{ "math_check": { "logmeanexp_absolute_difference": 9.928565702921333e-08, "sample_scaling": [ { "K": 2, "abs_error_mean": 0.7157605508005723, "bias": 0.436480942274036 }, { "K": 8, "abs_error_mean": 0.40372897433680116, "bias": 0.1729428094129632 }, { "K": 32, "abs_error_mean": 0.2113885934796732, "bias": 0.05154215594925562 }, { "K": 128, "abs_error_mean": 0.11300307392602017, "bias": 0.014449644809360049 } ] }, "runs": [ { "mode": "direct", "seed": 11, "device": "cuda", "seconds": 7.432541204998415, "estimator_variance": 5.50183676395477e-10, "value_target_rmse": 6.896183967590332, "ce_rmse": NaN, "checkpoints": [ { "step": 601, "ce_abs_error": NaN, "bellman_rmse": 4.506957530975342, "loss": 122.40324401855469 }, { "step": 1200, "ce_abs_error": NaN, "bellman_rmse": 6.88680362701416, "loss": 425.3627624511719 } ] }, { "mode": "aux", "seed": 11, "device": "cuda", "seconds": 9.682978046999779, "estimator_variance": 2.9493247843959125e-09, "value_target_rmse": 6.245853900909424, "ce_rmse": 9.322518348693848, "checkpoints": [ { "step": 601, "ce_abs_error": 5.964967250823975, "bellman_rmse": 1.4166280031204224, "loss": 66.29203033447266 }, { "step": 1200, "ce_abs_error": 9.322509765625, "bellman_rmse": 2.374951124191284, "loss": 234.13453674316406 } ] }, { "mode": "direct", "seed": 22, "device": "cuda", "seconds": 6.797823614997469, "estimator_variance": 5.42665135050413e-10, "value_target_rmse": 6.920567035675049, "ce_rmse": NaN, "checkpoints": [ { "step": 601, "ce_abs_error": NaN, "bellman_rmse": 4.586386203765869, "loss": 122.48664855957031 }, { "step": 1200, "ce_abs_error": NaN, "bellman_rmse": 6.939746379852295, "loss": 431.5095520019531 } ] }, { "mode": "aux", "seed": 22, "device": "cuda", "seconds": 9.508438042998023, "estimator_variance": 5.449813045288465e-09, "value_target_rmse": 6.109918117523193, "ce_rmse": 9.53398323059082, "checkpoints": [ { "step": 601, "ce_abs_error": 5.9523162841796875, "bellman_rmse": 1.4783697128295898, "loss": 52.603919982910156 }, { "step": 1200, "ce_abs_error": 9.533950805664062, "bellman_rmse": 2.669316530227661, "loss": 212.74722290039062 } ] }, { "mode": "direct", "seed": 33, "device": "cuda", "seconds": 7.001646349999646, "estimator_variance": 4.765146055518699e-10, "value_target_rmse": 6.83821439743042, "ce_rmse": NaN, "checkpoints": [ { "step": 601, "ce_abs_error": NaN, "bellman_rmse": 4.4487528800964355, "loss": 119.88429260253906 }, { "step": 1200, "ce_abs_error": NaN, "bellman_rmse": 6.826839923858643, "loss": 417.4648742675781 } ] }, { "mode": "aux", "seed": 33, "device": "cuda", "seconds": 9.544574978997844, "estimator_variance": 2.3244410840561613e-09, "value_target_rmse": 6.222592830657959, "ce_rmse": 9.921758651733398, "checkpoints": [ { "step": 601, "ce_abs_error": 6.286478042602539, "bellman_rmse": 1.7113144397735596, "loss": 60.145721435546875 }, { "step": 1200, "ce_abs_error": 9.921757698059082, "bellman_rmse": 2.9235434532165527, "loss": 220.10765075683594 } ] } ] }