Unverified50% confidenceFactExact time
GitHub user jjlalli compiled 136 entries of Tunisian dialect NLP resources, covering five categories: text corpora, speech data, pre-trained models, evaluation benchmarks, and dictionaries
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/9/2026
First Seen
Valid until: 11/7/2026
Sources
突尼斯方言NLP资源盘点：136个数据集的可用性真相
Related Claims
Unverified阿里建立了覆盖超过30种语言的高质量预训练语料库，总量据估计超过18万亿token66% similarUnverifiedDify's GitHub Readme has been translated into 17 languages, covering Chinese, English, Japanese, Korean, German, French, Arabic, and others.62% similarUnverified现代多语言模型如mT5、BLOOM和Llama系列在数百种语言的文本上联合训练，通过共享词汇表和跨语言注意力机制构建跨语言语义地图60% similarUnverifiedThe vocabulary of a standard autoregressive language model typically contains 32,000 to 128,000 tokens.60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/715010API
curl https://kongchang.com/api/v1/knowledge/claims/715010MCP
get_claim(id=715010)