LocalDoc/aznli-benchmark

AzNLI Benchmark数据集是阿塞拜疆语的自然语言推理（NLI）基准数据集，包含多对句子及其逻辑关系标签。每对句子由前提句和假设句组成，并标注了它们之间的逻辑关系，包括蕴含（标签0）、中立（标签1）和矛盾（标签2）。该数据集可用于阿塞拜疆语的NLI模型训练与评估、跨语言迁移学习实验、低资源语言的自然语言理解研究以及多语言模型在阿塞拜疆语NLI任务上的基准测试。

The AzNLI Benchmark dataset is a Natural Language Inference (NLI) benchmark dataset in Azerbaijani. It consists of multiple sentence pairs with their corresponding logical relationship labels. Each pair includes a premise sentence and a hypothesis sentence, labeled with their logical relationship, including Entailment (label 0), Neutral (label 1), and Contradiction (label 2). This dataset can be used for training and evaluating NLI models for Azerbaijani, cross-lingual transfer learning experiments, studying natural language understanding in low-resource languages, and benchmarking multilingual models on Azerbaijani NLI tasks.