Update api.py

api.py

@@ -101,13 +101,15 @@ retriever = db.as_retriever()
 # 5. RAG Pipeline
 llm = ChatGoogleGenerativeAI(model=gemini_model_name, temperature=0.7)
 
-def create_prompt(disease_name, severity):
-    prompt = """
-        You are an expert in plant disease treatment.
+def create_prompt(disease_name, severity,language="english"):
+    prompt = f"""
+    for the next prompt you should answer with the following language and it is obligatory : {language}
+    """ +"""
+        You are an expert in plant disease treatment for a platform called Growth.
         Use the following context to answer the user's question.
         If you don't know the answer, try to find the answer from the context try to make similair answer to the context.
         If the question is not related to the context, Generate from your exsisting knowlodge about the context.
-
+        
         Context:
         {context}
 
@@ -128,7 +130,19 @@ def create_prompt(disease_name, severity):
         - Organic:"""
     return prompt
 
-
+def create_prompt_chat(question,language):
+    prompt = f"""
+    for the next prompt you should answer with the following language and it is obligatory : {language}
+    """ +"""
+        You are a chatbot called Growth, and you assist users with plant leaf disease identification and solutions. 
+        Use the following context to answer the user's question.
+        If you don't know the answer, try to find the answer from the context try to make similair answer to the context.
+        If the question is not related to the context, Generate from your exsisting knowlodge about the context.
+        Context:
+        {context}
+        Answer the following question:
+        """+f"""{question}"""
+    return prompt
 
 
 # Load the saved model
@@ -383,10 +397,11 @@ async def classify(image: UploadFile = File(...)):
 class DiseaseQuery(BaseModel):
     disease: str
     severity: str  # "normal" or "severe"
+    language: str
 @app.post("/RAG")
 async def rag_classify(query: DiseaseQuery):
     try:
-        prompt_template = create_prompt(query.disease, query.severity)
+        prompt_template = create_prompt(query.disease, query.severity,query.language)
 
         prompt = PromptTemplate(
             input_variables=["context", "question"],
@@ -409,6 +424,35 @@ async def rag_classify(query: DiseaseQuery):
     except Exception as e:
         raise HTTPException(status_code=500, detail=str(e))
 
+class ChatQuery(BaseModel):
+    question: str
+    language: str
+@app.post("/RAGChat")
+async def rag_chat(query: ChatQuery):
+    try:
+        prompt_template = create_prompt(query.question,query.language)
+
+        prompt = PromptTemplate(
+            input_variables=["context", "question"],
+            template=prompt_template,
+        )
+
+        qa = RetrievalQA.from_chain_type(
+            llm=llm,
+            chain_type="stuff",
+            retriever=retriever,
+            chain_type_kwargs={"prompt": prompt},
+            verbose=True,
+        )
+
+        final_query = f"Answer the Provided questino like a human, and remember the chat history {query.question}"
+        result = qa.run(final_query)
+
+        return {"answer": result}
+
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
 class GeoSenQuery(BaseModel):
     region: str
     crop_type: str