Update app.py

app.py

@@ -10,13 +10,19 @@ __checkpoint = "gpt2"
 __tokenizer = GPT2Tokenizer.from_pretrained(__checkpoint)
 __model = GPT2LMHeadModel.from_pretrained(__checkpoint)
 __model_output_path = "/paavansundar/Models"
+
+#prepare data
+def prepareData():
+    df=pd.read_csv("paavansundar/Datasets/MedQuAD.csv")
+    
 # Create a Data collator object
 data_collator = DataCollatorForLanguageModeling(tokenizer=__tokenizer, mlm=False, return_tensors="pt")
 def queryGPT(question):
     return generate_response(__model, __tokenizer, question)
 
 def generate_response(model,tokenizer, prompt, max_length=200):
-
+    
+    prepareData()
     input_ids = tokenizer.encode(prompt, return_tensors="pt")      # 'pt' for returning pytorch tensor
 
     # Create the attention mask and pad token id