{"loss": 91.5, "grad_norm": 1460.72143555, "learning_rate": 1.515e-05, "token_acc": 0.09404563, "epoch": 0.00045866, "global_step/max_steps": "1/2181", "elapsed_time": "11s", "remaining_time": "6h 54m 42s", "memory(GiB)": 49.48, "train_speed(s/it)": 11.41369}
{"loss": 90.5, "grad_norm": 1222.9642334, "learning_rate": 3.03e-05, "token_acc": 0.09786629, "epoch": 0.00091733, "global_step/max_steps": "2/2181", "elapsed_time": "20s", "remaining_time": "6h 3m 40s", "memory(GiB)": 49.48, "train_speed(s/it)": 10.01375}
{"loss": 86.0, "grad_norm": 1073.49206543, "learning_rate": 4.545e-05, "token_acc": 0.09268565, "epoch": 0.00137599, "global_step/max_steps": "3/2181", "elapsed_time": "29s", "remaining_time": "5h 52m 2s", "memory(GiB)": 50.48, "train_speed(s/it)": 9.697769}
{"loss": 78.0, "grad_norm": 565.42718506, "learning_rate": 6.061e-05, "token_acc": 0.09655172, "epoch": 0.00183465, "global_step/max_steps": "4/2181", "elapsed_time": "38s", "remaining_time": "5h 48m 40s", "memory(GiB)": 50.48, "train_speed(s/it)": 9.609783}
{"loss": 80.5, "grad_norm": 674.19018555, "learning_rate": 7.576e-05, "token_acc": 0.08630137, "epoch": 0.00229331, "global_step/max_steps": "5/2181", "elapsed_time": "47s", "remaining_time": "5h 42m 52s", "memory(GiB)": 50.52, "train_speed(s/it)": 9.453935}
{"loss": 68.0, "grad_norm": 444.0715332, "learning_rate": 9.091e-05, "token_acc": 0.08351235, "epoch": 0.00275198, "global_step/max_steps": "6/2181", "elapsed_time": "56s", "remaining_time": "5h 40m 48s", "memory(GiB)": 50.52, "train_speed(s/it)": 9.401553}
{"loss": 68.0, "grad_norm": 292.11083984, "learning_rate": 0.00010606, "token_acc": 0.0822854, "epoch": 0.00321064, "global_step/max_steps": "7/2181", "elapsed_time": "1m 6s", "remaining_time": "5h 43m 28s", "memory(GiB)": 58.39, "train_speed(s/it)": 9.479473}
{"loss": 65.5, "grad_norm": 297.55383301, "learning_rate": 0.00012121, "token_acc": 0.08124318, "epoch": 0.0036693, "global_step/max_steps": "8/2181", "elapsed_time": "1m 15s", "remaining_time": "5h 40m 7s", "memory(GiB)": 58.39, "train_speed(s/it)": 9.391365}
{"loss": 58.5, "grad_norm": 321.20318604, "learning_rate": 0.00013636, "token_acc": 0.11724724, "epoch": 0.00412797, "global_step/max_steps": "9/2181", "elapsed_time": "1m 25s", "remaining_time": "5h 40m 1s", "memory(GiB)": 58.39, "train_speed(s/it)": 9.392739}
{"loss": 52.25, "grad_norm": 116.73591614, "learning_rate": 0.00015152, "token_acc": 0.17631954, "epoch": 0.00458663, "global_step/max_steps": "10/2181", "elapsed_time": "1m 35s", "remaining_time": "5h 42m 17s", "memory(GiB)": 58.39, "train_speed(s/it)": 9.459908}
{"loss": 59.5, "grad_norm": 965.50726318, "learning_rate": 0.00016667, "token_acc": 0.17268402, "epoch": 0.00504529, "global_step/max_steps": "11/2181", "elapsed_time": "1m 44s", "remaining_time": "5h 40m 18s", "memory(GiB)": 58.39, "train_speed(s/it)": 9.409398}
{"loss": 62.5, "grad_norm": 743.85742188, "learning_rate": 0.00018182, "token_acc": 0.16024567, "epoch": 0.00550396, "global_step/max_steps": "12/2181", "elapsed_time": "1m 53s", "remaining_time": "5h 39m 26s", "memory(GiB)": 58.39, "train_speed(s/it)": 9.38963}
{"loss": 50.5, "grad_norm": 545.18847656, "learning_rate": 0.00019697, "token_acc": 0.20463215, "epoch": 0.00596262, "global_step/max_steps": "13/2181", "elapsed_time": "2m 3s", "remaining_time": "5h 40m 34s", "memory(GiB)": 58.39, "train_speed(s/it)": 9.425341}
{"loss": 47.0, "grad_norm": 68.78133392, "learning_rate": 0.00021212, "token_acc": 0.21545278, "epoch": 0.00642128, "global_step/max_steps": "14/2181", "elapsed_time": "2m 11s", "remaining_time": "5h 38m 31s", "memory(GiB)": 58.39, "train_speed(s/it)": 9.372933}
{"loss": 46.0, "grad_norm": 67.8844223, "learning_rate": 0.00022727, "token_acc": 0.2352438, "epoch": 0.00687994, "global_step/max_steps": "15/2181", "elapsed_time": "2m 20s", "remaining_time": "5h 37m 26s", "memory(GiB)": 58.39, "train_speed(s/it)": 9.347114}
{"loss": 43.5, "grad_norm": 40.02203369, "learning_rate": 0.00024242, "token_acc": 0.25322359, "epoch": 0.00733861, "global_step/max_steps": "16/2181", "elapsed_time": "2m 29s", "remaining_time": "5h 36m 3s", "memory(GiB)": 58.39, "train_speed(s/it)": 9.31324}
{"loss": 43.25, "grad_norm": 25.75290489, "learning_rate": 0.00025758, "token_acc": 0.23896031, "epoch": 0.00779727, "global_step/max_steps": "17/2181", "elapsed_time": "2m 39s", "remaining_time": "5h 36m 58s", "memory(GiB)": 58.39, "train_speed(s/it)": 9.342963}
{"loss": 41.25, "grad_norm": 26.26945305, "learning_rate": 0.00027273, "token_acc": 0.25475717, "epoch": 0.00825593, "global_step/max_steps": "18/2181", "elapsed_time": "2m 48s", "remaining_time": "5h 36m 17s", "memory(GiB)": 58.39, "train_speed(s/it)": 9.328373}
{"loss": 42.0, "grad_norm": 37.69101334, "learning_rate": 0.00028788, "token_acc": 0.22839862, "epoch": 0.0087146, "global_step/max_steps": "19/2181", "elapsed_time": "2m 57s", "remaining_time": "5h 34m 57s", "memory(GiB)": 58.39, "train_speed(s/it)": 9.295619}
{"loss": 39.75, "grad_norm": 22.12291527, "learning_rate": 0.00030303, "token_acc": 0.25421853, "epoch": 0.00917326, "global_step/max_steps": "20/2181", "elapsed_time": "3m 6s", "remaining_time": "5h 34m 11s", "memory(GiB)": 58.39, "train_speed(s/it)": 9.278668}
{"loss": 39.5, "grad_norm": 15.31917572, "learning_rate": 0.00031818, "token_acc": 0.26930529, "epoch": 0.00963192, "global_step/max_steps": "21/2181", "elapsed_time": "3m 15s", "remaining_time": "5h 33m 44s", "memory(GiB)": 58.41, "train_speed(s/it)": 9.270349}
{"loss": 39.0, "grad_norm": 13.47049999, "learning_rate": 0.00033333, "token_acc": 0.25565801, "epoch": 0.01009059, "global_step/max_steps": "22/2181", "elapsed_time": "3m 24s", "remaining_time": "5h 32m 56s", "memory(GiB)": 58.41, "train_speed(s/it)": 9.252265}
{"loss": 37.5, "grad_norm": 14.24393845, "learning_rate": 0.00034848, "token_acc": 0.25448126, "epoch": 0.01054925, "global_step/max_steps": "23/2181", "elapsed_time": "3m 33s", "remaining_time": "5h 33m 20s", "memory(GiB)": 58.41, "train_speed(s/it)": 9.268039}
{"loss": 38.25, "grad_norm": 11.76596069, "learning_rate": 0.00036364, "token_acc": 0.25205092, "epoch": 0.01100791, "global_step/max_steps": "24/2181", "elapsed_time": "3m 42s", "remaining_time": "5h 32m 43s", "memory(GiB)": 58.41, "train_speed(s/it)": 9.255041}
{"loss": 37.25, "grad_norm": 11.01061726, "learning_rate": 0.00037879, "token_acc": 0.27642276, "epoch": 0.01146657, "global_step/max_steps": "25/2181", "elapsed_time": "3m 51s", "remaining_time": "5h 31m 39s", "memory(GiB)": 58.41, "train_speed(s/it)": 9.229725}
{"loss": 37.5, "grad_norm": 13.49042702, "learning_rate": 0.00039394, "token_acc": 0.26547922, "epoch": 0.01192524, "global_step/max_steps": "26/2181", "elapsed_time": "4m 1s", "remaining_time": "5h 32m 24s", "memory(GiB)": 58.41, "train_speed(s/it)": 9.254845}
{"loss": 37.25, "grad_norm": 9.88610649, "learning_rate": 0.00040909, "token_acc": 0.2570794, "epoch": 0.0123839, "global_step/max_steps": "27/2181", "elapsed_time": "4m 9s", "remaining_time": "5h 31m 6s", "memory(GiB)": 58.41, "train_speed(s/it)": 9.222714}
{"loss": 36.5, "grad_norm": 13.24606419, "learning_rate": 0.00042424, "token_acc": 0.27224841, "epoch": 0.01284256, "global_step/max_steps": "28/2181", "elapsed_time": "4m 18s", "remaining_time": "5h 31m 5s", "memory(GiB)": 58.41, "train_speed(s/it)": 9.226669}
{"loss": 36.75, "grad_norm": 6.53220034, "learning_rate": 0.00043939, "token_acc": 0.27373374, "epoch": 0.01330123, "global_step/max_steps": "29/2181", "elapsed_time": "4m 28s", "remaining_time": "5h 31m 1s", "memory(GiB)": 58.41, "train_speed(s/it)": 9.228963}
{"loss": 36.5, "grad_norm": 19.41822815, "learning_rate": 0.00045455, "token_acc": 0.27593943, "epoch": 0.01375989, "global_step/max_steps": "30/2181", "elapsed_time": "4m 37s", "remaining_time": "5h 30m 28s", "memory(GiB)": 58.41, "train_speed(s/it)": 9.21793}
{"loss": 37.0, "grad_norm": 8.41702366, "learning_rate": 0.0004697, "token_acc": 0.26605248, "epoch": 0.01421855, "global_step/max_steps": "31/2181", "elapsed_time": "4m 45s", "remaining_time": "5h 29m 29s", "memory(GiB)": 58.41, "train_speed(s/it)": 9.194706}
{"loss": 35.75, "grad_norm": 10.46466446, "learning_rate": 0.00048485, "token_acc": 0.28338576, "epoch": 0.01467722, "global_step/max_steps": "32/2181", "elapsed_time": "4m 54s", "remaining_time": "5h 29m 31s", "memory(GiB)": 58.45, "train_speed(s/it)": 9.19996}
{"loss": 37.25, "grad_norm": 11.5916872, "learning_rate": 0.0005, "token_acc": 0.26104891, "epoch": 0.01513588, "global_step/max_steps": "33/2181", "elapsed_time": "5m 4s", "remaining_time": "5h 29m 37s", "memory(GiB)": 58.45, "train_speed(s/it)": 9.207217}
{"loss": 35.25, "grad_norm": 5.74225712, "learning_rate": 0.00051515, "token_acc": 0.28634821, "epoch": 0.01559454, "global_step/max_steps": "34/2181", "elapsed_time": "5m 13s", "remaining_time": "5h 29m 17s", "memory(GiB)": 58.45, "train_speed(s/it)": 9.201973}
{"loss": 36.25, "grad_norm": 13.98093414, "learning_rate": 0.0005303, "token_acc": 0.25857143, "epoch": 0.0160532, "global_step/max_steps": "35/2181", "elapsed_time": "5m 22s", "remaining_time": "5h 29m 20s", "memory(GiB)": 58.45, "train_speed(s/it)": 9.207638}
{"loss": 36.0, "grad_norm": 10.69714546, "learning_rate": 0.00054545, "token_acc": 0.27213293, "epoch": 0.01651187, "global_step/max_steps": "36/2181", "elapsed_time": "5m 31s", "remaining_time": "5h 28m 54s", "memory(GiB)": 58.45, "train_speed(s/it)": 9.199886}
{"loss": 36.25, "grad_norm": 9.51211357, "learning_rate": 0.00056061, "token_acc": 0.27235326, "epoch": 0.01697053, "global_step/max_steps": "37/2181", "elapsed_time": "5m 41s", "remaining_time": "5h 29m 4s", "memory(GiB)": 58.45, "train_speed(s/it)": 9.209103}
{"loss": 36.0, "grad_norm": 6.66176271, "learning_rate": 0.00057576, "token_acc": 0.27795164, "epoch": 0.01742919, "global_step/max_steps": "38/2181", "elapsed_time": "5m 50s", "remaining_time": "5h 28m 59s", "memory(GiB)": 58.45, "train_speed(s/it)": 9.211005}
{"loss": 36.25, "grad_norm": 7.47650862, "learning_rate": 0.00059091, "token_acc": 0.27588137, "epoch": 0.01788786, "global_step/max_steps": "39/2181", "elapsed_time": "5m 59s", "remaining_time": "5h 28m 54s", "memory(GiB)": 58.45, "train_speed(s/it)": 9.212755}
{"loss": 36.0, "grad_norm": 10.20391464, "learning_rate": 0.00060606, "token_acc": 0.25903084, "epoch": 0.01834652, "global_step/max_steps": "40/2181", "elapsed_time": "6m 9s", "remaining_time": "5h 29m 10s", "memory(GiB)": 58.45, "train_speed(s/it)": 9.224671}
{"loss": 35.5, "grad_norm": 5.62328053, "learning_rate": 0.00062121, "token_acc": 0.27120067, "epoch": 0.01880518, "global_step/max_steps": "41/2181", "elapsed_time": "6m 18s", "remaining_time": "5h 28m 44s", "memory(GiB)": 58.45, "train_speed(s/it)": 9.216846}
{"loss": 36.5, "grad_norm": 15.06546497, "learning_rate": 0.00063636, "token_acc": 0.28984685, "epoch": 0.01926385, "global_step/max_steps": "42/2181", "elapsed_time": "6m 27s", "remaining_time": "5h 28m 15s", "memory(GiB)": 58.45, "train_speed(s/it)": 9.207751}
{"loss": 35.5, "grad_norm": 10.20448875, "learning_rate": 0.00065152, "token_acc": 0.28089888, "epoch": 0.01972251, "global_step/max_steps": "43/2181", "elapsed_time": "6m 36s", "remaining_time": "5h 27m 47s", "memory(GiB)": 58.45, "train_speed(s/it)": 9.198829}
{"loss": 36.0, "grad_norm": 11.27571487, "learning_rate": 0.00066667, "token_acc": 0.2564699, "epoch": 0.02018117, "global_step/max_steps": "44/2181", "elapsed_time": "6m 45s", "remaining_time": "5h 27m 33s", "memory(GiB)": 58.45, "train_speed(s/it)": 9.196405}
{"loss": 34.75, "grad_norm": 5.88678551, "learning_rate": 0.00068182, "token_acc": 0.28716124, "epoch": 0.02063983, "global_step/max_steps": "45/2181", "elapsed_time": "6m 54s", "remaining_time": "5h 27m 45s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.206563}
{"loss": 35.75, "grad_norm": 9.68039894, "learning_rate": 0.00069697, "token_acc": 0.27135678, "epoch": 0.0210985, "global_step/max_steps": "46/2181", "elapsed_time": "7m 3s", "remaining_time": "5h 27m 14s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.19647}
{"loss": 35.0, "grad_norm": 6.00964785, "learning_rate": 0.00071212, "token_acc": 0.28604719, "epoch": 0.02155716, "global_step/max_steps": "47/2181", "elapsed_time": "7m 12s", "remaining_time": "5h 26m 45s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.186807}
{"loss": 35.0, "grad_norm": 9.27452278, "learning_rate": 0.00072727, "token_acc": 0.26200116, "epoch": 0.02201582, "global_step/max_steps": "48/2181", "elapsed_time": "7m 21s", "remaining_time": "5h 26m 52s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.194692}
{"loss": 35.0, "grad_norm": 5.65792274, "learning_rate": 0.00074242, "token_acc": 0.27756757, "epoch": 0.02247449, "global_step/max_steps": "49/2181", "elapsed_time": "7m 31s", "remaining_time": "5h 26m 58s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.201725}
{"loss": 35.0, "grad_norm": 5.44071293, "learning_rate": 0.00075758, "token_acc": 0.2948362, "epoch": 0.02293315, "global_step/max_steps": "50/2181", "elapsed_time": "7m 39s", "remaining_time": "5h 26m 22s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.189194}
{"loss": 34.5, "grad_norm": 7.22914791, "learning_rate": 0.00077273, "token_acc": 0.27447109, "epoch": 0.02339181, "global_step/max_steps": "51/2181", "elapsed_time": "7m 49s", "remaining_time": "5h 26m 33s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.19836}
{"loss": 34.5, "grad_norm": 5.50582457, "learning_rate": 0.00078788, "token_acc": 0.27275329, "epoch": 0.02385048, "global_step/max_steps": "52/2181", "elapsed_time": "7m 59s", "remaining_time": "5h 26m 39s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.205783}
{"loss": 34.25, "grad_norm": 6.35494375, "learning_rate": 0.00080303, "token_acc": 0.29059588, "epoch": 0.02430914, "global_step/max_steps": "53/2181", "elapsed_time": "8m 9s", "remaining_time": "5h 27m 1s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.220413}
{"loss": 34.75, "grad_norm": 4.56995153, "learning_rate": 0.00081818, "token_acc": 0.27932652, "epoch": 0.0247678, "global_step/max_steps": "54/2181", "elapsed_time": "8m 18s", "remaining_time": "5h 26m 37s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.213644}
{"loss": 33.25, "grad_norm": 5.67364788, "learning_rate": 0.00083333, "token_acc": 0.28760331, "epoch": 0.02522646, "global_step/max_steps": "55/2181", "elapsed_time": "8m 27s", "remaining_time": "5h 26m 42s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.220104}
{"loss": 33.5, "grad_norm": 7.90001774, "learning_rate": 0.00084848, "token_acc": 0.311604, "epoch": 0.02568513, "global_step/max_steps": "56/2181", "elapsed_time": "8m 36s", "remaining_time": "5h 26m 34s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.220881}
{"loss": 34.25, "grad_norm": 4.18742561, "learning_rate": 0.00086364, "token_acc": 0.28019053, "epoch": 0.02614379, "global_step/max_steps": "57/2181", "elapsed_time": "8m 46s", "remaining_time": "5h 26m 42s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.228618}
{"loss": 33.75, "grad_norm": 5.14208794, "learning_rate": 0.00087879, "token_acc": 0.28091435, "epoch": 0.02660245, "global_step/max_steps": "58/2181", "elapsed_time": "8m 55s", "remaining_time": "5h 26m 32s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.228467}
{"loss": 34.0, "grad_norm": 5.25924873, "learning_rate": 0.00089394, "token_acc": 0.29904709, "epoch": 0.02706112, "global_step/max_steps": "59/2181", "elapsed_time": "9m 4s", "remaining_time": "5h 26m 12s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.223447}
{"loss": 34.0, "grad_norm": 4.16865301, "learning_rate": 0.00090909, "token_acc": 0.28337237, "epoch": 0.02751978, "global_step/max_steps": "60/2181", "elapsed_time": "9m 14s", "remaining_time": "5h 26m 36s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.239105}
{"loss": 34.0, "grad_norm": 6.02821732, "learning_rate": 0.00092424, "token_acc": 0.27825366, "epoch": 0.02797844, "global_step/max_steps": "61/2181", "elapsed_time": "9m 24s", "remaining_time": "5h 26m 28s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.239609}
{"loss": 33.5, "grad_norm": 7.73227787, "learning_rate": 0.00093939, "token_acc": 0.28470455, "epoch": 0.02843711, "global_step/max_steps": "62/2181", "elapsed_time": "9m 33s", "remaining_time": "5h 26m 25s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.242614}
{"loss": 33.0, "grad_norm": 4.6852932, "learning_rate": 0.00095455, "token_acc": 0.28872832, "epoch": 0.02889577, "global_step/max_steps": "63/2181", "elapsed_time": "9m 42s", "remaining_time": "5h 26m 13s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.241483}
{"loss": 33.25, "grad_norm": 5.42517757, "learning_rate": 0.0009697, "token_acc": 0.29638139, "epoch": 0.02935443, "global_step/max_steps": "64/2181", "elapsed_time": "9m 51s", "remaining_time": "5h 25m 49s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.234127}
{"loss": 33.75, "grad_norm": 4.31262493, "learning_rate": 0.00098485, "token_acc": 0.29322871, "epoch": 0.02981309, "global_step/max_steps": "65/2181", "elapsed_time": "10m 0s", "remaining_time": "5h 25m 43s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.235686}
{"loss": 32.25, "grad_norm": 4.08948803, "learning_rate": 0.001, "token_acc": 0.28990141, "epoch": 0.03027176, "global_step/max_steps": "66/2181", "elapsed_time": "10m 9s", "remaining_time": "5h 25m 24s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.230984}
{"loss": 32.25, "grad_norm": 4.56380939, "learning_rate": 0.001, "token_acc": 0.28463109, "epoch": 0.03073042, "global_step/max_steps": "67/2181", "elapsed_time": "10m 19s", "remaining_time": "5h 25m 26s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.236506}
{"loss": 33.0, "grad_norm": 6.24367428, "learning_rate": 0.001, "token_acc": 0.2962963, "epoch": 0.03118908, "global_step/max_steps": "68/2181", "elapsed_time": "10m 27s", "remaining_time": "5h 24m 55s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.226213}
{"loss": 32.5, "grad_norm": 3.75626087, "learning_rate": 0.001, "token_acc": 0.29094947, "epoch": 0.03164775, "global_step/max_steps": "69/2181", "elapsed_time": "10m 37s", "remaining_time": "5h 24m 44s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.225224}
{"loss": 32.5, "grad_norm": 5.60108662, "learning_rate": 0.00099999, "token_acc": 0.28700821, "epoch": 0.03210641, "global_step/max_steps": "70/2181", "elapsed_time": "10m 47s", "remaining_time": "5h 25m 8s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.240954}
{"loss": 32.75, "grad_norm": 5.0415163, "learning_rate": 0.00099999, "token_acc": 0.29722851, "epoch": 0.03256507, "global_step/max_steps": "71/2181", "elapsed_time": "10m 56s", "remaining_time": "5h 25m 7s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.24489}
{"loss": 32.75, "grad_norm": 4.22080421, "learning_rate": 0.00099998, "token_acc": 0.28899972, "epoch": 0.03302374, "global_step/max_steps": "72/2181", "elapsed_time": "11m 5s", "remaining_time": "5h 24m 52s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.24247}
{"loss": 32.5, "grad_norm": 4.06552124, "learning_rate": 0.00099997, "token_acc": 0.29405131, "epoch": 0.0334824, "global_step/max_steps": "73/2181", "elapsed_time": "11m 15s", "remaining_time": "5h 24m 43s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.24263}
{"loss": 31.625, "grad_norm": 3.30992365, "learning_rate": 0.00099996, "token_acc": 0.29782549, "epoch": 0.03394106, "global_step/max_steps": "74/2181", "elapsed_time": "11m 24s", "remaining_time": "5h 24m 23s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.237124}
{"loss": 30.75, "grad_norm": 3.54664755, "learning_rate": 0.00099996, "token_acc": 0.31747851, "epoch": 0.03439972, "global_step/max_steps": "75/2181", "elapsed_time": "11m 33s", "remaining_time": "5h 24m 14s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.237373}
{"loss": 32.25, "grad_norm": 4.39985132, "learning_rate": 0.00099994, "token_acc": 0.30922551, "epoch": 0.03485839, "global_step/max_steps": "76/2181", "elapsed_time": "11m 42s", "remaining_time": "5h 24m 9s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.239225}
{"loss": 31.75, "grad_norm": 3.61921859, "learning_rate": 0.00099993, "token_acc": 0.29671858, "epoch": 0.03531705, "global_step/max_steps": "77/2181", "elapsed_time": "11m 52s", "remaining_time": "5h 24m 12s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.245167}
{"loss": 33.0, "grad_norm": 3.84003973, "learning_rate": 0.00099992, "token_acc": 0.30471618, "epoch": 0.03577571, "global_step/max_steps": "78/2181", "elapsed_time": "12m 1s", "remaining_time": "5h 23m 49s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.238519}
{"loss": 31.75, "grad_norm": 3.57996964, "learning_rate": 0.00099991, "token_acc": 0.29468466, "epoch": 0.03623438, "global_step/max_steps": "79/2181", "elapsed_time": "12m 10s", "remaining_time": "5h 23m 50s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.24356}
{"loss": 31.0, "grad_norm": 4.4129138, "learning_rate": 0.00099989, "token_acc": 0.30651448, "epoch": 0.03669304, "global_step/max_steps": "80/2181", "elapsed_time": "12m 19s", "remaining_time": "5h 23m 31s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.238755}
{"loss": 32.25, "grad_norm": 5.69743681, "learning_rate": 0.00099988, "token_acc": 0.30803571, "epoch": 0.0371517, "global_step/max_steps": "81/2181", "elapsed_time": "12m 29s", "remaining_time": "5h 23m 27s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.241469}
{"loss": 31.0, "grad_norm": 4.04973984, "learning_rate": 0.00099986, "token_acc": 0.31462212, "epoch": 0.03761037, "global_step/max_steps": "82/2181", "elapsed_time": "12m 38s", "remaining_time": "5h 23m 22s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.243343}
{"loss": 31.375, "grad_norm": 3.37139845, "learning_rate": 0.00099984, "token_acc": 0.29342521, "epoch": 0.03806903, "global_step/max_steps": "83/2181", "elapsed_time": "12m 48s", "remaining_time": "5h 23m 24s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.24864}
{"loss": 31.125, "grad_norm": 5.13362122, "learning_rate": 0.00099982, "token_acc": 0.31567329, "epoch": 0.03852769, "global_step/max_steps": "84/2181", "elapsed_time": "12m 57s", "remaining_time": "5h 23m 22s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.252431}
{"loss": 31.625, "grad_norm": 4.99696922, "learning_rate": 0.0009998, "token_acc": 0.30732123, "epoch": 0.03898635, "global_step/max_steps": "85/2181", "elapsed_time": "13m 6s", "remaining_time": "5h 23m 7s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.249341}
{"loss": 31.25, "grad_norm": 4.64005756, "learning_rate": 0.00099978, "token_acc": 0.30803949, "epoch": 0.03944502, "global_step/max_steps": "86/2181", "elapsed_time": "13m 15s", "remaining_time": "5h 22m 57s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.248974}
{"loss": 31.25, "grad_norm": 3.65870881, "learning_rate": 0.00099976, "token_acc": 0.31212452, "epoch": 0.03990368, "global_step/max_steps": "87/2181", "elapsed_time": "13m 25s", "remaining_time": "5h 22m 48s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.249276}
{"loss": 30.375, "grad_norm": 3.48244357, "learning_rate": 0.00099973, "token_acc": 0.31801625, "epoch": 0.04036234, "global_step/max_steps": "88/2181", "elapsed_time": "13m 34s", "remaining_time": "5h 22m 41s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.250417}
{"loss": 32.0, "grad_norm": 4.16833878, "learning_rate": 0.00099971, "token_acc": 0.29181593, "epoch": 0.04082101, "global_step/max_steps": "89/2181", "elapsed_time": "13m 43s", "remaining_time": "5h 22m 33s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.25108}
{"loss": 31.25, "grad_norm": 4.03060198, "learning_rate": 0.00099968, "token_acc": 0.31536761, "epoch": 0.04127967, "global_step/max_steps": "90/2181", "elapsed_time": "13m 53s", "remaining_time": "5h 22m 33s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.255539}
{"loss": 31.0, "grad_norm": 4.32587337, "learning_rate": 0.00099966, "token_acc": 0.32040472, "epoch": 0.04173833, "global_step/max_steps": "91/2181", "elapsed_time": "14m 2s", "remaining_time": "5h 22m 14s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.250509}
{"loss": 31.625, "grad_norm": 3.16018248, "learning_rate": 0.00099963, "token_acc": 0.29056291, "epoch": 0.042197, "global_step/max_steps": "92/2181", "elapsed_time": "14m 11s", "remaining_time": "5h 22m 12s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.25395}
{"loss": 29.5, "grad_norm": 3.54427552, "learning_rate": 0.0009996, "token_acc": 0.3261236, "epoch": 0.04265566, "global_step/max_steps": "93/2181", "elapsed_time": "14m 21s", "remaining_time": "5h 21m 60s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.252742}
{"loss": 31.25, "grad_norm": 3.52672744, "learning_rate": 0.00099957, "token_acc": 0.30784044, "epoch": 0.04311432, "global_step/max_steps": "94/2181", "elapsed_time": "14m 29s", "remaining_time": "5h 21m 44s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.249831}
{"loss": 31.25, "grad_norm": 3.3421886, "learning_rate": 0.00099954, "token_acc": 0.3130171, "epoch": 0.04357298, "global_step/max_steps": "95/2181", "elapsed_time": "14m 39s", "remaining_time": "5h 21m 38s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.251061}
{"loss": 29.75, "grad_norm": 3.33516598, "learning_rate": 0.0009995, "token_acc": 0.32906348, "epoch": 0.04403165, "global_step/max_steps": "96/2181", "elapsed_time": "14m 48s", "remaining_time": "5h 21m 16s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.245281}
{"loss": 30.75, "grad_norm": 3.051265, "learning_rate": 0.00099947, "token_acc": 0.31127862, "epoch": 0.04449031, "global_step/max_steps": "97/2181", "elapsed_time": "14m 57s", "remaining_time": "5h 21m 4s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.243556}
{"loss": 30.875, "grad_norm": 3.52760649, "learning_rate": 0.00099944, "token_acc": 0.30180305, "epoch": 0.04494897, "global_step/max_steps": "98/2181", "elapsed_time": "15m 6s", "remaining_time": "5h 20m 48s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.240598}
{"loss": 30.75, "grad_norm": 4.2748003, "learning_rate": 0.0009994, "token_acc": 0.31609039, "epoch": 0.04540764, "global_step/max_steps": "99/2181", "elapsed_time": "15m 15s", "remaining_time": "5h 20m 41s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.241425}
{"loss": 30.125, "grad_norm": 3.1007309, "learning_rate": 0.00099936, "token_acc": 0.33177022, "epoch": 0.0458663, "global_step/max_steps": "100/2181", "elapsed_time": "15m 23s", "remaining_time": "5h 20m 17s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.234325}
{"loss": 31.375, "grad_norm": 3.62488246, "learning_rate": 0.00099932, "token_acc": 0.31628788, "epoch": 0.04632496, "global_step/max_steps": "101/2181", "elapsed_time": "15m 33s", "remaining_time": "5h 20m 20s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.24047}
{"loss": 30.5, "grad_norm": 4.68130922, "learning_rate": 0.00099929, "token_acc": 0.30445613, "epoch": 0.04678363, "global_step/max_steps": "102/2181", "elapsed_time": "15m 43s", "remaining_time": "5h 20m 14s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.242089}
{"loss": 29.75, "grad_norm": 3.16213179, "learning_rate": 0.00099925, "token_acc": 0.3252169, "epoch": 0.04724229, "global_step/max_steps": "103/2181", "elapsed_time": "15m 52s", "remaining_time": "5h 20m 3s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.241161}
{"loss": 30.0, "grad_norm": 3.83426738, "learning_rate": 0.0009992, "token_acc": 0.33140496, "epoch": 0.04770095, "global_step/max_steps": "104/2181", "elapsed_time": "16m 2s", "remaining_time": "5h 20m 5s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.246543}
{"loss": 30.0, "grad_norm": 3.40666533, "learning_rate": 0.00099916, "token_acc": 0.30814278, "epoch": 0.04815961, "global_step/max_steps": "105/2181", "elapsed_time": "16m 11s", "remaining_time": "5h 20m 4s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.250428}
{"loss": 29.875, "grad_norm": 2.82480955, "learning_rate": 0.00099912, "token_acc": 0.3142462, "epoch": 0.04861828, "global_step/max_steps": "106/2181", "elapsed_time": "16m 21s", "remaining_time": "5h 20m 2s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.253903}
{"loss": 30.375, "grad_norm": 3.22601032, "learning_rate": 0.00099907, "token_acc": 0.328304, "epoch": 0.04907694, "global_step/max_steps": "107/2181", "elapsed_time": "16m 30s", "remaining_time": "5h 19m 52s", "memory(GiB)": 68.78, "train_speed(s/it)": 9.253562}
{"loss": 30.125, "grad_norm": 3.19071913, "learning_rate": 0.00099903, "token_acc": 0.31848185, "epoch": 0.0495356, "global_step/max_steps": "108/2181", "elapsed_time": "16m 41s", "remaining_time": "5h 20m 15s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.269144}
{"loss": 30.75, "grad_norm": 3.90220165, "learning_rate": 0.00099898, "token_acc": 0.3179516, "epoch": 0.04999427, "global_step/max_steps": "109/2181", "elapsed_time": "16m 50s", "remaining_time": "5h 20m 7s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.26968}
{"loss": 29.75, "grad_norm": 2.86310554, "learning_rate": 0.00099893, "token_acc": 0.31409545, "epoch": 0.05045293, "global_step/max_steps": "110/2181", "elapsed_time": "16m 60s", "remaining_time": "5h 19m 59s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.270373}
{"loss": 29.75, "grad_norm": 3.2280519, "learning_rate": 0.00099888, "token_acc": 0.31086037, "epoch": 0.05091159, "global_step/max_steps": "111/2181", "elapsed_time": "17m 9s", "remaining_time": "5h 19m 50s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.270404}
{"loss": 30.125, "grad_norm": 3.00170255, "learning_rate": 0.00099883, "token_acc": 0.31341601, "epoch": 0.05137026, "global_step/max_steps": "112/2181", "elapsed_time": "17m 18s", "remaining_time": "5h 19m 38s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.269182}
{"loss": 30.375, "grad_norm": 2.84969616, "learning_rate": 0.00099878, "token_acc": 0.32378614, "epoch": 0.05182892, "global_step/max_steps": "113/2181", "elapsed_time": "17m 28s", "remaining_time": "5h 19m 43s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.276328}
{"loss": 29.5, "grad_norm": 3.11547232, "learning_rate": 0.00099873, "token_acc": 0.31430155, "epoch": 0.05228758, "global_step/max_steps": "114/2181", "elapsed_time": "17m 37s", "remaining_time": "5h 19m 28s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.273404}
{"loss": 29.5, "grad_norm": 3.28945184, "learning_rate": 0.00099868, "token_acc": 0.33692875, "epoch": 0.05274624, "global_step/max_steps": "115/2181", "elapsed_time": "17m 46s", "remaining_time": "5h 19m 13s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.270613}
{"loss": 29.75, "grad_norm": 2.80085349, "learning_rate": 0.00099862, "token_acc": 0.34637029, "epoch": 0.05320491, "global_step/max_steps": "116/2181", "elapsed_time": "17m 56s", "remaining_time": "5h 19m 8s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.272842}
{"loss": 29.75, "grad_norm": 2.81532502, "learning_rate": 0.00099857, "token_acc": 0.32038567, "epoch": 0.05366357, "global_step/max_steps": "117/2181", "elapsed_time": "18m 5s", "remaining_time": "5h 19m 2s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.274407}
{"loss": 29.375, "grad_norm": 3.48160195, "learning_rate": 0.00099851, "token_acc": 0.31740891, "epoch": 0.05412223, "global_step/max_steps": "118/2181", "elapsed_time": "18m 14s", "remaining_time": "5h 18m 53s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.274374}
{"loss": 29.875, "grad_norm": 3.10066819, "learning_rate": 0.00099845, "token_acc": 0.31464088, "epoch": 0.0545809, "global_step/max_steps": "119/2181", "elapsed_time": "18m 23s", "remaining_time": "5h 18m 40s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.272441}
{"loss": 30.125, "grad_norm": 3.35798216, "learning_rate": 0.00099839, "token_acc": 0.31805857, "epoch": 0.05503956, "global_step/max_steps": "120/2181", "elapsed_time": "18m 32s", "remaining_time": "5h 18m 27s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.270724}
{"loss": 29.375, "grad_norm": 3.06100035, "learning_rate": 0.00099833, "token_acc": 0.34467497, "epoch": 0.05549822, "global_step/max_steps": "121/2181", "elapsed_time": "18m 41s", "remaining_time": "5h 18m 12s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.267991}
{"loss": 29.875, "grad_norm": 2.87922454, "learning_rate": 0.00099827, "token_acc": 0.31127451, "epoch": 0.05595689, "global_step/max_steps": "122/2181", "elapsed_time": "18m 51s", "remaining_time": "5h 17m 60s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.266439}
{"loss": 30.0, "grad_norm": 2.96935868, "learning_rate": 0.00099821, "token_acc": 0.30864895, "epoch": 0.05641555, "global_step/max_steps": "123/2181", "elapsed_time": "19m 0s", "remaining_time": "5h 18m 0s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.271309}
{"loss": 28.875, "grad_norm": 2.77609372, "learning_rate": 0.00099815, "token_acc": 0.3289436, "epoch": 0.05687421, "global_step/max_steps": "124/2181", "elapsed_time": "19m 10s", "remaining_time": "5h 17m 57s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.274135}
{"loss": 29.375, "grad_norm": 3.35575318, "learning_rate": 0.00099808, "token_acc": 0.33439398, "epoch": 0.05733287, "global_step/max_steps": "125/2181", "elapsed_time": "19m 19s", "remaining_time": "5h 17m 42s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.271346}
{"loss": 29.125, "grad_norm": 3.53538799, "learning_rate": 0.00099802, "token_acc": 0.32962963, "epoch": 0.05779154, "global_step/max_steps": "126/2181", "elapsed_time": "19m 28s", "remaining_time": "5h 17m 33s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.271422}
{"loss": 29.0, "grad_norm": 3.18836713, "learning_rate": 0.00099795, "token_acc": 0.33220626, "epoch": 0.0582502, "global_step/max_steps": "127/2181", "elapsed_time": "19m 38s", "remaining_time": "5h 17m 34s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.276605}
{"loss": 29.25, "grad_norm": 3.69497013, "learning_rate": 0.00099788, "token_acc": 0.32158089, "epoch": 0.05870886, "global_step/max_steps": "128/2181", "elapsed_time": "19m 48s", "remaining_time": "5h 17m 27s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.277678}
{"loss": 29.5, "grad_norm": 3.50063539, "learning_rate": 0.00099781, "token_acc": 0.32777628, "epoch": 0.05916753, "global_step/max_steps": "129/2181", "elapsed_time": "19m 57s", "remaining_time": "5h 17m 22s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.279775}
{"loss": 29.375, "grad_norm": 2.96623826, "learning_rate": 0.00099774, "token_acc": 0.33597378, "epoch": 0.05962619, "global_step/max_steps": "130/2181", "elapsed_time": "20m 7s", "remaining_time": "5h 17m 21s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.283541}
{"loss": 29.625, "grad_norm": 3.69497299, "learning_rate": 0.00099767, "token_acc": 0.32928551, "epoch": 0.06008485, "global_step/max_steps": "131/2181", "elapsed_time": "20m 16s", "remaining_time": "5h 17m 9s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.282633}
{"loss": 29.25, "grad_norm": 3.00477982, "learning_rate": 0.0009976, "token_acc": 0.32593425, "epoch": 0.06054352, "global_step/max_steps": "132/2181", "elapsed_time": "20m 26s", "remaining_time": "5h 17m 9s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.287058}
{"loss": 29.0, "grad_norm": 4.08008242, "learning_rate": 0.00099753, "token_acc": 0.33453337, "epoch": 0.06100218, "global_step/max_steps": "133/2181", "elapsed_time": "20m 35s", "remaining_time": "5h 16m 52s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.283301}
{"loss": 29.0, "grad_norm": 2.99857306, "learning_rate": 0.00099745, "token_acc": 0.33351663, "epoch": 0.06146084, "global_step/max_steps": "134/2181", "elapsed_time": "20m 43s", "remaining_time": "5h 16m 34s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.278796}
{"loss": 29.125, "grad_norm": 3.39391637, "learning_rate": 0.00099738, "token_acc": 0.34273625, "epoch": 0.0619195, "global_step/max_steps": "135/2181", "elapsed_time": "20m 53s", "remaining_time": "5h 16m 24s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.278835}
{"loss": 29.125, "grad_norm": 3.07561159, "learning_rate": 0.0009973, "token_acc": 0.3324993, "epoch": 0.06237817, "global_step/max_steps": "136/2181", "elapsed_time": "21m 3s", "remaining_time": "5h 16m 25s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.283649}
{"loss": 28.625, "grad_norm": 3.23679328, "learning_rate": 0.00099722, "token_acc": 0.33678756, "epoch": 0.06283683, "global_step/max_steps": "137/2181", "elapsed_time": "21m 11s", "remaining_time": "5h 16m 9s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.280093}
{"loss": 29.25, "grad_norm": 3.46473169, "learning_rate": 0.00099714, "token_acc": 0.33130025, "epoch": 0.06329549, "global_step/max_steps": "138/2181", "elapsed_time": "21m 21s", "remaining_time": "5h 16m 9s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.28493}
{"loss": 28.0, "grad_norm": 3.12014413, "learning_rate": 0.00099706, "token_acc": 0.35387943, "epoch": 0.06375416, "global_step/max_steps": "139/2181", "elapsed_time": "21m 31s", "remaining_time": "5h 15m 60s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.284963}
{"loss": 28.875, "grad_norm": 3.03685045, "learning_rate": 0.00099698, "token_acc": 0.33371791, "epoch": 0.06421282, "global_step/max_steps": "140/2181", "elapsed_time": "21m 40s", "remaining_time": "5h 15m 51s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.285129}
{"loss": 28.75, "grad_norm": 3.32607126, "learning_rate": 0.0009969, "token_acc": 0.33700254, "epoch": 0.06467148, "global_step/max_steps": "141/2181", "elapsed_time": "21m 49s", "remaining_time": "5h 15m 43s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.285715}
{"loss": 28.25, "grad_norm": 3.41289377, "learning_rate": 0.00099682, "token_acc": 0.34623955, "epoch": 0.06513015, "global_step/max_steps": "142/2181", "elapsed_time": "21m 59s", "remaining_time": "5h 15m 34s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.286068}
{"loss": 28.75, "grad_norm": 3.31008601, "learning_rate": 0.00099673, "token_acc": 0.33305251, "epoch": 0.06558881, "global_step/max_steps": "143/2181", "elapsed_time": "22m 8s", "remaining_time": "5h 15m 22s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.284446}
{"loss": 28.875, "grad_norm": 3.23485351, "learning_rate": 0.00099665, "token_acc": 0.33408769, "epoch": 0.06604747, "global_step/max_steps": "144/2181", "elapsed_time": "22m 17s", "remaining_time": "5h 15m 10s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.283351}
{"loss": 28.375, "grad_norm": 3.05152321, "learning_rate": 0.00099656, "token_acc": 0.34055388, "epoch": 0.06650613, "global_step/max_steps": "145/2181", "elapsed_time": "22m 26s", "remaining_time": "5h 14m 56s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.280786}
{"loss": 28.5, "grad_norm": 2.97372818, "learning_rate": 0.00099647, "token_acc": 0.33590016, "epoch": 0.0669648, "global_step/max_steps": "146/2181", "elapsed_time": "22m 35s", "remaining_time": "5h 14m 49s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.282129}
{"loss": 27.875, "grad_norm": 3.5231061, "learning_rate": 0.00099639, "token_acc": 0.32807939, "epoch": 0.06742346, "global_step/max_steps": "147/2181", "elapsed_time": "22m 44s", "remaining_time": "5h 14m 33s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.278765}
{"loss": 28.25, "grad_norm": 3.29400086, "learning_rate": 0.0009963, "token_acc": 0.32663589, "epoch": 0.06788212, "global_step/max_steps": "148/2181", "elapsed_time": "22m 53s", "remaining_time": "5h 14m 18s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.27613}
{"loss": 27.125, "grad_norm": 3.29787779, "learning_rate": 0.0009962, "token_acc": 0.34501121, "epoch": 0.06834079, "global_step/max_steps": "149/2181", "elapsed_time": "23m 2s", "remaining_time": "5h 14m 1s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.271954}
{"loss": 27.875, "grad_norm": 3.44203997, "learning_rate": 0.00099611, "token_acc": 0.33054393, "epoch": 0.06879945, "global_step/max_steps": "150/2181", "elapsed_time": "23m 11s", "remaining_time": "5h 13m 50s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.271427}
{"loss": 27.5, "grad_norm": 4.04875946, "learning_rate": 0.00099602, "token_acc": 0.35895317, "epoch": 0.06925811, "global_step/max_steps": "151/2181", "elapsed_time": "23m 20s", "remaining_time": "5h 13m 41s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.271341}
{"loss": 27.875, "grad_norm": 3.74494958, "learning_rate": 0.00099593, "token_acc": 0.34673913, "epoch": 0.06971678, "global_step/max_steps": "152/2181", "elapsed_time": "23m 30s", "remaining_time": "5h 13m 37s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.274057}
{"loss": 27.125, "grad_norm": 3.7186408, "learning_rate": 0.00099583, "token_acc": 0.35694823, "epoch": 0.07017544, "global_step/max_steps": "153/2181", "elapsed_time": "23m 39s", "remaining_time": "5h 13m 30s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.275296}
{"loss": 27.5, "grad_norm": 3.42133403, "learning_rate": 0.00099573, "token_acc": 0.3452381, "epoch": 0.0706341, "global_step/max_steps": "154/2181", "elapsed_time": "23m 48s", "remaining_time": "5h 13m 20s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.274887}
{"loss": 28.375, "grad_norm": 3.68500948, "learning_rate": 0.00099564, "token_acc": 0.33532123, "epoch": 0.07109276, "global_step/max_steps": "155/2181", "elapsed_time": "23m 58s", "remaining_time": "5h 13m 15s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.277007}
{"loss": 28.0, "grad_norm": 3.77110767, "learning_rate": 0.00099554, "token_acc": 0.33396074, "epoch": 0.07155143, "global_step/max_steps": "156/2181", "elapsed_time": "24m 8s", "remaining_time": "5h 13m 12s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.279938}
{"loss": 27.375, "grad_norm": 4.3897233, "learning_rate": 0.00099544, "token_acc": 0.33175882, "epoch": 0.07201009, "global_step/max_steps": "157/2181", "elapsed_time": "24m 17s", "remaining_time": "5h 13m 4s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.28064}
{"loss": 26.625, "grad_norm": 4.33363724, "learning_rate": 0.00099534, "token_acc": 0.35259956, "epoch": 0.07246875, "global_step/max_steps": "158/2181", "elapsed_time": "24m 26s", "remaining_time": "5h 12m 51s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.278784}
{"loss": 27.625, "grad_norm": 5.1987648, "learning_rate": 0.00099524, "token_acc": 0.33996134, "epoch": 0.07292742, "global_step/max_steps": "159/2181", "elapsed_time": "24m 36s", "remaining_time": "5h 12m 45s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.280237}
{"loss": 27.0, "grad_norm": 4.0702529, "learning_rate": 0.00099513, "token_acc": 0.34477656, "epoch": 0.07338608, "global_step/max_steps": "160/2181", "elapsed_time": "24m 45s", "remaining_time": "5h 12m 31s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.278175}
{"loss": 26.875, "grad_norm": 3.43040347, "learning_rate": 0.00099503, "token_acc": 0.34544406, "epoch": 0.07384474, "global_step/max_steps": "161/2181", "elapsed_time": "24m 54s", "remaining_time": "5h 12m 21s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.27773}
{"loss": 27.125, "grad_norm": 3.59246111, "learning_rate": 0.00099493, "token_acc": 0.35525261, "epoch": 0.07430341, "global_step/max_steps": "162/2181", "elapsed_time": "25m 4s", "remaining_time": "5h 12m 19s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.281203}
{"loss": 27.0, "grad_norm": 3.58788848, "learning_rate": 0.00099482, "token_acc": 0.34979897, "epoch": 0.07476207, "global_step/max_steps": "163/2181", "elapsed_time": "25m 12s", "remaining_time": "5h 12m 4s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.278305}
{"loss": 27.5, "grad_norm": 3.60094142, "learning_rate": 0.00099471, "token_acc": 0.34439834, "epoch": 0.07522073, "global_step/max_steps": "164/2181", "elapsed_time": "25m 22s", "remaining_time": "5h 12m 4s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.282955}
{"loss": 27.125, "grad_norm": 3.62033796, "learning_rate": 0.0009946, "token_acc": 0.34577181, "epoch": 0.07567939, "global_step/max_steps": "165/2181", "elapsed_time": "25m 32s", "remaining_time": "5h 11m 55s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.283398}
{"loss": 27.125, "grad_norm": 3.48065829, "learning_rate": 0.00099449, "token_acc": 0.36220044, "epoch": 0.07613806, "global_step/max_steps": "166/2181", "elapsed_time": "25m 41s", "remaining_time": "5h 11m 44s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.282419}
{"loss": 27.125, "grad_norm": 3.74368525, "learning_rate": 0.00099438, "token_acc": 0.35959374, "epoch": 0.07659672, "global_step/max_steps": "167/2181", "elapsed_time": "25m 50s", "remaining_time": "5h 11m 36s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.283033}
{"loss": 26.875, "grad_norm": 2.96619105, "learning_rate": 0.00099427, "token_acc": 0.35076923, "epoch": 0.07705538, "global_step/max_steps": "168/2181", "elapsed_time": "25m 59s", "remaining_time": "5h 11m 26s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.282415}
{"loss": 26.125, "grad_norm": 3.68601012, "learning_rate": 0.00099416, "token_acc": 0.36257143, "epoch": 0.07751405, "global_step/max_steps": "169/2181", "elapsed_time": "26m 8s", "remaining_time": "5h 11m 11s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.279946}
{"loss": 26.75, "grad_norm": 3.33276844, "learning_rate": 0.00099405, "token_acc": 0.36019737, "epoch": 0.07797271, "global_step/max_steps": "170/2181", "elapsed_time": "26m 18s", "remaining_time": "5h 11m 2s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.279862}
{"loss": 26.75, "grad_norm": 4.05044127, "learning_rate": 0.00099393, "token_acc": 0.35580942, "epoch": 0.07843137, "global_step/max_steps": "171/2181", "elapsed_time": "26m 27s", "remaining_time": "5h 10m 51s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.278966}
{"loss": 26.25, "grad_norm": 3.1207633, "learning_rate": 0.00099382, "token_acc": 0.37095875, "epoch": 0.07889004, "global_step/max_steps": "172/2181", "elapsed_time": "26m 36s", "remaining_time": "5h 10m 44s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.280395}
{"loss": 27.125, "grad_norm": 3.3150084, "learning_rate": 0.0009937, "token_acc": 0.36904434, "epoch": 0.0793487, "global_step/max_steps": "173/2181", "elapsed_time": "26m 45s", "remaining_time": "5h 10m 33s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.279334}
{"loss": 26.0, "grad_norm": 2.99834943, "learning_rate": 0.00099358, "token_acc": 0.36630037, "epoch": 0.07980736, "global_step/max_steps": "174/2181", "elapsed_time": "26m 55s", "remaining_time": "5h 10m 24s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.279559}
{"loss": 27.125, "grad_norm": 2.68177223, "learning_rate": 0.00099346, "token_acc": 0.34602464, "epoch": 0.08026602, "global_step/max_steps": "175/2181", "elapsed_time": "27m 4s", "remaining_time": "5h 10m 15s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.279436}
{"loss": 26.25, "grad_norm": 2.88327789, "learning_rate": 0.00099334, "token_acc": 0.35995404, "epoch": 0.08072469, "global_step/max_steps": "176/2181", "elapsed_time": "27m 14s", "remaining_time": "5h 10m 11s", "memory(GiB)": 68.8, "train_speed(s/it)": 9.282531}
{"loss": 26.25, "grad_norm": 2.74282146, "learning_rate": 0.00099322, "token_acc": 0.3653208, "epoch": 0.08118335, "global_step/max_steps": "177/2181", "elapsed_time": "27m 24s", "remaining_time": "5h 10m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.287192}
{"loss": 26.125, "grad_norm": 2.96162367, "learning_rate": 0.0009931, "token_acc": 0.36040897, "epoch": 0.08164201, "global_step/max_steps": "178/2181", "elapsed_time": "27m 34s", "remaining_time": "5h 10m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.290034}
{"loss": 26.0, "grad_norm": 2.61794186, "learning_rate": 0.00099297, "token_acc": 0.36336977, "epoch": 0.08210068, "global_step/max_steps": "179/2181", "elapsed_time": "27m 43s", "remaining_time": "5h 9m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.290013}
{"loss": 26.0, "grad_norm": 2.66743302, "learning_rate": 0.00099285, "token_acc": 0.37321477, "epoch": 0.08255934, "global_step/max_steps": "180/2181", "elapsed_time": "27m 52s", "remaining_time": "5h 9m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.289641}
{"loss": 25.625, "grad_norm": 2.70880485, "learning_rate": 0.00099272, "token_acc": 0.3703385, "epoch": 0.083018, "global_step/max_steps": "181/2181", "elapsed_time": "28m 2s", "remaining_time": "5h 9m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.290942}
{"loss": 26.125, "grad_norm": 2.82023478, "learning_rate": 0.0009926, "token_acc": 0.37243736, "epoch": 0.08347667, "global_step/max_steps": "182/2181", "elapsed_time": "28m 11s", "remaining_time": "5h 9m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.291248}
{"loss": 26.375, "grad_norm": 2.59464717, "learning_rate": 0.00099247, "token_acc": 0.35078969, "epoch": 0.08393533, "global_step/max_steps": "183/2181", "elapsed_time": "28m 21s", "remaining_time": "5h 9m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.292668}
{"loss": 25.5, "grad_norm": 3.00852275, "learning_rate": 0.00099234, "token_acc": 0.38321995, "epoch": 0.08439399, "global_step/max_steps": "184/2181", "elapsed_time": "28m 30s", "remaining_time": "5h 9m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.290837}
{"loss": 26.25, "grad_norm": 3.1563344, "learning_rate": 0.00099221, "token_acc": 0.3659824, "epoch": 0.08485265, "global_step/max_steps": "185/2181", "elapsed_time": "28m 38s", "remaining_time": "5h 9m 0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.288729}
{"loss": 27.25, "grad_norm": 3.11177731, "learning_rate": 0.00099208, "token_acc": 0.35810432, "epoch": 0.08531132, "global_step/max_steps": "186/2181", "elapsed_time": "28m 48s", "remaining_time": "5h 8m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.290634}
{"loss": 25.125, "grad_norm": 2.71563244, "learning_rate": 0.00099195, "token_acc": 0.38429173, "epoch": 0.08576998, "global_step/max_steps": "187/2181", "elapsed_time": "28m 57s", "remaining_time": "5h 8m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.290923}
{"loss": 26.25, "grad_norm": 2.45415521, "learning_rate": 0.00099181, "token_acc": 0.35154062, "epoch": 0.08622864, "global_step/max_steps": "188/2181", "elapsed_time": "29m 7s", "remaining_time": "5h 8m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.293255}
{"loss": 26.25, "grad_norm": 2.42844963, "learning_rate": 0.00099168, "token_acc": 0.37517434, "epoch": 0.08668731, "global_step/max_steps": "189/2181", "elapsed_time": "29m 16s", "remaining_time": "5h 8m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.290198}
{"loss": 26.5, "grad_norm": 2.55922771, "learning_rate": 0.00099154, "token_acc": 0.35595568, "epoch": 0.08714597, "global_step/max_steps": "190/2181", "elapsed_time": "29m 26s", "remaining_time": "5h 8m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.292814}
{"loss": 26.625, "grad_norm": 2.83015084, "learning_rate": 0.00099141, "token_acc": 0.36198716, "epoch": 0.08760463, "global_step/max_steps": "191/2181", "elapsed_time": "29m 35s", "remaining_time": "5h 8m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.292945}
{"loss": 25.25, "grad_norm": 3.05878592, "learning_rate": 0.00099127, "token_acc": 0.38133705, "epoch": 0.0880633, "global_step/max_steps": "192/2181", "elapsed_time": "29m 44s", "remaining_time": "5h 8m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.291763}
{"loss": 26.625, "grad_norm": 3.31976748, "learning_rate": 0.00099113, "token_acc": 0.36642699, "epoch": 0.08852196, "global_step/max_steps": "193/2181", "elapsed_time": "29m 53s", "remaining_time": "5h 7m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.292221}
{"loss": 26.5, "grad_norm": 2.991328, "learning_rate": 0.00099099, "token_acc": 0.35860307, "epoch": 0.08898062, "global_step/max_steps": "194/2181", "elapsed_time": "30m 3s", "remaining_time": "5h 7m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.291696}
{"loss": 26.625, "grad_norm": 3.13119936, "learning_rate": 0.00099085, "token_acc": 0.3447002, "epoch": 0.08943928, "global_step/max_steps": "195/2181", "elapsed_time": "30m 12s", "remaining_time": "5h 7m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.290122}
{"loss": 26.0, "grad_norm": 2.84622931, "learning_rate": 0.00099071, "token_acc": 0.36779047, "epoch": 0.08989795, "global_step/max_steps": "196/2181", "elapsed_time": "30m 21s", "remaining_time": "5h 7m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.291153}
{"loss": 25.375, "grad_norm": 2.78480458, "learning_rate": 0.00099056, "token_acc": 0.37931034, "epoch": 0.09035661, "global_step/max_steps": "197/2181", "elapsed_time": "30m 30s", "remaining_time": "5h 7m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.289616}
{"loss": 25.375, "grad_norm": 2.48215008, "learning_rate": 0.00099042, "token_acc": 0.37149208, "epoch": 0.09081527, "global_step/max_steps": "198/2181", "elapsed_time": "30m 40s", "remaining_time": "5h 7m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.291373}
{"loss": 26.5, "grad_norm": 2.3912468, "learning_rate": 0.00099027, "token_acc": 0.37099359, "epoch": 0.09127394, "global_step/max_steps": "199/2181", "elapsed_time": "30m 48s", "remaining_time": "5h 6m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.288731}
{"loss": 25.5, "grad_norm": 2.77763557, "learning_rate": 0.00099013, "token_acc": 0.36944444, "epoch": 0.0917326, "global_step/max_steps": "200/2181", "elapsed_time": "30m 58s", "remaining_time": "5h 6m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.289285}
{"loss": 25.375, "grad_norm": 2.59954619, "learning_rate": 0.00098998, "token_acc": 0.36492103, "epoch": 0.09219126, "global_step/max_steps": "201/2181", "elapsed_time": "31m 7s", "remaining_time": "5h 6m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.289242}
{"loss": 25.875, "grad_norm": 2.77800536, "learning_rate": 0.00098983, "token_acc": 0.36778029, "epoch": 0.09264993, "global_step/max_steps": "202/2181", "elapsed_time": "31m 16s", "remaining_time": "5h 6m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.288249}
{"loss": 26.25, "grad_norm": 2.68178988, "learning_rate": 0.00098968, "token_acc": 0.36625068, "epoch": 0.09310859, "global_step/max_steps": "203/2181", "elapsed_time": "31m 26s", "remaining_time": "5h 6m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.291119}
{"loss": 25.5, "grad_norm": 2.49471045, "learning_rate": 0.00098953, "token_acc": 0.37254367, "epoch": 0.09356725, "global_step/max_steps": "204/2181", "elapsed_time": "31m 35s", "remaining_time": "5h 6m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.288443}
{"loss": 26.25, "grad_norm": 3.00087285, "learning_rate": 0.00098938, "token_acc": 0.36397268, "epoch": 0.09402591, "global_step/max_steps": "205/2181", "elapsed_time": "31m 44s", "remaining_time": "5h 5m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.286198}
{"loss": 25.125, "grad_norm": 2.4969523, "learning_rate": 0.00098923, "token_acc": 0.37880867, "epoch": 0.09448458, "global_step/max_steps": "206/2181", "elapsed_time": "31m 54s", "remaining_time": "5h 5m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.290269}
{"loss": 26.125, "grad_norm": 2.44160891, "learning_rate": 0.00098907, "token_acc": 0.35289256, "epoch": 0.09494324, "global_step/max_steps": "207/2181", "elapsed_time": "32m 2s", "remaining_time": "5h 5m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.2869}
{"loss": 25.125, "grad_norm": 2.60167813, "learning_rate": 0.00098892, "token_acc": 0.37316329, "epoch": 0.0954019, "global_step/max_steps": "208/2181", "elapsed_time": "32m 13s", "remaining_time": "5h 5m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.294611}
{"loss": 26.25, "grad_norm": 2.27404118, "learning_rate": 0.00098876, "token_acc": 0.36297943, "epoch": 0.09586057, "global_step/max_steps": "209/2181", "elapsed_time": "32m 22s", "remaining_time": "5h 5m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.293975}
{"loss": 26.125, "grad_norm": 2.49601078, "learning_rate": 0.00098861, "token_acc": 0.36734694, "epoch": 0.09631923, "global_step/max_steps": "210/2181", "elapsed_time": "32m 32s", "remaining_time": "5h 5m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.293347}
{"loss": 26.375, "grad_norm": 2.44581151, "learning_rate": 0.00098845, "token_acc": 0.35949721, "epoch": 0.09677789, "global_step/max_steps": "211/2181", "elapsed_time": "32m 41s", "remaining_time": "5h 5m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.293462}
{"loss": 25.75, "grad_norm": 2.42770672, "learning_rate": 0.00098829, "token_acc": 0.37976738, "epoch": 0.09723656, "global_step/max_steps": "212/2181", "elapsed_time": "32m 52s", "remaining_time": "5h 5m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.30074}
{"loss": 25.625, "grad_norm": 2.25797057, "learning_rate": 0.00098813, "token_acc": 0.3717362, "epoch": 0.09769522, "global_step/max_steps": "213/2181", "elapsed_time": "33m 0s", "remaining_time": "5h 4m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.298053}
{"loss": 26.0, "grad_norm": 2.32051945, "learning_rate": 0.00098797, "token_acc": 0.35802817, "epoch": 0.09815388, "global_step/max_steps": "214/2181", "elapsed_time": "33m 9s", "remaining_time": "5h 4m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.294858}
{"loss": 26.125, "grad_norm": 2.76535201, "learning_rate": 0.0009878, "token_acc": 0.35157159, "epoch": 0.09861254, "global_step/max_steps": "215/2181", "elapsed_time": "33m 20s", "remaining_time": "5h 4m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.300373}
{"loss": 25.75, "grad_norm": 2.29711199, "learning_rate": 0.00098764, "token_acc": 0.37442922, "epoch": 0.09907121, "global_step/max_steps": "216/2181", "elapsed_time": "33m 28s", "remaining_time": "5h 4m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.298126}
{"loss": 25.875, "grad_norm": 2.88779569, "learning_rate": 0.00098748, "token_acc": 0.38214784, "epoch": 0.09952987, "global_step/max_steps": "217/2181", "elapsed_time": "33m 37s", "remaining_time": "5h 4m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.2957}
{"loss": 26.0, "grad_norm": 2.40191555, "learning_rate": 0.00098731, "token_acc": 0.36967632, "epoch": 0.09998853, "global_step/max_steps": "218/2181", "elapsed_time": "33m 46s", "remaining_time": "5h 4m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.295645}
{"loss": 25.875, "grad_norm": 2.29713202, "learning_rate": 0.00098714, "token_acc": 0.36546294, "epoch": 0.1004472, "global_step/max_steps": "219/2181", "elapsed_time": "33m 56s", "remaining_time": "5h 4m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.298305}
{"loss": 26.5, "grad_norm": 2.28123975, "learning_rate": 0.00098698, "token_acc": 0.35664131, "epoch": 0.10090586, "global_step/max_steps": "220/2181", "elapsed_time": "34m 5s", "remaining_time": "5h 3m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.296002}
{"loss": 26.5, "grad_norm": 2.3795383, "learning_rate": 0.00098681, "token_acc": 0.37043189, "epoch": 0.10136452, "global_step/max_steps": "221/2181", "elapsed_time": "34m 14s", "remaining_time": "5h 3m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.294893}
{"loss": 25.75, "grad_norm": 2.1766715, "learning_rate": 0.00098664, "token_acc": 0.3655765, "epoch": 0.10182319, "global_step/max_steps": "222/2181", "elapsed_time": "34m 24s", "remaining_time": "5h 3m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.298364}
{"loss": 25.5, "grad_norm": 2.25565195, "learning_rate": 0.00098647, "token_acc": 0.36862004, "epoch": 0.10228185, "global_step/max_steps": "223/2181", "elapsed_time": "34m 33s", "remaining_time": "5h 3m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.29488}
{"loss": 25.5, "grad_norm": 2.31824255, "learning_rate": 0.00098629, "token_acc": 0.37490134, "epoch": 0.10274051, "global_step/max_steps": "224/2181", "elapsed_time": "34m 42s", "remaining_time": "5h 3m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.294462}
{"loss": 25.5, "grad_norm": 2.41421103, "learning_rate": 0.00098612, "token_acc": 0.38973923, "epoch": 0.10319917, "global_step/max_steps": "225/2181", "elapsed_time": "34m 51s", "remaining_time": "5h 3m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.295411}
{"loss": 25.5, "grad_norm": 2.0925653, "learning_rate": 0.00098595, "token_acc": 0.37783375, "epoch": 0.10365784, "global_step/max_steps": "226/2181", "elapsed_time": "35m 1s", "remaining_time": "5h 2m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.295601}
{"loss": 25.0, "grad_norm": 2.35998631, "learning_rate": 0.00098577, "token_acc": 0.39044159, "epoch": 0.1041165, "global_step/max_steps": "227/2181", "elapsed_time": "35m 11s", "remaining_time": "5h 2m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.298078}
{"loss": 26.125, "grad_norm": 2.23680997, "learning_rate": 0.00098559, "token_acc": 0.34719037, "epoch": 0.10457516, "global_step/max_steps": "228/2181", "elapsed_time": "35m 20s", "remaining_time": "5h 2m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.299987}
{"loss": 25.75, "grad_norm": 2.29078603, "learning_rate": 0.00098542, "token_acc": 0.36958322, "epoch": 0.10503383, "global_step/max_steps": "229/2181", "elapsed_time": "35m 30s", "remaining_time": "5h 2m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.300563}
{"loss": 25.25, "grad_norm": 2.27907538, "learning_rate": 0.00098524, "token_acc": 0.36697513, "epoch": 0.10549249, "global_step/max_steps": "230/2181", "elapsed_time": "35m 40s", "remaining_time": "5h 2m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.302637}
{"loss": 25.875, "grad_norm": 2.33717775, "learning_rate": 0.00098506, "token_acc": 0.36672424, "epoch": 0.10595115, "global_step/max_steps": "231/2181", "elapsed_time": "35m 49s", "remaining_time": "5h 2m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.301835}
{"loss": 26.125, "grad_norm": 2.63008285, "learning_rate": 0.00098488, "token_acc": 0.36196319, "epoch": 0.10640982, "global_step/max_steps": "232/2181", "elapsed_time": "35m 58s", "remaining_time": "5h 2m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.300111}
{"loss": 24.625, "grad_norm": 2.19610381, "learning_rate": 0.0009847, "token_acc": 0.37699218, "epoch": 0.10686848, "global_step/max_steps": "233/2181", "elapsed_time": "36m 7s", "remaining_time": "5h 1m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.298744}
{"loss": 26.375, "grad_norm": 2.4019506, "learning_rate": 0.00098451, "token_acc": 0.36327116, "epoch": 0.10732714, "global_step/max_steps": "234/2181", "elapsed_time": "36m 16s", "remaining_time": "5h 1m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.298837}
{"loss": 25.375, "grad_norm": 2.61027169, "learning_rate": 0.00098433, "token_acc": 0.37201953, "epoch": 0.1077858, "global_step/max_steps": "235/2181", "elapsed_time": "36m 25s", "remaining_time": "5h 1m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.299543}
{"loss": 25.625, "grad_norm": 2.19545984, "learning_rate": 0.00098414, "token_acc": 0.36900079, "epoch": 0.10824447, "global_step/max_steps": "236/2181", "elapsed_time": "36m 35s", "remaining_time": "5h 1m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.300145}
{"loss": 25.875, "grad_norm": 2.27695394, "learning_rate": 0.00098396, "token_acc": 0.37029813, "epoch": 0.10870313, "global_step/max_steps": "237/2181", "elapsed_time": "36m 44s", "remaining_time": "5h 1m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.299423}
{"loss": 25.25, "grad_norm": 2.12901187, "learning_rate": 0.00098377, "token_acc": 0.37230769, "epoch": 0.10916179, "global_step/max_steps": "238/2181", "elapsed_time": "36m 53s", "remaining_time": "5h 1m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.299887}
{"loss": 25.125, "grad_norm": 2.16444635, "learning_rate": 0.00098358, "token_acc": 0.36491128, "epoch": 0.10962046, "global_step/max_steps": "239/2181", "elapsed_time": "37m 2s", "remaining_time": "5h 0m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.298297}
{"loss": 26.125, "grad_norm": 2.48832822, "learning_rate": 0.00098339, "token_acc": 0.35562146, "epoch": 0.11007912, "global_step/max_steps": "240/2181", "elapsed_time": "37m 12s", "remaining_time": "5h 0m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.298652}
{"loss": 25.25, "grad_norm": 2.05288243, "learning_rate": 0.0009832, "token_acc": 0.36264045, "epoch": 0.11053778, "global_step/max_steps": "241/2181", "elapsed_time": "37m 21s", "remaining_time": "5h 0m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.29773}
{"loss": 25.75, "grad_norm": 2.34309745, "learning_rate": 0.00098301, "token_acc": 0.35756385, "epoch": 0.11099645, "global_step/max_steps": "242/2181", "elapsed_time": "37m 30s", "remaining_time": "5h 0m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.296716}
{"loss": 25.25, "grad_norm": 2.2441361, "learning_rate": 0.00098282, "token_acc": 0.38135826, "epoch": 0.11145511, "global_step/max_steps": "243/2181", "elapsed_time": "37m 39s", "remaining_time": "5h 0m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.295677}
{"loss": 26.5, "grad_norm": 2.65306592, "learning_rate": 0.00098262, "token_acc": 0.37462236, "epoch": 0.11191377, "global_step/max_steps": "244/2181", "elapsed_time": "37m 49s", "remaining_time": "5h 0m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.299416}
{"loss": 25.875, "grad_norm": 2.39977264, "learning_rate": 0.00098243, "token_acc": 0.3627535, "epoch": 0.11237243, "global_step/max_steps": "245/2181", "elapsed_time": "37m 58s", "remaining_time": "5h 0m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.298317}
{"loss": 25.75, "grad_norm": 2.3606782, "learning_rate": 0.00098223, "token_acc": 0.35530547, "epoch": 0.1128311, "global_step/max_steps": "246/2181", "elapsed_time": "38m 7s", "remaining_time": "4h 59m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.297307}
{"loss": 25.375, "grad_norm": 2.26597977, "learning_rate": 0.00098204, "token_acc": 0.37517512, "epoch": 0.11328976, "global_step/max_steps": "247/2181", "elapsed_time": "38m 17s", "remaining_time": "4h 59m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.29802}
{"loss": 24.875, "grad_norm": 2.24612474, "learning_rate": 0.00098184, "token_acc": 0.375, "epoch": 0.11374842, "global_step/max_steps": "248/2181", "elapsed_time": "38m 26s", "remaining_time": "4h 59m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.300251}
{"loss": 24.75, "grad_norm": 2.0895834, "learning_rate": 0.00098164, "token_acc": 0.37010078, "epoch": 0.11420709, "global_step/max_steps": "249/2181", "elapsed_time": "38m 35s", "remaining_time": "4h 59m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.298983}
{"loss": 25.0, "grad_norm": 2.37528014, "learning_rate": 0.00098144, "token_acc": 0.37369461, "epoch": 0.11466575, "global_step/max_steps": "250/2181", "elapsed_time": "38m 46s", "remaining_time": "4h 59m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.30265}
{"loss": 25.125, "grad_norm": 2.1619575, "learning_rate": 0.00098124, "token_acc": 0.38567493, "epoch": 0.11512441, "global_step/max_steps": "251/2181", "elapsed_time": "38m 55s", "remaining_time": "4h 59m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.302831}
{"loss": 25.0, "grad_norm": 2.48408675, "learning_rate": 0.00098104, "token_acc": 0.37870472, "epoch": 0.11558308, "global_step/max_steps": "252/2181", "elapsed_time": "39m 5s", "remaining_time": "4h 59m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.305743}
{"loss": 25.125, "grad_norm": 2.40520358, "learning_rate": 0.00098084, "token_acc": 0.37283811, "epoch": 0.11604174, "global_step/max_steps": "253/2181", "elapsed_time": "39m 15s", "remaining_time": "4h 59m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.308534}
{"loss": 25.25, "grad_norm": 2.37741899, "learning_rate": 0.00098063, "token_acc": 0.38687151, "epoch": 0.1165004, "global_step/max_steps": "254/2181", "elapsed_time": "39m 24s", "remaining_time": "4h 58m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.308286}
{"loss": 25.5, "grad_norm": 1.94094777, "learning_rate": 0.00098043, "token_acc": 0.38453008, "epoch": 0.11695906, "global_step/max_steps": "255/2181", "elapsed_time": "39m 34s", "remaining_time": "4h 58m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.309331}
{"loss": 25.5, "grad_norm": 2.27067494, "learning_rate": 0.00098022, "token_acc": 0.35712263, "epoch": 0.11741773, "global_step/max_steps": "256/2181", "elapsed_time": "39m 43s", "remaining_time": "4h 58m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.309163}
{"loss": 25.625, "grad_norm": 2.53515792, "learning_rate": 0.00098001, "token_acc": 0.35283019, "epoch": 0.11787639, "global_step/max_steps": "257/2181", "elapsed_time": "39m 53s", "remaining_time": "4h 58m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.309368}
{"loss": 25.75, "grad_norm": 2.1324048, "learning_rate": 0.0009798, "token_acc": 0.37248505, "epoch": 0.11833505, "global_step/max_steps": "258/2181", "elapsed_time": "40m 2s", "remaining_time": "4h 58m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.3098}
{"loss": 25.125, "grad_norm": 2.23881102, "learning_rate": 0.00097959, "token_acc": 0.37566285, "epoch": 0.11879372, "global_step/max_steps": "259/2181", "elapsed_time": "40m 12s", "remaining_time": "4h 58m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.312278}
{"loss": 25.0, "grad_norm": 2.17015457, "learning_rate": 0.00097938, "token_acc": 0.36303913, "epoch": 0.11925238, "global_step/max_steps": "260/2181", "elapsed_time": "40m 21s", "remaining_time": "4h 58m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.312843}
{"loss": 25.75, "grad_norm": 2.30102515, "learning_rate": 0.00097917, "token_acc": 0.36870897, "epoch": 0.11971104, "global_step/max_steps": "261/2181", "elapsed_time": "40m 30s", "remaining_time": "4h 57m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.311269}
{"loss": 25.5, "grad_norm": 1.92303181, "learning_rate": 0.00097896, "token_acc": 0.35505863, "epoch": 0.12016971, "global_step/max_steps": "262/2181", "elapsed_time": "40m 39s", "remaining_time": "4h 57m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.309943}
{"loss": 24.75, "grad_norm": 2.79741216, "learning_rate": 0.00097875, "token_acc": 0.37474777, "epoch": 0.12062837, "global_step/max_steps": "263/2181", "elapsed_time": "40m 49s", "remaining_time": "4h 57m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.31076}
{"loss": 25.375, "grad_norm": 2.13371229, "learning_rate": 0.00097853, "token_acc": 0.36099351, "epoch": 0.12108703, "global_step/max_steps": "264/2181", "elapsed_time": "40m 58s", "remaining_time": "4h 57m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.309161}
{"loss": 25.75, "grad_norm": 2.41432142, "learning_rate": 0.00097831, "token_acc": 0.36660978, "epoch": 0.12154569, "global_step/max_steps": "265/2181", "elapsed_time": "41m 7s", "remaining_time": "4h 57m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.310295}
{"loss": 26.25, "grad_norm": 2.09107447, "learning_rate": 0.0009781, "token_acc": 0.35980011, "epoch": 0.12200436, "global_step/max_steps": "266/2181", "elapsed_time": "41m 17s", "remaining_time": "4h 57m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.310616}
{"loss": 26.25, "grad_norm": 2.27696633, "learning_rate": 0.00097788, "token_acc": 0.35815217, "epoch": 0.12246302, "global_step/max_steps": "267/2181", "elapsed_time": "41m 26s", "remaining_time": "4h 56m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.309264}
{"loss": 25.875, "grad_norm": 2.27360988, "learning_rate": 0.00097766, "token_acc": 0.35820896, "epoch": 0.12292168, "global_step/max_steps": "268/2181", "elapsed_time": "41m 35s", "remaining_time": "4h 56m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.30869}
{"loss": 25.5, "grad_norm": 2.21320152, "learning_rate": 0.00097744, "token_acc": 0.38013699, "epoch": 0.12338035, "global_step/max_steps": "269/2181", "elapsed_time": "41m 44s", "remaining_time": "4h 56m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.308436}
{"loss": 24.875, "grad_norm": 2.25915933, "learning_rate": 0.00097722, "token_acc": 0.39292786, "epoch": 0.12383901, "global_step/max_steps": "270/2181", "elapsed_time": "41m 53s", "remaining_time": "4h 56m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.306676}
{"loss": 25.0, "grad_norm": 2.52968693, "learning_rate": 0.000977, "token_acc": 0.35921238, "epoch": 0.12429767, "global_step/max_steps": "271/2181", "elapsed_time": "42m 2s", "remaining_time": "4h 56m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.307602}
{"loss": 24.875, "grad_norm": 2.23921227, "learning_rate": 0.00097677, "token_acc": 0.37624051, "epoch": 0.12475634, "global_step/max_steps": "272/2181", "elapsed_time": "42m 12s", "remaining_time": "4h 56m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.309198}
{"loss": 25.5, "grad_norm": 2.32752275, "learning_rate": 0.00097655, "token_acc": 0.38767234, "epoch": 0.125215, "global_step/max_steps": "273/2181", "elapsed_time": "42m 22s", "remaining_time": "4h 56m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.310305}
{"loss": 25.375, "grad_norm": 2.74593425, "learning_rate": 0.00097633, "token_acc": 0.37728638, "epoch": 0.12567366, "global_step/max_steps": "274/2181", "elapsed_time": "42m 31s", "remaining_time": "4h 55m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.31084}
