Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Few-shot Learning with Multilingual Language Models

About

Large-scale generative language models such as GPT-3 are competitive few-shot learners. While these models are known to be able to jointly represent many different languages, their training data is dominated by English, potentially limiting their cross-lingual generalization. In this work, we train multilingual generative language models on a corpus covering a diverse set of languages, and study their few- and zero-shot learning capabilities in a wide range of tasks. Our largest model with 7.5 billion parameters sets new state of the art in few-shot learning in more than 20 representative languages, outperforming GPT-3 of comparable size in multilingual commonsense reasoning (with +7.4% absolute accuracy improvement in 0-shot settings and +9.4% in 4-shot settings) and natural language inference (+5.4% in each of 0-shot and 4-shot settings). On the FLORES-101 machine translation benchmark, our model outperforms GPT-3 on 171 out of 182 directions with 32 training examples, while surpassing the official supervised baseline in 45 directions. We conduct an in-depth analysis of different multilingual prompting approaches, showing in particular that strong few-shot learning performance across languages can be achieved via cross-lingual transfer through both templates and demonstration examples. Finally, we evaluate our models in social value tasks such as hate speech detection in five languages and find it has limitations similar to comparable sized GPT-3 models.

Xi Victoria Lin, Todor Mihaylov, Mikel Artetxe, Tianlu Wang, Shuohui Chen, Daniel Simig, Myle Ott, Naman Goyal, Shruti Bhosale, Jingfei Du, Ramakanth Pasunuru, Sam Shleifer, Punit Singh Koura, Vishrav Chaudhary, Brian O'Horo, Jeff Wang, Luke Zettlemoyer, Zornitsa Kozareva, Mona Diab, Veselin Stoyanov, Xian Li• 2021

Related benchmarks

TaskDatasetResultRank
Natural Language InferenceXNLI (test)--
167
Natural Language InferenceXNLI
Accuracy44.7
111
Paraphrase IdentificationPAWS-X
Accuracy50.3
57
Hate Speech DetectionHate Speech
Accuracy51.8
49
Commonsense ReasoningXStoryCloze--
32
Commonsense ReasoningXCOPA
Accuracy62.5
24
Winograd Schema ChallengeXWINO
Accuracy0.642
14
Machine TranslationWMT en-fr 14
BLEU33.2
14
Machine TranslationWMT en-de 16 (test)
BLEU23.5
13
Hate Speech DetectionMultilingual Hate Speech (test)
Performance (en)61.7
8
Showing 10 of 23 rows

Other info

Follow for update