async def train_comparison_models(
story_file: Text,
domain: Text,
output_path: Text = "",
exclusion_percentages: Optional[List] = None,
policy_configs: Optional[List] = None,
runs: int = 1,
dump_stories: bool = False,
additional_arguments: Optional[Dict] = None,
):
"""Train multiple models for comparison of policies"""
from rasa import model
from rasa.importers.importer import TrainingDataImporter
exclusion_percentages = exclusion_percentages or []
policy_configs = policy_configs or []
for r in range(runs):
logging.info("Starting run {}/{}".format(r + 1, runs))
for current_run, percentage in enumerate(exclusion_percentages, 1):
for policy_config in policy_configs:
file_importer = TrainingDataImporter.load_core_importer_from_config(
policy_config, domain, [story_file])
config_name = os.path.splitext(
os.path.basename(policy_config))[0]
logging.info("Starting to train {} round {}/{}"
" with {}% exclusion"
"".format(config_name, current_run,
len(exclusion_percentages), percentage))
with TempDirectoryPath(tempfile.mkdtemp()) as train_path:
_, new_fingerprint = await asyncio.gather(
train(
domain,
file_importer,
train_path,
policy_config=policy_config,
exclusion_percentage=percentage,
additional_arguments=additional_arguments,
dump_stories=dump_stories,
),
model.model_fingerprint(file_importer),
)
output_dir = os.path.join(output_path, "run_" + str(r + 1))
model_name = config_name + PERCENTAGE_KEY + str(percentage)
model.package_model(
fingerprint=new_fingerprint,
output_directory=output_dir,
train_path=train_path,
fixed_model_name=model_name,
)
async def _train_nlu_with_validated_data(
file_importer: TrainingDataImporter,
output: Text,
train_path: Optional[Text] = None,
fixed_model_name: Optional[Text] = None,
persist_nlu_training_data: bool = False,
retrain_nlu: Union[bool, List[Text]] = True) -> Optional[Text]:
"""Train NLU with validated training and config data."""
import rasa.nlu.train
import re
with ExitStack() as stack:
models = {}
from rasa.nlu import config as cfg_loader
if train_path:
# If the train path was provided, do nothing on exit.
_train_path = train_path
else:
# Otherwise, create a temp train path and clean it up on exit.
_train_path = stack.enter_context(
TempDirectoryPath(tempfile.mkdtemp()))
# bf mod
config = await file_importer.get_nlu_config(retrain_nlu)
for lang in config:
if config[lang]:
print_color("Start training {} NLU model ...".format(lang),
color=bcolors.OKBLUE)
_, models[lang], _ = await rasa.nlu.train(
config[lang],
file_importer, # config[lang]['path'],
_train_path,
fixed_model_name="nlu-{}".format(lang),
persist_nlu_training_data=persist_nlu_training_data,
)
else:
print_color(
"NLU data for language <{}> didn't change, skipping training..."
.format(lang),
color=bcolors.OKBLUE)
# /bf mod
print_color("NLU model training completed.", color=bcolors.OKBLUE)
if train_path is None:
# Only NLU was trained
new_fingerprint = await model.model_fingerprint(file_importer)
return model.package_model(
fingerprint=new_fingerprint,
output_directory=output,
train_path=_train_path,
fixed_model_name=fixed_model_name,
model_prefix="nlu-",
)
return _train_path
async def _train_nlu_with_validated_data(
file_importer: TrainingDataImporter,
output: Text,
train_path: Optional[Text] = None,
fixed_model_name: Optional[Text] = None,
persist_nlu_training_data: bool = False,
additional_arguments: Optional[Dict] = None,
) -> Optional[Text]:
"""Train NLU with validated training and config data."""
import rasa.nlu.train
if additional_arguments is None:
additional_arguments = {}
with ExitStack() as stack:
if train_path:
# If the train path was provided, do nothing on exit.
_train_path = train_path
else:
# Otherwise, create a temp train path and clean it up on exit.
_train_path = stack.enter_context(
TempDirectoryPath(tempfile.mkdtemp()))
config = await file_importer.get_config()
print_color("Training NLU model...",
color=rasa.shared.utils.io.bcolors.OKBLUE)
async with telemetry.track_model_training(file_importer,
model_type="nlu"):
await rasa.nlu.train(
config,
file_importer,
_train_path,
fixed_model_name="nlu",
persist_nlu_training_data=persist_nlu_training_data,
**additional_arguments,
)
print_color("NLU model training completed.",
color=rasa.shared.utils.io.bcolors.OKBLUE)
if train_path is None:
# Only NLU was trained
new_fingerprint = await model.model_fingerprint(file_importer)
return model.package_model(
fingerprint=new_fingerprint,
output_directory=output,
train_path=_train_path,
fixed_model_name=fixed_model_name,
model_prefix="nlu-",
)
return _train_path
async def _train_core_with_validated_data(
file_importer: TrainingDataImporter,
output: Text,
train_path: Optional[Text] = None,
fixed_model_name: Optional[Text] = None,
additional_arguments: Optional[Dict] = None,
interpreter: Optional[Interpreter] = None,
) -> Optional[Text]:
"""Train Core with validated training and config data."""
