{"step": 0, "total_steps": 46238, "tokens": 540672, "loss": 9.052214449102228, "ce_loss": 9.052203958684748, "z_loss": 0.10554811392318118, "muon_lr": 2.8571428571428574e-05, "residual_lr": 1.714285714285714e-07, "adamw_lr": 1.7142857142857143e-06, "weight_decay": 0.1, "grad_norm": 0.3953109681606293, "tokens_per_second": 11344.284782975496, "eta_seconds": 2203669.2257158756, "mean_data_wait_seconds": 2.6773874163627625}
{"step": 10, "total_steps": 46238, "tokens": 5947392, "loss": 9.032134901393544, "ce_loss": 9.032124410976063, "z_loss": 0.1053300601514903, "muon_lr": 0.0003142857142857143, "residual_lr": 1.8857142857142854e-06, "adamw_lr": 1.885714285714286e-05, "weight_decay": 0.1, "grad_norm": 0.4249423146247864, "tokens_per_second": 100117.16841716065, "eta_seconds": 249643.94158510727, "mean_data_wait_seconds": 6.848573684692383e-05}
{"step": 20, "total_steps": 46238, "tokens": 11354112, "loss": 8.983353759303238, "ce_loss": 8.983343268885758, "z_loss": 0.1047331120141528, "muon_lr": 0.0006000000000000001, "residual_lr": 3.5999999999999994e-06, "adamw_lr": 3.6e-05, "weight_decay": 0.1, "grad_norm": 0.47154128551483154, "tokens_per_second": 187555.45498287794, "eta_seconds": 133231.19728125844, "mean_data_wait_seconds": 6.753206253051758e-05}
{"step": 30, "total_steps": 46238, "tokens": 16760832, "loss": 8.879885745771004, "ce_loss": 8.879875255353523, "z_loss": 0.10429940954076522, "muon_lr": 0.0008857142857142858, "residual_lr": 5.314285714285714e-06, "adamw_lr": 5.314285714285714e-05, "weight_decay": 0.1, "grad_norm": 0.5881998538970947, "tokens_per_second": 271565.80123747507, "eta_seconds": 91995.49792410483, "mean_data_wait_seconds": 7.987022399902344e-05}
{"step": 40, "total_steps": 46238, "tokens": 22167552, "loss": 8.699208664171623, "ce_loss": 8.699198173754143, "z_loss": 0.10700192167000337, "muon_lr": 0.0011714285714285715, "residual_lr": 7.028571428571428e-06, "adamw_lr": 7.028571428571428e-05, "weight_decay": 0.1, "grad_norm": 0.8275405764579773, "tokens_per_second": 352338.9948708223, "eta_seconds": 70890.32082059907, "mean_data_wait_seconds": 7.230043411254883e-05}
{"step": 50, "total_steps": 46238, "tokens": 27574272, "loss": 8.427816383766405, "ce_loss": 8.427803032325976, "z_loss": 0.13336353331352724, "muon_lr": 0.0014571428571428574, "residual_lr": 8.742857142857143e-06, "adamw_lr": 8.742857142857143e-05, "weight_decay": 0.1, "grad_norm": 1.0486242771148682, "tokens_per_second": 430289.10679841833, "eta_seconds": 58035.44005518801, "mean_data_wait_seconds": 7.30752944946289e-05}
{"step": 60, "total_steps": 46238, "tokens": 32980992, "loss": 8.13443423762466, "ce_loss": 8.134410872603908, "z_loss": 0.2337838885459033, "muon_lr": 0.001742857142857143, "residual_lr": 1.0457142857142855e-05, "adamw_lr": 0.00010457142857142858, "weight_decay": 0.1, "grad_norm": 1.0130257606506348, "tokens_per_second": 504622.98396965035, "eta_seconds": 49475.770500183105, "mean_data_wait_seconds": 8.150935173034668e-05}
{"step": 70, "total_steps": 46238, "tokens": 38387712, "loss": 7.871237632000085, "ce_loss": 7.871192809307214, "z_loss": 0.44831768265276245, "muon_lr": 0.0020285714285714286, "residual_lr": 1.217142857142857e-05, "adamw_lr": 0.0001217142857142857, "weight_decay": 0.1, "grad_norm": 1.0109713077545166, "tokens_per_second": 576557.9273460183, "eta_seconds": 43293.48889347533, "mean_data_wait_seconds": 0.00010254979133605957}
{"step": 80, "total_steps": 46238, "tokens": 43794432, "loss": 7.607724359541228, "ce_loss": 7.607644185875401, "z_loss": 0.8018518334085291, "muon_lr": 0.002314285714285714, "residual_lr": 1.3885714285714284e-05, "adamw_lr": 0.00013885714285714285, "weight_decay": 0.1, "grad_norm": 0.9625251293182373, "tokens_per_second": 647619.6694159802, "eta_seconds": 38534.65032417097, "mean_data_wait_seconds": 0.0001068413257598877}
{"step": 90, "total_steps": 46238, "tokens": 49201152, "loss": 7.372094808202801, "ce_loss": 7.371967489069158, "z_loss": 1.2734035407051896, "muon_lr": 0.0026, "residual_lr": 1.56e-05, "adamw_lr": 0.00015600000000000002, "weight_decay": 0.1, "grad_norm": 0.886612594127655, "tokens_per_second": 715674.9060631958, "eta_seconds": 34862.743646061026, "mean_data_wait_seconds": 7.849931716918945e-05}
{"step": 100, "total_steps": 46238, "tokens": 54607872, "loss": 7.138011134032047, "ce_loss": 7.13783178546212, "z_loss": 1.793796546531446, "muon_lr": 0.002885714285714286, "residual_lr": 1.7314285714285714e-05, "adamw_lr": 0.00017314285714285714, "weight_decay": 0.1, "grad_norm": 0.7630711197853088, "tokens_per_second": 780923.6087470503, "eta_seconds": 31942.924742693947, "mean_data_wait_seconds": 0.00010332465171813965}
{"step": 110, "total_steps": 46238, "tokens": 60014592, "loss": 6.949020891478567, "ce_loss": 6.948795820727493, "z_loss": 2.2509782657478796, "muon_lr": 0.003171428571428572, "residual_lr": 1.9028571428571427e-05, "adamw_lr": 0.0001902857142857143, "weight_decay": 0.1, "grad_norm": 0.6479759812355042, "tokens_per_second": 843391.31844846, "eta_seconds": 29570.588170008617, "mean_data_wait_seconds": 0.0028472542762756348}
{"step": 120, "total_steps": 46238, "tokens": 65421312, "loss": 6.769602240938129, "ce_loss": 6.7693420720822886, "z_loss": 2.601500021688866, "muon_lr": 0.0034571428571428574, "residual_lr": 2.074285714285714e-05, "adamw_lr": 0.00020742857142857143, "weight_decay": 0.1, "grad_norm": 0.5374347567558289, "tokens_per_second": 904797.9967189356, "eta_seconds": 27557.720855283344, "mean_data_wait_seconds": 0.00013393163681030273}
{"step": 130, "total_steps": 46238, "tokens": 70828032, "loss": 6.577704389890035, "ce_loss": 6.5774198011918505, "z_loss": 2.8453644514083862, "muon_lr": 0.0037428571428571433, "residual_lr": 2.2457142857142855e-05, "adamw_lr": 0.00022457142857142858, "weight_decay": 0.1, "grad_norm": 0.46915215253829956, "tokens_per_second": 962560.2182910222, "eta_seconds": 25898.39412671737, "mean_data_wait_seconds": 0.000164717435836792}
{"step": 140, "total_steps": 46238, "tokens": 76234752, "loss": 6.439274220755606, "ce_loss": 6.4389699083386045, "z_loss": 3.0428148002335518, "muon_lr": 0.004028571428571429, "residual_lr": 2.417142857142857e-05, "adamw_lr": 0.0002417142857142857, "weight_decay": 0.1, "grad_norm": 0.38381829857826233, "tokens_per_second": 1019403.384317472, "eta_seconds": 24448.964529078057, "mean_data_wait_seconds": 0.0001106560230255127}
{"step": 150, "total_steps": 46238, "tokens": 81641472, "loss": 6.292080597444014, "ce_loss": 6.291761802904533, "z_loss": 3.1874968337290213, "muon_lr": 0.004314285714285715, "residual_lr": 2.5885714285714283e-05, "adamw_lr": 0.00025885714285714287, "weight_decay": 0.1, "grad_norm": 0.3728350102901459, "tokens_per_second": 1074356.0670840594, "eta_seconds": 23193.38181021356, "mean_data_wait_seconds": 6.517767906188965e-05}
{"step": 160, "total_steps": 46238, "tokens": 87048192, "loss": 6.203261570496992, "ce_loss": 6.202943263631878, "z_loss": 3.182531430865779, "muon_lr": 0.004600000000000001, "residual_lr": 2.7599999999999997e-05, "adamw_lr": 0.000276, "weight_decay": 0.1, "grad_norm": 0.36790215969085693, "tokens_per_second": 1128673.896459969, "eta_seconds": 22072.401800145275, "mean_data_wait_seconds": 9.456276893615723e-05}
{"step": 170, "total_steps": 46238, "tokens": 92454912, "loss": 6.021176916180235, "ce_loss": 6.020822506962401, "z_loss": 3.5438779935692297, "muon_lr": 0.004885714285714286, "residual_lr": 2.9314285714285714e-05, "adamw_lr": 0.00029314285714285716, "weight_decay": 0.1, "grad_norm": 0.26312363147735596, "tokens_per_second": 1180276.2746291952, "eta_seconds": 21102.802419564876, "mean_data_wait_seconds": 9.372830390930176e-05}
{"step": 180, "total_steps": 46238, "tokens": 97861632, "loss": 5.929763089526784, "ce_loss": 5.9293920704812715, "z_loss": 3.7095861037572226, "muon_lr": 0.005171428571428572, "residual_lr": 3.102857142857143e-05, "adamw_lr": 0.00031028571428571433, "weight_decay": 0.1, "grad_norm": 0.2812429964542389, "tokens_per_second": 1231421.820129556, "eta_seconds": 20221.933619285814, "mean_data_wait_seconds": 9.056925773620605e-05}
{"step": 190, "total_steps": 46238, "tokens": 103268352, "loss": 5.836227781844862, "ce_loss": 5.835880384300694, "z_loss": 3.474233612869725, "muon_lr": 0.0054571428571428575, "residual_lr": 3.274285714285714e-05, "adamw_lr": 0.00032742857142857145, "weight_decay": 0.1, "grad_norm": 0.29780805110931396, "tokens_per_second": 1278295.2819778176, "eta_seconds": 19476.19140507164, "mean_data_wait_seconds": 9.563565254211426e-05}
{"step": 200, "total_steps": 46238, "tokens": 108675072, "loss": 5.774684205199733, "ce_loss": 5.774328882044012, "z_loss": 3.553551738912409, "muon_lr": 0.005742857142857143, "residual_lr": 3.4457142857142855e-05, "adamw_lr": 0.00034457142857142857, "weight_decay": 0.1, "grad_norm": 0.38579097390174866, "tokens_per_second": 1326324.7456833012, "eta_seconds": 18766.834400859043, "mean_data_wait_seconds": 8.77678394317627e-05}
{"step": 210, "total_steps": 46238, "tokens": 114081792, "loss": 5.6659517577200225, "ce_loss": 5.6655723976366446, "z_loss": 3.793082560553695, "muon_lr": 0.006028571428571429, "residual_lr": 3.6171428571428566e-05, "adamw_lr": 0.00036171428571428574, "weight_decay": 0.1, "grad_norm": 0.36456620693206787, "tokens_per_second": 1372937.4941002524, "eta_seconds": 18125.741522055665, "mean_data_wait_seconds": 9.414553642272949e-05}
{"step": 220, "total_steps": 46238, "tokens": 119488512, "loss": 5.598135673638546, "ce_loss": 5.597749110424157, "z_loss": 3.8659336892041294, "muon_lr": 0.006314285714285715, "residual_lr": 3.788571428571428e-05, "adamw_lr": 0.00037885714285714286, "weight_decay": 0.1, "grad_norm": 0.3691854774951935, "tokens_per_second": 1418002.095588379, "eta_seconds": 17545.88621653367, "mean_data_wait_seconds": 7.888674736022949e-05}
{"step": 230, "total_steps": 46238, "tokens": 124895232, "loss": 5.536709380872322, "ce_loss": 5.536324566060847, "z_loss": 3.8490328933253433, "muon_lr": 0.006600000000000001, "residual_lr": 3.96e-05, "adamw_lr": 0.00039600000000000003, "weight_decay": 0.1, "grad_norm": 0.41115429997444153, "tokens_per_second": 1460913.3801457942, "eta_seconds": 17026.81147428302, "mean_data_wait_seconds": 7.346272468566895e-05}
{"step": 240, "total_steps": 46238, "tokens": 130301952, "loss": 5.43571455189676, "ce_loss": 5.435300736716299, "z_loss": 4.137660994674221, "muon_lr": 0.006885714285714286, "residual_lr": 4.131428571428571e-05, "adamw_lr": 0.00041314285714285715, "weight_decay": 0.1, "grad_norm": 0.49552541971206665, "tokens_per_second": 1504741.6876424737, "eta_seconds": 16527.28185059025, "mean_data_wait_seconds": 8.490681648254395e-05}
{"step": 250, "total_steps": 46238, "tokens": 135708672, "loss": 5.381461172392874, "ce_loss": 5.381046493848165, "z_loss": 4.1458047520030625, "muon_lr": 0.007171428571428572, "residual_lr": 4.302857142857142e-05, "adamw_lr": 0.00043028571428571427, "weight_decay": 0.1, "grad_norm": 0.612969696521759, "tokens_per_second": 1545982.0107016838, "eta_seconds": 16082.905940616272, "mean_data_wait_seconds": 8.022785186767578e-05}
{"step": 260, "total_steps": 46238, "tokens": 141115392, "loss": 5.281124927780845, "ce_loss": 5.280728578567505, "z_loss": 3.9635555112000667, "muon_lr": 0.0074571428571428575, "residual_lr": 4.474285714285714e-05, "adamw_lr": 0.00044742857142857144, "weight_decay": 0.1, "grad_norm": 0.3885248899459839, "tokens_per_second": 1586637.7233044836, "eta_seconds": 15667.392864092099, "mean_data_wait_seconds": 6.0617923736572266e-05}
{"step": 270, "total_steps": 46238, "tokens": 146522112, "loss": 5.226632927403306, "ce_loss": 5.226203094829213, "z_loss": 4.29835445953138, "muon_lr": 0.007742857142857143, "residual_lr": 4.6457142857142856e-05, "adamw_lr": 0.0004645714285714286, "weight_decay": 0.1, "grad_norm": 0.6426252722740173, "tokens_per_second": 1624216.9233877494, "eta_seconds": 15301.570539089147, "mean_data_wait_seconds": 7.140636444091797e-05}
{"step": 280, "total_steps": 46238, "tokens": 151928832, "loss": 5.115629181717381, "ce_loss": 5.115187760555383, "z_loss": 4.414462335181959, "muon_lr": 0.008028571428571428, "residual_lr": 4.8171428571428566e-05, "adamw_lr": 0.0004817142857142857, "weight_decay": 0.1, "grad_norm": 0.5316494703292847, "tokens_per_second": 1660430.51115586, "eta_seconds": 14964.590771524083, "mean_data_wait_seconds": 8.958578109741211e-05}
{"step": 290, "total_steps": 46238, "tokens": 157335552, "loss": 5.0240869558218755, "ce_loss": 5.0236753514318755, "z_loss": 4.115284258669073, "muon_lr": 0.008314285714285715, "residual_lr": 4.988571428571428e-05, "adamw_lr": 0.0004988571428571429, "weight_decay": 0.1, "grad_norm": 0.4625382721424103, "tokens_per_second": 1697404.606057911, "eta_seconds": 14635.43594458259, "mean_data_wait_seconds": 8.061528205871582e-05}
{"step": 300, "total_steps": 46238, "tokens": 162742272, "loss": 4.974108511751348, "ce_loss": 4.97370103633765, "z_loss": 4.074474193833091, "muon_lr": 0.0086, "residual_lr": 5.1599999999999994e-05, "adamw_lr": 0.000516, "weight_decay": 0.1, "grad_norm": 0.5326559543609619, "tokens_per_second": 1733261.2328470831, "eta_seconds": 14329.54778732493, "mean_data_wait_seconds": 9.140372276306152e-05}
{"step": 310, "total_steps": 46238, "tokens": 168148992, "loss": 4.923104593248079, "ce_loss": 4.922668771310286, "z_loss": 4.3581659685481675, "muon_lr": 0.008885714285714287, "residual_lr": 5.331428571428571e-05, "adamw_lr": 0.0005331428571428571, "weight_decay": 0.1, "grad_norm": 0.4334307610988617, "tokens_per_second": 1767175.8084765747, "eta_seconds": 14051.48419579509, "mean_data_wait_seconds": 7.906556129455566e-05}
{"step": 320, "total_steps": 46238, "tokens": 173555712, "loss": 4.827217571663134, "ce_loss": 4.826764305432637, "z_loss": 4.532740061933344, "muon_lr": 0.009171428571428572, "residual_lr": 5.502857142857143e-05, "adamw_lr": 0.0005502857142857143, "weight_decay": 0.1, "grad_norm": 0.5057864785194397, "tokens_per_second": 1799540.9755141486, "eta_seconds": 13795.760453249435, "mean_data_wait_seconds": 9.036064147949219e-05}
{"step": 330, "total_steps": 46238, "tokens": 178962432, "loss": 4.78076360803662, "ce_loss": 4.780308466969115, "z_loss": 4.552002906799316, "muon_lr": 0.009457142857142858, "residual_lr": 5.674285714285713e-05, "adamw_lr": 0.0005674285714285714, "weight_decay": 0.1, "grad_norm": 0.42452654242515564, "tokens_per_second": 1831615.900945791, "eta_seconds": 13551.219713250675, "mean_data_wait_seconds": 9.188055992126465e-05}
{"step": 340, "total_steps": 46238, "tokens": 184369152, "loss": 4.696901296124314, "ce_loss": 4.696412527199947, "z_loss": 4.887555494452968, "muon_lr": 0.009742857142857143, "residual_lr": 5.845714285714285e-05, "adamw_lr": 0.0005845714285714287, "weight_decay": 0.1, "grad_norm": 0.5704366564750671, "tokens_per_second": 1865494.244699479, "eta_seconds": 13302.224252103011, "mean_data_wait_seconds": 0.00011426210403442383}
{"step": 350, "total_steps": 46238, "tokens": 189775872, "loss": 4.612053795294329, "ce_loss": 4.61156221592065, "z_loss": 4.916486371647228, "muon_lr": 0.010028571428571428, "residual_lr": 6.0171428571428566e-05, "adamw_lr": 0.0006017142857142857, "weight_decay": 0.1, "grad_norm": 0.5638939142227173, "tokens_per_second": 1896501.1649908533, "eta_seconds": 13081.888122182965, "mean_data_wait_seconds": 0.00021395087242126465}
{"step": 360, "total_steps": 46238, "tokens": 195182592, "loss": 4.594254558736628, "ce_loss": 4.593751817038565, "z_loss": 5.027198581984549, "muon_lr": 0.010314285714285715, "residual_lr": 6.188571428571428e-05, "adamw_lr": 0.0006188571428571428, "weight_decay": 0.1, "grad_norm": 0.5701348781585693, "tokens_per_second": 1926930.8044833783, "eta_seconds": 12872.49614064383, "mean_data_wait_seconds": 0.000276714563369751}
{"step": 370, "total_steps": 46238, "tokens": 200589312, "loss": 4.482101194786303, "ce_loss": 4.4816241192095205, "z_loss": 4.770427331779942, "muon_lr": 0.0106, "residual_lr": 6.36e-05, "adamw_lr": 0.000636, "weight_decay": 0.1, "grad_norm": 0.49752551317214966, "tokens_per_second": 1955932.3242422258, "eta_seconds": 12678.865376187143, "mean_data_wait_seconds": 9.703636169433594e-05}
{"step": 380, "total_steps": 46238, "tokens": 205996032, "loss": 4.4479883006124785, "ce_loss": 4.447486812418157, "z_loss": 5.014239751931393, "muon_lr": 0.010885714285714287, "residual_lr": 6.53142857142857e-05, "adamw_lr": 0.0006531428571428571, "weight_decay": 0.1, "grad_norm": 0.5440066456794739, "tokens_per_second": 1982848.050729186, "eta_seconds": 12504.032215117157, "mean_data_wait_seconds": 8.484721183776855e-05}
{"step": 390, "total_steps": 46238, "tokens": 211402752, "loss": 4.351758093544931, "ce_loss": 4.351266199892217, "z_loss": 4.918577898632396, "muon_lr": 0.011171428571428572, "residual_lr": 6.702857142857142e-05, "adamw_lr": 0.0006702857142857143, "weight_decay": 0.1, "grad_norm": 0.42828017473220825, "tokens_per_second": 2009016.564302888, "eta_seconds": 12338.46929111871, "mean_data_wait_seconds": 0.00017514824867248535}
{"step": 400, "total_steps": 46238, "tokens": 216809472, "loss": 4.29128660216476, "ce_loss": 4.29074734991247, "z_loss": 5.392309261090828, "muon_lr": 0.011457142857142857, "residual_lr": 6.874285714285714e-05, "adamw_lr": 0.0006874285714285715, "weight_decay": 0.1, "grad_norm": 0.5227214694023132, "tokens_per_second": 2032028.2189948983, "eta_seconds": 12196.081841943269, "mean_data_wait_seconds": 0.00024449825286865234}
{"step": 410, "total_steps": 46238, "tokens": 222216192, "loss": 4.240909020105998, "ce_loss": 4.240399537664471, "z_loss": 5.094859675927595, "muon_lr": 0.011742857142857143, "residual_lr": 7.045714285714286e-05, "adamw_lr": 0.0007045714285714286, "weight_decay": 0.1, "grad_norm": 0.4517413377761841, "tokens_per_second": 2055497.419994368, "eta_seconds": 12054.199388908934, "mean_data_wait_seconds": 8.258223533630371e-05}
{"step": 420, "total_steps": 46238, "tokens": 227622912, "loss": 4.167386867783287, "ce_loss": 4.166840824213895, "z_loss": 5.45992982748783, "muon_lr": 0.012028571428571428, "residual_lr": 7.217142857142857e-05, "adamw_lr": 0.0007217142857142858, "weight_decay": 0.1, "grad_norm": 0.4295867383480072, "tokens_per_second": 2079870.2736118857, "eta_seconds": 11910.343322029023, "mean_data_wait_seconds": 0.00014215707778930664}
{"step": 430, "total_steps": 46238, "tokens": 233029632, "loss": 4.113530915794951, "ce_loss": 4.1129871101090405, "z_loss": 5.439461202332468, "muon_lr": 0.012314285714285715, "residual_lr": 7.388571428571429e-05, "adamw_lr": 0.0007388571428571428, "weight_decay": 0.1, "grad_norm": 0.3626275956630707, "tokens_per_second": 2103204.9355271384, "eta_seconds": 11775.629604916561, "mean_data_wait_seconds": 0.0001386106014251709}
{"step": 440, "total_steps": 46238, "tokens": 238436352, "loss": 4.022465886491718, "ce_loss": 4.021934556238579, "z_loss": 5.313172806393016, "muon_lr": 0.0126, "residual_lr": 7.56e-05, "adamw_lr": 0.0007559999999999999, "weight_decay": 0.1, "grad_norm": 0.4576810896396637, "tokens_per_second": 2124962.56108657, "eta_seconds": 11652.513807743855, "mean_data_wait_seconds": 0.00012630224227905273}
{"step": 450, "total_steps": 46238, "tokens": 243843072, "loss": 4.016290084882216, "ce_loss": 4.015733619530995, "z_loss": 5.564574787110994, "muon_lr": 0.012885714285714287, "residual_lr": 7.73142857142857e-05, "adamw_lr": 0.0007731428571428571, "weight_decay": 0.1, "grad_norm": 0.41140323877334595, "tokens_per_second": 2145908.3525819974, "eta_seconds": 11536.256352333694, "mean_data_wait_seconds": 9.942054748535156e-05}
{"step": 460, "total_steps": 46238, "tokens": 249249792, "loss": 4.005860944588979, "ce_loss": 4.0052911910143765, "z_loss": 5.697642441951867, "muon_lr": 0.013171428571428572, "residual_lr": 7.902857142857142e-05, "adamw_lr": 0.0007902857142857143, "weight_decay": 0.1, "grad_norm": 0.41040465235710144, "tokens_per_second": 2170043.950301031, "eta_seconds": 11405.456622464537, "mean_data_wait_seconds": 6.783008575439453e-05}
{"step": 470, "total_steps": 46238, "tokens": 254656512, "loss": 3.9645612583015906, "ce_loss": 3.9639642509547146, "z_loss": 5.970354860479182, "muon_lr": 0.013457142857142857, "residual_lr": 8.074285714285713e-05, "adamw_lr": 0.0008074285714285715, "weight_decay": 0.1, "grad_norm": 0.4186404049396515, "tokens_per_second": 2196792.2926308545, "eta_seconds": 11264.121558968935, "mean_data_wait_seconds": 6.198883056640625e-05}
{"step": 480, "total_steps": 46238, "tokens": 260063232, "loss": 3.867638288122235, "ce_loss": 3.8670596281687417, "z_loss": 5.786360639514345, "muon_lr": 0.013742857142857143, "residual_lr": 8.245714285714285e-05, "adamw_lr": 0.0008245714285714286, "weight_decay": 0.1, "grad_norm": 0.4207620918750763, "tokens_per_second": 2218833.9804816237, "eta_seconds": 11149.788096642544, "mean_data_wait_seconds": 7.686018943786621e-05}
{"step": 490, "total_steps": 46238, "tokens": 265469952, "loss": 3.908154623074965, "ce_loss": 3.907532016436259, "z_loss": 6.226537538297249, "muon_lr": 0.014028571428571428, "residual_lr": 8.417142857142857e-05, "adamw_lr": 0.0008417142857142858, "weight_decay": 0.1, "grad_norm": 0.4364689588546753, "tokens_per_second": 2243621.217862759, "eta_seconds": 11024.196859557856, "mean_data_wait_seconds": 7.912516593933105e-05}
{"step": 500, "total_steps": 46238, "tokens": 270876672, "loss": 3.890223631353089, "ce_loss": 3.889557179176446, "z_loss": 6.664539069840402, "muon_lr": 0.014314285714285715, "residual_lr": 8.588571428571428e-05, "adamw_lr": 0.0008588571428571429, "weight_decay": 0.1, "grad_norm": 0.45793312788009644, "tokens_per_second": 2267432.1800697725, "eta_seconds": 10906.04406224801, "mean_data_wait_seconds": 0.00015121698379516602}
{"step": 510, "total_steps": 46238, "tokens": 276283392, "loss": 3.7962850386446174, "ce_loss": 3.79566317616087, "z_loss": 6.218616146029848, "muon_lr": 0.0146, "residual_lr": 8.76e-05, "adamw_lr": 0.000876, "weight_decay": 0.1, "grad_norm": 0.30268368124961853, "tokens_per_second": 2291045.783911989, "eta_seconds": 10791.276506829403, "mean_data_wait_seconds": 6.178021430969238e-05}
{"step": 520, "total_steps": 46238, "tokens": 281690112, "loss": 3.7390191247968962, "ce_loss": 3.738355029713024, "z_loss": 6.641295451106447, "muon_lr": 0.014885714285714287, "residual_lr": 8.93142857142857e-05, "adamw_lr": 0.0008931428571428571, "weight_decay": 0.1, "grad_norm": 0.3748205602169037, "tokens_per_second": 2313620.7617779267, "eta_seconds": 10683.644542074935, "mean_data_wait_seconds": 7.417798042297363e-05}
{"step": 530, "total_steps": 46238, "tokens": 287096832, "loss": 3.7924333774682246, "ce_loss": 3.7917879115451467, "z_loss": 6.454664923928001, "muon_lr": 0.015171428571428572, "residual_lr": 9.102857142857142e-05, "adamw_lr": 0.0009102857142857143, "weight_decay": 0.1, "grad_norm": 0.369611531496048, "tokens_per_second": 2335223.2882456616, "eta_seconds": 10582.497711627946, "mean_data_wait_seconds": 0.0001061558723449707}
{"step": 540, "total_steps": 46238, "tokens": 292503552, "loss": 3.7409344080722695, "ce_loss": 3.7402730710578687, "z_loss": 6.613438895254424, "muon_lr": 0.015457142857142857, "residual_lr": 9.274285714285713e-05, "adamw_lr": 0.0009274285714285715, "weight_decay": 0.1, "grad_norm": 0.31902188062667847, "tokens_per_second": 2354935.8379185884, "eta_seconds": 10491.618491753634, "mean_data_wait_seconds": 7.516145706176758e-05}
{"step": 550, "total_steps": 46238, "tokens": 297910272, "loss": 3.85558359189467, "ce_loss": 3.854899874239257, "z_loss": 6.837251507874691, "muon_lr": 0.015742857142857142, "residual_lr": 9.445714285714284e-05, "adamw_lr": 0.0009445714285714286, "weight_decay": 0.1, "grad_norm": 0.415489137172699, "tokens_per_second": 2376840.1887098197, "eta_seconds": 10392.655670050917, "mean_data_wait_seconds": 0.00010734796524047852}
{"step": 560, "total_steps": 46238, "tokens": 303316992, "loss": 3.7667812018683464, "ce_loss": 3.766074413603002, "z_loss": 7.067843036218123, "muon_lr": 0.01602857142857143, "residual_lr": 9.617142857142856e-05, "adamw_lr": 0.0009617142857142858, "weight_decay": 0.1, "grad_norm": 0.3473696708679199, "tokens_per_second": 2397958.224223096, "eta_seconds": 10298.87622500231, "mean_data_wait_seconds": 6.61313533782959e-05}
{"step": 570, "total_steps": 46238, "tokens": 308723712, "loss": 3.7927884513681587, "ce_loss": 3.792060330058589, "z_loss": 7.2810338302092115, "muon_lr": 0.016314285714285715, "residual_lr": 9.788571428571428e-05, "adamw_lr": 0.0009788571428571429, "weight_decay": 0.1, "grad_norm": 0.312349408864975, "tokens_per_second": 2418792.39474479, "eta_seconds": 10207.93197367613, "mean_data_wait_seconds": 0.00010100007057189941}
{"step": 580, "total_steps": 46238, "tokens": 314130432, "loss": 3.794269789348949, "ce_loss": 3.7935274983897354, "z_loss": 7.4227275342652295, "muon_lr": 0.0166, "residual_lr": 9.96e-05, "adamw_lr": 0.0009960000000000001, "weight_decay": 0.1, "grad_norm": 0.30594298243522644, "tokens_per_second": 2437915.2191578117, "eta_seconds": 10125.643955956639, "mean_data_wait_seconds": 6.377696990966797e-05}
{"step": 590, "total_steps": 46238, "tokens": 319537152, "loss": 3.7661880525675686, "ce_loss": 3.7654012423573118, "z_loss": 7.868299726283912, "muon_lr": 0.016885714285714285, "residual_lr": 0.00010131428571428571, "adamw_lr": 0.0010131428571428572, "weight_decay": 0.1, "grad_norm": 0.31382685899734497, "tokens_per_second": 2456989.3972096043, "eta_seconds": 10044.835688761646, "mean_data_wait_seconds": 7.49826431274414e-05}
{"step": 600, "total_steps": 46238, "tokens": 324943872, "loss": 3.775128559632735, "ce_loss": 3.774371092969721, "z_loss": 7.574922438823815, "muon_lr": 0.01717142857142857, "residual_lr": 0.00010302857142857142, "adamw_lr": 0.0010302857142857143, "weight_decay": 0.1, "grad_norm": 0.36194613575935364, "tokens_per_second": 2476346.4937894135, "eta_seconds": 9964.13390689999, "mean_data_wait_seconds": 7.095932960510254e-05}
{"step": 610, "total_steps": 46238, "tokens": 330350592, "loss": 3.775815880659855, "ce_loss": 3.7750375559835723, "z_loss": 7.782880942026774, "muon_lr": 0.01745714285714286, "residual_lr": 0.00010474285714285713, "adamw_lr": 0.0010474285714285713, "weight_decay": 0.1, "grad_norm": 0.32553830742836, "tokens_per_second": 2493042.5181649076, "eta_seconds": 9895.234904440647, "mean_data_wait_seconds": 6.54458999633789e-05}
{"step": 620, "total_steps": 46238, "tokens": 335757312, "loss": 3.7576517336296313, "ce_loss": 3.7568403193444917, "z_loss": 8.114575165690798, "muon_lr": 0.017742857142857144, "residual_lr": 0.00010645714285714285, "adamw_lr": 0.0010645714285714286, "weight_decay": 0.1, "grad_norm": 0.3281985819339752, "tokens_per_second": 2511341.064478518, "eta_seconds": 9820.981694942127, "mean_data_wait_seconds": 6.905198097229004e-05}
{"step": 630, "total_steps": 46238, "tokens": 341164032, "loss": 3.6409137718605273, "ce_loss": 3.640098107583595, "z_loss": 8.156427719376303, "muon_lr": 0.01802857142857143, "residual_lr": 0.00010817142857142857, "adamw_lr": 0.0010817142857142857, "weight_decay": 0.1, "grad_norm": 0.3317421078681946, "tokens_per_second": 2529708.7460544775, "eta_seconds": 9747.536329017767, "mean_data_wait_seconds": 5.605816841125488e-05}
{"step": 640, "total_steps": 46238, "tokens": 346570752, "loss": 3.6960002978642783, "ce_loss": 3.6951471729712053, "z_loss": 8.53112227266485, "muon_lr": 0.018314285714285717, "residual_lr": 0.00010988571428571427, "adamw_lr": 0.001098857142857143, "weight_decay": 0.1, "grad_norm": 0.3365170955657959, "tokens_per_second": 2547490.1130512413, "eta_seconds": 9677.376590275355, "mean_data_wait_seconds": 0.00010558962821960449}
{"step": 650, "total_steps": 46238, "tokens": 351977472, "loss": 3.645277402617715, "ce_loss": 3.6444689252159814, "z_loss": 8.08448227969083, "muon_lr": 0.018600000000000002, "residual_lr": 0.00011159999999999999, "adamw_lr": 0.001116, "weight_decay": 0.1, "grad_norm": 0.2878842353820801, "tokens_per_second": 2563931.2573426557, "eta_seconds": 9613.211896151075, "mean_data_wait_seconds": 7.534027099609375e-05}
{"step": 660, "total_steps": 46238, "tokens": 357384192, "loss": 3.7351434790726863, "ce_loss": 3.7342790545839253, "z_loss": 8.644110997517904, "muon_lr": 0.018885714285714287, "residual_lr": 0.0001133142857142857, "adamw_lr": 0.0011331428571428573, "weight_decay": 0.1, "grad_norm": 0.33581894636154175, "tokens_per_second": 2581968.1151913847, "eta_seconds": 9543.962839437865, "mean_data_wait_seconds": 0.000206679105758667}
{"step": 670, "total_steps": 46238, "tokens": 362790912, "loss": 3.679190866874926, "ce_loss": 3.678334283106255, "z_loss": 8.565745346473925, "muon_lr": 0.019171428571428575, "residual_lr": 0.00011502857142857142, "adamw_lr": 0.0011502857142857141, "weight_decay": 0.1, "grad_norm": 0.3414471447467804, "tokens_per_second": 2597111.8340350683, "eta_seconds": 9486.230319825085, "mean_data_wait_seconds": 8.133053779602051e-05}
{"step": 680, "total_steps": 46238, "tokens": 368197632, "loss": 3.6527383797096484, "ce_loss": 3.6518543687733738, "z_loss": 8.84007673552542, "muon_lr": 0.01945714285714286, "residual_lr": 0.00011674285714285713, "adamw_lr": 0.0011674285714285714, "weight_decay": 0.1, "grad_norm": 0.3140312731266022, "tokens_per_second": 2611565.001545485, "eta_seconds": 9431.660437103235, "mean_data_wait_seconds": 0.00015419721603393555}
{"step": 690, "total_steps": 46238, "tokens": 373604352, "loss": 3.6681778304504626, "ce_loss": 3.667257263804927, "z_loss": 9.205871690403331, "muon_lr": 0.019742857142857145, "residual_lr": 0.00011845714285714284, "adamw_lr": 0.0011845714285714285, "weight_decay": 0.1, "grad_norm": 0.3939666152000427, "tokens_per_second": 2628979.7271603392, "eta_seconds": 9367.127227945368, "mean_data_wait_seconds": 9.468197822570801e-05}
{"step": 700, "total_steps": 46238, "tokens": 379011072, "loss": 3.6554380312110437, "ce_loss": 3.6545085112253823, "z_loss": 9.295054630799727, "muon_lr": 0.02002857142857143, "residual_lr": 0.00012017142857142856, "adamw_lr": 0.0012017142857142858, "weight_decay": 0.1, "grad_norm": 0.31567853689193726, "tokens_per_second": 2644425.6240799166, "eta_seconds": 9310.369949454078, "mean_data_wait_seconds": 6.398558616638184e-05}
{"step": 710, "total_steps": 46238, "tokens": 384417792, "loss": 3.6453957413182114, "ce_loss": 3.6444597298448738, "z_loss": 9.360314253604773, "muon_lr": 0.020314285714285715, "residual_lr": 0.00012188571428571428, "adamw_lr": 0.0012188571428571428, "weight_decay": 0.1, "grad_norm": 0.31086382269859314, "tokens_per_second": 2661350.3869364704, "eta_seconds": 9249.129413709024, "mean_data_wait_seconds": 0.0001087188720703125}
{"step": 720, "total_steps": 46238, "tokens": 389824512, "loss": 3.6566397854776094, "ce_loss": 3.6557150049643083, "z_loss": 9.248053319526441, "muon_lr": 0.020600000000000004, "residual_lr": 0.0001236, "adamw_lr": 0.0012360000000000001, "weight_decay": 0.1, "grad_norm": 0.3114159405231476, "tokens_per_second": 2677291.2060823175, "eta_seconds": 9192.039837911952, "mean_data_wait_seconds": 0.00013828277587890625}
{"step": 730, "total_steps": 46238, "tokens": 395231232, "loss": 3.558587451775869, "ce_loss": 3.5576875787792783, "z_loss": 8.998234474297726, "muon_lr": 0.02088571428571429, "residual_lr": 0.00012531428571428569, "adamw_lr": 0.0012531428571428572, "weight_decay": 0.1, "grad_norm": 0.31265509128570557, "tokens_per_second": 2693831.149675847, "eta_seconds": 9133.594251800334, "mean_data_wait_seconds": 8.445978164672852e-05}
{"step": 740, "total_steps": 46238, "tokens": 400637952, "loss": 3.6102102398872375, "ce_loss": 3.609273915941065, "z_loss": 9.363181684956405, "muon_lr": 0.021171428571428574, "residual_lr": 0.0001270285714285714, "adamw_lr": 0.0012702857142857142, "weight_decay": 0.1, "grad_norm": 0.27127939462661743, "tokens_per_second": 2708842.137779797, "eta_seconds": 9080.984691179394, "mean_data_wait_seconds": 0.00010344386100769043}
{"step": 750, "total_steps": 46238, "tokens": 406044672, "loss": 3.6089898383978642, "ce_loss": 3.6080095822160896, "z_loss": 9.802614067540024, "muon_lr": 0.02145714285714286, "residual_lr": 0.00012874285714285712, "adamw_lr": 0.0012874285714285715, "weight_decay": 0.1, "grad_norm": 0.30166977643966675, "tokens_per_second": 2722873.503919523, "eta_seconds": 9032.203379480563, "mean_data_wait_seconds": 7.414817810058594e-05}
{"step": 760, "total_steps": 46238, "tokens": 411451392, "loss": 3.601229053555113, "ce_loss": 3.6002613125425396, "z_loss": 9.677649346264927, "muon_lr": 0.021742857142857144, "residual_lr": 0.00013045714285714284, "adamw_lr": 0.0013045714285714286, "weight_decay": 0.1, "grad_norm": 0.2758495509624481, "tokens_per_second": 2738727.237067757, "eta_seconds": 8977.944284194404, "mean_data_wait_seconds": 6.350874900817871e-05}
{"step": 770, "total_steps": 46238, "tokens": 416858112, "loss": 3.631019626602982, "ce_loss": 3.6300423325914326, "z_loss": 9.773278525381377, "muon_lr": 0.022028571428571432, "residual_lr": 0.00013217142857142855, "adamw_lr": 0.0013217142857142859, "weight_decay": 0.1, "grad_norm": 0.30365970730781555, "tokens_per_second": 2754041.769079687, "eta_seconds": 8926.057004652754, "mean_data_wait_seconds": 0.00013375282287597656}
{"step": 780, "total_steps": 46238, "tokens": 422264832, "loss": 3.6220490011301907, "ce_loss": 3.621073740901369, "z_loss": 9.752623218478579, "muon_lr": 0.022314285714285717, "residual_lr": 0.00013388571428571427, "adamw_lr": 0.001338857142857143, "weight_decay": 0.1, "grad_norm": 0.25014927983283997, "tokens_per_second": 2766798.0281232595, "eta_seconds": 8882.949479572599, "mean_data_wait_seconds": 0.00014320015907287598}
{"step": 790, "total_steps": 46238, "tokens": 427671552, "loss": 3.6106043837287207, "ce_loss": 3.6096284100503633, "z_loss": 9.76001063260165, "muon_lr": 0.022600000000000002, "residual_lr": 0.0001356, "adamw_lr": 0.0013560000000000002, "weight_decay": 0.1, "grad_norm": 0.2565643787384033, "tokens_per_second": 2779423.394818039, "eta_seconds": 8840.653939163038, "mean_data_wait_seconds": 0.00012639164924621582}
{"step": 800, "total_steps": 46238, "tokens": 433078272, "loss": 3.5327758608442363, "ce_loss": 3.5318257808685303, "z_loss": 9.50088915680394, "muon_lr": 0.022885714285714287, "residual_lr": 0.0001373142857142857, "adamw_lr": 0.001373142857142857, "weight_decay": 0.1, "grad_norm": 0.3022971451282501, "tokens_per_second": 2794066.2353223315, "eta_seconds": 8792.38772275058, "mean_data_wait_seconds": 0.00010502338409423828}
{"step": 810, "total_steps": 46238, "tokens": 438484992, "loss": 3.59161873477878, "ce_loss": 3.590649315805146, "z_loss": 9.694518623930035, "muon_lr": 0.023171428571428575, "residual_lr": 0.00013902857142857142, "adamw_lr": 0.0013902857142857141, "weight_decay": 0.1, "grad_norm": 0.2686510384082794, "tokens_per_second": 2806788.6183568616, "eta_seconds": 8750.608002101155, "mean_data_wait_seconds": 6.112456321716309e-05}
{"step": 820, "total_steps": 46238, "tokens": 443891712, "loss": 3.5779100277207116, "ce_loss": 3.5768753199866326, "z_loss": 10.346984335870454, "muon_lr": 0.02345714285714286, "residual_lr": 0.00014074285714285714, "adamw_lr": 0.0014074285714285714, "weight_decay": 0.1, "grad_norm": 0.2964498698711395, "tokens_per_second": 2818620.978830824, "eta_seconds": 8711.955388264301, "mean_data_wait_seconds": 7.444620132446289e-05}
{"step": 830, "total_steps": 46238, "tokens": 449298432, "loss": 3.5412607120745108, "ce_loss": 3.54021061369867, "z_loss": 10.501328663392501, "muon_lr": 0.023742857142857145, "residual_lr": 0.00014245714285714283, "adamw_lr": 0.0014245714285714285, "weight_decay": 0.1, "grad_norm": 0.35820987820625305, "tokens_per_second": 2830207.97244766, "eta_seconds": 8674.377905439957, "mean_data_wait_seconds": 7.510185241699219e-05}
{"step": 840, "total_steps": 46238, "tokens": 454705152, "loss": 3.5269358880592114, "ce_loss": 3.525868610902266, "z_loss": 10.672737056558782, "muon_lr": 0.02402857142857143, "residual_lr": 0.00014417142857142855, "adamw_lr": 0.0014417142857142858, "weight_decay": 0.1, "grad_norm": 0.2574874758720398, "tokens_per_second": 2841845.2593439813, "eta_seconds": 8636.954001382188, "mean_data_wait_seconds": 6.505846977233887e-05}
{"step": 850, "total_steps": 46238, "tokens": 460111872, "loss": 3.581765263369589, "ce_loss": 3.5807577841209643, "z_loss": 10.074587525743427, "muon_lr": 0.024314285714285715, "residual_lr": 0.00014588571428571427, "adamw_lr": 0.0014588571428571428, "weight_decay": 0.1, "grad_norm": 0.27444562315940857, "tokens_per_second": 2852696.550436076, "eta_seconds": 8602.204836770594, "mean_data_wait_seconds": 7.134675979614258e-05}
{"step": 860, "total_steps": 46238, "tokens": 465518592, "loss": 3.4984573902505818, "ce_loss": 3.497417258493828, "z_loss": 10.401022528157089, "muon_lr": 0.024600000000000004, "residual_lr": 0.00014759999999999998, "adamw_lr": 0.001476, "weight_decay": 0.1, "grad_norm": 0.25474682450294495, "tokens_per_second": 2864456.717705098, "eta_seconds": 8565.000543508242, "mean_data_wait_seconds": 0.0003612637519836426}
{"step": 870, "total_steps": 46238, "tokens": 470925312, "loss": 3.5459305246671042, "ce_loss": 3.5449077270247717, "z_loss": 10.228458093874382, "muon_lr": 0.02488571428571429, "residual_lr": 0.0001493142857142857, "adamw_lr": 0.0014931428571428572, "weight_decay": 0.1, "grad_norm": 0.296743780374527, "tokens_per_second": 2872487.749730586, "eta_seconds": 8539.171882038687, "mean_data_wait_seconds": 0.0006667077541351318}
{"step": 880, "total_steps": 46238, "tokens": 476332032, "loss": 3.5617080330848694, "ce_loss": 3.5606449246406555, "z_loss": 10.631504210558804, "muon_lr": 0.025171428571428574, "residual_lr": 0.00015102857142857142, "adamw_lr": 0.0015102857142857144, "weight_decay": 0.1, "grad_norm": 0.2930905520915985, "tokens_per_second": 2879278.587753007, "eta_seconds": 8517.154265068177, "mean_data_wait_seconds": 0.00012812018394470215}
{"step": 890, "total_steps": 46238, "tokens": 481738752, "loss": 3.535235311045791, "ce_loss": 3.5341407728917673, "z_loss": 10.945257512005893, "muon_lr": 0.02545714285714286, "residual_lr": 0.00015274285714285713, "adamw_lr": 0.0015274285714285715, "weight_decay": 0.1, "grad_norm": 0.3061404228210449, "tokens_per_second": 2888817.7330953972, "eta_seconds": 8487.158225011612, "mean_data_wait_seconds": 0.0001958310604095459}
{"step": 900, "total_steps": 46238, "tokens": 487145472, "loss": 3.542224459575884, "ce_loss": 3.541111749230009, "z_loss": 11.126885103456901, "muon_lr": 0.025742857142857144, "residual_lr": 0.00015445714285714285, "adamw_lr": 0.0015445714285714286, "weight_decay": 0.1, "grad_norm": 0.30806365609169006, "tokens_per_second": 2897945.264416677, "eta_seconds": 8458.560886219524, "mean_data_wait_seconds": 7.396936416625977e-05}
{"step": 910, "total_steps": 46238, "tokens": 492552192, "loss": 3.537105423031431, "ce_loss": 3.5360331824331572, "z_loss": 10.722358905907834, "muon_lr": 0.026028571428571432, "residual_lr": 0.00015617142857142857, "adamw_lr": 0.0015617142857142859, "weight_decay": 0.1, "grad_norm": 0.2748248279094696, "tokens_per_second": 2907193.3881038465, "eta_seconds": 8429.793437300084, "mean_data_wait_seconds": 9.441375732421875e-05}
{"step": 920, "total_steps": 46238, "tokens": 497958912, "loss": 3.497484331781214, "ce_loss": 3.496426356561256, "z_loss": 10.579945210254554, "muon_lr": 0.026314285714285717, "residual_lr": 0.00015788571428571429, "adamw_lr": 0.0015788571428571427, "weight_decay": 0.1, "grad_norm": 0.30118125677108765, "tokens_per_second": 2916799.2025678265, "eta_seconds": 8400.178182450749, "mean_data_wait_seconds": 0.00010752677917480469}
{"step": 930, "total_steps": 46238, "tokens": 503365632, "loss": 3.530738496419155, "ce_loss": 3.529623112895272, "z_loss": 11.15395964275707, "muon_lr": 0.026600000000000002, "residual_lr": 0.0001596, "adamw_lr": 0.001596, "weight_decay": 0.1, "grad_norm": 0.2822984457015991, "tokens_per_second": 2926831.4645205587, "eta_seconds": 8369.537706884226, "mean_data_wait_seconds": 6.55055046081543e-05}
{"step": 940, "total_steps": 46238, "tokens": 508772352, "loss": 3.460087964029023, "ce_loss": 3.459034419420994, "z_loss": 10.535153677969268, "muon_lr": 0.026885714285714287, "residual_lr": 0.00016131428571428572, "adamw_lr": 0.001613142857142857, "weight_decay": 0.1, "grad_norm": 0.2771816849708557, "tokens_per_second": 2938059.502950187, "eta_seconds": 8335.712588328484, "mean_data_wait_seconds": 9.062886238098145e-05}
{"step": 950, "total_steps": 46238, "tokens": 514179072, "loss": 3.5082675381140276, "ce_loss": 3.5071054660912715, "z_loss": 11.620926221211752, "muon_lr": 0.027171428571428576, "residual_lr": 0.00016302857142857144, "adamw_lr": 0.0016302857142857143, "weight_decay": 0.1, "grad_norm": 0.3232291340827942, "tokens_per_second": 2948633.7298198147, "eta_seconds": 8303.985882131334, "mean_data_wait_seconds": 9.927153587341309e-05}
{"step": 960, "total_steps": 46238, "tokens": 519585792, "loss": 3.465458824779048, "ce_loss": 3.464367248795249, "z_loss": 10.915682749314742, "muon_lr": 0.02745714285714286, "residual_lr": 0.00016474285714285716, "adamw_lr": 0.0016474285714285714, "weight_decay": 0.1, "grad_norm": 0.2853988707065582, "tokens_per_second": 2958902.3820020244, "eta_seconds": 8273.340240253743, "mean_data_wait_seconds": 6.16610050201416e-05}
{"step": 970, "total_steps": 46238, "tokens": 524992512, "loss": 3.518980060562943, "ce_loss": 3.517861205520052, "z_loss": 11.188474850221114, "muon_lr": 0.027742857142857145, "residual_lr": 0.00016645714285714282, "adamw_lr": 0.0016645714285714287, "weight_decay": 0.1, "grad_norm": 0.2604736387729645, "tokens_per_second": 2969437.354084352, "eta_seconds": 8242.167288134935, "mean_data_wait_seconds": 5.9038400650024414e-05}
{"step": 980, "total_steps": 46238, "tokens": 530399232, "loss": 3.468756159146627, "ce_loss": 3.467618086121299, "z_loss": 11.381204532854484, "muon_lr": 0.02802857142857143, "residual_lr": 0.00016817142857142854, "adamw_lr": 0.0016817142857142857, "weight_decay": 0.1, "grad_norm": 0.2604101896286011, "tokens_per_second": 2978797.688084623, "eta_seconds": 8214.452697435043, "mean_data_wait_seconds": 0.00026297569274902344}
{"step": 990, "total_steps": 46238, "tokens": 535805952, "loss": 3.4642227490743003, "ce_loss": 3.463080841483492, "z_loss": 11.418714978478171, "muon_lr": 0.028314285714285715, "residual_lr": 0.00016988571428571425, "adamw_lr": 0.0016988571428571428, "weight_decay": 0.1, "grad_norm": 0.26113438606262207, "tokens_per_second": 2987427.0249758665, "eta_seconds": 8188.915002600817, "mean_data_wait_seconds": 7.37309455871582e-05}
{"step": 1000, "total_steps": 46238, "tokens": 541212672, "loss": 3.4695944280335396, "ce_loss": 3.46842144836079, "z_loss": 11.729695977586688, "muon_lr": 0.028600000000000004, "residual_lr": 0.00017159999999999997, "adamw_lr": 0.001716, "weight_decay": 0.1, "grad_norm": 0.28092503547668457, "tokens_per_second": 2997391.708207128, "eta_seconds": 8159.887543903841, "mean_data_wait_seconds": 0.00015562772750854492}
{"step": 1010, "total_steps": 46238, "tokens": 546619392, "loss": 3.3959062803875315, "ce_loss": 3.394798966971311, "z_loss": 11.07349250533364, "muon_lr": 0.02888571428571429, "residual_lr": 0.0001733142857142857, "adamw_lr": 0.0017331428571428571, "weight_decay": 0.1, "grad_norm": 0.259115606546402, "tokens_per_second": 3008137.663905604, "eta_seconds": 8128.94065235418, "mean_data_wait_seconds": 0.00013750791549682617}
{"step": 1020, "total_steps": 46238, "tokens": 552026112, "loss": 3.419576256564169, "ce_loss": 3.418389603947148, "z_loss": 11.866757862495653, "muon_lr": 0.029171428571428574, "residual_lr": 0.0001750285714285714, "adamw_lr": 0.0017502857142857144, "weight_decay": 0.1, "grad_norm": 0.3549405038356781, "tokens_per_second": 3017802.187916992, "eta_seconds": 8101.116077748849, "mean_data_wait_seconds": 0.00013056397438049316}
{"step": 1030, "total_steps": 46238, "tokens": 557432832, "loss": 3.474321020371986, "ce_loss": 3.473158173488848, "z_loss": 11.628327348015524, "muon_lr": 0.02945714285714286, "residual_lr": 0.00017674285714285712, "adamw_lr": 0.0017674285714285715, "weight_decay": 0.1, "grad_norm": 0.2753974199295044, "tokens_per_second": 3027534.3189328546, "eta_seconds": 8073.288864522392, "mean_data_wait_seconds": 0.00010085105895996094}
{"step": 1040, "total_steps": 46238, "tokens": 562839552, "loss": 3.4568745602260935, "ce_loss": 3.4556852542992793, "z_loss": 11.892928390791923, "muon_lr": 0.029742857142857144, "residual_lr": 0.00017845714285714284, "adamw_lr": 0.0017845714285714288, "weight_decay": 0.1, "grad_norm": 0.2553349733352661, "tokens_per_second": 3036888.1974606495, "eta_seconds": 8046.642087263286, "mean_data_wait_seconds": 0.0001157224178314209}
{"step": 1050, "total_steps": 46238, "tokens": 568246272, "loss": 3.3711810437115757, "ce_loss": 3.37000648541884, "z_loss": 11.745037505120942, "muon_lr": 0.030028571428571432, "residual_lr": 0.00018017142857142856, "adamw_lr": 0.0018017142857142858, "weight_decay": 0.1, "grad_norm": 0.2718960642814636, "tokens_per_second": 3046036.2559585324, "eta_seconds": 8020.700875181113, "mean_data_wait_seconds": 6.0886144638061523e-05}
{"step": 1060, "total_steps": 46238, "tokens": 573652992, "loss": 3.4106645999532756, "ce_loss": 3.409509657007275, "z_loss": 11.549420681866733, "muon_lr": 0.030314285714285717, "residual_lr": 0.00018188571428571427, "adamw_lr": 0.0018188571428571431, "weight_decay": 0.1, "grad_norm": 0.21588589251041412, "tokens_per_second": 3054976.220199272, "eta_seconds": 7995.459598833384, "mean_data_wait_seconds": 0.00010532140731811523}
{"step": 1070, "total_steps": 46238, "tokens": 579059712, "loss": 3.370537570028594, "ce_loss": 3.369422045621005, "z_loss": 11.155268004446318, "muon_lr": 0.030600000000000002, "residual_lr": 0.0001836, "adamw_lr": 0.001836, "weight_decay": 0.1, "grad_norm": 0.30497679114341736, "tokens_per_second": 3064805.8978725965, "eta_seconds": 7968.051823755384, "mean_data_wait_seconds": 7.778406143188477e-05}
{"step": 1080, "total_steps": 46238, "tokens": 584466432, "loss": 3.366933103763696, "ce_loss": 3.3657792445385093, "z_loss": 11.538481242728956, "muon_lr": 0.030885714285714287, "residual_lr": 0.0001853142857142857, "adamw_lr": 0.001853142857142857, "weight_decay": 0.1, "grad_norm": 0.27991053462028503, "tokens_per_second": 3074677.1006352296, "eta_seconds": 7940.712050366467, "mean_data_wait_seconds": 6.753206253051758e-05}
{"step": 1090, "total_steps": 46238, "tokens": 589873152, "loss": 3.4726954915306787, "ce_loss": 3.471499042077498, "z_loss": 11.96450348333879, "muon_lr": 0.031171428571428572, "residual_lr": 0.00018702857142857142, "adamw_lr": 0.0018702857142857143, "weight_decay": 0.1, "grad_norm": 0.26324713230133057, "tokens_per_second": 3083149.3970047664, "eta_seconds": 7917.137848627665, "mean_data_wait_seconds": 7.647275924682617e-05}
{"step": 1100, "total_steps": 46238, "tokens": 595279872, "loss": 3.4164850549264387, "ce_loss": 3.4153145804549707, "z_loss": 11.704873323440552, "muon_lr": 0.03145714285714286, "residual_lr": 0.00018874285714285714, "adamw_lr": 0.0018874285714285714, "weight_decay": 0.1, "grad_norm": 0.2876526117324829, "tokens_per_second": 3090910.2193331956, "eta_seconds": 7895.509844108238, "mean_data_wait_seconds": 7.921457290649414e-05}
{"step": 1110, "total_steps": 46238, "tokens": 600686592, "loss": 3.537984186952764, "ce_loss": 3.536846216880914, "z_loss": 11.379961996367484, "muon_lr": 0.03174285714285714, "residual_lr": 0.00019045714285714286, "adamw_lr": 0.0019045714285714286, "weight_decay": 0.1, "grad_norm": 0.31112438440322876, "tokens_per_second": 3100901.6248639445, "eta_seconds": 7868.326150163029, "mean_data_wait_seconds": 6.592273712158203e-05}
{"step": 1120, "total_steps": 46238, "tokens": 606093312, "loss": 3.4038231083841035, "ce_loss": 3.402599210088903, "z_loss": 12.238930969527274, "muon_lr": 0.032028571428571434, "residual_lr": 0.00019217142857142858, "adamw_lr": 0.0019217142857142857, "weight_decay": 0.1, "grad_norm": 0.258923202753067, "tokens_per_second": 3110653.110480266, "eta_seconds": 7841.921859372416, "mean_data_wait_seconds": 6.809830665588379e-05}
{"step": 1130, "total_steps": 46238, "tokens": 611500032, "loss": 3.486631747448083, "ce_loss": 3.485413190090295, "z_loss": 12.185909365162704, "muon_lr": 0.03231428571428572, "residual_lr": 0.0001938857142857143, "adamw_lr": 0.001938857142857143, "weight_decay": 0.1, "grad_norm": 0.25553324818611145, "tokens_per_second": 3120912.6158334096, "eta_seconds": 7814.4103683875155, "mean_data_wait_seconds": 9.885430335998535e-05}
{"step": 1140, "total_steps": 46238, "tokens": 616906752, "loss": 3.4494626142761926, "ce_loss": 3.4482458540887544, "z_loss": 12.167793245026559, "muon_lr": 0.032600000000000004, "residual_lr": 0.00019559999999999998, "adamw_lr": 0.001956, "weight_decay": 0.1, "grad_norm": 0.2522298991680145, "tokens_per_second": 3128530.6542053306, "eta_seconds": 7793.653915848678, "mean_data_wait_seconds": 0.0001741945743560791}
{"step": 1150, "total_steps": 46238, "tokens": 622313472, "loss": 3.432720608783491, "ce_loss": 3.4315326087402576, "z_loss": 11.879740303212946, "muon_lr": 0.03288571428571429, "residual_lr": 0.0001973142857142857, "adamw_lr": 0.0019731428571428573, "weight_decay": 0.1, "grad_norm": 0.2713463008403778, "tokens_per_second": 3138239.561285745, "eta_seconds": 7767.819501329773, "mean_data_wait_seconds": 9.289383888244629e-05}
{"step": 1160, "total_steps": 46238, "tokens": 627720192, "loss": 3.4696939984957376, "ce_loss": 3.4684925548958057, "z_loss": 12.014665134025343, "muon_lr": 0.033171428571428574, "residual_lr": 0.0001990285714285714, "adamw_lr": 0.0019902857142857144, "weight_decay": 0.1, "grad_norm": 0.24970538914203644, "tokens_per_second": 3147574.751370871, "eta_seconds": 7743.063682088967, "mean_data_wait_seconds": 6.99162483215332e-05}
{"step": 1170, "total_steps": 46238, "tokens": 633126912, "loss": 3.4705612045345884, "ce_loss": 3.4693785273667537, "z_loss": 11.826811617070978, "muon_lr": 0.03345714285714286, "residual_lr": 0.0002007428571428571, "adamw_lr": 0.0020074285714285715, "weight_decay": 0.1, "grad_norm": 0.3021051585674286, "tokens_per_second": 3156356.4551727646, "eta_seconds": 7719.807749871613, "mean_data_wait_seconds": 7.897615432739258e-05}
{"step": 1180, "total_steps": 46238, "tokens": 638533632, "loss": 3.406192757866599, "ce_loss": 3.405009686946869, "z_loss": 11.831011620434849, "muon_lr": 0.033742857142857144, "residual_lr": 0.00020245714285714283, "adamw_lr": 0.0020245714285714285, "weight_decay": 0.1, "grad_norm": 0.25465455651283264, "tokens_per_second": 3164572.1311023263, "eta_seconds": 7698.05752397694, "mean_data_wait_seconds": 0.00010010600090026855}
{"step": 1190, "total_steps": 46238, "tokens": 643940352, "loss": 3.437336788032994, "ce_loss": 3.436085439089573, "z_loss": 12.513240070054024, "muon_lr": 0.03402857142857143, "residual_lr": 0.00020417142857142854, "adamw_lr": 0.0020417142857142856, "weight_decay": 0.1, "grad_norm": 0.25343620777130127, "tokens_per_second": 3172870.044608195, "eta_seconds": 7676.220973937678, "mean_data_wait_seconds": 6.92605972290039e-05}
{"step": 1200, "total_steps": 46238, "tokens": 649347072, "loss": 3.439154209512653, "ce_loss": 3.4379660360740893, "z_loss": 11.881410844398268, "muon_lr": 0.034314285714285714, "residual_lr": 0.00020588571428571426, "adamw_lr": 0.002058857142857143, "weight_decay": 0.1, "grad_norm": 0.24985955655574799, "tokens_per_second": 3180914.31742467, "eta_seconds": 7655.108699600068, "mean_data_wait_seconds": 0.00010150671005249023}
{"step": 1210, "total_steps": 46238, "tokens": 654753792, "loss": 3.479338333462224, "ce_loss": 3.478121201197306, "z_loss": 12.171531742269343, "muon_lr": 0.034600000000000006, "residual_lr": 0.00020759999999999998, "adamw_lr": 0.002076, "weight_decay": 0.1, "grad_norm": 0.23958317935466766, "tokens_per_second": 3187591.2441115486, "eta_seconds": 7637.37765592509, "mean_data_wait_seconds": 8.091330528259277e-05}
{"step": 1220, "total_steps": 46238, "tokens": 660160512, "loss": 3.4803452834938513, "ce_loss": 3.479104210029949, "z_loss": 12.410801432349466, "muon_lr": 0.03488571428571429, "residual_lr": 0.0002093142857142857, "adamw_lr": 0.0020931428571428572, "weight_decay": 0.1, "grad_norm": 0.22083842754364014, "tokens_per_second": 3195982.2966710334, "eta_seconds": 7615.63399439108, "mean_data_wait_seconds": 8.293986320495605e-05}
{"step": 1230, "total_steps": 46238, "tokens": 665567232, "loss": 3.5055902726722485, "ce_loss": 3.504337032636007, "z_loss": 12.532276883269802, "muon_lr": 0.035171428571428576, "residual_lr": 0.0002110285714285714, "adamw_lr": 0.0021102857142857143, "weight_decay": 0.1, "grad_norm": 0.2304057776927948, "tokens_per_second": 3202894.8103211042, "eta_seconds": 7597.509798194218, "mean_data_wait_seconds": 7.861852645874023e-05}
{"step": 1240, "total_steps": 46238, "tokens": 670973952, "loss": 3.5086223591457713, "ce_loss": 3.5073887127818484, "z_loss": 12.33668877139236, "muon_lr": 0.03545714285714286, "residual_lr": 0.00021274285714285713, "adamw_lr": 0.0021274285714285713, "weight_decay": 0.1, "grad_norm": 0.22389572858810425, "tokens_per_second": 3209776.7255841093, "eta_seconds": 7579.5359191448815, "mean_data_wait_seconds": 9.66787338256836e-05}
{"step": 1250, "total_steps": 46238, "tokens": 676380672, "loss": 3.466648715915102, "ce_loss": 3.465415945558837, "z_loss": 12.327515031352188, "muon_lr": 0.035742857142857146, "residual_lr": 0.00021445714285714285, "adamw_lr": 0.002144571428571429, "weight_decay": 0.1, "grad_norm": 0.21136891841888428, "tokens_per_second": 3217160.0222940696, "eta_seconds": 7560.4604979070255, "mean_data_wait_seconds": 0.00016489624977111816}
{"step": 1260, "total_steps": 46238, "tokens": 681787392, "loss": 3.4881960821874216, "ce_loss": 3.4869422442985303, "z_loss": 12.53810006199461, "muon_lr": 0.03602857142857143, "residual_lr": 0.00021617142857142856, "adamw_lr": 0.0021617142857142855, "weight_decay": 0.1, "grad_norm": 0.26603227853775024, "tokens_per_second": 3224127.940386172, "eta_seconds": 7542.444032505521, "mean_data_wait_seconds": 0.00010192394256591797}
{"step": 1270, "total_steps": 46238, "tokens": 687194112, "loss": 3.480467935403188, "ce_loss": 3.4791498148080073, "z_loss": 13.180914503155332, "muon_lr": 0.036314285714285716, "residual_lr": 0.00021788571428571428, "adamw_lr": 0.002178857142857143, "weight_decay": 0.1, "grad_norm": 0.2906988561153412, "tokens_per_second": 3231579.2716387357, "eta_seconds": 7523.379679209035, "mean_data_wait_seconds": 8.738040924072266e-05}
{"step": 1280, "total_steps": 46238, "tokens": 692600832, "loss": 3.508553011850877, "ce_loss": 3.5072724837245364, "z_loss": 12.805801247105453, "muon_lr": 0.0366, "residual_lr": 0.0002196, "adamw_lr": 0.002196, "weight_decay": 0.1, "grad_norm": 0.2741829454898834, "tokens_per_second": 3236968.280499501, "eta_seconds": 7509.184211174647, "mean_data_wait_seconds": 0.00010356307029724121}
{"step": 1290, "total_steps": 46238, "tokens": 698007552, "loss": 3.49309119311246, "ce_loss": 3.4918367591771213, "z_loss": 12.544718699021773, "muon_lr": 0.036885714285714286, "residual_lr": 0.0002213142857142857, "adamw_lr": 0.002213142857142857, "weight_decay": 0.1, "grad_norm": 0.2462386041879654, "tokens_per_second": 3242474.262813198, "eta_seconds": 7494.7655445430555, "mean_data_wait_seconds": 9.176135063171387e-05}
{"step": 1300, "total_steps": 46238, "tokens": 703414272, "loss": 3.4848547632044014, "ce_loss": 3.483580455635533, "z_loss": 12.74289135499434, "muon_lr": 0.03717142857142857, "residual_lr": 0.0002230285714285714, "adamw_lr": 0.002230285714285714, "weight_decay": 0.1, "grad_norm": 0.20052245259284973, "tokens_per_second": 3248976.3372814455, "eta_seconds": 7478.102374955931, "mean_data_wait_seconds": 7.998943328857422e-05}
{"step": 1310, "total_steps": 46238, "tokens": 708820992, "loss": 3.4614190802429663, "ce_loss": 3.4601738705779566, "z_loss": 12.452067382407911, "muon_lr": 0.037457142857142856, "residual_lr": 0.00022474285714285712, "adamw_lr": 0.0022474285714285717, "weight_decay": 0.1, "grad_norm": 0.21331211924552917, "tokens_per_second": 3253570.071156588, "eta_seconds": 7465.882219455335, "mean_data_wait_seconds": 0.00015294551849365234}
{"step": 1320, "total_steps": 46238, "tokens": 714227712, "loss": 3.483688345461181, "ce_loss": 3.4824666994990725, "z_loss": 12.21670430356806, "muon_lr": 0.03774285714285714, "residual_lr": 0.00022645714285714284, "adamw_lr": 0.0022645714285714287, "weight_decay": 0.1, "grad_norm": 0.2284305840730667, "tokens_per_second": 3258229.51644487, "eta_seconds": 7453.546197843769, "mean_data_wait_seconds": 0.00012314319610595703}
{"step": 1330, "total_steps": 46238, "tokens": 719634432, "loss": 3.4486804189104023, "ce_loss": 3.447435203826789, "z_loss": 12.452074412143592, "muon_lr": 0.038028571428571425, "residual_lr": 0.00022817142857142856, "adamw_lr": 0.002281714285714286, "weight_decay": 0.1, "grad_norm": 0.2399924248456955, "tokens_per_second": 3263519.1122707785, "eta_seconds": 7439.808583534184, "mean_data_wait_seconds": 0.00011625885963439941}
{"step": 1340, "total_steps": 46238, "tokens": 725041152, "loss": 3.4369348829442803, "ce_loss": 3.4357188073071567, "z_loss": 12.160776333375411, "muon_lr": 0.03831428571428571, "residual_lr": 0.00022988571428571427, "adamw_lr": 0.002298857142857143, "weight_decay": 0.1, "grad_norm": 0.25334227085113525, "tokens_per_second": 3269381.1790216356, "eta_seconds": 7424.815111728322, "mean_data_wait_seconds": 0.00021958351135253906}
{"step": 1350, "total_steps": 46238, "tokens": 730447872, "loss": 3.465961288322102, "ce_loss": 3.4646735028787092, "z_loss": 12.87799564997355, "muon_lr": 0.038599999999999995, "residual_lr": 0.00023159999999999996, "adamw_lr": 0.002316, "weight_decay": 0.1, "grad_norm": 0.22006206214427948, "tokens_per_second": 3275108.1819386585, "eta_seconds": 7410.180890462736, "mean_data_wait_seconds": 9.900331497192383e-05}
{"step": 1360, "total_steps": 46238, "tokens": 735854592, "loss": 3.4863890409469604, "ce_loss": 3.4850996190851387, "z_loss": 12.893815329580596, "muon_lr": 0.03888571428571429, "residual_lr": 0.00023331428571428568, "adamw_lr": 0.0023331428571428574, "weight_decay": 0.1, "grad_norm": 0.247156023979187, "tokens_per_second": 3280567.4309118604, "eta_seconds": 7396.2013752162675, "mean_data_wait_seconds": 0.0001690983772277832}
{"step": 1370, "total_steps": 46238, "tokens": 741261312, "loss": 3.515314616940238, "ce_loss": 3.5140210552649065, "z_loss": 12.936194838899555, "muon_lr": 0.03917142857142857, "residual_lr": 0.0002350285714285714, "adamw_lr": 0.0023502857142857145, "weight_decay": 0.1, "grad_norm": 0.22689029574394226, "tokens_per_second": 3287403.3471486447, "eta_seconds": 7379.17683421496, "mean_data_wait_seconds": 7.346272468566895e-05}
{"step": 1380, "total_steps": 46238, "tokens": 746668032, "loss": 3.448874778819807, "ce_loss": 3.4475856134385774, "z_loss": 12.891487981333878, "muon_lr": 0.03945714285714286, "residual_lr": 0.00023674285714285712, "adamw_lr": 0.002367428571428571, "weight_decay": 0.1, "grad_norm": 0.2077181041240692, "tokens_per_second": 3293646.4953052667, "eta_seconds": 7363.547951660839, "mean_data_wait_seconds": 0.00010961294174194336}
{"step": 1390, "total_steps": 46238, "tokens": 752074752, "loss": 3.4657591888398835, "ce_loss": 3.464536325498061, "z_loss": 12.22877542177836, "muon_lr": 0.03974285714285714, "residual_lr": 0.00023845714285714283, "adamw_lr": 0.0023845714285714286, "weight_decay": 0.1, "grad_norm": 0.24885281920433044, "tokens_per_second": 3300144.7552363263, "eta_seconds": 7347.410184212848, "mean_data_wait_seconds": 7.590651512145996e-05}
{"step": 1400, "total_steps": 46238, "tokens": 757481472, "loss": 3.508916159470876, "ce_loss": 3.507583079916058, "z_loss": 13.330839294375796, "muon_lr": 0.04002857142857143, "residual_lr": 0.00024017142857142855, "adamw_lr": 0.0024017142857142857, "weight_decay": 0.1, "grad_norm": 0.265571266412735, "tokens_per_second": 3305606.299127683, "eta_seconds": 7333.635124787019, "mean_data_wait_seconds": 0.0001201927661895752}
{"step": 1410, "total_steps": 46238, "tokens": 762888192, "loss": 3.4301263708056826, "ce_loss": 3.4288228876662976, "z_loss": 13.035017815503208, "muon_lr": 0.04031428571428571, "residual_lr": 0.00024188571428571427, "adamw_lr": 0.0024188571428571427, "weight_decay": 0.1, "grad_norm": 0.2526622712612152, "tokens_per_second": 3311028.1269312827, "eta_seconds": 7319.993311703754, "mean_data_wait_seconds": 0.00010687112808227539}
{"step": 1420, "total_steps": 46238, "tokens": 768294912, "loss": 3.435164325165026, "ce_loss": 3.4338930682702498, "z_loss": 12.712812091365006, "muon_lr": 0.0406, "residual_lr": 0.00024359999999999999, "adamw_lr": 0.002436, "weight_decay": 0.1, "grad_norm": 0.2005140483379364, "tokens_per_second": 3316939.9065797073, "eta_seconds": 7305.316860258202, "mean_data_wait_seconds": 0.00010377168655395508}
{"step": 1430, "total_steps": 46238, "tokens": 773701632, "loss": 3.4704939470146643, "ce_loss": 3.469236787521478, "z_loss": 12.571522972800516, "muon_lr": 0.04088571428571428, "residual_lr": 0.0002453142857142857, "adamw_lr": 0.0024531428571428573, "weight_decay": 0.1, "grad_norm": 0.2290516197681427, "tokens_per_second": 3321717.0472085155, "eta_seconds": 7293.183001351306, "mean_data_wait_seconds": 7.811188697814941e-05}
{"step": 1440, "total_steps": 46238, "tokens": 779108352, "loss": 3.422162438883926, "ce_loss": 3.4209015820965623, "z_loss": 12.609164433045821, "muon_lr": 0.04117142857142857, "residual_lr": 0.0002470285714285714, "adamw_lr": 0.0024702857142857144, "weight_decay": 0.1, "grad_norm": 0.22147154808044434, "tokens_per_second": 3328794.260310349, "eta_seconds": 7276.053036014873, "mean_data_wait_seconds": 0.00012111663818359375}
{"step": 1450, "total_steps": 46238, "tokens": 784515072, "loss": 3.4428040475556343, "ce_loss": 3.441501442230109, "z_loss": 13.025907733223654, "muon_lr": 0.04145714285714285, "residual_lr": 0.0002487428571428571, "adamw_lr": 0.0024874285714285714, "weight_decay": 0.1, "grad_norm": 0.23751209676265717, "tokens_per_second": 3335626.59632546, "eta_seconds": 7259.5286566774075, "mean_data_wait_seconds": 7.972121238708496e-05}
{"step": 1460, "total_steps": 46238, "tokens": 789921792, "loss": 3.417783103205941, "ce_loss": 3.4165309721773323, "z_loss": 12.521298466306744, "muon_lr": 0.041742857142857144, "residual_lr": 0.0002504571428571428, "adamw_lr": 0.0025045714285714285, "weight_decay": 0.1, "grad_norm": 0.19974155724048615, "tokens_per_second": 3341400.389286271, "eta_seconds": 7245.366410330499, "mean_data_wait_seconds": 0.00011685490608215332}
{"step": 1470, "total_steps": 46238, "tokens": 795328512, "loss": 3.446620493224173, "ce_loss": 3.445254351153518, "z_loss": 13.661141287196767, "muon_lr": 0.04202857142857143, "residual_lr": 0.00025217142857142854, "adamw_lr": 0.002521714285714286, "weight_decay": 0.1, "grad_norm": 0.2603440582752228, "tokens_per_second": 3347872.7746913973, "eta_seconds": 7229.744095108607, "mean_data_wait_seconds": 7.641315460205078e-05}
{"step": 1480, "total_steps": 46238, "tokens": 800735232, "loss": 3.5028576670270977, "ce_loss": 3.5015007058779397, "z_loss": 13.569729711070206, "muon_lr": 0.042314285714285714, "residual_lr": 0.00025388571428571426, "adamw_lr": 0.002538857142857143, "weight_decay": 0.1, "grad_norm": 0.2515067756175995, "tokens_per_second": 3353757.617871426, "eta_seconds": 7215.445915068427, "mean_data_wait_seconds": 0.00010195374488830566}
{"step": 1490, "total_steps": 46238, "tokens": 806141952, "loss": 3.38147033886476, "ce_loss": 3.380147590781703, "z_loss": 13.227843754219286, "muon_lr": 0.0426, "residual_lr": 0.0002556, "adamw_lr": 0.002556, "weight_decay": 0.1, "grad_norm": 0.2198835015296936, "tokens_per_second": 3358759.7970683607, "eta_seconds": 7203.09026120798, "mean_data_wait_seconds": 7.605552673339844e-05}
{"step": 1500, "total_steps": 46238, "tokens": 811548672, "loss": 3.4402744914546157, "ce_loss": 3.439005770466544, "z_loss": 12.687035661755186, "muon_lr": 0.042885714285714284, "residual_lr": 0.0002573142857142857, "adamw_lr": 0.002573142857142857, "weight_decay": 0.1, "grad_norm": 0.21699558198451996, "tokens_per_second": 3364889.6712630647, "eta_seconds": 7188.361470086666, "mean_data_wait_seconds": 0.0001550912857055664}
{"step": 1510, "total_steps": 46238, "tokens": 816955392, "loss": 3.45277043725505, "ce_loss": 3.4514924956090525, "z_loss": 12.779436927853208, "muon_lr": 0.04317142857142857, "residual_lr": 0.0002590285714285714, "adamw_lr": 0.0025902857142857147, "weight_decay": 0.1, "grad_norm": 0.2536701261997223, "tokens_per_second": 3370103.4007903337, "eta_seconds": 7175.63637315367, "mean_data_wait_seconds": 0.00010645389556884766}
{"step": 1520, "total_steps": 46238, "tokens": 822362112, "loss": 3.4458591323910337, "ce_loss": 3.444493580948223, "z_loss": 13.65552925341057, "muon_lr": 0.043457142857142854, "residual_lr": 0.00026074285714285713, "adamw_lr": 0.0026074285714285713, "weight_decay": 0.1, "grad_norm": 0.24781936407089233, "tokens_per_second": 3374945.0562696806, "eta_seconds": 7163.740274552214, "mean_data_wait_seconds": 0.00010311603546142578}
{"step": 1530, "total_steps": 46238, "tokens": 827768832, "loss": 3.4326215892127068, "ce_loss": 3.431255053390156, "z_loss": 13.665238763346817, "muon_lr": 0.04374285714285714, "residual_lr": 0.00026245714285714285, "adamw_lr": 0.0026245714285714284, "weight_decay": 0.1, "grad_norm": 0.24475853145122528, "tokens_per_second": 3379799.380906025, "eta_seconds": 7151.851450283491, "mean_data_wait_seconds": 7.602572441101074e-05}
{"step": 1540, "total_steps": 46238, "tokens": 833175552, "loss": 3.3885484644860933, "ce_loss": 3.3872250752015547, "z_loss": 13.233841722661799, "muon_lr": 0.04402857142857144, "residual_lr": 0.0002641714285714285, "adamw_lr": 0.002641714285714286, "weight_decay": 0.1, "grad_norm": 0.20239129662513733, "tokens_per_second": 3385629.139360459, "eta_seconds": 7137.939623406302, "mean_data_wait_seconds": 0.00010076165199279785}
{"step": 1550, "total_steps": 46238, "tokens": 838582272, "loss": 3.4156967058326257, "ce_loss": 3.414417353543368, "z_loss": 12.794060092983823, "muon_lr": 0.04431428571428572, "residual_lr": 0.00026588571428571423, "adamw_lr": 0.002658857142857143, "weight_decay": 0.1, "grad_norm": 0.21897901594638824, "tokens_per_second": 3391479.5866363156, "eta_seconds": 7124.032165549018, "mean_data_wait_seconds": 6.99162483215332e-05}
{"step": 1560, "total_steps": 46238, "tokens": 843988992, "loss": 3.4088124307719143, "ce_loss": 3.4075024633696587, "z_loss": 13.09972123666243, "muon_lr": 0.04460000000000001, "residual_lr": 0.00026759999999999995, "adamw_lr": 0.002676, "weight_decay": 0.1, "grad_norm": 0.2238929271697998, "tokens_per_second": 3397186.0513985083, "eta_seconds": 7110.473956542929, "mean_data_wait_seconds": 6.362795829772949e-05}
{"step": 1570, "total_steps": 46238, "tokens": 849395712, "loss": 3.4184354561747927, "ce_loss": 3.4171111655957773, "z_loss": 13.242787837982178, "muon_lr": 0.04488571428571429, "residual_lr": 0.00026931428571428566, "adamw_lr": 0.002693142857142857, "weight_decay": 0.1, "grad_norm": 0.26395997405052185, "tokens_per_second": 3402057.467077909, "eta_seconds": 7098.703198786074, "mean_data_wait_seconds": 0.00011095404624938965}
{"step": 1580, "total_steps": 46238, "tokens": 854802432, "loss": 3.360869257739096, "ce_loss": 3.359540682850462, "z_loss": 13.285905397299564, "muon_lr": 0.04517142857142858, "residual_lr": 0.0002710285714285714, "adamw_lr": 0.0027102857142857146, "weight_decay": 0.1, "grad_norm": 0.2330450862646103, "tokens_per_second": 3407847.261234156, "eta_seconds": 7085.056240242392, "mean_data_wait_seconds": 6.41942024230957e-05}
{"step": 1590, "total_steps": 46238, "tokens": 860209152, "loss": 3.3179634755307976, "ce_loss": 3.3166052373972805, "z_loss": 13.58255285205263, "muon_lr": 0.04545714285714286, "residual_lr": 0.0002727428571428571, "adamw_lr": 0.002727428571428571, "weight_decay": 0.1, "grad_norm": 0.22231703996658325, "tokens_per_second": 3413597.078260599, "eta_seconds": 7071.53838914704, "mean_data_wait_seconds": 8.928775787353516e-05}
{"step": 1600, "total_steps": 46238, "tokens": 865615872, "loss": 3.4032477032054556, "ce_loss": 3.4018830024834834, "z_loss": 13.647081447370125, "muon_lr": 0.04574285714285715, "residual_lr": 0.0002744571428571428, "adamw_lr": 0.0027445714285714287, "weight_decay": 0.1, "grad_norm": 0.22695983946323395, "tokens_per_second": 3418428.733371021, "eta_seconds": 7059.961738678904, "mean_data_wait_seconds": 0.00019019842147827148}
{"step": 1610, "total_steps": 46238, "tokens": 871022592, "loss": 3.378206903284246, "ce_loss": 3.3768401579423384, "z_loss": 13.667116981564146, "muon_lr": 0.04602857142857143, "residual_lr": 0.00027617142857142853, "adamw_lr": 0.0027617142857142857, "weight_decay": 0.1, "grad_norm": 0.23828817903995514, "tokens_per_second": 3424566.762926016, "eta_seconds": 7045.728997084608, "mean_data_wait_seconds": 6.052851676940918e-05}
{"step": 1620, "total_steps": 46238, "tokens": 876429312, "loss": 3.3870214494791897, "ce_loss": 3.385667678081628, "z_loss": 13.537830966891665, "muon_lr": 0.04631428571428572, "residual_lr": 0.00027788571428571425, "adamw_lr": 0.0027788571428571432, "weight_decay": 0.1, "grad_norm": 0.22636477649211884, "tokens_per_second": 3429247.660448567, "eta_seconds": 7034.534980434903, "mean_data_wait_seconds": 0.00010347366333007812}
{"step": 1630, "total_steps": 46238, "tokens": 881836032, "loss": 3.3904809753100076, "ce_loss": 3.389174869566253, "z_loss": 13.060919963952268, "muon_lr": 0.04660000000000001, "residual_lr": 0.00027959999999999997, "adamw_lr": 0.002796, "weight_decay": 0.1, "grad_norm": 0.2011173814535141, "tokens_per_second": 3423408.563681424, "eta_seconds": 7044.9540145055125, "mean_data_wait_seconds": 6.085634231567383e-05}
{"step": 1640, "total_steps": 46238, "tokens": 887242752, "loss": 3.414105357545795, "ce_loss": 3.4127793871995173, "z_loss": 13.259436014926795, "muon_lr": 0.046885714285714294, "residual_lr": 0.0002813142857142857, "adamw_lr": 0.0028131428571428574, "weight_decay": 0.1, "grad_norm": 0.19718694686889648, "tokens_per_second": 3429299.2102481104, "eta_seconds": 7031.275985467441, "mean_data_wait_seconds": 5.474686622619629e-05}
{"step": 1650, "total_steps": 46238, "tokens": 892649472, "loss": 3.4383740479295906, "ce_loss": 3.4370849999514492, "z_loss": 12.890572995850533, "muon_lr": 0.04717142857142858, "residual_lr": 0.0002830285714285714, "adamw_lr": 0.0028302857142857144, "weight_decay": 0.1, "grad_norm": 0.2328481823205948, "tokens_per_second": 3425390.1119129523, "eta_seconds": 7037.721741579144, "mean_data_wait_seconds": 0.00010007619857788086}
{"step": 1660, "total_steps": 46238, "tokens": 898056192, "loss": 3.373906616008643, "ce_loss": 3.3726038300629817, "z_loss": 13.027710885712594, "muon_lr": 0.047457142857142864, "residual_lr": 0.0002847428571428571, "adamw_lr": 0.002847428571428572, "weight_decay": 0.1, "grad_norm": 0.20930340886116028, "tokens_per_second": 3429692.1622298807, "eta_seconds": 7027.3174978858515, "mean_data_wait_seconds": 9.495019912719727e-05}
{"step": 1670, "total_steps": 46238, "tokens": 903462912, "loss": 3.4219625014247317, "ce_loss": 3.4206554058826333, "z_loss": 13.070918220462222, "muon_lr": 0.04774285714285715, "residual_lr": 0.00028645714285714283, "adamw_lr": 0.0028645714285714286, "weight_decay": 0.1, "grad_norm": 0.21416185796260834, "tokens_per_second": 3432091.247399475, "eta_seconds": 7020.829950910496, "mean_data_wait_seconds": 8.180737495422363e-05}
{"step": 1680, "total_steps": 46238, "tokens": 908869632, "loss": 3.402523931228753, "ce_loss": 3.4012467174819023, "z_loss": 12.772131312977184, "muon_lr": 0.048028571428571434, "residual_lr": 0.00028817142857142855, "adamw_lr": 0.0028817142857142856, "weight_decay": 0.1, "grad_norm": 0.24153965711593628, "tokens_per_second": 3436058.5009891023, "eta_seconds": 7011.1502167571525, "mean_data_wait_seconds": 0.00012961030006408691}
{"step": 1690, "total_steps": 46238, "tokens": 914276352, "loss": 3.3527063727378845, "ce_loss": 3.3514526558644846, "z_loss": 12.536862907987652, "muon_lr": 0.04831428571428572, "residual_lr": 0.00028988571428571427, "adamw_lr": 0.002898857142857143, "weight_decay": 0.1, "grad_norm": 0.246310755610466, "tokens_per_second": 3439443.793229644, "eta_seconds": 7002.677476925373, "mean_data_wait_seconds": 0.0001500248908996582}
{"step": 1700, "total_steps": 46238, "tokens": 919683072, "loss": 3.4147503285697014, "ce_loss": 3.413402425520348, "z_loss": 13.479236349915013, "muon_lr": 0.048600000000000004, "residual_lr": 0.0002916, "adamw_lr": 0.0029159999999999998, "weight_decay": 0.1, "grad_norm": 0.2403583973646164, "tokens_per_second": 3443773.9267681376, "eta_seconds": 6992.302449597253, "mean_data_wait_seconds": 0.00010186433792114258}
{"step": 1710, "total_steps": 46238, "tokens": 925089792, "loss": 3.414756527452758, "ce_loss": 3.4133841250882004, "z_loss": 13.724292885173451, "muon_lr": 0.04888571428571429, "residual_lr": 0.0002933142857142857, "adamw_lr": 0.0029331428571428573, "weight_decay": 0.1, "grad_norm": 0.20169296860694885, "tokens_per_second": 3447992.2696248707, "eta_seconds": 6982.17984886005, "mean_data_wait_seconds": 0.0001061558723449707}
{"step": 1720, "total_steps": 46238, "tokens": 930496512, "loss": 3.398721998388117, "ce_loss": 3.3973921410965198, "z_loss": 13.298716285011984, "muon_lr": 0.04917142857142858, "residual_lr": 0.0002950285714285714, "adamw_lr": 0.0029502857142857143, "weight_decay": 0.1, "grad_norm": 0.21618150174617767, "tokens_per_second": 3451552.583140249, "eta_seconds": 6973.411195173435, "mean_data_wait_seconds": 7.262825965881348e-05}
{"step": 1730, "total_steps": 46238, "tokens": 935903232, "loss": 3.39344305341894, "ce_loss": 3.39210290258581, "z_loss": 13.401599884033203, "muon_lr": 0.049457142857142866, "residual_lr": 0.00029674285714285714, "adamw_lr": 0.002967428571428572, "weight_decay": 0.1, "grad_norm": 0.19978758692741394, "tokens_per_second": 3449450.498048647, "eta_seconds": 6976.093356786196, "mean_data_wait_seconds": 7.426738739013672e-05}
{"step": 1740, "total_steps": 46238, "tokens": 941309952, "loss": 3.46522363207557, "ce_loss": 3.4639787168213814, "z_loss": 12.44905861941251, "muon_lr": 0.04974285714285715, "residual_lr": 0.00029845714285714286, "adamw_lr": 0.0029845714285714284, "weight_decay": 0.1, "grad_norm": 0.28209295868873596, "tokens_per_second": 3445251.264069283, "eta_seconds": 6983.026821557301, "mean_data_wait_seconds": 6.082653999328613e-05}
{"step": 1750, "total_steps": 46238, "tokens": 946716672, "loss": 3.4352468288306035, "ce_loss": 3.4338687459627786, "z_loss": 13.780630588531494, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18005497753620148, "tokens_per_second": 3442011.555028571, "eta_seconds": 6988.028622059738, "mean_data_wait_seconds": 7.203221321105957e-05}
{"step": 1760, "total_steps": 46238, "tokens": 952123392, "loss": 3.3584633993379995, "ce_loss": 3.357138655402444, "z_loss": 13.247296918522228, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21155498921871185, "tokens_per_second": 3445880.2607680564, "eta_seconds": 6978.614091088595, "mean_data_wait_seconds": 0.00013133883476257324}
{"step": 1770, "total_steps": 46238, "tokens": 957530112, "loss": 3.492308562452143, "ce_loss": 3.491004055196589, "z_loss": 13.045048468040697, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.23488639295101166, "tokens_per_second": 3448535.9979003645, "eta_seconds": 6971.671990270065, "mean_data_wait_seconds": 0.00011387467384338379}
{"step": 1780, "total_steps": 46238, "tokens": 962936832, "loss": 3.4346147053169482, "ce_loss": 3.433255271478133, "z_loss": 13.594616470914898, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19827745854854584, "tokens_per_second": 3453901.0512506184, "eta_seconds": 6959.277277296812, "mean_data_wait_seconds": 9.506940841674805e-05}
{"step": 1790, "total_steps": 46238, "tokens": 968343552, "loss": 3.415206481109966, "ce_loss": 3.413857568394054, "z_loss": 13.48904419667793, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17726247012615204, "tokens_per_second": 3458700.2678910033, "eta_seconds": 6948.057513712638, "mean_data_wait_seconds": 0.00024196505546569824}
{"step": 1800, "total_steps": 46238, "tokens": 973750272, "loss": 3.511017615144903, "ce_loss": 3.5096335501381843, "z_loss": 13.84084710930333, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22702719271183014, "tokens_per_second": 3463201.741815242, "eta_seconds": 6937.465228752981, "mean_data_wait_seconds": 9.563565254211426e-05}
{"step": 1810, "total_steps": 46238, "tokens": 979156992, "loss": 3.3920922044551736, "ce_loss": 3.3907487338239495, "z_loss": 13.43463591373328, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19402948021888733, "tokens_per_second": 3466238.051148322, "eta_seconds": 6929.828416153451, "mean_data_wait_seconds": 8.469820022583008e-05}
{"step": 1820, "total_steps": 46238, "tokens": 984563712, "loss": 3.4012454188231267, "ce_loss": 3.3998569575223057, "z_loss": 13.884515285491943, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2170161008834839, "tokens_per_second": 3468734.8026090693, "eta_seconds": 6923.281712380168, "mean_data_wait_seconds": 6.830692291259766e-05}
{"step": 1830, "total_steps": 46238, "tokens": 989970432, "loss": 3.4863325808987473, "ce_loss": 3.4849815675706575, "z_loss": 13.510396379413027, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1857835054397583, "tokens_per_second": 3472950.8455976495, "eta_seconds": 6913.320277606249, "mean_data_wait_seconds": 0.00011149048805236816}
{"step": 1840, "total_steps": 46238, "tokens": 995377152, "loss": 3.4629869894547896, "ce_loss": 3.461562597390377, "z_loss": 14.244421749403983, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.24250531196594238, "tokens_per_second": 3476672.175143157, "eta_seconds": 6904.365316816674, "mean_data_wait_seconds": 8.374452590942383e-05}
{"step": 1850, "total_steps": 46238, "tokens": 1000783872, "loss": 3.4566193891294077, "ce_loss": 3.455264860933477, "z_loss": 13.54510177265514, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2292131930589676, "tokens_per_second": 3481326.996220809, "eta_seconds": 6893.580548466765, "mean_data_wait_seconds": 6.23166561126709e-05}
{"step": 1860, "total_steps": 46238, "tokens": 1006190592, "loss": 3.4575669024929856, "ce_loss": 3.45619207981861, "z_loss": 13.748175794428045, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19275951385498047, "tokens_per_second": 3486809.9163366137, "eta_seconds": 6881.189947173391, "mean_data_wait_seconds": 6.16908073425293e-05}
{"step": 1870, "total_steps": 46238, "tokens": 1011597312, "loss": 3.4384278853734336, "ce_loss": 3.437069168596557, "z_loss": 13.587232900388313, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2323630303144455, "tokens_per_second": 3490717.6554322774, "eta_seconds": 6871.9378053019345, "mean_data_wait_seconds": 0.0001011192798614502}
{"step": 1880, "total_steps": 46238, "tokens": 1017004032, "loss": 3.4420368219866897, "ce_loss": 3.4407211524067502, "z_loss": 13.15673390301791, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22479687631130219, "tokens_per_second": 3496113.009905147, "eta_seconds": 6859.786235757485, "mean_data_wait_seconds": 0.00013023614883422852}
{"step": 1890, "total_steps": 46238, "tokens": 1022410752, "loss": 3.479554174524365, "ce_loss": 3.4782473390752617, "z_loss": 13.068314675128821, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21198050677776337, "tokens_per_second": 3499898.484580814, "eta_seconds": 6850.821899444826, "mean_data_wait_seconds": 9.277462959289551e-05}
{"step": 1900, "total_steps": 46238, "tokens": 1027817472, "loss": 3.41043484210968, "ce_loss": 3.4090810147198765, "z_loss": 13.538162332592588, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17207162082195282, "tokens_per_second": 3503690.937064906, "eta_seconds": 6841.86330774983, "mean_data_wait_seconds": 9.369850158691406e-05}
{"step": 1910, "total_steps": 46238, "tokens": 1033224192, "loss": 3.4164377830245276, "ce_loss": 3.4151105989109385, "z_loss": 13.271885828538375, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22150124609470367, "tokens_per_second": 3507801.1649607522, "eta_seconds": 6832.305087129462, "mean_data_wait_seconds": 8.288025856018066e-05}
{"step": 1920, "total_steps": 46238, "tokens": 1038630912, "loss": 3.4171946355790803, "ce_loss": 3.415804442131158, "z_loss": 13.902028459491152, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18910792469978333, "tokens_per_second": 3512491.4826291907, "eta_seconds": 6821.642455931196, "mean_data_wait_seconds": 7.268786430358887e-05}
{"step": 1930, "total_steps": 46238, "tokens": 1044037632, "loss": 3.432517662192836, "ce_loss": 3.4311704400813943, "z_loss": 13.472473411849052, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20495766401290894, "tokens_per_second": 3516400.7841376425, "eta_seconds": 6812.5210334563235, "mean_data_wait_seconds": 9.149312973022461e-05}
{"step": 1940, "total_steps": 46238, "tokens": 1049444352, "loss": 3.4616172277566157, "ce_loss": 3.460224059495059, "z_loss": 13.931661078424165, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20423918962478638, "tokens_per_second": 3519579.2429757942, "eta_seconds": 6804.832603726296, "mean_data_wait_seconds": 9.891390800476074e-05}
{"step": 1950, "total_steps": 46238, "tokens": 1054851072, "loss": 3.3719019131226973, "ce_loss": 3.370490581700296, "z_loss": 14.113608967174184, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16032378375530243, "tokens_per_second": 3523488.3774457737, "eta_seconds": 6795.748502328785, "mean_data_wait_seconds": 0.00010979175567626953}
{"step": 1960, "total_steps": 46238, "tokens": 1060257792, "loss": 3.432850057428533, "ce_loss": 3.431527336438497, "z_loss": 13.227123816808065, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.25252681970596313, "tokens_per_second": 3527773.731882377, "eta_seconds": 6785.960768301958, "mean_data_wait_seconds": 0.00012600421905517578}
{"step": 1970, "total_steps": 46238, "tokens": 1065664512, "loss": 3.4160546877167444, "ce_loss": 3.4147019350167476, "z_loss": 13.527431516936332, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20774073898792267, "tokens_per_second": 3532085.1967705186, "eta_seconds": 6776.146692577925, "mean_data_wait_seconds": 8.338689804077148e-05}
{"step": 1980, "total_steps": 46238, "tokens": 1071071232, "loss": 3.4124796336347405, "ce_loss": 3.4111274643377825, "z_loss": 13.521967512188535, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19853806495666504, "tokens_per_second": 3536463.613733794, "eta_seconds": 6766.228446709874, "mean_data_wait_seconds": 5.6684017181396484e-05}
{"step": 1990, "total_steps": 46238, "tokens": 1076477952, "loss": 3.4229091279434436, "ce_loss": 3.421560041832201, "z_loss": 13.490712657119289, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17709651589393616, "tokens_per_second": 3539984.239631688, "eta_seconds": 6757.971890431083, "mean_data_wait_seconds": 6.893277168273926e-05}
{"step": 2000, "total_steps": 46238, "tokens": 1081884672, "loss": 3.417064509608529, "ce_loss": 3.4156623753634365, "z_loss": 14.021145965113785, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19148924946784973, "tokens_per_second": 3543520.4332287004, "eta_seconds": 6749.70208714367, "mean_data_wait_seconds": 7.155537605285645e-05}
{"step": 2010, "total_steps": 46238, "tokens": 1087291392, "loss": 3.463429396802729, "ce_loss": 3.46208412177635, "z_loss": 13.452694885658495, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2059415876865387, "tokens_per_second": 3546582.349421341, "eta_seconds": 6742.3502933469235, "mean_data_wait_seconds": 7.581710815429688e-05}
{"step": 2020, "total_steps": 46238, "tokens": 1092698112, "loss": 3.4194503914226186, "ce_loss": 3.418074808337472, "z_loss": 13.755792328805635, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1760665625333786, "tokens_per_second": 3549073.89170464, "eta_seconds": 6736.093570742024, "mean_data_wait_seconds": 9.131431579589844e-05}
{"step": 2030, "total_steps": 46238, "tokens": 1098104832, "loss": 3.3695272731058523, "ce_loss": 3.3681885238849754, "z_loss": 13.387442046945745, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.202083557844162, "tokens_per_second": 3552546.6728160786, "eta_seconds": 6727.9867951892265, "mean_data_wait_seconds": 6.872415542602539e-05}
{"step": 2040, "total_steps": 46238, "tokens": 1103511552, "loss": 3.405903671727036, "ce_loss": 3.404483645251303, "z_loss": 14.200160532286674, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16701668500900269, "tokens_per_second": 3555957.520683431, "eta_seconds": 6720.012892450789, "mean_data_wait_seconds": 5.701184272766113e-05}
{"step": 2050, "total_steps": 46238, "tokens": 1108918272, "loss": 3.3454629309249646, "ce_loss": 3.344072367205764, "z_loss": 13.905510808482315, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17836986482143402, "tokens_per_second": 3559415.0725169205, "eta_seconds": 6711.966201544041, "mean_data_wait_seconds": 7.891654968261719e-05}
{"step": 2060, "total_steps": 46238, "tokens": 1114324992, "loss": 3.4095678979700264, "ce_loss": 3.408154173330827, "z_loss": 14.136995142156428, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1905592978000641, "tokens_per_second": 3564082.2590455394, "eta_seconds": 6701.65984058866, "mean_data_wait_seconds": 6.48200511932373e-05}
{"step": 2070, "total_steps": 46238, "tokens": 1119731712, "loss": 3.3277058872309597, "ce_loss": 3.3262852773521887, "z_loss": 14.206134478251139, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18958327174186707, "tokens_per_second": 3566858.5346136787, "eta_seconds": 6694.927761295808, "mean_data_wait_seconds": 6.505846977233887e-05}
{"step": 2080, "total_steps": 46238, "tokens": 1125138432, "loss": 3.335522658897169, "ce_loss": 3.3341427806651955, "z_loss": 13.798813624815507, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19515502452850342, "tokens_per_second": 3570581.604662356, "eta_seconds": 6686.432673272464, "mean_data_wait_seconds": 0.00010582804679870605}
{"step": 2090, "total_steps": 46238, "tokens": 1130545152, "loss": 3.3343885324218054, "ce_loss": 3.332927877252752, "z_loss": 14.60675590688532, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18249817192554474, "tokens_per_second": 3574784.22946041, "eta_seconds": 6677.059439641445, "mean_data_wait_seconds": 6.368756294250488e-05}
{"step": 2100, "total_steps": 46238, "tokens": 1135951872, "loss": 3.3767003380891047, "ce_loss": 3.3753204291517083, "z_loss": 13.799031835613828, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17630605399608612, "tokens_per_second": 3578220.280459542, "eta_seconds": 6669.136663921443, "mean_data_wait_seconds": 0.00022488832473754883}
{"step": 2110, "total_steps": 46238, "tokens": 1141358592, "loss": 3.3166744763200935, "ce_loss": 3.315273992943041, "z_loss": 14.004634445363825, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1875944435596466, "tokens_per_second": 3580936.503024969, "eta_seconds": 6662.568108606768, "mean_data_wait_seconds": 8.577108383178711e-05}
{"step": 2120, "total_steps": 46238, "tokens": 1146765312, "loss": 3.346503859216517, "ce_loss": 3.3451360644716206, "z_loss": 13.678050330190947, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21158112585544586, "tokens_per_second": 3583589.1188346804, "eta_seconds": 6656.127651084205, "mean_data_wait_seconds": 0.00015988945960998535}
{"step": 2130, "total_steps": 46238, "tokens": 1152172032, "loss": 3.379041688008742, "ce_loss": 3.3776221672693887, "z_loss": 14.195161378744876, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17342153191566467, "tokens_per_second": 3586945.3963718615, "eta_seconds": 6648.392230369966, "mean_data_wait_seconds": 0.00011548399925231934}
{"step": 2140, "total_steps": 46238, "tokens": 1157578752, "loss": 3.3338667967102746, "ce_loss": 3.3324268207405554, "z_loss": 14.399292721892849, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19564084708690643, "tokens_per_second": 3589861.501834108, "eta_seconds": 6641.485520212632, "mean_data_wait_seconds": 8.383393287658691e-05}
{"step": 2150, "total_steps": 46238, "tokens": 1162985472, "loss": 3.2939616965525076, "ce_loss": 3.2925750450654463, "z_loss": 13.866478479269778, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17538219690322876, "tokens_per_second": 3591516.9523171075, "eta_seconds": 6636.918823011972, "mean_data_wait_seconds": 0.00016799569129943848}
{"step": 2160, "total_steps": 46238, "tokens": 1168392192, "loss": 3.3462985573392925, "ce_loss": 3.344939445004319, "z_loss": 13.591356949372726, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17645475268363953, "tokens_per_second": 3593308.728619909, "eta_seconds": 6632.104710121278, "mean_data_wait_seconds": 0.00016871094703674316}
{"step": 2170, "total_steps": 46238, "tokens": 1173798912, "loss": 3.3660065296924477, "ce_loss": 3.3646039655714324, "z_loss": 14.025949824940074, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.170454740524292, "tokens_per_second": 3595033.744622393, "eta_seconds": 6627.418465720843, "mean_data_wait_seconds": 8.153915405273438e-05}
{"step": 2180, "total_steps": 46238, "tokens": 1179205632, "loss": 3.3966421430761162, "ce_loss": 3.395194512425047, "z_loss": 14.475920156999068, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2034163624048233, "tokens_per_second": 3598117.8251062525, "eta_seconds": 6620.2352068047085, "mean_data_wait_seconds": 0.00010904669761657715}
{"step": 2190, "total_steps": 46238, "tokens": 1184612352, "loss": 3.353055087002841, "ce_loss": 3.3516265551249185, "z_loss": 14.285406184918953, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18624195456504822, "tokens_per_second": 3601821.9771726085, "eta_seconds": 6611.925779489663, "mean_data_wait_seconds": 6.54757022857666e-05}
{"step": 2200, "total_steps": 46238, "tokens": 1190019072, "loss": 3.363243184306405, "ce_loss": 3.361825061566902, "z_loss": 14.181133132992368, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1830615997314453, "tokens_per_second": 3605736.670605583, "eta_seconds": 6603.247835067552, "mean_data_wait_seconds": 8.687376976013184e-05}
{"step": 2210, "total_steps": 46238, "tokens": 1195425792, "loss": 3.327766436519045, "ce_loss": 3.3262928904909077, "z_loss": 14.735343651338058, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.23723307251930237, "tokens_per_second": 3608551.5160373184, "eta_seconds": 6596.598673514358, "mean_data_wait_seconds": 7.203221321105957e-05}
{"step": 2220, "total_steps": 46238, "tokens": 1200832512, "loss": 3.3524628408027417, "ce_loss": 3.351046905373082, "z_loss": 14.158968065724228, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16742071509361267, "tokens_per_second": 3611012.8114432883, "eta_seconds": 6590.6050924499095, "mean_data_wait_seconds": 0.0001327991485595703}
{"step": 2230, "total_steps": 46238, "tokens": 1206239232, "loss": 3.3405517430016487, "ce_loss": 3.339129388332367, "z_loss": 14.223787632855503, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18551596999168396, "tokens_per_second": 3615025.3712328617, "eta_seconds": 6581.794112245901, "mean_data_wait_seconds": 6.204843521118164e-05}
{"step": 2240, "total_steps": 46238, "tokens": 1211645952, "loss": 3.3871507102792915, "ce_loss": 3.385786305774342, "z_loss": 13.644201105291193, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2123870700597763, "tokens_per_second": 3617322.4859847706, "eta_seconds": 6576.119789199284, "mean_data_wait_seconds": 8.827447891235352e-05}
{"step": 2250, "total_steps": 46238, "tokens": 1217052672, "loss": 3.332081776676756, "ce_loss": 3.3307055776769463, "z_loss": 13.762096686796708, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17566822469234467, "tokens_per_second": 3618868.420449835, "eta_seconds": 6571.816518557966, "mean_data_wait_seconds": 0.00011780858039855957}
{"step": 2260, "total_steps": 46238, "tokens": 1222459392, "loss": 3.375756531050711, "ce_loss": 3.374304469787713, "z_loss": 14.520528793334961, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21705850958824158, "tokens_per_second": 3621543.642004156, "eta_seconds": 6565.469008359589, "mean_data_wait_seconds": 0.00010710954666137695}
{"step": 2270, "total_steps": 46238, "tokens": 1227866112, "loss": 3.368509130044417, "ce_loss": 3.367068789222024, "z_loss": 14.403319272128018, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20222684741020203, "tokens_per_second": 3623989.5791735183, "eta_seconds": 6559.545855377803, "mean_data_wait_seconds": 9.956955909729004e-05}
{"step": 2280, "total_steps": 46238, "tokens": 1233272832, "loss": 3.3820693872191687, "ce_loss": 3.3806093840888054, "z_loss": 14.6000597187967, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19278021156787872, "tokens_per_second": 3627057.026430249, "eta_seconds": 6552.5077027506295, "mean_data_wait_seconds": 8.720159530639648e-05}
{"step": 2290, "total_steps": 46238, "tokens": 1238679552, "loss": 3.360732811869997, "ce_loss": 3.3593458822279265, "z_loss": 13.868810704260161, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19178996980190277, "tokens_per_second": 3629414.9901077375, "eta_seconds": 6546.760965269135, "mean_data_wait_seconds": 9.581446647644043e-05}
{"step": 2300, "total_steps": 46238, "tokens": 1244086272, "loss": 3.305506939237768, "ce_loss": 3.3041580789016955, "z_loss": 13.488589185656924, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20834439992904663, "tokens_per_second": 3631981.3037199005, "eta_seconds": 6540.646461937854, "mean_data_wait_seconds": 0.00012978911399841309}
{"step": 2310, "total_steps": 46238, "tokens": 1249492992, "loss": 3.4174927870432534, "ce_loss": 3.4161423553120005, "z_loss": 13.504370797764171, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.209946870803833, "tokens_per_second": 3634880.886113906, "eta_seconds": 6533.941465518424, "mean_data_wait_seconds": 7.340312004089355e-05}
{"step": 2320, "total_steps": 46238, "tokens": 1254899712, "loss": 3.369710618799383, "ce_loss": 3.368265581853462, "z_loss": 14.450556841763584, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19592110812664032, "tokens_per_second": 3638379.1167983413, "eta_seconds": 6526.1731836495865, "mean_data_wait_seconds": 7.006525993347168e-05}
{"step": 2330, "total_steps": 46238, "tokens": 1260306432, "loss": 3.4010061675851997, "ce_loss": 3.3995680393594685, "z_loss": 14.381374619223855, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18881580233573914, "tokens_per_second": 3642059.5513807433, "eta_seconds": 6518.093723920628, "mean_data_wait_seconds": 5.799531936645508e-05}
{"step": 2340, "total_steps": 46238, "tokens": 1265713152, "loss": 3.380946598269723, "ce_loss": 3.3795335870800596, "z_loss": 14.129521637251884, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17450852692127228, "tokens_per_second": 3645493.7972099097, "eta_seconds": 6510.470214533022, "mean_data_wait_seconds": 0.00011271238327026367}
{"step": 2350, "total_steps": 46238, "tokens": 1271119872, "loss": 3.4191090699398154, "ce_loss": 3.417733517560092, "z_loss": 13.755572506875703, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21849603950977325, "tokens_per_second": 3648042.127087202, "eta_seconds": 6504.440255174937, "mean_data_wait_seconds": 8.928775787353516e-05}
{"step": 2360, "total_steps": 46238, "tokens": 1276526592, "loss": 3.420347542473764, "ce_loss": 3.4189152681466304, "z_loss": 14.322847366333008, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20286723971366882, "tokens_per_second": 3652006.539255507, "eta_seconds": 6495.89892268817, "mean_data_wait_seconds": 0.00011387467384338379}
{"step": 2370, "total_steps": 46238, "tokens": 1281933312, "loss": 3.3603278976498228, "ce_loss": 3.3588560855749883, "z_loss": 14.718001445134481, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1866539716720581, "tokens_per_second": 3655202.8112203637, "eta_seconds": 6488.739434975806, "mean_data_wait_seconds": 8.90493392944336e-05}
{"step": 2380, "total_steps": 46238, "tokens": 1287340032, "loss": 3.3850035631295405, "ce_loss": 3.3835689425468445, "z_loss": 14.34615168426976, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17349594831466675, "tokens_per_second": 3658687.99745505, "eta_seconds": 6481.080627944779, "mean_data_wait_seconds": 6.598234176635742e-05}
{"step": 2390, "total_steps": 46238, "tokens": 1292746752, "loss": 3.44716739654541, "ce_loss": 3.445792823126822, "z_loss": 13.745926430731108, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21422316133975983, "tokens_per_second": 3661817.2029040307, "eta_seconds": 6474.065708468221, "mean_data_wait_seconds": 0.0001119077205657959}
{"step": 2400, "total_steps": 46238, "tokens": 1298153472, "loss": 3.4259219151554685, "ce_loss": 3.424512474825888, "z_loss": 14.094246177962333, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1719534695148468, "tokens_per_second": 3663866.709103387, "eta_seconds": 6468.968536740291, "mean_data_wait_seconds": 0.0002282559871673584}
{"step": 2410, "total_steps": 46238, "tokens": 1303560192, "loss": 3.3613495031992593, "ce_loss": 3.3599630305261323, "z_loss": 13.864703879211888, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18252292275428772, "tokens_per_second": 3667397.0100442795, "eta_seconds": 6461.267127366147, "mean_data_wait_seconds": 0.0001163184642791748}
{"step": 2420, "total_steps": 46238, "tokens": 1308966912, "loss": 3.378290118593158, "ce_loss": 3.3768556732119936, "z_loss": 14.344323194388187, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17409655451774597, "tokens_per_second": 3670756.6506618885, "eta_seconds": 6453.880569753446, "mean_data_wait_seconds": 0.00013262033462524414}
{"step": 2430, "total_steps": 46238, "tokens": 1314373632, "loss": 3.385606718785835, "ce_loss": 3.384134643005602, "z_loss": 14.72103444735209, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2012355625629425, "tokens_per_second": 3673873.732154943, "eta_seconds": 6446.933136732281, "mean_data_wait_seconds": 7.840991020202637e-05}
{"step": 2440, "total_steps": 46238, "tokens": 1319780352, "loss": 3.3777915528326323, "ce_loss": 3.3763857863166113, "z_loss": 14.05822676600832, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18445934355258942, "tokens_per_second": 3675904.084280348, "eta_seconds": 6441.901377477298, "mean_data_wait_seconds": 0.00013774633407592773}
{"step": 2450, "total_steps": 46238, "tokens": 1325187072, "loss": 3.3863896781747993, "ce_loss": 3.3850350777308145, "z_loss": 13.546104777943004, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.220076784491539, "tokens_per_second": 3679458.3236456355, "eta_seconds": 6434.209272560321, "mean_data_wait_seconds": 7.12275505065918e-05}
{"step": 2460, "total_steps": 46238, "tokens": 1330593792, "loss": 3.4216629230614863, "ce_loss": 3.4202534664760935, "z_loss": 14.094605799877282, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1843060553073883, "tokens_per_second": 3682952.1509198067, "eta_seconds": 6426.637429456892, "mean_data_wait_seconds": 0.0005404055118560791}
{"step": 2470, "total_steps": 46238, "tokens": 1336000512, "loss": 3.415848784374468, "ce_loss": 3.4144721645297427, "z_loss": 13.766485344279896, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2059323638677597, "tokens_per_second": 3685873.3272519065, "eta_seconds": 6420.077230826316, "mean_data_wait_seconds": 0.0001575946807861328}
{"step": 2480, "total_steps": 46238, "tokens": 1341407232, "loss": 3.409594774246216, "ce_loss": 3.4081306331085437, "z_loss": 14.64133634711757, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20782940089702606, "tokens_per_second": 3689042.085570767, "eta_seconds": 6413.096992451257, "mean_data_wait_seconds": 9.626150131225586e-05}
{"step": 2490, "total_steps": 46238, "tokens": 1346813952, "loss": 3.3613780158938784, "ce_loss": 3.3600009444988137, "z_loss": 13.770974043643836, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2029178887605667, "tokens_per_second": 3691628.433342326, "eta_seconds": 6407.139399613208, "mean_data_wait_seconds": 6.705522537231445e-05}
{"step": 2500, "total_steps": 46238, "tokens": 1352220672, "loss": 3.36184197844881, "ce_loss": 3.3604224360350408, "z_loss": 14.195188211672233, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15836265683174133, "tokens_per_second": 3695442.5307055702, "eta_seconds": 6399.063459250985, "mean_data_wait_seconds": 6.48796558380127e-05}
{"step": 2510, "total_steps": 46238, "tokens": 1357627392, "loss": 3.430127608053612, "ce_loss": 3.4287773461052864, "z_loss": 13.502633210384484, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20393557846546173, "tokens_per_second": 3699324.0689499737, "eta_seconds": 6390.887660380238, "mean_data_wait_seconds": 6.091594696044922e-05}
{"step": 2520, "total_steps": 46238, "tokens": 1363034112, "loss": 3.401470549178846, "ce_loss": 3.400041990207903, "z_loss": 14.28560921640107, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1874476969242096, "tokens_per_second": 3703384.941125136, "eta_seconds": 6382.419921170525, "mean_data_wait_seconds": 6.532669067382812e-05}
{"step": 2530, "total_steps": 46238, "tokens": 1368440832, "loss": 3.360725746010289, "ce_loss": 3.359282007723144, "z_loss": 14.43788837664055, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18908129632472992, "tokens_per_second": 3707370.8422457487, "eta_seconds": 6374.099627348143, "mean_data_wait_seconds": 9.971857070922852e-05}
{"step": 2540, "total_steps": 46238, "tokens": 1373847552, "loss": 3.3838764609712544, "ce_loss": 3.382417675220605, "z_loss": 14.588022116458777, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20320352911949158, "tokens_per_second": 3710349.1568996, "eta_seconds": 6367.525907923415, "mean_data_wait_seconds": 0.0001678764820098877}
{"step": 2550, "total_steps": 46238, "tokens": 1379254272, "loss": 3.3252029418945312, "ce_loss": 3.3238314950104915, "z_loss": 13.714383016933095, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17185825109481812, "tokens_per_second": 3712512.2566349423, "eta_seconds": 6362.359510540635, "mean_data_wait_seconds": 0.0001450181007385254}
{"step": 2560, "total_steps": 46238, "tokens": 1384660992, "loss": 3.3207117969339546, "ce_loss": 3.319344117785945, "z_loss": 13.676965417283954, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22027458250522614, "tokens_per_second": 3714054.195048858, "eta_seconds": 6358.262347243252, "mean_data_wait_seconds": 0.0002015233039855957}
{"step": 2570, "total_steps": 46238, "tokens": 1390067712, "loss": 3.3319541559074866, "ce_loss": 3.330480604460745, "z_loss": 14.735544089115027, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16458293795585632, "tokens_per_second": 3715964.201744364, "eta_seconds": 6353.539200651372, "mean_data_wait_seconds": 0.0001977086067199707}
{"step": 2580, "total_steps": 46238, "tokens": 1395474432, "loss": 3.3497672080993652, "ce_loss": 3.3483679077842017, "z_loss": 13.993243824351918, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18194204568862915, "tokens_per_second": 3718402.469764133, "eta_seconds": 6347.918950660891, "mean_data_wait_seconds": 0.0001087486743927002}
{"step": 2590, "total_steps": 46238, "tokens": 1400881152, "loss": 3.3602321418848904, "ce_loss": 3.358827632485014, "z_loss": 14.045648791573264, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19843575358390808, "tokens_per_second": 3720233.724696823, "eta_seconds": 6343.340910905579, "mean_data_wait_seconds": 8.246302604675293e-05}
{"step": 2600, "total_steps": 46238, "tokens": 1406287872, "loss": 3.304272929827372, "ce_loss": 3.3028916608203542, "z_loss": 13.812745954051163, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17973551154136658, "tokens_per_second": 3722579.788159651, "eta_seconds": 6337.890765711144, "mean_data_wait_seconds": 7.81714916229248e-05}
{"step": 2610, "total_steps": 46238, "tokens": 1411694592, "loss": 3.2856323502280493, "ce_loss": 3.2841790589419277, "z_loss": 14.53242895819924, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18749776482582092, "tokens_per_second": 3724529.3642719886, "eta_seconds": 6333.121593903874, "mean_data_wait_seconds": 7.545948028564453e-05}
{"step": 2620, "total_steps": 46238, "tokens": 1417101312, "loss": 3.36999101711042, "ce_loss": 3.368538509715687, "z_loss": 14.525463031999992, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20338085293769836, "tokens_per_second": 3726303.267433229, "eta_seconds": 6328.655756525209, "mean_data_wait_seconds": 7.507205009460449e-05}
{"step": 2630, "total_steps": 46238, "tokens": 1422508032, "loss": 3.2983135599078555, "ce_loss": 3.2968606768232402, "z_loss": 14.529444094860192, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18742024898529053, "tokens_per_second": 3727877.469071822, "eta_seconds": 6324.532954638741, "mean_data_wait_seconds": 8.350610733032227e-05}
{"step": 2640, "total_steps": 46238, "tokens": 1427914752, "loss": 3.3299565730672893, "ce_loss": 3.328492533076893, "z_loss": 14.640182892481485, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19719606637954712, "tokens_per_second": 3729313.9538748427, "eta_seconds": 6320.647034693468, "mean_data_wait_seconds": 7.197260856628418e-05}
{"step": 2650, "total_steps": 46238, "tokens": 1433321472, "loss": 3.3986222057631523, "ce_loss": 3.3972116401701262, "z_loss": 14.10603920618693, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.177068293094635, "tokens_per_second": 3730823.1305106212, "eta_seconds": 6316.641030574556, "mean_data_wait_seconds": 0.00019475817680358887}
{"step": 2660, "total_steps": 46238, "tokens": 1438728192, "loss": 3.3626392790765474, "ce_loss": 3.361173526807265, "z_loss": 14.657582846554844, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17055097222328186, "tokens_per_second": 3731936.5287820147, "eta_seconds": 6313.307732403883, "mean_data_wait_seconds": 8.055567741394043e-05}
{"step": 2670, "total_steps": 46238, "tokens": 1444134912, "loss": 3.329199104598074, "ce_loss": 3.327791295268319, "z_loss": 14.077920913696289, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18893647193908691, "tokens_per_second": 3734437.9305539723, "eta_seconds": 6307.631151471768, "mean_data_wait_seconds": 8.472800254821777e-05}
{"step": 2680, "total_steps": 46238, "tokens": 1449541632, "loss": 3.298158288002014, "ce_loss": 3.296749938618053, "z_loss": 14.083305922421543, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1889648139476776, "tokens_per_second": 3737024.4123199196, "eta_seconds": 6301.818694671113, "mean_data_wait_seconds": 7.930397987365723e-05}
{"step": 2690, "total_steps": 46238, "tokens": 1454948352, "loss": 3.2776836030411, "ce_loss": 3.2762345816149856, "z_loss": 14.490191279035626, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1636555790901184, "tokens_per_second": 3739440.252225339, "eta_seconds": 6296.301584171212, "mean_data_wait_seconds": 7.125735282897949e-05}
{"step": 2700, "total_steps": 46238, "tokens": 1460355072, "loss": 3.3181618271452007, "ce_loss": 3.3167733206893457, "z_loss": 13.885091882763486, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22207584977149963, "tokens_per_second": 3741770.48632326, "eta_seconds": 6290.935521042643, "mean_data_wait_seconds": 8.589029312133789e-05}
{"step": 2710, "total_steps": 46238, "tokens": 1465761792, "loss": 3.2776904557690476, "ce_loss": 3.2762931856242092, "z_loss": 13.972452864502415, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15612278878688812, "tokens_per_second": 3744041.580188537, "eta_seconds": 6285.675423191993, "mean_data_wait_seconds": 0.00013002753257751465}
{"step": 2720, "total_steps": 46238, "tokens": 1471168512, "loss": 3.2884046272798018, "ce_loss": 3.2870172388625867, "z_loss": 13.874428922479803, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20694714784622192, "tokens_per_second": 3745399.4948848933, "eta_seconds": 6281.952954853778, "mean_data_wait_seconds": 7.960200309753418e-05}
{"step": 2730, "total_steps": 46238, "tokens": 1476575232, "loss": 3.2723586830225857, "ce_loss": 3.2709608511491255, "z_loss": 13.978268305460611, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18162700533866882, "tokens_per_second": 3747722.5908569978, "eta_seconds": 6276.616300626711, "mean_data_wait_seconds": 0.0002563595771789551}
{"step": 2740, "total_steps": 46238, "tokens": 1481981952, "loss": 3.36436743627895, "ce_loss": 3.3629720789013486, "z_loss": 13.953878128167355, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20060107111930847, "tokens_per_second": 3749210.70672525, "eta_seconds": 6272.68292545272, "mean_data_wait_seconds": 9.936094284057617e-05}
{"step": 2750, "total_steps": 46238, "tokens": 1487388672, "loss": 3.342721052242048, "ce_loss": 3.3412631117936336, "z_loss": 14.579406297568118, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1822504699230194, "tokens_per_second": 3751314.60119895, "eta_seconds": 6267.723655191519, "mean_data_wait_seconds": 0.0003097057342529297}
{"step": 2760, "total_steps": 46238, "tokens": 1492795392, "loss": 3.3422802199016917, "ce_loss": 3.3408065980130974, "z_loss": 14.736390764063055, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20276176929473877, "tokens_per_second": 3753267.7824495602, "eta_seconds": 6263.021427332945, "mean_data_wait_seconds": 0.00013270974159240723}
{"step": 2770, "total_steps": 46238, "tokens": 1498202112, "loss": 3.332904403859919, "ce_loss": 3.3314406564741423, "z_loss": 14.637805584705237, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16384005546569824, "tokens_per_second": 3755487.302545056, "eta_seconds": 6257.880251139005, "mean_data_wait_seconds": 8.377432823181152e-05}
{"step": 2780, "total_steps": 46238, "tokens": 1503608832, "loss": 3.3622001680460842, "ce_loss": 3.360831432270281, "z_loss": 13.687294606006507, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2071257382631302, "tokens_per_second": 3757602.816502665, "eta_seconds": 6252.91821711709, "mean_data_wait_seconds": 8.210539817810059e-05}
{"step": 2790, "total_steps": 46238, "tokens": 1509015552, "loss": 3.3045302286292566, "ce_loss": 3.3030469345323965, "z_loss": 14.833467562993368, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20889534056186676, "tokens_per_second": 3758791.458152472, "eta_seconds": 6249.502438623219, "mean_data_wait_seconds": 6.434321403503418e-05}
{"step": 2800, "total_steps": 46238, "tokens": 1514422272, "loss": 3.3596174102841, "ce_loss": 3.358214674573956, "z_loss": 14.027479157303318, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1952541321516037, "tokens_per_second": 3761317.3633455024, "eta_seconds": 6243.868143875827, "mean_data_wait_seconds": 6.979703903198242e-05}
{"step": 2810, "total_steps": 46238, "tokens": 1519828992, "loss": 3.293609030318983, "ce_loss": 3.2921891031843242, "z_loss": 14.199465318159623, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16367405652999878, "tokens_per_second": 3764570.465166588, "eta_seconds": 6237.0363793843835, "mean_data_wait_seconds": 6.714463233947754e-05}
{"step": 2820, "total_steps": 46238, "tokens": 1525235712, "loss": 3.3388620073145088, "ce_loss": 3.3373917106426125, "z_loss": 14.702778209339488, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2126803994178772, "tokens_per_second": 3767221.524493434, "eta_seconds": 6231.212067401988, "mean_data_wait_seconds": 0.00010827183723449707}
{"step": 2830, "total_steps": 46238, "tokens": 1530642432, "loss": 3.351118364117362, "ce_loss": 3.349662724769477, "z_loss": 14.55601723988851, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1874718815088272, "tokens_per_second": 3769299.949128632, "eta_seconds": 6226.341713512462, "mean_data_wait_seconds": 9.086728096008301e-05}
{"step": 2840, "total_steps": 46238, "tokens": 1536049152, "loss": 3.307605989051588, "ce_loss": 3.306147580797022, "z_loss": 14.58431152863936, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17175507545471191, "tokens_per_second": 3771532.601205496, "eta_seconds": 6221.222315962572, "mean_data_wait_seconds": 7.003545761108398e-05}
{"step": 2850, "total_steps": 46238, "tokens": 1541455872, "loss": 3.3913076917330423, "ce_loss": 3.3898497530908296, "z_loss": 14.579356655930027, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2066810429096222, "tokens_per_second": 3773694.9403462834, "eta_seconds": 6216.224796869094, "mean_data_wait_seconds": 7.522106170654297e-05}
{"step": 2860, "total_steps": 46238, "tokens": 1546862592, "loss": 3.3175930868495596, "ce_loss": 3.3161594434217974, "z_loss": 14.336667624386875, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15412305295467377, "tokens_per_second": 3776312.4757617153, "eta_seconds": 6210.484300367484, "mean_data_wait_seconds": 5.418062210083008e-05}
{"step": 2870, "total_steps": 46238, "tokens": 1552269312, "loss": 3.3560151096546287, "ce_loss": 3.3546297387643293, "z_loss": 13.85340336597327, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1925474852323532, "tokens_per_second": 3778715.5956371897, "eta_seconds": 6205.10383238995, "mean_data_wait_seconds": 6.80387020111084e-05}
{"step": 2880, "total_steps": 46238, "tokens": 1557676032, "loss": 3.377242026907025, "ce_loss": 3.3758742466117395, "z_loss": 13.677945722233165, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19388781487941742, "tokens_per_second": 3781352.0587978265, "eta_seconds": 6199.347624736294, "mean_data_wait_seconds": 8.568167686462402e-05}
{"step": 2890, "total_steps": 46238, "tokens": 1563082752, "loss": 3.337154224063411, "ce_loss": 3.335769331816471, "z_loss": 13.849117705316255, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17879502475261688, "tokens_per_second": 3784098.7683451762, "eta_seconds": 6193.418993196368, "mean_data_wait_seconds": 6.35683536529541e-05}
{"step": 2900, "total_steps": 46238, "tokens": 1568489472, "loss": 3.3814020770968813, "ce_loss": 3.379968094103264, "z_loss": 14.339824849909002, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17097878456115723, "tokens_per_second": 3786455.181341504, "eta_seconds": 6188.136751086167, "mean_data_wait_seconds": 0.00018516182899475098}
{"step": 2910, "total_steps": 46238, "tokens": 1573896192, "loss": 3.441555521704934, "ce_loss": 3.440162884466576, "z_loss": 13.926173571384314, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1914615035057068, "tokens_per_second": 3788468.9145315974, "eta_seconds": 6183.420340112869, "mean_data_wait_seconds": 0.00016421079635620117}
{"step": 2920, "total_steps": 46238, "tokens": 1579302912, "loss": 3.365973320874301, "ce_loss": 3.364587097456961, "z_loss": 13.862294399377072, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17278815805912018, "tokens_per_second": 3790931.636351193, "eta_seconds": 6177.977149316849, "mean_data_wait_seconds": 7.125735282897949e-05}
{"step": 2930, "total_steps": 46238, "tokens": 1584709632, "loss": 3.309361669150266, "ce_loss": 3.3079493768287427, "z_loss": 14.122515035398079, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15570354461669922, "tokens_per_second": 3792270.5676748743, "eta_seconds": 6174.370179065621, "mean_data_wait_seconds": 8.20159912109375e-05}
{"step": 2940, "total_steps": 46238, "tokens": 1590116352, "loss": 3.3292499545848733, "ce_loss": 3.3277868447881755, "z_loss": 14.630557479280414, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.159083291888237, "tokens_per_second": 3794756.431016201, "eta_seconds": 6168.9006948282995, "mean_data_wait_seconds": 9.459257125854492e-05}
{"step": 2950, "total_steps": 46238, "tokens": 1595523072, "loss": 3.3426746581539963, "ce_loss": 3.3412424885865413, "z_loss": 14.321613362341216, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17053604125976562, "tokens_per_second": 3796703.9833682287, "eta_seconds": 6164.312247286971, "mean_data_wait_seconds": 8.410215377807617e-05}
{"step": 2960, "total_steps": 46238, "tokens": 1600929792, "loss": 3.328032267816139, "ce_loss": 3.326562205950419, "z_loss": 14.700782414638635, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19599127769470215, "tokens_per_second": 3798593.8626999655, "eta_seconds": 6159.822026187526, "mean_data_wait_seconds": 8.407235145568848e-05}
{"step": 2970, "total_steps": 46238, "tokens": 1606336512, "loss": 3.334928041154688, "ce_loss": 3.3335256576538086, "z_loss": 14.02400127324191, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15788693726062775, "tokens_per_second": 3800383.0055551436, "eta_seconds": 6155.4994298746515, "mean_data_wait_seconds": 5.7756900787353516e-05}
{"step": 2980, "total_steps": 46238, "tokens": 1611743232, "loss": 3.3389505780104436, "ce_loss": 3.3375566096016853, "z_loss": 13.939616058811996, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17070627212524414, "tokens_per_second": 3801695.2696343577, "eta_seconds": 6151.952496247658, "mean_data_wait_seconds": 6.690621376037598e-05}
{"step": 2990, "total_steps": 46238, "tokens": 1617149952, "loss": 3.2874010331702954, "ce_loss": 3.2859991680492056, "z_loss": 14.018794565489799, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1834038347005844, "tokens_per_second": 3804544.322766491, "eta_seconds": 6145.92445252349, "mean_data_wait_seconds": 7.215142250061035e-05}
{"step": 3000, "total_steps": 46238, "tokens": 1622556672, "loss": 3.4263763174866186, "ce_loss": 3.4249564120263765, "z_loss": 14.198730114734534, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19371333718299866, "tokens_per_second": 3806217.8932717466, "eta_seconds": 6141.801630779887, "mean_data_wait_seconds": 7.581710815429688e-05}
{"step": 3010, "total_steps": 46238, "tokens": 1627963392, "loss": 3.361503088113033, "ce_loss": 3.3600467443466187, "z_loss": 14.563430554939039, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18167105317115784, "tokens_per_second": 3807722.588684097, "eta_seconds": 6137.954643401938, "mean_data_wait_seconds": 0.0001049339771270752}
{"step": 3020, "total_steps": 46238, "tokens": 1633370112, "loss": 3.3761402079553315, "ce_loss": 3.3747505599802192, "z_loss": 13.896566022526134, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1834682822227478, "tokens_per_second": 3809420.1457682527, "eta_seconds": 6133.80014014907, "mean_data_wait_seconds": 8.51750373840332e-05}
{"step": 3030, "total_steps": 46238, "tokens": 1638776832, "loss": 3.3373116345116585, "ce_loss": 3.3359009208101216, "z_loss": 14.107355977549698, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16508682072162628, "tokens_per_second": 3811309.3827655283, "eta_seconds": 6129.34106310969, "mean_data_wait_seconds": 0.00023511052131652832}
{"step": 3040, "total_steps": 46238, "tokens": 1644183552, "loss": 3.354283981250994, "ce_loss": 3.35286581696886, "z_loss": 14.181453495314628, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18008124828338623, "tokens_per_second": 3812548.2837503646, "eta_seconds": 6125.931174050739, "mean_data_wait_seconds": 0.00012832880020141602}
{"step": 3050, "total_steps": 46238, "tokens": 1649590272, "loss": 3.4095964124708464, "ce_loss": 3.4081790627855244, "z_loss": 14.173743284109866, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2151072919368744, "tokens_per_second": 3813458.4967814856, "eta_seconds": 6123.051210261532, "mean_data_wait_seconds": 0.00010272860527038574}
{"step": 3060, "total_steps": 46238, "tokens": 1654996992, "loss": 3.3319962602673154, "ce_loss": 3.330464372129151, "z_loss": 15.3196562203494, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19577819108963013, "tokens_per_second": 3816407.9207626213, "eta_seconds": 6116.902445620939, "mean_data_wait_seconds": 6.473064422607422e-05}
{"step": 3070, "total_steps": 46238, "tokens": 1660403712, "loss": 3.387013301704869, "ce_loss": 3.3855848041447727, "z_loss": 14.284982146638812, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1743386834859848, "tokens_per_second": 3819550.965815576, "eta_seconds": 6110.453410068967, "mean_data_wait_seconds": 6.458163261413574e-05}
{"step": 3080, "total_steps": 46238, "tokens": 1665810432, "loss": 3.406706634796027, "ce_loss": 3.405303799744808, "z_loss": 14.028403665080216, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16399340331554413, "tokens_per_second": 3822655.9777341657, "eta_seconds": 6104.075710687108, "mean_data_wait_seconds": 7.978081703186035e-05}
{"step": 3090, "total_steps": 46238, "tokens": 1671217152, "loss": 3.3334021785042505, "ce_loss": 3.3319166320742983, "z_loss": 14.855469985441728, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22454576194286346, "tokens_per_second": 3825016.4918002947, "eta_seconds": 6098.895216271392, "mean_data_wait_seconds": 6.058812141418457e-05}
{"step": 3100, "total_steps": 46238, "tokens": 1676623872, "loss": 3.3428955764481514, "ce_loss": 3.341428805481304, "z_loss": 14.667695659579653, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1674727201461792, "tokens_per_second": 3827491.442135319, "eta_seconds": 6093.538918793336, "mean_data_wait_seconds": 6.476044654846191e-05}
{"step": 3110, "total_steps": 46238, "tokens": 1682030592, "loss": 3.330956805836071, "ce_loss": 3.3295594941486013, "z_loss": 13.972681240601974, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1680527627468109, "tokens_per_second": 3829887.245414122, "eta_seconds": 6088.315360176796, "mean_data_wait_seconds": 7.581710815429688e-05}
{"step": 3120, "total_steps": 46238, "tokens": 1687437312, "loss": 3.3177816055037757, "ce_loss": 3.3163154504515906, "z_loss": 14.661804531559799, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1829400211572647, "tokens_per_second": 3832422.801127738, "eta_seconds": 6082.87650755551, "mean_data_wait_seconds": 8.872151374816895e-05}
{"step": 3130, "total_steps": 46238, "tokens": 1692844032, "loss": 3.3351108612436238, "ce_loss": 3.3337040543556213, "z_loss": 14.067982045086948, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17097951471805573, "tokens_per_second": 3834240.188966186, "eta_seconds": 6078.583175636716, "mean_data_wait_seconds": 5.5789947509765625e-05}
{"step": 3140, "total_steps": 46238, "tokens": 1698250752, "loss": 3.344611364783663, "ce_loss": 3.3431796178673254, "z_loss": 14.317472869699651, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1673688441514969, "tokens_per_second": 3836331.586503828, "eta_seconds": 6073.860055781899, "mean_data_wait_seconds": 9.748339653015137e-05}
{"step": 3150, "total_steps": 46238, "tokens": 1703657472, "loss": 3.2642096244927608, "ce_loss": 3.2627258011789033, "z_loss": 14.83818254326329, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17583730816841125, "tokens_per_second": 3838222.483930203, "eta_seconds": 6069.459121125723, "mean_data_wait_seconds": 8.58008861541748e-05}
{"step": 3160, "total_steps": 46238, "tokens": 1709064192, "loss": 3.322534385955695, "ce_loss": 3.3211318091912703, "z_loss": 14.025747660434607, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17298704385757446, "tokens_per_second": 3839821.212112178, "eta_seconds": 6065.524006829613, "mean_data_wait_seconds": 0.00011661648750305176}
{"step": 3170, "total_steps": 46238, "tokens": 1714470912, "loss": 3.3526771863301597, "ce_loss": 3.351299473733613, "z_loss": 13.777387366150364, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19099560379981995, "tokens_per_second": 3840532.044565646, "eta_seconds": 6062.993552403359, "mean_data_wait_seconds": 7.081031799316406e-05}
{"step": 3180, "total_steps": 46238, "tokens": 1719877632, "loss": 3.3077824946605796, "ce_loss": 3.306306737841982, "z_loss": 14.75705193750786, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1589420884847641, "tokens_per_second": 3842598.806159575, "eta_seconds": 6058.325492290085, "mean_data_wait_seconds": 0.00015872716903686523}
{"step": 3190, "total_steps": 46238, "tokens": 1725284352, "loss": 3.2495089624867295, "ce_loss": 3.2480238369016936, "z_loss": 14.8509894212087, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20879800617694855, "tokens_per_second": 3843693.0138636567, "eta_seconds": 6055.19418435678, "mean_data_wait_seconds": 0.0001544952392578125}
{"step": 3200, "total_steps": 46238, "tokens": 1730691072, "loss": 3.2705152251503686, "ce_loss": 3.2690588253917117, "z_loss": 14.563991488832416, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18927475810050964, "tokens_per_second": 3845339.255455286, "eta_seconds": 6051.195829077759, "mean_data_wait_seconds": 6.857514381408691e-05}
{"step": 3210, "total_steps": 46238, "tokens": 1736097792, "loss": 3.305633808627273, "ce_loss": 3.3041726314660274, "z_loss": 14.611586209499475, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1659340113401413, "tokens_per_second": 3847777.0897293473, "eta_seconds": 6045.9568216921725, "mean_data_wait_seconds": 7.873773574829102e-05}
{"step": 3220, "total_steps": 46238, "tokens": 1741504512, "loss": 3.317164538484631, "ce_loss": 3.315697998711557, "z_loss": 14.665281519745335, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.157099649310112, "tokens_per_second": 3849304.967147208, "eta_seconds": 6042.152446351115, "mean_data_wait_seconds": 0.00014713406562805176}
{"step": 3230, "total_steps": 46238, "tokens": 1746911232, "loss": 3.303584559397264, "ce_loss": 3.3021151026089988, "z_loss": 14.69444391221711, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1674063801765442, "tokens_per_second": 3851702.1698010266, "eta_seconds": 6036.9882402411195, "mean_data_wait_seconds": 0.00012448430061340332}
{"step": 3240, "total_steps": 46238, "tokens": 1752317952, "loss": 3.3419874039563267, "ce_loss": 3.3405056686112373, "z_loss": 14.81780775388082, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2015708088874817, "tokens_per_second": 3853616.099572289, "eta_seconds": 6032.586895871699, "mean_data_wait_seconds": 9.569525718688965e-05}
{"step": 3250, "total_steps": 46238, "tokens": 1757724672, "loss": 3.2960496147473655, "ce_loss": 3.2946475274635083, "z_loss": 14.020759459697839, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1292671412229538, "tokens_per_second": 3854749.888207838, "eta_seconds": 6029.409932691036, "mean_data_wait_seconds": 6.493926048278809e-05}
{"step": 3260, "total_steps": 46238, "tokens": 1763131392, "loss": 3.289334172552282, "ce_loss": 3.287876970840223, "z_loss": 14.572051467317523, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19795379042625427, "tokens_per_second": 3856686.072432706, "eta_seconds": 6024.98106083677, "mean_data_wait_seconds": 7.534027099609375e-05}
{"step": 3270, "total_steps": 46238, "tokens": 1768538112, "loss": 3.3092683770439844, "ce_loss": 3.3078594532879917, "z_loss": 14.08896287282308, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1595718115568161, "tokens_per_second": 3853417.355948943, "eta_seconds": 6028.688739914371, "mean_data_wait_seconds": 6.416440010070801e-05}
{"step": 3280, "total_steps": 46238, "tokens": 1773944832, "loss": 3.362907035784288, "ce_loss": 3.3614859021071233, "z_loss": 14.211799845550999, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17653369903564453, "tokens_per_second": 3849979.1944815875, "eta_seconds": 6032.668212153134, "mean_data_wait_seconds": 0.00017058849334716797}
{"step": 3290, "total_steps": 46238, "tokens": 1779351552, "loss": 3.322770736434243, "ce_loss": 3.3213743188164453, "z_loss": 13.964421163905751, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16564130783081055, "tokens_per_second": 3850829.4802367105, "eta_seconds": 6029.932123240277, "mean_data_wait_seconds": 8.815526962280273e-05}
{"step": 3300, "total_steps": 46238, "tokens": 1784758272, "loss": 3.304273480718786, "ce_loss": 3.3028834624723955, "z_loss": 13.900759906479806, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18461313843727112, "tokens_per_second": 3850357.5795423156, "eta_seconds": 6029.266940645939, "mean_data_wait_seconds": 7.635354995727539e-05}
{"step": 3310, "total_steps": 46238, "tokens": 1790164992, "loss": 3.3185077779220813, "ce_loss": 3.317106665986957, "z_loss": 14.011336384397564, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14391304552555084, "tokens_per_second": 3851583.892496327, "eta_seconds": 6025.943505791659, "mean_data_wait_seconds": 0.0001303553581237793}
{"step": 3320, "total_steps": 46238, "tokens": 1795571712, "loss": 3.2777337594465776, "ce_loss": 3.2762715924869883, "z_loss": 14.621811397147901, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.146871417760849, "tokens_per_second": 3851991.279822898, "eta_seconds": 6023.902583981666, "mean_data_wait_seconds": 8.338689804077148e-05}
{"step": 3330, "total_steps": 46238, "tokens": 1800978432, "loss": 3.3182826150547373, "ce_loss": 3.316815481041417, "z_loss": 14.671030528617628, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18187421560287476, "tokens_per_second": 3853714.565769467, "eta_seconds": 6019.805854346651, "mean_data_wait_seconds": 0.00010639429092407227}
{"step": 3340, "total_steps": 46238, "tokens": 1806385152, "loss": 3.225194611332633, "ce_loss": 3.2238284331379514, "z_loss": 13.661980296626236, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18417032063007355, "tokens_per_second": 3855828.4156981288, "eta_seconds": 6015.103444327069, "mean_data_wait_seconds": 0.00010433793067932129}
{"step": 3350, "total_steps": 46238, "tokens": 1811791872, "loss": 3.33672797499281, "ce_loss": 3.3352761575669954, "z_loss": 14.517909143910263, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15294893085956573, "tokens_per_second": 3855823.1279454576, "eta_seconds": 6013.70947125249, "mean_data_wait_seconds": 0.00011566281318664551}
{"step": 3360, "total_steps": 46238, "tokens": 1817198592, "loss": 3.327878132010951, "ce_loss": 3.3264017683086973, "z_loss": 14.763622428431656, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20793099701404572, "tokens_per_second": 3856404.1136245616, "eta_seconds": 6011.401466484617, "mean_data_wait_seconds": 0.00018486380577087402}
{"step": 3370, "total_steps": 46238, "tokens": 1822605312, "loss": 3.3203380884546223, "ce_loss": 3.318955826036858, "z_loss": 13.822221582586115, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22107534110546112, "tokens_per_second": 3857405.160252787, "eta_seconds": 6008.439782996802, "mean_data_wait_seconds": 7.292628288269043e-05}
{"step": 3380, "total_steps": 46238, "tokens": 1828012032, "loss": 3.287260104309429, "ce_loss": 3.2858672467145054, "z_loss": 13.928434653715653, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15855643153190613, "tokens_per_second": 3858164.255371333, "eta_seconds": 6005.85624931353, "mean_data_wait_seconds": 0.00012227892875671387}
{"step": 3390, "total_steps": 46238, "tokens": 1833418752, "loss": 3.2733381083517363, "ce_loss": 3.271906164559451, "z_loss": 14.319516109697746, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16089211404323578, "tokens_per_second": 3860079.0485142902, "eta_seconds": 6001.476366893691, "mean_data_wait_seconds": 0.00011590123176574707}
{"step": 3400, "total_steps": 46238, "tokens": 1838825472, "loss": 3.359282116095225, "ce_loss": 3.35783049735156, "z_loss": 14.516064766681556, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16806359589099884, "tokens_per_second": 3858526.7949040467, "eta_seconds": 6002.489472041092, "mean_data_wait_seconds": 6.765127182006836e-05}
{"step": 3410, "total_steps": 46238, "tokens": 1844232192, "loss": 3.321104215853142, "ce_loss": 3.3197199395208647, "z_loss": 13.842771147236679, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19758932292461395, "tokens_per_second": 3860269.110723331, "eta_seconds": 5998.3796672821045, "mean_data_wait_seconds": 6.458163261413574e-05}
{"step": 3420, "total_steps": 46238, "tokens": 1849638912, "loss": 3.2882751732161553, "ce_loss": 3.28690403880495, "z_loss": 13.711500529086951, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2146524041891098, "tokens_per_second": 3862059.3978954433, "eta_seconds": 5994.199114756011, "mean_data_wait_seconds": 0.0003420114517211914}
{"step": 3430, "total_steps": 46238, "tokens": 1855045632, "loss": 3.3523123715863083, "ce_loss": 3.350950250119874, "z_loss": 13.621059880112156, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17668448388576508, "tokens_per_second": 3863393.0252907523, "eta_seconds": 5990.7304673611825, "mean_data_wait_seconds": 0.00016829371452331543}
{"step": 3440, "total_steps": 46238, "tokens": 1860452352, "loss": 3.3107561974814446, "ce_loss": 3.309376136823134, "z_loss": 13.800631523132324, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16527201235294342, "tokens_per_second": 3864773.267662093, "eta_seconds": 5987.191998457777, "mean_data_wait_seconds": 6.979703903198242e-05}
{"step": 3450, "total_steps": 46238, "tokens": 1865859072, "loss": 3.3006625897956616, "ce_loss": 3.299248865156463, "z_loss": 14.137389327540543, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19593483209609985, "tokens_per_second": 3867066.311272599, "eta_seconds": 5982.243644636909, "mean_data_wait_seconds": 9.959936141967773e-05}
{"step": 3460, "total_steps": 46238, "tokens": 1871265792, "loss": 3.2830134702451303, "ce_loss": 3.281619127952691, "z_loss": 13.943330627499204, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17496290802955627, "tokens_per_second": 3865220.9333177796, "eta_seconds": 5983.700943104279, "mean_data_wait_seconds": 7.635354995727539e-05}
{"step": 3470, "total_steps": 46238, "tokens": 1876672512, "loss": 3.261788353775487, "ce_loss": 3.2603459069223115, "z_loss": 14.424779321208145, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15804798901081085, "tokens_per_second": 3862674.0770181525, "eta_seconds": 5986.246564672646, "mean_data_wait_seconds": 0.00010773539543151855}
{"step": 3480, "total_steps": 46238, "tokens": 1882079232, "loss": 3.273988660537835, "ce_loss": 3.272590711261287, "z_loss": 13.979398228905417, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1580304503440857, "tokens_per_second": 3863036.103187263, "eta_seconds": 5984.285957080884, "mean_data_wait_seconds": 0.00014898180961608887}
{"step": 3490, "total_steps": 46238, "tokens": 1887485952, "loss": 3.3475351134936013, "ce_loss": 3.3461691574616865, "z_loss": 13.659314971981626, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2086731344461441, "tokens_per_second": 3857576.3969625486, "eta_seconds": 5991.354053855795, "mean_data_wait_seconds": 9.435415267944336e-05}
{"step": 3500, "total_steps": 46238, "tokens": 1892892672, "loss": 3.3464717575998018, "ce_loss": 3.3451026714209355, "z_loss": 13.690685416712906, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18438000977039337, "tokens_per_second": 3858003.8266820083, "eta_seconds": 5989.2888400456595, "mean_data_wait_seconds": 0.00014907121658325195}
{"step": 3510, "total_steps": 46238, "tokens": 1898299392, "loss": 3.367899941675591, "ce_loss": 3.3665086157394177, "z_loss": 13.913276238874955, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16237086057662964, "tokens_per_second": 3858851.668727192, "eta_seconds": 5986.57179057099, "mean_data_wait_seconds": 0.00035309791564941406}
{"step": 3520, "total_steps": 46238, "tokens": 1903706112, "loss": 3.359457518115188, "ce_loss": 3.358069654667016, "z_loss": 13.878198963223081, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16944566369056702, "tokens_per_second": 3860425.2489513867, "eta_seconds": 5982.731003604738, "mean_data_wait_seconds": 0.00012877583503723145}
{"step": 3530, "total_steps": 46238, "tokens": 1909112832, "loss": 3.381641852133202, "ce_loss": 3.380168875058492, "z_loss": 14.72960461992206, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1789851188659668, "tokens_per_second": 3861506.2665801044, "eta_seconds": 5979.655996894134, "mean_data_wait_seconds": 0.00010204315185546875}
{"step": 3540, "total_steps": 46238, "tokens": 1914519552, "loss": 3.34854506001328, "ce_loss": 3.347124078057029, "z_loss": 14.210223609750921, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1542128622531891, "tokens_per_second": 3862001.4106759373, "eta_seconds": 5977.489371232413, "mean_data_wait_seconds": 0.00017243623733520508}
{"step": 3550, "total_steps": 46238, "tokens": 1919926272, "loss": 3.333540813489394, "ce_loss": 3.332087814807892, "z_loss": 14.529921025940865, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2285827249288559, "tokens_per_second": 3862274.285009353, "eta_seconds": 5975.667174539913, "mean_data_wait_seconds": 0.00011116266250610352}
{"step": 3560, "total_steps": 46238, "tokens": 1925332992, "loss": 3.3322528835498924, "ce_loss": 3.3308584130171575, "z_loss": 13.944675293835727, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18822188675403595, "tokens_per_second": 3863835.1224287613, "eta_seconds": 5971.853925665387, "mean_data_wait_seconds": 0.00011834502220153809}
{"step": 3570, "total_steps": 46238, "tokens": 1930739712, "loss": 3.272273488116987, "ce_loss": 3.270852986610297, "z_loss": 14.204555952187741, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15493595600128174, "tokens_per_second": 3865589.508609198, "eta_seconds": 5967.744938416897, "mean_data_wait_seconds": 8.085370063781738e-05}
{"step": 3580, "total_steps": 46238, "tokens": 1936146432, "loss": 3.3057281338807307, "ce_loss": 3.3043202793959416, "z_loss": 14.078582156788219, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15870821475982666, "tokens_per_second": 3866549.111371939, "eta_seconds": 5964.865527290966, "mean_data_wait_seconds": 0.0007784366607666016}
{"step": 3590, "total_steps": 46238, "tokens": 1941553152, "loss": 3.3417231314110034, "ce_loss": 3.3402684276754204, "z_loss": 14.547117428346114, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1829119324684143, "tokens_per_second": 3868423.7448696923, "eta_seconds": 5960.57730608742, "mean_data_wait_seconds": 9.822845458984375e-05}
{"step": 3600, "total_steps": 46238, "tokens": 1946959872, "loss": 3.3426878398114983, "ce_loss": 3.341240082726334, "z_loss": 14.477744391470244, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1819261908531189, "tokens_per_second": 3869776.608763132, "eta_seconds": 5957.096337756857, "mean_data_wait_seconds": 8.043646812438965e-05}
{"step": 3610, "total_steps": 46238, "tokens": 1952366592, "loss": 3.3317064317789944, "ce_loss": 3.3302812377611795, "z_loss": 14.251801158442643, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16467271745204926, "tokens_per_second": 3871893.859661748, "eta_seconds": 5952.442442730965, "mean_data_wait_seconds": 8.732080459594727e-05}
{"step": 3620, "total_steps": 46238, "tokens": 1957773312, "loss": 3.291280889149868, "ce_loss": 3.2898299423131077, "z_loss": 14.509148373748317, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18865570425987244, "tokens_per_second": 3873634.0141277425, "eta_seconds": 5948.372649548956, "mean_data_wait_seconds": 8.189678192138672e-05}
{"step": 3630, "total_steps": 46238, "tokens": 1963180032, "loss": 3.3357323426188845, "ce_loss": 3.334293737555995, "z_loss": 14.386066032178475, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16006410121917725, "tokens_per_second": 3875643.487703563, "eta_seconds": 5943.893440428334, "mean_data_wait_seconds": 9.85264778137207e-05}
{"step": 3640, "total_steps": 46238, "tokens": 1968586752, "loss": 3.4204534129662947, "ce_loss": 3.4190807992761787, "z_loss": 13.726186391079064, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2101934850215912, "tokens_per_second": 3877245.578545896, "eta_seconds": 5940.042929299681, "mean_data_wait_seconds": 7.411837577819824e-05}
{"step": 3650, "total_steps": 46238, "tokens": 1973993472, "loss": 3.328257732319109, "ce_loss": 3.3268391146804346, "z_loss": 14.186316576871006, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14457862079143524, "tokens_per_second": 3879197.4253854286, "eta_seconds": 5935.660379984973, "mean_data_wait_seconds": 8.356571197509766e-05}
{"step": 3660, "total_steps": 46238, "tokens": 1979400192, "loss": 3.383787830670675, "ce_loss": 3.3824280681032124, "z_loss": 13.598103739998557, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17414702475070953, "tokens_per_second": 3879833.582837638, "eta_seconds": 5933.293594299852, "mean_data_wait_seconds": 7.370114326477051e-05}
{"step": 3670, "total_steps": 46238, "tokens": 1984806912, "loss": 3.370279164025278, "ce_loss": 3.3689012455217764, "z_loss": 13.779035091400146, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17755138874053955, "tokens_per_second": 3881842.1245295787, "eta_seconds": 5928.830767889369, "mean_data_wait_seconds": 6.911158561706543e-05}
{"step": 3680, "total_steps": 46238, "tokens": 1990213632, "loss": 3.327399243008007, "ce_loss": 3.3260370203942964, "z_loss": 13.622166366288155, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17056314647197723, "tokens_per_second": 3883948.829929045, "eta_seconds": 5924.2228236102565, "mean_data_wait_seconds": 6.157159805297852e-05}
{"step": 3690, "total_steps": 46238, "tokens": 1995620352, "loss": 3.33141273801977, "ce_loss": 3.3299932118618125, "z_loss": 14.195205529530844, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18144819140434265, "tokens_per_second": 3884800.5600598776, "eta_seconds": 5921.532194086543, "mean_data_wait_seconds": 0.00023186206817626953}
{"step": 3700, "total_steps": 46238, "tokens": 2001027072, "loss": 3.330202904614535, "ce_loss": 3.32885853991364, "z_loss": 13.443418358311508, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17961396276950836, "tokens_per_second": 3887581.3140380066, "eta_seconds": 5915.905805224568, "mean_data_wait_seconds": 7.486343383789062e-05}
{"step": 3710, "total_steps": 46238, "tokens": 2006433792, "loss": 3.2721871047308952, "ce_loss": 3.270804475654255, "z_loss": 13.82624406525583, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17304617166519165, "tokens_per_second": 3889273.1442046086, "eta_seconds": 5911.942229684232, "mean_data_wait_seconds": 7.417798042297363e-05}
{"step": 3720, "total_steps": 46238, "tokens": 2011840512, "loss": 3.337778499632171, "ce_loss": 3.336366212729252, "z_loss": 14.122615279573383, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14533227682113647, "tokens_per_second": 3891719.689078456, "eta_seconds": 5906.836375834512, "mean_data_wait_seconds": 0.00013017654418945312}
{"step": 3730, "total_steps": 46238, "tokens": 2017247232, "loss": 3.3094956983219492, "ce_loss": 3.30809240991419, "z_loss": 14.032422448649552, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17542396485805511, "tokens_per_second": 3893472.4418315277, "eta_seconds": 5902.788589711677, "mean_data_wait_seconds": 0.00012496113777160645}
{"step": 3740, "total_steps": 46238, "tokens": 2022653952, "loss": 3.228658437728882, "ce_loss": 3.227268986629717, "z_loss": 13.894361214204269, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1641102284193039, "tokens_per_second": 3894225.358807819, "eta_seconds": 5900.258939054872, "mean_data_wait_seconds": 8.752942085266113e-05}
{"step": 3750, "total_steps": 46238, "tokens": 2028060672, "loss": 3.3123759139667857, "ce_loss": 3.3110027024240205, "z_loss": 13.732114762970895, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17723774909973145, "tokens_per_second": 3894618.9227716625, "eta_seconds": 5898.274444692518, "mean_data_wait_seconds": 0.000479280948638916}
{"step": 3760, "total_steps": 46238, "tokens": 2033467392, "loss": 3.2471441734920847, "ce_loss": 3.245693295290976, "z_loss": 14.508968555566037, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18240560591220856, "tokens_per_second": 3895255.5495487824, "eta_seconds": 5895.922424566045, "mean_data_wait_seconds": 7.864832878112793e-05}
{"step": 3770, "total_steps": 46238, "tokens": 2038874112, "loss": 3.306208281806021, "ce_loss": 3.304774976137913, "z_loss": 14.333106604489414, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22480815649032593, "tokens_per_second": 3896288.3381934846, "eta_seconds": 5892.971934065266, "mean_data_wait_seconds": 0.00014868378639221191}
{"step": 3780, "total_steps": 46238, "tokens": 2044280832, "loss": 3.290230935270136, "ce_loss": 3.2888233571341545, "z_loss": 14.07591842882561, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17949874699115753, "tokens_per_second": 3897106.067008534, "eta_seconds": 5890.348045266516, "mean_data_wait_seconds": 7.56680965423584e-05}
{"step": 3790, "total_steps": 46238, "tokens": 2049687552, "loss": 3.3104739008527813, "ce_loss": 3.3090882752880906, "z_loss": 13.856097011855155, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17616160213947296, "tokens_per_second": 3897404.0822896003, "eta_seconds": 5888.510377532541, "mean_data_wait_seconds": 0.0001157224178314209}
{"step": 3800, "total_steps": 46238, "tokens": 2055094272, "loss": 3.2498896483219033, "ce_loss": 3.2484537615920557, "z_loss": 14.358955087083759, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2043638527393341, "tokens_per_second": 3898754.0495677837, "eta_seconds": 5885.084663533374, "mean_data_wait_seconds": 0.00011041760444641113}
{"step": 3810, "total_steps": 46238, "tokens": 2060500992, "loss": 3.2647930022441978, "ce_loss": 3.263362821304437, "z_loss": 14.301684473500107, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22475308179855347, "tokens_per_second": 3900388.4125424633, "eta_seconds": 5881.232461422267, "mean_data_wait_seconds": 9.578466415405273e-05}
{"step": 3820, "total_steps": 46238, "tokens": 2065907712, "loss": 3.3089595996972285, "ce_loss": 3.307580303062092, "z_loss": 13.79258561856819, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16293105483055115, "tokens_per_second": 3901886.561713356, "eta_seconds": 5877.588664169057, "mean_data_wait_seconds": 0.00016584992408752441}
{"step": 3830, "total_steps": 46238, "tokens": 2071314432, "loss": 3.3129391941157253, "ce_loss": 3.3115288723598826, "z_loss": 14.103196787111687, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18787837028503418, "tokens_per_second": 3903185.650002096, "eta_seconds": 5874.247232894927, "mean_data_wait_seconds": 6.657838821411133e-05}
{"step": 3840, "total_steps": 46238, "tokens": 2076721152, "loss": 3.4250325416073655, "ce_loss": 3.4236737706444482, "z_loss": 13.587894627542207, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17615960538387299, "tokens_per_second": 3904070.4289864497, "eta_seconds": 5871.531059943273, "mean_data_wait_seconds": 0.00013688206672668457}
{"step": 3850, "total_steps": 46238, "tokens": 2082127872, "loss": 3.333454339793234, "ce_loss": 3.3320956844272036, "z_loss": 13.586560054258866, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15573401749134064, "tokens_per_second": 3906001.933910996, "eta_seconds": 5867.243399199558, "mean_data_wait_seconds": 0.00012803077697753906}
{"step": 3860, "total_steps": 46238, "tokens": 2087534592, "loss": 3.375711516900496, "ce_loss": 3.3742646213733787, "z_loss": 14.46896735827128, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2366429567337036, "tokens_per_second": 3907284.848401458, "eta_seconds": 5863.933199898068, "mean_data_wait_seconds": 0.00015547871589660645}
{"step": 3870, "total_steps": 46238, "tokens": 2092941312, "loss": 3.364957422921152, "ce_loss": 3.3635560870170593, "z_loss": 14.013103947494969, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1597951501607895, "tokens_per_second": 3908638.286930915, "eta_seconds": 5860.51942964168, "mean_data_wait_seconds": 0.00012165307998657227}
{"step": 3880, "total_steps": 46238, "tokens": 2098348032, "loss": 3.3252492023236826, "ce_loss": 3.32389070409717, "z_loss": 13.585141124147357, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18452082574367523, "tokens_per_second": 3910174.202445652, "eta_seconds": 5856.834687742613, "mean_data_wait_seconds": 6.857514381408691e-05}
{"step": 3890, "total_steps": 46238, "tokens": 2103754752, "loss": 3.3422007687164075, "ce_loss": 3.340758274901997, "z_loss": 14.424603816234704, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1478758007287979, "tokens_per_second": 3911654.742167696, "eta_seconds": 5853.235700273477, "mean_data_wait_seconds": 7.730722427368164e-05}
{"step": 3900, "total_steps": 46238, "tokens": 2109161472, "loss": 3.287814674955426, "ce_loss": 3.286394532882806, "z_loss": 14.201330567851212, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18568480014801025, "tokens_per_second": 3913619.476092554, "eta_seconds": 5848.915716980824, "mean_data_wait_seconds": 8.851289749145508e-05}
{"step": 3910, "total_steps": 46238, "tokens": 2114568192, "loss": 3.3162111621914487, "ce_loss": 3.3148214311310737, "z_loss": 13.89700232852589, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14926010370254517, "tokens_per_second": 3915376.5188894984, "eta_seconds": 5844.910095770504, "mean_data_wait_seconds": 9.894371032714844e-05}
{"step": 3920, "total_steps": 46238, "tokens": 2119974912, "loss": 3.279829540035941, "ce_loss": 3.278463287787004, "z_loss": 13.662325216062142, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16536693274974823, "tokens_per_second": 3916993.1286611347, "eta_seconds": 5841.117477737437, "mean_data_wait_seconds": 8.064508438110352e-05}
{"step": 3930, "total_steps": 46238, "tokens": 2125381632, "loss": 3.2862777186162546, "ce_loss": 3.2849427606120254, "z_loss": 13.349813244559549, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16870376467704773, "tokens_per_second": 3919439.1587427356, "eta_seconds": 5836.0927106054405, "mean_data_wait_seconds": 9.593367576599121e-05}
{"step": 3940, "total_steps": 46238, "tokens": 2130788352, "loss": 3.3281822114279778, "ce_loss": 3.3267806569735208, "z_loss": 14.015944856585879, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16380715370178223, "tokens_per_second": 3920560.099451564, "eta_seconds": 5833.045025173585, "mean_data_wait_seconds": 0.0001730024814605713}
{"step": 3950, "total_steps": 46238, "tokens": 2136195072, "loss": 3.2962684667471684, "ce_loss": 3.2948350021333406, "z_loss": 14.33443647442442, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18023356795310974, "tokens_per_second": 3921222.5722697354, "eta_seconds": 5830.680723324996, "mean_data_wait_seconds": 0.00010144710540771484}
{"step": 3960, "total_steps": 46238, "tokens": 2141601792, "loss": 3.311191204822425, "ce_loss": 3.3098025051030247, "z_loss": 13.887120427507343, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1788351684808731, "tokens_per_second": 3922439.9887864334, "eta_seconds": 5827.492634520088, "mean_data_wait_seconds": 7.969141006469727e-05}
{"step": 3970, "total_steps": 46238, "tokens": 2147008512, "loss": 3.286778352477334, "ce_loss": 3.285358867862008, "z_loss": 14.194791822722465, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20089273154735565, "tokens_per_second": 3924068.8860137537, "eta_seconds": 5823.695783081597, "mean_data_wait_seconds": 0.00033593177795410156}
{"step": 3980, "total_steps": 46238, "tokens": 2152415232, "loss": 3.329076461719744, "ce_loss": 3.327658255894979, "z_loss": 14.181868553161621, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17703212797641754, "tokens_per_second": 3924705.5739926375, "eta_seconds": 5821.373418530697, "mean_data_wait_seconds": 0.00016447901725769043}
{"step": 3990, "total_steps": 46238, "tokens": 2157821952, "loss": 3.19667663899335, "ce_loss": 3.1953247857816294, "z_loss": 13.518495993180709, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17006900906562805, "tokens_per_second": 3925989.05077689, "eta_seconds": 5818.093145084034, "mean_data_wait_seconds": 0.00011959671974182129}
{"step": 4000, "total_steps": 46238, "tokens": 2163228672, "loss": 3.3251280874917, "ce_loss": 3.3237162178212944, "z_loss": 14.119032874251857, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18825413286685944, "tokens_per_second": 3926567.9918999835, "eta_seconds": 5815.858355466797, "mean_data_wait_seconds": 0.00014218688011169434}
{"step": 4010, "total_steps": 46238, "tokens": 2168635392, "loss": 3.27069580374342, "ce_loss": 3.2692870606075632, "z_loss": 14.08748222842361, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18324142694473267, "tokens_per_second": 3927256.000596907, "eta_seconds": 5813.462769050427, "mean_data_wait_seconds": 0.00010848045349121094}
{"step": 4020, "total_steps": 46238, "tokens": 2174042112, "loss": 3.2452801195057956, "ce_loss": 3.243911062226151, "z_loss": 13.690159559249878, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.162391796708107, "tokens_per_second": 3928170.3386560343, "eta_seconds": 5810.733205579223, "mean_data_wait_seconds": 0.00019723176956176758}
{"step": 4030, "total_steps": 46238, "tokens": 2179448832, "loss": 3.349926551183065, "ce_loss": 3.3485844243656504, "z_loss": 13.421320431160204, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18585364520549774, "tokens_per_second": 3929697.418799725, "eta_seconds": 5807.099293403134, "mean_data_wait_seconds": 0.00013267993927001953}
{"step": 4040, "total_steps": 46238, "tokens": 2184855552, "loss": 3.263138458584294, "ce_loss": 3.2617763100248394, "z_loss": 13.621612881169174, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16657227277755737, "tokens_per_second": 3930575.659876667, "eta_seconds": 5804.426210871075, "mean_data_wait_seconds": 0.00010532140731811523}
{"step": 4050, "total_steps": 46238, "tokens": 2190262272, "loss": 3.288501930959297, "ce_loss": 3.2870962764277603, "z_loss": 14.056355172937566, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18257026374340057, "tokens_per_second": 3931409.996314735, "eta_seconds": 5801.8191146131385, "mean_data_wait_seconds": 9.706616401672363e-05}
{"step": 4060, "total_steps": 46238, "tokens": 2195668992, "loss": 3.3244028561043018, "ce_loss": 3.3230264656471484, "z_loss": 13.764221458724052, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1595294326543808, "tokens_per_second": 3932072.7402901906, "eta_seconds": 5799.466197646448, "mean_data_wait_seconds": 0.0001055002212524414}
{"step": 4070, "total_steps": 46238, "tokens": 2201075712, "loss": 3.328060603503025, "ce_loss": 3.326706638841918, "z_loss": 13.539872342889959, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16531257331371307, "tokens_per_second": 3933179.6215276127, "eta_seconds": 5796.459459724663, "mean_data_wait_seconds": 6.896257400512695e-05}
{"step": 4080, "total_steps": 46238, "tokens": 2206482432, "loss": 3.363628255598473, "ce_loss": 3.3622094176032324, "z_loss": 14.188272245002516, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22032995522022247, "tokens_per_second": 3934450.4896155074, "eta_seconds": 5793.212944008211, "mean_data_wait_seconds": 0.00015214085578918457}
{"step": 4090, "total_steps": 46238, "tokens": 2211889152, "loss": 3.334022218530828, "ce_loss": 3.3326111761006443, "z_loss": 14.110454935016055, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16145572066307068, "tokens_per_second": 3935515.51615783, "eta_seconds": 5790.271361005129, "mean_data_wait_seconds": 0.00011086463928222656}
{"step": 4100, "total_steps": 46238, "tokens": 2217295872, "loss": 3.3186792695161067, "ce_loss": 3.317343831062317, "z_loss": 13.354644081809305, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16711926460266113, "tokens_per_second": 3936531.16488595, "eta_seconds": 5787.403962965972, "mean_data_wait_seconds": 9.003281593322754e-05}
{"step": 4110, "total_steps": 46238, "tokens": 2222702592, "loss": 3.3131412719235276, "ce_loss": 3.3117273648579917, "z_loss": 14.138738718899814, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17157220840454102, "tokens_per_second": 3937926.5286619323, "eta_seconds": 5783.980269367636, "mean_data_wait_seconds": 5.78463077545166e-05}
{"step": 4120, "total_steps": 46238, "tokens": 2228109312, "loss": 3.3733567559357844, "ce_loss": 3.3719437971259607, "z_loss": 14.129315282359268, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20267237722873688, "tokens_per_second": 3938735.6806662623, "eta_seconds": 5781.419336102304, "mean_data_wait_seconds": 0.0003356039524078369}
{"step": 4130, "total_steps": 46238, "tokens": 2233516032, "loss": 3.3131147478566025, "ce_loss": 3.3117601492188196, "z_loss": 13.545936519449407, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16631916165351868, "tokens_per_second": 3939659.5021053962, "eta_seconds": 5778.691252843949, "mean_data_wait_seconds": 6.476044654846191e-05}
{"step": 4140, "total_steps": 46238, "tokens": 2238922752, "loss": 3.2922868421583464, "ce_loss": 3.2909112681042063, "z_loss": 13.755553556211067, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14238108694553375, "tokens_per_second": 3941218.810998461, "eta_seconds": 5775.033124393784, "mean_data_wait_seconds": 7.343292236328125e-05}
{"step": 4150, "total_steps": 46238, "tokens": 2244329472, "loss": 3.2876065409544744, "ce_loss": 3.2862695329117053, "z_loss": 13.370536291238034, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18198688328266144, "tokens_per_second": 3942582.0991408713, "eta_seconds": 5771.664835834009, "mean_data_wait_seconds": 7.453560829162598e-05}
{"step": 4160, "total_steps": 46238, "tokens": 2249736192, "loss": 3.3184870662111225, "ce_loss": 3.3171003984682486, "z_loss": 13.866608475193832, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18024587631225586, "tokens_per_second": 3943128.57703057, "eta_seconds": 5769.493766072449, "mean_data_wait_seconds": 6.67870044708252e-05}
{"step": 4170, "total_steps": 46238, "tokens": 2255142912, "loss": 3.3209427017154116, "ce_loss": 3.3196193846789273, "z_loss": 13.233196489738695, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17146971821784973, "tokens_per_second": 3944992.477643367, "eta_seconds": 5765.397311375084, "mean_data_wait_seconds": 6.48796558380127e-05}
{"step": 4180, "total_steps": 46238, "tokens": 2260549632, "loss": 3.323542564204245, "ce_loss": 3.3222271673607104, "z_loss": 13.153783184109312, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18753452599048615, "tokens_per_second": 3945848.451588323, "eta_seconds": 5762.776392196931, "mean_data_wait_seconds": 7.11977481842041e-05}
{"step": 4190, "total_steps": 46238, "tokens": 2265956352, "loss": 3.3422954281171164, "ce_loss": 3.3409614418492173, "z_loss": 13.339832472078728, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17177222669124603, "tokens_per_second": 3946426.382161432, "eta_seconds": 5760.562438655941, "mean_data_wait_seconds": 8.749961853027344e-05}
{"step": 4200, "total_steps": 46238, "tokens": 2271363072, "loss": 3.342468241850535, "ce_loss": 3.341067287054929, "z_loss": 14.009675163211245, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1635381579399109, "tokens_per_second": 3946529.2222081586, "eta_seconds": 5759.042334236948, "mean_data_wait_seconds": 8.088350296020508e-05}
{"step": 4210, "total_steps": 46238, "tokens": 2276769792, "loss": 3.291268213228746, "ce_loss": 3.2898744597579492, "z_loss": 13.937517426230691, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15072967112064362, "tokens_per_second": 3946371.4576366036, "eta_seconds": 5757.902515747518, "mean_data_wait_seconds": 9.608268737792969e-05}
{"step": 4220, "total_steps": 46238, "tokens": 2282176512, "loss": 3.284005091045842, "ce_loss": 3.2826211163491914, "z_loss": 13.839630632689506, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17436614632606506, "tokens_per_second": 3947064.370257925, "eta_seconds": 5755.521900068609, "mean_data_wait_seconds": 9.581446647644043e-05}
{"step": 4230, "total_steps": 46238, "tokens": 2287583232, "loss": 3.2489928368366128, "ce_loss": 3.247646071694114, "z_loss": 13.467860929893725, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.13501125574111938, "tokens_per_second": 3948184.7311203307, "eta_seconds": 5752.519259035601, "mean_data_wait_seconds": 7.042288780212402e-05}
{"step": 4240, "total_steps": 46238, "tokens": 2292989952, "loss": 3.299691704186526, "ce_loss": 3.298317855054682, "z_loss": 13.738637194488987, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1678764373064041, "tokens_per_second": 3948907.8489174056, "eta_seconds": 5750.096698312426, "mean_data_wait_seconds": 7.331371307373047e-05}
{"step": 4250, "total_steps": 46238, "tokens": 2298396672, "loss": 3.3234130974971885, "ce_loss": 3.322073661919796, "z_loss": 13.394454739310525, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14089329540729523, "tokens_per_second": 3949859.183042623, "eta_seconds": 5747.342933505038, "mean_data_wait_seconds": 9.176135063171387e-05}
{"step": 4260, "total_steps": 46238, "tokens": 2303803392, "loss": 3.2810645880121174, "ce_loss": 3.2797325181238577, "z_loss": 13.320686809944384, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17335885763168335, "tokens_per_second": 3949787.5433097007, "eta_seconds": 5746.078313109014, "mean_data_wait_seconds": 7.289648056030273e-05}
{"step": 4270, "total_steps": 46238, "tokens": 2309210112, "loss": 3.2585001244689478, "ce_loss": 3.2571789206880513, "z_loss": 13.212238290093161, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17317572236061096, "tokens_per_second": 3951030.490241046, "eta_seconds": 5742.902232732629, "mean_data_wait_seconds": 8.016824722290039e-05}
{"step": 4280, "total_steps": 46238, "tokens": 2314616832, "loss": 3.2830512794581326, "ce_loss": 3.281719496755889, "z_loss": 13.317790175929215, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17530763149261475, "tokens_per_second": 3952077.644892084, "eta_seconds": 5740.012505401937, "mean_data_wait_seconds": 0.00013369321823120117}
{"step": 4290, "total_steps": 46238, "tokens": 2320023552, "loss": 3.2605931957562766, "ce_loss": 3.2592838352376763, "z_loss": 13.093372814583056, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16749243438243866, "tokens_per_second": 3952994.735933045, "eta_seconds": 5737.313075031664, "mean_data_wait_seconds": 0.00010764598846435547}
{"step": 4300, "total_steps": 46238, "tokens": 2325430272, "loss": 3.305518215352839, "ce_loss": 3.304140840515946, "z_loss": 13.774079236117275, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1585158109664917, "tokens_per_second": 3954747.8315659272, "eta_seconds": 5733.4026415084745, "mean_data_wait_seconds": 0.0001716017723083496}
{"step": 4310, "total_steps": 46238, "tokens": 2330836992, "loss": 3.2017325054515493, "ce_loss": 3.2003649274508157, "z_loss": 13.675655523935953, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17503398656845093, "tokens_per_second": 3956085.581669602, "eta_seconds": 5730.097207460567, "mean_data_wait_seconds": 0.00013199448585510254}
{"step": 4320, "total_steps": 46238, "tokens": 2336243712, "loss": 3.247838405045596, "ce_loss": 3.246453137108774, "z_loss": 13.852726437828757, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1424909383058548, "tokens_per_second": 3957136.9676093077, "eta_seconds": 5727.208436177026, "mean_data_wait_seconds": 8.097290992736816e-05}
{"step": 4330, "total_steps": 46238, "tokens": 2341650432, "loss": 3.34943493568536, "ce_loss": 3.3480576782515556, "z_loss": 13.772626660086893, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18751806020736694, "tokens_per_second": 3958720.751580672, "eta_seconds": 5723.551350509617, "mean_data_wait_seconds": 0.0004062056541442871}
{"step": 4340, "total_steps": 46238, "tokens": 2347057152, "loss": 3.2790194460839936, "ce_loss": 3.2777040167288347, "z_loss": 13.153930114977287, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1594265103340149, "tokens_per_second": 3959858.2579636904, "eta_seconds": 5720.541824557325, "mean_data_wait_seconds": 7.843971252441406e-05}
{"step": 4350, "total_steps": 46238, "tokens": 2352463872, "loss": 3.3143211061304267, "ce_loss": 3.312941592751127, "z_loss": 13.794618071931781, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17858922481536865, "tokens_per_second": 3960426.6442590505, "eta_seconds": 5718.355646563683, "mean_data_wait_seconds": 6.99758529663086e-05}
{"step": 4360, "total_steps": 46238, "tokens": 2357870592, "loss": 3.306197263977744, "ce_loss": 3.3048621542526013, "z_loss": 13.351569898200758, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14283736050128937, "tokens_per_second": 3961740.528841569, "eta_seconds": 5715.094458904542, "mean_data_wait_seconds": 8.735060691833496e-05}
{"step": 4370, "total_steps": 46238, "tokens": 2363277312, "loss": 3.298352553988948, "ce_loss": 3.297001022281069, "z_loss": 13.515145822004838, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1952953189611435, "tokens_per_second": 3962337.6826941306, "eta_seconds": 5712.868623708211, "mean_data_wait_seconds": 8.195638656616211e-05}
{"step": 4380, "total_steps": 46238, "tokens": 2368684032, "loss": 3.2805627566395383, "ce_loss": 3.279196470072775, "z_loss": 13.66266064932852, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17573225498199463, "tokens_per_second": 3963303.1262675426, "eta_seconds": 5710.112798087375, "mean_data_wait_seconds": 6.282329559326172e-05}
{"step": 4390, "total_steps": 46238, "tokens": 2374090752, "loss": 3.2710079749425254, "ce_loss": 3.269641607096701, "z_loss": 13.663592107368238, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17706389725208282, "tokens_per_second": 3964928.568493638, "eta_seconds": 5706.408272721018, "mean_data_wait_seconds": 5.97834587097168e-05}
{"step": 4400, "total_steps": 46238, "tokens": 2379497472, "loss": 3.2824486367630237, "ce_loss": 3.281092016985922, "z_loss": 13.566445459019054, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18304027616977692, "tokens_per_second": 3966076.3986972277, "eta_seconds": 5703.393528029421, "mean_data_wait_seconds": 7.349252700805664e-05}
{"step": 4410, "total_steps": 46238, "tokens": 2384904192, "loss": 3.281354380376411, "ce_loss": 3.279947689085296, "z_loss": 14.067046143791892, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15515266358852386, "tokens_per_second": 3967425.8246535924, "eta_seconds": 5700.090875920674, "mean_data_wait_seconds": 7.936358451843262e-05}
{"step": 4420, "total_steps": 46238, "tokens": 2390310912, "loss": 3.3497711383935176, "ce_loss": 3.3484250903129578, "z_loss": 13.46022130503799, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1903311312198639, "tokens_per_second": 3967765.826903541, "eta_seconds": 5698.239767754733, "mean_data_wait_seconds": 0.0001023411750793457}
{"step": 4430, "total_steps": 46238, "tokens": 2395717632, "loss": 3.2922766371206804, "ce_loss": 3.290912536057559, "z_loss": 13.640861706300216, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16616332530975342, "tokens_per_second": 3969079.3322436446, "eta_seconds": 5694.9918134345935, "mean_data_wait_seconds": 0.00010803341865539551}
{"step": 4440, "total_steps": 46238, "tokens": 2401124352, "loss": 3.259844052069115, "ce_loss": 3.2584807710214094, "z_loss": 13.632912173415676, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17989908158779144, "tokens_per_second": 3969835.742794637, "eta_seconds": 5692.544741937208, "mean_data_wait_seconds": 7.852911949157715e-05}
{"step": 4450, "total_steps": 46238, "tokens": 2406531072, "loss": 3.263443471807422, "ce_loss": 3.2621260754989856, "z_loss": 13.173427112174757, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1724298596382141, "tokens_per_second": 3969939.819898334, "eta_seconds": 5691.03358966751, "mean_data_wait_seconds": 9.524822235107422e-05}
{"step": 4460, "total_steps": 46238, "tokens": 2411937792, "loss": 3.318016798207254, "ce_loss": 3.3167459043589504, "z_loss": 12.708634860587843, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17130877077579498, "tokens_per_second": 3969866.9546156046, "eta_seconds": 5689.776106410378, "mean_data_wait_seconds": 0.00010630488395690918}
{"step": 4470, "total_steps": 46238, "tokens": 2417344512, "loss": 3.288420501983527, "ce_loss": 3.2870828220338533, "z_loss": 13.377164941845518, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15831807255744934, "tokens_per_second": 3970515.220062187, "eta_seconds": 5687.48542000207, "mean_data_wait_seconds": 6.851553916931152e-05}
{"step": 4480, "total_steps": 46238, "tokens": 2422751232, "loss": 3.2513506213823953, "ce_loss": 3.2499729214292583, "z_loss": 13.77729480194323, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17324140667915344, "tokens_per_second": 3971381.441834532, "eta_seconds": 5684.883468048563, "mean_data_wait_seconds": 7.82310962677002e-05}
{"step": 4490, "total_steps": 46238, "tokens": 2428157952, "loss": 3.250526518532724, "ce_loss": 3.2491501407189802, "z_loss": 13.763999086437803, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16320490837097168, "tokens_per_second": 3971580.8146460126, "eta_seconds": 5683.2367355495435, "mean_data_wait_seconds": 9.682774543762207e-05}
{"step": 4500, "total_steps": 46238, "tokens": 2433564672, "loss": 3.2281190771045107, "ce_loss": 3.226752467227705, "z_loss": 13.666020480069248, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1797851175069809, "tokens_per_second": 3972722.517934957, "eta_seconds": 5680.242494190091, "mean_data_wait_seconds": 7.957220077514648e-05}
{"step": 4510, "total_steps": 46238, "tokens": 2438971392, "loss": 3.2868461645010747, "ce_loss": 3.2855100270473594, "z_loss": 13.361180551124342, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17001958191394806, "tokens_per_second": 3974108.9191879267, "eta_seconds": 5676.900407805144, "mean_data_wait_seconds": 7.179379463195801e-05}
{"step": 4520, "total_steps": 46238, "tokens": 2444378112, "loss": 3.2859446984348875, "ce_loss": 3.2846067692294265, "z_loss": 13.379364591656309, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18647295236587524, "tokens_per_second": 3975625.3120567496, "eta_seconds": 5673.375143174971, "mean_data_wait_seconds": 6.93202018737793e-05}
{"step": 4530, "total_steps": 46238, "tokens": 2449784832, "loss": 3.2391030084003103, "ce_loss": 3.2377916339671975, "z_loss": 13.113551479397398, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16776323318481445, "tokens_per_second": 3976845.2520597195, "eta_seconds": 5670.275224393211, "mean_data_wait_seconds": 0.0001043081283569336}
{"step": 4540, "total_steps": 46238, "tokens": 2455191552, "loss": 3.255036585258715, "ce_loss": 3.253736244909691, "z_loss": 13.00378114526922, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1889069676399231, "tokens_per_second": 3978512.24382064, "eta_seconds": 5666.540405654298, "mean_data_wait_seconds": 0.00012314319610595703}
{"step": 4550, "total_steps": 46238, "tokens": 2460598272, "loss": 3.2741296164917224, "ce_loss": 3.27286178596092, "z_loss": 12.678667444171328, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19846248626708984, "tokens_per_second": 3980437.193407807, "eta_seconds": 5662.441729096469, "mean_data_wait_seconds": 7.531046867370605e-05}
{"step": 4560, "total_steps": 46238, "tokens": 2466004992, "loss": 3.246398403789058, "ce_loss": 3.245110304066629, "z_loss": 12.880964149128307, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.13617847859859467, "tokens_per_second": 3981942.1352580325, "eta_seconds": 5658.943846641259, "mean_data_wait_seconds": 8.687376976013184e-05}
{"step": 4570, "total_steps": 46238, "tokens": 2471411712, "loss": 3.314114825292067, "ce_loss": 3.3127811803962244, "z_loss": 13.336265181050155, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16372330486774445, "tokens_per_second": 3983165.103080327, "eta_seconds": 5655.848964577977, "mean_data_wait_seconds": 0.00010496377944946289}
{"step": 4580, "total_steps": 46238, "tokens": 2476818432, "loss": 3.3139137679880317, "ce_loss": 3.312568399039182, "z_loss": 13.454052831187393, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19303801655769348, "tokens_per_second": 3983886.3448754163, "eta_seconds": 5653.467883934909, "mean_data_wait_seconds": 6.988644599914551e-05}
{"step": 4590, "total_steps": 46238, "tokens": 2482225152, "loss": 3.2571334820805173, "ce_loss": 3.2557969906113366, "z_loss": 13.364895972338589, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15807650983333588, "tokens_per_second": 3984787.9640960754, "eta_seconds": 5650.831860286429, "mean_data_wait_seconds": 9.09268856048584e-05}
{"step": 4600, "total_steps": 46238, "tokens": 2487631872, "loss": 3.3616333549672905, "ce_loss": 3.360308587551117, "z_loss": 13.24762124003786, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16454248130321503, "tokens_per_second": 3985120.782458201, "eta_seconds": 5649.00320288752, "mean_data_wait_seconds": 7.560849189758301e-05}
{"step": 4610, "total_steps": 46238, "tokens": 2493038592, "loss": 3.2868692405296094, "ce_loss": 3.2855501608415083, "z_loss": 13.19090310010043, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16048407554626465, "tokens_per_second": 3986607.5568326223, "eta_seconds": 5645.540230170425, "mean_data_wait_seconds": 5.6684017181396484e-05}
{"step": 4620, "total_steps": 46238, "tokens": 2498445312, "loss": 3.335456582632932, "ce_loss": 3.3341078415061487, "z_loss": 13.486978870449644, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18731442093849182, "tokens_per_second": 3987704.890653527, "eta_seconds": 5642.630846815846, "mean_data_wait_seconds": 7.385015487670898e-05}
{"step": 4630, "total_steps": 46238, "tokens": 2503852032, "loss": 3.342140714327494, "ce_loss": 3.340828602964228, "z_loss": 13.121134288383253, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1877896934747696, "tokens_per_second": 3988158.0936309393, "eta_seconds": 5640.633940747119, "mean_data_wait_seconds": 7.155537605285645e-05}
{"step": 4640, "total_steps": 46238, "tokens": 2509258752, "loss": 3.288210265564196, "ce_loss": 3.2868695927388742, "z_loss": 13.406888506629251, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16684550046920776, "tokens_per_second": 3988850.435663836, "eta_seconds": 5638.299441592649, "mean_data_wait_seconds": 8.079409599304199e-05}
{"step": 4650, "total_steps": 46238, "tokens": 2514665472, "loss": 3.246735556559129, "ce_loss": 3.245412927685362, "z_loss": 13.226259744528567, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15749317407608032, "tokens_per_second": 3989647.465991828, "eta_seconds": 5635.817865027891, "mean_data_wait_seconds": 6.54757022857666e-05}
{"step": 4660, "total_steps": 46238, "tokens": 2520072192, "loss": 3.342393873315869, "ce_loss": 3.3410692540082065, "z_loss": 13.246069857568452, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18933291733264923, "tokens_per_second": 3990079.8898355616, "eta_seconds": 5633.852044232232, "mean_data_wait_seconds": 9.128451347351074e-05}
{"step": 4670, "total_steps": 46238, "tokens": 2525478912, "loss": 3.308428518699877, "ce_loss": 3.3070839751850474, "z_loss": 13.445614821983106, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16897745430469513, "tokens_per_second": 3990307.5714655207, "eta_seconds": 5632.175620924862, "mean_data_wait_seconds": 8.377432823181152e-05}
{"step": 4680, "total_steps": 46238, "tokens": 2530885632, "loss": 3.317707155690049, "ce_loss": 3.3163884914282598, "z_loss": 13.186268206798669, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16207748651504517, "tokens_per_second": 3991361.0249650464, "eta_seconds": 5629.334496043681, "mean_data_wait_seconds": 7.56978988647461e-05}
{"step": 4690, "total_steps": 46238, "tokens": 2536292352, "loss": 3.3538515098167188, "ce_loss": 3.3525650410941155, "z_loss": 12.865051038337477, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1582772582769394, "tokens_per_second": 3992590.3146689134, "eta_seconds": 5626.247076107225, "mean_data_wait_seconds": 6.890296936035156e-05}
{"step": 4700, "total_steps": 46238, "tokens": 2541699072, "loss": 3.299591328158523, "ce_loss": 3.2982782649271414, "z_loss": 13.130222912990686, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1765286773443222, "tokens_per_second": 3994338.316199682, "eta_seconds": 5622.431323085078, "mean_data_wait_seconds": 0.00012105703353881836}
{"step": 4710, "total_steps": 46238, "tokens": 2547105792, "loss": 3.2789116899172464, "ce_loss": 3.2776890379009824, "z_loss": 12.226575569673019, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1907995641231537, "tokens_per_second": 3995311.0068726423, "eta_seconds": 5619.709230489879, "mean_data_wait_seconds": 0.00015103816986083984}
{"step": 4720, "total_steps": 46238, "tokens": 2552512512, "loss": 3.330324722058845, "ce_loss": 3.3290075876496057, "z_loss": 13.171353947032582, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1534404158592224, "tokens_per_second": 3996008.1396156196, "eta_seconds": 5617.375800981028, "mean_data_wait_seconds": 7.402896881103516e-05}
{"step": 4730, "total_steps": 46238, "tokens": 2557919232, "loss": 3.3079957058935454, "ce_loss": 3.306659570246032, "z_loss": 13.360888625636246, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16631478071212769, "tokens_per_second": 3997201.548365393, "eta_seconds": 5614.346044966696, "mean_data_wait_seconds": 8.478760719299316e-05}
{"step": 4740, "total_steps": 46238, "tokens": 2563325952, "loss": 3.3290134541916125, "ce_loss": 3.3277057987270933, "z_loss": 13.076196200919874, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.12914012372493744, "tokens_per_second": 3998064.7343853847, "eta_seconds": 5611.781567976309, "mean_data_wait_seconds": 9.53376293182373e-05}
{"step": 4750, "total_steps": 46238, "tokens": 2568732672, "loss": 3.1347901261214055, "ce_loss": 3.133537834340876, "z_loss": 12.522549405242458, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18741081655025482, "tokens_per_second": 3998758.8345803074, "eta_seconds": 5609.45538151071, "mean_data_wait_seconds": 6.860494613647461e-05}
{"step": 4760, "total_steps": 46238, "tokens": 2574139392, "loss": 3.265315539909132, "ce_loss": 3.2640410155961006, "z_loss": 12.745576988567006, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20657995343208313, "tokens_per_second": 3999672.3423901647, "eta_seconds": 5606.822415507859, "mean_data_wait_seconds": 8.025765419006348e-05}
{"step": 4770, "total_steps": 46238, "tokens": 2579546112, "loss": 3.293474607395403, "ce_loss": 3.292186657587687, "z_loss": 12.878999334393125, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15740302205085754, "tokens_per_second": 4000398.803344194, "eta_seconds": 5604.452687381474, "mean_data_wait_seconds": 6.628036499023438e-05}
{"step": 4780, "total_steps": 46238, "tokens": 2584952832, "loss": 3.2867753379272693, "ce_loss": 3.2854347987608477, "z_loss": 13.405394221797135, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17741279304027557, "tokens_per_second": 4001764.8590128343, "eta_seconds": 5601.18844902079, "mean_data_wait_seconds": 0.0001049339771270752}
{"step": 4790, "total_steps": 46238, "tokens": 2590359552, "loss": 3.230077660445011, "ce_loss": 3.228737159208818, "z_loss": 13.405236330899326, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1717788726091385, "tokens_per_second": 4002552.0973341907, "eta_seconds": 5598.735966216445, "mean_data_wait_seconds": 0.00010582804679870605}
{"step": 4800, "total_steps": 46238, "tokens": 2595766272, "loss": 3.233719399481109, "ce_loss": 3.2323772672450906, "z_loss": 13.421481226429794, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16803915798664093, "tokens_per_second": 4003847.3835491203, "eta_seconds": 5595.574335838104, "mean_data_wait_seconds": 9.450316429138184e-05}
{"step": 4810, "total_steps": 46238, "tokens": 2601172992, "loss": 3.267885152137641, "ce_loss": 3.266572401379094, "z_loss": 13.127395543185147, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1552276611328125, "tokens_per_second": 4005129.768660176, "eta_seconds": 5592.432764417737, "mean_data_wait_seconds": 0.00010800361633300781}
{"step": 4820, "total_steps": 46238, "tokens": 2606579712, "loss": 3.2586853197126677, "ce_loss": 3.257432036327593, "z_loss": 12.532854650959823, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1899901181459427, "tokens_per_second": 4005932.1794195916, "eta_seconds": 5589.962890296476, "mean_data_wait_seconds": 6.845593452453613e-05}
{"step": 4830, "total_steps": 46238, "tokens": 2611986432, "loss": 3.2524997808716516, "ce_loss": 3.251228894248153, "z_loss": 12.70939446218086, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1713312268257141, "tokens_per_second": 4007128.8443211606, "eta_seconds": 5586.944262031245, "mean_data_wait_seconds": 8.460879325866699e-05}
{"step": 4840, "total_steps": 46238, "tokens": 2617393152, "loss": 3.3373423724463493, "ce_loss": 3.3360438455234873, "z_loss": 12.985161499543624, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15837909281253815, "tokens_per_second": 4008365.986857418, "eta_seconds": 5583.871048049625, "mean_data_wait_seconds": 8.064508438110352e-05}
{"step": 4850, "total_steps": 46238, "tokens": 2622799872, "loss": 3.3340064955480173, "ce_loss": 3.3327080408732095, "z_loss": 12.984525333751332, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2063068300485611, "tokens_per_second": 4009938.757864927, "eta_seconds": 5580.3326223152635, "mean_data_wait_seconds": 8.627772331237793e-05}
{"step": 4860, "total_steps": 46238, "tokens": 2628206592, "loss": 3.26931447874416, "ce_loss": 3.268028542850957, "z_loss": 12.85938301230922, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15875445306301117, "tokens_per_second": 4010492.939853585, "eta_seconds": 5578.213371650203, "mean_data_wait_seconds": 8.463859558105469e-05}
{"step": 4870, "total_steps": 46238, "tokens": 2633613312, "loss": 3.3271853905735593, "ce_loss": 3.325894355773926, "z_loss": 12.910254189462373, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1683720350265503, "tokens_per_second": 4011504.2051028847, "eta_seconds": 5575.459349026501, "mean_data_wait_seconds": 0.00010690093040466309}
{"step": 4880, "total_steps": 46238, "tokens": 2639020032, "loss": 3.330589068658424, "ce_loss": 3.3292615269169663, "z_loss": 13.275539029728282, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17435507476329803, "tokens_per_second": 4012630.480831774, "eta_seconds": 5572.546989017763, "mean_data_wait_seconds": 9.420514106750488e-05}
{"step": 4890, "total_steps": 46238, "tokens": 2644426752, "loss": 3.3024026986324424, "ce_loss": 3.301098928307042, "z_loss": 13.03772748600353, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15314233303070068, "tokens_per_second": 4013726.3332302254, "eta_seconds": 5569.678480298552, "mean_data_wait_seconds": 8.004903793334961e-05}
{"step": 4900, "total_steps": 46238, "tokens": 2649833472, "loss": 3.2681251240499094, "ce_loss": 3.2668332403356377, "z_loss": 12.918315172195435, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16908828914165497, "tokens_per_second": 4014579.539337609, "eta_seconds": 5567.14800267418, "mean_data_wait_seconds": 7.906556129455566e-05}
{"step": 4910, "total_steps": 46238, "tokens": 2655240192, "loss": 3.2618513143423833, "ce_loss": 3.260541325265711, "z_loss": 13.100174405358054, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17415085434913635, "tokens_per_second": 4011525.2868218417, "eta_seconds": 5570.038862126297, "mean_data_wait_seconds": 6.22868537902832e-05}
{"step": 4920, "total_steps": 46238, "tokens": 2660646912, "loss": 3.326354915445501, "ce_loss": 3.3250822135896394, "z_loss": 12.727097077803178, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17004238069057465, "tokens_per_second": 4009661.8986996403, "eta_seconds": 5571.278972734501, "mean_data_wait_seconds": 0.06330934166908264}
{"step": 4930, "total_steps": 46238, "tokens": 2666053632, "loss": 3.302623560934356, "ce_loss": 3.301338448668971, "z_loss": 12.85142756953384, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1640615314245224, "tokens_per_second": 4010200.371752293, "eta_seconds": 5569.182642672082, "mean_data_wait_seconds": 0.00011655688285827637}
{"step": 4940, "total_steps": 46238, "tokens": 2671460352, "loss": 3.273482828429251, "ce_loss": 3.272176991809498, "z_loss": 13.058193611376213, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16434386372566223, "tokens_per_second": 4010336.1256459393, "eta_seconds": 5567.645924044245, "mean_data_wait_seconds": 7.161498069763184e-05}
{"step": 4950, "total_steps": 46238, "tokens": 2676867072, "loss": 3.2675787282712534, "ce_loss": 3.2662714268222, "z_loss": 13.072912281209772, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1638478934764862, "tokens_per_second": 4010656.129581378, "eta_seconds": 5565.853601697334, "mean_data_wait_seconds": 0.0001061856746673584}
{"step": 4960, "total_steps": 46238, "tokens": 2682273792, "loss": 3.2863667571183406, "ce_loss": 3.285023691076221, "z_loss": 13.431006727796612, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18180984258651733, "tokens_per_second": 4010878.5743903965, "eta_seconds": 5564.1969035155735, "mean_data_wait_seconds": 7.796287536621094e-05}
{"step": 4970, "total_steps": 46238, "tokens": 2687680512, "loss": 3.2999991630062913, "ce_loss": 3.2987264159953957, "z_loss": 12.727202986225937, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15109115839004517, "tokens_per_second": 4011507.264676778, "eta_seconds": 5561.977070431095, "mean_data_wait_seconds": 0.00022169947624206543}
{"step": 4980, "total_steps": 46238, "tokens": 2693087232, "loss": 3.3277018250841084, "ce_loss": 3.326453644217867, "z_loss": 12.48225322636691, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17828330397605896, "tokens_per_second": 4011934.220181413, "eta_seconds": 5560.037498070279, "mean_data_wait_seconds": 0.00010839104652404785}
{"step": 4990, "total_steps": 46238, "tokens": 2698493952, "loss": 3.271364983284112, "ce_loss": 3.270089917110674, "z_loss": 12.750413136048751, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16062858700752258, "tokens_per_second": 4012182.2351756147, "eta_seconds": 5558.346225772537, "mean_data_wait_seconds": 0.0002155303955078125}
{"step": 5000, "total_steps": 46238, "tokens": 2703900672, "loss": 3.296524056882569, "ce_loss": 3.295239401586128, "z_loss": 12.846630529923873, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15214936435222626, "tokens_per_second": 4013342.8152227267, "eta_seconds": 5555.391674848157, "mean_data_wait_seconds": 0.0001437366008758545}
{"step": 5010, "total_steps": 46238, "tokens": 2709307392, "loss": 3.2829208933945857, "ce_loss": 3.2816588246461116, "z_loss": 12.62056211269263, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.190052792429924, "tokens_per_second": 4013209.717268472, "eta_seconds": 5554.228688345629, "mean_data_wait_seconds": 8.669495582580566e-05}
{"step": 5020, "total_steps": 46238, "tokens": 2714714112, "loss": 3.35253924853874, "ce_loss": 3.3512773477669917, "z_loss": 12.619399692073014, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20312537252902985, "tokens_per_second": 4013558.0694951597, "eta_seconds": 5552.399501423702, "mean_data_wait_seconds": 7.560849189758301e-05}
{"step": 5030, "total_steps": 46238, "tokens": 2720120832, "loss": 3.2808138529459634, "ce_loss": 3.2795203689372903, "z_loss": 12.93427291060939, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1503116935491562, "tokens_per_second": 4014347.841886912, "eta_seconds": 5549.960287827901, "mean_data_wait_seconds": 5.888938903808594e-05}
{"step": 5040, "total_steps": 46238, "tokens": 2725527552, "loss": 3.302662856651075, "ce_loss": 3.3014164458621633, "z_loss": 12.464160738569317, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17369280755519867, "tokens_per_second": 4014855.251132437, "eta_seconds": 5547.91218879369, "mean_data_wait_seconds": 6.0945749282836914e-05}
{"step": 5050, "total_steps": 46238, "tokens": 2730934272, "loss": 3.2798881928126016, "ce_loss": 3.278604001709909, "z_loss": 12.841929442954786, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16883718967437744, "tokens_per_second": 4015908.857389263, "eta_seconds": 5545.110323663278, "mean_data_wait_seconds": 6.532669067382812e-05}
{"step": 5060, "total_steps": 46238, "tokens": 2736340992, "loss": 3.283783972263336, "ce_loss": 3.282457010312514, "z_loss": 13.2695769107703, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17422987520694733, "tokens_per_second": 4016965.238446808, "eta_seconds": 5542.3060998676365, "mean_data_wait_seconds": 7.385015487670898e-05}
{"step": 5070, "total_steps": 46238, "tokens": 2741747712, "loss": 3.2477234692284553, "ce_loss": 3.2464231884840764, "z_loss": 13.00308664639791, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.13895897567272186, "tokens_per_second": 4018153.8026397103, "eta_seconds": 5539.321120405545, "mean_data_wait_seconds": 0.00016555190086364746}
{"step": 5080, "total_steps": 46238, "tokens": 2747154432, "loss": 3.2752227476148894, "ce_loss": 3.2739113840189846, "z_loss": 13.113829121445164, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15958845615386963, "tokens_per_second": 4018815.716442205, "eta_seconds": 5537.0634221814325, "mean_data_wait_seconds": 9.810924530029297e-05}
{"step": 5090, "total_steps": 46238, "tokens": 2752561152, "loss": 3.267982938072898, "ce_loss": 3.2667185086192507, "z_loss": 12.64431584965099, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15780621767044067, "tokens_per_second": 4019511.8665729025, "eta_seconds": 5534.759324636143, "mean_data_wait_seconds": 0.00010785460472106934}
{"step": 5100, "total_steps": 46238, "tokens": 2757967872, "loss": 3.3435431411772063, "ce_loss": 3.342236309340506, "z_loss": 13.067954287384495, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19840894639492035, "tokens_per_second": 4020632.153043479, "eta_seconds": 5531.872406473161, "mean_data_wait_seconds": 0.00010284781455993652}
{"step": 5110, "total_steps": 46238, "tokens": 2763374592, "loss": 3.2928134925437695, "ce_loss": 3.2915108294198006, "z_loss": 13.026148976701679, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17668579518795013, "tokens_per_second": 4020980.6669814438, "eta_seconds": 5530.048310501518, "mean_data_wait_seconds": 0.00012254714965820312}
{"step": 5120, "total_steps": 46238, "tokens": 2768781312, "loss": 3.2251778606212502, "ce_loss": 3.223893743572813, "z_loss": 12.841278834776444, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1709548383951187, "tokens_per_second": 4022004.6688788314, "eta_seconds": 5527.296076013017, "mean_data_wait_seconds": 6.252527236938477e-05}
{"step": 5130, "total_steps": 46238, "tokens": 2774188032, "loss": 3.2940181385387075, "ce_loss": 3.2927509076667554, "z_loss": 12.672463048588146, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16851651668548584, "tokens_per_second": 4022713.0553809474, "eta_seconds": 5524.978689263054, "mean_data_wait_seconds": 0.00010794401168823242}
{"step": 5140, "total_steps": 46238, "tokens": 2779594752, "loss": 3.304904383240324, "ce_loss": 3.30362019033143, "z_loss": 12.841931010737564, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18591228127479553, "tokens_per_second": 4023724.8809807682, "eta_seconds": 5522.245640856031, "mean_data_wait_seconds": 7.337331771850586e-05}
{"step": 5150, "total_steps": 46238, "tokens": 2785001472, "loss": 3.2630656885378286, "ce_loss": 3.261741928982012, "z_loss": 13.238095919291178, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17125459015369415, "tokens_per_second": 4024767.644544277, "eta_seconds": 5519.471538714219, "mean_data_wait_seconds": 7.221102714538574e-05}
{"step": 5160, "total_steps": 46238, "tokens": 2790408192, "loss": 3.2837583097544583, "ce_loss": 3.2824812495347224, "z_loss": 12.770810437924935, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16738350689411163, "tokens_per_second": 4025868.976617152, "eta_seconds": 5516.618616500998, "mean_data_wait_seconds": 0.00015163421630859375}
{"step": 5170, "total_steps": 46238, "tokens": 2795814912, "loss": 3.2830372543045967, "ce_loss": 3.2817433494510073, "z_loss": 12.9390869140625, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17009417712688446, "tokens_per_second": 4026663.261015483, "eta_seconds": 5514.187699519834, "mean_data_wait_seconds": 0.00011444091796875}
{"step": 5180, "total_steps": 46238, "tokens": 2801221632, "loss": 3.3047154690280105, "ce_loss": 3.3034271760420366, "z_loss": 12.883382927287709, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16964583098888397, "tokens_per_second": 4027267.8143754127, "eta_seconds": 5512.017409113562, "mean_data_wait_seconds": 9.256601333618164e-05}
{"step": 5190, "total_steps": 46238, "tokens": 2806628352, "loss": 3.295466986569491, "ce_loss": 3.2942213253541426, "z_loss": 12.456434546094952, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1464109718799591, "tokens_per_second": 4024782.785101501, "eta_seconds": 5514.077347515865, "mean_data_wait_seconds": 0.00010246038436889648}
{"step": 5200, "total_steps": 46238, "tokens": 2812035072, "loss": 3.2353642095219004, "ce_loss": 3.23415812217828, "z_loss": 12.060984669309674, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1842942237854004, "tokens_per_second": 4025123.290235369, "eta_seconds": 5512.267641049719, "mean_data_wait_seconds": 8.803606033325195e-05}
{"step": 5210, "total_steps": 46238, "tokens": 2817441792, "loss": 3.325692908330397, "ce_loss": 3.324360166535233, "z_loss": 13.327489838455662, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18912789225578308, "tokens_per_second": 4025501.278494677, "eta_seconds": 5510.40693056119, "mean_data_wait_seconds": 6.625056266784668e-05}
{"step": 5220, "total_steps": 46238, "tokens": 2822848512, "loss": 3.354134489189495, "ce_loss": 3.3528347141814954, "z_loss": 12.997784246097911, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1586463749408722, "tokens_per_second": 4025891.222582793, "eta_seconds": 5508.530210553629, "mean_data_wait_seconds": 0.00011703372001647949}
{"step": 5230, "total_steps": 46238, "tokens": 2828255232, "loss": 3.2576843283393164, "ce_loss": 3.256411951599699, "z_loss": 12.723644263816603, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16783827543258667, "tokens_per_second": 4024073.588082744, "eta_seconds": 5509.674765804533, "mean_data_wait_seconds": 0.0001749098300933838}
{"step": 5240, "total_steps": 46238, "tokens": 2833661952, "loss": 3.2398045352011016, "ce_loss": 3.2385169087034282, "z_loss": 12.87562936002558, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15363390743732452, "tokens_per_second": 4022064.5272663315, "eta_seconds": 5511.082637718265, "mean_data_wait_seconds": 9.313225746154785e-05}
{"step": 5250, "total_steps": 46238, "tokens": 2839068672, "loss": 3.2363704948714287, "ce_loss": 3.2351494684363855, "z_loss": 12.209934783704353, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16717693209648132, "tokens_per_second": 4022001.2292116717, "eta_seconds": 5509.82508484801, "mean_data_wait_seconds": 8.744001388549805e-05}
{"step": 5260, "total_steps": 46238, "tokens": 2844475392, "loss": 3.2721526171221877, "ce_loss": 3.270866238709652, "z_loss": 12.863633545962246, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15717191994190216, "tokens_per_second": 4022947.1126966826, "eta_seconds": 5507.185633655738, "mean_data_wait_seconds": 8.14199447631836e-05}
{"step": 5270, "total_steps": 46238, "tokens": 2849882112, "loss": 3.305752568172686, "ce_loss": 3.30443615805019, "z_loss": 13.164197705008768, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19331413507461548, "tokens_per_second": 4022854.087149483, "eta_seconds": 5505.968982259274, "mean_data_wait_seconds": 0.00010201334953308105}
{"step": 5280, "total_steps": 46238, "tokens": 2855288832, "loss": 3.1987498492905586, "ce_loss": 3.197519994143284, "z_loss": 12.29847594463464, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19072867929935455, "tokens_per_second": 4023316.331124936, "eta_seconds": 5503.992547811511, "mean_data_wait_seconds": 0.00010344386100769043}
{"step": 5290, "total_steps": 46238, "tokens": 2860695552, "loss": 3.3505954308943315, "ce_loss": 3.3493221040928, "z_loss": 12.7332134030082, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1773228645324707, "tokens_per_second": 4024218.977492871, "eta_seconds": 5501.414437887462, "mean_data_wait_seconds": 9.241700172424316e-05}
{"step": 5300, "total_steps": 46238, "tokens": 2866102272, "loss": 3.3019531044093045, "ce_loss": 3.3006765734065664, "z_loss": 12.764772928122317, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1695604920387268, "tokens_per_second": 4025700.0158876968, "eta_seconds": 5498.047439364257, "mean_data_wait_seconds": 8.478760719299316e-05}
{"step": 5310, "total_steps": 46238, "tokens": 2871508992, "loss": 3.262954168247454, "ce_loss": 3.261685544794256, "z_loss": 12.68637028607455, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15663908421993256, "tokens_per_second": 4026462.5994748916, "eta_seconds": 5495.663351470301, "mean_data_wait_seconds": 7.12275505065918e-05}
{"step": 5320, "total_steps": 46238, "tokens": 2876915712, "loss": 3.230080758080338, "ce_loss": 3.228776503692974, "z_loss": 13.042640534314243, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1909833550453186, "tokens_per_second": 4027386.9578356394, "eta_seconds": 5493.059508711564, "mean_data_wait_seconds": 8.338689804077148e-05}
{"step": 5330, "total_steps": 46238, "tokens": 2882322432, "loss": 3.255011181036631, "ce_loss": 3.2537275408253525, "z_loss": 12.836357521288322, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17386677861213684, "tokens_per_second": 4027959.2181255184, "eta_seconds": 5490.9368010663875, "mean_data_wait_seconds": 0.0001468658447265625}
{"step": 5340, "total_steps": 46238, "tokens": 2887729152, "loss": 3.2356671022646353, "ce_loss": 3.234414321003538, "z_loss": 12.527819380615696, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1437101662158966, "tokens_per_second": 4028740.988233743, "eta_seconds": 5488.529257298856, "mean_data_wait_seconds": 6.857514381408691e-05}
{"step": 5350, "total_steps": 46238, "tokens": 2893135872, "loss": 3.268469432989756, "ce_loss": 3.2671889861424765, "z_loss": 12.80448985822273, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14660698175430298, "tokens_per_second": 4029335.570991275, "eta_seconds": 5486.377511754746, "mean_data_wait_seconds": 6.124377250671387e-05}
{"step": 5360, "total_steps": 46238, "tokens": 2898542592, "loss": 3.2945270755074243, "ce_loss": 3.293246944745382, "z_loss": 12.80142743659742, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16481560468673706, "tokens_per_second": 4030652.698072451, "eta_seconds": 5483.2432857758295, "mean_data_wait_seconds": 6.210803985595703e-05}
{"step": 5370, "total_steps": 46238, "tokens": 2903949312, "loss": 3.2947398785388833, "ce_loss": 3.293473557992415, "z_loss": 12.663091507824985, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18015053868293762, "tokens_per_second": 4031917.2652600342, "eta_seconds": 5480.182546001465, "mean_data_wait_seconds": 6.660819053649902e-05}
{"step": 5380, "total_steps": 46238, "tokens": 2909356032, "loss": 3.2964437116276133, "ce_loss": 3.295141733053959, "z_loss": 13.01971756328236, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.13037119805812836, "tokens_per_second": 4033394.1470680465, "eta_seconds": 5476.835414177865, "mean_data_wait_seconds": 0.00011086463928222656}
{"step": 5390, "total_steps": 46238, "tokens": 2914762752, "loss": 3.223884347713355, "ce_loss": 3.2225835540077905, "z_loss": 13.007728670582626, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17787228524684906, "tokens_per_second": 4033584.2721121605, "eta_seconds": 5475.236835063179, "mean_data_wait_seconds": 9.572505950927734e-05}
{"step": 5400, "total_steps": 46238, "tokens": 2920169472, "loss": 3.2933367508830447, "ce_loss": 3.2920744328787834, "z_loss": 12.623126232262814, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14395537972450256, "tokens_per_second": 4033972.8719235067, "eta_seconds": 5473.369099151115, "mean_data_wait_seconds": 0.00022152066230773926}
{"step": 5410, "total_steps": 46238, "tokens": 2925576192, "loss": 3.2564581018505674, "ce_loss": 3.2551597158114114, "z_loss": 12.983792492837617, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1562603861093521, "tokens_per_second": 4034334.718551416, "eta_seconds": 5471.538006624791, "mean_data_wait_seconds": 0.00011393427848815918}
{"step": 5420, "total_steps": 46238, "tokens": 2930982912, "loss": 3.3076307863900154, "ce_loss": 3.3064099098696853, "z_loss": 12.208864891167844, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18148159980773926, "tokens_per_second": 4034312.0885132775, "eta_seconds": 5470.228514753481, "mean_data_wait_seconds": 8.061528205871582e-05}
{"step": 5430, "total_steps": 46238, "tokens": 2936389632, "loss": 3.2472874702829304, "ce_loss": 3.246064187902393, "z_loss": 12.232757250467936, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.153345987200737, "tokens_per_second": 4034960.243781, "eta_seconds": 5468.00983677734, "mean_data_wait_seconds": 7.492303848266602e-05}
{"step": 5440, "total_steps": 46238, "tokens": 2941796352, "loss": 3.298574787197691, "ce_loss": 3.297302802403768, "z_loss": 12.719534902861625, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16756664216518402, "tokens_per_second": 4035047.7022848283, "eta_seconds": 5466.551379680064, "mean_data_wait_seconds": 5.888938903808594e-05}
{"step": 5450, "total_steps": 46238, "tokens": 2947203072, "loss": 3.283881501718001, "ce_loss": 3.2826664249102273, "z_loss": 12.150502356615933, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17537043988704681, "tokens_per_second": 4035725.1518431446, "eta_seconds": 5464.294032493396, "mean_data_wait_seconds": 0.00012376904487609863}
{"step": 5460, "total_steps": 46238, "tokens": 2952609792, "loss": 3.3351911721807537, "ce_loss": 3.3339797204191033, "z_loss": 12.114355954256924, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17843666672706604, "tokens_per_second": 4036170.312002477, "eta_seconds": 5462.351793837403, "mean_data_wait_seconds": 0.00011596083641052246}
{"step": 5470, "total_steps": 46238, "tokens": 2958016512, "loss": 3.3384179003310925, "ce_loss": 3.337176588448611, "z_loss": 12.4134963425723, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15279893577098846, "tokens_per_second": 4036511.930837498, "eta_seconds": 5460.550049563906, "mean_data_wait_seconds": 0.00018480420112609863}
{"step": 5480, "total_steps": 46238, "tokens": 2963423232, "loss": 3.339845252759529, "ce_loss": 3.3386017806602246, "z_loss": 12.435182910976987, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1721372753381729, "tokens_per_second": 4037132.734181743, "eta_seconds": 5458.371115079611, "mean_data_wait_seconds": 9.557604789733887e-05}
{"step": 5490, "total_steps": 46238, "tokens": 2968829952, "loss": 3.2874654405044788, "ce_loss": 3.2862293774431404, "z_loss": 12.3604624488137, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20061418414115906, "tokens_per_second": 4038489.597603573, "eta_seconds": 5455.1983982013935, "mean_data_wait_seconds": 7.343292236328125e-05}
{"step": 5500, "total_steps": 46238, "tokens": 2974236672, "loss": 3.351124133124496, "ce_loss": 3.3499005978757683, "z_loss": 12.235388539054178, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15932677686214447, "tokens_per_second": 4039630.745361042, "eta_seconds": 5452.318950016181, "mean_data_wait_seconds": 7.542967796325684e-05}
{"step": 5510, "total_steps": 46238, "tokens": 2979643392, "loss": 3.2992649150617197, "ce_loss": 3.2979720505801113, "z_loss": 12.92868635148713, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16997729241847992, "tokens_per_second": 4040633.8078870475, "eta_seconds": 5449.627358217547, "mean_data_wait_seconds": 6.371736526489258e-05}
{"step": 5520, "total_steps": 46238, "tokens": 2985050112, "loss": 3.2670630848768987, "ce_loss": 3.2657792604330815, "z_loss": 12.837812394806832, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1879594475030899, "tokens_per_second": 4041621.5401857095, "eta_seconds": 5446.957763142872, "mean_data_wait_seconds": 9.822845458984375e-05}
{"step": 5530, "total_steps": 46238, "tokens": 2990456832, "loss": 3.2859501405195757, "ce_loss": 3.284658919681202, "z_loss": 12.911844202966401, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18248848617076874, "tokens_per_second": 4042991.942631281, "eta_seconds": 5443.774169303909, "mean_data_wait_seconds": 6.216764450073242e-05}
{"step": 5540, "total_steps": 46238, "tokens": 2995863552, "loss": 3.2998053287014817, "ce_loss": 3.2985039443680733, "z_loss": 13.013691851587007, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19441799819469452, "tokens_per_second": 4043774.5487405234, "eta_seconds": 5441.38356844184, "mean_data_wait_seconds": 6.464123725891113e-05}
{"step": 5550, "total_steps": 46238, "tokens": 3001270272, "loss": 3.2924880584081015, "ce_loss": 3.291257805896528, "z_loss": 12.302673621611161, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1739121973514557, "tokens_per_second": 4045105.0407528994, "eta_seconds": 5438.2572126991145, "mean_data_wait_seconds": 9.98079776763916e-05}
{"step": 5560, "total_steps": 46238, "tokens": 3006676992, "loss": 3.302974870710662, "ce_loss": 3.3017827651717444, "z_loss": 11.921238761959653, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18473872542381287, "tokens_per_second": 4046198.917908529, "eta_seconds": 5435.450750248357, "mean_data_wait_seconds": 7.191300392150879e-05}
{"step": 5570, "total_steps": 46238, "tokens": 3012083712, "loss": 3.292598870667544, "ce_loss": 3.2913849859526665, "z_loss": 12.138987049911961, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14215977489948273, "tokens_per_second": 4047146.2212574426, "eta_seconds": 5432.842556691345, "mean_data_wait_seconds": 0.00011685490608215332}
{"step": 5580, "total_steps": 46238, "tokens": 3017490432, "loss": 3.267200190009493, "ce_loss": 3.26595373767795, "z_loss": 12.464210438005852, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15516246855258942, "tokens_per_second": 4047979.2946893997, "eta_seconds": 5430.388819636164, "mean_data_wait_seconds": 0.0004528164863586426}
{"step": 5590, "total_steps": 46238, "tokens": 3022897152, "loss": 3.3127274693864766, "ce_loss": 3.3114420735474788, "z_loss": 12.853901162292019, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18538805842399597, "tokens_per_second": 4048946.3631357844, "eta_seconds": 5427.756461308054, "mean_data_wait_seconds": 7.221102714538574e-05}
{"step": 5600, "total_steps": 46238, "tokens": 3028303872, "loss": 3.3156204693245166, "ce_loss": 3.3142959746447476, "z_loss": 13.245005166891849, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22856643795967102, "tokens_per_second": 4050190.344333579, "eta_seconds": 5424.754442649577, "mean_data_wait_seconds": 7.408857345581055e-05}
{"step": 5610, "total_steps": 46238, "tokens": 3033710592, "loss": 3.287819428877397, "ce_loss": 3.28659256660577, "z_loss": 12.268504345055783, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16902898252010345, "tokens_per_second": 4051112.3126785704, "eta_seconds": 5422.185229289853, "mean_data_wait_seconds": 0.00012809038162231445}
{"step": 5620, "total_steps": 46238, "tokens": 3039117312, "loss": 3.3413693688132544, "ce_loss": 3.340098023414612, "z_loss": 12.713039513790246, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18225082755088806, "tokens_per_second": 4052140.356749541, "eta_seconds": 5419.475311959772, "mean_data_wait_seconds": 0.00013527274131774902}
{"step": 5630, "total_steps": 46238, "tokens": 3044524032, "loss": 3.293162268219572, "ce_loss": 3.2919297976927324, "z_loss": 12.324401515902895, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19014304876327515, "tokens_per_second": 4052620.778453437, "eta_seconds": 5417.498725942599, "mean_data_wait_seconds": 0.00012138485908508301}
{"step": 5640, "total_steps": 46238, "tokens": 3049930752, "loss": 3.1768425052816216, "ce_loss": 3.1755838538661147, "z_loss": 12.586917248639194, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16787920892238617, "tokens_per_second": 4053663.4497159175, "eta_seconds": 5414.771466915498, "mean_data_wait_seconds": 0.00016584992408752441}
