l3lab/ntp-mathlib-instruct-st
收藏资源简介:
--- tags: - theorem-proving - math - lean configs: - config_name: default data_files: - split: train path: "state_tactic_mathlib_only/state_tactic_train*" - split: dev path: "state_tactic_mathlib_only/state_tactic_dev*" - split: test path: "state_tactic_mathlib_only/state_tactic_test*" --- ## [miniCTX: Neural Theorem Proving with (Long-)Contexts]() Lean 4 tactic prediction examples extracted from Mathlib. Examples contain: - prompt: - instruction, proof state - completion: tactic ### Version Generated using `ntptoolkit`'s `ntp-training-data` and `instruction_tuning.py`. It used the following config for `ntp-training-data`: ```json { "repo": "https://github.com/leanprover-community/mathlib4", "commit": "cf8e23a62939ed7cc530fbb68e83539730f32f86", "lean": "leanprover/lean4:v4.4.0", "name": "mathlib", "import_file": "Mathlib.lean", "imports": ["Mathlib"] } ``` #### Citation Please cite: ``` @misc{hu2024minictx, title={miniCTX: Neural Theorem Proving with (Long-)Contexts}, author={Jiewen Hu and Thomas Zhu and Sean Welleck}, year={2024}, eprint={2408.03350}, archivePrefix={arXiv}, primaryClass={cs.AI}, url={https://arxiv.org/abs/2408.03350}, } ```
--- 数据集标签: - 定理证明(theorem-proving) - 数学(math) - Lean(lean) 数据集配置: - 配置名称:default 数据文件: - 训练集(split: train):路径为`state_tactic_mathlib_only/state_tactic_train*` - 开发集(split: dev):路径为`state_tactic_mathlib_only/state_tactic_dev*` - 测试集(split: test):路径为`state_tactic_mathlib_only/state_tactic_test*` --- ## [miniCTX:基于(长)上下文的神经定理证明]() 本数据集收录从Mathlib(Mathlib)中提取的Lean 4(Lean 4)证明策略预测样本。 样本包含: - 提示(prompt): - 指令、证明状态 - 补全项(completion):证明策略(tactic) ### 版本说明 本数据集通过`ntptoolkit`的`ntp-training-data`模块与`instruction_tuning.py`脚本生成。 生成时为`ntp-training-data`配置了如下参数: json { "代码仓库": "https://github.com/leanprover-community/mathlib4", "提交哈希值": "cf8e23a62939ed7cc530fbb68e83539730f32f86", "Lean版本": "leanprover/lean4:v4.4.0", "数据集名称": "mathlib", "导入文件": "Mathlib.lean", "导入模块": ["Mathlib"] } #### 引用规范 请引用如下文献: bibtex @misc{hu2024minictx, title={miniCTX:基于(长)上下文的神经定理证明}, author={胡建文、朱托马斯、肖恩·韦莱克}, year={2024}, eprint={2408.03350}, archivePrefix={arXiv}, primaryClass={cs.AI}, url={https://arxiv.org/abs/2408.03350}, }
数据集概述
数据集名称
miniCTX: Neural Theorem Proving with (Long-)Contexts
数据集内容
- 类型: 定理证明
- 语言: Lean 4
- 来源: 从Mathlib提取的tactic预测示例
数据集结构
- 训练集:
state_tactic_mathlib_only/state_tactic_train* - 开发集:
state_tactic_mathlib_only/state_tactic_dev* - 测试集:
state_tactic_mathlib_only/state_tactic_test*
数据集特征
- 示例内容:
- prompt: 指令、证明状态
- completion: tactic
版本信息
- 生成工具:
ntptoolkit中的ntp-training-data和instruction_tuning.py - 配置详情:
- 仓库:
https://github.com/leanprover-community/mathlib4 - 提交:
cf8e23a62939ed7cc530fbb68e83539730f32f86 - Lean版本:
leanprover/lean4:v4.4.0 - 名称:
mathlib - 导入文件:
Mathlib.lean - 导入:
Mathlib
- 仓库:
引用信息
- 作者: Jiewen Hu, Thomas Zhu, Sean Welleck
- 标题: miniCTX: Neural Theorem Proving with (Long-)Contexts
- 年份: 2024
- 预印本: arXiv




