Update app.py

app.py

@@ -11,7 +11,7 @@ from langchain.document_loaders.blob_loaders.youtube_audio import YoutubeAudioLo
 from langchain.document_loaders.generic import GenericLoader
 from langchain.document_loaders.parsers import OpenAIWhisperParser
 from langchain.schema import AIMessage, HumanMessage
-from langchain.llms import HuggingFaceHub
+from langchain.llms import HuggingFaceHub, HuggingFaceChain
 from langchain.llms import HuggingFaceTextGenInference
 from langchain.embeddings import HuggingFaceInstructEmbeddings, HuggingFaceEmbeddings, HuggingFaceBgeEmbeddings, HuggingFaceInferenceAPIEmbeddings
 
@@ -87,7 +87,7 @@ MODEL_NAME  = "gpt-3.5-turbo-16k"
 #verfügbare Modelle anzeigen lassen
 
 
-#HuggingFace--------------------------------
+#HuggingFace Reop ID--------------------------------
 #repo_id = "meta-llama/Llama-2-13b-chat-hf"   
 repo_id = "HuggingFaceH4/zephyr-7b-alpha"   #das Modell ist echt gut!!! Vom MIT
 #repo_id = "TheBloke/Yi-34B-Chat-GGUF"
@@ -103,6 +103,8 @@ repo_id = "HuggingFaceH4/zephyr-7b-alpha"   #das Modell ist echt gut!!! Vom MIT
 #repo_id = "databricks/dolly-v2-3b"
 #repo_id = "google/flan-t5-xxl"
 
+#HuggingFace Model name--------------------------------
+MODEL_NAME_HF  = "mistralai/Mixtral-8x7B-Instruct-v0.1"
 
 ################################################
 #HF Hub Zugriff ermöglichen
@@ -318,7 +320,8 @@ def invoke (prompt, history, rag_option, model_option, openai_api_key,  temperat
             print("openAI")
         else:
             #oder an Hugging Face --------------------------
-            llm = HuggingFaceHub(repo_id=repo_id, model_kwargs={"temperature": 0.5, "max_length": 128}) 
+            #llm = HuggingFaceHub(repo_id=repo_id, model_kwargs={"temperature": 0.5, "max_length": 128}) 
+            llm = HuggingFaceChain(model=MODEL_NAME_HF, model_kwargs={"temperature": 0.5, "max_length": 128}) 
             #llm = HuggingFaceHub(url_??? = "https://wdgsjd6zf201mufn.us-east-1.aws.endpoints.huggingface.cloud", model_kwargs={"temperature": 0.5, "max_length": 64}) 
             #llm = HuggingFaceTextGenInference( inference_server_url="http://localhost:8010/", max_new_tokens=max_new_tokens,top_k=10,top_p=top_p,typical_p=0.95,temperature=temperature,repetition_penalty=repetition_penalty,)
             print("HF")