[Traffic Intersection Agent] Removing default models and related docs updates (open-edge-platform#1816)

Signed-off-by: Krishna Murti <krishna.murti@intel.com>

Author: krish918

metro-ai-suite/smart-traffic-intersection-agent/docker/agent-compose.yaml

@@ -20,7 +20,7 @@ services:
       http_proxy: ${http_proxy}
       https_proxy: ${https_proxy}
       # VLM Model configuration
-      VLM_MODEL_NAME: ${VLM_MODEL_NAME:-Qwen/Qwen2.5-VL-3B-Instruct}
+      VLM_MODEL_NAME: ${VLM_MODEL_NAME}
       VLM_COMPRESSION_WEIGHT_FORMAT: ${VLM_COMPRESSION_WEIGHT_FORMAT:-int4}
       VLM_DEVICE: ${VLM_DEVICE:-CPU}
       VLM_SEED: ${VLM_SEED:-42}
@@ -87,7 +87,7 @@ services:
       WEATHER_MOCK: ${WEATHER_MOCK:-false}
 
       # VLM configuration (overrides src/config/traffic_agent.json)
-      VLM_MODEL_NAME: ${VLM_MODEL_NAME:-Qwen/Qwen2.5-VL-3B-Instruct}
+      VLM_MODEL_NAME: ${VLM_MODEL_NAME}
       VLM_TIMEOUT_SECONDS: ${VLM_TIMEOUT_SECONDS:-}
       VLM_MAX_COMPLETION_TOKENS: ${VLM_MAX_COMPLETION_TOKENS:-}
       VLM_TEMPERATURE: ${VLM_TEMPERATURE:-}

metro-ai-suite/smart-traffic-intersection-agent/docs/user-guide/get-started.md

@@ -40,7 +40,14 @@ git clone https://github.com/open-edge-platform/edge-ai-suites.git
 cd metro-ai-suite/smart-traffic-intersection-agent/
 ```
 
-### 2. Run the Complete Setup
+### 2. Set the required environment variables
+
+```bash
+export VLM_MODEL_NAME=<supported_model_name>  # eg. microsoft/Phi-3.5-vision-instruct, Qwen/Qwen2.5-VL-3B-Instruct
+```
+> **IMPORTANT:** See this [disclaimer](#disclaimer-for-using-third-party-ai-models) before using any AI Model.
+
+### 3. Run the Complete Setup
 
 The easiest way to set up the service is to use default configurations without making any
 changes. Run the setup script with the `--setup` option to set up the agent quickly with
@@ -59,7 +66,7 @@ This single command will:
 - Build Docker images
 - Start services in the Smart Traffic Intersection Agent's application stack
 
-### 3. Run alternative setup options
+### 4. Run alternative setup options
 
 For a more granular control, run these commands:
 
@@ -83,7 +90,7 @@ source setup.sh --restart [service_type]
 source setup.sh --clean [option]
 ```
 
-### 4. Access Services
+### 5. Access Services
 
 When the script completes, it will show the URLs to access the services. Go to these URLs so
 that the respective services can access them in a web browser.
@@ -103,81 +110,92 @@ locations on the same machine for `n` required instances.
 
 ### Set up Instance #1
 
-1. Clone the repository into a new directory:
+1.  Clone the repository into a new directory:
 
-```bash
-# First instance
-git clone --depth 1 https://github.com/open-edge-platform/edge-ai-suites.git edge-ai-suites-instance1
-cd edge-ai-suites-instance1/metro-ai-suite/smart-traffic-intersection-agent/
-```
+    ```bash
+    git clone --depth 1 https://github.com/open-edge-platform/edge-ai-suites.git edge-ai-suites-instance1
+    cd edge-ai-suites-instance1/metro-ai-suite/smart-traffic-intersection-agent/
+    ```
 
-2. Edit the deployment configuration file for instance #1:
+2.  Edit the deployment configuration file for instance #1:
 
-```bash
-nano src/config/deployment_instance.json
-```
+    ```bash
+    nano src/config/deployment_instance.json
+    ```
 
-Update the `latitude` and `longitude` values as required. If not required, use the default
-values without updating this config file. Following is a sample value for the Instance #1
-deployment config:
-
-```json
-{
-    "name": "intersection_1",
-    "latitude": 37.7049108,
-    "longitude": -121.9096158,
-    "agent_backend_port": "8081",
-    "agent_ui_port": "7860"
-}
-```
+    Update `name`, `latitude` and `longitude` values as required. Following is a sample value for the Instance #1 deployment config:
 
-3. Run the setup for instance #1:
+    ```json
+    {
+        "name": "intersection_1",
+        "latitude": 37.5879818,
+        "longitude": -122.0534334,
+        "agent_backend_port": "8081",
+        "agent_ui_port": "7860"
+    }
+    ```
 
