Update inference_1.py

inference_1.py

@@ -2,7 +2,7 @@ from vllm import LLM, SamplingParams
 import argparse
 import json
 
-def setup_model(model_path):
+def setup_model(model_path,tensor_parallel_size):
     """
     Initialize the fine-tuned Qwen-2.5-7B model from a local path.
     
@@ -97,7 +97,7 @@ def main():
     args = parser.parse_args()
     
     # Initialize the model
-    llm = setup_model(args.model_path)
+    llm = setup_model(args.model_path, args.tensor_parallel_size)
     
     if args.mode == "single":
         if not args.prompt: