Traces

Every run: the agent's full trajectory, its submissions over the 20-hour budget, and the final score.

Multi-GPU Efficient Finetuning

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Kimi K30.3333$19.1518.0M131k19h 9m1676
2Kimi K30.3095$16.5615.3M148k19h 10m1343
3Kimi K30.1905$16.1311.8M128k18h 47m1161
4Kimi K30.0476$25.1335.1M233k19h 37m2361
5Kimi K30.0000$10.948.7M101k18h 40m1211