run integration test for each inference engine

.circleci/config.yml

@@ -3,6 +3,87 @@ version: 2.1
 orbs:
   python: circleci/python@2
 
+commands:
+  run_chatgpt_api_test:
+    parameters:
+      inference_engine:
+        type: string
+    steps:
+      - run:
+          name: Run chatgpt api integration test (<<parameters.inference_engine>>)
+          command: |
+            source env/bin/activate
+
+            # Start first instance
+            DEBUG_DISCOVERY=9 DEBUG=9 python3 main.py --inference-engine <<parameters.inference_engine>> --node-id "node1" --listen-port 5678 --broadcast-port 5679 --chatgpt-api-port 8000 --chatgpt-api-response-timeout-secs 900 > output1.log 2>&1 &
+            PID1=$!
+
+            # Start second instance
+            DEBUG_DISCOVERY=9 DEBUG=9 python3 main.py --inference-engine <<parameters.inference_engine>> --node-id "node2" --listen-port 5679 --broadcast-port 5678 --chatgpt-api-port 8001 --chatgpt-api-response-timeout-secs 900 > output2.log 2>&1 &
+            PID2=$!
+
+            # Wait for discovery
+            sleep 10
+
+            # Function to check if processes are still running
+            check_processes() {
+              if ! kill -0 $PID1 2>/dev/null; then
+                echo "First instance (PID $PID1) died unexpectedly. Log output:"
+                cat output1.log
+                exit 1
+              fi
+              if ! kill -0 $PID2 2>/dev/null; then
+                echo "Second instance (PID $PID2) died unexpectedly. Log output:"
+                cat output2.log
+                exit 1
+              fi
+            }
+
+            # Check processes before proceeding
+            check_processes
+
+            response_1=$(curl -s http://localhost:8000/v1/chat/completions \
+              -H "Content-Type: application/json" \
+              -d '{
+                "model": "llama-3.1-8b",
+                "messages": [{"role": "user", "content": "Keep responses concise. Who was the king of pop?"}],
+                "temperature": 0.7
+              }')
+            echo "Response 1: $response_1"
+
+            # Check processes after first response
+            check_processes
+
+            response_2=$(curl -s http://localhost:8001/v1/chat/completions \
+              -H "Content-Type: application/json" \
+              -d '{
+                "model": "llama-3.1-8b",
+                "messages": [{"role": "user", "content": "Keep responses concise. Who was the king of pop?"}],
+                "temperature": 0.7
+              }')
+            echo "Response 2: $response_2"
+
+            # Check processes after second response
+            check_processes
+
+            # Stop both instances
+            kill $PID1 $PID2
+
+            echo ""
+            if ! echo "$response_1" | grep -q "Michael Jackson" || ! echo "$response_2" | grep -q "Michael Jackson"; then
+              echo "Test failed: Response does not contain 'Michael Jackson'"
+              echo "Response 1: $response_1"
+              echo ""
+              echo "Response 2: $response_2"
+              echo "Output of first instance:"
+              cat output1.log
+              echo "Output of second instance:"
+              cat output2.log
+              exit 1
+            else
+              echo "Test passed: Response from both nodes contains 'Michael Jackson'"
+            fi
+
 jobs:
   unit_test:
     macos:
@@ -84,80 +165,10 @@ jobs:
             source env/bin/activate
             pip install --upgrade pip
             pip install .
-      - run:
-          name: Run chatgpt api integration test
-          command: |
-            source env/bin/activate
-
-            # Start first instance
-            DEBUG_DISCOVERY=9 DEBUG=9 python3 main.py --inference-engine mlx --node-id "node1" --listen-port 5678 --broadcast-port 5679 --chatgpt-api-port 8000 --chatgpt-api-response-timeout-secs 900 > output1.log 2>&1 &
-            PID1=$!
-
-            # Start second instance
-            DEBUG_DISCOVERY=9 DEBUG=9 python3 main.py --inference-engine mlx --node-id "node2" --listen-port 5679 --broadcast-port 5678 --chatgpt-api-port 8001 --chatgpt-api-response-timeout-secs 900 > output2.log 2>&1 &
-            PID2=$!
-
-            # Wait for discovery
-            sleep 10
-
-            # Function to check if processes are still running
-            check_processes() {
-              if ! kill -0 $PID1 2>/dev/null; then
-                echo "First instance (PID $PID1) died unexpectedly. Log output:"
-                cat output1.log
-                exit 1
-              fi
-              if ! kill -0 $PID2 2>/dev/null; then
-                echo "Second instance (PID $PID2) died unexpectedly. Log output:"
-                cat output2.log
-                exit 1
-              fi
-            }
-
-            # Check processes before proceeding
-            check_processes
-
-            response_1=$(curl -s http://localhost:8000/v1/chat/completions \
-              -H "Content-Type: application/json" \
-              -d '{
-                "model": "llama-3.1-8b",
-                "messages": [{"role": "user", "content": "Keep responses concise. Who was the king of pop?"}],
-                "temperature": 0.7
-              }')
-            echo "Response 1: $response_1"
-
-            # Check processes after first response
-            check_processes
-
-            response_2=$(curl -s http://localhost:8001/v1/chat/completions \
-              -H "Content-Type: application/json" \
-              -d '{
-                "model": "llama-3.1-8b",
-                "messages": [{"role": "user", "content": "Keep responses concise. Who was the king of pop?"}],
-                "temperature": 0.7
-              }')
-            echo "Response 2: $response_2"
-
-            # Check processes after second response
-            check_processes
-
-            # Stop both instances
-            kill $PID1 $PID2
-
-            echo ""
-            if ! echo "$response_1" | grep -q "Michael Jackson" || ! echo "$response_2" | grep -q "Michael Jackson"; then
-              echo "Test failed: Response does not contain 'Michael Jackson'"
-              echo "Response 1: $response_1"
-              echo ""
-              echo "Response 2: $response_2"
-              echo "Output of first instance:"
-              cat output1.log
-              echo "Output of second instance:"
-              cat output2.log
-              exit 1
-            else
-              echo "Test passed: Response from both nodes contains 'Michael Jackson'"
-            fi
+      - run_chatgpt_api_test:
+          inference_engine: mlx
+      - run_chatgpt_api_test:
+          inference_engine: tinygrad
 
 workflows:
   version: 2