Dataset Card for StorySummariserAI

This is a human written training set of prompts and responses to finetune a model to be a better summariser for single person rpg sessions.

Dataset Details

Dataset Description

This is a human-curated training dataset designed to fine-tune language models for tabletop role-playing game (RPG) single player scenarios. The dataset contains carefully crafted prompt-response pairs that demonstrate how an AI should respond to summarise older parts of the story in single-player role-playing scenarios.

Dataset Summary

• Curated by: Björn 'Idrinth' Büttner
• Language(s) (NLP): English
• License: MIT

Dataset Sources

Repository: https://github.com/bjoern-buettner/roleplay-ai

Data Structure

The dataset contains training examples with the following structure:

• messages: List of conversation turns with content (the message text) and role (user/assistant)
• text: Formatted text representation of the conversation

Each example demonstrates proper game master responses to player actions in role-playing scenarios.

Intended Uses

Primary Use Case

Fine-tuning language models to serve AI game masters for:

• Single-player tabletop RPG sessions
• Interactive storytelling applications
• Role-playing game assistance tools

Example Applications

• Digital D&D campaigns
• Solo adventure gaming
• Creative writing prompts
• Interactive fiction development

Usage Example

from datasets import load_dataset

Load the dataset

dataset = load_dataset("Idrinth/storysummariserai")

Access training examples

train_data = dataset["train"]
print(f"Number of examples: {len(train_data)}")

View a sample conversation

sample = train_data[0]
print("Sample conversation:")
for message in sample["messages"]:
    print(f"{message['role']}: {message['content']}")

Limitations and Considerations

• Small dataset size: Currently only 5 training examples, limiting models's exposure to diverse scenarios
• Domain-specific: Focused specifically on RPG/fantasy scenarios
• English only: No multilingual support
• Early stage: Dataset is actively being expanded with more training scenarios

Contributing

The dataset is actively being expanded. Contributors can help by:

• Adding new RPG scenario examples
• Improving existing conversation quality
• Suggesting additional use cases

For contributions, visit the GitHub Repository or join the Discord community.

Citation

If you use this dataset in your research or applications, please cite

@dataset{storysummariserai,
  title={Story Summariser AI Training Dataset},
  author={Björn Büttner},
  year={2025},
  url={https://huggingface.co/datasets/Idrinth/storysummariserai}
}