Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/amazon-science/ReFinED
/ types & classes
Types & classes
58 in github.com/amazon-science/ReFinED
⨍
Functions
427
◇
Types & classes
58
↓ 29 callers
Class
Date
src/refined/utilities/numeric_handling/date_utils.py:12
↓ 20 callers
Class
Span
Represents a span (entity-mention) of text in a document, `Doc`.
src/refined/data_types/base_types.py:81
↓ 8 callers
Class
BatchElementToken
This class represents a token used to produce tensors for a forward pass of a nn. The `acc_sum` is used for mean pooling contiguous token for
src/refined/data_types/modelling_types.py:14
↓ 8 callers
Class
Entity
src/refined/data_types/base_types.py:63
↓ 8 callers
Class
LmdbImmutableDict
src/refined/resource_management/lmdb_wrapper.py:17
↓ 7 callers
Class
ResourceManager
src/refined/resource_management/resource_manager.py:16
↓ 7 callers
Class
S3Manager
src/refined/resource_management/aws.py:17
↓ 6 callers
Class
BatchElement
Represents part of a document (`Doc`) that fits in single pass of a transformer model (< max_seq_len) as a singleton batch. If the document i
src/refined/data_types/modelling_types.py:110
↓ 3 callers
Class
BatchedElementsTns
Represents the tensors for a list/batch of `BatchElement` that have been collated (vertically stacking in dim=1). Equivalent to `BatchElement
src/refined/data_types/modelling_types.py:58
↓ 3 callers
Class
ClassHandler
src/refined/doc_preprocessing/class_handler.py:8
↓ 3 callers
Class
Infix
Infix operators. The calling sequence for the infix is: x |op| y
src/refined/offline_data_generation/wiki_extractor.py:1728
↓ 3 callers
Class
Metrics
src/refined/evaluation/metrics.py:8
↓ 3 callers
Class
SimpleNamespace
src/refined/offline_data_generation/wiki_extractor.py:83
↓ 2 callers
Class
BatchElementTns
Batch element as a tuple of tensors. This tuple is for a single `BatchElement`.
src/refined/data_types/modelling_types.py:34
↓ 2 callers
Class
DataParallelReFinED
src/refined/torch_overrides/data_parallel_refined.py:47
↓ 2 callers
Class
Extractor
An extraction task on a article.
src/refined/offline_data_generation/wiki_extractor.py:538
↓ 2 callers
Class
FineTuningArgs
src/refined/training/fine_tune/fine_tune_args.py:13
↓ 2 callers
Class
Frame
src/refined/offline_data_generation/wiki_extractor.py:508
↓ 2 callers
Class
ModelReturn
Result of RefinedModel(nn.Module).
src/refined/data_types/modelling_types.py:94
↓ 2 callers
Class
TemplateText
Fixed text of template
src/refined/offline_data_generation/wiki_extractor.py:447
↓ 2 callers
Class
WikidataMapper
src/refined/doc_preprocessing/wikidata_mapper.py:8
↓ 2 callers
Class
WikipediaDataset
This is a dataset reader for Wikipedia links dataset. It is a torch iterable dataset because it is too large to hold in memory. The datas
src/refined/dataset_reading/entity_linking/wikipedia_dataset.py:21
↓ 1 callers
Class
AdditionalEntity
src/refined/offline_data_generation/dataclasses_for_preprocessing.py:6
↓ 1 callers
Class
CandidateGeneratorExactMatch
src/refined/doc_preprocessing/candidate_generator.py:30
↓ 1 callers
Class
Datasets
src/refined/dataset_reading/entity_linking/dataset_factory.py:11
↓ 1 callers
Class
DateHandler
Handles resolving the string of a date mention to an instance of the Date class
src/refined/utilities/numeric_handling/date_utils.py:97
↓ 1 callers
Class
DescriptionEncoder
src/refined/model_components/description_encoder.py:12
↓ 1 callers
Class
DocDataset
Standard (go-to) dataset.
