| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 200, | |
| "global_step": 1004, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00199203187250996, | |
| "grad_norm": 5.302925716458761, | |
| "learning_rate": 9.99997552222299e-06, | |
| "loss": 0.4962, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00398406374501992, | |
| "grad_norm": 4.0068584615292755, | |
| "learning_rate": 9.999902089131626e-06, | |
| "loss": 0.5131, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00597609561752988, | |
| "grad_norm": 3.2685057084225533, | |
| "learning_rate": 9.999779701444897e-06, | |
| "loss": 0.3642, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.00796812749003984, | |
| "grad_norm": 2.969388810564799, | |
| "learning_rate": 9.999608360361114e-06, | |
| "loss": 0.4366, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0099601593625498, | |
| "grad_norm": 2.3257277817224877, | |
| "learning_rate": 9.9993880675579e-06, | |
| "loss": 0.2805, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01195219123505976, | |
| "grad_norm": 2.316100445863212, | |
| "learning_rate": 9.999118825192162e-06, | |
| "loss": 0.3663, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.013944223107569721, | |
| "grad_norm": 2.0546227682793825, | |
| "learning_rate": 9.998800635900085e-06, | |
| "loss": 0.2761, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.01593625498007968, | |
| "grad_norm": 2.6542621051493294, | |
| "learning_rate": 9.998433502797097e-06, | |
| "loss": 0.3528, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.017928286852589643, | |
| "grad_norm": 2.184358255514989, | |
| "learning_rate": 9.998017429477834e-06, | |
| "loss": 0.2891, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0199203187250996, | |
| "grad_norm": 2.210497121507057, | |
| "learning_rate": 9.99755242001612e-06, | |
| "loss": 0.3052, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.021912350597609563, | |
| "grad_norm": 2.391291694670009, | |
| "learning_rate": 9.99703847896491e-06, | |
| "loss": 0.3088, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02390438247011952, | |
| "grad_norm": 1.8867370891248774, | |
| "learning_rate": 9.996475611356265e-06, | |
| "loss": 0.2799, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.025896414342629483, | |
| "grad_norm": 1.472434411377998, | |
| "learning_rate": 9.995863822701278e-06, | |
| "loss": 0.1877, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.027888446215139442, | |
| "grad_norm": 1.550344203694589, | |
| "learning_rate": 9.99520311899004e-06, | |
| "loss": 0.1703, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.029880478087649404, | |
| "grad_norm": 2.6379971141320264, | |
| "learning_rate": 9.994493506691577e-06, | |
| "loss": 0.2986, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03187250996015936, | |
| "grad_norm": 2.1294276467465734, | |
| "learning_rate": 9.993734992753777e-06, | |
| "loss": 0.2666, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.03386454183266932, | |
| "grad_norm": 1.9769257581715476, | |
| "learning_rate": 9.992927584603339e-06, | |
| "loss": 0.2376, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.035856573705179286, | |
| "grad_norm": 2.262018155817254, | |
| "learning_rate": 9.992071290145684e-06, | |
| "loss": 0.2436, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.037848605577689244, | |
| "grad_norm": 2.1357744572584605, | |
| "learning_rate": 9.991166117764885e-06, | |
| "loss": 0.2337, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0398406374501992, | |
| "grad_norm": 2.065609976313799, | |
| "learning_rate": 9.990212076323587e-06, | |
| "loss": 0.2614, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04183266932270916, | |
| "grad_norm": 2.0110212882357894, | |
| "learning_rate": 9.989209175162912e-06, | |
| "loss": 0.2282, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.043824701195219126, | |
| "grad_norm": 2.0333703620396433, | |
| "learning_rate": 9.988157424102381e-06, | |
| "loss": 0.2166, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.045816733067729085, | |
| "grad_norm": 1.7747686699151117, | |
| "learning_rate": 9.9870568334398e-06, | |
| "loss": 0.1755, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.04780876494023904, | |
| "grad_norm": 1.610758994875989, | |
| "learning_rate": 9.98590741395118e-06, | |
| "loss": 0.1979, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.049800796812749, | |
| "grad_norm": 1.8977306151271889, | |
| "learning_rate": 9.98470917689061e-06, | |
| "loss": 0.2244, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05179282868525897, | |
| "grad_norm": 1.7649852344717443, | |
| "learning_rate": 9.983462133990163e-06, | |
| "loss": 0.1947, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.053784860557768925, | |
| "grad_norm": 2.1142885695917877, | |
| "learning_rate": 9.982166297459775e-06, | |
| "loss": 0.2156, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.055776892430278883, | |
| "grad_norm": 2.278915052614827, | |
| "learning_rate": 9.980821679987125e-06, | |
| "loss": 0.1901, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.05776892430278884, | |
| "grad_norm": 2.038139109307458, | |
| "learning_rate": 9.979428294737509e-06, | |
| "loss": 0.2118, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.05976095617529881, | |
| "grad_norm": 2.133948012877493, | |
| "learning_rate": 9.97798615535372e-06, | |
| "loss": 0.2146, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.061752988047808766, | |
| "grad_norm": 2.176892858037584, | |
| "learning_rate": 9.976495275955904e-06, | |
| "loss": 0.2097, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.06374501992031872, | |
| "grad_norm": 1.9943827766271132, | |
| "learning_rate": 9.974955671141425e-06, | |
| "loss": 0.1968, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.06573705179282868, | |
| "grad_norm": 1.9835967650496324, | |
| "learning_rate": 9.973367355984724e-06, | |
| "loss": 0.2116, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.06772908366533864, | |
| "grad_norm": 2.0204571026763434, | |
| "learning_rate": 9.971730346037172e-06, | |
| "loss": 0.2026, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0697211155378486, | |
| "grad_norm": 1.9322301469201182, | |
| "learning_rate": 9.970044657326913e-06, | |
| "loss": 0.2053, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.07171314741035857, | |
| "grad_norm": 1.947308039962251, | |
| "learning_rate": 9.968310306358715e-06, | |
| "loss": 0.1917, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.07370517928286853, | |
| "grad_norm": 2.046915884396975, | |
| "learning_rate": 9.966527310113798e-06, | |
| "loss": 0.2365, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.07569721115537849, | |
| "grad_norm": 2.09697417316823, | |
| "learning_rate": 9.964695686049676e-06, | |
| "loss": 0.2343, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.07768924302788845, | |
| "grad_norm": 2.6849483307841076, | |
| "learning_rate": 9.962815452099985e-06, | |
| "loss": 0.2169, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.0796812749003984, | |
| "grad_norm": 1.8234840582625904, | |
| "learning_rate": 9.960886626674302e-06, | |
| "loss": 0.2119, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08167330677290836, | |
| "grad_norm": 1.7256360371957784, | |
| "learning_rate": 9.95890922865797e-06, | |
| "loss": 0.1769, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.08366533864541832, | |
| "grad_norm": 1.682991846433481, | |
| "learning_rate": 9.956883277411914e-06, | |
| "loss": 0.1724, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.08565737051792828, | |
| "grad_norm": 2.167281430513665, | |
| "learning_rate": 9.954808792772447e-06, | |
| "loss": 0.2185, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.08764940239043825, | |
| "grad_norm": 1.7035143668249393, | |
| "learning_rate": 9.952685795051078e-06, | |
| "loss": 0.1701, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.08964143426294821, | |
| "grad_norm": 1.9522204858955863, | |
| "learning_rate": 9.95051430503431e-06, | |
| "loss": 0.1977, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.09163346613545817, | |
| "grad_norm": 1.9293284789264147, | |
| "learning_rate": 9.948294343983446e-06, | |
| "loss": 0.1995, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.09362549800796813, | |
| "grad_norm": 2.221989822189503, | |
| "learning_rate": 9.94602593363437e-06, | |
| "loss": 0.2135, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.09561752988047809, | |
| "grad_norm": 2.5630931672539123, | |
| "learning_rate": 9.943709096197334e-06, | |
| "loss": 0.1935, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.09760956175298804, | |
| "grad_norm": 2.2472057427928256, | |
| "learning_rate": 9.941343854356757e-06, | |
| "loss": 0.1932, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.099601593625498, | |
| "grad_norm": 2.2907511035454076, | |
| "learning_rate": 9.938930231270982e-06, | |
| "loss": 0.2136, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.10159362549800798, | |
| "grad_norm": 2.1818748427349854, | |
| "learning_rate": 9.93646825057206e-06, | |
| "loss": 0.2028, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.10358565737051793, | |
| "grad_norm": 1.9532878719288238, | |
| "learning_rate": 9.933957936365515e-06, | |
| "loss": 0.1867, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.10557768924302789, | |
| "grad_norm": 2.2638694550211667, | |
| "learning_rate": 9.931399313230112e-06, | |
| "loss": 0.2124, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.10756972111553785, | |
| "grad_norm": 2.0453202854269597, | |
| "learning_rate": 9.928792406217615e-06, | |
| "loss": 0.2013, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.10956175298804781, | |
| "grad_norm": 2.092666084016267, | |
| "learning_rate": 9.926137240852539e-06, | |
| "loss": 0.1815, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.11155378486055777, | |
| "grad_norm": 1.7643433151250012, | |
| "learning_rate": 9.9234338431319e-06, | |
| "loss": 0.1758, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.11354581673306773, | |
| "grad_norm": 1.7711425457667331, | |
| "learning_rate": 9.920682239524968e-06, | |
| "loss": 0.1732, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.11553784860557768, | |
| "grad_norm": 1.97738896833117, | |
| "learning_rate": 9.917882456972999e-06, | |
| "loss": 0.1995, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.11752988047808766, | |
| "grad_norm": 1.4375108861286383, | |
| "learning_rate": 9.915034522888972e-06, | |
| "loss": 0.1485, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.11952191235059761, | |
| "grad_norm": 2.0375847131853293, | |
| "learning_rate": 9.912138465157325e-06, | |
| "loss": 0.2021, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.12151394422310757, | |
| "grad_norm": 2.168757999506768, | |
| "learning_rate": 9.909194312133681e-06, | |
| "loss": 0.172, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.12350597609561753, | |
| "grad_norm": 1.859742755561109, | |
| "learning_rate": 9.90620209264457e-06, | |
| "loss": 0.1538, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.1254980079681275, | |
| "grad_norm": 1.800175258878161, | |
| "learning_rate": 9.90316183598714e-06, | |
| "loss": 0.1643, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.12749003984063745, | |
| "grad_norm": 1.780289250472305, | |
| "learning_rate": 9.900073571928887e-06, | |
| "loss": 0.1448, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.1294820717131474, | |
| "grad_norm": 1.8989241084218624, | |
| "learning_rate": 9.896937330707341e-06, | |
| "loss": 0.1661, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.13147410358565736, | |
| "grad_norm": 1.7382723265927287, | |
| "learning_rate": 9.893753143029792e-06, | |
| "loss": 0.1794, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.13346613545816732, | |
| "grad_norm": 1.6940222793391266, | |
| "learning_rate": 9.89052104007297e-06, | |
| "loss": 0.154, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.13545816733067728, | |
| "grad_norm": 2.091556995024388, | |
| "learning_rate": 9.887241053482756e-06, | |
| "loss": 0.1795, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.13745019920318724, | |
| "grad_norm": 2.378952761035815, | |
| "learning_rate": 9.883913215373862e-06, | |
| "loss": 0.172, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.1394422310756972, | |
| "grad_norm": 2.0726149359464143, | |
| "learning_rate": 9.880537558329518e-06, | |
| "loss": 0.16, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.14143426294820718, | |
| "grad_norm": 1.8509199481149878, | |
| "learning_rate": 9.877114115401159e-06, | |
| "loss": 0.151, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.14342629482071714, | |
| "grad_norm": 1.9383364231106874, | |
| "learning_rate": 9.87364292010809e-06, | |
| "loss": 0.161, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.1454183266932271, | |
| "grad_norm": 1.9835120874739343, | |
| "learning_rate": 9.870124006437172e-06, | |
| "loss": 0.1885, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.14741035856573706, | |
| "grad_norm": 1.7057694570966788, | |
| "learning_rate": 9.866557408842479e-06, | |
| "loss": 0.1619, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.14940239043824702, | |
| "grad_norm": 1.7963768846634989, | |
| "learning_rate": 9.86294316224496e-06, | |
| "loss": 0.1438, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.15139442231075698, | |
| "grad_norm": 1.8886735096348262, | |
| "learning_rate": 9.859281302032107e-06, | |
| "loss": 0.1753, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.15338645418326693, | |
| "grad_norm": 2.099235241932376, | |
| "learning_rate": 9.855571864057598e-06, | |
| "loss": 0.1837, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.1553784860557769, | |
| "grad_norm": 2.1695503922620683, | |
| "learning_rate": 9.85181488464095e-06, | |
| "loss": 0.1776, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.15737051792828685, | |
| "grad_norm": 2.2241373204323884, | |
| "learning_rate": 9.848010400567167e-06, | |
| "loss": 0.1762, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.1593625498007968, | |
| "grad_norm": 2.0073012916773556, | |
| "learning_rate": 9.844158449086372e-06, | |
| "loss": 0.1643, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.16135458167330677, | |
| "grad_norm": 1.900677106194213, | |
| "learning_rate": 9.84025906791345e-06, | |
| "loss": 0.1647, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.16334661354581673, | |
| "grad_norm": 2.2420248448336646, | |
| "learning_rate": 9.836312295227674e-06, | |
| "loss": 0.1904, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.16533864541832669, | |
| "grad_norm": 2.0352049260820966, | |
| "learning_rate": 9.832318169672334e-06, | |
