Fail
A pinned input contradicts the versioned expectation. Follow the cited source and publish a new revision after correcting the source workflow; the doctor makes no repair claim.
DATASET DOCTOR
Validate a LeRobot v3 or RLDS dataset before you fine-tune
Paste a public dataset repo to run deterministic, versioned checks against the LeRobot v3 and RLDS contracts, then download the exact typed findings as readiness-report.json.
DIRECT ANSWER
The doctor checks metadata, frame/action alignment evidence, FPS, action and state dimensions, stream lengths, NaN or missing-value signals, task IDs, v2.1-versus-v3 markers, and a license signal. It pins the repo's immutable Hugging Face revision (commit SHA) and hashes each inspected input. It never claims consent clearance, deployment readiness, legal certification, or completed repairs; a code license is a signal, never an “allowed” verdict.
Readiness report
No report yet
Paste a public repo to inspect format metadata and produce typed pass, fail, unknown, and not-applicable findings.
METHODOLOGY
How the readiness doctor works
The first request resolves owner/name to a Hugging Face commit SHA. The displayed run is repeated against that pinned revision, and every inspected input carries a SHA-256 hash. Unavailable or changed inputs become explicit blockers; the service does not silently substitute a different revision.
Every finding carries a stable rule ID, observed and expected values, source file, severity, next verification, ruleset version, checked date, retrieval hash, confidence, limitations, and review status. The report body contains no current timestamp or random value.
Automated structural facts stay separate from rights, consent, and target-domain judgment. Unknown means the inspected metadata cannot establish a result. Not applicable means the detected format is outside that rule.
INTERPRETATION RULES
How to read the result
Unknown
The metadata-only inspection cannot prove alignment, loadability, rights, consent, or deployment fit. Run the suggested bounded verification against the pinned revision.
Not applicable
An RLDS-specific rule does not grade LeRobot, and a LeRobot-specific rule does not grade RLDS. This status is distinct from pass and unknown.
CALIBRATION SOURCES
References behind the rubric
RLDS dataset format
The episode and step contract used to anchor RLDS structural findings.
LeRobotDataset v3 documentation
The v3 metadata, storage-layout, and v2.1 migration contract used by the versioned rules.
Hugging Face repository revisions
Commit revisions provide the pinned repository input used for repeatable inspection.
Format and next-step references
Read the canonical RLDS format and LeRobot format definitions, compare LeRobot datasets, or follow the RLDS and LeRobot interoperability guide. Then check licensing separately or turn missing evidence into a data specification.
TRUELABEL ROUTING
Convert a blocked repo into a cleared, spec-matched dataset
Use the failed and unknown findings to scope the missing format, provenance, consent, and target-domain evidence before requesting data.
TOOL FOLLOW-UP
Every tool output should route to evidence
A calculator or checker is useful only when it changes the buyer's next step. The output should send the user toward dataset research, rights review, format requirements, budget planning, or a request spec with concrete acceptance criteria.
The links below make that workflow explicit and keep tool pages from becoming isolated utilities — opening paths into deeper catalog, template, briefing, and provider research.
External references are included because tool outputs need calibration against the wider robotics data ecosystem. Buyers should be able to compare truelabel's workflow assumptions with public robotics datasets, developer tooling, and market signals.
Use the tool result as a draft memo, not a final answer. A buyer still needs a source link, a sample packet, a rights note, and a concrete acceptance rule before the output becomes a procurement decision. The links below are the evidence trail for that memo.
Where to go next
- Physical AI data toolsMove between cost estimation, dataset fit, license triage, and request-spec drafting from one workflow surface.
- Dataset catalogGround tool outputs in real dataset profiles before deciding whether public data or custom collection is the next step.
- Request templatesConvert calculator outputs into reusable scopes with capture requirements, QA gates, risk flags, and metadata fields.
- Data briefingsCheck whether licensing, dataset release, or teleoperation news changes the assumptions behind a tool result.
- Robot data formatsTranslate an output into loader, timestamp, manifest, and file-format requirements before sourcing data.
- Physical AI glossaryResolve vocabulary before turning a form result into procurement language a supplier can quote against.
- Physical AI data marketplaceUse truelabel when the result points to a scoped custom collection, dataset supplement, or evaluation package.
- Data annotation companiesCompare where tooling ends and managed labeling, curation, capture, or marketplace sourcing should begin.
Other places to verify the claims
- Scale AI physical AI data engineMarket context for why physical AI systems need custom, enriched, real-world data beyond generic labeling workflows.
- LeRobot documentationRobotics dataset and tooling context for Hugging Face based collection, sharing, conversion, and training workflows.
- Open X-EmbodimentA cross-embodiment robotics dataset reference for comparing trajectory scale, robot diversity, and VLA training assumptions.
- DROID datasetA large in-the-wild robot manipulation dataset reference for real-world trajectory capture and deployment transfer risk.