The Dataset Viewer has been disabled on this dataset.

simreadybench CodeBuddy runs (MuJoCo and Isaac)

Generation runs of the sim-ready benchmark with the CodeBuddy Code CLI (2.182.0): each budget a fresh session, budgets 900/1800/2700 s. The benchmark's items came in two splits; campaign in a run's run.json says which one it belongs to.

split campaign in run.json engine items runs
1 jiarui-codebuddy MuJoCo 111 5 x 333 = 1665
1 jiarui-codebuddy-isaac-a, jiarui-codebuddy-isaac-b Isaac Sim 111 5 x 333 = 1665
2 jiarui-2-codebuddy MuJoCo 214 2896 of 5 x 642 = 3210
2 jiarui-2-codebuddy Isaac Sim 214 2704 of 5 x 642 = 3210

Five models in both splits: Kimi K3.1 Pro, GLM 5.3 Flash, DeepSeek V4.1 Flash, Hunyuan 4 (hy4std) and Hunyuan 4 preview (hy4-preview). The model of a run is model_id in its run.json. Hunyuan 4 (hy4std) has a daily usage quota, so the second split's last 820 hy4std runs (MuJoCo 314, Isaac 506) are still running and will be added when they finish.

Layout: <track>/<engine>/<item>/<run_id>/ (track = scenes, objects or demos) with run.json, generation.jsonl (the CLI's stream-json log), submitted_prompt.txt, input/, output/. Each split's campaign state (every job and the run it produced) is in campaigns/jiarui-codebuddy/state.json and campaigns/jiarui-2-codebuddy/state.json, merged from the hosts that ran it.

Not evaluated yet. Model ids are given without deployment suffixes and some host-specific strings and credentials are redacted. The Hub refuses paths containing /x00, so files named x00* are stored as _x00*.

The Claude Code (Claude Opus 5.5) and Codex (GPT-6.1 Sol) runs of the same benchmark are in FlippyDora/simreadybench-claude-codex-runs.

Downloads last month
331