Update app.py

app.py

@@ -120,8 +120,8 @@ import transformers
 # model = load_checkpoint_and_dispatch(model, model_id, device_map=device_map, no_split_module_classes=["GPTJBlock"])
 # model.half()
 
-MODEL_ID = "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B"
-# MODEL_ID = "microsoft/phi-4"
+# MODEL_ID = "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B"
+MODEL_ID = "microsoft/phi-4"
 
 CHAT_TEMPLATE = "َAuto"
 MODEL_NAME = MODEL_ID.split("/")[-1]
@@ -134,7 +134,7 @@ quantization_config = BitsAndBytesConfig(
     bnb_4bit_compute_dtype=torch.bfloat16
 )
 tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
-model = AutoModel.from_pretrained(
+model = AutoModelForCausalLM.from_pretrained(
     MODEL_ID,
     device_map="auto",
      low_cpu_mem_usage=True,