Update app.py

app.py

@@ -61,7 +61,8 @@ def givetext(input_text,lmodel,ltokenizer):
         with torch.no_grad():
             print("BEFORE RETURNING")
             
-            return (ltokenizer.decode(lmodel.generate(**model_input, max_new_tokens=1000)[0], skip_special_tokens=True))
+            return (ltokenizer.decode(lmodel.generate(**model_input, max_new_tokens=1000, device="cpu")[0], skip_special_tokens=True))
+            #return (ltokenizer.decode(lmodel.generate(**model_input, max_new_tokens=1000)[0], skip_special_tokens=True))
             #return (ltokenizer.decode(lmodel.generate(**model_input, max_new_tokens=100)[0], skip_special_tokens=True))
     except Exception as error:
         print("Exception {error}".format(error = error))