spaCy plugin for Transformers , Udify, ELmo, etc.

Related tags

Text Data & NLPspacyspacy-extension
Overview

Camphr - spaCy plugin for Transformers, Udify, Elmo, etc.

Documentation Status Gitter PyPI version test and publish

Camphr is a Natural Language Processing library that helps in seamless integration for a wide variety of techniques from state-of-the-art to conventional ones. You can use Transformers , Udify, ELmo, etc. on spaCy.

Check the documentation for more information.

(For Japanese: https://qiita.com/tamurahey/items/53a1902625ccaac1bb2f)

Features

  • A spaCy plugin - Easily integration for a wide variety of methods
  • Transformers with spaCy - Fine-tuning pretrained model with Hydra. Embedding vector
  • Udify - BERT based multitask model in 75 languages
  • Elmo - Deep contextualized word representations
  • Rule base matching with Aho-Corasick, Regex
  • (for Japanese) KNP

License

Camphr is licensed under Apache 2.0.

Comments
  • NER Problem

    NER Problem

    Hello!

    First of all I would like to thank you for the great work on lib Camphr. It's been very useful to me! Can you help me with this doubt? I used lib to train a name recognition model (ner) but when I load the model using nlp = (spacy.load ("~ / outputs // 2020-04-30 // 22-28-36 // models // 9 "), and I pass a text (doc = nlp (" I live in Brazil ")), I can't get any entity recognition (doc.ents >> ()). Could you tell me why this is happening?

    opened by gabrielluz07 9
  • Gender and number subtags generation

    Gender and number subtags generation

    I was comparing the default morpho-syntactic tags generated by camphr-udify and https://github.com/Hyperparticle/udify.

    import spacy
import stanza
from spacy_conll import ConllFormatter

nlp=spacy.load("en_udify")
conllformatter = ConllFormatter(nlp)
nlp.add_pipe(conllformatter, last=True)

doc=nlp("Mother Teresa devoted her entire life to helping others") 
print(doc._.conll_str)


    1	Mother	Mother	PROPN		_	2	compound	_	_
2	Teresa	Teresa	PROPN		_	3	nsubj	_	_
3	devoted	devote	VERB		_	0	root	_	_
4	her	her	PRON		_	6	nmod:poss	_	_
5	entire	entire	ADJ		_	6	amod	_	_
6	life	life	NOUN		_	3	obj	_	_
7	to	to	SCONJ		_	8	mark	_	_
8	helping	help	VERB		_	3	advcl	_	_
9	others	other	NOUN		_	8	obj	_	SpaceAfter=No

    Tags returned by https://github.com/Hyperparticle/udify, for the same input.

    prediction:  1  Mother  Mother  PROPN   _       Number=Sing     2       compound        _       _
2       Teresa  Teresa  PROPN   _       Number=Sing     3       nsubj   _       _
3       devoted devote  VERB    _       Mood=Ind|Tense=Past|VerbForm=Fin        0       root    _       _
4       her     her     PRON    _       Gender=Fem|Number=Sing|Person=3|Poss=Yes|PronType=Prs   6       nmod:poss      _                                               _
5       entire  entire  ADJ     _       Degree=Pos      6       amod    _       _
6       life    life    NOUN    _       Number=Sing     3       obj     _       _
7       to      to      SCONJ   _       _       8       mark    _       _
8       helping help    VERB    _       VerbForm=Ger    3       advcl   _       _
9       others  other   NOUN    _       Number=Plur     8       obj     _       _

    Gender and number subtags are missing in camphr-udify. Could we have those included by default please?

    thanks, Ranjita

    enhancement 
    opened by ranjita-naik 6
  • Camphr+KNP returns an incorrect dependency tag when using a specific adposition.

    Camphr+KNP returns an incorrect dependency tag when using a specific adposition.

    Hello. I report a problem that is happened when analyzing universal dependencies in Japanese text using KNP. When I use a adposition “から”, camphr returns a following wrong result (that shows the conj dependency tag on NOUN→VERB, but an expectation result is the obl dependency tag on VERB→NOUN).

    例1 例2

    (Note that "再結晶", "留去" are the words I added manually, but other VERB words that existed in the original dictionary such as "除去", "撹拌" generates similarly incorrect results.) Same problems sometimes occur when using an adposition "と".

    But using other adpositions, such as “より”, “にて”, camphr returns a correct result.

