UnexpecTEDIntentPolicy doesn't train on e2e user data

[JEM-Mosig]

Rasa version:
intent-ted branch (#8913), b54a4be

Rasa Version      :         2.7.1
Minimum Compatible Version: 2.6.0
Rasa SDK Version  :         2.7.0
Rasa X Version    :         None
Python Version    :         3.8.0
Operating System  :         macOS-10.16-x86_64-i386-64bit
Python Path       :         /Users/johannes/miniconda3/envs/rasa-master/bin/python

Issue:

When trained on

- story: sad path 1
  steps:
  - intent: greet
  - action: utter_greet
  - user: "blah"
  - bot: "xyz"
  - action: utter_did_that_help
  - intent: affirm
  - action: utter_happy

UnexpecTEDIntentPolicy fails to train.

Error (including full traceback):

Traceback (most recent call last):
  File "/Users/johannes/miniconda3/envs/rasa-master/bin/rasa", line 5, in <module>
    main()
  File "/Users/johannes/rasa/rasa/rasa/__main__.py", line 117, in main
    cmdline_arguments.func(cmdline_arguments)
  File "/Users/johannes/rasa/rasa/rasa/cli/train.py", line 59, in <lambda>
    train_parser.set_defaults(func=lambda args: run_training(args, can_exit=True))
  File "/Users/johannes/rasa/rasa/rasa/cli/train.py", line 91, in run_training
    training_result = train_all(
  File "/Users/johannes/rasa/rasa/rasa/api.py", line 109, in train
    return rasa.utils.common.run_in_loop(
  File "/Users/johannes/rasa/rasa/rasa/utils/common.py", line 296, in run_in_loop
    result = loop.run_until_complete(f)
  File "uvloop/loop.pyx", line 1456, in uvloop.loop.Loop.run_until_complete
  File "/Users/johannes/rasa/rasa/rasa/model_training.py", line 108, in train_async
    return await _train_async_internal(
  File "/Users/johannes/rasa/rasa/rasa/model_training.py", line 288, in _train_async_internal
    await _do_training(
  File "/Users/johannes/rasa/rasa/rasa/model_training.py", line 352, in _do_training
    await _train_core_with_validated_data(
  File "/Users/johannes/rasa/rasa/rasa/model_training.py", line 549, in _train_core_with_validated_data
    await rasa.core.train.train(
  File "/Users/johannes/rasa/rasa/rasa/core/train.py", line 70, in train
    agent.train(training_data, **additional_arguments)
  File "/Users/johannes/rasa/rasa/rasa/core/agent.py", line 753, in train
    self.policy_ensemble.train(
  File "/Users/johannes/rasa/rasa/rasa/core/policies/ensemble.py", line 206, in train
    policy.train(
  File "/Users/johannes/rasa/rasa/rasa/core/policies/ted_policy.py", line 687, in train
    model_data, label_ids = self._prepare_for_training(
  File "/Users/johannes/rasa/rasa/rasa/core/policies/ted_policy.py", line 573, in _prepare_for_training
    tracker_state_features, label_ids, entity_tags = self._featurize_for_training(
  File "/Users/johannes/rasa/rasa/rasa/core/policies/policy.py", line 191, in _featurize_for_training
    state_features, label_ids, entity_tags = self.featurizer.featurize_trackers(
  File "/Users/johannes/rasa/rasa/rasa/core/featurizers/tracker_featurizers.py", line 391, in featurize_trackers
    label_ids = self._convert_labels_to_ids(trackers_as_labels, domain)
  File "/Users/johannes/rasa/rasa/rasa/core/featurizers/tracker_featurizers.py", line 959, in _convert_labels_to_ids
    label_ids = [
  File "/Users/johannes/rasa/rasa/rasa/core/featurizers/tracker_featurizers.py", line 960, in <listcomp>
    [domain.intents.index(intent) for intent in tracker_intents]
  File "/Users/johannes/rasa/rasa/rasa/core/featurizers/tracker_featurizers.py", line 960, in <listcomp>
    [domain.intents.index(intent) for intent in tracker_intents]
ValueError: 'blah' is not in list

Command or request that led to error:

rasa train

Content of configuration file (config.yml) (if relevant):

language: en

pipeline:
  - name: "SpacyNLP"
    model: "en_core_web_md"
  - name: "SpacyTokenizer"
  - name: "SpacyFeaturizer"
  - name: "DIETClassifier"
    entity_recognition: False
    epochs: 1

policies:
  - name: TEDPolicy
    max_history: 5
    epochs: 1
  - name: UnexpecTEDIntentPolicy
    max_history: 5
    epochs: 1
  - name: MemoizationPolicy
  - name: RulePolicy

Content of domain file (domain.yml) (if relevant):
Moodbot domain.

[dakshvar22]

@JEM-Mosig Moving it to the IntentTED Backlog as it's a known issue before the release.