Regenerate RAG Template

Signed-off-by: Maysun J Faisal <[email protected]>

Author: maysunfaisal

scripts/envs/audio-to-text

@@ -17,3 +17,6 @@ export SUPPORT_DETR=false
 
 # for gitlab case, since gitlab does not have pipeline webhook pre-set to trigger the initial build
 export APP_INTERFACE_CONTAINER="quay.io/redhat-ai-dev/audio-to-text:latest"
+
+# Database configurations
+export SUPPORT_DB=false

scripts/envs/chatbot

@@ -21,3 +21,6 @@ export VLLM_MAX_MODEL_LEN=4096
 
 # for gitlab case, since gitlab does not have pipeline webhook pre-set to trigger the initial build
 export APP_INTERFACE_CONTAINER="quay.io/redhat-ai-dev/chatbot:latest"
+
+# Database configurations
+export SUPPORT_DB=false

scripts/envs/codegen

@@ -22,3 +22,6 @@ export VLLM_MAX_MODEL_LEN=6144
 
 # for gitlab case, since gitlab does not have pipeline webhook pre-set to trigger the initial build
 export APP_INTERFACE_CONTAINER="quay.io/redhat-ai-dev/codegen:latest"
+
+# Database configurations
+export SUPPORT_DB=false

scripts/envs/object-detection

@@ -17,3 +17,6 @@ export SUPPORT_DETR=true
 
 # for gitlab case, since gitlab does not have pipeline webhook pre-set to trigger the initial build
 export APP_INTERFACE_CONTAINER="quay.io/redhat-ai-dev/object_detection:latest"
+
+# Database configurations
+export SUPPORT_DB=false

scripts/envs/rag

@@ -5,6 +5,7 @@ export APP_TAGS='["ai", "llamacpp", "vllm", "python", "rag", "database"]'
 export APP_RUN_COMMAND="streamlit run rag_app.py"
 export INIT_CONTAINER="quay.io/redhat-ai-dev/granite-7b-lab:latest"
 export INIT_CONTAINER_COMMAND="['/usr/bin/install', '/model/model.file', '/shared/']"
+export MODEL_PATH="/model/model.file"
 export APP_PORT=8501
 export MODEL_SERVICE_CONTAINER="quay.io/ai-lab/llamacpp_python:latest"
 export MODEL_SERVICE_PORT=8001
@@ -21,7 +22,7 @@ export VLLM_MAX_MODEL_LEN=4096
 # for gitlab case, since gitlab does not have pipeline webhook pre-set to trigger the initial build
 export APP_INTERFACE_CONTAINER="quay.io/redhat-ai-dev/rag:latest"
 
-# Database Required
+# Database configurations
 export SUPPORT_DB=true
 export DB_CONTAINER="quay.io/redhat-ai-dev/chroma:latest"
 export DB_PORT=8000

scripts/import-ai-lab-samples

@@ -62,6 +62,9 @@ for f in */; do
         if [ $SUPPORT_DETR == false ]; then
             sed -i '/# SED_DETR_MODEL_SERVER_START/,/# SED_DETR_MODEL_SERVER_END/d' $DEST/template.yaml
         fi
+        if [ $SUPPORT_DB == false ]; then
+            sed -i '/# SED_DB_START/,/# SED_DB_END/d' $DEST/template.yaml
+        fi
 
         source $ROOT_DIR/properties
         cat $DEST/template.yaml | envsubst > $DEST/new-template.yaml

skeleton/gitops-template/components/http/base/deployment-database.yaml

@@ -4,18 +4,18 @@ metadata:
   labels: 
     app.kubernetes.io/instance: ${{ values.name }}-database
     app.kubernetes.io/managed-by: kustomize
-    app.kubernetes.io/name:  ${{ values.name }}-database
+    app.kubernetes.io/name: ${{ values.name }}-database
     app.kubernetes.io/part-of: ${{ values.name }}  
   name: ${{ values.name }}-database
 spec:
   replicas: 1
   selector:
     matchLabels:
-      app.kubernetes.io/instance:  ${{ values.name }}-database
+      app.kubernetes.io/instance: ${{ values.name }}-database
   template:
     metadata: 
       labels:
-        app.kubernetes.io/instance:  ${{ values.name }}-database
+        app.kubernetes.io/instance: ${{ values.name }}-database
     spec:
       containers:
       - image: ${{ values.dbContainer }}

