Add rerank_model to API interface

src/rca_accelerator_chatbot/api.py

@@ -18,7 +18,7 @@
 from rca_accelerator_chatbot.settings import ModelSettings
 from rca_accelerator_chatbot.auth import authentification
 from rca_accelerator_chatbot.models import (
-    gen_model_provider, embed_model_provider, init_model_providers
+    gen_model_provider, embed_model_provider, rerank_model_provider, init_model_providers
 )
 
 app = FastAPI(title="RCAccelerator API")
@@ -34,6 +34,7 @@ class BaseModelSettings(BaseModel):
     max_tokens: int = Field(config.default_max_tokens, gt=1, le=1024)
     generative_model_name: str = Field("")
     embeddings_model_name: str = Field("")
+    rerank_model_name: str = Field("")
     profile_name: str = Field(CI_LOGS_PROFILE)
     enable_rerank: bool = Field(config.enable_rerank)
 
@@ -76,6 +77,15 @@ async def validate_settings(request: BaseModelSettings) -> BaseModelSettings:
             detail=f"Invalid embeddings model. Available: {available_embedding_models}"
         )
 
+    available_rerank_models = rerank_model_provider.all_model_names
+    if not request.rerank_model_name:
+        request.rerank_model_name = available_rerank_models[0]
+    elif request.rerank_model_name not in available_rerank_models:
+        raise HTTPException(
+            status_code=400,
+            detail=f"Invalid rerank model. Available: {available_rerank_models}"
+        )
+
     if request.profile_name not in [CI_LOGS_PROFILE, DOCS_PROFILE, RCA_FULL_PROFILE]:
         raise HTTPException(
             status_code=400,
@@ -224,12 +234,16 @@ async def process_prompt(
     embeddings_model_settings: ModelSettings = {
         "model": message_data.embeddings_model_name,
     }
+    rerank_model_settings: ModelSettings = {
+        "model": message_data.rerank_model_name,
+    }
 
     response = await handle_user_message_api(
         message_data.content,
         message_data.similarity_threshold,
         generative_model_settings,
         embeddings_model_settings,
+        rerank_model_settings,
         message_data.profile_name,
         message_data.enable_rerank,
         )
@@ -263,6 +277,9 @@ async def process_rca(
     embeddings_model_settings: ModelSettings = {
         "model": request.embeddings_model_name,
     }
+    rerank_model_settings: ModelSettings = {
+        "model": request.rerank_model_name,
+    }
 
     unique_items = {}
     for item in traceback_items:
@@ -278,6 +295,7 @@ async def process_rca(
             similarity_threshold=request.similarity_threshold,
             generative_model_settings=generative_model_settings,
             embeddings_model_settings=embeddings_model_settings,
+            rerank_model_settings=rerank_model_settings,
             profile_name=request.profile_name,
             enable_rerank=request.enable_rerank,
         )

src/rca_accelerator_chatbot/chat.py

@@ -1,8 +1,9 @@
 """Handler for chat messages and responses."""
 from dataclasses import dataclass
-import chainlit as cl
-import httpx
+
 from openai.types.chat import ChatCompletionAssistantMessageParam
+import httpx
+import chainlit as cl
 
 from rca_accelerator_chatbot import constants
 from rca_accelerator_chatbot.prompt import build_prompt
@@ -330,6 +331,7 @@ async def handle_user_message_api( # pylint: disable=too-many-arguments
     similarity_threshold: float,
     generative_model_settings: ModelSettings,
     embeddings_model_settings: ModelSettings,
+    rerank_model_settings: ModelSettings,
     profile_name: str,
     enable_rerank: bool = True,
     ) -> MockMessage:
@@ -362,6 +364,9 @@ async def handle_user_message_api( # pylint: disable=too-many-arguments
             settings={
                 "enable_rerank": enable_rerank,
                 "rerank_top_n": config.rerank_top_n,
+                "rerank_model": rerank_model_settings["model"],
+                "generative_model": generative_model_settings["model"],
+                "embeddings_model": embeddings_model_settings["model"],
             },
         )
     except httpx.HTTPStatusError:

src/rca_accelerator_chatbot/models/embeddings.py

@@ -1,7 +1,7 @@
 """Provider for the embedding model."""
 
-import chainlit as cl
 from openai import OpenAIError
+import chainlit as cl
 
 from rca_accelerator_chatbot.config import config
 from rca_accelerator_chatbot.models.model import ModelProvider

src/rca_accelerator_chatbot/models/generative.py

@@ -1,6 +1,6 @@
 """Provider for the generative model."""
-import chainlit as cl
 from openai import OpenAIError
+import chainlit as cl
 
 from rca_accelerator_chatbot.settings import ModelSettings, ThreadMessages
 from rca_accelerator_chatbot.config import config

src/rca_accelerator_chatbot/vectordb.py

@@ -1,9 +1,10 @@
 """Vector database client for RAG operations."""
 
 from typing import List
-import chainlit as cl
+
 from qdrant_client import QdrantClient
 from qdrant_client.http.exceptions import ApiException
+import chainlit as cl
 
 from rca_accelerator_chatbot.config import config