[ { "step": 500, "loss": 0.04259096086025238, "target_updates": 4, "greedy": { "3": 0.8, "6": 0.03, "10": 0.0, "15": 0.0, "20": 0.0 }, "mean": 0.166 }, { "step": 1000, "loss": 0.08486837148666382, "target_updates": 6, "greedy": { "3": 0.885, "6": 0.065, "10": 0.0, "15": 0.0, "20": 0.0 }, "mean": 0.19 }, { "step": 1500, "loss": 0.11638806760311127, "target_updates": 8, "greedy": { "3": 0.9, "6": 0.1, "10": 0.005, "15": 0.0, "20": 0.0 }, "mean": 0.20099999999999998 }, { "step": 2000, "loss": 0.08094704151153564, "target_updates": 9, "greedy": { "3": 0.97, "6": 0.29, "10": 0.005, "15": 0.0, "20": 0.0 }, "mean": 0.253 }, { "step": 2500, "loss": 0.08376012742519379, "target_updates": 10, "greedy": { "3": 1.0, "6": 0.325, "10": 0.005, "15": 0.0, "20": 0.0 }, "mean": 0.26599999999999996 }, { "step": 3000, "loss": 0.09172387421131134, "target_updates": 11, "greedy": { "3": 0.995, "6": 0.42, "10": 0.025, "15": 0.0, "20": 0.0 }, "mean": 0.288 }, { "step": 3500, "loss": 0.15869717299938202, "target_updates": 13, "greedy": { "3": 0.995, "6": 0.49, "10": 0.03, "15": 0.0, "20": 0.0 }, "mean": 0.303 }, { "step": 4000, "loss": 0.2646712064743042, "target_updates": 14, "greedy": { "3": 1.0, "6": 0.55, "10": 0.03, "15": 0.0, "20": 0.0 }, "mean": 0.316 }, { "step": 4500, "loss": 0.19903528690338135, "target_updates": 15, "greedy": { "3": 1.0, "6": 0.595, "10": 0.04, "15": 0.0, "20": 0.0 }, "mean": 0.327 }, { "step": 5000, "loss": 0.27375563979148865, "target_updates": 16, "greedy": { "3": 1.0, "6": 0.62, "10": 0.02, "15": 0.0, "20": 0.0 }, "mean": 0.328 }, { "step": 5500, "loss": 0.24008296430110931, "target_updates": 18, "greedy": { "3": 1.0, "6": 0.64, "10": 0.035, "15": 0.0, "20": 0.0 }, "mean": 0.335 }, { "step": 6000, "loss": 0.22554944455623627, "target_updates": 19, "greedy": { "3": 1.0, "6": 0.645, "10": 0.04, "15": 0.0, "20": 0.0 }, "mean": 0.337 }, { "step": 6500, "loss": 0.2247759848833084, "target_updates": 20, "greedy": { "3": 1.0, "6": 0.66, "10": 0.055, "15": 0.0, "20": 0.0 }, "mean": 0.343 }, { "step": 7000, "loss": 0.1801447868347168, "target_updates": 21, "greedy": { "3": 1.0, "6": 0.73, "10": 0.06, "15": 0.0, "20": 0.0 }, "mean": 0.358 }, { "step": 7500, "loss": 0.19227959215641022, "target_updates": 23, "greedy": { "3": 1.0, "6": 0.725, "10": 0.05, "15": 0.0, "20": 0.0 }, "mean": 0.35500000000000004 }, { "step": 8000, "loss": 0.17697712779045105, "target_updates": 24, "greedy": { "3": 1.0, "6": 0.765, "10": 0.085, "15": 0.0, "20": 0.0 }, "mean": 0.37 }, { "step": 8500, "loss": 0.20436029136180878, "target_updates": 25, "greedy": { "3": 1.0, "6": 0.77, "10": 0.085, "15": 0.0, "20": 0.0 }, "mean": 0.371 }, { "step": 9000, "loss": 0.14036798477172852, "target_updates": 26, "greedy": { "3": 1.0, "6": 0.81, "10": 0.075, "15": 0.0, "20": 0.0 }, "mean": 0.377 }, { "step": 9500, "loss": 0.19980084896087646, "target_updates": 28, "greedy": { "3": 1.0, "6": 0.765, "10": 0.085, "15": 0.0, "20": 0.0 }, "mean": 0.37 }, { "step": 10000, "loss": 0.16410405933856964, "target_updates": 29, "greedy": { "3": 1.0, "6": 0.8, "10": 0.075, "15": 0.005, "20": 0.0 }, "mean": 0.376 }, { "step": 10500, "loss": 0.16999322175979614, "target_updates": 30, "greedy": { "3": 1.0, "6": 0.815, "10": 0.11, "15": 0.0, "20": 0.0 }, "mean": 0.385 }, { "step": 11000, "loss": 0.16925780475139618, "target_updates": 31, "greedy": { "3": 1.0, "6": 0.8, "10": 0.105, "15": 0.0, "20": 0.0 }, "mean": 0.381 }, { "step": 11500, "loss": 0.18713811039924622, "target_updates": 33, "greedy": { "3": 1.0, "6": 0.815, "10": 0.125, "15": 0.005, "20": 0.0 }, "mean": 0.38899999999999996 }, { "step": 12000, "loss": 0.18486963212490082, "target_updates": 34, "greedy": { "3": 1.0, "6": 0.835, "10": 0.12, "15": 0.0, "20": 0.0 }, "mean": 0.391 }, { "step": 12500, "loss": 0.18077974021434784, "target_updates": 35, "greedy": { "3": 1.0, "6": 0.82, "10": 0.12, "15": 0.0, "20": 0.0 }, "mean": 0.388 }, { "step": 13000, "loss": 0.1808379739522934, "target_updates": 36, "greedy": { "3": 1.0, "6": 0.845, "10": 0.15, "15": 0.0, "20": 0.0 }, "mean": 0.39899999999999997 }, { "step": 13500, "loss": 0.17874817550182343, "target_updates": 38, "greedy": { "3": 1.0, "6": 0.865, "10": 0.14, "15": 0.005, "20": 0.0 }, "mean": 0.40199999999999997 }, { "step": 14000, "loss": 0.1975564956665039, "target_updates": 39, "greedy": { "3": 1.0, "6": 0.855, "10": 0.18, "15": 0.005, "20": 0.0 }, "mean": 0.40800000000000003 }, { "step": 14500, "loss": 0.1934536248445511, "target_updates": 40, "greedy": { "3": 1.0, "6": 0.865, "10": 0.185, "15": 0.005, "20": 0.0 }, "mean": 0.4109999999999999 }, { "step": 15000, "loss": 0.18869797885417938, "target_updates": 41, "greedy": { "3": 1.0, "6": 0.86, "10": 0.175, "15": 0.005, "20": 0.0 }, "mean": 0.4079999999999999 }, { "step": 15500, "loss": 0.19642984867095947, "target_updates": 43, "greedy": { "3": 1.0, "6": 0.865, "10": 0.18, "15": 0.0, "20": 0.0 }, "mean": 0.409 }, { "step": 16000, "loss": 0.1785489320755005, "target_updates": 44, "greedy": { "3": 1.0, "6": 0.885, "10": 0.19, "15": 0.005, "20": 0.0 }, "mean": 0.41600000000000004 }, { "step": 16500, "loss": 0.18260493874549866, "target_updates": 45, "greedy": { "3": 1.0, "6": 0.895, "10": 0.19, "15": 0.0, "20": 0.0 }, "mean": 0.417 }, { "step": 17000, "loss": 0.15584097802639008, "target_updates": 46, "greedy": { "3": 1.0, "6": 0.875, "10": 0.2, "15": 0.005, "20": 0.0 }, "mean": 0.41600000000000004 }, { "step": 17500, "loss": 0.16387194395065308, "target_updates": 48, "greedy": { "3": 1.0, "6": 0.9, "10": 0.23, "15": 0.0, "20": 0.0 }, "mean": 0.426 }, { "step": 18000, "loss": 0.17173488438129425, "target_updates": 49, "greedy": { "3": 1.0, "6": 0.915, "10": 0.205, "15": 0.005, "20": 0.0 }, "mean": 0.425 }, { "step": 18500, "loss": 0.18664591014385223, "target_updates": 50, "greedy": { "3": 