aisingapore · Syakyr · Mar 25, 2025 · Mar 25, 2025 · Mar 25, 2025
diff --git a/{{cookiecutter.repo_name}}/.gitlab-ci.yml b/{{cookiecutter.repo_name}}/.gitlab-ci.yml
@@ -217,9 +217,9 @@ build:cpu-image:
   script:
     - mkdir -p /kaniko/.docker
 {%- if cookiecutter.platform == 'onprem' %}
-    - cat $HARBOR_ROBOT_CREDS_JSON > /kaniko/.docker/config.json
+    - echo $HARBOR_ROBOT_CREDS_JSON > /kaniko/.docker/config.json
 {%- elif cookiecutter.platform == 'gcp' %}
-    - cat $GCP_SERVICE_ACCOUNT_KEY > /kaniko/.docker/config.json
+    - echo $GCP_SERVICE_ACCOUNT_KEY > /kaniko/.docker/config.json
 {%- endif %}
     - >-
       /kaniko/executor
@@ -259,9 +259,9 @@ build:gpu-image:
   script:
     - mkdir -p /kaniko/.docker
 {%- if cookiecutter.platform == 'onprem' %}
-    - cat $HARBOR_ROBOT_CREDS_JSON > /kaniko/.docker/config.json
+    - echo $HARBOR_ROBOT_CREDS_JSON > /kaniko/.docker/config.json
 {%- elif cookiecutter.platform == 'gcp' %}
-    - cat $GCP_SERVICE_ACCOUNT_KEY > /kaniko/.docker/config.json
+    - echo $GCP_SERVICE_ACCOUNT_KEY > /kaniko/.docker/config.json
 {%- endif %}
     - >-
       /kaniko/executor
@@ -302,11 +302,11 @@ build:retag-images:
 {%- endif %}
   script:
 {%- if cookiecutter.platform == 'onprem' %}
-    - cat $HARBOR_ROBOT_CREDS_JSON > /root/.docker/config.json
+    - echo $HARBOR_ROBOT_CREDS_JSON > /root/.docker/config.json
     - crane tag {{cookiecutter.registry_project_path}}/cpu:${CI_COMMIT_SHORT_SHA} ${CI_COMMIT_TAG}
     - crane tag {{cookiecutter.registry_project_path}}/gpu:${CI_COMMIT_SHORT_SHA} ${CI_COMMIT_TAG}
 {%- elif cookiecutter.platform == 'gcp' %}
-    - cat $GCP_SERVICE_ACCOUNT_KEY > /gcp-sa.json
+    - echo $GCP_SERVICE_ACCOUNT_KEY > /gcp-sa.json
     - gcloud container images add-tag "{{cookiecutter.registry_project_path}}/cpu:${CI_COMMIT_SHORT_SHA}" "{{cookiecutter.registry_project_path}}/data-prep:${CI_COMMIT_TAG}"
     - gcloud container images add-tag "{{cookiecutter.registry_project_path}}/gpu:${CI_COMMIT_SHORT_SHA}" "{{cookiecutter.registry_project_path}}/model-training:${CI_COMMIT_TAG}"
 {%- endif %}

diff --git a/{{cookiecutter.repo_name}}/aisg-context/guide-site/docs/appendix/09-cicd.md b/{{cookiecutter.repo_name}}/aisg-context/guide-site/docs/appendix/09-cicd.md
@@ -117,7 +117,7 @@ so:
     ```
 
 Using the output from above, copy and paste the following content
-into a CI/CD environment variable of type `File`
+into a CI/CD environment variable of type `variable`
 (under `Settings` -> `CI/CD` -> `Variables` -> `Add variable`):
 
 ```json
@@ -410,9 +410,9 @@ that builds a Docker image:
       script:
         - mkdir -p /kaniko/.docker
 {%- if cookiecutter.platform == 'onprem' %}
-        - cat $HARBOR_ROBOT_CREDS_JSON > /kaniko/.docker/config.json
+        - echo $HARBOR_ROBOT_CREDS_JSON > /kaniko/.docker/config.json
 {%- elif cookiecutter.platform == 'gcp' %}
-        - cat $GCP_SERVICE_ACCOUNT_KEY > /kaniko/.docker/config.json
+        - echo $GCP_SERVICE_ACCOUNT_KEY > /kaniko/.docker/config.json
 {%- endif %}
         - >-
           /kaniko/executor
@@ -531,11 +531,11 @@ the default branch before this.
