Benchmark suite of natural language understanding tasks across major Indian languages, used to evaluate Indic language models.
IndicGLUE and its successor IndicXTREME are benchmark suites developed by AI4Bharat to evaluate natural language understanding capabilities across major Indian languages. The benchmarks combine multiple tasks such as classification, natural language inference, named entity recognition, and question answering, adapted or newly created for Indic scripts and linguistic structures. They are modeled after the GLUE and XTREME benchmarks widely used for English and multilingual NLP evaluation.
Indicglue / Indicxtreme Is Used To Systematically Evaluate And Compare The Performance Of Language Models On Indian Language Understanding Tasks. It Provides A Standardized Way To Benchmark Progress In Indic Nlp, Helping Researchers Identify Gaps In Model Performance Across Languages And Tasks. The Benchmark Suite Is Widely Adopted When Releasing New Indic Language Models To Demonstrate Downstream Task Capability.
CC0 1.0 Public Domain
© 2026 - Copyright AIKosh. All rights reserved.