# BERT encoding

**URL:** https://forum.rasa.com/t/bert-encoding/5297
**Category:** Rasa Open Source
**Created:** [February 18, 2019, 6:27pm UTC](https://forum.rasa.com/t/bert-encoding/5297 "2019-02-18T18:27:36Z")
**Posts on this page:** 1
**Showing post:** 6

<div class="post-metadata">

### Author: ![yanvirin](https://avatars.discourse-cdn.com/v4/letter/y/ed8c4c/32.png) [@yanvirin](https://forum.rasa.com/u/yanvirin)
#### Post date: [March 21, 2019, 9:04pm UTC](https://forum.rasa.com/t/bert-encoding/5297/6 "2019-03-21T21:04:56Z")

</div>

When I am trying to run this:

python -m rasa\_nlu.evaluate --config conf/experiments/nlu4.yml --model /tmp/bert\_exp/order\_tracking/nlu/ --data data/nlu\_test/

with the innatis bert classifier configured in nlu4.yml, I get the following:

/Users/yan/environments/rasa/lib/python3.6/site-packages/rasa\_nlu/training\_data/training\_data.py:191: UserWarning: Intent ‘asking\_refund+damaged\_order’ has only 1 training examples! Minimum is 2, training may fail. self.MIN\_EXAMPLES\_PER\_INTENT)) /Users/yan/environments/rasa/lib/python3.6/site-packages/rasa\_nlu/training\_data/training\_data.py:191: UserWarning: Intent ‘cancel\_order+update\_order’ has only 1 training examples! Minimum is 2, training may fail. self.MIN\_EXAMPLES\_PER\_INTENT)) /Users/yan/environments/rasa/lib/python3.6/site-packages/rasa\_nlu/training\_data/training\_data.py:191: UserWarning: Intent ‘delayed\_order+cancel\_order’ has only 1 training examples! Minimum is 2, training may fail. self.MIN\_EXAMPLES\_PER\_INTENT)) /Users/yan/environments/rasa/lib/python3.6/site-packages/rasa\_nlu/training\_data/training\_data.py:191: UserWarning: Intent ‘general\_qa+get\_order\_info’ has only 1 training examples! Minimum is 2, training may fail. self.MIN\_EXAMPLES\_PER\_INTENT)) /Users/yan/environments/rasa/lib/python3.6/site-packages/rasa\_nlu/training\_data/training\_data.py:191: UserWarning: Intent ‘get\_order\_info+cancel\_order’ has only 1 training examples! Minimum is 2, training may fail. self.MIN\_EXAMPLES\_PER\_INTENT)) Traceback (most recent call last): File “/Users/yan/.pyenv/versions/3.6.6/lib/python3.6/runpy.py”, line 193, in \_run\_module\_as\_main “ **main** ”, mod\_spec) File “/Users/yan/.pyenv/versions/3.6.6/lib/python3.6/runpy.py”, line 85, in \_run\_code exec(code, run\_globals) File “/Users/yan/environments/rasa/lib/python3.6/site-packages/rasa\_nlu/evaluate.py”, line 978, in main() File “/Users/yan/environments/rasa/lib/python3.6/site-packages/rasa\_nlu/evaluate.py”, line 972, in main cmdline\_args.histogram) File “/Users/yan/environments/rasa/lib/python3.6/site-packages/rasa\_nlu/evaluate.py”, line 732, in run\_evaluation test\_data) File “/Users/yan/environments/rasa/lib/python3.6/site-packages/rasa\_nlu/evaluate.py”, line 636, in get\_entity\_predictions res = interpreter.parse(e.text, only\_output\_properties=False) File “/Users/yan/environments/rasa/lib/python3.6/site-packages/rasa\_nlu/model.py”, line 357, in parse component.process(message, \*\*self.context) File “/Users/yan/environments/rasa/lib/python3.6/site-packages/innatis/classifiers/bert\_intent\_classifier.py”, line 170, in process index = tf.argmax(probabilities, axis=0).eval() File “/Users/yan/environments/rasa/lib/python3.6/site-packages/tensorflow/python/framework/ops.py”, line 713, in eval return \_eval\_using\_default\_session(self, feed\_dict, self.graph, session) File “/Users/yan/environments/rasa/lib/python3.6/site-packages/tensorflow/python/framework/ops.py”, line 5148, in \_eval\_using\_default\_session raise ValueError("Cannot use the default session to evaluate tensor: " ValueError: Cannot use the default session to evaluate tensor: the tensor’s graph is different from the session’s graph. Pass an explicit session to `eval(session=sess)`.

Seems like some incorrect way of using the TF session, default vs. non default, but I am not really sure what’s going on. Have you seen something like that during testing?

---

_[View the full topic](https://forum.rasa.com/t/bert-encoding/5297)._