{"loss": 25.125, "grad_norm": 1.85645568, "learning_rate": 0.0009761, "token_acc": 0.35093349, "epoch": 0.12613232, "global_step/max_steps": "275/2181", "elapsed_time": "42m 40s", "remaining_time": "4h 55m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.309623}
{"loss": 25.25, "grad_norm": 2.10418224, "learning_rate": 0.00097587, "token_acc": 0.36581197, "epoch": 0.12659099, "global_step/max_steps": "276/2181", "elapsed_time": "42m 49s", "remaining_time": "4h 55m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.308287}
{"loss": 25.625, "grad_norm": 1.75749171, "learning_rate": 0.00097564, "token_acc": 0.37884982, "epoch": 0.12704965, "global_step/max_steps": "277/2181", "elapsed_time": "42m 59s", "remaining_time": "4h 55m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.311251}
{"loss": 25.0, "grad_norm": 1.85010016, "learning_rate": 0.00097541, "token_acc": 0.36126255, "epoch": 0.12750831, "global_step/max_steps": "278/2181", "elapsed_time": "43m 9s", "remaining_time": "4h 55m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.311206}
{"loss": 25.5, "grad_norm": 2.11826324, "learning_rate": 0.00097518, "token_acc": 0.3784546, "epoch": 0.12796698, "global_step/max_steps": "279/2181", "elapsed_time": "43m 17s", "remaining_time": "4h 55m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.309947}
{"loss": 25.25, "grad_norm": 1.76880741, "learning_rate": 0.00097495, "token_acc": 0.37482117, "epoch": 0.12842564, "global_step/max_steps": "280/2181", "elapsed_time": "43m 27s", "remaining_time": "4h 54m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.309802}
{"loss": 24.375, "grad_norm": 1.92863894, "learning_rate": 0.00097472, "token_acc": 0.39134055, "epoch": 0.1288843, "global_step/max_steps": "281/2181", "elapsed_time": "43m 36s", "remaining_time": "4h 54m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.309921}
{"loss": 25.0, "grad_norm": 2.11874771, "learning_rate": 0.00097448, "token_acc": 0.37272981, "epoch": 0.12934297, "global_step/max_steps": "282/2181", "elapsed_time": "43m 45s", "remaining_time": "4h 54m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.310162}
{"loss": 25.125, "grad_norm": 2.21482301, "learning_rate": 0.00097425, "token_acc": 0.36972973, "epoch": 0.12980163, "global_step/max_steps": "283/2181", "elapsed_time": "43m 54s", "remaining_time": "4h 54m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.308609}
{"loss": 25.375, "grad_norm": 1.91057932, "learning_rate": 0.00097401, "token_acc": 0.36726659, "epoch": 0.13026029, "global_step/max_steps": "284/2181", "elapsed_time": "44m 5s", "remaining_time": "4h 54m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.314774}
{"loss": 24.875, "grad_norm": 1.88264453, "learning_rate": 0.00097378, "token_acc": 0.39313085, "epoch": 0.13071895, "global_step/max_steps": "285/2181", "elapsed_time": "44m 15s", "remaining_time": "4h 54m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.31458}
{"loss": 25.375, "grad_norm": 1.83659446, "learning_rate": 0.00097354, "token_acc": 0.3855096, "epoch": 0.13117762, "global_step/max_steps": "286/2181", "elapsed_time": "44m 24s", "remaining_time": "4h 54m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.315533}
{"loss": 25.5, "grad_norm": 2.21096015, "learning_rate": 0.0009733, "token_acc": 0.37364168, "epoch": 0.13163628, "global_step/max_steps": "287/2181", "elapsed_time": "44m 34s", "remaining_time": "4h 54m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.315666}
{"loss": 24.875, "grad_norm": 1.65168524, "learning_rate": 0.00097306, "token_acc": 0.37091106, "epoch": 0.13209494, "global_step/max_steps": "288/2181", "elapsed_time": "44m 43s", "remaining_time": "4h 53m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.315927}
{"loss": 25.375, "grad_norm": 1.94249356, "learning_rate": 0.00097282, "token_acc": 0.37940529, "epoch": 0.13255361, "global_step/max_steps": "289/2181", "elapsed_time": "44m 51s", "remaining_time": "4h 53m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.313029}
{"loss": 25.875, "grad_norm": 1.92531991, "learning_rate": 0.00097258, "token_acc": 0.37905443, "epoch": 0.13301227, "global_step/max_steps": "290/2181", "elapsed_time": "45m 1s", "remaining_time": "4h 53m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.314786}
{"loss": 26.0, "grad_norm": 1.9514581, "learning_rate": 0.00097233, "token_acc": 0.36448598, "epoch": 0.13347093, "global_step/max_steps": "291/2181", "elapsed_time": "45m 11s", "remaining_time": "4h 53m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.314434}
{"loss": 25.25, "grad_norm": 1.98193109, "learning_rate": 0.00097209, "token_acc": 0.36290784, "epoch": 0.1339296, "global_step/max_steps": "292/2181", "elapsed_time": "45m 20s", "remaining_time": "4h 53m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.314008}
{"loss": 24.5, "grad_norm": 1.79064095, "learning_rate": 0.00097185, "token_acc": 0.38698257, "epoch": 0.13438826, "global_step/max_steps": "293/2181", "elapsed_time": "45m 29s", "remaining_time": "4h 53m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.314238}
{"loss": 25.25, "grad_norm": 1.90904927, "learning_rate": 0.0009716, "token_acc": 0.37139595, "epoch": 0.13484692, "global_step/max_steps": "294/2181", "elapsed_time": "45m 39s", "remaining_time": "4h 53m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.316867}
{"loss": 24.25, "grad_norm": 1.89990044, "learning_rate": 0.00097135, "token_acc": 0.40044126, "epoch": 0.13530558, "global_step/max_steps": "295/2181", "elapsed_time": "45m 49s", "remaining_time": "4h 52m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.317587}
{"loss": 24.75, "grad_norm": 1.87476945, "learning_rate": 0.0009711, "token_acc": 0.38513332, "epoch": 0.13576425, "global_step/max_steps": "296/2181", "elapsed_time": "45m 58s", "remaining_time": "4h 52m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.317733}
{"loss": 25.125, "grad_norm": 1.9074738, "learning_rate": 0.00097085, "token_acc": 0.36638889, "epoch": 0.13622291, "global_step/max_steps": "297/2181", "elapsed_time": "46m 7s", "remaining_time": "4h 52m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.31637}
{"loss": 24.875, "grad_norm": 1.84241629, "learning_rate": 0.0009706, "token_acc": 0.38079378, "epoch": 0.13668157, "global_step/max_steps": "298/2181", "elapsed_time": "46m 16s", "remaining_time": "4h 52m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.314985}
{"loss": 24.875, "grad_norm": 1.73319542, "learning_rate": 0.00097035, "token_acc": 0.37157442, "epoch": 0.13714024, "global_step/max_steps": "299/2181", "elapsed_time": "46m 26s", "remaining_time": "4h 52m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.316776}
{"loss": 24.625, "grad_norm": 2.05450678, "learning_rate": 0.0009701, "token_acc": 0.38474576, "epoch": 0.1375989, "global_step/max_steps": "300/2181", "elapsed_time": "46m 35s", "remaining_time": "4h 52m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.317003}
{"loss": 24.75, "grad_norm": 1.88315618, "learning_rate": 0.00096985, "token_acc": 0.37051793, "epoch": 0.13805756, "global_step/max_steps": "301/2181", "elapsed_time": "46m 44s", "remaining_time": "4h 51m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.317031}
{"loss": 25.625, "grad_norm": 1.90153992, "learning_rate": 0.00096959, "token_acc": 0.36760716, "epoch": 0.13851623, "global_step/max_steps": "302/2181", "elapsed_time": "46m 54s", "remaining_time": "4h 51m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.31937}
{"loss": 25.25, "grad_norm": 1.94753957, "learning_rate": 0.00096934, "token_acc": 0.36306995, "epoch": 0.13897489, "global_step/max_steps": "303/2181", "elapsed_time": "47m 4s", "remaining_time": "4h 51m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.319709}
{"loss": 24.75, "grad_norm": 2.06720924, "learning_rate": 0.00096908, "token_acc": 0.37611111, "epoch": 0.13943355, "global_step/max_steps": "304/2181", "elapsed_time": "47m 14s", "remaining_time": "4h 51m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.321356}
{"loss": 25.125, "grad_norm": 1.91122949, "learning_rate": 0.00096882, "token_acc": 0.37958487, "epoch": 0.13989221, "global_step/max_steps": "305/2181", "elapsed_time": "47m 23s", "remaining_time": "4h 51m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.320489}
{"loss": 25.125, "grad_norm": 1.8552289, "learning_rate": 0.00096856, "token_acc": 0.38144036, "epoch": 0.14035088, "global_step/max_steps": "306/2181", "elapsed_time": "47m 33s", "remaining_time": "4h 51m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.323101}
{"loss": 25.125, "grad_norm": 1.82439613, "learning_rate": 0.0009683, "token_acc": 0.378019, "epoch": 0.14080954, "global_step/max_steps": "307/2181", "elapsed_time": "47m 42s", "remaining_time": "4h 51m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.320984}
{"loss": 25.25, "grad_norm": 1.78592551, "learning_rate": 0.00096804, "token_acc": 0.3720529, "epoch": 0.1412682, "global_step/max_steps": "308/2181", "elapsed_time": "47m 51s", "remaining_time": "4h 50m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.320649}
{"loss": 25.875, "grad_norm": 1.81308043, "learning_rate": 0.00096778, "token_acc": 0.36305732, "epoch": 0.14172687, "global_step/max_steps": "309/2181", "elapsed_time": "48m 0s", "remaining_time": "4h 50m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.320489}
{"loss": 24.75, "grad_norm": 2.16359329, "learning_rate": 0.00096752, "token_acc": 0.37971175, "epoch": 0.14218553, "global_step/max_steps": "310/2181", "elapsed_time": "48m 10s", "remaining_time": "4h 50m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.321436}
{"loss": 25.875, "grad_norm": 1.78135931, "learning_rate": 0.00096725, "token_acc": 0.36376096, "epoch": 0.14264419, "global_step/max_steps": "311/2181", "elapsed_time": "48m 19s", "remaining_time": "4h 50m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.320562}
{"loss": 24.75, "grad_norm": 1.90835822, "learning_rate": 0.00096699, "token_acc": 0.36741389, "epoch": 0.14310286, "global_step/max_steps": "312/2181", "elapsed_time": "48m 29s", "remaining_time": "4h 50m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.322204}
{"loss": 25.5, "grad_norm": 2.11609149, "learning_rate": 0.00096672, "token_acc": 0.36062323, "epoch": 0.14356152, "global_step/max_steps": "313/2181", "elapsed_time": "48m 38s", "remaining_time": "4h 50m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.32254}
{"loss": 25.75, "grad_norm": 2.01651144, "learning_rate": 0.00096646, "token_acc": 0.3566657, "epoch": 0.14402018, "global_step/max_steps": "314/2181", "elapsed_time": "48m 48s", "remaining_time": "4h 50m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.325029}
{"loss": 24.625, "grad_norm": 2.20130849, "learning_rate": 0.00096619, "token_acc": 0.38229704, "epoch": 0.14447884, "global_step/max_steps": "315/2181", "elapsed_time": "48m 58s", "remaining_time": "4h 50m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.327301}
{"loss": 24.5, "grad_norm": 1.79373932, "learning_rate": 0.00096592, "token_acc": 0.3825259, "epoch": 0.14493751, "global_step/max_steps": "316/2181", "elapsed_time": "49m 7s", "remaining_time": "4h 49m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.326876}
{"loss": 25.125, "grad_norm": 2.04108405, "learning_rate": 0.00096565, "token_acc": 0.36908881, "epoch": 0.14539617, "global_step/max_steps": "317/2181", "elapsed_time": "49m 18s", "remaining_time": "4h 49m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.33041}
{"loss": 24.875, "grad_norm": 2.09422016, "learning_rate": 0.00096538, "token_acc": 0.37514029, "epoch": 0.14585483, "global_step/max_steps": "318/2181", "elapsed_time": "49m 27s", "remaining_time": "4h 49m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.328869}
{"loss": 25.5, "grad_norm": 1.91007841, "learning_rate": 0.00096511, "token_acc": 0.37690217, "epoch": 0.1463135, "global_step/max_steps": "319/2181", "elapsed_time": "49m 36s", "remaining_time": "4h 49m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.328893}
{"loss": 25.75, "grad_norm": 2.03102565, "learning_rate": 0.00096483, "token_acc": 0.37002614, "epoch": 0.14677216, "global_step/max_steps": "320/2181", "elapsed_time": "49m 45s", "remaining_time": "4h 49m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.327469}
{"loss": 25.125, "grad_norm": 1.87575245, "learning_rate": 0.00096456, "token_acc": 0.37559876, "epoch": 0.14723082, "global_step/max_steps": "321/2181", "elapsed_time": "49m 54s", "remaining_time": "4h 49m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.327599}
{"loss": 25.375, "grad_norm": 1.82449353, "learning_rate": 0.00096428, "token_acc": 0.37290542, "epoch": 0.14768949, "global_step/max_steps": "322/2181", "elapsed_time": "50m 4s", "remaining_time": "4h 49m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.328305}
{"loss": 25.375, "grad_norm": 1.77426541, "learning_rate": 0.00096401, "token_acc": 0.35934516, "epoch": 0.14814815, "global_step/max_steps": "323/2181", "elapsed_time": "50m 13s", "remaining_time": "4h 48m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.327491}
{"loss": 25.25, "grad_norm": 1.83995128, "learning_rate": 0.00096373, "token_acc": 0.37715223, "epoch": 0.14860681, "global_step/max_steps": "324/2181", "elapsed_time": "50m 22s", "remaining_time": "4h 48m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.327188}
{"loss": 24.125, "grad_norm": 1.88818634, "learning_rate": 0.00096345, "token_acc": 0.3853337, "epoch": 0.14906547, "global_step/max_steps": "325/2181", "elapsed_time": "50m 31s", "remaining_time": "4h 48m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.32609}
{"loss": 25.0, "grad_norm": 2.10813904, "learning_rate": 0.00096317, "token_acc": 0.37546047, "epoch": 0.14952414, "global_step/max_steps": "326/2181", "elapsed_time": "50m 40s", "remaining_time": "4h 48m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.324162}
{"loss": 24.75, "grad_norm": 2.07244039, "learning_rate": 0.00096289, "token_acc": 0.39123196, "epoch": 0.1499828, "global_step/max_steps": "327/2181", "elapsed_time": "50m 49s", "remaining_time": "4h 48m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.324476}
{"loss": 24.875, "grad_norm": 1.83663702, "learning_rate": 0.00096261, "token_acc": 0.38191882, "epoch": 0.15044146, "global_step/max_steps": "328/2181", "elapsed_time": "50m 59s", "remaining_time": "4h 48m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.327624}
{"loss": 25.5, "grad_norm": 1.80286825, "learning_rate": 0.00096233, "token_acc": 0.38138784, "epoch": 0.15090013, "global_step/max_steps": "329/2181", "elapsed_time": "51m 10s", "remaining_time": "4h 47m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.329846}
{"loss": 24.625, "grad_norm": 1.86954367, "learning_rate": 0.00096205, "token_acc": 0.38652678, "epoch": 0.15135879, "global_step/max_steps": "330/2181", "elapsed_time": "51m 18s", "remaining_time": "4h 47m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.328607}
{"loss": 24.25, "grad_norm": 1.87343311, "learning_rate": 0.00096176, "token_acc": 0.38918169, "epoch": 0.15181745, "global_step/max_steps": "331/2181", "elapsed_time": "51m 27s", "remaining_time": "4h 47m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.327503}
{"loss": 24.375, "grad_norm": 1.85669684, "learning_rate": 0.00096148, "token_acc": 0.3902371, "epoch": 0.15227612, "global_step/max_steps": "332/2181", "elapsed_time": "51m 37s", "remaining_time": "4h 47m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.327148}
{"loss": 25.5, "grad_norm": 1.8201313, "learning_rate": 0.00096119, "token_acc": 0.35640138, "epoch": 0.15273478, "global_step/max_steps": "333/2181", "elapsed_time": "51m 45s", "remaining_time": "4h 47m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.32538}
{"loss": 24.25, "grad_norm": 1.89534688, "learning_rate": 0.0009609, "token_acc": 0.38076709, "epoch": 0.15319344, "global_step/max_steps": "334/2181", "elapsed_time": "51m 54s", "remaining_time": "4h 47m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.324589}
{"loss": 24.25, "grad_norm": 1.70649517, "learning_rate": 0.00096061, "token_acc": 0.38976265, "epoch": 0.1536521, "global_step/max_steps": "335/2181", "elapsed_time": "52m 4s", "remaining_time": "4h 46m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.326434}
{"loss": 24.25, "grad_norm": 2.04278445, "learning_rate": 0.00096032, "token_acc": 0.38283456, "epoch": 0.15411077, "global_step/max_steps": "336/2181", "elapsed_time": "52m 13s", "remaining_time": "4h 46m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.324804}
{"loss": 25.0, "grad_norm": 1.78890157, "learning_rate": 0.00096003, "token_acc": 0.38187262, "epoch": 0.15456943, "global_step/max_steps": "337/2181", "elapsed_time": "52m 23s", "remaining_time": "4h 46m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.325736}
{"loss": 25.625, "grad_norm": 1.82443821, "learning_rate": 0.00095974, "token_acc": 0.36786007, "epoch": 0.15502809, "global_step/max_steps": "338/2181", "elapsed_time": "52m 32s", "remaining_time": "4h 46m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.326783}
{"loss": 24.625, "grad_norm": 1.71282446, "learning_rate": 0.00095945, "token_acc": 0.39254935, "epoch": 0.15548676, "global_step/max_steps": "339/2181", "elapsed_time": "52m 42s", "remaining_time": "4h 46m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.327353}
{"loss": 25.0, "grad_norm": 1.65207767, "learning_rate": 0.00095916, "token_acc": 0.36823425, "epoch": 0.15594542, "global_step/max_steps": "340/2181", "elapsed_time": "52m 51s", "remaining_time": "4h 46m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.326786}
{"loss": 24.375, "grad_norm": 2.10906792, "learning_rate": 0.00095886, "token_acc": 0.38457223, "epoch": 0.15640408, "global_step/max_steps": "341/2181", "elapsed_time": "53m 1s", "remaining_time": "4h 46m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.327129}
{"loss": 24.0, "grad_norm": 1.91656625, "learning_rate": 0.00095857, "token_acc": 0.40186916, "epoch": 0.15686275, "global_step/max_steps": "342/2181", "elapsed_time": "53m 10s", "remaining_time": "4h 45m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.32822}
{"loss": 25.75, "grad_norm": 1.85234106, "learning_rate": 0.00095827, "token_acc": 0.36177474, "epoch": 0.15732141, "global_step/max_steps": "343/2181", "elapsed_time": "53m 20s", "remaining_time": "4h 45m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.328001}
{"loss": 24.75, "grad_norm": 1.71080613, "learning_rate": 0.00095797, "token_acc": 0.3800554, "epoch": 0.15778007, "global_step/max_steps": "344/2181", "elapsed_time": "53m 29s", "remaining_time": "4h 45m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.329911}
{"loss": 25.75, "grad_norm": 2.02835989, "learning_rate": 0.00095767, "token_acc": 0.36921783, "epoch": 0.15823873, "global_step/max_steps": "345/2181", "elapsed_time": "53m 38s", "remaining_time": "4h 45m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.328267}
{"loss": 25.0, "grad_norm": 2.29868484, "learning_rate": 0.00095737, "token_acc": 0.36691052, "epoch": 0.1586974, "global_step/max_steps": "346/2181", "elapsed_time": "53m 47s", "remaining_time": "4h 45m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.327895}
{"loss": 25.0, "grad_norm": 2.06141162, "learning_rate": 0.00095707, "token_acc": 0.37332191, "epoch": 0.15915606, "global_step/max_steps": "347/2181", "elapsed_time": "53m 57s", "remaining_time": "4h 45m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.328304}
{"loss": 24.75, "grad_norm": 1.8937186, "learning_rate": 0.00095677, "token_acc": 0.38528139, "epoch": 0.15961472, "global_step/max_steps": "348/2181", "elapsed_time": "54m 6s", "remaining_time": "4h 44m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.328384}
{"loss": 24.625, "grad_norm": 1.74286783, "learning_rate": 0.00095647, "token_acc": 0.3841947, "epoch": 0.16007339, "global_step/max_steps": "349/2181", "elapsed_time": "54m 15s", "remaining_time": "4h 44m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.32799}
{"loss": 25.25, "grad_norm": 1.80728722, "learning_rate": 0.00095617, "token_acc": 0.36780673, "epoch": 0.16053205, "global_step/max_steps": "350/2181", "elapsed_time": "54m 24s", "remaining_time": "4h 44m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.327056}
{"loss": 25.375, "grad_norm": 1.6169486, "learning_rate": 0.00095586, "token_acc": 0.36096624, "epoch": 0.16099071, "global_step/max_steps": "351/2181", "elapsed_time": "54m 35s", "remaining_time": "4h 44m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.330785}
{"loss": 25.125, "grad_norm": 1.68589449, "learning_rate": 0.00095556, "token_acc": 0.38341245, "epoch": 0.16144938, "global_step/max_steps": "352/2181", "elapsed_time": "54m 44s", "remaining_time": "4h 44m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.330445}
{"loss": 24.75, "grad_norm": 1.86078179, "learning_rate": 0.00095525, "token_acc": 0.37100949, "epoch": 0.16190804, "global_step/max_steps": "353/2181", "elapsed_time": "54m 54s", "remaining_time": "4h 44m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.332714}
{"loss": 24.75, "grad_norm": 1.79370642, "learning_rate": 0.00095494, "token_acc": 0.38211608, "epoch": 0.1623667, "global_step/max_steps": "354/2181", "elapsed_time": "55m 3s", "remaining_time": "4h 44m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.331486}
{"loss": 24.875, "grad_norm": 1.9478085, "learning_rate": 0.00095463, "token_acc": 0.375, "epoch": 0.16282536, "global_step/max_steps": "355/2181", "elapsed_time": "55m 13s", "remaining_time": "4h 43m 60s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.331742}
{"loss": 25.0, "grad_norm": 1.88965619, "learning_rate": 0.00095432, "token_acc": 0.36550946, "epoch": 0.16328403, "global_step/max_steps": "356/2181", "elapsed_time": "55m 22s", "remaining_time": "4h 43m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.331995}
{"loss": 25.375, "grad_norm": 1.86808896, "learning_rate": 0.00095401, "token_acc": 0.36967489, "epoch": 0.16374269, "global_step/max_steps": "357/2181", "elapsed_time": "55m 31s", "remaining_time": "4h 43m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.330455}
{"loss": 25.0, "grad_norm": 1.65049386, "learning_rate": 0.0009537, "token_acc": 0.37915743, "epoch": 0.16420135, "global_step/max_steps": "358/2181", "elapsed_time": "55m 40s", "remaining_time": "4h 43m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.329971}
{"loss": 24.75, "grad_norm": 1.74211752, "learning_rate": 0.00095339, "token_acc": 0.36816901, "epoch": 0.16466002, "global_step/max_steps": "359/2181", "elapsed_time": "55m 49s", "remaining_time": "4h 43m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.329969}
{"loss": 24.625, "grad_norm": 1.86372137, "learning_rate": 0.00095308, "token_acc": 0.38146021, "epoch": 0.16511868, "global_step/max_steps": "360/2181", "elapsed_time": "55m 59s", "remaining_time": "4h 43m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.330575}
{"loss": 25.0, "grad_norm": 1.63342929, "learning_rate": 0.00095276, "token_acc": 0.37888027, "epoch": 0.16557734, "global_step/max_steps": "361/2181", "elapsed_time": "56m 8s", "remaining_time": "4h 43m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.330389}
{"loss": 24.375, "grad_norm": 2.13034868, "learning_rate": 0.00095245, "token_acc": 0.39757646, "epoch": 0.16603601, "global_step/max_steps": "362/2181", "elapsed_time": "56m 17s", "remaining_time": "4h 42m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.329304}
{"loss": 24.75, "grad_norm": 1.78066945, "learning_rate": 0.00095213, "token_acc": 0.36877828, "epoch": 0.16649467, "global_step/max_steps": "363/2181", "elapsed_time": "56m 27s", "remaining_time": "4h 42m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.330341}
{"loss": 24.75, "grad_norm": 1.86148214, "learning_rate": 0.00095181, "token_acc": 0.37907762, "epoch": 0.16695333, "global_step/max_steps": "364/2181", "elapsed_time": "56m 36s", "remaining_time": "4h 42m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.330048}
{"loss": 24.5, "grad_norm": 1.86329496, "learning_rate": 0.00095149, "token_acc": 0.3842378, "epoch": 0.16741199, "global_step/max_steps": "365/2181", "elapsed_time": "56m 46s", "remaining_time": "4h 42m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.330198}
{"loss": 24.5, "grad_norm": 1.66425943, "learning_rate": 0.00095117, "token_acc": 0.38499452, "epoch": 0.16787066, "global_step/max_steps": "366/2181", "elapsed_time": "56m 55s", "remaining_time": "4h 42m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.329928}
{"loss": 25.625, "grad_norm": 1.69504821, "learning_rate": 0.00095085, "token_acc": 0.3601765, "epoch": 0.16832932, "global_step/max_steps": "367/2181", "elapsed_time": "57m 5s", "remaining_time": "4h 42m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.331413}
{"loss": 25.125, "grad_norm": 1.73157442, "learning_rate": 0.00095053, "token_acc": 0.36993883, "epoch": 0.16878798, "global_step/max_steps": "368/2181", "elapsed_time": "57m 14s", "remaining_time": "4h 41m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.331743}
{"loss": 24.5, "grad_norm": 1.89332283, "learning_rate": 0.00095021, "token_acc": 0.38846907, "epoch": 0.16924665, "global_step/max_steps": "369/2181", "elapsed_time": "57m 23s", "remaining_time": "4h 41m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.330924}
{"loss": 24.875, "grad_norm": 1.67670131, "learning_rate": 0.00094988, "token_acc": 0.37472648, "epoch": 0.16970531, "global_step/max_steps": "370/2181", "elapsed_time": "57m 32s", "remaining_time": "4h 41m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.330085}
{"loss": 25.0, "grad_norm": 2.07303262, "learning_rate": 0.00094956, "token_acc": 0.37155833, "epoch": 0.17016397, "global_step/max_steps": "371/2181", "elapsed_time": "57m 41s", "remaining_time": "4h 41m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.329522}
{"loss": 25.0, "grad_norm": 1.7115463, "learning_rate": 0.00094923, "token_acc": 0.38934203, "epoch": 0.17062264, "global_step/max_steps": "372/2181", "elapsed_time": "57m 51s", "remaining_time": "4h 41m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.330618}
{"loss": 25.25, "grad_norm": 1.67472219, "learning_rate": 0.00094891, "token_acc": 0.37545328, "epoch": 0.1710813, "global_step/max_steps": "373/2181", "elapsed_time": "58m 0s", "remaining_time": "4h 41m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.33007}
{"loss": 24.25, "grad_norm": 1.82361782, "learning_rate": 0.00094858, "token_acc": 0.37723214, "epoch": 0.17153996, "global_step/max_steps": "374/2181", "elapsed_time": "58m 10s", "remaining_time": "4h 41m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.332641}
{"loss": 25.125, "grad_norm": 1.59564829, "learning_rate": 0.00094825, "token_acc": 0.37003311, "epoch": 0.17199862, "global_step/max_steps": "375/2181", "elapsed_time": "58m 20s", "remaining_time": "4h 40m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.333709}
{"loss": 25.375, "grad_norm": 1.68032885, "learning_rate": 0.00094792, "token_acc": 0.3830303, "epoch": 0.17245729, "global_step/max_steps": "376/2181", "elapsed_time": "58m 29s", "remaining_time": "4h 40m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.333733}
{"loss": 24.625, "grad_norm": 1.58902109, "learning_rate": 0.00094759, "token_acc": 0.38391877, "epoch": 0.17291595, "global_step/max_steps": "377/2181", "elapsed_time": "58m 40s", "remaining_time": "4h 40m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.335699}
{"loss": 24.0, "grad_norm": 1.56094992, "learning_rate": 0.00094726, "token_acc": 0.38292616, "epoch": 0.17337461, "global_step/max_steps": "378/2181", "elapsed_time": "58m 49s", "remaining_time": "4h 40m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.336442}
{"loss": 24.25, "grad_norm": 2.05067086, "learning_rate": 0.00094693, "token_acc": 0.38162931, "epoch": 0.17383328, "global_step/max_steps": "379/2181", "elapsed_time": "58m 58s", "remaining_time": "4h 40m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.336372}
{"loss": 24.625, "grad_norm": 1.69920981, "learning_rate": 0.00094659, "token_acc": 0.38093923, "epoch": 0.17429194, "global_step/max_steps": "380/2181", "elapsed_time": "59m 8s", "remaining_time": "4h 40m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.337703}
{"loss": 25.0, "grad_norm": 2.32598543, "learning_rate": 0.00094626, "token_acc": 0.38835227, "epoch": 0.1747506, "global_step/max_steps": "381/2181", "elapsed_time": "59m 17s", "remaining_time": "4h 40m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.335988}
{"loss": 24.75, "grad_norm": 1.72064948, "learning_rate": 0.00094592, "token_acc": 0.38581798, "epoch": 0.17520926, "global_step/max_steps": "382/2181", "elapsed_time": "59m 26s", "remaining_time": "4h 39m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.335452}
{"loss": 24.875, "grad_norm": 2.06569123, "learning_rate": 0.00094559, "token_acc": 0.36815359, "epoch": 0.17566793, "global_step/max_steps": "383/2181", "elapsed_time": "59m 36s", "remaining_time": "4h 39m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.336241}
{"loss": 24.875, "grad_norm": 1.65923178, "learning_rate": 0.00094525, "token_acc": 0.37085907, "epoch": 0.17612659, "global_step/max_steps": "384/2181", "elapsed_time": "59m 45s", "remaining_time": "4h 39m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.336205}
{"loss": 24.0, "grad_norm": 1.95258236, "learning_rate": 0.00094491, "token_acc": 0.39943503, "epoch": 0.17658525, "global_step/max_steps": "385/2181", "elapsed_time": "59m 54s", "remaining_time": "4h 39m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.33449}
{"loss": 25.125, "grad_norm": 1.83598936, "learning_rate": 0.00094457, "token_acc": 0.36540061, "epoch": 0.17704392, "global_step/max_steps": "386/2181", "elapsed_time": "1h 0m 3s", "remaining_time": "4h 39m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.332977}
{"loss": 24.5, "grad_norm": 2.23117638, "learning_rate": 0.00094423, "token_acc": 0.37697253, "epoch": 0.17750258, "global_step/max_steps": "387/2181", "elapsed_time": "1h 0m 12s", "remaining_time": "4h 39m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.333953}
{"loss": 24.75, "grad_norm": 1.8518393, "learning_rate": 0.00094389, "token_acc": 0.37063319, "epoch": 0.17796124, "global_step/max_steps": "388/2181", "elapsed_time": "1h 0m 21s", "remaining_time": "4h 38m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.332996}
{"loss": 25.0, "grad_norm": 1.90750313, "learning_rate": 0.00094355, "token_acc": 0.39229471, "epoch": 0.17841991, "global_step/max_steps": "389/2181", "elapsed_time": "1h 0m 31s", "remaining_time": "4h 38m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.333565}
{"loss": 24.875, "grad_norm": 1.71652949, "learning_rate": 0.0009432, "token_acc": 0.37852949, "epoch": 0.17887857, "global_step/max_steps": "390/2181", "elapsed_time": "1h 0m 41s", "remaining_time": "4h 38m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.334995}
{"loss": 24.625, "grad_norm": 2.01857352, "learning_rate": 0.00094286, "token_acc": 0.37688157, "epoch": 0.17933723, "global_step/max_steps": "391/2181", "elapsed_time": "1h 0m 49s", "remaining_time": "4h 38m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.33315}
{"loss": 24.25, "grad_norm": 1.72050464, "learning_rate": 0.00094252, "token_acc": 0.37409236, "epoch": 0.17979589, "global_step/max_steps": "392/2181", "elapsed_time": "1h 0m 59s", "remaining_time": "4h 38m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.333677}
{"loss": 25.0, "grad_norm": 2.03569436, "learning_rate": 0.00094217, "token_acc": 0.39312448, "epoch": 0.18025456, "global_step/max_steps": "393/2181", "elapsed_time": "1h 1m 8s", "remaining_time": "4h 38m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.334174}
{"loss": 25.5, "grad_norm": 1.74025309, "learning_rate": 0.00094182, "token_acc": 0.37101367, "epoch": 0.18071322, "global_step/max_steps": "394/2181", "elapsed_time": "1h 1m 19s", "remaining_time": "4h 38m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.336452}
{"loss": 25.5, "grad_norm": 1.99984932, "learning_rate": 0.00094147, "token_acc": 0.3652198, "epoch": 0.18117188, "global_step/max_steps": "395/2181", "elapsed_time": "1h 1m 28s", "remaining_time": "4h 37m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.336029}
{"loss": 24.5, "grad_norm": 1.8900547, "learning_rate": 0.00094112, "token_acc": 0.37857775, "epoch": 0.18163055, "global_step/max_steps": "396/2181", "elapsed_time": "1h 1m 37s", "remaining_time": "4h 37m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.33482}
{"loss": 24.75, "grad_norm": 2.04708123, "learning_rate": 0.00094077, "token_acc": 0.39327366, "epoch": 0.18208921, "global_step/max_steps": "397/2181", "elapsed_time": "1h 1m 46s", "remaining_time": "4h 37m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.335832}
{"loss": 25.75, "grad_norm": 1.87279618, "learning_rate": 0.00094042, "token_acc": 0.36040897, "epoch": 0.18254787, "global_step/max_steps": "398/2181", "elapsed_time": "1h 1m 56s", "remaining_time": "4h 37m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.33604}
{"loss": 24.625, "grad_norm": 1.95989394, "learning_rate": 0.00094007, "token_acc": 0.37211886, "epoch": 0.18300654, "global_step/max_steps": "399/2181", "elapsed_time": "1h 2m 5s", "remaining_time": "4h 37m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.335757}
{"loss": 24.75, "grad_norm": 1.96930444, "learning_rate": 0.00093972, "token_acc": 0.36785418, "epoch": 0.1834652, "global_step/max_steps": "400/2181", "elapsed_time": "1h 2m 15s", "remaining_time": "4h 37m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.336762}
{"loss": 25.0, "grad_norm": 1.91017461, "learning_rate": 0.00093936, "token_acc": 0.38231178, "epoch": 0.18392386, "global_step/max_steps": "401/2181", "elapsed_time": "1h 2m 23s", "remaining_time": "4h 36m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.334861}
{"loss": 25.5, "grad_norm": 2.07119441, "learning_rate": 0.00093901, "token_acc": 0.38092613, "epoch": 0.18438252, "global_step/max_steps": "402/2181", "elapsed_time": "1h 2m 33s", "remaining_time": "4h 36m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.336023}
{"loss": 25.875, "grad_norm": 1.67363107, "learning_rate": 0.00093865, "token_acc": 0.35200669, "epoch": 0.18484119, "global_step/max_steps": "403/2181", "elapsed_time": "1h 2m 42s", "remaining_time": "4h 36m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.335413}
{"loss": 24.5, "grad_norm": 2.12345123, "learning_rate": 0.0009383, "token_acc": 0.36516094, "epoch": 0.18529985, "global_step/max_steps": "404/2181", "elapsed_time": "1h 2m 52s", "remaining_time": "4h 36m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.335457}
{"loss": 24.75, "grad_norm": 1.73231173, "learning_rate": 0.00093794, "token_acc": 0.37435897, "epoch": 0.18575851, "global_step/max_steps": "405/2181", "elapsed_time": "1h 3m 1s", "remaining_time": "4h 36m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.335167}
{"loss": 25.125, "grad_norm": 1.70412576, "learning_rate": 0.00093758, "token_acc": 0.38626126, "epoch": 0.18621718, "global_step/max_steps": "406/2181", "elapsed_time": "1h 3m 10s", "remaining_time": "4h 36m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.335267}
{"loss": 25.0, "grad_norm": 1.90326178, "learning_rate": 0.00093722, "token_acc": 0.38996355, "epoch": 0.18667584, "global_step/max_steps": "407/2181", "elapsed_time": "1h 3m 20s", "remaining_time": "4h 36m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.336666}
{"loss": 24.75, "grad_norm": 1.61647761, "learning_rate": 0.00093686, "token_acc": 0.37831669, "epoch": 0.1871345, "global_step/max_steps": "408/2181", "elapsed_time": "1h 3m 30s", "remaining_time": "4h 35m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.339001}
{"loss": 24.25, "grad_norm": 1.79464328, "learning_rate": 0.0009365, "token_acc": 0.37429264, "epoch": 0.18759317, "global_step/max_steps": "409/2181", "elapsed_time": "1h 3m 40s", "remaining_time": "4h 35m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.338679}
{"loss": 24.5, "grad_norm": 1.63996089, "learning_rate": 0.00093613, "token_acc": 0.37832594, "epoch": 0.18805183, "global_step/max_steps": "410/2181", "elapsed_time": "1h 3m 49s", "remaining_time": "4h 35m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.338381}
{"loss": 25.5, "grad_norm": 2.03604198, "learning_rate": 0.00093577, "token_acc": 0.3598862, "epoch": 0.18851049, "global_step/max_steps": "411/2181", "elapsed_time": "1h 3m 58s", "remaining_time": "4h 35m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.337545}
{"loss": 24.5, "grad_norm": 1.64743376, "learning_rate": 0.00093541, "token_acc": 0.37843908, "epoch": 0.18896915, "global_step/max_steps": "412/2181", "elapsed_time": "1h 4m 8s", "remaining_time": "4h 35m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.338863}
{"loss": 24.5, "grad_norm": 2.13878083, "learning_rate": 0.00093504, "token_acc": 0.38148045, "epoch": 0.18942782, "global_step/max_steps": "413/2181", "elapsed_time": "1h 4m 17s", "remaining_time": "4h 35m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.33821}
{"loss": 25.375, "grad_norm": 2.08074188, "learning_rate": 0.00093467, "token_acc": 0.36634483, "epoch": 0.18988648, "global_step/max_steps": "414/2181", "elapsed_time": "1h 4m 25s", "remaining_time": "4h 34m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.336635}
{"loss": 25.5, "grad_norm": 1.61971033, "learning_rate": 0.00093431, "token_acc": 0.37625101, "epoch": 0.19034514, "global_step/max_steps": "415/2181", "elapsed_time": "1h 4m 34s", "remaining_time": "4h 34m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.335228}
{"loss": 24.875, "grad_norm": 1.78928816, "learning_rate": 0.00093394, "token_acc": 0.37946429, "epoch": 0.19080381, "global_step/max_steps": "416/2181", "elapsed_time": "1h 4m 44s", "remaining_time": "4h 34m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.336716}
{"loss": 25.0, "grad_norm": 1.66476536, "learning_rate": 0.00093357, "token_acc": 0.36498258, "epoch": 0.19126247, "global_step/max_steps": "417/2181", "elapsed_time": "1h 4m 53s", "remaining_time": "4h 34m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.335787}
{"loss": 26.0, "grad_norm": 1.8839159, "learning_rate": 0.0009332, "token_acc": 0.35280962, "epoch": 0.19172113, "global_step/max_steps": "418/2181", "elapsed_time": "1h 5m 2s", "remaining_time": "4h 34m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.335736}
{"loss": 24.25, "grad_norm": 1.67836452, "learning_rate": 0.00093283, "token_acc": 0.38985466, "epoch": 0.1921798, "global_step/max_steps": "419/2181", "elapsed_time": "1h 5m 12s", "remaining_time": "4h 34m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.336445}
{"loss": 24.5, "grad_norm": 1.72101498, "learning_rate": 0.00093245, "token_acc": 0.36440439, "epoch": 0.19263846, "global_step/max_steps": "420/2181", "elapsed_time": "1h 5m 21s", "remaining_time": "4h 34m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.33605}
{"loss": 23.75, "grad_norm": 1.69575572, "learning_rate": 0.00093208, "token_acc": 0.38126905, "epoch": 0.19309712, "global_step/max_steps": "421/2181", "elapsed_time": "1h 5m 30s", "remaining_time": "4h 33m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.33563}
{"loss": 24.5, "grad_norm": 1.90107167, "learning_rate": 0.00093171, "token_acc": 0.36215151, "epoch": 0.19355578, "global_step/max_steps": "422/2181", "elapsed_time": "1h 5m 39s", "remaining_time": "4h 33m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.335116}
{"loss": 25.25, "grad_norm": 1.77563417, "learning_rate": 0.00093133, "token_acc": 0.36049525, "epoch": 0.19401445, "global_step/max_steps": "423/2181", "elapsed_time": "1h 5m 48s", "remaining_time": "4h 33m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.334116}
{"loss": 24.625, "grad_norm": 1.6908834, "learning_rate": 0.00093096, "token_acc": 0.37626044, "epoch": 0.19447311, "global_step/max_steps": "424/2181", "elapsed_time": "1h 5m 58s", "remaining_time": "4h 33m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.334309}
{"loss": 24.25, "grad_norm": 1.6772989, "learning_rate": 0.00093058, "token_acc": 0.37964276, "epoch": 0.19493177, "global_step/max_steps": "425/2181", "elapsed_time": "1h 6m 7s", "remaining_time": "4h 33m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.334168}
{"loss": 25.25, "grad_norm": 1.66655195, "learning_rate": 0.0009302, "token_acc": 0.35985381, "epoch": 0.19539044, "global_step/max_steps": "426/2181", "elapsed_time": "1h 6m 17s", "remaining_time": "4h 33m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.335689}
{"loss": 25.25, "grad_norm": 1.62572205, "learning_rate": 0.00092982, "token_acc": 0.36649944, "epoch": 0.1958491, "global_step/max_steps": "427/2181", "elapsed_time": "1h 6m 27s", "remaining_time": "4h 32m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.336872}
{"loss": 24.75, "grad_norm": 1.55806816, "learning_rate": 0.00092944, "token_acc": 0.37954939, "epoch": 0.19630776, "global_step/max_steps": "428/2181", "elapsed_time": "1h 6m 36s", "remaining_time": "4h 32m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.336667}
{"loss": 24.5, "grad_norm": 1.55066359, "learning_rate": 0.00092906, "token_acc": 0.37721079, "epoch": 0.19676643, "global_step/max_steps": "429/2181", "elapsed_time": "1h 6m 45s", "remaining_time": "4h 32m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.33621}
{"loss": 25.5, "grad_norm": 1.45854032, "learning_rate": 0.00092868, "token_acc": 0.36989592, "epoch": 0.19722509, "global_step/max_steps": "430/2181", "elapsed_time": "1h 6m 55s", "remaining_time": "4h 32m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.336881}
{"loss": 24.875, "grad_norm": 1.80034053, "learning_rate": 0.0009283, "token_acc": 0.37482199, "epoch": 0.19768375, "global_step/max_steps": "431/2181", "elapsed_time": "1h 7m 4s", "remaining_time": "4h 32m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.337066}
{"loss": 25.25, "grad_norm": 1.6297909, "learning_rate": 0.00092791, "token_acc": 0.37971953, "epoch": 0.19814241, "global_step/max_steps": "432/2181", "elapsed_time": "1h 7m 14s", "remaining_time": "4h 32m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.337906}
{"loss": 24.125, "grad_norm": 1.61176288, "learning_rate": 0.00092753, "token_acc": 0.40800225, "epoch": 0.19860108, "global_step/max_steps": "433/2181", "elapsed_time": "1h 7m 23s", "remaining_time": "4h 32m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.338083}
{"loss": 24.75, "grad_norm": 1.66384459, "learning_rate": 0.00092714, "token_acc": 0.38048644, "epoch": 0.19905974, "global_step/max_steps": "434/2181", "elapsed_time": "1h 7m 33s", "remaining_time": "4h 31m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.33786}
{"loss": 25.25, "grad_norm": 1.82762682, "learning_rate": 0.00092676, "token_acc": 0.37599545, "epoch": 0.1995184, "global_step/max_steps": "435/2181", "elapsed_time": "1h 7m 42s", "remaining_time": "4h 31m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.338205}
{"loss": 24.25, "grad_norm": 1.5722909, "learning_rate": 0.00092637, "token_acc": 0.38801351, "epoch": 0.19997707, "global_step/max_steps": "436/2181", "elapsed_time": "1h 7m 52s", "remaining_time": "4h 31m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.339308}
{"loss": 24.875, "grad_norm": 1.79010665, "learning_rate": 0.00092598, "token_acc": 0.38521837, "epoch": 0.20043573, "global_step/max_steps": "437/2181", "elapsed_time": "1h 8m 1s", "remaining_time": "4h 31m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.338922}
{"loss": 24.375, "grad_norm": 1.60552335, "learning_rate": 0.00092559, "token_acc": 0.37174612, "epoch": 0.20089439, "global_step/max_steps": "438/2181", "elapsed_time": "1h 8m 11s", "remaining_time": "4h 31m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.339444}
{"loss": 24.25, "grad_norm": 1.96005118, "learning_rate": 0.0009252, "token_acc": 0.38670166, "epoch": 0.20135306, "global_step/max_steps": "439/2181", "elapsed_time": "1h 8m 21s", "remaining_time": "4h 31m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.340618}
{"loss": 24.5, "grad_norm": 1.62780821, "learning_rate": 0.00092481, "token_acc": 0.39397321, "epoch": 0.20181172, "global_step/max_steps": "440/2181", "elapsed_time": "1h 8m 29s", "remaining_time": "4h 31m 0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.339489}
{"loss": 24.375, "grad_norm": 2.01034999, "learning_rate": 0.00092442, "token_acc": 0.38884331, "epoch": 0.20227038, "global_step/max_steps": "441/2181", "elapsed_time": "1h 8m 38s", "remaining_time": "4h 30m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.338769}
{"loss": 24.625, "grad_norm": 1.53948438, "learning_rate": 0.00092402, "token_acc": 0.382287, "epoch": 0.20272904, "global_step/max_steps": "442/2181", "elapsed_time": "1h 8m 48s", "remaining_time": "4h 30m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.339167}
{"loss": 24.375, "grad_norm": 2.04122186, "learning_rate": 0.00092363, "token_acc": 0.3838915, "epoch": 0.20318771, "global_step/max_steps": "443/2181", "elapsed_time": "1h 8m 57s", "remaining_time": "4h 30m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.339269}
{"loss": 24.25, "grad_norm": 1.5767982, "learning_rate": 0.00092324, "token_acc": 0.3830076, "epoch": 0.20364637, "global_step/max_steps": "444/2181", "elapsed_time": "1h 9m 7s", "remaining_time": "4h 30m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.339239}
{"loss": 25.25, "grad_norm": 1.67319965, "learning_rate": 0.00092284, "token_acc": 0.37140487, "epoch": 0.20410503, "global_step/max_steps": "445/2181", "elapsed_time": "1h 9m 17s", "remaining_time": "4h 30m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.340721}
{"loss": 24.25, "grad_norm": 1.65660954, "learning_rate": 0.00092244, "token_acc": 0.38860399, "epoch": 0.2045637, "global_step/max_steps": "446/2181", "elapsed_time": "1h 9m 27s", "remaining_time": "4h 30m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.342194}
{"loss": 24.75, "grad_norm": 1.59069729, "learning_rate": 0.00092204, "token_acc": 0.37604924, "epoch": 0.20502236, "global_step/max_steps": "447/2181", "elapsed_time": "1h 9m 35s", "remaining_time": "4h 29m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.340868}
{"loss": 24.5, "grad_norm": 1.8557359, "learning_rate": 0.00092165, "token_acc": 0.3821368, "epoch": 0.20548102, "global_step/max_steps": "448/2181", "elapsed_time": "1h 9m 46s", "remaining_time": "4h 29m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.343169}
{"loss": 24.5, "grad_norm": 1.41730535, "learning_rate": 0.00092125, "token_acc": 0.37401253, "epoch": 0.20593969, "global_step/max_steps": "449/2181", "elapsed_time": "1h 9m 56s", "remaining_time": "4h 29m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.345743}
{"loss": 25.375, "grad_norm": 1.65188074, "learning_rate": 0.00092085, "token_acc": 0.36244053, "epoch": 0.20639835, "global_step/max_steps": "450/2181", "elapsed_time": "1h 10m 5s", "remaining_time": "4h 29m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.344444}
{"loss": 24.875, "grad_norm": 1.50820458, "learning_rate": 0.00092044, "token_acc": 0.37517955, "epoch": 0.20685701, "global_step/max_steps": "451/2181", "elapsed_time": "1h 10m 15s", "remaining_time": "4h 29m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.344956}
{"loss": 24.5, "grad_norm": 1.56512523, "learning_rate": 0.00092004, "token_acc": 0.36815359, "epoch": 0.20731567, "global_step/max_steps": "452/2181", "elapsed_time": "1h 10m 24s", "remaining_time": "4h 29m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.344475}
{"loss": 25.0, "grad_norm": 1.59376431, "learning_rate": 0.00091964, "token_acc": 0.35924197, "epoch": 0.20777434, "global_step/max_steps": "453/2181", "elapsed_time": "1h 10m 33s", "remaining_time": "4h 29m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.34382}
{"loss": 24.875, "grad_norm": 1.62537301, "learning_rate": 0.00091923, "token_acc": 0.38548753, "epoch": 0.208233, "global_step/max_steps": "454/2181", "elapsed_time": "1h 10m 43s", "remaining_time": "4h 28m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.344819}
{"loss": 24.5, "grad_norm": 1.61585617, "learning_rate": 0.00091883, "token_acc": 0.37992929, "epoch": 0.20869166, "global_step/max_steps": "455/2181", "elapsed_time": "1h 10m 52s", "remaining_time": "4h 28m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.345826}
{"loss": 24.25, "grad_norm": 1.65951872, "learning_rate": 0.00091842, "token_acc": 0.38548692, "epoch": 0.20915033, "global_step/max_steps": "456/2181", "elapsed_time": "1h 11m 2s", "remaining_time": "4h 28m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.346314}
{"loss": 25.25, "grad_norm": 1.50328958, "learning_rate": 0.00091802, "token_acc": 0.35858722, "epoch": 0.20960899, "global_step/max_steps": "457/2181", "elapsed_time": "1h 11m 11s", "remaining_time": "4h 28m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.346696}
{"loss": 24.5, "grad_norm": 1.76667881, "learning_rate": 0.00091761, "token_acc": 0.37916431, "epoch": 0.21006765, "global_step/max_steps": "458/2181", "elapsed_time": "1h 11m 21s", "remaining_time": "4h 28m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.346261}
{"loss": 24.5, "grad_norm": 1.49212158, "learning_rate": 0.0009172, "token_acc": 0.39081747, "epoch": 0.21052632, "global_step/max_steps": "459/2181", "elapsed_time": "1h 11m 30s", "remaining_time": "4h 28m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.345828}
{"loss": 25.25, "grad_norm": 1.87695456, "learning_rate": 0.00091679, "token_acc": 0.37612613, "epoch": 0.21098498, "global_step/max_steps": "460/2181", "elapsed_time": "1h 11m 39s", "remaining_time": "4h 28m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.345043}
{"loss": 25.25, "grad_norm": 1.71780026, "learning_rate": 0.00091638, "token_acc": 0.35937933, "epoch": 0.21144364, "global_step/max_steps": "461/2181", "elapsed_time": "1h 11m 49s", "remaining_time": "4h 27m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.347338}
{"loss": 25.375, "grad_norm": 1.45821893, "learning_rate": 0.00091597, "token_acc": 0.37650014, "epoch": 0.2119023, "global_step/max_steps": "462/2181", "elapsed_time": "1h 11m 59s", "remaining_time": "4h 27m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.347557}
{"loss": 24.375, "grad_norm": 1.47438431, "learning_rate": 0.00091555, "token_acc": 0.38235294, "epoch": 0.21236097, "global_step/max_steps": "463/2181", "elapsed_time": "1h 12m 8s", "remaining_time": "4h 27m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.347352}
{"loss": 24.625, "grad_norm": 1.54193425, "learning_rate": 0.00091514, "token_acc": 0.37510879, "epoch": 0.21281963, "global_step/max_steps": "464/2181", "elapsed_time": "1h 12m 16s", "remaining_time": "4h 27m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.345707}
{"loss": 23.875, "grad_norm": 1.55275297, "learning_rate": 0.00091473, "token_acc": 0.39321192, "epoch": 0.21327829, "global_step/max_steps": "465/2181", "elapsed_time": "1h 12m 26s", "remaining_time": "4h 27m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.347061}
{"loss": 24.625, "grad_norm": 1.56259418, "learning_rate": 0.00091431, "token_acc": 0.37308869, "epoch": 0.21373696, "global_step/max_steps": "466/2181", "elapsed_time": "1h 12m 35s", "remaining_time": "4h 27m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.346034}
{"loss": 25.0, "grad_norm": 1.6168083, "learning_rate": 0.00091389, "token_acc": 0.38216201, "epoch": 0.21419562, "global_step/max_steps": "467/2181", "elapsed_time": "1h 12m 45s", "remaining_time": "4h 27m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.347597}
{"loss": 25.0, "grad_norm": 1.65974176, "learning_rate": 0.00091348, "token_acc": 0.36724427, "epoch": 0.21465428, "global_step/max_steps": "468/2181", "elapsed_time": "1h 12m 55s", "remaining_time": "4h 26m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.347746}
{"loss": 24.5, "grad_norm": 1.61393416, "learning_rate": 0.00091306, "token_acc": 0.38805548, "epoch": 0.21511295, "global_step/max_steps": "469/2181", "elapsed_time": "1h 13m 4s", "remaining_time": "4h 26m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.346689}
{"loss": 24.5, "grad_norm": 1.49756372, "learning_rate": 0.00091264, "token_acc": 0.37406217, "epoch": 0.21557161, "global_step/max_steps": "470/2181", "elapsed_time": "1h 13m 14s", "remaining_time": "4h 26m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.349041}
{"loss": 24.5, "grad_norm": 1.64294004, "learning_rate": 0.00091222, "token_acc": 0.37646722, "epoch": 0.21603027, "global_step/max_steps": "471/2181", "elapsed_time": "1h 13m 23s", "remaining_time": "4h 26m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.348741}
{"loss": 25.125, "grad_norm": 1.5006094, "learning_rate": 0.0009118, "token_acc": 0.35689852, "epoch": 0.21648893, "global_step/max_steps": "472/2181", "elapsed_time": "1h 13m 33s", "remaining_time": "4h 26m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.348573}
{"loss": 25.0, "grad_norm": 1.62401402, "learning_rate": 0.00091138, "token_acc": 0.37642586, "epoch": 0.2169476, "global_step/max_steps": "473/2181", "elapsed_time": "1h 13m 42s", "remaining_time": "4h 26m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.348432}
{"loss": 24.25, "grad_norm": 1.60774493, "learning_rate": 0.00091096, "token_acc": 0.39685658, "epoch": 0.21740626, "global_step/max_steps": "474/2181", "elapsed_time": "1h 13m 51s", "remaining_time": "4h 25m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.348532}
{"loss": 25.75, "grad_norm": 1.79830801, "learning_rate": 0.00091053, "token_acc": 0.36714443, "epoch": 0.21786492, "global_step/max_steps": "475/2181", "elapsed_time": "1h 14m 1s", "remaining_time": "4h 25m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.349329}
{"loss": 23.75, "grad_norm": 1.70487142, "learning_rate": 0.00091011, "token_acc": 0.3891455, "epoch": 0.21832359, "global_step/max_steps": "476/2181", "elapsed_time": "1h 14m 10s", "remaining_time": "4h 25m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.349603}
{"loss": 24.25, "grad_norm": 1.37213922, "learning_rate": 0.00090968, "token_acc": 0.37621145, "epoch": 0.21878225, "global_step/max_steps": "477/2181", "elapsed_time": "1h 14m 19s", "remaining_time": "4h 25m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.348739}
{"loss": 24.25, "grad_norm": 1.39378643, "learning_rate": 0.00090926, "token_acc": 0.37273488, "epoch": 0.21924091, "global_step/max_steps": "478/2181", "elapsed_time": "1h 14m 29s", "remaining_time": "4h 25m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.348844}
{"loss": 24.75, "grad_norm": 1.47772992, "learning_rate": 0.00090883, "token_acc": 0.38038884, "epoch": 0.21969958, "global_step/max_steps": "479/2181", "elapsed_time": "1h 14m 37s", "remaining_time": "4h 25m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.347296}
{"loss": 24.875, "grad_norm": 1.52291119, "learning_rate": 0.0009084, "token_acc": 0.34879032, "epoch": 0.22015824, "global_step/max_steps": "480/2181", "elapsed_time": "1h 14m 46s", "remaining_time": "4h 24m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.346764}
{"loss": 25.25, "grad_norm": 1.52379227, "learning_rate": 0.00090797, "token_acc": 0.37257543, "epoch": 0.2206169, "global_step/max_steps": "481/2181", "elapsed_time": "1h 14m 57s", "remaining_time": "4h 24m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.349634}
{"loss": 24.75, "grad_norm": 1.41642141, "learning_rate": 0.00090754, "token_acc": 0.38282139, "epoch": 0.22107556, "global_step/max_steps": "482/2181", "elapsed_time": "1h 15m 6s", "remaining_time": "4h 24m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.348179}
{"loss": 24.625, "grad_norm": 1.4682883, "learning_rate": 0.00090711, "token_acc": 0.38790346, "epoch": 0.22153423, "global_step/max_steps": "483/2181", "elapsed_time": "1h 15m 15s", "remaining_time": "4h 24m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.347896}
{"loss": 25.625, "grad_norm": 1.41266632, "learning_rate": 0.00090668, "token_acc": 0.36122618, "epoch": 0.22199289, "global_step/max_steps": "484/2181", "elapsed_time": "1h 15m 25s", "remaining_time": "4h 24m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.348912}
{"loss": 24.0, "grad_norm": 1.60714841, "learning_rate": 0.00090625, "token_acc": 0.38898072, "epoch": 0.22245155, "global_step/max_steps": "485/2181", "elapsed_time": "1h 15m 34s", "remaining_time": "4h 24m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.34858}
{"loss": 24.75, "grad_norm": 1.52361357, "learning_rate": 0.00090581, "token_acc": 0.37710145, "epoch": 0.22291022, "global_step/max_steps": "486/2181", "elapsed_time": "1h 15m 43s", "remaining_time": "4h 24m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.34735}
{"loss": 25.125, "grad_norm": 1.66076052, "learning_rate": 0.00090538, "token_acc": 0.38472581, "epoch": 0.22336888, "global_step/max_steps": "487/2181", "elapsed_time": "1h 15m 53s", "remaining_time": "4h 23m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.349081}
{"loss": 24.875, "grad_norm": 1.80134165, "learning_rate": 0.00090494, "token_acc": 0.36530556, "epoch": 0.22382754, "global_step/max_steps": "488/2181", "elapsed_time": "1h 16m 2s", "remaining_time": "4h 23m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.348047}
{"loss": 24.125, "grad_norm": 1.62724912, "learning_rate": 0.00090451, "token_acc": 0.38053613, "epoch": 0.22428621, "global_step/max_steps": "489/2181", "elapsed_time": "1h 16m 12s", "remaining_time": "4h 23m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.348923}
{"loss": 25.0, "grad_norm": 1.49070191, "learning_rate": 0.00090407, "token_acc": 0.37359551, "epoch": 0.22474487, "global_step/max_steps": "490/2181", "elapsed_time": "1h 16m 21s", "remaining_time": "4h 23m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.348936}
{"loss": 24.125, "grad_norm": 1.57760978, "learning_rate": 0.00090363, "token_acc": 0.38748962, "epoch": 0.22520353, "global_step/max_steps": "491/2181", "elapsed_time": "1h 16m 30s", "remaining_time": "4h 23m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.348346}
{"loss": 25.5, "grad_norm": 1.57957304, "learning_rate": 0.00090319, "token_acc": 0.37348424, "epoch": 0.22566219, "global_step/max_steps": "492/2181", "elapsed_time": "1h 16m 41s", "remaining_time": "4h 23m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.350633}
{"loss": 24.625, "grad_norm": 1.49029231, "learning_rate": 0.00090276, "token_acc": 0.37447168, "epoch": 0.22612086, "global_step/max_steps": "493/2181", "elapsed_time": "1h 16m 50s", "remaining_time": "4h 23m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.351108}
{"loss": 24.875, "grad_norm": 1.73451746, "learning_rate": 0.00090231, "token_acc": 0.37241764, "epoch": 0.22657952, "global_step/max_steps": "494/2181", "elapsed_time": "1h 16m 60s", "remaining_time": "4h 22m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.351302}
{"loss": 24.625, "grad_norm": 1.72372127, "learning_rate": 0.00090187, "token_acc": 0.38180797, "epoch": 0.22703818, "global_step/max_steps": "495/2181", "elapsed_time": "1h 17m 10s", "remaining_time": "4h 22m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.353222}
{"loss": 25.0, "grad_norm": 1.66141701, "learning_rate": 0.00090143, "token_acc": 0.36594912, "epoch": 0.22749685, "global_step/max_steps": "496/2181", "elapsed_time": "1h 17m 19s", "remaining_time": "4h 22m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.353621}
{"loss": 24.875, "grad_norm": 1.54840505, "learning_rate": 0.00090099, "token_acc": 0.37647691, "epoch": 0.22795551, "global_step/max_steps": "497/2181", "elapsed_time": "1h 17m 29s", "remaining_time": "4h 22m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.354735}
{"loss": 25.0, "grad_norm": 1.57709432, "learning_rate": 0.00090054, "token_acc": 0.37809086, "epoch": 0.22841417, "global_step/max_steps": "498/2181", "elapsed_time": "1h 17m 38s", "remaining_time": "4h 22m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.35428}
{"loss": 24.75, "grad_norm": 1.57076752, "learning_rate": 0.0009001, "token_acc": 0.37087379, "epoch": 0.22887284, "global_step/max_steps": "499/2181", "elapsed_time": "1h 17m 48s", "remaining_time": "4h 22m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.3552}
{"loss": 24.625, "grad_norm": 1.64171243, "learning_rate": 0.00089965, "token_acc": 0.37549185, "epoch": 0.2293315, "global_step/max_steps": "500/2181", "elapsed_time": "1h 17m 58s", "remaining_time": "4h 22m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.356503}
{"loss": 25.25, "grad_norm": 1.70627022, "learning_rate": 0.00089921, "token_acc": 0.36880631, "epoch": 0.22979016, "global_step/max_steps": "501/2181", "elapsed_time": "1h 18m 27s", "remaining_time": "4h 23m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.395859}
{"loss": 23.75, "grad_norm": 1.54536831, "learning_rate": 0.00089876, "token_acc": 0.3916413, "epoch": 0.23024882, "global_step/max_steps": "502/2181", "elapsed_time": "1h 18m 37s", "remaining_time": "4h 22m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.396702}
{"loss": 24.5, "grad_norm": 1.46026707, "learning_rate": 0.00089831, "token_acc": 0.38584933, "epoch": 0.23070749, "global_step/max_steps": "503/2181", "elapsed_time": "1h 18m 47s", "remaining_time": "4h 22m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.397912}
{"loss": 25.25, "grad_norm": 1.61692715, "learning_rate": 0.00089786, "token_acc": 0.37862422, "epoch": 0.23116615, "global_step/max_steps": "504/2181", "elapsed_time": "1h 18m 56s", "remaining_time": "4h 22m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.397109}
{"loss": 25.125, "grad_norm": 1.38869488, "learning_rate": 0.00089741, "token_acc": 0.36825397, "epoch": 0.23162481, "global_step/max_steps": "505/2181", "elapsed_time": "1h 19m 6s", "remaining_time": "4h 22m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.397635}
{"loss": 24.625, "grad_norm": 1.56074786, "learning_rate": 0.00089696, "token_acc": 0.36902486, "epoch": 0.23208348, "global_step/max_steps": "506/2181", "elapsed_time": "1h 19m 15s", "remaining_time": "4h 22m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.397757}
{"loss": 24.75, "grad_norm": 1.44672239, "learning_rate": 0.00089651, "token_acc": 0.36237128, "epoch": 0.23254214, "global_step/max_steps": "507/2181", "elapsed_time": "1h 19m 25s", "remaining_time": "4h 22m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.39828}
{"loss": 24.25, "grad_norm": 1.70242035, "learning_rate": 0.00089605, "token_acc": 0.39314572, "epoch": 0.2330008, "global_step/max_steps": "508/2181", "elapsed_time": "1h 19m 35s", "remaining_time": "4h 22m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399356}
{"loss": 24.375, "grad_norm": 1.71860409, "learning_rate": 0.0008956, "token_acc": 0.38531588, "epoch": 0.23345947, "global_step/max_steps": "509/2181", "elapsed_time": "1h 19m 44s", "remaining_time": "4h 21m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.39961}
{"loss": 24.25, "grad_norm": 1.41450119, "learning_rate": 0.00089515, "token_acc": 0.37609742, "epoch": 0.23391813, "global_step/max_steps": "510/2181", "elapsed_time": "1h 19m 54s", "remaining_time": "4h 21m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.400682}
{"loss": 24.875, "grad_norm": 1.56862199, "learning_rate": 0.00089469, "token_acc": 0.35828427, "epoch": 0.23437679, "global_step/max_steps": "511/2181", "elapsed_time": "1h 20m 4s", "remaining_time": "4h 21m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.401218}
{"loss": 25.0, "grad_norm": 1.54583693, "learning_rate": 0.00089423, "token_acc": 0.37043236, "epoch": 0.23483545, "global_step/max_steps": "512/2181", "elapsed_time": "1h 20m 13s", "remaining_time": "4h 21m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.400261}
{"loss": 24.0, "grad_norm": 1.39614952, "learning_rate": 0.00089378, "token_acc": 0.38789359, "epoch": 0.23529412, "global_step/max_steps": "513/2181", "elapsed_time": "1h 20m 23s", "remaining_time": "4h 21m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.402234}
{"loss": 24.875, "grad_norm": 1.57235157, "learning_rate": 0.00089332, "token_acc": 0.37567645, "epoch": 0.23575278, "global_step/max_steps": "514/2181", "elapsed_time": "1h 20m 34s", "remaining_time": "4h 21m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.404794}
{"loss": 25.25, "grad_norm": 2.11495423, "learning_rate": 0.00089286, "token_acc": 0.37815126, "epoch": 0.23621144, "global_step/max_steps": "515/2181", "elapsed_time": "1h 20m 43s", "remaining_time": "4h 21m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403923}
{"loss": 24.125, "grad_norm": 1.74458206, "learning_rate": 0.0008924, "token_acc": 0.39833795, "epoch": 0.23667011, "global_step/max_steps": "516/2181", "elapsed_time": "1h 20m 52s", "remaining_time": "4h 20m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403418}
{"loss": 24.625, "grad_norm": 1.49531293, "learning_rate": 0.00089194, "token_acc": 0.37322745, "epoch": 0.23712877, "global_step/max_steps": "517/2181", "elapsed_time": "1h 21m 1s", "remaining_time": "4h 20m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.402798}
{"loss": 24.25, "grad_norm": 1.37980413, "learning_rate": 0.00089148, "token_acc": 0.38003452, "epoch": 0.23758743, "global_step/max_steps": "518/2181", "elapsed_time": "1h 21m 12s", "remaining_time": "4h 20m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.404925}
{"loss": 25.375, "grad_norm": 1.48494279, "learning_rate": 0.00089101, "token_acc": 0.35735573, "epoch": 0.2380461, "global_step/max_steps": "519/2181", "elapsed_time": "1h 21m 21s", "remaining_time": "4h 20m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.40528}
{"loss": 24.25, "grad_norm": 1.34332359, "learning_rate": 0.00089055, "token_acc": 0.37165269, "epoch": 0.23850476, "global_step/max_steps": "520/2181", "elapsed_time": "1h 21m 31s", "remaining_time": "4h 20m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.405712}
{"loss": 24.375, "grad_norm": 1.45580506, "learning_rate": 0.00089009, "token_acc": 0.37985961, "epoch": 0.23896342, "global_step/max_steps": "521/2181", "elapsed_time": "1h 21m 41s", "remaining_time": "4h 20m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406265}
{"loss": 24.75, "grad_norm": 1.58406699, "learning_rate": 0.00088962, "token_acc": 0.37742857, "epoch": 0.23942208, "global_step/max_steps": "522/2181", "elapsed_time": "1h 21m 50s", "remaining_time": "4h 20m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406572}
{"loss": 24.125, "grad_norm": 1.40309834, "learning_rate": 0.00088916, "token_acc": 0.39931643, "epoch": 0.23988075, "global_step/max_steps": "523/2181", "elapsed_time": "1h 21m 60s", "remaining_time": "4h 19m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406539}
{"loss": 24.625, "grad_norm": 1.48018074, "learning_rate": 0.00088869, "token_acc": 0.39082969, "epoch": 0.24033941, "global_step/max_steps": "524/2181", "elapsed_time": "1h 22m 9s", "remaining_time": "4h 19m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406973}
{"loss": 25.25, "grad_norm": 1.48950279, "learning_rate": 0.00088822, "token_acc": 0.36156727, "epoch": 0.24079807, "global_step/max_steps": "525/2181", "elapsed_time": "1h 22m 19s", "remaining_time": "4h 19m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.407249}
{"loss": 24.375, "grad_norm": 1.42634082, "learning_rate": 0.00088775, "token_acc": 0.38217768, "epoch": 0.24125674, "global_step/max_steps": "526/2181", "elapsed_time": "1h 22m 28s", "remaining_time": "4h 19m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.40683}
{"loss": 24.75, "grad_norm": 1.32061577, "learning_rate": 0.00088728, "token_acc": 0.38099174, "epoch": 0.2417154, "global_step/max_steps": "527/2181", "elapsed_time": "1h 22m 37s", "remaining_time": "4h 19m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.40637}
{"loss": 24.625, "grad_norm": 1.54265308, "learning_rate": 0.00088681, "token_acc": 0.38920135, "epoch": 0.24217406, "global_step/max_steps": "528/2181", "elapsed_time": "1h 22m 46s", "remaining_time": "4h 19m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.405882}
{"loss": 24.5, "grad_norm": 1.36204827, "learning_rate": 0.00088634, "token_acc": 0.36770921, "epoch": 0.24263273, "global_step/max_steps": "529/2181", "elapsed_time": "1h 22m 55s", "remaining_time": "4h 18m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.405005}
{"loss": 24.625, "grad_norm": 1.3517952, "learning_rate": 0.00088587, "token_acc": 0.37890847, "epoch": 0.24309139, "global_step/max_steps": "530/2181", "elapsed_time": "1h 23m 4s", "remaining_time": "4h 18m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.404217}
{"loss": 24.625, "grad_norm": 1.63936281, "learning_rate": 0.0008854, "token_acc": 0.36924365, "epoch": 0.24355005, "global_step/max_steps": "531/2181", "elapsed_time": "1h 23m 14s", "remaining_time": "4h 18m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403964}
{"loss": 24.75, "grad_norm": 1.52182531, "learning_rate": 0.00088493, "token_acc": 0.36595992, "epoch": 0.24400871, "global_step/max_steps": "532/2181", "elapsed_time": "1h 23m 23s", "remaining_time": "4h 18m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403869}
{"loss": 24.375, "grad_norm": 1.43005884, "learning_rate": 0.00088445, "token_acc": 0.38613861, "epoch": 0.24446738, "global_step/max_steps": "533/2181", "elapsed_time": "1h 23m 32s", "remaining_time": "4h 18m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403899}
{"loss": 25.25, "grad_norm": 1.30936718, "learning_rate": 0.00088398, "token_acc": 0.38056013, "epoch": 0.24492604, "global_step/max_steps": "534/2181", "elapsed_time": "1h 23m 41s", "remaining_time": "4h 18m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.402409}
{"loss": 24.5, "grad_norm": 1.86063576, "learning_rate": 0.0008835, "token_acc": 0.37514029, "epoch": 0.2453847, "global_step/max_steps": "535/2181", "elapsed_time": "1h 23m 51s", "remaining_time": "4h 17m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403925}
{"loss": 24.375, "grad_norm": 1.74388433, "learning_rate": 0.00088302, "token_acc": 0.36439459, "epoch": 0.24584337, "global_step/max_steps": "536/2181", "elapsed_time": "1h 24m 0s", "remaining_time": "4h 17m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403549}
{"loss": 25.0, "grad_norm": 1.87585521, "learning_rate": 0.00088254, "token_acc": 0.36424394, "epoch": 0.24630203, "global_step/max_steps": "537/2181", "elapsed_time": "1h 24m 9s", "remaining_time": "4h 17m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.402455}
{"loss": 24.5, "grad_norm": 1.41848958, "learning_rate": 0.00088207, "token_acc": 0.38280569, "epoch": 0.24676069, "global_step/max_steps": "538/2181", "elapsed_time": "1h 24m 19s", "remaining_time": "4h 17m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.40294}
{"loss": 23.75, "grad_norm": 1.69272697, "learning_rate": 0.00088159, "token_acc": 0.4184136, "epoch": 0.24721936, "global_step/max_steps": "539/2181", "elapsed_time": "1h 24m 28s", "remaining_time": "4h 17m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403309}
{"loss": 25.375, "grad_norm": 1.46914089, "learning_rate": 0.00088111, "token_acc": 0.37419181, "epoch": 0.24767802, "global_step/max_steps": "540/2181", "elapsed_time": "1h 24m 38s", "remaining_time": "4h 17m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.402803}
{"loss": 24.375, "grad_norm": 1.65193367, "learning_rate": 0.00088062, "token_acc": 0.38269504, "epoch": 0.24813668, "global_step/max_steps": "541/2181", "elapsed_time": "1h 24m 47s", "remaining_time": "4h 17m 0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.402564}
{"loss": 24.0, "grad_norm": 1.43850183, "learning_rate": 0.00088014, "token_acc": 0.39553255, "epoch": 0.24859534, "global_step/max_steps": "542/2181", "elapsed_time": "1h 24m 56s", "remaining_time": "4h 16m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.402085}
{"loss": 24.875, "grad_norm": 1.59021842, "learning_rate": 0.00087966, "token_acc": 0.3778728, "epoch": 0.24905401, "global_step/max_steps": "543/2181", "elapsed_time": "1h 25m 6s", "remaining_time": "4h 16m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.402937}
{"loss": 24.875, "grad_norm": 1.39534235, "learning_rate": 0.00087918, "token_acc": 0.36920575, "epoch": 0.24951267, "global_step/max_steps": "544/2181", "elapsed_time": "1h 25m 15s", "remaining_time": "4h 16m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403142}
{"loss": 24.375, "grad_norm": 1.75310314, "learning_rate": 0.00087869, "token_acc": 0.35953757, "epoch": 0.24997133, "global_step/max_steps": "545/2181", "elapsed_time": "1h 25m 24s", "remaining_time": "4h 16m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.402282}
{"loss": 24.875, "grad_norm": 1.50538194, "learning_rate": 0.00087821, "token_acc": 0.37443182, "epoch": 0.25043, "global_step/max_steps": "546/2181", "elapsed_time": "1h 25m 33s", "remaining_time": "4h 16m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.40184}
{"loss": 24.5, "grad_norm": 1.59097946, "learning_rate": 0.00087772, "token_acc": 0.37674801, "epoch": 0.25088866, "global_step/max_steps": "547/2181", "elapsed_time": "1h 25m 43s", "remaining_time": "4h 16m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.40263}
