Upload app.py

app.py

@@ -1,11 +1,20 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
+from openai import OpenAI
+import os
+import numpy as np
+from src.document_processing.processor import DocumentProcessor
+from src.rag.retriever import Retriever
+from src.rag.generator import Generator
+from src.api.openai_api import OpenAIAPI
 
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
+# Initialize OpenAI client
+api_key = os.environ.get("OPENAI_API_KEY", "")
+openai_api = OpenAIAPI(api_key=api_key)
 
+# Initialize RAG components with OpenAI API
+document_processor = DocumentProcessor(api_client=openai_api)
+retriever = Retriever(api_client=openai_api)
+generator = Generator(api_client=openai_api)
 
 def respond(
     message,
@@ -15,6 +24,30 @@ def respond(
     temperature,
     top_p,
 ):
+    # Check if we should use RAG
+    use_rag = "bruk dokumenter" in message.lower() or "bruk rag" in message.lower()
+    
+    if use_rag:
+        # Use our RAG implementation with GPT-4o
+        try:
+            # Retrieve relevant chunks
+            retrieved_chunks = retriever.retrieve(message)
+            
+            # Generate response using RAG
+            response = generator.generate(
+                query=message,
+                retrieved_chunks=retrieved_chunks,
+                temperature=temperature
+            )
+            
+            yield response
+            return
+        except Exception as e:
+            # If RAG fails, fall back to standard GPT-4o
+            print(f"RAG failed: {str(e)}, falling back to standard GPT-4o")
+    
+    # Standard GPT-4o approach
+    client = OpenAI(api_key=api_key)
     messages = [{"role": "system", "content": system_message}]
 
     for val in history:
@@ -27,17 +60,18 @@ def respond(
 
     response = ""
 
-    for message in client.chat_completion(
-        messages,
+    for chunk in client.chat.completions.create(
+        model="gpt-4o",
+        messages=messages,
         max_tokens=max_tokens,
         stream=True,
         temperature=temperature,
         top_p=top_p,
     ):
-        token = message.choices[0].delta.content
-
-        response += token
-        yield response
+        content = chunk.choices[0].delta.content
+        if content:
+            response += content
+            yield response
 
 
 """
@@ -46,9 +80,12 @@ For information on how to customize the ChatInterface, peruse the gradio docs: h
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
-        gr.Textbox(value="Du er en hjelpsom assistent", label="System message"),
+        gr.Textbox(
+            value="Du er en hjelpsom assistent som svarer på norsk. Bruk kunnskapen din til å svare på spørsmål. Hvis brukeren skriver 'bruk dokumenter' eller 'bruk RAG', vil du bruke Retrieval-Augmented Generation for å svare basert på opplastede dokumenter.", 
+            label="System message"
+        ),
         gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.2, step=0.1, label="Temperature"),
+        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
         gr.Slider(
             minimum=0.1,
             maximum=1.0,
@@ -57,8 +94,48 @@ demo = gr.ChatInterface(
             label="Top-p (nucleus sampling)",
         ),
     ],
+    title="Norwegian RAG Chatbot with GPT-4o",
+    description="En chatbot basert på Retrieval-Augmented Generation (RAG) for norsk språk med GPT-4o. Skriv 'bruk dokumenter' eller 'bruk RAG' i meldingen din for å aktivere RAG-funksjonalitet.",
 )
 
+# Create the document upload interface
+with gr.Blocks() as document_upload:
+    with gr.Tab("Last opp dokumenter"):
+        with gr.Row():
+            with gr.Column(scale=2):
+                file_output = gr.File(label="Opplastede dokumenter")
+                upload_button = gr.UploadButton(
+                    "Klikk for å laste opp dokument",
+                    file_types=["pdf", "txt", "html"],
+                    file_count="multiple"
+                )
+            
+            with gr.Column(scale=3):
+                documents_list = gr.Dataframe(
+                    headers=["Dokument ID", "Filnavn", "Dato", "Chunks"],
+                    label="Dokumentliste",
+                    interactive=False
+                )
+        
+        process_status = gr.Textbox(label="Status", interactive=False)
+        refresh_btn = gr.Button("Oppdater dokumentliste")
+        
+        # Set up event handlers
+        upload_button.upload(
+            fn=document_processor.process_document,
+            inputs=[upload_button],
+            outputs=[process_status, documents_list]
+        )
+        
+        refresh_btn.click(
+            fn=lambda: [[doc_id, meta.get("filename", "N/A"), meta.get("processed_date", "N/A"), meta.get("chunk_count", 0)] 
+                        for doc_id, meta in document_processor.get_all_documents().items()],
+            inputs=None,
+            outputs=[documents_list]
+        )
+
+# Combine the interfaces
+app = gr.TabbedInterface([demo, document_upload], ["Chat", "Dokumenter"])
 
 if __name__ == "__main__":
-    demo.launch()
+    app.launch()