revert

app.py

@@ -48,8 +48,9 @@ def load_llamaindex_stack(model_id: str, max_new_tokens: int, temperature: float
     # Tiny, fast sentence-transformers model for embeddings
     embed = HuggingFaceEmbedding(model_name="sentence-transformers/all-MiniLM-L6-v2")
 
-    #tok = AutoTokenizer.from_pretrained(model_id)
-    #mdl = AutoModelForSeq2SeqLM.from_pretrained(model_id)
+    """
+    tok = AutoTokenizer.from_pretrained(model_id)
+    mdl = AutoModelForSeq2SeqLM.from_pretrained(model_id)
     text2text = pipeline(
         task="text2text-generation",
         model=mdl,
@@ -57,6 +58,7 @@ def load_llamaindex_stack(model_id: str, max_new_tokens: int, temperature: float
         max_new_tokens=max_new_tokens,
         temperature=float(temperature)
     )
+    """
 
     # Wrap the same tiny HF model for LlamaIndex