andrewmonostate/synthetic-1756066683459

Fine-tuning Dataset for Style Transfer

This dataset was generated for fine-tuning language models on style transfer tasks.

Dataset Details

• Session ID: session_ce748190
• Repository: andrewmonostate/synthetic-1756066683459
• Number of Examples: 180
• Format: JSONL (JSON Lines)
• Generated: 2025-08-24T20:23:51.706100

Dataset Structure

Each example contains:

• task: The instruction for the model
• input: The source text to be transformed
• expected_output: The target text after transformation
• metadata: Additional information (source_style, target_style, sentiment, etc.)

Sample Examples

[
  {
    "task": "Convert casual writing to formal style while maintaining positive sentiment",
    "input": "This pizza place is super awesome! The food's always fresh and the staff is so friendly.",
    "expected_output": "This establishment offers exceptional cuisine. The meals are consistently fresh, and the personnel demonstrate remarkable hospitality.",
    "metadata": {
      "source_style": "general",
      "target_style": "improved",
      "sentiment": "positive",
      "negation": false
    }
  },
  {
    "task": "Transform general description into technical language while preserving positive sentiment",
    "input": "The new solar panels work great and save us lots of money on electricity bills.",
    "expected_output": "The newly installed photovoltaic system demonstrates optimal efficiency, resulting in significant reduction of electrical utility expenses.",
    "metadata": {
      "source_style": "general",
      "target_style": "improved",
      "sentiment": "positive",
      "negation": false
    }
  }
]

Usage

This dataset is designed for fine-tuning language models to perform style transfer tasks.

Loading the Dataset

import json

# Load from JSONL file
with open('data/finetune_dataset.jsonl', 'r') as f:
    examples = [json.loads(line) for line in f]

# Or load from JSON file
with open('data/finetune_dataset.json', 'r') as f:
    examples = json.load(f)

Fine-tuning Example

# Format for training
training_data = []
for example in examples:
    training_data.append({
        "instruction": example["task"],
        "input": example["input"],
        "output": example["expected_output"]
    })

Generated by Dataset Director

This dataset was created using the Dataset Director API with AI-powered generation.