def _generate_lookup_regex(
self, lookup_table: Dict[Text, Union[Text, List[Text]]]) -> Text:
"""creates a regex out of the contents of a lookup table file"""
lookup_elements = lookup_table["elements"]
elements_to_regex = []
# if it's a list, it should be the elements directly
if isinstance(lookup_elements, list):
elements_to_regex = lookup_elements
common_utils.raise_warning(
f"Directly including lookup tables as a list is deprecated since Rasa "
f"1.6.",
FutureWarning,
docs=DOCS_URL_TRAINING_DATA_NLU + "#lookup-tables",
)
# otherwise it's a file path.
else:
try:
f = open(lookup_elements,
"r",
encoding=rasa.utils.io.DEFAULT_ENCODING)
except OSError:
raise ValueError(
f"Could not load lookup table {lookup_elements}. "
f"Please make sure you've provided the correct path.")
with f:
for line in f:
new_element = line.strip()
if new_element:
elements_to_regex.append(new_element)
# sanitize the regex, escape special characters
elements_sanitized = [re.escape(e) for e in elements_to_regex]
# regex matching elements with word boundaries on either side
regex_string = "(?i)(\\b" + "\\b|\\b".join(elements_sanitized) + "\\b)"
return regex_string
def _get_events_from_request_body(request: Request) -> List[Event]:
events = request.json
if not isinstance(events, list):
events = [events]
events = [Event.from_parameters(event) for event in events]
events = [event for event in events if event]
if not events:
common_utils.raise_warning(
f"Append event called, but could not extract a valid event. "
f"Request JSON: {request.json}"
)
raise ErrorResponse(
400,
"BadRequest",
"Couldn't extract a proper event from the request body.",
{"parameter": "", "in": "body"},
)
return events
async def send_text_with_buttons(
self,
recipient_id: Text,
text: Text,
buttons: List[Dict[Text, Any]],
**kwargs: Any,
) -> None:
recipient = self.slack_channel or recipient_id
text_block = {
"type": "section",
"text": {
"type": "plain_text",
"text": text
}
}
if len(buttons) > 5:
raise_warning(
"Slack API currently allows only up to 5 buttons. "
"Since you added more than 5, slack will ignore all of them.")
return await self.send_text_message(recipient, text, **kwargs)
button_block = {"type": "actions", "elements": []}
for button in buttons:
button_block["elements"].append({
"type": "button",
"text": {
"type": "plain_text",
"text": button["title"]
},
"value": button["payload"],
})
await self.client.chat_postMessage(
channel=recipient,
as_user=True,
text=text,
blocks=[text_block, button_block],
)
def _load_from_module_string(
domain: Domain, store: EndpointConfig, event_broker: Optional[EventBroker] = None
) -> "TrackerStore":
"""Initializes a custom tracker.
Defaults to the InMemoryTrackerStore if the module path can not be found.
Args:
domain: defines the universe in which the assistant operates
store: the specific tracker store
event_broker: an event broker to publish events
Returns:
a tracker store from a specified type in a stores endpoint configuration
"""
try:
tracker_store_class = class_from_module_path(store.type)
init_args = arguments_of(tracker_store_class.__init__)
if "url" in init_args and "host" not in init_args:
raise_warning(
"The `url` initialization argument for custom tracker stores is "
"deprecated. Your custom tracker store should take a `host` "
"argument in its `__init__()` instead.",
DeprecationWarning,
)
store.kwargs["url"] = store.url
else:
store.kwargs["host"] = store.url
return tracker_store_class(
domain=domain, event_broker=event_broker, **store.kwargs
)
except (AttributeError, ImportError):
raise_warning(
f"Tracker store with type '{store.type}' not found. "
f"Using `InMemoryTrackerStore` instead."
