fix formatting issues

Author: estohlmann

.pre-commit-config.yaml

@@ -55,7 +55,7 @@ repos:
       name: isort (python)
 
 - repo: https://github.com/ambv/black
-  rev: '23.10.1'
+  rev: '24.10.0'
   hooks:
     - id: black
 
@@ -66,21 +66,19 @@ repos:
       args: [--exit-non-zero-on-fix]
 
 - repo: https://github.com/pycqa/flake8
-  rev: '6.1.0'
+  rev: '7.1.1'
   hooks:
     - id: flake8
       additional_dependencies:
-        - flake8-docstrings
         - flake8-broken-line
         - flake8-bugbear
         - flake8-comprehensions
         - flake8-debugger
         - flake8-string-format
       args:
-        - --docstring-convention=numpy
         - --max-line-length=120
         - --extend-immutable-calls=Query,fastapi.Depends,fastapi.params.Depends
-        - --ignore=B008 # Ignore error for function calls in argument defaults
+        - --ignore=B008,E203 # Ignore error for function calls in argument defaults
       exclude: ^(__init__.py$|.*\/__init__.py$)
 
 

lambda/models/domain_objects.py

@@ -98,7 +98,7 @@ class AutoScalingConfig(BaseModel):
     defaultInstanceWarmup: PositiveInt
     metricConfig: MetricConfig
 
-    @model_validator(mode="after")  # type: ignore
+    @model_validator(mode="after")
     def validate_auto_scaling_config(self) -> Self:
         """Validate autoScalingConfig values."""
         if self.minCapacity > self.maxCapacity:
@@ -115,7 +115,7 @@ class AutoScalingInstanceConfig(BaseModel):
     maxCapacity: Optional[PositiveInt] = None
     desiredCapacity: Optional[PositiveInt] = None
 
-    @model_validator(mode="after")  # type: ignore
+    @model_validator(mode="after")
     def validate_auto_scaling_instance_config(self) -> Self:
         """Validate autoScalingInstanceConfig values."""
         config_fields = [self.minCapacity, self.maxCapacity, self.desiredCapacity]
@@ -155,7 +155,7 @@ class ContainerConfig(BaseModel):
     healthCheckConfig: ContainerHealthCheckConfig
     environment: Optional[Dict[str, str]] = {}
 
-    @field_validator("environment")  # type: ignore
+    @field_validator("environment")
     @classmethod
     def validate_environment(cls, environment: Dict[str, str]) -> Dict[str, str]:
         """Validate that all keys in Dict are not empty."""
@@ -201,7 +201,7 @@ class CreateModelRequest(BaseModel):
     modelUrl: Optional[str] = None
     streaming: Optional[bool] = False
 
-    @model_validator(mode="after")  # type: ignore
+    @model_validator(mode="after")
     def validate_create_model_request(self) -> Self:
         """Validate whole request object."""
         # Validate that an embedding model cannot be set as streaming-enabled
@@ -252,7 +252,7 @@ class UpdateModelRequest(BaseModel):
     modelType: Optional[ModelType] = None
     streaming: Optional[bool] = None
 
-    @model_validator(mode="after")  # type: ignore
+    @model_validator(mode="after")
     def validate_update_model_request(self) -> Self:
         """Validate whole request object."""
         fields = [
@@ -273,7 +273,7 @@ def validate_update_model_request(self) -> Self:
             raise ValueError("Embedding model cannot be set with streaming enabled.")
         return self
 
-    @field_validator("autoScalingInstanceConfig")  # type: ignore
+    @field_validator("autoScalingInstanceConfig")
     @classmethod
     def validate_autoscaling_instance_config(cls, config: AutoScalingInstanceConfig) -> AutoScalingInstanceConfig:
         """Validate that the AutoScaling instance config has at least one positive value."""

lambda/models/lambda_functions.py

@@ -59,32 +59,32 @@
 stepfunctions = boto3.client("stepfunctions", region_name=os.environ["AWS_REGION"], config=retry_config)
 
 
-@app.exception_handler(ModelNotFoundError)  # type: ignore
+@app.exception_handler(ModelNotFoundError)
 async def model_not_found_handler(request: Request, exc: ModelNotFoundError) -> JSONResponse:
     """Handle exception when model cannot be found and translate to a 404 error."""
     return JSONResponse(status_code=404, content={"message": str(exc)})
 
