Update app.py

app.py

@@ -19,6 +19,7 @@ from langchain.embeddings.openai import OpenAIEmbeddings
 from langchain.prompts import PromptTemplate
 from langchain.text_splitter import RecursiveCharacterTextSplitter
 from langchain.vectorstores import Chroma
+from chromadb.errors import InvalidDimensionException
 #from langchain.vectorstores import MongoDBAtlasVectorSearch
 
 #from pymongo import MongoClient
@@ -75,6 +76,7 @@ YOUTUBE_URL_2 = "https://www.youtube.com/watch?v=hdhZwyf24mE"
 
 
 
+
 ################################################
 #LLM Model mit dem gearbeitet wird
 #openai
@@ -168,10 +170,12 @@ def document_loading_splitting():
 
 #Chroma DB die splits ablegen - vektorisiert...
 def document_storage_chroma(splits):
-    Chroma.from_documents(documents = splits, 
-                          embedding = OpenAIEmbeddings(disallowed_special = ()), 
-                          persist_directory = PATH_WORK + CHROMA_DIR)
+    #OpenAi embediings
+    #Chroma.from_documents(documents = splits, embedding = OpenAIEmbeddings(disallowed_special = ()),  persist_directory = PATH_WORK + CHROMA_DIR)  
 
+    #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...
 def document_storage_mongodb(splits):
     MongoDBAtlasVectorSearch.from_documents(documents = splits,
@@ -186,10 +190,9 @@ def document_retrieval_chroma(llm, prompt):
     #embeddings = HuggingFaceInstructEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2", model_kwargs={"device": "cpu"})
     #etwas weniger rechenaufwendig:
     embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-mpnet-base-v2", model_kwargs={"device": "cpu"}, encode_kwargs={'normalize_embeddings': False})
-    
-    db = Chroma(embedding_function = embeddings,
-                persist_directory = PATH_WORK + CHROMA_DIR)
-    
+
+    #ChromaDb für OpenAI embedinngs
+    db = Chroma(embedding_function = embeddings, persist_directory = PATH_WORK + CHROMA_DIR)
     return db
 
 #dokumente in mongo db vektorisiert ablegen können - die Db vorbereiten daüfür
@@ -304,6 +307,7 @@ def invoke (prompt, history, rag_option, openai_api_key,  temperature=0.9, max_n
         #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,)
         
+
         #zusätzliche Dokumenten Splits aus DB zum Prompt hinzufügen (aus VektorDB - Chroma oder Mongo DB)
         if (rag_option == "An"):
             #muss nur einmal ausgeführt werden...