ShareGPT4 is a cleaned version of the OpenChat-ShareGPT4 Dataset, designed for training conversational AI models. This dataset contains a collection of conversations, with each conversation consisting of two main features: role and value.
Features:
conversations:
role (string): The role of the speaker in the conversation.
value (string): The actual conversation text.
Splits:
train:
Number of examples: 6144
Size: 30,322,763 bytes… See the full description on the dataset page:
https://huggingface.co/datasets/erfanzar/ShareGPT4.
