def modelf_reader(resources_or_conf: Union[dict, SharedResources] = None):
"""Creates a knowledge_base_population model F."""
from jack.readers.knowledge_base_population.model_f import ModelFInputModule, ModelFModelModule, ModelFOutputModule
shared_resources = create_shared_resources(resources_or_conf)
input_module = ModelFInputModule(shared_resources)
model_module = ModelFModelModule(shared_resources)
output_module = ModelFOutputModule()
return TFReader(shared_resources, input_module, model_module,
output_module)
def bidaf_reader(resources_or_conf: Union[dict, SharedResources] = None):
"""Creates a FastQA model as described in https://arxiv.org/abs/1703.04816 (extractive qa model)."""
from jack.readers.extractive_qa.shared import XQAInputModule, XQAOutputModule
from jack.readers.extractive_qa.bidaf import BiDAF
shared_resources = create_shared_resources(resources_or_conf)
input_module = XQAInputModule(shared_resources)
model_module = BiDAF(shared_resources)
output_module = XQAOutputModule(shared_resources)
return TFReader(shared_resources, input_module, model_module,
output_module)
def fastqa_reader(resources_or_conf: Union[dict, SharedResources] = None):
"""Creates a FastQA reader instance (extractive qa model)."""
from jack.readers.extractive_qa.fastqa import FastQAModule
from jack.readers.extractive_qa.shared import XQAInputModule, XQAOutputModule
shared_resources = create_shared_resources(resources_or_conf)
input_module = XQAInputModule(shared_resources)
model_module = FastQAModule(shared_resources)
output_module = XQAOutputModule(shared_resources)
return TFReader(shared_resources, input_module, model_module,
output_module)
def complex_reader(resources_or_conf: Union[dict, SharedResources] = None):
""" Creates a knowledge_base_population Complex model. """
from jack.readers.knowledge_base_population.models import KnowledgeGraphEmbeddingInputModule, KnowledgeGraphEmbeddingModelModule, \
KnowledgeGraphEmbeddingOutputModule
shared_resources = create_shared_resources(resources_or_conf)
input_module = KnowledgeGraphEmbeddingInputModule(shared_resources)
model_module = KnowledgeGraphEmbeddingModelModule(shared_resources,
model_name='ComplEx')
output_module = KnowledgeGraphEmbeddingOutputModule()
return TFReader(shared_resources, input_module, model_module,
output_module)
def cbow_xqa_reader(resources_or_conf: Union[dict, SharedResources] = None):
"""Creates a CBow QA model as described in https://arxiv.org/abs/1703.04816. """
from jack.readers.extractive_qa.cbow import CbowXQAInputModule
from jack.readers.extractive_qa.cbow import CbowXQAModule
from jack.readers.extractive_qa.shared import XQANoScoreOutputModule
shared_resources = create_shared_resources(resources_or_conf)
input_module = CbowXQAInputModule(shared_resources)
model_module = CbowXQAModule(shared_resources)
output_module = XQANoScoreOutputModule(shared_resources)
return TFReader(shared_resources, input_module, model_module,
output_module)
def esim_snli_reader(resources_or_conf: Union[dict, SharedResources] = None):
"""Creates a SNLI reader instance (multiple choice qa model).
This particular reader uses an Enhanced LSTM Model (ESIM), as described in [1].
[1] Qian Chen et al. - Enhanced LSTM for Natural Language Inference. ACL 2017
"""
from jack.readers.multiple_choice.shared import SingleSupportFixedClassInputs
from jack.readers.natural_language_inference.esim import ESIMModel
from jack.readers.multiple_choice.shared import SimpleMCOutputModule
shared_resources = create_shared_resources(resources_or_conf)
input_module = SingleSupportFixedClassInputs(shared_resources)
model_module = ESIMModel(shared_resources)
output_module = SimpleMCOutputModule(shared_resources)
return TFReader(shared_resources, input_module, model_module,
output_module)
def dam_snli_reader(resources_or_conf: Union[dict, SharedResources] = None):
"""Creates a SNLI reader instance (multiple choice qa model).
This particular reader uses a Decomposable Attention Model, as described in [1].
[1] Ankur P. Parikh et al. - A Decomposable Attention Model for Natural Language Inference. EMNLP 2016
"""
from jack.readers.multiple_choice.shared import SingleSupportFixedClassInputs
from jack.readers.natural_language_inference.decomposable_attention import DecomposableAttentionModel
from jack.readers.multiple_choice.shared import SimpleMCOutputModule
shared_resources = create_shared_resources(resources_or_conf)
input_module = SingleSupportFixedClassInputs(shared_resources)
model_module = DecomposableAttentionModel(shared_resources)
output_module = SimpleMCOutputModule(shared_resources)
return TFReader(shared_resources, input_module, model_module,
output_module)
def cbilstm_snli_reader(resources_or_conf: Union[dict,
SharedResources] = None):
"""
Creates a SNLI reader instance (multiple choice qa model).
This particular reader uses a conditional Bidirectional LSTM, as described in [1].
[1] Tim Rocktäschel et al. - Reasoning about Entailment with Neural Attention. ICLR 2016
"""
from jack.readers.multiple_choice.shared import SingleSupportFixedClassInputs
from jack.readers.natural_language_inference.bilstm import PairOfBiLSTMOverSupportAndQuestionModel
from jack.readers.multiple_choice.shared import SimpleMCOutputModule
shared_resources = create_shared_resources(resources_or_conf)
input_module = SingleSupportFixedClassInputs(shared_resources)
model_module = PairOfBiLSTMOverSupportAndQuestionModel(shared_resources)
output_module = SimpleMCOutputModule(shared_resources)
return TFReader(shared_resources, input_module, model_module,
output_module)
