---
dataset_info:
  features:
  - name: instruction
    dtype: string
  - name: input
    dtype: string
  - name: output
    dtype: string
  splits:
  - name: train
    num_bytes: 87307
    num_examples: 32
  - name: validation
    num_bytes: 40154
    num_examples: 17
  download_size: 55998
  dataset_size: 127461
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
language:
- en
license: other
task_categories:
- text-generation
size_categories:
- n<1K
tags:
- finance
- instruction-tuning
- llm
- pipeline-test
pretty_name: Hedgefund Memos v2
---

# Hedgefund Memos v2

## Dataset Summary
Hedgefund Memos v2 is a small instruction-tuning dataset for generating hedge-fund-style stock memos from structured company facts.

This dataset is intended primarily for **pipeline testing and experimentation**:
- data schema validation
- smoke-test fine-tuning runs
- output formatting checks

It is **not** intended as a production benchmark or financial decision system.

## Dataset Structure

### Splits
- `train` (32 examples)
- `validation` (17 examples)

### Features
Each row contains:
- `instruction` (`string`)
- `input` (`string`)
- `output` (`string`)

## Data Construction
Examples are produced via a local pipeline that transforms structured financial/context inputs into instruction-tuning pairs.

Underlying inputs include:
- SEC company facts trend summaries
- Alpha Vantage valuation fields
- basic company context fields

## Intended Use
Use this dataset for:
- end-to-end workflow validation
- quick fine-tuning smoke tests
- debugging prompt/response formatting behavior

## Limitations
- very small dataset size
- limited ticker/scenario coverage
- template-driven outputs reduce linguistic diversity
- unsuitable for robust model benchmarking

## Risks and Safety
This dataset and models trained on it must not be treated as financial advice.
Outputs may be incomplete, stale, or incorrect and should be verified against trusted, current sources.

## Version
- Current version: `v2`
- Purpose: pipeline validation and experimentation