#!/usr/bin/env python
"""
Ensures models are automatically found by allennlp.
"""
import logging
from allennlp.common.plugins import import_plugins
from allennlp.models import Model
logging.basicConfig(level=logging.INFO)
import_plugins()
Model.by_name("copynet_seq2seq")
def __init__(
self,
id: str,
registered_model_name: Optional[str] = None,
model_class: Optional[type] = None,
display_name: Optional[str] = None,
archive_file: Optional[str] = None,
overrides: Optional[Dict] = None,
model_details: Optional[Union[str, ModelDetails]] = None,
intended_use: Optional[Union[str, IntendedUse]] = None,
factors: Optional[Union[str, Factors]] = None,
metrics: Optional[Union[str, Metrics]] = None,
evaluation_data: Optional[Union[str, EvaluationData]] = None,
training_data: Optional[Union[str, TrainingData]] = None,
quantitative_analyses: Optional[Union[str, QuantitativeAnalyses]] = None,
ethical_considerations: Optional[Union[str, EthicalConsiderations]] = None,
caveats_and_recommendations: Optional[Union[str, CaveatsAndRecommendations]] = None,
):
assert id
if not model_class and registered_model_name:
try:
model_class = Model.by_name(registered_model_name)
except ConfigurationError:
logger.warning("{} is not a registered model.".format(registered_model_name))
if model_class:
display_name = display_name or model_class.__name__
model_details = model_details or get_description(model_class)
if archive_file and not archive_file.startswith("https:"):
archive_file = os.path.join(self._storage_location, archive_file)
if isinstance(model_details, str):
model_details = ModelDetails(description=model_details)
if isinstance(intended_use, str):
intended_use = IntendedUse(primary_uses=intended_use)
if isinstance(factors, str):
factors = Factors(relevant_factors=factors)
if isinstance(metrics, str):
metrics = Metrics(model_performance_measures=metrics)
if isinstance(evaluation_data, str):
evaluation_data = EvaluationData(dataset=evaluation_data)
if isinstance(training_data, str):
training_data = TrainingData(dataset=training_data)
if isinstance(quantitative_analyses, str):
quantitative_analyses = QuantitativeAnalyses(unitary_results=quantitative_analyses)
if isinstance(ethical_considerations, str):
ethical_considerations = EthicalConsiderations(ethical_considerations)
if isinstance(caveats_and_recommendations, str):
caveats_and_recommendations = CaveatsAndRecommendations(caveats_and_recommendations)
self.id = id
self.registered_model_name = registered_model_name
self.display_name = display_name
self.archive_file = archive_file
self.model_details = model_details
self.intended_use = intended_use
self.factors = factors
self.metrics = metrics
self.evaluation_data = evaluation_data
self.training_data = training_data
self.quantitative_analyses = quantitative_analyses
self.ethical_considerations = ethical_considerations
self.caveats_and_recommendations = caveats_and_recommendations
def __init__(
self,
id: str,
registered_model_name: Optional[str] = None,
model_class: Optional[Callable[..., Model]] = None,
registered_predictor_name: Optional[str] = None,
display_name: Optional[str] = None,
task_id: Optional[str] = None,
model_usage: Optional[Union[str, ModelUsage]] = None,
model_details: Optional[Union[str, ModelDetails]] = None,
intended_use: Optional[Union[str, IntendedUse]] = None,
factors: Optional[Union[str, Factors]] = None,
metrics: Optional[Union[str, Metrics]] = None,
evaluation_data: Optional[Union[str, EvaluationData]] = None,
training_data: Optional[Union[str, TrainingData]] = None,
quantitative_analyses: Optional[Union[str,
QuantitativeAnalyses]] = None,
model_ethical_considerations: Optional[Union[
str, ModelEthicalConsiderations]] = None,
model_caveats_and_recommendations: Optional[Union[
str, ModelCaveatsAndRecommendations]] = None,
):
assert id
if not model_class and registered_model_name:
try:
model_class = Model.by_name(registered_model_name)
except ConfigurationError:
logger.warning("{} is not a registered model.".format(
registered_model_name))
if model_class:
display_name = display_name or model_class.__name__
model_details = model_details or get_description(model_class)
if not registered_predictor_name:
registered_predictor_name = model_class.default_predictor # type: ignore
if isinstance(model_usage, str):
model_usage = ModelUsage(archive_file=model_usage)
if isinstance(model_details, str):
model_details = ModelDetails(description=model_details)
if isinstance(intended_use, str):
intended_use = IntendedUse(primary_uses=intended_use)
if isinstance(factors, str):
factors = Factors(relevant_factors=factors)
if isinstance(metrics, str):
metrics = Metrics(model_performance_measures=metrics)
if isinstance(evaluation_data, str):
evaluation_data = EvaluationData(dataset=evaluation_data)
if isinstance(training_data, str):
training_data = TrainingData(dataset=training_data)
if isinstance(quantitative_analyses, str):
quantitative_analyses = QuantitativeAnalyses(
unitary_results=quantitative_analyses)
if isinstance(model_ethical_considerations, str):
model_ethical_considerations = ModelEthicalConsiderations(
model_ethical_considerations)
if isinstance(model_caveats_and_recommendations, str):
model_caveats_and_recommendations = ModelCaveatsAndRecommendations(
model_caveats_and_recommendations)
self.id = id
self.registered_model_name = registered_model_name
self.registered_predictor_name = registered_predictor_name
self.display_name = display_name
self.task_id = task_id
self.model_usage = model_usage
self.model_details = model_details
self.intended_use = intended_use
self.factors = factors
self.metrics = metrics
self.evaluation_data = evaluation_data
self.training_data = training_data
self.quantitative_analyses = quantitative_analyses
self.model_ethical_considerations = model_ethical_considerations
self.model_caveats_and_recommendations = model_caveats_and_recommendations
