Skip to content

Dataset · TheFinAI

Reasoning chains for financial questions (FinQA)

TheFinAI/Fino1_Reasoning_Path_FinQA

FinQA questions with step-by-step GPT-4o reasoning paths for training financial reasoning models.

What's inside

The dataset includes 1k to 10k English text examples used to train the Fino1-8B model. It is based on financial questions from the FinQA dataset, augmented with chain-of-thought reasoning paths generated using GPT-4o. The dataset accompanies a research paper on transferring reasoning and RL skills to the financial domain.

What you can do with it

  • Supervised fine-tuning language models for step-by-step financial and math problem solving
  • Creating a specialized LoRA for financial statement analysis with chain-of-thought explanation
Tasks
Question answering, Text generation
Languages
English
Size
1K–10K rows
Downloads this month
1K
Likes
42
Access
on request on Hugging Face

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("TheFinAI/Fino1_Reasoning_Path_FinQA")