{"loss": 24.125, "grad_norm": 1.30983543, "learning_rate": 0.00087723, "token_acc": 0.40481461, "epoch": 0.25134732, "global_step/max_steps": "548/2181", "elapsed_time": "1h 25m 52s", "remaining_time": "4h 15m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.40141}
{"loss": 25.125, "grad_norm": 1.35120416, "learning_rate": 0.00087675, "token_acc": 0.37920316, "epoch": 0.25180599, "global_step/max_steps": "549/2181", "elapsed_time": "1h 26m 1s", "remaining_time": "4h 15m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.40132}
{"loss": 24.875, "grad_norm": 1.36317706, "learning_rate": 0.00087626, "token_acc": 0.35800109, "epoch": 0.25226465, "global_step/max_steps": "550/2181", "elapsed_time": "1h 26m 10s", "remaining_time": "4h 15m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.400073}
{"loss": 24.5, "grad_norm": 1.3097657, "learning_rate": 0.00087577, "token_acc": 0.37747206, "epoch": 0.25272331, "global_step/max_steps": "551/2181", "elapsed_time": "1h 26m 19s", "remaining_time": "4h 15m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399107}
{"loss": 23.875, "grad_norm": 1.61924946, "learning_rate": 0.00087528, "token_acc": 0.38607955, "epoch": 0.25318197, "global_step/max_steps": "552/2181", "elapsed_time": "1h 26m 28s", "remaining_time": "4h 15m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399181}
{"loss": 25.0, "grad_norm": 1.40644884, "learning_rate": 0.00087479, "token_acc": 0.38278037, "epoch": 0.25364064, "global_step/max_steps": "553/2181", "elapsed_time": "1h 26m 37s", "remaining_time": "4h 15m 0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398057}
{"loss": 24.625, "grad_norm": 1.53222072, "learning_rate": 0.00087429, "token_acc": 0.38593622, "epoch": 0.2540993, "global_step/max_steps": "554/2181", "elapsed_time": "1h 26m 47s", "remaining_time": "4h 14m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399169}
{"loss": 23.75, "grad_norm": 1.53254616, "learning_rate": 0.0008738, "token_acc": 0.39424173, "epoch": 0.25455796, "global_step/max_steps": "555/2181", "elapsed_time": "1h 26m 56s", "remaining_time": "4h 14m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398876}
{"loss": 25.0, "grad_norm": 1.62437117, "learning_rate": 0.00087331, "token_acc": 0.39019662, "epoch": 0.25501663, "global_step/max_steps": "556/2181", "elapsed_time": "1h 27m 6s", "remaining_time": "4h 14m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399286}
{"loss": 24.875, "grad_norm": 1.40383041, "learning_rate": 0.00087281, "token_acc": 0.38484765, "epoch": 0.25547529, "global_step/max_steps": "557/2181", "elapsed_time": "1h 27m 16s", "remaining_time": "4h 14m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399903}
{"loss": 25.0, "grad_norm": 1.75867569, "learning_rate": 0.00087232, "token_acc": 0.36652361, "epoch": 0.25593395, "global_step/max_steps": "558/2181", "elapsed_time": "1h 27m 25s", "remaining_time": "4h 14m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399992}
{"loss": 24.0, "grad_norm": 1.55382884, "learning_rate": 0.00087182, "token_acc": 0.38765154, "epoch": 0.25639262, "global_step/max_steps": "559/2181", "elapsed_time": "1h 27m 35s", "remaining_time": "4h 14m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399878}
{"loss": 25.0, "grad_norm": 1.64414203, "learning_rate": 0.00087132, "token_acc": 0.37125071, "epoch": 0.25685128, "global_step/max_steps": "560/2181", "elapsed_time": "1h 27m 44s", "remaining_time": "4h 13m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399821}
{"loss": 24.625, "grad_norm": 1.51377845, "learning_rate": 0.00087083, "token_acc": 0.36720751, "epoch": 0.25730994, "global_step/max_steps": "561/2181", "elapsed_time": "1h 27m 53s", "remaining_time": "4h 13m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399896}
{"loss": 24.0, "grad_norm": 1.42218018, "learning_rate": 0.00087033, "token_acc": 0.38362439, "epoch": 0.2577686, "global_step/max_steps": "562/2181", "elapsed_time": "1h 28m 2s", "remaining_time": "4h 13m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399282}
{"loss": 24.875, "grad_norm": 1.38474393, "learning_rate": 0.00086983, "token_acc": 0.36723321, "epoch": 0.25822727, "global_step/max_steps": "563/2181", "elapsed_time": "1h 28m 12s", "remaining_time": "4h 13m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398826}
{"loss": 25.0, "grad_norm": 1.37287724, "learning_rate": 0.00086933, "token_acc": 0.36948016, "epoch": 0.25868593, "global_step/max_steps": "564/2181", "elapsed_time": "1h 28m 21s", "remaining_time": "4h 13m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398401}
{"loss": 24.0, "grad_norm": 1.63004184, "learning_rate": 0.00086883, "token_acc": 0.38530067, "epoch": 0.25914459, "global_step/max_steps": "565/2181", "elapsed_time": "1h 28m 30s", "remaining_time": "4h 13m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398672}
{"loss": 24.375, "grad_norm": 1.3532089, "learning_rate": 0.00086832, "token_acc": 0.39126831, "epoch": 0.25960326, "global_step/max_steps": "566/2181", "elapsed_time": "1h 28m 40s", "remaining_time": "4h 12m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398485}
{"loss": 25.125, "grad_norm": 1.41453278, "learning_rate": 0.00086782, "token_acc": 0.375, "epoch": 0.26006192, "global_step/max_steps": "567/2181", "elapsed_time": "1h 28m 49s", "remaining_time": "4h 12m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398619}
{"loss": 25.25, "grad_norm": 1.42649281, "learning_rate": 0.00086732, "token_acc": 0.37436762, "epoch": 0.26052058, "global_step/max_steps": "568/2181", "elapsed_time": "1h 28m 59s", "remaining_time": "4h 12m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398769}
{"loss": 24.0, "grad_norm": 1.39140522, "learning_rate": 0.00086681, "token_acc": 0.38285059, "epoch": 0.26097925, "global_step/max_steps": "569/2181", "elapsed_time": "1h 29m 8s", "remaining_time": "4h 12m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398941}
{"loss": 24.5, "grad_norm": 1.64426875, "learning_rate": 0.00086631, "token_acc": 0.38377316, "epoch": 0.26143791, "global_step/max_steps": "570/2181", "elapsed_time": "1h 29m 17s", "remaining_time": "4h 12m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398465}
{"loss": 25.625, "grad_norm": 1.62114811, "learning_rate": 0.0008658, "token_acc": 0.36435868, "epoch": 0.26189657, "global_step/max_steps": "571/2181", "elapsed_time": "1h 29m 27s", "remaining_time": "4h 12m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398924}
{"loss": 23.25, "grad_norm": 1.55612075, "learning_rate": 0.0008653, "token_acc": 0.39869459, "epoch": 0.26235523, "global_step/max_steps": "572/2181", "elapsed_time": "1h 29m 36s", "remaining_time": "4h 12m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398831}
{"loss": 24.375, "grad_norm": 1.4494226, "learning_rate": 0.00086479, "token_acc": 0.38649308, "epoch": 0.2628139, "global_step/max_steps": "573/2181", "elapsed_time": "1h 29m 46s", "remaining_time": "4h 11m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.39984}
{"loss": 24.625, "grad_norm": 1.66983986, "learning_rate": 0.00086428, "token_acc": 0.36796286, "epoch": 0.26327256, "global_step/max_steps": "574/2181", "elapsed_time": "1h 29m 55s", "remaining_time": "4h 11m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399702}
{"loss": 24.625, "grad_norm": 1.53707421, "learning_rate": 0.00086377, "token_acc": 0.37193764, "epoch": 0.26373122, "global_step/max_steps": "575/2181", "elapsed_time": "1h 30m 5s", "remaining_time": "4h 11m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399402}
{"loss": 24.75, "grad_norm": 1.36826694, "learning_rate": 0.00086326, "token_acc": 0.3660738, "epoch": 0.26418989, "global_step/max_steps": "576/2181", "elapsed_time": "1h 30m 14s", "remaining_time": "4h 11m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398674}
{"loss": 24.75, "grad_norm": 1.46177351, "learning_rate": 0.00086275, "token_acc": 0.38058678, "epoch": 0.26464855, "global_step/max_steps": "577/2181", "elapsed_time": "1h 30m 23s", "remaining_time": "4h 11m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399174}
{"loss": 24.25, "grad_norm": 1.82250106, "learning_rate": 0.00086224, "token_acc": 0.3891923, "epoch": 0.26510721, "global_step/max_steps": "578/2181", "elapsed_time": "1h 30m 32s", "remaining_time": "4h 11m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398598}
{"loss": 23.875, "grad_norm": 1.48687637, "learning_rate": 0.00086173, "token_acc": 0.3815393, "epoch": 0.26556588, "global_step/max_steps": "579/2181", "elapsed_time": "1h 30m 42s", "remaining_time": "4h 10m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399766}
{"loss": 25.0, "grad_norm": 1.6762594, "learning_rate": 0.00086121, "token_acc": 0.37456847, "epoch": 0.26602454, "global_step/max_steps": "580/2181", "elapsed_time": "1h 30m 52s", "remaining_time": "4h 10m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399422}
{"loss": 23.75, "grad_norm": 1.43142676, "learning_rate": 0.0008607, "token_acc": 0.38789359, "epoch": 0.2664832, "global_step/max_steps": "581/2181", "elapsed_time": "1h 31m 1s", "remaining_time": "4h 10m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398502}
{"loss": 24.625, "grad_norm": 1.35195744, "learning_rate": 0.00086019, "token_acc": 0.37094499, "epoch": 0.26694186, "global_step/max_steps": "582/2181", "elapsed_time": "1h 31m 9s", "remaining_time": "4h 10m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.397705}
{"loss": 24.5, "grad_norm": 1.44178355, "learning_rate": 0.00085967, "token_acc": 0.38510134, "epoch": 0.26740053, "global_step/max_steps": "583/2181", "elapsed_time": "1h 31m 19s", "remaining_time": "4h 10m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.397577}
{"loss": 24.5, "grad_norm": 1.34800088, "learning_rate": 0.00085915, "token_acc": 0.38736492, "epoch": 0.26785919, "global_step/max_steps": "584/2181", "elapsed_time": "1h 31m 28s", "remaining_time": "4h 10m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.396909}
{"loss": 24.0, "grad_norm": 1.46060693, "learning_rate": 0.00085864, "token_acc": 0.37443567, "epoch": 0.26831785, "global_step/max_steps": "585/2181", "elapsed_time": "1h 31m 37s", "remaining_time": "4h 9m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.396887}
{"loss": 24.75, "grad_norm": 1.4365406, "learning_rate": 0.00085812, "token_acc": 0.38358045, "epoch": 0.26877652, "global_step/max_steps": "586/2181", "elapsed_time": "1h 31m 47s", "remaining_time": "4h 9m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.396777}
{"loss": 24.5, "grad_norm": 1.41632867, "learning_rate": 0.0008576, "token_acc": 0.38465861, "epoch": 0.26923518, "global_step/max_steps": "587/2181", "elapsed_time": "1h 31m 57s", "remaining_time": "4h 9m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398116}
{"loss": 24.25, "grad_norm": 1.37565029, "learning_rate": 0.00085708, "token_acc": 0.3820132, "epoch": 0.26969384, "global_step/max_steps": "588/2181", "elapsed_time": "1h 32m 6s", "remaining_time": "4h 9m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398631}
{"loss": 25.0, "grad_norm": 1.3336153, "learning_rate": 0.00085656, "token_acc": 0.37165702, "epoch": 0.27015251, "global_step/max_steps": "589/2181", "elapsed_time": "1h 32m 16s", "remaining_time": "4h 9m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399253}
{"loss": 24.5, "grad_norm": 1.32579315, "learning_rate": 0.00085604, "token_acc": 0.36951118, "epoch": 0.27061117, "global_step/max_steps": "590/2181", "elapsed_time": "1h 32m 26s", "remaining_time": "4h 9m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399248}
{"loss": 24.625, "grad_norm": 1.66767299, "learning_rate": 0.00085552, "token_acc": 0.38025799, "epoch": 0.27106983, "global_step/max_steps": "591/2181", "elapsed_time": "1h 32m 35s", "remaining_time": "4h 9m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398814}
{"loss": 24.25, "grad_norm": 1.4168824, "learning_rate": 0.00085499, "token_acc": 0.39330897, "epoch": 0.27152849, "global_step/max_steps": "592/2181", "elapsed_time": "1h 32m 44s", "remaining_time": "4h 8m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398033}
{"loss": 24.5, "grad_norm": 1.5253911, "learning_rate": 0.00085447, "token_acc": 0.39415656, "epoch": 0.27198716, "global_step/max_steps": "593/2181", "elapsed_time": "1h 32m 53s", "remaining_time": "4h 8m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.398157}
{"loss": 24.375, "grad_norm": 1.50668025, "learning_rate": 0.00085395, "token_acc": 0.36219131, "epoch": 0.27244582, "global_step/max_steps": "594/2181", "elapsed_time": "1h 33m 2s", "remaining_time": "4h 8m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.39723}
{"loss": 23.625, "grad_norm": 1.49895954, "learning_rate": 0.00085342, "token_acc": 0.38726636, "epoch": 0.27290448, "global_step/max_steps": "595/2181", "elapsed_time": "1h 33m 11s", "remaining_time": "4h 8m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.397315}
{"loss": 24.75, "grad_norm": 1.37894166, "learning_rate": 0.0008529, "token_acc": 0.38176871, "epoch": 0.27336315, "global_step/max_steps": "596/2181", "elapsed_time": "1h 33m 21s", "remaining_time": "4h 8m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.396991}
{"loss": 24.0, "grad_norm": 1.29881048, "learning_rate": 0.00085237, "token_acc": 0.38614694, "epoch": 0.27382181, "global_step/max_steps": "597/2181", "elapsed_time": "1h 33m 32s", "remaining_time": "4h 8m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.399664}
{"loss": 24.5, "grad_norm": 1.44659364, "learning_rate": 0.00085184, "token_acc": 0.36389281, "epoch": 0.27428047, "global_step/max_steps": "598/2181", "elapsed_time": "1h 33m 41s", "remaining_time": "4h 7m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.39944}
{"loss": 24.5, "grad_norm": 1.37853265, "learning_rate": 0.00085131, "token_acc": 0.3847207, "epoch": 0.27473914, "global_step/max_steps": "599/2181", "elapsed_time": "1h 33m 51s", "remaining_time": "4h 7m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.400392}
{"loss": 24.875, "grad_norm": 1.36809635, "learning_rate": 0.00085079, "token_acc": 0.3850028, "epoch": 0.2751978, "global_step/max_steps": "600/2181", "elapsed_time": "1h 34m 1s", "remaining_time": "4h 7m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.401159}
{"loss": 24.875, "grad_norm": 1.30172932, "learning_rate": 0.00085026, "token_acc": 0.36986301, "epoch": 0.27565646, "global_step/max_steps": "601/2181", "elapsed_time": "1h 34m 11s", "remaining_time": "4h 7m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.40261}
{"loss": 24.0, "grad_norm": 1.42355442, "learning_rate": 0.00084973, "token_acc": 0.38294956, "epoch": 0.27611512, "global_step/max_steps": "602/2181", "elapsed_time": "1h 34m 20s", "remaining_time": "4h 7m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.402423}
{"loss": 23.875, "grad_norm": 1.59637976, "learning_rate": 0.00084919, "token_acc": 0.38025751, "epoch": 0.27657379, "global_step/max_steps": "603/2181", "elapsed_time": "1h 34m 30s", "remaining_time": "4h 7m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403323}
{"loss": 25.0, "grad_norm": 1.68493223, "learning_rate": 0.00084866, "token_acc": 0.38772124, "epoch": 0.27703245, "global_step/max_steps": "604/2181", "elapsed_time": "1h 34m 40s", "remaining_time": "4h 7m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403415}
{"loss": 23.75, "grad_norm": 1.36195612, "learning_rate": 0.00084813, "token_acc": 0.39583911, "epoch": 0.27749111, "global_step/max_steps": "605/2181", "elapsed_time": "1h 34m 49s", "remaining_time": "4h 6m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403102}
{"loss": 24.625, "grad_norm": 1.71735668, "learning_rate": 0.0008476, "token_acc": 0.36447255, "epoch": 0.27794978, "global_step/max_steps": "606/2181", "elapsed_time": "1h 34m 59s", "remaining_time": "4h 6m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403734}
{"loss": 24.375, "grad_norm": 1.45988214, "learning_rate": 0.00084706, "token_acc": 0.36782935, "epoch": 0.27840844, "global_step/max_steps": "607/2181", "elapsed_time": "1h 35m 8s", "remaining_time": "4h 6m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403326}
{"loss": 24.0, "grad_norm": 1.55787385, "learning_rate": 0.00084653, "token_acc": 0.40348964, "epoch": 0.2788671, "global_step/max_steps": "608/2181", "elapsed_time": "1h 35m 17s", "remaining_time": "4h 6m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.402788}
{"loss": 24.0, "grad_norm": 1.29042017, "learning_rate": 0.00084599, "token_acc": 0.38167053, "epoch": 0.27932577, "global_step/max_steps": "609/2181", "elapsed_time": "1h 35m 26s", "remaining_time": "4h 6m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.402349}
{"loss": 24.75, "grad_norm": 1.39519191, "learning_rate": 0.00084545, "token_acc": 0.36452933, "epoch": 0.27978443, "global_step/max_steps": "610/2181", "elapsed_time": "1h 35m 36s", "remaining_time": "4h 6m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403951}
{"loss": 25.0, "grad_norm": 1.45326436, "learning_rate": 0.00084492, "token_acc": 0.37199669, "epoch": 0.28024309, "global_step/max_steps": "611/2181", "elapsed_time": "1h 35m 45s", "remaining_time": "4h 6m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403217}
{"loss": 23.5, "grad_norm": 1.40307713, "learning_rate": 0.00084438, "token_acc": 0.39883914, "epoch": 0.28070175, "global_step/max_steps": "612/2181", "elapsed_time": "1h 35m 55s", "remaining_time": "4h 5m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403107}
{"loss": 24.375, "grad_norm": 1.70242023, "learning_rate": 0.00084384, "token_acc": 0.3931113, "epoch": 0.28116042, "global_step/max_steps": "613/2181", "elapsed_time": "1h 36m 4s", "remaining_time": "4h 5m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.402398}
{"loss": 24.75, "grad_norm": 1.32695723, "learning_rate": 0.0008433, "token_acc": 0.38038472, "epoch": 0.28161908, "global_step/max_steps": "614/2181", "elapsed_time": "1h 36m 14s", "remaining_time": "4h 5m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.403651}
{"loss": 24.625, "grad_norm": 1.58413064, "learning_rate": 0.00084276, "token_acc": 0.38106169, "epoch": 0.28207774, "global_step/max_steps": "615/2181", "elapsed_time": "1h 36m 23s", "remaining_time": "4h 5m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.404013}
{"loss": 24.25, "grad_norm": 1.28948271, "learning_rate": 0.00084222, "token_acc": 0.37882859, "epoch": 0.28253641, "global_step/max_steps": "616/2181", "elapsed_time": "1h 36m 33s", "remaining_time": "4h 5m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.404888}
{"loss": 23.375, "grad_norm": 1.38163078, "learning_rate": 0.00084168, "token_acc": 0.38525027, "epoch": 0.28299507, "global_step/max_steps": "617/2181", "elapsed_time": "1h 36m 43s", "remaining_time": "4h 5m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.405171}
{"loss": 24.625, "grad_norm": 1.42479229, "learning_rate": 0.00084113, "token_acc": 0.38300957, "epoch": 0.28345373, "global_step/max_steps": "618/2181", "elapsed_time": "1h 36m 53s", "remaining_time": "4h 5m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406794}
{"loss": 24.25, "grad_norm": 1.39270926, "learning_rate": 0.00084059, "token_acc": 0.38840903, "epoch": 0.2839124, "global_step/max_steps": "619/2181", "elapsed_time": "1h 37m 3s", "remaining_time": "4h 4m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406378}
{"loss": 24.375, "grad_norm": 1.3610779, "learning_rate": 0.00084005, "token_acc": 0.38550247, "epoch": 0.28437106, "global_step/max_steps": "620/2181", "elapsed_time": "1h 37m 12s", "remaining_time": "4h 4m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.407053}
{"loss": 25.0, "grad_norm": 1.33209705, "learning_rate": 0.0008395, "token_acc": 0.3856441, "epoch": 0.28482972, "global_step/max_steps": "621/2181", "elapsed_time": "1h 37m 21s", "remaining_time": "4h 4m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.40646}
{"loss": 24.0, "grad_norm": 1.40572715, "learning_rate": 0.00083896, "token_acc": 0.39683884, "epoch": 0.28528838, "global_step/max_steps": "622/2181", "elapsed_time": "1h 37m 30s", "remaining_time": "4h 4m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.405805}
{"loss": 24.25, "grad_norm": 1.45314026, "learning_rate": 0.00083841, "token_acc": 0.37507066, "epoch": 0.28574705, "global_step/max_steps": "623/2181", "elapsed_time": "1h 37m 40s", "remaining_time": "4h 4m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406162}
{"loss": 24.625, "grad_norm": 1.55468547, "learning_rate": 0.00083786, "token_acc": 0.38944591, "epoch": 0.28620571, "global_step/max_steps": "624/2181", "elapsed_time": "1h 37m 49s", "remaining_time": "4h 4m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406108}
{"loss": 23.875, "grad_norm": 1.3915565, "learning_rate": 0.00083732, "token_acc": 0.38643973, "epoch": 0.28666437, "global_step/max_steps": "625/2181", "elapsed_time": "1h 37m 59s", "remaining_time": "4h 3m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406174}
{"loss": 25.125, "grad_norm": 1.53028464, "learning_rate": 0.00083677, "token_acc": 0.36892386, "epoch": 0.28712304, "global_step/max_steps": "626/2181", "elapsed_time": "1h 38m 8s", "remaining_time": "4h 3m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406229}
{"loss": 24.25, "grad_norm": 1.35126364, "learning_rate": 0.00083622, "token_acc": 0.36693661, "epoch": 0.2875817, "global_step/max_steps": "627/2181", "elapsed_time": "1h 38m 18s", "remaining_time": "4h 3m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406275}
{"loss": 24.75, "grad_norm": 1.37399828, "learning_rate": 0.00083567, "token_acc": 0.37411829, "epoch": 0.28804036, "global_step/max_steps": "628/2181", "elapsed_time": "1h 38m 27s", "remaining_time": "4h 3m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.405619}
{"loss": 25.0, "grad_norm": 1.28006124, "learning_rate": 0.00083512, "token_acc": 0.36788124, "epoch": 0.28849903, "global_step/max_steps": "629/2181", "elapsed_time": "1h 38m 36s", "remaining_time": "4h 3m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406012}
{"loss": 24.0, "grad_norm": 1.4082638, "learning_rate": 0.00083457, "token_acc": 0.3822361, "epoch": 0.28895769, "global_step/max_steps": "630/2181", "elapsed_time": "1h 38m 46s", "remaining_time": "4h 3m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406241}
{"loss": 25.0, "grad_norm": 1.45631182, "learning_rate": 0.00083401, "token_acc": 0.3777533, "epoch": 0.28941635, "global_step/max_steps": "631/2181", "elapsed_time": "1h 38m 55s", "remaining_time": "4h 2m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.405325}
{"loss": 24.375, "grad_norm": 1.33939886, "learning_rate": 0.00083346, "token_acc": 0.38793346, "epoch": 0.28987501, "global_step/max_steps": "632/2181", "elapsed_time": "1h 39m 4s", "remaining_time": "4h 2m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.405787}
{"loss": 24.25, "grad_norm": 1.25336301, "learning_rate": 0.00083291, "token_acc": 0.3797433, "epoch": 0.29033368, "global_step/max_steps": "633/2181", "elapsed_time": "1h 39m 14s", "remaining_time": "4h 2m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406269}
{"loss": 25.0, "grad_norm": 1.45988154, "learning_rate": 0.00083235, "token_acc": 0.37671834, "epoch": 0.29079234, "global_step/max_steps": "634/2181", "elapsed_time": "1h 39m 24s", "remaining_time": "4h 2m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406317}
{"loss": 24.75, "grad_norm": 1.40342629, "learning_rate": 0.0008318, "token_acc": 0.37184703, "epoch": 0.291251, "global_step/max_steps": "635/2181", "elapsed_time": "1h 39m 32s", "remaining_time": "4h 2m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.405376}
{"loss": 24.0, "grad_norm": 1.39405036, "learning_rate": 0.00083124, "token_acc": 0.3771636, "epoch": 0.29170967, "global_step/max_steps": "636/2181", "elapsed_time": "1h 39m 42s", "remaining_time": "4h 2m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.405064}
{"loss": 24.25, "grad_norm": 1.3972199, "learning_rate": 0.00083068, "token_acc": 0.38628159, "epoch": 0.29216833, "global_step/max_steps": "637/2181", "elapsed_time": "1h 39m 52s", "remaining_time": "4h 2m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.407194}
{"loss": 24.375, "grad_norm": 1.33619535, "learning_rate": 0.00083013, "token_acc": 0.37838634, "epoch": 0.29262699, "global_step/max_steps": "638/2181", "elapsed_time": "1h 40m 1s", "remaining_time": "4h 1m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406467}
{"loss": 24.5, "grad_norm": 1.52783191, "learning_rate": 0.00082957, "token_acc": 0.37736917, "epoch": 0.29308566, "global_step/max_steps": "639/2181", "elapsed_time": "1h 40m 11s", "remaining_time": "4h 1m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406912}
{"loss": 24.25, "grad_norm": 1.42842615, "learning_rate": 0.00082901, "token_acc": 0.36942496, "epoch": 0.29354432, "global_step/max_steps": "640/2181", "elapsed_time": "1h 40m 20s", "remaining_time": "4h 1m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.406918}
{"loss": 24.75, "grad_norm": 1.37660789, "learning_rate": 0.00082845, "token_acc": 0.36500673, "epoch": 0.29400298, "global_step/max_steps": "641/2181", "elapsed_time": "1h 40m 31s", "remaining_time": "4h 1m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.408618}
{"loss": 25.25, "grad_norm": 1.37954044, "learning_rate": 0.00082789, "token_acc": 0.3604426, "epoch": 0.29446164, "global_step/max_steps": "642/2181", "elapsed_time": "1h 40m 40s", "remaining_time": "4h 1m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.408345}
{"loss": 25.125, "grad_norm": 1.36420023, "learning_rate": 0.00082733, "token_acc": 0.37751447, "epoch": 0.29492031, "global_step/max_steps": "643/2181", "elapsed_time": "1h 40m 50s", "remaining_time": "4h 1m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.409599}
{"loss": 23.75, "grad_norm": 1.51063097, "learning_rate": 0.00082677, "token_acc": 0.39067703, "epoch": 0.29537897, "global_step/max_steps": "644/2181", "elapsed_time": "1h 40m 60s", "remaining_time": "4h 1m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.409933}
{"loss": 24.75, "grad_norm": 1.3717016, "learning_rate": 0.0008262, "token_acc": 0.36995828, "epoch": 0.29583763, "global_step/max_steps": "645/2181", "elapsed_time": "1h 41m 11s", "remaining_time": "4h 0m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.411817}
{"loss": 24.5, "grad_norm": 1.41485941, "learning_rate": 0.00082564, "token_acc": 0.37245609, "epoch": 0.2962963, "global_step/max_steps": "646/2181", "elapsed_time": "1h 41m 20s", "remaining_time": "4h 0m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.411404}
{"loss": 24.125, "grad_norm": 1.46916199, "learning_rate": 0.00082508, "token_acc": 0.37349398, "epoch": 0.29675496, "global_step/max_steps": "647/2181", "elapsed_time": "1h 41m 29s", "remaining_time": "4h 0m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.411156}
{"loss": 24.125, "grad_norm": 1.33143091, "learning_rate": 0.00082451, "token_acc": 0.3746077, "epoch": 0.29721362, "global_step/max_steps": "648/2181", "elapsed_time": "1h 41m 39s", "remaining_time": "4h 0m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.412625}
{"loss": 24.875, "grad_norm": 1.25501263, "learning_rate": 0.00082395, "token_acc": 0.37631057, "epoch": 0.29767229, "global_step/max_steps": "649/2181", "elapsed_time": "1h 41m 49s", "remaining_time": "4h 0m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413255}
{"loss": 24.125, "grad_norm": 1.53851056, "learning_rate": 0.00082338, "token_acc": 0.3985025, "epoch": 0.29813095, "global_step/max_steps": "650/2181", "elapsed_time": "1h 41m 59s", "remaining_time": "4h 0m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413186}
{"loss": 24.625, "grad_norm": 1.32729328, "learning_rate": 0.00082281, "token_acc": 0.4001692, "epoch": 0.29858961, "global_step/max_steps": "651/2181", "elapsed_time": "1h 42m 8s", "remaining_time": "4h 0m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.412702}
{"loss": 25.0, "grad_norm": 1.29131424, "learning_rate": 0.00082225, "token_acc": 0.36931188, "epoch": 0.29904827, "global_step/max_steps": "652/2181", "elapsed_time": "1h 42m 17s", "remaining_time": "3h 59m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.412571}
{"loss": 23.75, "grad_norm": 1.40181446, "learning_rate": 0.00082168, "token_acc": 0.39045791, "epoch": 0.29950694, "global_step/max_steps": "653/2181", "elapsed_time": "1h 42m 26s", "remaining_time": "3h 59m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.411992}
{"loss": 23.75, "grad_norm": 1.4099623, "learning_rate": 0.00082111, "token_acc": 0.38792867, "epoch": 0.2999656, "global_step/max_steps": "654/2181", "elapsed_time": "1h 42m 36s", "remaining_time": "3h 59m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413221}
{"loss": 25.25, "grad_norm": 1.32266963, "learning_rate": 0.00082054, "token_acc": 0.37165623, "epoch": 0.30042426, "global_step/max_steps": "655/2181", "elapsed_time": "1h 42m 46s", "remaining_time": "3h 59m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413165}
{"loss": 24.0, "grad_norm": 1.35114288, "learning_rate": 0.00081997, "token_acc": 0.38855088, "epoch": 0.30088293, "global_step/max_steps": "656/2181", "elapsed_time": "1h 42m 54s", "remaining_time": "3h 59m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.411969}
{"loss": 24.125, "grad_norm": 1.60523474, "learning_rate": 0.0008194, "token_acc": 0.3901649, "epoch": 0.30134159, "global_step/max_steps": "657/2181", "elapsed_time": "1h 43m 3s", "remaining_time": "3h 59m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.411719}
{"loss": 24.375, "grad_norm": 1.43860281, "learning_rate": 0.00081883, "token_acc": 0.39676228, "epoch": 0.30180025, "global_step/max_steps": "658/2181", "elapsed_time": "1h 43m 14s", "remaining_time": "3h 58m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.412949}
{"loss": 23.75, "grad_norm": 1.25926173, "learning_rate": 0.00081825, "token_acc": 0.39988999, "epoch": 0.30225892, "global_step/max_steps": "659/2181", "elapsed_time": "1h 43m 23s", "remaining_time": "3h 58m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.412461}
{"loss": 24.875, "grad_norm": 1.35972679, "learning_rate": 0.00081768, "token_acc": 0.36549627, "epoch": 0.30271758, "global_step/max_steps": "660/2181", "elapsed_time": "1h 43m 32s", "remaining_time": "3h 58m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.412297}
{"loss": 24.0, "grad_norm": 1.47753203, "learning_rate": 0.00081711, "token_acc": 0.38311688, "epoch": 0.30317624, "global_step/max_steps": "661/2181", "elapsed_time": "1h 43m 41s", "remaining_time": "3h 58m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.411451}
{"loss": 25.375, "grad_norm": 1.2590766, "learning_rate": 0.00081653, "token_acc": 0.36589315, "epoch": 0.3036349, "global_step/max_steps": "662/2181", "elapsed_time": "1h 43m 51s", "remaining_time": "3h 58m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.411988}
{"loss": 23.75, "grad_norm": 1.3164115, "learning_rate": 0.00081596, "token_acc": 0.39827921, "epoch": 0.30409357, "global_step/max_steps": "663/2181", "elapsed_time": "1h 44m 1s", "remaining_time": "3h 58m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.412835}
{"loss": 24.25, "grad_norm": 1.44438124, "learning_rate": 0.00081538, "token_acc": 0.3819464, "epoch": 0.30455223, "global_step/max_steps": "664/2181", "elapsed_time": "1h 44m 11s", "remaining_time": "3h 58m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413894}
{"loss": 24.125, "grad_norm": 1.3659277, "learning_rate": 0.0008148, "token_acc": 0.39536199, "epoch": 0.30501089, "global_step/max_steps": "665/2181", "elapsed_time": "1h 44m 21s", "remaining_time": "3h 57m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.415267}
{"loss": 24.125, "grad_norm": 1.45076108, "learning_rate": 0.00081423, "token_acc": 0.39275766, "epoch": 0.30546956, "global_step/max_steps": "666/2181", "elapsed_time": "1h 44m 30s", "remaining_time": "3h 57m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.415071}
{"loss": 23.875, "grad_norm": 1.4189266, "learning_rate": 0.00081365, "token_acc": 0.38525269, "epoch": 0.30592822, "global_step/max_steps": "667/2181", "elapsed_time": "1h 44m 39s", "remaining_time": "3h 57m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.414331}
{"loss": 24.25, "grad_norm": 1.4015274, "learning_rate": 0.00081307, "token_acc": 0.37496492, "epoch": 0.30638688, "global_step/max_steps": "668/2181", "elapsed_time": "1h 44m 49s", "remaining_time": "3h 57m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.415046}
{"loss": 24.375, "grad_norm": 1.28702438, "learning_rate": 0.00081249, "token_acc": 0.3871061, "epoch": 0.30684555, "global_step/max_steps": "669/2181", "elapsed_time": "1h 44m 58s", "remaining_time": "3h 57m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413941}
{"loss": 24.5, "grad_norm": 1.28620863, "learning_rate": 0.00081191, "token_acc": 0.39053592, "epoch": 0.30730421, "global_step/max_steps": "670/2181", "elapsed_time": "1h 45m 7s", "remaining_time": "3h 57m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.414112}
{"loss": 24.25, "grad_norm": 1.44366097, "learning_rate": 0.00081133, "token_acc": 0.39437796, "epoch": 0.30776287, "global_step/max_steps": "671/2181", "elapsed_time": "1h 45m 17s", "remaining_time": "3h 56m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413997}
{"loss": 24.5, "grad_norm": 1.42250848, "learning_rate": 0.00081075, "token_acc": 0.38238542, "epoch": 0.30822153, "global_step/max_steps": "672/2181", "elapsed_time": "1h 45m 26s", "remaining_time": "3h 56m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413075}
{"loss": 23.75, "grad_norm": 1.38767648, "learning_rate": 0.00081017, "token_acc": 0.39237604, "epoch": 0.3086802, "global_step/max_steps": "673/2181", "elapsed_time": "1h 45m 36s", "remaining_time": "3h 56m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.414186}
{"loss": 24.0, "grad_norm": 1.34042501, "learning_rate": 0.00080958, "token_acc": 0.40130862, "epoch": 0.30913886, "global_step/max_steps": "674/2181", "elapsed_time": "1h 45m 45s", "remaining_time": "3h 56m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.414033}
{"loss": 25.375, "grad_norm": 1.37425828, "learning_rate": 0.000809, "token_acc": 0.36217009, "epoch": 0.30959752, "global_step/max_steps": "675/2181", "elapsed_time": "1h 45m 54s", "remaining_time": "3h 56m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413693}
{"loss": 24.375, "grad_norm": 1.32529855, "learning_rate": 0.00080842, "token_acc": 0.38369565, "epoch": 0.31005619, "global_step/max_steps": "676/2181", "elapsed_time": "1h 46m 3s", "remaining_time": "3h 56m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413184}
{"loss": 23.75, "grad_norm": 1.56001401, "learning_rate": 0.00080783, "token_acc": 0.38574598, "epoch": 0.31051485, "global_step/max_steps": "677/2181", "elapsed_time": "1h 46m 13s", "remaining_time": "3h 55m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413234}
{"loss": 24.125, "grad_norm": 1.2610482, "learning_rate": 0.00080725, "token_acc": 0.37458007, "epoch": 0.31097351, "global_step/max_steps": "678/2181", "elapsed_time": "1h 46m 22s", "remaining_time": "3h 55m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.412259}
{"loss": 24.5, "grad_norm": 1.58286536, "learning_rate": 0.00080666, "token_acc": 0.37850467, "epoch": 0.31143218, "global_step/max_steps": "679/2181", "elapsed_time": "1h 46m 31s", "remaining_time": "3h 55m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.412676}
{"loss": 24.5, "grad_norm": 1.34080338, "learning_rate": 0.00080607, "token_acc": 0.36899441, "epoch": 0.31189084, "global_step/max_steps": "680/2181", "elapsed_time": "1h 46m 42s", "remaining_time": "3h 55m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.414196}
{"loss": 25.0, "grad_norm": 1.40265965, "learning_rate": 0.00080548, "token_acc": 0.37617387, "epoch": 0.3123495, "global_step/max_steps": "681/2181", "elapsed_time": "1h 46m 52s", "remaining_time": "3h 55m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.415252}
{"loss": 23.75, "grad_norm": 1.30779338, "learning_rate": 0.0008049, "token_acc": 0.39043254, "epoch": 0.31280816, "global_step/max_steps": "682/2181", "elapsed_time": "1h 47m 1s", "remaining_time": "3h 55m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.414779}
{"loss": 24.75, "grad_norm": 1.42983127, "learning_rate": 0.00080431, "token_acc": 0.37604457, "epoch": 0.31326683, "global_step/max_steps": "683/2181", "elapsed_time": "1h 47m 10s", "remaining_time": "3h 55m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.414892}
{"loss": 24.5, "grad_norm": 1.37061989, "learning_rate": 0.00080372, "token_acc": 0.36811129, "epoch": 0.31372549, "global_step/max_steps": "684/2181", "elapsed_time": "1h 47m 20s", "remaining_time": "3h 54m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.414589}
{"loss": 23.75, "grad_norm": 1.39099622, "learning_rate": 0.00080313, "token_acc": 0.39612425, "epoch": 0.31418415, "global_step/max_steps": "685/2181", "elapsed_time": "1h 47m 28s", "remaining_time": "3h 54m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413755}
{"loss": 23.875, "grad_norm": 1.37300432, "learning_rate": 0.00080254, "token_acc": 0.3915893, "epoch": 0.31464282, "global_step/max_steps": "686/2181", "elapsed_time": "1h 47m 38s", "remaining_time": "3h 54m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413324}
{"loss": 24.75, "grad_norm": 1.28689384, "learning_rate": 0.00080194, "token_acc": 0.37602257, "epoch": 0.31510148, "global_step/max_steps": "687/2181", "elapsed_time": "1h 47m 47s", "remaining_time": "3h 54m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413438}
{"loss": 25.0, "grad_norm": 1.27228642, "learning_rate": 0.00080135, "token_acc": 0.36874831, "epoch": 0.31556014, "global_step/max_steps": "688/2181", "elapsed_time": "1h 47m 57s", "remaining_time": "3h 54m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413524}
{"loss": 24.0, "grad_norm": 1.42052686, "learning_rate": 0.00080076, "token_acc": 0.38485017, "epoch": 0.31601881, "global_step/max_steps": "689/2181", "elapsed_time": "1h 48m 6s", "remaining_time": "3h 54m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.41385}
{"loss": 24.5, "grad_norm": 1.26241469, "learning_rate": 0.00080017, "token_acc": 0.38649625, "epoch": 0.31647747, "global_step/max_steps": "690/2181", "elapsed_time": "1h 48m 15s", "remaining_time": "3h 53m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413575}
{"loss": 24.25, "grad_norm": 1.30287969, "learning_rate": 0.00079957, "token_acc": 0.3777536, "epoch": 0.31693613, "global_step/max_steps": "691/2181", "elapsed_time": "1h 48m 25s", "remaining_time": "3h 53m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.413759}
{"loss": 24.0, "grad_norm": 1.38964248, "learning_rate": 0.00079898, "token_acc": 0.38273341, "epoch": 0.31739479, "global_step/max_steps": "692/2181", "elapsed_time": "1h 48m 35s", "remaining_time": "3h 53m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.414675}
{"loss": 24.0, "grad_norm": 1.24562299, "learning_rate": 0.00079838, "token_acc": 0.39923643, "epoch": 0.31785346, "global_step/max_steps": "693/2181", "elapsed_time": "1h 48m 45s", "remaining_time": "3h 53m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.415288}
{"loss": 24.5, "grad_norm": 1.32923591, "learning_rate": 0.00079778, "token_acc": 0.37851144, "epoch": 0.31831212, "global_step/max_steps": "694/2181", "elapsed_time": "1h 48m 55s", "remaining_time": "3h 53m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.416201}
{"loss": 24.625, "grad_norm": 1.20829928, "learning_rate": 0.00079719, "token_acc": 0.37184116, "epoch": 0.31877078, "global_step/max_steps": "695/2181", "elapsed_time": "1h 49m 5s", "remaining_time": "3h 53m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.417733}
{"loss": 24.75, "grad_norm": 1.40228689, "learning_rate": 0.00079659, "token_acc": 0.3744089, "epoch": 0.31922945, "global_step/max_steps": "696/2181", "elapsed_time": "1h 49m 15s", "remaining_time": "3h 53m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.417426}
{"loss": 24.875, "grad_norm": 1.28417504, "learning_rate": 0.00079599, "token_acc": 0.3850812, "epoch": 0.31968811, "global_step/max_steps": "697/2181", "elapsed_time": "1h 49m 24s", "remaining_time": "3h 52m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.418061}
{"loss": 24.0, "grad_norm": 1.33255744, "learning_rate": 0.00079539, "token_acc": 0.38769144, "epoch": 0.32014677, "global_step/max_steps": "698/2181", "elapsed_time": "1h 49m 34s", "remaining_time": "3h 52m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.417929}
{"loss": 24.5, "grad_norm": 1.2723614, "learning_rate": 0.00079479, "token_acc": 0.37074924, "epoch": 0.32060544, "global_step/max_steps": "699/2181", "elapsed_time": "1h 49m 43s", "remaining_time": "3h 52m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.418074}
{"loss": 25.0, "grad_norm": 1.36904955, "learning_rate": 0.00079419, "token_acc": 0.37403952, "epoch": 0.3210641, "global_step/max_steps": "700/2181", "elapsed_time": "1h 49m 54s", "remaining_time": "3h 52m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.419742}
{"loss": 24.75, "grad_norm": 1.33916354, "learning_rate": 0.00079359, "token_acc": 0.36328872, "epoch": 0.32152276, "global_step/max_steps": "701/2181", "elapsed_time": "1h 50m 4s", "remaining_time": "3h 52m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.420812}
{"loss": 24.125, "grad_norm": 1.39952779, "learning_rate": 0.00079299, "token_acc": 0.38193869, "epoch": 0.32198142, "global_step/max_steps": "702/2181", "elapsed_time": "1h 50m 13s", "remaining_time": "3h 52m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.41982}
{"loss": 23.625, "grad_norm": 1.37805736, "learning_rate": 0.00079239, "token_acc": 0.39727149, "epoch": 0.32244009, "global_step/max_steps": "703/2181", "elapsed_time": "1h 50m 22s", "remaining_time": "3h 52m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.419763}
{"loss": 24.5, "grad_norm": 1.28790331, "learning_rate": 0.00079179, "token_acc": 0.3718452, "epoch": 0.32289875, "global_step/max_steps": "704/2181", "elapsed_time": "1h 50m 32s", "remaining_time": "3h 51m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.420748}
{"loss": 24.25, "grad_norm": 1.31188989, "learning_rate": 0.00079118, "token_acc": 0.37888707, "epoch": 0.32335741, "global_step/max_steps": "705/2181", "elapsed_time": "1h 50m 41s", "remaining_time": "3h 51m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.420001}
{"loss": 24.625, "grad_norm": 1.40843046, "learning_rate": 0.00079058, "token_acc": 0.37766554, "epoch": 0.32381608, "global_step/max_steps": "706/2181", "elapsed_time": "1h 50m 51s", "remaining_time": "3h 51m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.420422}
{"loss": 24.25, "grad_norm": 1.51514792, "learning_rate": 0.00078997, "token_acc": 0.37587696, "epoch": 0.32427474, "global_step/max_steps": "707/2181", "elapsed_time": "1h 51m 0s", "remaining_time": "3h 51m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.420174}
{"loss": 25.125, "grad_norm": 1.31028092, "learning_rate": 0.00078937, "token_acc": 0.36383866, "epoch": 0.3247334, "global_step/max_steps": "708/2181", "elapsed_time": "1h 51m 10s", "remaining_time": "3h 51m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.420811}
{"loss": 24.125, "grad_norm": 1.57432032, "learning_rate": 0.00078876, "token_acc": 0.39752066, "epoch": 0.32519207, "global_step/max_steps": "709/2181", "elapsed_time": "1h 51m 19s", "remaining_time": "3h 51m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.42047}
{"loss": 24.75, "grad_norm": 1.6856792, "learning_rate": 0.00078816, "token_acc": 0.3839756, "epoch": 0.32565073, "global_step/max_steps": "710/2181", "elapsed_time": "1h 51m 30s", "remaining_time": "3h 51m 0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.422485}
{"loss": 24.25, "grad_norm": 1.48385167, "learning_rate": 0.00078755, "token_acc": 0.37264284, "epoch": 0.32610939, "global_step/max_steps": "711/2181", "elapsed_time": "1h 51m 40s", "remaining_time": "3h 50m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.423339}
{"loss": 24.375, "grad_norm": 1.45170176, "learning_rate": 0.00078694, "token_acc": 0.37061814, "epoch": 0.32656805, "global_step/max_steps": "712/2181", "elapsed_time": "1h 51m 49s", "remaining_time": "3h 50m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.423249}
{"loss": 24.25, "grad_norm": 1.44864285, "learning_rate": 0.00078633, "token_acc": 0.36470911, "epoch": 0.32702672, "global_step/max_steps": "713/2181", "elapsed_time": "1h 51m 59s", "remaining_time": "3h 50m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.423078}
{"loss": 24.625, "grad_norm": 1.45638669, "learning_rate": 0.00078572, "token_acc": 0.38031987, "epoch": 0.32748538, "global_step/max_steps": "714/2181", "elapsed_time": "1h 52m 9s", "remaining_time": "3h 50m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.423799}
{"loss": 24.0, "grad_norm": 1.20179307, "learning_rate": 0.00078511, "token_acc": 0.37770383, "epoch": 0.32794404, "global_step/max_steps": "715/2181", "elapsed_time": "1h 52m 18s", "remaining_time": "3h 50m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.424155}
{"loss": 24.0, "grad_norm": 1.36428416, "learning_rate": 0.0007845, "token_acc": 0.38605162, "epoch": 0.32840271, "global_step/max_steps": "716/2181", "elapsed_time": "1h 52m 28s", "remaining_time": "3h 50m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.424714}
{"loss": 24.125, "grad_norm": 1.33101308, "learning_rate": 0.00078389, "token_acc": 0.38539611, "epoch": 0.32886137, "global_step/max_steps": "717/2181", "elapsed_time": "1h 52m 38s", "remaining_time": "3h 49m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.425287}
{"loss": 24.25, "grad_norm": 1.33616388, "learning_rate": 0.00078328, "token_acc": 0.38079191, "epoch": 0.32932003, "global_step/max_steps": "718/2181", "elapsed_time": "1h 52m 47s", "remaining_time": "3h 49m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.425255}
{"loss": 24.25, "grad_norm": 1.36943793, "learning_rate": 0.00078267, "token_acc": 0.38992485, "epoch": 0.3297787, "global_step/max_steps": "719/2181", "elapsed_time": "1h 52m 57s", "remaining_time": "3h 49m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.425588}
{"loss": 24.625, "grad_norm": 1.45002544, "learning_rate": 0.00078205, "token_acc": 0.36872983, "epoch": 0.33023736, "global_step/max_steps": "720/2181", "elapsed_time": "1h 53m 7s", "remaining_time": "3h 49m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.425855}
{"loss": 24.125, "grad_norm": 1.4078126, "learning_rate": 0.00078144, "token_acc": 0.39007092, "epoch": 0.33069602, "global_step/max_steps": "721/2181", "elapsed_time": "1h 53m 16s", "remaining_time": "3h 49m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.426019}
{"loss": 24.625, "grad_norm": 1.28924263, "learning_rate": 0.00078083, "token_acc": 0.3667407, "epoch": 0.33115468, "global_step/max_steps": "722/2181", "elapsed_time": "1h 53m 25s", "remaining_time": "3h 49m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.425782}
{"loss": 25.0, "grad_norm": 1.28694832, "learning_rate": 0.00078021, "token_acc": 0.36366197, "epoch": 0.33161335, "global_step/max_steps": "723/2181", "elapsed_time": "1h 53m 35s", "remaining_time": "3h 49m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.425879}
{"loss": 24.25, "grad_norm": 1.34153295, "learning_rate": 0.0007796, "token_acc": 0.3850812, "epoch": 0.33207201, "global_step/max_steps": "724/2181", "elapsed_time": "1h 53m 44s", "remaining_time": "3h 48m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.425661}
{"loss": 26.0, "grad_norm": 1.37470698, "learning_rate": 0.00077898, "token_acc": 0.34442877, "epoch": 0.33253067, "global_step/max_steps": "725/2181", "elapsed_time": "1h 53m 54s", "remaining_time": "3h 48m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.425878}
{"loss": 24.5, "grad_norm": 1.27244699, "learning_rate": 0.00077836, "token_acc": 0.38929169, "epoch": 0.33298934, "global_step/max_steps": "726/2181", "elapsed_time": "1h 54m 4s", "remaining_time": "3h 48m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.426382}
{"loss": 24.875, "grad_norm": 1.2207557, "learning_rate": 0.00077775, "token_acc": 0.37828571, "epoch": 0.333448, "global_step/max_steps": "727/2181", "elapsed_time": "1h 54m 13s", "remaining_time": "3h 48m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.425913}
{"loss": 24.25, "grad_norm": 1.33177638, "learning_rate": 0.00077713, "token_acc": 0.37641223, "epoch": 0.33390666, "global_step/max_steps": "728/2181", "elapsed_time": "1h 54m 22s", "remaining_time": "3h 48m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.425898}
{"loss": 24.0, "grad_norm": 1.281829, "learning_rate": 0.00077651, "token_acc": 0.37740113, "epoch": 0.33436533, "global_step/max_steps": "729/2181", "elapsed_time": "1h 54m 32s", "remaining_time": "3h 48m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.426314}
{"loss": 24.625, "grad_norm": 1.4303602, "learning_rate": 0.00077589, "token_acc": 0.36889798, "epoch": 0.33482399, "global_step/max_steps": "730/2181", "elapsed_time": "1h 54m 41s", "remaining_time": "3h 47m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.425928}
{"loss": 24.5, "grad_norm": 1.3454535, "learning_rate": 0.00077527, "token_acc": 0.38833852, "epoch": 0.33528265, "global_step/max_steps": "731/2181", "elapsed_time": "1h 54m 51s", "remaining_time": "3h 47m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.426346}
{"loss": 23.75, "grad_norm": 1.30069506, "learning_rate": 0.00077465, "token_acc": 0.38050661, "epoch": 0.33574131, "global_step/max_steps": "732/2181", "elapsed_time": "1h 55m 0s", "remaining_time": "3h 47m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.426825}
{"loss": 24.25, "grad_norm": 1.52869892, "learning_rate": 0.00077403, "token_acc": 0.37615207, "epoch": 0.33619998, "global_step/max_steps": "733/2181", "elapsed_time": "1h 55m 10s", "remaining_time": "3h 47m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.426982}
{"loss": 24.375, "grad_norm": 1.34929788, "learning_rate": 0.00077341, "token_acc": 0.38157895, "epoch": 0.33665864, "global_step/max_steps": "734/2181", "elapsed_time": "1h 55m 19s", "remaining_time": "3h 47m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.426722}
{"loss": 24.375, "grad_norm": 1.33509612, "learning_rate": 0.00077279, "token_acc": 0.37570781, "epoch": 0.3371173, "global_step/max_steps": "735/2181", "elapsed_time": "1h 55m 29s", "remaining_time": "3h 47m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.427282}
{"loss": 24.75, "grad_norm": 1.49999917, "learning_rate": 0.00077216, "token_acc": 0.39008681, "epoch": 0.33757597, "global_step/max_steps": "736/2181", "elapsed_time": "1h 55m 39s", "remaining_time": "3h 47m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.427926}
{"loss": 25.25, "grad_norm": 1.27626717, "learning_rate": 0.00077154, "token_acc": 0.370448, "epoch": 0.33803463, "global_step/max_steps": "737/2181", "elapsed_time": "1h 55m 48s", "remaining_time": "3h 46m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.427533}
{"loss": 24.875, "grad_norm": 1.40731454, "learning_rate": 0.00077092, "token_acc": 0.37847123, "epoch": 0.33849329, "global_step/max_steps": "738/2181", "elapsed_time": "1h 55m 58s", "remaining_time": "3h 46m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.428107}
{"loss": 25.0, "grad_norm": 1.23318374, "learning_rate": 0.00077029, "token_acc": 0.36879035, "epoch": 0.33895196, "global_step/max_steps": "739/2181", "elapsed_time": "1h 56m 8s", "remaining_time": "3h 46m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.429146}
{"loss": 24.875, "grad_norm": 1.23244905, "learning_rate": 0.00076967, "token_acc": 0.36373391, "epoch": 0.33941062, "global_step/max_steps": "740/2181", "elapsed_time": "1h 56m 17s", "remaining_time": "3h 46m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.428641}
{"loss": 24.375, "grad_norm": 1.26387787, "learning_rate": 0.00076904, "token_acc": 0.38526912, "epoch": 0.33986928, "global_step/max_steps": "741/2181", "elapsed_time": "1h 56m 27s", "remaining_time": "3h 46m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.429744}
{"loss": 24.625, "grad_norm": 1.49021554, "learning_rate": 0.00076841, "token_acc": 0.3726776, "epoch": 0.34032794, "global_step/max_steps": "742/2181", "elapsed_time": "1h 56m 38s", "remaining_time": "3h 46m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.43113}
{"loss": 24.75, "grad_norm": 1.3780663, "learning_rate": 0.00076779, "token_acc": 0.38157529, "epoch": 0.34078661, "global_step/max_steps": "743/2181", "elapsed_time": "1h 56m 47s", "remaining_time": "3h 46m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.430519}
{"loss": 24.375, "grad_norm": 1.32302451, "learning_rate": 0.00076716, "token_acc": 0.38294879, "epoch": 0.34124527, "global_step/max_steps": "744/2181", "elapsed_time": "1h 56m 56s", "remaining_time": "3h 45m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.430579}
{"loss": 24.125, "grad_norm": 1.34543514, "learning_rate": 0.00076653, "token_acc": 0.39157478, "epoch": 0.34170393, "global_step/max_steps": "745/2181", "elapsed_time": "1h 57m 6s", "remaining_time": "3h 45m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.430564}
{"loss": 24.25, "grad_norm": 1.37320542, "learning_rate": 0.0007659, "token_acc": 0.37956403, "epoch": 0.3421626, "global_step/max_steps": "746/2181", "elapsed_time": "1h 57m 15s", "remaining_time": "3h 45m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.430742}
{"loss": 24.0, "grad_norm": 1.33094656, "learning_rate": 0.00076527, "token_acc": 0.37768739, "epoch": 0.34262126, "global_step/max_steps": "747/2181", "elapsed_time": "1h 57m 25s", "remaining_time": "3h 45m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.430694}
{"loss": 25.0, "grad_norm": 1.30941892, "learning_rate": 0.00076464, "token_acc": 0.37755961, "epoch": 0.34307992, "global_step/max_steps": "748/2181", "elapsed_time": "1h 57m 34s", "remaining_time": "3h 45m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.430592}
{"loss": 24.5, "grad_norm": 1.25827146, "learning_rate": 0.00076401, "token_acc": 0.37317469, "epoch": 0.34353859, "global_step/max_steps": "749/2181", "elapsed_time": "1h 57m 44s", "remaining_time": "3h 45m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.431156}
{"loss": 24.625, "grad_norm": 1.33308387, "learning_rate": 0.00076338, "token_acc": 0.38474388, "epoch": 0.34399725, "global_step/max_steps": "750/2181", "elapsed_time": "1h 57m 54s", "remaining_time": "3h 44m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.431817}
{"loss": 24.5, "grad_norm": 1.35451162, "learning_rate": 0.00076275, "token_acc": 0.3816195, "epoch": 0.34445591, "global_step/max_steps": "751/2181", "elapsed_time": "1h 58m 3s", "remaining_time": "3h 44m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.431198}
{"loss": 24.125, "grad_norm": 1.29157472, "learning_rate": 0.00076212, "token_acc": 0.38237762, "epoch": 0.34491457, "global_step/max_steps": "752/2181", "elapsed_time": "1h 58m 13s", "remaining_time": "3h 44m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.431672}
{"loss": 25.25, "grad_norm": 1.53870058, "learning_rate": 0.00076149, "token_acc": 0.37286759, "epoch": 0.34537324, "global_step/max_steps": "753/2181", "elapsed_time": "1h 58m 22s", "remaining_time": "3h 44m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.431952}
{"loss": 24.125, "grad_norm": 1.25829852, "learning_rate": 0.00076085, "token_acc": 0.3843035, "epoch": 0.3458319, "global_step/max_steps": "754/2181", "elapsed_time": "1h 58m 31s", "remaining_time": "3h 44m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.431049}
{"loss": 24.875, "grad_norm": 1.33899045, "learning_rate": 0.00076022, "token_acc": 0.36825397, "epoch": 0.34629056, "global_step/max_steps": "755/2181", "elapsed_time": "1h 58m 41s", "remaining_time": "3h 44m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.432047}
{"loss": 25.0, "grad_norm": 1.19555175, "learning_rate": 0.00075959, "token_acc": 0.37838608, "epoch": 0.34674923, "global_step/max_steps": "756/2181", "elapsed_time": "1h 58m 51s", "remaining_time": "3h 44m 0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.431896}
{"loss": 24.375, "grad_norm": 1.29124951, "learning_rate": 0.00075895, "token_acc": 0.39093724, "epoch": 0.34720789, "global_step/max_steps": "757/2181", "elapsed_time": "1h 58m 59s", "remaining_time": "3h 43m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.431044}
{"loss": 23.5, "grad_norm": 1.1042366, "learning_rate": 0.00075831, "token_acc": 0.38908499, "epoch": 0.34766655, "global_step/max_steps": "758/2181", "elapsed_time": "1h 59m 9s", "remaining_time": "3h 43m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.431332}
{"loss": 24.25, "grad_norm": 1.37554955, "learning_rate": 0.00075768, "token_acc": 0.37937967, "epoch": 0.34812521, "global_step/max_steps": "759/2181", "elapsed_time": "1h 59m 19s", "remaining_time": "3h 43m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.432184}
{"loss": 25.125, "grad_norm": 1.18697286, "learning_rate": 0.00075704, "token_acc": 0.35979586, "epoch": 0.34858388, "global_step/max_steps": "760/2181", "elapsed_time": "1h 59m 28s", "remaining_time": "3h 43m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.431406}
{"loss": 23.625, "grad_norm": 1.23473907, "learning_rate": 0.0007564, "token_acc": 0.37940842, "epoch": 0.34904254, "global_step/max_steps": "761/2181", "elapsed_time": "1h 59m 37s", "remaining_time": "3h 43m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.431188}
{"loss": 24.25, "grad_norm": 1.21930516, "learning_rate": 0.00075577, "token_acc": 0.38541946, "epoch": 0.3495012, "global_step/max_steps": "762/2181", "elapsed_time": "1h 59m 47s", "remaining_time": "3h 43m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.432344}
{"loss": 24.5, "grad_norm": 1.39914131, "learning_rate": 0.00075513, "token_acc": 0.38361948, "epoch": 0.34995987, "global_step/max_steps": "763/2181", "elapsed_time": "1h 59m 57s", "remaining_time": "3h 42m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.432717}
{"loss": 23.625, "grad_norm": 1.21388006, "learning_rate": 0.00075449, "token_acc": 0.37647381, "epoch": 0.35041853, "global_step/max_steps": "764/2181", "elapsed_time": "2h 0m 6s", "remaining_time": "3h 42m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.431702}
{"loss": 24.25, "grad_norm": 1.46894753, "learning_rate": 0.00075385, "token_acc": 0.37625179, "epoch": 0.35087719, "global_step/max_steps": "765/2181", "elapsed_time": "2h 0m 16s", "remaining_time": "3h 42m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.432115}
{"loss": 24.125, "grad_norm": 1.21420157, "learning_rate": 0.00075321, "token_acc": 0.39433428, "epoch": 0.35133586, "global_step/max_steps": "766/2181", "elapsed_time": "2h 0m 26s", "remaining_time": "3h 42m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.432905}
{"loss": 25.0, "grad_norm": 1.37421989, "learning_rate": 0.00075257, "token_acc": 0.37981859, "epoch": 0.35179452, "global_step/max_steps": "767/2181", "elapsed_time": "2h 0m 35s", "remaining_time": "3h 42m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.432568}
{"loss": 24.625, "grad_norm": 1.38345253, "learning_rate": 0.00075193, "token_acc": 0.37851662, "epoch": 0.35225318, "global_step/max_steps": "768/2181", "elapsed_time": "2h 0m 44s", "remaining_time": "3h 42m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.432529}
{"loss": 24.625, "grad_norm": 1.35129929, "learning_rate": 0.00075129, "token_acc": 0.37569709, "epoch": 0.35271184, "global_step/max_steps": "769/2181", "elapsed_time": "2h 0m 54s", "remaining_time": "3h 41m 60s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.433086}
{"loss": 24.125, "grad_norm": 1.36333013, "learning_rate": 0.00075064, "token_acc": 0.39158128, "epoch": 0.35317051, "global_step/max_steps": "770/2181", "elapsed_time": "2h 1m 4s", "remaining_time": "3h 41m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.433484}
{"loss": 23.875, "grad_norm": 1.24982631, "learning_rate": 0.00075, "token_acc": 0.38907563, "epoch": 0.35362917, "global_step/max_steps": "771/2181", "elapsed_time": "2h 1m 13s", "remaining_time": "3h 41m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.433739}
{"loss": 24.75, "grad_norm": 1.30397069, "learning_rate": 0.00074936, "token_acc": 0.373394, "epoch": 0.35408783, "global_step/max_steps": "772/2181", "elapsed_time": "2h 1m 23s", "remaining_time": "3h 41m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.433969}
{"loss": 24.0, "grad_norm": 1.292974, "learning_rate": 0.00074871, "token_acc": 0.38026206, "epoch": 0.3545465, "global_step/max_steps": "773/2181", "elapsed_time": "2h 1m 33s", "remaining_time": "3h 41m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.434483}
{"loss": 24.125, "grad_norm": 1.32517111, "learning_rate": 0.00074807, "token_acc": 0.38348573, "epoch": 0.35500516, "global_step/max_steps": "774/2181", "elapsed_time": "2h 1m 42s", "remaining_time": "3h 41m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.434592}
{"loss": 24.625, "grad_norm": 1.27435613, "learning_rate": 0.00074742, "token_acc": 0.37392031, "epoch": 0.35546382, "global_step/max_steps": "775/2181", "elapsed_time": "2h 1m 52s", "remaining_time": "3h 41m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.435064}
{"loss": 24.5, "grad_norm": 1.18046474, "learning_rate": 0.00074678, "token_acc": 0.37361726, "epoch": 0.35592249, "global_step/max_steps": "776/2181", "elapsed_time": "2h 2m 2s", "remaining_time": "3h 40m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.435983}
{"loss": 23.75, "grad_norm": 1.46768701, "learning_rate": 0.00074613, "token_acc": 0.39446953, "epoch": 0.35638115, "global_step/max_steps": "777/2181", "elapsed_time": "2h 2m 12s", "remaining_time": "3h 40m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.436848}
{"loss": 24.25, "grad_norm": 1.30016971, "learning_rate": 0.00074548, "token_acc": 0.38244426, "epoch": 0.35683981, "global_step/max_steps": "778/2181", "elapsed_time": "2h 2m 22s", "remaining_time": "3h 40m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.436543}
{"loss": 24.0, "grad_norm": 1.2471813, "learning_rate": 0.00074484, "token_acc": 0.38974937, "epoch": 0.35729847, "global_step/max_steps": "779/2181", "elapsed_time": "2h 2m 31s", "remaining_time": "3h 40m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.436667}
{"loss": 23.625, "grad_norm": 1.45639944, "learning_rate": 0.00074419, "token_acc": 0.39896522, "epoch": 0.35775714, "global_step/max_steps": "780/2181", "elapsed_time": "2h 2m 41s", "remaining_time": "3h 40m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.437688}
{"loss": 24.0, "grad_norm": 1.26547348, "learning_rate": 0.00074354, "token_acc": 0.3917869, "epoch": 0.3582158, "global_step/max_steps": "781/2181", "elapsed_time": "2h 2m 52s", "remaining_time": "3h 40m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.438693}
{"loss": 23.0, "grad_norm": 1.27159119, "learning_rate": 0.00074289, "token_acc": 0.40717241, "epoch": 0.35867446, "global_step/max_steps": "782/2181", "elapsed_time": "2h 3m 1s", "remaining_time": "3h 40m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.438826}
{"loss": 24.0, "grad_norm": 1.20329392, "learning_rate": 0.00074224, "token_acc": 0.40095087, "epoch": 0.35913313, "global_step/max_steps": "783/2181", "elapsed_time": "2h 3m 11s", "remaining_time": "3h 39m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.43874}
{"loss": 24.375, "grad_norm": 1.29692459, "learning_rate": 0.00074159, "token_acc": 0.39795629, "epoch": 0.35959179, "global_step/max_steps": "784/2181", "elapsed_time": "2h 3m 20s", "remaining_time": "3h 39m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.438612}
{"loss": 24.75, "grad_norm": 1.34299684, "learning_rate": 0.00074094, "token_acc": 0.37885096, "epoch": 0.36005045, "global_step/max_steps": "785/2181", "elapsed_time": "2h 3m 29s", "remaining_time": "3h 39m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.438539}
{"loss": 24.25, "grad_norm": 1.31699026, "learning_rate": 0.00074029, "token_acc": 0.37791028, "epoch": 0.36050912, "global_step/max_steps": "786/2181", "elapsed_time": "2h 3m 38s", "remaining_time": "3h 39m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.438245}
{"loss": 24.625, "grad_norm": 1.30071688, "learning_rate": 0.00073964, "token_acc": 0.37546667, "epoch": 0.36096778, "global_step/max_steps": "787/2181", "elapsed_time": "2h 3m 47s", "remaining_time": "3h 39m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.437692}
{"loss": 24.25, "grad_norm": 1.37711477, "learning_rate": 0.00073899, "token_acc": 0.38083049, "epoch": 0.36142644, "global_step/max_steps": "788/2181", "elapsed_time": "2h 3m 57s", "remaining_time": "3h 39m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.438409}
{"loss": 24.0, "grad_norm": 1.20874918, "learning_rate": 0.00073833, "token_acc": 0.38948864, "epoch": 0.3618851, "global_step/max_steps": "789/2181", "elapsed_time": "2h 4m 7s", "remaining_time": "3h 38m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.438764}
{"loss": 24.375, "grad_norm": 1.47263336, "learning_rate": 0.00073768, "token_acc": 0.37462152, "epoch": 0.36234377, "global_step/max_steps": "790/2181", "elapsed_time": "2h 4m 17s", "remaining_time": "3h 38m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.439033}
{"loss": 24.625, "grad_norm": 1.22307265, "learning_rate": 0.00073703, "token_acc": 0.3612197, "epoch": 0.36280243, "global_step/max_steps": "791/2181", "elapsed_time": "2h 4m 26s", "remaining_time": "3h 38m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.438905}
{"loss": 23.875, "grad_norm": 1.35229909, "learning_rate": 0.00073637, "token_acc": 0.38059284, "epoch": 0.36326109, "global_step/max_steps": "792/2181", "elapsed_time": "2h 4m 36s", "remaining_time": "3h 38m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.439353}
{"loss": 24.375, "grad_norm": 1.32019734, "learning_rate": 0.00073572, "token_acc": 0.3825014, "epoch": 0.36371976, "global_step/max_steps": "793/2181", "elapsed_time": "2h 4m 46s", "remaining_time": "3h 38m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.4399}
{"loss": 24.25, "grad_norm": 1.30541885, "learning_rate": 0.00073506, "token_acc": 0.38133182, "epoch": 0.36417842, "global_step/max_steps": "794/2181", "elapsed_time": "2h 4m 56s", "remaining_time": "3h 38m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.441037}
{"loss": 24.375, "grad_norm": 1.18825197, "learning_rate": 0.00073441, "token_acc": 0.37710168, "epoch": 0.36463708, "global_step/max_steps": "795/2181", "elapsed_time": "2h 5m 6s", "remaining_time": "3h 38m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.441457}
{"loss": 24.875, "grad_norm": 1.19309831, "learning_rate": 0.00073375, "token_acc": 0.37882157, "epoch": 0.36509575, "global_step/max_steps": "796/2181", "elapsed_time": "2h 5m 16s", "remaining_time": "3h 37m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.441765}
{"loss": 23.375, "grad_norm": 1.362988, "learning_rate": 0.00073309, "token_acc": 0.40864333, "epoch": 0.36555441, "global_step/max_steps": "797/2181", "elapsed_time": "2h 5m 25s", "remaining_time": "3h 37m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.441365}
{"loss": 24.0, "grad_norm": 1.26924527, "learning_rate": 0.00073244, "token_acc": 0.38614146, "epoch": 0.36601307, "global_step/max_steps": "798/2181", "elapsed_time": "2h 5m 35s", "remaining_time": "3h 37m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.441759}
{"loss": 24.0, "grad_norm": 1.36909759, "learning_rate": 0.00073178, "token_acc": 0.39880103, "epoch": 0.36647173, "global_step/max_steps": "799/2181", "elapsed_time": "2h 5m 44s", "remaining_time": "3h 37m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.44216}
{"loss": 24.125, "grad_norm": 1.36963177, "learning_rate": 0.00073112, "token_acc": 0.3875, "epoch": 0.3669304, "global_step/max_steps": "800/2181", "elapsed_time": "2h 5m 54s", "remaining_time": "3h 37m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.44229}
{"loss": 24.625, "grad_norm": 1.37550509, "learning_rate": 0.00073046, "token_acc": 0.37214423, "epoch": 0.36738906, "global_step/max_steps": "801/2181", "elapsed_time": "2h 6m 4s", "remaining_time": "3h 37m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.442832}
{"loss": 24.625, "grad_norm": 1.29087353, "learning_rate": 0.0007298, "token_acc": 0.37496538, "epoch": 0.36784772, "global_step/max_steps": "802/2181", "elapsed_time": "2h 6m 14s", "remaining_time": "3h 37m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.443692}
{"loss": 24.25, "grad_norm": 1.24281871, "learning_rate": 0.00072914, "token_acc": 0.38790723, "epoch": 0.36830639, "global_step/max_steps": "803/2181", "elapsed_time": "2h 6m 23s", "remaining_time": "3h 36m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.443216}
{"loss": 24.625, "grad_norm": 1.34990275, "learning_rate": 0.00072848, "token_acc": 0.37782281, "epoch": 0.36876505, "global_step/max_steps": "804/2181", "elapsed_time": "2h 6m 33s", "remaining_time": "3h 36m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.443655}
{"loss": 24.5, "grad_norm": 1.31884074, "learning_rate": 0.00072782, "token_acc": 0.3688736, "epoch": 0.36922371, "global_step/max_steps": "805/2181", "elapsed_time": "2h 6m 42s", "remaining_time": "3h 36m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.44386}
{"loss": 23.625, "grad_norm": 1.28652096, "learning_rate": 0.00072716, "token_acc": 0.39066407, "epoch": 0.36968238, "global_step/max_steps": "806/2181", "elapsed_time": "2h 6m 52s", "remaining_time": "3h 36m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.443908}
{"loss": 24.25, "grad_norm": 1.4721334, "learning_rate": 0.0007265, "token_acc": 0.38169897, "epoch": 0.37014104, "global_step/max_steps": "807/2181", "elapsed_time": "2h 7m 1s", "remaining_time": "3h 36m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.444082}
{"loss": 25.125, "grad_norm": 1.28627801, "learning_rate": 0.00072584, "token_acc": 0.3646833, "epoch": 0.3705997, "global_step/max_steps": "808/2181", "elapsed_time": "2h 7m 11s", "remaining_time": "3h 36m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.444095}
{"loss": 23.625, "grad_norm": 1.13055384, "learning_rate": 0.00072517, "token_acc": 0.38335322, "epoch": 0.37105836, "global_step/max_steps": "809/2181", "elapsed_time": "2h 7m 20s", "remaining_time": "3h 35m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.443626}
{"loss": 24.5, "grad_norm": 1.30931354, "learning_rate": 0.00072451, "token_acc": 0.38580331, "epoch": 0.37151703, "global_step/max_steps": "810/2181", "elapsed_time": "2h 7m 30s", "remaining_time": "3h 35m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.444725}
{"loss": 24.25, "grad_norm": 1.29051745, "learning_rate": 0.00072385, "token_acc": 0.37677456, "epoch": 0.37197569, "global_step/max_steps": "811/2181", "elapsed_time": "2h 7m 40s", "remaining_time": "3h 35m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.445185}
{"loss": 24.5, "grad_norm": 1.28299427, "learning_rate": 0.00072318, "token_acc": 0.37959533, "epoch": 0.37243435, "global_step/max_steps": "812/2181", "elapsed_time": "2h 7m 50s", "remaining_time": "3h 35m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.446291}
{"loss": 24.5, "grad_norm": 1.19313133, "learning_rate": 0.00072252, "token_acc": 0.37225236, "epoch": 0.37289302, "global_step/max_steps": "813/2181", "elapsed_time": "2h 7m 60s", "remaining_time": "3h 35m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.4461}
{"loss": 25.0, "grad_norm": 1.17393184, "learning_rate": 0.00072185, "token_acc": 0.36439038, "epoch": 0.37335168, "global_step/max_steps": "814/2181", "elapsed_time": "2h 8m 9s", "remaining_time": "3h 35m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.445875}
{"loss": 24.125, "grad_norm": 1.31313837, "learning_rate": 0.00072119, "token_acc": 0.38001708, "epoch": 0.37381034, "global_step/max_steps": "815/2181", "elapsed_time": "2h 8m 18s", "remaining_time": "3h 35m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.445913}
{"loss": 24.625, "grad_norm": 1.26720583, "learning_rate": 0.00072052, "token_acc": 0.3762647, "epoch": 0.37426901, "global_step/max_steps": "816/2181", "elapsed_time": "2h 8m 28s", "remaining_time": "3h 34m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.446181}
{"loss": 23.125, "grad_norm": 1.26272821, "learning_rate": 0.00071985, "token_acc": 0.39318885, "epoch": 0.37472767, "global_step/max_steps": "817/2181", "elapsed_time": "2h 8m 38s", "remaining_time": "3h 34m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.446298}
{"loss": 24.0, "grad_norm": 1.26274872, "learning_rate": 0.00071919, "token_acc": 0.37951467, "epoch": 0.37518633, "global_step/max_steps": "818/2181", "elapsed_time": "2h 8m 48s", "remaining_time": "3h 34m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.446948}
{"loss": 24.375, "grad_norm": 1.34691679, "learning_rate": 0.00071852, "token_acc": 0.38967777, "epoch": 0.37564499, "global_step/max_steps": "819/2181", "elapsed_time": "2h 8m 57s", "remaining_time": "3h 34m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.446691}
