📌 FredT5-large

The model corrects spelling errors and typos by bringing all the words in the text to the norm of the Russian language. The proofreader was trained based on the FredT5-large model. An extensive dataset with “artificial” errors was taken as a training corpus: the corpus was assembled on the basis of the Russian-language Wikipedia and transcripts of Russian-language videos, then typos and spelling errors were automatically introduced into it using the functionality of the SAGE library.

Table of contents

Public references

Examples

Input

Output

Думю ешцъа лет череа 10 ретроспективно просматривотьэ то будкетцц мне невероя тна ин те р но

Думаю еще лет через 10 ретроспективно просматривать это будет мне невероятно интересно. Думаю это лет через 10 ретроспективно просматривать это будет мне невероятно интересно.

Основая цель мероприятия - практическая отработка навыков по оказанию помощи гражданам, попавшим в ДТП, а также повышение и совершенствование уровня профессиональной подготовки сотрудников МЧС при проведении аварийно-спасательных работ по ликвидации последствий дорожно-транспортных проишествий, сокращение временных показателей реагирования.

Основная цель мероприятия - практическая отработка навыков по оказанию помощи гражданам, попавшим в ДТП, а также повышение и совершенствование уровня профессиональной подготовки сотрудников МЧС при проведении аварийно-спасательных работ по ликвидации последствий дорожно-транспортных происшествий, сокращение временных показателей реагирования. Основная цель мероприятия

прийдя в МГТУ я был удивлен никого необноружив там…

прийдя в МГТУ я был удивлен никого не обнаружив там.. «при

Metrics

Below are automatic metrics for determining the correctness of the spell checkers. We compare our solution with both open automatic spell checkers and the ChatGPT family of models on all four available datasets: - RUSpellRU: texts collected from LiveJournal, with manually corrected typos and errors; - MultidomainGold: examples from 7 text sources, including the open web, news, social media, reviews, subtitles, policy documents and literary works; - MedSpellChecker: texts with errors from medical anamnesis; - GitHubTypoCorpusRu: spelling errors and typos in commits from GitHub;

RUSpellRU

Model

Precision

Recall

F1

FredT5-large-spell

58.5

42.4

49.2

ChatGPT gpt-3.5-turbo-0301

55.8

75.3

64.1

ChatGPT gpt-4-0314

57.0

75.9

63.9

ChatGPT text-davinci-003

55.9

75.3

64.2

Yandex.Speller

83.0

59.8

69.5

JamSpell

42.1

32.8

36.9

HunSpell

31.3

34.9

33.0

MultidomainGold

Model

Precision

Recall

F1

FredT5-large-spell

42.5

42.0

42.2

ChatGPT gpt-3.5-turbo-0301

33.8

72.1

46.0

ChatGPT gpt-4-0314

34.0

73.2

46.4

ChatGPT text-davinci-003

33.6

72.0

45.8

Yandex.Speller

52.9

51.4

52.2

JamSpell

25.7

30.6

28.0

HunSpell

16.2

40.1

23.0

MedSpellChecker

Model

Precision

Recall

F1

FredT5-large-spell

37.2

51.7

43.3

ChatGPT gpt-3.5-turbo-0301

53.2

67.6

59.6

ChatGPT gpt-4-0314

54.2

69.4

60.9

ChatGPT text-davinci-003

47.8

68.4

56.3

Yandex.Speller

80.6

47.8

60.0

JamSpell

24.6

29.7

26.9

HunSpell

10.3

40.2

16.4

GitHubTypoCorpusRu

Model

Precision

Recall

F1

FredT5-large-spell

52.7

41.7

46.6

ChatGPT gpt-3.5-turbo-0301

43.8

57.0

49.6

ChatGPT gpt-4-0314

45.2

58.2

51.0

ChatGPT text-davinci-003

46.5

58.1

51.7

Yandex.Speller

67.7

37.5

48.3

JamSpell

49.5

29.9

37.3

HunSpell

28.5

30.7

29.6

How to use

from transformers import T5ForConditionalGeneration, AutoTokenizer

path_to_model = "ai-forever/FRED-T5-large-spell"
model = T5ForConditionalGeneration.from_pretrained(path_to_model)
tokenizer = AutoTokenizer.from_pretrained(path_to_model, eos_token="</s>")

prefix = "Исправь: "
sentence = "прийдя в МГТУ я был удивлен никого необноружив там…"
sentence = prefix + sentence
encodings = tokenizer(sentence, return_tensors="pt")
generated_tokens = model.generate(
                **encodings, eos_token_id=tokenizer.eos_token_id, early_stopping=True)
answer = tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)
print(answer)

# ["прийдя в МГТУ я был удивлен никого не обнаружив там.. «при"]

API

class sage.spelling_correction.t5_correctors.T5ModelForSpellingCorruption(model_name_or_path: str | PathLike)

Bases: Corrector

T5-based models.

batch_correct(sentences: List[str], batch_size: int, prefix: str | None = '', **generation_params) → List[List[Any]]

Corrects multiple sentences.

Parameters:

  • sentences (list of str) – input sentences to correct;

  • batch_size (int) – size of subsample of input sentences;

  • prefix (str) – some models need some sort of a prompting;

  • generation_params (dict) – parameters passed to generate method of a HuggingFace model;

Returns:

corresponding corrections

Return type:

list of list of str

correct(sentence: str, prefix: str | None = '', **generation_params) → List[str]

Corrects a single input sentence.

Parameters:

  • sentence (str) – a source sentence;

  • prefix (str) – some models need some sort of a prompting;

  • generation_params (dict) – parameters passed to generate method of a HuggingFace model;

Returns:

corresponding corrected sentence

Return type:

list of str

evaluate(dataset_name_or_path: str | PathLike | None, metrics: List, batch_size: int, prefix: str = '', dataset_split: str = 'test', **generation_params) → Dict[str, float]

Evaluate the particular model on the spellcheck datasets.

Parameters:

  • dataset_name_or_path (str) – a path to a locally situated dataset or a name of a dataset on HuggingFace;

  • metrics (list of str) – set of metrics to be used to report performance;

  • batch_size (int) – size of subsample of input sentences;

  • prefix (str) – some models need some sort of a prompting;

  • dataset_split (str) – train / test / dev part to be evaluated on;

  • generation_params (dict) – parameters passed to generate method of a HuggingFace model;

Returns:

mapping between metric’s name and its corresponding value

Return type:

dict[str, float]

classmethod from_pretrained(model_name_or_path: str | PathLike)

Initialize the T5-type corrector from a pre-trained checkpoint. The latter can be either locally situated checkpoint or a name of a model on HuggingFace.

Parameters:

model_name_or_path (str or os.PathLike) – the aforementioned name or path to checkpoint;

Returns:

corrector initialized from pre-trained weights

Return type:

object of T5ModelForSpellingCorruption

Resources

License

Model [FRED-T5-large](https://huggingface.co/ai-forever/FRED-T5-large), on the basis of which our solution is made, and its source code are supplied under the MIT license. Our solution also comes with MIT license.

Specifications

  • File size: 3.5 Gb;

  • Framework: pytorch

  • Format: AI Service

  • Version: v1.0

  • Developer: SberDevices, AGI NLP

Contacts

nikita.martynov.98@list.ru