siddharthmb/2026.RA.TimeCost-GRPO download history
siddharthmb/2026.RA.TimeCost-GRPO is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 55 times (2 in the last 7 days), and 94 times in total. It ranks #173,280 among datasets by monthly downloads.
2026.RA.TimeCost-GRPO — pricing the clock in a negotiation GRPO reward Campaign artifacts for time-cost GRPO (Qwen3-8B LoRA negotiation policy). Every reward this program has trained on is blind to the clock: GameSpec.discount = 0.95 is declared in the spec and applied to nothing, so