MameLoshnLM：首个意第绪语开源语言模型及评估基准
原标题：MameLoshnLM: Yiddish Language Model and Evaluation Benchmark
AI 摘要
Hugging Face 每日论文介绍了 MameLoshnLM，这是首个专为意第绪语构建的开源 8B 参数语言模型。研究团队创建了高质量预训练语料库 Oytser 和多任务基准 Kashes，并基于 Llama 3.1 8B 继续预训练得到该模型。在基准测试中，MameLoshnLM 优于同规模开源模型，并更好地保留了意第绪语的词汇和形态特征，为低资源语言模型开发提供了模板。
正文节选
MameLoshnLM: Yiddish Language Model and Evaluation Benchmark Abstract We present MameLoshnLM, the first open-source 8B-parameter language model built specifically for Yiddish. Despite Yiddish's rich textual tradition, its limited digital presence and the scarcity of reliable evaluation resources have constrained progress in Yiddish language modeling. Existing multilingual corpora and benchmarks are often poor proxies for the language, containing substantial amounts of noisy, machine-translated,