    例3 例4

    Environment:

    • Docker(python:3.7-buster)
    • spacy = 2.3.2
    • camphr = 0.6.5
    • pyknp = 0.4.5
    • Juman++ ver.1.02
    • KNP ver.4.19
    opened by undermakingbook 6
  • Python 3.8

    Python 3.8

    Camphr is currently pinned at python < 3.8, is there a specific reason for this and if so, what can we do to help?

    Edit: sorry, I just saw #19, still, what can we do to help?

    opened by Evpok 5
  • Support multi labels textcat pipe for transformers

    Support multi labels textcat pipe for transformers

    closes #9

    • Add TrfForMultiLabelSequenceClassification for multiple text classification.
      • pipe name: transformers_multilabel_sequence_classifier
    • Add docs for fine-tuning multi textcat pipe
      • https://github.com/PKSHATechnology-Research/camphr/blob/feature%2Fmulti-textcat/docs/source/notes/finetune_transformers.rst#multilabel-text-classification
    enhancement 
    opened by tamuhey 5
  • unofficial-udify, allennlp, and transformers conflicting dependencies

    unofficial-udify, allennlp, and transformers conflicting dependencies

    I'm trying to install udify on WSL as shown below.

    $ pip install unofficial-udify==0.3.0 [email protected]://github.com/PKSHATechnology-Research/camphr_models/releases/download/0.7.0/en_udify-0.7.tar.gz

    ERROR: Cannot install unofficial-udify and unofficial-udify==0.3.0 because these package versions have conflicting dependencies.

    The conflict is caused by: unofficial-udify 0.3.0 depends on transformers<3.0.0 and >=2.3.0 allennlp 1.3.0 depends on transformers<4.1 and >=4.0 unofficial-udify 0.3.0 depends on transformers<3.0.0 and >=2.3.0 allennlp 1.2.2 depends on transformers<3.6 and >=3.4 unofficial-udify 0.3.0 depends on transformers<3.0.0 and >=2.3.0 allennlp 1.2.1 depends on transformers<3.5 and >=3.1 unofficial-udify 0.3.0 depends on transformers<3.0.0 and >=2.3.0 allennlp 1.2.0 depends on transformers<3.5 and >=3.1 unofficial-udify 0.3.0 depends on transformers<3.0.0 and >=2.3.0 allennlp 1.1.0 depends on transformers<3.1 and >=3.0

    Is this a known issue? Could you suggest a workaroudn please?

    bug 
    opened by ranjita-naik 3
  • Missing tag information

    Missing tag information

    I noticed that the spacy tag field is empty. Is this a known issue? It looks like Udify supports some level of ufeats tagging (https://universaldependencies.org/u/feat/index.html)? I wonder if I'm supposed to b getting any of this in Spacy and I have a bug in my setup, or if it just isn't implemented yet? Would it be souced in token.tag like I'm thinking (if it does exist)?

    I also noticed that displacy doesn't render the POS info. I am wondering if that is related?

    BTW, just have to say that this is awesome.

    opened by tslater 3
  • ImportError: cannot import name 'load_udify' from 'camphr.pipelines' following the example

    ImportError: cannot import name 'load_udify' from 'camphr.pipelines' following the example

    I followed the example here: https://camphr.readthedocs.io/en/latest/notes/udify.html

    I did only see the 0.7.0 model, so I went with that instead. Anyway, the German and English examples work great, but the Japanese one gives me this error:

    >>> from camphr.pipelines import load_udify
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ImportError: cannot import name 'load_udify' from 'camphr.pipelines' (/home/tyler/camphr/env/lib/python3.8/site-packages/camphr/pipelines/__init__.py)
    opened by tslater 3
  • doc.ents empty, doc.is_nered == False

    doc.ents empty, doc.is_nered == False

    I followed the documentation to fine-tune the bert-base-cased (en) ner model and then made a spacy doc with text "Bob Jones and Barack Obama went up the hill in Wisconsin." but the resulting doc has doc.ents = () and doc.is_nered = False.

    Am I missing something?

    Thank you!

    opened by jack-rory-staunton 3
  • Improvement for サ変 of KNP

    Improvement for サ変 of KNP

    Inside _get_child_dep(c), pos for 名詞,サ変名詞 is changed into VERB when it is followed by AUX. So now I think that _get_dep(tag[0]) should be done after _get_child_dep(c).

    opened by KoichiYasuoka 3
  • Bump transformers from 3.0.2 to 4.1.1

    Bump transformers from 3.0.2 to 4.1.1

    Bumps transformers from 3.0.2 to 4.1.1.

    Release notes

    Sourced from transformers's releases.

