---
datasets:
- SuperbEmphasis/Deepseek-R1-ERP-REASONING-Dataset
base_model:
- CreitinGameplays/Mistral-Nemo-12B-R1-v0.2
---

hah, this is pretty good... but it is using that deepseek long winded reasoning.

going to try a modified dataset with some sort of "concise" reasoning.