| "loss": 0.1652, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.16733067729083664, | |
| "grad_norm": 2.1040895543260834, | |
| "learning_rate": 9.828276730354353e-06, | |
| "loss": 0.1681, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.1693227091633466, | |
| "grad_norm": 2.0323467551141543, | |
| "learning_rate": 9.824188016843915e-06, | |
| "loss": 0.1646, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.17131474103585656, | |
| "grad_norm": 1.7648807229029235, | |
| "learning_rate": 9.820052069174062e-06, | |
| "loss": 0.1581, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.17330677290836655, | |
| "grad_norm": 2.256460995644407, | |
| "learning_rate": 9.81586892784032e-06, | |
| "loss": 0.1937, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.1752988047808765, | |
| "grad_norm": 2.1166732483303745, | |
| "learning_rate": 9.811638633800287e-06, | |
| "loss": 0.189, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.17729083665338646, | |
| "grad_norm": 2.1883580991082527, | |
| "learning_rate": 9.807361228473241e-06, | |
| "loss": 0.2019, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.17928286852589642, | |
| "grad_norm": 1.8188619314670527, | |
| "learning_rate": 9.803036753739733e-06, | |
| "loss": 0.1633, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.18127490039840638, | |
| "grad_norm": 2.130272928076839, | |
| "learning_rate": 9.798665251941172e-06, | |
| "loss": 0.1858, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.18326693227091634, | |
| "grad_norm": 1.9280374177349535, | |
| "learning_rate": 9.794246765879421e-06, | |
| "loss": 0.1367, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.1852589641434263, | |
| "grad_norm": 2.0068283142732564, | |
| "learning_rate": 9.789781338816362e-06, | |
| "loss": 0.1552, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.18725099601593626, | |
| "grad_norm": 2.1363237967962387, | |
| "learning_rate": 9.785269014473487e-06, | |
| "loss": 0.1635, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.1892430278884462, | |
| "grad_norm": 2.2218212298896978, | |
| "learning_rate": 9.780709837031464e-06, | |
| "loss": 0.1625, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.19123505976095617, | |
| "grad_norm": 1.9767577152960487, | |
| "learning_rate": 9.776103851129706e-06, | |
| "loss": 0.1632, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.19322709163346613, | |
| "grad_norm": 2.3121968917141094, | |
| "learning_rate": 9.77145110186593e-06, | |
| "loss": 0.1621, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.1952191235059761, | |
| "grad_norm": 1.7755655818003055, | |
| "learning_rate": 9.766751634795719e-06, | |
| "loss": 0.1669, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.19721115537848605, | |
| "grad_norm": 2.240571790906682, | |
| "learning_rate": 9.762005495932076e-06, | |
| "loss": 0.1637, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.199203187250996, | |
| "grad_norm": 1.6446940187832602, | |
| "learning_rate": 9.757212731744973e-06, | |
| "loss": 0.1302, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.20119521912350596, | |
| "grad_norm": 2.0617960985029504, | |
| "learning_rate": 9.752373389160896e-06, | |
| "loss": 0.1707, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.20318725099601595, | |
| "grad_norm": 2.1835252538751413, | |
| "learning_rate": 9.747487515562384e-06, | |
| "loss": 0.1649, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.2051792828685259, | |
| "grad_norm": 2.0129569799572407, | |
| "learning_rate": 9.742555158787567e-06, | |
| "loss": 0.1637, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.20717131474103587, | |
| "grad_norm": 2.524863713870793, | |
| "learning_rate": 9.737576367129694e-06, | |
| "loss": 0.1816, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.20916334661354583, | |
| "grad_norm": 1.9649026368356415, | |
| "learning_rate": 9.73255118933667e-06, | |
| "loss": 0.1216, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.21115537848605578, | |
| "grad_norm": 1.6210946660994265, | |
| "learning_rate": 9.727479674610565e-06, | |
| "loss": 0.1424, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.21314741035856574, | |
| "grad_norm": 2.371276035809818, | |
| "learning_rate": 9.722361872607142e-06, | |
| "loss": 0.1943, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.2151394422310757, | |
| "grad_norm": 1.760306649155103, | |
| "learning_rate": 9.717197833435367e-06, | |
| "loss": 0.1588, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.21713147410358566, | |
| "grad_norm": 1.850921085250863, | |
| "learning_rate": 9.71198760765692e-06, | |
| "loss": 0.135, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.21912350597609562, | |
| "grad_norm": 1.91926012170198, | |
| "learning_rate": 9.706731246285701e-06, | |
| "loss": 0.146, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.22111553784860558, | |
| "grad_norm": 1.9074743120890272, | |
| "learning_rate": 9.701428800787325e-06, | |
| "loss": 0.1546, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.22310756972111553, | |
| "grad_norm": 1.8574536892339737, | |
| "learning_rate": 9.696080323078621e-06, | |
| "loss": 0.15, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.2250996015936255, | |
| "grad_norm": 1.9674896794532148, | |
| "learning_rate": 9.690685865527132e-06, | |
| "loss": 0.1605, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.22709163346613545, | |
| "grad_norm": 1.683408319737727, | |
| "learning_rate": 9.685245480950584e-06, | |
| "loss": 0.1483, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.2290836653386454, | |
| "grad_norm": 1.8357325701673104, | |
| "learning_rate": 9.679759222616389e-06, | |
| "loss": 0.1342, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.23107569721115537, | |
| "grad_norm": 1.7645234238909926, | |
| "learning_rate": 9.67422714424111e-06, | |
| "loss": 0.1496, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.23306772908366533, | |
| "grad_norm": 2.3580150329396097, | |
| "learning_rate": 9.668649299989939e-06, | |
| "loss": 0.1602, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.2350597609561753, | |
| "grad_norm": 1.9878477683225815, | |
| "learning_rate": 9.663025744476167e-06, | |
| "loss": 0.1592, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.23705179282868527, | |
| "grad_norm": 2.1066728943604685, | |
| "learning_rate": 9.657356532760647e-06, | |
| "loss": 0.1586, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.23904382470119523, | |
| "grad_norm": 2.4857773499345814, | |
| "learning_rate": 9.651641720351262e-06, | |
| "loss": 0.1706, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2410358565737052, | |
| "grad_norm": 1.9669140495950204, | |
| "learning_rate": 9.645881363202371e-06, | |
| "loss": 0.135, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.24302788844621515, | |
| "grad_norm": 2.129283733459999, | |
| "learning_rate": 9.640075517714272e-06, | |
| "loss": 0.1785, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.2450199203187251, | |
| "grad_norm": 1.9937349989457298, | |
| "learning_rate": 9.634224240732641e-06, | |
| "loss": 0.1453, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.24701195219123506, | |
| "grad_norm": 2.1385569770971298, | |
| "learning_rate": 9.628327589547977e-06, | |
| "loss": 0.1475, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.24900398406374502, | |
| "grad_norm": 2.28108046376594, | |
| "learning_rate": 9.622385621895046e-06, | |
| "loss": 0.1813, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.250996015936255, | |
| "grad_norm": 2.148801989847443, | |
| "learning_rate": 9.616398395952313e-06, | |
| "loss": 0.1683, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.25298804780876494, | |
| "grad_norm": 1.8652803065183108, | |
| "learning_rate": 9.610365970341369e-06, | |
| "loss": 0.1494, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.2549800796812749, | |
| "grad_norm": 1.8989160712067141, | |
| "learning_rate": 9.604288404126362e-06, | |
| "loss": 0.146, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.25697211155378485, | |
| "grad_norm": 1.7384157916220273, | |
| "learning_rate": 9.598165756813418e-06, | |
| "loss": 0.1359, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.2589641434262948, | |
| "grad_norm": 2.2085278004769324, | |
| "learning_rate": 9.591998088350055e-06, | |
| "loss": 0.1784, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.26095617529880477, | |
| "grad_norm": 1.6343923035740249, | |
| "learning_rate": 9.585785459124595e-06, | |
| "loss": 0.1376, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.26294820717131473, | |
| "grad_norm": 2.1234396006125396, | |
| "learning_rate": 9.579527929965581e-06, | |
| "loss": 0.1702, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.2649402390438247, | |
| "grad_norm": 1.7801773475854494, | |
| "learning_rate": 9.573225562141174e-06, | |
| "loss": 0.1474, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.26693227091633465, | |
| "grad_norm": 1.4560862756744588, | |
| "learning_rate": 9.566878417358559e-06, | |
| "loss": 0.1246, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.2689243027888446, | |
| "grad_norm": 2.025058996733023, | |
| "learning_rate": 9.56048655776333e-06, | |
| "loss": 0.1639, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.27091633466135456, | |
| "grad_norm": 1.8323049535503093, | |
| "learning_rate": 9.554050045938893e-06, | |
| "loss": 0.1452, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.2729083665338645, | |
| "grad_norm": 1.8874907449560725, | |
| "learning_rate": 9.54756894490585e-06, | |
| "loss": 0.1502, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.2749003984063745, | |
| "grad_norm": 2.5711934437137316, | |
| "learning_rate": 9.541043318121379e-06, | |
| "loss": 0.1767, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.27689243027888444, | |
| "grad_norm": 2.6684694094439223, | |
| "learning_rate": 9.534473229478613e-06, | |
| "loss": 0.1809, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.2788844621513944, | |
| "grad_norm": 1.8024037893437692, | |
| "learning_rate": 9.52785874330602e-06, | |
| "loss": 0.139, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.28087649402390436, | |
| "grad_norm": 2.0746711127055155, | |
| "learning_rate": 9.521199924366766e-06, | |
| "loss": 0.1645, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.28286852589641437, | |
| "grad_norm": 1.8375760908575898, | |
| "learning_rate": 9.514496837858085e-06, | |
| "loss": 0.1409, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.2848605577689243, | |
| "grad_norm": 2.155790335928457, | |
| "learning_rate": 9.507749549410641e-06, | |
| "loss": 0.1527, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.2868525896414343, | |
| "grad_norm": 1.9735656021269132, | |
| "learning_rate": 9.500958125087882e-06, | |
| "loss": 0.1454, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.28884462151394424, | |
| "grad_norm": 1.9447279614136819, | |
| "learning_rate": 9.494122631385397e-06, | |
| "loss": 0.147, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.2908366533864542, | |
| "grad_norm": 1.8420179664244611, | |
| "learning_rate": 9.487243135230259e-06, | |
| "loss": 0.1266, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.29282868525896416, | |
| "grad_norm": 2.0895817258939613, | |
| "learning_rate": 9.480319703980382e-06, | |
| "loss": 0.1752, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.2948207171314741, | |
| "grad_norm": 2.368984572890111, | |
| "learning_rate": 9.473352405423845e-06, | |
| "loss": 0.1716, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.2968127490039841, | |
| "grad_norm": 2.3614919475438336, | |
| "learning_rate": 9.466341307778239e-06, | |
| "loss": 0.1853, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.29880478087649404, | |
| "grad_norm": 1.7497898226508322, | |
| "learning_rate": 9.459286479690002e-06, | |
| "loss": 0.1487, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.300796812749004, | |
| "grad_norm": 2.319602310333304, | |
| "learning_rate": 9.452187990233737e-06, | |
| "loss": 0.1564, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.30278884462151395, | |
| "grad_norm": 1.9314144164174814, | |
| "learning_rate": 9.445045908911536e-06, | |
| "loss": 0.1574, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.3047808764940239, | |
| "grad_norm": 2.0498544875744726, | |
| "learning_rate": 9.437860305652314e-06, | |
| "loss": 0.1609, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.30677290836653387, | |
| "grad_norm": 2.1271214704509895, | |
| "learning_rate": 9.430631250811107e-06, | |
| "loss": 0.1453, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.30876494023904383, | |
| "grad_norm": 1.5971982417533808, | |
| "learning_rate": 9.42335881516839e-06, | |
| "loss": 0.1293, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3107569721115538, | |
| "grad_norm": 1.7060046369161204, | |
| "learning_rate": 9.416043069929389e-06, | |
| "loss": 0.1308, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.31274900398406374, | |
| "grad_norm": 2.2558369441074744, | |
| "learning_rate": 9.408684086723375e-06, | |
| "loss": 0.1894, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.3147410358565737, | |
| "grad_norm": 2.040622936675236, | |
| "learning_rate": 9.401281937602966e-06, | |
| "loss": 0.1536, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.31673306772908366, | |
| "grad_norm": 2.021178408176851, | |
| "learning_rate": 9.393836695043429e-06, | |
| "loss": 0.1316, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.3187250996015936, | |
| "grad_norm": 1.9053653174415077, | |
| "learning_rate": 9.386348431941953e-06, | |
| "loss": 0.1445, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3207171314741036, | |
| "grad_norm": 2.3821069319286576, | |
| "learning_rate": 9.378817221616955e-06, | |
| "loss": 0.1498, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.32270916334661354, | |
| "grad_norm": 1.987196571819016, | |
| "learning_rate": 9.371243137807353e-06, | |
| "loss": 0.1351, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.3247011952191235, | |
| "grad_norm": 2.012726967199412, | |
| "learning_rate": 9.363626254671835e-06, | |
| "loss": 0.1263, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.32669322709163345, | |
| "grad_norm": 1.8276010793351352, | |
| "learning_rate": 9.355966646788152e-06, | |
| "loss": 0.1357, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.3286852589641434, | |
| "grad_norm": 1.8020618013754388, | |
| "learning_rate": 9.34826438915237e-06, | |
| "loss": 0.1235, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.33067729083665337, | |
| "grad_norm": 1.854178246940745, | |
| "learning_rate": 9.340519557178149e-06, | |