{"loss": 24.0, "grad_norm": 1.28419805, "learning_rate": 0.00071785, "token_acc": 0.38931931, "epoch": 0.37610366, "global_step/max_steps": "820/2181", "elapsed_time": "2h 9m 7s", "remaining_time": "3h 34m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.447925}
{"loss": 24.375, "grad_norm": 1.54067826, "learning_rate": 0.00071718, "token_acc": 0.37127748, "epoch": 0.37656232, "global_step/max_steps": "821/2181", "elapsed_time": "2h 9m 17s", "remaining_time": "3h 34m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.448755}
{"loss": 24.625, "grad_norm": 1.24039364, "learning_rate": 0.00071651, "token_acc": 0.37706271, "epoch": 0.37702098, "global_step/max_steps": "822/2181", "elapsed_time": "2h 9m 26s", "remaining_time": "3h 33m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.447748}
{"loss": 24.5, "grad_norm": 1.3682996, "learning_rate": 0.00071584, "token_acc": 0.38734315, "epoch": 0.37747965, "global_step/max_steps": "823/2181", "elapsed_time": "2h 9m 36s", "remaining_time": "3h 33m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.448081}
{"loss": 24.25, "grad_norm": 1.26911807, "learning_rate": 0.00071517, "token_acc": 0.38809723, "epoch": 0.37793831, "global_step/max_steps": "824/2181", "elapsed_time": "2h 9m 45s", "remaining_time": "3h 33m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.448007}
{"loss": 23.625, "grad_norm": 1.27654326, "learning_rate": 0.0007145, "token_acc": 0.39536785, "epoch": 0.37839697, "global_step/max_steps": "825/2181", "elapsed_time": "2h 9m 54s", "remaining_time": "3h 33m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.447425}
{"loss": 24.875, "grad_norm": 1.22835255, "learning_rate": 0.00071383, "token_acc": 0.37101607, "epoch": 0.37885564, "global_step/max_steps": "826/2181", "elapsed_time": "2h 10m 4s", "remaining_time": "3h 33m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.44742}
{"loss": 24.875, "grad_norm": 1.17969251, "learning_rate": 0.00071316, "token_acc": 0.36592971, "epoch": 0.3793143, "global_step/max_steps": "827/2181", "elapsed_time": "2h 10m 13s", "remaining_time": "3h 33m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.44785}
{"loss": 24.5, "grad_norm": 1.27108455, "learning_rate": 0.00071249, "token_acc": 0.38257783, "epoch": 0.37977296, "global_step/max_steps": "828/2181", "elapsed_time": "2h 10m 23s", "remaining_time": "3h 33m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.447796}
{"loss": 24.625, "grad_norm": 1.38360083, "learning_rate": 0.00071181, "token_acc": 0.37925926, "epoch": 0.38023162, "global_step/max_steps": "829/2181", "elapsed_time": "2h 10m 31s", "remaining_time": "3h 32m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.4469}
{"loss": 24.75, "grad_norm": 1.28792393, "learning_rate": 0.00071114, "token_acc": 0.37496523, "epoch": 0.38069029, "global_step/max_steps": "830/2181", "elapsed_time": "2h 10m 41s", "remaining_time": "3h 32m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.447384}
{"loss": 24.5, "grad_norm": 1.23429883, "learning_rate": 0.00071047, "token_acc": 0.37411251, "epoch": 0.38114895, "global_step/max_steps": "831/2181", "elapsed_time": "2h 10m 51s", "remaining_time": "3h 32m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.447678}
{"loss": 23.75, "grad_norm": 1.94476926, "learning_rate": 0.00070979, "token_acc": 0.38558856, "epoch": 0.38160761, "global_step/max_steps": "832/2181", "elapsed_time": "2h 11m 0s", "remaining_time": "3h 32m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.447704}
{"loss": 24.875, "grad_norm": 1.20433068, "learning_rate": 0.00070912, "token_acc": 0.37, "epoch": 0.38206628, "global_step/max_steps": "833/2181", "elapsed_time": "2h 11m 10s", "remaining_time": "3h 32m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.447739}
{"loss": 24.625, "grad_norm": 1.35232544, "learning_rate": 0.00070844, "token_acc": 0.37205339, "epoch": 0.38252494, "global_step/max_steps": "834/2181", "elapsed_time": "2h 11m 20s", "remaining_time": "3h 32m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.447891}
{"loss": 23.875, "grad_norm": 1.22702301, "learning_rate": 0.00070777, "token_acc": 0.39642567, "epoch": 0.3829836, "global_step/max_steps": "835/2181", "elapsed_time": "2h 11m 29s", "remaining_time": "3h 31m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.447869}
{"loss": 25.0, "grad_norm": 1.41758263, "learning_rate": 0.00070709, "token_acc": 0.38163037, "epoch": 0.38344227, "global_step/max_steps": "836/2181", "elapsed_time": "2h 11m 38s", "remaining_time": "3h 31m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.447721}
{"loss": 24.0, "grad_norm": 1.33322906, "learning_rate": 0.00070642, "token_acc": 0.38126362, "epoch": 0.38390093, "global_step/max_steps": "837/2181", "elapsed_time": "2h 11m 48s", "remaining_time": "3h 31m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.44803}
{"loss": 24.375, "grad_norm": 1.41735959, "learning_rate": 0.00070574, "token_acc": 0.37728027, "epoch": 0.38435959, "global_step/max_steps": "838/2181", "elapsed_time": "2h 11m 57s", "remaining_time": "3h 31m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.448013}
{"loss": 24.0, "grad_norm": 1.34941792, "learning_rate": 0.00070506, "token_acc": 0.36238136, "epoch": 0.38481825, "global_step/max_steps": "839/2181", "elapsed_time": "2h 12m 7s", "remaining_time": "3h 31m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.44867}
{"loss": 24.25, "grad_norm": 1.22184992, "learning_rate": 0.00070439, "token_acc": 0.37787482, "epoch": 0.38527692, "global_step/max_steps": "840/2181", "elapsed_time": "2h 12m 18s", "remaining_time": "3h 31m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.449627}
{"loss": 23.875, "grad_norm": 1.26132965, "learning_rate": 0.00070371, "token_acc": 0.38391123, "epoch": 0.38573558, "global_step/max_steps": "841/2181", "elapsed_time": "2h 12m 26s", "remaining_time": "3h 31m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.448865}
{"loss": 23.875, "grad_norm": 1.35689569, "learning_rate": 0.00070303, "token_acc": 0.39268222, "epoch": 0.38619424, "global_step/max_steps": "842/2181", "elapsed_time": "2h 12m 36s", "remaining_time": "3h 30m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.449202}
{"loss": 23.75, "grad_norm": 1.26905107, "learning_rate": 0.00070235, "token_acc": 0.3799835, "epoch": 0.38665291, "global_step/max_steps": "843/2181", "elapsed_time": "2h 12m 46s", "remaining_time": "3h 30m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.450035}
{"loss": 24.75, "grad_norm": 1.25970578, "learning_rate": 0.00070167, "token_acc": 0.3778352, "epoch": 0.38711157, "global_step/max_steps": "844/2181", "elapsed_time": "2h 12m 56s", "remaining_time": "3h 30m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.450283}
{"loss": 24.875, "grad_norm": 1.3580488, "learning_rate": 0.00070099, "token_acc": 0.37647059, "epoch": 0.38757023, "global_step/max_steps": "845/2181", "elapsed_time": "2h 13m 6s", "remaining_time": "3h 30m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.45061}
{"loss": 24.0, "grad_norm": 1.25601828, "learning_rate": 0.00070031, "token_acc": 0.39648545, "epoch": 0.3880289, "global_step/max_steps": "846/2181", "elapsed_time": "2h 13m 15s", "remaining_time": "3h 30m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.450934}
{"loss": 25.125, "grad_norm": 1.36827707, "learning_rate": 0.00069963, "token_acc": 0.36368593, "epoch": 0.38848756, "global_step/max_steps": "847/2181", "elapsed_time": "2h 13m 25s", "remaining_time": "3h 30m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.450986}
{"loss": 24.875, "grad_norm": 1.25382936, "learning_rate": 0.00069895, "token_acc": 0.37289944, "epoch": 0.38894622, "global_step/max_steps": "848/2181", "elapsed_time": "2h 13m 34s", "remaining_time": "3h 29m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.450587}
{"loss": 24.75, "grad_norm": 1.2221055, "learning_rate": 0.00069827, "token_acc": 0.38284398, "epoch": 0.38940488, "global_step/max_steps": "849/2181", "elapsed_time": "2h 13m 44s", "remaining_time": "3h 29m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.451266}
{"loss": 24.25, "grad_norm": 1.22137702, "learning_rate": 0.00069759, "token_acc": 0.38555896, "epoch": 0.38986355, "global_step/max_steps": "850/2181", "elapsed_time": "2h 13m 54s", "remaining_time": "3h 29m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.451756}
{"loss": 24.5, "grad_norm": 1.26398361, "learning_rate": 0.0006969, "token_acc": 0.38077028, "epoch": 0.39032221, "global_step/max_steps": "851/2181", "elapsed_time": "2h 14m 3s", "remaining_time": "3h 29m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.451347}
{"loss": 24.375, "grad_norm": 1.35958552, "learning_rate": 0.00069622, "token_acc": 0.37016875, "epoch": 0.39078087, "global_step/max_steps": "852/2181", "elapsed_time": "2h 14m 12s", "remaining_time": "3h 29m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.451185}
{"loss": 23.75, "grad_norm": 1.29658008, "learning_rate": 0.00069554, "token_acc": 0.38162642, "epoch": 0.39123954, "global_step/max_steps": "853/2181", "elapsed_time": "2h 14m 22s", "remaining_time": "3h 29m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.451796}
{"loss": 24.0, "grad_norm": 1.37595105, "learning_rate": 0.00069485, "token_acc": 0.38481375, "epoch": 0.3916982, "global_step/max_steps": "854/2181", "elapsed_time": "2h 14m 32s", "remaining_time": "3h 29m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.451639}
{"loss": 25.0, "grad_norm": 1.32317173, "learning_rate": 0.00069417, "token_acc": 0.36719174, "epoch": 0.39215686, "global_step/max_steps": "855/2181", "elapsed_time": "2h 14m 41s", "remaining_time": "3h 28m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.451474}
{"loss": 23.5, "grad_norm": 1.37776184, "learning_rate": 0.00069348, "token_acc": 0.3994429, "epoch": 0.39261553, "global_step/max_steps": "856/2181", "elapsed_time": "2h 14m 51s", "remaining_time": "3h 28m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.451599}
{"loss": 24.375, "grad_norm": 1.10167921, "learning_rate": 0.0006928, "token_acc": 0.37973274, "epoch": 0.39307419, "global_step/max_steps": "857/2181", "elapsed_time": "2h 14m 60s", "remaining_time": "3h 28m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.451016}
{"loss": 24.625, "grad_norm": 1.37353325, "learning_rate": 0.00069211, "token_acc": 0.38648422, "epoch": 0.39353285, "global_step/max_steps": "858/2181", "elapsed_time": "2h 15m 8s", "remaining_time": "3h 28m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.450373}
{"loss": 23.875, "grad_norm": 1.19727707, "learning_rate": 0.00069143, "token_acc": 0.38945625, "epoch": 0.39399151, "global_step/max_steps": "859/2181", "elapsed_time": "2h 15m 18s", "remaining_time": "3h 28m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.451091}
{"loss": 23.375, "grad_norm": 1.55880344, "learning_rate": 0.00069074, "token_acc": 0.4012792, "epoch": 0.39445018, "global_step/max_steps": "860/2181", "elapsed_time": "2h 15m 28s", "remaining_time": "3h 28m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.450826}
{"loss": 24.125, "grad_norm": 1.19558835, "learning_rate": 0.00069005, "token_acc": 0.37833424, "epoch": 0.39490884, "global_step/max_steps": "861/2181", "elapsed_time": "2h 15m 37s", "remaining_time": "3h 27m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.451001}
{"loss": 24.25, "grad_norm": 1.24253476, "learning_rate": 0.00068937, "token_acc": 0.38999447, "epoch": 0.3953675, "global_step/max_steps": "862/2181", "elapsed_time": "2h 15m 47s", "remaining_time": "3h 27m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.451657}
{"loss": 24.5, "grad_norm": 1.13615727, "learning_rate": 0.00068868, "token_acc": 0.38694178, "epoch": 0.39582617, "global_step/max_steps": "863/2181", "elapsed_time": "2h 15m 57s", "remaining_time": "3h 27m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.45143}
{"loss": 24.25, "grad_norm": 1.31643987, "learning_rate": 0.00068799, "token_acc": 0.37020567, "epoch": 0.39628483, "global_step/max_steps": "864/2181", "elapsed_time": "2h 16m 6s", "remaining_time": "3h 27m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.451928}
{"loss": 23.75, "grad_norm": 1.28695202, "learning_rate": 0.0006873, "token_acc": 0.38398182, "epoch": 0.39674349, "global_step/max_steps": "865/2181", "elapsed_time": "2h 16m 17s", "remaining_time": "3h 27m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.453014}
{"loss": 24.25, "grad_norm": 1.26635456, "learning_rate": 0.00068661, "token_acc": 0.38673442, "epoch": 0.39720216, "global_step/max_steps": "866/2181", "elapsed_time": "2h 16m 27s", "remaining_time": "3h 27m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.45351}
{"loss": 24.5, "grad_norm": 1.26918209, "learning_rate": 0.00068593, "token_acc": 0.37577365, "epoch": 0.39766082, "global_step/max_steps": "867/2181", "elapsed_time": "2h 16m 38s", "remaining_time": "3h 27m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.455032}
{"loss": 24.75, "grad_norm": 1.30185676, "learning_rate": 0.00068524, "token_acc": 0.36858432, "epoch": 0.39811948, "global_step/max_steps": "868/2181", "elapsed_time": "2h 16m 47s", "remaining_time": "3h 26m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.455094}
{"loss": 24.25, "grad_norm": 1.17605174, "learning_rate": 0.00068455, "token_acc": 0.38611482, "epoch": 0.39857814, "global_step/max_steps": "869/2181", "elapsed_time": "2h 16m 57s", "remaining_time": "3h 26m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.455427}
{"loss": 24.75, "grad_norm": 1.27960145, "learning_rate": 0.00068386, "token_acc": 0.3712393, "epoch": 0.39903681, "global_step/max_steps": "870/2181", "elapsed_time": "2h 17m 6s", "remaining_time": "3h 26m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.455261}
{"loss": 24.375, "grad_norm": 1.25352633, "learning_rate": 0.00068316, "token_acc": 0.37077726, "epoch": 0.39949547, "global_step/max_steps": "871/2181", "elapsed_time": "2h 17m 16s", "remaining_time": "3h 26m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.456147}
{"loss": 23.625, "grad_norm": 1.26093471, "learning_rate": 0.00068247, "token_acc": 0.39020857, "epoch": 0.39995413, "global_step/max_steps": "872/2181", "elapsed_time": "2h 17m 26s", "remaining_time": "3h 26m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.45615}
{"loss": 24.75, "grad_norm": 1.22613251, "learning_rate": 0.00068178, "token_acc": 0.37786885, "epoch": 0.4004128, "global_step/max_steps": "873/2181", "elapsed_time": "2h 17m 36s", "remaining_time": "3h 26m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.456862}
{"loss": 25.0, "grad_norm": 1.24749243, "learning_rate": 0.00068109, "token_acc": 0.37956605, "epoch": 0.40087146, "global_step/max_steps": "874/2181", "elapsed_time": "2h 17m 46s", "remaining_time": "3h 26m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.457359}
{"loss": 24.375, "grad_norm": 1.22874594, "learning_rate": 0.0006804, "token_acc": 0.37517591, "epoch": 0.40133012, "global_step/max_steps": "875/2181", "elapsed_time": "2h 17m 55s", "remaining_time": "3h 25m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.45727}
{"loss": 24.875, "grad_norm": 1.40018082, "learning_rate": 0.0006797, "token_acc": 0.36834812, "epoch": 0.40178879, "global_step/max_steps": "876/2181", "elapsed_time": "2h 18m 4s", "remaining_time": "3h 25m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.456452}
{"loss": 24.0, "grad_norm": 1.36077225, "learning_rate": 0.00067901, "token_acc": 0.38823206, "epoch": 0.40224745, "global_step/max_steps": "877/2181", "elapsed_time": "2h 18m 13s", "remaining_time": "3h 25m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.456419}
{"loss": 24.375, "grad_norm": 1.20215023, "learning_rate": 0.00067832, "token_acc": 0.37082153, "epoch": 0.40270611, "global_step/max_steps": "878/2181", "elapsed_time": "2h 18m 23s", "remaining_time": "3h 25m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.456502}
{"loss": 24.875, "grad_norm": 1.25218809, "learning_rate": 0.00067762, "token_acc": 0.37163121, "epoch": 0.40316477, "global_step/max_steps": "879/2181", "elapsed_time": "2h 18m 32s", "remaining_time": "3h 25m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.456555}
{"loss": 24.5, "grad_norm": 1.27008057, "learning_rate": 0.00067693, "token_acc": 0.38159752, "epoch": 0.40362344, "global_step/max_steps": "880/2181", "elapsed_time": "2h 18m 42s", "remaining_time": "3h 25m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.45669}
{"loss": 24.125, "grad_norm": 1.2257911, "learning_rate": 0.00067623, "token_acc": 0.38823865, "epoch": 0.4040821, "global_step/max_steps": "881/2181", "elapsed_time": "2h 18m 51s", "remaining_time": "3h 24m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.456518}
{"loss": 24.125, "grad_norm": 1.23250282, "learning_rate": 0.00067554, "token_acc": 0.37335203, "epoch": 0.40454076, "global_step/max_steps": "882/2181", "elapsed_time": "2h 19m 1s", "remaining_time": "3h 24m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.456805}
{"loss": 24.5, "grad_norm": 1.28635335, "learning_rate": 0.00067484, "token_acc": 0.36964874, "epoch": 0.40499943, "global_step/max_steps": "883/2181", "elapsed_time": "2h 19m 10s", "remaining_time": "3h 24m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.456362}
{"loss": 24.625, "grad_norm": 1.33061683, "learning_rate": 0.00067415, "token_acc": 0.37936819, "epoch": 0.40545809, "global_step/max_steps": "884/2181", "elapsed_time": "2h 19m 19s", "remaining_time": "3h 24m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.456398}
{"loss": 24.0, "grad_norm": 1.36874831, "learning_rate": 0.00067345, "token_acc": 0.38069476, "epoch": 0.40591675, "global_step/max_steps": "885/2181", "elapsed_time": "2h 19m 29s", "remaining_time": "3h 24m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.456516}
{"loss": 24.125, "grad_norm": 1.54927814, "learning_rate": 0.00067275, "token_acc": 0.38630137, "epoch": 0.40637542, "global_step/max_steps": "886/2181", "elapsed_time": "2h 19m 39s", "remaining_time": "3h 24m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.457132}
{"loss": 24.875, "grad_norm": 1.26016855, "learning_rate": 0.00067206, "token_acc": 0.38217479, "epoch": 0.40683408, "global_step/max_steps": "887/2181", "elapsed_time": "2h 19m 49s", "remaining_time": "3h 23m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.457565}
{"loss": 24.0, "grad_norm": 1.16958857, "learning_rate": 0.00067136, "token_acc": 0.38708776, "epoch": 0.40729274, "global_step/max_steps": "888/2181", "elapsed_time": "2h 19m 59s", "remaining_time": "3h 23m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.45834}
{"loss": 24.125, "grad_norm": 1.14641678, "learning_rate": 0.00067066, "token_acc": 0.37574553, "epoch": 0.4077514, "global_step/max_steps": "889/2181", "elapsed_time": "2h 20m 9s", "remaining_time": "3h 23m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.458626}
{"loss": 24.25, "grad_norm": 1.28392684, "learning_rate": 0.00066996, "token_acc": 0.38270209, "epoch": 0.40821007, "global_step/max_steps": "890/2181", "elapsed_time": "2h 20m 18s", "remaining_time": "3h 23m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.458695}
{"loss": 23.75, "grad_norm": 1.20410693, "learning_rate": 0.00066926, "token_acc": 0.39763346, "epoch": 0.40866873, "global_step/max_steps": "891/2181", "elapsed_time": "2h 20m 27s", "remaining_time": "3h 23m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.458359}
{"loss": 23.75, "grad_norm": 1.18618405, "learning_rate": 0.00066857, "token_acc": 0.38726717, "epoch": 0.40912739, "global_step/max_steps": "892/2181", "elapsed_time": "2h 20m 37s", "remaining_time": "3h 23m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.458191}
{"loss": 24.5, "grad_norm": 1.27556074, "learning_rate": 0.00066787, "token_acc": 0.3863074, "epoch": 0.40958606, "global_step/max_steps": "893/2181", "elapsed_time": "2h 20m 46s", "remaining_time": "3h 23m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.457963}
{"loss": 24.875, "grad_norm": 1.29510152, "learning_rate": 0.00066717, "token_acc": 0.3768773, "epoch": 0.41004472, "global_step/max_steps": "894/2181", "elapsed_time": "2h 20m 55s", "remaining_time": "3h 22m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.457979}
{"loss": 24.5, "grad_norm": 1.36657393, "learning_rate": 0.00066647, "token_acc": 0.36376365, "epoch": 0.41050338, "global_step/max_steps": "895/2181", "elapsed_time": "2h 21m 5s", "remaining_time": "3h 22m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.458003}
{"loss": 23.75, "grad_norm": 1.25814116, "learning_rate": 0.00066577, "token_acc": 0.3907448, "epoch": 0.41096205, "global_step/max_steps": "896/2181", "elapsed_time": "2h 21m 15s", "remaining_time": "3h 22m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.458546}
{"loss": 24.0, "grad_norm": 1.23808789, "learning_rate": 0.00066506, "token_acc": 0.38518316, "epoch": 0.41142071, "global_step/max_steps": "897/2181", "elapsed_time": "2h 21m 24s", "remaining_time": "3h 22m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.458318}
{"loss": 23.75, "grad_norm": 1.18211877, "learning_rate": 0.00066436, "token_acc": 0.38868664, "epoch": 0.41187937, "global_step/max_steps": "898/2181", "elapsed_time": "2h 21m 34s", "remaining_time": "3h 22m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.458433}
{"loss": 23.875, "grad_norm": 1.2962873, "learning_rate": 0.00066366, "token_acc": 0.37990196, "epoch": 0.41233803, "global_step/max_steps": "899/2181", "elapsed_time": "2h 21m 43s", "remaining_time": "3h 22m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.458538}
{"loss": 24.25, "grad_norm": 1.38339436, "learning_rate": 0.00066296, "token_acc": 0.3787234, "epoch": 0.4127967, "global_step/max_steps": "900/2181", "elapsed_time": "2h 21m 53s", "remaining_time": "3h 21m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.458786}
{"loss": 23.75, "grad_norm": 1.26020992, "learning_rate": 0.00066226, "token_acc": 0.39212469, "epoch": 0.41325536, "global_step/max_steps": "901/2181", "elapsed_time": "2h 22m 3s", "remaining_time": "3h 21m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.45945}
{"loss": 24.875, "grad_norm": 1.78459394, "learning_rate": 0.00066155, "token_acc": 0.37872573, "epoch": 0.41371402, "global_step/max_steps": "902/2181", "elapsed_time": "2h 22m 12s", "remaining_time": "3h 21m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.459174}
{"loss": 23.75, "grad_norm": 1.3940556, "learning_rate": 0.00066085, "token_acc": 0.39247005, "epoch": 0.41417269, "global_step/max_steps": "903/2181", "elapsed_time": "2h 22m 22s", "remaining_time": "3h 21m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.459675}
{"loss": 24.875, "grad_norm": 1.26461577, "learning_rate": 0.00066015, "token_acc": 0.38708799, "epoch": 0.41463135, "global_step/max_steps": "904/2181", "elapsed_time": "2h 22m 32s", "remaining_time": "3h 21m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.459647}
{"loss": 24.0, "grad_norm": 1.47335482, "learning_rate": 0.00065944, "token_acc": 0.37564416, "epoch": 0.41509001, "global_step/max_steps": "905/2181", "elapsed_time": "2h 22m 41s", "remaining_time": "3h 21m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.459803}
{"loss": 24.0, "grad_norm": 1.21115446, "learning_rate": 0.00065874, "token_acc": 0.38959602, "epoch": 0.41554868, "global_step/max_steps": "906/2181", "elapsed_time": "2h 22m 51s", "remaining_time": "3h 21m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.46051}
{"loss": 24.125, "grad_norm": 1.51210868, "learning_rate": 0.00065804, "token_acc": 0.38567275, "epoch": 0.41600734, "global_step/max_steps": "907/2181", "elapsed_time": "2h 23m 1s", "remaining_time": "3h 20m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.461387}
{"loss": 23.75, "grad_norm": 1.17326975, "learning_rate": 0.00065733, "token_acc": 0.38944724, "epoch": 0.416466, "global_step/max_steps": "908/2181", "elapsed_time": "2h 23m 11s", "remaining_time": "3h 20m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.461704}
{"loss": 24.375, "grad_norm": 1.31447136, "learning_rate": 0.00065663, "token_acc": 0.37959957, "epoch": 0.41692466, "global_step/max_steps": "909/2181", "elapsed_time": "2h 23m 21s", "remaining_time": "3h 20m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.461518}
{"loss": 24.125, "grad_norm": 1.24053717, "learning_rate": 0.00065592, "token_acc": 0.38258208, "epoch": 0.41738333, "global_step/max_steps": "910/2181", "elapsed_time": "2h 23m 30s", "remaining_time": "3h 20m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.461305}
{"loss": 23.875, "grad_norm": 1.13792574, "learning_rate": 0.00065521, "token_acc": 0.39805014, "epoch": 0.41784199, "global_step/max_steps": "911/2181", "elapsed_time": "2h 23m 40s", "remaining_time": "3h 20m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.461639}
{"loss": 24.5, "grad_norm": 1.45030427, "learning_rate": 0.00065451, "token_acc": 0.38104396, "epoch": 0.41830065, "global_step/max_steps": "912/2181", "elapsed_time": "2h 23m 49s", "remaining_time": "3h 20m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.461582}
{"loss": 23.75, "grad_norm": 1.17694986, "learning_rate": 0.0006538, "token_acc": 0.38408016, "epoch": 0.41875932, "global_step/max_steps": "913/2181", "elapsed_time": "2h 23m 59s", "remaining_time": "3h 19m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.461832}
{"loss": 24.875, "grad_norm": 1.23831904, "learning_rate": 0.0006531, "token_acc": 0.37472036, "epoch": 0.41921798, "global_step/max_steps": "914/2181", "elapsed_time": "2h 24m 9s", "remaining_time": "3h 19m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.462423}
{"loss": 24.625, "grad_norm": 1.26152122, "learning_rate": 0.00065239, "token_acc": 0.37858935, "epoch": 0.41967664, "global_step/max_steps": "915/2181", "elapsed_time": "2h 24m 18s", "remaining_time": "3h 19m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.462269}
{"loss": 23.875, "grad_norm": 1.21186113, "learning_rate": 0.00065168, "token_acc": 0.39401496, "epoch": 0.42013531, "global_step/max_steps": "916/2181", "elapsed_time": "2h 24m 28s", "remaining_time": "3h 19m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.462783}
{"loss": 24.75, "grad_norm": 1.18987787, "learning_rate": 0.00065097, "token_acc": 0.38774937, "epoch": 0.42059397, "global_step/max_steps": "917/2181", "elapsed_time": "2h 24m 38s", "remaining_time": "3h 19m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.462934}
{"loss": 23.5, "grad_norm": 1.17805541, "learning_rate": 0.00065026, "token_acc": 0.39494163, "epoch": 0.42105263, "global_step/max_steps": "918/2181", "elapsed_time": "2h 24m 48s", "remaining_time": "3h 19m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.463596}
{"loss": 24.75, "grad_norm": 1.41207457, "learning_rate": 0.00064956, "token_acc": 0.37444568, "epoch": 0.42151129, "global_step/max_steps": "919/2181", "elapsed_time": "2h 24m 57s", "remaining_time": "3h 19m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.463272}
{"loss": 24.5, "grad_norm": 1.5116508, "learning_rate": 0.00064885, "token_acc": 0.37197554, "epoch": 0.42196996, "global_step/max_steps": "920/2181", "elapsed_time": "2h 25m 6s", "remaining_time": "3h 18m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.463253}
{"loss": 23.875, "grad_norm": 1.23310411, "learning_rate": 0.00064814, "token_acc": 0.38728952, "epoch": 0.42242862, "global_step/max_steps": "921/2181", "elapsed_time": "2h 25m 16s", "remaining_time": "3h 18m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.463747}
{"loss": 24.125, "grad_norm": 1.33705962, "learning_rate": 0.00064743, "token_acc": 0.39640923, "epoch": 0.42288728, "global_step/max_steps": "922/2181", "elapsed_time": "2h 25m 26s", "remaining_time": "3h 18m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.464111}
{"loss": 23.5, "grad_norm": 1.2587657, "learning_rate": 0.00064672, "token_acc": 0.3956913, "epoch": 0.42334595, "global_step/max_steps": "923/2181", "elapsed_time": "2h 25m 36s", "remaining_time": "3h 18m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.464722}
{"loss": 24.75, "grad_norm": 1.14330983, "learning_rate": 0.00064601, "token_acc": 0.36916667, "epoch": 0.42380461, "global_step/max_steps": "924/2181", "elapsed_time": "2h 25m 45s", "remaining_time": "3h 18m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.464774}
{"loss": 23.875, "grad_norm": 1.38360548, "learning_rate": 0.0006453, "token_acc": 0.38171896, "epoch": 0.42426327, "global_step/max_steps": "925/2181", "elapsed_time": "2h 25m 55s", "remaining_time": "3h 18m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.464738}
{"loss": 24.25, "grad_norm": 1.20242965, "learning_rate": 0.00064459, "token_acc": 0.39270033, "epoch": 0.42472194, "global_step/max_steps": "926/2181", "elapsed_time": "2h 26m 4s", "remaining_time": "3h 17m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.464817}
{"loss": 23.75, "grad_norm": 1.26178432, "learning_rate": 0.00064388, "token_acc": 0.39671683, "epoch": 0.4251806, "global_step/max_steps": "927/2181", "elapsed_time": "2h 26m 14s", "remaining_time": "3h 17m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.465377}
{"loss": 24.875, "grad_norm": 1.25569677, "learning_rate": 0.00064316, "token_acc": 0.36389123, "epoch": 0.42563926, "global_step/max_steps": "928/2181", "elapsed_time": "2h 26m 24s", "remaining_time": "3h 17m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.465559}
{"loss": 24.875, "grad_norm": 1.20632458, "learning_rate": 0.00064245, "token_acc": 0.37406067, "epoch": 0.42609792, "global_step/max_steps": "929/2181", "elapsed_time": "2h 26m 33s", "remaining_time": "3h 17m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.465455}
{"loss": 24.5, "grad_norm": 1.22563934, "learning_rate": 0.00064174, "token_acc": 0.39038031, "epoch": 0.42655659, "global_step/max_steps": "930/2181", "elapsed_time": "2h 26m 43s", "remaining_time": "3h 17m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.465295}
{"loss": 24.625, "grad_norm": 1.16649103, "learning_rate": 0.00064103, "token_acc": 0.37127448, "epoch": 0.42701525, "global_step/max_steps": "931/2181", "elapsed_time": "2h 26m 53s", "remaining_time": "3h 17m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.465648}
{"loss": 24.25, "grad_norm": 1.12800491, "learning_rate": 0.00064032, "token_acc": 0.38517717, "epoch": 0.42747391, "global_step/max_steps": "932/2181", "elapsed_time": "2h 27m 2s", "remaining_time": "3h 17m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.465479}
{"loss": 24.125, "grad_norm": 1.17801034, "learning_rate": 0.0006396, "token_acc": 0.37447873, "epoch": 0.42793258, "global_step/max_steps": "933/2181", "elapsed_time": "2h 27m 12s", "remaining_time": "3h 16m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.465822}
{"loss": 24.625, "grad_norm": 1.30590069, "learning_rate": 0.00063889, "token_acc": 0.37175635, "epoch": 0.42839124, "global_step/max_steps": "934/2181", "elapsed_time": "2h 27m 22s", "remaining_time": "3h 16m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.466596}
{"loss": 23.875, "grad_norm": 1.47153294, "learning_rate": 0.00063818, "token_acc": 0.38315217, "epoch": 0.4288499, "global_step/max_steps": "935/2181", "elapsed_time": "2h 27m 30s", "remaining_time": "3h 16m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.465737}
{"loss": 24.625, "grad_norm": 1.17880082, "learning_rate": 0.00063746, "token_acc": 0.37832594, "epoch": 0.42930857, "global_step/max_steps": "936/2181", "elapsed_time": "2h 27m 40s", "remaining_time": "3h 16m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.465852}
{"loss": 24.625, "grad_norm": 1.20269787, "learning_rate": 0.00063675, "token_acc": 0.38085657, "epoch": 0.42976723, "global_step/max_steps": "937/2181", "elapsed_time": "2h 27m 51s", "remaining_time": "3h 16m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.466938}
{"loss": 23.375, "grad_norm": 1.42207074, "learning_rate": 0.00063603, "token_acc": 0.4, "epoch": 0.43022589, "global_step/max_steps": "938/2181", "elapsed_time": "2h 27m 60s", "remaining_time": "3h 16m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.466652}
{"loss": 24.125, "grad_norm": 1.33672774, "learning_rate": 0.00063532, "token_acc": 0.36598524, "epoch": 0.43068455, "global_step/max_steps": "939/2181", "elapsed_time": "2h 28m 9s", "remaining_time": "3h 15m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.466346}
{"loss": 25.0, "grad_norm": 1.342237, "learning_rate": 0.0006346, "token_acc": 0.35956618, "epoch": 0.43114322, "global_step/max_steps": "940/2181", "elapsed_time": "2h 28m 19s", "remaining_time": "3h 15m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.467305}
{"loss": 24.0, "grad_norm": 1.19482577, "learning_rate": 0.00063389, "token_acc": 0.38848139, "epoch": 0.43160188, "global_step/max_steps": "941/2181", "elapsed_time": "2h 28m 29s", "remaining_time": "3h 15m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.467277}
{"loss": 24.5, "grad_norm": 1.37551761, "learning_rate": 0.00063317, "token_acc": 0.38245906, "epoch": 0.43206054, "global_step/max_steps": "942/2181", "elapsed_time": "2h 28m 38s", "remaining_time": "3h 15m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.4676}
{"loss": 24.25, "grad_norm": 1.25284719, "learning_rate": 0.00063246, "token_acc": 0.38221709, "epoch": 0.43251921, "global_step/max_steps": "943/2181", "elapsed_time": "2h 28m 49s", "remaining_time": "3h 15m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.46843}
{"loss": 25.0, "grad_norm": 1.20927477, "learning_rate": 0.00063174, "token_acc": 0.37503514, "epoch": 0.43297787, "global_step/max_steps": "944/2181", "elapsed_time": "2h 28m 59s", "remaining_time": "3h 15m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.46943}
{"loss": 23.625, "grad_norm": 1.17289102, "learning_rate": 0.00063102, "token_acc": 0.37644676, "epoch": 0.43343653, "global_step/max_steps": "945/2181", "elapsed_time": "2h 29m 9s", "remaining_time": "3h 15m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.469381}
{"loss": 23.5, "grad_norm": 1.23288047, "learning_rate": 0.00063031, "token_acc": 0.39479575, "epoch": 0.4338952, "global_step/max_steps": "946/2181", "elapsed_time": "2h 29m 19s", "remaining_time": "3h 14m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.469973}
{"loss": 23.75, "grad_norm": 1.28673625, "learning_rate": 0.00062959, "token_acc": 0.39618995, "epoch": 0.43435386, "global_step/max_steps": "947/2181", "elapsed_time": "2h 29m 28s", "remaining_time": "3h 14m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.470428}
{"loss": 23.75, "grad_norm": 1.13250744, "learning_rate": 0.00062887, "token_acc": 0.38967922, "epoch": 0.43481252, "global_step/max_steps": "948/2181", "elapsed_time": "2h 29m 38s", "remaining_time": "3h 14m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.470002}
{"loss": 24.375, "grad_norm": 1.28483295, "learning_rate": 0.00062815, "token_acc": 0.38047138, "epoch": 0.43527118, "global_step/max_steps": "949/2181", "elapsed_time": "2h 29m 48s", "remaining_time": "3h 14m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.47087}
{"loss": 24.0, "grad_norm": 1.30351245, "learning_rate": 0.00062744, "token_acc": 0.37770193, "epoch": 0.43572985, "global_step/max_steps": "950/2181", "elapsed_time": "2h 29m 58s", "remaining_time": "3h 14m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.47137}
{"loss": 23.375, "grad_norm": 1.27505565, "learning_rate": 0.00062672, "token_acc": 0.37607573, "epoch": 0.43618851, "global_step/max_steps": "951/2181", "elapsed_time": "2h 30m 8s", "remaining_time": "3h 14m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.47186}
{"loss": 24.375, "grad_norm": 1.26531422, "learning_rate": 0.000626, "token_acc": 0.38154192, "epoch": 0.43664717, "global_step/max_steps": "952/2181", "elapsed_time": "2h 30m 17s", "remaining_time": "3h 14m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.471622}
{"loss": 24.625, "grad_norm": 1.19017303, "learning_rate": 0.00062528, "token_acc": 0.38505905, "epoch": 0.43710584, "global_step/max_steps": "953/2181", "elapsed_time": "2h 30m 27s", "remaining_time": "3h 13m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.4719}
{"loss": 23.375, "grad_norm": 1.25769019, "learning_rate": 0.00062456, "token_acc": 0.38466112, "epoch": 0.4375645, "global_step/max_steps": "954/2181", "elapsed_time": "2h 30m 36s", "remaining_time": "3h 13m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.471336}
{"loss": 24.375, "grad_norm": 1.21657133, "learning_rate": 0.00062384, "token_acc": 0.38626357, "epoch": 0.43802316, "global_step/max_steps": "955/2181", "elapsed_time": "2h 30m 46s", "remaining_time": "3h 13m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.472121}
{"loss": 24.375, "grad_norm": 1.18987477, "learning_rate": 0.00062312, "token_acc": 0.39357772, "epoch": 0.43848183, "global_step/max_steps": "956/2181", "elapsed_time": "2h 30m 56s", "remaining_time": "3h 13m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.472583}
{"loss": 23.875, "grad_norm": 1.3312459, "learning_rate": 0.0006224, "token_acc": 0.39386261, "epoch": 0.43894049, "global_step/max_steps": "957/2181", "elapsed_time": "2h 31m 5s", "remaining_time": "3h 13m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.472517}
{"loss": 24.0, "grad_norm": 1.20098317, "learning_rate": 0.00062168, "token_acc": 0.39446367, "epoch": 0.43939915, "global_step/max_steps": "958/2181", "elapsed_time": "2h 31m 15s", "remaining_time": "3h 13m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.472577}
{"loss": 23.875, "grad_norm": 1.17553163, "learning_rate": 0.00062096, "token_acc": 0.3753812, "epoch": 0.43985781, "global_step/max_steps": "959/2181", "elapsed_time": "2h 31m 25s", "remaining_time": "3h 12m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.473}
{"loss": 24.125, "grad_norm": 1.10719216, "learning_rate": 0.00062024, "token_acc": 0.37924208, "epoch": 0.44031648, "global_step/max_steps": "960/2181", "elapsed_time": "2h 31m 35s", "remaining_time": "3h 12m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.473628}
{"loss": 23.625, "grad_norm": 1.45763719, "learning_rate": 0.00061952, "token_acc": 0.40109589, "epoch": 0.44077514, "global_step/max_steps": "961/2181", "elapsed_time": "2h 31m 44s", "remaining_time": "3h 12m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.473366}
{"loss": 24.0, "grad_norm": 1.16410244, "learning_rate": 0.0006188, "token_acc": 0.39537507, "epoch": 0.4412338, "global_step/max_steps": "962/2181", "elapsed_time": "2h 31m 53s", "remaining_time": "3h 12m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.473386}
{"loss": 23.875, "grad_norm": 1.21795988, "learning_rate": 0.00061808, "token_acc": 0.38410042, "epoch": 0.44169247, "global_step/max_steps": "963/2181", "elapsed_time": "2h 32m 3s", "remaining_time": "3h 12m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.473576}
{"loss": 24.125, "grad_norm": 1.25407326, "learning_rate": 0.00061735, "token_acc": 0.37179487, "epoch": 0.44215113, "global_step/max_steps": "964/2181", "elapsed_time": "2h 32m 13s", "remaining_time": "3h 12m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.47367}
{"loss": 23.875, "grad_norm": 1.17024314, "learning_rate": 0.00061663, "token_acc": 0.38077475, "epoch": 0.44260979, "global_step/max_steps": "965/2181", "elapsed_time": "2h 32m 22s", "remaining_time": "3h 12m 0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.473696}
{"loss": 24.375, "grad_norm": 1.26292574, "learning_rate": 0.00061591, "token_acc": 0.37292092, "epoch": 0.44306846, "global_step/max_steps": "966/2181", "elapsed_time": "2h 32m 31s", "remaining_time": "3h 11m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.473261}
{"loss": 23.875, "grad_norm": 1.44677496, "learning_rate": 0.00061519, "token_acc": 0.38326118, "epoch": 0.44352712, "global_step/max_steps": "967/2181", "elapsed_time": "2h 32m 40s", "remaining_time": "3h 11m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.472551}
{"loss": 24.375, "grad_norm": 1.21989381, "learning_rate": 0.00061446, "token_acc": 0.36831849, "epoch": 0.44398578, "global_step/max_steps": "968/2181", "elapsed_time": "2h 32m 50s", "remaining_time": "3h 11m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.472801}
{"loss": 24.875, "grad_norm": 1.23942244, "learning_rate": 0.00061374, "token_acc": 0.38130949, "epoch": 0.44444444, "global_step/max_steps": "969/2181", "elapsed_time": "2h 32m 59s", "remaining_time": "3h 11m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.472754}
{"loss": 23.5, "grad_norm": 1.28729784, "learning_rate": 0.00061302, "token_acc": 0.39994387, "epoch": 0.44490311, "global_step/max_steps": "970/2181", "elapsed_time": "2h 33m 10s", "remaining_time": "3h 11m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.474153}
{"loss": 24.375, "grad_norm": 1.23510253, "learning_rate": 0.00061229, "token_acc": 0.38626126, "epoch": 0.44536177, "global_step/max_steps": "971/2181", "elapsed_time": "2h 33m 19s", "remaining_time": "3h 11m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.474071}
{"loss": 24.5, "grad_norm": 1.23400223, "learning_rate": 0.00061157, "token_acc": 0.37176274, "epoch": 0.44582043, "global_step/max_steps": "972/2181", "elapsed_time": "2h 33m 29s", "remaining_time": "3h 10m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.474257}
{"loss": 24.625, "grad_norm": 1.24106681, "learning_rate": 0.00061085, "token_acc": 0.37809473, "epoch": 0.4462791, "global_step/max_steps": "973/2181", "elapsed_time": "2h 33m 39s", "remaining_time": "3h 10m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.474745}
{"loss": 24.25, "grad_norm": 1.2989825, "learning_rate": 0.00061012, "token_acc": 0.38084313, "epoch": 0.44673776, "global_step/max_steps": "974/2181", "elapsed_time": "2h 33m 48s", "remaining_time": "3h 10m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.474648}
{"loss": 23.875, "grad_norm": 1.12695169, "learning_rate": 0.0006094, "token_acc": 0.38577994, "epoch": 0.44719642, "global_step/max_steps": "975/2181", "elapsed_time": "2h 33m 58s", "remaining_time": "3h 10m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.474784}
{"loss": 24.25, "grad_norm": 1.25731683, "learning_rate": 0.00060867, "token_acc": 0.38959212, "epoch": 0.44765509, "global_step/max_steps": "976/2181", "elapsed_time": "2h 34m 8s", "remaining_time": "3h 10m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.474956}
{"loss": 24.125, "grad_norm": 1.10764265, "learning_rate": 0.00060795, "token_acc": 0.39486229, "epoch": 0.44811375, "global_step/max_steps": "977/2181", "elapsed_time": "2h 34m 16s", "remaining_time": "3h 10m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.474126}
{"loss": 23.75, "grad_norm": 1.09924781, "learning_rate": 0.00060722, "token_acc": 0.40564719, "epoch": 0.44857241, "global_step/max_steps": "978/2181", "elapsed_time": "2h 34m 26s", "remaining_time": "3h 9m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.474107}
{"loss": 24.375, "grad_norm": 1.33472562, "learning_rate": 0.0006065, "token_acc": 0.38483736, "epoch": 0.44903107, "global_step/max_steps": "979/2181", "elapsed_time": "2h 34m 35s", "remaining_time": "3h 9m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.474444}
{"loss": 24.375, "grad_norm": 1.24491704, "learning_rate": 0.00060577, "token_acc": 0.37933009, "epoch": 0.44948974, "global_step/max_steps": "980/2181", "elapsed_time": "2h 34m 45s", "remaining_time": "3h 9m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.474954}
{"loss": 24.375, "grad_norm": 1.23400426, "learning_rate": 0.00060505, "token_acc": 0.37272225, "epoch": 0.4499484, "global_step/max_steps": "981/2181", "elapsed_time": "2h 34m 55s", "remaining_time": "3h 9m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.474554}
{"loss": 23.5, "grad_norm": 1.20206189, "learning_rate": 0.00060432, "token_acc": 0.3958986, "epoch": 0.45040706, "global_step/max_steps": "982/2181", "elapsed_time": "2h 35m 4s", "remaining_time": "3h 9m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.474576}
{"loss": 24.5, "grad_norm": 1.3043555, "learning_rate": 0.00060359, "token_acc": 0.37517395, "epoch": 0.45086573, "global_step/max_steps": "983/2181", "elapsed_time": "2h 35m 14s", "remaining_time": "3h 9m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.475068}
{"loss": 24.375, "grad_norm": 1.15127099, "learning_rate": 0.00060287, "token_acc": 0.37532355, "epoch": 0.45132439, "global_step/max_steps": "984/2181", "elapsed_time": "2h 35m 23s", "remaining_time": "3h 9m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.475044}
{"loss": 24.75, "grad_norm": 1.30198681, "learning_rate": 0.00060214, "token_acc": 0.35762332, "epoch": 0.45178305, "global_step/max_steps": "985/2181", "elapsed_time": "2h 35m 34s", "remaining_time": "3h 8m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.47571}
{"loss": 23.25, "grad_norm": 1.18803918, "learning_rate": 0.00060141, "token_acc": 0.39605884, "epoch": 0.45224172, "global_step/max_steps": "986/2181", "elapsed_time": "2h 35m 43s", "remaining_time": "3h 8m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.476013}
{"loss": 23.5, "grad_norm": 1.15429044, "learning_rate": 0.00060068, "token_acc": 0.394888, "epoch": 0.45270038, "global_step/max_steps": "987/2181", "elapsed_time": "2h 35m 52s", "remaining_time": "3h 8m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.475429}
{"loss": 24.625, "grad_norm": 1.10486329, "learning_rate": 0.00059996, "token_acc": 0.37378506, "epoch": 0.45315904, "global_step/max_steps": "988/2181", "elapsed_time": "2h 36m 2s", "remaining_time": "3h 8m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.475634}
{"loss": 24.5, "grad_norm": 1.29513144, "learning_rate": 0.00059923, "token_acc": 0.37125749, "epoch": 0.4536177, "global_step/max_steps": "989/2181", "elapsed_time": "2h 36m 12s", "remaining_time": "3h 8m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.476034}
{"loss": 24.25, "grad_norm": 1.16335833, "learning_rate": 0.0005985, "token_acc": 0.37576945, "epoch": 0.45407637, "global_step/max_steps": "990/2181", "elapsed_time": "2h 36m 21s", "remaining_time": "3h 8m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.476229}
{"loss": 24.5, "grad_norm": 1.18908584, "learning_rate": 0.00059777, "token_acc": 0.38981681, "epoch": 0.45453503, "global_step/max_steps": "991/2181", "elapsed_time": "2h 36m 31s", "remaining_time": "3h 7m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.476344}
{"loss": 24.0, "grad_norm": 1.43242908, "learning_rate": 0.00059704, "token_acc": 0.37712218, "epoch": 0.45499369, "global_step/max_steps": "992/2181", "elapsed_time": "2h 36m 41s", "remaining_time": "3h 7m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.476394}
{"loss": 23.75, "grad_norm": 1.10520589, "learning_rate": 0.00059632, "token_acc": 0.39564626, "epoch": 0.45545236, "global_step/max_steps": "993/2181", "elapsed_time": "2h 36m 50s", "remaining_time": "3h 7m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.476569}
{"loss": 24.125, "grad_norm": 1.31816709, "learning_rate": 0.00059559, "token_acc": 0.38547646, "epoch": 0.45591102, "global_step/max_steps": "994/2181", "elapsed_time": "2h 37m 1s", "remaining_time": "3h 7m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.478149}
{"loss": 24.875, "grad_norm": 1.24852705, "learning_rate": 0.00059486, "token_acc": 0.36988589, "epoch": 0.45636968, "global_step/max_steps": "995/2181", "elapsed_time": "2h 37m 10s", "remaining_time": "3h 7m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.477725}
{"loss": 23.75, "grad_norm": 1.44434488, "learning_rate": 0.00059413, "token_acc": 0.39493949, "epoch": 0.45682835, "global_step/max_steps": "996/2181", "elapsed_time": "2h 37m 20s", "remaining_time": "3h 7m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.477727}
{"loss": 24.125, "grad_norm": 1.23883867, "learning_rate": 0.0005934, "token_acc": 0.38512949, "epoch": 0.45728701, "global_step/max_steps": "997/2181", "elapsed_time": "2h 37m 29s", "remaining_time": "3h 7m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.47781}
{"loss": 24.125, "grad_norm": 1.24519861, "learning_rate": 0.00059267, "token_acc": 0.38204989, "epoch": 0.45774567, "global_step/max_steps": "998/2181", "elapsed_time": "2h 37m 39s", "remaining_time": "3h 6m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.477565}
{"loss": 24.375, "grad_norm": 1.27897096, "learning_rate": 0.00059194, "token_acc": 0.38144619, "epoch": 0.45820433, "global_step/max_steps": "999/2181", "elapsed_time": "2h 37m 49s", "remaining_time": "3h 6m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.47802}
{"loss": 24.375, "grad_norm": 1.08492231, "learning_rate": 0.00059121, "token_acc": 0.38478144, "epoch": 0.458663, "global_step/max_steps": "1000/2181", "elapsed_time": "2h 37m 58s", "remaining_time": "3h 6m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.477931}
{"loss": 23.875, "grad_norm": 1.2162956, "learning_rate": 0.00059048, "token_acc": 0.38815605, "epoch": 0.45912166, "global_step/max_steps": "1001/2181", "elapsed_time": "2h 38m 25s", "remaining_time": "3h 6m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.495912}
{"loss": 23.625, "grad_norm": 1.2941829, "learning_rate": 0.00058975, "token_acc": 0.39056964, "epoch": 0.45958032, "global_step/max_steps": "1002/2181", "elapsed_time": "2h 38m 35s", "remaining_time": "3h 6m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.496397}
{"loss": 24.0, "grad_norm": 1.1821506, "learning_rate": 0.00058902, "token_acc": 0.38014388, "epoch": 0.46003899, "global_step/max_steps": "1003/2181", "elapsed_time": "2h 38m 45s", "remaining_time": "3h 6m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.496377}
{"loss": 24.0, "grad_norm": 1.21583104, "learning_rate": 0.00058829, "token_acc": 0.38688801, "epoch": 0.46049765, "global_step/max_steps": "1004/2181", "elapsed_time": "2h 38m 54s", "remaining_time": "3h 6m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.49629}
{"loss": 24.125, "grad_norm": 1.1305995, "learning_rate": 0.00058756, "token_acc": 0.37192982, "epoch": 0.46095631, "global_step/max_steps": "1005/2181", "elapsed_time": "2h 39m 4s", "remaining_time": "3h 6m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.496454}
{"loss": 24.25, "grad_norm": 1.31277204, "learning_rate": 0.00058682, "token_acc": 0.37128995, "epoch": 0.46141498, "global_step/max_steps": "1006/2181", "elapsed_time": "2h 39m 13s", "remaining_time": "3h 5m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.496481}
{"loss": 24.5, "grad_norm": 1.1901114, "learning_rate": 0.00058609, "token_acc": 0.37907234, "epoch": 0.46187364, "global_step/max_steps": "1007/2181", "elapsed_time": "2h 39m 22s", "remaining_time": "3h 5m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.495651}
{"loss": 24.875, "grad_norm": 1.40960407, "learning_rate": 0.00058536, "token_acc": 0.38005618, "epoch": 0.4623323, "global_step/max_steps": "1008/2181", "elapsed_time": "2h 39m 31s", "remaining_time": "3h 5m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.495134}
{"loss": 24.5, "grad_norm": 1.231318, "learning_rate": 0.00058463, "token_acc": 0.37937909, "epoch": 0.46279096, "global_step/max_steps": "1009/2181", "elapsed_time": "2h 39m 41s", "remaining_time": "3h 5m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.495182}
{"loss": 24.0, "grad_norm": 1.15347242, "learning_rate": 0.0005839, "token_acc": 0.38572204, "epoch": 0.46324963, "global_step/max_steps": "1010/2181", "elapsed_time": "2h 39m 50s", "remaining_time": "3h 5m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.495013}
{"loss": 23.5, "grad_norm": 1.4564755, "learning_rate": 0.00058316, "token_acc": 0.37859379, "epoch": 0.46370829, "global_step/max_steps": "1011/2181", "elapsed_time": "2h 39m 60s", "remaining_time": "3h 5m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.495466}
{"loss": 23.875, "grad_norm": 1.29238558, "learning_rate": 0.00058243, "token_acc": 0.38259629, "epoch": 0.46416695, "global_step/max_steps": "1012/2181", "elapsed_time": "2h 40m 10s", "remaining_time": "3h 5m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.496308}
{"loss": 24.625, "grad_norm": 1.47578537, "learning_rate": 0.0005817, "token_acc": 0.38151724, "epoch": 0.46462562, "global_step/max_steps": "1013/2181", "elapsed_time": "2h 40m 20s", "remaining_time": "3h 4m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.49621}
{"loss": 24.125, "grad_norm": 1.30327737, "learning_rate": 0.00058097, "token_acc": 0.38352428, "epoch": 0.46508428, "global_step/max_steps": "1014/2181", "elapsed_time": "2h 40m 28s", "remaining_time": "3h 4m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.495189}
{"loss": 24.125, "grad_norm": 1.32165754, "learning_rate": 0.00058023, "token_acc": 0.38365304, "epoch": 0.46554294, "global_step/max_steps": "1015/2181", "elapsed_time": "2h 40m 38s", "remaining_time": "3h 4m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.495712}
{"loss": 24.5, "grad_norm": 1.37804222, "learning_rate": 0.0005795, "token_acc": 0.3758069, "epoch": 0.46600161, "global_step/max_steps": "1016/2181", "elapsed_time": "2h 40m 48s", "remaining_time": "3h 4m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.495764}
{"loss": 23.875, "grad_norm": 1.14412487, "learning_rate": 0.00057877, "token_acc": 0.38302503, "epoch": 0.46646027, "global_step/max_steps": "1017/2181", "elapsed_time": "2h 40m 56s", "remaining_time": "3h 4m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.494963}
{"loss": 23.75, "grad_norm": 1.20930314, "learning_rate": 0.00057803, "token_acc": 0.39134055, "epoch": 0.46691893, "global_step/max_steps": "1018/2181", "elapsed_time": "2h 41m 6s", "remaining_time": "3h 4m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.494969}
{"loss": 23.875, "grad_norm": 1.16091597, "learning_rate": 0.0005773, "token_acc": 0.38483067, "epoch": 0.46737759, "global_step/max_steps": "1019/2181", "elapsed_time": "2h 41m 15s", "remaining_time": "3h 3m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.494884}
{"loss": 23.875, "grad_norm": 1.21508944, "learning_rate": 0.00057657, "token_acc": 0.40032371, "epoch": 0.46783626, "global_step/max_steps": "1020/2181", "elapsed_time": "2h 41m 25s", "remaining_time": "3h 3m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.495092}
{"loss": 23.625, "grad_norm": 1.14990532, "learning_rate": 0.00057583, "token_acc": 0.38907104, "epoch": 0.46829492, "global_step/max_steps": "1021/2181", "elapsed_time": "2h 41m 35s", "remaining_time": "3h 3m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.495342}
{"loss": 24.25, "grad_norm": 1.15826786, "learning_rate": 0.0005751, "token_acc": 0.38648572, "epoch": 0.46875358, "global_step/max_steps": "1022/2181", "elapsed_time": "2h 41m 45s", "remaining_time": "3h 3m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.495763}
{"loss": 24.0, "grad_norm": 1.26403415, "learning_rate": 0.00057436, "token_acc": 0.3885126, "epoch": 0.46921225, "global_step/max_steps": "1023/2181", "elapsed_time": "2h 41m 54s", "remaining_time": "3h 3m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.495365}
{"loss": 24.625, "grad_norm": 1.112342, "learning_rate": 0.00057363, "token_acc": 0.37788018, "epoch": 0.46967091, "global_step/max_steps": "1024/2181", "elapsed_time": "2h 42m 3s", "remaining_time": "3h 3m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.49517}
{"loss": 24.125, "grad_norm": 1.36179471, "learning_rate": 0.00057289, "token_acc": 0.37289562, "epoch": 0.47012957, "global_step/max_steps": "1025/2181", "elapsed_time": "2h 42m 13s", "remaining_time": "3h 2m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.495391}
{"loss": 24.25, "grad_norm": 1.127702, "learning_rate": 0.00057216, "token_acc": 0.37028886, "epoch": 0.47058824, "global_step/max_steps": "1026/2181", "elapsed_time": "2h 42m 22s", "remaining_time": "3h 2m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.495422}
{"loss": 23.75, "grad_norm": 1.21386766, "learning_rate": 0.00057142, "token_acc": 0.38677065, "epoch": 0.4710469, "global_step/max_steps": "1027/2181", "elapsed_time": "2h 42m 32s", "remaining_time": "3h 2m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.496062}
{"loss": 23.625, "grad_norm": 1.20573854, "learning_rate": 0.00057069, "token_acc": 0.4, "epoch": 0.47150556, "global_step/max_steps": "1028/2181", "elapsed_time": "2h 42m 42s", "remaining_time": "3h 2m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.496025}
{"loss": 23.5, "grad_norm": 1.19477069, "learning_rate": 0.00056995, "token_acc": 0.3828169, "epoch": 0.47196422, "global_step/max_steps": "1029/2181", "elapsed_time": "2h 42m 52s", "remaining_time": "3h 2m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.496872}
{"loss": 23.625, "grad_norm": 1.25249445, "learning_rate": 0.00056922, "token_acc": 0.38376384, "epoch": 0.47242289, "global_step/max_steps": "1030/2181", "elapsed_time": "2h 43m 2s", "remaining_time": "3h 2m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.496809}
{"loss": 24.125, "grad_norm": 1.29487431, "learning_rate": 0.00056848, "token_acc": 0.37734807, "epoch": 0.47288155, "global_step/max_steps": "1031/2181", "elapsed_time": "2h 43m 11s", "remaining_time": "3h 2m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.49661}
{"loss": 24.0, "grad_norm": 1.21313429, "learning_rate": 0.00056775, "token_acc": 0.38797662, "epoch": 0.47334021, "global_step/max_steps": "1032/2181", "elapsed_time": "2h 43m 21s", "remaining_time": "3h 1m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.497281}
{"loss": 23.25, "grad_norm": 1.21329474, "learning_rate": 0.00056701, "token_acc": 0.4017192, "epoch": 0.47379888, "global_step/max_steps": "1033/2181", "elapsed_time": "2h 43m 30s", "remaining_time": "3h 1m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.49656}
{"loss": 23.625, "grad_norm": 1.2196418, "learning_rate": 0.00056628, "token_acc": 0.38383251, "epoch": 0.47425754, "global_step/max_steps": "1034/2181", "elapsed_time": "2h 43m 39s", "remaining_time": "3h 1m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.496251}
{"loss": 24.875, "grad_norm": 1.18282294, "learning_rate": 0.00056554, "token_acc": 0.36119568, "epoch": 0.4747162, "global_step/max_steps": "1035/2181", "elapsed_time": "2h 43m 49s", "remaining_time": "3h 1m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.4966}
{"loss": 23.5, "grad_norm": 1.40302002, "learning_rate": 0.0005648, "token_acc": 0.39001122, "epoch": 0.47517487, "global_step/max_steps": "1036/2181", "elapsed_time": "2h 43m 58s", "remaining_time": "3h 1m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.49646}
{"loss": 24.25, "grad_norm": 1.22121811, "learning_rate": 0.00056407, "token_acc": 0.38158265, "epoch": 0.47563353, "global_step/max_steps": "1037/2181", "elapsed_time": "2h 44m 8s", "remaining_time": "3h 1m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.496937}
{"loss": 24.5, "grad_norm": 1.17290902, "learning_rate": 0.00056333, "token_acc": 0.36998886, "epoch": 0.47609219, "global_step/max_steps": "1038/2181", "elapsed_time": "2h 44m 18s", "remaining_time": "3h 0m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.497509}
{"loss": 23.875, "grad_norm": 1.40294838, "learning_rate": 0.00056259, "token_acc": 0.39611872, "epoch": 0.47655085, "global_step/max_steps": "1039/2181", "elapsed_time": "2h 44m 28s", "remaining_time": "3h 0m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.497764}
{"loss": 23.25, "grad_norm": 1.43494749, "learning_rate": 0.00056186, "token_acc": 0.40311005, "epoch": 0.47700952, "global_step/max_steps": "1040/2181", "elapsed_time": "2h 44m 37s", "remaining_time": "3h 0m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.497566}
{"loss": 24.0, "grad_norm": 1.33089292, "learning_rate": 0.00056112, "token_acc": 0.39815591, "epoch": 0.47746818, "global_step/max_steps": "1041/2181", "elapsed_time": "2h 44m 47s", "remaining_time": "3h 0m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.497593}
{"loss": 23.375, "grad_norm": 1.46804535, "learning_rate": 0.00056038, "token_acc": 0.41518625, "epoch": 0.47792684, "global_step/max_steps": "1042/2181", "elapsed_time": "2h 44m 57s", "remaining_time": "3h 0m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.497816}
{"loss": 24.125, "grad_norm": 1.17176938, "learning_rate": 0.00055964, "token_acc": 0.37412791, "epoch": 0.47838551, "global_step/max_steps": "1043/2181", "elapsed_time": "2h 45m 6s", "remaining_time": "3h 0m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.49793}
{"loss": 24.5, "grad_norm": 1.56255114, "learning_rate": 0.00055891, "token_acc": 0.38382148, "epoch": 0.47884417, "global_step/max_steps": "1044/2181", "elapsed_time": "2h 45m 16s", "remaining_time": "2h 59m 60s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.498286}
{"loss": 23.5, "grad_norm": 1.3379916, "learning_rate": 0.00055817, "token_acc": 0.395218, "epoch": 0.47930283, "global_step/max_steps": "1045/2181", "elapsed_time": "2h 45m 26s", "remaining_time": "2h 59m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.498288}
{"loss": 24.125, "grad_norm": 1.2107594, "learning_rate": 0.00055743, "token_acc": 0.39428571, "epoch": 0.4797615, "global_step/max_steps": "1046/2181", "elapsed_time": "2h 45m 35s", "remaining_time": "2h 59m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.497973}
{"loss": 23.875, "grad_norm": 1.64887118, "learning_rate": 0.00055669, "token_acc": 0.39377239, "epoch": 0.48022016, "global_step/max_steps": "1047/2181", "elapsed_time": "2h 45m 45s", "remaining_time": "2h 59m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.498427}
{"loss": 25.25, "grad_norm": 1.29503191, "learning_rate": 0.00055596, "token_acc": 0.36268993, "epoch": 0.48067882, "global_step/max_steps": "1048/2181", "elapsed_time": "2h 45m 56s", "remaining_time": "2h 59m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.499625}
{"loss": 24.625, "grad_norm": 1.19203806, "learning_rate": 0.00055522, "token_acc": 0.36887115, "epoch": 0.48113748, "global_step/max_steps": "1049/2181", "elapsed_time": "2h 46m 5s", "remaining_time": "2h 59m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.499243}
{"loss": 24.5, "grad_norm": 1.37714779, "learning_rate": 0.00055448, "token_acc": 0.3729921, "epoch": 0.48159615, "global_step/max_steps": "1050/2181", "elapsed_time": "2h 46m 14s", "remaining_time": "2h 59m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.499522}
{"loss": 24.75, "grad_norm": 1.20720398, "learning_rate": 0.00055374, "token_acc": 0.37003913, "epoch": 0.48205481, "global_step/max_steps": "1051/2181", "elapsed_time": "2h 46m 24s", "remaining_time": "2h 58m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.499372}
{"loss": 23.75, "grad_norm": 1.28442013, "learning_rate": 0.000553, "token_acc": 0.40468037, "epoch": 0.48251347, "global_step/max_steps": "1052/2181", "elapsed_time": "2h 46m 33s", "remaining_time": "2h 58m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.498715}
{"loss": 24.125, "grad_norm": 1.40405011, "learning_rate": 0.00055226, "token_acc": 0.40865655, "epoch": 0.48297214, "global_step/max_steps": "1053/2181", "elapsed_time": "2h 46m 41s", "remaining_time": "2h 58m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.49798}
{"loss": 24.625, "grad_norm": 1.14009452, "learning_rate": 0.00055153, "token_acc": 0.37987789, "epoch": 0.4834308, "global_step/max_steps": "1054/2181", "elapsed_time": "2h 46m 52s", "remaining_time": "2h 58m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.498869}
{"loss": 24.125, "grad_norm": 1.19232881, "learning_rate": 0.00055079, "token_acc": 0.36763897, "epoch": 0.48388946, "global_step/max_steps": "1055/2181", "elapsed_time": "2h 47m 2s", "remaining_time": "2h 58m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.49942}
{"loss": 25.0, "grad_norm": 1.18288481, "learning_rate": 0.00055005, "token_acc": 0.35310655, "epoch": 0.48434813, "global_step/max_steps": "1056/2181", "elapsed_time": "2h 47m 11s", "remaining_time": "2h 58m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.498728}
{"loss": 24.25, "grad_norm": 1.29696238, "learning_rate": 0.00054931, "token_acc": 0.3869816, "epoch": 0.48480679, "global_step/max_steps": "1057/2181", "elapsed_time": "2h 47m 21s", "remaining_time": "2h 57m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.499562}
{"loss": 23.375, "grad_norm": 1.16387653, "learning_rate": 0.00054857, "token_acc": 0.40512541, "epoch": 0.48526545, "global_step/max_steps": "1058/2181", "elapsed_time": "2h 47m 31s", "remaining_time": "2h 57m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.499708}
{"loss": 24.25, "grad_norm": 1.13543785, "learning_rate": 0.00054783, "token_acc": 0.37817259, "epoch": 0.48572411, "global_step/max_steps": "1059/2181", "elapsed_time": "2h 47m 40s", "remaining_time": "2h 57m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.499359}
{"loss": 24.875, "grad_norm": 1.24539351, "learning_rate": 0.00054709, "token_acc": 0.36010499, "epoch": 0.48618278, "global_step/max_steps": "1060/2181", "elapsed_time": "2h 47m 50s", "remaining_time": "2h 57m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.50045}
{"loss": 24.125, "grad_norm": 1.25860143, "learning_rate": 0.00054635, "token_acc": 0.37025496, "epoch": 0.48664144, "global_step/max_steps": "1061/2181", "elapsed_time": "2h 48m 1s", "remaining_time": "2h 57m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.501103}
{"loss": 24.75, "grad_norm": 1.14863145, "learning_rate": 0.00054561, "token_acc": 0.37739412, "epoch": 0.4871001, "global_step/max_steps": "1062/2181", "elapsed_time": "2h 48m 10s", "remaining_time": "2h 57m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.501085}
{"loss": 24.375, "grad_norm": 1.22663248, "learning_rate": 0.00054487, "token_acc": 0.3952232, "epoch": 0.48755877, "global_step/max_steps": "1063/2181", "elapsed_time": "2h 48m 20s", "remaining_time": "2h 57m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.500974}
{"loss": 24.5, "grad_norm": 1.20335233, "learning_rate": 0.00054413, "token_acc": 0.38179348, "epoch": 0.48801743, "global_step/max_steps": "1064/2181", "elapsed_time": "2h 48m 30s", "remaining_time": "2h 56m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.501423}
{"loss": 23.75, "grad_norm": 1.36318004, "learning_rate": 0.00054339, "token_acc": 0.39326153, "epoch": 0.48847609, "global_step/max_steps": "1065/2181", "elapsed_time": "2h 48m 39s", "remaining_time": "2h 56m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.5015}
{"loss": 24.5, "grad_norm": 1.18820322, "learning_rate": 0.00054265, "token_acc": 0.36852533, "epoch": 0.48893476, "global_step/max_steps": "1066/2181", "elapsed_time": "2h 48m 48s", "remaining_time": "2h 56m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.501119}
{"loss": 24.125, "grad_norm": 1.32193005, "learning_rate": 0.00054191, "token_acc": 0.38039543, "epoch": 0.48939342, "global_step/max_steps": "1067/2181", "elapsed_time": "2h 48m 57s", "remaining_time": "2h 56m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.500821}
{"loss": 23.625, "grad_norm": 1.26083207, "learning_rate": 0.00054117, "token_acc": 0.40011631, "epoch": 0.48985208, "global_step/max_steps": "1068/2181", "elapsed_time": "2h 49m 8s", "remaining_time": "2h 56m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.501845}
{"loss": 24.125, "grad_norm": 1.20264161, "learning_rate": 0.00054043, "token_acc": 0.38575419, "epoch": 0.49031074, "global_step/max_steps": "1069/2181", "elapsed_time": "2h 49m 17s", "remaining_time": "2h 56m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.501806}
{"loss": 25.125, "grad_norm": 1.20805216, "learning_rate": 0.00053969, "token_acc": 0.36228995, "epoch": 0.49076941, "global_step/max_steps": "1070/2181", "elapsed_time": "2h 49m 27s", "remaining_time": "2h 55m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.501718}
{"loss": 23.375, "grad_norm": 1.21694005, "learning_rate": 0.00053895, "token_acc": 0.39006313, "epoch": 0.49122807, "global_step/max_steps": "1071/2181", "elapsed_time": "2h 49m 37s", "remaining_time": "2h 55m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.501987}
{"loss": 24.0, "grad_norm": 1.00168133, "learning_rate": 0.00053821, "token_acc": 0.38320569, "epoch": 0.49168673, "global_step/max_steps": "1072/2181", "elapsed_time": "2h 49m 45s", "remaining_time": "2h 55m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.50135}
{"loss": 24.625, "grad_norm": 1.22354352, "learning_rate": 0.00053747, "token_acc": 0.36969697, "epoch": 0.4921454, "global_step/max_steps": "1073/2181", "elapsed_time": "2h 49m 55s", "remaining_time": "2h 55m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.501449}
{"loss": 23.875, "grad_norm": 1.13473427, "learning_rate": 0.00053673, "token_acc": 0.38057016, "epoch": 0.49260406, "global_step/max_steps": "1074/2181", "elapsed_time": "2h 50m 5s", "remaining_time": "2h 55m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.501849}
{"loss": 24.125, "grad_norm": 1.07082021, "learning_rate": 0.00053599, "token_acc": 0.38406396, "epoch": 0.49306272, "global_step/max_steps": "1075/2181", "elapsed_time": "2h 50m 14s", "remaining_time": "2h 55m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.501596}
{"loss": 23.75, "grad_norm": 1.14566755, "learning_rate": 0.00053525, "token_acc": 0.40163703, "epoch": 0.49352139, "global_step/max_steps": "1076/2181", "elapsed_time": "2h 50m 25s", "remaining_time": "2h 55m 0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.502556}
{"loss": 24.25, "grad_norm": 1.16276872, "learning_rate": 0.00053451, "token_acc": 0.39302519, "epoch": 0.49398005, "global_step/max_steps": "1077/2181", "elapsed_time": "2h 50m 35s", "remaining_time": "2h 54m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.502958}
{"loss": 24.25, "grad_norm": 1.293625, "learning_rate": 0.00053377, "token_acc": 0.39546803, "epoch": 0.49443871, "global_step/max_steps": "1078/2181", "elapsed_time": "2h 50m 44s", "remaining_time": "2h 54m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.502828}
{"loss": 23.875, "grad_norm": 1.19878471, "learning_rate": 0.00053303, "token_acc": 0.39679415, "epoch": 0.49489737, "global_step/max_steps": "1079/2181", "elapsed_time": "2h 50m 54s", "remaining_time": "2h 54m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.502812}
{"loss": 24.75, "grad_norm": 1.14829469, "learning_rate": 0.00053228, "token_acc": 0.36419069, "epoch": 0.49535604, "global_step/max_steps": "1080/2181", "elapsed_time": "2h 51m 3s", "remaining_time": "2h 54m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.502953}
{"loss": 23.625, "grad_norm": 1.48171997, "learning_rate": 0.00053154, "token_acc": 0.3785073, "epoch": 0.4958147, "global_step/max_steps": "1081/2181", "elapsed_time": "2h 51m 13s", "remaining_time": "2h 54m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.503489}
{"loss": 24.125, "grad_norm": 1.16462219, "learning_rate": 0.0005308, "token_acc": 0.37290003, "epoch": 0.49627336, "global_step/max_steps": "1082/2181", "elapsed_time": "2h 51m 22s", "remaining_time": "2h 54m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.503109}
{"loss": 23.75, "grad_norm": 1.1365602, "learning_rate": 0.00053006, "token_acc": 0.38751045, "epoch": 0.49673203, "global_step/max_steps": "1083/2181", "elapsed_time": "2h 51m 32s", "remaining_time": "2h 53m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.503401}
{"loss": 24.125, "grad_norm": 1.09397686, "learning_rate": 0.00052932, "token_acc": 0.3957935, "epoch": 0.49719069, "global_step/max_steps": "1084/2181", "elapsed_time": "2h 51m 42s", "remaining_time": "2h 53m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.503711}
{"loss": 23.5, "grad_norm": 1.28354704, "learning_rate": 0.00052858, "token_acc": 0.39905973, "epoch": 0.49764935, "global_step/max_steps": "1085/2181", "elapsed_time": "2h 51m 51s", "remaining_time": "2h 53m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.503547}
{"loss": 24.5, "grad_norm": 1.32593489, "learning_rate": 0.00052784, "token_acc": 0.38169399, "epoch": 0.49810802, "global_step/max_steps": "1086/2181", "elapsed_time": "2h 52m 1s", "remaining_time": "2h 53m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.503511}
{"loss": 24.875, "grad_norm": 1.22278714, "learning_rate": 0.0005271, "token_acc": 0.37643521, "epoch": 0.49856668, "global_step/max_steps": "1087/2181", "elapsed_time": "2h 52m 10s", "remaining_time": "2h 53m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.503615}
{"loss": 24.25, "grad_norm": 1.23031723, "learning_rate": 0.00052635, "token_acc": 0.38619351, "epoch": 0.49902534, "global_step/max_steps": "1088/2181", "elapsed_time": "2h 52m 20s", "remaining_time": "2h 53m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.503251}
{"loss": 24.75, "grad_norm": 1.27074158, "learning_rate": 0.00052561, "token_acc": 0.37381616, "epoch": 0.499484, "global_step/max_steps": "1089/2181", "elapsed_time": "2h 52m 29s", "remaining_time": "2h 52m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.503228}
{"loss": 23.5, "grad_norm": 1.28375542, "learning_rate": 0.00052487, "token_acc": 0.38887328, "epoch": 0.49994267, "global_step/max_steps": "1090/2181", "elapsed_time": "2h 52m 39s", "remaining_time": "2h 52m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.504128}