1.0, "6": 0.9, "10": 0.2, "15": 0.005, "20": 0.0 }, "mean": 0.421 }, { "step": 19000, "loss": 0.19392891228199005, "target_updates": 51, "greedy": { "3": 1.0, "6": 0.93, "10": 0.215, "15": 0.0, "20": 0.0 }, "mean": 0.429 }, { "step": 19500, "loss": 0.20463189482688904, "target_updates": 53, "greedy": { "3": 1.0, "6": 0.905, "10": 0.225, "15": 0.005, "20": 0.0 }, "mean": 0.42699999999999994 }, { "step": 20000, "loss": 0.24079182744026184, "target_updates": 54, "greedy": { "3": 1.0, "6": 0.93, "10": 0.19, "15": 0.0, "20": 0.0 }, "mean": 0.42400000000000004 }, { "step": 20500, "loss": 0.2095763385295868, "target_updates": 55, "greedy": { "3": 1.0, "6": 0.915, "10": 0.2, "15": 0.0, "20": 0.0 }, "mean": 0.42300000000000004 }, { "step": 21000, "loss": 0.2016715556383133, "target_updates": 56, "greedy": { "3": 1.0, "6": 0.925, "10": 0.22, "15": 0.0, "20": 0.0 }, "mean": 0.429 }, { "step": 21500, "loss": 0.1898859441280365, "target_updates": 58, "greedy": { "3": 1.0, "6": 0.945, "10": 0.24, "15": 0.0, "20": 0.0 }, "mean": 0.43699999999999994 }, { "step": 22000, "loss": 0.17079637944698334, "target_updates": 59, "greedy": { "3": 1.0, "6": 0.94, "10": 0.22, "15": 0.0, "20": 0.0 }, "mean": 0.43200000000000005 }, { "step": 22500, "loss": 0.18184992671012878, "target_updates": 60, "greedy": { "3": 1.0, "6": 0.945, "10": 0.215, "15": 0.005, "20": 0.0 }, "mean": 0.43299999999999994 }, { "step": 23000, "loss": 0.1909194439649582, "target_updates": 61, "greedy": { "3": 1.0, "6": 0.935, "10": 0.265, "15": 0.0, "20": 0.0 }, "mean": 0.44000000000000006 }, { "step": 23500, "loss": 0.18325640261173248, "target_updates": 63, "greedy": { "3": 1.0, "6": 0.95, "10": 0.245, "15": 0.0, "20": 0.0 }, "mean": 0.43899999999999995 }, { "step": 24000, "loss": 0.18927478790283203, "target_updates": 64, "greedy": { "3": 1.0, "6": 0.93, "10": 0.285, "15": 0.0, "20": 0.0 }, "mean": 0.44300000000000006 }, { "step": 24500, "loss": 0.1842205673456192, "target_updates": 65, "greedy": { "3": 1.0, "6": 0.945, "10": 0.27, "15": 0.0, "20": 0.0 }, "mean": 0.44299999999999995 }, { "step": 25000, "loss": 0.1583680808544159, "target_updates": 66, "greedy": { "3": 1.0, "6": 0.955, "10": 0.28, "15": 0.005, "20": 0.0 }, "mean": 0.44800000000000006 }, { "step": 25500, "loss": 0.15866529941558838, "target_updates": 68, "greedy": { "3": 1.0, "6": 0.94, "10": 0.265, "15": 0.005, "20": 0.0 }, "mean": 0.442 }, { "step": 26000, "loss": 0.1440247893333435, "target_updates": 69, "greedy": { "3": 1.0, "6": 0.965, "10": 0.295, "15": 0.01, "20": 0.0 }, "mean": 0.4539999999999999 }, { "step": 26500, "loss": 0.1594139039516449, "target_updates": 70, "greedy": { "3": 1.0, "6": 0.955, "10": 0.295, "15": 0.005, "20": 0.0 }, "mean": 0.45099999999999996 }, { "step": 27000, "loss": 0.156402587890625, "target_updates": 71, "greedy": { "3": 1.0, "6": 0.95, "10": 0.345, "15": 0.01, "20": 0.0 }, "mean": 0.46099999999999997 } ]