| "loss": 0.1481, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.33266932270916333, | |
| "grad_norm": 1.9660957079874257, | |
| "learning_rate": 9.332732226695997e-06, | |
| "loss": 0.1354, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.3346613545816733, | |
| "grad_norm": 1.903678628134176, | |
| "learning_rate": 9.324902473952529e-06, | |
| "loss": 0.154, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.33665338645418325, | |
| "grad_norm": 1.8793502875666162, | |
| "learning_rate": 9.317030375609721e-06, | |
| "loss": 0.1415, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.3386454183266932, | |
| "grad_norm": 2.2655417937058355, | |
| "learning_rate": 9.309116008744164e-06, | |
| "loss": 0.1805, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.34063745019920316, | |
| "grad_norm": 2.224518864361946, | |
| "learning_rate": 9.301159450846296e-06, | |
| "loss": 0.1555, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.3426294820717131, | |
| "grad_norm": 2.594561985506128, | |
| "learning_rate": 9.293160779819658e-06, | |
| "loss": 0.1695, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.34462151394422313, | |
| "grad_norm": 2.0293944680745257, | |
| "learning_rate": 9.285120073980127e-06, | |
| "loss": 0.147, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.3466135458167331, | |
| "grad_norm": 1.8842194158472625, | |
| "learning_rate": 9.277037412055143e-06, | |
| "loss": 0.1397, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.34860557768924305, | |
| "grad_norm": 2.1876611599257294, | |
| "learning_rate": 9.268912873182945e-06, | |
| "loss": 0.1646, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.350597609561753, | |
| "grad_norm": 2.0439657178756083, | |
| "learning_rate": 9.260746536911792e-06, | |
| "loss": 0.1583, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.35258964143426297, | |
| "grad_norm": 2.3673755824284095, | |
| "learning_rate": 9.25253848319919e-06, | |
| "loss": 0.1779, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.3545816733067729, | |
| "grad_norm": 1.849959819313811, | |
| "learning_rate": 9.244288792411099e-06, | |
| "loss": 0.145, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.3565737051792829, | |
| "grad_norm": 1.8494333009850685, | |
| "learning_rate": 9.235997545321156e-06, | |
| "loss": 0.1402, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.35856573705179284, | |
| "grad_norm": 1.8409139333031814, | |
| "learning_rate": 9.227664823109884e-06, | |
| "loss": 0.1472, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3605577689243028, | |
| "grad_norm": 2.198400559444892, | |
| "learning_rate": 9.219290707363885e-06, | |
| "loss": 0.1683, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.36254980079681276, | |
| "grad_norm": 1.847947075455182, | |
| "learning_rate": 9.210875280075056e-06, | |
| "loss": 0.1472, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.3645418326693227, | |
| "grad_norm": 1.9180035039553798, | |
| "learning_rate": 9.202418623639779e-06, | |
| "loss": 0.1473, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.3665338645418327, | |
| "grad_norm": 2.1866401096601358, | |
| "learning_rate": 9.193920820858113e-06, | |
| "loss": 0.1744, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.36852589641434264, | |
| "grad_norm": 1.8712622384696984, | |
| "learning_rate": 9.185381954932984e-06, | |
| "loss": 0.1566, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.3705179282868526, | |
| "grad_norm": 1.7798834216487947, | |
| "learning_rate": 9.17680210946938e-06, | |
| "loss": 0.1236, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.37250996015936255, | |
| "grad_norm": 1.82049432105803, | |
| "learning_rate": 9.168181368473514e-06, | |
| "loss": 0.146, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.3745019920318725, | |
| "grad_norm": 2.0877726702358923, | |
| "learning_rate": 9.159519816352021e-06, | |
| "loss": 0.1389, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.37649402390438247, | |
| "grad_norm": 2.051030898440553, | |
| "learning_rate": 9.150817537911111e-06, | |
| "loss": 0.1448, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.3784860557768924, | |
| "grad_norm": 1.7874327218426942, | |
| "learning_rate": 9.142074618355763e-06, | |
| "loss": 0.1214, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3804780876494024, | |
| "grad_norm": 2.347061822643037, | |
| "learning_rate": 9.133291143288865e-06, | |
| "loss": 0.1751, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.38247011952191234, | |
| "grad_norm": 1.8742517441369677, | |
| "learning_rate": 9.124467198710401e-06, | |
| "loss": 0.1326, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.3844621513944223, | |
| "grad_norm": 1.8857929156382558, | |
| "learning_rate": 9.115602871016585e-06, | |
| "loss": 0.1261, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.38645418326693226, | |
| "grad_norm": 1.9824443306668558, | |
| "learning_rate": 9.106698246999036e-06, | |
| "loss": 0.1461, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.3884462151394422, | |
| "grad_norm": 2.0467435985376348, | |
| "learning_rate": 9.097753413843909e-06, | |
| "loss": 0.1384, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.3904382470119522, | |
| "grad_norm": 2.141133625722717, | |
| "learning_rate": 9.08876845913106e-06, | |
| "loss": 0.1497, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.39243027888446214, | |
| "grad_norm": 2.234122705708421, | |
| "learning_rate": 9.079743470833177e-06, | |
| "loss": 0.145, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.3944223107569721, | |
| "grad_norm": 2.682514937204839, | |
| "learning_rate": 9.070678537314919e-06, | |
| "loss": 0.1762, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.39641434262948205, | |
| "grad_norm": 2.2720011535768885, | |
| "learning_rate": 9.061573747332053e-06, | |
| "loss": 0.154, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.398406374501992, | |
| "grad_norm": 1.9823400380634495, | |
| "learning_rate": 9.052429190030589e-06, | |
| "loss": 0.1573, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.398406374501992, | |
| "eval_loss": 0.14404042065143585, | |
| "eval_runtime": 3.1342, | |
| "eval_samples_per_second": 13.082, | |
| "eval_steps_per_second": 3.51, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.40039840637450197, | |
| "grad_norm": 2.4201576423705538, | |
| "learning_rate": 9.0432449549459e-06, | |
| "loss": 0.1734, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.40239043824701193, | |
| "grad_norm": 1.8987324574568356, | |
| "learning_rate": 9.03402113200185e-06, | |
| "loss": 0.1357, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.4043824701195219, | |
| "grad_norm": 1.8455215459039969, | |
| "learning_rate": 9.02475781150991e-06, | |
| "loss": 0.1412, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.4063745019920319, | |
| "grad_norm": 2.1617514944433367, | |
| "learning_rate": 9.015455084168279e-06, | |
| "loss": 0.1658, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.40836653386454186, | |
| "grad_norm": 2.1331463500913665, | |
| "learning_rate": 9.00611304106099e-06, | |
| "loss": 0.1647, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4103585657370518, | |
| "grad_norm": 1.8588463567752136, | |
| "learning_rate": 8.996731773657022e-06, | |
| "loss": 0.1615, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.4123505976095618, | |
| "grad_norm": 1.5786356112405115, | |
| "learning_rate": 8.987311373809405e-06, | |
| "loss": 0.1381, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.41434262948207173, | |
| "grad_norm": 2.069308545418217, | |
| "learning_rate": 8.977851933754317e-06, | |
| "loss": 0.1485, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.4163346613545817, | |
| "grad_norm": 1.8545162202812966, | |
| "learning_rate": 8.968353546110181e-06, | |
| "loss": 0.1311, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.41832669322709165, | |
| "grad_norm": 1.7202119799614204, | |
| "learning_rate": 8.95881630387677e-06, | |
| "loss": 0.1265, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4203187250996016, | |
| "grad_norm": 2.1311229993111276, | |
| "learning_rate": 8.949240300434272e-06, | |
| "loss": 0.1544, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.42231075697211157, | |
| "grad_norm": 1.932887530350482, | |
| "learning_rate": 8.939625629542401e-06, | |
| "loss": 0.1358, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.4243027888446215, | |
| "grad_norm": 1.996095976241167, | |
| "learning_rate": 8.929972385339466e-06, | |
| "loss": 0.1395, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.4262948207171315, | |
| "grad_norm": 1.7960261511185895, | |
| "learning_rate": 8.92028066234145e-06, | |
| "loss": 0.1436, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.42828685258964144, | |
| "grad_norm": 1.6930133554882265, | |
| "learning_rate": 8.910550555441085e-06, | |
| "loss": 0.1246, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4302788844621514, | |
| "grad_norm": 2.225395808052067, | |
| "learning_rate": 8.900782159906927e-06, | |
| "loss": 0.165, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.43227091633466136, | |
| "grad_norm": 1.942615677966355, | |
| "learning_rate": 8.890975571382419e-06, | |
| "loss": 0.1418, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.4342629482071713, | |
| "grad_norm": 2.151401945627201, | |
| "learning_rate": 8.881130885884955e-06, | |
| "loss": 0.1405, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.4362549800796813, | |
| "grad_norm": 2.101742983984682, | |
| "learning_rate": 8.871248199804944e-06, | |
| "loss": 0.1415, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.43824701195219123, | |
| "grad_norm": 2.0803368131274973, | |
| "learning_rate": 8.861327609904859e-06, | |
| "loss": 0.1401, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4402390438247012, | |
| "grad_norm": 1.9987975497294423, | |
| "learning_rate": 8.851369213318293e-06, | |
| "loss": 0.1582, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.44223107569721115, | |
| "grad_norm": 1.882395487923144, | |
| "learning_rate": 8.841373107549014e-06, | |
| "loss": 0.1363, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.4442231075697211, | |
| "grad_norm": 1.7141338375852764, | |
| "learning_rate": 8.831339390469998e-06, | |
| "loss": 0.1304, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.44621513944223107, | |
| "grad_norm": 1.8845992727826126, | |
| "learning_rate": 8.821268160322482e-06, | |
| "loss": 0.1251, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.448207171314741, | |
| "grad_norm": 1.807320122447534, | |
| "learning_rate": 8.811159515714998e-06, | |
| "loss": 0.1202, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.450199203187251, | |
| "grad_norm": 1.993372692147205, | |
| "learning_rate": 8.801013555622403e-06, | |
| "loss": 0.1624, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.45219123505976094, | |
| "grad_norm": 1.8931180135396861, | |
| "learning_rate": 8.790830379384918e-06, | |
| "loss": 0.1408, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.4541832669322709, | |
| "grad_norm": 1.9138251654922296, | |
| "learning_rate": 8.780610086707149e-06, | |
| "loss": 0.1365, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.45617529880478086, | |
| "grad_norm": 1.9534031291438365, | |
| "learning_rate": 8.770352777657112e-06, | |
| "loss": 0.1363, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.4581673306772908, | |
| "grad_norm": 1.7954406324400658, | |
| "learning_rate": 8.760058552665262e-06, | |
| "loss": 0.1201, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.4601593625498008, | |
| "grad_norm": 2.0677561411750163, | |
| "learning_rate": 8.749727512523491e-06, | |
| "loss": 0.1479, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.46215139442231074, | |
| "grad_norm": 2.27753550624284, | |
| "learning_rate": 8.739359758384162e-06, | |
| "loss": 0.1417, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.4641434262948207, | |
| "grad_norm": 2.2517352339516523, | |
| "learning_rate": 8.728955391759102e-06, | |
| "loss": 0.1535, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.46613545816733065, | |
| "grad_norm": 2.070124167431417, | |
| "learning_rate": 8.718514514518617e-06, | |
| "loss": 0.1553, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.4681274900398406, | |
| "grad_norm": 2.064421588096535, | |
| "learning_rate": 8.708037228890494e-06, | |
| "loss": 0.1459, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.4701195219123506, | |
| "grad_norm": 1.8120445831055432, | |
| "learning_rate": 8.697523637458997e-06, | |
| "loss": 0.1391, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.4721115537848606, | |
| "grad_norm": 1.8340032254229388, | |
| "learning_rate": 8.686973843163868e-06, | |
| "loss": 0.1445, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.47410358565737054, | |
| "grad_norm": 2.139642376885461, | |
| "learning_rate": 8.676387949299307e-06, | |
| "loss": 0.1343, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.4760956175298805, | |
| "grad_norm": 2.0856211998803973, | |
| "learning_rate": 8.665766059512977e-06, | |
| "loss": 0.1422, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.47808764940239046, | |
| "grad_norm": 2.0329687486696906, | |
| "learning_rate": 8.655108277804975e-06, | |
| "loss": 0.124, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4800796812749004, | |
| "grad_norm": 1.8725092328941733, | |
| "learning_rate": 8.644414708526824e-06, | |
| "loss": 0.1319, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.4820717131474104, | |
| "grad_norm": 2.2032287252466585, | |
| "learning_rate": 8.63368545638045e-06, | |
| "loss": 0.1499, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.48406374501992033, | |
| "grad_norm": 2.04185058829803, | |
| "learning_rate": 8.622920626417141e-06, | |
| "loss": 0.1501, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.4860557768924303, | |
| "grad_norm": 2.003978218913384, | |
| "learning_rate": 8.612120324036548e-06, | |
| "loss": 0.1382, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.48804780876494025, | |
| "grad_norm": 1.8579294263600015, | |
| "learning_rate": 8.601284654985623e-06, | |
| "loss": 0.1342, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.4900398406374502, | |
| "grad_norm": 2.072837174686408, | |
| "learning_rate": 8.590413725357605e-06, | |
| "loss": 0.142, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.49203187250996017, | |
| "grad_norm": 1.9780709854474574, | |
| "learning_rate": 8.57950764159097e-06, | |
| "loss": 0.1395, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.4940239043824701, | |
| "grad_norm": 2.5022268847929054, | |
| "learning_rate": 8.568566510468392e-06, | |
| "loss": 0.1649, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.4960159362549801, | |