{"loss": 24.25, "grad_norm": 1.18300557, "learning_rate": 0.00052413, "token_acc": 0.37132753, "epoch": 0.50040133, "global_step/max_steps": "1091/2181", "elapsed_time": "2h 52m 49s", "remaining_time": "2h 52m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.504264}
{"loss": 24.375, "grad_norm": 1.22480977, "learning_rate": 0.00052339, "token_acc": 0.38282139, "epoch": 0.50085999, "global_step/max_steps": "1092/2181", "elapsed_time": "2h 52m 59s", "remaining_time": "2h 52m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.504637}
{"loss": 24.0, "grad_norm": 1.29374087, "learning_rate": 0.00052264, "token_acc": 0.38709677, "epoch": 0.50131866, "global_step/max_steps": "1093/2181", "elapsed_time": "2h 53m 10s", "remaining_time": "2h 52m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.505687}
{"loss": 24.0, "grad_norm": 1.20738602, "learning_rate": 0.0005219, "token_acc": 0.39001122, "epoch": 0.50177732, "global_step/max_steps": "1094/2181", "elapsed_time": "2h 53m 19s", "remaining_time": "2h 52m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.505649}
{"loss": 24.125, "grad_norm": 1.16622376, "learning_rate": 0.00052116, "token_acc": 0.38858397, "epoch": 0.50223598, "global_step/max_steps": "1095/2181", "elapsed_time": "2h 53m 28s", "remaining_time": "2h 52m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.505079}
{"loss": 23.125, "grad_norm": 1.15163767, "learning_rate": 0.00052042, "token_acc": 0.39497459, "epoch": 0.50269465, "global_step/max_steps": "1096/2181", "elapsed_time": "2h 53m 38s", "remaining_time": "2h 51m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.505639}
{"loss": 24.75, "grad_norm": 1.17078972, "learning_rate": 0.00051968, "token_acc": 0.38498313, "epoch": 0.50315331, "global_step/max_steps": "1097/2181", "elapsed_time": "2h 53m 48s", "remaining_time": "2h 51m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.50629}
{"loss": 23.875, "grad_norm": 1.25212657, "learning_rate": 0.00051893, "token_acc": 0.37296691, "epoch": 0.50361197, "global_step/max_steps": "1098/2181", "elapsed_time": "2h 53m 58s", "remaining_time": "2h 51m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.50593}
{"loss": 23.25, "grad_norm": 1.24752295, "learning_rate": 0.00051819, "token_acc": 0.39659901, "epoch": 0.50407063, "global_step/max_steps": "1099/2181", "elapsed_time": "2h 54m 7s", "remaining_time": "2h 51m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.505726}
{"loss": 24.25, "grad_norm": 1.14061224, "learning_rate": 0.00051745, "token_acc": 0.37642164, "epoch": 0.5045293, "global_step/max_steps": "1100/2181", "elapsed_time": "2h 54m 17s", "remaining_time": "2h 51m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.50611}
{"loss": 24.0, "grad_norm": 1.30817592, "learning_rate": 0.00051671, "token_acc": 0.3812825, "epoch": 0.50498796, "global_step/max_steps": "1101/2181", "elapsed_time": "2h 54m 26s", "remaining_time": "2h 51m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.505966}
{"loss": 24.25, "grad_norm": 1.09802186, "learning_rate": 0.00051597, "token_acc": 0.37364382, "epoch": 0.50544662, "global_step/max_steps": "1102/2181", "elapsed_time": "2h 54m 36s", "remaining_time": "2h 50m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.506604}
{"loss": 25.0, "grad_norm": 1.20557797, "learning_rate": 0.00051522, "token_acc": 0.36660269, "epoch": 0.50590529, "global_step/max_steps": "1103/2181", "elapsed_time": "2h 54m 46s", "remaining_time": "2h 50m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.506719}
{"loss": 24.5, "grad_norm": 1.19471037, "learning_rate": 0.00051448, "token_acc": 0.3740871, "epoch": 0.50636395, "global_step/max_steps": "1104/2181", "elapsed_time": "2h 54m 55s", "remaining_time": "2h 50m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.50658}
{"loss": 23.375, "grad_norm": 1.21978915, "learning_rate": 0.00051374, "token_acc": 0.39939693, "epoch": 0.50682261, "global_step/max_steps": "1105/2181", "elapsed_time": "2h 55m 5s", "remaining_time": "2h 50m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.506948}
{"loss": 23.25, "grad_norm": 1.06997228, "learning_rate": 0.000513, "token_acc": 0.40815775, "epoch": 0.50728128, "global_step/max_steps": "1106/2181", "elapsed_time": "2h 55m 14s", "remaining_time": "2h 50m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.506722}
{"loss": 24.25, "grad_norm": 1.29229617, "learning_rate": 0.00051225, "token_acc": 0.38738208, "epoch": 0.50773994, "global_step/max_steps": "1107/2181", "elapsed_time": "2h 55m 24s", "remaining_time": "2h 50m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.506974}
{"loss": 24.125, "grad_norm": 1.26257455, "learning_rate": 0.00051151, "token_acc": 0.38446272, "epoch": 0.5081986, "global_step/max_steps": "1108/2181", "elapsed_time": "2h 55m 33s", "remaining_time": "2h 50m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.506554}
{"loss": 24.0, "grad_norm": 1.38039505, "learning_rate": 0.00051077, "token_acc": 0.39188795, "epoch": 0.50865726, "global_step/max_steps": "1109/2181", "elapsed_time": "2h 55m 44s", "remaining_time": "2h 49m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.507578}
{"loss": 24.125, "grad_norm": 1.21609938, "learning_rate": 0.00051003, "token_acc": 0.3806114, "epoch": 0.50911593, "global_step/max_steps": "1110/2181", "elapsed_time": "2h 55m 53s", "remaining_time": "2h 49m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.507488}
{"loss": 24.875, "grad_norm": 1.28853786, "learning_rate": 0.00050928, "token_acc": 0.36881689, "epoch": 0.50957459, "global_step/max_steps": "1111/2181", "elapsed_time": "2h 56m 3s", "remaining_time": "2h 49m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.507326}
{"loss": 24.25, "grad_norm": 1.1431669, "learning_rate": 0.00050854, "token_acc": 0.37319244, "epoch": 0.51003325, "global_step/max_steps": "1112/2181", "elapsed_time": "2h 56m 12s", "remaining_time": "2h 49m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.507602}
{"loss": 24.625, "grad_norm": 1.26352143, "learning_rate": 0.0005078, "token_acc": 0.38335129, "epoch": 0.51049192, "global_step/max_steps": "1113/2181", "elapsed_time": "2h 56m 22s", "remaining_time": "2h 49m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.507785}
{"loss": 24.125, "grad_norm": 1.18770039, "learning_rate": 0.00050706, "token_acc": 0.37464868, "epoch": 0.51095058, "global_step/max_steps": "1114/2181", "elapsed_time": "2h 56m 31s", "remaining_time": "2h 49m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.507162}
{"loss": 24.125, "grad_norm": 1.18980587, "learning_rate": 0.00050631, "token_acc": 0.38451087, "epoch": 0.51140924, "global_step/max_steps": "1115/2181", "elapsed_time": "2h 56m 41s", "remaining_time": "2h 48m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.507624}
{"loss": 23.5, "grad_norm": 1.25758803, "learning_rate": 0.00050557, "token_acc": 0.39526015, "epoch": 0.51186791, "global_step/max_steps": "1116/2181", "elapsed_time": "2h 56m 51s", "remaining_time": "2h 48m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.508024}
{"loss": 23.875, "grad_norm": 1.25977898, "learning_rate": 0.00050483, "token_acc": 0.38676149, "epoch": 0.51232657, "global_step/max_steps": "1117/2181", "elapsed_time": "2h 57m 0s", "remaining_time": "2h 48m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.507856}
{"loss": 23.75, "grad_norm": 1.20722961, "learning_rate": 0.00050408, "token_acc": 0.39246425, "epoch": 0.51278523, "global_step/max_steps": "1118/2181", "elapsed_time": "2h 57m 11s", "remaining_time": "2h 48m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.50864}
{"loss": 23.5, "grad_norm": 1.09641445, "learning_rate": 0.00050334, "token_acc": 0.3906383, "epoch": 0.51324389, "global_step/max_steps": "1119/2181", "elapsed_time": "2h 57m 19s", "remaining_time": "2h 48m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.507996}
{"loss": 23.75, "grad_norm": 1.30035126, "learning_rate": 0.0005026, "token_acc": 0.38868966, "epoch": 0.51370256, "global_step/max_steps": "1120/2181", "elapsed_time": "2h 57m 29s", "remaining_time": "2h 48m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.50832}
{"loss": 23.75, "grad_norm": 1.24825346, "learning_rate": 0.00050186, "token_acc": 0.39302783, "epoch": 0.51416122, "global_step/max_steps": "1121/2181", "elapsed_time": "2h 57m 39s", "remaining_time": "2h 47m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.508149}
{"loss": 25.25, "grad_norm": 1.31522274, "learning_rate": 0.00050111, "token_acc": 0.35837823, "epoch": 0.51461988, "global_step/max_steps": "1122/2181", "elapsed_time": "2h 57m 48s", "remaining_time": "2h 47m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.508219}
{"loss": 24.0, "grad_norm": 1.2105397, "learning_rate": 0.00050037, "token_acc": 0.382469, "epoch": 0.51507855, "global_step/max_steps": "1123/2181", "elapsed_time": "2h 57m 58s", "remaining_time": "2h 47m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.508797}
{"loss": 24.25, "grad_norm": 1.23983836, "learning_rate": 0.00049963, "token_acc": 0.38596491, "epoch": 0.51553721, "global_step/max_steps": "1124/2181", "elapsed_time": "2h 58m 8s", "remaining_time": "2h 47m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.508783}
{"loss": 24.0, "grad_norm": 1.30918503, "learning_rate": 0.00049889, "token_acc": 0.38492176, "epoch": 0.51599587, "global_step/max_steps": "1125/2181", "elapsed_time": "2h 58m 18s", "remaining_time": "2h 47m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.509211}
{"loss": 23.25, "grad_norm": 1.13809252, "learning_rate": 0.00049814, "token_acc": 0.39630045, "epoch": 0.51645454, "global_step/max_steps": "1126/2181", "elapsed_time": "2h 58m 27s", "remaining_time": "2h 47m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.509276}
{"loss": 24.25, "grad_norm": 1.40573227, "learning_rate": 0.0004974, "token_acc": 0.39022989, "epoch": 0.5169132, "global_step/max_steps": "1127/2181", "elapsed_time": "2h 58m 37s", "remaining_time": "2h 47m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.509098}
{"loss": 23.75, "grad_norm": 1.21023691, "learning_rate": 0.00049666, "token_acc": 0.3967655, "epoch": 0.51737186, "global_step/max_steps": "1128/2181", "elapsed_time": "2h 58m 46s", "remaining_time": "2h 46m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.509035}
{"loss": 23.375, "grad_norm": 1.14709079, "learning_rate": 0.00049592, "token_acc": 0.3952164, "epoch": 0.51783052, "global_step/max_steps": "1129/2181", "elapsed_time": "2h 58m 56s", "remaining_time": "2h 46m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.509396}
{"loss": 23.375, "grad_norm": 1.12738001, "learning_rate": 0.00049517, "token_acc": 0.39065205, "epoch": 0.51828919, "global_step/max_steps": "1130/2181", "elapsed_time": "2h 59m 5s", "remaining_time": "2h 46m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.509232}
{"loss": 24.0, "grad_norm": 1.09856188, "learning_rate": 0.00049443, "token_acc": 0.37677924, "epoch": 0.51874785, "global_step/max_steps": "1131/2181", "elapsed_time": "2h 59m 16s", "remaining_time": "2h 46m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.509912}
{"loss": 24.0, "grad_norm": 1.07574141, "learning_rate": 0.00049369, "token_acc": 0.3956448, "epoch": 0.51920651, "global_step/max_steps": "1132/2181", "elapsed_time": "2h 59m 25s", "remaining_time": "2h 46m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.509745}
{"loss": 24.375, "grad_norm": 1.20255649, "learning_rate": 0.00049294, "token_acc": 0.36635354, "epoch": 0.51966518, "global_step/max_steps": "1133/2181", "elapsed_time": "2h 59m 35s", "remaining_time": "2h 46m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.509901}
{"loss": 24.875, "grad_norm": 1.07275879, "learning_rate": 0.0004922, "token_acc": 0.38222465, "epoch": 0.52012384, "global_step/max_steps": "1134/2181", "elapsed_time": "2h 59m 44s", "remaining_time": "2h 45m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.509969}
{"loss": 23.5, "grad_norm": 1.26921511, "learning_rate": 0.00049146, "token_acc": 0.39162844, "epoch": 0.5205825, "global_step/max_steps": "1135/2181", "elapsed_time": "2h 59m 54s", "remaining_time": "2h 45m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.510023}
{"loss": 23.875, "grad_norm": 1.12396097, "learning_rate": 0.00049072, "token_acc": 0.37822111, "epoch": 0.52104117, "global_step/max_steps": "1136/2181", "elapsed_time": "3h 0m 4s", "remaining_time": "2h 45m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.510178}
{"loss": 23.75, "grad_norm": 1.14188051, "learning_rate": 0.00048997, "token_acc": 0.3930974, "epoch": 0.52149983, "global_step/max_steps": "1137/2181", "elapsed_time": "3h 0m 14s", "remaining_time": "2h 45m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.510675}
{"loss": 24.375, "grad_norm": 1.27763724, "learning_rate": 0.00048923, "token_acc": 0.37687998, "epoch": 0.52195849, "global_step/max_steps": "1138/2181", "elapsed_time": "3h 0m 24s", "remaining_time": "2h 45m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511245}
{"loss": 23.75, "grad_norm": 1.13940036, "learning_rate": 0.00048849, "token_acc": 0.38592633, "epoch": 0.52241715, "global_step/max_steps": "1139/2181", "elapsed_time": "3h 0m 34s", "remaining_time": "2h 45m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511491}
{"loss": 23.5, "grad_norm": 1.10429406, "learning_rate": 0.00048775, "token_acc": 0.39042208, "epoch": 0.52287582, "global_step/max_steps": "1140/2181", "elapsed_time": "3h 0m 43s", "remaining_time": "2h 45m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511131}
{"loss": 24.25, "grad_norm": 1.20337188, "learning_rate": 0.000487, "token_acc": 0.37614679, "epoch": 0.52333448, "global_step/max_steps": "1141/2181", "elapsed_time": "3h 0m 52s", "remaining_time": "2h 44m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511118}
{"loss": 24.5, "grad_norm": 1.15233946, "learning_rate": 0.00048626, "token_acc": 0.38070511, "epoch": 0.52379314, "global_step/max_steps": "1142/2181", "elapsed_time": "3h 1m 2s", "remaining_time": "2h 44m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511313}
{"loss": 24.0, "grad_norm": 1.26442456, "learning_rate": 0.00048552, "token_acc": 0.379124, "epoch": 0.52425181, "global_step/max_steps": "1143/2181", "elapsed_time": "3h 1m 12s", "remaining_time": "2h 44m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511534}
{"loss": 23.625, "grad_norm": 1.22808588, "learning_rate": 0.00048478, "token_acc": 0.39201811, "epoch": 0.52471047, "global_step/max_steps": "1144/2181", "elapsed_time": "3h 1m 21s", "remaining_time": "2h 44m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511775}
{"loss": 24.75, "grad_norm": 1.22026646, "learning_rate": 0.00048403, "token_acc": 0.36407898, "epoch": 0.52516913, "global_step/max_steps": "1145/2181", "elapsed_time": "3h 1m 31s", "remaining_time": "2h 44m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511623}
{"loss": 24.75, "grad_norm": 1.14784992, "learning_rate": 0.00048329, "token_acc": 0.37617893, "epoch": 0.52562779, "global_step/max_steps": "1146/2181", "elapsed_time": "3h 1m 40s", "remaining_time": "2h 44m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.51153}
{"loss": 24.75, "grad_norm": 1.12676847, "learning_rate": 0.00048255, "token_acc": 0.37672158, "epoch": 0.52608646, "global_step/max_steps": "1147/2181", "elapsed_time": "3h 1m 50s", "remaining_time": "2h 43m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511401}
{"loss": 23.5, "grad_norm": 1.24659312, "learning_rate": 0.00048181, "token_acc": 0.38920781, "epoch": 0.52654512, "global_step/max_steps": "1148/2181", "elapsed_time": "3h 1m 59s", "remaining_time": "2h 43m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511408}
{"loss": 24.75, "grad_norm": 1.18484461, "learning_rate": 0.00048107, "token_acc": 0.36151927, "epoch": 0.52700378, "global_step/max_steps": "1149/2181", "elapsed_time": "3h 2m 9s", "remaining_time": "2h 43m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512063}
{"loss": 23.75, "grad_norm": 1.22705913, "learning_rate": 0.00048032, "token_acc": 0.40320856, "epoch": 0.52746245, "global_step/max_steps": "1150/2181", "elapsed_time": "3h 2m 19s", "remaining_time": "2h 43m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511798}
{"loss": 24.25, "grad_norm": 1.16987193, "learning_rate": 0.00047958, "token_acc": 0.3829376, "epoch": 0.52792111, "global_step/max_steps": "1151/2181", "elapsed_time": "3h 2m 28s", "remaining_time": "2h 43m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511477}
{"loss": 23.75, "grad_norm": 1.12271798, "learning_rate": 0.00047884, "token_acc": 0.39275843, "epoch": 0.52837977, "global_step/max_steps": "1152/2181", "elapsed_time": "3h 2m 37s", "remaining_time": "2h 43m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511178}
{"loss": 24.5, "grad_norm": 1.39248049, "learning_rate": 0.0004781, "token_acc": 0.37445534, "epoch": 0.52883844, "global_step/max_steps": "1153/2181", "elapsed_time": "3h 2m 46s", "remaining_time": "2h 42m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.510762}
{"loss": 23.5, "grad_norm": 1.335253, "learning_rate": 0.00047736, "token_acc": 0.39334638, "epoch": 0.5292971, "global_step/max_steps": "1154/2181", "elapsed_time": "3h 2m 55s", "remaining_time": "2h 42m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.510648}
{"loss": 24.125, "grad_norm": 1.26058376, "learning_rate": 0.00047661, "token_acc": 0.39504086, "epoch": 0.52975576, "global_step/max_steps": "1155/2181", "elapsed_time": "3h 3m 5s", "remaining_time": "2h 42m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.51101}
{"loss": 24.0, "grad_norm": 1.44064736, "learning_rate": 0.00047587, "token_acc": 0.39507198, "epoch": 0.53021442, "global_step/max_steps": "1156/2181", "elapsed_time": "3h 3m 15s", "remaining_time": "2h 42m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511134}
{"loss": 24.5, "grad_norm": 1.12311566, "learning_rate": 0.00047513, "token_acc": 0.37049362, "epoch": 0.53067309, "global_step/max_steps": "1157/2181", "elapsed_time": "3h 3m 25s", "remaining_time": "2h 42m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512053}
{"loss": 24.125, "grad_norm": 1.26742947, "learning_rate": 0.00047439, "token_acc": 0.38149973, "epoch": 0.53113175, "global_step/max_steps": "1158/2181", "elapsed_time": "3h 3m 35s", "remaining_time": "2h 42m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511659}
{"loss": 23.625, "grad_norm": 1.32966208, "learning_rate": 0.00047365, "token_acc": 0.38016768, "epoch": 0.53159041, "global_step/max_steps": "1159/2181", "elapsed_time": "3h 3m 45s", "remaining_time": "2h 42m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.51268}
{"loss": 23.75, "grad_norm": 1.13597429, "learning_rate": 0.0004729, "token_acc": 0.39431466, "epoch": 0.53204908, "global_step/max_steps": "1160/2181", "elapsed_time": "3h 3m 55s", "remaining_time": "2h 41m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512917}
{"loss": 24.125, "grad_norm": 1.16891146, "learning_rate": 0.00047216, "token_acc": 0.39143816, "epoch": 0.53250774, "global_step/max_steps": "1161/2181", "elapsed_time": "3h 4m 5s", "remaining_time": "2h 41m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.513037}
{"loss": 24.0, "grad_norm": 1.27919281, "learning_rate": 0.00047142, "token_acc": 0.38501144, "epoch": 0.5329664, "global_step/max_steps": "1162/2181", "elapsed_time": "3h 4m 14s", "remaining_time": "2h 41m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.513003}
{"loss": 22.75, "grad_norm": 1.08360279, "learning_rate": 0.00047068, "token_acc": 0.39983467, "epoch": 0.53342507, "global_step/max_steps": "1163/2181", "elapsed_time": "3h 4m 24s", "remaining_time": "2h 41m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.51294}
{"loss": 23.25, "grad_norm": 1.16300631, "learning_rate": 0.00046994, "token_acc": 0.38667772, "epoch": 0.53388373, "global_step/max_steps": "1164/2181", "elapsed_time": "3h 4m 33s", "remaining_time": "2h 41m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512553}
{"loss": 24.125, "grad_norm": 1.11009884, "learning_rate": 0.0004692, "token_acc": 0.38049605, "epoch": 0.53434239, "global_step/max_steps": "1165/2181", "elapsed_time": "3h 4m 42s", "remaining_time": "2h 41m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512316}
{"loss": 23.625, "grad_norm": 1.23107708, "learning_rate": 0.00046846, "token_acc": 0.39374821, "epoch": 0.53480105, "global_step/max_steps": "1166/2181", "elapsed_time": "3h 4m 52s", "remaining_time": "2h 40m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512508}
{"loss": 24.25, "grad_norm": 1.17914736, "learning_rate": 0.00046772, "token_acc": 0.37533656, "epoch": 0.53525972, "global_step/max_steps": "1167/2181", "elapsed_time": "3h 5m 1s", "remaining_time": "2h 40m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512082}
{"loss": 24.375, "grad_norm": 1.2096771, "learning_rate": 0.00046697, "token_acc": 0.37449007, "epoch": 0.53571838, "global_step/max_steps": "1168/2181", "elapsed_time": "3h 5m 10s", "remaining_time": "2h 40m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511644}
{"loss": 24.25, "grad_norm": 1.24277353, "learning_rate": 0.00046623, "token_acc": 0.37746479, "epoch": 0.53617704, "global_step/max_steps": "1169/2181", "elapsed_time": "3h 5m 20s", "remaining_time": "2h 40m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512434}
{"loss": 24.0, "grad_norm": 1.31742597, "learning_rate": 0.00046549, "token_acc": 0.37630761, "epoch": 0.53663571, "global_step/max_steps": "1170/2181", "elapsed_time": "3h 5m 30s", "remaining_time": "2h 40m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512574}
{"loss": 24.25, "grad_norm": 1.39147484, "learning_rate": 0.00046475, "token_acc": 0.38262032, "epoch": 0.53709437, "global_step/max_steps": "1171/2181", "elapsed_time": "3h 5m 39s", "remaining_time": "2h 40m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512172}
{"loss": 23.625, "grad_norm": 1.22353268, "learning_rate": 0.00046401, "token_acc": 0.40080107, "epoch": 0.53755303, "global_step/max_steps": "1172/2181", "elapsed_time": "3h 5m 48s", "remaining_time": "2h 39m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.51202}
{"loss": 24.25, "grad_norm": 1.19238245, "learning_rate": 0.00046327, "token_acc": 0.37838634, "epoch": 0.5380117, "global_step/max_steps": "1173/2181", "elapsed_time": "3h 5m 57s", "remaining_time": "2h 39m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511737}
{"loss": 23.25, "grad_norm": 1.16248012, "learning_rate": 0.00046253, "token_acc": 0.38322656, "epoch": 0.53847036, "global_step/max_steps": "1174/2181", "elapsed_time": "3h 6m 7s", "remaining_time": "2h 39m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511838}
{"loss": 24.375, "grad_norm": 1.2204771, "learning_rate": 0.00046179, "token_acc": 0.37677133, "epoch": 0.53892902, "global_step/max_steps": "1175/2181", "elapsed_time": "3h 6m 16s", "remaining_time": "2h 39m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511758}
{"loss": 24.375, "grad_norm": 1.07748055, "learning_rate": 0.00046105, "token_acc": 0.39113011, "epoch": 0.53938768, "global_step/max_steps": "1176/2181", "elapsed_time": "3h 6m 26s", "remaining_time": "2h 39m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512041}
{"loss": 23.375, "grad_norm": 1.28243756, "learning_rate": 0.00046031, "token_acc": 0.39948965, "epoch": 0.53984635, "global_step/max_steps": "1177/2181", "elapsed_time": "3h 6m 36s", "remaining_time": "2h 39m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512271}
{"loss": 24.125, "grad_norm": 1.22622347, "learning_rate": 0.00045957, "token_acc": 0.38364604, "epoch": 0.54030501, "global_step/max_steps": "1178/2181", "elapsed_time": "3h 6m 45s", "remaining_time": "2h 39m 0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511893}
{"loss": 24.25, "grad_norm": 1.0999583, "learning_rate": 0.00045883, "token_acc": 0.38770354, "epoch": 0.54076367, "global_step/max_steps": "1179/2181", "elapsed_time": "3h 6m 55s", "remaining_time": "2h 38m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512129}
{"loss": 24.625, "grad_norm": 1.13783264, "learning_rate": 0.00045809, "token_acc": 0.38452946, "epoch": 0.54122234, "global_step/max_steps": "1180/2181", "elapsed_time": "3h 7m 4s", "remaining_time": "2h 38m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512228}
{"loss": 24.25, "grad_norm": 1.08670044, "learning_rate": 0.00045735, "token_acc": 0.38461538, "epoch": 0.541681, "global_step/max_steps": "1181/2181", "elapsed_time": "3h 7m 14s", "remaining_time": "2h 38m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512115}
{"loss": 24.0, "grad_norm": 1.21151137, "learning_rate": 0.00045661, "token_acc": 0.38250577, "epoch": 0.54213966, "global_step/max_steps": "1182/2181", "elapsed_time": "3h 7m 23s", "remaining_time": "2h 38m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511938}
{"loss": 24.625, "grad_norm": 1.15156233, "learning_rate": 0.00045587, "token_acc": 0.37389625, "epoch": 0.54259833, "global_step/max_steps": "1183/2181", "elapsed_time": "3h 7m 33s", "remaining_time": "2h 38m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512603}
{"loss": 24.0, "grad_norm": 1.20461679, "learning_rate": 0.00045513, "token_acc": 0.37933931, "epoch": 0.54305699, "global_step/max_steps": "1184/2181", "elapsed_time": "3h 7m 43s", "remaining_time": "2h 38m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512441}
{"loss": 23.75, "grad_norm": 1.29987681, "learning_rate": 0.00045439, "token_acc": 0.39246694, "epoch": 0.54351565, "global_step/max_steps": "1185/2181", "elapsed_time": "3h 7m 53s", "remaining_time": "2h 37m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512715}
{"loss": 24.375, "grad_norm": 1.15694487, "learning_rate": 0.00045365, "token_acc": 0.38519343, "epoch": 0.54397431, "global_step/max_steps": "1186/2181", "elapsed_time": "3h 8m 2s", "remaining_time": "2h 37m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512722}
{"loss": 24.25, "grad_norm": 1.12986887, "learning_rate": 0.00045291, "token_acc": 0.38811881, "epoch": 0.54443298, "global_step/max_steps": "1187/2181", "elapsed_time": "3h 8m 11s", "remaining_time": "2h 37m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512121}
{"loss": 24.375, "grad_norm": 1.09610665, "learning_rate": 0.00045217, "token_acc": 0.3806114, "epoch": 0.54489164, "global_step/max_steps": "1188/2181", "elapsed_time": "3h 8m 20s", "remaining_time": "2h 37m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512167}
{"loss": 23.5, "grad_norm": 1.04850769, "learning_rate": 0.00045143, "token_acc": 0.38936346, "epoch": 0.5453503, "global_step/max_steps": "1189/2181", "elapsed_time": "3h 8m 30s", "remaining_time": "2h 37m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512402}
{"loss": 24.5, "grad_norm": 1.00064158, "learning_rate": 0.00045069, "token_acc": 0.38011364, "epoch": 0.54580897, "global_step/max_steps": "1190/2181", "elapsed_time": "3h 8m 40s", "remaining_time": "2h 37m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.51225}
{"loss": 24.0, "grad_norm": 1.15565336, "learning_rate": 0.00044995, "token_acc": 0.37790698, "epoch": 0.54626763, "global_step/max_steps": "1191/2181", "elapsed_time": "3h 8m 49s", "remaining_time": "2h 36m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.51213}
{"loss": 23.375, "grad_norm": 1.12150562, "learning_rate": 0.00044921, "token_acc": 0.39313528, "epoch": 0.54672629, "global_step/max_steps": "1192/2181", "elapsed_time": "3h 8m 58s", "remaining_time": "2h 36m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511765}
{"loss": 24.0, "grad_norm": 1.36559618, "learning_rate": 0.00044847, "token_acc": 0.38124098, "epoch": 0.54718496, "global_step/max_steps": "1193/2181", "elapsed_time": "3h 9m 8s", "remaining_time": "2h 36m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.5118}
{"loss": 24.0, "grad_norm": 1.12108397, "learning_rate": 0.00044774, "token_acc": 0.38081562, "epoch": 0.54764362, "global_step/max_steps": "1194/2181", "elapsed_time": "3h 9m 17s", "remaining_time": "2h 36m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.51195}
{"loss": 24.125, "grad_norm": 1.16626358, "learning_rate": 0.000447, "token_acc": 0.39039781, "epoch": 0.54810228, "global_step/max_steps": "1195/2181", "elapsed_time": "3h 9m 27s", "remaining_time": "2h 36m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511748}
{"loss": 24.375, "grad_norm": 1.08424163, "learning_rate": 0.00044626, "token_acc": 0.37655658, "epoch": 0.54856094, "global_step/max_steps": "1196/2181", "elapsed_time": "3h 9m 36s", "remaining_time": "2h 36m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511833}
{"loss": 24.75, "grad_norm": 1.20662272, "learning_rate": 0.00044552, "token_acc": 0.37471198, "epoch": 0.54901961, "global_step/max_steps": "1197/2181", "elapsed_time": "3h 9m 45s", "remaining_time": "2h 35m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511655}
{"loss": 23.5, "grad_norm": 1.26601875, "learning_rate": 0.00044478, "token_acc": 0.3836858, "epoch": 0.54947827, "global_step/max_steps": "1198/2181", "elapsed_time": "3h 9m 55s", "remaining_time": "2h 35m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511915}
{"loss": 24.125, "grad_norm": 1.21282315, "learning_rate": 0.00044404, "token_acc": 0.38934542, "epoch": 0.54993693, "global_step/max_steps": "1199/2181", "elapsed_time": "3h 10m 5s", "remaining_time": "2h 35m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512209}
{"loss": 25.25, "grad_norm": 1.27827084, "learning_rate": 0.00044331, "token_acc": 0.37073171, "epoch": 0.5503956, "global_step/max_steps": "1200/2181", "elapsed_time": "3h 10m 14s", "remaining_time": "2h 35m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511533}
{"loss": 24.25, "grad_norm": 1.16104102, "learning_rate": 0.00044257, "token_acc": 0.37980769, "epoch": 0.55085426, "global_step/max_steps": "1201/2181", "elapsed_time": "3h 10m 23s", "remaining_time": "2h 35m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511348}
{"loss": 23.125, "grad_norm": 1.2563864, "learning_rate": 0.00044183, "token_acc": 0.40336366, "epoch": 0.55131292, "global_step/max_steps": "1202/2181", "elapsed_time": "3h 10m 33s", "remaining_time": "2h 35m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511866}
{"loss": 24.25, "grad_norm": 1.07471824, "learning_rate": 0.00044109, "token_acc": 0.3952739, "epoch": 0.55177159, "global_step/max_steps": "1203/2181", "elapsed_time": "3h 10m 42s", "remaining_time": "2h 35m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511525}
{"loss": 23.625, "grad_norm": 1.09203601, "learning_rate": 0.00044036, "token_acc": 0.39596137, "epoch": 0.55223025, "global_step/max_steps": "1204/2181", "elapsed_time": "3h 10m 53s", "remaining_time": "2h 34m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512073}
{"loss": 23.25, "grad_norm": 1.23394668, "learning_rate": 0.00043962, "token_acc": 0.39926429, "epoch": 0.55268891, "global_step/max_steps": "1205/2181", "elapsed_time": "3h 11m 2s", "remaining_time": "2h 34m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512132}
{"loss": 23.5, "grad_norm": 1.21413386, "learning_rate": 0.00043888, "token_acc": 0.38910506, "epoch": 0.55314757, "global_step/max_steps": "1206/2181", "elapsed_time": "3h 11m 11s", "remaining_time": "2h 34m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511816}
{"loss": 24.5, "grad_norm": 1.33397388, "learning_rate": 0.00043814, "token_acc": 0.36444572, "epoch": 0.55360624, "global_step/max_steps": "1207/2181", "elapsed_time": "3h 11m 21s", "remaining_time": "2h 34m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512103}
{"loss": 23.75, "grad_norm": 1.34151125, "learning_rate": 0.00043741, "token_acc": 0.38593622, "epoch": 0.5540649, "global_step/max_steps": "1208/2181", "elapsed_time": "3h 11m 30s", "remaining_time": "2h 34m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511963}
{"loss": 24.125, "grad_norm": 1.25039423, "learning_rate": 0.00043667, "token_acc": 0.38014184, "epoch": 0.55452356, "global_step/max_steps": "1209/2181", "elapsed_time": "3h 11m 40s", "remaining_time": "2h 34m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511758}
{"loss": 24.125, "grad_norm": 1.26401591, "learning_rate": 0.00043593, "token_acc": 0.39197967, "epoch": 0.55498223, "global_step/max_steps": "1210/2181", "elapsed_time": "3h 11m 49s", "remaining_time": "2h 33m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511267}
{"loss": 23.75, "grad_norm": 1.16577125, "learning_rate": 0.0004352, "token_acc": 0.38782413, "epoch": 0.55544089, "global_step/max_steps": "1211/2181", "elapsed_time": "3h 11m 60s", "remaining_time": "2h 33m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512454}
{"loss": 24.375, "grad_norm": 1.18203664, "learning_rate": 0.00043446, "token_acc": 0.37279102, "epoch": 0.55589955, "global_step/max_steps": "1212/2181", "elapsed_time": "3h 12m 9s", "remaining_time": "2h 33m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512302}
{"loss": 24.125, "grad_norm": 1.12207472, "learning_rate": 0.00043372, "token_acc": 0.38303847, "epoch": 0.55635822, "global_step/max_steps": "1213/2181", "elapsed_time": "3h 12m 18s", "remaining_time": "2h 33m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511829}
{"loss": 23.25, "grad_norm": 1.1509167, "learning_rate": 0.00043299, "token_acc": 0.39105668, "epoch": 0.55681688, "global_step/max_steps": "1214/2181", "elapsed_time": "3h 12m 28s", "remaining_time": "2h 33m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511997}
{"loss": 23.5, "grad_norm": 1.06965661, "learning_rate": 0.00043225, "token_acc": 0.41049216, "epoch": 0.55727554, "global_step/max_steps": "1215/2181", "elapsed_time": "3h 12m 37s", "remaining_time": "2h 33m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511895}
{"loss": 23.875, "grad_norm": 1.14825618, "learning_rate": 0.00043152, "token_acc": 0.38424483, "epoch": 0.5577342, "global_step/max_steps": "1216/2181", "elapsed_time": "3h 12m 46s", "remaining_time": "2h 32m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.51152}
{"loss": 24.0, "grad_norm": 1.33978915, "learning_rate": 0.00043078, "token_acc": 0.38106236, "epoch": 0.55819287, "global_step/max_steps": "1217/2181", "elapsed_time": "3h 12m 56s", "remaining_time": "2h 32m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.51176}
{"loss": 24.75, "grad_norm": 1.12556303, "learning_rate": 0.00043005, "token_acc": 0.37697642, "epoch": 0.55865153, "global_step/max_steps": "1218/2181", "elapsed_time": "3h 13m 6s", "remaining_time": "2h 32m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.511955}
{"loss": 24.75, "grad_norm": 1.21879482, "learning_rate": 0.00042931, "token_acc": 0.38382193, "epoch": 0.55911019, "global_step/max_steps": "1219/2181", "elapsed_time": "3h 13m 15s", "remaining_time": "2h 32m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512247}
{"loss": 24.5, "grad_norm": 1.06712568, "learning_rate": 0.00042858, "token_acc": 0.37463672, "epoch": 0.55956886, "global_step/max_steps": "1220/2181", "elapsed_time": "3h 13m 25s", "remaining_time": "2h 32m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512382}
{"loss": 24.0, "grad_norm": 1.35844648, "learning_rate": 0.00042784, "token_acc": 0.38718022, "epoch": 0.56002752, "global_step/max_steps": "1221/2181", "elapsed_time": "3h 13m 35s", "remaining_time": "2h 32m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512688}
{"loss": 24.625, "grad_norm": 1.04676247, "learning_rate": 0.00042711, "token_acc": 0.37503437, "epoch": 0.56048618, "global_step/max_steps": "1222/2181", "elapsed_time": "3h 13m 44s", "remaining_time": "2h 32m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512575}
{"loss": 24.375, "grad_norm": 1.06369197, "learning_rate": 0.00042637, "token_acc": 0.37846656, "epoch": 0.56094485, "global_step/max_steps": "1223/2181", "elapsed_time": "3h 13m 54s", "remaining_time": "2h 31m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512337}
{"loss": 24.125, "grad_norm": 1.13346827, "learning_rate": 0.00042564, "token_acc": 0.38173302, "epoch": 0.56140351, "global_step/max_steps": "1224/2181", "elapsed_time": "3h 14m 3s", "remaining_time": "2h 31m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512408}
{"loss": 23.75, "grad_norm": 1.24907184, "learning_rate": 0.0004249, "token_acc": 0.38658329, "epoch": 0.56186217, "global_step/max_steps": "1225/2181", "elapsed_time": "3h 14m 13s", "remaining_time": "2h 31m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.513018}
{"loss": 24.625, "grad_norm": 1.15889931, "learning_rate": 0.00042417, "token_acc": 0.38697318, "epoch": 0.56232083, "global_step/max_steps": "1226/2181", "elapsed_time": "3h 14m 23s", "remaining_time": "2h 31m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512946}
{"loss": 23.625, "grad_norm": 1.23684788, "learning_rate": 0.00042343, "token_acc": 0.3928675, "epoch": 0.5627795, "global_step/max_steps": "1227/2181", "elapsed_time": "3h 14m 32s", "remaining_time": "2h 31m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.512883}
{"loss": 24.25, "grad_norm": 1.20171499, "learning_rate": 0.0004227, "token_acc": 0.38890392, "epoch": 0.56323816, "global_step/max_steps": "1228/2181", "elapsed_time": "3h 14m 42s", "remaining_time": "2h 31m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.513374}
{"loss": 23.75, "grad_norm": 1.2263633, "learning_rate": 0.00042197, "token_acc": 0.39602018, "epoch": 0.56369682, "global_step/max_steps": "1229/2181", "elapsed_time": "3h 14m 52s", "remaining_time": "2h 30m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.513811}
{"loss": 23.75, "grad_norm": 1.09281588, "learning_rate": 0.00042123, "token_acc": 0.3868113, "epoch": 0.56415549, "global_step/max_steps": "1230/2181", "elapsed_time": "3h 15m 3s", "remaining_time": "2h 30m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.514542}
{"loss": 24.875, "grad_norm": 1.12459683, "learning_rate": 0.0004205, "token_acc": 0.38257576, "epoch": 0.56461415, "global_step/max_steps": "1231/2181", "elapsed_time": "3h 15m 12s", "remaining_time": "2h 30m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.514542}
{"loss": 24.75, "grad_norm": 1.08039391, "learning_rate": 0.00041977, "token_acc": 0.37016726, "epoch": 0.56507281, "global_step/max_steps": "1232/2181", "elapsed_time": "3h 15m 22s", "remaining_time": "2h 30m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.514239}
{"loss": 24.125, "grad_norm": 1.16477609, "learning_rate": 0.00041903, "token_acc": 0.39579832, "epoch": 0.56553148, "global_step/max_steps": "1233/2181", "elapsed_time": "3h 15m 31s", "remaining_time": "2h 30m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.514245}
{"loss": 24.875, "grad_norm": 1.09204423, "learning_rate": 0.0004183, "token_acc": 0.36286201, "epoch": 0.56599014, "global_step/max_steps": "1234/2181", "elapsed_time": "3h 15m 41s", "remaining_time": "2h 30m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.514482}
{"loss": 24.0, "grad_norm": 1.06930566, "learning_rate": 0.00041757, "token_acc": 0.39371174, "epoch": 0.5664488, "global_step/max_steps": "1235/2181", "elapsed_time": "3h 15m 50s", "remaining_time": "2h 30m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.514288}
{"loss": 23.75, "grad_norm": 1.14008093, "learning_rate": 0.00041684, "token_acc": 0.38550027, "epoch": 0.56690746, "global_step/max_steps": "1236/2181", "elapsed_time": "3h 15m 59s", "remaining_time": "2h 29m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.513668}
{"loss": 24.0, "grad_norm": 1.09940696, "learning_rate": 0.0004161, "token_acc": 0.38478685, "epoch": 0.56736613, "global_step/max_steps": "1237/2181", "elapsed_time": "3h 16m 9s", "remaining_time": "2h 29m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.514164}
{"loss": 24.25, "grad_norm": 1.25034046, "learning_rate": 0.00041537, "token_acc": 0.38647746, "epoch": 0.56782479, "global_step/max_steps": "1238/2181", "elapsed_time": "3h 16m 19s", "remaining_time": "2h 29m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.514711}
{"loss": 24.25, "grad_norm": 1.1635679, "learning_rate": 0.00041464, "token_acc": 0.38532884, "epoch": 0.56828345, "global_step/max_steps": "1239/2181", "elapsed_time": "3h 16m 28s", "remaining_time": "2h 29m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.514073}
{"loss": 24.0, "grad_norm": 1.14518297, "learning_rate": 0.00041391, "token_acc": 0.39438944, "epoch": 0.56874212, "global_step/max_steps": "1240/2181", "elapsed_time": "3h 16m 38s", "remaining_time": "2h 29m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.514285}
{"loss": 24.125, "grad_norm": 1.17723668, "learning_rate": 0.00041318, "token_acc": 0.3823781, "epoch": 0.56920078, "global_step/max_steps": "1241/2181", "elapsed_time": "3h 16m 47s", "remaining_time": "2h 29m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.514504}
{"loss": 23.875, "grad_norm": 1.0888164, "learning_rate": 0.00041244, "token_acc": 0.39478022, "epoch": 0.56965944, "global_step/max_steps": "1242/2181", "elapsed_time": "3h 16m 56s", "remaining_time": "2h 28m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.513785}
{"loss": 24.625, "grad_norm": 1.20164251, "learning_rate": 0.00041171, "token_acc": 0.3658331, "epoch": 0.57011811, "global_step/max_steps": "1243/2181", "elapsed_time": "3h 17m 6s", "remaining_time": "2h 28m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.514178}
{"loss": 23.75, "grad_norm": 1.00116956, "learning_rate": 0.00041098, "token_acc": 0.39430449, "epoch": 0.57057677, "global_step/max_steps": "1244/2181", "elapsed_time": "3h 17m 17s", "remaining_time": "2h 28m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.515106}
{"loss": 23.25, "grad_norm": 1.11623871, "learning_rate": 0.00041025, "token_acc": 0.39424173, "epoch": 0.57103543, "global_step/max_steps": "1245/2181", "elapsed_time": "3h 17m 26s", "remaining_time": "2h 28m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.51489}
{"loss": 23.625, "grad_norm": 1.23598623, "learning_rate": 0.00040952, "token_acc": 0.38163972, "epoch": 0.57149409, "global_step/max_steps": "1246/2181", "elapsed_time": "3h 17m 35s", "remaining_time": "2h 28m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.514598}
{"loss": 23.875, "grad_norm": 1.14329457, "learning_rate": 0.00040879, "token_acc": 0.37220492, "epoch": 0.57195276, "global_step/max_steps": "1247/2181", "elapsed_time": "3h 17m 46s", "remaining_time": "2h 28m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.515304}
{"loss": 23.875, "grad_norm": 1.22007167, "learning_rate": 0.00040806, "token_acc": 0.38764783, "epoch": 0.57241142, "global_step/max_steps": "1248/2181", "elapsed_time": "3h 17m 55s", "remaining_time": "2h 27m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.515435}
{"loss": 24.375, "grad_norm": 1.11684453, "learning_rate": 0.00040733, "token_acc": 0.375, "epoch": 0.57287008, "global_step/max_steps": "1249/2181", "elapsed_time": "3h 18m 5s", "remaining_time": "2h 27m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.515755}
{"loss": 23.875, "grad_norm": 1.28183532, "learning_rate": 0.0004066, "token_acc": 0.39585666, "epoch": 0.57332875, "global_step/max_steps": "1250/2181", "elapsed_time": "3h 18m 15s", "remaining_time": "2h 27m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516359}
{"loss": 24.375, "grad_norm": 1.03312469, "learning_rate": 0.00040587, "token_acc": 0.38020973, "epoch": 0.57378741, "global_step/max_steps": "1251/2181", "elapsed_time": "3h 18m 24s", "remaining_time": "2h 27m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.515834}
{"loss": 24.25, "grad_norm": 1.0923878, "learning_rate": 0.00040514, "token_acc": 0.37615449, "epoch": 0.57424607, "global_step/max_steps": "1252/2181", "elapsed_time": "3h 18m 34s", "remaining_time": "2h 27m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.515674}
{"loss": 24.25, "grad_norm": 1.10662746, "learning_rate": 0.00040441, "token_acc": 0.38813651, "epoch": 0.57470474, "global_step/max_steps": "1253/2181", "elapsed_time": "3h 18m 44s", "remaining_time": "2h 27m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.51625}
{"loss": 24.25, "grad_norm": 1.11885834, "learning_rate": 0.00040368, "token_acc": 0.37341772, "epoch": 0.5751634, "global_step/max_steps": "1254/2181", "elapsed_time": "3h 18m 54s", "remaining_time": "2h 27m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516769}
{"loss": 23.875, "grad_norm": 1.35249662, "learning_rate": 0.00040296, "token_acc": 0.3800554, "epoch": 0.57562206, "global_step/max_steps": "1255/2181", "elapsed_time": "3h 19m 4s", "remaining_time": "2h 26m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516988}
{"loss": 23.875, "grad_norm": 1.06239784, "learning_rate": 0.00040223, "token_acc": 0.37517091, "epoch": 0.57608072, "global_step/max_steps": "1256/2181", "elapsed_time": "3h 19m 13s", "remaining_time": "2h 26m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516907}
{"loss": 24.75, "grad_norm": 1.26113605, "learning_rate": 0.0004015, "token_acc": 0.37943925, "epoch": 0.57653939, "global_step/max_steps": "1257/2181", "elapsed_time": "3h 19m 23s", "remaining_time": "2h 26m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.51674}
{"loss": 24.25, "grad_norm": 1.31799984, "learning_rate": 0.00040077, "token_acc": 0.39589117, "epoch": 0.57699805, "global_step/max_steps": "1258/2181", "elapsed_time": "3h 19m 31s", "remaining_time": "2h 26m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516258}
{"loss": 23.875, "grad_norm": 1.07562625, "learning_rate": 0.00040004, "token_acc": 0.38322237, "epoch": 0.57745671, "global_step/max_steps": "1259/2181", "elapsed_time": "3h 19m 41s", "remaining_time": "2h 26m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516412}
{"loss": 23.125, "grad_norm": 1.16787052, "learning_rate": 0.00039932, "token_acc": 0.39956391, "epoch": 0.57791538, "global_step/max_steps": "1260/2181", "elapsed_time": "3h 19m 51s", "remaining_time": "2h 26m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516486}
{"loss": 24.25, "grad_norm": 1.11987579, "learning_rate": 0.00039859, "token_acc": 0.3838728, "epoch": 0.57837404, "global_step/max_steps": "1261/2181", "elapsed_time": "3h 20m 0s", "remaining_time": "2h 25m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516533}
{"loss": 23.625, "grad_norm": 1.22806776, "learning_rate": 0.00039786, "token_acc": 0.38341245, "epoch": 0.5788327, "global_step/max_steps": "1262/2181", "elapsed_time": "3h 20m 9s", "remaining_time": "2h 25m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516097}
{"loss": 24.0, "grad_norm": 1.24911118, "learning_rate": 0.00039713, "token_acc": 0.3842532, "epoch": 0.57929137, "global_step/max_steps": "1263/2181", "elapsed_time": "3h 20m 19s", "remaining_time": "2h 25m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516433}
{"loss": 23.625, "grad_norm": 1.14543891, "learning_rate": 0.00039641, "token_acc": 0.39200705, "epoch": 0.57975003, "global_step/max_steps": "1264/2181", "elapsed_time": "3h 20m 29s", "remaining_time": "2h 25m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516529}
{"loss": 23.875, "grad_norm": 1.2914809, "learning_rate": 0.00039568, "token_acc": 0.37015782, "epoch": 0.58020869, "global_step/max_steps": "1265/2181", "elapsed_time": "3h 20m 38s", "remaining_time": "2h 25m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.515961}
{"loss": 24.375, "grad_norm": 1.32948184, "learning_rate": 0.00039495, "token_acc": 0.37610375, "epoch": 0.58066735, "global_step/max_steps": "1266/2181", "elapsed_time": "3h 20m 49s", "remaining_time": "2h 25m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.517042}
{"loss": 24.25, "grad_norm": 1.15816057, "learning_rate": 0.00039423, "token_acc": 0.38262527, "epoch": 0.58112602, "global_step/max_steps": "1267/2181", "elapsed_time": "3h 20m 58s", "remaining_time": "2h 24m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.517024}
{"loss": 24.25, "grad_norm": 1.16360414, "learning_rate": 0.0003935, "token_acc": 0.38955265, "epoch": 0.58158468, "global_step/max_steps": "1268/2181", "elapsed_time": "3h 21m 7s", "remaining_time": "2h 24m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.5168}
{"loss": 23.625, "grad_norm": 1.15169597, "learning_rate": 0.00039278, "token_acc": 0.40173724, "epoch": 0.58204334, "global_step/max_steps": "1269/2181", "elapsed_time": "3h 21m 17s", "remaining_time": "2h 24m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516765}
{"loss": 23.125, "grad_norm": 1.1181618, "learning_rate": 0.00039205, "token_acc": 0.38570602, "epoch": 0.58250201, "global_step/max_steps": "1270/2181", "elapsed_time": "3h 21m 26s", "remaining_time": "2h 24m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516726}
{"loss": 24.125, "grad_norm": 1.17909884, "learning_rate": 0.00039133, "token_acc": 0.38230337, "epoch": 0.58296067, "global_step/max_steps": "1271/2181", "elapsed_time": "3h 21m 36s", "remaining_time": "2h 24m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516702}
{"loss": 23.75, "grad_norm": 1.15285075, "learning_rate": 0.0003906, "token_acc": 0.37680764, "epoch": 0.58341933, "global_step/max_steps": "1272/2181", "elapsed_time": "3h 21m 45s", "remaining_time": "2h 24m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516518}
{"loss": 24.5, "grad_norm": 1.16483879, "learning_rate": 0.00038988, "token_acc": 0.39250276, "epoch": 0.583878, "global_step/max_steps": "1273/2181", "elapsed_time": "3h 21m 55s", "remaining_time": "2h 24m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.517092}
{"loss": 24.0, "grad_norm": 1.05397618, "learning_rate": 0.00038915, "token_acc": 0.38013319, "epoch": 0.58433666, "global_step/max_steps": "1274/2181", "elapsed_time": "3h 22m 4s", "remaining_time": "2h 23m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516767}
{"loss": 24.5, "grad_norm": 1.18668318, "learning_rate": 0.00038843, "token_acc": 0.38569343, "epoch": 0.58479532, "global_step/max_steps": "1275/2181", "elapsed_time": "3h 22m 14s", "remaining_time": "2h 23m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.516957}
{"loss": 23.875, "grad_norm": 1.12150109, "learning_rate": 0.00038771, "token_acc": 0.38590604, "epoch": 0.58525398, "global_step/max_steps": "1276/2181", "elapsed_time": "3h 22m 24s", "remaining_time": "2h 23m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.517335}
{"loss": 24.25, "grad_norm": 1.13511837, "learning_rate": 0.00038698, "token_acc": 0.38814318, "epoch": 0.58571265, "global_step/max_steps": "1277/2181", "elapsed_time": "3h 22m 34s", "remaining_time": "2h 23m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.517687}
{"loss": 24.25, "grad_norm": 1.34327674, "learning_rate": 0.00038626, "token_acc": 0.38929839, "epoch": 0.58617131, "global_step/max_steps": "1278/2181", "elapsed_time": "3h 22m 43s", "remaining_time": "2h 23m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.517395}
{"loss": 23.75, "grad_norm": 1.17927206, "learning_rate": 0.00038554, "token_acc": 0.3721316, "epoch": 0.58662997, "global_step/max_steps": "1279/2181", "elapsed_time": "3h 22m 53s", "remaining_time": "2h 23m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.517702}
{"loss": 23.75, "grad_norm": 1.06072235, "learning_rate": 0.00038481, "token_acc": 0.38470006, "epoch": 0.58708864, "global_step/max_steps": "1280/2181", "elapsed_time": "3h 23m 2s", "remaining_time": "2h 22m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.517307}
{"loss": 23.875, "grad_norm": 1.20626593, "learning_rate": 0.00038409, "token_acc": 0.38270176, "epoch": 0.5875473, "global_step/max_steps": "1281/2181", "elapsed_time": "3h 23m 12s", "remaining_time": "2h 22m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.517234}
{"loss": 24.25, "grad_norm": 1.0931021, "learning_rate": 0.00038337, "token_acc": 0.38817265, "epoch": 0.58800596, "global_step/max_steps": "1282/2181", "elapsed_time": "3h 23m 22s", "remaining_time": "2h 22m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.517915}
{"loss": 23.5, "grad_norm": 1.21949947, "learning_rate": 0.00038265, "token_acc": 0.39077533, "epoch": 0.58846463, "global_step/max_steps": "1283/2181", "elapsed_time": "3h 23m 32s", "remaining_time": "2h 22m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.518139}
{"loss": 24.0, "grad_norm": 1.13180566, "learning_rate": 0.00038192, "token_acc": 0.39271818, "epoch": 0.58892329, "global_step/max_steps": "1284/2181", "elapsed_time": "3h 23m 41s", "remaining_time": "2h 22m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.517739}
{"loss": 24.25, "grad_norm": 1.10013688, "learning_rate": 0.0003812, "token_acc": 0.38384417, "epoch": 0.58938195, "global_step/max_steps": "1285/2181", "elapsed_time": "3h 23m 50s", "remaining_time": "2h 22m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.517854}
{"loss": 24.25, "grad_norm": 1.08334541, "learning_rate": 0.00038048, "token_acc": 0.38008801, "epoch": 0.58984061, "global_step/max_steps": "1286/2181", "elapsed_time": "3h 24m 1s", "remaining_time": "2h 21m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.518301}
{"loss": 24.125, "grad_norm": 1.31310213, "learning_rate": 0.00037976, "token_acc": 0.38167939, "epoch": 0.59029928, "global_step/max_steps": "1287/2181", "elapsed_time": "3h 24m 10s", "remaining_time": "2h 21m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.517955}
{"loss": 23.625, "grad_norm": 1.12170923, "learning_rate": 0.00037904, "token_acc": 0.4059322, "epoch": 0.59075794, "global_step/max_steps": "1288/2181", "elapsed_time": "3h 24m 19s", "remaining_time": "2h 21m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.517708}
{"loss": 23.25, "grad_norm": 1.06162679, "learning_rate": 0.00037832, "token_acc": 0.41033435, "epoch": 0.5912166, "global_step/max_steps": "1289/2181", "elapsed_time": "3h 24m 29s", "remaining_time": "2h 21m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.518292}
{"loss": 23.125, "grad_norm": 1.14724433, "learning_rate": 0.0003776, "token_acc": 0.39706712, "epoch": 0.59167527, "global_step/max_steps": "1290/2181", "elapsed_time": "3h 24m 39s", "remaining_time": "2h 21m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.518752}
{"loss": 23.625, "grad_norm": 1.32439506, "learning_rate": 0.00037688, "token_acc": 0.39241949, "epoch": 0.59213393, "global_step/max_steps": "1291/2181", "elapsed_time": "3h 24m 49s", "remaining_time": "2h 21m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.51867}
{"loss": 24.625, "grad_norm": 1.12298727, "learning_rate": 0.00037616, "token_acc": 0.36630344, "epoch": 0.59259259, "global_step/max_steps": "1292/2181", "elapsed_time": "3h 24m 58s", "remaining_time": "2h 21m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.518832}
{"loss": 24.125, "grad_norm": 1.03426337, "learning_rate": 0.00037544, "token_acc": 0.37070211, "epoch": 0.59305126, "global_step/max_steps": "1293/2181", "elapsed_time": "3h 25m 8s", "remaining_time": "2h 20m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.518797}
{"loss": 23.625, "grad_norm": 1.22238445, "learning_rate": 0.00037472, "token_acc": 0.39145497, "epoch": 0.59350992, "global_step/max_steps": "1294/2181", "elapsed_time": "3h 25m 17s", "remaining_time": "2h 20m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.518549}
{"loss": 24.0, "grad_norm": 1.12589371, "learning_rate": 0.000374, "token_acc": 0.39436219, "epoch": 0.59396858, "global_step/max_steps": "1295/2181", "elapsed_time": "3h 25m 27s", "remaining_time": "2h 20m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.518974}
{"loss": 24.0, "grad_norm": 1.11613452, "learning_rate": 0.00037328, "token_acc": 0.37809496, "epoch": 0.59442724, "global_step/max_steps": "1296/2181", "elapsed_time": "3h 25m 37s", "remaining_time": "2h 20m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.519267}
{"loss": 23.75, "grad_norm": 1.14617729, "learning_rate": 0.00037256, "token_acc": 0.38765154, "epoch": 0.59488591, "global_step/max_steps": "1297/2181", "elapsed_time": "3h 25m 47s", "remaining_time": "2h 20m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.519299}
{"loss": 23.75, "grad_norm": 1.06136549, "learning_rate": 0.00037185, "token_acc": 0.39058013, "epoch": 0.59534457, "global_step/max_steps": "1298/2181", "elapsed_time": "3h 25m 56s", "remaining_time": "2h 20m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.519617}
{"loss": 24.25, "grad_norm": 1.19659555, "learning_rate": 0.00037113, "token_acc": 0.38154543, "epoch": 0.59580323, "global_step/max_steps": "1299/2181", "elapsed_time": "3h 26m 6s", "remaining_time": "2h 19m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.519562}
{"loss": 24.25, "grad_norm": 1.12303317, "learning_rate": 0.00037041, "token_acc": 0.38313526, "epoch": 0.5962619, "global_step/max_steps": "1300/2181", "elapsed_time": "3h 26m 15s", "remaining_time": "2h 19m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.519509}
{"loss": 24.75, "grad_norm": 1.1030283, "learning_rate": 0.00036969, "token_acc": 0.36430032, "epoch": 0.59672056, "global_step/max_steps": "1301/2181", "elapsed_time": "3h 26m 25s", "remaining_time": "2h 19m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.519281}
{"loss": 23.625, "grad_norm": 1.29539537, "learning_rate": 0.00036898, "token_acc": 0.40303968, "epoch": 0.59717922, "global_step/max_steps": "1302/2181", "elapsed_time": "3h 26m 35s", "remaining_time": "2h 19m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.519698}
{"loss": 24.25, "grad_norm": 1.20657194, "learning_rate": 0.00036826, "token_acc": 0.39702517, "epoch": 0.59763789, "global_step/max_steps": "1303/2181", "elapsed_time": "3h 26m 44s", "remaining_time": "2h 19m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.519571}
{"loss": 23.0, "grad_norm": 1.30042684, "learning_rate": 0.00036754, "token_acc": 0.40328416, "epoch": 0.59809655, "global_step/max_steps": "1304/2181", "elapsed_time": "3h 26m 54s", "remaining_time": "2h 19m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.519638}
{"loss": 23.5, "grad_norm": 1.12837172, "learning_rate": 0.00036683, "token_acc": 0.40178571, "epoch": 0.59855521, "global_step/max_steps": "1305/2181", "elapsed_time": "3h 27m 4s", "remaining_time": "2h 18m 60s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.520024}
{"loss": 23.5, "grad_norm": 1.23407543, "learning_rate": 0.00036611, "token_acc": 0.38996901, "epoch": 0.59901387, "global_step/max_steps": "1306/2181", "elapsed_time": "3h 27m 13s", "remaining_time": "2h 18m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.519982}
{"loss": 24.75, "grad_norm": 1.10471797, "learning_rate": 0.0003654, "token_acc": 0.36821154, "epoch": 0.59947254, "global_step/max_steps": "1307/2181", "elapsed_time": "3h 27m 22s", "remaining_time": "2h 18m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.519533}
{"loss": 23.0, "grad_norm": 1.16803718, "learning_rate": 0.00036468, "token_acc": 0.41369128, "epoch": 0.5999312, "global_step/max_steps": "1308/2181", "elapsed_time": "3h 27m 33s", "remaining_time": "2h 18m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.520266}
{"loss": 23.75, "grad_norm": 1.34741688, "learning_rate": 0.00036397, "token_acc": 0.3974884, "epoch": 0.60038986, "global_step/max_steps": "1309/2181", "elapsed_time": "3h 27m 42s", "remaining_time": "2h 18m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.520259}
{"loss": 24.375, "grad_norm": 1.19294953, "learning_rate": 0.00036325, "token_acc": 0.38807649, "epoch": 0.60084853, "global_step/max_steps": "1310/2181", "elapsed_time": "3h 27m 52s", "remaining_time": "2h 18m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.520589}
{"loss": 24.0, "grad_norm": 1.24853516, "learning_rate": 0.00036254, "token_acc": 0.38826243, "epoch": 0.60130719, "global_step/max_steps": "1311/2181", "elapsed_time": "3h 28m 3s", "remaining_time": "2h 18m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.521438}
{"loss": 23.875, "grad_norm": 1.20135403, "learning_rate": 0.00036182, "token_acc": 0.38157895, "epoch": 0.60176585, "global_step/max_steps": "1312/2181", "elapsed_time": "3h 28m 13s", "remaining_time": "2h 17m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.521973}
{"loss": 24.0, "grad_norm": 1.23040736, "learning_rate": 0.00036111, "token_acc": 0.39090649, "epoch": 0.60222452, "global_step/max_steps": "1313/2181", "elapsed_time": "3h 28m 22s", "remaining_time": "2h 17m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.522024}
{"loss": 24.125, "grad_norm": 1.16745114, "learning_rate": 0.0003604, "token_acc": 0.39314286, "epoch": 0.60268318, "global_step/max_steps": "1314/2181", "elapsed_time": "3h 28m 32s", "remaining_time": "2h 17m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.521758}
{"loss": 23.875, "grad_norm": 1.11343646, "learning_rate": 0.00035968, "token_acc": 0.39635158, "epoch": 0.60314184, "global_step/max_steps": "1315/2181", "elapsed_time": "3h 28m 42s", "remaining_time": "2h 17m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52215}
{"loss": 25.0, "grad_norm": 1.14848685, "learning_rate": 0.00035897, "token_acc": 0.37143664, "epoch": 0.6036005, "global_step/max_steps": "1316/2181", "elapsed_time": "3h 28m 51s", "remaining_time": "2h 17m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.521944}
{"loss": 24.0, "grad_norm": 1.24771583, "learning_rate": 0.00035826, "token_acc": 0.38268398, "epoch": 0.60405917, "global_step/max_steps": "1317/2181", "elapsed_time": "3h 29m 1s", "remaining_time": "2h 17m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.522342}
{"loss": 24.125, "grad_norm": 1.04401648, "learning_rate": 0.00035755, "token_acc": 0.37923437, "epoch": 0.60451783, "global_step/max_steps": "1318/2181", "elapsed_time": "3h 29m 11s", "remaining_time": "2h 16m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.522783}
{"loss": 24.0, "grad_norm": 1.20634604, "learning_rate": 0.00035684, "token_acc": 0.38109133, "epoch": 0.60497649, "global_step/max_steps": "1319/2181", "elapsed_time": "3h 29m 20s", "remaining_time": "2h 16m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.522695}
{"loss": 23.875, "grad_norm": 1.23008919, "learning_rate": 0.00035612, "token_acc": 0.37230265, "epoch": 0.60543516, "global_step/max_steps": "1320/2181", "elapsed_time": "3h 29m 30s", "remaining_time": "2h 16m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52248}
{"loss": 24.375, "grad_norm": 1.05746949, "learning_rate": 0.00035541, "token_acc": 0.365625, "epoch": 0.60589382, "global_step/max_steps": "1321/2181", "elapsed_time": "3h 29m 40s", "remaining_time": "2h 16m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523093}
{"loss": 24.125, "grad_norm": 1.12755549, "learning_rate": 0.0003547, "token_acc": 0.37905935, "epoch": 0.60635248, "global_step/max_steps": "1322/2181", "elapsed_time": "3h 29m 50s", "remaining_time": "2h 16m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523245}
{"loss": 23.625, "grad_norm": 1.07087934, "learning_rate": 0.00035399, "token_acc": 0.38857143, "epoch": 0.60681115, "global_step/max_steps": "1323/2181", "elapsed_time": "3h 29m 59s", "remaining_time": "2h 16m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523189}
{"loss": 23.375, "grad_norm": 1.19925761, "learning_rate": 0.00035328, "token_acc": 0.40406743, "epoch": 0.60726981, "global_step/max_steps": "1324/2181", "elapsed_time": "3h 30m 9s", "remaining_time": "2h 16m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523481}
{"loss": 24.125, "grad_norm": 1.34051645, "learning_rate": 0.00035257, "token_acc": 0.37471847, "epoch": 0.60772847, "global_step/max_steps": "1325/2181", "elapsed_time": "3h 30m 18s", "remaining_time": "2h 15m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523351}
{"loss": 23.625, "grad_norm": 1.12437057, "learning_rate": 0.00035186, "token_acc": 0.40382514, "epoch": 0.60818713, "global_step/max_steps": "1326/2181", "elapsed_time": "3h 30m 28s", "remaining_time": "2h 15m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523112}
{"loss": 22.875, "grad_norm": 1.13464975, "learning_rate": 0.00035115, "token_acc": 0.39885057, "epoch": 0.6086458, "global_step/max_steps": "1327/2181", "elapsed_time": "3h 30m 38s", "remaining_time": "2h 15m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523458}
{"loss": 24.0, "grad_norm": 1.26820982, "learning_rate": 0.00035044, "token_acc": 0.38616071, "epoch": 0.60910446, "global_step/max_steps": "1328/2181", "elapsed_time": "3h 30m 48s", "remaining_time": "2h 15m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524058}
{"loss": 23.5, "grad_norm": 1.20775366, "learning_rate": 0.00034974, "token_acc": 0.39977286, "epoch": 0.60956312, "global_step/max_steps": "1329/2181", "elapsed_time": "3h 30m 57s", "remaining_time": "2h 15m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523837}
{"loss": 23.25, "grad_norm": 1.10181725, "learning_rate": 0.00034903, "token_acc": 0.39555429, "epoch": 0.61002179, "global_step/max_steps": "1330/2181", "elapsed_time": "3h 31m 7s", "remaining_time": "2h 15m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523824}
{"loss": 24.375, "grad_norm": 1.18674982, "learning_rate": 0.00034832, "token_acc": 0.38256933, "epoch": 0.61048045, "global_step/max_steps": "1331/2181", "elapsed_time": "3h 31m 17s", "remaining_time": "2h 14m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524143}
{"loss": 24.125, "grad_norm": 1.23578787, "learning_rate": 0.00034761, "token_acc": 0.38470486, "epoch": 0.61093911, "global_step/max_steps": "1332/2181", "elapsed_time": "3h 31m 26s", "remaining_time": "2h 14m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523727}
{"loss": 24.125, "grad_norm": 1.25196373, "learning_rate": 0.0003469, "token_acc": 0.39257981, "epoch": 0.61139778, "global_step/max_steps": "1333/2181", "elapsed_time": "3h 31m 35s", "remaining_time": "2h 14m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523928}
{"loss": 23.875, "grad_norm": 1.01790059, "learning_rate": 0.0003462, "token_acc": 0.39145907, "epoch": 0.61185644, "global_step/max_steps": "1334/2181", "elapsed_time": "3h 31m 46s", "remaining_time": "2h 14m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524678}
{"loss": 23.625, "grad_norm": 1.3632437, "learning_rate": 0.00034549, "token_acc": 0.39534223, "epoch": 0.6123151, "global_step/max_steps": "1335/2181", "elapsed_time": "3h 31m 55s", "remaining_time": "2h 14m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524224}
{"loss": 23.625, "grad_norm": 1.18279874, "learning_rate": 0.00034479, "token_acc": 0.395964, "epoch": 0.61277376, "global_step/max_steps": "1336/2181", "elapsed_time": "3h 32m 5s", "remaining_time": "2h 14m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524449}
{"loss": 24.125, "grad_norm": 1.17234457, "learning_rate": 0.00034408, "token_acc": 0.38757396, "epoch": 0.61323243, "global_step/max_steps": "1337/2181", "elapsed_time": "3h 32m 14s", "remaining_time": "2h 13m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524566}
{"loss": 23.5, "grad_norm": 1.12708747, "learning_rate": 0.00034337, "token_acc": 0.39696373, "epoch": 0.61369109, "global_step/max_steps": "1338/2181", "elapsed_time": "3h 32m 24s", "remaining_time": "2h 13m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524544}
{"loss": 24.125, "grad_norm": 1.15951347, "learning_rate": 0.00034267, "token_acc": 0.39123196, "epoch": 0.61414975, "global_step/max_steps": "1339/2181", "elapsed_time": "3h 32m 34s", "remaining_time": "2h 13m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524807}
{"loss": 23.75, "grad_norm": 1.15078676, "learning_rate": 0.00034196, "token_acc": 0.39479575, "epoch": 0.61460842, "global_step/max_steps": "1340/2181", "elapsed_time": "3h 32m 43s", "remaining_time": "2h 13m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524755}
{"loss": 23.125, "grad_norm": 1.07043326, "learning_rate": 0.00034126, "token_acc": 0.39218443, "epoch": 0.61506708, "global_step/max_steps": "1341/2181", "elapsed_time": "3h 32m 53s", "remaining_time": "2h 13m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525117}
{"loss": 24.25, "grad_norm": 1.10913002, "learning_rate": 0.00034056, "token_acc": 0.38803557, "epoch": 0.61552574, "global_step/max_steps": "1342/2181", "elapsed_time": "3h 33m 2s", "remaining_time": "2h 13m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524709}
{"loss": 24.25, "grad_norm": 1.12931859, "learning_rate": 0.00033985, "token_acc": 0.38239265, "epoch": 0.61598441, "global_step/max_steps": "1343/2181", "elapsed_time": "3h 33m 12s", "remaining_time": "2h 13m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525166}
{"loss": 23.875, "grad_norm": 1.02042091, "learning_rate": 0.00033915, "token_acc": 0.38055872, "epoch": 0.61644307, "global_step/max_steps": "1344/2181", "elapsed_time": "3h 33m 22s", "remaining_time": "2h 12m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525608}
{"loss": 24.0, "grad_norm": 1.24257064, "learning_rate": 0.00033845, "token_acc": 0.38682106, "epoch": 0.61690173, "global_step/max_steps": "1345/2181", "elapsed_time": "3h 33m 31s", "remaining_time": "2h 12m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525274}
{"loss": 24.0, "grad_norm": 1.14564633, "learning_rate": 0.00033774, "token_acc": 0.37022187, "epoch": 0.61736039, "global_step/max_steps": "1346/2181", "elapsed_time": "3h 33m 41s", "remaining_time": "2h 12m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524982}
{"loss": 23.5, "grad_norm": 1.1486752, "learning_rate": 0.00033704, "token_acc": 0.40540541, "epoch": 0.61781906, "global_step/max_steps": "1347/2181", "elapsed_time": "3h 33m 50s", "remaining_time": "2h 12m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525235}
{"loss": 23.75, "grad_norm": 1.07775104, "learning_rate": 0.00033634, "token_acc": 0.39654162, "epoch": 0.61827772, "global_step/max_steps": "1348/2181", "elapsed_time": "3h 34m 1s", "remaining_time": "2h 12m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525858}
{"loss": 23.75, "grad_norm": 1.13902211, "learning_rate": 0.00033564, "token_acc": 0.38792867, "epoch": 0.61873638, "global_step/max_steps": "1349/2181", "elapsed_time": "3h 34m 10s", "remaining_time": "2h 12m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525493}
{"loss": 24.25, "grad_norm": 1.03494883, "learning_rate": 0.00033494, "token_acc": 0.39484622, "epoch": 0.61919505, "global_step/max_steps": "1350/2181", "elapsed_time": "3h 34m 20s", "remaining_time": "2h 11m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525621}
{"loss": 23.5, "grad_norm": 1.15037036, "learning_rate": 0.00033423, "token_acc": 0.38715795, "epoch": 0.61965371, "global_step/max_steps": "1351/2181", "elapsed_time": "3h 34m 29s", "remaining_time": "2h 11m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525672}
{"loss": 23.875, "grad_norm": 1.2323277, "learning_rate": 0.00033353, "token_acc": 0.38476672, "epoch": 0.62011237, "global_step/max_steps": "1352/2181", "elapsed_time": "3h 34m 39s", "remaining_time": "2h 11m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525791}
{"loss": 23.625, "grad_norm": 0.9731223, "learning_rate": 0.00033283, "token_acc": 0.39628314, "epoch": 0.62057104, "global_step/max_steps": "1353/2181", "elapsed_time": "3h 34m 48s", "remaining_time": "2h 11m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525736}
{"loss": 24.25, "grad_norm": 1.15950513, "learning_rate": 0.00033213, "token_acc": 0.3878295, "epoch": 0.6210297, "global_step/max_steps": "1354/2181", "elapsed_time": "3h 34m 59s", "remaining_time": "2h 11m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526233}
{"loss": 23.125, "grad_norm": 1.13727725, "learning_rate": 0.00033143, "token_acc": 0.39404795, "epoch": 0.62148836, "global_step/max_steps": "1355/2181", "elapsed_time": "3h 35m 8s", "remaining_time": "2h 11m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525955}
{"loss": 24.875, "grad_norm": 1.26149523, "learning_rate": 0.00033074, "token_acc": 0.36976744, "epoch": 0.62194702, "global_step/max_steps": "1356/2181", "elapsed_time": "3h 35m 17s", "remaining_time": "2h 10m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525702}
{"loss": 23.75, "grad_norm": 1.11508822, "learning_rate": 0.00033004, "token_acc": 0.39186947, "epoch": 0.62240569, "global_step/max_steps": "1357/2181", "elapsed_time": "3h 35m 28s", "remaining_time": "2h 10m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52665}
{"loss": 23.875, "grad_norm": 1.0421387, "learning_rate": 0.00032934, "token_acc": 0.39021018, "epoch": 0.62286435, "global_step/max_steps": "1358/2181", "elapsed_time": "3h 35m 37s", "remaining_time": "2h 10m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526818}
{"loss": 24.125, "grad_norm": 1.03243041, "learning_rate": 0.00032864, "token_acc": 0.38123085, "epoch": 0.62332301, "global_step/max_steps": "1359/2181", "elapsed_time": "3h 35m 47s", "remaining_time": "2h 10m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52693}
{"loss": 24.0, "grad_norm": 1.17370498, "learning_rate": 0.00032794, "token_acc": 0.39683933, "epoch": 0.62378168, "global_step/max_steps": "1360/2181", "elapsed_time": "3h 35m 57s", "remaining_time": "2h 10m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52708}
{"loss": 23.5, "grad_norm": 1.09317136, "learning_rate": 0.00032725, "token_acc": 0.39495565, "epoch": 0.62424034, "global_step/max_steps": "1361/2181", "elapsed_time": "3h 36m 6s", "remaining_time": "2h 10m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526605}
