffix" | sed '1!s/^/  /')
+      _ALL_ENV_VAR_KEYS: |-
+        ["ENV_SHORT","APP_DIR","ENV_TYPE","BUILD_INFO_BUILD_ID","BUILD_INFO_BUILD_TIME","BUILD_INFO_CURRENT_VERSION","HOSTNAME","ROOT_URL","HOSTNAME_INTERNAL","ROOT_URL_INTERNAL","DEPLOY_CLOUD_RUN_PROJECT_ID","DEPLOY_CLOUD_RUN_REGION","HF_TOKEN","GCLOUD_DEPLOY_credentialsKey","GCLOUD_RUN_canonicalHostSuffix"]
+
+      EOF
+    - collapseable_section_end "writeenvvars"
+    - collapseable_section_start "deploy" "Deploy to cloud run"
+    - gcloud beta run deploy $(printf %s "pan-my-ai-app-review-$([ -n "$CI_MERGE_REQUEST_IID" ] && echo "mr$CI_MERGE_REQUEST_IID" || { [ -n "$CI_COMMIT_REF_SLUG" ] && echo "$CI_COMMIT_REF_SLUG" || echo "unknown"; })-llm" | awk '{print tolower($0)}') --command="" --args="--model-id=meta-llama/Llama-3.2-1B-Instruct","--max-concurrent-requests=1" --image=us-docker.pkg.dev/deeplearning-platform-release/gcr.io/huggingface-text-generation-inference-cu124.2-3.ubuntu2204.py311 --project=google-project-id --region=europe-west4 --labels=customer-name=pan,component-name=llm,app-name=my-ai-app,env-type=review,env-name=review,build-type=disabled,cloud-run-service-name=$(printf %s "pan-my-ai-app-review-$([ -n "$CI_MERGE_REQUEST_IID" ] && echo "mr$CI_MERGE_REQUEST_IID" || { [ -n "$CI_COMMIT_REF_SLUG" ] && echo "$CI_COMMIT_REF_SLUG" || echo "unknown"; })-llm" | awk '{print tolower($0)}') --env-vars-file=____envvars.yaml --min-instances=0 --max-instances=100 --no-cpu-throttling --cpu=8 --memory=32Gi --allow-unauthenticated --ingress=all --cpu-boost --gpu=1 --gpu-type=nvidia-l4
+    - collapseable_section_end "deploy"
+    - collapseable_section_start "cleanup" "Cleanup"
+    - gcloud run revisions list --project=google-project-id --region=europe-west4 --service=$(printf %s "pan-my-ai-app-review-$([ -n "$CI_MERGE_REQUEST_IID" ] && echo "mr$CI_MERGE_REQUEST_IID" || { [ -n "$CI_COMMIT_REF_SLUG" ] && echo "$CI_COMMIT_REF_SLUG" || echo "unknown"; })-llm" | awk '{print tolower($0)}') --limit=unlimited --sort-by=metadata.creationTimestamp --format="value(name)" --filter='(status.conditions.status=False OR status.conditions.status=Unknown)' | while read -r revisionname; do gcloud run revisions delete --project=google-project-id --region=europe-west4 --quiet $revisionname ; done
+    - gcloud artifacts docker images list europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/review/llm/$([ -n "$CI_MERGE_REQUEST_IID" ] && echo "mr$CI_MERGE_REQUEST_IID" || { [ -n "$CI_COMMIT_REF_SLUG" ] && echo "$CI_COMMIT_REF_SLUG" || echo "unknown"; }) --sort-by=~CREATE_TIME --format="value(version)" | tail -n +2 | while read -r version; do gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/review/llm/$([ -n "$CI_MERGE_REQUEST_IID" ] && echo "mr$CI_MERGE_REQUEST_IID" || { [ -n "$CI_COMMIT_REF_SLUG" ] && echo "$CI_COMMIT_REF_SLUG" || echo "unknown"; })@$version --quiet --delete-tags; done
