Merge pull request #1088 from RasaHQ/enchanced-cli

Interactive learning enhancements.  Fixes #981

.gitignore

@@ -41,12 +41,10 @@ tmp/
 tokens.dat
 graph.png
 debug.md
-examples/moodbot/models
-examples/restaurantbot/models
-examples/concertbot/models
-examples/moodbot/*.png
-examples/moodbot/errors.json
+examples/concertbot/data*
+examples/concertbot/models*
+examples/moodbot/models*
 docs/key 
 docs/key.pub
 secrets.tar
-failed_stories.md
+failed_stories.md

CHANGELOG.rst

@@ -14,9 +14,13 @@ This project adheres to `Semantic Versioning`_ starting with version 0.2.0.
 Added
 -----
 - openapi documentation of server API
+- NLU data learned through interactive learning will now be stored in a separate markdown-format file (any previous NLU
+  data is merged)
+- Command line interface for interactive learning now displays policy confidence alongside the action name
 - added action prediction confidence & policy to ``ActionExecuted`` event
 - both the date and the time at which a model was trained are now included in the policy's metadata when it is persisted
 
+
 Changed
 -------
 - improved response format for ``/predict`` endpoint
@@ -26,6 +30,7 @@ Changed
 Removed
 -------
 
+
 Fixed
 -----
 - fixed an issue with boolean slots where False and None had the same value

data/test_trackers/tracker_moodbot.json

@@ -120,4 +120,4 @@
       "confidence": 1.0
     }
   ]
-}
+}

docs/interactive_learning.rst

@@ -74,7 +74,8 @@ so make sure you have the domain & data for it. You can download
 the data from :gh-code:`examples/concertbot`.
 
 If you ask ``/search_concerts``, the bot should suggest
-``action_search_concerts`` and then ``action_listen``.
+``action_search_concerts`` and then ``action_listen`` (the confidence at which
+the policy selected its next action will be displayed next to the action name).
 Now let's enter ``/compare_reviews`` as the next user message.
 The bot **might** choose the wrong one out of the two
 possibilities (depending on the training run, it might also be correct):
@@ -91,8 +92,8 @@ possibilities (depending on the training run, it might also be correct):
      2                                            /search_concerts
                                       intent: search_concerts 1.00
     ───────────────────────────────────────────────────────────────
-     3    action_search_concerts
-          action_listen
+     3    action_search_concerts 0.72
+          action_listen 0.78
     ───────────────────────────────────────────────────────────────
      4                                            /compare_reviews
                                       intent: compare_reviews 1.00

examples/concertbot/Makefile

@@ -21,6 +21,6 @@ run-actions:
 train-core:
 	python train.py
 
-run-online:
+run-interactive:
 	make run-actions&
-	python train_online.py
+	python train_interactive.py

examples/concertbot/train_online.py → examples/concertbot/train_interactive.py

@@ -6,7 +6,7 @@
 import logging
 
 from rasa_core import utils, train, run
-from rasa_core.training import online
+from rasa_core.training import interactive
 
 logger = logging.getLogger(__name__)
 
@@ -25,4 +25,4 @@ def train_agent():
 if __name__ == '__main__':
     utils.configure_colored_logging(loglevel="INFO")
     agent = train_agent()
-    online.run_online_learning(agent)
+    interactive.run_interactive_learning(agent)

rasa_core/events/__init__.py

@@ -709,6 +709,7 @@ def __str__(self):
         return ("ActionExecuted(action: {}, policy: {}, confidence: {})"
                 "".format(self.action_name, self.policy, self.confidence))
 
+
     def __hash__(self):
         return hash(self.action_name)
 

rasa_core/training/interactive.py

@@ -31,12 +31,17 @@
 from rasa_core.training.structures import Story
 from rasa_core.utils import EndpointConfig
 from rasa_nlu.training_data.formats import MarkdownWriter, MarkdownReader
+from rasa_nlu.training_data.loading import load_data, _guess_format
+from rasa_nlu.training_data.message import Message
+from rasa_nlu.training_data import TrainingData
 
 logger = logging.getLogger(__name__)
 