| "grad_norm": 1.7767948533765374, | |
| "learning_rate": 8.557590439115697e-06, | |
| "loss": 0.129, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.49800796812749004, | |
| "grad_norm": 1.8237813486682513, | |
| "learning_rate": 8.546579535000819e-06, | |
| "loss": 0.1429, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 2.108585214242547, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.151, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.50199203187251, | |
| "grad_norm": 1.8936980275001982, | |
| "learning_rate": 8.524453660060434e-06, | |
| "loss": 0.135, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.5039840637450199, | |
| "grad_norm": 1.953069313184316, | |
| "learning_rate": 8.513338905871819e-06, | |
| "loss": 0.1377, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.5059760956175299, | |
| "grad_norm": 1.6215779669447803, | |
| "learning_rate": 8.502189752192685e-06, | |
| "loss": 0.1269, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.5079681274900398, | |
| "grad_norm": 1.9115865609618454, | |
| "learning_rate": 8.491006308185632e-06, | |
| "loss": 0.1329, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5099601593625498, | |
| "grad_norm": 1.5921580761982146, | |
| "learning_rate": 8.479788683348996e-06, | |
| "loss": 0.1022, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.5119521912350598, | |
| "grad_norm": 2.0973700556988786, | |
| "learning_rate": 8.468536987515788e-06, | |
| "loss": 0.134, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.5139442231075697, | |
| "grad_norm": 2.045549212493722, | |
| "learning_rate": 8.457251330852608e-06, | |
| "loss": 0.1385, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.5159362549800797, | |
| "grad_norm": 1.9007334882981535, | |
| "learning_rate": 8.445931823858568e-06, | |
| "loss": 0.1566, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.5179282868525896, | |
| "grad_norm": 1.5993440106427084, | |
| "learning_rate": 8.434578577364218e-06, | |
| "loss": 0.1066, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5199203187250996, | |
| "grad_norm": 2.2603377167523973, | |
| "learning_rate": 8.423191702530453e-06, | |
| "loss": 0.1367, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.5219123505976095, | |
| "grad_norm": 2.096603538865956, | |
| "learning_rate": 8.411771310847426e-06, | |
| "loss": 0.1558, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.5239043824701195, | |
| "grad_norm": 2.104670890811567, | |
| "learning_rate": 8.400317514133454e-06, | |
| "loss": 0.1245, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.5258964143426295, | |
| "grad_norm": 2.1577997074465483, | |
| "learning_rate": 8.388830424533935e-06, | |
| "loss": 0.1587, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.5278884462151394, | |
| "grad_norm": 1.9717314334953373, | |
| "learning_rate": 8.377310154520232e-06, | |
| "loss": 0.1445, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5298804780876494, | |
| "grad_norm": 2.18540940693717, | |
| "learning_rate": 8.365756816888586e-06, | |
| "loss": 0.1369, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.5318725099601593, | |
| "grad_norm": 1.8217822120680458, | |
| "learning_rate": 8.354170524759008e-06, | |
| "loss": 0.1407, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.5338645418326693, | |
| "grad_norm": 1.4966007977041007, | |
| "learning_rate": 8.342551391574165e-06, | |
| "loss": 0.1101, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.5358565737051793, | |
| "grad_norm": 1.5671763413455242, | |
| "learning_rate": 8.33089953109828e-06, | |
| "loss": 0.1185, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.5378486055776892, | |
| "grad_norm": 1.8560918618066358, | |
| "learning_rate": 8.319215057416007e-06, | |
| "loss": 0.1246, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5398406374501992, | |
| "grad_norm": 1.8485098665344575, | |
| "learning_rate": 8.307498084931327e-06, | |
| "loss": 0.1324, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.5418326693227091, | |
| "grad_norm": 2.2839051894864237, | |
| "learning_rate": 8.295748728366414e-06, | |
| "loss": 0.1342, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.5438247011952191, | |
| "grad_norm": 1.6783837554851726, | |
| "learning_rate": 8.283967102760518e-06, | |
| "loss": 0.1192, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.545816733067729, | |
| "grad_norm": 2.0329474847075097, | |
| "learning_rate": 8.272153323468842e-06, | |
| "loss": 0.139, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.547808764940239, | |
| "grad_norm": 1.7759478728678548, | |
| "learning_rate": 8.260307506161407e-06, | |
| "loss": 0.1396, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.549800796812749, | |
| "grad_norm": 1.8489007641542803, | |
| "learning_rate": 8.248429766821925e-06, | |
| "loss": 0.1035, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.5517928286852589, | |
| "grad_norm": 1.3042776152655502, | |
| "learning_rate": 8.236520221746657e-06, | |
| "loss": 0.0979, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.5537848605577689, | |
| "grad_norm": 1.938450288353747, | |
| "learning_rate": 8.22457898754328e-06, | |
| "loss": 0.1355, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.5557768924302788, | |
| "grad_norm": 1.9473763473782129, | |
| "learning_rate": 8.212606181129737e-06, | |
| "loss": 0.1354, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.5577689243027888, | |
| "grad_norm": 2.3529581735987684, | |
| "learning_rate": 8.200601919733106e-06, | |
| "loss": 0.1478, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5597609561752988, | |
| "grad_norm": 2.2045125976242277, | |
| "learning_rate": 8.18856632088844e-06, | |
| "loss": 0.1403, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.5617529880478087, | |
| "grad_norm": 1.8741947740872564, | |
| "learning_rate": 8.176499502437621e-06, | |
| "loss": 0.1224, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.5637450199203188, | |
| "grad_norm": 1.6785392877999905, | |
| "learning_rate": 8.164401582528202e-06, | |
| "loss": 0.1114, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.5657370517928287, | |
| "grad_norm": 1.805219996685288, | |
| "learning_rate": 8.15227267961226e-06, | |
| "loss": 0.1281, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.5677290836653387, | |
| "grad_norm": 2.262380478016552, | |
| "learning_rate": 8.14011291244523e-06, | |
| "loss": 0.1495, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.5697211155378487, | |
| "grad_norm": 1.6184059179144714, | |
| "learning_rate": 8.127922400084736e-06, | |
| "loss": 0.1077, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.5717131474103586, | |
| "grad_norm": 1.7035754466739166, | |
| "learning_rate": 8.115701261889437e-06, | |
| "loss": 0.1081, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.5737051792828686, | |
| "grad_norm": 2.260155985593039, | |
| "learning_rate": 8.10344961751785e-06, | |
| "loss": 0.1367, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.5756972111553785, | |
| "grad_norm": 1.985868190263633, | |
| "learning_rate": 8.091167586927184e-06, | |
| "loss": 0.1184, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.5776892430278885, | |
| "grad_norm": 2.2887926342626144, | |
| "learning_rate": 8.078855290372161e-06, | |
| "loss": 0.1721, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5796812749003984, | |
| "grad_norm": 2.027120648554562, | |
| "learning_rate": 8.066512848403837e-06, | |
| "loss": 0.1342, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.5816733067729084, | |
| "grad_norm": 1.9925359431101095, | |
| "learning_rate": 8.054140381868435e-06, | |
| "loss": 0.1353, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.5836653386454184, | |
| "grad_norm": 2.0044583230657027, | |
| "learning_rate": 8.041738011906144e-06, | |
| "loss": 0.1321, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.5856573705179283, | |
| "grad_norm": 2.1494533777158993, | |
| "learning_rate": 8.02930585994994e-06, | |
| "loss": 0.132, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.5876494023904383, | |
| "grad_norm": 2.1658323964218322, | |
| "learning_rate": 8.016844047724404e-06, | |
| "loss": 0.156, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.5896414342629482, | |
| "grad_norm": 1.928648624837369, | |
| "learning_rate": 8.004352697244516e-06, | |
| "loss": 0.142, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.5916334661354582, | |
| "grad_norm": 1.612340205944657, | |
| "learning_rate": 7.991831930814475e-06, | |
| "loss": 0.1101, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.5936254980079682, | |
| "grad_norm": 1.6508868869854736, | |
| "learning_rate": 7.979281871026493e-06, | |
| "loss": 0.1281, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.5956175298804781, | |
| "grad_norm": 1.8982891507623567, | |
| "learning_rate": 7.966702640759598e-06, | |
| "loss": 0.1321, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.5976095617529881, | |
| "grad_norm": 1.8987961958988515, | |
| "learning_rate": 7.954094363178421e-06, | |
| "loss": 0.1229, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.599601593625498, | |
| "grad_norm": 2.1488376153773725, | |
| "learning_rate": 7.941457161732011e-06, | |
| "loss": 0.1544, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.601593625498008, | |
| "grad_norm": 1.563913063819495, | |
| "learning_rate": 7.928791160152603e-06, | |
| "loss": 0.0999, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.603585657370518, | |
| "grad_norm": 2.113124277431984, | |
| "learning_rate": 7.916096482454425e-06, | |
| "loss": 0.1452, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.6055776892430279, | |
| "grad_norm": 1.4930277954467297, | |
| "learning_rate": 7.903373252932474e-06, | |
| "loss": 0.1182, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.6075697211155379, | |
| "grad_norm": 1.9567309939982196, | |
| "learning_rate": 7.890621596161295e-06, | |
| "loss": 0.1485, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6095617529880478, | |
| "grad_norm": 2.115801942619243, | |
| "learning_rate": 7.877841636993777e-06, | |
| "loss": 0.1343, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.6115537848605578, | |
| "grad_norm": 1.6597909053499225, | |
| "learning_rate": 7.865033500559916e-06, | |
| "loss": 0.1221, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.6135458167330677, | |
| "grad_norm": 2.268140954664932, | |
| "learning_rate": 7.852197312265592e-06, | |
| "loss": 0.1548, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.6155378486055777, | |
| "grad_norm": 1.7438803215243461, | |
| "learning_rate": 7.83933319779135e-06, | |
| "loss": 0.1197, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.6175298804780877, | |
| "grad_norm": 1.7475479034468657, | |
| "learning_rate": 7.826441283091158e-06, | |
| "loss": 0.1205, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6195219123505976, | |
| "grad_norm": 1.917559079622195, | |
| "learning_rate": 7.813521694391183e-06, | |
| "loss": 0.1466, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.6215139442231076, | |
| "grad_norm": 1.9007500389944925, | |
| "learning_rate": 7.800574558188548e-06, | |
| "loss": 0.1297, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.6235059760956175, | |
| "grad_norm": 2.3138697129523624, | |
| "learning_rate": 7.787600001250098e-06, | |
| "loss": 0.1435, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.6254980079681275, | |
| "grad_norm": 1.803331449338951, | |
| "learning_rate": 7.77459815061116e-06, | |
| "loss": 0.1265, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.6274900398406374, | |
| "grad_norm": 1.729785988240638, | |
| "learning_rate": 7.761569133574291e-06, | |
| "loss": 0.1208, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6294820717131474, | |
| "grad_norm": 1.5580779702404448, | |
| "learning_rate": 7.748513077708044e-06, | |
| "loss": 0.1279, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.6314741035856574, | |
| "grad_norm": 1.5557075595562477, | |
| "learning_rate": 7.735430110845707e-06, | |
| "loss": 0.1218, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.6334661354581673, | |
| "grad_norm": 2.190732094534515, | |
| "learning_rate": 7.722320361084057e-06, | |
| "loss": 0.1435, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.6354581673306773, | |
| "grad_norm": 2.066394679381035, | |
| "learning_rate": 7.70918395678211e-06, | |
| "loss": 0.1552, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.6374501992031872, | |
| "grad_norm": 2.386874848230383, | |
| "learning_rate": 7.69602102655985e-06, | |
| "loss": 0.1511, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6394422310756972, | |
| "grad_norm": 1.9135694055276182, | |
| "learning_rate": 7.682831699296991e-06, | |
| "loss": 0.1494, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.6414342629482072, | |
| "grad_norm": 1.9705232701476278, | |
| "learning_rate": 7.669616104131697e-06, | |
| "loss": 0.1452, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.6434262948207171, | |
| "grad_norm": 1.853209913025916, | |
| "learning_rate": 7.656374370459321e-06, | |
| "loss": 0.1336, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.6454183266932271, | |
| "grad_norm": 1.772429733282671, | |
| "learning_rate": 7.643106627931148e-06, | |
| "loss": 0.1131, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.647410358565737, | |
| "grad_norm": 2.331769433940343, | |
| "learning_rate": 7.629813006453114e-06, | |
| "loss": 0.1417, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.649402390438247, | |
| "grad_norm": 2.1257162988182037, | |
| "learning_rate": 7.616493636184538e-06, | |
| "loss": 0.1537, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.651394422310757, | |
| "grad_norm": 1.6032004847610595, | |
| "learning_rate": 7.603148647536853e-06, | |
| "loss": 0.1208, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.6533864541832669, | |
| "grad_norm": 1.9414518467604416, | |
| "learning_rate": 7.5897781711723215e-06, | |
| "loss": 0.1189, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.6553784860557769, | |
| "grad_norm": 1.8035099668212278, | |
| "learning_rate": 7.576382338002759e-06, | |
| "loss": 0.1289, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.6573705179282868, | |
| "grad_norm": 2.5935545240432347, | |
| "learning_rate": 7.56296127918825e-06, | |
| "loss": 0.1589, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.6593625498007968, | |
| "grad_norm": 2.0001976484989887, | |
| "learning_rate": 7.549515126135871e-06, | |
| "loss": 0.1329, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.6613545816733067, | |
| "grad_norm": 2.2072498495488135, | |
| "learning_rate": 7.536044010498396e-06, | |
| "loss": 0.1562, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.6633466135458167, | |