+    - gcloud artifacts docker images list europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm --sort-by=~CREATE_TIME --format="value(version)" | tail -n +2 | while read -r version; do gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm@$version --quiet --delete-tags; done
+    - set +e
+    - gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/review/llm --quiet --delete-tags
+    - set -e
+    - collapseable_section_end "cleanup"
+    - echo "CL_GITLAB_ENVIRONMENT_URL=$ROOT_URL" >> gitlab_environment.env
+  environment:
+    name: review/$CI_COMMIT_REF_NAME/llm
+    url: $CL_GITLAB_ENVIRONMENT_URL
+    on_stop: 'llm 🛑 Stop ⚠️ | review '
+    auto_stop_in: 1 week
+  artifacts:
+    reports:
+      dotenv: gitlab_environment.env
+  rules:
+    - when: on_success
+      if: $CI_MERGE_REQUEST_ID
+  needs: []
+  retry: *a1
+  interruptible: true
+  allow_failure: false
+'llm 🛑 Stop ⚠️ | review ':
+  stage: stop review
+  image: path/to/docker/gcloud:the-version
+  variables:
+    KUBERNETES_CPU_REQUEST: '0.22'
+    KUBERNETES_MEMORY_REQUEST: 200Mi
+    KUBERNETES_MEMORY_LIMIT: 400Mi
+    GIT_STRATEGY: none
+  script:
+    - collapseable_section_start "injectvars" "Injecting variables"
+    - export CLOUDSDK_CORE_DISABLE_PROMPTS="1"
+    - collapseable_section_end "injectvars"
+    - set +e
+    - gcloud auth activate-service-account --key-file=<(echo "$CL_review_llm_GCLOUD_DEPLOY_credentialsKey")
+    - gcloud run services delete $(printf %s "pan-my-ai-app-review-$([ -n "$CI_MERGE_REQUEST_IID" ] && echo "mr$CI_MERGE_REQUEST_IID" || { [ -n "$CI_COMMIT_REF_SLUG" ] && echo "$CI_COMMIT_REF_SLUG" || echo "unknown"; })-llm" | awk '{print tolower($0)}') --project=google-project-id --region=europe-west4
+    - gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/review/llm/$([ -n "$CI_MERGE_REQUEST_IID" ] && echo "mr$CI_MERGE_REQUEST_IID" || { [ -n "$CI_COMMIT_REF_SLUG" ] && echo "$CI_COMMIT_REF_SLUG" || echo "unknown"; }) --quiet --delete-tags
+    - gcloud artifacts docker images list europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm --sort-by=~CREATE_TIME --format="value(version)" | tail -n +2 | while read -r version; do gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm@$version --quiet --delete-tags; done
+    - set +e
+    - gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/review/llm --quiet --delete-tags
+    - set -e
+    - set -e
+  environment:
+    name: review/$CI_COMMIT_REF_NAME/llm
+    action: stop
+  rules:
