Update app.py

app.py

@@ -79,11 +79,11 @@ YOUTUBE_URL_2 = "https://www.youtube.com/watch?v=hdhZwyf24mE"
 
 ################################################
 #LLM Model mit dem gearbeitet wird
-#openai
-#MODEL_NAME  = "gpt-3.5-turbo-16k"
-MODEL_NAME  ="gpt-4"
+#openai-------------------------------------
+MODEL_NAME  = "gpt-3.5-turbo-16k"
+#MODEL_NAME  ="gpt-4"
 
-#HuggingFace
+#HuggingFace--------------------------------
 #repo_id = "meta-llama/Llama-2-13b-chat-hf"   
 repo_id = "HuggingFaceH4/zephyr-7b-alpha"
 #repo_id = "meta-llama/Llama-2-70b-chat-hf"    
@@ -170,10 +170,10 @@ def document_loading_splitting():
 
 #Chroma DB die splits ablegen - vektorisiert...
 def document_storage_chroma(splits):
-    #OpenAi embediings
+    #OpenAi embediings----------------------------------
     Chroma.from_documents(documents = splits, embedding = OpenAIEmbeddings(disallowed_special = ()),  persist_directory = PATH_WORK + CHROMA_DIR)  
 
-    #HF embeddings
+    #HF embeddings--------------------------------------
     #Chroma.from_documents(documents = splits, embedding = HuggingFaceEmbeddings(model_name="sentence-transformers/all-mpnet-base-v2", model_kwargs={"device": "cpu"}, encode_kwargs={'normalize_embeddings': False}),  persist_directory = PATH_WORK + CHROMA_DIR)  
     
 #Mongo DB die splits ablegen - vektorisiert...
@@ -184,8 +184,11 @@ def document_storage_mongodb(splits):
                                             index_name = MONGODB_INDEX_NAME)
 
 #dokumente in chroma db vektorisiert ablegen können - die Db vorbereiten daüfur
-def document_retrieval_chroma(llm, prompt):      
+def document_retrieval_chroma(llm, prompt):  
+    #OpenAI embeddings -------------------------------
     embeddings = OpenAIEmbeddings()
+
+    #HF embeddings -----------------------------------
     #Alternative Embedding - für Vektorstore, um Ähnlichkeitsvektoren zu erzeugen - die ...InstructEmbedding ist sehr rechenaufwendig
     #embeddings = HuggingFaceInstructEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2", model_kwargs={"device": "cpu"})
     #etwas weniger rechenaufwendig:
@@ -300,10 +303,10 @@ def invoke (prompt, history, rag_option, openai_api_key,  temperature=0.9, max_n
         ###########################
         #LLM auswählen (OpenAI oder HF)
         ###########################
-        #Anfrage an OpenAI
+        #Anfrage an OpenAI ----------------------------
         #llm = ChatOpenAI(model_name = MODEL_NAME,  openai_api_key = openai_api_key, temperature=temperature)#, top_p = top_p)
-        #oder an Hugging Face
-        llm = HuggingFaceHub(repo_id=repo_id, model_kwargs={"temperature": 0.5, "max_length": 64}) 
+        #oder an Hugging Face --------------------------
+        llm = HuggingFaceHub(repo_id=repo_id, 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,)