OmniOpt Benchmark

What is OmniOpt?

A cross-domain benchmark for comparing optimizers in large-scale model training. It covers 24+ optimizers across two stages: Stage 1 sweeps on C4 with LLaMA-3…

Released

2026-07-04

Evaluates

General AI, Geometric reasoning, Language & Knowledge, cs.LG

Openness

unknown

Importer

Claire Radar

Review status

unreviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2607.04033

OmniOpt paper and code

• Paper https://arxiv.org/abs/2607.04033
• Code https://github.com/OpenRaiser/OmniOpt

Which sources cite OmniOpt?

• OmniOpt

Related General AI, Geometric reasoning, Language & Knowledge, cs.LG benchmarks

• $\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
• 3DCodeBench
• 5-Dialects-BN
• A Large-Scale Dataset and Benchmark
• A3S-Bench
• AARRI-Bench
• ABC-Bench
• ActBench