src/refined/dataset_reading/entity_linking/document_dataset.py:15
↓ 1 callers
Class
DownloadProgressBar
src/refined/offline_data_generation/preprocessing_utils.py:19
↓ 1 callers
Class
EDLayer
src/refined/model_components/ed_layer_2.py:11
↓ 1 callers
Class
EntityDisambiguation
src/refined/model_components/entity_disambiguation_layer.py:9
↓ 1 callers
Class
EntityTyping
src/refined/model_components/entity_typing_layer.py:7
↓ 1 callers
Class
LookupsInferenceOnly
src/refined/resource_management/data_lookups.py:17
↓ 1 callers
Class
MagicWords
One copy in each Extractor. @see https://doc.wikimedia.org/mediawiki-core/master/php/MagicWord_8php_source.html
src/refined/offline_data_generation/wiki_extractor.py:1541
↓ 1 callers
Class
MentionDetection
src/refined/model_components/mention_detection_layer.py:13
↓ 1 callers
Class
ModelConfig
src/refined/model_components/config.py:29
↓ 1 callers
Class
NextFile
Synchronous generation of next available file name.
src/refined/offline_data_generation/wiki_extractor.py:2658
↓ 1 callers
Class
OutputSplitter
File-like object, that splits output to multiple files of a given max size.
src/refined/offline_data_generation/wiki_extractor.py:2690
↓ 1 callers
Class
PreprocessorInferenceOnly
Implementation of Preprocessor that uses memmaped objects and only contains data needed for inference. See `Preprocessor` for additional docu
src/refined/doc_preprocessing/preprocessor.py:114
↓ 1 callers
Class
Refined
This class provides methods to process text. It wraps around document preprocessor and the ReFinED model class.
src/refined/inference/processor.py:30
↓ 1 callers
Class
RefinedModel
Model for MD, ET, ED built on top of a transformer.
src/refined/model_components/refined_model.py:27
↓ 1 callers
Class
SpanDateHandler
For Span instances which have been identified by the MD layer as having type "DATE", resolve the dates from their text format (e.g. "12th Jun
src/refined/utilities/numeric_handling/date_utils.py:288
↓ 1 callers
Class
Template
A Template is a list of TemplateText or TemplateArgs
src/refined/offline_data_generation/wiki_extractor.py:398
↓ 1 callers
Class
TemplateArg
parameter to a template. Has a name and a default value, both of which are Templates.
src/refined/offline_data_generation/wiki_extractor.py:455
↓ 1 callers
Class
Token
This class represents a token in tokenized text.
src/refined/data_types/base_types.py:7
↓ 1 callers
Class
TrainingArgs
src/refined/training/train/training_args.py:17
Class
CandidateGenerator
src/refined/doc_preprocessing/candidate_generator.py:10
Class
CoNLLNER
src/refined/dataset_reading/mention_detection/conll_reader.py:30
Class
Date
src/refined/data_types/base_types.py:19
Class
Doc
Represents a document of text and spans.
src/refined/data_types/doc_types.py:15
Class
DocIterDataset
Use this class instead of `DocDataset` when the dataset is very large and does not fit in main memory.
src/refined/dataset_reading/entity_linking/document_dataset.py:94
Class
MentionDetector
src/refined/inference/standalone_md.py:29
Class
OntoNotesNER
src/refined/dataset_reading/mention_detection/ontonotes_reader.py:15
Class
Package
Package Code and Dependencies into wheelhouse
package.py:13
Class
Preprocessor
Abstract class representing a preprocessor that provides methods to use resources such as Wikidata classes, and P(e|m) alias table. Implement
src/refined/doc_preprocessing/preprocessor.py:22
Class
ReFinEDConfig
src/refined/data_types/modelling_types.py:248
Class
Token
src/refined/inference/standalone_md.py:18
Class
WebQSPNER
src/refined/dataset_reading/mention_detection/webqsp_reader.py:16