---
dataset_info:
  features:
  - name: conversations
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  splits:
  - name: train
    num_bytes: 536575
    num_examples: 1024
  download_size: 133315
  dataset_size: 536575
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
license: apache-2.0
task_categories:
- table-question-answering
- text-generation
language:
- de
size_categories:
- n<1K
---

Dataset Name: Deutsche Bahn FAQ in Llama 3 Format

Dataset Description:

This dataset contains 1000 question-answer pairs extracted from the official Deutsche Bahn (German Railways) FAQ section. The data has been specifically formatted to be compatible with the Llama 3 instruct models for supervised fine-tuning (SFT).

Dataset Purpose:

The primary purpose of this dataset is to facilitate the fine-tuning of Llama 3 instruct models for tasks related to customer service and information retrieval in the domain of German railway services.

Dataset Structure:

Each sample in the dataset follows the Llama 3 instruct format, which includes:

System message: Provides context to the model ("You are a helpful AI assistant for Deutsche Bahn customer service.") User message: Represents a question from a customer. Assistant message: Represents the ideal answer from the AI assistant. Example:

<|begin_of_text|> <|start_header_id|>system<|end_header_id|> You are a helpful AI assistant for Deutsche Bahn customer service.<|eot_id|> <|start_header_id|>user<|end_header_id|> Wie kann ich eine BahnCard beantragen?<|eot_id|> <|start_header_id|>assistant<|end_header_id|> Sie können die BahnCard online, am Bahnhof oder in einem DB Reisezentrum beantragen.<|eot_id|> <|end_of_text|> Dataset Creation:

The data was originally scraped from the Deutsche Bahn website (https://www.bahn.de/faq/pk) and then manually cleaned and filtered to ensure quality and relevance. The Llama 3 format was applied using a custom Python script.

Potential Use Cases:

Fine-tuning Llama 3 instruct models for customer service chatbots or question-answering systems in the German railway domain. Educational purposes to demonstrate the process of SFT for LLMs.

Additional Notes: This dataset is a subset of the full Deutsche Bahn FAQ and is intended as a starting point for fine-tuning. The quality of the fine-tuned model will depend on the quality and diversity of the training data. This dataset was created as part of a master's thesis research project.