Welcome to the PapersWithCode API Documentation

This documentation details how to use the paperswithcode library to connect with PapersWithCode.

Contents

Installation

The library requires Python 3.6+ and can be installed via pip:

$ pip install paperswithcode-client

Usually you would want to create a virtual environment before installing the library:

$ python3 -m venv pwc
$ source pwc/bin/activate
$ pip install paperswithcode-client

Quickstart

Library is designed to work with python objects. All of the models used as return values in the library are described in the API part of the documentation.

To use the library you only need to import and instantiate the client and start calling methods on it:

>>> from paperswithcode import PapersWithCodeClient 
>>> client = PapersWithCodeClient()
>>> papers_page = client.paper_list()
>>> papers_page.count
175834
>>> papers_page.next_page
2
>>> paper = papers_page.results[0]
>>> paper.id
'efficient-methods-for-incorporating-knowledge'
>>> paper.title
'Efficient Methods for Incorporating Knowledge into Topic Models'

Same principle is used for all models.

>>> dataset_page = client.dataset_list()
>>> dataset_page.count
2782
>>> dataset_page.results[0].id
'hci'
>>> dataset_page.results[0].name
'HCI'

For nested queries you will need to provide the required id’s:

>>> conference_page = client.conference_list()
>>> conference = conference_page.results[0]
>>> conference.id
'eccv'
>>> proceedings_page = client.proceeding_list(conference_id=conference.id)
>>> proceeding = proceedings_page.results[0]
>>> proceeding.id
'eccv-2018'
>>> papers = client.proceeding_paper_list(
    conference_id=conference.id, proceeding_id=proceeding.id
)
>>> papers[0].title
'Person Search by Multi-Scale Matching'

Mirroring your competition on Papers with Code

For more information on how to mirror your competition, please refer to the README file in the paperswithcode-client repository.

API Documentation

PapersWithCode Client api documentation

PapersWithCode Client Models

Paper Models

class paperswithcode.models.paper.Paper(*, id: str, arxiv_id: str = None, nips_id: str = None, url_abs: str, url_pdf: str, title: str, abstract: str, authors: List[str], published: datetime.date = None, conference: str = None, conference_url_abs: str = None, conference_url_pdf: str = None, proceeding: str = None)

Paper object.

id

Paper ID.

Type

str

arxiv_id

ArXiv ID.

Type

str, optional

nips_id

NIPS Conference ID.

Type

str, optional

url_abs

URL to the paper abstract.

Type

str

url_pdf

URL to the paper PDF.

Type

str

title

Paper title.

Type

str

abstract

Paper abstract.

Type

str

authors

List of paper authors.

Type

List[str]

published

Paper publication date.

Type

date, optional

conference

ID of the conference in which the paper was published.

Type

str, optional

conference_url_abs

URL to the conference paper page.

Type

str, optional

conference_url_pdf

URL to the conference paper PDF.

Type

str, optional

proceeding

ID of the conference proceeding in which the paper was published.

Type

str, optional

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.paper.Papers(*, count: int, next_page: int = None, previous_page: int = None, results: List[paperswithcode.models.paper.Paper])

Object representing a paginated page of papers.

count

Number of elements matching the query.

Type

int

next_page

Number of the next page.

Type

int, optional

previous_page

Number of the previous page.

Type

int, optional

results

List of papers on this page.

Type

List[Paper]

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

Repository Models

class paperswithcode.models.repository.Repository(*, url: str, owner: str, name: str, description: str, stars: int, framework: str, is_official: bool = None)

Repository object.

url

URL of the repository.

Type

str

owner

Repository owner.

Type

str

name

Repository name.

Type

str

description

Repository description.

Type

str

stars

Number of repository stars.

Type

int

framework

Implementation framework (TensorFlow, PyTorch, MXNet, Torch, Jax, Caffee2…).

Type

str

is_official

Is this an official implementation of the paper. Available only when listing repositories for a specific paper.

Type

bool

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.repository.Repositories(*, count: int, next_page: int = None, previous_page: int = None, results: List[paperswithcode.models.repository.Repository])

Object representing a paginated page of repositories.

count

Number of elements matching the query.

Type

int

next_page

Number of the next page.

Type

int, optional

previous_page

Number of the previous page.

Type

int, optional

results

List of repositories on this page.

