Module topicnet.cooking_machine.recipes.wntm
Expand source code
from typing import List, Tuple
from .recipe_wrapper import BaseRecipe
from ..config_parser import parse
from .. import Dataset
from .. import DatasetCooc
from .. import Experiment
WNTM_template = '''
# This config follows a pipline for training a Word Net Topic Model
# https://link.springer.com/article/10.1007/s10115-015-0882-z
# Use .format(modality_list=modality_list, main_modality=main_modality, dataset_path=dataset_path,
# specific_topics=specific_topics, background_topics=background_topics)
# when loading the recipe to adjust for your dataset
topics:
# Describes number of model topics, better left to the user to define optimal topic number
specific_topics: {specific_topics}
background_topics: {background_topics}
# Here is example of model with one modality
regularizers:
- DecorrelatorPhiRegularizer:
name: decorrelation_phi
topic_names: specific_topics
class_ids: {modality_list}
tau: 0.2
scores:
- BleiLaffertyScore:
num_top_tokens: 30
model:
dataset_path: {dataset_path}
modalities_to_use: {modality_list}
main_modality: '{main_modality}'
stages:
- RegularizersModifierCube:
num_iter: 20
reg_search: add
regularizer_parameters:
name: decorrelation_phi
selection:
- PerplexityScore@all < 1.05 * MINIMUM(PerplexityScore@all) and BleiLaffertyScore -> max
strategy: PerplexityStrategy
# parameters of this strategy are intended for revision
strategy_params:
start_point: 0
step: 0.001
max_len: 50
tracked_score_function: PerplexityScore@all
verbose: false
use_relative_coefficients: true
'''
class WNTMRecipe(BaseRecipe):
"""
Class for baseline recipe creation and
unification of recipe interface
"""
def __init__(self):
super().__init__(recipe_template=WNTM_template)
def format_recipe(
self,
dataset_path: str,
modality_list: List[str] = None,
main_modality: str = None,
topic_number: int = 20,
background_topic_number: int = 0,
num_iter: int = 20,
):
self.dataset_path = dataset_path
if modality_list is None:
modality_list = list(Dataset(dataset_path).get_possible_modalities())
if main_modality is None:
main_modality = modality_list[0]
specific_topics = [f'topic_{i}' for i in range(topic_number)]
background_topics = [f'bcg_{i}' for i in range(
len(specific_topics), len(specific_topics) + background_topic_number)]
self._recipe = self.recipe_template.format(
dataset_path=dataset_path,
modality_list=modality_list,
main_modality=main_modality,
specific_topics=specific_topics,
background_topics=background_topics,
)
return self._recipe
def build_experiment_environment(
self,
save_path: str,
experiment_id: str = 'default_experiment_name',
force_separate_thread: bool = False
) -> Tuple[Experiment, Dataset]:
"""
Returns experiment and dataset instances
needed to perform the hyperparameter tuning on the data
according to recipe
Parameters
----------
save_path: path to the folder to save experiment logs and models
experiment_id: name of the experiment folder
force_separate_thread: train each model in dedicated process
this feature helps to handle resources in Jupyter notebooks
"""
if self._recipe is None:
raise ValueError(
'Recipe missing data specific parameters. '
'Provide them with "format_recipe" method!')
settings, regs, model, dataset = parse(
self._recipe,
force_separate_thread=force_separate_thread,
dataset_class=DatasetCooc
)
# TODO: handle dynamic addition of regularizers
experiment = Experiment(experiment_id=experiment_id, save_path=save_path, topic_model=model)
experiment.build(settings)
return experiment, datasetClasses
class WNTMRecipe-
Class for baseline recipe creation and unification of recipe interface
Expand source code
class WNTMRecipe(BaseRecipe): """ Class for baseline recipe creation and unification of recipe interface """ def __init__(self): super().__init__(recipe_template=WNTM_template) def format_recipe( self, dataset_path: str, modality_list: List[str] = None, main_modality: str = None, topic_number: int = 20, background_topic_number: int = 0, num_iter: int = 20, ): self.dataset_path = dataset_path if modality_list is None: modality_list = list(Dataset(dataset_path).get_possible_modalities()) if main_modality is None: main_modality = modality_list[0] specific_topics = [f'topic_{i}' for i in range(topic_number)] background_topics = [f'bcg_{i}' for i in range( len(specific_topics), len(specific_topics) + background_topic_number)] self._recipe = self.recipe_template.format( dataset_path=dataset_path, modality_list=modality_list, main_modality=main_modality, specific_topics=specific_topics, background_topics=background_topics, ) return self._recipe def build_experiment_environment( self, save_path: str, experiment_id: str = 'default_experiment_name', force_separate_thread: bool = False ) -> Tuple[Experiment, Dataset]: """ Returns experiment and dataset instances needed to perform the hyperparameter tuning on the data according to recipe Parameters ---------- save_path: path to the folder to save experiment logs and models experiment_id: name of the experiment folder force_separate_thread: train each model in dedicated process this feature helps to handle resources in Jupyter notebooks """ if self._recipe is None: raise ValueError( 'Recipe missing data specific parameters. ' 'Provide them with "format_recipe" method!') settings, regs, model, dataset = parse( self._recipe, force_separate_thread=force_separate_thread, dataset_class=DatasetCooc ) # TODO: handle dynamic addition of regularizers experiment = Experiment(experiment_id=experiment_id, save_path=save_path, topic_model=model) experiment.build(settings) return experiment, datasetAncestors
Methods
def build_experiment_environment(self, save_path: str, experiment_id: str = 'default_experiment_name', force_separate_thread: bool = False) ‑> Tuple[Experiment, Dataset]-
Returns experiment and dataset instances needed to perform the hyperparameter tuning on the data according to recipe
Parameters
save_path:path to the folder to save experiment logs and modelsexperiment_id:nameofthe experiment folderforce_separate_thread:train each model in dedicated process- this feature helps to handle resources in Jupyter notebooks
Expand source code
def build_experiment_environment( self, save_path: str, experiment_id: str = 'default_experiment_name', force_separate_thread: bool = False ) -> Tuple[Experiment, Dataset]: """ Returns experiment and dataset instances needed to perform the hyperparameter tuning on the data according to recipe Parameters ---------- save_path: path to the folder to save experiment logs and models experiment_id: name of the experiment folder force_separate_thread: train each model in dedicated process this feature helps to handle resources in Jupyter notebooks """ if self._recipe is None: raise ValueError( 'Recipe missing data specific parameters. ' 'Provide them with "format_recipe" method!') settings, regs, model, dataset = parse( self._recipe, force_separate_thread=force_separate_thread, dataset_class=DatasetCooc ) # TODO: handle dynamic addition of regularizers experiment = Experiment(experiment_id=experiment_id, save_path=save_path, topic_model=model) experiment.build(settings) return experiment, dataset
Inherited members