Update app.py

app.py

@@ -30,13 +30,18 @@ db = Chroma.from_documents(docs, embedding=DummyEmbeddings())
 retriever = db.as_retriever()
 
 # Step 4: Load a small open model instead of Mistral
-model_id = "google/flan-t5-base"  # Or "microsoft/phi-2" (smaller)
+
+
+from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, pipeline
+
+model_id = "google/flan-t5-base"
 tokenizer = AutoTokenizer.from_pretrained(model_id)
-model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", load_in_8bit=True)
+model = AutoModelForSeq2SeqLM.from_pretrained(model_id)
 
-llm_pipeline = pipeline("text-generation", model=model, tokenizer=tokenizer)
+llm_pipeline = pipeline("text2text-generation", model=model, tokenizer=tokenizer)
 llm = HuggingFacePipeline(pipeline=llm_pipeline)
 
+
 # Step 5: RAG Chain
 qa_chain = RetrievalQA.from_chain_type(llm=llm, retriever=retriever)