    Patch release: better error message & invalid trainer attribute

    This patch releases introduces:

    • A better error message when trying to instantiate a SentencePiece-based tokenizer without having SentencePiece installed. #8881
    • Fixes an incorrect attribute in the trainer. #8996

    Transformers v4.0.0: Fast tokenizers, model outputs, file reorganization

    Transformers v4.0.0-rc-1: Fast tokenizers, model outputs, file reorganization

    Breaking changes since v3.x

    Version v4.0.0 introduces several breaking changes that were necessary.

    1. AutoTokenizers and pipelines now use fast (rust) tokenizers by default.

    The python and rust tokenizers have roughly the same API, but the rust tokenizers have a more complete feature set. The main breaking change is the handling of overflowing tokens between the python and rust tokenizers.

    How to obtain the same behavior as v3.x in v4.x

    In version v3.x:

    from transformers import AutoTokenizer

    tokenizer = AutoTokenizer.from_pretrained("xxx")

    to obtain the same in version v4.x:

    from transformers import AutoTokenizer

    tokenizer = AutoTokenizer.from_pretrained("xxx", use_fast=False)

    2. SentencePiece is removed from the required dependencies

    The requirement on the SentencePiece dependency has been lifted from the setup.py. This is done so that we may have a channel on anaconda cloud without relying on conda-forge. This means that the tokenizers that depend on the SentencePiece library will not be available with a standard transformers installation.

    This includes the slow versions of:

    • XLNetTokenizer
    • AlbertTokenizer
    • CamembertTokenizer
    • MBartTokenizer
    • PegasusTokenizer
    • T5Tokenizer
    • ReformerTokenizer
    • XLMRobertaTokenizer

    How to obtain the same behavior as v3.x in v4.x

    Commits

    Dependabot compatibility score

    Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

    Dependabot commands and options

    You can trigger Dependabot actions by commenting on this PR:

    • @dependabot rebase will rebase this PR
    • @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
    • @dependabot merge will merge this PR after your CI passes on it
    • @dependabot squash and merge will squash and merge this PR after your CI passes on it
    • @dependabot cancel merge will cancel a previously requested merge and block automerging
    • @dependabot reopen will reopen this PR if it is closed
    • @dependabot close will close this PR and stop Dependabot recreating it. You can achieve the same result by closing it manually
    • @dependabot ignore this major version will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)
    • @dependabot ignore this minor version will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)
    • @dependabot ignore this dependency will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)
    • @dependabot use these labels will set the current labels as the default for future PRs for this repo and language
    • @dependabot use these reviewers will set the current reviewers as the default for future PRs for this repo and language
    • @dependabot use these assignees will set the current assignees as the default for future PRs for this repo and language
    • @dependabot use this milestone will set the current milestone as the default for future PRs for this repo and language
    • @dependabot badge me will comment on this PR with code to add a "Dependabot enabled" badge to your readme

    Additionally, you can set the following in your Dependabot dashboard:

    • Update frequency (including time of day and day of week)
    • Pull request limits (per update run and/or open at any time)
    • Out-of-range updates (receive only lockfile updates, if desired)
    • Security updates (receive only security updates, if desired)
    dependencies 
    opened by dependabot-preview[bot] 2
  • Bump certifi from 2021.5.30 to 2022.12.7 in /packages/camphr_pattern_search

    Bump certifi from 2021.5.30 to 2022.12.7 in /packages/camphr_pattern_search

    Bumps certifi from 2021.5.30 to 2022.12.7.

    Commits

    Dependabot compatibility score

    Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

    Dependabot commands and options

    You can trigger Dependabot actions by commenting on this PR:

    • @dependabot rebase will rebase this PR
    • @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
    • @dependabot merge will merge this PR after your CI passes on it
    • @dependabot squash and merge will squash and merge this PR after your CI passes on it
    • @dependabot cancel merge will cancel a previously requested merge and block automerging
    • @dependabot reopen will reopen this PR if it is closed
    • @dependabot close will close this PR and stop Dependabot recreating it. You can achieve the same result by closing it manually
    • @dependabot ignore this major version will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)
    • @dependabot ignore this minor version will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)
    • @dependabot ignore this dependency will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)
    • @dependabot use these labels will set the current labels as the default for future PRs for this repo and language
    • @dependabot use these reviewers will set the current reviewers as the default for future PRs for this repo and language
    • @dependabot use these assignees will set the current assignees as the default for future PRs for this repo and language
    • @dependabot use this milestone will set the current milestone as the default for future PRs for this repo and language

    You can disable automated security fix PRs for this repo from the Security Alerts page.

    dependencies 
    opened by dependabot[bot] 0
  • Bump numpy from 1.21.0 to 1.22.0 in /packages/camphr_pattern_search

    Bump numpy from 1.21.0 to 1.22.0 in /packages/camphr_pattern_search

    Bumps numpy from 1.21.0 to 1.22.0.

