Update app.py

app.py

@@ -206,12 +206,17 @@ def generate_text (prompt, chatbot, history, vektordatenbank, retriever, top_p=0
         #oder an Hugging Face --------------------------
         print("HF Anfrage.......................")
         model_kwargs={"temperature": 0.5, "max_length": 512, "num_return_sequences": 1, "top_k": top_k, "top_p": top_p, "repetition_penalty": repetition_penalty}
-        llm = HuggingFaceHub(repo_id=repo_id, model_kwargs=model_kwargs) 
-        #llm = HuggingFaceChain(model=MODEL_NAME_HF, model_kwargs={"temperature": 0.5, "max_length": 128}) 
+        #llm = HuggingFaceHub(repo_id=repo_id, model_kwargs=model_kwargs) 
+
         # Erstelle eine Pipeline mit den gewünschten Parametern
-        #pipe = pipeline("text-generation", model=MODEL_NAME_HF , model_kwargs=model_kwargs)
-        # Erstelle eine HuggingFacePipeline-Kette
-        #llm = HuggingFacePipeline(pipeline=pipe)
+        #pipe = pipeline("text-generation", model=MODEL_NAME_HF, config={"temperature": 0.5, "max_length": 512, "num_return_sequences": 1, "top_k": top_k, "top_p": top_p, "repetition_penalty": repetition_penalty})
+
+        # Erstelle eine HuggingFaceEndPoints-Instanz mit den entsprechenden Endpunkt-Parametern
+        llm = HuggingFaceEndPoints(
+            endpoint_url=f"https://api-inference.huggingface.co/models/{MODEL_NAME_HF}",
+            api_key=hf_token,
+            model_kwargs=model_kwargs
+        )
         
         #Prompt an history anhängen und einen Text daraus machen
         history_text_und_prompt = generate_prompt_with_history(prompt, history) 
@@ -222,6 +227,7 @@ def generate_text (prompt, chatbot, history, vektordatenbank, retriever, top_p=0
         result = rag_chain(llm, history_text_und_prompt, retriever)
         print("result regchain.....................")
         print(result)
+        print("Ene result............................")
            
     except Exception as e:
         raise gr.Error(e)