-```bash
-source setup.sh --setup
-```
+    > **TIPS:** Leave `agent_backend_port` and `agent_ui_port` empty to avoid port conflicts. Random ports would be assigned and application URLs with assigned ports will be shown when setup finishes.
+
+3.  Set the required environment variable.
+
+    ```bash
+    export VLM_MODEL_NAME=<supported_model_name>  # eg. microsoft/Phi-3.5-vision-instruct, Qwen/Qwen2.5-VL-3B-Instruct
+    ```
+    > **IMPORTANT:** See this [disclaimer](#disclaimer-for-using-third-party-ai-models) before using any AI Model.
+
+4.  Run the setup for instance #1:
+
+    ```bash
+    source setup.sh --setup
+    ```
 
 ### Set up Instance #2
 
-1. Clone the repository into a new directory:
+1.  Open a new terminal window and move to new directory. Clone the repository into the new directory:
 
-Open a new terminal window and move to different directory and run the following.
+    ```bash
+    git clone --depth 1 https://github.com/open-edge-platform/edge-ai-suites.git edge-ai-suites-instance2
+    cd edge-ai-suites-instance2/metro-ai-suite/smart-traffic-intersection-agent/
+    ```
 
-```bash
-git clone --depth 1 https://github.com/open-edge-platform/edge-ai-suites.git edge-ai-suites-instance2
-cd edge-ai-suites-instance2/metro-ai-suite/smart-traffic-intersection-agent/
-```
+2.  Edit the deployment configuration for instance #2:
 
-2. Edit the deployment configuration for instance #2:
+    ```bash
+    nano src/config/deployment_instance.json
+    ```
 
-```bash
-nano src/config/deployment_instance.json
-```
+    The following is a sample value for instance #2 deployment configuration:
 
-The following is a sample value for instance #2 deployment configuration:
+    ```json
+    {
+        "name": "intersection_2",
+        "latitude": 37.33874,
+        "longitude": -121.8852525,
+        "agent_backend_port": "8082",
+        "agent_ui_port": "7861"
+    }
+    ```
 
-```json
-{
-    "name": "intersection_2",
-    "latitude": 37.33874,
-    "longitude": -121.8852525,
-    "agent_backend_port": "8082",
-    "agent_ui_port": "7861"
-}
-```
-3. Run Setup for Instance #2
+    > **TIPS:** Leave `agent_backend_port` and `agent_ui_port` empty to avoid port conflicts. Random ports would be assigned and application URLs with assigned ports will be shown when setup finishes.
 
-```bash
-source setup.sh --setup
-```
+3.  Set the required environment variable.
+
+    ```bash
+    export VLM_MODEL_NAME=<supported_model_name>  # eg. microsoft/Phi-3.5-vision-instruct, Qwen/Qwen2.5-VL-3B-Instruct
+    ```
+    > **IMPORTANT:** See this [disclaimer](#disclaimer-for-using-third-party-ai-models) before using any AI Model. 
+
+4.  Run Setup for Instance #2
 
-> **Note:** Keep the `agent_backend_port` and `agent_ui_port` values empty to use random
-> ephemeral ports and avoid port conflicts.
+    ```bash
+    source setup.sh --setup
+    ```
 
 Ensure each instance has their `deployment_instance.json` updated with:
 
 - A unique value for `name` field
-- Unique latitude and longitude co-ordinates
+- Unique `latitude` and `longitude` co-ordinates
 - Different `agent_backend_port` and `agent_ui_port` values to avoid port conflicts. This is
 optional. If not specified, an ephemeral port is picked automatically.
 
@@ -191,6 +209,10 @@ are reached. Hence, deploy new instances only if you have the required resource
 To spin-up more instances - say `n number of new instances`, repeat the steps mentioned in
 [Set up Instance #2](#set-up-instance-2), by changing to a new directory `n` times.
 
+### Disclaimer for Using Third-Party AI Models
+
+Compliance with all license obligations and responsible use for a third-party AI Model is the user’s responsibility.
+
 ## Advanced Environment Configuration
 
 For advanced users who need more control over the configuration, you can configure the following environment variables before running the setup script to override the default behaviour:
@@ -202,9 +224,6 @@ export LOG_LEVEL=DEBUG
 # Select iGPU as the accelerator to perform VLM inference. By default, it is set to CPU 
 export VLM_DEVICE=GPU
 
-# Change the VLM Model name. Default value set in setup.sh is microsoft/Phi-3.5-vision-instruct.
-export VLM_MODEL_NAME=Qwen/Qwen2.5-VL-3B-Instruct
-
 # Other VLM related config, sample values
 export VLM_TIMEOUT_SECONDS=600          # Default 300
 export VLM_MAX_COMPLETION_TOKENS=1000   # Default 1500

metro-ai-suite/smart-traffic-intersection-agent/docs/user-guide/get-started/build-from-source.md