 {%- endif %}
       script:
 {%- if cookiecutter.platform == 'onprem' %}
-        - cat $HARBOR_ROBOT_CREDS_JSON > /root/.docker/config.json
+        - echo $HARBOR_ROBOT_CREDS_JSON > /root/.docker/config.json
         - crane tag {{cookiecutter.registry_project_path}}/cpu:${CI_COMMIT_SHORT_SHA} ${CI_COMMIT_TAG}
         - crane tag {{cookiecutter.registry_project_path}}/gpu:${CI_COMMIT_SHORT_SHA} ${CI_COMMIT_TAG}
 {%- elif cookiecutter.platform == 'gcp' %}
-        - cat $GCP_SERVICE_ACCOUNT_KEY > /gcp-sa.json
+        - echo $GCP_SERVICE_ACCOUNT_KEY > /gcp-sa.json
         - gcloud container images add-tag "{{cookiecutter.registry_project_path}}/cpu:${CI_COMMIT_SHORT_SHA}" "{{cookiecutter.registry_project_path}}/cpu:${CI_COMMIT_TAG}"
         - gcloud container images add-tag "{{cookiecutter.registry_project_path}}/gpu:${CI_COMMIT_SHORT_SHA}" "{{cookiecutter.registry_project_path}}/gpu:${CI_COMMIT_TAG}"
 {%- endif %}

diff --git a/...ide-site/docs/appendix/assets/screenshots/gitlab-settings-cicd-set-file-var.png b/...ide-site/docs/appendix/assets/screenshots/gitlab-settings-cicd-set-file-var.png
diff --git a/{{cookiecutter.repo_name}}/aisg-context/guide-site/docs/docker/07b-deployment.md b/{{cookiecutter.repo_name}}/aisg-context/guide-site/docs/docker/07b-deployment.md
@@ -125,12 +125,7 @@ that exists within the repository.
 
 ### Running the API Server
 
-Run the FastAPI server using [Gunicorn](https://gunicorn.org)
-(for Linux/macOS) or [`uvicorn`][uvicorn] (for Windows):
-
-!!! attention
-    Gunicorn is only executable on UNIX-based or UNIX-like systems;
-    this method would not be possible/applicable for Windows machines.
+Run the FastAPI server using [`uvicorn`][uvicorn]:
 
 === "Linux"
 
@@ -147,9 +142,8 @@ Run the FastAPI server using [Gunicorn](https://gunicorn.org)
         -v ./models:/home/aisg/{{cookiecutter.repo_name}}/models \
         -w /home/aisg/{{cookiecutter.repo_name}}/src \
         {{cookiecutter.registry_project_path}}/gpu:0.1.0 \
-        gunicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
-            -k uvicorn.workers.UvicornWorker \
-            -b 0.0.0.0:8080 -w 2 -t 90
+        uvicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
+            --host 0.0.0.0 --port 8080 --workers 2 --app-dir src
     ```
 
 === "macOS"
@@ -160,9 +154,8 @@ Run the FastAPI server using [Gunicorn](https://gunicorn.org)
         -v ./models:/home/aisg/{{cookiecutter.repo_name}}/models \
         -w /home/aisg/{{cookiecutter.repo_name}}/src \
         {{cookiecutter.registry_project_path}}/gpu:0.1.0 \
-        gunicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
-            -k uvicorn.workers.UvicornWorker \
-            -b 0.0.0.0:8080 -w 2 -t 90
+        uvicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
+            --host 0.0.0.0 --port 8080 --workers 2 --app-dir src
     ```
 
 === "Windows PowerShell"
@@ -181,9 +174,8 @@ Run the FastAPI server using [Gunicorn](https://gunicorn.org)
         -v .\models:/home/aisg/{{cookiecutter.repo_name}}/models `
         -w /home/aisg/{{cookiecutter.repo_name}}/src `
         {{cookiecutter.registry_project_path}}/gpu:0.1.0 `
-        gunicorn {{cookiecutter.src_package_name}}_fastapi.main:APP `
-            -k uvicorn.workers.UvicornWorker `
-            -b 0.0.0.0:8080 -w 2 -t 90
+        uvicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
+            --host 0.0.0.0 --port 8080 --workers 2 --app-dir src
     ```
 
 And with that, our document site for our server is viewable through
@@ -234,7 +226,6 @@ With the returned JSON object, we have successfully submitted a request
 to the FastAPI server and it returned predictions as part of the
 response.