)
return InMemoryTrackerStore(domain)
def _user_intent_from_step(self, step: Dict[Text, Any]) -> Text:
user_intent = step.get(KEY_USER_INTENT, "").strip()
if not user_intent:
common_utils.raise_warning(
f"Issue found in '{self.source_name}':\n"
f"User utterance cannot be empty. "
f"This {self._get_item_title()} step will be skipped:\n"
f"{step}",
docs=self._get_docs_link(),
)
if user_intent.startswith(INTENT_MESSAGE_PREFIX):
common_utils.raise_warning(
f"Issue found in '{self.source_name}':\n"
f"User intent '{user_intent}' starts with "
f"'{INTENT_MESSAGE_PREFIX}'. This is not required.",
docs=self._get_docs_link(),
)
# Remove leading slash
user_intent = user_intent[1:]
return user_intent
async def trigger_external_user_uttered(
self,
intent_name: Text,
entities: Optional[Union[List[Dict[Text, Any]], Dict[Text, Text]]],
tracker: DialogueStateTracker,
output_channel: OutputChannel,
) -> None:
"""Triggers an external message.
Triggers an external message (like a user message, but invisible;
used, e.g., by a reminder or the trigger_intent endpoint).
Args:
intent_name: Name of the intent to be triggered.
entities: Entities to be passed on.
tracker: The tracker to which the event should be added.
output_channel: The output channel.
"""
if isinstance(entities, list):
entity_list = entities
elif isinstance(entities, dict):
# Allow for a short-hand notation {"ent1": "val1", "ent2": "val2", ...}.
# Useful if properties like 'start', 'end', or 'extractor' are not given,
# e.g. for external events.
entity_list = [{
"entity": ent,
"value": val
} for ent, val in entities.items()]
elif not entities:
entity_list = []
else:
raise_warning(
f"Invalid entity specification: {entities}. Assuming no entities."
)
entity_list = []
tracker.update(UserUttered.create_external(intent_name, entity_list))
await self._predict_and_execute_next_action(output_channel, tracker)
# save tracker state to continue conversation from this state
self._save_tracker(tracker)
def _parse_step(self, step: Dict[Text, Any]) -> None:
if KEY_STORY_USER_UTTERANCE in step.keys():
self._parse_user_utterance(step)
elif KEY_OR in step.keys():
self._parse_or_statement(step)
elif KEY_SLOT_NAME in step.keys():
self._parse_slot(step)
elif KEY_ACTION in step.keys():
self._parse_action(step)
elif KEY_CHECKPOINT in step.keys():
self._parse_checkpoint(step)
elif KEY_METADATA in step.keys():
pass
else:
common_utils.raise_warning(
f"Issue found in '{self.source_name}': \n"
f"Found an unexpected step in the story description:\n"
f"{step}\n"
"It will be skipped.",
docs=DOCS_URL_STORIES,
)
def _check_OOV_present(self, all_tokens: List[List[Text]], attribute: Text) -> None:
"""Check if an OOV word is present"""
if not self.OOV_token or self.OOV_words or not all_tokens:
return
for tokens in all_tokens:
for text in tokens:
if self.OOV_token in text or (
self.lowercase and self.OOV_token in text.lower()
):
return
if any(text for tokens in all_tokens for text in tokens):
training_data_type = "NLU" if attribute == TEXT else "ResponseSelector"
# if there is some text in tokens, warn if there is no oov token
common_utils.raise_warning(
f"The out of vocabulary token '{self.OOV_token}' was configured, but "
f"could not be found in any one of the {training_data_type} "
f"training examples. All unseen words will be ignored during prediction.",
docs=DOCS_URL_COMPONENTS + "#countvectorsfeaturizer",
)
def validate_training_data_format_version(yaml_file_content: Dict[Text,
Any],
filename: Text) -> bool:
"""Validates version on the training data content using `version` field
and warns users if the file is not compatible with the current version of
Rasa Open Source.
Args:
yaml_file_content: Raw content of training data file as a dictionary.
filename: Name of the validated file.
Returns:
`True` if the file can be processed by current version of Rasa Open Source,
`False` otherwise.