Type

List[Repository]

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

Paper Repository Models

class paperswithcode.models.paper_repo.PaperRepo(*, paper: paperswithcode.models.paper.Paper, repository: paperswithcode.models.repository.Repository = None, is_official: bool)

Paper <-> Repository object.

paper

Paper objects.

Type

Paper

repository

Repository object.

Type

Repository, optional

is_official

Is this the official implementation.

Type

bool

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.paper_repo.PaperRepos(*, count: int, next_page: int = None, previous_page: int = None, results: List[paperswithcode.models.paper_repo.PaperRepo])

Object representing a paginated page of paper<->repos.

count

Number of elements matching the query.

Type

int

next_page

Number of the next page.

Type

int, optional

previous_page

Number of the previous page.

Type

int, optional

results

List of paper<->repos on this page.

Type

List[PaperRepo]

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

Author Models

class paperswithcode.models.author.Author(*, id: str, full_name: str)

Author object.

id

Author ID.

Type

str

full_name

Author full name.

Type

str, optional

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.author.Authors(*, count: int, next_page: int = None, previous_page: int = None, results: List[paperswithcode.models.author.Author])

Object representing a paginated page of authors.

count

Number of elements matching the query.

Type

int

next_page

Number of the next page.

Type

int, optional

previous_page

Number of the previous page.

Type

int, optional

results

List of authors on this page.

Type

List[Author]

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

Conference Models

class paperswithcode.models.conference.Conference(*, id: str, name: str)

Conference object.

id

Conference ID.

Type

str

name

Conerence name.

Type

str

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.conference.Conferences(*, count: int, next_page: int = None, previous_page: int = None, results: List[paperswithcode.models.conference.Conference])

Object representing a paginated page of conferences.

count

Number of elements matching the query.

Type

int

next_page

Number of the next page.

Type

int, optional

previous_page

Number of the previous page.

Type

int, optional

results

List of conferences on this page.

Type

List[Conference]

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.conference.Proceeding(*, id: str, year: int = None, month: int = None)

Conference proceeding object.

id

Proceeding ID.

Type

str

year

Year in which the proceeding was held.

Type

int, optinoal

month

Month in which the proceedingt was held.

Type

int, optional

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.conference.Proceedings(*, count: int, next_page: int = None, previous_page: int = None, results: List[paperswithcode.models.conference.Proceeding])

Object representing a paginated page of proceedings.

count

Number of elements matching the query.

Type

int

next_page

Number of the next page.

Type

int, optional

previous_page

Number of the previous page.

Type

int, optional

results

List of proceedings on this page.

Type

List[Proceeding]

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

Task Models

class paperswithcode.models.task.Area(*, id: str, name: str)

Area object.

Representing an area of research.

id

Area ID.

Type

str

name

Area name.

Type

str

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.task.Areas(*, count: int, next_page: int = None, previous_page: int = None, results: List[paperswithcode.models.task.Area])

Object representing a paginated page of areas.

count

Number of elements matching the query.

Type

int

next_page

Number of the next page.

Type

int, optional

previous_page

Number of the previous page.

Type

int, optional

results

List of areas on this page.

Type

List[Area]

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.task.Task(*, id: str, name: str, description: str)

Task object.

id

Task ID.

Type

str

name

Task name.

Type

str

description

Task description.

Type

str

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.task.TaskCreateRequest(*, name: str, description: str = '', area: str = None, parent_task: str = None)

Task object.

name

Task name.

Type

str

description

Task description.

Type

str

area

Task area ID or area name.

Type

str, optional

parent_task

ID of the parent task.

Type

str, optional

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.task.TaskUpdateRequest(*, name: str = None, description: str = None, area: str = None, parent_task: str = None)

Evaluation table row object.

name

Task name.

Type

str, optional

description

Task description.

Type

str, optional

area

Task area ID.

Type

str, optional

parent_task

ID of the parent task.

Type

str, optional

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.task.Tasks(*, count: int, next_page: int = None, previous_page: int = None, results: List[paperswithcode.models.task.Task])

Object representing a paginated page of tasks.

count

Number of elements matching the query.

Type

int

next_page

Number of the next page.

Type

int, optional

previous_page

Number of the previous page.

Type

int, optional

results

List of tasks on this page.

Type

List[Task]

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

Dataset Models

class paperswithcode.models.dataset.Dataset(*, id: str, name: str, full_name: str = None, url: str = None)

Dataset object.

id

Dataset ID.