| "grad_norm": 1.942689121900593, | |
| "learning_rate": 7.5225480641730084e-06, | |
| "loss": 0.138, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.6653386454183267, | |
| "grad_norm": 2.025836699652673, | |
| "learning_rate": 7.509027419300017e-06, | |
| "loss": 0.1442, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.6673306772908366, | |
| "grad_norm": 1.75497256540315, | |
| "learning_rate": 7.495482208261554e-06, | |
| "loss": 0.1324, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.6693227091633466, | |
| "grad_norm": 1.730691502621294, | |
| "learning_rate": 7.48191256368028e-06, | |
| "loss": 0.1037, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.6713147410358565, | |
| "grad_norm": 2.1818943000426825, | |
| "learning_rate": 7.468318618418089e-06, | |
| "loss": 0.1197, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.6733067729083665, | |
| "grad_norm": 2.282870874976595, | |
| "learning_rate": 7.454700505574805e-06, | |
| "loss": 0.1651, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.6752988047808764, | |
| "grad_norm": 1.9598715085355076, | |
| "learning_rate": 7.44105835848688e-06, | |
| "loss": 0.13, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.6772908366533864, | |
| "grad_norm": 1.32121304959289, | |
| "learning_rate": 7.427392310726088e-06, | |
| "loss": 0.088, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.6792828685258964, | |
| "grad_norm": 2.0337415726445296, | |
| "learning_rate": 7.413702496098218e-06, | |
| "loss": 0.1425, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.6812749003984063, | |
| "grad_norm": 2.1827633175836008, | |
| "learning_rate": 7.39998904864176e-06, | |
| "loss": 0.1507, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.6832669322709163, | |
| "grad_norm": 2.10204580287355, | |
| "learning_rate": 7.3862521026265986e-06, | |
| "loss": 0.1411, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.6852589641434262, | |
| "grad_norm": 2.1566650379770675, | |
| "learning_rate": 7.372491792552694e-06, | |
| "loss": 0.1427, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.6872509960159362, | |
| "grad_norm": 1.9617078978276397, | |
| "learning_rate": 7.3587082531487675e-06, | |
| "loss": 0.1167, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.6892430278884463, | |
| "grad_norm": 1.912009857403253, | |
| "learning_rate": 7.344901619370977e-06, | |
| "loss": 0.1234, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.6912350597609562, | |
| "grad_norm": 2.127034247138793, | |
| "learning_rate": 7.331072026401611e-06, | |
| "loss": 0.1176, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.6932270916334662, | |
| "grad_norm": 2.2559426016327078, | |
| "learning_rate": 7.31721960964774e-06, | |
| "loss": 0.1343, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.6952191235059761, | |
| "grad_norm": 1.9327539929872006, | |
| "learning_rate": 7.303344504739914e-06, | |
| "loss": 0.1277, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.6972111553784861, | |
| "grad_norm": 1.6413171465166834, | |
| "learning_rate": 7.289446847530822e-06, | |
| "loss": 0.0954, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6992031872509961, | |
| "grad_norm": 1.7965552697782228, | |
| "learning_rate": 7.2755267740939664e-06, | |
| "loss": 0.1182, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.701195219123506, | |
| "grad_norm": 2.264078406151102, | |
| "learning_rate": 7.261584420722328e-06, | |
| "loss": 0.1239, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.703187250996016, | |
| "grad_norm": 2.1024709135826423, | |
| "learning_rate": 7.2476199239270354e-06, | |
| "loss": 0.1252, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.7051792828685259, | |
| "grad_norm": 1.850775717751215, | |
| "learning_rate": 7.2336334204360206e-06, | |
| "loss": 0.096, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.7071713147410359, | |
| "grad_norm": 1.9294667982978444, | |
| "learning_rate": 7.21962504719269e-06, | |
| "loss": 0.1288, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7091633466135459, | |
| "grad_norm": 2.0263901644198645, | |
| "learning_rate": 7.20559494135458e-06, | |
| "loss": 0.1273, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.7111553784860558, | |
| "grad_norm": 2.519289066543413, | |
| "learning_rate": 7.19154324029201e-06, | |
| "loss": 0.1442, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.7131474103585658, | |
| "grad_norm": 2.3400623940562886, | |
| "learning_rate": 7.177470081586743e-06, | |
| "loss": 0.1531, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.7151394422310757, | |
| "grad_norm": 1.913600852089038, | |
| "learning_rate": 7.163375603030634e-06, | |
| "loss": 0.1142, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.7171314741035857, | |
| "grad_norm": 2.4654502175383657, | |
| "learning_rate": 7.149259942624287e-06, | |
| "loss": 0.1498, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7191235059760956, | |
| "grad_norm": 1.7809542538273848, | |
| "learning_rate": 7.135123238575693e-06, | |
| "loss": 0.0992, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.7211155378486056, | |
| "grad_norm": 2.236498970023622, | |
| "learning_rate": 7.120965629298891e-06, | |
| "loss": 0.1231, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.7231075697211156, | |
| "grad_norm": 1.5579483159438388, | |
| "learning_rate": 7.1067872534126004e-06, | |
| "loss": 0.1208, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.7250996015936255, | |
| "grad_norm": 2.623828745113357, | |
| "learning_rate": 7.092588249738871e-06, | |
| "loss": 0.1127, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.7270916334661355, | |
| "grad_norm": 1.926947918523208, | |
| "learning_rate": 7.0783687573017215e-06, | |
| "loss": 0.1114, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7290836653386454, | |
| "grad_norm": 1.5604394877235856, | |
| "learning_rate": 7.064128915325777e-06, | |
| "loss": 0.1116, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.7310756972111554, | |
| "grad_norm": 1.9819642302802614, | |
| "learning_rate": 7.049868863234911e-06, | |
| "loss": 0.1231, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.7330677290836654, | |
| "grad_norm": 2.04013986237716, | |
| "learning_rate": 7.03558874065087e-06, | |
| "loss": 0.1304, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.7350597609561753, | |
| "grad_norm": 1.8528858096630048, | |
| "learning_rate": 7.021288687391917e-06, | |
| "loss": 0.1232, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.7370517928286853, | |
| "grad_norm": 1.931351631379841, | |
| "learning_rate": 7.00696884347146e-06, | |
| "loss": 0.1202, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7390438247011952, | |
| "grad_norm": 2.3674440717388068, | |
| "learning_rate": 6.9926293490966755e-06, | |
| "loss": 0.1578, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.7410358565737052, | |
| "grad_norm": 1.6376485083203747, | |
| "learning_rate": 6.978270344667143e-06, | |
| "loss": 0.0929, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.7430278884462151, | |
| "grad_norm": 1.705119215989148, | |
| "learning_rate": 6.963891970773465e-06, | |
| "loss": 0.1146, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.7450199203187251, | |
| "grad_norm": 1.7904536633328327, | |
| "learning_rate": 6.949494368195896e-06, | |
| "loss": 0.0929, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.7470119521912351, | |
| "grad_norm": 1.7463821286816135, | |
| "learning_rate": 6.935077677902955e-06, | |
| "loss": 0.1146, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.749003984063745, | |
| "grad_norm": 1.8343889129423383, | |
| "learning_rate": 6.920642041050055e-06, | |
| "loss": 0.1231, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.750996015936255, | |
| "grad_norm": 2.112401907355922, | |
| "learning_rate": 6.9061875989781165e-06, | |
| "loss": 0.1323, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.7529880478087649, | |
| "grad_norm": 2.3269817857970456, | |
| "learning_rate": 6.891714493212183e-06, | |
| "loss": 0.1461, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.7549800796812749, | |
| "grad_norm": 2.164617790401808, | |
| "learning_rate": 6.877222865460037e-06, | |
| "loss": 0.1465, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.7569721115537849, | |
| "grad_norm": 1.8394467625888513, | |
| "learning_rate": 6.862712857610812e-06, | |
| "loss": 0.1078, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.7589641434262948, | |
| "grad_norm": 1.7772473532916493, | |
| "learning_rate": 6.848184611733602e-06, | |
| "loss": 0.1173, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.7609561752988048, | |
| "grad_norm": 1.90451509004974, | |
| "learning_rate": 6.833638270076071e-06, | |
| "loss": 0.1164, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.7629482071713147, | |
| "grad_norm": 1.3933704850807993, | |
| "learning_rate": 6.819073975063064e-06, | |
| "loss": 0.0916, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.7649402390438247, | |
| "grad_norm": 2.1083722503136073, | |
| "learning_rate": 6.804491869295207e-06, | |
| "loss": 0.1482, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.7669322709163346, | |
| "grad_norm": 2.1302855766956466, | |
| "learning_rate": 6.789892095547511e-06, | |
| "loss": 0.1255, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.7689243027888446, | |
| "grad_norm": 2.0218204203878747, | |
| "learning_rate": 6.7752747967679825e-06, | |
| "loss": 0.1414, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.7709163346613546, | |
| "grad_norm": 1.6484036328649272, | |
| "learning_rate": 6.7606401160762105e-06, | |
| "loss": 0.1006, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.7729083665338645, | |
| "grad_norm": 1.9790346630572022, | |
| "learning_rate": 6.745988196761976e-06, | |
| "loss": 0.1253, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.7749003984063745, | |
| "grad_norm": 2.4027541617777923, | |
| "learning_rate": 6.731319182283844e-06, | |
| "loss": 0.1516, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.7768924302788844, | |
| "grad_norm": 2.076146090352596, | |
| "learning_rate": 6.71663321626776e-06, | |
| "loss": 0.1324, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.7788844621513944, | |
| "grad_norm": 2.176712178015417, | |
| "learning_rate": 6.7019304425056484e-06, | |
| "loss": 0.1133, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.7808764940239044, | |
| "grad_norm": 2.270343647007193, | |
| "learning_rate": 6.687211004953992e-06, | |
| "loss": 0.1543, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.7828685258964143, | |
| "grad_norm": 2.2157431300656114, | |
| "learning_rate": 6.672475047732436e-06, | |
| "loss": 0.1419, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.7848605577689243, | |
| "grad_norm": 1.8472598515997243, | |
| "learning_rate": 6.657722715122372e-06, | |
| "loss": 0.1171, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.7868525896414342, | |
| "grad_norm": 1.9746467863711765, | |
| "learning_rate": 6.6429541515655215e-06, | |
| "loss": 0.1214, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.7888446215139442, | |
| "grad_norm": 1.838974412736928, | |
| "learning_rate": 6.628169501662527e-06, | |
| "loss": 0.1288, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.7908366533864541, | |
| "grad_norm": 2.315243867151159, | |
| "learning_rate": 6.613368910171533e-06, | |
| "loss": 0.1328, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.7928286852589641, | |
| "grad_norm": 2.081415971304626, | |
| "learning_rate": 6.598552522006772e-06, | |
| "loss": 0.1259, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.7948207171314741, | |
| "grad_norm": 2.3461895895379414, | |
| "learning_rate": 6.583720482237143e-06, | |
| "loss": 0.1449, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.796812749003984, | |
| "grad_norm": 2.2278972284384233, | |
| "learning_rate": 6.568872936084789e-06, | |
| "loss": 0.153, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.796812749003984, | |
| "eval_loss": 0.12241560965776443, | |
| "eval_runtime": 3.1403, | |
| "eval_samples_per_second": 13.056, | |
| "eval_steps_per_second": 3.503, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.798804780876494, | |
| "grad_norm": 1.5055565131366404, | |
| "learning_rate": 6.554010028923682e-06, | |
| "loss": 0.0904, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.8007968127490039, | |
| "grad_norm": 1.9134714393025904, | |
| "learning_rate": 6.539131906278189e-06, | |
| "loss": 0.1221, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.8027888446215139, | |
| "grad_norm": 1.676345510524098, | |
| "learning_rate": 6.524238713821661e-06, | |
| "loss": 0.1049, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.8047808764940239, | |
| "grad_norm": 1.8483113242007039, | |
| "learning_rate": 6.509330597374993e-06, | |
| "loss": 0.1209, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.8067729083665338, | |
| "grad_norm": 2.1209942031252096, | |
| "learning_rate": 6.494407702905207e-06, | |
| "loss": 0.1274, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8087649402390438, | |
| "grad_norm": 2.090229556416619, | |
| "learning_rate": 6.479470176524015e-06, | |
| "loss": 0.1378, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.8107569721115537, | |
| "grad_norm": 1.5536560670439392, | |
| "learning_rate": 6.464518164486395e-06, | |
| "loss": 0.1042, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.8127490039840638, | |
| "grad_norm": 2.136864457352028, | |
| "learning_rate": 6.44955181318915e-06, | |
| "loss": 0.1356, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.8147410358565738, | |
| "grad_norm": 1.6345271519484204, | |
| "learning_rate": 6.434571269169487e-06, | |
| "loss": 0.1103, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.8167330677290837, | |
| "grad_norm": 2.2602024940481877, | |
| "learning_rate": 6.419576679103571e-06, | |
| "loss": 0.1412, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8187250996015937, | |
| "grad_norm": 1.950297407773567, | |
| "learning_rate": 6.404568189805095e-06, | |
| "loss": 0.1177, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.8207171314741036, | |
| "grad_norm": 1.7261505783927313, | |
| "learning_rate": 6.389545948223841e-06, | |
| "loss": 0.1061, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.8227091633466136, | |
| "grad_norm": 1.7921074892509365, | |
| "learning_rate": 6.374510101444242e-06, | |
| "loss": 0.1149, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.8247011952191236, | |
| "grad_norm": 1.693484859639379, | |
| "learning_rate": 6.359460796683937e-06, | |
| "loss": 0.1013, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.8266932270916335, | |
| "grad_norm": 1.8929838681555156, | |