"""
if not isinstance(yaml_file_content, dict):
raise ValueError(f"Failed to validate {filename}.")
version_value = yaml_file_content.get(KEY_TRAINING_DATA_FORMAT_VERSION)
if not version_value:
raise_warning(
f"Training data file {filename} doesn't have a "
f"'{KEY_TRAINING_DATA_FORMAT_VERSION}' key. "
f"Rasa Open Source will read the file as a "
f"version '{LATEST_TRAINING_DATA_FORMAT_VERSION}' file.",
docs=DOCS_BASE_URL,
)
return True
try:
parsed_version = version.parse(version_value)
if isinstance(parsed_version, LegacyVersion):
raise TypeError
if version.parse(
LATEST_TRAINING_DATA_FORMAT_VERSION) >= parsed_version:
return True
except TypeError:
raise_warning(
f"Training data file {filename} must specify "
f"'{KEY_TRAINING_DATA_FORMAT_VERSION}' as string, for example:\n"
f"{KEY_TRAINING_DATA_FORMAT_VERSION}: '{LATEST_TRAINING_DATA_FORMAT_VERSION}'\n"
f"Rasa Open Source will read the file as a "
f"version '{LATEST_TRAINING_DATA_FORMAT_VERSION}' file.",
docs=DOCS_BASE_URL,
)
return True
raise_warning(
f"Training data file {filename} has a greater format version than "
f"your Rasa Open Source installation: "
f"{version_value} > {LATEST_TRAINING_DATA_FORMAT_VERSION}. "
f"Please consider updating to the latest version of Rasa Open Source."
f"This file will be skipped.",
docs=DOCS_BASE_URL,
)
return False
async def read_from_file(self, filename: Text) -> List[StoryStep]:
try:
yaml_content = rasa.utils.io.read_yaml_file(filename)
if not isinstance(yaml_content, dict):
common_utils.raise_warning(
f"Failed to read '{filename}'. It should be a Yaml dict.")
return []
stories = yaml_content.get(KEY_STORIES) # pytype: disable=attribute-error
if not stories:
return []
self._parse_stories(stories)
self._add_current_stories_to_result()
return self.story_steps
except ValueError as e:
common_utils.raise_warning(
f"Failed to read YAML from '{filename}', it will be skipped. Error: {e}"
)
return []
def read_from_json(self, js: Dict[Text, Any], **_) -> "TrainingData":
"""Loads training data stored in the rasa NLU data format."""
from rasa.nlu.training_data import Message, TrainingData
import rasa.nlu.schemas.data_schema as schema
import rasa.utils.validation as validation_utils
validation_utils.validate_training_data(js,
schema.rasa_nlu_data_schema())
data = js["rasa_nlu_data"]
common_examples = data.get("common_examples", [])
intent_examples = data.get("intent_examples", [])
entity_examples = data.get("entity_examples", [])
entity_synonyms = data.get("entity_synonyms", [])
regex_features = data.get("regex_features", [])
lookup_tables = data.get("lookup_tables", [])
entity_synonyms = transform_entity_synonyms(entity_synonyms)
if intent_examples or entity_examples:
raise_warning(
"Your rasa data "
"contains 'intent_examples' "
"or 'entity_examples' which will be "
"removed in the future. Consider "
"putting all your examples "
"into the 'common_examples' section.",
FutureWarning,
docs=DOCS_URL_TRAINING_DATA_NLU,
)
all_examples = common_examples + intent_examples + entity_examples
training_examples = []
for ex in all_examples:
msg = Message.build(**ex)
training_examples.append(msg)
return TrainingData(training_examples, entity_synonyms, regex_features,
lookup_tables)
def _parse_checkpoint(self, step: Dict[Text, Any]) -> None:
checkpoint_name = step.get(KEY_CHECKPOINT, "")
slots = step.get(KEY_CHECKPOINT_SLOTS, [])
slots_dict = {}
for slot in slots:
if not isinstance(slot, dict):
common_utils.raise_warning(
f"Issue found in '{self.source_name}':\n"
f"Checkpoint '{checkpoint_name}' has an invalid slot: "
f"{slots}\nItems under the '{KEY_CHECKPOINT_SLOTS}' key must be "
f"YAML dictionaries. The checkpoint will be skipped.",
docs=self._get_docs_link(),
)
return
for key, value in slot.items():
slots_dict[key] = value
self._add_checkpoint(checkpoint_name, slots_dict)
async def scheduler() -> AsyncIOScheduler:
"""Thread global scheduler to handle all recurring tasks.