 
-[reason]: https://fastapi.tiangolo.com/deployment/server-workers/
 [uvicorn]: https://www.uvicorn.org/
 [rocm-wsl]: https://rocm.docs.amd.com/projects/radeon/en/latest/docs/install/wsl/howto_wsl.html
 

diff --git a/{{cookiecutter.repo_name}}/aisg-context/guide-site/docs/local/07a-deployment.md b/{{cookiecutter.repo_name}}/aisg-context/guide-site/docs/local/07a-deployment.md
@@ -104,33 +104,22 @@ that exists within the repository.
 
 ### Running the API Server
 
-Run the FastAPI server using [Gunicorn](https://gunicorn.org)
-(for Linux/macOS) or [`uvicorn`][uvicorn] (for Windows):
-
-!!! attention
-    Gunicorn is only executable on UNIX-based or UNIX-like systems;
-    this method would not be possible/applicable for Windows machines.
+Run the FastAPI server using [`uvicorn`][uvicorn]:
 
 === "Linux/macOS"
 
     ```bash
     conda activate {{cookiecutter.repo_name}}
-    gunicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
-        -k uvicorn.workers.UvicornWorker \
-        -b 0.0.0.0:8080 -w 2 -t 90 --chdir src
+    uvicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
+        --host 0.0.0.0 --port 8080 --workers 2 --app-dir src
     ```
 
-    !!! info
-        See [here][reason] as to why Gunicorn is to be used instead of
-        just [Uvicorn][uvicorn]. TLDR: Gunicorn is needed to spin up 
-        multiple processes/workers to handle more requests i.e. better 
-        for the sake of production needs.
-
 === "Windows PowerShell"
 
     ```powershell
     conda activate {{cookiecutter.repo_name}}
-    uvicorn {{cookiecutter.src_package_name}}_fastapi.main:APP --app-dir src
+    uvicorn {{cookiecutter.src_package_name}}_fastapi.main:APP `
+        --host 0.0.0.0 --port 8080 --workers 2 --app-dir src
     ```
 
 And with that, our document site for our server is viewable through
@@ -169,7 +158,6 @@ With the returned JSON object, we have successfully submitted a request
 to the FastAPI server and it returned predictions as part of the
 response.
 
-[reason]: https://fastapi.tiangolo.com/deployment/server-workers/
 [uvicorn]: https://www.uvicorn.org/
 
 ### Pydantic Settings

diff --git a/{{cookiecutter.repo_name}}/aisg-context/guide-site/docs/runai/07c-deployment.md b/{{cookiecutter.repo_name}}/aisg-context/guide-site/docs/runai/07c-deployment.md
@@ -124,17 +124,16 @@ that exists within the repository.
 
 ### Running the API Server
 
-Run the FastAPI server using [Gunicorn](https://gunicorn.org):
+Run the FastAPI server using [`uvicorn`][uvicorn]:
 
 === "Coder Workspace Terminal"
 
     ```bash
     # Running in a working `{{cookiecutter.repo_name}}` repository
     conda activate {{cookiecutter.repo_name}}
     export MODEL_UUID=<MLFLOW_RUN_UUID>
-    gunicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
-        -k uvicorn.workers.UvicornWorker \
-        -b 0.0.0.0:8080 -w 2 -t 90 --chdir src
+    uvicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
+        --host 0.0.0.0 --port 8080 --workers 2 --app-dir src
     ```
 
     And with that, the link to our document site for our server would
@@ -153,9 +152,8 @@ Run the FastAPI server using [Gunicorn](https://gunicorn.org):
         --existing-pvc claimname=<NAME_OF_DATA_SOURCE>,path=/<NAME_OF_DATA_SOURCE> \
         --cpu 2 --cpu-limit 2 --memory 4G --memory-limit 4G --backoff-limit 1 \
         --service-type external-url,port=8080:8080 \
-        --command -- gunicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
-            -k uvicorn.workers.UvicornWorker \
-            -b 0.0.0.0:8080 -w 2 -t 90
+        --command -- uvicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
+            --host 0.0.0.0 --port 8080 --workers 2 --app-dir src
     ```
 
     And with that, the link to our document site for our server would
@@ -184,7 +182,6 @@ With the returned JSON object, we have successfully submitted a request
 to the FastAPI server and it returned predictions as part of the
 response.
