Update app.py

app.py

@@ -20,29 +20,33 @@ vector_store = ChromaVectorStore(chroma_collection=chroma_collection)
 embed_model = HuggingFaceEmbedding(model_name="BAAI/bge-small-en-v1.5", device="cpu")
 
 # Utwórz pipeline do przetwarzania dokumentów
-pipeline = IngestionPipeline(
-    transformations=[
-        SentenceSplitter(),
-        embed_model,
-    ],
-    vector_store=vector_store
-)
+# pipeline = IngestionPipeline(
+#     transformations=[
+#         SentenceSplitter(),
+#         embed_model,
+#     ],
+#     vector_store=vector_store
+# )
 
 # Utwórz indeks
 index = VectorStoreIndex.from_vector_store(vector_store, embed_model=embed_model)
 
 # Utwórz silnik zapytań
-from transformers import AutoTokenizer
-
-# Load the correct tokenizer for Qwen/Qwen2-7B-Instruct
-tokeni = AutoTokenizer.from_pretrained("Qwen/Qwen2-0.5B")
 
-llm = HuggingFaceLLM(model_name="Qwen/Qwen2-0.5B", tokenizer=tokeni)
+# Load the correct tokenizer and LLM
+from transformers import AutoTokenizer
+tokenizer = AutoTokenizer.from_pretrained("HuggingFaceH4/zephyr-7b-beta")
+llm = HuggingFaceLLM(model_name="HuggingFaceH4/zephyr-7b-beta", tokenizer=tokenizer)
 
+#Query Engine
 query_engine = index.as_query_engine(
     llm=llm,
     response_mode='compact')
 
+
+
+
+# app gui
 # Store LLM generated responses
 if "messages" not in st.session_state.keys():
     st.session_state.messages = [{"role": "assistant", "content": "Zadaj mi pytanie..."}]