import rasa.core.train
with ExitStack() as stack:
if train_path:
# If the train path was provided, do nothing on exit.
_train_path = train_path
else:
# Otherwise, create a temp train path and clean it up on exit.
_train_path = stack.enter_context(
TempDirectoryPath(tempfile.mkdtemp()))
# normal (not compare) training
print_color("Training Core model...",
color=rasa.shared.utils.io.bcolors.OKBLUE)
domain, config = await asyncio.gather(file_importer.get_domain(),
file_importer.get_config())
async with telemetry.track_model_training(file_importer,
model_type="core"):
await rasa.core.train(
domain_file=domain,
training_resource=file_importer,
output_path=os.path.join(_train_path,
DEFAULT_CORE_SUBDIRECTORY_NAME),
policy_config=config,
additional_arguments=additional_arguments,
interpreter=interpreter,
)
print_color("Core model training completed.",
color=rasa.shared.utils.io.bcolors.OKBLUE)
if train_path is None:
# Only Core was trained.
new_fingerprint = await model.model_fingerprint(file_importer)
return model.package_model(
fingerprint=new_fingerprint,
output_directory=output,
train_path=_train_path,
fixed_model_name=fixed_model_name,
model_prefix="core-",
)
return _train_path
def __init__(self, agentName, botconfig, data, **kwargs):
logger.info("Training Agent " + agentName + " in progress")
trainingData = load_data(data)
self.intents = list(trainingData.intents)
self.entities = list(trainingData.entities)
trainer = Trainer(config.load(botconfig))
self.interpreter = trainer.train(trainingData)
self.model_path = "./models/" + agentName + "/"
persist_path = trainer.persist(self.model_path)
self.tar_path = package_model(fingerprint=None,
train_path=persist_path,
output_directory=self.model_path)
self.model_name = self.tar_path.replace(self.model_path, "")
self.model_version = self.model_name[:self.model_name.index(".tar.gz")]
async def _train_core_with_validated_data(
file_importer: TrainingDataImporter,
output: Text,
train_path: Optional[Text] = None,
fixed_model_name: Optional[Text] = None,
kwargs: Optional[Dict] = None,
) -> Optional[Text]:
"""Train Core with validated training and config data."""
import rasa.core.train
with ExitStack() as stack:
if train_path:
# If the train path was provided, do nothing on exit.
_train_path = train_path
else:
# Otherwise, create a temp train path and clean it up on exit.
_train_path = stack.enter_context(
TempDirectoryPath(tempfile.mkdtemp()))
# normal (not compare) training
print_color("Training Core model...", color=bcolors.OKBLUE)
domain = await file_importer.get_domain()
# bf mod
# config = await file_importer.get_config()
config = await file_importer.get_core_config()
# /bf mod
await rasa.core.train(domain_file=domain,
training_resource=file_importer,
output_path=os.path.join(_train_path, "core"),
policy_config=config,
kwargs=kwargs)
print_color("Core model training completed.", color=bcolors.OKBLUE)
if train_path is None:
# Only Core was trained.
new_fingerprint = await model.model_fingerprint(file_importer)
return model.package_model(
fingerprint=new_fingerprint,
output_directory=output,
train_path=_train_path,
fixed_model_name=fixed_model_name,
model_prefix="core-",
)
return _train_path
async def _train_nlu_with_validated_data(
file_importer: TrainingDataImporter,
output: Text,
train_path: Optional[Text] = None,
fixed_model_name: Optional[Text] = None,
) -> Optional[Text]:
"""Train NLU with validated training and config data."""
