openai-format
openai-formate-function-calling-small
数据集内容说明:
包含700+个阿里云OpenAPI的信息;包括Dataworks,EMR，DataLake，Maxcompute，Hologram,实时计算Flink版，QuickBI,DTS等多个产品的公开Open API信息。
Functions信息与OpenAI functions calling 能力中，functions信息传入的格式保持一致
样例
{
"systemPrompt": 你是一个函数筛选助理，如果与问题相关的话,您可以使用下面的函数来获取更多数据以回答用户提出的问题:{""name"": ""UpdateTicketNum"", ""description"": ""对用于免登嵌入报表的指定的ticket进行更新票据数量操作。"", ""parameters"": {""type"": ""object"", ""properties"": [{""Ticket"": {""type"": ""string"", ""description"":… See the full description on the dataset page: https://huggingface.co/datasets/Deepexi/openai-formate-function-calling-small.OpenHermes-2.5-Formatted-OpenAI-Compatibleconversations_in_OpenAI_format
Dataset Card for Dataset Name
We convert some conversation datasets to the standard OpenAI format, to facilitate the standardized finetuning such as using LlamaFactory.
We add some new keys in the json for annotations such as strategies and emotions, to allow for some specific methods on affective computing.
Dataset Details
For ESConv, we omit some redundant attributes such as emotion/strategy intensities.
To keep a strict user-assistant interleaved… See the full description on the dataset page: https://huggingface.co/datasets/jiluoaaron/conversations_in_OpenAI_format.hermes-3-openai-formatopenai-too-calling-5k-formattedAReaL-boba-2-RL-Code-openai-format