 
-[reason]: https://fastapi.tiangolo.com/deployment/server-workers/
 [uvicorn]: https://www.uvicorn.org/
 
 ### Pydantic Settings

diff --git a/...po_name}}/problem-templates/cv/aisg-context/guide-site/docs/docker/07b-deployment.md.diff b/...po_name}}/problem-templates/cv/aisg-context/guide-site/docs/docker/07b-deployment.md.diff
@@ -56,37 +56,37 @@
  [beginner tutorials]: https://fastapi.tiangolo.com/tutorial/
 
  ### Running the API Server
-@@ -145,8 +142,9 @@
+@@ -140,8 +137,9 @@
      docker run --rm \
          -p 8080:8080 \
          -v ./models:/home/aisg/{{cookiecutter.repo_name}}/models \
          -w /home/aisg/{{cookiecutter.repo_name}}/src \
 +        -e PRED_MODEL_PATH="../models/$PRED_MODEL_UUID/model/model.pt" \
          {{cookiecutter.registry_project_path}}/gpu:0.1.0 \
-         gunicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
-             -k uvicorn.workers.UvicornWorker \
-             -b 0.0.0.0:8080 -w 2 -t 90
-@@ -158,8 +156,9 @@
+         uvicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
+             --host 0.0.0.0 --port 8080 --workers 2 --app-dir src
+     ```
+@@ -152,8 +150,9 @@
      docker run --rm \
          -p 8080:8080 \
          -v ./models:/home/aisg/{{cookiecutter.repo_name}}/models \
          -w /home/aisg/{{cookiecutter.repo_name}}/src \
 +        -e PRED_MODEL_PATH="../models/$PRED_MODEL_UUID/model/model.pt" \
          {{cookiecutter.registry_project_path}}/gpu:0.1.0 \
-         gunicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
-             -k uvicorn.workers.UvicornWorker \
-             -b 0.0.0.0:8080 -w 2 -t 90
-@@ -179,8 +178,9 @@
+         uvicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
+             --host 0.0.0.0 --port 8080 --workers 2 --app-dir src
+     ```
+@@ -172,8 +171,9 @@
      docker run --rm `
          -p 8080:8080 `
          -v .\models:/home/aisg/{{cookiecutter.repo_name}}/models `
          -w /home/aisg/{{cookiecutter.repo_name}}/src `
 +        -e PRED_MODEL_PATH="../models/$Env:PRED_MODEL_UUID/model/model.pt" `
          {{cookiecutter.registry_project_path}}/gpu:0.1.0 `
-         gunicorn {{cookiecutter.src_package_name}}_fastapi.main:APP `
-             -k uvicorn.workers.UvicornWorker `
-             -b 0.0.0.0:8080 -w 2 -t 90
-@@ -196,39 +196,39 @@
+         uvicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
+             --host 0.0.0.0 --port 8080 --workers 2 --app-dir src
+     ```
+@@ -188,39 +188,39 @@
 
  === "Linux"
 
@@ -136,7 +136,7 @@
 
  With the returned JSON object, we have successfully submitted a request
  to the FastAPI server and it returned predictions as part of the
-@@ -245,9 +245,10 @@
+@@ -236,9 +236,10 @@
  data and schema validation, as well as [settings management]. There's a 
  class called `Settings` under the module
  `src/{{cookiecutter.src_package_name}}_fastapi/config.py`. This class 
@@ -148,7 +148,7 @@
  `src/{{cookiecutter.src_package_name}}_fastapi/config.py`:
  ```python
  ...
-@@ -256,9 +257,12 @@
+@@ -247,9 +248,12 @@
      API_NAME: str = "{{cookiecutter.src_package_name}}_fastapi"
      API_V1_STR: str = "/api/v1"
      LOGGER_CONFIG_PATH: str = "../conf/logging.yaml"

diff --git a/...epo_name}}/problem-templates/cv/aisg-context/guide-site/docs/local/07a-deployment.md.diff b/...epo_name}}/problem-templates/cv/aisg-context/guide-site/docs/local/07a-deployment.md.diff
@@ -57,7 +57,7 @@
  [beginner tutorials]: https://fastapi.tiangolo.com/tutorial/
 
  ### Running the API Server
-@@ -145,25 +160,25 @@
+@@ -134,25 +149,25 @@
 
      ```bash
      curl -X POST \
@@ -88,7 +88,7 @@
 
  With the returned JSON object, we have successfully submitted a request
  to the FastAPI server and it returned predictions as part of the
-@@ -179,9 +194,10 @@
+@@ -167,9 +182,10 @@
  data and schema validation, as well as [settings management]. There's a 
  class called `Settings` under the module
  `src/{{cookiecutter.src_package_name}}_fastapi/config.py`. This class 
@@ -100,7 +100,7 @@
  `src/{{cookiecutter.src_package_name}}_fastapi/config.py`:
  ```python
  ...
-@@ -190,9 +206,12 @@
+@@ -178,9 +194,12 @@
      API_NAME: str = "{{cookiecutter.src_package_name}}_fastapi"
      API_V1_STR: str = "/api/v1"
      LOGGER_CONFIG_PATH: str = "../conf/logging.yaml"

diff --git a/...epo_name}}/problem-templates/cv/aisg-context/guide-site/docs/runai/07c-deployment.md.diff b/...epo_name}}/problem-templates/cv/aisg-context/guide-site/docs/runai/07c-deployment.md.diff
@@ -58,11 +58,11 @@
 -    export MODEL_UUID=<MLFLOW_RUN_UUID>
 +    export PRED_MODEL_UUID=<MLFLOW_RUN_UUID>
 +    export PRED_MODEL_PATH="<NAME_OF_DATA_SOURCE>/workspaces/<YOUR_HYPENATED_NAME>/models/$PRED_MODEL_UUID/model/model.pt"
-     gunicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
-         -k uvicorn.workers.UvicornWorker \
-         -b 0.0.0.0:8080 -w 2 -t 90 --chdir src
+     uvicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
+         --host 0.0.0.0 --port 8080 --workers 2 --app-dir src
      ```
-@@ -144,15 +148,16 @@
+
+@@ -143,15 +147,16 @@
 
  === "Using Run:ai"
 
@@ -77,10 +77,10 @@
          --cpu 2 --cpu-limit 2 --memory 4G --memory-limit 4G --backoff-limit 1 \
 +        -e PRED_MODEL_PATH=/<NAME_OF_DATA_SOURCE>/workspaces/<YOUR_HYPENATED_NAME>/models/$PRED_MODEL_UUID/model/model.pt \
          --service-type external-url,port=8080:8080 \
-         --command -- gunicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
-             -k uvicorn.workers.UvicornWorker \
-             -b 0.0.0.0:8080 -w 2 -t 90
-@@ -169,16 +174,16 @@
+         --command -- uvicorn {{cookiecutter.src_package_name}}_fastapi.main:APP \
+             --host 0.0.0.0 --port 8080 --workers 2 --app-dir src
+     ```
+@@ -167,16 +172,16 @@
 
      ```bash
      curl -X POST \
@@ -100,7 +100,7 @@
 
  With the returned JSON object, we have successfully submitted a request
  to the FastAPI server and it returned predictions as part of the
-@@ -194,9 +199,10 @@
+@@ -191,9 +196,10 @@
  data and schema validation, as well as [settings management]. There's a 
  class called `Settings` under the module
  `src/{{cookiecutter.src_package_name}}_fastapi/config.py`. This class 
@@ -112,7 +112,7 @@
  `src/{{cookiecutter.src_package_name}}_fastapi/config.py`:
  ```python
  ...
-@@ -205,9 +211,12 @@
+@@ -202,9 +208,12 @@
      API_NAME: str = "{{cookiecutter.src_package_name}}_fastapi"
      API_V1_STR: str = "/api/v1"
      LOGGER_CONFIG_PATH: str = "../conf/logging.yaml"