skeleton/template.yaml

@@ -312,10 +312,12 @@ spec:
           modelEndpoint: ${{ parameters.modelEndpoint }}
           # for RHOAI
           rhoaiSelected: ${{ parameters.rhoaiSelected }}
+          # SED_DB_START
           # Database is required for the RAG templates
           dbRequired: ${SUPPORT_DB}
           dbContainer: ${DB_CONTAINER}
           dbPort: ${DB_PORT}
+          # SED_DB_END
     - action: fs:rename
       id: renameComponentDir
       name: Rename Component Directory

templates/rag/content/Containerfile

@@ -1,4 +1,4 @@
-FROM registry.access.redhat.com/ubi9/python-311:1-66.1720018730
+FROM registry.access.redhat.com/ubi9/python-311:1-72.1722518949
 ### Update sqlite for chroma
 USER root
 RUN dnf remove sqlite3 -y
@@ -21,4 +21,4 @@ EXPOSE  8501
 ENV  HF_HUB_CACHE=/rag/models/
 RUN mkdir -p /rag/models/
 RUN chgrp -R 0 /rag/models/ && chmod -R g=u /rag/models/
-ENTRYPOINT [ "streamlit", "run" ,"rag_app.py" ]
+ENTRYPOINT [ "streamlit", "run" ,"rag_app.py" ]

templates/rag/content/docs/.assets/access-openshift-ai.png

@@ -0,0 +1,48 @@
+# Running Samples in OpenShift AI
+
+This document will outline how you can run build and run your sample applications within an OpenShift AI workbench.
+
+## Prerequisites
+
+- Red Hat OpenShift AI installed, and `Create workbench for OpenShift AI` selected during component creation.
+- `oc` cli installed
+   - `oc` can be downloaded from https://mirror.openshift.com/pub/openshift-v4/clients/ocp/stable/
+- Permissions to run `oc port-forward` on the cluster, specifically an account with the following roles:
+   - `get`, `create`, and `list` for the `pods/portforward` subresource
+
+## Running the Sample
+
+1) While on Console, navigate to the `OpenShift AI` by clicking the square "apps" icon on the top-right corner (next to the notifications icon). `Openshift AI` will be listed in the dropdown list
+
+![image](./.assets/access-openshift-ai.png)
+
+2) Go to the `Data Science Projects` section and access your application's project named `{{ values.namespace }}`
+
+![image](./.assets/data-science-projects.png)
+
+3) Access the `workbench` named `{{ values.name }}-notebook`
+
+![image](./.assets/access-workbench.png)
+
+4) Go to `File->Open` and select `Terminal`
+
+![image](./.assets/open-terminal.png)
+
+5) In the terminal, run `cd ${{ values.name }}` to navigate to your sample app's directory
+
+6) Run `pip install --upgrade -r requirements.txt` to install the dependencies for your application
+
+7) Run `${{ values.appRunCommand }}` to run the sample in the workbench.
+
+## Accessing the Sample
+
+With the sample app now running, the following steps will allow you to access the sample app in your browser:
+
+1) Navigate back to the OpenShift AI dashboard, and find the name of your workbench.
+![image](./.assets/workbench-name.png)
+
+2) In a terminal window on your machine, run `oc get pods -l app=<workbench-name>`. This will retrieve the name of the pod where the workbench is running.
+
+3) Run `oc port-forward <pod-name> ${{ values.appPort }}` to port forward the sample application's port to your local machine.
+
+4) Finally, visit `http://localhost:${{ values.appPort }}` in your browser to access the application.

templates/rag/content/docs/.assets/access-workbench.png

@@ -78,4 +78,4 @@ def clear_db(self):
                 self.client.drop_collection(self.collection_name)
             print("Cleared DB")
         except:
-            print("Couldn't clear the collection possibly because it doesn't exist") 
+            print("Couldn't clear the collection possibly because it doesn't exist") 

templates/rag/content/docs/.assets/data-science-projects.png

@@ -5,6 +5,7 @@ nav:
   - Source Component: source-component.md
   - Pipelines: pipelines.md
   - GitOps Application: gitops-application.md
+  - OpenShift AI: rhoai.md
 
 plugins:
   - techdocs-core

templates/rag/content/docs/.assets/open-terminal.png

@@ -1,7 +1,8 @@
 langchain-openai==0.1.7
 langchain==0.1.20
-chromadb==0.5.4
+chromadb==0.5.5
 sentence-transformers==2.7.0
 streamlit==1.34.0
 pypdf==4.2.0
 pymilvus==2.4.1
+

templates/rag/content/docs/.assets/workbench-name.png

@@ -2,9 +2,9 @@ apiVersion: scaffolder.backstage.io/v1beta3
 # https://backstage.io/docs/features/software-catalog/descriptor-format#kind-template
 kind: Template
 metadata:
-  name: rag2
-  title: RAG2 Chatbot Application
-  description: "RAG2 Chatbot Application example with an LLM and a Vector database enabled"
+  name: rag
+  title: RAG Chatbot Application
+  description: RAG Chatbot Application example with an LLM and a Vector database enabled
   tags: ["ai", "llamacpp", "vllm", "python", "rag", "database"]
 spec:
   type: service
@@ -177,8 +177,9 @@ spec:
           namespace: ${{ parameters.namespace }}
           repoURL: https://${{ parameters.githubServer if parameters.hostType === 'GitHub' else parameters.gitlabServer }}/${{ parameters.repoOwner }}/${{ parameters.repoName }}-gitops
           srcRepoURL: https://${{ parameters.githubServer if parameters.hostType === 'GitHub' else parameters.gitlabServer }}/${{ parameters.repoOwner }}/${{ parameters.repoName }}
-          appContainer: ${{ 'quay.io/redhat-ai-dev/ai-template-bootstrap-app:latest' if parameters.hostType === 'GitHub' else 'quay.io/redhat-ai-dev/chatbot:latest' }}
+          appContainer: ${{ 'quay.io/redhat-ai-dev/ai-template-bootstrap-app:latest' if parameters.hostType === 'GitHub' else 'quay.io/redhat-ai-dev/rag:latest' }}
           appPort: 8501
+          appRunCommand: "streamlit run rag_app.py"
           modelServiceContainer: quay.io/ai-lab/llamacpp_python:latest
           modelServicePort: 8001          
     # Renders all the template variables into the files and directory names and content, and places the result in the workspace.
@@ -191,15 +192,15 @@ spec:
         values:
           name: ${{ parameters.name }}
           namespace: ${{ parameters.namespace }}
-          description: Secure Supply Chain Example for RAG Application example with LLM enabled chat applications 
+          description: Secure Supply Chain Example for RAG Chatbot Application example with an LLM and a Vector database enabled 
           dockerfile: Containerfile
           buildContext: .
           gitopsSecretName: ${{ 'gitops-auth-secret' if parameters.hostType === 'GitHub' else 'gitlab-auth-secret' }}
           image: '${{ parameters.imageRegistry }}/${{ parameters.imageOrg }}/${{ parameters.imageName }}'
-          tags: '["ai", "llamacpp", "vllm", "python"]'
+          tags: '["ai", "llamacpp", "vllm", "python", "rag", "database"]'
           owner: ${{ parameters.owner }} 
           repoSlug: '${{ parameters.imageOrg }}/${{ parameters.imageName }}'
-          defaultBranch: ${{ parameters.branch }} 
+          defaultBranch: ${{ parameters.branch }}
     - id: fetch-github-action
       name: Fetch GitHub Action
       action: fetch:plain
@@ -270,7 +271,7 @@ spec:
           initContainer: quay.io/redhat-ai-dev/granite-7b-lab:latest
           modelInitCommand: "['/usr/bin/install', '/model/model.file', '/shared/']"
           modelPath: "/model/model.file"
-          appContainer: ${{ 'quay.io/redhat-ai-dev/ai-template-bootstrap-app:latest' if parameters.hostType === 'GitHub' else 'quay.io/redhat-ai-dev/chatbot:latest' }}
+          appContainer: ${{ 'quay.io/redhat-ai-dev/ai-template-bootstrap-app:latest' if parameters.hostType === 'GitHub' else 'quay.io/redhat-ai-dev/rag:latest' }}
           appPort: 8501
           modelServiceContainer: quay.io/ai-lab/llamacpp_python:latest
           modelServicePort: 8001
@@ -285,10 +286,12 @@ spec:
           modelEndpoint: ${{ parameters.modelEndpoint }}
           # for RHOAI
           rhoaiSelected: ${{ parameters.rhoaiSelected }}
+          # SED_DB_START
           # Database is required for the RAG templates
           dbRequired: true
           dbContainer: quay.io/redhat-ai-dev/chroma:latest
           dbPort: 8000
+          # SED_DB_END
     - action: fs:rename
       id: renameComponentDir
       name: Rename Component Directory