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 2026-10-08
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1GLM-5.3 Flash0.2857$0.6810.0M99k15h 21m1243
2GLM-5.3 Flash0.2619$1.1715.8M143k16h 53m1343
3GLM-5.3 Flash0.2619$0.9712.5M122k17h 37m1272
4GLM-5.3 Flash0.2381$1.3629.9M170k18h 53m2192
5GLM-5.3 Flash0.0000$2.3142.4M198k17h 43m2612