retry e2e tests

e2e_tests/poetry.lock

@@ -493,6 +493,23 @@ pytest = ">=4.6"
 [package.extras]
 testing = ["fields", "hunter", "process-tests", "pytest-xdist", "six", "virtualenv"]
 
+[[package]]
+name = "pytest-retry"
+version = "1.6.3"
+description = "Adds the ability to retry flaky tests in CI environments"
+optional = false
+python-versions = ">=3.9"
+files = [
+    {file = "pytest_retry-1.6.3-py3-none-any.whl", hash = "sha256:e96f7df77ee70b0838d1085f9c3b8b5b7d74bf8947a0baf32e2b8c71b27683c8"},
+    {file = "pytest_retry-1.6.3.tar.gz", hash = "sha256:36ccfa11c8c8f9ddad5e20375182146d040c20c4a791745139c5a99ddf1b557d"},
+]
+
+[package.dependencies]
+pytest = ">=7.0.0"
+
+[package.extras]
+dev = ["black", "flake8", "isort", "mypy"]
+
 [[package]]
 name = "pytest-sugar"
 version = "1.0.0"
@@ -699,4 +716,4 @@ h11 = ">=0.9.0,<1"
 [metadata]
 lock-version = "2.0"
 python-versions = "^3.10"
-content-hash = "6ae4fa6397091b87b63698201a08d7d97628ed65992d46514f118768b46b99ce"
+content-hash = "a5a13987eab6bb15765c050d6a7aaf3b0d0cf661eedb129d6b811d654880a3cb"

e2e_tests/pyproject.toml

@@ -14,6 +14,7 @@ requests = "^2.29.0"
 selenium = "^4.11.2"
 pytest-sugar = "^1.0.0"
 deepdiff = "^8.0.1"
+pytest-retry = "^1.6.3"
 
 [tool.poetry.dev-dependencies]
 pytest-cov = "^4.1.0"
@@ -21,3 +22,6 @@ pytest-cov = "^4.1.0"
 [tool.pytest.ini_options]
 python_files = ["test*.py"]
 addopts = ["-v", "-s"]
+retries = 3
+retry_delay = 0.5
+cumulative_timing = false

model_server/.vscode/launch.json

@@ -9,7 +9,7 @@
       "type": "debugpy",
       "request": "launch",
       "module": "uvicorn",
-      "args": ["app.main:app","--reload", "--port", "8000"]
+      "args": ["app.main:app","--reload", "--port", "51000"]
     }
   ]
 }

model_server/app/function_calling/model_utils.py

@@ -67,12 +67,16 @@ async def chat_completion(req: ChatMessage, res: Response):
         f"model_server => arch_function: {client_model_name}, messages: {json.dumps(messages)}"
     )
 
-    resp = const.arch_function_client.chat.completions.create(
-        messages=messages,
-        model=client_model_name,
-        stream=False,
-        extra_body=const.arch_function_generation_params,
-    )
+    try:
+        resp = const.arch_function_client.chat.completions.create(
+            messages=messages,
+            model=client_model_name,
+            stream=False,
+            extra_body=const.arch_function_generation_params,
+        )
+    except Exception as e:
+        logger.error(f"model_server <= arch_function: error: {e}")
+        raise
 
     tool_calls = const.arch_function_hanlder.extract_tool_calls(
         resp.choices[0].message.content