Update app.py

app.py

@@ -13,7 +13,7 @@ token_huggingface = os.getenv("HUGGINGFACE_TOKEN")
 hf_hub_download(
     repo_id="bartowski/gemma-2-2b-it-abliterated-GGUF",
     filename="gemma-2-2b-it-abliterated-Q4_K_M.gguf",
-    local_dir="./modelos",
+    local_dir="./models",
     token=token_huggingface
 )
 
@@ -36,7 +36,7 @@ def responder(
     
     if llm is None:
         llm = Llama(
-            model_path="modelos/2b_it_v2.gguf",
+            model_path="models/gemma-2-2b-it-abliterated-Q4_K_M.gguf",
             flash_attn=True,
             n_gpu_layers=81,
             n_batch=1024,