 MAX_VISUAL_HISTORY = 3
 
-DEFAULT_FILE_EXPORT_PATH = "stories.md"
+PATHS = {"stories": "data/stories.md",
+         "nlu": "data/nlu.md",
+         "backup": "data/nlu_interactive.md"}
 
 # choose other intent, making sure this doesn't clash with an existing intent
 OTHER_INTENT = uuid.uuid4().hex
@@ -138,7 +143,7 @@ def send_action(endpoint, sender_id, action_name):
 
 def send_event(endpoint, sender_id, evt):
     # type: (EndpointConfig, Text, Dict[Text, Any]) -> Dict[Text, Any]
-    """Log an event to a concersation."""
+    """Log an event to a conversation."""
 
     subpath = "/conversations/{}/tracker/events".format(sender_id)
 
@@ -151,7 +156,7 @@ def send_event(endpoint, sender_id, evt):
 
 def replace_events(endpoint, sender_id, evts):
     # type: (EndpointConfig, Text, List[Dict[Text, Any]]) -> Dict[Text, Any]
-    """Replace all the events of a concersation with the provided ones."""
+    """Replace all the events of a conversation with the provided ones."""
 
     subpath = "/conversations/{}/tracker/events".format(sender_id)
 
@@ -394,6 +399,8 @@ def add_user_cell(data, cell):
     for idx, evt in enumerate(evts):
         if evt.get("event") == "action":
             bot_column.append(colored(evt['name'], 'autocyan'))
+            if evt['confidence'] is not None:
+                bot_column[-1] += (colored(" {:03.2f}".format(evt['confidence']), 'autowhite'))
 
         elif evt.get("event") == 'user':
             if bot_column:
@@ -470,10 +477,15 @@ def _ask_if_quit(sender_id, endpoint):
 
     if not answers or answers["abort"] == "quit":
         # this is also the default answer if the user presses Ctrl-C
-        export_file_path = _request_export_stories_info()
-        _write_stories_to_file(export_file_path, sender_id, endpoint)
-        logger.info("Successfully wrote stories to "
-                    "{}.".format(export_file_path))
+        story_path, nlu_path = _request_export_info()
+
+        tracker = retrieve_tracker(endpoint, sender_id)
+        evts = tracker.get("events", [])
+
+        _write_stories_to_file(story_path, evts)
+        _write_nlu_to_file(nlu_path, evts)
+
+        logger.info("Successfully wrote stories and NLU data")
         sys.exit()
     elif answers["abort"] == "continue":
         # in this case we will just return, and the original
@@ -511,9 +523,9 @@ def _request_action_from_user(predictions, sender_id, endpoint):
     return action_name
 
 
-def _request_export_stories_info():
-    # type: () -> Text
-    """Request file path and export stories to that path"""
+def _request_export_info():
+    # type: () -> (Text, Text)
+    """Request file path and export stories & nlu data to that path"""
 
     def validate_path(path):
         try:
@@ -522,26 +534,31 @@ def validate_path(path):
         except Exception as e:
             return "Failed to open file. {}".format(e)
 
-    # export current stories and quit
+    # export training data and quit
     questions = [{
-        "name": "export",
+        "name": "export stories",
         "type": "input",
         "message": "Export stories to (if file exists, this "
                    "will append the stories)",
-        "default": DEFAULT_FILE_EXPORT_PATH,
+        "default": PATHS["stories"],
         "validate": validate_path
-    }]
+    }, {"name": "export nlu",
+        "type": "input",
+        "message": "Export NLU data to (if file exists, this "
+                   "will merge learned data with previous training examples)",
+        "default": PATHS["nlu"],
+        "validate": validate_path}]
 
     answers = prompt(questions)
     if not answers:
         sys.exit()
 
-    return answers["export"]
+    return answers["export stories"], answers["export nlu"]
 
 
 def _split_conversation_at_restarts(evts):
     # type: (List[Dict[Text, Any]]) -> List[List[Dict[Text, Any]]]
-    """"Split a conversation at restart events.
+    """Split a conversation at restart events.
 
