# Interactive Learning better than real Bot

**URL:** <https://forum.rasa.com/t/interactive-learning-better-than-real-bot/595>\
**Category:** Rasa Open Source\
**Created:** [August 22, 2018, 9:23am UTC](https://forum.rasa.com/t/interactive-learning-better-than-real-bot/595 "2018-08-22T09:23:37Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![mjoellnier](https://avatars.discourse-cdn.com/v4/letter/m/b782af/32.png) [@mjoellnier](https://forum.rasa.com/u/mjoellnier)\
**Post date:** [August 22, 2018, 9:23am UTC](https://forum.rasa.com/t/interactive-learning-better-than-real-bot/595/1 "2018-08-22T09:23:37Z")

</div>

Hey there!

I’m using the interactive learning script provided by the examples (`train_online.py`). When I’m using it to create training stories the prediction of the next action is nearly perfect. But as soon as I run the bot normally it behaves poorly even when I’m going along the path of a training story step by step!

(EDIT: It doesn’t seem to be a problem with training epochs. Even the flag `--epochs 200` hasn’t solved the issue)

---

<div class="post-metadata">

**Author:** ![mjoellnier](https://avatars.discourse-cdn.com/v4/letter/m/b782af/32.png) [@mjoellnier](https://forum.rasa.com/u/mjoellnier)\
**Post date:** [August 22, 2018, 9:59am UTC](https://forum.rasa.com/t/interactive-learning-better-than-real-bot/595/2 "2018-08-22T09:59:56Z")

</div>

Okay I think I was able to get near the solution: As soon as I’m setting the flag `--batch_size 50` for the learning command the result is **a lot** better!

---

<div class="post-metadata">

**Author:** ![akelad](https://dub1.discourse-cdn.com/flex013/user_avatar/forum.rasa.com/akelad/32/69_2.png) [@akelad](https://forum.rasa.com/u/akelad)\
**Post date:** [August 24, 2018, 1:03pm UTC](https://forum.rasa.com/t/interactive-learning-better-than-real-bot/595/3 "2018-08-24T13:03:28Z")

</div>

Have you retrained the bot after including the stories generated by interactive learning?

---

<div class="post-metadata">

**Author:** ![mjoellnier](https://avatars.discourse-cdn.com/v4/letter/m/b782af/32.png) [@mjoellnier](https://forum.rasa.com/u/mjoellnier)\
**Post date:** [August 27, 2018, 9:21am UTC](https://forum.rasa.com/t/interactive-learning-better-than-real-bot/595/4 "2018-08-27T09:21:02Z")

</div>

Yes I did a training after each interactive learning

---

<div class="post-metadata">

**Author:** ![akelad](https://dub1.discourse-cdn.com/flex013/user_avatar/forum.rasa.com/akelad/32/69_2.png) [@akelad](https://forum.rasa.com/u/akelad)\
**Post date:** [August 27, 2018, 10:28am UTC](https://forum.rasa.com/t/interactive-learning-better-than-real-bot/595/5 "2018-08-27T10:28:32Z")

</div>

I’d guess there’s a problem with some conflicts in your stories. Go through your stories and check that isn’t the case. You can also run Core in `--debug` mode and that should give you some insight as to what’s going wrong, there’s some more info on this here: [http://rasa.com/docs/core/debugging/](http://rasa.com/docs/core/debugging/)