nemotron-gym-agentic-function-calling-pivot
收藏资源简介:
该数据集名为nemo-nemotron-gym-agentic-function-calling-pivot,是一个Harbor任务二进制数据集,包含9,579个任务。它从nvidia/Nemotron-RL-Agentic-Function-Calling-Pivot-v1转换而来,属于nvidia/Nemotron-Post-Training-v3集合的一部分。数据集采用Apache-2.0许可证。每个数据样本代表一个有效的Harbor任务二进制,包含两列:path(字符串类型,表示路径)和task_binary(gzip tar格式,表示任务二进制内容)。转换过程使用了OpenThoughts-Agent框架中的data.nemotron_gym模块。该数据集适用于文本生成任务,特别是与智能体(agent)、Harbor平台、强化学习和Nemotron模型相关的场景。数据集的评估方法为单步评估,包括工具调用匹配(基于function_call)和大型语言模型(LLM)判断(基于message)。
The dataset is named nemotron-gym-agentic-function-calling-pivot and is a Harbor task binary dataset containing 9,579 tasks. It is converted from nvidia/Nemotron-RL-Agentic-Function-Calling-Pivot-v1 and is part of the nvidia/Nemotron-Post-Training-v3 collection. The dataset uses the Apache-2.0 license. Each data sample represents a valid Harbor task binary and includes two columns: path (string type, indicating the path) and task_binary (gzip tar format, representing the task binary content). The conversion process utilizes the data.nemotron_gym module from the OpenThoughts-Agent framework. This dataset is suitable for text generation tasks, particularly in scenarios related to agents, the Harbor platform, reinforcement learning, and Nemotron models. The evaluation method for the dataset is single-step assessment, which includes tool call matching (based on function_call) and large language model (LLM) judgment (based on message).
- 数据集名称: laion/nemotron-gym-agentic-function-calling-pivot
- 许可证: Apache-2.0
- 任务类别: 文本生成
- 标签: agent、harbor、reinforcement-learning、nemotron
- 数据集描述: 该数据集源自 nvidia/Nemotron-RL-Agentic-Function-Calling-Pivot-v1,属于 nvidia/Nemotron-Post-Training-v3 集合的一部分。它是一个 Harbor 任务二元数据集,包含 9,579 个任务。
- 数据格式: 每一行代表一个有效的 Harbor 任务二元数据,包含两列:
path(字符串)task_binary(gzip tar 格式)
- 数据转换: 使用 OpenThoughts-Agent 的
data.nemotron_gym框架进行转换。 - 评分方式: 单步评分,通过工具调用匹配(function_call)或大语言模型评判(message)进行。