If no scheduler exists yet, this will instantiate one."""
global __scheduler
if not __scheduler:
try:
__scheduler = AsyncIOScheduler(event_loop=asyncio.get_event_loop())
__scheduler.start()
return __scheduler
except UnknownTimeZoneError:
raise_warning(
"apscheduler could not find a timezone and is "
"defaulting to utc. This is probably because "
"your system timezone is not set. "
'Set it with e.g. echo "Europe/Berlin" > '
"/etc/timezone"
)
__scheduler = AsyncIOScheduler(
event_loop=asyncio.get_event_loop(), timezone=utc
)
__scheduler.start()
return __scheduler
else:
# scheduler already created, make sure it is running on
# the correct loop
# noinspection PyProtectedMember
if not __scheduler._eventloop == asyncio.get_event_loop():
raise RuntimeError(
"Detected inconsistent loop usage. "
"Trying to schedule a task on a new event "
"loop, but scheduler was created with a "
"different event loop. Make sure there "
"is only one event loop in use and that the "
"scheduler is running on that one."
)
return __scheduler
def get_tests_from_file(filename: Optional[Text] = None) -> List[Text]:
"""Returns an list of tests from a `filename`.
Args:
filename: Path to a test file.
"""
if not filename:
filename = utils.get_project_directory() / DEFAULT_FILENAME
try:
content = io_utils.read_file(filename)
return _split_tests(content)
except ValueError as e:
rasa_utils.raise_warning(
f"Unable to get tests from {filename}:\n{e} "
f"Please, make sure you have end-to-end tests added to your assistant. "
f"See https://rasa.com/docs/rasa/user-guide/testing-your-assistant/ "
f"for more information.",
UserWarning,
)
return []
def collect_templates(
yml_templates: Dict[Text, List[Any]]
) -> Dict[Text, List[Dict[Text, Any]]]:
"""Go through the templates and make sure they are all in dict format."""
templates = {}
for template_key, template_variations in yml_templates.items():
validated_variations = []
if template_variations is None:
raise InvalidDomain(
"Utterance '{}' does not have any defined templates.".format(
template_key
)
)
for t in template_variations:
# templates should be a dict with options
if isinstance(t, str):
raise_warning(
f"Templates should not be strings anymore. "
f"Utterance template '{template_key}' should contain "
f"either a '- text: ' or a '- custom: ' "
f"attribute to be a proper template.",
FutureWarning,
docs=DOCS_URL_DOMAINS + "#utterance-templates",
)
validated_variations.append({"text": t})
elif "text" not in t and "custom" not in t:
raise InvalidDomain(
f"Utter template '{template_key}' needs to contain either "
f"'- text: ' or '- custom: ' attribute to be a proper "
f"template."
)
else:
validated_variations.append(t)
templates[template_key] = validated_variations
return templates
def _parse_training_examples(
self, examples: Union[Text, List[Dict[Text, Any]]], intent: Text
) -> List[Tuple[Text, List[Dict[Text, Any]]]]:
import rasa.nlu.training_data.entities_parser as entities_parser
if isinstance(examples, list):
example_strings = [
# pytype: disable=attribute-error
example.get(KEY_INTENT_TEXT, "")
for example in examples
if example
]
# pytype: enable=attribute-error
elif isinstance(examples, str):
example_strings = self._parse_multiline_example(intent, examples)
else:
raise_warning(
f"Unexpected block found in '{self.filename}' "
f"while processing intent '{intent}':\n"
f"{examples}\n"
f"This block will be skipped.",
docs=DOCS_URL_TRAINING_DATA_NLU,
)
return []
if not example_strings:
raise_warning(
f"Issue found while processing '{self.filename}': "
f"Intent '{intent}' has no examples.",
docs=DOCS_URL_TRAINING_DATA_NLU,
)
results = []
for example in example_strings:
entities = entities_parser.find_entities_in_training_example(example)
results.append((example, entities))
return results
def read_from_json(self, js, **kwargs) -> "TrainingData":
"""Loads training data stored in the rasa NLU data format."""