 
-@app.exception_handler(RequestValidationError)  # type: ignore
-async def validation_exception_handler(request: Request, exc: RequestValidationError):
+@app.exception_handler(RequestValidationError)
+async def validation_exception_handler(request: Request, exc: RequestValidationError) -> JSONResponse:
     """Handle exception when request fails validation and and translate to a 422 error."""
     return JSONResponse(
         status_code=422, content={"detail": jsonable_encoder(exc.errors()), "type": "RequestValidationError"}
     )
 
 
-@app.exception_handler(InvalidStateTransitionError)  # type: ignore
-@app.exception_handler(ModelAlreadyExistsError)  # type: ignore
-@app.exception_handler(ValueError)  # type: ignore
+@app.exception_handler(InvalidStateTransitionError)
+@app.exception_handler(ModelAlreadyExistsError)
+@app.exception_handler(ValueError)
 async def user_error_handler(
     request: Request, exc: Union[InvalidStateTransitionError, ModelAlreadyExistsError, ValueError]
 ) -> JSONResponse:
     """Handle errors when customer requests options that cannot be processed."""
     return JSONResponse(status_code=400, content={"message": str(exc)})
 
 
-@app.post(path="", include_in_schema=False)  # type: ignore
-@app.post(path="/")  # type: ignore
+@app.post(path="", include_in_schema=False)
+@app.post(path="/")
 async def create_model(create_request: CreateModelRequest) -> CreateModelResponse:
     """Endpoint to create a model."""
     create_handler = CreateModelHandler(
@@ -95,8 +95,8 @@ async def create_model(create_request: CreateModelRequest) -> CreateModelRespons
     return create_handler(create_request=create_request)
 
 
-@app.get(path="", include_in_schema=False)  # type: ignore
-@app.get(path="/")  # type: ignore
+@app.get(path="", include_in_schema=False)
+@app.get(path="/")
 async def list_models() -> ListModelsResponse:
     """Endpoint to list models."""
     list_handler = ListModelsHandler(
@@ -107,7 +107,7 @@ async def list_models() -> ListModelsResponse:
     return list_handler()
 
 
-@app.get(path="/{model_id}")  # type: ignore
+@app.get(path="/{model_id}")
 async def get_model(
     model_id: Annotated[str, Path(title="The unique model ID of the model to get")], request: Request
 ) -> GetModelResponse:
@@ -120,7 +120,7 @@ async def get_model(
     return get_handler(model_id=model_id)
 
 
-@app.put(path="/{model_id}")  # type: ignore
+@app.put(path="/{model_id}")
 async def update_model(
     model_id: Annotated[str, Path(title="The unique model ID of the model to update")],
     update_request: UpdateModelRequest,
@@ -134,7 +134,7 @@ async def update_model(
     return update_handler(model_id=model_id, update_request=update_request)
 
 
-@app.delete(path="/{model_id}")  # type: ignore
+@app.delete(path="/{model_id}")
 async def delete_model(
     model_id: Annotated[str, Path(title="The unique model ID of the model to delete")], request: Request
 ) -> DeleteModelResponse:
@@ -147,7 +147,7 @@ async def delete_model(
     return delete_handler(model_id=model_id)
 
 
-@app.get(path="/metadata/instances")  # type: ignore
+@app.get(path="/metadata/instances")
 async def get_instances() -> list[str]:
     """Endpoint to list available instances in this region."""
     return list(sess.get_service_model("ec2").shape_for("InstanceType").enum)

lib/serve/rest-api/src/api/endpoints/v1/embeddings.py

@@ -27,7 +27,7 @@
 router = APIRouter()
 
 
-@router.post(f"/{RestApiResource.EMBEDDINGS.value}")  # type: ignore
+@router.post(f"/{RestApiResource.EMBEDDINGS.value}")
 async def embeddings(request: EmbeddingsRequest) -> JSONResponse:
     """Text embeddings."""
     response = await handle_embeddings(request.dict())

lib/serve/rest-api/src/api/endpoints/v1/generation.py

@@ -33,15 +33,15 @@
 router = APIRouter()
 
 
-@router.post(f"/{RestApiResource.GENERATE.value}")  # type: ignore
+@router.post(f"/{RestApiResource.GENERATE.value}")
 async def generate(request: GenerateRequest) -> JSONResponse:
     """Text generation."""
     response = await handle_generate(request.dict())
 
     return JSONResponse(content=response, status_code=200)
 
 
-@router.post(f"/{RestApiResource.GENERATE_STREAM.value}")  # type: ignore
+@router.post(f"/{RestApiResource.GENERATE_STREAM.value}")
 async def generate_stream(request: GenerateStreamRequest) -> StreamingResponse:
     """Text generation with streaming."""
     return StreamingResponse(
@@ -50,7 +50,7 @@ async def generate_stream(request: GenerateStreamRequest) -> StreamingResponse:
     )
 