{"loss": 24.0, "grad_norm": 1.20954645, "learning_rate": 0.00032655, "token_acc": 0.37741124, "epoch": 0.624699, "global_step/max_steps": "1362/2181", "elapsed_time": "3h 36m 15s", "remaining_time": "2h 10m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526506}
{"loss": 23.625, "grad_norm": 1.15231454, "learning_rate": 0.00032585, "token_acc": 0.38204127, "epoch": 0.62515767, "global_step/max_steps": "1363/2181", "elapsed_time": "3h 36m 25s", "remaining_time": "2h 9m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527002}
{"loss": 23.75, "grad_norm": 1.09363818, "learning_rate": 0.00032516, "token_acc": 0.38763736, "epoch": 0.62561633, "global_step/max_steps": "1364/2181", "elapsed_time": "3h 36m 35s", "remaining_time": "2h 9m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526839}
{"loss": 23.25, "grad_norm": 1.09285629, "learning_rate": 0.00032446, "token_acc": 0.39649816, "epoch": 0.62607499, "global_step/max_steps": "1365/2181", "elapsed_time": "3h 36m 44s", "remaining_time": "2h 9m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527023}
{"loss": 23.0, "grad_norm": 1.14682138, "learning_rate": 0.00032377, "token_acc": 0.40363327, "epoch": 0.62653365, "global_step/max_steps": "1366/2181", "elapsed_time": "3h 36m 54s", "remaining_time": "2h 9m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526985}
{"loss": 23.875, "grad_norm": 1.18793404, "learning_rate": 0.00032307, "token_acc": 0.38344828, "epoch": 0.62699232, "global_step/max_steps": "1367/2181", "elapsed_time": "3h 37m 3s", "remaining_time": "2h 9m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526637}
{"loss": 23.625, "grad_norm": 1.1210711, "learning_rate": 0.00032238, "token_acc": 0.38794306, "epoch": 0.62745098, "global_step/max_steps": "1368/2181", "elapsed_time": "3h 37m 12s", "remaining_time": "2h 9m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526488}
{"loss": 24.0, "grad_norm": 1.15183401, "learning_rate": 0.00032168, "token_acc": 0.3823043, "epoch": 0.62790964, "global_step/max_steps": "1369/2181", "elapsed_time": "3h 37m 22s", "remaining_time": "2h 8m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52676}
{"loss": 23.625, "grad_norm": 1.8147403, "learning_rate": 0.00032099, "token_acc": 0.37782805, "epoch": 0.62836831, "global_step/max_steps": "1370/2181", "elapsed_time": "3h 37m 32s", "remaining_time": "2h 8m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526828}
{"loss": 24.0, "grad_norm": 1.15623367, "learning_rate": 0.0003203, "token_acc": 0.37972936, "epoch": 0.62882697, "global_step/max_steps": "1371/2181", "elapsed_time": "3h 37m 41s", "remaining_time": "2h 8m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526364}
{"loss": 24.0, "grad_norm": 1.09592462, "learning_rate": 0.0003196, "token_acc": 0.38353863, "epoch": 0.62928563, "global_step/max_steps": "1372/2181", "elapsed_time": "3h 37m 51s", "remaining_time": "2h 8m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526759}
{"loss": 23.875, "grad_norm": 1.10376239, "learning_rate": 0.00031891, "token_acc": 0.39347357, "epoch": 0.6297443, "global_step/max_steps": "1373/2181", "elapsed_time": "3h 38m 0s", "remaining_time": "2h 8m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526919}
{"loss": 23.5, "grad_norm": 1.17862296, "learning_rate": 0.00031822, "token_acc": 0.39570017, "epoch": 0.63020296, "global_step/max_steps": "1374/2181", "elapsed_time": "3h 38m 9s", "remaining_time": "2h 8m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526355}
{"loss": 24.25, "grad_norm": 1.24836087, "learning_rate": 0.00031753, "token_acc": 0.37936963, "epoch": 0.63066162, "global_step/max_steps": "1375/2181", "elapsed_time": "3h 38m 18s", "remaining_time": "2h 7m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525979}
{"loss": 23.75, "grad_norm": 1.15879893, "learning_rate": 0.00031684, "token_acc": 0.39053254, "epoch": 0.63112028, "global_step/max_steps": "1376/2181", "elapsed_time": "3h 38m 28s", "remaining_time": "2h 7m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52647}
{"loss": 23.75, "grad_norm": 1.07364953, "learning_rate": 0.00031614, "token_acc": 0.40817485, "epoch": 0.63157895, "global_step/max_steps": "1377/2181", "elapsed_time": "3h 38m 37s", "remaining_time": "2h 7m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526113}
{"loss": 24.625, "grad_norm": 1.15273309, "learning_rate": 0.00031545, "token_acc": 0.38770424, "epoch": 0.63203761, "global_step/max_steps": "1378/2181", "elapsed_time": "3h 38m 46s", "remaining_time": "2h 7m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525558}
{"loss": 24.375, "grad_norm": 1.13189399, "learning_rate": 0.00031476, "token_acc": 0.37326678, "epoch": 0.63249627, "global_step/max_steps": "1379/2181", "elapsed_time": "3h 38m 57s", "remaining_time": "2h 7m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526431}
{"loss": 23.5, "grad_norm": 1.29981899, "learning_rate": 0.00031407, "token_acc": 0.39240131, "epoch": 0.63295494, "global_step/max_steps": "1380/2181", "elapsed_time": "3h 39m 6s", "remaining_time": "2h 7m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526336}
{"loss": 24.875, "grad_norm": 1.31610906, "learning_rate": 0.00031339, "token_acc": 0.36305198, "epoch": 0.6334136, "global_step/max_steps": "1381/2181", "elapsed_time": "3h 39m 15s", "remaining_time": "2h 7m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525872}
{"loss": 23.75, "grad_norm": 1.24826014, "learning_rate": 0.0003127, "token_acc": 0.38206897, "epoch": 0.63387226, "global_step/max_steps": "1382/2181", "elapsed_time": "3h 39m 25s", "remaining_time": "2h 6m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526192}
{"loss": 23.5, "grad_norm": 1.06857884, "learning_rate": 0.00031201, "token_acc": 0.38638206, "epoch": 0.63433093, "global_step/max_steps": "1383/2181", "elapsed_time": "3h 39m 35s", "remaining_time": "2h 6m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526553}
{"loss": 23.875, "grad_norm": 1.09935367, "learning_rate": 0.00031132, "token_acc": 0.3886202, "epoch": 0.63478959, "global_step/max_steps": "1384/2181", "elapsed_time": "3h 39m 44s", "remaining_time": "2h 6m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526307}
{"loss": 23.875, "grad_norm": 1.45730436, "learning_rate": 0.00031063, "token_acc": 0.39207537, "epoch": 0.63524825, "global_step/max_steps": "1385/2181", "elapsed_time": "3h 39m 54s", "remaining_time": "2h 6m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526611}
{"loss": 23.875, "grad_norm": 1.0793196, "learning_rate": 0.00030995, "token_acc": 0.39288668, "epoch": 0.63570691, "global_step/max_steps": "1386/2181", "elapsed_time": "3h 40m 4s", "remaining_time": "2h 6m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526505}
{"loss": 25.0, "grad_norm": 1.131814, "learning_rate": 0.00030926, "token_acc": 0.36518865, "epoch": 0.63616558, "global_step/max_steps": "1387/2181", "elapsed_time": "3h 40m 13s", "remaining_time": "2h 6m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526013}
{"loss": 23.875, "grad_norm": 1.15450346, "learning_rate": 0.00030857, "token_acc": 0.38250617, "epoch": 0.63662424, "global_step/max_steps": "1388/2181", "elapsed_time": "3h 40m 22s", "remaining_time": "2h 5m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525748}
{"loss": 24.5, "grad_norm": 1.10220909, "learning_rate": 0.00030789, "token_acc": 0.38345659, "epoch": 0.6370829, "global_step/max_steps": "1389/2181", "elapsed_time": "3h 40m 31s", "remaining_time": "2h 5m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525837}
{"loss": 24.25, "grad_norm": 1.1151309, "learning_rate": 0.0003072, "token_acc": 0.38772702, "epoch": 0.63754157, "global_step/max_steps": "1390/2181", "elapsed_time": "3h 40m 41s", "remaining_time": "2h 5m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525673}
{"loss": 23.625, "grad_norm": 1.07939291, "learning_rate": 0.00030652, "token_acc": 0.39792293, "epoch": 0.63800023, "global_step/max_steps": "1391/2181", "elapsed_time": "3h 40m 50s", "remaining_time": "2h 5m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525441}
{"loss": 23.5, "grad_norm": 1.114205, "learning_rate": 0.00030583, "token_acc": 0.40407139, "epoch": 0.63845889, "global_step/max_steps": "1392/2181", "elapsed_time": "3h 40m 60s", "remaining_time": "2h 5m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52584}
{"loss": 24.125, "grad_norm": 1.12501383, "learning_rate": 0.00030515, "token_acc": 0.39671949, "epoch": 0.63891756, "global_step/max_steps": "1393/2181", "elapsed_time": "3h 41m 9s", "remaining_time": "2h 5m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525775}
{"loss": 23.5, "grad_norm": 1.03502595, "learning_rate": 0.00030446, "token_acc": 0.3919598, "epoch": 0.63937622, "global_step/max_steps": "1394/2181", "elapsed_time": "3h 41m 19s", "remaining_time": "2h 4m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52604}
{"loss": 24.25, "grad_norm": 1.19422519, "learning_rate": 0.00030378, "token_acc": 0.37084765, "epoch": 0.63983488, "global_step/max_steps": "1395/2181", "elapsed_time": "3h 41m 29s", "remaining_time": "2h 4m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526084}
{"loss": 24.125, "grad_norm": 1.13296652, "learning_rate": 0.0003031, "token_acc": 0.38330976, "epoch": 0.64029354, "global_step/max_steps": "1396/2181", "elapsed_time": "3h 41m 39s", "remaining_time": "2h 4m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526186}
{"loss": 23.125, "grad_norm": 1.04960048, "learning_rate": 0.00030241, "token_acc": 0.39263278, "epoch": 0.64075221, "global_step/max_steps": "1397/2181", "elapsed_time": "3h 41m 47s", "remaining_time": "2h 4m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525659}
{"loss": 24.125, "grad_norm": 1.08801651, "learning_rate": 0.00030173, "token_acc": 0.37808989, "epoch": 0.64121087, "global_step/max_steps": "1398/2181", "elapsed_time": "3h 41m 57s", "remaining_time": "2h 4m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525999}
{"loss": 23.5, "grad_norm": 1.14354336, "learning_rate": 0.00030105, "token_acc": 0.3842684, "epoch": 0.64166953, "global_step/max_steps": "1399/2181", "elapsed_time": "3h 42m 8s", "remaining_time": "2h 4m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526743}
{"loss": 23.5, "grad_norm": 1.16332698, "learning_rate": 0.00030037, "token_acc": 0.37864612, "epoch": 0.6421282, "global_step/max_steps": "1400/2181", "elapsed_time": "3h 42m 16s", "remaining_time": "2h 3m 60s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525977}
{"loss": 24.25, "grad_norm": 1.25021434, "learning_rate": 0.00029969, "token_acc": 0.37503407, "epoch": 0.64258686, "global_step/max_steps": "1401/2181", "elapsed_time": "3h 42m 26s", "remaining_time": "2h 3m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526235}
{"loss": 23.625, "grad_norm": 1.05008757, "learning_rate": 0.00029901, "token_acc": 0.39487757, "epoch": 0.64304552, "global_step/max_steps": "1402/2181", "elapsed_time": "3h 42m 36s", "remaining_time": "2h 3m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526472}
{"loss": 24.375, "grad_norm": 1.12443948, "learning_rate": 0.00029833, "token_acc": 0.38626251, "epoch": 0.64350419, "global_step/max_steps": "1403/2181", "elapsed_time": "3h 42m 45s", "remaining_time": "2h 3m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526212}
{"loss": 23.75, "grad_norm": 1.08897829, "learning_rate": 0.00029765, "token_acc": 0.3986376, "epoch": 0.64396285, "global_step/max_steps": "1404/2181", "elapsed_time": "3h 42m 55s", "remaining_time": "2h 3m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526217}
{"loss": 24.125, "grad_norm": 1.19950271, "learning_rate": 0.00029697, "token_acc": 0.38980716, "epoch": 0.64442151, "global_step/max_steps": "1405/2181", "elapsed_time": "3h 43m 5s", "remaining_time": "2h 3m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526609}
{"loss": 23.25, "grad_norm": 1.02435303, "learning_rate": 0.00029629, "token_acc": 0.3992074, "epoch": 0.64488017, "global_step/max_steps": "1406/2181", "elapsed_time": "3h 43m 14s", "remaining_time": "2h 3m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526134}
{"loss": 23.125, "grad_norm": 1.06254315, "learning_rate": 0.00029561, "token_acc": 0.41547519, "epoch": 0.64533884, "global_step/max_steps": "1407/2181", "elapsed_time": "3h 43m 23s", "remaining_time": "2h 2m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526085}
{"loss": 24.0, "grad_norm": 1.12217271, "learning_rate": 0.00029494, "token_acc": 0.3720339, "epoch": 0.6457975, "global_step/max_steps": "1408/2181", "elapsed_time": "3h 43m 34s", "remaining_time": "2h 2m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526762}
{"loss": 24.5, "grad_norm": 1.12016416, "learning_rate": 0.00029426, "token_acc": 0.37471132, "epoch": 0.64625616, "global_step/max_steps": "1409/2181", "elapsed_time": "3h 43m 44s", "remaining_time": "2h 2m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527223}
{"loss": 23.5, "grad_norm": 1.30389106, "learning_rate": 0.00029358, "token_acc": 0.38684885, "epoch": 0.64671483, "global_step/max_steps": "1410/2181", "elapsed_time": "3h 43m 52s", "remaining_time": "2h 2m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52659}
{"loss": 24.25, "grad_norm": 1.16110396, "learning_rate": 0.00029291, "token_acc": 0.38297872, "epoch": 0.64717349, "global_step/max_steps": "1411/2181", "elapsed_time": "3h 44m 2s", "remaining_time": "2h 2m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526806}
{"loss": 24.0, "grad_norm": 1.00279415, "learning_rate": 0.00029223, "token_acc": 0.39591949, "epoch": 0.64763215, "global_step/max_steps": "1412/2181", "elapsed_time": "3h 44m 12s", "remaining_time": "2h 2m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527114}
{"loss": 24.125, "grad_norm": 1.12571907, "learning_rate": 0.00029156, "token_acc": 0.38430927, "epoch": 0.64809082, "global_step/max_steps": "1413/2181", "elapsed_time": "3h 44m 21s", "remaining_time": "2h 1m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526594}
{"loss": 23.5, "grad_norm": 1.1399405, "learning_rate": 0.00029088, "token_acc": 0.39949324, "epoch": 0.64854948, "global_step/max_steps": "1414/2181", "elapsed_time": "3h 44m 31s", "remaining_time": "2h 1m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526777}
{"loss": 23.875, "grad_norm": 1.11440611, "learning_rate": 0.00029021, "token_acc": 0.38221709, "epoch": 0.64900814, "global_step/max_steps": "1415/2181", "elapsed_time": "3h 44m 41s", "remaining_time": "2h 1m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527158}
{"loss": 23.5, "grad_norm": 1.1230756, "learning_rate": 0.00028953, "token_acc": 0.39550562, "epoch": 0.6494668, "global_step/max_steps": "1416/2181", "elapsed_time": "3h 44m 51s", "remaining_time": "2h 1m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527339}
{"loss": 23.75, "grad_norm": 1.06060243, "learning_rate": 0.00028886, "token_acc": 0.3770942, "epoch": 0.64992547, "global_step/max_steps": "1417/2181", "elapsed_time": "3h 45m 0s", "remaining_time": "2h 1m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527199}
{"loss": 24.0, "grad_norm": 1.08496714, "learning_rate": 0.00028819, "token_acc": 0.37952314, "epoch": 0.65038413, "global_step/max_steps": "1418/2181", "elapsed_time": "3h 45m 10s", "remaining_time": "2h 1m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527418}
{"loss": 23.5, "grad_norm": 1.15138328, "learning_rate": 0.00028751, "token_acc": 0.38910724, "epoch": 0.65084279, "global_step/max_steps": "1419/2181", "elapsed_time": "3h 45m 19s", "remaining_time": "2h 0m 60s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527068}
{"loss": 23.75, "grad_norm": 1.18528223, "learning_rate": 0.00028684, "token_acc": 0.37953795, "epoch": 0.65130146, "global_step/max_steps": "1420/2181", "elapsed_time": "3h 45m 28s", "remaining_time": "2h 0m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52692}
{"loss": 24.25, "grad_norm": 1.18682325, "learning_rate": 0.00028617, "token_acc": 0.37957497, "epoch": 0.65176012, "global_step/max_steps": "1421/2181", "elapsed_time": "3h 45m 39s", "remaining_time": "2h 0m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527516}
{"loss": 23.875, "grad_norm": 1.18777716, "learning_rate": 0.0002855, "token_acc": 0.39609484, "epoch": 0.65221878, "global_step/max_steps": "1422/2181", "elapsed_time": "3h 45m 48s", "remaining_time": "2h 0m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527396}
{"loss": 23.875, "grad_norm": 1.14807272, "learning_rate": 0.00028483, "token_acc": 0.38656633, "epoch": 0.65267745, "global_step/max_steps": "1423/2181", "elapsed_time": "3h 45m 57s", "remaining_time": "2h 0m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527255}
{"loss": 24.75, "grad_norm": 1.12563169, "learning_rate": 0.00028416, "token_acc": 0.37856743, "epoch": 0.65313611, "global_step/max_steps": "1424/2181", "elapsed_time": "3h 46m 7s", "remaining_time": "2h 0m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527156}
{"loss": 23.75, "grad_norm": 1.14283276, "learning_rate": 0.00028349, "token_acc": 0.3893087, "epoch": 0.65359477, "global_step/max_steps": "1425/2181", "elapsed_time": "3h 46m 16s", "remaining_time": "2h 0m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527366}
{"loss": 24.0, "grad_norm": 1.0607878, "learning_rate": 0.00028282, "token_acc": 0.3840539, "epoch": 0.65405343, "global_step/max_steps": "1426/2181", "elapsed_time": "3h 46m 25s", "remaining_time": "1h 59m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52688}
{"loss": 23.875, "grad_norm": 1.28814435, "learning_rate": 0.00028215, "token_acc": 0.38856508, "epoch": 0.6545121, "global_step/max_steps": "1427/2181", "elapsed_time": "3h 46m 35s", "remaining_time": "1h 59m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526971}
{"loss": 24.125, "grad_norm": 1.11851692, "learning_rate": 0.00028148, "token_acc": 0.39148581, "epoch": 0.65497076, "global_step/max_steps": "1428/2181", "elapsed_time": "3h 46m 45s", "remaining_time": "1h 59m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527193}
{"loss": 23.625, "grad_norm": 1.08183956, "learning_rate": 0.00028081, "token_acc": 0.38435195, "epoch": 0.65542942, "global_step/max_steps": "1429/2181", "elapsed_time": "3h 46m 55s", "remaining_time": "1h 59m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527406}
{"loss": 22.875, "grad_norm": 1.1373359, "learning_rate": 0.00028015, "token_acc": 0.40420097, "epoch": 0.65588809, "global_step/max_steps": "1430/2181", "elapsed_time": "3h 47m 4s", "remaining_time": "1h 59m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527232}
{"loss": 24.5, "grad_norm": 1.11305439, "learning_rate": 0.00027948, "token_acc": 0.38096567, "epoch": 0.65634675, "global_step/max_steps": "1431/2181", "elapsed_time": "3h 47m 14s", "remaining_time": "1h 59m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.5277}
{"loss": 24.375, "grad_norm": 1.16082847, "learning_rate": 0.00027881, "token_acc": 0.37618403, "epoch": 0.65680541, "global_step/max_steps": "1432/2181", "elapsed_time": "3h 47m 24s", "remaining_time": "1h 58m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527767}
{"loss": 23.875, "grad_norm": 1.05912209, "learning_rate": 0.00027815, "token_acc": 0.4002287, "epoch": 0.65726408, "global_step/max_steps": "1433/2181", "elapsed_time": "3h 47m 33s", "remaining_time": "1h 58m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527321}
{"loss": 24.0, "grad_norm": 1.0285424, "learning_rate": 0.00027748, "token_acc": 0.38132081, "epoch": 0.65772274, "global_step/max_steps": "1434/2181", "elapsed_time": "3h 47m 42s", "remaining_time": "1h 58m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527544}
{"loss": 25.0, "grad_norm": 1.17884362, "learning_rate": 0.00027682, "token_acc": 0.36610644, "epoch": 0.6581814, "global_step/max_steps": "1435/2181", "elapsed_time": "3h 47m 52s", "remaining_time": "1h 58m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527275}
{"loss": 24.0, "grad_norm": 1.19228554, "learning_rate": 0.00027615, "token_acc": 0.37656904, "epoch": 0.65864006, "global_step/max_steps": "1436/2181", "elapsed_time": "3h 48m 1s", "remaining_time": "1h 58m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527056}
{"loss": 23.75, "grad_norm": 1.02434146, "learning_rate": 0.00027549, "token_acc": 0.39018889, "epoch": 0.65909873, "global_step/max_steps": "1437/2181", "elapsed_time": "3h 48m 11s", "remaining_time": "1h 58m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527243}
{"loss": 24.0, "grad_norm": 1.18260348, "learning_rate": 0.00027483, "token_acc": 0.38378378, "epoch": 0.65955739, "global_step/max_steps": "1438/2181", "elapsed_time": "3h 48m 20s", "remaining_time": "1h 57m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.527278}
{"loss": 24.75, "grad_norm": 1.08147526, "learning_rate": 0.00027416, "token_acc": 0.37835594, "epoch": 0.66001605, "global_step/max_steps": "1439/2181", "elapsed_time": "3h 48m 29s", "remaining_time": "1h 57m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526835}
{"loss": 23.5, "grad_norm": 1.05940914, "learning_rate": 0.0002735, "token_acc": 0.40048544, "epoch": 0.66047472, "global_step/max_steps": "1440/2181", "elapsed_time": "3h 48m 39s", "remaining_time": "1h 57m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526765}
{"loss": 24.75, "grad_norm": 1.2454865, "learning_rate": 0.00027284, "token_acc": 0.37808989, "epoch": 0.66093338, "global_step/max_steps": "1441/2181", "elapsed_time": "3h 48m 47s", "remaining_time": "1h 57m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526342}
{"loss": 23.875, "grad_norm": 1.21303475, "learning_rate": 0.00027218, "token_acc": 0.38868458, "epoch": 0.66139204, "global_step/max_steps": "1442/2181", "elapsed_time": "3h 48m 57s", "remaining_time": "1h 57m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526036}
{"loss": 23.875, "grad_norm": 1.22040129, "learning_rate": 0.00027152, "token_acc": 0.38557559, "epoch": 0.66185071, "global_step/max_steps": "1443/2181", "elapsed_time": "3h 49m 5s", "remaining_time": "1h 57m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52562}
{"loss": 23.5, "grad_norm": 1.10563302, "learning_rate": 0.00027086, "token_acc": 0.3982808, "epoch": 0.66230937, "global_step/max_steps": "1444/2181", "elapsed_time": "3h 49m 16s", "remaining_time": "1h 57m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526125}
{"loss": 23.75, "grad_norm": 1.03840232, "learning_rate": 0.0002702, "token_acc": 0.39073569, "epoch": 0.66276803, "global_step/max_steps": "1445/2181", "elapsed_time": "3h 49m 25s", "remaining_time": "1h 56m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525951}
{"loss": 24.0, "grad_norm": 1.07352698, "learning_rate": 0.00026954, "token_acc": 0.37726524, "epoch": 0.66322669, "global_step/max_steps": "1446/2181", "elapsed_time": "3h 49m 34s", "remaining_time": "1h 56m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525858}
{"loss": 24.75, "grad_norm": 1.12913227, "learning_rate": 0.00026888, "token_acc": 0.37615191, "epoch": 0.66368536, "global_step/max_steps": "1447/2181", "elapsed_time": "3h 49m 44s", "remaining_time": "1h 56m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52578}
{"loss": 23.75, "grad_norm": 1.16288912, "learning_rate": 0.00026822, "token_acc": 0.3841497, "epoch": 0.66414402, "global_step/max_steps": "1448/2181", "elapsed_time": "3h 49m 53s", "remaining_time": "1h 56m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525696}
{"loss": 23.875, "grad_norm": 1.07966697, "learning_rate": 0.00026756, "token_acc": 0.38361831, "epoch": 0.66460268, "global_step/max_steps": "1449/2181", "elapsed_time": "3h 50m 3s", "remaining_time": "1h 56m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525642}
{"loss": 23.375, "grad_norm": 1.27655232, "learning_rate": 0.00026691, "token_acc": 0.4027111, "epoch": 0.66506135, "global_step/max_steps": "1450/2181", "elapsed_time": "3h 50m 13s", "remaining_time": "1h 56m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52616}
{"loss": 23.25, "grad_norm": 1.26025212, "learning_rate": 0.00026625, "token_acc": 0.38543417, "epoch": 0.66552001, "global_step/max_steps": "1451/2181", "elapsed_time": "3h 50m 23s", "remaining_time": "1h 55m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.526191}
{"loss": 24.0, "grad_norm": 1.1033951, "learning_rate": 0.00026559, "token_acc": 0.38606717, "epoch": 0.66597867, "global_step/max_steps": "1452/2181", "elapsed_time": "3h 50m 32s", "remaining_time": "1h 55m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525976}
{"loss": 24.625, "grad_norm": 1.18533111, "learning_rate": 0.00026494, "token_acc": 0.36301758, "epoch": 0.66643734, "global_step/max_steps": "1453/2181", "elapsed_time": "3h 50m 41s", "remaining_time": "1h 55m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525953}
{"loss": 24.0, "grad_norm": 1.23247159, "learning_rate": 0.00026428, "token_acc": 0.37686462, "epoch": 0.666896, "global_step/max_steps": "1454/2181", "elapsed_time": "3h 50m 51s", "remaining_time": "1h 55m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525937}
{"loss": 24.0, "grad_norm": 1.18284881, "learning_rate": 0.00026363, "token_acc": 0.39395575, "epoch": 0.66735466, "global_step/max_steps": "1455/2181", "elapsed_time": "3h 51m 0s", "remaining_time": "1h 55m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525821}
{"loss": 22.875, "grad_norm": 1.19333124, "learning_rate": 0.00026297, "token_acc": 0.39620472, "epoch": 0.66781332, "global_step/max_steps": "1456/2181", "elapsed_time": "3h 51m 9s", "remaining_time": "1h 55m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525676}
{"loss": 24.25, "grad_norm": 1.24242687, "learning_rate": 0.00026232, "token_acc": 0.3867121, "epoch": 0.66827199, "global_step/max_steps": "1457/2181", "elapsed_time": "3h 51m 19s", "remaining_time": "1h 54m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52575}
{"loss": 23.25, "grad_norm": 1.27845609, "learning_rate": 0.00026167, "token_acc": 0.39735099, "epoch": 0.66873065, "global_step/max_steps": "1458/2181", "elapsed_time": "3h 51m 28s", "remaining_time": "1h 54m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525214}
{"loss": 24.5, "grad_norm": 1.1681397, "learning_rate": 0.00026101, "token_acc": 0.38162038, "epoch": 0.66918931, "global_step/max_steps": "1459/2181", "elapsed_time": "3h 51m 37s", "remaining_time": "1h 54m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525148}
{"loss": 23.5, "grad_norm": 1.14153504, "learning_rate": 0.00026036, "token_acc": 0.39753294, "epoch": 0.66964798, "global_step/max_steps": "1460/2181", "elapsed_time": "3h 51m 47s", "remaining_time": "1h 54m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525244}
{"loss": 24.25, "grad_norm": 1.17588937, "learning_rate": 0.00025971, "token_acc": 0.37884512, "epoch": 0.67010664, "global_step/max_steps": "1461/2181", "elapsed_time": "3h 51m 56s", "remaining_time": "1h 54m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524683}
{"loss": 23.125, "grad_norm": 1.14484727, "learning_rate": 0.00025906, "token_acc": 0.39595619, "epoch": 0.6705653, "global_step/max_steps": "1462/2181", "elapsed_time": "3h 52m 5s", "remaining_time": "1h 54m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524824}
{"loss": 24.125, "grad_norm": 1.05594945, "learning_rate": 0.00025841, "token_acc": 0.3754162, "epoch": 0.67102397, "global_step/max_steps": "1463/2181", "elapsed_time": "3h 52m 15s", "remaining_time": "1h 53m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524638}
{"loss": 24.625, "grad_norm": 1.19645023, "learning_rate": 0.00025776, "token_acc": 0.38330976, "epoch": 0.67148263, "global_step/max_steps": "1464/2181", "elapsed_time": "3h 52m 24s", "remaining_time": "1h 53m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524338}
{"loss": 25.0, "grad_norm": 1.18794072, "learning_rate": 0.00025711, "token_acc": 0.3574352, "epoch": 0.67194129, "global_step/max_steps": "1465/2181", "elapsed_time": "3h 52m 34s", "remaining_time": "1h 53m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524632}
{"loss": 24.25, "grad_norm": 0.97205782, "learning_rate": 0.00025646, "token_acc": 0.38167939, "epoch": 0.67239995, "global_step/max_steps": "1466/2181", "elapsed_time": "3h 52m 44s", "remaining_time": "1h 53m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524922}
{"loss": 25.0, "grad_norm": 1.12398767, "learning_rate": 0.00025581, "token_acc": 0.36713666, "epoch": 0.67285862, "global_step/max_steps": "1467/2181", "elapsed_time": "3h 52m 53s", "remaining_time": "1h 53m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524846}
{"loss": 24.0, "grad_norm": 1.05900931, "learning_rate": 0.00025516, "token_acc": 0.38220038, "epoch": 0.67331728, "global_step/max_steps": "1468/2181", "elapsed_time": "3h 53m 3s", "remaining_time": "1h 53m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525184}
{"loss": 24.0, "grad_norm": 1.05521643, "learning_rate": 0.00025452, "token_acc": 0.37906725, "epoch": 0.67377594, "global_step/max_steps": "1469/2181", "elapsed_time": "3h 53m 13s", "remaining_time": "1h 53m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525212}
{"loss": 24.125, "grad_norm": 1.13711834, "learning_rate": 0.00025387, "token_acc": 0.37927202, "epoch": 0.67423461, "global_step/max_steps": "1470/2181", "elapsed_time": "3h 53m 22s", "remaining_time": "1h 52m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525499}
{"loss": 23.75, "grad_norm": 1.05158699, "learning_rate": 0.00025322, "token_acc": 0.38441781, "epoch": 0.67469327, "global_step/max_steps": "1471/2181", "elapsed_time": "3h 53m 31s", "remaining_time": "1h 52m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524922}
{"loss": 24.75, "grad_norm": 1.27154016, "learning_rate": 0.00025258, "token_acc": 0.37544338, "epoch": 0.67515193, "global_step/max_steps": "1472/2181", "elapsed_time": "3h 53m 41s", "remaining_time": "1h 52m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525038}
{"loss": 23.125, "grad_norm": 1.12216198, "learning_rate": 0.00025193, "token_acc": 0.39517937, "epoch": 0.6756106, "global_step/max_steps": "1473/2181", "elapsed_time": "3h 53m 51s", "remaining_time": "1h 52m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525611}
{"loss": 23.375, "grad_norm": 1.07719946, "learning_rate": 0.00025129, "token_acc": 0.37531346, "epoch": 0.67606926, "global_step/max_steps": "1474/2181", "elapsed_time": "3h 54m 0s", "remaining_time": "1h 52m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.525338}
{"loss": 23.875, "grad_norm": 1.21313429, "learning_rate": 0.00025064, "token_acc": 0.38585859, "epoch": 0.67652792, "global_step/max_steps": "1475/2181", "elapsed_time": "3h 54m 9s", "remaining_time": "1h 52m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524861}
{"loss": 23.0, "grad_norm": 1.04114676, "learning_rate": 0.00025, "token_acc": 0.39620566, "epoch": 0.67698658, "global_step/max_steps": "1476/2181", "elapsed_time": "3h 54m 19s", "remaining_time": "1h 51m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524944}
{"loss": 23.5, "grad_norm": 1.05384505, "learning_rate": 0.00024936, "token_acc": 0.3836056, "epoch": 0.67744525, "global_step/max_steps": "1477/2181", "elapsed_time": "3h 54m 28s", "remaining_time": "1h 51m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524557}
{"loss": 24.625, "grad_norm": 1.16572011, "learning_rate": 0.00024871, "token_acc": 0.36574746, "epoch": 0.67790391, "global_step/max_steps": "1478/2181", "elapsed_time": "3h 54m 37s", "remaining_time": "1h 51m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524608}
{"loss": 24.25, "grad_norm": 1.08743596, "learning_rate": 0.00024807, "token_acc": 0.38879457, "epoch": 0.67836257, "global_step/max_steps": "1479/2181", "elapsed_time": "3h 54m 47s", "remaining_time": "1h 51m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524676}
{"loss": 23.0, "grad_norm": 1.09845889, "learning_rate": 0.00024743, "token_acc": 0.41568852, "epoch": 0.67882124, "global_step/max_steps": "1480/2181", "elapsed_time": "3h 54m 56s", "remaining_time": "1h 51m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524255}
{"loss": 24.125, "grad_norm": 1.13227034, "learning_rate": 0.00024679, "token_acc": 0.38263666, "epoch": 0.6792799, "global_step/max_steps": "1481/2181", "elapsed_time": "3h 55m 5s", "remaining_time": "1h 51m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523965}
{"loss": 24.125, "grad_norm": 1.15753913, "learning_rate": 0.00024615, "token_acc": 0.38627451, "epoch": 0.67973856, "global_step/max_steps": "1482/2181", "elapsed_time": "3h 55m 14s", "remaining_time": "1h 50m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523875}
{"loss": 24.125, "grad_norm": 1.02595174, "learning_rate": 0.00024551, "token_acc": 0.38326955, "epoch": 0.68019723, "global_step/max_steps": "1483/2181", "elapsed_time": "3h 55m 24s", "remaining_time": "1h 50m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52376}
{"loss": 23.625, "grad_norm": 0.98370588, "learning_rate": 0.00024487, "token_acc": 0.37204059, "epoch": 0.68065589, "global_step/max_steps": "1484/2181", "elapsed_time": "3h 55m 34s", "remaining_time": "1h 50m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524286}
{"loss": 23.75, "grad_norm": 1.12157393, "learning_rate": 0.00024423, "token_acc": 0.40762621, "epoch": 0.68111455, "global_step/max_steps": "1485/2181", "elapsed_time": "3h 55m 43s", "remaining_time": "1h 50m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524014}
{"loss": 23.125, "grad_norm": 1.10892344, "learning_rate": 0.0002436, "token_acc": 0.40090472, "epoch": 0.68157321, "global_step/max_steps": "1486/2181", "elapsed_time": "3h 55m 53s", "remaining_time": "1h 50m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523947}
{"loss": 23.875, "grad_norm": 1.15242851, "learning_rate": 0.00024296, "token_acc": 0.38254665, "epoch": 0.68203188, "global_step/max_steps": "1487/2181", "elapsed_time": "3h 56m 2s", "remaining_time": "1h 50m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523804}
{"loss": 24.0, "grad_norm": 1.0210079, "learning_rate": 0.00024232, "token_acc": 0.38313526, "epoch": 0.68249054, "global_step/max_steps": "1488/2181", "elapsed_time": "3h 56m 12s", "remaining_time": "1h 50m 0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524077}
{"loss": 24.25, "grad_norm": 1.07490504, "learning_rate": 0.00024169, "token_acc": 0.38389923, "epoch": 0.6829492, "global_step/max_steps": "1489/2181", "elapsed_time": "3h 56m 22s", "remaining_time": "1h 49m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524215}
{"loss": 23.25, "grad_norm": 1.03555655, "learning_rate": 0.00024105, "token_acc": 0.39709724, "epoch": 0.68340787, "global_step/max_steps": "1490/2181", "elapsed_time": "3h 56m 31s", "remaining_time": "1h 49m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524332}
{"loss": 23.875, "grad_norm": 1.1493901, "learning_rate": 0.00024041, "token_acc": 0.37732704, "epoch": 0.68386653, "global_step/max_steps": "1491/2181", "elapsed_time": "3h 56m 41s", "remaining_time": "1h 49m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52429}
{"loss": 24.375, "grad_norm": 1.2176528, "learning_rate": 0.00023978, "token_acc": 0.37037037, "epoch": 0.68432519, "global_step/max_steps": "1492/2181", "elapsed_time": "3h 56m 50s", "remaining_time": "1h 49m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524223}
{"loss": 23.875, "grad_norm": 1.08034062, "learning_rate": 0.00023915, "token_acc": 0.38439636, "epoch": 0.68478386, "global_step/max_steps": "1493/2181", "elapsed_time": "3h 56m 59s", "remaining_time": "1h 49m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523902}
{"loss": 23.375, "grad_norm": 1.13312912, "learning_rate": 0.00023851, "token_acc": 0.39004598, "epoch": 0.68524252, "global_step/max_steps": "1494/2181", "elapsed_time": "3h 57m 9s", "remaining_time": "1h 49m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.523931}
{"loss": 23.375, "grad_norm": 1.13955545, "learning_rate": 0.00023788, "token_acc": 0.3911187, "epoch": 0.68570118, "global_step/max_steps": "1495/2181", "elapsed_time": "3h 57m 19s", "remaining_time": "1h 48m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524168}
{"loss": 23.0, "grad_norm": 1.21053159, "learning_rate": 0.00023725, "token_acc": 0.39239412, "epoch": 0.68615984, "global_step/max_steps": "1496/2181", "elapsed_time": "3h 57m 28s", "remaining_time": "1h 48m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524031}
{"loss": 22.75, "grad_norm": 1.02420378, "learning_rate": 0.00023662, "token_acc": 0.40370059, "epoch": 0.68661851, "global_step/max_steps": "1497/2181", "elapsed_time": "3h 57m 37s", "remaining_time": "1h 48m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524004}
{"loss": 25.0, "grad_norm": 1.06621385, "learning_rate": 0.00023599, "token_acc": 0.3635383, "epoch": 0.68707717, "global_step/max_steps": "1498/2181", "elapsed_time": "3h 57m 47s", "remaining_time": "1h 48m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.52381}
{"loss": 23.75, "grad_norm": 1.37480152, "learning_rate": 0.00023536, "token_acc": 0.39243949, "epoch": 0.68753583, "global_step/max_steps": "1499/2181", "elapsed_time": "3h 57m 57s", "remaining_time": "1h 48m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524148}
{"loss": 25.125, "grad_norm": 0.96767789, "learning_rate": 0.00023473, "token_acc": 0.36409845, "epoch": 0.6879945, "global_step/max_steps": "1500/2181", "elapsed_time": "3h 58m 6s", "remaining_time": "1h 48m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.524254}
{"loss": 23.625, "grad_norm": 1.23469877, "learning_rate": 0.0002341, "token_acc": 0.37699317, "epoch": 0.68845316, "global_step/max_steps": "1501/2181", "elapsed_time": "3h 58m 37s", "remaining_time": "1h 48m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538172}
{"loss": 23.5, "grad_norm": 1.0454036, "learning_rate": 0.00023347, "token_acc": 0.3918659, "epoch": 0.68891182, "global_step/max_steps": "1502/2181", "elapsed_time": "3h 58m 47s", "remaining_time": "1h 47m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.53882}
{"loss": 23.625, "grad_norm": 1.06807363, "learning_rate": 0.00023284, "token_acc": 0.38777841, "epoch": 0.68937049, "global_step/max_steps": "1503/2181", "elapsed_time": "3h 58m 57s", "remaining_time": "1h 47m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538817}
{"loss": 23.875, "grad_norm": 1.17182434, "learning_rate": 0.00023221, "token_acc": 0.37812062, "epoch": 0.68982915, "global_step/max_steps": "1504/2181", "elapsed_time": "3h 59m 6s", "remaining_time": "1h 47m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538805}
{"loss": 24.625, "grad_norm": 1.02629817, "learning_rate": 0.00023159, "token_acc": 0.37120794, "epoch": 0.69028781, "global_step/max_steps": "1505/2181", "elapsed_time": "3h 59m 16s", "remaining_time": "1h 47m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538844}
{"loss": 23.875, "grad_norm": 1.11708295, "learning_rate": 0.00023096, "token_acc": 0.39314059, "epoch": 0.69074647, "global_step/max_steps": "1506/2181", "elapsed_time": "3h 59m 25s", "remaining_time": "1h 47m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538606}
{"loss": 24.25, "grad_norm": 1.0511694, "learning_rate": 0.00023033, "token_acc": 0.37538036, "epoch": 0.69120514, "global_step/max_steps": "1507/2181", "elapsed_time": "3h 59m 34s", "remaining_time": "1h 47m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538305}
{"loss": 23.5, "grad_norm": 1.06564963, "learning_rate": 0.00022971, "token_acc": 0.38756537, "epoch": 0.6916638, "global_step/max_steps": "1508/2181", "elapsed_time": "3h 59m 44s", "remaining_time": "1h 46m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538244}
{"loss": 23.75, "grad_norm": 1.03589654, "learning_rate": 0.00022908, "token_acc": 0.39859228, "epoch": 0.69212246, "global_step/max_steps": "1509/2181", "elapsed_time": "3h 59m 54s", "remaining_time": "1h 46m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538457}
{"loss": 24.0, "grad_norm": 1.14107752, "learning_rate": 0.00022846, "token_acc": 0.38754228, "epoch": 0.69258113, "global_step/max_steps": "1510/2181", "elapsed_time": "4h 0m 3s", "remaining_time": "1h 46m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538455}
{"loss": 24.5, "grad_norm": 1.1765908, "learning_rate": 0.00022784, "token_acc": 0.38659058, "epoch": 0.69303979, "global_step/max_steps": "1511/2181", "elapsed_time": "4h 0m 13s", "remaining_time": "1h 46m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538828}
{"loss": 23.625, "grad_norm": 1.05052829, "learning_rate": 0.00022721, "token_acc": 0.39492754, "epoch": 0.69349845, "global_step/max_steps": "1512/2181", "elapsed_time": "4h 0m 22s", "remaining_time": "1h 46m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538631}
{"loss": 24.5, "grad_norm": 1.14992869, "learning_rate": 0.00022659, "token_acc": 0.37651225, "epoch": 0.69395712, "global_step/max_steps": "1513/2181", "elapsed_time": "4h 0m 32s", "remaining_time": "1h 46m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538776}
{"loss": 25.0, "grad_norm": 1.02156293, "learning_rate": 0.00022597, "token_acc": 0.36761488, "epoch": 0.69441578, "global_step/max_steps": "1514/2181", "elapsed_time": "4h 0m 41s", "remaining_time": "1h 46m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538571}
{"loss": 23.5, "grad_norm": 1.09675562, "learning_rate": 0.00022535, "token_acc": 0.39085587, "epoch": 0.69487444, "global_step/max_steps": "1515/2181", "elapsed_time": "4h 0m 51s", "remaining_time": "1h 45m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538831}
{"loss": 24.5, "grad_norm": 1.09535122, "learning_rate": 0.00022473, "token_acc": 0.37950293, "epoch": 0.6953331, "global_step/max_steps": "1516/2181", "elapsed_time": "4h 1m 0s", "remaining_time": "1h 45m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538473}
{"loss": 24.125, "grad_norm": 1.01589572, "learning_rate": 0.00022411, "token_acc": 0.38476672, "epoch": 0.69579177, "global_step/max_steps": "1517/2181", "elapsed_time": "4h 1m 10s", "remaining_time": "1h 45m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538543}
{"loss": 24.5, "grad_norm": 1.05874646, "learning_rate": 0.00022349, "token_acc": 0.38078673, "epoch": 0.69625043, "global_step/max_steps": "1518/2181", "elapsed_time": "4h 1m 19s", "remaining_time": "1h 45m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538493}
{"loss": 23.75, "grad_norm": 1.06382132, "learning_rate": 0.00022287, "token_acc": 0.38091119, "epoch": 0.69670909, "global_step/max_steps": "1519/2181", "elapsed_time": "4h 1m 29s", "remaining_time": "1h 45m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538551}
{"loss": 23.75, "grad_norm": 1.09327686, "learning_rate": 0.00022225, "token_acc": 0.38862292, "epoch": 0.69716776, "global_step/max_steps": "1520/2181", "elapsed_time": "4h 1m 38s", "remaining_time": "1h 45m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538237}
{"loss": 24.125, "grad_norm": 1.18729448, "learning_rate": 0.00022164, "token_acc": 0.38320066, "epoch": 0.69762642, "global_step/max_steps": "1521/2181", "elapsed_time": "4h 1m 48s", "remaining_time": "1h 44m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538347}
{"loss": 23.75, "grad_norm": 1.10758781, "learning_rate": 0.00022102, "token_acc": 0.38291774, "epoch": 0.69808508, "global_step/max_steps": "1522/2181", "elapsed_time": "4h 1m 56s", "remaining_time": "1h 44m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.537697}
{"loss": 24.375, "grad_norm": 1.17712498, "learning_rate": 0.0002204, "token_acc": 0.37614679, "epoch": 0.69854374, "global_step/max_steps": "1523/2181", "elapsed_time": "4h 2m 6s", "remaining_time": "1h 44m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.537569}
{"loss": 24.25, "grad_norm": 1.0893054, "learning_rate": 0.00021979, "token_acc": 0.37991266, "epoch": 0.69900241, "global_step/max_steps": "1524/2181", "elapsed_time": "4h 2m 16s", "remaining_time": "1h 44m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538237}
{"loss": 24.0, "grad_norm": 1.16069102, "learning_rate": 0.00021917, "token_acc": 0.38923513, "epoch": 0.69946107, "global_step/max_steps": "1525/2181", "elapsed_time": "4h 2m 25s", "remaining_time": "1h 44m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.537979}
{"loss": 23.625, "grad_norm": 1.21040154, "learning_rate": 0.00021856, "token_acc": 0.3936942, "epoch": 0.69991973, "global_step/max_steps": "1526/2181", "elapsed_time": "4h 2m 34s", "remaining_time": "1h 44m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.537519}
{"loss": 23.625, "grad_norm": 1.06988454, "learning_rate": 0.00021795, "token_acc": 0.38654279, "epoch": 0.7003784, "global_step/max_steps": "1527/2181", "elapsed_time": "4h 2m 45s", "remaining_time": "1h 43m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.53812}
{"loss": 23.125, "grad_norm": 1.05715466, "learning_rate": 0.00021733, "token_acc": 0.39801418, "epoch": 0.70083706, "global_step/max_steps": "1528/2181", "elapsed_time": "4h 2m 54s", "remaining_time": "1h 43m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.53786}
{"loss": 24.5, "grad_norm": 1.01286805, "learning_rate": 0.00021672, "token_acc": 0.37890949, "epoch": 0.70129572, "global_step/max_steps": "1529/2181", "elapsed_time": "4h 3m 3s", "remaining_time": "1h 43m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.53793}
{"loss": 24.375, "grad_norm": 1.04287553, "learning_rate": 0.00021611, "token_acc": 0.38858988, "epoch": 0.70175439, "global_step/max_steps": "1530/2181", "elapsed_time": "4h 3m 13s", "remaining_time": "1h 43m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.537842}
{"loss": 24.125, "grad_norm": 1.1257441, "learning_rate": 0.0002155, "token_acc": 0.38221614, "epoch": 0.70221305, "global_step/max_steps": "1531/2181", "elapsed_time": "4h 3m 23s", "remaining_time": "1h 43m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.537997}
{"loss": 23.125, "grad_norm": 1.0967263, "learning_rate": 0.00021489, "token_acc": 0.39618406, "epoch": 0.70267171, "global_step/max_steps": "1532/2181", "elapsed_time": "4h 3m 32s", "remaining_time": "1h 43m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.537755}
{"loss": 24.5, "grad_norm": 1.1351465, "learning_rate": 0.00021428, "token_acc": 0.37637131, "epoch": 0.70313037, "global_step/max_steps": "1533/2181", "elapsed_time": "4h 3m 42s", "remaining_time": "1h 43m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.537875}
{"loss": 24.0, "grad_norm": 1.15383375, "learning_rate": 0.00021367, "token_acc": 0.38478562, "epoch": 0.70358904, "global_step/max_steps": "1534/2181", "elapsed_time": "4h 3m 51s", "remaining_time": "1h 42m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.537828}
{"loss": 24.5, "grad_norm": 1.15107071, "learning_rate": 0.00021306, "token_acc": 0.37876654, "epoch": 0.7040477, "global_step/max_steps": "1535/2181", "elapsed_time": "4h 4m 1s", "remaining_time": "1h 42m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.537863}
{"loss": 24.25, "grad_norm": 1.23959625, "learning_rate": 0.00021245, "token_acc": 0.37945946, "epoch": 0.70450636, "global_step/max_steps": "1536/2181", "elapsed_time": "4h 4m 11s", "remaining_time": "1h 42m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538121}
{"loss": 23.5, "grad_norm": 1.1720885, "learning_rate": 0.00021184, "token_acc": 0.38368247, "epoch": 0.70496503, "global_step/max_steps": "1537/2181", "elapsed_time": "4h 4m 21s", "remaining_time": "1h 42m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538727}
{"loss": 23.375, "grad_norm": 1.14831328, "learning_rate": 0.00021124, "token_acc": 0.40151515, "epoch": 0.70542369, "global_step/max_steps": "1538/2181", "elapsed_time": "4h 4m 30s", "remaining_time": "1h 42m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538233}
{"loss": 24.0, "grad_norm": 1.19233549, "learning_rate": 0.00021063, "token_acc": 0.3803132, "epoch": 0.70588235, "global_step/max_steps": "1539/2181", "elapsed_time": "4h 4m 40s", "remaining_time": "1h 42m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538704}
{"loss": 23.125, "grad_norm": 1.24859762, "learning_rate": 0.00021003, "token_acc": 0.39334086, "epoch": 0.70634102, "global_step/max_steps": "1540/2181", "elapsed_time": "4h 4m 50s", "remaining_time": "1h 41m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.538908}
{"loss": 23.375, "grad_norm": 1.18096948, "learning_rate": 0.00020942, "token_acc": 0.39841763, "epoch": 0.70679968, "global_step/max_steps": "1541/2181", "elapsed_time": "4h 4m 60s", "remaining_time": "1h 41m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.539194}
{"loss": 24.125, "grad_norm": 1.06517768, "learning_rate": 0.00020882, "token_acc": 0.38631958, "epoch": 0.70725834, "global_step/max_steps": "1542/2181", "elapsed_time": "4h 5m 10s", "remaining_time": "1h 41m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.539558}
{"loss": 24.0, "grad_norm": 1.03240705, "learning_rate": 0.00020821, "token_acc": 0.37845228, "epoch": 0.707717, "global_step/max_steps": "1543/2181", "elapsed_time": "4h 5m 20s", "remaining_time": "1h 41m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.539609}
{"loss": 23.375, "grad_norm": 1.07280874, "learning_rate": 0.00020761, "token_acc": 0.3970346, "epoch": 0.70817567, "global_step/max_steps": "1544/2181", "elapsed_time": "4h 5m 29s", "remaining_time": "1h 41m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.539449}
{"loss": 23.5, "grad_norm": 1.04768479, "learning_rate": 0.00020701, "token_acc": 0.38853326, "epoch": 0.70863433, "global_step/max_steps": "1545/2181", "elapsed_time": "4h 5m 39s", "remaining_time": "1h 41m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.539835}
{"loss": 23.625, "grad_norm": 1.16440761, "learning_rate": 0.00020641, "token_acc": 0.39396503, "epoch": 0.70909299, "global_step/max_steps": "1546/2181", "elapsed_time": "4h 5m 49s", "remaining_time": "1h 40m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.539832}
{"loss": 23.875, "grad_norm": 1.0239116, "learning_rate": 0.00020581, "token_acc": 0.38395574, "epoch": 0.70955166, "global_step/max_steps": "1547/2181", "elapsed_time": "4h 5m 58s", "remaining_time": "1h 40m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.53981}
{"loss": 23.0, "grad_norm": 1.04283476, "learning_rate": 0.00020521, "token_acc": 0.40425532, "epoch": 0.71001032, "global_step/max_steps": "1548/2181", "elapsed_time": "4h 6m 8s", "remaining_time": "1h 40m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.539877}
{"loss": 24.75, "grad_norm": 1.06683898, "learning_rate": 0.00020461, "token_acc": 0.37699412, "epoch": 0.71046898, "global_step/max_steps": "1549/2181", "elapsed_time": "4h 6m 18s", "remaining_time": "1h 40m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.540614}
{"loss": 23.5, "grad_norm": 1.10573375, "learning_rate": 0.00020401, "token_acc": 0.38827945, "epoch": 0.71092765, "global_step/max_steps": "1550/2181", "elapsed_time": "4h 6m 29s", "remaining_time": "1h 40m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541013}
{"loss": 24.25, "grad_norm": 1.12151313, "learning_rate": 0.00020341, "token_acc": 0.38248049, "epoch": 0.71138631, "global_step/max_steps": "1551/2181", "elapsed_time": "4h 6m 38s", "remaining_time": "1h 40m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.540931}
{"loss": 24.625, "grad_norm": 1.13596916, "learning_rate": 0.00020281, "token_acc": 0.38479873, "epoch": 0.71184497, "global_step/max_steps": "1552/2181", "elapsed_time": "4h 6m 48s", "remaining_time": "1h 40m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541085}
{"loss": 22.75, "grad_norm": 1.1755482, "learning_rate": 0.00020222, "token_acc": 0.40540541, "epoch": 0.71230363, "global_step/max_steps": "1553/2181", "elapsed_time": "4h 6m 58s", "remaining_time": "1h 39m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541472}
{"loss": 23.625, "grad_norm": 1.08815646, "learning_rate": 0.00020162, "token_acc": 0.39417262, "epoch": 0.7127623, "global_step/max_steps": "1554/2181", "elapsed_time": "4h 7m 7s", "remaining_time": "1h 39m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541384}
{"loss": 24.5, "grad_norm": 1.08493567, "learning_rate": 0.00020102, "token_acc": 0.36475293, "epoch": 0.71322096, "global_step/max_steps": "1555/2181", "elapsed_time": "4h 7m 17s", "remaining_time": "1h 39m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541496}
{"loss": 23.5, "grad_norm": 1.14485335, "learning_rate": 0.00020043, "token_acc": 0.39457831, "epoch": 0.71367962, "global_step/max_steps": "1556/2181", "elapsed_time": "4h 7m 26s", "remaining_time": "1h 39m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541284}
{"loss": 24.25, "grad_norm": 1.14827204, "learning_rate": 0.00019983, "token_acc": 0.375, "epoch": 0.71413829, "global_step/max_steps": "1557/2181", "elapsed_time": "4h 7m 36s", "remaining_time": "1h 39m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541495}
{"loss": 23.75, "grad_norm": 1.38599193, "learning_rate": 0.00019924, "token_acc": 0.38879408, "epoch": 0.71459695, "global_step/max_steps": "1558/2181", "elapsed_time": "4h 7m 46s", "remaining_time": "1h 39m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541593}
{"loss": 24.0, "grad_norm": 1.05920351, "learning_rate": 0.00019865, "token_acc": 0.37819612, "epoch": 0.71505561, "global_step/max_steps": "1559/2181", "elapsed_time": "4h 7m 55s", "remaining_time": "1h 38m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54129}
{"loss": 24.375, "grad_norm": 1.18330014, "learning_rate": 0.00019806, "token_acc": 0.39287717, "epoch": 0.71551428, "global_step/max_steps": "1560/2181", "elapsed_time": "4h 8m 5s", "remaining_time": "1h 38m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541702}
{"loss": 24.875, "grad_norm": 1.19012368, "learning_rate": 0.00019746, "token_acc": 0.38141741, "epoch": 0.71597294, "global_step/max_steps": "1561/2181", "elapsed_time": "4h 8m 14s", "remaining_time": "1h 38m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541059}
{"loss": 23.875, "grad_norm": 1.13871729, "learning_rate": 0.00019687, "token_acc": 0.38922652, "epoch": 0.7164316, "global_step/max_steps": "1562/2181", "elapsed_time": "4h 8m 24s", "remaining_time": "1h 38m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541492}
{"loss": 24.375, "grad_norm": 1.12286937, "learning_rate": 0.00019628, "token_acc": 0.3877203, "epoch": 0.71689026, "global_step/max_steps": "1563/2181", "elapsed_time": "4h 8m 34s", "remaining_time": "1h 38m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542189}
{"loss": 24.0, "grad_norm": 1.09784114, "learning_rate": 0.00019569, "token_acc": 0.39168552, "epoch": 0.71734893, "global_step/max_steps": "1564/2181", "elapsed_time": "4h 8m 44s", "remaining_time": "1h 38m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542006}
{"loss": 23.75, "grad_norm": 1.23631454, "learning_rate": 0.0001951, "token_acc": 0.39248077, "epoch": 0.71780759, "global_step/max_steps": "1565/2181", "elapsed_time": "4h 8m 53s", "remaining_time": "1h 37m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541917}
{"loss": 24.0, "grad_norm": 1.13424933, "learning_rate": 0.00019452, "token_acc": 0.38299712, "epoch": 0.71826625, "global_step/max_steps": "1566/2181", "elapsed_time": "4h 9m 3s", "remaining_time": "1h 37m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542402}
{"loss": 23.125, "grad_norm": 1.10482979, "learning_rate": 0.00019393, "token_acc": 0.3906071, "epoch": 0.71872492, "global_step/max_steps": "1567/2181", "elapsed_time": "4h 9m 13s", "remaining_time": "1h 37m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54222}
{"loss": 23.875, "grad_norm": 1.19530177, "learning_rate": 0.00019334, "token_acc": 0.38606511, "epoch": 0.71918358, "global_step/max_steps": "1568/2181", "elapsed_time": "4h 9m 22s", "remaining_time": "1h 37m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542354}
{"loss": 24.25, "grad_norm": 1.20718551, "learning_rate": 0.00019275, "token_acc": 0.38963902, "epoch": 0.71964224, "global_step/max_steps": "1569/2181", "elapsed_time": "4h 9m 32s", "remaining_time": "1h 37m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54266}
{"loss": 24.5, "grad_norm": 1.19006133, "learning_rate": 0.00019217, "token_acc": 0.38987272, "epoch": 0.72010091, "global_step/max_steps": "1570/2181", "elapsed_time": "4h 9m 42s", "remaining_time": "1h 37m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542666}
{"loss": 24.375, "grad_norm": 1.25485778, "learning_rate": 0.00019158, "token_acc": 0.37740113, "epoch": 0.72055957, "global_step/max_steps": "1571/2181", "elapsed_time": "4h 9m 52s", "remaining_time": "1h 37m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542758}
{"loss": 23.625, "grad_norm": 1.15155649, "learning_rate": 0.000191, "token_acc": 0.39831697, "epoch": 0.72101823, "global_step/max_steps": "1572/2181", "elapsed_time": "4h 10m 1s", "remaining_time": "1h 36m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542924}
{"loss": 23.25, "grad_norm": 1.0610081, "learning_rate": 0.00019042, "token_acc": 0.3941886, "epoch": 0.72147689, "global_step/max_steps": "1573/2181", "elapsed_time": "4h 10m 10s", "remaining_time": "1h 36m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542475}
{"loss": 24.0, "grad_norm": 1.07726979, "learning_rate": 0.00018983, "token_acc": 0.38219178, "epoch": 0.72193556, "global_step/max_steps": "1574/2181", "elapsed_time": "4h 10m 19s", "remaining_time": "1h 36m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542247}
{"loss": 23.5, "grad_norm": 1.06884694, "learning_rate": 0.00018925, "token_acc": 0.38607416, "epoch": 0.72239422, "global_step/max_steps": "1575/2181", "elapsed_time": "4h 10m 30s", "remaining_time": "1h 36m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542579}
{"loss": 24.375, "grad_norm": 1.18044877, "learning_rate": 0.00018867, "token_acc": 0.37837838, "epoch": 0.72285288, "global_step/max_steps": "1576/2181", "elapsed_time": "4h 10m 39s", "remaining_time": "1h 36m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542494}
{"loss": 23.75, "grad_norm": 1.12573743, "learning_rate": 0.00018809, "token_acc": 0.39409142, "epoch": 0.72331155, "global_step/max_steps": "1577/2181", "elapsed_time": "4h 10m 48s", "remaining_time": "1h 36m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542479}
{"loss": 23.625, "grad_norm": 1.08162057, "learning_rate": 0.00018751, "token_acc": 0.38022165, "epoch": 0.72377021, "global_step/max_steps": "1578/2181", "elapsed_time": "4h 10m 58s", "remaining_time": "1h 35m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542696}
{"loss": 23.25, "grad_norm": 1.03186214, "learning_rate": 0.00018693, "token_acc": 0.39162692, "epoch": 0.72422887, "global_step/max_steps": "1579/2181", "elapsed_time": "4h 11m 8s", "remaining_time": "1h 35m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542843}
{"loss": 23.375, "grad_norm": 1.08541369, "learning_rate": 0.00018635, "token_acc": 0.38647474, "epoch": 0.72468754, "global_step/max_steps": "1580/2181", "elapsed_time": "4h 11m 18s", "remaining_time": "1h 35m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543277}
{"loss": 24.125, "grad_norm": 1.14411056, "learning_rate": 0.00018577, "token_acc": 0.38374343, "epoch": 0.7251462, "global_step/max_steps": "1581/2181", "elapsed_time": "4h 11m 27s", "remaining_time": "1h 35m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542974}
{"loss": 24.375, "grad_norm": 1.07576096, "learning_rate": 0.0001852, "token_acc": 0.37597984, "epoch": 0.72560486, "global_step/max_steps": "1582/2181", "elapsed_time": "4h 11m 38s", "remaining_time": "1h 35m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543413}
{"loss": 23.625, "grad_norm": 1.06351018, "learning_rate": 0.00018462, "token_acc": 0.38859962, "epoch": 0.72606352, "global_step/max_steps": "1583/2181", "elapsed_time": "4h 11m 47s", "remaining_time": "1h 35m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543408}
{"loss": 23.625, "grad_norm": 1.20242393, "learning_rate": 0.00018404, "token_acc": 0.38519553, "epoch": 0.72652219, "global_step/max_steps": "1584/2181", "elapsed_time": "4h 11m 57s", "remaining_time": "1h 34m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543782}
{"loss": 24.25, "grad_norm": 1.03802669, "learning_rate": 0.00018347, "token_acc": 0.38240132, "epoch": 0.72698085, "global_step/max_steps": "1585/2181", "elapsed_time": "4h 12m 7s", "remaining_time": "1h 34m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544078}
{"loss": 24.75, "grad_norm": 1.03899896, "learning_rate": 0.00018289, "token_acc": 0.3819978, "epoch": 0.72743951, "global_step/max_steps": "1586/2181", "elapsed_time": "4h 12m 18s", "remaining_time": "1h 34m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544656}
{"loss": 23.125, "grad_norm": 1.17496252, "learning_rate": 0.00018232, "token_acc": 0.41306241, "epoch": 0.72789818, "global_step/max_steps": "1587/2181", "elapsed_time": "4h 12m 28s", "remaining_time": "1h 34m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54495}
{"loss": 23.25, "grad_norm": 1.12891364, "learning_rate": 0.00018175, "token_acc": 0.39849204, "epoch": 0.72835684, "global_step/max_steps": "1588/2181", "elapsed_time": "4h 12m 38s", "remaining_time": "1h 34m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545164}
{"loss": 23.25, "grad_norm": 1.05327761, "learning_rate": 0.00018117, "token_acc": 0.38847532, "epoch": 0.7288155, "global_step/max_steps": "1589/2181", "elapsed_time": "4h 12m 47s", "remaining_time": "1h 34m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544888}
{"loss": 23.75, "grad_norm": 0.97074592, "learning_rate": 0.0001806, "token_acc": 0.38371777, "epoch": 0.72927417, "global_step/max_steps": "1590/2181", "elapsed_time": "4h 12m 56s", "remaining_time": "1h 34m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544562}
{"loss": 23.75, "grad_norm": 1.10351264, "learning_rate": 0.00018003, "token_acc": 0.38551468, "epoch": 0.72973283, "global_step/max_steps": "1591/2181", "elapsed_time": "4h 13m 6s", "remaining_time": "1h 33m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545066}
{"loss": 23.0, "grad_norm": 1.09732032, "learning_rate": 0.00017946, "token_acc": 0.39632107, "epoch": 0.73019149, "global_step/max_steps": "1592/2181", "elapsed_time": "4h 13m 15s", "remaining_time": "1h 33m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544561}
{"loss": 23.25, "grad_norm": 1.0726347, "learning_rate": 0.00017889, "token_acc": 0.38489409, "epoch": 0.73065015, "global_step/max_steps": "1593/2181", "elapsed_time": "4h 13m 24s", "remaining_time": "1h 33m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544272}
{"loss": 23.375, "grad_norm": 1.15470481, "learning_rate": 0.00017832, "token_acc": 0.39159853, "epoch": 0.73110882, "global_step/max_steps": "1594/2181", "elapsed_time": "4h 13m 34s", "remaining_time": "1h 33m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544764}
{"loss": 24.125, "grad_norm": 1.09534812, "learning_rate": 0.00017775, "token_acc": 0.37462069, "epoch": 0.73156748, "global_step/max_steps": "1595/2181", "elapsed_time": "4h 13m 44s", "remaining_time": "1h 33m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544593}
{"loss": 23.5, "grad_norm": 1.06487393, "learning_rate": 0.00017719, "token_acc": 0.39098797, "epoch": 0.73202614, "global_step/max_steps": "1596/2181", "elapsed_time": "4h 13m 54s", "remaining_time": "1h 33m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544874}
{"loss": 23.0, "grad_norm": 1.1256597, "learning_rate": 0.00017662, "token_acc": 0.39788026, "epoch": 0.73248481, "global_step/max_steps": "1597/2181", "elapsed_time": "4h 14m 3s", "remaining_time": "1h 32m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544828}
{"loss": 23.5, "grad_norm": 1.05694628, "learning_rate": 0.00017605, "token_acc": 0.3875035, "epoch": 0.73294347, "global_step/max_steps": "1598/2181", "elapsed_time": "4h 14m 13s", "remaining_time": "1h 32m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544797}
{"loss": 23.125, "grad_norm": 1.12000895, "learning_rate": 0.00017549, "token_acc": 0.39449801, "epoch": 0.73340213, "global_step/max_steps": "1599/2181", "elapsed_time": "4h 14m 22s", "remaining_time": "1h 32m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544527}
{"loss": 23.875, "grad_norm": 1.10593891, "learning_rate": 0.00017492, "token_acc": 0.40125036, "epoch": 0.7338608, "global_step/max_steps": "1600/2181", "elapsed_time": "4h 14m 31s", "remaining_time": "1h 32m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544666}
{"loss": 24.75, "grad_norm": 1.06815171, "learning_rate": 0.00017436, "token_acc": 0.372171, "epoch": 0.73431946, "global_step/max_steps": "1601/2181", "elapsed_time": "4h 14m 42s", "remaining_time": "1h 32m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545232}
{"loss": 23.875, "grad_norm": 1.03696036, "learning_rate": 0.0001738, "token_acc": 0.38517231, "epoch": 0.73477812, "global_step/max_steps": "1602/2181", "elapsed_time": "4h 14m 52s", "remaining_time": "1h 32m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545265}
{"loss": 24.125, "grad_norm": 1.09172904, "learning_rate": 0.00017323, "token_acc": 0.39384289, "epoch": 0.73523678, "global_step/max_steps": "1603/2181", "elapsed_time": "4h 15m 1s", "remaining_time": "1h 31m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545363}
{"loss": 24.375, "grad_norm": 0.98018616, "learning_rate": 0.00017267, "token_acc": 0.39130435, "epoch": 0.73569545, "global_step/max_steps": "1604/2181", "elapsed_time": "4h 15m 11s", "remaining_time": "1h 31m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545663}
{"loss": 24.25, "grad_norm": 1.14914215, "learning_rate": 0.00017211, "token_acc": 0.38408896, "epoch": 0.73615411, "global_step/max_steps": "1605/2181", "elapsed_time": "4h 15m 20s", "remaining_time": "1h 31m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545394}
{"loss": 24.25, "grad_norm": 1.07849133, "learning_rate": 0.00017155, "token_acc": 0.38321479, "epoch": 0.73661277, "global_step/max_steps": "1606/2181", "elapsed_time": "4h 15m 30s", "remaining_time": "1h 31m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54525}
{"loss": 24.125, "grad_norm": 1.01267147, "learning_rate": 0.00017099, "token_acc": 0.38330171, "epoch": 0.73707144, "global_step/max_steps": "1607/2181", "elapsed_time": "4h 15m 39s", "remaining_time": "1h 31m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545143}
{"loss": 24.75, "grad_norm": 1.13959408, "learning_rate": 0.00017043, "token_acc": 0.38267544, "epoch": 0.7375301, "global_step/max_steps": "1608/2181", "elapsed_time": "4h 15m 49s", "remaining_time": "1h 31m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545415}
{"loss": 23.75, "grad_norm": 1.102561, "learning_rate": 0.00016987, "token_acc": 0.39141344, "epoch": 0.73798876, "global_step/max_steps": "1609/2181", "elapsed_time": "4h 15m 58s", "remaining_time": "1h 30m 60s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54519}
{"loss": 24.125, "grad_norm": 1.04663026, "learning_rate": 0.00016932, "token_acc": 0.37880406, "epoch": 0.73844743, "global_step/max_steps": "1610/2181", "elapsed_time": "4h 16m 8s", "remaining_time": "1h 30m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545184}
{"loss": 24.25, "grad_norm": 1.03937328, "learning_rate": 0.00016876, "token_acc": 0.37001376, "epoch": 0.73890609, "global_step/max_steps": "1611/2181", "elapsed_time": "4h 16m 17s", "remaining_time": "1h 30m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545226}
{"loss": 24.0, "grad_norm": 1.20012045, "learning_rate": 0.0001682, "token_acc": 0.38599944, "epoch": 0.73936475, "global_step/max_steps": "1612/2181", "elapsed_time": "4h 16m 27s", "remaining_time": "1h 30m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545107}
{"loss": 24.125, "grad_norm": 1.12838209, "learning_rate": 0.00016765, "token_acc": 0.38379469, "epoch": 0.73982341, "global_step/max_steps": "1613/2181", "elapsed_time": "4h 16m 36s", "remaining_time": "1h 30m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544805}
{"loss": 24.0, "grad_norm": 1.03362918, "learning_rate": 0.00016709, "token_acc": 0.38245906, "epoch": 0.74028208, "global_step/max_steps": "1614/2181", "elapsed_time": "4h 16m 45s", "remaining_time": "1h 30m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544896}
{"loss": 24.125, "grad_norm": 1.05010021, "learning_rate": 0.00016654, "token_acc": 0.37372654, "epoch": 0.74074074, "global_step/max_steps": "1615/2181", "elapsed_time": "4h 16m 56s", "remaining_time": "1h 30m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54553}
{"loss": 24.0, "grad_norm": 1.12056124, "learning_rate": 0.00016599, "token_acc": 0.38601478, "epoch": 0.7411994, "global_step/max_steps": "1616/2181", "elapsed_time": "4h 17m 5s", "remaining_time": "1h 29m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545409}
{"loss": 23.5, "grad_norm": 1.08163428, "learning_rate": 0.00016543, "token_acc": 0.39004499, "epoch": 0.74165807, "global_step/max_steps": "1617/2181", "elapsed_time": "4h 17m 15s", "remaining_time": "1h 29m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545284}
{"loss": 24.375, "grad_norm": 1.20441496, "learning_rate": 0.00016488, "token_acc": 0.37777778, "epoch": 0.74211673, "global_step/max_steps": "1618/2181", "elapsed_time": "4h 17m 24s", "remaining_time": "1h 29m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54525}
{"loss": 22.5, "grad_norm": 1.10042012, "learning_rate": 0.00016433, "token_acc": 0.39880952, "epoch": 0.74257539, "global_step/max_steps": "1619/2181", "elapsed_time": "4h 17m 34s", "remaining_time": "1h 29m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54514}
{"loss": 24.125, "grad_norm": 1.07610846, "learning_rate": 0.00016378, "token_acc": 0.38139535, "epoch": 0.74303406, "global_step/max_steps": "1620/2181", "elapsed_time": "4h 17m 43s", "remaining_time": "1h 29m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545358}
{"loss": 23.25, "grad_norm": 1.12324154, "learning_rate": 0.00016323, "token_acc": 0.40022081, "epoch": 0.74349272, "global_step/max_steps": "1621/2181", "elapsed_time": "4h 17m 52s", "remaining_time": "1h 29m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544931}
{"loss": 23.5, "grad_norm": 1.11487269, "learning_rate": 0.00016268, "token_acc": 0.39213514, "epoch": 0.74395138, "global_step/max_steps": "1622/2181", "elapsed_time": "4h 18m 1s", "remaining_time": "1h 28m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544583}
{"loss": 23.875, "grad_norm": 1.1495831, "learning_rate": 0.00016214, "token_acc": 0.38202568, "epoch": 0.74441004, "global_step/max_steps": "1623/2181", "elapsed_time": "4h 18m 11s", "remaining_time": "1h 28m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544614}
{"loss": 24.125, "grad_norm": 1.03891897, "learning_rate": 0.00016159, "token_acc": 0.374605, "epoch": 0.74486871, "global_step/max_steps": "1624/2181", "elapsed_time": "4h 18m 21s", "remaining_time": "1h 28m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544984}
{"loss": 23.75, "grad_norm": 1.05044711, "learning_rate": 0.00016104, "token_acc": 0.38831909, "epoch": 0.74532737, "global_step/max_steps": "1625/2181", "elapsed_time": "4h 18m 31s", "remaining_time": "1h 28m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545091}
{"loss": 23.5, "grad_norm": 1.02659333, "learning_rate": 0.0001605, "token_acc": 0.37669152, "epoch": 0.74578603, "global_step/max_steps": "1626/2181", "elapsed_time": "4h 18m 40s", "remaining_time": "1h 28m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545152}
{"loss": 23.375, "grad_norm": 1.51165414, "learning_rate": 0.00015995, "token_acc": 0.39994347, "epoch": 0.7462447, "global_step/max_steps": "1627/2181", "elapsed_time": "4h 18m 50s", "remaining_time": "1h 28m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545263}
{"loss": 24.75, "grad_norm": 1.10536408, "learning_rate": 0.00015941, "token_acc": 0.37747561, "epoch": 0.74670336, "global_step/max_steps": "1628/2181", "elapsed_time": "4h 18m 59s", "remaining_time": "1h 27m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545136}
{"loss": 22.625, "grad_norm": 1.04685998, "learning_rate": 0.00015887, "token_acc": 0.40548258, "epoch": 0.74716202, "global_step/max_steps": "1629/2181", "elapsed_time": "4h 19m 9s", "remaining_time": "1h 27m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545176}
{"loss": 23.5, "grad_norm": 1.01086998, "learning_rate": 0.00015832, "token_acc": 0.38770208, "epoch": 0.74762069, "global_step/max_steps": "1630/2181", "elapsed_time": "4h 19m 19s", "remaining_time": "1h 27m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545605}
{"loss": 23.625, "grad_norm": 1.12170172, "learning_rate": 0.00015778, "token_acc": 0.3877154, "epoch": 0.74807935, "global_step/max_steps": "1631/2181", "elapsed_time": "4h 19m 29s", "remaining_time": "1h 27m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545551}
{"loss": 23.875, "grad_norm": 1.10216856, "learning_rate": 0.00015724, "token_acc": 0.38203168, "epoch": 0.74853801, "global_step/max_steps": "1632/2181", "elapsed_time": "4h 19m 38s", "remaining_time": "1h 27m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545592}