Type

str

name

Dataset name.

Type

str

full_name

Dataset full name.

Type

str, optional

url

URL for dataset download.

Type

str, optional

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.dataset.DatasetCreateRequest(*, name: str, full_name: str = None, url: str = None)

Task object.

name

Dataset name.

Type

str

full_name

Dataset full name.

Type

str, optional

url

Dataset url.

Type

str, optional

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.dataset.DatasetUpdateRequest(*, name: str = None, url: str = None)

Evaluation table row object.

name

Dataset name.

Type

str, optional

url

Dataset url.

Type

str, optional

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.dataset.Datasets(*, count: int, next_page: int = None, previous_page: int = None, results: List[paperswithcode.models.dataset.Dataset])

Object representing a paginated page of datasets.

count

Number of elements matching the query.

Type

int

next_page

Number of the next page.

Type

int, optional

previous_page

Number of the previous page.

Type

int, optional

results

List of datasets on this page.

Type

List[Dataset]

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

Method Models

class paperswithcode.models.method.Method(*, id: str, name: str, full_name: str, description: str, paper: str = None)

Method object.

id

Method ID.

Type

str

name

Method short name.

Type

str

full_name

Method full name.

Type

str

description

Method description.

Type

str

paper

ID of the paper that describes the method.

Type

str, optional

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.method.Methods(*, count: int, next_page: int = None, previous_page: int = None, results: List[paperswithcode.models.method.Method])

Object representing a paginated page of methods.

count

Number of elements matching the query.

Type

int

next_page

Number of the next page.

Type

int, optional

previous_page

Number of the previous page.

Type

int, optional

results

List of methods on this page.

Type

List[Method]

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

Evaluation Table Models

class paperswithcode.models.evaluation.Metric(*, id: str, name: str, description: str, is_loss: bool)

Metric object.

Metric used for evaluation.

id

Metric id.

Type

str

name

Metric name.

Type

str

description

Metric description.

Type

str

is_loss

Is this a loss metric.

Type

bool

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.Metrics(*, count: int, next_page: int = None, previous_page: int = None, results: List[paperswithcode.models.evaluation.metric.Metric])

Object representing a paginated page of metrics.

count

Number of elements matching the query.

Type

int

next_page

Number of the next page.

Type

int, optional

previous_page

Number of the previous page.

Type

int, optional

results

List of metrics on this page.

Type

List[Metric]

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.MetricCreateRequest(*, name: str, description: str, is_loss: bool)

Metric object.

Metric used for evaluation.

name

Metric name.

Type

str

description

Metric description.

Type

str

is_loss

Is this a loss metric.

Type

bool

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.MetricUpdateRequest(*, name: str = None, description: str = None, is_loss: bool = None)

Metric object.

Metric used for evaluation.

name

Metric name.

Type

str, optional

description

Metric description.

Type

str, optional

is_loss

Is this a loss metric.

Type

bool, optional

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.Result(*, id: str, best_rank: int = None, metrics: dict, methodology: str, uses_additional_data: bool, paper: str = None, best_metric: str = None, evaluated_on: str = None, external_source_url: str = None)

Evaluation table row object.

id

Result id.

Type

str

best_rank

Best rank of the row.

Type

int, optional

metrics

Dictionary of metrics and metric values.

Type

dict

methodology

Methodology used for this implementation.

Type

str

uses_additional_data

Does this evaluation uses additional data not provided in the dataset used for other evaluations.

Type

bool

paper

Paper describing the evaluation.

Type

str, optional

best_metric

Name of the best metric.

Type

str, optional

evaluated_on

Date of the result evaluation in YYYY-MM-DD format.

Type

str, optional

external_source_url

The URL to the external source (eg competition).

Type

str, option

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.Results(*, count: int, next_page: int = None, previous_page: int = None, results: List[paperswithcode.models.evaluation.result.Result])

Object representing a paginated page of results.

count

Number of elements matching the query.

Type

int

next_page

Number of the next page.

Type

int, optional

previous_page

Number of the previous page.

Type

int, optional

results

List of results on this page.

Type

List[Result]

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.ResultCreateRequest(*, metrics: dict, methodology: str, uses_additional_data: bool = False, paper: str = None, evaluated_on: str = None, external_source_url: str = None)

Evaluation table row object.

metrics

Dictionary of metrics and metric values.