+    - if: $CI_COMMIT_BRANCH =~ /^[0-9]+\\.([0-9]+|x)\\.x$/
+      when: on_success
+    - when: manual
+      if: $CI_MERGE_REQUEST_ID
+  needs: []
+  retry: *a1
+  interruptible: true
+  allow_failure: true
+'llm 🚀 Deploy | stage ':
+  stage: deploy stage
+  image: path/to/docker/gcloud:the-version
+  variables:
+    KUBERNETES_CPU_REQUEST: '0.22'
+    KUBERNETES_MEMORY_REQUEST: 200Mi
+    KUBERNETES_MEMORY_LIMIT: 400Mi
+  script:
+    - collapseable_section_start "injectvars" "Injecting variables"
+    - export ENV_SHORT="stage"
+    - export APP_DIR="llm"
+    - export ENV_TYPE="stage"
+    - export BUILD_INFO_BUILD_ID="$(git describe --tags 2>/dev/null || git rev-parse HEAD)"
+    - export BUILD_INFO_BUILD_TIME="$CI_JOB_STARTED_AT"
+    - export BUILD_INFO_CURRENT_VERSION="$(tag=$(git ls-remote origin "refs/tags/v*[0-9]" 2>/dev/null | cut -f 2- | sort -V | tail -1 | sed 's/refs\\/tags\\/v//'); [ -z "$tag" ] && echo "0.0.0" || echo "$tag")"
+    - export HOSTNAME="$(printf %s "pan-my-ai-app-stage-llm-$CL_stage_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')"
+    - export ROOT_URL="https://$(printf %s "pan-my-ai-app-stage-llm-$CL_stage_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')"
+    - export HOSTNAME_INTERNAL="$(printf %s "pan-my-ai-app-stage-llm-$CL_stage_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')"
+    - export ROOT_URL_INTERNAL="https://$(printf %s "pan-my-ai-app-stage-llm-$CL_stage_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')"
+    - export DEPLOY_CLOUD_RUN_PROJECT_ID="google-project-id"
+    - export DEPLOY_CLOUD_RUN_REGION="europe-west4"
+    - export HF_TOKEN="$CL_stage_llm_HF_TOKEN"
+    - export GCLOUD_DEPLOY_credentialsKey="$CL_stage_llm_GCLOUD_DEPLOY_credentialsKey"
+    - export GCLOUD_RUN_canonicalHostSuffix="$CL_stage_llm_GCLOUD_RUN_canonicalHostSuffix"
+    - export _ALL_ENV_VAR_KEYS="[\\"ENV_SHORT\\",\\"APP_DIR\\",\\"ENV_TYPE\\",\\"BUILD_INFO_BUILD_ID\\",\\"BUILD_INFO_BUILD_TIME\\",\\"BUILD_INFO_CURRENT_VERSION\\",\\"HOSTNAME\\",\\"ROOT_URL\\",\\"HOSTNAME_INTERNAL\\",\\"ROOT_URL_INTERNAL\\",\\"DEPLOY_CLOUD_RUN_PROJECT_ID\\",\\"DEPLOY_CLOUD_RUN_REGION\\",\\"HF_TOKEN\\",\\"GCLOUD_DEPLOY_credentialsKey\\",\\"GCLOUD_RUN_canonicalHostSuffix\\"]"
+    - export DOCKER_REGISTRY="europe-west4-docker.pkg.dev"
+    - export DOCKER_IMAGE="europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/stage/llm"
+    - export DOCKER_CACHE_IMAGE="europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm"
+    - export DOCKER_IMAGE_TAG="$CI_COMMIT_SHA"
+    - export CLOUDSDK_CORE_DISABLE_PROMPTS="1"
+    - collapseable_section_end "injectvars"
+    - collapseable_section_start "prepare" "Prepare..."
+    - gcloud auth activate-service-account --key-file=<(echo "$CL_stage_llm_GCLOUD_DEPLOY_credentialsKey")
+    - export GCLOUD_PROJECT_NUMBER=$(gcloud projects describe google-project-id --format="value(projectNumber)")
+    - 'echo "GCLOUD_PROJECT_NUMBER: $GCLOUD_PROJECT_NUMBER"'
+    - collapseable_section_end "prepare"
+    - collapseable_section_start "writeenvvars" "Write env vars to file"
+    - |
+      cat > ____envvars.yaml <<EOF
+      ENV_SHORT: |-
+        stage
+      APP_DIR: |-
+        llm
+      ENV_TYPE: |-
+        stage
+      BUILD_INFO_BUILD_ID: |-
+        $(printf %s "$(git describe --tags 2>/dev/null || git rev-parse HEAD)" | sed '1!s/^/  /')
+      BUILD_INFO_BUILD_TIME: |-
+        $(printf %s "$CI_JOB_STARTED_AT" | sed '1!s/^/  /')
+      BUILD_INFO_CURRENT_VERSION: |-
+        $(printf %s "$(tag=$(git ls-remote origin "refs/tags/v*[0-9]" 2>/dev/null | cut -f 2- | sort -V | tail -1 | sed 's/refs\\/tags\\/v//'); [ -z "$tag" ] && echo "0.0.0" || echo "$tag")" | sed '1!s/^/  /')
+      HOSTNAME: |-
+        $(printf %s "$(printf %s "pan-my-ai-app-stage-llm-$CL_stage_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')" | sed '1!s/^/  /')
+      ROOT_URL: |-
+        $(printf %s "https://$(printf %s "pan-my-ai-app-stage-llm-$CL_stage_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')" | sed '1!s/^/  /')
+      HOSTNAME_INTERNAL: |-
+        $(printf %s "$(printf %s "pan-my-ai-app-stage-llm-$CL_stage_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')" | sed '1!s/^/  /')
+      ROOT_URL_INTERNAL: |-
+        $(printf %s "https://$(printf %s "pan-my-ai-app-stage-llm-$CL_stage_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')" | sed '1!s/^/  /')
+      DEPLOY_CLOUD_RUN_PROJECT_ID: |-
+        google-project-id
+      DEPLOY_CLOUD_RUN_REGION: |-
+        europe-west4
+      HF_TOKEN: |-
+        $(printf %s "$CL_stage_llm_HF_TOKEN" | sed '1!s/^/  /')
+      GCLOUD_RUN_canonicalHostSuffix: |-
+        $(printf %s "$CL_stage_llm_GCLOUD_RUN_canonicalHostSuffix" | sed '1!s/^/  /')
+      _ALL_ENV_VAR_KEYS: |-
+        ["ENV_SHORT","APP_DIR","ENV_TYPE","BUILD_INFO_BUILD_ID","BUILD_INFO_BUILD_TIME","BUILD_INFO_CURRENT_VERSION","HOSTNAME","ROOT_URL","HOSTNAME_INTERNAL","ROOT_URL_INTERNAL","DEPLOY_CLOUD_RUN_PROJECT_ID","DEPLOY_CLOUD_RUN_REGION","HF_TOKEN","GCLOUD_DEPLOY_credentialsKey","GCLOUD_RUN_canonicalHostSuffix"]
+
+      EOF
+    - collapseable_section_end "writeenvvars"
+    - collapseable_section_start "deploy" "Deploy to cloud run"
+    - gcloud beta run deploy pan-my-ai-app-stage-llm --command="" --args="--model-id=meta-llama/Llama-3.2-1B-Instruct","--max-concurrent-requests=1" --image=us-docker.pkg.dev/deeplearning-platform-release/gcr.io/huggingface-text-generation-inference-cu124.2-3.ubuntu2204.py311 --project=google-project-id --region=europe-west4 --labels=customer-name=pan,component-name=llm,app-name=my-ai-app,env-type=stage,env-name=stage,build-type=disabled,cloud-run-service-name=pan-my-ai-app-stage-llm --env-vars-file=____envvars.yaml --min-instances=0 --max-instances=100 --no-cpu-throttling --cpu=8 --memory=32Gi --allow-unauthenticated --ingress=all --cpu-boost --gpu=1 --gpu-type=nvidia-l4
+    - collapseable_section_end "deploy"
