# Confidence score Different between AWS instance and Local system **URL:** \ **Category:** Rasa Open Source\ **Created:** [October 24, 2019, 10:45am UTC](https://forum.rasa.com/t/confidence-score-different-between-aws-instance-and-local-system/20279 "2019-10-24T10:45:23Z")\ **Posts on this page:** 7\ **Page:** 1
**Author:** ![Pavan\_gali](https://avatars.discourse-cdn.com/v4/letter/p/71c47a/32.png) [@Pavan\_gali](https://forum.rasa.com/u/Pavan_gali)\ **Post date:** [October 24, 2019, 10:45am UTC](https://forum.rasa.com/t/confidence-score-different-between-aws-instance-and-local-system/20279/1 "2019-10-24T10:45:23Z")
Hi team, I am facing issues with confidence score with RASA NLU 15.1 version. In my local I am having Python 3.7 and i have trained the model using Tensorflow embedding pipeline. after training I was getting good accuracy. With same training data I have trained the model in AWS instance with python version 3.6. the confidence scores are different with same utterance one is 0.82 and in Aws it is 0.62. the training accuracy is 0.99 with loss 0.21. please help me ---
**Author:** ![Ghostvv](https://dub1.discourse-cdn.com/flex013/user_avatar/forum.rasa.com/ghostvv/32/211_2.png) [@Ghostvv](https://forum.rasa.com/u/Ghostvv)\ **Post date:** [October 24, 2019, 1:08pm UTC](https://forum.rasa.com/t/confidence-score-different-between-aws-instance-and-local-system/20279/2 "2019-10-24T13:08:54Z")
these are statistical models, the numbers differ from training to training ---
**Author:** ![Ghostvv](https://dub1.discourse-cdn.com/flex013/user_avatar/forum.rasa.com/ghostvv/32/211_2.png) [@Ghostvv](https://forum.rasa.com/u/Ghostvv)\ **Post date:** [October 24, 2019, 1:09pm UTC](https://forum.rasa.com/t/confidence-score-different-between-aws-instance-and-local-system/20279/3 "2019-10-24T13:09:31Z")
if you want to get reproducible training, you need to fix all random seeds ---
**Author:** ![Pavan\_gali](https://avatars.discourse-cdn.com/v4/letter/p/71c47a/32.png) [@Pavan\_gali](https://forum.rasa.com/u/Pavan_gali)\ **Post date:** [October 24, 2019, 1:36pm UTC](https://forum.rasa.com/t/confidence-score-different-between-aws-instance-and-local-system/20279/4 "2019-10-24T13:36:08Z")
I have made random\_seed : 1. I mean constant. still i am facing the issue. ---
**Author:** ![Ghostvv](https://dub1.discourse-cdn.com/flex013/user_avatar/forum.rasa.com/ghostvv/32/211_2.png) [@Ghostvv](https://forum.rasa.com/u/Ghostvv)\ **Post date:** [October 24, 2019, 1:47pm UTC](https://forum.rasa.com/t/confidence-score-different-between-aws-instance-and-local-system/20279/5 "2019-10-24T13:47:00Z")
but setting `random_seed` doesn’t guarantee the same behavior on different machines, since effectively you use different random number generators ---
**Author:** ![Pavan\_gali](https://avatars.discourse-cdn.com/v4/letter/p/71c47a/32.png) [@Pavan\_gali](https://forum.rasa.com/u/Pavan_gali)\ **Post date:** [October 24, 2019, 3:47pm UTC](https://forum.rasa.com/t/confidence-score-different-between-aws-instance-and-local-system/20279/6 "2019-10-24T15:47:52Z")
Thanks for the response. we are actually moving our model to Production environment. One more question. Can we expect 100% accuracy from Training data means. all the utterances in Training data should give expected response with mapped intent. what is optimal number for the accuracy to go for Production deployment. I am training with 5000 utterances with 300+ intents. Shall i expect 100% from the model prediction accuracy. ---
**Author:** ![Ghostvv](https://dub1.discourse-cdn.com/flex013/user_avatar/forum.rasa.com/ghostvv/32/211_2.png) [@Ghostvv](https://forum.rasa.com/u/Ghostvv)\ **Post date:** [October 24, 2019, 4:06pm UTC](https://forum.rasa.com/t/confidence-score-different-between-aws-instance-and-local-system/20279/7 "2019-10-24T16:06:32Z")
what accuracy you want to achieve is defined by you. It depends on your systems and the way you handle possible mistakes