Type

dict

methodology

Methodology used for this implementation.

Type

str

uses_additional_data

Does this evaluation uses additional data not provided in the dataset used for other valuations.

Type

bool, optional

paper

Paper describing the evaluation.

Type

str, optional

evaluated_on

Date of the result evaluation: YYYY-MM-DD format.

Type

str, optional

external_source_url

The URL to the external source (eg competition).

Type

str, option

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.ResultUpdateRequest(*, metrics: dict = None, methodology: str = None, uses_additional_data: bool = None, paper: str = None, evaluated_on: str = None, external_source_url: str = None)

Evaluation table row object.

metrics

Dictionary of metrics and metric values.

Type

dict, optional

methodology

Methodology used for this implementation.

Type

str, optional

uses_additional_data

Does this evaluation uses additional data not provided in the dataset used for other evaluations.

Type

bool, optional

paper

Paper describing the evaluation.

Type

str, optional

evaluated_on

Date of the result evaluation: YYYY-MM-DD format.

Type

datetime, optional

external_source_url

The URL to the external source (eg competition).

Type

str, option

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.EvaluationTable(*, id: str, task: str, dataset: str, description: str = '', mirror_url: str = None)

Evaluation table object.

id

Evaluation table ID.

Type

str

task

ID of the task used in evaluation.

Type

str

dataset

ID of the dataset used in evaluation.

Type

str

description

Evaluation table description.

Type

str

mirror_url

URL to the evaluation table that this table is a mirror of.

Type

str, optional

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.EvaluationTables(*, count: int, next_page: int = None, previous_page: int = None, results: List[paperswithcode.models.evaluation.table.EvaluationTable])

Object representing a paginated page of evaluation tables.

count

Number of elements matching the query.

Type

int

next_page

Number of the next page.

Type

int, optional

previous_page

Number of the previous page.

Type

int, optional

results

List of evaluation tables on this page.

Type

List[SotaPartial]

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.EvaluationTableCreateRequest(*, task: str, dataset: str, description: str = '', mirror_url: str = None)

Evaluation table create request object.

task

ID of the task used in evaluation.

Type

str

dataset

ID of the dataset used in evaluation.

Type

str

description

Evaluation table description.

Type

str

mirror_url

URL to the evaluation table that this table is a mirror of.

Type

str, optional

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.EvaluationTableUpdateRequest(*, task: str = None, dataset: str = None, description: str = None, mirror_url: str = None)

Evaluation table update request object.

task

ID of the task used in evaluation.

Type

str, optional

dataset

ID of the dataset used in evaluation.

Type

str, optional

description

Evaluation table description.

Type

str, optional

mirror_url

URL to the evaluation table that this table is a mirror of.f

Type

str, optional

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.ResultSyncRequest(*, metrics: dict, methodology: str, paper: str = None, uses_additional_data: bool = False, external_id: str = '', evaluated_on: str, external_source_url: str = None)

Evaluation table row object.

metrics

Dictionary of metrics and metric values.

Type

dict

methodology

Methodology used for this implementation.

Type

str

uses_additional_data

Does this evaluation uses additional data not provided in the dataset used for other evaluations.

Type

bool

paper

Paper describing the evaluation.

Type

str, optional

external_id

Optional external ID used to identify rows when doing sync.

Type

str, optional

evaluated_on

Evaluation date in YYYY-MM-DD format

Type

str

external_source_url

The URL to the external source (eg competition).

Type

str, option

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.MetricSyncRequest(*, name: str, description: str = '', is_loss: bool = True)

Metric object.

Metric used for evaluation.

name

Metric name.

Type

str

description

Metric description.

Type

str

is_loss

Is this a loss metric.

Type

bool

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.EvaluationTableSyncRequest(*, task: str, dataset: str, description: str = '', mirror_url: str = None, external_id: str = None, metrics: List[paperswithcode.models.evaluation.synchronize.MetricSyncRequest] = None, results: List[paperswithcode.models.evaluation.synchronize.ResultSyncRequest] = None)

Evaluation table object.

task

ID of the task used in evaluation.

Type

str

dataset

ID of the dataset used in evaluation.

Type

str

description

Evaluation table description.

Type

str

mirror_url

URL to the evaluation table that this table is a mirror of.

Type

str, optional

external_id

Optional external ID used to identify rows when doing sync.

