remove btn

app.py

@@ -2,12 +2,12 @@ import gradio as gr
 from langchain_community.document_loaders import PyPDFLoader
 from langchain_community.vectorstores import FAISS
 from langchain_huggingface.embeddings import HuggingFaceEmbeddings
-from langchain_community.llms import HuggingFacePipeline
+from langchain_huggingface.llms import HuggingFacePipeline
 from langchain.chains import RetrievalQA
 from transformers import pipeline
 import os
 
-# 1. Ladda och indexera alla PDF:er i mappen "dokument/"
+# 1. Ladda och indexera alla PDF:er i mappen "document/"
 def load_vectorstore():
     all_docs = []
     for filename in os.listdir("document"):
@@ -19,32 +19,20 @@ def load_vectorstore():
     embedding = HuggingFaceEmbeddings(model_name="KBLab/sentence-bert-swedish-cased")
     return FAISS.from_documents(all_docs, embedding)
 
-vectorstore = load_vectorstore()
+with gr.Blocks() as demo:
+    gr.Markdown("### ⏳ Laddar dokument och språkmodell...")
+    vectorstore = load_vectorstore()
+    llm = pipeline("text-generation", model="tiiuae/falcon-rw-1b", device=-1)
+    llm = HuggingFacePipeline(pipeline=llm, model_kwargs={"temperature": 0.3, "max_new_tokens": 512})
+    qa_chain = RetrievalQA.from_chain_type(llm=llm, retriever=vectorstore.as_retriever())
 
-# 2. Initiera en mindre modell
+    def chat_fn(message, history):
+        svar = qa_chain.run(message)
+        return svar
 
-def load_model():
-    pipe = pipeline("text-generation", model="tiiuae/falcon-rw-1b")
-    return HuggingFacePipeline(pipeline=pipe, model_kwargs={"temperature": 0.3, "max_new_tokens": 512})
+    chatbot = gr.ChatInterface(fn=chat_fn,
+                               title="🌟 Dokumentassistent (Svenska)",
+                               examples=["Vad handlar dokumentet om?", "Står det något om diabetes?", "Vilken åtgärd rekommenderas?"])
+    chatbot.render()
 
-llm = load_model()
-
-# 3. Bygg QA-kedjan
-qa_chain = RetrievalQA.from_chain_type(llm=llm, retriever=vectorstore.as_retriever())
-
-# 4. Funktion för Gradio-chat
-chat_history = []
-
-def chat_fn(message, history):
-    svar = qa_chain.run(message)
-    return svar
-
-# 5. Starta Gradio-gränssnittet
-chatbot = gr.ChatInterface(fn=chat_fn,
-                           title="🌟 Dokumentagent på Svenska",
-                           theme="soft",
-                           examples=["Vad handlar dokumentet om?", "Finns det något om diabetes?", "Vilken åtgärd föreslås?"],
-                           retry_btn="↻ Pröva igen",
-                           submit_btn="Ställ fråga")
-
-chatbot.launch()
+demo.launch()