French 4-grams

该数据集包含从书籍中提取的n-gram（主要是4-gram），这些n-gram与一个轻量级的法语GPT-2模型计算出的概率相关联。为了确保文本生成的质量，这些n-gram经过了基于概率阈值的筛选。在特定条件下，为5-gram生成了5,052个解决方案，为4-gram生成了56,652个解决方案。该数据集的任务是在约束编程中使用n-gram进行文本生成。

This dataset comprises n-grams (primarily 4-grams) extracted from books, which are associated with probabilities calculated by a lightweight French GPT-2 model. To ensure the quality of text generation, these n-grams were filtered using probability thresholds. Under specific conditions, 5,052 solutions were generated for 5-grams and 56,652 for 4-grams. The core task of this dataset is to enable text generation via n-grams in constraint programming.