{"step": 5650, "total_steps": 46238, "tokens": 3055337472, "loss": 3.310646962035786, "ce_loss": 3.3094262426549736, "z_loss": 12.207089149590695, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1737755686044693, "tokens_per_second": 4054934.142953853, "eta_seconds": 5411.74127380883, "mean_data_wait_seconds": 0.0001857280731201172}
{"step": 5660, "total_steps": 46238, "tokens": 3060744192, "loss": 3.3330292376604946, "ce_loss": 3.3317944154594885, "z_loss": 12.347931110497678, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2178938239812851, "tokens_per_second": 4055847.4005123903, "eta_seconds": 5409.18964091903, "mean_data_wait_seconds": 0.00017139315605163574}
{"step": 5670, "total_steps": 46238, "tokens": 3066150912, "loss": 3.315011629552552, "ce_loss": 3.313774526119232, "z_loss": 12.370922543785788, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18394432961940765, "tokens_per_second": 4055881.200206425, "eta_seconds": 5407.811506629852, "mean_data_wait_seconds": 0.00011560320854187012}
{"step": 5680, "total_steps": 46238, "tokens": 3071557632, "loss": 3.270482644890294, "ce_loss": 3.2692222252036585, "z_loss": 12.604104887355458, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15503694117069244, "tokens_per_second": 4056567.1957179913, "eta_seconds": 5405.564174345904, "mean_data_wait_seconds": 0.00014653801918029785}
{"step": 5690, "total_steps": 46238, "tokens": 3076964352, "loss": 3.2597087477192734, "ce_loss": 3.2584209604696794, "z_loss": 12.878249255093662, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1701619029045105, "tokens_per_second": 4057734.164195332, "eta_seconds": 5402.677133815482, "mean_data_wait_seconds": 0.0005337893962860107}
{"step": 5700, "total_steps": 46238, "tokens": 3082371072, "loss": 3.2163593769073486, "ce_loss": 3.215166799949877, "z_loss": 11.92620901627974, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16521093249320984, "tokens_per_second": 4057970.1309761857, "eta_seconds": 5401.030603132505, "mean_data_wait_seconds": 7.131695747375488e-05}
{"step": 5710, "total_steps": 46238, "tokens": 3087777792, "loss": 3.2951382470853403, "ce_loss": 3.2938796733364915, "z_loss": 12.585946465983536, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15068961679935455, "tokens_per_second": 4058744.5303803957, "eta_seconds": 5398.667982176836, "mean_data_wait_seconds": 8.183717727661133e-05}
{"step": 5720, "total_steps": 46238, "tokens": 3093184512, "loss": 3.2554993701703623, "ce_loss": 3.254261865760341, "z_loss": 12.375226230332345, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17237220704555511, "tokens_per_second": 4058554.1938359826, "eta_seconds": 5397.58898803688, "mean_data_wait_seconds": 7.048249244689941e-05}
{"step": 5730, "total_steps": 46238, "tokens": 3098591232, "loss": 3.3149500182180693, "ce_loss": 3.313696310375676, "z_loss": 12.537186709317295, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1771841198205948, "tokens_per_second": 4059275.977604046, "eta_seconds": 5395.297295584935, "mean_data_wait_seconds": 7.912516593933105e-05}
{"step": 5740, "total_steps": 46238, "tokens": 3103997952, "loss": 3.2882296695853723, "ce_loss": 3.2869765903010513, "z_loss": 12.530682282014327, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17730630934238434, "tokens_per_second": 4059861.0729900114, "eta_seconds": 5393.187991005393, "mean_data_wait_seconds": 9.655952453613281e-05}
{"step": 5750, "total_steps": 46238, "tokens": 3109404672, "loss": 3.3015658421949907, "ce_loss": 3.3003500067826472, "z_loss": 12.158126628760135, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1899886429309845, "tokens_per_second": 4060556.8699374828, "eta_seconds": 5390.932319176464, "mean_data_wait_seconds": 6.443262100219727e-05}
{"step": 5760, "total_steps": 46238, "tokens": 3114811392, "loss": 3.274249904083483, "ce_loss": 3.2730185985565186, "z_loss": 12.313149112643618, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.138099804520607, "tokens_per_second": 4061385.7462070365, "eta_seconds": 5388.500849602476, "mean_data_wait_seconds": 8.71121883392334e-05}
{"step": 5770, "total_steps": 46238, "tokens": 3120218112, "loss": 3.253359850608941, "ce_loss": 3.2521251151056, "z_loss": 12.347158735448664, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14923503994941711, "tokens_per_second": 4059767.011506284, "eta_seconds": 5389.317604184916, "mean_data_wait_seconds": 6.0230493545532227e-05}
{"step": 5780, "total_steps": 46238, "tokens": 3125624832, "loss": 3.3080132748141433, "ce_loss": 3.3067769570784136, "z_loss": 12.363207838752054, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18725822865962982, "tokens_per_second": 4060297.918351449, "eta_seconds": 5387.281313801034, "mean_data_wait_seconds": 0.0001672208309173584}
{"step": 5790, "total_steps": 46238, "tokens": 3131031552, "loss": 3.334389740770513, "ce_loss": 3.3331678997386587, "z_loss": 12.21824768817786, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1676424890756607, "tokens_per_second": 4060624.384017127, "eta_seconds": 5385.5166880433535, "mean_data_wait_seconds": 0.00014448165893554688}
{"step": 5800, "total_steps": 46238, "tokens": 3136438272, "loss": 3.2377161221070723, "ce_loss": 3.2364820910222605, "z_loss": 12.340312892740423, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1875617653131485, "tokens_per_second": 4060845.9045667313, "eta_seconds": 5383.891479214519, "mean_data_wait_seconds": 7.56382942199707e-05}
{"step": 5810, "total_steps": 46238, "tokens": 3141844992, "loss": 3.3200087962728557, "ce_loss": 3.318777451009461, "z_loss": 12.31365865649599, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1451372653245926, "tokens_per_second": 4061646.712229654, "eta_seconds": 5381.498808891017, "mean_data_wait_seconds": 8.228421211242676e-05}
{"step": 5820, "total_steps": 46238, "tokens": 3147251712, "loss": 3.279882990952694, "ce_loss": 3.2786372448458816, "z_loss": 12.457422697182858, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17640088498592377, "tokens_per_second": 4062503.0921364366, "eta_seconds": 5379.033499395575, "mean_data_wait_seconds": 7.191300392150879e-05}
{"step": 5830, "total_steps": 46238, "tokens": 3152658432, "loss": 3.3152284893122586, "ce_loss": 3.313967562083042, "z_loss": 12.609011339418815, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16354869306087494, "tokens_per_second": 4063119.638599098, "eta_seconds": 5376.886591390769, "mean_data_wait_seconds": 0.00010886788368225098}
{"step": 5840, "total_steps": 46238, "tokens": 3158065152, "loss": 3.249252619165363, "ce_loss": 3.248003990361185, "z_loss": 12.48619346185164, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14827673137187958, "tokens_per_second": 4063475.748870625, "eta_seconds": 5375.084812569753, "mean_data_wait_seconds": 0.00015938282012939453}
{"step": 5850, "total_steps": 46238, "tokens": 3163471872, "loss": 3.362396077676253, "ce_loss": 3.3611688631953616, "z_loss": 12.271945411508733, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19054721295833588, "tokens_per_second": 4063620.9626178243, "eta_seconds": 5373.562215786228, "mean_data_wait_seconds": 0.00012892484664916992}
{"step": 5860, "total_steps": 46238, "tokens": 3168878592, "loss": 3.3282581621950325, "ce_loss": 3.3270467953248457, "z_loss": 12.113247698003596, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1881798803806305, "tokens_per_second": 4064062.6336805364, "eta_seconds": 5371.647858741157, "mean_data_wait_seconds": 8.863210678100586e-05}
{"step": 5870, "total_steps": 46238, "tokens": 3174285312, "loss": 3.283824212623365, "ce_loss": 3.2826043728626138, "z_loss": 12.198165965802742, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16102203726768494, "tokens_per_second": 4064131.202151495, "eta_seconds": 5370.226879596305, "mean_data_wait_seconds": 0.00012427568435668945}
{"step": 5880, "total_steps": 46238, "tokens": 3179692032, "loss": 3.251146107008963, "ce_loss": 3.2499531904856362, "z_loss": 11.929439508553708, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18859727680683136, "tokens_per_second": 4064443.610870819, "eta_seconds": 5368.483854872579, "mean_data_wait_seconds": 0.00020816922187805176}
{"step": 5890, "total_steps": 46238, "tokens": 3185098752, "loss": 3.258202110276078, "ce_loss": 3.2569568662932427, "z_loss": 12.452896371032253, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19542816281318665, "tokens_per_second": 4064748.7264835425, "eta_seconds": 5366.750727263761, "mean_data_wait_seconds": 9.527802467346191e-05}
{"step": 5900, "total_steps": 46238, "tokens": 3190505472, "loss": 3.2361582337003765, "ce_loss": 3.234908934795495, "z_loss": 12.493358958851207, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16274122893810272, "tokens_per_second": 4065075.2291046004, "eta_seconds": 5364.98963361222, "mean_data_wait_seconds": 0.0001398324966430664}
{"step": 5910, "total_steps": 46238, "tokens": 3195912192, "loss": 3.215006705486413, "ce_loss": 3.213800533251329, "z_loss": 12.061572566176906, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16914807260036469, "tokens_per_second": 4065447.7778958473, "eta_seconds": 5363.168077708016, "mean_data_wait_seconds": 0.00012445449829101562}
{"step": 5920, "total_steps": 46238, "tokens": 3201318912, "loss": 3.2579635255264514, "ce_loss": 3.2567527330282964, "z_loss": 12.108068892450044, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.207213893532753, "tokens_per_second": 4066573.3268925007, "eta_seconds": 5360.35410448563, "mean_data_wait_seconds": 0.0003425776958465576}
{"step": 5930, "total_steps": 46238, "tokens": 3206725632, "loss": 3.264766722014456, "ce_loss": 3.2634943705616575, "z_loss": 12.72369376818339, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.184527188539505, "tokens_per_second": 4066647.92267039, "eta_seconds": 5358.92625041648, "mean_data_wait_seconds": 9.837746620178223e-05}
{"step": 5940, "total_steps": 46238, "tokens": 3212132352, "loss": 3.2645009936708393, "ce_loss": 3.263255574486472, "z_loss": 12.454709566000735, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17933163046836853, "tokens_per_second": 4067133.7205140837, "eta_seconds": 5356.956786079331, "mean_data_wait_seconds": 7.790327072143555e-05}
{"step": 5950, "total_steps": 46238, "tokens": 3217539072, "loss": 3.2970398646412473, "ce_loss": 3.295863749402942, "z_loss": 11.761153264479203, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16297949850559235, "tokens_per_second": 4068065.7061501183, "eta_seconds": 5354.4004540216265, "mean_data_wait_seconds": 0.00011855363845825195}
{"step": 5960, "total_steps": 46238, "tokens": 3222945792, "loss": 3.290781040986379, "ce_loss": 3.289539971134879, "z_loss": 12.41042330048301, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15864020586013794, "tokens_per_second": 4069069.3200716404, "eta_seconds": 5351.75108386126, "mean_data_wait_seconds": 0.00013580918312072754}
{"step": 5970, "total_steps": 46238, "tokens": 3228352512, "loss": 3.2260518543648, "ce_loss": 3.224833067619439, "z_loss": 12.18775581591057, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1614438146352768, "tokens_per_second": 4070081.2163743246, "eta_seconds": 5349.092135167286, "mean_data_wait_seconds": 6.943941116333008e-05}
{"step": 5980, "total_steps": 46238, "tokens": 3233759232, "loss": 3.254365066687266, "ce_loss": 3.2531199075958948, "z_loss": 12.45137256564516, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16035661101341248, "tokens_per_second": 4070495.5893967417, "eta_seconds": 5347.21933139983, "mean_data_wait_seconds": 6.80088996887207e-05}
{"step": 5990, "total_steps": 46238, "tokens": 3239165952, "loss": 3.3332105531837, "ce_loss": 3.3319396087617585, "z_loss": 12.709509863997951, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19506514072418213, "tokens_per_second": 4071542.3235758753, "eta_seconds": 5344.516710043352, "mean_data_wait_seconds": 0.0001385509967803955}
{"step": 6000, "total_steps": 46238, "tokens": 3244572672, "loss": 3.3011919783823416, "ce_loss": 3.2999346147884023, "z_loss": 12.573754606824933, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1697269231081009, "tokens_per_second": 4072416.0943694217, "eta_seconds": 5342.042355170629, "mean_data_wait_seconds": 9.41157341003418e-05}
{"step": 6010, "total_steps": 46238, "tokens": 3249979392, "loss": 3.2972392439842224, "ce_loss": 3.296000133861195, "z_loss": 12.391108744072191, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18297119438648224, "tokens_per_second": 4073235.7422638014, "eta_seconds": 5339.640011091555, "mean_data_wait_seconds": 9.021162986755371e-05}
{"step": 6020, "total_steps": 46238, "tokens": 3255386112, "loss": 3.281142007220875, "ce_loss": 3.2799145236159815, "z_loss": 12.274734179178873, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17049427330493927, "tokens_per_second": 4073960.829914956, "eta_seconds": 5337.362515695545, "mean_data_wait_seconds": 8.729100227355957e-05}
{"step": 6030, "total_steps": 46238, "tokens": 3260792832, "loss": 3.221083828897187, "ce_loss": 3.219872682383566, "z_loss": 12.111572634090077, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16177450120449066, "tokens_per_second": 4074963.958035607, "eta_seconds": 5334.7218105162065, "mean_data_wait_seconds": 7.221102714538574e-05}
{"step": 6040, "total_steps": 46238, "tokens": 3266199552, "loss": 3.29409226323619, "ce_loss": 3.2928712620879663, "z_loss": 12.210158109664917, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1641196310520172, "tokens_per_second": 4075880.0343528558, "eta_seconds": 5332.196286648241, "mean_data_wait_seconds": 0.0001506507396697998}
{"step": 6050, "total_steps": 46238, "tokens": 3271606272, "loss": 3.2234468153028777, "ce_loss": 3.2222318757664072, "z_loss": 12.14909542690624, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1736394315958023, "tokens_per_second": 4076533.4219341134, "eta_seconds": 5330.015337809041, "mean_data_wait_seconds": 6.836652755737305e-05}
{"step": 6060, "total_steps": 46238, "tokens": 3277012992, "loss": 3.2860090660326406, "ce_loss": 3.284811460610592, "z_loss": 11.975744760397708, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18369460105895996, "tokens_per_second": 4076660.9570471747, "eta_seconds": 5328.522330621822, "mean_data_wait_seconds": 6.103515625e-05}
{"step": 6070, "total_steps": 46238, "tokens": 3282419712, "loss": 3.2675856044798186, "ce_loss": 3.266334907575087, "z_loss": 12.506904912717415, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17648477852344513, "tokens_per_second": 4077626.127270831, "eta_seconds": 5325.935126508368, "mean_data_wait_seconds": 6.207823753356934e-05}
{"step": 6080, "total_steps": 46238, "tokens": 3287826432, "loss": 3.285594241185622, "ce_loss": 3.284333420522285, "z_loss": 12.608048395677047, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17020170390605927, "tokens_per_second": 4078189.1102101333, "eta_seconds": 5323.874130712216, "mean_data_wait_seconds": 7.62641429901123e-05}
{"step": 6090, "total_steps": 46238, "tokens": 3293233152, "loss": 3.3270168593435576, "ce_loss": 3.325750242580067, "z_loss": 12.665881272518273, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19281105697155, "tokens_per_second": 4078420.1931403647, "eta_seconds": 5322.246790683479, "mean_data_wait_seconds": 6.431341171264648e-05}
{"step": 6100, "total_steps": 46238, "tokens": 3298639872, "loss": 3.281232826637499, "ce_loss": 3.279981100197994, "z_loss": 12.517083601518111, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15868611633777618, "tokens_per_second": 4079203.6461440623, "eta_seconds": 5319.899163287227, "mean_data_wait_seconds": 9.605288505554199e-05}
{"step": 6110, "total_steps": 46238, "tokens": 3304046592, "loss": 3.2513813683480928, "ce_loss": 3.2501924922972014, "z_loss": 11.888690435525143, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19253365695476532, "tokens_per_second": 4079815.3036029395, "eta_seconds": 5317.776352483499, "mean_data_wait_seconds": 0.00010007619857788086}
{"step": 6120, "total_steps": 46238, "tokens": 3309453312, "loss": 3.2584334377086526, "ce_loss": 3.2572013500965005, "z_loss": 12.320773926648227, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17334051430225372, "tokens_per_second": 4080645.5874556364, "eta_seconds": 5315.369384363574, "mean_data_wait_seconds": 0.0001010596752166748}
{"step": 6130, "total_steps": 46238, "tokens": 3314860032, "loss": 3.2902291489369944, "ce_loss": 3.288974966063644, "z_loss": 12.54182878407565, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15880392491817474, "tokens_per_second": 4080989.346073217, "eta_seconds": 5313.5967935974495, "mean_data_wait_seconds": 8.741021156311035e-05}
{"step": 6140, "total_steps": 46238, "tokens": 3320266752, "loss": 3.293205609827331, "ce_loss": 3.2919850475860364, "z_loss": 12.20593961802396, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15271060168743134, "tokens_per_second": 4081309.3467242583, "eta_seconds": 5311.855422427184, "mean_data_wait_seconds": 0.00011992454528808594}
{"step": 6150, "total_steps": 46238, "tokens": 3325673472, "loss": 3.3003338250246914, "ce_loss": 3.2990967775836135, "z_loss": 12.370391318292329, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15106429159641266, "tokens_per_second": 4081706.4011780047, "eta_seconds": 5310.014080813058, "mean_data_wait_seconds": 8.779764175415039e-05}
{"step": 6160, "total_steps": 46238, "tokens": 3331080192, "loss": 3.3122572754368638, "ce_loss": 3.311010344461961, "z_loss": 12.469084053328544, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1682126671075821, "tokens_per_second": 4082298.616534378, "eta_seconds": 5307.919331583646, "mean_data_wait_seconds": 0.00013512372970581055}
{"step": 6170, "total_steps": 46238, "tokens": 3336486912, "loss": 3.3011276198155954, "ce_loss": 3.2998841097860625, "z_loss": 12.435158967971802, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17748922109603882, "tokens_per_second": 4082527.928439557, "eta_seconds": 5306.296834637987, "mean_data_wait_seconds": 0.00012132525444030762}
{"step": 6180, "total_steps": 46238, "tokens": 3341893632, "loss": 3.2752903013518364, "ce_loss": 3.274033165339268, "z_loss": 12.571458346915968, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16116641461849213, "tokens_per_second": 4082624.6841665935, "eta_seconds": 5304.846754097626, "mean_data_wait_seconds": 8.442997932434082e-05}
{"step": 6190, "total_steps": 46238, "tokens": 3347300352, "loss": 3.3479044925082815, "ce_loss": 3.3466612750833686, "z_loss": 12.43222915765011, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17022103071212769, "tokens_per_second": 4082430.8764285166, "eta_seconds": 5303.77420693583, "mean_data_wait_seconds": 9.891390800476074e-05}
{"step": 6200, "total_steps": 46238, "tokens": 3352707072, "loss": 3.2469341700727288, "ce_loss": 3.245698166616035, "z_loss": 12.36032324848753, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15849919617176056, "tokens_per_second": 4082428.492280132, "eta_seconds": 5302.452916183158, "mean_data_wait_seconds": 0.00010439753532409668}
{"step": 6210, "total_steps": 46238, "tokens": 3358113792, "loss": 3.347080516092705, "ce_loss": 3.3458198579874905, "z_loss": 12.60666907194889, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20252083241939545, "tokens_per_second": 4082287.904701115, "eta_seconds": 5301.311090547515, "mean_data_wait_seconds": 7.703900337219238e-05}
{"step": 6220, "total_steps": 46238, "tokens": 3363520512, "loss": 3.289556561094342, "ce_loss": 3.288297550244765, "z_loss": 12.590478557528872, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1674225926399231, "tokens_per_second": 4082663.7826891015, "eta_seconds": 5299.4987036990615, "mean_data_wait_seconds": 0.00010213255882263184}
{"step": 6230, "total_steps": 46238, "tokens": 3368927232, "loss": 3.2645694812138877, "ce_loss": 3.263349746212815, "z_loss": 12.197018442731915, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.182524174451828, "tokens_per_second": 4082781.895534463, "eta_seconds": 5298.021118311147, "mean_data_wait_seconds": 7.811188697814941e-05}
{"step": 6240, "total_steps": 46238, "tokens": 3374333952, "loss": 3.3016842693993538, "ce_loss": 3.3004364624167932, "z_loss": 12.478403185353134, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14524556696414948, "tokens_per_second": 4083099.3435933497, "eta_seconds": 5296.285043353512, "mean_data_wait_seconds": 7.167458534240723e-05}
{"step": 6250, "total_steps": 46238, "tokens": 3379740672, "loss": 3.2021537495381907, "ce_loss": 3.2009345672347327, "z_loss": 12.192030711607499, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21240299940109253, "tokens_per_second": 4083714.584490002, "eta_seconds": 5294.16314894103, "mean_data_wait_seconds": 8.034706115722656e-05}
{"step": 6260, "total_steps": 46238, "tokens": 3385147392, "loss": 3.2942278276790273, "ce_loss": 3.2930174668629966, "z_loss": 12.103704922127001, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18221424520015717, "tokens_per_second": 4084281.603400855, "eta_seconds": 5292.104375467725, "mean_data_wait_seconds": 9.706616401672363e-05}
{"step": 6270, "total_steps": 46238, "tokens": 3390554112, "loss": 3.289186943661083, "ce_loss": 3.287976947697726, "z_loss": 12.10029194571755, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15341468155384064, "tokens_per_second": 4085196.197086783, "eta_seconds": 5289.596088288181, "mean_data_wait_seconds": 0.00012561678886413574}
{"step": 6280, "total_steps": 46238, "tokens": 3395960832, "loss": 3.2762417070793384, "ce_loss": 3.2749987985148574, "z_loss": 12.428794232281772, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15388600528240204, "tokens_per_second": 4085738.196827107, "eta_seconds": 5287.571073637781, "mean_data_wait_seconds": 6.821751594543457e-05}
{"step": 6290, "total_steps": 46238, "tokens": 3401367552, "loss": 3.228793030435389, "ce_loss": 3.2276001355864783, "z_loss": 11.92903420419404, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17770008742809296, "tokens_per_second": 4086217.6385473814, "eta_seconds": 5285.627515346442, "mean_data_wait_seconds": 7.75456428527832e-05}
{"step": 6300, "total_steps": 46238, "tokens": 3406774272, "loss": 3.1864047249158225, "ce_loss": 3.1852516008145884, "z_loss": 11.53125685634035, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15451279282569885, "tokens_per_second": 4086830.4285537233, "eta_seconds": 5283.512012912515, "mean_data_wait_seconds": 6.628036499023438e-05}
{"step": 6310, "total_steps": 46238, "tokens": 3412180992, "loss": 3.287701008897839, "ce_loss": 3.286480876532468, "z_loss": 12.201216827739369, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17016978561878204, "tokens_per_second": 4087403.231797866, "eta_seconds": 5281.448812307334, "mean_data_wait_seconds": 5.799531936645508e-05}
{"step": 6320, "total_steps": 46238, "tokens": 3417587712, "loss": 3.2559143557693018, "ce_loss": 3.2547041231935676, "z_loss": 12.102501601883858, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1688738912343979, "tokens_per_second": 4088670.3185384735, "eta_seconds": 5278.489714894561, "mean_data_wait_seconds": 5.942583084106445e-05}
{"step": 6330, "total_steps": 46238, "tokens": 3422994432, "loss": 3.2530215472886055, "ce_loss": 3.2517918528932515, "z_loss": 12.29691048824426, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15926721692085266, "tokens_per_second": 4089885.656046986, "eta_seconds": 5275.599202168042, "mean_data_wait_seconds": 7.069110870361328e-05}
{"step": 6340, "total_steps": 46238, "tokens": 3428401152, "loss": 3.2389757886077417, "ce_loss": 3.237739709290591, "z_loss": 12.360564730384134, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1700906902551651, "tokens_per_second": 4090823.9114644607, "eta_seconds": 5273.067541124692, "mean_data_wait_seconds": 0.0001163482666015625}
{"step": 6350, "total_steps": 46238, "tokens": 3433807872, "loss": 3.2915226492014797, "ce_loss": 3.290273716955474, "z_loss": 12.489202159823794, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1763523519039154, "tokens_per_second": 4091618.028081896, "eta_seconds": 5270.722710670477, "mean_data_wait_seconds": 0.00010010600090026855}
{"step": 6360, "total_steps": 46238, "tokens": 3439214592, "loss": 3.3703949902996873, "ce_loss": 3.369147761301561, "z_loss": 12.472468260562781, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20716775953769684, "tokens_per_second": 4092309.4258380607, "eta_seconds": 5268.511028973492, "mean_data_wait_seconds": 0.0001340806484222412}
{"step": 6370, "total_steps": 46238, "tokens": 3444621312, "loss": 3.274446355574059, "ce_loss": 3.2731993902813303, "z_loss": 12.469445842685122, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17869989573955536, "tokens_per_second": 4093154.388059237, "eta_seconds": 5266.102516651041, "mean_data_wait_seconds": 9.340047836303711e-05}
{"step": 6380, "total_steps": 46238, "tokens": 3450028032, "loss": 3.261438949541612, "ce_loss": 3.2602341518257605, "z_loss": 12.048388878504435, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1678398996591568, "tokens_per_second": 4093518.0719657834, "eta_seconds": 5264.313855502658, "mean_data_wait_seconds": 7.918477058410645e-05}
{"step": 6390, "total_steps": 46238, "tokens": 3455434752, "loss": 3.278416684179595, "ce_loss": 3.2771948738531633, "z_loss": 12.218197771997163, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1740543097257614, "tokens_per_second": 4094579.6608247748, "eta_seconds": 5261.628535433193, "mean_data_wait_seconds": 7.042288780212402e-05}
{"step": 6400, "total_steps": 46238, "tokens": 3460841472, "loss": 3.2561128320115986, "ce_loss": 3.2549390919280774, "z_loss": 11.737409887891827, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19097980856895447, "tokens_per_second": 4095198.3130216612, "eta_seconds": 5259.513414896758, "mean_data_wait_seconds": 8.764863014221191e-05}
{"step": 6410, "total_steps": 46238, "tokens": 3466248192, "loss": 3.2274393656037073, "ce_loss": 3.226246167312969, "z_loss": 11.931648630084414, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14125052094459534, "tokens_per_second": 4095521.311936904, "eta_seconds": 5257.778461862328, "mean_data_wait_seconds": 7.125735282897949e-05}
{"step": 6420, "total_steps": 46238, "tokens": 3471654912, "loss": 3.285628622228449, "ce_loss": 3.28438329335415, "z_loss": 12.453093499848336, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16357645392417908, "tokens_per_second": 4096092.972051157, "eta_seconds": 5255.724704222152, "mean_data_wait_seconds": 6.124377250671387e-05}
{"step": 6430, "total_steps": 46238, "tokens": 3477061632, "loss": 3.2467222159559075, "ce_loss": 3.245466573671861, "z_loss": 12.556316209561897, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15743331611156464, "tokens_per_second": 4096244.321047944, "eta_seconds": 5254.210593203552, "mean_data_wait_seconds": 8.20159912109375e-05}
{"step": 6440, "total_steps": 46238, "tokens": 3482468352, "loss": 3.199234855897499, "ce_loss": 3.1980299895459954, "z_loss": 12.04854077281374, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17767752707004547, "tokens_per_second": 4096409.851993159, "eta_seconds": 5252.6784090050405, "mean_data_wait_seconds": 7.152557373046875e-05}
{"step": 6450, "total_steps": 46238, "tokens": 3487875072, "loss": 3.3301958188866125, "ce_loss": 3.328945082245451, "z_loss": 12.507321119308472, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1759834587574005, "tokens_per_second": 4096534.220366772, "eta_seconds": 5251.19911291111, "mean_data_wait_seconds": 6.04093074798584e-05}
{"step": 6460, "total_steps": 46238, "tokens": 3493281792, "loss": 3.217277640646154, "ce_loss": 3.216049400242892, "z_loss": 12.282738439964525, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17378397285938263, "tokens_per_second": 4097394.1002441505, "eta_seconds": 5248.777544419881, "mean_data_wait_seconds": 6.091594696044922e-05}
{"step": 6470, "total_steps": 46238, "tokens": 3498688512, "loss": 3.2826560880198623, "ce_loss": 3.2814302137403777, "z_loss": 12.258824933658946, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14629428088665009, "tokens_per_second": 4097556.6250254614, "eta_seconds": 5247.249859266166, "mean_data_wait_seconds": 7.003545761108398e-05}
{"step": 6480, "total_steps": 46238, "tokens": 3504095232, "loss": 3.282959009661819, "ce_loss": 3.2816925355882356, "z_loss": 12.664670496275932, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1623067408800125, "tokens_per_second": 4098389.0967094344, "eta_seconds": 5244.864798527444, "mean_data_wait_seconds": 6.711483001708984e-05}
{"step": 6490, "total_steps": 46238, "tokens": 3509501952, "loss": 3.2314044291322883, "ce_loss": 3.230206881508683, "z_loss": 11.975436449050903, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.181245356798172, "tokens_per_second": 4099454.1772193154, "eta_seconds": 5242.183240739834, "mean_data_wait_seconds": 6.595253944396973e-05}
{"step": 6500, "total_steps": 46238, "tokens": 3514908672, "loss": 3.3102748412074465, "ce_loss": 3.3091003768371814, "z_loss": 11.744805032556707, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14540746808052063, "tokens_per_second": 4100252.030605003, "eta_seconds": 5239.844551904259, "mean_data_wait_seconds": 6.389617919921875e-05}
{"step": 6510, "total_steps": 46238, "tokens": 3520315392, "loss": 3.2860915931788357, "ce_loss": 3.284861127535502, "z_loss": 12.304499712857334, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19952188432216644, "tokens_per_second": 4101048.4087511147, "eta_seconds": 5237.508657095088, "mean_data_wait_seconds": 6.401538848876953e-05}
{"step": 6520, "total_steps": 46238, "tokens": 3525722112, "loss": 3.2775908293146077, "ce_loss": 3.2763475595098552, "z_loss": 12.432840296716401, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1694699376821518, "tokens_per_second": 4102031.890801445, "eta_seconds": 5234.93487999297, "mean_data_wait_seconds": 9.360909461975098e-05}
{"step": 6530, "total_steps": 46238, "tokens": 3531128832, "loss": 3.335434743852326, "ce_loss": 3.3342352455312554, "z_loss": 11.995071721799446, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16738612949848175, "tokens_per_second": 4102283.96846828, "eta_seconds": 5233.295225053848, "mean_data_wait_seconds": 0.00012692809104919434}
{"step": 6540, "total_steps": 46238, "tokens": 3536535552, "loss": 3.3204565210775896, "ce_loss": 3.3191979148171167, "z_loss": 12.586032809633197, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17658716440200806, "tokens_per_second": 4103124.244302195, "eta_seconds": 5230.905794238301, "mean_data_wait_seconds": 8.58008861541748e-05}
{"step": 6550, "total_steps": 46238, "tokens": 3541942272, "loss": 3.281830653999791, "ce_loss": 3.28059281724872, "z_loss": 12.378673315048218, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21835295855998993, "tokens_per_second": 4099606.087309941, "eta_seconds": 5234.0759592539225, "mean_data_wait_seconds": 0.00014340877532958984}
{"step": 6560, "total_steps": 46238, "tokens": 3547348992, "loss": 3.273836475430113, "ce_loss": 3.272665287509109, "z_loss": 11.711660905317826, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18266764283180237, "tokens_per_second": 4099881.530430302, "eta_seconds": 5232.405567033174, "mean_data_wait_seconds": 9.72747802734375e-05}
{"step": 6570, "total_steps": 46238, "tokens": 3552755712, "loss": 3.2771043506535618, "ce_loss": 3.2758501695864126, "z_loss": 12.541628129554518, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18960891664028168, "tokens_per_second": 4098134.3177125254, "eta_seconds": 5233.317056326031, "mean_data_wait_seconds": 0.00014540553092956543}
{"step": 6580, "total_steps": 46238, "tokens": 3558162432, "loss": 3.282989364681822, "ce_loss": 3.281759359619834, "z_loss": 12.300156578873143, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1747254878282547, "tokens_per_second": 4098379.3609075746, "eta_seconds": 5231.684921244541, "mean_data_wait_seconds": 0.00010612607002258301}
{"step": 6590, "total_steps": 46238, "tokens": 3563569152, "loss": 3.314456748239922, "ce_loss": 3.313280515598528, "z_loss": 11.76244466232531, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17939946055412292, "tokens_per_second": 4098307.6305199834, "eta_seconds": 5230.4572317525735, "mean_data_wait_seconds": 8.320808410644531e-05}
{"step": 6600, "total_steps": 46238, "tokens": 3568975872, "loss": 3.302395284175873, "ce_loss": 3.301176342097196, "z_loss": 12.189263661702475, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18241485953330994, "tokens_per_second": 4098440.0031020907, "eta_seconds": 5228.969082816697, "mean_data_wait_seconds": 0.000284045934677124}
{"step": 6610, "total_steps": 46238, "tokens": 3574382592, "loss": 3.2407186717698067, "ce_loss": 3.2394804990652837, "z_loss": 12.381586103728324, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18344683945178986, "tokens_per_second": 4099103.3372716494, "eta_seconds": 5226.8039083543945, "mean_data_wait_seconds": 0.00011214613914489746}
{"step": 6620, "total_steps": 46238, "tokens": 3579789312, "loss": 3.2873345468983506, "ce_loss": 3.286110312649698, "z_loss": 12.242310133847324, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1819113940000534, "tokens_per_second": 4099427.2673938647, "eta_seconds": 5225.071998317766, "mean_data_wait_seconds": 8.669495582580566e-05}
{"step": 6630, "total_steps": 46238, "tokens": 3585196032, "loss": 3.3574999751466694, "ce_loss": 3.356306047150583, "z_loss": 11.939307501821807, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19601857662200928, "tokens_per_second": 4099405.1045783325, "eta_seconds": 5223.781343318276, "mean_data_wait_seconds": 8.463859558105469e-05}
{"step": 6640, "total_steps": 46238, "tokens": 3590602752, "loss": 3.291652209831007, "ce_loss": 3.2904545213236953, "z_loss": 11.976411617163455, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16599740087985992, "tokens_per_second": 4099729.5561043588, "eta_seconds": 5222.049135441809, "mean_data_wait_seconds": 0.00013875961303710938}
{"step": 6650, "total_steps": 46238, "tokens": 3596009472, "loss": 3.291850203817541, "ce_loss": 3.290671512936101, "z_loss": 11.787001393058084, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17737402021884918, "tokens_per_second": 4099829.278771995, "eta_seconds": 5220.603349223099, "mean_data_wait_seconds": 0.00016412138938903809}
{"step": 6660, "total_steps": 46238, "tokens": 3601416192, "loss": 3.277919296062354, "ce_loss": 3.2767741933013452, "z_loss": 11.450692798152115, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21596291661262512, "tokens_per_second": 4100401.457899798, "eta_seconds": 5218.556271551035, "mean_data_wait_seconds": 0.0001322031021118164}
{"step": 6670, "total_steps": 46238, "tokens": 3606822912, "loss": 3.2370907266934714, "ce_loss": 3.235949301358425, "z_loss": 11.414434237913651, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17795559763908386, "tokens_per_second": 4100396.075763264, "eta_seconds": 5217.24453655806, "mean_data_wait_seconds": 7.429718971252441e-05}
{"step": 6680, "total_steps": 46238, "tokens": 3612229632, "loss": 3.3460416902195322, "ce_loss": 3.344820559024811, "z_loss": 12.211275794289328, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20198482275009155, "tokens_per_second": 4100873.396906504, "eta_seconds": 5215.318844062235, "mean_data_wait_seconds": 6.192922592163086e-05}
{"step": 6690, "total_steps": 46238, "tokens": 3617636352, "loss": 3.2606789180726716, "ce_loss": 3.2594649358229204, "z_loss": 12.139435666980166, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1732555329799652, "tokens_per_second": 4101892.553788779, "eta_seconds": 5212.704941344749, "mean_data_wait_seconds": 6.765127182006836e-05}
{"step": 6700, "total_steps": 46238, "tokens": 3623043072, "loss": 3.29327420574246, "ce_loss": 3.2920403516653813, "z_loss": 12.338785005338265, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20237864553928375, "tokens_per_second": 4102705.3744042655, "eta_seconds": 5210.354366989852, "mean_data_wait_seconds": 7.18235969543457e-05}
{"step": 6710, "total_steps": 46238, "tokens": 3628449792, "loss": 3.2873159339933684, "ce_loss": 3.286116777044354, "z_loss": 11.991995233477969, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18181046843528748, "tokens_per_second": 4101850.2679805923, "eta_seconds": 5210.122444211344, "mean_data_wait_seconds": 9.307265281677246e-05}
{"step": 6720, "total_steps": 46238, "tokens": 3633856512, "loss": 3.3038815277995486, "ce_loss": 3.3026666207747026, "z_loss": 12.149334885857321, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17554378509521484, "tokens_per_second": 4101908.1773312483, "eta_seconds": 5208.730790726966, "mean_data_wait_seconds": 9.870529174804688e-05}
{"step": 6730, "total_steps": 46238, "tokens": 3639263232, "loss": 3.2535055008801548, "ce_loss": 3.2522722753611477, "z_loss": 12.332141760623816, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15187114477157593, "tokens_per_second": 4102069.0107242647, "eta_seconds": 5207.208520421406, "mean_data_wait_seconds": 7.295608520507812e-05}
{"step": 6740, "total_steps": 46238, "tokens": 3644669952, "loss": 3.3176819555687183, "ce_loss": 3.3164345531752617, "z_loss": 12.473467949664954, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18422602117061615, "tokens_per_second": 4102919.286791665, "eta_seconds": 5204.811621019915, "mean_data_wait_seconds": 0.00010380148887634277}
{"step": 6750, "total_steps": 46238, "tokens": 3650076672, "loss": 3.287017255118399, "ce_loss": 3.285793788505323, "z_loss": 12.235194762547811, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14741671085357666, "tokens_per_second": 4103520.464556963, "eta_seconds": 5202.7315200205785, "mean_data_wait_seconds": 6.571412086486816e-05}
{"step": 6760, "total_steps": 46238, "tokens": 3655483392, "loss": 3.2755755384763083, "ce_loss": 3.2743626074357466, "z_loss": 12.129150014935117, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17193393409252167, "tokens_per_second": 4103934.040143832, "eta_seconds": 5200.889764605463, "mean_data_wait_seconds": 8.559226989746094e-05}
{"step": 6770, "total_steps": 46238, "tokens": 3660890112, "loss": 3.237028192390095, "ce_loss": 3.235810955365499, "z_loss": 12.17205236897324, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19016633927822113, "tokens_per_second": 4104217.2157160183, "eta_seconds": 5199.213565570814, "mean_data_wait_seconds": 6.321072578430176e-05}
{"step": 6780, "total_steps": 46238, "tokens": 3666296832, "loss": 3.258132166934736, "ce_loss": 3.2569269266995517, "z_loss": 12.052851568568837, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16132350265979767, "tokens_per_second": 4105286.3245457876, "eta_seconds": 5196.5425593939135, "mean_data_wait_seconds": 0.00012236833572387695}
{"step": 6790, "total_steps": 46238, "tokens": 3671703552, "loss": 3.221734339540655, "ce_loss": 3.220553042310657, "z_loss": 11.812918504079184, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15623024106025696, "tokens_per_second": 4106161.4759909753, "eta_seconds": 5194.118280224904, "mean_data_wait_seconds": 6.440281867980957e-05}
{"step": 6800, "total_steps": 46238, "tokens": 3677110272, "loss": 3.271984779473507, "ce_loss": 3.2707753091147453, "z_loss": 12.0948986862645, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15517480671405792, "tokens_per_second": 4106275.434993126, "eta_seconds": 5192.657434105049, "mean_data_wait_seconds": 0.00011214613914489746}
{"step": 6810, "total_steps": 46238, "tokens": 3682516992, "loss": 3.29585582559759, "ce_loss": 3.2946822011109553, "z_loss": 11.736440954786358, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18372640013694763, "tokens_per_second": 4107195.39175634, "eta_seconds": 5190.177946436652, "mean_data_wait_seconds": 7.018446922302246e-05}
{"step": 6820, "total_steps": 46238, "tokens": 3687923712, "loss": 3.265844444433848, "ce_loss": 3.2646283507347107, "z_loss": 12.161505858103434, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20313185453414917, "tokens_per_second": 4107984.460779603, "eta_seconds": 5187.864858660036, "mean_data_wait_seconds": 0.00018045306205749512}
{"step": 6830, "total_steps": 46238, "tokens": 3693330432, "loss": 3.2092990893306155, "ce_loss": 3.2080737099503027, "z_loss": 12.253708658796368, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18668034672737122, "tokens_per_second": 4108196.487332325, "eta_seconds": 5186.281028596885, "mean_data_wait_seconds": 6.759166717529297e-05}
{"step": 6840, "total_steps": 46238, "tokens": 3698737152, "loss": 3.21686234257438, "ce_loss": 3.2156579946026658, "z_loss": 12.043724088957816, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17517313361167908, "tokens_per_second": 4109147.539232892, "eta_seconds": 5183.764900292801, "mean_data_wait_seconds": 6.99162483215332e-05}
{"step": 6850, "total_steps": 46238, "tokens": 3704143872, "loss": 3.2518868482474126, "ce_loss": 3.2506871819496155, "z_loss": 11.996907133044619, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18165434896945953, "tokens_per_second": 4109658.4307662747, "eta_seconds": 5181.804868398592, "mean_data_wait_seconds": 6.967782974243164e-05}
{"step": 6860, "total_steps": 46238, "tokens": 3709550592, "loss": 3.342182726571054, "ce_loss": 3.341000051209421, "z_loss": 11.826778556361344, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18708853423595428, "tokens_per_second": 4110639.4572298103, "eta_seconds": 5179.252903475877, "mean_data_wait_seconds": 9.194016456604004e-05}
{"step": 6870, "total_steps": 46238, "tokens": 3714957312, "loss": 3.227655963464217, "ce_loss": 3.2265053644324793, "z_loss": 11.505665013284395, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1811959594488144, "tokens_per_second": 4111305.8790245294, "eta_seconds": 5177.0982870897715, "mean_data_wait_seconds": 0.0001423358917236328}
{"step": 6880, "total_steps": 46238, "tokens": 3720364032, "loss": 3.2525192336602644, "ce_loss": 3.2512916561329, "z_loss": 12.276106111931078, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1909390091896057, "tokens_per_second": 4111741.8123169136, "eta_seconds": 5175.234456661915, "mean_data_wait_seconds": 7.340312004089355e-05}
{"step": 6890, "total_steps": 46238, "tokens": 3725770752, "loss": 3.257754497455828, "ce_loss": 3.256545760414817, "z_loss": 12.08725576689749, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19304364919662476, "tokens_per_second": 4112512.363643533, "eta_seconds": 5172.950085711642, "mean_data_wait_seconds": 8.13603401184082e-05}
{"step": 6900, "total_steps": 46238, "tokens": 3731177472, "loss": 3.2534604253190937, "ce_loss": 3.252281246763287, "z_loss": 11.792130983237064, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1490117460489273, "tokens_per_second": 4112951.2575777057, "eta_seconds": 5171.083519362173, "mean_data_wait_seconds": 8.377432823181152e-05}
{"step": 6910, "total_steps": 46238, "tokens": 3736584192, "loss": 3.260276835976225, "ce_loss": 3.2590790390968323, "z_loss": 11.978037118911743, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.13532321155071259, "tokens_per_second": 4113446.7494558124, "eta_seconds": 5169.146226777577, "mean_data_wait_seconds": 6.383657455444336e-05}
{"step": 6920, "total_steps": 46238, "tokens": 3741990912, "loss": 3.2089455958568687, "ce_loss": 3.207762969262672, "z_loss": 11.826137347654862, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16206464171409607, "tokens_per_second": 4113819.218203302, "eta_seconds": 5167.363925458103, "mean_data_wait_seconds": 8.20159912109375e-05}
{"step": 6930, "total_steps": 46238, "tokens": 3747397632, "loss": 3.2416082512248647, "ce_loss": 3.240369090528199, "z_loss": 12.391891074903084, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1835915744304657, "tokens_per_second": 4112087.538722544, "eta_seconds": 5168.225166384026, "mean_data_wait_seconds": 0.14514243602752686}
{"step": 6940, "total_steps": 46238, "tokens": 3752804352, "loss": 3.2317462635762766, "ce_loss": 3.2305578643625434, "z_loss": 11.884015740770282, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1820816993713379, "tokens_per_second": 4112593.2191738393, "eta_seconds": 5166.275012306754, "mean_data_wait_seconds": 7.387995719909668e-05}
{"step": 6950, "total_steps": 46238, "tokens": 3758211072, "loss": 3.2469596194498465, "ce_loss": 3.2457887999939197, "z_loss": 11.708293387384126, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15595796704292297, "tokens_per_second": 4113131.5582886594, "eta_seconds": 5164.284332504514, "mean_data_wait_seconds": 0.0001131594181060791}
{"step": 6960, "total_steps": 46238, "tokens": 3763617792, "loss": 3.2639719934174507, "ce_loss": 3.262782658591415, "z_loss": 11.893171418796886, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14047253131866455, "tokens_per_second": 4113475.5472260346, "eta_seconds": 5162.538077641108, "mean_data_wait_seconds": 9.548664093017578e-05}
{"step": 6970, "total_steps": 46238, "tokens": 3769024512, "loss": 3.283348869193684, "ce_loss": 3.2821630058866558, "z_loss": 11.85883819695675, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18133966624736786, "tokens_per_second": 4111532.6295569693, "eta_seconds": 5163.662637960789, "mean_data_wait_seconds": 0.00013110041618347168}
{"step": 6980, "total_steps": 46238, "tokens": 3774431232, "loss": 3.238763478669253, "ce_loss": 3.2375763763080943, "z_loss": 11.871017506628325, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16688333451747894, "tokens_per_second": 4111862.7149828034, "eta_seconds": 5161.933210138503, "mean_data_wait_seconds": 0.00012060999870300293}
{"step": 6990, "total_steps": 46238, "tokens": 3779837952, "loss": 3.3192455967267356, "ce_loss": 3.3180699222015613, "z_loss": 11.756757512237087, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1581389158964157, "tokens_per_second": 4109586.7408883534, "eta_seconds": 5163.476359526361, "mean_data_wait_seconds": 8.571147918701172e-05}
{"step": 7000, "total_steps": 46238, "tokens": 3785244672, "loss": 3.2557417768420596, "ce_loss": 3.2545674407120906, "z_loss": 11.743259075916175, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19293749332427979, "tokens_per_second": 4109853.4793932154, "eta_seconds": 5161.825688036965, "mean_data_wait_seconds": 0.0001245737075805664}
{"step": 7010, "total_steps": 46238, "tokens": 3790651392, "loss": 3.282978352272149, "ce_loss": 3.281782222516609, "z_loss": 11.961533199657094, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20094403624534607, "tokens_per_second": 4109366.2733199606, "eta_seconds": 5161.121967077731, "mean_data_wait_seconds": 0.0001698136329650879}
{"step": 7020, "total_steps": 46238, "tokens": 3796058112, "loss": 3.2549281933090906, "ce_loss": 3.253756066163381, "z_loss": 11.72137182409113, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17353342473506927, "tokens_per_second": 4109724.046178515, "eta_seconds": 5159.357072579218, "mean_data_wait_seconds": 0.00013646483421325684}
{"step": 7030, "total_steps": 46238, "tokens": 3801464832, "loss": 3.2607758767677075, "ce_loss": 3.2595538063482805, "z_loss": 12.220708406332767, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16383948922157288, "tokens_per_second": 4109607.8197367955, "eta_seconds": 5158.1873584612895, "mean_data_wait_seconds": 0.0001506507396697998}
{"step": 7040, "total_steps": 46238, "tokens": 3806871552, "loss": 3.285831259958672, "ce_loss": 3.284605504888477, "z_loss": 12.257286866505941, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1790281981229782, "tokens_per_second": 4110090.5032460187, "eta_seconds": 5156.266113182341, "mean_data_wait_seconds": 0.00013107061386108398}
{"step": 7050, "total_steps": 46238, "tokens": 3812278272, "loss": 3.182866121783401, "ce_loss": 3.181663363268881, "z_loss": 12.02725345438177, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19368618726730347, "tokens_per_second": 4110362.4411656912, "eta_seconds": 5154.609591554977, "mean_data_wait_seconds": 0.00026726722717285156}
{"step": 7060, "total_steps": 46238, "tokens": 3817684992, "loss": 3.2450647986296453, "ce_loss": 3.243882330981168, "z_loss": 11.824390685919559, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1540517956018448, "tokens_per_second": 4110424.4678284493, "eta_seconds": 5153.2164402452745, "mean_data_wait_seconds": 0.0001284480094909668}
{"step": 7070, "total_steps": 46238, "tokens": 3823091712, "loss": 3.3142575332612703, "ce_loss": 3.313063321691571, "z_loss": 11.941944042841593, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1869831383228302, "tokens_per_second": 4110463.274793769, "eta_seconds": 5151.852433242449, "mean_data_wait_seconds": 0.00017073750495910645}
{"step": 7080, "total_steps": 46238, "tokens": 3828498432, "loss": 3.286751586379427, "ce_loss": 3.285525067286058, "z_loss": 12.265325408993345, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17664870619773865, "tokens_per_second": 4110678.9733571555, "eta_seconds": 5150.266815097399, "mean_data_wait_seconds": 9.757280349731445e-05}
{"step": 7090, "total_steps": 46238, "tokens": 3833905152, "loss": 3.226745186430035, "ce_loss": 3.225573763702855, "z_loss": 11.714085008158829, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17621494829654694, "tokens_per_second": 4110319.6325360048, "eta_seconds": 5149.401670969587, "mean_data_wait_seconds": 9.775161743164062e-05}
{"step": 7100, "total_steps": 46238, "tokens": 3839311872, "loss": 3.2427956768960664, "ce_loss": 3.2416301279356987, "z_loss": 11.655750700921724, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21912461519241333, "tokens_per_second": 4110734.6640765257, "eta_seconds": 5147.566504089518, "mean_data_wait_seconds": 9.143352508544922e-05}
{"step": 7110, "total_steps": 46238, "tokens": 3844718592, "loss": 3.2837456880193767, "ce_loss": 3.2826044685912854, "z_loss": 11.412001205213143, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16395032405853271, "tokens_per_second": 4111632.635356797, "eta_seconds": 5145.127305899067, "mean_data_wait_seconds": 7.957220077514648e-05}
{"step": 7120, "total_steps": 46238, "tokens": 3850125312, "loss": 3.251048805135669, "ce_loss": 3.249840743613966, "z_loss": 12.080627809871327, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15313759446144104, "tokens_per_second": 4111807.014354384, "eta_seconds": 5143.594178950246, "mean_data_wait_seconds": 8.100271224975586e-05}
{"step": 7130, "total_steps": 46238, "tokens": 3855532032, "loss": 3.336531998533191, "ce_loss": 3.3353535424579275, "z_loss": 11.78466122078173, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20682042837142944, "tokens_per_second": 4112208.9029565253, "eta_seconds": 5141.776695439331, "mean_data_wait_seconds": 6.976723670959473e-05}
{"step": 7140, "total_steps": 46238, "tokens": 3860938752, "loss": 3.2833954059716426, "ce_loss": 3.2822112477186955, "z_loss": 11.841616023670543, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16901825368404388, "tokens_per_second": 4112942.5015617586, "eta_seconds": 5139.545027914509, "mean_data_wait_seconds": 0.0004684627056121826}
{"step": 7150, "total_steps": 46238, "tokens": 3866345472, "loss": 3.251684954672149, "ce_loss": 3.250500556194421, "z_loss": 11.84450196497368, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16048818826675415, "tokens_per_second": 4113361.10171777, "eta_seconds": 5137.707568434631, "mean_data_wait_seconds": 5.951523780822754e-05}
{"step": 7160, "total_steps": 46238, "tokens": 3871752192, "loss": 3.3227003252867497, "ce_loss": 3.321481729998733, "z_loss": 12.185984474239927, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18020471930503845, "tokens_per_second": 4113479.471962874, "eta_seconds": 5136.245333909056, "mean_data_wait_seconds": 0.0001023709774017334}
{"step": 7170, "total_steps": 46238, "tokens": 3877158912, "loss": 3.261687154119665, "ce_loss": 3.260508905757557, "z_loss": 11.78287704785665, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1641150414943695, "tokens_per_second": 4114060.8158919024, "eta_seconds": 5134.205343393979, "mean_data_wait_seconds": 0.00010034441947937012}
{"step": 7180, "total_steps": 46238, "tokens": 3882565632, "loss": 3.275435315840172, "ce_loss": 3.2742581620360864, "z_loss": 11.771147894136833, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1881714016199112, "tokens_per_second": 4114683.5943273916, "eta_seconds": 5132.1142488604655, "mean_data_wait_seconds": 0.00010940432548522949}
{"step": 7190, "total_steps": 46238, "tokens": 3887972352, "loss": 3.316775173851938, "ce_loss": 3.3155716585390493, "z_loss": 12.035296115008267, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17326730489730835, "tokens_per_second": 4115565.827694538, "eta_seconds": 5129.700378483882, "mean_data_wait_seconds": 8.183717727661133e-05}
{"step": 7200, "total_steps": 46238, "tokens": 3893379072, "loss": 3.280055831779133, "ce_loss": 3.278864293387442, "z_loss": 11.915537863066703, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1885032057762146, "tokens_per_second": 4116401.5966403857, "eta_seconds": 5127.345417712864, "mean_data_wait_seconds": 9.039044380187988e-05}
{"step": 7210, "total_steps": 46238, "tokens": 3898785792, "loss": 3.3038256204489507, "ce_loss": 3.302638519893993, "z_loss": 11.870981584895741, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15858270227909088, "tokens_per_second": 4116365.2743302584, "eta_seconds": 5126.0771913476865, "mean_data_wait_seconds": 0.00011458992958068848}
{"step": 7220, "total_steps": 46238, "tokens": 3904192512, "loss": 3.2416202678824915, "ce_loss": 3.2403999639279917, "z_loss": 12.20273215120489, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1881023794412613, "tokens_per_second": 4116731.092152452, "eta_seconds": 5124.308329055851, "mean_data_wait_seconds": 8.779764175415039e-05}
{"step": 7230, "total_steps": 46238, "tokens": 3909599232, "loss": 3.281532018473654, "ce_loss": 3.280344029267629, "z_loss": 11.879970998475045, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18203943967819214, "tokens_per_second": 4117731.282391312, "eta_seconds": 5121.750609173383, "mean_data_wait_seconds": 0.00017458200454711914}
{"step": 7240, "total_steps": 46238, "tokens": 3915005952, "loss": 3.29604477593393, "ce_loss": 3.294863083145835, "z_loss": 11.81713360006159, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19546517729759216, "tokens_per_second": 4118152.4658283526, "eta_seconds": 5119.9138834600935, "mean_data_wait_seconds": 8.311867713928223e-05}
{"step": 7250, "total_steps": 46238, "tokens": 3920412672, "loss": 3.242460725885449, "ce_loss": 3.241281574422663, "z_loss": 11.791540384292603, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15593722462654114, "tokens_per_second": 4118612.5298240404, "eta_seconds": 5118.02921769399, "mean_data_wait_seconds": 8.299946784973145e-05}
{"step": 7260, "total_steps": 46238, "tokens": 3925819392, "loss": 3.333059258533247, "ce_loss": 3.331851303577423, "z_loss": 12.07988708669489, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21084006130695343, "tokens_per_second": 4119183.335154803, "eta_seconds": 5116.007428984228, "mean_data_wait_seconds": 9.72449779510498e-05}
{"step": 7270, "total_steps": 46238, "tokens": 3931226112, "loss": 3.3101730780168013, "ce_loss": 3.3089567495114878, "z_loss": 12.163087050120035, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1721070408821106, "tokens_per_second": 4119334.3552863915, "eta_seconds": 5114.507346790801, "mean_data_wait_seconds": 7.12275505065918e-05}
{"step": 7280, "total_steps": 46238, "tokens": 3936632832, "loss": 3.3533868193626404, "ce_loss": 3.3522238063089773, "z_loss": 11.629825628165042, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16923709213733673, "tokens_per_second": 4119673.750938053, "eta_seconds": 5112.773578054318, "mean_data_wait_seconds": 0.0001666545867919922}
{"step": 7290, "total_steps": 46238, "tokens": 3942039552, "loss": 3.3039420788938347, "ce_loss": 3.3027495525100012, "z_loss": 11.925234845190337, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2077074646949768, "tokens_per_second": 4120048.875267176, "eta_seconds": 5110.995772503904, "mean_data_wait_seconds": 0.00010025501251220703}
{"step": 7300, "total_steps": 46238, "tokens": 3947446272, "loss": 3.2434699842424104, "ce_loss": 3.2423013539025276, "z_loss": 11.68623255960869, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19130147993564606, "tokens_per_second": 4120400.9492456964, "eta_seconds": 5109.246872650567, "mean_data_wait_seconds": 9.420514106750488e-05}
{"step": 7310, "total_steps": 46238, "tokens": 3952852992, "loss": 3.2891518329129075, "ce_loss": 3.287928145943266, "z_loss": 12.23640751838684, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16713325679302216, "tokens_per_second": 4120809.5392085994, "eta_seconds": 5107.42822344612, "mean_data_wait_seconds": 9.992718696594238e-05}
{"step": 7320, "total_steps": 46238, "tokens": 3958259712, "loss": 3.2792542414231733, "ce_loss": 3.278045016707796, "z_loss": 12.091965552532312, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14285415410995483, "tokens_per_second": 4121099.390693846, "eta_seconds": 5105.757039375211, "mean_data_wait_seconds": 0.00018405914306640625}
{"step": 7330, "total_steps": 46238, "tokens": 3963666432, "loss": 3.272488017876943, "ce_loss": 3.2713223099708557, "z_loss": 11.657242919459488, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16767661273479462, "tokens_per_second": 4121695.5752469976, "eta_seconds": 5103.706743974995, "mean_data_wait_seconds": 0.0001106560230255127}
{"step": 7340, "total_steps": 46238, "tokens": 3969073152, "loss": 3.301700655258063, "ce_loss": 3.300492472720869, "z_loss": 12.08217883832527, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17798107862472534, "tokens_per_second": 4121929.583295299, "eta_seconds": 5102.105302630385, "mean_data_wait_seconds": 8.702278137207031e-05}
{"step": 7350, "total_steps": 46238, "tokens": 3974479872, "loss": 3.265772850224466, "ce_loss": 3.2645818175691548, "z_loss": 11.910240664626613, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1479302942752838, "tokens_per_second": 4122007.7426653383, "eta_seconds": 5100.696887678556, "mean_data_wait_seconds": 0.00015291571617126465}
{"step": 7360, "total_steps": 46238, "tokens": 3979886592, "loss": 3.335757295290629, "ce_loss": 3.334548959226319, "z_loss": 12.082973920937741, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14414913952350616, "tokens_per_second": 4122766.359108583, "eta_seconds": 5098.446895386243, "mean_data_wait_seconds": 9.328126907348633e-05}
{"step": 7370, "total_steps": 46238, "tokens": 3985293312, "loss": 3.2548988949168813, "ce_loss": 3.2537002166112265, "z_loss": 11.986907720565796, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17932935059070587, "tokens_per_second": 4123159.7466969113, "eta_seconds": 5096.649151378305, "mean_data_wait_seconds": 0.00011608004570007324}
{"step": 7380, "total_steps": 46238, "tokens": 3990700032, "loss": 3.2547795898986585, "ce_loss": 3.2535896536075706, "z_loss": 11.899525931387236, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16095302999019623, "tokens_per_second": 4123793.0156307346, "eta_seconds": 5094.555382476365, "mean_data_wait_seconds": 7.644295692443848e-05}
{"step": 7390, "total_steps": 46238, "tokens": 3996106752, "loss": 3.2336209271893357, "ce_loss": 3.232426076224356, "z_loss": 11.94863399592313, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1832219958305359, "tokens_per_second": 4124332.1344579766, "eta_seconds": 5092.578507079983, "mean_data_wait_seconds": 6.124377250671387e-05}
{"step": 7400, "total_steps": 46238, "tokens": 4001513472, "loss": 3.285854697227478, "ce_loss": 3.284657203789913, "z_loss": 11.974658250808716, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17317359149456024, "tokens_per_second": 4124565.1664932766, "eta_seconds": 5090.979925491796, "mean_data_wait_seconds": 6.860494613647461e-05}
{"step": 7410, "total_steps": 46238, "tokens": 4006920192, "loss": 3.239088961572358, "ce_loss": 3.237847218007752, "z_loss": 12.417024446256233, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18587107956409454, "tokens_per_second": 4125431.5881730197, "eta_seconds": 5088.600136815448, "mean_data_wait_seconds": 0.00012239813804626465}
{"step": 7420, "total_steps": 46238, "tokens": 4012326912, "loss": 3.2383929472981077, "ce_loss": 3.237155365221428, "z_loss": 12.37565936464252, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19709692895412445, "tokens_per_second": 4125920.9494115165, "eta_seconds": 5086.686168088952, "mean_data_wait_seconds": 0.0007914304733276367}
{"step": 7430, "total_steps": 46238, "tokens": 4017733632, "loss": 3.2540124456087747, "ce_loss": 3.2528049060792634, "z_loss": 12.075455672813185, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1659083217382431, "tokens_per_second": 4126332.8084772276, "eta_seconds": 5084.8681572398655, "mean_data_wait_seconds": 7.602572441101074e-05}
{"step": 7440, "total_steps": 46238, "tokens": 4023140352, "loss": 3.25242748224374, "ce_loss": 3.2512575098962495, "z_loss": 11.699481154933121, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14895480871200562, "tokens_per_second": 4126627.672127188, "eta_seconds": 5083.194620557344, "mean_data_wait_seconds": 0.00012859702110290527}
{"step": 7450, "total_steps": 46238, "tokens": 4028547072, "loss": 3.2438352126063723, "ce_loss": 3.2426364132852266, "z_loss": 11.988014864199089, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14883296191692352, "tokens_per_second": 4126524.6794287222, "eta_seconds": 5082.01125478383, "mean_data_wait_seconds": 0.0001323223114013672}
{"step": 7460, "total_steps": 46238, "tokens": 4033953792, "loss": 3.22138606598883, "ce_loss": 3.2201834664200293, "z_loss": 12.025770418571703, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15072287619113922, "tokens_per_second": 4127025.2423937325, "eta_seconds": 5080.0847856795845, "mean_data_wait_seconds": 0.0001226365566253662}
{"step": 7470, "total_steps": 46238, "tokens": 4039360512, "loss": 3.2057864015752617, "ce_loss": 3.2046195124134873, "z_loss": 11.668809413909912, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18356439471244812, "tokens_per_second": 4127424.4573560776, "eta_seconds": 5078.283476913491, "mean_data_wait_seconds": 0.00013250112533569336}
{"step": 7480, "total_steps": 46238, "tokens": 4044767232, "loss": 3.3009167620629976, "ce_loss": 3.299759729342027, "z_loss": 11.570319067348134, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19837428629398346, "tokens_per_second": 4127484.0170771414, "eta_seconds": 5076.900265949197, "mean_data_wait_seconds": 0.00014394521713256836}
{"step": 7490, "total_steps": 46238, "tokens": 4050173952, "loss": 3.172688614238392, "ce_loss": 3.1714654691291577, "z_loss": 12.231495084184589, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16163644194602966, "tokens_per_second": 4127895.086865863, "eta_seconds": 5075.084890276611, "mean_data_wait_seconds": 0.00011467933654785156}
{"step": 7500, "total_steps": 46238, "tokens": 4055580672, "loss": 3.1915892886392996, "ce_loss": 3.190388525977279, "z_loss": 12.007594029108683, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1476895958185196, "tokens_per_second": 4128289.4342204626, "eta_seconds": 5073.290426390567, "mean_data_wait_seconds": 9.47117805480957e-05}
{"step": 7510, "total_steps": 46238, "tokens": 4060987392, "loss": 3.241067080786734, "ce_loss": 3.2398514585061506, "z_loss": 12.156781962423613, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18555611371994019, "tokens_per_second": 4128633.889830325, "eta_seconds": 5071.55759089613, "mean_data_wait_seconds": 8.866190910339355e-05}
{"step": 7520, "total_steps": 46238, "tokens": 4066394112, "loss": 3.2363828203894873, "ce_loss": 3.235191397594683, "z_loss": 11.913913589535337, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15853042900562286, "tokens_per_second": 4128522.1462720353, "eta_seconds": 5070.38525708339, "mean_data_wait_seconds": 6.458163261413574e-05}
{"step": 7530, "total_steps": 46238, "tokens": 4071800832, "loss": 3.2121759978207676, "ce_loss": 3.2109974640788455, "z_loss": 11.786051952477658, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18376260995864868, "tokens_per_second": 4128850.8471072502, "eta_seconds": 5068.672102471902, "mean_data_wait_seconds": 7.420778274536133e-05}
{"step": 7540, "total_steps": 46238, "tokens": 4077207552, "loss": 3.280390206611518, "ce_loss": 3.279216199210196, "z_loss": 11.740008354187012, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16282114386558533, "tokens_per_second": 4129648.01796724, "eta_seconds": 5066.38442137709, "mean_data_wait_seconds": 7.700920104980469e-05}
{"step": 7550, "total_steps": 46238, "tokens": 4082614272, "loss": 3.2149686361804153, "ce_loss": 3.2137659679759634, "z_loss": 12.026503671299327, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17435956001281738, "tokens_per_second": 4129982.639450662, "eta_seconds": 5064.664791613316, "mean_data_wait_seconds": 6.070733070373535e-05}
{"step": 7560, "total_steps": 46238, "tokens": 4088020992, "loss": 3.2510073853261545, "ce_loss": 3.24985121235703, "z_loss": 11.561505064819798, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18849113583564758, "tokens_per_second": 4130102.757233106, "eta_seconds": 5063.208392909178, "mean_data_wait_seconds": 6.73532485961914e-05}
{"step": 7570, "total_steps": 46238, "tokens": 4093427712, "loss": 3.230387772574569, "ce_loss": 3.2292376214807685, "z_loss": 11.501528617107507, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17988745868206024, "tokens_per_second": 4131234.412118283, "eta_seconds": 5060.512703582076, "mean_data_wait_seconds": 8.130073547363281e-05}
{"step": 7580, "total_steps": 46238, "tokens": 4098834432, "loss": 3.2567980632637488, "ce_loss": 3.2556416735504614, "z_loss": 11.563656026666814, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1800573170185089, "tokens_per_second": 4131506.1372297644, "eta_seconds": 5058.871222690296, "mean_data_wait_seconds": 0.0001221299171447754}
{"step": 7590, "total_steps": 46238, "tokens": 4104241152, "loss": 3.2510428482835945, "ce_loss": 3.2498772740364075, "z_loss": 11.656301086599177, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15922869741916656, "tokens_per_second": 4132072.6528830607, "eta_seconds": 5056.869164539191, "mean_data_wait_seconds": 7.522106170654297e-05}
{"step": 7600, "total_steps": 46238, "tokens": 4109647872, "loss": 3.30355251196659, "ce_loss": 3.30239276452498, "z_loss": 11.597356890187118, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1947646290063858, "tokens_per_second": 4132375.6919971993, "eta_seconds": 5055.189949078366, "mean_data_wait_seconds": 6.154179573059082e-05}
{"step": 7610, "total_steps": 46238, "tokens": 4115054592, "loss": 3.2017839605158027, "ce_loss": 3.2005834958770056, "z_loss": 12.00445344953826, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.167434424161911, "tokens_per_second": 4133133.855930513, "eta_seconds": 5052.954506671345, "mean_data_wait_seconds": 6.0111284255981445e-05}
{"step": 7620, "total_steps": 46238, "tokens": 4120461312, "loss": 3.234679299773592, "ce_loss": 3.233500468008446, "z_loss": 11.788367235299313, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1830424964427948, "tokens_per_second": 4133825.7619106914, "eta_seconds": 5050.80083838596, "mean_data_wait_seconds": 7.346272468566895e-05}
{"step": 7630, "total_steps": 46238, "tokens": 4125868032, "loss": 3.237632671991984, "ce_loss": 3.23642652684992, "z_loss": 12.06115932898088, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1497470587491989, "tokens_per_second": 4134584.8341489234, "eta_seconds": 5048.5658757312, "mean_data_wait_seconds": 0.00010854005813598633}
{"step": 7640, "total_steps": 46238, "tokens": 4131274752, "loss": 3.2649616400400796, "ce_loss": 3.263794355320208, "z_loss": 11.672702348593509, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1774672567844391, "tokens_per_second": 4135044.280522203, "eta_seconds": 5046.697391439929, "mean_data_wait_seconds": 7.912516593933105e-05}
{"step": 7650, "total_steps": 46238, "tokens": 4136681472, "loss": 3.270686214620417, "ce_loss": 3.26950916557601, "z_loss": 11.770636334563747, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1507648229598999, "tokens_per_second": 4135661.468368361, "eta_seconds": 5044.636903569147, "mean_data_wait_seconds": 7.474422454833984e-05}
{"step": 7660, "total_steps": 46238, "tokens": 4142088192, "loss": 3.2805000453284294, "ce_loss": 3.279319167137146, "z_loss": 11.808759783253524, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16170980036258698, "tokens_per_second": 4135913.2664644006, "eta_seconds": 5043.022520109593, "mean_data_wait_seconds": 7.888674736022949e-05}
{"step": 7670, "total_steps": 46238, "tokens": 4147494912, "loss": 3.278148526495153, "ce_loss": 3.276954699646343, "z_loss": 11.938150145790793, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17039182782173157, "tokens_per_second": 4136062.6929686363, "eta_seconds": 5041.53311298904, "mean_data_wait_seconds": 7.858872413635254e-05}
{"step": 7680, "total_steps": 46238, "tokens": 4152901632, "loss": 3.2997017864025002, "ce_loss": 3.2985102931658425, "z_loss": 11.915072412201852, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1800454556941986, "tokens_per_second": 4136309.7539002527, "eta_seconds": 5039.924847103875, "mean_data_wait_seconds": 9.748339653015137e-05}
{"step": 7690, "total_steps": 46238, "tokens": 4158308352, "loss": 3.2226623206427605, "ce_loss": 3.2215155308896843, "z_loss": 11.467462062835693, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14227233827114105, "tokens_per_second": 4136869.80291574, "eta_seconds": 5037.9355833994805, "mean_data_wait_seconds": 6.273388862609863e-05}
{"step": 7700, "total_steps": 46238, "tokens": 4163715072, "loss": 3.330818398432298, "ce_loss": 3.329638473915331, "z_loss": 11.79933857195305, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1904565542936325, "tokens_per_second": 4137346.881808367, "eta_seconds": 5036.047848831321, "mean_data_wait_seconds": 6.22868537902832e-05}
{"step": 7710, "total_steps": 46238, "tokens": 4169121792, "loss": 3.257076915466424, "ce_loss": 3.2558844649430476, "z_loss": 11.924331874558419, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1628011018037796, "tokens_per_second": 4137627.124363848, "eta_seconds": 5034.400036035786, "mean_data_wait_seconds": 6.878376007080078e-05}
{"step": 7720, "total_steps": 46238, "tokens": 4174528512, "loss": 3.2468426462375755, "ce_loss": 3.2456800124861975, "z_loss": 11.626434759660201, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16785281896591187, "tokens_per_second": 4138070.5645718235, "eta_seconds": 5032.553964230144, "mean_data_wait_seconds": 0.0004264712333679199}
{"step": 7730, "total_steps": 46238, "tokens": 4179935232, "loss": 3.291956417488329, "ce_loss": 3.290744673122059, "z_loss": 12.117533560955163, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20703530311584473, "tokens_per_second": 4138330.91997843, "eta_seconds": 5030.930852699551, "mean_data_wait_seconds": 9.950995445251465e-05}
{"step": 7740, "total_steps": 46238, "tokens": 4185341952, "loss": 3.2878322131706006, "ce_loss": 3.2866039908293523, "z_loss": 12.282571561408766, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.164496511220932, "tokens_per_second": 4138711.1235392257, "eta_seconds": 5029.162307467514, "mean_data_wait_seconds": 6.747245788574219e-05}
{"step": 7750, "total_steps": 46238, "tokens": 4190748672, "loss": 3.3167758818828696, "ce_loss": 3.31558811303341, "z_loss": 11.87765478365349, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20287464559078217, "tokens_per_second": 4139089.2431241316, "eta_seconds": 5027.396618366642, "mean_data_wait_seconds": 8.058547973632812e-05}
{"step": 7760, "total_steps": 46238, "tokens": 4196155392, "loss": 3.255867479425488, "ce_loss": 3.2546981789849023, "z_loss": 11.69284025105563, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17196831107139587, "tokens_per_second": 4139790.860242361, "eta_seconds": 5025.238531683235, "mean_data_wait_seconds": 7.581710815429688e-05}
{"step": 7770, "total_steps": 46238, "tokens": 4201562112, "loss": 3.2367800347732776, "ce_loss": 3.2355994979540506, "z_loss": 11.805171728134155, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20086009800434113, "tokens_per_second": 4140691.6531684194, "eta_seconds": 5022.839555822888, "mean_data_wait_seconds": 9.351968765258789e-05}
{"step": 7780, "total_steps": 46238, "tokens": 4206968832, "loss": 3.3389086199529245, "ce_loss": 3.3377373868768867, "z_loss": 11.712360815568404, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21731051802635193, "tokens_per_second": 4141028.9188320576, "eta_seconds": 5021.124824664201, "mean_data_wait_seconds": 6.240606307983398e-05}
{"step": 7790, "total_steps": 46238, "tokens": 4212375552, "loss": 3.295807715618249, "ce_loss": 3.294610579808553, "z_loss": 11.97137594945503, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1533203274011612, "tokens_per_second": 4141116.4825779945, "eta_seconds": 5019.713034263457, "mean_data_wait_seconds": 7.846951484680176e-05}
{"step": 7800, "total_steps": 46238, "tokens": 4217782272, "loss": 3.291487935817603, "ce_loss": 3.29032762484117, "z_loss": 11.603098573106708, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1728127896785736, "tokens_per_second": 4141245.975086382, "eta_seconds": 5018.250494904861, "mean_data_wait_seconds": 6.058812141418457e-05}
{"step": 7810, "total_steps": 46238, "tokens": 4223188992, "loss": 3.3643076690760525, "ce_loss": 3.3631587100751474, "z_loss": 11.489358938101565, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1663757562637329, "tokens_per_second": 4141781.151928048, "eta_seconds": 5016.296656410332, "mean_data_wait_seconds": 5.856156349182129e-05}
{"step": 7820, "total_steps": 46238, "tokens": 4228595712, "loss": 3.270447870095571, "ce_loss": 3.269308529116891, "z_loss": 11.39304553378712, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15301063656806946, "tokens_per_second": 4142181.4253813694, "eta_seconds": 5014.506630908283, "mean_data_wait_seconds": 9.891390800476074e-05}
{"step": 7830, "total_steps": 46238, "tokens": 4234002432, "loss": 3.322502831617991, "ce_loss": 3.321299211545424, "z_loss": 12.035977667028254, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14788508415222168, "tokens_per_second": 4142396.8890945576, "eta_seconds": 5012.940589702627, "mean_data_wait_seconds": 6.243586540222168e-05}
{"step": 7840, "total_steps": 46238, "tokens": 4239409152, "loss": 3.3260126330635766, "ce_loss": 3.324802774371523, "z_loss": 12.098674246759126, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15812808275222778, "tokens_per_second": 4143231.7249802933, "eta_seconds": 5010.625560437063, "mean_data_wait_seconds": 6.118416786193848e-05}
{"step": 7850, "total_steps": 46238, "tokens": 4244815872, "loss": 3.2932724699829565, "ce_loss": 3.2920768134521716, "z_loss": 11.956733999830304, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1901978701353073, "tokens_per_second": 4143513.838130831, "eta_seconds": 5008.979546056646, "mean_data_wait_seconds": 6.461143493652344e-05}
{"step": 7860, "total_steps": 46238, "tokens": 4250222592, "loss": 3.2736669327273513, "ce_loss": 3.2725192723852214, "z_loss": 11.477033015453454, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15671710669994354, "tokens_per_second": 4144217.2952043447, "eta_seconds": 5006.824658545536, "mean_data_wait_seconds": 7.766485214233398e-05}
{"step": 7870, "total_steps": 46238, "tokens": 4255629312, "loss": 3.290046914057298, "ce_loss": 3.2889160694498005, "z_loss": 11.308412833647294, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20756807923316956, "tokens_per_second": 4144765.744758497, "eta_seconds": 5004.857669033039, "mean_data_wait_seconds": 8.812546730041504e-05}
{"step": 7880, "total_steps": 46238, "tokens": 4261036032, "loss": 3.3107885501601477, "ce_loss": 3.3096427231123955, "z_loss": 11.45848332751881, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18027010560035706, "tokens_per_second": 4144876.778916127, "eta_seconds": 5003.419163023484, "mean_data_wait_seconds": 8.493661880493164e-05}
{"step": 7890, "total_steps": 46238, "tokens": 4266442752, "loss": 3.2743662758307024, "ce_loss": 3.2731959639173565, "z_loss": 11.70314114744013, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15955594182014465, "tokens_per_second": 4145042.692747567, "eta_seconds": 5001.914508691563, "mean_data_wait_seconds": 5.97834587097168e-05}
{"step": 7900, "total_steps": 46238, "tokens": 4271849472, "loss": 3.3081518086520108, "ce_loss": 3.306986077265306, "z_loss": 11.657434405702533, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1770450621843338, "tokens_per_second": 4145346.876267552, "eta_seconds": 5000.24318415137, "mean_data_wait_seconds": 8.291006088256836e-05}
{"step": 7910, "total_steps": 46238, "tokens": 4277256192, "loss": 3.258465947526874, "ce_loss": 3.257284451614727, "z_loss": 11.814889907836914, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19527091085910797, "tokens_per_second": 4145573.1602001446, "eta_seconds": 4998.666033190822, "mean_data_wait_seconds": 8.729100227355957e-05}
{"step": 7920, "total_steps": 46238, "tokens": 4282662912, "loss": 3.2883846470803926, "ce_loss": 3.2871854612321565, "z_loss": 11.99186186356978, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17146526277065277, "tokens_per_second": 4145283.2686431967, "eta_seconds": 4997.711297732594, "mean_data_wait_seconds": 7.331371307373047e-05}
{"step": 7930, "total_steps": 46238, "tokens": 4288069632, "loss": 3.2871594988938533, "ce_loss": 3.2859900051897224, "z_loss": 11.694686513958555, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15649345517158508, "tokens_per_second": 4145410.1080888375, "eta_seconds": 4996.254113335159, "mean_data_wait_seconds": 0.00011399388313293457}
{"step": 7940, "total_steps": 46238, "tokens": 4293476352, "loss": 3.2743607795599736, "ce_loss": 3.2731871261741174, "z_loss": 11.736769676208496, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16499480605125427, "tokens_per_second": 4145073.3162521305, "eta_seconds": 4995.355691976503, "mean_data_wait_seconds": 7.098913192749023e-05}
{"step": 7950, "total_steps": 46238, "tokens": 4298883072, "loss": 3.321493780974186, "ce_loss": 3.3203667998313904, "z_loss": 11.270139332973596, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20084553956985474, "tokens_per_second": 4145087.011572653, "eta_seconds": 4994.03481910169, "mean_data_wait_seconds": 6.720423698425293e-05}
{"step": 7960, "total_steps": 46238, "tokens": 4304289792, "loss": 3.251645568645362, "ce_loss": 3.2504768190961895, "z_loss": 11.687436681805234, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.194074347615242, "tokens_per_second": 4145292.2405835255, "eta_seconds": 4992.483266049961, "mean_data_wait_seconds": 6.446242332458496e-05}
{"step": 7970, "total_steps": 46238, "tokens": 4309696512, "loss": 3.320146253614715, "ce_loss": 3.318945443991459, "z_loss": 12.008037646611532, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17133782804012299, "tokens_per_second": 4145835.4322969713, "eta_seconds": 4990.525012840875, "mean_data_wait_seconds": 8.362531661987305e-05}
{"step": 7980, "total_steps": 46238, "tokens": 4315103232, "loss": 3.2592556476593018, "ce_loss": 3.2581054333484536, "z_loss": 11.502011075164333, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17830637097358704, "tokens_per_second": 4146098.2299016975, "eta_seconds": 4988.904641675705, "mean_data_wait_seconds": 0.00010401010513305664}
{"step": 7990, "total_steps": 46238, "tokens": 4320509952, "loss": 3.2629853017402417, "ce_loss": 3.2617718342578774, "z_loss": 12.134607885823105, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1982961744070053, "tokens_per_second": 4146310.120336846, "eta_seconds": 4987.345708313789, "mean_data_wait_seconds": 7.939338684082031e-05}
{"step": 8000, "total_steps": 46238, "tokens": 4325916672, "loss": 3.285402404539513, "ce_loss": 3.2842118017601245, "z_loss": 11.906071438933864, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2014296054840088, "tokens_per_second": 4146598.6168626775, "eta_seconds": 4985.694824651664, "mean_data_wait_seconds": 6.172060966491699e-05}
{"step": 8010, "total_steps": 46238, "tokens": 4331323392, "loss": 3.245116221182274, "ce_loss": 3.2439493013150766, "z_loss": 11.668855681563869, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17574548721313477, "tokens_per_second": 4146973.254667141, "eta_seconds": 4983.940641705188, "mean_data_wait_seconds": 5.7309865951538086e-05}
{"step": 8020, "total_steps": 46238, "tokens": 4336730112, "loss": 3.2145172465931284, "ce_loss": 3.2133240392713835, "z_loss": 11.932152581937386, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1543562114238739, "tokens_per_second": 4147540.13117451, "eta_seconds": 4981.955851057345, "mean_data_wait_seconds": 6.300210952758789e-05}
{"step": 8030, "total_steps": 46238, "tokens": 4342136832, "loss": 3.197414669123563, "ce_loss": 3.1962305036458103, "z_loss": 11.841433691255974, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1666758805513382, "tokens_per_second": 4148034.3225003663, "eta_seconds": 4980.05886594208, "mean_data_wait_seconds": 7.635354995727539e-05}
{"step": 8040, "total_steps": 46238, "tokens": 4347543552, "loss": 3.2516699704256924, "ce_loss": 3.2504634405627395, "z_loss": 12.065566503640378, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17869833111763, "tokens_per_second": 4148355.823681081, "eta_seconds": 4978.369566589931, "mean_data_wait_seconds": 6.246566772460938e-05}
{"step": 8050, "total_steps": 46238, "tokens": 4352950272, "loss": 3.242919871301362, "ce_loss": 3.241755565007528, "z_loss": 11.643133307948258, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20328764617443085, "tokens_per_second": 4149016.6485300977, "eta_seconds": 4976.273515632827, "mean_data_wait_seconds": 0.00011029839515686035}
{"step": 8060, "total_steps": 46238, "tokens": 4358356992, "loss": 3.1950938773877695, "ce_loss": 3.19392563718738, "z_loss": 11.682283878326416, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1893848180770874, "tokens_per_second": 4149528.552260744, "eta_seconds": 4974.356648962989, "mean_data_wait_seconds": 7.593631744384766e-05}
{"step": 8070, "total_steps": 46238, "tokens": 4363763712, "loss": 3.1819354927901067, "ce_loss": 3.180778629852064, "z_loss": 11.568944562565196, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1828315258026123, "tokens_per_second": 4150133.718067123, "eta_seconds": 4972.328514178791, "mean_data_wait_seconds": 5.8323144912719727e-05}
{"step": 8080, "total_steps": 46238, "tokens": 4369170432, "loss": 3.155547985524842, "ce_loss": 3.154376436363567, "z_loss": 11.715224475571603, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18440671265125275, "tokens_per_second": 4150238.460399048, "eta_seconds": 4970.900274972723, "mean_data_wait_seconds": 6.127357482910156e-05}
{"step": 8090, "total_steps": 46238, "tokens": 4374577152, "loss": 3.199982327042204, "ce_loss": 3.1987706838232097, "z_loss": 12.116166035334269, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18783175945281982, "tokens_per_second": 4150693.0365369646, "eta_seconds": 4969.053264706852, "mean_data_wait_seconds": 6.204843521118164e-05}
{"step": 8100, "total_steps": 46238, "tokens": 4379983872, "loss": 3.2418058713277182, "ce_loss": 3.240628639856974, "z_loss": 11.772191900195498, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17395658791065216, "tokens_per_second": 4149714.957219719, "eta_seconds": 4968.92154679824, "mean_data_wait_seconds": 5.91278076171875e-05}
{"step": 8110, "total_steps": 46238, "tokens": 4385390592, "loss": 3.2370120359189585, "ce_loss": 3.235820006240498, "z_loss": 11.920294479890304, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2181982398033142, "tokens_per_second": 4150533.267562161, "eta_seconds": 4966.639228049813, "mean_data_wait_seconds": 6.362795829772949e-05}
{"step": 8120, "total_steps": 46238, "tokens": 4390797312, "loss": 3.1870191530747847, "ce_loss": 3.1858008792906096, "z_loss": 12.182857701272676, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19847315549850464, "tokens_per_second": 4151361.2259453586, "eta_seconds": 4964.346271579128, "mean_data_wait_seconds": 0.00011998414993286133}
{"step": 8130, "total_steps": 46238, "tokens": 4396204032, "loss": 3.239503587737228, "ce_loss": 3.238283036333142, "z_loss": 12.205176815842137, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1758720576763153, "tokens_per_second": 4151493.7810001597, "eta_seconds": 4962.88540724643, "mean_data_wait_seconds": 0.00010070204734802246}
{"step": 8140, "total_steps": 46238, "tokens": 4401610752, "loss": 3.302676908897631, "ce_loss": 3.3014712965849675, "z_loss": 12.056545091397835, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1923643797636032, "tokens_per_second": 4151954.916044409, "eta_seconds": 4961.0319958926275, "mean_data_wait_seconds": 0.00010213255882263184}
{"step": 8150, "total_steps": 46238, "tokens": 4407017472, "loss": 3.2183507933761133, "ce_loss": 3.217206008506544, "z_loss": 11.44815587274956, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14972156286239624, "tokens_per_second": 4152736.340717433, "eta_seconds": 4958.796507760566, "mean_data_wait_seconds": 0.00015214085578918457}
{"step": 8160, "total_steps": 46238, "tokens": 4412424192, "loss": 3.22693339622382, "ce_loss": 3.2257856618274343, "z_loss": 11.477464784275401, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16652607917785645, "tokens_per_second": 4153231.940880494, "eta_seconds": 4956.90297027704, "mean_data_wait_seconds": 0.00010949373245239258}
{"step": 8170, "total_steps": 46238, "tokens": 4417830912, "loss": 3.2659039081949177, "ce_loss": 3.2647060228116587, "z_loss": 11.978980273911446, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18100033700466156, "tokens_per_second": 4151978.8044905374, "eta_seconds": 4957.096843013739, "mean_data_wait_seconds": 9.188055992126465e-05}
{"step": 8180, "total_steps": 46238, "tokens": 4423237632, "loss": 3.2695738980264375, "ce_loss": 3.2683677944270046, "z_loss": 12.0611321998365, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17982058227062225, "tokens_per_second": 4151844.0132707264, "eta_seconds": 4955.955531621821, "mean_data_wait_seconds": 7.638335227966309e-05}
{"step": 8190, "total_steps": 46238, "tokens": 4428644352, "loss": 3.2784366607666016, "ce_loss": 3.27727721676682, "z_loss": 11.594874049677994, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15443813800811768, "tokens_per_second": 4149344.0170207866, "eta_seconds": 4957.638484448889, "mean_data_wait_seconds": 8.177757263183594e-05}
{"step": 8200, "total_steps": 46238, "tokens": 4434051072, "loss": 3.33720265193419, "ce_loss": 3.3360307650132612, "z_loss": 11.71871151346149, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15869538486003876, "tokens_per_second": 4147913.453600592, "eta_seconds": 4958.0448324321, "mean_data_wait_seconds": 8.785724639892578e-05}
{"step": 8210, "total_steps": 46238, "tokens": 4439457792, "loss": 3.2956405950315073, "ce_loss": 3.2944905053485525, "z_loss": 11.501081719542995, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.196111798286438, "tokens_per_second": 4148068.9124303474, "eta_seconds": 4956.555587200659, "mean_data_wait_seconds": 8.90493392944336e-05}
{"step": 8220, "total_steps": 46238, "tokens": 4444864512, "loss": 3.2829059832023852, "ce_loss": 3.281765950448585, "z_loss": 11.40052699320244, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1950874626636505, "tokens_per_second": 4148423.9925346817, "eta_seconds": 4954.8280168539595, "mean_data_wait_seconds": 9.718537330627441e-05}
{"step": 8230, "total_steps": 46238, "tokens": 4450271232, "loss": 3.318156762556596, "ce_loss": 3.317012729066791, "z_loss": 11.440586017839836, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1852729171514511, "tokens_per_second": 4148654.1714920956, "eta_seconds": 4953.24986237869, "mean_data_wait_seconds": 9.331107139587402e-05}
{"step": 8240, "total_steps": 46238, "tokens": 4455677952, "loss": 3.2457069917158647, "ce_loss": 3.244503053751859, "z_loss": 12.039370356184063, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18174046277999878, "tokens_per_second": 4149060.5102817155, "eta_seconds": 4951.46164609807, "mean_data_wait_seconds": 9.691715240478516e-05}
{"step": 8250, "total_steps": 46238, "tokens": 4461084672, "loss": 3.304019180211154, "ce_loss": 3.3028488719102107, "z_loss": 11.703111619660348, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1707552820444107, "tokens_per_second": 4149723.862241281, "eta_seconds": 4949.367221969097, "mean_data_wait_seconds": 9.09268856048584e-05}
{"step": 8260, "total_steps": 46238, "tokens": 4466491392, "loss": 3.2638823010704736, "ce_loss": 3.2627291390390107, "z_loss": 11.531708161036173, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19343067705631256, "tokens_per_second": 4148285.7841225117, "eta_seconds": 4949.779646954428, "mean_data_wait_seconds": 9.322166442871094e-05}
{"step": 8270, "total_steps": 46238, "tokens": 4471898112, "loss": 3.3404090639316673, "ce_loss": 3.3392196965940073, "z_loss": 11.893690571640477, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15458612143993378, "tokens_per_second": 4148532.752835496, "eta_seconds": 4948.181693869827, "mean_data_wait_seconds": 9.569525718688965e-05}
{"step": 8280, "total_steps": 46238, "tokens": 4477304832, "loss": 3.284545063972473, "ce_loss": 3.283371672485814, "z_loss": 11.73434539274736, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16522175073623657, "tokens_per_second": 4148902.969735197, "eta_seconds": 4946.436986765644, "mean_data_wait_seconds": 0.0001437067985534668}
{"step": 8290, "total_steps": 46238, "tokens": 4482711552, "loss": 3.3081913283376982, "ce_loss": 3.3070240255558128, "z_loss": 11.673168594186956, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18105943500995636, "tokens_per_second": 4148876.6819599126, "eta_seconds": 4945.165151138671, "mean_data_wait_seconds": 0.00017240643501281738}
{"step": 8300, "total_steps": 46238, "tokens": 4488118272, "loss": 3.2181365038409377, "ce_loss": 3.2169669848499876, "z_loss": 11.695334579005387, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1696968525648117, "tokens_per_second": 4147035.553819102, "eta_seconds": 4946.056863464916, "mean_data_wait_seconds": 8.660554885864258e-05}
{"step": 8310, "total_steps": 46238, "tokens": 4493524992, "loss": 3.3058358683730615, "ce_loss": 3.3046697468468635, "z_loss": 11.661651553529682, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1805831342935562, "tokens_per_second": 4147927.710388335, "eta_seconds": 4943.689566393189, "mean_data_wait_seconds": 8.955597877502441e-05}
{"step": 8320, "total_steps": 46238, "tokens": 4498931712, "loss": 3.3133359786235923, "ce_loss": 3.312178293863932, "z_loss": 11.576979470975472, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1762242317199707, "tokens_per_second": 4148480.825670605, "eta_seconds": 4941.727125058135, "mean_data_wait_seconds": 8.133053779602051e-05}
{"step": 8330, "total_steps": 46238, "tokens": 4504338432, "loss": 3.2889718160484778, "ce_loss": 3.287784865408233, "z_loss": 11.869357506434122, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20171880722045898, "tokens_per_second": 4148579.69577819, "eta_seconds": 4940.306082309817, "mean_data_wait_seconds": 0.00010147690773010254}
{"step": 8340, "total_steps": 46238, "tokens": 4509745152, "loss": 3.31224863095717, "ce_loss": 3.311083419756456, "z_loss": 11.652232148430564, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17511510848999023, "tokens_per_second": 4149057.5930193583, "eta_seconds": 4938.433927375083, "mean_data_wait_seconds": 6.628036499023438e-05}
{"step": 8350, "total_steps": 46238, "tokens": 4515151872, "loss": 3.240559767593037, "ce_loss": 3.239410745375084, "z_loss": 11.490004033753365, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14997854828834534, "tokens_per_second": 4149213.7783670244, "eta_seconds": 4936.944963115859, "mean_data_wait_seconds": 6.607174873352051e-05}
{"step": 8360, "total_steps": 46238, "tokens": 4520558592, "loss": 3.2796933127172068, "ce_loss": 3.278514007727305, "z_loss": 11.793057918548584, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1622406244277954, "tokens_per_second": 4149572.7155608027, "eta_seconds": 4935.214960134111, "mean_data_wait_seconds": 0.00011092424392700195}
{"step": 8370, "total_steps": 46238, "tokens": 4525965312, "loss": 3.3284614808631665, "ce_loss": 3.3273125778545034, "z_loss": 11.489228198022554, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16390365362167358, "tokens_per_second": 4149531.1073547653, "eta_seconds": 4933.961475240389, "mean_data_wait_seconds": 8.240342140197754e-05}
{"step": 8380, "total_steps": 46238, "tokens": 4531372032, "loss": 3.2936178048451743, "ce_loss": 3.2924506772648203, "z_loss": 11.671392541943174, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17654582858085632, "tokens_per_second": 4150130.558851743, "eta_seconds": 4931.94602284106, "mean_data_wait_seconds": 0.00014981627464294434}
{"step": 8390, "total_steps": 46238, "tokens": 4536778752, "loss": 3.368427986448461, "ce_loss": 3.3672624862555303, "z_loss": 11.654765302484686, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16562888026237488, "tokens_per_second": 4150417.861667574, "eta_seconds": 4930.301927666232, "mean_data_wait_seconds": 7.781386375427246e-05}
{"step": 8400, "total_steps": 46238, "tokens": 4542185472, "loss": 3.3277387673204597, "ce_loss": 3.3265641096866494, "z_loss": 11.746600468953451, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18407964706420898, "tokens_per_second": 4150889.6042365385, "eta_seconds": 4928.439061140166, "mean_data_wait_seconds": 6.973743438720703e-05}
{"step": 8410, "total_steps": 46238, "tokens": 4547592192, "loss": 3.2809283950112085, "ce_loss": 3.2797152508388865, "z_loss": 12.131752982284084, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18496163189411163, "tokens_per_second": 4151369.5713973916, "eta_seconds": 4926.566857576994, "mean_data_wait_seconds": 6.371736526489258e-05}
{"step": 8420, "total_steps": 46238, "tokens": 4552998912, "loss": 3.2943513465650156, "ce_loss": 3.293152675484166, "z_loss": 11.98666711287065, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1682218760251999, "tokens_per_second": 4151715.1288518785, "eta_seconds": 4924.854521426265, "mean_data_wait_seconds": 8.246302604675293e-05}
{"step": 8430, "total_steps": 46238, "tokens": 4558405632, "loss": 3.2950083252155418, "ce_loss": 3.2938451658595693, "z_loss": 11.631683241237294, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19737166166305542, "tokens_per_second": 4151915.9263601545, "eta_seconds": 4923.314119686451, "mean_data_wait_seconds": 7.617473602294922e-05}
{"step": 8440, "total_steps": 46238, "tokens": 4563812352, "loss": 3.2984357133056177, "ce_loss": 3.2972329276980776, "z_loss": 12.02771027882894, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1701963245868683, "tokens_per_second": 4152671.850358845, "eta_seconds": 4921.115927383976, "mean_data_wait_seconds": 0.00010466575622558594}
{"step": 8450, "total_steps": 46238, "tokens": 4569219072, "loss": 3.2754589210857046, "ce_loss": 3.274306024565841, "z_loss": 11.528739315090757, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17008507251739502, "tokens_per_second": 4153254.363987888, "eta_seconds": 4919.1239142846725, "mean_data_wait_seconds": 7.474422454833984e-05}
{"step": 8460, "total_steps": 46238, "tokens": 4574625792, "loss": 3.260104571328019, "ce_loss": 3.2589322436939585, "z_loss": 11.722849694165317, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18695580959320068, "tokens_per_second": 4153688.973163511, "eta_seconds": 4917.307549015651, "mean_data_wait_seconds": 6.189942359924316e-05}
{"step": 8470, "total_steps": 46238, "tokens": 4580032512, "loss": 3.2843412774981875, "ce_loss": 3.2831800941264992, "z_loss": 11.611526734901197, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1767999231815338, "tokens_per_second": 4154503.066434218, "eta_seconds": 4915.042568863951, "mean_data_wait_seconds": 5.8770179748535156e-05}
{"step": 8480, "total_steps": 46238, "tokens": 4585439232, "loss": 3.2453066128672976, "ce_loss": 3.244086216796528, "z_loss": 12.203938874331387, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16960176825523376, "tokens_per_second": 4154883.1639148043, "eta_seconds": 4913.2916374874485, "mean_data_wait_seconds": 7.420778274536133e-05}
{"step": 8490, "total_steps": 46238, "tokens": 4590845952, "loss": 3.2821959618366128, "ce_loss": 3.281013609785022, "z_loss": 11.823780132062508, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15571328997612, "tokens_per_second": 4155482.841947493, "eta_seconds": 4911.281494892495, "mean_data_wait_seconds": 0.00010737776756286621}
{"step": 8500, "total_steps": 46238, "tokens": 4596252672, "loss": 3.2759574254353843, "ce_loss": 3.274773547143647, "z_loss": 11.838463624318441, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1680941879749298, "tokens_per_second": 4155485.143455093, "eta_seconds": 4909.977670389545, "mean_data_wait_seconds": 0.00010955333709716797}
{"step": 8510, "total_steps": 46238, "tokens": 4601659392, "loss": 3.253616549751975, "ce_loss": 3.2524215651280954, "z_loss": 11.950141242056182, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1787746250629425, "tokens_per_second": 4156241.2874453827, "eta_seconds": 4907.783531628768, "mean_data_wait_seconds": 7.998943328857422e-05}
{"step": 8520, "total_steps": 46238, "tokens": 4607066112, "loss": 3.274670051805901, "ce_loss": 3.273468568469539, "z_loss": 12.014676729838053, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17657959461212158, "tokens_per_second": 4156684.0660579437, "eta_seconds": 4905.960015224244, "mean_data_wait_seconds": 0.00010231137275695801}
{"step": 8530, "total_steps": 46238, "tokens": 4612472832, "loss": 3.295877090006164, "ce_loss": 3.2947101683327644, "z_loss": 11.669002402912486, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17816129326820374, "tokens_per_second": 4157165.6274311095, "eta_seconds": 4904.09113591129, "mean_data_wait_seconds": 6.344914436340332e-05}
{"step": 8540, "total_steps": 46238, "tokens": 4617879552, "loss": 3.256506916248437, "ce_loss": 3.2552993803313286, "z_loss": 12.075580148985892, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1624167412519455, "tokens_per_second": 4157718.5706064515, "eta_seconds": 4902.138525702833, "mean_data_wait_seconds": 8.258223533630371e-05}
{"step": 8550, "total_steps": 46238, "tokens": 4623286272, "loss": 3.264697793758277, "ce_loss": 3.263519605000814, "z_loss": 11.781461578426939, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17484231293201447, "tokens_per_second": 4158073.4721520054, "eta_seconds": 4900.419821936015, "mean_data_wait_seconds": 9.703636169433594e-05}
{"step": 8560, "total_steps": 46238, "tokens": 4628692992, "loss": 3.2191551255457327, "ce_loss": 3.217974046866099, "z_loss": 11.810719186609441, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1436523199081421, "tokens_per_second": 4158683.85703716, "eta_seconds": 4898.400466178541, "mean_data_wait_seconds": 0.00013840198516845703}
{"step": 8570, "total_steps": 46238, "tokens": 4634099712, "loss": 3.2950439019636675, "ce_loss": 3.293863888942834, "z_loss": 11.800603122422189, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17079496383666992, "tokens_per_second": 4158919.729964355, "eta_seconds": 4896.822623738051, "mean_data_wait_seconds": 9.161233901977539e-05}
{"step": 8580, "total_steps": 46238, "tokens": 4639506432, "loss": 3.2451988061269126, "ce_loss": 3.244014716509617, "z_loss": 11.840983845970847, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15598151087760925, "tokens_per_second": 4159285.701244584, "eta_seconds": 4895.091841829391, "mean_data_wait_seconds": 0.0020197033882141113}
{"step": 8590, "total_steps": 46238, "tokens": 4644913152, "loss": 3.261673477562991, "ce_loss": 3.2605364431034434, "z_loss": 11.37077817772374, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19430510699748993, "tokens_per_second": 4159515.848142515, "eta_seconds": 4893.521151768094, "mean_data_wait_seconds": 6.0677528381347656e-05}
{"step": 8600, "total_steps": 46238, "tokens": 4650319872, "loss": 3.26293154557546, "ce_loss": 3.261744045850002, "z_loss": 11.874942230455803, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15976305305957794, "tokens_per_second": 4160217.286387056, "eta_seconds": 4891.39644955236, "mean_data_wait_seconds": 7.179379463195801e-05}
{"step": 8610, "total_steps": 46238, "tokens": 4655726592, "loss": 3.2458785266587227, "ce_loss": 3.24469237797188, "z_loss": 11.86130191340591, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19191205501556396, "tokens_per_second": 4160844.2460351945, "eta_seconds": 4889.359981062824, "mean_data_wait_seconds": 6.625056266784668e-05}
{"step": 8620, "total_steps": 46238, "tokens": 4661133312, "loss": 3.1692697821241436, "ce_loss": 3.1680705854386995, "z_loss": 11.991817958427198, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18108253180980682, "tokens_per_second": 4161305.1493887873, "eta_seconds": 4887.519154173857, "mean_data_wait_seconds": 6.958842277526855e-05}
{"step": 8630, "total_steps": 46238, "tokens": 4666540032, "loss": 3.1835696588863027, "ce_loss": 3.182364156751922, "z_loss": 12.05542262395223, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17138859629631042, "tokens_per_second": 4161976.3284295113, "eta_seconds": 4885.431895686085, "mean_data_wait_seconds": 9.894371032714844e-05}
{"step": 8640, "total_steps": 46238, "tokens": 4671946752, "loss": 3.1900248924891152, "ce_loss": 3.1887982072252217, "z_loss": 12.26667419346896, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15825355052947998, "tokens_per_second": 4162388.231370263, "eta_seconds": 4883.649494969891, "mean_data_wait_seconds": 9.09268856048584e-05}
{"step": 8650, "total_steps": 46238, "tokens": 4677353472, "loss": 3.2340099197445493, "ce_loss": 3.2328156196709834, "z_loss": 11.943075851960616, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21382392942905426, "tokens_per_second": 4160309.53197251, "eta_seconds": 4884.790015699794, "mean_data_wait_seconds": 9.310245513916016e-05}
{"step": 8660, "total_steps": 46238, "tokens": 4682760192, "loss": 3.247596464373849, "ce_loss": 3.246367228753639, "z_loss": 12.292298475901285, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19147904217243195, "tokens_per_second": 4160252.719879235, "eta_seconds": 4883.557108662803, "mean_data_wait_seconds": 7.298588752746582e-05}
{"step": 8670, "total_steps": 46238, "tokens": 4688166912, "loss": 3.2170371937029287, "ce_loss": 3.2158666144717825, "z_loss": 11.705637072071884, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15408235788345337, "tokens_per_second": 4160507.1865034653, "eta_seconds": 4881.958884697887, "mean_data_wait_seconds": 8.624792098999023e-05}
{"step": 8680, "total_steps": 46238, "tokens": 4693573632, "loss": 3.220892971212214, "ce_loss": 3.219724956786994, "z_loss": 11.680472706303451, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15096034109592438, "tokens_per_second": 4158951.2255725097, "eta_seconds": 4882.485319650444, "mean_data_wait_seconds": 8.64267349243164e-05}
{"step": 8690, "total_steps": 46238, "tokens": 4698980352, "loss": 3.2463275808276553, "ce_loss": 3.2451770305633545, "z_loss": 11.505295110471321, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19410347938537598, "tokens_per_second": 4159216.4492899473, "eta_seconds": 4880.874037576399, "mean_data_wait_seconds": 0.00010901689529418945}
{"step": 8700, "total_steps": 46238, "tokens": 4704387072, "loss": 3.2228871421380476, "ce_loss": 3.221725005092043, "z_loss": 11.621506784901474, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17206449806690216, "tokens_per_second": 4159374.899030135, "eta_seconds": 4879.388214977291, "mean_data_wait_seconds": 8.958578109741211e-05}
{"step": 8710, "total_steps": 46238, "tokens": 4709793792, "loss": 3.204673292058887, "ce_loss": 3.203543863513253, "z_loss": 11.29443471359484, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2151210457086563, "tokens_per_second": 4159306.0516278576, "eta_seconds": 4878.169072472807, "mean_data_wait_seconds": 8.425116539001465e-05}
{"step": 8720, "total_steps": 46238, "tokens": 4715200512, "loss": 3.2236840454014866, "ce_loss": 3.222527628595179, "z_loss": 11.563719554380937, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1719006597995758, "tokens_per_second": 4158021.5367574557, "eta_seconds": 4878.375747860688, "mean_data_wait_seconds": 0.056555718183517456}
{"step": 8730, "total_steps": 46238, "tokens": 4720607232, "loss": 3.206611024610924, "ce_loss": 3.2054411552169104, "z_loss": 11.698514006354593, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16445520520210266, "tokens_per_second": 4157135.7635637377, "eta_seconds": 4878.114610001498, "mean_data_wait_seconds": 8.153915405273438e-05}
{"step": 8740, "total_steps": 46238, "tokens": 4726013952, "loss": 3.2622885487296363, "ce_loss": 3.261115648529746, "z_loss": 11.728701165228179, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16742806136608124, "tokens_per_second": 4157400.2428212846, "eta_seconds": 4876.503776370108, "mean_data_wait_seconds": 0.00012347102165222168}
{"step": 8750, "total_steps": 46238, "tokens": 4731420672, "loss": 3.283230797810988, "ce_loss": 3.282092031204339, "z_loss": 11.387771201856209, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21478933095932007, "tokens_per_second": 4157501.5735737, "eta_seconds": 4875.08444803255, "mean_data_wait_seconds": 7.107853889465332e-05}
{"step": 8760, "total_steps": 46238, "tokens": 4736827392, "loss": 3.2022052244706587, "ce_loss": 3.201026529976816, "z_loss": 11.78701569817283, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16221725940704346, "tokens_per_second": 4157701.2269205023, "eta_seconds": 4873.549934950012, "mean_data_wait_seconds": 7.086992263793945e-05}
{"step": 8770, "total_steps": 46238, "tokens": 4742234112, "loss": 3.3534488786350596, "ce_loss": 3.35231560829914, "z_loss": 11.33303552685362, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17407891154289246, "tokens_per_second": 4158025.5798423514, "eta_seconds": 4871.86945703399, "mean_data_wait_seconds": 6.896257400512695e-05}
{"step": 8780, "total_steps": 46238, "tokens": 4747640832, "loss": 3.2401293516159058, "ce_loss": 3.238979270963958, "z_loss": 11.500550306204593, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15500982105731964, "tokens_per_second": 4158174.207783822, "eta_seconds": 4870.395056101717, "mean_data_wait_seconds": 7.88271427154541e-05}
{"step": 8790, "total_steps": 46238, "tokens": 4753047552, "loss": 3.24244509863131, "ce_loss": 3.2412693283774634, "z_loss": 11.75804727005236, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20381996035575867, "tokens_per_second": 4158002.552289791, "eta_seconds": 4869.295804748925, "mean_data_wait_seconds": 0.00012412667274475098}
{"step": 8800, "total_steps": 46238, "tokens": 4758454272, "loss": 3.1988034013545876, "ce_loss": 3.1976282795270285, "z_loss": 11.75112397742994, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17866294085979462, "tokens_per_second": 4158533.3246435863, "eta_seconds": 4867.374163878992, "mean_data_wait_seconds": 8.639693260192871e-05}
{"step": 8810, "total_steps": 46238, "tokens": 4763860992, "loss": 3.309940825809132, "ce_loss": 3.308763249353929, "z_loss": 11.775619651332047, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16274519264698029, "tokens_per_second": 4159412.084525306, "eta_seconds": 4865.0459566834215, "mean_data_wait_seconds": 0.00013205409049987793}
{"step": 8820, "total_steps": 46238, "tokens": 4769267712, "loss": 3.2793900081605623, "ce_loss": 3.2782401713457974, "z_loss": 11.497927622361617, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21339401602745056, "tokens_per_second": 4158265.457423669, "eta_seconds": 4865.087241576462, "mean_data_wait_seconds": 0.0001271069049835205}
{"step": 8830, "total_steps": 46238, "tokens": 4774674432, "loss": 3.2070774869485335, "ce_loss": 3.2059029683922278, "z_loss": 11.745273944103356, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15693403780460358, "tokens_per_second": 4158956.204593355, "eta_seconds": 4862.9791969587495, "mean_data_wait_seconds": 8.657574653625488e-05}
{"step": 8840, "total_steps": 46238, "tokens": 4780081152, "loss": 3.271540656234279, "ce_loss": 3.2703619093605965, "z_loss": 11.78723602583914, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.169581338763237, "tokens_per_second": 4159693.028085271, "eta_seconds": 4860.818009281601, "mean_data_wait_seconds": 8.344650268554688e-05}
{"step": 8850, "total_steps": 46238, "tokens": 4785487872, "loss": 3.275193577462977, "ce_loss": 3.274039017431664, "z_loss": 11.54553551384897, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17237932980060577, "tokens_per_second": 4160297.6797040873, "eta_seconds": 4858.811945744657, "mean_data_wait_seconds": 0.0001067817211151123}
{"step": 8860, "total_steps": 46238, "tokens": 4790894592, "loss": 3.287651580391508, "ce_loss": 3.2864837917414578, "z_loss": 11.67824782747211, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2042796015739441, "tokens_per_second": 4161062.2567225113, "eta_seconds": 4856.619799752171, "mean_data_wait_seconds": 0.0001392662525177002}
{"step": 8870, "total_steps": 46238, "tokens": 4796301312, "loss": 3.2906770236564404, "ce_loss": 3.2895375743056787, "z_loss": 11.394723429824367, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22263331711292267, "tokens_per_second": 4161585.187686187, "eta_seconds": 4854.710335806652, "mean_data_wait_seconds": 9.28044319152832e-05}
{"step": 8880, "total_steps": 46238, "tokens": 4801708032, "loss": 3.350206721912731, "ce_loss": 3.349045583696076, "z_loss": 11.611085508808944, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1793460249900818, "tokens_per_second": 4161693.0307343476, "eta_seconds": 4853.285370842453, "mean_data_wait_seconds": 7.796287536621094e-05}
{"step": 8890, "total_steps": 46238, "tokens": 4807114752, "loss": 3.301748472632784, "ce_loss": 3.3005685047669844, "z_loss": 11.799396587140633, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18579578399658203, "tokens_per_second": 4162317.4698507804, "eta_seconds": 4851.258302679133, "mean_data_wait_seconds": 0.00011086463928222656}
{"step": 8900, "total_steps": 46238, "tokens": 4812521472, "loss": 3.3047327453439888, "ce_loss": 3.303569425236095, "z_loss": 11.633404623378407, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19292816519737244, "tokens_per_second": 4162589.3777894503, "eta_seconds": 4849.642525806948, "mean_data_wait_seconds": 6.35385513305664e-05}
{"step": 8910, "total_steps": 46238, "tokens": 4817928192, "loss": 3.320816982876171, "ce_loss": 3.319636390064702, "z_loss": 11.80605318329551, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1814039945602417, "tokens_per_second": 4163321.385425115, "eta_seconds": 4847.491191684511, "mean_data_wait_seconds": 6.961822509765625e-05}
{"step": 8920, "total_steps": 46238, "tokens": 4823334912, "loss": 3.285794243668065, "ce_loss": 3.2846021598035637, "z_loss": 11.920703201582938, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1698838472366333, "tokens_per_second": 4164021.745935675, "eta_seconds": 4845.377439177206, "mean_data_wait_seconds": 6.720423698425293e-05}
{"step": 8930, "total_steps": 46238, "tokens": 4828741632, "loss": 3.328879177570343, "ce_loss": 3.327690940914732, "z_loss": 11.882527625922, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15537405014038086, "tokens_per_second": 4164444.970527916, "eta_seconds": 4843.586707652663, "mean_data_wait_seconds": 9.050965309143066e-05}
{"step": 8940, "total_steps": 46238, "tokens": 4834148352, "loss": 3.250308961579294, "ce_loss": 3.2491426684639673, "z_loss": 11.662978692488236, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16369393467903137, "tokens_per_second": 4164588.0134640858, "eta_seconds": 4842.122082377718, "mean_data_wait_seconds": 0.00011685490608215332}
{"step": 8950, "total_steps": 46238, "tokens": 4839555072, "loss": 3.3166842966368706, "ce_loss": 3.315511253747073, "z_loss": 11.730132926594127, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16189703345298767, "tokens_per_second": 4164900.4327359996, "eta_seconds": 4840.460699982809, "mean_data_wait_seconds": 8.112192153930664e-05}
{"step": 8960, "total_steps": 46238, "tokens": 4844961792, "loss": 3.2636368545618923, "ce_loss": 3.262460314866268, "z_loss": 11.765376647313436, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17786003649234772, "tokens_per_second": 4165515.5135266, "eta_seconds": 4838.447985261908, "mean_data_wait_seconds": 8.335709571838379e-05}
{"step": 8970, "total_steps": 46238, "tokens": 4850368512, "loss": 3.274906721982089, "ce_loss": 3.2737120751178628, "z_loss": 11.946809869824033, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18998700380325317, "tokens_per_second": 4166228.0054137865, "eta_seconds": 4836.322783538774, "mean_data_wait_seconds": 8.037686347961426e-05}
{"step": 8980, "total_steps": 46238, "tokens": 4855775232, "loss": 3.3019450306892395, "ce_loss": 3.300744177717151, "z_loss": 12.008759411898525, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17867819964885712, "tokens_per_second": 4166546.967437197, "eta_seconds": 4834.654898031851, "mean_data_wait_seconds": 8.159875869750977e-05}
{"step": 8990, "total_steps": 46238, "tokens": 4861181952, "loss": 3.3328663490035315, "ce_loss": 3.331668367891601, "z_loss": 11.97957233949141, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16958677768707275, "tokens_per_second": 4167006.4396764175, "eta_seconds": 4832.824300977998, "mean_data_wait_seconds": 8.243322372436523e-05}
{"step": 9000, "total_steps": 46238, "tokens": 4866588672, "loss": 3.266312149437991, "ce_loss": 3.265186147256331, "z_loss": 11.259747873653065, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17912495136260986, "tokens_per_second": 4167754.304468012, "eta_seconds": 4830.659821385477, "mean_data_wait_seconds": 9.28640365600586e-05}
{"step": 9010, "total_steps": 46238, "tokens": 4871995392, "loss": 3.297692268183737, "ce_loss": 3.2965294899362507, "z_loss": 11.628070311112838, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18080994486808777, "tokens_per_second": 4167993.1198828476, "eta_seconds": 4829.085837014467, "mean_data_wait_seconds": 6.538629531860352e-05}
{"step": 9020, "total_steps": 46238, "tokens": 4877402112, "loss": 3.275618470076359, "ce_loss": 3.274461699254585, "z_loss": 11.567785718224265, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15404696762561798, "tokens_per_second": 4168497.705424415, "eta_seconds": 4827.204246224063, "mean_data_wait_seconds": 6.267428398132324e-05}
{"step": 9030, "total_steps": 46238, "tokens": 4882808832, "loss": 3.2591370597030176, "ce_loss": 3.2579239245617027, "z_loss": 12.131615140221335, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18591928482055664, "tokens_per_second": 4168855.8058673977, "eta_seconds": 4825.492662923701, "mean_data_wait_seconds": 8.168816566467285e-05}
{"step": 9040, "total_steps": 46238, "tokens": 4888215552, "loss": 3.2406414674990103, "ce_loss": 3.2394342296051257, "z_loss": 12.072381893793741, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16575665771961212, "tokens_per_second": 4169087.395796898, "eta_seconds": 4823.927750777175, "mean_data_wait_seconds": 5.894899368286133e-05}
{"step": 9050, "total_steps": 46238, "tokens": 4893622272, "loss": 3.2144798076514043, "ce_loss": 3.2133227424188093, "z_loss": 11.571023760419903, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15045380592346191, "tokens_per_second": 4169861.862982799, "eta_seconds": 4821.735185639395, "mean_data_wait_seconds": 6.705522537231445e-05}
{"step": 9060, "total_steps": 46238, "tokens": 4899028992, "loss": 3.276612321535746, "ce_loss": 3.275420768694444, "z_loss": 11.915647448915424, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15750889480113983, "tokens_per_second": 4170193.213410774, "eta_seconds": 4820.055550270267, "mean_data_wait_seconds": 7.387995719909668e-05}
{"step": 9070, "total_steps": 46238, "tokens": 4904435712, "loss": 3.245385746161143, "ce_loss": 3.2442327268195874, "z_loss": 11.52993072162975, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17540812492370605, "tokens_per_second": 4170877.115647536, "eta_seconds": 4817.9688988224225, "mean_data_wait_seconds": 5.888938903808594e-05}
{"step": 9080, "total_steps": 46238, "tokens": 4909842432, "loss": 3.273968344384974, "ce_loss": 3.272787392139435, "z_loss": 11.80960935534853, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1492665708065033, "tokens_per_second": 4171563.252266846, "eta_seconds": 4815.88035206781, "mean_data_wait_seconds": 5.8591365814208984e-05}
{"step": 9090, "total_steps": 46238, "tokens": 4915249152, "loss": 3.259914203123613, "ce_loss": 3.258727429491101, "z_loss": 11.867399302395908, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19955849647521973, "tokens_per_second": 4171786.7912321067, "eta_seconds": 4814.3262800993325, "mean_data_wait_seconds": 0.0001278519630432129}
{"step": 9100, "total_steps": 46238, "tokens": 4920655872, "loss": 3.2697802312446362, "ce_loss": 3.268596537185438, "z_loss": 11.83725314429312, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18672478199005127, "tokens_per_second": 4171612.5454487456, "eta_seconds": 4813.2312973088165, "mean_data_wait_seconds": 7.429718971252441e-05}
{"step": 9110, "total_steps": 46238, "tokens": 4926062592, "loss": 3.230864026329734, "ce_loss": 3.229690367525274, "z_loss": 11.736664042328343, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17893226444721222, "tokens_per_second": 4171713.8863975136, "eta_seconds": 4811.818329500662, "mean_data_wait_seconds": 6.914138793945312e-05}
{"step": 9120, "total_steps": 46238, "tokens": 4931469312, "loss": 3.2203943115292173, "ce_loss": 3.2191863836664143, "z_loss": 12.079450795144746, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16858334839344025, "tokens_per_second": 4172116.387086635, "eta_seconds": 4810.058196390215, "mean_data_wait_seconds": 9.72747802734375e-05}
{"step": 9130, "total_steps": 46238, "tokens": 4936876032, "loss": 3.181619714606892, "ce_loss": 3.180441397609133, "z_loss": 11.783008727160366, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15849840641021729, "tokens_per_second": 4172327.9431098895, "eta_seconds": 4808.51845242204, "mean_data_wait_seconds": 0.0001119375228881836}
{"step": 9140, "total_steps": 46238, "tokens": 4942282752, "loss": 3.188897925795931, "ce_loss": 3.187729654890118, "z_loss": 11.682828823725382, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18387264013290405, "tokens_per_second": 4172465.728832241, "eta_seconds": 4807.063853251466, "mean_data_wait_seconds": 0.00016561150550842285}
{"step": 9150, "total_steps": 46238, "tokens": 4947689472, "loss": 3.1771178209420405, "ce_loss": 3.175942879734617, "z_loss": 11.749596003330115, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1790837049484253, "tokens_per_second": 4172634.5762612624, "eta_seconds": 4805.57357648289, "mean_data_wait_seconds": 0.00012409687042236328}
{"step": 9160, "total_steps": 46238, "tokens": 4953096192, "loss": 3.231586248585672, "ce_loss": 3.230442803917509, "z_loss": 11.434512080568256, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1901606172323227, "tokens_per_second": 4172764.684520067, "eta_seconds": 4804.128020534582, "mean_data_wait_seconds": 0.00012090802192687988}
{"step": 9170, "total_steps": 46238, "tokens": 4958502912, "loss": 3.2150617187673394, "ce_loss": 3.213911108898394, "z_loss": 11.50613192355994, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15497605502605438, "tokens_per_second": 4172616.6381121087, "eta_seconds": 4803.0027108044, "mean_data_wait_seconds": 8.285045623779297e-05}
{"step": 9180, "total_steps": 46238, "tokens": 4963909632, "loss": 3.214933236440023, "ce_loss": 3.213733004801201, "z_loss": 12.002185828758009, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17032746970653534, "tokens_per_second": 4172449.822954562, "eta_seconds": 4801.898921294275, "mean_data_wait_seconds": 0.00011226534843444824}
{"step": 9190, "total_steps": 46238, "tokens": 4969316352, "loss": 3.231642464796702, "ce_loss": 3.2304631850936194, "z_loss": 11.792689699115176, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1705877184867859, "tokens_per_second": 4172486.1682371297, "eta_seconds": 4800.561290407529, "mean_data_wait_seconds": 8.606910705566406e-05}
{"step": 9200, "total_steps": 46238, "tokens": 4974723072, "loss": 3.2412048506014273, "ce_loss": 3.239992826273947, "z_loss": 12.12059364896832, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1816314309835434, "tokens_per_second": 4172652.0039225924, "eta_seconds": 4799.074747948112, "mean_data_wait_seconds": 0.0001207888126373291}
{"step": 9210, "total_steps": 46238, "tokens": 4980129792, "loss": 3.1461202971863025, "ce_loss": 3.1449173056718074, "z_loss": 12.029917146220352, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17395931482315063, "tokens_per_second": 4172657.364063666, "eta_seconds": 4797.772833306268, "mean_data_wait_seconds": 8.64267349243164e-05}
{"step": 9220, "total_steps": 46238, "tokens": 4985536512, "loss": 3.113785373442101, "ce_loss": 3.1126321481935904, "z_loss": 11.532239270932747, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16770265996456146, "tokens_per_second": 4172810.6421374246, "eta_seconds": 4796.300896545899, "mean_data_wait_seconds": 0.00012159347534179688}
{"step": 9230, "total_steps": 46238, "tokens": 4990943232, "loss": 3.2326763940579966, "ce_loss": 3.231517528042649, "z_loss": 11.589169849048961, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21179863810539246, "tokens_per_second": 4172500.533136936, "eta_seconds": 4795.3615691828945, "mean_data_wait_seconds": 0.0001347064971923828}
{"step": 9240, "total_steps": 46238, "tokens": 4996349952, "loss": 3.1850605299978545, "ce_loss": 3.18392618858453, "z_loss": 11.343309980450254, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18292637169361115, "tokens_per_second": 4172480.9170390703, "eta_seconds": 4794.088309023342, "mean_data_wait_seconds": 8.100271224975586e-05}
{"step": 9250, "total_steps": 46238, "tokens": 5001756672, "loss": 3.212024141441692, "ce_loss": 3.210831739685752, "z_loss": 11.923696597417196, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1998092085123062, "tokens_per_second": 4173036.6484098216, "eta_seconds": 4792.15423895699, "mean_data_wait_seconds": 0.0001768171787261963}
{"step": 9260, "total_steps": 46238, "tokens": 5007163392, "loss": 3.2940601724566836, "ce_loss": 3.2928897232720344, "z_loss": 11.705049225778291, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2078394591808319, "tokens_per_second": 4173449.5629819077, "eta_seconds": 4790.384606856376, "mean_data_wait_seconds": 9.119510650634766e-05}
{"step": 9270, "total_steps": 46238, "tokens": 5012570112, "loss": 3.220407738830104, "ce_loss": 3.219231887297197, "z_loss": 11.758775862780483, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16638725996017456, "tokens_per_second": 4173572.6579285683, "eta_seconds": 4788.947854071858, "mean_data_wait_seconds": 9.694695472717285e-05}
{"step": 9280, "total_steps": 46238, "tokens": 5017976832, "loss": 3.2941647497090427, "ce_loss": 3.2930073232361763, "z_loss": 11.574162056951812, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1676868498325348, "tokens_per_second": 4173614.681024764, "eta_seconds": 4787.604182735392, "mean_data_wait_seconds": 9.22083854675293e-05}
{"step": 9290, "total_steps": 46238, "tokens": 5023383552, "loss": 3.249428265022509, "ce_loss": 3.2482660430850405, "z_loss": 11.622177210721103, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18253998458385468, "tokens_per_second": 4173988.757784687, "eta_seconds": 4785.87977668685, "mean_data_wait_seconds": 7.027387619018555e-05}
{"step": 9300, "total_steps": 46238, "tokens": 5028790272, "loss": 3.256240431106452, "ce_loss": 3.2550248720429162, "z_loss": 12.155423605080808, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1784856617450714, "tokens_per_second": 4174516.88450959, "eta_seconds": 4783.979132556823, "mean_data_wait_seconds": 6.92605972290039e-05}
{"step": 9310, "total_steps": 46238, "tokens": 5034196992, "loss": 3.2343141454638857, "ce_loss": 3.2331208568630796, "z_loss": 11.932913036057442, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17452473938465118, "tokens_per_second": 4175035.5277652894, "eta_seconds": 4782.089831625119, "mean_data_wait_seconds": 7.557868957519531e-05}
{"step": 9320, "total_steps": 46238, "tokens": 5039603712, "loss": 3.28002462423209, "ce_loss": 3.278907223181291, "z_loss": 11.174225641019417, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17474213242530823, "tokens_per_second": 4175518.641646193, "eta_seconds": 4780.241674632017, "mean_data_wait_seconds": 6.407499313354492e-05}
{"step": 9330, "total_steps": 46238, "tokens": 5045010432, "loss": 3.229884483597495, "ce_loss": 3.228696483554262, "z_loss": 11.879801497314915, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1774875819683075, "tokens_per_second": 4175910.985226151, "eta_seconds": 4778.497811518685, "mean_data_wait_seconds": 6.064772605895996e-05}
{"step": 9340, "total_steps": 46238, "tokens": 5050417152, "loss": 3.279262262763399, "ce_loss": 3.2780539122494785, "z_loss": 12.08340426647302, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16148445010185242, "tokens_per_second": 4176500.6892617526, "eta_seconds": 4776.528550633679, "mean_data_wait_seconds": 6.4849853515625e-05}
{"step": 9350, "total_steps": 46238, "tokens": 5055823872, "loss": 3.251033354889263, "ce_loss": 3.2498795715245334, "z_loss": 11.538061821099484, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18777617812156677, "tokens_per_second": 4176930.901406414, "eta_seconds": 4774.742157521623, "mean_data_wait_seconds": 7.647275924682617e-05}
{"step": 9360, "total_steps": 46238, "tokens": 5061230592, "loss": 3.2536231532241358, "ce_loss": 3.252443644133481, "z_loss": 11.795067346457278, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20924106240272522, "tokens_per_second": 4177570.8819148154, "eta_seconds": 4772.716467915711, "mean_data_wait_seconds": 7.066130638122559e-05}
{"step": 9370, "total_steps": 46238, "tokens": 5066637312, "loss": 3.2571246353062717, "ce_loss": 3.255926226124619, "z_loss": 11.984436945481734, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16529597342014313, "tokens_per_second": 4177985.5071229525, "eta_seconds": 4770.948723976366, "mean_data_wait_seconds": 0.00010824203491210938}
{"step": 9380, "total_steps": 46238, "tokens": 5072044032, "loss": 3.264230925025362, "ce_loss": 3.263044713121472, "z_loss": 11.862395799521243, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17480646073818207, "tokens_per_second": 4178498.7605144805, "eta_seconds": 4769.06876036656, "mean_data_wait_seconds": 7.483363151550293e-05}
{"step": 9390, "total_steps": 46238, "tokens": 5077450752, "loss": 3.3260784528472205, "ce_loss": 3.324914664933176, "z_loss": 11.638149738311768, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16860733926296234, "tokens_per_second": 4179112.264741927, "eta_seconds": 4767.074900590223, "mean_data_wait_seconds": 6.67273998260498e-05}
{"step": 9400, "total_steps": 46238, "tokens": 5082857472, "loss": 3.264982378844059, "ce_loss": 3.2638021978464993, "z_loss": 11.801577510255756, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1848120540380478, "tokens_per_second": 4179871.2768813213, "eta_seconds": 4764.91574612801, "mean_data_wait_seconds": 7.125735282897949e-05}
{"step": 9410, "total_steps": 46238, "tokens": 5088264192, "loss": 3.265590994647055, "ce_loss": 3.264431259848855, "z_loss": 11.597310716455633, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1796000450849533, "tokens_per_second": 4180593.2714592717, "eta_seconds": 4762.799548077008, "mean_data_wait_seconds": 6.940960884094238e-05}
{"step": 9420, "total_steps": 46238, "tokens": 5093670912, "loss": 3.2784679423679006, "ce_loss": 3.2772767688288833, "z_loss": 11.911601268883908, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15537667274475098, "tokens_per_second": 4181197.03355298, "eta_seconds": 4760.818699587784, "mean_data_wait_seconds": 0.0001061558723449707}
{"step": 9430, "total_steps": 46238, "tokens": 5099077632, "loss": 3.280204176902771, "ce_loss": 3.279033601284027, "z_loss": 11.705716082544038, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18120697140693665, "tokens_per_second": 4181583.994746726, "eta_seconds": 4759.085152660039, "mean_data_wait_seconds": 6.93202018737793e-05}
{"step": 9440, "total_steps": 46238, "tokens": 5104484352, "loss": 3.249185834870194, "ce_loss": 3.2479507561885947, "z_loss": 12.35063169219277, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.23604795336723328, "tokens_per_second": 4181903.962560568, "eta_seconds": 4757.4281385023205, "mean_data_wait_seconds": 6.523728370666504e-05}
{"step": 9450, "total_steps": 46238, "tokens": 5109891072, "loss": 3.310957681049, "ce_loss": 3.3097616668903465, "z_loss": 11.960032679817893, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19189463555812836, "tokens_per_second": 4182388.766186377, "eta_seconds": 4755.583943989981, "mean_data_wait_seconds": 5.924701690673828e-05}
{"step": 9460, "total_steps": 46238, "tokens": 5115297792, "loss": 3.326032716216463, "ce_loss": 3.324829784306613, "z_loss": 12.029422023079611, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1837274432182312, "tokens_per_second": 4182868.505374271, "eta_seconds": 4753.745932594362, "mean_data_wait_seconds": 7.432699203491211e-05}
{"step": 9470, "total_steps": 46238, "tokens": 5120704512, "loss": 3.2851020368662747, "ce_loss": 3.283864559549274, "z_loss": 12.374801202253861, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22820143401622772, "tokens_per_second": 4183402.565114139, "eta_seconds": 4751.846640285652, "mean_data_wait_seconds": 8.422136306762695e-05}
{"step": 9480, "total_steps": 46238, "tokens": 5126111232, "loss": 3.25462939341863, "ce_loss": 3.2534369880502876, "z_loss": 11.923982844208226, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1866021454334259, "tokens_per_second": 4183669.6095008124, "eta_seconds": 4750.25098991296, "mean_data_wait_seconds": 8.744001388549805e-05}
{"step": 9490, "total_steps": 46238, "tokens": 5131517952, "loss": 3.3403966842275676, "ce_loss": 3.3392180421135644, "z_loss": 11.786567305073593, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1978183090686798, "tokens_per_second": 4183854.9145520325, "eta_seconds": 4748.748316987776, "mean_data_wait_seconds": 0.00010198354721069336}
{"step": 9500, "total_steps": 46238, "tokens": 5136924672, "loss": 3.261923237280412, "ce_loss": 3.2607586401881594, "z_loss": 11.645847775719382, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1537354588508606, "tokens_per_second": 4184347.7073202366, "eta_seconds": 4746.896924758808, "mean_data_wait_seconds": 0.00015464425086975098}
{"step": 9510, "total_steps": 46238, "tokens": 5142331392, "loss": 3.311097797119256, "ce_loss": 3.309898909294244, "z_loss": 11.988936091914322, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1850462108850479, "tokens_per_second": 4184846.8434215626, "eta_seconds": 4745.038776082079, "mean_data_wait_seconds": 0.0004252791404724121}
{"step": 9520, "total_steps": 46238, "tokens": 5147738112, "loss": 3.2883018760970146, "ce_loss": 3.2871073484420776, "z_loss": 11.94495565963514, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17431490123271942, "tokens_per_second": 4185360.243993712, "eta_seconds": 4743.164904977729, "mean_data_wait_seconds": 0.0001659691333770752}
{"step": 9530, "total_steps": 46238, "tokens": 5153144832, "loss": 3.2738016283873357, "ce_loss": 3.2726219224207327, "z_loss": 11.797010407303318, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15538707375526428, "tokens_per_second": 4185738.2796664233, "eta_seconds": 4741.444824778112, "mean_data_wait_seconds": 7.492303848266602e-05}
{"step": 9540, "total_steps": 46238, "tokens": 5158551552, "loss": 3.2237214500253852, "ce_loss": 3.222573878187122, "z_loss": 11.475823633598559, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20981931686401367, "tokens_per_second": 4185684.9943499537, "eta_seconds": 4740.213468233378, "mean_data_wait_seconds": 9.584426879882812e-05}
{"step": 9550, "total_steps": 46238, "tokens": 5163958272, "loss": 3.291440001039794, "ce_loss": 3.2902524868647256, "z_loss": 11.875172485004772, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16277514398097992, "tokens_per_second": 4185700.153292841, "eta_seconds": 4738.904588852487, "mean_data_wait_seconds": 0.00035250186920166016}
{"step": 9560, "total_steps": 46238, "tokens": 5169364992, "loss": 3.308047319903518, "ce_loss": 3.3068771434552744, "z_loss": 11.70179693626635, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17939580976963043, "tokens_per_second": 4185450.8467883472, "eta_seconds": 4737.8950726936555, "mean_data_wait_seconds": 7.370114326477051e-05}
{"step": 9570, "total_steps": 46238, "tokens": 5174771712, "loss": 3.3023203484939807, "ce_loss": 3.3011129408171684, "z_loss": 12.07378167817087, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19770631194114685, "tokens_per_second": 4185610.8137990274, "eta_seconds": 4736.422258524844, "mean_data_wait_seconds": 0.00010463595390319824}
{"step": 9580, "total_steps": 46238, "tokens": 5180178432, "loss": 3.3073752775336756, "ce_loss": 3.306165088306774, "z_loss": 12.102119229056619, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16309021413326263, "tokens_per_second": 4185979.1363010644, "eta_seconds": 4734.713876647126, "mean_data_wait_seconds": 5.739927291870117e-05}
{"step": 9590, "total_steps": 46238, "tokens": 5185585152, "loss": 3.315443667498502, "ce_loss": 3.3142550190289817, "z_loss": 11.886686151677912, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19156943261623383, "tokens_per_second": 4186477.482426728, "eta_seconds": 4732.858797681778, "mean_data_wait_seconds": 8.928775787353516e-05}
{"step": 9600, "total_steps": 46238, "tokens": 5190991872, "loss": 3.2839278416200117, "ce_loss": 3.2827304113994944, "z_loss": 11.974339528517289, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17626507580280304, "tokens_per_second": 4186868.811059493, "eta_seconds": 4731.125086049067, "mean_data_wait_seconds": 0.00012040138244628906}
{"step": 9610, "total_steps": 46238, "tokens": 5196398592, "loss": 3.2938076239643674, "ce_loss": 3.2926003264658377, "z_loss": 12.072804429314353, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19962745904922485, "tokens_per_second": 4187457.7038412085, "eta_seconds": 4729.168565890057, "mean_data_wait_seconds": 0.00010383129119873047}
{"step": 9620, "total_steps": 46238, "tokens": 5201805312, "loss": 3.2393344568483755, "ce_loss": 3.2381309379230845, "z_loss": 12.035684318253487, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18534278869628906, "tokens_per_second": 4187766.828429849, "eta_seconds": 4727.528402392674, "mean_data_wait_seconds": 7.593631744384766e-05}
{"step": 9630, "total_steps": 46238, "tokens": 5207212032, "loss": 3.3192085497307056, "ce_loss": 3.3180703900077124, "z_loss": 11.381427533698805, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18276292085647583, "tokens_per_second": 4188192.0181822856, "eta_seconds": 4725.757514955123, "mean_data_wait_seconds": 0.0001665055751800537}
{"step": 9640, "total_steps": 46238, "tokens": 5212618752, "loss": 3.306330939133962, "ce_loss": 3.305154614376299, "z_loss": 11.763041546850493, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16450762748718262, "tokens_per_second": 4188188.4393007713, "eta_seconds": 4724.470608419779, "mean_data_wait_seconds": 9.658932685852051e-05}
{"step": 9650, "total_steps": 46238, "tokens": 5218025472, "loss": 3.221184062235283, "ce_loss": 3.2200262094988967, "z_loss": 11.57840465776848, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.170515775680542, "tokens_per_second": 4188226.8715401436, "eta_seconds": 4723.136322537774, "mean_data_wait_seconds": 8.022785186767578e-05}
{"step": 9660, "total_steps": 46238, "tokens": 5223432192, "loss": 3.207830168984153, "ce_loss": 3.20667258717797, "z_loss": 11.575759013493856, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.166733056306839, "tokens_per_second": 4188584.0399664515, "eta_seconds": 4721.442748981682, "mean_data_wait_seconds": 7.677078247070312e-05}
{"step": 9670, "total_steps": 46238, "tokens": 5228838912, "loss": 3.2137344587932932, "ce_loss": 3.2125539201678652, "z_loss": 11.80509834578543, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22400154173374176, "tokens_per_second": 4188609.42087535, "eta_seconds": 4720.12332433427, "mean_data_wait_seconds": 7.456541061401367e-05}
{"step": 9680, "total_steps": 46238, "tokens": 5234245632, "loss": 3.228918644514951, "ce_loss": 3.227743974237731, "z_loss": 11.746877359621452, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1846984326839447, "tokens_per_second": 4188535.2947903597, "eta_seconds": 4718.916020257452, "mean_data_wait_seconds": 7.50422477722168e-05}
{"step": 9690, "total_steps": 46238, "tokens": 5239652352, "loss": 3.244665212703474, "ce_loss": 3.243503314076048, "z_loss": 11.619230487129904, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2012806236743927, "tokens_per_second": 4188846.4287747755, "eta_seconds": 4717.274772419795, "mean_data_wait_seconds": 0.00013265013694763184}
{"step": 9700, "total_steps": 46238, "tokens": 5245059072, "loss": 3.2095106334397285, "ce_loss": 3.208369088895393, "z_loss": 11.41539977536057, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18844975531101227, "tokens_per_second": 4189009.499840601, "eta_seconds": 4715.800445129497, "mean_data_wait_seconds": 9.498000144958496e-05}
{"step": 9710, "total_steps": 46238, "tokens": 5250465792, "loss": 3.2036186799858557, "ce_loss": 3.2024692983338325, "z_loss": 11.493887417244188, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16582384705543518, "tokens_per_second": 4189305.5056085587, "eta_seconds": 4714.176638003665, "mean_data_wait_seconds": 6.347894668579102e-05}
{"step": 9720, "total_steps": 46238, "tokens": 5255872512, "loss": 3.2419275803999468, "ce_loss": 3.2407458532940256, "z_loss": 11.817301909128824, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21033215522766113, "tokens_per_second": 4189784.1774271103, "eta_seconds": 4712.347602621468, "mean_data_wait_seconds": 6.690621376037598e-05}
{"step": 9730, "total_steps": 46238, "tokens": 5261279232, "loss": 3.1750622807127056, "ce_loss": 3.173886243141059, "z_loss": 11.76016853794907, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17173421382904053, "tokens_per_second": 4190037.261675935, "eta_seconds": 4710.772594920803, "mean_data_wait_seconds": 6.777048110961914e-05}
{"step": 9740, "total_steps": 46238, "tokens": 5266685952, "loss": 3.209207012797847, "ce_loss": 3.2080363739620554, "z_loss": 11.70654042561849, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2397216409444809, "tokens_per_second": 4190544.663152528, "eta_seconds": 4708.911984046251, "mean_data_wait_seconds": 6.261467933654785e-05}
{"step": 9750, "total_steps": 46238, "tokens": 5272092672, "loss": 3.243627333279812, "ce_loss": 3.242432065082319, "z_loss": 11.952576117082076, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1893896460533142, "tokens_per_second": 4191005.8286034837, "eta_seconds": 4707.103752841486, "mean_data_wait_seconds": 6.052851676940918e-05}
{"step": 9760, "total_steps": 46238, "tokens": 5277499392, "loss": 3.2192251736467536, "ce_loss": 3.218010804869912, "z_loss": 12.143450570828987, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17193199694156647, "tokens_per_second": 4191462.431173669, "eta_seconds": 4705.301041784009, "mean_data_wait_seconds": 7.051229476928711e-05}
{"step": 9770, "total_steps": 46238, "tokens": 5282906112, "loss": 3.1777393041235027, "ce_loss": 3.1765343005006965, "z_loss": 12.050207304232048, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17786861956119537, "tokens_per_second": 4192088.965726952, "eta_seconds": 4703.308060777503, "mean_data_wait_seconds": 6.210803985595703e-05}
{"step": 9780, "total_steps": 46238, "tokens": 5288312832, "loss": 3.2422455964666423, "ce_loss": 3.2410437482776064, "z_loss": 12.01822549646551, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18257124722003937, "tokens_per_second": 4192623.5780633106, "eta_seconds": 4701.418750572688, "mean_data_wait_seconds": 6.893277168273926e-05}
{"step": 9790, "total_steps": 46238, "tokens": 5293719552, "loss": 3.277913669745127, "ce_loss": 3.2767361691503814, "z_loss": 11.775113575386278, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19167867302894592, "tokens_per_second": 4193327.0567080756, "eta_seconds": 4699.340670906284, "mean_data_wait_seconds": 6.73830509185791e-05}
{"step": 9800, "total_steps": 46238, "tokens": 5299126272, "loss": 3.2696981971914116, "ce_loss": 3.268517194372235, "z_loss": 11.810045076139046, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17472122609615326, "tokens_per_second": 4193839.180314175, "eta_seconds": 4697.477613465419, "mean_data_wait_seconds": 6.157159805297852e-05}
{"step": 9810, "total_steps": 46238, "tokens": 5304532992, "loss": 3.2244164130904456, "ce_loss": 3.2232856714364253, "z_loss": 11.307452743703669, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18707288801670074, "tokens_per_second": 4194359.833781604, "eta_seconds": 4695.605461738146, "mean_data_wait_seconds": 0.00012230873107910156}
{"step": 9820, "total_steps": 46238, "tokens": 5309939712, "loss": 3.2402974421327766, "ce_loss": 3.239131571668567, "z_loss": 11.658681443243315, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17503070831298828, "tokens_per_second": 4195266.552818635, "eta_seconds": 4693.301838180293, "mean_data_wait_seconds": 0.00013405084609985352}
{"step": 9830, "total_steps": 46238, "tokens": 5315346432, "loss": 3.259061332904931, "ce_loss": 3.257913482911659, "z_loss": 11.478361411528153, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16801971197128296, "tokens_per_second": 4192907.108277945, "eta_seconds": 4694.653374299163, "mean_data_wait_seconds": 7.328391075134277e-05}
{"step": 9840, "total_steps": 46238, "tokens": 5320753152, "loss": 3.2388079925016924, "ce_loss": 3.2376133420250635, "z_loss": 11.946383013869777, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19121094048023224, "tokens_per_second": 4191240.155831919, "eta_seconds": 4695.230540921831, "mean_data_wait_seconds": 0.00011169910430908203}
{"step": 9850, "total_steps": 46238, "tokens": 5326159872, "loss": 3.2826490998268127, "ce_loss": 3.281475928696719, "z_loss": 11.731809211499764, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18282988667488098, "tokens_per_second": 4191649.647099439, "eta_seconds": 4693.481974957934, "mean_data_wait_seconds": 0.0001258552074432373}
{"step": 9860, "total_steps": 46238, "tokens": 5331566592, "loss": 3.2970911715969895, "ce_loss": 3.2958947889732593, "z_loss": 11.963836287007187, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16715063154697418, "tokens_per_second": 4192048.041139175, "eta_seconds": 4691.74617060335, "mean_data_wait_seconds": 0.0001118779182434082}
{"step": 9870, "total_steps": 46238, "tokens": 5336973312, "loss": 3.343761895642136, "ce_loss": 3.3425915765039846, "z_loss": 11.703047051574245, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19326896965503693, "tokens_per_second": 4191799.1765805623, "eta_seconds": 4690.734883926304, "mean_data_wait_seconds": 8.51750373840332e-05}
{"step": 9880, "total_steps": 46238, "tokens": 5342380032, "loss": 3.2800482367024277, "ce_loss": 3.278865458387317, "z_loss": 11.827781265432185, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19998985528945923, "tokens_per_second": 4192116.3229256882, "eta_seconds": 4689.090280367312, "mean_data_wait_seconds": 8.115172386169434e-05}
{"step": 9890, "total_steps": 46238, "tokens": 5347786752, "loss": 3.210891064369317, "ce_loss": 3.209751869692947, "z_loss": 11.39179928375013, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17597250640392303, "tokens_per_second": 4192240.975373085, "eta_seconds": 4687.661157706018, "mean_data_wait_seconds": 0.0001176595687866211}
{"step": 9900, "total_steps": 46238, "tokens": 5353193472, "loss": 3.268458026828188, "ce_loss": 3.2672155011783945, "z_loss": 12.425135388518825, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21482545137405396, "tokens_per_second": 4192366.743120744, "eta_seconds": 4686.230873345654, "mean_data_wait_seconds": 7.736682891845703e-05}
{"step": 9910, "total_steps": 46238, "tokens": 5358600192, "loss": 3.1878540461713616, "ce_loss": 3.186639047030247, "z_loss": 12.14999734994137, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19456249475479126, "tokens_per_second": 4192636.512014608, "eta_seconds": 4684.639769680937, "mean_data_wait_seconds": 0.00015717744827270508}
{"step": 9920, "total_steps": 46238, "tokens": 5364006912, "loss": 3.2535355054970942, "ce_loss": 3.2523358157186797, "z_loss": 11.997309822024722, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2173006236553192, "tokens_per_second": 4192871.496824658, "eta_seconds": 4683.0877213552585, "mean_data_wait_seconds": 7.82310962677002e-05}
{"step": 9930, "total_steps": 46238, "tokens": 5369413632, "loss": 3.337105680595745, "ce_loss": 3.3359561110987808, "z_loss": 11.495755766377304, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18437756597995758, "tokens_per_second": 4192973.291531706, "eta_seconds": 4681.684556313745, "mean_data_wait_seconds": 8.00788402557373e-05}
{"step": 9940, "total_steps": 46238, "tokens": 5374820352, "loss": 3.2831530119433547, "ce_loss": 3.281987797130238, "z_loss": 11.652245709390352, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18064163625240326, "tokens_per_second": 4193321.5222499557, "eta_seconds": 4680.006405392495, "mean_data_wait_seconds": 6.920099258422852e-05}
{"step": 9950, "total_steps": 46238, "tokens": 5380227072, "loss": 3.2599985978820105, "ce_loss": 3.2588500615322227, "z_loss": 11.485687682122895, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1664622724056244, "tokens_per_second": 4193457.3213611557, "eta_seconds": 4678.565527318099, "mean_data_wait_seconds": 8.624792098999023e-05}
{"step": 9960, "total_steps": 46238, "tokens": 5385633792, "loss": 3.2631574164737356, "ce_loss": 3.2619873321417607, "z_loss": 11.701003616506403, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17166881263256073, "tokens_per_second": 4193650.342526062, "eta_seconds": 4677.060923535521, "mean_data_wait_seconds": 6.732344627380371e-05}
{"step": 9970, "total_steps": 46238, "tokens": 5391040512, "loss": 3.2834468122684592, "ce_loss": 3.2822508179780208, "z_loss": 11.960082668246645, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1687512844800949, "tokens_per_second": 4193648.84044534, "eta_seconds": 4675.773334878867, "mean_data_wait_seconds": 8.696317672729492e-05}
{"step": 9980, "total_steps": 46238, "tokens": 5396447232, "loss": 3.2686219558571326, "ce_loss": 3.267458529183359, "z_loss": 11.633972189643167, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1795347034931183, "tokens_per_second": 4193804.5030133408, "eta_seconds": 4674.310566912384, "mean_data_wait_seconds": 0.00030991435050964355}
{"step": 9990, "total_steps": 46238, "tokens": 5401853952, "loss": 3.238166343082081, "ce_loss": 3.2369540857546255, "z_loss": 12.122610287232833, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2070358693599701, "tokens_per_second": 4193800.4780182517, "eta_seconds": 4673.02583580723, "mean_data_wait_seconds": 0.00011333823204040527}
{"step": 10000, "total_steps": 46238, "tokens": 5407260672, "loss": 3.2470476211923542, "ce_loss": 3.2458397366783838, "z_loss": 12.078502799525406, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20072464644908905, "tokens_per_second": 4194172.8994386825, "eta_seconds": 4671.321791865589, "mean_data_wait_seconds": 0.0003879368305206299}
{"step": 10010, "total_steps": 46238, "tokens": 5412667392, "loss": 3.2896106748869927, "ce_loss": 3.288429713610447, "z_loss": 11.809454715613162, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16353391110897064, "tokens_per_second": 4194245.382341974, "eta_seconds": 4669.95198384485, "mean_data_wait_seconds": 7.286667823791504e-05}
{"step": 10020, "total_steps": 46238, "tokens": 5418074112, "loss": 3.3215773665543757, "ce_loss": 3.3203965877041672, "z_loss": 11.807836496468747, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18289129436016083, "tokens_per_second": 4194247.68730522, "eta_seconds": 4668.660337649495, "mean_data_wait_seconds": 6.285309791564941e-05}
{"step": 10030, "total_steps": 46238, "tokens": 5423480832, "loss": 3.27565008763111, "ce_loss": 3.274451161875869, "z_loss": 11.989022356091123, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2358812838792801, "tokens_per_second": 4194537.973059935, "eta_seconds": 4667.048249349651, "mean_data_wait_seconds": 7.852911949157715e-05}
{"step": 10040, "total_steps": 46238, "tokens": 5428887552, "loss": 3.3416595296426252, "ce_loss": 3.340482103102135, "z_loss": 11.774510976040002, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1722194254398346, "tokens_per_second": 4194674.518971006, "eta_seconds": 4665.607377995297, "mean_data_wait_seconds": 6.368756294250488e-05}
{"step": 10050, "total_steps": 46238, "tokens": 5434294272, "loss": 3.3380067763906536, "ce_loss": 3.3367991158456514, "z_loss": 12.07661062298399, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20922721922397614, "tokens_per_second": 4194970.099550032, "eta_seconds": 4663.989778162816, "mean_data_wait_seconds": 0.00011548399925231934}
{"step": 10060, "total_steps": 46238, "tokens": 5439700992, "loss": 3.257473522966558, "ce_loss": 3.2562717217387576, "z_loss": 12.018326217477972, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1897994577884674, "tokens_per_second": 4195220.061999192, "eta_seconds": 4662.423104136024, "mean_data_wait_seconds": 8.696317672729492e-05}
{"step": 10070, "total_steps": 46238, "tokens": 5445107712, "loss": 3.282131527409409, "ce_loss": 3.280944253459121, "z_loss": 11.872809171676636, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17537878453731537, "tokens_per_second": 4195407.078218772, "eta_seconds": 4660.92654644187, "mean_data_wait_seconds": 0.00011372566223144531}
{"step": 10080, "total_steps": 46238, "tokens": 5450514432, "loss": 3.248959920623086, "ce_loss": 3.2477894028027854, "z_loss": 11.704795237743493, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.192342147231102, "tokens_per_second": 4195418.328791563, "eta_seconds": 4659.625327429711, "mean_data_wait_seconds": 0.0001226663589477539}
{"step": 10090, "total_steps": 46238, "tokens": 5455921152, "loss": 3.248728228337837, "ce_loss": 3.247578781662565, "z_loss": 11.49484447999434, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18606528639793396, "tokens_per_second": 4195919.695450963, "eta_seconds": 4657.7799868735365, "mean_data_wait_seconds": 8.681416511535645e-05}
{"step": 10100, "total_steps": 46238, "tokens": 5461327872, "loss": 3.324061186024637, "ce_loss": 3.3229018865209636, "z_loss": 11.593522830442948, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17755942046642303, "tokens_per_second": 4196002.801990447, "eta_seconds": 4656.399193711617, "mean_data_wait_seconds": 7.221102714538574e-05}
{"step": 10110, "total_steps": 46238, "tokens": 5466734592, "loss": 3.271805743376414, "ce_loss": 3.2706333886493337, "z_loss": 11.723219893195413, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1514316201210022, "tokens_per_second": 4196083.972823561, "eta_seconds": 4655.020602663551, "mean_data_wait_seconds": 7.367134094238281e-05}
{"step": 10120, "total_steps": 46238, "tokens": 5472141312, "loss": 3.2044803445989434, "ce_loss": 3.203314358537847, "z_loss": 11.660178632447213, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1673964411020279, "tokens_per_second": 4195998.963442628, "eta_seconds": 4653.82636986607, "mean_data_wait_seconds": 7.015466690063477e-05}
{"step": 10130, "total_steps": 46238, "tokens": 5477548032, "loss": 3.287219417817665, "ce_loss": 3.2860587401823564, "z_loss": 11.606964421994759, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19508713483810425, "tokens_per_second": 4196020.978343096, "eta_seconds": 4652.513418011739, "mean_data_wait_seconds": 6.783008575439453e-05}
{"step": 10140, "total_steps": 46238, "tokens": 5482954752, "loss": 3.282773079294147, "ce_loss": 3.281586074467861, "z_loss": 11.870159734379161, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1898808479309082, "tokens_per_second": 4196721.138440661, "eta_seconds": 4650.448895742362, "mean_data_wait_seconds": 8.079409599304199e-05}
{"step": 10150, "total_steps": 46238, "tokens": 5488361472, "loss": 3.2903032194484365, "ce_loss": 3.289130539605112, "z_loss": 11.726574218634402, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19812250137329102, "tokens_per_second": 4196742.54445196, "eta_seconds": 4649.1368620630765, "mean_data_wait_seconds": 9.506940841674805e-05}
{"step": 10160, "total_steps": 46238, "tokens": 5493768192, "loss": 3.2984572034893613, "ce_loss": 3.297314409053687, "z_loss": 11.428179119572494, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21615903079509735, "tokens_per_second": 4197189.51918016, "eta_seconds": 4647.35358145326, "mean_data_wait_seconds": 0.00012984871864318848}
{"step": 10170, "total_steps": 46238, "tokens": 5499174912, "loss": 3.2329040241963938, "ce_loss": 3.2317434964757976, "z_loss": 11.604995741988674, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15304823219776154, "tokens_per_second": 4197355.62243592, "eta_seconds": 4645.881544981648, "mean_data_wait_seconds": 7.510185241699219e-05}
{"step": 10180, "total_steps": 46238, "tokens": 5504581632, "loss": 3.2550056287736604, "ce_loss": 3.253839763728055, "z_loss": 11.658874836834995, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1530553549528122, "tokens_per_second": 4197509.900633589, "eta_seconds": 4644.422709058374, "mean_data_wait_seconds": 7.361173629760742e-05}
{"step": 10190, "total_steps": 46238, "tokens": 5509988352, "loss": 3.257415843732429, "ce_loss": 3.2562776388544026, "z_loss": 11.382203708995473, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17333558201789856, "tokens_per_second": 4198105.915611736, "eta_seconds": 4642.475434343593, "mean_data_wait_seconds": 6.452202796936035e-05}
{"step": 10200, "total_steps": 46238, "tokens": 5515395072, "loss": 3.230503564531153, "ce_loss": 3.2293136444958774, "z_loss": 11.899312330014778, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15673156082630157, "tokens_per_second": 4198381.177249607, "eta_seconds": 4640.883245566629, "mean_data_wait_seconds": 0.00010946393013000488}
{"step": 10210, "total_steps": 46238, "tokens": 5520801792, "loss": 3.2330458236463144, "ce_loss": 3.2318599404710713, "z_loss": 11.859272068197077, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18763358891010284, "tokens_per_second": 4198776.082189583, "eta_seconds": 4639.159069859752, "mean_data_wait_seconds": 6.461143493652344e-05}
{"step": 10220, "total_steps": 46238, "tokens": 5526208512, "loss": 3.2488424452868374, "ce_loss": 3.24767954602386, "z_loss": 11.628625653006814, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20416121184825897, "tokens_per_second": 4199240.629010889, "eta_seconds": 4637.358309372917, "mean_data_wait_seconds": 7.042288780212402e-05}
{"step": 10230, "total_steps": 46238, "tokens": 5531615232, "loss": 3.262112082857074, "ce_loss": 3.2609760869633067, "z_loss": 11.359574751420455, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17608362436294556, "tokens_per_second": 4199604.61054993, "eta_seconds": 4635.668952047062, "mean_data_wait_seconds": 8.362531661987305e-05}
{"step": 10240, "total_steps": 46238, "tokens": 5537021952, "loss": 3.216387410958608, "ce_loss": 3.2152362194928257, "z_loss": 11.512478113174438, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20534706115722656, "tokens_per_second": 4199791.83113461, "eta_seconds": 4634.174922603728, "mean_data_wait_seconds": 0.00010377168655395508}
{"step": 10250, "total_steps": 46238, "tokens": 5542428672, "loss": 3.237369730617061, "ce_loss": 3.2361874453949206, "z_loss": 11.82295736399564, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18149788677692413, "tokens_per_second": 4200010.264593598, "eta_seconds": 4632.646598039378, "mean_data_wait_seconds": 6.198883056640625e-05}
{"step": 10260, "total_steps": 46238, "tokens": 5547835392, "loss": 3.1901890472932295, "ce_loss": 3.189001168265487, "z_loss": 11.879059228030117, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17162853479385376, "tokens_per_second": 4200463.660748327, "eta_seconds": 4630.859380065343, "mean_data_wait_seconds": 0.00012424588203430176}
{"step": 10270, "total_steps": 46238, "tokens": 5553242112, "loss": 3.204520897431807, "ce_loss": 3.20334115895358, "z_loss": 11.797516136458427, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21226459741592407, "tokens_per_second": 4200953.661607033, "eta_seconds": 4629.03221278594, "mean_data_wait_seconds": 0.00011533498764038086}
{"step": 10280, "total_steps": 46238, "tokens": 5558648832, "loss": 3.2022030516104265, "ce_loss": 3.200988592523517, "z_loss": 12.145189162456628, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17698726058006287, "tokens_per_second": 4201312.622929182, "eta_seconds": 4627.349794894732, "mean_data_wait_seconds": 0.00010710954666137695}
{"step": 10290, "total_steps": 46238, "tokens": 5564055552, "loss": 3.236116262999448, "ce_loss": 3.234934026544744, "z_loss": 11.822450695615826, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18103580176830292, "tokens_per_second": 4201661.100831058, "eta_seconds": 4625.679205815955, "mean_data_wait_seconds": 9.134411811828613e-05}
{"step": 10300, "total_steps": 46238, "tokens": 5569462272, "loss": 3.123806458531004, "ce_loss": 3.1226112553567598, "z_loss": 11.952296141422156, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17086626589298248, "tokens_per_second": 4201449.195318677, "eta_seconds": 4624.625637660778, "mean_data_wait_seconds": 0.0001035928726196289}
{"step": 10310, "total_steps": 46238, "tokens": 5574868992, "loss": 3.252909961974982, "ce_loss": 3.251759496602145, "z_loss": 11.5048533930923, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19741198420524597, "tokens_per_second": 4201916.114605119, "eta_seconds": 4622.825019396053, "mean_data_wait_seconds": 7.164478302001953e-05}
{"step": 10320, "total_steps": 46238, "tokens": 5580275712, "loss": 3.1663130395340198, "ce_loss": 3.1651362812880315, "z_loss": 11.767499692512281, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17927198112010956, "tokens_per_second": 4202315.1070479, "eta_seconds": 4621.099496187459, "mean_data_wait_seconds": 8.01384449005127e-05}
{"step": 10330, "total_steps": 46238, "tokens": 5585682432, "loss": 3.2058020450852136, "ce_loss": 3.204618634599628, "z_loss": 11.833601575909238, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17236016690731049, "tokens_per_second": 4202756.633418932, "eta_seconds": 4619.3275503099585, "mean_data_wait_seconds": 7.611513137817383e-05}
{"step": 10340, "total_steps": 46238, "tokens": 5591089152, "loss": 3.165681385632717, "ce_loss": 3.1645485722657405, "z_loss": 11.327805548003226, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19633977115154266, "tokens_per_second": 4202944.110250507, "eta_seconds": 4617.8350877102675, "mean_data_wait_seconds": 6.73532485961914e-05}
{"step": 10350, "total_steps": 46238, "tokens": 5596495872, "loss": 3.1834172985770484, "ce_loss": 3.1822481588883833, "z_loss": 11.691698623426033, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17370182275772095, "tokens_per_second": 4203196.34025783, "eta_seconds": 4616.27164026551, "mean_data_wait_seconds": 8.040666580200195e-05}
{"step": 10360, "total_steps": 46238, "tokens": 5601902592, "loss": 3.1919506174145322, "ce_loss": 3.1907428033424146, "z_loss": 12.078310713623509, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18151187896728516, "tokens_per_second": 4203478.984012356, "eta_seconds": 4614.674991305483, "mean_data_wait_seconds": 8.884072303771973e-05}
{"step": 10370, "total_steps": 46238, "tokens": 5607309312, "loss": 3.185994693727204, "ce_loss": 3.184788330034776, "z_loss": 12.06376184839191, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15993261337280273, "tokens_per_second": 4203729.401192575, "eta_seconds": 4613.113921771111, "mean_data_wait_seconds": 8.937716484069824e-05}
{"step": 10380, "total_steps": 46238, "tokens": 5612716032, "loss": 3.250509065209013, "ce_loss": 3.249313898158796, "z_loss": 11.95191775668751, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1697932481765747, "tokens_per_second": 4203681.3987631835, "eta_seconds": 4611.880412655452, "mean_data_wait_seconds": 7.924437522888184e-05}
{"step": 10390, "total_steps": 46238, "tokens": 5618122752, "loss": 3.2399102250734964, "ce_loss": 3.238724065549446, "z_loss": 11.861731233018817, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17793788015842438, "tokens_per_second": 4203721.537949991, "eta_seconds": 4610.550201536819, "mean_data_wait_seconds": 7.724761962890625e-05}
{"step": 10400, "total_steps": 46238, "tokens": 5623529472, "loss": 3.243667645887895, "ce_loss": 3.242444578445319, "z_loss": 12.230711062749227, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1816655397415161, "tokens_per_second": 4201360.959171597, "eta_seconds": 4611.853790306195, "mean_data_wait_seconds": 7.110834121704102e-05}
{"step": 10410, "total_steps": 46238, "tokens": 5628936192, "loss": 3.273011153394526, "ce_loss": 3.2718058372988845, "z_loss": 12.053391124262955, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20096129179000854, "tokens_per_second": 4201878.418471916, "eta_seconds": 4609.999103935155, "mean_data_wait_seconds": 8.353590965270996e-05}
{"step": 10420, "total_steps": 46238, "tokens": 5634342912, "loss": 3.2446960680412524, "ce_loss": 3.2435530369932, "z_loss": 11.430361950036252, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19898556172847748, "tokens_per_second": 4200315.008828541, "eta_seconds": 4610.42778536768, "mean_data_wait_seconds": 0.00015285611152648926}
{"step": 10430, "total_steps": 46238, "tokens": 5639749632, "loss": 3.1531325524503533, "ce_loss": 3.1520046664006784, "z_loss": 11.279137734210853, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15451067686080933, "tokens_per_second": 4200365.47763953, "eta_seconds": 4609.085187244136, "mean_data_wait_seconds": 0.00011727213859558105}
{"step": 10440, "total_steps": 46238, "tokens": 5645156352, "loss": 3.2392369400371206, "ce_loss": 3.2380741961074597, "z_loss": 11.627351551344901, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20037971436977386, "tokens_per_second": 4200330.6691665, "eta_seconds": 4607.836170155773, "mean_data_wait_seconds": 8.612871170043945e-05}
{"step": 10450, "total_steps": 46238, "tokens": 5650563072, "loss": 3.2379899115273445, "ce_loss": 3.2367974411357534, "z_loss": 11.924639557347152, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1915852427482605, "tokens_per_second": 4200758.523339061, "eta_seconds": 4606.079772616879, "mean_data_wait_seconds": 5.6356191635131836e-05}
{"step": 10460, "total_steps": 46238, "tokens": 5655969792, "loss": 3.202043358123664, "ce_loss": 3.200869498830853, "z_loss": 11.738640344504153, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1639224737882614, "tokens_per_second": 4198998.819789014, "eta_seconds": 4606.722453180388, "mean_data_wait_seconds": 7.769465446472168e-05}
{"step": 10470, "total_steps": 46238, "tokens": 5661376512, "loss": 3.2178539128014534, "ce_loss": 3.2166369629628733, "z_loss": 12.16966413006638, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18840070068836212, "tokens_per_second": 4199099.814806684, "eta_seconds": 4605.324063936376, "mean_data_wait_seconds": 8.067488670349121e-05}
{"step": 10480, "total_steps": 46238, "tokens": 5666783232, "loss": 3.30626473643563, "ce_loss": 3.3050501960696597, "z_loss": 12.14536781022043, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.217402845621109, "tokens_per_second": 4199670.991308554, "eta_seconds": 4603.410301428443, "mean_data_wait_seconds": 8.475780487060547e-05}
{"step": 10490, "total_steps": 46238, "tokens": 5672189952, "loss": 3.2497960581923975, "ce_loss": 3.2486446843002783, "z_loss": 11.51332339373502, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17813503742218018, "tokens_per_second": 4199828.208454104, "eta_seconds": 4601.950609573656, "mean_data_wait_seconds": 0.0001831352710723877}
{"step": 10500, "total_steps": 46238, "tokens": 5677596672, "loss": 3.2316024321498293, "ce_loss": 3.230435346112107, "z_loss": 11.670825127399329, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18702934682369232, "tokens_per_second": 4200113.329910682, "eta_seconds": 4600.350930152376, "mean_data_wait_seconds": 0.00015527009963989258}
{"step": 10510, "total_steps": 46238, "tokens": 5683003392, "loss": 3.2524656960458467, "ce_loss": 3.2512889540556706, "z_loss": 11.767561616319599, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16580495238304138, "tokens_per_second": 4200111.1538775, "eta_seconds": 4599.066033327979, "mean_data_wait_seconds": 9.363889694213867e-05}
{"step": 10520, "total_steps": 46238, "tokens": 5688410112, "loss": 3.288448761809956, "ce_loss": 3.2872762860673848, "z_loss": 11.72501557523554, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1839449107646942, "tokens_per_second": 4200508.720168332, "eta_seconds": 4597.343586332591, "mean_data_wait_seconds": 6.985664367675781e-05}
{"step": 10530, "total_steps": 46238, "tokens": 5693816832, "loss": 3.2709269144318323, "ce_loss": 3.2697753689505835, "z_loss": 11.515164353630759, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18222391605377197, "tokens_per_second": 4200875.156265688, "eta_seconds": 4595.655520780011, "mean_data_wait_seconds": 0.00010779500007629395}
{"step": 10540, "total_steps": 46238, "tokens": 5699223552, "loss": 3.259448970809127, "ce_loss": 3.2582825097170742, "z_loss": 11.664614287289707, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18907497823238373, "tokens_per_second": 4201265.693763723, "eta_seconds": 4593.941395482103, "mean_data_wait_seconds": 7.385015487670898e-05}
{"step": 10550, "total_steps": 46238, "tokens": 5704630272, "loss": 3.2712661786512895, "ce_loss": 3.2701018217838174, "z_loss": 11.643201741305264, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16580413281917572, "tokens_per_second": 4201387.476950895, "eta_seconds": 4592.52134440194, "mean_data_wait_seconds": 9.709596633911133e-05}
{"step": 10560, "total_steps": 46238, "tokens": 5710036992, "loss": 3.2811096400925606, "ce_loss": 3.279935638109843, "z_loss": 11.7399110288331, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1851486712694168, "tokens_per_second": 4201912.004704451, "eta_seconds": 4590.661328081944, "mean_data_wait_seconds": 9.608268737792969e-05}
{"step": 10570, "total_steps": 46238, "tokens": 5715443712, "loss": 3.2450872153946846, "ce_loss": 3.2438965891346787, "z_loss": 11.906383341008967, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21952787041664124, "tokens_per_second": 4201714.203797366, "eta_seconds": 4589.5906500665005, "mean_data_wait_seconds": 7.912516593933105e-05}
{"step": 10580, "total_steps": 46238, "tokens": 5720850432, "loss": 3.2473690256927954, "ce_loss": 3.2462005723606455, "z_loss": 11.684749270930435, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18524084985256195, "tokens_per_second": 4202028.084095934, "eta_seconds": 4587.961126906132, "mean_data_wait_seconds": 9.098649024963379e-05}
{"step": 10590, "total_steps": 46238, "tokens": 5726257152, "loss": 3.2308603525161743, "ce_loss": 3.2296945452690125, "z_loss": 11.658049785729611, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15011654794216156, "tokens_per_second": 4202149.995165922, "eta_seconds": 4586.541367198149, "mean_data_wait_seconds": 0.00010001659393310547}
{"step": 10600, "total_steps": 46238, "tokens": 5731663872, "loss": 3.2941217639229516, "ce_loss": 3.2929503321647644, "z_loss": 11.714457006165475, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16657666862010956, "tokens_per_second": 4202264.5756943645, "eta_seconds": 4585.129688274387, "mean_data_wait_seconds": 0.00010201334953308105}
{"step": 10610, "total_steps": 46238, "tokens": 5737070592, "loss": 3.27069886525472, "ce_loss": 3.2695645545468186, "z_loss": 11.343114643385917, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17459310591220856, "tokens_per_second": 4202403.639628297, "eta_seconds": 4583.69138136949, "mean_data_wait_seconds": 8.165836334228516e-05}
{"step": 10620, "total_steps": 46238, "tokens": 5742477312, "loss": 3.29477093075261, "ce_loss": 3.2935979077310273, "z_loss": 11.730707486470541, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22245930135250092, "tokens_per_second": 4202884.3586329315, "eta_seconds": 4581.8806754568295, "mean_data_wait_seconds": 8.14199447631836e-05}
{"step": 10630, "total_steps": 46238, "tokens": 5747884032, "loss": 3.2850415923378686, "ce_loss": 3.2839035138939368, "z_loss": 11.380680358771121, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1978612095117569, "tokens_per_second": 4203544.065521099, "eta_seconds": 4579.875363246236, "mean_data_wait_seconds": 6.619095802307129e-05}
{"step": 10640, "total_steps": 46238, "tokens": 5753290752, "loss": 3.265259457357002, "ce_loss": 3.264035754131548, "z_loss": 12.237166180755153, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18691986799240112, "tokens_per_second": 4203909.93539464, "eta_seconds": 4578.190655788458, "mean_data_wait_seconds": 8.359551429748535e-05}
{"step": 10650, "total_steps": 46238, "tokens": 5758697472, "loss": 3.2892641659938926, "ce_loss": 3.288084676771453, "z_loss": 11.79473210103584, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1635485589504242, "tokens_per_second": 4204324.760258148, "eta_seconds": 4576.452952891916, "mean_data_wait_seconds": 0.00014713406562805176}
{"step": 10660, "total_steps": 46238, "tokens": 5764104192, "loss": 3.2721620346560623, "ce_loss": 3.2709485762047046, "z_loss": 12.134994145595666, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2012052834033966, "tokens_per_second": 4204405.343273453, "eta_seconds": 4575.079273641987, "mean_data_wait_seconds": 0.00021532177925109863}
{"step": 10670, "total_steps": 46238, "tokens": 5769510912, "loss": 3.2464135198882134, "ce_loss": 3.245252681500984, "z_loss": 11.608980518398862, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1746329516172409, "tokens_per_second": 4204625.553881181, "eta_seconds": 4573.553763009695, "mean_data_wait_seconds": 9.912252426147461e-05}
{"step": 10680, "total_steps": 46238, "tokens": 5774917632, "loss": 3.270671687342904, "ce_loss": 3.2694756731842505, "z_loss": 11.96024503852382, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15716801583766937, "tokens_per_second": 4204910.377661392, "eta_seconds": 4571.958157807876, "mean_data_wait_seconds": 0.00010097026824951172}
{"step": 10690, "total_steps": 46238, "tokens": 5780324352, "loss": 3.3003408601789763, "ce_loss": 3.2991729974746704, "z_loss": 11.678369189753678, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19789065420627594, "tokens_per_second": 4205009.735964797, "eta_seconds": 4570.564348429583, "mean_data_wait_seconds": 7.724761962890625e-05}
{"step": 10700, "total_steps": 46238, "tokens": 5785731072, "loss": 3.313475509484609, "ce_loss": 3.3122975609519263, "z_loss": 11.779264970259232, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18994982540607452, "tokens_per_second": 4205571.207523862, "eta_seconds": 4568.66853891951, "mean_data_wait_seconds": 0.00013139843940734863}
{"step": 10710, "total_steps": 46238, "tokens": 5791137792, "loss": 3.2612003503423748, "ce_loss": 3.259997674913117, "z_loss": 12.02616363583189, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18649977445602417, "tokens_per_second": 4206099.486023509, "eta_seconds": 4566.809274918001, "mean_data_wait_seconds": 7.230043411254883e-05}
{"step": 10720, "total_steps": 46238, "tokens": 5796544512, "loss": 3.2769321365789934, "ce_loss": 3.2757117874694592, "z_loss": 12.203813473383585, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21540123224258423, "tokens_per_second": 4206425.511027119, "eta_seconds": 4565.1699700040635, "mean_data_wait_seconds": 6.35385513305664e-05}
{"step": 10730, "total_steps": 46238, "tokens": 5801951232, "loss": 3.2242495345346853, "ce_loss": 3.2230544126395024, "z_loss": 11.950702002554229, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18656666576862335, "tokens_per_second": 4207050.215586115, "eta_seconds": 4563.206931278674, "mean_data_wait_seconds": 0.00011676549911499023}
{"step": 10740, "total_steps": 46238, "tokens": 5807357952, "loss": 3.242366566802516, "ce_loss": 3.2411596901489026, "z_loss": 12.068149689472083, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17368699610233307, "tokens_per_second": 4207210.624493366, "eta_seconds": 4561.74784125792, "mean_data_wait_seconds": 9.554624557495117e-05}
{"step": 10750, "total_steps": 46238, "tokens": 5812764672, "loss": 3.268236517906189, "ce_loss": 3.267042931282159, "z_loss": 11.935938994089762, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19862906634807587, "tokens_per_second": 4207898.676275411, "eta_seconds": 4559.717032203606, "mean_data_wait_seconds": 0.00021278858184814453}
{"step": 10760, "total_steps": 46238, "tokens": 5818171392, "loss": 3.261079105463895, "ce_loss": 3.2598894093975876, "z_loss": 11.896798343369454, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16358759999275208, "tokens_per_second": 4208125.121944163, "eta_seconds": 4558.186838118098, "mean_data_wait_seconds": 0.00010269880294799805}
{"step": 10770, "total_steps": 46238, "tokens": 5823578112, "loss": 3.237129070542075, "ce_loss": 3.2359297275543213, "z_loss": 11.993257919947306, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17541025578975677, "tokens_per_second": 4208354.338511437, "eta_seconds": 4556.653808477274, "mean_data_wait_seconds": 0.00020614266395568848}
{"step": 10780, "total_steps": 46238, "tokens": 5828984832, "loss": 3.2359375285379812, "ce_loss": 3.2347622622143137, "z_loss": 11.75309070673856, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15795420110225677, "tokens_per_second": 4208553.147522314, "eta_seconds": 4555.153857397819, "mean_data_wait_seconds": 0.00013008713722229004}
{"step": 10790, "total_steps": 46238, "tokens": 5834391552, "loss": 3.2334837769017075, "ce_loss": 3.2322926593549326, "z_loss": 11.911111434300741, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18096409738063812, "tokens_per_second": 4209017.858668422, "eta_seconds": 4553.3663737086545, "mean_data_wait_seconds": 0.0001342296600341797}
{"step": 10800, "total_steps": 46238, "tokens": 5839798272, "loss": 3.197823844172738, "ce_loss": 3.1966497428489453, "z_loss": 11.7412290428624, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17879532277584076, "tokens_per_second": 4209131.766962017, "eta_seconds": 4551.958628234814, "mean_data_wait_seconds": 9.712576866149902e-05}
{"step": 10810, "total_steps": 46238, "tokens": 5845204992, "loss": 3.2532535990079245, "ce_loss": 3.2520365335724573, "z_loss": 12.170667106455022, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18749983608722687, "tokens_per_second": 4209528.787926364, "eta_seconds": 4550.244910769586, "mean_data_wait_seconds": 0.00010573863983154297}
{"step": 10820, "total_steps": 46238, "tokens": 5850611712, "loss": 3.198821150895321, "ce_loss": 3.197674223870942, "z_loss": 11.468883991241455, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17488060891628265, "tokens_per_second": 4209649.438787664, "eta_seconds": 4548.830134775952, "mean_data_wait_seconds": 6.064772605895996e-05}
{"step": 10830, "total_steps": 46238, "tokens": 5856018432, "loss": 3.231472367590124, "ce_loss": 3.2302689480059072, "z_loss": 12.034158432122434, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1877109557390213, "tokens_per_second": 4209952.710176416, "eta_seconds": 4547.218180794672, "mean_data_wait_seconds": 7.778406143188477e-05}
{"step": 10840, "total_steps": 46238, "tokens": 5861425152, "loss": 3.2312995195388794, "ce_loss": 3.2301137591853286, "z_loss": 11.857594779043486, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1820105016231537, "tokens_per_second": 4210382.15240057, "eta_seconds": 4545.470242668656, "mean_data_wait_seconds": 7.393956184387207e-05}
{"step": 10850, "total_steps": 46238, "tokens": 5866831872, "loss": 3.2080376310782, "ce_loss": 3.206868796637564, "z_loss": 11.688504695892334, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17424428462982178, "tokens_per_second": 4210381.819860899, "eta_seconds": 4544.186461605067, "mean_data_wait_seconds": 7.638335227966309e-05}
{"step": 10860, "total_steps": 46238, "tokens": 5872238592, "loss": 3.205844673243436, "ce_loss": 3.2046715147567517, "z_loss": 11.731687242334539, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18113207817077637, "tokens_per_second": 4210735.900572327, "eta_seconds": 4542.520308956017, "mean_data_wait_seconds": 6.29723072052002e-05}
{"step": 10870, "total_steps": 46238, "tokens": 5877645312, "loss": 3.1898762547608577, "ce_loss": 3.188668243812792, "z_loss": 12.08008725715406, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1849588006734848, "tokens_per_second": 4210865.88069615, "eta_seconds": 4541.096098942652, "mean_data_wait_seconds": 6.794929504394531e-05}
{"step": 10880, "total_steps": 46238, "tokens": 5883052032, "loss": 3.191639172308373, "ce_loss": 3.190446573676485, "z_loss": 11.925985249606045, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1800868958234787, "tokens_per_second": 4211170.004737122, "eta_seconds": 4539.484248438299, "mean_data_wait_seconds": 7.420778274536133e-05}
{"step": 10890, "total_steps": 46238, "tokens": 5888458752, "loss": 3.152281611254721, "ce_loss": 3.151086615793633, "z_loss": 11.949983517328898, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17533016204833984, "tokens_per_second": 4211302.214970893, "eta_seconds": 4538.057875794622, "mean_data_wait_seconds": 8.64267349243164e-05}
{"step": 10900, "total_steps": 46238, "tokens": 5893865472, "loss": 3.20429221608422, "ce_loss": 3.203115549954501, "z_loss": 11.766852588364571, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1598021388053894, "tokens_per_second": 4211711.366092007, "eta_seconds": 4536.3332867057225, "mean_data_wait_seconds": 6.213784217834473e-05}
{"step": 10910, "total_steps": 46238, "tokens": 5899272192, "loss": 3.1543951251290063, "ce_loss": 3.153217563123414, "z_loss": 11.775986606424505, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1927797496318817, "tokens_per_second": 4212086.401199425, "eta_seconds": 4534.645760960894, "mean_data_wait_seconds": 7.76052474975586e-05}
{"step": 10920, "total_steps": 46238, "tokens": 5904678912, "loss": 3.232856040651148, "ce_loss": 3.231667419274648, "z_loss": 11.885919267481023, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17517387866973877, "tokens_per_second": 4212537.424133143, "eta_seconds": 4532.876767956395, "mean_data_wait_seconds": 9.083747863769531e-05}
{"step": 10930, "total_steps": 46238, "tokens": 5910085632, "loss": 3.3272333777312078, "ce_loss": 3.32605728597352, "z_loss": 11.760814283833358, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16433614492416382, "tokens_per_second": 4212658.886860599, "eta_seconds": 4531.46262649955, "mean_data_wait_seconds": 8.291006088256836e-05}
{"step": 10940, "total_steps": 46238, "tokens": 5915492352, "loss": 3.2566421031951904, "ce_loss": 3.255505439006921, "z_loss": 11.36655068397522, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.186399444937706, "tokens_per_second": 4212730.552839019, "eta_seconds": 4530.102114206889, "mean_data_wait_seconds": 0.00010865926742553711}
{"step": 10950, "total_steps": 46238, "tokens": 5920899072, "loss": 3.2069432735443115, "ce_loss": 3.2057764385685776, "z_loss": 11.668407620805683, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18210606276988983, "tokens_per_second": 4212842.998865577, "eta_seconds": 4528.69781027621, "mean_data_wait_seconds": 0.00014278292655944824}
{"step": 10960, "total_steps": 46238, "tokens": 5926305792, "loss": 3.2611880790103567, "ce_loss": 3.260044332706567, "z_loss": 11.437311707120953, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19190248847007751, "tokens_per_second": 4212774.609763919, "eta_seconds": 4527.487917296589, "mean_data_wait_seconds": 8.675456047058105e-05}
{"step": 10970, "total_steps": 46238, "tokens": 5931712512, "loss": 3.228487740863453, "ce_loss": 3.227339253281102, "z_loss": 11.48482060432434, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20431040227413177, "tokens_per_second": 4212800.800640731, "eta_seconds": 4526.176367299384, "mean_data_wait_seconds": 7.367134094238281e-05}
{"step": 10980, "total_steps": 46238, "tokens": 5937119232, "loss": 3.1699045914592165, "ce_loss": 3.168739662025914, "z_loss": 11.64923629616246, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1811765730381012, "tokens_per_second": 4213055.832021799, "eta_seconds": 4524.619056579682, "mean_data_wait_seconds": 8.568167686462402e-05}
{"step": 10990, "total_steps": 46238, "tokens": 5942525952, "loss": 3.262511251550732, "ce_loss": 3.2613720442309524, "z_loss": 11.392219803550027, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17053893208503723, "tokens_per_second": 4213141.543208683, "eta_seconds": 4523.243709843735, "mean_data_wait_seconds": 8.252263069152832e-05}
{"step": 11000, "total_steps": 46238, "tokens": 5947932672, "loss": 3.1799682920629326, "ce_loss": 3.178817573821906, "z_loss": 11.50740021647829, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17654643952846527, "tokens_per_second": 4213592.713980621, "eta_seconds": 4521.476221654493, "mean_data_wait_seconds": 7.897615432739258e-05}
{"step": 11010, "total_steps": 46238, "tokens": 5953339392, "loss": 3.292661442901149, "ce_loss": 3.291471242904663, "z_loss": 11.90187738158486, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19556641578674316, "tokens_per_second": 4213912.965595395, "eta_seconds": 4519.849531659443, "mean_data_wait_seconds": 0.00012260675430297852}
{"step": 11020, "total_steps": 46238, "tokens": 5958746112, "loss": 3.2706494981592353, "ce_loss": 3.2694542371865474, "z_loss": 11.952607176520608, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1984638124704361, "tokens_per_second": 4214273.780135496, "eta_seconds": 4518.179600421643, "mean_data_wait_seconds": 7.572770118713379e-05}
{"step": 11030, "total_steps": 46238, "tokens": 5964152832, "loss": 3.2246049985741125, "ce_loss": 3.2234058091134736, "z_loss": 11.991858359539147, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15367266535758972, "tokens_per_second": 4214693.033120903, "eta_seconds": 4516.447331848651, "mean_data_wait_seconds": 9.745359420776367e-05}
{"step": 11040, "total_steps": 46238, "tokens": 5969559552, "loss": 3.3055403142264397, "ce_loss": 3.304325403589191, "z_loss": 12.149184342586633, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22918696701526642, "tokens_per_second": 4214600.20081218, "eta_seconds": 4515.26395797466, "mean_data_wait_seconds": 6.86943531036377e-05}
{"step": 11050, "total_steps": 46238, "tokens": 5974966272, "loss": 3.2711570606087195, "ce_loss": 3.2699854518427993, "z_loss": 11.71617851112828, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.181278795003891, "tokens_per_second": 4214709.0098743215, "eta_seconds": 4513.864567975785, "mean_data_wait_seconds": 0.00010126829147338867}
{"step": 11060, "total_steps": 46238, "tokens": 5980372992, "loss": 3.288560612635179, "ce_loss": 3.287401703270999, "z_loss": 11.589101054451682, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18125547468662262, "tokens_per_second": 4215121.713507596, "eta_seconds": 4512.1399182974565, "mean_data_wait_seconds": 8.83638858795166e-05}
{"step": 11070, "total_steps": 46238, "tokens": 5985779712, "loss": 3.304593133203911, "ce_loss": 3.3034274289102266, "z_loss": 11.656979575301662, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1797851175069809, "tokens_per_second": 4215311.344084807, "eta_seconds": 4510.654296196031, "mean_data_wait_seconds": 7.253885269165039e-05}
{"step": 11080, "total_steps": 46238, "tokens": 5991186432, "loss": 3.278977031057531, "ce_loss": 3.2778356129472908, "z_loss": 11.414334326079398, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20461049675941467, "tokens_per_second": 4215812.0144557245, "eta_seconds": 4508.836124291479, "mean_data_wait_seconds": 0.0001099705696105957}
{"step": 11090, "total_steps": 46238, "tokens": 5996593152, "loss": 3.268165880983526, "ce_loss": 3.266994761698174, "z_loss": 11.711527520960027, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19188345968723297, "tokens_per_second": 4215788.762092765, "eta_seconds": 4507.578499869309, "mean_data_wait_seconds": 8.153915405273438e-05}
{"step": 11100, "total_steps": 46238, "tokens": 6001999872, "loss": 3.2253517472382747, "ce_loss": 3.2241762966820686, "z_loss": 11.754648526509603, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16395719349384308, "tokens_per_second": 4216163.642165103, "eta_seconds": 4505.895329585517, "mean_data_wait_seconds": 7.912516593933105e-05}
{"step": 11110, "total_steps": 46238, "tokens": 6007406592, "loss": 3.2856987263217117, "ce_loss": 3.2844968654892663, "z_loss": 12.018582495776089, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2040071189403534, "tokens_per_second": 4216645.670489795, "eta_seconds": 4504.098003044661, "mean_data_wait_seconds": 7.522106170654297e-05}
{"step": 11120, "total_steps": 46238, "tokens": 6012813312, "loss": 3.2641734426671807, "ce_loss": 3.2630058815985015, "z_loss": 11.676063053535692, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16002658009529114, "tokens_per_second": 4216813.768147096, "eta_seconds": 4502.636271827331, "mean_data_wait_seconds": 0.0001055598258972168}
{"step": 11130, "total_steps": 46238, "tokens": 6018220032, "loss": 3.248464716203285, "ce_loss": 3.247271644346642, "z_loss": 11.930498867323905, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16083428263664246, "tokens_per_second": 4217018.129263362, "eta_seconds": 4501.135950135388, "mean_data_wait_seconds": 7.787346839904785e-05}
{"step": 11140, "total_steps": 46238, "tokens": 6023626752, "loss": 3.293163680669033, "ce_loss": 3.2919819535631123, "z_loss": 11.817168575344663, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19730454683303833, "tokens_per_second": 4217394.652681011, "eta_seconds": 4499.452089914545, "mean_data_wait_seconds": 0.00012084841728210449}
{"step": 11150, "total_steps": 46238, "tokens": 6029033472, "loss": 3.303567194577419, "ce_loss": 3.302393356959025, "z_loss": 11.73892111489267, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19118055701255798, "tokens_per_second": 4217645.97630189, "eta_seconds": 4497.9020454992615, "mean_data_wait_seconds": 8.279085159301758e-05}
{"step": 11160, "total_steps": 46238, "tokens": 6034440192, "loss": 3.262900025555582, "ce_loss": 3.2617382912924797, "z_loss": 11.616912047068277, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1820058375597, "tokens_per_second": 4217983.033835205, "eta_seconds": 4496.260793812609, "mean_data_wait_seconds": 6.940960884094238e-05}
{"step": 11170, "total_steps": 46238, "tokens": 6039846912, "loss": 3.3008943525227634, "ce_loss": 3.2997194438269646, "z_loss": 11.749535914623376, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17452450096607208, "tokens_per_second": 4218607.175180101, "eta_seconds": 4494.313937440874, "mean_data_wait_seconds": 6.762146949768066e-05}
{"step": 11180, "total_steps": 46238, "tokens": 6045253632, "loss": 3.2669716947006457, "ce_loss": 3.2657741488832417, "z_loss": 11.975413929332387, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19957321882247925, "tokens_per_second": 4219100.937874731, "eta_seconds": 4492.50648019523, "mean_data_wait_seconds": 0.00018385052680969238}
{"step": 11190, "total_steps": 46238, "tokens": 6050660352, "loss": 3.292755871108084, "ce_loss": 3.291584188287908, "z_loss": 11.716942982240157, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17463889718055725, "tokens_per_second": 4219125.210642933, "eta_seconds": 4491.199155739789, "mean_data_wait_seconds": 6.875395774841309e-05}
{"step": 11200, "total_steps": 46238, "tokens": 6056067072, "loss": 3.2833130793137983, "ce_loss": 3.2821520512754265, "z_loss": 11.610282002073346, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17333030700683594, "tokens_per_second": 4219614.997319446, "eta_seconds": 4489.396515092982, "mean_data_wait_seconds": 0.00014019012451171875}
{"step": 11210, "total_steps": 46238, "tokens": 6061473792, "loss": 3.248165692343856, "ce_loss": 3.246956326744773, "z_loss": 12.093330181006229, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15862342715263367, "tokens_per_second": 4219881.024946121, "eta_seconds": 4487.832247413137, "mean_data_wait_seconds": 7.325410842895508e-05}
{"step": 11220, "total_steps": 46238, "tokens": 6066880512, "loss": 3.313549493298386, "ce_loss": 3.312359981464617, "z_loss": 11.895454869125828, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20257876813411713, "tokens_per_second": 4220222.464500786, "eta_seconds": 4486.188011000877, "mean_data_wait_seconds": 8.487701416015625e-05}
{"step": 11230, "total_steps": 46238, "tokens": 6072287232, "loss": 3.2793084346886836, "ce_loss": 3.278080889672944, "z_loss": 12.275334148695976, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21800455451011658, "tokens_per_second": 4220337.421194429, "eta_seconds": 4484.784702035328, "mean_data_wait_seconds": 8.270144462585449e-05}
{"step": 11240, "total_steps": 46238, "tokens": 6077693952, "loss": 3.316376068375327, "ce_loss": 3.3151978651682534, "z_loss": 11.781838922789603, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1718853861093521, "tokens_per_second": 4220257.158169833, "eta_seconds": 4483.588860780635, "mean_data_wait_seconds": 0.00010201334953308105}
{"step": 11250, "total_steps": 46238, "tokens": 6083100672, "loss": 3.2767025738051445, "ce_loss": 3.27552590045062, "z_loss": 11.767259583328709, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15805667638778687, "tokens_per_second": 4220691.836549043, "eta_seconds": 4481.846104042189, "mean_data_wait_seconds": 8.279085159301758e-05}
{"step": 11260, "total_steps": 46238, "tokens": 6088507392, "loss": 3.279080300620108, "ce_loss": 3.277904116746151, "z_loss": 11.762224847620184, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18009991943836212, "tokens_per_second": 4221006.638620976, "eta_seconds": 4480.23094087773, "mean_data_wait_seconds": 6.35981559753418e-05}
{"step": 11270, "total_steps": 46238, "tokens": 6093914112, "loss": 3.237974880319653, "ce_loss": 3.236777758959568, "z_loss": 11.97091446500836, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21327608823776245, "tokens_per_second": 4221276.504140223, "eta_seconds": 4478.66369460928, "mean_data_wait_seconds": 0.0001251697540283203}
{"step": 11280, "total_steps": 46238, "tokens": 6099320832, "loss": 3.3062449513059673, "ce_loss": 3.305042183760441, "z_loss": 12.02751567146995, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1944521814584732, "tokens_per_second": 4221574.08844972, "eta_seconds": 4477.067252168185, "mean_data_wait_seconds": 8.678436279296875e-05}
{"step": 11290, "total_steps": 46238, "tokens": 6104727552, "loss": 3.262703731204524, "ce_loss": 3.2615295901443018, "z_loss": 11.741507125623299, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17318004369735718, "tokens_per_second": 4221748.183287339, "eta_seconds": 4475.601945847746, "mean_data_wait_seconds": 0.0001354813575744629}
{"step": 11300, "total_steps": 46238, "tokens": 6110134272, "loss": 3.2774267810763735, "ce_loss": 3.2762064428040474, "z_loss": 12.203149029702852, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18290524184703827, "tokens_per_second": 4221905.177162594, "eta_seconds": 4474.15488300829, "mean_data_wait_seconds": 8.091330528259277e-05}
{"step": 11310, "total_steps": 46238, "tokens": 6115540992, "loss": 3.26359084519473, "ce_loss": 3.2623424150727014, "z_loss": 12.484463294347128, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16614340245723724, "tokens_per_second": 4222193.847116727, "eta_seconds": 4472.568439010832, "mean_data_wait_seconds": 0.00010979175567626953}
{"step": 11320, "total_steps": 46238, "tokens": 6120947712, "loss": 3.2409230001044995, "ce_loss": 3.2397253603646248, "z_loss": 11.976569269642685, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16387970745563507, "tokens_per_second": 4222590.460383093, "eta_seconds": 4470.867918904748, "mean_data_wait_seconds": 7.170438766479492e-05}
{"step": 11330, "total_steps": 46238, "tokens": 6126354432, "loss": 3.2032382813366977, "ce_loss": 3.202098071575165, "z_loss": 11.401578982671102, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18374568223953247, "tokens_per_second": 4223125.7183702765, "eta_seconds": 4469.02099596582, "mean_data_wait_seconds": 7.933378219604492e-05}
{"step": 11340, "total_steps": 46238, "tokens": 6131761152, "loss": 3.237678587436676, "ce_loss": 3.2364871303240457, "z_loss": 11.914488293907858, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19915883243083954, "tokens_per_second": 4223578.831982024, "eta_seconds": 4467.261423210084, "mean_data_wait_seconds": 8.541345596313477e-05}
{"step": 11350, "total_steps": 46238, "tokens": 6137167872, "loss": 3.1857661677129343, "ce_loss": 3.1845756859490364, "z_loss": 11.904826828927705, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19055180251598358, "tokens_per_second": 4223874.547153411, "eta_seconds": 4465.668630407577, "mean_data_wait_seconds": 7.864832878112793e-05}
{"step": 11360, "total_steps": 46238, "tokens": 6142574592, "loss": 3.2885541012792876, "ce_loss": 3.2873617699652007, "z_loss": 11.923410379525388, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1814982295036316, "tokens_per_second": 4224427.487765662, "eta_seconds": 4463.804242968234, "mean_data_wait_seconds": 9.903311729431152e-05}
{"step": 11370, "total_steps": 46238, "tokens": 6147981312, "loss": 3.1484919786453247, "ce_loss": 3.147324135809234, "z_loss": 11.678251114758579, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16167813539505005, "tokens_per_second": 4224749.730865759, "eta_seconds": 4462.183993118291, "mean_data_wait_seconds": 5.486607551574707e-05}
{"step": 11380, "total_steps": 46238, "tokens": 6153388032, "loss": 3.2071642568617156, "ce_loss": 3.2059791991204927, "z_loss": 11.850177909388687, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18446457386016846, "tokens_per_second": 4225114.8783221105, "eta_seconds": 4460.518695170783, "mean_data_wait_seconds": 6.48796558380127e-05}
{"step": 11390, "total_steps": 46238, "tokens": 6158794752, "loss": 3.2522522048516707, "ce_loss": 3.2510378234314197, "z_loss": 12.14338297554941, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16518084704875946, "tokens_per_second": 4225538.05843433, "eta_seconds": 4458.792448074884, "mean_data_wait_seconds": 6.058812141418457e-05}
{"step": 11400, "total_steps": 46238, "tokens": 6164201472, "loss": 3.228784888079672, "ce_loss": 3.2275911190293054, "z_loss": 11.93760070656285, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18274743854999542, "tokens_per_second": 4226039.110281753, "eta_seconds": 4456.98441790905, "mean_data_wait_seconds": 5.766749382019043e-05}
{"step": 11410, "total_steps": 46238, "tokens": 6169608192, "loss": 3.2300698161125183, "ce_loss": 3.228870299729434, "z_loss": 11.995558240196921, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16618146002292633, "tokens_per_second": 4226447.906963409, "eta_seconds": 4455.274064297848, "mean_data_wait_seconds": 9.074807167053223e-05}
{"step": 11420, "total_steps": 46238, "tokens": 6175014912, "loss": 3.2493017478422686, "ce_loss": 3.2480943618398723, "z_loss": 12.073780312682644, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18065467476844788, "tokens_per_second": 4226631.877457756, "eta_seconds": 4453.800938851256, "mean_data_wait_seconds": 8.51750373840332e-05}
{"step": 11430, "total_steps": 46238, "tokens": 6180421632, "loss": 3.233509488178022, "ce_loss": 3.232283575968309, "z_loss": 12.259057832486702, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18793010711669922, "tokens_per_second": 4226650.420431019, "eta_seconds": 4452.50220198738, "mean_data_wait_seconds": 0.0001016855239868164}
{"step": 11440, "total_steps": 46238, "tokens": 6185828352, "loss": 3.1809316367814033, "ce_loss": 3.17971689231468, "z_loss": 12.147756330894701, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19245445728302002, "tokens_per_second": 4226679.125922873, "eta_seconds": 4451.192774159812, "mean_data_wait_seconds": 6.455183029174805e-05}
{"step": 11450, "total_steps": 46238, "tokens": 6191235072, "loss": 3.235684821099946, "ce_loss": 3.2345189488295354, "z_loss": 11.658735217470111, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1644384264945984, "tokens_per_second": 4226890.214160204, "eta_seconds": 4449.6913596174, "mean_data_wait_seconds": 0.00010994076728820801}
{"step": 11460, "total_steps": 46238, "tokens": 6196641792, "loss": 3.2519684253316936, "ce_loss": 3.2507676409952566, "z_loss": 12.008007295203932, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18326550722122192, "tokens_per_second": 4223790.889929393, "eta_seconds": 4451.676381241098, "mean_data_wait_seconds": 7.042288780212402e-05}
{"step": 11470, "total_steps": 46238, "tokens": 6202048512, "loss": 3.260316940871152, "ce_loss": 3.2591248299136306, "z_loss": 11.920931628256133, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2008724808692932, "tokens_per_second": 4221267.000700995, "eta_seconds": 4453.05720317583, "mean_data_wait_seconds": 8.20159912109375e-05}
{"step": 11480, "total_steps": 46238, "tokens": 6207455232, "loss": 3.2457083879095134, "ce_loss": 3.2445216341452165, "z_loss": 11.867587754220674, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17253851890563965, "tokens_per_second": 4221628.057794613, "eta_seconds": 4451.3956338012995, "mean_data_wait_seconds": 9.739398956298828e-05}
{"step": 11490, "total_steps": 46238, "tokens": 6212861952, "loss": 3.1821374730630354, "ce_loss": 3.180950607314254, "z_loss": 11.868289001060255, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15896032750606537, "tokens_per_second": 4220048.142634809, "eta_seconds": 4451.7809629229505, "mean_data_wait_seconds": 8.216500282287598e-05}
{"step": 11500, "total_steps": 46238, "tokens": 6218268672, "loss": 3.2848701838291055, "ce_loss": 3.2837265314477864, "z_loss": 11.436144684300277, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1855623722076416, "tokens_per_second": 4220183.975792869, "eta_seconds": 4450.356518040532, "mean_data_wait_seconds": 0.0001195073127746582}
{"step": 11510, "total_steps": 46238, "tokens": 6223675392, "loss": 3.1939230796062583, "ce_loss": 3.192716813448704, "z_loss": 12.062470132654363, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14750289916992188, "tokens_per_second": 4220332.534060106, "eta_seconds": 4448.918750470338, "mean_data_wait_seconds": 0.00010266900062561035}
{"step": 11520, "total_steps": 46238, "tokens": 6229082112, "loss": 3.2306086872563218, "ce_loss": 3.229417455918861, "z_loss": 11.91258808338281, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21339456737041473, "tokens_per_second": 4220412.540483324, "eta_seconds": 4447.553324218487, "mean_data_wait_seconds": 0.0001321732997894287}
{"step": 11530, "total_steps": 46238, "tokens": 6234488832, "loss": 3.235695446982528, "ce_loss": 3.234518365426497, "z_loss": 11.770734324599758, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18794231116771698, "tokens_per_second": 4220642.64042716, "eta_seconds": 4446.029835423554, "mean_data_wait_seconds": 0.00016006827354431152}
{"step": 11540, "total_steps": 46238, "tokens": 6239895552, "loss": 3.2075982924663657, "ce_loss": 3.2063724290240896, "z_loss": 12.25875355980613, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20775604248046875, "tokens_per_second": 4220690.4655412985, "eta_seconds": 4444.69845328828, "mean_data_wait_seconds": 0.00015428662300109863}
{"step": 11550, "total_steps": 46238, "tokens": 6245302272, "loss": 3.260248774831945, "ce_loss": 3.2590523940144163, "z_loss": 11.964423049579967, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18944771587848663, "tokens_per_second": 4220924.35379771, "eta_seconds": 4443.171232653371, "mean_data_wait_seconds": 0.00010842084884643555}
{"step": 11560, "total_steps": 46238, "tokens": 6250708992, "loss": 3.295922626148571, "ce_loss": 3.294766635605783, "z_loss": 11.559609853860104, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15095268189907074, "tokens_per_second": 4221052.503889442, "eta_seconds": 4441.755445288598, "mean_data_wait_seconds": 0.00013265013694763184}
{"step": 11570, "total_steps": 46238, "tokens": 6256115712, "loss": 3.2644022323868493, "ce_loss": 3.2632166797464546, "z_loss": 11.855994593013417, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2107626497745514, "tokens_per_second": 4221042.458628942, "eta_seconds": 4440.485118950488, "mean_data_wait_seconds": 8.672475814819336e-05}
{"step": 11580, "total_steps": 46238, "tokens": 6261522432, "loss": 3.250922777435996, "ce_loss": 3.249750276406606, "z_loss": 11.725256471922904, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19355542957782745, "tokens_per_second": 4221231.8693770105, "eta_seconds": 4439.005030719967, "mean_data_wait_seconds": 0.00014480948448181152}
{"step": 11590, "total_steps": 46238, "tokens": 6266929152, "loss": 3.1912350437857886, "ce_loss": 3.190090627381296, "z_loss": 11.444554921352502, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.31723901629447937, "tokens_per_second": 4221389.902393178, "eta_seconds": 4437.558059581308, "mean_data_wait_seconds": 0.00014013051986694336}
{"step": 11600, "total_steps": 46238, "tokens": 6272335872, "loss": 3.2652205915162056, "ce_loss": 3.2640018607630874, "z_loss": 12.187336314808238, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15768010914325714, "tokens_per_second": 4221512.824810357, "eta_seconds": 4436.148092204668, "mean_data_wait_seconds": 8.58306884765625e-05}
{"step": 11610, "total_steps": 46238, "tokens": 6277742592, "loss": 3.270443022251129, "ce_loss": 3.2692390391320894, "z_loss": 12.039820338740494, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20221833884716034, "tokens_per_second": 4221891.570517081, "eta_seconds": 4434.469486317722, "mean_data_wait_seconds": 7.173418998718262e-05}
{"step": 11620, "total_steps": 46238, "tokens": 6283149312, "loss": 3.3082340865424187, "ce_loss": 3.307049738638329, "z_loss": 11.843323122371327, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19493919610977173, "tokens_per_second": 4222228.463954651, "eta_seconds": 4432.835121022723, "mean_data_wait_seconds": 7.012486457824707e-05}
{"step": 11630, "total_steps": 46238, "tokens": 6288556032, "loss": 3.2904398766430942, "ce_loss": 3.289278577674519, "z_loss": 11.613278179457694, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18288661539554596, "tokens_per_second": 4222513.019567616, "eta_seconds": 4431.255941021587, "mean_data_wait_seconds": 0.00015917420387268066}
{"step": 11640, "total_steps": 46238, "tokens": 6293962752, "loss": 3.243023966297959, "ce_loss": 3.2418076594670615, "z_loss": 12.162881627227321, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18476060032844543, "tokens_per_second": 4222833.280928207, "eta_seconds": 4429.639519154395, "mean_data_wait_seconds": 8.222460746765137e-05}
{"step": 11650, "total_steps": 46238, "tokens": 6299369472, "loss": 3.2829016717997463, "ce_loss": 3.2816780841711797, "z_loss": 12.235965540914824, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1809864342212677, "tokens_per_second": 4223029.26754749, "eta_seconds": 4428.153649728336, "mean_data_wait_seconds": 8.83638858795166e-05}
{"step": 11660, "total_steps": 46238, "tokens": 6304776192, "loss": 3.3340791229045754, "ce_loss": 3.332882928125786, "z_loss": 11.962022449031021, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15663787722587585, "tokens_per_second": 4223447.777384287, "eta_seconds": 4426.434687817611, "mean_data_wait_seconds": 6.648898124694824e-05}
{"step": 11670, "total_steps": 46238, "tokens": 6310182912, "loss": 3.2897578369487417, "ce_loss": 3.2885576865889807, "z_loss": 12.001280734033296, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17870907485485077, "tokens_per_second": 4223887.897897749, "eta_seconds": 4424.69342837006, "mean_data_wait_seconds": 6.988644599914551e-05}
{"step": 11680, "total_steps": 46238, "tokens": 6315589632, "loss": 3.2282159743887004, "ce_loss": 3.2269922531012334, "z_loss": 12.237007249485362, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1665511578321457, "tokens_per_second": 4224348.186044557, "eta_seconds": 4422.931415957607, "mean_data_wait_seconds": 6.517767906188965e-05}
{"step": 11690, "total_steps": 46238, "tokens": 6320996352, "loss": 3.3274189017035742, "ce_loss": 3.326211008158597, "z_loss": 12.078992648558183, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1967349499464035, "tokens_per_second": 4224860.578764819, "eta_seconds": 4421.115261858151, "mean_data_wait_seconds": 5.760788917541504e-05}
{"step": 11700, "total_steps": 46238, "tokens": 6326403072, "loss": 3.283470049048915, "ce_loss": 3.282239075863, "z_loss": 12.309934536616007, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1980038583278656, "tokens_per_second": 4225354.427783156, "eta_seconds": 4419.3189430967905, "mean_data_wait_seconds": 7.396936416625977e-05}
{"step": 11710, "total_steps": 46238, "tokens": 6331809792, "loss": 3.2832979252844146, "ce_loss": 3.2820693146098745, "z_loss": 12.286400216998477, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19005459547042847, "tokens_per_second": 4225592.957223786, "eta_seconds": 4417.789960598744, "mean_data_wait_seconds": 0.00014662742614746094}
{"step": 11720, "total_steps": 46238, "tokens": 6337216512, "loss": 3.2115381432302073, "ce_loss": 3.2103411681724316, "z_loss": 11.969943126042685, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19484034180641174, "tokens_per_second": 4225791.3910707645, "eta_seconds": 4416.303053537903, "mean_data_wait_seconds": 7.346272468566895e-05}
{"step": 11730, "total_steps": 46238, "tokens": 6342623232, "loss": 3.2933087583744163, "ce_loss": 3.29212008823048, "z_loss": 11.886531728686709, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1689530611038208, "tokens_per_second": 4225962.510351665, "eta_seconds": 4414.844821339281, "mean_data_wait_seconds": 6.181001663208008e-05}