import rasa.nlu.train
with ExitStack() as stack:
if train_path:
# If the train path was provided, do nothing on exit.
_train_path = train_path
else:
# Otherwise, create a temp train path and clean it up on exit.
_train_path = stack.enter_context(
TempDirectoryPath(tempfile.mkdtemp()))
config = await file_importer.get_config()
print_color("Training NLU model...", color=bcolors.OKBLUE)
_, nlu_model, _ = await rasa.nlu.train(config,
file_importer,
_train_path,
fixed_model_name="nlu")
print_color("NLU model training completed.", color=bcolors.OKBLUE)
if train_path is None:
# Only NLU was trained
new_fingerprint = await model.model_fingerprint(file_importer)
return model.package_model(
fingerprint=new_fingerprint,
output_directory=output,
train_path=_train_path,
fixed_model_name=fixed_model_name,
model_prefix="nlu-",
)
return _train_path
async def _train_async_internal(
file_importer: TrainingDataImporter,
train_path: Text,
output_path: Text,
dry_run: bool,
force_training: bool,
fixed_model_name: Optional[Text],
persist_nlu_training_data: bool,
core_additional_arguments: Optional[Dict] = None,
nlu_additional_arguments: Optional[Dict] = None,
model_to_finetune: Optional[Text] = None,
finetuning_epoch_fraction: float = 1.0,
) -> TrainingResult:
"""Trains a Rasa model (Core and NLU). Use only from `train_async`.
Args:
file_importer: `TrainingDataImporter` which supplies the training data.
train_path: Directory in which to train the model.
output_path: Output path.
dry_run: If `True` then no training will be done, and the information about
whether the training needs to be done will be printed.
force_training: If `True` retrain model even if data has not changed.
fixed_model_name: Name of model to be stored.
persist_nlu_training_data: `True` if the NLU training data should be persisted
with the model.
core_additional_arguments: Additional training parameters for core training.
nlu_additional_arguments: Additional training parameters forwarded to training
method of each NLU component.
model_to_finetune: Optional path to a model which should be finetuned or
a directory in case the latest trained model should be used.
finetuning_epoch_fraction: The fraction currently specified training epochs
in the model configuration which should be used for finetuning.
Returns:
An instance of `TrainingResult`.
"""
stories, nlu_data = await asyncio.gather(file_importer.get_stories(),
file_importer.get_nlu_data())
new_fingerprint = await model.model_fingerprint(file_importer)
old_model = model.get_latest_model(output_path)
fingerprint_comparison = model.should_retrain(
new_fingerprint, old_model, train_path, force_training=force_training)
if dry_run:
code, texts = dry_run_result(fingerprint_comparison)
for text in texts:
print_warning(text) if code > 0 else print_success(text)
return TrainingResult(code=code)
if nlu_data.has_e2e_examples():
rasa.shared.utils.common.mark_as_experimental_feature(
"end-to-end training")
if stories.is_empty() and nlu_data.contains_no_pure_nlu_data():
rasa.shared.utils.cli.print_error(
"No training data given. Please provide stories and NLU data in "
"order to train a Rasa model using the '--data' argument.")
return TrainingResult()
if stories.is_empty():
rasa.shared.utils.cli.print_warning(
"No stories present. Just a Rasa NLU model will be trained.")
trained_model = await _train_nlu_with_validated_data(
file_importer,
output=output_path,
fixed_model_name=fixed_model_name,
persist_nlu_training_data=persist_nlu_training_data,
additional_arguments=nlu_additional_arguments,
model_to_finetune=model_to_finetune,
finetuning_epoch_fraction=finetuning_epoch_fraction,
)
return TrainingResult(model=trained_model)
# We will train nlu if there are any nlu example, including from e2e stories.
if nlu_data.contains_no_pure_nlu_data(
) and not nlu_data.has_e2e_examples():
rasa.shared.utils.cli.print_warning(
"No NLU data present. Just a Rasa Core model will be trained.")
trained_model = await _train_core_with_validated_data(
file_importer,
output=output_path,
fixed_model_name=fixed_model_name,
additional_arguments=core_additional_arguments,
model_to_finetune=model_to_finetune,
finetuning_epoch_fraction=finetuning_epoch_fraction,
)
return TrainingResult(model=trained_model)
new_fingerprint = await model.model_fingerprint(file_importer)
old_model = model.get_latest_model(output_path)
if not force_training:
fingerprint_comparison = model.should_retrain(
new_fingerprint,
old_model,
train_path,
has_e2e_examples=nlu_data.has_e2e_examples(),
)
else:
fingerprint_comparison = FingerprintComparisonResult(
force_training=True)
if fingerprint_comparison.is_training_required():
async with telemetry.track_model_training(
file_importer,
model_type="rasa",
):
await _do_training(
file_importer,
output_path=output_path,
train_path=train_path,
fingerprint_comparison_result=fingerprint_comparison,
fixed_model_name=fixed_model_name,
persist_nlu_training_data=persist_nlu_training_data,
core_additional_arguments=core_additional_arguments,
nlu_additional_arguments=nlu_additional_arguments,
old_model_zip_path=old_model,
model_to_finetune=model_to_finetune,
finetuning_epoch_fraction=finetuning_epoch_fraction,
)
trained_model = model.package_model(
fingerprint=new_fingerprint,
output_directory=output_path,
train_path=train_path,
fixed_model_name=fixed_model_name,
)
return TrainingResult(model=trained_model)
rasa.shared.utils.cli.print_success(
"Nothing changed. You can use the old model stored at '{}'."
