This is the dataset behind LLMango: a blog dedicated to showcasing different behaviours of Large Language Models through prompting them in great volumes and data visualization. LLMango is divided into experiments, where each one is different and tries to answer specific questions you might have about LLMs.
One row is one call. It holds the prompt that produced it, the provider's verbatim response, and the usage, cost and timing the provider reported.
Charts… See the full description on the dataset page:
https://huggingface.co/datasets/rafalkwiecien/llmango.
