Update app.py

app.py

@@ -134,8 +134,8 @@ huggingface_token = os.getenv("HUGGINGFACE_TOKEN")
 
 
 hf_hub_download(
-    repo_id="unsloth/Qwen3.5-122B-A10B-GGUF",
-    filename="Qwen3.5-122B-A10B-UD-IQ3_S.gguf",
+    repo_id="unsloth/Qwen3.5-35B-A3B-GGUF",
+    filename="Qwen3.5-35B-A3B-UD-Q4_K_XL.gguf",
     local_dir="./models"
 )
 from huggingface_hub import snapshot_download
@@ -149,7 +149,7 @@ from huggingface_hub import snapshot_download
 # )
 
 hf_hub_download(
-    repo_id="unsloth/Qwen3.5-122B-A10B-GGUF",
+    repo_id="unsloth/Qwen3.5-35B-A3B-GGUF",
     filename="mmproj-BF16.gguf",
     local_dir="./models"
 )
@@ -380,7 +380,7 @@ def load_llama_model(model_path, enable_thinking):
         print("Loading model:", key)
 
         llm = Llama(
-            model_path="models/Qwen3.5-122B-A10B-UD-IQ3_S.gguf",
+            model_path="models/Qwen3.5-35B-A3B-UD-Q4_K_XL.gguf",
             flash_attn=True,
             n_gpu_layers=-1,
             n_batch=4096,