Fix run mode of model (#61)

added code to run the model in inference_mode.

backend/models/interfaces/model_search.py

@@ -104,7 +104,8 @@ def answer_question(question, answer_text, model, tokenizer):
     # There should be a segment_id for every input token.
     assert len(segment_ids) == len(input_ids)
 
-    outputs = model(torch.tensor([input_ids]),  # The tokens representing our input text.
+    with torch.inference_mode(): ## Run the model in inference mode
+        outputs = model(torch.tensor([input_ids]),  # The tokens representing our input text.
                     # The segment IDs to differentiate question from answer_text
                     token_type_ids=torch.tensor([segment_ids]),
                     return_dict=True)
@@ -656,4 +657,4 @@ def partial_ratio(s1, s2):
         max_distance = max(max_distance, len(shorter) - distance)
 
     ratio = (max_distance / len(shorter)) * 100
-    return ratio
+    return ratio