@@ -17,15 +17,22 @@ git clone https://github.com/open-edge-platform/edge-ai-suites.git
 cd edge-ai-suites/metro-ai-suite/smart-traffic-intersection-agent/
 ```
 
-### 2. Build the Docker Image
+### 2. Set the required environment variables
+
+```bash
+export VLM_MODEL_NAME=<supported_model_name>  # eg. microsoft/Phi-3.5-vision-instruct, Qwen/Qwen2.5-VL-3B-Instruct
+```
+> **IMPORTANT:** See this [disclaimer](../get-started.md#disclaimer-for-using-third-party-ai-models) before using any AI Model. 
+
+### 3. Build the Docker Image
 
 Build the Traffic Intersection Agent:
 
 ```bash
 docker compose -f docker/agent-compose.yaml build traffic-agent
 ```
 
-### 3. Run the Service
+### 4. Run the Service
 
 ```bash
 # Using setup script (recommended)

metro-ai-suite/smart-traffic-intersection-agent/setup.sh

@@ -127,6 +127,12 @@ fi
 # Dependencies: Setup Smart Intersection RI before running the agent Backend/UI
 # ============================================================================
 
+# Check if VLM Model name is set or not
+if [ -z "$VLM_MODEL_NAME" ]; then
+    echo -e "${RED}Error: VLM_MODEL_NAME environment variable is not set. Please check docs for some possible VLM model names.${NC}"
+    return 1
+fi
+
 export SAMPLE_APP="smart-intersection"
 SUBMODULE="deps/metro-vision"
 SUBMODULE_PATH="$APP_DIR/$SUBMODULE"
@@ -206,7 +212,7 @@ export VIDEO_GROUP_ID=$(getent group video | awk -F: '{printf "%s\n", $3}' 2>/de
 export RENDER_GROUP_ID=$(getent group render | awk -F: '{printf "%s\n", $3}' 2>/dev/null || echo "109")
 
 # VLM Service Configuration
-export VLM_MODEL_NAME=${VLM_MODEL_NAME:-microsoft/Phi-3.5-vision-instruct}
+export VLM_MODEL_NAME=${VLM_MODEL_NAME}
 
 # VLM OpenVINO Configuration
 export VLM_DEVICE=${VLM_DEVICE:-CPU}
@@ -263,11 +269,11 @@ print_all_service_host_endpoints() {
             *traffic-agent*)
                 BACKEND_SERVICE_NAME="Traffic Intersection Agent API Docs"
                 PORT=$(docker port "$CONTAINER_NAME" 8081 | cut -d: -f2)
-                echo -e "${BLUE}Access $BACKEND_SERVICE_NAME -> http://$HOST_IP:$PORT/docs${NC}"
+                echo -e "${CYAN}Access $BACKEND_SERVICE_NAME -> http://$HOST_IP:$PORT/docs${NC}"
 
                 UI_SERVICE_NAME="Traffic Intersection Agent UI"
                 PORT=$(docker port "$CONTAINER_NAME" 7860 | cut -d: -f2)
-                echo -e "${BLUE}Access $UI_SERVICE_NAME -> http://$HOST_IP:$PORT${NC}"
+                echo -e "${CYAN}Access $UI_SERVICE_NAME -> http://$HOST_IP:$PORT${NC}"
                 ;;
             *vlm*)
                 SERVICE_NAME="VLM OpenVINO Serving API"

metro-ai-suite/smart-traffic-intersection-agent/src/config/deployment_instance.json

@@ -1,7 +1,7 @@
 {
     "name": "intersection_1",
-    "latitude": 33.3091336,
-    "longitude": -111.9353095,
+    "latitude": 37.51358,
+    "longitude": -122.25591,
     "agent_backend_port": "",
     "agent_ui_port": ""
 }

metro-ai-suite/smart-traffic-intersection-agent/src/docker-entrypoint.sh

@@ -15,7 +15,6 @@ echo "========================================"
 export REFRESH_INTERVAL=${REFRESH_INTERVAL:-15}
 export LOG_LEVEL=${LOG_LEVEL:-INFO}
 export WEATHER_MOCK=${WEATHER_MOCK:-false}
-export VLM_MODEL=${VLM_MODEL:-Qwen/Qwen2.5-VL-3B-Instruct}
 export HIGH_DENSITY_THRESHOLD=${HIGH_DENSITY_THRESHOLD:-10}
 export USE_API=${USE_API:-true}
 
@@ -26,7 +25,7 @@ echo "INTERSECTION_NAME: ${INTERSECTION_NAME}"
 echo "INTERSECTION_LATITUDE: ${INTERSECTION_LATITUDE}"
 echo "INTERSECTION_LONGITUDE: ${INTERSECTION_LONGITUDE}"
 echo "WEATHER_MOCK: ${WEATHER_MOCK}"
-echo "VLM_MODEL: ${VLM_MODEL}"
+echo "VLM_MODEL_NAME: ${VLM_MODEL_NAME}"
 echo "HIGH_DENSITY_THRESHOLD: ${HIGH_DENSITY_THRESHOLD}"
 echo "========================================"