+    - collapseable_section_start "cleanup" "Cleanup"
+    - gcloud run revisions list --project=google-project-id --region=europe-west4 --service=pan-my-ai-app-stage-llm --limit=unlimited --sort-by=metadata.creationTimestamp --format="value(name)" --filter='(status.conditions.status=False OR status.conditions.status=Unknown)' | while read -r revisionname; do gcloud run revisions delete --project=google-project-id --region=europe-west4 --quiet $revisionname ; done
+    - gcloud artifacts docker images list europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/stage/llm --sort-by=~CREATE_TIME --format="value(version)" | tail -n +2 | while read -r version; do gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/stage/llm@$version --quiet --delete-tags; done
+    - gcloud artifacts docker images list europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm --sort-by=~CREATE_TIME --format="value(version)" | tail -n +2 | while read -r version; do gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm@$version --quiet --delete-tags; done
+    - collapseable_section_end "cleanup"
+    - echo "CL_GITLAB_ENVIRONMENT_URL=$ROOT_URL" >> gitlab_environment.env
+  environment:
+    name: stage/llm
+    url: $CL_GITLAB_ENVIRONMENT_URL
+    on_stop: 'llm 🛑 Stop ⚠️ | stage '
+  artifacts:
+    reports:
+      dotenv: gitlab_environment.env
+  rules:
+    - when: on_success
+      if: $CI_COMMIT_TAG
+  needs: []
+  retry: *a1
+  interruptible: true
+  allow_failure: false
+'llm 🛑 Stop ⚠️ | stage ':
+  stage: stop stage
+  image: path/to/docker/gcloud:the-version
+  variables:
+    KUBERNETES_CPU_REQUEST: '0.22'
+    KUBERNETES_MEMORY_REQUEST: 200Mi
+    KUBERNETES_MEMORY_LIMIT: 400Mi
+    GIT_STRATEGY: none
+  script:
+    - collapseable_section_start "injectvars" "Injecting variables"
+    - export CLOUDSDK_CORE_DISABLE_PROMPTS="1"
+    - collapseable_section_end "injectvars"
+    - set +e
+    - gcloud auth activate-service-account --key-file=<(echo "$CL_stage_llm_GCLOUD_DEPLOY_credentialsKey")
+    - gcloud run services delete pan-my-ai-app-stage-llm --project=google-project-id --region=europe-west4
+    - gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/stage/llm --quiet --delete-tags
+    - gcloud artifacts docker images list europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm --sort-by=~CREATE_TIME --format="value(version)" | tail -n +2 | while read -r version; do gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm@$version --quiet --delete-tags; done
+    - set -e
+  environment:
+    name: stage/llm
+    action: stop
+  rules:
+    - if: $CI_COMMIT_BRANCH =~ /^[0-9]+\\.([0-9]+|x)\\.x$/
+      when: on_success
+    - when: manual
+      if: $CI_COMMIT_TAG
+  needs: []
+  retry: *a1
+  interruptible: true
+  allow_failure: true
+'llm 🚀 Deploy | prod ':
+  stage: deploy prod
+  image: path/to/docker/gcloud:the-version
+  variables:
+    KUBERNETES_CPU_REQUEST: '0.22'
+    KUBERNETES_MEMORY_REQUEST: 200Mi
+    KUBERNETES_MEMORY_LIMIT: 400Mi
+  script:
+    - collapseable_section_start "injectvars" "Injecting variables"
+    - export ENV_SHORT="prod"
+    - export APP_DIR="llm"
+    - export ENV_TYPE="prod"
