datikz-v3

# Dataset Card for DaTi*k*Z_v3
DaTi*k*Z_v3 is a dataset of Ti*k*Z drawings aligned with captions. In compliance with licensing agreements, certain Ti*k*Z drawings are excluded from this public version of the dataset. Check out the [DeTi*k*Zify](https://github.com/potamides/DeTikZify) project and the [DaTi*k*Z](https://github.com/potamides/DeTikZify) repository for more information as well as tools and methods to recreate the complete dataset from scratch.

## Usage
```python
from datasets import load_dataset
# full dataset
ds = load_dataset("nllg/datikz-v3")
# only the train split
ds = load_dataset("nllg/datikz-v3", split="train")
for sample in ds:
print(sample["code"])
```

# DaTi*k*Z_v3 数据集卡片
DaTi*k*Z_v3 是一组与说明文字对齐的Ti*k*Z绘图数据集。根据许可协议要求，本公开版本的数据集未包含部分Ti*k*Z绘图。可访问 [DeTi*k*Zify](https://github.com/potamides/DeTikZify) 项目与 [DaTi*k*Z](https://github.com/potamides/DeTikZify) 代码仓库以获取更多信息，以及从零重建完整数据集所需的工具与方法。

## 使用方法
python
from datasets import load_dataset
# 完整数据集
ds = load_dataset("nllg/datikz-v3")
# 仅训练划分集
ds = load_dataset("nllg/datikz-v3", split="train")
for sample in ds:
print(sample["code"])

搜集汇总

背景与挑战

背景概述

DaTi _k_ Zv3是一个TikZ绘图与标题对齐的数据集，根据许可协议，公开版本中排除了部分绘图。该数据集由nllg维护，采用Apache License 2.0许可，更新于2025年4月22日。