Update app.py

app.py

@@ -16,13 +16,9 @@ MODEL_FILE = "gemma-2-27b-it-Q4_K_M.gguf"
 
 os.environ["HF_HUB_ENABLE_HF_TRANSFER"] = "1"
 
-llm = Llama(
-    model_path=hf_hub_download(
-        repo_id=REPO_ID,
-        filename=MODEL_FILE,
-    ),
-    n_ctx=4096,
-    n_gpu_layers=-1,
+llm = llama_cpp.Llama.from_pretrained(
+    repo_id=REPO_ID,
+    filename=MODEL_FILE,
     tokenizer=llama_cpp.llama_tokenizer.LlamaHFTokenizer.from_pretrained(MODEL_ID),
     verbose=False,
 )