mindchain/distilabel-generator

Distilabel Generator API

Phase 1 of HF Spaces synthetic data pipeline. Generates synthetic instruction-response data with LLM Judge scoring using HF Inference API.

Models

• —Generator: Qwen2.5-72B via Cerebras
• —Judge: Qwen2.5-7B via Together (UltraFeedback-style)

API Endpoints

• —POST /generate - Generate synthetic data with judge scoring
• —POST /load-seeds - Load seed dataset from HF Hub
• —GET /health - Health check
• —GET /models - List available models