Optimized_Reasoning was created because even modern LLM's are not very good at handling reasoning very well, and if they are, they still waste tons of tokens in the process. With this dataset I hope to accomplish 2 things:
So how does this dataset accomplish that? By Adding a "system_prompt" like reasoning tag to the beggining of every data… See the full description on the dataset page:
https://huggingface.co/datasets/Rombo-Org/Optimized_Reasoning.