| "learning_rate": 6.344398181292338e-06, | |
| "loss": 0.1117, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8286852589641435, | |
| "grad_norm": 1.9178716343933822, | |
| "learning_rate": 6.329322402749181e-06, | |
| "loss": 0.1133, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.8306772908366534, | |
| "grad_norm": 2.06968861690023, | |
| "learning_rate": 6.314233608663085e-06, | |
| "loss": 0.1347, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.8326693227091634, | |
| "grad_norm": 2.150136253809127, | |
| "learning_rate": 6.299131946770104e-06, | |
| "loss": 0.1102, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.8346613545816733, | |
| "grad_norm": 2.0515375853571105, | |
| "learning_rate": 6.284017564932284e-06, | |
| "loss": 0.1321, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.8366533864541833, | |
| "grad_norm": 1.972608566259217, | |
| "learning_rate": 6.2688906111362115e-06, | |
| "loss": 0.1317, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8386454183266933, | |
| "grad_norm": 2.1835844876053176, | |
| "learning_rate": 6.253751233491565e-06, | |
| "loss": 0.1368, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.8406374501992032, | |
| "grad_norm": 1.9358655395330886, | |
| "learning_rate": 6.238599580229673e-06, | |
| "loss": 0.1242, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.8426294820717132, | |
| "grad_norm": 1.9756294879268275, | |
| "learning_rate": 6.2234357997020475e-06, | |
| "loss": 0.1228, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.8446215139442231, | |
| "grad_norm": 2.0407229239665616, | |
| "learning_rate": 6.208260040378946e-06, | |
| "loss": 0.1069, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.8466135458167331, | |
| "grad_norm": 2.469118035586134, | |
| "learning_rate": 6.193072450847909e-06, | |
| "loss": 0.1353, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.848605577689243, | |
| "grad_norm": 1.3150969469878047, | |
| "learning_rate": 6.1778731798123105e-06, | |
| "loss": 0.0766, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.850597609561753, | |
| "grad_norm": 1.7880637008772997, | |
| "learning_rate": 6.162662376089894e-06, | |
| "loss": 0.1111, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.852589641434263, | |
| "grad_norm": 1.7519177689555412, | |
| "learning_rate": 6.147440188611324e-06, | |
| "loss": 0.1172, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.8545816733067729, | |
| "grad_norm": 1.6790695470314452, | |
| "learning_rate": 6.132206766418728e-06, | |
| "loss": 0.1307, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.8565737051792829, | |
| "grad_norm": 2.329181770786997, | |
| "learning_rate": 6.116962258664228e-06, | |
| "loss": 0.1398, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.8585657370517928, | |
| "grad_norm": 2.1360997891269298, | |
| "learning_rate": 6.10170681460849e-06, | |
| "loss": 0.1348, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.8605577689243028, | |
| "grad_norm": 2.0074048591988745, | |
| "learning_rate": 6.0864405836192575e-06, | |
| "loss": 0.1257, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.8625498007968128, | |
| "grad_norm": 1.9966192443343993, | |
| "learning_rate": 6.071163715169889e-06, | |
| "loss": 0.1233, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.8645418326693227, | |
| "grad_norm": 1.6688986826934915, | |
| "learning_rate": 6.055876358837894e-06, | |
| "loss": 0.1034, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.8665338645418327, | |
| "grad_norm": 2.1720920201705574, | |
| "learning_rate": 6.040578664303476e-06, | |
| "loss": 0.122, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.8685258964143426, | |
| "grad_norm": 2.2349377112258884, | |
| "learning_rate": 6.025270781348055e-06, | |
| "loss": 0.1305, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.8705179282868526, | |
| "grad_norm": 1.9735977301953636, | |
| "learning_rate": 6.009952859852809e-06, | |
| "loss": 0.1209, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.8725099601593626, | |
| "grad_norm": 2.1227111042570446, | |
| "learning_rate": 5.994625049797206e-06, | |
| "loss": 0.1313, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.8745019920318725, | |
| "grad_norm": 1.8425960578334477, | |
| "learning_rate": 5.979287501257531e-06, | |
| "loss": 0.1045, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.8764940239043825, | |
| "grad_norm": 1.9096751566518075, | |
| "learning_rate": 5.963940364405425e-06, | |
| "loss": 0.1094, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.8784860557768924, | |
| "grad_norm": 2.0206284611558396, | |
| "learning_rate": 5.9485837895064e-06, | |
| "loss": 0.1147, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.8804780876494024, | |
| "grad_norm": 1.7978035110634043, | |
| "learning_rate": 5.933217926918386e-06, | |
| "loss": 0.1039, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.8824701195219123, | |
| "grad_norm": 1.3448754139603254, | |
| "learning_rate": 5.9178429270902445e-06, | |
| "loss": 0.0751, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.8844621513944223, | |
| "grad_norm": 1.8906404995180646, | |
| "learning_rate": 5.902458940560304e-06, | |
| "loss": 0.1118, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.8864541832669323, | |
| "grad_norm": 2.248607528732412, | |
| "learning_rate": 5.88706611795488e-06, | |
| "loss": 0.1344, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.8884462151394422, | |
| "grad_norm": 2.0500336844672566, | |
| "learning_rate": 5.871664609986804e-06, | |
| "loss": 0.1172, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.8904382470119522, | |
| "grad_norm": 1.7749390516298367, | |
| "learning_rate": 5.85625456745395e-06, | |
| "loss": 0.096, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.8924302788844621, | |
| "grad_norm": 1.6650651187511873, | |
| "learning_rate": 5.8408361412377475e-06, | |
| "loss": 0.112, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.8944223107569721, | |
| "grad_norm": 1.8698848014452956, | |
| "learning_rate": 5.8254094823017195e-06, | |
| "loss": 0.1094, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.896414342629482, | |
| "grad_norm": 2.31893715626561, | |
| "learning_rate": 5.80997474168999e-06, | |
| "loss": 0.1524, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.898406374501992, | |
| "grad_norm": 1.7326523317597042, | |
| "learning_rate": 5.794532070525817e-06, | |
| "loss": 0.1145, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.900398406374502, | |
| "grad_norm": 1.8551714873826095, | |
| "learning_rate": 5.779081620010104e-06, | |
| "loss": 0.1064, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.9023904382470119, | |
| "grad_norm": 2.000481014536217, | |
| "learning_rate": 5.763623541419925e-06, | |
| "loss": 0.1152, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.9043824701195219, | |
| "grad_norm": 1.514470739576286, | |
| "learning_rate": 5.748157986107038e-06, | |
| "loss": 0.0937, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.9063745019920318, | |
| "grad_norm": 1.7848046901968333, | |
| "learning_rate": 5.73268510549641e-06, | |
| "loss": 0.1182, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9083665338645418, | |
| "grad_norm": 2.123090333270882, | |
| "learning_rate": 5.717205051084731e-06, | |
| "loss": 0.1207, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.9103585657370518, | |
| "grad_norm": 1.9230718298635634, | |
| "learning_rate": 5.7017179744389276e-06, | |
| "loss": 0.1105, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.9123505976095617, | |
| "grad_norm": 2.218085940556976, | |
| "learning_rate": 5.686224027194682e-06, | |
| "loss": 0.1313, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.9143426294820717, | |
| "grad_norm": 1.8057081110378266, | |
| "learning_rate": 5.6707233610549505e-06, | |
| "loss": 0.1248, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.9163346613545816, | |
| "grad_norm": 2.3840484640685715, | |
| "learning_rate": 5.655216127788472e-06, | |
| "loss": 0.1432, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9183266932270916, | |
| "grad_norm": 2.020040126603478, | |
| "learning_rate": 5.639702479228286e-06, | |
| "loss": 0.1113, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.9203187250996016, | |
| "grad_norm": 2.074741713662112, | |
| "learning_rate": 5.6241825672702444e-06, | |
| "loss": 0.1067, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.9223107569721115, | |
| "grad_norm": 2.0160262890682854, | |
| "learning_rate": 5.608656543871524e-06, | |
| "loss": 0.1058, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.9243027888446215, | |
| "grad_norm": 1.576835822191225, | |
| "learning_rate": 5.593124561049141e-06, | |
| "loss": 0.1148, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.9262948207171314, | |
| "grad_norm": 1.7157626959132268, | |
| "learning_rate": 5.57758677087846e-06, | |
| "loss": 0.101, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.9282868525896414, | |
| "grad_norm": 2.0879882110494465, | |
| "learning_rate": 5.5620433254917075e-06, | |
| "loss": 0.1223, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.9302788844621513, | |
| "grad_norm": 1.6905070173375571, | |
| "learning_rate": 5.546494377076478e-06, | |
| "loss": 0.0903, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.9322709163346613, | |
| "grad_norm": 1.8563993306347122, | |
| "learning_rate": 5.530940077874248e-06, | |
| "loss": 0.1065, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.9342629482071713, | |
| "grad_norm": 1.9938558563881068, | |
| "learning_rate": 5.515380580178887e-06, | |
| "loss": 0.1157, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.9362549800796812, | |
| "grad_norm": 1.9502947540268583, | |
| "learning_rate": 5.499816036335157e-06, | |
| "loss": 0.1213, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9382470119521913, | |
| "grad_norm": 2.0454369850534997, | |
| "learning_rate": 5.484246598737234e-06, | |
| "loss": 0.1321, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.9402390438247012, | |
| "grad_norm": 2.215669709261003, | |
| "learning_rate": 5.468672419827208e-06, | |
| "loss": 0.1102, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.9422310756972112, | |
| "grad_norm": 1.9756062583329823, | |
| "learning_rate": 5.453093652093588e-06, | |
| "loss": 0.1232, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.9442231075697212, | |
| "grad_norm": 1.799263076671508, | |
| "learning_rate": 5.437510448069815e-06, | |
| "loss": 0.1081, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.9462151394422311, | |
| "grad_norm": 2.3780944126238475, | |
| "learning_rate": 5.421922960332767e-06, | |
| "loss": 0.1121, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.9482071713147411, | |
| "grad_norm": 1.8280247291319718, | |
| "learning_rate": 5.406331341501264e-06, | |
| "loss": 0.1324, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.950199203187251, | |
| "grad_norm": 2.15631656729653, | |
| "learning_rate": 5.390735744234573e-06, | |
| "loss": 0.1287, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.952191235059761, | |
| "grad_norm": 1.7263537522584287, | |
| "learning_rate": 5.375136321230915e-06, | |
| "loss": 0.086, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.954183266932271, | |
| "grad_norm": 2.3095667041223753, | |
| "learning_rate": 5.359533225225971e-06, | |
| "loss": 0.1374, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.9561752988047809, | |
| "grad_norm": 1.7602753374015587, | |
| "learning_rate": 5.34392660899138e-06, | |
| "loss": 0.1078, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.9581673306772909, | |
| "grad_norm": 2.1582500460632263, | |
| "learning_rate": 5.328316625333251e-06, | |
| "loss": 0.1415, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.9601593625498008, | |
| "grad_norm": 1.8078245176783199, | |
| "learning_rate": 5.312703427090665e-06, | |
| "loss": 0.1202, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.9621513944223108, | |
| "grad_norm": 1.6747521348323762, | |
| "learning_rate": 5.297087167134176e-06, | |
| "loss": 0.1105, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.9641434262948207, | |
| "grad_norm": 2.108324269500925, | |
| "learning_rate": 5.281467998364314e-06, | |
| "loss": 0.1166, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.9661354581673307, | |
| "grad_norm": 1.899301062782763, | |
| "learning_rate": 5.265846073710093e-06, | |
| "loss": 0.1032, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.9681274900398407, | |
| "grad_norm": 1.7986533480735443, | |
| "learning_rate": 5.250221546127508e-06, | |
| "loss": 0.1023, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.9701195219123506, | |
| "grad_norm": 1.7143716184742455, | |
| "learning_rate": 5.2345945685980404e-06, | |
| "loss": 0.1056, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.9721115537848606, | |
| "grad_norm": 1.8989132351788962, | |
| "learning_rate": 5.218965294127155e-06, | |
| "loss": 0.1246, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.9741035856573705, | |
| "grad_norm": 1.692065620829879, | |
| "learning_rate": 5.203333875742814e-06, | |
| "loss": 0.1002, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.9760956175298805, | |
| "grad_norm": 2.1756954006047207, | |
| "learning_rate": 5.187700466493966e-06, | |
| "loss": 0.1316, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.9780876494023905, | |
| "grad_norm": 1.5992575019291566, | |
| "learning_rate": 5.1720652194490504e-06, | |
| "loss": 0.0879, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.9800796812749004, | |
| "grad_norm": 1.8188244998440222, | |
| "learning_rate": 5.156428287694508e-06, | |
| "loss": 0.113, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.9820717131474104, | |
| "grad_norm": 1.8310353091664167, | |
| "learning_rate": 5.140789824333266e-06, | |
| "loss": 0.0925, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.9840637450199203, | |
| "grad_norm": 2.2645161516308674, | |
| "learning_rate": 5.125149982483255e-06, | |
| "loss": 0.1333, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.9860557768924303, | |
| "grad_norm": 2.38898722297588, | |
| "learning_rate": 5.109508915275898e-06, | |
| "loss": 0.1262, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.9880478087649402, | |
| "grad_norm": 1.894617587966868, | |
| "learning_rate": 5.093866775854618e-06, | |
| "loss": 0.0988, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.9900398406374502, | |
| "grad_norm": 1.8643326954112833, | |
| "learning_rate": 5.078223717373334e-06, | |
| "loss": 0.1151, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.9920318725099602, | |
| "grad_norm": 2.381725171456074, | |
| "learning_rate": 5.062579892994966e-06, | |
| "loss": 0.1243, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.9940239043824701, | |