+    - export BUILD_INFO_BUILD_ID="$(git describe --tags 2>/dev/null || git rev-parse HEAD)"
+    - export BUILD_INFO_BUILD_TIME="$CI_JOB_STARTED_AT"
+    - export BUILD_INFO_CURRENT_VERSION="$(tag=$(git ls-remote origin "refs/tags/v*[0-9]" 2>/dev/null | cut -f 2- | sort -V | tail -1 | sed 's/refs\\/tags\\/v//'); [ -z "$tag" ] && echo "0.0.0" || echo "$tag")"
+    - export HOSTNAME="$(printf %s "pan-my-ai-app-prod-llm-$CL_prod_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')"
+    - export ROOT_URL="https://$(printf %s "pan-my-ai-app-prod-llm-$CL_prod_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')"
+    - export HOSTNAME_INTERNAL="$(printf %s "pan-my-ai-app-prod-llm-$CL_prod_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')"
+    - export ROOT_URL_INTERNAL="https://$(printf %s "pan-my-ai-app-prod-llm-$CL_prod_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')"
+    - export DEPLOY_CLOUD_RUN_PROJECT_ID="google-project-id"
+    - export DEPLOY_CLOUD_RUN_REGION="europe-west4"
+    - export HF_TOKEN="$CL_prod_llm_HF_TOKEN"
+    - export GCLOUD_DEPLOY_credentialsKey="$CL_prod_llm_GCLOUD_DEPLOY_credentialsKey"
+    - export GCLOUD_RUN_canonicalHostSuffix="$CL_prod_llm_GCLOUD_RUN_canonicalHostSuffix"
+    - export _ALL_ENV_VAR_KEYS="[\\"ENV_SHORT\\",\\"APP_DIR\\",\\"ENV_TYPE\\",\\"BUILD_INFO_BUILD_ID\\",\\"BUILD_INFO_BUILD_TIME\\",\\"BUILD_INFO_CURRENT_VERSION\\",\\"HOSTNAME\\",\\"ROOT_URL\\",\\"HOSTNAME_INTERNAL\\",\\"ROOT_URL_INTERNAL\\",\\"DEPLOY_CLOUD_RUN_PROJECT_ID\\",\\"DEPLOY_CLOUD_RUN_REGION\\",\\"HF_TOKEN\\",\\"GCLOUD_DEPLOY_credentialsKey\\",\\"GCLOUD_RUN_canonicalHostSuffix\\"]"
+    - export DOCKER_REGISTRY="europe-west4-docker.pkg.dev"
+    - export DOCKER_IMAGE="europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/prod/llm"
+    - export DOCKER_CACHE_IMAGE="europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm"
+    - export DOCKER_IMAGE_TAG="$CI_COMMIT_SHA"
+    - export CLOUDSDK_CORE_DISABLE_PROMPTS="1"
+    - collapseable_section_end "injectvars"
+    - collapseable_section_start "prepare" "Prepare..."
+    - gcloud auth activate-service-account --key-file=<(echo "$CL_prod_llm_GCLOUD_DEPLOY_credentialsKey")
+    - export GCLOUD_PROJECT_NUMBER=$(gcloud projects describe google-project-id --format="value(projectNumber)")
+    - 'echo "GCLOUD_PROJECT_NUMBER: $GCLOUD_PROJECT_NUMBER"'
+    - collapseable_section_end "prepare"
+    - collapseable_section_start "writeenvvars" "Write env vars to file"
+    - |
+      cat > ____envvars.yaml <<EOF
+      ENV_SHORT: |-
+        prod
+      APP_DIR: |-
+        llm
+      ENV_TYPE: |-
+        prod
+      BUILD_INFO_BUILD_ID: |-
+        $(printf %s "$(git describe --tags 2>/dev/null || git rev-parse HEAD)" | sed '1!s/^/  /')
+      BUILD_INFO_BUILD_TIME: |-
+        $(printf %s "$CI_JOB_STARTED_AT" | sed '1!s/^/  /')
+      BUILD_INFO_CURRENT_VERSION: |-
+        $(printf %s "$(tag=$(git ls-remote origin "refs/tags/v*[0-9]" 2>/dev/null | cut -f 2- | sort -V | tail -1 | sed 's/refs\\/tags\\/v//'); [ -z "$tag" ] && echo "0.0.0" || echo "$tag")" | sed '1!s/^/  /')
+      HOSTNAME: |-
+        $(printf %s "$(printf %s "pan-my-ai-app-prod-llm-$CL_prod_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')" | sed '1!s/^/  /')