"".format(os.path.abspath(old_model)))
return TrainingResult(model=old_model)
async def _train_async_internal(
file_importer: TrainingDataImporter,
train_path: Text,
output_path: Text,
force_training: bool,
fixed_model_name: Optional[Text],
persist_nlu_training_data: bool,
kwargs: Optional[Dict],
) -> Optional[Text]:
"""Trains a Rasa model (Core and NLU). Use only from `train_async`.
Args:
file_importer: `TrainingDataImporter` which supplies the training data.
train_path: Directory in which to train the model.
output_path: Output path.
force_training: If `True` retrain model even if data has not changed.
persist_nlu_training_data: `True` if the NLU training data should be persisted
with the model.
fixed_model_name: Name of model to be stored.
kwargs: Additional training parameters.
Returns:
Path of the trained model archive.
"""
stories = await file_importer.get_stories()
nlu_data = await file_importer.get_nlu_data()
if stories.is_empty() and nlu_data.is_empty():
print_error(
"No training data given. Please provide stories and NLU data in "
"order to train a Rasa model using the '--data' argument.")
return
if stories.is_empty():
print_warning(
"No stories present. Just a Rasa NLU model will be trained.")
return await _train_nlu_with_validated_data(
file_importer,
output=output_path,
fixed_model_name=fixed_model_name,
persist_nlu_training_data=persist_nlu_training_data,
)
if nlu_data.is_empty():
print_warning(
"No NLU data present. Just a Rasa Core model will be trained.")
return await _train_core_with_validated_data(
file_importer,
output=output_path,
fixed_model_name=fixed_model_name,
kwargs=kwargs,
)
new_fingerprint = await model.model_fingerprint(file_importer)
old_model = model.get_latest_model(output_path)
fingerprint_comparison = FingerprintComparisonResult(
force_training=force_training)
if not force_training:
fingerprint_comparison = model.should_retrain(new_fingerprint,
old_model, train_path)
if fingerprint_comparison.is_training_required():
await _do_training(
file_importer,
output_path=output_path,
train_path=train_path,
fingerprint_comparison_result=fingerprint_comparison,
fixed_model_name=fixed_model_name,
persist_nlu_training_data=persist_nlu_training_data,
kwargs=kwargs,
)
return model.package_model(
fingerprint=new_fingerprint,
output_directory=output_path,
train_path=train_path,
fixed_model_name=fixed_model_name,
)
print_success("Nothing changed. You can use the old model stored at '{}'."
"".format(os.path.abspath(old_model)))
return old_model
async def _train_async_internal(
file_importer: TrainingDataImporter,
train_path: Text,
output_path: Text,
force_training: bool,
fixed_model_name: Optional[Text],
kwargs: Optional[Dict],
) -> Optional[Text]:
"""Trains a Rasa model (Core and NLU). Use only from `train_async`.
Args:
domain: Path to the domain file.
config: Path to the config for Core and NLU.
train_path: Directory in which to train the model.
nlu_data_directory: Path to NLU training files.
story_directory: Path to Core training files.
output_path: Output path.
force_training: If `True` retrain model even if data has not changed.
fixed_model_name: Name of model to be stored.
kwargs: Additional training parameters.
Returns:
Path of the trained model archive.
"""
new_fingerprint = await model.model_fingerprint(file_importer)
stories = await file_importer.get_stories()
nlu_data = await file_importer.get_nlu_data()
if stories.is_empty() and nlu_data.is_empty():
print_error(
"No training data given. Please provide stories and NLU data in "
"order to train a Rasa model using the '--data' argument.")
return
if stories.is_empty():
print_warning(
"No stories present. Just a Rasa NLU model will be trained.")
return await _train_nlu_with_validated_data(
file_importer,
output=output_path,
fixed_model_name=fixed_model_name)
if nlu_data.is_empty():
print_warning(
"No NLU data present. Just a Rasa Core model will be trained.")
return await _train_core_with_validated_data(
file_importer,
output=output_path,
fixed_model_name=fixed_model_name,
kwargs=kwargs,
)
old_model = model.get_latest_model(output_path)
retrain_core, retrain_nlu = model.should_retrain(new_fingerprint,
old_model, train_path)
if force_training or retrain_core or retrain_nlu:
await _do_training(
file_importer,
output_path=output_path,
train_path=train_path,
force_training=force_training,
retrain_core=retrain_core,
retrain_nlu=retrain_nlu,
fixed_model_name=fixed_model_name,
kwargs=kwargs,
)
return model.package_model(
fingerprint=new_fingerprint,
output_directory=output_path,
train_path=train_path,
fixed_model_name=fixed_model_name,
)
print_success("Nothing changed. You can use the old model stored at '{}'."
