Add n8n workflow training datasets (3 formats)
This dataset contains 4,000+ training examples extracted from 6,837 publicly available n8n workflows from the n8n marketplace. The data is designed for fine-tuning Large Language Models to generate n8n workflow configurations from natural language descriptions.
## Dataset Contents
**Three format variations:**
1. **training_data_alpaca.json** - Alpaca format for Llama/Mistral models
- Format: instruction-input-output triplets
- Use case: Fine-tuning with Unsloth, Axolotl, or similar frameworks
2. **training_data_openai.jsonl** - OpenAI format for GPT models
- Format: messages array with system/user/assistant roles
- Use case: OpenAI fine-tuning API
3. **training_data_simple.json** - Simplified format
- Format: Basic instruction-output pairs
- Use case: Custom training pipelines or quick prototyping
## Data Statistics
- Total examples: 4,000+
- Source workflows: 6,837 from n8n marketplace
- Coverage: Diverse work…
