OmniEvaluator Dashboard

Internal

Submission

0 models

Models are auto-detected from configured sources. Choose a source to see its configuration and detected models.

Internal Path

Set OMNI_INTERNAL_OUTPUTS_PATH to change.

Loading...

Internal Path

Leaderboard

Filters
Model Mode
Filter models by debug prefix.
Sources
Filter rows by data source.
Modality
Filter benchmarks by input modality.
Inference Engine
Filter rows by inference backend.
Evaluation Engine
Filter benchmark columns only.
Value
Choose score/time/coverage and sub-metrics.
Options
Avg, OmniScore, normalization, and merge toggles.
Benchmarks

Inference Viewer

Select models by source, then choose a benchmark to view sample-level predictions.

Model Selection

Internal Path

Select one or more models above to begin.

Select Benchmark

Benchmarks
Selected Select benchmark

Inference Viewer