Type

str, optional

metric

List of MetricSyncRequest objects used in the evaluation.

Type

list

results

List of ResultSyncRequest objects - results of the evaluation.

Type

list

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.ResultSyncResponse(*, id: str, metrics: dict, methodology: str, paper: str = None, uses_additional_data: bool = False, external_id: str = '', evaluated_on: str = None, external_source_url: str = None)

Evaluation table row object.

id

Result id.

Type

str

metrics

Dictionary of metrics and metric values.

Type

dict

methodology

Methodology used for this implementation.

Type

str

uses_additional_data

Does this evaluation uses additional data not provided in the dataset used for other evaluations.

Type

bool

paper

Paper describing the evaluation.

Type

str, optional

external_id

Optional external ID used to identify rows when doing sync.

Type

str, optional

evaluated_on

Evaluation date in YYYY-MM-DD format

Type

str, optional

external_source_url

The URL to the external source (eg competition)

Type

str, option

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.MetricSyncResponse(*, name: str, description: str = '', is_loss: bool = True)

Metric object.

Metric used for evaluation.

name

Metric name.

Type

str

description

Metric description.

Type

str

is_loss

Is this a loss metric.

Type

bool

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

class paperswithcode.models.evaluation.EvaluationTableSyncResponse(*, id: str, task: str, dataset: str, description: str = '', mirror_url: str = None, external_id: str = '', metrics: List[paperswithcode.models.evaluation.synchronize.MetricSyncResponse] = None, results: List[paperswithcode.models.evaluation.synchronize.ResultSyncResponse] = None)

Evaluation table object.

id

Evaluation table ID.

Type

str

task

ID of the task used in evaluation.

Type

str

dataset

ID of the dataset used in evaluation.

Type

str

description

Evaluation table description.

Type

str

mirror_url

URL to the evaluation table that this table is a mirror of.

Type

str, optional

external_id

Optional external ID used to identify rows when doing sync.

Type

str, optional

metric

List of metrics sync objects used in the evaluation.

Type

list

results

List of result sync objects - results of the evaluation.

Type

list

Create a new model by parsing and validating input data from keyword arguments.

Raises ValidationError if the input data cannot be parsed to form a valid model.

PapersWithCode Client Class

class paperswithcode.client.PapersWithCodeClient(token=None, url=None)

PapersWithCode client.

__init__(token=None, url=None)

Initialize self. See help(type(self)) for accurate signature.

search(q: Optional[str] = None, page: int = 1, items_per_page: int = 50) → paperswithcode.models.paper_repo.PaperRepos

Search in a similar fashion to the frontpage search.

Parameters

• q (str, optional) – Filter papers by querying the paper title and abstract.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

PaperRepos object.

Return type

PaperRepos

paper_list(q: Optional[str] = None, arxiv_id: Optional[str] = None, title: Optional[str] = None, abstract: Optional[str] = None, page: int = 1, items_per_page: int = 50) → paperswithcode.models.paper.Papers

Return a paginated list of papers.

Parameters

• q (str, optional) – Filter papers by querying the paper title and abstract.

• arxiv_id (str, optional) – Filter papers by arxiv id.

• title (str, optional) – Filter papers by part of the title.

• abstract (str, optional) – Filter papers by part of the abstract.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Papers object.

Return type

Papers

paper_get(paper_id: str) → paperswithcode.models.paper.Paper

Return a paper by it’s ID.

Parameters

paper_id (str) – ID of the paper.

Returns

Paper object.

Return type

Paper

paper_dataset_list(paper_id: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.repository.Repositories

Return a list of datasets mentioned in the paper..

Parameters

• paper_id (str) – ID of the paper.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Datasets object.

Return type

Datasets

paper_repository_list(paper_id: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.repository.Repositories

Return a list of paper implementations.

Parameters

• paper_id (str) – ID of the paper.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Repositories object.

Return type

Repositories

paper_task_list(paper_id: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.task.Tasks

Return a list of tasks mentioned in the paper.

Parameters

• paper_id (str) – ID of the paper.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Tasks object.

Return type

Tasks

paper_method_list(paper_id: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.method.Methods

Return a list of methods mentioned in the paper.

Parameters

• paper_id (str) – ID of the paper.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Methods object.

Return type

Methods

paper_result_list(paper_id: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.evaluation.result.Results

Return a list of evaluation results for the paper.