{"loss": 24.625, "grad_norm": 1.13015902, "learning_rate": 0.0001567, "token_acc": 0.37851732, "epoch": 0.74899667, "global_step/max_steps": "1633/2181", "elapsed_time": "4h 19m 49s", "remaining_time": "1h 27m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545947}
{"loss": 24.125, "grad_norm": 0.9938367, "learning_rate": 0.00015616, "token_acc": 0.36955927, "epoch": 0.74945534, "global_step/max_steps": "1634/2181", "elapsed_time": "4h 19m 57s", "remaining_time": "1h 27m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545535}
{"loss": 24.5, "grad_norm": 0.94960129, "learning_rate": 0.00015562, "token_acc": 0.37275799, "epoch": 0.749914, "global_step/max_steps": "1635/2181", "elapsed_time": "4h 20m 7s", "remaining_time": "1h 26m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545296}
{"loss": 23.75, "grad_norm": 1.10976088, "learning_rate": 0.00015508, "token_acc": 0.38601583, "epoch": 0.75037266, "global_step/max_steps": "1636/2181", "elapsed_time": "4h 20m 16s", "remaining_time": "1h 26m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544933}
{"loss": 23.25, "grad_norm": 0.99418378, "learning_rate": 0.00015455, "token_acc": 0.38529574, "epoch": 0.75083133, "global_step/max_steps": "1637/2181", "elapsed_time": "4h 20m 24s", "remaining_time": "1h 26m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54458}
{"loss": 24.5, "grad_norm": 0.99040049, "learning_rate": 0.00015401, "token_acc": 0.38602065, "epoch": 0.75128999, "global_step/max_steps": "1638/2181", "elapsed_time": "4h 20m 34s", "remaining_time": "1h 26m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544417}
{"loss": 23.625, "grad_norm": 0.99766487, "learning_rate": 0.00015347, "token_acc": 0.39483395, "epoch": 0.75174865, "global_step/max_steps": "1639/2181", "elapsed_time": "4h 20m 43s", "remaining_time": "1h 26m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544525}
{"loss": 24.375, "grad_norm": 1.19279397, "learning_rate": 0.00015294, "token_acc": 0.38481375, "epoch": 0.75220732, "global_step/max_steps": "1640/2181", "elapsed_time": "4h 20m 53s", "remaining_time": "1h 26m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544258}
{"loss": 23.5, "grad_norm": 1.25106835, "learning_rate": 0.0001524, "token_acc": 0.39976656, "epoch": 0.75266598, "global_step/max_steps": "1641/2181", "elapsed_time": "4h 21m 3s", "remaining_time": "1h 25m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544492}
{"loss": 22.625, "grad_norm": 1.15039492, "learning_rate": 0.00015187, "token_acc": 0.4084587, "epoch": 0.75312464, "global_step/max_steps": "1642/2181", "elapsed_time": "4h 21m 12s", "remaining_time": "1h 25m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544255}
{"loss": 23.625, "grad_norm": 1.07084513, "learning_rate": 0.00015134, "token_acc": 0.38497783, "epoch": 0.7535833, "global_step/max_steps": "1643/2181", "elapsed_time": "4h 21m 22s", "remaining_time": "1h 25m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544684}
{"loss": 23.5, "grad_norm": 0.97486031, "learning_rate": 0.00015081, "token_acc": 0.38637665, "epoch": 0.75404197, "global_step/max_steps": "1644/2181", "elapsed_time": "4h 21m 31s", "remaining_time": "1h 25m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544429}
{"loss": 23.625, "grad_norm": 1.06310165, "learning_rate": 0.00015027, "token_acc": 0.39146515, "epoch": 0.75450063, "global_step/max_steps": "1645/2181", "elapsed_time": "4h 21m 41s", "remaining_time": "1h 25m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544537}
{"loss": 24.0, "grad_norm": 1.02098036, "learning_rate": 0.00014974, "token_acc": 0.38441998, "epoch": 0.75495929, "global_step/max_steps": "1646/2181", "elapsed_time": "4h 21m 50s", "remaining_time": "1h 25m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544213}
{"loss": 23.875, "grad_norm": 1.07239068, "learning_rate": 0.00014921, "token_acc": 0.37417219, "epoch": 0.75541796, "global_step/max_steps": "1647/2181", "elapsed_time": "4h 21m 59s", "remaining_time": "1h 24m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544131}
{"loss": 24.0, "grad_norm": 1.12986457, "learning_rate": 0.00014869, "token_acc": 0.37608757, "epoch": 0.75587662, "global_step/max_steps": "1648/2181", "elapsed_time": "4h 22m 9s", "remaining_time": "1h 24m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.5443}
{"loss": 24.375, "grad_norm": 1.20870316, "learning_rate": 0.00014816, "token_acc": 0.37704464, "epoch": 0.75633528, "global_step/max_steps": "1649/2181", "elapsed_time": "4h 22m 19s", "remaining_time": "1h 24m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544806}
{"loss": 24.0, "grad_norm": 1.05052102, "learning_rate": 0.00014763, "token_acc": 0.37184633, "epoch": 0.75679395, "global_step/max_steps": "1650/2181", "elapsed_time": "4h 22m 30s", "remaining_time": "1h 24m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545252}
{"loss": 23.625, "grad_norm": 1.13494861, "learning_rate": 0.0001471, "token_acc": 0.39262776, "epoch": 0.75725261, "global_step/max_steps": "1651/2181", "elapsed_time": "4h 22m 39s", "remaining_time": "1h 24m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545337}
{"loss": 23.625, "grad_norm": 1.07315755, "learning_rate": 0.00014658, "token_acc": 0.4085592, "epoch": 0.75771127, "global_step/max_steps": "1652/2181", "elapsed_time": "4h 22m 48s", "remaining_time": "1h 24m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545012}
{"loss": 23.5, "grad_norm": 1.03512585, "learning_rate": 0.00014605, "token_acc": 0.39571746, "epoch": 0.75816993, "global_step/max_steps": "1653/2181", "elapsed_time": "4h 22m 57s", "remaining_time": "1h 23m 60s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544702}
{"loss": 24.25, "grad_norm": 1.07655251, "learning_rate": 0.00014553, "token_acc": 0.38588631, "epoch": 0.7586286, "global_step/max_steps": "1654/2181", "elapsed_time": "4h 23m 7s", "remaining_time": "1h 23m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544647}
{"loss": 23.25, "grad_norm": 1.02893674, "learning_rate": 0.00014501, "token_acc": 0.40756414, "epoch": 0.75908726, "global_step/max_steps": "1655/2181", "elapsed_time": "4h 23m 16s", "remaining_time": "1h 23m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544453}
{"loss": 23.875, "grad_norm": 0.99826556, "learning_rate": 0.00014448, "token_acc": 0.391339, "epoch": 0.75954592, "global_step/max_steps": "1656/2181", "elapsed_time": "4h 23m 26s", "remaining_time": "1h 23m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544413}
{"loss": 23.125, "grad_norm": 1.24864984, "learning_rate": 0.00014396, "token_acc": 0.40283401, "epoch": 0.76000459, "global_step/max_steps": "1657/2181", "elapsed_time": "4h 23m 35s", "remaining_time": "1h 23m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544061}
{"loss": 23.25, "grad_norm": 1.09988654, "learning_rate": 0.00014344, "token_acc": 0.40513393, "epoch": 0.76046325, "global_step/max_steps": "1658/2181", "elapsed_time": "4h 23m 44s", "remaining_time": "1h 23m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544098}
{"loss": 24.125, "grad_norm": 1.12327635, "learning_rate": 0.00014292, "token_acc": 0.37972557, "epoch": 0.76092191, "global_step/max_steps": "1659/2181", "elapsed_time": "4h 23m 54s", "remaining_time": "1h 23m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544297}
{"loss": 24.125, "grad_norm": 1.12613809, "learning_rate": 0.0001424, "token_acc": 0.37604225, "epoch": 0.76138058, "global_step/max_steps": "1660/2181", "elapsed_time": "4h 24m 3s", "remaining_time": "1h 22m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544271}
{"loss": 23.625, "grad_norm": 0.9980222, "learning_rate": 0.00014188, "token_acc": 0.39711191, "epoch": 0.76183924, "global_step/max_steps": "1661/2181", "elapsed_time": "4h 24m 13s", "remaining_time": "1h 22m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544011}
{"loss": 24.375, "grad_norm": 1.13486814, "learning_rate": 0.00014136, "token_acc": 0.38567416, "epoch": 0.7622979, "global_step/max_steps": "1662/2181", "elapsed_time": "4h 24m 22s", "remaining_time": "1h 22m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543752}
{"loss": 24.0, "grad_norm": 1.11037254, "learning_rate": 0.00014085, "token_acc": 0.38841846, "epoch": 0.76275656, "global_step/max_steps": "1663/2181", "elapsed_time": "4h 24m 30s", "remaining_time": "1h 22m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54327}
{"loss": 23.75, "grad_norm": 1.05422461, "learning_rate": 0.00014033, "token_acc": 0.38576014, "epoch": 0.76321523, "global_step/max_steps": "1664/2181", "elapsed_time": "4h 24m 40s", "remaining_time": "1h 22m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543254}
{"loss": 24.0, "grad_norm": 1.18157375, "learning_rate": 0.00013981, "token_acc": 0.38704273, "epoch": 0.76367389, "global_step/max_steps": "1665/2181", "elapsed_time": "4h 24m 49s", "remaining_time": "1h 22m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543121}
{"loss": 24.25, "grad_norm": 1.07004201, "learning_rate": 0.0001393, "token_acc": 0.37851872, "epoch": 0.76413255, "global_step/max_steps": "1666/2181", "elapsed_time": "4h 24m 58s", "remaining_time": "1h 21m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542848}
{"loss": 24.0, "grad_norm": 1.09286821, "learning_rate": 0.00013879, "token_acc": 0.39077087, "epoch": 0.76459122, "global_step/max_steps": "1667/2181", "elapsed_time": "4h 25m 8s", "remaining_time": "1h 21m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542805}
{"loss": 23.25, "grad_norm": 1.08403826, "learning_rate": 0.00013827, "token_acc": 0.39540165, "epoch": 0.76504988, "global_step/max_steps": "1668/2181", "elapsed_time": "4h 25m 17s", "remaining_time": "1h 21m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54271}
{"loss": 23.625, "grad_norm": 1.23485577, "learning_rate": 0.00013776, "token_acc": 0.39616963, "epoch": 0.76550854, "global_step/max_steps": "1669/2181", "elapsed_time": "4h 25m 27s", "remaining_time": "1h 21m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542672}
{"loss": 23.75, "grad_norm": 1.04923975, "learning_rate": 0.00013725, "token_acc": 0.38410017, "epoch": 0.76596721, "global_step/max_steps": "1670/2181", "elapsed_time": "4h 25m 37s", "remaining_time": "1h 21m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543047}
{"loss": 23.75, "grad_norm": 1.02434921, "learning_rate": 0.00013674, "token_acc": 0.38355377, "epoch": 0.76642587, "global_step/max_steps": "1671/2181", "elapsed_time": "4h 25m 46s", "remaining_time": "1h 21m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54291}
{"loss": 24.0, "grad_norm": 1.22231424, "learning_rate": 0.00013623, "token_acc": 0.39052751, "epoch": 0.76688453, "global_step/max_steps": "1672/2181", "elapsed_time": "4h 25m 56s", "remaining_time": "1h 20m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543255}
{"loss": 23.875, "grad_norm": 1.02303183, "learning_rate": 0.00013572, "token_acc": 0.38381568, "epoch": 0.76734319, "global_step/max_steps": "1673/2181", "elapsed_time": "4h 26m 5s", "remaining_time": "1h 20m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542933}
{"loss": 23.75, "grad_norm": 1.06479347, "learning_rate": 0.00013521, "token_acc": 0.3818743, "epoch": 0.76780186, "global_step/max_steps": "1674/2181", "elapsed_time": "4h 26m 15s", "remaining_time": "1h 20m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543028}
{"loss": 23.875, "grad_norm": 1.10048723, "learning_rate": 0.0001347, "token_acc": 0.39563863, "epoch": 0.76826052, "global_step/max_steps": "1675/2181", "elapsed_time": "4h 26m 25s", "remaining_time": "1h 20m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543131}
{"loss": 23.625, "grad_norm": 1.12892163, "learning_rate": 0.0001342, "token_acc": 0.38951724, "epoch": 0.76871918, "global_step/max_steps": "1676/2181", "elapsed_time": "4h 26m 34s", "remaining_time": "1h 20m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543223}
{"loss": 24.875, "grad_norm": 1.07918262, "learning_rate": 0.00013369, "token_acc": 0.37198336, "epoch": 0.76917785, "global_step/max_steps": "1677/2181", "elapsed_time": "4h 26m 44s", "remaining_time": "1h 20m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543286}
{"loss": 24.5, "grad_norm": 1.18959665, "learning_rate": 0.00013319, "token_acc": 0.38450899, "epoch": 0.76963651, "global_step/max_steps": "1678/2181", "elapsed_time": "4h 26m 53s", "remaining_time": "1h 20m 0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543123}
{"loss": 23.25, "grad_norm": 1.07601988, "learning_rate": 0.00013268, "token_acc": 0.39965147, "epoch": 0.77009517, "global_step/max_steps": "1679/2181", "elapsed_time": "4h 27m 3s", "remaining_time": "1h 19m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542921}
{"loss": 24.0, "grad_norm": 1.00719774, "learning_rate": 0.00013218, "token_acc": 0.39422558, "epoch": 0.77055384, "global_step/max_steps": "1680/2181", "elapsed_time": "4h 27m 13s", "remaining_time": "1h 19m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543501}
{"loss": 24.0, "grad_norm": 1.06482267, "learning_rate": 0.00013168, "token_acc": 0.38502227, "epoch": 0.7710125, "global_step/max_steps": "1681/2181", "elapsed_time": "4h 27m 22s", "remaining_time": "1h 19m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54337}
{"loss": 24.0, "grad_norm": 1.06347501, "learning_rate": 0.00013117, "token_acc": 0.38802661, "epoch": 0.77147116, "global_step/max_steps": "1682/2181", "elapsed_time": "4h 27m 31s", "remaining_time": "1h 19m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542971}
{"loss": 23.25, "grad_norm": 0.99625409, "learning_rate": 0.00013067, "token_acc": 0.38896402, "epoch": 0.77192982, "global_step/max_steps": "1683/2181", "elapsed_time": "4h 27m 41s", "remaining_time": "1h 19m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542843}
{"loss": 24.125, "grad_norm": 1.08220303, "learning_rate": 0.00013017, "token_acc": 0.38470554, "epoch": 0.77238849, "global_step/max_steps": "1684/2181", "elapsed_time": "4h 27m 50s", "remaining_time": "1h 19m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542844}
{"loss": 23.875, "grad_norm": 1.02598178, "learning_rate": 0.00012967, "token_acc": 0.38611416, "epoch": 0.77284715, "global_step/max_steps": "1685/2181", "elapsed_time": "4h 27m 60s", "remaining_time": "1h 18m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543025}
{"loss": 24.0, "grad_norm": 1.09610343, "learning_rate": 0.00012917, "token_acc": 0.38626609, "epoch": 0.77330581, "global_step/max_steps": "1686/2181", "elapsed_time": "4h 28m 9s", "remaining_time": "1h 18m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542642}
{"loss": 23.5, "grad_norm": 1.08648491, "learning_rate": 0.00012868, "token_acc": 0.38989272, "epoch": 0.77376448, "global_step/max_steps": "1687/2181", "elapsed_time": "4h 28m 18s", "remaining_time": "1h 18m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542478}
{"loss": 24.25, "grad_norm": 1.07214606, "learning_rate": 0.00012818, "token_acc": 0.37679558, "epoch": 0.77422314, "global_step/max_steps": "1688/2181", "elapsed_time": "4h 28m 27s", "remaining_time": "1h 18m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542198}
{"loss": 23.625, "grad_norm": 1.12798464, "learning_rate": 0.00012768, "token_acc": 0.39005525, "epoch": 0.7746818, "global_step/max_steps": "1689/2181", "elapsed_time": "4h 28m 37s", "remaining_time": "1h 18m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542133}
{"loss": 23.5, "grad_norm": 1.02672887, "learning_rate": 0.00012719, "token_acc": 0.38195233, "epoch": 0.77514047, "global_step/max_steps": "1690/2181", "elapsed_time": "4h 28m 46s", "remaining_time": "1h 18m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542052}
{"loss": 24.125, "grad_norm": 1.12624693, "learning_rate": 0.00012669, "token_acc": 0.38655929, "epoch": 0.77559913, "global_step/max_steps": "1691/2181", "elapsed_time": "4h 28m 55s", "remaining_time": "1h 17m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541787}
{"loss": 23.75, "grad_norm": 1.11761391, "learning_rate": 0.0001262, "token_acc": 0.39512472, "epoch": 0.77605779, "global_step/max_steps": "1692/2181", "elapsed_time": "4h 29m 5s", "remaining_time": "1h 17m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541808}
{"loss": 24.125, "grad_norm": 1.01529503, "learning_rate": 0.00012571, "token_acc": 0.38476295, "epoch": 0.77651645, "global_step/max_steps": "1693/2181", "elapsed_time": "4h 29m 15s", "remaining_time": "1h 17m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542204}
{"loss": 24.0, "grad_norm": 1.09003377, "learning_rate": 0.00012521, "token_acc": 0.38537271, "epoch": 0.77697512, "global_step/max_steps": "1694/2181", "elapsed_time": "4h 29m 24s", "remaining_time": "1h 17m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541937}
{"loss": 24.0, "grad_norm": 1.05206585, "learning_rate": 0.00012472, "token_acc": 0.37771509, "epoch": 0.77743378, "global_step/max_steps": "1695/2181", "elapsed_time": "4h 29m 34s", "remaining_time": "1h 17m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541951}
{"loss": 24.25, "grad_norm": 1.12550592, "learning_rate": 0.00012423, "token_acc": 0.37615708, "epoch": 0.77789244, "global_step/max_steps": "1696/2181", "elapsed_time": "4h 29m 44s", "remaining_time": "1h 17m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542538}
{"loss": 23.625, "grad_norm": 1.15777266, "learning_rate": 0.00012374, "token_acc": 0.39439714, "epoch": 0.77835111, "global_step/max_steps": "1697/2181", "elapsed_time": "4h 29m 54s", "remaining_time": "1h 16m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542609}
{"loss": 23.375, "grad_norm": 1.09410775, "learning_rate": 0.00012325, "token_acc": 0.3920439, "epoch": 0.77880977, "global_step/max_steps": "1698/2181", "elapsed_time": "4h 30m 3s", "remaining_time": "1h 16m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54263}
{"loss": 23.875, "grad_norm": 1.05412197, "learning_rate": 0.00012277, "token_acc": 0.4, "epoch": 0.77926843, "global_step/max_steps": "1699/2181", "elapsed_time": "4h 30m 12s", "remaining_time": "1h 16m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542363}
{"loss": 24.5, "grad_norm": 1.13209581, "learning_rate": 0.00012228, "token_acc": 0.37808516, "epoch": 0.7797271, "global_step/max_steps": "1700/2181", "elapsed_time": "4h 30m 22s", "remaining_time": "1h 16m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542406}
{"loss": 24.5, "grad_norm": 0.99654567, "learning_rate": 0.00012179, "token_acc": 0.38560411, "epoch": 0.78018576, "global_step/max_steps": "1701/2181", "elapsed_time": "4h 30m 32s", "remaining_time": "1h 16m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542369}
{"loss": 24.125, "grad_norm": 1.1169579, "learning_rate": 0.00012131, "token_acc": 0.38851541, "epoch": 0.78064442, "global_step/max_steps": "1702/2181", "elapsed_time": "4h 30m 41s", "remaining_time": "1h 16m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542315}
{"loss": 23.75, "grad_norm": 1.12118411, "learning_rate": 0.00012082, "token_acc": 0.39293937, "epoch": 0.78110308, "global_step/max_steps": "1703/2181", "elapsed_time": "4h 30m 51s", "remaining_time": "1h 16m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542452}
{"loss": 24.125, "grad_norm": 1.11475909, "learning_rate": 0.00012034, "token_acc": 0.37936681, "epoch": 0.78156175, "global_step/max_steps": "1704/2181", "elapsed_time": "4h 30m 60s", "remaining_time": "1h 15m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542142}
{"loss": 23.625, "grad_norm": 1.10777402, "learning_rate": 0.00011986, "token_acc": 0.3856078, "epoch": 0.78202041, "global_step/max_steps": "1705/2181", "elapsed_time": "4h 31m 9s", "remaining_time": "1h 15m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541969}
{"loss": 24.5, "grad_norm": 1.0259676, "learning_rate": 0.00011938, "token_acc": 0.37753173, "epoch": 0.78247907, "global_step/max_steps": "1706/2181", "elapsed_time": "4h 31m 19s", "remaining_time": "1h 15m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542416}
{"loss": 23.5, "grad_norm": 1.11964536, "learning_rate": 0.00011889, "token_acc": 0.3864215, "epoch": 0.78293774, "global_step/max_steps": "1707/2181", "elapsed_time": "4h 31m 28s", "remaining_time": "1h 15m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542059}
{"loss": 24.25, "grad_norm": 1.09111965, "learning_rate": 0.00011841, "token_acc": 0.37978298, "epoch": 0.7833964, "global_step/max_steps": "1708/2181", "elapsed_time": "4h 31m 38s", "remaining_time": "1h 15m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541937}
{"loss": 24.375, "grad_norm": 1.10481727, "learning_rate": 0.00011793, "token_acc": 0.38350748, "epoch": 0.78385506, "global_step/max_steps": "1709/2181", "elapsed_time": "4h 31m 46s", "remaining_time": "1h 15m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541247}
{"loss": 23.75, "grad_norm": 1.05151784, "learning_rate": 0.00011746, "token_acc": 0.39099591, "epoch": 0.78431373, "global_step/max_steps": "1710/2181", "elapsed_time": "4h 31m 55s", "remaining_time": "1h 14m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541088}
{"loss": 23.5, "grad_norm": 1.06149757, "learning_rate": 0.00011698, "token_acc": 0.39601218, "epoch": 0.78477239, "global_step/max_steps": "1711/2181", "elapsed_time": "4h 32m 5s", "remaining_time": "1h 14m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541346}
{"loss": 23.5, "grad_norm": 1.16068459, "learning_rate": 0.0001165, "token_acc": 0.40083449, "epoch": 0.78523105, "global_step/max_steps": "1712/2181", "elapsed_time": "4h 32m 15s", "remaining_time": "1h 14m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54127}
{"loss": 23.5, "grad_norm": 1.21174395, "learning_rate": 0.00011602, "token_acc": 0.38662873, "epoch": 0.78568971, "global_step/max_steps": "1713/2181", "elapsed_time": "4h 32m 24s", "remaining_time": "1h 14m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541149}
{"loss": 23.75, "grad_norm": 1.07454705, "learning_rate": 0.00011555, "token_acc": 0.38398245, "epoch": 0.78614838, "global_step/max_steps": "1714/2181", "elapsed_time": "4h 32m 34s", "remaining_time": "1h 14m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541562}
{"loss": 23.25, "grad_norm": 1.0830636, "learning_rate": 0.00011507, "token_acc": 0.39612977, "epoch": 0.78660704, "global_step/max_steps": "1715/2181", "elapsed_time": "4h 32m 44s", "remaining_time": "1h 14m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54166}
{"loss": 24.125, "grad_norm": 1.03574312, "learning_rate": 0.0001146, "token_acc": 0.37792171, "epoch": 0.7870657, "global_step/max_steps": "1716/2181", "elapsed_time": "4h 32m 54s", "remaining_time": "1h 13m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541709}
{"loss": 24.625, "grad_norm": 2.09851265, "learning_rate": 0.00011413, "token_acc": 0.36909288, "epoch": 0.78752437, "global_step/max_steps": "1717/2181", "elapsed_time": "4h 33m 3s", "remaining_time": "1h 13m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541794}
{"loss": 23.75, "grad_norm": 1.10843539, "learning_rate": 0.00011366, "token_acc": 0.38479329, "epoch": 0.78798303, "global_step/max_steps": "1718/2181", "elapsed_time": "4h 33m 13s", "remaining_time": "1h 13m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542008}
{"loss": 25.0, "grad_norm": 1.04392684, "learning_rate": 0.00011319, "token_acc": 0.37244616, "epoch": 0.78844169, "global_step/max_steps": "1719/2181", "elapsed_time": "4h 33m 23s", "remaining_time": "1h 13m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542424}
{"loss": 24.0, "grad_norm": 1.0574851, "learning_rate": 0.00011272, "token_acc": 0.3797125, "epoch": 0.78890036, "global_step/max_steps": "1720/2181", "elapsed_time": "4h 33m 32s", "remaining_time": "1h 13m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541953}
{"loss": 24.5, "grad_norm": 1.11830819, "learning_rate": 0.00011225, "token_acc": 0.36647009, "epoch": 0.78935902, "global_step/max_steps": "1721/2181", "elapsed_time": "4h 33m 41s", "remaining_time": "1h 13m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541737}
{"loss": 24.25, "grad_norm": 1.09285617, "learning_rate": 0.00011178, "token_acc": 0.38062756, "epoch": 0.78981768, "global_step/max_steps": "1722/2181", "elapsed_time": "4h 33m 51s", "remaining_time": "1h 12m 60s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541719}
{"loss": 23.5, "grad_norm": 1.17258883, "learning_rate": 0.00011131, "token_acc": 0.39988273, "epoch": 0.79027634, "global_step/max_steps": "1723/2181", "elapsed_time": "4h 34m 1s", "remaining_time": "1h 12m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.5419}
{"loss": 24.0, "grad_norm": 1.05390286, "learning_rate": 0.00011084, "token_acc": 0.37248505, "epoch": 0.79073501, "global_step/max_steps": "1724/2181", "elapsed_time": "4h 34m 10s", "remaining_time": "1h 12m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.5416}
{"loss": 23.875, "grad_norm": 1.07843089, "learning_rate": 0.00011038, "token_acc": 0.37370048, "epoch": 0.79119367, "global_step/max_steps": "1725/2181", "elapsed_time": "4h 34m 19s", "remaining_time": "1h 12m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541307}
{"loss": 22.875, "grad_norm": 1.20331073, "learning_rate": 0.00010991, "token_acc": 0.40708215, "epoch": 0.79165233, "global_step/max_steps": "1726/2181", "elapsed_time": "4h 34m 29s", "remaining_time": "1h 12m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541616}
{"loss": 23.25, "grad_norm": 1.16737878, "learning_rate": 0.00010945, "token_acc": 0.38582236, "epoch": 0.792111, "global_step/max_steps": "1727/2181", "elapsed_time": "4h 34m 39s", "remaining_time": "1h 12m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542227}
{"loss": 23.25, "grad_norm": 1.01282716, "learning_rate": 0.00010899, "token_acc": 0.39737226, "epoch": 0.79256966, "global_step/max_steps": "1728/2181", "elapsed_time": "4h 34m 49s", "remaining_time": "1h 12m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542268}
{"loss": 24.5, "grad_norm": 1.03495562, "learning_rate": 0.00010852, "token_acc": 0.37944984, "epoch": 0.79302832, "global_step/max_steps": "1729/2181", "elapsed_time": "4h 34m 59s", "remaining_time": "1h 11m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542365}
{"loss": 24.25, "grad_norm": 1.2040298, "learning_rate": 0.00010806, "token_acc": 0.38311136, "epoch": 0.79348699, "global_step/max_steps": "1730/2181", "elapsed_time": "4h 35m 8s", "remaining_time": "1h 11m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542222}
{"loss": 24.25, "grad_norm": 1.06989801, "learning_rate": 0.0001076, "token_acc": 0.38615216, "epoch": 0.79394565, "global_step/max_steps": "1731/2181", "elapsed_time": "4h 35m 17s", "remaining_time": "1h 11m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541999}
{"loss": 23.125, "grad_norm": 1.13883424, "learning_rate": 0.00010714, "token_acc": 0.40122768, "epoch": 0.79440431, "global_step/max_steps": "1732/2181", "elapsed_time": "4h 35m 27s", "remaining_time": "1h 11m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542057}
{"loss": 23.375, "grad_norm": 1.10453522, "learning_rate": 0.00010668, "token_acc": 0.40320796, "epoch": 0.79486297, "global_step/max_steps": "1733/2181", "elapsed_time": "4h 35m 36s", "remaining_time": "1h 11m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541938}
{"loss": 24.25, "grad_norm": 1.10371208, "learning_rate": 0.00010622, "token_acc": 0.37629298, "epoch": 0.79532164, "global_step/max_steps": "1734/2181", "elapsed_time": "4h 35m 46s", "remaining_time": "1h 11m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542285}
{"loss": 23.75, "grad_norm": 1.09125412, "learning_rate": 0.00010577, "token_acc": 0.37834758, "epoch": 0.7957803, "global_step/max_steps": "1735/2181", "elapsed_time": "4h 35m 55s", "remaining_time": "1h 10m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54201}
{"loss": 23.875, "grad_norm": 1.07389176, "learning_rate": 0.00010531, "token_acc": 0.38350634, "epoch": 0.79623896, "global_step/max_steps": "1736/2181", "elapsed_time": "4h 36m 5s", "remaining_time": "1h 10m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542236}
{"loss": 24.0, "grad_norm": 1.08804893, "learning_rate": 0.00010485, "token_acc": 0.38091216, "epoch": 0.79669763, "global_step/max_steps": "1737/2181", "elapsed_time": "4h 36m 15s", "remaining_time": "1h 10m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542497}
{"loss": 24.0, "grad_norm": 1.06920183, "learning_rate": 0.0001044, "token_acc": 0.37961677, "epoch": 0.79715629, "global_step/max_steps": "1738/2181", "elapsed_time": "4h 36m 25s", "remaining_time": "1h 10m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542397}
{"loss": 23.5, "grad_norm": 1.07380688, "learning_rate": 0.00010395, "token_acc": 0.39849841, "epoch": 0.79761495, "global_step/max_steps": "1739/2181", "elapsed_time": "4h 36m 34s", "remaining_time": "1h 10m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542374}
{"loss": 24.125, "grad_norm": 1.06683874, "learning_rate": 0.00010349, "token_acc": 0.38707926, "epoch": 0.79807362, "global_step/max_steps": "1740/2181", "elapsed_time": "4h 36m 43s", "remaining_time": "1h 10m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542229}
{"loss": 23.375, "grad_norm": 1.07488942, "learning_rate": 0.00010304, "token_acc": 0.3902638, "epoch": 0.79853228, "global_step/max_steps": "1741/2181", "elapsed_time": "4h 36m 53s", "remaining_time": "1h 9m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542399}
{"loss": 23.25, "grad_norm": 1.16299486, "learning_rate": 0.00010259, "token_acc": 0.39353024, "epoch": 0.79899094, "global_step/max_steps": "1742/2181", "elapsed_time": "4h 37m 3s", "remaining_time": "1h 9m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542469}
{"loss": 23.625, "grad_norm": 1.02580464, "learning_rate": 0.00010214, "token_acc": 0.39009731, "epoch": 0.7994496, "global_step/max_steps": "1743/2181", "elapsed_time": "4h 37m 14s", "remaining_time": "1h 9m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543227}
{"loss": 23.625, "grad_norm": 1.04329884, "learning_rate": 0.00010169, "token_acc": 0.38664374, "epoch": 0.79990827, "global_step/max_steps": "1744/2181", "elapsed_time": "4h 37m 23s", "remaining_time": "1h 9m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543131}
{"loss": 23.75, "grad_norm": 0.99909276, "learning_rate": 0.00010124, "token_acc": 0.39729658, "epoch": 0.80036693, "global_step/max_steps": "1745/2181", "elapsed_time": "4h 37m 32s", "remaining_time": "1h 9m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542908}
{"loss": 24.125, "grad_norm": 1.18391144, "learning_rate": 0.00010079, "token_acc": 0.3768546, "epoch": 0.80082559, "global_step/max_steps": "1746/2181", "elapsed_time": "4h 37m 42s", "remaining_time": "1h 9m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542757}
{"loss": 23.75, "grad_norm": 1.10524225, "learning_rate": 0.00010035, "token_acc": 0.37823982, "epoch": 0.80128426, "global_step/max_steps": "1747/2181", "elapsed_time": "4h 37m 51s", "remaining_time": "1h 9m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542789}
{"loss": 24.125, "grad_norm": 1.02459395, "learning_rate": 9.99e-05, "token_acc": 0.38476454, "epoch": 0.80174292, "global_step/max_steps": "1748/2181", "elapsed_time": "4h 38m 1s", "remaining_time": "1h 8m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543177}
{"loss": 23.0, "grad_norm": 1.09081566, "learning_rate": 9.946e-05, "token_acc": 0.39266304, "epoch": 0.80220158, "global_step/max_steps": "1749/2181", "elapsed_time": "4h 38m 11s", "remaining_time": "1h 8m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543091}
{"loss": 23.25, "grad_norm": 1.05090857, "learning_rate": 9.901e-05, "token_acc": 0.40095937, "epoch": 0.80266025, "global_step/max_steps": "1750/2181", "elapsed_time": "4h 38m 21s", "remaining_time": "1h 8m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543238}
{"loss": 23.625, "grad_norm": 1.07734275, "learning_rate": 9.857e-05, "token_acc": 0.38868766, "epoch": 0.80311891, "global_step/max_steps": "1751/2181", "elapsed_time": "4h 38m 30s", "remaining_time": "1h 8m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542967}
{"loss": 24.0, "grad_norm": 1.09922862, "learning_rate": 9.813e-05, "token_acc": 0.39086294, "epoch": 0.80357757, "global_step/max_steps": "1752/2181", "elapsed_time": "4h 38m 40s", "remaining_time": "1h 8m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543187}
{"loss": 24.875, "grad_norm": 1.02639723, "learning_rate": 9.769e-05, "token_acc": 0.36268286, "epoch": 0.80403623, "global_step/max_steps": "1753/2181", "elapsed_time": "4h 38m 49s", "remaining_time": "1h 8m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542967}
{"loss": 23.25, "grad_norm": 1.109676, "learning_rate": 9.724e-05, "token_acc": 0.40567067, "epoch": 0.8044949, "global_step/max_steps": "1754/2181", "elapsed_time": "4h 38m 58s", "remaining_time": "1h 7m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542737}
{"loss": 22.875, "grad_norm": 1.06682837, "learning_rate": 9.681e-05, "token_acc": 0.38857483, "epoch": 0.80495356, "global_step/max_steps": "1755/2181", "elapsed_time": "4h 39m 8s", "remaining_time": "1h 7m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54274}
{"loss": 22.5, "grad_norm": 1.07262504, "learning_rate": 9.637e-05, "token_acc": 0.41977716, "epoch": 0.80541222, "global_step/max_steps": "1756/2181", "elapsed_time": "4h 39m 17s", "remaining_time": "1h 7m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542607}
{"loss": 24.25, "grad_norm": 1.0730679, "learning_rate": 9.593e-05, "token_acc": 0.38017917, "epoch": 0.80587089, "global_step/max_steps": "1757/2181", "elapsed_time": "4h 39m 27s", "remaining_time": "1h 7m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542698}
{"loss": 24.25, "grad_norm": 1.05593216, "learning_rate": 9.549e-05, "token_acc": 0.37568757, "epoch": 0.80632955, "global_step/max_steps": "1758/2181", "elapsed_time": "4h 39m 36s", "remaining_time": "1h 7m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542434}
{"loss": 23.25, "grad_norm": 1.05609691, "learning_rate": 9.506e-05, "token_acc": 0.39552035, "epoch": 0.80678821, "global_step/max_steps": "1759/2181", "elapsed_time": "4h 39m 45s", "remaining_time": "1h 7m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542149}
{"loss": 24.0, "grad_norm": 1.092345, "learning_rate": 9.462e-05, "token_acc": 0.39320524, "epoch": 0.80724688, "global_step/max_steps": "1760/2181", "elapsed_time": "4h 39m 54s", "remaining_time": "1h 6m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542215}
{"loss": 23.25, "grad_norm": 1.05111682, "learning_rate": 9.419e-05, "token_acc": 0.39645815, "epoch": 0.80770554, "global_step/max_steps": "1761/2181", "elapsed_time": "4h 40m 3s", "remaining_time": "1h 6m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542011}
{"loss": 24.75, "grad_norm": 1.16741133, "learning_rate": 9.375e-05, "token_acc": 0.37628152, "epoch": 0.8081642, "global_step/max_steps": "1762/2181", "elapsed_time": "4h 40m 13s", "remaining_time": "1h 6m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.541953}
{"loss": 24.5, "grad_norm": 1.07301652, "learning_rate": 9.332e-05, "token_acc": 0.38080752, "epoch": 0.80862286, "global_step/max_steps": "1763/2181", "elapsed_time": "4h 40m 23s", "remaining_time": "1h 6m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542336}
{"loss": 24.125, "grad_norm": 1.07365668, "learning_rate": 9.289e-05, "token_acc": 0.37854535, "epoch": 0.80908153, "global_step/max_steps": "1764/2181", "elapsed_time": "4h 40m 34s", "remaining_time": "1h 6m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.542974}
{"loss": 23.75, "grad_norm": 1.07159197, "learning_rate": 9.246e-05, "token_acc": 0.38963964, "epoch": 0.80954019, "global_step/max_steps": "1765/2181", "elapsed_time": "4h 40m 44s", "remaining_time": "1h 6m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543349}
{"loss": 22.875, "grad_norm": 1.07387853, "learning_rate": 9.203e-05, "token_acc": 0.40213626, "epoch": 0.80999885, "global_step/max_steps": "1766/2181", "elapsed_time": "4h 40m 53s", "remaining_time": "1h 6m 0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54327}
{"loss": 24.25, "grad_norm": 1.06754303, "learning_rate": 9.16e-05, "token_acc": 0.38269915, "epoch": 0.81045752, "global_step/max_steps": "1767/2181", "elapsed_time": "4h 41m 3s", "remaining_time": "1h 5m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54342}
{"loss": 24.0, "grad_norm": 1.08348882, "learning_rate": 9.117e-05, "token_acc": 0.37866518, "epoch": 0.81091618, "global_step/max_steps": "1768/2181", "elapsed_time": "4h 41m 13s", "remaining_time": "1h 5m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543638}
{"loss": 24.5, "grad_norm": 1.15035152, "learning_rate": 9.074e-05, "token_acc": 0.38774389, "epoch": 0.81137484, "global_step/max_steps": "1769/2181", "elapsed_time": "4h 41m 22s", "remaining_time": "1h 5m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543415}
{"loss": 23.875, "grad_norm": 0.99857938, "learning_rate": 9.032e-05, "token_acc": 0.38607764, "epoch": 0.81183351, "global_step/max_steps": "1770/2181", "elapsed_time": "4h 41m 32s", "remaining_time": "1h 5m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543656}
{"loss": 23.375, "grad_norm": 0.96822333, "learning_rate": 8.989e-05, "token_acc": 0.40194715, "epoch": 0.81229217, "global_step/max_steps": "1771/2181", "elapsed_time": "4h 41m 42s", "remaining_time": "1h 5m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543505}
{"loss": 23.75, "grad_norm": 1.02419829, "learning_rate": 8.947e-05, "token_acc": 0.38488913, "epoch": 0.81275083, "global_step/max_steps": "1772/2181", "elapsed_time": "4h 41m 51s", "remaining_time": "1h 5m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543494}
{"loss": 24.0, "grad_norm": 1.14347696, "learning_rate": 8.904e-05, "token_acc": 0.39404255, "epoch": 0.81320949, "global_step/max_steps": "1773/2181", "elapsed_time": "4h 42m 1s", "remaining_time": "1h 4m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543932}
{"loss": 23.75, "grad_norm": 1.07149112, "learning_rate": 8.862e-05, "token_acc": 0.38896629, "epoch": 0.81366816, "global_step/max_steps": "1774/2181", "elapsed_time": "4h 42m 11s", "remaining_time": "1h 4m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543823}
{"loss": 23.625, "grad_norm": 1.13004124, "learning_rate": 8.82e-05, "token_acc": 0.38672831, "epoch": 0.81412682, "global_step/max_steps": "1775/2181", "elapsed_time": "4h 42m 20s", "remaining_time": "1h 4m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543927}
{"loss": 24.375, "grad_norm": 0.97908962, "learning_rate": 8.778e-05, "token_acc": 0.3799458, "epoch": 0.81458548, "global_step/max_steps": "1776/2181", "elapsed_time": "4h 42m 30s", "remaining_time": "1h 4m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544021}
{"loss": 23.5, "grad_norm": 1.04705691, "learning_rate": 8.736e-05, "token_acc": 0.38019169, "epoch": 0.81504415, "global_step/max_steps": "1777/2181", "elapsed_time": "4h 42m 40s", "remaining_time": "1h 4m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543996}
{"loss": 24.0, "grad_norm": 1.00384974, "learning_rate": 8.694e-05, "token_acc": 0.37901033, "epoch": 0.81550281, "global_step/max_steps": "1778/2181", "elapsed_time": "4h 42m 49s", "remaining_time": "1h 4m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543949}
{"loss": 22.75, "grad_norm": 1.09632242, "learning_rate": 8.652e-05, "token_acc": 0.40683144, "epoch": 0.81596147, "global_step/max_steps": "1779/2181", "elapsed_time": "4h 42m 59s", "remaining_time": "1h 3m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544377}
{"loss": 24.75, "grad_norm": 1.11642098, "learning_rate": 8.611e-05, "token_acc": 0.36949548, "epoch": 0.81642014, "global_step/max_steps": "1780/2181", "elapsed_time": "4h 43m 9s", "remaining_time": "1h 3m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54437}
{"loss": 23.125, "grad_norm": 0.89365423, "learning_rate": 8.569e-05, "token_acc": 0.39007861, "epoch": 0.8168788, "global_step/max_steps": "1781/2181", "elapsed_time": "4h 43m 19s", "remaining_time": "1h 3m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54464}
{"loss": 24.0, "grad_norm": 1.12404823, "learning_rate": 8.527e-05, "token_acc": 0.38251956, "epoch": 0.81733746, "global_step/max_steps": "1782/2181", "elapsed_time": "4h 43m 30s", "remaining_time": "1h 3m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545485}
{"loss": 24.125, "grad_norm": 1.05888486, "learning_rate": 8.486e-05, "token_acc": 0.38509142, "epoch": 0.81779612, "global_step/max_steps": "1783/2181", "elapsed_time": "4h 43m 39s", "remaining_time": "1h 3m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545326}
{"loss": 24.125, "grad_norm": 1.08377171, "learning_rate": 8.445e-05, "token_acc": 0.38161784, "epoch": 0.81825479, "global_step/max_steps": "1784/2181", "elapsed_time": "4h 43m 49s", "remaining_time": "1h 3m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545325}
{"loss": 23.75, "grad_norm": 0.97159362, "learning_rate": 8.403e-05, "token_acc": 0.38867403, "epoch": 0.81871345, "global_step/max_steps": "1785/2181", "elapsed_time": "4h 43m 59s", "remaining_time": "1h 3m 0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545789}
{"loss": 24.25, "grad_norm": 1.06982684, "learning_rate": 8.362e-05, "token_acc": 0.37612108, "epoch": 0.81917211, "global_step/max_steps": "1786/2181", "elapsed_time": "4h 44m 8s", "remaining_time": "1h 2m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545494}
{"loss": 24.25, "grad_norm": 1.09465408, "learning_rate": 8.321e-05, "token_acc": 0.38756119, "epoch": 0.81963078, "global_step/max_steps": "1787/2181", "elapsed_time": "4h 44m 18s", "remaining_time": "1h 2m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545331}
{"loss": 23.5, "grad_norm": 1.13868868, "learning_rate": 8.28e-05, "token_acc": 0.39345157, "epoch": 0.82008944, "global_step/max_steps": "1788/2181", "elapsed_time": "4h 44m 27s", "remaining_time": "1h 2m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545139}
{"loss": 23.875, "grad_norm": 1.1091013, "learning_rate": 8.239e-05, "token_acc": 0.37675753, "epoch": 0.8205481, "global_step/max_steps": "1789/2181", "elapsed_time": "4h 44m 36s", "remaining_time": "1h 2m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54518}
{"loss": 23.875, "grad_norm": 0.96073252, "learning_rate": 8.198e-05, "token_acc": 0.38106236, "epoch": 0.82100677, "global_step/max_steps": "1790/2181", "elapsed_time": "4h 44m 45s", "remaining_time": "1h 2m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544882}
{"loss": 23.875, "grad_norm": 0.9741444, "learning_rate": 8.158e-05, "token_acc": 0.37936866, "epoch": 0.82146543, "global_step/max_steps": "1791/2181", "elapsed_time": "4h 44m 55s", "remaining_time": "1h 2m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544762}
{"loss": 24.0, "grad_norm": 1.08846986, "learning_rate": 8.117e-05, "token_acc": 0.37886813, "epoch": 0.82192409, "global_step/max_steps": "1792/2181", "elapsed_time": "4h 45m 4s", "remaining_time": "1h 1m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544828}
{"loss": 23.125, "grad_norm": 1.03830791, "learning_rate": 8.077e-05, "token_acc": 0.38702334, "epoch": 0.82238275, "global_step/max_steps": "1793/2181", "elapsed_time": "4h 45m 14s", "remaining_time": "1h 1m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544774}
{"loss": 24.0, "grad_norm": 1.1102643, "learning_rate": 8.036e-05, "token_acc": 0.38459422, "epoch": 0.82284142, "global_step/max_steps": "1794/2181", "elapsed_time": "4h 45m 24s", "remaining_time": "1h 1m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545292}
{"loss": 23.375, "grad_norm": 1.05507946, "learning_rate": 7.996e-05, "token_acc": 0.38575342, "epoch": 0.82330008, "global_step/max_steps": "1795/2181", "elapsed_time": "4h 45m 33s", "remaining_time": "1h 1m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544861}
{"loss": 23.75, "grad_norm": 1.08092427, "learning_rate": 7.956e-05, "token_acc": 0.37414966, "epoch": 0.82375874, "global_step/max_steps": "1796/2181", "elapsed_time": "4h 45m 43s", "remaining_time": "1h 1m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54489}
{"loss": 23.75, "grad_norm": 1.05223465, "learning_rate": 7.915e-05, "token_acc": 0.40083449, "epoch": 0.82421741, "global_step/max_steps": "1797/2181", "elapsed_time": "4h 45m 52s", "remaining_time": "1h 1m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544595}
{"loss": 23.75, "grad_norm": 1.04662681, "learning_rate": 7.875e-05, "token_acc": 0.37728027, "epoch": 0.82467607, "global_step/max_steps": "1798/2181", "elapsed_time": "4h 46m 1s", "remaining_time": "1h 0m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544524}
{"loss": 24.5, "grad_norm": 1.17635667, "learning_rate": 7.835e-05, "token_acc": 0.38225539, "epoch": 0.82513473, "global_step/max_steps": "1799/2181", "elapsed_time": "4h 46m 10s", "remaining_time": "1h 0m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544332}
{"loss": 23.375, "grad_norm": 1.02159107, "learning_rate": 7.796e-05, "token_acc": 0.3979442, "epoch": 0.8255934, "global_step/max_steps": "1800/2181", "elapsed_time": "4h 46m 20s", "remaining_time": "1h 0m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544629}
{"loss": 24.125, "grad_norm": 1.05300522, "learning_rate": 7.756e-05, "token_acc": 0.38303063, "epoch": 0.82605206, "global_step/max_steps": "1801/2181", "elapsed_time": "4h 46m 30s", "remaining_time": "1h 0m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544516}
{"loss": 23.875, "grad_norm": 1.06388402, "learning_rate": 7.716e-05, "token_acc": 0.38453069, "epoch": 0.82651072, "global_step/max_steps": "1802/2181", "elapsed_time": "4h 46m 39s", "remaining_time": "1h 0m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544579}
{"loss": 24.5, "grad_norm": 1.07078147, "learning_rate": 7.676e-05, "token_acc": 0.37827015, "epoch": 0.82696938, "global_step/max_steps": "1803/2181", "elapsed_time": "4h 46m 48s", "remaining_time": "1h 0m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544236}
{"loss": 23.875, "grad_norm": 1.2049607, "learning_rate": 7.637e-05, "token_acc": 0.36597015, "epoch": 0.82742805, "global_step/max_steps": "1804/2181", "elapsed_time": "4h 46m 58s", "remaining_time": "59m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544352}
{"loss": 23.75, "grad_norm": 1.06418407, "learning_rate": 7.598e-05, "token_acc": 0.36842105, "epoch": 0.82788671, "global_step/max_steps": "1805/2181", "elapsed_time": "4h 47m 8s", "remaining_time": "59m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.5445}
{"loss": 23.5, "grad_norm": 1.05465317, "learning_rate": 7.558e-05, "token_acc": 0.40102244, "epoch": 0.82834537, "global_step/max_steps": "1806/2181", "elapsed_time": "4h 47m 17s", "remaining_time": "59m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544148}
{"loss": 23.75, "grad_norm": 1.07214081, "learning_rate": 7.519e-05, "token_acc": 0.38537118, "epoch": 0.82880404, "global_step/max_steps": "1807/2181", "elapsed_time": "4h 47m 27s", "remaining_time": "59m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544453}
{"loss": 23.25, "grad_norm": 1.24558032, "learning_rate": 7.48e-05, "token_acc": 0.39227468, "epoch": 0.8292627, "global_step/max_steps": "1808/2181", "elapsed_time": "4h 47m 36s", "remaining_time": "59m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544175}
{"loss": 24.25, "grad_norm": 1.07628679, "learning_rate": 7.441e-05, "token_acc": 0.38416504, "epoch": 0.82972136, "global_step/max_steps": "1809/2181", "elapsed_time": "4h 47m 46s", "remaining_time": "59m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544257}
{"loss": 23.375, "grad_norm": 1.00655329, "learning_rate": 7.402e-05, "token_acc": 0.40106502, "epoch": 0.83018003, "global_step/max_steps": "1810/2181", "elapsed_time": "4h 47m 56s", "remaining_time": "59m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544508}
{"loss": 23.0, "grad_norm": 1.10395992, "learning_rate": 7.363e-05, "token_acc": 0.40051828, "epoch": 0.83063869, "global_step/max_steps": "1811/2181", "elapsed_time": "4h 48m 5s", "remaining_time": "58m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544414}
{"loss": 23.75, "grad_norm": 1.01424098, "learning_rate": 7.324e-05, "token_acc": 0.38806802, "epoch": 0.83109735, "global_step/max_steps": "1812/2181", "elapsed_time": "4h 48m 14s", "remaining_time": "58m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544136}
{"loss": 24.25, "grad_norm": 1.13836384, "learning_rate": 7.286e-05, "token_acc": 0.37914559, "epoch": 0.83155601, "global_step/max_steps": "1813/2181", "elapsed_time": "4h 48m 23s", "remaining_time": "58m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543999}
{"loss": 23.75, "grad_norm": 1.1509136, "learning_rate": 7.247e-05, "token_acc": 0.38624484, "epoch": 0.83201468, "global_step/max_steps": "1814/2181", "elapsed_time": "4h 48m 33s", "remaining_time": "58m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544014}
{"loss": 23.25, "grad_norm": 1.17834854, "learning_rate": 7.209e-05, "token_acc": 0.39053254, "epoch": 0.83247334, "global_step/max_steps": "1815/2181", "elapsed_time": "4h 48m 42s", "remaining_time": "58m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.543894}
{"loss": 23.375, "grad_norm": 1.05631864, "learning_rate": 7.17e-05, "token_acc": 0.40802961, "epoch": 0.832932, "global_step/max_steps": "1816/2181", "elapsed_time": "4h 48m 53s", "remaining_time": "58m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544502}
{"loss": 23.0, "grad_norm": 1.11215889, "learning_rate": 7.132e-05, "token_acc": 0.4090025, "epoch": 0.83339067, "global_step/max_steps": "1817/2181", "elapsed_time": "4h 49m 3s", "remaining_time": "57m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544638}
{"loss": 24.0, "grad_norm": 1.08216465, "learning_rate": 7.094e-05, "token_acc": 0.3892522, "epoch": 0.83384933, "global_step/max_steps": "1818/2181", "elapsed_time": "4h 49m 13s", "remaining_time": "57m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545111}
{"loss": 23.25, "grad_norm": 1.02189922, "learning_rate": 7.056e-05, "token_acc": 0.40078147, "epoch": 0.83430799, "global_step/max_steps": "1819/2181", "elapsed_time": "4h 49m 22s", "remaining_time": "57m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544886}
{"loss": 23.625, "grad_norm": 0.99703354, "learning_rate": 7.018e-05, "token_acc": 0.38789052, "epoch": 0.83476666, "global_step/max_steps": "1820/2181", "elapsed_time": "4h 49m 32s", "remaining_time": "57m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545127}
{"loss": 22.875, "grad_norm": 1.09374297, "learning_rate": 6.98e-05, "token_acc": 0.40132718, "epoch": 0.83522532, "global_step/max_steps": "1821/2181", "elapsed_time": "4h 49m 41s", "remaining_time": "57m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545009}
{"loss": 23.875, "grad_norm": 0.98956823, "learning_rate": 6.942e-05, "token_acc": 0.37962436, "epoch": 0.83568398, "global_step/max_steps": "1822/2181", "elapsed_time": "4h 49m 51s", "remaining_time": "57m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544849}
{"loss": 23.5, "grad_norm": 1.10402727, "learning_rate": 6.904e-05, "token_acc": 0.398557, "epoch": 0.83614264, "global_step/max_steps": "1823/2181", "elapsed_time": "4h 50m 0s", "remaining_time": "56m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544933}
{"loss": 24.5, "grad_norm": 1.07926702, "learning_rate": 6.867e-05, "token_acc": 0.36769948, "epoch": 0.83660131, "global_step/max_steps": "1824/2181", "elapsed_time": "4h 50m 10s", "remaining_time": "56m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544891}
{"loss": 23.75, "grad_norm": 0.9784652, "learning_rate": 6.829e-05, "token_acc": 0.38628258, "epoch": 0.83705997, "global_step/max_steps": "1825/2181", "elapsed_time": "4h 50m 19s", "remaining_time": "56m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544873}
{"loss": 24.125, "grad_norm": 1.0225774, "learning_rate": 6.792e-05, "token_acc": 0.3791655, "epoch": 0.83751863, "global_step/max_steps": "1826/2181", "elapsed_time": "4h 50m 29s", "remaining_time": "56m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544795}
{"loss": 24.75, "grad_norm": 1.11880934, "learning_rate": 6.755e-05, "token_acc": 0.37475345, "epoch": 0.8379773, "global_step/max_steps": "1827/2181", "elapsed_time": "4h 50m 39s", "remaining_time": "56m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545318}
{"loss": 23.625, "grad_norm": 0.99781799, "learning_rate": 6.717e-05, "token_acc": 0.39212513, "epoch": 0.83843596, "global_step/max_steps": "1828/2181", "elapsed_time": "4h 50m 48s", "remaining_time": "56m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545038}
{"loss": 24.125, "grad_norm": 1.063151, "learning_rate": 6.68e-05, "token_acc": 0.38054116, "epoch": 0.83889462, "global_step/max_steps": "1829/2181", "elapsed_time": "4h 50m 57s", "remaining_time": "55m 60s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544675}
{"loss": 24.0, "grad_norm": 1.07603729, "learning_rate": 6.643e-05, "token_acc": 0.38541068, "epoch": 0.83935329, "global_step/max_steps": "1830/2181", "elapsed_time": "4h 51m 8s", "remaining_time": "55m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545083}
{"loss": 23.625, "grad_norm": 1.08076262, "learning_rate": 6.606e-05, "token_acc": 0.38171589, "epoch": 0.83981195, "global_step/max_steps": "1831/2181", "elapsed_time": "4h 51m 17s", "remaining_time": "55m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545287}
{"loss": 24.375, "grad_norm": 1.05368686, "learning_rate": 6.569e-05, "token_acc": 0.37858935, "epoch": 0.84027061, "global_step/max_steps": "1832/2181", "elapsed_time": "4h 51m 27s", "remaining_time": "55m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545195}
{"loss": 23.5, "grad_norm": 1.07818758, "learning_rate": 6.533e-05, "token_acc": 0.38561721, "epoch": 0.84072927, "global_step/max_steps": "1833/2181", "elapsed_time": "4h 51m 36s", "remaining_time": "55m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545101}
{"loss": 24.25, "grad_norm": 1.24015725, "learning_rate": 6.496e-05, "token_acc": 0.3788269, "epoch": 0.84118794, "global_step/max_steps": "1834/2181", "elapsed_time": "4h 51m 46s", "remaining_time": "55m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544987}
{"loss": 23.25, "grad_norm": 1.14917922, "learning_rate": 6.459e-05, "token_acc": 0.41410437, "epoch": 0.8416466, "global_step/max_steps": "1835/2181", "elapsed_time": "4h 51m 55s", "remaining_time": "55m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544835}
{"loss": 24.5, "grad_norm": 1.13490713, "learning_rate": 6.423e-05, "token_acc": 0.36960651, "epoch": 0.84210526, "global_step/max_steps": "1836/2181", "elapsed_time": "4h 52m 4s", "remaining_time": "54m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544701}
{"loss": 24.25, "grad_norm": 1.06073463, "learning_rate": 6.387e-05, "token_acc": 0.37683977, "epoch": 0.84256393, "global_step/max_steps": "1837/2181", "elapsed_time": "4h 52m 14s", "remaining_time": "54m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544684}
{"loss": 23.75, "grad_norm": 0.95760596, "learning_rate": 6.35e-05, "token_acc": 0.38257576, "epoch": 0.84302259, "global_step/max_steps": "1838/2181", "elapsed_time": "4h 52m 23s", "remaining_time": "54m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544716}
{"loss": 24.25, "grad_norm": 1.21670246, "learning_rate": 6.314e-05, "token_acc": 0.38559205, "epoch": 0.84348125, "global_step/max_steps": "1839/2181", "elapsed_time": "4h 52m 33s", "remaining_time": "54m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544926}
{"loss": 24.0, "grad_norm": 0.99530441, "learning_rate": 6.278e-05, "token_acc": 0.37805907, "epoch": 0.84393992, "global_step/max_steps": "1840/2181", "elapsed_time": "4h 52m 43s", "remaining_time": "54m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545078}
{"loss": 24.125, "grad_norm": 1.01986301, "learning_rate": 6.242e-05, "token_acc": 0.37245314, "epoch": 0.84439858, "global_step/max_steps": "1841/2181", "elapsed_time": "4h 52m 52s", "remaining_time": "54m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545034}
{"loss": 23.625, "grad_norm": 1.0368284, "learning_rate": 6.206e-05, "token_acc": 0.38496072, "epoch": 0.84485724, "global_step/max_steps": "1842/2181", "elapsed_time": "4h 53m 2s", "remaining_time": "53m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545075}
{"loss": 23.875, "grad_norm": 0.9664945, "learning_rate": 6.17e-05, "token_acc": 0.38700722, "epoch": 0.8453159, "global_step/max_steps": "1843/2181", "elapsed_time": "4h 53m 12s", "remaining_time": "53m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545166}
{"loss": 23.5, "grad_norm": 1.04358435, "learning_rate": 6.135e-05, "token_acc": 0.39806087, "epoch": 0.84577457, "global_step/max_steps": "1844/2181", "elapsed_time": "4h 53m 21s", "remaining_time": "53m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545008}
{"loss": 24.0, "grad_norm": 1.11549318, "learning_rate": 6.099e-05, "token_acc": 0.38137033, "epoch": 0.84623323, "global_step/max_steps": "1845/2181", "elapsed_time": "4h 53m 30s", "remaining_time": "53m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544782}
{"loss": 24.25, "grad_norm": 1.10427129, "learning_rate": 6.064e-05, "token_acc": 0.39185751, "epoch": 0.84669189, "global_step/max_steps": "1846/2181", "elapsed_time": "4h 53m 40s", "remaining_time": "53m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544782}
{"loss": 23.25, "grad_norm": 0.99717218, "learning_rate": 6.028e-05, "token_acc": 0.39869281, "epoch": 0.84715056, "global_step/max_steps": "1847/2181", "elapsed_time": "4h 53m 49s", "remaining_time": "53m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.5446}
{"loss": 24.25, "grad_norm": 1.12456608, "learning_rate": 5.993e-05, "token_acc": 0.37774687, "epoch": 0.84760922, "global_step/max_steps": "1848/2181", "elapsed_time": "4h 53m 59s", "remaining_time": "52m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544797}
{"loss": 24.125, "grad_norm": 1.07419574, "learning_rate": 5.958e-05, "token_acc": 0.38477366, "epoch": 0.84806788, "global_step/max_steps": "1849/2181", "elapsed_time": "4h 54m 9s", "remaining_time": "52m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544951}
{"loss": 23.625, "grad_norm": 1.03108811, "learning_rate": 5.923e-05, "token_acc": 0.3898007, "epoch": 0.84852655, "global_step/max_steps": "1850/2181", "elapsed_time": "4h 54m 18s", "remaining_time": "52m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545084}
{"loss": 23.625, "grad_norm": 1.14805079, "learning_rate": 5.888e-05, "token_acc": 0.38231178, "epoch": 0.84898521, "global_step/max_steps": "1851/2181", "elapsed_time": "4h 54m 28s", "remaining_time": "52m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545157}
{"loss": 24.375, "grad_norm": 0.98422551, "learning_rate": 5.853e-05, "token_acc": 0.38197425, "epoch": 0.84944387, "global_step/max_steps": "1852/2181", "elapsed_time": "4h 54m 37s", "remaining_time": "52m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545041}
{"loss": 23.0, "grad_norm": 1.07004619, "learning_rate": 5.818e-05, "token_acc": 0.40069384, "epoch": 0.84990253, "global_step/max_steps": "1853/2181", "elapsed_time": "4h 54m 47s", "remaining_time": "52m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544961}
{"loss": 24.125, "grad_norm": 1.03874671, "learning_rate": 5.783e-05, "token_acc": 0.37426431, "epoch": 0.8503612, "global_step/max_steps": "1854/2181", "elapsed_time": "4h 54m 57s", "remaining_time": "52m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545321}
{"loss": 23.875, "grad_norm": 1.03164208, "learning_rate": 5.748e-05, "token_acc": 0.3722063, "epoch": 0.85081986, "global_step/max_steps": "1855/2181", "elapsed_time": "4h 55m 6s", "remaining_time": "51m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545068}
{"loss": 23.75, "grad_norm": 1.11696541, "learning_rate": 5.714e-05, "token_acc": 0.38315217, "epoch": 0.85127852, "global_step/max_steps": "1856/2181", "elapsed_time": "4h 55m 16s", "remaining_time": "51m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545233}
{"loss": 23.25, "grad_norm": 1.03361273, "learning_rate": 5.68e-05, "token_acc": 0.40180279, "epoch": 0.85173719, "global_step/max_steps": "1857/2181", "elapsed_time": "4h 55m 25s", "remaining_time": "51m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545075}
{"loss": 23.5, "grad_norm": 1.11318922, "learning_rate": 5.645e-05, "token_acc": 0.39246837, "epoch": 0.85219585, "global_step/max_steps": "1858/2181", "elapsed_time": "4h 55m 34s", "remaining_time": "51m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544885}
{"loss": 24.25, "grad_norm": 1.0218333, "learning_rate": 5.611e-05, "token_acc": 0.38213802, "epoch": 0.85265451, "global_step/max_steps": "1859/2181", "elapsed_time": "4h 55m 45s", "remaining_time": "51m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545443}
{"loss": 23.5, "grad_norm": 1.22325599, "learning_rate": 5.577e-05, "token_acc": 0.39961122, "epoch": 0.85311318, "global_step/max_steps": "1860/2181", "elapsed_time": "4h 55m 54s", "remaining_time": "51m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545085}
{"loss": 23.25, "grad_norm": 1.01669872, "learning_rate": 5.543e-05, "token_acc": 0.4038671, "epoch": 0.85357184, "global_step/max_steps": "1861/2181", "elapsed_time": "4h 56m 3s", "remaining_time": "50m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544888}
{"loss": 24.0, "grad_norm": 0.99213696, "learning_rate": 5.509e-05, "token_acc": 0.39364035, "epoch": 0.8540305, "global_step/max_steps": "1862/2181", "elapsed_time": "4h 56m 13s", "remaining_time": "50m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544937}
{"loss": 25.125, "grad_norm": 1.23065245, "learning_rate": 5.475e-05, "token_acc": 0.36361126, "epoch": 0.85448916, "global_step/max_steps": "1863/2181", "elapsed_time": "4h 56m 23s", "remaining_time": "50m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545251}
{"loss": 24.5, "grad_norm": 1.05071485, "learning_rate": 5.441e-05, "token_acc": 0.37075575, "epoch": 0.85494783, "global_step/max_steps": "1864/2181", "elapsed_time": "4h 56m 32s", "remaining_time": "50m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545155}
{"loss": 24.25, "grad_norm": 1.09338307, "learning_rate": 5.408e-05, "token_acc": 0.377274, "epoch": 0.85540649, "global_step/max_steps": "1865/2181", "elapsed_time": "4h 56m 42s", "remaining_time": "50m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545213}
{"loss": 23.375, "grad_norm": 0.97509617, "learning_rate": 5.374e-05, "token_acc": 0.3920175, "epoch": 0.85586515, "global_step/max_steps": "1866/2181", "elapsed_time": "4h 56m 51s", "remaining_time": "50m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545276}
{"loss": 23.375, "grad_norm": 0.99504501, "learning_rate": 5.341e-05, "token_acc": 0.40279361, "epoch": 0.85632382, "global_step/max_steps": "1867/2181", "elapsed_time": "4h 57m 1s", "remaining_time": "49m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545168}
{"loss": 24.125, "grad_norm": 1.10896266, "learning_rate": 5.307e-05, "token_acc": 0.38120617, "epoch": 0.85678248, "global_step/max_steps": "1868/2181", "elapsed_time": "4h 57m 10s", "remaining_time": "49m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545124}
{"loss": 24.25, "grad_norm": 1.08804786, "learning_rate": 5.274e-05, "token_acc": 0.39377239, "epoch": 0.85724114, "global_step/max_steps": "1869/2181", "elapsed_time": "4h 57m 20s", "remaining_time": "49m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.544984}
{"loss": 23.875, "grad_norm": 1.01424909, "learning_rate": 5.241e-05, "token_acc": 0.38350056, "epoch": 0.85769981, "global_step/max_steps": "1870/2181", "elapsed_time": "4h 57m 30s", "remaining_time": "49m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545417}
{"loss": 23.625, "grad_norm": 1.03454483, "learning_rate": 5.208e-05, "token_acc": 0.39056604, "epoch": 0.85815847, "global_step/max_steps": "1871/2181", "elapsed_time": "4h 57m 40s", "remaining_time": "49m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545855}
{"loss": 24.0, "grad_norm": 1.17545581, "learning_rate": 5.175e-05, "token_acc": 0.38625526, "epoch": 0.85861713, "global_step/max_steps": "1872/2181", "elapsed_time": "4h 57m 50s", "remaining_time": "49m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546033}
{"loss": 23.875, "grad_norm": 1.01768279, "learning_rate": 5.142e-05, "token_acc": 0.38396861, "epoch": 0.85907579, "global_step/max_steps": "1873/2181", "elapsed_time": "4h 57m 59s", "remaining_time": "49m 0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545906}
{"loss": 23.875, "grad_norm": 0.96611738, "learning_rate": 5.109e-05, "token_acc": 0.38830219, "epoch": 0.85953446, "global_step/max_steps": "1874/2181", "elapsed_time": "4h 58m 9s", "remaining_time": "48m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546002}
{"loss": 23.5, "grad_norm": 1.06572556, "learning_rate": 5.077e-05, "token_acc": 0.39270687, "epoch": 0.85999312, "global_step/max_steps": "1875/2181", "elapsed_time": "4h 58m 19s", "remaining_time": "48m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545921}
{"loss": 23.0, "grad_norm": 1.13275707, "learning_rate": 5.044e-05, "token_acc": 0.39566086, "epoch": 0.86045178, "global_step/max_steps": "1876/2181", "elapsed_time": "4h 58m 28s", "remaining_time": "48m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545955}
{"loss": 23.875, "grad_norm": 1.08025753, "learning_rate": 5.012e-05, "token_acc": 0.3818602, "epoch": 0.86091045, "global_step/max_steps": "1877/2181", "elapsed_time": "4h 58m 37s", "remaining_time": "48m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545789}
{"loss": 24.375, "grad_norm": 1.33217931, "learning_rate": 4.979e-05, "token_acc": 0.37292291, "epoch": 0.86136911, "global_step/max_steps": "1878/2181", "elapsed_time": "4h 58m 48s", "remaining_time": "48m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546095}
{"loss": 23.5, "grad_norm": 1.10312021, "learning_rate": 4.947e-05, "token_acc": 0.39634146, "epoch": 0.86182777, "global_step/max_steps": "1879/2181", "elapsed_time": "4h 58m 57s", "remaining_time": "48m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54628}
{"loss": 23.25, "grad_norm": 1.09017181, "learning_rate": 4.915e-05, "token_acc": 0.39051406, "epoch": 0.86228644, "global_step/max_steps": "1880/2181", "elapsed_time": "4h 59m 7s", "remaining_time": "47m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54615}
{"loss": 23.25, "grad_norm": 1.04010701, "learning_rate": 4.883e-05, "token_acc": 0.39391401, "epoch": 0.8627451, "global_step/max_steps": "1881/2181", "elapsed_time": "4h 59m 16s", "remaining_time": "47m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546225}
{"loss": 23.625, "grad_norm": 1.27151167, "learning_rate": 4.851e-05, "token_acc": 0.38717134, "epoch": 0.86320376, "global_step/max_steps": "1882/2181", "elapsed_time": "4h 59m 27s", "remaining_time": "47m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546531}
{"loss": 23.875, "grad_norm": 1.09092522, "learning_rate": 4.819e-05, "token_acc": 0.38850376, "epoch": 0.86366242, "global_step/max_steps": "1883/2181", "elapsed_time": "4h 59m 36s", "remaining_time": "47m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546473}
{"loss": 23.625, "grad_norm": 1.08615315, "learning_rate": 4.787e-05, "token_acc": 0.39698217, "epoch": 0.86412109, "global_step/max_steps": "1884/2181", "elapsed_time": "4h 59m 46s", "remaining_time": "47m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546582}
{"loss": 23.25, "grad_norm": 0.99666327, "learning_rate": 4.755e-05, "token_acc": 0.39554553, "epoch": 0.86457975, "global_step/max_steps": "1885/2181", "elapsed_time": "4h 59m 55s", "remaining_time": "47m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546375}
{"loss": 23.625, "grad_norm": 1.14125764, "learning_rate": 4.724e-05, "token_acc": 0.38604914, "epoch": 0.86503841, "global_step/max_steps": "1886/2181", "elapsed_time": "5h 0m 4s", "remaining_time": "46m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545945}
{"loss": 23.5, "grad_norm": 1.14842021, "learning_rate": 4.692e-05, "token_acc": 0.39977318, "epoch": 0.86549708, "global_step/max_steps": "1887/2181", "elapsed_time": "5h 0m 14s", "remaining_time": "46m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546227}
{"loss": 24.0, "grad_norm": 1.06799746, "learning_rate": 4.661e-05, "token_acc": 0.37843193, "epoch": 0.86595574, "global_step/max_steps": "1888/2181", "elapsed_time": "5h 0m 23s", "remaining_time": "46m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545933}
{"loss": 23.375, "grad_norm": 1.05275393, "learning_rate": 4.63e-05, "token_acc": 0.39728353, "epoch": 0.8664144, "global_step/max_steps": "1889/2181", "elapsed_time": "5h 0m 32s", "remaining_time": "46m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545714}
{"loss": 25.0, "grad_norm": 1.02792311, "learning_rate": 4.599e-05, "token_acc": 0.37784011, "epoch": 0.86687307, "global_step/max_steps": "1890/2181", "elapsed_time": "5h 0m 41s", "remaining_time": "46m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545662}
{"loss": 23.5, "grad_norm": 1.07136488, "learning_rate": 4.568e-05, "token_acc": 0.40960135, "epoch": 0.86733173, "global_step/max_steps": "1891/2181", "elapsed_time": "5h 0m 51s", "remaining_time": "46m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545718}
{"loss": 24.5, "grad_norm": 1.15275502, "learning_rate": 4.537e-05, "token_acc": 0.37569677, "epoch": 0.86779039, "global_step/max_steps": "1892/2181", "elapsed_time": "5h 1m 1s", "remaining_time": "45m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546066}
{"loss": 24.375, "grad_norm": 1.22250664, "learning_rate": 4.506e-05, "token_acc": 0.3761236, "epoch": 0.86824905, "global_step/max_steps": "1893/2181", "elapsed_time": "5h 1m 11s", "remaining_time": "45m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546026}
{"loss": 24.125, "grad_norm": 1.20747042, "learning_rate": 4.475e-05, "token_acc": 0.38472622, "epoch": 0.86870772, "global_step/max_steps": "1894/2181", "elapsed_time": "5h 1m 20s", "remaining_time": "45m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546142}
{"loss": 24.75, "grad_norm": 1.20543981, "learning_rate": 4.444e-05, "token_acc": 0.38663685, "epoch": 0.86916638, "global_step/max_steps": "1895/2181", "elapsed_time": "5h 1m 30s", "remaining_time": "45m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546012}
{"loss": 23.875, "grad_norm": 1.06569421, "learning_rate": 4.414e-05, "token_acc": 0.39135255, "epoch": 0.86962504, "global_step/max_steps": "1896/2181", "elapsed_time": "5h 1m 39s", "remaining_time": "45m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545967}
{"loss": 24.5, "grad_norm": 1.12561357, "learning_rate": 4.383e-05, "token_acc": 0.39251548, "epoch": 0.87008371, "global_step/max_steps": "1897/2181", "elapsed_time": "5h 1m 49s", "remaining_time": "45m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546005}
{"loss": 24.125, "grad_norm": 1.10021722, "learning_rate": 4.353e-05, "token_acc": 0.3851728, "epoch": 0.87054237, "global_step/max_steps": "1898/2181", "elapsed_time": "5h 1m 59s", "remaining_time": "45m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546205}
{"loss": 23.125, "grad_norm": 1.03705621, "learning_rate": 4.323e-05, "token_acc": 0.390625, "epoch": 0.87100103, "global_step/max_steps": "1899/2181", "elapsed_time": "5h 2m 8s", "remaining_time": "44m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.545986}
{"loss": 23.75, "grad_norm": 1.05563343, "learning_rate": 4.293e-05, "token_acc": 0.3914641, "epoch": 0.87145969, "global_step/max_steps": "1900/2181", "elapsed_time": "5h 2m 18s", "remaining_time": "44m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546242}
{"loss": 23.5, "grad_norm": 1.02244794, "learning_rate": 4.263e-05, "token_acc": 0.40288969, "epoch": 0.87191836, "global_step/max_steps": "1901/2181", "elapsed_time": "5h 2m 27s", "remaining_time": "44m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54619}
{"loss": 23.75, "grad_norm": 1.02669096, "learning_rate": 4.233e-05, "token_acc": 0.38426712, "epoch": 0.87237702, "global_step/max_steps": "1902/2181", "elapsed_time": "5h 2m 37s", "remaining_time": "44m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546332}
{"loss": 23.0, "grad_norm": 1.06594241, "learning_rate": 4.203e-05, "token_acc": 0.39510086, "epoch": 0.87283568, "global_step/max_steps": "1903/2181", "elapsed_time": "5h 2m 46s", "remaining_time": "44m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546119}
{"loss": 24.0, "grad_norm": 1.05532503, "learning_rate": 4.173e-05, "token_acc": 0.38977901, "epoch": 0.87329435, "global_step/max_steps": "1904/2181", "elapsed_time": "5h 2m 56s", "remaining_time": "44m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546352}
{"loss": 23.625, "grad_norm": 1.04627335, "learning_rate": 4.143e-05, "token_acc": 0.39314572, "epoch": 0.87375301, "global_step/max_steps": "1905/2181", "elapsed_time": "5h 3m 6s", "remaining_time": "43m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546506}
