Upload experiments/weak_to_strong/reasoning/knights_knaves/distill_baseline/260814-delta-sampledrkl-knights_knaves-q3-8b-base-t4b-s105-tr8k-va8k-save30-best-latestopt-b64-mb32-dp4/global_step_15/actor/model_world_size_4_rank_0.pt with huggingface_hub