Independent project. Not a U.S. government website.

Support our project (opens Buy Me a Coffee)
USASI
Generated orange rounded-tile grid artwork for the dataset jingyaogong/minimind_dataset, labeled USASI · AI on Hugging Face, with Task: Text generation · License: Apache 2.0.
DatasetRank #646 of 836 120 likes

minimind_dataset

by jingyaogong (personal account)

分词器可以粗略理解成 LLM 使用的一本“词典”,负责把自然语言映射成 token id,再把 token id 解码回文本;项目中也提供了train_tokenizer.py作为词表训练示例。不建议重新训练 tokenizer,因为词表和切分规则一旦变化,模型权重、数据格式、推理接口与社区生态的兼容性都会下降,也会削弱模型的传播性。同时,tokenizer 还会影响 PPL 这类按 token 统计的指标,因此跨 tokenizer 比较时,BPB(Bits Per Byte)往往更有参考价值,可参考这篇。 对 MiniMind… (excerpt from the dataset card on Hugging Face)

Dataset facts

Hub id
jingyaogong/minimind_dataset
Likes
120
Downloads (last 30 days)
3,902
Tasks
Text generation
Modality
Not given
Size
Not given
License
Apache 2.0 (apache-2.0)
Access
Open download
Author
jingyaogong · Personal account
Languages
multilingual
Created
Sep 9, 2024
Last modified
Apr 9, 2026

Rank #646 of 836 datasets in AI on Hugging Face, by likes. Data from the Hugging Face Hub API, fetched Oct 18, 2026. Likes measure popularity, not quality or safety. Downloads are for the last 30 days, as the Hub reports them; counts change daily. A listing is not an endorsement. How this list is made

Support Us

Help keep USASI useful.

Optional. No USASI account required. Payment takes place on the linked provider’s website (Buy Me a Coffee).

About supporting this project