Tools•formatter
OpenAI Fine-Tuning API for Fine-Tuning & Custom Models
Converts raw text datasets into structured LoRA-compatible training formats with tokenization guidance
Try the tool
client runnerFormatted training data
Run the tool to see output.
Examples
Basic JSONL conversion
{
"raw_text": "Hello world\\nThis is a test",
"format_type": "JSONL"
}Expected output
["Hello world\nThis is a test"]
CSV with tokens
{
"raw_text": "System: Keep it concise\\nUser: Explain AI\\nAssistant: Artificial intelligence...",
"format_type": "CSV",
"add_special_tokens": true
}Expected output
text\n"<s>[INST] Keep it concise\nExplain AI[/INST] Artificial intelligence..."
How it works
Processes raw text inputs through tokenization rules optimized for LoRA training, adding model-specific special tokens when requested. Outputs structured datasets in selected formats with proper formatting for training frameworks.
.