{"loss": 23.375, "grad_norm": 1.04773128, "learning_rate": 4.114e-05, "token_acc": 0.40067152, "epoch": 0.87421167, "global_step/max_steps": "1906/2181", "elapsed_time": "5h 3m 15s", "remaining_time": "43m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546294}
{"loss": 24.375, "grad_norm": 1.12716699, "learning_rate": 4.084e-05, "token_acc": 0.37877551, "epoch": 0.87467034, "global_step/max_steps": "1907/2181", "elapsed_time": "5h 3m 25s", "remaining_time": "43m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546361}
{"loss": 23.25, "grad_norm": 0.97240341, "learning_rate": 4.055e-05, "token_acc": 0.40098334, "epoch": 0.875129, "global_step/max_steps": "1908/2181", "elapsed_time": "5h 3m 35s", "remaining_time": "43m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546433}
{"loss": 24.0, "grad_norm": 1.05968177, "learning_rate": 4.026e-05, "token_acc": 0.38579111, "epoch": 0.87558766, "global_step/max_steps": "1909/2181", "elapsed_time": "5h 3m 44s", "remaining_time": "43m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54646}
{"loss": 23.125, "grad_norm": 1.08412504, "learning_rate": 3.997e-05, "token_acc": 0.40227968, "epoch": 0.87604632, "global_step/max_steps": "1910/2181", "elapsed_time": "5h 3m 54s", "remaining_time": "43m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546649}
{"loss": 23.75, "grad_norm": 1.09628272, "learning_rate": 3.968e-05, "token_acc": 0.37611, "epoch": 0.87650499, "global_step/max_steps": "1911/2181", "elapsed_time": "5h 4m 4s", "remaining_time": "42m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546678}
{"loss": 23.75, "grad_norm": 1.09054756, "learning_rate": 3.939e-05, "token_acc": 0.37831669, "epoch": 0.87696365, "global_step/max_steps": "1912/2181", "elapsed_time": "5h 4m 14s", "remaining_time": "42m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546838}
{"loss": 23.5, "grad_norm": 0.96721745, "learning_rate": 3.91e-05, "token_acc": 0.38661088, "epoch": 0.87742231, "global_step/max_steps": "1913/2181", "elapsed_time": "5h 4m 23s", "remaining_time": "42m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546777}
{"loss": 24.25, "grad_norm": 0.93909508, "learning_rate": 3.881e-05, "token_acc": 0.37956605, "epoch": 0.87788098, "global_step/max_steps": "1914/2181", "elapsed_time": "5h 4m 33s", "remaining_time": "42m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546765}
{"loss": 24.125, "grad_norm": 1.07481623, "learning_rate": 3.852e-05, "token_acc": 0.37441021, "epoch": 0.87833964, "global_step/max_steps": "1915/2181", "elapsed_time": "5h 4m 42s", "remaining_time": "42m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.546964}
{"loss": 23.875, "grad_norm": 1.14126432, "learning_rate": 3.824e-05, "token_acc": 0.38995929, "epoch": 0.8787983, "global_step/max_steps": "1916/2181", "elapsed_time": "5h 4m 52s", "remaining_time": "42m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547198}
{"loss": 23.5, "grad_norm": 1.10929275, "learning_rate": 3.795e-05, "token_acc": 0.39596574, "epoch": 0.87925697, "global_step/max_steps": "1917/2181", "elapsed_time": "5h 5m 2s", "remaining_time": "42m 0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547082}
{"loss": 23.75, "grad_norm": 1.03293121, "learning_rate": 3.767e-05, "token_acc": 0.39240155, "epoch": 0.87971563, "global_step/max_steps": "1918/2181", "elapsed_time": "5h 5m 12s", "remaining_time": "41m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547353}
{"loss": 23.25, "grad_norm": 1.06487489, "learning_rate": 3.739e-05, "token_acc": 0.39095979, "epoch": 0.88017429, "global_step/max_steps": "1919/2181", "elapsed_time": "5h 5m 21s", "remaining_time": "41m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547367}
{"loss": 24.75, "grad_norm": 1.08352923, "learning_rate": 3.711e-05, "token_acc": 0.38842258, "epoch": 0.88063295, "global_step/max_steps": "1920/2181", "elapsed_time": "5h 5m 31s", "remaining_time": "41m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54761}
{"loss": 24.125, "grad_norm": 1.07120621, "learning_rate": 3.683e-05, "token_acc": 0.37068239, "epoch": 0.88109162, "global_step/max_steps": "1921/2181", "elapsed_time": "5h 5m 40s", "remaining_time": "41m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547279}
{"loss": 24.25, "grad_norm": 1.07450008, "learning_rate": 3.655e-05, "token_acc": 0.37299312, "epoch": 0.88155028, "global_step/max_steps": "1922/2181", "elapsed_time": "5h 5m 50s", "remaining_time": "41m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547386}
{"loss": 23.5, "grad_norm": 1.03781021, "learning_rate": 3.627e-05, "token_acc": 0.39014202, "epoch": 0.88200894, "global_step/max_steps": "1923/2181", "elapsed_time": "5h 5m 60s", "remaining_time": "41m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547436}
{"loss": 24.0, "grad_norm": 1.00042295, "learning_rate": 3.599e-05, "token_acc": 0.39021679, "epoch": 0.88246761, "global_step/max_steps": "1924/2181", "elapsed_time": "5h 6m 10s", "remaining_time": "40m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547684}
{"loss": 24.25, "grad_norm": 1.17154872, "learning_rate": 3.572e-05, "token_acc": 0.38607595, "epoch": 0.88292627, "global_step/max_steps": "1925/2181", "elapsed_time": "5h 6m 19s", "remaining_time": "40m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547369}
{"loss": 23.125, "grad_norm": 0.99896485, "learning_rate": 3.544e-05, "token_acc": 0.40497676, "epoch": 0.88338493, "global_step/max_steps": "1926/2181", "elapsed_time": "5h 6m 28s", "remaining_time": "40m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547335}
{"loss": 23.0, "grad_norm": 1.09684443, "learning_rate": 3.517e-05, "token_acc": 0.39248957, "epoch": 0.8838436, "global_step/max_steps": "1927/2181", "elapsed_time": "5h 6m 38s", "remaining_time": "40m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547343}
{"loss": 24.25, "grad_norm": 1.01800215, "learning_rate": 3.489e-05, "token_acc": 0.37673611, "epoch": 0.88430226, "global_step/max_steps": "1928/2181", "elapsed_time": "5h 6m 47s", "remaining_time": "40m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547407}
{"loss": 23.625, "grad_norm": 1.05804253, "learning_rate": 3.462e-05, "token_acc": 0.38916256, "epoch": 0.88476092, "global_step/max_steps": "1929/2181", "elapsed_time": "5h 6m 57s", "remaining_time": "40m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547469}
{"loss": 23.5, "grad_norm": 1.09355891, "learning_rate": 3.435e-05, "token_acc": 0.40178323, "epoch": 0.88521958, "global_step/max_steps": "1930/2181", "elapsed_time": "5h 7m 6s", "remaining_time": "39m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547182}
{"loss": 23.0, "grad_norm": 1.01097178, "learning_rate": 3.408e-05, "token_acc": 0.40441989, "epoch": 0.88567825, "global_step/max_steps": "1931/2181", "elapsed_time": "5h 7m 16s", "remaining_time": "39m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547315}
{"loss": 24.0, "grad_norm": 1.10184312, "learning_rate": 3.381e-05, "token_acc": 0.37440493, "epoch": 0.88613691, "global_step/max_steps": "1932/2181", "elapsed_time": "5h 7m 26s", "remaining_time": "39m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547485}
{"loss": 24.5, "grad_norm": 1.06348169, "learning_rate": 3.354e-05, "token_acc": 0.37676349, "epoch": 0.88659557, "global_step/max_steps": "1933/2181", "elapsed_time": "5h 7m 36s", "remaining_time": "39m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547928}
{"loss": 23.75, "grad_norm": 0.96254063, "learning_rate": 3.328e-05, "token_acc": 0.38922821, "epoch": 0.88705424, "global_step/max_steps": "1934/2181", "elapsed_time": "5h 7m 47s", "remaining_time": "39m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548384}
{"loss": 23.625, "grad_norm": 1.08432281, "learning_rate": 3.301e-05, "token_acc": 0.39852189, "epoch": 0.8875129, "global_step/max_steps": "1935/2181", "elapsed_time": "5h 7m 57s", "remaining_time": "39m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548639}
{"loss": 23.875, "grad_norm": 1.07377553, "learning_rate": 3.275e-05, "token_acc": 0.38935811, "epoch": 0.88797156, "global_step/max_steps": "1936/2181", "elapsed_time": "5h 8m 7s", "remaining_time": "38m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548878}
{"loss": 24.125, "grad_norm": 1.11085963, "learning_rate": 3.248e-05, "token_acc": 0.38901537, "epoch": 0.88843023, "global_step/max_steps": "1937/2181", "elapsed_time": "5h 8m 16s", "remaining_time": "38m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548895}
{"loss": 23.625, "grad_norm": 1.08507764, "learning_rate": 3.222e-05, "token_acc": 0.3966305, "epoch": 0.88888889, "global_step/max_steps": "1938/2181", "elapsed_time": "5h 8m 26s", "remaining_time": "38m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548996}
{"loss": 23.75, "grad_norm": 1.03146064, "learning_rate": 3.196e-05, "token_acc": 0.38254486, "epoch": 0.88934755, "global_step/max_steps": "1939/2181", "elapsed_time": "5h 8m 35s", "remaining_time": "38m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548828}
{"loss": 24.0, "grad_norm": 0.9964084, "learning_rate": 3.17e-05, "token_acc": 0.37150056, "epoch": 0.88980621, "global_step/max_steps": "1940/2181", "elapsed_time": "5h 8m 45s", "remaining_time": "38m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548967}
{"loss": 23.75, "grad_norm": 1.05289412, "learning_rate": 3.144e-05, "token_acc": 0.40131579, "epoch": 0.89026488, "global_step/max_steps": "1941/2181", "elapsed_time": "5h 8m 54s", "remaining_time": "38m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548833}
{"loss": 23.25, "grad_norm": 1.11287117, "learning_rate": 3.118e-05, "token_acc": 0.39450305, "epoch": 0.89072354, "global_step/max_steps": "1942/2181", "elapsed_time": "5h 9m 3s", "remaining_time": "38m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548464}
{"loss": 23.875, "grad_norm": 1.10911655, "learning_rate": 3.092e-05, "token_acc": 0.3750705, "epoch": 0.8911822, "global_step/max_steps": "1943/2181", "elapsed_time": "5h 9m 13s", "remaining_time": "37m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548584}
{"loss": 23.75, "grad_norm": 0.97167921, "learning_rate": 3.066e-05, "token_acc": 0.38728484, "epoch": 0.89164087, "global_step/max_steps": "1944/2181", "elapsed_time": "5h 9m 23s", "remaining_time": "37m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548772}
{"loss": 24.5, "grad_norm": 1.32144237, "learning_rate": 3.041e-05, "token_acc": 0.38896629, "epoch": 0.89209953, "global_step/max_steps": "1945/2181", "elapsed_time": "5h 9m 32s", "remaining_time": "37m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548591}
{"loss": 23.875, "grad_norm": 1.01386368, "learning_rate": 3.015e-05, "token_acc": 0.38912926, "epoch": 0.89255819, "global_step/max_steps": "1946/2181", "elapsed_time": "5h 9m 42s", "remaining_time": "37m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548793}
{"loss": 23.125, "grad_norm": 0.98648834, "learning_rate": 2.99e-05, "token_acc": 0.3758051, "epoch": 0.89301686, "global_step/max_steps": "1947/2181", "elapsed_time": "5h 9m 51s", "remaining_time": "37m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548624}
{"loss": 24.125, "grad_norm": 0.98678297, "learning_rate": 2.965e-05, "token_acc": 0.39179207, "epoch": 0.89347552, "global_step/max_steps": "1948/2181", "elapsed_time": "5h 10m 1s", "remaining_time": "37m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548768}
{"loss": 23.625, "grad_norm": 0.97898394, "learning_rate": 2.94e-05, "token_acc": 0.39197361, "epoch": 0.89393418, "global_step/max_steps": "1949/2181", "elapsed_time": "5h 10m 10s", "remaining_time": "36m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548606}
{"loss": 22.875, "grad_norm": 1.0784837, "learning_rate": 2.915e-05, "token_acc": 0.40963855, "epoch": 0.89439284, "global_step/max_steps": "1950/2181", "elapsed_time": "5h 10m 20s", "remaining_time": "36m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54852}
{"loss": 23.75, "grad_norm": 1.06540322, "learning_rate": 2.89e-05, "token_acc": 0.38987688, "epoch": 0.89485151, "global_step/max_steps": "1951/2181", "elapsed_time": "5h 10m 29s", "remaining_time": "36m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548483}
{"loss": 24.0, "grad_norm": 1.12973762, "learning_rate": 2.865e-05, "token_acc": 0.37156751, "epoch": 0.89531017, "global_step/max_steps": "1952/2181", "elapsed_time": "5h 10m 39s", "remaining_time": "36m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548645}
{"loss": 23.5, "grad_norm": 0.90450555, "learning_rate": 2.84e-05, "token_acc": 0.3933199, "epoch": 0.89576883, "global_step/max_steps": "1953/2181", "elapsed_time": "5h 10m 49s", "remaining_time": "36m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548976}
{"loss": 24.375, "grad_norm": 0.99631119, "learning_rate": 2.815e-05, "token_acc": 0.38565629, "epoch": 0.8962275, "global_step/max_steps": "1954/2181", "elapsed_time": "5h 10m 59s", "remaining_time": "36m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548898}
{"loss": 23.0, "grad_norm": 1.00934088, "learning_rate": 2.791e-05, "token_acc": 0.40420694, "epoch": 0.89668616, "global_step/max_steps": "1955/2181", "elapsed_time": "5h 11m 8s", "remaining_time": "35m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548893}
{"loss": 24.125, "grad_norm": 0.99677891, "learning_rate": 2.767e-05, "token_acc": 0.37702703, "epoch": 0.89714482, "global_step/max_steps": "1956/2181", "elapsed_time": "5h 11m 17s", "remaining_time": "35m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548586}
{"loss": 22.875, "grad_norm": 1.03847623, "learning_rate": 2.742e-05, "token_acc": 0.39861111, "epoch": 0.89760349, "global_step/max_steps": "1957/2181", "elapsed_time": "5h 11m 26s", "remaining_time": "35m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548536}
{"loss": 23.375, "grad_norm": 1.06317866, "learning_rate": 2.718e-05, "token_acc": 0.39940266, "epoch": 0.89806215, "global_step/max_steps": "1958/2181", "elapsed_time": "5h 11m 36s", "remaining_time": "35m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548276}
{"loss": 24.125, "grad_norm": 1.09471416, "learning_rate": 2.694e-05, "token_acc": 0.37136648, "epoch": 0.89852081, "global_step/max_steps": "1959/2181", "elapsed_time": "5h 11m 45s", "remaining_time": "35m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548274}
{"loss": 24.125, "grad_norm": 1.00324881, "learning_rate": 2.67e-05, "token_acc": 0.39816105, "epoch": 0.89897947, "global_step/max_steps": "1960/2181", "elapsed_time": "5h 11m 54s", "remaining_time": "35m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548}
{"loss": 24.0, "grad_norm": 1.02188814, "learning_rate": 2.646e-05, "token_acc": 0.38847532, "epoch": 0.89943814, "global_step/max_steps": "1961/2181", "elapsed_time": "5h 12m 3s", "remaining_time": "35m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547753}
{"loss": 24.0, "grad_norm": 1.09208393, "learning_rate": 2.622e-05, "token_acc": 0.37951807, "epoch": 0.8998968, "global_step/max_steps": "1962/2181", "elapsed_time": "5h 12m 13s", "remaining_time": "34m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54798}
{"loss": 23.375, "grad_norm": 1.08095372, "learning_rate": 2.599e-05, "token_acc": 0.39161438, "epoch": 0.90035546, "global_step/max_steps": "1963/2181", "elapsed_time": "5h 12m 22s", "remaining_time": "34m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547863}
{"loss": 23.5, "grad_norm": 1.13341033, "learning_rate": 2.575e-05, "token_acc": 0.38836612, "epoch": 0.90081413, "global_step/max_steps": "1964/2181", "elapsed_time": "5h 12m 33s", "remaining_time": "34m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548197}
{"loss": 24.125, "grad_norm": 1.02767217, "learning_rate": 2.552e-05, "token_acc": 0.395189, "epoch": 0.90127279, "global_step/max_steps": "1965/2181", "elapsed_time": "5h 12m 42s", "remaining_time": "34m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.54832}
{"loss": 23.5, "grad_norm": 1.04606998, "learning_rate": 2.528e-05, "token_acc": 0.39439038, "epoch": 0.90173145, "global_step/max_steps": "1966/2181", "elapsed_time": "5h 12m 52s", "remaining_time": "34m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548568}
{"loss": 23.5, "grad_norm": 1.25872314, "learning_rate": 2.505e-05, "token_acc": 0.3923946, "epoch": 0.90219012, "global_step/max_steps": "1967/2181", "elapsed_time": "5h 13m 2s", "remaining_time": "34m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548655}
{"loss": 23.5, "grad_norm": 1.06762922, "learning_rate": 2.482e-05, "token_acc": 0.39467724, "epoch": 0.90264878, "global_step/max_steps": "1968/2181", "elapsed_time": "5h 13m 12s", "remaining_time": "33m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548689}
{"loss": 24.625, "grad_norm": 1.02491474, "learning_rate": 2.459e-05, "token_acc": 0.36859414, "epoch": 0.90310744, "global_step/max_steps": "1969/2181", "elapsed_time": "5h 13m 21s", "remaining_time": "33m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548389}
{"loss": 23.75, "grad_norm": 1.01204729, "learning_rate": 2.436e-05, "token_acc": 0.38603352, "epoch": 0.9035661, "global_step/max_steps": "1970/2181", "elapsed_time": "5h 13m 31s", "remaining_time": "33m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548551}
{"loss": 23.5, "grad_norm": 1.00660646, "learning_rate": 2.413e-05, "token_acc": 0.39288691, "epoch": 0.90402477, "global_step/max_steps": "1971/2181", "elapsed_time": "5h 13m 40s", "remaining_time": "33m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548467}
{"loss": 22.75, "grad_norm": 1.08403325, "learning_rate": 2.39e-05, "token_acc": 0.40128996, "epoch": 0.90448343, "global_step/max_steps": "1972/2181", "elapsed_time": "5h 13m 49s", "remaining_time": "33m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548227}
{"loss": 24.25, "grad_norm": 1.09982228, "learning_rate": 2.367e-05, "token_acc": 0.37991747, "epoch": 0.90494209, "global_step/max_steps": "1973/2181", "elapsed_time": "5h 13m 59s", "remaining_time": "33m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548169}
{"loss": 24.75, "grad_norm": 1.02156413, "learning_rate": 2.345e-05, "token_acc": 0.37196314, "epoch": 0.90540076, "global_step/max_steps": "1974/2181", "elapsed_time": "5h 14m 8s", "remaining_time": "32m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548124}
{"loss": 24.0, "grad_norm": 1.05927205, "learning_rate": 2.323e-05, "token_acc": 0.38824481, "epoch": 0.90585942, "global_step/max_steps": "1975/2181", "elapsed_time": "5h 14m 17s", "remaining_time": "32m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548059}
{"loss": 23.75, "grad_norm": 1.04236662, "learning_rate": 2.3e-05, "token_acc": 0.38835227, "epoch": 0.90631808, "global_step/max_steps": "1976/2181", "elapsed_time": "5h 14m 27s", "remaining_time": "32m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547989}
{"loss": 23.75, "grad_norm": 1.10253787, "learning_rate": 2.278e-05, "token_acc": 0.37921925, "epoch": 0.90677675, "global_step/max_steps": "1977/2181", "elapsed_time": "5h 14m 37s", "remaining_time": "32m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548057}
{"loss": 24.125, "grad_norm": 1.06308424, "learning_rate": 2.256e-05, "token_acc": 0.37428494, "epoch": 0.90723541, "global_step/max_steps": "1978/2181", "elapsed_time": "5h 14m 46s", "remaining_time": "32m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547826}
{"loss": 23.625, "grad_norm": 1.08575833, "learning_rate": 2.234e-05, "token_acc": 0.39088663, "epoch": 0.90769407, "global_step/max_steps": "1979/2181", "elapsed_time": "5h 14m 55s", "remaining_time": "32m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547869}
{"loss": 24.5, "grad_norm": 1.10060751, "learning_rate": 2.212e-05, "token_acc": 0.37795714, "epoch": 0.90815273, "global_step/max_steps": "1980/2181", "elapsed_time": "5h 15m 4s", "remaining_time": "31m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547674}
{"loss": 24.125, "grad_norm": 1.5378294, "learning_rate": 2.19e-05, "token_acc": 0.37517053, "epoch": 0.9086114, "global_step/max_steps": "1981/2181", "elapsed_time": "5h 15m 14s", "remaining_time": "31m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547724}
{"loss": 23.875, "grad_norm": 0.93274587, "learning_rate": 2.169e-05, "token_acc": 0.38823206, "epoch": 0.90907006, "global_step/max_steps": "1982/2181", "elapsed_time": "5h 15m 24s", "remaining_time": "31m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547692}
{"loss": 23.5, "grad_norm": 0.99405098, "learning_rate": 2.147e-05, "token_acc": 0.3891455, "epoch": 0.90952872, "global_step/max_steps": "1983/2181", "elapsed_time": "5h 15m 33s", "remaining_time": "31m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547873}
{"loss": 24.125, "grad_norm": 1.09600985, "learning_rate": 2.125e-05, "token_acc": 0.38302101, "epoch": 0.90998739, "global_step/max_steps": "1984/2181", "elapsed_time": "5h 15m 43s", "remaining_time": "31m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548032}
{"loss": 23.875, "grad_norm": 1.0168488, "learning_rate": 2.104e-05, "token_acc": 0.39134276, "epoch": 0.91044605, "global_step/max_steps": "1985/2181", "elapsed_time": "5h 15m 53s", "remaining_time": "31m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547886}
{"loss": 23.125, "grad_norm": 1.14965081, "learning_rate": 2.083e-05, "token_acc": 0.40289934, "epoch": 0.91090471, "global_step/max_steps": "1986/2181", "elapsed_time": "5h 16m 2s", "remaining_time": "31m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547933}
{"loss": 23.5, "grad_norm": 1.02347505, "learning_rate": 2.062e-05, "token_acc": 0.37973274, "epoch": 0.91136338, "global_step/max_steps": "1987/2181", "elapsed_time": "5h 16m 12s", "remaining_time": "30m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548038}
{"loss": 23.875, "grad_norm": 1.10749733, "learning_rate": 2.041e-05, "token_acc": 0.38668516, "epoch": 0.91182204, "global_step/max_steps": "1988/2181", "elapsed_time": "5h 16m 22s", "remaining_time": "30m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548127}
{"loss": 24.0, "grad_norm": 1.06548524, "learning_rate": 2.02e-05, "token_acc": 0.38783375, "epoch": 0.9122807, "global_step/max_steps": "1989/2181", "elapsed_time": "5h 16m 31s", "remaining_time": "30m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.548241}
{"loss": 24.0, "grad_norm": 1.03842366, "learning_rate": 1.999e-05, "token_acc": 0.37775915, "epoch": 0.91273936, "global_step/max_steps": "1990/2181", "elapsed_time": "5h 16m 40s", "remaining_time": "30m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547871}
{"loss": 23.625, "grad_norm": 1.01057839, "learning_rate": 1.978e-05, "token_acc": 0.37685291, "epoch": 0.91319803, "global_step/max_steps": "1991/2181", "elapsed_time": "5h 16m 49s", "remaining_time": "30m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547712}
{"loss": 23.375, "grad_norm": 1.02660251, "learning_rate": 1.957e-05, "token_acc": 0.39460993, "epoch": 0.91365669, "global_step/max_steps": "1992/2181", "elapsed_time": "5h 16m 59s", "remaining_time": "30m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547754}
{"loss": 23.625, "grad_norm": 1.05795753, "learning_rate": 1.937e-05, "token_acc": 0.39146205, "epoch": 0.91411535, "global_step/max_steps": "1993/2181", "elapsed_time": "5h 17m 8s", "remaining_time": "29m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547637}
{"loss": 23.375, "grad_norm": 1.01508725, "learning_rate": 1.916e-05, "token_acc": 0.38594092, "epoch": 0.91457402, "global_step/max_steps": "1994/2181", "elapsed_time": "5h 17m 18s", "remaining_time": "29m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547412}
{"loss": 23.625, "grad_norm": 1.00821233, "learning_rate": 1.896e-05, "token_acc": 0.39368061, "epoch": 0.91503268, "global_step/max_steps": "1995/2181", "elapsed_time": "5h 17m 27s", "remaining_time": "29m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547378}
{"loss": 23.5, "grad_norm": 1.02041233, "learning_rate": 1.876e-05, "token_acc": 0.38279017, "epoch": 0.91549134, "global_step/max_steps": "1996/2181", "elapsed_time": "5h 17m 37s", "remaining_time": "29m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547638}
{"loss": 24.375, "grad_norm": 1.04273379, "learning_rate": 1.856e-05, "token_acc": 0.38968321, "epoch": 0.91595001, "global_step/max_steps": "1997/2181", "elapsed_time": "5h 17m 47s", "remaining_time": "29m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547732}
{"loss": 23.5, "grad_norm": 1.02539682, "learning_rate": 1.836e-05, "token_acc": 0.40080506, "epoch": 0.91640867, "global_step/max_steps": "1998/2181", "elapsed_time": "5h 17m 56s", "remaining_time": "29m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547581}
{"loss": 23.875, "grad_norm": 1.02512205, "learning_rate": 1.816e-05, "token_acc": 0.39171532, "epoch": 0.91686733, "global_step/max_steps": "1999/2181", "elapsed_time": "5h 18m 6s", "remaining_time": "28m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547596}
{"loss": 23.5, "grad_norm": 0.99337494, "learning_rate": 1.796e-05, "token_acc": 0.38635082, "epoch": 0.91732599, "global_step/max_steps": "2000/2181", "elapsed_time": "5h 18m 15s", "remaining_time": "28m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.547547}
{"loss": 24.0, "grad_norm": 1.0657171, "learning_rate": 1.777e-05, "token_acc": 0.38115981, "epoch": 0.91778466, "global_step/max_steps": "2001/2181", "elapsed_time": "5h 18m 45s", "remaining_time": "28m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.557601}
{"loss": 24.0, "grad_norm": 1.066589, "learning_rate": 1.757e-05, "token_acc": 0.39119245, "epoch": 0.91824332, "global_step/max_steps": "2002/2181", "elapsed_time": "5h 18m 55s", "remaining_time": "28m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558133}
{"loss": 23.5, "grad_norm": 1.03825343, "learning_rate": 1.738e-05, "token_acc": 0.39036408, "epoch": 0.91870198, "global_step/max_steps": "2003/2181", "elapsed_time": "5h 19m 4s", "remaining_time": "28m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.557861}
{"loss": 23.75, "grad_norm": 1.08353746, "learning_rate": 1.718e-05, "token_acc": 0.39425514, "epoch": 0.91916065, "global_step/max_steps": "2004/2181", "elapsed_time": "5h 19m 14s", "remaining_time": "28m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.557911}
{"loss": 23.625, "grad_norm": 1.0187099, "learning_rate": 1.699e-05, "token_acc": 0.39581613, "epoch": 0.91961931, "global_step/max_steps": "2005/2181", "elapsed_time": "5h 19m 24s", "remaining_time": "28m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558045}
{"loss": 23.625, "grad_norm": 1.092489, "learning_rate": 1.68e-05, "token_acc": 0.37940923, "epoch": 0.92007797, "global_step/max_steps": "2006/2181", "elapsed_time": "5h 19m 34s", "remaining_time": "27m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558453}
{"loss": 23.875, "grad_norm": 1.04190123, "learning_rate": 1.661e-05, "token_acc": 0.37228109, "epoch": 0.92053664, "global_step/max_steps": "2007/2181", "elapsed_time": "5h 19m 43s", "remaining_time": "27m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558285}
{"loss": 22.875, "grad_norm": 1.11884534, "learning_rate": 1.642e-05, "token_acc": 0.40157259, "epoch": 0.9209953, "global_step/max_steps": "2008/2181", "elapsed_time": "5h 19m 52s", "remaining_time": "27m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.557877}
{"loss": 23.75, "grad_norm": 1.08600283, "learning_rate": 1.623e-05, "token_acc": 0.39393939, "epoch": 0.92145396, "global_step/max_steps": "2009/2181", "elapsed_time": "5h 20m 2s", "remaining_time": "27m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.557849}
{"loss": 25.0, "grad_norm": 1.06761599, "learning_rate": 1.604e-05, "token_acc": 0.36783576, "epoch": 0.92191262, "global_step/max_steps": "2010/2181", "elapsed_time": "5h 20m 11s", "remaining_time": "27m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.557924}
{"loss": 23.875, "grad_norm": 0.97107774, "learning_rate": 1.586e-05, "token_acc": 0.3843009, "epoch": 0.92237129, "global_step/max_steps": "2011/2181", "elapsed_time": "5h 20m 20s", "remaining_time": "27m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.557513}
{"loss": 24.75, "grad_norm": 1.05206847, "learning_rate": 1.567e-05, "token_acc": 0.36995638, "epoch": 0.92282995, "global_step/max_steps": "2012/2181", "elapsed_time": "5h 20m 31s", "remaining_time": "26m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558132}
{"loss": 23.75, "grad_norm": 1.03702879, "learning_rate": 1.549e-05, "token_acc": 0.38446656, "epoch": 0.92328861, "global_step/max_steps": "2013/2181", "elapsed_time": "5h 20m 42s", "remaining_time": "26m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558882}
{"loss": 24.125, "grad_norm": 1.00699687, "learning_rate": 1.53e-05, "token_acc": 0.37716067, "epoch": 0.92374728, "global_step/max_steps": "2014/2181", "elapsed_time": "5h 20m 51s", "remaining_time": "26m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558641}
{"loss": 23.625, "grad_norm": 1.03625047, "learning_rate": 1.512e-05, "token_acc": 0.37890411, "epoch": 0.92420594, "global_step/max_steps": "2015/2181", "elapsed_time": "5h 21m 1s", "remaining_time": "26m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558629}
{"loss": 24.125, "grad_norm": 1.07987082, "learning_rate": 1.494e-05, "token_acc": 0.39146868, "epoch": 0.9246646, "global_step/max_steps": "2016/2181", "elapsed_time": "5h 21m 11s", "remaining_time": "26m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558815}
{"loss": 24.25, "grad_norm": 1.07989478, "learning_rate": 1.476e-05, "token_acc": 0.38428693, "epoch": 0.92512327, "global_step/max_steps": "2017/2181", "elapsed_time": "5h 21m 20s", "remaining_time": "26m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558677}
{"loss": 23.875, "grad_norm": 1.11219835, "learning_rate": 1.458e-05, "token_acc": 0.37560011, "epoch": 0.92558193, "global_step/max_steps": "2018/2181", "elapsed_time": "5h 21m 29s", "remaining_time": "25m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558721}
{"loss": 23.625, "grad_norm": 1.0351994, "learning_rate": 1.441e-05, "token_acc": 0.38083049, "epoch": 0.92604059, "global_step/max_steps": "2019/2181", "elapsed_time": "5h 21m 39s", "remaining_time": "25m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558707}
{"loss": 24.375, "grad_norm": 1.00553989, "learning_rate": 1.423e-05, "token_acc": 0.3684356, "epoch": 0.92649925, "global_step/max_steps": "2020/2181", "elapsed_time": "5h 21m 48s", "remaining_time": "25m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.55864}
{"loss": 24.25, "grad_norm": 0.99647969, "learning_rate": 1.405e-05, "token_acc": 0.35867391, "epoch": 0.92695792, "global_step/max_steps": "2021/2181", "elapsed_time": "5h 21m 58s", "remaining_time": "25m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558815}
{"loss": 23.375, "grad_norm": 1.01283002, "learning_rate": 1.388e-05, "token_acc": 0.3839779, "epoch": 0.92741658, "global_step/max_steps": "2022/2181", "elapsed_time": "5h 22m 8s", "remaining_time": "25m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558798}
{"loss": 23.75, "grad_norm": 1.05964518, "learning_rate": 1.371e-05, "token_acc": 0.38746203, "epoch": 0.92787524, "global_step/max_steps": "2023/2181", "elapsed_time": "5h 22m 17s", "remaining_time": "25m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558682}
{"loss": 23.25, "grad_norm": 1.02208269, "learning_rate": 1.353e-05, "token_acc": 0.38888889, "epoch": 0.92833391, "global_step/max_steps": "2024/2181", "elapsed_time": "5h 22m 27s", "remaining_time": "25m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558614}
{"loss": 23.125, "grad_norm": 1.0832932, "learning_rate": 1.336e-05, "token_acc": 0.4020979, "epoch": 0.92879257, "global_step/max_steps": "2025/2181", "elapsed_time": "5h 22m 37s", "remaining_time": "24m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558805}
{"loss": 23.5, "grad_norm": 1.06540489, "learning_rate": 1.319e-05, "token_acc": 0.39623686, "epoch": 0.92925123, "global_step/max_steps": "2026/2181", "elapsed_time": "5h 22m 46s", "remaining_time": "24m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558637}
{"loss": 24.125, "grad_norm": 1.16559029, "learning_rate": 1.302e-05, "token_acc": 0.36669424, "epoch": 0.9297099, "global_step/max_steps": "2027/2181", "elapsed_time": "5h 22m 55s", "remaining_time": "24m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558619}
{"loss": 23.75, "grad_norm": 1.02978301, "learning_rate": 1.286e-05, "token_acc": 0.37732657, "epoch": 0.93016856, "global_step/max_steps": "2028/2181", "elapsed_time": "5h 23m 5s", "remaining_time": "24m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.55862}
{"loss": 23.5, "grad_norm": 1.06401658, "learning_rate": 1.269e-05, "token_acc": 0.40449119, "epoch": 0.93062722, "global_step/max_steps": "2029/2181", "elapsed_time": "5h 23m 15s", "remaining_time": "24m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558711}
{"loss": 24.25, "grad_norm": 0.96454793, "learning_rate": 1.252e-05, "token_acc": 0.37863043, "epoch": 0.93108588, "global_step/max_steps": "2030/2181", "elapsed_time": "5h 23m 24s", "remaining_time": "24m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558523}
{"loss": 23.25, "grad_norm": 0.99166375, "learning_rate": 1.236e-05, "token_acc": 0.40278157, "epoch": 0.93154455, "global_step/max_steps": "2031/2181", "elapsed_time": "5h 23m 33s", "remaining_time": "23m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558544}
{"loss": 24.625, "grad_norm": 1.1019994, "learning_rate": 1.22e-05, "token_acc": 0.38134888, "epoch": 0.93200321, "global_step/max_steps": "2032/2181", "elapsed_time": "5h 23m 44s", "remaining_time": "23m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559092}
{"loss": 24.25, "grad_norm": 1.01192963, "learning_rate": 1.203e-05, "token_acc": 0.38463695, "epoch": 0.93246187, "global_step/max_steps": "2033/2181", "elapsed_time": "5h 23m 53s", "remaining_time": "23m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.55901}
{"loss": 24.25, "grad_norm": 0.99247497, "learning_rate": 1.187e-05, "token_acc": 0.37680339, "epoch": 0.93292054, "global_step/max_steps": "2034/2181", "elapsed_time": "5h 24m 3s", "remaining_time": "23m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558828}
{"loss": 22.875, "grad_norm": 1.0222373, "learning_rate": 1.171e-05, "token_acc": 0.40489914, "epoch": 0.9333792, "global_step/max_steps": "2035/2181", "elapsed_time": "5h 24m 12s", "remaining_time": "23m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558799}
{"loss": 23.75, "grad_norm": 1.02635682, "learning_rate": 1.155e-05, "token_acc": 0.38778409, "epoch": 0.93383786, "global_step/max_steps": "2036/2181", "elapsed_time": "5h 24m 22s", "remaining_time": "23m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558803}
{"loss": 23.75, "grad_norm": 1.01281786, "learning_rate": 1.139e-05, "token_acc": 0.39575874, "epoch": 0.93429653, "global_step/max_steps": "2037/2181", "elapsed_time": "5h 24m 31s", "remaining_time": "22m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.55882}
{"loss": 23.625, "grad_norm": 1.07867992, "learning_rate": 1.124e-05, "token_acc": 0.38294956, "epoch": 0.93475519, "global_step/max_steps": "2038/2181", "elapsed_time": "5h 24m 41s", "remaining_time": "22m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559037}
{"loss": 23.375, "grad_norm": 0.97606528, "learning_rate": 1.108e-05, "token_acc": 0.39433609, "epoch": 0.93521385, "global_step/max_steps": "2039/2181", "elapsed_time": "5h 24m 51s", "remaining_time": "22m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559295}
{"loss": 24.125, "grad_norm": 1.08493698, "learning_rate": 1.093e-05, "token_acc": 0.37934969, "epoch": 0.93567251, "global_step/max_steps": "2040/2181", "elapsed_time": "5h 25m 1s", "remaining_time": "22m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559302}
{"loss": 23.625, "grad_norm": 1.04585695, "learning_rate": 1.077e-05, "token_acc": 0.38791209, "epoch": 0.93613118, "global_step/max_steps": "2041/2181", "elapsed_time": "5h 25m 10s", "remaining_time": "22m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559036}
{"loss": 24.0, "grad_norm": 1.09981596, "learning_rate": 1.062e-05, "token_acc": 0.37961926, "epoch": 0.93658984, "global_step/max_steps": "2042/2181", "elapsed_time": "5h 25m 19s", "remaining_time": "22m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558953}
{"loss": 24.25, "grad_norm": 1.10201192, "learning_rate": 1.047e-05, "token_acc": 0.37729455, "epoch": 0.9370485, "global_step/max_steps": "2043/2181", "elapsed_time": "5h 25m 29s", "remaining_time": "21m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559052}
{"loss": 23.625, "grad_norm": 1.03404498, "learning_rate": 1.032e-05, "token_acc": 0.3785073, "epoch": 0.93750717, "global_step/max_steps": "2044/2181", "elapsed_time": "5h 25m 39s", "remaining_time": "21m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558975}
{"loss": 23.875, "grad_norm": 1.03443813, "learning_rate": 1.017e-05, "token_acc": 0.38193476, "epoch": 0.93796583, "global_step/max_steps": "2045/2181", "elapsed_time": "5h 25m 49s", "remaining_time": "21m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559211}
{"loss": 24.375, "grad_norm": 1.04344809, "learning_rate": 1.002e-05, "token_acc": 0.38808812, "epoch": 0.93842449, "global_step/max_steps": "2046/2181", "elapsed_time": "5h 25m 58s", "remaining_time": "21m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559273}
{"loss": 23.75, "grad_norm": 1.13599038, "learning_rate": 9.87e-06, "token_acc": 0.38525469, "epoch": 0.93888316, "global_step/max_steps": "2047/2181", "elapsed_time": "5h 26m 8s", "remaining_time": "21m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559175}
{"loss": 24.0, "grad_norm": 1.0727706, "learning_rate": 9.73e-06, "token_acc": 0.38232888, "epoch": 0.93934182, "global_step/max_steps": "2048/2181", "elapsed_time": "5h 26m 17s", "remaining_time": "21m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.55929}
{"loss": 23.75, "grad_norm": 1.21448421, "learning_rate": 9.58e-06, "token_acc": 0.39384784, "epoch": 0.93980048, "global_step/max_steps": "2049/2181", "elapsed_time": "5h 26m 27s", "remaining_time": "21m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559245}
{"loss": 23.75, "grad_norm": 1.03577995, "learning_rate": 9.44e-06, "token_acc": 0.38150127, "epoch": 0.94025914, "global_step/max_steps": "2050/2181", "elapsed_time": "5h 26m 36s", "remaining_time": "20m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559171}
{"loss": 23.375, "grad_norm": 1.09392035, "learning_rate": 9.29e-06, "token_acc": 0.3917768, "epoch": 0.94071781, "global_step/max_steps": "2051/2181", "elapsed_time": "5h 26m 46s", "remaining_time": "20m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559299}
{"loss": 24.0, "grad_norm": 0.95522308, "learning_rate": 9.15e-06, "token_acc": 0.39983306, "epoch": 0.94117647, "global_step/max_steps": "2052/2181", "elapsed_time": "5h 26m 56s", "remaining_time": "20m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559388}
{"loss": 24.125, "grad_norm": 1.06728911, "learning_rate": 9.01e-06, "token_acc": 0.36972304, "epoch": 0.94163513, "global_step/max_steps": "2053/2181", "elapsed_time": "5h 27m 5s", "remaining_time": "20m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559328}
{"loss": 24.0, "grad_norm": 1.04584754, "learning_rate": 8.87e-06, "token_acc": 0.37834852, "epoch": 0.9420938, "global_step/max_steps": "2054/2181", "elapsed_time": "5h 27m 15s", "remaining_time": "20m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559334}
{"loss": 23.875, "grad_norm": 1.03351617, "learning_rate": 8.73e-06, "token_acc": 0.38282548, "epoch": 0.94255246, "global_step/max_steps": "2055/2181", "elapsed_time": "5h 27m 25s", "remaining_time": "20m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559722}
{"loss": 23.625, "grad_norm": 0.99713379, "learning_rate": 8.59e-06, "token_acc": 0.38405594, "epoch": 0.94301112, "global_step/max_steps": "2056/2181", "elapsed_time": "5h 27m 34s", "remaining_time": "19m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559541}
{"loss": 24.125, "grad_norm": 1.00834787, "learning_rate": 8.46e-06, "token_acc": 0.38525499, "epoch": 0.94346979, "global_step/max_steps": "2057/2181", "elapsed_time": "5h 27m 45s", "remaining_time": "19m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559842}
{"loss": 23.75, "grad_norm": 1.10754061, "learning_rate": 8.32e-06, "token_acc": 0.40056417, "epoch": 0.94392845, "global_step/max_steps": "2058/2181", "elapsed_time": "5h 27m 54s", "remaining_time": "19m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559769}
{"loss": 23.25, "grad_norm": 1.06269979, "learning_rate": 8.19e-06, "token_acc": 0.38746439, "epoch": 0.94438711, "global_step/max_steps": "2059/2181", "elapsed_time": "5h 28m 3s", "remaining_time": "19m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559403}
{"loss": 23.5, "grad_norm": 1.03007722, "learning_rate": 8.05e-06, "token_acc": 0.40087977, "epoch": 0.94484577, "global_step/max_steps": "2060/2181", "elapsed_time": "5h 28m 12s", "remaining_time": "19m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.55943}
{"loss": 23.75, "grad_norm": 1.07347846, "learning_rate": 7.92e-06, "token_acc": 0.38239497, "epoch": 0.94530444, "global_step/max_steps": "2061/2181", "elapsed_time": "5h 28m 22s", "remaining_time": "19m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559466}
{"loss": 24.0, "grad_norm": 1.06495512, "learning_rate": 7.79e-06, "token_acc": 0.37600443, "epoch": 0.9457631, "global_step/max_steps": "2062/2181", "elapsed_time": "5h 28m 31s", "remaining_time": "18m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559155}
{"loss": 24.125, "grad_norm": 1.00590289, "learning_rate": 7.66e-06, "token_acc": 0.38510876, "epoch": 0.94622176, "global_step/max_steps": "2063/2181", "elapsed_time": "5h 28m 41s", "remaining_time": "18m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559486}
{"loss": 24.25, "grad_norm": 1.16199303, "learning_rate": 7.53e-06, "token_acc": 0.38890415, "epoch": 0.94668043, "global_step/max_steps": "2064/2181", "elapsed_time": "5h 28m 50s", "remaining_time": "18m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559276}
{"loss": 23.375, "grad_norm": 1.18406844, "learning_rate": 7.4e-06, "token_acc": 0.3916501, "epoch": 0.94713909, "global_step/max_steps": "2065/2181", "elapsed_time": "5h 29m 0s", "remaining_time": "18m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559376}
{"loss": 23.5, "grad_norm": 1.05847669, "learning_rate": 7.28e-06, "token_acc": 0.39168552, "epoch": 0.94759775, "global_step/max_steps": "2066/2181", "elapsed_time": "5h 29m 9s", "remaining_time": "18m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559234}
{"loss": 24.0, "grad_norm": 1.16978395, "learning_rate": 7.15e-06, "token_acc": 0.3893534, "epoch": 0.94805642, "global_step/max_steps": "2067/2181", "elapsed_time": "5h 29m 19s", "remaining_time": "18m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559393}
{"loss": 24.375, "grad_norm": 1.06031239, "learning_rate": 7.03e-06, "token_acc": 0.37678768, "epoch": 0.94851508, "global_step/max_steps": "2068/2181", "elapsed_time": "5h 29m 28s", "remaining_time": "18m 0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559221}
{"loss": 24.125, "grad_norm": 1.17737818, "learning_rate": 6.9e-06, "token_acc": 0.38440189, "epoch": 0.94897374, "global_step/max_steps": "2069/2181", "elapsed_time": "5h 29m 37s", "remaining_time": "17m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558952}
{"loss": 24.25, "grad_norm": 1.07078063, "learning_rate": 6.78e-06, "token_acc": 0.39903299, "epoch": 0.9494324, "global_step/max_steps": "2070/2181", "elapsed_time": "5h 29m 47s", "remaining_time": "17m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558936}
{"loss": 24.125, "grad_norm": 1.03681004, "learning_rate": 6.66e-06, "token_acc": 0.38140162, "epoch": 0.94989107, "global_step/max_steps": "2071/2181", "elapsed_time": "5h 29m 57s", "remaining_time": "17m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558977}
{"loss": 24.25, "grad_norm": 0.94813651, "learning_rate": 6.54e-06, "token_acc": 0.37601797, "epoch": 0.95034973, "global_step/max_steps": "2072/2181", "elapsed_time": "5h 30m 6s", "remaining_time": "17m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558968}
{"loss": 23.875, "grad_norm": 1.01365876, "learning_rate": 6.42e-06, "token_acc": 0.37369942, "epoch": 0.95080839, "global_step/max_steps": "2073/2181", "elapsed_time": "5h 30m 16s", "remaining_time": "17m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558906}
{"loss": 23.75, "grad_norm": 0.9785037, "learning_rate": 6.3e-06, "token_acc": 0.37851519, "epoch": 0.95126706, "global_step/max_steps": "2074/2181", "elapsed_time": "5h 30m 25s", "remaining_time": "17m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558981}
{"loss": 24.375, "grad_norm": 1.01687729, "learning_rate": 6.18e-06, "token_acc": 0.38042553, "epoch": 0.95172572, "global_step/max_steps": "2075/2181", "elapsed_time": "5h 30m 35s", "remaining_time": "16m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559024}
{"loss": 23.875, "grad_norm": 1.02541685, "learning_rate": 6.07e-06, "token_acc": 0.38799571, "epoch": 0.95218438, "global_step/max_steps": "2076/2181", "elapsed_time": "5h 30m 44s", "remaining_time": "16m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558804}
{"loss": 23.625, "grad_norm": 1.01818371, "learning_rate": 5.95e-06, "token_acc": 0.38641553, "epoch": 0.95264305, "global_step/max_steps": "2077/2181", "elapsed_time": "5h 30m 54s", "remaining_time": "16m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559124}
{"loss": 23.625, "grad_norm": 0.98890013, "learning_rate": 5.84e-06, "token_acc": 0.38823874, "epoch": 0.95310171, "global_step/max_steps": "2078/2181", "elapsed_time": "5h 31m 4s", "remaining_time": "16m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558971}
{"loss": 23.625, "grad_norm": 1.07132876, "learning_rate": 5.73e-06, "token_acc": 0.3888587, "epoch": 0.95356037, "global_step/max_steps": "2079/2181", "elapsed_time": "5h 31m 13s", "remaining_time": "16m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558952}
{"loss": 23.875, "grad_norm": 0.96186936, "learning_rate": 5.62e-06, "token_acc": 0.37694524, "epoch": 0.95401903, "global_step/max_steps": "2080/2181", "elapsed_time": "5h 31m 23s", "remaining_time": "16m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559358}
{"loss": 24.5, "grad_norm": 1.09835744, "learning_rate": 5.51e-06, "token_acc": 0.36901798, "epoch": 0.9544777, "global_step/max_steps": "2081/2181", "elapsed_time": "5h 31m 33s", "remaining_time": "15m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559266}
{"loss": 24.125, "grad_norm": 1.10834968, "learning_rate": 5.4e-06, "token_acc": 0.39136691, "epoch": 0.95493636, "global_step/max_steps": "2082/2181", "elapsed_time": "5h 31m 42s", "remaining_time": "15m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559247}
{"loss": 24.0, "grad_norm": 1.06905866, "learning_rate": 5.29e-06, "token_acc": 0.37973967, "epoch": 0.95539502, "global_step/max_steps": "2083/2181", "elapsed_time": "5h 31m 52s", "remaining_time": "15m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559383}
{"loss": 24.25, "grad_norm": 1.06451595, "learning_rate": 5.18e-06, "token_acc": 0.38407821, "epoch": 0.95585369, "global_step/max_steps": "2084/2181", "elapsed_time": "5h 32m 2s", "remaining_time": "15m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559272}
{"loss": 24.0, "grad_norm": 0.96859515, "learning_rate": 5.07e-06, "token_acc": 0.37706667, "epoch": 0.95631235, "global_step/max_steps": "2085/2181", "elapsed_time": "5h 32m 11s", "remaining_time": "15m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559182}
{"loss": 23.5, "grad_norm": 1.06436467, "learning_rate": 4.97e-06, "token_acc": 0.38169209, "epoch": 0.95677101, "global_step/max_steps": "2086/2181", "elapsed_time": "5h 32m 21s", "remaining_time": "15m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559441}
{"loss": 23.625, "grad_norm": 0.97714514, "learning_rate": 4.87e-06, "token_acc": 0.40411144, "epoch": 0.95722968, "global_step/max_steps": "2087/2181", "elapsed_time": "5h 32m 30s", "remaining_time": "14m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559248}
{"loss": 23.0, "grad_norm": 1.04697776, "learning_rate": 4.76e-06, "token_acc": 0.3927467, "epoch": 0.95768834, "global_step/max_steps": "2088/2181", "elapsed_time": "5h 32m 41s", "remaining_time": "14m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559653}
{"loss": 23.75, "grad_norm": 1.11495006, "learning_rate": 4.66e-06, "token_acc": 0.400271, "epoch": 0.958147, "global_step/max_steps": "2089/2181", "elapsed_time": "5h 32m 50s", "remaining_time": "14m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.55961}
{"loss": 23.75, "grad_norm": 0.98074955, "learning_rate": 4.56e-06, "token_acc": 0.3922916, "epoch": 0.95860566, "global_step/max_steps": "2090/2181", "elapsed_time": "5h 32m 59s", "remaining_time": "14m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559411}
{"loss": 24.375, "grad_norm": 0.99976653, "learning_rate": 4.46e-06, "token_acc": 0.36480807, "epoch": 0.95906433, "global_step/max_steps": "2091/2181", "elapsed_time": "5h 33m 9s", "remaining_time": "14m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559474}
{"loss": 24.5, "grad_norm": 1.04608059, "learning_rate": 4.36e-06, "token_acc": 0.38517915, "epoch": 0.95952299, "global_step/max_steps": "2092/2181", "elapsed_time": "5h 33m 19s", "remaining_time": "14m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559603}
{"loss": 24.625, "grad_norm": 1.02726912, "learning_rate": 4.27e-06, "token_acc": 0.37453679, "epoch": 0.95998165, "global_step/max_steps": "2093/2181", "elapsed_time": "5h 33m 28s", "remaining_time": "14m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559381}
{"loss": 25.0, "grad_norm": 0.99497962, "learning_rate": 4.17e-06, "token_acc": 0.37342956, "epoch": 0.96044032, "global_step/max_steps": "2094/2181", "elapsed_time": "5h 33m 37s", "remaining_time": "13m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559235}
{"loss": 23.875, "grad_norm": 1.19207609, "learning_rate": 4.07e-06, "token_acc": 0.37944891, "epoch": 0.96089898, "global_step/max_steps": "2095/2181", "elapsed_time": "5h 33m 47s", "remaining_time": "13m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559219}
{"loss": 23.375, "grad_norm": 1.07075191, "learning_rate": 3.98e-06, "token_acc": 0.39949679, "epoch": 0.96135764, "global_step/max_steps": "2096/2181", "elapsed_time": "5h 33m 56s", "remaining_time": "13m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559075}
{"loss": 24.0, "grad_norm": 1.06343174, "learning_rate": 3.89e-06, "token_acc": 0.37503546, "epoch": 0.96181631, "global_step/max_steps": "2097/2181", "elapsed_time": "5h 34m 6s", "remaining_time": "13m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559277}
{"loss": 24.0, "grad_norm": 1.08467948, "learning_rate": 3.8e-06, "token_acc": 0.37702587, "epoch": 0.96227497, "global_step/max_steps": "2098/2181", "elapsed_time": "5h 34m 16s", "remaining_time": "13m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559448}
{"loss": 23.625, "grad_norm": 1.07239354, "learning_rate": 3.7e-06, "token_acc": 0.38814857, "epoch": 0.96273363, "global_step/max_steps": "2099/2181", "elapsed_time": "5h 34m 25s", "remaining_time": "13m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559492}
{"loss": 23.75, "grad_norm": 1.10127556, "learning_rate": 3.61e-06, "token_acc": 0.39044863, "epoch": 0.96319229, "global_step/max_steps": "2100/2181", "elapsed_time": "5h 34m 35s", "remaining_time": "12m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559456}
{"loss": 23.5, "grad_norm": 0.95304483, "learning_rate": 3.53e-06, "token_acc": 0.37797702, "epoch": 0.96365096, "global_step/max_steps": "2101/2181", "elapsed_time": "5h 34m 45s", "remaining_time": "12m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559672}
{"loss": 24.0, "grad_norm": 0.98445553, "learning_rate": 3.44e-06, "token_acc": 0.38528711, "epoch": 0.96410962, "global_step/max_steps": "2102/2181", "elapsed_time": "5h 34m 54s", "remaining_time": "12m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559449}
{"loss": 23.25, "grad_norm": 1.05818379, "learning_rate": 3.35e-06, "token_acc": 0.39691445, "epoch": 0.96456828, "global_step/max_steps": "2103/2181", "elapsed_time": "5h 35m 4s", "remaining_time": "12m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559456}
{"loss": 23.875, "grad_norm": 1.06866086, "learning_rate": 3.27e-06, "token_acc": 0.38136078, "epoch": 0.96502695, "global_step/max_steps": "2104/2181", "elapsed_time": "5h 35m 14s", "remaining_time": "12m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559711}
{"loss": 24.25, "grad_norm": 1.04229939, "learning_rate": 3.18e-06, "token_acc": 0.37759454, "epoch": 0.96548561, "global_step/max_steps": "2105/2181", "elapsed_time": "5h 35m 23s", "remaining_time": "12m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.55985}
{"loss": 24.25, "grad_norm": 1.0418781, "learning_rate": 3.1e-06, "token_acc": 0.38836773, "epoch": 0.96594427, "global_step/max_steps": "2106/2181", "elapsed_time": "5h 35m 33s", "remaining_time": "11m 57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559967}
{"loss": 23.875, "grad_norm": 1.01088703, "learning_rate": 3.02e-06, "token_acc": 0.3884829, "epoch": 0.96640294, "global_step/max_steps": "2107/2181", "elapsed_time": "5h 35m 43s", "remaining_time": "11m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560012}
{"loss": 24.5, "grad_norm": 1.01565766, "learning_rate": 2.94e-06, "token_acc": 0.38100654, "epoch": 0.9668616, "global_step/max_steps": "2108/2181", "elapsed_time": "5h 35m 53s", "remaining_time": "11m 38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560037}
{"loss": 23.625, "grad_norm": 1.02427697, "learning_rate": 2.86e-06, "token_acc": 0.38924234, "epoch": 0.96732026, "global_step/max_steps": "2109/2181", "elapsed_time": "5h 36m 2s", "remaining_time": "11m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560015}
{"loss": 23.875, "grad_norm": 1.01096261, "learning_rate": 2.78e-06, "token_acc": 0.39192264, "epoch": 0.96777892, "global_step/max_steps": "2110/2181", "elapsed_time": "5h 36m 12s", "remaining_time": "11m 19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560217}
{"loss": 23.625, "grad_norm": 0.99413741, "learning_rate": 2.7e-06, "token_acc": 0.3920501, "epoch": 0.96823759, "global_step/max_steps": "2111/2181", "elapsed_time": "5h 36m 22s", "remaining_time": "11m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560172}
{"loss": 24.625, "grad_norm": 0.97438633, "learning_rate": 2.62e-06, "token_acc": 0.38003787, "epoch": 0.96869625, "global_step/max_steps": "2112/2181", "elapsed_time": "5h 36m 32s", "remaining_time": "10m 60s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560392}
{"loss": 23.125, "grad_norm": 1.09159708, "learning_rate": 2.55e-06, "token_acc": 0.39282664, "epoch": 0.96915491, "global_step/max_steps": "2113/2181", "elapsed_time": "5h 36m 41s", "remaining_time": "10m 50s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560358}
{"loss": 23.875, "grad_norm": 1.09769177, "learning_rate": 2.47e-06, "token_acc": 0.39143563, "epoch": 0.96961358, "global_step/max_steps": "2114/2181", "elapsed_time": "5h 36m 50s", "remaining_time": "10m 41s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560087}
{"loss": 23.75, "grad_norm": 0.98852122, "learning_rate": 2.4e-06, "token_acc": 0.39345172, "epoch": 0.97007224, "global_step/max_steps": "2115/2181", "elapsed_time": "5h 36m 59s", "remaining_time": "10m 31s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559871}
{"loss": 23.5, "grad_norm": 1.03924429, "learning_rate": 2.33e-06, "token_acc": 0.39518994, "epoch": 0.9705309, "global_step/max_steps": "2116/2181", "elapsed_time": "5h 37m 10s", "remaining_time": "10m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560283}
{"loss": 24.25, "grad_norm": 1.037117, "learning_rate": 2.26e-06, "token_acc": 0.38802779, "epoch": 0.97098957, "global_step/max_steps": "2117/2181", "elapsed_time": "5h 37m 19s", "remaining_time": "10m 12s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560346}
{"loss": 23.875, "grad_norm": 1.02565503, "learning_rate": 2.19e-06, "token_acc": 0.38233681, "epoch": 0.97144823, "global_step/max_steps": "2118/2181", "elapsed_time": "5h 37m 29s", "remaining_time": "10m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560232}
{"loss": 24.0, "grad_norm": 0.94971842, "learning_rate": 2.12e-06, "token_acc": 0.38791733, "epoch": 0.97190689, "global_step/max_steps": "2119/2181", "elapsed_time": "5h 37m 39s", "remaining_time": "9m 53s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560532}
{"loss": 23.625, "grad_norm": 1.05986285, "learning_rate": 2.05e-06, "token_acc": 0.3933518, "epoch": 0.97236555, "global_step/max_steps": "2120/2181", "elapsed_time": "5h 37m 48s", "remaining_time": "9m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.56036}
{"loss": 23.75, "grad_norm": 1.05978954, "learning_rate": 1.98e-06, "token_acc": 0.39572486, "epoch": 0.97282422, "global_step/max_steps": "2121/2181", "elapsed_time": "5h 37m 58s", "remaining_time": "9m 34s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560471}
{"loss": 24.375, "grad_norm": 1.0225879, "learning_rate": 1.92e-06, "token_acc": 0.37360179, "epoch": 0.97328288, "global_step/max_steps": "2122/2181", "elapsed_time": "5h 38m 8s", "remaining_time": "9m 24s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560592}
{"loss": 23.75, "grad_norm": 1.13698983, "learning_rate": 1.85e-06, "token_acc": 0.38904035, "epoch": 0.97374154, "global_step/max_steps": "2123/2181", "elapsed_time": "5h 38m 17s", "remaining_time": "9m 15s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560407}
{"loss": 24.125, "grad_norm": 1.14418757, "learning_rate": 1.79e-06, "token_acc": 0.38753316, "epoch": 0.97420021, "global_step/max_steps": "2124/2181", "elapsed_time": "5h 38m 27s", "remaining_time": "9m 5s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560544}
{"loss": 24.0, "grad_norm": 0.96010989, "learning_rate": 1.73e-06, "token_acc": 0.39517937, "epoch": 0.97465887, "global_step/max_steps": "2125/2181", "elapsed_time": "5h 38m 35s", "remaining_time": "8m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560219}
{"loss": 23.5, "grad_norm": 0.98698038, "learning_rate": 1.67e-06, "token_acc": 0.39132867, "epoch": 0.97511753, "global_step/max_steps": "2126/2181", "elapsed_time": "5h 38m 45s", "remaining_time": "8m 46s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560091}
{"loss": 25.0, "grad_norm": 1.02491307, "learning_rate": 1.61e-06, "token_acc": 0.362961, "epoch": 0.9755762, "global_step/max_steps": "2127/2181", "elapsed_time": "5h 38m 54s", "remaining_time": "8m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560061}
{"loss": 24.0, "grad_norm": 1.13284326, "learning_rate": 1.55e-06, "token_acc": 0.36811024, "epoch": 0.97603486, "global_step/max_steps": "2128/2181", "elapsed_time": "5h 39m 4s", "remaining_time": "8m 27s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560062}
{"loss": 24.0, "grad_norm": 1.12943232, "learning_rate": 1.49e-06, "token_acc": 0.38817112, "epoch": 0.97649352, "global_step/max_steps": "2129/2181", "elapsed_time": "5h 39m 13s", "remaining_time": "8m 17s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559978}
{"loss": 23.75, "grad_norm": 1.10630739, "learning_rate": 1.43e-06, "token_acc": 0.38360381, "epoch": 0.97695218, "global_step/max_steps": "2130/2181", "elapsed_time": "5h 39m 23s", "remaining_time": "8m 8s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559965}
{"loss": 23.125, "grad_norm": 1.0338906, "learning_rate": 1.38e-06, "token_acc": 0.4, "epoch": 0.97741085, "global_step/max_steps": "2131/2181", "elapsed_time": "5h 39m 32s", "remaining_time": "7m 58s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.559976}
{"loss": 24.5, "grad_norm": 1.12701285, "learning_rate": 1.32e-06, "token_acc": 0.37907762, "epoch": 0.97786951, "global_step/max_steps": "2132/2181", "elapsed_time": "5h 39m 43s", "remaining_time": "7m 48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560443}
{"loss": 24.375, "grad_norm": 1.07662714, "learning_rate": 1.27e-06, "token_acc": 0.3773224, "epoch": 0.97832817, "global_step/max_steps": "2133/2181", "elapsed_time": "5h 39m 52s", "remaining_time": "7m 39s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560178}
{"loss": 23.625, "grad_norm": 1.05820644, "learning_rate": 1.22e-06, "token_acc": 0.39221725, "epoch": 0.97878684, "global_step/max_steps": "2134/2181", "elapsed_time": "5h 40m 2s", "remaining_time": "7m 29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560677}
{"loss": 24.0, "grad_norm": 1.0499301, "learning_rate": 1.17e-06, "token_acc": 0.37535572, "epoch": 0.9792455, "global_step/max_steps": "2135/2181", "elapsed_time": "5h 40m 12s", "remaining_time": "7m 20s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560724}
{"loss": 23.25, "grad_norm": 1.05797064, "learning_rate": 1.12e-06, "token_acc": 0.39462249, "epoch": 0.97970416, "global_step/max_steps": "2136/2181", "elapsed_time": "5h 40m 22s", "remaining_time": "7m 10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.561082}
{"loss": 24.125, "grad_norm": 1.04763591, "learning_rate": 1.07e-06, "token_acc": 0.38459474, "epoch": 0.98016283, "global_step/max_steps": "2137/2181", "elapsed_time": "5h 40m 32s", "remaining_time": "7m 1s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.561105}
{"loss": 24.0, "grad_norm": 1.13200808, "learning_rate": 1.02e-06, "token_acc": 0.37736885, "epoch": 0.98062149, "global_step/max_steps": "2138/2181", "elapsed_time": "5h 40m 41s", "remaining_time": "6m 51s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560885}
{"loss": 23.125, "grad_norm": 1.02898514, "learning_rate": 9.7e-07, "token_acc": 0.39890411, "epoch": 0.98108015, "global_step/max_steps": "2139/2181", "elapsed_time": "5h 40m 51s", "remaining_time": "6m 42s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.56087}
{"loss": 23.25, "grad_norm": 1.14173472, "learning_rate": 9.3e-07, "token_acc": 0.39634495, "epoch": 0.98153881, "global_step/max_steps": "2140/2181", "elapsed_time": "5h 40m 60s", "remaining_time": "6m 32s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560745}
{"loss": 24.75, "grad_norm": 0.98348534, "learning_rate": 8.8e-07, "token_acc": 0.37569061, "epoch": 0.98199748, "global_step/max_steps": "2141/2181", "elapsed_time": "5h 41m 9s", "remaining_time": "6m 22s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560406}
{"loss": 24.25, "grad_norm": 1.08150053, "learning_rate": 8.4e-07, "token_acc": 0.38155668, "epoch": 0.98245614, "global_step/max_steps": "2142/2181", "elapsed_time": "5h 41m 19s", "remaining_time": "6m 13s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560513}
{"loss": 24.0, "grad_norm": 1.02817619, "learning_rate": 8e-07, "token_acc": 0.3836132, "epoch": 0.9829148, "global_step/max_steps": "2143/2181", "elapsed_time": "5h 41m 28s", "remaining_time": "6m 3s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560406}
{"loss": 24.375, "grad_norm": 0.97348201, "learning_rate": 7.5e-07, "token_acc": 0.37537212, "epoch": 0.98337347, "global_step/max_steps": "2144/2181", "elapsed_time": "5h 41m 37s", "remaining_time": "5m 54s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.56036}
{"loss": 24.0, "grad_norm": 1.12296712, "learning_rate": 7.1e-07, "token_acc": 0.38962443, "epoch": 0.98383213, "global_step/max_steps": "2145/2181", "elapsed_time": "5h 41m 47s", "remaining_time": "5m 44s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560449}
{"loss": 22.75, "grad_norm": 1.00252306, "learning_rate": 6.8e-07, "token_acc": 0.40476843, "epoch": 0.98429079, "global_step/max_steps": "2146/2181", "elapsed_time": "5h 41m 57s", "remaining_time": "5m 35s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560728}
{"loss": 23.5, "grad_norm": 1.06691539, "learning_rate": 6.4e-07, "token_acc": 0.38881234, "epoch": 0.98474946, "global_step/max_steps": "2147/2181", "elapsed_time": "5h 42m 7s", "remaining_time": "5m 25s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560975}
{"loss": 24.0, "grad_norm": 1.07549524, "learning_rate": 6e-07, "token_acc": 0.38245807, "epoch": 0.98520812, "global_step/max_steps": "2148/2181", "elapsed_time": "5h 42m 17s", "remaining_time": "5m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560841}
{"loss": 23.5, "grad_norm": 1.41929007, "learning_rate": 5.6e-07, "token_acc": 0.38735403, "epoch": 0.98566678, "global_step/max_steps": "2149/2181", "elapsed_time": "5h 42m 27s", "remaining_time": "5m 6s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.561007}
{"loss": 23.875, "grad_norm": 1.07385325, "learning_rate": 5.3e-07, "token_acc": 0.39129221, "epoch": 0.98612544, "global_step/max_steps": "2150/2181", "elapsed_time": "5h 42m 36s", "remaining_time": "4m 56s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560825}
{"loss": 24.25, "grad_norm": 0.939381, "learning_rate": 5e-07, "token_acc": 0.38390612, "epoch": 0.98658411, "global_step/max_steps": "2151/2181", "elapsed_time": "5h 42m 45s", "remaining_time": "4m 47s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560477}
{"loss": 23.5, "grad_norm": 1.1496774, "learning_rate": 4.6e-07, "token_acc": 0.38920135, "epoch": 0.98704277, "global_step/max_steps": "2152/2181", "elapsed_time": "5h 42m 54s", "remaining_time": "4m 37s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560563}
{"loss": 23.75, "grad_norm": 1.06352353, "learning_rate": 4.3e-07, "token_acc": 0.38568644, "epoch": 0.98750143, "global_step/max_steps": "2153/2181", "elapsed_time": "5h 43m 4s", "remaining_time": "4m 28s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560702}
{"loss": 24.5, "grad_norm": 1.04254329, "learning_rate": 4e-07, "token_acc": 0.38137283, "epoch": 0.9879601, "global_step/max_steps": "2154/2181", "elapsed_time": "5h 43m 14s", "remaining_time": "4m 18s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560663}
{"loss": 23.375, "grad_norm": 0.99847406, "learning_rate": 3.7e-07, "token_acc": 0.40142779, "epoch": 0.98841876, "global_step/max_steps": "2155/2181", "elapsed_time": "5h 43m 23s", "remaining_time": "4m 9s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560592}
{"loss": 24.125, "grad_norm": 1.02071667, "learning_rate": 3.4e-07, "token_acc": 0.38344788, "epoch": 0.98887742, "global_step/max_steps": "2156/2181", "elapsed_time": "5h 43m 32s", "remaining_time": "3m 59s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560314}
{"loss": 23.625, "grad_norm": 0.92721426, "learning_rate": 3.2e-07, "token_acc": 0.38452977, "epoch": 0.98933609, "global_step/max_steps": "2157/2181", "elapsed_time": "5h 43m 42s", "remaining_time": "3m 49s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560376}
{"loss": 23.75, "grad_norm": 1.02330065, "learning_rate": 2.9e-07, "token_acc": 0.38727524, "epoch": 0.98979475, "global_step/max_steps": "2158/2181", "elapsed_time": "5h 43m 51s", "remaining_time": "3m 40s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560463}
{"loss": 24.0, "grad_norm": 1.11357248, "learning_rate": 2.7e-07, "token_acc": 0.37749494, "epoch": 0.99025341, "global_step/max_steps": "2159/2181", "elapsed_time": "5h 44m 1s", "remaining_time": "3m 30s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560482}
{"loss": 23.25, "grad_norm": 1.09092748, "learning_rate": 2.4e-07, "token_acc": 0.39712644, "epoch": 0.99071207, "global_step/max_steps": "2160/2181", "elapsed_time": "5h 44m 11s", "remaining_time": "3m 21s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560718}
{"loss": 23.75, "grad_norm": 1.11071849, "learning_rate": 2.2e-07, "token_acc": 0.38194055, "epoch": 0.99117074, "global_step/max_steps": "2161/2181", "elapsed_time": "5h 44m 21s", "remaining_time": "3m 11s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560621}
{"loss": 24.375, "grad_norm": 1.00707877, "learning_rate": 2e-07, "token_acc": 0.3805646, "epoch": 0.9916294, "global_step/max_steps": "2162/2181", "elapsed_time": "5h 44m 30s", "remaining_time": "3m 2s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560406}
{"loss": 23.875, "grad_norm": 1.02938199, "learning_rate": 1.8e-07, "token_acc": 0.37386105, "epoch": 0.99208806, "global_step/max_steps": "2163/2181", "elapsed_time": "5h 44m 40s", "remaining_time": "2m 52s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560588}
{"loss": 24.25, "grad_norm": 1.8068378, "learning_rate": 1.6e-07, "token_acc": 0.37609991, "epoch": 0.99254673, "global_step/max_steps": "2164/2181", "elapsed_time": "5h 44m 49s", "remaining_time": "2m 43s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.56052}
{"loss": 23.875, "grad_norm": 1.045959, "learning_rate": 1.4e-07, "token_acc": 0.39696724, "epoch": 0.99300539, "global_step/max_steps": "2165/2181", "elapsed_time": "5h 44m 59s", "remaining_time": "2m 33s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560615}
{"loss": 24.25, "grad_norm": 1.02469611, "learning_rate": 1.2e-07, "token_acc": 0.38757729, "epoch": 0.99346405, "global_step/max_steps": "2166/2181", "elapsed_time": "5h 45m 9s", "remaining_time": "2m 23s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560784}
{"loss": 24.0, "grad_norm": 1.10749257, "learning_rate": 1.1e-07, "token_acc": 0.38221355, "epoch": 0.99392272, "global_step/max_steps": "2167/2181", "elapsed_time": "5h 45m 19s", "remaining_time": "2m 14s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560943}
{"loss": 23.875, "grad_norm": 1.0991652, "learning_rate": 9e-08, "token_acc": 0.37984716, "epoch": 0.99438138, "global_step/max_steps": "2168/2181", "elapsed_time": "5h 45m 28s", "remaining_time": "2m 4s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560916}
{"loss": 23.25, "grad_norm": 1.01441276, "learning_rate": 8e-08, "token_acc": 0.38787879, "epoch": 0.99484004, "global_step/max_steps": "2169/2181", "elapsed_time": "5h 45m 39s", "remaining_time": "1m 55s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.561325}
{"loss": 23.75, "grad_norm": 1.03842926, "learning_rate": 7e-08, "token_acc": 0.38871824, "epoch": 0.9952987, "global_step/max_steps": "2170/2181", "elapsed_time": "5h 45m 48s", "remaining_time": "1m 45s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.561263}
{"loss": 23.625, "grad_norm": 1.06383049, "learning_rate": 6e-08, "token_acc": 0.39343365, "epoch": 0.99575737, "global_step/max_steps": "2171/2181", "elapsed_time": "5h 45m 58s", "remaining_time": "1m 36s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.561403}
{"loss": 23.5, "grad_norm": 1.02303839, "learning_rate": 4e-08, "token_acc": 0.38188529, "epoch": 0.99621603, "global_step/max_steps": "2172/2181", "elapsed_time": "5h 46m 7s", "remaining_time": "1m 26s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.561252}
{"loss": 24.625, "grad_norm": 1.13398051, "learning_rate": 4e-08, "token_acc": 0.37004646, "epoch": 0.99667469, "global_step/max_steps": "2173/2181", "elapsed_time": "5h 46m 17s", "remaining_time": "1m 16s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.561315}
{"loss": 23.75, "grad_norm": 0.98532456, "learning_rate": 3e-08, "token_acc": 0.39261168, "epoch": 0.99713336, "global_step/max_steps": "2174/2181", "elapsed_time": "5h 46m 26s", "remaining_time": "1m 7s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.561193}
{"loss": 24.25, "grad_norm": 1.06413496, "learning_rate": 2e-08, "token_acc": 0.38070078, "epoch": 0.99759202, "global_step/max_steps": "2175/2181", "elapsed_time": "5h 46m 35s", "remaining_time": "57s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.561081}
{"loss": 23.75, "grad_norm": 0.929281, "learning_rate": 1e-08, "token_acc": 0.39236981, "epoch": 0.99805068, "global_step/max_steps": "2176/2181", "elapsed_time": "5h 46m 45s", "remaining_time": "48s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.561054}
{"loss": 23.125, "grad_norm": 1.02126443, "learning_rate": 1e-08, "token_acc": 0.39344729, "epoch": 0.99850935, "global_step/max_steps": "2177/2181", "elapsed_time": "5h 46m 54s", "remaining_time": "38s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560975}
{"loss": 24.5, "grad_norm": 1.13666964, "learning_rate": 0.0, "token_acc": 0.37935693, "epoch": 0.99896801, "global_step/max_steps": "2178/2181", "elapsed_time": "5h 47m 4s", "remaining_time": "29s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.560984}
{"loss": 23.5, "grad_norm": 0.9544096, "learning_rate": 0.0, "token_acc": 0.38686534, "epoch": 0.99942667, "global_step/max_steps": "2179/2181", "elapsed_time": "5h 47m 14s", "remaining_time": "19s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.561461}
{"loss": 23.25, "grad_norm": 0.96375948, "learning_rate": 0.0, "token_acc": 0.39860335, "epoch": 0.99988533, "global_step/max_steps": "2180/2181", "elapsed_time": "5h 47m 25s", "remaining_time": "10s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.561759}
{"loss": 7.9375, "grad_norm": 1.14567924, "learning_rate": 0.0, "token_acc": 0.39201183, "epoch": 1.0, "global_step/max_steps": "2181/2181", "elapsed_time": "5h 47m 26s", "remaining_time": "0s", "memory(GiB)": 68.92, "train_speed(s/it)": 9.558043}