Describe the task. We build a clean, answer-verified dataset and deliver it HuggingFace-ready — drop the repo straight into Gradients, TRL, Axolotl, or Unsloth. Every checkable answer is verified in code, not trusted from a model.
Most synthetic datasets trust the language model to be right. This one doesn't. The correct answer is computed independently before the model writes a word — then the model's answer is checked against it.
Each problem is built in code with a known-correct answer as ground truth.
The model writes step-by-step reasoning and a final answer for the problem.
The model's answer is checked against ground truth in code. Mismatches are discarded.
Deduplicated, split train/val/test, documented, pushed to a HuggingFace repo.
Describe your dataset and we'll price it instantly for standard work, or quote it within 24h for specialized domains. Every order gets a scope review — if we can't build it as described, you get a full refund.
A Gradients run costs $100–500 and you still have to bring the data. We supply the verified dataset for a flat per-build price. No setup fee, no per-record meter.
After payment, every order gets a scope review within 24 hours. If your request needs domain expertise, sources, or verification depth beyond the quoted price, we propose an adjusted scope first — reduce rows, simplify verification, or refocus the task. If you accept, we build immediately.
If no acceptable adjustment exists, you get a full refund within 2 business days — no partial charges, no quibbling. Revisions for errors in the delivered dataset (wrong format, bad split, missing rows) are always included. Changes to the task definition itself are a new order.
To keep every promise we make, we don't fulfill requests requiring: HIPAA-regulated patient data, classified or restricted-source material, real-time data streams, or expert credentials we don't hold (e.g. board-certified clinical sign-off). These are flagged at scope review and automatically refunded.
HSH·Intelligence