Parameters

• paper_id (str) – ID of the paper.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Results object.

Return type

Results

repository_list(q: Optional[str] = None, owner: Optional[str] = None, name: Optional[str] = None, stars: Optional[int] = None, framework: Optional[str] = None, page: int = 1, items_per_page: int = 50) → paperswithcode.models.paper.Papers

Return a paginated list of repositories.

Parameters

• q (str, optional) – Search all searchable fields.

• owner (str, optional) – Filter repositories by owner.

• name (str, optional) – Filter repositories by name.

• stars (int, optional) – Filter repositories by minimum number of stars.

• framework (str, optional) – Filter repositories by framework. Available values: tf, pytorch, mxnet, torch, caffe2, jax, paddle, mindspore.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Repositories object.

Return type

Repositories

repository_owner_list(owner: str) → paperswithcode.models.repository.Repositories

List all repositories for a specific repo owner.

Parameters

owner (str) – Repository owner.

Returns

Repositories object.

Return type

Repositories

repository_get(owner: str, name: str) → paperswithcode.models.repository.Repository

Return a repository by it’s owner/name pair.

Parameters

• owner (str) – Owner name.

• name (str) – Repository name.

Returns

Repository object.

Return type

Repository

repository_paper_list(owner: str, name: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.paper.Papers

List all papers connected to the repository.

Parameters

• owner (str) – Owner name.

• name (str) – Repository name.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Papers object.

Return type

Papers

author_list(q: Optional[str] = None, full_name: Optional[str] = None, page: int = 1, items_per_page: int = 50) → paperswithcode.models.author.Authors

Return a paginated list of paper authors.

Parameters

• q (str, optional) – Search all searchable fields.

• full_name (str, optional) – Filter authors by part of their full name.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Repositories object.

Return type

Repositories

author_get(author_id: str) → paperswithcode.models.author.Author

Return a specific author selected by its id.

Parameters

author_id (str) – Author id.

Returns

Author object.

Return type

Author

author_paper_list(author_id: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.paper.Papers

List all papers connected to the author.

Parameters

• author_id (str) – Author id.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Papers object.

Return type

Papers

conference_list(q: Optional[str] = None, name: Optional[str] = None, page: int = 1, items_per_page: int = 50) → paperswithcode.models.conference.Conferences

Return a paginated list of conferences.

Parameters

• q (str, optional) – Search all searchable fields.

• name (str, optional) – Filter conferences by part of the name.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Conferences object.

Return type

Conferences

conference_get(conference_id: str) → paperswithcode.models.conference.Conference

Return a conference by it’s ID.

Parameters

conference_id (str) – ID of the conference.

Returns

Conference object.

Return type

Conference

proceeding_list(conference_id: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.conference.Proceedings

Return a paginated list of conference proceedings.

Parameters

• conference_id (str) – ID of the conference.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Proceedings object.

Return type

Proceedings

proceeding_get(conference_id: str, proceeding_id: str) → paperswithcode.models.conference.Proceeding

Return a conference proceeding by it’s ID.

Parameters

• conference_id (str) – ID of the conference.

• proceeding_id (str) – ID of the proceeding.

Returns

Proceeding object.

Return type

Proceeding

proceeding_paper_list(conference_id: str, proceeding_id: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.paper.Papers

Return a list of papers published in a confernce proceeding.

Parameters

• conference_id (str) – ID of the conference.

• proceeding_id (str) – ID of the proceding.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Papers object.

Return type

Papers

area_list(q: Optional[str] = None, name: Optional[str] = None, page: int = 1, items_per_page: int = 50) → paperswithcode.models.task.Areas

Return a paginated list of areas.

Parameters

• q (str, optional) – Filter areas by querying the area name.

• name (str, optional) – Filter areas by part of the name.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Areas object.

Return type

Areas

area_get(area_id: str) → paperswithcode.models.task.Area

Return an area by it’s ID.

Parameters

area_id (str) – ID of the area.

Returns

Area object.

Return type

Area

area_task_list(area_id: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.task.Tasks

Return a paginated list of tasks in an area.

Parameters

• area_id (str) – ID of the area.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Tasks object.

Return type

Tasks

task_list(q: Optional[str] = None, name: Optional[str] = None, page: int = 1, items_per_page: int = 50) → paperswithcode.models.task.Tasks

Return a paginated list of tasks.

Parameters

• q (str, optional) – Filter tasks by querying the task name.

