mzio/strl-rollouts-Eenv-Ggen-S0-h50e9202c
收藏官方服务：
资源简介：
该数据集是一个用于强化学习或对话系统训练的数据集，包含状态消息、动作消息、观察消息等特征，以及奖励、完成标志、时间步等元数据。数据集支持工具调用和任务提示，适用于训练AI代理进行交互式任务。数据分为训练集，包含201,059个示例，总大小约2.5GB。
This dataset is designed for reinforcement learning or dialogue system training, featuring state messages, action messages, observation messages, and metadata such as rewards, done flags, and timesteps. It supports tool usage and task prompts, suitable for training AI agents in interactive tasks. The dataset includes a training split with 201,059 examples and a total size of approximately 2.5GB.
提供机构：
mzio