| "grad_norm": 1.7622591670635168, | |
| "learning_rate": 5.046935455889933e-06, | |
| "loss": 0.1052, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.9960159362549801, | |
| "grad_norm": 2.133367063969825, | |
| "learning_rate": 5.03129055923465e-06, | |
| "loss": 0.1098, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.99800796812749, | |
| "grad_norm": 1.800851217571952, | |
| "learning_rate": 5.0156453562100325e-06, | |
| "loss": 0.1187, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.515740716232249, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0727, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.00199203187251, | |
| "grad_norm": 1.454807758568271, | |
| "learning_rate": 4.984354643789968e-06, | |
| "loss": 0.0645, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 1.00398406374502, | |
| "grad_norm": 1.0906220710604444, | |
| "learning_rate": 4.968709440765352e-06, | |
| "loss": 0.0463, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 1.0059760956175299, | |
| "grad_norm": 1.5597035712618907, | |
| "learning_rate": 4.953064544110069e-06, | |
| "loss": 0.0638, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.0079681274900398, | |
| "grad_norm": 1.3271308187362474, | |
| "learning_rate": 4.9374201070050345e-06, | |
| "loss": 0.0564, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 1.0099601593625498, | |
| "grad_norm": 1.6371457735090003, | |
| "learning_rate": 4.9217762826266665e-06, | |
| "loss": 0.0652, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 1.0119521912350598, | |
| "grad_norm": 1.3010690622611256, | |
| "learning_rate": 4.906133224145384e-06, | |
| "loss": 0.0478, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 1.0139442231075697, | |
| "grad_norm": 1.6385417730692136, | |
| "learning_rate": 4.8904910847241025e-06, | |
| "loss": 0.0571, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 1.0159362549800797, | |
| "grad_norm": 1.460130563576597, | |
| "learning_rate": 4.874850017516746e-06, | |
| "loss": 0.0478, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0179282868525896, | |
| "grad_norm": 1.218172055244865, | |
| "learning_rate": 4.8592101756667345e-06, | |
| "loss": 0.0406, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 1.0199203187250996, | |
| "grad_norm": 1.380650876589631, | |
| "learning_rate": 4.843571712305493e-06, | |
| "loss": 0.0488, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 1.0219123505976095, | |
| "grad_norm": 1.4404478206417628, | |
| "learning_rate": 4.82793478055095e-06, | |
| "loss": 0.0423, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 1.0239043824701195, | |
| "grad_norm": 1.4907418600568456, | |
| "learning_rate": 4.8122995335060365e-06, | |
| "loss": 0.0524, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 1.0258964143426295, | |
| "grad_norm": 1.3967454054494113, | |
| "learning_rate": 4.796666124257187e-06, | |
| "loss": 0.0427, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.0278884462151394, | |
| "grad_norm": 1.3774126162481908, | |
| "learning_rate": 4.781034705872846e-06, | |
| "loss": 0.0424, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 1.0298804780876494, | |
| "grad_norm": 1.4995564243577384, | |
| "learning_rate": 4.765405431401961e-06, | |
| "loss": 0.0429, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 1.0318725099601593, | |
| "grad_norm": 1.4776829368859978, | |
| "learning_rate": 4.7497784538724925e-06, | |
| "loss": 0.0439, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 1.0338645418326693, | |
| "grad_norm": 1.5766123608089488, | |
| "learning_rate": 4.7341539262899075e-06, | |
| "loss": 0.047, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 1.0358565737051793, | |
| "grad_norm": 2.010013759539753, | |
| "learning_rate": 4.7185320016356865e-06, | |
| "loss": 0.0495, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.0378486055776892, | |
| "grad_norm": 1.7137673308746173, | |
| "learning_rate": 4.7029128328658255e-06, | |
| "loss": 0.0519, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 1.0398406374501992, | |
| "grad_norm": 1.8846039575121651, | |
| "learning_rate": 4.687296572909336e-06, | |
| "loss": 0.0518, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.0418326693227091, | |
| "grad_norm": 1.517157653051372, | |
| "learning_rate": 4.671683374666751e-06, | |
| "loss": 0.0376, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 1.043824701195219, | |
| "grad_norm": 1.5477788120437967, | |
| "learning_rate": 4.656073391008622e-06, | |
| "loss": 0.0509, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 1.045816733067729, | |
| "grad_norm": 1.5961149461761268, | |
| "learning_rate": 4.64046677477403e-06, | |
| "loss": 0.0406, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.047808764940239, | |
| "grad_norm": 1.7872243106969083, | |
| "learning_rate": 4.624863678769086e-06, | |
| "loss": 0.0484, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 1.049800796812749, | |
| "grad_norm": 2.0305402714348038, | |
| "learning_rate": 4.609264255765429e-06, | |
| "loss": 0.06, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.051792828685259, | |
| "grad_norm": 1.5196304776495169, | |
| "learning_rate": 4.593668658498737e-06, | |
| "loss": 0.0379, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.0537848605577689, | |
| "grad_norm": 1.8563914910463752, | |
| "learning_rate": 4.578077039667235e-06, | |
| "loss": 0.0579, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.0557768924302788, | |
| "grad_norm": 2.0399962964080927, | |
| "learning_rate": 4.562489551930187e-06, | |
| "loss": 0.0554, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.0577689243027888, | |
| "grad_norm": 1.5558523714376704, | |
| "learning_rate": 4.546906347906414e-06, | |
| "loss": 0.0487, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 1.0597609561752988, | |
| "grad_norm": 2.067131962038375, | |
| "learning_rate": 4.531327580172794e-06, | |
| "loss": 0.0616, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 1.0617529880478087, | |
| "grad_norm": 1.8515752338818168, | |
| "learning_rate": 4.515753401262767e-06, | |
| "loss": 0.0476, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.0637450199203187, | |
| "grad_norm": 1.6512814928931692, | |
| "learning_rate": 4.5001839636648456e-06, | |
| "loss": 0.0457, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 1.0657370517928286, | |
| "grad_norm": 1.6309985963797304, | |
| "learning_rate": 4.484619419821116e-06, | |
| "loss": 0.0479, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.0677290836653386, | |
| "grad_norm": 1.3534892174684972, | |
| "learning_rate": 4.469059922125753e-06, | |
| "loss": 0.0351, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 1.0697211155378485, | |
| "grad_norm": 1.5618587225235019, | |
| "learning_rate": 4.453505622923524e-06, | |
| "loss": 0.0514, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 1.0717131474103585, | |
| "grad_norm": 1.8799695188430705, | |
| "learning_rate": 4.437956674508295e-06, | |
| "loss": 0.0564, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 1.0737051792828685, | |
| "grad_norm": 1.6371009327782438, | |
| "learning_rate": 4.422413229121541e-06, | |
| "loss": 0.05, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 1.0756972111553784, | |
| "grad_norm": 1.7149994790253227, | |
| "learning_rate": 4.4068754389508616e-06, | |
| "loss": 0.0526, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.0776892430278884, | |
| "grad_norm": 1.5057791531519986, | |
| "learning_rate": 4.391343456128479e-06, | |
| "loss": 0.0401, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 1.0796812749003983, | |
| "grad_norm": 1.6556428738570654, | |
| "learning_rate": 4.375817432729759e-06, | |
| "loss": 0.0501, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 1.0816733067729083, | |
| "grad_norm": 1.8035332314518027, | |
| "learning_rate": 4.360297520771716e-06, | |
| "loss": 0.0598, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 1.0836653386454183, | |
| "grad_norm": 1.9373725794619105, | |
| "learning_rate": 4.34478387221153e-06, | |
| "loss": 0.0493, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 1.0856573705179282, | |
| "grad_norm": 1.2034769948690562, | |
| "learning_rate": 4.329276638945051e-06, | |
| "loss": 0.0397, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.0876494023904382, | |
| "grad_norm": 1.329498987282073, | |
| "learning_rate": 4.3137759728053206e-06, | |
| "loss": 0.0489, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.0896414342629481, | |
| "grad_norm": 1.3953725126886292, | |
| "learning_rate": 4.298282025561076e-06, | |
| "loss": 0.0478, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 1.091633466135458, | |
| "grad_norm": 1.561219438568973, | |
| "learning_rate": 4.282794948915271e-06, | |
| "loss": 0.0476, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 1.093625498007968, | |
| "grad_norm": 1.7067688119812607, | |
| "learning_rate": 4.267314894503591e-06, | |
| "loss": 0.0413, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 1.095617529880478, | |
| "grad_norm": 1.4480366291730635, | |
| "learning_rate": 4.2518420138929645e-06, | |
| "loss": 0.0452, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.097609561752988, | |
| "grad_norm": 1.8458558711170783, | |
| "learning_rate": 4.2363764585800775e-06, | |
| "loss": 0.0598, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 1.099601593625498, | |
| "grad_norm": 1.60919784354473, | |
| "learning_rate": 4.220918379989898e-06, | |
| "loss": 0.0578, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 1.1015936254980079, | |
| "grad_norm": 1.5149797237579112, | |
| "learning_rate": 4.205467929474186e-06, | |
| "loss": 0.0414, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 1.1035856573705178, | |
| "grad_norm": 1.9355788387540067, | |
| "learning_rate": 4.190025258310013e-06, | |
| "loss": 0.0605, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 1.1055776892430278, | |
| "grad_norm": 1.2712828264468141, | |
| "learning_rate": 4.174590517698284e-06, | |
| "loss": 0.0476, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.1075697211155378, | |
| "grad_norm": 1.7052649700043325, | |
| "learning_rate": 4.159163858762255e-06, | |
| "loss": 0.0448, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 1.1095617529880477, | |
| "grad_norm": 1.4892279816307104, | |
| "learning_rate": 4.143745432546053e-06, | |
| "loss": 0.0454, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 1.1115537848605577, | |
| "grad_norm": 1.2192774756681568, | |
| "learning_rate": 4.1283353900131965e-06, | |
| "loss": 0.0389, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 1.1135458167330676, | |
| "grad_norm": 1.748484313536612, | |
| "learning_rate": 4.112933882045121e-06, | |
| "loss": 0.0527, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 1.1155378486055776, | |
| "grad_norm": 1.5189552368858088, | |
| "learning_rate": 4.097541059439698e-06, | |
| "loss": 0.0538, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.1175298804780875, | |
| "grad_norm": 1.4923045394934673, | |
| "learning_rate": 4.082157072909757e-06, | |
| "loss": 0.0494, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 1.1195219123505975, | |
| "grad_norm": 1.7886352200727111, | |
| "learning_rate": 4.066782073081616e-06, | |
| "loss": 0.0516, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 1.1215139442231075, | |
| "grad_norm": 1.6403729404088143, | |
| "learning_rate": 4.0514162104936025e-06, | |
| "loss": 0.0504, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 1.1235059760956174, | |
| "grad_norm": 1.794559862930424, | |
| "learning_rate": 4.036059635594578e-06, | |
| "loss": 0.0539, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 1.1254980079681274, | |
| "grad_norm": 1.27443198183184, | |
| "learning_rate": 4.020712498742469e-06, | |
| "loss": 0.033, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.1274900398406373, | |
| "grad_norm": 1.528559195651957, | |
| "learning_rate": 4.005374950202795e-06, | |
| "loss": 0.0485, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 1.1294820717131473, | |
| "grad_norm": 1.6682443502200057, | |
| "learning_rate": 3.990047140147192e-06, | |
| "loss": 0.0472, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 1.1314741035856573, | |
| "grad_norm": 1.5977930972583627, | |
| "learning_rate": 3.974729218651946e-06, | |
| "loss": 0.0435, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 1.1334661354581672, | |
| "grad_norm": 1.8138279047378771, | |
| "learning_rate": 3.959421335696524e-06, | |
| "loss": 0.0525, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 1.1354581673306772, | |
| "grad_norm": 1.861112915058331, | |
| "learning_rate": 3.944123641162106e-06, | |
| "loss": 0.0539, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.1374501992031871, | |
| "grad_norm": 1.5212525579613898, | |
| "learning_rate": 3.928836284830113e-06, | |
| "loss": 0.0493, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 1.139442231075697, | |
| "grad_norm": 1.3446527061200657, | |
| "learning_rate": 3.913559416380743e-06, | |
| "loss": 0.0499, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.1414342629482073, | |
| "grad_norm": 1.2939076783357535, | |
| "learning_rate": 3.898293185391509e-06, | |
| "loss": 0.0323, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 1.1434262948207172, | |
| "grad_norm": 1.5273046564926365, | |
| "learning_rate": 3.883037741335772e-06, | |
| "loss": 0.0437, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 1.1454183266932272, | |
| "grad_norm": 1.3539627956837923, | |
| "learning_rate": 3.867793233581272e-06, | |
| "loss": 0.047, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.1474103585657371, | |
| "grad_norm": 1.2884254132839787, | |
| "learning_rate": 3.852559811388676e-06, | |
| "loss": 0.0435, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 1.149402390438247, | |
| "grad_norm": 1.3084774623960684, | |
| "learning_rate": 3.8373376239101076e-06, | |
| "loss": 0.0483, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 1.151394422310757, | |
| "grad_norm": 1.242637903895707, | |
| "learning_rate": 3.822126820187691e-06, | |
| "loss": 0.0313, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 1.153386454183267, | |
| "grad_norm": 1.4593599661421381, | |
| "learning_rate": 3.806927549152091e-06, | |
| "loss": 0.0401, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 1.155378486055777, | |
| "grad_norm": 1.5535605047130172, | |
| "learning_rate": 3.791739959621054e-06, | |
| "loss": 0.0527, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.157370517928287, | |
| "grad_norm": 1.7540657914683675, | |
| "learning_rate": 3.776564200297953e-06, | |
| "loss": 0.0527, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 1.159362549800797, | |