• name (str, optional) – Filter tasks by part of th name.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Tasks object.

Return type

Tasks

task_get(task_id: str) → paperswithcode.models.task.Task

Return a task by it’s ID.

Parameters

task_id (str) – ID of the task.

Returns

Task object.

Return type

Task

task_add(task: paperswithcode.models.task.TaskCreateRequest) → paperswithcode.models.task.Task

Add a task.

Parameters

task (TaskCreateRequest) – Task create request.

Returns

Created task.

Return type

Task

task_update(task_id: str, task: paperswithcode.models.task.TaskUpdateRequest) → paperswithcode.models.task.Task

Update a task.

Parameters

• task_id (str) – ID of the task.

• task (TaskUpdateRequest) – Task update request.

Returns

Updated task.

Return type

Task

task_delete(task_id: str)

Delete a task.

Parameters

task_id (str) – ID of the task.

task_parent_list(task_id: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.task.Tasks

Return a paginated list of parent tasks for a selected task.

Parameters

• task_id (str) – ID of the task.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Tasks object.

Return type

Tasks

task_child_list(task_id: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.task.Tasks

Return a paginated list of child tasks for a selected task.

Parameters

• task_id (str) – ID of the task.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Tasks object.

Return type

Tasks

task_paper_list(task_id: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.paper.Papers

Return a paginated list of papers for a selected task.

Parameters

• task_id (str) – ID of the task.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Papers object.

Return type

Papers

task_evaluation_list(task_id: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.evaluation.table.EvaluationTables

Return a list of evaluation tables for a selected task.

Parameters

• task_id (str) – ID of the task.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

EvaluationTables object.

Return type

EvaluationTables

dataset_list(q: Optional[str] = None, name: Optional[str] = None, full_name: Optional[str] = None, page: int = 1, items_per_page: int = 50) → paperswithcode.models.dataset.Datasets

Return a paginated list of datasets.

Parameters

• q (str, optional) – Filter datasets by querying the dataset name.

• name (str, optional) – Filter datasets by their name.

• full_name (str, optional) – Filter datasets by their full name.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Datasets object.

Return type

Datasets

dataset_get(dataset_id: str) → paperswithcode.models.dataset.Dataset

Return a dastaset by it’s ID.

Parameters

dataset_id (str) – ID of the dataset.

Returns

Dataset object.

Return type

Dataset

dataset_add(dataset: paperswithcode.models.dataset.DatasetCreateRequest) → paperswithcode.models.dataset.Dataset

Add a dataset.

Parameters

dataset (DatasetCreateRequest) – Dataset create request.

Returns

Created dataset.

Return type

Dataset

dataset_update(dataset_id: str, dataset: paperswithcode.models.dataset.DatasetUpdateRequest) → paperswithcode.models.dataset.Dataset

Update a dataset.

Parameters

• dataset_id (str) – ID of the dataset.

• dataset (DatasetUpdateRequest) – Dataset update request.

Returns

Updated dataset.

Return type

Dataset

dataset_delete(dataset_id: str)

Delete a dataset.

Parameters

dataset_id (str) – ID of the dataset.

dataset_evaluation_list(dataset_id: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.evaluation.table.EvaluationTables

Return a list of evaluation tables for a selected dataset.

Parameters

• dataset_id (str) – ID of the dasaset.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

EvaluationTables object.

Return type

EvaluationTables

method_list(q: Optional[str] = None, name: Optional[str] = None, full_name: Optional[str] = None, page: int = 1, items_per_page: int = 50) → paperswithcode.models.method.Methods

Return a paginated list of methods.

Parameters

• q (str, optional) – Search all searchable fields.

• name (str, optional) – Filter methods by part of the name.

• full_name (str, optional) – Filter methods by part of the full name.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Methods object.

Return type

Methods

method_get(method_id) → paperswithcode.models.method.Method

Return a method by it’s ID.

Parameters

method_id (str) – ID of the method.

Returns

Method object.

Return type

Method

evaluation_list(page: int = 1, items_per_page: int = 50) → paperswithcode.models.evaluation.table.EvaluationTables

Return a paginated list of evaluation tables.

Parameters

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Evaluation table page object.

Return type

EvaluationTables

evaluation_get(evaluation_id: str) → paperswithcode.models.evaluation.table.EvaluationTable

Return a evaluation table by it’s ID.

