Local state
Results, trial manifests, and suggestion logs are saved as files you can inspect and carry between runs.
Optimization workflow
Looptimum connects a guided parameter search to your existing evaluation process. Each completed trial informs the next suggestion, with search state and results recorded locally.
Suggest, evaluate, and ingest workflow
- 01
Suggest
Looptimum selects the next candidate from your parameter bounds, using the results collected so far.
- 02
Evaluate
Your script, cluster, solver, or lab workflow runs the trial in its existing environment.
- 03
Ingest
Record the score or failure in local files, then repeat. The saved history lets you review decisions and resume interrupted work.
Integration and records
The controller keeps its state and decision history in local files. You can inspect the record, resume after interruptions, and work in restricted environments.
Read the codeEvaluator interface
The evaluator accepts parameters within defined bounds and returns a single score or an explicit failure.
Failed runs
Failures and timeouts are recorded with their reasons, so they remain part of the campaign history.
Decision trace
Suggestion logs preserve the context for each choice and the observations available at the time.
Constraints, scores, and failures
Define which candidates are allowed, how successful runs are scored, and how failures are reported.
Hard constraints
Parameter bounds and feasibility rules filter candidates before they are suggested.
Objective score
A single score ranks valid candidates. If several measurements matter, define how they combine into that score.
Failure outcomes
A failed or timed-out evaluation is recorded as a completed failure with a reason, separately from successful measurements.
Case studies
Review the methods, recorded results, and validation checks from simulation and model-training campaigns.