Update app.py

app.py

@@ -1,15 +1,23 @@
+import logging
 from fastapi import FastAPI, HTTPException
 from pydantic import BaseModel
 from sentence_transformers import SentenceTransformer, util
 from transformers import pipeline
 
+# Set up logging
+logging.basicConfig(level=logging.INFO, format="%(asctime)s - %(levelname)s - %(message)s")
+logger = logging.getLogger(__name__)
+
 # Initialize FastAPI app
 app = FastAPI()
 
+# Log model loading
+logger.info("Loading models...")
 # Load models
 model = SentenceTransformer("sentence-transformers/all-MiniLM-L6-v2")
 question_model = "deepset/tinyroberta-squad2"
 nlp = pipeline('question-answering', model=question_model, tokenizer=question_model)
+logger.info("Models loaded successfully.")
 
 # Define request models
 class ModifyQueryRequest(BaseModel):
@@ -19,7 +27,7 @@ class AnswerQuestionRequest(BaseModel):
     question: str
     context: dict
 
-# Define response models (if needed)
+# Define response models
 class ModifyQueryResponse(BaseModel):
     embeddings: list
 
@@ -30,38 +38,50 @@ class AnswerQuestionResponse(BaseModel):
 # Define API endpoints
 @app.post("/modify_query", response_model=ModifyQueryResponse)
 async def modify_query(request: ModifyQueryRequest):
+    logger.info(f"Received /modify_query request: {request.query_string}")
     try:
         binary_embeddings = model.encode([request.query_string], precision="binary")
+        logger.info("Embeddings generated successfully.")
         return ModifyQueryResponse(embeddings=binary_embeddings[0].tolist())
     except Exception as e:
+        logger.error(f"Error generating embeddings: {str(e)}")
         raise HTTPException(status_code=500, detail=str(e))
 
 @app.post("/answer_question", response_model=AnswerQuestionResponse)
 async def answer_question(request: AnswerQuestionRequest):
+    logger.info(f"Received /answer_question request: {request.question}")
     try:
         res_locs = []
         context_string = ''
+        
         corpus_embeddings = model.encode(request.context['context'], convert_to_tensor=True)
         query_embeddings = model.encode(request.question, convert_to_tensor=True)
         hits = util.semantic_search(query_embeddings, corpus_embeddings)
+
         for hit in hits:
-            if hit['score'] > .5:
+            if hit['score'] > 0.5:
                 loc = hit['corpus_id']
                 res_locs.append(request.context['locations'][loc])
                 context_string += request.context['context'][loc] + ' '
+
         if len(res_locs) == 0:
             ans = "Sorry, I couldn't find any results for your query."
+            logger.info("No relevant context found.")
         else:
             QA_input = {
                 'question': request.question,
-                'context': context_string.replace('\n',' ')
+                'context': context_string.replace('\n', ' ')
             }
             result = nlp(QA_input)
             ans = result['answer']
-        return AnswerQuestionResponse(answer=ans, locations = res_locs)
+            logger.info("Answer generated successfully.")
+        
+        return AnswerQuestionResponse(answer=ans, locations=res_locs)
     except Exception as e:
+        logger.error(f"Error answering question: {str(e)}")
         raise HTTPException(status_code=500, detail=str(e))
 
 if __name__ == "__main__":
     import uvicorn
+    logger.info("Starting FastAPI server...")
     uvicorn.run(app, host="0.0.0.0", port=8000)