Parameters

evaluation_id (str) – ID of the evaluation table.

Returns

Evaluation table object.

Return type

EvaluationTable

evaluation_create(evaluation: paperswithcode.models.evaluation.table.EvaluationTableCreateRequest) → paperswithcode.models.evaluation.table.EvaluationTable

Create an evaluation table.

Parameters

evaluation (EvaluationTableCreateRequest) – Evaluation table create request object.

Returns

The new created evaluation table.

Return type

EvaluationTable

evaluation_update(evaluation_id: str, evaluation: paperswithcode.models.evaluation.table.EvaluationTableUpdateRequest) → paperswithcode.models.evaluation.table.EvaluationTable

Update an evaluation table.

Parameters

• evaluation_id (str) – ID of the evaluation table.

• evaluation (EvaluationTableUpdateRequest) – Evaluation table update request object.

Returns

The updated evaluation table.

Return type

EvaluationTable

evaluation_delete(evaluation_id: str)

Delete an evaluation table.

Parameters

evaluation_id (str) – ID of the evaluation table.

evaluation_metric_list(evaluation_id: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.evaluation.metric.Metrics

List all metrics used in the evaluation table.

Parameters

• evaluation_id (str) – ID of the evaluation table.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Metrics object.

Return type

Metrics

evaluation_metric_get(evaluation_id: str, metric_id: str) → paperswithcode.models.evaluation.metric.Metric

Get a metrics used in the evaluation table.

Parameters

• evaluation_id (str) – ID of the evaluation table.

• metric_id (str) – ID of the metric.

Returns

Requested metric.

Return type

Metric

evaluation_metric_add(evaluation_id: str, metric: paperswithcode.models.evaluation.metric.MetricCreateRequest) → paperswithcode.models.evaluation.metric.Metric

Add a metrics to the evaluation table.

Parameters

• evaluation_id (str) – ID of the evaluation table.

• metric (MetricCreateRequest) – Metric create request.

Returns

Created metric.

Return type

Metric

evaluation_metric_update(evaluation_id: str, metric_id: str, metric: paperswithcode.models.evaluation.metric.MetricUpdateRequest) → paperswithcode.models.evaluation.metric.Metric

Update a metrics in the evaluation table.

Parameters

• evaluation_id (str) – ID of the evaluation table.

• metric_id (str) – ID of the metric.

• metric (MetricCreateRequest) – Metric update request.

Returns

Updated metric.

Return type

Metric

evaluation_metric_delete(evaluation_id: str, metric_id: str)

Delete a metrics from the evaluation table.

Parameters

• evaluation_id (str) – ID of the evaluation table.

• metric_id (str) – ID of the metric.

evaluation_result_list(evaluation_id: str, page: int = 1, items_per_page: int = 50) → paperswithcode.models.evaluation.result.Results

List all results from the evaluation table.

Parameters

• evaluation_id (str) – ID of the evaluation table.

• page (int) – Desired page.

• items_per_page (int) – Desired number of items per page. Default: 50.

Returns

Results object.

Return type

Results

evaluation_result_get(evaluation_id: str, result_id: str) → paperswithcode.models.evaluation.result.Result

Get a result from the evaluation table.

Parameters

• evaluation_id (str) – ID of the evaluation table.

• result_id (str) – ID of the result.

Returns

Requested result.

Return type

Result

evaluation_result_add(evaluation_id: str, result: paperswithcode.models.evaluation.result.ResultCreateRequest) → paperswithcode.models.evaluation.result.Result

Add a result to the evaluation table.

Parameters

• evaluation_id (str) – ID of the evaluation table.

• result (ResultCreateRequest) – Result create request.

Returns

Created result.

Return type

Result

evaluation_result_update(evaluation_id: str, result_id: str, result: paperswithcode.models.evaluation.result.ResultUpdateRequest) → paperswithcode.models.evaluation.result.Result

Update a result in the evaluation table.

Parameters

• evaluation_id (str) – ID of the evaluation table.

• result_id (str) – ID of the result.

• result (ResultUpdateRequest) – Result update request.

Returns

Updated result.

Return type

Result

evaluation_result_delete(evaluation_id: str, result_id: str)

Delete a result from the evaluation table.

Parameters

• evaluation_id (str) – ID of the evaluation table.

• result_id (str) – ID of the result.

Indices and tables

• Index

• Module Index