{"step": 11740, "total_steps": 46238, "tokens": 6348029952, "loss": 3.281044114719738, "ce_loss": 3.2798173373395745, "z_loss": 12.267899578267878, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1613175868988037, "tokens_per_second": 4226431.440514882, "eta_seconds": 4413.075722749163, "mean_data_wait_seconds": 6.437301635742188e-05}
{"step": 11750, "total_steps": 46238, "tokens": 6353436672, "loss": 3.226388095003186, "ce_loss": 3.225237752452041, "z_loss": 11.503266681324352, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17813949286937714, "tokens_per_second": 4226872.377958059, "eta_seconds": 4411.336230834509, "mean_data_wait_seconds": 9.688735008239746e-05}
{"step": 11760, "total_steps": 46238, "tokens": 6358843392, "loss": 3.272114679668889, "ce_loss": 3.270903433814193, "z_loss": 12.112591425577799, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19511105120182037, "tokens_per_second": 4227187.67466959, "eta_seconds": 4409.728164117298, "mean_data_wait_seconds": 0.00014671683311462402}
{"step": 11770, "total_steps": 46238, "tokens": 6364250112, "loss": 3.274322562145464, "ce_loss": 3.2730997836951055, "z_loss": 12.227914116599344, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.269988089799881, "tokens_per_second": 4227562.7880403595, "eta_seconds": 4408.057965861273, "mean_data_wait_seconds": 5.793571472167969e-05}
{"step": 11780, "total_steps": 46238, "tokens": 6369656832, "loss": 3.261764618483457, "ce_loss": 3.260530224352172, "z_loss": 12.344312465552127, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19881826639175415, "tokens_per_second": 4227923.145123019, "eta_seconds": 4406.403443139677, "mean_data_wait_seconds": 7.355213165283203e-05}
{"step": 11790, "total_steps": 46238, "tokens": 6375063552, "loss": 3.2677722013357915, "ce_loss": 3.2665701815576265, "z_loss": 12.020138148105506, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14585144817829132, "tokens_per_second": 4228186.476976858, "eta_seconds": 4404.850279289594, "mean_data_wait_seconds": 8.493661880493164e-05}
{"step": 11800, "total_steps": 46238, "tokens": 6380470272, "loss": 3.253602972536376, "ce_loss": 3.252450455318798, "z_loss": 11.525435548840147, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2026498019695282, "tokens_per_second": 4228643.961286886, "eta_seconds": 4403.095137461921, "mean_data_wait_seconds": 6.35981559753418e-05}
{"step": 11810, "total_steps": 46238, "tokens": 6385876992, "loss": 3.2572856996998643, "ce_loss": 3.256073357480945, "z_loss": 12.123351862936309, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22024644911289215, "tokens_per_second": 4229058.94607447, "eta_seconds": 4401.3846062083785, "mean_data_wait_seconds": 5.7369470596313477e-05}
{"step": 11820, "total_steps": 46238, "tokens": 6391283712, "loss": 3.3160876469178633, "ce_loss": 3.3148668318083794, "z_loss": 12.20791972767223, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17000974714756012, "tokens_per_second": 4229399.588061744, "eta_seconds": 4399.751746447737, "mean_data_wait_seconds": 8.621811866760254e-05}
{"step": 11830, "total_steps": 46238, "tokens": 6396690432, "loss": 3.27916776230841, "ce_loss": 3.2779701189561323, "z_loss": 11.975940003539577, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1631908267736435, "tokens_per_second": 4229499.970176344, "eta_seconds": 4398.368988101535, "mean_data_wait_seconds": 8.526444435119629e-05}
{"step": 11840, "total_steps": 46238, "tokens": 6402097152, "loss": 3.257583515210585, "ce_loss": 3.256391859415806, "z_loss": 11.916532198588053, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20537927746772766, "tokens_per_second": 4229699.770799722, "eta_seconds": 4396.882944834573, "mean_data_wait_seconds": 0.0001856386661529541}
{"step": 11850, "total_steps": 46238, "tokens": 6407503872, "loss": 3.23491381334536, "ce_loss": 3.2337244640697134, "z_loss": 11.893755183075413, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.186368927359581, "tokens_per_second": 4229779.006031032, "eta_seconds": 4395.522328114652, "mean_data_wait_seconds": 0.00012022256851196289}
{"step": 11860, "total_steps": 46238, "tokens": 6412910592, "loss": 3.2603256196686714, "ce_loss": 3.259133600827419, "z_loss": 11.920390223011825, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18688246607780457, "tokens_per_second": 4229791.01605315, "eta_seconds": 4394.231599967644, "mean_data_wait_seconds": 6.61015510559082e-05}
{"step": 11870, "total_steps": 46238, "tokens": 6418317312, "loss": 3.26191810947476, "ce_loss": 3.2606916770790564, "z_loss": 12.26423157345165, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18852297961711884, "tokens_per_second": 4229955.66432004, "eta_seconds": 4392.782359572773, "mean_data_wait_seconds": 6.350874900817871e-05}
{"step": 11880, "total_steps": 46238, "tokens": 6423724032, "loss": 3.251172813502225, "ce_loss": 3.2500101273710076, "z_loss": 11.626679210951835, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19064897298812866, "tokens_per_second": 4230323.055182081, "eta_seconds": 4391.122772821061, "mean_data_wait_seconds": 0.0001131892204284668}
{"step": 11890, "total_steps": 46238, "tokens": 6429130752, "loss": 3.2365295923117436, "ce_loss": 3.23533894257112, "z_loss": 11.906836646975894, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1925235390663147, "tokens_per_second": 4230391.297108681, "eta_seconds": 4389.773871909258, "mean_data_wait_seconds": 8.466839790344238e-05}
{"step": 11900, "total_steps": 46238, "tokens": 6434537472, "loss": 3.226344074263717, "ce_loss": 3.225148795229016, "z_loss": 11.95263594569582, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17160660028457642, "tokens_per_second": 4230550.472764116, "eta_seconds": 4388.330687346734, "mean_data_wait_seconds": 6.398558616638184e-05}
{"step": 11910, "total_steps": 46238, "tokens": 6439944192, "loss": 3.2165311087261546, "ce_loss": 3.215285422223987, "z_loss": 12.45691711252386, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16588102281093597, "tokens_per_second": 4231049.836521874, "eta_seconds": 4386.534893490388, "mean_data_wait_seconds": 6.237626075744629e-05}
{"step": 11920, "total_steps": 46238, "tokens": 6445350912, "loss": 3.2593339031392876, "ce_loss": 3.2581384633526658, "z_loss": 11.954462759422533, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18814314901828766, "tokens_per_second": 4231365.4158362895, "eta_seconds": 4384.929969545759, "mean_data_wait_seconds": 8.031725883483887e-05}
{"step": 11930, "total_steps": 46238, "tokens": 6450757632, "loss": 3.252426109530709, "ce_loss": 3.251222951845689, "z_loss": 12.031506603414362, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20810358226299286, "tokens_per_second": 4231629.819169778, "eta_seconds": 4383.3782955143215, "mean_data_wait_seconds": 6.350874900817871e-05}
{"step": 11940, "total_steps": 46238, "tokens": 6456164352, "loss": 3.242749658497897, "ce_loss": 3.241535352938103, "z_loss": 12.143059354839902, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17610307037830353, "tokens_per_second": 4231869.650788197, "eta_seconds": 4381.852257794905, "mean_data_wait_seconds": 0.00010642409324645996}
{"step": 11950, "total_steps": 46238, "tokens": 6461571072, "loss": 3.2323272770101372, "ce_loss": 3.2311271501309946, "z_loss": 12.00122409878355, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19278956949710846, "tokens_per_second": 4231975.612121606, "eta_seconds": 4380.464956107433, "mean_data_wait_seconds": 9.629130363464355e-05}
{"step": 11960, "total_steps": 46238, "tokens": 6466977792, "loss": 3.240957415465153, "ce_loss": 3.2397726846463755, "z_loss": 11.847402211391564, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1725715547800064, "tokens_per_second": 4232113.471005914, "eta_seconds": 4379.04471866513, "mean_data_wait_seconds": 8.577108383178711e-05}
{"step": 11970, "total_steps": 46238, "tokens": 6472384512, "loss": 3.2545667380997627, "ce_loss": 3.25336135336847, "z_loss": 12.053858395778771, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2154734879732132, "tokens_per_second": 4232121.480363379, "eta_seconds": 4377.75888758496, "mean_data_wait_seconds": 6.651878356933594e-05}
{"step": 11980, "total_steps": 46238, "tokens": 6477791232, "loss": 3.1991182041890696, "ce_loss": 3.1979412707415493, "z_loss": 11.769429257421782, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1516897827386856, "tokens_per_second": 4232258.190560711, "eta_seconds": 4376.339975030242, "mean_data_wait_seconds": 0.000312119722366333}
{"step": 11990, "total_steps": 46238, "tokens": 6483197952, "loss": 3.1975236842126558, "ce_loss": 3.196350547400388, "z_loss": 11.73133863102306, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17953088879585266, "tokens_per_second": 4232174.915563923, "eta_seconds": 4375.1485591735645, "mean_data_wait_seconds": 6.499886512756348e-05}
{"step": 12000, "total_steps": 46238, "tokens": 6488604672, "loss": 3.2134525920405532, "ce_loss": 3.2122638044935283, "z_loss": 11.888210166584361, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18244807422161102, "tokens_per_second": 4232110.479622968, "eta_seconds": 4373.937625949953, "mean_data_wait_seconds": 0.00011983513832092285}
{"step": 12010, "total_steps": 46238, "tokens": 6494011392, "loss": 3.2723138025312712, "ce_loss": 3.2711293173558786, "z_loss": 11.844763510154955, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16622881591320038, "tokens_per_second": 4232419.051337087, "eta_seconds": 4372.341282736123, "mean_data_wait_seconds": 0.0001366138458251953}
{"step": 12020, "total_steps": 46238, "tokens": 6499418112, "loss": 3.271610666405071, "ce_loss": 3.2704559492342398, "z_loss": 11.547404520439379, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17242419719696045, "tokens_per_second": 4232500.483400215, "eta_seconds": 4370.979730907846, "mean_data_wait_seconds": 0.00010639429092407227}
{"step": 12030, "total_steps": 46238, "tokens": 6504824832, "loss": 3.2335596734827217, "ce_loss": 3.232360632130594, "z_loss": 11.990283604824182, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19421902298927307, "tokens_per_second": 4232758.555122582, "eta_seconds": 4369.435880441894, "mean_data_wait_seconds": 0.00020825862884521484}
{"step": 12040, "total_steps": 46238, "tokens": 6510231552, "loss": 3.282737262321241, "ce_loss": 3.281539732759649, "z_loss": 11.975590669747556, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20422391593456268, "tokens_per_second": 4232640.886535906, "eta_seconds": 4368.2799650721445, "mean_data_wait_seconds": 7.554888725280762e-05}
{"step": 12050, "total_steps": 46238, "tokens": 6515638272, "loss": 3.248221951903719, "ce_loss": 3.247022580016743, "z_loss": 11.99346113204956, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17224614322185516, "tokens_per_second": 4233075.234419429, "eta_seconds": 4366.554488260848, "mean_data_wait_seconds": 0.00010940432548522949}
{"step": 12060, "total_steps": 46238, "tokens": 6521044992, "loss": 3.2055363817648455, "ce_loss": 3.2042877945032986, "z_loss": 12.48613386443167, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1952366679906845, "tokens_per_second": 4233295.889081323, "eta_seconds": 4365.049698430145, "mean_data_wait_seconds": 7.298588752746582e-05}
{"step": 12070, "total_steps": 46238, "tokens": 6526451712, "loss": 3.2284216555682095, "ce_loss": 3.227196783730478, "z_loss": 12.248717373067683, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17752742767333984, "tokens_per_second": 4233869.582446365, "eta_seconds": 4363.181213845059, "mean_data_wait_seconds": 6.985664367675781e-05}
{"step": 12080, "total_steps": 46238, "tokens": 6531858432, "loss": 3.1976924845666597, "ce_loss": 3.196496605873108, "z_loss": 11.958946148554483, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1761780083179474, "tokens_per_second": 4234159.058341046, "eta_seconds": 4361.605988235054, "mean_data_wait_seconds": 7.05718994140625e-05}
{"step": 12090, "total_steps": 46238, "tokens": 6537265152, "loss": 3.1861547376170303, "ce_loss": 3.1849628415974705, "z_loss": 11.91886466922182, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1710847169160843, "tokens_per_second": 4234618.730955518, "eta_seconds": 4359.855740739635, "mean_data_wait_seconds": 0.0001010596752166748}
{"step": 12100, "total_steps": 46238, "tokens": 6542671872, "loss": 3.251613075082952, "ce_loss": 3.2504185474280156, "z_loss": 11.945412469632698, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18006937205791473, "tokens_per_second": 4234692.311919709, "eta_seconds": 4358.503216880224, "mean_data_wait_seconds": 8.207559585571289e-05}
{"step": 12110, "total_steps": 46238, "tokens": 6548078592, "loss": 3.2593913692416567, "ce_loss": 3.2581878701845803, "z_loss": 12.03501439817024, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17977933585643768, "tokens_per_second": 4234842.198116863, "eta_seconds": 4357.072230980641, "mean_data_wait_seconds": 6.74128532409668e-05}
{"step": 12120, "total_steps": 46238, "tokens": 6553485312, "loss": 3.208048603751443, "ce_loss": 3.206825613975525, "z_loss": 12.229870947924526, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18866081535816193, "tokens_per_second": 4234711.303903342, "eta_seconds": 4355.930144989415, "mean_data_wait_seconds": 9.644031524658203e-05}
{"step": 12130, "total_steps": 46238, "tokens": 6558892032, "loss": 3.2357398741173022, "ce_loss": 3.2345266685341345, "z_loss": 12.132151639822757, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17498216032981873, "tokens_per_second": 4233321.199743594, "eta_seconds": 4356.083328880626, "mean_data_wait_seconds": 6.034970283508301e-05}
{"step": 12140, "total_steps": 46238, "tokens": 6564298752, "loss": 3.227893236911658, "ce_loss": 3.226646678014235, "z_loss": 12.466059475234061, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16293859481811523, "tokens_per_second": 4233331.962368062, "eta_seconds": 4354.795075812475, "mean_data_wait_seconds": 7.56680965423584e-05}
{"step": 12150, "total_steps": 46238, "tokens": 6569705472, "loss": 3.2742727796236673, "ce_loss": 3.2730931260369043, "z_loss": 11.796481500972401, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19816909730434418, "tokens_per_second": 4233368.836791876, "eta_seconds": 4353.479976473419, "mean_data_wait_seconds": 7.11977481842041e-05}
{"step": 12160, "total_steps": 46238, "tokens": 6575112192, "loss": 3.2754812439282737, "ce_loss": 3.2742916092728125, "z_loss": 11.89653119173917, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17596101760864258, "tokens_per_second": 4233764.287370571, "eta_seconds": 4351.796296019763, "mean_data_wait_seconds": 5.8591365814208984e-05}
{"step": 12170, "total_steps": 46238, "tokens": 6580518912, "loss": 3.2961263927546414, "ce_loss": 3.2949405312538147, "z_loss": 11.85860345580361, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15511120855808258, "tokens_per_second": 4232649.428228224, "eta_seconds": 4351.665153546671, "mean_data_wait_seconds": 0.00011244416236877441}
{"step": 12180, "total_steps": 46238, "tokens": 6585925632, "loss": 3.2011311054229736, "ce_loss": 3.1999323765436807, "z_loss": 11.98761880759037, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17686200141906738, "tokens_per_second": 4232865.380149272, "eta_seconds": 4350.165821562377, "mean_data_wait_seconds": 7.808208465576172e-05}
{"step": 12190, "total_steps": 46238, "tokens": 6591332352, "loss": 3.278234915299849, "ce_loss": 3.2770572954958137, "z_loss": 11.776472113349222, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1959625631570816, "tokens_per_second": 4233183.786964163, "eta_seconds": 4348.561392653713, "mean_data_wait_seconds": 6.16908073425293e-05}
{"step": 12200, "total_steps": 46238, "tokens": 6596739072, "loss": 3.2891014940810925, "ce_loss": 3.287941419717037, "z_loss": 11.600988286914248, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18291956186294556, "tokens_per_second": 4231935.272453514, "eta_seconds": 4348.566714568565, "mean_data_wait_seconds": 7.918477058410645e-05}
{"step": 12210, "total_steps": 46238, "tokens": 6602145792, "loss": 3.247171205101591, "ce_loss": 3.245950854185856, "z_loss": 12.203272971239956, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18331073224544525, "tokens_per_second": 4230493.83307035, "eta_seconds": 4348.770349263871, "mean_data_wait_seconds": 9.542703628540039e-05}
{"step": 12220, "total_steps": 46238, "tokens": 6607552512, "loss": 3.3052379940495347, "ce_loss": 3.30400497082508, "z_loss": 12.330790317419803, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18423070013523102, "tokens_per_second": 4230546.653081752, "eta_seconds": 4347.43803394823, "mean_data_wait_seconds": 6.988644599914551e-05}
{"step": 12230, "total_steps": 46238, "tokens": 6612959232, "loss": 3.2529488025289592, "ce_loss": 3.2517187902421663, "z_loss": 12.299949804941813, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18855836987495422, "tokens_per_second": 4230579.27992536, "eta_seconds": 4346.1264964935, "mean_data_wait_seconds": 8.121132850646973e-05}
{"step": 12240, "total_steps": 46238, "tokens": 6618365952, "loss": 3.279722161365278, "ce_loss": 3.2784738233595183, "z_loss": 12.483464320500692, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19055941700935364, "tokens_per_second": 4230535.23378653, "eta_seconds": 4344.893723423249, "mean_data_wait_seconds": 9.527802467346191e-05}
{"step": 12250, "total_steps": 46238, "tokens": 6623772672, "loss": 3.2753369013468423, "ce_loss": 3.2741342132741753, "z_loss": 12.027242284832578, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17644037306308746, "tokens_per_second": 4230614.893816745, "eta_seconds": 4343.533912967871, "mean_data_wait_seconds": 9.226799011230469e-05}
{"step": 12260, "total_steps": 46238, "tokens": 6629179392, "loss": 3.287026746706529, "ce_loss": 3.28580591172883, "z_loss": 12.208651925578263, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19076275825500488, "tokens_per_second": 4230684.369931762, "eta_seconds": 4342.184606008862, "mean_data_wait_seconds": 6.341934204101562e-05}
{"step": 12270, "total_steps": 46238, "tokens": 6634586112, "loss": 3.280929280049873, "ce_loss": 3.2797287214886057, "z_loss": 12.005618044824311, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19336435198783875, "tokens_per_second": 4230925.126299396, "eta_seconds": 4340.6596136252265, "mean_data_wait_seconds": 8.955597877502441e-05}
{"step": 12280, "total_steps": 46238, "tokens": 6639992832, "loss": 3.227628525459405, "ce_loss": 3.2264025807380676, "z_loss": 12.259095466498172, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15515145659446716, "tokens_per_second": 4230974.1782523515, "eta_seconds": 4339.33139993391, "mean_data_wait_seconds": 0.00010761618614196777}
{"step": 12290, "total_steps": 46238, "tokens": 6645399552, "loss": 3.3418186884937864, "ce_loss": 3.3406195586377923, "z_loss": 11.991274840904005, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17396555840969086, "tokens_per_second": 4231418.9674488045, "eta_seconds": 4337.5975116607415, "mean_data_wait_seconds": 8.699297904968262e-05}
{"step": 12300, "total_steps": 46238, "tokens": 6650806272, "loss": 3.2282161911328635, "ce_loss": 3.227068897449609, "z_loss": 11.472802675131595, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17835630476474762, "tokens_per_second": 4231458.856076263, "eta_seconds": 4336.278878772891, "mean_data_wait_seconds": 8.45491886138916e-05}
{"step": 12310, "total_steps": 46238, "tokens": 6656212992, "loss": 3.326514325358651, "ce_loss": 3.325298137737043, "z_loss": 12.1616364031127, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.207190603017807, "tokens_per_second": 4231529.5610387195, "eta_seconds": 4334.928701170937, "mean_data_wait_seconds": 7.578730583190918e-05}
{"step": 12320, "total_steps": 46238, "tokens": 6661619712, "loss": 3.2669869011098687, "ce_loss": 3.265770393790621, "z_loss": 12.165189425150553, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1696358323097229, "tokens_per_second": 4231801.396135335, "eta_seconds": 4333.372601263149, "mean_data_wait_seconds": 6.473064422607422e-05}
{"step": 12330, "total_steps": 46238, "tokens": 6667026432, "loss": 3.275142622716499, "ce_loss": 3.27393812121767, "z_loss": 12.04485028440302, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16078895330429077, "tokens_per_second": 4232103.371135519, "eta_seconds": 4331.7858512234725, "mean_data_wait_seconds": 6.48200511932373e-05}
{"step": 12340, "total_steps": 46238, "tokens": 6672433152, "loss": 3.2700492956421594, "ce_loss": 3.2688256375717395, "z_loss": 12.23639514951995, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17020577192306519, "tokens_per_second": 4232318.995859267, "eta_seconds": 4330.287674896567, "mean_data_wait_seconds": 6.914138793945312e-05}
{"step": 12350, "total_steps": 46238, "tokens": 6677839872, "loss": 3.222267477801352, "ce_loss": 3.22108368620728, "z_loss": 11.838216463724772, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18007604777812958, "tokens_per_second": 4232544.374661824, "eta_seconds": 4328.77967533746, "mean_data_wait_seconds": 8.931756019592285e-05}
{"step": 12360, "total_steps": 46238, "tokens": 6683246592, "loss": 3.2432535381028145, "ce_loss": 3.2420742493687253, "z_loss": 11.792763153711954, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17740760743618011, "tokens_per_second": 4233091.049085383, "eta_seconds": 4326.943392336788, "mean_data_wait_seconds": 0.00014677643775939941}
{"step": 12370, "total_steps": 46238, "tokens": 6688653312, "loss": 3.2883783759492817, "ce_loss": 3.287167950110002, "z_loss": 12.10399517146024, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20420487225055695, "tokens_per_second": 4233451.79512576, "eta_seconds": 4325.297537362428, "mean_data_wait_seconds": 0.00010097026824951172}
{"step": 12380, "total_steps": 46238, "tokens": 6694060032, "loss": 3.2349707303625164, "ce_loss": 3.2337515733458777, "z_loss": 12.191214489214348, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16187936067581177, "tokens_per_second": 4233536.586203224, "eta_seconds": 4323.933791822266, "mean_data_wait_seconds": 0.00042626261711120605}
{"step": 12390, "total_steps": 46238, "tokens": 6699466752, "loss": 3.288802450353449, "ce_loss": 3.2875547896731985, "z_loss": 12.47642984534755, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18465135991573334, "tokens_per_second": 4233891.525236143, "eta_seconds": 4322.294294721998, "mean_data_wait_seconds": 8.463859558105469e-05}
{"step": 12400, "total_steps": 46238, "tokens": 6704873472, "loss": 3.234129862351851, "ce_loss": 3.2329074487541662, "z_loss": 12.224405389843565, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15372027456760406, "tokens_per_second": 4234152.181323563, "eta_seconds": 4320.75128161341, "mean_data_wait_seconds": 0.0002320408821105957}
{"step": 12410, "total_steps": 46238, "tokens": 6710280192, "loss": 3.280495285987854, "ce_loss": 3.2792821689085527, "z_loss": 12.131399046291005, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1909887045621872, "tokens_per_second": 4234220.376381711, "eta_seconds": 4319.404782523118, "mean_data_wait_seconds": 0.00024011731147766113}
{"step": 12420, "total_steps": 46238, "tokens": 6715686912, "loss": 3.313077615969109, "ce_loss": 3.3118524930693884, "z_loss": 12.251241286595663, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19725844264030457, "tokens_per_second": 4234496.8808240965, "eta_seconds": 4317.845906747174, "mean_data_wait_seconds": 0.00016996264457702637}
{"step": 12430, "total_steps": 46238, "tokens": 6721093632, "loss": 3.251582207101764, "ce_loss": 3.2503449555599326, "z_loss": 12.37211790229335, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16493527591228485, "tokens_per_second": 4234225.57793898, "eta_seconds": 4316.845658680543, "mean_data_wait_seconds": 0.00016608834266662598}
{"step": 12440, "total_steps": 46238, "tokens": 6726500352, "loss": 3.251170564781536, "ce_loss": 3.2499741803516042, "z_loss": 11.964145877144553, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.190871000289917, "tokens_per_second": 4234134.579010564, "eta_seconds": 4315.661498947932, "mean_data_wait_seconds": 0.00010991096496582031}
{"step": 12450, "total_steps": 46238, "tokens": 6731907072, "loss": 3.2853240244316333, "ce_loss": 3.2840841701536467, "z_loss": 12.39807925079808, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22716140747070312, "tokens_per_second": 4234379.01369148, "eta_seconds": 4314.135509583129, "mean_data_wait_seconds": 0.00010102987289428711}
{"step": 12460, "total_steps": 46238, "tokens": 6737313792, "loss": 3.2611478494875357, "ce_loss": 3.2598914287307044, "z_loss": 12.564284779808737, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21301019191741943, "tokens_per_second": 4234478.9816110395, "eta_seconds": 4312.756828716618, "mean_data_wait_seconds": 8.031725883483887e-05}
{"step": 12470, "total_steps": 46238, "tokens": 6742720512, "loss": 3.2009202892130073, "ce_loss": 3.19970872185447, "z_loss": 12.115464109362978, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18947936594486237, "tokens_per_second": 4234682.509097471, "eta_seconds": 4311.272777776922, "mean_data_wait_seconds": 8.133053779602051e-05}
{"step": 12480, "total_steps": 46238, "tokens": 6748127232, "loss": 3.238986645684098, "ce_loss": 3.23780045003602, "z_loss": 11.862558733333241, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1959400326013565, "tokens_per_second": 4234748.409533027, "eta_seconds": 4309.9289353089625, "mean_data_wait_seconds": 7.486343383789062e-05}
{"step": 12490, "total_steps": 46238, "tokens": 6753533952, "loss": 3.196889711148811, "ce_loss": 3.1956646045049033, "z_loss": 12.25138385368116, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1792088747024536, "tokens_per_second": 4234779.554254045, "eta_seconds": 4308.620496117899, "mean_data_wait_seconds": 8.687376976013184e-05}
{"step": 12500, "total_steps": 46238, "tokens": 6758940672, "loss": 3.2182871789643257, "ce_loss": 3.2171029845873513, "z_loss": 11.842155261473222, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20746131241321564, "tokens_per_second": 4234630.793770651, "eta_seconds": 4307.495069188296, "mean_data_wait_seconds": 9.009242057800293e-05}
{"step": 12510, "total_steps": 46238, "tokens": 6764347392, "loss": 3.200565674088218, "ce_loss": 3.199384449106274, "z_loss": 11.812445163726807, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17617569863796234, "tokens_per_second": 4235105.702818462, "eta_seconds": 4305.7353992049, "mean_data_wait_seconds": 7.259845733642578e-05}
{"step": 12520, "total_steps": 46238, "tokens": 6769754112, "loss": 3.2296613960555107, "ce_loss": 3.228427975466757, "z_loss": 12.334723429246383, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1949291229248047, "tokens_per_second": 4235366.814004856, "eta_seconds": 4304.19338502639, "mean_data_wait_seconds": 8.481740951538086e-05}
{"step": 12530, "total_steps": 46238, "tokens": 6775160832, "loss": 3.233781778451168, "ce_loss": 3.2326073917475613, "z_loss": 11.744003129727913, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.175028994679451, "tokens_per_second": 4235725.342064242, "eta_seconds": 4302.552604867078, "mean_data_wait_seconds": 8.019804954528809e-05}
{"step": 12540, "total_steps": 46238, "tokens": 6780567552, "loss": 3.2369141000689883, "ce_loss": 3.2357093493143716, "z_loss": 12.046934178381255, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15411581099033356, "tokens_per_second": 4236059.14463235, "eta_seconds": 4300.937206480208, "mean_data_wait_seconds": 0.00012916326522827148}
{"step": 12550, "total_steps": 46238, "tokens": 6785974272, "loss": 3.252889931201935, "ce_loss": 3.2517164367617983, "z_loss": 11.734765414035682, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17009706795215607, "tokens_per_second": 4236442.494396905, "eta_seconds": 4299.271780058204, "mean_data_wait_seconds": 7.781386375427246e-05}
{"step": 12560, "total_steps": 46238, "tokens": 6791380992, "loss": 3.215209617759242, "ce_loss": 3.214011497569807, "z_loss": 11.981501918850523, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1659686416387558, "tokens_per_second": 4236849.159150926, "eta_seconds": 4297.583005680798, "mean_data_wait_seconds": 8.812546730041504e-05}
{"step": 12570, "total_steps": 46238, "tokens": 6796787712, "loss": 3.1833259065945945, "ce_loss": 3.1820901415564795, "z_loss": 12.357669888120709, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19626422226428986, "tokens_per_second": 4237034.667880308, "eta_seconds": 4296.118783730993, "mean_data_wait_seconds": 9.191036224365234e-05}
{"step": 12580, "total_steps": 46238, "tokens": 6802194432, "loss": 3.289697430350564, "ce_loss": 3.288504961765174, "z_loss": 11.924553892829202, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20877955853939056, "tokens_per_second": 4237245.607844468, "eta_seconds": 4294.62891419627, "mean_data_wait_seconds": 9.703636169433594e-05}
{"step": 12590, "total_steps": 46238, "tokens": 6807601152, "loss": 3.268875441767953, "ce_loss": 3.2676768393227547, "z_loss": 11.98596986134847, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17022444307804108, "tokens_per_second": 4237376.814747483, "eta_seconds": 4293.219975312512, "mean_data_wait_seconds": 9.560585021972656e-05}
{"step": 12600, "total_steps": 46238, "tokens": 6813007872, "loss": 3.24456136154406, "ce_loss": 3.2433055097406562, "z_loss": 12.558543978315411, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19436553120613098, "tokens_per_second": 4237625.5668209335, "eta_seconds": 4291.692075485465, "mean_data_wait_seconds": 6.109476089477539e-05}
{"step": 12610, "total_steps": 46238, "tokens": 6818414592, "loss": 3.300992012023926, "ce_loss": 3.2997847867734507, "z_loss": 12.072002822702581, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19998504221439362, "tokens_per_second": 4238159.66304936, "eta_seconds": 4289.8755095315655, "mean_data_wait_seconds": 0.00011053681373596191}
{"step": 12620, "total_steps": 46238, "tokens": 6823821312, "loss": 3.3251021432154104, "ce_loss": 3.323888695601261, "z_loss": 12.134547883814031, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20305541157722473, "tokens_per_second": 4238309.118458758, "eta_seconds": 4288.4485571947935, "mean_data_wait_seconds": 7.537007331848145e-05}
{"step": 12630, "total_steps": 46238, "tokens": 6829228032, "loss": 3.2353357365637114, "ce_loss": 3.2341259681817256, "z_loss": 12.097762628035111, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18620339035987854, "tokens_per_second": 4238703.128223105, "eta_seconds": 4286.774363369285, "mean_data_wait_seconds": 9.134411811828613e-05}
{"step": 12640, "total_steps": 46238, "tokens": 6834634752, "loss": 3.2651441729429997, "ce_loss": 3.2638977151928525, "z_loss": 12.464268966154618, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17019341886043549, "tokens_per_second": 4238648.031327814, "eta_seconds": 4285.5545093017745, "mean_data_wait_seconds": 8.505582809448242e-05}
{"step": 12650, "total_steps": 46238, "tokens": 6840041472, "loss": 3.2527906786311758, "ce_loss": 3.2515552531589162, "z_loss": 12.353850884871049, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20833085477352142, "tokens_per_second": 4238841.59806633, "eta_seconds": 4284.0832911246325, "mean_data_wait_seconds": 8.124113082885742e-05}
{"step": 12660, "total_steps": 46238, "tokens": 6845448192, "loss": 3.206567641460534, "ce_loss": 3.2053722287669326, "z_loss": 11.95396739063841, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18251757323741913, "tokens_per_second": 4239117.505872836, "eta_seconds": 4282.529021394053, "mean_data_wait_seconds": 0.00014483928680419922}
{"step": 12670, "total_steps": 46238, "tokens": 6850854912, "loss": 3.2738249771522754, "ce_loss": 3.2726209705526177, "z_loss": 12.04011057362412, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16588523983955383, "tokens_per_second": 4239300.963276415, "eta_seconds": 4281.0683131997885, "mean_data_wait_seconds": 0.00011163949966430664}
{"step": 12680, "total_steps": 46238, "tokens": 6856261632, "loss": 3.254810676430211, "ce_loss": 3.2535658514860906, "z_loss": 12.448193875226108, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19329889118671417, "tokens_per_second": 4239403.546448585, "eta_seconds": 4279.689372623881, "mean_data_wait_seconds": 0.00010877847671508789}
{"step": 12690, "total_steps": 46238, "tokens": 6861668352, "loss": 3.2266860604286194, "ce_loss": 3.225480930371718, "z_loss": 12.05087636456345, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18083734810352325, "tokens_per_second": 4239362.561050773, "eta_seconds": 4278.455386345705, "mean_data_wait_seconds": 8.997321128845215e-05}
{"step": 12700, "total_steps": 46238, "tokens": 6867075072, "loss": 3.3047053199825864, "ce_loss": 3.30348704439221, "z_loss": 12.182662436456392, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22950461506843567, "tokens_per_second": 4239229.4426280875, "eta_seconds": 4277.314334927539, "mean_data_wait_seconds": 0.00015908479690551758}
{"step": 12710, "total_steps": 46238, "tokens": 6872481792, "loss": 3.2674897692420264, "ce_loss": 3.266249513987339, "z_loss": 12.402517672741052, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1585710644721985, "tokens_per_second": 4239434.835115907, "eta_seconds": 4275.831767444634, "mean_data_wait_seconds": 8.544325828552246e-05}
{"step": 12720, "total_steps": 46238, "tokens": 6877888512, "loss": 3.2605632814494045, "ce_loss": 3.2593079155141655, "z_loss": 12.55335369977084, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18150950968265533, "tokens_per_second": 4239462.119666718, "eta_seconds": 4274.52891722609, "mean_data_wait_seconds": 7.334351539611816e-05}
{"step": 12730, "total_steps": 46238, "tokens": 6883295232, "loss": 3.2494108279546103, "ce_loss": 3.248169750878305, "z_loss": 12.410776008259166, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.172689288854599, "tokens_per_second": 4239879.58334422, "eta_seconds": 4272.8328359058505, "mean_data_wait_seconds": 0.0001252293586730957}
{"step": 12740, "total_steps": 46238, "tokens": 6888701952, "loss": 3.247241800481623, "ce_loss": 3.246082210179531, "z_loss": 11.595852136611938, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1774192452430725, "tokens_per_second": 4240358.755020612, "eta_seconds": 4271.074932647289, "mean_data_wait_seconds": 0.00013020634651184082}
{"step": 12750, "total_steps": 46238, "tokens": 6894108672, "loss": 3.299953924887108, "ce_loss": 3.2987065369432624, "z_loss": 12.47408788131945, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22636711597442627, "tokens_per_second": 4240290.379369295, "eta_seconds": 4269.868722220158, "mean_data_wait_seconds": 0.00010144710540771484}
{"step": 12760, "total_steps": 46238, "tokens": 6899515392, "loss": 3.265469384915901, "ce_loss": 3.264261802037557, "z_loss": 12.07588799794515, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17059969902038574, "tokens_per_second": 4240529.277472968, "eta_seconds": 4268.353160572037, "mean_data_wait_seconds": 9.47415828704834e-05}
{"step": 12770, "total_steps": 46238, "tokens": 6904922112, "loss": 3.2530816847627815, "ce_loss": 3.251882870992025, "z_loss": 11.988228487245964, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18325573205947876, "tokens_per_second": 4241068.077785325, "eta_seconds": 4266.536045195717, "mean_data_wait_seconds": 9.456276893615723e-05}
{"step": 12780, "total_steps": 46238, "tokens": 6910328832, "loss": 3.2718512036583642, "ce_loss": 3.270634033463218, "z_loss": 12.17170237049912, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1844455897808075, "tokens_per_second": 4241421.126675739, "eta_seconds": 4264.906163226867, "mean_data_wait_seconds": 0.00010585784912109375}
{"step": 12790, "total_steps": 46238, "tokens": 6915735552, "loss": 3.243598166740302, "ce_loss": 3.2424017479925444, "z_loss": 11.963813984032834, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15758271515369415, "tokens_per_second": 4241872.652202585, "eta_seconds": 4263.177579036936, "mean_data_wait_seconds": 0.00011417269706726074}
{"step": 12800, "total_steps": 46238, "tokens": 6921142272, "loss": 3.3173231807622043, "ce_loss": 3.3161156502636997, "z_loss": 12.075375903736461, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.27890369296073914, "tokens_per_second": 4242200.0101351, "eta_seconds": 4261.574093821253, "mean_data_wait_seconds": 0.000199049711227417}
{"step": 12810, "total_steps": 46238, "tokens": 6926548992, "loss": 3.277348099332867, "ce_loss": 3.2761542616468486, "z_loss": 11.938327341368705, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17740991711616516, "tokens_per_second": 4242713.0799928615, "eta_seconds": 4259.784388726661, "mean_data_wait_seconds": 0.00011610984802246094}
{"step": 12820, "total_steps": 46238, "tokens": 6931955712, "loss": 3.2496427116972026, "ce_loss": 3.2484532793362937, "z_loss": 11.894206502220847, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2108064591884613, "tokens_per_second": 4243022.214896452, "eta_seconds": 4258.199771042427, "mean_data_wait_seconds": 5.990266799926758e-05}
{"step": 12830, "total_steps": 46238, "tokens": 6937362432, "loss": 3.298900586186033, "ce_loss": 3.297679987820712, "z_loss": 12.205804550286496, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1977820247411728, "tokens_per_second": 4242959.977372848, "eta_seconds": 4256.987951883476, "mean_data_wait_seconds": 0.00011733174324035645}
{"step": 12840, "total_steps": 46238, "tokens": 6942769152, "loss": 3.293289991942319, "ce_loss": 3.292098876201745, "z_loss": 11.911188067811908, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1631714403629303, "tokens_per_second": 4243163.721749856, "eta_seconds": 4255.509324668121, "mean_data_wait_seconds": 7.43567943572998e-05}
{"step": 12850, "total_steps": 46238, "tokens": 6948175872, "loss": 3.268574024691726, "ce_loss": 3.2673946962212073, "z_loss": 11.793432401888298, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18282705545425415, "tokens_per_second": 4243470.001597395, "eta_seconds": 4253.928048791389, "mean_data_wait_seconds": 6.768107414245605e-05}
{"step": 12860, "total_steps": 46238, "tokens": 6953582592, "loss": 3.238300095904957, "ce_loss": 3.237098607149991, "z_loss": 12.014892448078502, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20204253494739532, "tokens_per_second": 4243613.47852662, "eta_seconds": 4252.510139133963, "mean_data_wait_seconds": 9.709596633911133e-05}
{"step": 12870, "total_steps": 46238, "tokens": 6958989312, "loss": 3.2651000889864834, "ce_loss": 3.263884809884158, "z_loss": 12.152715372316765, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18322470784187317, "tokens_per_second": 4243719.301504707, "eta_seconds": 4251.130044723574, "mean_data_wait_seconds": 9.268522262573242e-05}
{"step": 12880, "total_steps": 46238, "tokens": 6964396032, "loss": 3.2342258312485437, "ce_loss": 3.233015035137986, "z_loss": 12.10799452752778, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16061978042125702, "tokens_per_second": 4243843.425165701, "eta_seconds": 4249.731692986721, "mean_data_wait_seconds": 8.469820022583008e-05}
{"step": 12890, "total_steps": 46238, "tokens": 6969802752, "loss": 3.2390153191306372, "ce_loss": 3.237780847332694, "z_loss": 12.344376910816539, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20115354657173157, "tokens_per_second": 4243660.980431661, "eta_seconds": 4248.640328984534, "mean_data_wait_seconds": 0.00015360116958618164}
{"step": 12900, "total_steps": 46238, "tokens": 6975209472, "loss": 3.2485869653297192, "ce_loss": 3.24738653681495, "z_loss": 12.004356521548647, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16689614951610565, "tokens_per_second": 4243794.91615101, "eta_seconds": 4247.232210822183, "mean_data_wait_seconds": 0.00010541081428527832}
{"step": 12910, "total_steps": 46238, "tokens": 6980616192, "loss": 3.2193929065357554, "ce_loss": 3.2182192025762615, "z_loss": 11.736930146361843, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19862483441829681, "tokens_per_second": 4243851.298253187, "eta_seconds": 4245.901771208806, "mean_data_wait_seconds": 5.695223808288574e-05}
{"step": 12920, "total_steps": 46238, "tokens": 6986022912, "loss": 3.2946970426675044, "ce_loss": 3.2935082027406404, "z_loss": 11.888316999782216, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19891178607940674, "tokens_per_second": 4244051.478921469, "eta_seconds": 4244.427550765182, "mean_data_wait_seconds": 6.857514381408691e-05}
{"step": 12930, "total_steps": 46238, "tokens": 6991429632, "loss": 3.2835119872382195, "ce_loss": 3.2822834596489416, "z_loss": 12.28498726902586, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19161368906497955, "tokens_per_second": 4244211.6925515095, "eta_seconds": 4242.993424574909, "mean_data_wait_seconds": 7.358193397521973e-05}
{"step": 12940, "total_steps": 46238, "tokens": 6996836352, "loss": 3.15719055406975, "ce_loss": 3.1559734868280813, "z_loss": 12.17060746568622, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1538398563861847, "tokens_per_second": 4244502.353779598, "eta_seconds": 4241.429049501905, "mean_data_wait_seconds": 9.873509407043457e-05}
{"step": 12950, "total_steps": 46238, "tokens": 7002243072, "loss": 3.207320404775215, "ce_loss": 3.206157431457982, "z_loss": 11.630005952083703, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1630643606185913, "tokens_per_second": 4244509.3329381645, "eta_seconds": 4240.148260327124, "mean_data_wait_seconds": 0.00010716915130615234}
{"step": 12960, "total_steps": 46238, "tokens": 7007649792, "loss": 3.2641426342906374, "ce_loss": 3.262939062985507, "z_loss": 12.03549489107999, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18559375405311584, "tokens_per_second": 4244728.8525627805, "eta_seconds": 4238.6552283869105, "mean_data_wait_seconds": 0.00010558962821960449}
{"step": 12970, "total_steps": 46238, "tokens": 7013056512, "loss": 3.236390032551505, "ce_loss": 3.235165370233131, "z_loss": 12.246418555577597, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17900361120700836, "tokens_per_second": 4244996.679655291, "eta_seconds": 4237.114132551117, "mean_data_wait_seconds": 9.647011756896973e-05}
{"step": 12980, "total_steps": 46238, "tokens": 7018463232, "loss": 3.2647291259332136, "ce_loss": 3.2634901621124963, "z_loss": 12.389807982878251, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20873072743415833, "tokens_per_second": 4245056.505410332, "eta_seconds": 4235.78076783737, "mean_data_wait_seconds": 0.00010210275650024414}
{"step": 12990, "total_steps": 46238, "tokens": 7023869952, "loss": 3.2315661419521677, "ce_loss": 3.230352313229532, "z_loss": 12.13819694519043, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19228515028953552, "tokens_per_second": 4244928.347630082, "eta_seconds": 4234.634960101444, "mean_data_wait_seconds": 0.00010690093040466309}
{"step": 13000, "total_steps": 46238, "tokens": 7029276672, "loss": 3.226498423200665, "ce_loss": 3.2252693700067923, "z_loss": 12.290371295177575, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14186938107013702, "tokens_per_second": 4245154.229287411, "eta_seconds": 4233.136016595677, "mean_data_wait_seconds": 0.00010097026824951172}
{"step": 13010, "total_steps": 46238, "tokens": 7034683392, "loss": 3.2555165110212383, "ce_loss": 3.254325135187669, "z_loss": 11.913671811421713, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1634427309036255, "tokens_per_second": 4245191.694083341, "eta_seconds": 4231.825047862565, "mean_data_wait_seconds": 0.0001061558723449707}
{"step": 13020, "total_steps": 46238, "tokens": 7040090112, "loss": 3.2339732267639856, "ce_loss": 3.2327748771869773, "z_loss": 11.983332619522557, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1589515209197998, "tokens_per_second": 4245422.002248225, "eta_seconds": 4230.32193607356, "mean_data_wait_seconds": 7.084012031555176e-05}
{"step": 13030, "total_steps": 46238, "tokens": 7045496832, "loss": 3.235428421786337, "ce_loss": 3.234181507067247, "z_loss": 12.469010056871356, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19938239455223083, "tokens_per_second": 4245600.484735082, "eta_seconds": 4228.870608186843, "mean_data_wait_seconds": 7.69197940826416e-05}
{"step": 13040, "total_steps": 46238, "tokens": 7050903552, "loss": 3.2922169493906424, "ce_loss": 3.291006731264519, "z_loss": 12.102824904701926, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17651580274105072, "tokens_per_second": 4245920.197281168, "eta_seconds": 4227.278787644964, "mean_data_wait_seconds": 0.00012439489364624023}
{"step": 13050, "total_steps": 46238, "tokens": 7056310272, "loss": 3.246380672310338, "ce_loss": 3.245204999591365, "z_loss": 11.75694301634124, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21774448454380035, "tokens_per_second": 4246204.954287655, "eta_seconds": 4225.721993443007, "mean_data_wait_seconds": 9.948015213012695e-05}
{"step": 13060, "total_steps": 46238, "tokens": 7061716992, "loss": 3.2835430719635705, "ce_loss": 3.2823263840241865, "z_loss": 12.16736094156901, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17389562726020813, "tokens_per_second": 4246511.84656811, "eta_seconds": 4224.143389237638, "mean_data_wait_seconds": 0.00010520219802856445}
{"step": 13070, "total_steps": 46238, "tokens": 7067123712, "loss": 3.2940730741529753, "ce_loss": 3.292855578841585, "z_loss": 12.17498796636408, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17856548726558685, "tokens_per_second": 4246579.082109033, "eta_seconds": 4222.803314684526, "mean_data_wait_seconds": 7.736682891845703e-05}
{"step": 13080, "total_steps": 46238, "tokens": 7072530432, "loss": 3.2614882227146262, "ce_loss": 3.26025857889291, "z_loss": 12.2965507218332, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17471912503242493, "tokens_per_second": 4246774.466037361, "eta_seconds": 4221.335897954485, "mean_data_wait_seconds": 7.432699203491211e-05}
{"step": 13090, "total_steps": 46238, "tokens": 7077937152, "loss": 3.2914753086639172, "ce_loss": 3.2902323675878122, "z_loss": 12.429492524175933, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19563338160514832, "tokens_per_second": 4247111.345438737, "eta_seconds": 4219.728028380346, "mean_data_wait_seconds": 0.00010961294174194336}
{"step": 13100, "total_steps": 46238, "tokens": 7083343872, "loss": 3.2719548579418296, "ce_loss": 3.2707239913217947, "z_loss": 12.309070717204701, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18861240148544312, "tokens_per_second": 4244533.743177153, "eta_seconds": 4221.0167589783805, "mean_data_wait_seconds": 8.293986320495605e-05}
{"step": 13110, "total_steps": 46238, "tokens": 7088750592, "loss": 3.2644138318119627, "ce_loss": 3.263206456646775, "z_loss": 12.073754238359856, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15412187576293945, "tokens_per_second": 4242305.82025195, "eta_seconds": 4221.959024853205, "mean_data_wait_seconds": 0.0001112520694732666}
{"step": 13120, "total_steps": 46238, "tokens": 7094157312, "loss": 3.294233686996229, "ce_loss": 3.2929880673235115, "z_loss": 12.456208272413773, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21827076375484467, "tokens_per_second": 4242176.884054029, "eta_seconds": 4220.812831097392, "mean_data_wait_seconds": 0.00012663006782531738}
{"step": 13130, "total_steps": 46238, "tokens": 7099564032, "loss": 3.2953339797077756, "ce_loss": 3.2941138419238003, "z_loss": 12.201314492659135, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17715877294540405, "tokens_per_second": 4242449.99809429, "eta_seconds": 4219.266676576201, "mean_data_wait_seconds": 7.507205009460449e-05}
{"step": 13140, "total_steps": 46238, "tokens": 7104970752, "loss": 3.2625683130639973, "ce_loss": 3.261378709113959, "z_loss": 11.895830725178573, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1950904130935669, "tokens_per_second": 4240570.590613285, "eta_seconds": 4219.861643999192, "mean_data_wait_seconds": 8.916854858398438e-05}
{"step": 13150, "total_steps": 46238, "tokens": 7110377472, "loss": 3.267894367376963, "ce_loss": 3.2667095299923057, "z_loss": 11.848315383448746, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22217530012130737, "tokens_per_second": 4240633.689503262, "eta_seconds": 4218.523874929528, "mean_data_wait_seconds": 9.530782699584961e-05}
{"step": 13160, "total_steps": 46238, "tokens": 7115784192, "loss": 3.258470594882965, "ce_loss": 3.2572345679456536, "z_loss": 12.36001824610161, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19430267810821533, "tokens_per_second": 4240553.573539509, "eta_seconds": 4217.328571343275, "mean_data_wait_seconds": 0.00011035799980163574}
{"step": 13170, "total_steps": 46238, "tokens": 7121190912, "loss": 3.219680018497236, "ce_loss": 3.218459938511704, "z_loss": 12.200691562710386, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18424823880195618, "tokens_per_second": 4240568.770925245, "eta_seconds": 4216.038458468186, "mean_data_wait_seconds": 0.00010767579078674316}
{"step": 13180, "total_steps": 46238, "tokens": 7126597632, "loss": 3.3243027130762735, "ce_loss": 3.3230676073016543, "z_loss": 12.351040746226456, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18322502076625824, "tokens_per_second": 4240564.155758591, "eta_seconds": 4214.76804677719, "mean_data_wait_seconds": 0.000167161226272583}
{"step": 13190, "total_steps": 46238, "tokens": 7132004352, "loss": 3.2587803331288425, "ce_loss": 3.2575421116568823, "z_loss": 12.381955493580211, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16651491820812225, "tokens_per_second": 4240406.596330413, "eta_seconds": 4213.649606021827, "mean_data_wait_seconds": 0.00015366077423095703}
{"step": 13200, "total_steps": 46238, "tokens": 7137411072, "loss": 3.2927612915183557, "ce_loss": 3.2915714943047725, "z_loss": 11.898308876788978, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1977221518754959, "tokens_per_second": 4240291.007476893, "eta_seconds": 4212.4893863425095, "mean_data_wait_seconds": 0.00025835633277893066}
{"step": 13210, "total_steps": 46238, "tokens": 7142817792, "loss": 3.163686681877483, "ce_loss": 3.162504443616578, "z_loss": 11.822097424304847, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18692149221897125, "tokens_per_second": 4240020.52912307, "eta_seconds": 4211.482944799132, "mean_data_wait_seconds": 9.28044319152832e-05}
{"step": 13220, "total_steps": 46238, "tokens": 7148224512, "loss": 3.21572676933173, "ce_loss": 3.2145330779480212, "z_loss": 11.937355922930168, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16520632803440094, "tokens_per_second": 4240284.596442726, "eta_seconds": 4209.945586901391, "mean_data_wait_seconds": 9.435415267944336e-05}
{"step": 13230, "total_steps": 46238, "tokens": 7153631232, "loss": 3.287942032019297, "ce_loss": 3.2867211645299736, "z_loss": 12.20916370189551, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19209593534469604, "tokens_per_second": 4240406.149263944, "eta_seconds": 4208.5498595689305, "mean_data_wait_seconds": 0.00010776519775390625}
{"step": 13240, "total_steps": 46238, "tokens": 7159037952, "loss": 3.2673717683011834, "ce_loss": 3.266205850875739, "z_loss": 11.659152565580426, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20501647889614105, "tokens_per_second": 4240650.6737646, "eta_seconds": 4207.032211913415, "mean_data_wait_seconds": 7.042288780212402e-05}
{"step": 13250, "total_steps": 46238, "tokens": 7164444672, "loss": 3.2154416351607353, "ce_loss": 3.214234373786233, "z_loss": 12.072765530961933, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1693381816148758, "tokens_per_second": 4241096.243910112, "eta_seconds": 4205.315380335897, "mean_data_wait_seconds": 6.103515625e-05}
{"step": 13260, "total_steps": 46238, "tokens": 7169851392, "loss": 3.2524077621373264, "ce_loss": 3.251204176382585, "z_loss": 12.03590075897448, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18949046730995178, "tokens_per_second": 4241118.27050279, "eta_seconds": 4204.018706105609, "mean_data_wait_seconds": 6.684660911560059e-05}
{"step": 13270, "total_steps": 46238, "tokens": 7175258112, "loss": 3.31441115249287, "ce_loss": 3.313201523188389, "z_loss": 12.096268393776633, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19309984147548676, "tokens_per_second": 4241410.328138035, "eta_seconds": 4202.454477406062, "mean_data_wait_seconds": 7.066130638122559e-05}
{"step": 13280, "total_steps": 46238, "tokens": 7180664832, "loss": 3.1682563532482493, "ce_loss": 3.167046075517481, "z_loss": 12.103010647224657, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16684745252132416, "tokens_per_second": 4241492.062757859, "eta_seconds": 4201.098773815449, "mean_data_wait_seconds": 8.386373519897461e-05}
{"step": 13290, "total_steps": 46238, "tokens": 7186071552, "loss": 3.2609342499212786, "ce_loss": 3.2597299163991753, "z_loss": 12.043125766696352, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17224560678005219, "tokens_per_second": 4241606.2530748835, "eta_seconds": 4199.710987102204, "mean_data_wait_seconds": 0.00010657310485839844}
{"step": 13300, "total_steps": 46238, "tokens": 7191478272, "loss": 3.265081282817956, "ce_loss": 3.263902395060568, "z_loss": 11.789130781636093, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19856788218021393, "tokens_per_second": 4241993.63457345, "eta_seconds": 4198.052896369016, "mean_data_wait_seconds": 0.00014507770538330078}
{"step": 13310, "total_steps": 46238, "tokens": 7196884992, "loss": 3.2125710447629294, "ce_loss": 3.2113875024246448, "z_loss": 11.835553190924905, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1622433066368103, "tokens_per_second": 4241912.796647798, "eta_seconds": 4196.858303656953, "mean_data_wait_seconds": 8.109211921691895e-05}
{"step": 13320, "total_steps": 46238, "tokens": 7202291712, "loss": 3.168349291339065, "ce_loss": 3.1671841541926065, "z_loss": 11.65157200350906, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16246150434017181, "tokens_per_second": 4242072.618622436, "eta_seconds": 4195.425638370959, "mean_data_wait_seconds": 0.00011351704597473145}
{"step": 13330, "total_steps": 46238, "tokens": 7207698432, "loss": 3.2830799980597063, "ce_loss": 3.281905340425896, "z_loss": 11.74620725891807, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19111159443855286, "tokens_per_second": 4242168.577929627, "eta_seconds": 4194.056218454963, "mean_data_wait_seconds": 0.00012922286987304688}
{"step": 13340, "total_steps": 46238, "tokens": 7213105152, "loss": 3.295915141250148, "ce_loss": 3.294716076417403, "z_loss": 11.990476615501173, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17801639437675476, "tokens_per_second": 4242312.803048651, "eta_seconds": 4192.639159285498, "mean_data_wait_seconds": 9.205937385559082e-05}
{"step": 13350, "total_steps": 46238, "tokens": 7218511872, "loss": 3.2539991862846143, "ce_loss": 3.252813458442688, "z_loss": 11.857448606780082, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18794620037078857, "tokens_per_second": 4242572.709149712, "eta_seconds": 4191.107915641038, "mean_data_wait_seconds": 7.608532905578613e-05}
{"step": 13360, "total_steps": 46238, "tokens": 7223918592, "loss": 3.2611079559181677, "ce_loss": 3.2598868463978623, "z_loss": 12.211224938883927, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2039901167154312, "tokens_per_second": 4242761.119641063, "eta_seconds": 4189.647458988646, "mean_data_wait_seconds": 8.58306884765625e-05}
{"step": 13370, "total_steps": 46238, "tokens": 7229325312, "loss": 3.2370788021521135, "ce_loss": 3.235810444210515, "z_loss": 12.6835089524587, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17476938664913177, "tokens_per_second": 4243146.590016264, "eta_seconds": 4187.992624580026, "mean_data_wait_seconds": 0.00012353062629699707}
{"step": 13380, "total_steps": 46238, "tokens": 7234732032, "loss": 3.2623772566968743, "ce_loss": 3.261155746199868, "z_loss": 12.214990167906791, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17428620159626007, "tokens_per_second": 4243341.504095615, "eta_seconds": 4186.526087248363, "mean_data_wait_seconds": 6.592273712158203e-05}
{"step": 13390, "total_steps": 46238, "tokens": 7240138752, "loss": 3.301714051853527, "ce_loss": 3.3004947702089944, "z_loss": 12.193212436907219, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1860770583152771, "tokens_per_second": 4243527.509795283, "eta_seconds": 4185.068470277633, "mean_data_wait_seconds": 0.00011944770812988281}
{"step": 13400, "total_steps": 46238, "tokens": 7245545472, "loss": 3.2658615563855027, "ce_loss": 3.264650243701357, "z_loss": 12.112939003742103, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17850293219089508, "tokens_per_second": 4243985.726852313, "eta_seconds": 4183.342642193063, "mean_data_wait_seconds": 0.00015205144882202148}
{"step": 13410, "total_steps": 46238, "tokens": 7250952192, "loss": 3.303752489162214, "ce_loss": 3.3025172930775266, "z_loss": 12.35197040528962, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19388356804847717, "tokens_per_second": 4244309.182505798, "eta_seconds": 4181.749957603555, "mean_data_wait_seconds": 0.00011873245239257812}
{"step": 13420, "total_steps": 46238, "tokens": 7256358912, "loss": 3.2965138428138965, "ce_loss": 3.295331306529768, "z_loss": 11.825216885769006, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2165369838476181, "tokens_per_second": 4244312.041678477, "eta_seconds": 4180.473266283026, "mean_data_wait_seconds": 7.596611976623535e-05}
{"step": 13430, "total_steps": 46238, "tokens": 7261765632, "loss": 3.2789580821990967, "ce_loss": 3.277784425200838, "z_loss": 11.736523136948094, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18674780428409576, "tokens_per_second": 4244459.305331729, "eta_seconds": 4179.054392562655, "mean_data_wait_seconds": 6.717443466186523e-05}
{"step": 13440, "total_steps": 46238, "tokens": 7267172352, "loss": 3.2116897106170654, "ce_loss": 3.2105045011549285, "z_loss": 11.851972421010336, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17249807715415955, "tokens_per_second": 4244813.353118933, "eta_seconds": 4177.432105694558, "mean_data_wait_seconds": 0.00011330842971801758}
{"step": 13450, "total_steps": 46238, "tokens": 7272579072, "loss": 3.2805236541863643, "ce_loss": 3.2793613311016196, "z_loss": 11.62324917677677, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19186607003211975, "tokens_per_second": 4244984.940851067, "eta_seconds": 4175.989576171724, "mean_data_wait_seconds": 6.747245788574219e-05}
{"step": 13460, "total_steps": 46238, "tokens": 7277985792, "loss": 3.310243369955005, "ce_loss": 3.309015393257141, "z_loss": 12.279837781732732, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1784311980009079, "tokens_per_second": 4245233.5178953605, "eta_seconds": 4174.471455880184, "mean_data_wait_seconds": 0.00010970234870910645}
{"step": 13470, "total_steps": 46238, "tokens": 7283392512, "loss": 3.2838642091462105, "ce_loss": 3.2826202006051033, "z_loss": 12.440232305815726, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17896436154842377, "tokens_per_second": 4245558.999439992, "eta_seconds": 4172.877924046479, "mean_data_wait_seconds": 0.00016993284225463867}
{"step": 13480, "total_steps": 46238, "tokens": 7288799232, "loss": 3.2818008263905845, "ce_loss": 3.2805964892560784, "z_loss": 12.043710737517387, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1844145953655243, "tokens_per_second": 4245912.16075957, "eta_seconds": 4171.257443260823, "mean_data_wait_seconds": 8.729100227355957e-05}
{"step": 13490, "total_steps": 46238, "tokens": 7294205952, "loss": 3.2875390992020117, "ce_loss": 3.28634018609018, "z_loss": 11.989265095103871, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17083680629730225, "tokens_per_second": 4246102.547609149, "eta_seconds": 4169.797075195314, "mean_data_wait_seconds": 7.617473602294922e-05}
{"step": 13500, "total_steps": 46238, "tokens": 7299612672, "loss": 3.226290437308225, "ce_loss": 3.2251114375663525, "z_loss": 11.790028362563163, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17549686133861542, "tokens_per_second": 4246517.378689674, "eta_seconds": 4168.116525985251, "mean_data_wait_seconds": 7.927417755126953e-05}
{"step": 13510, "total_steps": 46238, "tokens": 7305019392, "loss": 3.2076766888300576, "ce_loss": 3.206513887101954, "z_loss": 11.627824125867901, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1739802211523056, "tokens_per_second": 4246736.629748375, "eta_seconds": 4166.628186935253, "mean_data_wait_seconds": 7.990002632141113e-05}
{"step": 13520, "total_steps": 46238, "tokens": 7310426112, "loss": 3.2838674061226123, "ce_loss": 3.2826547171130325, "z_loss": 12.127260186455466, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1935005635023117, "tokens_per_second": 4247289.371378819, "eta_seconds": 4164.812961226957, "mean_data_wait_seconds": 0.00010135769844055176}
{"step": 13530, "total_steps": 46238, "tokens": 7315832832, "loss": 3.2309535055449516, "ce_loss": 3.229726448203578, "z_loss": 12.270120678525982, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18201731145381927, "tokens_per_second": 4247583.890104355, "eta_seconds": 4163.251288620351, "mean_data_wait_seconds": 7.298588752746582e-05}
{"step": 13540, "total_steps": 46238, "tokens": 7321239552, "loss": 3.239646236101786, "ce_loss": 3.2384441314321575, "z_loss": 12.02098914348718, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21686400473117828, "tokens_per_second": 4248107.571154165, "eta_seconds": 4161.465332008291, "mean_data_wait_seconds": 6.836652755737305e-05}
{"step": 13550, "total_steps": 46238, "tokens": 7326646272, "loss": 3.227309199896726, "ce_loss": 3.2261422764171255, "z_loss": 11.669585769826716, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20624831318855286, "tokens_per_second": 4248357.015566078, "eta_seconds": 4159.948328082108, "mean_data_wait_seconds": 0.00011283159255981445}
{"step": 13560, "total_steps": 46238, "tokens": 7332052992, "loss": 3.281902475790544, "ce_loss": 3.2806965148810185, "z_loss": 12.059811844970241, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16439907252788544, "tokens_per_second": 4248781.617443579, "eta_seconds": 4158.260069537361, "mean_data_wait_seconds": 8.106231689453125e-05}
{"step": 13570, "total_steps": 46238, "tokens": 7337459712, "loss": 3.254059377944831, "ce_loss": 3.252863022414121, "z_loss": 11.963246143225467, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16455310583114624, "tokens_per_second": 4248960.241211965, "eta_seconds": 4156.812778027334, "mean_data_wait_seconds": 7.766485214233398e-05}
{"step": 13580, "total_steps": 46238, "tokens": 7342866432, "loss": 3.2255946831269697, "ce_loss": 3.224390356829672, "z_loss": 12.043123404184977, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1722879707813263, "tokens_per_second": 4249178.821712697, "eta_seconds": 4155.326533629664, "mean_data_wait_seconds": 0.00012746453285217285}
{"step": 13590, "total_steps": 46238, "tokens": 7348273152, "loss": 3.2864811871991013, "ce_loss": 3.285300820162802, "z_loss": 11.803409496943155, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17949511110782623, "tokens_per_second": 4249074.793279709, "eta_seconds": 4154.155820442873, "mean_data_wait_seconds": 0.00014194846153259277}
{"step": 13600, "total_steps": 46238, "tokens": 7353679872, "loss": 3.2299907062992905, "ce_loss": 3.228808444557768, "z_loss": 11.82260911392443, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21284446120262146, "tokens_per_second": 4249052.227188521, "eta_seconds": 4152.905429377554, "mean_data_wait_seconds": 0.003443419933319092}
{"step": 13610, "total_steps": 46238, "tokens": 7359086592, "loss": 3.2549830512567, "ce_loss": 3.2538246783343228, "z_loss": 11.583605910792496, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20265346765518188, "tokens_per_second": 4249292.729540557, "eta_seconds": 4151.3980012168595, "mean_data_wait_seconds": 0.00014793872833251953}
{"step": 13620, "total_steps": 46238, "tokens": 7364493312, "loss": 3.284898102283478, "ce_loss": 3.2837027112642923, "z_loss": 11.953513506687049, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21812081336975098, "tokens_per_second": 4249421.512655512, "eta_seconds": 4149.999846209566, "mean_data_wait_seconds": 8.39531421661377e-05}
{"step": 13630, "total_steps": 46238, "tokens": 7369900032, "loss": 3.271208293510206, "ce_loss": 3.2699872435945454, "z_loss": 12.210180260918357, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17768090963363647, "tokens_per_second": 4249581.493591824, "eta_seconds": 4148.571319454581, "mean_data_wait_seconds": 8.144974708557129e-05}
{"step": 13640, "total_steps": 46238, "tokens": 7375306752, "loss": 3.2645116195534216, "ce_loss": 3.263289350451845, "z_loss": 12.222021984331535, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17126700282096863, "tokens_per_second": 4249589.857915265, "eta_seconds": 4147.290861769424, "mean_data_wait_seconds": 7.990002632141113e-05}
{"step": 13650, "total_steps": 46238, "tokens": 7380713472, "loss": 3.2197323748559663, "ce_loss": 3.218541858774243, "z_loss": 11.905002449498031, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18623659014701843, "tokens_per_second": 4249595.107704798, "eta_seconds": 4146.013447741364, "mean_data_wait_seconds": 8.755922317504883e-05}
{"step": 13660, "total_steps": 46238, "tokens": 7386120192, "loss": 3.264268288106629, "ce_loss": 3.26306688785553, "z_loss": 12.014621098836264, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17114576697349548, "tokens_per_second": 4249704.332175773, "eta_seconds": 4144.634630377266, "mean_data_wait_seconds": 8.594989776611328e-05}
{"step": 13670, "total_steps": 46238, "tokens": 7391526912, "loss": 3.296296061891498, "ce_loss": 3.2950711647669473, "z_loss": 12.248920447898634, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16053801774978638, "tokens_per_second": 4249652.461104347, "eta_seconds": 4143.412946155186, "mean_data_wait_seconds": 0.00011348724365234375}
{"step": 13680, "total_steps": 46238, "tokens": 7396933632, "loss": 3.243184832009402, "ce_loss": 3.24197762482094, "z_loss": 12.071866223306367, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17459137737751007, "tokens_per_second": 4249677.762454166, "eta_seconds": 4142.116011599562, "mean_data_wait_seconds": 8.32974910736084e-05}
{"step": 13690, "total_steps": 46238, "tokens": 7402340352, "loss": 3.260990366791234, "ce_loss": 3.2597964442137517, "z_loss": 11.93954160719207, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18895351886749268, "tokens_per_second": 4249859.520256551, "eta_seconds": 4140.666650303234, "mean_data_wait_seconds": 6.586313247680664e-05}
{"step": 13700, "total_steps": 46238, "tokens": 7407747072, "loss": 3.2891055797085618, "ce_loss": 3.2879024671785757, "z_loss": 12.03100848197937, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20138594508171082, "tokens_per_second": 4250070.336803521, "eta_seconds": 4139.189112157337, "mean_data_wait_seconds": 7.37607479095459e-05}
{"step": 13710, "total_steps": 46238, "tokens": 7413153792, "loss": 3.2144143599452395, "ce_loss": 3.2132423772956384, "z_loss": 11.719734928824685, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1954885721206665, "tokens_per_second": 4250128.180913334, "eta_seconds": 4137.860646880714, "mean_data_wait_seconds": 0.00011104345321655273}
{"step": 13720, "total_steps": 46238, "tokens": 7418560512, "loss": 3.26533561222481, "ce_loss": 3.2641162926500495, "z_loss": 12.193160808447635, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1983160823583603, "tokens_per_second": 4250333.113354419, "eta_seconds": 4136.389067661762, "mean_data_wait_seconds": 7.900595664978027e-05}
{"step": 13730, "total_steps": 46238, "tokens": 7423967232, "loss": 3.238284802798069, "ce_loss": 3.2370773752530417, "z_loss": 12.074379920959473, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17874175310134888, "tokens_per_second": 4250729.478802538, "eta_seconds": 4134.731412960014, "mean_data_wait_seconds": 7.915496826171875e-05}
{"step": 13740, "total_steps": 46238, "tokens": 7429373952, "loss": 3.2080240502501978, "ce_loss": 3.2068160266587227, "z_loss": 12.080455100897586, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18111881613731384, "tokens_per_second": 4250920.7207209915, "eta_seconds": 4133.273504338596, "mean_data_wait_seconds": 9.605288505554199e-05}
{"step": 13750, "total_steps": 46238, "tokens": 7434780672, "loss": 3.2667332327727117, "ce_loss": 3.2655123959888113, "z_loss": 12.208263527263295, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19975689053535461, "tokens_per_second": 4251221.2874975735, "eta_seconds": 4131.709472677038, "mean_data_wait_seconds": 0.00013080239295959473}
{"step": 13760, "total_steps": 46238, "tokens": 7440187392, "loss": 3.23551101214958, "ce_loss": 3.234241626479409, "z_loss": 12.693724458867853, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2060462385416031, "tokens_per_second": 4251461.361793715, "eta_seconds": 4130.204428481879, "mean_data_wait_seconds": 0.0001429915428161621}
{"step": 13770, "total_steps": 46238, "tokens": 7445594112, "loss": 3.239286953752691, "ce_loss": 3.2380467310096277, "z_loss": 12.402079625563188, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18687599897384644, "tokens_per_second": 4251653.356402149, "eta_seconds": 4128.7462435212765, "mean_data_wait_seconds": 5.8203935623168945e-05}
{"step": 13780, "total_steps": 46238, "tokens": 7451000832, "loss": 3.2790928320451216, "ce_loss": 3.277877393997077, "z_loss": 12.154786370017312, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19383585453033447, "tokens_per_second": 4251976.919408905, "eta_seconds": 4127.160480080767, "mean_data_wait_seconds": 0.00010088086128234863}
{"step": 13790, "total_steps": 46238, "tokens": 7456407552, "loss": 3.229075691916726, "ce_loss": 3.227868204767054, "z_loss": 12.074800036170267, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.170037180185318, "tokens_per_second": 4252352.903999717, "eta_seconds": 4125.524099257866, "mean_data_wait_seconds": 0.00010660290718078613}
{"step": 13800, "total_steps": 46238, "tokens": 7461814272, "loss": 3.2874211957960418, "ce_loss": 3.286201845515858, "z_loss": 12.193215955387462, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21976107358932495, "tokens_per_second": 4252671.8187625725, "eta_seconds": 4123.9433493607985, "mean_data_wait_seconds": 7.829070091247559e-05}
{"step": 13810, "total_steps": 46238, "tokens": 7467220992, "loss": 3.2487280838417285, "ce_loss": 3.247516189560746, "z_loss": 12.118765339706883, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16908258199691772, "tokens_per_second": 4252958.573765136, "eta_seconds": 4122.394008761441, "mean_data_wait_seconds": 9.91523265838623e-05}
{"step": 13820, "total_steps": 46238, "tokens": 7472627712, "loss": 3.2422978643215066, "ce_loss": 3.241102007302371, "z_loss": 11.958315769831339, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21183748543262482, "tokens_per_second": 4253107.582254186, "eta_seconds": 4120.978339962552, "mean_data_wait_seconds": 5.9545040130615234e-05}
{"step": 13830, "total_steps": 46238, "tokens": 7478034432, "loss": 3.2548567472082195, "ce_loss": 3.2536546642130073, "z_loss": 12.020560481331565, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18526428937911987, "tokens_per_second": 4253213.58297117, "eta_seconds": 4119.604426674467, "mean_data_wait_seconds": 0.00018721818923950195}
{"step": 13840, "total_steps": 46238, "tokens": 7483441152, "loss": 3.2496979760401175, "ce_loss": 3.248480869062019, "z_loss": 12.170695160374496, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1831117570400238, "tokens_per_second": 4253504.429274821, "eta_seconds": 4118.0516149095265, "mean_data_wait_seconds": 8.511543273925781e-05}
{"step": 13850, "total_steps": 46238, "tokens": 7488847872, "loss": 3.29939783522577, "ce_loss": 3.2981644254742246, "z_loss": 12.334348844759392, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18709325790405273, "tokens_per_second": 4253757.240573103, "eta_seconds": 4116.535823196343, "mean_data_wait_seconds": 8.079409599304199e-05}
{"step": 13860, "total_steps": 46238, "tokens": 7494254592, "loss": 3.270613307302648, "ce_loss": 3.269385997093085, "z_loss": 12.273399222980846, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16470617055892944, "tokens_per_second": 4254128.16078699, "eta_seconds": 4114.905964836191, "mean_data_wait_seconds": 0.00011903047561645508}
{"step": 13870, "total_steps": 46238, "tokens": 7499661312, "loss": 3.2749817154624243, "ce_loss": 3.2737356875882004, "z_loss": 12.460418043714581, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2054322063922882, "tokens_per_second": 4254399.181590227, "eta_seconds": 4113.372976312675, "mean_data_wait_seconds": 0.00013309717178344727}
{"step": 13880, "total_steps": 46238, "tokens": 7505068032, "loss": 3.2804476618766785, "ce_loss": 3.279221175294934, "z_loss": 12.264434005274918, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14792975783348083, "tokens_per_second": 4253012.723001686, "eta_seconds": 4113.442644877097, "mean_data_wait_seconds": 0.00013747811317443848}
{"step": 13890, "total_steps": 46238, "tokens": 7510474752, "loss": 3.2706444118962144, "ce_loss": 3.26943536418857, "z_loss": 12.090761531483043, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17181693017482758, "tokens_per_second": 4253137.59616624, "eta_seconds": 4112.050642275156, "mean_data_wait_seconds": 0.00012823939323425293}
{"step": 13900, "total_steps": 46238, "tokens": 7515881472, "loss": 3.2735191692005503, "ce_loss": 3.2723190820578374, "z_loss": 12.000969409942627, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18804696202278137, "tokens_per_second": 4253388.907048622, "eta_seconds": 4110.536526539198, "mean_data_wait_seconds": 0.0001443326473236084}
{"step": 13910, "total_steps": 46238, "tokens": 7521288192, "loss": 3.2636686617677864, "ce_loss": 3.26246127576539, "z_loss": 12.073960679950137, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17518071830272675, "tokens_per_second": 4253666.933201439, "eta_seconds": 4108.996782887582, "mean_data_wait_seconds": 0.00010907649993896484}
{"step": 13920, "total_steps": 46238, "tokens": 7526694912, "loss": 3.266884657469663, "ce_loss": 3.2656704927935745, "z_loss": 12.142148581418125, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15264709293842316, "tokens_per_second": 4254073.790155216, "eta_seconds": 4107.332849852253, "mean_data_wait_seconds": 9.426474571228027e-05}
{"step": 13930, "total_steps": 46238, "tokens": 7532101632, "loss": 3.2656811475753784, "ce_loss": 3.2644722064336142, "z_loss": 12.089523741693208, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15071246027946472, "tokens_per_second": 4252435.989931683, "eta_seconds": 4107.643323816526, "mean_data_wait_seconds": 7.045269012451172e-05}
{"step": 13940, "total_steps": 46238, "tokens": 7537508352, "loss": 3.341457699284409, "ce_loss": 3.340278993953358, "z_loss": 11.78695001746669, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17389622330665588, "tokens_per_second": 4252649.14532742, "eta_seconds": 4106.1660596163665, "mean_data_wait_seconds": 6.03795051574707e-05}
{"step": 13950, "total_steps": 46238, "tokens": 7542915072, "loss": 3.2741439830173147, "ce_loss": 3.2729354934258894, "z_loss": 12.084595629663179, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2049596607685089, "tokens_per_second": 4251637.859758332, "eta_seconds": 4105.871064237878, "mean_data_wait_seconds": 7.838010787963867e-05}
{"step": 13960, "total_steps": 46238, "tokens": 7548321792, "loss": 3.2502634940725383, "ce_loss": 3.249043123288588, "z_loss": 12.203705534790501, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1621810644865036, "tokens_per_second": 4251653.330831899, "eta_seconds": 4104.584448936103, "mean_data_wait_seconds": 0.00013718008995056152}
{"step": 13970, "total_steps": 46238, "tokens": 7553728512, "loss": 3.277044294458447, "ce_loss": 3.275801828413299, "z_loss": 12.424701314983945, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20739799737930298, "tokens_per_second": 4251525.316355935, "eta_seconds": 4103.436325989748, "mean_data_wait_seconds": 7.456541061401367e-05}
{"step": 13980, "total_steps": 46238, "tokens": 7559135232, "loss": 3.2949754812500696, "ce_loss": 3.293759766853217, "z_loss": 12.156830765984274, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15255114436149597, "tokens_per_second": 4251673.261187917, "eta_seconds": 4102.021870591048, "mean_data_wait_seconds": 8.380413055419922e-05}
{"step": 13990, "total_steps": 46238, "tokens": 7564541952, "loss": 3.318831983840827, "ce_loss": 3.3176103884523567, "z_loss": 12.216235832734542, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1980717033147812, "tokens_per_second": 4251609.609746509, "eta_seconds": 4100.811594750234, "mean_data_wait_seconds": 8.13603401184082e-05}
{"step": 14000, "total_steps": 46238, "tokens": 7569948672, "loss": 3.2505929325566147, "ce_loss": 3.249381332686453, "z_loss": 12.1158113118374, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15934015810489655, "tokens_per_second": 4251663.944976686, "eta_seconds": 4099.487515844947, "mean_data_wait_seconds": 0.00012060999870300293}
{"step": 14010, "total_steps": 46238, "tokens": 7575355392, "loss": 3.26836402127237, "ce_loss": 3.2671741752913506, "z_loss": 11.898576245163426, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18768846988677979, "tokens_per_second": 4251653.800495798, "eta_seconds": 4098.225622690189, "mean_data_wait_seconds": 8.308887481689453e-05}
{"step": 14020, "total_steps": 46238, "tokens": 7580762112, "loss": 3.254947992888364, "ce_loss": 3.2537542834426416, "z_loss": 11.937171401399555, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.180908203125, "tokens_per_second": 4251813.3098553615, "eta_seconds": 4096.800248408027, "mean_data_wait_seconds": 6.896257400512695e-05}
{"step": 14030, "total_steps": 46238, "tokens": 7586168832, "loss": 3.2583481488805828, "ce_loss": 3.2571656216274607, "z_loss": 11.825274156801628, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17935451865196228, "tokens_per_second": 4252022.486341857, "eta_seconds": 4095.327143714447, "mean_data_wait_seconds": 7.829070091247559e-05}
{"step": 14040, "total_steps": 46238, "tokens": 7591575552, "loss": 3.2565974123550183, "ce_loss": 3.255394818204822, "z_loss": 12.025910717068296, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18149085342884064, "tokens_per_second": 4252191.638932948, "eta_seconds": 4093.8927174901264, "mean_data_wait_seconds": 9.080767631530762e-05}
{"step": 14050, "total_steps": 46238, "tokens": 7596982272, "loss": 3.2717746712944726, "ce_loss": 3.2705546816190085, "z_loss": 12.199641950202711, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18008297681808472, "tokens_per_second": 4252462.94206546, "eta_seconds": 4092.3600983921556, "mean_data_wait_seconds": 0.00020569562911987305}
{"step": 14060, "total_steps": 46238, "tokens": 7602388992, "loss": 3.21861030477466, "ce_loss": 3.217399430997444, "z_loss": 12.108344995614255, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17024005949497223, "tokens_per_second": 4252465.165625189, "eta_seconds": 4091.086526617625, "mean_data_wait_seconds": 0.00011038780212402344}
{"step": 14070, "total_steps": 46238, "tokens": 7607795712, "loss": 3.215166743957635, "ce_loss": 3.2139626723347288, "z_loss": 12.040490822358565, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14698593318462372, "tokens_per_second": 4252623.471905043, "eta_seconds": 4089.6628490386934, "mean_data_wait_seconds": 0.0004171133041381836}
{"step": 14080, "total_steps": 46238, "tokens": 7613202432, "loss": 3.2215198440985247, "ce_loss": 3.220336374008294, "z_loss": 11.834612246715661, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17244793474674225, "tokens_per_second": 4252752.35974093, "eta_seconds": 4088.2675578737785, "mean_data_wait_seconds": 0.00012370944023132324}
{"step": 14090, "total_steps": 46238, "tokens": 7618609152, "loss": 3.2402184677846506, "ce_loss": 3.239014128843943, "z_loss": 12.043609084505023, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18928110599517822, "tokens_per_second": 4252710.759663816, "eta_seconds": 4087.0361908586506, "mean_data_wait_seconds": 0.0001004338264465332}
{"step": 14100, "total_steps": 46238, "tokens": 7624015872, "loss": 3.2226746533856248, "ce_loss": 3.2214246139381872, "z_loss": 12.500162399176395, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19251681864261627, "tokens_per_second": 4252668.911736635, "eta_seconds": 4085.805037877837, "mean_data_wait_seconds": 8.499622344970703e-05}
{"step": 14110, "total_steps": 46238, "tokens": 7629422592, "loss": 3.244194310722929, "ce_loss": 3.2429572903748713, "z_loss": 12.370484366561428, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18487082421779633, "tokens_per_second": 4252876.018301944, "eta_seconds": 4084.334758231543, "mean_data_wait_seconds": 8.004903793334961e-05}
{"step": 14120, "total_steps": 46238, "tokens": 7634829312, "loss": 3.2675673004352683, "ce_loss": 3.26633134393981, "z_loss": 12.359480677228985, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17334206402301788, "tokens_per_second": 4252912.869330758, "eta_seconds": 4083.028069825595, "mean_data_wait_seconds": 0.0001271665096282959}
{"step": 14130, "total_steps": 46238, "tokens": 7640236032, "loss": 3.253490498571685, "ce_loss": 3.2522820523290923, "z_loss": 12.08453563487891, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16670124232769012, "tokens_per_second": 4252685.536517964, "eta_seconds": 4081.9749673317215, "mean_data_wait_seconds": 0.00010654330253601074}
{"step": 14140, "total_steps": 46238, "tokens": 7645642752, "loss": 3.235669013225671, "ce_loss": 3.234470425230084, "z_loss": 11.986020622831402, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16238249838352203, "tokens_per_second": 4253056.828460148, "eta_seconds": 4080.3473557824836, "mean_data_wait_seconds": 7.793307304382324e-05}
{"step": 14150, "total_steps": 46238, "tokens": 7651049472, "loss": 3.2049810525142783, "ce_loss": 3.203739977244175, "z_loss": 12.410831054051718, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17361611127853394, "tokens_per_second": 4253229.118205065, "eta_seconds": 4078.9108655687423, "mean_data_wait_seconds": 7.289648056030273e-05}
{"step": 14160, "total_steps": 46238, "tokens": 7656456192, "loss": 3.2187592549757524, "ce_loss": 3.2175747950871787, "z_loss": 11.844604383815419, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1639488786458969, "tokens_per_second": 4253410.712290117, "eta_seconds": 4077.465572249459, "mean_data_wait_seconds": 9.882450103759766e-05}
{"step": 14170, "total_steps": 46238, "tokens": 7661862912, "loss": 3.262659296844945, "ce_loss": 3.2614017110882383, "z_loss": 12.575994766119754, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2071128636598587, "tokens_per_second": 4253774.485220913, "eta_seconds": 4075.8458362654815, "mean_data_wait_seconds": 0.0001106560230255127}
{"step": 14180, "total_steps": 46238, "tokens": 7667269632, "loss": 3.2664721915216157, "ce_loss": 3.2652751586653967, "z_loss": 11.970023856018528, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1875585913658142, "tokens_per_second": 4253850.669229329, "eta_seconds": 4074.5018224017963, "mean_data_wait_seconds": 7.462501525878906e-05}
{"step": 14190, "total_steps": 46238, "tokens": 7672676352, "loss": 3.2265313231583796, "ce_loss": 3.225306252638499, "z_loss": 12.250544316840895, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16388411819934845, "tokens_per_second": 4254243.199521995, "eta_seconds": 4072.854975932463, "mean_data_wait_seconds": 7.2479248046875e-05}
{"step": 14200, "total_steps": 46238, "tokens": 7678083072, "loss": 3.2484291141683403, "ce_loss": 3.247193065556613, "z_loss": 12.36056416684931, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16256310045719147, "tokens_per_second": 4254115.741480334, "eta_seconds": 4071.7060645774804, "mean_data_wait_seconds": 6.085634231567383e-05}
{"step": 14210, "total_steps": 46238, "tokens": 7683489792, "loss": 3.267410706390034, "ce_loss": 3.266154581850225, "z_loss": 12.561465725754246, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14772073924541473, "tokens_per_second": 4254367.818292177, "eta_seconds": 4070.1939473938505, "mean_data_wait_seconds": 0.0001499950885772705}
{"step": 14220, "total_steps": 46238, "tokens": 7688896512, "loss": 3.248297494469267, "ce_loss": 3.2470750375227495, "z_loss": 12.224930243058639, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18517449498176575, "tokens_per_second": 4254643.046478952, "eta_seconds": 4068.659869909873, "mean_data_wait_seconds": 8.746981620788574e-05}
{"step": 14230, "total_steps": 46238, "tokens": 7694303232, "loss": 3.219783259160591, "ce_loss": 3.2185378923560632, "z_loss": 12.453480749419242, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1771656721830368, "tokens_per_second": 4255015.952614129, "eta_seconds": 4067.0326261334585, "mean_data_wait_seconds": 0.002723187208175659}
{"step": 14240, "total_steps": 46238, "tokens": 7699709952, "loss": 3.231308037584478, "ce_loss": 3.2300606604778404, "z_loss": 12.473795435645364, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1505119651556015, "tokens_per_second": 4255168.147811087, "eta_seconds": 4065.616535717697, "mean_data_wait_seconds": 9.205937385559082e-05}
{"step": 14250, "total_steps": 46238, "tokens": 7705116672, "loss": 3.1414154984734277, "ce_loss": 3.140219717314749, "z_loss": 11.957791942538638, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16784526407718658, "tokens_per_second": 4255424.334553609, "eta_seconds": 4064.1012280657033, "mean_data_wait_seconds": 6.607174873352051e-05}
{"step": 14260, "total_steps": 46238, "tokens": 7710523392, "loss": 3.1856649138710718, "ce_loss": 3.1844252781434492, "z_loss": 12.396408102729104, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1868564486503601, "tokens_per_second": 4255865.451497405, "eta_seconds": 4062.4095712229173, "mean_data_wait_seconds": 9.357929229736328e-05}
{"step": 14270, "total_steps": 46238, "tokens": 7715930112, "loss": 3.2612795450470666, "ce_loss": 3.260092495065747, "z_loss": 11.870416128274167, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19903281331062317, "tokens_per_second": 4256184.201008566, "eta_seconds": 4060.835012710301, "mean_data_wait_seconds": 0.0001316070556640625}
{"step": 14280, "total_steps": 46238, "tokens": 7721336832, "loss": 3.2519401185440295, "ce_loss": 3.2507383967890884, "z_loss": 12.016895474809589, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.148210346698761, "tokens_per_second": 4256348.801439757, "eta_seconds": 4059.4077013038586, "mean_data_wait_seconds": 0.0001499950885772705}
{"step": 14290, "total_steps": 46238, "tokens": 7726743552, "loss": 3.1900117993354797, "ce_loss": 3.1887756802818994, "z_loss": 12.361204812020967, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1714981347322464, "tokens_per_second": 4256416.4230480585, "eta_seconds": 4058.072957915794, "mean_data_wait_seconds": 7.656216621398926e-05}
{"step": 14300, "total_steps": 46238, "tokens": 7732150272, "loss": 3.262621917507865, "ce_loss": 3.2613792870983933, "z_loss": 12.426222208774451, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17497947812080383, "tokens_per_second": 4256432.369670428, "eta_seconds": 4056.787507547548, "mean_data_wait_seconds": 5.888938903808594e-05}
{"step": 14310, "total_steps": 46238, "tokens": 7737556992, "loss": 3.2664825211871755, "ce_loss": 3.265214600346305, "z_loss": 12.679416013486458, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2022022008895874, "tokens_per_second": 4256717.670412255, "eta_seconds": 4055.2454450962455, "mean_data_wait_seconds": 0.00013497471809387207}
{"step": 14320, "total_steps": 46238, "tokens": 7742963712, "loss": 3.195390163045941, "ce_loss": 3.194216338070956, "z_loss": 11.73779093135487, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18093939125537872, "tokens_per_second": 4256767.070176736, "eta_seconds": 4053.92823697152, "mean_data_wait_seconds": 6.073713302612305e-05}
{"step": 14330, "total_steps": 46238, "tokens": 7748370432, "loss": 3.2864698532855874, "ce_loss": 3.2852590156323984, "z_loss": 12.108429908752441, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19609394669532776, "tokens_per_second": 4257132.077197154, "eta_seconds": 4052.310614557677, "mean_data_wait_seconds": 6.034970283508301e-05}
{"step": 14340, "total_steps": 46238, "tokens": 7753777152, "loss": 3.2058599826061362, "ce_loss": 3.20462735674598, "z_loss": 12.326254042712124, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17526818811893463, "tokens_per_second": 4257286.343478286, "eta_seconds": 4050.8937836466584, "mean_data_wait_seconds": 7.480382919311523e-05}
{"step": 14350, "total_steps": 46238, "tokens": 7759183872, "loss": 3.2591281208125027, "ce_loss": 3.2579220316626807, "z_loss": 12.061158389756173, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1793433427810669, "tokens_per_second": 4257660.504199565, "eta_seconds": 4049.2679129758785, "mean_data_wait_seconds": 6.306171417236328e-05}
{"step": 14360, "total_steps": 46238, "tokens": 7764590592, "loss": 3.2246464508952517, "ce_loss": 3.22344402291558, "z_loss": 12.02433526877201, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18821509182453156, "tokens_per_second": 4257929.639821992, "eta_seconds": 4047.742166242214, "mean_data_wait_seconds": 6.324052810668945e-05}
{"step": 14370, "total_steps": 46238, "tokens": 7769997312, "loss": 3.2923476497332254, "ce_loss": 3.291147923830784, "z_loss": 11.997451160893295, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1896858513355255, "tokens_per_second": 4258262.893485595, "eta_seconds": 4046.155687183687, "mean_data_wait_seconds": 0.0001487135887145996}
{"step": 14380, "total_steps": 46238, "tokens": 7775404032, "loss": 3.289011250842701, "ce_loss": 3.287753731915445, "z_loss": 12.57503799236182, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20524875819683075, "tokens_per_second": 4258229.255375942, "eta_seconds": 4044.917939130392, "mean_data_wait_seconds": 0.00013259053230285645}
{"step": 14390, "total_steps": 46238, "tokens": 7780810752, "loss": 3.217886050542196, "ce_loss": 3.2167198025819026, "z_loss": 11.662785219423698, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18261566758155823, "tokens_per_second": 4258335.124405125, "eta_seconds": 4043.5476966847236, "mean_data_wait_seconds": 6.514787673950195e-05}
{"step": 14400, "total_steps": 46238, "tokens": 7786217472, "loss": 3.232960446314378, "ce_loss": 3.231699246348757, "z_loss": 12.612233978329282, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1804082691669464, "tokens_per_second": 4258314.002020848, "eta_seconds": 4042.2980681629233, "mean_data_wait_seconds": 6.785988807678223e-05}
{"step": 14410, "total_steps": 46238, "tokens": 7791624192, "loss": 3.2708468852621135, "ce_loss": 3.2696680769775854, "z_loss": 11.788202748154148, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1976923644542694, "tokens_per_second": 4258352.708891109, "eta_seconds": 4040.991651084022, "mean_data_wait_seconds": 6.0230493545532227e-05}
{"step": 14420, "total_steps": 46238, "tokens": 7797030912, "loss": 3.2557045835437197, "ce_loss": 3.2544591173981177, "z_loss": 12.45454310648369, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16922251880168915, "tokens_per_second": 4258469.575967391, "eta_seconds": 4039.611113129091, "mean_data_wait_seconds": 6.002187728881836e-05}
{"step": 14430, "total_steps": 46238, "tokens": 7802437632, "loss": 3.2495738141464465, "ce_loss": 3.2483637459350354, "z_loss": 12.100536563179709, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22095519304275513, "tokens_per_second": 4258472.569342639, "eta_seconds": 4038.3386352667394, "mean_data_wait_seconds": 8.413195610046387e-05}
{"step": 14440, "total_steps": 46238, "tokens": 7807844352, "loss": 3.258640143004331, "ce_loss": 3.2574501380775915, "z_loss": 11.900312568202164, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20820368826389313, "tokens_per_second": 4258383.719819982, "eta_seconds": 4037.1532288139506, "mean_data_wait_seconds": 6.86347484588623e-05}
{"step": 14450, "total_steps": 46238, "tokens": 7813251072, "loss": 3.240264710151788, "ce_loss": 3.239010324983886, "z_loss": 12.543599569436276, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22927263379096985, "tokens_per_second": 4258484.744220502, "eta_seconds": 4035.787820380202, "mean_data_wait_seconds": 9.930133819580078e-05}
{"step": 14460, "total_steps": 46238, "tokens": 7818657792, "loss": 3.2298308484482043, "ce_loss": 3.22858970634865, "z_loss": 12.411403576532999, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18459452688694, "tokens_per_second": 4258513.493809655, "eta_seconds": 4034.4909483027095, "mean_data_wait_seconds": 6.261467933654785e-05}
{"step": 14470, "total_steps": 46238, "tokens": 7824064512, "loss": 3.215484888264627, "ce_loss": 3.21426271127932, "z_loss": 12.221708088210136, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16999638080596924, "tokens_per_second": 4258498.987099131, "eta_seconds": 4033.2350614693673, "mean_data_wait_seconds": 5.6684017181396484e-05}
{"step": 14480, "total_steps": 46238, "tokens": 7829471232, "loss": 3.229008902202953, "ce_loss": 3.227803799239072, "z_loss": 12.051402019731926, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15396305918693542, "tokens_per_second": 4258911.835834166, "eta_seconds": 4031.5745819229896, "mean_data_wait_seconds": 5.8144330978393555e-05}
{"step": 14490, "total_steps": 46238, "tokens": 7834877952, "loss": 3.2149538217168865, "ce_loss": 3.2137774156801626, "z_loss": 11.763799400040597, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17646753787994385, "tokens_per_second": 4259175.26630932, "eta_seconds": 4030.055799716748, "mean_data_wait_seconds": 6.189942359924316e-05}
{"step": 14500, "total_steps": 46238, "tokens": 7840284672, "loss": 3.2804729306336604, "ce_loss": 3.2792868397452613, "z_loss": 11.860819708217274, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17070122063159943, "tokens_per_second": 4259051.19158543, "eta_seconds": 4028.9037375041403, "mean_data_wait_seconds": 9.909272193908691e-05}
{"step": 14510, "total_steps": 46238, "tokens": 7845691392, "loss": 3.2963839119130913, "ce_loss": 3.2951960852651885, "z_loss": 11.877987139152758, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1792125552892685, "tokens_per_second": 4258977.3727566805, "eta_seconds": 4027.7040807326257, "mean_data_wait_seconds": 7.50124454498291e-05}
{"step": 14520, "total_steps": 46238, "tokens": 7851098112, "loss": 3.2321802323514763, "ce_loss": 3.2309495011965432, "z_loss": 12.307431929039232, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18271008133888245, "tokens_per_second": 4258823.993697591, "eta_seconds": 4026.5796025797617, "mean_data_wait_seconds": 7.686018943786621e-05}
{"step": 14530, "total_steps": 46238, "tokens": 7856504832, "loss": 3.2709634159550522, "ce_loss": 3.269752244154612, "z_loss": 12.112004706353853, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1826445758342743, "tokens_per_second": 4258842.624055265, "eta_seconds": 4025.2924602497687, "mean_data_wait_seconds": 7.507205009460449e-05}
{"step": 14540, "total_steps": 46238, "tokens": 7861911552, "loss": 3.2220490177472434, "ce_loss": 3.2208031849427656, "z_loss": 12.458294456655329, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19011902809143066, "tokens_per_second": 4259164.616268816, "eta_seconds": 4023.718716703003, "mean_data_wait_seconds": 9.384751319885254e-05}
{"step": 14550, "total_steps": 46238, "tokens": 7867318272, "loss": 3.247550090154012, "ce_loss": 3.24631400903066, "z_loss": 12.36059401252053, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22315779328346252, "tokens_per_second": 4259365.607932962, "eta_seconds": 4022.259472652822, "mean_data_wait_seconds": 5.7816505432128906e-05}
{"step": 14560, "total_steps": 46238, "tokens": 7872724992, "loss": 3.254325312195402, "ce_loss": 3.2531349550593984, "z_loss": 11.903468001972545, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16344760358333588, "tokens_per_second": 4259888.091573371, "eta_seconds": 4020.4969181888223, "mean_data_wait_seconds": 0.00012135505676269531}
{"step": 14570, "total_steps": 46238, "tokens": 7878131712, "loss": 3.225845304402438, "ce_loss": 3.224637204950506, "z_loss": 12.081000089645386, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18352283537387848, "tokens_per_second": 4260276.857077548, "eta_seconds": 4018.860932842033, "mean_data_wait_seconds": 6.881356239318848e-05}
{"step": 14580, "total_steps": 46238, "tokens": 7883538432, "loss": 3.225853802579822, "ce_loss": 3.224614437782403, "z_loss": 12.393750248533307, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1679178774356842, "tokens_per_second": 4260745.397168185, "eta_seconds": 4017.150030925534, "mean_data_wait_seconds": 7.551908493041992e-05}
{"step": 14590, "total_steps": 46238, "tokens": 7888945152, "loss": 3.264489381602316, "ce_loss": 3.2632202740871543, "z_loss": 12.690674593954375, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18945784866809845, "tokens_per_second": 4261112.052421015, "eta_seconds": 4015.5355159642727, "mean_data_wait_seconds": 6.386637687683105e-05}
{"step": 14600, "total_steps": 46238, "tokens": 7894351872, "loss": 3.181215192332412, "ce_loss": 3.180008028492783, "z_loss": 12.07171679988052, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17671535909175873, "tokens_per_second": 4261420.093620132, "eta_seconds": 4013.976488637823, "mean_data_wait_seconds": 5.677342414855957e-05}
{"step": 14610, "total_steps": 46238, "tokens": 7899758592, "loss": 3.198841331583081, "ce_loss": 3.1976277485038294, "z_loss": 12.136117313847397, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15980495512485504, "tokens_per_second": 4261677.899218694, "eta_seconds": 4012.4649840699976, "mean_data_wait_seconds": 5.6862831115722656e-05}
{"step": 14620, "total_steps": 46238, "tokens": 7905165312, "loss": 3.1947541923233955, "ce_loss": 3.1934922698772317, "z_loss": 12.618935881238995, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16726894676685333, "tokens_per_second": 4262096.776291279, "eta_seconds": 4010.8020819918934, "mean_data_wait_seconds": 7.039308547973633e-05}
{"step": 14630, "total_steps": 46238, "tokens": 7910572032, "loss": 3.2553094553224966, "ce_loss": 3.254095778320775, "z_loss": 12.13654098366246, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2124595195055008, "tokens_per_second": 4262290.018920836, "eta_seconds": 4009.3517400598535, "mean_data_wait_seconds": 0.00010356307029724121}
{"step": 14640, "total_steps": 46238, "tokens": 7915978752, "loss": 3.256508343147509, "ce_loss": 3.255293056820378, "z_loss": 12.152896563212076, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1919715255498886, "tokens_per_second": 4262382.309532987, "eta_seconds": 4007.996454422172, "mean_data_wait_seconds": 7.236003875732422e-05}
{"step": 14650, "total_steps": 46238, "tokens": 7921385472, "loss": 3.1616999904314675, "ce_loss": 3.160489385778254, "z_loss": 12.106170849366622, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15667842328548431, "tokens_per_second": 4262584.6911573885, "eta_seconds": 4006.537746787356, "mean_data_wait_seconds": 0.00015863776206970215}
{"step": 14660, "total_steps": 46238, "tokens": 7926792192, "loss": 3.2603276028777612, "ce_loss": 3.2590551737583047, "z_loss": 12.724008170041172, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2027384638786316, "tokens_per_second": 4262917.448051568, "eta_seconds": 4004.956688007971, "mean_data_wait_seconds": 0.00011307001113891602}
{"step": 14670, "total_steps": 46238, "tokens": 7932198912, "loss": 3.2248677737785107, "ce_loss": 3.2235946275971155, "z_loss": 12.730990756641734, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1701269894838333, "tokens_per_second": 4263419.079177112, "eta_seconds": 4003.2173021316494, "mean_data_wait_seconds": 7.134675979614258e-05}
{"step": 14680, "total_steps": 46238, "tokens": 7937605632, "loss": 3.220441921190782, "ce_loss": 3.2192212469650037, "z_loss": 12.206685528610691, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1750975400209427, "tokens_per_second": 4263829.351739224, "eta_seconds": 4001.5640628395186, "mean_data_wait_seconds": 7.215142250061035e-05}
{"step": 14690, "total_steps": 46238, "tokens": 7943012352, "loss": 3.2889630288788765, "ce_loss": 3.2877190763300117, "z_loss": 12.439618493571427, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20122551918029785, "tokens_per_second": 4264193.587215274, "eta_seconds": 3999.9543255114686, "mean_data_wait_seconds": 6.857514381408691e-05}
{"step": 14700, "total_steps": 46238, "tokens": 7948419072, "loss": 3.2445342233686736, "ce_loss": 3.243289640455535, "z_loss": 12.445818929961234, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17892757058143616, "tokens_per_second": 4264215.096911723, "eta_seconds": 3998.6662202732196, "mean_data_wait_seconds": 0.0001156926155090332}
{"step": 14710, "total_steps": 46238, "tokens": 7953825792, "loss": 3.268826428687934, "ce_loss": 3.26760976784157, "z_loss": 12.166583335760867, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18365155160427094, "tokens_per_second": 4264386.658295525, "eta_seconds": 3997.23747161642, "mean_data_wait_seconds": 9.146332740783691e-05}
{"step": 14720, "total_steps": 46238, "tokens": 7959232512, "loss": 3.2623881354476465, "ce_loss": 3.2611208232966336, "z_loss": 12.672902685223203, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1854012906551361, "tokens_per_second": 4264682.032298048, "eta_seconds": 3995.6928312467194, "mean_data_wait_seconds": 7.352232933044434e-05}
{"step": 14730, "total_steps": 46238, "tokens": 7964639232, "loss": 3.25249147234541, "ce_loss": 3.2512830134594077, "z_loss": 12.08467901836742, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1623399704694748, "tokens_per_second": 4265006.35085804, "eta_seconds": 3994.121298453139, "mean_data_wait_seconds": 5.97834587097168e-05}
{"step": 14740, "total_steps": 46238, "tokens": 7970045952, "loss": 3.238861757697481, "ce_loss": 3.237676199638482, "z_loss": 11.855729088638768, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18216265738010406, "tokens_per_second": 4263429.513452473, "eta_seconds": 3994.330369545545, "mean_data_wait_seconds": 0.00010356307029724121}
{"step": 14750, "total_steps": 46238, "tokens": 7975452672, "loss": 3.286842998230096, "ce_loss": 3.285647182753592, "z_loss": 11.958033207691077, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18672209978103638, "tokens_per_second": 4262289.635414298, "eta_seconds": 3994.130085048817, "mean_data_wait_seconds": 7.343292236328125e-05}
{"step": 14760, "total_steps": 46238, "tokens": 7980859392, "loss": 3.287458725047834, "ce_loss": 3.286237985798807, "z_loss": 12.207441951289322, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16920705139636993, "tokens_per_second": 4262554.187521499, "eta_seconds": 3992.613769889855, "mean_data_wait_seconds": 7.18832015991211e-05}
{"step": 14770, "total_steps": 46238, "tokens": 7986266112, "loss": 3.261909154328433, "ce_loss": 3.260697668248957, "z_loss": 12.115066969033444, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1696195900440216, "tokens_per_second": 4262760.203340187, "eta_seconds": 3991.152448751118, "mean_data_wait_seconds": 8.26716423034668e-05}
{"step": 14780, "total_steps": 46238, "tokens": 7991672832, "loss": 3.2713205633741436, "ce_loss": 3.2701195297819194, "z_loss": 12.010488943620162, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1794177144765854, "tokens_per_second": 4261687.520796502, "eta_seconds": 3990.8883560804215, "mean_data_wait_seconds": 0.0682709813117981}
{"step": 14790, "total_steps": 46238, "tokens": 7997079552, "loss": 3.2797319256898128, "ce_loss": 3.2784956711711306, "z_loss": 12.362782312161995, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16490404307842255, "tokens_per_second": 4261820.70947627, "eta_seconds": 3989.4949935820787, "mean_data_wait_seconds": 0.00014290213584899902}
{"step": 14800, "total_steps": 46238, "tokens": 8002486272, "loss": 3.2814178611292983, "ce_loss": 3.280184704245943, "z_loss": 12.33188435525605, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1857306957244873, "tokens_per_second": 4261812.8065095935, "eta_seconds": 3988.2337483331553, "mean_data_wait_seconds": 9.530782699584961e-05}
{"step": 14810, "total_steps": 46238, "tokens": 8007892992, "loss": 3.293957618149844, "ce_loss": 3.292694395238703, "z_loss": 12.632463614145914, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22204995155334473, "tokens_per_second": 4261884.353093101, "eta_seconds": 3986.898173731092, "mean_data_wait_seconds": 0.00011959671974182129}
{"step": 14820, "total_steps": 46238, "tokens": 8013299712, "loss": 3.338571181803039, "ce_loss": 3.337376108675292, "z_loss": 11.950615080920132, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18407122790813446, "tokens_per_second": 4261973.226645244, "eta_seconds": 3985.5464407434897, "mean_data_wait_seconds": 0.00017565488815307617}
{"step": 14830, "total_steps": 46238, "tokens": 8018706432, "loss": 3.3157830075784163, "ce_loss": 3.3145573590741013, "z_loss": 12.256301598115401, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1908298134803772, "tokens_per_second": 4262018.323238262, "eta_seconds": 3984.2356874472566, "mean_data_wait_seconds": 0.00017467141151428223}
{"step": 14840, "total_steps": 46238, "tokens": 8024113152, "loss": 3.259530518994187, "ce_loss": 3.2582592205567793, "z_loss": 12.712558406772036, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2130134105682373, "tokens_per_second": 4261953.5164791, "eta_seconds": 3983.0276699084793, "mean_data_wait_seconds": 0.00012955069541931152}
{"step": 14850, "total_steps": 46238, "tokens": 8029519872, "loss": 3.2749723557269936, "ce_loss": 3.273757091074279, "z_loss": 12.152738535042966, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16514506936073303, "tokens_per_second": 4262166.551625685, "eta_seconds": 3981.560048967875, "mean_data_wait_seconds": 0.000149458646774292}
{"step": 14860, "total_steps": 46238, "tokens": 8034926592, "loss": 3.2600456711017722, "ce_loss": 3.2588138219081992, "z_loss": 12.318458795547485, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18039079010486603, "tokens_per_second": 4262095.30540968, "eta_seconds": 3980.3580465381747, "mean_data_wait_seconds": 0.0001105964183807373}
{"step": 14870, "total_steps": 46238, "tokens": 8040333312, "loss": 3.2594803246584805, "ce_loss": 3.258232461683678, "z_loss": 12.478832700035788, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1972837746143341, "tokens_per_second": 4262239.80814574, "eta_seconds": 3978.9545842982534, "mean_data_wait_seconds": 0.00012442469596862793}
{"step": 14880, "total_steps": 46238, "tokens": 8045740032, "loss": 3.2955055345188486, "ce_loss": 3.2943121194839478, "z_loss": 11.934148521134347, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18671730160713196, "tokens_per_second": 4262498.500425415, "eta_seconds": 3977.4446612257893, "mean_data_wait_seconds": 0.0001201629638671875}
{"step": 14890, "total_steps": 46238, "tokens": 8051146752, "loss": 3.237777151844718, "ce_loss": 3.2365650354009685, "z_loss": 12.120819250742594, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17831192910671234, "tokens_per_second": 4262489.5762802325, "eta_seconds": 3976.184546776178, "mean_data_wait_seconds": 7.686018943786621e-05}
{"step": 14900, "total_steps": 46238, "tokens": 8056553472, "loss": 3.2715766393777095, "ce_loss": 3.270329217116038, "z_loss": 12.474400787642509, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19797460734844208, "tokens_per_second": 4262621.178569294, "eta_seconds": 3974.7933851552725, "mean_data_wait_seconds": 8.219480514526367e-05}
{"step": 14910, "total_steps": 46238, "tokens": 8061960192, "loss": 3.2951023452209705, "ce_loss": 3.2938664537487607, "z_loss": 12.358935218868833, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18620984256267548, "tokens_per_second": 4262471.593684936, "eta_seconds": 3973.664427252476, "mean_data_wait_seconds": 9.417533874511719e-05}
{"step": 14920, "total_steps": 46238, "tokens": 8067366912, "loss": 3.2374059937217017, "ce_loss": 3.236211429942738, "z_loss": 11.945470203052867, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2174767106771469, "tokens_per_second": 4262340.609536333, "eta_seconds": 3972.518054074971, "mean_data_wait_seconds": 9.277462959289551e-05}
{"step": 14930, "total_steps": 46238, "tokens": 8072773632, "loss": 3.197022577126821, "ce_loss": 3.1958743315754514, "z_loss": 11.482344822450118, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17235031723976135, "tokens_per_second": 4262372.962126055, "eta_seconds": 3971.2194250493203, "mean_data_wait_seconds": 0.00011354684829711914}
{"step": 14940, "total_steps": 46238, "tokens": 8078180352, "loss": 3.2097917957739397, "ce_loss": 3.2085706681916206, "z_loss": 12.211098678184278, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18620198965072632, "tokens_per_second": 4262605.34604725, "eta_seconds": 3969.7345192163684, "mean_data_wait_seconds": 0.00012022256851196289}
{"step": 14950, "total_steps": 46238, "tokens": 8083587072, "loss": 3.23123371962345, "ce_loss": 3.230009203607386, "z_loss": 12.245202223459879, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17506900429725647, "tokens_per_second": 4262800.053913729, "eta_seconds": 3968.2848480001326, "mean_data_wait_seconds": 0.00017645955085754395}
{"step": 14960, "total_steps": 46238, "tokens": 8088993792, "loss": 3.263290499195908, "ce_loss": 3.2620643413428105, "z_loss": 12.261301943750093, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2099909782409668, "tokens_per_second": 4263030.424210124, "eta_seconds": 3966.802124602074, "mean_data_wait_seconds": 9.97781753540039e-05}
{"step": 14970, "total_steps": 46238, "tokens": 8094400512, "loss": 3.2081681616378552, "ce_loss": 3.2069707187739285, "z_loss": 11.97413512432214, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17736601829528809, "tokens_per_second": 4263355.433705046, "eta_seconds": 3965.2315381334824, "mean_data_wait_seconds": 0.00013244152069091797}
{"step": 14980, "total_steps": 46238, "tokens": 8099807232, "loss": 3.2023596402370567, "ce_loss": 3.2011582616603738, "z_loss": 12.014002604918046, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19081835448741913, "tokens_per_second": 4263633.602459534, "eta_seconds": 3963.7047363195406, "mean_data_wait_seconds": 0.00012135505676269531}
{"step": 14990, "total_steps": 46238, "tokens": 8105213952, "loss": 3.261160543470672, "ce_loss": 3.2599184982704394, "z_loss": 12.420615687514797, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19428883492946625, "tokens_per_second": 4264008.976580143, "eta_seconds": 3962.08780909973, "mean_data_wait_seconds": 0.00011542439460754395}
{"step": 15000, "total_steps": 46238, "tokens": 8110620672, "loss": 3.2408782081170515, "ce_loss": 3.2396514072562708, "z_loss": 12.267789414434722, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16195014119148254, "tokens_per_second": 4264264.197315337, "eta_seconds": 3960.5827600064813, "mean_data_wait_seconds": 0.00015059113502502441}
{"step": 15010, "total_steps": 46238, "tokens": 8116027392, "loss": 3.2203314611406038, "ce_loss": 3.219106314760266, "z_loss": 12.251855756297257, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17242899537086487, "tokens_per_second": 4264556.034980096, "eta_seconds": 3959.0438970697687, "mean_data_wait_seconds": 8.630752563476562e-05}
{"step": 15020, "total_steps": 46238, "tokens": 8121434112, "loss": 3.2412176222512215, "ce_loss": 3.2399873823830574, "z_loss": 12.30233206170978, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21268758177757263, "tokens_per_second": 4264733.401167376, "eta_seconds": 3957.611469776746, "mean_data_wait_seconds": 8.043646812438965e-05}
{"step": 15030, "total_steps": 46238, "tokens": 8126840832, "loss": 3.2331055800120034, "ce_loss": 3.2317925384550383, "z_loss": 13.130541230693009, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1761181801557541, "tokens_per_second": 4264797.363490203, "eta_seconds": 3956.2843591217575, "mean_data_wait_seconds": 9.28044319152832e-05}
{"step": 15040, "total_steps": 46238, "tokens": 8132247552, "loss": 3.191845827030413, "ce_loss": 3.1905922745213364, "z_loss": 12.535141272978349, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18788881599903107, "tokens_per_second": 4264899.20368843, "eta_seconds": 3954.9221630871248, "mean_data_wait_seconds": 5.885958671569824e-05}
{"step": 15050, "total_steps": 46238, "tokens": 8137654272, "loss": 3.2109331073183003, "ce_loss": 3.2096729404998547, "z_loss": 12.601548888466574, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18407754600048065, "tokens_per_second": 4265190.865770049, "eta_seconds": 3953.3840793208437, "mean_data_wait_seconds": 6.22868537902832e-05}
{"step": 15060, "total_steps": 46238, "tokens": 8143060992, "loss": 3.2266889413197837, "ce_loss": 3.225486061789773, "z_loss": 12.028478463490805, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2070043981075287, "tokens_per_second": 4265285.992208192, "eta_seconds": 3952.0282988745525, "mean_data_wait_seconds": 6.654858589172363e-05}
{"step": 15070, "total_steps": 46238, "tokens": 8148467712, "loss": 3.230958656831221, "ce_loss": 3.229715732010928, "z_loss": 12.429204261664188, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.13647247850894928, "tokens_per_second": 4265501.55171945, "eta_seconds": 3950.561034776135, "mean_data_wait_seconds": 5.945563316345215e-05}
{"step": 15080, "total_steps": 46238, "tokens": 8153874432, "loss": 3.251238086006858, "ce_loss": 3.250037431716919, "z_loss": 12.006343400839603, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2001623809337616, "tokens_per_second": 4265643.557715644, "eta_seconds": 3949.1620141419626, "mean_data_wait_seconds": 0.00011107325553894043}
{"step": 15090, "total_steps": 46238, "tokens": 8159281152, "loss": 3.2629012176484773, "ce_loss": 3.2617077791329585, "z_loss": 11.93434797633778, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17282332479953766, "tokens_per_second": 4265893.295200139, "eta_seconds": 3947.663389271418, "mean_data_wait_seconds": 0.00012251734733581543}
{"step": 15100, "total_steps": 46238, "tokens": 8164687872, "loss": 3.20456596215566, "ce_loss": 3.2033302061485522, "z_loss": 12.357531829313798, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16953839361667633, "tokens_per_second": 4266115.149537972, "eta_seconds": 3946.190731823835, "mean_data_wait_seconds": 8.907914161682129e-05}
{"step": 15110, "total_steps": 46238, "tokens": 8170094592, "loss": 3.2503978790658894, "ce_loss": 3.2491927056601555, "z_loss": 12.05182559562452, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16249096393585205, "tokens_per_second": 4266296.6870345455, "eta_seconds": 3944.7555054353224, "mean_data_wait_seconds": 0.00015112757682800293}
{"step": 15120, "total_steps": 46238, "tokens": 8175501312, "loss": 3.2617795593810803, "ce_loss": 3.260568329782197, "z_loss": 12.112551436279759, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18061664700508118, "tokens_per_second": 4266536.384158417, "eta_seconds": 3943.266647500672, "mean_data_wait_seconds": 8.320808410644531e-05}
{"step": 15130, "total_steps": 46238, "tokens": 8180908032, "loss": 3.2556294509858796, "ce_loss": 3.2543998830246204, "z_loss": 12.295344020381119, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21845632791519165, "tokens_per_second": 4266728.166270708, "eta_seconds": 3941.8222226939306, "mean_data_wait_seconds": 7.644295692443848e-05}
{"step": 15140, "total_steps": 46238, "tokens": 8186314752, "loss": 3.194086000774846, "ce_loss": 3.1928845878803367, "z_loss": 12.014102603449967, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17550906538963318, "tokens_per_second": 4266837.85756454, "eta_seconds": 3940.4537376999883, "mean_data_wait_seconds": 6.195902824401855e-05}
{"step": 15150, "total_steps": 46238, "tokens": 8191721472, "loss": 3.225119471549988, "ce_loss": 3.2239043893236103, "z_loss": 12.150983066269845, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16856884956359863, "tokens_per_second": 4266874.621872766, "eta_seconds": 3939.1526476638983, "mean_data_wait_seconds": 0.0001361370086669922}
{"step": 15160, "total_steps": 46238, "tokens": 8197128192, "loss": 3.2613924434690764, "ce_loss": 3.2601769656846016, "z_loss": 12.154807740991766, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1958893984556198, "tokens_per_second": 4267256.371188704, "eta_seconds": 3937.5332256682386, "mean_data_wait_seconds": 7.456541061401367e-05}
{"step": 15170, "total_steps": 46238, "tokens": 8202534912, "loss": 3.240154038776051, "ce_loss": 3.2389287641554168, "z_loss": 12.252459121472908, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18410168588161469, "tokens_per_second": 4267503.956184591, "eta_seconds": 3936.037832995378, "mean_data_wait_seconds": 6.949901580810547e-05}
{"step": 15180, "total_steps": 46238, "tokens": 8207941632, "loss": 3.2623338320038537, "ce_loss": 3.2611168044986147, "z_loss": 12.170197934815377, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19880296289920807, "tokens_per_second": 4267607.365138647, "eta_seconds": 3934.675537672025, "mean_data_wait_seconds": 7.972121238708496e-05}
{"step": 15190, "total_steps": 46238, "tokens": 8213348352, "loss": 3.2430775436488064, "ce_loss": 3.2418556953921462, "z_loss": 12.218561251958212, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19658219814300537, "tokens_per_second": 4267579.512845886, "eta_seconds": 3933.4342883293803, "mean_data_wait_seconds": 6.386637687683105e-05}
{"step": 15200, "total_steps": 46238, "tokens": 8218755072, "loss": 3.2175640283208904, "ce_loss": 3.2163408994674683, "z_loss": 12.23159898411144, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17596007883548737, "tokens_per_second": 4267751.521919336, "eta_seconds": 3932.0088758244183, "mean_data_wait_seconds": 8.830428123474121e-05}
{"step": 15210, "total_steps": 46238, "tokens": 8224161792, "loss": 3.214339989604372, "ce_loss": 3.2130715485775108, "z_loss": 12.684296969211463, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1686103492975235, "tokens_per_second": 4267846.330777565, "eta_seconds": 3930.654677752575, "mean_data_wait_seconds": 8.374452590942383e-05}
{"step": 15220, "total_steps": 46238, "tokens": 8229568512, "loss": 3.1805078459508493, "ce_loss": 3.1792825586868054, "z_loss": 12.252722053816825, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17136144638061523, "tokens_per_second": 4267753.785048208, "eta_seconds": 3929.4730363201043, "mean_data_wait_seconds": 0.00010851025581359863}
{"step": 15230, "total_steps": 46238, "tokens": 8234975232, "loss": 3.261758414181796, "ce_loss": 3.260490655899048, "z_loss": 12.67777041232947, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20834915339946747, "tokens_per_second": 4267935.697078509, "eta_seconds": 3928.038727358458, "mean_data_wait_seconds": 0.00010788440704345703}
{"step": 15240, "total_steps": 46238, "tokens": 8240381952, "loss": 3.266579795967449, "ce_loss": 3.265301711631544, "z_loss": 12.780930237336593, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2004939615726471, "tokens_per_second": 4268122.611217121, "eta_seconds": 3926.5999388009272, "mean_data_wait_seconds": 8.58306884765625e-05}
{"step": 15250, "total_steps": 46238, "tokens": 8245788672, "loss": 3.2366432746251426, "ce_loss": 3.2353855642405422, "z_loss": 12.577204270796342, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18017663061618805, "tokens_per_second": 4268202.765609919, "eta_seconds": 3925.2594555699156, "mean_data_wait_seconds": 8.764863014221191e-05}
{"step": 15260, "total_steps": 46238, "tokens": 8251195392, "loss": 3.170611681360187, "ce_loss": 3.169393423831824, "z_loss": 12.182537346175224, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1907057911157608, "tokens_per_second": 4268201.213364133, "eta_seconds": 3923.9941386922487, "mean_data_wait_seconds": 7.724761962890625e-05}
{"step": 15270, "total_steps": 46238, "tokens": 8256602112, "loss": 3.184022215279666, "ce_loss": 3.182815757664767, "z_loss": 12.06453751072739, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19123299419879913, "tokens_per_second": 4268222.115670974, "eta_seconds": 3922.7081839361967, "mean_data_wait_seconds": 9.897351264953613e-05}
{"step": 15280, "total_steps": 46238, "tokens": 8262008832, "loss": 3.1737229932438242, "ce_loss": 3.172462172580488, "z_loss": 12.607902230638446, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1638374775648117, "tokens_per_second": 4268253.22710144, "eta_seconds": 3921.412862227589, "mean_data_wait_seconds": 0.00012350082397460938}
{"step": 15290, "total_steps": 46238, "tokens": 8267415552, "loss": 3.270948913964358, "ce_loss": 3.269720642855673, "z_loss": 12.282775062503237, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1529020518064499, "tokens_per_second": 4268269.672171551, "eta_seconds": 3920.1310294640393, "mean_data_wait_seconds": 0.00016483664512634277}
{"step": 15300, "total_steps": 46238, "tokens": 8272822272, "loss": 3.245228622898911, "ce_loss": 3.2439988273562808, "z_loss": 12.29783255403692, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18314675986766815, "tokens_per_second": 4268281.8251912715, "eta_seconds": 3918.8531472498153, "mean_data_wait_seconds": 0.00011819601058959961}
{"step": 15310, "total_steps": 46238, "tokens": 8278228992, "loss": 3.2605557243029275, "ce_loss": 3.259360781221679, "z_loss": 11.949525479114417, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17623235285282135, "tokens_per_second": 4268133.339851377, "eta_seconds": 3917.722716831116, "mean_data_wait_seconds": 0.00011542439460754395}
{"step": 15320, "total_steps": 46238, "tokens": 8283635712, "loss": 3.2623534058079575, "ce_loss": 3.2611382639769353, "z_loss": 12.151314359722715, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18329699337482452, "tokens_per_second": 4268208.370397883, "eta_seconds": 3916.3871051688457, "mean_data_wait_seconds": 7.873773574829102e-05}
{"step": 15330, "total_steps": 46238, "tokens": 8289042432, "loss": 3.3008822852914985, "ce_loss": 3.299650358431267, "z_loss": 12.31965541117119, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17693260312080383, "tokens_per_second": 4268274.929886221, "eta_seconds": 3915.0593104941936, "mean_data_wait_seconds": 7.849931716918945e-05}
{"step": 15340, "total_steps": 46238, "tokens": 8294449152, "loss": 3.266177504351645, "ce_loss": 3.2649669647216797, "z_loss": 12.105696165200436, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20222590863704681, "tokens_per_second": 4268591.329528518, "eta_seconds": 3913.5024869773947, "mean_data_wait_seconds": 9.420514106750488e-05}
{"step": 15350, "total_steps": 46238, "tokens": 8299855872, "loss": 3.3109774932716833, "ce_loss": 3.309733166839137, "z_loss": 12.443187070615364, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.23715440928936005, "tokens_per_second": 4268783.230594767, "eta_seconds": 3912.0599856913404, "mean_data_wait_seconds": 7.316470146179199e-05}
{"step": 15360, "total_steps": 46238, "tokens": 8305262592, "loss": 3.251269508491863, "ce_loss": 3.2500164328199443, "z_loss": 12.531119043176824, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17508918046951294, "tokens_per_second": 4268764.505723016, "eta_seconds": 3910.8105686360464, "mean_data_wait_seconds": 6.878376007080078e-05}
{"step": 15370, "total_steps": 46238, "tokens": 8310669312, "loss": 3.251425878568129, "ce_loss": 3.25016091447888, "z_loss": 12.64977594577905, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1942598819732666, "tokens_per_second": 4268915.172611152, "eta_seconds": 3909.4060081292146, "mean_data_wait_seconds": 7.358193397521973e-05}
{"step": 15380, "total_steps": 46238, "tokens": 8316076032, "loss": 3.269463219425895, "ce_loss": 3.2682142095132307, "z_loss": 12.490128076437747, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20188815891742706, "tokens_per_second": 4269100.413355979, "eta_seconds": 3907.9698973126133, "mean_data_wait_seconds": 7.575750350952148e-05}
{"step": 15390, "total_steps": 46238, "tokens": 8321482752, "loss": 3.3042880856629573, "ce_loss": 3.3030280579220164, "z_loss": 12.60035126137011, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21530936658382416, "tokens_per_second": 4269185.053096116, "eta_seconds": 3906.62596644871, "mean_data_wait_seconds": 8.445978164672852e-05}
{"step": 15400, "total_steps": 46238, "tokens": 8326889472, "loss": 3.2972097595532737, "ce_loss": 3.295988055792722, "z_loss": 12.21687818296028, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21150992810726166, "tokens_per_second": 4269263.577439249, "eta_seconds": 3905.287682893655, "mean_data_wait_seconds": 6.580352783203125e-05}
{"step": 15410, "total_steps": 46238, "tokens": 8332296192, "loss": 3.2534215648969016, "ce_loss": 3.2521922678658455, "z_loss": 12.29309252536658, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17652066051959991, "tokens_per_second": 4269500.765320537, "eta_seconds": 3903.8043696775603, "mean_data_wait_seconds": 7.295608520507812e-05}
{"step": 15420, "total_steps": 46238, "tokens": 8337702912, "loss": 3.202895877939282, "ce_loss": 3.2016354781208616, "z_loss": 12.60427678715099, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18430237472057343, "tokens_per_second": 4269640.345529615, "eta_seconds": 3902.4104316995404, "mean_data_wait_seconds": 6.273388862609863e-05}
{"step": 15430, "total_steps": 46238, "tokens": 8343109632, "loss": 3.2720482277147696, "ce_loss": 3.270826854489066, "z_loss": 12.214058428099662, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20626351237297058, "tokens_per_second": 4270002.965949498, "eta_seconds": 3900.812818357419, "mean_data_wait_seconds": 9.965896606445312e-05}
{"step": 15440, "total_steps": 46238, "tokens": 8348516352, "loss": 3.321595175699754, "ce_loss": 3.320360355304949, "z_loss": 12.34832501411438, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16780707240104675, "tokens_per_second": 4270072.663348406, "eta_seconds": 3899.48295890191, "mean_data_wait_seconds": 9.02414321899414e-05}
{"step": 15450, "total_steps": 46238, "tokens": 8353923072, "loss": 3.2211702755003264, "ce_loss": 3.2199680732958242, "z_loss": 12.022096930128155, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18547122180461884, "tokens_per_second": 4270160.84947329, "eta_seconds": 3898.13626483255, "mean_data_wait_seconds": 6.437301635742188e-05}
{"step": 15460, "total_steps": 46238, "tokens": 8359329792, "loss": 3.273796263969306, "ce_loss": 3.272585695440119, "z_loss": 12.106130607200392, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19222253561019897, "tokens_per_second": 4270515.734935829, "eta_seconds": 3896.546266735637, "mean_data_wait_seconds": 6.428360939025879e-05}
{"step": 15470, "total_steps": 46238, "tokens": 8364736512, "loss": 3.2865041693051658, "ce_loss": 3.2852188402956184, "z_loss": 12.853220809589732, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18994641304016113, "tokens_per_second": 4270589.400972242, "eta_seconds": 3895.2130167823934, "mean_data_wait_seconds": 7.104873657226562e-05}
{"step": 15480, "total_steps": 46238, "tokens": 8370143232, "loss": 3.2341382666067644, "ce_loss": 3.2328923471046216, "z_loss": 12.4589304635019, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21712741255760193, "tokens_per_second": 4271005.446157486, "eta_seconds": 3893.567665421988, "mean_data_wait_seconds": 7.471442222595215e-05}
{"step": 15490, "total_steps": 46238, "tokens": 8375549952, "loss": 3.157122933503353, "ce_loss": 3.155891655069409, "z_loss": 12.313024549773246, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20328250527381897, "tokens_per_second": 4271054.790517643, "eta_seconds": 3892.2567841807527, "mean_data_wait_seconds": 7.322430610656738e-05}
{"step": 15500, "total_steps": 46238, "tokens": 8380956672, "loss": 3.225940036051201, "ce_loss": 3.224727101398237, "z_loss": 12.129341920216879, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2119966447353363, "tokens_per_second": 4271410.525881406, "eta_seconds": 3890.6668331934084, "mean_data_wait_seconds": 7.939338684082031e-05}
{"step": 15510, "total_steps": 46238, "tokens": 8386363392, "loss": 3.242562375285409, "ce_loss": 3.241298606901458, "z_loss": 12.637530991525361, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1484956443309784, "tokens_per_second": 4271601.795221736, "eta_seconds": 3889.2268850021915, "mean_data_wait_seconds": 0.00010263919830322266}
{"step": 15520, "total_steps": 46238, "tokens": 8391770112, "loss": 3.1817935289758625, "ce_loss": 3.1805518594655124, "z_loss": 12.41666593696132, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18301065266132355, "tokens_per_second": 4271655.443621925, "eta_seconds": 3887.9123195194493, "mean_data_wait_seconds": 0.00019338726997375488}
{"step": 15530, "total_steps": 46238, "tokens": 8397176832, "loss": 3.251622324640101, "ce_loss": 3.2504017678174106, "z_loss": 12.205563761971213, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15168042480945587, "tokens_per_second": 4271689.732033168, "eta_seconds": 3886.615401745917, "mean_data_wait_seconds": 0.00011843442916870117}
{"step": 15540, "total_steps": 46238, "tokens": 8402583552, "loss": 3.2354460990790166, "ce_loss": 3.234262132283413, "z_loss": 11.83985507849491, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21552971005439758, "tokens_per_second": 4271729.382352455, "eta_seconds": 3885.3136279105715, "mean_data_wait_seconds": 0.0004722177982330322}
{"step": 15550, "total_steps": 46238, "tokens": 8407990272, "loss": 3.266187729257526, "ce_loss": 3.2649672013340574, "z_loss": 12.205163363254432, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22560055553913116, "tokens_per_second": 4272101.229074708, "eta_seconds": 3883.7098594673434, "mean_data_wait_seconds": 0.00010162591934204102}
{"step": 15560, "total_steps": 46238, "tokens": 8413396992, "loss": 3.183585811745037, "ce_loss": 3.182371766278238, "z_loss": 12.1402390870181, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18853212893009186, "tokens_per_second": 4272031.789815722, "eta_seconds": 3882.5073782317195, "mean_data_wait_seconds": 8.019804954528809e-05}
{"step": 15570, "total_steps": 46238, "tokens": 8418803712, "loss": 3.2575354106498486, "ce_loss": 3.256288358659455, "z_loss": 12.47030269015919, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16619378328323364, "tokens_per_second": 4272305.368237623, "eta_seconds": 3880.9932331311265, "mean_data_wait_seconds": 8.842349052429199e-05}
{"step": 15580, "total_steps": 46238, "tokens": 8424210432, "loss": 3.250960501757535, "ce_loss": 3.2497394012682363, "z_loss": 12.210910963289665, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16395433247089386, "tokens_per_second": 4272402.169880143, "eta_seconds": 3879.6398009658815, "mean_data_wait_seconds": 0.0001119077205657959}
{"step": 15590, "total_steps": 46238, "tokens": 8429617152, "loss": 3.2416569193204245, "ce_loss": 3.2404387069470957, "z_loss": 12.182496684970278, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17951489984989166, "tokens_per_second": 4272692.922133927, "eta_seconds": 3878.1103828365917, "mean_data_wait_seconds": 8.45193862915039e-05}
{"step": 15600, "total_steps": 46238, "tokens": 8435023872, "loss": 3.219611249186776, "ce_loss": 3.2183785275979475, "z_loss": 12.327534552776452, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19435235857963562, "tokens_per_second": 4272789.785205899, "eta_seconds": 3876.7570830077193, "mean_data_wait_seconds": 0.00010949373245239258}
{"step": 15610, "total_steps": 46238, "tokens": 8440430592, "loss": 3.2206102591572385, "ce_loss": 3.2193635268644853, "z_loss": 12.46722971309315, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1701163947582245, "tokens_per_second": 4271880.586937083, "eta_seconds": 3876.316532497655, "mean_data_wait_seconds": 0.00017201900482177734}
{"step": 15620, "total_steps": 46238, "tokens": 8445837312, "loss": 3.2954871437766333, "ce_loss": 3.2942571784510757, "z_loss": 12.299748341242472, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19053903222084045, "tokens_per_second": 4272026.236224346, "eta_seconds": 3874.91876422331, "mean_data_wait_seconds": 0.00013303756713867188}
{"step": 15630, "total_steps": 46238, "tokens": 8451244032, "loss": 3.25766982273622, "ce_loss": 3.2564845193516123, "z_loss": 11.852988163630167, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17430365085601807, "tokens_per_second": 4272068.469881819, "eta_seconds": 3873.6148590937237, "mean_data_wait_seconds": 0.00013694167137145996}
{"step": 15640, "total_steps": 46238, "tokens": 8456650752, "loss": 3.2255930521271448, "ce_loss": 3.224363518483711, "z_loss": 12.295669721834587, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16855835914611816, "tokens_per_second": 4272066.159329361, "eta_seconds": 3872.3513557657425, "mean_data_wait_seconds": 8.64267349243164e-05}
{"step": 15650, "total_steps": 46238, "tokens": 8462057472, "loss": 3.236201286315918, "ce_loss": 3.2349898779031, "z_loss": 12.114005507844867, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16342775523662567, "tokens_per_second": 4271244.706921837, "eta_seconds": 3871.8302506058294, "mean_data_wait_seconds": 6.324052810668945e-05}
{"step": 15660, "total_steps": 46238, "tokens": 8467464192, "loss": 3.298149889165705, "ce_loss": 3.296905613306797, "z_loss": 12.442363955757834, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15650057792663574, "tokens_per_second": 4270298.271532189, "eta_seconds": 3871.4222503404308, "mean_data_wait_seconds": 0.0001163184642791748}
{"step": 15670, "total_steps": 46238, "tokens": 8472870912, "loss": 3.1844881718808953, "ce_loss": 3.18328162576213, "z_loss": 12.065470507650664, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17862799763679504, "tokens_per_second": 4270315.086275417, "eta_seconds": 3870.1408889278614, "mean_data_wait_seconds": 7.659196853637695e-05}
{"step": 15680, "total_steps": 46238, "tokens": 8478277632, "loss": 3.2431247360778577, "ce_loss": 3.2418954769770303, "z_loss": 12.2926435398333, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19507043063640594, "tokens_per_second": 4270307.930628785, "eta_seconds": 3868.8812545579835, "mean_data_wait_seconds": 8.973479270935059e-05}
{"step": 15690, "total_steps": 46238, "tokens": 8483684352, "loss": 3.2526432528640283, "ce_loss": 3.251426436684348, "z_loss": 12.168207204703128, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1768270879983902, "tokens_per_second": 4269184.798087206, "eta_seconds": 3868.6326231180947, "mean_data_wait_seconds": 0.00019636750221252441}
{"step": 15700, "total_steps": 46238, "tokens": 8489091072, "loss": 3.257925490538279, "ce_loss": 3.2566790454315417, "z_loss": 12.464493231339889, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1590021401643753, "tokens_per_second": 4269040.700848398, "eta_seconds": 3867.496709675039, "mean_data_wait_seconds": 0.00011348724365234375}
{"step": 15710, "total_steps": 46238, "tokens": 8494497792, "loss": 3.2126244125944194, "ce_loss": 3.2113733092943826, "z_loss": 12.510801004640983, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19830822944641113, "tokens_per_second": 4268869.283590136, "eta_seconds": 3866.3854635809203, "mean_data_wait_seconds": 0.0001690685749053955}
{"step": 15720, "total_steps": 46238, "tokens": 8499904512, "loss": 3.222335575204907, "ce_loss": 3.221052986202818, "z_loss": 12.825636639739528, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1899067461490631, "tokens_per_second": 4268807.594648127, "eta_seconds": 3865.17477261939, "mean_data_wait_seconds": 9.849667549133301e-05}
{"step": 15730, "total_steps": 46238, "tokens": 8505311232, "loss": 3.2555772933092983, "ce_loss": 3.254312518871192, "z_loss": 12.648224577759251, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1775217056274414, "tokens_per_second": 4268749.101794953, "eta_seconds": 3863.9611536467132, "mean_data_wait_seconds": 9.801983833312988e-05}
{"step": 15740, "total_steps": 46238, "tokens": 8510717952, "loss": 3.2166755723230764, "ce_loss": 3.215434137618903, "z_loss": 12.414297024408976, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16281020641326904, "tokens_per_second": 4268662.7867729245, "eta_seconds": 3862.7726779199297, "mean_data_wait_seconds": 8.71419906616211e-05}
{"step": 15750, "total_steps": 46238, "tokens": 8516124672, "loss": 3.204049162792437, "ce_loss": 3.2027591105663413, "z_loss": 12.900429010391235, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19362890720367432, "tokens_per_second": 4268538.2039572075, "eta_seconds": 3861.6187735461226, "mean_data_wait_seconds": 8.240342140197754e-05}
{"step": 15760, "total_steps": 46238, "tokens": 8521531392, "loss": 3.209100033297683, "ce_loss": 3.207875233708006, "z_loss": 12.2480913364526, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15813207626342773, "tokens_per_second": 4268519.631136253, "eta_seconds": 3860.3689259860907, "mean_data_wait_seconds": 0.00014269351959228516}
{"step": 15770, "total_steps": 46238, "tokens": 8526938112, "loss": 3.2211249055284443, "ce_loss": 3.2198468699599756, "z_loss": 12.780334703850023, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16652323305606842, "tokens_per_second": 4268699.439288294, "eta_seconds": 3858.9397211686637, "mean_data_wait_seconds": 0.0001284480094909668}
{"step": 15780, "total_steps": 46238, "tokens": 8532344832, "loss": 3.198273353504412, "ce_loss": 3.197032484141263, "z_loss": 12.408761631358754, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1521262377500534, "tokens_per_second": 4268833.18513557, "eta_seconds": 3857.5522607302423, "mean_data_wait_seconds": 7.24494457244873e-05}
{"step": 15790, "total_steps": 46238, "tokens": 8537751552, "loss": 3.238966399973089, "ce_loss": 3.2377099683790496, "z_loss": 12.564602505077016, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19056262075901031, "tokens_per_second": 4268986.918866149, "eta_seconds": 3856.1468322260157, "mean_data_wait_seconds": 9.638071060180664e-05}
{"step": 15800, "total_steps": 46238, "tokens": 8543158272, "loss": 3.288258129900152, "ce_loss": 3.2869447018160964, "z_loss": 13.134299451654607, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18580438196659088, "tokens_per_second": 4269123.1121110255, "eta_seconds": 3854.7573428639093, "mean_data_wait_seconds": 9.202957153320312e-05}
{"step": 15810, "total_steps": 46238, "tokens": 8548564992, "loss": 3.275056161663749, "ce_loss": 3.273847404754523, "z_loss": 12.08773069670706, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18728768825531006, "tokens_per_second": 4269474.174144913, "eta_seconds": 3853.174014641931, "mean_data_wait_seconds": 9.188055992126465e-05}
{"step": 15820, "total_steps": 46238, "tokens": 8553971712, "loss": 3.228381684332183, "ce_loss": 3.227165453361742, "z_loss": 12.162509007887406, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.182403102517128, "tokens_per_second": 4269545.615667045, "eta_seconds": 3851.843194660574, "mean_data_wait_seconds": 8.004903793334961e-05}
{"step": 15830, "total_steps": 46238, "tokens": 8559378432, "loss": 3.245131801475178, "ce_loss": 3.2439051433043047, "z_loss": 12.266782464403095, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1570020169019699, "tokens_per_second": 4269711.307304307, "eta_seconds": 3850.427422546179, "mean_data_wait_seconds": 6.943941116333008e-05}
{"step": 15840, "total_steps": 46238, "tokens": 8564785152, "loss": 3.223250936378132, "ce_loss": 3.2220020962483957, "z_loss": 12.488145120216139, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1769208312034607, "tokens_per_second": 4269861.230363958, "eta_seconds": 3849.0259746917154, "mean_data_wait_seconds": 0.00015291571617126465}
{"step": 15850, "total_steps": 46238, "tokens": 8570191872, "loss": 3.2435072100523747, "ce_loss": 3.2423025170962014, "z_loss": 12.04690939729864, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1616736352443695, "tokens_per_second": 4269893.323784101, "eta_seconds": 3847.7308021924537, "mean_data_wait_seconds": 6.4849853515625e-05}
{"step": 15860, "total_steps": 46238, "tokens": 8575598592, "loss": 3.2436368808601843, "ce_loss": 3.24237254894141, "z_loss": 12.643250183625655, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1861743927001953, "tokens_per_second": 4270153.752800135, "eta_seconds": 3846.2299708131204, "mean_data_wait_seconds": 7.82310962677002e-05}
{"step": 15870, "total_steps": 46238, "tokens": 8581005312, "loss": 3.2538140470331367, "ce_loss": 3.2525696736393552, "z_loss": 12.444079724225132, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19426526129245758, "tokens_per_second": 4270394.93878943, "eta_seconds": 3844.7466474036087, "mean_data_wait_seconds": 6.365776062011719e-05}
{"step": 15880, "total_steps": 46238, "tokens": 8586412032, "loss": 3.2650656772382334, "ce_loss": 3.2638209787282078, "z_loss": 12.447447942964958, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1909971535205841, "tokens_per_second": 4270809.45761311, "eta_seconds": 3843.107510858861, "mean_data_wait_seconds": 6.642937660217285e-05}
{"step": 15890, "total_steps": 46238, "tokens": 8591818752, "loss": 3.2763113632346643, "ce_loss": 3.275062434601061, "z_loss": 12.489255211570047, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2054579108953476, "tokens_per_second": 4270864.110043899, "eta_seconds": 3841.7923776627367, "mean_data_wait_seconds": 5.78463077545166e-05}
{"step": 15900, "total_steps": 46238, "tokens": 8597225472, "loss": 3.205738472216057, "ce_loss": 3.204484522342682, "z_loss": 12.53971497217814, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17600303888320923, "tokens_per_second": 4271079.595852643, "eta_seconds": 3840.332659669286, "mean_data_wait_seconds": 6.237626075744629e-05}
{"step": 15910, "total_steps": 46238, "tokens": 8602632192, "loss": 3.24877789887515, "ce_loss": 3.2475965384281045, "z_loss": 11.813422997792562, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21176369488239288, "tokens_per_second": 4271367.478281585, "eta_seconds": 3838.808022810687, "mean_data_wait_seconds": 7.101893424987793e-05}
{"step": 15920, "total_steps": 46238, "tokens": 8608038912, "loss": 3.293753784714323, "ce_loss": 3.2925346060232683, "z_loss": 12.19189443732753, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17618952691555023, "tokens_per_second": 4271619.434262326, "eta_seconds": 3837.3158649210727, "mean_data_wait_seconds": 8.296966552734375e-05}
{"step": 15930, "total_steps": 46238, "tokens": 8613445632, "loss": 3.2363489216024224, "ce_loss": 3.23511281338605, "z_loss": 12.361196496269919, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1631513237953186, "tokens_per_second": 4271789.904172048, "eta_seconds": 3835.8970528949594, "mean_data_wait_seconds": 0.0001131594181060791}
{"step": 15940, "total_steps": 46238, "tokens": 8618852352, "loss": 3.244095453710267, "ce_loss": 3.242838529023257, "z_loss": 12.569458643595377, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17024411261081696, "tokens_per_second": 4271977.490531188, "eta_seconds": 3834.4629905723536, "mean_data_wait_seconds": 9.623169898986816e-05}
{"step": 15950, "total_steps": 46238, "tokens": 8624259072, "loss": 3.219406565030416, "ce_loss": 3.2181231704625217, "z_loss": 12.834425160379121, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2116353064775467, "tokens_per_second": 4272044.455941025, "eta_seconds": 3833.1372795587918, "mean_data_wait_seconds": 7.933378219604492e-05}
{"step": 15960, "total_steps": 46238, "tokens": 8629665792, "loss": 3.2274098576921406, "ce_loss": 3.226155241330465, "z_loss": 12.54571117776813, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17903126776218414, "tokens_per_second": 4272080.499356235, "eta_seconds": 3831.8393453650515, "mean_data_wait_seconds": 6.54458999633789e-05}
{"step": 15970, "total_steps": 46238, "tokens": 8635072512, "loss": 3.3188800089287036, "ce_loss": 3.3176632957025007, "z_loss": 12.166950796589706, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16440781950950623, "tokens_per_second": 4272233.899859628, "eta_seconds": 3830.43620915458, "mean_data_wait_seconds": 6.118416786193848e-05}
{"step": 15980, "total_steps": 46238, "tokens": 8640479232, "loss": 3.245102618679856, "ce_loss": 3.2438533414493906, "z_loss": 12.492987452131329, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18312843143939972, "tokens_per_second": 4272296.768979863, "eta_seconds": 3829.1143121844084, "mean_data_wait_seconds": 6.559491157531738e-05}
{"step": 15990, "total_steps": 46238, "tokens": 8645885952, "loss": 3.2861314506241768, "ce_loss": 3.284872470480023, "z_loss": 12.589673432436856, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.188130721449852, "tokens_per_second": 4272562.361794807, "eta_seconds": 3827.6108337784863, "mean_data_wait_seconds": 0.00024706125259399414}
{"step": 16000, "total_steps": 46238, "tokens": 8651292672, "loss": 3.291496562235283, "ce_loss": 3.290208592559352, "z_loss": 12.87971219149503, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17554819583892822, "tokens_per_second": 4272656.118523878, "eta_seconds": 3826.261419242892, "mean_data_wait_seconds": 0.00016990303993225098}
{"step": 16010, "total_steps": 46238, "tokens": 8656699392, "loss": 3.26872807560545, "ce_loss": 3.2674653132756553, "z_loss": 12.627485080198808, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17112557590007782, "tokens_per_second": 4272689.486403011, "eta_seconds": 3824.966124032187, "mean_data_wait_seconds": 0.00013199448585510254}
{"step": 16020, "total_steps": 46238, "tokens": 8662106112, "loss": 3.2200614593245764, "ce_loss": 3.2187865304224417, "z_loss": 12.749197179620916, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19073882699012756, "tokens_per_second": 4272899.699399191, "eta_seconds": 3823.512596445267, "mean_data_wait_seconds": 0.0001277327537536621}
{"step": 16030, "total_steps": 46238, "tokens": 8667512832, "loss": 3.155050640756434, "ce_loss": 3.1538245009653494, "z_loss": 12.26095284837665, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15704604983329773, "tokens_per_second": 4273195.045133651, "eta_seconds": 3821.983066885539, "mean_data_wait_seconds": 9.039044380187988e-05}
{"step": 16040, "total_steps": 46238, "tokens": 8672919552, "loss": 3.225465933481852, "ce_loss": 3.224197613470482, "z_loss": 12.682959455432314, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18066588044166565, "tokens_per_second": 4273557.474148897, "eta_seconds": 3820.3937779616613, "mean_data_wait_seconds": 0.00010019540786743164}
{"step": 16050, "total_steps": 46238, "tokens": 8678326272, "loss": 3.253914116006909, "ce_loss": 3.252624244400949, "z_loss": 12.898787317853985, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18567046523094177, "tokens_per_second": 4273858.312607765, "eta_seconds": 3818.8597913629264, "mean_data_wait_seconds": 6.282329559326172e-05}
{"step": 16060, "total_steps": 46238, "tokens": 8683732992, "loss": 3.264030534209627, "ce_loss": 3.2627679272131487, "z_loss": 12.626168922944503, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17314283549785614, "tokens_per_second": 4274054.168391796, "eta_seconds": 3817.4197848641656, "mean_data_wait_seconds": 8.499622344970703e-05}
{"step": 16070, "total_steps": 46238, "tokens": 8689139712, "loss": 3.3248951182220923, "ce_loss": 3.3236501361384536, "z_loss": 12.44993635379907, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20412838459014893, "tokens_per_second": 4274134.9659909, "eta_seconds": 3816.0826351487576, "mean_data_wait_seconds": 0.00014504790306091309}
{"step": 16080, "total_steps": 46238, "tokens": 8694546432, "loss": 3.2477946082750955, "ce_loss": 3.2464994950727983, "z_loss": 12.950963461037839, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21007601916790009, "tokens_per_second": 4274364.0135860015, "eta_seconds": 3814.6132271782794, "mean_data_wait_seconds": 7.43567943572998e-05}
{"step": 16090, "total_steps": 46238, "tokens": 8699953152, "loss": 3.235179046789805, "ce_loss": 3.2339672392064873, "z_loss": 12.117954514243387, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1999446302652359, "tokens_per_second": 4274411.445503758, "eta_seconds": 3813.305993541063, "mean_data_wait_seconds": 9.188055992126465e-05}
{"step": 16100, "total_steps": 46238, "tokens": 8705359872, "loss": 3.2005964210539153, "ce_loss": 3.1993286790269795, "z_loss": 12.677860267234571, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.162455752491951, "tokens_per_second": 4274585.560136488, "eta_seconds": 3811.885815540846, "mean_data_wait_seconds": 0.00010177493095397949}
{"step": 16110, "total_steps": 46238, "tokens": 8710766592, "loss": 3.223800120931683, "ce_loss": 3.222535653547807, "z_loss": 12.644597176349524, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19336774945259094, "tokens_per_second": 4274769.982627071, "eta_seconds": 3810.456564961107, "mean_data_wait_seconds": 0.00010856986045837402}
{"step": 16120, "total_steps": 46238, "tokens": 8716173312, "loss": 3.2040816816416653, "ce_loss": 3.2028066660418655, "z_loss": 12.750616268678145, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19388705492019653, "tokens_per_second": 4275196.667622396, "eta_seconds": 3808.8115915976905, "mean_data_wait_seconds": 0.00010138750076293945}
{"step": 16130, "total_steps": 46238, "tokens": 8721580032, "loss": 3.2238015965981917, "ce_loss": 3.2225541111194724, "z_loss": 12.474345257787993, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1737951934337616, "tokens_per_second": 4275138.645395585, "eta_seconds": 3807.598596020216, "mean_data_wait_seconds": 0.00010281801223754883}
{"step": 16140, "total_steps": 46238, "tokens": 8726986752, "loss": 3.2482272350426875, "ce_loss": 3.247002491445252, "z_loss": 12.248025193358913, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20877401530742645, "tokens_per_second": 4275062.211009852, "eta_seconds": 3806.401961143882, "mean_data_wait_seconds": 0.00011500716209411621}
{"step": 16150, "total_steps": 46238, "tokens": 8732393472, "loss": 3.2250381491400977, "ce_loss": 3.2237922007387336, "z_loss": 12.459439292098537, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2003527730703354, "tokens_per_second": 4275196.952569029, "eta_seconds": 3805.017323055678, "mean_data_wait_seconds": 9.638071060180664e-05}
{"step": 16160, "total_steps": 46238, "tokens": 8737800192, "loss": 3.2296871199752344, "ce_loss": 3.2284351135745193, "z_loss": 12.520226348530162, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17213433980941772, "tokens_per_second": 4275185.485695398, "eta_seconds": 3803.762853895185, "mean_data_wait_seconds": 0.00011789798736572266}
{"step": 16170, "total_steps": 46238, "tokens": 8743206912, "loss": 3.263456778092818, "ce_loss": 3.2622427506880327, "z_loss": 12.140019171165697, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20388227701187134, "tokens_per_second": 4274932.769857325, "eta_seconds": 3802.722966457916, "mean_data_wait_seconds": 9.453296661376953e-05}
{"step": 16180, "total_steps": 46238, "tokens": 8748613632, "loss": 3.2390465917009297, "ce_loss": 3.237803197268284, "z_loss": 12.433847405693747, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1674884855747223, "tokens_per_second": 4274645.773225974, "eta_seconds": 3801.7134439038605, "mean_data_wait_seconds": 8.514523506164551e-05}
{"step": 16190, "total_steps": 46238, "tokens": 8754020352, "loss": 3.183993489453287, "ce_loss": 3.18276203401161, "z_loss": 12.314601276860092, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16728809475898743, "tokens_per_second": 4274874.307441332, "eta_seconds": 3800.2454377947697, "mean_data_wait_seconds": 6.35981559753418e-05}
{"step": 16200, "total_steps": 46238, "tokens": 8759427072, "loss": 3.1915708473234465, "ce_loss": 3.1903158065044517, "z_loss": 12.550434676083652, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15587960183620453, "tokens_per_second": 4274657.946911055, "eta_seconds": 3799.1729550513946, "mean_data_wait_seconds": 8.350610733032227e-05}
{"step": 16210, "total_steps": 46238, "tokens": 8764833792, "loss": 3.1879102154211565, "ce_loss": 3.18664586724657, "z_loss": 12.643444993279196, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16574761271476746, "tokens_per_second": 4274749.71496752, "eta_seconds": 3797.826592550192, "mean_data_wait_seconds": 0.00015872716903686523}
{"step": 16220, "total_steps": 46238, "tokens": 8770240512, "loss": 3.1591151266387016, "ce_loss": 3.1578568996805134, "z_loss": 12.58206894903472, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19969290494918823, "tokens_per_second": 4274739.539255432, "eta_seconds": 3796.5708261202753, "mean_data_wait_seconds": 0.00011304020881652832}
{"step": 16230, "total_steps": 46238, "tokens": 8775647232, "loss": 3.222185765251969, "ce_loss": 3.2209275671930024, "z_loss": 12.581991802562367, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15548385679721832, "tokens_per_second": 4274966.668047744, "eta_seconds": 3795.104374792474, "mean_data_wait_seconds": 0.002660810947418213}
{"step": 16240, "total_steps": 46238, "tokens": 8781053952, "loss": 3.256780434738506, "ce_loss": 3.2555828780838936, "z_loss": 11.975669918638287, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.155303955078125, "tokens_per_second": 4275010.557964735, "eta_seconds": 3793.800685189744, "mean_data_wait_seconds": 9.307265281677246e-05}
{"step": 16250, "total_steps": 46238, "tokens": 8786460672, "loss": 3.1984258528911704, "ce_loss": 3.197197063402696, "z_loss": 12.287870385430075, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16094979643821716, "tokens_per_second": 4275191.912643453, "eta_seconds": 3792.3750781926965, "mean_data_wait_seconds": 0.00011542439460754395}
{"step": 16260, "total_steps": 46238, "tokens": 8791867392, "loss": 3.2423157023661062, "ce_loss": 3.2410688436392583, "z_loss": 12.46896928729433, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18178120255470276, "tokens_per_second": 4275256.378233688, "eta_seconds": 3791.053239875215, "mean_data_wait_seconds": 9.60230827331543e-05}
{"step": 16270, "total_steps": 46238, "tokens": 8797274112, "loss": 3.2241465918945544, "ce_loss": 3.222911052631609, "z_loss": 12.355337070696281, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17017832398414612, "tokens_per_second": 4275437.12012899, "eta_seconds": 3789.6283745394376, "mean_data_wait_seconds": 0.0001277625560760498}
{"step": 16280, "total_steps": 46238, "tokens": 8802680832, "loss": 3.1509444695530515, "ce_loss": 3.1496994224461643, "z_loss": 12.450220093582615, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16762183606624603, "tokens_per_second": 4275623.306224263, "eta_seconds": 3788.198806106528, "mean_data_wait_seconds": 9.128451347351074e-05}
{"step": 16290, "total_steps": 46238, "tokens": 8808087552, "loss": 3.197782451456243, "ce_loss": 3.1965500098286252, "z_loss": 12.324655554511331, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17684032022953033, "tokens_per_second": 4275781.772505863, "eta_seconds": 3786.7939117273554, "mean_data_wait_seconds": 0.00010541081428527832}
{"step": 16300, "total_steps": 46238, "tokens": 8813494272, "loss": 3.273438410325484, "ce_loss": 3.272190298094894, "z_loss": 12.48087814359954, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17800697684288025, "tokens_per_second": 4276115.934613242, "eta_seconds": 3785.2335884957643, "mean_data_wait_seconds": 6.74128532409668e-05}
{"step": 16310, "total_steps": 46238, "tokens": 8818900992, "loss": 3.1955664735851865, "ce_loss": 3.1943134846109333, "z_loss": 12.53017369183627, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17502747476100922, "tokens_per_second": 4276300.002123443, "eta_seconds": 3783.8063129259644, "mean_data_wait_seconds": 0.00010612607002258301}
{"step": 16320, "total_steps": 46238, "tokens": 8824307712, "loss": 3.200988021763888, "ce_loss": 3.1997101162419175, "z_loss": 12.77896836309722, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1955590546131134, "tokens_per_second": 4276484.449872957, "eta_seconds": 3782.3788239147057, "mean_data_wait_seconds": 0.00014141201972961426}
{"step": 16330, "total_steps": 46238, "tokens": 8829714432, "loss": 3.228079049876242, "ce_loss": 3.226875962633075, "z_loss": 12.030835339517305, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18301080167293549, "tokens_per_second": 4276797.817378425, "eta_seconds": 3780.837485067683, "mean_data_wait_seconds": 8.687376976013184e-05}
{"step": 16340, "total_steps": 46238, "tokens": 8835121152, "loss": 3.2146137844432485, "ce_loss": 3.2133966503721294, "z_loss": 12.17144293496103, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18941614031791687, "tokens_per_second": 4277027.993271482, "eta_seconds": 3779.3698824112344, "mean_data_wait_seconds": 0.00010734796524047852}
{"step": 16350, "total_steps": 46238, "tokens": 8840527872, "loss": 3.2310262484983965, "ce_loss": 3.22979186520432, "z_loss": 12.344050551905777, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17251846194267273, "tokens_per_second": 4277034.198066796, "eta_seconds": 3778.100271282338, "mean_data_wait_seconds": 0.00010389089584350586}
{"step": 16360, "total_steps": 46238, "tokens": 8845934592, "loss": 3.220197477123954, "ce_loss": 3.2189437837311714, "z_loss": 12.53718496091438, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17137685418128967, "tokens_per_second": 4277241.848042487, "eta_seconds": 3776.6527865130765, "mean_data_wait_seconds": 0.00010582804679870605}
{"step": 16370, "total_steps": 46238, "tokens": 8851341312, "loss": 3.2774379975867993, "ce_loss": 3.276139188896526, "z_loss": 12.988312656229192, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2020643949508667, "tokens_per_second": 4277265.014548725, "eta_seconds": 3775.3682713306757, "mean_data_wait_seconds": 6.714463233947754e-05}
{"step": 16380, "total_steps": 46238, "tokens": 8856748032, "loss": 3.263515241218336, "ce_loss": 3.2622463775403574, "z_loss": 12.68889500878074, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18996085226535797, "tokens_per_second": 4275455.886360273, "eta_seconds": 3775.701196099236, "mean_data_wait_seconds": 8.651614189147949e-05}
{"step": 16390, "total_steps": 46238, "tokens": 8862154752, "loss": 3.247300007126548, "ce_loss": 3.246082947109685, "z_loss": 12.170616749561194, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19000890851020813, "tokens_per_second": 4274576.8674061755, "eta_seconds": 3775.2127718297975, "mean_data_wait_seconds": 9.271502494812012e-05}
{"step": 16400, "total_steps": 46238, "tokens": 8867561472, "loss": 3.2305951262965347, "ce_loss": 3.229365238637635, "z_loss": 12.29888601736589, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19362448155879974, "tokens_per_second": 4274662.945160317, "eta_seconds": 3773.8719218235306, "mean_data_wait_seconds": 0.00010180473327636719}
{"step": 16410, "total_steps": 46238, "tokens": 8872968192, "loss": 3.263947848117713, "ce_loss": 3.262738930456566, "z_loss": 12.088631723866317, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18756304681301117, "tokens_per_second": 4274734.264804756, "eta_seconds": 3772.5441501184323, "mean_data_wait_seconds": 7.793307304382324e-05}
{"step": 16420, "total_steps": 46238, "tokens": 8878374912, "loss": 3.189524717403181, "ce_loss": 3.1882400621067393, "z_loss": 12.846347483721646, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19056741893291473, "tokens_per_second": 4273335.158797671, "eta_seconds": 3772.5140727168714, "mean_data_wait_seconds": 0.0001221001148223877}
{"step": 16430, "total_steps": 46238, "tokens": 8883781632, "loss": 3.295461454174735, "ce_loss": 3.29421455932386, "z_loss": 12.469252976504238, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19233323633670807, "tokens_per_second": 4273524.28028064, "eta_seconds": 3771.081956492753, "mean_data_wait_seconds": 6.583333015441895e-05}
{"step": 16440, "total_steps": 46238, "tokens": 8889188352, "loss": 3.2347747647401057, "ce_loss": 3.233501652876536, "z_loss": 12.731287840640906, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1608322560787201, "tokens_per_second": 4273480.848543432, "eta_seconds": 3769.8551028937104, "mean_data_wait_seconds": 8.174777030944824e-05}
{"step": 16450, "total_steps": 46238, "tokens": 8894595072, "loss": 3.2249174786336496, "ce_loss": 3.2236531936761104, "z_loss": 12.642892945896495, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1944378912448883, "tokens_per_second": 4273636.811985561, "eta_seconds": 3768.452391376119, "mean_data_wait_seconds": 6.93202018737793e-05}
{"step": 16460, "total_steps": 46238, "tokens": 8900001792, "loss": 3.213578912344846, "ce_loss": 3.2123566631114846, "z_loss": 12.222494725025062, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21804410219192505, "tokens_per_second": 4273458.793736364, "eta_seconds": 3767.3441867737847, "mean_data_wait_seconds": 7.772445678710938e-05}
{"step": 16470, "total_steps": 46238, "tokens": 8905408512, "loss": 3.3177769671786916, "ce_loss": 3.3165463407834372, "z_loss": 12.30654046752236, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17769530415534973, "tokens_per_second": 4273673.748959473, "eta_seconds": 3765.889576367056, "mean_data_wait_seconds": 6.729364395141602e-05}
{"step": 16480, "total_steps": 46238, "tokens": 8910815232, "loss": 3.2250293240402685, "ce_loss": 3.223755236827966, "z_loss": 12.740939689405037, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16853182017803192, "tokens_per_second": 4273672.377506196, "eta_seconds": 3764.625662154346, "mean_data_wait_seconds": 7.089972496032715e-05}
{"step": 16490, "total_steps": 46238, "tokens": 8916221952, "loss": 3.284267707304521, "ce_loss": 3.283013752012542, "z_loss": 12.539397983840018, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19039614498615265, "tokens_per_second": 4273739.223836438, "eta_seconds": 3763.3016760349565, "mean_data_wait_seconds": 0.0001322329044342041}
{"step": 16500, "total_steps": 46238, "tokens": 8921628672, "loss": 3.2532233704220164, "ce_loss": 3.2519906235463694, "z_loss": 12.327531099319458, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16147828102111816, "tokens_per_second": 4273816.87289383, "eta_seconds": 3761.9682223570576, "mean_data_wait_seconds": 6.315112113952637e-05}
{"step": 16510, "total_steps": 46238, "tokens": 8927035392, "loss": 3.262776203227766, "ce_loss": 3.2615159731922727, "z_loss": 12.602318778182521, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17478139698505402, "tokens_per_second": 4273779.987668366, "eta_seconds": 3760.7355994871086, "mean_data_wait_seconds": 8.353590965270996e-05}
{"step": 16520, "total_steps": 46238, "tokens": 8932442112, "loss": 3.2865161263581477, "ce_loss": 3.2852739511114177, "z_loss": 12.421917727499297, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.177329882979393, "tokens_per_second": 4273817.443654916, "eta_seconds": 3759.437560407253, "mean_data_wait_seconds": 0.00010883808135986328}
{"step": 16530, "total_steps": 46238, "tokens": 8937848832, "loss": 3.2409253319104514, "ce_loss": 3.239716652667884, "z_loss": 12.086924263925264, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20351777970790863, "tokens_per_second": 4273861.615503956, "eta_seconds": 3758.1336386124585, "mean_data_wait_seconds": 6.091594696044922e-05}
{"step": 16540, "total_steps": 46238, "tokens": 8943255552, "loss": 3.2748819047754463, "ce_loss": 3.2736461451559355, "z_loss": 12.357490864667026, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16421931982040405, "tokens_per_second": 4273920.623258197, "eta_seconds": 3756.8167028239172, "mean_data_wait_seconds": 7.459521293640137e-05}
{"step": 16550, "total_steps": 46238, "tokens": 8948662272, "loss": 3.3742317647644966, "ce_loss": 3.3729736949458267, "z_loss": 12.580496954195427, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20287585258483887, "tokens_per_second": 4274037.712075592, "eta_seconds": 3755.448768889131, "mean_data_wait_seconds": 0.00015106797218322754}
{"step": 16560, "total_steps": 46238, "tokens": 8954068992, "loss": 3.2260977734218943, "ce_loss": 3.2248268127441406, "z_loss": 12.709637345689716, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19985200464725494, "tokens_per_second": 4274184.577154955, "eta_seconds": 3754.0547569614914, "mean_data_wait_seconds": 6.279349327087402e-05}
{"step": 16570, "total_steps": 46238, "tokens": 8959475712, "loss": 3.3139147379181604, "ce_loss": 3.3126986676996406, "z_loss": 12.160860097769534, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18076056241989136, "tokens_per_second": 4274360.705457687, "eta_seconds": 3752.6351492795848, "mean_data_wait_seconds": 6.434321403503418e-05}
{"step": 16580, "total_steps": 46238, "tokens": 8964882432, "loss": 3.287792787407384, "ce_loss": 3.2865745840650615, "z_loss": 12.182131630001646, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18680357933044434, "tokens_per_second": 4274592.493199531, "eta_seconds": 3751.1668140319093, "mean_data_wait_seconds": 6.064772605895996e-05}
{"step": 16590, "total_steps": 46238, "tokens": 8970289152, "loss": 3.2786725560824075, "ce_loss": 3.27744948683363, "z_loss": 12.23051052382498, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1705487221479416, "tokens_per_second": 4274696.986710953, "eta_seconds": 3749.810298561841, "mean_data_wait_seconds": 0.00011053681373596191}
{"step": 16600, "total_steps": 46238, "tokens": 8975695872, "loss": 3.3241721427801885, "ce_loss": 3.322933964657061, "z_loss": 12.38184067697236, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16354289650917053, "tokens_per_second": 4274694.87306514, "eta_seconds": 3748.547332574916, "mean_data_wait_seconds": 8.633732795715332e-05}
{"step": 16610, "total_steps": 46238, "tokens": 8981102592, "loss": 3.3084415504426667, "ce_loss": 3.3071603377660117, "z_loss": 12.811824336196437, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20641405880451202, "tokens_per_second": 4274836.12279736, "eta_seconds": 3747.1586942420263, "mean_data_wait_seconds": 7.24494457244873e-05}
{"step": 16620, "total_steps": 46238, "tokens": 8986509312, "loss": 3.3255231832012986, "ce_loss": 3.3242884621475683, "z_loss": 12.347225586573282, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16726699471473694, "tokens_per_second": 4274891.13878095, "eta_seconds": 3745.845707913483, "mean_data_wait_seconds": 9.045004844665527e-05}
{"step": 16630, "total_steps": 46238, "tokens": 8991916032, "loss": 3.3202195275913584, "ce_loss": 3.3190080649925, "z_loss": 12.114960511525473, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19060976803302765, "tokens_per_second": 4275036.964723123, "eta_seconds": 3744.4532143447213, "mean_data_wait_seconds": 6.711483001708984e-05}
{"step": 16640, "total_steps": 46238, "tokens": 8997322752, "loss": 3.213472147782644, "ce_loss": 3.212238266612544, "z_loss": 12.338468226519497, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20357473194599152, "tokens_per_second": 4275255.90532213, "eta_seconds": 3742.996802619297, "mean_data_wait_seconds": 7.197260856628418e-05}
{"step": 16650, "total_steps": 46238, "tokens": 9002729472, "loss": 3.242260315201499, "ce_loss": 3.241028261907173, "z_loss": 12.320333683129514, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15460903942584991, "tokens_per_second": 4275501.809471231, "eta_seconds": 3741.516943944037, "mean_data_wait_seconds": 9.429454803466797e-05}
{"step": 16660, "total_steps": 46238, "tokens": 9008136192, "loss": 3.252689581928831, "ce_loss": 3.2514104481899375, "z_loss": 12.791254130276767, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20133785903453827, "tokens_per_second": 4275423.010405547, "eta_seconds": 3740.3212980516573, "mean_data_wait_seconds": 7.385015487670898e-05}
{"step": 16670, "total_steps": 46238, "tokens": 9013542912, "loss": 3.1967054930600254, "ce_loss": 3.195472894292889, "z_loss": 12.325367140047478, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18810710310935974, "tokens_per_second": 4275514.063728372, "eta_seconds": 3738.977064680663, "mean_data_wait_seconds": 0.0001507103443145752}
{"step": 16680, "total_steps": 46238, "tokens": 9018949632, "loss": 3.2645559347037114, "ce_loss": 3.2633391347798435, "z_loss": 12.168299732786236, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19250212609767914, "tokens_per_second": 4275627.340817181, "eta_seconds": 3737.613461173558, "mean_data_wait_seconds": 0.0001169443130493164}
{"step": 16690, "total_steps": 46238, "tokens": 9024356352, "loss": 3.2383744608272207, "ce_loss": 3.2371100133115593, "z_loss": 12.644571535515063, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17741526663303375, "tokens_per_second": 4275943.880165393, "eta_seconds": 3736.072322675591, "mean_data_wait_seconds": 7.036328315734863e-05}
{"step": 16700, "total_steps": 46238, "tokens": 9029763072, "loss": 3.1759996034882287, "ce_loss": 3.1747070803786768, "z_loss": 12.92542365103057, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17819707095623016, "tokens_per_second": 4276289.742221078, "eta_seconds": 3734.505804488676, "mean_data_wait_seconds": 6.464123725891113e-05}
{"step": 16710, "total_steps": 46238, "tokens": 9035169792, "loss": 3.22053149432847, "ce_loss": 3.219317517497323, "z_loss": 12.140229167360248, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20312748849391937, "tokens_per_second": 4276440.185552741, "eta_seconds": 3733.110122277217, "mean_data_wait_seconds": 5.519390106201172e-05}
{"step": 16720, "total_steps": 46238, "tokens": 9040576512, "loss": 3.180934859044624, "ce_loss": 3.179668892513622, "z_loss": 12.659718657984879, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14462196826934814, "tokens_per_second": 4276581.515299605, "eta_seconds": 3731.7224907104237, "mean_data_wait_seconds": 9.423494338989258e-05}
{"step": 16730, "total_steps": 46238, "tokens": 9045983232, "loss": 3.1862840056419373, "ce_loss": 3.185020795374206, "z_loss": 12.631679115873395, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16822834312915802, "tokens_per_second": 4276857.862933779, "eta_seconds": 3730.2171863753188, "mean_data_wait_seconds": 6.654858589172363e-05}
{"step": 16740, "total_steps": 46238, "tokens": 9051389952, "loss": 3.2182574904326238, "ce_loss": 3.217033140587084, "z_loss": 12.24341953161991, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18835574388504028, "tokens_per_second": 4277065.718967649, "eta_seconds": 3728.771786992649, "mean_data_wait_seconds": 5.841255187988281e-05}
{"step": 16750, "total_steps": 46238, "tokens": 9056796672, "loss": 3.2262544216531697, "ce_loss": 3.225053122549346, "z_loss": 12.012670719262326, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19895653426647186, "tokens_per_second": 4277364.484838557, "eta_seconds": 3727.2473085963206, "mean_data_wait_seconds": 5.53131103515625e-05}
{"step": 16760, "total_steps": 46238, "tokens": 9062203392, "loss": 3.2244681806275337, "ce_loss": 3.223246126463919, "z_loss": 12.220634359301943, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18624714016914368, "tokens_per_second": 4277525.192140064, "eta_seconds": 3725.843292117809, "mean_data_wait_seconds": 0.00013828277587890625}
{"step": 16770, "total_steps": 46238, "tokens": 9067610112, "loss": 3.1970680663079927, "ce_loss": 3.1958090247529927, "z_loss": 12.590357867154209, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1682700514793396, "tokens_per_second": 4277820.027005301, "eta_seconds": 3724.3226043694094, "mean_data_wait_seconds": 9.956955909729004e-05}
{"step": 16780, "total_steps": 46238, "tokens": 9073016832, "loss": 3.2046980315988716, "ce_loss": 3.203442192438877, "z_loss": 12.558075232939286, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20361876487731934, "tokens_per_second": 4278074.637522932, "eta_seconds": 3722.837129653661, "mean_data_wait_seconds": 9.807944297790527e-05}
{"step": 16790, "total_steps": 46238, "tokens": 9078423552, "loss": 3.1996781013228675, "ce_loss": 3.1984508254311304, "z_loss": 12.272492112535419, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.13526266813278198, "tokens_per_second": 4278076.876340163, "eta_seconds": 3721.571361200116, "mean_data_wait_seconds": 0.00011864304542541504}
{"step": 16800, "total_steps": 46238, "tokens": 9083830272, "loss": 3.2089899868676155, "ce_loss": 3.2077316443125405, "z_loss": 12.583661657391172, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17565634846687317, "tokens_per_second": 4278291.230726283, "eta_seconds": 3720.1211431551233, "mean_data_wait_seconds": 0.00010818243026733398}
{"step": 16810, "total_steps": 46238, "tokens": 9089236992, "loss": 3.1178925001260005, "ce_loss": 3.116633035919883, "z_loss": 12.594085801731456, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18962371349334717, "tokens_per_second": 4278216.247032793, "eta_seconds": 3718.9225661593923, "mean_data_wait_seconds": 0.00012579560279846191}
{"step": 16820, "total_steps": 46238, "tokens": 9094643712, "loss": 3.3043344978130227, "ce_loss": 3.303071865529725, "z_loss": 12.626303911209106, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2102130502462387, "tokens_per_second": 4278363.837818529, "eta_seconds": 3717.530538989803, "mean_data_wait_seconds": 7.906556129455566e-05}
{"step": 16830, "total_steps": 46238, "tokens": 9100050432, "loss": 3.1667636903849514, "ce_loss": 3.1655130639220728, "z_loss": 12.50624418258667, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.202444389462471, "tokens_per_second": 4278495.296515178, "eta_seconds": 3716.152619578694, "mean_data_wait_seconds": 0.00010690093040466309}
{"step": 16840, "total_steps": 46238, "tokens": 9105457152, "loss": 3.196663571126533, "ce_loss": 3.1954113497878565, "z_loss": 12.522165724725435, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17483535408973694, "tokens_per_second": 4278698.6101234555, "eta_seconds": 3714.7124002598066, "mean_data_wait_seconds": 0.00010904669761657715}
{"step": 16850, "total_steps": 46238, "tokens": 9110863872, "loss": 3.2003580349864382, "ce_loss": 3.1990678563262476, "z_loss": 12.901332819100583, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16555652022361755, "tokens_per_second": 4278820.85110275, "eta_seconds": 3713.3426747476733, "mean_data_wait_seconds": 7.715821266174316e-05}
{"step": 16860, "total_steps": 46238, "tokens": 9116270592, "loss": 3.210874803138502, "ce_loss": 3.209612882498539, "z_loss": 12.619589249293009, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19795231521129608, "tokens_per_second": 4279001.100160604, "eta_seconds": 3711.922706307285, "mean_data_wait_seconds": 8.58008861541748e-05}
{"step": 16870, "total_steps": 46238, "tokens": 9121677312, "loss": 3.1886307326230137, "ce_loss": 3.187398632367452, "z_loss": 12.3211290547342, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14821971952915192, "tokens_per_second": 4278959.623509594, "eta_seconds": 3710.695127096564, "mean_data_wait_seconds": 0.00011199712753295898}
{"step": 16880, "total_steps": 46238, "tokens": 9127084032, "loss": 3.2432378602750376, "ce_loss": 3.2419862656882317, "z_loss": 12.51620449441852, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18306100368499756, "tokens_per_second": 4278864.464920742, "eta_seconds": 3709.514062463768, "mean_data_wait_seconds": 9.050965309143066e-05}
{"step": 16890, "total_steps": 46238, "tokens": 9132490752, "loss": 3.1970332662264505, "ce_loss": 3.1958242401932226, "z_loss": 12.090356313821042, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16965758800506592, "tokens_per_second": 4278826.883009533, "eta_seconds": 3708.2830452910957, "mean_data_wait_seconds": 8.806586265563965e-05}
{"step": 16900, "total_steps": 46238, "tokens": 9137897472, "loss": 3.2064387437069053, "ce_loss": 3.205181213942441, "z_loss": 12.575232852589, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1816762387752533, "tokens_per_second": 4279033.30143429, "eta_seconds": 3706.8406218487053, "mean_data_wait_seconds": 8.866190910339355e-05}
{"step": 16910, "total_steps": 46238, "tokens": 9143304192, "loss": 3.2565030148535064, "ce_loss": 3.2552466302207024, "z_loss": 12.563691890600955, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1886463165283203, "tokens_per_second": 4279111.498892695, "eta_seconds": 3705.5093675645353, "mean_data_wait_seconds": 6.979703903198242e-05}
{"step": 16920, "total_steps": 46238, "tokens": 9148710912, "loss": 3.23816560976433, "ce_loss": 3.2368969917297363, "z_loss": 12.685663880723896, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18141967058181763, "tokens_per_second": 4279209.5902277045, "eta_seconds": 3704.16094135659, "mean_data_wait_seconds": 0.00037980079650878906}
{"step": 16930, "total_steps": 46238, "tokens": 9154117632, "loss": 3.1598001248908765, "ce_loss": 3.158586409958926, "z_loss": 12.137004483829845, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17738474905490875, "tokens_per_second": 4279066.664187288, "eta_seconds": 3703.0211369725157, "mean_data_wait_seconds": 7.405877113342285e-05}
{"step": 16940, "total_steps": 46238, "tokens": 9159524352, "loss": 3.2016578154130415, "ce_loss": 3.200390940362757, "z_loss": 12.669014851252237, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20789994299411774, "tokens_per_second": 4278990.744054227, "eta_seconds": 3701.8232876549687, "mean_data_wait_seconds": 7.852911949157715e-05}
{"step": 16950, "total_steps": 46238, "tokens": 9164931072, "loss": 3.2847109653732995, "ce_loss": 3.2834882772330083, "z_loss": 12.2269738370722, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1785736232995987, "tokens_per_second": 4278980.633374263, "eta_seconds": 3700.5684813098374, "mean_data_wait_seconds": 0.0001563429832458496}
{"step": 16960, "total_steps": 46238, "tokens": 9170337792, "loss": 3.3036456234527356, "ce_loss": 3.30239007509116, "z_loss": 12.555322921637332, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19285959005355835, "tokens_per_second": 4278892.810071211, "eta_seconds": 3699.380855426608, "mean_data_wait_seconds": 0.00010818243026733398}
{"step": 16970, "total_steps": 46238, "tokens": 9175744512, "loss": 3.2630657842665007, "ce_loss": 3.2618483937147893, "z_loss": 12.17388317079255, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18764284253120422, "tokens_per_second": 4279052.335972782, "eta_seconds": 3697.97940795755, "mean_data_wait_seconds": 0.00017312169075012207}
{"step": 16980, "total_steps": 46238, "tokens": 9181151232, "loss": 3.3468492789701982, "ce_loss": 3.3455896178881326, "z_loss": 12.596787922310106, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17082342505455017, "tokens_per_second": 4279254.584070187, "eta_seconds": 3696.5411599686563, "mean_data_wait_seconds": 0.00018960237503051758}
{"step": 16990, "total_steps": 46238, "tokens": 9186557952, "loss": 3.2821397456255825, "ce_loss": 3.280908824819507, "z_loss": 12.30973668531938, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.161579892039299, "tokens_per_second": 4279430.802847723, "eta_seconds": 3695.1255231133327, "mean_data_wait_seconds": 9.79006290435791e-05}
{"step": 17000, "total_steps": 46238, "tokens": 9191964672, "loss": 3.2559713360035056, "ce_loss": 3.2547497044910085, "z_loss": 12.216158317797111, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14960472285747528, "tokens_per_second": 4279608.497186445, "eta_seconds": 3693.7087292897127, "mean_data_wait_seconds": 7.617473602294922e-05}
{"step": 17010, "total_steps": 46238, "tokens": 9197371392, "loss": 3.263198379314307, "ce_loss": 3.2619508360371445, "z_loss": 12.475456360614661, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16815334558486938, "tokens_per_second": 4279864.88289022, "eta_seconds": 3692.2241651069744, "mean_data_wait_seconds": 6.937980651855469e-05}
{"step": 17020, "total_steps": 46238, "tokens": 9202778112, "loss": 3.3132159944736594, "ce_loss": 3.3119633017164287, "z_loss": 12.526831157279737, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20396070182323456, "tokens_per_second": 4280233.076775198, "eta_seconds": 3690.6433693329604, "mean_data_wait_seconds": 6.827712059020996e-05}
{"step": 17030, "total_steps": 46238, "tokens": 9208184832, "loss": 3.2834811481562527, "ce_loss": 3.282203936215603, "z_loss": 12.772128220760461, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17639441788196564, "tokens_per_second": 4280658.24330004, "eta_seconds": 3689.013746592885, "mean_data_wait_seconds": 7.465481758117676e-05}
{"step": 17040, "total_steps": 46238, "tokens": 9213591552, "loss": 3.232301414012909, "ce_loss": 3.231062919804544, "z_loss": 12.384849707285563, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15754826366901398, "tokens_per_second": 4280804.932947824, "eta_seconds": 3687.624320954408, "mean_data_wait_seconds": 5.793571472167969e-05}
{"step": 17050, "total_steps": 46238, "tokens": 9218998272, "loss": 3.231976288737673, "ce_loss": 3.230746222264839, "z_loss": 12.301093686710704, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1759214848279953, "tokens_per_second": 4281111.886554593, "eta_seconds": 3686.096995867142, "mean_data_wait_seconds": 5.710124969482422e-05}
{"step": 17060, "total_steps": 46238, "tokens": 9224404992, "loss": 3.2399327231175974, "ce_loss": 3.238650031162031, "z_loss": 12.826318343480429, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21065905690193176, "tokens_per_second": 4281171.638749232, "eta_seconds": 3684.782642493822, "mean_data_wait_seconds": 0.0002535879611968994}
{"step": 17070, "total_steps": 46238, "tokens": 9229811712, "loss": 3.302032505020951, "ce_loss": 3.300762861063986, "z_loss": 12.696425452376857, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1864309012889862, "tokens_per_second": 4281228.82983139, "eta_seconds": 3683.4705293295597, "mean_data_wait_seconds": 7.832050323486328e-05}
{"step": 17080, "total_steps": 46238, "tokens": 9235218432, "loss": 3.2687370578447976, "ce_loss": 3.2674948374430337, "z_loss": 12.422385894890988, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1679706871509552, "tokens_per_second": 4281261.447828704, "eta_seconds": 3682.1795856441104, "mean_data_wait_seconds": 0.00011008977890014648}
{"step": 17090, "total_steps": 46238, "tokens": 9240625152, "loss": 3.2482446414051633, "ce_loss": 3.2469691616116148, "z_loss": 12.755173141306097, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17457526922225952, "tokens_per_second": 4281308.468852232, "eta_seconds": 3680.876278514169, "mean_data_wait_seconds": 8.881092071533203e-05}
{"step": 17100, "total_steps": 46238, "tokens": 9246031872, "loss": 3.258249338829156, "ce_loss": 3.256970105749188, "z_loss": 12.79216725176031, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19236765801906586, "tokens_per_second": 4281500.6258666115, "eta_seconds": 3679.448268400369, "mean_data_wait_seconds": 0.00021955370903015137}
{"step": 17110, "total_steps": 46238, "tokens": 9251438592, "loss": 3.262081317829363, "ce_loss": 3.2608459573803525, "z_loss": 12.353576212218314, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2060783952474594, "tokens_per_second": 4281311.439592411, "eta_seconds": 3678.347993646371, "mean_data_wait_seconds": 0.00011739134788513184}
{"step": 17120, "total_steps": 46238, "tokens": 9256845312, "loss": 3.2793823426420037, "ce_loss": 3.2780971400665515, "z_loss": 12.851878932028105, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20336869359016418, "tokens_per_second": 4281331.323592406, "eta_seconds": 3677.0680505965793, "mean_data_wait_seconds": 8.606910705566406e-05}
{"step": 17130, "total_steps": 46238, "tokens": 9262252032, "loss": 3.270768797758854, "ce_loss": 3.269530410116369, "z_loss": 12.383997758229574, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16901585459709167, "tokens_per_second": 4281591.473982643, "eta_seconds": 3675.581848391871, "mean_data_wait_seconds": 6.809830665588379e-05}
{"step": 17140, "total_steps": 46238, "tokens": 9267658752, "loss": 3.2515956922010942, "ce_loss": 3.250357631481055, "z_loss": 12.3807570067319, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19337546825408936, "tokens_per_second": 4281805.314183489, "eta_seconds": 3674.1355642415447, "mean_data_wait_seconds": 9.986758232116699e-05}
{"step": 17150, "total_steps": 46238, "tokens": 9273065472, "loss": 3.2910960406968086, "ce_loss": 3.2898714072776563, "z_loss": 12.246147647048488, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16333436965942383, "tokens_per_second": 4282087.237126957, "eta_seconds": 3672.63103087821, "mean_data_wait_seconds": 7.787346839904785e-05}
{"step": 17160, "total_steps": 46238, "tokens": 9278472192, "loss": 3.3053717884150418, "ce_loss": 3.3040793682589675, "z_loss": 12.924322576233834, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19621621072292328, "tokens_per_second": 4282349.739125091, "eta_seconds": 3671.143344590981, "mean_data_wait_seconds": 8.64267349243164e-05}
{"step": 17170, "total_steps": 46238, "tokens": 9283878912, "loss": 3.235052769834345, "ce_loss": 3.2338297547716084, "z_loss": 12.230355826291172, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20217351615428925, "tokens_per_second": 4282328.920897472, "eta_seconds": 3669.8986262611443, "mean_data_wait_seconds": 0.00011682510375976562}
{"step": 17180, "total_steps": 46238, "tokens": 9289285632, "loss": 3.272862419937596, "ce_loss": 3.271607854149558, "z_loss": 12.545423623287316, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18016250431537628, "tokens_per_second": 4282588.329645323, "eta_seconds": 3668.4138410523115, "mean_data_wait_seconds": 8.171796798706055e-05}
{"step": 17190, "total_steps": 46238, "tokens": 9294692352, "loss": 3.2309132398981038, "ce_loss": 3.2296534325137283, "z_loss": 12.598237998557813, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16601809859275818, "tokens_per_second": 4282776.815711775, "eta_seconds": 3666.9899599682803, "mean_data_wait_seconds": 0.00011527538299560547}
{"step": 17200, "total_steps": 46238, "tokens": 9300099072, "loss": 3.2906304941032873, "ce_loss": 3.2893748608502476, "z_loss": 12.556421944589326, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20246970653533936, "tokens_per_second": 4283145.354670078, "eta_seconds": 3665.412112825506, "mean_data_wait_seconds": 0.00014969706535339355}
{"step": 17210, "total_steps": 46238, "tokens": 9305505792, "loss": 3.317868521719268, "ce_loss": 3.3166100780169168, "z_loss": 12.584537484429099, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17853198945522308, "tokens_per_second": 4283143.417849227, "eta_seconds": 3664.151445080669, "mean_data_wait_seconds": 9.542703628540039e-05}
{"step": 17220, "total_steps": 46238, "tokens": 9310912512, "loss": 3.283008136532523, "ce_loss": 3.2817894491282376, "z_loss": 12.187416575171731, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17342881858348846, "tokens_per_second": 4283232.440851305, "eta_seconds": 3662.812990107497, "mean_data_wait_seconds": 7.426738739013672e-05}
{"step": 17230, "total_steps": 46238, "tokens": 9316319232, "loss": 3.2316717993129385, "ce_loss": 3.230379977009513, "z_loss": 12.918082533460675, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.23227737843990326, "tokens_per_second": 4283272.955046265, "eta_seconds": 3661.5160576033377, "mean_data_wait_seconds": 8.222460746765137e-05}
{"step": 17240, "total_steps": 46238, "tokens": 9321725952, "loss": 3.156365035158215, "ce_loss": 3.1550639687162456, "z_loss": 13.010745864925962, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17853021621704102, "tokens_per_second": 4283351.983902937, "eta_seconds": 3660.1862380019784, "mean_data_wait_seconds": 7.680058479309082e-05}
{"step": 17250, "total_steps": 46238, "tokens": 9327132672, "loss": 3.24185894836079, "ce_loss": 3.2406058130842266, "z_loss": 12.531132156198675, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.193584144115448, "tokens_per_second": 4283660.192413632, "eta_seconds": 3658.6607153751243, "mean_data_wait_seconds": 6.705522537231445e-05}
{"step": 17260, "total_steps": 46238, "tokens": 9332539392, "loss": 3.1795391982251946, "ce_loss": 3.1782929572192105, "z_loss": 12.462301124225963, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18841049075126648, "tokens_per_second": 4283799.894283559, "eta_seconds": 3657.279268554682, "mean_data_wait_seconds": 0.00014594197273254395}
{"step": 17270, "total_steps": 46238, "tokens": 9337946112, "loss": 3.220428096525597, "ce_loss": 3.2191787777525005, "z_loss": 12.493159583120635, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19380320608615875, "tokens_per_second": 4284122.016344158, "eta_seconds": 3655.7422417592147, "mean_data_wait_seconds": 8.171796798706055e-05}
{"step": 17280, "total_steps": 46238, "tokens": 9343352832, "loss": 3.2390890952312583, "ce_loss": 3.2378350098927817, "z_loss": 12.540901350252556, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18857453763484955, "tokens_per_second": 4284133.20798395, "eta_seconds": 3654.4706581071964, "mean_data_wait_seconds": 5.9545040130615234e-05}
{"step": 17290, "total_steps": 46238, "tokens": 9348759552, "loss": 3.18021404020714, "ce_loss": 3.178941062002471, "z_loss": 12.729479016679706, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17892558872699738, "tokens_per_second": 4284198.182194701, "eta_seconds": 3653.153219905999, "mean_data_wait_seconds": 7.346272468566895e-05}
{"step": 17300, "total_steps": 46238, "tokens": 9354166272, "loss": 3.1603818012006357, "ce_loss": 3.1591360767682395, "z_loss": 12.457333788727269, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18745383620262146, "tokens_per_second": 4284513.111829721, "eta_seconds": 3651.6227761801733, "mean_data_wait_seconds": 6.285309791564941e-05}
{"step": 17310, "total_steps": 46238, "tokens": 9359572992, "loss": 3.1695997353756065, "ce_loss": 3.1683538194858665, "z_loss": 12.459178541645858, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19143171608448029, "tokens_per_second": 4284903.505120816, "eta_seconds": 3650.028273754328, "mean_data_wait_seconds": 0.00010824203491210938}
{"step": 17320, "total_steps": 46238, "tokens": 9364979712, "loss": 3.20156782504284, "ce_loss": 3.200350786700393, "z_loss": 12.170366792967826, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.23001645505428314, "tokens_per_second": 4285081.585987793, "eta_seconds": 3648.6148303745595, "mean_data_wait_seconds": 0.0001055002212524414}
{"step": 17330, "total_steps": 46238, "tokens": 9370386432, "loss": 3.2087065252390774, "ce_loss": 3.207461427558552, "z_loss": 12.451038013805043, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19119027256965637, "tokens_per_second": 4285141.171818233, "eta_seconds": 3647.3023588551587, "mean_data_wait_seconds": 0.00017458200454711914}
{"step": 17340, "total_steps": 46238, "tokens": 9375793152, "loss": 3.1874237512097214, "ce_loss": 3.1861472310441914, "z_loss": 12.765194177627563, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19909749925136566, "tokens_per_second": 4283540.783026399, "eta_seconds": 3647.402832233922, "mean_data_wait_seconds": 0.0019762516021728516}
{"step": 17350, "total_steps": 46238, "tokens": 9381199872, "loss": 3.1905787135615493, "ce_loss": 3.189308625279051, "z_loss": 12.700935645536942, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18908832967281342, "tokens_per_second": 4283687.589849697, "eta_seconds": 3646.0156667372667, "mean_data_wait_seconds": 7.605552673339844e-05}
{"step": 17360, "total_steps": 46238, "tokens": 9386606592, "loss": 3.2093266898935493, "ce_loss": 3.2080664869510764, "z_loss": 12.60198898748918, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20595136284828186, "tokens_per_second": 4283841.740260589, "eta_seconds": 3644.62234850213, "mean_data_wait_seconds": 6.139278411865234e-05}
{"step": 17370, "total_steps": 46238, "tokens": 9392013312, "loss": 3.1870635639537466, "ce_loss": 3.1858172741803257, "z_loss": 12.463275837175773, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20129556953907013, "tokens_per_second": 4283894.24416939, "eta_seconds": 3643.3155755987095, "mean_data_wait_seconds": 9.301304817199707e-05}
{"step": 17380, "total_steps": 46238, "tokens": 9397420032, "loss": 3.181595321857568, "ce_loss": 3.180339556751829, "z_loss": 12.557172255082564, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1735721379518509, "tokens_per_second": 4284262.853431427, "eta_seconds": 3641.7401167399507, "mean_data_wait_seconds": 7.137656211853027e-05}
{"step": 17390, "total_steps": 46238, "tokens": 9402826752, "loss": 3.1914327505863076, "ce_loss": 3.190201304175637, "z_loss": 12.314594268798828, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16708096861839294, "tokens_per_second": 4283241.680460214, "eta_seconds": 3641.346052255497, "mean_data_wait_seconds": 0.00011205673217773438}
{"step": 17400, "total_steps": 46238, "tokens": 9408233472, "loss": 3.1252786697763386, "ce_loss": 3.1240418896530615, "z_loss": 12.367612130714186, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15853068232536316, "tokens_per_second": 4283266.542279217, "eta_seconds": 3640.062627459907, "mean_data_wait_seconds": 0.00010639429092407227}
{"step": 17410, "total_steps": 46238, "tokens": 9413640192, "loss": 3.1514228943622475, "ce_loss": 3.150174476883628, "z_loss": 12.484037269245494, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19994157552719116, "tokens_per_second": 4283391.293632534, "eta_seconds": 3638.6943605103884, "mean_data_wait_seconds": 7.730722427368164e-05}
{"step": 17420, "total_steps": 46238, "tokens": 9419046912, "loss": 3.256967974431587, "ce_loss": 3.2557148445736277, "z_loss": 12.531044035246879, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1870957762002945, "tokens_per_second": 4283428.316563315, "eta_seconds": 3637.400668934411, "mean_data_wait_seconds": 0.00015184283256530762}
{"step": 17430, "total_steps": 46238, "tokens": 9424453632, "loss": 3.243506548982678, "ce_loss": 3.2422253796548555, "z_loss": 12.811783783363573, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17086799442768097, "tokens_per_second": 4283619.558027326, "eta_seconds": 3635.976092884541, "mean_data_wait_seconds": 9.748339653015137e-05}
{"step": 17440, "total_steps": 46238, "tokens": 9429860352, "loss": 3.241689526673519, "ce_loss": 3.2404299161650916, "z_loss": 12.596198349288017, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1885022521018982, "tokens_per_second": 4282802.6627646955, "eta_seconds": 3635.4071877664346, "mean_data_wait_seconds": 0.00011959671974182129}
{"step": 17450, "total_steps": 46238, "tokens": 9435267072, "loss": 3.255079247734763, "ce_loss": 3.253797077771389, "z_loss": 12.821562600858284, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2154439240694046, "tokens_per_second": 4282955.591316675, "eta_seconds": 3634.015000191768, "mean_data_wait_seconds": 8.818507194519043e-05}
{"step": 17460, "total_steps": 46238, "tokens": 9440673792, "loss": 3.2134252533768164, "ce_loss": 3.2121553059780235, "z_loss": 12.699268637281476, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1741618812084198, "tokens_per_second": 4282933.929394134, "eta_seconds": 3632.7709930844, "mean_data_wait_seconds": 8.729100227355957e-05}
{"step": 17470, "total_steps": 46238, "tokens": 9446080512, "loss": 3.246118075919874, "ce_loss": 3.2448287064378913, "z_loss": 12.893818631316677, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19773314893245697, "tokens_per_second": 4282913.943029667, "eta_seconds": 3631.525552670266, "mean_data_wait_seconds": 9.712576866149902e-05}
{"step": 17480, "total_steps": 46238, "tokens": 9451487232, "loss": 3.2428460012782705, "ce_loss": 3.2415713649807554, "z_loss": 12.746402646556046, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1681985855102539, "tokens_per_second": 4283005.783169881, "eta_seconds": 3630.185316372079, "mean_data_wait_seconds": 0.00010067224502563477}
{"step": 17490, "total_steps": 46238, "tokens": 9456893952, "loss": 3.286512470606602, "ce_loss": 3.2852756146228677, "z_loss": 12.368318875630697, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19647358357906342, "tokens_per_second": 4283107.418266857, "eta_seconds": 3628.8368388129975, "mean_data_wait_seconds": 8.249282836914062e-05}
{"step": 17500, "total_steps": 46238, "tokens": 9462300672, "loss": 3.262555566701022, "ce_loss": 3.2613413044900605, "z_loss": 12.14215171698368, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19725638628005981, "tokens_per_second": 4283168.171504637, "eta_seconds": 3627.5230487954186, "mean_data_wait_seconds": 0.00010508298873901367}
{"step": 17510, "total_steps": 46238, "tokens": 9467707392, "loss": 3.2958028316497803, "ce_loss": 3.29454879507874, "z_loss": 12.540495858047947, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19077140092849731, "tokens_per_second": 4283119.787313509, "eta_seconds": 3626.3016948545414, "mean_data_wait_seconds": 8.815526962280273e-05}
{"step": 17520, "total_steps": 46238, "tokens": 9473114112, "loss": 3.2588748173280195, "ce_loss": 3.2576278141050627, "z_loss": 12.469790689872973, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20153926312923431, "tokens_per_second": 4283237.639763489, "eta_seconds": 3624.9396204076447, "mean_data_wait_seconds": 7.212162017822266e-05}
{"step": 17530, "total_steps": 46238, "tokens": 9478520832, "loss": 3.296315614021186, "ce_loss": 3.2950434377699187, "z_loss": 12.721788977131698, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1920236051082611, "tokens_per_second": 4283106.778269325, "eta_seconds": 3623.788036933228, "mean_data_wait_seconds": 6.774067878723145e-05}
{"step": 17540, "total_steps": 46238, "tokens": 9483927552, "loss": 3.270203203865976, "ce_loss": 3.2689381675286726, "z_loss": 12.650374029621933, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15090931951999664, "tokens_per_second": 4283435.443468712, "eta_seconds": 3622.247746877554, "mean_data_wait_seconds": 0.00010082125663757324}
{"step": 17550, "total_steps": 46238, "tokens": 9489334272, "loss": 3.246924198035038, "ce_loss": 3.245624775236303, "z_loss": 12.994089076013276, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19981972873210907, "tokens_per_second": 4283681.535211049, "eta_seconds": 3620.7774869603695, "mean_data_wait_seconds": 0.00012996792793273926}
{"step": 17560, "total_steps": 46238, "tokens": 9494740992, "loss": 3.3039435942967734, "ce_loss": 3.3026562640161226, "z_loss": 12.873006878477154, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1746615469455719, "tokens_per_second": 4283821.568088758, "eta_seconds": 3619.3970027835553, "mean_data_wait_seconds": 0.00016057491302490234}
{"step": 17570, "total_steps": 46238, "tokens": 9500147712, "loss": 3.244200704675732, "ce_loss": 3.2429361487879897, "z_loss": 12.645289370507905, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1647080034017563, "tokens_per_second": 4283912.545523424, "eta_seconds": 3618.058039069101, "mean_data_wait_seconds": 9.748339653015137e-05}
{"step": 17580, "total_steps": 46238, "tokens": 9505554432, "loss": 3.265367309252421, "ce_loss": 3.264123842571721, "z_loss": 12.4347546823097, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18579883873462677, "tokens_per_second": 4284016.410440444, "eta_seconds": 3616.708252153274, "mean_data_wait_seconds": 7.444620132446289e-05}
{"step": 17590, "total_steps": 46238, "tokens": 9510961152, "loss": 3.248764126589804, "ce_loss": 3.2474754416581355, "z_loss": 12.886920098102454, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19176630675792694, "tokens_per_second": 4284375.873548258, "eta_seconds": 3615.142844871951, "mean_data_wait_seconds": 0.00013652443885803223}
{"step": 17600, "total_steps": 46238, "tokens": 9516367872, "loss": 3.3098823970014397, "ce_loss": 3.3086360349799646, "z_loss": 12.463885076118238, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15948614478111267, "tokens_per_second": 4284516.532621785, "eta_seconds": 3613.7622404097697, "mean_data_wait_seconds": 7.253885269165039e-05}
{"step": 17610, "total_steps": 46238, "tokens": 9521774592, "loss": 3.3657085570422085, "ce_loss": 3.364482581615448, "z_loss": 12.259797522515962, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2030402570962906, "tokens_per_second": 4284847.979048553, "eta_seconds": 3612.2208815064746, "mean_data_wait_seconds": 7.44163990020752e-05}
{"step": 17620, "total_steps": 46238, "tokens": 9527181312, "loss": 3.30685570565137, "ce_loss": 3.3056330807281262, "z_loss": 12.22642529372013, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19958524405956268, "tokens_per_second": 4284937.508527773, "eta_seconds": 3610.88361106952, "mean_data_wait_seconds": 6.86347484588623e-05}
{"step": 17630, "total_steps": 46238, "tokens": 9532588032, "loss": 3.2557925527746026, "ce_loss": 3.2545896606011824, "z_loss": 12.028419588551376, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20774269104003906, "tokens_per_second": 4285212.569315115, "eta_seconds": 3609.3901186498238, "mean_data_wait_seconds": 7.25090503692627e-05}
{"step": 17640, "total_steps": 46238, "tokens": 9537994752, "loss": 3.273176171562888, "ce_loss": 3.2719219092166787, "z_loss": 12.542714884786895, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16763897240161896, "tokens_per_second": 4285337.948699883, "eta_seconds": 3608.022837193237, "mean_data_wait_seconds": 0.00010067224502563477}
{"step": 17650, "total_steps": 46238, "tokens": 9543401472, "loss": 3.287030442194505, "ce_loss": 3.285794603102135, "z_loss": 12.358450398300633, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18468955159187317, "tokens_per_second": 4285506.380604, "eta_seconds": 3606.619403007773, "mean_data_wait_seconds": 6.267428398132324e-05}
{"step": 17660, "total_steps": 46238, "tokens": 9548808192, "loss": 3.2828217708703242, "ce_loss": 3.2816235170219885, "z_loss": 11.982233213655876, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17470099031925201, "tokens_per_second": 4285779.094469161, "eta_seconds": 3605.1283566946754, "mean_data_wait_seconds": 6.747245788574219e-05}
{"step": 17670, "total_steps": 46238, "tokens": 9554214912, "loss": 3.2314136895266445, "ce_loss": 3.2301567738706414, "z_loss": 12.569047696662672, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17699451744556427, "tokens_per_second": 4285850.43524519, "eta_seconds": 3603.8068190581603, "mean_data_wait_seconds": 6.797909736633301e-05}
{"step": 17680, "total_steps": 46238, "tokens": 9559621632, "loss": 3.303455477411097, "ce_loss": 3.302232964472337, "z_loss": 12.224646705569644, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19349856674671173, "tokens_per_second": 4286029.013452885, "eta_seconds": 3602.395190405243, "mean_data_wait_seconds": 9.489059448242188e-05}
{"step": 17690, "total_steps": 46238, "tokens": 9565028352, "loss": 3.246500410816886, "ce_loss": 3.2452138553966177, "z_loss": 12.865612665812174, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17218448221683502, "tokens_per_second": 4286040.3129805345, "eta_seconds": 3601.124221173441, "mean_data_wait_seconds": 8.26120376586914e-05}
{"step": 17700, "total_steps": 46238, "tokens": 9570435072, "loss": 3.2273562984033064, "ce_loss": 3.226096545204972, "z_loss": 12.597515633611968, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17341256141662598, "tokens_per_second": 4286356.31168296, "eta_seconds": 3599.5973601042097, "mean_data_wait_seconds": 5.97536563873291e-05}
{"step": 17710, "total_steps": 46238, "tokens": 9575841792, "loss": 3.203253630435828, "ce_loss": 3.2020384777675974, "z_loss": 12.151781125502152, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19093866646289825, "tokens_per_second": 4286401.757643912, "eta_seconds": 3598.297830224367, "mean_data_wait_seconds": 0.00010558962821960449}
{"step": 17720, "total_steps": 46238, "tokens": 9581248512, "loss": 3.258360376863769, "ce_loss": 3.2571130124005405, "z_loss": 12.47361995234634, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17172130942344666, "tokens_per_second": 4286564.63173801, "eta_seconds": 3596.8997900653494, "mean_data_wait_seconds": 0.00010693073272705078}
{"step": 17730, "total_steps": 46238, "tokens": 9586655232, "loss": 3.2326490915182866, "ce_loss": 3.231394796660452, "z_loss": 12.543154774290143, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16641095280647278, "tokens_per_second": 4286561.528898177, "eta_seconds": 3595.641075041739, "mean_data_wait_seconds": 9.146332740783691e-05}
{"step": 17740, "total_steps": 46238, "tokens": 9592061952, "loss": 3.277411796829917, "ce_loss": 3.2761303746339046, "z_loss": 12.813975630384503, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15768729150295258, "tokens_per_second": 4286829.47437983, "eta_seconds": 3594.155092028471, "mean_data_wait_seconds": 7.086992263793945e-05}
{"step": 17750, "total_steps": 46238, "tokens": 9597468672, "loss": 3.2279333183259675, "ce_loss": 3.226657804214593, "z_loss": 12.754869070920078, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.152539923787117, "tokens_per_second": 4286988.563341587, "eta_seconds": 3592.7605209178537, "mean_data_wait_seconds": 6.446242332458496e-05}
{"step": 17760, "total_steps": 46238, "tokens": 9602875392, "loss": 3.2447961045034006, "ce_loss": 3.2435123324394226, "z_loss": 12.837512608730432, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20345136523246765, "tokens_per_second": 4287321.511193734, "eta_seconds": 3591.220416710255, "mean_data_wait_seconds": 5.647540092468262e-05}
{"step": 17770, "total_steps": 46238, "tokens": 9608282112, "loss": 3.2455382708347207, "ce_loss": 3.2443068912535002, "z_loss": 12.313685099283854, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16319800913333893, "tokens_per_second": 4287509.318653417, "eta_seconds": 3589.8020692428413, "mean_data_wait_seconds": 9.897351264953613e-05}
{"step": 17780, "total_steps": 46238, "tokens": 9613688832, "loss": 3.2199914491537847, "ce_loss": 3.218760943774021, "z_loss": 12.30524292859164, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18366633355617523, "tokens_per_second": 4287508.160410742, "eta_seconds": 3588.541998839259, "mean_data_wait_seconds": 0.00017049908638000488}
{"step": 17790, "total_steps": 46238, "tokens": 9619095552, "loss": 3.22860968654806, "ce_loss": 3.2273602286974588, "z_loss": 12.494408015048865, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16664616763591766, "tokens_per_second": 4287648.24399428, "eta_seconds": 3587.1637570883995, "mean_data_wait_seconds": 7.599592208862305e-05}
{"step": 17800, "total_steps": 46238, "tokens": 9624502272, "loss": 3.2126399387012827, "ce_loss": 3.2113700689691487, "z_loss": 12.698467688126998, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1557992845773697, "tokens_per_second": 4288039.108916722, "eta_seconds": 3585.575894592103, "mean_data_wait_seconds": 8.100271224975586e-05}
{"step": 17810, "total_steps": 46238, "tokens": 9629908992, "loss": 3.2899975686362297, "ce_loss": 3.2887422767552463, "z_loss": 12.552764285694469, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1719181388616562, "tokens_per_second": 4288173.230224867, "eta_seconds": 3584.2029038537767, "mean_data_wait_seconds": 0.00012952089309692383}
{"step": 17820, "total_steps": 46238, "tokens": 9635315712, "loss": 3.2425936948169363, "ce_loss": 3.241331443642125, "z_loss": 12.622469107309977, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1940731704235077, "tokens_per_second": 4288522.419082786, "eta_seconds": 3582.6503216196447, "mean_data_wait_seconds": 0.00010919570922851562}
{"step": 17830, "total_steps": 46238, "tokens": 9640722432, "loss": 3.184773421648777, "ce_loss": 3.1835060011256826, "z_loss": 12.674314181009928, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19581328332424164, "tokens_per_second": 4288635.524663388, "eta_seconds": 3581.2951265438924, "mean_data_wait_seconds": 7.2479248046875e-05}
{"step": 17840, "total_steps": 46238, "tokens": 9646129152, "loss": 3.2061512434121333, "ce_loss": 3.204928706992756, "z_loss": 12.22527704816876, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21725060045719147, "tokens_per_second": 4288957.334098938, "eta_seconds": 3579.7658004041655, "mean_data_wait_seconds": 5.882978439331055e-05}
{"step": 17850, "total_steps": 46238, "tokens": 9651535872, "loss": 3.240405250679363, "ce_loss": 3.2391806028105994, "z_loss": 12.246650659676755, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17475947737693787, "tokens_per_second": 4289205.881559726, "eta_seconds": 3578.297821977908, "mean_data_wait_seconds": 7.37905502319336e-05}
{"step": 17860, "total_steps": 46238, "tokens": 9656942592, "loss": 3.1217932538552717, "ce_loss": 3.120512853969227, "z_loss": 12.803740913217718, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22488723695278168, "tokens_per_second": 4289184.667127563, "eta_seconds": 3577.0549730783373, "mean_data_wait_seconds": 7.805228233337402e-05}
{"step": 17870, "total_steps": 46238, "tokens": 9662349312, "loss": 3.098234980395346, "ce_loss": 3.0969613483457854, "z_loss": 12.736254713752054, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1925695389509201, "tokens_per_second": 4289190.064328648, "eta_seconds": 3575.7899262971023, "mean_data_wait_seconds": 9.331107139587402e-05}
{"step": 17880, "total_steps": 46238, "tokens": 9667756032, "loss": 3.0931288094231575, "ce_loss": 3.0918399113597292, "z_loss": 12.888771548415676, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19115832448005676, "tokens_per_second": 4289394.895998878, "eta_seconds": 3574.3586859539196, "mean_data_wait_seconds": 0.00012186169624328613}
{"step": 17890, "total_steps": 46238, "tokens": 9673162752, "loss": 3.1111062692873404, "ce_loss": 3.109809935092926, "z_loss": 12.96351000034448, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17078082263469696, "tokens_per_second": 4289532.208314418, "eta_seconds": 3572.9838219404714, "mean_data_wait_seconds": 0.00011858344078063965}
{"step": 17900, "total_steps": 46238, "tokens": 9678569472, "loss": 3.113672642996817, "ce_loss": 3.112378949468786, "z_loss": 12.937193350358443, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1823950707912445, "tokens_per_second": 4289564.507093382, "eta_seconds": 3571.696483096266, "mean_data_wait_seconds": 6.744265556335449e-05}
{"step": 17910, "total_steps": 46238, "tokens": 9683976192, "loss": 3.1432365720922295, "ce_loss": 3.1420143318898752, "z_loss": 12.222614533973463, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21174271404743195, "tokens_per_second": 4289543.542223136, "eta_seconds": 3570.4534977309954, "mean_data_wait_seconds": 7.06017017364502e-05}
{"step": 17920, "total_steps": 46238, "tokens": 9689382912, "loss": 3.1565887061032383, "ce_loss": 3.1553132823019316, "z_loss": 12.7542202761679, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1923462152481079, "tokens_per_second": 4289677.295228179, "eta_seconds": 3569.081767766312, "mean_data_wait_seconds": 9.754300117492676e-05}
{"step": 17930, "total_steps": 46238, "tokens": 9694789632, "loss": 3.212134478670178, "ce_loss": 3.210890421361634, "z_loss": 12.440674789024122, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16766196489334106, "tokens_per_second": 4289738.194126225, "eta_seconds": 3567.770714995214, "mean_data_wait_seconds": 0.00011366605758666992}
{"step": 17940, "total_steps": 46238, "tokens": 9700196352, "loss": 3.1906225446498757, "ce_loss": 3.1893923300685305, "z_loss": 12.301909706809305, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20440803468227386, "tokens_per_second": 4289616.949297606, "eta_seconds": 3566.6111368067877, "mean_data_wait_seconds": 8.112192153930664e-05}
{"step": 17950, "total_steps": 46238, "tokens": 9705603072, "loss": 3.180188460783525, "ce_loss": 3.178922096888224, "z_loss": 12.664005438486734, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18829211592674255, "tokens_per_second": 4289600.48674105, "eta_seconds": 3565.3643996155324, "mean_data_wait_seconds": 6.756186485290527e-05}
{"step": 17960, "total_steps": 46238, "tokens": 9711009792, "loss": 3.205002403620518, "ce_loss": 3.203721624432188, "z_loss": 12.807737617781669, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18221160769462585, "tokens_per_second": 4289797.069871721, "eta_seconds": 3563.9406468374455, "mean_data_wait_seconds": 7.82310962677002e-05}
{"step": 17970, "total_steps": 46238, "tokens": 9716416512, "loss": 3.20001066272909, "ce_loss": 3.198746227856838, "z_loss": 12.644141442847975, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.171881303191185, "tokens_per_second": 4289821.969324115, "eta_seconds": 3562.6596006285895, "mean_data_wait_seconds": 0.0001061558723449707}
{"step": 17980, "total_steps": 46238, "tokens": 9721823232, "loss": 3.2093717004313613, "ce_loss": 3.2081098701014663, "z_loss": 12.618731180826822, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1764831840991974, "tokens_per_second": 4289959.440560516, "eta_seconds": 3561.2851160205473, "mean_data_wait_seconds": 7.361173629760742e-05}
{"step": 17990, "total_steps": 46238, "tokens": 9727229952, "loss": 3.2746639450391135, "ce_loss": 3.27344267837929, "z_loss": 12.212374333179358, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1938706636428833, "tokens_per_second": 4290111.978444252, "eta_seconds": 3559.8982172811034, "mean_data_wait_seconds": 8.133053779602051e-05}
{"step": 18000, "total_steps": 46238, "tokens": 9732636672, "loss": 3.1841674609617754, "ce_loss": 3.182925242366213, "z_loss": 12.422522906101111, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1595906913280487, "tokens_per_second": 4290309.546639082, "eta_seconds": 3558.474067671816, "mean_data_wait_seconds": 9.59932804107666e-05}
{"step": 18010, "total_steps": 46238, "tokens": 9738043392, "loss": 3.190865843585043, "ce_loss": 3.1896077647353662, "z_loss": 12.580480720057633, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17744125425815582, "tokens_per_second": 4290481.204604851, "eta_seconds": 3557.0715302563767, "mean_data_wait_seconds": 0.00013595819473266602}
{"step": 18020, "total_steps": 46238, "tokens": 9743450112, "loss": 3.2660745924169365, "ce_loss": 3.2648003498713174, "z_loss": 12.74229056184942, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18663787841796875, "tokens_per_second": 4289229.663448493, "eta_seconds": 3556.8489031977438, "mean_data_wait_seconds": 0.0001398324966430664}
{"step": 18030, "total_steps": 46238, "tokens": 9748856832, "loss": 3.2127043749346877, "ce_loss": 3.2115118088144246, "z_loss": 11.925914273117527, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2054629772901535, "tokens_per_second": 4288388.635545336, "eta_seconds": 3556.2856821302603, "mean_data_wait_seconds": 9.617209434509277e-05}
{"step": 18040, "total_steps": 46238, "tokens": 9754263552, "loss": 3.230515496297316, "ce_loss": 3.2292759815851846, "z_loss": 12.395156000599716, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19144664704799652, "tokens_per_second": 4288573.2723897565, "eta_seconds": 3554.8718456440693, "mean_data_wait_seconds": 0.00013875961303710938}
{"step": 18050, "total_steps": 46238, "tokens": 9759670272, "loss": 3.1870414379871255, "ce_loss": 3.1858195626374446, "z_loss": 12.218894409410881, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18528953194618225, "tokens_per_second": 4287556.906651293, "eta_seconds": 3554.453502496559, "mean_data_wait_seconds": 0.07107460498809814}
{"step": 18060, "total_steps": 46238, "tokens": 9765076992, "loss": 3.1920550429459773, "ce_loss": 3.1907886989188916, "z_loss": 12.663564898750998, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16255471110343933, "tokens_per_second": 4287117.794992789, "eta_seconds": 3553.556415406502, "mean_data_wait_seconds": 9.274482727050781e-05}
{"step": 18070, "total_steps": 46238, "tokens": 9770483712, "loss": 3.2310598889986673, "ce_loss": 3.229831357796987, "z_loss": 12.285299525116429, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17722028493881226, "tokens_per_second": 4287001.26483439, "eta_seconds": 3552.3918196437276, "mean_data_wait_seconds": 0.00015211105346679688}
{"step": 18080, "total_steps": 46238, "tokens": 9775890432, "loss": 3.304199128439932, "ce_loss": 3.3029037966872705, "z_loss": 12.953222701043794, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18199969828128815, "tokens_per_second": 4286866.853862732, "eta_seconds": 3551.241972976722, "mean_data_wait_seconds": 8.553266525268555e-05}
{"step": 18090, "total_steps": 46238, "tokens": 9781297152, "loss": 3.246167321999868, "ce_loss": 3.244911914521998, "z_loss": 12.554065003539577, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19636324048042297, "tokens_per_second": 4286809.382561796, "eta_seconds": 3550.0283371371997, "mean_data_wait_seconds": 7.18533992767334e-05}
{"step": 18100, "total_steps": 46238, "tokens": 9786703872, "loss": 3.2493856567325015, "ce_loss": 3.248138714920391, "z_loss": 12.469550147201076, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17919516563415527, "tokens_per_second": 4286928.999113408, "eta_seconds": 3548.6680715137154, "mean_data_wait_seconds": 7.385015487670898e-05}
{"step": 18110, "total_steps": 46238, "tokens": 9792110592, "loss": 3.247839320789684, "ce_loss": 3.2465827772111604, "z_loss": 12.565536961410984, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16110216081142426, "tokens_per_second": 4286947.907775681, "eta_seconds": 3547.39121425213, "mean_data_wait_seconds": 7.69197940826416e-05}
{"step": 18120, "total_steps": 46238, "tokens": 9797517312, "loss": 3.226892016150735, "ce_loss": 3.2256061127691558, "z_loss": 12.858612696329752, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.180129274725914, "tokens_per_second": 4286879.236146382, "eta_seconds": 3546.1868148321455, "mean_data_wait_seconds": 8.466839790344238e-05}
{"step": 18130, "total_steps": 46238, "tokens": 9802924032, "loss": 3.2623629678379404, "ce_loss": 3.261139797441887, "z_loss": 12.231669368165912, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16014808416366577, "tokens_per_second": 4286989.427945351, "eta_seconds": 3544.834471701366, "mean_data_wait_seconds": 0.00011247396469116211}
{"step": 18140, "total_steps": 46238, "tokens": 9808330752, "loss": 3.3945063171964702, "ce_loss": 3.393273272297599, "z_loss": 12.330193642413978, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18889489769935608, "tokens_per_second": 4287070.496576869, "eta_seconds": 3543.5062698712063, "mean_data_wait_seconds": 8.39531421661377e-05}
{"step": 18150, "total_steps": 46238, "tokens": 9813737472, "loss": 3.268917699654897, "ce_loss": 3.2676227941657556, "z_loss": 12.949187098127423, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.175770565867424, "tokens_per_second": 4287040.289694291, "eta_seconds": 3542.270059953858, "mean_data_wait_seconds": 8.103251457214355e-05}
{"step": 18160, "total_steps": 46238, "tokens": 9819144192, "loss": 3.2801207502683005, "ce_loss": 3.2788318811040935, "z_loss": 12.888851274143565, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1774555891752243, "tokens_per_second": 4287309.08903982, "eta_seconds": 3540.7868732412276, "mean_data_wait_seconds": 7.387995719909668e-05}
{"step": 18170, "total_steps": 46238, "tokens": 9824550912, "loss": 3.217370912884221, "ce_loss": 3.2160811948053762, "z_loss": 12.897694168668805, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19625188410282135, "tokens_per_second": 4287465.333218338, "eta_seconds": 3539.3967868211366, "mean_data_wait_seconds": 6.809830665588379e-05}
{"step": 18180, "total_steps": 46238, "tokens": 9829957632, "loss": 3.30005432620193, "ce_loss": 3.298785984516144, "z_loss": 12.683402386578647, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1816171109676361, "tokens_per_second": 4287781.451310463, "eta_seconds": 3537.8748838431925, "mean_data_wait_seconds": 6.830692291259766e-05}
{"step": 18190, "total_steps": 46238, "tokens": 9835364352, "loss": 3.2127163699178984, "ce_loss": 3.211463476672317, "z_loss": 12.52886829954205, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1712895631790161, "tokens_per_second": 4288141.248395225, "eta_seconds": 3536.3171839721917, "mean_data_wait_seconds": 6.133317947387695e-05}
{"step": 18200, "total_steps": 46238, "tokens": 9840771072, "loss": 3.2617494590354688, "ce_loss": 3.2605250134612573, "z_loss": 12.24422295888265, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1668134480714798, "tokens_per_second": 4288344.274773601, "eta_seconds": 3534.888967094484, "mean_data_wait_seconds": 6.306171417236328e-05}
{"step": 18210, "total_steps": 46238, "tokens": 9846177792, "loss": 3.281804538134373, "ce_loss": 3.280537500525966, "z_loss": 12.670287031115908, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18652500212192535, "tokens_per_second": 4288779.19767528, "eta_seconds": 3533.2698293756557, "mean_data_wait_seconds": 5.97834587097168e-05}
{"step": 18220, "total_steps": 46238, "tokens": 9851584512, "loss": 3.203870070703102, "ce_loss": 3.2026382449901467, "z_loss": 12.317958268252285, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1574157327413559, "tokens_per_second": 4289045.862844322, "eta_seconds": 3531.7895654196745, "mean_data_wait_seconds": 0.00011008977890014648}
{"step": 18230, "total_steps": 46238, "tokens": 9856991232, "loss": 3.236829501209837, "ce_loss": 3.2355366385344304, "z_loss": 12.929396925550519, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1593070924282074, "tokens_per_second": 4289217.077878568, "eta_seconds": 3530.3880472959127, "mean_data_wait_seconds": 7.370114326477051e-05}
{"step": 18240, "total_steps": 46238, "tokens": 9862397952, "loss": 3.284625049793359, "ce_loss": 3.2833729350205623, "z_loss": 12.521241852731416, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18678854405879974, "tokens_per_second": 4289504.881575879, "eta_seconds": 3528.89072326662, "mean_data_wait_seconds": 6.0558319091796875e-05}
{"step": 18250, "total_steps": 46238, "tokens": 9867804672, "loss": 3.24831155574683, "ce_loss": 3.2470409274101257, "z_loss": 12.706701784422904, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1697390079498291, "tokens_per_second": 4289724.371512247, "eta_seconds": 3527.449773810438, "mean_data_wait_seconds": 6.407499313354492e-05}
{"step": 18260, "total_steps": 46238, "tokens": 9873211392, "loss": 3.2612191727667144, "ce_loss": 3.25994830420523, "z_loss": 12.70857985091932, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18187621235847473, "tokens_per_second": 4290040.757936268, "eta_seconds": 3525.9293320272727, "mean_data_wait_seconds": 5.561113357543945e-05}
{"step": 18270, "total_steps": 46238, "tokens": 9878618112, "loss": 3.243964236794096, "ce_loss": 3.242694945046396, "z_loss": 12.69299791798447, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21054519712924957, "tokens_per_second": 4290288.093081016, "eta_seconds": 3524.4658391089683, "mean_data_wait_seconds": 7.480382919311523e-05}
{"step": 18280, "total_steps": 46238, "tokens": 9884024832, "loss": 3.2529521494200737, "ce_loss": 3.2517075881813513, "z_loss": 12.44577779191913, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1867986023426056, "tokens_per_second": 4290432.197144136, "eta_seconds": 3523.0872810579453, "mean_data_wait_seconds": 0.00012031197547912598}
{"step": 18290, "total_steps": 46238, "tokens": 9889431552, "loss": 3.290299339727922, "ce_loss": 3.2890066883780738, "z_loss": 12.926747329307325, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19257758557796478, "tokens_per_second": 4290717.190164447, "eta_seconds": 3521.5931776246675, "mean_data_wait_seconds": 8.016824722290039e-05}
{"step": 18300, "total_steps": 46238, "tokens": 9894838272, "loss": 3.2248288609764795, "ce_loss": 3.2235390543937683, "z_loss": 12.898513685573231, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16640689969062805, "tokens_per_second": 4290955.55925801, "eta_seconds": 3520.1375207465226, "mean_data_wait_seconds": 6.619095802307129e-05}
{"step": 18310, "total_steps": 46238, "tokens": 9900244992, "loss": 3.2795946670301035, "ce_loss": 3.278354792883902, "z_loss": 12.398565704172308, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21742375195026398, "tokens_per_second": 4291225.445225453, "eta_seconds": 3518.6561826529037, "mean_data_wait_seconds": 6.115436553955078e-05}
{"step": 18320, "total_steps": 46238, "tokens": 9905651712, "loss": 3.2325794299443564, "ce_loss": 3.231316651358749, "z_loss": 12.627455805287216, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1846909075975418, "tokens_per_second": 4291431.580535836, "eta_seconds": 3517.227279693772, "mean_data_wait_seconds": 9.21487808227539e-05}
{"step": 18330, "total_steps": 46238, "tokens": 9911058432, "loss": 3.2645486809990625, "ce_loss": 3.263290018746347, "z_loss": 12.586475018298987, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14822877943515778, "tokens_per_second": 4291715.660591601, "eta_seconds": 3515.7346612100782, "mean_data_wait_seconds": 0.00013107061386108398}
{"step": 18340, "total_steps": 46238, "tokens": 9916465152, "loss": 3.254371975407456, "ce_loss": 3.2531043056285744, "z_loss": 12.676526734323213, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1882745921611786, "tokens_per_second": 4291799.88981233, "eta_seconds": 3514.405883602264, "mean_data_wait_seconds": 7.933378219604492e-05}
{"step": 18350, "total_steps": 46238, "tokens": 9921871872, "loss": 3.2353314847657173, "ce_loss": 3.234097507866946, "z_loss": 12.33924810814135, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19925116002559662, "tokens_per_second": 4291910.121615552, "eta_seconds": 3513.0558741347722, "mean_data_wait_seconds": 0.00010386109352111816}
{"step": 18360, "total_steps": 46238, "tokens": 9927278592, "loss": 3.246106778130387, "ce_loss": 3.2448502851255014, "z_loss": 12.565086870482475, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17220433056354523, "tokens_per_second": 4291810.135387086, "eta_seconds": 3511.877941599716, "mean_data_wait_seconds": 0.00012260675430297852}
{"step": 18370, "total_steps": 46238, "tokens": 9932685312, "loss": 3.248756159435619, "ce_loss": 3.247476111758839, "z_loss": 12.80030437671777, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22173653542995453, "tokens_per_second": 4291847.216246001, "eta_seconds": 3510.587834293585, "mean_data_wait_seconds": 0.00048279762268066406}
{"step": 18380, "total_steps": 46238, "tokens": 9938092032, "loss": 3.2703526814778647, "ce_loss": 3.269082410769029, "z_loss": 12.702956929351345, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16840943694114685, "tokens_per_second": 4291721.678517815, "eta_seconds": 3509.4307208666955, "mean_data_wait_seconds": 7.513165473937988e-05}
{"step": 18390, "total_steps": 46238, "tokens": 9943498752, "loss": 3.2476312987732165, "ce_loss": 3.2463629588936316, "z_loss": 12.683698097864786, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16071605682373047, "tokens_per_second": 4291656.224720904, "eta_seconds": 3508.22442330621, "mean_data_wait_seconds": 9.888410568237305e-05}
{"step": 18400, "total_steps": 46238, "tokens": 9948905472, "loss": 3.219025068210833, "ce_loss": 3.2177108164989585, "z_loss": 13.142623121088201, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19083988666534424, "tokens_per_second": 4291818.065570332, "eta_seconds": 3506.832357303091, "mean_data_wait_seconds": 0.00015553832054138184}
{"step": 18410, "total_steps": 46238, "tokens": 9954312192, "loss": 3.1304794622190073, "ce_loss": 3.1292214628421897, "z_loss": 12.58032422354727, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18390317261219025, "tokens_per_second": 4292015.480566806, "eta_seconds": 3505.411341622913, "mean_data_wait_seconds": 0.00015437602996826172}
{"step": 18420, "total_steps": 46238, "tokens": 9959718912, "loss": 3.168743888537089, "ce_loss": 3.167489718307148, "z_loss": 12.541529698805375, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1963685154914856, "tokens_per_second": 4292014.928154393, "eta_seconds": 3504.1520767653265, "mean_data_wait_seconds": 0.0001068115234375}
{"step": 18430, "total_steps": 46238, "tokens": 9965125632, "loss": 3.1689981438896875, "ce_loss": 3.1676901524717156, "z_loss": 13.080160083192768, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19213227927684784, "tokens_per_second": 4292326.896716037, "eta_seconds": 3502.637768689643, "mean_data_wait_seconds": 0.00012236833572387695}
{"step": 18440, "total_steps": 46238, "tokens": 9970532352, "loss": 3.252791439041947, "ce_loss": 3.2514820153062995, "z_loss": 13.093943783731172, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1761469841003418, "tokens_per_second": 4292448.370210542, "eta_seconds": 3501.279057495765, "mean_data_wait_seconds": 7.447600364685059e-05}
{"step": 18450, "total_steps": 46238, "tokens": 9975939072, "loss": 3.273529690323454, "ce_loss": 3.2722655914046546, "z_loss": 12.640743407336148, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16544106602668762, "tokens_per_second": 4292623.087966705, "eta_seconds": 3499.8770113581722, "mean_data_wait_seconds": 8.219480514526367e-05}
{"step": 18460, "total_steps": 46238, "tokens": 9981345792, "loss": 3.2268332116531604, "ce_loss": 3.225575667439085, "z_loss": 12.575595183805985, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22296768426895142, "tokens_per_second": 4292844.665873913, "eta_seconds": 3498.436890434905, "mean_data_wait_seconds": 6.514787673950195e-05}
{"step": 18470, "total_steps": 46238, "tokens": 9986752512, "loss": 3.181413330815055, "ce_loss": 3.1801695119250906, "z_loss": 12.438412709669633, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17453840374946594, "tokens_per_second": 4293061.225444861, "eta_seconds": 3497.00100595335, "mean_data_wait_seconds": 7.104873657226562e-05}
{"step": 18480, "total_steps": 46238, "tokens": 9992159232, "loss": 3.181558021993348, "ce_loss": 3.1803227735288218, "z_loss": 12.352706865830855, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21645067632198334, "tokens_per_second": 4293339.871981799, "eta_seconds": 3495.514716162593, "mean_data_wait_seconds": 6.014108657836914e-05}
{"step": 18490, "total_steps": 46238, "tokens": 9997565952, "loss": 3.229851139314247, "ce_loss": 3.2285657055450208, "z_loss": 12.854023268728545, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.3310339152812958, "tokens_per_second": 4293621.180782358, "eta_seconds": 3494.026452810264, "mean_data_wait_seconds": 7.298588752746582e-05}
{"step": 18500, "total_steps": 46238, "tokens": 10002972672, "loss": 3.2430059097029944, "ce_loss": 3.24170468973391, "z_loss": 13.01213627873045, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18910686671733856, "tokens_per_second": 4293789.857230187, "eta_seconds": 3492.6299988220503, "mean_data_wait_seconds": 6.0945749282836914e-05}
{"step": 18510, "total_steps": 46238, "tokens": 10008379392, "loss": 3.239175301609617, "ce_loss": 3.237893155126861, "z_loss": 12.821129264253559, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17872072756290436, "tokens_per_second": 4293869.14748666, "eta_seconds": 3491.3063321398236, "mean_data_wait_seconds": 8.514523506164551e-05}
{"step": 18520, "total_steps": 46238, "tokens": 10013786112, "loss": 3.2486048683975683, "ce_loss": 3.2473172021634653, "z_loss": 12.876350576227361, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22109216451644897, "tokens_per_second": 4293998.153860634, "eta_seconds": 3489.94230715414, "mean_data_wait_seconds": 9.047985076904297e-05}
{"step": 18530, "total_steps": 46238, "tokens": 10019192832, "loss": 3.267115887367364, "ce_loss": 3.2658798387556365, "z_loss": 12.360370693784771, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1754925400018692, "tokens_per_second": 4293895.845149439, "eta_seconds": 3488.76629621151, "mean_data_wait_seconds": 6.809830665588379e-05}
{"step": 18540, "total_steps": 46238, "tokens": 10024599552, "loss": 3.1905198621027395, "ce_loss": 3.189250299424836, "z_loss": 12.695872617490364, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.190697580575943, "tokens_per_second": 4293969.836295863, "eta_seconds": 3487.447037335964, "mean_data_wait_seconds": 6.893277168273926e-05}
{"step": 18550, "total_steps": 46238, "tokens": 10030006272, "loss": 3.2690267382246074, "ce_loss": 3.267747275757067, "z_loss": 12.794709400697188, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.197560653090477, "tokens_per_second": 4293913.550471128, "eta_seconds": 3486.2335927460713, "mean_data_wait_seconds": 0.00011229515075683594}
{"step": 18560, "total_steps": 46238, "tokens": 10035412992, "loss": 3.281067035414956, "ce_loss": 3.2797620206168205, "z_loss": 13.05020903818535, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17448098957538605, "tokens_per_second": 4293978.344547794, "eta_seconds": 3484.9218471258737, "mean_data_wait_seconds": 0.00010368227958679199}
{"step": 18570, "total_steps": 46238, "tokens": 10040819712, "loss": 3.2177034110733955, "ce_loss": 3.216427232279922, "z_loss": 12.761827924034812, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1874438226222992, "tokens_per_second": 4293972.435385984, "eta_seconds": 3483.6675011527777, "mean_data_wait_seconds": 0.00011980533599853516}
{"step": 18580, "total_steps": 46238, "tokens": 10046226432, "loss": 3.2696252555558174, "ce_loss": 3.2683617039160295, "z_loss": 12.635981054017037, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18120789527893066, "tokens_per_second": 4294054.141233149, "eta_seconds": 3482.3420972763406, "mean_data_wait_seconds": 7.656216621398926e-05}
{"step": 18590, "total_steps": 46238, "tokens": 10051633152, "loss": 3.1947651957020615, "ce_loss": 3.1934981779618696, "z_loss": 12.67020412647363, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19750389456748962, "tokens_per_second": 4294145.15758216, "eta_seconds": 3481.009196348761, "mean_data_wait_seconds": 7.537007331848145e-05}
{"step": 18600, "total_steps": 46238, "tokens": 10057039872, "loss": 3.2434327403704324, "ce_loss": 3.242183416178732, "z_loss": 12.492805921670163, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19025763869285583, "tokens_per_second": 4294108.070654378, "eta_seconds": 3479.78015879859, "mean_data_wait_seconds": 6.428360939025879e-05}
{"step": 18610, "total_steps": 46238, "tokens": 10062446592, "loss": 3.228595245968212, "ce_loss": 3.2273304643053, "z_loss": 12.648107702081854, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18179908394813538, "tokens_per_second": 4294179.386871906, "eta_seconds": 3478.46328676105, "mean_data_wait_seconds": 0.00010368227958679199}
{"step": 18620, "total_steps": 46238, "tokens": 10067853312, "loss": 3.3017269228443955, "ce_loss": 3.300475517908732, "z_loss": 12.514105825713187, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17742615938186646, "tokens_per_second": 4294393.69876063, "eta_seconds": 3477.030675671243, "mean_data_wait_seconds": 0.00012370944023132324}
{"step": 18630, "total_steps": 46238, "tokens": 10073260032, "loss": 3.2075107025377676, "ce_loss": 3.206264781229424, "z_loss": 12.459171663631093, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18848104774951935, "tokens_per_second": 4294573.646569164, "eta_seconds": 3475.6260184114676, "mean_data_wait_seconds": 7.072091102600098e-05}
{"step": 18640, "total_steps": 46238, "tokens": 10078666752, "loss": 3.2279209928079085, "ce_loss": 3.226629168698282, "z_loss": 12.91814002123746, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18744595348834991, "tokens_per_second": 4294911.034004575, "eta_seconds": 3474.0941234556217, "mean_data_wait_seconds": 5.7578086853027344e-05}
{"step": 18650, "total_steps": 46238, "tokens": 10084073472, "loss": 3.2928286881157844, "ce_loss": 3.291569931940599, "z_loss": 12.587935678886645, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1719726324081421, "tokens_per_second": 4295074.45775287, "eta_seconds": 3472.70311858659, "mean_data_wait_seconds": 6.008148193359375e-05}
{"step": 18660, "total_steps": 46238, "tokens": 10089480192, "loss": 3.2468609033208904, "ce_loss": 3.245601966525569, "z_loss": 12.589389100219265, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17093677818775177, "tokens_per_second": 4295194.595874373, "eta_seconds": 3471.3472023645877, "mean_data_wait_seconds": 6.529688835144043e-05}
{"step": 18670, "total_steps": 46238, "tokens": 10094886912, "loss": 3.2543433687903662, "ce_loss": 3.25310165051258, "z_loss": 12.417652397444755, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19129161536693573, "tokens_per_second": 4295301.087524649, "eta_seconds": 3470.0023863960314, "mean_data_wait_seconds": 0.00015783309936523438}
{"step": 18680, "total_steps": 46238, "tokens": 10100293632, "loss": 3.3105765942371255, "ce_loss": 3.3093328259207984, "z_loss": 12.437483303474657, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1369549036026001, "tokens_per_second": 4295410.395436093, "eta_seconds": 3468.6553629033024, "mean_data_wait_seconds": 7.393956184387207e-05}
{"step": 18690, "total_steps": 46238, "tokens": 10105700352, "loss": 3.258896453814073, "ce_loss": 3.257663147016005, "z_loss": 12.332799087871205, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1820228546857834, "tokens_per_second": 4295416.3229923975, "eta_seconds": 3467.3918577522622, "mean_data_wait_seconds": 7.581710815429688e-05}
{"step": 18700, "total_steps": 46238, "tokens": 10111107072, "loss": 3.287683568217538, "ce_loss": 3.2864547552484455, "z_loss": 12.288311098561142, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16657604277133942, "tokens_per_second": 4295683.762301646, "eta_seconds": 3465.917345838951, "mean_data_wait_seconds": 6.338953971862793e-05}
{"step": 18710, "total_steps": 46238, "tokens": 10116513792, "loss": 3.2340352444937737, "ce_loss": 3.232768551869826, "z_loss": 12.666886062333077, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19277305901050568, "tokens_per_second": 4295665.858343279, "eta_seconds": 3464.6731460952124, "mean_data_wait_seconds": 5.4001808166503906e-05}
{"step": 18720, "total_steps": 46238, "tokens": 10121920512, "loss": 3.302417820150202, "ce_loss": 3.301138157194311, "z_loss": 12.79661589680296, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19347992539405823, "tokens_per_second": 4295835.489808917, "eta_seconds": 3463.277739404721, "mean_data_wait_seconds": 0.00010162591934204102}
{"step": 18730, "total_steps": 46238, "tokens": 10127327232, "loss": 3.1644110083580017, "ce_loss": 3.1631802392728403, "z_loss": 12.307747168974442, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18554216623306274, "tokens_per_second": 4295988.916745096, "eta_seconds": 3461.895501180235, "mean_data_wait_seconds": 0.00012490153312683105}
{"step": 18740, "total_steps": 46238, "tokens": 10132733952, "loss": 3.279920677344004, "ce_loss": 3.2786688208580017, "z_loss": 12.518640279769897, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18520595133304596, "tokens_per_second": 4296126.8447594475, "eta_seconds": 3460.5258460036084, "mean_data_wait_seconds": 9.655952453613281e-05}
{"step": 18750, "total_steps": 46238, "tokens": 10138140672, "loss": 3.2814540230866633, "ce_loss": 3.2801662105502505, "z_loss": 12.877978787277684, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17735880613327026, "tokens_per_second": 4296270.766674216, "eta_seconds": 3459.1514527619943, "mean_data_wait_seconds": 0.00011497735977172852}
{"step": 18760, "total_steps": 46238, "tokens": 10143547392, "loss": 3.290973663330078, "ce_loss": 3.2897173726197444, "z_loss": 12.56291963114883, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16844910383224487, "tokens_per_second": 4296477.318881595, "eta_seconds": 3457.7267471452965, "mean_data_wait_seconds": 0.00012981891632080078}
{"step": 18770, "total_steps": 46238, "tokens": 10148954112, "loss": 3.3001466845021104, "ce_loss": 3.2988935763185676, "z_loss": 12.530856782739813, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22828394174575806, "tokens_per_second": 4296625.710302774, "eta_seconds": 3456.348964348935, "mean_data_wait_seconds": 7.683038711547852e-05}
{"step": 18780, "total_steps": 46238, "tokens": 10154360832, "loss": 3.269389793728337, "ce_loss": 3.268118744546717, "z_loss": 12.711143031264797, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16681407392024994, "tokens_per_second": 4296795.910850991, "eta_seconds": 3454.9537404162784, "mean_data_wait_seconds": 0.00010800361633300781}
{"step": 18790, "total_steps": 46238, "tokens": 10159767552, "loss": 3.240771340601372, "ce_loss": 3.2394699075005273, "z_loss": 13.014153718948364, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17645297944545746, "tokens_per_second": 4296972.049663438, "eta_seconds": 3453.5538543152343, "mean_data_wait_seconds": 6.258487701416016e-05}
{"step": 18800, "total_steps": 46238, "tokens": 10165174272, "loss": 3.2204729463114883, "ce_loss": 3.219231235258507, "z_loss": 12.417006882754238, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1595563292503357, "tokens_per_second": 4297231.873105702, "eta_seconds": 3452.0868554572194, "mean_data_wait_seconds": 9.268522262573242e-05}
{"step": 18810, "total_steps": 46238, "tokens": 10170580992, "loss": 3.252836362882094, "ce_loss": 3.2515482595472625, "z_loss": 12.88154300776395, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19828404486179352, "tokens_per_second": 4297400.346412675, "eta_seconds": 3450.6933840545616, "mean_data_wait_seconds": 7.659196853637695e-05}
{"step": 18820, "total_steps": 46238, "tokens": 10175987712, "loss": 3.260361190998193, "ce_loss": 3.2590646039355886, "z_loss": 12.965945648424553, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17034967243671417, "tokens_per_second": 4297536.594599611, "eta_seconds": 3449.3258865154753, "mean_data_wait_seconds": 7.382035255432129e-05}
{"step": 18830, "total_steps": 46238, "tokens": 10181394432, "loss": 3.222313693075469, "ce_loss": 3.221084762703289, "z_loss": 12.289261695110437, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16705800592899323, "tokens_per_second": 4297487.896512438, "eta_seconds": 3448.1068616913353, "mean_data_wait_seconds": 7.984042167663574e-05}
{"step": 18840, "total_steps": 46238, "tokens": 10186801152, "loss": 3.249448850299373, "ce_loss": 3.248181837977785, "z_loss": 12.669992988759821, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16465014219284058, "tokens_per_second": 4297617.560629471, "eta_seconds": 3446.7447545123987, "mean_data_wait_seconds": 0.00014415383338928223}
{"step": 18850, "total_steps": 46238, "tokens": 10192207872, "loss": 3.293452324289264, "ce_loss": 3.292156559048277, "z_loss": 12.957583232359452, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19595184922218323, "tokens_per_second": 4297745.907664954, "eta_seconds": 3445.3837853911496, "mean_data_wait_seconds": 7.957220077514648e-05}
{"step": 18860, "total_steps": 46238, "tokens": 10197614592, "loss": 3.35768615657633, "ce_loss": 3.3564503662513965, "z_loss": 12.358387051206646, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21618640422821045, "tokens_per_second": 4297928.153356443, "eta_seconds": 3443.979707394708, "mean_data_wait_seconds": 6.312131881713867e-05}
{"step": 18870, "total_steps": 46238, "tokens": 10203021312, "loss": 3.244376489610383, "ce_loss": 3.2431617415312566, "z_loss": 12.147278879628036, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17462287843227386, "tokens_per_second": 4298239.798776834, "eta_seconds": 3442.4721087480307, "mean_data_wait_seconds": 9.447336196899414e-05}
{"step": 18880, "total_steps": 46238, "tokens": 10208428032, "loss": 3.2281207153291414, "ce_loss": 3.2268800590977524, "z_loss": 12.406302495436234, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1745551973581314, "tokens_per_second": 4298219.341037446, "eta_seconds": 3441.230595837836, "mean_data_wait_seconds": 9.235739707946777e-05}
{"step": 18890, "total_steps": 46238, "tokens": 10213834752, "loss": 3.2681617429762175, "ce_loss": 3.266912285125617, "z_loss": 12.494527592803493, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17300091683864594, "tokens_per_second": 4298436.499896565, "eta_seconds": 3439.798909290807, "mean_data_wait_seconds": 6.479024887084961e-05}
{"step": 18900, "total_steps": 46238, "tokens": 10219241472, "loss": 3.2139818578055412, "ce_loss": 3.2127549106424507, "z_loss": 12.269434394258441, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17286373674869537, "tokens_per_second": 4298604.426242924, "eta_seconds": 3438.40674749371, "mean_data_wait_seconds": 0.00011742115020751953}
{"step": 18910, "total_steps": 46238, "tokens": 10224648192, "loss": 3.2196634447935857, "ce_loss": 3.2183975107742078, "z_loss": 12.659395571910974, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1705377846956253, "tokens_per_second": 4298795.247185681, "eta_seconds": 3436.9963895518877, "mean_data_wait_seconds": 7.000565528869629e-05}
{"step": 18920, "total_steps": 46238, "tokens": 10230054912, "loss": 3.2252353194988137, "ce_loss": 3.2239829554702295, "z_loss": 12.523293683023164, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21700622141361237, "tokens_per_second": 4299057.309820977, "eta_seconds": 3435.529224106817, "mean_data_wait_seconds": 0.0001055002212524414}
{"step": 18930, "total_steps": 46238, "tokens": 10235461632, "loss": 3.2397474104707893, "ce_loss": 3.2384968743179785, "z_loss": 12.505276203155518, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18492093682289124, "tokens_per_second": 4299347.8416272355, "eta_seconds": 3434.039498049083, "mean_data_wait_seconds": 0.00011110305786132812}
{"step": 18940, "total_steps": 46238, "tokens": 10240868352, "loss": 3.1499294024525266, "ce_loss": 3.1486571358911917, "z_loss": 12.722611629601682, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15792176127433777, "tokens_per_second": 4299522.756671839, "eta_seconds": 3432.64227665686, "mean_data_wait_seconds": 8.90493392944336e-05}
{"step": 18950, "total_steps": 46238, "tokens": 10246275072, "loss": 3.1134918711402197, "ce_loss": 3.112199732751557, "z_loss": 12.921202861901486, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20039811730384827, "tokens_per_second": 4299646.481334965, "eta_seconds": 3431.286020384484, "mean_data_wait_seconds": 0.00015679001808166504}
{"step": 18960, "total_steps": 46238, "tokens": 10251681792, "loss": 3.200804703163378, "ce_loss": 3.1994959874586626, "z_loss": 13.086930852947813, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18487544357776642, "tokens_per_second": 4299657.407890929, "eta_seconds": 3430.0198236571027, "mean_data_wait_seconds": 0.00010624527931213379}
{"step": 18970, "total_steps": 46238, "tokens": 10257088512, "loss": 3.225003470074047, "ce_loss": 3.2237112413753164, "z_loss": 12.922268267833825, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17733576893806458, "tokens_per_second": 4299833.456238186, "eta_seconds": 3428.6219626975594, "mean_data_wait_seconds": 0.00012010335922241211}
{"step": 18980, "total_steps": 46238, "tokens": 10262495232, "loss": 3.150886448946866, "ce_loss": 3.149627681934472, "z_loss": 12.58763319795782, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1582632064819336, "tokens_per_second": 4299818.877532222, "eta_seconds": 3427.3761578669105, "mean_data_wait_seconds": 9.560585021972656e-05}
{"step": 18990, "total_steps": 46238, "tokens": 10267901952, "loss": 3.071858393423485, "ce_loss": 3.0705960880626333, "z_loss": 12.623341025728168, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19786487519741058, "tokens_per_second": 4299772.238458557, "eta_seconds": 3426.155890824865, "mean_data_wait_seconds": 0.00012257695198059082}
{"step": 19000, "total_steps": 46238, "tokens": 10273308672, "loss": 3.176349356318965, "ce_loss": 3.1750939163294705, "z_loss": 12.554442542971987, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2007060945034027, "tokens_per_second": 4299850.595584008, "eta_seconds": 3424.8360347971275, "mean_data_wait_seconds": 7.647275924682617e-05}
{"step": 19010, "total_steps": 46238, "tokens": 10278715392, "loss": 3.136453355803634, "ce_loss": 3.1351714802510813, "z_loss": 12.819108977462307, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19147059321403503, "tokens_per_second": 4299879.402039635, "eta_seconds": 3423.5556785655886, "mean_data_wait_seconds": 0.00013500452041625977}
{"step": 19020, "total_steps": 46238, "tokens": 10284122112, "loss": 3.1446072188290684, "ce_loss": 3.143305354046099, "z_loss": 13.018970713470921, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19042424857616425, "tokens_per_second": 4299904.541167807, "eta_seconds": 3422.2782582990644, "mean_data_wait_seconds": 0.00011172890663146973}
{"step": 19030, "total_steps": 46238, "tokens": 10289528832, "loss": 3.0896308349840567, "ce_loss": 3.088314970334371, "z_loss": 13.158790747324625, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1858636438846588, "tokens_per_second": 4299979.780899639, "eta_seconds": 3420.960993663642, "mean_data_wait_seconds": 0.00010657310485839844}
{"step": 19040, "total_steps": 46238, "tokens": 10294935552, "loss": 3.146919026519313, "ce_loss": 3.14571465326078, "z_loss": 12.043733748522671, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1997966766357422, "tokens_per_second": 4299969.078476072, "eta_seconds": 3419.712122490749, "mean_data_wait_seconds": 0.00013068318367004395}
{"step": 19050, "total_steps": 46238, "tokens": 10300342272, "loss": 3.162381780870033, "ce_loss": 3.1611531683892915, "z_loss": 12.286095171263725, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18435512483119965, "tokens_per_second": 4300147.461536615, "eta_seconds": 3418.312928912296, "mean_data_wait_seconds": 6.955862045288086e-05}
{"step": 19060, "total_steps": 46238, "tokens": 10305748992, "loss": 3.2527062658107644, "ce_loss": 3.2514698180285366, "z_loss": 12.364852247816144, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16754131019115448, "tokens_per_second": 4300250.686990774, "eta_seconds": 3416.9735705065245, "mean_data_wait_seconds": 6.0230493545532227e-05}
{"step": 19070, "total_steps": 46238, "tokens": 10311155712, "loss": 3.2320996486779414, "ce_loss": 3.230834597891027, "z_loss": 12.65069352496754, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1833706647157669, "tokens_per_second": 4300448.471493089, "eta_seconds": 3415.559172808846, "mean_data_wait_seconds": 5.78761100769043e-05}
{"step": 19080, "total_steps": 46238, "tokens": 10316562432, "loss": 3.164664134834752, "ce_loss": 3.163391785188155, "z_loss": 12.723749406409986, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18346406519412994, "tokens_per_second": 4300862.113099608, "eta_seconds": 3413.9735517858817, "mean_data_wait_seconds": 6.085634231567383e-05}
{"step": 19090, "total_steps": 46238, "tokens": 10321969152, "loss": 3.2114512541077356, "ce_loss": 3.2101887265841165, "z_loss": 12.625276406606039, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16860295832157135, "tokens_per_second": 4300916.750557649, "eta_seconds": 3412.673073036563, "mean_data_wait_seconds": 6.008148193359375e-05}
{"step": 19100, "total_steps": 46238, "tokens": 10327375872, "loss": 3.254939464005557, "ce_loss": 3.2536849668531707, "z_loss": 12.544687993598707, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16651488840579987, "tokens_per_second": 4299716.17444033, "eta_seconds": 3412.3685073026477, "mean_data_wait_seconds": 0.00011056661605834961}
{"step": 19110, "total_steps": 46238, "tokens": 10332782592, "loss": 3.2855123967835396, "ce_loss": 3.284236241470684, "z_loss": 12.761749932260225, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21430526673793793, "tokens_per_second": 4299981.000155169, "eta_seconds": 3410.900965253273, "mean_data_wait_seconds": 6.660819053649902e-05}
{"step": 19120, "total_steps": 46238, "tokens": 10338189312, "loss": 3.258815839435115, "ce_loss": 3.25758306907885, "z_loss": 12.327838240247784, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18524755537509918, "tokens_per_second": 4299246.771382442, "eta_seconds": 3410.2258845880488, "mean_data_wait_seconds": 0.07422971725463867}
{"step": 19130, "total_steps": 46238, "tokens": 10343596032, "loss": 3.3006614952376396, "ce_loss": 3.2993466998591567, "z_loss": 13.14802545489687, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20754559338092804, "tokens_per_second": 4299251.196388155, "eta_seconds": 3408.964778869551, "mean_data_wait_seconds": 9.784102439880371e-05}
{"step": 19140, "total_steps": 46238, "tokens": 10349002752, "loss": 3.258424363353036, "ce_loss": 3.2571907404697305, "z_loss": 12.33621722279173, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20504818856716156, "tokens_per_second": 4299272.287350728, "eta_seconds": 3407.690465920198, "mean_data_wait_seconds": 5.888938903808594e-05}
{"step": 19150, "total_steps": 46238, "tokens": 10354409472, "loss": 3.2267573800953953, "ce_loss": 3.2254986781062502, "z_loss": 12.587012623295639, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15097388625144958, "tokens_per_second": 4298111.7579197055, "eta_seconds": 3407.3526443361484, "mean_data_wait_seconds": 6.842613220214844e-05}
{"step": 19160, "total_steps": 46238, "tokens": 10359816192, "loss": 3.283323351180915, "ce_loss": 3.282090909553297, "z_loss": 12.32431057727698, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17157557606697083, "tokens_per_second": 4297978.075904131, "eta_seconds": 3406.2006565541515, "mean_data_wait_seconds": 8.195638656616211e-05}
{"step": 19170, "total_steps": 46238, "tokens": 10365222912, "loss": 3.1992869648066433, "ce_loss": 3.1979882897752705, "z_loss": 12.986901196566494, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1743500828742981, "tokens_per_second": 4297190.929270307, "eta_seconds": 3405.5663955534365, "mean_data_wait_seconds": 0.00012153387069702148}
{"step": 19180, "total_steps": 46238, "tokens": 10370629632, "loss": 3.203631908604593, "ce_loss": 3.202362600601081, "z_loss": 12.693130038001321, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19456464052200317, "tokens_per_second": 4297163.78983567, "eta_seconds": 3404.3296973233205, "mean_data_wait_seconds": 9.632110595703125e-05}
{"step": 19190, "total_steps": 46238, "tokens": 10376036352, "loss": 3.1867347160975137, "ce_loss": 3.185474780472842, "z_loss": 12.599498293616556, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17644187808036804, "tokens_per_second": 4297382.975475131, "eta_seconds": 3402.8979189091656, "mean_data_wait_seconds": 0.00011840462684631348}
{"step": 19200, "total_steps": 46238, "tokens": 10381443072, "loss": 3.306737999121348, "ce_loss": 3.305474243380807, "z_loss": 12.637599901719527, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17543266713619232, "tokens_per_second": 4297204.766066619, "eta_seconds": 3401.780845873095, "mean_data_wait_seconds": 9.843707084655762e-05}
{"step": 19210, "total_steps": 46238, "tokens": 10386849792, "loss": 3.322022866119038, "ce_loss": 3.320771715857766, "z_loss": 12.511116620266076, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21068978309631348, "tokens_per_second": 4297251.315851675, "eta_seconds": 3400.485815222536, "mean_data_wait_seconds": 0.00029546022415161133}
{"step": 19220, "total_steps": 46238, "tokens": 10392256512, "loss": 3.2184213905623467, "ce_loss": 3.2171208786241934, "z_loss": 13.00532687071598, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16435419023036957, "tokens_per_second": 4297376.466437441, "eta_seconds": 3399.128640016404, "mean_data_wait_seconds": 9.447336196899414e-05}
{"step": 19230, "total_steps": 46238, "tokens": 10397663232, "loss": 3.273766010096579, "ce_loss": 3.2724648677941524, "z_loss": 13.011396436980277, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17235197126865387, "tokens_per_second": 4297269.841294897, "eta_seconds": 3397.954804625441, "mean_data_wait_seconds": 7.975101470947266e-05}
{"step": 19240, "total_steps": 46238, "tokens": 10403069952, "loss": 3.2255024765477036, "ce_loss": 3.2242160854917583, "z_loss": 12.863870599053122, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17642056941986084, "tokens_per_second": 4297259.293843569, "eta_seconds": 3396.7049660958505, "mean_data_wait_seconds": 7.957220077514648e-05}
{"step": 19250, "total_steps": 46238, "tokens": 10408476672, "loss": 3.2297660943233604, "ce_loss": 3.228529545393857, "z_loss": 12.36565079833522, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16762587428092957, "tokens_per_second": 4297233.664193136, "eta_seconds": 3395.467038616268, "mean_data_wait_seconds": 0.00012832880020141602}
{"step": 19260, "total_steps": 46238, "tokens": 10413883392, "loss": 3.245539356361736, "ce_loss": 3.2442513144377507, "z_loss": 12.880210486325351, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1935289353132248, "tokens_per_second": 4297370.598341917, "eta_seconds": 3394.100697209522, "mean_data_wait_seconds": 0.0003542602062225342}
{"step": 19270, "total_steps": 46238, "tokens": 10419290112, "loss": 3.2657633152875034, "ce_loss": 3.264503811344956, "z_loss": 12.595169551444776, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1562463939189911, "tokens_per_second": 4297460.644207482, "eta_seconds": 3392.771459967339, "mean_data_wait_seconds": 7.915496826171875e-05}
{"step": 19280, "total_steps": 46238, "tokens": 10424696832, "loss": 3.2691005937980884, "ce_loss": 3.267864612015811, "z_loss": 12.359896024068197, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1718531996011734, "tokens_per_second": 4297691.315765771, "eta_seconds": 3391.3313063069577, "mean_data_wait_seconds": 9.906291961669922e-05}
{"step": 19290, "total_steps": 46238, "tokens": 10430103552, "loss": 3.299945233446179, "ce_loss": 3.2986369837414133, "z_loss": 13.082868597724222, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19583696126937866, "tokens_per_second": 4297759.111060786, "eta_seconds": 3390.0197771679937, "mean_data_wait_seconds": 7.781386375427246e-05}
{"step": 19300, "total_steps": 46238, "tokens": 10435510272, "loss": 3.2589516856453637, "ce_loss": 3.25764539566907, "z_loss": 13.06333117051558, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18163347244262695, "tokens_per_second": 4297800.16390681, "eta_seconds": 3388.729375160356, "mean_data_wait_seconds": 7.200241088867188e-05}
{"step": 19310, "total_steps": 46238, "tokens": 10440916992, "loss": 3.288448396957282, "ce_loss": 3.2871768546826914, "z_loss": 12.715382901105015, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17401929199695587, "tokens_per_second": 4297904.574804064, "eta_seconds": 3387.3890614855522, "mean_data_wait_seconds": 0.00012138485908508301}
{"step": 19320, "total_steps": 46238, "tokens": 10446323712, "loss": 3.256311217943827, "ce_loss": 3.255041989413175, "z_loss": 12.692196022380482, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17610272765159607, "tokens_per_second": 4298028.434532405, "eta_seconds": 3386.0334908610935, "mean_data_wait_seconds": 8.606910705566406e-05}
{"step": 19330, "total_steps": 46238, "tokens": 10451730432, "loss": 3.2810486555099487, "ce_loss": 3.2797932787375017, "z_loss": 12.55363046761715, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19757217168807983, "tokens_per_second": 4298159.533782464, "eta_seconds": 3384.672297446716, "mean_data_wait_seconds": 9.53376293182373e-05}
{"step": 19340, "total_steps": 46238, "tokens": 10457137152, "loss": 3.2338422482663933, "ce_loss": 3.232622161056056, "z_loss": 12.200352950529618, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18763616681098938, "tokens_per_second": 4298222.116554243, "eta_seconds": 3383.3651192643006, "mean_data_wait_seconds": 7.763504981994629e-05}
{"step": 19350, "total_steps": 46238, "tokens": 10462543872, "loss": 3.226690086451444, "ce_loss": 3.2254165754173743, "z_loss": 12.735079563025272, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18354901671409607, "tokens_per_second": 4298461.417291242, "eta_seconds": 3381.9189362785537, "mean_data_wait_seconds": 0.00011307001113891602}
{"step": 19360, "total_steps": 46238, "tokens": 10467950592, "loss": 3.211187346415086, "ce_loss": 3.2099495548190493, "z_loss": 12.3777916431427, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16089016199111938, "tokens_per_second": 4298739.654289607, "eta_seconds": 3380.442295336316, "mean_data_wait_seconds": 6.207823753356934e-05}
{"step": 19370, "total_steps": 46238, "tokens": 10473357312, "loss": 3.269615964456038, "ce_loss": 3.268376350402832, "z_loss": 12.39657770503651, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1855759620666504, "tokens_per_second": 4299007.936951436, "eta_seconds": 3378.9736695161855, "mean_data_wait_seconds": 0.00017893314361572266}
{"step": 19380, "total_steps": 46238, "tokens": 10478764032, "loss": 3.262618913795009, "ce_loss": 3.2613822005011817, "z_loss": 12.367267767588297, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20213177800178528, "tokens_per_second": 4299135.312148458, "eta_seconds": 3377.6159273161684, "mean_data_wait_seconds": 0.0001379251480102539}
{"step": 19390, "total_steps": 46238, "tokens": 10484170752, "loss": 3.220494149309216, "ce_loss": 3.219237910978722, "z_loss": 12.562066952387491, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17573630809783936, "tokens_per_second": 4299316.157109179, "eta_seconds": 3376.216275697212, "mean_data_wait_seconds": 0.00010073184967041016}
{"step": 19400, "total_steps": 46238, "tokens": 10489577472, "loss": 3.2786288261413574, "ce_loss": 3.277389741305149, "z_loss": 12.391025283119895, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16247548162937164, "tokens_per_second": 4299457.897660956, "eta_seconds": 3374.8474364393514, "mean_data_wait_seconds": 6.788969039916992e-05}
{"step": 19410, "total_steps": 46238, "tokens": 10494984192, "loss": 3.228547540577975, "ce_loss": 3.22726527669213, "z_loss": 12.822966481700089, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18771445751190186, "tokens_per_second": 4299410.953357711, "eta_seconds": 3373.6267366282664, "mean_data_wait_seconds": 7.50124454498291e-05}
{"step": 19420, "total_steps": 46238, "tokens": 10500390912, "loss": 3.260390296126857, "ce_loss": 3.259075874632055, "z_loss": 13.14439481677431, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18714438378810883, "tokens_per_second": 4299536.628731641, "eta_seconds": 3372.2706133282204, "mean_data_wait_seconds": 9.009242057800293e-05}
{"step": 19430, "total_steps": 46238, "tokens": 10505797632, "loss": 3.2263184641346787, "ce_loss": 3.2250251589399395, "z_loss": 12.932986266685255, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17948971688747406, "tokens_per_second": 4299383.851795221, "eta_seconds": 3371.132888715687, "mean_data_wait_seconds": 6.285309791564941e-05}
{"step": 19440, "total_steps": 46238, "tokens": 10511204352, "loss": 3.2292660799893467, "ce_loss": 3.2279889005603213, "z_loss": 12.771429466478752, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14556635916233063, "tokens_per_second": 4299584.642035001, "eta_seconds": 3369.7179588823305, "mean_data_wait_seconds": 7.37905502319336e-05}
{"step": 19450, "total_steps": 46238, "tokens": 10516611072, "loss": 3.1844900449117026, "ce_loss": 3.183190790089694, "z_loss": 12.992386955203433, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17385929822921753, "tokens_per_second": 4299798.3648841735, "eta_seconds": 3368.2930302686736, "mean_data_wait_seconds": 9.08970832824707e-05}
{"step": 19460, "total_steps": 46238, "tokens": 10522017792, "loss": 3.2506354613737627, "ce_loss": 3.2493422356518833, "z_loss": 12.931880748633182, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17068128287792206, "tokens_per_second": 4299975.575183052, "eta_seconds": 3366.8968325206556, "mean_data_wait_seconds": 0.00010210275650024414}
{"step": 19470, "total_steps": 46238, "tokens": 10527424512, "loss": 3.2156792842980586, "ce_loss": 3.2144234361070576, "z_loss": 12.558733463287354, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20424483716487885, "tokens_per_second": 4299986.5112735, "eta_seconds": 3365.630888854549, "mean_data_wait_seconds": 7.078051567077637e-05}
{"step": 19480, "total_steps": 46238, "tokens": 10532831232, "loss": 3.1863579298510696, "ce_loss": 3.1850626450596433, "z_loss": 12.952808344002927, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15060120820999146, "tokens_per_second": 4300301.683400778, "eta_seconds": 3364.1269308713595, "mean_data_wait_seconds": 0.00010761618614196777}
{"step": 19490, "total_steps": 46238, "tokens": 10538237952, "loss": 3.1814238663875694, "ce_loss": 3.1801233436122085, "z_loss": 13.004830613280788, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1788789927959442, "tokens_per_second": 4300367.549125973, "eta_seconds": 3362.81813561243, "mean_data_wait_seconds": 9.149312973022461e-05}
{"step": 19500, "total_steps": 46238, "tokens": 10543644672, "loss": 3.1578622207497107, "ce_loss": 3.1565929597074334, "z_loss": 12.692581805315884, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16336967051029205, "tokens_per_second": 4300453.223474115, "eta_seconds": 3361.49389675765, "mean_data_wait_seconds": 0.00011143088340759277}
{"step": 19510, "total_steps": 46238, "tokens": 10549051392, "loss": 3.1584952347206348, "ce_loss": 3.1571765650402415, "z_loss": 13.186451016050396, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17065806686878204, "tokens_per_second": 4300376.182217799, "eta_seconds": 3360.296851180944, "mean_data_wait_seconds": 7.411837577819824e-05}
{"step": 19520, "total_steps": 46238, "tokens": 10554458112, "loss": 3.196525971094767, "ce_loss": 3.195278424205202, "z_loss": 12.475693052465266, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1740848869085312, "tokens_per_second": 4300560.677127341, "eta_seconds": 3358.895480960627, "mean_data_wait_seconds": 0.00010734796524047852}
{"step": 19530, "total_steps": 46238, "tokens": 10559864832, "loss": 3.167407882936073, "ce_loss": 3.1661048621842354, "z_loss": 13.03058588143551, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16853700578212738, "tokens_per_second": 4300787.770785967, "eta_seconds": 3357.460975425241, "mean_data_wait_seconds": 0.00011172890663146973}
{"step": 19540, "total_steps": 46238, "tokens": 10565271552, "loss": 3.168223227515365, "ce_loss": 3.1669577811703538, "z_loss": 12.654607339338822, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1676676720380783, "tokens_per_second": 4301017.502119884, "eta_seconds": 3356.024563230821, "mean_data_wait_seconds": 5.8084726333618164e-05}
{"step": 19550, "total_steps": 46238, "tokens": 10570678272, "loss": 3.2125951304580225, "ce_loss": 3.211335868546457, "z_loss": 12.592279441428907, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1770995557308197, "tokens_per_second": 4301430.462649881, "eta_seconds": 3354.4454081703598, "mean_data_wait_seconds": 0.0001271665096282959}
{"step": 19560, "total_steps": 46238, "tokens": 10576084992, "loss": 3.1249280499689505, "ce_loss": 3.1236391898357505, "z_loss": 12.88858413696289, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2526777386665344, "tokens_per_second": 4301516.111186877, "eta_seconds": 3353.1216834197226, "mean_data_wait_seconds": 8.738040924072266e-05}
{"step": 19570, "total_steps": 46238, "tokens": 10581491712, "loss": 3.129682300668774, "ce_loss": 3.128367871949167, "z_loss": 13.144191481850363, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21122366189956665, "tokens_per_second": 4301736.48200336, "eta_seconds": 3351.693039385191, "mean_data_wait_seconds": 0.00013270974159240723}
{"step": 19580, "total_steps": 46238, "tokens": 10586898432, "loss": 3.1773878661069004, "ce_loss": 3.1761027628725227, "z_loss": 12.85101691159335, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.180515319108963, "tokens_per_second": 4302022.328201548, "eta_seconds": 3350.2135517797738, "mean_data_wait_seconds": 7.835030555725098e-05}
{"step": 19590, "total_steps": 46238, "tokens": 10592305152, "loss": 3.159355943853205, "ce_loss": 3.1580955042983545, "z_loss": 12.604490504120335, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19299198687076569, "tokens_per_second": 4302261.429059636, "eta_seconds": 3348.7706457552636, "mean_data_wait_seconds": 0.0001049041748046875}
{"step": 19600, "total_steps": 46238, "tokens": 10597711872, "loss": 3.14577558185115, "ce_loss": 3.144493393825762, "z_loss": 12.822063381021673, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1554165929555893, "tokens_per_second": 4302529.138561905, "eta_seconds": 3347.305642841908, "mean_data_wait_seconds": 5.969405174255371e-05}
{"step": 19610, "total_steps": 46238, "tokens": 10603118592, "loss": 3.2093843348098523, "ce_loss": 3.208140882578763, "z_loss": 12.434353438290684, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17752841114997864, "tokens_per_second": 4302812.515672574, "eta_seconds": 3345.828639189426, "mean_data_wait_seconds": 9.992718696594238e-05}
{"step": 19620, "total_steps": 46238, "tokens": 10608525312, "loss": 3.1946100430055098, "ce_loss": 3.1933148087877217, "z_loss": 12.952726826523289, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1675579845905304, "tokens_per_second": 4303056.466704605, "eta_seconds": 3344.382472168919, "mean_data_wait_seconds": 7.298588752746582e-05}
{"step": 19630, "total_steps": 46238, "tokens": 10613932032, "loss": 3.2504716587789133, "ce_loss": 3.2492177775411895, "z_loss": 12.53876661531853, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22136926651000977, "tokens_per_second": 4303291.657706471, "eta_seconds": 3342.9432741881915, "mean_data_wait_seconds": 8.71419906616211e-05}
{"step": 19640, "total_steps": 46238, "tokens": 10619338752, "loss": 3.1637276392994504, "ce_loss": 3.1624565034201653, "z_loss": 12.71105942581639, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17973463237285614, "tokens_per_second": 4303275.408263702, "eta_seconds": 3341.6994776549027, "mean_data_wait_seconds": 0.00011178851127624512}
{"step": 19650, "total_steps": 46238, "tokens": 10624745472, "loss": 3.1650148178591873, "ce_loss": 3.1637658458767515, "z_loss": 12.48946545340798, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17091116309165955, "tokens_per_second": 4303361.9401984215, "eta_seconds": 3340.375888377448, "mean_data_wait_seconds": 9.170174598693848e-05}
{"step": 19660, "total_steps": 46238, "tokens": 10630152192, "loss": 3.2046474055810408, "ce_loss": 3.203373645291184, "z_loss": 12.737689654032389, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17105340957641602, "tokens_per_second": 4301841.362184587, "eta_seconds": 3340.299777280217, "mean_data_wait_seconds": 0.00010254979133605957}
{"step": 19670, "total_steps": 46238, "tokens": 10635558912, "loss": 3.185862365997199, "ce_loss": 3.1845383048057556, "z_loss": 13.240829214905247, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19675026834011078, "tokens_per_second": 4301104.935067927, "eta_seconds": 3339.6146434109605, "mean_data_wait_seconds": 0.00013530254364013672}
{"step": 19680, "total_steps": 46238, "tokens": 10640965632, "loss": 3.203217976021044, "ce_loss": 3.2019220012606997, "z_loss": 12.959846713326193, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16969862580299377, "tokens_per_second": 4301121.838850094, "eta_seconds": 3338.3444696462684, "mean_data_wait_seconds": 8.025765419006348e-05}
{"step": 19690, "total_steps": 46238, "tokens": 10646372352, "loss": 3.2074919668110935, "ce_loss": 3.206232235287175, "z_loss": 12.597037647709701, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17842990159988403, "tokens_per_second": 4301056.037217131, "eta_seconds": 3337.138474784165, "mean_data_wait_seconds": 0.00014317035675048828}
{"step": 19700, "total_steps": 46238, "tokens": 10651779072, "loss": 3.2544087966283164, "ce_loss": 3.253163800095067, "z_loss": 12.449949791937163, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16706357896327972, "tokens_per_second": 4300211.234481283, "eta_seconds": 3336.5367610204657, "mean_data_wait_seconds": 0.00010311603546142578}
{"step": 19710, "total_steps": 46238, "tokens": 10657185792, "loss": 3.2813721136613325, "ce_loss": 3.2800897793336348, "z_loss": 12.823187647443829, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14789536595344543, "tokens_per_second": 4300117.662382185, "eta_seconds": 3335.352022915246, "mean_data_wait_seconds": 0.00012028217315673828}
{"step": 19720, "total_steps": 46238, "tokens": 10662592512, "loss": 3.263196531570319, "ce_loss": 3.261953353881836, "z_loss": 12.43184708103989, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17981165647506714, "tokens_per_second": 4300173.651546201, "eta_seconds": 3334.0512699632227, "mean_data_wait_seconds": 8.466839790344238e-05}
{"step": 19730, "total_steps": 46238, "tokens": 10667999232, "loss": 3.26708517110709, "ce_loss": 3.265864417408452, "z_loss": 12.207541826999549, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18685093522071838, "tokens_per_second": 4300263.498300151, "eta_seconds": 3332.724310885863, "mean_data_wait_seconds": 0.00016891956329345703}
{"step": 19740, "total_steps": 46238, "tokens": 10673405952, "loss": 3.2040549661173965, "ce_loss": 3.2028195316141304, "z_loss": 12.354598941224994, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1882946789264679, "tokens_per_second": 4300245.043977154, "eta_seconds": 3331.4813080396425, "mean_data_wait_seconds": 0.00013571977615356445}
{"step": 19750, "total_steps": 46238, "tokens": 10678812672, "loss": 3.1540623343352117, "ce_loss": 3.152780350410577, "z_loss": 12.819697986949574, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1685519963502884, "tokens_per_second": 4300266.528440671, "eta_seconds": 3330.2073648892847, "mean_data_wait_seconds": 0.0001308917999267578}
{"step": 19760, "total_steps": 46238, "tokens": 10684219392, "loss": 3.2212835532246213, "ce_loss": 3.2199848998676646, "z_loss": 12.986532269102154, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21841883659362793, "tokens_per_second": 4300240.800093759, "eta_seconds": 3328.9699831897506, "mean_data_wait_seconds": 0.00011247396469116211}
{"step": 19770, "total_steps": 46238, "tokens": 10689626112, "loss": 3.2789676514538852, "ce_loss": 3.2776825120954802, "z_loss": 12.85136918588118, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15968744456768036, "tokens_per_second": 4300155.075680214, "eta_seconds": 3327.7790154431577, "mean_data_wait_seconds": 0.00014701485633850098}
{"step": 19780, "total_steps": 46238, "tokens": 10695032832, "loss": 3.2747758771433975, "ce_loss": 3.2734836087082373, "z_loss": 12.92248181140784, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1925479620695114, "tokens_per_second": 4300251.451407123, "eta_seconds": 3326.4471312065434, "mean_data_wait_seconds": 8.958578109741211e-05}
{"step": 19790, "total_steps": 46238, "tokens": 10700439552, "loss": 3.236466463768121, "ce_loss": 3.2351948564702813, "z_loss": 12.71612618186257, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20848000049591064, "tokens_per_second": 4300477.252643097, "eta_seconds": 3325.0152352768896, "mean_data_wait_seconds": 0.00010773539543151855}
{"step": 19800, "total_steps": 46238, "tokens": 10705846272, "loss": 3.2624334584582937, "ce_loss": 3.261145384022684, "z_loss": 12.880738019943237, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16534948348999023, "tokens_per_second": 4300450.913791622, "eta_seconds": 3323.778354999869, "mean_data_wait_seconds": 0.0001379847526550293}
{"step": 19810, "total_steps": 46238, "tokens": 10711252992, "loss": 3.2369060570543464, "ce_loss": 3.2356420303836013, "z_loss": 12.639962861032197, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17063869535923004, "tokens_per_second": 4300354.0746933045, "eta_seconds": 3322.595929503555, "mean_data_wait_seconds": 0.00010243058204650879}
{"step": 19820, "total_steps": 46238, "tokens": 10716659712, "loss": 3.3267834674228323, "ce_loss": 3.3255275126659507, "z_loss": 12.559676517139781, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1642342507839203, "tokens_per_second": 4300605.706182982, "eta_seconds": 3321.144322406824, "mean_data_wait_seconds": 9.918212890625e-05}
{"step": 19830, "total_steps": 46238, "tokens": 10722066432, "loss": 3.263999006964944, "ce_loss": 3.2627274105043123, "z_loss": 12.716016668261904, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17481429874897003, "tokens_per_second": 4300878.989465846, "eta_seconds": 3319.676172933482, "mean_data_wait_seconds": 9.047985076904297e-05}
{"step": 19840, "total_steps": 46238, "tokens": 10727473152, "loss": 3.2574242227005237, "ce_loss": 3.2561558123790855, "z_loss": 12.683683576005878, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20839695632457733, "tokens_per_second": 4300867.4256034745, "eta_seconds": 3318.4279754908775, "mean_data_wait_seconds": 0.00010293722152709961}
{"step": 19850, "total_steps": 46238, "tokens": 10732879872, "loss": 3.2324177315740874, "ce_loss": 3.231117689248287, "z_loss": 13.000558332963424, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15871472656726837, "tokens_per_second": 4301192.868350375, "eta_seconds": 3316.919864017089, "mean_data_wait_seconds": 8.147954940795898e-05}
{"step": 19860, "total_steps": 46238, "tokens": 10738286592, "loss": 3.2630713040178474, "ce_loss": 3.261807199680444, "z_loss": 12.641467282266328, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1717744618654251, "tokens_per_second": 4301233.229398132, "eta_seconds": 3315.6317231361973, "mean_data_wait_seconds": 6.744265556335449e-05}
{"step": 19870, "total_steps": 46238, "tokens": 10743693312, "loss": 3.2785898248354592, "ce_loss": 3.277327779567603, "z_loss": 12.620421366258102, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20877793431282043, "tokens_per_second": 4301384.468472806, "eta_seconds": 3314.2581716396803, "mean_data_wait_seconds": 8.046627044677734e-05}
{"step": 19880, "total_steps": 46238, "tokens": 10749100032, "loss": 3.275875808614673, "ce_loss": 3.274611324974985, "z_loss": 12.64497496142532, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.185916006565094, "tokens_per_second": 4301672.739582273, "eta_seconds": 3312.7791830542265, "mean_data_wait_seconds": 0.00012814998626708984}
{"step": 19890, "total_steps": 46238, "tokens": 10754506752, "loss": 3.3115006161458567, "ce_loss": 3.3102056257652515, "z_loss": 12.95002539952596, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18305276334285736, "tokens_per_second": 4301637.439642703, "eta_seconds": 3311.54947014391, "mean_data_wait_seconds": 6.887316703796387e-05}
{"step": 19900, "total_steps": 46238, "tokens": 10759913472, "loss": 3.270792796756282, "ce_loss": 3.269473585215482, "z_loss": 13.191954041972306, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18663525581359863, "tokens_per_second": 4301938.393425283, "eta_seconds": 3310.060991520175, "mean_data_wait_seconds": 0.00013300776481628418}
{"step": 19910, "total_steps": 46238, "tokens": 10765320192, "loss": 3.241501768430074, "ce_loss": 3.240217443668481, "z_loss": 12.843222683126276, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16786913573741913, "tokens_per_second": 4302010.211281634, "eta_seconds": 3308.74894408012, "mean_data_wait_seconds": 6.809830665588379e-05}
{"step": 19920, "total_steps": 46238, "tokens": 10770726912, "loss": 3.2879918416341147, "ce_loss": 3.2866791847980386, "z_loss": 13.126651055885084, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17664773762226105, "tokens_per_second": 4302164.260616146, "eta_seconds": 3307.3737221651727, "mean_data_wait_seconds": 9.775161743164062e-05}
{"step": 19930, "total_steps": 46238, "tokens": 10776133632, "loss": 3.274403716578628, "ce_loss": 3.2731067339579263, "z_loss": 12.969865733926946, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16491082310676575, "tokens_per_second": 4302373.650999968, "eta_seconds": 3305.9560739672506, "mean_data_wait_seconds": 6.699562072753906e-05}
{"step": 19940, "total_steps": 46238, "tokens": 10781540352, "loss": 3.273970912803303, "ce_loss": 3.2726910945140952, "z_loss": 12.79807110266252, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2101140320301056, "tokens_per_second": 4302500.249890972, "eta_seconds": 3304.6021518209777, "mean_data_wait_seconds": 8.541345596313477e-05}
{"step": 19950, "total_steps": 46238, "tokens": 10786947072, "loss": 3.2363880204431936, "ce_loss": 3.23509076869849, "z_loss": 12.972504521861222, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14914153516292572, "tokens_per_second": 4302390.815325535, "eta_seconds": 3303.42952885014, "mean_data_wait_seconds": 8.45491886138916e-05}
{"step": 19960, "total_steps": 46238, "tokens": 10792353792, "loss": 3.262051127173684, "ce_loss": 3.2607146718285303, "z_loss": 13.364884586045235, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22032126784324646, "tokens_per_second": 4302437.038523592, "eta_seconds": 3302.1373739556925, "mean_data_wait_seconds": 9.515881538391113e-05}
{"step": 19970, "total_steps": 46238, "tokens": 10797760512, "loss": 3.21596558527513, "ce_loss": 3.214713717951919, "z_loss": 12.519154425823327, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17359551787376404, "tokens_per_second": 4302576.521017861, "eta_seconds": 3300.7737002758226, "mean_data_wait_seconds": 0.00014099478721618652}
{"step": 19980, "total_steps": 46238, "tokens": 10803167232, "loss": 3.2917058522051033, "ce_loss": 3.290462166974039, "z_loss": 12.43697672179251, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17609304189682007, "tokens_per_second": 4302747.639237331, "eta_seconds": 3299.3858562702826, "mean_data_wait_seconds": 7.840991020202637e-05}
{"step": 19990, "total_steps": 46238, "tokens": 10808573952, "loss": 3.244349788535725, "ce_loss": 3.2430192792054378, "z_loss": 13.304747884923762, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22000133991241455, "tokens_per_second": 4303043.670523574, "eta_seconds": 3297.9023850513945, "mean_data_wait_seconds": 9.462237358093262e-05}
{"step": 20000, "total_steps": 46238, "tokens": 10813980672, "loss": 3.246922250949975, "ce_loss": 3.2455831585508403, "z_loss": 13.390677921699755, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1863124668598175, "tokens_per_second": 4303263.892624881, "eta_seconds": 3296.4771898632366, "mean_data_wait_seconds": 7.608532905578613e-05}
{"step": 20010, "total_steps": 46238, "tokens": 10819387392, "loss": 3.259407850829038, "ce_loss": 3.258086327350501, "z_loss": 13.21529290892861, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17936843633651733, "tokens_per_second": 4303453.392315516, "eta_seconds": 3295.0756639588467, "mean_data_wait_seconds": 0.00011464953422546387}
{"step": 20020, "total_steps": 46238, "tokens": 10824794112, "loss": 3.2682333516352102, "ce_loss": 3.2669787406921387, "z_loss": 12.546190052321464, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19318494200706482, "tokens_per_second": 4303652.538431528, "eta_seconds": 3293.666878870761, "mean_data_wait_seconds": 7.322430610656738e-05}
{"step": 20030, "total_steps": 46238, "tokens": 10830200832, "loss": 3.2613479404738457, "ce_loss": 3.260023091778611, "z_loss": 13.249091617988817, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20690912008285522, "tokens_per_second": 4303768.1843090365, "eta_seconds": 3292.3220994243384, "mean_data_wait_seconds": 6.437301635742188e-05}
{"step": 20040, "total_steps": 46238, "tokens": 10835607552, "loss": 3.2214310711080376, "ce_loss": 3.2201473062688653, "z_loss": 12.837816469597094, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17985916137695312, "tokens_per_second": 4303944.4350758055, "eta_seconds": 3290.9310511929343, "mean_data_wait_seconds": 8.448958396911621e-05}
{"step": 20050, "total_steps": 46238, "tokens": 10841014272, "loss": 3.2400054100788003, "ce_loss": 3.2387657057155264, "z_loss": 12.397068030906446, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16212651133537292, "tokens_per_second": 4304013.067674255, "eta_seconds": 3289.6223690256643, "mean_data_wait_seconds": 6.344914436340332e-05}
{"step": 20060, "total_steps": 46238, "tokens": 10846420992, "loss": 3.237489225286426, "ce_loss": 3.23620108943997, "z_loss": 12.881030480066935, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18470428884029388, "tokens_per_second": 4304108.938621103, "eta_seconds": 3288.2929186579136, "mean_data_wait_seconds": 0.00014090538024902344}
{"step": 20070, "total_steps": 46238, "tokens": 10851827712, "loss": 3.1299034916993342, "ce_loss": 3.128648309996634, "z_loss": 12.551580457976371, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15565553307533264, "tokens_per_second": 4304099.094579821, "eta_seconds": 3287.044260160359, "mean_data_wait_seconds": 0.00013828277587890625}
{"step": 20080, "total_steps": 46238, "tokens": 10857234432, "loss": 3.2738565116217644, "ce_loss": 3.2726173816305217, "z_loss": 12.390751954280969, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17739717662334442, "tokens_per_second": 4304002.102315401, "eta_seconds": 3285.8621273423432, "mean_data_wait_seconds": 0.0001304149627685547}
{"step": 20090, "total_steps": 46238, "tokens": 10862641152, "loss": 3.15448156450734, "ce_loss": 3.1532390045397207, "z_loss": 12.425564946550312, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21559295058250427, "tokens_per_second": 4304112.316640555, "eta_seconds": 3284.5218116970914, "mean_data_wait_seconds": 0.00015649199485778809}
{"step": 20100, "total_steps": 46238, "tokens": 10868047872, "loss": 3.151442061771046, "ce_loss": 3.1501190897190208, "z_loss": 13.229372956536032, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20975954830646515, "tokens_per_second": 4304157.266581682, "eta_seconds": 3283.23134791567, "mean_data_wait_seconds": 0.00010290741920471191}
{"step": 20110, "total_steps": 46238, "tokens": 10873454592, "loss": 3.18354994058609, "ce_loss": 3.182279189427694, "z_loss": 12.707220084739454, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15900452435016632, "tokens_per_second": 4304214.5790613275, "eta_seconds": 3281.9314847171627, "mean_data_wait_seconds": 8.890032768249512e-05}
{"step": 20120, "total_steps": 46238, "tokens": 10878861312, "loss": 3.1650517041032966, "ce_loss": 3.1637800841620476, "z_loss": 12.716356219667377, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18186643719673157, "tokens_per_second": 4304233.019353572, "eta_seconds": 3280.661284021447, "mean_data_wait_seconds": 9.688735008239746e-05}
{"step": 20130, "total_steps": 46238, "tokens": 10884268032, "loss": 3.1779270244367197, "ce_loss": 3.176686032251878, "z_loss": 12.410018581332583, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18661805987358093, "tokens_per_second": 4304109.364059451, "eta_seconds": 3279.4993598134397, "mean_data_wait_seconds": 9.796023368835449e-05}
{"step": 20140, "total_steps": 46238, "tokens": 10889674752, "loss": 3.1096839399048775, "ce_loss": 3.108415242397424, "z_loss": 12.686737486810395, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19866475462913513, "tokens_per_second": 4304122.816489793, "eta_seconds": 3278.23293748557, "mean_data_wait_seconds": 9.694695472717285e-05}
{"step": 20150, "total_steps": 46238, "tokens": 10895081472, "loss": 3.1551449461416765, "ce_loss": 3.1538668383251536, "z_loss": 12.780600424968835, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17803949117660522, "tokens_per_second": 4304165.339509194, "eta_seconds": 3276.9443902469475, "mean_data_wait_seconds": 8.612871170043945e-05}
{"step": 20160, "total_steps": 46238, "tokens": 10900488192, "loss": 3.185185356573625, "ce_loss": 3.18389648921562, "z_loss": 12.888914332245335, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.214839369058609, "tokens_per_second": 4304279.645525454, "eta_seconds": 3275.6012399558726, "mean_data_wait_seconds": 0.00016492605209350586}
{"step": 20170, "total_steps": 46238, "tokens": 10905894912, "loss": 3.209539866808689, "ce_loss": 3.2082716353011853, "z_loss": 12.682937600395896, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1926773488521576, "tokens_per_second": 4304085.211111986, "eta_seconds": 3274.4930299274465, "mean_data_wait_seconds": 0.00019347667694091797}
{"step": 20180, "total_steps": 46238, "tokens": 10911301632, "loss": 3.2208053451595884, "ce_loss": 3.2195087291977624, "z_loss": 12.965767976009484, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17541682720184326, "tokens_per_second": 4303872.324425791, "eta_seconds": 3273.398753965039, "mean_data_wait_seconds": 0.00012797117233276367}
{"step": 20190, "total_steps": 46238, "tokens": 10916708352, "loss": 3.2289838086475027, "ce_loss": 3.2276569442315535, "z_loss": 13.268628546685884, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.23009595274925232, "tokens_per_second": 4303805.256834819, "eta_seconds": 3272.193499377127, "mean_data_wait_seconds": 0.0001729130744934082}
{"step": 20200, "total_steps": 46238, "tokens": 10922115072, "loss": 3.2113185922304788, "ce_loss": 3.2100309368335838, "z_loss": 12.876515178969413, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20105703175067902, "tokens_per_second": 4303887.274966584, "eta_seconds": 3270.874900902069, "mean_data_wait_seconds": 8.824467658996582e-05}
{"step": 20210, "total_steps": 46238, "tokens": 10927521792, "loss": 3.1931997049938547, "ce_loss": 3.191931430137519, "z_loss": 12.682991244576193, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18974360823631287, "tokens_per_second": 4303981.272895513, "eta_seconds": 3269.54725212663, "mean_data_wait_seconds": 0.0001308917999267578}
{"step": 20220, "total_steps": 46238, "tokens": 10932928512, "loss": 3.174447113817388, "ce_loss": 3.173158858761643, "z_loss": 12.882590452829996, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1855442374944687, "tokens_per_second": 4304025.633582972, "eta_seconds": 3268.2573528935804, "mean_data_wait_seconds": 7.861852645874023e-05}
{"step": 20230, "total_steps": 46238, "tokens": 10938335232, "loss": 3.2435899105939, "ce_loss": 3.242301030592485, "z_loss": 12.88900985139789, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1717498004436493, "tokens_per_second": 4304006.307262974, "eta_seconds": 3267.0158220427675, "mean_data_wait_seconds": 9.331107139587402e-05}
{"step": 20240, "total_steps": 46238, "tokens": 10943741952, "loss": 3.2360349080779334, "ce_loss": 3.2347297993573276, "z_loss": 13.050902084870772, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2316228598356247, "tokens_per_second": 4304213.74846989, "eta_seconds": 3265.602222704839, "mean_data_wait_seconds": 0.00010243058204650879}
{"step": 20250, "total_steps": 46238, "tokens": 10949148672, "loss": 3.2270727518833047, "ce_loss": 3.2257747650146484, "z_loss": 12.979857777104232, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17707832157611847, "tokens_per_second": 4304317.076664393, "eta_seconds": 3264.267713959473, "mean_data_wait_seconds": 9.936094284057617e-05}
{"step": 20260, "total_steps": 46238, "tokens": 10954555392, "loss": 3.254612908218846, "ce_loss": 3.253359429764025, "z_loss": 12.534746300090443, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1528450846672058, "tokens_per_second": 4304525.941013497, "eta_seconds": 3262.853270363404, "mean_data_wait_seconds": 0.0001061558723449707}
{"step": 20270, "total_steps": 46238, "tokens": 10959962112, "loss": 3.294167697429657, "ce_loss": 3.2928779052965567, "z_loss": 12.897799636378434, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21324697136878967, "tokens_per_second": 4304686.779748561, "eta_seconds": 3261.4753505526974, "mean_data_wait_seconds": 7.063150405883789e-05}
{"step": 20280, "total_steps": 46238, "tokens": 10965368832, "loss": 3.20979401920781, "ce_loss": 3.2084946325331023, "z_loss": 12.994366436293632, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1677245795726776, "tokens_per_second": 4304989.254418317, "eta_seconds": 3259.9902751432724, "mean_data_wait_seconds": 7.176399230957031e-05}
{"step": 20290, "total_steps": 46238, "tokens": 10970775552, "loss": 3.238453978841955, "ce_loss": 3.2371938264731206, "z_loss": 12.601842815225774, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19840538501739502, "tokens_per_second": 4304927.172286121, "eta_seconds": 3258.7813504287533, "mean_data_wait_seconds": 8.14199447631836e-05}
{"step": 20300, "total_steps": 46238, "tokens": 10976182272, "loss": 3.24784481344801, "ce_loss": 3.2465392856886894, "z_loss": 13.05556210604581, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19892121851444244, "tokens_per_second": 4305002.393063668, "eta_seconds": 3257.468494464691, "mean_data_wait_seconds": 6.073713302612305e-05}
{"step": 20310, "total_steps": 46238, "tokens": 10981588992, "loss": 3.272357003255324, "ce_loss": 3.2711083798697502, "z_loss": 12.48620962374138, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17388606071472168, "tokens_per_second": 4305186.735526506, "eta_seconds": 3256.073152024533, "mean_data_wait_seconds": 6.446242332458496e-05}
{"step": 20320, "total_steps": 46238, "tokens": 10986995712, "loss": 3.212623493237929, "ce_loss": 3.21125110171058, "z_loss": 13.723775487957578, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20458850264549255, "tokens_per_second": 4305381.457803266, "eta_seconds": 3254.6700823925703, "mean_data_wait_seconds": 7.519125938415527e-05}
{"step": 20330, "total_steps": 46238, "tokens": 10992402432, "loss": 3.2794535051692617, "ce_loss": 3.2781934647849114, "z_loss": 12.600519216421878, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16605184972286224, "tokens_per_second": 4305525.430160497, "eta_seconds": 3253.3054864520573, "mean_data_wait_seconds": 5.987286567687988e-05}
{"step": 20340, "total_steps": 46238, "tokens": 10997809152, "loss": 3.2642193526932686, "ce_loss": 3.262945467775518, "z_loss": 12.738947507106896, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20044605433940887, "tokens_per_second": 4305684.430601087, "eta_seconds": 3251.9296315557676, "mean_data_wait_seconds": 0.00010153651237487793}
{"step": 20350, "total_steps": 46238, "tokens": 11003215872, "loss": 3.235452798279849, "ce_loss": 3.2341602264028606, "z_loss": 12.925773252140392, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16796670854091644, "tokens_per_second": 4305690.178063598, "eta_seconds": 3250.6695756485205, "mean_data_wait_seconds": 6.142258644104004e-05}
{"step": 20360, "total_steps": 46238, "tokens": 11008622592, "loss": 3.28750054583405, "ce_loss": 3.2861915302999094, "z_loss": 13.090075514533304, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17357805371284485, "tokens_per_second": 4305733.472029266, "eta_seconds": 3249.381187871376, "mean_data_wait_seconds": 7.94827938079834e-05}
{"step": 20370, "total_steps": 46238, "tokens": 11014029312, "loss": 3.2801843267498594, "ce_loss": 3.2789371952866064, "z_loss": 12.471121130567608, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19118835031986237, "tokens_per_second": 4305867.232009784, "eta_seconds": 3248.0245837659445, "mean_data_wait_seconds": 9.40859317779541e-05}
{"step": 20380, "total_steps": 46238, "tokens": 11019436032, "loss": 3.2639752951535312, "ce_loss": 3.2626855590126733, "z_loss": 12.897769024877837, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1691085696220398, "tokens_per_second": 4305963.956447795, "eta_seconds": 3246.6959884942767, "mean_data_wait_seconds": 7.846951484680176e-05}
{"step": 20390, "total_steps": 46238, "tokens": 11024842752, "loss": 3.240612880750136, "ce_loss": 3.2392811558463355, "z_loss": 13.316951137600523, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19130820035934448, "tokens_per_second": 4306221.280604649, "eta_seconds": 3245.2464175361106, "mean_data_wait_seconds": 6.070733070373535e-05}
{"step": 20400, "total_steps": 46238, "tokens": 11030249472, "loss": 3.2948720906720017, "ce_loss": 3.2936103506521746, "z_loss": 12.617695317123875, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18437844514846802, "tokens_per_second": 4306335.250314497, "eta_seconds": 3243.905003211211, "mean_data_wait_seconds": 0.00010630488395690918}
{"step": 20410, "total_steps": 46238, "tokens": 11035656192, "loss": 3.2687851750489436, "ce_loss": 3.2674719763524607, "z_loss": 13.132525783596616, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16644495725631714, "tokens_per_second": 4306475.796372702, "eta_seconds": 3242.543649208866, "mean_data_wait_seconds": 0.0013710260391235352}
{"step": 20420, "total_steps": 46238, "tokens": 11041062912, "loss": 3.2654058156591472, "ce_loss": 3.2640994787216187, "z_loss": 13.063519810185287, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16964967548847198, "tokens_per_second": 4306610.458378203, "eta_seconds": 3241.1868124373027, "mean_data_wait_seconds": 9.086728096008301e-05}
{"step": 20430, "total_steps": 46238, "tokens": 11046469632, "loss": 3.2801036383166458, "ce_loss": 3.278837471297293, "z_loss": 12.661766803625857, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18503150343894958, "tokens_per_second": 4306660.957559447, "eta_seconds": 3239.8933748216696, "mean_data_wait_seconds": 7.539987564086914e-05}
{"step": 20440, "total_steps": 46238, "tokens": 11051876352, "loss": 3.2631700996196633, "ce_loss": 3.261869127100164, "z_loss": 13.00939318628022, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2308608740568161, "tokens_per_second": 4306670.602958679, "eta_seconds": 3238.630689428147, "mean_data_wait_seconds": 6.955862045288086e-05}
{"step": 20450, "total_steps": 46238, "tokens": 11057283072, "loss": 3.285218731923537, "ce_loss": 3.283937394618988, "z_loss": 12.813263763080943, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15722313523292542, "tokens_per_second": 4306928.109821036, "eta_seconds": 3237.1817008525218, "mean_data_wait_seconds": 6.935000419616699e-05}
{"step": 20460, "total_steps": 46238, "tokens": 11062689792, "loss": 3.3002327752835825, "ce_loss": 3.298965188589963, "z_loss": 12.67613368323355, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17483775317668915, "tokens_per_second": 4307175.163513303, "eta_seconds": 3235.7407383989143, "mean_data_wait_seconds": 6.034970283508301e-05}
{"step": 20470, "total_steps": 46238, "tokens": 11068096512, "loss": 3.3067200852162912, "ce_loss": 3.3054544239333183, "z_loss": 12.656402696262706, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19518142938613892, "tokens_per_second": 4307293.615443652, "eta_seconds": 3234.396506903803, "mean_data_wait_seconds": 9.623169898986816e-05}
{"step": 20480, "total_steps": 46238, "tokens": 11073503232, "loss": 3.2816240805568118, "ce_loss": 3.280299051241441, "z_loss": 13.25017896565524, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16485048830509186, "tokens_per_second": 4307346.425593702, "eta_seconds": 3233.101619422334, "mean_data_wait_seconds": 0.0001169741153717041}
{"step": 20490, "total_steps": 46238, "tokens": 11078909952, "loss": 3.3006684455004605, "ce_loss": 3.2994543024987886, "z_loss": 12.141471140312426, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21043117344379425, "tokens_per_second": 4307322.195258001, "eta_seconds": 3231.8645675787848, "mean_data_wait_seconds": 0.00015565752983093262}
{"step": 20500, "total_steps": 46238, "tokens": 11084316672, "loss": 3.302869722698674, "ce_loss": 3.3016227248943215, "z_loss": 12.469957423932625, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19635961949825287, "tokens_per_second": 4307547.742263631, "eta_seconds": 3230.440170742594, "mean_data_wait_seconds": 0.0029430389404296875}
{"step": 20510, "total_steps": 46238, "tokens": 11089723392, "loss": 3.2725570292183845, "ce_loss": 3.2713036591356452, "z_loss": 12.533386179895112, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20586560666561127, "tokens_per_second": 4307786.299610566, "eta_seconds": 3229.0061708161993, "mean_data_wait_seconds": 0.00010368227958679199}
{"step": 20520, "total_steps": 46238, "tokens": 11095130112, "loss": 3.264942340778582, "ce_loss": 3.263624476663994, "z_loss": 13.178257234168775, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20132194459438324, "tokens_per_second": 4308000.874429316, "eta_seconds": 3227.590297516347, "mean_data_wait_seconds": 0.0001234114170074463}
{"step": 20530, "total_steps": 46238, "tokens": 11100536832, "loss": 3.2811759854808, "ce_loss": 3.27986992669828, "z_loss": 13.060627063115438, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1494138538837433, "tokens_per_second": 4308001.423992741, "eta_seconds": 3226.3348444110034, "mean_data_wait_seconds": 0.0001704096794128418}
{"step": 20540, "total_steps": 46238, "tokens": 11105943552, "loss": 3.207204600175222, "ce_loss": 3.2059043754230845, "z_loss": 13.00241825797341, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18597160279750824, "tokens_per_second": 4308048.109652488, "eta_seconds": 3225.044853345601, "mean_data_wait_seconds": 7.298588752746582e-05}
{"step": 20550, "total_steps": 46238, "tokens": 11111350272, "loss": 3.122347197749398, "ce_loss": 3.121076787963058, "z_loss": 12.703833616141116, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14034709334373474, "tokens_per_second": 4308180.60376077, "eta_seconds": 3223.6906809051698, "mean_data_wait_seconds": 9.02712345123291e-05}
{"step": 20560, "total_steps": 46238, "tokens": 11116756992, "loss": 3.2449316057291897, "ce_loss": 3.2436378905267427, "z_loss": 12.937475305615049, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1794792115688324, "tokens_per_second": 4308198.881649846, "eta_seconds": 3222.4220202860038, "mean_data_wait_seconds": 0.00011515617370605469}
{"step": 20570, "total_steps": 46238, "tokens": 11122163712, "loss": 3.222293870015578, "ce_loss": 3.2209656021811743, "z_loss": 13.28265202406681, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1794506311416626, "tokens_per_second": 4308255.662476293, "eta_seconds": 3221.1245829416616, "mean_data_wait_seconds": 9.813904762268066e-05}
{"step": 20580, "total_steps": 46238, "tokens": 11127570432, "loss": 3.1816284186912305, "ce_loss": 3.180330345124909, "z_loss": 12.980385361295758, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1642259806394577, "tokens_per_second": 4308284.3226178, "eta_seconds": 3219.8481959916426, "mean_data_wait_seconds": 0.00013330578804016113}
{"step": 20590, "total_steps": 46238, "tokens": 11132977152, "loss": 3.175772609132709, "ce_loss": 3.174473173690565, "z_loss": 12.99414680220864, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1933799833059311, "tokens_per_second": 4308387.9476313945, "eta_seconds": 3218.515823679108, "mean_data_wait_seconds": 7.346272468566895e-05}
{"step": 20600, "total_steps": 46238, "tokens": 11138383872, "loss": 3.279857371792649, "ce_loss": 3.2786242203278975, "z_loss": 12.331663760271939, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2176513522863388, "tokens_per_second": 4308513.254052941, "eta_seconds": 3217.1673258660658, "mean_data_wait_seconds": 7.227063179016113e-05}
{"step": 20610, "total_steps": 46238, "tokens": 11143790592, "loss": 3.252740847341942, "ce_loss": 3.251460026610981, "z_loss": 12.808667211821586, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.178509920835495, "tokens_per_second": 4308476.767839081, "eta_seconds": 3215.939667454534, "mean_data_wait_seconds": 6.455183029174805e-05}
{"step": 20620, "total_steps": 46238, "tokens": 11149197312, "loss": 3.202168993877642, "ce_loss": 3.2008284258119986, "z_loss": 13.40572654839718, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1729438453912735, "tokens_per_second": 4308529.090354976, "eta_seconds": 3214.645725615578, "mean_data_wait_seconds": 7.387995719909668e-05}
{"step": 20630, "total_steps": 46238, "tokens": 11154604032, "loss": 3.1706732457334343, "ce_loss": 3.1693602041764692, "z_loss": 13.130523761113485, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18805186450481415, "tokens_per_second": 4308674.026459712, "eta_seconds": 3213.282745219867, "mean_data_wait_seconds": 0.00011390447616577148}
{"step": 20640, "total_steps": 46238, "tokens": 11160010752, "loss": 3.2047641927545722, "ce_loss": 3.2034485737482705, "z_loss": 13.156337550192168, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2033519148826599, "tokens_per_second": 4308641.881641729, "eta_seconds": 3212.0518632490016, "mean_data_wait_seconds": 9.256601333618164e-05}
{"step": 20650, "total_steps": 46238, "tokens": 11165417472, "loss": 3.1621474099881723, "ce_loss": 3.160858490250327, "z_loss": 12.8892840905623, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1832709014415741, "tokens_per_second": 4308564.0166130215, "eta_seconds": 3210.855034451849, "mean_data_wait_seconds": 9.244680404663086e-05}
{"step": 20660, "total_steps": 46238, "tokens": 11170824192, "loss": 3.17666754397479, "ce_loss": 3.175395002870849, "z_loss": 12.725601492506085, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17632007598876953, "tokens_per_second": 4308582.481303179, "eta_seconds": 3209.5864020264353, "mean_data_wait_seconds": 0.00011157989501953125}
{"step": 20670, "total_steps": 46238, "tokens": 11176230912, "loss": 3.1671349279808276, "ce_loss": 3.165914349483721, "z_loss": 12.206158695798932, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1886621117591858, "tokens_per_second": 4308645.240623077, "eta_seconds": 3208.2847976597373, "mean_data_wait_seconds": 0.00010403990745544434}
{"step": 20680, "total_steps": 46238, "tokens": 11181637632, "loss": 3.1970580491152676, "ce_loss": 3.195782704786821, "z_loss": 12.753007744297836, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19098104536533356, "tokens_per_second": 4308655.599481364, "eta_seconds": 3207.022233492804, "mean_data_wait_seconds": 6.940960884094238e-05}
{"step": 20690, "total_steps": 46238, "tokens": 11187044352, "loss": 3.241735583001917, "ce_loss": 3.2404723167419434, "z_loss": 12.632787422700362, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.174884632229805, "tokens_per_second": 4308845.814488501, "eta_seconds": 3205.625863324068, "mean_data_wait_seconds": 8.803606033325195e-05}
{"step": 20700, "total_steps": 46238, "tokens": 11192451072, "loss": 3.2322666681174077, "ce_loss": 3.2309610536604216, "z_loss": 13.056220307494655, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19965775310993195, "tokens_per_second": 4308871.914475979, "eta_seconds": 3204.351658171567, "mean_data_wait_seconds": 0.00010409951210021973}
{"step": 20710, "total_steps": 46238, "tokens": 11197857792, "loss": 3.16524799303575, "ce_loss": 3.1638932246150393, "z_loss": 13.54760192379807, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20234717428684235, "tokens_per_second": 4308888.090250125, "eta_seconds": 3203.084845770229, "mean_data_wait_seconds": 7.024407386779785e-05}
{"step": 20720, "total_steps": 46238, "tokens": 11203264512, "loss": 3.1924675811420786, "ce_loss": 3.191187806201704, "z_loss": 12.797640121344363, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19432732462882996, "tokens_per_second": 4308963.816906041, "eta_seconds": 3201.773793010441, "mean_data_wait_seconds": 0.00012123584747314453}
{"step": 20730, "total_steps": 46238, "tokens": 11208671232, "loss": 3.1483398657856565, "ce_loss": 3.14705839843461, "z_loss": 12.814958088325731, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16166995465755463, "tokens_per_second": 4309077.253248673, "eta_seconds": 3200.4347783746125, "mean_data_wait_seconds": 8.174777030944824e-05}
{"step": 20740, "total_steps": 46238, "tokens": 11214077952, "loss": 3.215264508218476, "ce_loss": 3.2139855442625103, "z_loss": 12.789425084085176, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18560990691184998, "tokens_per_second": 4309258.5782579295, "eta_seconds": 3199.045435229594, "mean_data_wait_seconds": 0.0001105964183807373}
{"step": 20750, "total_steps": 46238, "tokens": 11219484672, "loss": 3.287166158358256, "ce_loss": 3.2858857945962385, "z_loss": 12.80375533392935, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21784251928329468, "tokens_per_second": 4309354.399072283, "eta_seconds": 3197.7196554004886, "mean_data_wait_seconds": 6.893277168273926e-05}
{"step": 20760, "total_steps": 46238, "tokens": 11224891392, "loss": 3.1761312394431145, "ce_loss": 3.1748177969094478, "z_loss": 13.133986234664917, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17747598886489868, "tokens_per_second": 4309373.880743398, "eta_seconds": 3196.4505575978856, "mean_data_wait_seconds": 7.316470146179199e-05}
{"step": 20770, "total_steps": 46238, "tokens": 11230298112, "loss": 3.256673433563926, "ce_loss": 3.2553739222613247, "z_loss": 12.99496899951588, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2046739161014557, "tokens_per_second": 4309539.63881496, "eta_seconds": 3195.073018933013, "mean_data_wait_seconds": 7.036328315734863e-05}
{"step": 20780, "total_steps": 46238, "tokens": 11235704832, "loss": 3.290362657922687, "ce_loss": 3.28908458442399, "z_loss": 12.78085934754574, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17443493008613586, "tokens_per_second": 4309703.366171038, "eta_seconds": 3193.6970910897157, "mean_data_wait_seconds": 6.935000419616699e-05}
{"step": 20790, "total_steps": 46238, "tokens": 11241111552, "loss": 3.2358081846526177, "ce_loss": 3.2345466071909126, "z_loss": 12.61613070603573, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15836282074451447, "tokens_per_second": 4309932.762238048, "eta_seconds": 3192.2726276721637, "mean_data_wait_seconds": 6.160140037536621e-05}
{"step": 20800, "total_steps": 46238, "tokens": 11246518272, "loss": 3.2410458109595557, "ce_loss": 3.2397415619907957, "z_loss": 13.042995236136697, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2058483213186264, "tokens_per_second": 4309983.241844198, "eta_seconds": 3190.980774699068, "mean_data_wait_seconds": 7.209181785583496e-05}
{"step": 20810, "total_steps": 46238, "tokens": 11251924992, "loss": 3.2860147645979216, "ce_loss": 3.28473327737866, "z_loss": 12.814759348378036, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18640761077404022, "tokens_per_second": 4310234.142297494, "eta_seconds": 3189.540635180447, "mean_data_wait_seconds": 9.40561294555664e-05}
{"step": 20820, "total_steps": 46238, "tokens": 11257331712, "loss": 3.1826560804338166, "ce_loss": 3.1814067417925056, "z_loss": 12.493637749643037, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18545246124267578, "tokens_per_second": 4310394.343780177, "eta_seconds": 3188.167747071643, "mean_data_wait_seconds": 0.00011745095252990723}
{"step": 20830, "total_steps": 46238, "tokens": 11262738432, "loss": 3.1921352545420327, "ce_loss": 3.190902413743915, "z_loss": 12.328357978300614, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16451209783554077, "tokens_per_second": 4309290.138405687, "eta_seconds": 3187.7300118581106, "mean_data_wait_seconds": 9.754300117492676e-05}
{"step": 20840, "total_steps": 46238, "tokens": 11268145152, "loss": 3.220284384308439, "ce_loss": 3.2189984773144578, "z_loss": 12.859478560360996, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22405405342578888, "tokens_per_second": 4309508.672099365, "eta_seconds": 3186.3137607542544, "mean_data_wait_seconds": 0.00011613965034484863}
{"step": 20850, "total_steps": 46238, "tokens": 11273551872, "loss": 3.2335713054194595, "ce_loss": 3.232294635339217, "z_loss": 12.7661255272952, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16554418206214905, "tokens_per_second": 4308842.18027347, "eta_seconds": 3185.551823373779, "mean_data_wait_seconds": 0.00011271238327026367}
{"step": 20860, "total_steps": 46238, "tokens": 11278958592, "loss": 3.2423549764084094, "ce_loss": 3.241045411789056, "z_loss": 13.09516395222057, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20666810870170593, "tokens_per_second": 4308858.851092277, "eta_seconds": 3184.2847069640907, "mean_data_wait_seconds": 0.00018611550331115723}
{"step": 20870, "total_steps": 46238, "tokens": 11284365312, "loss": 3.24729596304171, "ce_loss": 3.246025452108094, "z_loss": 12.705553589445172, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18623924255371094, "tokens_per_second": 4308807.797291281, "eta_seconds": 3183.0676301277667, "mean_data_wait_seconds": 0.00011280179023742676}
{"step": 20880, "total_steps": 46238, "tokens": 11289772032, "loss": 3.2767280773683027, "ce_loss": 3.2754638465968045, "z_loss": 12.642381689765237, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21877917647361755, "tokens_per_second": 4307717.387618998, "eta_seconds": 3182.618233824717, "mean_data_wait_seconds": 0.10471302270889282}
{"step": 20890, "total_steps": 46238, "tokens": 11295178752, "loss": 3.2891870682889763, "ce_loss": 3.2879346121441233, "z_loss": 12.524454730929751, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20578226447105408, "tokens_per_second": 4307918.87722051, "eta_seconds": 3181.214311269054, "mean_data_wait_seconds": 0.00013631582260131836}
{"step": 20900, "total_steps": 46238, "tokens": 11300585472, "loss": 3.149128440654639, "ce_loss": 3.147856268015775, "z_loss": 12.721691731250647, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19217278063297272, "tokens_per_second": 4308001.110393973, "eta_seconds": 3179.89854527851, "mean_data_wait_seconds": 0.00013780593872070312}
{"step": 20910, "total_steps": 46238, "tokens": 11305992192, "loss": 3.269336046594562, "ce_loss": 3.2680576046307883, "z_loss": 12.78466771588181, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18097791075706482, "tokens_per_second": 4308205.23478495, "eta_seconds": 3178.4928984896737, "mean_data_wait_seconds": 0.00011938810348510742}
{"step": 20920, "total_steps": 46238, "tokens": 11311398912, "loss": 3.2651168505350747, "ce_loss": 3.263844022245118, "z_loss": 12.728632045514656, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.176130473613739, "tokens_per_second": 4307570.222176406, "eta_seconds": 3177.706297979751, "mean_data_wait_seconds": 0.00012874603271484375}
{"step": 20930, "total_steps": 46238, "tokens": 11316805632, "loss": 3.250096109780398, "ce_loss": 3.2488354101325525, "z_loss": 12.6068844722979, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1747722029685974, "tokens_per_second": 4307512.376305551, "eta_seconds": 3176.49378775213, "mean_data_wait_seconds": 0.00012859702110290527}
{"step": 20940, "total_steps": 46238, "tokens": 11322212352, "loss": 3.2221921158559397, "ce_loss": 3.220878492702137, "z_loss": 13.136684836763324, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17972634732723236, "tokens_per_second": 4307573.080327898, "eta_seconds": 3175.1938571774294, "mean_data_wait_seconds": 0.00014647841453552246}
{"step": 20950, "total_steps": 46238, "tokens": 11327619072, "loss": 3.287366619615844, "ce_loss": 3.2860869548537512, "z_loss": 12.796765471949723, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16340506076812744, "tokens_per_second": 4307605.9971356215, "eta_seconds": 3173.9144371818807, "mean_data_wait_seconds": 0.00013592839241027832}
{"step": 20960, "total_steps": 46238, "tokens": 11333025792, "loss": 3.262197230801438, "ce_loss": 3.260886519244223, "z_loss": 13.107190717350353, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19056294858455658, "tokens_per_second": 4307653.682281927, "eta_seconds": 3172.624159693428, "mean_data_wait_seconds": 8.735060691833496e-05}
{"step": 20970, "total_steps": 46238, "tokens": 11338432512, "loss": 3.288025865049073, "ce_loss": 3.28674092798522, "z_loss": 12.848910057183469, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16626188158988953, "tokens_per_second": 4307737.716900895, "eta_seconds": 3171.3071504799536, "mean_data_wait_seconds": 9.050965309143066e-05}
{"step": 20980, "total_steps": 46238, "tokens": 11343839232, "loss": 3.266311231887702, "ce_loss": 3.264989045533267, "z_loss": 13.221846089218602, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18043950200080872, "tokens_per_second": 4307868.998830949, "eta_seconds": 3169.95542522436, "mean_data_wait_seconds": 9.363889694213867e-05}
{"step": 20990, "total_steps": 46238, "tokens": 11349245952, "loss": 3.2586338736794214, "ce_loss": 3.2573673779314216, "z_loss": 12.664873571106881, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2115209847688675, "tokens_per_second": 4307929.299347229, "eta_seconds": 3168.6559911901077, "mean_data_wait_seconds": 6.899237632751465e-05}
{"step": 21000, "total_steps": 46238, "tokens": 11354652672, "loss": 3.2582774848648994, "ce_loss": 3.2570228377978006, "z_loss": 12.546896378199259, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18531037867069244, "tokens_per_second": 4308027.0642592795, "eta_seconds": 3167.329048882404, "mean_data_wait_seconds": 0.0001271963119506836}
{"step": 21010, "total_steps": 46238, "tokens": 11360059392, "loss": 3.2176020271850354, "ce_loss": 3.2163126920208787, "z_loss": 12.89331874702916, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1620412915945053, "tokens_per_second": 4308131.741497221, "eta_seconds": 3165.9970870017546, "mean_data_wait_seconds": 0.0001431107521057129}
{"step": 21020, "total_steps": 46238, "tokens": 11365466112, "loss": 3.2952665289243064, "ce_loss": 3.294004825028506, "z_loss": 12.61752773776199, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1912335604429245, "tokens_per_second": 4308136.357456973, "eta_seconds": 3164.738692729776, "mean_data_wait_seconds": 7.37905502319336e-05}
{"step": 21030, "total_steps": 46238, "tokens": 11370872832, "loss": 3.1670840816064314, "ce_loss": 3.1658317735700896, "z_loss": 12.523272189227017, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22323131561279297, "tokens_per_second": 4308096.919511495, "eta_seconds": 3163.5126503944557, "mean_data_wait_seconds": 9.223818778991699e-05}
{"step": 21040, "total_steps": 46238, "tokens": 11376279552, "loss": 3.2743117050691084, "ce_loss": 3.2730127464641225, "z_loss": 12.98912874135104, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17727801203727722, "tokens_per_second": 4308114.347302625, "eta_seconds": 3162.244844436351, "mean_data_wait_seconds": 8.308887481689453e-05}
{"step": 21050, "total_steps": 46238, "tokens": 11381686272, "loss": 3.302171770370368, "ce_loss": 3.300914603652376, "z_loss": 12.572002894950636, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19465135037899017, "tokens_per_second": 4308230.130187678, "eta_seconds": 3160.9048849502306, "mean_data_wait_seconds": 7.531046867370605e-05}
{"step": 21060, "total_steps": 46238, "tokens": 11387092992, "loss": 3.2623681407986265, "ce_loss": 3.261093784462322, "z_loss": 12.743962880336877, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18118180334568024, "tokens_per_second": 4308391.519027136, "eta_seconds": 3159.53155229351, "mean_data_wait_seconds": 6.4849853515625e-05}
{"step": 21070, "total_steps": 46238, "tokens": 11392499712, "loss": 3.2872007669824543, "ce_loss": 3.2859235098867705, "z_loss": 12.772600116151752, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1787765622138977, "tokens_per_second": 4308293.504817057, "eta_seconds": 3158.348475744759, "mean_data_wait_seconds": 9.28044319152832e-05}
{"step": 21080, "total_steps": 46238, "tokens": 11397906432, "loss": 3.2648879289627075, "ce_loss": 3.2636375878796433, "z_loss": 12.503675626985954, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17386750876903534, "tokens_per_second": 4308276.987860694, "eta_seconds": 3157.105623042593, "mean_data_wait_seconds": 7.659196853637695e-05}
{"step": 21090, "total_steps": 46238, "tokens": 11403313152, "loss": 3.203163607554002, "ce_loss": 3.2018433610598245, "z_loss": 13.202187133557869, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18548031151294708, "tokens_per_second": 4308406.300972632, "eta_seconds": 3155.755941803959, "mean_data_wait_seconds": 7.706880569458008e-05}
{"step": 21100, "total_steps": 46238, "tokens": 11408719872, "loss": 3.149576145591158, "ce_loss": 3.1482934572479944, "z_loss": 12.826966740868308, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1548643261194229, "tokens_per_second": 4308506.943717965, "eta_seconds": 3154.4273321448914, "mean_data_wait_seconds": 9.202957153320312e-05}
{"step": 21110, "total_steps": 46238, "tokens": 11414126592, "loss": 3.2188036983663384, "ce_loss": 3.2175020557461362, "z_loss": 13.016652836944118, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19297930598258972, "tokens_per_second": 4308681.4016656615, "eta_seconds": 3153.0447664912276, "mean_data_wait_seconds": 6.812810897827148e-05}
{"step": 21120, "total_steps": 46238, "tokens": 11419533312, "loss": 3.2378997080253833, "ce_loss": 3.2366559162284387, "z_loss": 12.437530423655655, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18173541128635406, "tokens_per_second": 4308949.843434994, "eta_seconds": 3151.5935709231403, "mean_data_wait_seconds": 9.760260581970215e-05}
{"step": 21130, "total_steps": 46238, "tokens": 11424940032, "loss": 3.1970482035116716, "ce_loss": 3.195812281334039, "z_loss": 12.359133395281704, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18946421146392822, "tokens_per_second": 4309048.28364088, "eta_seconds": 3150.2668363071243, "mean_data_wait_seconds": 6.946921348571777e-05}
{"step": 21140, "total_steps": 46238, "tokens": 11430346752, "loss": 3.12280028516596, "ce_loss": 3.1214500322486414, "z_loss": 13.502484646710483, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16470247507095337, "tokens_per_second": 4309248.126272805, "eta_seconds": 3148.8660634950343, "mean_data_wait_seconds": 6.416440010070801e-05}
{"step": 21150, "total_steps": 46238, "tokens": 11435753472, "loss": 3.2009796247337805, "ce_loss": 3.199695753328728, "z_loss": 12.838237755226366, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18286766111850739, "tokens_per_second": 4309319.406683451, "eta_seconds": 3147.5593206118447, "mean_data_wait_seconds": 0.00011396408081054688}
{"step": 21160, "total_steps": 46238, "tokens": 11441160192, "loss": 3.1590318733995613, "ce_loss": 3.1577559980479153, "z_loss": 12.758929360996593, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22136206924915314, "tokens_per_second": 4309491.850674088, "eta_seconds": 3146.1787639485146, "mean_data_wait_seconds": 7.534027099609375e-05}
{"step": 21170, "total_steps": 46238, "tokens": 11446566912, "loss": 3.177073140939077, "ce_loss": 3.175801846114072, "z_loss": 12.712968038790153, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18376123905181885, "tokens_per_second": 4309631.67175608, "eta_seconds": 3144.8221231577872, "mean_data_wait_seconds": 5.97536563873291e-05}
{"step": 21180, "total_steps": 46238, "tokens": 11451973632, "loss": 3.1873777400363577, "ce_loss": 3.1860484769850066, "z_loss": 13.292489246888595, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18941082060337067, "tokens_per_second": 4309703.439512898, "eta_seconds": 3143.515207981738, "mean_data_wait_seconds": 6.160140037536621e-05}
{"step": 21190, "total_steps": 46238, "tokens": 11457380352, "loss": 3.1993120150132612, "ce_loss": 3.1980047713626516, "z_loss": 13.072095408584133, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17590710520744324, "tokens_per_second": 4309861.261808606, "eta_seconds": 3142.1455961942256, "mean_data_wait_seconds": 5.84721565246582e-05}
{"step": 21200, "total_steps": 46238, "tokens": 11462787072, "loss": 3.180382304119341, "ce_loss": 3.179099014311126, "z_loss": 12.832736701676339, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1598748415708542, "tokens_per_second": 4310162.338901517, "eta_seconds": 3140.6716962428786, "mean_data_wait_seconds": 6.321072578430176e-05}
{"step": 21210, "total_steps": 46238, "tokens": 11468193792, "loss": 3.2468307198900166, "ce_loss": 3.2455635793281323, "z_loss": 12.671217448783644, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1741604208946228, "tokens_per_second": 4310349.986716969, "eta_seconds": 3139.2806119454713, "mean_data_wait_seconds": 7.784366607666016e-05}
{"step": 21220, "total_steps": 46238, "tokens": 11473600512, "loss": 3.237869027889136, "ce_loss": 3.2365685791680305, "z_loss": 13.004786050680911, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20089547336101532, "tokens_per_second": 4310497.56328231, "eta_seconds": 3137.918819214081, "mean_data_wait_seconds": 8.38935375213623e-05}
{"step": 21230, "total_steps": 46238, "tokens": 11479007232, "loss": 3.2656579487251514, "ce_loss": 3.264401323867567, "z_loss": 12.566087758902347, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18869106471538544, "tokens_per_second": 4310601.767884814, "eta_seconds": 3136.588678808636, "mean_data_wait_seconds": 8.672475814819336e-05}
{"step": 21240, "total_steps": 46238, "tokens": 11484413952, "loss": 3.1717591845627986, "ce_loss": 3.170463316368334, "z_loss": 12.958544174830118, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2138138860464096, "tokens_per_second": 4310900.675638013, "eta_seconds": 3135.1169977953055, "mean_data_wait_seconds": 9.810924530029297e-05}
{"step": 21250, "total_steps": 46238, "tokens": 11489820672, "loss": 3.2667347662376636, "ce_loss": 3.26545818827369, "z_loss": 12.766066551208496, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1876033991575241, "tokens_per_second": 4310991.042782041, "eta_seconds": 3133.7971083515977, "mean_data_wait_seconds": 7.620453834533691e-05}
{"step": 21260, "total_steps": 46238, "tokens": 11495227392, "loss": 3.1768512310403767, "ce_loss": 3.1755662217284693, "z_loss": 12.850400411721433, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16135917603969574, "tokens_per_second": 4311341.60867285, "eta_seconds": 3132.2882224953214, "mean_data_wait_seconds": 8.246302604675293e-05}
{"step": 21270, "total_steps": 46238, "tokens": 11500634112, "loss": 3.2194997361212065, "ce_loss": 3.2181903304475727, "z_loss": 13.09449153957945, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15117698907852173, "tokens_per_second": 4311521.805725752, "eta_seconds": 3130.903294069677, "mean_data_wait_seconds": 9.73045825958252e-05}
{"step": 21280, "total_steps": 46238, "tokens": 11506040832, "loss": 3.2990147868792215, "ce_loss": 3.2977013009967227, "z_loss": 13.134584202910915, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18759454786777496, "tokens_per_second": 4310459.89961811, "eta_seconds": 3130.4202842011073, "mean_data_wait_seconds": 9.173154830932617e-05}
{"step": 21290, "total_steps": 46238, "tokens": 11511447552, "loss": 3.307558086785403, "ce_loss": 3.3062469652204802, "z_loss": 13.111251946651574, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19031915068626404, "tokens_per_second": 4309897.379882946, "eta_seconds": 3129.574371528616, "mean_data_wait_seconds": 0.00010564923286437988}
{"step": 21300, "total_steps": 46238, "tokens": 11516854272, "loss": 3.1866492159438855, "ce_loss": 3.1854056390849026, "z_loss": 12.436190251148108, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1986781507730484, "tokens_per_second": 4309142.405129855, "eta_seconds": 3128.867973346474, "mean_data_wait_seconds": 0.0001176595687866211}
{"step": 21310, "total_steps": 46238, "tokens": 11522260992, "loss": 3.2525252772100046, "ce_loss": 3.251268668608232, "z_loss": 12.565857973965732, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19466976821422577, "tokens_per_second": 4309334.602447701, "eta_seconds": 3127.4737720168864, "mean_data_wait_seconds": 9.813904762268066e-05}
{"step": 21320, "total_steps": 46238, "tokens": 11527667712, "loss": 3.1812674312880547, "ce_loss": 3.1800045822605942, "z_loss": 12.62910869627288, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17880277335643768, "tokens_per_second": 4309267.647031975, "eta_seconds": 3126.267692673682, "mean_data_wait_seconds": 0.00014773011207580566}
{"step": 21330, "total_steps": 46238, "tokens": 11533074432, "loss": 3.2286004586653276, "ce_loss": 3.2273172573609785, "z_loss": 12.832088838924061, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16163039207458496, "tokens_per_second": 4309303.028633228, "eta_seconds": 3124.9873621143665, "mean_data_wait_seconds": 7.617473602294922e-05}
{"step": 21340, "total_steps": 46238, "tokens": 11538481152, "loss": 3.2524401184284324, "ce_loss": 3.2511387196454136, "z_loss": 13.014191078417229, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17783135175704956, "tokens_per_second": 4308576.024527685, "eta_seconds": 3124.259780347182, "mean_data_wait_seconds": 6.914138793945312e-05}
{"step": 21350, "total_steps": 46238, "tokens": 11543887872, "loss": 3.2201172041170523, "ce_loss": 3.218838348533168, "z_loss": 12.788497216773756, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17159403860569, "tokens_per_second": 4308548.308122262, "eta_seconds": 3123.0249962925964, "mean_data_wait_seconds": 8.109211921691895e-05}
{"step": 21360, "total_steps": 46238, "tokens": 11549294592, "loss": 3.266053714535453, "ce_loss": 3.264780889857899, "z_loss": 12.728298678542629, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17286860942840576, "tokens_per_second": 4308532.647133937, "eta_seconds": 3121.781461479054, "mean_data_wait_seconds": 8.094310760498047e-05}
{"step": 21370, "total_steps": 46238, "tokens": 11554701312, "loss": 3.2209091276833504, "ce_loss": 3.219629573099541, "z_loss": 12.795854785225607, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18325205147266388, "tokens_per_second": 4308448.753816774, "eta_seconds": 3120.5873371684934, "mean_data_wait_seconds": 9.098649024963379e-05}
{"step": 21380, "total_steps": 46238, "tokens": 11560108032, "loss": 3.218921067136707, "ce_loss": 3.2176488348931978, "z_loss": 12.722391533129143, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19666539132595062, "tokens_per_second": 4308408.103028387, "eta_seconds": 3119.361857701773, "mean_data_wait_seconds": 9.006261825561523e-05}
{"step": 21390, "total_steps": 46238, "tokens": 11565514752, "loss": 3.187493445295276, "ce_loss": 3.1862461295994846, "z_loss": 12.473173322099628, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17489373683929443, "tokens_per_second": 4308314.893486088, "eta_seconds": 3118.174394427741, "mean_data_wait_seconds": 7.218122482299805e-05}
{"step": 21400, "total_steps": 46238, "tokens": 11570921472, "loss": 3.2352479443405615, "ce_loss": 3.2339684945164304, "z_loss": 12.794560461333305, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18036223948001862, "tokens_per_second": 4308286.0075654, "eta_seconds": 3116.940342497945, "mean_data_wait_seconds": 0.00011330842971801758}
{"step": 21410, "total_steps": 46238, "tokens": 11576328192, "loss": 3.2890035293319007, "ce_loss": 3.287722515337395, "z_loss": 12.809996778314764, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16790466010570526, "tokens_per_second": 4308150.294787047, "eta_seconds": 3115.7835324924554, "mean_data_wait_seconds": 0.00010734796524047852}
{"step": 21420, "total_steps": 46238, "tokens": 11581734912, "loss": 3.2992396318551265, "ce_loss": 3.297966075666023, "z_loss": 12.735104582526468, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20331284403800964, "tokens_per_second": 4308169.82543173, "eta_seconds": 3114.5144150986134, "mean_data_wait_seconds": 6.794929504394531e-05}
{"step": 21430, "total_steps": 46238, "tokens": 11587141632, "loss": 3.2550380609252234, "ce_loss": 3.253720532764088, "z_loss": 13.175200426217282, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1697601079940796, "tokens_per_second": 4308260.624883178, "eta_seconds": 3113.193808780705, "mean_data_wait_seconds": 8.195638656616211e-05}
{"step": 21440, "total_steps": 46238, "tokens": 11592548352, "loss": 3.258301642808047, "ce_loss": 3.2570360519669275, "z_loss": 12.656365770282168, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1997719556093216, "tokens_per_second": 4308407.51277848, "eta_seconds": 3111.832746608929, "mean_data_wait_seconds": 7.453560829162598e-05}
{"step": 21450, "total_steps": 46238, "tokens": 11597955072, "loss": 3.2819243705633916, "ce_loss": 3.280677697875283, "z_loss": 12.466600743207065, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21014615893363953, "tokens_per_second": 4308560.487899941, "eta_seconds": 3110.467382699358, "mean_data_wait_seconds": 6.410479545593262e-05}
{"step": 21460, "total_steps": 46238, "tokens": 11603361792, "loss": 3.2807205644520847, "ce_loss": 3.2794517892779727, "z_loss": 12.687890175617103, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18425317108631134, "tokens_per_second": 4308599.950768914, "eta_seconds": 3109.1840266138665, "mean_data_wait_seconds": 9.28044319152832e-05}
{"step": 21470, "total_steps": 46238, "tokens": 11608768512, "loss": 3.3292030583728445, "ce_loss": 3.3278828714833115, "z_loss": 13.201667843442975, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18445353209972382, "tokens_per_second": 4308622.7019975735, "eta_seconds": 3107.9127484965707, "mean_data_wait_seconds": 6.255507469177246e-05}
{"step": 21480, "total_steps": 46238, "tokens": 11614175232, "loss": 3.265548106395837, "ce_loss": 3.264269935362267, "z_loss": 12.781629779122092, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1839311569929123, "tokens_per_second": 4308693.355129851, "eta_seconds": 3106.6069457144285, "mean_data_wait_seconds": 7.548928260803223e-05}
{"step": 21490, "total_steps": 46238, "tokens": 11619581952, "loss": 3.2789383801546963, "ce_loss": 3.2777165228670295, "z_loss": 12.218710986050693, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22859029471874237, "tokens_per_second": 4308741.335093873, "eta_seconds": 3105.317526262805, "mean_data_wait_seconds": 7.545948028564453e-05}
{"step": 21500, "total_steps": 46238, "tokens": 11624988672, "loss": 3.268060247103373, "ce_loss": 3.266807662718224, "z_loss": 12.526244019017074, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.15625108778476715, "tokens_per_second": 4308683.992671784, "eta_seconds": 3104.1040110501363, "mean_data_wait_seconds": 9.390711784362793e-05}
{"step": 21510, "total_steps": 46238, "tokens": 11630395392, "loss": 3.2366219289375073, "ce_loss": 3.2353363560907766, "z_loss": 12.855775825905077, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18781639635562897, "tokens_per_second": 4308822.496599147, "eta_seconds": 3102.749429699644, "mean_data_wait_seconds": 6.312131881713867e-05}
{"step": 21520, "total_steps": 46238, "tokens": 11635802112, "loss": 3.2191975965644373, "ce_loss": 3.2178827578371223, "z_loss": 13.148242733695291, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1517609804868698, "tokens_per_second": 4308937.043838865, "eta_seconds": 3101.412178464807, "mean_data_wait_seconds": 7.492303848266602e-05}
{"step": 21530, "total_steps": 46238, "tokens": 11641208832, "loss": 3.272591645067388, "ce_loss": 3.2712819359519263, "z_loss": 13.097227450573083, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.23753798007965088, "tokens_per_second": 4309103.556510362, "eta_seconds": 3100.037613117381, "mean_data_wait_seconds": 6.657838821411133e-05}
{"step": 21540, "total_steps": 46238, "tokens": 11646615552, "loss": 3.3092381195588545, "ce_loss": 3.307945414022966, "z_loss": 12.927052252220385, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17741484940052032, "tokens_per_second": 4309202.640014595, "eta_seconds": 3098.7116409904484, "mean_data_wait_seconds": 8.538365364074707e-05}
{"step": 21550, "total_steps": 46238, "tokens": 11652022272, "loss": 3.23182307770758, "ce_loss": 3.230516466227445, "z_loss": 13.066130356355147, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19942927360534668, "tokens_per_second": 4309229.325455011, "eta_seconds": 3097.4377680841185, "mean_data_wait_seconds": 8.541345596313477e-05}
{"step": 21560, "total_steps": 46238, "tokens": 11657428992, "loss": 3.3276623090108237, "ce_loss": 3.326328535874685, "z_loss": 13.337648478421299, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2063422054052353, "tokens_per_second": 4309321.076381943, "eta_seconds": 3096.117162660325, "mean_data_wait_seconds": 7.349252700805664e-05}
{"step": 21570, "total_steps": 46238, "tokens": 11662835712, "loss": 3.256934982357603, "ce_loss": 3.255630460652438, "z_loss": 13.045449871005435, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16954125463962555, "tokens_per_second": 4309432.962073938, "eta_seconds": 3094.782153794455, "mean_data_wait_seconds": 7.095932960510254e-05}
{"step": 21580, "total_steps": 46238, "tokens": 11668242432, "loss": 3.2480022582140835, "ce_loss": 3.246768908067183, "z_loss": 12.333327372868856, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19605377316474915, "tokens_per_second": 4309588.8790311925, "eta_seconds": 3093.4156083577336, "mean_data_wait_seconds": 6.002187728881836e-05}
{"step": 21590, "total_steps": 46238, "tokens": 11673649152, "loss": 3.27559797149716, "ce_loss": 3.2742947484507705, "z_loss": 13.03217339515686, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18721286952495575, "tokens_per_second": 4309506.705586805, "eta_seconds": 3092.219990451429, "mean_data_wait_seconds": 6.29127025604248e-05}
{"step": 21600, "total_steps": 46238, "tokens": 11679055872, "loss": 3.2360961202419167, "ce_loss": 3.234777614925847, "z_loss": 13.185395659822406, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20752839744091034, "tokens_per_second": 4309770.653581987, "eta_seconds": 3090.776084089041, "mean_data_wait_seconds": 7.051229476928711e-05}
{"step": 21610, "total_steps": 46238, "tokens": 11684462592, "loss": 3.1870034102237588, "ce_loss": 3.185708999633789, "z_loss": 12.943844419537168, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19198238849639893, "tokens_per_second": 4310017.577798152, "eta_seconds": 3089.3445568735397, "mean_data_wait_seconds": 8.419156074523926e-05}
{"step": 21620, "total_steps": 46238, "tokens": 11689869312, "loss": 3.2213036201216956, "ce_loss": 3.220015321717118, "z_loss": 12.883248639829231, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1885405331850052, "tokens_per_second": 4310086.680586649, "eta_seconds": 3088.0405918398847, "mean_data_wait_seconds": 0.0003033578395843506}
{"step": 21630, "total_steps": 46238, "tokens": 11695276032, "loss": 3.283390957297701, "ce_loss": 3.282115558783213, "z_loss": 12.754584543632738, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16526389122009277, "tokens_per_second": 4310240.06490194, "eta_seconds": 3086.676311219032, "mean_data_wait_seconds": 0.00012362003326416016}
{"step": 21640, "total_steps": 46238, "tokens": 11700682752, "loss": 3.241106235619747, "ce_loss": 3.2397909850785225, "z_loss": 13.151991403464114, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19384954869747162, "tokens_per_second": 4310328.5068303235, "eta_seconds": 3085.358613137259, "mean_data_wait_seconds": 6.979703903198242e-05}
{"step": 21650, "total_steps": 46238, "tokens": 11706089472, "loss": 3.224170885302804, "ce_loss": 3.222855979746038, "z_loss": 13.149187456477772, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19499564170837402, "tokens_per_second": 4310535.964916867, "eta_seconds": 3083.955817140799, "mean_data_wait_seconds": 9.480118751525879e-05}
{"step": 21660, "total_steps": 46238, "tokens": 11711496192, "loss": 3.2381815711657205, "ce_loss": 3.2368678540894478, "z_loss": 13.136783701000791, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1959611028432846, "tokens_per_second": 4310620.066150478, "eta_seconds": 3082.6413694739504, "mean_data_wait_seconds": 9.396672248840332e-05}
{"step": 21670, "total_steps": 46238, "tokens": 11716902912, "loss": 3.2222497716094507, "ce_loss": 3.2208958340413645, "z_loss": 13.53941695617907, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2222379595041275, "tokens_per_second": 4310775.352605617, "eta_seconds": 3081.2760901519437, "mean_data_wait_seconds": 6.181001663208008e-05}
{"step": 21680, "total_steps": 46238, "tokens": 11722309632, "loss": 3.19218700762951, "ce_loss": 3.19093890804233, "z_loss": 12.481102928970799, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.14788296818733215, "tokens_per_second": 4311073.940471142, "eta_seconds": 3079.8085320125533, "mean_data_wait_seconds": 6.639957427978516e-05}
{"step": 21690, "total_steps": 46238, "tokens": 11727716352, "loss": 3.2756310430440037, "ce_loss": 3.274305446581407, "z_loss": 13.256219863891602, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2048179805278778, "tokens_per_second": 4311270.556571418, "eta_seconds": 3078.4139872109054, "mean_data_wait_seconds": 7.236003875732422e-05}
{"step": 21700, "total_steps": 46238, "tokens": 11733123072, "loss": 3.230832327495922, "ce_loss": 3.2294917702674866, "z_loss": 13.405427499250932, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16908496618270874, "tokens_per_second": 4311446.588486259, "eta_seconds": 3077.034260247634, "mean_data_wait_seconds": 6.74128532409668e-05}
{"step": 21710, "total_steps": 46238, "tokens": 11738529792, "loss": 3.167599643721725, "ce_loss": 3.1662984508456606, "z_loss": 13.012341831669662, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19638247787952423, "tokens_per_second": 4311675.606548167, "eta_seconds": 3075.6168492500565, "mean_data_wait_seconds": 0.00011813640594482422}
{"step": 21720, "total_steps": 46238, "tokens": 11743936512, "loss": 3.1990187384865503, "ce_loss": 3.1977392742128083, "z_loss": 12.794722159703573, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16276997327804565, "tokens_per_second": 4311784.696172276, "eta_seconds": 3074.2850949323874, "mean_data_wait_seconds": 5.888938903808594e-05}
{"step": 21730, "total_steps": 46238, "tokens": 11749343232, "loss": 3.1903680021112617, "ce_loss": 3.1891059098821697, "z_loss": 12.621131441809915, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16055530309677124, "tokens_per_second": 4311693.5657787835, "eta_seconds": 3073.0961052439084, "mean_data_wait_seconds": 7.641315460205078e-05}
{"step": 21740, "total_steps": 46238, "tokens": 11754749952, "loss": 3.205752699664145, "ce_loss": 3.204395612080892, "z_loss": 13.57100696997209, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20633938908576965, "tokens_per_second": 4311800.218035875, "eta_seconds": 3071.7661566503034, "mean_data_wait_seconds": 0.00010210275650024414}
{"step": 21750, "total_steps": 46238, "tokens": 11760156672, "loss": 3.206386817224098, "ce_loss": 3.2051080772371003, "z_loss": 12.787458073009144, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1715753972530365, "tokens_per_second": 4311965.87595717, "eta_seconds": 3070.394257482641, "mean_data_wait_seconds": 8.663535118103027e-05}
{"step": 21760, "total_steps": 46238, "tokens": 11765563392, "loss": 3.177090657479835, "ce_loss": 3.1758375492962925, "z_loss": 12.531050956610477, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1999838501214981, "tokens_per_second": 4312111.468095613, "eta_seconds": 3069.036744971863, "mean_data_wait_seconds": 0.00012120604515075684}
{"step": 21770, "total_steps": 46238, "tokens": 11770970112, "loss": 3.216374872308789, "ce_loss": 3.2150975050348225, "z_loss": 12.773527947339145, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16504746675491333, "tokens_per_second": 4312120.627109057, "eta_seconds": 3067.776383813448, "mean_data_wait_seconds": 8.720159530639648e-05}
{"step": 21780, "total_steps": 46238, "tokens": 11776376832, "loss": 3.188557727770372, "ce_loss": 3.1872441064227712, "z_loss": 13.13614222497651, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20379102230072021, "tokens_per_second": 4312102.215086615, "eta_seconds": 3066.5356349245058, "mean_data_wait_seconds": 0.0001138150691986084}
{"step": 21790, "total_steps": 46238, "tokens": 11781783552, "loss": 3.2613733410835266, "ce_loss": 3.260048169078249, "z_loss": 13.25142148769263, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18747450411319733, "tokens_per_second": 4312188.776219884, "eta_seconds": 3065.220255868967, "mean_data_wait_seconds": 0.00015464425086975098}
{"step": 21800, "total_steps": 46238, "tokens": 11787190272, "loss": 3.2253030664993054, "ce_loss": 3.223981257640954, "z_loss": 13.217987710779363, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19244623184204102, "tokens_per_second": 4312116.516459514, "eta_seconds": 3064.0177772487723, "mean_data_wait_seconds": 9.012222290039062e-05}
{"step": 21810, "total_steps": 46238, "tokens": 11792596992, "loss": 3.240723709265391, "ce_loss": 3.2393827980214898, "z_loss": 13.408861189177543, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1793576180934906, "tokens_per_second": 4311924.506761949, "eta_seconds": 3062.900318242777, "mean_data_wait_seconds": 7.733702659606934e-05}
{"step": 21820, "total_steps": 46238, "tokens": 11798003712, "loss": 3.249502461968046, "ce_loss": 3.2482046195954988, "z_loss": 12.978753985780658, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16747014224529266, "tokens_per_second": 4311917.422682056, "eta_seconds": 3061.6514487395907, "mean_data_wait_seconds": 7.18533992767334e-05}
{"step": 21830, "total_steps": 46238, "tokens": 11803410432, "loss": 3.223056231484269, "ce_loss": 3.2217362595326975, "z_loss": 13.199862436814742, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1683284193277359, "tokens_per_second": 4311765.065188435, "eta_seconds": 3060.5056872279506, "mean_data_wait_seconds": 6.109476089477539e-05}
{"step": 21840, "total_steps": 46238, "tokens": 11808817152, "loss": 3.1717652118567266, "ce_loss": 3.170484559102492, "z_loss": 12.806582552013975, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18738308548927307, "tokens_per_second": 4311801.69697882, "eta_seconds": 3059.2257508601274, "mean_data_wait_seconds": 0.00011163949966430664}
{"step": 21850, "total_steps": 46238, "tokens": 11814223872, "loss": 3.245193351398815, "ce_loss": 3.2439147757761404, "z_loss": 12.785897702881783, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1803310513496399, "tokens_per_second": 4311763.040079368, "eta_seconds": 3057.9992317382294, "mean_data_wait_seconds": 9.080767631530762e-05}
{"step": 21860, "total_steps": 46238, "tokens": 11819630592, "loss": 3.2050210255565066, "ce_loss": 3.2036948962645098, "z_loss": 13.261337424769547, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19832906126976013, "tokens_per_second": 4311947.262302857, "eta_seconds": 3056.6146898932743, "mean_data_wait_seconds": 0.00014549493789672852}
{"step": 21870, "total_steps": 46238, "tokens": 11825037312, "loss": 3.1663913799054697, "ce_loss": 3.1651062694462864, "z_loss": 12.851154869252985, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20336192846298218, "tokens_per_second": 4311965.033973825, "eta_seconds": 3055.348204402896, "mean_data_wait_seconds": 0.00010138750076293945}
{"step": 21880, "total_steps": 46238, "tokens": 11830444032, "loss": 3.1571722500252, "ce_loss": 3.155861977374915, "z_loss": 13.10283011378664, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.21266601979732513, "tokens_per_second": 4312077.0767864585, "eta_seconds": 3054.014960654229, "mean_data_wait_seconds": 0.00038123130798339844}
{"step": 21890, "total_steps": 46238, "tokens": 11835850752, "loss": 3.223274068398909, "ce_loss": 3.221941462068847, "z_loss": 13.32604904608293, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19887277483940125, "tokens_per_second": 4312130.850062334, "eta_seconds": 3052.7230368739183, "mean_data_wait_seconds": 0.00011387467384338379}
{"step": 21900, "total_steps": 46238, "tokens": 11841257472, "loss": 3.182423806551731, "ce_loss": 3.181113687428561, "z_loss": 13.101166840755578, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18629150092601776, "tokens_per_second": 4312361.199719938, "eta_seconds": 3051.3061996881324, "mean_data_wait_seconds": 7.855892181396484e-05}
{"step": 21910, "total_steps": 46238, "tokens": 11846664192, "loss": 3.1947828621575325, "ce_loss": 3.1934741229721992, "z_loss": 13.087494351647116, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17257747054100037, "tokens_per_second": 4312368.384836965, "eta_seconds": 3050.0473452704027, "mean_data_wait_seconds": 6.568431854248047e-05}
{"step": 21920, "total_steps": 46238, "tokens": 11852070912, "loss": 3.256792523644187, "ce_loss": 3.2555540294358223, "z_loss": 12.384732073003596, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19538477063179016, "tokens_per_second": 4312468.635531316, "eta_seconds": 3048.7227004214874, "mean_data_wait_seconds": 8.508563041687012e-05}
{"step": 21930, "total_steps": 46238, "tokens": 11857477632, "loss": 3.248772753007484, "ce_loss": 3.2474971558108474, "z_loss": 12.756076595999978, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1883898675441742, "tokens_per_second": 4312414.946963553, "eta_seconds": 3047.506899412264, "mean_data_wait_seconds": 9.974837303161621e-05}
{"step": 21940, "total_steps": 46238, "tokens": 11862884352, "loss": 3.2648980473027085, "ce_loss": 3.2636751640926707, "z_loss": 12.22839314287359, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.200348898768425, "tokens_per_second": 4312556.350576217, "eta_seconds": 3046.153259480251, "mean_data_wait_seconds": 7.531046867370605e-05}
{"step": 21950, "total_steps": 46238, "tokens": 11868291072, "loss": 3.2633578470259, "ce_loss": 3.2620486490654224, "z_loss": 13.091967019167813, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17276914417743683, "tokens_per_second": 4312615.715127291, "eta_seconds": 3044.8576296607075, "mean_data_wait_seconds": 7.617473602294922e-05}
{"step": 21960, "total_steps": 46238, "tokens": 11873697792, "loss": 3.242266349720232, "ce_loss": 3.2410075375528047, "z_loss": 12.587837154215032, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1694488227367401, "tokens_per_second": 4312751.205332818, "eta_seconds": 3043.5083126913332, "mean_data_wait_seconds": 8.112192153930664e-05}
{"step": 21970, "total_steps": 46238, "tokens": 11879104512, "loss": 3.20981128468658, "ce_loss": 3.208488605239175, "z_loss": 13.226653178532919, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18394888937473297, "tokens_per_second": 4312751.098409534, "eta_seconds": 3042.2547289683844, "mean_data_wait_seconds": 8.934736251831055e-05}
{"step": 21980, "total_steps": 46238, "tokens": 11884511232, "loss": 3.2406894943930884, "ce_loss": 3.2393887494549607, "z_loss": 13.007313684983687, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1708691418170929, "tokens_per_second": 4312860.960276335, "eta_seconds": 3040.9236061159013, "mean_data_wait_seconds": 6.791949272155762e-05}
{"step": 21990, "total_steps": 46238, "tokens": 11889917952, "loss": 3.2968304175319094, "ce_loss": 3.2955706787831858, "z_loss": 12.59688631693522, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1771150380373001, "tokens_per_second": 4312789.307846913, "eta_seconds": 3039.720479771079, "mean_data_wait_seconds": 7.906556129455566e-05}
{"step": 22000, "total_steps": 46238, "tokens": 11895324672, "loss": 3.2358467398267803, "ce_loss": 3.2345673748941133, "z_loss": 12.793601064971, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17246899008750916, "tokens_per_second": 4313083.892190146, "eta_seconds": 3038.259303912071, "mean_data_wait_seconds": 5.835294723510742e-05}
{"step": 22010, "total_steps": 46238, "tokens": 11900731392, "loss": 3.248362927725821, "ce_loss": 3.247034470240275, "z_loss": 13.284112070546005, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18201255798339844, "tokens_per_second": 4313310.06140577, "eta_seconds": 3036.8464955034774, "mean_data_wait_seconds": 6.270408630371094e-05}
{"step": 22020, "total_steps": 46238, "tokens": 11906138112, "loss": 3.256436571930394, "ce_loss": 3.255102175654787, "z_loss": 13.343864672111742, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2054738998413086, "tokens_per_second": 4313483.86175556, "eta_seconds": 3035.4706876476057, "mean_data_wait_seconds": 6.714463233947754e-05}
{"step": 22030, "total_steps": 46238, "tokens": 11911544832, "loss": 3.2029921827894268, "ce_loss": 3.2016744252407188, "z_loss": 13.177700794104373, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1821095496416092, "tokens_per_second": 4313859.691485524, "eta_seconds": 3033.952896018505, "mean_data_wait_seconds": 0.00010406970977783203}
{"step": 22040, "total_steps": 46238, "tokens": 11916951552, "loss": 3.2832433165925923, "ce_loss": 3.281900852015524, "z_loss": 13.424631732882876, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1669762134552002, "tokens_per_second": 4313944.464269231, "eta_seconds": 3032.6399638100484, "mean_data_wait_seconds": 6.476044654846191e-05}
{"step": 22050, "total_steps": 46238, "tokens": 11922358272, "loss": 3.260833857637463, "ce_loss": 3.2595172398018115, "z_loss": 13.166036576935738, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.2127273827791214, "tokens_per_second": 4314042.395747905, "eta_seconds": 3031.3178370452392, "mean_data_wait_seconds": 5.796551704406738e-05}
{"step": 22060, "total_steps": 46238, "tokens": 11927764992, "loss": 3.216312742594517, "ce_loss": 3.2150450005675806, "z_loss": 12.677460236982865, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18611930310726166, "tokens_per_second": 4314042.919083287, "eta_seconds": 3030.064185540764, "mean_data_wait_seconds": 6.172060966491699e-05}
{"step": 22070, "total_steps": 46238, "tokens": 11933171712, "loss": 3.2100695172945657, "ce_loss": 3.208817384459756, "z_loss": 12.521649505152848, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1956552118062973, "tokens_per_second": 4314211.891299341, "eta_seconds": 3028.692274097992, "mean_data_wait_seconds": 0.00011178851127624512}
{"step": 22080, "total_steps": 46238, "tokens": 11938578432, "loss": 3.2280024777759206, "ce_loss": 3.226723044207602, "z_loss": 12.794414787581474, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.180217906832695, "tokens_per_second": 4314420.939375555, "eta_seconds": 3027.292349895367, "mean_data_wait_seconds": 7.981061935424805e-05}
{"step": 22090, "total_steps": 46238, "tokens": 11943985152, "loss": 3.2572143547462695, "ce_loss": 3.255902107917901, "z_loss": 13.122064193089804, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1686088591814041, "tokens_per_second": 4314550.548648573, "eta_seconds": 3025.948273590014, "mean_data_wait_seconds": 8.186697959899902e-05}
{"step": 22100, "total_steps": 46238, "tokens": 11949391872, "loss": 3.25338398326527, "ce_loss": 3.252085900667942, "z_loss": 12.98110487244346, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18304350972175598, "tokens_per_second": 4314659.938240664, "eta_seconds": 3024.618452160408, "mean_data_wait_seconds": 6.207823753356934e-05}
{"step": 22110, "total_steps": 46238, "tokens": 11954798592, "loss": 3.2294924981666333, "ce_loss": 3.2281958695613975, "z_loss": 12.96615117968935, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1730073094367981, "tokens_per_second": 4314855.0103793405, "eta_seconds": 3023.2286629842442, "mean_data_wait_seconds": 0.00015750527381896973}
{"step": 22120, "total_steps": 46238, "tokens": 11960205312, "loss": 3.28245534138246, "ce_loss": 3.2811859900301155, "z_loss": 12.694121418577252, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16912703216075897, "tokens_per_second": 4314991.722092723, "eta_seconds": 3021.879869951649, "mean_data_wait_seconds": 6.386637687683105e-05}
{"step": 22130, "total_steps": 46238, "tokens": 11965612032, "loss": 3.311424445022236, "ce_loss": 3.3101111686590947, "z_loss": 13.132410157810558, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.20671550929546356, "tokens_per_second": 4315142.2457816135, "eta_seconds": 3020.521494219971, "mean_data_wait_seconds": 7.426738739013672e-05}
{"step": 22140, "total_steps": 46238, "tokens": 11971018752, "loss": 3.2876005028233384, "ce_loss": 3.2862697857798953, "z_loss": 13.307367837790286, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.19969207048416138, "tokens_per_second": 4315348.930818862, "eta_seconds": 3019.123920884829, "mean_data_wait_seconds": 0.00010094046592712402}
{"step": 22150, "total_steps": 46238, "tokens": 11976425472, "loss": 3.287075774236159, "ce_loss": 3.285783260157614, "z_loss": 12.925158890810879, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.16662287712097168, "tokens_per_second": 4315621.498063558, "eta_seconds": 3017.6804128544554, "mean_data_wait_seconds": 6.851553916931152e-05}
{"step": 22160, "total_steps": 46238, "tokens": 11981832192, "loss": 3.2347316832253425, "ce_loss": 3.233455255176082, "z_loss": 12.763810374520041, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.17901009321212769, "tokens_per_second": 4315718.308458594, "eta_seconds": 3016.359922862861, "mean_data_wait_seconds": 9.959936141967773e-05}
{"step": 22170, "total_steps": 46238, "tokens": 11987238912, "loss": 3.251772185166677, "ce_loss": 3.2504805525143943, "z_loss": 12.916224494124904, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.1730746328830719, "tokens_per_second": 4315797.053959829, "eta_seconds": 3015.052112346411, "mean_data_wait_seconds": 0.00010463595390319824}
{"step": 22180, "total_steps": 46238, "tokens": 11992645632, "loss": 3.2280679074200718, "ce_loss": 3.2268266117934026, "z_loss": 12.412748495737711, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.22529011964797974, "tokens_per_second": 4315784.689352633, "eta_seconds": 3013.807972415565, "mean_data_wait_seconds": 0.0001024007797241211}
{"step": 22190, "total_steps": 46238, "tokens": 11998052352, "loss": 3.192684769630432, "ce_loss": 3.191402437108936, "z_loss": 12.823230996276394, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.1, "grad_norm": 0.18768827617168427, "tokens_per_second": 4315830.104463663, "eta_seconds": 3012.5234935807853, "mean_data_wait_seconds": 9.524822235107422e-05}
{"step": 22200, "total_steps": 46238, "tokens": 12003459072, "loss": 3.1930714430231038, "ce_loss": 3.1930714430231038, "z_loss": 13.020881674506448, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17164914309978485, "tokens_per_second": 4303130.602432367, "eta_seconds": 3020.1576630404543, "mean_data_wait_seconds": 8.064508438110352e-05}
{"step": 22210, "total_steps": 46238, "tokens": 12008865792, "loss": 3.189373563636433, "ce_loss": 3.189373563636433, "z_loss": 13.390790072354404, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17297130823135376, "tokens_per_second": 4303170.616963466, "eta_seconds": 3018.8731287552137, "mean_data_wait_seconds": 9.608268737792969e-05}
{"step": 22220, "total_steps": 46238, "tokens": 12014272512, "loss": 3.2393430092118005, "ce_loss": 3.2393430092118005, "z_loss": 14.06059497052973, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16526766121387482, "tokens_per_second": 4303082.972627289, "eta_seconds": 3017.6781406731016, "mean_data_wait_seconds": 0.00012347102165222168}
{"step": 22230, "total_steps": 46238, "tokens": 12019679232, "loss": 3.2419105461149504, "ce_loss": 3.2419105461149504, "z_loss": 13.893281185265744, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16031543910503387, "tokens_per_second": 4303183.7675541155, "eta_seconds": 3016.3510101214306, "mean_data_wait_seconds": 0.0001506805419921875}
{"step": 22240, "total_steps": 46238, "tokens": 12025085952, "loss": 3.1483965479966365, "ce_loss": 3.1483965479966365, "z_loss": 14.518057693134654, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16544897854328156, "tokens_per_second": 4303109.366600413, "eta_seconds": 3015.1466947841614, "mean_data_wait_seconds": 8.654594421386719e-05}
{"step": 22250, "total_steps": 46238, "tokens": 12030492672, "loss": 3.2120541388338264, "ce_loss": 3.2120541388338264, "z_loss": 15.124961058298746, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16111251711845398, "tokens_per_second": 4303160.052954847, "eta_seconds": 3013.854726387535, "mean_data_wait_seconds": 0.0001176595687866211}
{"step": 22260, "total_steps": 46238, "tokens": 12035899392, "loss": 3.1621401165470933, "ce_loss": 3.1621401165470933, "z_loss": 14.748066548145179, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15886664390563965, "tokens_per_second": 4303110.381909881, "eta_seconds": 3012.6330475971267, "mean_data_wait_seconds": 0.004377007484436035}
{"step": 22270, "total_steps": 46238, "tokens": 12041306112, "loss": 3.13354363044103, "ce_loss": 3.13354363044103, "z_loss": 15.010689323598688, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1314958781003952, "tokens_per_second": 4303318.299326336, "eta_seconds": 3011.2310832383832, "mean_data_wait_seconds": 9.244680404663086e-05}
{"step": 22280, "total_steps": 46238, "tokens": 12046712832, "loss": 3.1452734795483677, "ce_loss": 3.1452734795483677, "z_loss": 14.488497611248132, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1585489958524704, "tokens_per_second": 4303329.701895074, "eta_seconds": 3009.9667005053993, "mean_data_wait_seconds": 0.00010752677917480469}
{"step": 22290, "total_steps": 46238, "tokens": 12052119552, "loss": 3.1530298102985728, "ce_loss": 3.1530298102985728, "z_loss": 14.916781187057495, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1419956088066101, "tokens_per_second": 4303384.898403005, "eta_seconds": 3008.6717060342044, "mean_data_wait_seconds": 0.00010710954666137695}
{"step": 22300, "total_steps": 46238, "tokens": 12057526272, "loss": 3.1148116949832803, "ce_loss": 3.1148116949832803, "z_loss": 15.767718307899706, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1305389255285263, "tokens_per_second": 4303431.535350824, "eta_seconds": 3007.382726479216, "mean_data_wait_seconds": 9.697675704956055e-05}
{"step": 22310, "total_steps": 46238, "tokens": 12062932992, "loss": 3.173183858394623, "ce_loss": 3.173183858394623, "z_loss": 15.613172061515577, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1391395628452301, "tokens_per_second": 4303653.995872752, "eta_seconds": 3005.970962444097, "mean_data_wait_seconds": 0.00015777349472045898}
{"step": 22320, "total_steps": 46238, "tokens": 12068339712, "loss": 3.192966296817317, "ce_loss": 3.192966296817317, "z_loss": 15.728248278299967, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1448303759098053, "tokens_per_second": 4303680.57077163, "eta_seconds": 3004.696099385807, "mean_data_wait_seconds": 6.085634231567383e-05}
{"step": 22330, "total_steps": 46238, "tokens": 12073746432, "loss": 3.2245239904432585, "ce_loss": 3.2245239904432585, "z_loss": 15.665911869569259, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15686561167240143, "tokens_per_second": 4303820.199479678, "eta_seconds": 3003.342357462495, "mean_data_wait_seconds": 7.545948028564453e-05}
{"step": 22340, "total_steps": 46238, "tokens": 12079153152, "loss": 3.1830131176746255, "ce_loss": 3.1830131176746255, "z_loss": 15.626882148511482, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13810034096240997, "tokens_per_second": 4303974.021887285, "eta_seconds": 3001.978803379118, "mean_data_wait_seconds": 7.876753807067871e-05}
{"step": 22350, "total_steps": 46238, "tokens": 12084559872, "loss": 3.2045903133623526, "ce_loss": 3.2045903133623526, "z_loss": 15.501897443424571, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14131371676921844, "tokens_per_second": 4304063.346799424, "eta_seconds": 3000.6603117502536, "mean_data_wait_seconds": 7.200241088867188e-05}
{"step": 22360, "total_steps": 46238, "tokens": 12089966592, "loss": 3.163914214481007, "ce_loss": 3.163914214481007, "z_loss": 16.03535106687835, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12635144591331482, "tokens_per_second": 4304137.9255765965, "eta_seconds": 2999.352150702881, "mean_data_wait_seconds": 6.571412086486816e-05}
{"step": 22370, "total_steps": 46238, "tokens": 12095373312, "loss": 3.1462002848133896, "ce_loss": 3.1462002848133896, "z_loss": 16.39586187131477, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15461799502372742, "tokens_per_second": 4304186.29090086, "eta_seconds": 2998.062293744067, "mean_data_wait_seconds": 0.00013825297355651855}
{"step": 22380, "total_steps": 46238, "tokens": 12100780032, "loss": 3.127339572617502, "ce_loss": 3.127339572617502, "z_loss": 16.521342985557787, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14465394616127014, "tokens_per_second": 4304271.083710733, "eta_seconds": 2996.747103781361, "mean_data_wait_seconds": 0.00011047720909118652}
{"step": 22390, "total_steps": 46238, "tokens": 12106186752, "loss": 3.1689538124835854, "ce_loss": 3.1689538124835854, "z_loss": 16.199348962668218, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13961336016654968, "tokens_per_second": 4304179.645483202, "eta_seconds": 2995.5546110930372, "mean_data_wait_seconds": 9.581446647644043e-05}
{"step": 22400, "total_steps": 46238, "tokens": 12111593472, "loss": 3.0701732888366235, "ce_loss": 3.0701732888366235, "z_loss": 16.36047364726211, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14623074233531952, "tokens_per_second": 4304445.213781045, "eta_seconds": 2994.113718241316, "mean_data_wait_seconds": 9.03010368347168e-05}
{"step": 22410, "total_steps": 46238, "tokens": 12117000192, "loss": 3.0802645285924277, "ce_loss": 3.0802645285924277, "z_loss": 16.563104398322828, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14060918986797333, "tokens_per_second": 4303978.620841435, "eta_seconds": 2993.1820947292326, "mean_data_wait_seconds": 0.00013116002082824707}
{"step": 22420, "total_steps": 46238, "tokens": 12122406912, "loss": 3.17918332417806, "ce_loss": 3.17918332417806, "z_loss": 16.793244087334834, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16336901485919952, "tokens_per_second": 4304078.56724574, "eta_seconds": 2991.8564038296236, "mean_data_wait_seconds": 5.9098005294799805e-05}
{"step": 22430, "total_steps": 46238, "tokens": 12127813632, "loss": 3.091122480955991, "ce_loss": 3.091122480955991, "z_loss": 17.06676058335738, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1314115822315216, "tokens_per_second": 4304298.644245399, "eta_seconds": 2990.4473104367025, "mean_data_wait_seconds": 6.073713302612305e-05}
{"step": 22440, "total_steps": 46238, "tokens": 12133220352, "loss": 3.051359409635717, "ce_loss": 3.051359409635717, "z_loss": 17.183593880046498, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14039398729801178, "tokens_per_second": 4304506.954143132, "eta_seconds": 2989.046532173908, "mean_data_wait_seconds": 7.677078247070312e-05}
{"step": 22450, "total_steps": 46238, "tokens": 12138627072, "loss": 3.0942014708663477, "ce_loss": 3.0942014708663477, "z_loss": 16.971869945526123, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1156948134303093, "tokens_per_second": 4304443.12164229, "eta_seconds": 2987.8347792160184, "mean_data_wait_seconds": 7.31050968170166e-05}
{"step": 22460, "total_steps": 46238, "tokens": 12144033792, "loss": 3.136084036393599, "ce_loss": 3.136084036393599, "z_loss": 17.195040356029164, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17470650374889374, "tokens_per_second": 4304518.647084081, "eta_seconds": 2986.526298987802, "mean_data_wait_seconds": 7.870793342590332e-05}
{"step": 22470, "total_steps": 46238, "tokens": 12149440512, "loss": 3.1315372441754197, "ce_loss": 3.1315372441754197, "z_loss": 17.40861275701812, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.11806433647871017, "tokens_per_second": 4304508.063270083, "eta_seconds": 2985.2775822744993, "mean_data_wait_seconds": 7.557868957519531e-05}
{"step": 22480, "total_steps": 46238, "tokens": 12154847232, "loss": 3.119620570630738, "ce_loss": 3.119620570630738, "z_loss": 16.95385035601529, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13731588423252106, "tokens_per_second": 4304532.182317975, "eta_seconds": 2984.0048023716136, "mean_data_wait_seconds": 8.368492126464844e-05}
{"step": 22490, "total_steps": 46238, "tokens": 12160253952, "loss": 3.095534917080041, "ce_loss": 3.095534917080041, "z_loss": 16.73614620439934, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14570817351341248, "tokens_per_second": 4304501.854122844, "eta_seconds": 2982.7697650315813, "mean_data_wait_seconds": 8.004903793334961e-05}
{"step": 22500, "total_steps": 46238, "tokens": 12165660672, "loss": 3.1101063164797695, "ce_loss": 3.1101063164797695, "z_loss": 17.16783942598285, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14550140500068665, "tokens_per_second": 4304752.142139277, "eta_seconds": 2981.3403513685425, "mean_data_wait_seconds": 8.71419906616211e-05}
{"step": 22510, "total_steps": 46238, "tokens": 12171067392, "loss": 3.154459052013628, "ce_loss": 3.154459052013628, "z_loss": 17.847253250353265, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12617486715316772, "tokens_per_second": 4304832.729628839, "eta_seconds": 2980.028574793444, "mean_data_wait_seconds": 5.9485435485839844e-05}
{"step": 22520, "total_steps": 46238, "tokens": 12176474112, "loss": 3.120338159980196, "ce_loss": 3.120338159980196, "z_loss": 17.383286519484088, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14742232859134674, "tokens_per_second": 4304862.96305767, "eta_seconds": 2978.751689436349, "mean_data_wait_seconds": 7.367134094238281e-05}
{"step": 22530, "total_steps": 46238, "tokens": 12181880832, "loss": 3.1063419128909255, "ce_loss": 3.1063419128909255, "z_loss": 17.358778968001857, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.11858118325471878, "tokens_per_second": 4305005.301372637, "eta_seconds": 2977.3972868077803, "mean_data_wait_seconds": 7.396936416625977e-05}
{"step": 22540, "total_steps": 46238, "tokens": 12187287552, "loss": 3.191684849334486, "ce_loss": 3.191684849334486, "z_loss": 17.53145777095448, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12650518119335175, "tokens_per_second": 4304965.918457617, "eta_seconds": 2976.1685984706683, "mean_data_wait_seconds": 7.662177085876465e-05}
{"step": 22550, "total_steps": 46238, "tokens": 12192694272, "loss": 3.1366942723592124, "ce_loss": 3.1366942723592124, "z_loss": 17.378084096041594, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1351752132177353, "tokens_per_second": 4304969.063228455, "eta_seconds": 2974.910498984082, "mean_data_wait_seconds": 7.06017017364502e-05}
{"step": 22560, "total_steps": 46238, "tokens": 12198100992, "loss": 3.198485533396403, "ce_loss": 3.198485533396403, "z_loss": 18.126418922886703, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14117242395877838, "tokens_per_second": 4305212.272533415, "eta_seconds": 2973.486586404002, "mean_data_wait_seconds": 9.179115295410156e-05}
{"step": 22570, "total_steps": 46238, "tokens": 12203507712, "loss": 3.1417480996160796, "ce_loss": 3.1417480996160796, "z_loss": 17.324174953229498, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14586742222309113, "tokens_per_second": 4304233.860160332, "eta_seconds": 2972.9063614409065, "mean_data_wait_seconds": 0.00013399124145507812}
{"step": 22580, "total_steps": 46238, "tokens": 12208914432, "loss": 3.070852615616538, "ce_loss": 3.070852615616538, "z_loss": 17.6256805044232, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1264999955892563, "tokens_per_second": 4304286.282402499, "eta_seconds": 2971.6140295530486, "mean_data_wait_seconds": 8.723139762878418e-05}
{"step": 22590, "total_steps": 46238, "tokens": 12214321152, "loss": 3.0922934503266304, "ce_loss": 3.0922934503266304, "z_loss": 17.610438000072133, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1420062780380249, "tokens_per_second": 4304349.791363409, "eta_seconds": 2970.3140784825127, "mean_data_wait_seconds": 8.988380432128906e-05}
{"step": 22600, "total_steps": 46238, "tokens": 12219727872, "loss": 3.1024362155885408, "ce_loss": 3.1024362155885408, "z_loss": 18.008016268412273, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13496166467666626, "tokens_per_second": 4304499.615861955, "eta_seconds": 2968.954629919486, "mean_data_wait_seconds": 7.829070091247559e-05}
{"step": 22610, "total_steps": 46238, "tokens": 12225134592, "loss": 3.1057425863815076, "ce_loss": 3.1057425863815076, "z_loss": 17.523914004817154, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.11913365125656128, "tokens_per_second": 4303890.863692444, "eta_seconds": 2968.1183256213867, "mean_data_wait_seconds": 0.00011372566223144531}
{"step": 22620, "total_steps": 46238, "tokens": 12230541312, "loss": 3.1248217929493296, "ce_loss": 3.1248217929493296, "z_loss": 17.969104564551152, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1342347413301468, "tokens_per_second": 4304040.640726007, "eta_seconds": 2966.7588412562277, "mean_data_wait_seconds": 8.258223533630371e-05}
{"step": 22630, "total_steps": 46238, "tokens": 12235948032, "loss": 3.103072256752939, "ce_loss": 3.103072256752939, "z_loss": 18.206525166829426, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.11347638070583344, "tokens_per_second": 4303904.149335112, "eta_seconds": 2965.596691081466, "mean_data_wait_seconds": 0.00011974573135375977}
{"step": 22640, "total_steps": 46238, "tokens": 12241354752, "loss": 3.141548465598713, "ce_loss": 3.141548465598713, "z_loss": 17.920441497455943, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15167514979839325, "tokens_per_second": 4303912.279522059, "eta_seconds": 2964.3348552207894, "mean_data_wait_seconds": 0.00015118718147277832}
{"step": 22650, "total_steps": 46238, "tokens": 12246761472, "loss": 3.151855595184095, "ce_loss": 3.151855595184095, "z_loss": 17.750769398429178, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12263074517250061, "tokens_per_second": 4304082.470715132, "eta_seconds": 2962.9614559595307, "mean_data_wait_seconds": 0.00013914704322814941}
{"step": 22660, "total_steps": 46238, "tokens": 12252168192, "loss": 3.0849980639688894, "ce_loss": 3.0849980639688894, "z_loss": 17.97901306730328, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14686264097690582, "tokens_per_second": 4303276.505931533, "eta_seconds": 2962.2599724719657, "mean_data_wait_seconds": 6.720423698425293e-05}
{"step": 22670, "total_steps": 46238, "tokens": 12257574912, "loss": 3.076900092038241, "ce_loss": 3.076900092038241, "z_loss": 17.728756803454775, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13207592070102692, "tokens_per_second": 4303417.854530842, "eta_seconds": 2960.9062969761585, "mean_data_wait_seconds": 5.97536563873291e-05}
{"step": 22680, "total_steps": 46238, "tokens": 12262981632, "loss": 3.109686410788334, "ce_loss": 3.109686410788334, "z_loss": 17.919835495226312, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1288301646709442, "tokens_per_second": 4303473.111628575, "eta_seconds": 2959.6119166131025, "mean_data_wait_seconds": 6.702542304992676e-05}
{"step": 22690, "total_steps": 46238, "tokens": 12268388352, "loss": 3.1337460911635198, "ce_loss": 3.1337460911635198, "z_loss": 17.885648091634113, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12309083342552185, "tokens_per_second": 4303477.125707816, "eta_seconds": 2958.3527952192912, "mean_data_wait_seconds": 6.181001663208008e-05}
{"step": 22700, "total_steps": 46238, "tokens": 12273795072, "loss": 3.1978946400411203, "ce_loss": 3.1978946400411203, "z_loss": 17.836496208653305, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15347401797771454, "tokens_per_second": 4303603.483581685, "eta_seconds": 2957.0096112592887, "mean_data_wait_seconds": 0.00011295080184936523}
{"step": 22710, "total_steps": 46238, "tokens": 12279201792, "loss": 3.118797986796408, "ce_loss": 3.118797986796408, "z_loss": 18.151756662310977, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12458577752113342, "tokens_per_second": 4303552.572601121, "eta_seconds": 2955.78825386851, "mean_data_wait_seconds": 0.00010764598846435547}
{"step": 22720, "total_steps": 46238, "tokens": 12284608512, "loss": 3.0378565138036553, "ce_loss": 3.0378565138036553, "z_loss": 18.104312188697584, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1276123821735382, "tokens_per_second": 4303762.27099889, "eta_seconds": 2954.3879571788925, "mean_data_wait_seconds": 7.838010787963867e-05}
{"step": 22730, "total_steps": 46238, "tokens": 12290015232, "loss": 3.108329800042239, "ce_loss": 3.108329800042239, "z_loss": 18.53858652981845, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15187481045722961, "tokens_per_second": 4303981.451193336, "eta_seconds": 2952.9812914217136, "mean_data_wait_seconds": 6.16610050201416e-05}
{"step": 22740, "total_steps": 46238, "tokens": 12295421952, "loss": 3.1095235474181897, "ce_loss": 3.1095235474181897, "z_loss": 18.355959372086957, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12317167222499847, "tokens_per_second": 4304099.660528204, "eta_seconds": 2951.6440105945244, "mean_data_wait_seconds": 8.70823860168457e-05}
{"step": 22750, "total_steps": 46238, "tokens": 12300828672, "loss": 3.0615679820378623, "ce_loss": 3.0615679820378623, "z_loss": 18.647030223499645, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13669556379318237, "tokens_per_second": 4304273.085487881, "eta_seconds": 2950.268956403964, "mean_data_wait_seconds": 5.790591239929199e-05}
{"step": 22760, "total_steps": 46238, "tokens": 12306235392, "loss": 3.100607754606189, "ce_loss": 3.100607754606189, "z_loss": 18.864118258158367, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.11135402321815491, "tokens_per_second": 4304387.045612436, "eta_seconds": 2948.934751799014, "mean_data_wait_seconds": 7.298588752746582e-05}
{"step": 22770, "total_steps": 46238, "tokens": 12311642112, "loss": 3.0451686797720012, "ce_loss": 3.0451686797720012, "z_loss": 18.718534281759553, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13182735443115234, "tokens_per_second": 4304645.197883301, "eta_seconds": 2947.5018824406184, "mean_data_wait_seconds": 7.584691047668457e-05}
{"step": 22780, "total_steps": 46238, "tokens": 12317048832, "loss": 3.1176319953167075, "ce_loss": 3.1176319953167075, "z_loss": 18.515937848524615, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14118216931819916, "tokens_per_second": 4304903.2017935375, "eta_seconds": 2946.0692864629605, "mean_data_wait_seconds": 0.0001310110092163086}
{"step": 22790, "total_steps": 46238, "tokens": 12322455552, "loss": 3.0409149906852027, "ce_loss": 3.0409149906852027, "z_loss": 18.307021141052246, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12207658588886261, "tokens_per_second": 4305163.905597855, "eta_seconds": 2944.6350155255086, "mean_data_wait_seconds": 0.00011223554611206055}
{"step": 22800, "total_steps": 46238, "tokens": 12327862272, "loss": 3.0754784508184954, "ce_loss": 3.0754784508184954, "z_loss": 19.166194366686273, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1274603307247162, "tokens_per_second": 4305431.557097575, "eta_seconds": 2943.196168827825, "mean_data_wait_seconds": 9.247660636901855e-05}
{"step": 22810, "total_steps": 46238, "tokens": 12333268992, "loss": 3.037587133320895, "ce_loss": 3.037587133320895, "z_loss": 18.859124747189608, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1550484150648117, "tokens_per_second": 4305367.5033702245, "eta_seconds": 2941.9841474821496, "mean_data_wait_seconds": 7.721781730651855e-05}
{"step": 22820, "total_steps": 46238, "tokens": 12338675712, "loss": 3.062802316564502, "ce_loss": 3.062802316564502, "z_loss": 19.188371109239984, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1140400618314743, "tokens_per_second": 4305556.003535649, "eta_seconds": 2940.599591226562, "mean_data_wait_seconds": 9.1552734375e-05}
{"step": 22830, "total_steps": 46238, "tokens": 12344082432, "loss": 3.0181021690368652, "ce_loss": 3.0181021690368652, "z_loss": 19.157985094821814, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1231839656829834, "tokens_per_second": 4305645.459739921, "eta_seconds": 2939.2827677837754, "mean_data_wait_seconds": 0.00012916326522827148}
{"step": 22840, "total_steps": 46238, "tokens": 12349489152, "loss": 3.08104608817534, "ce_loss": 3.08104608817534, "z_loss": 18.68727543859771, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12846243381500244, "tokens_per_second": 4305834.262992108, "eta_seconds": 2937.8982123686037, "mean_data_wait_seconds": 6.535649299621582e-05}
{"step": 22850, "total_steps": 46238, "tokens": 12354895872, "loss": 3.1152024196855947, "ce_loss": 3.1152024196855947, "z_loss": 18.812464959693678, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1190275326371193, "tokens_per_second": 4305970.744314105, "eta_seconds": 2936.549460002001, "mean_data_wait_seconds": 6.920099258422852e-05}
{"step": 22860, "total_steps": 46238, "tokens": 12360302592, "loss": 3.0912709597385293, "ce_loss": 3.0912709597385293, "z_loss": 19.110251238851838, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.119440458714962, "tokens_per_second": 4305965.329750733, "eta_seconds": 2935.2975177651215, "mean_data_wait_seconds": 9.563565254211426e-05}
{"step": 22870, "total_steps": 46238, "tokens": 12365709312, "loss": 3.18647981412483, "ce_loss": 3.18647981412483, "z_loss": 19.543462305357963, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15234515070915222, "tokens_per_second": 4305930.948260112, "eta_seconds": 2934.0653103378136, "mean_data_wait_seconds": 8.881092071533203e-05}
{"step": 22880, "total_steps": 46238, "tokens": 12371116032, "loss": 3.0943546259041987, "ce_loss": 3.0943546259041987, "z_loss": 19.304396455938164, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12244407087564468, "tokens_per_second": 4306064.948543385, "eta_seconds": 2932.718399491824, "mean_data_wait_seconds": 9.769201278686523e-05}
{"step": 22890, "total_steps": 46238, "tokens": 12376522752, "loss": 3.0906619194782143, "ce_loss": 3.0906619194782143, "z_loss": 18.802439068302963, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13240396976470947, "tokens_per_second": 4306180.659344671, "eta_seconds": 2931.3840227783244, "mean_data_wait_seconds": 0.00010764598846435547}
{"step": 22900, "total_steps": 46238, "tokens": 12381929472, "loss": 3.1350304448243342, "ce_loss": 3.1350304448243342, "z_loss": 18.806839364947695, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13683070242404938, "tokens_per_second": 4306241.99378628, "eta_seconds": 2930.086716493578, "mean_data_wait_seconds": 8.657574653625488e-05}
{"step": 22910, "total_steps": 46238, "tokens": 12387336192, "loss": 3.059022338101358, "ce_loss": 3.059022338101358, "z_loss": 18.70132569110755, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13412803411483765, "tokens_per_second": 4306436.139159936, "eta_seconds": 2928.699122996932, "mean_data_wait_seconds": 0.00030553340911865234}
{"step": 22920, "total_steps": 46238, "tokens": 12392742912, "loss": 3.0917085481412485, "ce_loss": 3.0917085481412485, "z_loss": 19.085005167758826, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15034836530685425, "tokens_per_second": 4304581.332897362, "eta_seconds": 2928.7050351803855, "mean_data_wait_seconds": 0.00013700127601623535}
{"step": 22930, "total_steps": 46238, "tokens": 12398149632, "loss": 3.0945724736560476, "ce_loss": 3.0945724736560476, "z_loss": 19.254445018190324, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1364116221666336, "tokens_per_second": 4304007.186407607, "eta_seconds": 2927.8395128605607, "mean_data_wait_seconds": 7.107853889465332e-05}
{"step": 22940, "total_steps": 46238, "tokens": 12403556352, "loss": 3.051014026006063, "ce_loss": 3.051014026006063, "z_loss": 18.872712395407937, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12149541825056076, "tokens_per_second": 4304040.34252945, "eta_seconds": 2926.5607618811055, "mean_data_wait_seconds": 8.317828178405762e-05}
{"step": 22950, "total_steps": 46238, "tokens": 12408963072, "loss": 3.0687973083871785, "ce_loss": 3.0687973083871785, "z_loss": 19.10368364507502, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1412234902381897, "tokens_per_second": 4304185.298384907, "eta_seconds": 2925.2060474079685, "mean_data_wait_seconds": 7.63237476348877e-05}
{"step": 22960, "total_steps": 46238, "tokens": 12414369792, "loss": 3.120415418437033, "ce_loss": 3.120415418437033, "z_loss": 19.255217479937006, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16083060204982758, "tokens_per_second": 4304178.593533004, "eta_seconds": 2923.9544481051976, "mean_data_wait_seconds": 0.00010582804679870605}
{"step": 22970, "total_steps": 46238, "tokens": 12419776512, "loss": 3.066165978258306, "ce_loss": 3.066165978258306, "z_loss": 19.538398699326947, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13780993223190308, "tokens_per_second": 4303298.129580183, "eta_seconds": 2923.2962823394364, "mean_data_wait_seconds": 0.0003072023391723633}
{"step": 22980, "total_steps": 46238, "tokens": 12425183232, "loss": 3.1291001186226355, "ce_loss": 3.1291001186226355, "z_loss": 19.236223408670135, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15486375987529755, "tokens_per_second": 4303284.287605293, "eta_seconds": 2922.0492683269717, "mean_data_wait_seconds": 6.22868537902832e-05}
{"step": 22990, "total_steps": 46238, "tokens": 12430589952, "loss": 3.0937295826998623, "ce_loss": 3.0937295826998623, "z_loss": 19.249293500726875, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12293906509876251, "tokens_per_second": 4303398.863642894, "eta_seconds": 2920.715086437548, "mean_data_wait_seconds": 0.00010731816291809082}
{"step": 23000, "total_steps": 46238, "tokens": 12435996672, "loss": 3.0991531664674934, "ce_loss": 3.0991531664674934, "z_loss": 19.557026097268768, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13224785029888153, "tokens_per_second": 4303301.582247602, "eta_seconds": 2919.5247007155076, "mean_data_wait_seconds": 7.146596908569336e-05}
{"step": 23010, "total_steps": 46238, "tokens": 12441403392, "loss": 3.044697384039561, "ce_loss": 3.044697384039561, "z_loss": 19.486756671558727, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16193653643131256, "tokens_per_second": 4303273.549195344, "eta_seconds": 2918.287299292934, "mean_data_wait_seconds": 9.381771087646484e-05}
{"step": 23020, "total_steps": 46238, "tokens": 12446810112, "loss": 3.049720509485765, "ce_loss": 3.049720509485765, "z_loss": 19.690836039456453, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1261831670999527, "tokens_per_second": 4303295.463070155, "eta_seconds": 2917.0160245153857, "mean_data_wait_seconds": 9.128451347351074e-05}
{"step": 23030, "total_steps": 46238, "tokens": 12452216832, "loss": 3.093645894166195, "ce_loss": 3.093645894166195, "z_loss": 19.47070494565097, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12898671627044678, "tokens_per_second": 4303317.088665914, "eta_seconds": 2915.7449580109505, "mean_data_wait_seconds": 0.00010249018669128418}
{"step": 23040, "total_steps": 46238, "tokens": 12457623552, "loss": 3.063504813295422, "ce_loss": 3.063504813295422, "z_loss": 19.485526258295234, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1386585831642151, "tokens_per_second": 4303403.070347063, "eta_seconds": 2914.430319209794, "mean_data_wait_seconds": 0.00011089444160461426}
{"step": 23050, "total_steps": 46238, "tokens": 12463030272, "loss": 3.051253602360234, "ce_loss": 3.051253602360234, "z_loss": 19.166276787266586, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14439910650253296, "tokens_per_second": 4303344.732265907, "eta_seconds": 2913.21342908053, "mean_data_wait_seconds": 0.00011241436004638672}
{"step": 23060, "total_steps": 46238, "tokens": 12468436992, "loss": 3.0740630283500208, "ce_loss": 3.0740630283500208, "z_loss": 19.235851605733234, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1369461864233017, "tokens_per_second": 4303542.570956141, "eta_seconds": 2911.8231636816104, "mean_data_wait_seconds": 7.575750350952148e-05}
{"step": 23070, "total_steps": 46238, "tokens": 12473843712, "loss": 3.085005449526238, "ce_loss": 3.085005449526238, "z_loss": 19.158017418601297, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15713219344615936, "tokens_per_second": 4303664.7777322475, "eta_seconds": 2910.4841735838586, "mean_data_wait_seconds": 6.198883056640625e-05}
{"step": 23080, "total_steps": 46238, "tokens": 12479250432, "loss": 3.111167625947432, "ce_loss": 3.111167625947432, "z_loss": 19.559132084702, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1406235247850418, "tokens_per_second": 4303760.372843618, "eta_seconds": 2909.1632477965895, "mean_data_wait_seconds": 7.11679458618164e-05}
{"step": 23090, "total_steps": 46238, "tokens": 12484657152, "loss": 3.092702601895188, "ce_loss": 3.092702601895188, "z_loss": 19.030942830172453, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14817211031913757, "tokens_per_second": 4303807.912403169, "eta_seconds": 2907.874849138396, "mean_data_wait_seconds": 9.199976921081543e-05}
{"step": 23100, "total_steps": 46238, "tokens": 12490063872, "loss": 3.0206503850041013, "ce_loss": 3.0206503850041013, "z_loss": 19.18980292117957, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13138432800769806, "tokens_per_second": 4303872.087576833, "eta_seconds": 2906.5752442106423, "mean_data_wait_seconds": 9.784102439880371e-05}
{"step": 23110, "total_steps": 46238, "tokens": 12495470592, "loss": 3.103205054095297, "ce_loss": 3.103205054095297, "z_loss": 20.158759637312457, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16108864545822144, "tokens_per_second": 4304024.853343232, "eta_seconds": 2905.2158781767234, "mean_data_wait_seconds": 0.00011721253395080566}
{"step": 23120, "total_steps": 46238, "tokens": 12500877312, "loss": 3.103450890743371, "ce_loss": 3.103450890743371, "z_loss": 19.00101494066643, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13593877851963043, "tokens_per_second": 4304131.369705788, "eta_seconds": 2903.887811596782, "mean_data_wait_seconds": 0.00013244152069091797}
{"step": 23130, "total_steps": 46238, "tokens": 12506284032, "loss": 3.0415051182111106, "ce_loss": 3.0415051182111106, "z_loss": 19.77251367857962, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13629858195781708, "tokens_per_second": 4304066.461774358, "eta_seconds": 2902.6754152048143, "mean_data_wait_seconds": 0.00010284781455993652}
{"step": 23140, "total_steps": 46238, "tokens": 12511690752, "loss": 3.086759433601842, "ce_loss": 3.086759433601842, "z_loss": 19.56539096254291, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12625281512737274, "tokens_per_second": 4304200.770126273, "eta_seconds": 2901.328690490811, "mean_data_wait_seconds": 8.034706115722656e-05}
{"step": 23150, "total_steps": 46238, "tokens": 12517097472, "loss": 3.062978365204551, "ce_loss": 3.062978365204551, "z_loss": 19.6625294974356, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1384977400302887, "tokens_per_second": 4304418.385056194, "eta_seconds": 2899.9259243329902, "mean_data_wait_seconds": 6.54757022857666e-05}
{"step": 23160, "total_steps": 46238, "tokens": 12522504192, "loss": 3.0614316499594487, "ce_loss": 3.0614316499594487, "z_loss": 19.60001464323564, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14115485548973083, "tokens_per_second": 4304544.130260257, "eta_seconds": 2898.5851617336357, "mean_data_wait_seconds": 7.638335227966309e-05}
{"step": 23170, "total_steps": 46238, "tokens": 12527910912, "loss": 3.030085715380582, "ce_loss": 3.030085715380582, "z_loss": 19.548381343032375, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1442224532365799, "tokens_per_second": 4304698.03493627, "eta_seconds": 2897.2255249454774, "mean_data_wait_seconds": 6.747245788574219e-05}
{"step": 23180, "total_steps": 46238, "tokens": 12533317632, "loss": 3.0978478662895435, "ce_loss": 3.0978478662895435, "z_loss": 19.80810211644028, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.161993145942688, "tokens_per_second": 4304854.776919234, "eta_seconds": 2895.8640767253664, "mean_data_wait_seconds": 5.939602851867676e-05}
{"step": 23190, "total_steps": 46238, "tokens": 12538724352, "loss": 3.050324347886172, "ce_loss": 3.050324347886172, "z_loss": 19.78000175591671, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13985306024551392, "tokens_per_second": 4304962.667630652, "eta_seconds": 2894.5355734892264, "mean_data_wait_seconds": 7.721781730651855e-05}
{"step": 23200, "total_steps": 46238, "tokens": 12544131072, "loss": 3.1110704974694685, "ce_loss": 3.1110704974694685, "z_loss": 20.00368204983798, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1316402554512024, "tokens_per_second": 4304833.147625722, "eta_seconds": 2893.3666966557475, "mean_data_wait_seconds": 8.192658424377441e-05}
{"step": 23210, "total_steps": 46238, "tokens": 12549537792, "loss": 3.0275991414532517, "ce_loss": 3.0275991414532517, "z_loss": 19.972344412948146, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13374873995780945, "tokens_per_second": 4304942.191182345, "eta_seconds": 2892.037474858777, "mean_data_wait_seconds": 7.081031799316406e-05}
{"step": 23220, "total_steps": 46238, "tokens": 12554944512, "loss": 3.0612540768854544, "ce_loss": 3.0612540768854544, "z_loss": 19.78933942679203, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.132558673620224, "tokens_per_second": 4304922.748317624, "eta_seconds": 2890.7945976181345, "mean_data_wait_seconds": 6.660819053649902e-05}
{"step": 23230, "total_steps": 46238, "tokens": 12560351232, "loss": 3.048684786666523, "ce_loss": 3.048684786666523, "z_loss": 20.150365684971664, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13994956016540527, "tokens_per_second": 4304960.87917418, "eta_seconds": 2889.5130648402587, "mean_data_wait_seconds": 8.356571197509766e-05}
{"step": 23240, "total_steps": 46238, "tokens": 12565757952, "loss": 3.1239031264276216, "ce_loss": 3.1239031264276216, "z_loss": 20.24154988202182, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15019361674785614, "tokens_per_second": 4305180.737062452, "eta_seconds": 2888.1096389193544, "mean_data_wait_seconds": 7.56680965423584e-05}
{"step": 23250, "total_steps": 46238, "tokens": 12571164672, "loss": 3.0796743215936604, "ce_loss": 3.0796743215936604, "z_loss": 20.368212049657647, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1361832469701767, "tokens_per_second": 4305246.678576441, "eta_seconds": 2886.8095586359163, "mean_data_wait_seconds": 6.404519081115723e-05}
{"step": 23260, "total_steps": 46238, "tokens": 12576571392, "loss": 3.0959476413148823, "ce_loss": 3.0959476413148823, "z_loss": 19.906835050293893, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.164918452501297, "tokens_per_second": 4305451.320443963, "eta_seconds": 2885.41656132788, "mean_data_wait_seconds": 7.12275505065918e-05}
{"step": 23270, "total_steps": 46238, "tokens": 12581978112, "loss": 3.07869851769823, "ce_loss": 3.07869851769823, "z_loss": 19.976687546932336, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13339272141456604, "tokens_per_second": 4305524.041911757, "eta_seconds": 2884.112062346371, "mean_data_wait_seconds": 6.67572021484375e-05}
{"step": 23280, "total_steps": 46238, "tokens": 12587384832, "loss": 3.034616383639249, "ce_loss": 3.034616383639249, "z_loss": 19.904083093007404, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13333068788051605, "tokens_per_second": 4305682.50504859, "eta_seconds": 2882.7502003332056, "mean_data_wait_seconds": 0.004065722227096558}
{"step": 23290, "total_steps": 46238, "tokens": 12592791552, "loss": 3.0821599996451177, "ce_loss": 3.0821599996451177, "z_loss": 20.660376577666312, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12787121534347534, "tokens_per_second": 4305728.064481608, "eta_seconds": 2881.4639935914597, "mean_data_wait_seconds": 7.423758506774902e-05}
{"step": 23300, "total_steps": 46238, "tokens": 12598198272, "loss": 3.0503007769584656, "ce_loss": 3.0503007769584656, "z_loss": 20.047808531558875, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12493855506181717, "tokens_per_second": 4305925.832177929, "eta_seconds": 2880.0760039397614, "mean_data_wait_seconds": 0.00013682246208190918}
{"step": 23310, "total_steps": 46238, "tokens": 12603604992, "loss": 3.084351402340513, "ce_loss": 3.084351402340513, "z_loss": 20.699108889608674, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.21148312091827393, "tokens_per_second": 4306022.7213969, "eta_seconds": 2878.7555816655486, "mean_data_wait_seconds": 0.00014898180961608887}
{"step": 23320, "total_steps": 46238, "tokens": 12609011712, "loss": 3.064645741925095, "ce_loss": 3.064645741925095, "z_loss": 19.927475755864922, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12843342125415802, "tokens_per_second": 4306119.888104277, "eta_seconds": 2877.435033387986, "mean_data_wait_seconds": 9.906291961669922e-05}
{"step": 23330, "total_steps": 46238, "tokens": 12614418432, "loss": 3.0516544237281336, "ce_loss": 3.0516544237281336, "z_loss": 20.421060880025227, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13621479272842407, "tokens_per_second": 4306228.00136185, "eta_seconds": 2876.107233542482, "mean_data_wait_seconds": 7.590651512145996e-05}
{"step": 23340, "total_steps": 46238, "tokens": 12619825152, "loss": 3.0515761519923355, "ce_loss": 3.0515761519923355, "z_loss": 20.209502046758477, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14717891812324524, "tokens_per_second": 4306388.114974073, "eta_seconds": 2874.744786925582, "mean_data_wait_seconds": 8.612871170043945e-05}
{"step": 23350, "total_steps": 46238, "tokens": 12625231872, "loss": 3.0471368359796926, "ce_loss": 3.0471368359796926, "z_loss": 20.03062174536965, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1361619383096695, "tokens_per_second": 4306629.897687443, "eta_seconds": 2873.32795201295, "mean_data_wait_seconds": 5.716085433959961e-05}
{"step": 23360, "total_steps": 46238, "tokens": 12630638592, "loss": 3.074026907935287, "ce_loss": 3.074026907935287, "z_loss": 20.106243292490642, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12027285248041153, "tokens_per_second": 4306797.200875919, "eta_seconds": 2871.96094152852, "mean_data_wait_seconds": 0.0004527568817138672}
{"step": 23370, "total_steps": 46238, "tokens": 12636045312, "loss": 3.0902574820951982, "ce_loss": 3.0902574820951982, "z_loss": 20.615480610818572, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17013798654079437, "tokens_per_second": 4306684.648131189, "eta_seconds": 2870.7805734893886, "mean_data_wait_seconds": 8.761882781982422e-05}
{"step": 23380, "total_steps": 46238, "tokens": 12641452032, "loss": 3.0502477541114343, "ce_loss": 3.0502477541114343, "z_loss": 20.682806304006867, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13293859362602234, "tokens_per_second": 4306754.513723865, "eta_seconds": 2869.478598006843, "mean_data_wait_seconds": 6.80088996887207e-05}
{"step": 23390, "total_steps": 46238, "tokens": 12646858752, "loss": 3.0169744058088823, "ce_loss": 3.0169744058088823, "z_loss": 20.92933381687511, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13981182873249054, "tokens_per_second": 4306672.2195801, "eta_seconds": 2868.2780007818637, "mean_data_wait_seconds": 7.346272468566895e-05}
{"step": 23400, "total_steps": 46238, "tokens": 12652265472, "loss": 3.052927093072371, "ce_loss": 3.052927093072371, "z_loss": 20.184423345507998, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15952666103839874, "tokens_per_second": 4306808.808560912, "eta_seconds": 2866.931645411436, "mean_data_wait_seconds": 0.00010892748832702637}
{"step": 23410, "total_steps": 46238, "tokens": 12657672192, "loss": 3.0353084098209036, "ce_loss": 3.0353084098209036, "z_loss": 20.087143898010254, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1601443737745285, "tokens_per_second": 4306908.867487446, "eta_seconds": 2865.6096805688853, "mean_data_wait_seconds": 8.359551429748535e-05}
{"step": 23420, "total_steps": 46238, "tokens": 12663078912, "loss": 3.0642219077457082, "ce_loss": 3.0642219077457082, "z_loss": 20.546577150171455, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1236247792840004, "tokens_per_second": 4307019.860691452, "eta_seconds": 2864.2805055511135, "mean_data_wait_seconds": 6.496906280517578e-05}
{"step": 23430, "total_steps": 46238, "tokens": 12668485632, "loss": 3.056300685261235, "ce_loss": 3.056300685261235, "z_loss": 21.06307890920928, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15617452561855316, "tokens_per_second": 4307106.27678503, "eta_seconds": 2862.967735545257, "mean_data_wait_seconds": 0.00012132525444030762}
{"step": 23440, "total_steps": 46238, "tokens": 12673892352, "loss": 3.081148451024836, "ce_loss": 3.081148451024836, "z_loss": 20.408402399583295, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14762678742408752, "tokens_per_second": 4307106.650212913, "eta_seconds": 2861.712185230125, "mean_data_wait_seconds": 0.00013139843940734863}
{"step": 23450, "total_steps": 46238, "tokens": 12679299072, "loss": 3.0510709249612056, "ce_loss": 3.0510709249612056, "z_loss": 21.157745173483185, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14721977710723877, "tokens_per_second": 4307213.472003193, "eta_seconds": 2860.385941881375, "mean_data_wait_seconds": 7.241964340209961e-05}
{"step": 23460, "total_steps": 46238, "tokens": 12684705792, "loss": 3.0000159921068135, "ce_loss": 3.0000159921068135, "z_loss": 20.76344361449733, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12116490304470062, "tokens_per_second": 4307179.166785292, "eta_seconds": 2859.15344292291, "mean_data_wait_seconds": 9.635090827941895e-05}
{"step": 23470, "total_steps": 46238, "tokens": 12690112512, "loss": 3.0579720572991804, "ce_loss": 3.0579720572991804, "z_loss": 20.7460285244566, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1543359011411667, "tokens_per_second": 4307288.432885563, "eta_seconds": 2857.8256635935486, "mean_data_wait_seconds": 6.753206253051758e-05}
{"step": 23480, "total_steps": 46238, "tokens": 12695519232, "loss": 3.0821100798520176, "ce_loss": 3.0821100798520176, "z_loss": 20.486318906148274, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1549915373325348, "tokens_per_second": 4307358.803312718, "eta_seconds": 2856.523745952425, "mean_data_wait_seconds": 0.0001379549503326416}
{"step": 23490, "total_steps": 46238, "tokens": 12700925952, "loss": 3.0722747336734426, "ce_loss": 3.0722747336734426, "z_loss": 20.641537883064963, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13562020659446716, "tokens_per_second": 4307443.012939008, "eta_seconds": 2855.212697430095, "mean_data_wait_seconds": 0.00010699033737182617}
{"step": 23500, "total_steps": 46238, "tokens": 12706332672, "loss": 3.080124753894228, "ce_loss": 3.080124753894228, "z_loss": 20.812094327175256, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13774526119232178, "tokens_per_second": 4307480.8692180095, "eta_seconds": 2853.9324113659377, "mean_data_wait_seconds": 9.66787338256836e-05}
{"step": 23510, "total_steps": 46238, "tokens": 12711739392, "loss": 3.020983822417982, "ce_loss": 3.020983822417982, "z_loss": 20.553671836853027, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15710535645484924, "tokens_per_second": 4307477.249366717, "eta_seconds": 2852.6796156164382, "mean_data_wait_seconds": 0.0001302957534790039}
{"step": 23520, "total_steps": 46238, "tokens": 12717146112, "loss": 3.091134734225996, "ce_loss": 3.091134734225996, "z_loss": 20.813431494163744, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2163083702325821, "tokens_per_second": 4307666.5570068825, "eta_seconds": 2851.299110888999, "mean_data_wait_seconds": 7.262825965881348e-05}
{"step": 23530, "total_steps": 46238, "tokens": 12722552832, "loss": 3.074288263465419, "ce_loss": 3.074288263465419, "z_loss": 20.310810681545373, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13739824295043945, "tokens_per_second": 4307716.690812092, "eta_seconds": 2850.010802749781, "mean_data_wait_seconds": 0.00012168288230895996}
{"step": 23540, "total_steps": 46238, "tokens": 12727959552, "loss": 3.034687775554079, "ce_loss": 3.034687775554079, "z_loss": 20.97440585223111, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12949880957603455, "tokens_per_second": 4307873.350312551, "eta_seconds": 2848.652080987862, "mean_data_wait_seconds": 7.063150405883789e-05}
{"step": 23550, "total_steps": 46238, "tokens": 12733366272, "loss": 3.0629931579936636, "ce_loss": 3.0629931579936636, "z_loss": 20.288735230763752, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1439313292503357, "tokens_per_second": 4307929.09791352, "eta_seconds": 2847.360155008345, "mean_data_wait_seconds": 9.068846702575684e-05}
{"step": 23560, "total_steps": 46238, "tokens": 12738772992, "loss": 3.0597276019327566, "ce_loss": 3.0597276019327566, "z_loss": 20.65486345869122, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14764592051506042, "tokens_per_second": 4307911.213112561, "eta_seconds": 2846.116908510119, "mean_data_wait_seconds": 8.52048397064209e-05}
{"step": 23570, "total_steps": 46238, "tokens": 12744179712, "loss": 3.0709336851582383, "ce_loss": 3.0709336851582383, "z_loss": 20.342804504163336, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16101029515266418, "tokens_per_second": 4307868.661168766, "eta_seconds": 2844.889941624866, "mean_data_wait_seconds": 0.00013947486877441406}
{"step": 23580, "total_steps": 46238, "tokens": 12749586432, "loss": 3.0452379656560495, "ce_loss": 3.0452379656560495, "z_loss": 20.986343528285172, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13091054558753967, "tokens_per_second": 4307843.948940047, "eta_seconds": 2843.651174275739, "mean_data_wait_seconds": 7.089972496032715e-05}
{"step": 23590, "total_steps": 46238, "tokens": 12754993152, "loss": 3.091262703592127, "ce_loss": 3.091262703592127, "z_loss": 20.5361027284102, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14169853925704956, "tokens_per_second": 4307972.413850866, "eta_seconds": 2842.3113260037426, "mean_data_wait_seconds": 7.18235969543457e-05}
{"step": 23600, "total_steps": 46238, "tokens": 12760399872, "loss": 3.0266928564418447, "ce_loss": 3.0266928564418447, "z_loss": 20.799565026254363, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15553335845470428, "tokens_per_second": 4308049.122249812, "eta_seconds": 2841.0056888135646, "mean_data_wait_seconds": 8.490681648254395e-05}
{"step": 23610, "total_steps": 46238, "tokens": 12765806592, "loss": 3.0331417430530894, "ce_loss": 3.0331417430530894, "z_loss": 20.36058576179273, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1262650489807129, "tokens_per_second": 4308206.184352497, "eta_seconds": 2839.6471339819777, "mean_data_wait_seconds": 0.00012120604515075684}
{"step": 23620, "total_steps": 46238, "tokens": 12771213312, "loss": 3.070237401759986, "ce_loss": 3.070237401759986, "z_loss": 20.792586745637834, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12531816959381104, "tokens_per_second": 4308215.962496837, "eta_seconds": 2838.3857101056315, "mean_data_wait_seconds": 8.621811866760254e-05}
{"step": 23630, "total_steps": 46238, "tokens": 12776620032, "loss": 3.009948887608268, "ce_loss": 3.009948887608268, "z_loss": 21.029709671482895, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14098864793777466, "tokens_per_second": 4308276.440712144, "eta_seconds": 2837.090904496273, "mean_data_wait_seconds": 6.553530693054199e-05}
{"step": 23640, "total_steps": 46238, "tokens": 12782026752, "loss": 3.046532898238211, "ce_loss": 3.046532898238211, "z_loss": 20.99488557468761, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15154887735843658, "tokens_per_second": 4308331.251257453, "eta_seconds": 2835.799865768008, "mean_data_wait_seconds": 0.00011599063873291016}
{"step": 23650, "total_steps": 46238, "tokens": 12787433472, "loss": 3.0588048100471497, "ce_loss": 3.0588048100471497, "z_loss": 20.933960047635164, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1614185869693756, "tokens_per_second": 4308347.274428501, "eta_seconds": 2834.534378526841, "mean_data_wait_seconds": 0.00014725327491760254}
{"step": 23660, "total_steps": 46238, "tokens": 12792840192, "loss": 3.047629891019879, "ce_loss": 3.047629891019879, "z_loss": 21.73634211222331, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.180001363158226, "tokens_per_second": 4308500.297113454, "eta_seconds": 2833.1788098466886, "mean_data_wait_seconds": 7.319450378417969e-05}
{"step": 23670, "total_steps": 46238, "tokens": 12798246912, "loss": 3.089791400866075, "ce_loss": 3.089791400866075, "z_loss": 20.66947911002419, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16535212099552155, "tokens_per_second": 4308500.622703034, "eta_seconds": 2831.923699791695, "mean_data_wait_seconds": 0.00012621283531188965}
{"step": 23680, "total_steps": 46238, "tokens": 12803653632, "loss": 3.035750074820085, "ce_loss": 3.035750074820085, "z_loss": 20.83399753859549, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13532540202140808, "tokens_per_second": 4308514.3509715, "eta_seconds": 2830.6597844451912, "mean_data_wait_seconds": 7.990002632141113e-05}
{"step": 23690, "total_steps": 46238, "tokens": 12809060352, "loss": 3.023588724208601, "ce_loss": 3.023588724208601, "z_loss": 20.43635167497577, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.318042129278183, "tokens_per_second": 4308667.48471601, "eta_seconds": 2829.3043330085366, "mean_data_wait_seconds": 9.02712345123291e-05}
{"step": 23700, "total_steps": 46238, "tokens": 12814467072, "loss": 3.1022415829427317, "ce_loss": 3.1022415829427317, "z_loss": 21.506032611384537, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16589097678661346, "tokens_per_second": 4308885.654511312, "eta_seconds": 2827.9062943437434, "mean_data_wait_seconds": 7.778406143188477e-05}
{"step": 23710, "total_steps": 46238, "tokens": 12819873792, "loss": 2.979147670846997, "ce_loss": 2.979147670846997, "z_loss": 20.7731658039671, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16831138730049133, "tokens_per_second": 4309035.950016765, "eta_seconds": 2826.5529193258676, "mean_data_wait_seconds": 7.170438766479492e-05}
{"step": 23720, "total_steps": 46238, "tokens": 12825280512, "loss": 2.9934226003560154, "ce_loss": 2.9934226003560154, "z_loss": 20.93172719261863, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1375035047531128, "tokens_per_second": 4309189.707217508, "eta_seconds": 2825.1973691501944, "mean_data_wait_seconds": 8.881092071533203e-05}
{"step": 23730, "total_steps": 46238, "tokens": 12830687232, "loss": 3.074738466378414, "ce_loss": 3.074738466378414, "z_loss": 21.01230554869681, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.11725173890590668, "tokens_per_second": 4309245.3424908165, "eta_seconds": 2823.9062148562116, "mean_data_wait_seconds": 0.00012320280075073242}
{"step": 23740, "total_steps": 46238, "tokens": 12836093952, "loss": 3.1066924008456143, "ce_loss": 3.1066924008456143, "z_loss": 20.750729965441156, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12459222972393036, "tokens_per_second": 4309302.534838466, "eta_seconds": 2822.6140740095307, "mean_data_wait_seconds": 0.00013780593872070312}
{"step": 23750, "total_steps": 46238, "tokens": 12841500672, "loss": 3.0832345449563228, "ce_loss": 3.0832345449563228, "z_loss": 21.119056138125334, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13728941977024078, "tokens_per_second": 4309392.017243126, "eta_seconds": 2821.3008274373633, "mean_data_wait_seconds": 0.00016874074935913086}
{"step": 23760, "total_steps": 46238, "tokens": 12846907392, "loss": 3.0401944626461375, "ce_loss": 3.0401944626461375, "z_loss": 20.676362427798185, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14058050513267517, "tokens_per_second": 4309639.563741421, "eta_seconds": 2819.8842070796345, "mean_data_wait_seconds": 8.657574653625488e-05}
{"step": 23770, "total_steps": 46238, "tokens": 12852314112, "loss": 3.0785840081446096, "ce_loss": 3.0785840081446096, "z_loss": 20.597418944040935, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16171124577522278, "tokens_per_second": 4309775.030458953, "eta_seconds": 2818.541046377175, "mean_data_wait_seconds": 8.270144462585449e-05}
{"step": 23780, "total_steps": 46238, "tokens": 12857720832, "loss": 3.0052905154950693, "ce_loss": 3.0052905154950693, "z_loss": 20.190044446425006, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1422456055879593, "tokens_per_second": 4309921.763788818, "eta_seconds": 2817.190605642497, "mean_data_wait_seconds": 6.124377250671387e-05}
{"step": 23790, "total_steps": 46238, "tokens": 12863127552, "loss": 3.086978281989242, "ce_loss": 3.086978281989242, "z_loss": 20.76213651714903, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15639275312423706, "tokens_per_second": 4309912.20118034, "eta_seconds": 2815.9423713263186, "mean_data_wait_seconds": 0.00017699599266052246}
{"step": 23800, "total_steps": 46238, "tokens": 12868534272, "loss": 3.0947791356028933, "ce_loss": 3.0947791356028933, "z_loss": 20.846773638869777, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15805195271968842, "tokens_per_second": 4310050.026074736, "eta_seconds": 2814.597879516503, "mean_data_wait_seconds": 8.51750373840332e-05}
{"step": 23810, "total_steps": 46238, "tokens": 12873940992, "loss": 3.043073791446108, "ce_loss": 3.043073791446108, "z_loss": 20.936314207134824, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13790419697761536, "tokens_per_second": 4309980.264610746, "eta_seconds": 2813.3889715374653, "mean_data_wait_seconds": 6.341934204101562e-05}
{"step": 23820, "total_steps": 46238, "tokens": 12879347712, "loss": 3.0345795281005628, "ce_loss": 3.0345795281005628, "z_loss": 21.3045899217779, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.156361386179924, "tokens_per_second": 4310133.232260147, "eta_seconds": 2812.0347030767743, "mean_data_wait_seconds": 6.964802742004395e-05}
{"step": 23830, "total_steps": 46238, "tokens": 12884754432, "loss": 3.0522274122093664, "ce_loss": 3.0522274122093664, "z_loss": 20.915842778754957, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14319337904453278, "tokens_per_second": 4310227.613325499, "eta_seconds": 2810.7187347939052, "mean_data_wait_seconds": 0.0001016855239868164}
{"step": 23840, "total_steps": 46238, "tokens": 12890161152, "loss": 3.089680314064026, "ce_loss": 3.089680314064026, "z_loss": 21.458348462075897, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13878850638866425, "tokens_per_second": 4310309.235359434, "eta_seconds": 2809.4111403100296, "mean_data_wait_seconds": 7.94529914855957e-05}
{"step": 23850, "total_steps": 46238, "tokens": 12895567872, "loss": 3.0021979104388845, "ce_loss": 3.0021979104388845, "z_loss": 21.060769615751326, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14200642704963684, "tokens_per_second": 4310377.534507359, "eta_seconds": 2808.1122748760313, "mean_data_wait_seconds": 0.00013190507888793945}
{"step": 23860, "total_steps": 46238, "tokens": 12900974592, "loss": 2.9897408124172324, "ce_loss": 2.9897408124172324, "z_loss": 21.679401094263252, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1301816999912262, "tokens_per_second": 4310428.962448273, "eta_seconds": 2806.824436593459, "mean_data_wait_seconds": 8.001923561096191e-05}
{"step": 23870, "total_steps": 46238, "tokens": 12906381312, "loss": 3.1250366857557585, "ce_loss": 3.1250366857557585, "z_loss": 21.454065120581426, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13887199759483337, "tokens_per_second": 4310604.622793521, "eta_seconds": 2805.4557729683174, "mean_data_wait_seconds": 7.715821266174316e-05}
{"step": 23880, "total_steps": 46238, "tokens": 12911788032, "loss": 3.0146733269546973, "ce_loss": 3.0146733269546973, "z_loss": 21.768000862815164, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1297435462474823, "tokens_per_second": 4310762.767158952, "eta_seconds": 2804.098614771738, "mean_data_wait_seconds": 0.0001004636287689209}
{"step": 23890, "total_steps": 46238, "tokens": 12917194752, "loss": 3.0557882876107185, "ce_loss": 3.0557882876107185, "z_loss": 20.91129256739761, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1590702384710312, "tokens_per_second": 4310735.93188928, "eta_seconds": 2802.861825661543, "mean_data_wait_seconds": 8.39531421661377e-05}
{"step": 23900, "total_steps": 46238, "tokens": 12922601472, "loss": 3.014875417405909, "ce_loss": 3.014875417405909, "z_loss": 20.981705954580598, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14541660249233246, "tokens_per_second": 4310707.731262418, "eta_seconds": 2801.6259085287547, "mean_data_wait_seconds": 6.961822509765625e-05}
{"step": 23910, "total_steps": 46238, "tokens": 12928008192, "loss": 3.0064880179636404, "ce_loss": 3.0064880179636404, "z_loss": 21.58875458168261, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15016783773899078, "tokens_per_second": 4310716.367569246, "eta_seconds": 2800.366044682963, "mean_data_wait_seconds": 7.089972496032715e-05}
{"step": 23920, "total_steps": 46238, "tokens": 12933414912, "loss": 3.023566843885364, "ce_loss": 3.023566843885364, "z_loss": 21.552076946605336, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15020476281642914, "tokens_per_second": 4310747.013355909, "eta_seconds": 2799.091894424698, "mean_data_wait_seconds": 0.00011089444160461426}
{"step": 23930, "total_steps": 46238, "tokens": 12938821632, "loss": 2.968402876998439, "ce_loss": 2.968402876998439, "z_loss": 21.413941484509092, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17000363767147064, "tokens_per_second": 4310814.428123491, "eta_seconds": 2797.793898367851, "mean_data_wait_seconds": 9.050965309143066e-05}
{"step": 23940, "total_steps": 46238, "tokens": 12944228352, "loss": 3.060943634221048, "ce_loss": 3.060943634221048, "z_loss": 21.587776718717635, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13469038903713226, "tokens_per_second": 4310970.304349386, "eta_seconds": 2796.4385585855716, "mean_data_wait_seconds": 7.161498069763184e-05}
{"step": 23950, "total_steps": 46238, "tokens": 12949635072, "loss": 3.043872239011707, "ce_loss": 3.043872239011707, "z_loss": 21.88736778317076, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1413198709487915, "tokens_per_second": 4310980.40426204, "eta_seconds": 2795.17783288619, "mean_data_wait_seconds": 0.00010305643081665039}
{"step": 23960, "total_steps": 46238, "tokens": 12955041792, "loss": 3.013848889957775, "ce_loss": 3.013848889957775, "z_loss": 21.043596441095527, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15627378225326538, "tokens_per_second": 4310974.875996427, "eta_seconds": 2793.92724162329, "mean_data_wait_seconds": 6.362795829772949e-05}
{"step": 23970, "total_steps": 46238, "tokens": 12960448512, "loss": 3.064416686693827, "ce_loss": 3.064416686693827, "z_loss": 22.062697063792836, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16300149261951447, "tokens_per_second": 4311147.845545033, "eta_seconds": 2792.561019785199, "mean_data_wait_seconds": 6.276369094848633e-05}
{"step": 23980, "total_steps": 46238, "tokens": 12965855232, "loss": 3.0536360000119065, "ce_loss": 3.0536360000119065, "z_loss": 21.081169446309406, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12927423417568207, "tokens_per_second": 4311408.1915998245, "eta_seconds": 2791.138340240215, "mean_data_wait_seconds": 0.00013047456741333008}
{"step": 23990, "total_steps": 46238, "tokens": 12971261952, "loss": 3.0514449639753862, "ce_loss": 3.0514449639753862, "z_loss": 21.590623465451326, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13250616192817688, "tokens_per_second": 4311533.302999603, "eta_seconds": 2789.8033341483642, "mean_data_wait_seconds": 0.00011980533599853516}
{"step": 24000, "total_steps": 46238, "tokens": 12976668672, "loss": 3.0385832858808115, "ce_loss": 3.0385832858808115, "z_loss": 21.611926570083156, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16087359189987183, "tokens_per_second": 4311758.754093822, "eta_seconds": 2788.4035145947746, "mean_data_wait_seconds": 0.0005414783954620361}
{"step": 24010, "total_steps": 46238, "tokens": 12982075392, "loss": 3.0373496413230896, "ce_loss": 3.0373496413230896, "z_loss": 20.93281517606793, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17118193209171295, "tokens_per_second": 4311865.2873853175, "eta_seconds": 2787.080704760916, "mean_data_wait_seconds": 0.00023537874221801758}
{"step": 24020, "total_steps": 46238, "tokens": 12987482112, "loss": 3.0043069991198452, "ce_loss": 3.0043069991198452, "z_loss": 20.781564380183365, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.22980450093746185, "tokens_per_second": 4311927.215038366, "eta_seconds": 2785.7867781502246, "mean_data_wait_seconds": 0.00012499094009399414}
{"step": 24030, "total_steps": 46238, "tokens": 12992888832, "loss": 3.0243965513778455, "ce_loss": 3.0243965513778455, "z_loss": 21.195992253043435, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13782551884651184, "tokens_per_second": 4311923.147690741, "eta_seconds": 2784.5355060259394, "mean_data_wait_seconds": 6.327033042907715e-05}
{"step": 24040, "total_steps": 46238, "tokens": 12998295552, "loss": 3.0674765326760034, "ce_loss": 3.0674765326760034, "z_loss": 21.33345314950654, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16092389822006226, "tokens_per_second": 4312071.84172802, "eta_seconds": 2783.185629669518, "mean_data_wait_seconds": 6.952881813049316e-05}
{"step": 24050, "total_steps": 46238, "tokens": 13003702272, "loss": 3.075624285322247, "ce_loss": 3.075624285322247, "z_loss": 21.15579183173902, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12757371366024017, "tokens_per_second": 4312126.782453638, "eta_seconds": 2781.896328468857, "mean_data_wait_seconds": 9.760260581970215e-05}
{"step": 24060, "total_steps": 46238, "tokens": 13009108992, "loss": 3.063920320886554, "ce_loss": 3.063920320886554, "z_loss": 21.89861436323686, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.11943267285823822, "tokens_per_second": 4312234.129830904, "eta_seconds": 2780.5732673587886, "mean_data_wait_seconds": 6.428360939025879e-05}
{"step": 24070, "total_steps": 46238, "tokens": 13014515712, "loss": 3.070235841202013, "ce_loss": 3.070235841202013, "z_loss": 20.506970636772387, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14824464917182922, "tokens_per_second": 4312462.295977669, "eta_seconds": 2779.172408110965, "mean_data_wait_seconds": 9.897351264953613e-05}
{"step": 24080, "total_steps": 46238, "tokens": 13019922432, "loss": 3.0589099327723184, "ce_loss": 3.0589099327723184, "z_loss": 21.18218806295684, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1654336303472519, "tokens_per_second": 4312507.937751812, "eta_seconds": 2777.8892646503086, "mean_data_wait_seconds": 9.813904762268066e-05}
{"step": 24090, "total_steps": 46238, "tokens": 13025329152, "loss": 3.036664968187159, "ce_loss": 3.036664968187159, "z_loss": 22.225222630934283, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1885560005903244, "tokens_per_second": 4312592.317261464, "eta_seconds": 2776.581207565608, "mean_data_wait_seconds": 8.133053779602051e-05}
{"step": 24100, "total_steps": 46238, "tokens": 13030735872, "loss": 3.0385429190866873, "ce_loss": 3.0385429190866873, "z_loss": 21.55359719016335, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1659080535173416, "tokens_per_second": 4312646.33348265, "eta_seconds": 2775.292741042975, "mean_data_wait_seconds": 9.810924530029297e-05}
{"step": 24110, "total_steps": 46238, "tokens": 13036142592, "loss": 3.0988060091481064, "ce_loss": 3.0988060091481064, "z_loss": 21.642581780751545, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14098763465881348, "tokens_per_second": 4312750.1695633875, "eta_seconds": 2773.972262161235, "mean_data_wait_seconds": 9.271502494812012e-05}
{"step": 24120, "total_steps": 46238, "tokens": 13041549312, "loss": 3.038476080605478, "ce_loss": 3.038476080605478, "z_loss": 21.139179619875822, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16263392567634583, "tokens_per_second": 4312790.5347195165, "eta_seconds": 2772.692651714349, "mean_data_wait_seconds": 9.948015213012695e-05}
{"step": 24130, "total_steps": 46238, "tokens": 13046956032, "loss": 3.0207465327147283, "ce_loss": 3.0207465327147283, "z_loss": 21.713966860915676, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1732841581106186, "tokens_per_second": 4312869.0094539905, "eta_seconds": 2771.3885763280355, "mean_data_wait_seconds": 8.618831634521484e-05}
{"step": 24140, "total_steps": 46238, "tokens": 13052362752, "loss": 3.045048256715139, "ce_loss": 3.045048256715139, "z_loss": 21.849372415831596, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15816730260849, "tokens_per_second": 4312894.368570709, "eta_seconds": 2770.118663481041, "mean_data_wait_seconds": 0.00015240907669067383}
{"step": 24150, "total_steps": 46238, "tokens": 13057769472, "loss": 3.02076907230146, "ce_loss": 3.02076907230146, "z_loss": 21.759123267549455, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15382246673107147, "tokens_per_second": 4312987.56089186, "eta_seconds": 2768.8052180541445, "mean_data_wait_seconds": 0.00012803077697753906}
{"step": 24160, "total_steps": 46238, "tokens": 13063176192, "loss": 3.0576427470554006, "ce_loss": 3.0576427470554006, "z_loss": 21.583526105591744, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1462908536195755, "tokens_per_second": 4313062.612904808, "eta_seconds": 2767.503469178931, "mean_data_wait_seconds": 0.00011974573135375977}
{"step": 24170, "total_steps": 46238, "tokens": 13068582912, "loss": 2.990974258292805, "ce_loss": 2.990974258292805, "z_loss": 22.550330118699506, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17890453338623047, "tokens_per_second": 4313202.508169473, "eta_seconds": 2766.160179449476, "mean_data_wait_seconds": 0.00013464689254760742}
{"step": 24180, "total_steps": 46238, "tokens": 13073989632, "loss": 3.044359319137804, "ce_loss": 3.044359319137804, "z_loss": 21.248859232122246, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16776080429553986, "tokens_per_second": 4313277.678938978, "eta_seconds": 2764.858465345448, "mean_data_wait_seconds": 0.00012874603271484375}
{"step": 24190, "total_steps": 46238, "tokens": 13079396352, "loss": 3.013513498234026, "ce_loss": 3.013513498234026, "z_loss": 21.287346623160623, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13382156193256378, "tokens_per_second": 4313468.892670188, "eta_seconds": 2763.482450112439, "mean_data_wait_seconds": 0.00021529197692871094}
{"step": 24200, "total_steps": 46238, "tokens": 13084803072, "loss": 3.004488291162433, "ce_loss": 3.004488291162433, "z_loss": 22.371312329263397, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14783045649528503, "tokens_per_second": 4313538.959118859, "eta_seconds": 2762.184131619359, "mean_data_wait_seconds": 7.30752944946289e-05}
{"step": 24210, "total_steps": 46238, "tokens": 13090209792, "loss": 3.0239006678263345, "ce_loss": 3.0239006678263345, "z_loss": 22.055369261539344, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1804289072751999, "tokens_per_second": 4313553.939980812, "eta_seconds": 2760.9211127780577, "mean_data_wait_seconds": 9.417533874511719e-05}
{"step": 24220, "total_steps": 46238, "tokens": 13095616512, "loss": 2.9802413763421955, "ce_loss": 2.9802413763421955, "z_loss": 21.66774745420976, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1322697252035141, "tokens_per_second": 4313678.876818638, "eta_seconds": 2759.587758831795, "mean_data_wait_seconds": 7.343292236328125e-05}
{"step": 24230, "total_steps": 46238, "tokens": 13101023232, "loss": 3.011353268767848, "ce_loss": 3.011353268767848, "z_loss": 22.171369726007637, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18339642882347107, "tokens_per_second": 4313890.922478199, "eta_seconds": 2758.198785694988, "mean_data_wait_seconds": 5.9038400650024414e-05}
{"step": 24240, "total_steps": 46238, "tokens": 13106429952, "loss": 3.0289856285759895, "ce_loss": 3.0289856285759895, "z_loss": 21.921082641139176, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14750617742538452, "tokens_per_second": 4313996.544088779, "eta_seconds": 2756.877958165385, "mean_data_wait_seconds": 7.578730583190918e-05}
{"step": 24250, "total_steps": 46238, "tokens": 13111836672, "loss": 2.994272941892797, "ce_loss": 2.994272941892797, "z_loss": 22.516975431731254, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12968194484710693, "tokens_per_second": 4314075.375393923, "eta_seconds": 2755.574307255704, "mean_data_wait_seconds": 7.703900337219238e-05}
{"step": 24260, "total_steps": 46238, "tokens": 13117243392, "loss": 3.066355849757339, "ce_loss": 3.066355849757339, "z_loss": 22.07386029850353, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16283918917179108, "tokens_per_second": 4314354.542875088, "eta_seconds": 2754.1428099883506, "mean_data_wait_seconds": 0.00011152029037475586}
{"step": 24270, "total_steps": 46238, "tokens": 13122650112, "loss": 3.051590682882251, "ce_loss": 3.051590682882251, "z_loss": 21.735758131200615, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15698428452014923, "tokens_per_second": 4314612.260047804, "eta_seconds": 2752.7251832053175, "mean_data_wait_seconds": 9.611248970031738e-05}
{"step": 24280, "total_steps": 46238, "tokens": 13128056832, "loss": 3.0341692675243723, "ce_loss": 3.0341692675243723, "z_loss": 22.009242751381613, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1348823755979538, "tokens_per_second": 4314858.540615486, "eta_seconds": 2751.31501815274, "mean_data_wait_seconds": 0.00017270445823669434}
{"step": 24290, "total_steps": 46238, "tokens": 13133463552, "loss": 3.0231038440357554, "ce_loss": 3.0231038440357554, "z_loss": 21.46821607242931, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15101948380470276, "tokens_per_second": 4314946.106679088, "eta_seconds": 2750.006162448348, "mean_data_wait_seconds": 6.80088996887207e-05}
{"step": 24300, "total_steps": 46238, "tokens": 13138870272, "loss": 3.029375430309411, "ce_loss": 3.029375430309411, "z_loss": 22.0929342472192, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12708555161952972, "tokens_per_second": 4315181.20906978, "eta_seconds": 2748.6033817237553, "mean_data_wait_seconds": 0.00011920928955078125}
{"step": 24310, "total_steps": 46238, "tokens": 13144276992, "loss": 3.0446560220284895, "ce_loss": 3.0446560220284895, "z_loss": 22.694986935817834, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16869701445102692, "tokens_per_second": 4315254.093897984, "eta_seconds": 2747.3040256804557, "mean_data_wait_seconds": 0.00011560320854187012}
{"step": 24320, "total_steps": 46238, "tokens": 13149683712, "loss": 3.053142016584223, "ce_loss": 3.053142016584223, "z_loss": 21.902205553921785, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14971815049648285, "tokens_per_second": 4314571.355734722, "eta_seconds": 2746.4856290416124, "mean_data_wait_seconds": 0.0001093745231628418}
{"step": 24330, "total_steps": 46238, "tokens": 13155090432, "loss": 3.039282516999678, "ce_loss": 3.039282516999678, "z_loss": 22.04370605584347, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16942930221557617, "tokens_per_second": 4314048.966953187, "eta_seconds": 2745.5649193442564, "mean_data_wait_seconds": 9.569525718688965e-05}
{"step": 24340, "total_steps": 46238, "tokens": 13160497152, "loss": 3.0481807734027053, "ce_loss": 3.0481807734027053, "z_loss": 22.372702251781117, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15763802826404572, "tokens_per_second": 4314132.585895344, "eta_seconds": 2744.258445534758, "mean_data_wait_seconds": 0.00014570355415344238}
{"step": 24350, "total_steps": 46238, "tokens": 13165903872, "loss": 2.981015281243758, "ce_loss": 2.981015281243758, "z_loss": 21.71934426914562, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1543019562959671, "tokens_per_second": 4314125.763924769, "eta_seconds": 2743.009525349192, "mean_data_wait_seconds": 0.00013121962547302246}
{"step": 24360, "total_steps": 46238, "tokens": 13171310592, "loss": 2.955452572215687, "ce_loss": 2.955452572215687, "z_loss": 22.40732127969915, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15046942234039307, "tokens_per_second": 4314281.99874242, "eta_seconds": 2741.656977325046, "mean_data_wait_seconds": 7.730722427368164e-05}
{"step": 24370, "total_steps": 46238, "tokens": 13176717312, "loss": 2.994891123338179, "ce_loss": 2.994891123338179, "z_loss": 22.268120621189926, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13262344896793365, "tokens_per_second": 4313702.781396151, "eta_seconds": 2740.771727479442, "mean_data_wait_seconds": 8.884072303771973e-05}
{"step": 24380, "total_steps": 46238, "tokens": 13182124032, "loss": 2.9884939609151897, "ce_loss": 2.9884939609151897, "z_loss": 22.371998338988334, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12686502933502197, "tokens_per_second": 4313817.350767469, "eta_seconds": 2739.4455868414457, "mean_data_wait_seconds": 8.594989776611328e-05}
{"step": 24390, "total_steps": 46238, "tokens": 13187530752, "loss": 3.019489949399775, "ce_loss": 3.019489949399775, "z_loss": 21.725118030201305, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1941843181848526, "tokens_per_second": 4313950.20364228, "eta_seconds": 2738.1079118685802, "mean_data_wait_seconds": 7.924437522888184e-05}
{"step": 24400, "total_steps": 46238, "tokens": 13192937472, "loss": 3.0145803274530354, "ce_loss": 3.0145803274530354, "z_loss": 22.296899506540008, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17688797414302826, "tokens_per_second": 4314107.449732392, "eta_seconds": 2736.7548447900567, "mean_data_wait_seconds": 0.00013506412506103516}
{"step": 24410, "total_steps": 46238, "tokens": 13198344192, "loss": 3.016831939870661, "ce_loss": 3.016831939870661, "z_loss": 22.015430190346457, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1688448041677475, "tokens_per_second": 4313525.23501887, "eta_seconds": 2735.8708019586616, "mean_data_wait_seconds": 9.655952453613281e-05}
{"step": 24420, "total_steps": 46238, "tokens": 13203750912, "loss": 3.013354427886732, "ce_loss": 3.013354427886732, "z_loss": 21.260804682066947, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1635541319847107, "tokens_per_second": 4313495.88408252, "eta_seconds": 2734.6359753184215, "mean_data_wait_seconds": 7.939338684082031e-05}
{"step": 24430, "total_steps": 46238, "tokens": 13209157632, "loss": 3.0344761339100925, "ce_loss": 3.0344761339100925, "z_loss": 22.089648925896846, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15148048102855682, "tokens_per_second": 4313401.8755790675, "eta_seconds": 2733.442105349192, "mean_data_wait_seconds": 7.465481758117676e-05}
{"step": 24440, "total_steps": 46238, "tokens": 13214564352, "loss": 3.010548279140935, "ce_loss": 3.010548279140935, "z_loss": 22.252755251797762, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14867079257965088, "tokens_per_second": 4313301.125345779, "eta_seconds": 2732.2524538685543, "mean_data_wait_seconds": 6.467103958129883e-05}
{"step": 24450, "total_steps": 46238, "tokens": 13219971072, "loss": 3.0211204633568274, "ce_loss": 3.0211204633568274, "z_loss": 22.274873473427512, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1605524867773056, "tokens_per_second": 4313431.425046141, "eta_seconds": 2730.9164568146557, "mean_data_wait_seconds": 7.18832015991211e-05}
{"step": 24460, "total_steps": 46238, "tokens": 13225377792, "loss": 3.019256541223237, "ce_loss": 3.019256541223237, "z_loss": 22.878284945632473, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14347848296165466, "tokens_per_second": 4313574.646226545, "eta_seconds": 2729.5723639093435, "mean_data_wait_seconds": 9.468197822570801e-05}
{"step": 24470, "total_steps": 46238, "tokens": 13230784512, "loss": 3.0246182517571882, "ce_loss": 3.0246182517571882, "z_loss": 22.433296593752775, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14685510098934174, "tokens_per_second": 4313554.924441909, "eta_seconds": 2728.331418087288, "mean_data_wait_seconds": 6.133317947387695e-05}
{"step": 24480, "total_steps": 46238, "tokens": 13236191232, "loss": 3.0408139012076636, "ce_loss": 3.0408139012076636, "z_loss": 22.203203909324877, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16272073984146118, "tokens_per_second": 4313713.255479397, "eta_seconds": 2726.977897535912, "mean_data_wait_seconds": 9.655952453613281e-05}
{"step": 24490, "total_steps": 46238, "tokens": 13241597952, "loss": 3.008113073580193, "ce_loss": 3.008113073580193, "z_loss": 22.278561895543877, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13719525933265686, "tokens_per_second": 4313812.229128575, "eta_seconds": 2725.661980511194, "mean_data_wait_seconds": 5.4389238357543945e-05}
{"step": 24500, "total_steps": 46238, "tokens": 13247004672, "loss": 3.023963308695591, "ce_loss": 3.023963308695591, "z_loss": 22.15044576471502, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1635449230670929, "tokens_per_second": 4313902.625135656, "eta_seconds": 2724.3515408812514, "mean_data_wait_seconds": 6.398558616638184e-05}
{"step": 24510, "total_steps": 46238, "tokens": 13252411392, "loss": 3.0375549558437234, "ce_loss": 3.0375549558437234, "z_loss": 22.001363075140752, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17123684287071228, "tokens_per_second": 4314026.521922849, "eta_seconds": 2723.0200102627196, "mean_data_wait_seconds": 6.625056266784668e-05}
{"step": 24520, "total_steps": 46238, "tokens": 13257818112, "loss": 3.0499838771242085, "ce_loss": 3.0499838771242085, "z_loss": 22.034948681340072, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.143620103597641, "tokens_per_second": 4314082.105499116, "eta_seconds": 2721.7316538859754, "mean_data_wait_seconds": 9.959936141967773e-05}
{"step": 24530, "total_steps": 46238, "tokens": 13263224832, "loss": 3.0704684727119678, "ce_loss": 3.0704684727119678, "z_loss": 22.623380169723973, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14460952579975128, "tokens_per_second": 4313131.431365157, "eta_seconds": 2721.0780127526286, "mean_data_wait_seconds": 8.091330528259277e-05}
{"step": 24540, "total_steps": 46238, "tokens": 13268631552, "loss": 3.06321094975327, "ce_loss": 3.06321094975327, "z_loss": 22.3674054145813, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16243097186088562, "tokens_per_second": 4313145.399959758, "eta_seconds": 2719.815655671949, "mean_data_wait_seconds": 0.00011980533599853516}
{"step": 24550, "total_steps": 46238, "tokens": 13274038272, "loss": 3.0070519176396457, "ce_loss": 3.0070519176396457, "z_loss": 21.92770952166933, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15677125751972198, "tokens_per_second": 4313105.420446292, "eta_seconds": 2718.5873102973487, "mean_data_wait_seconds": 8.39531421661377e-05}
{"step": 24560, "total_steps": 46238, "tokens": 13279444992, "loss": 3.1393521464232244, "ce_loss": 3.1393521464232244, "z_loss": 22.4338922067122, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16968871653079987, "tokens_per_second": 4312230.628742391, "eta_seconds": 2717.8850003711505, "mean_data_wait_seconds": 0.00010758638381958008}
{"step": 24570, "total_steps": 46238, "tokens": 13284851712, "loss": 3.0334612131118774, "ce_loss": 3.0334612131118774, "z_loss": 22.246175390301328, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1557537466287613, "tokens_per_second": 4312186.3538287105, "eta_seconds": 2716.659082601729, "mean_data_wait_seconds": 8.544325828552246e-05}
{"step": 24580, "total_steps": 46238, "tokens": 13290258432, "loss": 3.0399006189722004, "ce_loss": 3.0399006189722004, "z_loss": 22.581960519154865, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16811344027519226, "tokens_per_second": 4312144.063255382, "eta_seconds": 2715.4318900839858, "mean_data_wait_seconds": 8.58306884765625e-05}
{"step": 24590, "total_steps": 46238, "tokens": 13295665152, "loss": 3.021291333617586, "ce_loss": 3.021291333617586, "z_loss": 22.596709236954197, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14931870996952057, "tokens_per_second": 4312192.794241588, "eta_seconds": 2714.1473821924615, "mean_data_wait_seconds": 8.657574653625488e-05}
{"step": 24600, "total_steps": 46238, "tokens": 13301071872, "loss": 3.083395367318934, "ce_loss": 3.083395367318934, "z_loss": 22.16944393967137, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1497279852628708, "tokens_per_second": 4311294.064348119, "eta_seconds": 2713.459088940353, "mean_data_wait_seconds": 9.673833847045898e-05}
{"step": 24610, "total_steps": 46238, "tokens": 13306478592, "loss": 3.033781080534964, "ce_loss": 3.033781080534964, "z_loss": 22.284806713913426, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16504070162773132, "tokens_per_second": 4311279.286477879, "eta_seconds": 2712.2143027650495, "mean_data_wait_seconds": 0.00011354684829711914}
{"step": 24620, "total_steps": 46238, "tokens": 13311885312, "loss": 3.023195044560866, "ce_loss": 3.023195044560866, "z_loss": 22.474632624423865, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15819446742534637, "tokens_per_second": 4311227.220593886, "eta_seconds": 2710.992955363178, "mean_data_wait_seconds": 7.030367851257324e-05}
{"step": 24630, "total_steps": 46238, "tokens": 13317292032, "loss": 3.0218635887810676, "ce_loss": 3.0218635887810676, "z_loss": 22.552743391557172, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1657370626926422, "tokens_per_second": 4311213.13418313, "eta_seconds": 2709.7477068281178, "mean_data_wait_seconds": 0.00016608834266662598}
{"step": 24640, "total_steps": 46238, "tokens": 13322698752, "loss": 3.0056220600099275, "ce_loss": 3.0056220600099275, "z_loss": 23.104766860152736, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13859744369983673, "tokens_per_second": 4311316.291575152, "eta_seconds": 2708.428793966729, "mean_data_wait_seconds": 7.423758506774902e-05}
{"step": 24650, "total_steps": 46238, "tokens": 13328105472, "loss": 2.9975099581660647, "ce_loss": 2.9975099581660647, "z_loss": 22.309154308203496, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13836529850959778, "tokens_per_second": 4311318.2571977535, "eta_seconds": 2707.1734833109185, "mean_data_wait_seconds": 7.88569450378418e-05}
{"step": 24660, "total_steps": 46238, "tokens": 13333512192, "loss": 3.0756519671642417, "ce_loss": 3.0756519671642417, "z_loss": 22.26200385527177, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15109378099441528, "tokens_per_second": 4311311.133706376, "eta_seconds": 2705.923878421372, "mean_data_wait_seconds": 6.553530693054199e-05}
{"step": 24670, "total_steps": 46238, "tokens": 13338918912, "loss": 2.993989084706162, "ce_loss": 2.993989084706162, "z_loss": 22.172973098176897, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17054662108421326, "tokens_per_second": 4311363.191355506, "eta_seconds": 2704.6371429296933, "mean_data_wait_seconds": 0.00013890862464904785}
{"step": 24680, "total_steps": 46238, "tokens": 13344325632, "loss": 3.0809174703829214, "ce_loss": 3.0809174703829214, "z_loss": 22.52065454829823, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14957773685455322, "tokens_per_second": 4311511.979400926, "eta_seconds": 2703.289787825075, "mean_data_wait_seconds": 0.0001760721206665039}
{"step": 24690, "total_steps": 46238, "tokens": 13349732352, "loss": 3.0435824448412117, "ce_loss": 3.0435824448412117, "z_loss": 21.982942523378313, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.131186380982399, "tokens_per_second": 4311666.648388611, "eta_seconds": 2701.9388403678827, "mean_data_wait_seconds": 0.00011712312698364258}
{"step": 24700, "total_steps": 46238, "tokens": 13355139072, "loss": 3.0893828417315627, "ce_loss": 3.0893828417315627, "z_loss": 22.63786928581469, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1631658971309662, "tokens_per_second": 4311665.740440285, "eta_seconds": 2700.685434583556, "mean_data_wait_seconds": 9.271502494812012e-05}
{"step": 24710, "total_steps": 46238, "tokens": 13360545792, "loss": 3.0594845179355508, "ce_loss": 3.0594845179355508, "z_loss": 23.058158787814055, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1682807207107544, "tokens_per_second": 4311798.543343984, "eta_seconds": 2699.3483176450595, "mean_data_wait_seconds": 9.09268856048584e-05}
{"step": 24720, "total_steps": 46238, "tokens": 13365952512, "loss": 3.0158251812963774, "ce_loss": 3.0158251812963774, "z_loss": 22.741026112527557, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14456655085086823, "tokens_per_second": 4311924.214288641, "eta_seconds": 2698.015745603557, "mean_data_wait_seconds": 5.957484245300293e-05}
{"step": 24730, "total_steps": 46238, "tokens": 13371359232, "loss": 3.030581329808091, "ce_loss": 3.030581329808091, "z_loss": 22.553187731540564, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1401195079088211, "tokens_per_second": 4312038.813424482, "eta_seconds": 2696.690175375586, "mean_data_wait_seconds": 0.0001036524772644043}
{"step": 24740, "total_steps": 46238, "tokens": 13376765952, "loss": 3.0471980987173137, "ce_loss": 3.0471980987173137, "z_loss": 22.350656177058365, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1499389111995697, "tokens_per_second": 4312205.303068216, "eta_seconds": 2695.332241192259, "mean_data_wait_seconds": 0.00011873245239257812}
{"step": 24750, "total_steps": 46238, "tokens": 13382172672, "loss": 3.011522654331092, "ce_loss": 3.011522654331092, "z_loss": 22.18000079646255, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1588457077741623, "tokens_per_second": 4312234.565136586, "eta_seconds": 2694.0601417938005, "mean_data_wait_seconds": 7.903575897216797e-05}
{"step": 24760, "total_steps": 46238, "tokens": 13387579392, "loss": 3.0466459357377254, "ce_loss": 3.0466459357377254, "z_loss": 23.45515467903831, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17403320968151093, "tokens_per_second": 4312354.139612325, "eta_seconds": 2692.7316653645485, "mean_data_wait_seconds": 6.812810897827148e-05}
{"step": 24770, "total_steps": 46238, "tokens": 13392986112, "loss": 2.997561702222535, "ce_loss": 2.997561702222535, "z_loss": 22.095226071097635, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14940422773361206, "tokens_per_second": 4312559.304954659, "eta_seconds": 2691.3498466363762, "mean_data_wait_seconds": 9.992718696594238e-05}
{"step": 24780, "total_steps": 46238, "tokens": 13398392832, "loss": 3.006885709184589, "ce_loss": 3.006885709184589, "z_loss": 22.493452072143555, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1890193521976471, "tokens_per_second": 4312722.469831333, "eta_seconds": 2689.9943562688177, "mean_data_wait_seconds": 8.845329284667969e-05}
{"step": 24790, "total_steps": 46238, "tokens": 13403799552, "loss": 2.996930302995624, "ce_loss": 2.996930302995624, "z_loss": 23.11919206561464, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1458248794078827, "tokens_per_second": 4312889.723987609, "eta_seconds": 2688.6364192216743, "mean_data_wait_seconds": 6.35385513305664e-05}
{"step": 24800, "total_steps": 46238, "tokens": 13409206272, "loss": 2.9939119255903996, "ce_loss": 2.9939119255903996, "z_loss": 22.74784501393636, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17060516774654388, "tokens_per_second": 4313044.124812135, "eta_seconds": 2687.2865958691873, "mean_data_wait_seconds": 6.332993507385254e-05}
{"step": 24810, "total_steps": 46238, "tokens": 13414612992, "loss": 2.97675585566145, "ce_loss": 2.97675585566145, "z_loss": 23.174511591593426, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14670245349407196, "tokens_per_second": 4313227.568613888, "eta_seconds": 2685.91878348839, "mean_data_wait_seconds": 6.16312026977539e-05}
{"step": 24820, "total_steps": 46238, "tokens": 13420019712, "loss": 3.051080757921392, "ce_loss": 3.051080757921392, "z_loss": 21.810595440142084, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14119261503219604, "tokens_per_second": 4313407.804153175, "eta_seconds": 2684.5530841880013, "mean_data_wait_seconds": 6.476044654846191e-05}
{"step": 24830, "total_steps": 46238, "tokens": 13425426432, "loss": 3.0183604377688784, "ce_loss": 3.0183604377688784, "z_loss": 23.3684876759847, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16039974987506866, "tokens_per_second": 4313424.2490982115, "eta_seconds": 2683.2893857866543, "mean_data_wait_seconds": 6.0677528381347656e-05}
{"step": 24840, "total_steps": 46238, "tokens": 13430833152, "loss": 2.9735074332266143, "ce_loss": 2.9735074332266143, "z_loss": 22.73801766019879, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19833175837993622, "tokens_per_second": 4313554.361676835, "eta_seconds": 2681.9550222389694, "mean_data_wait_seconds": 0.00015231966972351074}
{"step": 24850, "total_steps": 46238, "tokens": 13436239872, "loss": 3.0098585883776345, "ce_loss": 3.0098585883776345, "z_loss": 22.537251472473145, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1522054225206375, "tokens_per_second": 4313524.290833954, "eta_seconds": 2680.7202844717035, "mean_data_wait_seconds": 8.243322372436523e-05}
{"step": 24860, "total_steps": 46238, "tokens": 13441646592, "loss": 2.9973354700839883, "ce_loss": 2.9973354700839883, "z_loss": 23.35069214213978, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15301427245140076, "tokens_per_second": 4313739.697063862, "eta_seconds": 2679.333051057042, "mean_data_wait_seconds": 7.194280624389648e-05}
{"step": 24870, "total_steps": 46238, "tokens": 13447053312, "loss": 3.044744818499594, "ce_loss": 3.044744818499594, "z_loss": 22.989806652069092, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16149631142616272, "tokens_per_second": 4313861.936897437, "eta_seconds": 2678.0037917274367, "mean_data_wait_seconds": 0.00010648369789123535}
{"step": 24880, "total_steps": 46238, "tokens": 13452460032, "loss": 3.023152942007238, "ce_loss": 3.023152942007238, "z_loss": 23.325105580416594, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1427069455385208, "tokens_per_second": 4313864.209884149, "eta_seconds": 2676.7490449844513, "mean_data_wait_seconds": 7.399916648864746e-05}
{"step": 24890, "total_steps": 46238, "tokens": 13457866752, "loss": 3.0235818714806526, "ce_loss": 3.0235818714806526, "z_loss": 22.56048400474317, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14285308122634888, "tokens_per_second": 4313927.717667235, "eta_seconds": 2675.4563217951204, "mean_data_wait_seconds": 0.00012472271919250488}
{"step": 24900, "total_steps": 46238, "tokens": 13463273472, "loss": 3.0357627868652344, "ce_loss": 3.0357627868652344, "z_loss": 23.283104636452414, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15438370406627655, "tokens_per_second": 4314098.323230996, "eta_seconds": 2674.097250375138, "mean_data_wait_seconds": 0.0005294084548950195}
{"step": 24910, "total_steps": 46238, "tokens": 13468680192, "loss": 3.0174263408689788, "ce_loss": 3.0174263408689788, "z_loss": 23.04407428972649, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13511711359024048, "tokens_per_second": 4314212.961748462, "eta_seconds": 2672.772959109269, "mean_data_wait_seconds": 0.00016313791275024414}
{"step": 24920, "total_steps": 46238, "tokens": 13474086912, "loss": 3.0328534678979353, "ce_loss": 3.0328534678979353, "z_loss": 23.26682931726629, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16997268795967102, "tokens_per_second": 4314199.969660105, "eta_seconds": 2671.527769935068, "mean_data_wait_seconds": 9.72449779510498e-05}
{"step": 24930, "total_steps": 46238, "tokens": 13479493632, "loss": 3.0027689662846653, "ce_loss": 3.0027689662846653, "z_loss": 22.521155429608893, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14363491535186768, "tokens_per_second": 4314306.792174871, "eta_seconds": 2670.2084156126134, "mean_data_wait_seconds": 0.00010824203491210938}
{"step": 24940, "total_steps": 46238, "tokens": 13484900352, "loss": 3.0584995656302483, "ce_loss": 3.0584995656302483, "z_loss": 23.40058812228116, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16862165927886963, "tokens_per_second": 4314519.739456519, "eta_seconds": 2668.823479632627, "mean_data_wait_seconds": 0.00013905763626098633}
{"step": 24950, "total_steps": 46238, "tokens": 13490307072, "loss": 2.9613184387033638, "ce_loss": 2.9613184387033638, "z_loss": 23.47446766766635, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13834604620933533, "tokens_per_second": 4314602.732233337, "eta_seconds": 2667.519022786724, "mean_data_wait_seconds": 0.0001061558723449707}
{"step": 24960, "total_steps": 46238, "tokens": 13495713792, "loss": 3.007304740674568, "ce_loss": 3.007304740674568, "z_loss": 22.86486325119481, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1433468759059906, "tokens_per_second": 4314836.295767186, "eta_seconds": 2666.1215757560017, "mean_data_wait_seconds": 6.935000419616699e-05}
{"step": 24970, "total_steps": 46238, "tokens": 13501120512, "loss": 3.040237571253921, "ce_loss": 3.040237571253921, "z_loss": 22.91404965429595, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2066289633512497, "tokens_per_second": 4314918.1884953175, "eta_seconds": 2664.817945948056, "mean_data_wait_seconds": 7.039308547973633e-05}
{"step": 24980, "total_steps": 46238, "tokens": 13506527232, "loss": 3.0446983937061196, "ce_loss": 3.0446983937061196, "z_loss": 22.872183611898713, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14159750938415527, "tokens_per_second": 4315018.47817742, "eta_seconds": 2663.5030098073757, "mean_data_wait_seconds": 8.368492126464844e-05}
{"step": 24990, "total_steps": 46238, "tokens": 13511933952, "loss": 3.006074627240499, "ce_loss": 3.006074627240499, "z_loss": 22.960516033750594, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1338035762310028, "tokens_per_second": 4315139.578597476, "eta_seconds": 2662.175295783541, "mean_data_wait_seconds": 0.00014194846153259277}
{"step": 25000, "total_steps": 46238, "tokens": 13517340672, "loss": 2.992124830231522, "ce_loss": 2.992124830231522, "z_loss": 23.158333185947303, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1583697348833084, "tokens_per_second": 4315309.485774998, "eta_seconds": 2660.8175617183742, "mean_data_wait_seconds": 0.0001201927661895752}
{"step": 25010, "total_steps": 46238, "tokens": 13522747392, "loss": 3.0532673452839707, "ce_loss": 3.0532673452839707, "z_loss": 22.799310366312664, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17671774327754974, "tokens_per_second": 4315353.082785537, "eta_seconds": 2659.537776823527, "mean_data_wait_seconds": 6.073713302612305e-05}
{"step": 25020, "total_steps": 46238, "tokens": 13528154112, "loss": 3.087859706445174, "ce_loss": 3.087859706445174, "z_loss": 23.216931054086395, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13957086205482483, "tokens_per_second": 4315376.120088523, "eta_seconds": 2658.270682501872, "mean_data_wait_seconds": 0.00011092424392700195}
{"step": 25030, "total_steps": 46238, "tokens": 13533560832, "loss": 2.9994135878302832, "ce_loss": 2.9994135878302832, "z_loss": 23.401947440523088, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17346586287021637, "tokens_per_second": 4315301.120253974, "eta_seconds": 2657.063964825976, "mean_data_wait_seconds": 7.37905502319336e-05}
{"step": 25040, "total_steps": 46238, "tokens": 13538967552, "loss": 2.9540593533804924, "ce_loss": 2.9540593533804924, "z_loss": 23.097889437819973, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1575431227684021, "tokens_per_second": 4315245.801202484, "eta_seconds": 2655.8450924872895, "mean_data_wait_seconds": 8.183717727661133e-05}
{"step": 25050, "total_steps": 46238, "tokens": 13544374272, "loss": 3.0416870279745623, "ce_loss": 3.0416870279745623, "z_loss": 22.902444377089992, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1598166674375534, "tokens_per_second": 4315278.441485376, "eta_seconds": 2654.5720790283376, "mean_data_wait_seconds": 6.4849853515625e-05}
{"step": 25060, "total_steps": 46238, "tokens": 13549780992, "loss": 3.0670742085485747, "ce_loss": 3.0670742085485747, "z_loss": 23.14240737394853, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1500391960144043, "tokens_per_second": 4315428.764763718, "eta_seconds": 2653.2267285906437, "mean_data_wait_seconds": 7.56680965423584e-05}
{"step": 25070, "total_steps": 46238, "tokens": 13555187712, "loss": 3.0576987374912608, "ce_loss": 3.0576987374912608, "z_loss": 22.834707086736504, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1362299919128418, "tokens_per_second": 4315519.0149075575, "eta_seconds": 2651.9183867493975, "mean_data_wait_seconds": 7.450580596923828e-05}
{"step": 25080, "total_steps": 46238, "tokens": 13560594432, "loss": 3.0688431642272254, "ce_loss": 3.0688431642272254, "z_loss": 23.10509540095474, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1582556664943695, "tokens_per_second": 4315695.362576126, "eta_seconds": 2650.557220325169, "mean_data_wait_seconds": 8.431077003479004e-05}
{"step": 25090, "total_steps": 46238, "tokens": 13566001152, "loss": 3.053029128999421, "ce_loss": 3.053029128999421, "z_loss": 22.957329620014537, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13632942736148834, "tokens_per_second": 4315741.961692091, "eta_seconds": 2649.2758106226497, "mean_data_wait_seconds": 0.00011831521987915039}
{"step": 25100, "total_steps": 46238, "tokens": 13571407872, "loss": 3.051134739861344, "ce_loss": 3.051134739861344, "z_loss": 22.807884707595363, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1616070717573166, "tokens_per_second": 4315751.590411224, "eta_seconds": 2648.017112335947, "mean_data_wait_seconds": 0.00031891465187072754}
{"step": 25110, "total_steps": 46238, "tokens": 13576814592, "loss": 3.0287617697860254, "ce_loss": 3.0287617697860254, "z_loss": 23.434150796948057, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1443636417388916, "tokens_per_second": 4315661.391570606, "eta_seconds": 2646.819643058903, "mean_data_wait_seconds": 9.438395500183105e-05}
{"step": 25120, "total_steps": 46238, "tokens": 13582221312, "loss": 2.996374215140487, "ce_loss": 2.996374215140487, "z_loss": 22.766937458153926, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1334412544965744, "tokens_per_second": 4315654.3381753, "eta_seconds": 2645.5711531399834, "mean_data_wait_seconds": 0.0001239776611328125}
{"step": 25130, "total_steps": 46238, "tokens": 13587628032, "loss": 3.0338601325497483, "ce_loss": 3.0338601325497483, "z_loss": 23.40875176227454, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13233457505702972, "tokens_per_second": 4315688.0616523195, "eta_seconds": 2644.297674200015, "mean_data_wait_seconds": 6.091594696044922e-05}
{"step": 25140, "total_steps": 46238, "tokens": 13593034752, "loss": 3.0012263944654753, "ce_loss": 3.0012263944654753, "z_loss": 22.795681736686014, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14370431005954742, "tokens_per_second": 4315668.562586756, "eta_seconds": 2643.056809988915, "mean_data_wait_seconds": 0.00010010600090026855}
{"step": 25150, "total_steps": 46238, "tokens": 13598441472, "loss": 3.0461854826320303, "ce_loss": 3.0461854826320303, "z_loss": 22.89632253935843, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1366313099861145, "tokens_per_second": 4315764.576733121, "eta_seconds": 2641.745225275996, "mean_data_wait_seconds": 0.00011152029037475586}
{"step": 25160, "total_steps": 46238, "tokens": 13603848192, "loss": 3.018800423000798, "ce_loss": 3.018800423000798, "z_loss": 23.43141537001639, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16946515440940857, "tokens_per_second": 4315825.457362782, "eta_seconds": 2640.4551937008723, "mean_data_wait_seconds": 6.604194641113281e-05}
{"step": 25170, "total_steps": 46238, "tokens": 13609254912, "loss": 3.054397149519487, "ce_loss": 3.054397149519487, "z_loss": 23.542500857150916, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14998862147331238, "tokens_per_second": 4315916.358810087, "eta_seconds": 2639.146840913376, "mean_data_wait_seconds": 6.0677528381347656e-05}
{"step": 25180, "total_steps": 46238, "tokens": 13614661632, "loss": 3.051417699365905, "ce_loss": 3.051417699365905, "z_loss": 22.575233806263316, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.21472613513469696, "tokens_per_second": 4316019.177532839, "eta_seconds": 2637.831259709081, "mean_data_wait_seconds": 6.845593452453613e-05}
{"step": 25190, "total_steps": 46238, "tokens": 13620068352, "loss": 3.0377234455310935, "ce_loss": 3.0377234455310935, "z_loss": 23.312078923890084, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15938863158226013, "tokens_per_second": 4315969.824263267, "eta_seconds": 2636.608699168205, "mean_data_wait_seconds": 0.00018236041069030762}
{"step": 25200, "total_steps": 46238, "tokens": 13625475072, "loss": 3.043004886670546, "ce_loss": 3.043004886670546, "z_loss": 23.229560317415178, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1586497277021408, "tokens_per_second": 4316131.963651532, "eta_seconds": 2635.25697540936, "mean_data_wait_seconds": 6.112456321716309e-05}
{"step": 25210, "total_steps": 46238, "tokens": 13630881792, "loss": 3.0742954593716245, "ce_loss": 3.0742954593716245, "z_loss": 23.97227530045943, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16873174905776978, "tokens_per_second": 4316303.360119172, "eta_seconds": 2633.8997043261834, "mean_data_wait_seconds": 7.256865501403809e-05}
{"step": 25220, "total_steps": 46238, "tokens": 13636288512, "loss": 2.982820136980577, "ce_loss": 2.982820136980577, "z_loss": 24.13520720510772, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14654742181301117, "tokens_per_second": 4316541.673835262, "eta_seconds": 2632.501730002682, "mean_data_wait_seconds": 7.471442222595215e-05}
{"step": 25230, "total_steps": 46238, "tokens": 13641695232, "loss": 3.029532672780933, "ce_loss": 3.029532672780933, "z_loss": 23.23438410325484, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20305775105953217, "tokens_per_second": 4316696.087625927, "eta_seconds": 2631.15504854699, "mean_data_wait_seconds": 7.274746894836426e-05}
{"step": 25240, "total_steps": 46238, "tokens": 13647101952, "loss": 3.0450868967807656, "ce_loss": 3.0450868967807656, "z_loss": 22.949952819130637, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.131815567612648, "tokens_per_second": 4316891.369298322, "eta_seconds": 2629.783567114699, "mean_data_wait_seconds": 8.302927017211914e-05}
{"step": 25250, "total_steps": 46238, "tokens": 13652508672, "loss": 3.051262328118989, "ce_loss": 3.051262328118989, "z_loss": 23.036352504383434, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14065206050872803, "tokens_per_second": 4317119.428840362, "eta_seconds": 2628.3922534540548, "mean_data_wait_seconds": 6.970763206481934e-05}
{"step": 25260, "total_steps": 46238, "tokens": 13657915392, "loss": 3.046821196873983, "ce_loss": 3.046821196873983, "z_loss": 23.213280923438795, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15487223863601685, "tokens_per_second": 4317341.863249444, "eta_seconds": 2627.0045095441424, "mean_data_wait_seconds": 6.580352783203125e-05}
{"step": 25270, "total_steps": 46238, "tokens": 13663322112, "loss": 3.0704549478762075, "ce_loss": 3.0704549478762075, "z_loss": 23.899062142227635, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15445098280906677, "tokens_per_second": 4317527.636033646, "eta_seconds": 2625.639203647163, "mean_data_wait_seconds": 7.137656211853027e-05}
{"step": 25280, "total_steps": 46238, "tokens": 13668728832, "loss": 3.0302464853633535, "ce_loss": 3.0302464853633535, "z_loss": 23.62270668781165, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15246757864952087, "tokens_per_second": 4317698.645238965, "eta_seconds": 2624.2829884606012, "mean_data_wait_seconds": 0.0001544654369354248}
{"step": 25290, "total_steps": 46238, "tokens": 13674135552, "loss": 3.14720665324818, "ce_loss": 3.14720665324818, "z_loss": 23.0226756587173, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14909540116786957, "tokens_per_second": 4317905.024368982, "eta_seconds": 2622.905395112321, "mean_data_wait_seconds": 0.0001647472381591797}
{"step": 25300, "total_steps": 46238, "tokens": 13679542272, "loss": 3.0178711649143333, "ce_loss": 3.0178711649143333, "z_loss": 23.326578906088166, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16430430114269257, "tokens_per_second": 4317976.140025266, "eta_seconds": 2621.6100545506397, "mean_data_wait_seconds": 0.002638787031173706}
{"step": 25310, "total_steps": 46238, "tokens": 13684948992, "loss": 3.0069319876757534, "ce_loss": 3.0069319876757534, "z_loss": 23.36734741384333, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15555629134178162, "tokens_per_second": 4318050.468762639, "eta_seconds": 2620.312806867742, "mean_data_wait_seconds": 0.00011682510375976562}
{"step": 25320, "total_steps": 46238, "tokens": 13690355712, "loss": 3.0036653966614693, "ce_loss": 3.0036653966614693, "z_loss": 23.766487425023858, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17797690629959106, "tokens_per_second": 4318172.7044121185, "eta_seconds": 2618.9865478156353, "mean_data_wait_seconds": 7.924437522888184e-05}
{"step": 25330, "total_steps": 46238, "tokens": 13695762432, "loss": 3.0248746619080054, "ce_loss": 3.0248746619080054, "z_loss": 23.645618453170314, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15977884829044342, "tokens_per_second": 4318306.985206916, "eta_seconds": 2617.6530623513245, "mean_data_wait_seconds": 0.00011643767356872559}
{"step": 25340, "total_steps": 46238, "tokens": 13701169152, "loss": 2.997570140795274, "ce_loss": 2.997570140795274, "z_loss": 23.724395477410518, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13923120498657227, "tokens_per_second": 4318351.826988051, "eta_seconds": 2616.373847399179, "mean_data_wait_seconds": 6.440281867980957e-05}
{"step": 25350, "total_steps": 46238, "tokens": 13706575872, "loss": 3.0197102391358577, "ce_loss": 3.0197102391358577, "z_loss": 23.098399090044428, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15559962391853333, "tokens_per_second": 4318439.598116577, "eta_seconds": 2615.0686625153403, "mean_data_wait_seconds": 7.835030555725098e-05}
{"step": 25360, "total_steps": 46238, "tokens": 13711982592, "loss": 2.952521669142174, "ce_loss": 2.952521669142174, "z_loss": 23.535452553720184, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1552252173423767, "tokens_per_second": 4318540.3925983375, "eta_seconds": 2613.7556484005886, "mean_data_wait_seconds": 9.140372276306152e-05}
{"step": 25370, "total_steps": 46238, "tokens": 13717389312, "loss": 2.9904340505599976, "ce_loss": 2.9904340505599976, "z_loss": 23.485156015916303, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12739184498786926, "tokens_per_second": 4318610.341086959, "eta_seconds": 2612.461355140543, "mean_data_wait_seconds": 0.00010350346565246582}
{"step": 25380, "total_steps": 46238, "tokens": 13722796032, "loss": 2.9868776328635938, "ce_loss": 2.9868776328635938, "z_loss": 23.626406308376428, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1327628493309021, "tokens_per_second": 4318589.8821246335, "eta_seconds": 2611.2217672431793, "mean_data_wait_seconds": 8.857250213623047e-05}
{"step": 25390, "total_steps": 46238, "tokens": 13728202752, "loss": 2.999662413741603, "ce_loss": 2.999662413741603, "z_loss": 24.272317712957207, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18881364166736603, "tokens_per_second": 4318532.468624128, "eta_seconds": 2610.004501735524, "mean_data_wait_seconds": 8.481740951538086e-05}
{"step": 25400, "total_steps": 46238, "tokens": 13733609472, "loss": 3.027623982140512, "ce_loss": 3.027623982140512, "z_loss": 23.87374798456828, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14111840724945068, "tokens_per_second": 4318589.630892389, "eta_seconds": 2608.7179905704556, "mean_data_wait_seconds": 6.54458999633789e-05}
{"step": 25410, "total_steps": 46238, "tokens": 13739016192, "loss": 2.9936372511314624, "ce_loss": 2.9936372511314624, "z_loss": 23.905310023914684, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14712265133857727, "tokens_per_second": 4318752.70563864, "eta_seconds": 2607.367569182184, "mean_data_wait_seconds": 7.450580596923828e-05}
{"step": 25420, "total_steps": 46238, "tokens": 13744422912, "loss": 3.0042639392794985, "ce_loss": 3.0042639392794985, "z_loss": 24.085270737156723, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15600799024105072, "tokens_per_second": 4318839.512180261, "eta_seconds": 2606.063270528453, "mean_data_wait_seconds": 9.948015213012695e-05}
{"step": 25430, "total_steps": 46238, "tokens": 13749829632, "loss": 3.0267471743352488, "ce_loss": 3.0267471743352488, "z_loss": 24.072802514740914, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16422706842422485, "tokens_per_second": 4318950.240093957, "eta_seconds": 2604.7445973249432, "mean_data_wait_seconds": 0.00016006827354431152}
{"step": 25440, "total_steps": 46238, "tokens": 13755236352, "loss": 3.00725048419201, "ce_loss": 3.00725048419201, "z_loss": 23.645710453842625, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14245770871639252, "tokens_per_second": 4319102.750318205, "eta_seconds": 2603.4008066077113, "mean_data_wait_seconds": 7.471442222595215e-05}
{"step": 25450, "total_steps": 46238, "tokens": 13760643072, "loss": 2.982527752717336, "ce_loss": 2.982527752717336, "z_loss": 23.2646102327289, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15750549733638763, "tokens_per_second": 4319219.164650768, "eta_seconds": 2602.0788562852954, "mean_data_wait_seconds": 0.0001081228256225586}
{"step": 25460, "total_steps": 46238, "tokens": 13766049792, "loss": 2.9728587938077524, "ce_loss": 2.9728587938077524, "z_loss": 23.401765606620096, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1489401012659073, "tokens_per_second": 4319215.50372307, "eta_seconds": 2600.8292789088505, "mean_data_wait_seconds": 6.401538848876953e-05}
{"step": 25470, "total_steps": 46238, "tokens": 13771456512, "loss": 3.0049290277741174, "ce_loss": 3.0049290277741174, "z_loss": 23.43738891139175, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17375078797340393, "tokens_per_second": 4319331.765836984, "eta_seconds": 2599.5075240126303, "mean_data_wait_seconds": 8.687376976013184e-05}
{"step": 25480, "total_steps": 46238, "tokens": 13776863232, "loss": 3.000504611116467, "ce_loss": 3.000504611116467, "z_loss": 23.715750000693582, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14386898279190063, "tokens_per_second": 4319334.603704389, "eta_seconds": 2598.254067738826, "mean_data_wait_seconds": 0.0001016855239868164}
{"step": 25490, "total_steps": 46238, "tokens": 13782269952, "loss": 2.975474605054566, "ce_loss": 2.975474605054566, "z_loss": 24.884267243472014, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18067167699337006, "tokens_per_second": 4319338.39401588, "eta_seconds": 2597.0000404554453, "mean_data_wait_seconds": 7.826089859008789e-05}
{"step": 25500, "total_steps": 46238, "tokens": 13787676672, "loss": 3.0035501303094807, "ce_loss": 3.0035501303094807, "z_loss": 24.36615550879276, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15380465984344482, "tokens_per_second": 4319330.660173945, "eta_seconds": 2595.75294093101, "mean_data_wait_seconds": 0.00010383129119873047}
{"step": 25510, "total_steps": 46238, "tokens": 13793083392, "loss": 3.01529115077221, "ce_loss": 3.01529115077221, "z_loss": 24.288776542201187, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14934127032756805, "tokens_per_second": 4319367.813507475, "eta_seconds": 2594.4788746527074, "mean_data_wait_seconds": 7.471442222595215e-05}
{"step": 25520, "total_steps": 46238, "tokens": 13798490112, "loss": 2.9533903978087683, "ce_loss": 2.9533903978087683, "z_loss": 23.142983147592254, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17060382664203644, "tokens_per_second": 4319264.551837438, "eta_seconds": 2593.2891328073415, "mean_data_wait_seconds": 0.00010699033737182617}
{"step": 25530, "total_steps": 46238, "tokens": 13803896832, "loss": 3.0078932787432815, "ce_loss": 3.0078932787432815, "z_loss": 23.988462375872064, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16885749995708466, "tokens_per_second": 4319391.764593562, "eta_seconds": 2591.9610246452075, "mean_data_wait_seconds": 7.519125938415527e-05}
{"step": 25540, "total_steps": 46238, "tokens": 13809303552, "loss": 2.9864416284994646, "ce_loss": 2.9864416284994646, "z_loss": 24.047714493491434, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16758425533771515, "tokens_per_second": 4319224.16261514, "eta_seconds": 2590.8098220178204, "mean_data_wait_seconds": 7.009506225585938e-05}
{"step": 25550, "total_steps": 46238, "tokens": 13814710272, "loss": 3.0247667052529077, "ce_loss": 3.0247667052529077, "z_loss": 23.8197797572974, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1651618927717209, "tokens_per_second": 4319246.378499273, "eta_seconds": 2589.5447223564493, "mean_data_wait_seconds": 0.00013932585716247559}
{"step": 25560, "total_steps": 46238, "tokens": 13820116992, "loss": 3.0151988845882993, "ce_loss": 3.0151988845882993, "z_loss": 24.29409290082527, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16727609932422638, "tokens_per_second": 4319214.000797792, "eta_seconds": 2588.312350796942, "mean_data_wait_seconds": 7.930397987365723e-05}
{"step": 25570, "total_steps": 46238, "tokens": 13825523712, "loss": 2.994376666618116, "ce_loss": 2.994376666618116, "z_loss": 23.464980428869072, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16247782111167908, "tokens_per_second": 4319343.911169446, "eta_seconds": 2586.98275798434, "mean_data_wait_seconds": 7.11679458618164e-05}
{"step": 25580, "total_steps": 46238, "tokens": 13830930432, "loss": 2.960641698403792, "ce_loss": 2.960641698403792, "z_loss": 23.576536482030694, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14331316947937012, "tokens_per_second": 4319388.606463557, "eta_seconds": 2585.70425621977, "mean_data_wait_seconds": 0.0001074075698852539}
{"step": 25590, "total_steps": 46238, "tokens": 13836337152, "loss": 3.0561457456964436, "ce_loss": 3.0561457456964436, "z_loss": 24.052131595033586, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20427770912647247, "tokens_per_second": 4319355.494033161, "eta_seconds": 2584.4723360744747, "mean_data_wait_seconds": 0.00011897087097167969}
{"step": 25600, "total_steps": 46238, "tokens": 13841743872, "loss": 3.024368607636654, "ce_loss": 3.024368607636654, "z_loss": 23.69030339790113, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.150217205286026, "tokens_per_second": 4319453.511860878, "eta_seconds": 2583.161974856641, "mean_data_wait_seconds": 8.291006088256836e-05}
{"step": 25610, "total_steps": 46238, "tokens": 13847150592, "loss": 3.0622944235801697, "ce_loss": 3.0622944235801697, "z_loss": 22.81895910609852, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15731200575828552, "tokens_per_second": 4319450.450911051, "eta_seconds": 2581.9120906104495, "mean_data_wait_seconds": 0.00010958313941955566}
{"step": 25620, "total_steps": 46238, "tokens": 13852557312, "loss": 3.0516913840264985, "ce_loss": 3.0516913840264985, "z_loss": 23.985536922108043, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13686420023441315, "tokens_per_second": 4319582.8046148345, "eta_seconds": 2580.5813033821332, "mean_data_wait_seconds": 6.434321403503418e-05}
{"step": 25630, "total_steps": 46238, "tokens": 13857964032, "loss": 3.010043518109755, "ce_loss": 3.010043518109755, "z_loss": 23.422503572521787, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13781270384788513, "tokens_per_second": 4319728.764592504, "eta_seconds": 2579.2424735841096, "mean_data_wait_seconds": 6.327033042907715e-05}
{"step": 25640, "total_steps": 46238, "tokens": 13863370752, "loss": 3.1119243520678896, "ce_loss": 3.1119243520678896, "z_loss": 22.576237245039508, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1809183955192566, "tokens_per_second": 4319769.054051031, "eta_seconds": 2577.9667951360907, "mean_data_wait_seconds": 6.732344627380371e-05}
{"step": 25650, "total_steps": 46238, "tokens": 13868777472, "loss": 3.0458857037804346, "ce_loss": 3.0458857037804346, "z_loss": 24.23092964923743, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1745501160621643, "tokens_per_second": 4319904.675557939, "eta_seconds": 2576.6342778298445, "mean_data_wait_seconds": 0.0001755356788635254}
{"step": 25660, "total_steps": 46238, "tokens": 13874184192, "loss": 3.0934899937022817, "ce_loss": 3.0934899937022817, "z_loss": 24.093409755013205, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1516750007867813, "tokens_per_second": 4319966.294242799, "eta_seconds": 2575.3459601818613, "mean_data_wait_seconds": 0.0001112222671508789}
{"step": 25670, "total_steps": 46238, "tokens": 13879590912, "loss": 3.073921656969822, "ce_loss": 3.073921656969822, "z_loss": 23.784387415105645, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14712299406528473, "tokens_per_second": 4320167.207133259, "eta_seconds": 2573.9746845073896, "mean_data_wait_seconds": 0.0001335442066192627}
{"step": 25680, "total_steps": 46238, "tokens": 13884997632, "loss": 3.082255533247283, "ce_loss": 3.082255533247283, "z_loss": 23.87869271365079, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1484740674495697, "tokens_per_second": 4320274.522858166, "eta_seconds": 2572.659270885155, "mean_data_wait_seconds": 0.00012007355690002441}
{"step": 25690, "total_steps": 46238, "tokens": 13890404352, "loss": 3.0468426527399006, "ce_loss": 3.0468426527399006, "z_loss": 23.726769172784053, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16159313917160034, "tokens_per_second": 4320228.704414313, "eta_seconds": 2571.4350660762198, "mean_data_wait_seconds": 0.00014907121658325195}
{"step": 25700, "total_steps": 46238, "tokens": 13895811072, "loss": 3.069992838483868, "ce_loss": 3.069992838483868, "z_loss": 23.49699115753174, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1613423228263855, "tokens_per_second": 4320341.248264618, "eta_seconds": 2570.1166241116102, "mean_data_wait_seconds": 0.00015681982040405273}
{"step": 25710, "total_steps": 46238, "tokens": 13901217792, "loss": 3.090261822397059, "ce_loss": 3.090261822397059, "z_loss": 23.250594037951846, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18818649649620056, "tokens_per_second": 4320354.118028793, "eta_seconds": 2568.857515101968, "mean_data_wait_seconds": 7.620453834533691e-05}
{"step": 25720, "total_steps": 46238, "tokens": 13906624512, "loss": 3.0729578643134148, "ce_loss": 3.0729578643134148, "z_loss": 23.982393553762726, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14831019937992096, "tokens_per_second": 4320380.958058684, "eta_seconds": 2567.5901110777286, "mean_data_wait_seconds": 6.824731826782227e-05}
{"step": 25730, "total_steps": 46238, "tokens": 13912031232, "loss": 3.077347018501975, "ce_loss": 3.077347018501975, "z_loss": 23.550440383679938, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1620502769947052, "tokens_per_second": 4320499.060396093, "eta_seconds": 2566.268514124736, "mean_data_wait_seconds": 8.240342140197754e-05}
{"step": 25740, "total_steps": 46238, "tokens": 13917437952, "loss": 2.991950750350952, "ce_loss": 2.991950750350952, "z_loss": 23.226713050495494, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17106802761554718, "tokens_per_second": 4320462.010947393, "eta_seconds": 2565.039099040683, "mean_data_wait_seconds": 9.244680404663086e-05}
{"step": 25750, "total_steps": 46238, "tokens": 13922844672, "loss": 3.025089068846269, "ce_loss": 3.025089068846269, "z_loss": 24.2230142102097, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16255846619606018, "tokens_per_second": 4320414.84482755, "eta_seconds": 2563.8156662805673, "mean_data_wait_seconds": 5.91278076171875e-05}
{"step": 25760, "total_steps": 46238, "tokens": 13928251392, "loss": 3.058416926499569, "ce_loss": 3.058416926499569, "z_loss": 24.077419916788738, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16556741297245026, "tokens_per_second": 4320538.483170853, "eta_seconds": 2562.4908994849916, "mean_data_wait_seconds": 5.969405174255371e-05}
{"step": 25770, "total_steps": 46238, "tokens": 13933658112, "loss": 3.0252723440979468, "ce_loss": 3.0252723440979468, "z_loss": 22.94742984482736, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1605474352836609, "tokens_per_second": 4320525.923694283, "eta_seconds": 2561.246945264948, "mean_data_wait_seconds": 0.00011834502220153809}
{"step": 25780, "total_steps": 46238, "tokens": 13939064832, "loss": 3.0642183061802024, "ce_loss": 3.0642183061802024, "z_loss": 24.469972494876746, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.4847138226032257, "tokens_per_second": 4320614.969720215, "eta_seconds": 2559.9427816444, "mean_data_wait_seconds": 0.0001035928726196289}
{"step": 25790, "total_steps": 46238, "tokens": 13944471552, "loss": 3.0478528828331917, "ce_loss": 3.0478528828331917, "z_loss": 23.542971712170225, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14373181760311127, "tokens_per_second": 4320753.2923383955, "eta_seconds": 2558.6094914521163, "mean_data_wait_seconds": 8.854269981384277e-05}
{"step": 25800, "total_steps": 46238, "tokens": 13949878272, "loss": 3.017756863073869, "ce_loss": 3.017756863073869, "z_loss": 24.143792007908676, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1411009132862091, "tokens_per_second": 4320913.647663476, "eta_seconds": 2557.2632468540783, "mean_data_wait_seconds": 7.718801498413086e-05}
{"step": 25810, "total_steps": 46238, "tokens": 13955284992, "loss": 3.015101645932053, "ce_loss": 3.015101645932053, "z_loss": 23.638872002110336, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17665845155715942, "tokens_per_second": 4321040.023758633, "eta_seconds": 2555.937201061417, "mean_data_wait_seconds": 7.426738739013672e-05}
{"step": 25820, "total_steps": 46238, "tokens": 13960691712, "loss": 3.0049100030552256, "ce_loss": 3.0049100030552256, "z_loss": 23.892516959797252, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16550301015377045, "tokens_per_second": 4321071.378190711, "eta_seconds": 2554.667409502995, "mean_data_wait_seconds": 8.043646812438965e-05}
{"step": 25830, "total_steps": 46238, "tokens": 13966098432, "loss": 3.041362554738016, "ce_loss": 3.041362554738016, "z_loss": 24.94651293032097, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16553491353988647, "tokens_per_second": 4321174.141555213, "eta_seconds": 2553.3554405722207, "mean_data_wait_seconds": 0.000110626220703125}
{"step": 25840, "total_steps": 46238, "tokens": 13971505152, "loss": 3.028923932350043, "ce_loss": 3.028923932350043, "z_loss": 24.41522537578236, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14661917090415955, "tokens_per_second": 4321111.490207889, "eta_seconds": 2552.141227781521, "mean_data_wait_seconds": 6.198883056640625e-05}
{"step": 25850, "total_steps": 46238, "tokens": 13976911872, "loss": 3.0586953217333015, "ce_loss": 3.0586953217333015, "z_loss": 24.248034679528438, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1479031890630722, "tokens_per_second": 4321163.306349141, "eta_seconds": 2550.8594058003396, "mean_data_wait_seconds": 0.00013780593872070312}
{"step": 25860, "total_steps": 46238, "tokens": 13982318592, "loss": 2.973573825576089, "ce_loss": 2.973573825576089, "z_loss": 23.52557172197284, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16651225090026855, "tokens_per_second": 4321247.479799006, "eta_seconds": 2549.558523436488, "mean_data_wait_seconds": 7.49826431274414e-05}
{"step": 25870, "total_steps": 46238, "tokens": 13987725312, "loss": 3.0464327931404114, "ce_loss": 3.0464327931404114, "z_loss": 24.327717073035963, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12637878954410553, "tokens_per_second": 4321329.874894953, "eta_seconds": 2548.258740434086, "mean_data_wait_seconds": 0.00011751055717468262}
{"step": 25880, "total_steps": 46238, "tokens": 13993132032, "loss": 3.0381088690324263, "ce_loss": 3.0381088690324263, "z_loss": 24.04300014900439, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18838046491146088, "tokens_per_second": 4321431.611370334, "eta_seconds": 2546.947607603081, "mean_data_wait_seconds": 7.417798042297363e-05}
{"step": 25890, "total_steps": 46238, "tokens": 13998538752, "loss": 2.9441075433384287, "ce_loss": 2.9441075433384287, "z_loss": 24.5213428699609, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15947452187538147, "tokens_per_second": 4321561.030628004, "eta_seconds": 2545.6202298273083, "mean_data_wait_seconds": 9.584426879882812e-05}
{"step": 25900, "total_steps": 46238, "tokens": 14003945472, "loss": 3.0219329541379754, "ce_loss": 3.0219329541379754, "z_loss": 24.654514616185967, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18638311326503754, "tokens_per_second": 4321603.38453178, "eta_seconds": 2544.3441902504233, "mean_data_wait_seconds": 8.171796798706055e-05}
{"step": 25910, "total_steps": 46238, "tokens": 14009352192, "loss": 3.02843483469703, "ce_loss": 3.02843483469703, "z_loss": 24.76153799981782, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1673293113708496, "tokens_per_second": 4321709.259754524, "eta_seconds": 2543.030797176822, "mean_data_wait_seconds": 6.920099258422852e-05}
{"step": 25920, "total_steps": 46238, "tokens": 14014758912, "loss": 2.974864136088978, "ce_loss": 2.974864136088978, "z_loss": 24.673293446049545, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18048837780952454, "tokens_per_second": 4321982.106585205, "eta_seconds": 2541.6192740046095, "mean_data_wait_seconds": 0.00010988116264343262}
{"step": 25930, "total_steps": 46238, "tokens": 14020165632, "loss": 2.96890191598372, "ce_loss": 2.96890191598372, "z_loss": 25.378194433270078, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17554309964179993, "tokens_per_second": 4321976.3396166675, "eta_seconds": 2540.37168213045, "mean_data_wait_seconds": 0.00014165043830871582}
{"step": 25940, "total_steps": 46238, "tokens": 14025572352, "loss": 2.999523861841722, "ce_loss": 2.999523861841722, "z_loss": 24.77224063873291, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15440542995929718, "tokens_per_second": 4322076.226920147, "eta_seconds": 2539.0620173813863, "mean_data_wait_seconds": 8.779764175415039e-05}
{"step": 25950, "total_steps": 46238, "tokens": 14030979072, "loss": 2.9185614134326126, "ce_loss": 2.9185614134326126, "z_loss": 24.408531420158617, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19470670819282532, "tokens_per_second": 4322212.556261658, "eta_seconds": 2537.7310165159734, "mean_data_wait_seconds": 8.308887481689453e-05}
{"step": 25960, "total_steps": 46238, "tokens": 14036385792, "loss": 2.9174845471526636, "ce_loss": 2.9174845471526636, "z_loss": 23.804713841640588, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14975056052207947, "tokens_per_second": 4322269.163755039, "eta_seconds": 2536.446882099204, "mean_data_wait_seconds": 8.943676948547363e-05}
{"step": 25970, "total_steps": 46238, "tokens": 14041792512, "loss": 2.9539823965592817, "ce_loss": 2.9539823965592817, "z_loss": 24.09627628326416, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15105365216732025, "tokens_per_second": 4322210.448982347, "eta_seconds": 2535.230422799053, "mean_data_wait_seconds": 7.429718971252441e-05}
{"step": 25980, "total_steps": 46238, "tokens": 14047199232, "loss": 3.024602655208472, "ce_loss": 3.024602655208472, "z_loss": 24.506746075370096, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14399367570877075, "tokens_per_second": 4322239.5886496045, "eta_seconds": 2533.9624237308535, "mean_data_wait_seconds": 9.498000144958496e-05}
{"step": 25990, "total_steps": 46238, "tokens": 14052605952, "loss": 2.9339827877102476, "ce_loss": 2.9339827877102476, "z_loss": 25.1008381988063, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15865127742290497, "tokens_per_second": 4322347.695530013, "eta_seconds": 2532.6481706506174, "mean_data_wait_seconds": 7.662177085876465e-05}
{"step": 26000, "total_steps": 46238, "tokens": 14058012672, "loss": 2.9316388657598784, "ce_loss": 2.9316388657598784, "z_loss": 24.818132039272424, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15486890077590942, "tokens_per_second": 4322475.621155548, "eta_seconds": 2531.322377030535, "mean_data_wait_seconds": 0.00011664628982543945}
{"step": 26010, "total_steps": 46238, "tokens": 14063419392, "loss": 2.9453731684973747, "ce_loss": 2.9453731684973747, "z_loss": 24.56333791848385, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15505219995975494, "tokens_per_second": 4322730.2605899125, "eta_seconds": 2529.922499144688, "mean_data_wait_seconds": 9.050965309143066e-05}
{"step": 26020, "total_steps": 46238, "tokens": 14068826112, "loss": 3.004785044626756, "ce_loss": 3.004785044626756, "z_loss": 25.01208685383652, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1655709594488144, "tokens_per_second": 4322959.047430888, "eta_seconds": 2528.5379075002106, "mean_data_wait_seconds": 5.665421485900879e-05}
{"step": 26030, "total_steps": 46238, "tokens": 14074232832, "loss": 2.9155736529465877, "ce_loss": 2.9155736529465877, "z_loss": 24.238628242955063, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16473396122455597, "tokens_per_second": 4323047.523063216, "eta_seconds": 2527.2354850863476, "mean_data_wait_seconds": 0.0001691281795501709}
{"step": 26040, "total_steps": 46238, "tokens": 14079639552, "loss": 2.970993020317771, "ce_loss": 2.970993020317771, "z_loss": 24.947498364882037, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19732831418514252, "tokens_per_second": 4323166.05935864, "eta_seconds": 2525.915552182148, "mean_data_wait_seconds": 0.00014129281044006348}
{"step": 26050, "total_steps": 46238, "tokens": 14085046272, "loss": 3.002161367373033, "ce_loss": 3.002161367373033, "z_loss": 24.499379288066518, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14755138754844666, "tokens_per_second": 4322404.838418253, "eta_seconds": 2525.1095332370774, "mean_data_wait_seconds": 0.00010570883750915527}
{"step": 26060, "total_steps": 46238, "tokens": 14090452992, "loss": 2.9578376520763743, "ce_loss": 2.9578376520763743, "z_loss": 25.419965541724004, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1735624372959137, "tokens_per_second": 4322488.45359842, "eta_seconds": 2523.809851919505, "mean_data_wait_seconds": 0.0001048743724822998}
{"step": 26070, "total_steps": 46238, "tokens": 14095859712, "loss": 2.9672574744080054, "ce_loss": 2.9672574744080054, "z_loss": 25.59772974072081, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16746948659420013, "tokens_per_second": 4321812.910111867, "eta_seconds": 2522.9533186150265, "mean_data_wait_seconds": 6.103515625e-05}
{"step": 26080, "total_steps": 46238, "tokens": 14101266432, "loss": 2.9919991782217314, "ce_loss": 2.9919991782217314, "z_loss": 24.29515803943981, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17220841348171234, "tokens_per_second": 4321981.8133881325, "eta_seconds": 2521.603739802985, "mean_data_wait_seconds": 6.380677223205566e-05}
{"step": 26090, "total_steps": 46238, "tokens": 14106673152, "loss": 2.8914920488993325, "ce_loss": 2.8914920488993325, "z_loss": 24.837145574165113, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16965195536613464, "tokens_per_second": 4322105.307234535, "eta_seconds": 2520.2807450727637, "mean_data_wait_seconds": 0.00011214613914489746}
{"step": 26100, "total_steps": 46238, "tokens": 14112079872, "loss": 2.988978239622983, "ce_loss": 2.988978239622983, "z_loss": 25.43407188762318, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1715967357158661, "tokens_per_second": 4322272.767879935, "eta_seconds": 2518.932202730995, "mean_data_wait_seconds": 0.00013911724090576172}
{"step": 26110, "total_steps": 46238, "tokens": 14117486592, "loss": 3.037976181868351, "ce_loss": 3.037976181868351, "z_loss": 24.638341253454033, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18502677977085114, "tokens_per_second": 4321111.8542674, "eta_seconds": 2518.3577076934866, "mean_data_wait_seconds": 0.002684175968170166}
{"step": 26120, "total_steps": 46238, "tokens": 14122893312, "loss": 2.95603336168058, "ce_loss": 2.95603336168058, "z_loss": 24.412544308286726, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1672057956457138, "tokens_per_second": 4321059.611022806, "eta_seconds": 2517.1369069415496, "mean_data_wait_seconds": 7.531046867370605e-05}
{"step": 26130, "total_steps": 46238, "tokens": 14128300032, "loss": 3.0170489513512813, "ce_loss": 3.0170489513512813, "z_loss": 24.53528847838893, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17772716283798218, "tokens_per_second": 4321048.753843936, "eta_seconds": 2515.891979772057, "mean_data_wait_seconds": 7.44163990020752e-05}
{"step": 26140, "total_steps": 46238, "tokens": 14133706752, "loss": 2.9935189250743752, "ce_loss": 2.9935189250743752, "z_loss": 23.86754069183812, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17949655652046204, "tokens_per_second": 4321115.303328172, "eta_seconds": 2514.6020000047142, "mean_data_wait_seconds": 9.509921073913574e-05}
{"step": 26150, "total_steps": 46238, "tokens": 14139113472, "loss": 3.02450092272325, "ce_loss": 3.02450092272325, "z_loss": 24.2901435620857, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18083961308002472, "tokens_per_second": 4320435.440760262, "eta_seconds": 2513.746267688447, "mean_data_wait_seconds": 0.00015211105346679688}
{"step": 26160, "total_steps": 46238, "tokens": 14144520192, "loss": 3.0463402795069143, "ce_loss": 3.0463402795069143, "z_loss": 24.269288207545426, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15837089717388153, "tokens_per_second": 4319496.801327331, "eta_seconds": 2513.04081083342, "mean_data_wait_seconds": 8.589029312133789e-05}
{"step": 26170, "total_steps": 46238, "tokens": 14149926912, "loss": 3.0545085271199546, "ce_loss": 3.0545085271199546, "z_loss": 24.808400472005207, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17241144180297852, "tokens_per_second": 4319358.561778113, "eta_seconds": 2511.8694984038584, "mean_data_wait_seconds": 8.240342140197754e-05}
{"step": 26180, "total_steps": 46238, "tokens": 14155333632, "loss": 3.0614807930859653, "ce_loss": 3.0614807930859653, "z_loss": 23.87474759419759, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15112143754959106, "tokens_per_second": 4319383.061165258, "eta_seconds": 2510.603516853747, "mean_data_wait_seconds": 6.48200511932373e-05}
{"step": 26190, "total_steps": 46238, "tokens": 14160740352, "loss": 3.0190450964551983, "ce_loss": 3.0190450964551983, "z_loss": 24.31499482646133, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15131278336048126, "tokens_per_second": 4318168.413225547, "eta_seconds": 2510.0576324913854, "mean_data_wait_seconds": 0.06695529818534851}
{"step": 26200, "total_steps": 46238, "tokens": 14166147072, "loss": 3.0252521850846033, "ce_loss": 3.0252521850846033, "z_loss": 24.03314565889763, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20925560593605042, "tokens_per_second": 4318195.369092185, "eta_seconds": 2508.7898851314635, "mean_data_wait_seconds": 7.268786430358887e-05}
{"step": 26210, "total_steps": 46238, "tokens": 14171553792, "loss": 3.003798352949547, "ce_loss": 3.003798352949547, "z_loss": 24.298797607421875, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1937832087278366, "tokens_per_second": 4318177.011621772, "eta_seconds": 2507.5484666001053, "mean_data_wait_seconds": 5.9992074966430664e-05}
{"step": 26220, "total_steps": 46238, "tokens": 14176960512, "loss": 3.0314352277553445, "ce_loss": 3.0314352277553445, "z_loss": 24.327682350621078, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1548580378293991, "tokens_per_second": 4318284.300689604, "eta_seconds": 2506.2341129952215, "mean_data_wait_seconds": 0.00010141730308532715}
{"step": 26230, "total_steps": 46238, "tokens": 14182367232, "loss": 3.043582932515578, "ce_loss": 3.043582932515578, "z_loss": 24.338810429428563, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16560916602611542, "tokens_per_second": 4317726.336934645, "eta_seconds": 2505.305769721305, "mean_data_wait_seconds": 8.627772331237793e-05}
{"step": 26240, "total_steps": 46238, "tokens": 14187773952, "loss": 3.07266442342238, "ce_loss": 3.07266442342238, "z_loss": 24.603999528017912, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13263490796089172, "tokens_per_second": 4317593.724083191, "eta_seconds": 2504.130465933501, "mean_data_wait_seconds": 9.250640869140625e-05}
{"step": 26250, "total_steps": 46238, "tokens": 14193180672, "loss": 3.0669398054932104, "ce_loss": 3.0669398054932104, "z_loss": 24.342116688237045, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14144358038902283, "tokens_per_second": 4317441.701866032, "eta_seconds": 2502.9663421579926, "mean_data_wait_seconds": 0.00015667080879211426}
{"step": 26260, "total_steps": 46238, "tokens": 14198587392, "loss": 3.0363033937685415, "ce_loss": 3.0363033937685415, "z_loss": 25.663899739583332, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2077682763338089, "tokens_per_second": 4317473.144458033, "eta_seconds": 2501.6958259171374, "mean_data_wait_seconds": 9.199976921081543e-05}
{"step": 26270, "total_steps": 46238, "tokens": 14203994112, "loss": 3.024689271594539, "ce_loss": 3.024689271594539, "z_loss": 23.856564825231377, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14577560126781464, "tokens_per_second": 4317406.167989588, "eta_seconds": 2500.482327570074, "mean_data_wait_seconds": 0.00011980533599853516}
{"step": 26280, "total_steps": 46238, "tokens": 14209400832, "loss": 3.045140232100631, "ce_loss": 3.045140232100631, "z_loss": 24.23886354041822, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16651961207389832, "tokens_per_second": 4317377.40879186, "eta_seconds": 2499.2466681339865, "mean_data_wait_seconds": 0.00010535120964050293}
{"step": 26290, "total_steps": 46238, "tokens": 14214807552, "loss": 2.9919069535804517, "ce_loss": 2.9919069535804517, "z_loss": 24.19326420986291, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17121431231498718, "tokens_per_second": 4317340.391500496, "eta_seconds": 2498.015770364527, "mean_data_wait_seconds": 7.37905502319336e-05}
{"step": 26300, "total_steps": 46238, "tokens": 14220214272, "loss": 3.0862753409327883, "ce_loss": 3.0862753409327883, "z_loss": 23.637604800137606, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15533797442913055, "tokens_per_second": 4317341.088170926, "eta_seconds": 2496.7630409221997, "mean_data_wait_seconds": 6.139278411865234e-05}
{"step": 26310, "total_steps": 46238, "tokens": 14225620992, "loss": 2.9942435387409096, "ce_loss": 2.9942435387409096, "z_loss": 24.66820053620772, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13645878434181213, "tokens_per_second": 4317433.145981052, "eta_seconds": 2495.4575044269336, "mean_data_wait_seconds": 8.124113082885742e-05}
{"step": 26320, "total_steps": 46238, "tokens": 14231027712, "loss": 3.0504169897599653, "ce_loss": 3.0504169897599653, "z_loss": 24.847312652703486, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18333223462104797, "tokens_per_second": 4317551.0785886785, "eta_seconds": 2494.1370763168898, "mean_data_wait_seconds": 0.00010663270950317383}
{"step": 26330, "total_steps": 46238, "tokens": 14236434432, "loss": 3.01503227335034, "ce_loss": 3.01503227335034, "z_loss": 24.9416573119886, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16661758720874786, "tokens_per_second": 4317532.46235808, "eta_seconds": 2492.8955596367546, "mean_data_wait_seconds": 8.404254913330078e-05}
{"step": 26340, "total_steps": 46238, "tokens": 14241841152, "loss": 3.033972467436935, "ce_loss": 3.033972467436935, "z_loss": 24.17797257683494, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1873282939195633, "tokens_per_second": 4317632.941724266, "eta_seconds": 2491.5853036140315, "mean_data_wait_seconds": 6.961822509765625e-05}
{"step": 26350, "total_steps": 46238, "tokens": 14247247872, "loss": 3.013517571218086, "ce_loss": 3.013517571218086, "z_loss": 24.769252603704278, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15554644167423248, "tokens_per_second": 4317705.964135561, "eta_seconds": 2490.2909446157028, "mean_data_wait_seconds": 0.00023061037063598633}
{"step": 26360, "total_steps": 46238, "tokens": 14252654592, "loss": 3.044934751409473, "ce_loss": 3.044934751409473, "z_loss": 24.30000657746286, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1444016546010971, "tokens_per_second": 4317751.624484724, "eta_seconds": 2489.012402440478, "mean_data_wait_seconds": 0.00013071298599243164}
{"step": 26370, "total_steps": 46238, "tokens": 14258061312, "loss": 3.0505775646729902, "ce_loss": 3.0505775646729902, "z_loss": 24.241343050292045, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14469192922115326, "tokens_per_second": 4317842.460447413, "eta_seconds": 2487.7078592827975, "mean_data_wait_seconds": 7.18235969543457e-05}
{"step": 26380, "total_steps": 46238, "tokens": 14263468032, "loss": 3.0143705516150505, "ce_loss": 3.0143705516150505, "z_loss": 24.639085813002154, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14464417099952698, "tokens_per_second": 4317960.829373036, "eta_seconds": 2486.3875167572737, "mean_data_wait_seconds": 0.0001519918441772461}
{"step": 26390, "total_steps": 46238, "tokens": 14268874752, "loss": 3.034579316775004, "ce_loss": 3.034579316775004, "z_loss": 25.544984962000992, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18351364135742188, "tokens_per_second": 4318137.374665956, "eta_seconds": 2485.03376640029, "mean_data_wait_seconds": 7.236003875732422e-05}
{"step": 26400, "total_steps": 46238, "tokens": 14274281472, "loss": 3.0213553020448396, "ce_loss": 3.0213553020448396, "z_loss": 24.138429526126746, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1586713343858719, "tokens_per_second": 4318206.910695453, "eta_seconds": 2483.7416746833637, "mean_data_wait_seconds": 7.358193397521973e-05}
{"step": 26410, "total_steps": 46238, "tokens": 14279688192, "loss": 3.097358611497012, "ce_loss": 3.097358611497012, "z_loss": 25.031725449995562, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15381522476673126, "tokens_per_second": 4318216.314357458, "eta_seconds": 2482.484193382772, "mean_data_wait_seconds": 8.031725883483887e-05}
{"step": 26420, "total_steps": 46238, "tokens": 14285094912, "loss": 3.0508439540863037, "ce_loss": 3.0508439540863037, "z_loss": 24.919226733121004, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18578819930553436, "tokens_per_second": 4317802.405647982, "eta_seconds": 2481.469974166651, "mean_data_wait_seconds": 9.828805923461914e-05}
{"step": 26430, "total_steps": 46238, "tokens": 14290501632, "loss": 3.0137854742281363, "ce_loss": 3.0137854742281363, "z_loss": 25.040197401335746, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.166099414229393, "tokens_per_second": 4317796.179609338, "eta_seconds": 2480.221357963434, "mean_data_wait_seconds": 0.0001379847526550293}
{"step": 26440, "total_steps": 46238, "tokens": 14295908352, "loss": 3.0403951424540896, "ce_loss": 3.0403951424540896, "z_loss": 24.490228219465777, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18060161173343658, "tokens_per_second": 4317899.417098569, "eta_seconds": 2478.9098934574968, "mean_data_wait_seconds": 6.181001663208008e-05}
{"step": 26450, "total_steps": 46238, "tokens": 14301315072, "loss": 2.9927668661782234, "ce_loss": 2.9927668661782234, "z_loss": 24.764300866560504, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14875604212284088, "tokens_per_second": 4318000.291539427, "eta_seconds": 2477.599847540982, "mean_data_wait_seconds": 0.00015917420387268066}
{"step": 26460, "total_steps": 46238, "tokens": 14306721792, "loss": 3.063364039767872, "ce_loss": 3.063364039767872, "z_loss": 24.302166866533685, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17932671308517456, "tokens_per_second": 4318169.679926053, "eta_seconds": 2476.250572947173, "mean_data_wait_seconds": 8.445978164672852e-05}
{"step": 26470, "total_steps": 46238, "tokens": 14312128512, "loss": 3.020784235361851, "ce_loss": 3.020784235361851, "z_loss": 25.608164397152986, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1549815684556961, "tokens_per_second": 4318291.383898503, "eta_seconds": 2474.928733121173, "mean_data_wait_seconds": 0.00010880827903747559}
{"step": 26480, "total_steps": 46238, "tokens": 14317535232, "loss": 2.9775223822304695, "ce_loss": 2.9775223822304695, "z_loss": 25.153663418509744, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18567915260791779, "tokens_per_second": 4318299.086625717, "eta_seconds": 2473.672269964717, "mean_data_wait_seconds": 7.322430610656738e-05}
{"step": 26490, "total_steps": 46238, "tokens": 14322941952, "loss": 3.02932506980318, "ce_loss": 3.02932506980318, "z_loss": 24.753168120528713, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1457759439945221, "tokens_per_second": 4318507.071164601, "eta_seconds": 2472.301146683258, "mean_data_wait_seconds": 6.67572021484375e-05}
{"step": 26500, "total_steps": 46238, "tokens": 14328348672, "loss": 2.9677705710584465, "ce_loss": 2.9677705710584465, "z_loss": 24.601709380294338, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16285531222820282, "tokens_per_second": 4318673.944441477, "eta_seconds": 2470.9536772820866, "mean_data_wait_seconds": 6.723403930664062e-05}
{"step": 26510, "total_steps": 46238, "tokens": 14333755392, "loss": 3.008478446440263, "ce_loss": 3.008478446440263, "z_loss": 24.42238694971258, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16556575894355774, "tokens_per_second": 4318854.123768617, "eta_seconds": 2469.598703346115, "mean_data_wait_seconds": 0.00010928511619567871}
{"step": 26520, "total_steps": 46238, "tokens": 14339162112, "loss": 3.003974285992709, "ce_loss": 3.003974285992709, "z_loss": 24.633963339256518, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1449917107820511, "tokens_per_second": 4318897.473348908, "eta_seconds": 2468.322040470624, "mean_data_wait_seconds": 6.55055046081543e-05}
{"step": 26530, "total_steps": 46238, "tokens": 14344568832, "loss": 2.9695930914445356, "ce_loss": 2.9695930914445356, "z_loss": 24.860661246559836, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15222059190273285, "tokens_per_second": 4318942.278035965, "eta_seconds": 2467.0445720440057, "mean_data_wait_seconds": 0.00010058283805847168}
{"step": 26540, "total_steps": 46238, "tokens": 14349975552, "loss": 2.990841394120997, "ce_loss": 2.990841394120997, "z_loss": 24.150368574893836, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1755441427230835, "tokens_per_second": 4319032.054475867, "eta_seconds": 2465.741455417926, "mean_data_wait_seconds": 0.00015208125114440918}
{"step": 26550, "total_steps": 46238, "tokens": 14355382272, "loss": 2.9805977181954817, "ce_loss": 2.9805977181954817, "z_loss": 24.427894751230877, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15744362771511078, "tokens_per_second": 4318973.254181449, "eta_seconds": 2464.5231719586873, "mean_data_wait_seconds": 0.00010204315185546875}
{"step": 26560, "total_steps": 46238, "tokens": 14360788992, "loss": 2.9895638534517, "ce_loss": 2.9895638534517, "z_loss": 25.40010962341771, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17858047783374786, "tokens_per_second": 4319005.154771212, "eta_seconds": 2463.253124911717, "mean_data_wait_seconds": 7.149577140808105e-05}
{"step": 26570, "total_steps": 46238, "tokens": 14366195712, "loss": 3.0487335486845537, "ce_loss": 3.0487335486845537, "z_loss": 25.260816689693566, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19800546765327454, "tokens_per_second": 4319122.727907852, "eta_seconds": 2461.9342616250988, "mean_data_wait_seconds": 8.07344913482666e-05}
{"step": 26580, "total_steps": 46238, "tokens": 14371602432, "loss": 3.028370013742736, "ce_loss": 3.028370013742736, "z_loss": 24.904923424576268, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15491102635860443, "tokens_per_second": 4319105.686624252, "eta_seconds": 2460.692160627974, "mean_data_wait_seconds": 7.641315460205078e-05}
{"step": 26590, "total_steps": 46238, "tokens": 14377009152, "loss": 2.957627914168618, "ce_loss": 2.957627914168618, "z_loss": 24.08368376529578, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20110853016376495, "tokens_per_second": 4319148.572028589, "eta_seconds": 2459.415925813095, "mean_data_wait_seconds": 0.0001233816146850586}
{"step": 26600, "total_steps": 46238, "tokens": 14382415872, "loss": 2.956330799695217, "ce_loss": 2.956330799695217, "z_loss": 25.447005214113176, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1543889045715332, "tokens_per_second": 4319177.9380364, "eta_seconds": 2458.1474105294255, "mean_data_wait_seconds": 0.00012946128845214844}
{"step": 26610, "total_steps": 46238, "tokens": 14387822592, "loss": 3.000397022926446, "ce_loss": 3.000397022926446, "z_loss": 26.09414633837613, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1781906634569168, "tokens_per_second": 4319174.402650478, "eta_seconds": 2456.8976278170308, "mean_data_wait_seconds": 8.946657180786133e-05}
{"step": 26620, "total_steps": 46238, "tokens": 14393229312, "loss": 2.9536255381324072, "ce_loss": 2.9536255381324072, "z_loss": 24.925843152132902, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13858164846897125, "tokens_per_second": 4319137.444724045, "eta_seconds": 2455.666845461468, "mean_data_wait_seconds": 0.00013241171836853027}
{"step": 26630, "total_steps": 46238, "tokens": 14398636032, "loss": 2.9466157320773965, "ce_loss": 2.9466157320773965, "z_loss": 25.351352633851945, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18024873733520508, "tokens_per_second": 4319127.371886689, "eta_seconds": 2454.420764018652, "mean_data_wait_seconds": 7.560849189758301e-05}
{"step": 26640, "total_steps": 46238, "tokens": 14404042752, "loss": 3.000665003603155, "ce_loss": 3.000665003603155, "z_loss": 24.765480373844955, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16765224933624268, "tokens_per_second": 4319111.302487389, "eta_seconds": 2453.178082699094, "mean_data_wait_seconds": 8.133053779602051e-05}
{"step": 26650, "total_steps": 46238, "tokens": 14409449472, "loss": 2.989784154024991, "ce_loss": 2.989784154024991, "z_loss": 24.992049795208555, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14724403619766235, "tokens_per_second": 4319092.411165991, "eta_seconds": 2451.936994129066, "mean_data_wait_seconds": 6.940960884094238e-05}
{"step": 26660, "total_steps": 46238, "tokens": 14414856192, "loss": 2.9896470019311616, "ce_loss": 2.9896470019311616, "z_loss": 25.29196281144113, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16717292368412018, "tokens_per_second": 4319089.231809815, "eta_seconds": 2450.6869795706234, "mean_data_wait_seconds": 7.2479248046875e-05}
{"step": 26670, "total_steps": 46238, "tokens": 14420262912, "loss": 3.0059699325850517, "ce_loss": 3.0059699325850517, "z_loss": 25.015965577327844, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1622648686170578, "tokens_per_second": 4319188.297006513, "eta_seconds": 2449.3789796874994, "mean_data_wait_seconds": 7.62641429901123e-05}
{"step": 26680, "total_steps": 46238, "tokens": 14425669632, "loss": 3.0714918158271094, "ce_loss": 3.0714918158271094, "z_loss": 24.252860921801943, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15460827946662903, "tokens_per_second": 4319102.498404846, "eta_seconds": 2448.175820764897, "mean_data_wait_seconds": 8.937716484069824e-05}
{"step": 26690, "total_steps": 46238, "tokens": 14431076352, "loss": 3.026465945171587, "ce_loss": 3.026465945171587, "z_loss": 24.957757660836883, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16383479535579681, "tokens_per_second": 4319082.863661337, "eta_seconds": 2446.935128964149, "mean_data_wait_seconds": 7.358193397521973e-05}
{"step": 26700, "total_steps": 46238, "tokens": 14436483072, "loss": 3.019659931009466, "ce_loss": 3.019659931009466, "z_loss": 24.918743017948035, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15990279614925385, "tokens_per_second": 4319063.9930441445, "eta_seconds": 2445.6939931920188, "mean_data_wait_seconds": 0.00013306736946105957}
{"step": 26710, "total_steps": 46238, "tokens": 14441889792, "loss": 3.0324397267717305, "ce_loss": 3.0324397267717305, "z_loss": 25.075670184511125, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17567315697669983, "tokens_per_second": 4319061.863527142, "eta_seconds": 2444.44337163953, "mean_data_wait_seconds": 0.00013637542724609375}
{"step": 26720, "total_steps": 46238, "tokens": 14447296512, "loss": 3.024227787147869, "ce_loss": 3.024227787147869, "z_loss": 24.703911261125043, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18640978634357452, "tokens_per_second": 4319168.518103784, "eta_seconds": 2443.1312137440527, "mean_data_wait_seconds": 0.00014093518257141113}
{"step": 26730, "total_steps": 46238, "tokens": 14452703232, "loss": 3.032478392124176, "ce_loss": 3.032478392124176, "z_loss": 24.762214675094143, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16294452548027039, "tokens_per_second": 4319045.586324732, "eta_seconds": 2441.9489197785515, "mean_data_wait_seconds": 9.459257125854492e-05}
{"step": 26740, "total_steps": 46238, "tokens": 14458109952, "loss": 2.986574279539513, "ce_loss": 2.986574279539513, "z_loss": 24.930205923138242, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1794317066669464, "tokens_per_second": 4319050.489514058, "eta_seconds": 2440.6943168626944, "mean_data_wait_seconds": 0.00022602081298828125}
{"step": 26750, "total_steps": 46238, "tokens": 14463516672, "loss": 3.048710311904098, "ce_loss": 3.048710311904098, "z_loss": 25.188052755413633, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15646232664585114, "tokens_per_second": 4319183.1537527, "eta_seconds": 2439.367558387929, "mean_data_wait_seconds": 0.00012317299842834473}
{"step": 26760, "total_steps": 46238, "tokens": 14468923392, "loss": 2.9749271978031504, "ce_loss": 2.9749271978031504, "z_loss": 25.52451075929584, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17130403220653534, "tokens_per_second": 4319172.184818843, "eta_seconds": 2438.12195795609, "mean_data_wait_seconds": 8.413195610046387e-05}
{"step": 26770, "total_steps": 46238, "tokens": 14474330112, "loss": 2.9458087068615537, "ce_loss": 2.9458087068615537, "z_loss": 24.734425241296943, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1759253591299057, "tokens_per_second": 4319280.556395974, "eta_seconds": 2436.8090209871257, "mean_data_wait_seconds": 7.596611976623535e-05}
{"step": 26780, "total_steps": 46238, "tokens": 14479736832, "loss": 3.011026548616814, "ce_loss": 3.011026548616814, "z_loss": 25.05806317473903, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18200060725212097, "tokens_per_second": 4319491.748239726, "eta_seconds": 2435.4381758657228, "mean_data_wait_seconds": 0.00010097026824951172}
{"step": 26790, "total_steps": 46238, "tokens": 14485143552, "loss": 3.0566943938081916, "ce_loss": 3.0566943938081916, "z_loss": 24.960348562760785, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1523878425359726, "tokens_per_second": 4319729.554108377, "eta_seconds": 2434.0524684004804, "mean_data_wait_seconds": 8.219480514526367e-05}
{"step": 26800, "total_steps": 46238, "tokens": 14490550272, "loss": 3.0221025293523613, "ce_loss": 3.0221025293523613, "z_loss": 25.18915216850512, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1650664508342743, "tokens_per_second": 4319917.185700506, "eta_seconds": 2432.6951680431075, "mean_data_wait_seconds": 7.826089859008789e-05}
{"step": 26810, "total_steps": 46238, "tokens": 14495956992, "loss": 2.99293773282658, "ce_loss": 2.99293773282658, "z_loss": 25.050018729585588, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16195669770240784, "tokens_per_second": 4319996.043561896, "eta_seconds": 2431.3992045556615, "mean_data_wait_seconds": 7.700920104980469e-05}
{"step": 26820, "total_steps": 46238, "tokens": 14501363712, "loss": 3.0294842521349588, "ce_loss": 3.0294842521349588, "z_loss": 24.441657687678482, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18426886200904846, "tokens_per_second": 4320207.794097048, "eta_seconds": 2430.028536670005, "mean_data_wait_seconds": 0.00017827749252319336}
{"step": 26830, "total_steps": 46238, "tokens": 14506770432, "loss": 3.052038575663711, "ce_loss": 3.052038575663711, "z_loss": 25.353740446495287, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1396118849515915, "tokens_per_second": 4320341.212698742, "eta_seconds": 2428.702037042477, "mean_data_wait_seconds": 0.00010135769844055176}
{"step": 26840, "total_steps": 46238, "tokens": 14512177152, "loss": 3.0221964662725274, "ce_loss": 3.0221964662725274, "z_loss": 25.75591532389323, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13736766576766968, "tokens_per_second": 4320508.711638055, "eta_seconds": 2427.3564721094745, "mean_data_wait_seconds": 8.675456047058105e-05}
{"step": 26850, "total_steps": 46238, "tokens": 14517583872, "loss": 3.0292765353665208, "ce_loss": 3.0292765353665208, "z_loss": 24.65517775217692, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14144913852214813, "tokens_per_second": 4320645.349881558, "eta_seconds": 2426.028339559817, "mean_data_wait_seconds": 8.484721183776855e-05}
{"step": 26860, "total_steps": 46238, "tokens": 14522990592, "loss": 3.0687387061841562, "ce_loss": 3.0687387061841562, "z_loss": 25.812082420695912, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17646098136901855, "tokens_per_second": 4320594.858906959, "eta_seconds": 2424.805307167914, "mean_data_wait_seconds": 0.00012195110321044922}
{"step": 26870, "total_steps": 46238, "tokens": 14528397312, "loss": 3.023912493026618, "ce_loss": 3.023912493026618, "z_loss": 25.221820296663225, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18286900222301483, "tokens_per_second": 4320736.4930713475, "eta_seconds": 2423.474479591943, "mean_data_wait_seconds": 8.189678192138672e-05}
{"step": 26880, "total_steps": 46238, "tokens": 14533804032, "loss": 3.026342668316581, "ce_loss": 3.026342668316581, "z_loss": 25.218645861654572, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17632697522640228, "tokens_per_second": 4320928.072595697, "eta_seconds": 2422.115741841757, "mean_data_wait_seconds": 6.0051679611206055e-05}
{"step": 26890, "total_steps": 46238, "tokens": 14539210752, "loss": 3.0524784525235495, "ce_loss": 3.0524784525235495, "z_loss": 24.292185205401797, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16233409941196442, "tokens_per_second": 4321049.86906521, "eta_seconds": 2420.7962187353637, "mean_data_wait_seconds": 5.957484245300293e-05}
{"step": 26900, "total_steps": 46238, "tokens": 14544617472, "loss": 3.074709410017187, "ce_loss": 3.074709410017187, "z_loss": 25.033655426718973, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19389435648918152, "tokens_per_second": 4321157.842993296, "eta_seconds": 2419.4845094475345, "mean_data_wait_seconds": 0.0001220405101776123}
{"step": 26910, "total_steps": 46238, "tokens": 14550024192, "loss": 3.024922150554079, "ce_loss": 3.024922150554079, "z_loss": 26.105091586257473, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1639508455991745, "tokens_per_second": 4321279.870299172, "eta_seconds": 2418.1650014898373, "mean_data_wait_seconds": 7.107853889465332e-05}
{"step": 26920, "total_steps": 46238, "tokens": 14555430912, "loss": 3.022514234889637, "ce_loss": 3.022514234889637, "z_loss": 25.27569442806822, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1787383258342743, "tokens_per_second": 4321402.142358859, "eta_seconds": 2416.845431168089, "mean_data_wait_seconds": 6.976723670959473e-05}
{"step": 26930, "total_steps": 46238, "tokens": 14560837632, "loss": 3.028394198778904, "ce_loss": 3.028394198778904, "z_loss": 25.84132294221358, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15923558175563812, "tokens_per_second": 4321537.641128971, "eta_seconds": 2415.5185424401275, "mean_data_wait_seconds": 0.00011840462684631348}
{"step": 26940, "total_steps": 46238, "tokens": 14566244352, "loss": 3.0503047144774236, "ce_loss": 3.0503047144774236, "z_loss": 25.318949193665475, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13311977684497833, "tokens_per_second": 4321685.422594361, "eta_seconds": 2414.184875524034, "mean_data_wait_seconds": 0.00010886788368225098}
{"step": 26950, "total_steps": 46238, "tokens": 14571651072, "loss": 3.0616403980688616, "ce_loss": 3.0616403980688616, "z_loss": 25.518210194327615, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15253840386867523, "tokens_per_second": 4321735.810901278, "eta_seconds": 2412.905675005919, "mean_data_wait_seconds": 9.66489315032959e-05}
{"step": 26960, "total_steps": 46238, "tokens": 14577057792, "loss": 3.0615647561622388, "ce_loss": 3.0615647561622388, "z_loss": 24.658954056826506, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12815046310424805, "tokens_per_second": 4321786.974689887, "eta_seconds": 2411.6260715853255, "mean_data_wait_seconds": 0.00012344121932983398}
{"step": 26970, "total_steps": 46238, "tokens": 14582464512, "loss": 3.019073766289335, "ce_loss": 3.019073766289335, "z_loss": 25.018687667268694, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1964646428823471, "tokens_per_second": 4321877.662153671, "eta_seconds": 2410.3244557850244, "mean_data_wait_seconds": 0.00013819336891174316}
{"step": 26980, "total_steps": 46238, "tokens": 14587871232, "loss": 3.0404583902070015, "ce_loss": 3.0404583902070015, "z_loss": 25.616905819286, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17264506220817566, "tokens_per_second": 4321975.324037556, "eta_seconds": 2409.0190071408024, "mean_data_wait_seconds": 0.00017699599266052246}
{"step": 26990, "total_steps": 46238, "tokens": 14593277952, "loss": 3.0660879322976777, "ce_loss": 3.0660879322976777, "z_loss": 25.467356927467115, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1842532753944397, "tokens_per_second": 4322086.699508475, "eta_seconds": 2407.7059780368236, "mean_data_wait_seconds": 8.749961853027344e-05}
{"step": 27000, "total_steps": 46238, "tokens": 14598684672, "loss": 3.0603826678160466, "ce_loss": 3.0603826678160466, "z_loss": 25.5808775063717, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1561451256275177, "tokens_per_second": 4322165.195216046, "eta_seconds": 2406.411322619543, "mean_data_wait_seconds": 6.181001663208008e-05}
{"step": 27010, "total_steps": 46238, "tokens": 14604091392, "loss": 2.9979824188983804, "ce_loss": 2.9979824188983804, "z_loss": 25.53181789860581, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16032390296459198, "tokens_per_second": 4322169.999773805, "eta_seconds": 2405.1577204376586, "mean_data_wait_seconds": 7.030367851257324e-05}
{"step": 27020, "total_steps": 46238, "tokens": 14609498112, "loss": 2.9959549181389087, "ce_loss": 2.9959549181389087, "z_loss": 25.18884781635169, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14857296645641327, "tokens_per_second": 4322296.650144412, "eta_seconds": 2403.8363548353063, "mean_data_wait_seconds": 7.212162017822266e-05}
{"step": 27030, "total_steps": 46238, "tokens": 14614904832, "loss": 2.9538877895384124, "ce_loss": 2.9538877895384124, "z_loss": 25.8521728515625, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15846805274486542, "tokens_per_second": 4322309.954099246, "eta_seconds": 2402.578069199142, "mean_data_wait_seconds": 7.897615432739258e-05}
{"step": 27040, "total_steps": 46238, "tokens": 14620311552, "loss": 2.953508209098469, "ce_loss": 2.953508209098469, "z_loss": 24.69227274981412, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16082149744033813, "tokens_per_second": 4322336.672649093, "eta_seconds": 2401.3123386889483, "mean_data_wait_seconds": 7.772445678710938e-05}
{"step": 27050, "total_steps": 46238, "tokens": 14625718272, "loss": 3.001789027994329, "ce_loss": 3.001789027994329, "z_loss": 25.71803244677457, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13781268894672394, "tokens_per_second": 4322439.437500465, "eta_seconds": 2400.004398904637, "mean_data_wait_seconds": 8.222460746765137e-05}
{"step": 27060, "total_steps": 46238, "tokens": 14631124992, "loss": 3.049964265389876, "ce_loss": 3.049964265389876, "z_loss": 26.227447639812123, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.24624085426330566, "tokens_per_second": 4322575.216489332, "eta_seconds": 2398.678201005595, "mean_data_wait_seconds": 8.410215377807617e-05}
{"step": 27070, "total_steps": 46238, "tokens": 14636531712, "loss": 2.983892099423842, "ce_loss": 2.983892099423842, "z_loss": 25.64396215207649, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13394160568714142, "tokens_per_second": 4322431.262557502, "eta_seconds": 2397.507234821444, "mean_data_wait_seconds": 8.562207221984863e-05}
{"step": 27080, "total_steps": 46238, "tokens": 14641938432, "loss": 2.9495997085715784, "ce_loss": 2.9495997085715784, "z_loss": 25.46987841346047, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16795293986797333, "tokens_per_second": 4322448.613241959, "eta_seconds": 2396.246764454064, "mean_data_wait_seconds": 8.943676948547363e-05}
{"step": 27090, "total_steps": 46238, "tokens": 14647345152, "loss": 2.985337658361955, "ce_loss": 2.985337658361955, "z_loss": 26.57041595921372, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19116400182247162, "tokens_per_second": 4322471.829591183, "eta_seconds": 2394.9830541704446, "mean_data_wait_seconds": 0.00010603666305541992}
{"step": 27100, "total_steps": 46238, "tokens": 14652751872, "loss": 2.9883332974982983, "ce_loss": 2.9883332974982983, "z_loss": 25.671597133983266, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15799692273139954, "tokens_per_second": 4322326.912717007, "eta_seconds": 2393.8124702131786, "mean_data_wait_seconds": 0.0001271069049835205}
{"step": 27110, "total_steps": 46238, "tokens": 14658158592, "loss": 2.970743269631357, "ce_loss": 2.970743269631357, "z_loss": 24.80615335522276, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1821182668209076, "tokens_per_second": 4322344.151186631, "eta_seconds": 2392.5520463614457, "mean_data_wait_seconds": 9.500980377197266e-05}
{"step": 27120, "total_steps": 46238, "tokens": 14663565312, "loss": 2.9244925054636868, "ce_loss": 2.9244925054636868, "z_loss": 25.64778806224014, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.180758535861969, "tokens_per_second": 4322401.4897736525, "eta_seconds": 2391.269447887697, "mean_data_wait_seconds": 8.350610733032227e-05}
{"step": 27130, "total_steps": 46238, "tokens": 14668972032, "loss": 2.9751574126156894, "ce_loss": 2.9751574126156894, "z_loss": 26.318250916220926, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1546752154827118, "tokens_per_second": 4322536.331485831, "eta_seconds": 2389.9440309502147, "mean_data_wait_seconds": 8.150935173034668e-05}
{"step": 27140, "total_steps": 46238, "tokens": 14674378752, "loss": 2.947943116679336, "ce_loss": 2.947943116679336, "z_loss": 26.006974798260313, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14263854920864105, "tokens_per_second": 4322619.760063853, "eta_seconds": 2388.647106875641, "mean_data_wait_seconds": 7.930397987365723e-05}
{"step": 27150, "total_steps": 46238, "tokens": 14679785472, "loss": 2.9772439472603076, "ce_loss": 2.9772439472603076, "z_loss": 25.341937585310504, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1928824931383133, "tokens_per_second": 4322770.114204394, "eta_seconds": 2387.313271665699, "mean_data_wait_seconds": 0.0001080930233001709}
{"step": 27160, "total_steps": 46238, "tokens": 14685192192, "loss": 3.0124609560677498, "ce_loss": 3.0124609560677498, "z_loss": 25.191596999312893, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15719188749790192, "tokens_per_second": 4322925.801366019, "eta_seconds": 2385.976585751416, "mean_data_wait_seconds": 9.313225746154785e-05}
{"step": 27170, "total_steps": 46238, "tokens": 14690598912, "loss": 3.053375323613485, "ce_loss": 3.053375323613485, "z_loss": 25.735964673938174, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1871529221534729, "tokens_per_second": 4322904.239052148, "eta_seconds": 2384.73777209101, "mean_data_wait_seconds": 0.00013014674186706543}
{"step": 27180, "total_steps": 46238, "tokens": 14696005632, "loss": 2.987578991687659, "ce_loss": 2.987578991687659, "z_loss": 26.266594727834065, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14751219749450684, "tokens_per_second": 4322899.825081174, "eta_seconds": 2383.4894910632174, "mean_data_wait_seconds": 7.489323616027832e-05}
{"step": 27190, "total_steps": 46238, "tokens": 14701412352, "loss": 2.9421574690125207, "ce_loss": 2.9421574690125207, "z_loss": 25.394409613175824, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17337921261787415, "tokens_per_second": 4322918.377140079, "eta_seconds": 2382.2285515399867, "mean_data_wait_seconds": 8.38935375213623e-05}
{"step": 27200, "total_steps": 46238, "tokens": 14706819072, "loss": 2.9859513893271936, "ce_loss": 2.9859513893271936, "z_loss": 25.27018185817834, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14619705080986023, "tokens_per_second": 4322923.604490347, "eta_seconds": 2380.9749617847965, "mean_data_wait_seconds": 6.452202796936035e-05}
{"step": 27210, "total_steps": 46238, "tokens": 14712225792, "loss": 2.9858843467452307, "ce_loss": 2.9858843467452307, "z_loss": 25.797288331118498, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1656106412410736, "tokens_per_second": 4323059.901631273, "eta_seconds": 2379.649224873831, "mean_data_wait_seconds": 9.745359420776367e-05}
{"step": 27220, "total_steps": 46238, "tokens": 14717632512, "loss": 3.0145045266006933, "ce_loss": 3.0145045266006933, "z_loss": 25.55619290380767, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15069229900836945, "tokens_per_second": 4323088.720900044, "eta_seconds": 2378.382699918162, "mean_data_wait_seconds": 8.177757263183594e-05}
{"step": 27230, "total_steps": 46238, "tokens": 14723039232, "loss": 2.9711547096570334, "ce_loss": 2.9711547096570334, "z_loss": 25.305021719499067, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15713746845722198, "tokens_per_second": 4323112.358046354, "eta_seconds": 2377.1190413020054, "mean_data_wait_seconds": 0.00010451674461364746}
{"step": 27240, "total_steps": 46238, "tokens": 14728445952, "loss": 2.992018795374668, "ce_loss": 2.992018795374668, "z_loss": 25.917286280429725, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18669891357421875, "tokens_per_second": 4323269.918254463, "eta_seconds": 2375.781799010832, "mean_data_wait_seconds": 0.0001341700553894043}
{"step": 27250, "total_steps": 46238, "tokens": 14733852672, "loss": 3.074693242708842, "ce_loss": 3.074693242708842, "z_loss": 24.804649642019562, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1597856730222702, "tokens_per_second": 4323400.694627322, "eta_seconds": 2374.4593640735648, "mean_data_wait_seconds": 0.00011578202247619629}
{"step": 27260, "total_steps": 46238, "tokens": 14739259392, "loss": 2.9795998714186926, "ce_loss": 2.9795998714186926, "z_loss": 25.69224416848385, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16149482131004333, "tokens_per_second": 4323586.101505612, "eta_seconds": 2373.107023456066, "mean_data_wait_seconds": 0.00010412931442260742}
{"step": 27270, "total_steps": 46238, "tokens": 14744666112, "loss": 3.017043502041788, "ce_loss": 3.017043502041788, "z_loss": 25.51745453747836, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15247146785259247, "tokens_per_second": 4323767.996853502, "eta_seconds": 2371.756725028431, "mean_data_wait_seconds": 7.572770118713379e-05}
{"step": 27280, "total_steps": 46238, "tokens": 14750072832, "loss": 3.0383480985959372, "ce_loss": 3.0383480985959372, "z_loss": 26.2684708219586, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17251309752464294, "tokens_per_second": 4323998.872458299, "eta_seconds": 2370.379689339026, "mean_data_wait_seconds": 7.30752944946289e-05}
{"step": 27290, "total_steps": 46238, "tokens": 14755479552, "loss": 2.957383594729684, "ce_loss": 2.957383594729684, "z_loss": 26.110954544760965, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1682378649711609, "tokens_per_second": 4324193.6524866335, "eta_seconds": 2369.0225755983683, "mean_data_wait_seconds": 7.078051567077637e-05}
{"step": 27300, "total_steps": 46238, "tokens": 14760886272, "loss": 3.0395627762332107, "ce_loss": 3.0395627762332107, "z_loss": 26.130187713738643, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16485954821109772, "tokens_per_second": 4324299.445836249, "eta_seconds": 2367.714306616433, "mean_data_wait_seconds": 7.125735282897949e-05}
{"step": 27310, "total_steps": 46238, "tokens": 14766292992, "loss": 2.96709125511574, "ce_loss": 2.96709125511574, "z_loss": 25.381481358499236, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14968715608119965, "tokens_per_second": 4324322.084116703, "eta_seconds": 2366.4516067355516, "mean_data_wait_seconds": 9.259581565856934e-05}
{"step": 27320, "total_steps": 46238, "tokens": 14771699712, "loss": 3.0382665702790925, "ce_loss": 3.0382665702790925, "z_loss": 25.117274833448004, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15911856293678284, "tokens_per_second": 4324367.938296347, "eta_seconds": 2365.176222268783, "mean_data_wait_seconds": 7.06017017364502e-05}
{"step": 27330, "total_steps": 46238, "tokens": 14777106432, "loss": 3.041697559934674, "ce_loss": 3.041697559934674, "z_loss": 25.98459264003869, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15455183386802673, "tokens_per_second": 4324492.411473888, "eta_seconds": 2363.8578892813775, "mean_data_wait_seconds": 0.000152587890625}
{"step": 27340, "total_steps": 46238, "tokens": 14782513152, "loss": 3.0112242499987283, "ce_loss": 3.0112242499987283, "z_loss": 25.58026188070124, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16342604160308838, "tokens_per_second": 4324648.214546087, "eta_seconds": 2362.522516775941, "mean_data_wait_seconds": 8.818507194519043e-05}
{"step": 27350, "total_steps": 46238, "tokens": 14787919872, "loss": 3.026986194379402, "ce_loss": 3.026986194379402, "z_loss": 25.595702951604668, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14924243092536926, "tokens_per_second": 4324636.338200092, "eta_seconds": 2361.2787909584285, "mean_data_wait_seconds": 7.858872413635254e-05}
{"step": 27360, "total_steps": 46238, "tokens": 14793326592, "loss": 3.0740561557538584, "ce_loss": 3.0740561557538584, "z_loss": 25.857541835669316, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1595579981803894, "tokens_per_second": 4324871.019417204, "eta_seconds": 2359.9005145303363, "mean_data_wait_seconds": 0.00013393163681030273}
{"step": 27370, "total_steps": 46238, "tokens": 14798733312, "loss": 3.0028540166941555, "ce_loss": 3.0028540166941555, "z_loss": 26.196152773770418, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14608895778656006, "tokens_per_second": 4325077.512023052, "eta_seconds": 2358.5377592986893, "mean_data_wait_seconds": 8.270144462585449e-05}
{"step": 27380, "total_steps": 46238, "tokens": 14804140032, "loss": 3.0142728957262905, "ce_loss": 3.0142728957262905, "z_loss": 25.729643662770588, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16684715449810028, "tokens_per_second": 4325221.951251971, "eta_seconds": 2357.2089522593965, "mean_data_wait_seconds": 0.00011023879051208496}
{"step": 27390, "total_steps": 46238, "tokens": 14809546752, "loss": 3.0525594462047922, "ce_loss": 3.0525594462047922, "z_loss": 26.1349734248537, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16260680556297302, "tokens_per_second": 4325378.77490047, "eta_seconds": 2355.8734886136026, "mean_data_wait_seconds": 8.493661880493164e-05}
{"step": 27400, "total_steps": 46238, "tokens": 14814953472, "loss": 3.0115347775545986, "ce_loss": 3.0115347775545986, "z_loss": 26.328674056313254, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16612893342971802, "tokens_per_second": 4325470.546325931, "eta_seconds": 2354.5735325029236, "mean_data_wait_seconds": 8.356571197509766e-05}
{"step": 27410, "total_steps": 46238, "tokens": 14820360192, "loss": 3.0150810769110015, "ce_loss": 3.0150810769110015, "z_loss": 26.021853865999162, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19312599301338196, "tokens_per_second": 4325514.303508234, "eta_seconds": 2353.299753452225, "mean_data_wait_seconds": 7.978081703186035e-05}
{"step": 27420, "total_steps": 46238, "tokens": 14825766912, "loss": 3.043982298085184, "ce_loss": 3.043982298085184, "z_loss": 25.404684442462344, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.151100754737854, "tokens_per_second": 4325579.571959774, "eta_seconds": 2352.014303459128, "mean_data_wait_seconds": 0.00010710954666137695}
{"step": 27430, "total_steps": 46238, "tokens": 14831173632, "loss": 3.0760530106949084, "ce_loss": 3.0760530106949084, "z_loss": 25.591444969177246, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18375541269779205, "tokens_per_second": 4325633.571970256, "eta_seconds": 2350.735015996385, "mean_data_wait_seconds": 0.00010880827903747559}
{"step": 27440, "total_steps": 46238, "tokens": 14836580352, "loss": 2.991968519759901, "ce_loss": 2.991968519759901, "z_loss": 25.72369413664847, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15729007124900818, "tokens_per_second": 4325521.314436835, "eta_seconds": 2349.5460651367016, "mean_data_wait_seconds": 0.0001455247402191162}
{"step": 27450, "total_steps": 46238, "tokens": 14841987072, "loss": 3.042447507381439, "ce_loss": 3.042447507381439, "z_loss": 25.62198612906716, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17728197574615479, "tokens_per_second": 4325628.344627069, "eta_seconds": 2348.238002605314, "mean_data_wait_seconds": 0.00014281272888183594}
{"step": 27460, "total_steps": 46238, "tokens": 14847393792, "loss": 3.064710487018932, "ce_loss": 3.064710487018932, "z_loss": 25.936723506811894, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14976520836353302, "tokens_per_second": 4325644.237920278, "eta_seconds": 2346.9794522170564, "mean_data_wait_seconds": 7.802248001098633e-05}
{"step": 27470, "total_steps": 46238, "tokens": 14852800512, "loss": 3.0188853072397634, "ce_loss": 3.0188853072397634, "z_loss": 25.003352613160104, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1569792628288269, "tokens_per_second": 4325590.423236986, "eta_seconds": 2345.758712958961, "mean_data_wait_seconds": 8.291006088256836e-05}
{"step": 27480, "total_steps": 46238, "tokens": 14858207232, "loss": 3.015170693397522, "ce_loss": 3.015170693397522, "z_loss": 25.632399544571385, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1438370794057846, "tokens_per_second": 4325642.141843296, "eta_seconds": 2344.480743309577, "mean_data_wait_seconds": 0.0001240074634552002}
{"step": 27490, "total_steps": 46238, "tokens": 14863613952, "loss": 3.0294795018253904, "ce_loss": 3.0294795018253904, "z_loss": 25.7176620165507, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17076550424098969, "tokens_per_second": 4325727.90878944, "eta_seconds": 2343.184360580036, "mean_data_wait_seconds": 7.30752944946289e-05}
{"step": 27500, "total_steps": 46238, "tokens": 14869020672, "loss": 2.9243520371841663, "ce_loss": 2.9243520371841663, "z_loss": 25.358741557959355, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14836467802524567, "tokens_per_second": 4325672.418834118, "eta_seconds": 2341.9645047302156, "mean_data_wait_seconds": 6.955862045288086e-05}
{"step": 27510, "total_steps": 46238, "tokens": 14874427392, "loss": 2.9777372298818645, "ce_loss": 2.9777372298818645, "z_loss": 26.138770666989412, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1617298126220703, "tokens_per_second": 4325780.53403233, "eta_seconds": 2340.6560883849793, "mean_data_wait_seconds": 5.963444709777832e-05}
{"step": 27520, "total_steps": 46238, "tokens": 14879834112, "loss": 3.029464461586692, "ce_loss": 3.029464461586692, "z_loss": 25.956362825451475, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18094637989997864, "tokens_per_second": 4325856.186755222, "eta_seconds": 2339.3652925828587, "mean_data_wait_seconds": 9.042024612426758e-05}
{"step": 27530, "total_steps": 46238, "tokens": 14885240832, "loss": 2.980151987437046, "ce_loss": 2.980151987437046, "z_loss": 26.38449813380386, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12882980704307556, "tokens_per_second": 4325970.727833973, "eta_seconds": 2338.0535237843105, "mean_data_wait_seconds": 9.265542030334473e-05}
{"step": 27540, "total_steps": 46238, "tokens": 14890647552, "loss": 3.0533189827745613, "ce_loss": 3.0533189827745613, "z_loss": 26.52895553184278, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16285443305969238, "tokens_per_second": 4326044.158487522, "eta_seconds": 2336.764030521201, "mean_data_wait_seconds": 0.0001945197582244873}
{"step": 27550, "total_steps": 46238, "tokens": 14896054272, "loss": 2.9981941978136697, "ce_loss": 2.9981941978136697, "z_loss": 25.771894642800994, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14486655592918396, "tokens_per_second": 4326145.257716582, "eta_seconds": 2335.459644120417, "mean_data_wait_seconds": 0.00011366605758666992}
{"step": 27560, "total_steps": 46238, "tokens": 14901460992, "loss": 3.038193469697779, "ce_loss": 3.038193469697779, "z_loss": 26.70220149647106, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15095144510269165, "tokens_per_second": 4326319.214971576, "eta_seconds": 2334.116009991728, "mean_data_wait_seconds": 9.02414321899414e-05}
{"step": 27570, "total_steps": 46238, "tokens": 14906867712, "loss": 3.0302782383832065, "ce_loss": 3.0302782383832065, "z_loss": 25.387024734959457, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18951217830181122, "tokens_per_second": 4326326.485777607, "eta_seconds": 2332.8623619088585, "mean_data_wait_seconds": 9.15229320526123e-05}
{"step": 27580, "total_steps": 46238, "tokens": 14912274432, "loss": 3.012983818848928, "ce_loss": 3.012983818848928, "z_loss": 26.542996926741168, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1494879275560379, "tokens_per_second": 4326388.788564997, "eta_seconds": 2331.579059806556, "mean_data_wait_seconds": 8.64267349243164e-05}
{"step": 27590, "total_steps": 46238, "tokens": 14917681152, "loss": 2.9212267742012488, "ce_loss": 2.9212267742012488, "z_loss": 26.387073068907767, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19697721302509308, "tokens_per_second": 4326517.274168245, "eta_seconds": 2330.260148086016, "mean_data_wait_seconds": 6.815791130065918e-05}
{"step": 27600, "total_steps": 46238, "tokens": 14923087872, "loss": 3.0150699814160666, "ce_loss": 3.0150699814160666, "z_loss": 25.44817931724317, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15165814757347107, "tokens_per_second": 4326728.991173839, "eta_seconds": 2328.896513868841, "mean_data_wait_seconds": 5.6296586990356445e-05}
{"step": 27610, "total_steps": 46238, "tokens": 14928494592, "loss": 2.967973651308002, "ce_loss": 2.967973651308002, "z_loss": 25.70192779194225, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16392719745635986, "tokens_per_second": 4326957.392516216, "eta_seconds": 2327.5240383505247, "mean_data_wait_seconds": 6.288290023803711e-05}
{"step": 27620, "total_steps": 46238, "tokens": 14933901312, "loss": 2.9715927044550576, "ce_loss": 2.9715927044550576, "z_loss": 26.794854034077037, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1549195796251297, "tokens_per_second": 4327096.375148902, "eta_seconds": 2326.199777247537, "mean_data_wait_seconds": 0.00014337897300720215}
{"step": 27630, "total_steps": 46238, "tokens": 14939308032, "loss": 2.9125193899328057, "ce_loss": 2.9125193899328057, "z_loss": 26.58627713810314, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1598491221666336, "tokens_per_second": 4327198.1765843695, "eta_seconds": 2324.8955775676964, "mean_data_wait_seconds": 0.0001195073127746582}
{"step": 27640, "total_steps": 46238, "tokens": 14944714752, "loss": 2.8957177581209126, "ce_loss": 2.8957177581209126, "z_loss": 27.728717298218697, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17476961016654968, "tokens_per_second": 4327402.1988078095, "eta_seconds": 2323.5365519687766, "mean_data_wait_seconds": 0.00011703372001647949}
{"step": 27650, "total_steps": 46238, "tokens": 14950121472, "loss": 2.9473760200269297, "ce_loss": 2.9473760200269297, "z_loss": 26.125072334751938, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15019552409648895, "tokens_per_second": 4327387.069736636, "eta_seconds": 2322.2952562483874, "mean_data_wait_seconds": 0.00014641880989074707}
{"step": 27660, "total_steps": 46238, "tokens": 14955528192, "loss": 2.88864216298768, "ce_loss": 2.88864216298768, "z_loss": 25.679309555978485, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18055354058742523, "tokens_per_second": 4327657.454897706, "eta_seconds": 2320.900821906066, "mean_data_wait_seconds": 0.00023499131202697754}
{"step": 27670, "total_steps": 46238, "tokens": 14960934912, "loss": 2.9249465465545654, "ce_loss": 2.9249465465545654, "z_loss": 26.13422357674801, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1988288313150406, "tokens_per_second": 4327638.210554686, "eta_seconds": 2319.661796015364, "mean_data_wait_seconds": 0.0001100301742553711}
{"step": 27680, "total_steps": 46238, "tokens": 14966341632, "loss": 2.973313989061298, "ce_loss": 2.973313989061298, "z_loss": 25.87884457906087, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18412639200687408, "tokens_per_second": 4327927.862455216, "eta_seconds": 2318.257286827368, "mean_data_wait_seconds": 7.137656211853027e-05}
{"step": 27690, "total_steps": 46238, "tokens": 14971748352, "loss": 2.9079441446246523, "ce_loss": 2.9079441446246523, "z_loss": 27.35618543624878, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15017451345920563, "tokens_per_second": 4327922.525192877, "eta_seconds": 2317.010881231776, "mean_data_wait_seconds": 0.00011649727821350098}
{"step": 27700, "total_steps": 46238, "tokens": 14977155072, "loss": 2.9148543458996397, "ce_loss": 2.9148543458996397, "z_loss": 26.34602604490338, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.22838547825813293, "tokens_per_second": 4328030.150518448, "eta_seconds": 2315.7040305736846, "mean_data_wait_seconds": 7.942318916320801e-05}
{"step": 27710, "total_steps": 46238, "tokens": 14982561792, "loss": 2.9352048472924666, "ce_loss": 2.9352048472924666, "z_loss": 26.640255133310955, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1444450467824936, "tokens_per_second": 4328177.073113714, "eta_seconds": 2314.376231560622, "mean_data_wait_seconds": 0.0001518726348876953}
{"step": 27720, "total_steps": 46238, "tokens": 14987968512, "loss": 2.9280506643381985, "ce_loss": 2.9280506643381985, "z_loss": 26.79846513632572, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16850732266902924, "tokens_per_second": 4328407.443580986, "eta_seconds": 2313.003929158103, "mean_data_wait_seconds": 0.00011998414993286133}
{"step": 27730, "total_steps": 46238, "tokens": 14993375232, "loss": 2.9334135272286157, "ce_loss": 2.9334135272286157, "z_loss": 26.14905407934478, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1844053864479065, "tokens_per_second": 4328572.666084291, "eta_seconds": 2311.6665644547934, "mean_data_wait_seconds": 0.00010764598846435547}
{"step": 27740, "total_steps": 46238, "tokens": 14998781952, "loss": 2.916528409177607, "ce_loss": 2.916528409177607, "z_loss": 27.484068726048324, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19720017910003662, "tokens_per_second": 4328724.668692315, "eta_seconds": 2310.3363575723083, "mean_data_wait_seconds": 8.946657180786133e-05}
{"step": 27750, "total_steps": 46238, "tokens": 15004188672, "loss": 2.9975050687789917, "ce_loss": 2.9975050687789917, "z_loss": 26.046109488516144, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16352123022079468, "tokens_per_second": 4328766.875986808, "eta_seconds": 2309.064809991967, "mean_data_wait_seconds": 0.00012126564979553223}
{"step": 27760, "total_steps": 46238, "tokens": 15009595392, "loss": 3.0082324038852346, "ce_loss": 3.0082324038852346, "z_loss": 27.183785033948492, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1467643827199936, "tokens_per_second": 4328853.280684987, "eta_seconds": 2307.769724738559, "mean_data_wait_seconds": 7.599592208862305e-05}
{"step": 27770, "total_steps": 46238, "tokens": 15015002112, "loss": 2.988978700204329, "ce_loss": 2.988978700204329, "z_loss": 26.371437058304295, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.6310884952545166, "tokens_per_second": 4328982.333150061, "eta_seconds": 2306.451968524098, "mean_data_wait_seconds": 9.09566879272461e-05}
{"step": 27780, "total_steps": 46238, "tokens": 15020408832, "loss": 3.0073590116067366, "ce_loss": 3.0073590116067366, "z_loss": 26.99198344259551, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2799627184867859, "tokens_per_second": 4329083.287901181, "eta_seconds": 2305.1492522422895, "mean_data_wait_seconds": 0.00012227892875671387}
{"step": 27790, "total_steps": 46238, "tokens": 15025815552, "loss": 3.026407373673988, "ce_loss": 3.026407373673988, "z_loss": 26.014893054962158, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1997058242559433, "tokens_per_second": 4329251.272477812, "eta_seconds": 2303.8109262462813, "mean_data_wait_seconds": 9.867548942565918e-05}
{"step": 27800, "total_steps": 46238, "tokens": 15031222272, "loss": 2.9941233974514585, "ce_loss": 2.9941233974514585, "z_loss": 25.959898789723713, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14746330678462982, "tokens_per_second": 4328591.845907392, "eta_seconds": 2302.9128221975743, "mean_data_wait_seconds": 0.00013566017150878906}
{"step": 27810, "total_steps": 46238, "tokens": 15036628992, "loss": 3.0139503009391553, "ce_loss": 3.0139503009391553, "z_loss": 26.0379695892334, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1629265695810318, "tokens_per_second": 4327861.65831877, "eta_seconds": 2302.0520826606735, "mean_data_wait_seconds": 0.00012367963790893555}
{"step": 27820, "total_steps": 46238, "tokens": 15042035712, "loss": 3.03828479123838, "ce_loss": 3.03828479123838, "z_loss": 25.913220044338342, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18103334307670593, "tokens_per_second": 4327868.887884565, "eta_seconds": 2300.798957166928, "mean_data_wait_seconds": 0.00011807680130004883}
{"step": 27830, "total_steps": 46238, "tokens": 15047442432, "loss": 3.065989243261742, "ce_loss": 3.065989243261742, "z_loss": 27.070212291948724, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18304672837257385, "tokens_per_second": 4328033.543394366, "eta_seconds": 2299.462193214608, "mean_data_wait_seconds": 0.00015652179718017578}
{"step": 27840, "total_steps": 46238, "tokens": 15052849152, "loss": 3.0148033409407646, "ce_loss": 3.0148033409407646, "z_loss": 26.28484393611099, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1747671365737915, "tokens_per_second": 4327347.013222202, "eta_seconds": 2298.5775704161797, "mean_data_wait_seconds": 0.00013697147369384766}
{"step": 27850, "total_steps": 46238, "tokens": 15058255872, "loss": 2.981788192734574, "ce_loss": 2.981788192734574, "z_loss": 26.710947889270205, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15024219453334808, "tokens_per_second": 4327320.408218913, "eta_seconds": 2297.342264076019, "mean_data_wait_seconds": 6.4849853515625e-05}
{"step": 27860, "total_steps": 46238, "tokens": 15063662592, "loss": 2.995758856787826, "ce_loss": 2.995758856787826, "z_loss": 26.369818904183127, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1387697160243988, "tokens_per_second": 4326614.527010249, "eta_seconds": 2296.4674301285318, "mean_data_wait_seconds": 0.0001665651798248291}
{"step": 27870, "total_steps": 46238, "tokens": 15069069312, "loss": 3.0160898241129788, "ce_loss": 3.0160898241129788, "z_loss": 26.853735403581098, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1542081981897354, "tokens_per_second": 4326164.050075182, "eta_seconds": 2295.4567855158944, "mean_data_wait_seconds": 0.00017574429512023926}
{"step": 27880, "total_steps": 46238, "tokens": 15074476032, "loss": 3.035550989887931, "ce_loss": 3.035550989887931, "z_loss": 25.861696734572902, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16057667136192322, "tokens_per_second": 4326000.923100024, "eta_seconds": 2294.2935243036504, "mean_data_wait_seconds": 0.00011366605758666992}
{"step": 27890, "total_steps": 46238, "tokens": 15079882752, "loss": 3.0190164121714504, "ce_loss": 3.0190164121714504, "z_loss": 26.184731136668812, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18334412574768066, "tokens_per_second": 4325962.319893966, "eta_seconds": 2293.064167106093, "mean_data_wait_seconds": 9.608268737792969e-05}
{"step": 27900, "total_steps": 46238, "tokens": 15085289472, "loss": 2.9817665706981313, "ce_loss": 2.9817665706981313, "z_loss": 25.823749513337106, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16805750131607056, "tokens_per_second": 4325844.3973717615, "eta_seconds": 2291.876811385911, "mean_data_wait_seconds": 7.94827938079834e-05}
{"step": 27910, "total_steps": 46238, "tokens": 15090696192, "loss": 2.983434608488372, "ce_loss": 2.983434608488372, "z_loss": 26.64013197927764, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1658133566379547, "tokens_per_second": 4325864.850377752, "eta_seconds": 2290.616116482399, "mean_data_wait_seconds": 0.00015369057655334473}
{"step": 27920, "total_steps": 46238, "tokens": 15096102912, "loss": 3.018956077821327, "ce_loss": 3.018956077821327, "z_loss": 25.296042774662826, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16877347230911255, "tokens_per_second": 4325866.837470369, "eta_seconds": 2289.365206116989, "mean_data_wait_seconds": 0.00017243623733520508}
{"step": 27930, "total_steps": 46238, "tokens": 15101509632, "loss": 3.0162738344886084, "ce_loss": 3.0162738344886084, "z_loss": 26.369471448840518, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1470247358083725, "tokens_per_second": 4325764.066301919, "eta_seconds": 2288.1697088167452, "mean_data_wait_seconds": 8.377432823181152e-05}
{"step": 27940, "total_steps": 46238, "tokens": 15106916352, "loss": 3.015364504221714, "ce_loss": 3.015364504221714, "z_loss": 26.094760995922666, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16777487099170685, "tokens_per_second": 4325743.764660074, "eta_seconds": 2286.930553959288, "mean_data_wait_seconds": 8.293986320495605e-05}
{"step": 27950, "total_steps": 46238, "tokens": 15112323072, "loss": 3.027639925479889, "ce_loss": 3.027639925479889, "z_loss": 26.215295141393486, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18535427749156952, "tokens_per_second": 4325819.423552861, "eta_seconds": 2285.640683512266, "mean_data_wait_seconds": 0.0005464255809783936}
{"step": 27960, "total_steps": 46238, "tokens": 15117729792, "loss": 3.0237260804031836, "ce_loss": 3.0237260804031836, "z_loss": 26.52599879467126, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18422195315361023, "tokens_per_second": 4325839.234836969, "eta_seconds": 2284.3803496947166, "mean_data_wait_seconds": 8.282065391540527e-05}
{"step": 27970, "total_steps": 46238, "tokens": 15123136512, "loss": 3.095861225417166, "ce_loss": 3.095861225417166, "z_loss": 26.141704631574225, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15615762770175934, "tokens_per_second": 4325920.1236589635, "eta_seconds": 2283.0877921172214, "mean_data_wait_seconds": 0.0001621842384338379}
{"step": 27980, "total_steps": 46238, "tokens": 15128543232, "loss": 3.040743547858614, "ce_loss": 3.040743547858614, "z_loss": 27.14441072579586, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15461465716362, "tokens_per_second": 4326146.234137067, "eta_seconds": 2281.718686739902, "mean_data_wait_seconds": 6.183981895446777e-05}
{"step": 27990, "total_steps": 46238, "tokens": 15133949952, "loss": 3.002589948249586, "ce_loss": 3.002589948249586, "z_loss": 25.874350114302203, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15594834089279175, "tokens_per_second": 4326308.846679971, "eta_seconds": 2280.383193532505, "mean_data_wait_seconds": 7.852911949157715e-05}
{"step": 28000, "total_steps": 46238, "tokens": 15139356672, "loss": 3.016947876323353, "ce_loss": 3.016947876323353, "z_loss": 25.965175252972227, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15300263464450836, "tokens_per_second": 4326259.583955397, "eta_seconds": 2279.15941534535, "mean_data_wait_seconds": 7.176399230957031e-05}
{"step": 28010, "total_steps": 46238, "tokens": 15144763392, "loss": 2.9860939437692817, "ce_loss": 2.9860939437692817, "z_loss": 25.85794818762577, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1642482727766037, "tokens_per_second": 4326431.014364133, "eta_seconds": 2277.8194107986697, "mean_data_wait_seconds": 7.963180541992188e-05}
{"step": 28020, "total_steps": 46238, "tokens": 15150170112, "loss": 3.0577872756755715, "ce_loss": 3.0577872756755715, "z_loss": 26.50764513015747, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.157074436545372, "tokens_per_second": 4326489.832126924, "eta_seconds": 2276.5387661059117, "mean_data_wait_seconds": 0.00011041760444641113}
{"step": 28030, "total_steps": 46238, "tokens": 15155576832, "loss": 3.0092909065159885, "ce_loss": 3.0092909065159885, "z_loss": 25.67548234534986, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17187339067459106, "tokens_per_second": 4326410.562114849, "eta_seconds": 2275.330776556726, "mean_data_wait_seconds": 7.233023643493652e-05}
{"step": 28040, "total_steps": 46238, "tokens": 15160983552, "loss": 3.0572001590873255, "ce_loss": 3.0572001590873255, "z_loss": 26.694544387586188, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17618624866008759, "tokens_per_second": 4326581.401010362, "eta_seconds": 2273.9912813618726, "mean_data_wait_seconds": 0.00012093782424926758}
{"step": 28050, "total_steps": 46238, "tokens": 15166390272, "loss": 3.0268509062853726, "ce_loss": 3.0268509062853726, "z_loss": 26.592928713018242, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17633065581321716, "tokens_per_second": 4326697.649554048, "eta_seconds": 2272.680566208158, "mean_data_wait_seconds": 7.826089859008789e-05}
{"step": 28060, "total_steps": 46238, "tokens": 15171796992, "loss": 2.9817348483837014, "ce_loss": 2.9817348483837014, "z_loss": 26.166012388287168, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16147258877754211, "tokens_per_second": 4326666.596064003, "eta_seconds": 2271.4472506248594, "mean_data_wait_seconds": 8.592009544372559e-05}
{"step": 28070, "total_steps": 46238, "tokens": 15177203712, "loss": 3.0269457011511833, "ce_loss": 3.0269457011511833, "z_loss": 27.73228983445601, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16614750027656555, "tokens_per_second": 4326837.610437663, "eta_seconds": 2270.107895962025, "mean_data_wait_seconds": 6.23166561126709e-05}
{"step": 28080, "total_steps": 46238, "tokens": 15182610432, "loss": 3.0179058331431765, "ce_loss": 3.0179058331431765, "z_loss": 26.539732947494045, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1855880618095398, "tokens_per_second": 4326885.892310834, "eta_seconds": 2268.833000991645, "mean_data_wait_seconds": 7.170438766479492e-05}
{"step": 28090, "total_steps": 46238, "tokens": 15188017152, "loss": 2.9851691181009468, "ce_loss": 2.9851691181009468, "z_loss": 26.381073677178584, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16953547298908234, "tokens_per_second": 4326914.724762569, "eta_seconds": 2267.5683271152034, "mean_data_wait_seconds": 7.218122482299805e-05}
{"step": 28100, "total_steps": 46238, "tokens": 15193423872, "loss": 3.043547975294518, "ce_loss": 3.043547975294518, "z_loss": 25.874521082097832, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18371793627738953, "tokens_per_second": 4327044.359873846, "eta_seconds": 2266.2508743695653, "mean_data_wait_seconds": 5.552172660827637e-05}
{"step": 28110, "total_steps": 46238, "tokens": 15198830592, "loss": 3.0604779268756057, "ce_loss": 3.0604779268756057, "z_loss": 26.944199056336373, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1698814034461975, "tokens_per_second": 4327128.617363419, "eta_seconds": 2264.9572524081204, "mean_data_wait_seconds": 9.763240814208984e-05}
{"step": 28120, "total_steps": 46238, "tokens": 15204237312, "loss": 2.957068394530903, "ce_loss": 2.957068394530903, "z_loss": 27.10093290155584, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14768123626708984, "tokens_per_second": 4327275.831187783, "eta_seconds": 2263.6307474098085, "mean_data_wait_seconds": 0.0002701282501220703}
{"step": 28130, "total_steps": 46238, "tokens": 15209644032, "loss": 2.9973794799862485, "ce_loss": 2.9973794799862485, "z_loss": 27.029200366049103, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14736810326576233, "tokens_per_second": 4327296.49545033, "eta_seconds": 2262.3704925911684, "mean_data_wait_seconds": 7.471442222595215e-05}
{"step": 28140, "total_steps": 46238, "tokens": 15215050752, "loss": 2.949214528907429, "ce_loss": 2.949214528907429, "z_loss": 26.293572989377108, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17493726313114166, "tokens_per_second": 4327472.841128713, "eta_seconds": 2261.0289060642485, "mean_data_wait_seconds": 9.062886238098145e-05}
{"step": 28150, "total_steps": 46238, "tokens": 15220457472, "loss": 2.991948593746532, "ce_loss": 2.991948593746532, "z_loss": 27.62813104282726, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15705019235610962, "tokens_per_second": 4327578.665754625, "eta_seconds": 2259.7242521285875, "mean_data_wait_seconds": 8.597970008850098e-05}
{"step": 28160, "total_steps": 46238, "tokens": 15225864192, "loss": 3.034692206166007, "ce_loss": 3.034692206166007, "z_loss": 26.772374297633316, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16814212501049042, "tokens_per_second": 4327602.050019873, "eta_seconds": 2258.4626846535293, "mean_data_wait_seconds": 0.00011351704597473145}
{"step": 28170, "total_steps": 46238, "tokens": 15231270912, "loss": 3.017375015851223, "ce_loss": 3.017375015851223, "z_loss": 27.134342655991063, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14944888651371002, "tokens_per_second": 4327767.823448747, "eta_seconds": 2257.126865973078, "mean_data_wait_seconds": 0.00014096498489379883}
{"step": 28180, "total_steps": 46238, "tokens": 15236677632, "loss": 3.0111918900952195, "ce_loss": 3.0111918900952195, "z_loss": 26.753722407601096, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1470198780298233, "tokens_per_second": 4327861.824210909, "eta_seconds": 2255.8285593556475, "mean_data_wait_seconds": 6.371736526489258e-05}
{"step": 28190, "total_steps": 46238, "tokens": 15242084352, "loss": 2.993257229978388, "ce_loss": 2.993257229978388, "z_loss": 26.79515384905266, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1782088726758957, "tokens_per_second": 4327934.30404547, "eta_seconds": 2254.541519930032, "mean_data_wait_seconds": 9.647011756896973e-05}
{"step": 28200, "total_steps": 46238, "tokens": 15247491072, "loss": 2.9681878378897, "ce_loss": 2.9681878378897, "z_loss": 27.530146353172533, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18220309913158417, "tokens_per_second": 4328137.663739721, "eta_seconds": 2253.186386768879, "mean_data_wait_seconds": 7.462501525878906e-05}
{"step": 28210, "total_steps": 46238, "tokens": 15252897792, "loss": 3.0444904060074776, "ce_loss": 3.0444904060074776, "z_loss": 27.8504469611428, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.23424267768859863, "tokens_per_second": 4328246.152284362, "eta_seconds": 2251.880739004617, "mean_data_wait_seconds": 8.663535118103027e-05}
{"step": 28220, "total_steps": 46238, "tokens": 15258304512, "loss": 3.0148901867144033, "ce_loss": 3.0148901867144033, "z_loss": 27.349563280741375, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17824114859104156, "tokens_per_second": 4328480.973933507, "eta_seconds": 2250.5094703344867, "mean_data_wait_seconds": 5.59687614440918e-05}
{"step": 28230, "total_steps": 46238, "tokens": 15263711232, "loss": 3.046481860406471, "ce_loss": 3.046481860406471, "z_loss": 26.596551779544715, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1550510823726654, "tokens_per_second": 4328602.07718836, "eta_seconds": 2249.197438431193, "mean_data_wait_seconds": 0.00010672211647033691}
{"step": 28240, "total_steps": 46238, "tokens": 15269117952, "loss": 2.9534485033064177, "ce_loss": 2.9534485033064177, "z_loss": 26.32394722736243, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16662012040615082, "tokens_per_second": 4328723.525983292, "eta_seconds": 2247.8853005031483, "mean_data_wait_seconds": 0.0001277923583984375}
{"step": 28250, "total_steps": 46238, "tokens": 15274524672, "loss": 2.9752448616605816, "ce_loss": 2.9752448616605816, "z_loss": 27.57637169866851, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15655890107154846, "tokens_per_second": 4328891.987012687, "eta_seconds": 2246.548838173055, "mean_data_wait_seconds": 0.0001341402530670166}
{"step": 28260, "total_steps": 46238, "tokens": 15279931392, "loss": 2.9778739069447373, "ce_loss": 2.9778739069447373, "z_loss": 27.844015887289338, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16868919134140015, "tokens_per_second": 4329084.938030983, "eta_seconds": 2245.1997785058097, "mean_data_wait_seconds": 0.00011986494064331055}
{"step": 28270, "total_steps": 46238, "tokens": 15285338112, "loss": 2.974876309886123, "ce_loss": 2.974876309886123, "z_loss": 27.446770191192627, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1667211800813675, "tokens_per_second": 4329182.020907125, "eta_seconds": 2243.900528341495, "mean_data_wait_seconds": 0.00015148520469665527}
{"step": 28280, "total_steps": 46238, "tokens": 15290744832, "loss": 2.969803040677851, "ce_loss": 2.969803040677851, "z_loss": 26.165976524353027, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15476083755493164, "tokens_per_second": 4329333.1084269015, "eta_seconds": 2242.57336195777, "mean_data_wait_seconds": 0.00013896822929382324}
{"step": 28290, "total_steps": 46238, "tokens": 15296151552, "loss": 2.932590815153989, "ce_loss": 2.932590815153989, "z_loss": 26.59631418459343, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15242663025856018, "tokens_per_second": 4329180.073421864, "eta_seconds": 2241.403734525236, "mean_data_wait_seconds": 0.0001188516616821289}
{"step": 28300, "total_steps": 46238, "tokens": 15301558272, "loss": 3.0073627251567263, "ce_loss": 3.0073627251567263, "z_loss": 26.933121233275443, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17313097417354584, "tokens_per_second": 4329281.488959994, "eta_seconds": 2240.1023561832935, "mean_data_wait_seconds": 0.00012177228927612305}
{"step": 28310, "total_steps": 46238, "tokens": 15306964992, "loss": 2.934407980153055, "ce_loss": 2.934407980153055, "z_loss": 27.235597639372855, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20108895003795624, "tokens_per_second": 4329214.753703824, "eta_seconds": 2238.887995959903, "mean_data_wait_seconds": 9.143352508544922e-05}
{"step": 28320, "total_steps": 46238, "tokens": 15312371712, "loss": 3.063856912381721, "ce_loss": 3.063856912381721, "z_loss": 27.242765802325625, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1793736219406128, "tokens_per_second": 4329258.9220987875, "eta_seconds": 2237.6162752824494, "mean_data_wait_seconds": 0.00029531121253967285}
{"step": 28330, "total_steps": 46238, "tokens": 15317778432, "loss": 2.9872836019053604, "ce_loss": 2.9872836019053604, "z_loss": 27.49210801269069, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17170758545398712, "tokens_per_second": 4329142.350961294, "eta_seconds": 2236.4276152411885, "mean_data_wait_seconds": 8.618831634521484e-05}
{"step": 28340, "total_steps": 46238, "tokens": 15323185152, "loss": 2.9691455310041253, "ce_loss": 2.9691455310041253, "z_loss": 26.82043728683934, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1448512226343155, "tokens_per_second": 4328919.189002237, "eta_seconds": 2235.2939293907893, "mean_data_wait_seconds": 0.00012427568435668945}
{"step": 28350, "total_steps": 46238, "tokens": 15328591872, "loss": 2.944118268562086, "ce_loss": 2.944118268562086, "z_loss": 26.203822569413617, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18510760366916656, "tokens_per_second": 4328820.156013318, "eta_seconds": 2234.0960620795645, "mean_data_wait_seconds": 9.110569953918457e-05}
{"step": 28360, "total_steps": 46238, "tokens": 15333998592, "loss": 2.9412420608780603, "ce_loss": 2.9412420608780603, "z_loss": 26.731102192040645, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17199887335300446, "tokens_per_second": 4328733.675625186, "eta_seconds": 2232.89166492878, "mean_data_wait_seconds": 0.00010636448860168457}
{"step": 28370, "total_steps": 46238, "tokens": 15339405312, "loss": 2.9909940325852595, "ce_loss": 2.9909940325852595, "z_loss": 27.1909969069741, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1817796379327774, "tokens_per_second": 4328655.27664427, "eta_seconds": 2231.6830531926594, "mean_data_wait_seconds": 7.49826431274414e-05}
{"step": 28380, "total_steps": 46238, "tokens": 15344812032, "loss": 3.001854116266424, "ce_loss": 3.001854116266424, "z_loss": 27.063026052532773, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13987699151039124, "tokens_per_second": 4328606.142340939, "eta_seconds": 2230.459317968493, "mean_data_wait_seconds": 0.00010585784912109375}
{"step": 28390, "total_steps": 46238, "tokens": 15350218752, "loss": 2.898365318775177, "ce_loss": 2.898365318775177, "z_loss": 26.34275482640122, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16672973334789276, "tokens_per_second": 4328681.013159996, "eta_seconds": 2229.1716933320126, "mean_data_wait_seconds": 8.130073547363281e-05}
{"step": 28400, "total_steps": 46238, "tokens": 15355625472, "loss": 2.9853690880717654, "ce_loss": 2.9853690880717654, "z_loss": 26.349715738585502, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16512462496757507, "tokens_per_second": 4328838.16646971, "eta_seconds": 2227.841765649772, "mean_data_wait_seconds": 9.91523265838623e-05}
{"step": 28410, "total_steps": 46238, "tokens": 15361032192, "loss": 3.032579024632772, "ce_loss": 3.032579024632772, "z_loss": 26.12599536144372, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14758408069610596, "tokens_per_second": 4329047.385783138, "eta_seconds": 2226.4851559845783, "mean_data_wait_seconds": 6.449222564697266e-05}
{"step": 28420, "total_steps": 46238, "tokens": 15366438912, "loss": 3.003032436876586, "ce_loss": 3.003032436876586, "z_loss": 26.6212866378553, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15583236515522003, "tokens_per_second": 4329231.727695032, "eta_seconds": 2225.1414638709766, "mean_data_wait_seconds": 5.576014518737793e-05}
{"step": 28430, "total_steps": 46238, "tokens": 15371845632, "loss": 3.0610862880042107, "ce_loss": 3.0610862880042107, "z_loss": 26.840670715678822, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.154868945479393, "tokens_per_second": 4329424.106795204, "eta_seconds": 2223.7937579016266, "mean_data_wait_seconds": 6.511807441711426e-05}
{"step": 28440, "total_steps": 46238, "tokens": 15377252352, "loss": 3.0274901245579575, "ce_loss": 3.0274901245579575, "z_loss": 27.35001497557669, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20502141118049622, "tokens_per_second": 4329541.38565794, "eta_seconds": 2222.484722256036, "mean_data_wait_seconds": 6.631016731262207e-05}
{"step": 28450, "total_steps": 46238, "tokens": 15382659072, "loss": 2.9659336534413425, "ce_loss": 2.9659336534413425, "z_loss": 26.526988997603908, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12857666611671448, "tokens_per_second": 4329742.222462811, "eta_seconds": 2221.1328919553484, "mean_data_wait_seconds": 6.282329559326172e-05}
{"step": 28460, "total_steps": 46238, "tokens": 15388065792, "loss": 3.035746131882523, "ce_loss": 3.035746131882523, "z_loss": 26.638135837786127, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14752572774887085, "tokens_per_second": 4329797.607569412, "eta_seconds": 2219.8557565824776, "mean_data_wait_seconds": 6.639957427978516e-05}
{"step": 28470, "total_steps": 46238, "tokens": 15393472512, "loss": 3.029228475960818, "ce_loss": 3.029228475960818, "z_loss": 26.509433023857348, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14974942803382874, "tokens_per_second": 4329937.738696354, "eta_seconds": 2218.535231615636, "mean_data_wait_seconds": 7.259845733642578e-05}
{"step": 28480, "total_steps": 46238, "tokens": 15398879232, "loss": 3.0137812639727737, "ce_loss": 3.0137812639727737, "z_loss": 26.41849303968025, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16699126362800598, "tokens_per_second": 4329997.288166701, "eta_seconds": 2217.2560546948735, "mean_data_wait_seconds": 8.958578109741211e-05}
{"step": 28490, "total_steps": 46238, "tokens": 15404285952, "loss": 3.063775203444741, "ce_loss": 3.063775203444741, "z_loss": 27.078733921051025, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16804465651512146, "tokens_per_second": 4330056.546164591, "eta_seconds": 2215.9770621238604, "mean_data_wait_seconds": 6.240606307983398e-05}
{"step": 28500, "total_steps": 46238, "tokens": 15409692672, "loss": 3.079899074453296, "ce_loss": 3.079899074453296, "z_loss": 27.151622280930027, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1447151154279709, "tokens_per_second": 4330250.290135805, "eta_seconds": 2214.62932196911, "mean_data_wait_seconds": 0.00010842084884643555}
{"step": 28510, "total_steps": 46238, "tokens": 15415099392, "loss": 3.0709664297826365, "ce_loss": 3.0709664297826365, "z_loss": 26.936335650357332, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14932313561439514, "tokens_per_second": 4330418.825423405, "eta_seconds": 2213.2945865953, "mean_data_wait_seconds": 0.0001214444637298584}
{"step": 28520, "total_steps": 46238, "tokens": 15420506112, "loss": 3.0398394898934797, "ce_loss": 3.0398394898934797, "z_loss": 26.692260005257346, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16856873035430908, "tokens_per_second": 4330404.057208033, "eta_seconds": 2212.053586097917, "mean_data_wait_seconds": 5.8203935623168945e-05}
{"step": 28530, "total_steps": 46238, "tokens": 15425912832, "loss": 3.020913277611588, "ce_loss": 3.020913277611588, "z_loss": 26.866685925108015, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1593952178955078, "tokens_per_second": 4330433.898052031, "eta_seconds": 2210.7898029124863, "mean_data_wait_seconds": 0.00012385845184326172}
{"step": 28540, "total_steps": 46238, "tokens": 15431319552, "loss": 3.0560620354883596, "ce_loss": 3.0560620354883596, "z_loss": 26.574264844258625, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18210287392139435, "tokens_per_second": 4330133.87767735, "eta_seconds": 2209.694354561699, "mean_data_wait_seconds": 0.00015121698379516602}
{"step": 28550, "total_steps": 46238, "tokens": 15436726272, "loss": 3.0245312109138025, "ce_loss": 3.0245312109138025, "z_loss": 27.233578234007865, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15533049404621124, "tokens_per_second": 4330256.274058249, "eta_seconds": 2208.3833054614643, "mean_data_wait_seconds": 9.772181510925293e-05}
{"step": 28560, "total_steps": 46238, "tokens": 15442132992, "loss": 3.0429705905191824, "ce_loss": 3.0429705905191824, "z_loss": 26.980484066587508, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13664942979812622, "tokens_per_second": 4330345.866415427, "eta_seconds": 2207.089049889558, "mean_data_wait_seconds": 0.00011411309242248535}
{"step": 28570, "total_steps": 46238, "tokens": 15447539712, "loss": 3.027596901763569, "ce_loss": 3.027596901763569, "z_loss": 26.95270210323912, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14426860213279724, "tokens_per_second": 4330395.87523835, "eta_seconds": 2205.81501072907, "mean_data_wait_seconds": 0.00011789798736572266}
{"step": 28580, "total_steps": 46238, "tokens": 15452946432, "loss": 3.0046568856094824, "ce_loss": 3.0046568856094824, "z_loss": 26.778294563293457, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15550477802753448, "tokens_per_second": 4330510.81136027, "eta_seconds": 2204.5079483363015, "mean_data_wait_seconds": 0.0021984875202178955}
{"step": 28590, "total_steps": 46238, "tokens": 15458353152, "loss": 3.0337001211715466, "ce_loss": 3.0337001211715466, "z_loss": 27.319687684377033, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15129299461841583, "tokens_per_second": 4330613.120492472, "eta_seconds": 2203.207379308679, "mean_data_wait_seconds": 8.133053779602051e-05}
{"step": 28600, "total_steps": 46238, "tokens": 15463759872, "loss": 3.0554792122407393, "ce_loss": 3.0554792122407393, "z_loss": 27.601222255013205, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17968837916851044, "tokens_per_second": 4330528.428744171, "eta_seconds": 2202.0019544740267, "mean_data_wait_seconds": 6.988644599914551e-05}
{"step": 28610, "total_steps": 46238, "tokens": 15469166592, "loss": 3.0624511765711233, "ce_loss": 3.0624511765711233, "z_loss": 26.509265451720267, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16967913508415222, "tokens_per_second": 4330561.898756008, "eta_seconds": 2200.7364325487874, "mean_data_wait_seconds": 0.00010529160499572754}
{"step": 28620, "total_steps": 46238, "tokens": 15474573312, "loss": 2.955602896936012, "ce_loss": 2.955602896936012, "z_loss": 26.684499596104477, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16660913825035095, "tokens_per_second": 4330684.38411196, "eta_seconds": 2199.425721011802, "mean_data_wait_seconds": 7.477402687072754e-05}
{"step": 28630, "total_steps": 46238, "tokens": 15479980032, "loss": 2.9749364256858826, "ce_loss": 2.9749364256858826, "z_loss": 25.995862151637223, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16803023219108582, "tokens_per_second": 4330841.657820535, "eta_seconds": 2198.0974268153404, "mean_data_wait_seconds": 9.247660636901855e-05}
{"step": 28640, "total_steps": 46238, "tokens": 15485386752, "loss": 3.030778796383829, "ce_loss": 3.030778796383829, "z_loss": 27.42861996275006, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15518532693386078, "tokens_per_second": 4330879.855365906, "eta_seconds": 2196.8296285596607, "mean_data_wait_seconds": 6.589293479919434e-05}
{"step": 28650, "total_steps": 46238, "tokens": 15490793472, "loss": 3.0356187802372556, "ce_loss": 3.0356187802372556, "z_loss": 26.936258402737703, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1598770022392273, "tokens_per_second": 4331017.03563585, "eta_seconds": 2195.511674454539, "mean_data_wait_seconds": 0.00022050738334655762}
{"step": 28660, "total_steps": 46238, "tokens": 15496200192, "loss": 3.0342570434917104, "ce_loss": 3.0342570434917104, "z_loss": 26.899877317024, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16413207352161407, "tokens_per_second": 4331044.10790703, "eta_seconds": 2194.249586756691, "mean_data_wait_seconds": 8.100271224975586e-05}
{"step": 28670, "total_steps": 46238, "tokens": 15501606912, "loss": 2.968791475801757, "ce_loss": 2.968791475801757, "z_loss": 26.73659883845936, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19374768435955048, "tokens_per_second": 4331192.082832682, "eta_seconds": 2192.9262988927835, "mean_data_wait_seconds": 6.669759750366211e-05}
{"step": 28680, "total_steps": 46238, "tokens": 15507013632, "loss": 3.0361269802758186, "ce_loss": 3.0361269802758186, "z_loss": 27.255613124731816, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15957403182983398, "tokens_per_second": 4331316.751958705, "eta_seconds": 2191.6148939481914, "mean_data_wait_seconds": 6.276369094848633e-05}
{"step": 28690, "total_steps": 46238, "tokens": 15512420352, "loss": 3.070816366961508, "ce_loss": 3.070816366961508, "z_loss": 27.48052592711015, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16081787645816803, "tokens_per_second": 4331597.376555131, "eta_seconds": 2190.224704481892, "mean_data_wait_seconds": 0.0001291036605834961}
{"step": 28700, "total_steps": 46238, "tokens": 15517827072, "loss": 2.983096632090482, "ce_loss": 2.983096632090482, "z_loss": 26.73622463688706, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15998291969299316, "tokens_per_second": 4331614.174274788, "eta_seconds": 2188.968011119657, "mean_data_wait_seconds": 9.191036224365234e-05}
{"step": 28710, "total_steps": 46238, "tokens": 15523233792, "loss": 3.056073981704134, "ce_loss": 3.056073981704134, "z_loss": 27.32554919791944, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17750270664691925, "tokens_per_second": 4331745.804411072, "eta_seconds": 2187.653332370081, "mean_data_wait_seconds": 7.978081703186035e-05}
{"step": 28720, "total_steps": 46238, "tokens": 15528640512, "loss": 2.996165383945812, "ce_loss": 2.996165383945812, "z_loss": 26.37797500147964, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16109295189380646, "tokens_per_second": 4331868.1483626915, "eta_seconds": 2186.343420350806, "mean_data_wait_seconds": 5.939602851867676e-05}
{"step": 28730, "total_steps": 46238, "tokens": 15534047232, "loss": 2.985491743593505, "ce_loss": 2.985491743593505, "z_loss": 26.94899878357396, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.169672429561615, "tokens_per_second": 4332050.914873841, "eta_seconds": 2185.0031059193257, "mean_data_wait_seconds": 6.538629531860352e-05}
{"step": 28740, "total_steps": 46238, "tokens": 15539453952, "loss": 2.9672362370924517, "ce_loss": 2.9672362370924517, "z_loss": 27.90619168137059, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.192288339138031, "tokens_per_second": 4332168.952671467, "eta_seconds": 2183.6955315803943, "mean_data_wait_seconds": 7.295608520507812e-05}
{"step": 28750, "total_steps": 46238, "tokens": 15544860672, "loss": 2.9205452023130474, "ce_loss": 2.9205452023130474, "z_loss": 27.159018357594807, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14898622035980225, "tokens_per_second": 4332091.078959634, "eta_seconds": 2182.4867233102095, "mean_data_wait_seconds": 8.26716423034668e-05}
{"step": 28760, "total_steps": 46238, "tokens": 15550267392, "loss": 2.9717516285000425, "ce_loss": 2.9717516285000425, "z_loss": 27.617371544693455, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1589374989271164, "tokens_per_second": 4332174.287263342, "eta_seconds": 2181.1967657398172, "mean_data_wait_seconds": 0.00014126300811767578}
{"step": 28770, "total_steps": 46238, "tokens": 15555674112, "loss": 2.946976253480622, "ce_loss": 2.946976253480622, "z_loss": 28.04171371459961, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13233840465545654, "tokens_per_second": 4332152.750383063, "eta_seconds": 2179.9595647140877, "mean_data_wait_seconds": 0.0001087486743927002}
{"step": 28780, "total_steps": 46238, "tokens": 15561080832, "loss": 3.0014273235292146, "ce_loss": 3.0014273235292146, "z_loss": 28.2026328607039, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.21102434396743774, "tokens_per_second": 4332186.613430551, "eta_seconds": 2178.694489923156, "mean_data_wait_seconds": 8.499622344970703e-05}
{"step": 28790, "total_steps": 46238, "tokens": 15566487552, "loss": 2.905941793412873, "ce_loss": 2.905941793412873, "z_loss": 27.633722261949018, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15314224362373352, "tokens_per_second": 4332113.664112575, "eta_seconds": 2177.483121494309, "mean_data_wait_seconds": 8.231401443481445e-05}
{"step": 28800, "total_steps": 46238, "tokens": 15571894272, "loss": 2.9143229578480576, "ce_loss": 2.9143229578480576, "z_loss": 27.105604952031914, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17348115146160126, "tokens_per_second": 4332203.970193792, "eta_seconds": 2176.189701330769, "mean_data_wait_seconds": 0.0003535449504852295}
{"step": 28810, "total_steps": 46238, "tokens": 15577300992, "loss": 2.910711185498671, "ce_loss": 2.910711185498671, "z_loss": 26.98765977223714, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.21039234101772308, "tokens_per_second": 4332027.831563779, "eta_seconds": 2175.030103765223, "mean_data_wait_seconds": 6.967782974243164e-05}
{"step": 28820, "total_steps": 46238, "tokens": 15582707712, "loss": 2.920878737261801, "ce_loss": 2.920878737261801, "z_loss": 28.408843473954633, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19044756889343262, "tokens_per_second": 4331946.992153957, "eta_seconds": 2173.8225885625807, "mean_data_wait_seconds": 0.00013774633407592773}
{"step": 28830, "total_steps": 46238, "tokens": 15588114432, "loss": 2.9554163018862405, "ce_loss": 2.9554163018862405, "z_loss": 27.088677579706367, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19444215297698975, "tokens_per_second": 4331800.8615735, "eta_seconds": 2172.647775082934, "mean_data_wait_seconds": 0.00010532140731811523}
{"step": 28840, "total_steps": 46238, "tokens": 15593521152, "loss": 2.9497404911301355, "ce_loss": 2.9497404911301355, "z_loss": 27.40675790382154, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16384384036064148, "tokens_per_second": 4331898.059348921, "eta_seconds": 2171.350907877486, "mean_data_wait_seconds": 0.00013524293899536133}
{"step": 28850, "total_steps": 46238, "tokens": 15598927872, "loss": 3.0067516204082603, "ce_loss": 3.0067516204082603, "z_loss": 27.64475172216242, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17170563340187073, "tokens_per_second": 4332072.299096449, "eta_seconds": 2170.0155064265014, "mean_data_wait_seconds": 0.00015786290168762207}
{"step": 28860, "total_steps": 46238, "tokens": 15604334592, "loss": 2.9595971775777414, "ce_loss": 2.9595971775777414, "z_loss": 27.78597018212983, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14918310940265656, "tokens_per_second": 4332178.5047693495, "eta_seconds": 2168.7142701199045, "mean_data_wait_seconds": 0.00013700127601623535}
{"step": 28870, "total_steps": 46238, "tokens": 15609741312, "loss": 2.8660083069945825, "ce_loss": 2.8660083069945825, "z_loss": 27.5813870863481, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1499185860157013, "tokens_per_second": 4332284.455738087, "eta_seconds": 2167.41322503955, "mean_data_wait_seconds": 6.985664367675781e-05}
{"step": 28880, "total_steps": 46238, "tokens": 15615148032, "loss": 3.0100515520933904, "ce_loss": 3.0100515520933904, "z_loss": 27.566978873628557, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16218425333499908, "tokens_per_second": 4332471.172238313, "eta_seconds": 2166.071863128325, "mean_data_wait_seconds": 0.00012254714965820312}
{"step": 28890, "total_steps": 46238, "tokens": 15620554752, "loss": 2.9432319113702485, "ce_loss": 2.9432319113702485, "z_loss": 27.161867820855342, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2148132622241974, "tokens_per_second": 4332601.859938087, "eta_seconds": 2164.758610922543, "mean_data_wait_seconds": 0.00010496377944946289}
{"step": 28900, "total_steps": 46238, "tokens": 15625961472, "loss": 2.931533230073524, "ce_loss": 2.931533230073524, "z_loss": 26.97822445089167, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14350014925003052, "tokens_per_second": 4332783.189966749, "eta_seconds": 2163.420151210459, "mean_data_wait_seconds": 7.259845733642578e-05}
{"step": 28910, "total_steps": 46238, "tokens": 15631368192, "loss": 2.9699401259422302, "ce_loss": 2.9699401259422302, "z_loss": 26.869978572383072, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17522406578063965, "tokens_per_second": 4332701.927685253, "eta_seconds": 2162.212840938489, "mean_data_wait_seconds": 7.548928260803223e-05}
{"step": 28920, "total_steps": 46238, "tokens": 15636774912, "loss": 3.029899234121496, "ce_loss": 3.029899234121496, "z_loss": 27.374511588703502, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1741616576910019, "tokens_per_second": 4332888.211655466, "eta_seconds": 2160.8720480750067, "mean_data_wait_seconds": 9.861588478088379e-05}
{"step": 28930, "total_steps": 46238, "tokens": 15642181632, "loss": 2.97070269873648, "ce_loss": 2.97070269873648, "z_loss": 26.728174671982273, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.45695778727531433, "tokens_per_second": 4333054.20830866, "eta_seconds": 2159.5414814006026, "mean_data_wait_seconds": 6.002187728881836e-05}
{"step": 28940, "total_steps": 46238, "tokens": 15647588352, "loss": 2.91444263494376, "ce_loss": 2.91444263494376, "z_loss": 27.458539948318943, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13817772269248962, "tokens_per_second": 4333112.485731552, "eta_seconds": 2158.264668825258, "mean_data_wait_seconds": 6.154179573059082e-05}
{"step": 28950, "total_steps": 46238, "tokens": 15652995072, "loss": 2.944031030842752, "ce_loss": 2.944031030842752, "z_loss": 27.756217566403475, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17531666159629822, "tokens_per_second": 4333268.9933368815, "eta_seconds": 2156.9389941801305, "mean_data_wait_seconds": 6.118416786193848e-05}
{"step": 28960, "total_steps": 46238, "tokens": 15658401792, "loss": 3.032496936393507, "ce_loss": 3.032496936393507, "z_loss": 26.913950602213543, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16567684710025787, "tokens_per_second": 4333404.7114240555, "eta_seconds": 2155.623756851982, "mean_data_wait_seconds": 6.213784217834473e-05}
{"step": 28970, "total_steps": 46238, "tokens": 15663808512, "loss": 3.007562684290337, "ce_loss": 3.007562684290337, "z_loss": 26.55678094517101, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.22366663813591003, "tokens_per_second": 4333586.412779449, "eta_seconds": 2154.2857427440276, "mean_data_wait_seconds": 0.00011461973190307617}
{"step": 28980, "total_steps": 46238, "tokens": 15669215232, "loss": 2.944309171402093, "ce_loss": 2.944309171402093, "z_loss": 27.886462009314336, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14083144068717957, "tokens_per_second": 4333622.991457405, "eta_seconds": 2153.0199379116225, "mean_data_wait_seconds": 0.00010719895362854004}
{"step": 28990, "total_steps": 46238, "tokens": 15674621952, "loss": 2.9782824913660684, "ce_loss": 2.9782824913660684, "z_loss": 27.822738083926115, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1661810427904129, "tokens_per_second": 4333693.082791579, "eta_seconds": 2151.7375148295587, "mean_data_wait_seconds": 0.00010263919830322266}
{"step": 29000, "total_steps": 46238, "tokens": 15680028672, "loss": 2.990330134377335, "ce_loss": 2.990330134377335, "z_loss": 27.7233986565561, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16491730511188507, "tokens_per_second": 4333746.552643113, "eta_seconds": 2150.463381001384, "mean_data_wait_seconds": 0.00010949373245239258}
{"step": 29010, "total_steps": 46238, "tokens": 15685435392, "loss": 2.9870392336989893, "ce_loss": 2.9870392336989893, "z_loss": 27.535159443363998, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15866076946258545, "tokens_per_second": 4333927.335529557, "eta_seconds": 2149.126144234699, "mean_data_wait_seconds": 8.308887481689453e-05}
{"step": 29020, "total_steps": 46238, "tokens": 15690842112, "loss": 2.9625659747557207, "ce_loss": 2.9625659747557207, "z_loss": 26.908070390874688, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15054510533809662, "tokens_per_second": 4334030.375882699, "eta_seconds": 2147.827545418187, "mean_data_wait_seconds": 0.00017130374908447266}
{"step": 29030, "total_steps": 46238, "tokens": 15696248832, "loss": 3.037124615727049, "ce_loss": 3.037124615727049, "z_loss": 26.93921034263842, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1384785920381546, "tokens_per_second": 4334263.309411753, "eta_seconds": 2146.4646791988857, "mean_data_wait_seconds": 0.0001430213451385498}
{"step": 29040, "total_steps": 46238, "tokens": 15701655552, "loss": 2.999634092504328, "ce_loss": 2.999634092504328, "z_loss": 27.63770565842137, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1769166886806488, "tokens_per_second": 4334422.115346549, "eta_seconds": 2145.1386451447643, "mean_data_wait_seconds": 0.00010919570922851562}
{"step": 29050, "total_steps": 46238, "tokens": 15707062272, "loss": 3.027235493515477, "ce_loss": 3.027235493515477, "z_loss": 27.801991563854795, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15461547672748566, "tokens_per_second": 4334482.08334594, "eta_seconds": 2143.8615929926204, "mean_data_wait_seconds": 9.927153587341309e-05}
{"step": 29060, "total_steps": 46238, "tokens": 15712468992, "loss": 3.0665996995839206, "ce_loss": 3.0665996995839206, "z_loss": 26.75367512847438, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14130628108978271, "tokens_per_second": 4334518.893046586, "eta_seconds": 2142.5960234937165, "mean_data_wait_seconds": 7.686018943786621e-05}
{"step": 29070, "total_steps": 46238, "tokens": 15717875712, "loss": 3.0246684352556863, "ce_loss": 3.0246684352556863, "z_loss": 28.407530336668998, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16978251934051514, "tokens_per_second": 4334565.044104375, "eta_seconds": 2141.325860739927, "mean_data_wait_seconds": 9.21487808227539e-05}
{"step": 29080, "total_steps": 46238, "tokens": 15723282432, "loss": 3.064334652640603, "ce_loss": 3.064334652640603, "z_loss": 27.180397741722338, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16021381318569183, "tokens_per_second": 4334655.183234489, "eta_seconds": 2140.034007751935, "mean_data_wait_seconds": 8.767843246459961e-05}
{"step": 29090, "total_steps": 46238, "tokens": 15728689152, "loss": 2.9853946927822, "ce_loss": 2.9853946927822, "z_loss": 27.535307942014754, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14450424909591675, "tokens_per_second": 4334717.198453524, "eta_seconds": 2138.756084781617, "mean_data_wait_seconds": 0.00011068582534790039}
{"step": 29100, "total_steps": 46238, "tokens": 15734095872, "loss": 3.0750508272286616, "ce_loss": 3.0750508272286616, "z_loss": 27.814772721492883, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1974109411239624, "tokens_per_second": 4334792.224003754, "eta_seconds": 2137.471783005574, "mean_data_wait_seconds": 5.879998207092285e-05}
{"step": 29110, "total_steps": 46238, "tokens": 15739502592, "loss": 3.0460512005921565, "ce_loss": 3.0460512005921565, "z_loss": 27.94724276571563, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14660346508026123, "tokens_per_second": 4334997.002321261, "eta_seconds": 2136.12358648495, "mean_data_wait_seconds": 6.225705146789551e-05}
{"step": 29120, "total_steps": 46238, "tokens": 15744909312, "loss": 3.0427588964953567, "ce_loss": 3.0427588964953567, "z_loss": 27.93745266307484, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13998141884803772, "tokens_per_second": 4335200.0696127005, "eta_seconds": 2134.776359889383, "mean_data_wait_seconds": 6.604194641113281e-05}
{"step": 29130, "total_steps": 46238, "tokens": 15750316032, "loss": 3.0046150016062185, "ce_loss": 3.0046150016062185, "z_loss": 28.354346997810133, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1606622189283371, "tokens_per_second": 4335245.4683753345, "eta_seconds": 2133.5068501821734, "mean_data_wait_seconds": 6.765127182006836e-05}
{"step": 29140, "total_steps": 46238, "tokens": 15755722752, "loss": 2.9979260166486106, "ce_loss": 2.9979260166486106, "z_loss": 27.751606970122367, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1608615219593048, "tokens_per_second": 4335418.590599314, "eta_seconds": 2132.174550352278, "mean_data_wait_seconds": 7.092952728271484e-05}
{"step": 29150, "total_steps": 46238, "tokens": 15761129472, "loss": 3.082006194374778, "ce_loss": 3.082006194374778, "z_loss": 28.29614292491566, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16416950523853302, "tokens_per_second": 4335381.040243043, "eta_seconds": 2130.945902619459, "mean_data_wait_seconds": 0.00012034177780151367}
{"step": 29160, "total_steps": 46238, "tokens": 15766536192, "loss": 3.0226517084873086, "ce_loss": 3.0226517084873086, "z_loss": 26.762194474538166, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16893477737903595, "tokens_per_second": 4335416.106089923, "eta_seconds": 2129.6815618298792, "mean_data_wait_seconds": 9.503960609436035e-05}
{"step": 29170, "total_steps": 46238, "tokens": 15771942912, "loss": 3.035108972679485, "ce_loss": 3.035108972679485, "z_loss": 27.04028508157441, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1804407685995102, "tokens_per_second": 4335572.404128986, "eta_seconds": 2128.3577262398017, "mean_data_wait_seconds": 5.882978439331055e-05}
{"step": 29180, "total_steps": 46238, "tokens": 15777349632, "loss": 3.024401231245561, "ce_loss": 3.024401231245561, "z_loss": 27.276890956994258, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15671125054359436, "tokens_per_second": 4335729.137804235, "eta_seconds": 2127.0337723796247, "mean_data_wait_seconds": 0.00015410780906677246}
{"step": 29190, "total_steps": 46238, "tokens": 15782756352, "loss": 3.0399361505652918, "ce_loss": 3.0399361505652918, "z_loss": 27.676741686734285, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1698482781648636, "tokens_per_second": 4335687.396501322, "eta_seconds": 2125.807222964809, "mean_data_wait_seconds": 8.541345596313477e-05}
{"step": 29200, "total_steps": 46238, "tokens": 15788163072, "loss": 2.9953219150051926, "ce_loss": 2.9953219150051926, "z_loss": 26.57326483004021, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18138684332370758, "tokens_per_second": 4335774.5300308205, "eta_seconds": 2124.5174997452004, "mean_data_wait_seconds": 7.414817810058594e-05}
{"step": 29210, "total_steps": 46238, "tokens": 15793569792, "loss": 3.035292002287778, "ce_loss": 3.035292002287778, "z_loss": 28.04214026711204, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16642199456691742, "tokens_per_second": 4335661.021417586, "eta_seconds": 2123.3260853474203, "mean_data_wait_seconds": 8.165836334228516e-05}
{"step": 29220, "total_steps": 46238, "tokens": 15798976512, "loss": 3.0570789702010877, "ce_loss": 3.0570789702010877, "z_loss": 27.490018526713055, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14424292743206024, "tokens_per_second": 4335770.621888164, "eta_seconds": 2122.0254082521706, "mean_data_wait_seconds": 6.285309791564941e-05}
{"step": 29230, "total_steps": 46238, "tokens": 15804383232, "loss": 3.0270748391295923, "ce_loss": 3.0270748391295923, "z_loss": 27.176523078571666, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1529327780008316, "tokens_per_second": 4335672.586276065, "eta_seconds": 2120.8263587767406, "mean_data_wait_seconds": 8.118152618408203e-05}
{"step": 29240, "total_steps": 46238, "tokens": 15809789952, "loss": 2.985278429407062, "ce_loss": 2.985278429407062, "z_loss": 29.11735468199759, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20114220678806305, "tokens_per_second": 4335687.580288351, "eta_seconds": 2119.571997249124, "mean_data_wait_seconds": 0.00011101365089416504}
{"step": 29250, "total_steps": 46238, "tokens": 15815196672, "loss": 3.00504192980853, "ce_loss": 3.00504192980853, "z_loss": 28.245233636913877, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15971630811691284, "tokens_per_second": 4335733.47229867, "eta_seconds": 2118.3025485029925, "mean_data_wait_seconds": 0.0001423954963684082}
{"step": 29260, "total_steps": 46238, "tokens": 15820603392, "loss": 2.9872371590498723, "ce_loss": 2.9872371590498723, "z_loss": 27.417773203416303, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17476347088813782, "tokens_per_second": 4335724.586837484, "eta_seconds": 2117.0598731907085, "mean_data_wait_seconds": 9.652972221374512e-05}
{"step": 29270, "total_steps": 46238, "tokens": 15826010112, "loss": 3.0039344195163613, "ce_loss": 3.0039344195163613, "z_loss": 28.394459464333274, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19614127278327942, "tokens_per_second": 4335808.337029985, "eta_seconds": 2115.771987809746, "mean_data_wait_seconds": 0.00010412931442260742}
{"step": 29280, "total_steps": 46238, "tokens": 15831416832, "loss": 2.9389294638778223, "ce_loss": 2.9389294638778223, "z_loss": 28.341070652008057, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16693641245365143, "tokens_per_second": 4335935.019031665, "eta_seconds": 2114.4632158365484, "mean_data_wait_seconds": 6.16908073425293e-05}
{"step": 29290, "total_steps": 46238, "tokens": 15836823552, "loss": 2.960483541994384, "ce_loss": 2.960483541994384, "z_loss": 28.337742747682512, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16789966821670532, "tokens_per_second": 4336031.064084744, "eta_seconds": 2113.1694511819856, "mean_data_wait_seconds": 6.61313533782959e-05}
{"step": 29300, "total_steps": 46238, "tokens": 15842230272, "loss": 3.0145781202749773, "ce_loss": 3.0145781202749773, "z_loss": 28.62392422647187, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16235221922397614, "tokens_per_second": 4336176.970785778, "eta_seconds": 2111.8514594067765, "mean_data_wait_seconds": 5.8978796005249023e-05}
{"step": 29310, "total_steps": 46238, "tokens": 15847636992, "loss": 2.9344468387690457, "ce_loss": 2.9344468387690457, "z_loss": 28.227320179794773, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14245615899562836, "tokens_per_second": 4336219.6142701, "eta_seconds": 2110.5838168071, "mean_data_wait_seconds": 5.951523780822754e-05}
{"step": 29320, "total_steps": 46238, "tokens": 15853043712, "loss": 2.9583890149087617, "ce_loss": 2.9583890149087617, "z_loss": 27.75632787473274, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15351703763008118, "tokens_per_second": 4336270.176665644, "eta_seconds": 2109.31234709946, "mean_data_wait_seconds": 0.0001087188720703125}
{"step": 29330, "total_steps": 46238, "tokens": 15858450432, "loss": 2.9403053467923943, "ce_loss": 2.9403053467923943, "z_loss": 28.988055460380785, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17485731840133667, "tokens_per_second": 4336383.323903905, "eta_seconds": 2108.010482747297, "mean_data_wait_seconds": 7.218122482299805e-05}
{"step": 29340, "total_steps": 46238, "tokens": 15863857152, "loss": 2.936578504966967, "ce_loss": 2.936578504966967, "z_loss": 28.651392170877166, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15535223484039307, "tokens_per_second": 4336549.523733817, "eta_seconds": 2106.6829132240673, "mean_data_wait_seconds": 7.158517837524414e-05}
{"step": 29350, "total_steps": 46238, "tokens": 15869263872, "loss": 2.882027694673249, "ce_loss": 2.882027694673249, "z_loss": 28.995636145273846, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1691722571849823, "tokens_per_second": 4336588.31730323, "eta_seconds": 2105.4172994862065, "mean_data_wait_seconds": 6.064772605895996e-05}
{"step": 29360, "total_steps": 46238, "tokens": 15874670592, "loss": 2.897081653277079, "ce_loss": 2.897081653277079, "z_loss": 28.872539765907057, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17287978529930115, "tokens_per_second": 4336705.973641105, "eta_seconds": 2104.113444504217, "mean_data_wait_seconds": 7.647275924682617e-05}
{"step": 29370, "total_steps": 46238, "tokens": 15880077312, "loss": 2.966088343750347, "ce_loss": 2.966088343750347, "z_loss": 27.573995806954123, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16619379818439484, "tokens_per_second": 4336999.985620326, "eta_seconds": 2102.724153616898, "mean_data_wait_seconds": 9.086728096008301e-05}
{"step": 29380, "total_steps": 46238, "tokens": 15885484032, "loss": 2.9087715022491687, "ce_loss": 2.9087715022491687, "z_loss": 27.31863784790039, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1583484262228012, "tokens_per_second": 4337141.020464161, "eta_seconds": 2101.4091681585687, "mean_data_wait_seconds": 7.721781730651855e-05}
{"step": 29390, "total_steps": 46238, "tokens": 15890890752, "loss": 2.9286412253524317, "ce_loss": 2.9286412253524317, "z_loss": 28.489454298308402, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19353294372558594, "tokens_per_second": 4337282.131920309, "eta_seconds": 2100.094231123298, "mean_data_wait_seconds": 9.41157341003418e-05}
{"step": 29400, "total_steps": 46238, "tokens": 15896297472, "loss": 2.9525972695061653, "ce_loss": 2.9525972695061653, "z_loss": 29.02156568296028, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14925013482570648, "tokens_per_second": 4337289.648382365, "eta_seconds": 2098.8440251840602, "mean_data_wait_seconds": 8.377432823181152e-05}
{"step": 29410, "total_steps": 46238, "tokens": 15901704192, "loss": 2.983147725914464, "ce_loss": 2.983147725914464, "z_loss": 28.160891272804953, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15955105423927307, "tokens_per_second": 4337417.700712938, "eta_seconds": 2097.5355319144355, "mean_data_wait_seconds": 7.489323616027832e-05}
{"step": 29420, "total_steps": 46238, "tokens": 15907110912, "loss": 2.9811149365974194, "ce_loss": 2.9811149365974194, "z_loss": 28.422226053295713, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14436395466327667, "tokens_per_second": 4337028.255845521, "eta_seconds": 2096.4772391660117, "mean_data_wait_seconds": 6.669759750366211e-05}
{"step": 29430, "total_steps": 46238, "tokens": 15912517632, "loss": 2.9084630572434627, "ce_loss": 2.9084630572434627, "z_loss": 28.266563805666838, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15933524072170258, "tokens_per_second": 4337234.541384852, "eta_seconds": 2095.1309451433435, "mean_data_wait_seconds": 8.448958396911621e-05}
{"step": 29440, "total_steps": 46238, "tokens": 15917924352, "loss": 2.982482025117585, "ce_loss": 2.982482025117585, "z_loss": 28.68597516146573, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2095661759376526, "tokens_per_second": 4336553.699881867, "eta_seconds": 2094.2131038864795, "mean_data_wait_seconds": 7.43567943572998e-05}
{"step": 29450, "total_steps": 46238, "tokens": 15923331072, "loss": 3.018776799693252, "ce_loss": 3.018776799693252, "z_loss": 26.83432811679262, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19368702173233032, "tokens_per_second": 4336054.175497441, "eta_seconds": 2093.2074408315607, "mean_data_wait_seconds": 6.681680679321289e-05}
{"step": 29460, "total_steps": 46238, "tokens": 15928737792, "loss": 2.904625227957061, "ce_loss": 2.904625227957061, "z_loss": 28.148865483023904, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14597439765930176, "tokens_per_second": 4336123.951257356, "eta_seconds": 2091.926855866217, "mean_data_wait_seconds": 8.45193862915039e-05}
{"step": 29470, "total_steps": 46238, "tokens": 15934144512, "loss": 3.001391053199768, "ce_loss": 3.001391053199768, "z_loss": 28.33879784381751, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1657777577638626, "tokens_per_second": 4336227.391831132, "eta_seconds": 2090.6300811341403, "mean_data_wait_seconds": 7.405877113342285e-05}
{"step": 29480, "total_steps": 46238, "tokens": 15939551232, "loss": 2.9989197019374734, "ce_loss": 2.9989197019374734, "z_loss": 27.600983518542666, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17554062604904175, "tokens_per_second": 4336242.300712317, "eta_seconds": 2089.376025530608, "mean_data_wait_seconds": 0.00012561678886413574}
{"step": 29490, "total_steps": 46238, "tokens": 15944957952, "loss": 3.038117710388068, "ce_loss": 3.038117710388068, "z_loss": 28.543313416567717, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1458759903907776, "tokens_per_second": 4335723.471360979, "eta_seconds": 2088.379031506306, "mean_data_wait_seconds": 8.702278137207031e-05}
{"step": 29500, "total_steps": 46238, "tokens": 15950364672, "loss": 3.0134759364706096, "ce_loss": 3.0134759364706096, "z_loss": 27.438786029815674, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1643628478050232, "tokens_per_second": 4335777.2819853015, "eta_seconds": 2087.1061116535175, "mean_data_wait_seconds": 0.0001392662525177002}
{"step": 29510, "total_steps": 46238, "tokens": 15955771392, "loss": 3.0074061968109826, "ce_loss": 3.0074061968109826, "z_loss": 28.324195356080025, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17477302253246307, "tokens_per_second": 4335205.773476474, "eta_seconds": 2086.134088335929, "mean_data_wait_seconds": 0.00013381242752075195}
{"step": 29520, "total_steps": 46238, "tokens": 15961178112, "loss": 2.9809935580600393, "ce_loss": 2.9809935580600393, "z_loss": 28.23580854589289, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13644558191299438, "tokens_per_second": 4335224.769446113, "eta_seconds": 2084.8777871221628, "mean_data_wait_seconds": 7.703900337219238e-05}
{"step": 29530, "total_steps": 46238, "tokens": 15966584832, "loss": 2.9735052206299524, "ce_loss": 2.9735052206299524, "z_loss": 28.458650227748986, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1479041427373886, "tokens_per_second": 4335162.125421629, "eta_seconds": 2083.6607357842395, "mean_data_wait_seconds": 9.495019912719727e-05}
{"step": 29540, "total_steps": 46238, "tokens": 15971991552, "loss": 3.000397234252005, "ce_loss": 3.000397234252005, "z_loss": 28.067651242920846, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1503232717514038, "tokens_per_second": 4334988.22318465, "eta_seconds": 2082.4970955441204, "mean_data_wait_seconds": 6.499886512756348e-05}
{"step": 29550, "total_steps": 46238, "tokens": 15977398272, "loss": 3.0348934585397895, "ce_loss": 3.0348934585397895, "z_loss": 28.811938343626082, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15562964975833893, "tokens_per_second": 4334308.372359937, "eta_seconds": 2081.5763182737296, "mean_data_wait_seconds": 0.00010454654693603516}
{"step": 29560, "total_steps": 46238, "tokens": 15982804992, "loss": 2.9699841936429343, "ce_loss": 2.9699841936429343, "z_loss": 28.303505305087928, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15951839089393616, "tokens_per_second": 4334184.796674814, "eta_seconds": 2080.388208393347, "mean_data_wait_seconds": 9.328126907348633e-05}
{"step": 29570, "total_steps": 46238, "tokens": 15988211712, "loss": 2.9966693141243677, "ce_loss": 2.9966693141243677, "z_loss": 28.040333849011045, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1613217443227768, "tokens_per_second": 4333608.491276164, "eta_seconds": 2079.4172436528347, "mean_data_wait_seconds": 0.06597661972045898}
{"step": 29580, "total_steps": 46238, "tokens": 15993618432, "loss": 3.054735774343664, "ce_loss": 3.054735774343664, "z_loss": 28.805724852012865, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1984003633260727, "tokens_per_second": 4333531.391175294, "eta_seconds": 2078.206592050899, "mean_data_wait_seconds": 0.00019285082817077637}
{"step": 29590, "total_steps": 46238, "tokens": 15999025152, "loss": 3.0736261678464487, "ce_loss": 3.0736261678464487, "z_loss": 28.007532134200588, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16576893627643585, "tokens_per_second": 4333455.825457922, "eta_seconds": 2076.9951619499657, "mean_data_wait_seconds": 0.00012487173080444336}
{"step": 29600, "total_steps": 46238, "tokens": 16004431872, "loss": 3.0282061063882075, "ce_loss": 3.0282061063882075, "z_loss": 28.487573074572015, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.165996253490448, "tokens_per_second": 4332865.718339476, "eta_seconds": 2076.03019542625, "mean_data_wait_seconds": 0.00010177493095397949}
{"step": 29610, "total_steps": 46238, "tokens": 16009838592, "loss": 2.9848184766191426, "ce_loss": 2.9848184766191426, "z_loss": 27.764081304723565, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15477529168128967, "tokens_per_second": 4332698.885868155, "eta_seconds": 2074.8622465599055, "mean_data_wait_seconds": 0.00011396408081054688}
{"step": 29620, "total_steps": 46238, "tokens": 16015245312, "loss": 2.977878016052824, "ce_loss": 2.977878016052824, "z_loss": 27.839310429312967, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15914537012577057, "tokens_per_second": 4332674.301910015, "eta_seconds": 2073.626125102305, "mean_data_wait_seconds": 0.00012549757957458496}
{"step": 29630, "total_steps": 46238, "tokens": 16020652032, "loss": 2.9926004626534204, "ce_loss": 2.9926004626534204, "z_loss": 28.07928550604618, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1644071489572525, "tokens_per_second": 4332593.892789785, "eta_seconds": 2072.4166922135423, "mean_data_wait_seconds": 7.662177085876465e-05}
{"step": 29640, "total_steps": 46238, "tokens": 16026058752, "loss": 3.0131102475253018, "ce_loss": 3.0131102475253018, "z_loss": 28.382654681350246, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14984862506389618, "tokens_per_second": 4332627.907701728, "eta_seconds": 2071.152514170106, "mean_data_wait_seconds": 7.924437522888184e-05}
{"step": 29650, "total_steps": 46238, "tokens": 16031465472, "loss": 3.08934644316182, "ce_loss": 3.08934644316182, "z_loss": 27.913638533967912, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16551902890205383, "tokens_per_second": 4332557.014195641, "eta_seconds": 2069.938476196827, "mean_data_wait_seconds": 0.00011533498764038086}
{"step": 29660, "total_steps": 46238, "tokens": 16036872192, "loss": 3.00916397932804, "ce_loss": 3.00916397932804, "z_loss": 28.32641501860185, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16293910145759583, "tokens_per_second": 4332646.3622818785, "eta_seconds": 2068.6478873571386, "mean_data_wait_seconds": 8.589029312133789e-05}
{"step": 29670, "total_steps": 46238, "tokens": 16042278912, "loss": 3.07011515082735, "ce_loss": 3.07011515082735, "z_loss": 28.43549039147117, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1709735095500946, "tokens_per_second": 4332709.95283855, "eta_seconds": 2067.3696419793037, "mean_data_wait_seconds": 7.471442222595215e-05}
{"step": 29680, "total_steps": 46238, "tokens": 16047685632, "loss": 2.972332771980401, "ce_loss": 2.972332771980401, "z_loss": 26.914984081730697, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17363610863685608, "tokens_per_second": 4332796.916793138, "eta_seconds": 2066.08028853234, "mean_data_wait_seconds": 0.00011858344078063965}
{"step": 29690, "total_steps": 46238, "tokens": 16053092352, "loss": 3.010591604492881, "ce_loss": 3.010591604492881, "z_loss": 28.21520607399218, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15383675694465637, "tokens_per_second": 4332794.398904286, "eta_seconds": 2064.833629369182, "mean_data_wait_seconds": 0.00013747811317443848}
{"step": 29700, "total_steps": 46238, "tokens": 16058499072, "loss": 3.0319158362619802, "ce_loss": 3.0319158362619802, "z_loss": 27.909944505402535, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1832006573677063, "tokens_per_second": 4332931.408885721, "eta_seconds": 2063.520517694818, "mean_data_wait_seconds": 9.295344352722168e-05}
{"step": 29710, "total_steps": 46238, "tokens": 16063905792, "loss": 3.0015176299846535, "ce_loss": 3.0015176299846535, "z_loss": 27.026377793514367, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17485228180885315, "tokens_per_second": 4332938.712455834, "eta_seconds": 2062.269221189932, "mean_data_wait_seconds": 7.459521293640137e-05}
{"step": 29720, "total_steps": 46238, "tokens": 16069312512, "loss": 3.0814428022413543, "ce_loss": 3.0814428022413543, "z_loss": 27.40707810719808, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15409749746322632, "tokens_per_second": 4332973.92270203, "eta_seconds": 2061.004654842488, "mean_data_wait_seconds": 6.732344627380371e-05}
{"step": 29730, "total_steps": 46238, "tokens": 16074719232, "loss": 3.076881321993741, "ce_loss": 3.076881321993741, "z_loss": 28.001344016104035, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1542099267244339, "tokens_per_second": 4332986.506951202, "eta_seconds": 2059.750864601645, "mean_data_wait_seconds": 7.006525993347168e-05}
{"step": 29740, "total_steps": 46238, "tokens": 16080125952, "loss": 3.044639603658156, "ce_loss": 3.044639603658156, "z_loss": 28.507680242711846, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1756255030632019, "tokens_per_second": 4333160.529014751, "eta_seconds": 2058.420389522946, "mean_data_wait_seconds": 8.067488670349121e-05}
{"step": 29750, "total_steps": 46238, "tokens": 16085532672, "loss": 3.078638084007032, "ce_loss": 3.078638084007032, "z_loss": 27.61750874374852, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1891300231218338, "tokens_per_second": 4333125.889255152, "eta_seconds": 2057.1890805444136, "mean_data_wait_seconds": 0.00015565752983093262}
{"step": 29760, "total_steps": 46238, "tokens": 16090939392, "loss": 3.043510402693893, "ce_loss": 3.043510402693893, "z_loss": 28.41337151960893, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1378173530101776, "tokens_per_second": 4333174.385531577, "eta_seconds": 2055.918306391244, "mean_data_wait_seconds": 7.104873657226562e-05}
{"step": 29770, "total_steps": 46238, "tokens": 16096346112, "loss": 3.0520237919056052, "ce_loss": 3.0520237919056052, "z_loss": 28.26241566918113, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2277306765317917, "tokens_per_second": 4333286.101668448, "eta_seconds": 2054.617584694437, "mean_data_wait_seconds": 0.00022408366203308105}
{"step": 29780, "total_steps": 46238, "tokens": 16101752832, "loss": 3.047562938747984, "ce_loss": 3.047562938747984, "z_loss": 28.3316638397448, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15551620721817017, "tokens_per_second": 4333434.823302949, "eta_seconds": 2053.2993957015506, "mean_data_wait_seconds": 7.954239845275879e-05}
{"step": 29790, "total_steps": 46238, "tokens": 16107159552, "loss": 3.059575008623528, "ce_loss": 3.059575008623528, "z_loss": 27.534105040810324, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14434683322906494, "tokens_per_second": 4333516.76445351, "eta_seconds": 2052.012918685779, "mean_data_wait_seconds": 7.575750350952148e-05}
{"step": 29800, "total_steps": 46238, "tokens": 16112566272, "loss": 3.0457959825342353, "ce_loss": 3.0457959825342353, "z_loss": 28.11326963251287, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1589365005493164, "tokens_per_second": 4333753.194900529, "eta_seconds": 2050.653386182039, "mean_data_wait_seconds": 7.641315460205078e-05}
{"step": 29810, "total_steps": 46238, "tokens": 16117972992, "loss": 3.00552203799739, "ce_loss": 3.00552203799739, "z_loss": 28.649043704524185, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16844820976257324, "tokens_per_second": 4333715.689568928, "eta_seconds": 2049.4235386455284, "mean_data_wait_seconds": 6.48200511932373e-05}
{"step": 29820, "total_steps": 46238, "tokens": 16123379712, "loss": 3.0193121993180476, "ce_loss": 3.0193121993180476, "z_loss": 27.734657952279754, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16390781104564667, "tokens_per_second": 4333822.913160828, "eta_seconds": 2048.125269965456, "mean_data_wait_seconds": 0.00015145540237426758}
{"step": 29830, "total_steps": 46238, "tokens": 16128786432, "loss": 3.007813143007683, "ce_loss": 3.007813143007683, "z_loss": 28.1319372437217, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15224553644657135, "tokens_per_second": 4333952.855037458, "eta_seconds": 2046.8163361973927, "mean_data_wait_seconds": 8.150935173034668e-05}
{"step": 29840, "total_steps": 46238, "tokens": 16134193152, "loss": 3.0173315911582024, "ce_loss": 3.0173315911582024, "z_loss": 27.768377217379484, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14645767211914062, "tokens_per_second": 4334067.348369763, "eta_seconds": 2045.514772015408, "mean_data_wait_seconds": 8.064508438110352e-05}
{"step": 29850, "total_steps": 46238, "tokens": 16139599872, "loss": 2.974791071631692, "ce_loss": 2.974791071631692, "z_loss": 28.184234893683232, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14599764347076416, "tokens_per_second": 4334078.345427734, "eta_seconds": 2044.2620916963604, "mean_data_wait_seconds": 0.00013193488121032715}
{"step": 29860, "total_steps": 46238, "tokens": 16145006592, "loss": 3.0711795163877085, "ce_loss": 3.0711795163877085, "z_loss": 28.20359537818215, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16954198479652405, "tokens_per_second": 4334164.918405696, "eta_seconds": 2042.9737932669902, "mean_data_wait_seconds": 8.270144462585449e-05}
{"step": 29870, "total_steps": 46238, "tokens": 16150413312, "loss": 2.978929143963438, "ce_loss": 2.978929143963438, "z_loss": 29.8121686704231, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2281360775232315, "tokens_per_second": 4334204.992279601, "eta_seconds": 2041.707450331213, "mean_data_wait_seconds": 0.00014328956604003906}
{"step": 29880, "total_steps": 46238, "tokens": 16155820032, "loss": 2.905470873370315, "ce_loss": 2.905470873370315, "z_loss": 28.140527840816613, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17852284014225006, "tokens_per_second": 4334301.824160644, "eta_seconds": 2040.4144110828356, "mean_data_wait_seconds": 6.872415542602539e-05}
{"step": 29890, "total_steps": 46238, "tokens": 16161226752, "loss": 3.025568981965383, "ce_loss": 3.025568981965383, "z_loss": 28.088121500882234, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1548399180173874, "tokens_per_second": 4334395.090945981, "eta_seconds": 2039.123106811896, "mean_data_wait_seconds": 0.00016897916793823242}
{"step": 29900, "total_steps": 46238, "tokens": 16166633472, "loss": 2.9830052256584167, "ce_loss": 2.9830052256584167, "z_loss": 27.89459904757413, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15509377419948578, "tokens_per_second": 4334411.77501613, "eta_seconds": 2037.8678636196555, "mean_data_wait_seconds": 0.00011262297630310059}
{"step": 29910, "total_steps": 46238, "tokens": 16172040192, "loss": 3.013235735170769, "ce_loss": 3.013235735170769, "z_loss": 28.372251019333348, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16598835587501526, "tokens_per_second": 4334492.98118537, "eta_seconds": 2036.5823136218105, "mean_data_wait_seconds": 9.682774543762207e-05}
{"step": 29920, "total_steps": 46238, "tokens": 16177446912, "loss": 2.9589608943823613, "ce_loss": 2.9589608943823613, "z_loss": 29.43293711633393, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1653480976819992, "tokens_per_second": 4334484.565239874, "eta_seconds": 2035.3388946747295, "mean_data_wait_seconds": 0.00014463067054748535}
{"step": 29930, "total_steps": 46238, "tokens": 16182853632, "loss": 2.969034579667178, "ce_loss": 2.969034579667178, "z_loss": 29.228536504687685, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17879006266593933, "tokens_per_second": 4334551.118143091, "eta_seconds": 2034.0602899100345, "mean_data_wait_seconds": 8.627772331237793e-05}
{"step": 29940, "total_steps": 46238, "tokens": 16188260352, "loss": 3.014274748888883, "ce_loss": 3.014274748888883, "z_loss": 28.923513600320526, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1534310281276703, "tokens_per_second": 4334490.734087928, "eta_seconds": 2032.8412550763237, "mean_data_wait_seconds": 9.03010368347168e-05}
{"step": 29950, "total_steps": 46238, "tokens": 16193667072, "loss": 2.9821317647442673, "ce_loss": 2.9821317647442673, "z_loss": 29.025105823170055, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17789077758789062, "tokens_per_second": 4334502.319756061, "eta_seconds": 2031.5884533880198, "mean_data_wait_seconds": 8.907914161682129e-05}
{"step": 29960, "total_steps": 46238, "tokens": 16199073792, "loss": 2.9282401529225437, "ce_loss": 2.9282401529225437, "z_loss": 29.315457893140387, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1573300063610077, "tokens_per_second": 4334464.775166151, "eta_seconds": 2030.3586718299387, "mean_data_wait_seconds": 0.00023296475410461426}
{"step": 29970, "total_steps": 46238, "tokens": 16204480512, "loss": 2.9945869951537163, "ce_loss": 2.9945869951537163, "z_loss": 28.50253488078262, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.21787245571613312, "tokens_per_second": 4334370.306628761, "eta_seconds": 2029.1555178267101, "mean_data_wait_seconds": 9.194016456604004e-05}
{"step": 29980, "total_steps": 46238, "tokens": 16209887232, "loss": 2.9770279508648496, "ce_loss": 2.9770279508648496, "z_loss": 29.39327557881673, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1413915455341339, "tokens_per_second": 4334369.182679555, "eta_seconds": 2027.908637576208, "mean_data_wait_seconds": 9.116530418395996e-05}
{"step": 29990, "total_steps": 46238, "tokens": 16215293952, "loss": 2.9618097453406365, "ce_loss": 2.9618097453406365, "z_loss": 28.640618194233287, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1897413730621338, "tokens_per_second": 4334393.736631991, "eta_seconds": 2026.6497502891314, "mean_data_wait_seconds": 8.577108383178711e-05}
{"step": 30000, "total_steps": 46238, "tokens": 16220700672, "loss": 3.0249854416558235, "ce_loss": 3.0249854416558235, "z_loss": 29.102264071955826, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1685846596956253, "tokens_per_second": 4334295.690891636, "eta_seconds": 2025.4481673801167, "mean_data_wait_seconds": 8.624792098999023e-05}
{"step": 30010, "total_steps": 46238, "tokens": 16226107392, "loss": 3.0349820454915366, "ce_loss": 3.0349820454915366, "z_loss": 28.507309061108213, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16918818652629852, "tokens_per_second": 4334421.478752544, "eta_seconds": 2024.1419961136378, "mean_data_wait_seconds": 0.00013235211372375488}
{"step": 30020, "total_steps": 46238, "tokens": 16231514112, "loss": 2.9219034567023767, "ce_loss": 2.9219034567023767, "z_loss": 27.772740653066926, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1370873749256134, "tokens_per_second": 4334442.563350628, "eta_seconds": 2022.8847644994667, "mean_data_wait_seconds": 9.694695472717285e-05}
{"step": 30030, "total_steps": 46238, "tokens": 16236920832, "loss": 2.978530791672793, "ce_loss": 2.978530791672793, "z_loss": 28.46147766980258, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14645668864250183, "tokens_per_second": 4334557.857379995, "eta_seconds": 2021.583606060471, "mean_data_wait_seconds": 7.894635200500488e-05}
{"step": 30040, "total_steps": 46238, "tokens": 16242327552, "loss": 3.020840664704641, "ce_loss": 3.020840664704641, "z_loss": 29.091699065584123, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2190931886434555, "tokens_per_second": 4334632.6230983585, "eta_seconds": 2020.301406244754, "mean_data_wait_seconds": 7.554888725280762e-05}
{"step": 30050, "total_steps": 46238, "tokens": 16247734272, "loss": 3.007336313074285, "ce_loss": 3.007336313074285, "z_loss": 28.644392923875287, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1671583503484726, "tokens_per_second": 4334677.908544534, "eta_seconds": 2019.032982069626, "mean_data_wait_seconds": 5.894899368286133e-05}
{"step": 30060, "total_steps": 46238, "tokens": 16253140992, "loss": 2.9516763325893516, "ce_loss": 2.9516763325893516, "z_loss": 27.799647836974174, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15494218468666077, "tokens_per_second": 4334747.299602101, "eta_seconds": 2017.7533635704353, "mean_data_wait_seconds": 0.00010940432548522949}
{"step": 30070, "total_steps": 46238, "tokens": 16258547712, "loss": 3.014696645014214, "ce_loss": 3.014696645014214, "z_loss": 27.662056821765322, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20043128728866577, "tokens_per_second": 4334844.473559858, "eta_seconds": 2016.4608620483414, "mean_data_wait_seconds": 0.0001055598258972168}
{"step": 30080, "total_steps": 46238, "tokens": 16263954432, "loss": 2.9980537981698006, "ce_loss": 2.9980537981698006, "z_loss": 28.13798813386397, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18450124561786652, "tokens_per_second": 4334906.993526571, "eta_seconds": 2015.1845280752627, "mean_data_wait_seconds": 8.90195369720459e-05}
{"step": 30090, "total_steps": 46238, "tokens": 16269361152, "loss": 2.946697137572549, "ce_loss": 2.946697137572549, "z_loss": 28.220883788484514, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13913889229297638, "tokens_per_second": 4334965.619994286, "eta_seconds": 2013.9100397321045, "mean_data_wait_seconds": 9.861588478088379e-05}
{"step": 30100, "total_steps": 46238, "tokens": 16274767872, "loss": 3.051053565559965, "ce_loss": 3.051053565559965, "z_loss": 28.510394833304666, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16146886348724365, "tokens_per_second": 4334876.481639725, "eta_seconds": 2012.7041914467004, "mean_data_wait_seconds": 8.171796798706055e-05}
{"step": 30110, "total_steps": 46238, "tokens": 16280174592, "loss": 3.1277958154678345, "ce_loss": 3.1277958154678345, "z_loss": 27.51923776395393, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18987542390823364, "tokens_per_second": 4334853.329793636, "eta_seconds": 2011.4676739051504, "mean_data_wait_seconds": 7.417798042297363e-05}
{"step": 30120, "total_steps": 46238, "tokens": 16285581312, "loss": 3.013323456952066, "ce_loss": 3.013323456952066, "z_loss": 28.4194122950236, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16232271492481232, "tokens_per_second": 4334985.445911548, "eta_seconds": 2010.1591418763444, "mean_data_wait_seconds": 5.91278076171875e-05}
{"step": 30130, "total_steps": 46238, "tokens": 16290988032, "loss": 2.9816547252915124, "ce_loss": 2.9816547252915124, "z_loss": 28.776147986903336, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1313590109348297, "tokens_per_second": 4335176.868807894, "eta_seconds": 2008.8232078048363, "mean_data_wait_seconds": 6.401538848876953e-05}
{"step": 30140, "total_steps": 46238, "tokens": 16296394752, "loss": 3.007136915669297, "ce_loss": 3.007136915669297, "z_loss": 29.266955332322553, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14771609008312225, "tokens_per_second": 4335234.817530418, "eta_seconds": 2007.5491986747347, "mean_data_wait_seconds": 7.677078247070312e-05}
{"step": 30150, "total_steps": 46238, "tokens": 16301801472, "loss": 3.062573967557965, "ce_loss": 3.062573967557965, "z_loss": 28.28168181217078, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18372969329357147, "tokens_per_second": 4335326.354179692, "eta_seconds": 2006.2596799925918, "mean_data_wait_seconds": 0.00011801719665527344}
{"step": 30160, "total_steps": 46238, "tokens": 16307208192, "loss": 3.0296566233490454, "ce_loss": 3.0296566233490454, "z_loss": 28.945705818407465, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15935371816158295, "tokens_per_second": 4335439.456470606, "eta_seconds": 2004.960242502451, "mean_data_wait_seconds": 9.641051292419434e-05}
{"step": 30170, "total_steps": 46238, "tokens": 16312614912, "loss": 3.026018458785433, "ce_loss": 3.026018458785433, "z_loss": 29.174777507781982, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15220624208450317, "tokens_per_second": 4335578.2951852, "eta_seconds": 2003.6489788795118, "mean_data_wait_seconds": 7.265806198120117e-05}
{"step": 30180, "total_steps": 46238, "tokens": 16318021632, "loss": 2.9888864611134385, "ce_loss": 2.9888864611134385, "z_loss": 28.77796764084787, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18409882485866547, "tokens_per_second": 4335720.764452401, "eta_seconds": 2002.3361225608076, "mean_data_wait_seconds": 0.00011774897575378418}
{"step": 30190, "total_steps": 46238, "tokens": 16323428352, "loss": 3.007866783575578, "ce_loss": 3.007866783575578, "z_loss": 29.2532664790298, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20349743962287903, "tokens_per_second": 4335840.793590698, "eta_seconds": 2001.033708808042, "mean_data_wait_seconds": 5.6415796279907227e-05}
{"step": 30200, "total_steps": 46238, "tokens": 16328835072, "loss": 3.050670387166919, "ce_loss": 3.050670387166919, "z_loss": 27.81577155084321, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1585141122341156, "tokens_per_second": 4335949.807731258, "eta_seconds": 1999.736447257651, "mean_data_wait_seconds": 6.222724914550781e-05}
{"step": 30210, "total_steps": 46238, "tokens": 16334241792, "loss": 3.06183455026511, "ce_loss": 3.06183455026511, "z_loss": 28.937318700732607, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.175747349858284, "tokens_per_second": 4336016.221431923, "eta_seconds": 1998.4588851787921, "mean_data_wait_seconds": 7.18832015991211e-05}
{"step": 30220, "total_steps": 46238, "tokens": 16339648512, "loss": 3.0092245430657356, "ce_loss": 3.0092245430657356, "z_loss": 29.265600768002596, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14923375844955444, "tokens_per_second": 4336125.111719186, "eta_seconds": 1997.1617978906763, "mean_data_wait_seconds": 9.518861770629883e-05}
{"step": 30230, "total_steps": 46238, "tokens": 16345055232, "loss": 3.0469374602491204, "ce_loss": 3.0469374602491204, "z_loss": 29.281040639588326, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14851872622966766, "tokens_per_second": 4336282.0498691695, "eta_seconds": 1995.842660710024, "mean_data_wait_seconds": 0.00012561678886413574}
{"step": 30240, "total_steps": 46238, "tokens": 16350461952, "loss": 3.018205413312623, "ce_loss": 3.018205413312623, "z_loss": 28.771775404612224, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17666038870811462, "tokens_per_second": 4336247.874180689, "eta_seconds": 1994.6115247468888, "mean_data_wait_seconds": 0.00019696354866027832}
{"step": 30250, "total_steps": 46238, "tokens": 16355868672, "loss": 3.0088125918850754, "ce_loss": 3.0088125918850754, "z_loss": 27.966766285173822, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1687978059053421, "tokens_per_second": 4336251.44422107, "eta_seconds": 1993.363017617326, "mean_data_wait_seconds": 7.978081703186035e-05}
{"step": 30260, "total_steps": 46238, "tokens": 16361275392, "loss": 2.9702690334031074, "ce_loss": 2.9702690334031074, "z_loss": 28.519890423977014, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15644274652004242, "tokens_per_second": 4336162.840662993, "eta_seconds": 1992.156858822953, "mean_data_wait_seconds": 6.434321403503418e-05}
{"step": 30270, "total_steps": 46238, "tokens": 16366682112, "loss": 3.0057085752487183, "ce_loss": 3.0057085752487183, "z_loss": 27.722982160972826, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13521349430084229, "tokens_per_second": 4336205.828339801, "eta_seconds": 1990.8902311736604, "mean_data_wait_seconds": 7.027387619018555e-05}
{"step": 30280, "total_steps": 46238, "tokens": 16372088832, "loss": 2.9921947302240315, "ce_loss": 2.9921947302240315, "z_loss": 28.94414739897757, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1612139791250229, "tokens_per_second": 4336267.450724502, "eta_seconds": 1989.6150784146212, "mean_data_wait_seconds": 7.012486457824707e-05}
{"step": 30290, "total_steps": 46238, "tokens": 16377495552, "loss": 2.9161902593843863, "ce_loss": 2.9161902593843863, "z_loss": 28.562463688127924, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16609902679920197, "tokens_per_second": 4336258.353114162, "eta_seconds": 1988.3723897142538, "mean_data_wait_seconds": 5.939602851867676e-05}
{"step": 30300, "total_steps": 46238, "tokens": 16382902272, "loss": 3.013383202480547, "ce_loss": 3.013383202480547, "z_loss": 28.48972362460512, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1621459275484085, "tokens_per_second": 4336251.734695983, "eta_seconds": 1987.1285596854586, "mean_data_wait_seconds": 5.570054054260254e-05}
{"step": 30310, "total_steps": 46238, "tokens": 16388308992, "loss": 3.105092091993852, "ce_loss": 3.105092091993852, "z_loss": 28.03887738603534, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17301523685455322, "tokens_per_second": 4336307.525802846, "eta_seconds": 1985.8561443715096, "mean_data_wait_seconds": 0.003241986036300659}
{"step": 30320, "total_steps": 46238, "tokens": 16393715712, "loss": 3.0339631203449136, "ce_loss": 3.0339631203449136, "z_loss": 29.171397628206194, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1662001758813858, "tokens_per_second": 4336359.834149253, "eta_seconds": 1984.5853557234557, "mean_data_wait_seconds": 6.058812141418457e-05}
{"step": 30330, "total_steps": 46238, "tokens": 16399122432, "loss": 3.021809281724872, "ce_loss": 3.021809281724872, "z_loss": 29.59935406482581, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15929296612739563, "tokens_per_second": 4336479.720988519, "eta_seconds": 1983.2836903107868, "mean_data_wait_seconds": 5.796551704406738e-05}
{"step": 30340, "total_steps": 46238, "tokens": 16404529152, "loss": 2.9874605951887188, "ce_loss": 2.9874605951887188, "z_loss": 28.817010055888783, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16463437676429749, "tokens_per_second": 4336638.245031201, "eta_seconds": 1981.9644384329226, "mean_data_wait_seconds": 0.00011816620826721191}
{"step": 30350, "total_steps": 46238, "tokens": 16409935872, "loss": 2.9884488961913367, "ce_loss": 2.9884488961913367, "z_loss": 28.839563283053312, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16108736395835876, "tokens_per_second": 4336686.463505182, "eta_seconds": 1980.6956616036523, "mean_data_wait_seconds": 9.79006290435791e-05}
{"step": 30360, "total_steps": 46238, "tokens": 16415342592, "loss": 2.983229131409616, "ce_loss": 2.983229131409616, "z_loss": 27.869987429994524, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1525062620639801, "tokens_per_second": 4336919.6669065505, "eta_seconds": 1979.3424834458592, "mean_data_wait_seconds": 0.00015938282012939453}
{"step": 30370, "total_steps": 46238, "tokens": 16420749312, "loss": 2.9930262944915076, "ce_loss": 2.9930262944915076, "z_loss": 29.356756658265084, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17800790071487427, "tokens_per_second": 4337020.968296073, "eta_seconds": 1978.049607486784, "mean_data_wait_seconds": 0.00011673569679260254}
{"step": 30380, "total_steps": 46238, "tokens": 16426156032, "loss": 2.9320040200695847, "ce_loss": 2.9320040200695847, "z_loss": 28.260524634158973, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13633815944194794, "tokens_per_second": 4336979.687878325, "eta_seconds": 1976.8217794430516, "mean_data_wait_seconds": 0.00012603402137756348}
{"step": 30390, "total_steps": 46238, "tokens": 16431562752, "loss": 3.0190705368013093, "ce_loss": 3.0190705368013093, "z_loss": 28.74903337883227, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15761658549308777, "tokens_per_second": 4337013.692541011, "eta_seconds": 1975.5596342099814, "mean_data_wait_seconds": 0.0001450181007385254}
{"step": 30400, "total_steps": 46238, "tokens": 16436969472, "loss": 2.9568642124985205, "ce_loss": 2.9568642124985205, "z_loss": 28.964128306417756, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1496325582265854, "tokens_per_second": 4337140.985372251, "eta_seconds": 1974.2550433289828, "mean_data_wait_seconds": 0.00010946393013000488}
{"step": 30410, "total_steps": 46238, "tokens": 16442376192, "loss": 2.9780742688612505, "ce_loss": 2.9780742688612505, "z_loss": 29.132072679924242, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19513368606567383, "tokens_per_second": 4337174.919339233, "eta_seconds": 1972.9929973180995, "mean_data_wait_seconds": 8.437037467956543e-05}
{"step": 30420, "total_steps": 46238, "tokens": 16447782912, "loss": 2.966597613060113, "ce_loss": 2.966597613060113, "z_loss": 29.073507670200232, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17238213121891022, "tokens_per_second": 4337180.022523724, "eta_seconds": 1971.7440778545003, "mean_data_wait_seconds": 7.849931716918945e-05}
{"step": 30430, "total_steps": 46238, "tokens": 16453189632, "loss": 2.996049743710142, "ce_loss": 2.996049743710142, "z_loss": 29.33303911035711, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14671771228313446, "tokens_per_second": 4337258.97247967, "eta_seconds": 1970.461611406594, "mean_data_wait_seconds": 9.110569953918457e-05}
{"step": 30440, "total_steps": 46238, "tokens": 16458596352, "loss": 2.975624691356312, "ce_loss": 2.975624691356312, "z_loss": 28.93280988751036, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1512478142976761, "tokens_per_second": 4337173.9183389805, "eta_seconds": 1969.2536533722791, "mean_data_wait_seconds": 0.0002808868885040283}
{"step": 30450, "total_steps": 46238, "tokens": 16464003072, "loss": 2.9539306217973884, "ce_loss": 2.9539306217973884, "z_loss": 29.512036063454367, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1329578161239624, "tokens_per_second": 4336994.225775069, "eta_seconds": 1968.0885930795991, "mean_data_wait_seconds": 7.578730583190918e-05}
{"step": 30460, "total_steps": 46238, "tokens": 16469409792, "loss": 2.9957773793827402, "ce_loss": 2.9957773793827402, "z_loss": 28.674164945429023, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17384357750415802, "tokens_per_second": 4336979.243071519, "eta_seconds": 1966.8487363934873, "mean_data_wait_seconds": 7.197260856628418e-05}
{"step": 30470, "total_steps": 46238, "tokens": 16474816512, "loss": 3.0106333584496467, "ce_loss": 3.0106333584496467, "z_loss": 27.99174872311679, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15835288166999817, "tokens_per_second": 4337003.604657459, "eta_seconds": 1965.5910395936355, "mean_data_wait_seconds": 8.96155834197998e-05}
{"step": 30480, "total_steps": 46238, "tokens": 16480223232, "loss": 2.9741868737972146, "ce_loss": 2.9741868737972146, "z_loss": 28.698778932744805, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14723704755306244, "tokens_per_second": 4337068.672132034, "eta_seconds": 1964.3149205225782, "mean_data_wait_seconds": 0.00010579824447631836}
{"step": 30490, "total_steps": 46238, "tokens": 16485629952, "loss": 2.9983442136735627, "ce_loss": 2.9983442136735627, "z_loss": 28.828056638891045, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14867736399173737, "tokens_per_second": 4337049.668909561, "eta_seconds": 1963.0768918864183, "mean_data_wait_seconds": 7.784366607666016e-05}
{"step": 30500, "total_steps": 46238, "tokens": 16491036672, "loss": 2.968755841255188, "ce_loss": 2.968755841255188, "z_loss": 29.87123694564357, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18718986213207245, "tokens_per_second": 4337183.840121518, "eta_seconds": 1961.7695669920254, "mean_data_wait_seconds": 9.182095527648926e-05}
{"step": 30510, "total_steps": 46238, "tokens": 16496443392, "loss": 2.9701513196482803, "ce_loss": 2.9701513196482803, "z_loss": 28.920424967101127, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15034975111484528, "tokens_per_second": 4337206.894070249, "eta_seconds": 1960.5125491304902, "mean_data_wait_seconds": 7.534027099609375e-05}
{"step": 30520, "total_steps": 46238, "tokens": 16501850112, "loss": 3.0167729077917156, "ce_loss": 3.0167729077917156, "z_loss": 29.01469562992905, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17593805491924286, "tokens_per_second": 4337326.515634192, "eta_seconds": 1959.2119231442005, "mean_data_wait_seconds": 8.761882781982422e-05}
{"step": 30530, "total_steps": 46238, "tokens": 16507256832, "loss": 2.9644645885987715, "ce_loss": 2.9644645885987715, "z_loss": 29.453127297488127, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17608530819416046, "tokens_per_second": 4337267.859278537, "eta_seconds": 1957.9918463722963, "mean_data_wait_seconds": 8.079409599304199e-05}
{"step": 30540, "total_steps": 46238, "tokens": 16512663552, "loss": 2.969101652954564, "ce_loss": 2.969101652954564, "z_loss": 29.07746258648959, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1294524371623993, "tokens_per_second": 4337318.515762545, "eta_seconds": 1956.7224203519004, "mean_data_wait_seconds": 0.00032970309257507324}
{"step": 30550, "total_steps": 46238, "tokens": 16518070272, "loss": 3.0017172116221804, "ce_loss": 3.0017172116221804, "z_loss": 29.844230044971813, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19313132762908936, "tokens_per_second": 4337165.089849903, "eta_seconds": 1955.545036514513, "mean_data_wait_seconds": 0.000614851713180542}
{"step": 30560, "total_steps": 46238, "tokens": 16523476992, "loss": 2.944148343620878, "ce_loss": 2.944148343620878, "z_loss": 29.128386728691332, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1571495234966278, "tokens_per_second": 4337201.022000937, "eta_seconds": 1954.2822435492292, "mean_data_wait_seconds": 9.521842002868652e-05}
{"step": 30570, "total_steps": 46238, "tokens": 16528883712, "loss": 2.951540652549628, "ce_loss": 2.951540652549628, "z_loss": 27.664068987875275, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15838170051574707, "tokens_per_second": 4337151.30356882, "eta_seconds": 1953.058039969666, "mean_data_wait_seconds": 0.00011628866195678711}
{"step": 30580, "total_steps": 46238, "tokens": 16534290432, "loss": 2.936417028759465, "ce_loss": 2.936417028759465, "z_loss": 29.64767425710505, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20315490663051605, "tokens_per_second": 4337238.705376985, "eta_seconds": 1951.772101799551, "mean_data_wait_seconds": 7.042288780212402e-05}
{"step": 30590, "total_steps": 46238, "tokens": 16539697152, "loss": 2.97220180793242, "ce_loss": 2.97220180793242, "z_loss": 29.29198681224476, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17974554002285004, "tokens_per_second": 4337285.111826305, "eta_seconds": 1950.5046511544138, "mean_data_wait_seconds": 8.305907249450684e-05}
{"step": 30600, "total_steps": 46238, "tokens": 16545103872, "loss": 2.9770770795417554, "ce_loss": 2.9770770795417554, "z_loss": 30.043301914677475, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16769731044769287, "tokens_per_second": 4337336.067135014, "eta_seconds": 1949.2351833332878, "mean_data_wait_seconds": 0.0002785325050354004}
{"step": 30610, "total_steps": 46238, "tokens": 16550510592, "loss": 3.003557441812573, "ce_loss": 3.003557441812573, "z_loss": 29.195266535787873, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1668033003807068, "tokens_per_second": 4337435.08726656, "eta_seconds": 1947.9441591654088, "mean_data_wait_seconds": 8.425116539001465e-05}
{"step": 30620, "total_steps": 46238, "tokens": 16555917312, "loss": 2.999023294810093, "ce_loss": 2.999023294810093, "z_loss": 28.92732523426865, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1412491649389267, "tokens_per_second": 4337548.389751536, "eta_seconds": 1946.6467841488845, "mean_data_wait_seconds": 0.0001232922077178955}
{"step": 30630, "total_steps": 46238, "tokens": 16561324032, "loss": 2.994974223050204, "ce_loss": 2.994974223050204, "z_loss": 29.455806457635127, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20305094122886658, "tokens_per_second": 4337615.957258645, "eta_seconds": 1945.3699882948947, "mean_data_wait_seconds": 9.050965309143066e-05}
{"step": 30640, "total_steps": 46238, "tokens": 16566730752, "loss": 2.9785910682244734, "ce_loss": 2.9785910682244734, "z_loss": 29.109204003305145, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15485666692256927, "tokens_per_second": 4337324.7282987265, "eta_seconds": 1944.2540534214756, "mean_data_wait_seconds": 8.499622344970703e-05}
{"step": 30650, "total_steps": 46238, "tokens": 16572137472, "loss": 3.0334224809299815, "ce_loss": 3.0334224809299815, "z_loss": 29.784581025441486, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16918708384037018, "tokens_per_second": 4337487.933412684, "eta_seconds": 1942.934388146961, "mean_data_wait_seconds": 0.0003713071346282959}
{"step": 30660, "total_steps": 46238, "tokens": 16577544192, "loss": 3.0225951328422083, "ce_loss": 3.0225951328422083, "z_loss": 29.94787430040764, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1773068606853485, "tokens_per_second": 4337666.436710466, "eta_seconds": 1941.6079744451226, "mean_data_wait_seconds": 7.316470146179199e-05}
{"step": 30670, "total_steps": 46238, "tokens": 16582950912, "loss": 3.0047527623899057, "ce_loss": 3.0047527623899057, "z_loss": 28.3748170968258, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17909815907478333, "tokens_per_second": 4337836.960812362, "eta_seconds": 1940.2852389417112, "mean_data_wait_seconds": 7.331371307373047e-05}
{"step": 30680, "total_steps": 46238, "tokens": 16588357632, "loss": 2.997716403368748, "ce_loss": 2.997716403368748, "z_loss": 28.36107438983339, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1537027657032013, "tokens_per_second": 4337989.632293652, "eta_seconds": 1938.9705870626244, "mean_data_wait_seconds": 0.00011104345321655273}
{"step": 30690, "total_steps": 46238, "tokens": 16593764352, "loss": 3.043464830427459, "ce_loss": 3.043464830427459, "z_loss": 29.271349863572553, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15282978117465973, "tokens_per_second": 4338073.29448627, "eta_seconds": 1937.686851599276, "mean_data_wait_seconds": 6.860494613647461e-05}
{"step": 30700, "total_steps": 46238, "tokens": 16599171072, "loss": 3.0065294142925376, "ce_loss": 3.0065294142925376, "z_loss": 29.317062638022684, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16211046278476715, "tokens_per_second": 4338141.726664027, "eta_seconds": 1936.4099638256428, "mean_data_wait_seconds": 5.677342414855957e-05}
{"step": 30710, "total_steps": 46238, "tokens": 16604577792, "loss": 3.0001785917715593, "ce_loss": 3.0001785917715593, "z_loss": 29.17523362419822, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14696060121059418, "tokens_per_second": 4338196.801705981, "eta_seconds": 1935.1390745340761, "mean_data_wait_seconds": 6.681680679321289e-05}
{"step": 30720, "total_steps": 46238, "tokens": 16609984512, "loss": 3.0331509293931904, "ce_loss": 3.0331509293931904, "z_loss": 28.793933319322992, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16544926166534424, "tokens_per_second": 4338134.160199563, "eta_seconds": 1933.9206935947, "mean_data_wait_seconds": 0.00013300776481628418}
{"step": 30730, "total_steps": 46238, "tokens": 16615391232, "loss": 3.056092011206078, "ce_loss": 3.056092011206078, "z_loss": 29.236056009928387, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1819477379322052, "tokens_per_second": 4338119.837093482, "eta_seconds": 1932.6807508428287, "mean_data_wait_seconds": 0.0001291334629058838}
{"step": 30740, "total_steps": 46238, "tokens": 16620797952, "loss": 3.0307194102894175, "ce_loss": 3.0307194102894175, "z_loss": 28.62077507828221, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1800418198108673, "tokens_per_second": 4338193.452443052, "eta_seconds": 1931.401648140307, "mean_data_wait_seconds": 0.00010898709297180176}
{"step": 30750, "total_steps": 46238, "tokens": 16626204672, "loss": 2.997811284932223, "ce_loss": 2.997811284932223, "z_loss": 29.474821986574113, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1470007598400116, "tokens_per_second": 4338173.841859121, "eta_seconds": 1930.1640665491616, "mean_data_wait_seconds": 7.763504981994629e-05}
{"step": 30760, "total_steps": 46238, "tokens": 16631611392, "loss": 3.072243155855121, "ce_loss": 3.072243155855121, "z_loss": 29.150086720784504, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16018016636371613, "tokens_per_second": 4338185.185648818, "eta_seconds": 1928.912710246252, "mean_data_wait_seconds": 7.209181785583496e-05}
{"step": 30770, "total_steps": 46238, "tokens": 16637018112, "loss": 3.0372438701716336, "ce_loss": 3.0372438701716336, "z_loss": 29.77903038082701, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16539102792739868, "tokens_per_second": 4338398.583663293, "eta_seconds": 1927.5715826319351, "mean_data_wait_seconds": 6.61611557006836e-05}
{"step": 30780, "total_steps": 46238, "tokens": 16642424832, "loss": 3.0346833648103657, "ce_loss": 3.0346833648103657, "z_loss": 28.978260285926588, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2020215541124344, "tokens_per_second": 4338584.458893152, "eta_seconds": 1926.242806422641, "mean_data_wait_seconds": 0.00010180473327636719}
{"step": 30790, "total_steps": 46238, "tokens": 16647831552, "loss": 3.001335064570109, "ce_loss": 3.001335064570109, "z_loss": 29.334719108812738, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13802853226661682, "tokens_per_second": 4338713.617279303, "eta_seconds": 1924.939307065207, "mean_data_wait_seconds": 0.00010716915130615234}
{"step": 30800, "total_steps": 46238, "tokens": 16653238272, "loss": 2.9770790374640264, "ce_loss": 2.9770790374640264, "z_loss": 29.526857376098633, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15150168538093567, "tokens_per_second": 4338779.004747957, "eta_seconds": 1923.6641587106708, "mean_data_wait_seconds": 9.47713851928711e-05}
{"step": 30810, "total_steps": 46238, "tokens": 16658644992, "loss": 3.0255210200945535, "ce_loss": 3.0255210200945535, "z_loss": 29.407458406506162, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1566586196422577, "tokens_per_second": 4338877.425533238, "eta_seconds": 1922.3744130026714, "mean_data_wait_seconds": 8.550286293029785e-05}
{"step": 30820, "total_steps": 46238, "tokens": 16664051712, "loss": 3.052275897878589, "ce_loss": 3.052275897878589, "z_loss": 28.525024298465613, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19049470126628876, "tokens_per_second": 4338996.504068704, "eta_seconds": 1921.0755796147132, "mean_data_wait_seconds": 0.00012433528900146484}
{"step": 30830, "total_steps": 46238, "tokens": 16669458432, "loss": 2.982300360997518, "ce_loss": 2.982300360997518, "z_loss": 29.658127669132117, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14077450335025787, "tokens_per_second": 4338961.169490412, "eta_seconds": 1919.845137719527, "mean_data_wait_seconds": 7.283687591552734e-05}
{"step": 30840, "total_steps": 46238, "tokens": 16674865152, "loss": 3.069993647662076, "ce_loss": 3.069993647662076, "z_loss": 28.189817775379527, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1856461465358734, "tokens_per_second": 4338967.449747389, "eta_seconds": 1918.5962744395924, "mean_data_wait_seconds": 7.531046867370605e-05}
{"step": 30850, "total_steps": 46238, "tokens": 16680271872, "loss": 3.07071051272479, "ce_loss": 3.07071051272479, "z_loss": 28.692761132211395, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16053438186645508, "tokens_per_second": 4338946.222391966, "eta_seconds": 1917.3595701800934, "mean_data_wait_seconds": 9.003281593322754e-05}
{"step": 30860, "total_steps": 46238, "tokens": 16685678592, "loss": 3.0032374913042243, "ce_loss": 3.0032374913042243, "z_loss": 29.22805777463046, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16537408530712128, "tokens_per_second": 4338959.093559437, "eta_seconds": 1916.1077956095073, "mean_data_wait_seconds": 7.525086402893066e-05}
{"step": 30870, "total_steps": 46238, "tokens": 16691085312, "loss": 2.995936540040103, "ce_loss": 2.995936540040103, "z_loss": 29.717487797592625, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1568235158920288, "tokens_per_second": 4339025.63807981, "eta_seconds": 1914.832341870384, "mean_data_wait_seconds": 7.647275924682617e-05}
{"step": 30880, "total_steps": 46238, "tokens": 16696492032, "loss": 2.994104047616323, "ce_loss": 2.994104047616323, "z_loss": 29.87746439558087, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15104182064533234, "tokens_per_second": 4339202.728664357, "eta_seconds": 1913.5081772396845, "mean_data_wait_seconds": 7.370114326477051e-05}
{"step": 30890, "total_steps": 46238, "tokens": 16701898752, "loss": 3.0195788542429605, "ce_loss": 3.0195788542429605, "z_loss": 28.433879953442197, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2017165869474411, "tokens_per_second": 4339328.986282452, "eta_seconds": 1912.2065209231162, "mean_data_wait_seconds": 7.221102714538574e-05}
{"step": 30900, "total_steps": 46238, "tokens": 16707305472, "loss": 2.9994874000549316, "ce_loss": 2.9994874000549316, "z_loss": 28.840741446523957, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17221011221408844, "tokens_per_second": 4339449.56908435, "eta_seconds": 1910.90743929298, "mean_data_wait_seconds": 6.86347484588623e-05}
{"step": 30910, "total_steps": 46238, "tokens": 16712712192, "loss": 2.9867844834472193, "ce_loss": 2.9867844834472193, "z_loss": 29.199680024927314, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15835875272750854, "tokens_per_second": 4339513.270463186, "eta_seconds": 1909.6334606013277, "mean_data_wait_seconds": 6.267428398132324e-05}
{"step": 30920, "total_steps": 46238, "tokens": 16718118912, "loss": 2.8916636633150503, "ce_loss": 2.8916636633150503, "z_loss": 30.509244803226355, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15524068474769592, "tokens_per_second": 4339696.709359011, "eta_seconds": 1908.3068653484784, "mean_data_wait_seconds": 5.829334259033203e-05}
{"step": 30930, "total_steps": 46238, "tokens": 16723525632, "loss": 2.909564789497491, "ce_loss": 2.909564789497491, "z_loss": 29.938388997858222, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18679046630859375, "tokens_per_second": 4339767.227953189, "eta_seconds": 1907.0300016766867, "mean_data_wait_seconds": 9.15825366973877e-05}
{"step": 30940, "total_steps": 46238, "tokens": 16728932352, "loss": 2.972155413844369, "ce_loss": 2.972155413844369, "z_loss": 28.57426797982418, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1684967279434204, "tokens_per_second": 4339844.338130807, "eta_seconds": 1905.750284942758, "mean_data_wait_seconds": 6.064772605895996e-05}
{"step": 30950, "total_steps": 46238, "tokens": 16734339072, "loss": 2.993171518499201, "ce_loss": 2.993171518499201, "z_loss": 29.816497788284764, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1619691699743271, "tokens_per_second": 4339965.716343336, "eta_seconds": 1904.4511879148986, "mean_data_wait_seconds": 5.385279655456543e-05}
{"step": 30960, "total_steps": 46238, "tokens": 16739745792, "loss": 2.928824888937401, "ce_loss": 2.928824888937401, "z_loss": 29.807159423828125, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17058712244033813, "tokens_per_second": 4340079.170125489, "eta_seconds": 1903.155638462967, "mean_data_wait_seconds": 5.9932470321655273e-05}
{"step": 30970, "total_steps": 46238, "tokens": 16745152512, "loss": 2.957420085415696, "ce_loss": 2.957420085415696, "z_loss": 29.214580839330498, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1669585257768631, "tokens_per_second": 4340198.232358517, "eta_seconds": 1901.8576991388793, "mean_data_wait_seconds": 6.720423698425293e-05}
{"step": 30980, "total_steps": 46238, "tokens": 16750559232, "loss": 2.964605790196043, "ce_loss": 2.964605790196043, "z_loss": 28.771738269112326, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19349278509616852, "tokens_per_second": 4340391.025417709, "eta_seconds": 1900.5275459498796, "mean_data_wait_seconds": 0.0001277923583984375}
{"step": 30990, "total_steps": 46238, "tokens": 16755965952, "loss": 2.9941498763633496, "ce_loss": 2.9941498763633496, "z_loss": 30.144886667078193, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1753874123096466, "tokens_per_second": 4340484.502896243, "eta_seconds": 1899.240966878084, "mean_data_wait_seconds": 8.296966552734375e-05}
{"step": 31000, "total_steps": 46238, "tokens": 16761372672, "loss": 2.9251806826302498, "ce_loss": 2.9251806826302498, "z_loss": 30.103518587170225, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18101459741592407, "tokens_per_second": 4340478.10385377, "eta_seconds": 1897.9981160797822, "mean_data_wait_seconds": 8.767843246459961e-05}
{"step": 31010, "total_steps": 46238, "tokens": 16766779392, "loss": 3.0175616361878137, "ce_loss": 3.0175616361878137, "z_loss": 29.607638084527217, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14095686376094818, "tokens_per_second": 4340540.122869608, "eta_seconds": 1896.7253638833183, "mean_data_wait_seconds": 6.213784217834473e-05}
{"step": 31020, "total_steps": 46238, "tokens": 16772186112, "loss": 2.9979870933474917, "ce_loss": 2.9979870933474917, "z_loss": 30.076464811960857, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1616532802581787, "tokens_per_second": 4340609.501265258, "eta_seconds": 1895.4494343713177, "mean_data_wait_seconds": 6.690621376037598e-05}
{"step": 31030, "total_steps": 46238, "tokens": 16777592832, "loss": 2.9535984378872495, "ce_loss": 2.9535984378872495, "z_loss": 29.477987997459643, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16189120709896088, "tokens_per_second": 4340603.776461728, "eta_seconds": 1894.206319541614, "mean_data_wait_seconds": 5.882978439331055e-05}
{"step": 31040, "total_steps": 46238, "tokens": 16782999552, "loss": 3.0356117685635886, "ce_loss": 3.0356117685635886, "z_loss": 30.167381517814867, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14191894233226776, "tokens_per_second": 4340693.053376151, "eta_seconds": 1892.9217714690076, "mean_data_wait_seconds": 6.29425048828125e-05}
{"step": 31050, "total_steps": 46238, "tokens": 16788406272, "loss": 3.022041384017829, "ce_loss": 3.022041384017829, "z_loss": 29.16204533432469, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19468404352664948, "tokens_per_second": 4340828.998358789, "eta_seconds": 1891.6169393230055, "mean_data_wait_seconds": 5.811452865600586e-05}
{"step": 31060, "total_steps": 46238, "tokens": 16793812992, "loss": 2.9911965637496025, "ce_loss": 2.9911965637496025, "z_loss": 28.374953428904217, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14808028936386108, "tokens_per_second": 4340907.290626858, "eta_seconds": 1890.3372946292586, "mean_data_wait_seconds": 0.00012800097465515137}
{"step": 31070, "total_steps": 46238, "tokens": 16799219712, "loss": 3.0100045601526895, "ce_loss": 3.0100045601526895, "z_loss": 29.69872018062707, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1374351680278778, "tokens_per_second": 4340941.488000477, "eta_seconds": 1889.0768849725393, "mean_data_wait_seconds": 0.00010398030281066895}
{"step": 31080, "total_steps": 46238, "tokens": 16804626432, "loss": 3.024941007296244, "ce_loss": 3.024941007296244, "z_loss": 29.401813131390195, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16082268953323364, "tokens_per_second": 4340328.339796816, "eta_seconds": 1888.098056743705, "mean_data_wait_seconds": 0.00011670589447021484}
{"step": 31090, "total_steps": 46238, "tokens": 16810033152, "loss": 2.9782511880903533, "ce_loss": 2.9782511880903533, "z_loss": 29.04222602555246, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1453254371881485, "tokens_per_second": 4339786.767494972, "eta_seconds": 1887.0878277568481, "mean_data_wait_seconds": 0.00013399124145507812}
{"step": 31100, "total_steps": 46238, "tokens": 16815439872, "loss": 2.9858791918465584, "ce_loss": 2.9858791918465584, "z_loss": 28.788574045354668, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19555123150348663, "tokens_per_second": 4339842.062235576, "eta_seconds": 1885.8179506615757, "mean_data_wait_seconds": 0.0001423954963684082}
{"step": 31110, "total_steps": 46238, "tokens": 16820846592, "loss": 2.9621841763005112, "ce_loss": 2.9621841763005112, "z_loss": 29.17309164278435, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18831677734851837, "tokens_per_second": 4339908.9338921355, "eta_seconds": 1884.5430788026013, "mean_data_wait_seconds": 0.0001048743724822998}
{"step": 31120, "total_steps": 46238, "tokens": 16826253312, "loss": 2.9447745283444724, "ce_loss": 2.9447745283444724, "z_loss": 29.608137202985358, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16117414832115173, "tokens_per_second": 4339438.040253318, "eta_seconds": 1883.5016304375843, "mean_data_wait_seconds": 0.06393745541572571}
{"step": 31130, "total_steps": 46238, "tokens": 16831660032, "loss": 2.945211634491429, "ce_loss": 2.945211634491429, "z_loss": 29.915948781100187, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14497114717960358, "tokens_per_second": 4339479.063518655, "eta_seconds": 1882.2378871847939, "mean_data_wait_seconds": 9.968876838684082e-05}
{"step": 31140, "total_steps": 46238, "tokens": 16837066752, "loss": 2.9664095134446113, "ce_loss": 2.9664095134446113, "z_loss": 30.042607813170463, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15130844712257385, "tokens_per_second": 4339075.018285132, "eta_seconds": 1881.1671034961626, "mean_data_wait_seconds": 0.0001201927661895752}
{"step": 31150, "total_steps": 46238, "tokens": 16842473472, "loss": 2.964709971890305, "ce_loss": 2.964709971890305, "z_loss": 29.97626238158255, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15291135013103485, "tokens_per_second": 4338990.009527066, "eta_seconds": 1879.9578810021496, "mean_data_wait_seconds": 0.00013506412506103516}
{"step": 31160, "total_steps": 46238, "tokens": 16847880192, "loss": 2.9260083545338023, "ce_loss": 2.9260083545338023, "z_loss": 29.67996379100915, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14784032106399536, "tokens_per_second": 4338577.073853477, "eta_seconds": 1878.89061441975, "mean_data_wait_seconds": 0.00011119246482849121}
{"step": 31170, "total_steps": 46238, "tokens": 16853286912, "loss": 2.9208427432811623, "ce_loss": 2.9208427432811623, "z_loss": 30.317954381306965, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2020561844110489, "tokens_per_second": 4338580.436898513, "eta_seconds": 1877.642962365701, "mean_data_wait_seconds": 0.0001621544361114502}
{"step": 31180, "total_steps": 46238, "tokens": 16858693632, "loss": 2.9421068213202735, "ce_loss": 2.9421068213202735, "z_loss": 29.03235067020763, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16122967004776, "tokens_per_second": 4338603.715155384, "eta_seconds": 1876.3866991499224, "mean_data_wait_seconds": 0.00010669231414794922}
{"step": 31190, "total_steps": 46238, "tokens": 16864100352, "loss": 2.958191187092752, "ce_loss": 2.958191187092752, "z_loss": 30.35651144836888, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17036327719688416, "tokens_per_second": 4338475.382312424, "eta_seconds": 1875.1959771784514, "mean_data_wait_seconds": 0.00012385845184326172}
{"step": 31200, "total_steps": 46238, "tokens": 16869507072, "loss": 2.9717152805039375, "ce_loss": 2.9717152805039375, "z_loss": 29.906415448044285, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15131135284900665, "tokens_per_second": 4338397.312025177, "eta_seconds": 1873.9834734511328, "mean_data_wait_seconds": 8.639693260192871e-05}
{"step": 31210, "total_steps": 46238, "tokens": 16874913792, "loss": 3.0193375710285073, "ce_loss": 3.0193375710285073, "z_loss": 29.459933801130816, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1582096964120865, "tokens_per_second": 4338446.373380628, "eta_seconds": 1872.7160473506194, "mean_data_wait_seconds": 0.0001513063907623291}
{"step": 31220, "total_steps": 46238, "tokens": 16880320512, "loss": 2.9986849427223206, "ce_loss": 2.9986849427223206, "z_loss": 29.48050090038415, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13549557328224182, "tokens_per_second": 4338243.337369983, "eta_seconds": 1871.5574006787335, "mean_data_wait_seconds": 9.495019912719727e-05}
{"step": 31230, "total_steps": 46238, "tokens": 16885727232, "loss": 3.0327325011744644, "ce_loss": 3.0327325011744644, "z_loss": 28.921110384392016, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16855190694332123, "tokens_per_second": 4338231.838395764, "eta_seconds": 1870.3160656809039, "mean_data_wait_seconds": 0.0001335442066192627}
{"step": 31240, "total_steps": 46238, "tokens": 16891133952, "loss": 2.950931319684693, "ce_loss": 2.950931319684693, "z_loss": 29.704408501133774, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14803160727024078, "tokens_per_second": 4338351.979066954, "eta_seconds": 1869.0180103237913, "mean_data_wait_seconds": 0.0001144707202911377}
{"step": 31250, "total_steps": 46238, "tokens": 16896540672, "loss": 2.9912010666095847, "ce_loss": 2.9912010666095847, "z_loss": 29.975047010363955, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14335276186466217, "tokens_per_second": 4338355.255924216, "eta_seconds": 1867.7703382946627, "mean_data_wait_seconds": 9.304285049438477e-05}
{"step": 31260, "total_steps": 46238, "tokens": 16901947392, "loss": 2.9989092729308386, "ce_loss": 2.9989092729308386, "z_loss": 28.921473387515906, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16172686219215393, "tokens_per_second": 4338503.278174444, "eta_seconds": 1866.4603953940834, "mean_data_wait_seconds": 9.945034980773926e-05}
{"step": 31270, "total_steps": 46238, "tokens": 16907354112, "loss": 3.0465496453371914, "ce_loss": 3.0465496453371914, "z_loss": 28.768001773140647, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1520698219537735, "tokens_per_second": 4338613.12859351, "eta_seconds": 1865.1669517773619, "mean_data_wait_seconds": 8.636713027954102e-05}
{"step": 31280, "total_steps": 46238, "tokens": 16912760832, "loss": 3.0077369773026668, "ce_loss": 3.0077369773026668, "z_loss": 29.206344084306195, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1481248289346695, "tokens_per_second": 4337967.956555634, "eta_seconds": 1864.1979804804691, "mean_data_wait_seconds": 0.08950993418693542}
{"step": 31290, "total_steps": 46238, "tokens": 16918167552, "loss": 3.054652595158779, "ce_loss": 3.054652595158779, "z_loss": 29.929202715555828, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1478918492794037, "tokens_per_second": 4338003.018412398, "eta_seconds": 1862.936551611161, "mean_data_wait_seconds": 0.00011566281318664551}
{"step": 31300, "total_steps": 46238, "tokens": 16923574272, "loss": 3.0715916427699, "ce_loss": 3.0715916427699, "z_loss": 28.47385646357681, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17892935872077942, "tokens_per_second": 4338041.644119173, "eta_seconds": 1861.673613702667, "mean_data_wait_seconds": 0.00011646747589111328}
{"step": 31310, "total_steps": 46238, "tokens": 16928980992, "loss": 3.00892199530746, "ce_loss": 3.00892199530746, "z_loss": 29.585976340553977, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15269255638122559, "tokens_per_second": 4338002.5183471395, "eta_seconds": 1860.4440430511909, "mean_data_wait_seconds": 6.80088996887207e-05}
{"step": 31320, "total_steps": 46238, "tokens": 16934387712, "loss": 3.0288672212398415, "ce_loss": 3.0288672212398415, "z_loss": 29.912813966924492, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1686432957649231, "tokens_per_second": 4337990.312700111, "eta_seconds": 1859.202912553289, "mean_data_wait_seconds": 0.00015094876289367676}
{"step": 31330, "total_steps": 46238, "tokens": 16939794432, "loss": 3.0301420309326867, "ce_loss": 3.0301420309326867, "z_loss": 29.68460967323997, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20873622596263885, "tokens_per_second": 4337575.871174691, "eta_seconds": 1858.1340692070169, "mean_data_wait_seconds": 0.00013700127601623535}
{"step": 31340, "total_steps": 46238, "tokens": 16945201152, "loss": 3.0155509168451484, "ce_loss": 3.0155509168451484, "z_loss": 30.147907517173074, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1537509262561798, "tokens_per_second": 4337027.454220159, "eta_seconds": 1857.1223883221326, "mean_data_wait_seconds": 8.913874626159668e-05}
{"step": 31350, "total_steps": 46238, "tokens": 16950607872, "loss": 3.0485784953290764, "ce_loss": 3.0485784953290764, "z_loss": 30.360007979653098, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16017290949821472, "tokens_per_second": 4336976.497738309, "eta_seconds": 1855.8975517154558, "mean_data_wait_seconds": 0.00014474987983703613}
{"step": 31360, "total_steps": 46238, "tokens": 16956014592, "loss": 3.036594067559098, "ce_loss": 3.036594067559098, "z_loss": 29.502407550811768, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14427128434181213, "tokens_per_second": 4337007.857557081, "eta_seconds": 1854.6374846852893, "mean_data_wait_seconds": 8.541345596313477e-05}
{"step": 31370, "total_steps": 46238, "tokens": 16961421312, "loss": 3.0287497133919685, "ce_loss": 3.0287497133919685, "z_loss": 29.80991604833892, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13943980634212494, "tokens_per_second": 4337055.265862584, "eta_seconds": 1853.3705777902537, "mean_data_wait_seconds": 9.763240814208984e-05}
{"step": 31380, "total_steps": 46238, "tokens": 16966828032, "loss": 3.048173675031373, "ce_loss": 3.048173675031373, "z_loss": 30.302466782656584, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18668346107006073, "tokens_per_second": 4337141.954755983, "eta_seconds": 1852.086924476038, "mean_data_wait_seconds": 0.00017395615577697754}
{"step": 31390, "total_steps": 46238, "tokens": 16972234752, "loss": 3.065606193108992, "ce_loss": 3.065606193108992, "z_loss": 29.613335797280975, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19989462196826935, "tokens_per_second": 4337154.962282433, "eta_seconds": 1850.8347646807606, "mean_data_wait_seconds": 8.100271224975586e-05}
{"step": 31400, "total_steps": 46238, "tokens": 16977641472, "loss": 2.986131081075379, "ce_loss": 2.986131081075379, "z_loss": 29.864191835576836, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18410281836986542, "tokens_per_second": 4337130.639904391, "eta_seconds": 1849.5985318479682, "mean_data_wait_seconds": 0.00020948052406311035}
{"step": 31410, "total_steps": 46238, "tokens": 16983048192, "loss": 3.022226991075458, "ce_loss": 3.022226991075458, "z_loss": 30.239678570718475, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17151467502117157, "tokens_per_second": 4337319.986024866, "eta_seconds": 1848.2712296602137, "mean_data_wait_seconds": 0.00013685226440429688}
{"step": 31420, "total_steps": 46238, "tokens": 16988454912, "loss": 3.032152889352856, "ce_loss": 3.032152889352856, "z_loss": 29.97583432631059, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16708920896053314, "tokens_per_second": 4337390.978085762, "eta_seconds": 1846.9944407768348, "mean_data_wait_seconds": 0.00016030669212341309}
{"step": 31430, "total_steps": 46238, "tokens": 16993861632, "loss": 3.0455432010419443, "ce_loss": 3.0455432010419443, "z_loss": 30.047699191353537, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14936698973178864, "tokens_per_second": 4337457.543641178, "eta_seconds": 1845.7195772985956, "mean_data_wait_seconds": 0.0005159378051757812}
{"step": 31440, "total_steps": 46238, "tokens": 16999268352, "loss": 3.0498681646404844, "ce_loss": 3.0498681646404844, "z_loss": 28.893955259612113, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17651720345020294, "tokens_per_second": 4337447.248873683, "eta_seconds": 1844.4774368327976, "mean_data_wait_seconds": 6.759166717529297e-05}
{"step": 31450, "total_steps": 46238, "tokens": 17004675072, "loss": 3.019513756939859, "ce_loss": 3.019513756939859, "z_loss": 30.31480328241984, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14769527316093445, "tokens_per_second": 4337132.053246716, "eta_seconds": 1843.364870113908, "mean_data_wait_seconds": 7.56382942199707e-05}
{"step": 31460, "total_steps": 46238, "tokens": 17010081792, "loss": 3.01772663268176, "ce_loss": 3.01772663268176, "z_loss": 29.780957467628248, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1812523454427719, "tokens_per_second": 4337329.679054403, "eta_seconds": 1842.0343241562912, "mean_data_wait_seconds": 9.259581565856934e-05}
{"step": 31470, "total_steps": 46238, "tokens": 17015488512, "loss": 3.060454245769616, "ce_loss": 3.060454245769616, "z_loss": 29.130455739570387, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.23889434337615967, "tokens_per_second": 4337263.960462569, "eta_seconds": 1840.8156609284383, "mean_data_wait_seconds": 0.00011536478996276855}
{"step": 31480, "total_steps": 46238, "tokens": 17020895232, "loss": 3.024941157210957, "ce_loss": 3.024941157210957, "z_loss": 29.459354920820758, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19425508379936218, "tokens_per_second": 4337395.009798172, "eta_seconds": 1839.5135066038788, "mean_data_wait_seconds": 0.000144273042678833}
{"step": 31490, "total_steps": 46238, "tokens": 17026301952, "loss": 2.9266654722618335, "ce_loss": 2.9266654722618335, "z_loss": 29.26898884050774, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13521285355091095, "tokens_per_second": 4337445.89201145, "eta_seconds": 1838.2454058239473, "mean_data_wait_seconds": 6.195902824401855e-05}
{"step": 31500, "total_steps": 46238, "tokens": 17031708672, "loss": 2.9685889572808235, "ce_loss": 2.9685889572808235, "z_loss": 29.50381528969967, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15869155526161194, "tokens_per_second": 4337499.882967543, "eta_seconds": 1836.9760182099867, "mean_data_wait_seconds": 9.22083854675293e-05}
{"step": 31510, "total_steps": 46238, "tokens": 17037115392, "loss": 2.960137052969499, "ce_loss": 2.960137052969499, "z_loss": 29.306865692138672, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17544393241405487, "tokens_per_second": 4337505.196411522, "eta_seconds": 1835.7272633557805, "mean_data_wait_seconds": 6.702542304992676e-05}
{"step": 31520, "total_steps": 46238, "tokens": 17042522112, "loss": 2.9749594962958135, "ce_loss": 2.9749594962958135, "z_loss": 30.35608317635276, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16145406663417816, "tokens_per_second": 4337513.321958598, "eta_seconds": 1834.4773222291792, "mean_data_wait_seconds": 0.00011521577835083008}
{"step": 31530, "total_steps": 46238, "tokens": 17047928832, "loss": 2.9499473047978952, "ce_loss": 2.9499473047978952, "z_loss": 29.60383367538452, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18662996590137482, "tokens_per_second": 4337713.043632656, "eta_seconds": 1833.1464124101692, "mean_data_wait_seconds": 6.893277168273926e-05}
{"step": 31540, "total_steps": 46238, "tokens": 17053335552, "loss": 2.960045939142054, "ce_loss": 2.960045939142054, "z_loss": 31.046838269089207, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1664460450410843, "tokens_per_second": 4337827.231808718, "eta_seconds": 1831.8517449775647, "mean_data_wait_seconds": 8.973479270935059e-05}
{"step": 31550, "total_steps": 46238, "tokens": 17058742272, "loss": 2.9884974179845867, "ce_loss": 2.9884974179845867, "z_loss": 30.288767236651797, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13782544434070587, "tokens_per_second": 4337897.02125002, "eta_seconds": 1830.5758816081218, "mean_data_wait_seconds": 8.96155834197998e-05}
{"step": 31560, "total_steps": 46238, "tokens": 17064148992, "loss": 2.976720665440415, "ce_loss": 2.976720665440415, "z_loss": 31.006486054622766, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1690482497215271, "tokens_per_second": 4338089.587798634, "eta_seconds": 1829.2482862316465, "mean_data_wait_seconds": 7.56680965423584e-05}
{"step": 31570, "total_steps": 46238, "tokens": 17069555712, "loss": 3.0264711380004883, "ce_loss": 3.0264711380004883, "z_loss": 30.133723721359715, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17528420686721802, "tokens_per_second": 4338167.041733605, "eta_seconds": 1827.969312318371, "mean_data_wait_seconds": 7.584691047668457e-05}
{"step": 31580, "total_steps": 46238, "tokens": 17074962432, "loss": 3.015677258823857, "ce_loss": 3.015677258823857, "z_loss": 29.866779269594133, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15891607105731964, "tokens_per_second": 4338072.0979671925, "eta_seconds": 1826.7629778936725, "mean_data_wait_seconds": 9.167194366455078e-05}
{"step": 31590, "total_steps": 46238, "tokens": 17080369152, "loss": 2.979077983986248, "ce_loss": 2.979077983986248, "z_loss": 29.86967978332982, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1679127961397171, "tokens_per_second": 4338114.609810938, "eta_seconds": 1825.4987468726956, "mean_data_wait_seconds": 0.00010687112808227539}
{"step": 31600, "total_steps": 46238, "tokens": 17085775872, "loss": 2.9531080180948432, "ce_loss": 2.9531080180948432, "z_loss": 29.21823648972945, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17557039856910706, "tokens_per_second": 4338112.929017781, "eta_seconds": 1824.253124224642, "mean_data_wait_seconds": 8.0108642578125e-05}
{"step": 31610, "total_steps": 46238, "tokens": 17091182592, "loss": 2.9849895293062385, "ce_loss": 2.9849895293062385, "z_loss": 30.033967567212652, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14998352527618408, "tokens_per_second": 4338253.45937497, "eta_seconds": 1822.9477410799777, "mean_data_wait_seconds": 0.00014075636863708496}
{"step": 31620, "total_steps": 46238, "tokens": 17096589312, "loss": 2.995802949775349, "ce_loss": 2.995802949775349, "z_loss": 29.633484609199293, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16557516157627106, "tokens_per_second": 4338213.48058325, "eta_seconds": 1821.7182394024287, "mean_data_wait_seconds": 0.0001157522201538086}
{"step": 31630, "total_steps": 46238, "tokens": 17101996032, "loss": 2.9951509255351443, "ce_loss": 2.9951509255351443, "z_loss": 30.936363249114066, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16434043645858765, "tokens_per_second": 4338254.308574916, "eta_seconds": 1820.4548056091946, "mean_data_wait_seconds": 8.64565372467041e-05}
{"step": 31640, "total_steps": 46238, "tokens": 17107402752, "loss": 3.0104977000843394, "ce_loss": 3.0104977000843394, "z_loss": 30.534362691821475, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17024070024490356, "tokens_per_second": 4338170.3197235875, "eta_seconds": 1819.243736954722, "mean_data_wait_seconds": 0.00015342235565185547}
{"step": 31650, "total_steps": 46238, "tokens": 17112809472, "loss": 3.022761518304998, "ce_loss": 3.022761518304998, "z_loss": 29.778625444932416, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15358664095401764, "tokens_per_second": 4338151.783727654, "eta_seconds": 1818.0051914235019, "mean_data_wait_seconds": 7.429718971252441e-05}
{"step": 31660, "total_steps": 46238, "tokens": 17118216192, "loss": 3.01394076238979, "ce_loss": 3.01394076238979, "z_loss": 30.70679116971565, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15950331091880798, "tokens_per_second": 4338019.030406361, "eta_seconds": 1816.8144696363208, "mean_data_wait_seconds": 0.00010201334953308105}
{"step": 31670, "total_steps": 46238, "tokens": 17123622912, "loss": 3.025771729873888, "ce_loss": 3.025771729873888, "z_loss": 29.1717228455977, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17124855518341064, "tokens_per_second": 4338008.302927331, "eta_seconds": 1815.5726024510413, "mean_data_wait_seconds": 8.526444435119629e-05}
{"step": 31680, "total_steps": 46238, "tokens": 17129029632, "loss": 2.9758081237475076, "ce_loss": 2.9758081237475076, "z_loss": 30.026138363462508, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17130105197429657, "tokens_per_second": 4337972.418501695, "eta_seconds": 1814.3412508644851, "mean_data_wait_seconds": 7.772445678710938e-05}
{"step": 31690, "total_steps": 46238, "tokens": 17134436352, "loss": 3.0479657306815637, "ce_loss": 3.0479657306815637, "z_loss": 30.23493229259144, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1467728316783905, "tokens_per_second": 4338063.739285861, "eta_seconds": 1813.0567130151883, "mean_data_wait_seconds": 7.146596908569336e-05}
{"step": 31700, "total_steps": 46238, "tokens": 17139843072, "loss": 3.039763165242744, "ce_loss": 3.039763165242744, "z_loss": 29.61359840450865, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16656962037086487, "tokens_per_second": 4338074.444499905, "eta_seconds": 1811.805897882897, "mean_data_wait_seconds": 7.56680965423584e-05}
{"step": 31710, "total_steps": 46238, "tokens": 17145249792, "loss": 2.965077768672596, "ce_loss": 2.965077768672596, "z_loss": 30.389342394742098, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16223622858524323, "tokens_per_second": 4338133.933903851, "eta_seconds": 1810.5347284499219, "mean_data_wait_seconds": 7.915496826171875e-05}
{"step": 31720, "total_steps": 46238, "tokens": 17150656512, "loss": 2.8778731443665246, "ce_loss": 2.8778731443665246, "z_loss": 29.961268309390906, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1478242427110672, "tokens_per_second": 4338170.253996697, "eta_seconds": 1809.273256799657, "mean_data_wait_seconds": 8.541345596313477e-05}
{"step": 31730, "total_steps": 46238, "tokens": 17156063232, "loss": 2.9780899972626655, "ce_loss": 2.9780899972626655, "z_loss": 29.842784809343744, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1597258746623993, "tokens_per_second": 4338287.285056881, "eta_seconds": 1807.9781694072758, "mean_data_wait_seconds": 8.055567741394043e-05}
{"step": 31740, "total_steps": 46238, "tokens": 17161469952, "loss": 2.975266032146685, "ce_loss": 2.975266032146685, "z_loss": 29.568276968869295, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14823675155639648, "tokens_per_second": 4338402.421955655, "eta_seconds": 1806.6839406904878, "mean_data_wait_seconds": 9.545683860778809e-05}
{"step": 31750, "total_steps": 46238, "tokens": 17166876672, "loss": 2.9731499697222854, "ce_loss": 2.9731499697222854, "z_loss": 29.8200627673756, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14494061470031738, "tokens_per_second": 4338503.68816665, "eta_seconds": 1805.3955527026235, "mean_data_wait_seconds": 0.00012177228927612305}
{"step": 31760, "total_steps": 46238, "tokens": 17172283392, "loss": 2.9584872090455256, "ce_loss": 2.9584872090455256, "z_loss": 29.770665082064543, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18031933903694153, "tokens_per_second": 4338595.425560089, "eta_seconds": 1804.1111872028348, "mean_data_wait_seconds": 7.006525993347168e-05}
{"step": 31770, "total_steps": 46238, "tokens": 17177690112, "loss": 3.0110208428267278, "ce_loss": 3.0110208428267278, "z_loss": 29.96113782940489, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1701003611087799, "tokens_per_second": 4338690.105343699, "eta_seconds": 1802.8256533847027, "mean_data_wait_seconds": 9.825825691223145e-05}
{"step": 31780, "total_steps": 46238, "tokens": 17183096832, "loss": 2.9654740167386606, "ce_loss": 2.9654740167386606, "z_loss": 29.91431964527477, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17330418527126312, "tokens_per_second": 4338782.292930279, "eta_seconds": 1801.5412104765878, "mean_data_wait_seconds": 8.77082347869873e-05}
{"step": 31790, "total_steps": 46238, "tokens": 17188503552, "loss": 2.98550205339085, "ce_loss": 2.98550205339085, "z_loss": 29.08242992921309, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1501777023077011, "tokens_per_second": 4338895.773105444, "eta_seconds": 1800.2479876140076, "mean_data_wait_seconds": 0.0001627504825592041}
{"step": 31800, "total_steps": 46238, "tokens": 17193910272, "loss": 2.989054085630359, "ce_loss": 2.989054085630359, "z_loss": 30.559986316796504, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17223060131072998, "tokens_per_second": 4338920.466617535, "eta_seconds": 1798.9916441324, "mean_data_wait_seconds": 0.00012382864952087402}
{"step": 31810, "total_steps": 46238, "tokens": 17199316992, "loss": 3.025199563214273, "ce_loss": 3.025199563214273, "z_loss": 30.669486912814055, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15999768674373627, "tokens_per_second": 4338943.047438406, "eta_seconds": 1797.7361902929492, "mean_data_wait_seconds": 0.00010779500007629395}
{"step": 31820, "total_steps": 46238, "tokens": 17204723712, "loss": 2.9872257456634985, "ce_loss": 2.9872257456634985, "z_loss": 29.953764901016697, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14355747401714325, "tokens_per_second": 4338883.744043192, "eta_seconds": 1796.5146530375453, "mean_data_wait_seconds": 9.652972221374512e-05}
{"step": 31830, "total_steps": 46238, "tokens": 17210130432, "loss": 3.0517856045202776, "ce_loss": 3.0517856045202776, "z_loss": 30.34000450192076, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15804626047611237, "tokens_per_second": 4338859.779269724, "eta_seconds": 1795.278460303469, "mean_data_wait_seconds": 9.250640869140625e-05}
{"step": 31840, "total_steps": 46238, "tokens": 17215537152, "loss": 2.988544904824459, "ce_loss": 2.988544904824459, "z_loss": 30.728844946080986, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17294059693813324, "tokens_per_second": 4338800.872911785, "eta_seconds": 1794.056701840777, "mean_data_wait_seconds": 0.00025343894958496094}
{"step": 31850, "total_steps": 46238, "tokens": 17220943872, "loss": 2.994067638209372, "ce_loss": 2.994067638209372, "z_loss": 29.749269153132584, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1746867597103119, "tokens_per_second": 4338815.367257318, "eta_seconds": 1792.8045804164035, "mean_data_wait_seconds": 0.00015479326248168945}
{"step": 31860, "total_steps": 46238, "tokens": 17226350592, "loss": 3.062352229248394, "ce_loss": 3.062352229248394, "z_loss": 30.164964401360713, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16521725058555603, "tokens_per_second": 4338907.641633473, "eta_seconds": 1791.520351669343, "mean_data_wait_seconds": 0.00011426210403442383}
{"step": 31870, "total_steps": 46238, "tokens": 17231757312, "loss": 2.9938127741669165, "ce_loss": 2.9938127741669165, "z_loss": 30.86418883005778, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1577746570110321, "tokens_per_second": 4338875.00609002, "eta_seconds": 1790.2877158473366, "mean_data_wait_seconds": 0.0001716315746307373}
{"step": 31880, "total_steps": 46238, "tokens": 17237164032, "loss": 3.031641853578163, "ce_loss": 3.031641853578163, "z_loss": 30.322674866878625, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17261242866516113, "tokens_per_second": 4338830.658169701, "eta_seconds": 1789.0598909141374, "mean_data_wait_seconds": 9.036064147949219e-05}
{"step": 31890, "total_steps": 46238, "tokens": 17242570752, "loss": 3.052604837851091, "ce_loss": 3.052604837851091, "z_loss": 30.091839226809416, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1560887098312378, "tokens_per_second": 4338851.747943096, "eta_seconds": 1787.805077156034, "mean_data_wait_seconds": 8.460879325866699e-05}
{"step": 31900, "total_steps": 46238, "tokens": 17247977472, "loss": 3.0325750509897866, "ce_loss": 3.0325750509897866, "z_loss": 30.49542438622677, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19670987129211426, "tokens_per_second": 4339024.746267688, "eta_seconds": 1786.4877287615677, "mean_data_wait_seconds": 0.00010696053504943848}
{"step": 31910, "total_steps": 46238, "tokens": 17253384192, "loss": 3.0182382012858535, "ce_loss": 3.0182382012858535, "z_loss": 30.285564133615203, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15152819454669952, "tokens_per_second": 4339037.158836654, "eta_seconds": 1785.2365537419935, "mean_data_wait_seconds": 7.790327072143555e-05}
{"step": 31920, "total_steps": 46238, "tokens": 17258790912, "loss": 3.0073367429502085, "ce_loss": 3.0073367429502085, "z_loss": 29.531399062185578, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1508232057094574, "tokens_per_second": 4339215.759012265, "eta_seconds": 1783.9170610317929, "mean_data_wait_seconds": 0.00014266371726989746}
{"step": 31930, "total_steps": 46238, "tokens": 17264197632, "loss": 3.020349708470431, "ce_loss": 3.020349708470431, "z_loss": 30.208605694048334, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14087721705436707, "tokens_per_second": 4339288.092602719, "eta_seconds": 1782.6413316937171, "mean_data_wait_seconds": 7.650256156921387e-05}
{"step": 31940, "total_steps": 46238, "tokens": 17269604352, "loss": 3.012275659676754, "ce_loss": 3.012275659676754, "z_loss": 29.571876670375016, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15477634966373444, "tokens_per_second": 4339338.974021291, "eta_seconds": 1781.3744513341335, "mean_data_wait_seconds": 0.00013777613639831543}
{"step": 31950, "total_steps": 46238, "tokens": 17275011072, "loss": 2.9856681010939856, "ce_loss": 2.9856681010939856, "z_loss": 30.44717080665357, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15668514370918274, "tokens_per_second": 4339479.046780831, "eta_seconds": 1780.0710133006287, "mean_data_wait_seconds": 0.00011420249938964844}
{"step": 31960, "total_steps": 46238, "tokens": 17280417792, "loss": 3.0256092313564187, "ce_loss": 3.0256092313564187, "z_loss": 29.207106937061656, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.140598326921463, "tokens_per_second": 4339626.714232069, "eta_seconds": 1778.7645464261936, "mean_data_wait_seconds": 0.00012752413749694824}
{"step": 31970, "total_steps": 46238, "tokens": 17285824512, "loss": 2.9971622680172776, "ce_loss": 2.9971622680172776, "z_loss": 30.07637692942764, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.149903804063797, "tokens_per_second": 4339574.983948785, "eta_seconds": 1777.5398403142415, "mean_data_wait_seconds": 0.0001163780689239502}
{"step": 31980, "total_steps": 46238, "tokens": 17291231232, "loss": 3.01728302601612, "ce_loss": 3.01728302601612, "z_loss": 29.57509238792188, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15647545456886292, "tokens_per_second": 4339684.31572433, "eta_seconds": 1776.249179247825, "mean_data_wait_seconds": 0.0001588761806488037}
{"step": 31990, "total_steps": 46238, "tokens": 17296637952, "loss": 2.9748875372337573, "ce_loss": 2.9748875372337573, "z_loss": 29.96748434413563, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17076271772384644, "tokens_per_second": 4339705.72163312, "eta_seconds": 1774.994545275577, "mean_data_wait_seconds": 9.196996688842773e-05}
{"step": 32000, "total_steps": 46238, "tokens": 17302044672, "loss": 3.003046812433185, "ce_loss": 3.003046812433185, "z_loss": 28.62760250496142, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14447614550590515, "tokens_per_second": 4339765.183925124, "eta_seconds": 1773.7243693535308, "mean_data_wait_seconds": 0.00010603666305541992}
{"step": 32010, "total_steps": 46238, "tokens": 17307451392, "loss": 3.0453668290918525, "ce_loss": 3.0453668290918525, "z_loss": 30.17809350562818, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16294175386428833, "tokens_per_second": 4339736.893154764, "eta_seconds": 1772.4900687258514, "mean_data_wait_seconds": 7.984042167663574e-05}
{"step": 32020, "total_steps": 46238, "tokens": 17312858112, "loss": 2.9648646170442756, "ce_loss": 2.9648646170442756, "z_loss": 31.008511861165363, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16434764862060547, "tokens_per_second": 4339815.390154796, "eta_seconds": 1771.2121675585431, "mean_data_wait_seconds": 6.92605972290039e-05}
{"step": 32030, "total_steps": 46238, "tokens": 17318264832, "loss": 2.9806612675840203, "ce_loss": 2.9806612675840203, "z_loss": 30.56319453499534, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1802733689546585, "tokens_per_second": 4339898.925497919, "eta_seconds": 1769.9322578390963, "mean_data_wait_seconds": 7.370114326477051e-05}
{"step": 32040, "total_steps": 46238, "tokens": 17323671552, "loss": 2.974721919406544, "ce_loss": 2.974721919406544, "z_loss": 31.634862278447006, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2528797686100006, "tokens_per_second": 4339936.533283777, "eta_seconds": 1768.6711142275806, "mean_data_wait_seconds": 6.639957427978516e-05}
{"step": 32050, "total_steps": 46238, "tokens": 17329078272, "loss": 2.992459026249972, "ce_loss": 2.992459026249972, "z_loss": 31.2813828497222, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15817786753177643, "tokens_per_second": 4339962.709268624, "eta_seconds": 1767.4146479688634, "mean_data_wait_seconds": 7.709860801696777e-05}
{"step": 32060, "total_steps": 46238, "tokens": 17334484992, "loss": 3.0307080980503196, "ce_loss": 3.0307080980503196, "z_loss": 31.292413046865754, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16378970444202423, "tokens_per_second": 4339957.5849461695, "eta_seconds": 1766.1709346164207, "mean_data_wait_seconds": 7.164478302001953e-05}
{"step": 32070, "total_steps": 46238, "tokens": 17339891712, "loss": 2.999602821740237, "ce_loss": 2.999602821740237, "z_loss": 30.73127102129387, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.168134406208992, "tokens_per_second": 4340071.084392021, "eta_seconds": 1764.8789789517948, "mean_data_wait_seconds": 9.420514106750488e-05}
{"step": 32080, "total_steps": 46238, "tokens": 17345298432, "loss": 2.978007506240498, "ce_loss": 2.978007506240498, "z_loss": 30.42605530131947, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1740567982196808, "tokens_per_second": 4340063.732364828, "eta_seconds": 1763.6361989157483, "mean_data_wait_seconds": 9.077787399291992e-05}
{"step": 32090, "total_steps": 46238, "tokens": 17350705152, "loss": 3.0079535282019414, "ce_loss": 3.0079535282019414, "z_loss": 30.46765730597756, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1799226552248001, "tokens_per_second": 4339940.859748233, "eta_seconds": 1762.4403260747945, "mean_data_wait_seconds": 7.87973403930664e-05}
{"step": 32100, "total_steps": 46238, "tokens": 17356111872, "loss": 2.955516901883212, "ce_loss": 2.955516901883212, "z_loss": 30.20511377218998, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1479235589504242, "tokens_per_second": 4339902.07412442, "eta_seconds": 1761.2102608425976, "mean_data_wait_seconds": 6.300210952758789e-05}
{"step": 32110, "total_steps": 46238, "tokens": 17361518592, "loss": 2.9913784500324363, "ce_loss": 2.9913784500324363, "z_loss": 30.22194025733254, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14900967478752136, "tokens_per_second": 4339906.1324921865, "eta_seconds": 1759.962798922069, "mean_data_wait_seconds": 0.00010782480239868164}
{"step": 32120, "total_steps": 46238, "tokens": 17366925312, "loss": 3.0101238705895166, "ce_loss": 3.0101238705895166, "z_loss": 30.725906603264086, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17459896206855774, "tokens_per_second": 4339979.041791628, "eta_seconds": 1758.68743846493, "mean_data_wait_seconds": 0.00010728836059570312}
{"step": 32130, "total_steps": 46238, "tokens": 17372332032, "loss": 2.944967078440117, "ce_loss": 2.944967078440117, "z_loss": 30.245144280520353, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1938566267490387, "tokens_per_second": 4340021.541843673, "eta_seconds": 1757.4244345247844, "mean_data_wait_seconds": 9.813904762268066e-05}
{"step": 32140, "total_steps": 46238, "tokens": 17377738752, "loss": 2.98615023764697, "ce_loss": 2.98615023764697, "z_loss": 30.695050008369215, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1574639230966568, "tokens_per_second": 4340099.446236195, "eta_seconds": 1756.147129441883, "mean_data_wait_seconds": 0.00011876225471496582}
{"step": 32150, "total_steps": 46238, "tokens": 17383145472, "loss": 2.94695220932816, "ce_loss": 2.94695220932816, "z_loss": 30.03233947175922, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15564881265163422, "tokens_per_second": 4340250.825881815, "eta_seconds": 1754.84016236609, "mean_data_wait_seconds": 6.818771362304688e-05}
{"step": 32160, "total_steps": 46238, "tokens": 17388552192, "loss": 2.9854556087291604, "ce_loss": 2.9854556087291604, "z_loss": 29.723784201072924, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14983046054840088, "tokens_per_second": 4340368.528307442, "eta_seconds": 1753.546892242346, "mean_data_wait_seconds": 6.747245788574219e-05}
{"step": 32170, "total_steps": 46238, "tokens": 17393958912, "loss": 2.95745554114833, "ce_loss": 2.95745554114833, "z_loss": 30.47500318469423, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1639409363269806, "tokens_per_second": 4340482.84895281, "eta_seconds": 1752.2550574839718, "mean_data_wait_seconds": 6.321072578430176e-05}
{"step": 32180, "total_steps": 46238, "tokens": 17399365632, "loss": 2.9880212075782544, "ce_loss": 2.9880212075782544, "z_loss": 30.15427975221114, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17359308898448944, "tokens_per_second": 4340710.489817377, "eta_seconds": 1750.9175794674475, "mean_data_wait_seconds": 6.145238876342773e-05}
{"step": 32190, "total_steps": 46238, "tokens": 17404772352, "loss": 2.950852300181533, "ce_loss": 2.950852300181533, "z_loss": 31.144958423845697, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15981459617614746, "tokens_per_second": 4340796.992093734, "eta_seconds": 1749.6371283506453, "mean_data_wait_seconds": 6.708502769470215e-05}
{"step": 32200, "total_steps": 46238, "tokens": 17410179072, "loss": 3.008628796447407, "ce_loss": 3.008628796447407, "z_loss": 30.918232990033697, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12679046392440796, "tokens_per_second": 4340963.626685668, "eta_seconds": 1748.324454354972, "mean_data_wait_seconds": 6.332993507385254e-05}
{"step": 32210, "total_steps": 46238, "tokens": 17415585792, "loss": 3.008186694347497, "ce_loss": 3.008186694347497, "z_loss": 31.18296290888931, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19252614676952362, "tokens_per_second": 4341180.680759999, "eta_seconds": 1746.9915909310387, "mean_data_wait_seconds": 5.558133125305176e-05}
{"step": 32220, "total_steps": 46238, "tokens": 17420992512, "loss": 2.990299125512441, "ce_loss": 2.990299125512441, "z_loss": 30.210110490972344, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15922948718070984, "tokens_per_second": 4341367.238976793, "eta_seconds": 1745.671123133592, "mean_data_wait_seconds": 0.00017246603965759277}
{"step": 32230, "total_steps": 46238, "tokens": 17426399232, "loss": 2.9889615322604324, "ce_loss": 2.9889615322604324, "z_loss": 30.958425001664594, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1567726582288742, "tokens_per_second": 4341437.254516127, "eta_seconds": 1744.3975946265443, "mean_data_wait_seconds": 0.00013524293899536133}
{"step": 32240, "total_steps": 46238, "tokens": 17431805952, "loss": 2.986244302807432, "ce_loss": 2.986244302807432, "z_loss": 32.06172227859497, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17428113520145416, "tokens_per_second": 4341356.563562022, "eta_seconds": 1743.184618263822, "mean_data_wait_seconds": 6.112456321716309e-05}
{"step": 32250, "total_steps": 46238, "tokens": 17437212672, "loss": 2.949983990553654, "ce_loss": 2.949983990553654, "z_loss": 30.83830839214903, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17299456894397736, "tokens_per_second": 4341418.355660061, "eta_seconds": 1741.9144262244752, "mean_data_wait_seconds": 9.933114051818848e-05}
{"step": 32260, "total_steps": 46238, "tokens": 17442619392, "loss": 2.9768506902636904, "ce_loss": 2.9768506902636904, "z_loss": 29.747422738508746, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1428033411502838, "tokens_per_second": 4341540.2655754415, "eta_seconds": 1740.6201674369072, "mean_data_wait_seconds": 9.28640365600586e-05}
{"step": 32270, "total_steps": 46238, "tokens": 17448026112, "loss": 2.976759278413021, "ce_loss": 2.976759278413021, "z_loss": 30.206083398876768, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1728065460920334, "tokens_per_second": 4341539.204137477, "eta_seconds": 1739.3752466414157, "mean_data_wait_seconds": 6.854534149169922e-05}
{"step": 32280, "total_steps": 46238, "tokens": 17453432832, "loss": 2.8937967744740574, "ce_loss": 2.8937967744740574, "z_loss": 29.989173036633115, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18273788690567017, "tokens_per_second": 4341549.661320027, "eta_seconds": 1738.1257137815687, "mean_data_wait_seconds": 7.954239845275879e-05}
{"step": 32290, "total_steps": 46238, "tokens": 17458839552, "loss": 2.984273495096149, "ce_loss": 2.984273495096149, "z_loss": 30.620767535585344, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13295869529247284, "tokens_per_second": 4341644.107921541, "eta_seconds": 1736.842586945699, "mean_data_wait_seconds": 8.815526962280273e-05}
{"step": 32300, "total_steps": 46238, "tokens": 17464246272, "loss": 2.9934585257010027, "ce_loss": 2.9934585257010027, "z_loss": 30.272959131183047, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1650683730840683, "tokens_per_second": 4341603.698218035, "eta_seconds": 1735.6134248487035, "mean_data_wait_seconds": 0.00012579560279846191}
{"step": 32310, "total_steps": 46238, "tokens": 17469652992, "loss": 2.9986713004834726, "ce_loss": 2.9986713004834726, "z_loss": 30.315618702859588, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16942918300628662, "tokens_per_second": 4341676.266593906, "eta_seconds": 1734.3391081314596, "mean_data_wait_seconds": 0.00010016560554504395}
{"step": 32320, "total_steps": 46238, "tokens": 17475059712, "loss": 3.002742259791403, "ce_loss": 3.002742259791403, "z_loss": 30.656753959077776, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1825912445783615, "tokens_per_second": 4341612.652712168, "eta_seconds": 1733.1191946152749, "mean_data_wait_seconds": 0.00013390183448791504}
{"step": 32330, "total_steps": 46238, "tokens": 17480466432, "loss": 2.963635791431774, "ce_loss": 2.963635791431774, "z_loss": 31.132972991827764, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16117222607135773, "tokens_per_second": 4341510.998911684, "eta_seconds": 1731.91442009127, "mean_data_wait_seconds": 7.402896881103516e-05}
{"step": 32340, "total_steps": 46238, "tokens": 17485873152, "loss": 3.034314740787853, "ce_loss": 3.034314740787853, "z_loss": 30.4115622549346, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12853659689426422, "tokens_per_second": 4341473.697641132, "eta_seconds": 1730.6839352919387, "mean_data_wait_seconds": 0.00010821223258972168}
{"step": 32350, "total_steps": 46238, "tokens": 17491279872, "loss": 2.9793357325322702, "ce_loss": 2.9793357325322702, "z_loss": 29.945758371642142, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1613156944513321, "tokens_per_second": 4341535.915093963, "eta_seconds": 1729.4137860051537, "mean_data_wait_seconds": 7.677078247070312e-05}
{"step": 32360, "total_steps": 46238, "tokens": 17496686592, "loss": 3.020281220927383, "ce_loss": 3.020281220927383, "z_loss": 30.659323677872166, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18779006600379944, "tokens_per_second": 4341519.9445661465, "eta_seconds": 1728.1747958778005, "mean_data_wait_seconds": 0.00393986701965332}
{"step": 32370, "total_steps": 46238, "tokens": 17502093312, "loss": 3.0325119531515874, "ce_loss": 3.0325119531515874, "z_loss": 30.342236027573094, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15650513768196106, "tokens_per_second": 4341548.009627652, "eta_seconds": 1726.9182806164602, "mean_data_wait_seconds": 0.00012940168380737305}
{"step": 32380, "total_steps": 46238, "tokens": 17507500032, "loss": 3.018401366291624, "ce_loss": 3.018401366291624, "z_loss": 30.374145767905496, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1624227911233902, "tokens_per_second": 4341644.665398078, "eta_seconds": 1725.634519036132, "mean_data_wait_seconds": 9.274482727050781e-05}
{"step": 32390, "total_steps": 46238, "tokens": 17512906752, "loss": 3.004744117910212, "ce_loss": 3.004744117910212, "z_loss": 31.1443402261445, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16524738073349, "tokens_per_second": 4341720.081484882, "eta_seconds": 1724.3592501337696, "mean_data_wait_seconds": 0.00014534592628479004}
{"step": 32400, "total_steps": 46238, "tokens": 17518313472, "loss": 3.0431328415870667, "ce_loss": 3.0431328415870667, "z_loss": 30.36629545327389, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13250912725925446, "tokens_per_second": 4341774.070099493, "eta_seconds": 1723.0925292776842, "mean_data_wait_seconds": 7.551908493041992e-05}
{"step": 32410, "total_steps": 46238, "tokens": 17523720192, "loss": 2.961918182445295, "ce_loss": 2.961918182445295, "z_loss": 30.005741971911807, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15988808870315552, "tokens_per_second": 4341853.914776026, "eta_seconds": 1721.8155863232541, "mean_data_wait_seconds": 7.581710815429688e-05}
{"step": 32420, "total_steps": 46238, "tokens": 17529126912, "loss": 3.0103510798829975, "ce_loss": 3.0103510798829975, "z_loss": 30.876147067908086, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17117814719676971, "tokens_per_second": 4341993.9270017985, "eta_seconds": 1720.5148486143669, "mean_data_wait_seconds": 7.352232933044434e-05}
{"step": 32430, "total_steps": 46238, "tokens": 17534533632, "loss": 3.0390693418907397, "ce_loss": 3.0390693418907397, "z_loss": 30.433543248610064, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16137133538722992, "tokens_per_second": 4342085.228166697, "eta_seconds": 1719.2334815481906, "mean_data_wait_seconds": 8.088350296020508e-05}
{"step": 32440, "total_steps": 46238, "tokens": 17539940352, "loss": 3.059097019108859, "ce_loss": 3.059097019108859, "z_loss": 31.257339968825832, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15610697865486145, "tokens_per_second": 4342178.6970974635, "eta_seconds": 1717.951310706401, "mean_data_wait_seconds": 7.063150405883789e-05}
{"step": 32450, "total_steps": 46238, "tokens": 17545347072, "loss": 2.9966860124559114, "ce_loss": 2.9966860124559114, "z_loss": 31.244142749092795, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14982473850250244, "tokens_per_second": 4342326.024003479, "eta_seconds": 1716.6479031732022, "mean_data_wait_seconds": 5.793571472167969e-05}
{"step": 32460, "total_steps": 46238, "tokens": 17550753792, "loss": 3.0279985250848713, "ce_loss": 3.0279985250848713, "z_loss": 30.378908027302135, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15870624780654907, "tokens_per_second": 4342322.062877641, "eta_seconds": 1715.404347291477, "mean_data_wait_seconds": 7.179379463195801e-05}
{"step": 32470, "total_steps": 46238, "tokens": 17556160512, "loss": 3.013517009489464, "ce_loss": 3.013517009489464, "z_loss": 30.952172366055574, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18167394399642944, "tokens_per_second": 4342500.639846244, "eta_seconds": 1714.0887339658634, "mean_data_wait_seconds": 7.30752944946289e-05}
{"step": 32480, "total_steps": 46238, "tokens": 17561567232, "loss": 2.940927969686913, "ce_loss": 2.940927969686913, "z_loss": 30.341699166731402, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19071677327156067, "tokens_per_second": 4342378.239045555, "eta_seconds": 1712.8919441238868, "mean_data_wait_seconds": 0.00010696053504943848}
{"step": 32490, "total_steps": 46238, "tokens": 17566973952, "loss": 2.974347094694773, "ce_loss": 2.974347094694773, "z_loss": 30.308227264519893, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15463808178901672, "tokens_per_second": 4342352.234216046, "eta_seconds": 1711.657088854714, "mean_data_wait_seconds": 0.00010716915130615234}
{"step": 32500, "total_steps": 46238, "tokens": 17572380672, "loss": 2.977283568093271, "ce_loss": 2.977283568093271, "z_loss": 31.526719310066916, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1441451907157898, "tokens_per_second": 4342436.981446822, "eta_seconds": 1710.3785951835243, "mean_data_wait_seconds": 8.589029312133789e-05}
{"step": 32510, "total_steps": 46238, "tokens": 17577787392, "loss": 2.9506603786439607, "ce_loss": 2.9506603786439607, "z_loss": 31.22334484620528, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1604994833469391, "tokens_per_second": 4342433.016890085, "eta_seconds": 1709.1350667085856, "mean_data_wait_seconds": 8.016824722290039e-05}
{"step": 32520, "total_steps": 46238, "tokens": 17583194112, "loss": 2.9759236267118743, "ce_loss": 2.9759236267118743, "z_loss": 31.225982348124187, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14006386697292328, "tokens_per_second": 4342344.98796539, "eta_seconds": 1707.9245993936931, "mean_data_wait_seconds": 8.532404899597168e-05}
{"step": 32530, "total_steps": 46238, "tokens": 17588600832, "loss": 2.9780314058968513, "ce_loss": 2.9780314058968513, "z_loss": 31.11198329925537, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14421361684799194, "tokens_per_second": 4342521.151695878, "eta_seconds": 1706.6102490038068, "mean_data_wait_seconds": 7.659196853637695e-05}
{"step": 32540, "total_steps": 46238, "tokens": 17594007552, "loss": 3.017596190625971, "ce_loss": 3.017596190625971, "z_loss": 30.603746732076008, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19167912006378174, "tokens_per_second": 4342574.941284057, "eta_seconds": 1705.3440606393408, "mean_data_wait_seconds": 7.233023643493652e-05}
{"step": 32550, "total_steps": 46238, "tokens": 17599414272, "loss": 2.94568388931679, "ce_loss": 2.94568388931679, "z_loss": 30.372077609553482, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15977367758750916, "tokens_per_second": 4342709.82475643, "eta_seconds": 1704.0460824285108, "mean_data_wait_seconds": 7.399916648864746e-05}
{"step": 32560, "total_steps": 46238, "tokens": 17604820992, "loss": 2.9749307469888167, "ce_loss": 2.9749307469888167, "z_loss": 31.367367484352805, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14011867344379425, "tokens_per_second": 4342965.426852761, "eta_seconds": 1702.7008546459479, "mean_data_wait_seconds": 7.015466690063477e-05}
{"step": 32570, "total_steps": 46238, "tokens": 17610227712, "loss": 3.0059891180558638, "ce_loss": 3.0059891180558638, "z_loss": 31.752127329508465, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1624024659395218, "tokens_per_second": 4343123.701861749, "eta_seconds": 1701.3939116752376, "mean_data_wait_seconds": 6.815791130065918e-05}
{"step": 32580, "total_steps": 46238, "tokens": 17615634432, "loss": 2.979980521129839, "ce_loss": 2.979980521129839, "z_loss": 31.388602459069453, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18081887066364288, "tokens_per_second": 4343183.062239638, "eta_seconds": 1700.1257829073256, "mean_data_wait_seconds": 0.00011673569679260254}
{"step": 32590, "total_steps": 46238, "tokens": 17621041152, "loss": 2.9631179950454016, "ce_loss": 2.9631179950454016, "z_loss": 30.913213816556063, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16703295707702637, "tokens_per_second": 4343226.116519229, "eta_seconds": 1698.86406695154, "mean_data_wait_seconds": 8.919835090637207e-05}
{"step": 32600, "total_steps": 46238, "tokens": 17626447872, "loss": 2.955165397037159, "ce_loss": 2.955165397037159, "z_loss": 31.930378061352354, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16482014954090118, "tokens_per_second": 4343213.19343516, "eta_seconds": 1697.6242555038816, "mean_data_wait_seconds": 0.00022420287132263184}
{"step": 32610, "total_steps": 46238, "tokens": 17631854592, "loss": 2.9112403645659937, "ce_loss": 2.9112403645659937, "z_loss": 31.09755229949951, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1815948486328125, "tokens_per_second": 4343344.793113973, "eta_seconds": 1696.3279902808915, "mean_data_wait_seconds": 7.212162017822266e-05}
{"step": 32620, "total_steps": 46238, "tokens": 17637261312, "loss": 2.9526909427209334, "ce_loss": 2.9526909427209334, "z_loss": 31.243277925433535, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15336467325687408, "tokens_per_second": 4343497.792832418, "eta_seconds": 1695.0234523312567, "mean_data_wait_seconds": 5.793571472167969e-05}
{"step": 32630, "total_steps": 46238, "tokens": 17642668032, "loss": 2.9405639550902625, "ce_loss": 2.9405639550902625, "z_loss": 31.25059785264911, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18947702646255493, "tokens_per_second": 4343513.60153365, "eta_seconds": 1693.7725028424788, "mean_data_wait_seconds": 6.809830665588379e-05}
{"step": 32640, "total_steps": 46238, "tokens": 17648074752, "loss": 2.897468225522475, "ce_loss": 2.897468225522475, "z_loss": 31.43121931769631, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1570623219013214, "tokens_per_second": 4343591.173706487, "eta_seconds": 1692.4974957361792, "mean_data_wait_seconds": 0.00010010600090026855}
{"step": 32650, "total_steps": 46238, "tokens": 17653481472, "loss": 3.0077840776154487, "ce_loss": 3.0077840776154487, "z_loss": 31.729174469456527, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1595342755317688, "tokens_per_second": 4343608.348533373, "eta_seconds": 1691.2460504134601, "mean_data_wait_seconds": 0.00012728571891784668}
{"step": 32660, "total_steps": 46238, "tokens": 17658888192, "loss": 2.9684140050049983, "ce_loss": 2.9684140050049983, "z_loss": 31.70856020667336, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14743180572986603, "tokens_per_second": 4343736.638065251, "eta_seconds": 1689.951384177295, "mean_data_wait_seconds": 0.0001359879970550537}
{"step": 32670, "total_steps": 46238, "tokens": 17664294912, "loss": 2.9729698499043784, "ce_loss": 2.9729698499043784, "z_loss": 31.414431803154223, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18837028741836548, "tokens_per_second": 4343819.447502115, "eta_seconds": 1688.6744747685393, "mean_data_wait_seconds": 0.00014603137969970703}
{"step": 32680, "total_steps": 46238, "tokens": 17669701632, "loss": 3.0385880596709973, "ce_loss": 3.0385880596709973, "z_loss": 30.76309243115512, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17074140906333923, "tokens_per_second": 4343903.331471036, "eta_seconds": 1687.397196640142, "mean_data_wait_seconds": 0.0001259148120880127}
{"step": 32690, "total_steps": 46238, "tokens": 17675108352, "loss": 2.9577321536613232, "ce_loss": 2.9577321536613232, "z_loss": 31.633101896806195, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14447543025016785, "tokens_per_second": 4344063.191981436, "eta_seconds": 1686.0904780390913, "mean_data_wait_seconds": 7.110834121704102e-05}
{"step": 32700, "total_steps": 46238, "tokens": 17680515072, "loss": 2.9569074963078354, "ce_loss": 2.9569074963078354, "z_loss": 30.92630617546313, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16992813348770142, "tokens_per_second": 4344309.132637994, "eta_seconds": 1684.750472523496, "mean_data_wait_seconds": 6.541609764099121e-05}
{"step": 32710, "total_steps": 46238, "tokens": 17685921792, "loss": 3.0067751841111616, "ce_loss": 3.0067751841111616, "z_loss": 31.007972486091383, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15323640406131744, "tokens_per_second": 4343575.675491858, "eta_seconds": 1683.7901973865837, "mean_data_wait_seconds": 7.042288780212402e-05}
{"step": 32720, "total_steps": 46238, "tokens": 17691328512, "loss": 2.996211871956334, "ce_loss": 2.996211871956334, "z_loss": 31.080990458979752, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16518329083919525, "tokens_per_second": 4343736.849500392, "eta_seconds": 1682.4830041996174, "mean_data_wait_seconds": 6.35981559753418e-05}
{"step": 32730, "total_steps": 46238, "tokens": 17696735232, "loss": 2.9945066661545723, "ce_loss": 2.9945066661545723, "z_loss": 30.415233901052765, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16420692205429077, "tokens_per_second": 4343822.480813724, "eta_seconds": 1681.205145066601, "mean_data_wait_seconds": 7.173418998718262e-05}
{"step": 32740, "total_steps": 46238, "tokens": 17702141952, "loss": 2.970458780274247, "ce_loss": 2.970458780274247, "z_loss": 31.852368846084133, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14260418713092804, "tokens_per_second": 4342900.06891421, "eta_seconds": 1680.31726915247, "mean_data_wait_seconds": 6.771087646484375e-05}
{"step": 32750, "total_steps": 46238, "tokens": 17707548672, "loss": 2.9857918078249153, "ce_loss": 2.9857918078249153, "z_loss": 31.115804383248992, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14411765336990356, "tokens_per_second": 4342419.244332153, "eta_seconds": 1679.258232267135, "mean_data_wait_seconds": 0.06486782431602478}
{"step": 32760, "total_steps": 46238, "tokens": 17712955392, "loss": 3.0191019665111196, "ce_loss": 3.0191019665111196, "z_loss": 31.144093816930596, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19342994689941406, "tokens_per_second": 4341787.269366321, "eta_seconds": 1678.2573838684355, "mean_data_wait_seconds": 0.00021252036094665527}
{"step": 32770, "total_steps": 46238, "tokens": 17718362112, "loss": 3.0003523537606904, "ce_loss": 3.0003523537606904, "z_loss": 30.901254986271713, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1275482177734375, "tokens_per_second": 4341740.111707162, "eta_seconds": 1677.030323479458, "mean_data_wait_seconds": 0.0001201927661895752}
{"step": 32780, "total_steps": 46238, "tokens": 17723768832, "loss": 2.9844665744087915, "ce_loss": 2.9844665744087915, "z_loss": 31.37744176749027, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14605507254600525, "tokens_per_second": 4341686.443962999, "eta_seconds": 1675.8057491961083, "mean_data_wait_seconds": 8.663535118103027e-05}
{"step": 32790, "total_steps": 46238, "tokens": 17729175552, "loss": 3.0171350872877873, "ce_loss": 3.0171350872877873, "z_loss": 31.1967316829797, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17368817329406738, "tokens_per_second": 4341726.107851538, "eta_seconds": 1674.5451471137817, "mean_data_wait_seconds": 0.00014889240264892578}
{"step": 32800, "total_steps": 46238, "tokens": 17734582272, "loss": 2.9341074119914663, "ce_loss": 2.9341074119914663, "z_loss": 32.34178684697007, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.166841521859169, "tokens_per_second": 4341764.385606071, "eta_seconds": 1673.285102269747, "mean_data_wait_seconds": 9.40561294555664e-05}
{"step": 32810, "total_steps": 46238, "tokens": 17739988992, "loss": 2.936324195428328, "ce_loss": 2.936324195428328, "z_loss": 32.28932133587924, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16525417566299438, "tokens_per_second": 4341707.576697421, "eta_seconds": 1672.061698250557, "mean_data_wait_seconds": 7.31348991394043e-05}
{"step": 32820, "total_steps": 46238, "tokens": 17745395712, "loss": 3.010105828444163, "ce_loss": 3.010105828444163, "z_loss": 31.253675374117766, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16044913232326508, "tokens_per_second": 4341711.51158454, "eta_seconds": 1670.814885937119, "mean_data_wait_seconds": 7.319450378417969e-05}
{"step": 32830, "total_steps": 46238, "tokens": 17750802432, "loss": 2.9629000172470557, "ce_loss": 2.9629000172470557, "z_loss": 31.764355370492645, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18371152877807617, "tokens_per_second": 4341642.675240747, "eta_seconds": 1669.5960598825766, "mean_data_wait_seconds": 6.008148193359375e-05}
{"step": 32840, "total_steps": 46238, "tokens": 17756209152, "loss": 3.044846717155341, "ce_loss": 3.044846717155341, "z_loss": 31.011688405817207, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18646122515201569, "tokens_per_second": 4341746.611023079, "eta_seconds": 1668.3108050594383, "mean_data_wait_seconds": 6.306171417236328e-05}
{"step": 32850, "total_steps": 46238, "tokens": 17761615872, "loss": 3.030364170218959, "ce_loss": 3.030364170218959, "z_loss": 31.08828641429092, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14643888175487518, "tokens_per_second": 4341742.951243085, "eta_seconds": 1667.0669234178624, "mean_data_wait_seconds": 8.213520050048828e-05}
{"step": 32860, "total_steps": 46238, "tokens": 17767022592, "loss": 2.9621754866657835, "ce_loss": 2.9621754866657835, "z_loss": 31.524029645052824, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.166377454996109, "tokens_per_second": 4341836.997130274, "eta_seconds": 1665.785553161104, "mean_data_wait_seconds": 8.910894393920898e-05}
{"step": 32870, "total_steps": 46238, "tokens": 17772429312, "loss": 2.9705587444883403, "ce_loss": 2.9705587444883403, "z_loss": 31.20546141537753, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16952677071094513, "tokens_per_second": 4341819.530225465, "eta_seconds": 1664.5469885812372, "mean_data_wait_seconds": 8.532404899597168e-05}
{"step": 32880, "total_steps": 46238, "tokens": 17777836032, "loss": 3.001775582631429, "ce_loss": 3.001775582631429, "z_loss": 31.62173047210231, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13583603501319885, "tokens_per_second": 4341742.213594874, "eta_seconds": 1663.3313422863337, "mean_data_wait_seconds": 8.752942085266113e-05}
{"step": 32890, "total_steps": 46238, "tokens": 17783242752, "loss": 3.0118547172257393, "ce_loss": 3.0118547172257393, "z_loss": 30.7675160928206, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19411028921604156, "tokens_per_second": 4341915.683679406, "eta_seconds": 1662.0196497884906, "mean_data_wait_seconds": 6.282329559326172e-05}
{"step": 32900, "total_steps": 46238, "tokens": 17788649472, "loss": 2.9109417976755085, "ce_loss": 2.9109417976755085, "z_loss": 31.380724184440844, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.171645849943161, "tokens_per_second": 4342070.992358612, "eta_seconds": 1660.7150082737403, "mean_data_wait_seconds": 6.338953971862793e-05}
{"step": 32910, "total_steps": 46238, "tokens": 17794056192, "loss": 2.996892546162461, "ce_loss": 2.996892546162461, "z_loss": 31.337428772088252, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1698954850435257, "tokens_per_second": 4342202.399146694, "eta_seconds": 1659.4195944012174, "mean_data_wait_seconds": 9.766221046447754e-05}
{"step": 32920, "total_steps": 46238, "tokens": 17799462912, "loss": 2.9921549450267446, "ce_loss": 2.9921549450267446, "z_loss": 32.12060884995894, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14518919587135315, "tokens_per_second": 4342160.460119917, "eta_seconds": 1658.1904538371562, "mean_data_wait_seconds": 0.00011089444160461426}
{"step": 32930, "total_steps": 46238, "tokens": 17804869632, "loss": 2.935763252503944, "ce_loss": 2.935763252503944, "z_loss": 31.263721928452, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17657774686813354, "tokens_per_second": 4342303.4502246445, "eta_seconds": 1656.8907232007907, "mean_data_wait_seconds": 0.00011470913887023926}
{"step": 32940, "total_steps": 46238, "tokens": 17810276352, "loss": 3.007880626302777, "ce_loss": 3.007880626302777, "z_loss": 32.20292857199004, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17214694619178772, "tokens_per_second": 4342285.992445735, "eta_seconds": 1655.6522524097297, "mean_data_wait_seconds": 0.00013834238052368164}
{"step": 32950, "total_steps": 46238, "tokens": 17815683072, "loss": 2.963935111508225, "ce_loss": 2.963935111508225, "z_loss": 29.906603972117107, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16639944911003113, "tokens_per_second": 4342317.860566174, "eta_seconds": 1654.3949785986706, "mean_data_wait_seconds": 0.00016134977340698242}
{"step": 32960, "total_steps": 46238, "tokens": 17821089792, "loss": 2.9833785476106587, "ce_loss": 2.9833785476106587, "z_loss": 31.286048542369496, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14563710987567902, "tokens_per_second": 4342379.955987405, "eta_seconds": 1653.1262157522774, "mean_data_wait_seconds": 7.405877113342285e-05}
{"step": 32970, "total_steps": 46238, "tokens": 17826496512, "loss": 3.0117631753285727, "ce_loss": 3.0117631753285727, "z_loss": 31.35362536979444, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17741024494171143, "tokens_per_second": 4342467.177738576, "eta_seconds": 1651.8479312342272, "mean_data_wait_seconds": 7.024407386779785e-05}
{"step": 32980, "total_steps": 46238, "tokens": 17831903232, "loss": 3.0441638538331697, "ce_loss": 3.0441638538331697, "z_loss": 31.500944339867793, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14247485995292664, "tokens_per_second": 4342559.498194405, "eta_seconds": 1650.5677600917747, "mean_data_wait_seconds": 6.520748138427734e-05}
{"step": 32990, "total_steps": 46238, "tokens": 17837309952, "loss": 3.004845886519461, "ce_loss": 3.004845886519461, "z_loss": 31.702420595920447, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15139436721801758, "tokens_per_second": 4342606.182788367, "eta_seconds": 1649.304975520744, "mean_data_wait_seconds": 7.718801498413086e-05}
{"step": 33000, "total_steps": 46238, "tokens": 17842716672, "loss": 3.003922392021526, "ce_loss": 3.003922392021526, "z_loss": 30.934451637846053, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1341206580400467, "tokens_per_second": 4342654.585904037, "eta_seconds": 1648.0415659193188, "mean_data_wait_seconds": 7.69197940826416e-05}
{"step": 33010, "total_steps": 46238, "tokens": 17848123392, "loss": 3.0022135250496143, "ce_loss": 3.0022135250496143, "z_loss": 31.124261191397004, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1307682991027832, "tokens_per_second": 4342741.643134696, "eta_seconds": 1646.7635267471026, "mean_data_wait_seconds": 0.00017446279525756836}
{"step": 33020, "total_steps": 46238, "tokens": 17853530112, "loss": 2.9952186887914483, "ce_loss": 2.9952186887914483, "z_loss": 30.62426419691606, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15106524527072906, "tokens_per_second": 4342885.655633121, "eta_seconds": 1645.463958907346, "mean_data_wait_seconds": 0.00012037158012390137}
{"step": 33030, "total_steps": 46238, "tokens": 17858936832, "loss": 2.9901637597517534, "ce_loss": 2.9901637597517534, "z_loss": 30.78477600849036, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17688900232315063, "tokens_per_second": 4342348.787882686, "eta_seconds": 1644.422282227992, "mean_data_wait_seconds": 0.00010865926742553711}
{"step": 33040, "total_steps": 46238, "tokens": 17864343552, "loss": 3.0330411900173533, "ce_loss": 3.0330411900173533, "z_loss": 31.899806022644043, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14284363389015198, "tokens_per_second": 4341577.908035282, "eta_seconds": 1643.4689265380366, "mean_data_wait_seconds": 7.355213165283203e-05}
{"step": 33050, "total_steps": 46238, "tokens": 17869750272, "loss": 3.0171171589331194, "ce_loss": 3.0171171589331194, "z_loss": 31.580358837590072, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1334076225757599, "tokens_per_second": 4341597.076668794, "eta_seconds": 1642.2163406905922, "mean_data_wait_seconds": 8.26120376586914e-05}
{"step": 33060, "total_steps": 46238, "tokens": 17875156992, "loss": 3.0133975238511055, "ce_loss": 3.0133975238511055, "z_loss": 30.361777970285125, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17686307430267334, "tokens_per_second": 4341710.898880333, "eta_seconds": 1640.9279912758109, "mean_data_wait_seconds": 9.134411811828613e-05}
{"step": 33070, "total_steps": 46238, "tokens": 17880563712, "loss": 2.988470702460318, "ce_loss": 2.988470702460318, "z_loss": 31.49495519291271, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17633262276649475, "tokens_per_second": 4341193.42791399, "eta_seconds": 1639.8781446190483, "mean_data_wait_seconds": 7.423758506774902e-05}
{"step": 33080, "total_steps": 46238, "tokens": 17885970432, "loss": 2.9826683564619585, "ce_loss": 2.9826683564619585, "z_loss": 31.20908170757872, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15682463347911835, "tokens_per_second": 4341107.224943584, "eta_seconds": 1638.665238012509, "mean_data_wait_seconds": 8.103251457214355e-05}
{"step": 33090, "total_steps": 46238, "tokens": 17891377152, "loss": 2.9518802725907527, "ce_loss": 2.9518802725907527, "z_loss": 32.168696865890965, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18014760315418243, "tokens_per_second": 4341029.062098366, "eta_seconds": 1637.4492504696643, "mean_data_wait_seconds": 9.816884994506836e-05}
{"step": 33100, "total_steps": 46238, "tokens": 17896783872, "loss": 3.0488023378632287, "ce_loss": 3.0488023378632287, "z_loss": 31.3004468571056, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1850072741508484, "tokens_per_second": 4341020.781668246, "eta_seconds": 1636.20687880476, "mean_data_wait_seconds": 0.00016957521438598633}
{"step": 33110, "total_steps": 46238, "tokens": 17902190592, "loss": 3.032220011407679, "ce_loss": 3.032220011407679, "z_loss": 29.300799918897223, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.21904447674751282, "tokens_per_second": 4340989.769565879, "eta_seconds": 1634.9730639217275, "mean_data_wait_seconds": 0.00011071562767028809}
{"step": 33120, "total_steps": 46238, "tokens": 17907597312, "loss": 2.9358957662726892, "ce_loss": 2.9358957662726892, "z_loss": 31.109569549560547, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1591947376728058, "tokens_per_second": 4340892.949511717, "eta_seconds": 1633.7639989020552, "mean_data_wait_seconds": 0.00010845065116882324}
{"step": 33130, "total_steps": 46238, "tokens": 17913004032, "loss": 2.9758917075214963, "ce_loss": 2.9758917075214963, "z_loss": 30.816847425518613, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14847563207149506, "tokens_per_second": 4340849.607164282, "eta_seconds": 1632.534767457518, "mean_data_wait_seconds": 7.683038711547852e-05}
{"step": 33140, "total_steps": 46238, "tokens": 17918410752, "loss": 2.967590705914931, "ce_loss": 2.967590705914931, "z_loss": 31.393767111229174, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1497524231672287, "tokens_per_second": 4340922.688439248, "eta_seconds": 1631.2617598232312, "mean_data_wait_seconds": 7.870793342590332e-05}
{"step": 33150, "total_steps": 46238, "tokens": 17923817472, "loss": 2.919822902390451, "ce_loss": 2.919822902390451, "z_loss": 30.549169872746322, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15495432913303375, "tokens_per_second": 4340987.919979196, "eta_seconds": 1629.991742532633, "mean_data_wait_seconds": 7.233023643493652e-05}
{"step": 33160, "total_steps": 46238, "tokens": 17929224192, "loss": 2.9413571068734834, "ce_loss": 2.9413571068734834, "z_loss": 31.362096584204473, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15499670803546906, "tokens_per_second": 4341090.725201012, "eta_seconds": 1628.7076662450106, "mean_data_wait_seconds": 9.953975677490234e-05}
{"step": 33170, "total_steps": 46238, "tokens": 17934630912, "loss": 2.988280247558247, "ce_loss": 2.988280247558247, "z_loss": 31.21257494435166, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14371851086616516, "tokens_per_second": 4341100.798653067, "eta_seconds": 1627.4584147394314, "mean_data_wait_seconds": 7.981061935424805e-05}
{"step": 33180, "total_steps": 46238, "tokens": 17940037632, "loss": 2.9594989545417554, "ce_loss": 2.9594989545417554, "z_loss": 30.472744956161037, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20836322009563446, "tokens_per_second": 4341110.411015066, "eta_seconds": 1626.2093417590108, "mean_data_wait_seconds": 0.00011712312698364258}
{"step": 33190, "total_steps": 46238, "tokens": 17945444352, "loss": 2.978663469805862, "ce_loss": 2.978663469805862, "z_loss": 31.25731199437922, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15224216878414154, "tokens_per_second": 4341184.853979938, "eta_seconds": 1624.9360073973482, "mean_data_wait_seconds": 7.233023643493652e-05}
{"step": 33200, "total_steps": 46238, "tokens": 17950851072, "loss": 2.9542914899912747, "ce_loss": 2.9542914899912747, "z_loss": 32.1129636331038, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18247443437576294, "tokens_per_second": 4341217.484525603, "eta_seconds": 1623.678355006503, "mean_data_wait_seconds": 7.256865501403809e-05}
{"step": 33210, "total_steps": 46238, "tokens": 17956257792, "loss": 2.985338922702905, "ce_loss": 2.985338922702905, "z_loss": 31.56363277724295, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15560157597064972, "tokens_per_second": 4341314.964166893, "eta_seconds": 1622.396486349299, "mean_data_wait_seconds": 9.530782699584961e-05}
{"step": 33220, "total_steps": 46238, "tokens": 17961664512, "loss": 2.9909550475351736, "ce_loss": 2.9909550475351736, "z_loss": 30.947650013547957, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15097516775131226, "tokens_per_second": 4341324.832909379, "eta_seconds": 1621.1473904576424, "mean_data_wait_seconds": 8.38935375213623e-05}
{"step": 33230, "total_steps": 46238, "tokens": 17967071232, "loss": 2.9566527514746697, "ce_loss": 2.9566527514746697, "z_loss": 31.216238310842805, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1599034070968628, "tokens_per_second": 4341343.048779563, "eta_seconds": 1619.8951856561946, "mean_data_wait_seconds": 0.00012725591659545898}
{"step": 33240, "total_steps": 46238, "tokens": 17972477952, "loss": 3.064694249268734, "ce_loss": 3.064694249268734, "z_loss": 30.1772755131577, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17658072710037231, "tokens_per_second": 4341253.8951261155, "eta_seconds": 1618.6830242500384, "mean_data_wait_seconds": 0.00026530027389526367}
{"step": 33250, "total_steps": 46238, "tokens": 17977884672, "loss": 3.030626533609448, "ce_loss": 3.030626533609448, "z_loss": 31.638831196409285, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15903586149215698, "tokens_per_second": 4341273.100793008, "eta_seconds": 1617.4304405584078, "mean_data_wait_seconds": 6.404519081115723e-05}
{"step": 33260, "total_steps": 46238, "tokens": 17983291392, "loss": 3.0061644478277727, "ce_loss": 3.0061644478277727, "z_loss": 30.71239997401382, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1589082032442093, "tokens_per_second": 4341243.000872186, "eta_seconds": 1616.1962236599923, "mean_data_wait_seconds": 7.68899917602539e-05}
{"step": 33270, "total_steps": 46238, "tokens": 17988698112, "loss": 3.001983283144055, "ce_loss": 3.001983283144055, "z_loss": 31.304016069932416, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16442914307117462, "tokens_per_second": 4341306.8395783305, "eta_seconds": 1614.927044567291, "mean_data_wait_seconds": 6.845593452453613e-05}
{"step": 33280, "total_steps": 46238, "tokens": 17994104832, "loss": 2.990672134991848, "ce_loss": 2.990672134991848, "z_loss": 30.470251242319744, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14045444130897522, "tokens_per_second": 4341320.514397216, "eta_seconds": 1613.676548591045, "mean_data_wait_seconds": 5.9485435485839844e-05}
{"step": 33290, "total_steps": 46238, "tokens": 17999511552, "loss": 3.029585347031102, "ce_loss": 3.029585347031102, "z_loss": 29.741519046552252, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14970049262046814, "tokens_per_second": 4341417.307286922, "eta_seconds": 1612.3951899879798, "mean_data_wait_seconds": 0.00010758638381958008}
{"step": 33300, "total_steps": 46238, "tokens": 18004918272, "loss": 3.0881358764388342, "ce_loss": 3.0881358764388342, "z_loss": 30.412000728375983, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1688951700925827, "tokens_per_second": 4341397.83699844, "eta_seconds": 1611.157034351863, "mean_data_wait_seconds": 0.0001240074634552002}
{"step": 33310, "total_steps": 46238, "tokens": 18010324992, "loss": 3.0403542988228076, "ce_loss": 3.0403542988228076, "z_loss": 30.66362239375259, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1978583186864853, "tokens_per_second": 4341421.083237425, "eta_seconds": 1609.903027141532, "mean_data_wait_seconds": 9.08970832824707e-05}
{"step": 33320, "total_steps": 46238, "tokens": 18015731712, "loss": 3.04794877948183, "ce_loss": 3.04794877948183, "z_loss": 31.628730167042125, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16620469093322754, "tokens_per_second": 4341558.252496178, "eta_seconds": 1608.6068222128843, "mean_data_wait_seconds": 9.906291961669922e-05}
{"step": 33330, "total_steps": 46238, "tokens": 18021138432, "loss": 2.9440992546804026, "ce_loss": 2.9440992546804026, "z_loss": 30.902972394769844, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13551875948905945, "tokens_per_second": 4341598.333877927, "eta_seconds": 1607.3466422599317, "mean_data_wait_seconds": 7.31050968170166e-05}
{"step": 33340, "total_steps": 46238, "tokens": 18026545152, "loss": 3.0299016616561194, "ce_loss": 3.0299016616561194, "z_loss": 30.524898023316354, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18111863732337952, "tokens_per_second": 4341773.3780734, "eta_seconds": 1606.0365608244138, "mean_data_wait_seconds": 0.00011742115020751953}
{"step": 33350, "total_steps": 46238, "tokens": 18031951872, "loss": 3.0400223406878384, "ce_loss": 3.0400223406878384, "z_loss": 30.284757007252086, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16821011900901794, "tokens_per_second": 4341840.008429958, "eta_seconds": 1604.7666543382265, "mean_data_wait_seconds": 0.00010964274406433105}
{"step": 33360, "total_steps": 46238, "tokens": 18037358592, "loss": 3.058913756500591, "ce_loss": 3.058913756500591, "z_loss": 31.62344116153139, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17224618792533875, "tokens_per_second": 4342008.620582145, "eta_seconds": 1603.4591251149002, "mean_data_wait_seconds": 7.390975952148438e-05}
{"step": 33370, "total_steps": 46238, "tokens": 18042765312, "loss": 3.021270762790333, "ce_loss": 3.021270762790333, "z_loss": 29.84306544968576, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18123766779899597, "tokens_per_second": 4341982.732470119, "eta_seconds": 1602.2234662463331, "mean_data_wait_seconds": 6.86943531036377e-05}
{"step": 33380, "total_steps": 46238, "tokens": 18048172032, "loss": 3.0198074940479165, "ce_loss": 3.0198074940479165, "z_loss": 31.13836668476914, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15268905460834503, "tokens_per_second": 4342137.615121804, "eta_seconds": 1600.9211407282864, "mean_data_wait_seconds": 0.00011768937110900879}
{"step": 33390, "total_steps": 46238, "tokens": 18053578752, "loss": 3.0934120217959085, "ce_loss": 3.0934120217959085, "z_loss": 31.092253641648725, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1833483725786209, "tokens_per_second": 4342259.228885896, "eta_seconds": 1599.6311638405236, "mean_data_wait_seconds": 0.00012353062629699707}
{"step": 33400, "total_steps": 46238, "tokens": 18058985472, "loss": 3.082545042037964, "ce_loss": 3.082545042037964, "z_loss": 31.991619413549248, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18491533398628235, "tokens_per_second": 4342290.871938512, "eta_seconds": 1598.3743762659392, "mean_data_wait_seconds": 9.244680404663086e-05}
{"step": 33410, "total_steps": 46238, "tokens": 18064392192, "loss": 3.042649167956728, "ce_loss": 3.042649167956728, "z_loss": 31.74535775907112, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14948436617851257, "tokens_per_second": 4342479.746244552, "eta_seconds": 1597.0597790347038, "mean_data_wait_seconds": 8.758902549743652e-05}
{"step": 33420, "total_steps": 46238, "tokens": 18069798912, "loss": 3.056444971850424, "ce_loss": 3.056444971850424, "z_loss": 31.406319025791053, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14751672744750977, "tokens_per_second": 4342501.371362802, "eta_seconds": 1595.806755456529, "mean_data_wait_seconds": 0.00015917420387268066}
{"step": 33430, "total_steps": 46238, "tokens": 18075205632, "loss": 3.0086624676531013, "ce_loss": 3.0086624676531013, "z_loss": 30.4182423100327, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18659250438213348, "tokens_per_second": 4342663.38912212, "eta_seconds": 1594.5021945161127, "mean_data_wait_seconds": 0.00016498565673828125}
{"step": 33440, "total_steps": 46238, "tokens": 18080612352, "loss": 3.0523014574339897, "ce_loss": 3.0523014574339897, "z_loss": 32.04987365549261, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1523769348859787, "tokens_per_second": 4342794.106504379, "eta_seconds": 1593.2092137725717, "mean_data_wait_seconds": 8.568167686462402e-05}
{"step": 33450, "total_steps": 46238, "tokens": 18086019072, "loss": 3.0271690198869416, "ce_loss": 3.0271690198869416, "z_loss": 31.251801505233303, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.140501007437706, "tokens_per_second": 4342979.524529224, "eta_seconds": 1591.8962603788527, "mean_data_wait_seconds": 9.781122207641602e-05}
{"step": 33460, "total_steps": 46238, "tokens": 18091425792, "loss": 3.0445899096402256, "ce_loss": 3.0445899096402256, "z_loss": 31.140289783477783, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1759169101715088, "tokens_per_second": 4342958.570183818, "eta_seconds": 1590.6590017752824, "mean_data_wait_seconds": 5.760788917541504e-05}
{"step": 33470, "total_steps": 46238, "tokens": 18096832512, "loss": 3.0013300198497195, "ce_loss": 3.0013300198497195, "z_loss": 31.752324797890402, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16994938254356384, "tokens_per_second": 4343162.17530101, "eta_seconds": 1589.3395515495788, "mean_data_wait_seconds": 0.0001232624053955078}
{"step": 33480, "total_steps": 46238, "tokens": 18102239232, "loss": 2.99681837089134, "ce_loss": 2.99681837089134, "z_loss": 31.77060079574585, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17491039633750916, "tokens_per_second": 4343302.10001722, "eta_seconds": 1588.0435081807123, "mean_data_wait_seconds": 9.644031524658203e-05}
{"step": 33490, "total_steps": 46238, "tokens": 18107645952, "loss": 2.954763145157785, "ce_loss": 2.954763145157785, "z_loss": 30.069747823657412, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1541275680065155, "tokens_per_second": 4343422.684356741, "eta_seconds": 1586.7546137800528, "mean_data_wait_seconds": 6.240606307983398e-05}
{"step": 33500, "total_steps": 46238, "tokens": 18113052672, "loss": 3.0333412741169785, "ce_loss": 3.0333412741169785, "z_loss": 31.75001270120794, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1712867170572281, "tokens_per_second": 4343427.152174096, "eta_seconds": 1585.5081765450939, "mean_data_wait_seconds": 0.00010460615158081055}
{"step": 33510, "total_steps": 46238, "tokens": 18118459392, "loss": 2.9378144470128147, "ce_loss": 2.9378144470128147, "z_loss": 31.737408883643873, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16351209580898285, "tokens_per_second": 4343502.65251922, "eta_seconds": 1584.235833264419, "mean_data_wait_seconds": 0.00012406706809997559}
{"step": 33520, "total_steps": 46238, "tokens": 18123866112, "loss": 2.9375074289061804, "ce_loss": 2.9375074289061804, "z_loss": 31.8429499539462, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14302244782447815, "tokens_per_second": 4343583.310774445, "eta_seconds": 1582.9616544810979, "mean_data_wait_seconds": 8.213520050048828e-05}
{"step": 33530, "total_steps": 46238, "tokens": 18129272832, "loss": 2.9307239561369927, "ce_loss": 2.9307239561369927, "z_loss": 31.616863135135535, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13701775670051575, "tokens_per_second": 4343577.302895186, "eta_seconds": 1581.7190819697464, "mean_data_wait_seconds": 9.372830390930176e-05}
{"step": 33540, "total_steps": 46238, "tokens": 18134679552, "loss": 2.9756924618374216, "ce_loss": 2.9756924618374216, "z_loss": 31.33916117928245, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1552068591117859, "tokens_per_second": 4343631.257242439, "eta_seconds": 1580.4546881261278, "mean_data_wait_seconds": 0.00011184811592102051}
{"step": 33550, "total_steps": 46238, "tokens": 18140086272, "loss": 2.962952920884797, "ce_loss": 2.962952920884797, "z_loss": 31.952636111866344, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1629582643508911, "tokens_per_second": 4343642.817936657, "eta_seconds": 1579.20573848161, "mean_data_wait_seconds": 8.07046890258789e-05}
{"step": 33560, "total_steps": 46238, "tokens": 18145492992, "loss": 3.0104314323627586, "ce_loss": 3.0104314323627586, "z_loss": 30.995696125608504, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17245464026927948, "tokens_per_second": 4343670.432260547, "eta_seconds": 1577.9509635663053, "mean_data_wait_seconds": 9.492039680480957e-05}
{"step": 33570, "total_steps": 46238, "tokens": 18150899712, "loss": 2.973693437648542, "ce_loss": 2.973693437648542, "z_loss": 30.883682626666445, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1617470532655716, "tokens_per_second": 4343717.035482835, "eta_seconds": 1576.6893119543913, "mean_data_wait_seconds": 7.089972496032715e-05}
{"step": 33580, "total_steps": 46238, "tokens": 18156306432, "loss": 2.9965348641077676, "ce_loss": 2.9965348641077676, "z_loss": 32.3633198593602, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2379591017961502, "tokens_per_second": 4343669.515720717, "eta_seconds": 1575.4618253604726, "mean_data_wait_seconds": 0.0001226663589477539}
{"step": 33590, "total_steps": 46238, "tokens": 18161713152, "loss": 2.9947658867547005, "ce_loss": 2.9947658867547005, "z_loss": 31.630522135532264, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1712244749069214, "tokens_per_second": 4343628.125854576, "eta_seconds": 1574.2320902885072, "mean_data_wait_seconds": 7.674098014831543e-05}
{"step": 33600, "total_steps": 46238, "tokens": 18167119872, "loss": 2.9707572839476843, "ce_loss": 2.9707572839476843, "z_loss": 32.179316997528076, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17395447194576263, "tokens_per_second": 4343649.733002277, "eta_seconds": 1572.9795181430247, "mean_data_wait_seconds": 7.486343383789062e-05}
{"step": 33610, "total_steps": 46238, "tokens": 18172526592, "loss": 2.9483310634439643, "ce_loss": 2.9483310634439643, "z_loss": 31.78323088270245, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14220917224884033, "tokens_per_second": 4343571.2984582465, "eta_seconds": 1571.763158676196, "mean_data_wait_seconds": 8.565187454223633e-05}
{"step": 33620, "total_steps": 46238, "tokens": 18177933312, "loss": 2.9878022092761416, "ce_loss": 2.9878022092761416, "z_loss": 31.44425665248524, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16617877781391144, "tokens_per_second": 4343662.151429699, "eta_seconds": 1570.485545648314, "mean_data_wait_seconds": 0.00013658404350280762}
{"step": 33630, "total_steps": 46238, "tokens": 18183340032, "loss": 2.9095653855439387, "ce_loss": 2.9095653855439387, "z_loss": 31.370625712654807, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15624935925006866, "tokens_per_second": 4343595.3998385975, "eta_seconds": 1569.2649237664455, "mean_data_wait_seconds": 9.894371032714844e-05}
{"step": 33640, "total_steps": 46238, "tokens": 18188746752, "loss": 2.9234361522125476, "ce_loss": 2.9234361522125476, "z_loss": 31.082095030582312, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1594625562429428, "tokens_per_second": 4343636.922183655, "eta_seconds": 1568.0051776924342, "mean_data_wait_seconds": 0.00018292665481567383}
{"step": 33650, "total_steps": 46238, "tokens": 18194153472, "loss": 2.932613618446119, "ce_loss": 2.932613618446119, "z_loss": 32.53741689161821, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16022220253944397, "tokens_per_second": 4343606.279867636, "eta_seconds": 1566.7714856069745, "mean_data_wait_seconds": 0.00013259053230285645}
{"step": 33660, "total_steps": 46238, "tokens": 18199560192, "loss": 2.926799506852121, "ce_loss": 2.926799506852121, "z_loss": 31.832558516300086, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1480671614408493, "tokens_per_second": 4343665.483772852, "eta_seconds": 1565.505393866928, "mean_data_wait_seconds": 0.0001805424690246582}
{"step": 33670, "total_steps": 46238, "tokens": 18204966912, "loss": 2.928456703821818, "ce_loss": 2.928456703821818, "z_loss": 32.0533496538798, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1665390580892563, "tokens_per_second": 4343726.56742955, "eta_seconds": 1564.2386597139784, "mean_data_wait_seconds": 0.00012546777725219727}
{"step": 33680, "total_steps": 46238, "tokens": 18210373632, "loss": 2.9682016408804692, "ce_loss": 2.9682016408804692, "z_loss": 31.84630398316817, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14422498643398285, "tokens_per_second": 4343706.449773302, "eta_seconds": 1563.0011794085046, "mean_data_wait_seconds": 0.00016441941261291504}
{"step": 33690, "total_steps": 46238, "tokens": 18215780352, "loss": 2.939255091277036, "ce_loss": 2.939255091277036, "z_loss": 31.408476439389315, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15085142850875854, "tokens_per_second": 4343770.109116835, "eta_seconds": 1561.7335663694387, "mean_data_wait_seconds": 0.00010892748832702637}
{"step": 33700, "total_steps": 46238, "tokens": 18221187072, "loss": 2.9894382718837624, "ce_loss": 2.9894382718837624, "z_loss": 32.83920155149518, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1791415959596634, "tokens_per_second": 4343659.309228167, "eta_seconds": 1560.5286652199404, "mean_data_wait_seconds": 0.00026041269302368164}
{"step": 33710, "total_steps": 46238, "tokens": 18226593792, "loss": 2.984510351311077, "ce_loss": 2.984510351311077, "z_loss": 31.75481742801088, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1576698124408722, "tokens_per_second": 4343519.608010141, "eta_seconds": 1559.3340781769498, "mean_data_wait_seconds": 6.633996963500977e-05}
{"step": 33720, "total_steps": 46238, "tokens": 18232000512, "loss": 2.94253599282467, "ce_loss": 2.94253599282467, "z_loss": 32.01544157663981, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16491951048374176, "tokens_per_second": 4343585.131769333, "eta_seconds": 1558.0657955800818, "mean_data_wait_seconds": 0.00010073184967041016}
{"step": 33730, "total_steps": 46238, "tokens": 18237407232, "loss": 2.986321843031681, "ce_loss": 2.986321843031681, "z_loss": 32.45019927169337, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15165849030017853, "tokens_per_second": 4343588.252791109, "eta_seconds": 1556.8199171859226, "mean_data_wait_seconds": 5.906820297241211e-05}
{"step": 33740, "total_steps": 46238, "tokens": 18242813952, "loss": 2.9870892853447883, "ce_loss": 2.9870892853447883, "z_loss": 32.2606470079133, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18574199080467224, "tokens_per_second": 4343669.283649152, "eta_seconds": 1555.546139167293, "mean_data_wait_seconds": 8.392333984375e-05}
{"step": 33750, "total_steps": 46238, "tokens": 18248220672, "loss": 2.9392286557139773, "ce_loss": 2.9392286557139773, "z_loss": 30.73748355923277, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16089311242103577, "tokens_per_second": 4343753.229705488, "eta_seconds": 1554.2713655623, "mean_data_wait_seconds": 0.00011566281318664551}
{"step": 33760, "total_steps": 46238, "tokens": 18253627392, "loss": 3.0047265164779895, "ce_loss": 3.0047265164779895, "z_loss": 31.393845081329346, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16390538215637207, "tokens_per_second": 4343745.963713249, "eta_seconds": 1553.0292517919752, "mean_data_wait_seconds": 7.140636444091797e-05}
{"step": 33770, "total_steps": 46238, "tokens": 18259034112, "loss": 3.0410190802631956, "ce_loss": 3.0410190802631956, "z_loss": 31.453648841742314, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15800173580646515, "tokens_per_second": 4343809.0670074, "eta_seconds": 1551.7619950648068, "mean_data_wait_seconds": 0.00011837482452392578}
{"step": 33780, "total_steps": 46238, "tokens": 18264440832, "loss": 3.0073536363514988, "ce_loss": 3.0073536363514988, "z_loss": 31.358717716101445, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16614149510860443, "tokens_per_second": 4343914.830885861, "eta_seconds": 1550.4795481053413, "mean_data_wait_seconds": 0.00014886260032653809}
{"step": 33790, "total_steps": 46238, "tokens": 18269847552, "loss": 2.974642079887968, "ce_loss": 2.974642079887968, "z_loss": 30.495990391933557, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15576139092445374, "tokens_per_second": 4343928.575497315, "eta_seconds": 1549.2299808887958, "mean_data_wait_seconds": 9.694695472717285e-05}
{"step": 33800, "total_steps": 46238, "tokens": 18275254272, "loss": 3.0353347677173037, "ce_loss": 3.0353347677173037, "z_loss": 31.79653463941632, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15705139935016632, "tokens_per_second": 4344050.164457755, "eta_seconds": 1547.9419917885234, "mean_data_wait_seconds": 0.0001239180564880371}
{"step": 33810, "total_steps": 46238, "tokens": 18280660992, "loss": 3.0504342949751653, "ce_loss": 3.0504342949751653, "z_loss": 31.797387932286117, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15689995884895325, "tokens_per_second": 4344105.301523447, "eta_seconds": 1546.677733995932, "mean_data_wait_seconds": 7.849931716918945e-05}
{"step": 33820, "total_steps": 46238, "tokens": 18286067712, "loss": 3.075493003382827, "ce_loss": 3.075493003382827, "z_loss": 31.717505570613977, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18885384500026703, "tokens_per_second": 4344202.894782801, "eta_seconds": 1545.3984048633297, "mean_data_wait_seconds": 0.00010859966278076172}
{"step": 33830, "total_steps": 46238, "tokens": 18291474432, "loss": 3.0703757351095025, "ce_loss": 3.0703757351095025, "z_loss": 32.09119019363866, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16098901629447937, "tokens_per_second": 4344169.071735839, "eta_seconds": 1544.1658446593508, "mean_data_wait_seconds": 6.860494613647461e-05}
{"step": 33840, "total_steps": 46238, "tokens": 18296881152, "loss": 3.0180384823770234, "ce_loss": 3.0180384823770234, "z_loss": 31.697577823292125, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13578012585639954, "tokens_per_second": 4344217.526675309, "eta_seconds": 1542.9040426365757, "mean_data_wait_seconds": 7.411837577819824e-05}
{"step": 33850, "total_steps": 46238, "tokens": 18302287872, "loss": 3.028244340058529, "ce_loss": 3.028244340058529, "z_loss": 31.336007248271596, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14266538619995117, "tokens_per_second": 4344268.620294775, "eta_seconds": 1541.6413323781906, "mean_data_wait_seconds": 6.973743438720703e-05}
{"step": 33860, "total_steps": 46238, "tokens": 18307694592, "loss": 3.0403980179266497, "ce_loss": 3.0403980179266497, "z_loss": 32.12554304527514, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18628321588039398, "tokens_per_second": 4344321.392384488, "eta_seconds": 1540.37805668125, "mean_data_wait_seconds": 6.189942359924316e-05}
{"step": 33870, "total_steps": 46238, "tokens": 18313101312, "loss": 3.0714105187040386, "ce_loss": 3.0714105187040386, "z_loss": 31.19941907940489, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15655504167079926, "tokens_per_second": 4344491.823735294, "eta_seconds": 1539.073128753436, "mean_data_wait_seconds": 6.377696990966797e-05}
{"step": 33880, "total_steps": 46238, "tokens": 18318508032, "loss": 3.0253060333656543, "ce_loss": 3.0253060333656543, "z_loss": 31.274792324412953, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13881206512451172, "tokens_per_second": 4344678.878759661, "eta_seconds": 1537.7624193729473, "mean_data_wait_seconds": 6.03795051574707e-05}
{"step": 33890, "total_steps": 46238, "tokens": 18323914752, "loss": 3.1015972874381323, "ce_loss": 3.1015972874381323, "z_loss": 32.1411888960636, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1874481737613678, "tokens_per_second": 4344724.792525827, "eta_seconds": 1536.501735503266, "mean_data_wait_seconds": 0.00010985136032104492}
{"step": 33900, "total_steps": 46238, "tokens": 18329321472, "loss": 2.9969770456805374, "ce_loss": 2.9969770456805374, "z_loss": 31.929271322308164, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18017931282520294, "tokens_per_second": 4344905.999166357, "eta_seconds": 1535.1932735207165, "mean_data_wait_seconds": 6.210803985595703e-05}
{"step": 33910, "total_steps": 46238, "tokens": 18334728192, "loss": 2.9815481196750295, "ce_loss": 2.9815481196750295, "z_loss": 32.382374532294996, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1532447785139084, "tokens_per_second": 4344989.62740189, "eta_seconds": 1533.919368177017, "mean_data_wait_seconds": 6.273388862609863e-05}
{"step": 33920, "total_steps": 46238, "tokens": 18340134912, "loss": 2.9997397137410715, "ce_loss": 2.9997397137410715, "z_loss": 31.9174361373439, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15979993343353271, "tokens_per_second": 4345034.078659238, "eta_seconds": 1532.6593309608588, "mean_data_wait_seconds": 0.00013825297355651855}
{"step": 33930, "total_steps": 46238, "tokens": 18345541632, "loss": 3.04107972166755, "ce_loss": 3.04107972166755, "z_loss": 31.58688821214618, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16508199274539948, "tokens_per_second": 4345095.758940447, "eta_seconds": 1531.3932472739316, "mean_data_wait_seconds": 8.32974910736084e-05}
{"step": 33940, "total_steps": 46238, "tokens": 18350948352, "loss": 3.0213907541650715, "ce_loss": 3.0213907541650715, "z_loss": 32.06497218392112, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1732444167137146, "tokens_per_second": 4345179.454994466, "eta_seconds": 1530.1194468177534, "mean_data_wait_seconds": 8.96751880645752e-05}
{"step": 33950, "total_steps": 46238, "tokens": 18356355072, "loss": 3.0210759802298113, "ce_loss": 3.0210759802298113, "z_loss": 31.4567258719242, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16669848561286926, "tokens_per_second": 4345192.931317699, "eta_seconds": 1528.8704020756586, "mean_data_wait_seconds": 6.663799285888672e-05}
{"step": 33960, "total_steps": 46238, "tokens": 18361761792, "loss": 2.9936749465537793, "ce_loss": 2.9936749465537793, "z_loss": 31.097523414727412, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18589900434017181, "tokens_per_second": 4345209.944534574, "eta_seconds": 1527.6201216351112, "mean_data_wait_seconds": 0.00017687678337097168}
{"step": 33970, "total_steps": 46238, "tokens": 18367168512, "loss": 3.0242671966552734, "ce_loss": 3.0242671966552734, "z_loss": 31.449775450157397, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15574657917022705, "tokens_per_second": 4345332.669580684, "eta_seconds": 1526.3327179596622, "mean_data_wait_seconds": 7.811188697814941e-05}
{"step": 33980, "total_steps": 46238, "tokens": 18372575232, "loss": 3.010102685653802, "ce_loss": 3.010102685653802, "z_loss": 31.535968173633922, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19140960276126862, "tokens_per_second": 4345312.237065162, "eta_seconds": 1525.095630061307, "mean_data_wait_seconds": 0.0001163482666015625}
{"step": 33990, "total_steps": 46238, "tokens": 18377981952, "loss": 3.029675928029147, "ce_loss": 3.029675928029147, "z_loss": 32.239252003756434, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14775869250297546, "tokens_per_second": 4345387.99788932, "eta_seconds": 1523.8247970529458, "mean_data_wait_seconds": 0.00016257166862487793}
{"step": 34000, "total_steps": 46238, "tokens": 18383388672, "loss": 2.9528751174608865, "ce_loss": 2.9528751174608865, "z_loss": 32.728099100517504, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.146729975938797, "tokens_per_second": 4345317.765148235, "eta_seconds": 1522.6051629792137, "mean_data_wait_seconds": 0.00011110305786132812}
{"step": 34010, "total_steps": 46238, "tokens": 18388795392, "loss": 3.0117998628905327, "ce_loss": 3.0117998628905327, "z_loss": 31.846594550392844, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1790631264448166, "tokens_per_second": 4345356.256319974, "eta_seconds": 1521.3474233292436, "mean_data_wait_seconds": 8.499622344970703e-05}
{"step": 34020, "total_steps": 46238, "tokens": 18394202112, "loss": 3.0080084240797795, "ce_loss": 3.0080084240797795, "z_loss": 31.911050232973967, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17068707942962646, "tokens_per_second": 4345431.449672803, "eta_seconds": 1520.07686705019, "mean_data_wait_seconds": 0.00014275312423706055}
{"step": 34030, "total_steps": 46238, "tokens": 18399608832, "loss": 2.984290066993598, "ce_loss": 2.984290066993598, "z_loss": 31.955037680539217, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17131586372852325, "tokens_per_second": 4345480.745917005, "eta_seconds": 1518.8154061438001, "mean_data_wait_seconds": 0.00016021728515625}
{"step": 34040, "total_steps": 46238, "tokens": 18405015552, "loss": 2.9712450413992912, "ce_loss": 2.9712450413992912, "z_loss": 32.0089681076281, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1674523502588272, "tokens_per_second": 4345453.691483985, "eta_seconds": 1517.580637649813, "mean_data_wait_seconds": 0.00010526180267333984}
{"step": 34050, "total_steps": 46238, "tokens": 18410422272, "loss": 2.933521512782935, "ce_loss": 2.933521512782935, "z_loss": 31.509123296448678, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13567879796028137, "tokens_per_second": 4345462.940143551, "eta_seconds": 1516.3331858451725, "mean_data_wait_seconds": 0.00011309981346130371}
{"step": 34060, "total_steps": 46238, "tokens": 18415828992, "loss": 2.949231718525742, "ce_loss": 2.949231718525742, "z_loss": 33.14926105557066, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16443471610546112, "tokens_per_second": 4345327.890961237, "eta_seconds": 1515.1360516878267, "mean_data_wait_seconds": 7.838010787963867e-05}
{"step": 34070, "total_steps": 46238, "tokens": 18421235712, "loss": 2.960991241715171, "ce_loss": 2.960991241715171, "z_loss": 32.96258724097049, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1645917445421219, "tokens_per_second": 4345261.905562414, "eta_seconds": 1513.9147805058608, "mean_data_wait_seconds": 0.0001086890697479248}
{"step": 34080, "total_steps": 46238, "tokens": 18426642432, "loss": 2.9553560867454065, "ce_loss": 2.9553560867454065, "z_loss": 32.755494088837594, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15896517038345337, "tokens_per_second": 4345243.727466815, "eta_seconds": 1512.6768292539232, "mean_data_wait_seconds": 0.00013938546180725098}
{"step": 34090, "total_steps": 46238, "tokens": 18432049152, "loss": 2.9183824188781506, "ce_loss": 2.9183824188781506, "z_loss": 31.80041771224051, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19016927480697632, "tokens_per_second": 4345176.525864296, "eta_seconds": 1511.455920123672, "mean_data_wait_seconds": 9.584426879882812e-05}
{"step": 34100, "total_steps": 46238, "tokens": 18437455872, "loss": 2.940593564149105, "ce_loss": 2.940593564149105, "z_loss": 33.24351541923754, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17274846136569977, "tokens_per_second": 4345275.196480054, "eta_seconds": 1510.177323018745, "mean_data_wait_seconds": 0.00012534856796264648}
{"step": 34110, "total_steps": 46238, "tokens": 18442862592, "loss": 2.9553768905726345, "ce_loss": 2.9553768905726345, "z_loss": 31.737036733916312, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15558432042598724, "tokens_per_second": 4345140.547392664, "eta_seconds": 1508.979806863651, "mean_data_wait_seconds": 7.49826431274414e-05}
{"step": 34120, "total_steps": 46238, "tokens": 18448269312, "loss": 2.9564180175463357, "ce_loss": 2.9564180175463357, "z_loss": 31.114693294871937, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2184598594903946, "tokens_per_second": 4345256.3873959575, "eta_seconds": 1507.6952980273052, "mean_data_wait_seconds": 9.486079216003418e-05}
{"step": 34130, "total_steps": 46238, "tokens": 18453676032, "loss": 2.976518618338036, "ce_loss": 2.976518618338036, "z_loss": 31.869170232252642, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14111651480197906, "tokens_per_second": 4345386.534491428, "eta_seconds": 1506.4058978509527, "mean_data_wait_seconds": 0.00010719895362854004}
{"step": 34140, "total_steps": 46238, "tokens": 18459082752, "loss": 2.982632293845668, "ce_loss": 2.982632293845668, "z_loss": 32.362144874803946, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16583743691444397, "tokens_per_second": 4345492.053462984, "eta_seconds": 1505.125105173711, "mean_data_wait_seconds": 0.00012114644050598145}
{"step": 34150, "total_steps": 46238, "tokens": 18464489472, "loss": 2.908105696692611, "ce_loss": 2.908105696692611, "z_loss": 32.467238411758885, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12983325123786926, "tokens_per_second": 4345562.435768297, "eta_seconds": 1503.8565342450527, "mean_data_wait_seconds": 7.349252700805664e-05}
{"step": 34160, "total_steps": 46238, "tokens": 18469896192, "loss": 2.968793300065127, "ce_loss": 2.968793300065127, "z_loss": 32.21163215059222, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16645152866840363, "tokens_per_second": 4345500.922643138, "eta_seconds": 1502.6336112312529, "mean_data_wait_seconds": 8.401274681091309e-05}
{"step": 34170, "total_steps": 46238, "tokens": 18475302912, "loss": 2.925856985829093, "ce_loss": 2.925856985829093, "z_loss": 33.06047543612394, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15695799887180328, "tokens_per_second": 4345505.531974724, "eta_seconds": 1501.387807700057, "mean_data_wait_seconds": 8.186697959899902e-05}
{"step": 34180, "total_steps": 46238, "tokens": 18480709632, "loss": 2.8846045750560183, "ce_loss": 2.8846045750560183, "z_loss": 32.455724123752475, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15678803622722626, "tokens_per_second": 4345608.278954618, "eta_seconds": 1500.1081288367268, "mean_data_wait_seconds": 0.00010800361633300781}
{"step": 34190, "total_steps": 46238, "tokens": 18486116352, "loss": 3.0148321950074397, "ce_loss": 3.0148321950074397, "z_loss": 33.13642071232651, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1825621873140335, "tokens_per_second": 4345618.34390362, "eta_seconds": 1498.860477045257, "mean_data_wait_seconds": 0.00011757016181945801}
{"step": 34200, "total_steps": 46238, "tokens": 18491523072, "loss": 2.9881847988475454, "ce_loss": 2.9881847988475454, "z_loss": 32.088220076127485, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16498808562755585, "tokens_per_second": 4345637.767961754, "eta_seconds": 1497.6096056557647, "mean_data_wait_seconds": 6.368756294250488e-05}
{"step": 34210, "total_steps": 46238, "tokens": 18496929792, "loss": 2.988548614762046, "ce_loss": 2.988548614762046, "z_loss": 32.66005533391779, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18139097094535828, "tokens_per_second": 4345808.323385582, "eta_seconds": 1496.306707548052, "mean_data_wait_seconds": 0.00011366605758666992}
{"step": 34220, "total_steps": 46238, "tokens": 18502336512, "loss": 2.933828800013571, "ce_loss": 2.933828800013571, "z_loss": 31.645970936977502, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16258762776851654, "tokens_per_second": 4345855.510532303, "eta_seconds": 1495.0463512313559, "mean_data_wait_seconds": 0.00011527538299560547}
{"step": 34230, "total_steps": 46238, "tokens": 18507743232, "loss": 2.918971462683244, "ce_loss": 2.918971462683244, "z_loss": 32.88224217385957, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14278070628643036, "tokens_per_second": 4345997.944652322, "eta_seconds": 1493.7532844414047, "mean_data_wait_seconds": 7.215142250061035e-05}
{"step": 34240, "total_steps": 46238, "tokens": 18513149952, "loss": 2.887865050272508, "ce_loss": 2.887865050272508, "z_loss": 33.17577789769028, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16421447694301605, "tokens_per_second": 4346210.565054494, "eta_seconds": 1492.4362008950827, "mean_data_wait_seconds": 0.00012305378913879395}
{"step": 34250, "total_steps": 46238, "tokens": 18518556672, "loss": 2.9567550547195203, "ce_loss": 2.9567550547195203, "z_loss": 33.00850782972394, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.149801567196846, "tokens_per_second": 4346268.740498765, "eta_seconds": 1491.172233232926, "mean_data_wait_seconds": 8.350610733032227e-05}
{"step": 34260, "total_steps": 46238, "tokens": 18523963392, "loss": 3.0330639337048386, "ce_loss": 3.0330639337048386, "z_loss": 31.165345163056344, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1511562615633011, "tokens_per_second": 4346455.7793612415, "eta_seconds": 1489.8641267096161, "mean_data_wait_seconds": 0.00015690922737121582}
{"step": 34270, "total_steps": 46238, "tokens": 18529370112, "loss": 2.9734035947106103, "ce_loss": 2.9734035947106103, "z_loss": 33.093242674162894, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1765049695968628, "tokens_per_second": 4346501.827951156, "eta_seconds": 1488.6044180153765, "mean_data_wait_seconds": 8.705258369445801e-05}
{"step": 34280, "total_steps": 46238, "tokens": 18534776832, "loss": 2.9431869080572417, "ce_loss": 2.9431869080572417, "z_loss": 32.08845495455193, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17130668461322784, "tokens_per_second": 4346555.838726182, "eta_seconds": 1487.3420114383262, "mean_data_wait_seconds": 7.76052474975586e-05}
{"step": 34290, "total_steps": 46238, "tokens": 18540183552, "loss": 3.0144755135882986, "ce_loss": 3.0144755135882986, "z_loss": 32.06260212984952, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1652076691389084, "tokens_per_second": 4346692.787435493, "eta_seconds": 1486.0512807970927, "mean_data_wait_seconds": 7.399916648864746e-05}
{"step": 34300, "total_steps": 46238, "tokens": 18545590272, "loss": 2.9722472176407324, "ce_loss": 2.9722472176407324, "z_loss": 32.19190687121767, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15166887640953064, "tokens_per_second": 4346833.206055217, "eta_seconds": 1484.7594462583609, "mean_data_wait_seconds": 6.899237632751465e-05}
{"step": 34310, "total_steps": 46238, "tokens": 18550996992, "loss": 2.999841444420092, "ce_loss": 2.999841444420092, "z_loss": 31.52032777034875, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16210323572158813, "tokens_per_second": 4346931.261551672, "eta_seconds": 1483.4821523490396, "mean_data_wait_seconds": 7.134675979614258e-05}
{"step": 34320, "total_steps": 46238, "tokens": 18556403712, "loss": 3.0110146331064627, "ce_loss": 3.0110146331064627, "z_loss": 32.09631526831424, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1615419238805771, "tokens_per_second": 4347039.355203689, "eta_seconds": 1482.201493365153, "mean_data_wait_seconds": 7.697939872741699e-05}
{"step": 34330, "total_steps": 46238, "tokens": 18561810432, "loss": 3.000867430007819, "ce_loss": 3.000867430007819, "z_loss": 31.90480675841823, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13644888997077942, "tokens_per_second": 4347069.286579023, "eta_seconds": 1480.947525698696, "mean_data_wait_seconds": 8.574128150939941e-05}
{"step": 34340, "total_steps": 46238, "tokens": 18567217152, "loss": 3.004207594828172, "ce_loss": 3.004207594828172, "z_loss": 31.78182432868264, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1458217203617096, "tokens_per_second": 4347162.207187902, "eta_seconds": 1479.6721349307513, "mean_data_wait_seconds": 0.00010076165199279785}
{"step": 34350, "total_steps": 46238, "tokens": 18572623872, "loss": 3.0382737300612708, "ce_loss": 3.0382737300612708, "z_loss": 31.847984487360176, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17094369232654572, "tokens_per_second": 4346450.516335355, "eta_seconds": 1478.6704783237249, "mean_data_wait_seconds": 0.00014027953147888184}
{"step": 34360, "total_steps": 46238, "tokens": 18578030592, "loss": 3.027925061457085, "ce_loss": 3.027925061457085, "z_loss": 32.73962104681766, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17487332224845886, "tokens_per_second": 4345982.536061678, "eta_seconds": 1477.585630111438, "mean_data_wait_seconds": 6.61313533782959e-05}
{"step": 34370, "total_steps": 46238, "tokens": 18583437312, "loss": 3.0064955696915137, "ce_loss": 3.0064955696915137, "z_loss": 32.91236062483354, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16416069865226746, "tokens_per_second": 4346098.657364845, "eta_seconds": 1476.3021113492819, "mean_data_wait_seconds": 8.350610733032227e-05}
{"step": 34380, "total_steps": 46238, "tokens": 18588844032, "loss": 2.975228699770841, "ce_loss": 2.975228699770841, "z_loss": 32.73008476604115, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14968286454677582, "tokens_per_second": 4346116.887818499, "eta_seconds": 1475.0518841240435, "mean_data_wait_seconds": 0.0001080632209777832}
{"step": 34390, "total_steps": 46238, "tokens": 18594250752, "loss": 3.038146298943144, "ce_loss": 3.038146298943144, "z_loss": 33.10316970131614, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16219337284564972, "tokens_per_second": 4345580.875495687, "eta_seconds": 1473.9896385588181, "mean_data_wait_seconds": 9.1552734375e-05}
{"step": 34400, "total_steps": 46238, "tokens": 18599657472, "loss": 3.0142754262143914, "ce_loss": 3.0142754262143914, "z_loss": 31.701557217222273, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1413963884115219, "tokens_per_second": 4344974.174157971, "eta_seconds": 1472.951094177739, "mean_data_wait_seconds": 0.00011539459228515625}
{"step": 34410, "total_steps": 46238, "tokens": 18605064192, "loss": 3.0644081541986177, "ce_loss": 3.0644081541986177, "z_loss": 32.02123072653106, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17033925652503967, "tokens_per_second": 4344949.512749698, "eta_seconds": 1471.7150855806442, "mean_data_wait_seconds": 0.00011175870895385742}
{"step": 34420, "total_steps": 46238, "tokens": 18610470912, "loss": 3.0302519129984304, "ce_loss": 3.0302519129984304, "z_loss": 31.700004071900338, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16041137278079987, "tokens_per_second": 4344907.871412942, "eta_seconds": 1470.4848095944299, "mean_data_wait_seconds": 0.00016555190086364746}
{"step": 34430, "total_steps": 46238, "tokens": 18615877632, "loss": 2.934092156814806, "ce_loss": 2.934092156814806, "z_loss": 32.43623734965469, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13610143959522247, "tokens_per_second": 4344830.576297883, "eta_seconds": 1469.266566762057, "mean_data_wait_seconds": 0.00013306736946105957}
{"step": 34440, "total_steps": 46238, "tokens": 18621284352, "loss": 3.011707983233712, "ce_loss": 3.011707983233712, "z_loss": 32.75553868033669, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16334308683872223, "tokens_per_second": 4344785.554700224, "eta_seconds": 1468.0373757687294, "mean_data_wait_seconds": 0.00011011958122253418}
{"step": 34450, "total_steps": 46238, "tokens": 18626691072, "loss": 3.049065154610258, "ce_loss": 3.049065154610258, "z_loss": 32.920508442503035, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15407556295394897, "tokens_per_second": 4344719.377729879, "eta_seconds": 1466.8153015051225, "mean_data_wait_seconds": 0.001994490623474121}
{"step": 34460, "total_steps": 46238, "tokens": 18632097792, "loss": 3.011059325752836, "ce_loss": 3.011059325752836, "z_loss": 32.6536063858957, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1911466121673584, "tokens_per_second": 4344696.497029517, "eta_seconds": 1465.578584914614, "mean_data_wait_seconds": 8.064508438110352e-05}
{"step": 34470, "total_steps": 46238, "tokens": 18637504512, "loss": 3.0471649206045903, "ce_loss": 3.0471649206045903, "z_loss": 31.669456626429703, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17664605379104614, "tokens_per_second": 4344704.69639684, "eta_seconds": 1464.3313800535675, "mean_data_wait_seconds": 0.00011560320854187012}
{"step": 34480, "total_steps": 46238, "tokens": 18642911232, "loss": 3.061918889031266, "ce_loss": 3.061918889031266, "z_loss": 31.557277708342582, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15705232322216034, "tokens_per_second": 4344603.003516125, "eta_seconds": 1463.1211871039734, "mean_data_wait_seconds": 8.392333984375e-05}
{"step": 34490, "total_steps": 46238, "tokens": 18648317952, "loss": 2.9833528164661294, "ce_loss": 2.9833528164661294, "z_loss": 31.799615180853642, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16874532401561737, "tokens_per_second": 4344543.649390009, "eta_seconds": 1461.8966907817219, "mean_data_wait_seconds": 8.100271224975586e-05}
{"step": 34500, "total_steps": 46238, "tokens": 18653724672, "loss": 3.041322247548537, "ce_loss": 3.041322247548537, "z_loss": 32.87494737451727, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17508307099342346, "tokens_per_second": 4344569.007786324, "eta_seconds": 1460.6436801042762, "mean_data_wait_seconds": 7.715821266174316e-05}
{"step": 34510, "total_steps": 46238, "tokens": 18659131392, "loss": 3.023471373500246, "ce_loss": 3.023471373500246, "z_loss": 32.117609255241625, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16786426305770874, "tokens_per_second": 4344481.0068134675, "eta_seconds": 1459.4287635407381, "mean_data_wait_seconds": 0.00011351704597473145}
{"step": 34520, "total_steps": 46238, "tokens": 18664538112, "loss": 2.9777855403495557, "ce_loss": 2.9777855403495557, "z_loss": 32.82732316219445, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17114472389221191, "tokens_per_second": 4344520.714369315, "eta_seconds": 1458.170933112848, "mean_data_wait_seconds": 0.0001150667667388916}
{"step": 34530, "total_steps": 46238, "tokens": 18669944832, "loss": 3.0043383529691985, "ce_loss": 3.0043383529691985, "z_loss": 33.35017004880038, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16691218316555023, "tokens_per_second": 4344602.786671647, "eta_seconds": 1456.8989191412534, "mean_data_wait_seconds": 7.849931716918945e-05}
{"step": 34540, "total_steps": 46238, "tokens": 18675351552, "loss": 3.0191502553043943, "ce_loss": 3.0191502553043943, "z_loss": 32.51592875971939, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15037061274051666, "tokens_per_second": 4344674.527458118, "eta_seconds": 1455.6304146676875, "mean_data_wait_seconds": 6.154179573059082e-05}
{"step": 34550, "total_steps": 46238, "tokens": 18680758272, "loss": 2.9880220998417246, "ce_loss": 2.9880220998417246, "z_loss": 31.289579839417428, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1607673466205597, "tokens_per_second": 4344628.962088222, "eta_seconds": 1454.4012202512429, "mean_data_wait_seconds": 5.689263343811035e-05}
{"step": 34560, "total_steps": 46238, "tokens": 18686164992, "loss": 3.0052353703614436, "ce_loss": 3.0052353703614436, "z_loss": 31.81748439326431, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13879533112049103, "tokens_per_second": 4344730.362737552, "eta_seconds": 1453.122844664174, "mean_data_wait_seconds": 5.981326103210449e-05}
{"step": 34570, "total_steps": 46238, "tokens": 18691571712, "loss": 3.0008756536425967, "ce_loss": 3.0008756536425967, "z_loss": 33.10516181136622, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15994945168495178, "tokens_per_second": 4344870.727813531, "eta_seconds": 1451.831508730384, "mean_data_wait_seconds": 0.00011304020881652832}
{"step": 34580, "total_steps": 46238, "tokens": 18696978432, "loss": 3.024057814569184, "ce_loss": 3.024057814569184, "z_loss": 32.778969995903246, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15106405317783356, "tokens_per_second": 4344908.428198586, "eta_seconds": 1450.574530661187, "mean_data_wait_seconds": 0.00010672211647033691}
{"step": 34590, "total_steps": 46238, "tokens": 18702385152, "loss": 2.9839710250045313, "ce_loss": 2.9839710250045313, "z_loss": 33.03686555226644, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1581423580646515, "tokens_per_second": 4345001.334609069, "eta_seconds": 1449.2991598969384, "mean_data_wait_seconds": 6.595253944396973e-05}
{"step": 34600, "total_steps": 46238, "tokens": 18707791872, "loss": 2.936630312240485, "ce_loss": 2.936630312240485, "z_loss": 32.70829330791127, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1452060341835022, "tokens_per_second": 4345115.063693491, "eta_seconds": 1448.0169044480408, "mean_data_wait_seconds": 6.93202018737793e-05}
{"step": 34610, "total_steps": 46238, "tokens": 18713198592, "loss": 3.013605690363682, "ce_loss": 3.013605690363682, "z_loss": 33.31944621693004, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17038485407829285, "tokens_per_second": 4345135.013680347, "eta_seconds": 1446.7659403465575, "mean_data_wait_seconds": 7.402896881103516e-05}
{"step": 34620, "total_steps": 46238, "tokens": 18718605312, "loss": 2.983086008014101, "ce_loss": 2.983086008014101, "z_loss": 33.461280707157016, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1654236614704132, "tokens_per_second": 4345202.638871231, "eta_seconds": 1445.499127661313, "mean_data_wait_seconds": 6.195902824401855e-05}
{"step": 34630, "total_steps": 46238, "tokens": 18724012032, "loss": 2.983977558034839, "ce_loss": 2.983977558034839, "z_loss": 32.35374155911532, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15528127551078796, "tokens_per_second": 4345222.127794193, "eta_seconds": 1444.2483535785852, "mean_data_wait_seconds": 8.189678192138672e-05}
{"step": 34640, "total_steps": 46238, "tokens": 18729418752, "loss": 2.9557288162636035, "ce_loss": 2.9557288162636035, "z_loss": 32.671589677984066, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19807614386081696, "tokens_per_second": 4345328.31738404, "eta_seconds": 1442.968799138922, "mean_data_wait_seconds": 0.00012001395225524902}
{"step": 34650, "total_steps": 46238, "tokens": 18734825472, "loss": 2.961652231938911, "ce_loss": 2.961652231938911, "z_loss": 32.569179115873396, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18030773103237152, "tokens_per_second": 4345318.832780315, "eta_seconds": 1441.727685604958, "mean_data_wait_seconds": 8.755922317504883e-05}
{"step": 34660, "total_steps": 46238, "tokens": 18740232192, "loss": 2.9234463391882, "ce_loss": 2.9234463391882, "z_loss": 33.50632592403527, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1618366241455078, "tokens_per_second": 4345429.755041017, "eta_seconds": 1440.4466524257316, "mean_data_wait_seconds": 7.036328315734863e-05}
{"step": 34670, "total_steps": 46238, "tokens": 18745638912, "loss": 2.9582877628731006, "ce_loss": 2.9582877628731006, "z_loss": 32.137506716179125, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16607573628425598, "tokens_per_second": 4345505.530917187, "eta_seconds": 1439.1773245954205, "mean_data_wait_seconds": 0.00011134147644042969}
{"step": 34680, "total_steps": 46238, "tokens": 18751045632, "loss": 2.922007071249413, "ce_loss": 2.922007071249413, "z_loss": 32.230439735181406, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15577739477157593, "tokens_per_second": 4345612.810663356, "eta_seconds": 1437.8976168026718, "mean_data_wait_seconds": 7.766485214233398e-05}
{"step": 34690, "total_steps": 46238, "tokens": 18756452352, "loss": 2.964327441923546, "ce_loss": 2.964327441923546, "z_loss": 32.485308372613154, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16732823848724365, "tokens_per_second": 4345682.866033213, "eta_seconds": 1436.6302780163078, "mean_data_wait_seconds": 6.341934204101562e-05}
{"step": 34700, "total_steps": 46238, "tokens": 18761859072, "loss": 2.9104768760276563, "ce_loss": 2.9104768760276563, "z_loss": 32.91150403745247, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15803484618663788, "tokens_per_second": 4345867.118704535, "eta_seconds": 1435.3252627428271, "mean_data_wait_seconds": 0.00010779500007629395}
{"step": 34710, "total_steps": 46238, "tokens": 18767265792, "loss": 2.8710123842412774, "ce_loss": 2.8710123842412774, "z_loss": 33.253523667653404, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1469612568616867, "tokens_per_second": 4345899.867890775, "eta_seconds": 1434.0703498593898, "mean_data_wait_seconds": 6.324052810668945e-05}
{"step": 34720, "total_steps": 46238, "tokens": 18772672512, "loss": 2.90696390469869, "ce_loss": 2.90696390469869, "z_loss": 33.04797978834672, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15267905592918396, "tokens_per_second": 4345954.266712009, "eta_seconds": 1432.8083182318117, "mean_data_wait_seconds": 8.919835090637207e-05}
{"step": 34730, "total_steps": 46238, "tokens": 18778079232, "loss": 2.981963710351424, "ce_loss": 2.981963710351424, "z_loss": 31.539441975680266, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.156734436750412, "tokens_per_second": 4345963.324607212, "eta_seconds": 1431.5612533527994, "mean_data_wait_seconds": 7.283687591552734e-05}
{"step": 34740, "total_steps": 46238, "tokens": 18783485952, "loss": 3.0043092875769646, "ce_loss": 3.0043092875769646, "z_loss": 32.06882241277984, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20353853702545166, "tokens_per_second": 4345933.471707792, "eta_seconds": 1430.3269998188212, "mean_data_wait_seconds": 0.00010856986045837402}
{"step": 34750, "total_steps": 46238, "tokens": 18788892672, "loss": 2.951942077188781, "ce_loss": 2.951942077188781, "z_loss": 32.96268627860329, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14610007405281067, "tokens_per_second": 4345936.953980003, "eta_seconds": 1429.0817675834553, "mean_data_wait_seconds": 6.538629531860352e-05}
{"step": 34760, "total_steps": 46238, "tokens": 18794299392, "loss": 2.984926628343987, "ce_loss": 2.984926628343987, "z_loss": 33.26129112821637, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1724894940853119, "tokens_per_second": 4345023.658573639, "eta_seconds": 1428.1378035205084, "mean_data_wait_seconds": 0.00010690093040466309}
{"step": 34770, "total_steps": 46238, "tokens": 18799706112, "loss": 2.9802292838241113, "ce_loss": 2.9802292838241113, "z_loss": 32.659235607494004, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.22244292497634888, "tokens_per_second": 4345036.008581499, "eta_seconds": 1426.8894001695614, "mean_data_wait_seconds": 0.00014275312423706055}
{"step": 34780, "total_steps": 46238, "tokens": 18805112832, "loss": 2.9834037098017605, "ce_loss": 2.9834037098017605, "z_loss": 32.97647161194772, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20078115165233612, "tokens_per_second": 4344987.319430965, "eta_seconds": 1425.661031574944, "mean_data_wait_seconds": 0.00010281801223754883}
{"step": 34790, "total_steps": 46238, "tokens": 18810519552, "loss": 2.994180917739868, "ce_loss": 2.994180917739868, "z_loss": 32.83368214693937, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1811031848192215, "tokens_per_second": 4344385.254138487, "eta_seconds": 1424.6140758590072, "mean_data_wait_seconds": 9.450316429138184e-05}
{"step": 34800, "total_steps": 46238, "tokens": 18815926272, "loss": 2.9887133222637754, "ce_loss": 2.9887133222637754, "z_loss": 33.41877029881333, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13986815512180328, "tokens_per_second": 4344358.762483621, "eta_seconds": 1423.3782249753397, "mean_data_wait_seconds": 7.876753807067871e-05}
{"step": 34810, "total_steps": 46238, "tokens": 18821332992, "loss": 2.9737368207989316, "ce_loss": 2.9737368207989316, "z_loss": 33.5897585550944, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15050512552261353, "tokens_per_second": 4343931.771721398, "eta_seconds": 1422.2734768119299, "mean_data_wait_seconds": 0.05791357159614563}
{"step": 34820, "total_steps": 46238, "tokens": 18826739712, "loss": 2.9606845812364058, "ce_loss": 2.9606845812364058, "z_loss": 33.04165785240404, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2076982855796814, "tokens_per_second": 4343961.821842299, "eta_seconds": 1421.018986161821, "mean_data_wait_seconds": 0.00011444091796875}
{"step": 34830, "total_steps": 46238, "tokens": 18832146432, "loss": 2.983764077677871, "ce_loss": 2.983764077677871, "z_loss": 32.09103315526789, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14480793476104736, "tokens_per_second": 4343889.5576626565, "eta_seconds": 1419.7979534540825, "mean_data_wait_seconds": 8.887052536010742e-05}
{"step": 34840, "total_steps": 46238, "tokens": 18837553152, "loss": 3.0021208452455923, "ce_loss": 3.0021208452455923, "z_loss": 32.26711929205692, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1986391395330429, "tokens_per_second": 4343785.801330132, "eta_seconds": 1418.5871647062088, "mean_data_wait_seconds": 0.00011593103408813477}
{"step": 34850, "total_steps": 46238, "tokens": 18842959872, "loss": 3.0426951574556753, "ce_loss": 3.0426951574556753, "z_loss": 32.89071191440929, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17188119888305664, "tokens_per_second": 4343743.172955093, "eta_seconds": 1417.356371880426, "mean_data_wait_seconds": 0.00013247132301330566}
{"step": 34860, "total_steps": 46238, "tokens": 18848366592, "loss": 2.993355205564788, "ce_loss": 2.993355205564788, "z_loss": 31.995446277387213, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15602682530879974, "tokens_per_second": 4343609.125702024, "eta_seconds": 1416.1553597449504, "mean_data_wait_seconds": 0.0001487433910369873}
{"step": 34870, "total_steps": 46238, "tokens": 18853773312, "loss": 2.8717280572110955, "ce_loss": 2.8717280572110955, "z_loss": 32.21173498847268, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1611478328704834, "tokens_per_second": 4343720.429214628, "eta_seconds": 1414.8743511817593, "mean_data_wait_seconds": 0.00012058019638061523}
{"step": 34880, "total_steps": 46238, "tokens": 18859180032, "loss": 2.983410044149919, "ce_loss": 2.983410044149919, "z_loss": 33.605589693242855, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1673043966293335, "tokens_per_second": 4343640.937865758, "eta_seconds": 1413.6555004974887, "mean_data_wait_seconds": 7.796287536621094e-05}
{"step": 34890, "total_steps": 46238, "tokens": 18864586752, "loss": 2.9254614530187664, "ce_loss": 2.9254614530187664, "z_loss": 32.62382542003285, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15315702557563782, "tokens_per_second": 4343466.372346152, "eta_seconds": 1412.4675220372749, "mean_data_wait_seconds": 0.00013267993927001953}
{"step": 34900, "total_steps": 46238, "tokens": 18869993472, "loss": 2.9717947804566585, "ce_loss": 2.9717947804566585, "z_loss": 32.9996107563828, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1340925395488739, "tokens_per_second": 4343422.171611393, "eta_seconds": 1411.23708951505, "mean_data_wait_seconds": 7.143616676330566e-05}
{"step": 34910, "total_steps": 46238, "tokens": 18875400192, "loss": 3.000648709860715, "ce_loss": 3.000648709860715, "z_loss": 32.198477456063934, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18736019730567932, "tokens_per_second": 4343448.2891365485, "eta_seconds": 1409.9838046460206, "mean_data_wait_seconds": 6.490945816040039e-05}
{"step": 34920, "total_steps": 46238, "tokens": 18880806912, "loss": 2.9944819753820244, "ce_loss": 2.9944819753820244, "z_loss": 32.3249695373304, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13520726561546326, "tokens_per_second": 4343501.029007887, "eta_seconds": 1408.7219004061367, "mean_data_wait_seconds": 9.587407112121582e-05}
{"step": 34930, "total_steps": 46238, "tokens": 18886213632, "loss": 3.0392637794668023, "ce_loss": 3.0392637794668023, "z_loss": 32.68632734183109, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18064793944358826, "tokens_per_second": 4343457.226420763, "eta_seconds": 1407.4913105654653, "mean_data_wait_seconds": 7.581710815429688e-05}
{"step": 34940, "total_steps": 46238, "tokens": 18891620352, "loss": 2.988961198113181, "ce_loss": 2.988961198113181, "z_loss": 32.26753022454002, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1807771772146225, "tokens_per_second": 4343423.877489215, "eta_seconds": 1406.2573113473811, "mean_data_wait_seconds": 7.212162017822266e-05}
{"step": 34950, "total_steps": 46238, "tokens": 18897027072, "loss": 2.9859877120364797, "ce_loss": 2.9859877120364797, "z_loss": 33.3908383918531, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1528148055076599, "tokens_per_second": 4343361.410786652, "eta_seconds": 1405.0327124158725, "mean_data_wait_seconds": 7.265806198120117e-05}
{"step": 34960, "total_steps": 46238, "tokens": 18902433792, "loss": 2.9860884782039756, "ce_loss": 2.9860884782039756, "z_loss": 32.37837395523534, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1648753434419632, "tokens_per_second": 4343530.927526746, "eta_seconds": 1403.7331023384213, "mean_data_wait_seconds": 0.0001989305019378662}
{"step": 34970, "total_steps": 46238, "tokens": 18907840512, "loss": 2.9846309622128806, "ce_loss": 2.9846309622128806, "z_loss": 32.79009258385861, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18369586765766144, "tokens_per_second": 4343660.889256165, "eta_seconds": 1402.4463647859002, "mean_data_wait_seconds": 6.264448165893555e-05}
{"step": 34980, "total_steps": 46238, "tokens": 18913247232, "loss": 3.0648876121549895, "ce_loss": 3.0648876121549895, "z_loss": 32.47399507869374, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1548537313938141, "tokens_per_second": 4343796.039620869, "eta_seconds": 1401.1580305532077, "mean_data_wait_seconds": 6.455183029174805e-05}
{"step": 34990, "total_steps": 46238, "tokens": 18918653952, "loss": 2.956272916360335, "ce_loss": 2.956272916360335, "z_loss": 32.66574954986572, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.21626774966716766, "tokens_per_second": 4343844.055443806, "eta_seconds": 1399.8978569175908, "mean_data_wait_seconds": 9.760260581970215e-05}
{"step": 35000, "total_steps": 46238, "tokens": 18924060672, "loss": 3.068052620598764, "ce_loss": 3.068052620598764, "z_loss": 32.41705440752434, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16057553887367249, "tokens_per_second": 4343851.602656191, "eta_seconds": 1398.6507412649448, "mean_data_wait_seconds": 7.653236389160156e-05}
{"step": 35010, "total_steps": 46238, "tokens": 18929467392, "loss": 2.992811499219952, "ce_loss": 2.992811499219952, "z_loss": 33.23375930208148, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15656471252441406, "tokens_per_second": 4343866.825528224, "eta_seconds": 1397.401160718563, "mean_data_wait_seconds": 8.738040924072266e-05}
{"step": 35020, "total_steps": 46238, "tokens": 18934874112, "loss": 3.034324212507768, "ce_loss": 3.034324212507768, "z_loss": 33.67565602967233, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18713131546974182, "tokens_per_second": 4343911.096686445, "eta_seconds": 1396.14225268703, "mean_data_wait_seconds": 0.00012034177780151367}
{"step": 35030, "total_steps": 46238, "tokens": 18940280832, "loss": 3.0442464550336203, "ce_loss": 3.0442464550336203, "z_loss": 32.353252844376996, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15834079682826996, "tokens_per_second": 4343892.1614992265, "eta_seconds": 1394.9036667403648, "mean_data_wait_seconds": 6.622076034545898e-05}
{"step": 35040, "total_steps": 46238, "tokens": 18945687552, "loss": 3.015541369264776, "ce_loss": 3.015541369264776, "z_loss": 32.106469790140785, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17135021090507507, "tokens_per_second": 4344008.296514169, "eta_seconds": 1393.6217361412337, "mean_data_wait_seconds": 7.164478302001953e-05}
{"step": 35050, "total_steps": 46238, "tokens": 18951094272, "loss": 3.04389294349786, "ce_loss": 3.04389294349786, "z_loss": 32.49935658772787, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12731589376926422, "tokens_per_second": 4344086.237778703, "eta_seconds": 1392.3521157104901, "mean_data_wait_seconds": 5.4508447647094727e-05}
{"step": 35060, "total_steps": 46238, "tokens": 18956500992, "loss": 3.0066436944585857, "ce_loss": 3.0066436944585857, "z_loss": 33.04651439551151, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18272440135478973, "tokens_per_second": 4344058.129893982, "eta_seconds": 1391.1165005859357, "mean_data_wait_seconds": 6.595253944396973e-05}
{"step": 35070, "total_steps": 46238, "tokens": 18961907712, "loss": 3.0466820832454795, "ce_loss": 3.0466820832454795, "z_loss": 32.09131482153228, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14143750071525574, "tokens_per_second": 4344105.471081314, "eta_seconds": 1389.8567298130376, "mean_data_wait_seconds": 6.788969039916992e-05}
{"step": 35080, "total_steps": 46238, "tokens": 18967314432, "loss": 3.047737565907565, "ce_loss": 3.047737565907565, "z_loss": 32.25778543587887, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15123862028121948, "tokens_per_second": 4344237.427428319, "eta_seconds": 1388.5699400115336, "mean_data_wait_seconds": 9.077787399291992e-05}
{"step": 35090, "total_steps": 46238, "tokens": 18972721152, "loss": 3.027632377364419, "ce_loss": 3.027632377364419, "z_loss": 32.83389484521115, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1660800278186798, "tokens_per_second": 4344215.01657997, "eta_seconds": 1387.3325240574118, "mean_data_wait_seconds": 0.00011676549911499023}
{"step": 35100, "total_steps": 46238, "tokens": 18978127872, "loss": 3.050096408887343, "ce_loss": 3.050096408887343, "z_loss": 32.749669233957924, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19347840547561646, "tokens_per_second": 4344300.436009137, "eta_seconds": 1386.060690943276, "mean_data_wait_seconds": 0.00011840462684631348}
{"step": 35110, "total_steps": 46238, "tokens": 18983534592, "loss": 3.029828813942996, "ce_loss": 3.029828813942996, "z_loss": 34.03087483030377, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1566954106092453, "tokens_per_second": 4344363.566512825, "eta_seconds": 1384.7960125558795, "mean_data_wait_seconds": 7.969141006469727e-05}
{"step": 35120, "total_steps": 46238, "tokens": 18988941312, "loss": 2.9627025037100823, "ce_loss": 2.9627025037100823, "z_loss": 31.94751768401175, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17709632217884064, "tokens_per_second": 4344454.189490015, "eta_seconds": 1383.522615692623, "mean_data_wait_seconds": 7.236003875732422e-05}
{"step": 35130, "total_steps": 46238, "tokens": 18994348032, "loss": 3.004341347651048, "ce_loss": 3.004341347651048, "z_loss": 33.13230410489169, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15136753022670746, "tokens_per_second": 4344515.319079004, "eta_seconds": 1382.2586555577054, "mean_data_wait_seconds": 9.649991989135742e-05}
{"step": 35140, "total_steps": 46238, "tokens": 18999754752, "loss": 2.9892759883042537, "ce_loss": 2.9892759883042537, "z_loss": 32.535035725795865, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16417594254016876, "tokens_per_second": 4344513.334750704, "eta_seconds": 1381.0147930744658, "mean_data_wait_seconds": 0.0001119077205657959}
{"step": 35150, "total_steps": 46238, "tokens": 19005161472, "loss": 2.9868219928307966, "ce_loss": 2.9868219928307966, "z_loss": 32.31190771045107, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1857401430606842, "tokens_per_second": 4344544.862753753, "eta_seconds": 1379.7602863745044, "mean_data_wait_seconds": 0.00014573335647583008}
{"step": 35160, "total_steps": 46238, "tokens": 19010568192, "loss": 2.995932082335154, "ce_loss": 2.995932082335154, "z_loss": 33.1843099305124, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14922019839286804, "tokens_per_second": 4344596.023245655, "eta_seconds": 1378.4995686493921, "mean_data_wait_seconds": 9.140372276306152e-05}
{"step": 35170, "total_steps": 46238, "tokens": 19015974912, "loss": 2.9957555387959336, "ce_loss": 2.9957555387959336, "z_loss": 32.75926913637103, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17710046470165253, "tokens_per_second": 4344620.830080206, "eta_seconds": 1377.247234688956, "mean_data_wait_seconds": 6.622076034545898e-05}
{"step": 35180, "total_steps": 46238, "tokens": 19021381632, "loss": 2.983159332564383, "ce_loss": 2.983159332564383, "z_loss": 31.185293717817828, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15097057819366455, "tokens_per_second": 4344737.789702545, "eta_seconds": 1375.9657298925945, "mean_data_wait_seconds": 0.00010856986045837402}
{"step": 35190, "total_steps": 46238, "tokens": 19026788352, "loss": 2.993885708577705, "ce_loss": 2.993885708577705, "z_loss": 32.731159470298074, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1517915278673172, "tokens_per_second": 4344830.1243905295, "eta_seconds": 1374.692085306289, "mean_data_wait_seconds": 0.00015482306480407715}
{"step": 35200, "total_steps": 46238, "tokens": 19032195072, "loss": 2.910878519217173, "ce_loss": 2.910878519217173, "z_loss": 33.16097970442338, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14767107367515564, "tokens_per_second": 4344824.170704347, "eta_seconds": 1373.4495642507472, "mean_data_wait_seconds": 7.161498069763184e-05}
{"step": 35210, "total_steps": 46238, "tokens": 19037601792, "loss": 2.9559852480888367, "ce_loss": 2.9559852480888367, "z_loss": 33.72176141449899, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.22007416188716888, "tokens_per_second": 4344831.4517744705, "eta_seconds": 1372.202859921083, "mean_data_wait_seconds": 6.42240047454834e-05}
{"step": 35220, "total_steps": 46238, "tokens": 19043008512, "loss": 2.919172864971739, "ce_loss": 2.919172864971739, "z_loss": 34.08512020111084, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17413555085659027, "tokens_per_second": 4344952.546493602, "eta_seconds": 1370.9202483251497, "mean_data_wait_seconds": 0.00011295080184936523}
{"step": 35230, "total_steps": 46238, "tokens": 19048415232, "loss": 2.953414261341095, "ce_loss": 2.953414261341095, "z_loss": 32.625301736773864, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16500073671340942, "tokens_per_second": 4345011.282882589, "eta_seconds": 1369.657364860015, "mean_data_wait_seconds": 0.00013589859008789062}
{"step": 35240, "total_steps": 46238, "tokens": 19053821952, "loss": 2.915190678654295, "ce_loss": 2.915190678654295, "z_loss": 33.3089369282578, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1364239752292633, "tokens_per_second": 4345021.591674811, "eta_seconds": 1368.4097670290682, "mean_data_wait_seconds": 7.876753807067871e-05}
{"step": 35250, "total_steps": 46238, "tokens": 19059228672, "loss": 2.9239614244663352, "ce_loss": 2.9239614244663352, "z_loss": 32.047961697433934, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18874326348304749, "tokens_per_second": 4345018.208312993, "eta_seconds": 1367.166483361279, "mean_data_wait_seconds": 6.553530693054199e-05}
{"step": 35260, "total_steps": 46238, "tokens": 19064635392, "loss": 2.9287525325110466, "ce_loss": 2.9287525325110466, "z_loss": 33.23292012648149, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17675650119781494, "tokens_per_second": 4344989.813276585, "eta_seconds": 1365.9310606126394, "mean_data_wait_seconds": 8.568167686462402e-05}
{"step": 35270, "total_steps": 46238, "tokens": 19070042112, "loss": 2.927390535672506, "ce_loss": 2.927390535672506, "z_loss": 33.73304598259203, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14322684705257416, "tokens_per_second": 4345046.369944208, "eta_seconds": 1364.6689400178109, "mean_data_wait_seconds": 8.195638656616211e-05}
{"step": 35280, "total_steps": 46238, "tokens": 19075448832, "loss": 2.899074800086744, "ce_loss": 2.899074800086744, "z_loss": 33.17404605403091, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17802532017230988, "tokens_per_second": 4345010.637297061, "eta_seconds": 1363.4358114449367, "mean_data_wait_seconds": 9.936094284057617e-05}
{"step": 35290, "total_steps": 46238, "tokens": 19080855552, "loss": 2.9286209977034368, "ce_loss": 2.9286209977034368, "z_loss": 32.49417579535282, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14843034744262695, "tokens_per_second": 4345052.948340259, "eta_seconds": 1362.1781953798431, "mean_data_wait_seconds": 7.745623588562012e-05}
{"step": 35300, "total_steps": 46238, "tokens": 19086262272, "loss": 2.909720373876167, "ce_loss": 2.909720373876167, "z_loss": 32.75457203026974, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17058879137039185, "tokens_per_second": 4344910.166410119, "eta_seconds": 1360.9785789623702, "mean_data_wait_seconds": 9.34898853302002e-05}
{"step": 35310, "total_steps": 46238, "tokens": 19091668992, "loss": 2.923395109899116, "ce_loss": 2.923395109899116, "z_loss": 32.97589738441236, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13685224950313568, "tokens_per_second": 4344895.934895646, "eta_seconds": 1359.7386525534573, "mean_data_wait_seconds": 8.958578109741211e-05}
{"step": 35320, "total_steps": 46238, "tokens": 19097075712, "loss": 2.958381697987065, "ce_loss": 2.958381697987065, "z_loss": 33.0340819792314, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16854265332221985, "tokens_per_second": 4344721.624950224, "eta_seconds": 1358.548771019967, "mean_data_wait_seconds": 0.00016582012176513672}
{"step": 35330, "total_steps": 46238, "tokens": 19102482432, "loss": 2.9547932436971953, "ce_loss": 2.9547932436971953, "z_loss": 32.737192269527554, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1473018229007721, "tokens_per_second": 4344562.528650184, "eta_seconds": 1357.354040852573, "mean_data_wait_seconds": 9.042024612426758e-05}
{"step": 35340, "total_steps": 46238, "tokens": 19107889152, "loss": 2.8952919587944494, "ce_loss": 2.8952919587944494, "z_loss": 33.28004342859442, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14735867083072662, "tokens_per_second": 4344540.155572153, "eta_seconds": 1356.1165446804564, "mean_data_wait_seconds": 9.965896606445312e-05}
{"step": 35350, "total_steps": 46238, "tokens": 19113295872, "loss": 2.959614229924751, "ce_loss": 2.959614229924751, "z_loss": 32.939100814588144, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1370573192834854, "tokens_per_second": 4344567.222237631, "eta_seconds": 1354.8636176857947, "mean_data_wait_seconds": 0.00010162591934204102}
{"step": 35360, "total_steps": 46238, "tokens": 19118702592, "loss": 2.8751828399571506, "ce_loss": 2.8751828399571506, "z_loss": 33.108047889940664, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17410914599895477, "tokens_per_second": 4344531.728371204, "eta_seconds": 1353.6301980708029, "mean_data_wait_seconds": 0.00011268258094787598}
{"step": 35370, "total_steps": 46238, "tokens": 19124109312, "loss": 2.9736596382025517, "ce_loss": 2.9736596382025517, "z_loss": 33.052349697459825, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14893917739391327, "tokens_per_second": 4344516.988760789, "eta_seconds": 1352.3902977476669, "mean_data_wait_seconds": 7.867813110351562e-05}
{"step": 35380, "total_steps": 46238, "tokens": 19129516032, "loss": 2.9875631494955583, "ce_loss": 2.9875631494955583, "z_loss": 32.29876687309959, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1493454873561859, "tokens_per_second": 4344617.726240123, "eta_seconds": 1351.1144763201123, "mean_data_wait_seconds": 5.8323144912719727e-05}
{"step": 35390, "total_steps": 46238, "tokens": 19134922752, "loss": 2.9550193205024256, "ce_loss": 2.9550193205024256, "z_loss": 33.370333382577606, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15714207291603088, "tokens_per_second": 4344768.169453225, "eta_seconds": 1349.8232714078388, "mean_data_wait_seconds": 6.073713302612305e-05}
{"step": 35400, "total_steps": 46238, "tokens": 19140329472, "loss": 2.9519279057329353, "ce_loss": 2.9519279057329353, "z_loss": 32.39104148113366, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16442139446735382, "tokens_per_second": 4344829.688986815, "eta_seconds": 1348.5597557142316, "mean_data_wait_seconds": 7.30752944946289e-05}
{"step": 35410, "total_steps": 46238, "tokens": 19145736192, "loss": 3.0262058214707808, "ce_loss": 3.0262058214707808, "z_loss": 32.61913727269028, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.109794020652771, "tokens_per_second": 4344793.719248242, "eta_seconds": 1347.3265066800143, "mean_data_wait_seconds": 0.00011342763900756836}
{"step": 35420, "total_steps": 46238, "tokens": 19151142912, "loss": 2.9818312471563164, "ce_loss": 2.9818312471563164, "z_loss": 32.78752295176188, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16760112345218658, "tokens_per_second": 4344835.1242915215, "eta_seconds": 1346.0692653909764, "mean_data_wait_seconds": 9.530782699584961e-05}
{"step": 35430, "total_steps": 46238, "tokens": 19156549632, "loss": 3.0576030846798057, "ce_loss": 3.0576030846798057, "z_loss": 33.20267746665261, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.154981791973114, "tokens_per_second": 4344923.881889231, "eta_seconds": 1344.7973918151513, "mean_data_wait_seconds": 0.0001310408115386963}
{"step": 35440, "total_steps": 46238, "tokens": 19161956352, "loss": 3.025076510328235, "ce_loss": 3.025076510328235, "z_loss": 33.052918029553965, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16079257428646088, "tokens_per_second": 4344981.9421262015, "eta_seconds": 1343.5350622293665, "mean_data_wait_seconds": 6.598234176635742e-05}
{"step": 35450, "total_steps": 46238, "tokens": 19167363072, "loss": 3.045247249531023, "ce_loss": 3.045247249531023, "z_loss": 33.67604853890159, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17958830296993256, "tokens_per_second": 4345075.639331759, "eta_seconds": 1342.2617574724095, "mean_data_wait_seconds": 0.0001067817211151123}
{"step": 35460, "total_steps": 46238, "tokens": 19172769792, "loss": 3.03043827866063, "ce_loss": 3.03043827866063, "z_loss": 33.013029329704516, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14364959299564362, "tokens_per_second": 4345177.85266985, "eta_seconds": 1340.9858794202794, "mean_data_wait_seconds": 8.89897346496582e-05}
{"step": 35470, "total_steps": 46238, "tokens": 19178176512, "loss": 2.9950829672090933, "ce_loss": 2.9950829672090933, "z_loss": 34.13477443926262, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16137376427650452, "tokens_per_second": 4345165.090608339, "eta_seconds": 1339.745510839723, "mean_data_wait_seconds": 9.1552734375e-05}
{"step": 35480, "total_steps": 46238, "tokens": 19183583232, "loss": 3.0606928034262224, "ce_loss": 3.0606928034262224, "z_loss": 31.85216369050922, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17598243057727814, "tokens_per_second": 4345400.602791997, "eta_seconds": 1338.4286595493893, "mean_data_wait_seconds": 0.00012701749801635742}
{"step": 35490, "total_steps": 46238, "tokens": 19188989952, "loss": 3.032970742745833, "ce_loss": 3.032970742745833, "z_loss": 32.71346950531006, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1615002155303955, "tokens_per_second": 4345446.404331907, "eta_seconds": 1337.1703257477766, "mean_data_wait_seconds": 7.259845733642578e-05}
{"step": 35500, "total_steps": 46238, "tokens": 19194396672, "loss": 3.0481757124265036, "ce_loss": 3.0481757124265036, "z_loss": 32.50537339123812, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1538456380367279, "tokens_per_second": 4345516.043556, "eta_seconds": 1335.9046902170733, "mean_data_wait_seconds": 8.118152618408203e-05}
{"step": 35510, "total_steps": 46238, "tokens": 19199803392, "loss": 3.019715991887179, "ce_loss": 3.019715991887179, "z_loss": 33.41793077642267, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15101350843906403, "tokens_per_second": 4345661.430437655, "eta_seconds": 1334.6158316378317, "mean_data_wait_seconds": 8.818507194519043e-05}
{"step": 35520, "total_steps": 46238, "tokens": 19205210112, "loss": 3.014533032070507, "ce_loss": 3.014533032070507, "z_loss": 33.402687246149235, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14895398914813995, "tokens_per_second": 4345797.333656873, "eta_seconds": 1333.3299689620321, "mean_data_wait_seconds": 7.554888725280762e-05}
{"step": 35530, "total_steps": 46238, "tokens": 19210616832, "loss": 2.9991752035690076, "ce_loss": 2.9991752035690076, "z_loss": 33.73468624461781, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16693618893623352, "tokens_per_second": 4345928.18591823, "eta_seconds": 1332.04573484614, "mean_data_wait_seconds": 6.029009819030762e-05}
{"step": 35540, "total_steps": 46238, "tokens": 19216023552, "loss": 3.0358543666926296, "ce_loss": 3.0358543666926296, "z_loss": 32.538257512179285, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.153916597366333, "tokens_per_second": 4346014.181505599, "eta_seconds": 1330.7753133001481, "mean_data_wait_seconds": 7.811188697814941e-05}
{"step": 35550, "total_steps": 46238, "tokens": 19221430272, "loss": 3.0554010326212104, "ce_loss": 3.0554010326212104, "z_loss": 34.24147027911562, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15861865878105164, "tokens_per_second": 4346054.176841034, "eta_seconds": 1329.5190140036188, "mean_data_wait_seconds": 7.408857345581055e-05}
{"step": 35560, "total_steps": 46238, "tokens": 19226836992, "loss": 3.0122378342079394, "ce_loss": 3.0122378342079394, "z_loss": 33.340635270783395, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15947377681732178, "tokens_per_second": 4346126.981777494, "eta_seconds": 1328.2527105637025, "mean_data_wait_seconds": 7.939338684082031e-05}
{"step": 35570, "total_steps": 46238, "tokens": 19232243712, "loss": 3.0189499295119084, "ce_loss": 3.0189499295119084, "z_loss": 31.94805516618671, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15859495103359222, "tokens_per_second": 4346182.259370578, "eta_seconds": 1326.9918010376393, "mean_data_wait_seconds": 8.368492126464844e-05}
{"step": 35580, "total_steps": 46238, "tokens": 19237650432, "loss": 3.0346518140850645, "ce_loss": 3.0346518140850645, "z_loss": 33.83210664806944, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17528711259365082, "tokens_per_second": 4346279.469330939, "eta_seconds": 1325.7181330971764, "mean_data_wait_seconds": 6.902217864990234e-05}
{"step": 35590, "total_steps": 46238, "tokens": 19243057152, "loss": 3.045540486321305, "ce_loss": 3.045540486321305, "z_loss": 33.05811968716708, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14771012961864471, "tokens_per_second": 4346354.721898605, "eta_seconds": 1324.4512131042518, "mean_data_wait_seconds": 0.0002887845039367676}
{"step": 35600, "total_steps": 46238, "tokens": 19248463872, "loss": 3.0189093531984272, "ce_loss": 3.0189093531984272, "z_loss": 33.328203490286164, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12807777523994446, "tokens_per_second": 4346406.513507715, "eta_seconds": 1323.1914792430728, "mean_data_wait_seconds": 0.00018095970153808594}
{"step": 35610, "total_steps": 46238, "tokens": 19253870592, "loss": 2.982298784183733, "ce_loss": 2.982298784183733, "z_loss": 32.321542248581395, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17884889245033264, "tokens_per_second": 4346533.037258359, "eta_seconds": 1321.9090467616002, "mean_data_wait_seconds": 0.00011241436004638672}
{"step": 35620, "total_steps": 46238, "tokens": 19259277312, "loss": 3.016909881071611, "ce_loss": 3.016909881071611, "z_loss": 33.15758791836825, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17616215348243713, "tokens_per_second": 4346643.080196996, "eta_seconds": 1320.631696251407, "mean_data_wait_seconds": 7.531046867370605e-05}
{"step": 35630, "total_steps": 46238, "tokens": 19264684032, "loss": 3.042728723901691, "ce_loss": 3.042728723901691, "z_loss": 32.26116621133053, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1514958292245865, "tokens_per_second": 4346714.259679853, "eta_seconds": 1319.3662066073769, "mean_data_wait_seconds": 0.00017154216766357422}
{"step": 35640, "total_steps": 46238, "tokens": 19270090752, "loss": 2.977333477049163, "ce_loss": 2.977333477049163, "z_loss": 32.95157111774791, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16258563101291656, "tokens_per_second": 4346826.68165817, "eta_seconds": 1318.0882523281066, "mean_data_wait_seconds": 6.428360939025879e-05}
{"step": 35650, "total_steps": 46238, "tokens": 19275497472, "loss": 3.0351321697235107, "ce_loss": 3.0351321697235107, "z_loss": 33.36709155458392, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17300091683864594, "tokens_per_second": 4346920.304205986, "eta_seconds": 1316.8160590525413, "mean_data_wait_seconds": 7.206201553344727e-05}
{"step": 35660, "total_steps": 46238, "tokens": 19280904192, "loss": 3.0147880604772856, "ce_loss": 3.0147880604772856, "z_loss": 33.83056345852938, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16412527859210968, "tokens_per_second": 4347016.977618964, "eta_seconds": 1315.5429972883048, "mean_data_wait_seconds": 9.953975677490234e-05}
{"step": 35670, "total_steps": 46238, "tokens": 19286310912, "loss": 3.046701151313204, "ce_loss": 3.046701151313204, "z_loss": 33.60307719490745, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1605365127325058, "tokens_per_second": 4347062.035891595, "eta_seconds": 1314.2855972213401, "mean_data_wait_seconds": 9.28640365600586e-05}
{"step": 35680, "total_steps": 46238, "tokens": 19291717632, "loss": 3.0377067797111743, "ce_loss": 3.0377067797111743, "z_loss": 33.3972464763757, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16831181943416595, "tokens_per_second": 4347186.813480888, "eta_seconds": 1313.004144726318, "mean_data_wait_seconds": 6.145238876342773e-05}
{"step": 35690, "total_steps": 46238, "tokens": 19297124352, "loss": 3.006680905818939, "ce_loss": 3.006680905818939, "z_loss": 33.392638726667926, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16567200422286987, "tokens_per_second": 4347163.5564254075, "eta_seconds": 1311.767434094206, "mean_data_wait_seconds": 7.334351539611816e-05}
{"step": 35700, "total_steps": 46238, "tokens": 19302531072, "loss": 2.9810507225267813, "ce_loss": 2.9810507225267813, "z_loss": 33.21605838428844, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1519889533519745, "tokens_per_second": 4347293.026650939, "eta_seconds": 1310.4846692123933, "mean_data_wait_seconds": 9.715557098388672e-05}
{"step": 35710, "total_steps": 46238, "tokens": 19307937792, "loss": 2.9234429633978642, "ce_loss": 2.9234429633978642, "z_loss": 32.70411205291748, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14383569359779358, "tokens_per_second": 4347290.59865043, "eta_seconds": 1309.2417023529351, "mean_data_wait_seconds": 8.508563041687012e-05}
{"step": 35720, "total_steps": 46238, "tokens": 19313344512, "loss": 2.9528514778975286, "ce_loss": 2.9528514778975286, "z_loss": 34.44867310379491, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15771213173866272, "tokens_per_second": 4347226.5385542605, "eta_seconds": 1308.0172780439118, "mean_data_wait_seconds": 7.393956184387207e-05}
{"step": 35730, "total_steps": 46238, "tokens": 19318751232, "loss": 2.9435981819123933, "ce_loss": 2.9435981819123933, "z_loss": 33.91666296756629, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17118719220161438, "tokens_per_second": 4347265.476202772, "eta_seconds": 1306.7618564123377, "mean_data_wait_seconds": 0.00010132789611816406}
{"step": 35740, "total_steps": 46238, "tokens": 19324157952, "loss": 2.9508822614496406, "ce_loss": 2.9508822614496406, "z_loss": 33.896350889494926, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17835664749145508, "tokens_per_second": 4347192.812201151, "eta_seconds": 1305.5399723865273, "mean_data_wait_seconds": 8.320808410644531e-05}
{"step": 35750, "total_steps": 46238, "tokens": 19329564672, "loss": 2.9392889268470532, "ce_loss": 2.9392889268470532, "z_loss": 32.71376757188277, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16435478627681732, "tokens_per_second": 4347035.927596474, "eta_seconds": 1304.3433177086768, "mean_data_wait_seconds": 7.352232933044434e-05}
{"step": 35760, "total_steps": 46238, "tokens": 19334971392, "loss": 2.973769010919513, "ce_loss": 2.973769010919513, "z_loss": 33.997916626207754, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2108435183763504, "tokens_per_second": 4347003.594590359, "eta_seconds": 1303.109238522221, "mean_data_wait_seconds": 6.523728370666504e-05}
{"step": 35770, "total_steps": 46238, "tokens": 19340378112, "loss": 2.907010419802232, "ce_loss": 2.907010419802232, "z_loss": 34.15891231190074, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15749327838420868, "tokens_per_second": 4347059.056035418, "eta_seconds": 1301.8488479338228, "mean_data_wait_seconds": 6.458163261413574e-05}
{"step": 35780, "total_steps": 46238, "tokens": 19345784832, "loss": 2.8914998065341604, "ce_loss": 2.8914998065341604, "z_loss": 34.48197139393199, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1765865534543991, "tokens_per_second": 4347132.544140997, "eta_seconds": 1300.5830962343487, "mean_data_wait_seconds": 6.732344627380371e-05}
{"step": 35790, "total_steps": 46238, "tokens": 19351191552, "loss": 2.9010696916869194, "ce_loss": 2.9010696916869194, "z_loss": 33.55052433591901, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18322719633579254, "tokens_per_second": 4347240.77822668, "eta_seconds": 1299.307002338179, "mean_data_wait_seconds": 6.687641143798828e-05}
{"step": 35800, "total_steps": 46238, "tokens": 19356598272, "loss": 2.8499569007844636, "ce_loss": 2.8499569007844636, "z_loss": 35.76773868907582, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20853891968727112, "tokens_per_second": 4347309.667444583, "eta_seconds": 1298.042719675187, "mean_data_wait_seconds": 6.937980651855469e-05}
{"step": 35810, "total_steps": 46238, "tokens": 19362004992, "loss": 2.9100740028150156, "ce_loss": 2.9100740028150156, "z_loss": 33.9811368711067, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16522672772407532, "tokens_per_second": 4347403.297439373, "eta_seconds": 1296.7710972019888, "mean_data_wait_seconds": 6.008148193359375e-05}
{"step": 35820, "total_steps": 46238, "tokens": 19367411712, "loss": 2.7849553805409055, "ce_loss": 2.7849553805409055, "z_loss": 34.381846572413586, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1521826833486557, "tokens_per_second": 4347451.160240794, "eta_seconds": 1295.5131676943436, "mean_data_wait_seconds": 6.16610050201416e-05}
{"step": 35830, "total_steps": 46238, "tokens": 19372818432, "loss": 2.8560033187721716, "ce_loss": 2.8560033187721716, "z_loss": 34.675162228670985, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19716711342334747, "tokens_per_second": 4347557.560242126, "eta_seconds": 1294.23783952998, "mean_data_wait_seconds": 9.593367576599121e-05}
{"step": 35840, "total_steps": 46238, "tokens": 19378225152, "loss": 2.9044915600256487, "ce_loss": 2.9044915600256487, "z_loss": 34.44583858143199, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15000738203525543, "tokens_per_second": 4347668.9273174945, "eta_seconds": 1292.9610966188209, "mean_data_wait_seconds": 0.00010368227958679199}
{"step": 35850, "total_steps": 46238, "tokens": 19383631872, "loss": 2.7998887824289724, "ce_loss": 2.7998887824289724, "z_loss": 35.118701356830016, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15197384357452393, "tokens_per_second": 4347783.281280888, "eta_seconds": 1291.6835317388445, "mean_data_wait_seconds": 0.00010123848915100098}
{"step": 35860, "total_steps": 46238, "tokens": 19389038592, "loss": 2.918578700585799, "ce_loss": 2.918578700585799, "z_loss": 34.88068626866196, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16353310644626617, "tokens_per_second": 4347830.306424347, "eta_seconds": 1290.4260167904565, "mean_data_wait_seconds": 7.295608520507812e-05}
{"step": 35870, "total_steps": 46238, "tokens": 19394445312, "loss": 2.9774045817779773, "ce_loss": 2.9774045817779773, "z_loss": 34.202938571120754, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17872506380081177, "tokens_per_second": 4347995.207709595, "eta_seconds": 1289.1335790944072, "mean_data_wait_seconds": 5.695223808288574e-05}
{"step": 35880, "total_steps": 46238, "tokens": 19399852032, "loss": 2.926295858440977, "ce_loss": 2.926295858440977, "z_loss": 34.369226831378356, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19207042455673218, "tokens_per_second": 4348079.074291142, "eta_seconds": 1287.865240793238, "mean_data_wait_seconds": 6.29425048828125e-05}
{"step": 35890, "total_steps": 46238, "tokens": 19405258752, "loss": 2.9075862729188167, "ce_loss": 2.9075862729188167, "z_loss": 33.51492020578095, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16495242714881897, "tokens_per_second": 4348288.777849194, "eta_seconds": 1286.5597180431841, "mean_data_wait_seconds": 6.511807441711426e-05}
{"step": 35900, "total_steps": 46238, "tokens": 19410665472, "loss": 2.997662446715615, "ce_loss": 2.997662446715615, "z_loss": 33.30157499602347, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13953833281993866, "tokens_per_second": 4348432.689067463, "eta_seconds": 1285.2737672705161, "mean_data_wait_seconds": 8.982419967651367e-05}
{"step": 35910, "total_steps": 46238, "tokens": 19416072192, "loss": 2.9817207509821113, "ce_loss": 2.9817207509821113, "z_loss": 34.73175912914854, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1613125503063202, "tokens_per_second": 4348486.33097383, "eta_seconds": 1284.0145556464445, "mean_data_wait_seconds": 0.00012251734733581543}
{"step": 35920, "total_steps": 46238, "tokens": 19421478912, "loss": 3.012298531604536, "ce_loss": 3.012298531604536, "z_loss": 33.909167752121434, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1709161251783371, "tokens_per_second": 4348483.955191294, "eta_seconds": 1282.7718996963883, "mean_data_wait_seconds": 6.0051679611206055e-05}
{"step": 35930, "total_steps": 46238, "tokens": 19426885632, "loss": 2.958931296160727, "ce_loss": 2.958931296160727, "z_loss": 33.21794119748202, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17076504230499268, "tokens_per_second": 4348528.952954947, "eta_seconds": 1281.5152812109461, "mean_data_wait_seconds": 7.966160774230957e-05}
{"step": 35940, "total_steps": 46238, "tokens": 19432292352, "loss": 3.0409628333467427, "ce_loss": 3.0409628333467427, "z_loss": 33.94532339500658, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15606699883937836, "tokens_per_second": 4348630.149092952, "eta_seconds": 1280.2421436463712, "mean_data_wait_seconds": 7.104873657226562e-05}
{"step": 35950, "total_steps": 46238, "tokens": 19437699072, "loss": 3.0162435282360422, "ce_loss": 3.0162435282360422, "z_loss": 32.651232517126836, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1669214963912964, "tokens_per_second": 4348641.753478044, "eta_seconds": 1278.9954149595326, "mean_data_wait_seconds": 7.534027099609375e-05}
{"step": 35960, "total_steps": 46238, "tokens": 19443105792, "loss": 3.005918445009174, "ce_loss": 3.005918445009174, "z_loss": 34.63130127299916, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16929565370082855, "tokens_per_second": 4348739.485407056, "eta_seconds": 1277.7233868907865, "mean_data_wait_seconds": 6.875395774841309e-05}
{"step": 35970, "total_steps": 46238, "tokens": 19448512512, "loss": 2.9811572577014114, "ce_loss": 2.9811572577014114, "z_loss": 34.14696048967766, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1572190523147583, "tokens_per_second": 4348851.477908567, "eta_seconds": 1276.447230308634, "mean_data_wait_seconds": 7.25090503692627e-05}
{"step": 35980, "total_steps": 46238, "tokens": 19453919232, "loss": 2.9893157175092986, "ce_loss": 2.9893157175092986, "z_loss": 34.91471429304643, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15345916152000427, "tokens_per_second": 4348929.455491296, "eta_seconds": 1275.1811131352345, "mean_data_wait_seconds": 7.903575897216797e-05}
{"step": 35990, "total_steps": 46238, "tokens": 19459325952, "loss": 3.0708493355548745, "ce_loss": 3.0708493355548745, "z_loss": 33.93046867486203, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1572837382555008, "tokens_per_second": 4347849.335799089, "eta_seconds": 1274.2543625840149, "mean_data_wait_seconds": 7.751584053039551e-05}
{"step": 36000, "total_steps": 46238, "tokens": 19464732672, "loss": 3.0707017092993767, "ce_loss": 3.0707017092993767, "z_loss": 33.11650030540697, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15154998004436493, "tokens_per_second": 4347875.241610533, "eta_seconds": 1273.003238692237, "mean_data_wait_seconds": 9.34898853302002e-05}
{"step": 36010, "total_steps": 46238, "tokens": 19470139392, "loss": 3.0250534505555122, "ce_loss": 3.0250534505555122, "z_loss": 33.57790458563602, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16366708278656006, "tokens_per_second": 4348025.588399656, "eta_seconds": 1271.7157320215272, "mean_data_wait_seconds": 9.143352508544922e-05}
{"step": 36020, "total_steps": 46238, "tokens": 19475546112, "loss": 2.990126306360418, "ce_loss": 2.990126306360418, "z_loss": 33.06388268326268, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14842729270458221, "tokens_per_second": 4348114.387172952, "eta_seconds": 1270.4462974332218, "mean_data_wait_seconds": 7.107853889465332e-05}
{"step": 36030, "total_steps": 46238, "tokens": 19480952832, "loss": 3.024258391423659, "ce_loss": 3.024258391423659, "z_loss": 33.86310103445342, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19460846483707428, "tokens_per_second": 4347493.238048805, "eta_seconds": 1269.3841719410739, "mean_data_wait_seconds": 9.334087371826172e-05}
{"step": 36040, "total_steps": 46238, "tokens": 19486359552, "loss": 3.0240106781323752, "ce_loss": 3.0240106781323752, "z_loss": 32.944007873535156, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15632446110248566, "tokens_per_second": 4346983.6047224635, "eta_seconds": 1268.289205878428, "mean_data_wait_seconds": 9.113550186157227e-05}
{"step": 36050, "total_steps": 46238, "tokens": 19491766272, "loss": 3.052426898118221, "ce_loss": 3.052426898118221, "z_loss": 33.988498254255816, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16769708693027496, "tokens_per_second": 4346895.569001107, "eta_seconds": 1267.0710801699036, "mean_data_wait_seconds": 8.857250213623047e-05}
{"step": 36060, "total_steps": 46238, "tokens": 19497172992, "loss": 3.062637403155818, "ce_loss": 3.062637403155818, "z_loss": 33.36913634791519, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13569983839988708, "tokens_per_second": 4346919.956900549, "eta_seconds": 1265.8201665906329, "mean_data_wait_seconds": 7.459521293640137e-05}
{"step": 36070, "total_steps": 46238, "tokens": 19502579712, "loss": 3.0352022720105722, "ce_loss": 3.0352022720105722, "z_loss": 33.113099878484554, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1857300102710724, "tokens_per_second": 4346972.2474271925, "eta_seconds": 1264.5611499483284, "mean_data_wait_seconds": 9.530782699584961e-05}
{"step": 36080, "total_steps": 46238, "tokens": 19507986432, "loss": 3.071371264529951, "ce_loss": 3.071371264529951, "z_loss": 33.77243622866544, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.144353449344635, "tokens_per_second": 4346977.975649495, "eta_seconds": 1263.3156953548823, "mean_data_wait_seconds": 7.82310962677002e-05}
{"step": 36090, "total_steps": 46238, "tokens": 19513393152, "loss": 3.100156941197135, "ce_loss": 3.100156941197135, "z_loss": 33.33089533719149, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18329350650310516, "tokens_per_second": 4347004.350806587, "eta_seconds": 1262.0642495980098, "mean_data_wait_seconds": 0.00010040402412414551}
{"step": 36100, "total_steps": 46238, "tokens": 19518799872, "loss": 3.0565536419550576, "ce_loss": 3.0565536419550576, "z_loss": 33.62103768550988, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15386465191841125, "tokens_per_second": 4347032.880913877, "eta_seconds": 1260.8121940057129, "mean_data_wait_seconds": 0.0001278221607208252}
{"step": 36110, "total_steps": 46238, "tokens": 19524206592, "loss": 3.0429672400156655, "ce_loss": 3.0429672400156655, "z_loss": 34.4106238972057, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1708848625421524, "tokens_per_second": 4347056.90134051, "eta_seconds": 1259.561461528499, "mean_data_wait_seconds": 0.00010219216346740723}
{"step": 36120, "total_steps": 46238, "tokens": 19529613312, "loss": 3.0627736593737747, "ce_loss": 3.0627736593737747, "z_loss": 33.55329458641283, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1539405733346939, "tokens_per_second": 4347012.775755994, "eta_seconds": 1258.3304688007754, "mean_data_wait_seconds": 0.00016358494758605957}
{"step": 36130, "total_steps": 46238, "tokens": 19535020032, "loss": 3.008197833191265, "ce_loss": 3.008197833191265, "z_loss": 34.03083026770389, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16769134998321533, "tokens_per_second": 4347103.882306933, "eta_seconds": 1257.0603445298955, "mean_data_wait_seconds": 9.950995445251465e-05}
{"step": 36140, "total_steps": 46238, "tokens": 19540426752, "loss": 3.022874223463463, "ce_loss": 3.022874223463463, "z_loss": 33.07301128271854, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2048964500427246, "tokens_per_second": 4347168.56436479, "eta_seconds": 1255.7979068837178, "mean_data_wait_seconds": 0.0001310110092163086}
{"step": 36150, "total_steps": 46238, "tokens": 19545833472, "loss": 3.057927996823282, "ce_loss": 3.057927996823282, "z_loss": 32.9831026395162, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15575383603572845, "tokens_per_second": 4347270.258403058, "eta_seconds": 1254.524825885429, "mean_data_wait_seconds": 0.00016260147094726562}
{"step": 36160, "total_steps": 46238, "tokens": 19551240192, "loss": 3.004885079282703, "ce_loss": 3.004885079282703, "z_loss": 33.0366956392924, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1520383059978485, "tokens_per_second": 4347390.667496242, "eta_seconds": 1253.2464093312842, "mean_data_wait_seconds": 6.589293479919434e-05}
{"step": 36170, "total_steps": 46238, "tokens": 19556646912, "loss": 3.016544997692108, "ce_loss": 3.016544997692108, "z_loss": 33.36086209615072, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1495128870010376, "tokens_per_second": 4347436.99848143, "eta_seconds": 1251.9893964883754, "mean_data_wait_seconds": 9.128451347351074e-05}
{"step": 36180, "total_steps": 46238, "tokens": 19562053632, "loss": 3.0502977280905754, "ce_loss": 3.0502977280905754, "z_loss": 33.49886001240123, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14691118896007538, "tokens_per_second": 4347432.12748147, "eta_seconds": 1250.7471409680281, "mean_data_wait_seconds": 9.652972221374512e-05}
{"step": 36190, "total_steps": 46238, "tokens": 19567460352, "loss": 3.0185420260284888, "ce_loss": 3.0185420260284888, "z_loss": 33.23881660808217, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15424640476703644, "tokens_per_second": 4347417.220056092, "eta_seconds": 1249.5077672646087, "mean_data_wait_seconds": 0.00015616416931152344}
{"step": 36200, "total_steps": 46238, "tokens": 19572867072, "loss": 3.0206369269977915, "ce_loss": 3.0206369269977915, "z_loss": 33.32785768219919, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17407017946243286, "tokens_per_second": 4347359.3031817805, "eta_seconds": 1248.2807344744301, "mean_data_wait_seconds": 0.00012451410293579102}
{"step": 36210, "total_steps": 46238, "tokens": 19578273792, "loss": 3.008933455655069, "ce_loss": 3.008933455655069, "z_loss": 33.684073188088156, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15037934482097626, "tokens_per_second": 4347261.46485524, "eta_seconds": 1247.065120841662, "mean_data_wait_seconds": 0.00012114644050598145}
{"step": 36220, "total_steps": 46238, "tokens": 19583680512, "loss": 3.0620802640914917, "ce_loss": 3.0620802640914917, "z_loss": 34.23666404955315, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17060445249080658, "tokens_per_second": 4347240.311183168, "eta_seconds": 1245.8274758972266, "mean_data_wait_seconds": 0.00016239285469055176}
{"step": 36230, "total_steps": 46238, "tokens": 19589087232, "loss": 3.0256742130626333, "ce_loss": 3.0256742130626333, "z_loss": 34.08591744394013, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1429155021905899, "tokens_per_second": 4347294.800229481, "eta_seconds": 1244.5681631055697, "mean_data_wait_seconds": 9.575486183166504e-05}
{"step": 36240, "total_steps": 46238, "tokens": 19594493952, "loss": 3.0056635954163293, "ce_loss": 3.0056635954163293, "z_loss": 34.252161690683074, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16003811359405518, "tokens_per_second": 4347208.513289744, "eta_seconds": 1243.3491440486944, "mean_data_wait_seconds": 6.341934204101562e-05}
{"step": 36250, "total_steps": 46238, "tokens": 19599900672, "loss": 3.0213091572125754, "ce_loss": 3.0213091572125754, "z_loss": 33.74423116626161, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16624094545841217, "tokens_per_second": 4347227.933274812, "eta_seconds": 1242.0998730407855, "mean_data_wait_seconds": 7.843971252441406e-05}
{"step": 36260, "total_steps": 46238, "tokens": 19605307392, "loss": 3.060982751123833, "ce_loss": 3.060982751123833, "z_loss": 33.72216938481186, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17925594747066498, "tokens_per_second": 4347313.814778147, "eta_seconds": 1240.8316431316293, "mean_data_wait_seconds": 6.529688835144043e-05}
{"step": 36270, "total_steps": 46238, "tokens": 19610714112, "loss": 3.003153273553559, "ce_loss": 3.003153273553559, "z_loss": 34.51769640951446, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1491488367319107, "tokens_per_second": 4347417.359433846, "eta_seconds": 1239.5584270983777, "mean_data_wait_seconds": 5.6177377700805664e-05}
{"step": 36280, "total_steps": 46238, "tokens": 19616120832, "loss": 2.9538422389463945, "ce_loss": 2.9538422389463945, "z_loss": 34.31255360805627, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16366584599018097, "tokens_per_second": 4347392.226594982, "eta_seconds": 1238.3219234434039, "mean_data_wait_seconds": 7.167458534240723e-05}
{"step": 36290, "total_steps": 46238, "tokens": 19621527552, "loss": 2.947889642281966, "ce_loss": 2.947889642281966, "z_loss": 34.9866715633508, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16243751347064972, "tokens_per_second": 4347591.839219382, "eta_seconds": 1237.0214552996404, "mean_data_wait_seconds": 6.350874900817871e-05}
{"step": 36300, "total_steps": 46238, "tokens": 19626934272, "loss": 2.9547693422346404, "ce_loss": 2.9547693422346404, "z_loss": 33.33117782708371, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17066825926303864, "tokens_per_second": 4347669.286428626, "eta_seconds": 1235.7558291682637, "mean_data_wait_seconds": 6.717443466186523e-05}
{"step": 36310, "total_steps": 46238, "tokens": 19632340992, "loss": 2.9573465043848213, "ce_loss": 2.9573465043848213, "z_loss": 34.42823976458925, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16700604557991028, "tokens_per_second": 4347728.274139456, "eta_seconds": 1234.4954895007409, "mean_data_wait_seconds": 7.387995719909668e-05}
{"step": 36320, "total_steps": 46238, "tokens": 19637747712, "loss": 2.8969638781114058, "ce_loss": 2.8969638781114058, "z_loss": 35.13758165186102, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16028504073619843, "tokens_per_second": 4347764.784947596, "eta_seconds": 1233.2415595625712, "mean_data_wait_seconds": 9.262561798095703e-05}
{"step": 36330, "total_steps": 46238, "tokens": 19643154432, "loss": 2.92101765943296, "ce_loss": 2.92101765943296, "z_loss": 33.4808826157541, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16711023449897766, "tokens_per_second": 4347817.600264668, "eta_seconds": 1231.9830306758804, "mean_data_wait_seconds": 7.614493370056152e-05}
{"step": 36340, "total_steps": 46238, "tokens": 19648561152, "loss": 2.86556191516645, "ce_loss": 2.86556191516645, "z_loss": 34.730665842692055, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14541611075401306, "tokens_per_second": 4347883.703102132, "eta_seconds": 1230.7207711609537, "mean_data_wait_seconds": 8.460879325866699e-05}
{"step": 36350, "total_steps": 46238, "tokens": 19653967872, "loss": 2.934261918067932, "ce_loss": 2.934261918067932, "z_loss": 33.94655114954168, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1858111321926117, "tokens_per_second": 4347912.588958801, "eta_seconds": 1229.4690738665752, "mean_data_wait_seconds": 7.933378219604492e-05}
{"step": 36360, "total_steps": 46238, "tokens": 19659374592, "loss": 2.8734726797450674, "ce_loss": 2.8734726797450674, "z_loss": 33.896157495903246, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17821356654167175, "tokens_per_second": 4347888.296385851, "eta_seconds": 1228.2324153633422, "mean_data_wait_seconds": 7.334351539611816e-05}
{"step": 36370, "total_steps": 46238, "tokens": 19664781312, "loss": 2.88601855617581, "ce_loss": 2.88601855617581, "z_loss": 34.06337171612364, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15970848500728607, "tokens_per_second": 4347926.805075894, "eta_seconds": 1226.9780203686937, "mean_data_wait_seconds": 6.398558616638184e-05}
{"step": 36380, "total_steps": 46238, "tokens": 19670188032, "loss": 2.904839179732583, "ce_loss": 2.904839179732583, "z_loss": 33.19342705697724, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.23939140141010284, "tokens_per_second": 4348016.104850009, "eta_seconds": 1225.709329377897, "mean_data_wait_seconds": 7.358193397521973e-05}
{"step": 36390, "total_steps": 46238, "tokens": 19675594752, "loss": 2.871799788691781, "ce_loss": 2.871799788691781, "z_loss": 35.76479643041437, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17229881882667542, "tokens_per_second": 4348125.569914222, "eta_seconds": 1224.4350119136577, "mean_data_wait_seconds": 6.839632987976074e-05}
{"step": 36400, "total_steps": 46238, "tokens": 19681001472, "loss": 2.874602081197681, "ce_loss": 2.874602081197681, "z_loss": 33.877280697678074, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.144016832113266, "tokens_per_second": 4348248.714321285, "eta_seconds": 1223.1569106161803, "mean_data_wait_seconds": 5.7637691497802734e-05}
{"step": 36410, "total_steps": 46238, "tokens": 19686408192, "loss": 2.894210513794061, "ce_loss": 2.894210513794061, "z_loss": 35.49774019645922, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15611472725868225, "tokens_per_second": 4348369.098888391, "eta_seconds": 1221.8796572163694, "mean_data_wait_seconds": 5.930662155151367e-05}
{"step": 36420, "total_steps": 46238, "tokens": 19691814912, "loss": 2.8936059655565205, "ce_loss": 2.8936059655565205, "z_loss": 33.021701870542586, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15863603353500366, "tokens_per_second": 4348530.423639259, "eta_seconds": 1220.5909829091072, "mean_data_wait_seconds": 7.423758506774902e-05}
{"step": 36430, "total_steps": 46238, "tokens": 19697221632, "loss": 2.9257418459111992, "ce_loss": 2.9257418459111992, "z_loss": 34.0529980803981, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.22634410858154297, "tokens_per_second": 4348670.271478428, "eta_seconds": 1219.3084260208445, "mean_data_wait_seconds": 6.699562072753906e-05}
{"step": 36440, "total_steps": 46238, "tokens": 19702628352, "loss": 2.954919650699153, "ce_loss": 2.954919650699153, "z_loss": 34.352510741262726, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14659683406352997, "tokens_per_second": 4348781.030883932, "eta_seconds": 1218.0340988388052, "mean_data_wait_seconds": 6.338953971862793e-05}
{"step": 36450, "total_steps": 46238, "tokens": 19708035072, "loss": 2.9154871973124417, "ce_loss": 2.9154871973124417, "z_loss": 34.868663585547246, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1746084988117218, "tokens_per_second": 4348916.332550597, "eta_seconds": 1216.7529700201321, "mean_data_wait_seconds": 0.00011494755744934082}
{"step": 36460, "total_steps": 46238, "tokens": 19713441792, "loss": 2.9854181788184424, "ce_loss": 2.9854181788184424, "z_loss": 33.70570114887122, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15990789234638214, "tokens_per_second": 4348859.18606159, "eta_seconds": 1215.525708659985, "mean_data_wait_seconds": 6.592273712158203e-05}
{"step": 36470, "total_steps": 46238, "tokens": 19718848512, "loss": 2.9828896648956067, "ce_loss": 2.9828896648956067, "z_loss": 34.166157520178594, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15130141377449036, "tokens_per_second": 4348835.881563657, "eta_seconds": 1214.2889655567476, "mean_data_wait_seconds": 8.338689804077148e-05}
{"step": 36480, "total_steps": 46238, "tokens": 19724255232, "loss": 2.9678719061793704, "ce_loss": 2.9678719061793704, "z_loss": 34.05447720036362, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.157487154006958, "tokens_per_second": 4348973.507247558, "eta_seconds": 1213.0073212008901, "mean_data_wait_seconds": 9.85562801361084e-05}
{"step": 36490, "total_steps": 46238, "tokens": 19729661952, "loss": 2.9670105341709023, "ce_loss": 2.9670105341709023, "z_loss": 33.39324139103745, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.11187849193811417, "tokens_per_second": 4348932.371579681, "eta_seconds": 1211.7755655247822, "mean_data_wait_seconds": 0.0001303553581237793}
{"step": 36500, "total_steps": 46238, "tokens": 19735068672, "loss": 3.015618510318525, "ce_loss": 3.015618510318525, "z_loss": 33.21327501354796, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1740282028913498, "tokens_per_second": 4348262.906738372, "eta_seconds": 1210.7187115668023, "mean_data_wait_seconds": 0.00014722347259521484}
{"step": 36510, "total_steps": 46238, "tokens": 19740475392, "loss": 2.9438236843455923, "ce_loss": 2.9438236843455923, "z_loss": 34.57524981643214, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12984204292297363, "tokens_per_second": 4348112.957104767, "eta_seconds": 1209.5170010260804, "mean_data_wait_seconds": 0.0001049041748046875}
{"step": 36520, "total_steps": 46238, "tokens": 19745882112, "loss": 2.9620653228326277, "ce_loss": 2.9620653228326277, "z_loss": 34.038379293499574, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1630341112613678, "tokens_per_second": 4347394.935241264, "eta_seconds": 1208.4730976272435, "mean_data_wait_seconds": 0.00019189715385437012}
{"step": 36530, "total_steps": 46238, "tokens": 19751288832, "loss": 2.961545610066616, "ce_loss": 2.961545610066616, "z_loss": 34.91567802429199, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1439785361289978, "tokens_per_second": 4347292.710188832, "eta_seconds": 1207.2578162725167, "mean_data_wait_seconds": 7.191300392150879e-05}
{"step": 36540, "total_steps": 46238, "tokens": 19756695552, "loss": 2.955764788569826, "ce_loss": 2.955764788569826, "z_loss": 33.66660820354115, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18047092854976654, "tokens_per_second": 4346682.058189412, "eta_seconds": 1206.1835473156052, "mean_data_wait_seconds": 0.00017112493515014648}
{"step": 36550, "total_steps": 46238, "tokens": 19762102272, "loss": 3.024787982304891, "ce_loss": 3.024787982304891, "z_loss": 34.57317164449981, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15223665535449982, "tokens_per_second": 4346610.807709068, "eta_seconds": 1204.9594260224278, "mean_data_wait_seconds": 0.00013139843940734863}
{"step": 36560, "total_steps": 46238, "tokens": 19767508992, "loss": 3.0045472907297537, "ce_loss": 3.0045472907297537, "z_loss": 34.436448097229004, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14814382791519165, "tokens_per_second": 4346570.454029722, "eta_seconds": 1203.7267080645881, "mean_data_wait_seconds": 8.577108383178711e-05}
{"step": 36570, "total_steps": 46238, "tokens": 19772915712, "loss": 2.9843814770380654, "ce_loss": 2.9843814770380654, "z_loss": 34.7954437082464, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14685966074466705, "tokens_per_second": 4346543.84022111, "eta_seconds": 1202.4901660106384, "mean_data_wait_seconds": 7.534027099609375e-05}
{"step": 36580, "total_steps": 46238, "tokens": 19778322432, "loss": 3.006584983883482, "ce_loss": 3.006584983883482, "z_loss": 33.946670243234344, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16155581176280975, "tokens_per_second": 4346646.790567976, "eta_seconds": 1201.2178020376339, "mean_data_wait_seconds": 0.00014504790306091309}
{"step": 36590, "total_steps": 46238, "tokens": 19783729152, "loss": 2.9835120981389824, "ce_loss": 2.9835120981389824, "z_loss": 33.978692863926746, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14584945142269135, "tokens_per_second": 4346529.892956516, "eta_seconds": 1200.0061917099029, "mean_data_wait_seconds": 0.00015792250633239746}
{"step": 36600, "total_steps": 46238, "tokens": 19789135872, "loss": 3.0125863046357124, "ce_loss": 3.0125863046357124, "z_loss": 34.26568080439712, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17960289120674133, "tokens_per_second": 4346437.435851612, "eta_seconds": 1198.787775252791, "mean_data_wait_seconds": 0.00016680359840393066}
{"step": 36610, "total_steps": 46238, "tokens": 19794542592, "loss": 3.0355164589303913, "ce_loss": 3.0355164589303913, "z_loss": 32.81326487570098, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18336988985538483, "tokens_per_second": 4346406.664135267, "eta_seconds": 1197.5523107282836, "mean_data_wait_seconds": 0.00034546852111816406}
{"step": 36620, "total_steps": 46238, "tokens": 19799949312, "loss": 3.0280905239509814, "ce_loss": 3.0280905239509814, "z_loss": 34.05144561420787, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17614245414733887, "tokens_per_second": 4346474.047987619, "eta_seconds": 1196.289812522265, "mean_data_wait_seconds": 8.419156074523926e-05}
{"step": 36630, "total_steps": 46238, "tokens": 19805356032, "loss": 3.043449965390292, "ce_loss": 3.043449965390292, "z_loss": 33.99519038922859, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17902633547782898, "tokens_per_second": 4346436.306058093, "eta_seconds": 1195.0562571825194, "mean_data_wait_seconds": 7.203221321105957e-05}
{"step": 36640, "total_steps": 46238, "tokens": 19810762752, "loss": 3.022813679593982, "ce_loss": 3.022813679593982, "z_loss": 33.73326547218092, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15496748685836792, "tokens_per_second": 4346521.669973941, "eta_seconds": 1193.7888679687887, "mean_data_wait_seconds": 6.92605972290039e-05}
{"step": 36650, "total_steps": 46238, "tokens": 19816169472, "loss": 3.016947146618005, "ce_loss": 3.016947146618005, "z_loss": 33.648341496785484, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15746675431728363, "tokens_per_second": 4346586.922112984, "eta_seconds": 1192.5270463658896, "mean_data_wait_seconds": 8.255243301391602e-05}
{"step": 36660, "total_steps": 46238, "tokens": 19821576192, "loss": 3.0321812720009773, "ce_loss": 3.0321812720009773, "z_loss": 33.91341379917029, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1577710509300232, "tokens_per_second": 4346642.869574595, "eta_seconds": 1191.2678127399895, "mean_data_wait_seconds": 9.453296661376953e-05}
{"step": 36670, "total_steps": 46238, "tokens": 19826982912, "loss": 3.0591649051868552, "ce_loss": 3.0591649051868552, "z_loss": 33.26958011858391, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1476181149482727, "tokens_per_second": 4346651.63396756, "eta_seconds": 1190.0215291186146, "mean_data_wait_seconds": 0.00010409951210021973}
{"step": 36680, "total_steps": 46238, "tokens": 19832389632, "loss": 3.069662392139435, "ce_loss": 3.069662392139435, "z_loss": 34.023876768169984, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17367765307426453, "tokens_per_second": 4346686.464327308, "eta_seconds": 1188.7681217420577, "mean_data_wait_seconds": 7.414817810058594e-05}
{"step": 36690, "total_steps": 46238, "tokens": 19837796352, "loss": 3.0216867977922615, "ce_loss": 3.0216867977922615, "z_loss": 33.77570738936915, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1410892903804779, "tokens_per_second": 4346733.926898284, "eta_seconds": 1187.5112833702528, "mean_data_wait_seconds": 7.43865966796875e-05}
{"step": 36700, "total_steps": 46238, "tokens": 19843203072, "loss": 3.018915748957432, "ce_loss": 3.018915748957432, "z_loss": 33.62399942224676, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1890457719564438, "tokens_per_second": 4346700.458981345, "eta_seconds": 1186.276559118685, "mean_data_wait_seconds": 0.00018039345741271973}
{"step": 36710, "total_steps": 46238, "tokens": 19848609792, "loss": 3.021064973238743, "ce_loss": 3.021064973238743, "z_loss": 33.58091721390233, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15961900353431702, "tokens_per_second": 4346821.527479301, "eta_seconds": 1184.9996857329054, "mean_data_wait_seconds": 0.00013047456741333008}
{"step": 36720, "total_steps": 46238, "tokens": 19854016512, "loss": 3.0337765180703364, "ce_loss": 3.0337765180703364, "z_loss": 33.81413974183978, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13906580209732056, "tokens_per_second": 4346908.340337963, "eta_seconds": 1183.7322117539618, "mean_data_wait_seconds": 0.00013500452041625977}
{"step": 36730, "total_steps": 46238, "tokens": 19859423232, "loss": 2.9433532826828235, "ce_loss": 2.9433532826828235, "z_loss": 34.272414872140594, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14799676835536957, "tokens_per_second": 4346953.107037714, "eta_seconds": 1182.476225859918, "mean_data_wait_seconds": 0.00010657310485839844}
{"step": 36740, "total_steps": 46238, "tokens": 19864829952, "loss": 3.026519871119297, "ce_loss": 3.026519871119297, "z_loss": 33.67072197885224, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1520403027534485, "tokens_per_second": 4347002.806227036, "eta_seconds": 1181.2189255190974, "mean_data_wait_seconds": 0.00014287233352661133}
{"step": 36750, "total_steps": 46238, "tokens": 19870236672, "loss": 3.0901503219748987, "ce_loss": 3.0901503219748987, "z_loss": 34.43967851003011, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1559799164533615, "tokens_per_second": 4347038.340314752, "eta_seconds": 1179.9654989076087, "mean_data_wait_seconds": 0.00011005997657775879}
{"step": 36760, "total_steps": 46238, "tokens": 19875643392, "loss": 2.97517164367618, "ce_loss": 2.97517164367618, "z_loss": 33.76157728830973, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17205247282981873, "tokens_per_second": 4347207.0658856705, "eta_seconds": 1178.6759789313328, "mean_data_wait_seconds": 9.819865226745605e-05}
{"step": 36770, "total_steps": 46238, "tokens": 19881050112, "loss": 2.932528083974665, "ce_loss": 2.932528083974665, "z_loss": 34.70801521070076, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13699746131896973, "tokens_per_second": 4347278.490390852, "eta_seconds": 1177.4129113913302, "mean_data_wait_seconds": 0.00012600421905517578}
{"step": 36780, "total_steps": 46238, "tokens": 19886456832, "loss": 2.937737240935817, "ce_loss": 2.937737240935817, "z_loss": 33.27104397976037, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.7153345942497253, "tokens_per_second": 4347340.082460076, "eta_seconds": 1176.152545467889, "mean_data_wait_seconds": 0.00011157989501953125}
{"step": 36790, "total_steps": 46238, "tokens": 19891863552, "loss": 3.021569974494703, "ce_loss": 3.021569974494703, "z_loss": 33.61914776310776, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17127929627895355, "tokens_per_second": 4347319.977157581, "eta_seconds": 1174.9142945165952, "mean_data_wait_seconds": 7.271766662597656e-05}
{"step": 36800, "total_steps": 46238, "tokens": 19897270272, "loss": 3.039400521552924, "ce_loss": 3.039400521552924, "z_loss": 33.89540602944114, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16608783602714539, "tokens_per_second": 4347406.056765582, "eta_seconds": 1173.6473652051875, "mean_data_wait_seconds": 6.008148193359375e-05}
{"step": 36810, "total_steps": 46238, "tokens": 19902676992, "loss": 3.007948485287753, "ce_loss": 3.007948485287753, "z_loss": 33.8316924644239, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14198170602321625, "tokens_per_second": 4347433.505387571, "eta_seconds": 1172.3962971908902, "mean_data_wait_seconds": 6.300210952758789e-05}
{"step": 36820, "total_steps": 46238, "tokens": 19908083712, "loss": 3.0524760340199326, "ce_loss": 3.0524760340199326, "z_loss": 34.005260640924625, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1837378591299057, "tokens_per_second": 4347529.918228133, "eta_seconds": 1171.1266672720405, "mean_data_wait_seconds": 6.312131881713867e-05}
{"step": 36830, "total_steps": 46238, "tokens": 19913490432, "loss": 3.012709987886024, "ce_loss": 3.012709987886024, "z_loss": 33.49997517556855, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15415889024734497, "tokens_per_second": 4347548.299150399, "eta_seconds": 1169.8780908297051, "mean_data_wait_seconds": 0.00047665834426879883}
{"step": 36840, "total_steps": 46238, "tokens": 19918897152, "loss": 3.033885890787298, "ce_loss": 3.033885890787298, "z_loss": 33.52151740681041, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18002170324325562, "tokens_per_second": 4347564.891559836, "eta_seconds": 1168.630005698921, "mean_data_wait_seconds": 0.00013965368270874023}
{"step": 36850, "total_steps": 46238, "tokens": 19924303872, "loss": 3.02828772501512, "ce_loss": 3.02828772501512, "z_loss": 34.1381981011593, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15919332206249237, "tokens_per_second": 4347625.570955419, "eta_seconds": 1167.370092288024, "mean_data_wait_seconds": 7.745623588562012e-05}
{"step": 36860, "total_steps": 46238, "tokens": 19929710592, "loss": 3.015958244150335, "ce_loss": 3.015958244150335, "z_loss": 33.00934615279689, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16083957254886627, "tokens_per_second": 4347715.61083548, "eta_seconds": 1166.10233920653, "mean_data_wait_seconds": 7.203221321105957e-05}
{"step": 36870, "total_steps": 46238, "tokens": 19935117312, "loss": 2.974355683182225, "ce_loss": 2.974355683182225, "z_loss": 34.295245719678476, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14707933366298676, "tokens_per_second": 4347742.678815071, "eta_seconds": 1164.8515098828864, "mean_data_wait_seconds": 8.040666580200195e-05}
{"step": 36880, "total_steps": 46238, "tokens": 19940524032, "loss": 2.9595674294413943, "ce_loss": 2.9595674294413943, "z_loss": 34.17342200423732, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1788129210472107, "tokens_per_second": 4347882.4606361985, "eta_seconds": 1163.5705311269473, "mean_data_wait_seconds": 9.936094284057617e-05}
{"step": 36890, "total_steps": 46238, "tokens": 19945930752, "loss": 2.934329428456046, "ce_loss": 2.934329428456046, "z_loss": 35.892651268930145, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14992108941078186, "tokens_per_second": 4347935.11648915, "eta_seconds": 1162.3129252399485, "mean_data_wait_seconds": 8.404254913330078e-05}
{"step": 36900, "total_steps": 46238, "tokens": 19951337472, "loss": 2.8472586021278845, "ce_loss": 2.8472586021278845, "z_loss": 34.56449014490301, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1536976546049118, "tokens_per_second": 4347998.203376974, "eta_seconds": 1161.0525643913918, "mean_data_wait_seconds": 9.1552734375e-05}
{"step": 36910, "total_steps": 46238, "tokens": 19956744192, "loss": 2.8936349894061233, "ce_loss": 2.8936349894061233, "z_loss": 34.40231901226622, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1524553745985031, "tokens_per_second": 4347983.833152155, "eta_seconds": 1159.8129012232525, "mean_data_wait_seconds": 9.045004844665527e-05}
{"step": 36920, "total_steps": 46238, "tokens": 19962150912, "loss": 2.936009269772154, "ce_loss": 2.936009269772154, "z_loss": 34.219360409360945, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20568899810314178, "tokens_per_second": 4348038.573517554, "eta_seconds": 1158.5548147344794, "mean_data_wait_seconds": 0.0001672208309173584}
{"step": 36930, "total_steps": 46238, "tokens": 19967557632, "loss": 2.95507690039548, "ce_loss": 2.95507690039548, "z_loss": 35.35257530212402, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17813070118427277, "tokens_per_second": 4348071.132060932, "eta_seconds": 1157.3026639090142, "mean_data_wait_seconds": 0.00010132789611816406}
{"step": 36940, "total_steps": 46238, "tokens": 19972964352, "loss": 2.8267335891723633, "ce_loss": 2.8267335891723633, "z_loss": 35.180346171061196, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1447352170944214, "tokens_per_second": 4347997.741261288, "eta_seconds": 1156.0787017662647, "mean_data_wait_seconds": 0.00030726194381713867}
{"step": 36950, "total_steps": 46238, "tokens": 19978371072, "loss": 2.8529679016633467, "ce_loss": 2.8529679016633467, "z_loss": 36.06008093284838, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1385454684495926, "tokens_per_second": 4347953.384133862, "eta_seconds": 1154.8469867048163, "mean_data_wait_seconds": 0.0002789497375488281}
{"step": 36960, "total_steps": 46238, "tokens": 19983777792, "loss": 2.918585426879652, "ce_loss": 2.918585426879652, "z_loss": 34.39842510223389, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14235204458236694, "tokens_per_second": 4347915.816767487, "eta_seconds": 1153.6134450112402, "mean_data_wait_seconds": 0.00014442205429077148}
{"step": 36970, "total_steps": 46238, "tokens": 19989184512, "loss": 2.9286495374910757, "ce_loss": 2.9286495374910757, "z_loss": 32.788941267764926, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1958932876586914, "tokens_per_second": 4347827.580189693, "eta_seconds": 1152.3933117378585, "mean_data_wait_seconds": 0.00015437602996826172}
{"step": 36980, "total_steps": 46238, "tokens": 19994591232, "loss": 2.8667338653044268, "ce_loss": 2.8667338653044268, "z_loss": 35.28340455257531, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17706885933876038, "tokens_per_second": 4347819.941862916, "eta_seconds": 1151.1517889251645, "mean_data_wait_seconds": 0.00010627508163452148}
{"step": 36990, "total_steps": 46238, "tokens": 19999997952, "loss": 2.856381578878923, "ce_loss": 2.856381578878923, "z_loss": 34.51890563964844, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16600674390792847, "tokens_per_second": 4347929.849595533, "eta_seconds": 1149.8791739855435, "mean_data_wait_seconds": 0.00010928511619567871}
{"step": 37000, "total_steps": 46238, "tokens": 20005404672, "loss": 2.914353329123873, "ce_loss": 2.914353329123873, "z_loss": 35.66644654129491, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1979162096977234, "tokens_per_second": 4347916.08274693, "eta_seconds": 1148.639294998713, "mean_data_wait_seconds": 0.0001188814640045166}
{"step": 37010, "total_steps": 46238, "tokens": 20010811392, "loss": 2.932558196963686, "ce_loss": 2.932558196963686, "z_loss": 34.37597982811205, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1571289449930191, "tokens_per_second": 4347821.098842006, "eta_seconds": 1147.4208415173077, "mean_data_wait_seconds": 0.00017276406288146973}
{"step": 37020, "total_steps": 46238, "tokens": 20016218112, "loss": 2.980086973219207, "ce_loss": 2.980086973219207, "z_loss": 34.2564750440193, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1743788719177246, "tokens_per_second": 4347903.9610752575, "eta_seconds": 1146.1554506754992, "mean_data_wait_seconds": 0.00015181303024291992}
{"step": 37030, "total_steps": 46238, "tokens": 20021624832, "loss": 2.951400910363053, "ce_loss": 2.951400910363053, "z_loss": 34.631649508620754, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16553638875484467, "tokens_per_second": 4347921.679539761, "eta_seconds": 1144.9072616521767, "mean_data_wait_seconds": 0.00010833144187927246}
{"step": 37040, "total_steps": 46238, "tokens": 20027031552, "loss": 2.9994168570547393, "ce_loss": 2.9994168570547393, "z_loss": 33.935626954743356, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17178551852703094, "tokens_per_second": 4347989.340545961, "eta_seconds": 1143.6459463296694, "mean_data_wait_seconds": 0.0001462996006011963}
{"step": 37050, "total_steps": 46238, "tokens": 20032438272, "loss": 3.0316574537392817, "ce_loss": 3.0316574537392817, "z_loss": 33.766683433995105, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15333960950374603, "tokens_per_second": 4347876.0167660685, "eta_seconds": 1142.4322231926355, "mean_data_wait_seconds": 7.447600364685059e-05}
{"step": 37060, "total_steps": 46238, "tokens": 20037844992, "loss": 3.03025685115294, "ce_loss": 3.03025685115294, "z_loss": 33.580196900801226, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19707205891609192, "tokens_per_second": 4347895.801378968, "eta_seconds": 1141.1834990218358, "mean_data_wait_seconds": 0.00014314055442810059}
{"step": 37070, "total_steps": 46238, "tokens": 20043251712, "loss": 2.974452845978014, "ce_loss": 2.974452845978014, "z_loss": 34.101737629283555, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16947731375694275, "tokens_per_second": 4347884.362679581, "eta_seconds": 1139.9429723897786, "mean_data_wait_seconds": 8.150935173034668e-05}
{"step": 37080, "total_steps": 46238, "tokens": 20048658432, "loss": 2.98537041201736, "ce_loss": 2.98537041201736, "z_loss": 35.19144503275553, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16870634257793427, "tokens_per_second": 4347843.827206806, "eta_seconds": 1138.7100596896641, "mean_data_wait_seconds": 7.581710815429688e-05}
{"step": 37090, "total_steps": 46238, "tokens": 20054065152, "loss": 2.9821742140885554, "ce_loss": 2.9821742140885554, "z_loss": 33.962857130802035, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16223672032356262, "tokens_per_second": 4347897.600254334, "eta_seconds": 1137.4524514355414, "mean_data_wait_seconds": 0.00011596083641052246}
{"step": 37100, "total_steps": 46238, "tokens": 20059471872, "loss": 2.941637272184545, "ce_loss": 2.941637272184545, "z_loss": 34.67325173002301, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16296327114105225, "tokens_per_second": 4347961.84198755, "eta_seconds": 1136.1921386462216, "mean_data_wait_seconds": 0.00011289119720458984}
{"step": 37110, "total_steps": 46238, "tokens": 20064878592, "loss": 2.997081109971711, "ce_loss": 2.997081109971711, "z_loss": 33.754663958693996, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15852266550064087, "tokens_per_second": 4347976.640332673, "eta_seconds": 1134.9447690736524, "mean_data_wait_seconds": 0.00010830163955688477}
{"step": 37120, "total_steps": 46238, "tokens": 20070285312, "loss": 3.0292183630394214, "ce_loss": 3.0292183630394214, "z_loss": 34.66416306929155, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15884104371070862, "tokens_per_second": 4348037.2358685965, "eta_seconds": 1133.685466935815, "mean_data_wait_seconds": 0.00015240907669067383}
{"step": 37130, "total_steps": 46238, "tokens": 20075692032, "loss": 3.018047388755914, "ce_loss": 3.018047388755914, "z_loss": 35.44282184947621, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16279719769954681, "tokens_per_second": 4348098.706419182, "eta_seconds": 1132.425972007202, "mean_data_wait_seconds": 7.888674736022949e-05}
{"step": 37140, "total_steps": 46238, "tokens": 20081098752, "loss": 3.0168624989914172, "ce_loss": 3.0168624989914172, "z_loss": 35.051031748453774, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16024594008922577, "tokens_per_second": 4348180.360371023, "eta_seconds": 1131.1612620365897, "mean_data_wait_seconds": 7.471442222595215e-05}
{"step": 37150, "total_steps": 46238, "tokens": 20086505472, "loss": 2.98683855750344, "ce_loss": 2.98683855750344, "z_loss": 34.239749532757386, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17545129358768463, "tokens_per_second": 4348163.2489086855, "eta_seconds": 1129.922264356818, "mean_data_wait_seconds": 0.00010776519775390625}
{"step": 37160, "total_steps": 46238, "tokens": 20091912192, "loss": 3.000398541941787, "ce_loss": 3.000398541941787, "z_loss": 33.56784112525709, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14861315488815308, "tokens_per_second": 4348267.686251113, "eta_seconds": 1128.6517064066006, "mean_data_wait_seconds": 7.233023643493652e-05}
{"step": 37170, "total_steps": 46238, "tokens": 20097318912, "loss": 3.0041752168626497, "ce_loss": 3.0041752168626497, "z_loss": 33.427810090960875, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14686496555805206, "tokens_per_second": 4348306.149252401, "eta_seconds": 1127.398314592647, "mean_data_wait_seconds": 8.055567741394043e-05}
{"step": 37180, "total_steps": 46238, "tokens": 20102725632, "loss": 3.0654919960282068, "ce_loss": 3.0654919960282068, "z_loss": 35.13986746470133, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18972265720367432, "tokens_per_second": 4348392.570006422, "eta_seconds": 1126.1325248729252, "mean_data_wait_seconds": 5.8025121688842773e-05}
{"step": 37190, "total_steps": 46238, "tokens": 20108132352, "loss": 3.0323030695770727, "ce_loss": 3.0323030695770727, "z_loss": 34.45101685957475, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16049373149871826, "tokens_per_second": 4348505.6111033615, "eta_seconds": 1124.8598993434143, "mean_data_wait_seconds": 8.806586265563965e-05}
{"step": 37200, "total_steps": 46238, "tokens": 20113539072, "loss": 3.03257428335421, "ce_loss": 3.03257428335421, "z_loss": 34.6655851132942, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16460634768009186, "tokens_per_second": 4348468.479387635, "eta_seconds": 1123.6261426662265, "mean_data_wait_seconds": 0.00010588765144348145}
{"step": 37210, "total_steps": 46238, "tokens": 20118945792, "loss": 2.992737703251116, "ce_loss": 2.992737703251116, "z_loss": 35.130617748607285, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13897405564785004, "tokens_per_second": 4348431.032593401, "eta_seconds": 1122.3924462449588, "mean_data_wait_seconds": 9.271502494812012e-05}
{"step": 37220, "total_steps": 46238, "tokens": 20124352512, "loss": 3.0088407559828325, "ce_loss": 3.0088407559828325, "z_loss": 34.831143552606754, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16309620440006256, "tokens_per_second": 4348473.202567328, "eta_seconds": 1121.1382011326805, "mean_data_wait_seconds": 0.00010523200035095215}
{"step": 37230, "total_steps": 46238, "tokens": 20129759232, "loss": 3.058666032372099, "ce_loss": 3.058666032372099, "z_loss": 35.182292100154996, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15252473950386047, "tokens_per_second": 4348543.249973316, "eta_seconds": 1119.8768010482322, "mean_data_wait_seconds": 9.796023368835449e-05}
{"step": 37240, "total_steps": 46238, "tokens": 20135165952, "loss": 3.0391535379669885, "ce_loss": 3.0391535379669885, "z_loss": 35.12213452657064, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16928160190582275, "tokens_per_second": 4348711.11379326, "eta_seconds": 1118.5902803639895, "mean_data_wait_seconds": 0.00016137957572937012}
{"step": 37250, "total_steps": 46238, "tokens": 20140572672, "loss": 2.998922299255024, "ce_loss": 2.998922299255024, "z_loss": 35.25311481591427, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16485707461833954, "tokens_per_second": 4348876.639882526, "eta_seconds": 1117.3044596020673, "mean_data_wait_seconds": 9.384751319885254e-05}
{"step": 37260, "total_steps": 46238, "tokens": 20145979392, "loss": 3.067137710975878, "ce_loss": 3.067137710975878, "z_loss": 34.23377444527366, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17863549292087555, "tokens_per_second": 4348867.980233343, "eta_seconds": 1116.0634367520106, "mean_data_wait_seconds": 6.890296936035156e-05}
{"step": 37270, "total_steps": 46238, "tokens": 20151386112, "loss": 3.025937739646796, "ce_loss": 3.025937739646796, "z_loss": 34.20694411884654, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15273521840572357, "tokens_per_second": 4348969.231993128, "eta_seconds": 1114.7942340760303, "mean_data_wait_seconds": 0.0001379549503326416}
{"step": 37280, "total_steps": 46238, "tokens": 20156792832, "loss": 3.000700751940409, "ce_loss": 3.000700751940409, "z_loss": 34.13391101721561, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16246066987514496, "tokens_per_second": 4348949.969329812, "eta_seconds": 1113.5559475627383, "mean_data_wait_seconds": 9.417533874511719e-05}
{"step": 37290, "total_steps": 46238, "tokens": 20162199552, "loss": 3.02041938088157, "ce_loss": 3.02041938088157, "z_loss": 34.91642058979381, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1487126350402832, "tokens_per_second": 4348967.758471904, "eta_seconds": 1112.3081734916593, "mean_data_wait_seconds": 0.0001538693904876709}
{"step": 37300, "total_steps": 46238, "tokens": 20167606272, "loss": 3.019644417546012, "ce_loss": 3.019644417546012, "z_loss": 34.9062697092692, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16382476687431335, "tokens_per_second": 4349070.865453702, "eta_seconds": 1111.038613415631, "mean_data_wait_seconds": 0.00012451410293579102}
{"step": 37310, "total_steps": 46238, "tokens": 20173012992, "loss": 2.9918573426477835, "ce_loss": 2.9918573426477835, "z_loss": 34.51128844058875, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14698511362075806, "tokens_per_second": 4349179.012163795, "eta_seconds": 1109.767827560331, "mean_data_wait_seconds": 9.796023368835449e-05}
{"step": 37320, "total_steps": 46238, "tokens": 20178419712, "loss": 3.040077030658722, "ce_loss": 3.040077030658722, "z_loss": 34.553713451732285, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14317680895328522, "tokens_per_second": 4349342.864591729, "eta_seconds": 1108.4829074409063, "mean_data_wait_seconds": 0.00011089444160461426}
{"step": 37330, "total_steps": 46238, "tokens": 20183826432, "loss": 2.98542417901935, "ce_loss": 2.98542417901935, "z_loss": 34.20086291342071, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13802237808704376, "tokens_per_second": 4349461.1742242845, "eta_seconds": 1107.2096774973236, "mean_data_wait_seconds": 7.37309455871582e-05}
{"step": 37340, "total_steps": 46238, "tokens": 20189233152, "loss": 3.0297718499646042, "ce_loss": 3.0297718499646042, "z_loss": 34.10062460465865, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17785662412643433, "tokens_per_second": 4349492.288053858, "eta_seconds": 1105.9586879167343, "mean_data_wait_seconds": 0.0001302659511566162}
{"step": 37350, "total_steps": 46238, "tokens": 20194639872, "loss": 2.9742517453251462, "ce_loss": 2.9742517453251462, "z_loss": 33.72857160279245, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16461515426635742, "tokens_per_second": 4349565.2194335805, "eta_seconds": 1104.6970953629527, "mean_data_wait_seconds": 7.301568984985352e-05}
{"step": 37360, "total_steps": 46238, "tokens": 20200046592, "loss": 2.953028108134414, "ce_loss": 2.953028108134414, "z_loss": 34.209591316454336, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13710692524909973, "tokens_per_second": 4349570.390687347, "eta_seconds": 1103.452735073807, "mean_data_wait_seconds": 0.00011894106864929199}
{"step": 37370, "total_steps": 46238, "tokens": 20205453312, "loss": 3.0069170413595256, "ce_loss": 3.0069170413595256, "z_loss": 34.490806203899965, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15999941527843475, "tokens_per_second": 4349590.840449512, "eta_seconds": 1102.2045060920134, "mean_data_wait_seconds": 0.00013521313667297363}
{"step": 37380, "total_steps": 46238, "tokens": 20210860032, "loss": 2.9948463584437515, "ce_loss": 2.9948463584437515, "z_loss": 33.92967383066813, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16398964822292328, "tokens_per_second": 4349611.139528403, "eta_seconds": 1100.9563269877058, "mean_data_wait_seconds": 0.0039460062980651855}
{"step": 37390, "total_steps": 46238, "tokens": 20216266752, "loss": 3.0069520274798074, "ce_loss": 3.0069520274798074, "z_loss": 34.40305579792369, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17545704543590546, "tokens_per_second": 4349678.113996083, "eta_seconds": 1099.6963588198764, "mean_data_wait_seconds": 8.612871170043945e-05}
{"step": 37400, "total_steps": 46238, "tokens": 20221673472, "loss": 2.961967276804375, "ce_loss": 2.961967276804375, "z_loss": 34.784512519836426, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.167518749833107, "tokens_per_second": 4349708.280017041, "eta_seconds": 1098.4457247282987, "mean_data_wait_seconds": 8.893013000488281e-05}
{"step": 37410, "total_steps": 46238, "tokens": 20227080192, "loss": 3.019424010406841, "ce_loss": 3.019424010406841, "z_loss": 34.574744282346785, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16781137883663177, "tokens_per_second": 4349806.747223802, "eta_seconds": 1097.1778796945366, "mean_data_wait_seconds": 0.00017344951629638672}
{"step": 37420, "total_steps": 46238, "tokens": 20232486912, "loss": 2.992996262781548, "ce_loss": 2.992996262781548, "z_loss": 33.481989138054125, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1910700798034668, "tokens_per_second": 4349857.4808661835, "eta_seconds": 1095.9221181312657, "mean_data_wait_seconds": 6.464123725891113e-05}
{"step": 37430, "total_steps": 46238, "tokens": 20237893632, "loss": 2.957974562139222, "ce_loss": 2.957974562139222, "z_loss": 33.64311891613585, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17976528406143188, "tokens_per_second": 4349952.207615332, "eta_seconds": 1094.6553149857225, "mean_data_wait_seconds": 9.763240814208984e-05}
{"step": 37440, "total_steps": 46238, "tokens": 20243300352, "loss": 2.942629191008481, "ce_loss": 2.942629191008481, "z_loss": 35.00154449000503, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18217791616916656, "tokens_per_second": 4350021.662111288, "eta_seconds": 1093.3949192546156, "mean_data_wait_seconds": 0.00011691451072692871}
{"step": 37450, "total_steps": 46238, "tokens": 20248707072, "loss": 2.8734561999638877, "ce_loss": 2.8734561999638877, "z_loss": 34.45325614466812, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1416136473417282, "tokens_per_second": 4350179.323946073, "eta_seconds": 1092.1124188716076, "mean_data_wait_seconds": 0.00014072656631469727}
{"step": 37460, "total_steps": 46238, "tokens": 20254113792, "loss": 2.913370450337728, "ce_loss": 2.913370450337728, "z_loss": 35.4042749115915, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19163738191127777, "tokens_per_second": 4350251.886498549, "eta_seconds": 1090.8513501776933, "mean_data_wait_seconds": 0.000164717435836792}
{"step": 37470, "total_steps": 46238, "tokens": 20259520512, "loss": 2.956959560061946, "ce_loss": 2.956959560061946, "z_loss": 34.625736727859035, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16671645641326904, "tokens_per_second": 4350274.27205017, "eta_seconds": 1089.6028911221106, "mean_data_wait_seconds": 0.00012168288230895996}
{"step": 37480, "total_steps": 46238, "tokens": 20264927232, "loss": 2.9073735638098284, "ce_loss": 2.9073735638098284, "z_loss": 34.9707163030451, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17555445432662964, "tokens_per_second": 4350296.338622862, "eta_seconds": 1088.3545247170023, "mean_data_wait_seconds": 0.00013715028762817383}
{"step": 37490, "total_steps": 46238, "tokens": 20270333952, "loss": 2.8983778339443784, "ce_loss": 2.8983778339443784, "z_loss": 33.87532098365553, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1899380087852478, "tokens_per_second": 4350319.404020178, "eta_seconds": 1087.1059213789315, "mean_data_wait_seconds": 0.00010314583778381348}
{"step": 37500, "total_steps": 46238, "tokens": 20275740672, "loss": 2.91457851005323, "ce_loss": 2.91457851005323, "z_loss": 35.7102476755778, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17028464376926422, "tokens_per_second": 4350311.688376154, "eta_seconds": 1085.8650143671148, "mean_data_wait_seconds": 0.00010770559310913086}
{"step": 37510, "total_steps": 46238, "tokens": 20281147392, "loss": 2.865623647516424, "ce_loss": 2.865623647516424, "z_loss": 36.55281615979744, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20337171852588654, "tokens_per_second": 4350271.159412907, "eta_seconds": 1084.6322840796847, "mean_data_wait_seconds": 0.0002053380012512207}
{"step": 37520, "total_steps": 46238, "tokens": 20286554112, "loss": 2.872145889383374, "ce_loss": 2.872145889383374, "z_loss": 34.14281313347094, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14473411440849304, "tokens_per_second": 4350237.505080706, "eta_seconds": 1083.397818738764, "mean_data_wait_seconds": 0.0002599954605102539}
{"step": 37530, "total_steps": 46238, "tokens": 20291960832, "loss": 2.882002445784482, "ce_loss": 2.882002445784482, "z_loss": 34.8834705063791, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15930800139904022, "tokens_per_second": 4350386.22753582, "eta_seconds": 1082.1179678721385, "mean_data_wait_seconds": 0.0001361370086669922}
{"step": 37540, "total_steps": 46238, "tokens": 20297367552, "loss": 2.954141976255359, "ce_loss": 2.954141976255359, "z_loss": 34.919287334788926, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16408896446228027, "tokens_per_second": 4350270.190468106, "eta_seconds": 1080.9039848382433, "mean_data_wait_seconds": 0.00018277764320373535}
{"step": 37550, "total_steps": 46238, "tokens": 20302774272, "loss": 2.879948964624694, "ce_loss": 2.879948964624694, "z_loss": 34.31849985411673, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19785165786743164, "tokens_per_second": 4350348.9385865275, "eta_seconds": 1079.6415943420952, "mean_data_wait_seconds": 0.0001067817211151123}
{"step": 37560, "total_steps": 46238, "tokens": 20308180992, "loss": 3.01903775063428, "ce_loss": 3.01903775063428, "z_loss": 35.520099899985574, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16007453203201294, "tokens_per_second": 4350402.336503175, "eta_seconds": 1078.385533364467, "mean_data_wait_seconds": 9.638071060180664e-05}
{"step": 37570, "total_steps": 46238, "tokens": 20313587712, "loss": 2.9553099798433706, "ce_loss": 2.9553099798433706, "z_loss": 34.00830442255194, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18474841117858887, "tokens_per_second": 4350338.286743175, "eta_seconds": 1077.1585828807158, "mean_data_wait_seconds": 9.521842002868652e-05}
{"step": 37580, "total_steps": 46238, "tokens": 20318994432, "loss": 2.9602060137373027, "ce_loss": 2.9602060137373027, "z_loss": 33.455205194877855, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16098809242248535, "tokens_per_second": 4350359.000203804, "eta_seconds": 1075.9106326123258, "mean_data_wait_seconds": 7.748603820800781e-05}
{"step": 37590, "total_steps": 46238, "tokens": 20324401152, "loss": 3.0370626359274895, "ce_loss": 3.0370626359274895, "z_loss": 35.27024309562914, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1537868082523346, "tokens_per_second": 4350455.895845657, "eta_seconds": 1074.643875476232, "mean_data_wait_seconds": 9.065866470336914e-05}
{"step": 37600, "total_steps": 46238, "tokens": 20329807872, "loss": 2.967005359404015, "ce_loss": 2.967005359404015, "z_loss": 35.43976272236217, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15388882160186768, "tokens_per_second": 4350553.084764733, "eta_seconds": 1073.3771024086998, "mean_data_wait_seconds": 6.976723670959473e-05}
{"step": 37610, "total_steps": 46238, "tokens": 20335214592, "loss": 3.0374682238607695, "ce_loss": 3.0374682238607695, "z_loss": 34.36821322007613, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15072397887706757, "tokens_per_second": 4350627.619744468, "eta_seconds": 1072.115968471225, "mean_data_wait_seconds": 7.048249244689941e-05}
{"step": 37620, "total_steps": 46238, "tokens": 20340621312, "loss": 3.0065181598518835, "ce_loss": 3.0065181598518835, "z_loss": 34.46522053805265, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1608245074748993, "tokens_per_second": 4350115.503116149, "eta_seconds": 1070.999292010202, "mean_data_wait_seconds": 8.541345596313477e-05}
{"step": 37630, "total_steps": 46238, "tokens": 20346028032, "loss": 3.0024521549542746, "ce_loss": 3.0024521549542746, "z_loss": 34.86872147068833, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1650862842798233, "tokens_per_second": 4350214.454125303, "eta_seconds": 1069.7320679413933, "mean_data_wait_seconds": 0.00013133883476257324}
{"step": 37640, "total_steps": 46238, "tokens": 20351434752, "loss": 3.0297954913341636, "ce_loss": 3.0297954913341636, "z_loss": 34.46113517067649, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14056038856506348, "tokens_per_second": 4349691.22522143, "eta_seconds": 1068.6177347596383, "mean_data_wait_seconds": 9.936094284057617e-05}
{"step": 37650, "total_steps": 46238, "tokens": 20356841472, "loss": 2.988118588924408, "ce_loss": 2.988118588924408, "z_loss": 35.10418574015299, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12680518627166748, "tokens_per_second": 4349747.976409311, "eta_seconds": 1067.3607963449322, "mean_data_wait_seconds": 7.858872413635254e-05}
{"step": 37660, "total_steps": 46238, "tokens": 20362248192, "loss": 3.0248745210242993, "ce_loss": 3.0248745210242993, "z_loss": 35.483859062194824, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.158778116106987, "tokens_per_second": 4349277.371950213, "eta_seconds": 1066.2331572383985, "mean_data_wait_seconds": 8.544325828552246e-05}
{"step": 37670, "total_steps": 46238, "tokens": 20367654912, "loss": 2.99981689453125, "ce_loss": 2.99981689453125, "z_loss": 34.88320382436117, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1453053206205368, "tokens_per_second": 4348819.166417958, "eta_seconds": 1065.1022373540634, "mean_data_wait_seconds": 9.73045825958252e-05}
{"step": 37680, "total_steps": 46238, "tokens": 20373061632, "loss": 2.9865963729945095, "ce_loss": 2.9865963729945095, "z_loss": 35.29973584955389, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16081255674362183, "tokens_per_second": 4348841.218314728, "eta_seconds": 1063.8535811599218, "mean_data_wait_seconds": 8.70823860168457e-05}
{"step": 37690, "total_steps": 46238, "tokens": 20378468352, "loss": 2.994412404118162, "ce_loss": 2.994412404118162, "z_loss": 35.621042367183804, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15467828512191772, "tokens_per_second": 4348881.641154279, "eta_seconds": 1062.6004488761996, "mean_data_wait_seconds": 9.423494338989258e-05}
{"step": 37700, "total_steps": 46238, "tokens": 20383875072, "loss": 3.005711205077894, "ce_loss": 3.005711205077894, "z_loss": 34.753261334968336, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1532886028289795, "tokens_per_second": 4348928.269877279, "eta_seconds": 1061.3458253544038, "mean_data_wait_seconds": 0.00017702579498291016}
{"step": 37710, "total_steps": 46238, "tokens": 20389281792, "loss": 3.04866602565303, "ce_loss": 3.04866602565303, "z_loss": 33.85065778096517, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.12313593178987503, "tokens_per_second": 4348486.701200406, "eta_seconds": 1060.210243422687, "mean_data_wait_seconds": 8.684396743774414e-05}
{"step": 37720, "total_steps": 46238, "tokens": 20394688512, "loss": 2.993833491296479, "ce_loss": 2.993833491296479, "z_loss": 35.08582282788826, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16322587430477142, "tokens_per_second": 4348533.872948323, "eta_seconds": 1058.955399346552, "mean_data_wait_seconds": 8.627772331237793e-05}
{"step": 37730, "total_steps": 46238, "tokens": 20400095232, "loss": 2.9957933552337415, "ce_loss": 2.9957933552337415, "z_loss": 34.23403560754025, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14807440340518951, "tokens_per_second": 4348561.965706064, "eta_seconds": 1057.7052230767033, "mean_data_wait_seconds": 0.00010642409324645996}
{"step": 37740, "total_steps": 46238, "tokens": 20405501952, "loss": 3.10380558534102, "ce_loss": 3.10380558534102, "z_loss": 34.573872450626254, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15041175484657288, "tokens_per_second": 4348532.874880932, "eta_seconds": 1056.4689554349502, "mean_data_wait_seconds": 9.065866470336914e-05}
{"step": 37750, "total_steps": 46238, "tokens": 20410908672, "loss": 3.028851097280329, "ce_loss": 3.028851097280329, "z_loss": 34.2072196151271, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1678514927625656, "tokens_per_second": 4348588.915852458, "eta_seconds": 1055.2120130905673, "mean_data_wait_seconds": 8.082389831542969e-05}
{"step": 37760, "total_steps": 46238, "tokens": 20416315392, "loss": 3.033939014781605, "ce_loss": 3.033939014781605, "z_loss": 33.83868613387599, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1850094050168991, "tokens_per_second": 4348614.323086374, "eta_seconds": 1053.962527710914, "mean_data_wait_seconds": 0.00011911988258361816}
{"step": 37770, "total_steps": 46238, "tokens": 20421722112, "loss": 3.0568570043101455, "ce_loss": 3.0568570043101455, "z_loss": 34.704852335380785, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1736392378807068, "tokens_per_second": 4348619.393375479, "eta_seconds": 1052.7179800958788, "mean_data_wait_seconds": 7.75456428527832e-05}
{"step": 37780, "total_steps": 46238, "tokens": 20427128832, "loss": 3.0021117672775732, "ce_loss": 3.0021117672775732, "z_loss": 34.42413569941665, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15781673789024353, "tokens_per_second": 4348700.634771537, "eta_seconds": 1051.4550179516366, "mean_data_wait_seconds": 0.00011286139488220215}
{"step": 37790, "total_steps": 46238, "tokens": 20432535552, "loss": 3.043541178558812, "ce_loss": 3.043541178558812, "z_loss": 34.90313316114021, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13258618116378784, "tokens_per_second": 4348689.770792272, "eta_seconds": 1050.2143460943971, "mean_data_wait_seconds": 0.00013554096221923828}
{"step": 37800, "total_steps": 46238, "tokens": 20437942272, "loss": 3.0016243945468557, "ce_loss": 3.0016243945468557, "z_loss": 34.16947850314054, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16816075146198273, "tokens_per_second": 4348805.748966167, "eta_seconds": 1048.9430724939673, "mean_data_wait_seconds": 0.00011280179023742676}
{"step": 37810, "total_steps": 46238, "tokens": 20443348992, "loss": 3.0093353444879707, "ce_loss": 3.0093353444879707, "z_loss": 34.98143941705877, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20145545899868011, "tokens_per_second": 4348982.330018475, "eta_seconds": 1047.6572674372408, "mean_data_wait_seconds": 6.371736526489258e-05}
{"step": 37820, "total_steps": 46238, "tokens": 20448755712, "loss": 3.054330009402651, "ce_loss": 3.054330009402651, "z_loss": 35.375314221237645, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1405201405286789, "tokens_per_second": 4349127.565008874, "eta_seconds": 1046.3791084478605, "mean_data_wait_seconds": 6.389617919921875e-05}
{"step": 37830, "total_steps": 46238, "tokens": 20454162432, "loss": 2.924401814287359, "ce_loss": 2.924401814287359, "z_loss": 34.50644354386763, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14624613523483276, "tokens_per_second": 4349231.792230723, "eta_seconds": 1045.1108888056406, "mean_data_wait_seconds": 0.0001232624053955078}
{"step": 37840, "total_steps": 46238, "tokens": 20459569152, "loss": 2.961753962617932, "ce_loss": 2.961753962617932, "z_loss": 34.658514774206914, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14117157459259033, "tokens_per_second": 4349300.561542256, "eta_seconds": 1043.85123993135, "mean_data_wait_seconds": 6.109476089477539e-05}
{"step": 37850, "total_steps": 46238, "tokens": 20464975872, "loss": 3.0237454808119573, "ce_loss": 3.0237454808119573, "z_loss": 34.342229814240426, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1718580722808838, "tokens_per_second": 4349397.94402377, "eta_seconds": 1042.5847720443069, "mean_data_wait_seconds": 6.091594696044922e-05}
{"step": 37860, "total_steps": 46238, "tokens": 20470382592, "loss": 3.0040815273920694, "ce_loss": 3.0040815273920694, "z_loss": 35.182271899599016, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1849462240934372, "tokens_per_second": 4349499.098343885, "eta_seconds": 1041.3174578480869, "mean_data_wait_seconds": 9.79304313659668e-05}
{"step": 37870, "total_steps": 46238, "tokens": 20475789312, "loss": 3.086084893255523, "ce_loss": 3.086084893255523, "z_loss": 34.33102272496079, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17415063083171844, "tokens_per_second": 4349502.738949906, "eta_seconds": 1040.0735200116635, "mean_data_wait_seconds": 9.429454803466797e-05}
{"step": 37880, "total_steps": 46238, "tokens": 20481196032, "loss": 3.0175186359521113, "ce_loss": 3.0175186359521113, "z_loss": 34.730838082053445, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16533561050891876, "tokens_per_second": 4349581.019895051, "eta_seconds": 1038.8117575768304, "mean_data_wait_seconds": 6.961822509765625e-05}
{"step": 37890, "total_steps": 46238, "tokens": 20486602752, "loss": 2.98393566680677, "ce_loss": 2.98393566680677, "z_loss": 35.57244841257731, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1443963348865509, "tokens_per_second": 4349616.812878548, "eta_seconds": 1037.5601755625303, "mean_data_wait_seconds": 0.00011411309242248535}
{"step": 37900, "total_steps": 46238, "tokens": 20492009472, "loss": 2.9823623191226614, "ce_loss": 2.9823623191226614, "z_loss": 35.40173585487135, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.153632253408432, "tokens_per_second": 4349613.662086263, "eta_seconds": 1036.3178926189892, "mean_data_wait_seconds": 8.147954940795898e-05}
{"step": 37910, "total_steps": 46238, "tokens": 20497416192, "loss": 2.9877541426456338, "ce_loss": 2.9877541426456338, "z_loss": 34.76580322150028, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14952678978443146, "tokens_per_second": 4349742.7831304, "eta_seconds": 1035.0441321405901, "mean_data_wait_seconds": 6.839632987976074e-05}
{"step": 37920, "total_steps": 46238, "tokens": 20502822912, "loss": 2.9758019826628943, "ce_loss": 2.9758019826628943, "z_loss": 34.499950900222316, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.21148622035980225, "tokens_per_second": 4349811.292895836, "eta_seconds": 1033.7848520794862, "mean_data_wait_seconds": 6.29723072052002e-05}
{"step": 37930, "total_steps": 46238, "tokens": 20508229632, "loss": 2.9765878066872107, "ce_loss": 2.9765878066872107, "z_loss": 35.33950375065659, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17964942753314972, "tokens_per_second": 4349900.097016209, "eta_seconds": 1032.5207944616536, "mean_data_wait_seconds": 9.918212890625e-05}
{"step": 37940, "total_steps": 46238, "tokens": 20513636352, "loss": 3.0026718829617356, "ce_loss": 3.0026718829617356, "z_loss": 34.713043530782066, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1677299290895462, "tokens_per_second": 4349883.868497107, "eta_seconds": 1031.281689262639, "mean_data_wait_seconds": 7.554888725280762e-05}
{"step": 37950, "total_steps": 46238, "tokens": 20519043072, "loss": 3.0333369934197627, "ce_loss": 3.0333369934197627, "z_loss": 34.90043862660726, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15294180810451508, "tokens_per_second": 4349845.556918256, "eta_seconds": 1030.0478040821163, "mean_data_wait_seconds": 8.651614189147949e-05}
{"step": 37960, "total_steps": 46238, "tokens": 20524449792, "loss": 2.9872229424389927, "ce_loss": 2.9872229424389927, "z_loss": 34.23474912932425, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14663708209991455, "tokens_per_second": 4349820.865108851, "eta_seconds": 1028.8106758364204, "mean_data_wait_seconds": 6.249547004699707e-05}
{"step": 37970, "total_steps": 46238, "tokens": 20529856512, "loss": 3.0064875609946973, "ce_loss": 3.0064875609946973, "z_loss": 35.25502834898053, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16529464721679688, "tokens_per_second": 4349945.720218576, "eta_seconds": 1027.5382065630477, "mean_data_wait_seconds": 0.00012430548667907715}
{"step": 37980, "total_steps": 46238, "tokens": 20535263232, "loss": 2.991032791860176, "ce_loss": 2.991032791860176, "z_loss": 35.02482509613037, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16201435029506683, "tokens_per_second": 4349997.204570535, "eta_seconds": 1026.2831202073733, "mean_data_wait_seconds": 0.00010877847671508789}
{"step": 37990, "total_steps": 46238, "tokens": 20540669952, "loss": 2.935630451549183, "ce_loss": 2.935630451549183, "z_loss": 36.25899005658699, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15821701288223267, "tokens_per_second": 4350069.504666348, "eta_seconds": 1025.02315864537, "mean_data_wait_seconds": 9.769201278686523e-05}
{"step": 38000, "total_steps": 46238, "tokens": 20546076672, "loss": 2.957469389294133, "ce_loss": 2.957469389294133, "z_loss": 34.713213140314274, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18097639083862305, "tokens_per_second": 4350140.237233446, "eta_seconds": 1023.7636078675702, "mean_data_wait_seconds": 0.00010752677917480469}
{"step": 38010, "total_steps": 46238, "tokens": 20551483392, "loss": 2.9392129995606164, "ce_loss": 2.9392129995606164, "z_loss": 33.10173104026101, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19640514254570007, "tokens_per_second": 4350257.62778852, "eta_seconds": 1022.4931313461595, "mean_data_wait_seconds": 0.00013247132301330566}
{"step": 38020, "total_steps": 46238, "tokens": 20556890112, "loss": 2.936788002649943, "ce_loss": 2.936788002649943, "z_loss": 34.91704322352554, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1377660185098648, "tokens_per_second": 4350365.372956243, "eta_seconds": 1021.224987587884, "mean_data_wait_seconds": 9.822845458984375e-05}
{"step": 38030, "total_steps": 46238, "tokens": 20562296832, "loss": 2.921311454339461, "ce_loss": 2.921311454339461, "z_loss": 36.495296304876156, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2093029022216797, "tokens_per_second": 4350433.346391934, "eta_seconds": 1019.9662311066335, "mean_data_wait_seconds": 6.720423698425293e-05}
{"step": 38040, "total_steps": 46238, "tokens": 20567703552, "loss": 2.9237803907105415, "ce_loss": 2.9237803907105415, "z_loss": 34.11629925352155, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17940634489059448, "tokens_per_second": 4350413.986821844, "eta_seconds": 1018.7279641489191, "mean_data_wait_seconds": 6.508827209472656e-05}
{"step": 38050, "total_steps": 46238, "tokens": 20573110272, "loss": 2.8448697527249656, "ce_loss": 2.8448697527249656, "z_loss": 35.78094959259033, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19724434614181519, "tokens_per_second": 4350522.072986667, "eta_seconds": 1017.4598794671064, "mean_data_wait_seconds": 6.031990051269531e-05}
{"step": 38060, "total_steps": 46238, "tokens": 20578516992, "loss": 2.8828627123977197, "ce_loss": 2.8828627123977197, "z_loss": 37.119102391329676, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15549881756305695, "tokens_per_second": 4350636.091655048, "eta_seconds": 1016.1904721197117, "mean_data_wait_seconds": 6.67870044708252e-05}
{"step": 38070, "total_steps": 46238, "tokens": 20583923712, "loss": 2.8539070198030183, "ce_loss": 2.8539070198030183, "z_loss": 35.32077485864813, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20414908230304718, "tokens_per_second": 4350715.0897632, "eta_seconds": 1014.9293007923292, "mean_data_wait_seconds": 6.0677528381347656e-05}
{"step": 38080, "total_steps": 46238, "tokens": 20589330432, "loss": 2.8871656385335056, "ce_loss": 2.8871656385335056, "z_loss": 34.677937998916164, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15375111997127533, "tokens_per_second": 4350795.363774267, "eta_seconds": 1013.6678779978627, "mean_data_wait_seconds": 5.441904067993164e-05}
{"step": 38090, "total_steps": 46238, "tokens": 20594737152, "loss": 2.8951456420349353, "ce_loss": 2.8951456420349353, "z_loss": 34.859723929202914, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15683047473430634, "tokens_per_second": 4350957.454042557, "eta_seconds": 1012.3874642597036, "mean_data_wait_seconds": 8.314847946166992e-05}
{"step": 38100, "total_steps": 46238, "tokens": 20600143872, "loss": 2.8690297025622744, "ce_loss": 2.8690297025622744, "z_loss": 35.83461154590953, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18474531173706055, "tokens_per_second": 4350996.221793244, "eta_seconds": 1011.1358042473287, "mean_data_wait_seconds": 7.402896881103516e-05}
{"step": 38110, "total_steps": 46238, "tokens": 20605550592, "loss": 2.9318743257811577, "ce_loss": 2.9318743257811577, "z_loss": 34.75986943100438, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13394486904144287, "tokens_per_second": 4351084.560271577, "eta_seconds": 1009.8726612028294, "mean_data_wait_seconds": 8.177757263183594e-05}
{"step": 38120, "total_steps": 46238, "tokens": 20610957312, "loss": 2.909025732314948, "ce_loss": 2.909025732314948, "z_loss": 35.62953186035156, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16073206067085266, "tokens_per_second": 4351095.521776462, "eta_seconds": 1008.6275058857387, "mean_data_wait_seconds": 8.112192153930664e-05}
{"step": 38130, "total_steps": 46238, "tokens": 20616364032, "loss": 2.8539400191018074, "ce_loss": 2.8539400191018074, "z_loss": 35.228703932328656, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16283802688121796, "tokens_per_second": 4351081.7982400395, "eta_seconds": 1007.388072036008, "mean_data_wait_seconds": 6.872415542602539e-05}
{"step": 38140, "total_steps": 46238, "tokens": 20621770752, "loss": 2.830966228788549, "ce_loss": 2.830966228788549, "z_loss": 34.428894187464856, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19320613145828247, "tokens_per_second": 4351134.966148089, "eta_seconds": 1006.1331625103635, "mean_data_wait_seconds": 7.325410842895508e-05}
{"step": 38150, "total_steps": 46238, "tokens": 20627177472, "loss": 2.9278685053189597, "ce_loss": 2.9278685053189597, "z_loss": 34.36946189764774, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1536974161863327, "tokens_per_second": 4351104.207035445, "eta_seconds": 1004.8976664199625, "mean_data_wait_seconds": 8.374452590942383e-05}
{"step": 38160, "total_steps": 46238, "tokens": 20632584192, "loss": 3.009448891336268, "ce_loss": 3.009448891336268, "z_loss": 35.149261272314824, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17301948368549347, "tokens_per_second": 4351088.329223047, "eta_seconds": 1003.6587202034108, "mean_data_wait_seconds": 9.065866470336914e-05}
{"step": 38170, "total_steps": 46238, "tokens": 20637990912, "loss": 2.9700081709659463, "ce_loss": 2.9700081709659463, "z_loss": 35.12684761394154, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16721010208129883, "tokens_per_second": 4351094.1735351, "eta_seconds": 1002.4147605282382, "mean_data_wait_seconds": 9.524822235107422e-05}
{"step": 38180, "total_steps": 46238, "tokens": 20643397632, "loss": 2.959854037472696, "ce_loss": 2.959854037472696, "z_loss": 35.511054154598355, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1618424952030182, "tokens_per_second": 4351180.499915267, "eta_seconds": 1001.1522859336288, "mean_data_wait_seconds": 9.417533874511719e-05}
{"step": 38190, "total_steps": 46238, "tokens": 20648804352, "loss": 2.9808044108477505, "ce_loss": 2.9808044108477505, "z_loss": 34.4623445742058, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16550235450267792, "tokens_per_second": 4351167.907080819, "eta_seconds": 999.9125928741569, "mean_data_wait_seconds": 0.0001392066478729248}
{"step": 38200, "total_steps": 46238, "tokens": 20654211072, "loss": 2.958622208147338, "ce_loss": 2.958622208147338, "z_loss": 35.37144175442782, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19900599122047424, "tokens_per_second": 4351156.252569206, "eta_seconds": 998.6726772761158, "mean_data_wait_seconds": 0.00010082125663757324}
{"step": 38210, "total_steps": 46238, "tokens": 20659617792, "loss": 2.98670703533924, "ce_loss": 2.98670703533924, "z_loss": 35.25264043518991, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1520913541316986, "tokens_per_second": 4351126.58089646, "eta_seconds": 997.4368852091261, "mean_data_wait_seconds": 0.00010654330253601074}
{"step": 38220, "total_steps": 46238, "tokens": 20665024512, "loss": 3.0009535099520828, "ce_loss": 3.0009535099520828, "z_loss": 36.09462477944114, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1604672521352768, "tokens_per_second": 4351104.295688958, "eta_seconds": 996.1993851295767, "mean_data_wait_seconds": 0.00013631582260131836}
{"step": 38230, "total_steps": 46238, "tokens": 20670431232, "loss": 2.96396066383882, "ce_loss": 2.96396066383882, "z_loss": 35.25773663954301, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14270862936973572, "tokens_per_second": 4351161.419743469, "eta_seconds": 994.9437141900459, "mean_data_wait_seconds": 0.00010484457015991211}
{"step": 38240, "total_steps": 46238, "tokens": 20675837952, "loss": 3.0132561777577256, "ce_loss": 3.0132561777577256, "z_loss": 35.34873095425692, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17121492326259613, "tokens_per_second": 4351219.997832122, "eta_seconds": 993.6877441623714, "mean_data_wait_seconds": 6.255507469177246e-05}
{"step": 38250, "total_steps": 46238, "tokens": 20681244672, "loss": 3.030797971017433, "ce_loss": 3.030797971017433, "z_loss": 34.53779382416696, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14302687346935272, "tokens_per_second": 4350494.854495838, "eta_seconds": 992.6105899280362, "mean_data_wait_seconds": 7.662177085876465e-05}
{"step": 38260, "total_steps": 46238, "tokens": 20686651392, "loss": 3.005375927144831, "ce_loss": 3.005375927144831, "z_loss": 34.38130063721628, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16671688854694366, "tokens_per_second": 4350110.002015354, "eta_seconds": 991.4555130794085, "mean_data_wait_seconds": 0.00018027424812316895}
{"step": 38270, "total_steps": 46238, "tokens": 20692058112, "loss": 3.01748065514998, "ce_loss": 3.01748065514998, "z_loss": 35.454740755485766, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16233693063259125, "tokens_per_second": 4350120.995170508, "eta_seconds": 990.2101180133178, "mean_data_wait_seconds": 0.00014215707778930664}
{"step": 38280, "total_steps": 46238, "tokens": 20697464832, "loss": 3.013069479754477, "ce_loss": 3.013069479754477, "z_loss": 34.51171927018599, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17206810414791107, "tokens_per_second": 4350137.069991709, "eta_seconds": 988.9635739703714, "mean_data_wait_seconds": 0.00016495585441589355}
{"step": 38290, "total_steps": 46238, "tokens": 20702871552, "loss": 3.0001995581569094, "ce_loss": 3.0001995581569094, "z_loss": 34.440532597628504, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14864714443683624, "tokens_per_second": 4349683.048945213, "eta_seconds": 987.8237875382538, "mean_data_wait_seconds": 0.00016251206398010254}
{"step": 38300, "total_steps": 46238, "tokens": 20708278272, "loss": 3.033450616128517, "ce_loss": 3.033450616128517, "z_loss": 34.36760081666888, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18200241029262543, "tokens_per_second": 4349721.955460716, "eta_seconds": 986.5719482627185, "mean_data_wait_seconds": 0.00012546777725219727}
{"step": 38310, "total_steps": 46238, "tokens": 20713684992, "loss": 3.046660820643107, "ce_loss": 3.046660820643107, "z_loss": 36.070689750440195, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2835122346878052, "tokens_per_second": 4349747.273543264, "eta_seconds": 985.323209481258, "mean_data_wait_seconds": 0.00012257695198059082}
{"step": 38320, "total_steps": 46238, "tokens": 20719091712, "loss": 3.0163518154259883, "ce_loss": 3.0163518154259883, "z_loss": 34.61389608094186, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16336475312709808, "tokens_per_second": 4349619.453405473, "eta_seconds": 984.1091318112077, "mean_data_wait_seconds": 0.00015273690223693848}
{"step": 38330, "total_steps": 46238, "tokens": 20724498432, "loss": 3.0126194177251873, "ce_loss": 3.0126194177251873, "z_loss": 34.479025436170176, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13492555916309357, "tokens_per_second": 4349642.450810222, "eta_seconds": 982.8609023262738, "mean_data_wait_seconds": 0.00012233853340148926}
{"step": 38340, "total_steps": 46238, "tokens": 20729905152, "loss": 3.0301516724355295, "ce_loss": 3.0301516724355295, "z_loss": 35.35431096048066, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18236121535301208, "tokens_per_second": 4349653.008181647, "eta_seconds": 981.6154934586203, "mean_data_wait_seconds": 8.487701416015625e-05}
{"step": 38350, "total_steps": 46238, "tokens": 20735311872, "loss": 3.0178803530606357, "ce_loss": 3.0178803530606357, "z_loss": 35.266922025969535, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17959676682949066, "tokens_per_second": 4349648.707111536, "eta_seconds": 980.3734395901994, "mean_data_wait_seconds": 0.0001131594181060791}
{"step": 38360, "total_steps": 46238, "tokens": 20740718592, "loss": 2.986253624612635, "ce_loss": 2.986253624612635, "z_loss": 34.543433131593645, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1643580049276352, "tokens_per_second": 4349596.309581337, "eta_seconds": 979.1422101905202, "mean_data_wait_seconds": 7.274746894836426e-05}
{"step": 38370, "total_steps": 46238, "tokens": 20746125312, "loss": 3.044207159316901, "ce_loss": 3.044207159316901, "z_loss": 34.638266534516305, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1734100580215454, "tokens_per_second": 4349615.705887825, "eta_seconds": 977.8948099351228, "mean_data_wait_seconds": 0.0001150369644165039}
{"step": 38380, "total_steps": 46238, "tokens": 20751532032, "loss": 3.0211473161523994, "ce_loss": 3.0211473161523994, "z_loss": 35.78406016031901, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1531369388103485, "tokens_per_second": 4349645.327490284, "eta_seconds": 976.6451248683998, "mean_data_wait_seconds": 0.00017562508583068848}
{"step": 38390, "total_steps": 46238, "tokens": 20756938752, "loss": 3.0134104472218137, "ce_loss": 3.0134104472218137, "z_loss": 34.48003644654245, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1791105568408966, "tokens_per_second": 4349585.180766905, "eta_seconds": 975.4155873898644, "mean_data_wait_seconds": 8.195638656616211e-05}
{"step": 38400, "total_steps": 46238, "tokens": 20762345472, "loss": 3.0491160623955005, "ce_loss": 3.0491160623955005, "z_loss": 33.91787309357614, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1644909530878067, "tokens_per_second": 4349704.47326139, "eta_seconds": 974.1458276182452, "mean_data_wait_seconds": 6.383657455444336e-05}
{"step": 38410, "total_steps": 46238, "tokens": 20767752192, "loss": 3.030420218453263, "ce_loss": 3.030420218453263, "z_loss": 35.94306162631873, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18303422629833221, "tokens_per_second": 4349776.763783209, "eta_seconds": 972.8866500080723, "mean_data_wait_seconds": 0.0001361370086669922}
{"step": 38420, "total_steps": 46238, "tokens": 20773158912, "loss": 3.0101674560344582, "ce_loss": 3.0101674560344582, "z_loss": 34.76078559413101, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14760971069335938, "tokens_per_second": 4349796.970277217, "eta_seconds": 971.6391484199881, "mean_data_wait_seconds": 0.0001004636287689209}
{"step": 38430, "total_steps": 46238, "tokens": 20778565632, "loss": 3.0426993911916558, "ce_loss": 3.0426993911916558, "z_loss": 36.22609953446822, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14267343282699585, "tokens_per_second": 4349862.276646197, "eta_seconds": 970.3815972892063, "mean_data_wait_seconds": 0.00010696053504943848}
{"step": 38440, "total_steps": 46238, "tokens": 20783972352, "loss": 3.0478512428023596, "ce_loss": 3.0478512428023596, "z_loss": 35.64065355243105, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14923913776874542, "tokens_per_second": 4349916.752393729, "eta_seconds": 969.1264968875953, "mean_data_wait_seconds": 7.787346839904785e-05}
{"step": 38450, "total_steps": 46238, "tokens": 20789379072, "loss": 2.9566132390137874, "ce_loss": 2.9566132390137874, "z_loss": 34.96382664189194, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17756198346614838, "tokens_per_second": 4350052.538200072, "eta_seconds": 967.8533367190241, "mean_data_wait_seconds": 5.632638931274414e-05}
{"step": 38460, "total_steps": 46238, "tokens": 20794785792, "loss": 3.0078907374179726, "ce_loss": 3.0078907374179726, "z_loss": 35.54127014044559, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14210094511508942, "tokens_per_second": 4350071.420031776, "eta_seconds": 966.6062319430345, "mean_data_wait_seconds": 0.00011879205703735352}
{"step": 38470, "total_steps": 46238, "tokens": 20800192512, "loss": 3.0742615100109214, "ce_loss": 3.0742615100109214, "z_loss": 34.983732628099844, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14668472111225128, "tokens_per_second": 4350142.050881656, "eta_seconds": 965.3476541412931, "mean_data_wait_seconds": 5.7250261306762695e-05}
{"step": 38480, "total_steps": 46238, "tokens": 20805599232, "loss": 3.012823001904921, "ce_loss": 3.012823001904921, "z_loss": 34.03862785570549, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17532068490982056, "tokens_per_second": 4350218.561113066, "eta_seconds": 964.0878142285584, "mean_data_wait_seconds": 7.343292236328125e-05}
{"step": 38490, "total_steps": 46238, "tokens": 20811005952, "loss": 2.975134004246105, "ce_loss": 2.975134004246105, "z_loss": 35.67127216223514, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1689494252204895, "tokens_per_second": 4350325.780035411, "eta_seconds": 962.8212220846378, "mean_data_wait_seconds": 7.167458534240723e-05}
{"step": 38500, "total_steps": 46238, "tokens": 20816412672, "loss": 3.012777485630729, "ce_loss": 3.012777485630729, "z_loss": 35.29952078154593, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17146773636341095, "tokens_per_second": 4350397.87912506, "eta_seconds": 961.5624547981138, "mean_data_wait_seconds": 8.893013000488281e-05}
{"step": 38510, "total_steps": 46238, "tokens": 20821819392, "loss": 2.9168519865382803, "ce_loss": 2.9168519865382803, "z_loss": 34.94979474038789, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16452790796756744, "tokens_per_second": 4350455.922239435, "eta_seconds": 960.3068318985416, "mean_data_wait_seconds": 9.617209434509277e-05}
{"step": 38520, "total_steps": 46238, "tokens": 20827226112, "loss": 2.8493492657488044, "ce_loss": 2.8493492657488044, "z_loss": 35.419067527308606, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1615534871816635, "tokens_per_second": 4350579.927365579, "eta_seconds": 959.036701694734, "mean_data_wait_seconds": 6.368756294250488e-05}
{"step": 38530, "total_steps": 46238, "tokens": 20832632832, "loss": 2.906372744025606, "ce_loss": 2.906372744025606, "z_loss": 34.79678249359131, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1448432207107544, "tokens_per_second": 4350646.968139536, "eta_seconds": 957.779184225999, "mean_data_wait_seconds": 9.012222290039062e-05}
{"step": 38540, "total_steps": 46238, "tokens": 20838039552, "loss": 2.955872432752089, "ce_loss": 2.955872432752089, "z_loss": 36.42189915974935, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1922266036272049, "tokens_per_second": 4350819.814304457, "eta_seconds": 956.4984443432497, "mean_data_wait_seconds": 9.080767631530762e-05}
{"step": 38550, "total_steps": 46238, "tokens": 20843446272, "loss": 2.952316217350237, "ce_loss": 2.952316217350237, "z_loss": 35.61742568738533, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1604641079902649, "tokens_per_second": 4350884.520432404, "eta_seconds": 955.2415478926454, "mean_data_wait_seconds": 6.851553916931152e-05}
{"step": 38560, "total_steps": 46238, "tokens": 20848852992, "loss": 2.9038452271259194, "ce_loss": 2.9038452271259194, "z_loss": 34.236895474520594, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19244131445884705, "tokens_per_second": 4350927.51848077, "eta_seconds": 953.9894485416133, "mean_data_wait_seconds": 0.00015538930892944336}
{"step": 38570, "total_steps": 46238, "tokens": 20854259712, "loss": 2.9787835063356343, "ce_loss": 2.9787835063356343, "z_loss": 35.13856858918161, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14176791906356812, "tokens_per_second": 4350975.851950004, "eta_seconds": 952.736205635837, "mean_data_wait_seconds": 0.00015679001808166504}
{"step": 38580, "total_steps": 46238, "tokens": 20859666432, "loss": 3.0171425378683843, "ce_loss": 3.0171425378683843, "z_loss": 35.788321408358485, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15793132781982422, "tokens_per_second": 4351004.947904564, "eta_seconds": 951.487197456252, "mean_data_wait_seconds": 0.00012248754501342773}
{"step": 38590, "total_steps": 46238, "tokens": 20865073152, "loss": 2.9565554188959524, "ce_loss": 2.9565554188959524, "z_loss": 35.43601983966249, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15963156521320343, "tokens_per_second": 4351034.277509045, "eta_seconds": 950.2381549535851, "mean_data_wait_seconds": 0.0001170039176940918}
{"step": 38600, "total_steps": 46238, "tokens": 20870479872, "loss": 2.8987447185949846, "ce_loss": 2.8987447185949846, "z_loss": 35.637909629128195, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13738159835338593, "tokens_per_second": 4351016.262325928, "eta_seconds": 948.9994555416108, "mean_data_wait_seconds": 0.00012695789337158203}
{"step": 38610, "total_steps": 46238, "tokens": 20875886592, "loss": 2.942987991101814, "ce_loss": 2.942987991101814, "z_loss": 34.68871657053629, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16742341220378876, "tokens_per_second": 4350914.7292736275, "eta_seconds": 947.7789385884932, "mean_data_wait_seconds": 8.574128150939941e-05}
{"step": 38620, "total_steps": 46238, "tokens": 20881293312, "loss": 2.862173916715564, "ce_loss": 2.862173916715564, "z_loss": 36.53329407085072, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.162004753947258, "tokens_per_second": 4350943.4043944795, "eta_seconds": 946.5300375639208, "mean_data_wait_seconds": 7.414817810058594e-05}
{"step": 38630, "total_steps": 46238, "tokens": 20886700032, "loss": 2.9365347930879304, "ce_loss": 2.9365347930879304, "z_loss": 36.46848932902018, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15200099349021912, "tokens_per_second": 4350865.61301252, "eta_seconds": 945.3042842093788, "mean_data_wait_seconds": 9.846687316894531e-05}
{"step": 38640, "total_steps": 46238, "tokens": 20892106752, "loss": 2.91003263538534, "ce_loss": 2.91003263538534, "z_loss": 35.102327317902535, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17756715416908264, "tokens_per_second": 4350820.00672183, "eta_seconds": 944.0715032233261, "mean_data_wait_seconds": 7.915496826171875e-05}
{"step": 38650, "total_steps": 46238, "tokens": 20897513472, "loss": 2.9347674900835212, "ce_loss": 2.9347674900835212, "z_loss": 35.1978008963845, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1626802384853363, "tokens_per_second": 4350871.607533939, "eta_seconds": 942.8176315055745, "mean_data_wait_seconds": 0.00019937753677368164}
{"step": 38660, "total_steps": 46238, "tokens": 20902920192, "loss": 2.942601203918457, "ce_loss": 2.942601203918457, "z_loss": 35.39230739709103, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14632071554660797, "tokens_per_second": 4350835.649181098, "eta_seconds": 941.5827381967564, "mean_data_wait_seconds": 9.268522262573242e-05}
{"step": 38670, "total_steps": 46238, "tokens": 20908326912, "loss": 2.962667268334013, "ce_loss": 2.962667268334013, "z_loss": 36.565316171357125, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1319502741098404, "tokens_per_second": 4350849.60986537, "eta_seconds": 940.3370354891667, "mean_data_wait_seconds": 8.347630500793457e-05}
{"step": 38680, "total_steps": 46238, "tokens": 20913733632, "loss": 2.935855381416552, "ce_loss": 2.935855381416552, "z_loss": 34.950423963142164, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14872656762599945, "tokens_per_second": 4350865.618307278, "eta_seconds": 939.090898787543, "mean_data_wait_seconds": 0.00028762221336364746}
{"step": 38690, "total_steps": 46238, "tokens": 20919140352, "loss": 2.9532473936225427, "ce_loss": 2.9532473936225427, "z_loss": 35.60825631112763, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18131951987743378, "tokens_per_second": 4350763.675365824, "eta_seconds": 937.8701966975727, "mean_data_wait_seconds": 7.075071334838867e-05}
{"step": 38700, "total_steps": 46238, "tokens": 20924547072, "loss": 2.9705623334104363, "ce_loss": 2.9705623334104363, "z_loss": 36.12204358071992, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.19288738071918488, "tokens_per_second": 4350755.516245374, "eta_seconds": 936.6292472155944, "mean_data_wait_seconds": 0.00010845065116882324}
{"step": 38710, "total_steps": 46238, "tokens": 20929953792, "loss": 2.9469518589250967, "ce_loss": 2.9469518589250967, "z_loss": 35.49188567652847, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14998595416545868, "tokens_per_second": 4350881.668116724, "eta_seconds": 935.3594177985401, "mean_data_wait_seconds": 8.147954940795898e-05}
{"step": 38720, "total_steps": 46238, "tokens": 20935360512, "loss": 3.0721378922462463, "ce_loss": 3.0721378922462463, "z_loss": 34.25100150252833, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16982436180114746, "tokens_per_second": 4350948.428092361, "eta_seconds": 934.1024126507355, "mean_data_wait_seconds": 8.58604907989502e-05}
{"step": 38730, "total_steps": 46238, "tokens": 20940767232, "loss": 2.9334552613171665, "ce_loss": 2.9334552613171665, "z_loss": 35.9915216330326, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16398972272872925, "tokens_per_second": 4350974.30853142, "eta_seconds": 932.8542106170171, "mean_data_wait_seconds": 6.648898124694824e-05}
{"step": 38740, "total_steps": 46238, "tokens": 20946173952, "loss": 2.956957732186173, "ce_loss": 2.956957732186173, "z_loss": 35.61692038449374, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14496786892414093, "tokens_per_second": 4351082.414944287, "eta_seconds": 931.5884181090377, "mean_data_wait_seconds": 0.0001436173915863037}
{"step": 38750, "total_steps": 46238, "tokens": 20951580672, "loss": 2.9882038325974434, "ce_loss": 2.9882038325974434, "z_loss": 35.700671080386996, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1747312694787979, "tokens_per_second": 4351194.173596949, "eta_seconds": 930.32190761868, "mean_data_wait_seconds": 0.0004674196243286133}
{"step": 38760, "total_steps": 46238, "tokens": 20956987392, "loss": 2.976921731775457, "ce_loss": 2.976921731775457, "z_loss": 35.05718953681715, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16815216839313507, "tokens_per_second": 4351238.8479216695, "eta_seconds": 929.0697857073311, "mean_data_wait_seconds": 7.233023643493652e-05}
{"step": 38770, "total_steps": 46238, "tokens": 20962394112, "loss": 2.994276852318735, "ce_loss": 2.994276852318735, "z_loss": 35.33357787854744, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15037979185581207, "tokens_per_second": 4351197.56080262, "eta_seconds": 927.8360192993168, "mean_data_wait_seconds": 7.82310962677002e-05}
{"step": 38780, "total_steps": 46238, "tokens": 20967800832, "loss": 3.053171244534579, "ce_loss": 3.053171244534579, "z_loss": 34.83365038669471, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1686573177576065, "tokens_per_second": 4351054.531350596, "eta_seconds": 926.6238965634167, "mean_data_wait_seconds": 8.654594421386719e-05}
{"step": 38790, "total_steps": 46238, "tokens": 20973207552, "loss": 3.0100984374682107, "ce_loss": 3.0100984374682107, "z_loss": 35.589264031612515, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1576756238937378, "tokens_per_second": 4351031.528481455, "eta_seconds": 925.3861659341368, "mean_data_wait_seconds": 0.00010150671005249023}
{"step": 38800, "total_steps": 46238, "tokens": 20978614272, "loss": 3.0556983983877934, "ce_loss": 3.0556983983877934, "z_loss": 34.65222029252486, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13978272676467896, "tokens_per_second": 4350969.569499754, "eta_seconds": 924.1566965181752, "mean_data_wait_seconds": 6.818771362304688e-05}
{"step": 38810, "total_steps": 46238, "tokens": 20984020992, "loss": 3.047506885095076, "ce_loss": 3.047506885095076, "z_loss": 35.07307934038567, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16553398966789246, "tokens_per_second": 4350940.450586612, "eta_seconds": 922.9202260073689, "mean_data_wait_seconds": 6.896257400512695e-05}
{"step": 38820, "total_steps": 46238, "tokens": 20989427712, "loss": 3.001878695054488, "ce_loss": 3.001878695054488, "z_loss": 36.04719502998121, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16862145066261292, "tokens_per_second": 4350941.563998526, "eta_seconds": 921.6773346674528, "mean_data_wait_seconds": 8.577108383178711e-05}
{"step": 38830, "total_steps": 46238, "tokens": 20994834432, "loss": 3.0576531399380076, "ce_loss": 3.0576531399380076, "z_loss": 34.95137844663678, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18220683932304382, "tokens_per_second": 4351033.697274712, "eta_seconds": 920.4151892706316, "mean_data_wait_seconds": 6.22868537902832e-05}
{"step": 38840, "total_steps": 46238, "tokens": 21000241152, "loss": 2.994549415328286, "ce_loss": 2.994549415328286, "z_loss": 36.353456815083824, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1524830311536789, "tokens_per_second": 4351131.11174642, "eta_seconds": 919.1519817004031, "mean_data_wait_seconds": 6.0945749282836914e-05}
{"step": 38850, "total_steps": 46238, "tokens": 21005647872, "loss": 2.9764032887689993, "ce_loss": 2.9764032887689993, "z_loss": 36.672586903427586, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15374992787837982, "tokens_per_second": 4351212.092972059, "eta_seconds": 917.8922972867475, "mean_data_wait_seconds": 5.8263540267944336e-05}
{"step": 38860, "total_steps": 46238, "tokens": 21011054592, "loss": 3.0096378127733865, "ce_loss": 3.0096378127733865, "z_loss": 34.67850358558424, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16889429092407227, "tokens_per_second": 4351286.724201718, "eta_seconds": 916.6339974370989, "mean_data_wait_seconds": 7.933378219604492e-05}
{"step": 38870, "total_steps": 46238, "tokens": 21016461312, "loss": 3.032099723815918, "ce_loss": 3.032099723815918, "z_loss": 35.10903601212935, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18952146172523499, "tokens_per_second": 4351327.463563688, "eta_seconds": 915.3828704810603, "mean_data_wait_seconds": 9.942054748535156e-05}
{"step": 38880, "total_steps": 46238, "tokens": 21021868032, "loss": 3.0004799221501206, "ce_loss": 3.0004799221501206, "z_loss": 34.950218027288265, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1543351411819458, "tokens_per_second": 4351286.558890181, "eta_seconds": 914.1489189842142, "mean_data_wait_seconds": 6.583333015441895e-05}
{"step": 38890, "total_steps": 46238, "tokens": 21027274752, "loss": 2.9708792783997278, "ce_loss": 2.9708792783997278, "z_loss": 36.539483937350184, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15004442632198334, "tokens_per_second": 4351342.060562815, "eta_seconds": 912.8947181610929, "mean_data_wait_seconds": 9.968876838684082e-05}
{"step": 38900, "total_steps": 46238, "tokens": 21032681472, "loss": 3.0339323787978203, "ce_loss": 3.0339323787978203, "z_loss": 36.014768282572426, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.166036918759346, "tokens_per_second": 4351347.6402830025, "eta_seconds": 911.6510083625496, "mean_data_wait_seconds": 7.522106170654297e-05}
{"step": 38910, "total_steps": 46238, "tokens": 21038088192, "loss": 2.991200642152266, "ce_loss": 2.991200642152266, "z_loss": 34.593267411896676, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1725490540266037, "tokens_per_second": 4351436.066779107, "eta_seconds": 910.3899685540523, "mean_data_wait_seconds": 7.316470146179199e-05}
{"step": 38920, "total_steps": 46238, "tokens": 21043494912, "loss": 3.0061602646654304, "ce_loss": 3.0061602646654304, "z_loss": 36.33596374049331, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16138401627540588, "tokens_per_second": 4351521.213401495, "eta_seconds": 909.1296652343791, "mean_data_wait_seconds": 8.755922317504883e-05}
{"step": 38930, "total_steps": 46238, "tokens": 21048901632, "loss": 2.9941019759033667, "ce_loss": 2.9941019759033667, "z_loss": 34.88930852485426, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.11715377867221832, "tokens_per_second": 4351603.147714968, "eta_seconds": 907.870081414596, "mean_data_wait_seconds": 9.655952453613281e-05}
{"step": 38940, "total_steps": 46238, "tokens": 21054308352, "loss": 2.990355022025831, "ce_loss": 2.990355022025831, "z_loss": 35.390068516586766, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1713232547044754, "tokens_per_second": 4351611.959576626, "eta_seconds": 906.62577928567, "mean_data_wait_seconds": 7.44163990020752e-05}
{"step": 38950, "total_steps": 46238, "tokens": 21059715072, "loss": 2.967629454352639, "ce_loss": 2.967629454352639, "z_loss": 35.12974294026693, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1551562249660492, "tokens_per_second": 4351678.927173802, "eta_seconds": 905.3693826990939, "mean_data_wait_seconds": 8.749961853027344e-05}
{"step": 38960, "total_steps": 46238, "tokens": 21065121792, "loss": 3.023912884972312, "ce_loss": 3.023912884972312, "z_loss": 35.46648863590125, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17633754014968872, "tokens_per_second": 4351743.511574284, "eta_seconds": 904.1135199111651, "mean_data_wait_seconds": 7.590651512145996e-05}
{"step": 38970, "total_steps": 46238, "tokens": 21070528512, "loss": 2.9995925281987046, "ce_loss": 2.9995925281987046, "z_loss": 36.24386177640973, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16283266246318817, "tokens_per_second": 4351785.662226488, "eta_seconds": 902.8623486915455, "mean_data_wait_seconds": 7.578730583190918e-05}
{"step": 38980, "total_steps": 46238, "tokens": 21075935232, "loss": 2.985789542848414, "ce_loss": 2.985789542848414, "z_loss": 34.952983393813625, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14888998866081238, "tokens_per_second": 4351890.109244452, "eta_seconds": 901.5982953395854, "mean_data_wait_seconds": 6.1005353927612305e-05}
{"step": 38990, "total_steps": 46238, "tokens": 21081341952, "loss": 3.0154137846195335, "ce_loss": 3.0154137846195335, "z_loss": 35.00829115780917, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13548077642917633, "tokens_per_second": 4351989.110090059, "eta_seconds": 900.3354293593158, "mean_data_wait_seconds": 0.00015878677368164062}
{"step": 39000, "total_steps": 46238, "tokens": 21086748672, "loss": 2.9831078377636997, "ce_loss": 2.9831078377636997, "z_loss": 36.131257664073594, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17289647459983826, "tokens_per_second": 4351969.234601346, "eta_seconds": 899.097179476828, "mean_data_wait_seconds": 0.0001970529556274414}
{"step": 39010, "total_steps": 46238, "tokens": 21092155392, "loss": 2.9593736908652564, "ce_loss": 2.9593736908652564, "z_loss": 35.2688227855798, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16479690372943878, "tokens_per_second": 4351926.4115278805, "eta_seconds": 897.8636526687435, "mean_data_wait_seconds": 9.28640365600586e-05}
{"step": 39020, "total_steps": 46238, "tokens": 21097562112, "loss": 2.9733482563134395, "ce_loss": 2.9733482563134395, "z_loss": 34.83027397502553, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1714484989643097, "tokens_per_second": 4351912.258538863, "eta_seconds": 896.6241946500298, "mean_data_wait_seconds": 0.00017008185386657715}
{"step": 39030, "total_steps": 46238, "tokens": 21102968832, "loss": 3.006121586669575, "ce_loss": 3.006121586669575, "z_loss": 35.00714059309526, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17589671909809113, "tokens_per_second": 4351906.965942972, "eta_seconds": 895.3829055846277, "mean_data_wait_seconds": 7.18533992767334e-05}
{"step": 39040, "total_steps": 46238, "tokens": 21108375552, "loss": 2.9892741387540642, "ce_loss": 2.9892741387540642, "z_loss": 34.46919319846413, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14198267459869385, "tokens_per_second": 4351947.155746962, "eta_seconds": 894.1322687848946, "mean_data_wait_seconds": 6.967782974243164e-05}
{"step": 39050, "total_steps": 46238, "tokens": 21113782272, "loss": 3.015395704543952, "ce_loss": 3.015395704543952, "z_loss": 36.284778797265254, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.187875434756279, "tokens_per_second": 4351931.957468258, "eta_seconds": 892.8930190031222, "mean_data_wait_seconds": 0.00011545419692993164}
{"step": 39060, "total_steps": 46238, "tokens": 21119188992, "loss": 2.9601498155882866, "ce_loss": 2.9601498155882866, "z_loss": 37.092536955168754, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2010209858417511, "tokens_per_second": 4351959.887453747, "eta_seconds": 891.644924206862, "mean_data_wait_seconds": 0.00015395879745483398}
{"step": 39070, "total_steps": 46238, "tokens": 21124595712, "loss": 2.9131917863181145, "ce_loss": 2.9131917863181145, "z_loss": 35.65672253117417, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13196463882923126, "tokens_per_second": 4352079.871149643, "eta_seconds": 890.3780120598713, "mean_data_wait_seconds": 8.183717727661133e-05}
{"step": 39080, "total_steps": 46238, "tokens": 21130002432, "loss": 2.9352131792993257, "ce_loss": 2.9352131792993257, "z_loss": 35.892443050037734, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2318151742219925, "tokens_per_second": 4352131.724767901, "eta_seconds": 889.1250882821946, "mean_data_wait_seconds": 0.0001386106014251709}
{"step": 39090, "total_steps": 46238, "tokens": 21135409152, "loss": 2.946600104823257, "ce_loss": 2.946600104823257, "z_loss": 35.54676659901937, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.16565027832984924, "tokens_per_second": 4352190.53456065, "eta_seconds": 887.8707752600924, "mean_data_wait_seconds": 0.00010696053504943848}
{"step": 39100, "total_steps": 46238, "tokens": 21140815872, "loss": 2.9201171867775195, "ce_loss": 2.9201171867775195, "z_loss": 37.44418202024518, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.20378126204013824, "tokens_per_second": 4352226.268737295, "eta_seconds": 886.6211969993788, "mean_data_wait_seconds": 0.00011181831359863281}
{"step": 39110, "total_steps": 46238, "tokens": 21146222592, "loss": 2.9202642513043955, "ce_loss": 2.9202642513043955, "z_loss": 35.61780588554613, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14547792077064514, "tokens_per_second": 4352262.129172309, "eta_seconds": 885.3716135734714, "mean_data_wait_seconds": 0.00010702013969421387}
{"step": 39120, "total_steps": 46238, "tokens": 21151629312, "loss": 2.992793746066816, "ce_loss": 2.992793746066816, "z_loss": 34.91832426822547, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15747670829296112, "tokens_per_second": 4352355.906984361, "eta_seconds": 884.1102856099279, "mean_data_wait_seconds": 0.00010091066360473633}
{"step": 39130, "total_steps": 46238, "tokens": 21157036032, "loss": 2.8985343557415586, "ce_loss": 2.8985343557415586, "z_loss": 36.94872269485936, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1354416310787201, "tokens_per_second": 4352404.98094677, "eta_seconds": 882.858079802155, "mean_data_wait_seconds": 5.906820297241211e-05}
{"step": 39140, "total_steps": 46238, "tokens": 21162442752, "loss": 2.9093178203611663, "ce_loss": 2.9093178203611663, "z_loss": 35.21728090806441, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.13945764303207397, "tokens_per_second": 4352532.807487111, "eta_seconds": 881.5899508900745, "mean_data_wait_seconds": 7.322430610656738e-05}
{"step": 39150, "total_steps": 46238, "tokens": 21167849472, "loss": 2.940474504774267, "ce_loss": 2.940474504774267, "z_loss": 35.17320745641535, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18429437279701233, "tokens_per_second": 4352645.589940945, "eta_seconds": 880.3249391255831, "mean_data_wait_seconds": 0.00010079145431518555}
{"step": 39160, "total_steps": 46238, "tokens": 21173256192, "loss": 2.952872771205324, "ce_loss": 2.952872771205324, "z_loss": 35.90868103142941, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1772632896900177, "tokens_per_second": 4352786.547156196, "eta_seconds": 879.054302926905, "mean_data_wait_seconds": 6.565451622009277e-05}
{"step": 39170, "total_steps": 46238, "tokens": 21178662912, "loss": 2.966035586414915, "ce_loss": 2.966035586414915, "z_loss": 36.571203058416195, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1560622751712799, "tokens_per_second": 4352929.193839062, "eta_seconds": 877.783408332938, "mean_data_wait_seconds": 7.739663124084473e-05}
{"step": 39180, "total_steps": 46238, "tokens": 21184069632, "loss": 3.0069260488856924, "ce_loss": 3.0069260488856924, "z_loss": 34.92819872769442, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1936711072921753, "tokens_per_second": 4353044.623080408, "eta_seconds": 876.5180774324263, "mean_data_wait_seconds": 0.00011077523231506348}
{"step": 39190, "total_steps": 46238, "tokens": 21189476352, "loss": 2.8998872583562676, "ce_loss": 2.8998872583562676, "z_loss": 34.668351144501656, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18607454001903534, "tokens_per_second": 4353046.085578464, "eta_seconds": 875.2757285577151, "mean_data_wait_seconds": 7.489323616027832e-05}
{"step": 39200, "total_steps": 46238, "tokens": 21194883072, "loss": 2.9343154773567663, "ce_loss": 2.9343154773567663, "z_loss": 35.41613966045958, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.2004527747631073, "tokens_per_second": 4353115.583904697, "eta_seconds": 874.0197200523718, "mean_data_wait_seconds": 7.608532905578613e-05}
{"step": 39210, "total_steps": 46238, "tokens": 21200289792, "loss": 2.9981900652249656, "ce_loss": 2.9981900652249656, "z_loss": 34.89196679086396, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1567467600107193, "tokens_per_second": 4353086.054978633, "eta_seconds": 872.7836059327911, "mean_data_wait_seconds": 7.045269012451172e-05}
{"step": 39220, "total_steps": 46238, "tokens": 21205696512, "loss": 2.952862027919654, "ce_loss": 2.952862027919654, "z_loss": 35.780254392912894, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14177514612674713, "tokens_per_second": 4353077.717456589, "eta_seconds": 871.5432322252891, "mean_data_wait_seconds": 7.888674736022949e-05}
{"step": 39230, "total_steps": 46238, "tokens": 21211103232, "loss": 2.966177786841537, "ce_loss": 2.966177786841537, "z_loss": 35.0933300076109, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17305058240890503, "tokens_per_second": 4353080.217773862, "eta_seconds": 870.3006869782448, "mean_data_wait_seconds": 8.359551429748535e-05}
{"step": 39240, "total_steps": 46238, "tokens": 21216509952, "loss": 2.990933308095643, "ce_loss": 2.990933308095643, "z_loss": 36.55354878396699, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15285727381706238, "tokens_per_second": 4353060.248216864, "eta_seconds": 869.0626291123945, "mean_data_wait_seconds": 0.00010719895362854004}
{"step": 39250, "total_steps": 46238, "tokens": 21221916672, "loss": 2.9795346639373084, "ce_loss": 2.9795346639373084, "z_loss": 36.50637921420011, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17941682040691376, "tokens_per_second": 4353089.070626709, "eta_seconds": 867.8148328024294, "mean_data_wait_seconds": 8.478760719299316e-05}
{"step": 39260, "total_steps": 46238, "tokens": 21227323392, "loss": 2.93476326899095, "ce_loss": 2.93476326899095, "z_loss": 35.67744951537161, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.17646639049053192, "tokens_per_second": 4352496.518414263, "eta_seconds": 866.6907665613352, "mean_data_wait_seconds": 8.216500282287598e-05}
{"step": 39270, "total_steps": 46238, "tokens": 21232730112, "loss": 2.9017954938339465, "ce_loss": 2.9017954938339465, "z_loss": 36.19937012412331, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.1746920943260193, "tokens_per_second": 4352073.271380388, "eta_seconds": 865.532721788305, "mean_data_wait_seconds": 0.0001589655876159668}
{"step": 39280, "total_steps": 46238, "tokens": 21238136832, "loss": 2.904569360342893, "ce_loss": 2.904569360342893, "z_loss": 35.64177229910186, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.14525873959064484, "tokens_per_second": 4352120.574807645, "eta_seconds": 864.2809957456772, "mean_data_wait_seconds": 0.0001685023307800293}
{"step": 39290, "total_steps": 46238, "tokens": 21243543552, "loss": 2.963048476161379, "ce_loss": 2.963048476161379, "z_loss": 36.45491839900161, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.18519151210784912, "tokens_per_second": 4352209.8860385865, "eta_seconds": 863.0209668998255, "mean_data_wait_seconds": 0.00010222196578979492}
{"step": 39300, "total_steps": 46238, "tokens": 21248950272, "loss": 2.969556122115164, "ce_loss": 2.969556122115164, "z_loss": 36.57046081080581, "muon_lr": 0.05, "residual_lr": 0.0003, "adamw_lr": 0.003, "weight_decay": 0.01, "grad_norm": 0.15167218446731567, "tokens_per_second": 4351825.389935648, "eta_seconds": 861.854814458781, "mean_data_wait_seconds": 0.00010842084884643555}
{"step": 39310, "total_steps": 46238, "tokens": 21254356992, "loss": 2.9648671186331548, "ce_loss": 2.9648671186331548, "z_loss": 35.86238716587876, "muon_lr": 0.049999874306579845, "residual_lr": 0.000299999245839479, "adamw_lr": 0.0029999924583947904, "weight_decay": 0.01, "grad_norm": 0.17165958881378174, "tokens_per_second": 4351881.4931373885, "eta_seconds": 860.601316903039, "mean_data_wait_seconds": 0.00010114908218383789}
{"step": 39320, "total_steps": 46238, "tokens": 21259763712, "loss": 2.980864830089338, "ce_loss": 2.980864830089338, "z_loss": 36.02076833898371, "muon_lr": 0.04999925866838089, "residual_lr": 0.0002999955520102853, "adamw_lr": 0.002999955520102853, "weight_decay": 0.01, "grad_norm": 0.15323780477046967, "tokens_per_second": 4351413.610490001, "eta_seconds": 859.4513320876588, "mean_data_wait_seconds": 8.690357208251953e-05}
{"step": 39330, "total_steps": 46238, "tokens": 21265170432, "loss": 2.93268084526062, "ce_loss": 2.93268084526062, "z_loss": 35.070823698332816, "muon_lr": 0.049998130011474524, "residual_lr": 0.00029998878006884714, "adamw_lr": 0.0029998878006884714, "weight_decay": 0.01, "grad_norm": 0.1548638790845871, "tokens_per_second": 4351452.944816447, "eta_seconds": 858.2010540751752, "mean_data_wait_seconds": 8.410215377807617e-05}
{"step": 39340, "total_steps": 46238, "tokens": 21270577152, "loss": 3.2084243749127244, "ce_loss": 3.2084243749127244, "z_loss": 34.09319733128403, "muon_lr": 0.04999648835902232, "residual_lr": 0.0002999789301541339, "adamw_lr": 0.0029997893015413393, "weight_decay": 0.01, "grad_norm": 0.2008768618106842, "tokens_per_second": 4351447.881175521, "eta_seconds": 856.9595421633836, "mean_data_wait_seconds": 8.413195610046387e-05}
{"step": 39350, "total_steps": 46238, "tokens": 21275983872, "loss": 3.005756201166095, "ce_loss": 3.005756201166095, "z_loss": 36.74916004412102, "muon_lr": 0.049994333744713196, "residual_lr": 0.00029996600246827914, "adamw_lr": 0.0029996600246827917, "weight_decay": 0.01, "grad_norm": 0.1615646481513977, "tokens_per_second": 4351433.774431569, "eta_seconds": 855.719805706205, "mean_data_wait_seconds": 0.00011390447616577148}
{"step": 39360, "total_steps": 46238, "tokens": 21281390592, "loss": 3.0546450741363294, "ce_loss": 3.0546450741363294, "z_loss": 34.93028213038589, "muon_lr": 0.049991666212762784, "residual_lr": 0.0002999499972765767, "adamw_lr": 0.002999499972765767, "weight_decay": 0.01, "grad_norm": 0.14020667970180511, "tokens_per_second": 4351435.082425577, "eta_seconds": 854.4770342586382, "mean_data_wait_seconds": 7.843971252441406e-05}
{"step": 39370, "total_steps": 46238, "tokens": 21286797312, "loss": 3.004742654887113, "ce_loss": 3.004742654887113, "z_loss": 35.96047170234449, "muon_lr": 0.049988485817912445, "residual_lr": 0.0002999309149074746, "adamw_lr": 0.002999309149074747, "weight_decay": 0.01, "grad_norm": 0.1584450751543045, "tokens_per_second": 4351450.895072125, "eta_seconds": 853.2314194800217, "mean_data_wait_seconds": 8.356571197509766e-05}
{"step": 39380, "total_steps": 46238, "tokens": 21292204032, "loss": 3.0021079435493006, "ce_loss": 3.0021079435493006, "z_loss": 36.074402115561746, "muon_lr": 0.0499847926254282, "residual_lr": 0.0002999087557525692, "adamw_lr": 0.002999087557525692, "weight_decay": 0.01, "grad_norm": 0.13303498923778534, "tokens_per_second": 4351528.711111256, "eta_seconds": 851.97367410986, "mean_data_wait_seconds": 6.985664367675781e-05}
{"step": 39390, "total_steps": 46238, "tokens": 21297610752, "loss": 2.9715647769696787, "ce_loss": 2.9715647769696787, "z_loss": 36.53139807961204, "muon_lr": 0.04998058671109937, "residual_lr": 0.00029988352026659616, "adamw_lr": 0.0029988352026659617, "weight_decay": 0.01, "grad_norm": 0.16598165035247803, "tokens_per_second": 4351602.150424243, "eta_seconds": 850.7168293496429, "mean_data_wait_seconds": 6.222724914550781e-05}
{"step": 39400, "total_steps": 46238, "tokens": 21303017472, "loss": 3.0666509957024544, "ce_loss": 3.0666509957024544, "z_loss": 34.4419079693881, "muon_lr": 0.049975868161237016, "residual_lr": 0.00029985520896742206, "adamw_lr": 0.002998552089674221, "weight_decay": 0.01, "grad_norm": 0.17874740064144135, "tokens_per_second": 4351658.0153145, "eta_seconds": 849.4634576041803, "mean_data_wait_seconds": 6.186962127685547e-05}
{"step": 39410, "total_steps": 46238, "tokens": 21308424192, "loss": 2.9798638116229665, "ce_loss": 2.9798638116229665, "z_loss": 35.87934031631007, "muon_lr": 0.0499706370726722, "residual_lr": 0.0002998238224360332, "adamw_lr": 0.002998238224360332, "weight_decay": 0.01, "grad_norm": 0.13866961002349854, "tokens_per_second": 4351650.228041303, "eta_seconds": 848.2225249204853, "mean_data_wait_seconds": 6.866455078125e-05}
{"step": 39420, "total_steps": 46238, "tokens": 21313830912, "loss": 3.000629518971299, "ce_loss": 3.000629518971299, "z_loss": 35.911454142946184, "muon_lr": 0.04996489355275394, "residual_lr": 0.00029978936131652364, "adamw_lr": 0.0029978936131652366, "weight_decay": 0.01, "grad_norm": 0.1435914933681488, "tokens_per_second": 4351707.57354245, "eta_seconds": 846.9689108727624, "mean_data_wait_seconds": 6.99758529663086e-05}
{"step": 39430, "total_steps": 46238, "tokens": 21319237632, "loss": 3.000394116748463, "ce_loss": 3.000394116748463, "z_loss": 36.1391373836633, "muon_lr": 0.04995863771934708, "residual_lr": 0.0002997518263160824, "adamw_lr": 0.0029975182631608245, "weight_decay": 0.01, "grad_norm": 0.16554349660873413, "tokens_per_second": 4351784.421280268, "eta_seconds": 845.7115398462828, "mean_data_wait_seconds": 5.957484245300293e-05}
{"step": 39440, "total_steps": 46238, "tokens": 21324644352, "loss": 3.0187144243355952, "ce_loss": 3.0187144243355952, "z_loss": 35.80712780807958, "muon_lr": 0.04995186970082978, "residual_lr": 0.00029971121820497864, "adamw_lr": 0.002997112182049787, "weight_decay": 0.01, "grad_norm": 0.1684526801109314, "tokens_per_second": 4351796.916665917, "eta_seconds": 844.4667006234109, "mean_data_wait_seconds": 0.00010004639625549316}
{"step": 39450, "total_steps": 46238, "tokens": 21330051072, "loss": 3.0196267637339504, "ce_loss": 3.0196267637339504, "z_loss": 35.716811064517856, "muon_lr": 0.04994458963609099, "residual_lr": 0.0002996675378165459, "adamw_lr": 0.0029966753781654593, "weight_decay": 0.01, "grad_norm": 0.19672764837741852, "tokens_per_second": 4351894.213112544, "eta_seconds": 843.2054375180884, "mean_data_wait_seconds": 0.00010123848915100098}
{"step": 39460, "total_steps": 46238, "tokens": 21335457792, "loss": 3.047783315181732, "ce_loss": 3.047783315181732, "z_loss": 35.41035079956055, "muon_lr": 0.049936797674527494, "residual_lr": 0.0002996207860471649, "adamw_lr": 0.0029962078604716495, "weight_decay": 0.01, "grad_norm": 0.1455022245645523, "tokens_per_second": 4351994.966736431, "eta_seconds": 841.9435619769893, "mean_data_wait_seconds": 8.19861888885498e-05}
{"step": 39470, "total_steps": 46238, "tokens": 21340864512, "loss": 3.026491287982825, "ce_loss": 3.026491287982825, "z_loss": 36.29866490219579, "muon_lr": 0.04992849397604093, "residual_lr": 0.00029957096385624556, "adamw_lr": 0.0029957096385624556, "weight_decay": 0.01, "grad_norm": 0.1516106277704239, "tokens_per_second": 4352068.1733847475, "eta_seconds": 840.6870660655314, "mean_data_wait_seconds": 7.581710815429688e-05}
{"step": 39480, "total_steps": 46238, "tokens": 21346271232, "loss": 3.012485738956567, "ce_loss": 3.012485738956567, "z_loss": 34.95573428182891, "muon_lr": 0.049919678711034456, "residual_lr": 0.0002995180722662067, "adamw_lr": 0.0029951807226620674, "weight_decay": 0.01, "grad_norm": 0.16999295353889465, "tokens_per_second": 4352181.091103195, "eta_seconds": 839.422953118422, "mean_data_wait_seconds": 0.0001036226749420166}
{"step": 39490, "total_steps": 46238, "tokens": 21351677952, "loss": 3.000470221042633, "ce_loss": 3.000470221042633, "z_loss": 36.94458198547363, "muon_lr": 0.04991035206040926, "residual_lr": 0.0002994621123624555, "adamw_lr": 0.0029946211236245557, "weight_decay": 0.01, "grad_norm": 0.148947611451149, "tokens_per_second": 4352144.312925325, "eta_seconds": 838.1877349898879, "mean_data_wait_seconds": 9.97781753540039e-05}
{"step": 39500, "total_steps": 46238, "tokens": 21357084672, "loss": 3.000264532638319, "ce_loss": 3.000264532638319, "z_loss": 36.26224543831565, "muon_lr": 0.04990051421556088, "residual_lr": 0.00029940308529336526, "adamw_lr": 0.002994030852933653, "weight_decay": 0.01, "grad_norm": 0.1588149517774582, "tokens_per_second": 4352174.963480085, "eta_seconds": 836.9395289860726, "mean_data_wait_seconds": 7.826089859008789e-05}
{"step": 39510, "total_steps": 46238, "tokens": 21362491392, "loss": 3.018111628113371, "ce_loss": 3.018111628113371, "z_loss": 36.11977860421845, "muon_lr": 0.04989016537837524, "residual_lr": 0.0002993409922702514, "adamw_lr": 0.0029934099227025145, "weight_decay": 0.01, "grad_norm": 0.16085319221019745, "tokens_per_second": 4352246.385770675, "eta_seconds": 835.6835118276421, "mean_data_wait_seconds": 6.720423698425293e-05}
{"step": 39520, "total_steps": 46238, "tokens": 21367898112, "loss": 2.9956862818111074, "ce_loss": 2.9956862818111074, "z_loss": 35.42202088327119, "muon_lr": 0.04987930576122451, "residual_lr": 0.00029927583456734703, "adamw_lr": 0.0029927583456734706, "weight_decay": 0.01, "grad_norm": 0.16595110297203064, "tokens_per_second": 4352147.203313562, "eta_seconds": 834.4602455622282, "mean_data_wait_seconds": 0.00010675191879272461}
{"step": 39530, "total_steps": 46238, "tokens": 21373304832, "loss": 3.004744356328791, "ce_loss": 3.004744356328791, "z_loss": 36.48270095478404, "muon_lr": 0.049867935586962786, "residual_lr": 0.00029920761352177663, "adamw_lr": 0.002992076135217767, "weight_decay": 0.01, "grad_norm": 0.1621471792459488, "tokens_per_second": 4352239.522155459, "eta_seconds": 833.2002605876964, "mean_data_wait_seconds": 0.00010094046592712402}
{"step": 39540, "total_steps": 46238, "tokens": 21378711552, "loss": 3.014596012505618, "ce_loss": 3.014596012505618, "z_loss": 35.3275170470729, "muon_lr": 0.04985605508892145, "residual_lr": 0.00029913633053352864, "adamw_lr": 0.002991363305335287, "weight_decay": 0.01, "grad_norm": 0.1633499264717102, "tokens_per_second": 4352320.481219232, "eta_seconds": 831.9425004717642, "mean_data_wait_seconds": 8.133053779602051e-05}
{"step": 39550, "total_steps": 46238, "tokens": 21384118272, "loss": 2.9651235323963743, "ce_loss": 2.9651235323963743, "z_loss": 35.18016627340606, "muon_lr": 0.049843664510904453, "residual_lr": 0.00029906198706542667, "adamw_lr": 0.002990619870654267, "weight_decay": 0.01, "grad_norm": 0.18552963435649872, "tokens_per_second": 4352381.167628028, "eta_seconds": 830.6886563362211, "mean_data_wait_seconds": 5.4955482482910156e-05}
{"step": 39560, "total_steps": 46238, "tokens": 21389524992, "loss": 3.030215620994568, "ce_loss": 3.030215620994568, "z_loss": 35.91841596545595, "muon_lr": 0.04983076410718327, "residual_lr": 0.0002989845846430996, "adamw_lr": 0.002989845846430996, "weight_decay": 0.01, "grad_norm": 0.16145679354667664, "tokens_per_second": 4352420.9062747015, "eta_seconds": 829.4388391516819, "mean_data_wait_seconds": 6.181001663208008e-05}
{"step": 39570, "total_steps": 46238, "tokens": 21394931712, "loss": 3.017123545661117, "ce_loss": 3.017123545661117, "z_loss": 35.59213649865353, "muon_lr": 0.049817354142491686, "residual_lr": 0.00029890412485495006, "adamw_lr": 0.002989041248549501, "weight_decay": 0.01, "grad_norm": 0.14243797957897186, "tokens_per_second": 4352504.299228788, "eta_seconds": 828.1807383024762, "mean_data_wait_seconds": 7.238984107971191e-05}
{"step": 39580, "total_steps": 46238, "tokens": 21400338432, "loss": 2.9943175695159217, "ce_loss": 2.9943175695159217, "z_loss": 36.0112598881577, "muon_lr": 0.04980343489202036, "residual_lr": 0.00029882060935212213, "adamw_lr": 0.0029882060935212218, "weight_decay": 0.01, "grad_norm": 0.16944469511508942, "tokens_per_second": 4352640.287156828, "eta_seconds": 826.9126935713439, "mean_data_wait_seconds": 0.00016182661056518555}
{"step": 39590, "total_steps": 46238, "tokens": 21405745152, "loss": 2.975775057619268, "ce_loss": 2.975775057619268, "z_loss": 36.967835224036016, "muon_lr": 0.049789006641411226, "residual_lr": 0.0002987340398484673, "adamw_lr": 0.0029873403984846734, "weight_decay": 0.01, "grad_norm": 0.1737080067396164, "tokens_per_second": 4352745.505925324, "eta_seconds": 825.6505644788451, "mean_data_wait_seconds": 0.00018212199211120605}
{"step": 39600, "total_steps": 46238, "tokens": 21411151872, "loss": 2.9284716085954146, "ce_loss": 2.9284716085954146, "z_loss": 35.42736706589208, "muon_lr": 0.049774069686751535, "residual_lr": 0.00029864441812050917, "adamw_lr": 0.002986444181205092, "weight_decay": 0.01, "grad_norm": 0.16565777361392975, "tokens_per_second": 4352772.100999459, "eta_seconds": 824.4033872520096, "mean_data_wait_seconds": 0.0001507103443145752}
{"step": 39610, "total_steps": 46238, "tokens": 21416558592, "loss": 2.9961098902153247, "ce_loss": 2.9961098902153247, "z_loss": 36.296116424329355, "muon_lr": 0.049758624334567875, "residual_lr": 0.0002985517460074072, "adamw_lr": 0.002985517460074072, "weight_decay": 0.01, "grad_norm": 0.16646720468997955, "tokens_per_second": 4352830.783003131, "eta_seconds": 823.1501573621872, "mean_data_wait_seconds": 0.00010699033737182617}
{"step": 39620, "total_steps": 46238, "tokens": 21421965312, "loss": 2.94853981516578, "ce_loss": 2.94853981516578, "z_loss": 35.96800726110285, "muon_lr": 0.04974267090181983, "residual_lr": 0.0002984560254109189, "adamw_lr": 0.0029845602541091897, "weight_decay": 0.01, "grad_norm": 0.16333355009555817, "tokens_per_second": 4352892.521514861, "eta_seconds": 821.8963841438799, "mean_data_wait_seconds": 0.00046506524085998535}
{"step": 39630, "total_steps": 46238, "tokens": 21427372032, "loss": 2.963523151296558, "ce_loss": 2.963523151296558, "z_loss": 36.657971006451234, "muon_lr": 0.04972620971589348, "residual_lr": 0.0002983572582953608, "adamw_lr": 0.002983572582953609, "weight_decay": 0.01, "grad_norm": 0.16462236642837524, "tokens_per_second": 4352955.162925393, "eta_seconds": 820.6424762710623, "mean_data_wait_seconds": 0.00010463595390319824}
{"step": 39640, "total_steps": 46238, "tokens": 21432778752, "loss": 2.9423530463016396, "ce_loss": 2.9423530463016396, "z_loss": 36.04117358814586, "muon_lr": 0.049709241114594696, "residual_lr": 0.00029825544668756816, "adamw_lr": 0.0029825544668756815, "weight_decay": 0.01, "grad_norm": 0.1509174257516861, "tokens_per_second": 4353037.59992335, "eta_seconds": 819.3848782888542, "mean_data_wait_seconds": 0.0001360476016998291}
{"step": 39650, "total_steps": 46238, "tokens": 21438185472, "loss": 2.957871279933236, "ce_loss": 2.957871279933236, "z_loss": 36.47729168516217, "muon_lr": 0.04969176544614221, "residual_lr": 0.00029815059267685324, "adamw_lr": 0.0029815059267685328, "weight_decay": 0.01, "grad_norm": 0.1594589203596115, "tokens_per_second": 4353054.87728967, "eta_seconds": 818.139574251687, "mean_data_wait_seconds": 0.00015777349472045898}
{"step": 39660, "total_steps": 46238, "tokens": 21443592192, "loss": 2.9575952724976973, "ce_loss": 2.9575952724976973, "z_loss": 36.59935873205011, "muon_lr": 0.04967378306916045, "residual_lr": 0.0002980426984149626, "adamw_lr": 0.002980426984149627, "weight_decay": 0.01, "grad_norm": 0.13466405868530273, "tokens_per_second": 4352982.229969424, "eta_seconds": 816.9111556481079, "mean_data_wait_seconds": 0.00014922022819519043}
{"step": 39670, "total_steps": 46238, "tokens": 21448998912, "loss": 2.9754691774194892, "ce_loss": 2.9754691774194892, "z_loss": 35.70189984639486, "muon_lr": 0.04965529435267217, "residual_lr": 0.000297931766116033, "adamw_lr": 0.00297931766116033, "weight_decay": 0.01, "grad_norm": 0.16342990100383759, "tokens_per_second": 4352998.578020791, "eta_seconds": 815.666019724356, "mean_data_wait_seconds": 6.54757022857666e-05}
{"step": 39680, "total_steps": 46238, "tokens": 21454405632, "loss": 2.9569694219213543, "ce_loss": 2.9569694219213543, "z_loss": 36.56633033174457, "muon_lr": 0.04963629967609094, "residual_lr": 0.0002978177980565456, "adamw_lr": 0.0029781779805654563, "weight_decay": 0.01, "grad_norm": 0.14529989659786224, "tokens_per_second": 4353103.377088822, "eta_seconds": 814.4043448770278, "mean_data_wait_seconds": 0.00010320544242858887}
{"step": 39690, "total_steps": 46238, "tokens": 21459812352, "loss": 2.9847676302447463, "ce_loss": 2.9847676302447463, "z_loss": 36.50167364062685, "muon_lr": 0.04961679942921329, "residual_lr": 0.0002977007965752797, "adamw_lr": 0.0029770079657527974, "weight_decay": 0.01, "grad_norm": 0.1710987240076065, "tokens_per_second": 4353123.016623073, "eta_seconds": 813.1586381737443, "mean_data_wait_seconds": 0.00010496377944946289}
{"step": 39700, "total_steps": 46238, "tokens": 21465219072, "loss": 2.9092736605441933, "ce_loss": 2.9092736605441933, "z_loss": 35.685749227350406, "muon_lr": 0.04959679401221073, "residual_lr": 0.0002975807640732644, "adamw_lr": 0.002975807640732644, "weight_decay": 0.01, "grad_norm": 0.18792109191417694, "tokens_per_second": 4353125.516393366, "eta_seconds": 811.9161394933276, "mean_data_wait_seconds": 0.00012171268463134766}
{"step": 39710, "total_steps": 46238, "tokens": 21470625792, "loss": 2.8933452566464744, "ce_loss": 2.8933452566464744, "z_loss": 36.15659037503329, "muon_lr": 0.049576283835621585, "residual_lr": 0.0002974577030137295, "adamw_lr": 0.002974577030137295, "weight_decay": 0.01, "grad_norm": 0.14332465827465057, "tokens_per_second": 4353186.25029939, "eta_seconds": 810.6627975674818, "mean_data_wait_seconds": 9.992718696594238e-05}
{"step": 39720, "total_steps": 46238, "tokens": 21476032512, "loss": 2.9953370094299316, "ce_loss": 2.9953370094299316, "z_loss": 35.653032996437766, "muon_lr": 0.049555269320342495, "residual_lr": 0.0002973316159220549, "adamw_lr": 0.0029733161592205497, "weight_decay": 0.01, "grad_norm": 0.15716828405857086, "tokens_per_second": 4353212.649115998, "eta_seconds": 809.4158746679938, "mean_data_wait_seconds": 0.00016489624977111816}
{"step": 39730, "total_steps": 46238, "tokens": 21481439232, "loss": 2.968711990298647, "ce_loss": 2.968711990298647, "z_loss": 35.66756476777972, "muon_lr": 0.04953375089761983, "residual_lr": 0.0002972025053857189, "adamw_lr": 0.002972025053857189, "weight_decay": 0.01, "grad_norm": 0.17580735683441162, "tokens_per_second": 4353272.491613344, "eta_seconds": 808.1627581957672, "mean_data_wait_seconds": 7.56680965423584e-05}
{"step": 39740, "total_steps": 46238, "tokens": 21486845952, "loss": 2.9663787827347265, "ce_loss": 2.9663787827347265, "z_loss": 35.70039922540838, "muon_lr": 0.04951172900904081, "residual_lr": 0.0002970703740542448, "adamw_lr": 0.0029707037405424484, "weight_decay": 0.01, "grad_norm": 0.13628613948822021, "tokens_per_second": 4353289.051429619, "eta_seconds": 806.9176988939926, "mean_data_wait_seconds": 7.87973403930664e-05}
{"step": 39750, "total_steps": 46238, "tokens": 21492252672, "loss": 2.950499845273567, "ce_loss": 2.950499845273567, "z_loss": 36.28296548669989, "muon_lr": 0.04948920410652447, "residual_lr": 0.0002969352246391468, "adamw_lr": 0.002969352246391468, "weight_decay": 0.01, "grad_norm": 0.15524610877037048, "tokens_per_second": 4353302.21832578, "eta_seconds": 805.6732769977258, "mean_data_wait_seconds": 0.00010702013969421387}
{"step": 39760, "total_steps": 46238, "tokens": 21497659392, "loss": 2.903080116618763, "ce_loss": 2.903080116618763, "z_loss": 36.091158202200226, "muon_lr": 0.04946617665231239, "residual_lr": 0.00029679705991387433, "adamw_lr": 0.0029679705991387433, "weight_decay": 0.01, "grad_norm": 0.18558068573474884, "tokens_per_second": 4353403.4988199435, "eta_seconds": 804.4125808575411, "mean_data_wait_seconds": 7.396936416625977e-05}
{"step": 39770, "total_steps": 46238, "tokens": 21503066112, "loss": 2.9411415349353445, "ce_loss": 2.9411415349353445, "z_loss": 36.05505859490597, "muon_lr": 0.049442647118959154, "residual_lr": 0.00029665588271375484, "adamw_lr": 0.0029665588271375488, "weight_decay": 0.01, "grad_norm": 0.1711980253458023, "tokens_per_second": 4353508.183034214, "eta_seconds": 803.1513154439662, "mean_data_wait_seconds": 0.0003578662872314453}
{"step": 39780, "total_steps": 46238, "tokens": 21508472832, "loss": 3.005272056117202, "ce_loss": 3.005272056117202, "z_loss": 36.73764555382006, "muon_lr": 0.04941861598932273, "residual_lr": 0.00029651169593593633, "adamw_lr": 0.0029651169593593635, "weight_decay": 0.01, "grad_norm": 0.1735108345746994, "tokens_per_second": 4353431.438400831, "eta_seconds": 801.9235293808626, "mean_data_wait_seconds": 8.848309516906738e-05}
{"step": 39790, "total_steps": 46238, "tokens": 21513879552, "loss": 2.9448740861632605, "ce_loss": 2.9448740861632605, "z_loss": 36.12298685131651, "muon_lr": 0.049394083756554484, "residual_lr": 0.00029636450253932685, "adamw_lr": 0.0029636450253932686, "weight_decay": 0.01, "grad_norm": 0.16576263308525085, "tokens_per_second": 4353407.315045176, "eta_seconds": 800.6860217176412, "mean_data_wait_seconds": 9.98079776763916e-05}
{"step": 39800, "total_steps": 46238, "tokens": 21519286272, "loss": 2.9795605702833696, "ce_loss": 2.9795605702833696, "z_loss": 36.56560744661273, "muon_lr": 0.049369050924089136, "residual_lr": 0.00029621430554453476, "adamw_lr": 0.002962143055445348, "weight_decay": 0.01, "grad_norm": 0.1557600200176239, "tokens_per_second": 4353333.743527536, "eta_seconds": 799.4575810261413, "mean_data_wait_seconds": 0.00014659762382507324}
{"step": 39810, "total_steps": 46238, "tokens": 21524692992, "loss": 2.9255098989515593, "ce_loss": 2.9255098989515593, "z_loss": 36.823407259854406, "muon_lr": 0.04934351800563436, "residual_lr": 0.0002960611080338061, "adamw_lr": 0.0029606110803380614, "weight_decay": 0.01, "grad_norm": 0.135102316737175, "tokens_per_second": 4353272.879538716, "eta_seconds": 798.2267687221596, "mean_data_wait_seconds": 8.90493392944336e-05}
{"step": 39820, "total_steps": 46238, "tokens": 21530099712, "loss": 2.938514460216869, "ce_loss": 2.938514460216869, "z_loss": 35.50064170721805, "muon_lr": 0.0493174855251603, "residual_lr": 0.00029590491315096174, "adamw_lr": 0.0029590491315096175, "weight_decay": 0.01, "grad_norm": 0.16262711584568024, "tokens_per_second": 4353301.046733026, "eta_seconds": 796.9796223037944, "mean_data_wait_seconds": 8.878111839294434e-05}
{"step": 39830, "total_steps": 46238, "tokens": 21535506432, "loss": 2.9163954655329385, "ce_loss": 2.9163954655329385, "z_loss": 37.21144395886046, "muon_lr": 0.04929095401688876, "residual_lr": 0.0002957457241013325, "adamw_lr": 0.0029574572410133253, "weight_decay": 0.01, "grad_norm": 0.1536838710308075, "tokens_per_second": 4353298.079540995, "eta_seconds": 795.7381830295544, "mean_data_wait_seconds": 7.510185241699219e-05}
{"step": 39840, "total_steps": 46238, "tokens": 21540913152, "loss": 2.9419043335047634, "ce_loss": 2.9419043335047634, "z_loss": 36.64310738534638, "muon_lr": 0.049263924025282305, "residual_lr": 0.0002955835441516938, "adamw_lr": 0.002955835441516938, "weight_decay": 0.01, "grad_norm": 0.16396379470825195, "tokens_per_second": 4353372.95906952, "eta_seconds": 794.4825349260335, "mean_data_wait_seconds": 9.879469871520996e-05}
{"step": 39850, "total_steps": 46238, "tokens": 21546319872, "loss": 2.981675847010179, "ce_loss": 2.981675847010179, "z_loss": 35.62903794375333, "muon_lr": 0.049236396105033026, "residual_lr": 0.0002954183766301981, "adamw_lr": 0.002954183766301981, "weight_decay": 0.01, "grad_norm": 0.16520832479000092, "tokens_per_second": 4353369.601730449, "eta_seconds": 793.2411855467858, "mean_data_wait_seconds": 0.00012040138244628906}
{"step": 39860, "total_steps": 46238, "tokens": 21551726592, "loss": 2.9661545446424773, "ce_loss": 2.9661545446424773, "z_loss": 36.26016405856971, "muon_lr": 0.0492083708210512, "residual_lr": 0.0002952502249263071, "adamw_lr": 0.0029525022492630716, "weight_decay": 0.01, "grad_norm": 0.17169156670570374, "tokens_per_second": 4353381.937043589, "eta_seconds": 791.9969793280919, "mean_data_wait_seconds": 8.991360664367676e-05}
{"step": 39870, "total_steps": 46238, "tokens": 21557133312, "loss": 2.932556522615028, "ce_loss": 2.932556522615028, "z_loss": 36.39637932632909, "muon_lr": 0.049179848748453686, "residual_lr": 0.00029507909249072207, "adamw_lr": 0.0029507909249072207, "weight_decay": 0.01, "grad_norm": 0.173211470246315, "tokens_per_second": 4353428.345405508, "eta_seconds": 790.7465911625901, "mean_data_wait_seconds": 9.161233901977539e-05}
{"step": 39880, "total_steps": 46238, "tokens": 21562540032, "loss": 2.9519210692607993, "ce_loss": 2.9519210692607993, "z_loss": 36.85059397148363, "muon_lr": 0.0491508304725521, "residual_lr": 0.00029490498283531256, "adamw_lr": 0.002949049828353126, "weight_decay": 0.01, "grad_norm": 0.26866060495376587, "tokens_per_second": 4353485.65806524, "eta_seconds": 789.4942521821657, "mean_data_wait_seconds": 8.881092071533203e-05}
{"step": 39890, "total_steps": 46238, "tokens": 21567946752, "loss": 2.9820503050630744, "ce_loss": 2.9820503050630744, "z_loss": 35.73777088974462, "muon_lr": 0.04912131658884085, "residual_lr": 0.00029472789953304506, "adamw_lr": 0.0029472789953304507, "weight_decay": 0.01, "grad_norm": 0.1597980558872223, "tokens_per_second": 4353525.6193271205, "eta_seconds": 788.2450877894211, "mean_data_wait_seconds": 0.0002785623073577881}
{"step": 39900, "total_steps": 46238, "tokens": 21573353472, "loss": 2.9730257735107886, "ce_loss": 2.9730257735107886, "z_loss": 35.100395173737496, "muon_lr": 0.049091307702984865, "residual_lr": 0.00029454784621790915, "adamw_lr": 0.002945478462179092, "weight_decay": 0.01, "grad_norm": 0.16233497858047485, "tokens_per_second": 4353603.301656237, "eta_seconds": 786.9891275340956, "mean_data_wait_seconds": 0.00016105175018310547}
{"step": 39910, "total_steps": 46238, "tokens": 21578760192, "loss": 2.982797861099243, "ce_loss": 2.982797861099243, "z_loss": 36.50420535694469, "muon_lr": 0.04906080443080721, "residual_lr": 0.0002943648265848432, "adamw_lr": 0.0029436482658484323, "weight_decay": 0.01, "grad_norm": 0.15292946994304657, "tokens_per_second": 4353559.844549117, "eta_seconds": 785.7550754201895, "mean_data_wait_seconds": 8.341670036315918e-05}
{"step": 39920, "total_steps": 46238, "tokens": 21584166912, "loss": 2.995039856795109, "ce_loss": 2.995039856795109, "z_loss": 35.63105167042125, "muon_lr": 0.0490298073982764, "residual_lr": 0.0002941788443896584, "adamw_lr": 0.002941788443896584, "weight_decay": 0.01, "grad_norm": 0.16351303458213806, "tokens_per_second": 4353684.067265121, "eta_seconds": 784.4907832610571, "mean_data_wait_seconds": 0.00013902783393859863}
{"step": 39930, "total_steps": 46238, "tokens": 21589573632, "loss": 3.036884748574459, "ce_loss": 3.036884748574459, "z_loss": 36.50545825380267, "muon_lr": 0.048998317241493616, "residual_lr": 0.00029398990344896167, "adamw_lr": 0.002939899034489617, "weight_decay": 0.01, "grad_norm": 0.15358392894268036, "tokens_per_second": 4353801.395368418, "eta_seconds": 783.2278035529099, "mean_data_wait_seconds": 7.402896881103516e-05}
{"step": 39940, "total_steps": 46238, "tokens": 21594980352, "loss": 3.003407218239524, "ce_loss": 3.003407218239524, "z_loss": 36.89515347914262, "muon_lr": 0.048966334606679596, "residual_lr": 0.00029379800764007754, "adamw_lr": 0.0029379800764007757, "weight_decay": 0.01, "grad_norm": 0.13434362411499023, "tokens_per_second": 4353834.449461203, "eta_seconds": 781.980027839903, "mean_data_wait_seconds": 0.00013017654418945312}
{"step": 39950, "total_steps": 46238, "tokens": 21600387072, "loss": 3.018755974191608, "ce_loss": 3.018755974191608, "z_loss": 36.55049228668213, "muon_lr": 0.04893386015016141, "residual_lr": 0.00029360316090096836, "adamw_lr": 0.0029360316090096842, "weight_decay": 0.01, "grad_norm": 0.19284749031066895, "tokens_per_second": 4353869.060518085, "eta_seconds": 780.7319918793136, "mean_data_wait_seconds": 0.00011500716209411621}
{"step": 39960, "total_steps": 46238, "tokens": 21605793792, "loss": 3.001724640528361, "ce_loss": 3.001724640528361, "z_loss": 36.74251374331388, "muon_lr": 0.04890089453835894, "residual_lr": 0.0002934053672301536, "adamw_lr": 0.0029340536723015363, "weight_decay": 0.01, "grad_norm": 0.15424467623233795, "tokens_per_second": 4353874.106991916, "eta_seconds": 779.4892687755662, "mean_data_wait_seconds": 7.659196853637695e-05}
{"step": 39970, "total_steps": 46238, "tokens": 21611200512, "loss": 3.0061748027801514, "ce_loss": 3.0061748027801514, "z_loss": 36.6911346840136, "muon_lr": 0.048867438447771305, "residual_lr": 0.0002932046306866278, "adamw_lr": 0.002932046306866278, "weight_decay": 0.01, "grad_norm": 0.16679981350898743, "tokens_per_second": 4353955.746785749, "eta_seconds": 778.2328579020207, "mean_data_wait_seconds": 0.00011491775512695312}
{"step": 39980, "total_steps": 46238, "tokens": 21616607232, "loss": 3.059351207632007, "ce_loss": 3.059351207632007, "z_loss": 37.37338756792473, "muon_lr": 0.04883349256496285, "residual_lr": 0.0002930009553897771, "adamw_lr": 0.002930009553897771, "weight_decay": 0.01, "grad_norm": 0.17915718257427216, "tokens_per_second": 4353239.100760536, "eta_seconds": 777.1189741011407, "mean_data_wait_seconds": 9.337067604064941e-05}
{"step": 39990, "total_steps": 46238, "tokens": 21622013952, "loss": 2.9818747260353784, "ce_loss": 2.9818747260353784, "z_loss": 35.90605406327681, "muon_lr": 0.04879905758654917, "residual_lr": 0.00029279434551929495, "adamw_lr": 0.00292794345519295, "weight_decay": 0.01, "grad_norm": 0.15624067187309265, "tokens_per_second": 4353237.950778546, "eta_seconds": 775.8771797429415, "mean_data_wait_seconds": 0.00011423230171203613}
{"step": 40000, "total_steps": 46238, "tokens": 21627420672, "loss": 2.971883051323168, "ce_loss": 2.971883051323168, "z_loss": 35.882026209975734, "muon_lr": 0.04876413421918272, "residual_lr": 0.00029258480531509627, "adamw_lr": 0.0029258480531509632, "weight_decay": 0.01, "grad_norm": 0.13555173575878143, "tokens_per_second": 4352951.434965981, "eta_seconds": 774.6861673926196, "mean_data_wait_seconds": 9.41157341003418e-05}
{"step": 40010, "total_steps": 46238, "tokens": 21632827392, "loss": 2.952491343021393, "ce_loss": 2.952491343021393, "z_loss": 37.359422828211926, "muon_lr": 0.048728723179538413, "residual_lr": 0.00029237233907723043, "adamw_lr": 0.0029237233907723047, "weight_decay": 0.01, "grad_norm": 0.1599736511707306, "tokens_per_second": 4352974.5474264845, "eta_seconds": 773.4399793333181, "mean_data_wait_seconds": 0.0001640021800994873}
{"step": 40020, "total_steps": 46238, "tokens": 21638234112, "loss": 3.0144794655568674, "ce_loss": 3.0144794655568674, "z_loss": 36.00011236017401, "muon_lr": 0.0486928251942988, "residual_lr": 0.0002921569511657928, "adamw_lr": 0.002921569511657928, "weight_decay": 0.01, "grad_norm": 0.14258164167404175, "tokens_per_second": 4353009.693933345, "eta_seconds": 772.191669750844, "mean_data_wait_seconds": 9.912252426147461e-05}
{"step": 40030, "total_steps": 46238, "tokens": 21643640832, "loss": 2.959613505638007, "ce_loss": 2.959613505638007, "z_loss": 36.750880443688594, "muon_lr": 0.048656441000139265, "residual_lr": 0.00029193864600083556, "adamw_lr": 0.0029193864600083556, "weight_decay": 0.01, "grad_norm": 0.15670229494571686, "tokens_per_second": 4352544.884504848, "eta_seconds": 771.0319348910696, "mean_data_wait_seconds": 0.00011533498764038086}
{"step": 40040, "total_steps": 46238, "tokens": 21649047552, "loss": 3.016857344092745, "ce_loss": 3.016857344092745, "z_loss": 36.55522855122884, "muon_lr": 0.048619571343712836, "residual_lr": 0.00029171742806227696, "adamw_lr": 0.00291717428062277, "weight_decay": 0.01, "grad_norm": 0.16073541343212128, "tokens_per_second": 4352622.993873384, "eta_seconds": 769.7759233262614, "mean_data_wait_seconds": 8.285045623779297e-05}
{"step": 40050, "total_steps": 46238, "tokens": 21654454272, "loss": 3.006758032423077, "ce_loss": 3.006758032423077, "z_loss": 36.33268523938728, "muon_lr": 0.048582216981634875, "residual_lr": 0.0002914933018898092, "adamw_lr": 0.0029149330188980925, "weight_decay": 0.01, "grad_norm": 0.16458366811275482, "tokens_per_second": 4352613.695771499, "eta_seconds": 768.5353899542597, "mean_data_wait_seconds": 0.00011610984802246094}
{"step": 40060, "total_steps": 46238, "tokens": 21659860992, "loss": 3.0494876485882383, "ce_loss": 3.0494876485882383, "z_loss": 35.8478797565807, "muon_lr": 0.048544378680467584, "residual_lr": 0.00029126627208280544, "adamw_lr": 0.002912662720828055, "weight_decay": 0.01, "grad_norm": 0.16637268662452698, "tokens_per_second": 4352615.676717136, "eta_seconds": 767.2928629708282, "mean_data_wait_seconds": 8.70823860168457e-05}
{"step": 40070, "total_steps": 46238, "tokens": 21665267712, "loss": 3.03263535824689, "ce_loss": 3.03263535824689, "z_loss": 35.8542500409213, "muon_lr": 0.04850605721670425, "residual_lr": 0.00029103634330022543, "adamw_lr": 0.0029103634330022547, "weight_decay": 0.01, "grad_norm": 0.18746764957904816, "tokens_per_second": 4352569.969967567, "eta_seconds": 766.0587301310736, "mean_data_wait_seconds": 0.00011375546455383301}
{"step": 40080, "total_steps": 46238, "tokens": 21670674432, "loss": 3.020233311436393, "ce_loss": 3.020233311436393, "z_loss": 36.788569016890094, "muon_lr": 0.0484672533767533, "residual_lr": 0.00029080352026051974, "adamw_lr": 0.0029080352026051976, "weight_decay": 0.01, "grad_norm": 0.18508881330490112, "tokens_per_second": 4352606.763620397, "eta_seconds": 764.8100746944307, "mean_data_wait_seconds": 8.091330528259277e-05}
{"step": 40090, "total_steps": 46238, "tokens": 21676081152, "loss": 2.964549998442332, "ce_loss": 2.964549998442332, "z_loss": 36.0717435316606, "muon_lr": 0.04842796795692217, "residual_lr": 0.000290567807741533, "adamw_lr": 0.0029056780774153303, "weight_decay": 0.01, "grad_norm": 0.15850262343883514, "tokens_per_second": 4352602.324173648, "eta_seconds": 763.5686737430065, "mean_data_wait_seconds": 0.00010114908218383789}
{"step": 40100, "total_steps": 46238, "tokens": 21681487872, "loss": 2.9756832845283276, "ce_loss": 2.9756832845283276, "z_loss": 36.5481426932595, "muon_lr": 0.04838820176340099, "residual_lr": 0.0002903292105804059, "adamw_lr": 0.002903292105804059, "weight_decay": 0.01, "grad_norm": 0.14577890932559967, "tokens_per_second": 4352689.222459075, "eta_seconds": 762.3112734259073, "mean_data_wait_seconds": 9.515881538391113e-05}
{"step": 40110, "total_steps": 46238, "tokens": 21686894592, "loss": 3.0516363978385925, "ce_loss": 3.0516363978385925, "z_loss": 36.84328865282463, "muon_lr": 0.048347955612246, "residual_lr": 0.00029008773367347594, "adamw_lr": 0.00290087733673476, "weight_decay": 0.01, "grad_norm": 0.13809509575366974, "tokens_per_second": 4352779.581463515, "eta_seconds": 761.0533182307814, "mean_data_wait_seconds": 9.366869926452637e-05}
{"step": 40120, "total_steps": 46238, "tokens": 21692301312, "loss": 2.9937523115764963, "ce_loss": 2.9937523115764963, "z_loss": 36.99712591460257, "muon_lr": 0.04830723032936284, "residual_lr": 0.000289843381976177, "adamw_lr": 0.0028984338197617704, "weight_decay": 0.01, "grad_norm": 0.14985214173793793, "tokens_per_second": 4352840.7988739405, "eta_seconds": 759.800501974614, "mean_data_wait_seconds": 7.396936416625977e-05}
{"step": 40130, "total_steps": 46238, "tokens": 21697708032, "loss": 3.0015392646645056, "ce_loss": 3.0015392646645056, "z_loss": 35.96539907744437, "muon_lr": 0.048266026750489556, "residual_lr": 0.0002895961605029373, "adamw_lr": 0.0028959616050293732, "weight_decay": 0.01, "grad_norm": 0.1780097335577011, "tokens_per_second": 4352797.068298696, "eta_seconds": 758.5660098991364, "mean_data_wait_seconds": 0.00011703372001647949}
{"step": 40140, "total_steps": 46238, "tokens": 21703114752, "loss": 2.916209143219572, "ce_loss": 2.916209143219572, "z_loss": 36.120373928185664, "muon_lr": 0.0482243457211795, "residual_lr": 0.00028934607432707694, "adamw_lr": 0.0028934607432707696, "weight_decay": 0.01, "grad_norm": 0.1561841368675232, "tokens_per_second": 4352906.566618186, "eta_seconds": 757.3048338046602, "mean_data_wait_seconds": 0.00012832880020141602}
{"step": 40150, "total_steps": 46238, "tokens": 21708521472, "loss": 2.957322844953248, "ce_loss": 2.957322844953248, "z_loss": 36.01339135025487, "muon_lr": 0.04818218809678394, "residual_lr": 0.0002890931285807036, "adamw_lr": 0.0028909312858070363, "weight_decay": 0.01, "grad_norm": 0.17084668576717377, "tokens_per_second": 4352952.531996747, "eta_seconds": 756.0547558946962, "mean_data_wait_seconds": 8.311867713928223e-05}
{"step": 40160, "total_steps": 46238, "tokens": 21713928192, "loss": 2.9144693197626057, "ce_loss": 2.9144693197626057, "z_loss": 36.461345094622985, "muon_lr": 0.04813955474243455, "residual_lr": 0.0002888373284546073, "adamw_lr": 0.002888373284546073, "weight_decay": 0.01, "grad_norm": 0.17121854424476624, "tokens_per_second": 4353074.446291289, "eta_seconds": 754.791535164142, "mean_data_wait_seconds": 9.238719940185547e-05}
{"step": 40170, "total_steps": 46238, "tokens": 21719334912, "loss": 3.006292397325689, "ce_loss": 3.006292397325689, "z_loss": 35.371581771157004, "muon_lr": 0.04809644653302558, "residual_lr": 0.00028857867919815345, "adamw_lr": 0.002885786791981535, "weight_decay": 0.01, "grad_norm": 0.13511767983436584, "tokens_per_second": 4353161.5932530705, "eta_seconds": 753.5344033825997, "mean_data_wait_seconds": 5.885958671569824e-05}
{"step": 40180, "total_steps": 46238, "tokens": 21724741632, "loss": 2.9693512230208428, "ce_loss": 2.9693512230208428, "z_loss": 36.68025979128751, "muon_lr": 0.048052864353196, "residual_lr": 0.00028831718611917596, "adamw_lr": 0.00288317186119176, "weight_decay": 0.01, "grad_norm": 0.12915349006652832, "tokens_per_second": 4353244.775204032, "eta_seconds": 752.2780071209094, "mean_data_wait_seconds": 7.048249244689941e-05}
{"step": 40190, "total_steps": 46238, "tokens": 21730148352, "loss": 2.9201733072598777, "ce_loss": 2.9201733072598777, "z_loss": 37.72318666631525, "muon_lr": 0.048008809097311256, "residual_lr": 0.0002880528545838675, "adamw_lr": 0.0028805285458386753, "weight_decay": 0.01, "grad_norm": 0.17038045823574066, "tokens_per_second": 4353330.701062796, "eta_seconds": 751.0211855032786, "mean_data_wait_seconds": 5.951523780822754e-05}
{"step": 40200, "total_steps": 46238, "tokens": 21735555072, "loss": 2.9215636560411165, "ce_loss": 2.9215636560411165, "z_loss": 37.46114878220992, "muon_lr": 0.04796428166944499, "residual_lr": 0.00028778569001666987, "adamw_lr": 0.0028778569001666992, "weight_decay": 0.01, "grad_norm": 0.19360597431659698, "tokens_per_second": 4353386.221962904, "eta_seconds": 749.7696500101189, "mean_data_wait_seconds": 7.963180541992188e-05}
{"step": 40210, "total_steps": 46238, "tokens": 21740961792, "loss": 2.978170450889703, "ce_loss": 2.978170450889703, "z_loss": 35.56074156905665, "muon_lr": 0.04791928298336043, "residual_lr": 0.00028751569790016253, "adamw_lr": 0.0028751569790016252, "weight_decay": 0.01, "grad_norm": 0.1672568917274475, "tokens_per_second": 4353472.579294439, "eta_seconds": 748.512844550435, "mean_data_wait_seconds": 6.398558616638184e-05}
{"step": 40220, "total_steps": 46238, "tokens": 21746368512, "loss": 2.909943958123525, "ce_loss": 2.909943958123525, "z_loss": 37.52635583010587, "muon_lr": 0.04787381396249166, "residual_lr": 0.00028724288377494987, "adamw_lr": 0.0028724288377494993, "weight_decay": 0.01, "grad_norm": 0.15926827490329742, "tokens_per_second": 4353504.519647331, "eta_seconds": 747.2654293381869, "mean_data_wait_seconds": 7.078051567077637e-05}
{"step": 40230, "total_steps": 46238, "tokens": 21751775232, "loss": 2.9463423924012617, "ce_loss": 2.9463423924012617, "z_loss": 36.372679739287406, "muon_lr": 0.04782787553992468, "residual_lr": 0.00028696725323954804, "adamw_lr": 0.0028696725323954808, "weight_decay": 0.01, "grad_norm": 0.14118875563144684, "tokens_per_second": 4353502.394061186, "eta_seconds": 746.0238699835095, "mean_data_wait_seconds": 9.480118751525879e-05}
{"step": 40240, "total_steps": 46238, "tokens": 21757181952, "loss": 2.838613159728773, "ce_loss": 2.838613159728773, "z_loss": 37.357923767783426, "muon_lr": 0.04778146865837828, "residual_lr": 0.00028668881195026966, "adamw_lr": 0.002866888119502697, "weight_decay": 0.01, "grad_norm": 0.16279472410678864, "tokens_per_second": 4353477.502310976, "eta_seconds": 744.7862041962584, "mean_data_wait_seconds": 5.97536563873291e-05}
{"step": 40250, "total_steps": 46238, "tokens": 21762588672, "loss": 2.9877165375333843, "ce_loss": 2.9877165375333843, "z_loss": 35.67393938700358, "muon_lr": 0.04773459427018461, "residual_lr": 0.0002864075656211076, "adamw_lr": 0.002864075656211076, "weight_decay": 0.01, "grad_norm": 0.1495196521282196, "tokens_per_second": 4353483.7638762165, "eta_seconds": 743.5432034591685, "mean_data_wait_seconds": 6.446242332458496e-05}
{"step": 40260, "total_steps": 46238, "tokens": 21767995392, "loss": 2.920729100704193, "ce_loss": 2.920729100704193, "z_loss": 36.234171751773715, "muon_lr": 0.047687253337269725, "residual_lr": 0.0002861235200236183, "adamw_lr": 0.0028612352002361833, "weight_decay": 0.01, "grad_norm": 0.18526539206504822, "tokens_per_second": 4353487.186942097, "eta_seconds": 742.3006902817793, "mean_data_wait_seconds": 0.00011083483695983887}
{"step": 40270, "total_steps": 46238, "tokens": 21773402112, "loss": 2.9339411854743958, "ce_loss": 2.9339411854743958, "z_loss": 36.06728397716176, "muon_lr": 0.047639446831133796, "residual_lr": 0.00028583668098680273, "adamw_lr": 0.0028583668098680274, "weight_decay": 0.01, "grad_norm": 0.13288040459156036, "tokens_per_second": 4353456.206107362, "eta_seconds": 741.0640353919383, "mean_data_wait_seconds": 0.00010266900062561035}
{"step": 40280, "total_steps": 46238, "tokens": 21778808832, "loss": 2.9695820952906753, "ce_loss": 2.9695820952906753, "z_loss": 37.49104155916156, "muon_lr": 0.0475911757328312, "residual_lr": 0.0002855470543969871, "adamw_lr": 0.0028554705439698715, "weight_decay": 0.01, "grad_norm": 0.2146417200565338, "tokens_per_second": 4353434.930460259, "eta_seconds": 739.8257135910582, "mean_data_wait_seconds": 0.0001246035099029541}
{"step": 40290, "total_steps": 46238, "tokens": 21784215552, "loss": 2.93532632336472, "ce_loss": 2.93532632336472, "z_loss": 35.863453720555164, "muon_lr": 0.047542441032950356, "residual_lr": 0.0002852546461977021, "adamw_lr": 0.0028525464619770214, "weight_decay": 0.01, "grad_norm": 0.1448214054107666, "tokens_per_second": 4353361.476992927, "eta_seconds": 738.5962321284226, "mean_data_wait_seconds": 0.0001226365566253662}
{"step": 40300, "total_steps": 46238, "tokens": 21789622272, "loss": 3.0008875547033367, "ce_loss": 3.0008875547033367, "z_loss": 35.67596996191776, "muon_lr": 0.04749324373159345, "residual_lr": 0.00028495946238956066, "adamw_lr": 0.002849594623895607, "weight_decay": 0.01, "grad_norm": 0.15602466464042664, "tokens_per_second": 4353323.327851595, "eta_seconds": 737.3607293222002, "mean_data_wait_seconds": 0.00011479854583740234}
{"step": 40310, "total_steps": 46238, "tokens": 21795028992, "loss": 3.029029871478225, "ce_loss": 3.029029871478225, "z_loss": 36.43362045288086, "muon_lr": 0.04744358483835585, "residual_lr": 0.00028466150903013506, "adamw_lr": 0.002846615090301351, "weight_decay": 0.01, "grad_norm": 0.14647948741912842, "tokens_per_second": 4353454.485044442, "eta_seconds": 736.0965768698708, "mean_data_wait_seconds": 7.861852645874023e-05}
{"step": 40320, "total_steps": 46238, "tokens": 21800435712, "loss": 2.9602528575694924, "ce_loss": 2.9602528575694924, "z_loss": 35.46148360859264, "muon_lr": 0.04739346537230541, "residual_lr": 0.00028436079223383246, "adamw_lr": 0.002843607922338325, "weight_decay": 0.01, "grad_norm": 0.1696060448884964, "tokens_per_second": 4353488.838755712, "eta_seconds": 734.8488402038406, "mean_data_wait_seconds": 6.473064422607422e-05}
{"step": 40330, "total_steps": 46238, "tokens": 21805842432, "loss": 2.984920308445439, "ce_loss": 2.984920308445439, "z_loss": 36.490937955451734, "muon_lr": 0.0473428863619616, "residual_lr": 0.0002840573181717696, "adamw_lr": 0.002840573181717696, "weight_decay": 0.01, "grad_norm": 0.12104889750480652, "tokens_per_second": 4353551.152172852, "eta_seconds": 733.5964118409412, "mean_data_wait_seconds": 7.11977481842041e-05}
{"step": 40340, "total_steps": 46238, "tokens": 21811249152, "loss": 2.961737430456913, "ce_loss": 2.961737430456913, "z_loss": 37.42780708544182, "muon_lr": 0.04729184884527432, "residual_lr": 0.0002837510930716459, "adamw_lr": 0.0028375109307164593, "weight_decay": 0.01, "grad_norm": 0.1477394551038742, "tokens_per_second": 4353597.26509209, "eta_seconds": 732.3467445104981, "mean_data_wait_seconds": 5.6803226470947266e-05}
{"step": 40350, "total_steps": 46238, "tokens": 21816655872, "loss": 3.0078758200009665, "ce_loss": 3.0078758200009665, "z_loss": 37.18659210205078, "muon_lr": 0.04724035386960268, "residual_lr": 0.00028344212321761603, "adamw_lr": 0.0028344212321761604, "weight_decay": 0.01, "grad_norm": 0.15040062367916107, "tokens_per_second": 4353708.044931867, "eta_seconds": 731.0862444497725, "mean_data_wait_seconds": 6.264448165893555e-05}
{"step": 40360, "total_steps": 46238, "tokens": 21822062592, "loss": 2.908779471209555, "ce_loss": 2.908779471209555, "z_loss": 37.16101461468321, "muon_lr": 0.047188402491693426, "residual_lr": 0.00028313041495016053, "adamw_lr": 0.0028313041495016057, "weight_decay": 0.01, "grad_norm": 0.1452818214893341, "tokens_per_second": 4353784.187003438, "eta_seconds": 729.8316148708752, "mean_data_wait_seconds": 8.040666580200195e-05}
{"step": 40370, "total_steps": 46238, "tokens": 21827469312, "loss": 2.953498894518072, "ce_loss": 2.953498894518072, "z_loss": 36.58645349560362, "muon_lr": 0.04713599577765934, "residual_lr": 0.00028281597466595597, "adamw_lr": 0.00282815974665956, "weight_decay": 0.01, "grad_norm": 0.1766158789396286, "tokens_per_second": 4353815.178858231, "eta_seconds": 728.5845847117184, "mean_data_wait_seconds": 5.939602851867676e-05}
{"step": 40380, "total_steps": 46238, "tokens": 21832876032, "loss": 2.9336294080271865, "ce_loss": 2.9336294080271865, "z_loss": 34.800921873612836, "muon_lr": 0.04708313480295729, "residual_lr": 0.0002824988088177437, "adamw_lr": 0.002824988088177437, "weight_decay": 0.01, "grad_norm": 0.17483913898468018, "tokens_per_second": 4353939.493307923, "eta_seconds": 727.3219825097007, "mean_data_wait_seconds": 0.00011947751045227051}
{"step": 40390, "total_steps": 46238, "tokens": 21838282752, "loss": 2.9582460450403616, "ce_loss": 2.9582460450403616, "z_loss": 36.561915310946375, "muon_lr": 0.04702982065236621, "residual_lr": 0.0002821789239141972, "adamw_lr": 0.002821789239141972, "weight_decay": 0.01, "grad_norm": 0.14454206824302673, "tokens_per_second": 4354035.05953568, "eta_seconds": 726.0642463308795, "mean_data_wait_seconds": 8.082389831542969e-05}
{"step": 40400, "total_steps": 46238, "tokens": 21843689472, "loss": 2.997111911123449, "ce_loss": 2.997111911123449, "z_loss": 36.93299241499467, "muon_lr": 0.0469760544199648, "residual_lr": 0.0002818563265197888, "adamw_lr": 0.002818563265197888, "weight_decay": 0.01, "grad_norm": 0.13229435682296753, "tokens_per_second": 4354117.996146152, "eta_seconds": 724.8086677470162, "mean_data_wait_seconds": 0.0001150965690612793}
{"step": 40410, "total_steps": 46238, "tokens": 21849096192, "loss": 2.9722525748339566, "ce_loss": 2.9722525748339566, "z_loss": 35.03070345791903, "muon_lr": 0.0469218372091091, "residual_lr": 0.00028153102325465456, "adamw_lr": 0.002815310232546546, "weight_decay": 0.01, "grad_norm": 0.15145224332809448, "tokens_per_second": 4354229.163048165, "eta_seconds": 723.5484458963352, "mean_data_wait_seconds": 6.61313533782959e-05}
{"step": 40420, "total_steps": 46238, "tokens": 21854502912, "loss": 2.898106892903646, "ce_loss": 2.898106892903646, "z_loss": 37.00477793722442, "muon_lr": 0.04686717013240985, "residual_lr": 0.00028120302079445903, "adamw_lr": 0.0028120302079445905, "weight_decay": 0.01, "grad_norm": 0.14969824254512787, "tokens_per_second": 4354289.211526808, "eta_seconds": 722.2967679028355, "mean_data_wait_seconds": 5.9604644775390625e-05}
{"step": 40430, "total_steps": 46238, "tokens": 21859909632, "loss": 3.022947519114523, "ce_loss": 3.022947519114523, "z_loss": 36.359147331931375, "muon_lr": 0.04681205431170962, "residual_lr": 0.0002808723258702577, "adamw_lr": 0.002808723258702577, "weight_decay": 0.01, "grad_norm": 0.18607103824615479, "tokens_per_second": 4354408.7909337, "eta_seconds": 721.0352667248702, "mean_data_wait_seconds": 6.35981559753418e-05}
{"step": 40440, "total_steps": 46238, "tokens": 21865316352, "loss": 2.9500484358180654, "ce_loss": 2.9500484358180654, "z_loss": 36.15171233090487, "muon_lr": 0.046756490878059845, "residual_lr": 0.000280538945268359, "adamw_lr": 0.0028053894526835904, "weight_decay": 0.01, "grad_norm": 0.1565168797969818, "tokens_per_second": 4354504.500382438, "eta_seconds": 719.7777803937806, "mean_data_wait_seconds": 8.425116539001465e-05}
{"step": 40450, "total_steps": 46238, "tokens": 21870723072, "loss": 2.97244533625516, "ce_loss": 2.97244533625516, "z_loss": 36.442281954216234, "muon_lr": 0.04670048097169758, "residual_lr": 0.0002802028858301854, "adamw_lr": 0.0028020288583018547, "weight_decay": 0.01, "grad_norm": 0.14376531541347504, "tokens_per_second": 4354522.684629561, "eta_seconds": 718.5331414265379, "mean_data_wait_seconds": 8.115172386169434e-05}
{"step": 40460, "total_steps": 46238, "tokens": 21876129792, "loss": 2.965916447567217, "ce_loss": 2.965916447567217, "z_loss": 37.56561548059637, "muon_lr": 0.04664402574202208, "residual_lr": 0.00027986415445213246, "adamw_lr": 0.002798641544521325, "weight_decay": 0.01, "grad_norm": 0.18164241313934326, "tokens_per_second": 4354595.05476645, "eta_seconds": 717.2795873593625, "mean_data_wait_seconds": 0.00020870566368103027}
{"step": 40470, "total_steps": 46238, "tokens": 21881536512, "loss": 2.9569734316883665, "ce_loss": 2.9569734316883665, "z_loss": 36.57957496065082, "muon_lr": 0.04658712634757127, "residual_lr": 0.0002795227580854276, "adamw_lr": 0.002795227580854276, "weight_decay": 0.01, "grad_norm": 0.13147442042827606, "tokens_per_second": 4354691.736262033, "eta_seconds": 716.0220775297557, "mean_data_wait_seconds": 9.545683860778809e-05}
{"step": 40480, "total_steps": 46238, "tokens": 21886943232, "loss": 2.968142724398411, "ce_loss": 2.968142724398411, "z_loss": 36.75976490251946, "muon_lr": 0.046529783955997925, "residual_lr": 0.0002791787037359875, "adamw_lr": 0.002791787037359875, "weight_decay": 0.01, "grad_norm": 0.1642725020647049, "tokens_per_second": 4354744.201231314, "eta_seconds": 714.7718810027674, "mean_data_wait_seconds": 0.0001029670238494873}
{"step": 40490, "total_steps": 46238, "tokens": 21892349952, "loss": 2.931816886771809, "ce_loss": 2.931816886771809, "z_loss": 36.64886512178363, "muon_lr": 0.04647199974404573, "residual_lr": 0.0002788319984642743, "adamw_lr": 0.0027883199846427432, "weight_decay": 0.01, "grad_norm": 0.1352575719356537, "tokens_per_second": 4354829.259720854, "eta_seconds": 713.5163742789712, "mean_data_wait_seconds": 9.447336196899414e-05}
{"step": 40500, "total_steps": 46238, "tokens": 21897756672, "loss": 2.966761370499929, "ce_loss": 2.966761370499929, "z_loss": 36.10630110538367, "muon_lr": 0.046413774897525094, "residual_lr": 0.0002784826493851505, "adamw_lr": 0.0027848264938515055, "weight_decay": 0.01, "grad_norm": 0.14702104032039642, "tokens_per_second": 4354796.583021228, "eta_seconds": 712.2801731069695, "mean_data_wait_seconds": 0.00014474987983703613}
{"step": 40510, "total_steps": 46238, "tokens": 21903163392, "loss": 2.9678146694645737, "ce_loss": 2.9678146694645737, "z_loss": 36.049371256972805, "muon_lr": 0.046355110611288886, "residual_lr": 0.0002781306636677333, "adamw_lr": 0.002781306636677333, "weight_decay": 0.01, "grad_norm": 0.16982237994670868, "tokens_per_second": 4354798.706881712, "eta_seconds": 711.0382712079985, "mean_data_wait_seconds": 0.00013139843940734863}
{"step": 40520, "total_steps": 46238, "tokens": 21908570112, "loss": 2.9353145668000886, "ce_loss": 2.9353145668000886, "z_loss": 36.67009451895049, "muon_lr": 0.04629600808920786, "residual_lr": 0.00027777604853524714, "adamw_lr": 0.002777760485352472, "weight_decay": 0.01, "grad_norm": 0.16103097796440125, "tokens_per_second": 4354792.221288614, "eta_seconds": 709.7977737926024, "mean_data_wait_seconds": 8.875131607055664e-05}
{"step": 40530, "total_steps": 46238, "tokens": 21913976832, "loss": 3.036839479749853, "ce_loss": 3.036839479749853, "z_loss": 36.994631825071394, "muon_lr": 0.04623646854414596, "residual_lr": 0.00027741881126487573, "adamw_lr": 0.0027741881126487574, "weight_decay": 0.01, "grad_norm": 0.1517072468996048, "tokens_per_second": 4354677.657471282, "eta_seconds": 708.5748582805061, "mean_data_wait_seconds": 0.002887636423110962}
{"step": 40540, "total_steps": 46238, "tokens": 21919383552, "loss": 3.0087032932223696, "ce_loss": 3.0087032932223696, "z_loss": 36.55327406796542, "muon_lr": 0.04617649319793543, "residual_lr": 0.00027705895918761253, "adamw_lr": 0.0027705895918761258, "weight_decay": 0.01, "grad_norm": 0.1699703484773636, "tokens_per_second": 4354627.050904174, "eta_seconds": 707.3414894073283, "mean_data_wait_seconds": 9.039044380187988e-05}
{"step": 40550, "total_steps": 46238, "tokens": 21924790272, "loss": 2.9571722745895386, "ce_loss": 2.9571722745895386, "z_loss": 36.75633202177106, "muon_lr": 0.04611608328135174, "residual_lr": 0.0002766964996881104, "adamw_lr": 0.0027669649968811044, "weight_decay": 0.01, "grad_norm": 0.15928930044174194, "tokens_per_second": 4354642.278289785, "eta_seconds": 706.0974168485726, "mean_data_wait_seconds": 0.00030809640884399414}
{"step": 40560, "total_steps": 46238, "tokens": 21930196992, "loss": 2.978096853603016, "ce_loss": 2.978096853603016, "z_loss": 36.560633081378356, "muon_lr": 0.04605524003408837, "residual_lr": 0.00027633144020453016, "adamw_lr": 0.002763314402045302, "weight_decay": 0.01, "grad_norm": 0.17466920614242554, "tokens_per_second": 4354573.984235583, "eta_seconds": 704.866872192737, "mean_data_wait_seconds": 0.00010147690773010254}
{"step": 40570, "total_steps": 46238, "tokens": 21935603712, "loss": 2.992260678247972, "ce_loss": 2.992260678247972, "z_loss": 37.00807166822029, "muon_lr": 0.04599396470473127, "residual_lr": 0.0002759637882283876, "adamw_lr": 0.0027596378822838763, "weight_decay": 0.01, "grad_norm": 0.14585936069488525, "tokens_per_second": 4354572.840743461, "eta_seconds": 703.6254383740845, "mean_data_wait_seconds": 8.693337440490723e-05}
{"step": 40580, "total_steps": 46238, "tokens": 21941010432, "loss": 2.922983841462569, "ce_loss": 2.922983841462569, "z_loss": 36.68226935646751, "muon_lr": 0.04593225855073338, "residual_lr": 0.00027559355130440027, "adamw_lr": 0.002755935513044003, "weight_decay": 0.01, "grad_norm": 0.1662355661392212, "tokens_per_second": 4354640.312173724, "eta_seconds": 702.3729366233775, "mean_data_wait_seconds": 9.34898853302002e-05}
{"step": 40590, "total_steps": 46238, "tokens": 21946417152, "loss": 2.971403136397853, "ce_loss": 2.971403136397853, "z_loss": 36.44051083651456, "muon_lr": 0.04587012283838868, "residual_lr": 0.000275220737030332, "adamw_lr": 0.0027522073703033206, "weight_decay": 0.01, "grad_norm": 0.1483970284461975, "tokens_per_second": 4354682.510859959, "eta_seconds": 701.1245426930244, "mean_data_wait_seconds": 0.0001189112663269043}
{"step": 40600, "total_steps": 46238, "tokens": 21951823872, "loss": 2.9665687300942163, "ce_loss": 2.9665687300942163, "z_loss": 37.18691109166001, "muon_lr": 0.045807558842806295, "residual_lr": 0.0002748453530568377, "adamw_lr": 0.0027484535305683775, "weight_decay": 0.01, "grad_norm": 0.12880273163318634, "tokens_per_second": 4354729.861333559, "eta_seconds": 699.8753449810259, "mean_data_wait_seconds": 0.00012475252151489258}
{"step": 40610, "total_steps": 46238, "tokens": 21957230592, "loss": 3.0258824879472908, "ce_loss": 3.0258824879472908, "z_loss": 36.45061140349417, "muon_lr": 0.045744567847884304, "residual_lr": 0.0002744674070873058, "adamw_lr": 0.002744674070873058, "weight_decay": 0.01, "grad_norm": 0.13656744360923767, "tokens_per_second": 4354799.272486402, "eta_seconds": 698.6226353121767, "mean_data_wait_seconds": 0.0001424252986907959}
{"step": 40620, "total_steps": 46238, "tokens": 21962637312, "loss": 2.9860842282121833, "ce_loss": 2.9860842282121833, "z_loss": 36.216252413663, "muon_lr": 0.0456811511462834, "residual_lr": 0.00027408690687770037, "adamw_lr": 0.0027408690687770036, "weight_decay": 0.01, "grad_norm": 0.1748780459165573, "tokens_per_second": 4354833.826381021, "eta_seconds": 697.3755475128628, "mean_data_wait_seconds": 7.69495964050293e-05}
{"step": 40630, "total_steps": 46238, "tokens": 21968044032, "loss": 2.9786651315111103, "ce_loss": 2.9786651315111103, "z_loss": 37.41917948289351, "muon_lr": 0.04561731003940033, "residual_lr": 0.00027370386023640195, "adamw_lr": 0.0027370386023640196, "weight_decay": 0.01, "grad_norm": 0.1373664140701294, "tokens_per_second": 4354875.852781575, "eta_seconds": 696.1272850209197, "mean_data_wait_seconds": 7.88569450378418e-05}
{"step": 40640, "total_steps": 46238, "tokens": 21973450752, "loss": 2.9811002774672075, "ce_loss": 2.9811002774672075, "z_loss": 36.4512438918605, "muon_lr": 0.04555304583734124, "residual_lr": 0.00027331827502404736, "adamw_lr": 0.002733182750240474, "weight_decay": 0.01, "grad_norm": 0.14250220358371735, "tokens_per_second": 4354940.372510037, "eta_seconds": 694.8754575612792, "mean_data_wait_seconds": 7.987022399902344e-05}
{"step": 40650, "total_steps": 46238, "tokens": 21978857472, "loss": 2.9834030920808967, "ce_loss": 2.9834030920808967, "z_loss": 36.476976625847094, "muon_lr": 0.04548835985889474, "residual_lr": 0.0002729301591533684, "adamw_lr": 0.002729301591533684, "weight_decay": 0.01, "grad_norm": 0.1458350419998169, "tokens_per_second": 4354931.751103273, "eta_seconds": 693.6353166119609, "mean_data_wait_seconds": 0.00011214613914489746}
{"step": 40660, "total_steps": 46238, "tokens": 21984264192, "loss": 2.9493393482583943, "ce_loss": 2.9493393482583943, "z_loss": 37.66816847252123, "muon_lr": 0.045423253431504884, "residual_lr": 0.00027253952058902927, "adamw_lr": 0.002725395205890293, "weight_decay": 0.01, "grad_norm": 0.18096128106117249, "tokens_per_second": 4354889.89697135, "eta_seconds": 692.4004545090885, "mean_data_wait_seconds": 7.37905502319336e-05}
{"step": 40670, "total_steps": 46238, "tokens": 21989670912, "loss": 2.9360352248856514, "ce_loss": 2.9360352248856514, "z_loss": 37.11179441394228, "muon_lr": 0.045357727891243906, "residual_lr": 0.0002721463673474634, "adamw_lr": 0.0027214636734746345, "weight_decay": 0.01, "grad_norm": 0.16875410079956055, "tokens_per_second": 4354908.183482992, "eta_seconds": 691.1560237746985, "mean_data_wait_seconds": 7.173418998718262e-05}
{"step": 40680, "total_steps": 46238, "tokens": 21995077632, "loss": 3.0104390635634912, "ce_loss": 3.0104390635634912, "z_loss": 36.3014023809722, "muon_lr": 0.04529178458278478, "residual_lr": 0.00027175070749670864, "adamw_lr": 0.002717507074967087, "weight_decay": 0.01, "grad_norm": 0.16695068776607513, "tokens_per_second": 4354882.155888918, "eta_seconds": 689.9186238454527, "mean_data_wait_seconds": 7.325410842895508e-05}
{"step": 40690, "total_steps": 46238, "tokens": 22000484352, "loss": 2.994349013675343, "ce_loss": 2.994349013675343, "z_loss": 37.09300483356822, "muon_lr": 0.04522542485937369, "residual_lr": 0.0002713525491562421, "adamw_lr": 0.002713525491562421, "weight_decay": 0.01, "grad_norm": 0.14473769068717957, "tokens_per_second": 4354896.338321716, "eta_seconds": 688.674850330833, "mean_data_wait_seconds": 6.335973739624023e-05}
{"step": 40700, "total_steps": 46238, "tokens": 22005891072, "loss": 2.989015201727549, "ce_loss": 2.989015201727549, "z_loss": 37.38578510284424, "muon_lr": 0.045158650082802176, "residual_lr": 0.00027095190049681304, "adamw_lr": 0.0027095190049681306, "weight_decay": 0.01, "grad_norm": 0.1573217809200287, "tokens_per_second": 4354977.047588445, "eta_seconds": 687.4205836877493, "mean_data_wait_seconds": 0.0001226365566253662}
{"step": 40710, "total_steps": 46238, "tokens": 22011297792, "loss": 2.9712820847829184, "ce_loss": 2.9712820847829184, "z_loss": 36.89698730815541, "muon_lr": 0.04509146162337928, "residual_lr": 0.00027054876974027565, "adamw_lr": 0.0027054876974027565, "weight_decay": 0.01, "grad_norm": 0.18394409120082855, "tokens_per_second": 4354944.475256787, "eta_seconds": 686.1842122163445, "mean_data_wait_seconds": 9.205937385559082e-05}
{"step": 40720, "total_steps": 46238, "tokens": 22016704512, "loss": 2.955257175546704, "ce_loss": 2.955257175546704, "z_loss": 37.91738758665143, "muon_lr": 0.045023860859903335, "residual_lr": 0.00027014316515941997, "adamw_lr": 0.0027014316515942, "weight_decay": 0.01, "grad_norm": 0.1165325939655304, "tokens_per_second": 4355011.704132352, "eta_seconds": 684.932125709242, "mean_data_wait_seconds": 7.459521293640137e-05}
{"step": 40730, "total_steps": 46238, "tokens": 22022111232, "loss": 3.0729022333116243, "ce_loss": 3.0729022333116243, "z_loss": 36.86265821167917, "muon_lr": 0.044955849179633724, "residual_lr": 0.0002697350950778023, "adamw_lr": 0.0026973509507780234, "weight_decay": 0.01, "grad_norm": 0.1707708090543747, "tokens_per_second": 4355057.69118579, "eta_seconds": 683.6834125127962, "mean_data_wait_seconds": 7.271766662597656e-05}
{"step": 40740, "total_steps": 46238, "tokens": 22027517952, "loss": 2.9592201709747314, "ce_loss": 2.9592201709747314, "z_loss": 37.12967557618112, "muon_lr": 0.04488742797826242, "residual_lr": 0.00026932456786957454, "adamw_lr": 0.0026932456786957454, "weight_decay": 0.01, "grad_norm": 0.1703481376171112, "tokens_per_second": 4355214.146643066, "eta_seconds": 682.4174159819054, "mean_data_wait_seconds": 7.62939453125e-05}
{"step": 40750, "total_steps": 46238, "tokens": 22032924672, "loss": 3.0071111917495728, "ce_loss": 3.0071111917495728, "z_loss": 37.329076102285676, "muon_lr": 0.044818598659885314, "residual_lr": 0.00026891159195931185, "adamw_lr": 0.002689115919593119, "weight_decay": 0.01, "grad_norm": 0.17325438559055328, "tokens_per_second": 4355250.034053747, "eta_seconds": 681.1703669832035, "mean_data_wait_seconds": 7.769465446472168e-05}
{"step": 40760, "total_steps": 46238, "tokens": 22038331392, "loss": 2.8845651655486138, "ce_loss": 2.8845651655486138, "z_loss": 36.99918801856764, "muon_lr": 0.0447493626369734, "residual_lr": 0.00026849617582184034, "adamw_lr": 0.002684961758218404, "weight_decay": 0.01, "grad_norm": 0.13124100863933563, "tokens_per_second": 4355244.930499123, "eta_seconds": 679.9297378805815, "mean_data_wait_seconds": 8.130073547363281e-05}
{"step": 40770, "total_steps": 46238, "tokens": 22043738112, "loss": 2.939956571116592, "ce_loss": 2.939956571116592, "z_loss": 37.04611376560096, "muon_lr": 0.044679721330343806, "residual_lr": 0.0002680783279820628, "adamw_lr": 0.002680783279820628, "weight_decay": 0.01, "grad_norm": 0.1683848351240158, "tokens_per_second": 4355350.345395879, "eta_seconds": 678.6718839104844, "mean_data_wait_seconds": 6.321072578430176e-05}
{"step": 40780, "total_steps": 46238, "tokens": 22049144832, "loss": 2.9432414517258154, "ce_loss": 2.9432414517258154, "z_loss": 36.23588238340436, "muon_lr": 0.044609676169130634, "residual_lr": 0.0002676580570147838, "adamw_lr": 0.002676580570147838, "weight_decay": 0.01, "grad_norm": 0.12920673191547394, "tokens_per_second": 4355453.788815256, "eta_seconds": 677.414397456519, "mean_data_wait_seconds": 7.545948028564453e-05}
{"step": 40790, "total_steps": 46238, "tokens": 22054551552, "loss": 2.9878160231041186, "ce_loss": 2.9878160231041186, "z_loss": 36.70492007515647, "muon_lr": 0.04453922859075563, "residual_lr": 0.00026723537154453374, "adamw_lr": 0.002672353715445338, "weight_decay": 0.01, "grad_norm": 0.15215247869491577, "tokens_per_second": 4355468.530922784, "eta_seconds": 676.1707410100471, "mean_data_wait_seconds": 6.687641143798828e-05}
{"step": 40800, "total_steps": 46238, "tokens": 22059958272, "loss": 2.952592965328332, "ce_loss": 2.952592965328332, "z_loss": 37.60267514893503, "muon_lr": 0.0444683800408987, "residual_lr": 0.0002668102802453922, "adamw_lr": 0.002668102802453922, "weight_decay": 0.01, "grad_norm": 0.16280238330364227, "tokens_per_second": 4355532.716386411, "eta_seconds": 674.9194313110065, "mean_data_wait_seconds": 0.000132828950881958}
{"step": 40810, "total_steps": 46238, "tokens": 22065364992, "loss": 2.9728335575623945, "ce_loss": 2.9728335575623945, "z_loss": 36.497085108901516, "muon_lr": 0.04439713197346822, "residual_lr": 0.0002663827918408093, "adamw_lr": 0.002663827918408093, "weight_decay": 0.01, "grad_norm": 0.1517658531665802, "tokens_per_second": 4355570.141036998, "eta_seconds": 673.67229753793, "mean_data_wait_seconds": 0.0001106560230255127}
{"step": 40820, "total_steps": 46238, "tokens": 22070771712, "loss": 2.9800171508933557, "ce_loss": 2.9800171508933557, "z_loss": 36.90730944546786, "muon_lr": 0.04432548585057118, "residual_lr": 0.000265952915103427, "adamw_lr": 0.002659529151034271, "weight_decay": 0.01, "grad_norm": 0.17200405895709991, "tokens_per_second": 4355704.074775754, "eta_seconds": 672.4102863096331, "mean_data_wait_seconds": 9.429454803466797e-05}
{"step": 40830, "total_steps": 46238, "tokens": 22076178432, "loss": 2.9762485045375247, "ce_loss": 2.9762485045375247, "z_loss": 38.34121013410164, "muon_lr": 0.044253443142483244, "residual_lr": 0.0002655206588548994, "adamw_lr": 0.0026552065885489945, "weight_decay": 0.01, "grad_norm": 0.16141360998153687, "tokens_per_second": 4355667.098305174, "eta_seconds": 671.1746876012458, "mean_data_wait_seconds": 6.884336471557617e-05}
{"step": 40840, "total_steps": 46238, "tokens": 22081585152, "loss": 2.981475886070367, "ce_loss": 2.981475886070367, "z_loss": 36.16216619086988, "muon_lr": 0.04418100532761854, "residual_lr": 0.00026508603196571115, "adamw_lr": 0.002650860319657112, "weight_decay": 0.01, "grad_norm": 0.12484649568796158, "tokens_per_second": 4355735.461301367, "eta_seconds": 669.9228660521511, "mean_data_wait_seconds": 6.374716758728027e-05}
{"step": 40850, "total_steps": 46238, "tokens": 22086991872, "loss": 3.037020602009513, "ce_loss": 3.037020602009513, "z_loss": 37.137479319716945, "muon_lr": 0.0441081738924993, "residual_lr": 0.00026464904335499575, "adamw_lr": 0.0026464904335499577, "weight_decay": 0.01, "grad_norm": 0.13343699276447296, "tokens_per_second": 4355809.526900194, "eta_seconds": 668.6702083763402, "mean_data_wait_seconds": 7.262825965881348e-05}
{"step": 40860, "total_steps": 46238, "tokens": 22092398592, "loss": 2.969195945696397, "ce_loss": 2.969195945696397, "z_loss": 37.22872141635779, "muon_lr": 0.0440349503317254, "residual_lr": 0.00026420970199035235, "adamw_lr": 0.002642097019903524, "weight_decay": 0.01, "grad_norm": 0.16142572462558746, "tokens_per_second": 4355868.950464246, "eta_seconds": 667.4198367905796, "mean_data_wait_seconds": 6.589293479919434e-05}
{"step": 40870, "total_steps": 46238, "tokens": 22097805312, "loss": 2.961833271113309, "ce_loss": 2.961833271113309, "z_loss": 36.36772040164832, "muon_lr": 0.04396133614794367, "residual_lr": 0.00026376801688766195, "adamw_lr": 0.00263768016887662, "weight_decay": 0.01, "grad_norm": 0.1407613903284073, "tokens_per_second": 4355916.47674219, "eta_seconds": 666.171318824336, "mean_data_wait_seconds": 5.817413330078125e-05}
{"step": 40880, "total_steps": 46238, "tokens": 22103212032, "loss": 3.017343947381684, "ce_loss": 3.017343947381684, "z_loss": 37.32179552136046, "muon_lr": 0.04388733285181701, "residual_lr": 0.000263323997110902, "adamw_lr": 0.0026332399711090203, "weight_decay": 0.01, "grad_norm": 0.17277376353740692, "tokens_per_second": 4355927.518613745, "eta_seconds": 664.9283973673097, "mean_data_wait_seconds": 7.799267768859863e-05}
{"step": 40890, "total_steps": 46238, "tokens": 22108618752, "loss": 2.9651995608300874, "ce_loss": 2.9651995608300874, "z_loss": 36.91326583515514, "muon_lr": 0.04381294196199349, "residual_lr": 0.0002628776517719609, "adamw_lr": 0.0026287765177196096, "weight_decay": 0.01, "grad_norm": 0.15442247688770294, "tokens_per_second": 4355450.517981759, "eta_seconds": 663.7598503448564, "mean_data_wait_seconds": 0.00012803077697753906}
{"step": 40900, "total_steps": 46238, "tokens": 22114025472, "loss": 2.9701225902094985, "ce_loss": 2.9701225902094985, "z_loss": 37.719021392591074, "muon_lr": 0.04373816500507511, "residual_lr": 0.0002624289900304506, "adamw_lr": 0.0026242899003045064, "weight_decay": 0.01, "grad_norm": 0.13872778415679932, "tokens_per_second": 4355078.225349037, "eta_seconds": 662.5751168381681, "mean_data_wait_seconds": 7.739663124084473e-05}
{"step": 40910, "total_steps": 46238, "tokens": 22119432192, "loss": 2.990390136386409, "ce_loss": 2.990390136386409, "z_loss": 37.42704209414396, "muon_lr": 0.04366300351558647, "residual_lr": 0.00026197802109351875, "adamw_lr": 0.0026197802109351877, "weight_decay": 0.01, "grad_norm": 0.14838865399360657, "tokens_per_second": 4355026.1625527935, "eta_seconds": 661.341548017643, "mean_data_wait_seconds": 7.510185241699219e-05}
{"step": 40920, "total_steps": 46238, "tokens": 22124838912, "loss": 2.938598730347373, "ce_loss": 2.938598730347373, "z_loss": 37.64142287861217, "muon_lr": 0.04358745903594332, "residual_lr": 0.0002615247542156599, "adamw_lr": 0.002615247542156599, "weight_decay": 0.01, "grad_norm": 0.1437455117702484, "tokens_per_second": 4355074.27082244, "eta_seconds": 660.0927665596649, "mean_data_wait_seconds": 0.00010839104652404785}
{"step": 40930, "total_steps": 46238, "tokens": 22130245632, "loss": 2.9090037219452136, "ce_loss": 2.9090037219452136, "z_loss": 38.18679208466501, "muon_lr": 0.04351153311642088, "residual_lr": 0.00026106919869852525, "adamw_lr": 0.0026106919869852527, "weight_decay": 0.01, "grad_norm": 0.1498052328824997, "tokens_per_second": 4354992.129397179, "eta_seconds": 658.8637174867126, "mean_data_wait_seconds": 7.292628288269043e-05}
{"step": 40940, "total_steps": 46238, "tokens": 22135652352, "loss": 2.9349393591736304, "ce_loss": 2.9349393591736304, "z_loss": 37.451119394013375, "muon_lr": 0.04343522731512204, "residual_lr": 0.0002606113638907322, "adamw_lr": 0.002606113638907322, "weight_decay": 0.01, "grad_norm": 0.163513645529747, "tokens_per_second": 4354539.739864856, "eta_seconds": 657.6905379416478, "mean_data_wait_seconds": 0.06397491693496704}
{"step": 40950, "total_steps": 46238, "tokens": 22141059072, "loss": 2.9674172293056142, "ce_loss": 2.9674172293056142, "z_loss": 39.362328182567246, "muon_lr": 0.04335854319794538, "residual_lr": 0.0002601512591876722, "adamw_lr": 0.002601512591876723, "weight_decay": 0.01, "grad_norm": 0.17018462717533112, "tokens_per_second": 4354559.872252703, "eta_seconds": 656.4458746369751, "mean_data_wait_seconds": 0.00015872716903686523}
{"step": 40960, "total_steps": 46238, "tokens": 22146465792, "loss": 2.9056678089228543, "ce_loss": 2.9056678089228543, "z_loss": 38.15885084325617, "muon_lr": 0.04328148233855303, "residual_lr": 0.0002596888940313181, "adamw_lr": 0.0025968889403131814, "weight_decay": 0.01, "grad_norm": 0.15783801674842834, "tokens_per_second": 4354230.976561844, "eta_seconds": 655.253742706333, "mean_data_wait_seconds": 9.086728096008301e-05}
{"step": 40970, "total_steps": 46238, "tokens": 22151872512, "loss": 2.8525548631494697, "ce_loss": 2.8525548631494697, "z_loss": 37.38207710150516, "muon_lr": 0.043204046318338366, "residual_lr": 0.0002592242779100302, "adamw_lr": 0.002592242779100302, "weight_decay": 0.01, "grad_norm": 0.16600821912288666, "tokens_per_second": 4354170.711674144, "eta_seconds": 654.0210783111613, "mean_data_wait_seconds": 8.171796798706055e-05}
{"step": 40980, "total_steps": 46238, "tokens": 22157279232, "loss": 2.9544535568266204, "ce_loss": 2.9544535568266204, "z_loss": 37.58206023591937, "muon_lr": 0.04312623672639361, "residual_lr": 0.00025875742035836165, "adamw_lr": 0.0025875742035836166, "weight_decay": 0.01, "grad_norm": 0.19048216938972473, "tokens_per_second": 4354169.388340728, "eta_seconds": 652.7795431227214, "mean_data_wait_seconds": 0.00011220574378967285}
{"step": 40990, "total_steps": 46238, "tokens": 22162685952, "loss": 2.888014950535514, "ce_loss": 2.888014950535514, "z_loss": 37.647689906033605, "muon_lr": 0.043048055159477155, "residual_lr": 0.0002582883309568629, "adamw_lr": 0.002582883309568629, "weight_decay": 0.01, "grad_norm": 0.1338760405778885, "tokens_per_second": 4354186.1297619855, "eta_seconds": 651.5353040626848, "mean_data_wait_seconds": 0.00010892748832702637}
{"step": 41000, "total_steps": 46238, "tokens": 22168092672, "loss": 2.966154929363366, "ce_loss": 2.966154929363366, "z_loss": 38.48942389632716, "muon_lr": 0.04296950322198082, "residual_lr": 0.0002578170193318849, "adamw_lr": 0.002578170193318849, "weight_decay": 0.01, "grad_norm": 0.13756631314754486, "tokens_per_second": 4354142.70592085, "eta_seconds": 650.300060250591, "mean_data_wait_seconds": 9.647011756896973e-05}
{"step": 41010, "total_steps": 46238, "tokens": 22173499392, "loss": 2.978682819640998, "ce_loss": 2.978682819640998, "z_loss": 37.03068854592063, "muon_lr": 0.042890582525896975, "residual_lr": 0.0002573434951553818, "adamw_lr": 0.0025734349515538183, "weight_decay": 0.01, "grad_norm": 0.14497779309749603, "tokens_per_second": 4354144.407996276, "eta_seconds": 649.0580649575959, "mean_data_wait_seconds": 9.807944297790527e-05}
{"step": 41020, "total_steps": 46238, "tokens": 22178906112, "loss": 2.977530412601702, "ce_loss": 2.977530412601702, "z_loss": 37.6491756439209, "muon_lr": 0.04281129469078536, "residual_lr": 0.0002568677681447121, "adamw_lr": 0.0025686776814471215, "weight_decay": 0.01, "grad_norm": 0.13523094356060028, "tokens_per_second": 4354038.089529661, "eta_seconds": 647.8321424847021, "mean_data_wait_seconds": 7.194280624389648e-05}
{"step": 41030, "total_steps": 46238, "tokens": 22184312832, "loss": 2.9575073393908413, "ce_loss": 2.9575073393908413, "z_loss": 37.08359056530577, "muon_lr": 0.04273164134373994, "residual_lr": 0.0002563898480624396, "adamw_lr": 0.002563898480624396, "weight_decay": 0.01, "grad_norm": 0.1515979766845703, "tokens_per_second": 4354048.654502165, "eta_seconds": 646.5888021460097, "mean_data_wait_seconds": 9.259581565856934e-05}
{"step": 41040, "total_steps": 46238, "tokens": 22189719552, "loss": 2.988948881626129, "ce_loss": 2.988948881626129, "z_loss": 37.342825918486625, "muon_lr": 0.04265162411935546, "residual_lr": 0.00025590974471613275, "adamw_lr": 0.0025590974471613277, "weight_decay": 0.01, "grad_norm": 0.15139329433441162, "tokens_per_second": 4354071.233542977, "eta_seconds": 645.3436871572636, "mean_data_wait_seconds": 7.835030555725098e-05}
{"step": 41050, "total_steps": 46238, "tokens": 22195126272, "loss": 2.9931919032877143, "ce_loss": 2.9931919032877143, "z_loss": 37.85343025669907, "muon_lr": 0.042571244659693946, "residual_lr": 0.0002554274679581636, "adamw_lr": 0.0025542746795816367, "weight_decay": 0.01, "grad_norm": 0.15420277416706085, "tokens_per_second": 4354115.282609863, "eta_seconds": 644.0954090492062, "mean_data_wait_seconds": 7.671117782592773e-05}
{"step": 41060, "total_steps": 46238, "tokens": 22200532992, "loss": 2.9893084638046497, "ce_loss": 2.9893084638046497, "z_loss": 38.11585134448427, "muon_lr": 0.04249050461425097, "residual_lr": 0.0002549430276855058, "adamw_lr": 0.002549430276855058, "weight_decay": 0.01, "grad_norm": 0.13787661492824554, "tokens_per_second": 4354110.980850926, "eta_seconds": 642.8542947825778, "mean_data_wait_seconds": 0.00011628866195678711}
{"step": 41070, "total_steps": 46238, "tokens": 22205939712, "loss": 2.9985780734004397, "ce_loss": 2.9985780734004397, "z_loss": 37.61458081910104, "muon_lr": 0.04240940563992181, "residual_lr": 0.0002544564338395308, "adamw_lr": 0.0025445643383953083, "weight_decay": 0.01, "grad_norm": 0.12844623625278473, "tokens_per_second": 4354146.147270332, "eta_seconds": 641.6073621578768, "mean_data_wait_seconds": 0.0001474618911743164}
{"step": 41080, "total_steps": 46238, "tokens": 22211346432, "loss": 2.992967300342791, "ce_loss": 2.992967300342791, "z_loss": 36.98417429490523, "muon_lr": 0.04232794940096746, "residual_lr": 0.00025396769640580473, "adamw_lr": 0.0025396769640580476, "weight_decay": 0.01, "grad_norm": 0.15813478827476501, "tokens_per_second": 4354176.378797989, "eta_seconds": 640.3611754399625, "mean_data_wait_seconds": 0.00017383694648742676}
{"step": 41090, "total_steps": 46238, "tokens": 22216753152, "loss": 3.041733568364924, "ce_loss": 3.041733568364924, "z_loss": 37.215642900177926, "muon_lr": 0.04224613756898046, "residual_lr": 0.00025347682541388273, "adamw_lr": 0.0025347682541388273, "weight_decay": 0.01, "grad_norm": 0.14746665954589844, "tokens_per_second": 4354260.282483629, "eta_seconds": 639.1071280683053, "mean_data_wait_seconds": 8.779764175415039e-05}
{"step": 41100, "total_steps": 46238, "tokens": 22222159872, "loss": 3.000900134895787, "ce_loss": 3.000900134895787, "z_loss": 36.529285055218324, "muon_lr": 0.0421639718228506, "residual_lr": 0.00025298383093710355, "adamw_lr": 0.002529838309371036, "weight_decay": 0.01, "grad_norm": 0.13692663609981537, "tokens_per_second": 4354348.064869094, "eta_seconds": 637.8525608479347, "mean_data_wait_seconds": 7.146596908569336e-05}
{"step": 41110, "total_steps": 46238, "tokens": 22227566592, "loss": 3.019708302888003, "ce_loss": 3.019708302888003, "z_loss": 37.46120617606423, "muon_lr": 0.04208145384873049, "residual_lr": 0.00025248872309238294, "adamw_lr": 0.0025248872309238295, "weight_decay": 0.01, "grad_norm": 0.1690712571144104, "tokens_per_second": 4354421.246612578, "eta_seconds": 636.6001787623175, "mean_data_wait_seconds": 8.025765419006348e-05}
{"step": 41120, "total_steps": 46238, "tokens": 22232973312, "loss": 2.994122422102726, "ce_loss": 2.994122422102726, "z_loss": 37.547558293198094, "muon_lr": 0.04199858534000092, "residual_lr": 0.0002519915120400055, "adamw_lr": 0.002519915120400055, "weight_decay": 0.01, "grad_norm": 0.16507449746131897, "tokens_per_second": 4354524.0416603945, "eta_seconds": 635.3435180358033, "mean_data_wait_seconds": 7.092952728271484e-05}
{"step": 41130, "total_steps": 46238, "tokens": 22238380032, "loss": 3.060389630722277, "ce_loss": 3.060389630722277, "z_loss": 36.968787742383554, "muon_lr": 0.04191536799723612, "residual_lr": 0.0002514922079834167, "adamw_lr": 0.0025149220798341672, "weight_decay": 0.01, "grad_norm": 0.1564350128173828, "tokens_per_second": 4354607.079629378, "eta_seconds": 634.0897935239217, "mean_data_wait_seconds": 9.53972339630127e-05}
{"step": 41140, "total_steps": 46238, "tokens": 22243786752, "loss": 2.971014192610076, "ce_loss": 2.971014192610076, "z_loss": 36.51867802937826, "muon_lr": 0.04183180352816888, "residual_lr": 0.00025099082116901325, "adamw_lr": 0.002509908211690133, "weight_decay": 0.01, "grad_norm": 0.13880018889904022, "tokens_per_second": 4354691.964793621, "eta_seconds": 632.8358483860303, "mean_data_wait_seconds": 0.0026664137840270996}
{"step": 41150, "total_steps": 46238, "tokens": 22249193472, "loss": 3.02517551725561, "ce_loss": 3.02517551725561, "z_loss": 37.90404620315089, "muon_lr": 0.04174789364765548, "residual_lr": 0.0002504873618859328, "adamw_lr": 0.0025048736188593287, "weight_decay": 0.01, "grad_norm": 0.13344386219978333, "tokens_per_second": 4354722.747504175, "eta_seconds": 631.5897988170057, "mean_data_wait_seconds": 0.00010392069816589355}
{"step": 41160, "total_steps": 46238, "tokens": 22254600192, "loss": 2.9662964073094455, "ce_loss": 2.9662964073094455, "z_loss": 37.04709576115464, "muon_lr": 0.04166364007764052, "residual_lr": 0.0002499818404658431, "adamw_lr": 0.0024998184046584313, "weight_decay": 0.01, "grad_norm": 0.15244828164577484, "tokens_per_second": 4354754.282028369, "eta_seconds": 630.3436580402031, "mean_data_wait_seconds": 0.00012692809104919434}
{"step": 41170, "total_steps": 46238, "tokens": 22260006912, "loss": 3.007456143697103, "ce_loss": 3.007456143697103, "z_loss": 36.54906275778106, "muon_lr": 0.04157904454712158, "residual_lr": 0.00024947426728272944, "adamw_lr": 0.0024947426728272946, "weight_decay": 0.01, "grad_norm": 0.1635628640651703, "tokens_per_second": 4354749.328706931, "eta_seconds": 629.1028064325974, "mean_data_wait_seconds": 8.32974910736084e-05}
{"step": 41180, "total_steps": 46238, "tokens": 22265413632, "loss": 2.940343566013105, "ce_loss": 2.940343566013105, "z_loss": 37.118636478077285, "muon_lr": 0.04149410879211373, "residual_lr": 0.0002489646527526823, "adamw_lr": 0.0024896465275268237, "weight_decay": 0.01, "grad_norm": 0.12569257616996765, "tokens_per_second": 4354789.054955001, "eta_seconds": 627.8555102201736, "mean_data_wait_seconds": 7.700920104980469e-05}
{"step": 41190, "total_steps": 46238, "tokens": 22270820352, "loss": 2.994750783298955, "ce_loss": 2.994750783298955, "z_loss": 37.25322671370073, "muon_lr": 0.04140883455561389, "residual_lr": 0.0002484530073336833, "adamw_lr": 0.002484530073336833, "weight_decay": 0.01, "grad_norm": 0.1236664429306984, "tokens_per_second": 4354841.213488757, "eta_seconds": 626.6064479108579, "mean_data_wait_seconds": 9.581446647644043e-05}
{"step": 41200, "total_steps": 46238, "tokens": 22276227072, "loss": 2.991616357456554, "ce_loss": 2.991616357456554, "z_loss": 38.17916190985477, "muon_lr": 0.04132322358756511, "residual_lr": 0.0002479393415253906, "adamw_lr": 0.0024793934152539062, "weight_decay": 0.01, "grad_norm": 0.12034694105386734, "tokens_per_second": 4354928.225899489, "eta_seconds": 625.352410587089, "mean_data_wait_seconds": 8.490681648254395e-05}
{"step": 41210, "total_steps": 46238, "tokens": 22281633792, "loss": 2.9622458454334373, "ce_loss": 2.9622458454334373, "z_loss": 36.449678392121285, "muon_lr": 0.0412372776448206, "residual_lr": 0.00024742366586892356, "adamw_lr": 0.002474236658689236, "weight_decay": 0.01, "grad_norm": 0.15722785890102386, "tokens_per_second": 4354998.401688226, "eta_seconds": 624.1008361670987, "mean_data_wait_seconds": 7.018446922302246e-05}
{"step": 41220, "total_steps": 46238, "tokens": 22287040512, "loss": 2.9308794483993994, "ce_loss": 2.9308794483993994, "z_loss": 37.32564203666918, "muon_lr": 0.04115099849110773, "residual_lr": 0.00024690599094664636, "adamw_lr": 0.0024690599094664635, "weight_decay": 0.01, "grad_norm": 0.17779052257537842, "tokens_per_second": 4355173.935671563, "eta_seconds": 622.8342344223107, "mean_data_wait_seconds": 6.648898124694824e-05}
{"step": 41230, "total_steps": 46238, "tokens": 22292447232, "loss": 3.000725359627695, "ce_loss": 3.000725359627695, "z_loss": 37.66938443617387, "muon_lr": 0.041064387896991794, "residual_lr": 0.00024638632738195073, "adamw_lr": 0.0024638632738195076, "weight_decay": 0.01, "grad_norm": 0.16777704656124115, "tokens_per_second": 4355314.061168527, "eta_seconds": 621.5727880881396, "mean_data_wait_seconds": 7.465481758117676e-05}
{"step": 41240, "total_steps": 46238, "tokens": 22297853952, "loss": 2.951025973666798, "ce_loss": 2.951025973666798, "z_loss": 38.925417466597125, "muon_lr": 0.0409774476398397, "residual_lr": 0.00024586468583903817, "adamw_lr": 0.002458646858390382, "weight_decay": 0.01, "grad_norm": 0.13957442343235016, "tokens_per_second": 4355376.423445424, "eta_seconds": 620.3224982934371, "mean_data_wait_seconds": 9.861588478088379e-05}
{"step": 41250, "total_steps": 46238, "tokens": 22303260672, "loss": 2.9794030460444363, "ce_loss": 2.9794030460444363, "z_loss": 38.49344793955485, "muon_lr": 0.040890179503783486, "residual_lr": 0.0002453410770227009, "adamw_lr": 0.0024534107702270094, "weight_decay": 0.01, "grad_norm": 0.1386006772518158, "tokens_per_second": 4355497.585146879, "eta_seconds": 619.0638867979243, "mean_data_wait_seconds": 8.237361907958984e-05}
{"step": 41260, "total_steps": 46238, "tokens": 22308667392, "loss": 2.961985353267554, "ce_loss": 2.961985353267554, "z_loss": 37.92571038910837, "muon_lr": 0.04080258527968372, "residual_lr": 0.0002448155116781023, "adamw_lr": 0.0024481551167810233, "weight_decay": 0.01, "grad_norm": 0.14929242432117462, "tokens_per_second": 4355582.220395326, "eta_seconds": 617.8105263171369, "mean_data_wait_seconds": 9.21487808227539e-05}
{"step": 41270, "total_steps": 46238, "tokens": 22314074112, "loss": 2.9211909120733086, "ce_loss": 2.9211909120733086, "z_loss": 37.43775124983354, "muon_lr": 0.04071466676509272, "residual_lr": 0.00024428800059055625, "adamw_lr": 0.002442880005905563, "weight_decay": 0.01, "grad_norm": 0.1569899022579193, "tokens_per_second": 4355660.034917404, "eta_seconds": 616.5581800396241, "mean_data_wait_seconds": 9.47713851928711e-05}
{"step": 41280, "total_steps": 46238, "tokens": 22319480832, "loss": 2.8922645250956216, "ce_loss": 2.8922645250956216, "z_loss": 37.48032809748794, "muon_lr": 0.04062642576421769, "residual_lr": 0.0002437585545853061, "adamw_lr": 0.0024375855458530615, "weight_decay": 0.01, "grad_norm": 0.1435798704624176, "tokens_per_second": 4355670.333884668, "eta_seconds": 615.3154161255598, "mean_data_wait_seconds": 9.98079776763916e-05}
{"step": 41290, "total_steps": 46238, "tokens": 22324887552, "loss": 2.9919756849606833, "ce_loss": 2.9919756849606833, "z_loss": 39.44381447994348, "muon_lr": 0.040537864087883714, "residual_lr": 0.00024322718452730226, "adamw_lr": 0.0024322718452730228, "weight_decay": 0.01, "grad_norm": 0.14647378027439117, "tokens_per_second": 4355723.581447854, "eta_seconds": 614.0666031683584, "mean_data_wait_seconds": 0.00010910630226135254}
{"step": 41300, "total_steps": 46238, "tokens": 22330294272, "loss": 2.9469216014399673, "ce_loss": 2.9469216014399673, "z_loss": 38.039103796987824, "muon_lr": 0.040448983553496534, "residual_lr": 0.00024269390132097916, "adamw_lr": 0.002426939013209792, "weight_decay": 0.01, "grad_norm": 0.18272629380226135, "tokens_per_second": 4355738.750918057, "eta_seconds": 612.8231780286166, "mean_data_wait_seconds": 0.0005549192428588867}
{"step": 41310, "total_steps": 46238, "tokens": 22335700992, "loss": 3.011887962167913, "ce_loss": 3.011887962167913, "z_loss": 39.12816911755186, "muon_lr": 0.040359785985005316, "residual_lr": 0.00024215871591003185, "adamw_lr": 0.002421587159100319, "weight_decay": 0.01, "grad_norm": 0.14723528921604156, "tokens_per_second": 4355867.6302974, "eta_seconds": 611.5637962621286, "mean_data_wait_seconds": 7.349252700805664e-05}
{"step": 41320, "total_steps": 46238, "tokens": 22341107712, "loss": 3.0016258629885586, "ce_loss": 3.0016258629885586, "z_loss": 38.30654667363022, "muon_lr": 0.0402702732128652, "residual_lr": 0.00024162163927719116, "adamw_lr": 0.002416216392771912, "weight_decay": 0.01, "grad_norm": 0.14014138281345367, "tokens_per_second": 4355989.002979258, "eta_seconds": 610.3055407581934, "mean_data_wait_seconds": 0.0004757046699523926}
{"step": 41330, "total_steps": 46238, "tokens": 22346514432, "loss": 2.9248464956428064, "ce_loss": 2.9248464956428064, "z_loss": 38.1045444950913, "muon_lr": 0.04018044707399972, "residual_lr": 0.00024108268244399828, "adamw_lr": 0.002410826824439983, "weight_decay": 0.01, "grad_norm": 0.11091688275337219, "tokens_per_second": 4356068.105189059, "eta_seconds": 609.0532654527569, "mean_data_wait_seconds": 0.00012156367301940918}
{"step": 41340, "total_steps": 46238, "tokens": 22351921152, "loss": 2.9761805534362793, "ce_loss": 2.9761805534362793, "z_loss": 37.489214059078336, "muon_lr": 0.040090309411763114, "residual_lr": 0.00024054185647057864, "adamw_lr": 0.002405418564705787, "weight_decay": 0.01, "grad_norm": 0.14683479070663452, "tokens_per_second": 4356132.099447517, "eta_seconds": 607.8031436043459, "mean_data_wait_seconds": 0.00013700127601623535}
{"step": 41350, "total_steps": 46238, "tokens": 22357327872, "loss": 2.9569552053104746, "ce_loss": 2.9569552053104746, "z_loss": 38.86151490066991, "muon_lr": 0.03999986207590252, "residual_lr": 0.00023999917245541506, "adamw_lr": 0.002399991724554151, "weight_decay": 0.01, "grad_norm": 0.15902076661586761, "tokens_per_second": 4356146.18308324, "eta_seconds": 606.5600080780187, "mean_data_wait_seconds": 6.908178329467773e-05}
{"step": 41360, "total_steps": 46238, "tokens": 22362734592, "loss": 2.952982434720704, "ce_loss": 2.952982434720704, "z_loss": 38.22619302345045, "muon_lr": 0.03990910692251998, "residual_lr": 0.00023945464153511984, "adamw_lr": 0.0023945464153511987, "weight_decay": 0.01, "grad_norm": 0.14476172626018524, "tokens_per_second": 4356228.690876462, "eta_seconds": 605.3073727562892, "mean_data_wait_seconds": 8.529424667358398e-05}
{"step": 41370, "total_steps": 46238, "tokens": 22368141312, "loss": 2.928454567085613, "ce_loss": 2.928454567085613, "z_loss": 38.8061608690204, "muon_lr": 0.03981804581403439, "residual_lr": 0.00023890827488420632, "adamw_lr": 0.0023890827488420637, "weight_decay": 0.01, "grad_norm": 0.12659859657287598, "tokens_per_second": 4356271.035034672, "eta_seconds": 604.0603541049084, "mean_data_wait_seconds": 7.846951484680176e-05}
{"step": 41380, "total_steps": 46238, "tokens": 22373548032, "loss": 2.93587466803464, "ce_loss": 2.93587466803464, "z_loss": 38.510009765625, "muon_lr": 0.03972668061914325, "residual_lr": 0.0002383600837148595, "adamw_lr": 0.0023836008371485953, "weight_decay": 0.01, "grad_norm": 0.11623295396566391, "tokens_per_second": 4356415.099701535, "eta_seconds": 602.7992842509233, "mean_data_wait_seconds": 6.476044654846191e-05}
{"step": 41390, "total_steps": 46238, "tokens": 22378954752, "loss": 2.9657338839588743, "ce_loss": 2.9657338839588743, "z_loss": 37.52311718102658, "muon_lr": 0.039635013212784335, "residual_lr": 0.00023781007927670597, "adamw_lr": 0.00237810079276706, "weight_decay": 0.01, "grad_norm": 0.12951479852199554, "tokens_per_second": 4356455.207594327, "eta_seconds": 601.5526521268055, "mean_data_wait_seconds": 6.866455078125e-05}
{"step": 41400, "total_steps": 46238, "tokens": 22384361472, "loss": 2.9333565903432444, "ce_loss": 2.9333565903432444, "z_loss": 37.99983695059112, "muon_lr": 0.03954304547609718, "residual_lr": 0.00023725827285658304, "adamw_lr": 0.0023725827285658307, "weight_decay": 0.01, "grad_norm": 0.1218949556350708, "tokens_per_second": 4356533.404140775, "eta_seconds": 600.3007945524507, "mean_data_wait_seconds": 7.238984107971191e-05}
{"step": 41410, "total_steps": 46238, "tokens": 22389768192, "loss": 2.9530728152303984, "ce_loss": 2.9530728152303984, "z_loss": 38.27465424393163, "muon_lr": 0.03945077929638453, "residual_lr": 0.00023670467577830716, "adamw_lr": 0.002367046757783072, "weight_decay": 0.01, "grad_norm": 0.14230024814605713, "tokens_per_second": 4356554.1803296795, "eta_seconds": 599.0568775165567, "mean_data_wait_seconds": 0.00013944506645202637}
{"step": 41420, "total_steps": 46238, "tokens": 22395174912, "loss": 2.974270511757244, "ce_loss": 2.974270511757244, "z_loss": 39.09900211565422, "muon_lr": 0.03935821656707359, "residual_lr": 0.0002361492994024415, "adamw_lr": 0.0023614929940244155, "weight_decay": 0.01, "grad_norm": 0.14612410962581635, "tokens_per_second": 4356532.909533076, "eta_seconds": 597.8187421242586, "mean_data_wait_seconds": 8.824467658996582e-05}
{"step": 41430, "total_steps": 46238, "tokens": 22400581632, "loss": 3.0078079917214136, "ce_loss": 3.0078079917214136, "z_loss": 37.8823966401996, "muon_lr": 0.03926535918767715, "residual_lr": 0.0002355921551260629, "adamw_lr": 0.002355921551260629, "weight_decay": 0.01, "grad_norm": 0.12309423089027405, "tokens_per_second": 4356576.209225655, "eta_seconds": 596.5717524913796, "mean_data_wait_seconds": 0.00010335445404052734}
{"step": 41440, "total_steps": 46238, "tokens": 22405988352, "loss": 2.996592693256609, "ce_loss": 2.996592693256609, "z_loss": 37.83284545667244, "muon_lr": 0.03917220906375464, "residual_lr": 0.0002350332543825278, "adamw_lr": 0.0023503325438252783, "weight_decay": 0.01, "grad_norm": 0.1461659073829651, "tokens_per_second": 4356494.233634978, "eta_seconds": 595.341906796453, "mean_data_wait_seconds": 8.958578109741211e-05}
{"step": 41450, "total_steps": 46238, "tokens": 22411395072, "loss": 2.9305038289590315, "ce_loss": 2.9305038289590315, "z_loss": 37.932065356861465, "muon_lr": 0.03907876810687297, "residual_lr": 0.00023447260864123775, "adamw_lr": 0.002344726086412378, "weight_decay": 0.01, "grad_norm": 0.10558564960956573, "tokens_per_second": 4356459.236306377, "eta_seconds": 594.1056081576933, "mean_data_wait_seconds": 7.858872413635254e-05}
{"step": 41460, "total_steps": 46238, "tokens": 22416801792, "loss": 2.949789917830265, "ce_loss": 2.949789917830265, "z_loss": 38.05657753799901, "muon_lr": 0.038985038234567365, "residual_lr": 0.00023391022940740414, "adamw_lr": 0.0023391022940740416, "weight_decay": 0.01, "grad_norm": 0.12938739359378815, "tokens_per_second": 4356390.713399531, "eta_seconds": 592.8738522134317, "mean_data_wait_seconds": 6.529688835144043e-05}
