Upload fun_advaitbert.py

fun_advaitbert.py

@@ -225,6 +225,7 @@ def predict_CTH(txt):
 
 def llm_model_function(txt,history,chatbot=[], temperature=0.9, max_new_tokens=1024, top_p=0.95, repetition_penalty=1.0,):
     system_prompt=[]
+    chatbot=[]
     if (txt!='') and len(txt)>=3 and (count_special_character(txt)):
         valid_data = tf.data.Dataset.from_tensor_slices(([txt] , [1])) # 1 refers to 'entertainment' and 2 refers to 'sport'
         valid_data = (valid_data.map(to_feature_map).batch(1))
@@ -248,7 +249,9 @@ def llm_model_function(txt,history,chatbot=[], temperature=0.9, max_new_tokens=1
             pred_desc=''
             pred_CTH=''
 
-            return{'Not a adequate description':float(1.0)}
+            #return{'Not a adequate description':float(1.0)}
+            chatbot.append(('Not a adequate description', 'Not a adequate description'))
+            return "", chatbot
         else:
             Var_CTH=[]
             Var_desc=[]
@@ -306,11 +309,14 @@ def llm_model_function(txt,history,chatbot=[], temperature=0.9, max_new_tokens=1
             chatbot.append((txt, output))
             return "", chatbot 
     else:
-        warning_msg = f"Unexpected response"
-        raise gr.Error(warning_msg)
+        # warning_msg = f"Unexpected response"
+        # raise gr.Error(warning_msg)
+        chatbot.append(('Not a adequate description', 'Not a adequate description'))
+        return "", chatbot
 
 def product_explaination(txt,history,chatbot=[], temperature=0.9, max_new_tokens=1024, top_p=0.95, repetition_penalty=1.0,):
     print('Input Descrption is:',txt)
+    chatbot=[]
     prompt=f'What is the product- {txt}?'
     print('prompt',prompt)
     temperature = float(temperature)