Member Junction
    Preparing search index...

    Retrains from Training Snapshots: Retrain re-runs a completed run through the protocol trainer from its verified snapshot files, with its reads split into protected steps a subclass can replace.

    Index

    Constructors

    Properties

    DefaultTolerance: 1e-9

    The default largest holdout-metric difference a retraining may show and still reproduce its run.

    Methods

    • Re-runs a completed unit run from its snapshot files and logged trainer request. The run and snapshot are read through deps.snapshots and the files through deps.snapshotFiles (defaults: the RunView and storage-provider stores on deps.provider, as deps.contextUser).

      Parameters

      • trainingRunId: string

        the MJ: ML Training Runs row to retrain

      • deps: TrainingDeps

        the training seams; only the snapshot stores, the trainer, the provider and the user are used

      • tolerance: number = SnapshotRetrainer.DefaultTolerance

        the largest holdout-metric difference that still counts as reproduced

      Returns Promise<RetrainResult>

      both metric sets, their largest difference and whether the run reproduced

      Error when the run logged no trainer request (only runs trained through /trainer are retrained)

      SnapshotIntegrityError when a stored file or the content hash does not match the snapshot row

    • The largest |original − retrained| over every metric name on either side. A name present on one side only makes the difference Infinity (the retraining did not measure what the run measured), and so does a non-finite value.

      Parameters

      • original: Readonly<Record<string, number>>
      • retrained: Readonly<Record<string, number>>

      Returns MetricDifference