from rasa.nlu.training_data import Message, TrainingData
validate_rasa_nlu_data(js)
data = js["rasa_nlu_data"]
common_examples = data.get("common_examples", [])
intent_examples = data.get("intent_examples", [])
entity_examples = data.get("entity_examples", [])
entity_synonyms = data.get("entity_synonyms", [])
regex_features = data.get("regex_features", [])
lookup_tables = data.get("lookup_tables", [])
gazette = data.get("gazette", [])
entity_synonyms = transform_entity_synonyms(entity_synonyms)
if intent_examples or entity_examples:
raise_warning(
"Your rasa data "
"contains 'intent_examples' "
"or 'entity_examples' which will be "
"removed in the future. Consider "
"putting all your examples "
"into the 'common_examples' section.",
FutureWarning,
docs=DOCS_URL_TRAINING_DATA_NLU,
)
all_examples = common_examples + intent_examples + entity_examples
training_examples = []
for ex in all_examples:
msg = Message.build(ex["text"], ex.get("intent"),
ex.get("entities"))
training_examples.append(msg)
return TrainingData(training_examples, entity_synonyms, regex_features,
lookup_tables, gazette)
async def handle_message(sid: Text, data: Dict) -> Any:
output_channel = SocketIOOutput(sio, sid, self.bot_message_evt)
if self.session_persistence:
if not data.get("session_id"):
raise_warning("A message without a valid session_id "
"was received. This message will be "
"ignored. Make sure to set a proper "
"session id using the "
"`session_request` socketIO event.")
return
sender_id = data["session_id"]
else:
sender_id = sid
message = UserMessage(
data["message"],
output_channel,
sender_id,
input_channel=self.name(),
metadata=data["metadata"],
)
await on_new_message(message)
def add_checkpoint(self, name: Text, conditions: Optional[Dict[Text, Any]]) -> None:
# Depending on the state of the story part this
# is either a start or an end check point
if not self.current_steps:
self.start_checkpoints.append(Checkpoint(name, conditions))
else:
if conditions:
common_utils.raise_warning(
f"End or intermediate checkpoints "
f"do not support conditions! "
f"(checkpoint: {name})",
docs=DOCS_URL_STORIES + "#checkpoints",
)
additional_steps = []
for t in self.current_steps:
if t.end_checkpoints:
tcp = t.create_copy(use_new_id=True)
tcp.end_checkpoints = [Checkpoint(name)]
additional_steps.append(tcp)
else:
t.end_checkpoints = [Checkpoint(name)]
self.current_steps.extend(additional_steps)
def _setup_gpu_environment() -> None:
"""Set configuration for TensorFlow GPU environment based on the environment variable set."""
gpu_memory_config = os.getenv(ENV_GPU_CONFIG)
if not gpu_memory_config:
return
# Import from tensorflow only if necessary (environment variable was set)
from tensorflow import config as tf_config
parsed_gpu_config = _parse_gpu_config(gpu_memory_config)
physical_gpus = tf_config.list_physical_devices("GPU")
# Logic taken from https://www.tensorflow.org/guide/gpu
if physical_gpus:
for gpu_id, gpu_id_memory in parsed_gpu_config.items():
_allocate_gpu_memory(physical_gpus[gpu_id], gpu_id_memory)
else:
rasa_utils.raise_warning(
f"You have an environment variable '{ENV_GPU_CONFIG}' set but no GPUs were detected to configure."
)
def verify_example_repetition_in_intents(self,
ignore_warnings: bool = True
) -> bool:
"""Checks if there is no duplicated example in different intents."""
everything_is_alright = True
duplication_hash = defaultdict(set)
for example in self.intents.intent_examples:
text = example.text
duplication_hash[text].add(example.get("intent"))
for text, intents in duplication_hash.items():
if len(duplication_hash[text]) > 1:
everything_is_alright = ignore_warnings and everything_is_alright
intents_string = ", ".join(sorted(intents))
raise_warning(
f"The example '{text}' was found labeled with multiple "
f"different intents in the training data. Each annotated message "
f"should only appear with one intent. You should fix that "
f"conflict The example is labeled with: {intents_string}.")
return everything_is_alright
def as_feature(self) -> List[float]:
r = [0.0] * self.feature_dimensionality()
try:
for i, v in enumerate(self.values):
if v == str(self.value).lower():
r[i] = 1.0
break
else:
if self.value is not None:
raise_warning(
f"Categorical slot '{self.name}' is set to a value "
f"('{self.value}') "
"that is not specified in the domain. "
"Value will be ignored and the slot will "
"behave as if no value is set. "
"Make sure to add all values a categorical "
"slot should store to the domain."
