Dataset Viewer
The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.
llm101-v3.1-data
配套 42ailab/OLMo3-190M-zh-v3.1 模型的训练数据包。
目录
tokenizer/— 48k 中文 BPE（5 MB）tokenized/full_v31.bin— v3.1 完整训练用 tokenized.bin（3.3 GB, uint16, 1.65B tokens）tokenized/nano_0.5B.bin— nano 子集（1 GB，供学员 T4 快速训练）tokenized/meta.json— 文件元数据eval/prompts.yaml— 7-prompt 评测 promptseval/v3_samples.jsonl— v3 基线生成对照
按需下载
from huggingface_hub import snapshot_download
# 只要 tokenizer + nano（学员大多数用法）
snapshot_download("42ailab/llm101-v3.1-data",
repo_type="dataset",
allow_patterns=["tokenizer/*", "tokenized/nano_*", "tokenized/meta.json"])
License
- 代码和 metadata: Apache-2.0
- tokenized 内容来自 Ultra-FineWeb / Cosmopedia-Chinese (Apache-2.0) 和 Wikipedia-zh (CC-BY-SA-3.0)
- Downloads last month
- 43