Skip to content

Latest commit

 

History

History
356 lines (267 loc) · 6.58 KB

modules.md

File metadata and controls

356 lines (267 loc) · 6.58 KB
Raw

Available Modules

The Benchmark, Reranker, and Searcher module types are most often configured by the end user. For a complete list of modules, run the command capreolus modules or see the API Reference.

.. important:: When using Capreolus' configuration system, modules are selected by specifying their ``module_name``.
   For example, the ``NF`` benchmark can be selected with the ``benchmark.name=nf`` config string or the equivalent config dictionary ``{"benchmark": {"name": "nf"}}``.
   
   The corresponding class can be created as ``benchmark.nf.NF(config=..., provide=...)`` or created by name with ``Benchmark.create("nf", config=..., provide=...)``.

Benchmarks

ANTIQUE

.. autoapiclass:: capreolus.benchmark.antique.ANTIQUE
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

CodeSearchNet

.. autoapiclass:: capreolus.benchmark.codesearchnet.CodeSearchNetCorpus
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

.. autoapiclass:: capreolus.benchmark.codesearchnet.CodeSearchNetChallenge
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

(TREC) COVID

.. autoapiclass:: capreolus.benchmark.covid.COVID
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

Dummy

.. autoapiclass:: capreolus.benchmark.dummy.DummyBenchmark
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

NF Corpus

.. autoapiclass:: capreolus.benchmark.nf.NF
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

(TREC) Robust04

.. autoapiclass:: capreolus.benchmark.robust04.Robust04
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

.. autoapiclass:: capreolus.benchmark.robust04.Robust04Yang19
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

Searchers

.. note:: Some searchers (e.g., BM25) automatically perform a cross-validated grid search when their parameters are provided as lists. For example, ``searcher.b=0.4,0.6,0.8 searcher.k1=1.0,1.5``.

BM25

.. autoapiclass:: capreolus.searcher.anserini.BM25
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

BM25 with Axiomatic expansion

.. autoapiclass:: capreolus.searcher.anserini.AxiomaticSemanticMatching
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

BM25 with RM3 expansion

.. autoapiclass:: capreolus.searcher.anserini.BM25RM3
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

BM25 PRF

.. autoapiclass:: capreolus.searcher.anserini.BM25PRF
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

F2Exp

.. autoapiclass:: capreolus.searcher.anserini.F2Exp
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

F2Log

.. autoapiclass:: capreolus.searcher.anserini.F2Log
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

I(n)L2

.. autoapiclass:: capreolus.searcher.anserini.INL2
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

QL with Dirichlet smoothing

.. autoapiclass:: capreolus.searcher.anserini.DirichletQL
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

QL with J-M smoothing

.. autoapiclass:: capreolus.searcher.anserini.QLJM
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

SDM

.. autoapiclass:: capreolus.searcher.anserini.SDM
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

SPL

.. autoapiclass:: capreolus.searcher.anserini.SPL
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

Rerankers

.. note:: Rerankers are implemented in PyTorch or TensorFlow. Rerankers with TensorFlow implementations can run on both GPUs and TPUs.

CEDR-KNRM

.. autoapiclass:: capreolus.reranker.CEDRKNRM.CEDRKNRM
   :noindex:

   .. autoapiattribute:: module_name
      :noindex:

CDSSM

.. autoapiclass:: capreolus.reranker.CDSSM.CDSSM
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

ConvKNRM

.. autoapiclass:: capreolus.reranker.ConvKNRM.ConvKNRM
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

DRMM

.. autoapiclass:: capreolus.reranker.DRMM.DRMM
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

DRMMTKS

.. autoapiclass:: capreolus.reranker.DRMMTKS.DRMMTKS
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

DSSM

.. autoapiclass:: capreolus.reranker.DSSM.DSSM
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

DUET

.. autoapiclass:: capreolus.reranker.DUET.DUET
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

DeepTileBars

.. autoapiclass:: capreolus.reranker.DeepTileBar.DeepTileBar
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

HiNT

.. autoapiclass:: capreolus.reranker.HINT.HINT
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

KNRM

.. autoapiclass:: capreolus.reranker.KNRM.KNRM
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

PACRR

.. autoapiclass:: capreolus.reranker.PACRR.PACRR
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

POSITDRMM

.. autoapiclass:: capreolus.reranker.POSITDRMM.POSITDRMM
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

TK

.. autoapiclass:: capreolus.reranker.TK.TK
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

TensorFlow KNRM

.. autoapiclass:: capreolus.reranker.TFKNRM.TFKNRM
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

TensorFlow BERT-MaxP

.. autoapiclass:: capreolus.reranker.TFBERTMaxP.TFBERTMaxP
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex:

TensorFlow CEDR-KNRM

.. autoapiclass:: capreolus.reranker.TFCEDRKNRM.TFCEDRKNRM
   :noindex:

   .. autoapiattribute:: module_name
      :noindex:

TensorFlow PARADE

.. autoapiclass:: capreolus.reranker.parade.TFParade
   :noindex:
   
   .. autoapiattribute:: module_name
      :noindex: