{ "updated_unix": 1793192568.9132428, "cases": 20, "epochs_each": 2000, "rows": [ { "variant": "S001", "model": "TS3", "layers": 16, "width": 256, "heads": 8, "slices": 1, "parameters": 7593108, "role": "slice_count_ablation", "phase": "completed", "best_checkpoint_epoch": 2000, "mean_U_L2_percent": 13.341607534989047, "worst_U_L2_percent": 47.910896132053196, "mean_node_VM_L2_percent": 20.196433052079183, "worst_node_VM_L2_percent": 35.24903499224592, "elapsed_seconds": 5090.240893125534 }, { "variant": "S002", "model": "TS3", "layers": 16, "width": 256, "heads": 8, "slices": 2, "parameters": 7593636, "role": "slice_count_ablation", "phase": "completed", "best_checkpoint_epoch": 2000, "mean_U_L2_percent": 14.42519259266328, "worst_U_L2_percent": 50.16383400210016, "mean_node_VM_L2_percent": 21.128552856737773, "worst_node_VM_L2_percent": 35.822720320770465, "elapsed_seconds": 6145.8671045303345 }, { "variant": "S004", "model": "TS3", "layers": 16, "width": 256, "heads": 8, "slices": 4, "parameters": 7594692, "role": "slice_count_ablation", "phase": "completed", "best_checkpoint_epoch": 2000, "mean_U_L2_percent": 14.157596762532588, "worst_U_L2_percent": 49.41444096102204, "mean_node_VM_L2_percent": 22.38332572001397, "worst_node_VM_L2_percent": 36.94918798409384, "elapsed_seconds": 5195.455749750137 }, { "variant": "S008", "model": "TS3", "layers": 16, "width": 256, "heads": 8, "slices": 8, "parameters": 7596804, "role": "slice_count_ablation", "phase": "completed", "best_checkpoint_epoch": 2000, "mean_U_L2_percent": 13.400675251909947, "worst_U_L2_percent": 48.64786710016021, "mean_node_VM_L2_percent": 19.521377020706026, "worst_node_VM_L2_percent": 34.87050221982171, "elapsed_seconds": 5172.21451997757 }, { "variant": "S016", "model": "TS3", "layers": 16, "width": 256, "heads": 8, "slices": 16, "parameters": 7601028, "role": "slice_count_ablation", "phase": "completed", "best_checkpoint_epoch": 2000, "mean_U_L2_percent": 14.337526767621211, "worst_U_L2_percent": 49.48300514922367, "mean_node_VM_L2_percent": 21.192432646871502, "worst_node_VM_L2_percent": 35.875370086811635, "elapsed_seconds": 5192.929567813873 }, { "variant": "S032", "model": "TS3", "layers": 16, "width": 256, "heads": 8, "slices": 32, "parameters": 7609476, "role": "slice_count_ablation", "phase": "completed", "best_checkpoint_epoch": 2000, "mean_U_L2_percent": 13.913043360101751, "worst_U_L2_percent": 48.201670125340904, "mean_node_VM_L2_percent": 21.188621225860306, "worst_node_VM_L2_percent": 35.735227035870416, "elapsed_seconds": 5386.258642673492 }, { "variant": "S064", "model": "TS3", "layers": 16, "width": 256, "heads": 8, "slices": 64, "parameters": 7626372, "role": "reused_64_baseline", "phase": "completed_reused", "best_checkpoint_epoch": 2000, "mean_U_L2_percent": 14.875156625607161, "worst_U_L2_percent": 51.994215887763104, "mean_node_VM_L2_percent": 20.88300502849246, "worst_node_VM_L2_percent": 35.95954078414134, "elapsed_seconds": 7205.401588201523 }, { "variant": "H064", "model": "TS3", "layers": 8, "width": 256, "heads": 8, "slices": 64, "parameters": 3894340, "role": "half_parameter_comparison", "phase": "completed", "best_checkpoint_epoch": 2000, "mean_U_L2_percent": 13.655987851600553, "worst_U_L2_percent": 47.88284152119077, "mean_node_VM_L2_percent": 21.09179323956881, "worst_node_VM_L2_percent": 36.774165845452416, "elapsed_seconds": 3620.7709844112396 }, { "model": "MGN", "width": 256, "parameters": 9298436, "role": "existing_20_case_baseline_not_retrained", "phase": "completed_reused", "best_checkpoint_epoch": 2000, "mean_U_L2_percent": 2.1165523057834945, "worst_U_L2_percent": 4.08942404721866, "mean_node_VM_L2_percent": 11.238286817850183, "worst_node_VM_L2_percent": 15.89784101523684, "elapsed_seconds": 10857.828392505646 } ], "curves": [ { "variant": "S001", "points": [ { "step": 50, "elapsed_seconds": 127.05299854278564, "loss": 2.8137219503521917, "learning_rate": 0.0003, "mean_training_objective": 1.7094289028920073, "mean_relative_U_L2": 1.3443739098657155, "worst_relative_U_L2": 4.391310716683112, "mean_relative_VM_L2": 0.7022894679648186, "worst_relative_VM_L2": 0.9363746065153492 }, { "step": 100, "elapsed_seconds": 254.3136396408081, "loss": 2.160243607312441, "learning_rate": 0.00029952980967946045, "mean_training_objective": 1.353310453120708, "mean_relative_U_L2": 1.189607502125808, "worst_relative_U_L2": 3.982582741365088, "mean_relative_VM_L2": 0.6415273562413304, "worst_relative_VM_L2": 0.8833083152171557 }, { "step": 150, "elapsed_seconds": 382.3522620201111, "loss": 2.077729442715645, "learning_rate": 0.0002981222880824973, "mean_training_objective": 2.4734370067802276, "mean_relative_U_L2": 1.46000794953093, "worst_relative_U_L2": 5.982190594297814, "mean_relative_VM_L2": 0.5618173365546328, "worst_relative_VM_L2": 0.8319425098995294 }, { "step": 200, "elapsed_seconds": 508.59428215026855, "loss": 0.7482014108449222, "learning_rate": 0.0002957865635267775, "mean_training_objective": 1.0125825968919218, "mean_relative_U_L2": 0.9817511503133168, "worst_relative_U_L2": 3.66603668468503, "mean_relative_VM_L2": 0.4616333545607957, "worst_relative_VM_L2": 0.7057210878786349 }, { "step": 250, "elapsed_seconds": 648.484540939331, "loss": 0.37335884775966405, "learning_rate": 0.00029253778408233605, "mean_training_objective": 0.40028637606216544, "mean_relative_U_L2": 0.6173111037048729, "worst_relative_U_L2": 2.005762679856052, "mean_relative_VM_L2": 0.3664797026704318, "worst_relative_VM_L2": 0.5495309760288942 }, { "step": 300, "elapsed_seconds": 782.1295154094696, "loss": 0.2572360089980065, "learning_rate": 0.0002883970193305295, "mean_training_objective": 0.3534941021176102, "mean_relative_U_L2": 0.5733653714830055, "worst_relative_U_L2": 1.9227859569725982, "mean_relative_VM_L2": 0.33249063757857444, "worst_relative_VM_L2": 0.5216339855211611 }, { "step": 350, "elapsed_seconds": 911.184086561203, "loss": 0.2622101817280055, "learning_rate": 0.0002833911237197154, "mean_training_objective": 0.2421368303052273, "mean_relative_U_L2": 0.46666305135772124, "worst_relative_U_L2": 1.4695313901904248, "mean_relative_VM_L2": 0.31508637214620777, "worst_relative_VM_L2": 0.48156279713510375 }, { "step": 400, "elapsed_seconds": 1036.0214009284973, "loss": 0.2978905342519284, "learning_rate": 0.00027755256240385024, "mean_training_objective": 0.3749187805330196, "mean_relative_U_L2": 0.5774988386680516, "worst_relative_U_L2": 2.1132461350355003, "mean_relative_VM_L2": 0.3330695687313718, "worst_relative_VM_L2": 0.4900442548080578 }, { "step": 450, "elapsed_seconds": 1163.4766857624054, "loss": 0.9267067855224014, "learning_rate": 0.0002709192006935076, "mean_training_objective": 0.23972577352460905, "mean_relative_U_L2": 0.5002626382751321, "worst_relative_U_L2": 0.9874759857419775, "mean_relative_VM_L2": 0.3527184210019624, "worst_relative_VM_L2": 0.5980599580516716 }, { "step": 500, "elapsed_seconds": 1288.61785197258, "loss": 0.24408338014036413, "learning_rate": 0.0002635340584848096, "mean_training_objective": 0.7752267245468402, "mean_relative_U_L2": 0.7928066347448228, "worst_relative_U_L2": 3.4738604950376493, "mean_relative_VM_L2": 0.3134323703344376, "worst_relative_VM_L2": 0.41507507460407816 }, { "step": 550, "elapsed_seconds": 1416.9484462738037, "loss": 0.3238934841938316, "learning_rate": 0.0002554450312588919, "mean_training_objective": 0.30037474775086803, "mean_relative_U_L2": 0.5206236053592652, "worst_relative_U_L2": 1.8986667275913163, "mean_relative_VM_L2": 0.2964103326241453, "worst_relative_VM_L2": 0.4351678726340087 }, { "step": 600, "elapsed_seconds": 1544.1486909389496, "loss": 0.18545181881636383, "learning_rate": 0.00024670457946132966, "mean_training_objective": 0.17041268403456145, "mean_relative_U_L2": 0.37716336769163095, "worst_relative_U_L2": 1.2357711116181482, "mean_relative_VM_L2": 0.27750891976617836, "worst_relative_VM_L2": 0.3987744126618612 }, { "step": 650, "elapsed_seconds": 1667.1263890266418, "loss": 0.1550300186499953, "learning_rate": 0.0002373693882760176, "mean_training_objective": 0.14167836663976272, "mean_relative_U_L2": 0.3356588771792698, "worst_relative_U_L2": 1.05290669484823, "mean_relative_VM_L2": 0.28223206717678295, "worst_relative_VM_L2": 0.38688709600868115 }, { "step": 700, "elapsed_seconds": 1791.2722125053406, "loss": 0.13489659735932943, "learning_rate": 0.00022749999999999997, "mean_training_objective": 0.12427084313994538, "mean_relative_U_L2": 0.3146890757022938, "worst_relative_U_L2": 0.9888783648739252, "mean_relative_VM_L2": 0.2673232491236438, "worst_relative_VM_L2": 0.3821132090174356 }, { "step": 750, "elapsed_seconds": 1915.7069401741028, "loss": 0.11579944901168346, "learning_rate": 0.0002171604214034428, "mean_training_objective": 0.1063712079560094, "mean_relative_U_L2": 0.27808655694445295, "worst_relative_U_L2": 0.8716257373067985, "mean_relative_VM_L2": 0.26071840297257987, "worst_relative_VM_L2": 0.38130970986104346 }, { "step": 800, "elapsed_seconds": 2042.7660808563232, "loss": 0.11177034862339497, "learning_rate": 0.00020641770862116766, "mean_training_objective": 0.11982328770861768, "mean_relative_U_L2": 0.34097652401189393, "worst_relative_U_L2": 0.8335028783892633, "mean_relative_VM_L2": 0.25788721835572453, "worst_relative_VM_L2": 0.38073687086720315 }, { "step": 850, "elapsed_seconds": 2167.5666830539703, "loss": 0.10189055297523737, "learning_rate": 0.00019534153226788558, "mean_training_objective": 0.09693461159643053, "mean_relative_U_L2": 0.274232166679295, "worst_relative_U_L2": 0.7983228066362493, "mean_relative_VM_L2": 0.2543305140211011, "worst_relative_VM_L2": 0.3803670876109113 }, { "step": 900, "elapsed_seconds": 2292.349007129669, "loss": 0.09836873896420001, "learning_rate": 0.00018400372559752647, "mean_training_objective": 0.08567603693744187, "mean_relative_U_L2": 0.2494962340114907, "worst_relative_U_L2": 0.6835809184185557, "mean_relative_VM_L2": 0.24765388667391425, "worst_relative_VM_L2": 0.3778832705722717 }, { "step": 950, "elapsed_seconds": 2417.7809085845947, "loss": 0.1043436750769615, "learning_rate": 0.00017247781863702187, "mean_training_objective": 0.16611265500880598, "mean_relative_U_L2": 0.39534718114848133, "worst_relative_U_L2": 1.1083440051948588, "mean_relative_VM_L2": 0.3060488727541101, "worst_relative_VM_L2": 0.404209965525861 }, { "step": 1000, "elapsed_seconds": 2542.505879878998, "loss": 0.13460614848881958, "learning_rate": 0.0001608385613158652, "mean_training_objective": 0.12104163955810576, "mean_relative_U_L2": 0.30709046900142994, "worst_relative_U_L2": 1.0413003989055662, "mean_relative_VM_L2": 0.2587424679929458, "worst_relative_VM_L2": 0.38231143827529024 }, { "step": 1050, "elapsed_seconds": 2666.7103741168976, "loss": 0.10204384448006751, "learning_rate": 0.00014916143868413484, "mean_training_objective": 0.09428104057199845, "mean_relative_U_L2": 0.26309150151491395, "worst_relative_U_L2": 0.8252356537689635, "mean_relative_VM_L2": 0.24856898607955621, "worst_relative_VM_L2": 0.3772817615705904 }, { "step": 1100, "elapsed_seconds": 2791.458160161972, "loss": 0.08384233801625669, "learning_rate": 0.00013752218136297816, "mean_training_objective": 0.07826204805707293, "mean_relative_U_L2": 0.2313724179861624, "worst_relative_U_L2": 0.6564320442301247, "mean_relative_VM_L2": 0.24282082673823804, "worst_relative_VM_L2": 0.3761064213711508 }, { "step": 1150, "elapsed_seconds": 2924.39519238472, "loss": 0.07948410669341684, "learning_rate": 0.00012599627440247358, "mean_training_objective": 0.07556411044357278, "mean_relative_U_L2": 0.22608166883201095, "worst_relative_U_L2": 0.6275598031840921, "mean_relative_VM_L2": 0.23926414551009584, "worst_relative_VM_L2": 0.3750035223543296 }, { "step": 1200, "elapsed_seconds": 3047.7118515968323, "loss": 0.06938800825737416, "learning_rate": 0.00011465846773211439, "mean_training_objective": 0.06591459033479936, "mean_relative_U_L2": 0.20378418577594073, "worst_relative_U_L2": 0.5294485669684571, "mean_relative_VM_L2": 0.23383123866659003, "worst_relative_VM_L2": 0.3737149296167319 }, { "step": 1250, "elapsed_seconds": 3176.022289276123, "loss": 0.06697877906262875, "learning_rate": 0.00010358229137883234, "mean_training_objective": 0.06437402866008285, "mean_relative_U_L2": 0.20137329644343396, "worst_relative_U_L2": 0.5270097792990156, "mean_relative_VM_L2": 0.22982814870477863, "worst_relative_VM_L2": 0.37222084632685115 }, { "step": 1300, "elapsed_seconds": 3305.042543888092, "loss": 0.06091966493986546, "learning_rate": 9.283957859655712e-05, "mean_training_objective": 0.05925558828643316, "mean_relative_U_L2": 0.18758974395258915, "worst_relative_U_L2": 0.5206975888689388, "mean_relative_VM_L2": 0.22587173124814672, "worst_relative_VM_L2": 0.37051757660204315 }, { "step": 1350, "elapsed_seconds": 3430.739282131195, "loss": 0.06198491612449288, "learning_rate": 8.250000000000003e-05, "mean_training_objective": 0.05660738808655352, "mean_relative_U_L2": 0.18467668432383538, "worst_relative_U_L2": 0.5171221305970373, "mean_relative_VM_L2": 0.22228592328779442, "worst_relative_VM_L2": 0.36913084101450433 }, { "step": 1400, "elapsed_seconds": 3559.145592212677, "loss": 0.054925853991881005, "learning_rate": 7.263061172398237e-05, "mean_training_objective": 0.053555021058697985, "mean_relative_U_L2": 0.1735657636787877, "worst_relative_U_L2": 0.5126871280537127, "mean_relative_VM_L2": 0.21861040990515632, "worst_relative_VM_L2": 0.36705602156269684 }, { "step": 1450, "elapsed_seconds": 3687.7562975883484, "loss": 0.05126221608370543, "learning_rate": 6.32954205386703e-05, "mean_training_objective": 0.05033340529581576, "mean_relative_U_L2": 0.16260254183825135, "worst_relative_U_L2": 0.5083105053471842, "mean_relative_VM_L2": 0.21575261086570716, "worst_relative_VM_L2": 0.3648081907900573 }, { "step": 1500, "elapsed_seconds": 3812.9915862083435, "loss": 0.04969895845279097, "learning_rate": 5.455496874110809e-05, "mean_training_objective": 0.0489693381058701, "mean_relative_U_L2": 0.15873881552800845, "worst_relative_U_L2": 0.5039268389206093, "mean_relative_VM_L2": 0.21304351866931034, "worst_relative_VM_L2": 0.362480304340242 }, { "step": 1550, "elapsed_seconds": 3941.5642478466034, "loss": 0.04711375338956714, "learning_rate": 4.646594151519032e-05, "mean_training_objective": 0.04668620335806864, "mean_relative_U_L2": 0.14987286408619524, "worst_relative_U_L2": 0.4995386103091235, "mean_relative_VM_L2": 0.21077172945507544, "worst_relative_VM_L2": 0.36015587770099733 }, { "step": 1600, "elapsed_seconds": 4068.4328451156616, "loss": 0.056455945782363404, "learning_rate": 3.908079930649232e-05, "mean_training_objective": 0.05355517200400698, "mean_relative_U_L2": 0.1814388605889797, "worst_relative_U_L2": 0.5006749053284013, "mean_relative_VM_L2": 0.21810943111730013, "worst_relative_VM_L2": 0.3646065716137248 }, { "step": 1650, "elapsed_seconds": 4193.76366686821, "loss": 0.05295168254524468, "learning_rate": 3.244743759614975e-05, "mean_training_objective": 0.05165519054827078, "mean_relative_U_L2": 0.1737033820933255, "worst_relative_U_L2": 0.4980002376059822, "mean_relative_VM_L2": 0.2266600727447928, "worst_relative_VM_L2": 0.36741894668940406 }, { "step": 1700, "elapsed_seconds": 4323.505651712418, "loss": 0.05177939063869417, "learning_rate": 2.6608876280284544e-05, "mean_training_objective": 0.04916816352803158, "mean_relative_U_L2": 0.15666535913694013, "worst_relative_U_L2": 0.4963314422396308, "mean_relative_VM_L2": 0.2218655675590749, "worst_relative_VM_L2": 0.3678821567231228 }, { "step": 1750, "elapsed_seconds": 4448.888975381851, "loss": 0.045087520685046914, "learning_rate": 2.160298066947052e-05, "mean_training_objective": 0.04485765797220786, "mean_relative_U_L2": 0.14413180069883738, "worst_relative_U_L2": 0.489577975526909, "mean_relative_VM_L2": 0.208834933398641, "worst_relative_VM_L2": 0.3617636429115258 }, { "step": 1800, "elapsed_seconds": 4577.379599809647, "loss": 0.04407362346537411, "learning_rate": 1.7462215917663923e-05, "mean_training_objective": 0.043919468267514514, "mean_relative_U_L2": 0.1405208416249073, "worst_relative_U_L2": 0.4875244105053997, "mean_relative_VM_L2": 0.2069245975524084, "worst_relative_VM_L2": 0.3589758154265501 }, { "step": 1850, "elapsed_seconds": 4702.994608163834, "loss": 0.043214503908529885, "learning_rate": 1.4213436473222475e-05, "mean_training_objective": 0.04312334312630993, "mean_relative_U_L2": 0.1372761428607002, "worst_relative_U_L2": 0.48549009202946486, "mean_relative_VM_L2": 0.2055332716361999, "worst_relative_VM_L2": 0.35696447866059966 }, { "step": 1900, "elapsed_seconds": 4831.86917090416, "loss": 0.042690380755811934, "learning_rate": 1.1877711917502638e-05, "mean_training_objective": 0.042610849877534024, "mean_relative_U_L2": 0.1360003366745433, "worst_relative_U_L2": 0.4833598407712092, "mean_relative_VM_L2": 0.20429091609933483, "worst_relative_VM_L2": 0.3552851939322031 }, { "step": 1950, "elapsed_seconds": 4959.2043035030365, "loss": 0.042135941702872506, "learning_rate": 1.047019032053956e-05, "mean_training_objective": 0.04207759893918966, "mean_relative_U_L2": 0.1343091320306975, "worst_relative_U_L2": 0.4812656309714803, "mean_relative_VM_L2": 0.20313263663003012, "worst_relative_VM_L2": 0.3538142068233671 }, { "step": 2000, "elapsed_seconds": 5087.553075551987, "loss": 0.04167998423799873, "learning_rate": 1e-05, "mean_training_objective": 0.04163504900424085, "mean_relative_U_L2": 0.13341607534989047, "worst_relative_U_L2": 0.479108961320532, "mean_relative_VM_L2": 0.20196433052079182, "worst_relative_VM_L2": 0.3524903499224592 } ] }, { "variant": "S002", "points": [ { "step": 50, "elapsed_seconds": 130.96780228614807, "loss": 3.0123793825507166, "learning_rate": 0.0003, "mean_training_objective": 2.132396106185777, "mean_relative_U_L2": 1.4656678851376852, "worst_relative_U_L2": 5.210188409963773, "mean_relative_VM_L2": 0.7025753499495987, "worst_relative_VM_L2": 0.9358532428359109 }, { "step": 100, "elapsed_seconds": 268.20361948013306, "loss": 1.4646827057003973, "learning_rate": 0.00029952980967946045, "mean_training_objective": 1.3890707396764888, "mean_relative_U_L2": 1.2477928495459945, "worst_relative_U_L2": 3.5341503832248846, "mean_relative_VM_L2": 0.6399018667854823, "worst_relative_VM_L2": 0.869078093412952 }, { "step": 150, "elapsed_seconds": 405.497092962265, "loss": 0.5816525191068649, "learning_rate": 0.0002981222880824973, "mean_training_objective": 0.3143460997242066, "mean_relative_U_L2": 0.5047261125819527, "worst_relative_U_L2": 0.9950100360300043, "mean_relative_VM_L2": 0.5211408112149838, "worst_relative_VM_L2": 0.82153863490224 }, { "step": 200, "elapsed_seconds": 538.0120069980621, "loss": 0.7451013114303352, "learning_rate": 0.0002957865635267775, "mean_training_objective": 1.3506763462342164, "mean_relative_U_L2": 1.131762074070576, "worst_relative_U_L2": 4.353786831141096, "mean_relative_VM_L2": 0.48818583022347467, "worst_relative_VM_L2": 0.7755674561523234 }, { "step": 250, "elapsed_seconds": 671.4423723220825, "loss": 0.8690464075654745, "learning_rate": 0.00029253778408233605, "mean_training_objective": 0.3020267336076134, "mean_relative_U_L2": 0.5420501724902598, "worst_relative_U_L2": 1.3232691697847112, "mean_relative_VM_L2": 0.39858039654900046, "worst_relative_VM_L2": 0.7130394359632641 }, { "step": 300, "elapsed_seconds": 807.771196603775, "loss": 0.3773607680574058, "learning_rate": 0.0002883970193305295, "mean_training_objective": 0.30464111292754914, "mean_relative_U_L2": 0.5400838358523518, "worst_relative_U_L2": 1.6085244379685273, "mean_relative_VM_L2": 0.35647476643701775, "worst_relative_VM_L2": 0.5392351796989237 }, { "step": 350, "elapsed_seconds": 941.7279863357544, "loss": 0.3913132032379508, "learning_rate": 0.0002833911237197154, "mean_training_objective": 0.40231224187317266, "mean_relative_U_L2": 0.6174059979557036, "worst_relative_U_L2": 2.0632736462392702, "mean_relative_VM_L2": 0.35556715058860966, "worst_relative_VM_L2": 0.5186644799912725 }, { "step": 400, "elapsed_seconds": 1076.9511501789093, "loss": 0.3955012330785394, "learning_rate": 0.00027755256240385024, "mean_training_objective": 0.32282012520131986, "mean_relative_U_L2": 0.5577089479373402, "worst_relative_U_L2": 1.8112178261393326, "mean_relative_VM_L2": 0.3398946002068069, "worst_relative_VM_L2": 0.4998098529101826 }, { "step": 450, "elapsed_seconds": 1209.0350513458252, "loss": 0.17730890791863205, "learning_rate": 0.0002709192006935076, "mean_training_objective": 0.24486975217051388, "mean_relative_U_L2": 0.4877678919533646, "worst_relative_U_L2": 1.4789862184992804, "mean_relative_VM_L2": 0.3074272481038173, "worst_relative_VM_L2": 0.46138862944839915 }, { "step": 500, "elapsed_seconds": 1344.8975224494934, "loss": 0.408481109328568, "learning_rate": 0.0002635340584848096, "mean_training_objective": 0.17368266160959284, "mean_relative_U_L2": 0.42218471342909103, "worst_relative_U_L2": 0.8076044993052949, "mean_relative_VM_L2": 0.3448369707079409, "worst_relative_VM_L2": 0.4699331765331076 }, { "step": 550, "elapsed_seconds": 1476.66526222229, "loss": 1.0210404425859452, "learning_rate": 0.0002554450312588919, "mean_training_objective": 0.6022625157322447, "mean_relative_U_L2": 0.8550501673887991, "worst_relative_U_L2": 2.415406052568963, "mean_relative_VM_L2": 0.4399229711322474, "worst_relative_VM_L2": 0.6008822711382155 }, { "step": 600, "elapsed_seconds": 1612.9336638450623, "loss": 0.1731933425180614, "learning_rate": 0.00024670457946132966, "mean_training_objective": 0.19562579630461588, "mean_relative_U_L2": 0.4378229562341335, "worst_relative_U_L2": 1.0629813136151558, "mean_relative_VM_L2": 0.3165325257249277, "worst_relative_VM_L2": 0.4490267177381898 }, { "step": 650, "elapsed_seconds": 1743.9283680915833, "loss": 0.38719163853675126, "learning_rate": 0.0002373693882760176, "mean_training_objective": 0.15616721856765006, "mean_relative_U_L2": 0.3743470790354039, "worst_relative_U_L2": 0.8166484041382298, "mean_relative_VM_L2": 0.3119051878441573, "worst_relative_VM_L2": 0.44476703132820955 }, { "step": 700, "elapsed_seconds": 1915.0711534023285, "loss": 0.14729822343215349, "learning_rate": 0.00022749999999999997, "mean_training_objective": 0.2051847524739419, "mean_relative_U_L2": 0.4265080632150492, "worst_relative_U_L2": 1.3817069752230788, "mean_relative_VM_L2": 0.2911879172997056, "worst_relative_VM_L2": 0.41657131598241814 }, { "step": 750, "elapsed_seconds": 2106.6072056293488, "loss": 0.15597967701032756, "learning_rate": 0.0002171604214034428, "mean_training_objective": 0.13653550793808228, "mean_relative_U_L2": 0.33117368436568345, "worst_relative_U_L2": 0.9403480266648513, "mean_relative_VM_L2": 0.28418813790053815, "worst_relative_VM_L2": 0.4074499444879504 }, { "step": 800, "elapsed_seconds": 2301.0383155345917, "loss": 0.10803215811029078, "learning_rate": 0.00020641770862116766, "mean_training_objective": 0.1559480789341216, "mean_relative_U_L2": 0.3581250743919717, "worst_relative_U_L2": 1.1537005455115756, "mean_relative_VM_L2": 0.28100017107173675, "worst_relative_VM_L2": 0.4070464044803173 }, { "step": 850, "elapsed_seconds": 2495.8888218402863, "loss": 0.11386342551559211, "learning_rate": 0.00019534153226788558, "mean_training_objective": 0.11550007791399668, "mean_relative_U_L2": 0.2953840599739252, "worst_relative_U_L2": 0.8790479993375684, "mean_relative_VM_L2": 0.2755767489843189, "worst_relative_VM_L2": 0.3857230376365248 }, { "step": 900, "elapsed_seconds": 2695.4550251960754, "loss": 0.10521957939490677, "learning_rate": 0.00018400372559752647, "mean_training_objective": 0.11063237045380907, "mean_relative_U_L2": 0.2882628086143445, "worst_relative_U_L2": 0.8341560213452607, "mean_relative_VM_L2": 0.2736751219660116, "worst_relative_VM_L2": 0.3901026366012816 }, { "step": 950, "elapsed_seconds": 2884.369094133377, "loss": 0.09186916230246425, "learning_rate": 0.00017247781863702187, "mean_training_objective": 0.09756703760292114, "mean_relative_U_L2": 0.2618520644771854, "worst_relative_U_L2": 0.7617884961291672, "mean_relative_VM_L2": 0.2677622541876855, "worst_relative_VM_L2": 0.3823534238617672 }, { "step": 1000, "elapsed_seconds": 3070.1424753665924, "loss": 0.08766920305788514, "learning_rate": 0.0001608385613158652, "mean_training_objective": 0.10096095794562161, "mean_relative_U_L2": 0.2796912988637061, "worst_relative_U_L2": 0.7413156254407941, "mean_relative_VM_L2": 0.26750622067108865, "worst_relative_VM_L2": 0.38109547809659644 }, { "step": 1050, "elapsed_seconds": 3253.1657683849335, "loss": 0.09862167942337692, "learning_rate": 0.00014916143868413484, "mean_training_objective": 0.07468194661262925, "mean_relative_U_L2": 0.21615420146384148, "worst_relative_U_L2": 0.5541994271285146, "mean_relative_VM_L2": 0.2608754017553354, "worst_relative_VM_L2": 0.3800836731624621 }, { "step": 1100, "elapsed_seconds": 3439.1131978034973, "loss": 0.10265074763447046, "learning_rate": 0.00013752218136297816, "mean_training_objective": 0.1391176041456403, "mean_relative_U_L2": 0.35695318999861625, "worst_relative_U_L2": 0.8830943977499349, "mean_relative_VM_L2": 0.28351356826949803, "worst_relative_VM_L2": 0.3872866157086461 }, { "step": 1150, "elapsed_seconds": 3626.73645567894, "loss": 0.08569276197813452, "learning_rate": 0.00012599627440247358, "mean_training_objective": 0.08115031360618044, "mean_relative_U_L2": 0.23343826889976013, "worst_relative_U_L2": 0.5654522352389587, "mean_relative_VM_L2": 0.26161340300956787, "worst_relative_VM_L2": 0.37851022451090616 }, { "step": 1200, "elapsed_seconds": 3822.15940785408, "loss": 0.077160199964419, "learning_rate": 0.00011465846773211439, "mean_training_objective": 0.07360981329629793, "mean_relative_U_L2": 0.21300170310366653, "worst_relative_U_L2": 0.5386848392623304, "mean_relative_VM_L2": 0.2555008176317308, "worst_relative_VM_L2": 0.377069905096196 }, { "step": 1250, "elapsed_seconds": 4022.161125421524, "loss": 0.07225975543260572, "learning_rate": 0.00010358229137883234, "mean_training_objective": 0.06972537189646305, "mean_relative_U_L2": 0.20452224801741486, "worst_relative_U_L2": 0.5362320107486863, "mean_relative_VM_L2": 0.25062494465764795, "worst_relative_VM_L2": 0.3753721330415745 }, { "step": 1300, "elapsed_seconds": 4212.56346654892, "loss": 0.06732577132061124, "learning_rate": 9.283957859655712e-05, "mean_training_objective": 0.0653668000924787, "mean_relative_U_L2": 0.19394736724261713, "worst_relative_U_L2": 0.5340617413724805, "mean_relative_VM_L2": 0.2457889688928223, "worst_relative_VM_L2": 0.3736009352809318 }, { "step": 1350, "elapsed_seconds": 4396.154632091522, "loss": 0.06382881654426456, "learning_rate": 8.250000000000003e-05, "mean_training_objective": 0.06216418575346312, "mean_relative_U_L2": 0.18644247153547056, "worst_relative_U_L2": 0.5328785971433164, "mean_relative_VM_L2": 0.241105230295893, "worst_relative_VM_L2": 0.37147418943978033 }, { "step": 1400, "elapsed_seconds": 4564.948605537415, "loss": 0.06041427482850849, "learning_rate": 7.263061172398237e-05, "mean_training_objective": 0.05922594351624866, "mean_relative_U_L2": 0.1805365884540236, "worst_relative_U_L2": 0.5282994628698784, "mean_relative_VM_L2": 0.23581026791188653, "worst_relative_VM_L2": 0.3695176910510895 }, { "step": 1450, "elapsed_seconds": 4697.475116968155, "loss": 0.06484281746670603, "learning_rate": 6.32954205386703e-05, "mean_training_objective": 0.061234580909583045, "mean_relative_U_L2": 0.18791981388582116, "worst_relative_U_L2": 0.5508191754308872, "mean_relative_VM_L2": 0.2364827991531379, "worst_relative_VM_L2": 0.36728180838014635 }, { "step": 1500, "elapsed_seconds": 4831.311518192291, "loss": 0.054791502933949236, "learning_rate": 5.455496874110809e-05, "mean_training_objective": 0.054055802011137556, "mean_relative_U_L2": 0.16763571867286034, "worst_relative_U_L2": 0.5226528863453452, "mean_relative_VM_L2": 0.22681612315440902, "worst_relative_VM_L2": 0.3652151301826413 }, { "step": 1550, "elapsed_seconds": 4960.989387512207, "loss": 0.06990323364734649, "learning_rate": 4.646594151519032e-05, "mean_training_objective": 0.06996750592848396, "mean_relative_U_L2": 0.21720354252376423, "worst_relative_U_L2": 0.6441263565481415, "mean_relative_VM_L2": 0.22890207854701322, "worst_relative_VM_L2": 0.36291569059129086 }, { "step": 1600, "elapsed_seconds": 5093.171133518219, "loss": 0.08721775729209186, "learning_rate": 3.908079930649232e-05, "mean_training_objective": 0.084416114752445, "mean_relative_U_L2": 0.25742551170435457, "worst_relative_U_L2": 0.5319818318251799, "mean_relative_VM_L2": 0.2770319085369799, "worst_relative_VM_L2": 0.404963752746223 }, { "step": 1650, "elapsed_seconds": 5224.839302301407, "loss": 0.07032587146386506, "learning_rate": 3.244743759614975e-05, "mean_training_objective": 0.05702858970808816, "mean_relative_U_L2": 0.17744055012828325, "worst_relative_U_L2": 0.5168670401228788, "mean_relative_VM_L2": 0.23355851252988397, "worst_relative_VM_L2": 0.36805850148312397 }, { "step": 1700, "elapsed_seconds": 5353.815454483032, "loss": 0.05566726089455187, "learning_rate": 2.6608876280284544e-05, "mean_training_objective": 0.06147048321490347, "mean_relative_U_L2": 0.19737432492815546, "worst_relative_U_L2": 0.5143571023567193, "mean_relative_VM_L2": 0.22758387305274425, "worst_relative_VM_L2": 0.36647208846768925 }, { "step": 1750, "elapsed_seconds": 5485.352037906647, "loss": 0.0514924175105989, "learning_rate": 2.160298066947052e-05, "mean_training_objective": 0.052416468191172974, "mean_relative_U_L2": 0.1653544290018155, "worst_relative_U_L2": 0.512901205692323, "mean_relative_VM_L2": 0.224041782651129, "worst_relative_VM_L2": 0.36528195276626274 }, { "step": 1800, "elapsed_seconds": 5615.651978969574, "loss": 0.05287833777256311, "learning_rate": 1.7462215917663923e-05, "mean_training_objective": 0.04983693387773965, "mean_relative_U_L2": 0.1543024043073392, "worst_relative_U_L2": 0.5112987435113512, "mean_relative_VM_L2": 0.22130664952457285, "worst_relative_VM_L2": 0.36412906466983197 }, { "step": 1850, "elapsed_seconds": 5749.316082000732, "loss": 0.0477710472419858, "learning_rate": 1.4213436473222475e-05, "mean_training_objective": 0.04889097892437019, "mean_relative_U_L2": 0.15383650253631306, "worst_relative_U_L2": 0.5087606375469664, "mean_relative_VM_L2": 0.21784432456734165, "worst_relative_VM_L2": 0.3623383954181486 }, { "step": 1900, "elapsed_seconds": 5878.879971981049, "loss": 0.048287335596978655, "learning_rate": 1.1877711917502638e-05, "mean_training_objective": 0.04731874177131575, "mean_relative_U_L2": 0.14728764936795752, "worst_relative_U_L2": 0.507414230379862, "mean_relative_VM_L2": 0.21509080316219437, "worst_relative_VM_L2": 0.36068474182602295 }, { "step": 1950, "elapsed_seconds": 6011.07030415535, "loss": 0.047092309407889846, "learning_rate": 1.047019032053956e-05, "mean_training_objective": 0.04700699782935587, "mean_relative_U_L2": 0.14905276916598725, "worst_relative_U_L2": 0.5038192221792082, "mean_relative_VM_L2": 0.21328746140527008, "worst_relative_VM_L2": 0.35936269880537275 }, { "step": 2000, "elapsed_seconds": 6143.070252895355, "loss": 0.04575190115720034, "learning_rate": 1e-05, "mean_training_objective": 0.04584610343659522, "mean_relative_U_L2": 0.1442519259266328, "worst_relative_U_L2": 0.5016383400210016, "mean_relative_VM_L2": 0.21128552856737773, "worst_relative_VM_L2": 0.3582272032077047 } ] }, { "variant": "S004", "points": [ { "step": 50, "elapsed_seconds": 127.47364091873169, "loss": 4.5111404180526735, "learning_rate": 0.0003, "mean_training_objective": 3.748453792501757, "mean_relative_U_L2": 1.8358427825345938, "worst_relative_U_L2": 7.4029023679092525, "mean_relative_VM_L2": 0.7034555734243492, "worst_relative_VM_L2": 0.9387812299378842 }, { "step": 100, "elapsed_seconds": 258.74081587791443, "loss": 0.9144223347306254, "learning_rate": 0.00029952980967946045, "mean_training_objective": 1.136092132336011, "mean_relative_U_L2": 1.0391259692264136, "worst_relative_U_L2": 3.7382578376186224, "mean_relative_VM_L2": 0.6253967084808113, "worst_relative_VM_L2": 0.8873396768891042 }, { "step": 150, "elapsed_seconds": 388.10922861099243, "loss": 3.0986712574958797, "learning_rate": 0.0002981222880824973, "mean_training_objective": 3.165236629283352, "mean_relative_U_L2": 1.7328168892939029, "worst_relative_U_L2": 5.823446322477776, "mean_relative_VM_L2": 0.6102250687902477, "worst_relative_VM_L2": 0.8449789314242703 }, { "step": 200, "elapsed_seconds": 518.9672265052795, "loss": 0.5721815302968024, "learning_rate": 0.0002957865635267775, "mean_training_objective": 0.5096982562844456, "mean_relative_U_L2": 0.7142370475905369, "worst_relative_U_L2": 1.9949835395461388, "mean_relative_VM_L2": 0.5003980285412524, "worst_relative_VM_L2": 0.8167692952567833 }, { "step": 250, "elapsed_seconds": 648.2994568347931, "loss": 0.4521676996722818, "learning_rate": 0.00029253778408233605, "mean_training_objective": 0.32984892186082543, "mean_relative_U_L2": 0.5619003075944891, "worst_relative_U_L2": 1.5077197412282732, "mean_relative_VM_L2": 0.40889767797240484, "worst_relative_VM_L2": 0.7404505883088256 }, { "step": 300, "elapsed_seconds": 779.4502210617065, "loss": 1.8738376885652543, "learning_rate": 0.0002883970193305295, "mean_training_objective": 0.9982536095156125, "mean_relative_U_L2": 1.02893554463959, "worst_relative_U_L2": 3.381505671644152, "mean_relative_VM_L2": 0.5715431799588077, "worst_relative_VM_L2": 0.7459748700064469 }, { "step": 350, "elapsed_seconds": 907.3353185653687, "loss": 1.0729691937565804, "learning_rate": 0.0002833911237197154, "mean_training_objective": 0.6516347318923348, "mean_relative_U_L2": 0.8076687569583856, "worst_relative_U_L2": 2.750772852802111, "mean_relative_VM_L2": 0.3928273324031439, "worst_relative_VM_L2": 0.638049239247238 }, { "step": 400, "elapsed_seconds": 1039.5915274620056, "loss": 0.23232358321547505, "learning_rate": 0.00027755256240385024, "mean_training_objective": 0.24276516806680865, "mean_relative_U_L2": 0.48089574407008173, "worst_relative_U_L2": 1.2053184354902742, "mean_relative_VM_L2": 0.35367724484138324, "worst_relative_VM_L2": 0.5377050036002845 }, { "step": 450, "elapsed_seconds": 1168.7973189353943, "loss": 0.21546615362167357, "learning_rate": 0.0002709192006935076, "mean_training_objective": 0.26608645576274104, "mean_relative_U_L2": 0.506456164735541, "worst_relative_U_L2": 1.50155124709507, "mean_relative_VM_L2": 0.3232574833733767, "worst_relative_VM_L2": 0.45305590191878753 }, { "step": 500, "elapsed_seconds": 1295.788866519928, "loss": 0.22876557279378176, "learning_rate": 0.0002635340584848096, "mean_training_objective": 0.21388413978613857, "mean_relative_U_L2": 0.45096143035058905, "worst_relative_U_L2": 1.1797785571467845, "mean_relative_VM_L2": 0.31811874474929447, "worst_relative_VM_L2": 0.4891732265914137 }, { "step": 550, "elapsed_seconds": 1430.5190517902374, "loss": 0.16143481750041244, "learning_rate": 0.0002554450312588919, "mean_training_objective": 0.17551288906313117, "mean_relative_U_L2": 0.4048607607229437, "worst_relative_U_L2": 1.1535085815506174, "mean_relative_VM_L2": 0.2912727561712343, "worst_relative_VM_L2": 0.4125777818472701 }, { "step": 600, "elapsed_seconds": 1559.0274362564087, "loss": 0.14932873351499437, "learning_rate": 0.00024670457946132966, "mean_training_objective": 0.1494522656612615, "mean_relative_U_L2": 0.3667267786211476, "worst_relative_U_L2": 0.9934384455717207, "mean_relative_VM_L2": 0.28496075590983966, "worst_relative_VM_L2": 0.39914009746078777 }, { "step": 650, "elapsed_seconds": 1690.5703027248383, "loss": 0.23936713468283413, "learning_rate": 0.0002373693882760176, "mean_training_objective": 0.3897243583954613, "mean_relative_U_L2": 0.5953206623108426, "worst_relative_U_L2": 2.2279505612435315, "mean_relative_VM_L2": 0.31614495780695356, "worst_relative_VM_L2": 0.4645788962851592 }, { "step": 700, "elapsed_seconds": 1821.612999677658, "loss": 0.20721479505300522, "learning_rate": 0.00022749999999999997, "mean_training_objective": 0.18496116285370884, "mean_relative_U_L2": 0.39965900122658987, "worst_relative_U_L2": 1.3266559996377196, "mean_relative_VM_L2": 0.2803888183539138, "worst_relative_VM_L2": 0.38878789631587685 }, { "step": 750, "elapsed_seconds": 1949.6924324035645, "loss": 0.14164074696600437, "learning_rate": 0.0002171604214034428, "mean_training_objective": 0.17329257035075282, "mean_relative_U_L2": 0.38037069768049026, "worst_relative_U_L2": 1.3036172842215852, "mean_relative_VM_L2": 0.27561007216805045, "worst_relative_VM_L2": 0.385858804741692 }, { "step": 800, "elapsed_seconds": 2082.084330558777, "loss": 0.12966027995571494, "learning_rate": 0.00020641770862116766, "mean_training_objective": 0.13735114786250235, "mean_relative_U_L2": 0.33083975231348456, "worst_relative_U_L2": 1.0773569542985648, "mean_relative_VM_L2": 0.2725393770207754, "worst_relative_VM_L2": 0.3820582508922859 }, { "step": 850, "elapsed_seconds": 2211.344109773636, "loss": 0.11118633523583413, "learning_rate": 0.00019534153226788558, "mean_training_objective": 0.13341279014762192, "mean_relative_U_L2": 0.33903012729658705, "worst_relative_U_L2": 1.025693908341028, "mean_relative_VM_L2": 0.27379579682955796, "worst_relative_VM_L2": 0.3883756621984367 }, { "step": 900, "elapsed_seconds": 2342.391256570816, "loss": 0.1066462797112763, "learning_rate": 0.00018400372559752647, "mean_training_objective": 0.10581392112538623, "mean_relative_U_L2": 0.2836724253352003, "worst_relative_U_L2": 0.8498807762567061, "mean_relative_VM_L2": 0.2660202272744138, "worst_relative_VM_L2": 0.3800237798553163 }, { "step": 950, "elapsed_seconds": 2471.2367825508118, "loss": 0.23294039051979779, "learning_rate": 0.00017247781863702187, "mean_training_objective": 0.11397981612707372, "mean_relative_U_L2": 0.3035917484673369, "worst_relative_U_L2": 0.6177590552941891, "mean_relative_VM_L2": 0.3156594066354366, "worst_relative_VM_L2": 0.6192357850314764 }, { "step": 1000, "elapsed_seconds": 2601.152111053467, "loss": 0.32361583523452286, "learning_rate": 0.0001608385613158652, "mean_training_objective": 0.42562645098956586, "mean_relative_U_L2": 0.641604204949964, "worst_relative_U_L2": 2.064619929113984, "mean_relative_VM_L2": 0.4529209881026549, "worst_relative_VM_L2": 0.8024514679302144 }, { "step": 1050, "elapsed_seconds": 2724.680283308029, "loss": 0.3065861424431205, "learning_rate": 0.00014916143868413484, "mean_training_objective": 0.17344159759298397, "mean_relative_U_L2": 0.41808409144143505, "worst_relative_U_L2": 1.0686388435782914, "mean_relative_VM_L2": 0.3212363183011523, "worst_relative_VM_L2": 0.43111924508017263 }, { "step": 1100, "elapsed_seconds": 2855.574877977371, "loss": 0.08481204817071557, "learning_rate": 0.00013752218136297816, "mean_training_objective": 0.07608586815451913, "mean_relative_U_L2": 0.21096350940886327, "worst_relative_U_L2": 0.5497810571636668, "mean_relative_VM_L2": 0.2772260143728834, "worst_relative_VM_L2": 0.39206515770956385 }, { "step": 1150, "elapsed_seconds": 2986.829628944397, "loss": 0.129347687959671, "learning_rate": 0.00012599627440247358, "mean_training_objective": 0.09755059433928978, "mean_relative_U_L2": 0.2736942665588161, "worst_relative_U_L2": 0.6976117865414528, "mean_relative_VM_L2": 0.28123171104430755, "worst_relative_VM_L2": 0.44157845642759413 }, { "step": 1200, "elapsed_seconds": 3116.144548177719, "loss": 0.09577487651258705, "learning_rate": 0.00011465846773211439, "mean_training_objective": 0.100609695980173, "mean_relative_U_L2": 0.2751271214807186, "worst_relative_U_L2": 0.7300464471024208, "mean_relative_VM_L2": 0.27281921370599466, "worst_relative_VM_L2": 0.41495017863850603 }, { "step": 1250, "elapsed_seconds": 3255.125456094742, "loss": 0.12647038670256733, "learning_rate": 0.00010358229137883234, "mean_training_objective": 0.08074278081123103, "mean_relative_U_L2": 0.22828680815836327, "worst_relative_U_L2": 0.5915530033611708, "mean_relative_VM_L2": 0.26256210441236955, "worst_relative_VM_L2": 0.38329259349463746 }, { "step": 1300, "elapsed_seconds": 3382.358488559723, "loss": 0.07859864742495119, "learning_rate": 9.283957859655712e-05, "mean_training_objective": 0.10095692460750068, "mean_relative_U_L2": 0.2720234818724674, "worst_relative_U_L2": 0.8467355596933234, "mean_relative_VM_L2": 0.2598767711822585, "worst_relative_VM_L2": 0.38264590553994066 }, { "step": 1350, "elapsed_seconds": 3512.942933320999, "loss": 0.06687489207834005, "learning_rate": 8.250000000000003e-05, "mean_training_objective": 0.08003212861927282, "mean_relative_U_L2": 0.22878200313428057, "worst_relative_U_L2": 0.617995173760368, "mean_relative_VM_L2": 0.2573006150004334, "worst_relative_VM_L2": 0.38176735741844015 }, { "step": 1400, "elapsed_seconds": 3642.6796028614044, "loss": 0.09078276250511408, "learning_rate": 7.263061172398237e-05, "mean_training_objective": 0.0683825497660864, "mean_relative_U_L2": 0.19782168310766027, "worst_relative_U_L2": 0.5231882823265248, "mean_relative_VM_L2": 0.2548998482221403, "worst_relative_VM_L2": 0.38106385377187396 }, { "step": 1450, "elapsed_seconds": 3772.8471200466156, "loss": 0.06636449703946709, "learning_rate": 6.32954205386703e-05, "mean_training_objective": 0.07650999806969153, "mean_relative_U_L2": 0.22295639144781002, "worst_relative_U_L2": 0.5954123769300718, "mean_relative_VM_L2": 0.25259183021637377, "worst_relative_VM_L2": 0.3805338899273714 }, { "step": 1500, "elapsed_seconds": 3900.0254068374634, "loss": 0.06013870076276363, "learning_rate": 5.455496874110809e-05, "mean_training_objective": 0.0648121528173361, "mean_relative_U_L2": 0.1897937253432886, "worst_relative_U_L2": 0.5185240064531471, "mean_relative_VM_L2": 0.250206284896919, "worst_relative_VM_L2": 0.37956877898003544 }, { "step": 1550, "elapsed_seconds": 4030.9067842960358, "loss": 0.06706237480975688, "learning_rate": 4.646594151519032e-05, "mean_training_objective": 0.05901187546553446, "mean_relative_U_L2": 0.1687093591040963, "worst_relative_U_L2": 0.5163185428983991, "mean_relative_VM_L2": 0.24781014841253152, "worst_relative_VM_L2": 0.37871193842510753 }, { "step": 1600, "elapsed_seconds": 4159.464012145996, "loss": 0.05838449159637094, "learning_rate": 3.908079930649232e-05, "mean_training_objective": 0.0623517599951555, "mean_relative_U_L2": 0.18587456656772028, "worst_relative_U_L2": 0.5141179414216078, "mean_relative_VM_L2": 0.2454027914937884, "worst_relative_VM_L2": 0.37797566369686564 }, { "step": 1650, "elapsed_seconds": 4289.541310071945, "loss": 0.05507381507195533, "learning_rate": 3.244743759614975e-05, "mean_training_objective": 0.056959205929746445, "mean_relative_U_L2": 0.16573767870837738, "worst_relative_U_L2": 0.5117896097502098, "mean_relative_VM_L2": 0.2428705356576913, "worst_relative_VM_L2": 0.3768992843466317 }, { "step": 1700, "elapsed_seconds": 4418.440986156464, "loss": 0.05781228099949657, "learning_rate": 2.6608876280284544e-05, "mean_training_objective": 0.05444239225805134, "mean_relative_U_L2": 0.1560811467914362, "worst_relative_U_L2": 0.5093309474639636, "mean_relative_VM_L2": 0.2401938207078917, "worst_relative_VM_L2": 0.37589415668417003 }, { "step": 1750, "elapsed_seconds": 4548.72092294693, "loss": 0.055443656863644726, "learning_rate": 2.160298066947052e-05, "mean_training_objective": 0.0559122165643549, "mean_relative_U_L2": 0.16885785725702251, "worst_relative_U_L2": 0.5069728437913997, "mean_relative_VM_L2": 0.23804032961658858, "worst_relative_VM_L2": 0.37486140756644765 }, { "step": 1800, "elapsed_seconds": 4677.462767362595, "loss": 0.05237363418564201, "learning_rate": 1.7462215917663923e-05, "mean_training_objective": 0.0530952635070381, "mean_relative_U_L2": 0.15548802110296395, "worst_relative_U_L2": 0.5032773031665403, "mean_relative_VM_L2": 0.23617789793315008, "worst_relative_VM_L2": 0.3741653569103513 }, { "step": 1850, "elapsed_seconds": 4803.82675909996, "loss": 0.05053789778612554, "learning_rate": 1.4213436473222475e-05, "mean_training_objective": 0.050946113982586846, "mean_relative_U_L2": 0.1482250762846941, "worst_relative_U_L2": 0.5013817510897957, "mean_relative_VM_L2": 0.23232769902486092, "worst_relative_VM_L2": 0.3727653615527865 }, { "step": 1900, "elapsed_seconds": 4931.907533168793, "loss": 0.04964394979178906, "learning_rate": 1.1877711917502638e-05, "mean_training_objective": 0.04965019154112784, "mean_relative_U_L2": 0.14427099552917222, "worst_relative_U_L2": 0.49897289453804977, "mean_relative_VM_L2": 0.2294884917698498, "worst_relative_VM_L2": 0.3717383157903594 }, { "step": 1950, "elapsed_seconds": 5061.409984111786, "loss": 0.04861853779293597, "learning_rate": 1.047019032053956e-05, "mean_training_objective": 0.048762787004826745, "mean_relative_U_L2": 0.1432428468011693, "worst_relative_U_L2": 0.4966528813177644, "mean_relative_VM_L2": 0.2266856876613729, "worst_relative_VM_L2": 0.3706604825289766 }, { "step": 2000, "elapsed_seconds": 5192.525388240814, "loss": 0.04775690557435155, "learning_rate": 1e-05, "mean_training_objective": 0.04779932982927203, "mean_relative_U_L2": 0.14157596762532587, "worst_relative_U_L2": 0.4941444096102204, "mean_relative_VM_L2": 0.22383325720013972, "worst_relative_VM_L2": 0.3694918798409384 } ] }, { "variant": "S008", "points": [ { "step": 50, "elapsed_seconds": 127.17322397232056, "loss": 2.888490588963032, "learning_rate": 0.0003, "mean_training_objective": 1.1902020628999979, "mean_relative_U_L2": 1.1389696043301898, "worst_relative_U_L2": 3.0862535923655687, "mean_relative_VM_L2": 0.7161251874102682, "worst_relative_VM_L2": 0.9487413953297255 }, { "step": 100, "elapsed_seconds": 254.311537027359, "loss": 7.4757161319255845, "learning_rate": 0.00029952980967946045, "mean_training_objective": 3.245589727279738, "mean_relative_U_L2": 1.6452203308913511, "worst_relative_U_L2": 7.084253337477424, "mean_relative_VM_L2": 0.6409889393783741, "worst_relative_VM_L2": 0.8864676305040041 }, { "step": 150, "elapsed_seconds": 387.29032707214355, "loss": 0.42540514878928665, "learning_rate": 0.0002981222880824973, "mean_training_objective": 0.6296265018395042, "mean_relative_U_L2": 0.7773175770791325, "worst_relative_U_L2": 2.3802376597864803, "mean_relative_VM_L2": 0.5498606110666173, "worst_relative_VM_L2": 0.8428124882412776 }, { "step": 200, "elapsed_seconds": 516.3828279972076, "loss": 0.5766274489462376, "learning_rate": 0.0002957865635267775, "mean_training_objective": 0.49226095627957395, "mean_relative_U_L2": 0.6973479719978358, "worst_relative_U_L2": 2.28233820025586, "mean_relative_VM_L2": 0.40143569730504874, "worst_relative_VM_L2": 0.6461959041894344 }, { "step": 250, "elapsed_seconds": 645.6914541721344, "loss": 0.47956685293465845, "learning_rate": 0.00029253778408233605, "mean_training_objective": 0.882609452180813, "mean_relative_U_L2": 0.9991457608809899, "worst_relative_U_L2": 2.9932905581111986, "mean_relative_VM_L2": 0.3904115411708988, "worst_relative_VM_L2": 0.5456623963988309 }, { "step": 300, "elapsed_seconds": 772.6171596050262, "loss": 0.7601324677467346, "learning_rate": 0.0002883970193305295, "mean_training_objective": 1.2749174392821487, "mean_relative_U_L2": 1.0285440366277454, "worst_relative_U_L2": 4.510843708482927, "mean_relative_VM_L2": 0.35450869528116635, "worst_relative_VM_L2": 0.5424207528938325 }, { "step": 350, "elapsed_seconds": 902.1982605457306, "loss": 0.5038176823407412, "learning_rate": 0.0002833911237197154, "mean_training_objective": 0.5887824813591713, "mean_relative_U_L2": 0.8020952653970672, "worst_relative_U_L2": 2.24821723080745, "mean_relative_VM_L2": 0.4523473681248637, "worst_relative_VM_L2": 0.6219774313865144 }, { "step": 400, "elapsed_seconds": 1030.7725791931152, "loss": 0.5509488569572568, "learning_rate": 0.00027755256240385024, "mean_training_objective": 0.2348319600580008, "mean_relative_U_L2": 0.47628251444737224, "worst_relative_U_L2": 1.2631542707984578, "mean_relative_VM_L2": 0.331090876860637, "worst_relative_VM_L2": 0.48275429542376813 }, { "step": 450, "elapsed_seconds": 1156.8118524551392, "loss": 0.3501533973962069, "learning_rate": 0.0002709192006935076, "mean_training_objective": 0.5626365688696154, "mean_relative_U_L2": 0.727999926594283, "worst_relative_U_L2": 2.708323905259378, "mean_relative_VM_L2": 0.31122937897099046, "worst_relative_VM_L2": 0.4553163317996727 }, { "step": 500, "elapsed_seconds": 1289.2403726577759, "loss": 0.19183632740750906, "learning_rate": 0.0002635340584848096, "mean_training_objective": 0.2770096785988506, "mean_relative_U_L2": 0.5249082922968494, "worst_relative_U_L2": 1.6684523477576294, "mean_relative_VM_L2": 0.3009807949186655, "worst_relative_VM_L2": 0.4508335478531261 }, { "step": 550, "elapsed_seconds": 1419.4097170829773, "loss": 0.4322753858752549, "learning_rate": 0.0002554450312588919, "mean_training_objective": 0.2322020243991731, "mean_relative_U_L2": 0.47482442949943743, "worst_relative_U_L2": 1.34956960664341, "mean_relative_VM_L2": 0.3150897429649278, "worst_relative_VM_L2": 0.48252726149146696 }, { "step": 600, "elapsed_seconds": 1546.9826755523682, "loss": 0.19040794158354404, "learning_rate": 0.00024670457946132966, "mean_training_objective": 0.2847577104113902, "mean_relative_U_L2": 0.5100699248779706, "worst_relative_U_L2": 1.7534655985673502, "mean_relative_VM_L2": 0.2867220128789274, "worst_relative_VM_L2": 0.4361429183356932 }, { "step": 650, "elapsed_seconds": 1678.2803871631622, "loss": 0.12857229853980243, "learning_rate": 0.0002373693882760176, "mean_training_objective": 0.20975846146634541, "mean_relative_U_L2": 0.42891664166866805, "worst_relative_U_L2": 1.4323328824952923, "mean_relative_VM_L2": 0.2751666869455776, "worst_relative_VM_L2": 0.40276465899641606 }, { "step": 700, "elapsed_seconds": 1805.7505390644073, "loss": 0.28381707966327663, "learning_rate": 0.00022749999999999997, "mean_training_objective": 0.15528868816264205, "mean_relative_U_L2": 0.3737767198239883, "worst_relative_U_L2": 0.8727000575974758, "mean_relative_VM_L2": 0.30832647272284824, "worst_relative_VM_L2": 0.5049460048541788 }, { "step": 750, "elapsed_seconds": 1936.3359298706055, "loss": 0.1461351855657995, "learning_rate": 0.0002171604214034428, "mean_training_objective": 0.1841897374564599, "mean_relative_U_L2": 0.3903999713358034, "worst_relative_U_L2": 1.2947950237086372, "mean_relative_VM_L2": 0.27378484568732364, "worst_relative_VM_L2": 0.43430275887236874 }, { "step": 800, "elapsed_seconds": 2065.091139793396, "loss": 0.12069933731108905, "learning_rate": 0.00020641770862116766, "mean_training_objective": 0.1398837446094839, "mean_relative_U_L2": 0.3316799996061496, "worst_relative_U_L2": 1.0761998232360088, "mean_relative_VM_L2": 0.2659366743628155, "worst_relative_VM_L2": 0.3922425196117094 }, { "step": 850, "elapsed_seconds": 2192.262484073639, "loss": 0.11106854192912577, "learning_rate": 0.00019534153226788558, "mean_training_objective": 0.10948093041201865, "mean_relative_U_L2": 0.28273677004771214, "worst_relative_U_L2": 0.8792182839671969, "mean_relative_VM_L2": 0.26041908160590743, "worst_relative_VM_L2": 0.38358335476683836 }, { "step": 900, "elapsed_seconds": 2321.6127948760986, "loss": 0.09534197715111076, "learning_rate": 0.00018400372559752647, "mean_training_objective": 0.10057153373750942, "mean_relative_U_L2": 0.27322022766620657, "worst_relative_U_L2": 0.8351630522818569, "mean_relative_VM_L2": 0.2564904520545611, "worst_relative_VM_L2": 0.3823971622347467 }, { "step": 950, "elapsed_seconds": 2452.8723669052124, "loss": 0.16239084526896475, "learning_rate": 0.00017247781863702187, "mean_training_objective": 0.15744192568191342, "mean_relative_U_L2": 0.39072433082337704, "worst_relative_U_L2": 0.9208753561007792, "mean_relative_VM_L2": 0.3206498653204367, "worst_relative_VM_L2": 0.4162106381017435 }, { "step": 1000, "elapsed_seconds": 2582.981183052063, "loss": 0.10871516047045587, "learning_rate": 0.0001608385613158652, "mean_training_objective": 0.11426403383616227, "mean_relative_U_L2": 0.29610195675924167, "worst_relative_U_L2": 0.9717211332603882, "mean_relative_VM_L2": 0.2566817830147742, "worst_relative_VM_L2": 0.3820333263302134 }, { "step": 1050, "elapsed_seconds": 2712.385975599289, "loss": 0.09724138542078438, "learning_rate": 0.00014916143868413484, "mean_training_objective": 0.09158603413815819, "mean_relative_U_L2": 0.2574277235063893, "worst_relative_U_L2": 0.7607972633746913, "mean_relative_VM_L2": 0.24813323880982727, "worst_relative_VM_L2": 0.37952379052311636 }, { "step": 1100, "elapsed_seconds": 2841.51708483696, "loss": 0.07567087272182108, "learning_rate": 0.00013752218136297816, "mean_training_objective": 0.09509390736024939, "mean_relative_U_L2": 0.265143542518489, "worst_relative_U_L2": 0.8267981338847641, "mean_relative_VM_L2": 0.24276800946253987, "worst_relative_VM_L2": 0.37841729653973355 }, { "step": 1150, "elapsed_seconds": 2969.6832599639893, "loss": 0.08100714911706747, "learning_rate": 0.00012599627440247358, "mean_training_objective": 0.07269458352620349, "mean_relative_U_L2": 0.2182831552359143, "worst_relative_U_L2": 0.5620424704002579, "mean_relative_VM_L2": 0.23762392899662954, "worst_relative_VM_L2": 0.3771947708593857 }, { "step": 1200, "elapsed_seconds": 3097.0778930187225, "loss": 0.07241167500615119, "learning_rate": 0.00011465846773211439, "mean_training_objective": 0.0727985998935704, "mean_relative_U_L2": 0.22508028784494835, "worst_relative_U_L2": 0.5681634569080225, "mean_relative_VM_L2": 0.23640786878892484, "worst_relative_VM_L2": 0.3761827134199181 }, { "step": 1250, "elapsed_seconds": 3227.2778899669647, "loss": 0.06591827953234312, "learning_rate": 0.00010358229137883234, "mean_training_objective": 0.0657168205331106, "mean_relative_U_L2": 0.20347379331132903, "worst_relative_U_L2": 0.5522874414914677, "mean_relative_VM_L2": 0.22942843354660059, "worst_relative_VM_L2": 0.3742264067053524 }, { "step": 1300, "elapsed_seconds": 3357.4784309864044, "loss": 0.06077727079391479, "learning_rate": 9.283957859655712e-05, "mean_training_objective": 0.06296459211703387, "mean_relative_U_L2": 0.19813857586715944, "worst_relative_U_L2": 0.547426470515716, "mean_relative_VM_L2": 0.22520125395461096, "worst_relative_VM_L2": 0.3724839327485648 }, { "step": 1350, "elapsed_seconds": 3485.688883304596, "loss": 0.057254203641787176, "learning_rate": 8.250000000000003e-05, "mean_training_objective": 0.059243993002647724, "mean_relative_U_L2": 0.18870522750607194, "worst_relative_U_L2": 0.5427385447198944, "mean_relative_VM_L2": 0.22113443627644264, "worst_relative_VM_L2": 0.3706664234910867 }, { "step": 1400, "elapsed_seconds": 3616.103651046753, "loss": 0.05441266600973904, "learning_rate": 7.263061172398237e-05, "mean_training_objective": 0.05581123635940812, "mean_relative_U_L2": 0.17955008781888307, "worst_relative_U_L2": 0.5379593008386586, "mean_relative_VM_L2": 0.21721688239091944, "worst_relative_VM_L2": 0.36859074795771773 }, { "step": 1450, "elapsed_seconds": 3746.566905975342, "loss": 0.052115212101489304, "learning_rate": 6.32954205386703e-05, "mean_training_objective": 0.05301220920773375, "mean_relative_U_L2": 0.17161674998940624, "worst_relative_U_L2": 0.5332726114833584, "mean_relative_VM_L2": 0.21345080620815007, "worst_relative_VM_L2": 0.36648684560803574 }, { "step": 1500, "elapsed_seconds": 3871.02632188797, "loss": 0.049882724927738306, "learning_rate": 5.455496874110809e-05, "mean_training_objective": 0.05103050711376424, "mean_relative_U_L2": 0.16657005367012043, "worst_relative_U_L2": 0.5278769937964128, "mean_relative_VM_L2": 0.2100099694193703, "worst_relative_VM_L2": 0.3641366681233971 }, { "step": 1550, "elapsed_seconds": 4005.354216814041, "loss": 0.04840149134397507, "learning_rate": 4.646594151519032e-05, "mean_training_objective": 0.04866444973935783, "mean_relative_U_L2": 0.15860483791962054, "worst_relative_U_L2": 0.5226486259763561, "mean_relative_VM_L2": 0.2069850444157506, "worst_relative_VM_L2": 0.3617478091796933 }, { "step": 1600, "elapsed_seconds": 4134.440596818924, "loss": 0.04807284278795123, "learning_rate": 3.908079930649232e-05, "mean_training_objective": 0.047910472610121296, "mean_relative_U_L2": 0.15831802119625146, "worst_relative_U_L2": 0.5190222443048227, "mean_relative_VM_L2": 0.20787601637408634, "worst_relative_VM_L2": 0.3604794343114375 }, { "step": 1650, "elapsed_seconds": 4263.55733704567, "loss": 0.048241162812337285, "learning_rate": 3.244743759614975e-05, "mean_training_objective": 0.04852292064109642, "mean_relative_U_L2": 0.16718548100649486, "worst_relative_U_L2": 0.5138549178635154, "mean_relative_VM_L2": 0.20694532607120603, "worst_relative_VM_L2": 0.36052614540511563 }, { "step": 1700, "elapsed_seconds": 4390.78746843338, "loss": 0.04564168178476393, "learning_rate": 2.6608876280284544e-05, "mean_training_objective": 0.0453368666046148, "mean_relative_U_L2": 0.1477004713077556, "worst_relative_U_L2": 0.5097849957255379, "mean_relative_VM_L2": 0.20233221028514067, "worst_relative_VM_L2": 0.3570706778928518 }, { "step": 1750, "elapsed_seconds": 4517.838570356369, "loss": 0.044210850959643715, "learning_rate": 2.160298066947052e-05, "mean_training_objective": 0.04402254672677025, "mean_relative_U_L2": 0.1424921944314752, "worst_relative_U_L2": 0.5056766767115831, "mean_relative_VM_L2": 0.20067532424932502, "worst_relative_VM_L2": 0.35490528565317353 }, { "step": 1800, "elapsed_seconds": 4642.821153640747, "loss": 0.04323487696237861, "learning_rate": 1.7462215917663923e-05, "mean_training_objective": 0.04310180187796289, "mean_relative_U_L2": 0.139435997188421, "worst_relative_U_L2": 0.5014937601269557, "mean_relative_VM_L2": 0.19912527262332247, "worst_relative_VM_L2": 0.3528547002711068 }, { "step": 1850, "elapsed_seconds": 4778.296600818634, "loss": 0.042426899867132314, "learning_rate": 1.4213436473222475e-05, "mean_training_objective": 0.0423443110099065, "mean_relative_U_L2": 0.137663123950295, "worst_relative_U_L2": 0.49727206827038156, "mean_relative_VM_L2": 0.19762490482985395, "worst_relative_VM_L2": 0.35120948612671066 }, { "step": 1900, "elapsed_seconds": 4908.841392993927, "loss": 0.04143998217768967, "learning_rate": 1.1877711917502638e-05, "mean_training_objective": 0.04195505053796743, "mean_relative_U_L2": 0.139761938574243, "worst_relative_U_L2": 0.49302753485169765, "mean_relative_VM_L2": 0.1962481097170451, "worst_relative_VM_L2": 0.34981729758920455 }, { "step": 1950, "elapsed_seconds": 5036.748710632324, "loss": 0.04355704784393311, "learning_rate": 1.047019032053956e-05, "mean_training_objective": 0.041409422977111, "mean_relative_U_L2": 0.13696327505299905, "worst_relative_U_L2": 0.4906028760839966, "mean_relative_VM_L2": 0.19588205216181173, "worst_relative_VM_L2": 0.34942264414205215 }, { "step": 2000, "elapsed_seconds": 5164.919672012329, "loss": 0.04078074404969811, "learning_rate": 1e-05, "mean_training_objective": 0.04083708295121814, "mean_relative_U_L2": 0.13400675251909946, "worst_relative_U_L2": 0.48647867100160214, "mean_relative_VM_L2": 0.19521377020706027, "worst_relative_VM_L2": 0.3487050221982171 } ] }, { "variant": "S016", "points": [ { "step": 50, "elapsed_seconds": 129.75081181526184, "loss": 3.580676904320717, "learning_rate": 0.0003, "mean_training_objective": 7.575577067406942, "mean_relative_U_L2": 2.471743234499308, "worst_relative_U_L2": 11.244262123608348, "mean_relative_VM_L2": 0.7109431966040776, "worst_relative_VM_L2": 0.9279844694521316 }, { "step": 100, "elapsed_seconds": 261.298828125, "loss": 1.2138869650661945, "learning_rate": 0.00029952980967946045, "mean_training_objective": 1.5502267350000032, "mean_relative_U_L2": 1.2969080518944873, "worst_relative_U_L2": 4.0477239543079, "mean_relative_VM_L2": 0.7038706701705745, "worst_relative_VM_L2": 0.9486682758553093 }, { "step": 150, "elapsed_seconds": 389.82175517082214, "loss": 0.6229446530342102, "learning_rate": 0.0002981222880824973, "mean_training_objective": 3.7027958635328693, "mean_relative_U_L2": 1.8656071084950618, "worst_relative_U_L2": 7.3196472591553485, "mean_relative_VM_L2": 0.6920787207017016, "worst_relative_VM_L2": 0.9109716879637176 }, { "step": 200, "elapsed_seconds": 521.131974697113, "loss": 0.5231303829699756, "learning_rate": 0.0002957865635267775, "mean_training_objective": 0.6279426226005952, "mean_relative_U_L2": 0.8011367018666566, "worst_relative_U_L2": 1.9541051650662122, "mean_relative_VM_L2": 0.6203906720755055, "worst_relative_VM_L2": 0.8492954157442492 }, { "step": 250, "elapsed_seconds": 651.1593656539917, "loss": 0.46573294922709463, "learning_rate": 0.00029253778408233605, "mean_training_objective": 0.834559910615741, "mean_relative_U_L2": 0.9844138218803368, "worst_relative_U_L2": 2.833763848871366, "mean_relative_VM_L2": 0.5442249871966082, "worst_relative_VM_L2": 0.8230235331771134 }, { "step": 300, "elapsed_seconds": 782.4058139324188, "loss": 0.3028106451034545, "learning_rate": 0.0002883970193305295, "mean_training_objective": 0.6916717691509434, "mean_relative_U_L2": 0.8510685089349014, "worst_relative_U_L2": 2.7885098172141896, "mean_relative_VM_L2": 0.4172263359538884, "worst_relative_VM_L2": 0.7698347773891173 }, { "step": 350, "elapsed_seconds": 910.3782515525818, "loss": 0.3216007027775049, "learning_rate": 0.0002833911237197154, "mean_training_objective": 0.43088848235066646, "mean_relative_U_L2": 0.6535778261176233, "worst_relative_U_L2": 2.034376716558616, "mean_relative_VM_L2": 0.3888877259818396, "worst_relative_VM_L2": 0.7507823611686734 }, { "step": 400, "elapsed_seconds": 1041.4516820907593, "loss": 0.32084273491054777, "learning_rate": 0.00027755256240385024, "mean_training_objective": 0.29680068196084197, "mean_relative_U_L2": 0.5400729320735789, "worst_relative_U_L2": 1.510846668795786, "mean_relative_VM_L2": 0.36803033243117944, "worst_relative_VM_L2": 0.6488362206921985 }, { "step": 450, "elapsed_seconds": 1173.0446548461914, "loss": 0.2203962763771415, "learning_rate": 0.0002709192006935076, "mean_training_objective": 0.3307597674635912, "mean_relative_U_L2": 0.5841439440845674, "worst_relative_U_L2": 1.749488223168945, "mean_relative_VM_L2": 0.3534605633045026, "worst_relative_VM_L2": 0.578323306412672 }, { "step": 500, "elapsed_seconds": 1303.833659172058, "loss": 0.9790562067180871, "learning_rate": 0.0002635340584848096, "mean_training_objective": 1.551439546226512, "mean_relative_U_L2": 1.1330153583370266, "worst_relative_U_L2": 4.885709942493034, "mean_relative_VM_L2": 0.35434546376658094, "worst_relative_VM_L2": 0.5075239861498219 }, { "step": 550, "elapsed_seconds": 1433.1610443592072, "loss": 0.6728323521092534, "learning_rate": 0.0002554450312588919, "mean_training_objective": 0.5050670122650993, "mean_relative_U_L2": 0.6854885961548678, "worst_relative_U_L2": 2.558048099296737, "mean_relative_VM_L2": 0.3254924601661343, "worst_relative_VM_L2": 0.4737110518814835 }, { "step": 600, "elapsed_seconds": 1560.1924068927765, "loss": 0.7064588446170093, "learning_rate": 0.00024670457946132966, "mean_training_objective": 0.33142609036117787, "mean_relative_U_L2": 0.5952485055953065, "worst_relative_U_L2": 1.642965471176387, "mean_relative_VM_L2": 0.37902982188559775, "worst_relative_VM_L2": 0.5369998136873677 }, { "step": 650, "elapsed_seconds": 1692.1271991729736, "loss": 0.22428971640765663, "learning_rate": 0.0002373693882760176, "mean_training_objective": 0.17731038394639254, "mean_relative_U_L2": 0.3987206103848783, "worst_relative_U_L2": 1.0500650680259336, "mean_relative_VM_L2": 0.32911011064805984, "worst_relative_VM_L2": 0.44438224123551295 }, { "step": 700, "elapsed_seconds": 1821.5864081382751, "loss": 0.18948574438691138, "learning_rate": 0.00022749999999999997, "mean_training_objective": 0.16861026077695182, "mean_relative_U_L2": 0.39275117673544646, "worst_relative_U_L2": 1.0958886905426788, "mean_relative_VM_L2": 0.3085956139265947, "worst_relative_VM_L2": 0.42895239053072354 }, { "step": 750, "elapsed_seconds": 1949.2483904361725, "loss": 0.14767606118693946, "learning_rate": 0.0002171604214034428, "mean_training_objective": 0.1351950546082012, "mean_relative_U_L2": 0.33328432246197576, "worst_relative_U_L2": 0.9378348406107346, "mean_relative_VM_L2": 0.2939209076727708, "worst_relative_VM_L2": 0.4107689033747507 }, { "step": 800, "elapsed_seconds": 2074.158988237381, "loss": 0.13894756864756347, "learning_rate": 0.00020641770862116766, "mean_training_objective": 0.12760087359931743, "mean_relative_U_L2": 0.32483102397380537, "worst_relative_U_L2": 0.904926479768789, "mean_relative_VM_L2": 0.28790008836099557, "worst_relative_VM_L2": 0.40675855129798627 }, { "step": 850, "elapsed_seconds": 2210.4412682056427, "loss": 0.12225043736398218, "learning_rate": 0.00019534153226788558, "mean_training_objective": 0.11321301394351554, "mean_relative_U_L2": 0.29668411588248805, "worst_relative_U_L2": 0.8167489740948825, "mean_relative_VM_L2": 0.28122428417110745, "worst_relative_VM_L2": 0.38875087520683127 }, { "step": 900, "elapsed_seconds": 2336.557732820511, "loss": 0.10714612035080791, "learning_rate": 0.00018400372559752647, "mean_training_objective": 0.09946266255329114, "mean_relative_U_L2": 0.2737754797338285, "worst_relative_U_L2": 0.7165241316542452, "mean_relative_VM_L2": 0.27723625216525916, "worst_relative_VM_L2": 0.38290816455376925 }, { "step": 950, "elapsed_seconds": 2468.359470129013, "loss": 0.1115085059776902, "learning_rate": 0.00017247781863702187, "mean_training_objective": 0.10865895484109198, "mean_relative_U_L2": 0.28875397160561767, "worst_relative_U_L2": 0.7816961025410104, "mean_relative_VM_L2": 0.27972132742837635, "worst_relative_VM_L2": 0.4090615623689592 }, { "step": 1000, "elapsed_seconds": 2597.2925939559937, "loss": 0.14055289411917327, "learning_rate": 0.0001608385613158652, "mean_training_objective": 0.12451976924019947, "mean_relative_U_L2": 0.3172041929713569, "worst_relative_U_L2": 1.0109460488186635, "mean_relative_VM_L2": 0.27026454239089975, "worst_relative_VM_L2": 0.3800888574550105 }, { "step": 1050, "elapsed_seconds": 2726.6822118759155, "loss": 0.12205930063501001, "learning_rate": 0.00014916143868413484, "mean_training_objective": 0.1106639228875794, "mean_relative_U_L2": 0.29327385947735285, "worst_relative_U_L2": 0.9055316105633816, "mean_relative_VM_L2": 0.2664832997073107, "worst_relative_VM_L2": 0.3793824516723603 }, { "step": 1100, "elapsed_seconds": 2856.632980823517, "loss": 0.11048722518607976, "learning_rate": 0.00013752218136297816, "mean_training_objective": 0.10121374059650656, "mean_relative_U_L2": 0.27698558539287943, "worst_relative_U_L2": 0.8243630639574286, "mean_relative_VM_L2": 0.2641742850229706, "worst_relative_VM_L2": 0.378694083552363 }, { "step": 1150, "elapsed_seconds": 2984.582623243332, "loss": 0.09913996867835523, "learning_rate": 0.00012599627440247358, "mean_training_objective": 0.09227683277101348, "mean_relative_U_L2": 0.2595491067673108, "worst_relative_U_L2": 0.7318604827077516, "mean_relative_VM_L2": 0.26163378363264594, "worst_relative_VM_L2": 0.3779989788693076 }, { "step": 1200, "elapsed_seconds": 3116.6070404052734, "loss": 0.08775026379153132, "learning_rate": 0.00011465846773211439, "mean_training_objective": 0.08228552660602832, "mean_relative_U_L2": 0.238971438314251, "worst_relative_U_L2": 0.6267813207596242, "mean_relative_VM_L2": 0.25877180961262675, "worst_relative_VM_L2": 0.3770779716437189 }, { "step": 1250, "elapsed_seconds": 3244.3487372398376, "loss": 0.08436467582359909, "learning_rate": 0.00010358229137883234, "mean_training_objective": 0.07981138348523328, "mean_relative_U_L2": 0.23315582182564448, "worst_relative_U_L2": 0.6050247427851986, "mean_relative_VM_L2": 0.2560786839489179, "worst_relative_VM_L2": 0.3760692639779202 }, { "step": 1300, "elapsed_seconds": 3375.463374853134, "loss": 0.08962533809244633, "learning_rate": 9.283957859655712e-05, "mean_training_objective": 0.07737730452847975, "mean_relative_U_L2": 0.22540111356085485, "worst_relative_U_L2": 0.5805446863359092, "mean_relative_VM_L2": 0.2560249767458996, "worst_relative_VM_L2": 0.37483011271375366 }, { "step": 1350, "elapsed_seconds": 3505.822018623352, "loss": 0.07033880222588777, "learning_rate": 8.250000000000003e-05, "mean_training_objective": 0.0681286081964357, "mean_relative_U_L2": 0.2044743610302418, "worst_relative_U_L2": 0.5278270898336956, "mean_relative_VM_L2": 0.2506967523533354, "worst_relative_VM_L2": 0.37368318514570253 }, { "step": 1400, "elapsed_seconds": 3635.043041944504, "loss": 0.06618916150182487, "learning_rate": 7.263061172398237e-05, "mean_training_objective": 0.0643557835571689, "mean_relative_U_L2": 0.1953167706779776, "worst_relative_U_L2": 0.5245390264555551, "mean_relative_VM_L2": 0.24660269644420368, "worst_relative_VM_L2": 0.37228193890337397 }, { "step": 1450, "elapsed_seconds": 3766.09392952919, "loss": 0.06639860412105916, "learning_rate": 6.32954205386703e-05, "mean_training_objective": 0.06421100092254277, "mean_relative_U_L2": 0.19667304578464737, "worst_relative_U_L2": 0.5317141495547402, "mean_relative_VM_L2": 0.24429624359516455, "worst_relative_VM_L2": 0.37104808508452014 }, { "step": 1500, "elapsed_seconds": 3895.5968623161316, "loss": 0.06079560914076865, "learning_rate": 5.455496874110809e-05, "mean_training_objective": 0.05956331046492721, "mean_relative_U_L2": 0.18012404380911776, "worst_relative_U_L2": 0.5193377358964418, "mean_relative_VM_L2": 0.2414722973025988, "worst_relative_VM_L2": 0.36976272639077457 }, { "step": 1550, "elapsed_seconds": 4023.368581533432, "loss": 0.05725652608089149, "learning_rate": 4.646594151519032e-05, "mean_training_objective": 0.05637896954796212, "mean_relative_U_L2": 0.17161179771666807, "worst_relative_U_L2": 0.5170994070052898, "mean_relative_VM_L2": 0.23656148439160646, "worst_relative_VM_L2": 0.36840624431865976 }, { "step": 1600, "elapsed_seconds": 4154.039636135101, "loss": 0.05471747820265591, "learning_rate": 3.908079930649232e-05, "mean_training_objective": 0.05417870606994259, "mean_relative_U_L2": 0.16544952371542257, "worst_relative_U_L2": 0.5145627082144745, "mean_relative_VM_L2": 0.23275505874814645, "worst_relative_VM_L2": 0.36696818384819724 }, { "step": 1650, "elapsed_seconds": 4284.38018321991, "loss": 0.05316787557676434, "learning_rate": 3.244743759614975e-05, "mean_training_objective": 0.052308951394674584, "mean_relative_U_L2": 0.15952512279889913, "worst_relative_U_L2": 0.5125459714805258, "mean_relative_VM_L2": 0.22954177791937386, "worst_relative_VM_L2": 0.36559191320265627 }, { "step": 1700, "elapsed_seconds": 4412.049046039581, "loss": 0.05081545803695917, "learning_rate": 2.6608876280284544e-05, "mean_training_objective": 0.05053780849026063, "mean_relative_U_L2": 0.15387404529772591, "worst_relative_U_L2": 0.5087613812241166, "mean_relative_VM_L2": 0.2262904737600774, "worst_relative_VM_L2": 0.36431149314819106 }, { "step": 1750, "elapsed_seconds": 4542.963385820389, "loss": 0.05039173653349281, "learning_rate": 2.160298066947052e-05, "mean_training_objective": 0.05031502338868765, "mean_relative_U_L2": 0.15532135617033946, "worst_relative_U_L2": 0.5087141062044588, "mean_relative_VM_L2": 0.2254647939193209, "worst_relative_VM_L2": 0.36334146322926797 }, { "step": 1800, "elapsed_seconds": 4671.880819797516, "loss": 0.04882182204164565, "learning_rate": 1.7462215917663923e-05, "mean_training_objective": 0.048673528215576224, "mean_relative_U_L2": 0.15017007179914282, "worst_relative_U_L2": 0.5032321193437747, "mean_relative_VM_L2": 0.22150137564994243, "worst_relative_VM_L2": 0.362193312590239 }, { "step": 1850, "elapsed_seconds": 4800.96030330658, "loss": 0.04762083985842765, "learning_rate": 1.4213436473222475e-05, "mean_training_objective": 0.047532535625861155, "mean_relative_U_L2": 0.14697941519225569, "worst_relative_U_L2": 0.5009278406478063, "mean_relative_VM_L2": 0.21883217504008212, "worst_relative_VM_L2": 0.36128731628781025 }, { "step": 1900, "elapsed_seconds": 4931.458275079727, "loss": 0.04654569597914814, "learning_rate": 1.1877711917502638e-05, "mean_training_objective": 0.046483085292339, "mean_relative_U_L2": 0.14462297884069833, "worst_relative_U_L2": 0.4986359667030497, "mean_relative_VM_L2": 0.215951424308186, "worst_relative_VM_L2": 0.3603810804449936 }, { "step": 1950, "elapsed_seconds": 5062.266132593155, "loss": 0.04567444035783411, "learning_rate": 1.047019032053956e-05, "mean_training_objective": 0.045670845470524704, "mean_relative_U_L2": 0.14346423109072032, "worst_relative_U_L2": 0.4963344043887045, "mean_relative_VM_L2": 0.21336365322002485, "worst_relative_VM_L2": 0.35950254261599257 }, { "step": 2000, "elapsed_seconds": 5190.053691864014, "loss": 0.04542811377905309, "learning_rate": 1e-05, "mean_training_objective": 0.04522835525197016, "mean_relative_U_L2": 0.14337526767621211, "worst_relative_U_L2": 0.49483005149223674, "mean_relative_VM_L2": 0.211924326468715, "worst_relative_VM_L2": 0.35875370086811637 } ] }, { "variant": "S032", "points": [ { "step": 50, "elapsed_seconds": 135.424414396286, "loss": 12.195220711827277, "learning_rate": 0.0003, "mean_training_objective": 2.964616181277523, "mean_relative_U_L2": 1.6919470381172217, "worst_relative_U_L2": 6.074741229045488, "mean_relative_VM_L2": 0.7058406146987087, "worst_relative_VM_L2": 0.9429412987510922 }, { "step": 100, "elapsed_seconds": 268.9628086090088, "loss": 1.0448783613741397, "learning_rate": 0.00029952980967946045, "mean_training_objective": 0.6856735325287759, "mean_relative_U_L2": 0.8664279516057924, "worst_relative_U_L2": 1.9531925042415346, "mean_relative_VM_L2": 0.6494781718875958, "worst_relative_VM_L2": 0.904795150386059 }, { "step": 150, "elapsed_seconds": 404.9490213394165, "loss": 1.455021657049656, "learning_rate": 0.0002981222880824973, "mean_training_objective": 1.8216266079460937, "mean_relative_U_L2": 1.2639064217928877, "worst_relative_U_L2": 5.129611238404301, "mean_relative_VM_L2": 0.5701545527005683, "worst_relative_VM_L2": 0.8420804038178986 }, { "step": 200, "elapsed_seconds": 537.5612869262695, "loss": 0.31559976171702137, "learning_rate": 0.0002957865635267775, "mean_training_objective": 0.5770729439642331, "mean_relative_U_L2": 0.7852255290922943, "worst_relative_U_L2": 2.269890642447186, "mean_relative_VM_L2": 0.42862755357886567, "worst_relative_VM_L2": 0.7375250573679344 }, { "step": 250, "elapsed_seconds": 673.8498566150665, "loss": 0.8181314416229724, "learning_rate": 0.00029253778408233605, "mean_training_objective": 0.7409698229295271, "mean_relative_U_L2": 0.8873081051998758, "worst_relative_U_L2": 1.4550072008450712, "mean_relative_VM_L2": 0.7789095615814198, "worst_relative_VM_L2": 0.9539104969094088 }, { "step": 300, "elapsed_seconds": 805.8735764026642, "loss": 0.5438422571867705, "learning_rate": 0.0002883970193305295, "mean_training_objective": 0.6676502747382793, "mean_relative_U_L2": 0.8427042392130127, "worst_relative_U_L2": 2.452279764170299, "mean_relative_VM_L2": 0.5369597333608043, "worst_relative_VM_L2": 0.7946734083281002 }, { "step": 350, "elapsed_seconds": 942.2297360897064, "loss": 0.4820216249674559, "learning_rate": 0.0002833911237197154, "mean_training_objective": 0.8584081487225996, "mean_relative_U_L2": 0.9559578372882209, "worst_relative_U_L2": 3.185489863173293, "mean_relative_VM_L2": 0.4234259578755136, "worst_relative_VM_L2": 0.6378038816561008 }, { "step": 400, "elapsed_seconds": 1075.9332749843597, "loss": 0.7797899097204207, "learning_rate": 0.00027755256240385024, "mean_training_objective": 1.357394418660462, "mean_relative_U_L2": 1.1438093847627484, "worst_relative_U_L2": 4.2504032120309665, "mean_relative_VM_L2": 0.4028573673681036, "worst_relative_VM_L2": 0.7606744126345341 }, { "step": 450, "elapsed_seconds": 1210.919175863266, "loss": 0.27361058248206976, "learning_rate": 0.0002709192006935076, "mean_training_objective": 0.46309535375777156, "mean_relative_U_L2": 0.6791485431454835, "worst_relative_U_L2": 2.2288819173374628, "mean_relative_VM_L2": 0.3479644366872993, "worst_relative_VM_L2": 0.6038243133296995 }, { "step": 500, "elapsed_seconds": 1345.9736194610596, "loss": 0.25219249445945024, "learning_rate": 0.0002635340584848096, "mean_training_objective": 0.25889408942985415, "mean_relative_U_L2": 0.48990040594993145, "worst_relative_U_L2": 1.5097352119921863, "mean_relative_VM_L2": 0.3233838010219511, "worst_relative_VM_L2": 0.5397076307800772 }, { "step": 550, "elapsed_seconds": 1478.4916965961456, "loss": 0.20447823870927098, "learning_rate": 0.0002554450312588919, "mean_training_objective": 0.2212140162120991, "mean_relative_U_L2": 0.4604111403778354, "worst_relative_U_L2": 1.372974352006688, "mean_relative_VM_L2": 0.31045769086854536, "worst_relative_VM_L2": 0.46545451864938525 }, { "step": 600, "elapsed_seconds": 1615.7122371196747, "loss": 0.17425315491855142, "learning_rate": 0.00024670457946132966, "mean_training_objective": 0.18408141664415853, "mean_relative_U_L2": 0.41281034566569386, "worst_relative_U_L2": 1.1970196967088755, "mean_relative_VM_L2": 0.2984837858456427, "worst_relative_VM_L2": 0.405271231553049 }, { "step": 650, "elapsed_seconds": 1750.1756298542023, "loss": 0.33105936683714393, "learning_rate": 0.0002373693882760176, "mean_training_objective": 0.19728265494580882, "mean_relative_U_L2": 0.4517776039336301, "worst_relative_U_L2": 0.9236794050764873, "mean_relative_VM_L2": 0.33988805598908356, "worst_relative_VM_L2": 0.46239469796673094 }, { "step": 700, "elapsed_seconds": 1882.467379808426, "loss": 0.15242958022281528, "learning_rate": 0.00022749999999999997, "mean_training_objective": 0.42434104173966924, "mean_relative_U_L2": 0.6513973639276487, "worst_relative_U_L2": 2.1891413778447943, "mean_relative_VM_L2": 0.33789573251692584, "worst_relative_VM_L2": 0.6230386438459719 }, { "step": 750, "elapsed_seconds": 2014.9248797893524, "loss": 0.29153419155627486, "learning_rate": 0.0002171604214034428, "mean_training_objective": 0.25011702765889166, "mean_relative_U_L2": 0.5318464633641873, "worst_relative_U_L2": 1.2155114689160644, "mean_relative_VM_L2": 0.3410484063779647, "worst_relative_VM_L2": 0.5423112566442848 }, { "step": 800, "elapsed_seconds": 2153.292413711548, "loss": 0.36557070612907405, "learning_rate": 0.00020641770862116766, "mean_training_objective": 0.27136702661205075, "mean_relative_U_L2": 0.5701072011888109, "worst_relative_U_L2": 1.206822726845334, "mean_relative_VM_L2": 0.39864370795445175, "worst_relative_VM_L2": 0.5368427314231775 }, { "step": 850, "elapsed_seconds": 2288.3990144729614, "loss": 0.1560046640224755, "learning_rate": 0.00019534153226788558, "mean_training_objective": 0.13777357946889787, "mean_relative_U_L2": 0.36495191647041364, "worst_relative_U_L2": 0.8193433516387676, "mean_relative_VM_L2": 0.29383499226926274, "worst_relative_VM_L2": 0.40640841329074584 }, { "step": 900, "elapsed_seconds": 2421.0872719287872, "loss": 0.10147722410038112, "learning_rate": 0.00018400372559752647, "mean_training_objective": 0.11190956896110568, "mean_relative_U_L2": 0.29732510095668474, "worst_relative_U_L2": 0.722599865970103, "mean_relative_VM_L2": 0.28163442384177745, "worst_relative_VM_L2": 0.40893298048784527 }, { "step": 950, "elapsed_seconds": 2556.3046686649323, "loss": 0.08750225491821766, "learning_rate": 0.00017247781863702187, "mean_training_objective": 0.09370415404004319, "mean_relative_U_L2": 0.25487416991836914, "worst_relative_U_L2": 0.6747138482248297, "mean_relative_VM_L2": 0.273889173525043, "worst_relative_VM_L2": 0.38352633972651196 }, { "step": 1000, "elapsed_seconds": 2691.685780286789, "loss": 0.08295822609215975, "learning_rate": 0.0001608385613158652, "mean_training_objective": 0.08565739118837765, "mean_relative_U_L2": 0.2381329030545994, "worst_relative_U_L2": 0.6056542702595589, "mean_relative_VM_L2": 0.2700046755716358, "worst_relative_VM_L2": 0.3804355232325456 }, { "step": 1050, "elapsed_seconds": 2826.3357243537903, "loss": 0.08123063491657376, "learning_rate": 0.00014916143868413484, "mean_training_objective": 0.0791687061423223, "mean_relative_U_L2": 0.2228169290909059, "worst_relative_U_L2": 0.5528046130516888, "mean_relative_VM_L2": 0.2661057564689161, "worst_relative_VM_L2": 0.3787724309992188 }, { "step": 1100, "elapsed_seconds": 2960.4426925182343, "loss": 0.11891473392024636, "learning_rate": 0.00013752218136297816, "mean_training_objective": 0.13080080347757955, "mean_relative_U_L2": 0.34925413641745495, "worst_relative_U_L2": 0.7892484565923515, "mean_relative_VM_L2": 0.2814788049968665, "worst_relative_VM_L2": 0.4736212582616792 }, { "step": 1150, "elapsed_seconds": 3095.288285970688, "loss": 0.08686991175636649, "learning_rate": 0.00012599627440247358, "mean_training_objective": 0.08834383258887449, "mean_relative_U_L2": 0.2476851565672152, "worst_relative_U_L2": 0.6911065058752746, "mean_relative_VM_L2": 0.2635987174721209, "worst_relative_VM_L2": 0.3768384388459262 }, { "step": 1200, "elapsed_seconds": 3229.3235006332397, "loss": 0.0785286127589643, "learning_rate": 0.00011465846773211439, "mean_training_objective": 0.0813996559107914, "mean_relative_U_L2": 0.2323181246264347, "worst_relative_U_L2": 0.6282732149429724, "mean_relative_VM_L2": 0.25857460450595643, "worst_relative_VM_L2": 0.37474538793412276 }, { "step": 1250, "elapsed_seconds": 3359.5334420204163, "loss": 0.10170157970860598, "learning_rate": 0.00010358229137883234, "mean_training_objective": 0.07876490989399625, "mean_relative_U_L2": 0.2366794791423513, "worst_relative_U_L2": 0.5910741574761592, "mean_relative_VM_L2": 0.25643430633896325, "worst_relative_VM_L2": 0.37291206483153333 }, { "step": 1300, "elapsed_seconds": 3497.882667541504, "loss": 0.14510974036529659, "learning_rate": 9.283957859655712e-05, "mean_training_objective": 0.0906016208712707, "mean_relative_U_L2": 0.25338179549626416, "worst_relative_U_L2": 0.758599454394297, "mean_relative_VM_L2": 0.2564958763401055, "worst_relative_VM_L2": 0.37330244100428833 }, { "step": 1350, "elapsed_seconds": 3634.3210003376007, "loss": 0.06342796413227916, "learning_rate": 8.250000000000003e-05, "mean_training_objective": 0.09147207563984097, "mean_relative_U_L2": 0.25521628581665023, "worst_relative_U_L2": 0.7653746674918067, "mean_relative_VM_L2": 0.2553991354297184, "worst_relative_VM_L2": 0.37016908186352787 }, { "step": 1400, "elapsed_seconds": 3766.7607543468475, "loss": 0.07727114791050553, "learning_rate": 7.263061172398237e-05, "mean_training_objective": 0.0755794134595476, "mean_relative_U_L2": 0.22467920385589743, "worst_relative_U_L2": 0.6128333336316178, "mean_relative_VM_L2": 0.24660005542422417, "worst_relative_VM_L2": 0.369033467719531 }, { "step": 1450, "elapsed_seconds": 3902.92391872406, "loss": 0.11216420345008372, "learning_rate": 6.32954205386703e-05, "mean_training_objective": 0.09984938016581814, "mean_relative_U_L2": 0.3012762127277414, "worst_relative_U_L2": 0.7017212918589064, "mean_relative_VM_L2": 0.26724829505055914, "worst_relative_VM_L2": 0.40516239007287735 }, { "step": 1500, "elapsed_seconds": 4037.669469833374, "loss": 0.09062657495960592, "learning_rate": 5.455496874110809e-05, "mean_training_objective": 0.07791892653948432, "mean_relative_U_L2": 0.24486505674754264, "worst_relative_U_L2": 0.5459425495120025, "mean_relative_VM_L2": 0.25474418928914017, "worst_relative_VM_L2": 0.38768219239928364 }, { "step": 1550, "elapsed_seconds": 4171.217752695084, "loss": 0.053989971661940216, "learning_rate": 4.646594151519032e-05, "mean_training_objective": 0.06155495839925518, "mean_relative_U_L2": 0.19059083319191605, "worst_relative_U_L2": 0.5113295395884604, "mean_relative_VM_L2": 0.2374880737417809, "worst_relative_VM_L2": 0.36765978402398014 }, { "step": 1600, "elapsed_seconds": 4305.27675318718, "loss": 0.053829326992854476, "learning_rate": 3.908079930649232e-05, "mean_training_objective": 0.05173204936039327, "mean_relative_U_L2": 0.15397606360461324, "worst_relative_U_L2": 0.5085010364664988, "mean_relative_VM_L2": 0.23126805311098653, "worst_relative_VM_L2": 0.36571566468339933 }, { "step": 1650, "elapsed_seconds": 4442.898301362991, "loss": 0.05096699646674096, "learning_rate": 3.244743759614975e-05, "mean_training_objective": 0.049754552158257984, "mean_relative_U_L2": 0.14796735398881908, "worst_relative_U_L2": 0.5046779672449087, "mean_relative_VM_L2": 0.22702828573800948, "worst_relative_VM_L2": 0.36428247161299265 }, { "step": 1700, "elapsed_seconds": 4576.7877452373505, "loss": 0.051167679950594895, "learning_rate": 2.6608876280284544e-05, "mean_training_objective": 0.050896231856801935, "mean_relative_U_L2": 0.15982164762443699, "worst_relative_U_L2": 0.5017078701826713, "mean_relative_VM_L2": 0.22418051872708933, "worst_relative_VM_L2": 0.3633037349445788 }, { "step": 1750, "elapsed_seconds": 4709.00955080986, "loss": 0.05346709946170449, "learning_rate": 2.160298066947052e-05, "mean_training_objective": 0.0493294924302678, "mean_relative_U_L2": 0.1500986335892641, "worst_relative_U_L2": 0.5003310834489697, "mean_relative_VM_L2": 0.22584997490935743, "worst_relative_VM_L2": 0.36327777774885034 }, { "step": 1800, "elapsed_seconds": 4842.566228866577, "loss": 0.050753841968253256, "learning_rate": 1.7462215917663923e-05, "mean_training_objective": 0.047695731990120775, "mean_relative_U_L2": 0.14715595765785952, "worst_relative_U_L2": 0.4951776216384098, "mean_relative_VM_L2": 0.2208656793071814, "worst_relative_VM_L2": 0.3618055007154412 }, { "step": 1850, "elapsed_seconds": 4980.868582248688, "loss": 0.046622951282188296, "learning_rate": 1.4213436473222475e-05, "mean_training_objective": 0.046573681328189453, "mean_relative_U_L2": 0.14335971129893849, "worst_relative_U_L2": 0.49124799944969016, "mean_relative_VM_L2": 0.2180532730004346, "worst_relative_VM_L2": 0.36040743607501935 }, { "step": 1900, "elapsed_seconds": 5114.47919344902, "loss": 0.04578052796423435, "learning_rate": 1.1877711917502638e-05, "mean_training_objective": 0.04566782966105404, "mean_relative_U_L2": 0.14107407035776554, "worst_relative_U_L2": 0.48811543387991996, "mean_relative_VM_L2": 0.21592118990783332, "worst_relative_VM_L2": 0.3593597317246937 }, { "step": 1950, "elapsed_seconds": 5247.332569122314, "loss": 0.04489049902185799, "learning_rate": 1.047019032053956e-05, "mean_training_objective": 0.044823994616174465, "mean_relative_U_L2": 0.13886924924576943, "worst_relative_U_L2": 0.4850566469444439, "mean_relative_VM_L2": 0.21395821579615398, "worst_relative_VM_L2": 0.3583744280346663 }, { "step": 2000, "elapsed_seconds": 5383.330991983414, "loss": 0.04432221017777919, "learning_rate": 1e-05, "mean_training_objective": 0.044258376962392515, "mean_relative_U_L2": 0.1391304336010175, "worst_relative_U_L2": 0.482016701253409, "mean_relative_VM_L2": 0.21188621225860307, "worst_relative_VM_L2": 0.35735227035870415 } ] }, { "variant": "S064", "points": [ { "step": 50, "elapsed_seconds": 179.2006766796112, "loss": 2.963256362080574, "learning_rate": 0.0003, "mean_training_objective": 2.970490675217195, "mean_relative_U_L2": 1.565623906068734, "worst_relative_U_L2": 6.749394146234147, "mean_relative_VM_L2": 0.7103888504522822, "worst_relative_VM_L2": 0.9456042084092691 }, { "step": 100, "elapsed_seconds": 357.97281193733215, "loss": 0.8880303777754307, "learning_rate": 0.00029952980967946045, "mean_training_objective": 0.7567905047119696, "mean_relative_U_L2": 0.8996077435507654, "worst_relative_U_L2": 2.448705700851271, "mean_relative_VM_L2": 0.6506906011512816, "worst_relative_VM_L2": 0.9002421439530092 }, { "step": 150, "elapsed_seconds": 538.7892134189606, "loss": 1.972909282892943, "learning_rate": 0.0002981222880824973, "mean_training_objective": 4.439018768209851, "mean_relative_U_L2": 1.963801594242265, "worst_relative_U_L2": 8.353495548207794, "mean_relative_VM_L2": 0.6232536318160954, "worst_relative_VM_L2": 0.8831875558529104 }, { "step": 200, "elapsed_seconds": 717.37242436409, "loss": 2.422786055505276, "learning_rate": 0.0002957865635267775, "mean_training_objective": 0.8590521668330273, "mean_relative_U_L2": 0.9420454085671587, "worst_relative_U_L2": 2.833633121118695, "mean_relative_VM_L2": 0.5892354776085806, "worst_relative_VM_L2": 0.861516205153449 }, { "step": 250, "elapsed_seconds": 896.5475528240204, "loss": 0.541191589832306, "learning_rate": 0.00029253778408233605, "mean_training_objective": 1.1635550005936945, "mean_relative_U_L2": 1.074050215191662, "worst_relative_U_L2": 3.866620394201137, "mean_relative_VM_L2": 0.48451437192016994, "worst_relative_VM_L2": 0.8342810194336213 }, { "step": 300, "elapsed_seconds": 1076.90163564682, "loss": 0.3385904246941209, "learning_rate": 0.0002883970193305295, "mean_training_objective": 1.4621068620125677, "mean_relative_U_L2": 1.1311786031591662, "worst_relative_U_L2": 4.592214612132656, "mean_relative_VM_L2": 0.3994079971394384, "worst_relative_VM_L2": 0.7984611356155169 }, { "step": 350, "elapsed_seconds": 1255.8521614074707, "loss": 1.0466882729902864, "learning_rate": 0.0002833911237197154, "mean_training_objective": 0.43521002777372564, "mean_relative_U_L2": 0.6526125375280403, "worst_relative_U_L2": 1.987741631335389, "mean_relative_VM_L2": 0.3765456469888029, "worst_relative_VM_L2": 0.7626523316481519 }, { "step": 400, "elapsed_seconds": 1436.7388229370117, "loss": 0.4046552770771087, "learning_rate": 0.00027755256240385024, "mean_training_objective": 0.6811564743211538, "mean_relative_U_L2": 0.8110346236437582, "worst_relative_U_L2": 2.928825682741071, "mean_relative_VM_L2": 0.347781357859625, "worst_relative_VM_L2": 0.6877016042667092 }, { "step": 450, "elapsed_seconds": 1617.7532868385315, "loss": 0.30398767367005347, "learning_rate": 0.0002709192006935076, "mean_training_objective": 0.29851204516652585, "mean_relative_U_L2": 0.5432658733936429, "worst_relative_U_L2": 1.5634987097648725, "mean_relative_VM_L2": 0.3218434883484264, "worst_relative_VM_L2": 0.46916878121718664 }, { "step": 500, "elapsed_seconds": 1801.5529854297638, "loss": 0.33726030746474867, "learning_rate": 0.0002635340584848096, "mean_training_objective": 0.5365874245567197, "mean_relative_U_L2": 0.735740843975044, "worst_relative_U_L2": 2.6086933879000336, "mean_relative_VM_L2": 0.34060050881513393, "worst_relative_VM_L2": 0.6013978347978376 }, { "step": 550, "elapsed_seconds": 1985.0439586639404, "loss": 0.21173200402408837, "learning_rate": 0.0002554450312588919, "mean_training_objective": 0.24820045724385126, "mean_relative_U_L2": 0.49397868417728563, "worst_relative_U_L2": 1.274063265181114, "mean_relative_VM_L2": 0.35333279160945513, "worst_relative_VM_L2": 0.4715063040360503 }, { "step": 600, "elapsed_seconds": 2167.82085108757, "loss": 0.20218459516763684, "learning_rate": 0.00024670457946132966, "mean_training_objective": 0.12524571064272155, "mean_relative_U_L2": 0.3038407194837102, "worst_relative_U_L2": 0.8322110249214382, "mean_relative_VM_L2": 0.2972798678651697, "worst_relative_VM_L2": 0.4093361277253206 }, { "step": 650, "elapsed_seconds": 2347.176343202591, "loss": 0.19846822591498492, "learning_rate": 0.0002373693882760176, "mean_training_objective": 0.16973194099479033, "mean_relative_U_L2": 0.3854931663648319, "worst_relative_U_L2": 1.0859112736432475, "mean_relative_VM_L2": 0.29359695852587814, "worst_relative_VM_L2": 0.43356483783575017 }, { "step": 700, "elapsed_seconds": 2527.3217208385468, "loss": 0.1895087950862944, "learning_rate": 0.00022749999999999997, "mean_training_objective": 0.17670006236701474, "mean_relative_U_L2": 0.3888554203060847, "worst_relative_U_L2": 1.1644208717072069, "mean_relative_VM_L2": 0.2853337645874815, "worst_relative_VM_L2": 0.4541470303033214 }, { "step": 750, "elapsed_seconds": 2710.062234401703, "loss": 0.14459396284073592, "learning_rate": 0.0002171604214034428, "mean_training_objective": 0.13056754252724062, "mean_relative_U_L2": 0.3214272750762998, "worst_relative_U_L2": 0.9387545904320381, "mean_relative_VM_L2": 0.27468340788230283, "worst_relative_VM_L2": 0.3864798608438502 }, { "step": 800, "elapsed_seconds": 2890.8336639404297, "loss": 0.13253616718575356, "learning_rate": 0.00020641770862116766, "mean_training_objective": 0.12198961877730208, "mean_relative_U_L2": 0.31281366813006417, "worst_relative_U_L2": 0.8677477747938658, "mean_relative_VM_L2": 0.27106783695639786, "worst_relative_VM_L2": 0.38622091315808516 }, { "step": 850, "elapsed_seconds": 3071.7022111415863, "loss": 0.11725206859409808, "learning_rate": 0.00019534153226788558, "mean_training_objective": 0.10895492245873806, "mean_relative_U_L2": 0.28878263224763023, "worst_relative_U_L2": 0.8128954269497359, "mean_relative_VM_L2": 0.2670135751489595, "worst_relative_VM_L2": 0.3855641222745536 }, { "step": 900, "elapsed_seconds": 3250.6403620243073, "loss": 0.10284697581082582, "learning_rate": 0.00018400372559752647, "mean_training_objective": 0.0954452261990987, "mean_relative_U_L2": 0.26345924068309345, "worst_relative_U_L2": 0.7180079589143384, "mean_relative_VM_L2": 0.26006706851212924, "worst_relative_VM_L2": 0.3843996372836651 }, { "step": 950, "elapsed_seconds": 3431.3681972026825, "loss": 0.09349920018576087, "learning_rate": 0.00017247781863702187, "mean_training_objective": 0.08664736047008266, "mean_relative_U_L2": 0.24648310782030464, "worst_relative_U_L2": 0.6724206362677496, "mean_relative_VM_L2": 0.2537602242143784, "worst_relative_VM_L2": 0.3830846588589727 }, { "step": 1000, "elapsed_seconds": 3610.2476065158844, "loss": 0.4543005123734475, "learning_rate": 0.0001608385613158652, "mean_training_objective": 0.5352026861661255, "mean_relative_U_L2": 0.7946679242182066, "worst_relative_U_L2": 2.0517961741662236, "mean_relative_VM_L2": 0.536336184928194, "worst_relative_VM_L2": 0.8397460588009695 }, { "step": 1050, "elapsed_seconds": 3790.8245220184326, "loss": 0.8149201029911638, "learning_rate": 0.00014916143868413484, "mean_training_objective": 0.7761299205631939, "mean_relative_U_L2": 0.9098567103307411, "worst_relative_U_L2": 3.0050137265511605, "mean_relative_VM_L2": 0.4236351619913649, "worst_relative_VM_L2": 0.7724860914939711 }, { "step": 1100, "elapsed_seconds": 3971.7329561710358, "loss": 0.18392920941114427, "learning_rate": 0.00013752218136297816, "mean_training_objective": 0.21251593109300754, "mean_relative_U_L2": 0.48637120445529447, "worst_relative_U_L2": 0.9746715513701953, "mean_relative_VM_L2": 0.29825252143427194, "worst_relative_VM_L2": 0.400549383997177 }, { "step": 1150, "elapsed_seconds": 4153.854551792145, "loss": 0.09388500591740015, "learning_rate": 0.00012599627440247358, "mean_training_objective": 0.12671701902941876, "mean_relative_U_L2": 0.3288844957503988, "worst_relative_U_L2": 0.9170879923676213, "mean_relative_VM_L2": 0.28376771597266187, "worst_relative_VM_L2": 0.384420537510991 }, { "step": 1200, "elapsed_seconds": 4338.551786899567, "loss": 0.13049623956903816, "learning_rate": 0.00011465846773211439, "mean_training_objective": 0.08626796052994634, "mean_relative_U_L2": 0.24532331035926758, "worst_relative_U_L2": 0.5987075351469103, "mean_relative_VM_L2": 0.2684806766270059, "worst_relative_VM_L2": 0.38049410790351823 }, { "step": 1250, "elapsed_seconds": 4518.754544019699, "loss": 0.08683793209493161, "learning_rate": 0.00010358229137883234, "mean_training_objective": 0.11083012605522809, "mean_relative_U_L2": 0.2952584860320674, "worst_relative_U_L2": 0.9046721129832619, "mean_relative_VM_L2": 0.26183617634560913, "worst_relative_VM_L2": 0.3793853633875699 }, { "step": 1300, "elapsed_seconds": 4697.476194858551, "loss": 0.07417388195171952, "learning_rate": 9.283957859655712e-05, "mean_training_objective": 0.07559793540241785, "mean_relative_U_L2": 0.22173483928807958, "worst_relative_U_L2": 0.5597174446407939, "mean_relative_VM_L2": 0.2561366919271958, "worst_relative_VM_L2": 0.3784078232839726 }, { "step": 1350, "elapsed_seconds": 4879.282232999802, "loss": 0.08383740857243538, "learning_rate": 8.250000000000003e-05, "mean_training_objective": 0.0656022115028589, "mean_relative_U_L2": 0.1914314011590677, "worst_relative_U_L2": 0.5572552588623345, "mean_relative_VM_L2": 0.25083449738515, "worst_relative_VM_L2": 0.3773103047210915 }, { "step": 1400, "elapsed_seconds": 5057.49857878685, "loss": 0.07089812252670526, "learning_rate": 7.263061172398237e-05, "mean_training_objective": 0.0829239993102578, "mean_relative_U_L2": 0.24467205535393682, "worst_relative_U_L2": 0.6702450462656698, "mean_relative_VM_L2": 0.24591028639329862, "worst_relative_VM_L2": 0.37601191817168605 }, { "step": 1450, "elapsed_seconds": 5236.212889909744, "loss": 0.06307391677983105, "learning_rate": 6.32954205386703e-05, "mean_training_objective": 0.0627555926515504, "mean_relative_U_L2": 0.18862785835763166, "worst_relative_U_L2": 0.5534078928645643, "mean_relative_VM_L2": 0.24124059737736073, "worst_relative_VM_L2": 0.37454299218420056 }, { "step": 1500, "elapsed_seconds": 5415.8405911922455, "loss": 0.0664056467823684, "learning_rate": 5.455496874110809e-05, "mean_training_objective": 0.05841726561094958, "mean_relative_U_L2": 0.17272896960038137, "worst_relative_U_L2": 0.551270148363933, "mean_relative_VM_L2": 0.23741816001224877, "worst_relative_VM_L2": 0.37297602070455016 }, { "step": 1550, "elapsed_seconds": 5595.352372169495, "loss": 0.05972972749732435, "learning_rate": 4.646594151519032e-05, "mean_training_objective": 0.06605510662009995, "mean_relative_U_L2": 0.20866896541857946, "worst_relative_U_L2": 0.5479951943510212, "mean_relative_VM_L2": 0.2355645679872575, "worst_relative_VM_L2": 0.37138559915830915 }, { "step": 1600, "elapsed_seconds": 5776.8753390312195, "loss": 0.056525486940518024, "learning_rate": 3.908079930649232e-05, "mean_training_objective": 0.05702337013860822, "mean_relative_U_L2": 0.1714681544535648, "worst_relative_U_L2": 0.5594727752766318, "mean_relative_VM_L2": 0.23124069392678148, "worst_relative_VM_L2": 0.3698384609121407 }, { "step": 1650, "elapsed_seconds": 5954.931214332581, "loss": 0.056737126968801024, "learning_rate": 3.244743759614975e-05, "mean_training_objective": 0.05370313073720614, "mean_relative_U_L2": 0.16086755364369504, "worst_relative_U_L2": 0.5419619272037706, "mean_relative_VM_L2": 0.22722222099578318, "worst_relative_VM_L2": 0.36831758003396836 }, { "step": 1700, "elapsed_seconds": 6132.969147205353, "loss": 0.053393528657034034, "learning_rate": 2.6608876280284544e-05, "mean_training_objective": 0.05586001775698901, "mean_relative_U_L2": 0.17422617987115457, "worst_relative_U_L2": 0.5390590225076746, "mean_relative_VM_L2": 0.22660414449016142, "worst_relative_VM_L2": 0.3668071964296406 }, { "step": 1750, "elapsed_seconds": 6310.459774017334, "loss": 0.05120803536847233, "learning_rate": 2.160298066947052e-05, "mean_training_objective": 0.05184330124926012, "mean_relative_U_L2": 0.1585114177807186, "worst_relative_U_L2": 0.5354674042120645, "mean_relative_VM_L2": 0.22192697706404835, "worst_relative_VM_L2": 0.36540435601634796 }, { "step": 1800, "elapsed_seconds": 6488.040977239609, "loss": 0.05127139491960407, "learning_rate": 1.7462215917663923e-05, "mean_training_objective": 0.05019239158950038, "mean_relative_U_L2": 0.1521576352621419, "worst_relative_U_L2": 0.5340027914327273, "mean_relative_VM_L2": 0.21886251468529627, "worst_relative_VM_L2": 0.3640879978360921 }, { "step": 1850, "elapsed_seconds": 6665.7494242191315, "loss": 0.049258766882121566, "learning_rate": 1.4213436473222475e-05, "mean_training_objective": 0.05094714836022938, "mean_relative_U_L2": 0.15779024405375253, "worst_relative_U_L2": 0.5306778383911792, "mean_relative_VM_L2": 0.21930784509961443, "worst_relative_VM_L2": 0.3629658743191142 }, { "step": 1900, "elapsed_seconds": 6843.55020236969, "loss": 0.04836572734639049, "learning_rate": 1.1877711917502638e-05, "mean_training_objective": 0.04816746253745578, "mean_relative_U_L2": 0.14791685593777257, "worst_relative_U_L2": 0.5258646364848225, "mean_relative_VM_L2": 0.2137544842224887, "worst_relative_VM_L2": 0.36171227024345953 }, { "step": 1950, "elapsed_seconds": 7021.829658269882, "loss": 0.047657296760007746, "learning_rate": 1.047019032053956e-05, "mean_training_objective": 0.04734234253140242, "mean_relative_U_L2": 0.14657582515996587, "worst_relative_U_L2": 0.5234674494584732, "mean_relative_VM_L2": 0.21112070037139802, "worst_relative_VM_L2": 0.3606574107926237 }, { "step": 2000, "elapsed_seconds": 7199.823485851288, "loss": 0.04663099129684269, "learning_rate": 1e-05, "mean_training_objective": 0.046935005608081316, "mean_relative_U_L2": 0.1487515662560716, "worst_relative_U_L2": 0.5199421588776311, "mean_relative_VM_L2": 0.20883005028492457, "worst_relative_VM_L2": 0.3595954078414134 } ] }, { "variant": "H064", "points": [ { "step": 50, "elapsed_seconds": 87.92495918273926, "loss": 2.484228190779686, "learning_rate": 0.0003, "mean_training_objective": 2.2762286717728353, "mean_relative_U_L2": 1.4809033875207782, "worst_relative_U_L2": 5.561315242740274, "mean_relative_VM_L2": 0.7338205545104948, "worst_relative_VM_L2": 0.9541218076829444 }, { "step": 100, "elapsed_seconds": 179.84996604919434, "loss": 2.6297349177300924, "learning_rate": 0.00029952980967946045, "mean_training_objective": 2.7449136129709593, "mean_relative_U_L2": 1.5861553552857017, "worst_relative_U_L2": 6.274483283053437, "mean_relative_VM_L2": 0.6684802195995768, "worst_relative_VM_L2": 0.9452264286086307 }, { "step": 150, "elapsed_seconds": 268.6702814102173, "loss": 1.1470502991229297, "learning_rate": 0.0002981222880824973, "mean_training_objective": 1.475317057248374, "mean_relative_U_L2": 1.1762323178878427, "worst_relative_U_L2": 4.361099415591946, "mean_relative_VM_L2": 0.6329801239575842, "worst_relative_VM_L2": 0.9391065581123366 }, { "step": 200, "elapsed_seconds": 360.41180992126465, "loss": 0.5867032080888749, "learning_rate": 0.0002957865635267775, "mean_training_objective": 1.2187492997687135, "mean_relative_U_L2": 1.124531932257916, "worst_relative_U_L2": 3.757474679742097, "mean_relative_VM_L2": 0.6112920130708547, "worst_relative_VM_L2": 0.8866721488488659 }, { "step": 250, "elapsed_seconds": 450.49268198013306, "loss": 1.4128484837710857, "learning_rate": 0.00029253778408233605, "mean_training_objective": 1.183658058539801, "mean_relative_U_L2": 1.0440477655125089, "worst_relative_U_L2": 3.957077960360485, "mean_relative_VM_L2": 0.48015144693098133, "worst_relative_VM_L2": 0.8194092684540185 }, { "step": 300, "elapsed_seconds": 542.8340320587158, "loss": 0.7212586216628551, "learning_rate": 0.0002883970193305295, "mean_training_objective": 1.0616371150048032, "mean_relative_U_L2": 1.0439171619191865, "worst_relative_U_L2": 3.6466764609688753, "mean_relative_VM_L2": 0.4642082704043798, "worst_relative_VM_L2": 0.7468424828538297 }, { "step": 350, "elapsed_seconds": 632.2816390991211, "loss": 0.3545133432373406, "learning_rate": 0.0002833911237197154, "mean_training_objective": 0.6933173030254551, "mean_relative_U_L2": 0.7999405182935823, "worst_relative_U_L2": 2.9107424054986795, "mean_relative_VM_L2": 0.39522498755534574, "worst_relative_VM_L2": 0.6698511826243745 }, { "step": 400, "elapsed_seconds": 724.8604698181152, "loss": 0.2458243016153574, "learning_rate": 0.00027755256240385024, "mean_training_objective": 0.3370644206527874, "mean_relative_U_L2": 0.5690933586357696, "worst_relative_U_L2": 1.7225130446710757, "mean_relative_VM_L2": 0.3613377344531255, "worst_relative_VM_L2": 0.6245723212073452 }, { "step": 450, "elapsed_seconds": 813.5448212623596, "loss": 0.23641777131706473, "learning_rate": 0.0002709192006935076, "mean_training_objective": 0.26020325679138084, "mean_relative_U_L2": 0.4994212121399514, "worst_relative_U_L2": 1.390487422278035, "mean_relative_VM_L2": 0.34223195927023775, "worst_relative_VM_L2": 0.5112023904353735 }, { "step": 500, "elapsed_seconds": 905.6203491687775, "loss": 0.13381979344412684, "learning_rate": 0.0002635340584848096, "mean_training_objective": 0.15889477345408914, "mean_relative_U_L2": 0.36237766713226965, "worst_relative_U_L2": 0.8587441647124597, "mean_relative_VM_L2": 0.3220714502194554, "worst_relative_VM_L2": 0.44994622456091 }, { "step": 550, "elapsed_seconds": 994.5465085506439, "loss": 0.34641841948032376, "learning_rate": 0.0002554450312588919, "mean_training_objective": 0.3759113276474223, "mean_relative_U_L2": 0.6601787557492805, "worst_relative_U_L2": 1.2994489849023565, "mean_relative_VM_L2": 0.45199677011502065, "worst_relative_VM_L2": 0.6903739946218251 }, { "step": 600, "elapsed_seconds": 1086.4568610191345, "loss": 0.36196318119764326, "learning_rate": 0.00024670457946132966, "mean_training_objective": 0.5109515236677515, "mean_relative_U_L2": 0.7113810661102578, "worst_relative_U_L2": 2.3713728395059097, "mean_relative_VM_L2": 0.350027416119943, "worst_relative_VM_L2": 0.5670053582038453 }, { "step": 650, "elapsed_seconds": 1174.779706478119, "loss": 0.1803417693823576, "learning_rate": 0.0002373693882760176, "mean_training_objective": 0.1717984355126798, "mean_relative_U_L2": 0.37717606060945374, "worst_relative_U_L2": 0.9417797451900544, "mean_relative_VM_L2": 0.34403232355656677, "worst_relative_VM_L2": 0.5571551157372983 }, { "step": 700, "elapsed_seconds": 1266.7221820354462, "loss": 0.16110318480059502, "learning_rate": 0.00022749999999999997, "mean_training_objective": 0.1565121909158257, "mean_relative_U_L2": 0.36251381529710036, "worst_relative_U_L2": 1.0906955268482597, "mean_relative_VM_L2": 0.2976792655294337, "worst_relative_VM_L2": 0.4134982186875229 }, { "step": 750, "elapsed_seconds": 1355.3910160064697, "loss": 0.1481035907752812, "learning_rate": 0.0002171604214034428, "mean_training_objective": 0.1382916297107206, "mean_relative_U_L2": 0.3394564906490445, "worst_relative_U_L2": 0.8878482896182955, "mean_relative_VM_L2": 0.29536365674348625, "worst_relative_VM_L2": 0.4074912612532424 }, { "step": 800, "elapsed_seconds": 1448.5585646629333, "loss": 0.10995412170886991, "learning_rate": 0.00020641770862116766, "mean_training_objective": 0.13604705150242408, "mean_relative_U_L2": 0.3310072759711887, "worst_relative_U_L2": 1.008514492656141, "mean_relative_VM_L2": 0.2850939399869331, "worst_relative_VM_L2": 0.3884288875559366 }, { "step": 850, "elapsed_seconds": 1539.0262038707733, "loss": 0.11209040172398092, "learning_rate": 0.00019534153226788558, "mean_training_objective": 0.115534954654685, "mean_relative_U_L2": 0.30102370981677523, "worst_relative_U_L2": 0.832166726979308, "mean_relative_VM_L2": 0.282588232549135, "worst_relative_VM_L2": 0.38408605372684335 }, { "step": 900, "elapsed_seconds": 1627.6292598247528, "loss": 0.09539618492126464, "learning_rate": 0.00018400372559752647, "mean_training_objective": 0.1039161163203249, "mean_relative_U_L2": 0.2756076590634808, "worst_relative_U_L2": 0.7921866546482922, "mean_relative_VM_L2": 0.27515894719299816, "worst_relative_VM_L2": 0.3831524818552191 }, { "step": 950, "elapsed_seconds": 1719.258152961731, "loss": 0.09440401419997214, "learning_rate": 0.00017247781863702187, "mean_training_objective": 0.09295273337991243, "mean_relative_U_L2": 0.25404671686770375, "worst_relative_U_L2": 0.6834644979691006, "mean_relative_VM_L2": 0.2710228501632787, "worst_relative_VM_L2": 0.3825700063710336 }, { "step": 1000, "elapsed_seconds": 1808.949234008789, "loss": 0.08206208888441326, "learning_rate": 0.0001608385613158652, "mean_training_objective": 0.09177842759673982, "mean_relative_U_L2": 0.2541039131995691, "worst_relative_U_L2": 0.6940332799232263, "mean_relative_VM_L2": 0.2671646291997814, "worst_relative_VM_L2": 0.3818983566611201 }, { "step": 1050, "elapsed_seconds": 1900.780734539032, "loss": 0.08241394544020296, "learning_rate": 0.00014916143868413484, "mean_training_objective": 0.08402407185185504, "mean_relative_U_L2": 0.23968506579670773, "worst_relative_U_L2": 0.5814945603291751, "mean_relative_VM_L2": 0.2644983505697678, "worst_relative_VM_L2": 0.381440501565035 }, { "step": 1100, "elapsed_seconds": 1989.668140888214, "loss": 0.07372363479807974, "learning_rate": 0.00013752218136297816, "mean_training_objective": 0.07723824750964746, "mean_relative_U_L2": 0.22309858832966234, "worst_relative_U_L2": 0.5639600532736129, "mean_relative_VM_L2": 0.2561759738459108, "worst_relative_VM_L2": 0.38075807077800466 }, { "step": 1150, "elapsed_seconds": 2081.5413262844086, "loss": 0.06925943214446305, "learning_rate": 0.00012599627440247358, "mean_training_objective": 0.0720235263765687, "mean_relative_U_L2": 0.2122731099034924, "worst_relative_U_L2": 0.5493920109937113, "mean_relative_VM_L2": 0.25057801670084456, "worst_relative_VM_L2": 0.3799743134857815 }, { "step": 1200, "elapsed_seconds": 2170.31147480011, "loss": 0.06595684606581927, "learning_rate": 0.00011465846773211439, "mean_training_objective": 0.0690360322118376, "mean_relative_U_L2": 0.20477177909974098, "worst_relative_U_L2": 0.5449635075291588, "mean_relative_VM_L2": 0.247940566625178, "worst_relative_VM_L2": 0.3791095111211088 }, { "step": 1250, "elapsed_seconds": 2262.035770893097, "loss": 0.062627184856683, "learning_rate": 0.00010358229137883234, "mean_training_objective": 0.06440824794982054, "mean_relative_U_L2": 0.19441802505598832, "worst_relative_U_L2": 0.5381406610267113, "mean_relative_VM_L2": 0.241892394733164, "worst_relative_VM_L2": 0.3780136199040201 }, { "step": 1300, "elapsed_seconds": 2354.03679394722, "loss": 0.06016041194088757, "learning_rate": 9.283957859655712e-05, "mean_training_objective": 0.06175121055922811, "mean_relative_U_L2": 0.18729915267252, "worst_relative_U_L2": 0.537303938286725, "mean_relative_VM_L2": 0.23945252850528317, "worst_relative_VM_L2": 0.3769890283329945 }, { "step": 1350, "elapsed_seconds": 2443.337594270706, "loss": 0.05798735814169049, "learning_rate": 8.250000000000003e-05, "mean_training_objective": 0.05938492110825342, "mean_relative_U_L2": 0.18087714953468503, "worst_relative_U_L2": 0.5337310058509502, "mean_relative_VM_L2": 0.23809009828036762, "worst_relative_VM_L2": 0.37566367629229136 }, { "step": 1400, "elapsed_seconds": 2535.1468806266785, "loss": 0.12109324503690003, "learning_rate": 7.263061172398237e-05, "mean_training_objective": 0.11543988089079074, "mean_relative_U_L2": 0.31933386914372147, "worst_relative_U_L2": 0.5987020548180465, "mean_relative_VM_L2": 0.32707536989073943, "worst_relative_VM_L2": 0.4645745893902273 }, { "step": 1450, "elapsed_seconds": 2623.5256192684174, "loss": 0.06478857304900884, "learning_rate": 6.32954205386703e-05, "mean_training_objective": 0.06350628101842232, "mean_relative_U_L2": 0.18379690355406994, "worst_relative_U_L2": 0.5302690325123196, "mean_relative_VM_L2": 0.2523506793959778, "worst_relative_VM_L2": 0.3780302117993259 }, { "step": 1500, "elapsed_seconds": 2715.0375509262085, "loss": 0.05838191946968436, "learning_rate": 5.455496874110809e-05, "mean_training_objective": 0.057475292893322795, "mean_relative_U_L2": 0.16850106072049104, "worst_relative_U_L2": 0.5224791059878041, "mean_relative_VM_L2": 0.24230929372007184, "worst_relative_VM_L2": 0.3771414146809052 }, { "step": 1550, "elapsed_seconds": 2803.669611930847, "loss": 0.05523433410562575, "learning_rate": 4.646594151519032e-05, "mean_training_objective": 0.0545750181546311, "mean_relative_U_L2": 0.1642483744819527, "worst_relative_U_L2": 0.5157588756399241, "mean_relative_VM_L2": 0.23411828791045566, "worst_relative_VM_L2": 0.37600366177865724 }, { "step": 1600, "elapsed_seconds": 2895.583115339279, "loss": 0.05286637153476476, "learning_rate": 3.908079930649232e-05, "mean_training_objective": 0.052320485178484734, "mean_relative_U_L2": 0.1567169767920758, "worst_relative_U_L2": 0.5110736751123567, "mean_relative_VM_L2": 0.2306966479366829, "worst_relative_VM_L2": 0.37509644544953885 }, { "step": 1650, "elapsed_seconds": 2984.4970693588257, "loss": 0.050921671278774734, "learning_rate": 3.244743759614975e-05, "mean_training_objective": 0.05059533001916651, "mean_relative_U_L2": 0.15170065674805983, "worst_relative_U_L2": 0.5060906864478586, "mean_relative_VM_L2": 0.22722748517420585, "worst_relative_VM_L2": 0.37403930385687245 }, { "step": 1700, "elapsed_seconds": 3076.272294282913, "loss": 0.04931133151985704, "learning_rate": 2.6608876280284544e-05, "mean_training_objective": 0.04908296440574879, "mean_relative_U_L2": 0.14736492051325586, "worst_relative_U_L2": 0.501943294401375, "mean_relative_VM_L2": 0.22406250898106078, "worst_relative_VM_L2": 0.3731374267720533 }, { "step": 1750, "elapsed_seconds": 3168.3154697418213, "loss": 0.05669584218412637, "learning_rate": 2.160298066947052e-05, "mean_training_objective": 0.052749258178281765, "mean_relative_U_L2": 0.15567643105127738, "worst_relative_U_L2": 0.5276067937086789, "mean_relative_VM_L2": 0.22999533925672364, "worst_relative_VM_L2": 0.37221346483494927 }, { "step": 1800, "elapsed_seconds": 3258.391053915024, "loss": 0.047032877476885906, "learning_rate": 1.7462215917663923e-05, "mean_training_objective": 0.04696032213006914, "mean_relative_U_L2": 0.142051117906576, "worst_relative_U_L2": 0.49413139183165883, "mean_relative_VM_L2": 0.21939169876388537, "worst_relative_VM_L2": 0.3713025475485811 }, { "step": 1850, "elapsed_seconds": 3347.298699617386, "loss": 0.04589824811555445, "learning_rate": 1.4213436473222475e-05, "mean_training_objective": 0.045835459041699333, "mean_relative_U_L2": 0.13991944372739482, "worst_relative_U_L2": 0.4902318819817173, "mean_relative_VM_L2": 0.21621466526629818, "worst_relative_VM_L2": 0.3704398966485011 }, { "step": 1900, "elapsed_seconds": 3439.342064142227, "loss": 0.045107859838753935, "learning_rate": 1.1877711917502638e-05, "mean_training_objective": 0.04502805534378283, "mean_relative_U_L2": 0.1387207200811075, "worst_relative_U_L2": 0.4864653526596369, "mean_relative_VM_L2": 0.21388483185221605, "worst_relative_VM_L2": 0.36959889251714695 }, { "step": 1950, "elapsed_seconds": 3528.312733888626, "loss": 0.045725555857643485, "learning_rate": 1.047019032053956e-05, "mean_training_objective": 0.04444801889724631, "mean_relative_U_L2": 0.13747263415092956, "worst_relative_U_L2": 0.48322795209084235, "mean_relative_VM_L2": 0.21278924172573083, "worst_relative_VM_L2": 0.368670141704416 }, { "step": 2000, "elapsed_seconds": 3618.6009981632233, "loss": 0.04367800313048064, "learning_rate": 1e-05, "mean_training_objective": 0.04366698628289756, "mean_relative_U_L2": 0.13655987851600554, "worst_relative_U_L2": 0.4788284152119077, "mean_relative_VM_L2": 0.21091793239568807, "worst_relative_VM_L2": 0.36774165845452417 } ] }, { "variant": "MGN", "points": [ { "step": 50, "elapsed_seconds": 272.7831811904907, "loss": 4.763434594869613, "learning_rate": 0.0003, "mean_training_objective": 1.855480426899112, "mean_relative_U_L2": 1.4342729142046298, "worst_relative_U_L2": 3.8512958592857864, "mean_relative_VM_L2": 0.8151652708209817, "worst_relative_VM_L2": 0.96780490690235 }, { "step": 100, "elapsed_seconds": 544.0952396392822, "loss": 0.939474505186081, "learning_rate": 0.00029952980967946045, "mean_training_objective": 6.7953780252863, "mean_relative_U_L2": 2.384952148083512, "worst_relative_U_L2": 10.558868172040661, "mean_relative_VM_L2": 0.6737368314377836, "worst_relative_VM_L2": 0.9083579508874353 }, { "step": 150, "elapsed_seconds": 816.9968283176422, "loss": 1.6753847852349282, "learning_rate": 0.0002981222880824973, "mean_training_objective": 1.2645891806097183, "mean_relative_U_L2": 1.1120725491419832, "worst_relative_U_L2": 3.7552710922573933, "mean_relative_VM_L2": 0.6319447059095201, "worst_relative_VM_L2": 0.8591526351666368 }, { "step": 200, "elapsed_seconds": 1090.3913328647614, "loss": 0.4420451126992702, "learning_rate": 0.0002957865635267775, "mean_training_objective": 0.488020540570945, "mean_relative_U_L2": 0.772339868326007, "worst_relative_U_L2": 1.5300186964093985, "mean_relative_VM_L2": 0.5246778172963233, "worst_relative_VM_L2": 0.7346055557549324 }, { "step": 250, "elapsed_seconds": 1363.8325843811035, "loss": 0.4692907094955445, "learning_rate": 0.00029253778408233605, "mean_training_objective": 0.3462909496559626, "mean_relative_U_L2": 0.506468955727365, "worst_relative_U_L2": 0.9606587140666579, "mean_relative_VM_L2": 0.6140864357149789, "worst_relative_VM_L2": 0.9771215938093267 }, { "step": 300, "elapsed_seconds": 1637.0684776306152, "loss": 0.13269720273092392, "learning_rate": 0.0002883970193305295, "mean_training_objective": 0.11991987180003194, "mean_relative_U_L2": 0.28580251558928227, "worst_relative_U_L2": 0.9061846999408714, "mean_relative_VM_L2": 0.32267306413174013, "worst_relative_VM_L2": 0.45345034393730926 }, { "step": 350, "elapsed_seconds": 1907.6814105510712, "loss": 0.10645618773996833, "learning_rate": 0.0002833911237197154, "mean_training_objective": 0.11762732489582498, "mean_relative_U_L2": 0.2674052287687598, "worst_relative_U_L2": 0.6175907031778436, "mean_relative_VM_L2": 0.37385521233892793, "worst_relative_VM_L2": 0.5442829066379824 }, { "step": 400, "elapsed_seconds": 2186.5799481868744, "loss": 0.040934174973517645, "learning_rate": 0.00027755256240385024, "mean_training_objective": 0.04216928084197981, "mean_relative_U_L2": 0.1202000426680077, "worst_relative_U_L2": 0.29612031758274476, "mean_relative_VM_L2": 0.2485208347491444, "worst_relative_VM_L2": 0.3988905347763434 }, { "step": 450, "elapsed_seconds": 2466.576562166214, "loss": 0.0621743619441986, "learning_rate": 0.0002709192006935076, "mean_training_objective": 0.09655318141284182, "mean_relative_U_L2": 0.2022617275297076, "worst_relative_U_L2": 0.6595832290498865, "mean_relative_VM_L2": 0.3434234755793456, "worst_relative_VM_L2": 0.5841993090559997 }, { "step": 500, "elapsed_seconds": 2745.9383606910706, "loss": 0.03788096467033029, "learning_rate": 0.0002635340584848096, "mean_training_objective": 0.03704253705189435, "mean_relative_U_L2": 0.11132060351021547, "worst_relative_U_L2": 0.3495673115124293, "mean_relative_VM_L2": 0.22657571293303205, "worst_relative_VM_L2": 0.37132614288193094 }, { "step": 550, "elapsed_seconds": 3019.76917719841, "loss": 0.04040515432134271, "learning_rate": 0.0002554450312588919, "mean_training_objective": 0.06696265595689718, "mean_relative_U_L2": 0.15299827918421133, "worst_relative_U_L2": 0.5074728339629729, "mean_relative_VM_L2": 0.2950306115752549, "worst_relative_VM_L2": 0.5163341245253504 }, { "step": 600, "elapsed_seconds": 3293.2419109344482, "loss": 0.0389953875914216, "learning_rate": 0.00024670457946132966, "mean_training_objective": 0.04637249311583116, "mean_relative_U_L2": 0.13369251894300288, "worst_relative_U_L2": 0.2519200222510693, "mean_relative_VM_L2": 0.2595446087406791, "worst_relative_VM_L2": 0.48076800456783947 }, { "step": 650, "elapsed_seconds": 3563.151460647583, "loss": 0.02658536988310516, "learning_rate": 0.0002373693882760176, "mean_training_objective": 0.033390866359996385, "mean_relative_U_L2": 0.12146551492589566, "worst_relative_U_L2": 0.445212227021131, "mean_relative_VM_L2": 0.1915579061939636, "worst_relative_VM_L2": 0.3386095696304781 }, { "step": 700, "elapsed_seconds": 3837.8308959007263, "loss": 0.028858510591089723, "learning_rate": 0.00022749999999999997, "mean_training_objective": 0.02619629757843859, "mean_relative_U_L2": 0.09831429578445784, "worst_relative_U_L2": 0.19090242746433522, "mean_relative_VM_L2": 0.19911033511530057, "worst_relative_VM_L2": 0.3304985117825815 }, { "step": 750, "elapsed_seconds": 4116.728692770004, "loss": 0.0338193085975945, "learning_rate": 0.0002171604214034428, "mean_training_objective": 0.029215384989889053, "mean_relative_U_L2": 0.0969096935123813, "worst_relative_U_L2": 0.26023689696384306, "mean_relative_VM_L2": 0.20699718118019902, "worst_relative_VM_L2": 0.31871208946597906 }, { "step": 800, "elapsed_seconds": 4395.002585411072, "loss": 0.01946368967182934, "learning_rate": 0.00020641770862116766, "mean_training_objective": 0.01849015631730998, "mean_relative_U_L2": 0.05641186455816925, "worst_relative_U_L2": 0.16557462183592322, "mean_relative_VM_L2": 0.1716336033721629, "worst_relative_VM_L2": 0.30251239510168487 }, { "step": 850, "elapsed_seconds": 4665.156551837921, "loss": 0.01831136189866811, "learning_rate": 0.00019534153226788558, "mean_training_objective": 0.01750146161567935, "mean_relative_U_L2": 0.052823633071165756, "worst_relative_U_L2": 0.16902473297431192, "mean_relative_VM_L2": 0.1673112286360744, "worst_relative_VM_L2": 0.28659768116348916 }, { "step": 900, "elapsed_seconds": 4932.366221666336, "loss": 0.015962740522809327, "learning_rate": 0.00018400372559752647, "mean_training_objective": 0.015690019474659943, "mean_relative_U_L2": 0.04410669079744238, "worst_relative_U_L2": 0.12499161397008005, "mean_relative_VM_L2": 0.1623075414959965, "worst_relative_VM_L2": 0.2607263016095473 }, { "step": 950, "elapsed_seconds": 5202.388907909393, "loss": 0.03443790376186372, "learning_rate": 0.00017247781863702187, "mean_training_objective": 0.03193855728340167, "mean_relative_U_L2": 0.12798686030763112, "worst_relative_U_L2": 0.24965366396684074, "mean_relative_VM_L2": 0.2088229483007411, "worst_relative_VM_L2": 0.3505095620132485 }, { "step": 1000, "elapsed_seconds": 5468.595334529877, "loss": 0.018761745514348147, "learning_rate": 0.0001608385613158652, "mean_training_objective": 0.017339713340372362, "mean_relative_U_L2": 0.05402671614544026, "worst_relative_U_L2": 0.11009556624169665, "mean_relative_VM_L2": 0.1678766857080993, "worst_relative_VM_L2": 0.3147744370097255 }, { "step": 1050, "elapsed_seconds": 5738.538566350937, "loss": 0.01624065425712615, "learning_rate": 0.00014916143868413484, "mean_training_objective": 0.0160190606479249, "mean_relative_U_L2": 0.046821572671453295, "worst_relative_U_L2": 0.13102811000507203, "mean_relative_VM_L2": 0.16184925042314047, "worst_relative_VM_L2": 0.281214432019637 }, { "step": 1100, "elapsed_seconds": 6007.47891330719, "loss": 0.015870346664451064, "learning_rate": 0.00013752218136297816, "mean_training_objective": 0.01482071990024016, "mean_relative_U_L2": 0.04090200908331187, "worst_relative_U_L2": 0.10270016517180579, "mean_relative_VM_L2": 0.15888542392350652, "worst_relative_VM_L2": 0.2605903332748718 }, { "step": 1150, "elapsed_seconds": 6274.428282737732, "loss": 0.016653000283986332, "learning_rate": 0.00012599627440247358, "mean_training_objective": 0.018403346867956312, "mean_relative_U_L2": 0.04486488806251101, "worst_relative_U_L2": 0.10831560812285118, "mean_relative_VM_L2": 0.17659192265196158, "worst_relative_VM_L2": 0.30848186493745083 }, { "step": 1200, "elapsed_seconds": 6543.742869853973, "loss": 0.01200055761728436, "learning_rate": 0.00011465846773211439, "mean_training_objective": 0.012179762785860318, "mean_relative_U_L2": 0.038871761989289134, "worst_relative_U_L2": 0.10733106560948166, "mean_relative_VM_L2": 0.14440911844485624, "worst_relative_VM_L2": 0.2135476125149908 }, { "step": 1250, "elapsed_seconds": 6810.2671501636505, "loss": 0.01208982248790562, "learning_rate": 0.00010358229137883234, "mean_training_objective": 0.01152698597053093, "mean_relative_U_L2": 0.03633658003836123, "worst_relative_U_L2": 0.09572971162980273, "mean_relative_VM_L2": 0.14123729764771237, "worst_relative_VM_L2": 0.20316138833062758 }, { "step": 1300, "elapsed_seconds": 7078.8772559165955, "loss": 0.027786271832883357, "learning_rate": 9.283957859655712e-05, "mean_training_objective": 0.02109336559331739, "mean_relative_U_L2": 0.06968445365456141, "worst_relative_U_L2": 0.19446314211435792, "mean_relative_VM_L2": 0.1794055177347915, "worst_relative_VM_L2": 0.32752687906707967 }, { "step": 1350, "elapsed_seconds": 7347.199320554733, "loss": 0.015231057722121478, "learning_rate": 8.250000000000003e-05, "mean_training_objective": 0.014806500668382655, "mean_relative_U_L2": 0.052585865866739676, "worst_relative_U_L2": 0.11837246587144136, "mean_relative_VM_L2": 0.15717503274075914, "worst_relative_VM_L2": 0.25951773177420256 }, { "step": 1400, "elapsed_seconds": 7612.891181468964, "loss": 0.010366187861654906, "learning_rate": 7.263061172398237e-05, "mean_training_objective": 0.010133862084157896, "mean_relative_U_L2": 0.028250667453317158, "worst_relative_U_L2": 0.06015613866358916, "mean_relative_VM_L2": 0.1351349761310698, "worst_relative_VM_L2": 0.19853982017099067 }, { "step": 1450, "elapsed_seconds": 7883.062077045441, "loss": 0.01172805419191718, "learning_rate": 6.32954205386703e-05, "mean_training_objective": 0.01216356453114321, "mean_relative_U_L2": 0.051281297490093594, "worst_relative_U_L2": 0.16742553110064123, "mean_relative_VM_L2": 0.1361626753614894, "worst_relative_VM_L2": 0.21429771371071765 }, { "step": 1500, "elapsed_seconds": 8149.574891328812, "loss": 0.009732101194094865, "learning_rate": 5.455496874110809e-05, "mean_training_objective": 0.009315717286988457, "mean_relative_U_L2": 0.02637849396440011, "worst_relative_U_L2": 0.057117021116396736, "mean_relative_VM_L2": 0.12993608171307194, "worst_relative_VM_L2": 0.18291912458003382 }, { "step": 1550, "elapsed_seconds": 8419.251989841461, "loss": 0.008683631010353566, "learning_rate": 4.646594151519032e-05, "mean_training_objective": 0.008671136953079884, "mean_relative_U_L2": 0.024230975839099674, "worst_relative_U_L2": 0.0488752712602132, "mean_relative_VM_L2": 0.1259215055598655, "worst_relative_VM_L2": 0.17353951354216338 }, { "step": 1600, "elapsed_seconds": 8685.848690748215, "loss": 0.008286534622311591, "learning_rate": 3.908079930649232e-05, "mean_training_objective": 0.00828073069353009, "mean_relative_U_L2": 0.023322712696910653, "worst_relative_U_L2": 0.046259089486284724, "mean_relative_VM_L2": 0.12323985541297286, "worst_relative_VM_L2": 0.1709275271486924 }, { "step": 1650, "elapsed_seconds": 8955.629194021225, "loss": 0.007950804138090461, "learning_rate": 3.244743759614975e-05, "mean_training_objective": 0.00794536364243836, "mean_relative_U_L2": 0.022648090934381963, "worst_relative_U_L2": 0.04465408131313908, "mean_relative_VM_L2": 0.12082698541322068, "worst_relative_VM_L2": 0.16845596390464623 }, { "step": 1700, "elapsed_seconds": 9224.555995464325, "loss": 0.0077296980540268135, "learning_rate": 2.6608876280284544e-05, "mean_training_objective": 0.007720277517824258, "mean_relative_U_L2": 0.022237747576538534, "worst_relative_U_L2": 0.04361635941532164, "mean_relative_VM_L2": 0.11914726000893705, "worst_relative_VM_L2": 0.16607889285399313 }, { "step": 1750, "elapsed_seconds": 9490.8275411129, "loss": 0.007498510763980448, "learning_rate": 2.160298066947052e-05, "mean_training_objective": 0.007494663259860389, "mean_relative_U_L2": 0.021976737634703418, "worst_relative_U_L2": 0.04292326488414796, "mean_relative_VM_L2": 0.11740541884111885, "worst_relative_VM_L2": 0.1637240768008511 }, { "step": 1800, "elapsed_seconds": 9762.646355628967, "loss": 0.007325060979928822, "learning_rate": 1.7462215917663923e-05, "mean_training_objective": 0.007321874166061486, "mean_relative_U_L2": 0.02179210551197701, "worst_relative_U_L2": 0.042297559157619284, "mean_relative_VM_L2": 0.11603798431784104, "worst_relative_VM_L2": 0.1614547009149057 }, { "step": 1850, "elapsed_seconds": 10035.813342094421, "loss": 0.007179439964238554, "learning_rate": 1.4213436473222475e-05, "mean_training_objective": 0.007177197005997449, "mean_relative_U_L2": 0.021606207143345744, "worst_relative_U_L2": 0.04189421473266117, "mean_relative_VM_L2": 0.11488368991427049, "worst_relative_VM_L2": 0.16060585044536202 }, { "step": 1900, "elapsed_seconds": 10309.976961374283, "loss": 0.007061271870043128, "learning_rate": 1.1877711917502638e-05, "mean_training_objective": 0.007059084155738233, "mean_relative_U_L2": 0.021427030838268184, "worst_relative_U_L2": 0.041489217209366186, "mean_relative_VM_L2": 0.11394257587218777, "worst_relative_VM_L2": 0.16003256344525438 }, { "step": 1950, "elapsed_seconds": 10583.519711017609, "loss": 0.006960467901080847, "learning_rate": 1.047019032053956e-05, "mean_training_objective": 0.006958425793938515, "mean_relative_U_L2": 0.02128401240486035, "worst_relative_U_L2": 0.041162157268561865, "mean_relative_VM_L2": 0.11313663746547035, "worst_relative_VM_L2": 0.15951338267991144 }, { "step": 2000, "elapsed_seconds": 10854.554028749466, "loss": 0.0068663904210552575, "learning_rate": 1e-05, "mean_training_objective": 0.006864725835916842, "mean_relative_U_L2": 0.021166705038592366, "worst_relative_U_L2": 0.04089674737765141, "mean_relative_VM_L2": 0.11238280734405437, "worst_relative_VM_L2": 0.15897842200912704 } ] } ], "half_parameter_ratio": 0.5106412328168624, "queue": { "phase": "handoff_to_original_precision_queue", "supervisor_pid": 22676, "updated_unix": 1793191864.9011025 }, "error_threshold": null, "note": "Nominal trainable parameters; one-slice softmax is identically1. Original64 baseline reused; no new training result is invented." }