OpenMol/Yield_Regression
收藏资源简介:
该数据集结合了Buchwald-Hartwig和Suzuki-Miyaura两个主要的产率预测数据集。数据集包含9515条训练数据和200条测试数据。每条数据记录包含系统、用户和助手的交互信息,涉及化学反应方程式的预测和产率的预测。
该数据集结合了Buchwald-Hartwig和Suzuki-Miyaura两个主要的产率预测数据集。数据集包含9515条训练数据和200条测试数据。每条数据记录包含系统、用户和助手的交互信息,涉及化学反应方程式的预测和产率的预测。
数据来源
结合Buchwald-Hartwig和Suzuki-Miyaura两种主要产率预测数据集。
数据统计
- 训练集:9515条
- 测试集:200条
示例
json { "messages": [ { "content": "You are a chemist. Now you are given a reaction equation. Please predict the possible reagents of the reaction. The reaction equation has the following format:
reactant1.reactant2. ... .reactantN>>product
The return value should be in range of 0-1. The higher the value, the more likely the reaction is to occur. We provide the structure of the reactions.", "role": "system" }, { "content": "<molecule_2d>.<molecule_2d>.<molecule_2d>.<molecule_2d>.<molecule_2d>.<molecule_2d>>><molecule_2d> Using the chemical reaction information, what is the ratio of the reactions yield?", "role": "user" }, { "content": "0.0368", "role": "assistant" } ] }