{"step": 41470, "total_steps": 46238, "tokens": 22422208512, "loss": 2.913711444898085, "ce_loss": 2.913711444898085, "z_loss": 37.72485842849269, "muon_lr": 0.03889102137030198, "residual_lr": 0.00023334612822181186, "adamw_lr": 0.002333461282218119, "weight_decay": 0.01, "grad_norm": 0.1586856245994568, "tokens_per_second": 4356464.553303047, "eta_seconds": 591.6227235329718, "mean_data_wait_seconds": 0.0001302957534790039}
{"step": 41480, "total_steps": 46238, "tokens": 22427615232, "loss": 2.9614856134761465, "ce_loss": 2.9614856134761465, "z_loss": 36.77756139003869, "muon_lr": 0.038796719443430445, "residual_lr": 0.00023278031666058263, "adamw_lr": 0.0023278031666058266, "weight_decay": 0.01, "grad_norm": 0.13874930143356323, "tokens_per_second": 4356386.68672217, "eta_seconds": 590.3921963215817, "mean_data_wait_seconds": 8.916854858398438e-05}
{"step": 41490, "total_steps": 46238, "tokens": 22433021952, "loss": 2.944641225265734, "ce_loss": 2.944641225265734, "z_loss": 36.63027130473744, "muon_lr": 0.03870213438915624, "residual_lr": 0.0002322128063349374, "adamw_lr": 0.0023221280633493745, "weight_decay": 0.01, "grad_norm": 0.14894628524780273, "tokens_per_second": 4356298.857744589, "eta_seconds": 589.1629724708108, "mean_data_wait_seconds": 9.614229202270508e-05}
{"step": 41500, "total_steps": 46238, "tokens": 22438428672, "loss": 2.8623100501118284, "ce_loss": 2.8623100501118284, "z_loss": 38.58195793267453, "muon_lr": 0.03860726814849301, "residual_lr": 0.00023164360889095802, "adamw_lr": 0.0023164360889095805, "weight_decay": 0.01, "grad_norm": 0.15544359385967255, "tokens_per_second": 4356320.878838705, "eta_seconds": 587.9188735708439, "mean_data_wait_seconds": 7.557868957519531e-05}
{"step": 41510, "total_steps": 46238, "tokens": 22443835392, "loss": 2.9753876707770606, "ce_loss": 2.9753876707770606, "z_loss": 38.840000297083996, "muon_lr": 0.038512122668224735, "residual_lr": 0.00023107273600934838, "adamw_lr": 0.0023107273600934844, "weight_decay": 0.01, "grad_norm": 0.16680985689163208, "tokens_per_second": 4356376.645043984, "eta_seconds": 586.6702427825076, "mean_data_wait_seconds": 6.213784217834473e-05}
{"step": 41520, "total_steps": 46238, "tokens": 22449242112, "loss": 2.942100445429484, "ce_loss": 2.942100445429484, "z_loss": 38.095639286619246, "muon_lr": 0.03841669990086577, "residual_lr": 0.00023050019940519455, "adamw_lr": 0.002305001994051946, "weight_decay": 0.01, "grad_norm": 0.1609802395105362, "tokens_per_second": 4356413.478129268, "eta_seconds": 585.4241882235596, "mean_data_wait_seconds": 9.444355964660645e-05}
{"step": 41530, "total_steps": 46238, "tokens": 22454648832, "loss": 2.96016033671119, "ce_loss": 2.96016033671119, "z_loss": 37.63471958853982, "muon_lr": 0.03832100180462076, "residual_lr": 0.00022992601082772452, "adamw_lr": 0.0022992601082772457, "weight_decay": 0.01, "grad_norm": 0.11996142566204071, "tokens_per_second": 4356566.506737417, "eta_seconds": 584.1625739132534, "mean_data_wait_seconds": 6.920099258422852e-05}
{"step": 41540, "total_steps": 46238, "tokens": 22460055552, "loss": 2.947544567512743, "ce_loss": 2.947544567512743, "z_loss": 37.61905254017223, "muon_lr": 0.0382250303433445, "residual_lr": 0.00022935018206006696, "adamw_lr": 0.0022935018206006698, "weight_decay": 0.01, "grad_norm": 0.1426171064376831, "tokens_per_second": 4356672.559537973, "eta_seconds": 582.9073333593193, "mean_data_wait_seconds": 6.401538848876953e-05}
{"step": 41550, "total_steps": 46238, "tokens": 22465462272, "loss": 2.98837073282762, "ce_loss": 2.98837073282762, "z_loss": 38.18336001309481, "muon_lr": 0.038128787486501586, "residual_lr": 0.00022877272491900948, "adamw_lr": 0.002287727249190095, "weight_decay": 0.01, "grad_norm": 0.13201260566711426, "tokens_per_second": 4356765.322513778, "eta_seconds": 581.6539281803343, "mean_data_wait_seconds": 6.29127025604248e-05}
{"step": 41560, "total_steps": 46238, "tokens": 22470868992, "loss": 2.9596243699391684, "ce_loss": 2.9596243699391684, "z_loss": 39.18715991395892, "muon_lr": 0.03803227520912601, "residual_lr": 0.00022819365125475603, "adamw_lr": 0.0022819365125475604, "weight_decay": 0.01, "grad_norm": 0.16094546020030975, "tokens_per_second": 4356773.456560486, "eta_seconds": 580.4118504698049, "mean_data_wait_seconds": 0.0001068115234375}
{"step": 41570, "total_steps": 46238, "tokens": 22476275712, "loss": 2.9404154564395095, "ce_loss": 2.9404154564395095, "z_loss": 37.88202288656524, "muon_lr": 0.037935495491780685, "residual_lr": 0.00022761297295068405, "adamw_lr": 0.002276129729506841, "weight_decay": 0.01, "grad_norm": 0.1314976066350937, "tokens_per_second": 4356792.891839746, "eta_seconds": 579.1682750690675, "mean_data_wait_seconds": 5.9992074966430664e-05}
{"step": 41580, "total_steps": 46238, "tokens": 22481682432, "loss": 2.980764981472131, "ce_loss": 2.980764981472131, "z_loss": 38.39625922116366, "muon_lr": 0.037838450320516706, "residual_lr": 0.0002270307019231002, "adamw_lr": 0.0022703070192310023, "weight_decay": 0.01, "grad_norm": 0.16975605487823486, "tokens_per_second": 4356793.728147804, "eta_seconds": 577.927177899798, "mean_data_wait_seconds": 7.840991020202637e-05}
{"step": 41590, "total_steps": 46238, "tokens": 22487089152, "loss": 2.9631817665967075, "ce_loss": 2.9631817665967075, "z_loss": 38.711107196229875, "muon_lr": 0.03774114168683265, "residual_lr": 0.00022644685012099586, "adamw_lr": 0.002264468501209959, "weight_decay": 0.01, "grad_norm": 0.13848380744457245, "tokens_per_second": 4356718.200379414, "eta_seconds": 576.6961892970709, "mean_data_wait_seconds": 0.0001100301742553711}
{"step": 41600, "total_steps": 46238, "tokens": 22492495872, "loss": 2.9006027579307556, "ce_loss": 2.9006027579307556, "z_loss": 37.5510859633937, "muon_lr": 0.03764357158763373, "residual_lr": 0.00022586142952580232, "adamw_lr": 0.0022586142952580233, "weight_decay": 0.01, "grad_norm": 0.13437439501285553, "tokens_per_second": 4356815.786254795, "eta_seconds": 575.4422924902109, "mean_data_wait_seconds": 6.16908073425293e-05}
{"step": 41610, "total_steps": 46238, "tokens": 22497902592, "loss": 2.9467701189445727, "ce_loss": 2.9467701189445727, "z_loss": 37.8358160943696, "muon_lr": 0.037545742025190745, "residual_lr": 0.00022527445215114443, "adamw_lr": 0.002252744521511445, "weight_decay": 0.01, "grad_norm": 0.15401767194271088, "tokens_per_second": 4356886.603366062, "eta_seconds": 574.1919796735665, "mean_data_wait_seconds": 6.559491157531738e-05}
{"step": 41620, "total_steps": 46238, "tokens": 22503309312, "loss": 2.9260257067102375, "ce_loss": 2.9260257067102375, "z_loss": 38.40666617769183, "muon_lr": 0.03744765500709908, "residual_lr": 0.00022468593004259445, "adamw_lr": 0.002246859300425945, "weight_decay": 0.01, "grad_norm": 0.1427965760231018, "tokens_per_second": 4356979.003675472, "eta_seconds": 572.9388693161429, "mean_data_wait_seconds": 6.23464584350586e-05}
{"step": 41630, "total_steps": 46238, "tokens": 22508716032, "loss": 2.9192576119394014, "ce_loss": 2.9192576119394014, "z_loss": 37.684222076878406, "muon_lr": 0.03734931254623741, "residual_lr": 0.00022409587527742446, "adamw_lr": 0.0022409587527742448, "weight_decay": 0.01, "grad_norm": 0.13966557383537292, "tokens_per_second": 4357050.879525829, "eta_seconds": 571.6885051090057, "mean_data_wait_seconds": 9.849667549133301e-05}
{"step": 41640, "total_steps": 46238, "tokens": 22514122752, "loss": 3.009534194613948, "ce_loss": 3.009534194613948, "z_loss": 37.874572002526484, "muon_lr": 0.037250716660726484, "residual_lr": 0.00022350429996435885, "adamw_lr": 0.002235042999643589, "weight_decay": 0.01, "grad_norm": 0.12009128928184509, "tokens_per_second": 4357084.008498449, "eta_seconds": 570.4432549733071, "mean_data_wait_seconds": 0.00011369585990905762}
{"step": 41650, "total_steps": 46238, "tokens": 22519529472, "loss": 2.999102093956687, "ce_loss": 2.999102093956687, "z_loss": 38.297607364076555, "muon_lr": 0.03715186937388763, "residual_lr": 0.00022291121624332577, "adamw_lr": 0.002229112162433258, "weight_decay": 0.01, "grad_norm": 0.1630558967590332, "tokens_per_second": 4357175.939214556, "eta_seconds": 569.1903422304924, "mean_data_wait_seconds": 0.00012725591659545898}
{"step": 41660, "total_steps": 46238, "tokens": 22524936192, "loss": 2.98011889963439, "ce_loss": 2.98011889963439, "z_loss": 38.75227266369444, "muon_lr": 0.037052772714201314, "residual_lr": 0.00022231663628520786, "adamw_lr": 0.002223166362852079, "weight_decay": 0.01, "grad_norm": 0.14609283208847046, "tokens_per_second": 4357241.593776661, "eta_seconds": 567.9409072782397, "mean_data_wait_seconds": 0.0001575946807861328}
{"step": 41670, "total_steps": 46238, "tokens": 22530342912, "loss": 2.9515468658822956, "ce_loss": 2.9515468658822956, "z_loss": 38.59080271287398, "muon_lr": 0.036953428715265434, "residual_lr": 0.00022172057229159255, "adamw_lr": 0.0022172057229159258, "weight_decay": 0.01, "grad_norm": 0.15387281775474548, "tokens_per_second": 4357268.156790494, "eta_seconds": 566.6965940922985, "mean_data_wait_seconds": 0.0002873837947845459}
{"step": 41680, "total_steps": 46238, "tokens": 22535749632, "loss": 2.960205856597785, "ce_loss": 2.960205856597785, "z_loss": 38.74485989772912, "muon_lr": 0.036853839415753646, "residual_lr": 0.00022112303649452183, "adamw_lr": 0.0022112303649452184, "weight_decay": 0.01, "grad_norm": 0.11706569045782089, "tokens_per_second": 4357425.031797685, "eta_seconds": 565.4353858116809, "mean_data_wait_seconds": 0.00013110041618347168}
{"step": 41690, "total_steps": 46238, "tokens": 22541156352, "loss": 2.9891032070824592, "ce_loss": 2.9891032070824592, "z_loss": 38.73582877534808, "muon_lr": 0.03675400685937349, "residual_lr": 0.00022052404115624092, "adamw_lr": 0.0022052404115624094, "weight_decay": 0.01, "grad_norm": 0.13027624785900116, "tokens_per_second": 4357454.80399652, "eta_seconds": 564.1907247655673, "mean_data_wait_seconds": 0.00012683868408203125}
{"step": 41700, "total_steps": 46238, "tokens": 22546563072, "loss": 2.960832068414399, "ce_loss": 2.960832068414399, "z_loss": 37.74943568489768, "muon_lr": 0.036653933094824495, "residual_lr": 0.00021992359856894695, "adamw_lr": 0.0021992359856894696, "weight_decay": 0.01, "grad_norm": 0.13072657585144043, "tokens_per_second": 4357513.036244349, "eta_seconds": 562.9424039805548, "mean_data_wait_seconds": 6.935000419616699e-05}
{"step": 41710, "total_steps": 46238, "tokens": 22551969792, "loss": 2.9702752575729834, "ce_loss": 2.9702752575729834, "z_loss": 38.212213747429125, "muon_lr": 0.03655362017575608, "residual_lr": 0.00021932172105453641, "adamw_lr": 0.0021932172105453645, "weight_decay": 0.01, "grad_norm": 0.13964363932609558, "tokens_per_second": 4357520.747529461, "eta_seconds": 561.7006288237877, "mean_data_wait_seconds": 0.00010666251182556152}
{"step": 41720, "total_steps": 46238, "tokens": 22557376512, "loss": 2.9853255532004614, "ce_loss": 2.9853255532004614, "z_loss": 38.04058730963504, "muon_lr": 0.03645307016072546, "residual_lr": 0.0002187184209643527, "adamw_lr": 0.0021871842096435274, "weight_decay": 0.01, "grad_norm": 0.13843934237957, "tokens_per_second": 4357566.2975217085, "eta_seconds": 560.4539913458043, "mean_data_wait_seconds": 0.00018486380577087402}
{"step": 41730, "total_steps": 46238, "tokens": 22562783232, "loss": 3.0040811878262144, "ce_loss": 3.0040811878262144, "z_loss": 37.68480338472308, "muon_lr": 0.03635228511315535, "residual_lr": 0.0002181137106789321, "adamw_lr": 0.0021811371067893214, "weight_decay": 0.01, "grad_norm": 0.12681204080581665, "tokens_per_second": 4356862.752014321, "eta_seconds": 559.3035270329282, "mean_data_wait_seconds": 0.00011676549911499023}
{"step": 41740, "total_steps": 46238, "tokens": 22568189952, "loss": 2.933817233100082, "ce_loss": 2.933817233100082, "z_loss": 38.053642533042215, "muon_lr": 0.03625126710129171, "residual_lr": 0.00021750760260775022, "adamw_lr": 0.0021750760260775023, "weight_decay": 0.01, "grad_norm": 0.12815643846988678, "tokens_per_second": 4356286.9134353595, "eta_seconds": 558.1363285556876, "mean_data_wait_seconds": 0.00012806057929992676}
{"step": 41750, "total_steps": 46238, "tokens": 22573596672, "loss": 3.021150706392346, "ce_loss": 3.021150706392346, "z_loss": 38.51432612448028, "muon_lr": 0.036150018198161174, "residual_lr": 0.00021690010918896702, "adamw_lr": 0.00216900109188967, "weight_decay": 0.01, "grad_norm": 0.13847501575946808, "tokens_per_second": 4356344.176713258, "eta_seconds": 556.8878779064576, "mean_data_wait_seconds": 9.66489315032959e-05}
{"step": 41760, "total_steps": 46238, "tokens": 22579003392, "loss": 2.9856350313533437, "ce_loss": 2.9856350313533437, "z_loss": 38.337733499931566, "muon_lr": 0.03604854048152863, "residual_lr": 0.00021629124288917177, "adamw_lr": 0.0021629124288917176, "weight_decay": 0.01, "grad_norm": 0.16185487806797028, "tokens_per_second": 4356283.457608556, "eta_seconds": 555.65450860923, "mean_data_wait_seconds": 0.00010436773300170898}
{"step": 41770, "total_steps": 46238, "tokens": 22584410112, "loss": 2.9162008328871294, "ce_loss": 2.9162008328871294, "z_loss": 38.53014714790113, "muon_lr": 0.035946836033854505, "residual_lr": 0.00021568101620312699, "adamw_lr": 0.0021568101620312702, "weight_decay": 0.01, "grad_norm": 0.1363743394613266, "tokens_per_second": 4355837.18609139, "eta_seconds": 554.4701789387145, "mean_data_wait_seconds": 6.398558616638184e-05}
{"step": 41780, "total_steps": 46238, "tokens": 22589816832, "loss": 2.949561561598922, "ce_loss": 2.949561561598922, "z_loss": 39.61611863338586, "muon_lr": 0.03584490694225206, "residual_lr": 0.0002150694416535123, "adamw_lr": 0.002150694416535123, "weight_decay": 0.01, "grad_norm": 0.14955081045627594, "tokens_per_second": 4355826.913646501, "eta_seconds": 553.2302251153147, "mean_data_wait_seconds": 7.173418998718262e-05}
{"step": 41790, "total_steps": 46238, "tokens": 22595223552, "loss": 2.8898650353605095, "ce_loss": 2.8898650353605095, "z_loss": 39.1357276800907, "muon_lr": 0.035742755298444555, "residual_lr": 0.0002144565317906673, "adamw_lr": 0.002144565317906673, "weight_decay": 0.01, "grad_norm": 0.12647494673728943, "tokens_per_second": 4355862.674682129, "eta_seconds": 551.9844319186348, "mean_data_wait_seconds": 8.20457935333252e-05}
{"step": 41800, "total_steps": 46238, "tokens": 22600630272, "loss": 2.879632480216749, "ce_loss": 2.879632480216749, "z_loss": 38.85586793494947, "muon_lr": 0.03564038319872231, "residual_lr": 0.00021384229919233383, "adamw_lr": 0.0021384229919233387, "weight_decay": 0.01, "grad_norm": 0.1417161524295807, "tokens_per_second": 4355806.527554897, "eta_seconds": 550.7502798446471, "mean_data_wait_seconds": 0.00012561678886413574}
{"step": 41810, "total_steps": 46238, "tokens": 22606036992, "loss": 2.952289985887932, "ce_loss": 2.952289985887932, "z_loss": 38.617572264237836, "muon_lr": 0.03553779274389973, "residual_lr": 0.00021322675646339839, "adamw_lr": 0.002132267564633984, "weight_decay": 0.01, "grad_norm": 0.1642017960548401, "tokens_per_second": 4355815.087602865, "eta_seconds": 549.5079326972176, "mean_data_wait_seconds": 8.356571197509766e-05}
{"step": 41820, "total_steps": 46238, "tokens": 22611443712, "loss": 2.9149150758078606, "ce_loss": 2.9149150758078606, "z_loss": 38.198739340811066, "muon_lr": 0.035434986039272136, "residual_lr": 0.0002126099162356328, "adamw_lr": 0.0021260991623563283, "weight_decay": 0.01, "grad_norm": 0.1557641476392746, "tokens_per_second": 4355884.615423319, "eta_seconds": 548.2579165536303, "mean_data_wait_seconds": 0.00010713934898376465}
{"step": 41830, "total_steps": 46238, "tokens": 22616850432, "loss": 2.9295912258552783, "ce_loss": 2.9295912258552783, "z_loss": 39.79437088243889, "muon_lr": 0.03533196519457261, "residual_lr": 0.00021199179116743563, "adamw_lr": 0.0021199179116743565, "weight_decay": 0.01, "grad_norm": 0.15621210634708405, "tokens_per_second": 4355971.978926186, "eta_seconds": 547.0057005709624, "mean_data_wait_seconds": 0.00011220574378967285}
{"step": 41840, "total_steps": 46238, "tokens": 22622257152, "loss": 2.9840196642008694, "ce_loss": 2.9840196642008694, "z_loss": 39.5541857806119, "muon_lr": 0.03522873232392864, "residual_lr": 0.00021137239394357184, "adamw_lr": 0.0021137239394357185, "weight_decay": 0.01, "grad_norm": 0.14396828413009644, "tokens_per_second": 4355911.867753915, "eta_seconds": 545.7720119635594, "mean_data_wait_seconds": 8.058547973632812e-05}
{"step": 41850, "total_steps": 46238, "tokens": 22627663872, "loss": 2.9193665078192046, "ce_loss": 2.9193665078192046, "z_loss": 38.28340365669944, "muon_lr": 0.03512528954581883, "residual_lr": 0.00021075173727491296, "adamw_lr": 0.00210751737274913, "weight_decay": 0.01, "grad_norm": 0.13206084072589874, "tokens_per_second": 4355932.652905391, "eta_seconds": 544.5281763982124, "mean_data_wait_seconds": 8.472800254821777e-05}
{"step": 41860, "total_steps": 46238, "tokens": 22633070592, "loss": 2.8868944428183814, "ce_loss": 2.8868944428183814, "z_loss": 38.69149656006784, "muon_lr": 0.03502163898302932, "residual_lr": 0.0002101298338981759, "adamw_lr": 0.002101298338981759, "weight_decay": 0.01, "grad_norm": 0.13459986448287964, "tokens_per_second": 4356040.482944119, "eta_seconds": 543.27349648517, "mean_data_wait_seconds": 8.133053779602051e-05}
{"step": 41870, "total_steps": 46238, "tokens": 22638477312, "loss": 2.9494300430471245, "ce_loss": 2.9494300430471245, "z_loss": 37.50964335239295, "muon_lr": 0.0349177827626103, "residual_lr": 0.00020950669657566177, "adamw_lr": 0.002095066965756618, "weight_decay": 0.01, "grad_norm": 0.15330076217651367, "tokens_per_second": 4356029.92827594, "eta_seconds": 542.0336092443926, "mean_data_wait_seconds": 0.0001125633716583252}
{"step": 41880, "total_steps": 46238, "tokens": 22643884032, "loss": 2.884754962993391, "ce_loss": 2.884754962993391, "z_loss": 39.4816918517604, "muon_lr": 0.034813723015832326, "residual_lr": 0.00020888233809499395, "adamw_lr": 0.0020888233809499396, "weight_decay": 0.01, "grad_norm": 0.15790264308452606, "tokens_per_second": 4355939.564769721, "eta_seconds": 540.8036243323169, "mean_data_wait_seconds": 9.575486183166504e-05}
{"step": 41890, "total_steps": 46238, "tokens": 22649290752, "loss": 2.950840110128576, "ce_loss": 2.950840110128576, "z_loss": 38.314317067464195, "muon_lr": 0.034709461878142604, "residual_lr": 0.0002082567712688556, "adamw_lr": 0.002082567712688556, "weight_decay": 0.01, "grad_norm": 0.12851475179195404, "tokens_per_second": 4355987.366761749, "eta_seconds": 539.5564739085136, "mean_data_wait_seconds": 0.0001207590103149414}
{"step": 41900, "total_steps": 46238, "tokens": 22654697472, "loss": 2.866181658976006, "ce_loss": 2.866181658976006, "z_loss": 39.325952616604894, "muon_lr": 0.034605001489121115, "residual_lr": 0.00020763000893472665, "adamw_lr": 0.0020763000893472667, "weight_decay": 0.01, "grad_norm": 0.12643256783485413, "tokens_per_second": 4356034.90036373, "eta_seconds": 538.3093840235763, "mean_data_wait_seconds": 0.0001169741153717041}
{"step": 41910, "total_steps": 46238, "tokens": 22660104192, "loss": 2.986599786715074, "ce_loss": 2.986599786715074, "z_loss": 38.879317717118695, "muon_lr": 0.03450034399243681, "residual_lr": 0.0002070020639546208, "adamw_lr": 0.002070020639546208, "weight_decay": 0.01, "grad_norm": 0.132830411195755, "tokens_per_second": 4356019.104869358, "eta_seconds": 537.0701293262955, "mean_data_wait_seconds": 8.299946784973145e-05}
{"step": 41920, "total_steps": 46238, "tokens": 22665510912, "loss": 2.950210878343293, "ce_loss": 2.950210878343293, "z_loss": 39.076858058120266, "muon_lr": 0.03439549153580349, "residual_lr": 0.00020637294921482092, "adamw_lr": 0.0020637294921482096, "weight_decay": 0.01, "grad_norm": 0.1196967363357544, "tokens_per_second": 4356009.141632186, "eta_seconds": 535.8301482180603, "mean_data_wait_seconds": 7.62939453125e-05}
{"step": 41930, "total_steps": 46238, "tokens": 22670917632, "loss": 3.0007319071076135, "ce_loss": 3.0007319071076135, "z_loss": 37.54182936928489, "muon_lr": 0.03429044627093586, "residual_lr": 0.00020574267762561515, "adamw_lr": 0.0020574267762561516, "weight_decay": 0.01, "grad_norm": 0.12354446202516556, "tokens_per_second": 4355971.387782002, "eta_seconds": 534.5935720633205, "mean_data_wait_seconds": 0.00011077523231506348}
{"step": 41940, "total_steps": 46238, "tokens": 22676324352, "loss": 3.0107680306290137, "ce_loss": 3.0107680306290137, "z_loss": 38.937095353097625, "muon_lr": 0.03418521035350527, "residual_lr": 0.0002051112621210316, "adamw_lr": 0.002051112621210316, "weight_decay": 0.01, "grad_norm": 0.13616542518138885, "tokens_per_second": 4355946.1812504595, "eta_seconds": 533.3554381365337, "mean_data_wait_seconds": 6.264448165893555e-05}
{"step": 41950, "total_steps": 46238, "tokens": 22681731072, "loss": 3.009536893078775, "ce_loss": 3.009536893078775, "z_loss": 37.56511766260321, "muon_lr": 0.034079785943095534, "residual_lr": 0.00020447871565857316, "adamw_lr": 0.002044787156585732, "weight_decay": 0.01, "grad_norm": 0.12924695014953613, "tokens_per_second": 4355899.112571271, "eta_seconds": 532.1199605634978, "mean_data_wait_seconds": 7.453560829162598e-05}
{"step": 41960, "total_steps": 46238, "tokens": 22687137792, "loss": 2.979289461265911, "ce_loss": 2.979289461265911, "z_loss": 38.46634645173044, "muon_lr": 0.03397417520315858, "residual_lr": 0.00020384505121895147, "adamw_lr": 0.002038450512189515, "weight_decay": 0.01, "grad_norm": 0.13563700020313263, "tokens_per_second": 4355938.351335195, "eta_seconds": 530.8739374815027, "mean_data_wait_seconds": 8.699297904968262e-05}
{"step": 41970, "total_steps": 46238, "tokens": 22692544512, "loss": 2.9730049751021643, "ce_loss": 2.9730049751021643, "z_loss": 37.84497720544989, "muon_lr": 0.03386838030097011, "residual_lr": 0.00020321028180582065, "adamw_lr": 0.0020321028180582067, "weight_decay": 0.01, "grad_norm": 0.12570743262767792, "tokens_per_second": 4356028.735297794, "eta_seconds": 529.6217183567962, "mean_data_wait_seconds": 8.025765419006348e-05}
{"step": 41980, "total_steps": 46238, "tokens": 22697951232, "loss": 2.962673808589126, "ce_loss": 2.962673808589126, "z_loss": 38.56558270887895, "muon_lr": 0.03376240340758506, "residual_lr": 0.00020257442044551032, "adamw_lr": 0.0020257442044551034, "weight_decay": 0.01, "grad_norm": 0.12034713476896286, "tokens_per_second": 4356103.815606572, "eta_seconds": 528.3714074384393, "mean_data_wait_seconds": 0.00013145804405212402}
{"step": 41990, "total_steps": 46238, "tokens": 22703357952, "loss": 3.0043282689470234, "ce_loss": 3.0043282689470234, "z_loss": 38.382293701171875, "muon_lr": 0.03365624669779306, "residual_lr": 0.00020193748018675832, "adamw_lr": 0.0020193748018675836, "weight_decay": 0.01, "grad_norm": 0.14272895455360413, "tokens_per_second": 4356069.494230104, "eta_seconds": 527.134378145599, "mean_data_wait_seconds": 7.271766662597656e-05}
{"step": 42000, "total_steps": 46238, "tokens": 22708764672, "loss": 2.9589949665647564, "ce_loss": 2.9589949665647564, "z_loss": 39.168254389907375, "muon_lr": 0.03354991235007383, "residual_lr": 0.00020129947410044296, "adamw_lr": 0.00201299474100443, "weight_decay": 0.01, "grad_norm": 0.11966510117053986, "tokens_per_second": 4356166.315219579, "eta_seconds": 525.8814972229836, "mean_data_wait_seconds": 7.635354995727539e-05}
{"step": 42010, "total_steps": 46238, "tokens": 22714171392, "loss": 2.98765557643139, "ce_loss": 2.98765557643139, "z_loss": 38.25963115692139, "muon_lr": 0.033443402546552474, "residual_lr": 0.00020066041527931484, "adamw_lr": 0.0020066041527931487, "weight_decay": 0.01, "grad_norm": 0.13951818645000458, "tokens_per_second": 4356182.867822156, "eta_seconds": 524.6383389645395, "mean_data_wait_seconds": 7.43567943572998e-05}
{"step": 42020, "total_steps": 46238, "tokens": 22719578112, "loss": 2.9830859556342615, "ce_loss": 2.9830859556342615, "z_loss": 38.3207591085723, "muon_lr": 0.033336719472954676, "residual_lr": 0.000200020316837728, "adamw_lr": 0.0020002031683772803, "weight_decay": 0.01, "grad_norm": 0.11698548495769501, "tokens_per_second": 4356270.842849365, "eta_seconds": 523.3866089255093, "mean_data_wait_seconds": 5.97834587097168e-05}
{"step": 42030, "total_steps": 46238, "tokens": 22724984832, "loss": 2.8737327926086658, "ce_loss": 2.8737327926086658, "z_loss": 38.29096083207564, "muon_lr": 0.03322986531856184, "residual_lr": 0.00019937919191137103, "adamw_lr": 0.0019937919191137105, "weight_decay": 0.01, "grad_norm": 0.1258978396654129, "tokens_per_second": 4356343.022394513, "eta_seconds": 522.1368226301281, "mean_data_wait_seconds": 0.00015291571617126465}
{"step": 42040, "total_steps": 46238, "tokens": 22730391552, "loss": 3.064686451897477, "ce_loss": 3.064686451897477, "z_loss": 38.40280879627574, "muon_lr": 0.033122842276166214, "residual_lr": 0.00019873705365699727, "adamw_lr": 0.001987370536569973, "weight_decay": 0.01, "grad_norm": 0.12641949951648712, "tokens_per_second": 4356387.639879, "eta_seconds": 520.8903733054912, "mean_data_wait_seconds": 0.00010529160499572754}
{"step": 42050, "total_steps": 46238, "tokens": 22735798272, "loss": 2.98744491555474, "ce_loss": 2.98744491555474, "z_loss": 39.41110524264249, "muon_lr": 0.033015652542025826, "residual_lr": 0.00019809391525215492, "adamw_lr": 0.0019809391525215493, "weight_decay": 0.01, "grad_norm": 0.11465343087911606, "tokens_per_second": 4356484.898585718, "eta_seconds": 519.6376704381356, "mean_data_wait_seconds": 6.0051679611206055e-05}
{"step": 42060, "total_steps": 46238, "tokens": 22741204992, "loss": 2.954233184005275, "ce_loss": 2.954233184005275, "z_loss": 39.48302948113644, "muon_lr": 0.03290829831581947, "residual_lr": 0.0001974497898949168, "adamw_lr": 0.001974497898949168, "weight_decay": 0.01, "grad_norm": 0.13625195622444153, "tokens_per_second": 4356578.482734472, "eta_seconds": 518.3854607348861, "mean_data_wait_seconds": 0.00016102194786071777}
{"step": 42070, "total_steps": 46238, "tokens": 22746611712, "loss": 2.9879992243015403, "ce_loss": 2.9879992243015403, "z_loss": 37.9553450382117, "muon_lr": 0.03280078180060152, "residual_lr": 0.00019680469080360905, "adamw_lr": 0.001968046908036091, "weight_decay": 0.01, "grad_norm": 0.1167425736784935, "tokens_per_second": 4356648.467269287, "eta_seconds": 517.136106097665, "mean_data_wait_seconds": 6.026029586791992e-05}
{"step": 42080, "total_steps": 46238, "tokens": 22752018432, "loss": 2.946931537353631, "ce_loss": 2.946931537353631, "z_loss": 38.85497182788271, "muon_lr": 0.03269310520275675, "residual_lr": 0.00019615863121654045, "adamw_lr": 0.0019615863121654046, "weight_decay": 0.01, "grad_norm": 0.13423612713813782, "tokens_per_second": 4356623.091334833, "eta_seconds": 515.8980836488571, "mean_data_wait_seconds": 6.136298179626465e-05}
{"step": 42090, "total_steps": 46238, "tokens": 22757425152, "loss": 2.9177181630423576, "ce_loss": 2.9177181630423576, "z_loss": 39.434853900562636, "muon_lr": 0.032585270731955045, "residual_lr": 0.00019551162439173023, "adamw_lr": 0.0019551162439173024, "weight_decay": 0.01, "grad_norm": 0.12045305967330933, "tokens_per_second": 4356718.731697508, "eta_seconds": 514.645751098645, "mean_data_wait_seconds": 7.086992263793945e-05}
{"step": 42100, "total_steps": 46238, "tokens": 22762831872, "loss": 2.9443052104025176, "ce_loss": 2.9443052104025176, "z_loss": 40.03626528653231, "muon_lr": 0.03247728060110606, "residual_lr": 0.0001948636836066363, "adamw_lr": 0.0019486368360663634, "weight_decay": 0.01, "grad_norm": 0.11952099949121475, "tokens_per_second": 4356738.200633704, "eta_seconds": 513.4024494918366, "mean_data_wait_seconds": 5.933642387390137e-05}
{"step": 42110, "total_steps": 46238, "tokens": 22768238592, "loss": 2.9056890823624353, "ce_loss": 2.9056890823624353, "z_loss": 39.80491964744799, "muon_lr": 0.0323691370263138, "residual_lr": 0.00019421482215788277, "adamw_lr": 0.001942148221578828, "weight_decay": 0.01, "grad_norm": 0.11086852848529816, "tokens_per_second": 4356837.156568281, "eta_seconds": 512.1498150639061, "mean_data_wait_seconds": 6.517767906188965e-05}
{"step": 42120, "total_steps": 46238, "tokens": 22773645312, "loss": 2.935319797559218, "ce_loss": 2.935319797559218, "z_loss": 40.487241138111465, "muon_lr": 0.032260842226831156, "residual_lr": 0.00019356505336098691, "adamw_lr": 0.0019356505336098693, "weight_decay": 0.01, "grad_norm": 0.142150416970253, "tokens_per_second": 4356930.002103233, "eta_seconds": 510.89795404687766, "mean_data_wait_seconds": 9.98079776763916e-05}
{"step": 42130, "total_steps": 46238, "tokens": 22779052032, "loss": 2.9388402899106345, "ce_loss": 2.9388402899106345, "z_loss": 38.507273731809676, "muon_lr": 0.03215239842501435, "residual_lr": 0.0001929143905500861, "adamw_lr": 0.001929143905500861, "weight_decay": 0.01, "grad_norm": 0.12167667597532272, "tokens_per_second": 4357001.648926336, "eta_seconds": 509.6486260332702, "mean_data_wait_seconds": 6.0170888900756836e-05}
{"step": 42140, "total_steps": 46238, "tokens": 22784458752, "loss": 2.9547845540624675, "ce_loss": 2.9547845540624675, "z_loss": 38.77996343554872, "muon_lr": 0.032043807846277333, "residual_lr": 0.000192262847077664, "adamw_lr": 0.0019226284707766401, "weight_decay": 0.01, "grad_norm": 0.14258816838264465, "tokens_per_second": 4357070.229905092, "eta_seconds": 508.3996968412995, "mean_data_wait_seconds": 0.00011298060417175293}
{"step": 42150, "total_steps": 46238, "tokens": 22789865472, "loss": 2.955782523660949, "ce_loss": 2.955782523660949, "z_loss": 38.453683968746304, "muon_lr": 0.031935072719046116, "residual_lr": 0.00019161043631427666, "adamw_lr": 0.001916104363142767, "weight_decay": 0.01, "grad_norm": 0.1062307208776474, "tokens_per_second": 4357131.512181136, "eta_seconds": 507.1516565020626, "mean_data_wait_seconds": 0.00010564923286437988}
{"step": 42160, "total_steps": 46238, "tokens": 22795272192, "loss": 2.9064070510141775, "ce_loss": 2.9064070510141775, "z_loss": 39.63840475949374, "muon_lr": 0.03182619527471305, "residual_lr": 0.00019095717164827828, "adamw_lr": 0.001909571716482783, "weight_decay": 0.01, "grad_norm": 0.11426805704832077, "tokens_per_second": 4357271.692494779, "eta_seconds": 505.89449076513864, "mean_data_wait_seconds": 8.416175842285156e-05}
{"step": 42170, "total_steps": 46238, "tokens": 22800678912, "loss": 2.936706654953234, "ce_loss": 2.936706654953234, "z_loss": 38.637873245008066, "muon_lr": 0.03171717774759101, "residual_lr": 0.00019030306648554603, "adamw_lr": 0.0019030306648554604, "weight_decay": 0.01, "grad_norm": 0.11354298889636993, "tokens_per_second": 4357263.457286324, "eta_seconds": 504.65459469129024, "mean_data_wait_seconds": 0.00010564923286437988}
{"step": 42180, "total_steps": 46238, "tokens": 22806085632, "loss": 2.921108285586039, "ce_loss": 2.921108285586039, "z_loss": 39.13436696023652, "muon_lr": 0.03160802237486759, "residual_lr": 0.00018964813424920547, "adamw_lr": 0.001896481342492055, "weight_decay": 0.01, "grad_norm": 0.11794934421777725, "tokens_per_second": 4357211.360532662, "eta_seconds": 503.41976151734065, "mean_data_wait_seconds": 5.778670310974121e-05}
{"step": 42190, "total_steps": 46238, "tokens": 22811492352, "loss": 2.890125765944972, "ce_loss": 2.890125765944972, "z_loss": 40.22381216106993, "muon_lr": 0.031498731396559114, "residual_lr": 0.0001889923883793547, "adamw_lr": 0.001889923883793547, "weight_decay": 0.01, "grad_norm": 0.17472045123577118, "tokens_per_second": 4357263.139157242, "eta_seconds": 502.1729269311952, "mean_data_wait_seconds": 5.888938903808594e-05}
{"step": 42200, "total_steps": 46238, "tokens": 22816899072, "loss": 2.8905138283064873, "ce_loss": 2.8905138283064873, "z_loss": 39.40223945270885, "muon_lr": 0.03138930705546476, "residual_lr": 0.00018833584233278854, "adamw_lr": 0.0018833584233278857, "weight_decay": 0.01, "grad_norm": 0.10948338359594345, "tokens_per_second": 4357202.134680877, "eta_seconds": 500.93908809669233, "mean_data_wait_seconds": 7.349252700805664e-05}
{"step": 42210, "total_steps": 46238, "tokens": 22822305792, "loss": 2.8610427722786413, "ce_loss": 2.8610427722786413, "z_loss": 40.384101607582785, "muon_lr": 0.031279751597120486, "residual_lr": 0.0001876785095827229, "adamw_lr": 0.0018767850958272291, "weight_decay": 0.01, "grad_norm": 0.14723031222820282, "tokens_per_second": 4357337.993697195, "eta_seconds": 499.68263814957714, "mean_data_wait_seconds": 7.513165473937988e-05}
{"step": 42220, "total_steps": 46238, "tokens": 22827712512, "loss": 2.9158073230223223, "ce_loss": 2.9158073230223223, "z_loss": 38.6677144946474, "muon_lr": 0.031170067269752928, "residual_lr": 0.00018702040361851755, "adamw_lr": 0.0018702040361851756, "weight_decay": 0.01, "grad_norm": 0.13844464719295502, "tokens_per_second": 4357379.646643953, "eta_seconds": 498.43704247179335, "mean_data_wait_seconds": 7.450580596923828e-05}
{"step": 42230, "total_steps": 46238, "tokens": 22833119232, "loss": 2.897742997516285, "ce_loss": 2.897742997516285, "z_loss": 37.95976375811028, "muon_lr": 0.031060256324233322, "residual_lr": 0.0001863615379453999, "adamw_lr": 0.0018636153794539994, "weight_decay": 0.01, "grad_norm": 0.13535922765731812, "tokens_per_second": 4357491.197787566, "eta_seconds": 497.1834951955808, "mean_data_wait_seconds": 0.00010499358177185059}
{"step": 42240, "total_steps": 46238, "tokens": 22838525952, "loss": 2.9317294016028894, "ce_loss": 2.9317294016028894, "z_loss": 38.93830027724757, "muon_lr": 0.030950321014031248, "residual_lr": 0.00018570192608418745, "adamw_lr": 0.0018570192608418747, "weight_decay": 0.01, "grad_norm": 0.1304534673690796, "tokens_per_second": 4357556.660243195, "eta_seconds": 495.9352574154231, "mean_data_wait_seconds": 0.00010213255882263184}
{"step": 42250, "total_steps": 46238, "tokens": 22843932672, "loss": 2.957958400249481, "ce_loss": 2.957958400249481, "z_loss": 39.279775417212285, "muon_lr": 0.03084026359516844, "residual_lr": 0.0001850415815710106, "adamw_lr": 0.0018504158157101062, "weight_decay": 0.01, "grad_norm": 0.11432596296072006, "tokens_per_second": 4357557.183473784, "eta_seconds": 494.6944292952545, "mean_data_wait_seconds": 7.76052474975586e-05}
{"step": 42260, "total_steps": 46238, "tokens": 22849339392, "loss": 2.944255402593902, "ce_loss": 2.944255402593902, "z_loss": 39.309210979577266, "muon_lr": 0.030730086326172464, "residual_lr": 0.00018438051795703476, "adamw_lr": 0.0018438051795703478, "weight_decay": 0.01, "grad_norm": 0.10809838026762009, "tokens_per_second": 4357531.846137533, "eta_seconds": 493.4565299633919, "mean_data_wait_seconds": 0.00016993284225463867}
{"step": 42270, "total_steps": 46238, "tokens": 22854746112, "loss": 2.9593290650483333, "ce_loss": 2.9593290650483333, "z_loss": 38.69586941690156, "muon_lr": 0.03061979146803037, "residual_lr": 0.0001837187488081822, "adamw_lr": 0.001837187488081822, "weight_decay": 0.01, "grad_norm": 0.11397019028663635, "tokens_per_second": 4357557.561846697, "eta_seconds": 492.2128494135215, "mean_data_wait_seconds": 8.809566497802734e-05}
{"step": 42280, "total_steps": 46238, "tokens": 22860152832, "loss": 2.9365720748901367, "ce_loss": 2.9365720748901367, "z_loss": 38.559241439356946, "muon_lr": 0.03050938128414229, "residual_lr": 0.00018305628770485372, "adamw_lr": 0.0018305628770485375, "weight_decay": 0.01, "grad_norm": 0.11904724687337875, "tokens_per_second": 4357536.753785165, "eta_seconds": 490.9744254346405, "mean_data_wait_seconds": 7.06017017364502e-05}
{"step": 42290, "total_steps": 46238, "tokens": 22865559552, "loss": 2.8995217608683035, "ce_loss": 2.8995217608683035, "z_loss": 39.565321806705356, "muon_lr": 0.03039885804027502, "residual_lr": 0.0001823931482416501, "adamw_lr": 0.001823931482416501, "weight_decay": 0.01, "grad_norm": 0.1063595861196518, "tokens_per_second": 4357549.95273708, "eta_seconds": 489.7321676506689, "mean_data_wait_seconds": 7.069110870361328e-05}
{"step": 42300, "total_steps": 46238, "tokens": 22870966272, "loss": 2.9196118730487246, "ce_loss": 2.9196118730487246, "z_loss": 40.63991144931678, "muon_lr": 0.030288224004515463, "residual_lr": 0.00018172934402709275, "adamw_lr": 0.0018172934402709278, "weight_decay": 0.01, "grad_norm": 0.1410936564207077, "tokens_per_second": 4357505.436429108, "eta_seconds": 488.496387452443, "mean_data_wait_seconds": 0.0001869797706604004}
{"step": 42310, "total_steps": 46238, "tokens": 22876372992, "loss": 2.9452110890186196, "ce_loss": 2.9452110890186196, "z_loss": 39.26906210003477, "muon_lr": 0.03017748144722417, "residual_lr": 0.000181064888683345, "adamw_lr": 0.0018106488868334502, "weight_decay": 0.01, "grad_norm": 0.13103662431240082, "tokens_per_second": 4357513.907101378, "eta_seconds": 487.254656959286, "mean_data_wait_seconds": 9.402632713317871e-05}
{"step": 42320, "total_steps": 46238, "tokens": 22881779712, "loss": 2.952796681360765, "ce_loss": 2.952796681360765, "z_loss": 40.54345760923444, "muon_lr": 0.03006663264098865, "residual_lr": 0.00018039979584593189, "adamw_lr": 0.001803997958459319, "weight_decay": 0.01, "grad_norm": 0.11706917732954025, "tokens_per_second": 4357443.2439445965, "eta_seconds": 486.021757585267, "mean_data_wait_seconds": 0.00013813376426696777}
{"step": 42330, "total_steps": 46238, "tokens": 22887186432, "loss": 2.915870301651232, "ce_loss": 2.915870301651232, "z_loss": 40.02026745767304, "muon_lr": 0.029955679860576824, "residual_lr": 0.0001797340791634609, "adamw_lr": 0.0017973407916346093, "weight_decay": 0.01, "grad_norm": 0.12482287734746933, "tokens_per_second": 4357387.288330883, "eta_seconds": 484.78718190991157, "mean_data_wait_seconds": 8.067488670349121e-05}
{"step": 42340, "total_steps": 46238, "tokens": 22892593152, "loss": 2.8981850309805437, "ce_loss": 2.8981850309805437, "z_loss": 38.86640548706055, "muon_lr": 0.029844625382890285, "residual_lr": 0.00017906775229734168, "adamw_lr": 0.001790677522973417, "weight_decay": 0.01, "grad_norm": 0.10995305329561234, "tokens_per_second": 4357430.113093521, "eta_seconds": 483.54161267411666, "mean_data_wait_seconds": 0.000125199556350708}
{"step": 42350, "total_steps": 46238, "tokens": 22897999872, "loss": 2.8421069130753027, "ce_loss": 2.8421069130753027, "z_loss": 39.36030122005578, "muon_lr": 0.029733471486917597, "residual_lr": 0.00017840082892150555, "adamw_lr": 0.0017840082892150557, "weight_decay": 0.01, "grad_norm": 0.11234342306852341, "tokens_per_second": 4357475.460334843, "eta_seconds": 482.29578872683004, "mean_data_wait_seconds": 6.80088996887207e-05}
{"step": 42360, "total_steps": 46238, "tokens": 22903406592, "loss": 2.8882514581535803, "ce_loss": 2.8882514581535803, "z_loss": 39.628291968143344, "muon_lr": 0.02962222045368751, "residual_lr": 0.00017773332272212503, "adamw_lr": 0.0017773332272212506, "weight_decay": 0.01, "grad_norm": 0.13190516829490662, "tokens_per_second": 4357605.145514755, "eta_seconds": 481.04068037407785, "mean_data_wait_seconds": 0.0001913011074066162}
{"step": 42370, "total_steps": 46238, "tokens": 22908813312, "loss": 2.9044903390335315, "ce_loss": 2.9044903390335315, "z_loss": 39.28322688016024, "muon_lr": 0.029510874566222197, "residual_lr": 0.00017706524739733316, "adamw_lr": 0.0017706524739733317, "weight_decay": 0.01, "grad_norm": 0.12318076938390732, "tokens_per_second": 4357665.000088325, "eta_seconds": 479.79333518240213, "mean_data_wait_seconds": 7.94827938079834e-05}
{"step": 42380, "total_steps": 46238, "tokens": 22914220032, "loss": 2.8676160953261634, "ce_loss": 2.8676160953261634, "z_loss": 39.313808007673785, "muon_lr": 0.029399436109490324, "residual_lr": 0.00017639661665694192, "adamw_lr": 0.0017639661665694193, "weight_decay": 0.01, "grad_norm": 0.12047666311264038, "tokens_per_second": 4357626.412650603, "eta_seconds": 478.5568349654682, "mean_data_wait_seconds": 8.174777030944824e-05}
{"step": 42390, "total_steps": 46238, "tokens": 22919626752, "loss": 2.991904762658206, "ce_loss": 2.991904762658206, "z_loss": 39.99300575256348, "muon_lr": 0.029287907370360234, "residual_lr": 0.0001757274442221614, "adamw_lr": 0.001757274442221614, "weight_decay": 0.01, "grad_norm": 0.13908395171165466, "tokens_per_second": 4357703.808142653, "eta_seconds": 477.3076086799314, "mean_data_wait_seconds": 7.832050323486328e-05}
{"step": 42400, "total_steps": 46238, "tokens": 22925033472, "loss": 2.927641687971173, "ce_loss": 2.927641687971173, "z_loss": 40.68578280824603, "muon_lr": 0.029176290637552983, "residual_lr": 0.00017505774382531788, "adamw_lr": 0.0017505774382531789, "weight_decay": 0.01, "grad_norm": 0.11947553604841232, "tokens_per_second": 4357723.511922097, "eta_seconds": 476.0647292845244, "mean_data_wait_seconds": 6.604194641113281e-05}
{"step": 42410, "total_steps": 46238, "tokens": 22930440192, "loss": 2.922448644132325, "ce_loss": 2.922448644132325, "z_loss": 39.49851232586485, "muon_lr": 0.029064588201595368, "residual_lr": 0.00017438752920957218, "adamw_lr": 0.001743875292095722, "weight_decay": 0.01, "grad_norm": 0.1331910938024521, "tokens_per_second": 4357791.2981475685, "eta_seconds": 474.81662209926515, "mean_data_wait_seconds": 9.968876838684082e-05}
{"step": 42420, "total_steps": 46238, "tokens": 22935846912, "loss": 2.8816693985100947, "ce_loss": 2.8816693985100947, "z_loss": 41.003713896780305, "muon_lr": 0.02895280235477293, "residual_lr": 0.00017371681412863755, "adamw_lr": 0.0017371681412863757, "weight_decay": 0.01, "grad_norm": 0.14048391580581665, "tokens_per_second": 4357899.050584716, "eta_seconds": 473.5642106540076, "mean_data_wait_seconds": 0.000122755765914917}
{"step": 42430, "total_steps": 46238, "tokens": 22941253632, "loss": 2.9355775117874146, "ce_loss": 2.9355775117874146, "z_loss": 39.2099521521366, "muon_lr": 0.028840935391082918, "residual_lr": 0.0001730456123464975, "adamw_lr": 0.001730456123464975, "weight_decay": 0.01, "grad_norm": 0.13842357695102692, "tokens_per_second": 4357941.897348087, "eta_seconds": 472.31889558980777, "mean_data_wait_seconds": 7.426738739013672e-05}
{"step": 42440, "total_steps": 46238, "tokens": 22946660352, "loss": 2.9602168527516453, "ce_loss": 2.9602168527516453, "z_loss": 39.96043407555783, "muon_lr": 0.02872898960618723, "residual_lr": 0.00017237393763712335, "adamw_lr": 0.0017237393763712336, "weight_decay": 0.01, "grad_norm": 0.12853416800498962, "tokens_per_second": 4357955.04522623, "eta_seconds": 471.0768153170401, "mean_data_wait_seconds": 0.00010079145431518555}
{"step": 42450, "total_steps": 46238, "tokens": 22952067072, "loss": 2.971902106747483, "ce_loss": 2.971902106747483, "z_loss": 38.70642445304177, "muon_lr": 0.028616967297365233, "residual_lr": 0.00017170180378419137, "adamw_lr": 0.001717018037841914, "weight_decay": 0.01, "grad_norm": 0.12106724828481674, "tokens_per_second": 4358089.286090562, "eta_seconds": 469.8216878059281, "mean_data_wait_seconds": 7.238984107971191e-05}
{"step": 42460, "total_steps": 46238, "tokens": 22957473792, "loss": 2.9045714501178628, "ce_loss": 2.9045714501178628, "z_loss": 39.82826784885291, "muon_lr": 0.028504870763466722, "residual_lr": 0.0001710292245808003, "adamw_lr": 0.0017102922458080033, "weight_decay": 0.01, "grad_norm": 0.12244654446840286, "tokens_per_second": 4358120.195190228, "eta_seconds": 468.5777474090211, "mean_data_wait_seconds": 0.00010246038436889648}
{"step": 42470, "total_steps": 46238, "tokens": 22962880512, "loss": 2.981994432030302, "ce_loss": 2.981994432030302, "z_loss": 39.63031725449996, "muon_lr": 0.02839270230486466, "residual_lr": 0.00017035621382918792, "adamw_lr": 0.0017035621382918795, "weight_decay": 0.01, "grad_norm": 0.1262810081243515, "tokens_per_second": 4358154.988133492, "eta_seconds": 467.33340818433845, "mean_data_wait_seconds": 0.00014472007751464844}
{"step": 42480, "total_steps": 46238, "tokens": 22968287232, "loss": 2.951869834553112, "ce_loss": 2.951869834553112, "z_loss": 39.79676448937619, "muon_lr": 0.02828046422340801, "residual_lr": 0.00016968278534044805, "adamw_lr": 0.0016968278534044805, "weight_decay": 0.01, "grad_norm": 0.11487355083227158, "tokens_per_second": 4358193.444302056, "eta_seconds": 466.08869706225346, "mean_data_wait_seconds": 0.00010034441947937012}
{"step": 42490, "total_steps": 46238, "tokens": 22973693952, "loss": 3.002022558992559, "ce_loss": 3.002022558992559, "z_loss": 39.63756769353693, "muon_lr": 0.028168158822374487, "residual_lr": 0.0001690089529342469, "adamw_lr": 0.001690089529342469, "weight_decay": 0.01, "grad_norm": 0.1106916218996048, "tokens_per_second": 4358337.037093974, "eta_seconds": 464.83279442537474, "mean_data_wait_seconds": 7.936358451843262e-05}
{"step": 42500, "total_steps": 46238, "tokens": 22979100672, "loss": 3.0591144579829592, "ce_loss": 3.0591144579829592, "z_loss": 38.49380348667954, "muon_lr": 0.028055788406423306, "residual_lr": 0.0001683347304385398, "adamw_lr": 0.0016833473043853984, "weight_decay": 0.01, "grad_norm": 0.12249629944562912, "tokens_per_second": 4358340.718875647, "eta_seconds": 463.5918562423089, "mean_data_wait_seconds": 8.925795555114746e-05}
{"step": 42510, "total_steps": 46238, "tokens": 22984507392, "loss": 2.9908699447458442, "ce_loss": 2.9908699447458442, "z_loss": 40.35152845671683, "muon_lr": 0.02794335528154786, "residual_lr": 0.00016766013168928714, "adamw_lr": 0.0016766013168928718, "weight_decay": 0.01, "grad_norm": 0.13043399155139923, "tokens_per_second": 4358397.399794737, "eta_seconds": 462.3452978599204, "mean_data_wait_seconds": 0.00010418891906738281}
{"step": 42520, "total_steps": 46238, "tokens": 22989914112, "loss": 2.9196865703120376, "ce_loss": 2.9196865703120376, "z_loss": 39.39996525735566, "muon_lr": 0.027830861755028437, "residual_lr": 0.0001669851705301706, "adamw_lr": 0.0016698517053017061, "weight_decay": 0.01, "grad_norm": 0.1294998675584793, "tokens_per_second": 4358467.598773661, "eta_seconds": 461.0973417733934, "mean_data_wait_seconds": 7.084012031555176e-05}
{"step": 42530, "total_steps": 46238, "tokens": 22995320832, "loss": 2.9449037692763587, "ce_loss": 2.9449037692763587, "z_loss": 38.71575806357644, "muon_lr": 0.02771831013538481, "residual_lr": 0.00016630986081230884, "adamw_lr": 0.0016630986081230886, "weight_decay": 0.01, "grad_norm": 0.10724949836730957, "tokens_per_second": 4357923.405522389, "eta_seconds": 459.9142567444335, "mean_data_wait_seconds": 0.00011608004570007324}
{"step": 42540, "total_steps": 46238, "tokens": 23000727552, "loss": 2.959579081246347, "ce_loss": 2.959579081246347, "z_loss": 39.460943366542004, "muon_lr": 0.027605702732328957, "residual_lr": 0.00016563421639397374, "adamw_lr": 0.0016563421639397375, "weight_decay": 0.01, "grad_norm": 0.0881686881184578, "tokens_per_second": 4357709.793301176, "eta_seconds": 458.6960763364105, "mean_data_wait_seconds": 0.00013777613639831543}
{"step": 42550, "total_steps": 46238, "tokens": 23006134272, "loss": 2.951682085340673, "ce_loss": 2.951682085340673, "z_loss": 40.21570694085324, "muon_lr": 0.027493041856717554, "residual_lr": 0.00016495825114030532, "adamw_lr": 0.0016495825114030533, "weight_decay": 0.01, "grad_norm": 0.10522226989269257, "tokens_per_second": 4357685.58088345, "eta_seconds": 457.45789295699, "mean_data_wait_seconds": 9.810924530029297e-05}
{"step": 42560, "total_steps": 46238, "tokens": 23011540992, "loss": 2.9717954993247986, "ce_loss": 2.9717954993247986, "z_loss": 40.27713712056478, "muon_lr": 0.027380329820504645, "residual_lr": 0.00016428197892302785, "adamw_lr": 0.0016428197892302788, "weight_decay": 0.01, "grad_norm": 0.10468573123216629, "tokens_per_second": 4357713.805918786, "eta_seconds": 456.2142060132003, "mean_data_wait_seconds": 0.00011077523231506348}
{"step": 42570, "total_steps": 46238, "tokens": 23016947712, "loss": 2.910078480388179, "ce_loss": 2.910078480388179, "z_loss": 39.40732788317131, "muon_lr": 0.02726756893669413, "residual_lr": 0.00016360541362016477, "adamw_lr": 0.001636054136201648, "weight_decay": 0.01, "grad_norm": 0.11899839341640472, "tokens_per_second": 4357349.433203231, "eta_seconds": 455.01152808451565, "mean_data_wait_seconds": 8.487701416015625e-05}
{"step": 42580, "total_steps": 46238, "tokens": 23022354432, "loss": 2.936287858269431, "ce_loss": 2.936287858269431, "z_loss": 39.65166297103419, "muon_lr": 0.027154761519292355, "residual_lr": 0.0001629285691157541, "adamw_lr": 0.0016292856911575412, "weight_decay": 0.01, "grad_norm": 0.11056956648826599, "tokens_per_second": 4357320.15907137, "eta_seconds": 453.7737489598166, "mean_data_wait_seconds": 9.426474571228027e-05}
{"step": 42590, "total_steps": 46238, "tokens": 23027761152, "loss": 2.9665198109366675, "ce_loss": 2.9665198109366675, "z_loss": 40.07945421970252, "muon_lr": 0.027041909883260568, "residual_lr": 0.0001622514592995634, "adamw_lr": 0.0016225145929956342, "weight_decay": 0.01, "grad_norm": 0.11265020817518234, "tokens_per_second": 4356694.106889793, "eta_seconds": 452.5979413798398, "mean_data_wait_seconds": 0.09913620352745056}
{"step": 42600, "total_steps": 46238, "tokens": 23033167872, "loss": 2.948647537014701, "ce_loss": 2.948647537014701, "z_loss": 39.950391509316184, "muon_lr": 0.026929016344467472, "residual_lr": 0.0001615740980668048, "adamw_lr": 0.0016157409806680483, "weight_decay": 0.01, "grad_norm": 0.10987798869609833, "tokens_per_second": 4356685.259686096, "eta_seconds": 451.357843587187, "mean_data_wait_seconds": 9.012222290039062e-05}
{"step": 42610, "total_steps": 46238, "tokens": 23038574592, "loss": 2.9986028237776323, "ce_loss": 2.9986028237776323, "z_loss": 39.96761238213742, "muon_lr": 0.026816083219641657, "residual_lr": 0.00016089649931784992, "adamw_lr": 0.0016089649931784993, "weight_decay": 0.01, "grad_norm": 0.08986227959394455, "tokens_per_second": 4356707.237621236, "eta_seconds": 450.1145560266557, "mean_data_wait_seconds": 0.00010469555854797363}
{"step": 42620, "total_steps": 46238, "tokens": 23043981312, "loss": 2.9507330963105867, "ce_loss": 2.9507330963105867, "z_loss": 40.298624501083836, "muon_lr": 0.026703112826324066, "residual_lr": 0.00016021867695794438, "adamw_lr": 0.001602186769579444, "weight_decay": 0.01, "grad_norm": 0.12424439936876297, "tokens_per_second": 4356671.518549229, "eta_seconds": 448.87722557775436, "mean_data_wait_seconds": 7.197260856628418e-05}
{"step": 42630, "total_steps": 46238, "tokens": 23049388032, "loss": 2.9642664139921013, "ce_loss": 2.9642664139921013, "z_loss": 39.99195824247418, "muon_lr": 0.02659010748282047, "residual_lr": 0.0001595406448969228, "adamw_lr": 0.001595406448969228, "weight_decay": 0.01, "grad_norm": 0.1414390355348587, "tokens_per_second": 4356621.759160039, "eta_seconds": 447.641317472555, "mean_data_wait_seconds": 0.00012704730033874512}
{"step": 42640, "total_steps": 46238, "tokens": 23054794752, "loss": 2.9581708492654744, "ce_loss": 2.9581708492654744, "z_loss": 39.62007990750399, "muon_lr": 0.026477069508153823, "residual_lr": 0.00015886241704892291, "adamw_lr": 0.0015886241704892294, "weight_decay": 0.01, "grad_norm": 0.12725797295570374, "tokens_per_second": 4356582.197680396, "eta_seconds": 446.4043361870875, "mean_data_wait_seconds": 0.00011673569679260254}
{"step": 42650, "total_steps": 46238, "tokens": 23060201472, "loss": 2.9790385889284536, "ce_loss": 2.9790385889284536, "z_loss": 40.00843712777802, "muon_lr": 0.02636400122201676, "residual_lr": 0.00015818400733210055, "adamw_lr": 0.0015818400733210055, "weight_decay": 0.01, "grad_norm": 0.11406750231981277, "tokens_per_second": 4356545.125393013, "eta_seconds": 445.1670780811765, "mean_data_wait_seconds": 0.00017952919006347656}
{"step": 42660, "total_steps": 46238, "tokens": 23065608192, "loss": 2.913061275626674, "ce_loss": 2.913061275626674, "z_loss": 39.68513676614472, "muon_lr": 0.026250904944723902, "residual_lr": 0.0001575054296683434, "adamw_lr": 0.001575054296683434, "weight_decay": 0.01, "grad_norm": 0.12257567793130875, "tokens_per_second": 4356472.456114085, "eta_seconds": 443.93342629442157, "mean_data_wait_seconds": 7.203221321105957e-05}
{"step": 42670, "total_steps": 46238, "tokens": 23071014912, "loss": 2.9227050145467124, "ce_loss": 2.9227050145467124, "z_loss": 40.04944133758545, "muon_lr": 0.026137782997164336, "residual_lr": 0.000156826697982986, "adamw_lr": 0.00156826697982986, "weight_decay": 0.01, "grad_norm": 0.10926556587219238, "tokens_per_second": 4356440.448842169, "eta_seconds": 442.6956012936127, "mean_data_wait_seconds": 0.00013011693954467773}
{"step": 42680, "total_steps": 46238, "tokens": 23076421632, "loss": 2.8826540798851936, "ce_loss": 2.8826540798851936, "z_loss": 40.340119968761094, "muon_lr": 0.02602463770075389, "residual_lr": 0.00015614782620452332, "adamw_lr": 0.0015614782620452334, "weight_decay": 0.01, "grad_norm": 0.10610350221395493, "tokens_per_second": 4356354.108627222, "eta_seconds": 441.4632640150622, "mean_data_wait_seconds": 0.00017276406288146973}
{"step": 42690, "total_steps": 46238, "tokens": 23081828352, "loss": 2.9438328435926726, "ce_loss": 2.9438328435926726, "z_loss": 39.6120960929177, "muon_lr": 0.025911471377387574, "residual_lr": 0.00015546882826432542, "adamw_lr": 0.0015546882826432545, "weight_decay": 0.01, "grad_norm": 0.13238705694675446, "tokens_per_second": 4356304.879078804, "eta_seconds": 440.22712763059315, "mean_data_wait_seconds": 0.00014966726303100586}
{"step": 42700, "total_steps": 46238, "tokens": 23087235072, "loss": 2.8390473651163504, "ce_loss": 2.8390473651163504, "z_loss": 40.219162651986785, "muon_lr": 0.025798286349391894, "residual_lr": 0.00015478971809635133, "adamw_lr": 0.0015478971809635136, "weight_decay": 0.01, "grad_norm": 0.12297278642654419, "tokens_per_second": 4356316.423832909, "eta_seconds": 438.98483901162786, "mean_data_wait_seconds": 7.584691047668457e-05}
{"step": 42710, "total_steps": 46238, "tokens": 23092641792, "loss": 2.838299693483295, "ce_loss": 2.838299693483295, "z_loss": 39.39589991714015, "muon_lr": 0.025685084939477194, "residual_lr": 0.00015411050963686315, "adamw_lr": 0.0015411050963686317, "weight_decay": 0.01, "grad_norm": 0.12601745128631592, "tokens_per_second": 4356187.7968843905, "eta_seconds": 437.75664248540403, "mean_data_wait_seconds": 0.00010776519775390625}
{"step": 42720, "total_steps": 46238, "tokens": 23098048512, "loss": 2.8809792959328853, "ce_loss": 2.8809792959328853, "z_loss": 40.24380308209044, "muon_lr": 0.02557186947068999, "residual_lr": 0.00015343121682413993, "adamw_lr": 0.0015343121682413995, "weight_decay": 0.01, "grad_norm": 0.10853011161088943, "tokens_per_second": 4356124.490864686, "eta_seconds": 436.5218275987668, "mean_data_wait_seconds": 0.0003815293312072754}
{"step": 42730, "total_steps": 46238, "tokens": 23103455232, "loss": 2.943259414398309, "ce_loss": 2.943259414398309, "z_loss": 39.649741953069515, "muon_lr": 0.02545864226636533, "residual_lr": 0.00015275185359819196, "adamw_lr": 0.0015275185359819197, "weight_decay": 0.01, "grad_norm": 0.11250130087137222, "tokens_per_second": 4356145.888194906, "eta_seconds": 435.2785128566295, "mean_data_wait_seconds": 0.0005146265029907227}
{"step": 42740, "total_steps": 46238, "tokens": 23108861952, "loss": 2.8619895071694343, "ce_loss": 2.8619895071694343, "z_loss": 39.96681580399022, "muon_lr": 0.02534540565007906, "residual_lr": 0.00015207243390047436, "adamw_lr": 0.0015207243390047438, "weight_decay": 0.01, "grad_norm": 0.09621980786323547, "tokens_per_second": 4356142.168123512, "eta_seconds": 434.03771296437435, "mean_data_wait_seconds": 0.00011771917343139648}
{"step": 42750, "total_steps": 46238, "tokens": 23114268672, "loss": 2.8433472550276555, "ce_loss": 2.8433472550276555, "z_loss": 40.34066425670277, "muon_lr": 0.02523216194560021, "residual_lr": 0.00015139297167360123, "adamw_lr": 0.0015139297167360123, "weight_decay": 0.01, "grad_norm": 0.10138173401355743, "tokens_per_second": 4356138.068707364, "eta_seconds": 432.7969486420455, "mean_data_wait_seconds": 0.00012308359146118164}
{"step": 42760, "total_steps": 46238, "tokens": 23119675392, "loss": 2.9231104254722595, "ce_loss": 2.9231104254722595, "z_loss": 41.03379301591353, "muon_lr": 0.02511891347684322, "residual_lr": 0.0001507134808610593, "adamw_lr": 0.0015071348086105932, "weight_decay": 0.01, "grad_norm": 0.12594017386436462, "tokens_per_second": 4356008.470526054, "eta_seconds": 431.5686153321396, "mean_data_wait_seconds": 9.968876838684082e-05}
{"step": 42770, "total_steps": 46238, "tokens": 23125082112, "loss": 2.9094408931154194, "ce_loss": 2.9094408931154194, "z_loss": 40.831130114468664, "muon_lr": 0.025005662567820348, "residual_lr": 0.00015003397540692208, "adamw_lr": 0.0015003397540692208, "weight_decay": 0.01, "grad_norm": 0.12103308737277985, "tokens_per_second": 4355955.395697038, "eta_seconds": 430.3326489182385, "mean_data_wait_seconds": 0.00015288591384887695}
{"step": 42780, "total_steps": 46238, "tokens": 23130488832, "loss": 2.9642838907964304, "ce_loss": 2.9642838907964304, "z_loss": 39.4120048754143, "muon_lr": 0.024892411542593897, "residual_lr": 0.00014935446925556337, "adamw_lr": 0.0014935446925556338, "weight_decay": 0.01, "grad_norm": 0.10508096218109131, "tokens_per_second": 4355879.639166815, "eta_seconds": 429.0988867537944, "mean_data_wait_seconds": 0.00011563301086425781}
{"step": 42790, "total_steps": 46238, "tokens": 23135895552, "loss": 2.8937973976135254, "ce_loss": 2.8937973976135254, "z_loss": 39.27199167193788, "muon_lr": 0.02477916272522857, "residual_lr": 0.0001486749763513714, "adamw_lr": 0.0014867497635137142, "weight_decay": 0.01, "grad_norm": 0.12878185510635376, "tokens_per_second": 4355810.258847326, "eta_seconds": 427.86445534778375, "mean_data_wait_seconds": 0.0001296401023864746}
{"step": 42800, "total_steps": 46238, "tokens": 23141302272, "loss": 2.9560918500929168, "ce_loss": 2.9560918500929168, "z_loss": 40.07003243764242, "muon_lr": 0.024665918439743764, "residual_lr": 0.00014799551063846255, "adamw_lr": 0.0014799551063846257, "weight_decay": 0.01, "grad_norm": 0.105243980884552, "tokens_per_second": 4355859.432928056, "eta_seconds": 426.61837293285595, "mean_data_wait_seconds": 7.62641429901123e-05}
{"step": 42810, "total_steps": 46238, "tokens": 23146708992, "loss": 2.890613868381038, "ce_loss": 2.890613868381038, "z_loss": 40.365157185178816, "muon_lr": 0.024552681010065858, "residual_lr": 0.00014731608606039512, "adamw_lr": 0.0014731608606039514, "weight_decay": 0.01, "grad_norm": 0.1108323335647583, "tokens_per_second": 4355914.742419314, "eta_seconds": 425.37171950498106, "mean_data_wait_seconds": 6.872415542602539e-05}
{"step": 42820, "total_steps": 46238, "tokens": 23152115712, "loss": 2.9600251226714165, "ce_loss": 2.9600251226714165, "z_loss": 40.97334763498017, "muon_lr": 0.024439452759980565, "residual_lr": 0.00014663671655988335, "adamw_lr": 0.0014663671655988338, "weight_decay": 0.01, "grad_norm": 0.0986773818731308, "tokens_per_second": 4355993.198435602, "eta_seconds": 424.1228440539111, "mean_data_wait_seconds": 6.470084190368652e-05}
{"step": 42830, "total_steps": 46238, "tokens": 23157522432, "loss": 2.924571360602523, "ce_loss": 2.924571360602523, "z_loss": 39.89160014643814, "muon_lr": 0.02432623601308519, "residual_lr": 0.00014595741607851113, "adamw_lr": 0.0014595741607851115, "weight_decay": 0.01, "grad_norm": 0.09505172818899155, "tokens_per_second": 4356047.160009512, "eta_seconds": 422.8763914475108, "mean_data_wait_seconds": 5.620718002319336e-05}
{"step": 42840, "total_steps": 46238, "tokens": 23162929152, "loss": 2.9873333555279356, "ce_loss": 2.9873333555279356, "z_loss": 39.997404329704516, "muon_lr": 0.02421303309274101, "residual_lr": 0.00014527819855644604, "adamw_lr": 0.0014527819855644605, "weight_decay": 0.01, "grad_norm": 0.10246816277503967, "tokens_per_second": 4356066.313021827, "eta_seconds": 421.6333388932955, "mean_data_wait_seconds": 6.4849853515625e-05}
{"step": 42850, "total_steps": 46238, "tokens": 23168335872, "loss": 2.9396676077987207, "ce_loss": 2.9396676077987207, "z_loss": 40.66696669838645, "muon_lr": 0.02409984632202554, "residual_lr": 0.00014459907793215322, "adamw_lr": 0.0014459907793215323, "weight_decay": 0.01, "grad_norm": 0.0986001268029213, "tokens_per_second": 4356185.976567833, "eta_seconds": 420.38059758018335, "mean_data_wait_seconds": 9.042024612426758e-05}
{"step": 42860, "total_steps": 46238, "tokens": 23173742592, "loss": 2.991766317324205, "ce_loss": 2.991766317324205, "z_loss": 40.354109504006125, "muon_lr": 0.02398667802368489, "residual_lr": 0.00014392006814210932, "adamw_lr": 0.0014392006814210935, "weight_decay": 0.01, "grad_norm": 0.12277045845985413, "tokens_per_second": 4356175.333128221, "eta_seconds": 419.14046253249313, "mean_data_wait_seconds": 6.335973739624023e-05}
{"step": 42870, "total_steps": 46238, "tokens": 23179149312, "loss": 2.8814912167462436, "ce_loss": 2.8814912167462436, "z_loss": 40.54587291948723, "muon_lr": 0.02387353052008609, "residual_lr": 0.00014324118312051653, "adamw_lr": 0.0014324118312051654, "weight_decay": 0.01, "grad_norm": 0.11684069782495499, "tokens_per_second": 4356256.040881225, "eta_seconds": 417.8915580067106, "mean_data_wait_seconds": 7.289648056030273e-05}
{"step": 42880, "total_steps": 46238, "tokens": 23184556032, "loss": 2.8440123200416565, "ce_loss": 2.8440123200416565, "z_loss": 40.068283370046906, "muon_lr": 0.023760406133169442, "residual_lr": 0.00014256243679901663, "adamw_lr": 0.0014256243679901665, "weight_decay": 0.01, "grad_norm": 0.13108140230178833, "tokens_per_second": 4356255.11020787, "eta_seconds": 416.6505078517752, "mean_data_wait_seconds": 8.508563041687012e-05}
{"step": 42890, "total_steps": 46238, "tokens": 23189962752, "loss": 2.904963791370392, "ce_loss": 2.904963791370392, "z_loss": 40.58681722120805, "muon_lr": 0.02364730718440084, "residual_lr": 0.000141883843106405, "adamw_lr": 0.0014188384310640505, "weight_decay": 0.01, "grad_norm": 0.11880675703287125, "tokens_per_second": 4356326.402277714, "eta_seconds": 415.40257016871647, "mean_data_wait_seconds": 5.918741226196289e-05}
{"step": 42900, "total_steps": 46238, "tokens": 23195369472, "loss": 2.8613405950141675, "ce_loss": 2.8613405950141675, "z_loss": 41.84246794382731, "muon_lr": 0.02353423599472419, "residual_lr": 0.0001412054159683451, "adamw_lr": 0.0014120541596834512, "weight_decay": 0.01, "grad_norm": 0.09584756940603256, "tokens_per_second": 4356310.928576918, "eta_seconds": 414.1629221561069, "mean_data_wait_seconds": 0.00010764598846435547}
{"step": 42910, "total_steps": 46238, "tokens": 23200776192, "loss": 2.9640111218799245, "ce_loss": 2.9640111218799245, "z_loss": 38.697882348840885, "muon_lr": 0.0234211948845137, "residual_lr": 0.00014052716930708218, "adamw_lr": 0.001405271693070822, "weight_decay": 0.01, "grad_norm": 0.10885933041572571, "tokens_per_second": 4356376.888343803, "eta_seconds": 412.91554658942044, "mean_data_wait_seconds": 0.00010195374488830566}
{"step": 42920, "total_steps": 46238, "tokens": 23206182912, "loss": 2.9058087088844995, "ce_loss": 2.9058087088844995, "z_loss": 41.33323897737445, "muon_lr": 0.023308186173526337, "residual_lr": 0.000139849117041158, "adamw_lr": 0.0013984911704115802, "weight_decay": 0.01, "grad_norm": 0.09872636944055557, "tokens_per_second": 4356318.581231084, "eta_seconds": 411.67995190406566, "mean_data_wait_seconds": 0.0001062154769897461}
{"step": 42930, "total_steps": 46238, "tokens": 23211589632, "loss": 2.8917879769296357, "ce_loss": 2.8917879769296357, "z_loss": 40.88084197766853, "muon_lr": 0.023195212180854186, "residual_lr": 0.0001391712730851251, "adamw_lr": 0.001391712730851251, "weight_decay": 0.01, "grad_norm": 0.10262062400579453, "tokens_per_second": 4356359.661401194, "eta_seconds": 410.43496014397056, "mean_data_wait_seconds": 8.806586265563965e-05}
{"step": 42940, "total_steps": 46238, "tokens": 23216996352, "loss": 2.9115928212801614, "ce_loss": 2.9115928212801614, "z_loss": 41.22917464285186, "muon_lr": 0.023082275224876856, "residual_lr": 0.00013849365134926112, "adamw_lr": 0.0013849365134926113, "weight_decay": 0.01, "grad_norm": 0.10664236545562744, "tokens_per_second": 4356380.4182954505, "eta_seconds": 409.1919008068372, "mean_data_wait_seconds": 0.00011929869651794434}
{"step": 42950, "total_steps": 46238, "tokens": 23222403072, "loss": 2.8837165958953626, "ce_loss": 2.8837165958953626, "z_loss": 41.277598323244035, "muon_lr": 0.022969377623213924, "residual_lr": 0.00013781626573928352, "adamw_lr": 0.0013781626573928355, "weight_decay": 0.01, "grad_norm": 0.09360045939683914, "tokens_per_second": 4356361.211032904, "eta_seconds": 407.9525957349676, "mean_data_wait_seconds": 0.00016701221466064453}
{"step": 42960, "total_steps": 46238, "tokens": 23227809792, "loss": 2.810534392342423, "ce_loss": 2.810534392342423, "z_loss": 40.56081156297164, "muon_lr": 0.022856521692677354, "residual_lr": 0.0001371391301560641, "adamw_lr": 0.0013713913015606413, "weight_decay": 0.01, "grad_norm": 0.10739859193563461, "tokens_per_second": 4356422.407443332, "eta_seconds": 406.70577329066026, "mean_data_wait_seconds": 9.036064147949219e-05}
{"step": 42970, "total_steps": 46238, "tokens": 23233216512, "loss": 2.9082266543850754, "ce_loss": 2.9082266543850754, "z_loss": 41.24557610714074, "muon_lr": 0.022743709749223975, "residual_lr": 0.00013646225849534382, "adamw_lr": 0.0013646225849534383, "weight_decay": 0.01, "grad_norm": 0.09023477882146835, "tokens_per_second": 4356453.714721162, "eta_seconds": 405.46176768299676, "mean_data_wait_seconds": 8.791685104370117e-05}
{"step": 42980, "total_steps": 46238, "tokens": 23238623232, "loss": 2.8874367294889507, "ce_loss": 2.8874367294889507, "z_loss": 41.16336715582645, "muon_lr": 0.02263094410790792, "residual_lr": 0.00013578566464744751, "adamw_lr": 0.0013578566464744752, "weight_decay": 0.01, "grad_norm": 0.11215700209140778, "tokens_per_second": 4356494.396338586, "eta_seconds": 404.2169101560203, "mean_data_wait_seconds": 7.623434066772461e-05}
{"step": 42990, "total_steps": 46238, "tokens": 23244029952, "loss": 2.858023034803795, "ce_loss": 2.858023034803795, "z_loss": 41.01013362768924, "muon_lr": 0.02251822708283316, "residual_lr": 0.00013510936249699895, "adamw_lr": 0.0013510936249699896, "weight_decay": 0.01, "grad_norm": 0.13367879390716553, "tokens_per_second": 4356539.525258742, "eta_seconds": 402.97166451066096, "mean_data_wait_seconds": 0.0001112818717956543}
{"step": 43000, "total_steps": 46238, "tokens": 23249436672, "loss": 2.8729556184826475, "ce_loss": 2.8729556184826475, "z_loss": 40.73386998610063, "muon_lr": 0.022405560987105982, "residual_lr": 0.00013443336592263588, "adamw_lr": 0.001344333659226359, "weight_decay": 0.01, "grad_norm": 0.09449595212936401, "tokens_per_second": 4356609.86747339, "eta_seconds": 401.72411972591897, "mean_data_wait_seconds": 0.00016430020332336426}
{"step": 43010, "total_steps": 46238, "tokens": 23254843392, "loss": 2.8836842305732495, "ce_loss": 2.8836842305732495, "z_loss": 40.38471013849432, "muon_lr": 0.02229294813278754, "residual_lr": 0.00013375768879672522, "adamw_lr": 0.0013375768879672525, "weight_decay": 0.01, "grad_norm": 0.0867091715335846, "tokens_per_second": 4356735.431338736, "eta_seconds": 400.47153918269356, "mean_data_wait_seconds": 8.413195610046387e-05}
{"step": 43020, "total_steps": 46238, "tokens": 23260250112, "loss": 2.917706973624952, "ce_loss": 2.917706973624952, "z_loss": 41.03328363823168, "muon_lr": 0.0221803908308464, "residual_lr": 0.0001330823449850784, "adamw_lr": 0.001330823449850784, "weight_decay": 0.01, "grad_norm": 0.09620331227779388, "tokens_per_second": 4356741.008760512, "eta_seconds": 399.23002549440986, "mean_data_wait_seconds": 6.300210952758789e-05}
{"step": 43030, "total_steps": 46238, "tokens": 23265656832, "loss": 2.8873977588884756, "ce_loss": 2.8873977588884756, "z_loss": 40.45760449496183, "muon_lr": 0.022067891391111122, "residual_lr": 0.00013240734834666672, "adamw_lr": 0.0013240734834666674, "weight_decay": 0.01, "grad_norm": 0.10014185309410095, "tokens_per_second": 4356727.421791074, "eta_seconds": 397.99026565843087, "mean_data_wait_seconds": 7.94529914855957e-05}
{"step": 43040, "total_steps": 46238, "tokens": 23271063552, "loss": 2.923078878359361, "ce_loss": 2.923078878359361, "z_loss": 40.810891498218886, "muon_lr": 0.021955452122222843, "residual_lr": 0.00013173271273333702, "adamw_lr": 0.0013173271273333705, "weight_decay": 0.01, "grad_norm": 0.0880386158823967, "tokens_per_second": 4356742.760139832, "eta_seconds": 396.7478639809622, "mean_data_wait_seconds": 7.259845733642578e-05}
{"step": 43050, "total_steps": 46238, "tokens": 23276470272, "loss": 2.9600769949681833, "ce_loss": 2.9600769949681833, "z_loss": 40.412704930160984, "muon_lr": 0.021843075331587927, "residual_lr": 0.00013105845198952754, "adamw_lr": 0.0013105845198952758, "weight_decay": 0.01, "grad_norm": 0.09360173344612122, "tokens_per_second": 4356711.495925429, "eta_seconds": 395.5097016663904, "mean_data_wait_seconds": 8.055567741394043e-05}
{"step": 43060, "total_steps": 46238, "tokens": 23281876992, "loss": 2.977069332744136, "ce_loss": 2.977069332744136, "z_loss": 41.02535311381022, "muon_lr": 0.02173076332533059, "residual_lr": 0.0001303845799519835, "adamw_lr": 0.0013038457995198352, "weight_decay": 0.01, "grad_norm": 0.1062752977013588, "tokens_per_second": 4356706.446809287, "eta_seconds": 394.26914917758563, "mean_data_wait_seconds": 0.0003027021884918213}
{"step": 43070, "total_steps": 46238, "tokens": 23287283712, "loss": 2.9407565918835727, "ce_loss": 2.9407565918835727, "z_loss": 40.89211001540675, "muon_lr": 0.021618518408245593, "residual_lr": 0.00012971111044947353, "adamw_lr": 0.0012971111044947354, "weight_decay": 0.01, "grad_norm": 0.09461318701505661, "tokens_per_second": 4356611.391789505, "eta_seconds": 393.0367136318437, "mean_data_wait_seconds": 7.954239845275879e-05}
{"step": 43080, "total_steps": 46238, "tokens": 23292690432, "loss": 2.8840942563432637, "ce_loss": 2.8840942563432637, "z_loss": 40.702562187657215, "muon_lr": 0.021506342883750917, "residual_lr": 0.00012903805730250548, "adamw_lr": 0.001290380573025055, "weight_decay": 0.01, "grad_norm": 0.11425542831420898, "tokens_per_second": 4356576.7304697465, "eta_seconds": 391.7987928599972, "mean_data_wait_seconds": 0.0001226663589477539}
{"step": 43090, "total_steps": 46238, "tokens": 23298097152, "loss": 2.945831520990892, "ce_loss": 2.945831520990892, "z_loss": 40.7422898610433, "muon_lr": 0.02139423905384055, "residual_lr": 0.00012836543432304326, "adamw_lr": 0.0012836543432304327, "weight_decay": 0.01, "grad_norm": 0.10976286977529526, "tokens_per_second": 4356564.614365272, "eta_seconds": 390.558831238154, "mean_data_wait_seconds": 0.00011053681373596191}
{"step": 43100, "total_steps": 46238, "tokens": 23303503872, "loss": 2.9487586689717844, "ce_loss": 2.9487586689717844, "z_loss": 41.2004280090332, "muon_lr": 0.02128220921903716, "residual_lr": 0.00012769325531422293, "adamw_lr": 0.0012769325531422294, "weight_decay": 0.01, "grad_norm": 0.09301849454641342, "tokens_per_second": 4356569.534390522, "eta_seconds": 389.31734030896865, "mean_data_wait_seconds": 0.00033536553382873535}
{"step": 43110, "total_steps": 46238, "tokens": 23308910592, "loss": 2.9441832015008638, "ce_loss": 2.9441832015008638, "z_loss": 40.88220662781686, "muon_lr": 0.021170255678344983, "residual_lr": 0.00012702153407006988, "adamw_lr": 0.0012702153407006989, "weight_decay": 0.01, "grad_norm": 0.09980151057243347, "tokens_per_second": 4356441.609967033, "eta_seconds": 388.08768609039015, "mean_data_wait_seconds": 0.00010126829147338867}
{"step": 43120, "total_steps": 46238, "tokens": 23314317312, "loss": 2.8925279252456897, "ce_loss": 2.8925279252456897, "z_loss": 41.31956401015773, "muon_lr": 0.021058380729202568, "residual_lr": 0.0001263502843752154, "adamw_lr": 0.001263502843752154, "weight_decay": 0.01, "grad_norm": 0.08743523061275482, "tokens_per_second": 4356433.329985351, "eta_seconds": 386.8473350436117, "mean_data_wait_seconds": 7.858872413635254e-05}
{"step": 43130, "total_steps": 46238, "tokens": 23319724032, "loss": 2.9300634084325847, "ce_loss": 2.9300634084325847, "z_loss": 40.27066531325831, "muon_lr": 0.02094658666743566, "residual_lr": 0.00012567952000461393, "adamw_lr": 0.0012567952000461394, "weight_decay": 0.01, "grad_norm": 0.10681764781475067, "tokens_per_second": 4356471.393885671, "eta_seconds": 385.60287721794816, "mean_data_wait_seconds": 0.00010865926742553711}
{"step": 43140, "total_steps": 46238, "tokens": 23325130752, "loss": 2.9304337736332053, "ce_loss": 2.9304337736332053, "z_loss": 40.62824419772986, "muon_lr": 0.0208348757872101, "residual_lr": 0.00012500925472326056, "adamw_lr": 0.0012500925472326058, "weight_decay": 0.01, "grad_norm": 0.08751016110181808, "tokens_per_second": 4356429.981197104, "eta_seconds": 384.3654531869406, "mean_data_wait_seconds": 0.00010380148887634277}
{"step": 43150, "total_steps": 46238, "tokens": 23330537472, "loss": 2.9885202357263276, "ce_loss": 2.9885202357263276, "z_loss": 40.68685531616211, "muon_lr": 0.02072325038098471, "residual_lr": 0.00012433950228590824, "adamw_lr": 0.0012433950228590825, "weight_decay": 0.01, "grad_norm": 0.10102641582489014, "tokens_per_second": 4356394.182323483, "eta_seconds": 383.1275119162449, "mean_data_wait_seconds": 6.788969039916992e-05}
{"step": 43160, "total_steps": 46238, "tokens": 23335944192, "loss": 2.8283469478289285, "ce_loss": 2.8283469478289285, "z_loss": 40.509265234976105, "muon_lr": 0.0206117127394643, "residual_lr": 0.00012367027643678578, "adamw_lr": 0.0012367027643678578, "weight_decay": 0.01, "grad_norm": 0.09141969680786133, "tokens_per_second": 4356493.033551558, "eta_seconds": 381.877746891228, "mean_data_wait_seconds": 9.918212890625e-05}
{"step": 43170, "total_steps": 46238, "tokens": 23341350912, "loss": 2.9924011591709023, "ce_loss": 2.9924011591709023, "z_loss": 40.68004142876828, "muon_lr": 0.02050026515155259, "residual_lr": 0.0001230015909093155, "adamw_lr": 0.0012300159090931552, "weight_decay": 0.01, "grad_norm": 0.12638190388679504, "tokens_per_second": 4356628.818239312, "eta_seconds": 380.6248117943088, "mean_data_wait_seconds": 7.614493370056152e-05}
{"step": 43180, "total_steps": 46238, "tokens": 23346757632, "loss": 2.9521044059233232, "ce_loss": 2.9521044059233232, "z_loss": 40.57895330949263, "muon_lr": 0.02038890990430532, "residual_lr": 0.00012233345942583187, "adamw_lr": 0.001223334594258319, "weight_decay": 0.01, "grad_norm": 0.10760513693094254, "tokens_per_second": 4356525.683027615, "eta_seconds": 379.392760253704, "mean_data_wait_seconds": 8.171796798706055e-05}
{"step": 43190, "total_steps": 46238, "tokens": 23352164352, "loss": 2.944310921611208, "ce_loss": 2.944310921611208, "z_loss": 40.64906438191732, "muon_lr": 0.020277649282883226, "residual_lr": 0.00012166589569729934, "adamw_lr": 0.0012166589569729935, "weight_decay": 0.01, "grad_norm": 0.08848864585161209, "tokens_per_second": 4356502.005395195, "eta_seconds": 378.15375316246536, "mean_data_wait_seconds": 7.361173629760742e-05}
{"step": 43200, "total_steps": 46238, "tokens": 23357571072, "loss": 2.960575419845003, "ce_loss": 2.960575419845003, "z_loss": 40.69127030806108, "muon_lr": 0.020166485570505252, "residual_lr": 0.00012099891342303149, "adamw_lr": 0.001209989134230315, "weight_decay": 0.01, "grad_norm": 0.09952151775360107, "tokens_per_second": 4356612.25097113, "eta_seconds": 376.9031461622452, "mean_data_wait_seconds": 6.732344627380371e-05}
{"step": 43210, "total_steps": 46238, "tokens": 23362977792, "loss": 2.9248403455271865, "ce_loss": 2.9248403455271865, "z_loss": 40.88520720510772, "muon_lr": 0.020055421048401575, "residual_lr": 0.00012033252629040943, "adamw_lr": 0.0012033252629040944, "weight_decay": 0.01, "grad_norm": 0.113357774913311, "tokens_per_second": 4356628.1246640235, "eta_seconds": 375.66073972086224, "mean_data_wait_seconds": 8.562207221984863e-05}
{"step": 43220, "total_steps": 46238, "tokens": 23368384512, "loss": 2.925139656572631, "ce_loss": 2.925139656572631, "z_loss": 40.64952937039462, "muon_lr": 0.01994445799576691, "residual_lr": 0.00011966674797460143, "adamw_lr": 0.0011966674797460144, "weight_decay": 0.01, "grad_norm": 0.08321890980005264, "tokens_per_second": 4356655.088993412, "eta_seconds": 374.41738918489506, "mean_data_wait_seconds": 5.53131103515625e-05}
{"step": 43230, "total_steps": 46238, "tokens": 23373791232, "loss": 2.9261369109153748, "ce_loss": 2.9261369109153748, "z_loss": 40.30924320220947, "muon_lr": 0.019833598689713627, "residual_lr": 0.00011900159213828175, "adamw_lr": 0.0011900159213828177, "weight_decay": 0.01, "grad_norm": 0.08171024918556213, "tokens_per_second": 4356677.796723954, "eta_seconds": 373.174418641318, "mean_data_wait_seconds": 0.00011742115020751953}
{"step": 43240, "total_steps": 46238, "tokens": 23379197952, "loss": 2.92531577204213, "ce_loss": 2.92531577204213, "z_loss": 40.9525565812082, "muon_lr": 0.01972284540522513, "residual_lr": 0.00011833707243135075, "adamw_lr": 0.0011833707243135078, "weight_decay": 0.01, "grad_norm": 0.12613923847675323, "tokens_per_second": 4356728.115448648, "eta_seconds": 371.9291039195671, "mean_data_wait_seconds": 0.0001074075698852539}
{"step": 43250, "total_steps": 46238, "tokens": 23384604672, "loss": 2.9027016686670706, "ce_loss": 2.9027016686670706, "z_loss": 41.5090338678071, "muon_lr": 0.01961220041510907, "residual_lr": 0.00011767320249065438, "adamw_lr": 0.001176732024906544, "weight_decay": 0.01, "grad_norm": 0.11500392109155655, "tokens_per_second": 4356797.849399732, "eta_seconds": 370.68216608271337, "mean_data_wait_seconds": 5.626678466796875e-05}
{"step": 43260, "total_steps": 46238, "tokens": 23390011392, "loss": 2.8765889749382483, "ce_loss": 2.8765889749382483, "z_loss": 41.13176987387917, "muon_lr": 0.019501665989950775, "residual_lr": 0.00011700999593970464, "adamw_lr": 0.0011700999593970466, "weight_decay": 0.01, "grad_norm": 0.09454596042633057, "tokens_per_second": 4356906.64756102, "eta_seconds": 369.43195578932983, "mean_data_wait_seconds": 6.723403930664062e-05}
{"step": 43270, "total_steps": 46238, "tokens": 23395418112, "loss": 2.9191230553569216, "ce_loss": 2.9191230553569216, "z_loss": 41.7080552072236, "muon_lr": 0.019391244398066627, "residual_lr": 0.00011634746638839975, "adamw_lr": 0.0011634746638839975, "weight_decay": 0.01, "grad_norm": 0.09341598302125931, "tokens_per_second": 4357058.526446993, "eta_seconds": 368.1781675097533, "mean_data_wait_seconds": 9.140372276306152e-05}
{"step": 43280, "total_steps": 46238, "tokens": 23400824832, "loss": 2.8651676701776907, "ce_loss": 2.8651676701776907, "z_loss": 40.75361621741092, "muon_lr": 0.019280937905457516, "residual_lr": 0.00011568562743274508, "adamw_lr": 0.0011568562743274509, "weight_decay": 0.01, "grad_norm": 0.099680595099926, "tokens_per_second": 4357130.983809689, "eta_seconds": 366.9311549138021, "mean_data_wait_seconds": 6.887316703796387e-05}
{"step": 43290, "total_steps": 46238, "tokens": 23406231552, "loss": 2.863587343331539, "ce_loss": 2.863587343331539, "z_loss": 41.65078203605883, "muon_lr": 0.01917074877576233, "residual_lr": 0.00011502449265457396, "adamw_lr": 0.0011502449265457398, "weight_decay": 0.01, "grad_norm": 0.08606056869029999, "tokens_per_second": 4357221.869425758, "eta_seconds": 365.68263718229946, "mean_data_wait_seconds": 7.191300392150879e-05}
{"step": 43300, "total_steps": 46238, "tokens": 23411638272, "loss": 2.850184207612818, "ce_loss": 2.850184207612818, "z_loss": 40.04904998432506, "muon_lr": 0.019060679270211524, "residual_lr": 0.00011436407562126913, "adamw_lr": 0.0011436407562126914, "weight_decay": 0.01, "grad_norm": 0.11309091746807098, "tokens_per_second": 4357342.819830819, "eta_seconds": 364.43165701193436, "mean_data_wait_seconds": 7.686018943786621e-05}
{"step": 43310, "total_steps": 46238, "tokens": 23417044992, "loss": 2.923730989297231, "ce_loss": 2.923730989297231, "z_loss": 40.46438520604914, "muon_lr": 0.018950731647580685, "residual_lr": 0.0001137043898854841, "adamw_lr": 0.0011370438988548412, "weight_decay": 0.01, "grad_norm": 0.08503202348947525, "tokens_per_second": 4357473.208267169, "eta_seconds": 363.1799596604587, "mean_data_wait_seconds": 7.12275505065918e-05}
{"step": 43320, "total_steps": 46238, "tokens": 23422451712, "loss": 2.8637268489057366, "ce_loss": 2.8637268489057366, "z_loss": 41.63083417487867, "muon_lr": 0.018840908164144226, "residual_lr": 0.00011304544898486533, "adamw_lr": 0.0011304544898486535, "weight_decay": 0.01, "grad_norm": 0.10749123990535736, "tokens_per_second": 4357523.258326119, "eta_seconds": 361.9350099822155, "mean_data_wait_seconds": 6.628036499023438e-05}
{"step": 43330, "total_steps": 46238, "tokens": 23427858432, "loss": 2.8588864875562265, "ce_loss": 2.8588864875562265, "z_loss": 40.77991898854574, "muon_lr": 0.018731211073629028, "residual_lr": 0.00011238726644177414, "adamw_lr": 0.0011238726644177415, "weight_decay": 0.01, "grad_norm": 0.09305170178413391, "tokens_per_second": 4357598.409316316, "eta_seconds": 360.68801123107545, "mean_data_wait_seconds": 6.86943531036377e-05}
{"step": 43340, "total_steps": 46238, "tokens": 23433265152, "loss": 2.89019286813158, "ce_loss": 2.89019286813158, "z_loss": 41.29658473621715, "muon_lr": 0.018621642627168224, "residual_lr": 0.00011172985576300932, "adamw_lr": 0.0011172985576300934, "weight_decay": 0.01, "grad_norm": 0.08162171393632889, "tokens_per_second": 4357715.574504764, "eta_seconds": 359.43758999874757, "mean_data_wait_seconds": 8.0108642578125e-05}
{"step": 43350, "total_steps": 46238, "tokens": 23438671872, "loss": 2.9465680899042073, "ce_loss": 2.9465680899042073, "z_loss": 41.18736041675914, "muon_lr": 0.018512205073255007, "residual_lr": 0.00011107323043953002, "adamw_lr": 0.0011107323043953003, "weight_decay": 0.01, "grad_norm": 0.08344994485378265, "tokens_per_second": 4357739.504573594, "eta_seconds": 358.1948995257202, "mean_data_wait_seconds": 7.31348991394043e-05}
{"step": 43360, "total_steps": 46238, "tokens": 23444078592, "loss": 2.9219683643543357, "ce_loss": 2.9219683643543357, "z_loss": 41.218528660860926, "muon_lr": 0.018402900657696463, "residual_lr": 0.00011041740394617878, "adamw_lr": 0.0011041740394617879, "weight_decay": 0.01, "grad_norm": 0.0907207503914833, "tokens_per_second": 4357825.177398977, "eta_seconds": 356.9471653124983, "mean_data_wait_seconds": 0.00013202428817749023}
{"step": 43370, "total_steps": 46238, "tokens": 23449485312, "loss": 2.9774089383356497, "ce_loss": 2.9774089383356497, "z_loss": 41.3111299456972, "muon_lr": 0.01829373162356752, "residual_lr": 0.00010976238974140512, "adamw_lr": 0.0010976238974140513, "weight_decay": 0.01, "grad_norm": 0.08964315056800842, "tokens_per_second": 4357894.186254293, "eta_seconds": 355.7008403024928, "mean_data_wait_seconds": 0.0001258254051208496}
{"step": 43380, "total_steps": 46238, "tokens": 23454892032, "loss": 2.943392065438357, "ce_loss": 2.943392065438357, "z_loss": 40.99077837394945, "muon_lr": 0.01818470021116487, "residual_lr": 0.0001091082012669892, "adamw_lr": 0.001091082012669892, "weight_decay": 0.01, "grad_norm": 0.1005631610751152, "tokens_per_second": 4357930.631866387, "eta_seconds": 354.4572033122165, "mean_data_wait_seconds": 9.816884994506836e-05}
{"step": 43390, "total_steps": 46238, "tokens": 23460298752, "loss": 2.8910391493277117, "ce_loss": 2.8910391493277117, "z_loss": 42.510288296323836, "muon_lr": 0.01807580865796103, "residual_lr": 0.00010845485194776617, "adamw_lr": 0.001084548519477662, "weight_decay": 0.01, "grad_norm": 0.09578970819711685, "tokens_per_second": 4357959.357017623, "eta_seconds": 353.2142128680654, "mean_data_wait_seconds": 6.505846977233887e-05}
{"step": 43400, "total_steps": 46238, "tokens": 23465705472, "loss": 2.9423417087757224, "ce_loss": 2.9423417087757224, "z_loss": 41.36211551319469, "muon_lr": 0.017967059198558427, "residual_lr": 0.00010780235519135056, "adamw_lr": 0.0010780235519135056, "weight_decay": 0.01, "grad_norm": 0.092515729367733, "tokens_per_second": 4358091.468201773, "eta_seconds": 351.9628890746777, "mean_data_wait_seconds": 5.91278076171875e-05}
{"step": 43410, "total_steps": 46238, "tokens": 23471112192, "loss": 2.8910715182622275, "ce_loss": 2.8910715182622275, "z_loss": 41.492828051249184, "muon_lr": 0.01785845406464352, "residual_lr": 0.00010715072438786109, "adamw_lr": 0.001071507243878611, "weight_decay": 0.01, "grad_norm": 0.09027975797653198, "tokens_per_second": 4358158.638982874, "eta_seconds": 350.71686705666207, "mean_data_wait_seconds": 5.888938903808594e-05}
{"step": 43420, "total_steps": 46238, "tokens": 23476518912, "loss": 2.8492440725817825, "ce_loss": 2.8492440725817825, "z_loss": 41.5930818788933, "muon_lr": 0.017749995484941, "residual_lr": 0.00010649997290964599, "adamw_lr": 0.0010649997290964601, "weight_decay": 0.01, "grad_norm": 0.11656133830547333, "tokens_per_second": 4358299.357551514, "eta_seconds": 349.4649860067575, "mean_data_wait_seconds": 6.440281867980957e-05}
{"step": 43430, "total_steps": 46238, "tokens": 23481925632, "loss": 2.917062380097129, "ce_loss": 2.917062380097129, "z_loss": 41.93540301467433, "muon_lr": 0.017641685685168104, "residual_lr": 0.00010585011411100862, "adamw_lr": 0.0010585011411100862, "weight_decay": 0.01, "grad_norm": 0.11008638143539429, "tokens_per_second": 4358330.240008745, "eta_seconds": 348.2219612612363, "mean_data_wait_seconds": 6.1005353927612305e-05}
{"step": 43440, "total_steps": 46238, "tokens": 23487332352, "loss": 2.7689316074053445, "ce_loss": 2.7689316074053445, "z_loss": 41.546136104699336, "muon_lr": 0.017533526887988867, "residual_lr": 0.00010520116132793317, "adamw_lr": 0.0010520116132793319, "weight_decay": 0.01, "grad_norm": 0.10088436305522919, "tokens_per_second": 4358372.45541994, "eta_seconds": 346.9780518916872, "mean_data_wait_seconds": 5.799531936645508e-05}
{"step": 43450, "total_steps": 46238, "tokens": 23492739072, "loss": 2.8789861021619854, "ce_loss": 2.8789861021619854, "z_loss": 41.702436822833434, "muon_lr": 0.017425521312968564, "residual_lr": 0.00010455312787781135, "adamw_lr": 0.0010455312787781138, "weight_decay": 0.01, "grad_norm": 0.11231325566768646, "tokens_per_second": 4357788.6297925925, "eta_seconds": 345.78383487859026, "mean_data_wait_seconds": 0.09964057803153992}
{"step": 43460, "total_steps": 46238, "tokens": 23498145792, "loss": 2.867392281691233, "ce_loss": 2.867392281691233, "z_loss": 42.23254802010276, "muon_lr": 0.01731767117652813, "residual_lr": 0.00010390602705916875, "adamw_lr": 0.0010390602705916876, "weight_decay": 0.01, "grad_norm": 0.09884797781705856, "tokens_per_second": 4357849.005126777, "eta_seconds": 344.5383587714096, "mean_data_wait_seconds": 6.008148193359375e-05}
{"step": 43470, "total_steps": 46238, "tokens": 23503552512, "loss": 2.8546313896323694, "ce_loss": 2.8546313896323694, "z_loss": 41.897132844635934, "muon_lr": 0.01720997869189871, "residual_lr": 0.00010325987215139224, "adamw_lr": 0.0010325987215139226, "weight_decay": 0.01, "grad_norm": 0.09551650285720825, "tokens_per_second": 4358000.538744114, "eta_seconds": 343.2857363599885, "mean_data_wait_seconds": 5.397200584411621e-05}
{"step": 43480, "total_steps": 46238, "tokens": 23508959232, "loss": 2.88389185703162, "ce_loss": 2.88389185703162, "z_loss": 42.58397342219497, "muon_lr": 0.0171024460690762, "residual_lr": 0.00010261467641445719, "adamw_lr": 0.001026146764144572, "weight_decay": 0.01, "grad_norm": 0.10728240013122559, "tokens_per_second": 4357546.363934035, "eta_seconds": 342.0807444155895, "mean_data_wait_seconds": 0.00017204880714416504}
{"step": 43490, "total_steps": 46238, "tokens": 23514365952, "loss": 2.818380276362101, "ce_loss": 2.818380276362101, "z_loss": 41.98907964879816, "muon_lr": 0.016995075514775932, "residual_lr": 0.00010197045308865558, "adamw_lr": 0.0010197045308865558, "weight_decay": 0.01, "grad_norm": 0.07871540635824203, "tokens_per_second": 4356944.953166228, "eta_seconds": 340.8870205993018, "mean_data_wait_seconds": 0.00016501545906066895}
{"step": 43500, "total_steps": 46238, "tokens": 23519772672, "loss": 2.880612763491544, "ce_loss": 2.880612763491544, "z_loss": 42.9128833250566, "muon_lr": 0.01688786923238735, "residual_lr": 0.00010132721539432408, "adamw_lr": 0.001013272153943241, "weight_decay": 0.01, "grad_norm": 0.10524638742208481, "tokens_per_second": 4356904.070771003, "eta_seconds": 339.64926469866697, "mean_data_wait_seconds": 0.00015911459922790527}
{"step": 43510, "total_steps": 46238, "tokens": 23525179392, "loss": 2.8639211040554624, "ce_loss": 2.8639211040554624, "z_loss": 41.7613417596528, "muon_lr": 0.01678082942192884, "residual_lr": 0.00010068497653157303, "adamw_lr": 0.0010068497653157303, "weight_decay": 0.01, "grad_norm": 0.09092912822961807, "tokens_per_second": 4356934.005084558, "eta_seconds": 338.4059850985475, "mean_data_wait_seconds": 0.00013339519500732422}
{"step": 43520, "total_steps": 46238, "tokens": 23530586112, "loss": 2.9193853609489673, "ce_loss": 2.9193853609489673, "z_loss": 41.728013443224356, "muon_lr": 0.01667395828000254, "residual_lr": 0.00010004374968001523, "adamw_lr": 0.0010004374968001525, "weight_decay": 0.01, "grad_norm": 0.08735767006874084, "tokens_per_second": 4356828.948877389, "eta_seconds": 337.17316911845126, "mean_data_wait_seconds": 0.00015622377395629883}
{"step": 43530, "total_steps": 46238, "tokens": 23535992832, "loss": 2.876736509077477, "ce_loss": 2.876736509077477, "z_loss": 41.6386556336374, "muon_lr": 0.01656725799974929, "residual_lr": 9.940354799849571e-05, "adamw_lr": 0.0009940354799849573, "weight_decay": 0.01, "grad_norm": 0.09559149295091629, "tokens_per_second": 4356763.972316013, "eta_seconds": 335.93720323159147, "mean_data_wait_seconds": 0.00011134147644042969}
{"step": 43540, "total_steps": 46238, "tokens": 23541399552, "loss": 2.975189945914529, "ce_loss": 2.975189945914529, "z_loss": 42.05358895388517, "muon_lr": 0.01646073077080361, "residual_lr": 9.876438462482166e-05, "adamw_lr": 0.0009876438462482168, "weight_decay": 0.01, "grad_norm": 0.09701483696699142, "tokens_per_second": 4356701.94079697, "eta_seconds": 334.7009742266769, "mean_data_wait_seconds": 7.942318916320801e-05}
{"step": 43550, "total_steps": 46238, "tokens": 23546806272, "loss": 2.9263392289479575, "ce_loss": 2.9263392289479575, "z_loss": 42.17526707504735, "muon_lr": 0.016354378779248786, "residual_lr": 9.81262726754927e-05, "adamw_lr": 0.0009812627267549272, "weight_decay": 0.01, "grad_norm": 0.0955820232629776, "tokens_per_second": 4356611.02448817, "eta_seconds": 333.46692092408654, "mean_data_wait_seconds": 0.0001239478588104248}
{"step": 43560, "total_steps": 46238, "tokens": 23552212992, "loss": 2.889765258991357, "ce_loss": 2.889765258991357, "z_loss": 42.15691023161917, "muon_lr": 0.016248204207571976, "residual_lr": 9.748922524543185e-05, "adamw_lr": 0.0009748922524543186, "weight_decay": 0.01, "grad_norm": 0.08163529634475708, "tokens_per_second": 4356463.18465088, "eta_seconds": 332.23715721954176, "mean_data_wait_seconds": 0.00013595819473266602}
{"step": 43570, "total_steps": 46238, "tokens": 23557619712, "loss": 2.9404430497776377, "ce_loss": 2.9404430497776377, "z_loss": 41.42701570915453, "muon_lr": 0.01614220923461945, "residual_lr": 9.68532554077167e-05, "adamw_lr": 0.0009685325540771671, "weight_decay": 0.01, "grad_norm": 0.0857987031340599, "tokens_per_second": 4356359.483150475, "eta_seconds": 331.00395630279354, "mean_data_wait_seconds": 7.843971252441406e-05}
{"step": 43580, "total_steps": 46238, "tokens": 23563026432, "loss": 2.8815341031912602, "ce_loss": 2.8815341031912602, "z_loss": 41.62083013129957, "muon_lr": 0.016036396035551882, "residual_lr": 9.621837621331128e-05, "adamw_lr": 0.0009621837621331128, "weight_decay": 0.01, "grad_norm": 0.09332066774368286, "tokens_per_second": 4356379.793353969, "eta_seconds": 329.76130919338203, "mean_data_wait_seconds": 5.581974983215332e-05}
{"step": 43590, "total_steps": 46238, "tokens": 23568433152, "loss": 2.928259959726623, "ce_loss": 2.928259959726623, "z_loss": 41.229866750312574, "muon_lr": 0.01593076678179967, "residual_lr": 9.5584600690798e-05, "adamw_lr": 0.0009558460069079802, "weight_decay": 0.01, "grad_norm": 0.09317793697118759, "tokens_per_second": 4356412.606057528, "eta_seconds": 328.51773085267325, "mean_data_wait_seconds": 9.340047836303711e-05}
{"step": 43600, "total_steps": 46238, "tokens": 23573839872, "loss": 2.8465383341818145, "ce_loss": 2.8465383341818145, "z_loss": 41.47209473812219, "muon_lr": 0.01582532364101843, "residual_lr": 9.495194184611059e-05, "adamw_lr": 0.000949519418461106, "weight_decay": 0.01, "grad_norm": 0.08294977247714996, "tokens_per_second": 4356362.585247808, "eta_seconds": 327.28039415913247, "mean_data_wait_seconds": 7.304549217224121e-05}
{"step": 43610, "total_steps": 46238, "tokens": 23579246592, "loss": 2.8787492242726413, "ce_loss": 2.8787492242726413, "z_loss": 41.714372316996254, "muon_lr": 0.015720068777044478, "residual_lr": 9.432041266226686e-05, "adamw_lr": 0.0009432041266226686, "weight_decay": 0.01, "grad_norm": 0.0893341526389122, "tokens_per_second": 4356379.39177661, "eta_seconds": 326.0380275145773, "mean_data_wait_seconds": 6.759166717529297e-05}
{"step": 43620, "total_steps": 46238, "tokens": 23584653312, "loss": 2.9287916999874692, "ce_loss": 2.9287916999874692, "z_loss": 42.01570768067331, "muon_lr": 0.015615004349850437, "residual_lr": 9.369002609910261e-05, "adamw_lr": 0.0009369002609910262, "weight_decay": 0.01, "grad_norm": 0.08185373991727829, "tokens_per_second": 4356398.080851697, "eta_seconds": 324.79553010072317, "mean_data_wait_seconds": 6.192922592163086e-05}
{"step": 43630, "total_steps": 46238, "tokens": 23590060032, "loss": 2.930824814420758, "ce_loss": 2.930824814420758, "z_loss": 41.75959575537479, "muon_lr": 0.015510132515500902, "residual_lr": 9.30607950930054e-05, "adamw_lr": 0.0009306079509300542, "weight_decay": 0.01, "grad_norm": 0.0835556760430336, "tokens_per_second": 4356505.772683337, "eta_seconds": 323.546433207597, "mean_data_wait_seconds": 0.00010702013969421387}
{"step": 43640, "total_steps": 46238, "tokens": 23595466752, "loss": 2.9438784935257654, "ce_loss": 2.9438784935257654, "z_loss": 42.10836448091449, "muon_lr": 0.015405455426108211, "residual_lr": 9.243273255664926e-05, "adamw_lr": 0.0009243273255664927, "weight_decay": 0.01, "grad_norm": 0.10845933854579926, "tokens_per_second": 4356588.029934855, "eta_seconds": 322.29927970054035, "mean_data_wait_seconds": 5.951523780822754e-05}
{"step": 43650, "total_steps": 46238, "tokens": 23600873472, "loss": 2.896605356173082, "ce_loss": 2.896605356173082, "z_loss": 41.717835802020446, "muon_lr": 0.015300975229788256, "residual_lr": 9.180585137872953e-05, "adamw_lr": 0.0009180585137872953, "weight_decay": 0.01, "grad_norm": 0.0784652829170227, "tokens_per_second": 4356651.8667471325, "eta_seconds": 321.05353073445013, "mean_data_wait_seconds": 6.133317947387695e-05}
{"step": 43660, "total_steps": 46238, "tokens": 23606280192, "loss": 2.930001119772593, "ce_loss": 2.930001119772593, "z_loss": 41.65930424314557, "muon_lr": 0.01519669407061644, "residual_lr": 9.118016442369863e-05, "adamw_lr": 0.0009118016442369864, "weight_decay": 0.01, "grad_norm": 0.08790801465511322, "tokens_per_second": 4356715.354699479, "eta_seconds": 319.80784388336724, "mean_data_wait_seconds": 6.645917892456055e-05}
{"step": 43670, "total_steps": 46238, "tokens": 23611686912, "loss": 2.876259308872801, "ce_loss": 2.876259308872801, "z_loss": 41.605753696326055, "muon_lr": 0.015092614088583634, "residual_lr": 9.05556845315018e-05, "adamw_lr": 0.000905556845315018, "weight_decay": 0.01, "grad_norm": 0.09975333511829376, "tokens_per_second": 4356811.146196914, "eta_seconds": 318.5598313600328, "mean_data_wait_seconds": 8.511543273925781e-05}
{"step": 43680, "total_steps": 46238, "tokens": 23617093632, "loss": 2.9108428756395974, "ce_loss": 2.9108428756395974, "z_loss": 42.26343062429717, "muon_lr": 0.014988737419552287, "residual_lr": 8.993242451731371e-05, "adamw_lr": 0.0008993242451731372, "weight_decay": 0.01, "grad_norm": 0.093689925968647, "tokens_per_second": 4356909.116388581, "eta_seconds": 317.311715041223, "mean_data_wait_seconds": 8.788704872131348e-05}
{"step": 43690, "total_steps": 46238, "tokens": 23622500352, "loss": 2.8115360791032966, "ce_loss": 2.8115360791032966, "z_loss": 41.57961804939039, "muon_lr": 0.014885066195212588, "residual_lr": 8.931039717127552e-05, "adamw_lr": 0.0008931039717127553, "weight_decay": 0.01, "grad_norm": 0.09076696634292603, "tokens_per_second": 4356930.137221124, "eta_seconds": 316.0692369692935, "mean_data_wait_seconds": 0.00010070204734802246}
{"step": 43700, "total_steps": 46238, "tokens": 23627907072, "loss": 2.9172003251133543, "ce_loss": 2.9172003251133543, "z_loss": 41.70981768405799, "muon_lr": 0.01478160254303873, "residual_lr": 8.868961525823237e-05, "adamw_lr": 0.0008868961525823238, "weight_decay": 0.01, "grad_norm": 0.08579033613204956, "tokens_per_second": 4357023.9973355355, "eta_seconds": 314.82150771692574, "mean_data_wait_seconds": 8.717179298400879e-05}
{"step": 43710, "total_steps": 46238, "tokens": 23633313792, "loss": 2.954594550710736, "ce_loss": 2.954594550710736, "z_loss": 42.118818398677945, "muon_lr": 0.014678348586245227, "residual_lr": 8.807009151747135e-05, "adamw_lr": 0.0008807009151747137, "weight_decay": 0.01, "grad_norm": 0.15207481384277344, "tokens_per_second": 4357071.343462931, "eta_seconds": 313.5771797838187, "mean_data_wait_seconds": 6.517767906188965e-05}
{"step": 43720, "total_steps": 46238, "tokens": 23638720512, "loss": 2.887294790961526, "ce_loss": 2.887294790961526, "z_loss": 41.16461320356889, "muon_lr": 0.01457530644374338, "residual_lr": 8.745183866246027e-05, "adamw_lr": 0.0008745183866246028, "weight_decay": 0.01, "grad_norm": 0.09299255162477493, "tokens_per_second": 4357093.073523373, "eta_seconds": 312.33471514977947, "mean_data_wait_seconds": 5.939602851867676e-05}
{"step": 43730, "total_steps": 46238, "tokens": 23644127232, "loss": 2.93976527633089, "ce_loss": 2.93976527633089, "z_loss": 42.11109256744385, "muon_lr": 0.014472478230097757, "residual_lr": 8.683486938058654e-05, "adamw_lr": 0.0008683486938058654, "weight_decay": 0.01, "grad_norm": 0.07083312422037125, "tokens_per_second": 4357199.963486517, "eta_seconds": 311.0861827225833, "mean_data_wait_seconds": 0.00011748075485229492}
{"step": 43740, "total_steps": 46238, "tokens": 23649533952, "loss": 2.9195863712917673, "ce_loss": 2.9195863712917673, "z_loss": 42.786812262101606, "muon_lr": 0.014369866055482834, "residual_lr": 8.621919633289698e-05, "adamw_lr": 0.00086219196332897, "weight_decay": 0.01, "grad_norm": 0.07169721275568008, "tokens_per_second": 4357297.325517001, "eta_seconds": 309.83838906146104, "mean_data_wait_seconds": 0.00011238455772399902}
{"step": 43750, "total_steps": 46238, "tokens": 23654940672, "loss": 2.917693616765918, "ce_loss": 2.917693616765918, "z_loss": 41.73378308614095, "muon_lr": 0.01426747202563964, "residual_lr": 8.560483215383782e-05, "adamw_lr": 0.0008560483215383783, "weight_decay": 0.01, "grad_norm": 0.09039810299873352, "tokens_per_second": 4357324.594185858, "eta_seconds": 308.5956152530428, "mean_data_wait_seconds": 0.0001201629638671875}
{"step": 43760, "total_steps": 46238, "tokens": 23660347392, "loss": 2.9440313920830237, "ce_loss": 2.9440313920830237, "z_loss": 41.714427427812055, "muon_lr": 0.014165298241832606, "residual_lr": 8.499178945099562e-05, "adamw_lr": 0.0008499178945099563, "weight_decay": 0.01, "grad_norm": 0.07739850878715515, "tokens_per_second": 4357394.56716606, "eta_seconds": 307.34984481127924, "mean_data_wait_seconds": 0.00017639994621276855}
{"step": 43770, "total_steps": 46238, "tokens": 23665754112, "loss": 2.9171962160052676, "ce_loss": 2.9171962160052676, "z_loss": 41.92355647231593, "muon_lr": 0.01406334680080641, "residual_lr": 8.438008080483845e-05, "adamw_lr": 0.0008438008080483845, "weight_decay": 0.01, "grad_norm": 0.08106812834739685, "tokens_per_second": 4357501.169138448, "eta_seconds": 306.1015412793847, "mean_data_wait_seconds": 0.00010693073272705078}
{"step": 43780, "total_steps": 46238, "tokens": 23671160832, "loss": 2.9094389550613635, "ce_loss": 2.9094389550613635, "z_loss": 41.905796975800484, "muon_lr": 0.01396161979474296, "residual_lr": 8.376971876845776e-05, "adamw_lr": 0.0008376971876845776, "weight_decay": 0.01, "grad_norm": 0.08378926664590836, "tokens_per_second": 4357593.217788154, "eta_seconds": 304.8543169603818, "mean_data_wait_seconds": 7.680058479309082e-05}
{"step": 43790, "total_steps": 46238, "tokens": 23676567552, "loss": 2.9495774417212517, "ce_loss": 2.9495774417212517, "z_loss": 41.411670193527684, "muon_lr": 0.013860119311218428, "residual_lr": 8.316071586731056e-05, "adamw_lr": 0.0008316071586731056, "weight_decay": 0.01, "grad_norm": 0.07827972620725632, "tokens_per_second": 4357578.290228424, "eta_seconds": 303.6145987248911, "mean_data_wait_seconds": 9.40561294555664e-05}
{"step": 43800, "total_steps": 46238, "tokens": 23681974272, "loss": 2.91178405826742, "ce_loss": 2.91178405826742, "z_loss": 42.54432981664484, "muon_lr": 0.013758847433160463, "residual_lr": 8.255308459896277e-05, "adamw_lr": 0.0008255308459896277, "weight_decay": 0.01, "grad_norm": 0.08231548219919205, "tokens_per_second": 4357545.017389864, "eta_seconds": 302.3761449948813, "mean_data_wait_seconds": 9.79006290435791e-05}
{"step": 43810, "total_steps": 46238, "tokens": 23687380992, "loss": 2.937168486190565, "ce_loss": 2.937168486190565, "z_loss": 42.73478311480898, "muon_lr": 0.01365780623880541, "residual_lr": 8.194683743283245e-05, "adamw_lr": 0.0008194683743283246, "weight_decay": 0.01, "grad_norm": 0.09507038444280624, "tokens_per_second": 4357615.4123587655, "eta_seconds": 301.1305082771643, "mean_data_wait_seconds": 8.39531421661377e-05}
{"step": 43820, "total_steps": 46238, "tokens": 23692787712, "loss": 2.8617968198024863, "ce_loss": 2.8617968198024863, "z_loss": 41.803938403274074, "muon_lr": 0.013556997801655682, "residual_lr": 8.134198680993408e-05, "adamw_lr": 0.0008134198680993409, "weight_decay": 0.01, "grad_norm": 0.08026670664548874, "tokens_per_second": 4357634.387550091, "eta_seconds": 299.8884504247497, "mean_data_wait_seconds": 7.024407386779785e-05}
{"step": 43830, "total_steps": 46238, "tokens": 23698194432, "loss": 2.916741105643186, "ce_loss": 2.916741105643186, "z_loss": 42.624411871938996, "muon_lr": 0.01345642419043717, "residual_lr": 8.0738545142623e-05, "adamw_lr": 0.0008073854514262301, "weight_decay": 0.01, "grad_norm": 0.09337704628705978, "tokens_per_second": 4357688.167644323, "eta_seconds": 298.6440180972171, "mean_data_wait_seconds": 0.00011655688285827637}
{"step": 43840, "total_steps": 46238, "tokens": 23703601152, "loss": 2.861925036618204, "ce_loss": 2.861925036618204, "z_loss": 42.665102236198656, "muon_lr": 0.013356087469056845, "residual_lr": 8.013652481434105e-05, "adamw_lr": 0.0008013652481434106, "weight_decay": 0.01, "grad_norm": 0.0861789807677269, "tokens_per_second": 4357805.948460385, "eta_seconds": 297.3952487393052, "mean_data_wait_seconds": 6.052851676940918e-05}
{"step": 43850, "total_steps": 46238, "tokens": 23709007872, "loss": 2.839434670679497, "ce_loss": 2.839434670679497, "z_loss": 42.446265480735086, "muon_lr": 0.01325598969656037, "residual_lr": 7.95359381793622e-05, "adamw_lr": 0.0007953593817936221, "weight_decay": 0.01, "grad_norm": 0.0698467493057251, "tokens_per_second": 4357893.742897771, "eta_seconds": 296.1485846467264, "mean_data_wait_seconds": 0.00010418891906738281}
{"step": 43860, "total_steps": 46238, "tokens": 23714414592, "loss": 2.831415624329538, "ce_loss": 2.831415624329538, "z_loss": 42.18934651577111, "muon_lr": 0.013156132927089859, "residual_lr": 7.893679756253914e-05, "adamw_lr": 0.0007893679756253916, "weight_decay": 0.01, "grad_norm": 0.07327625900506973, "tokens_per_second": 4357908.420092553, "eta_seconds": 294.90691866643346, "mean_data_wait_seconds": 7.236003875732422e-05}
{"step": 43870, "total_steps": 46238, "tokens": 23719821312, "loss": 2.8794119159380593, "ce_loss": 2.8794119159380593, "z_loss": 42.206769076260656, "muon_lr": 0.013056519209841695, "residual_lr": 7.833911525905016e-05, "adamw_lr": 0.0007833911525905017, "weight_decay": 0.01, "grad_norm": 0.07580963522195816, "tokens_per_second": 4358036.364772585, "eta_seconds": 293.657628546838, "mean_data_wait_seconds": 8.597970008850098e-05}
{"step": 43880, "total_steps": 46238, "tokens": 23725228032, "loss": 2.886254252809467, "ce_loss": 2.886254252809467, "z_loss": 42.664571212999746, "muon_lr": 0.012957150589024517, "residual_lr": 7.77429035341471e-05, "adamw_lr": 0.000777429035341471, "weight_decay": 0.01, "grad_norm": 0.07140523195266724, "tokens_per_second": 4358127.386376201, "eta_seconds": 292.41088913182006, "mean_data_wait_seconds": 8.07046890258789e-05}
{"step": 43890, "total_steps": 46238, "tokens": 23730634752, "loss": 2.8839521606763205, "ce_loss": 2.8839521606763205, "z_loss": 42.887966213804305, "muon_lr": 0.012858029103817253, "residual_lr": 7.714817462290351e-05, "adamw_lr": 0.0007714817462290352, "weight_decay": 0.01, "grad_norm": 0.07707361876964569, "tokens_per_second": 4358282.387570436, "eta_seconds": 291.15992750240116, "mean_data_wait_seconds": 6.392598152160645e-05}
{"step": 43900, "total_steps": 46238, "tokens": 23736041472, "loss": 2.854200187957648, "ce_loss": 2.854200187957648, "z_loss": 41.89064650102095, "muon_lr": 0.01275915678832728, "residual_lr": 7.655494072996368e-05, "adamw_lr": 0.0007655494072996369, "weight_decay": 0.01, "grad_norm": 0.07410594820976257, "tokens_per_second": 4358357.615632561, "eta_seconds": 289.9143611960377, "mean_data_wait_seconds": 8.097290992736816e-05}
{"step": 43910, "total_steps": 46238, "tokens": 23741448192, "loss": 2.8782246221195567, "ce_loss": 2.8782246221195567, "z_loss": 43.03337348591197, "muon_lr": 0.012660535671548645, "residual_lr": 7.596321402929187e-05, "adamw_lr": 0.0007596321402929188, "weight_decay": 0.01, "grad_norm": 0.07354718446731567, "tokens_per_second": 4358416.786868511, "eta_seconds": 288.6699013712194, "mean_data_wait_seconds": 6.920099258422852e-05}
{"step": 43920, "total_steps": 46238, "tokens": 23746854912, "loss": 2.8878170360218394, "ce_loss": 2.8878170360218394, "z_loss": 43.21764683000969, "muon_lr": 0.012562167777320483, "residual_lr": 7.537300666392288e-05, "adamw_lr": 0.000753730066639229, "weight_decay": 0.01, "grad_norm": 0.0796256884932518, "tokens_per_second": 4358497.229583434, "eta_seconds": 287.42407256726216, "mean_data_wait_seconds": 6.839632987976074e-05}
{"step": 43930, "total_steps": 46238, "tokens": 23752261632, "loss": 2.8026916475007027, "ce_loss": 2.8026916475007027, "z_loss": 43.43107105023933, "muon_lr": 0.012464055124285462, "residual_lr": 7.478433074571277e-05, "adamw_lr": 0.0007478433074571277, "weight_decay": 0.01, "grad_norm": 0.07288237661123276, "tokens_per_second": 4358413.957326467, "eta_seconds": 286.1890394562557, "mean_data_wait_seconds": 8.293986320495605e-05}
{"step": 43940, "total_steps": 46238, "tokens": 23757668352, "loss": 2.8282725287206247, "ce_loss": 2.8282725287206247, "z_loss": 42.7271627368349, "muon_lr": 0.012366199725848324, "residual_lr": 7.419719835508993e-05, "adamw_lr": 0.0007419719835508994, "weight_decay": 0.01, "grad_norm": 0.07970011234283447, "tokens_per_second": 4358419.888156561, "eta_seconds": 284.94812704364847, "mean_data_wait_seconds": 9.888410568237305e-05}
{"step": 43950, "total_steps": 46238, "tokens": 23763075072, "loss": 2.7888407598842275, "ce_loss": 2.7888407598842275, "z_loss": 43.07947976661451, "muon_lr": 0.012268603590134628, "residual_lr": 7.361162154080776e-05, "adamw_lr": 0.0007361162154080776, "weight_decay": 0.01, "grad_norm": 0.07629566639661789, "tokens_per_second": 4358420.300488134, "eta_seconds": 283.7075772296473, "mean_data_wait_seconds": 9.97781753540039e-05}
{"step": 43960, "total_steps": 46238, "tokens": 23768481792, "loss": 2.8606242880676733, "ce_loss": 2.8606242880676733, "z_loss": 42.98477051474831, "muon_lr": 0.012171268719949497, "residual_lr": 7.302761231969697e-05, "adamw_lr": 0.0007302761231969697, "weight_decay": 0.01, "grad_norm": 0.07801707088947296, "tokens_per_second": 4358339.438624934, "eta_seconds": 282.47229508778645, "mean_data_wait_seconds": 7.280707359313965e-05}
{"step": 43970, "total_steps": 46238, "tokens": 23773888512, "loss": 2.8508018815156184, "ce_loss": 2.8508018815156184, "z_loss": 42.694780638723664, "muon_lr": 0.012074197112736546, "residual_lr": 7.244518267641927e-05, "adamw_lr": 0.0007244518267641928, "weight_decay": 0.01, "grad_norm": 0.08564302325248718, "tokens_per_second": 4358306.520223623, "eta_seconds": 281.23387336627934, "mean_data_wait_seconds": 0.0001169443130493164}
{"step": 43980, "total_steps": 46238, "tokens": 23779295232, "loss": 2.8944120461290535, "ce_loss": 2.8944120461290535, "z_loss": 42.48575242360433, "muon_lr": 0.011977390760536845, "residual_lr": 7.186434456322106e-05, "adamw_lr": 0.0007186434456322106, "weight_decay": 0.01, "grad_norm": 0.08879625052213669, "tokens_per_second": 4358253.021496023, "eta_seconds": 279.9967551175169, "mean_data_wait_seconds": 0.0001646578311920166}
{"step": 43990, "total_steps": 46238, "tokens": 23784701952, "loss": 2.8418007637515212, "ce_loss": 2.8418007637515212, "z_loss": 42.93255988034335, "muon_lr": 0.011880851649948097, "residual_lr": 7.128510989968857e-05, "adamw_lr": 0.0007128510989968858, "weight_decay": 0.01, "grad_norm": 0.07635673135519028, "tokens_per_second": 4358332.828354233, "eta_seconds": 278.75108025165656, "mean_data_wait_seconds": 0.00011909008026123047}
{"step": 44000, "total_steps": 46238, "tokens": 23790108672, "loss": 2.8224073106592353, "ce_loss": 2.8224073106592353, "z_loss": 43.40630597779245, "muon_lr": 0.011784581762083836, "residual_lr": 7.0707490572503e-05, "adamw_lr": 0.00070707490572503, "weight_decay": 0.01, "grad_norm": 0.08782390505075455, "tokens_per_second": 4358342.672929224, "eta_seconds": 277.50990566033477, "mean_data_wait_seconds": 9.244680404663086e-05}
{"step": 44010, "total_steps": 46238, "tokens": 23795515392, "loss": 2.8916216655211016, "ce_loss": 2.8916216655211016, "z_loss": 42.646352652347446, "muon_lr": 0.011688583072532796, "residual_lr": 7.013149843519677e-05, "adamw_lr": 0.0007013149843519677, "weight_decay": 0.01, "grad_norm": 0.07850513607263565, "tokens_per_second": 4358281.638011025, "eta_seconds": 276.27322968267384, "mean_data_wait_seconds": 7.420778274536133e-05}
{"step": 44020, "total_steps": 46238, "tokens": 23800922112, "loss": 2.818087687998107, "ce_loss": 2.818087687998107, "z_loss": 43.568509275263004, "muon_lr": 0.01159285755131832, "residual_lr": 6.955714530790991e-05, "adamw_lr": 0.0006955714530790992, "weight_decay": 0.01, "grad_norm": 0.07605432718992233, "tokens_per_second": 4358302.5914410055, "eta_seconds": 275.0313450823703, "mean_data_wait_seconds": 8.934736251831055e-05}
{"step": 44030, "total_steps": 46238, "tokens": 23806328832, "loss": 2.8396302425500117, "ce_loss": 2.8396302425500117, "z_loss": 43.25453804478501, "muon_lr": 0.011497407162857992, "residual_lr": 6.898444297714795e-05, "adamw_lr": 0.0006898444297714796, "weight_decay": 0.01, "grad_norm": 0.06811317801475525, "tokens_per_second": 4358266.167265411, "eta_seconds": 273.79307692644016, "mean_data_wait_seconds": 0.00010010600090026855}
{"step": 44040, "total_steps": 46238, "tokens": 23811735552, "loss": 2.8729909188819653, "ce_loss": 2.8729909188819653, "z_loss": 43.94868830478553, "muon_lr": 0.011402233865923296, "residual_lr": 6.841340319553976e-05, "adamw_lr": 0.0006841340319553977, "weight_decay": 0.01, "grad_norm": 0.076206274330616, "tokens_per_second": 4358308.389721154, "eta_seconds": 272.5498697617401, "mean_data_wait_seconds": 0.00010433793067932129}
{"step": 44050, "total_steps": 46238, "tokens": 23817142272, "loss": 2.8428585529327393, "ce_loss": 2.8428585529327393, "z_loss": 43.35128879547119, "muon_lr": 0.011307339613599428, "residual_lr": 6.784403768159655e-05, "adamw_lr": 0.0006784403768159656, "weight_decay": 0.01, "grad_norm": 0.07843104749917984, "tokens_per_second": 4358424.879013119, "eta_seconds": 271.30206366382134, "mean_data_wait_seconds": 0.00017696619033813477}
{"step": 44060, "total_steps": 46238, "tokens": 23822548992, "loss": 2.822651339299751, "ce_loss": 2.822651339299751, "z_loss": 42.76829066421046, "muon_lr": 0.011212726353245175, "residual_lr": 6.727635811947105e-05, "adamw_lr": 0.0006727635811947106, "weight_decay": 0.01, "grad_norm": 0.09386283159255981, "tokens_per_second": 4358573.357959262, "eta_seconds": 270.0523422074754, "mean_data_wait_seconds": 0.00011664628982543945}
{"step": 44070, "total_steps": 46238, "tokens": 23827955712, "loss": 2.8586588158751978, "ce_loss": 2.8586588158751978, "z_loss": 42.76528956673362, "muon_lr": 0.011118396026453032, "residual_lr": 6.671037615871818e-05, "adamw_lr": 0.0006671037615871819, "weight_decay": 0.01, "grad_norm": 0.06827277690172195, "tokens_per_second": 4358595.480107631, "eta_seconds": 268.81049855332475, "mean_data_wait_seconds": 9.891390800476074e-05}
{"step": 44080, "total_steps": 46238, "tokens": 23833362432, "loss": 2.939364102753726, "ce_loss": 2.939364102753726, "z_loss": 42.94937992095947, "muon_lr": 0.011024350569009286, "residual_lr": 6.614610341405571e-05, "adamw_lr": 0.0006614610341405571, "weight_decay": 0.01, "grad_norm": 0.06848685443401337, "tokens_per_second": 4358658.396048575, "eta_seconds": 267.5661632619036, "mean_data_wait_seconds": 8.487701416015625e-05}
{"step": 44090, "total_steps": 46238, "tokens": 23838769152, "loss": 2.86748667196794, "ce_loss": 2.86748667196794, "z_loss": 43.44790490468343, "muon_lr": 0.01093059191085434, "residual_lr": 6.558355146512603e-05, "adamw_lr": 0.0006558355146512604, "weight_decay": 0.01, "grad_norm": 0.07518245279788971, "tokens_per_second": 4358747.403506827, "eta_seconds": 266.32026968712637, "mean_data_wait_seconds": 0.00011268258094787598}
{"step": 44100, "total_steps": 46238, "tokens": 23844175872, "loss": 2.9147861112247813, "ce_loss": 2.9147861112247813, "z_loss": 43.93427791017474, "muon_lr": 0.010837121976043055, "residual_lr": 6.502273185625832e-05, "adamw_lr": 0.0006502273185625833, "weight_decay": 0.01, "grad_norm": 0.06969047337770462, "tokens_per_second": 4358786.520901138, "eta_seconds": 265.0774609996565, "mean_data_wait_seconds": 0.00010383129119873047}
{"step": 44110, "total_steps": 46238, "tokens": 23849582592, "loss": 2.8883055159539888, "ce_loss": 2.8883055159539888, "z_loss": 43.506506573070176, "muon_lr": 0.010743942682705322, "residual_lr": 6.446365609623192e-05, "adamw_lr": 0.0006446365609623192, "weight_decay": 0.01, "grad_norm": 0.07843399792909622, "tokens_per_second": 4358722.172845714, "eta_seconds": 263.8409374115222, "mean_data_wait_seconds": 0.00010418891906738281}
{"step": 44120, "total_steps": 46238, "tokens": 23854989312, "loss": 2.9434774608323067, "ce_loss": 2.9434774608323067, "z_loss": 42.97682970220392, "muon_lr": 0.010651055943006669, "residual_lr": 6.390633565804e-05, "adamw_lr": 0.0006390633565804001, "weight_decay": 0.01, "grad_norm": 0.0756802037358284, "tokens_per_second": 4358769.3725189995, "eta_seconds": 262.5976568561866, "mean_data_wait_seconds": 6.750226020812988e-05}
{"step": 44130, "total_steps": 46238, "tokens": 23860396032, "loss": 2.930470154140935, "ce_loss": 2.930470154140935, "z_loss": 43.13518460591634, "muon_lr": 0.010558463663109037, "residual_lr": 6.335078197865421e-05, "adamw_lr": 0.0006335078197865422, "weight_decay": 0.01, "grad_norm": 0.06952010840177536, "tokens_per_second": 4358863.802908182, "eta_seconds": 261.3515713062523, "mean_data_wait_seconds": 5.97536563873291e-05}
{"step": 44140, "total_steps": 46238, "tokens": 23865802752, "loss": 2.9395655917398855, "ce_loss": 2.9395655917398855, "z_loss": 42.77902652278091, "muon_lr": 0.010466167743131617, "residual_lr": 6.27970064587897e-05, "adamw_lr": 0.0006279700645878969, "weight_decay": 0.01, "grad_norm": 0.07955276966094971, "tokens_per_second": 4358941.661627688, "eta_seconds": 260.1065286055303, "mean_data_wait_seconds": 6.371736526489258e-05}
{"step": 44150, "total_steps": 46238, "tokens": 23871209472, "loss": 2.8848557616725112, "ce_loss": 2.8848557616725112, "z_loss": 42.79540472319632, "muon_lr": 0.01037417007711193, "residual_lr": 6.224502046267158e-05, "adamw_lr": 0.0006224502046267158, "weight_decay": 0.01, "grad_norm": 0.07036682218313217, "tokens_per_second": 4358453.488062479, "eta_seconds": 258.89514872432767, "mean_data_wait_seconds": 6.216764450073242e-05}
{"step": 44160, "total_steps": 46238, "tokens": 23876616192, "loss": 2.899032372416872, "ce_loss": 2.899032372416872, "z_loss": 43.168752352396645, "muon_lr": 0.010282472552966916, "residual_lr": 6.169483531780148e-05, "adamw_lr": 0.0006169483531780149, "weight_decay": 0.01, "grad_norm": 0.07726267725229263, "tokens_per_second": 4358481.340272222, "eta_seconds": 257.6529888114334, "mean_data_wait_seconds": 0.00011083483695983887}
{"step": 44170, "total_steps": 46238, "tokens": 23882022912, "loss": 2.9212485642144173, "ce_loss": 2.9212485642144173, "z_loss": 42.856784184773765, "muon_lr": 0.010191077052454166, "residual_lr": 6.114646231472498e-05, "adamw_lr": 0.0006114646231472499, "weight_decay": 0.01, "grad_norm": 0.07778802514076233, "tokens_per_second": 4358497.291696983, "eta_seconds": 256.41154489851107, "mean_data_wait_seconds": 7.656216621398926e-05}
{"step": 44180, "total_steps": 46238, "tokens": 23887429632, "loss": 2.8931654167897776, "ce_loss": 2.8931654167897776, "z_loss": 42.970690524939336, "muon_lr": 0.01009998545113337, "residual_lr": 6.0599912706800214e-05, "adamw_lr": 0.0006059991270680022, "weight_decay": 0.01, "grad_norm": 0.08260523527860641, "tokens_per_second": 4358026.494527268, "eta_seconds": 255.1986100581613, "mean_data_wait_seconds": 7.706880569458008e-05}
{"step": 44190, "total_steps": 46238, "tokens": 23892836352, "loss": 2.974780877431234, "ce_loss": 2.974780877431234, "z_loss": 42.71010930610426, "muon_lr": 0.010009199618327781, "residual_lr": 6.005519770996667e-05, "adamw_lr": 0.0006005519770996668, "weight_decay": 0.01, "grad_norm": 0.07252418249845505, "tokens_per_second": 4358114.166922262, "eta_seconds": 253.95286621910606, "mean_data_wait_seconds": 0.00015240907669067383}
{"step": 44200, "total_steps": 46238, "tokens": 23898243072, "loss": 2.9000802365216343, "ce_loss": 2.9000802365216343, "z_loss": 43.107839266459145, "muon_lr": 0.00991872141708588, "residual_lr": 5.951232850251527e-05, "adamw_lr": 0.0005951232850251528, "weight_decay": 0.01, "grad_norm": 0.0799633041024208, "tokens_per_second": 4358166.693409345, "eta_seconds": 252.70921042683364, "mean_data_wait_seconds": 8.291006088256836e-05}
{"step": 44210, "total_steps": 46238, "tokens": 23903649792, "loss": 2.8806838555769487, "ce_loss": 2.8806838555769487, "z_loss": 42.70010627399791, "muon_lr": 0.009828552704143095, "residual_lr": 5.897131622485857e-05, "adamw_lr": 0.0005897131622485857, "weight_decay": 0.01, "grad_norm": 0.0813906341791153, "tokens_per_second": 4357805.779989032, "eta_seconds": 251.48944201059788, "mean_data_wait_seconds": 0.00011652708053588867}
{"step": 44220, "total_steps": 46238, "tokens": 23909056512, "loss": 2.8922019763426348, "ce_loss": 2.8922019763426348, "z_loss": 42.59969037951845, "muon_lr": 0.009738695329883777, "residual_lr": 5.843217197930265e-05, "adamw_lr": 0.0005843217197930266, "weight_decay": 0.01, "grad_norm": 0.06720244884490967, "tokens_per_second": 4357853.032667695, "eta_seconds": 250.2460307461126, "mean_data_wait_seconds": 7.596611976623535e-05}
{"step": 44230, "total_steps": 46238, "tokens": 23914463232, "loss": 2.932489042932337, "ce_loss": 2.932489042932337, "z_loss": 43.206041278261125, "muon_lr": 0.009649151138303155, "residual_lr": 5.7894906829818924e-05, "adamw_lr": 0.0005789490682981893, "weight_decay": 0.01, "grad_norm": 0.07371750473976135, "tokens_per_second": 4357908.855022505, "eta_seconds": 249.00215679118332, "mean_data_wait_seconds": 0.00011259317398071289}
{"step": 44240, "total_steps": 46238, "tokens": 23919869952, "loss": 2.92435298724608, "ce_loss": 2.92435298724608, "z_loss": 42.78956430608576, "muon_lr": 0.009559921966969554, "residual_lr": 5.7359531801817316e-05, "adamw_lr": 0.0005735953180181732, "weight_decay": 0.01, "grad_norm": 0.07359886914491653, "tokens_per_second": 4357616.859644289, "eta_seconds": 247.77809035926518, "mean_data_wait_seconds": 0.00012859702110290527}
{"step": 44250, "total_steps": 46238, "tokens": 23925276672, "loss": 2.873839111039133, "ce_loss": 2.873839111039133, "z_loss": 42.057019435998164, "muon_lr": 0.009471009646986618, "residual_lr": 5.68260578819197e-05, "adamw_lr": 0.000568260578819197, "weight_decay": 0.01, "grad_norm": 0.07829795032739639, "tokens_per_second": 4357336.766767882, "eta_seconds": 246.55318638519856, "mean_data_wait_seconds": 0.00011900067329406738}
{"step": 44260, "total_steps": 46238, "tokens": 23930683392, "loss": 2.9733788190465984, "ce_loss": 2.9733788190465984, "z_loss": 42.381141547000766, "muon_lr": 0.009382416002955796, "residual_lr": 5.6294496017734766e-05, "adamw_lr": 0.0005629449601773478, "weight_decay": 0.01, "grad_norm": 0.07399067282676697, "tokens_per_second": 4357329.538979164, "eta_seconds": 245.31276196530777, "mean_data_wait_seconds": 0.0001646876335144043}
{"step": 44270, "total_steps": 46238, "tokens": 23936090112, "loss": 2.91433106769215, "ce_loss": 2.91433106769215, "z_loss": 43.64877183509596, "muon_lr": 0.009294142852938872, "residual_lr": 5.5764857117633226e-05, "adamw_lr": 0.0005576485711763324, "weight_decay": 0.01, "grad_norm": 0.07067802548408508, "tokens_per_second": 4357399.370231314, "eta_seconds": 244.06801709881913, "mean_data_wait_seconds": 0.00015747547149658203}
{"step": 44280, "total_steps": 46238, "tokens": 23941496832, "loss": 2.916416636019042, "ce_loss": 2.916416636019042, "z_loss": 42.63680721051765, "muon_lr": 0.009206192008420664, "residual_lr": 5.523715205052397e-05, "adamw_lr": 0.0005523715205052398, "weight_decay": 0.01, "grad_norm": 0.08079122751951218, "tokens_per_second": 4357335.901094792, "eta_seconds": 242.83074062161486, "mean_data_wait_seconds": 9.423494338989258e-05}
{"step": 44290, "total_steps": 46238, "tokens": 23946903552, "loss": 2.9508257309595742, "ce_loss": 2.9508257309595742, "z_loss": 42.99138849431818, "muon_lr": 0.009118565274271817, "residual_lr": 5.4711391645630896e-05, "adamw_lr": 0.0005471139164563091, "weight_decay": 0.01, "grad_norm": 0.06517712026834488, "tokens_per_second": 4357338.366412328, "eta_seconds": 241.58977235149746, "mean_data_wait_seconds": 8.147954940795898e-05}
{"step": 44300, "total_steps": 46238, "tokens": 23952310272, "loss": 2.962637888662743, "ce_loss": 2.962637888662743, "z_loss": 42.99505311792547, "muon_lr": 0.00903126444871182, "residual_lr": 5.418758669227092e-05, "adamw_lr": 0.0005418758669227092, "weight_decay": 0.01, "grad_norm": 0.06925395131111145, "tokens_per_second": 4357261.903580546, "eta_seconds": 240.353159202894, "mean_data_wait_seconds": 7.560849189758301e-05}
{"step": 44310, "total_steps": 46238, "tokens": 23957716992, "loss": 2.9259190902565466, "ce_loss": 2.9259190902565466, "z_loss": 43.341155803564824, "muon_lr": 0.008944291323272064, "residual_lr": 5.366574793963238e-05, "adamw_lr": 0.0005366574793963238, "weight_decay": 0.01, "grad_norm": 0.06883492320775986, "tokens_per_second": 4357245.167461218, "eta_seconds": 239.11322497537046, "mean_data_wait_seconds": 9.882450103759766e-05}
{"step": 44320, "total_steps": 46238, "tokens": 23963123712, "loss": 2.8796436533783423, "ce_loss": 2.8796436533783423, "z_loss": 43.470967697374746, "muon_lr": 0.008857647682759099, "residual_lr": 5.3145886096554584e-05, "adamw_lr": 0.0005314588609655459, "weight_decay": 0.01, "grad_norm": 0.0673772543668747, "tokens_per_second": 4357235.492700145, "eta_seconds": 237.87289572400613, "mean_data_wait_seconds": 0.00010022521018981934}
{"step": 44330, "total_steps": 46238, "tokens": 23968530432, "loss": 2.970880596926718, "ce_loss": 2.970880596926718, "z_loss": 42.908846970760465, "muon_lr": 0.008771335305217967, "residual_lr": 5.2628011831307796e-05, "adamw_lr": 0.000526280118313078, "weight_decay": 0.01, "grad_norm": 0.0770459994673729, "tokens_per_second": 4357140.171317337, "eta_seconds": 236.63721235947042, "mean_data_wait_seconds": 0.00012060999870300293}
{"step": 44340, "total_steps": 46238, "tokens": 23973937152, "loss": 2.8604928453763327, "ce_loss": 2.8604928453763327, "z_loss": 43.52047009901567, "muon_lr": 0.008685355961895782, "residual_lr": 5.211213577137469e-05, "adamw_lr": 0.000521121357713747, "weight_decay": 0.01, "grad_norm": 0.072691410779953, "tokens_per_second": 4357102.13710046, "eta_seconds": 235.3983798696413, "mean_data_wait_seconds": 7.88271427154541e-05}
{"step": 44350, "total_steps": 46238, "tokens": 23979343872, "loss": 2.884519484910098, "ce_loss": 2.884519484910098, "z_loss": 42.967796527978145, "muon_lr": 0.008599711417205334, "residual_lr": 5.1598268503231994e-05, "adamw_lr": 0.0005159826850323199, "weight_decay": 0.01, "grad_norm": 0.07186892628669739, "tokens_per_second": 4357127.836063052, "eta_seconds": 234.1561006210597, "mean_data_wait_seconds": 6.16610050201416e-05}
{"step": 44360, "total_steps": 46238, "tokens": 23984750592, "loss": 2.9159127275149026, "ce_loss": 2.9159127275149026, "z_loss": 43.19227224407774, "muon_lr": 0.00851440342868886, "residual_lr": 5.1086420572133155e-05, "adamw_lr": 0.0005108642057213317, "weight_decay": 0.01, "grad_norm": 0.06230101361870766, "tokens_per_second": 4357196.016579865, "eta_seconds": 232.9115651759429, "mean_data_wait_seconds": 6.306171417236328e-05}
{"step": 44370, "total_steps": 46238, "tokens": 23990157312, "loss": 2.8719908558961116, "ce_loss": 2.8719908558961116, "z_loss": 43.812188350793086, "muon_lr": 0.008429433746982051, "residual_lr": 5.05766024818923e-05, "adamw_lr": 0.0005057660248189231, "weight_decay": 0.01, "grad_norm": 0.07449357956647873, "tokens_per_second": 4357270.678100362, "eta_seconds": 231.666724097132, "mean_data_wait_seconds": 7.417798042297363e-05}
{"step": 44380, "total_steps": 46238, "tokens": 23995564032, "loss": 2.8448764916622276, "ce_loss": 2.8448764916622276, "z_loss": 44.01002109411991, "muon_lr": 0.008344804115778056, "residual_lr": 5.006882469466834e-05, "adamw_lr": 0.0005006882469466834, "weight_decay": 0.01, "grad_norm": 0.0646526888012886, "tokens_per_second": 4357311.483762219, "eta_seconds": 230.42371603259716, "mean_data_wait_seconds": 0.00010612607002258301}
{"step": 44390, "total_steps": 46238, "tokens": 24000970752, "loss": 2.911375737551487, "ce_loss": 2.911375737551487, "z_loss": 43.49327991947983, "muon_lr": 0.008260516271791751, "residual_lr": 4.95630976307505e-05, "adamw_lr": 0.000495630976307505, "weight_decay": 0.01, "grad_norm": 0.06321091949939728, "tokens_per_second": 4357375.79776069, "eta_seconds": 229.17949480354758, "mean_data_wait_seconds": 0.00013238191604614258}
{"step": 44400, "total_steps": 46238, "tokens": 24006377472, "loss": 2.8177843003562004, "ce_loss": 2.8177843003562004, "z_loss": 44.03184292533181, "muon_lr": 0.008176571944724034, "residual_lr": 4.90594316683442e-05, "adamw_lr": 0.0004905943166834421, "weight_decay": 0.01, "grad_norm": 0.06292066723108292, "tokens_per_second": 4357405.415317563, "eta_seconds": 227.9371252692161, "mean_data_wait_seconds": 6.681680679321289e-05}
{"step": 44410, "total_steps": 46238, "tokens": 24011784192, "loss": 2.8569523619882986, "ce_loss": 2.8569523619882986, "z_loss": 43.31291716026537, "muon_lr": 0.008092972857226405, "residual_lr": 4.855783714335842e-05, "adamw_lr": 0.0004855783714335843, "weight_decay": 0.01, "grad_norm": 0.06645388156175613, "tokens_per_second": 4357431.710131286, "eta_seconds": 226.6949454889422, "mean_data_wait_seconds": 6.574392318725586e-05}
{"step": 44420, "total_steps": 46238, "tokens": 24017190912, "loss": 2.874489153876449, "ce_loss": 2.874489153876449, "z_loss": 43.97993038639878, "muon_lr": 0.008009720724865572, "residual_lr": 4.8058324349193425e-05, "adamw_lr": 0.0004805832434919343, "weight_decay": 0.01, "grad_norm": 0.07807699590921402, "tokens_per_second": 4357463.334244608, "eta_seconds": 225.45250496532407, "mean_data_wait_seconds": 7.104873657226562e-05}
{"step": 44430, "total_steps": 46238, "tokens": 24022597632, "loss": 2.859006784178994, "ce_loss": 2.859006784178994, "z_loss": 44.21046493992661, "muon_lr": 0.00792681725608825, "residual_lr": 4.7560903536529495e-05, "adamw_lr": 0.00047560903536529505, "weight_decay": 0.01, "grad_norm": 0.06979278475046158, "tokens_per_second": 4357473.829911956, "eta_seconds": 224.21116962158337, "mean_data_wait_seconds": 0.00013706088066101074}
{"step": 44440, "total_steps": 46238, "tokens": 24028004352, "loss": 2.878368121204954, "ce_loss": 2.878368121204954, "z_loss": 43.93999738404245, "muon_lr": 0.007844264152186095, "residual_lr": 4.706558491311657e-05, "adamw_lr": 0.0004706558491311657, "weight_decay": 0.01, "grad_norm": 0.07038005441427231, "tokens_per_second": 4357510.923578197, "eta_seconds": 222.96847926250862, "mean_data_wait_seconds": 0.00014927983283996582}
{"step": 44450, "total_steps": 46238, "tokens": 24033411072, "loss": 2.8661548242424475, "ce_loss": 2.8661548242424475, "z_loss": 43.904617656360976, "muon_lr": 0.007762063107260811, "residual_lr": 4.657237864356486e-05, "adamw_lr": 0.00046572378643564867, "weight_decay": 0.01, "grad_norm": 0.06486499309539795, "tokens_per_second": 4357507.535227741, "eta_seconds": 221.72786993230145, "mean_data_wait_seconds": 7.414817810058594e-05}
{"step": 44460, "total_steps": 46238, "tokens": 24038817792, "loss": 2.7847510052449778, "ce_loss": 2.7847510052449778, "z_loss": 44.06122817415179, "muon_lr": 0.007680215808189378, "residual_lr": 4.6081294849136265e-05, "adamw_lr": 0.00046081294849136266, "weight_decay": 0.01, "grad_norm": 0.07365355640649796, "tokens_per_second": 4357514.567760146, "eta_seconds": 220.48673138317426, "mean_data_wait_seconds": 6.815791130065918e-05}
{"step": 44470, "total_steps": 46238, "tokens": 24044224512, "loss": 2.8388618754618093, "ce_loss": 2.8388618754618093, "z_loss": 44.45208266287139, "muon_lr": 0.0075987239345894275, "residual_lr": 4.5592343607536555e-05, "adamw_lr": 0.0004559234360753656, "weight_decay": 0.01, "grad_norm": 0.06429721415042877, "tokens_per_second": 4357519.863077495, "eta_seconds": 219.24568424692674, "mean_data_wait_seconds": 0.00011217594146728516}
{"step": 44480, "total_steps": 46238, "tokens": 24049631232, "loss": 2.84788460261894, "ce_loss": 2.84788460261894, "z_loss": 43.846271601590246, "muon_lr": 0.007517589158784763, "residual_lr": 4.510553495270857e-05, "adamw_lr": 0.0004510553495270858, "weight_decay": 0.01, "grad_norm": 0.06499931216239929, "tokens_per_second": 4357387.166873062, "eta_seconds": 218.0115439871983, "mean_data_wait_seconds": 9.85562801361084e-05}
{"step": 44490, "total_steps": 46238, "tokens": 24055037952, "loss": 2.8786395282456367, "ce_loss": 2.8786395282456367, "z_loss": 43.804297793995254, "muon_lr": 0.007436813145771074, "residual_lr": 4.462087887462644e-05, "adamw_lr": 0.0004462087887462644, "weight_decay": 0.01, "grad_norm": 0.07894352078437805, "tokens_per_second": 4357348.317742949, "eta_seconds": 216.77265968245266, "mean_data_wait_seconds": 0.00011864304542541504}
{"step": 44500, "total_steps": 46238, "tokens": 24060444672, "loss": 2.801916830467455, "ce_loss": 2.801916830467455, "z_loss": 45.1084849617698, "muon_lr": 0.007356397553181751, "residual_lr": 4.41383853190905e-05, "adamw_lr": 0.0004413838531909051, "weight_decay": 0.01, "grad_norm": 0.07532551884651184, "tokens_per_second": 4357301.739125974, "eta_seconds": 215.53413562503994, "mean_data_wait_seconds": 0.00011080503463745117}
{"step": 44510, "total_steps": 46238, "tokens": 24065851392, "loss": 2.858948806921641, "ce_loss": 2.858948806921641, "z_loss": 43.609448635216914, "muon_lr": 0.007276344031253871, "residual_lr": 4.365806418752322e-05, "adamw_lr": 0.0004365806418752322, "weight_decay": 0.01, "grad_norm": 0.06650784611701965, "tokens_per_second": 4357304.602603858, "eta_seconds": 214.29315348805565, "mean_data_wait_seconds": 8.526444435119629e-05}
{"step": 44520, "total_steps": 46238, "tokens": 24071258112, "loss": 2.814429745529637, "ce_loss": 2.814429745529637, "z_loss": 44.274650342536695, "muon_lr": 0.0071966542227943, "residual_lr": 4.317992533676579e-05, "adamw_lr": 0.000431799253367658, "weight_decay": 0.01, "grad_norm": 0.06467734277248383, "tokens_per_second": 4357300.430729317, "eta_seconds": 213.0525169788711, "mean_data_wait_seconds": 8.946657180786133e-05}
{"step": 44530, "total_steps": 46238, "tokens": 24076664832, "loss": 2.80650380524722, "ce_loss": 2.80650380524722, "z_loss": 43.669811422174625, "muon_lr": 0.007117329763146049, "residual_lr": 4.270397857887629e-05, "adamw_lr": 0.00042703978578876293, "weight_decay": 0.01, "grad_norm": 0.07082290947437286, "tokens_per_second": 4357364.037423084, "eta_seconds": 211.80858337138454, "mean_data_wait_seconds": 8.001923561096191e-05}
{"step": 44540, "total_steps": 46238, "tokens": 24082071552, "loss": 2.852828970461181, "ce_loss": 2.852828970461181, "z_loss": 44.17965022000399, "muon_lr": 0.007038372280154662, "residual_lr": 4.223023368092797e-05, "adamw_lr": 0.0004223023368092797, "weight_decay": 0.01, "grad_norm": 0.06410755962133408, "tokens_per_second": 4357478.103952538, "eta_seconds": 210.5622477294251, "mean_data_wait_seconds": 6.759166717529297e-05}
{"step": 44550, "total_steps": 46238, "tokens": 24087478272, "loss": 2.792447637427937, "ce_loss": 2.792447637427937, "z_loss": 44.104498660925664, "muon_lr": 0.006959783394134805, "residual_lr": 4.175870036480883e-05, "adamw_lr": 0.0004175870036480883, "weight_decay": 0.01, "grad_norm": 0.07691316306591034, "tokens_per_second": 4357486.290450572, "eta_seconds": 209.32106338438663, "mean_data_wait_seconds": 7.236003875732422e-05}
{"step": 44560, "total_steps": 46238, "tokens": 24092884992, "loss": 2.910168609835885, "ce_loss": 2.910168609835885, "z_loss": 43.36414640600031, "muon_lr": 0.00688156471783706, "residual_lr": 4.128938830702236e-05, "adamw_lr": 0.0004128938830702236, "weight_decay": 0.01, "grad_norm": 0.06827614456415176, "tokens_per_second": 4357510.941982355, "eta_seconds": 208.0790974646442, "mean_data_wait_seconds": 6.467103958129883e-05}
{"step": 44570, "total_steps": 46238, "tokens": 24098291712, "loss": 2.853244022889571, "ce_loss": 2.853244022889571, "z_loss": 44.37337600823605, "muon_lr": 0.006803717856414796, "residual_lr": 4.082230713848877e-05, "adamw_lr": 0.00040822307138488773, "weight_decay": 0.01, "grad_norm": 0.0672723576426506, "tokens_per_second": 4357612.620843902, "eta_seconds": 206.83348944070497, "mean_data_wait_seconds": 7.221102714538574e-05}
{"step": 44580, "total_steps": 46238, "tokens": 24103698432, "loss": 2.8483121214490947, "ce_loss": 2.8483121214490947, "z_loss": 43.71921628894228, "muon_lr": 0.0067262444073912355, "residual_lr": 4.035746644434741e-05, "adamw_lr": 0.0004035746644434741, "weight_decay": 0.01, "grad_norm": 0.0670047327876091, "tokens_per_second": 4357695.616024581, "eta_seconds": 205.58882100565384, "mean_data_wait_seconds": 7.030367851257324e-05}
{"step": 44590, "total_steps": 46238, "tokens": 24109105152, "loss": 2.8460718086271575, "ce_loss": 2.8460718086271575, "z_loss": 44.202992988355234, "muon_lr": 0.006649145960626657, "residual_lr": 3.989487576375994e-05, "adamw_lr": 0.0003989487576375994, "weight_decay": 0.01, "grad_norm": 0.06574346125125885, "tokens_per_second": 4357727.766545755, "eta_seconds": 204.34658420754522, "mean_data_wait_seconds": 6.988644599914551e-05}
{"step": 44600, "total_steps": 46238, "tokens": 24114511872, "loss": 2.8495017850037776, "ce_loss": 2.8495017850037776, "z_loss": 44.26635976271196, "muon_lr": 0.0065724240982858095, "residual_lr": 3.943454458971485e-05, "adamw_lr": 0.00039434544589714857, "weight_decay": 0.01, "grad_norm": 0.07388684153556824, "tokens_per_second": 4357795.442161748, "eta_seconds": 203.10271001636164, "mean_data_wait_seconds": 6.052851676940918e-05}
{"step": 44610, "total_steps": 46238, "tokens": 24119918592, "loss": 2.8874210010875356, "ce_loss": 2.8874210010875356, "z_loss": 43.937113877498746, "muon_lr": 0.006496080394805398, "residual_lr": 3.8976482368832385e-05, "adamw_lr": 0.0003897648236883239, "weight_decay": 0.01, "grad_norm": 0.07199782133102417, "tokens_per_second": 4357931.776568602, "eta_seconds": 201.8556941918552, "mean_data_wait_seconds": 6.270408630371094e-05}
{"step": 44620, "total_steps": 46238, "tokens": 24125325312, "loss": 2.853298794139515, "ce_loss": 2.853298794139515, "z_loss": 44.68235819267504, "muon_lr": 0.006420116416861816, "residual_lr": 3.852069850117089e-05, "adamw_lr": 0.00038520698501170894, "weight_decay": 0.01, "grad_norm": 0.06116843596100807, "tokens_per_second": 4357990.080373701, "eta_seconds": 200.61234832481102, "mean_data_wait_seconds": 5.8650970458984375e-05}
{"step": 44630, "total_steps": 46238, "tokens": 24130732032, "loss": 2.89786350365841, "ce_loss": 2.89786350365841, "z_loss": 43.57943393244888, "muon_lr": 0.006344533723338944, "residual_lr": 3.806720234003366e-05, "adamw_lr": 0.00038067202340033666, "weight_decay": 0.01, "grad_norm": 0.06563835591077805, "tokens_per_second": 4358112.292531489, "eta_seconds": 199.36611213276169, "mean_data_wait_seconds": 9.176135063171387e-05}
{"step": 44640, "total_steps": 46238, "tokens": 24136138752, "loss": 2.8536661447900715, "ce_loss": 2.8536661447900715, "z_loss": 43.53914867747914, "muon_lr": 0.006269333865296212, "residual_lr": 3.761600319177727e-05, "adamw_lr": 0.0003761600319177727, "weight_decay": 0.01, "grad_norm": 0.06410752236843109, "tokens_per_second": 4358120.016815489, "eta_seconds": 198.125150447539, "mean_data_wait_seconds": 0.00016957521438598633}
{"step": 44650, "total_steps": 46238, "tokens": 24141545472, "loss": 2.832412067687873, "ce_loss": 2.832412067687873, "z_loss": 43.568819132718176, "muon_lr": 0.0061945183859367405, "residual_lr": 3.7167110315620434e-05, "adamw_lr": 0.0003716711031562044, "weight_decay": 0.01, "grad_norm": 0.06474684178829193, "tokens_per_second": 4358169.329254535, "eta_seconds": 196.88231437918196, "mean_data_wait_seconds": 9.307265281677246e-05}
{"step": 44660, "total_steps": 46238, "tokens": 24146952192, "loss": 2.9234623674190408, "ce_loss": 2.9234623674190408, "z_loss": 43.98231344512015, "muon_lr": 0.006120088820575684, "residual_lr": 3.67205329234541e-05, "adamw_lr": 0.00036720532923454106, "weight_decay": 0.01, "grad_norm": 0.060232024639844894, "tokens_per_second": 4358134.551176283, "eta_seconds": 195.64328131398977, "mean_data_wait_seconds": 0.0003418922424316406}
{"step": 44670, "total_steps": 46238, "tokens": 24152358912, "loss": 2.9365271149259624, "ce_loss": 2.9365271149259624, "z_loss": 43.72526319099195, "muon_lr": 0.006046046696608698, "residual_lr": 3.627628017965219e-05, "adamw_lr": 0.0003627628017965219, "weight_decay": 0.01, "grad_norm": 0.06848631054162979, "tokens_per_second": 4358123.255798842, "eta_seconds": 194.40318097306832, "mean_data_wait_seconds": 0.00015208125114440918}
{"step": 44680, "total_steps": 46238, "tokens": 24157765632, "loss": 2.880439140579917, "ce_loss": 2.880439140579917, "z_loss": 44.4324589064627, "muon_lr": 0.005972393533480633, "residual_lr": 3.5834361200883794e-05, "adamw_lr": 0.000358343612008838, "weight_decay": 0.01, "grad_norm": 0.07177964597940445, "tokens_per_second": 4358108.370104695, "eta_seconds": 193.1632333364342, "mean_data_wait_seconds": 7.343292236328125e-05}
{"step": 44690, "total_steps": 46238, "tokens": 24163172352, "loss": 2.8972024953726567, "ce_loss": 2.8972024953726567, "z_loss": 43.520408456975765, "muon_lr": 0.005899130842654335, "residual_lr": 3.539478505592601e-05, "adamw_lr": 0.0003539478505592601, "weight_decay": 0.01, "grad_norm": 0.06621315330266953, "tokens_per_second": 4358061.140958058, "eta_seconds": 191.9247015924437, "mean_data_wait_seconds": 0.00010317564010620117}
{"step": 44700, "total_steps": 46238, "tokens": 24168579072, "loss": 2.8988770354877818, "ce_loss": 2.8988770354877818, "z_loss": 43.68707073096073, "muon_lr": 0.005826260127579627, "residual_lr": 3.495756076547776e-05, "adamw_lr": 0.00034957560765477765, "weight_decay": 0.01, "grad_norm": 0.0632246881723404, "tokens_per_second": 4357975.610006791, "eta_seconds": 190.68781892487578, "mean_data_wait_seconds": 9.199976921081543e-05}
{"step": 44710, "total_steps": 46238, "tokens": 24173985792, "loss": 2.9353554555864045, "ce_loss": 2.9353554555864045, "z_loss": 43.8825329578284, "muon_lr": 0.0057537828836624425, "residual_lr": 3.452269730197465e-05, "adamw_lr": 0.00034522697301974655, "weight_decay": 0.01, "grad_norm": 0.07692604511976242, "tokens_per_second": 4357898.98516991, "eta_seconds": 189.45050053008754, "mean_data_wait_seconds": 7.405877113342285e-05}
{"step": 44720, "total_steps": 46238, "tokens": 24179392512, "loss": 2.9416035973664485, "ce_loss": 2.9416035973664485, "z_loss": 44.00746380199086, "muon_lr": 0.005681700598234172, "residual_lr": 3.4090203589405026e-05, "adamw_lr": 0.0003409020358940503, "weight_decay": 0.01, "grad_norm": 0.07314436137676239, "tokens_per_second": 4357873.528285917, "eta_seconds": 188.2109287193126, "mean_data_wait_seconds": 0.0001106560230255127}
{"step": 44730, "total_steps": 46238, "tokens": 24184799232, "loss": 2.9289957465547505, "ce_loss": 2.9289957465547505, "z_loss": 44.18155228007924, "muon_lr": 0.0056100147505211174, "residual_lr": 3.3660088503126695e-05, "adamw_lr": 0.000336600885031267, "weight_decay": 0.01, "grad_norm": 0.06685971468687057, "tokens_per_second": 4357944.754131889, "eta_seconds": 186.96719439306162, "mean_data_wait_seconds": 0.00015267729759216309}
{"step": 44740, "total_steps": 46238, "tokens": 24190205952, "loss": 2.8681255109382398, "ce_loss": 2.8681255109382398, "z_loss": 44.611137447935164, "muon_lr": 0.005538726811614142, "residual_lr": 3.3232360869684846e-05, "adamw_lr": 0.0003323236086968485, "weight_decay": 0.01, "grad_norm": 0.06981837004423141, "tokens_per_second": 4357899.9718254665, "eta_seconds": 185.72844471713722, "mean_data_wait_seconds": 0.0027215182781219482}
{"step": 44750, "total_steps": 46238, "tokens": 24195612672, "loss": 2.8938909191073794, "ce_loss": 2.8938909191073794, "z_loss": 44.03493652921735, "muon_lr": 0.005467838244438467, "residual_lr": 3.2807029466630795e-05, "adamw_lr": 0.000328070294666308, "weight_decay": 0.01, "grad_norm": 0.06016810983419418, "tokens_per_second": 4357931.54610365, "eta_seconds": 184.48643708477334, "mean_data_wait_seconds": 0.00015738606452941895}
{"step": 44760, "total_steps": 46238, "tokens": 24201019392, "loss": 2.9158800677819685, "ce_loss": 2.9158800677819685, "z_loss": 44.05556930195201, "muon_lr": 0.005397350503723678, "residual_lr": 3.238410302234206e-05, "adamw_lr": 0.00032384103022342067, "weight_decay": 0.01, "grad_norm": 0.071255162358284, "tokens_per_second": 4357833.445499985, "eta_seconds": 183.2499002054857, "mean_data_wait_seconds": 9.60230827331543e-05}
{"step": 44770, "total_steps": 46238, "tokens": 24206426112, "loss": 2.8955339482336333, "ce_loss": 2.8955339482336333, "z_loss": 43.64616905559193, "muon_lr": 0.0053272650359738605, "residual_lr": 3.196359021584316e-05, "adamw_lr": 0.00031963590215843164, "weight_decay": 0.01, "grad_norm": 0.06922334432601929, "tokens_per_second": 4357870.560174558, "eta_seconds": 182.0076601743373, "mean_data_wait_seconds": 9.97781753540039e-05}
{"step": 44780, "total_steps": 46238, "tokens": 24211832832, "loss": 2.859173449602994, "ce_loss": 2.859173449602994, "z_loss": 44.05659204540831, "muon_lr": 0.005257583279437894, "residual_lr": 3.154549967662736e-05, "adamw_lr": 0.00031545499676627364, "weight_decay": 0.01, "grad_norm": 0.06560295075178146, "tokens_per_second": 4357846.6223625075, "eta_seconds": 180.76797378723126, "mean_data_wait_seconds": 8.615851402282715e-05}
{"step": 44790, "total_steps": 46238, "tokens": 24217239552, "loss": 2.863978245041587, "ce_loss": 2.863978245041587, "z_loss": 44.39600164240057, "muon_lr": 0.005188306664079979, "residual_lr": 3.112983998447987e-05, "adamw_lr": 0.0003112983998447987, "weight_decay": 0.01, "grad_norm": 0.06713129580020905, "tokens_per_second": 4357877.464619118, "eta_seconds": 179.5260170465528, "mean_data_wait_seconds": 7.236003875732422e-05}
{"step": 44800, "total_steps": 46238, "tokens": 24222646272, "loss": 2.8814386815735786, "ce_loss": 2.8814386815735786, "z_loss": 44.06234715201638, "muon_lr": 0.0051194366115502495, "residual_lr": 3.071661966930149e-05, "adamw_lr": 0.00030716619669301495, "weight_decay": 0.01, "grad_norm": 0.07667805254459381, "tokens_per_second": 4357926.8398846015, "eta_seconds": 178.28331969441084, "mean_data_wait_seconds": 9.018182754516602e-05}
{"step": 44810, "total_steps": 46238, "tokens": 24228052992, "loss": 2.8261040846506753, "ce_loss": 2.8261040846506753, "z_loss": 43.64597664457379, "muon_lr": 0.005050974535155642, "residual_lr": 3.030584721093385e-05, "adamw_lr": 0.00030305847210933853, "weight_decay": 0.01, "grad_norm": 0.06698603183031082, "tokens_per_second": 4358034.652106981, "eta_seconds": 177.03827656050962, "mean_data_wait_seconds": 8.639693260192871e-05}
{"step": 44820, "total_steps": 46238, "tokens": 24233459712, "loss": 2.9432519204688794, "ce_loss": 2.9432519204688794, "z_loss": 43.77777489748868, "muon_lr": 0.00498292183983084, "residual_lr": 2.9897531038985035e-05, "adamw_lr": 0.00029897531038985037, "weight_decay": 0.01, "grad_norm": 0.06370395421981812, "tokens_per_second": 4358038.022686879, "eta_seconds": 175.79750796383675, "mean_data_wait_seconds": 6.67870044708252e-05}
{"step": 44830, "total_steps": 46238, "tokens": 24238866432, "loss": 2.945727198412924, "ce_loss": 2.945727198412924, "z_loss": 43.12360619053696, "muon_lr": 0.004915279922109492, "residual_lr": 2.9491679532656947e-05, "adamw_lr": 0.0002949167953265695, "weight_decay": 0.01, "grad_norm": 0.059626493602991104, "tokens_per_second": 4358022.534237759, "eta_seconds": 174.557496668166, "mean_data_wait_seconds": 6.464123725891113e-05}
{"step": 44840, "total_steps": 46238, "tokens": 24244273152, "loss": 2.883760358348037, "ce_loss": 2.883760358348037, "z_loss": 44.033271153767906, "muon_lr": 0.0048480501700955275, "residual_lr": 2.9088301020573163e-05, "adamw_lr": 0.00029088301020573163, "weight_decay": 0.01, "grad_norm": 0.062399107962846756, "tokens_per_second": 4358037.763964242, "eta_seconds": 173.316254908478, "mean_data_wait_seconds": 6.395578384399414e-05}
{"step": 44850, "total_steps": 46238, "tokens": 24249679872, "loss": 2.9542138215267295, "ce_loss": 2.9542138215267295, "z_loss": 44.18555875258012, "muon_lr": 0.004781233963434684, "residual_lr": 2.86874037806081e-05, "adamw_lr": 0.000286874037806081, "weight_decay": 0.01, "grad_norm": 0.062330473214387894, "tokens_per_second": 4358112.886647898, "eta_seconds": 172.0726570203199, "mean_data_wait_seconds": 8.678436279296875e-05}
{"step": 44860, "total_steps": 46238, "tokens": 24255086592, "loss": 2.8592705419569304, "ce_loss": 2.8592705419569304, "z_loss": 44.0378364794182, "muon_lr": 0.0047148326732861685, "residual_lr": 2.8288996039717004e-05, "adamw_lr": 0.0002828899603971701, "weight_decay": 0.01, "grad_norm": 0.06527531892061234, "tokens_per_second": 4358176.028033749, "eta_seconds": 170.82957163983434, "mean_data_wait_seconds": 6.833672523498535e-05}
{"step": 44870, "total_steps": 46238, "tokens": 24260493312, "loss": 2.908667940081972, "ce_loss": 2.908667940081972, "z_loss": 43.14366912841797, "muon_lr": 0.004648847662294553, "residual_lr": 2.7893085973767316e-05, "adamw_lr": 0.0002789308597376732, "weight_decay": 0.01, "grad_norm": 0.06051008030772209, "tokens_per_second": 4358222.699111884, "eta_seconds": 169.58716316874145, "mean_data_wait_seconds": 5.894899368286133e-05}
{"step": 44880, "total_steps": 46238, "tokens": 24265900032, "loss": 2.9240027196479565, "ce_loss": 2.9240027196479565, "z_loss": 43.76313663251472, "muon_lr": 0.004583280284561797, "residual_lr": 2.749968170737078e-05, "adamw_lr": 0.00027499681707370785, "weight_decay": 0.01, "grad_norm": 0.06410238891839981, "tokens_per_second": 4358302.204255562, "eta_seconds": 168.34351305047267, "mean_data_wait_seconds": 9.34898853302002e-05}
{"step": 44890, "total_steps": 46238, "tokens": 24271306752, "loss": 2.8834564432953345, "ce_loss": 2.8834564432953345, "z_loss": 43.929756193450004, "muon_lr": 0.004518131885619456, "residual_lr": 2.7108791313716734e-05, "adamw_lr": 0.0002710879131371674, "weight_decay": 0.01, "grad_norm": 0.06015048548579216, "tokens_per_second": 4358391.987893748, "eta_seconds": 167.09951422977758, "mean_data_wait_seconds": 6.219744682312012e-05}
{"step": 44900, "total_steps": 46238, "tokens": 24276713472, "loss": 2.966399792468909, "ce_loss": 2.966399792468909, "z_loss": 43.96443598198168, "muon_lr": 0.004453403802401063, "residual_lr": 2.6720422814406372e-05, "adamw_lr": 0.00026720422814406375, "weight_decay": 0.01, "grad_norm": 0.059188976883888245, "tokens_per_second": 4358389.605195804, "eta_seconds": 165.85907398875696, "mean_data_wait_seconds": 7.152557373046875e-05}
{"step": 44910, "total_steps": 46238, "tokens": 24282120192, "loss": 2.889390513752446, "ce_loss": 2.889390513752446, "z_loss": 43.878509897174254, "muon_lr": 0.004389097363214714, "residual_lr": 2.6334584179288282e-05, "adamw_lr": 0.00026334584179288286, "weight_decay": 0.01, "grad_norm": 0.0681362971663475, "tokens_per_second": 4358407.295812064, "eta_seconds": 164.61787421506227, "mean_data_wait_seconds": 0.00011688470840454102}
{"step": 44920, "total_steps": 46238, "tokens": 24287526912, "loss": 2.9094959515513796, "ce_loss": 2.9094959515513796, "z_loss": 42.93213457049745, "muon_lr": 0.004325213887715793, "residual_lr": 2.5951283326294754e-05, "adamw_lr": 0.00025951283326294757, "weight_decay": 0.01, "grad_norm": 0.05964013934135437, "tokens_per_second": 4358354.266298847, "eta_seconds": 163.37933552260125, "mean_data_wait_seconds": 8.213520050048828e-05}
{"step": 44930, "total_steps": 46238, "tokens": 24292933632, "loss": 2.9098858417886677, "ce_loss": 2.9098858417886677, "z_loss": 43.99926136479233, "muon_lr": 0.004261754686879904, "residual_lr": 2.5570528121279423e-05, "adamw_lr": 0.0002557052812127942, "weight_decay": 0.01, "grad_norm": 0.06032320857048035, "tokens_per_second": 4358366.050993707, "eta_seconds": 162.1383554598132, "mean_data_wait_seconds": 0.00021073222160339355}
{"step": 44940, "total_steps": 46238, "tokens": 24298340352, "loss": 2.9083700089743645, "ce_loss": 2.9083700089743645, "z_loss": 44.824317238547586, "muon_lr": 0.004198721062975939, "residual_lr": 2.5192326377855625e-05, "adamw_lr": 0.00025192326377855627, "weight_decay": 0.01, "grad_norm": 0.0692044198513031, "tokens_per_second": 4358370.688208744, "eta_seconds": 160.89764597058834, "mean_data_wait_seconds": 7.393956184387207e-05}
{"step": 44950, "total_steps": 46238, "tokens": 24303747072, "loss": 2.8638637499375776, "ce_loss": 2.8638637499375776, "z_loss": 44.92272353894783, "muon_lr": 0.004136114309539388, "residual_lr": 2.4816685857236324e-05, "adamw_lr": 0.00024816685857236324, "weight_decay": 0.01, "grad_norm": 0.06043270230293274, "tokens_per_second": 4358404.952501854, "eta_seconds": 159.65585382344162, "mean_data_wait_seconds": 6.416440010070801e-05}
{"step": 44960, "total_steps": 46238, "tokens": 24309153792, "loss": 2.8111231471552993, "ce_loss": 2.8111231471552993, "z_loss": 44.459605448173754, "muon_lr": 0.004073935711345786, "residual_lr": 2.4443614268074712e-05, "adamw_lr": 0.00024443614268074716, "weight_decay": 0.01, "grad_norm": 0.06335291266441345, "tokens_per_second": 4358424.080429637, "eta_seconds": 158.41463135729077, "mean_data_wait_seconds": 6.80387020111084e-05}
{"step": 44970, "total_steps": 46238, "tokens": 24314560512, "loss": 2.873513888229023, "ce_loss": 2.873513888229023, "z_loss": 44.4519262602835, "muon_lr": 0.004012186544384322, "residual_lr": 2.407311926630593e-05, "adamw_lr": 0.00024073119266305933, "weight_decay": 0.01, "grad_norm": 0.06258811056613922, "tokens_per_second": 4358472.231076941, "eta_seconds": 157.1723731805754, "mean_data_wait_seconds": 6.42538070678711e-05}
{"step": 44980, "total_steps": 46238, "tokens": 24319967232, "loss": 2.828762366916194, "ce_loss": 2.828762366916194, "z_loss": 44.41500316966664, "muon_lr": 0.003950868075831693, "residual_lr": 2.3705208454990154e-05, "adamw_lr": 0.00023705208454990158, "weight_decay": 0.01, "grad_norm": 0.07037448137998581, "tokens_per_second": 4358443.122109098, "eta_seconds": 155.93290653546998, "mean_data_wait_seconds": 5.6415796279907227e-05}
{"step": 44990, "total_steps": 46238, "tokens": 24325373952, "loss": 2.848743057612217, "ce_loss": 2.848743057612217, "z_loss": 44.29514457240249, "muon_lr": 0.0038899815640260657, "residual_lr": 2.333988938415639e-05, "adamw_lr": 0.00023339889384156392, "weight_decay": 0.01, "grad_norm": 0.06109324097633362, "tokens_per_second": 4358542.185779637, "eta_seconds": 154.6888742294917, "mean_data_wait_seconds": 6.16312026977539e-05}
{"step": 45000, "total_steps": 46238, "tokens": 24330780672, "loss": 2.8148418412064062, "ce_loss": 2.8148418412064062, "z_loss": 44.530117786291875, "muon_lr": 0.003829528258441284, "residual_lr": 2.29771695506477e-05, "adamw_lr": 0.00022977169550647704, "weight_decay": 0.01, "grad_norm": 0.0636911392211914, "tokens_per_second": 4358546.543865231, "eta_seconds": 153.448232631901, "mean_data_wait_seconds": 5.8978796005249023e-05}
{"step": 45010, "total_steps": 46238, "tokens": 24336187392, "loss": 2.79793950102546, "ce_loss": 2.79793950102546, "z_loss": 44.597156726952754, "muon_lr": 0.003769509399661192, "residual_lr": 2.2617056397967148e-05, "adamw_lr": 0.00022617056397967151, "weight_decay": 0.01, "grad_norm": 0.077520452439785, "tokens_per_second": 4358569.041816465, "eta_seconds": 152.20696004474016, "mean_data_wait_seconds": 7.838010787963867e-05}
{"step": 45020, "total_steps": 46238, "tokens": 24341594112, "loss": 2.7301086115114614, "ce_loss": 2.7301086115114614, "z_loss": 44.34268962975705, "muon_lr": 0.0037099262193542124, "residual_lr": 2.225955731612527e-05, "adamw_lr": 0.00022259557316125272, "weight_decay": 0.01, "grad_norm": 0.06586119532585144, "tokens_per_second": 4358679.724635933, "eta_seconds": 150.96264593172432, "mean_data_wait_seconds": 6.955862045288086e-05}
{"step": 45030, "total_steps": 46238, "tokens": 24347000832, "loss": 2.829498970147335, "ce_loss": 2.829498970147335, "z_loss": 45.01837808435614, "muon_lr": 0.0036507799402480477, "residual_lr": 2.1904679641488283e-05, "adamw_lr": 0.00021904679641488283, "weight_decay": 0.01, "grad_norm": 0.06472299993038177, "tokens_per_second": 4358796.295712807, "eta_seconds": 149.71819275928786, "mean_data_wait_seconds": 6.002187728881836e-05}
{"step": 45040, "total_steps": 46238, "tokens": 24352407552, "loss": 2.8209600430546384, "ce_loss": 2.8209600430546384, "z_loss": 44.800209045410156, "muon_lr": 0.0035920717761046025, "residual_lr": 2.155243065662761e-05, "adamw_lr": 0.00021552430656627614, "weight_decay": 0.01, "grad_norm": 0.06748241931200027, "tokens_per_second": 4358896.159520008, "eta_seconds": 148.47437523523539, "mean_data_wait_seconds": 9.071826934814453e-05}
{"step": 45050, "total_steps": 46238, "tokens": 24357814272, "loss": 2.826849240245241, "ce_loss": 2.826849240245241, "z_loss": 44.986689740961246, "muon_lr": 0.0035338029316950488, "residual_lr": 2.120281759017029e-05, "adamw_lr": 0.00021202817590170294, "weight_decay": 0.01, "grad_norm": 0.06381949037313461, "tokens_per_second": 4359001.779568062, "eta_seconds": 147.2304202783772, "mean_data_wait_seconds": 7.027387619018555e-05}
{"step": 45060, "total_steps": 46238, "tokens": 24363220992, "loss": 2.8764482014106982, "ce_loss": 2.8764482014106982, "z_loss": 43.49881305116595, "muon_lr": 0.003475974602775142, "residual_lr": 2.085584761665085e-05, "adamw_lr": 0.00020855847616650853, "weight_decay": 0.01, "grad_norm": 0.06625930964946747, "tokens_per_second": 4359034.392680839, "eta_seconds": 145.9889706464617, "mean_data_wait_seconds": 8.982419967651367e-05}
{"step": 45070, "total_steps": 46238, "tokens": 24368627712, "loss": 2.7692465402863244, "ce_loss": 2.7692465402863244, "z_loss": 44.89421104662346, "muon_lr": 0.0034185879760606527, "residual_lr": 2.0511527856363912e-05, "adamw_lr": 0.00020511527856363916, "weight_decay": 0.01, "grad_norm": 0.06496027857065201, "tokens_per_second": 4359067.2018845165, "eta_seconds": 144.74753307937553, "mean_data_wait_seconds": 5.823373794555664e-05}
{"step": 45080, "total_steps": 46238, "tokens": 24374034432, "loss": 2.8505253755685054, "ce_loss": 2.8505253755685054, "z_loss": 44.97303662155614, "muon_lr": 0.0033616442292030285, "residual_lr": 2.016986537521817e-05, "adamw_lr": 0.00020169865375218172, "weight_decay": 0.01, "grad_norm": 0.060497358441352844, "tokens_per_second": 4359132.259244339, "eta_seconds": 143.5050525648518, "mean_data_wait_seconds": 6.201863288879395e-05}
{"step": 45090, "total_steps": 46238, "tokens": 24379441152, "loss": 2.808548674438939, "ce_loss": 2.808548674438939, "z_loss": 44.939114079330906, "muon_lr": 0.003305144530765211, "residual_lr": 1.9830867184591264e-05, "adamw_lr": 0.00019830867184591266, "weight_decay": 0.01, "grad_norm": 0.062233295291662216, "tokens_per_second": 4359147.597574491, "eta_seconds": 142.2642317376597, "mean_data_wait_seconds": 7.325410842895508e-05}
{"step": 45100, "total_steps": 46238, "tokens": 24384847872, "loss": 2.8151653950864617, "ce_loss": 2.8151653950864617, "z_loss": 45.200738849061906, "muon_lr": 0.0032490900401976747, "residual_lr": 1.9494540241186046e-05, "adamw_lr": 0.00019494540241186048, "weight_decay": 0.01, "grad_norm": 0.05857604742050171, "tokens_per_second": 4359198.588498463, "eta_seconds": 141.0222662537038, "mean_data_wait_seconds": 0.00011879205703735352}
{"step": 45110, "total_steps": 46238, "tokens": 24390254592, "loss": 2.8335152214223687, "ce_loss": 2.8335152214223687, "z_loss": 44.93386259945956, "muon_lr": 0.003193481907814622, "residual_lr": 1.9160891446887726e-05, "adamw_lr": 0.0001916089144688773, "weight_decay": 0.01, "grad_norm": 0.06390456855297089, "tokens_per_second": 4359174.944806015, "eta_seconds": 139.78272304166856, "mean_data_wait_seconds": 7.519125938415527e-05}
{"step": 45120, "total_steps": 46238, "tokens": 24395661312, "loss": 2.8146558476216867, "ce_loss": 2.8146558476216867, "z_loss": 45.51169776916504, "muon_lr": 0.0031383212747703865, "residual_lr": 1.8829927648622318e-05, "adamw_lr": 0.0001882992764862232, "weight_decay": 0.01, "grad_norm": 0.06444278359413147, "tokens_per_second": 4359127.93649623, "eta_seconds": 138.54390896483437, "mean_data_wait_seconds": 7.736682891845703e-05}
{"step": 45130, "total_steps": 46238, "tokens": 24401068032, "loss": 2.877218793738972, "ce_loss": 2.877218793738972, "z_loss": 45.23504225413004, "muon_lr": 0.0030836092730359894, "residual_lr": 1.8501655638215934e-05, "adamw_lr": 0.00018501655638215937, "weight_decay": 0.01, "grad_norm": 0.06326630711555481, "tokens_per_second": 4359246.7556788735, "eta_seconds": 137.29984502948625, "mean_data_wait_seconds": 6.130337715148926e-05}
{"step": 45140, "total_steps": 46238, "tokens": 24406474752, "loss": 2.8712229186838325, "ce_loss": 2.8712229186838325, "z_loss": 44.814963803146824, "muon_lr": 0.003029347025375948, "residual_lr": 1.8176082152255684e-05, "adamw_lr": 0.00018176082152255684, "weight_decay": 0.01, "grad_norm": 0.062385644763708115, "tokens_per_second": 4359014.5087905675, "eta_seconds": 136.06680656921318, "mean_data_wait_seconds": 7.12275505065918e-05}
{"step": 45150, "total_steps": 46238, "tokens": 24411881472, "loss": 2.878064699245222, "ce_loss": 2.878064699245222, "z_loss": 45.346966309980914, "muon_lr": 0.0029755356453252087, "residual_lr": 1.785321387195125e-05, "adamw_lr": 0.0001785321387195125, "weight_decay": 0.01, "grad_norm": 0.07223938405513763, "tokens_per_second": 4358925.1406970965, "eta_seconds": 134.82921707299866, "mean_data_wait_seconds": 8.168816566467285e-05}
{"step": 45160, "total_steps": 46238, "tokens": 24417288192, "loss": 2.8432955434828093, "ce_loss": 2.8432955434828093, "z_loss": 45.07661293492173, "muon_lr": 0.002922176237166299, "residual_lr": 1.7533057422997794e-05, "adamw_lr": 0.00017533057422997796, "weight_decay": 0.01, "grad_norm": 0.0700899288058281, "tokens_per_second": 4358919.485982402, "eta_seconds": 133.58901119247489, "mean_data_wait_seconds": 8.702278137207031e-05}
{"step": 45170, "total_steps": 46238, "tokens": 24422694912, "loss": 2.8318412087180396, "ce_loss": 2.8318412087180396, "z_loss": 45.72957715121183, "muon_lr": 0.0028692698959066733, "residual_lr": 1.7215619375440036e-05, "adamw_lr": 0.0001721561937544004, "weight_decay": 0.01, "grad_norm": 0.068634532392025, "tokens_per_second": 4358930.965022548, "eta_seconds": 132.34828186755092, "mean_data_wait_seconds": 0.00016543269157409668}
{"step": 45180, "total_steps": 46238, "tokens": 24428101632, "loss": 2.785345010685198, "ce_loss": 2.785345010685198, "z_loss": 44.55796967130719, "muon_lr": 0.002816817707256242, "residual_lr": 1.690090624353745e-05, "adamw_lr": 0.00016900906243537452, "weight_decay": 0.01, "grad_norm": 0.0801551416516304, "tokens_per_second": 4358879.628600882, "eta_seconds": 131.10944845784547, "mean_data_wait_seconds": 7.542967796325684e-05}
{"step": 45190, "total_steps": 46238, "tokens": 24433508352, "loss": 2.8591692917274707, "ce_loss": 2.8591692917274707, "z_loss": 45.002674131682426, "muon_lr": 0.0027648207476050875, "residual_lr": 1.6588924485630523e-05, "adamw_lr": 0.00016588924485630525, "weight_decay": 0.01, "grad_norm": 0.06897249817848206, "tokens_per_second": 4358889.9952571625, "eta_seconds": 129.86874746000618, "mean_data_wait_seconds": 9.122490882873535e-05}
{"step": 45200, "total_steps": 46238, "tokens": 24438915072, "loss": 2.8868726780920317, "ce_loss": 2.8868726780920317, "z_loss": 45.20051375302401, "muon_lr": 0.002713280084001363, "residual_lr": 1.6279680504008176e-05, "adamw_lr": 0.00016279680504008176, "weight_decay": 0.01, "grad_norm": 0.06504840403795242, "tokens_per_second": 4358300.073208041, "eta_seconds": 128.64576889660998, "mean_data_wait_seconds": 0.00010073184967041016}
{"step": 45210, "total_steps": 46238, "tokens": 24444321792, "loss": 2.8686018784840903, "ce_loss": 2.8686018784840903, "z_loss": 44.569129423661664, "muon_lr": 0.0026621967741294245, "residual_lr": 1.5973180644776547e-05, "adamw_lr": 0.00015973180644776548, "weight_decay": 0.01, "grad_norm": 0.06940539181232452, "tokens_per_second": 4357467.782748024, "eta_seconds": 127.42954662761053, "mean_data_wait_seconds": 0.00012090802192687988}
{"step": 45220, "total_steps": 46238, "tokens": 24449728512, "loss": 2.850764375744444, "ce_loss": 2.850764375744444, "z_loss": 45.2147159576416, "muon_lr": 0.0026115718662880955, "residual_lr": 1.566943119772857e-05, "adamw_lr": 0.00015669431197728572, "weight_decay": 0.01, "grad_norm": 0.06358088552951813, "tokens_per_second": 4357096.714860314, "eta_seconds": 126.19949934198979, "mean_data_wait_seconds": 0.0001575946807861328}
{"step": 45230, "total_steps": 46238, "tokens": 24455135232, "loss": 2.8466709527102383, "ce_loss": 2.8466709527102383, "z_loss": 45.20720424074115, "muon_lr": 0.0025614063993691875, "residual_lr": 1.5368438396215122e-05, "adamw_lr": 0.00015368438396215123, "weight_decay": 0.01, "grad_norm": 0.06315295398235321, "tokens_per_second": 4357089.247840991, "eta_seconds": 124.9588137928979, "mean_data_wait_seconds": 8.907914161682129e-05}
{"step": 45240, "total_steps": 46238, "tokens": 24460541952, "loss": 2.882068420901443, "ce_loss": 2.882068420901443, "z_loss": 45.079365730285645, "muon_lr": 0.002511701402836128, "residual_lr": 1.5070208417016767e-05, "adamw_lr": 0.00015070208417016767, "weight_decay": 0.01, "grad_norm": 0.058953907340765, "tokens_per_second": 4357073.881292413, "eta_seconds": 123.71834829665656, "mean_data_wait_seconds": 0.00010439753532409668}
{"step": 45250, "total_steps": 46238, "tokens": 24465948672, "loss": 2.9448813488989165, "ce_loss": 2.9448813488989165, "z_loss": 44.72630284049294, "muon_lr": 0.0024624578967028994, "residual_lr": 1.4774747380217394e-05, "adamw_lr": 0.00014774747380217395, "weight_decay": 0.01, "grad_norm": 0.08172480016946793, "tokens_per_second": 4357062.678346704, "eta_seconds": 122.4777570109439, "mean_data_wait_seconds": 9.244680404663086e-05}
{"step": 45260, "total_steps": 46238, "tokens": 24471355392, "loss": 2.8809220899235117, "ce_loss": 2.8809220899235117, "z_loss": 44.94321505228678, "muon_lr": 0.0024136768915130587, "residual_lr": 1.448206134907835e-05, "adamw_lr": 0.00014482061349078352, "weight_decay": 0.01, "grad_norm": 0.061263442039489746, "tokens_per_second": 4357064.006747015, "eta_seconds": 121.23681065552708, "mean_data_wait_seconds": 0.00011670589447021484}
{"step": 45270, "total_steps": 46238, "tokens": 24476762112, "loss": 2.857015232245127, "ce_loss": 2.857015232245127, "z_loss": 44.65977596514153, "muon_lr": 0.0023653593883190245, "residual_lr": 1.4192156329914145e-05, "adamw_lr": 0.00014192156329914147, "weight_decay": 0.01, "grad_norm": 0.06690295785665512, "tokens_per_second": 4357056.799414452, "eta_seconds": 119.99610013582183, "mean_data_wait_seconds": 0.00011178851127624512}
{"step": 45280, "total_steps": 46238, "tokens": 24482168832, "loss": 2.8942964167305916, "ce_loss": 2.8942964167305916, "z_loss": 45.11509262431752, "muon_lr": 0.00231750637866151, "residual_lr": 1.390503827196906e-05, "adamw_lr": 0.00013905038271969062, "weight_decay": 0.01, "grad_norm": 0.06001053377985954, "tokens_per_second": 4356942.344911206, "eta_seconds": 118.75830870342745, "mean_data_wait_seconds": 9.384751319885254e-05}
{"step": 45290, "total_steps": 46238, "tokens": 24487575552, "loss": 2.865162092627901, "ce_loss": 2.865162092627901, "z_loss": 44.584797974788785, "muon_lr": 0.0022701188445492073, "residual_lr": 1.3620713067295242e-05, "adamw_lr": 0.00013620713067295242, "weight_decay": 0.01, "grad_norm": 0.06667324900627136, "tokens_per_second": 4356863.582737409, "eta_seconds": 117.51948948520925, "mean_data_wait_seconds": 8.821487426757812e-05}
{"step": 45300, "total_steps": 46238, "tokens": 24492982272, "loss": 2.880167056213726, "ce_loss": 2.880167056213726, "z_loss": 44.99849715377345, "muon_lr": 0.0022231977584386137, "residual_lr": 1.3339186550631681e-05, "adamw_lr": 0.00013339186550631683, "weight_decay": 0.01, "grad_norm": 0.06523837894201279, "tokens_per_second": 4356808.632069522, "eta_seconds": 116.27998996121985, "mean_data_wait_seconds": 8.806586265563965e-05}
{"step": 45310, "total_steps": 46238, "tokens": 24498388992, "loss": 2.865583311427723, "ce_loss": 2.865583311427723, "z_loss": 44.97010996847442, "muon_lr": 0.0021767440832140827, "residual_lr": 1.3060464499284495e-05, "adamw_lr": 0.00013060464499284498, "weight_decay": 0.01, "grad_norm": 0.062136322259902954, "tokens_per_second": 4356747.339352839, "eta_seconds": 115.04062663281499, "mean_data_wait_seconds": 7.641315460205078e-05}
{"step": 45320, "total_steps": 46238, "tokens": 24503795712, "loss": 2.88047802809513, "ce_loss": 2.88047802809513, "z_loss": 44.87865606943766, "muon_lr": 0.002130758772168054, "residual_lr": 1.2784552633008321e-05, "adamw_lr": 0.00012784552633008322, "weight_decay": 0.01, "grad_norm": 0.06111311912536621, "tokens_per_second": 4356821.232855466, "eta_seconds": 113.7976973351864, "mean_data_wait_seconds": 0.00011155009269714355}
{"step": 45330, "total_steps": 46238, "tokens": 24509202432, "loss": 2.9523348248366155, "ce_loss": 2.9523348248366155, "z_loss": 44.18121017109264, "muon_lr": 0.002085242768981502, "residual_lr": 1.2511456613889009e-05, "adamw_lr": 0.00012511456613889012, "weight_decay": 0.01, "grad_norm": 0.0670076459646225, "tokens_per_second": 4356897.980231899, "eta_seconds": 112.55473647191039, "mean_data_wait_seconds": 6.660819053649902e-05}
{"step": 45340, "total_steps": 46238, "tokens": 24514609152, "loss": 2.8863973147941357, "ce_loss": 2.8863973147941357, "z_loss": 45.213907877604164, "muon_lr": 0.002040197007704575, "residual_lr": 1.224118204622745e-05, "adamw_lr": 0.0001224118204622745, "weight_decay": 0.01, "grad_norm": 0.05923696979880333, "tokens_per_second": 4356920.175089573, "eta_seconds": 111.3132131207865, "mean_data_wait_seconds": 0.00012493133544921875}
{"step": 45350, "total_steps": 46238, "tokens": 24520015872, "loss": 2.8940659996235008, "ce_loss": 2.8940659996235008, "z_loss": 44.36521874052106, "muon_lr": 0.0019956224127374135, "residual_lr": 1.1973734476424479e-05, "adamw_lr": 0.0001197373447642448, "weight_decay": 0.01, "grad_norm": 0.05577755719423294, "tokens_per_second": 4357000.74562823, "eta_seconds": 110.07022766319278, "mean_data_wait_seconds": 5.8203935623168945e-05}
{"step": 45360, "total_steps": 46238, "tokens": 24525422592, "loss": 2.8868918527256358, "ce_loss": 2.8868918527256358, "z_loss": 44.394014820908055, "muon_lr": 0.0019515198988111827, "residual_lr": 1.1709119392867094e-05, "adamw_lr": 0.00011709119392867095, "weight_decay": 0.01, "grad_norm": 0.05877229571342468, "tokens_per_second": 4357112.444435348, "eta_seconds": 108.82651068727448, "mean_data_wait_seconds": 0.00013759732246398926}
{"step": 45370, "total_steps": 46238, "tokens": 24530829312, "loss": 2.8907393740885183, "ce_loss": 2.8907393740885183, "z_loss": 44.85115346041593, "muon_lr": 0.0019078903709693062, "residual_lr": 1.1447342225815836e-05, "adamw_lr": 0.00011447342225815837, "weight_decay": 0.01, "grad_norm": 0.06983163952827454, "tokens_per_second": 4357059.378186908, "eta_seconds": 107.58692579375979, "mean_data_wait_seconds": 7.972121238708496e-05}
{"step": 45380, "total_steps": 46238, "tokens": 24536236032, "loss": 2.927717909668431, "ce_loss": 2.927717909668431, "z_loss": 44.262122501026504, "muon_lr": 0.0018647347245488999, "residual_lr": 1.1188408347293398e-05, "adamw_lr": 0.000111884083472934, "weight_decay": 0.01, "grad_norm": 0.0637325569987297, "tokens_per_second": 4357113.856300734, "eta_seconds": 106.34468579010172, "mean_data_wait_seconds": 6.240606307983398e-05}
{"step": 45390, "total_steps": 46238, "tokens": 24541642752, "loss": 2.9171893813393335, "ce_loss": 2.9171893813393335, "z_loss": 44.54098149501916, "muon_lr": 0.0018220538451623747, "residual_lr": 1.0932323070974248e-05, "adamw_lr": 0.00010932323070974249, "weight_decay": 0.01, "grad_norm": 0.060320619493722916, "tokens_per_second": 4357212.260122447, "eta_seconds": 105.10141729637257, "mean_data_wait_seconds": 0.00013148784637451172}
{"step": 45400, "total_steps": 46238, "tokens": 24547049472, "loss": 2.928933165290139, "ce_loss": 2.928933165290139, "z_loss": 44.49494150913123, "muon_lr": 0.001779848608679288, "residual_lr": 1.0679091652075727e-05, "adamw_lr": 0.00010679091652075729, "weight_decay": 0.01, "grad_norm": 0.058466535061597824, "tokens_per_second": 4357300.007819566, "eta_seconds": 103.85845895115598, "mean_data_wait_seconds": 0.0001881718635559082}
{"step": 45410, "total_steps": 46238, "tokens": 24552456192, "loss": 2.9277579657959216, "ce_loss": 2.9277579657959216, "z_loss": 44.248848250418, "muon_lr": 0.001738119881208361, "residual_lr": 1.0428719287250164e-05, "adamw_lr": 0.00010428719287250166, "weight_decay": 0.01, "grad_norm": 0.061445631086826324, "tokens_per_second": 4357359.321783288, "eta_seconds": 102.61622027925979, "mean_data_wait_seconds": 8.437037467956543e-05}
{"step": 45420, "total_steps": 46238, "tokens": 24557862912, "loss": 2.9114364602349023, "ce_loss": 2.9114364602349023, "z_loss": 44.96549250862815, "muon_lr": 0.0016968685190797, "residual_lr": 1.0181211114478199e-05, "adamw_lr": 0.00010181211114478201, "weight_decay": 0.01, "grad_norm": 0.0722062885761261, "tokens_per_second": 4357378.685911935, "eta_seconds": 101.37494485576313, "mean_data_wait_seconds": 0.00016388297080993652}
{"step": 45430, "total_steps": 46238, "tokens": 24563269632, "loss": 2.934162029714295, "ce_loss": 2.934162029714295, "z_loss": 44.51054526820327, "muon_lr": 0.0016560953688272207, "residual_lr": 9.936572212963323e-06, "adamw_lr": 9.936572212963324e-05, "weight_decay": 0.01, "grad_norm": 0.055657561868429184, "tokens_per_second": 4357507.356050777, "eta_seconds": 100.13116865864347, "mean_data_wait_seconds": 9.334087371826172e-05}
{"step": 45440, "total_steps": 46238, "tokens": 24568676352, "loss": 2.9015632741379016, "ce_loss": 2.9015632741379016, "z_loss": 44.61905725074537, "muon_lr": 0.0016158012671712936, "residual_lr": 9.69480760302776e-06, "adamw_lr": 9.694807603027761e-05, "weight_decay": 0.01, "grad_norm": 0.061607472598552704, "tokens_per_second": 4357547.271342537, "eta_seconds": 98.88948006001486, "mean_data_wait_seconds": 9.372830390930176e-05}
{"step": 45450, "total_steps": 46238, "tokens": 24574083072, "loss": 2.902810826446071, "ce_loss": 2.902810826446071, "z_loss": 43.86749501661821, "muon_lr": 0.0015759870410015531, "residual_lr": 9.455922246009317e-06, "adamw_lr": 9.455922246009318e-05, "weight_decay": 0.01, "grad_norm": 0.06675766408443451, "tokens_per_second": 4357622.776701491, "eta_seconds": 97.64701668878496, "mean_data_wait_seconds": 9.15229320526123e-05}
{"step": 45460, "total_steps": 46238, "tokens": 24579489792, "loss": 2.8926025556795523, "ce_loss": 2.8926025556795523, "z_loss": 44.67040365392511, "muon_lr": 0.0015366535073599475, "residual_lr": 9.219921044159684e-06, "adamw_lr": 9.219921044159685e-05, "weight_decay": 0.01, "grad_norm": 0.06338609755039215, "tokens_per_second": 4357592.827074989, "eta_seconds": 96.40692939225148, "mean_data_wait_seconds": 0.00011172890663146973}
{"step": 45470, "total_steps": 46238, "tokens": 24584896512, "loss": 2.876551310221354, "ce_loss": 2.876551310221354, "z_loss": 44.746921279213645, "muon_lr": 0.0014978014734239487, "residual_lr": 8.986808840543691e-06, "adamw_lr": 8.986808840543692e-05, "weight_decay": 0.01, "grad_norm": 0.058332741260528564, "tokens_per_second": 4357666.650711288, "eta_seconds": 95.16455875125524, "mean_data_wait_seconds": 0.00010704994201660156}
{"step": 45480, "total_steps": 46238, "tokens": 24590303232, "loss": 2.885583142439524, "ce_loss": 2.885583142439524, "z_loss": 44.06983413118304, "muon_lr": 0.0014594317364900078, "residual_lr": 8.756590418940046e-06, "adamw_lr": 8.756590418940047e-05, "weight_decay": 0.01, "grad_norm": 0.05753554031252861, "tokens_per_second": 4357675.9968267465, "eta_seconds": 93.92361990612507, "mean_data_wait_seconds": 0.00015038251876831055}
{"step": 45490, "total_steps": 46238, "tokens": 24595709952, "loss": 2.846922392194921, "ce_loss": 2.846922392194921, "z_loss": 44.74491223421964, "muon_lr": 0.0014215450839571932, "residual_lr": 8.529270503743158e-06, "adamw_lr": 8.529270503743159e-05, "weight_decay": 0.01, "grad_norm": 0.06317934393882751, "tokens_per_second": 4357701.285815112, "eta_seconds": 92.68234729963908, "mean_data_wait_seconds": 0.00017186999320983887}
{"step": 45500, "total_steps": 46238, "tokens": 24601116672, "loss": 2.9711504018667974, "ce_loss": 2.9711504018667974, "z_loss": 44.298184279239535, "muon_lr": 0.0013841422933110192, "residual_lr": 8.304853759866115e-06, "adamw_lr": 8.304853759866115e-05, "weight_decay": 0.01, "grad_norm": 0.05815594270825386, "tokens_per_second": 4357799.9777748715, "eta_seconds": 91.43954886232864, "mean_data_wait_seconds": 9.79304313659668e-05}
{"step": 45510, "total_steps": 46238, "tokens": 24606523392, "loss": 2.949210667248928, "ce_loss": 2.949210667248928, "z_loss": 44.53784517808394, "muon_lr": 0.0013472241321074942, "residual_lr": 8.083344792644963e-06, "adamw_lr": 8.083344792644964e-05, "weight_decay": 0.01, "grad_norm": 0.05931288003921509, "tokens_per_second": 4357877.163279078, "eta_seconds": 90.19725184365595, "mean_data_wait_seconds": 6.669759750366211e-05}
{"step": 45520, "total_steps": 46238, "tokens": 24611930112, "loss": 2.930785997347398, "ce_loss": 2.930785997347398, "z_loss": 44.63047221212676, "muon_lr": 0.0013107913579573805, "residual_lr": 7.864748147744282e-06, "adamw_lr": 7.864748147744283e-05, "weight_decay": 0.01, "grad_norm": 0.05939493700861931, "tokens_per_second": 4357987.198856169, "eta_seconds": 88.95432829673035, "mean_data_wait_seconds": 6.881356239318848e-05}
{"step": 45530, "total_steps": 46238, "tokens": 24617336832, "loss": 2.912443843754855, "ce_loss": 2.912443843754855, "z_loss": 44.23868008815881, "muon_lr": 0.001274844718510637, "residual_lr": 7.649068311063822e-06, "adamw_lr": 7.649068311063823e-05, "weight_decay": 0.01, "grad_norm": 0.07219294458627701, "tokens_per_second": 4358055.12171019, "eta_seconds": 87.71231508655065, "mean_data_wait_seconds": 7.331371307373047e-05}
{"step": 45540, "total_steps": 46238, "tokens": 24622743552, "loss": 2.7965194113326795, "ce_loss": 2.7965194113326795, "z_loss": 44.80287762844201, "muon_lr": 0.0012393849514410805, "residual_lr": 7.436309708646482e-06, "adamw_lr": 7.436309708646482e-05, "weight_decay": 0.01, "grad_norm": 0.06014491990208626, "tokens_per_second": 4358168.45954572, "eta_seconds": 86.46943951296488, "mean_data_wait_seconds": 5.97536563873291e-05}
{"step": 45550, "total_steps": 46238, "tokens": 24628150272, "loss": 2.859447562333309, "ce_loss": 2.859447562333309, "z_loss": 44.74058159914884, "muon_lr": 0.0012044127844312387, "residual_lr": 7.226476706587431e-06, "adamw_lr": 7.226476706587432e-05, "weight_decay": 0.01, "grad_norm": 0.056852877140045166, "tokens_per_second": 4358270.061044555, "eta_seconds": 85.22685808758162, "mean_data_wait_seconds": 0.00013846158981323242}
{"step": 45560, "total_steps": 46238, "tokens": 24633556992, "loss": 2.8250268643552605, "ce_loss": 2.8250268643552605, "z_loss": 44.609426787405305, "muon_lr": 0.00116992893515743, "residual_lr": 7.019573610944578e-06, "adamw_lr": 7.019573610944579e-05, "weight_decay": 0.01, "grad_norm": 0.055540379136800766, "tokens_per_second": 4358314.070398978, "eta_seconds": 83.98544439145746, "mean_data_wait_seconds": 0.0001023411750793457}
{"step": 45570, "total_steps": 46238, "tokens": 24638963712, "loss": 2.814998922926007, "ce_loss": 2.814998922926007, "z_loss": 45.303102897875235, "muon_lr": 0.001135934111275036, "residual_lr": 6.815604667650215e-06, "adamw_lr": 6.815604667650216e-05, "weight_decay": 0.01, "grad_norm": 0.058435603976249695, "tokens_per_second": 4358302.535411853, "eta_seconds": 82.74511029692003, "mean_data_wait_seconds": 0.00010582804679870605}
{"step": 45580, "total_steps": 46238, "tokens": 24644370432, "loss": 2.803095599015554, "ce_loss": 2.803095599015554, "z_loss": 45.430999958153926, "muon_lr": 0.001102429010403963, "residual_lr": 6.614574062423778e-06, "adamw_lr": 6.614574062423779e-05, "weight_decay": 0.01, "grad_norm": 0.05691802501678467, "tokens_per_second": 4358340.083500049, "eta_seconds": 81.50385174043888, "mean_data_wait_seconds": 0.00018733739852905273}
{"step": 45590, "total_steps": 46238, "tokens": 24649777152, "loss": 2.750214063760006, "ce_loss": 2.750214063760006, "z_loss": 45.3352977290298, "muon_lr": 0.0010694143201143396, "residual_lr": 6.4164859206860375e-06, "adamw_lr": 6.416485920686038e-05, "weight_decay": 0.01, "grad_norm": 0.06014742702245712, "tokens_per_second": 4358459.157886712, "eta_seconds": 80.26111323470904, "mean_data_wait_seconds": 9.205937385559082e-05}
{"step": 45600, "total_steps": 46238, "tokens": 24655183872, "loss": 2.771144576144941, "ce_loss": 2.771144576144941, "z_loss": 44.76686804222338, "muon_lr": 0.001036890717912406, "residual_lr": 6.221344307474435e-06, "adamw_lr": 6.221344307474436e-05, "weight_decay": 0.01, "grad_norm": 0.0585709847509861, "tokens_per_second": 4358487.027674476, "eta_seconds": 79.0200961510635, "mean_data_wait_seconds": 7.635354995727539e-05}
{"step": 45610, "total_steps": 46238, "tokens": 24660590592, "loss": 2.7937157840439766, "ce_loss": 2.7937157840439766, "z_loss": 46.341520540642016, "muon_lr": 0.001004858871226616, "residual_lr": 6.029153227359696e-06, "adamw_lr": 6.029153227359696e-05, "weight_decay": 0.01, "grad_norm": 0.06688112765550613, "tokens_per_second": 4358565.45227024, "eta_seconds": 77.77819278208266, "mean_data_wait_seconds": 6.464123725891113e-05}
{"step": 45620, "total_steps": 46238, "tokens": 24665997312, "loss": 2.7477850317955017, "ce_loss": 2.7477850317955017, "z_loss": 45.68766007278905, "muon_lr": 0.0009733194373939158, "residual_lr": 5.839916624363494e-06, "adamw_lr": 5.839916624363495e-05, "weight_decay": 0.01, "grad_norm": 0.0607377290725708, "tokens_per_second": 4358665.639017676, "eta_seconds": 76.53595196973703, "mean_data_wait_seconds": 6.625056266784668e-05}
{"step": 45630, "total_steps": 46238, "tokens": 24671404032, "loss": 2.803661528861884, "ce_loss": 2.803661528861884, "z_loss": 45.63508279395826, "muon_lr": 0.0009422730636462845, "residual_lr": 5.653638381877706e-06, "adamw_lr": 5.653638381877707e-05, "weight_decay": 0.01, "grad_norm": 0.05963090434670448, "tokens_per_second": 4358783.231672215, "eta_seconds": 75.29346759326985, "mean_data_wait_seconds": 6.631016731262207e-05}
{"step": 45640, "total_steps": 46238, "tokens": 24676810752, "loss": 2.791204683708422, "ce_loss": 2.791204683708422, "z_loss": 45.63890344446356, "muon_lr": 0.000911720387097434, "residual_lr": 5.470322322584603e-06, "adamw_lr": 5.470322322584603e-05, "weight_decay": 0.01, "grad_norm": 0.060333557426929474, "tokens_per_second": 4358881.7964692665, "eta_seconds": 74.05137351085217, "mean_data_wait_seconds": 5.906820297241211e-05}
{"step": 45650, "total_steps": 46238, "tokens": 24682217472, "loss": 2.808571595134157, "ce_loss": 2.808571595134157, "z_loss": 45.47535720016017, "muon_lr": 0.0008816620347297494, "residual_lr": 5.289972208378496e-06, "adamw_lr": 5.289972208378496e-05, "weight_decay": 0.01, "grad_norm": 0.05935253947973251, "tokens_per_second": 4359027.455762974, "eta_seconds": 72.80854897585154, "mean_data_wait_seconds": 7.084012031555176e-05}
{"step": 45660, "total_steps": 46238, "tokens": 24687624192, "loss": 2.7572078596461904, "ce_loss": 2.7572078596461904, "z_loss": 46.29456297556559, "muon_lr": 0.0008520986233813921, "residual_lr": 5.112591740288352e-06, "adamw_lr": 5.112591740288353e-05, "weight_decay": 0.01, "grad_norm": 0.06518849730491638, "tokens_per_second": 4359148.596330857, "eta_seconds": 71.56621003068963, "mean_data_wait_seconds": 8.922815322875977e-05}
{"step": 45670, "total_steps": 46238, "tokens": 24693030912, "loss": 2.7531490885850154, "ce_loss": 2.7531490885850154, "z_loss": 46.32531215205337, "muon_lr": 0.0008230307597336868, "residual_lr": 4.938184558402119e-06, "adamw_lr": 4.93818455840212e-05, "weight_decay": 0.01, "grad_norm": 0.06328018754720688, "tokens_per_second": 4359190.068788437, "eta_seconds": 70.325225365822, "mean_data_wait_seconds": 7.396936416625977e-05}
{"step": 45680, "total_steps": 46238, "tokens": 24698437632, "loss": 2.7285713390870527, "ce_loss": 2.7285713390870527, "z_loss": 44.98312848987001, "muon_lr": 0.0007944590402986257, "residual_lr": 4.766754241791754e-06, "adamw_lr": 4.7667542417917544e-05, "weight_decay": 0.01, "grad_norm": 0.059680648148059845, "tokens_per_second": 4359240.782037851, "eta_seconds": 69.08411786770283, "mean_data_wait_seconds": 7.209181785583496e-05}
{"step": 45690, "total_steps": 46238, "tokens": 24703844352, "loss": 2.723246005448428, "ce_loss": 2.723246005448428, "z_loss": 45.80709983363296, "muon_lr": 0.0007663840514066712, "residual_lr": 4.5983043084400265e-06, "adamw_lr": 4.598304308440027e-05, "weight_decay": 0.01, "grad_norm": 0.057841405272483826, "tokens_per_second": 4359358.321245324, "eta_seconds": 67.84199925908241, "mean_data_wait_seconds": 7.858872413635254e-05}
{"step": 45700, "total_steps": 46238, "tokens": 24709251072, "loss": 2.8629488186402754, "ce_loss": 2.8629488186402754, "z_loss": 45.293765010255754, "muon_lr": 0.0007388063691946751, "residual_lr": 4.43283821516805e-06, "adamw_lr": 4.43283821516805e-05, "weight_decay": 0.01, "grad_norm": 0.06171698868274689, "tokens_per_second": 4359257.1247629635, "eta_seconds": 66.60328943450139, "mean_data_wait_seconds": 8.723139762878418e-05}
{"step": 45710, "total_steps": 46238, "tokens": 24714657792, "loss": 2.810622052712874, "ce_loss": 2.810622052712874, "z_loss": 45.80326762343898, "muon_lr": 0.0007117265595941003, "residual_lr": 4.270359357564601e-06, "adamw_lr": 4.270359357564601e-05, "weight_decay": 0.01, "grad_norm": 0.05383312329649925, "tokens_per_second": 4359325.072039393, "eta_seconds": 65.36198592473886, "mean_data_wait_seconds": 9.933114051818848e-05}
{"step": 45720, "total_steps": 46238, "tokens": 24720064512, "loss": 2.8410961013851743, "ce_loss": 2.8410961013851743, "z_loss": 45.806116682110414, "muon_lr": 0.0006851451783193796, "residual_lr": 4.110871069916277e-06, "adamw_lr": 4.1108710699162776e-05, "weight_decay": 0.01, "grad_norm": 0.06106872484087944, "tokens_per_second": 4359366.4461173, "eta_seconds": 64.12111196776382, "mean_data_wait_seconds": 0.00012621283531188965}
{"step": 45730, "total_steps": 46238, "tokens": 24725471232, "loss": 2.8490282875118833, "ce_loss": 2.8490282875118833, "z_loss": 45.490770079872824, "muon_lr": 0.0006590627708565306, "residual_lr": 3.954376625139183e-06, "adamw_lr": 3.9543766251391836e-05, "weight_decay": 0.01, "grad_norm": 0.055551573634147644, "tokens_per_second": 4359317.957743979, "eta_seconds": 62.88155777053301, "mean_data_wait_seconds": 0.00018042325973510742}
{"step": 45740, "total_steps": 46238, "tokens": 24730877952, "loss": 2.800144233486869, "ce_loss": 2.800144233486869, "z_loss": 44.931873523827754, "muon_lr": 0.0006334798724519337, "residual_lr": 3.8008792347116014e-06, "adamw_lr": 3.800879234711602e-05, "weight_decay": 0.01, "grad_norm": 0.07958647608757019, "tokens_per_second": 4359383.982499187, "eta_seconds": 61.64035677489213, "mean_data_wait_seconds": 0.00015479326248168945}
{"step": 45750, "total_steps": 46238, "tokens": 24736284672, "loss": 2.8354938608227354, "ce_loss": 2.8354938608227354, "z_loss": 44.72413904016668, "muon_lr": 0.0006083970081013801, "residual_lr": 3.65038204860828e-06, "adamw_lr": 3.650382048608281e-05, "weight_decay": 0.01, "grad_norm": 0.0644553080201149, "tokens_per_second": 4359427.76709334, "eta_seconds": 60.399501509704066, "mean_data_wait_seconds": 0.0001373291015625}
{"step": 45760, "total_steps": 46238, "tokens": 24741691392, "loss": 2.8186250907002073, "ce_loss": 2.8186250907002073, "z_loss": 44.847213398326524, "muon_lr": 0.0005838146925392834, "residual_lr": 3.5028881552357003e-06, "adamw_lr": 3.5028881552357e-05, "weight_decay": 0.01, "grad_norm": 0.06293365359306335, "tokens_per_second": 4359432.284202739, "eta_seconds": 59.159204040065816, "mean_data_wait_seconds": 0.00018864870071411133}
{"step": 45770, "total_steps": 46238, "tokens": 24747098112, "loss": 2.9054181991201458, "ce_loss": 2.9054181991201458, "z_loss": 45.413062153440535, "muon_lr": 0.0005597334302281043, "residual_lr": 3.3584005813686256e-06, "adamw_lr": 3.358400581368626e-05, "weight_decay": 0.01, "grad_norm": 0.07168937474489212, "tokens_per_second": 4358963.52386452, "eta_seconds": 57.92519772593713, "mean_data_wait_seconds": 9.587407112121582e-05}
{"step": 45780, "total_steps": 46238, "tokens": 24752504832, "loss": 2.848755048983025, "ce_loss": 2.848755048983025, "z_loss": 45.07751571770871, "muon_lr": 0.0005361537153480228, "residual_lr": 3.2169222920881364e-06, "adamw_lr": 3.216922292088137e-05, "weight_decay": 0.01, "grad_norm": 0.0626247301697731, "tokens_per_second": 4358889.088336712, "eta_seconds": 56.68579745723349, "mean_data_wait_seconds": 0.00010028481483459473}
{"step": 45790, "total_steps": 46238, "tokens": 24757911552, "loss": 2.9036002321676775, "ce_loss": 2.9036002321676775, "z_loss": 44.94862446640477, "muon_lr": 0.0005130760317867827, "residual_lr": 3.0784561907206962e-06, "adamw_lr": 3.0784561907206964e-05, "weight_decay": 0.01, "grad_norm": 0.0657467320561409, "tokens_per_second": 4358891.978956434, "eta_seconds": 55.44537124727301, "mean_data_wait_seconds": 6.318092346191406e-05}
{"step": 45800, "total_steps": 46238, "tokens": 24763318272, "loss": 2.872802830103672, "ce_loss": 2.872802830103672, "z_loss": 45.17219523227576, "muon_lr": 0.0004905008531297661, "residual_lr": 2.9430051187785962e-06, "adamw_lr": 2.943005118778597e-05, "weight_decay": 0.01, "grad_norm": 0.061873212456703186, "tokens_per_second": 4359024.238890731, "eta_seconds": 54.203337960819894, "mean_data_wait_seconds": 7.036328315734863e-05}
{"step": 45810, "total_steps": 46238, "tokens": 24768724992, "loss": 2.803528856147419, "ce_loss": 2.803528856147419, "z_loss": 44.79607434706254, "muon_lr": 0.0004684286426502649, "residual_lr": 2.810571855901589e-06, "adamw_lr": 2.810571855901589e-05, "weight_decay": 0.01, "grad_norm": 0.06337124854326248, "tokens_per_second": 4359093.143571943, "eta_seconds": 52.96214978577453, "mean_data_wait_seconds": 7.063150405883789e-05}
{"step": 45820, "total_steps": 46238, "tokens": 24774131712, "loss": 2.8455320398012796, "ce_loss": 2.8455320398012796, "z_loss": 45.33693050615715, "muon_lr": 0.0004468598532999857, "residual_lr": 2.6811591197999137e-06, "adamw_lr": 2.681159119799914e-05, "weight_decay": 0.01, "grad_norm": 0.05736478045582771, "tokens_per_second": 4359171.620434415, "eta_seconds": 51.72088727663622, "mean_data_wait_seconds": 6.780028343200684e-05}
{"step": 45830, "total_steps": 46238, "tokens": 24779538432, "loss": 2.8773633169405386, "ce_loss": 2.8773633169405386, "z_loss": 45.3280448624582, "muon_lr": 0.0004257949276997547, "residual_lr": 2.554769566198528e-06, "adamw_lr": 2.554769566198528e-05, "weight_decay": 0.01, "grad_norm": 0.0578363798558712, "tokens_per_second": 4358729.10673009, "eta_seconds": 50.48570319734407, "mean_data_wait_seconds": 7.05718994140625e-05}
{"step": 45840, "total_steps": 46238, "tokens": 24784945152, "loss": 2.840378837151961, "ce_loss": 2.840378837151961, "z_loss": 45.74482227094246, "muon_lr": 0.0004052342981304247, "residual_lr": 2.431405788782548e-06, "adamw_lr": 2.4314057887825482e-05, "weight_decay": 0.01, "grad_norm": 0.06644150614738464, "tokens_per_second": 4358775.96194172, "eta_seconds": 49.24473886113213, "mean_data_wait_seconds": 7.978081703186035e-05}
{"step": 45850, "total_steps": 46238, "tokens": 24790351872, "loss": 2.7930596705639, "ce_loss": 2.7930596705639, "z_loss": 45.357877644625574, "muon_lr": 0.00038517838652401, "residual_lr": 2.3110703191440593e-06, "adamw_lr": 2.3110703191440597e-05, "weight_decay": 0.01, "grad_norm": 0.08693435043096542, "tokens_per_second": 4358795.7711679675, "eta_seconds": 48.00409906425434, "mean_data_wait_seconds": 7.995963096618652e-05}
{"step": 45860, "total_steps": 46238, "tokens": 24795758592, "loss": 2.913505615610065, "ce_loss": 2.913505615610065, "z_loss": 45.43135515848795, "muon_lr": 0.0003656276044550322, "residual_lr": 2.193765626730193e-06, "adamw_lr": 2.1937656267301932e-05, "weight_decay": 0.01, "grad_norm": 0.05918214097619057, "tokens_per_second": 4358826.068049804, "eta_seconds": 46.763358027542886, "mean_data_wait_seconds": 0.00015804171562194824}
{"step": 45870, "total_steps": 46238, "tokens": 24801165312, "loss": 2.894219154661352, "ce_loss": 2.894219154661352, "z_loss": 44.93966515858968, "muon_lr": 0.0003465823531320633, "residual_lr": 2.0794941187923796e-06, "adamw_lr": 2.07949411879238e-05, "weight_decay": 0.01, "grad_norm": 0.05874667316675186, "tokens_per_second": 4358825.179778188, "eta_seconds": 45.522959929789515, "mean_data_wait_seconds": 0.00012987852096557617}
{"step": 45880, "total_steps": 46238, "tokens": 24806572032, "loss": 2.9008820418155556, "ce_loss": 2.9008820418155556, "z_loss": 45.1422443967877, "muon_lr": 0.00032804302338950434, "residual_lr": 1.9682581403370256e-06, "adamw_lr": 1.9682581403370257e-05, "weight_decay": 0.01, "grad_norm": 0.06015162914991379, "tokens_per_second": 4358767.296949735, "eta_seconds": 44.2831403583016, "mean_data_wait_seconds": 0.00012674927711486816}
{"step": 45890, "total_steps": 46238, "tokens": 24811978752, "loss": 2.9162139838392083, "ce_loss": 2.9162139838392083, "z_loss": 45.104498660925664, "muon_lr": 0.0003100099956795527, "residual_lr": 1.860059974077316e-06, "adamw_lr": 1.8600599740773162e-05, "weight_decay": 0.01, "grad_norm": 0.06180707365274429, "tokens_per_second": 4358438.972362902, "eta_seconds": 43.04595869981555, "mean_data_wait_seconds": 8.884072303771973e-05}
{"step": 45900, "total_steps": 46238, "tokens": 24817385472, "loss": 2.8644847039020425, "ce_loss": 2.8644847039020425, "z_loss": 45.36476817275538, "muon_lr": 0.000292483640064406, "residual_lr": 1.7549018403864357e-06, "adamw_lr": 1.7549018403864357e-05, "weight_decay": 0.01, "grad_norm": 0.061001818627119064, "tokens_per_second": 4358465.308565022, "eta_seconds": 41.80518854697263, "mean_data_wait_seconds": 8.866190910339355e-05}
{"step": 45910, "total_steps": 46238, "tokens": 24822792192, "loss": 2.8988072817975823, "ce_loss": 2.8988072817975823, "z_loss": 44.97627148483739, "muon_lr": 0.0002754643162086623, "residual_lr": 1.6527858972519737e-06, "adamw_lr": 1.6527858972519736e-05, "weight_decay": 0.01, "grad_norm": 0.05894298106431961, "tokens_per_second": 4358471.562641899, "eta_seconds": 40.56462029382437, "mean_data_wait_seconds": 0.000110626220703125}
{"step": 45920, "total_steps": 46238, "tokens": 24828198912, "loss": 2.9428369836373762, "ce_loss": 2.9428369836373762, "z_loss": 45.26771808393074, "muon_lr": 0.0002589523733719429, "residual_lr": 1.5537142402316572e-06, "adamw_lr": 1.5537142402316573e-05, "weight_decay": 0.01, "grad_norm": 0.05829582363367081, "tokens_per_second": 4358505.480962238, "eta_seconds": 39.323806003832566, "mean_data_wait_seconds": 8.985400199890137e-05}
{"step": 45930, "total_steps": 46238, "tokens": 24833605632, "loss": 2.9233027190873115, "ce_loss": 2.9233027190873115, "z_loss": 44.94516658782959, "muon_lr": 0.0002429481504017117, "residual_lr": 1.45768890241027e-06, "adamw_lr": 1.4576889024102701e-05, "weight_decay": 0.01, "grad_norm": 0.05779004842042923, "tokens_per_second": 4358513.535446506, "eta_seconds": 38.08323701419815, "mean_data_wait_seconds": 7.724761962890625e-05}
{"step": 45940, "total_steps": 46238, "tokens": 24839012352, "loss": 2.9050752303817053, "ce_loss": 2.9050752303817053, "z_loss": 44.70131460825602, "muon_lr": 0.00022745197572634213, "residual_lr": 1.3647118543580527e-06, "adamw_lr": 1.3647118543580529e-05, "weight_decay": 0.01, "grad_norm": 0.0588325671851635, "tokens_per_second": 4358557.991672601, "eta_seconds": 36.8423649075683, "mean_data_wait_seconds": 9.223818778991699e-05}
{"step": 45950, "total_steps": 46238, "tokens": 24844419072, "loss": 2.950034448594758, "ce_loss": 2.950034448594758, "z_loss": 45.00439010966908, "muon_lr": 0.00021246416734836405, "residual_lr": 1.274785004090184e-06, "adamw_lr": 1.2747850040901842e-05, "weight_decay": 0.01, "grad_norm": 0.057162895798683167, "tokens_per_second": 4358587.900466781, "eta_seconds": 35.60163693919809, "mean_data_wait_seconds": 0.00010961294174194336}
{"step": 45960, "total_steps": 46238, "tokens": 24849825792, "loss": 2.9305383292111484, "ce_loss": 2.9305383292111484, "z_loss": 44.7413602886778, "muon_lr": 0.000197985032837944, "residual_lr": 1.1879101970276638e-06, "adamw_lr": 1.187910197027664e-05, "weight_decay": 0.01, "grad_norm": 0.060212478041648865, "tokens_per_second": 4358599.269755804, "eta_seconds": 34.36107215435541, "mean_data_wait_seconds": 0.00011038780212402344}
{"step": 45970, "total_steps": 46238, "tokens": 24855232512, "loss": 2.921827834663969, "ce_loss": 2.921827834663969, "z_loss": 45.00463702461936, "muon_lr": 0.00018401486932656254, "residual_lr": 1.104089215959375e-06, "adamw_lr": 1.1040892159593751e-05, "weight_decay": 0.01, "grad_norm": 0.05865864455699921, "tokens_per_second": 4358608.213423986, "eta_seconds": 33.12053227344234, "mean_data_wait_seconds": 0.00010102987289428711}
{"step": 45980, "total_steps": 46238, "tokens": 24860639232, "loss": 2.8841487458257964, "ce_loss": 2.8841487458257964, "z_loss": 44.9943245974454, "muon_lr": 0.00017055396350093013, "residual_lr": 1.0233237810055804e-06, "adamw_lr": 1.0233237810055807e-05, "weight_decay": 0.01, "grad_norm": 0.05522632226347923, "tokens_per_second": 4358610.976267506, "eta_seconds": 31.88004269171828, "mean_data_wait_seconds": 0.00016838312149047852}
{"step": 45990, "total_steps": 46238, "tokens": 24866045952, "loss": 2.916102398525585, "ce_loss": 2.916102398525585, "z_loss": 44.31156094868978, "muon_lr": 0.0001576025915971058, "residual_lr": 9.456155495826345e-07, "adamw_lr": 9.456155495826346e-06, "weight_decay": 0.01, "grad_norm": 0.06245172768831253, "tokens_per_second": 4358713.864463066, "eta_seconds": 30.638850852039365, "mean_data_wait_seconds": 0.00010380148887634277}
{"step": 46000, "total_steps": 46238, "tokens": 24871452672, "loss": 2.891634346860828, "ce_loss": 2.891634346860828, "z_loss": 44.883897636875965, "muon_lr": 0.0001451610193948044, "residual_lr": 8.709661163688264e-07, "adamw_lr": 8.709661163688266e-06, "weight_decay": 0.01, "grad_norm": 0.06093263253569603, "tokens_per_second": 4358704.233210706, "eta_seconds": 29.398476506769107, "mean_data_wait_seconds": 6.35981559753418e-05}
{"step": 46010, "total_steps": 46238, "tokens": 24876859392, "loss": 2.947849306193265, "ce_loss": 2.947849306193265, "z_loss": 44.61817221208052, "muon_lr": 0.00013322950221197338, "residual_lr": 7.993770132718403e-07, "adamw_lr": 7.993770132718403e-06, "weight_decay": 0.01, "grad_norm": 0.05752739682793617, "tokens_per_second": 4358777.910275991, "eta_seconds": 28.157558500664415, "mean_data_wait_seconds": 7.361173629760742e-05}
{"step": 46020, "total_steps": 46238, "tokens": 24882266112, "loss": 2.892707721753554, "ce_loss": 2.892707721753554, "z_loss": 44.93367229808461, "muon_lr": 0.00012180828489953278, "residual_lr": 7.308497093971966e-07, "adamw_lr": 7.308497093971966e-06, "weight_decay": 0.01, "grad_norm": 0.05555659532546997, "tokens_per_second": 4358731.513246725, "eta_seconds": 26.917423943969084, "mean_data_wait_seconds": 0.00014287233352661133}
{"step": 46030, "total_steps": 46238, "tokens": 24887672832, "loss": 2.9265862089214902, "ce_loss": 2.9265862089214902, "z_loss": 44.19421248002486, "muon_lr": 0.00011089760183636, "residual_lr": 6.653856110181599e-07, "adamw_lr": 6.6538561101816e-06, "weight_decay": 0.01, "grad_norm": 0.05473090708255768, "tokens_per_second": 4358815.709062084, "eta_seconds": 25.676493678619508, "mean_data_wait_seconds": 0.00011360645294189453}
{"step": 46040, "total_steps": 46238, "tokens": 24893079552, "loss": 2.918538631814899, "ce_loss": 2.918538631814899, "z_loss": 44.681462114507504, "muon_lr": 0.00010049767692447421, "residual_lr": 6.029860615468451e-07, "adamw_lr": 6.029860615468452e-06, "weight_decay": 0.01, "grad_norm": 0.05571117252111435, "tokens_per_second": 4358726.69633393, "eta_seconds": 24.436582383012503, "mean_data_wait_seconds": 6.812810897827148e-05}
{"step": 46050, "total_steps": 46238, "tokens": 24898486272, "loss": 2.905450089411302, "ce_loss": 2.905450089411302, "z_loss": 44.09922969702518, "muon_lr": 9.06087235844455e-05, "residual_lr": 5.436523415066729e-07, "adamw_lr": 5.4365234150667296e-06, "weight_decay": 0.01, "grad_norm": 0.06638909876346588, "tokens_per_second": 4358738.970357029, "eta_seconds": 23.196081409692294, "mean_data_wait_seconds": 7.072091102600098e-05}
{"step": 46060, "total_steps": 46238, "tokens": 24903892992, "loss": 2.8709993579170923, "ce_loss": 2.8709993579170923, "z_loss": 44.90661707791415, "muon_lr": 8.123094475101511e-05, "residual_lr": 4.873856685060906e-07, "adamw_lr": 4.873856685060907e-06, "weight_decay": 0.01, "grad_norm": 0.0568418875336647, "tokens_per_second": 4358813.6356819, "eta_seconds": 21.95527315428082, "mean_data_wait_seconds": 0.00010791420936584473}
{"step": 46070, "total_steps": 46238, "tokens": 24909299712, "loss": 2.882285531723138, "ce_loss": 2.882285531723138, "z_loss": 44.79954245596221, "muon_lr": 7.236453286892653e-05, "residual_lr": 4.341871972135591e-07, "adamw_lr": 4.341871972135592e-06, "weight_decay": 0.01, "grad_norm": 0.05425570532679558, "tokens_per_second": 4358875.604482231, "eta_seconds": 20.714567744753378, "mean_data_wait_seconds": 6.16610050201416e-05}
{"step": 46080, "total_steps": 46238, "tokens": 24914706432, "loss": 2.896205773859313, "ce_loss": 2.896205773859313, "z_loss": 44.92525554425789, "muon_lr": 6.400966988898416e-05, "residual_lr": 3.840580193339049e-07, "adamw_lr": 3.840580193339049e-06, "weight_decay": 0.01, "grad_norm": 0.06017562374472618, "tokens_per_second": 4358865.550324027, "eta_seconds": 19.474219385750455, "mean_data_wait_seconds": 9.402632713317871e-05}
{"step": 46090, "total_steps": 46238, "tokens": 24920113152, "loss": 2.8945900057301377, "ce_loss": 2.8945900057301377, "z_loss": 45.03507406061346, "muon_lr": 5.6166527264309156e-05, "residual_lr": 3.369991635858549e-07, "adamw_lr": 3.3699916358585494e-06, "weight_decay": 0.01, "grad_norm": 0.054795004427433014, "tokens_per_second": 4358947.204541218, "eta_seconds": 18.23348168043829, "mean_data_wait_seconds": 6.625056266784668e-05}
{"step": 46100, "total_steps": 46238, "tokens": 24925519872, "loss": 2.827968516133048, "ce_loss": 2.827968516133048, "z_loss": 45.016131950147226, "muon_lr": 4.883526594683385e-05, "residual_lr": 2.9301159568100307e-07, "adamw_lr": 2.930115956810031e-06, "weight_decay": 0.01, "grad_norm": 0.0629286915063858, "tokens_per_second": 4359014.622826686, "eta_seconds": 16.992845954704908, "mean_data_wait_seconds": 7.337331771850586e-05}
{"step": 46110, "total_steps": 46238, "tokens": 24930926592, "loss": 2.886820356051127, "ce_loss": 2.886820356051127, "z_loss": 45.25811834046335, "muon_lr": 4.201603638398499e-05, "residual_lr": 2.520962183039099e-07, "adamw_lr": 2.5209621830390993e-06, "weight_decay": 0.01, "grad_norm": 0.054536860436201096, "tokens_per_second": 4359107.275450241, "eta_seconds": 15.752157416889387, "mean_data_wait_seconds": 7.69197940826416e-05}
{"step": 46120, "total_steps": 46238, "tokens": 24936333312, "loss": 2.888951073993336, "ce_loss": 2.888951073993336, "z_loss": 45.11541626670144, "muon_lr": 3.570897851560562e-05, "residual_lr": 2.1425387109363368e-07, "adamw_lr": 2.142538710936337e-06, "weight_decay": 0.01, "grad_norm": 0.056394513696432114, "tokens_per_second": 4359098.980041332, "eta_seconds": 14.511857677386395, "mean_data_wait_seconds": 7.805228233337402e-05}
{"step": 46130, "total_steps": 46238, "tokens": 24941740032, "loss": 2.817921817302704, "ce_loss": 2.817921817302704, "z_loss": 45.40990161895752, "muon_lr": 2.9914221771082386e-05, "residual_lr": 1.794853306264943e-07, "adamw_lr": 1.794853306264943e-06, "weight_decay": 0.01, "grad_norm": 0.05606565251946449, "tokens_per_second": 4359199.764036434, "eta_seconds": 13.271221125785617, "mean_data_wait_seconds": 7.608532905578613e-05}
{"step": 46140, "total_steps": 46238, "tokens": 24947146752, "loss": 2.7971823847655095, "ce_loss": 2.7971823847655095, "z_loss": 45.245298674612336, "muon_lr": 2.463188506668379e-05, "residual_lr": 1.4779131040010273e-07, "adamw_lr": 1.4779131040010274e-06, "weight_decay": 0.01, "grad_norm": 0.05824029818177223, "tokens_per_second": 4359196.293639269, "eta_seconds": 12.030929663921194, "mean_data_wait_seconds": 8.508563041687012e-05}
{"step": 46150, "total_steps": 46238, "tokens": 24952553472, "loss": 2.911639195499998, "ce_loss": 2.911639195499998, "z_loss": 44.0382242491751, "muon_lr": 1.986207680313157e-05, "residual_lr": 1.1917246081878939e-07, "adamw_lr": 1.191724608187894e-06, "weight_decay": 0.01, "grad_norm": 0.07412780821323395, "tokens_per_second": 4359219.155596153, "eta_seconds": 10.790571045186915, "mean_data_wait_seconds": 0.000110626220703125}
{"step": 46160, "total_steps": 46238, "tokens": 24957960192, "loss": 2.856398217605822, "ce_loss": 2.856398217605822, "z_loss": 45.22540523066665, "muon_lr": 1.5604894863355256e-05, "residual_lr": 9.362936918013153e-08, "adamw_lr": 9.362936918013154e-07, "weight_decay": 0.01, "grad_norm": 0.05963176488876343, "tokens_per_second": 4359261.782610665, "eta_seconds": 9.550182135441215, "mean_data_wait_seconds": 0.00012508034706115723}
{"step": 46170, "total_steps": 46238, "tokens": 24963366912, "loss": 2.7958305351661914, "ce_loss": 2.7958305351661914, "z_loss": 45.63210438237046, "muon_lr": 1.1860426610510456e-05, "residual_lr": 7.116255966306273e-08, "adamw_lr": 7.116255966306273e-07, "weight_decay": 0.01, "grad_norm": 0.061492715030908585, "tokens_per_second": 4359328.796140539, "eta_seconds": 8.309770997790126, "mean_data_wait_seconds": 0.00022491812705993652}
{"step": 46180, "total_steps": 46238, "tokens": 24968773632, "loss": 2.7885118090745173, "ce_loss": 2.7885118090745173, "z_loss": 45.05338330702348, "muon_lr": 8.628748886163607e-06, "residual_lr": 5.1772493316981636e-08, "adamw_lr": 5.177249331698164e-07, "weight_decay": 0.01, "grad_norm": 0.06081140413880348, "tokens_per_second": 4359348.320405195, "eta_seconds": 7.069475007478982, "mean_data_wait_seconds": 0.00010210275650024414}
{"step": 46190, "total_steps": 46238, "tokens": 24974180352, "loss": 2.768550567554705, "ce_loss": 2.768550567554705, "z_loss": 45.08809520259048, "muon_lr": 5.909928008726584e-06, "residual_lr": 3.5459568052359496e-08, "adamw_lr": 3.54595680523595e-07, "weight_decay": 0.01, "grad_norm": 0.06061793863773346, "tokens_per_second": 4359392.848438702, "eta_seconds": 5.829156693024593, "mean_data_wait_seconds": 0.00016620755195617676}
{"step": 46200, "total_steps": 46238, "tokens": 24979587072, "loss": 2.8249430656433105, "ce_loss": 2.8249430656433105, "z_loss": 46.01676903348981, "muon_lr": 3.7040197720938917e-06, "residual_lr": 2.2224118632563347e-08, "adamw_lr": 2.222411863256335e-07, "weight_decay": 0.01, "grad_norm": 0.06011546030640602, "tokens_per_second": 4359400.587558662, "eta_seconds": 4.588902441563202, "mean_data_wait_seconds": 9.745359420776367e-05}
{"step": 46210, "total_steps": 46238, "tokens": 24984993792, "loss": 2.731505634206714, "ce_loss": 2.731505634206714, "z_loss": 46.565733331622496, "muon_lr": 2.0110694444963607e-06, "residual_lr": 1.2066416666978163e-08, "adamw_lr": 1.2066416666978164e-07, "weight_decay": 0.01, "grad_norm": 0.07150918245315552, "tokens_per_second": 4359439.108347618, "eta_seconds": 3.3486289490881807, "mean_data_wait_seconds": 7.429718971252441e-05}
{"step": 46220, "total_steps": 46238, "tokens": 24990400512, "loss": 2.7602971972841206, "ce_loss": 2.7602971972841206, "z_loss": 45.92346000671387, "muon_lr": 8.311117675685598e-07, "residual_lr": 4.986670605411358e-09, "adamw_lr": 4.9866706054113585e-08, "weight_decay": 0.01, "grad_norm": 0.0650131031870842, "tokens_per_second": 4359469.988649119, "eta_seconds": 2.10838107016036, "mean_data_wait_seconds": 9.524822235107422e-05}
{"step": 46230, "total_steps": 46238, "tokens": 24995807232, "loss": 2.7536270853244895, "ce_loss": 2.7536270853244895, "z_loss": 45.903192433443934, "muon_lr": 1.6417095564380535e-07, "residual_lr": 9.85025733862832e-10, "adamw_lr": 9.850257338628322e-09, "weight_decay": 0.01, "grad_norm": 0.06820283085107803, "tokens_per_second": 4359463.230355016, "eta_seconds": 0.8681582571099676, "mean_data_wait_seconds": 7.200241088867188e-05}
{"step": 46237, "total_steps": 46238, "tokens": 24999591936, "loss": 2.711019122239315, "ce_loss": 2.711019122239315, "z_loss": 46.08332200483842, "muon_lr": 2.5651739454790602e-09, "residual_lr": 1.539104367287436e-11, "adamw_lr": 1.5391043672874362e-10, "weight_decay": 0.01, "grad_norm": 0.05836516246199608, "tokens_per_second": 4359466.770430391, "eta_seconds": 0.0, "mean_data_wait_seconds": 7.703900337219238e-05}