    Release notes

    Sourced from numpy's releases.

    v1.22.0

    NumPy 1.22.0 Release Notes

    NumPy 1.22.0 is a big release featuring the work of 153 contributors spread over 609 pull requests. There have been many improvements, highlights are:

    • Annotations of the main namespace are essentially complete. Upstream is a moving target, so there will likely be further improvements, but the major work is done. This is probably the most user visible enhancement in this release.
    • A preliminary version of the proposed Array-API is provided. This is a step in creating a standard collection of functions that can be used across application such as CuPy and JAX.
    • NumPy now has a DLPack backend. DLPack provides a common interchange format for array (tensor) data.
    • New methods for quantile, percentile, and related functions. The new methods provide a complete set of the methods commonly found in the literature.
    • A new configurable allocator for use by downstream projects.

    These are in addition to the ongoing work to provide SIMD support for commonly used functions, improvements to F2PY, and better documentation.

    The Python versions supported in this release are 3.8-3.10, Python 3.7 has been dropped. Note that 32 bit wheels are only provided for Python 3.8 and 3.9 on Windows, all other wheels are 64 bits on account of Ubuntu, Fedora, and other Linux distributions dropping 32 bit support. All 64 bit wheels are also linked with 64 bit integer OpenBLAS, which should fix the occasional problems encountered by folks using truly huge arrays.

    Expired deprecations

    Deprecated numeric style dtype strings have been removed

    Using the strings "Bytes0", "Datetime64", "Str0", "Uint32", and "Uint64" as a dtype will now raise a TypeError.

    (gh-19539)

    Expired deprecations for loads, ndfromtxt, and mafromtxt in npyio

    numpy.loads was deprecated in v1.15, with the recommendation that users use pickle.loads instead. ndfromtxt and mafromtxt were both deprecated in v1.17 - users should use numpy.genfromtxt instead with the appropriate value for the usemask parameter.

    (gh-19615)

    ... (truncated)

    Commits

    Dependabot compatibility score

    Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

    Dependabot commands and options

    You can trigger Dependabot actions by commenting on this PR:

    • @dependabot rebase will rebase this PR
    • @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
    • @dependabot merge will merge this PR after your CI passes on it
    • @dependabot squash and merge will squash and merge this PR after your CI passes on it
    • @dependabot cancel merge will cancel a previously requested merge and block automerging
    • @dependabot reopen will reopen this PR if it is closed
    • @dependabot close will close this PR and stop Dependabot recreating it. You can achieve the same result by closing it manually
    • @dependabot ignore this major version will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)
    • @dependabot ignore this minor version will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)
    • @dependabot ignore this dependency will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)
    • @dependabot use these labels will set the current labels as the default for future PRs for this repo and language
    • @dependabot use these reviewers will set the current reviewers as the default for future PRs for this repo and language
    • @dependabot use these assignees will set the current assignees as the default for future PRs for this repo and language
    • @dependabot use this milestone will set the current milestone as the default for future PRs for this repo and language

    You can disable automated security fix PRs for this repo from the Security Alerts page.

    dependencies 
    opened by dependabot[bot] 0
Releases(0.7.0)
Owner
GitHub Repository https://camphr.readthedocs.io/en/latest/
Mysticbbs-rjam - rJAM splitscreen message reader for MysticBBS A46+

rJAM splitscreen message reader for MysticBBS A46+

Robbert Langezaal 4 Nov 22, 2022
Pre-Training with Whole Word Masking for Chinese BERT

Pre-Training with Whole Word Masking for Chinese BERT

Yiming Cui 7.7k Dec 31, 2022
2021搜狐校园文本匹配算法大赛baseline

sohu2021-baseline 2021搜狐校园文本匹配算法大赛baseline 简介 分享了一个搜狐文本匹配的baseline，主要是通过条件LayerNorm来增加模型的多样性，以实现同一模型处理不同类型的数据、形成不同输出的目的。 线下验证集F1约0.74，线上测试集F1约0.73。

苏剑林(Jianlin Su) 45 Sep 06, 2022
Guide to using pre-trained large language models of source code

Large Models of Source Code I occasionally train and publicly release large neural language models on programs, including PolyCoder. Here, I describe

Vincent Hellendoorn 947 Dec 28, 2022
An example project using OpenPrompt under pytorch-lightning for prompt-based SST2 sentiment analysis model

pl_prompt_sst An example project using OpenPrompt under the framework of pytorch-lightning for a training prompt-based text classification model on SS

Zhiling Zhang 5 Oct 21, 2022
Universal End2End Training Platform, including pre-training, classification tasks, machine translation, and etc.