 
-@router.post(f"/{RestApiResource.OPENAI_CHAT_COMPLETIONS.value}")  # type: ignore
+@router.post(f"/{RestApiResource.OPENAI_CHAT_COMPLETIONS.value}")
 async def openai_chat_completion_generate_stream(request: OpenAIChatCompletionsRequest) -> StreamingResponse:
     """Text generation with streaming."""
     return StreamingResponse(
@@ -59,7 +59,7 @@ async def openai_chat_completion_generate_stream(request: OpenAIChatCompletionsR
     )
 
 
-@router.post(f"/{RestApiResource.OPENAI_COMPLETIONS.value}")  # type: ignore
+@router.post(f"/{RestApiResource.OPENAI_COMPLETIONS.value}")
 async def openai_completion_generate_stream(request: OpenAICompletionsRequest) -> StreamingResponse:
     """Text generation with streaming."""
     return StreamingResponse(

lib/serve/rest-api/src/api/endpoints/v1/models.py

@@ -33,7 +33,7 @@
 router = APIRouter()
 
 
-@router.get(f"/{RestApiResource.DESCRIBE_MODEL.value}")  # type: ignore
+@router.get(f"/{RestApiResource.DESCRIBE_MODEL.value}")
 async def describe_model(
     provider: str = Query(
         None,
@@ -52,7 +52,7 @@ async def describe_model(
     return JSONResponse(content=response, status_code=200)
 
 
-@router.get(f"/{RestApiResource.DESCRIBE_MODELS.value}")  # type: ignore
+@router.get(f"/{RestApiResource.DESCRIBE_MODELS.value}")
 async def describe_models(
     model_types: Optional[List[ModelType]] = Query(
         None,
@@ -69,7 +69,7 @@ async def describe_models(
     return JSONResponse(content=response, status_code=200)
 
 
-@router.get(f"/{RestApiResource.LIST_MODELS.value}")  # type: ignore
+@router.get(f"/{RestApiResource.LIST_MODELS.value}")
 async def list_models(
     model_types: Optional[List[ModelType]] = Query(
         None,
@@ -86,7 +86,7 @@ async def list_models(
     return JSONResponse(content=response, status_code=200)
 
 
-@router.get(f"/{RestApiResource.OPENAI_LIST_MODELS.value}")  # type: ignore
+@router.get(f"/{RestApiResource.OPENAI_LIST_MODELS.value}")
 async def openai_list_models() -> JSONResponse:
     """List models for OpenAI Compatibility. Only returns TEXTGEN models."""
     response = await handle_openai_list_models()

lib/serve/rest-api/src/api/endpoints/v2/litellm_passthrough.py

@@ -82,9 +82,7 @@ def generate_response(iterator: Iterator[Union[str, bytes]]) -> Iterator[str]:
             yield f"{line}\n\n"
 
 
-@router.api_route(
-    "/{api_path:path}", methods=["GET", "POST", "OPTIONS", "PUT", "PATCH", "DELETE", "HEAD"]
-)  # type: ignore
+@router.api_route("/{api_path:path}", methods=["GET", "POST", "OPTIONS", "PUT", "PATCH", "DELETE", "HEAD"])
 async def litellm_passthrough(request: Request, api_path: str) -> Response:
     """
     Pass requests directly to LiteLLM. LiteLLM and deployed models will respond here directly.

lib/serve/rest-api/src/api/routes.py

@@ -40,7 +40,7 @@
 )
 
 
-@router.get("/health")  # type: ignore
+@router.get("/health")
 async def health_check() -> JSONResponse:
     """Health check path.
 

lib/serve/rest-api/src/handlers/generation.py

@@ -31,7 +31,7 @@ async def handle_generate(request_data: Dict[str, Any]) -> Dict[str, Any]:
     return response.dict()  # type: ignore
 
 
-@handle_stream_exceptions  # type: ignore
+@handle_stream_exceptions
 async def handle_generate_stream(request_data: Dict[str, Any]) -> AsyncGenerator[str, None]:
     """Handle for generate_stream endpoint."""
     model, model_kwargs, text = await validate_and_prepare_llm_request(request_data, RestApiResource.GENERATE_STREAM)
@@ -57,7 +57,7 @@ def parse_model_provider_names(model_string: str) -> Tuple[str, str]:
     return model_name, provider
 
