Fine-tuning dataset for creating an argumentative agent, following Schopenhauer's stratagems.
Credits (GitHub): @basileplus, @vdeva, @mcosson , @yanisgomes, @raphaaal
This dataset contains 1,000 conversations, generated synthetically using Mistral-Large.
Each conversation starts with a claim from an Opponent and contains between 1 and 5 tweets debating this claim.
Topic: the Silicon Valley Bank run debate on Twitter.
Input: the beggining of a conversation between two Users (Opponent and You)… See the full description on the dataset page:
https://huggingface.co/datasets/raphaaal/schopenhauer-debate.