     Returns an array of event lists, without the restart events."""
 
@@ -561,22 +578,63 @@ def _split_conversation_at_restarts(evts):
     return sub_conversations
 
 
-def _write_stories_to_file(export_file_path, sender_id, endpoint):
-    # type: (Text, Text, EndpointConfig) -> None
-    """Write the conversation of the sender_id to the file path."""
+def _collect_messages(evts):
+    # type: (List[Dict[Text, Any]]) -> List[Dict[Text, Any]]
+    """Collect the message text and parsed data from the UserMessage events into a list"""
+
+    msgs = []
+
+    for evt in evts:
+        if evt.get("event") == "user":
+            data = evt.get("parse_data")
+            msg = Message.build(data["text"], data["intent"]["name"], data["entities"])
+            msgs.append(msg)
+
+    return msgs
+
 
-    tracker = retrieve_tracker(endpoint, sender_id)
-    evts = tracker.get("events", [])
+def _write_stories_to_file(export_story_path, evts):
+    # type: (Text, List[Dict[Text, Any]]) -> None
+    """Write the conversation of the sender_id to the file paths."""
 
     sub_conversations = _split_conversation_at_restarts(evts)
 
-    with io.open(export_file_path, 'a', encoding="utf-8") as f:
+    with io.open(export_story_path, 'a', encoding="utf-8") as f:
         for conversation in sub_conversations:
             parsed_events = events.deserialise_events(conversation)
             s = Story.from_events(parsed_events)
             f.write(s.as_story_string(flat=True) + "\n")
 
 
+def _write_nlu_to_file(export_nlu_path, evts):
+    # type: (Text, List[Dict[Text, Any]]) -> None
+    """Write the nlu data of the sender_id to the file paths."""
+
+    msgs = _collect_messages(evts)
+
+    try:
+        previous_examples = load_data(export_nlu_path)
+
+    except:
+        questions = [{"name": "export nlu",
+                     "type": "input",
+                     "message": "Could not load existing NLU data, please specify where to store NLU data "
+                                "learned in this session (this will overwrite any existing file)",
+                     "default": PATHS["backup"]}]
+
+        answers = prompt(questions)
+        export_nlu_path = answers["export nlu"]
+        previous_examples = TrainingData()
+
+    nlu_data = previous_examples.merge(TrainingData(msgs))
+
+    with io.open(export_nlu_path, 'w', encoding="utf-8") as f:
+        if _guess_format(export_nlu_path) in ["md", "unk"]:
+            f.write(nlu_data.as_markdown())
+        else:
+            f.write(nlu_data.as_json())
+
+
 def _predict_till_next_listen(endpoint,  # type: EndpointConfig
                               sender_id,  # type: Text
                               finetune  # type: bool
@@ -784,8 +842,8 @@ def _enter_user_message(sender_id, endpoint, exit_text):
     }]
 
     answers = _ask_questions(
-            questions, sender_id, endpoint,
-            is_abort=lambda a: a["message"] == exit_text)
+        questions, sender_id, endpoint,
+        is_abort=lambda a: a["message"] == exit_text)
 
     send_message(endpoint, sender_id, answers["message"])
 

tests/test_server.py

@@ -88,6 +88,8 @@ def test_requesting_non_existent_tracker(app):
                                   "policy": None,
                                   "confidence": None,
                                   "name": "action_listen",
+                                  "policy": None,
+                                  "confidence": None,
                                   "timestamp": 1514764800}]
     assert content["latest_message"] == {"text": None,
                                          "intent": {},