Concept information
Terme préférentiel
low-resource setting
Définition
- A setting where available resources such as annotated data and algorithms are inadequate for particular analysis tasks in NLP. (Based on Dima and Massey, Keyphrase Extraction for Technical Language Processing, in the Journal of Research of National Institute of Standards and Technology, 2021)
Concept générique
Exemple
- Our observations summarized in Table 10 indicate that the model soup approach yields mixed results over vanilla finetuning in low-resource settings. (Somayajula, Liang, Zhang, Singh & Xie, 2024)
- We extensively evaluate our method on several datasets of the GLUE benchmark in low-resource settings demonstrating improvements over several baselines. (Somayajula, Liang, Zhang, Singh & Xie, 2024)
Traductions
-
français
URI
http://data.loterre.fr/ark:/67375/8LP-WBJ0G6H3-0
{{label}}
{{#each values }} {{! loop through ConceptPropertyValue objects }}
{{#if prefLabel }}
{{/if}}
{{/each}}
{{#if notation }}{{ notation }} {{/if}}{{ prefLabel }}
{{#ifDifferentLabelLang lang }} ({{ lang }}){{/ifDifferentLabelLang}}
{{#if vocabName }}
{{ vocabName }}
{{/if}}