Update app.py

app.py

@@ -1,13 +1,15 @@
 import gradio as gr
 from transformers import AutoModelForCausalLM, AutoTokenizer
 
-# 加载模型（使用量化版本节省显存）
-model_name = "Qwen/Qwen2.5-7B-Instruct-GPTQ-Int8"
-tokenizer = AutoTokenizer.from_pretrained(model_name)
+from huggingface_hub import login
+login(token="你的Token")  # 添加在模型加载前
+
 model = AutoModelForCausalLM.from_pretrained(
-    model_name,
-    device_map="auto",  # 自动分配GPU/CPU
-    torch_dtype="auto"
+    "Qwen/Qwen2.5-7B-Instruct-AWQ",
+    device_map="auto",
+    use_auth_token=True,  # 启用认证
+    resume_download=True,  # 断点续传
+    local_files_only=False
 )
 
 # 定义生成函数