more finishReason fixes

src/leapfrogai_api/backend/grpc_client.py

@@ -48,7 +48,7 @@ async def completion(model: Model, request: lfai.CompletionRequest):
                 CompletionChoice(
                     index=0,
                     text=response.choices[0].text,
-                    finish_reason=str(response.choices[0].finish_reason),
+                    finish_reason=response.choices[0].finish_reason,
                     logprobs=None,
                 )
             ],

src/leapfrogai_api/backend/types.py

@@ -131,8 +131,8 @@ class CompletionChoice(BaseModel):
         None,
         description="Log probabilities for the generated tokens. Only returned if requested.",
     )
-    finish_reason: str = Field(
-        "", description="The reason why the completion finished.", example="length"
+    finish_reason: str | None = Field(
+        None, description="The reason why the completion finished.", example="length"
     )
 
 
@@ -252,7 +252,7 @@ class ChatChoice(BaseModel):
         default=ChatMessage(), description="The message content for this choice."
     )
     finish_reason: str | None = Field(
-        default="",
+        default=None,
         description="The reason why the model stopped generating tokens.",
         examples=["stop", "length"],
     )

src/leapfrogai_sdk/llm.py

@@ -21,7 +21,7 @@
 
 
 class FinishReason(Enum):
-    NONE = ""
+    NONE = None
     STOP = "stop"
     LENGTH = "length"
 

tests/pytest/leapfrogai_api/test_api.py

@@ -260,7 +260,7 @@ def test_chat_completion(dummy_auth_middleware):
 
         # parse finish reason
         assert "finish_reason" in response_choices[0]
-        assert "STOP" == response_choices[0].get("finish_reason")
+        assert "stop" == response_choices[0].get("finish_reason")
 
         # parse usage data
         response_usage = response_obj.get("usage")
@@ -324,9 +324,9 @@ def test_stream_chat_completion(dummy_auth_middleware):
                     assert "finish_reason" in choices[0]
                     # in streaming responses, the stop reason is not STOP until the last iteration (token) is sent back
                     if iter_length == input_length:
-                        assert "STOP" == choices[0].get("finish_reason")
+                        assert "stop" == choices[0].get("finish_reason")
                     else:
-                        assert "NONE" == choices[0].get("finish_reason")
+                        assert None is choices[0].get("finish_reason")
                     # parse usage data
                     response_usage = stream_response.get("usage")
                     prompt_tokens = response_usage.get("prompt_tokens")