Skip to content

Model Training Overview ​

Model training turns frozen dataset versions into robot models. Each training run is a training job, which produces one or more model versions.

How training jobs, checkpoint models, model versions, and deployments relate

Key Concepts ​

ConceptDescription
Training inputOne or more Exported dataset versions in the same format
Model familyThe model architecture used for training, such as ACT, EasyDP, Pi 0.5, or R0
Training strategyThe platform's unified "long-horizon, quality-first" strategy, which runs the full step budget (currently 80,000 steps)
Parameter snapshotSteps, learning rate, batch size, optimizer, random seed, and dataset versions, locked at submission
Compute tierLight, Standard, or High-performance instance configuration
CheckpointA snapshot of the weights at a given step, which can be registered as a model
Model versionA model registered when training completes or when a checkpoint is exported

Why a Unified Training Strategy ​

NexCore doesn't offer fast, standard, or fine-grained training tiers. Every job trains at the same intensity:

  • Comparable results: Results from different dataset versions and model families are produced under the same conditions and can be compared directly.
  • No misleading conclusions: A short run only proves the pipeline works; it says little about model quality.
  • Controlled pacing: You can export checkpoints at any time to check direction early, without waiting for training to finish.

Supported Model Families ​

Model familyCharacteristicsRecommended data format
ACTImitation learning with action chunking; well suited to precise bimanual manipulationHDF5
EasyDPDiffusion policy; well suited to multimodal tasks with complex action distributionsHDF5 / LeRobot
Pi 0.5Vision-language-action (VLA) model; well suited to language-conditioned tasks or tasks that require generalizationLeRobot
R0Lumos Robotics' general-purpose manipulation modelHDF5

Supported combinations of model family and target device are listed in the platform's training templates.

Training Workflow ​

text
Prepare dataset versions → Create training job → Queue and train → Monitor and export checkpoints → Training complete → Model version

NexCore is evolving rapidly; the latest version prevails.NexCore Docs · Embodied AI Skill Learning Platform by Lumos Robotics