"".format(os.path.abspath(old_model)))
return old_model
async def _train_async_internal(
file_importer: TrainingDataImporter,
train_path: Text,
output_path: Text,
force_training: bool,
fixed_model_name: Optional[Text],
persist_nlu_training_data: bool,
additional_arguments: Optional[Dict],
) -> Optional[Text]:
"""Trains a Rasa model (Core and NLU). Use only from `train_async`.
Args:
file_importer: `TrainingDataImporter` which supplies the training data.
train_path: Directory in which to train the model.
output_path: Output path.
force_training: If `True` retrain model even if data has not changed.
persist_nlu_training_data: `True` if the NLU training data should be persisted
with the model.
fixed_model_name: Name of model to be stored.
additional_arguments: Additional training parameters.
Returns:
Path of the trained model archive.
"""
stories, nlu_data = await asyncio.gather(file_importer.get_stories(),
file_importer.get_nlu_data())
# if stories.is_empty() and nlu_data.is_empty():
# print_error(
# "No training data given. Please provide stories and NLU data in "
# "order to train a Rasa model using the '--data' argument."
# )
# return
# if nlu_data.is_empty():
# print_warning("No NLU data present. Just a Rasa Core model will be trained.")
# return await _train_core_with_validated_data(
# file_importer,
# output=output_path,
# fixed_model_name=fixed_model_name,
# additional_arguments=additional_arguments,
# )
new_fingerprint = await model.model_fingerprint(file_importer)
old_model = model.get_latest_model(output_path)
fingerprint_comparison = FingerprintComparisonResult(
force_training=force_training)
if not force_training:
fingerprint_comparison = model.should_retrain(new_fingerprint,
old_model, train_path)
# bf mod >
if fingerprint_comparison.nlu == True: # replace True with list of all langs
fingerprint_comparison.nlu = list(
new_fingerprint.get("nlu-config", {}).keys())
domain = await file_importer.get_domain()
core_untrainable = domain.is_empty() or stories.is_empty()
nlu_untrainable = [l for l, d in nlu_data.items() if d.is_empty()]
fingerprint_comparison.core = fingerprint_comparison.core and not core_untrainable
fingerprint_comparison.nlu = [
l for l in fingerprint_comparison.nlu if l not in nlu_untrainable
]
if core_untrainable:
print_color(
"Skipping Core training since domain or stories are empty.",
color=bcolors.OKBLUE)
for lang in nlu_untrainable:
print_color(
"No NLU data found for language <{}>, skipping training...".format(
lang),
color=bcolors.OKBLUE)
# </ bf mod
if fingerprint_comparison.is_training_required():
await _do_training(
file_importer,
output_path=output_path,
train_path=train_path,
fingerprint_comparison_result=fingerprint_comparison,
fixed_model_name=fixed_model_name,
persist_nlu_training_data=persist_nlu_training_data,
additional_arguments=additional_arguments,
)
return model.package_model(
fingerprint=new_fingerprint,
output_directory=output_path,
train_path=train_path,
fixed_model_name=fixed_model_name,
)
print_success("Nothing changed. You can use the old model stored at '{}'."
"".format(os.path.abspath(old_model)))
return old_model
async def _train_async_internal(
file_importer: TrainingDataImporter,
train_path: Text,
output_path: Text,
force_training: bool,
fixed_model_name: Optional[Text],
persist_nlu_training_data: bool,
kwargs: Optional[Dict],
) -> Optional[Text]:
"""Trains a Rasa model (Core and NLU). Use only from `train_async`.
Args:
file_importer: `TrainingDataImporter` which supplies the training data.
train_path: Directory in which to train the model.
output_path: Output path.
force_training: If `True` retrain model even if data has not changed.
fixed_model_name: Name of model to be stored.
kwargs: Additional training parameters.
Returns:
Path of the trained model archive.