+      ROOT_URL: |-
+        $(printf %s "https://$(printf %s "pan-my-ai-app-prod-llm-$CL_prod_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')" | sed '1!s/^/  /')
+      HOSTNAME_INTERNAL: |-
+        $(printf %s "$(printf %s "pan-my-ai-app-prod-llm-$CL_prod_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')" | sed '1!s/^/  /')
+      ROOT_URL_INTERNAL: |-
+        $(printf %s "https://$(printf %s "pan-my-ai-app-prod-llm-$CL_prod_llm_GCLOUD_RUN_canonicalHostSuffix" | awk '{print tolower($0)}')" | sed '1!s/^/  /')
+      DEPLOY_CLOUD_RUN_PROJECT_ID: |-
+        google-project-id
+      DEPLOY_CLOUD_RUN_REGION: |-
+        europe-west4
+      HF_TOKEN: |-
+        $(printf %s "$CL_prod_llm_HF_TOKEN" | sed '1!s/^/  /')
+      GCLOUD_RUN_canonicalHostSuffix: |-
+        $(printf %s "$CL_prod_llm_GCLOUD_RUN_canonicalHostSuffix" | sed '1!s/^/  /')
+      _ALL_ENV_VAR_KEYS: |-
+        ["ENV_SHORT","APP_DIR","ENV_TYPE","BUILD_INFO_BUILD_ID","BUILD_INFO_BUILD_TIME","BUILD_INFO_CURRENT_VERSION","HOSTNAME","ROOT_URL","HOSTNAME_INTERNAL","ROOT_URL_INTERNAL","DEPLOY_CLOUD_RUN_PROJECT_ID","DEPLOY_CLOUD_RUN_REGION","HF_TOKEN","GCLOUD_DEPLOY_credentialsKey","GCLOUD_RUN_canonicalHostSuffix"]
+
+      EOF
+    - collapseable_section_end "writeenvvars"
+    - collapseable_section_start "deploy" "Deploy to cloud run"
+    - gcloud beta run deploy pan-my-ai-app-prod-llm --command="" --args="--model-id=meta-llama/Llama-3.2-1B-Instruct","--max-concurrent-requests=1" --image=us-docker.pkg.dev/deeplearning-platform-release/gcr.io/huggingface-text-generation-inference-cu124.2-3.ubuntu2204.py311 --project=google-project-id --region=europe-west4 --labels=customer-name=pan,component-name=llm,app-name=my-ai-app,env-type=prod,env-name=prod,build-type=disabled,cloud-run-service-name=pan-my-ai-app-prod-llm --env-vars-file=____envvars.yaml --min-instances=0 --max-instances=100 --no-cpu-throttling --cpu=8 --memory=32Gi --allow-unauthenticated --ingress=all --cpu-boost --gpu=1 --gpu-type=nvidia-l4
+    - collapseable_section_end "deploy"
+    - collapseable_section_start "cleanup" "Cleanup"
+    - gcloud run revisions list --project=google-project-id --region=europe-west4 --service=pan-my-ai-app-prod-llm --limit=unlimited --sort-by=metadata.creationTimestamp --format="value(name)" --filter='(status.conditions.status=False OR status.conditions.status=Unknown)' | tail -n +6 | while read -r revisionname; do gcloud run revisions delete --project=google-project-id --region=europe-west4 --quiet $revisionname ; done
+    - gcloud artifacts docker images list europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/prod/llm --sort-by=~CREATE_TIME --format="value(version)" | tail -n +7 | while read -r version; do gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/prod/llm@$version --quiet --delete-tags; done
+    - gcloud artifacts docker images list europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm --sort-by=~CREATE_TIME --format="value(version)" | tail -n +2 | while read -r version; do gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm@$version --quiet --delete-tags; done
+    - collapseable_section_end "cleanup"
+    - echo "CL_GITLAB_ENVIRONMENT_URL=$ROOT_URL" >> gitlab_environment.env
