Dataset Summary:
A new and improved verison of InstructMix that has nearly twice as many examples.
Dataset Contents:
The dataset contains a collection of instructional data with corresponding inputs and outputs. Each entry has an "Input" field that contains the instructional content, and an "Output" field that represents the corresponding response or completion. Here is a list of the datasets used:
Locutusque/ColumnedChatCombined
TokenBender/code_instructions_120k_alpaca_style… See the full description on the dataset page:
https://huggingface.co/datasets/Locutusque/InstructMix-V2.
