fix(config): Add some missing config settings to the config schema

Author: stronk7

env_template

@@ -43,4 +43,4 @@ AUTH_URL="http://0.0.0.0:4000/key/info"
 # LANGSMITH_API_KEY="<<<YOUR_LANGSMITH_API_KEY_HERE>>>
 
 # User agent for the crawler.
-USER_AGENT="Moodle Research Wiki-RAG Crawler/{version} (https://github.com/moodlehq/wiwki-rag)"
+USER_AGENT="Moodle Research Wiki-RAG Crawler/{version} (https://github.com/moodlehq/wiki-rag)"

wiki_rag/search/main.py

@@ -25,7 +25,7 @@
 
 
 async def run():
-    """Make an index from the json information present in the specified file."""
+    """Perform a search with all the configuration in place."""
     setup_logging(level=LOG_LEVEL)
     logger = logging.getLogger(__name__)
     logger.info("wiki_rag-search starting up...")
@@ -136,6 +136,7 @@ async def run():
         stream=stream,
         wrapper_chat_max_turns=0,
         wrapper_chat_max_tokens=0,
+        wrapper_model_name=llm_model,
     ).items()
 
     # Prepare the configuration.

wiki_rag/search/util.py

@@ -47,8 +47,8 @@ class ConfigSchema(TypedDict):
     llm_model: str
     search_distance_cutoff: float
     max_completion_tokens: int
-    top_p: float
     temperature: float
+    top_p: float
     stream: bool
     wrapper_chat_max_turns: int
     wrapper_chat_max_tokens: int

wiki_rag/server/main.py

@@ -22,7 +22,7 @@
 
 
 def main():
-    """Make an index from the json information present in the specified file."""
+    """Run the OpenAI server with all the configuration in place."""
     setup_logging(level=LOG_LEVEL)
     logger = logging.getLogger(__name__)
     logger.info("wiki_rag-server starting up...")

wiki_rag/server/server.py

@@ -8,9 +8,12 @@
 import time
 import uuid
 
+from typing import Any
+
 from fastapi import Depends, FastAPI, HTTPException
 from fastapi.responses import StreamingResponse
 from langchain_core.messages import AIMessageChunk, BaseMessage
+from langchain_core.runnables import RunnableConfig
 
 from wiki_rag import __version__, server
 from wiki_rag.server.util import (
@@ -182,9 +185,10 @@ async def open_ai_langgraph_stream(question: str, history: list[BaseMessage]):
         return StreamingResponse(open_ai_langgraph_stream(question, history), media_type="text/event-stream")
     else:
         logger.info("Running the search (non-streaming)")
-        completion = await server.graph.ainvoke(
-            {"question": question, "history": history},
-            config=server.config
+        completion = await invoke_graph(
+            question=question,
+            history=history,
+            config=server.config,
         )
 
         return ChatCompletionResponse(
@@ -199,3 +203,16 @@ async def open_ai_langgraph_stream(question: str, history: list[BaseMessage]):
                 ),
             ],
         )
+
+
+async def invoke_graph(
+    question: str,
+    history: list[BaseMessage],
+    config: RunnableConfig,
+) -> Any:
+    """Invoke the graph with the given question, history and configuration. No streaming."""
+    assert server.graph is not None
+    return await server.graph.ainvoke(
+        {"question": question, "history": history},
+        config=config
+    )