 
-@handle_stream_exceptions  # type: ignore
+@handle_stream_exceptions
 async def handle_openai_generate_stream(
     request_data: Dict[str, Any], is_text_completion: bool = False
 ) -> AsyncGenerator[str, None]:

lib/serve/rest-api/src/lisa_serve/ecs/textgen/tgi.py

@@ -211,16 +211,18 @@ async def openai_generate_stream(
                 object="text_completion" if is_text_completion else "chat.completion.chunk",
                 system_fingerprint=fingerprint,
                 choices=[
-                    OpenAICompletionsChoice(
-                        index=0,
-                        finish_reason=resp.details.finish_reason if resp.details else None,
-                        text=resp.token.text,
-                    )
-                    if is_text_completion
-                    else OpenAIChatCompletionsChoice(
-                        index=0,
-                        finish_reason=resp.details.finish_reason if resp.details else None,
-                        delta=OpenAIChatCompletionsDelta(content=resp.token.text, role="assistant"),
+                    (
+                        OpenAICompletionsChoice(
+                            index=0,
+                            finish_reason=resp.details.finish_reason if resp.details else None,
+                            text=resp.token.text,
+                        )
+                        if is_text_completion
+                        else OpenAIChatCompletionsChoice(
+                            index=0,
+                            finish_reason=resp.details.finish_reason if resp.details else None,
+                            delta=OpenAIChatCompletionsDelta(content=resp.token.text, role="assistant"),
+                        )
                     )
                 ],
             )

lib/serve/rest-api/src/main.py

@@ -147,7 +147,7 @@ async def lifespan(app: FastAPI):  # type: ignore
 ##############
 
 
-@app.middleware("http")  # type: ignore
+@app.middleware("http")
 async def process_request(request: Request, call_next: Any) -> Any:
     """Middleware for processing all HTTP requests."""
     event = "process_request"

lib/serve/rest-api/src/utils/generate_litellm_config.py

@@ -25,10 +25,8 @@
 secrets_client = boto3.client("secretsmanager", region_name=os.environ["AWS_REGION"])
 
 
-@click.command()  # type: ignore
-@click.option(
-    "-f", "--filepath", type=click.Path(exists=True, file_okay=True, dir_okay=False, writable=True)
-)  # type: ignore
+@click.command()
+@click.option("-f", "--filepath", type=click.Path(exists=True, file_okay=True, dir_okay=False, writable=True))
 def generate_config(filepath: str) -> None:
     """Read LiteLLM configuration and rewrite it with LISA-deployed model information."""
     with open(filepath, "r") as fp:

lisa-sdk/lisapy/main.py

@@ -51,7 +51,7 @@ class Lisa(BaseModel):
 
     _session: Session
 
-    @field_validator("url")  # type: ignore
+    @field_validator("url")
     def validate_url(cls: "Lisa", v: str) -> str:
         """Validate URL is properly formatted."""
         url = v.rstrip("/")

lisa-sdk/tests/test_client.py

@@ -22,13 +22,13 @@
 from lisapy.types import ModelKwargs, ModelType
 
 
-@pytest.fixture(scope="session")  # type: ignore
+@pytest.fixture(scope="session")
 def url(pytestconfig: pytest.Config) -> Any:
     """Get the url argument."""
     return pytestconfig.getoption("url")
 
 
-@pytest.fixture(scope="session")  # type: ignore
+@pytest.fixture(scope="session")
 def verify(pytestconfig: pytest.Config) -> Union[bool, Any]:
     """Get the verify argument."""
     if pytestconfig.getoption("verify") == "false":
@@ -114,7 +114,7 @@ def test_generate_stream(url: str, verify: Union[bool, str]) -> None:
     assert response.generated_tokens == 1
 
 
-@pytest.mark.asyncio  # type: ignore
+@pytest.mark.asyncio
 async def test_generate_async(url: str, verify: Union[bool, str]) -> None:
     """Generates a batch async response from a textgen.tgi model."""
     client = Lisa(url=url, verify=verify)
@@ -127,7 +127,7 @@ async def test_generate_async(url: str, verify: Union[bool, str]) -> None:
     assert response.generated_tokens == 1
 
 
-@pytest.mark.asyncio  # type: ignore
+@pytest.mark.asyncio
 async def test_generate_stream_async(url: str, verify: Union[bool, str]) -> None:
     """Generates a streaming async response from a textgen.tgi model."""
     client = Lisa(url=url, verify=verify)

pyproject.toml

@@ -27,7 +27,7 @@ skip_glob = [
 [tool.mypy]
 ignore_missing_imports = true
 disallow_untyped_defs = true
-disallow_untyped_decorators = true
+disallow_untyped_decorators = false
 disallow_incomplete_defs = true
 disallow_any_unimported = false
 no_implicit_optional = true