ash0t/

finance-slm-distillation-data

	Finance Instruction SFT Dataset




	
	Overview

This repository provides a curated finance-oriented instruction dataset designed for conversational supervised fine-tuning (SFT) of large language models. The dataset is derived from: heladell/Finance_DeepSeek-R1-Distill-dataset and was further processed through a custom preprocessing pipeline focused on:

finance-domain filtering reasoning-quality filtering answer-quality filtering conversational formatting train/test split… See the full description on the dataset page: https://huggingface.co/datasets/ash0t/finance-slm-distillation-data.