"""
new_fingerprint = await model.model_fingerprint(file_importer)
stories = await file_importer.get_stories()
nlu_data = await file_importer.get_nlu_data()
# if stories.is_empty() and nlu_data.is_empty():
# print_error(
# "No training data given. Please provide stories and NLU data in "
# "order to train a Rasa model using the '--data' argument."
# )
# return
# if stories.is_empty():
# print_warning("No stories present. Just a Rasa NLU model will be trained.")
# return await _train_nlu_with_validated_data(
# file_importer,
# output=output_path,
# fixed_model_name=fixed_model_name,
# persist_nlu_training_data=persist_nlu_training_data,
# )
# if nlu_data.is_empty():
# print_warning("No NLU data present. Just a Rasa Core model will be trained.")
# return await _train_core_with_validated_data(
# file_importer,
# output=output_path,
# fixed_model_name=fixed_model_name,
# kwargs=kwargs,
# )
old_model = model.get_latest_model(output_path)
retrain_core, retrain_nlu = model.should_retrain(new_fingerprint,
old_model, train_path)
# bf mod
domain = await file_importer.get_domain()
core_untrainable = domain.is_empty() or stories.is_empty()
nlu_untrainable = [l for l, d in nlu_data.items() if d.is_empty()]
retrain_core = retrain_core and not core_untrainable
if retrain_nlu is True:
from rasa.model import FINGERPRINT_NLU_DATA_KEY
possible_retrains = new_fingerprint[FINGERPRINT_NLU_DATA_KEY].keys()
else:
possible_retrains = retrain_nlu
if core_untrainable:
print_color(
"Skipping Core training since domain or stories are empty.",
color=bcolors.OKBLUE)
for lang in nlu_untrainable:
print_color(
"No NLU data found for language <{}>, skipping training...".format(
lang),
color=bcolors.OKBLUE)
retrain_nlu = [l for l in possible_retrains if l not in nlu_untrainable]
# /bf mod
if force_training or retrain_core or retrain_nlu:
await _do_training(
file_importer,
output_path=output_path,
train_path=train_path,
force_training=force_training,
retrain_core=retrain_core,
retrain_nlu=retrain_nlu,
fixed_model_name=fixed_model_name,
persist_nlu_training_data=persist_nlu_training_data,
kwargs=kwargs,
)
return model.package_model(
fingerprint=new_fingerprint,
output_directory=output_path,
train_path=train_path,
fixed_model_name=fixed_model_name,
)
print_success("Nothing changed. You can use the old model stored at '{}'."
"".format(os.path.abspath(old_model)))
return old_model
async def _train_async_internal(
file_importer: TrainingDataImporter,
train_path: Text,
output_path: Text,
dry_run: bool,
force_training: bool,
fixed_model_name: Optional[Text],
persist_nlu_training_data: bool,
core_additional_arguments: Optional[Dict] = None,
nlu_additional_arguments: Optional[Dict] = None,
) -> TrainingResult:
"""Trains a Rasa model (Core and NLU). Use only from `train_async`.
Args:
file_importer: `TrainingDataImporter` which supplies the training data.
train_path: Directory in which to train the model.
output_path: Output path.
dry_run: If `True` then no training will be done, and the information about
whether the training needs to be done will be printed.
force_training: If `True` retrain model even if data has not changed.
fixed_model_name: Name of model to be stored.
persist_nlu_training_data: `True` if the NLU training data should be persisted
with the model.
core_additional_arguments: Additional training parameters for core training.
nlu_additional_arguments: Additional training parameters forwarded to training
method of each NLU component.
Returns:
An instance of `TrainingResult`.
"""
stories, nlu_data = await asyncio.gather(file_importer.get_stories(),
file_importer.get_nlu_data())
new_fingerprint = await model.model_fingerprint(file_importer)
old_model = model.get_latest_model(output_path)
fingerprint_comparison = model.should_retrain(new_fingerprint, old_model,
train_path, force_training)
if dry_run:
code, texts = dry_run_result(fingerprint_comparison)
for text in texts:
print_warning(text) if code > 0 else print_success(text)
return TrainingResult(code=code)
if stories.is_empty() and nlu_data.can_train_nlu_model():
print_error(
"No training data given. Please provide stories and NLU data in "
"order to train a Rasa model using the '--data' argument.")
return TrainingResult()
if stories.is_empty():
print_warning(
"No stories present. Just a Rasa NLU model will be trained.")
trained_model = await _train_nlu_with_validated_data(
file_importer,
output=output_path,
fixed_model_name=fixed_model_name,
persist_nlu_training_data=persist_nlu_training_data,
additional_arguments=nlu_additional_arguments,
)
return TrainingResult(model=trained_model)
if nlu_data.can_train_nlu_model():
print_warning(
"No NLU data present. Just a Rasa Core model will be trained.")