)
except (TypeError, ValueError):
logger.exception("Failed to featurize categorical slot.")
return r
return r
def __init__(
self,
component_config: Optional[Dict[Text, Any]] = None,
clf: "sklearn.model_selection.GridSearchCV" = None,
le: Optional["sklearn.preprocessing.LabelEncoder"] = None,
) -> None:
"""Construct a new intent classifier using the sklearn framework."""
from sklearn.preprocessing import LabelEncoder
super().__init__(component_config)
if le is not None:
self.le = le
else:
self.le = LabelEncoder()
self.clf = clf
common_utils.raise_warning(
"'SklearnIntentClassifier' is deprecated and will be removed in version "
"2.0. Use 'DIETClassifier' instead.",
category=FutureWarning,
docs=DOCS_URL_MIGRATION_GUIDE,
)
def validate_required_components_from_data(
pipeline: List["Component"], data: TrainingData
) -> None:
"""Validates that all components are present in the pipeline based on data.
Args:
pipeline: The list of the :class:`rasa.nlu.components.Component`.
data: The :class:`rasa.nlu.training_data.training_data.TrainingData`.
"""
from rasa.nlu.selectors.response_selector import ResponseSelector
response_selector_exists = False
for component in pipeline:
# check if a response selector is part of NLU pipeline
if isinstance(component, ResponseSelector):
response_selector_exists = True
if len(data.response_examples) and not response_selector_exists:
raise_warning(
"Training data consists examples for training a response selector but "
"no response selector component specified inside NLU pipeline."
)
def process(self, message: Message, **kwargs: Any) -> None:
if self._url() is not None:
reference_time = self._reference_time_from_message(message)
matches = self._duckling_parse(message.text, reference_time)
all_extracted = convert_duckling_format_to_rasa(matches)
dimensions = self.component_config["dimensions"]
extracted = DucklingHTTPExtractor.filter_irrelevant_entities(
all_extracted, dimensions)
else:
extracted = []
raise_warning(
"Duckling HTTP component in pipeline, but no "
"`url` configuration in the config "
"file nor is `RASA_DUCKLING_HTTP_URL` "
"set as an environment variable. No entities will be extracted!",
docs=DOCS_URL_COMPONENTS + "#ducklinghttpextractor",
)
extracted = self.add_extractor_name(extracted)
message.set(ENTITIES,
message.get(ENTITIES, []) + extracted,
add_to_output=True)
def handle_channels(
self,
channels: List[InputChannel],
http_port: int = constants.DEFAULT_SERVER_PORT,
route: Text = "/webhooks/",
cors: Union[Text, List[Text], None] = None,
) -> Sanic:
"""Start a webserver attaching the input channels and handling msgs."""
from rasa.core import run
raise_warning(
"Using `handle_channels` is deprecated. "
"Please use `rasa.run(...)` or see "
"`rasa.core.run.configure_app(...)` if you want to implement "
"this on a more detailed level.",
DeprecationWarning,
)
app = run.configure_app(channels, cors, None, enable_api=False, route=route)
app.agent = self
update_sanic_log_level()
app.run(
host="0.0.0.0",
port=http_port,
backlog=int(os.environ.get(ENV_SANIC_BACKLOG, "100")),
workers=rasa.core.utils.number_of_sanic_workers(self.lock_store),
)
# this might seem unnecessary (as run does not return until the server
# is killed) - but we use it for tests where we mock `.run` to directly
# return and need the app to inspect if we created a properly
# configured server
return app
def create(
obj: Union["NaturalLanguageInterpreter", EndpointConfig, Text, None],
# this second parameter is deprecated!
endpoint: Optional[EndpointConfig] = None,
) -> "NaturalLanguageInterpreter":
"""Factory to create an natural language interpreter."""
