# Fatih Çoban YouTube Script Dataset

This dataset contains 270 training pairs for fine-tuning a language model to generate YouTube scripts in Fatih Çoban's style.

Each example consists of:
- Input: Analysis of successful YouTube videos, including strategic analysis and intro techniques
- Output: A complete script for a YouTube video

## Usage

```python
from datasets import load_dataset

# Load from Hugging Face
dataset = load_dataset("your-username/fatih-script-dataset")  # Replace with your username

# Or load locally
dataset = load_dataset("json", data_files="train.jsonl", split="train")

# Inspect a sample
print(dataset[0])
```