+  environment:
+    name: prod/llm
+    url: $CL_GITLAB_ENVIRONMENT_URL
+    on_stop: 'llm 🛑 Stop ⚠️ | prod '
+  artifacts:
+    reports:
+      dotenv: gitlab_environment.env
+  rules:
+    - when: manual
+      if: $CI_COMMIT_TAG
+  needs: []
+  retry: *a1
+  interruptible: true
+  allow_failure: true
+'llm 🛑 Stop ⚠️ | prod ':
+  stage: stop prod
+  image: path/to/docker/gcloud:the-version
+  variables:
+    KUBERNETES_CPU_REQUEST: '0.22'
+    KUBERNETES_MEMORY_REQUEST: 200Mi
+    KUBERNETES_MEMORY_LIMIT: 400Mi
+    GIT_STRATEGY: none
+  script:
+    - collapseable_section_start "injectvars" "Injecting variables"
+    - export CLOUDSDK_CORE_DISABLE_PROMPTS="1"
+    - collapseable_section_end "injectvars"
+    - set +e
+    - gcloud auth activate-service-account --key-file=<(echo "$CL_prod_llm_GCLOUD_DEPLOY_credentialsKey")
+    - gcloud run services delete pan-my-ai-app-prod-llm --project=google-project-id --region=europe-west4
+    - gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/prod/llm --quiet --delete-tags
+    - gcloud artifacts docker images list europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm --sort-by=~CREATE_TIME --format="value(version)" | tail -n +2 | while read -r version; do gcloud artifacts docker images delete europe-west4-docker.pkg.dev/google-project-id/catladder-deploy/pan-my-ai-app/caches/llm@$version --quiet --delete-tags; done
+    - set -e
+  environment:
+    name: prod/llm
+    action: stop
+  rules:
+    - if: $CI_COMMIT_BRANCH =~ /^[0-9]+\\.([0-9]+|x)\\.x$/
+      when: on_success
+    - when: manual
+      if: $CI_COMMIT_TAG
+  needs: []
+  retry: *a1
+  interruptible: true
+  allow_failure: true
+create release:
+  stage: release
+  image: path/to/docker/semantic-release:the-version
+  script:
+    - semanticRelease
+    - echo '👉 The project access token might be invald - run \`project-renew-token\` in catladder CLI to fix.'
+  rules:
+    - &a2
+      if: $CI_COMMIT_MESSAGE =~ /^chore\\(release\\).*/
+      when: never
+    - &a3
+      if: $CI_PIPELINE_SOURCE  == "schedule"
+      when: never
+    - if: $CI_COMMIT_BRANCH == $CI_DEFAULT_BRANCH && $AUTO_RELEASE == "true"
+      when: on_success
+    - if: $CI_COMMIT_BRANCH == $CI_DEFAULT_BRANCH
+      when: manual
+    - if: $CI_COMMIT_BRANCH =~ /^[0-9]+.([0-9]+|x).x$/
+      when: manual
+⚠️ force create release:
+  stage: release
+  image: path/to/docker/semantic-release:the-version
+  script:
+    - semanticRelease
+    - echo '👉 The project access token might be invald - run \`project-renew-token\` in catladder CLI to fix.'
+  rules:
+    - *a2
+    - *a3
+    - if: $CI_COMMIT_BRANCH == $CI_DEFAULT_BRANCH
+      when: manual
+    - if: $CI_COMMIT_BRANCH =~ /^[0-9]+.([0-9]+|x).x$/
+      when: manual
+  needs: []
+"
+`;

package/examples/cloud-run-llama.test.ts

@@ -0,0 +1,11 @@
+import { createYamlLocalPipeline } from "./__utils__/helpers";
+import config from "./cloud-run-llama";
+
+/**
+ * This test is auto-generated.
+ * Modifications will be overwritten on every `yarn test` run!
+ */
+
+it("matches snapshot for cloud-run-llama local pipeline YAML", async () => {
+  expect(await createYamlLocalPipeline(config)).toMatchSnapshot();
+});