Update qa_summary.py

qa_summary.py

@@ -16,8 +16,9 @@ def generate_answer(llm_name, texts, query, queries, mode='validate'):
         tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2", use_fast=True)
         llm_model = AutoModelForCausalLM.from_pretrained(
         "mistralai/Mistral-7B-Instruct-v0.2",
-        device_map="auto", #device_map="cuda"
-        #torch_dtype=torch.float16,
+        #device_map="auto", 
+        device_map="cuda",
+        torch_dtype=torch.float16,
         )
 
     elif llm_name == 'phi3mini':