trained_model = await _train_core_with_validated_data(
file_importer,
output=output_path,
fixed_model_name=fixed_model_name,
additional_arguments=core_additional_arguments,
)
return TrainingResult(model=trained_model)
if fingerprint_comparison.is_training_required():
async with telemetry.track_model_training(file_importer,
model_type="rasa"):
await _do_training(
file_importer,
output_path=output_path,
train_path=train_path,
fingerprint_comparison_result=fingerprint_comparison,
fixed_model_name=fixed_model_name,
persist_nlu_training_data=persist_nlu_training_data,
core_additional_arguments=core_additional_arguments,
nlu_additional_arguments=nlu_additional_arguments,
old_model_zip_path=old_model,
)
trained_model = model.package_model(
fingerprint=new_fingerprint,
output_directory=output_path,
train_path=train_path,
fixed_model_name=fixed_model_name,
)
return TrainingResult(model=trained_model)
print_success("Nothing changed. You can use the old model stored at '{}'."
"".format(os.path.abspath(old_model)))
return TrainingResult(model=old_model)
async def train_comparison_models(
story_file: Text,
domain: Text,
output_path: Text = "",
exclusion_percentages: Optional[List] = None,
policy_configs: Optional[List] = None,
runs: int = 1,
dump_stories: bool = False,
kwargs: Optional[Dict] = None,
):
"""Train multiple models for comparison of policies"""
from rasa.core import config
from rasa import model
from rasa.importers.importer import TrainingDataImporter
exclusion_percentages = exclusion_percentages or []
policy_configs = policy_configs or []
for r in range(runs):
logging.info("Starting run {}/{}".format(r + 1, runs))
for current_run, percentage in enumerate(exclusion_percentages, 1):
for policy_config in policy_configs:
policies = config.load(policy_config)
if len(policies) > 1:
raise ValueError(
"You can only specify one policy per model for comparison"
)
file_importer = TrainingDataImporter.load_core_importer_from_config(
policy_config, domain, [story_file])
policy_name = type(policies[0]).__name__
logging.info("Starting to train {} round {}/{}"
" with {}% exclusion"
"".format(policy_name, current_run,
len(exclusion_percentages), percentage))
with TempDirectoryPath(tempfile.mkdtemp()) as train_path:
await train(
domain,
file_importer,
train_path,
policy_config=policy_config,
exclusion_percentage=current_run,
kwargs=kwargs,
dump_stories=dump_stories,
)
new_fingerprint = await model.model_fingerprint(
file_importer)
output_dir = os.path.join(output_path, "run_" + str(r + 1))
model_name = policy_name + str(current_run)
model.package_model(
fingerprint=new_fingerprint,
output_directory=output_dir,
train_path=train_path,
fixed_model_name=model_name,
)
async def _train_core_with_validated_data(
file_importer: TrainingDataImporter,
output: Text,
train_path: Optional[Text] = None,
fixed_model_name: Optional[Text] = None,
additional_arguments: Optional[Dict] = None,
interpreter: Optional[Interpreter] = None,
model_to_finetune: Optional["Text"] = None,
finetuning_epoch_fraction: float = 1.0,
) -> Optional[Text]:
"""Train Core with validated training and config data."""
import rasa.core.train
with ExitStack() as stack:
if train_path:
# If the train path was provided, do nothing on exit.
_train_path = train_path
else:
# Otherwise, create a temp train path and clean it up on exit.
_train_path = stack.enter_context(
TempDirectoryPath(tempfile.mkdtemp()))
# normal (not compare) training
rasa.shared.utils.cli.print_color(
"Training Core model...",
color=rasa.shared.utils.io.bcolors.OKBLUE)
domain, config = await asyncio.gather(file_importer.get_domain(),
file_importer.get_config())
if model_to_finetune:
rasa.shared.utils.common.mark_as_experimental_feature(
"Incremental Training feature")
model_to_finetune = await _core_model_for_finetuning(
model_to_finetune,
file_importer=file_importer,
finetuning_epoch_fraction=finetuning_epoch_fraction,
)
if not model_to_finetune:
rasa.shared.utils.cli.print_error_and_exit(
f"No Core model for finetuning found. Please make sure to either "
f"specify a path to a previous model or to have a finetunable "
f"model within the directory '{output}'.")
async with telemetry.track_model_training(
file_importer,
model_type="core",
is_finetuning=model_to_finetune is not None,
):
await rasa.core.train.train(
domain_file=domain,
training_resource=file_importer,
output_path=os.path.join(_train_path,
DEFAULT_CORE_SUBDIRECTORY_NAME),
policy_config=config,
additional_arguments=additional_arguments,
interpreter=interpreter,
model_to_finetune=model_to_finetune,
)
rasa.shared.utils.cli.print_color(
"Core model training completed.",
color=rasa.shared.utils.io.bcolors.OKBLUE)
if train_path is None:
# Only Core was trained.
new_fingerprint = await model.model_fingerprint(file_importer)
return model.package_model(
fingerprint=new_fingerprint,
output_directory=output,
train_path=_train_path,
fixed_model_name=fixed_model_name,
model_prefix="core-",
)
return _train_path
async def _train_nlu_with_validated_data(
file_importer: TrainingDataImporter,
output: Text,
train_path: Optional[Text] = None,
fixed_model_name: Optional[Text] = None,
persist_nlu_training_data: bool = False,
additional_arguments: Optional[Dict] = None,
model_to_finetune: Optional["Text"] = None,
finetuning_epoch_fraction: float = 1.0,
) -> Optional[Text]:
"""Train NLU with validated training and config data."""
import rasa.nlu.train
if additional_arguments is None:
additional_arguments = {}
with ExitStack() as stack:
if train_path:
# If the train path was provided, do nothing on exit.
_train_path = train_path
else:
# Otherwise, create a temp train path and clean it up on exit.
_train_path = stack.enter_context(
TempDirectoryPath(tempfile.mkdtemp()))
config = await file_importer.get_config()
rasa.shared.utils.cli.print_color(
"Training NLU model...", color=rasa.shared.utils.io.bcolors.OKBLUE)
if model_to_finetune:
rasa.shared.utils.common.mark_as_experimental_feature(
"Incremental Training feature")
model_to_finetune = await _nlu_model_for_finetuning(
model_to_finetune,
file_importer,
finetuning_epoch_fraction,
called_from_combined_training=train_path is not None,
)
if not model_to_finetune:
rasa.shared.utils.cli.print_error_and_exit(
f"No NLU model for finetuning found. Please make sure to either "
f"specify a path to a previous model or to have a finetunable "
f"model within the directory '{output}'.")
async with telemetry.track_model_training(
file_importer,
model_type="nlu",
is_finetuning=model_to_finetune is not None,
):
await rasa.nlu.train.train(
config,
file_importer,
_train_path,
fixed_model_name="nlu",
persist_nlu_training_data=persist_nlu_training_data,
model_to_finetune=model_to_finetune,
**additional_arguments,
)
rasa.shared.utils.cli.print_color(
"NLU model training completed.",
color=rasa.shared.utils.io.bcolors.OKBLUE)
if train_path is None:
# Only NLU was trained
new_fingerprint = await model.model_fingerprint(file_importer)
return model.package_model(
fingerprint=new_fingerprint,
output_directory=output,
train_path=_train_path,
fixed_model_name=fixed_model_name,
model_prefix="nlu-",
)
return _train_path
async def _train_nlu_with_validated_data(
file_importer: TrainingDataImporter,
output: Text,
train_path: Optional[Text] = None,
fixed_model_name: Optional[Text] = None,
persist_nlu_training_data: bool = False,
retrain_nlu: Union[bool, List[Text]] = True,
additional_arguments: Optional[Dict] = None,
) -> Optional[Text]:
"""Train NLU with validated training and config data."""
import rasa.nlu.train
if additional_arguments is None:
additional_arguments = {}
with ExitStack() as stack:
if train_path:
# If the train path was provided, do nothing on exit.
_train_path = train_path
else:
# Otherwise, create a temp train path and clean it up on exit.
_train_path = stack.enter_context(TempDirectoryPath(tempfile.mkdtemp()))
# bf mod
config = await file_importer.get_nlu_config(retrain_nlu)
async with telemetry.track_model_training(file_importer, model_type="nlu"):
for lang in config:
if config[lang]:
print_color(
"Start training <{}> NLU model ...".format(lang),
color=rasa.shared.utils.io.bcolors.OKBLUE,
)
await rasa.nlu.train(
config[lang],
file_importer,
_train_path,
fixed_model_name="nlu-{}".format(lang),
persist_nlu_training_data=persist_nlu_training_data,
**additional_arguments,
)
else:
print_color(
"NLU data for language <{}> didn't change, skipping training...".format(
lang
),
color=rasa.shared.utils.io.bcolors.OKBLUE,
)
# /bf mod
print_color(
"NLU model training completed.", color=rasa.shared.utils.io.bcolors.OKBLUE
)
if train_path is None:
# Only NLU was trained
new_fingerprint = await model.model_fingerprint(file_importer)
return model.package_model(
fingerprint=new_fingerprint,
output_directory=output,
train_path=_train_path,
fixed_model_name=fixed_model_name,
model_prefix="nlu-",
)
return _train_path