natural-language-processing-11

XLM-R

3260 papers • 126 benchmarks • 313 datasets

XLM-R

(Image credit: Papersgraph)

Benchmarks

These leaderboards are used to track progress in xlm-r-11

Trend

Dataset

Best Model

Actions

No benchmarks available.

Libraries

i

Use these libraries to find xlm-r-11 models and implementations

huggingface/transformers

2 papers 114,634

Datasets

Belebele

Subtasks

No subtasks available.

Most implemented papers

Unsupervised Cross-lingual Representation Learning at Scale

Edouard Grave, Alexis Conneau, Luke Zettlemoyer, Veselin Stoyanov, Francisco (Paco) Guzmán, Myle Ott, Vishrav Chaudhary, Guillaume Wenzek, Naman Goyal, Kartikay Khandelwal•Mon Nov 04 2019

It is shown that pretraining multilingual language models at scale leads to significant performance gains for a wide range of cross-lingual transfer tasks, and the possibility of multilingual modeling without sacrificing per-language performance is shown for the first time.

Content

Adapter-Hub/adapter-transformers

2 papers 1,961

ai4bharat/indicbert

2 papers 40

7766

0

Paper Graph

MASSIVE: A 1M-Example Multilingual Natural Language Understanding Dataset with 51 Typologically-Diverse Languages

P. Natarajan, Jack G. M. FitzGerald, C. Hench, Charith Peris, Scott Mackie, Kay Rottmann, A. Sánchez, Aaron Nash, Liam Urbach, Vishesh Kakarala, Richa Singh, Swetha Ranganath, Laurie Crist, Misha Britan, Wouter Leeuwis, Gokhan Tur•Sun Apr 17 2022

The MASSIVE dataset–Multilingual Amazon Slu resource package (SLURP) for Slot-filling, Intent classification, and Virtual assistant Evaluation is presented and modeling results on XLM-R and mT5 are presented, including exact match accuracy, intent classification accuracy, and slot- filling F1 score.

173 0

Paper Graph

AdapterHub: A Framework for Adapting Transformers

Sebastian Ruder, Kyunghyun Cho, Ivan Vulic, Iryna Gurevych, Jonas Pfeiffer, Clifton A. Poth, Aishwarya Kamath, Andreas Rücklé•Tue Jul 14 2020

AdaptersHub is proposed, a framework that allows dynamic “stiching-in” of pre-trained adapters for different tasks and languages that enables scalable and easy access to sharing of task-specific models, particularly in low-resource scenarios.

709 0

Paper Graph

Emotion Classification in a Resource Constrained Language Using Transformer-based Approach

Iqbal H. Sarker, Omar Sharif, M. M. Hoque, Avishek Das•Fri Apr 16 2021

Experimental outcomes indicate that XLM-R outdoes all other techniques by achieving the highest weighted f_1-score of 69.73% on the test data.

44 0

Paper Graph

AltCLIP: Altering the Language Encoder in CLIP for Extended Language Capabilities

Ledell Yu Wu, Zhongzhi Chen, Guangyi Liu, Fulong Ye, Qinghong Yang•Fri Nov 11 2022

very close performances with CLIP on almost all tasks are obtained, suggesting that one can simply alter the text encoder in CLIP for extended capabilities such as multilingual understanding.

108 0

Paper Graph

MAD-X: An Adapter-based Framework for Multi-task Cross-lingual Transfer

Sebastian Ruder, Ivan Vulic, Iryna Gurevych, Jonas Pfeiffer•Wed Apr 29 2020

MAD-X is proposed, an adapter-based framework that enables high portability and parameter-efficient transfer to arbitrary tasks and languages by learning modular language and task representations and introduces a novel invertible adapter architecture and a strong baseline method for adapting a pretrained multilingual model to a new language.

711 0

Paper Graph

BERTweet: A pre-trained language model for English Tweets

A. Nguyen, Dat Quoc Nguyen, Thanh Vu•Tue May 19 2020

BERTweet is presented, the first public large-scale pre-trained language model for English Tweets, trained using the RoBERTa pre-training procedure, producing better performance results than the previous state-of-the-art models on three Tweet NLP tasks.

1026 0

Paper Graph

ARBERT & MARBERT: Deep Bidirectional Transformers for Arabic

AbdelRahim Elmadany, El Moatez Billah Nagoudi, Muhammad Abdul-Mageed•Sat Dec 26 2020

This work introduces two powerful deep bidirectional transformer-based models, ARBERT and MARBERT, for a collection of diverse Arabic varieties, and introduces ARLUE, a new benchmark for multi-dialectal Arabic language understanding evaluation.

576 0

Paper Graph

Adding a benchmark result helps the community track progress.