if endpoint is not None:
raise_warning(
"Calling `NaturalLanguageInterpreter.create` with two parameters"
"is deprecated. The `endpoint` parameter will be removed in the "
"future. You should replace a call "
"`NaturalLanguageInterpreter.create(s, e)` "
"with the single parameter version "
"`NaturalLanguageInterpreter.create(e or s)`.",
category=DeprecationWarning,
)
obj = endpoint or obj
# <bf
from rasa.nlu.model import UnsupportedModelError
try:
if isinstance(obj, NaturalLanguageInterpreter):
return obj
elif isinstance(obj, str) and os.path.exists(obj):
return RasaNLUInterpreter(model_directory=obj)
elif isinstance(obj, str) and not os.path.exists(obj):
# user passed in a string, but file does not exist
logger.warning(
f"No local NLU model '{obj}' found. Using RegexInterpreter instead."
)
return RegexInterpreter()
else:
return _create_from_endpoint_config(obj)
except UnsupportedModelError as e:
logger.warning(e.message)
return RegexInterpreter()
def train(
self,
training_data: TrainingData,
config: Optional[RasaNLUModelConfig] = None,
**kwargs: Any,
) -> None:
"""Train the intent classifier on a data set."""
num_threads = kwargs.get("num_threads", 1)
labels = [e.get("intent") for e in training_data.intent_examples]
if len(set(labels)) < 2:
common_utils.raise_warning(
"Can not train an intent classifier as there are not "
"enough intents. Need at least 2 different intents. "
"Skipping training of intent classifier.",
docs=DOCS_URL_TRAINING_DATA_NLU,
)
else:
y = self.transform_labels_str2num(labels)
X = np.stack([
sequence_to_sentence_features(
example.get(DENSE_FEATURE_NAMES[TEXT]))
for example in training_data.intent_examples
])
# reduce dimensionality
X = np.reshape(X, (len(X), -1))
self.clf = self._create_classifier(num_threads, y)
with warnings.catch_warnings():
# sklearn raises lots of
# "UndefinedMetricWarning: F - score is ill - defined"
# if there are few intent examples, this is needed to prevent it
warnings.simplefilter("ignore")
self.clf.fit(X, y)
def handle_deprecated_request_parameters(
request: Request, old_name: Text, new_name: Text
) -> None:
""" Modify a request to account for a deprecated request parameter following a rename.
Replace the deprecated parameter in the input request with the corresponding new
one. Also do this when the parameter is used as fields parameter.
Args:
request: The request to fix.
old_name: The deprecated name of the parameter.
new_name: The new name of the parameter.
"""
if request.args.get(old_name):
rasa_utils.raise_warning(
f"Your request includes the parameter '{old_name}'. This has been "
f"deprecated and renamed to '{new_name}'. The '{old_name}' parameter will "
f"no longer work in future versions of Rasa X. Please use '{new_name}' "
f"instead.",
FutureWarning,
)
request.args[new_name] = request.args.pop(old_name)
keys = [k for k in request.args.keys() if "fields" in k]
for k in keys:
new_k = "fields"
# get the content between brackets, [<CONTENT>], as a list, e.g. "fields[a][b]"
# becomes [a, b]
d = re.findall(r"\[(.*?)\]", k)
for field in d:
if field == old_name:
new_k += f"[{new_name}]"
else:
new_k += f"[{field}]"
request.args[new_k] = request.args.pop(k)
def _check_deprecated_attributes(component: "Component") -> None:
"""Checks that the component doesn't have deprecated attributes.
Args:
component: The :class:`rasa.nlu.components.Component`.
"""
if hasattr(component, "provides"):
raise_warning(
f"'{component.name}' contains property 'provides', "
f"which is deprecated. There is no need to specify "
f"the list of attributes that a component provides.",
category=FutureWarning,
docs=DOCS_URL_MIGRATION_GUIDE,
)
if hasattr(component, "requires"):
raise_warning(
f"'{component.name}' contains property 'requires', "
f"which is deprecated. Use 'required_components()' method "
f"to specify which components are required to be present "
f"in the pipeline by this component.",
category=FutureWarning,
docs=DOCS_URL_MIGRATION_GUIDE,
)