背景 安装教程 快速上手 （一）预训练模型 （二）机器翻译 （三）文本分类 TenTrans 进阶 1. 多语言机器翻译 2. 跨语言预训练 背景 TrenTrans是一个统一的端到端的多语言多任务预训练平台，支持多种预训练方式，以及序列生成和自然语言理解任务。 安装教程 git clone git

Tencent Minority-Mandarin Translation Team 42 Dec 20, 2022
A single model that parses Universal Dependencies across 75 languages.

A single model that parses Universal Dependencies across 75 languages. Given a sentence, jointly predicts part-of-speech tags, morphology tags, lemmas, and dependency trees.

Dan Kondratyuk 189 Nov 29, 2022
Text Analysis & Topic Extraction on Android App user reviews

AndroidApp_TextAnalysis Hi, there! This is code archive for Text Analysis and Topic Extraction from user_reviews of Android App. Dataset Source : http

Fitrie Ratnasari 1 Feb 14, 2022
Simple Text-Generator with OpenAI gpt-2 Pytorch Implementation

GPT2-Pytorch with Text-Generator Better Language Models and Their Implications Our model, called GPT-2 (a successor to GPT), was trained simply to pre

Tae-Hwan Jung 775 Jan 08, 2023
DziriBERT: a Pre-trained Language Model for the Algerian Dialect

DziriBERT is the first Transformer-based Language Model that has been pre-trained specifically for the Algerian Dialect.

117 Jan 07, 2023
RuCLIP tiny (Russian Contrastive Language–Image Pretraining) is a neural network trained to work with different pairs (images, texts).

RuCLIPtiny Zero-shot image classification model for Russian language RuCLIP tiny (Russian Contrastive Language–Image Pretraining) is a neural network

Shahmatov Arseniy 26 Sep 20, 2022
NeuralQA: A Usable Library for Question Answering on Large Datasets with BERT

NeuralQA: A Usable Library for (Extractive) Question Answering on Large Datasets with BERT Still in alpha, lots of changes anticipated. View demo on n

Victor Dibia 220 Dec 11, 2022
Partially offline multi-language translator built upon Huggingface transformers.

Translate Command-line interface to translation pipelines, powered by Huggingface transformers. This tool can download translation models, and then us

Richard Jarry 8 Oct 25, 2022
This is a GUI program that will generate a word search puzzle image

Word Search Puzzle Generator Table of Contents About The Project Built With Getting Started Prerequisites Installation Usage Roadmap Contributing Cont

11 Feb 22, 2022
Russian GPT3 models.

Russian GPT-3 models (ruGPT3XL, ruGPT3Large, ruGPT3Medium, ruGPT3Small) trained with 2048 sequence length with sparse and dense attention blocks. We also provide Russian GPT-2 large model (ruGPT2Larg

Sberbank AI 1.6k Jan 05, 2023
Spacy-ginza-ner-webapi - Named Entity Recognition API with spaCy and GiNZA

Named Entity Recognition API with spaCy and GiNZA I wrote a blog post about this

Yuki Okuda 3 Feb 27, 2022
DANeS is an open-source E-newspaper dataset by collaboration between DATASET JSC (dataset.vn) and AIV Group (aivgroup.vn)

DANeS - Open-source E-newspaper dataset Source: Technology vector created by macrovector - www.freepik.com. DANeS is an open-source E-newspaper datase

DATASET .JSC 64 Aug 17, 2022
Wrapper to display a script output or a text file content on the desktop in sway or other wlroots-based compositors

nwg-wrapper This program is a part of the nwg-shell project. This program is a GTK3-based wrapper to display a script output, or a text file content o

Piotr Miller 94 Dec 27, 2022
Binaural Speech Synthesis

Binaural Speech Synthesis This repository contains code to train a mono-to-binaural neural sound renderer. If you use this code or the provided datase

Facebook Research 135 Dec 18, 2022
Unsupervised text tokenizer focused on computational efficiency

YouTokenToMe YouTokenToMe is an unsupervised text tokenizer focused on computational efficiency. It currently implements fast Byte Pair Encoding (BPE)

VK.com 847 Dec 19, 2022