Merge branch '561-long-lived-api-keys' of github.com:defenseunicorns/…

…leapfrogai into 561-long-lived-api-keys

src/leapfrogai_api/backend/grpc_client.py

@@ -30,7 +30,7 @@ async def stream_completion(model: Model, request: lfai.CompletionRequest):
 
         await stream.wait_for_connection()
         return StreamingResponse(
-            recv_completion(stream), media_type="text/event-stream"
+            recv_completion(stream, model.name), media_type="text/event-stream"
         )
 
 
@@ -66,7 +66,9 @@ async def stream_chat_completion(model: Model, request: lfai.ChatCompletionReque
         stream = stub.ChatCompleteStream(request)
 
         await stream.wait_for_connection()
-        return StreamingResponse(recv_chat(stream), media_type="text/event-stream")
+        return StreamingResponse(
+            recv_chat(stream, model.name), media_type="text/event-stream"
+        )
 
 
 async def stream_chat_completion_raw(

src/leapfrogai_api/backend/helpers.py

@@ -1,5 +1,7 @@
 """Helper functions for the OpenAI backend."""
 
+import time
+import uuid
 from typing import BinaryIO, Iterator, AsyncGenerator, Any
 import grpc
 import leapfrogai_sdk as lfai
@@ -15,15 +17,16 @@
 
 async def recv_completion(
     stream: grpc.aio.UnaryStreamCall[lfai.CompletionRequest, lfai.CompletionResponse],
+    model: str,
 ):
     async for c in stream:
         yield (
             "data: "
             + CompletionResponse(
-                id="foo",
+                id=str(uuid.uuid4()),
                 object="completion.chunk",
-                created=55,
-                model="mpt-7b-8k-chat",
+                created=int(time.time()),
+                model=model,
                 choices=[
                     CompletionChoice(
                         index=0,
@@ -48,16 +51,17 @@ async def recv_chat(
     stream: grpc.aio.UnaryStreamCall[
         lfai.ChatCompletionRequest, lfai.ChatCompletionResponse
     ],
+    model: str,
 ) -> AsyncGenerator[str, Any]:
     """Generator that yields chat completion responses as Server-Sent Events."""
     async for c in stream:
         yield (
             "data: "
             + ChatCompletionResponse(
-                id="foo",
+                id=str(uuid.uuid4()),
                 object="chat.completion.chunk",
-                created=55,
-                model="mpt-7b-8k-chat",
+                created=int(time.time()),
+                model=model,
                 choices=[
                     ChatStreamChoice(
                         index=0,