| "grad_norm": 1.57101285496225, | |
| "learning_rate": 3.761400419770328e-06, | |
| "loss": 0.0395, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 1.1613545816733069, | |
| "grad_norm": 1.7555114294741245, | |
| "learning_rate": 3.746248766508435e-06, | |
| "loss": 0.0434, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 1.1633466135458168, | |
| "grad_norm": 2.1592609021335227, | |
| "learning_rate": 3.7311093888637906e-06, | |
| "loss": 0.0495, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 1.1653386454183268, | |
| "grad_norm": 1.2940156736695578, | |
| "learning_rate": 3.7159824350677177e-06, | |
| "loss": 0.0387, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.1673306772908367, | |
| "grad_norm": 1.5202353281570202, | |
| "learning_rate": 3.7008680532298962e-06, | |
| "loss": 0.0432, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 1.1693227091633467, | |
| "grad_norm": 1.4400071402140493, | |
| "learning_rate": 3.685766391336916e-06, | |
| "loss": 0.0463, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 1.1713147410358566, | |
| "grad_norm": 1.7100573375487627, | |
| "learning_rate": 3.670677597250819e-06, | |
| "loss": 0.0511, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 1.1733067729083666, | |
| "grad_norm": 1.68459514966065, | |
| "learning_rate": 3.6556018187076624e-06, | |
| "loss": 0.0478, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 1.1752988047808766, | |
| "grad_norm": 1.6034087639429946, | |
| "learning_rate": 3.6405392033160637e-06, | |
| "loss": 0.04, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.1772908366533865, | |
| "grad_norm": 1.7092195009028075, | |
| "learning_rate": 3.6254898985557598e-06, | |
| "loss": 0.0474, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 1.1792828685258965, | |
| "grad_norm": 1.5626916752981477, | |
| "learning_rate": 3.6104540517761594e-06, | |
| "loss": 0.0551, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.1812749003984064, | |
| "grad_norm": 1.3965756525646038, | |
| "learning_rate": 3.5954318101949047e-06, | |
| "loss": 0.0369, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 1.1832669322709164, | |
| "grad_norm": 1.8211192187451488, | |
| "learning_rate": 3.580423320896429e-06, | |
| "loss": 0.0583, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 1.1852589641434264, | |
| "grad_norm": 1.348709853436774, | |
| "learning_rate": 3.5654287308305137e-06, | |
| "loss": 0.0385, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.1872509960159363, | |
| "grad_norm": 1.5559379215487086, | |
| "learning_rate": 3.55044818681085e-06, | |
| "loss": 0.0456, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 1.1892430278884463, | |
| "grad_norm": 2.0539321664502976, | |
| "learning_rate": 3.5354818355136058e-06, | |
| "loss": 0.0677, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 1.1912350597609562, | |
| "grad_norm": 1.4411394411205194, | |
| "learning_rate": 3.5205298234759854e-06, | |
| "loss": 0.0451, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 1.1932270916334662, | |
| "grad_norm": 1.7774567393386194, | |
| "learning_rate": 3.5055922970947943e-06, | |
| "loss": 0.0492, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 1.1952191235059761, | |
| "grad_norm": 1.5120643301720722, | |
| "learning_rate": 3.4906694026250075e-06, | |
| "loss": 0.0457, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.1952191235059761, | |
| "eval_loss": 0.11691030859947205, | |
| "eval_runtime": 3.1386, | |
| "eval_samples_per_second": 13.063, | |
| "eval_steps_per_second": 3.505, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.197211155378486, | |
| "grad_norm": 1.364354387569068, | |
| "learning_rate": 3.475761286178341e-06, | |
| "loss": 0.0498, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 1.199203187250996, | |
| "grad_norm": 1.3894575596183438, | |
| "learning_rate": 3.460868093721812e-06, | |
| "loss": 0.038, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 1.201195219123506, | |
| "grad_norm": 1.8217847559393374, | |
| "learning_rate": 3.44598997107632e-06, | |
| "loss": 0.0567, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 1.203187250996016, | |
| "grad_norm": 2.1685604089404857, | |
| "learning_rate": 3.431127063915213e-06, | |
| "loss": 0.071, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 1.205179282868526, | |
| "grad_norm": 1.7472889111167405, | |
| "learning_rate": 3.416279517762858e-06, | |
| "loss": 0.0504, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.207171314741036, | |
| "grad_norm": 1.789030781154645, | |
| "learning_rate": 3.4014474779932295e-06, | |
| "loss": 0.0535, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.2091633466135459, | |
| "grad_norm": 1.5379198221444872, | |
| "learning_rate": 3.386631089828468e-06, | |
| "loss": 0.05, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 1.2111553784860558, | |
| "grad_norm": 1.3874042612095352, | |
| "learning_rate": 3.371830498337475e-06, | |
| "loss": 0.0373, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 1.2131474103585658, | |
| "grad_norm": 1.5231435277419703, | |
| "learning_rate": 3.35704584843448e-06, | |
| "loss": 0.037, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 1.2151394422310757, | |
| "grad_norm": 1.6029887156514784, | |
| "learning_rate": 3.342277284877629e-06, | |
| "loss": 0.0461, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.2171314741035857, | |
| "grad_norm": 1.5816632438715723, | |
| "learning_rate": 3.3275249522675656e-06, | |
| "loss": 0.046, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 1.2191235059760956, | |
| "grad_norm": 1.5428519312912603, | |
| "learning_rate": 3.3127889950460094e-06, | |
| "loss": 0.0472, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 1.2211155378486056, | |
| "grad_norm": 1.625742641074798, | |
| "learning_rate": 3.2980695574943532e-06, | |
| "loss": 0.0437, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 1.2231075697211156, | |
| "grad_norm": 1.3751070319547616, | |
| "learning_rate": 3.28336678373224e-06, | |
| "loss": 0.0375, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 1.2250996015936255, | |
| "grad_norm": 1.9737969684729024, | |
| "learning_rate": 3.268680817716158e-06, | |
| "loss": 0.0585, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.2270916334661355, | |
| "grad_norm": 1.9490150643596276, | |
| "learning_rate": 3.254011803238026e-06, | |
| "loss": 0.0504, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 1.2290836653386454, | |
| "grad_norm": 1.332216726824679, | |
| "learning_rate": 3.2393598839237903e-06, | |
| "loss": 0.0342, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 1.2310756972111554, | |
| "grad_norm": 2.1004912830561744, | |
| "learning_rate": 3.22472520323202e-06, | |
| "loss": 0.0488, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.2330677290836654, | |
| "grad_norm": 2.068005000040964, | |
| "learning_rate": 3.2101079044524895e-06, | |
| "loss": 0.065, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 1.2350597609561753, | |
| "grad_norm": 1.2584068242022144, | |
| "learning_rate": 3.195508130704795e-06, | |
| "loss": 0.0378, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.2370517928286853, | |
| "grad_norm": 1.3495383443316602, | |
| "learning_rate": 3.1809260249369373e-06, | |
| "loss": 0.0377, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 1.2390438247011952, | |
| "grad_norm": 1.74209098701691, | |
| "learning_rate": 3.1663617299239303e-06, | |
| "loss": 0.0582, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 1.2410358565737052, | |
| "grad_norm": 2.3541586375494767, | |
| "learning_rate": 3.1518153882663994e-06, | |
| "loss": 0.0658, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 1.2430278884462151, | |
| "grad_norm": 1.5107261868901234, | |
| "learning_rate": 3.1372871423891894e-06, | |
| "loss": 0.0413, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.245019920318725, | |
| "grad_norm": 2.007995750567458, | |
| "learning_rate": 3.1227771345399647e-06, | |
| "loss": 0.0477, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.247011952191235, | |
| "grad_norm": 1.3321234729078801, | |
| "learning_rate": 3.1082855067878182e-06, | |
| "loss": 0.0353, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 1.249003984063745, | |
| "grad_norm": 1.5773251846473662, | |
| "learning_rate": 3.093812401021885e-06, | |
| "loss": 0.0446, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 1.250996015936255, | |
| "grad_norm": 1.7010297581371503, | |
| "learning_rate": 3.079357958949946e-06, | |
| "loss": 0.0489, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 1.252988047808765, | |
| "grad_norm": 1.4948785976133703, | |
| "learning_rate": 3.0649223220970458e-06, | |
| "loss": 0.0346, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 1.254980079681275, | |
| "grad_norm": 1.580233346726061, | |
| "learning_rate": 3.050505631804105e-06, | |
| "loss": 0.0397, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.2569721115537849, | |
| "grad_norm": 1.4427856316940244, | |
| "learning_rate": 3.0361080292265354e-06, | |
| "loss": 0.0462, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 1.2589641434262948, | |
| "grad_norm": 1.5405358822480004, | |
| "learning_rate": 3.021729655332858e-06, | |
| "loss": 0.0446, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 1.2609561752988048, | |
| "grad_norm": 1.4069898266648972, | |
| "learning_rate": 3.0073706509033257e-06, | |
| "loss": 0.0481, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 1.2629482071713147, | |
| "grad_norm": 1.46434016713666, | |
| "learning_rate": 2.993031156528542e-06, | |
| "loss": 0.0405, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 1.2649402390438247, | |
| "grad_norm": 1.535938715676213, | |
| "learning_rate": 2.978711312608084e-06, | |
| "loss": 0.0404, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.2669322709163346, | |
| "grad_norm": 1.6527329484718531, | |
| "learning_rate": 2.9644112593491315e-06, | |
| "loss": 0.0378, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 1.2689243027888446, | |
| "grad_norm": 1.3342944497270683, | |
| "learning_rate": 2.9501311367650908e-06, | |
| "loss": 0.0458, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 1.2709163346613546, | |
| "grad_norm": 2.0538614960525874, | |
| "learning_rate": 2.9358710846742237e-06, | |
| "loss": 0.0536, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 1.2729083665338645, | |
| "grad_norm": 1.905491659479835, | |
| "learning_rate": 2.92163124269828e-06, | |
| "loss": 0.0531, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 1.2749003984063745, | |
| "grad_norm": 1.374868869463802, | |
| "learning_rate": 2.90741175026113e-06, | |
| "loss": 0.0371, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.2768924302788844, | |
| "grad_norm": 1.4720536952560632, | |
| "learning_rate": 2.8932127465874004e-06, | |
| "loss": 0.0444, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 1.2788844621513944, | |
| "grad_norm": 1.756270327258482, | |
| "learning_rate": 2.8790343707011114e-06, | |
| "loss": 0.0508, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 1.2808764940239044, | |
| "grad_norm": 1.1919142011447958, | |
| "learning_rate": 2.864876761424309e-06, | |
| "loss": 0.0361, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 1.2828685258964143, | |
| "grad_norm": 1.392440748957012, | |
| "learning_rate": 2.850740057375716e-06, | |
| "loss": 0.0379, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 1.2848605577689243, | |
| "grad_norm": 1.3950602927829037, | |
| "learning_rate": 2.8366243969693674e-06, | |
| "loss": 0.0338, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.2868525896414342, | |
| "grad_norm": 1.5201761557299762, | |
| "learning_rate": 2.822529918413259e-06, | |
| "loss": 0.0411, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 1.2888446215139442, | |
| "grad_norm": 1.3120599984550645, | |
| "learning_rate": 2.8084567597079915e-06, | |
| "loss": 0.0402, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 1.2908366533864541, | |
| "grad_norm": 1.4249084044129727, | |
| "learning_rate": 2.7944050586454215e-06, | |
| "loss": 0.0416, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.292828685258964, | |
| "grad_norm": 1.7782238234601455, | |
| "learning_rate": 2.7803749528073108e-06, | |
| "loss": 0.0488, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 1.294820717131474, | |
| "grad_norm": 1.6503484254463683, | |
| "learning_rate": 2.7663665795639815e-06, | |
| "loss": 0.0495, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.296812749003984, | |
| "grad_norm": 1.5251619493297126, | |
| "learning_rate": 2.752380076072967e-06, | |
| "loss": 0.0503, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 1.298804780876494, | |
| "grad_norm": 1.3125100617788958, | |
| "learning_rate": 2.7384155792776724e-06, | |
| "loss": 0.0356, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 1.300796812749004, | |
| "grad_norm": 1.5295843374836597, | |
| "learning_rate": 2.7244732259060335e-06, | |
| "loss": 0.044, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 1.302788844621514, | |
| "grad_norm": 1.5090295425620295, | |
| "learning_rate": 2.710553152469178e-06, | |
| "loss": 0.0426, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 1.3047808764940239, | |
| "grad_norm": 1.2597916390063753, | |
| "learning_rate": 2.6966554952600886e-06, | |
| "loss": 0.0359, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.3067729083665338, | |
| "grad_norm": 1.8988044363314096, | |
| "learning_rate": 2.682780390352262e-06, | |
| "loss": 0.0575, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 1.3087649402390438, | |
| "grad_norm": 1.6686947788798556, | |
| "learning_rate": 2.668927973598392e-06, | |
| "loss": 0.0424, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 1.3107569721115537, | |
| "grad_norm": 1.681750415123868, | |
| "learning_rate": 2.655098380629024e-06, | |
| "loss": 0.047, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 1.3127490039840637, | |
| "grad_norm": 1.5736992626946802, | |
| "learning_rate": 2.6412917468512354e-06, | |
| "loss": 0.034, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 1.3147410358565736, | |
| "grad_norm": 1.2994141414494456, | |
| "learning_rate": 2.627508207447308e-06, | |
| "loss": 0.0323, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.3167330677290836, | |
| "grad_norm": 1.5957841782185134, | |
| "learning_rate": 2.613747897373403e-06, | |
| "loss": 0.0537, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 1.3187250996015936, | |
| "grad_norm": 1.62545382658515, | |
| "learning_rate": 2.6000109513582417e-06, | |
| "loss": 0.0523, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 1.3207171314741035, | |
| "grad_norm": 2.230115579239022, | |
| "learning_rate": 2.5862975039017835e-06, | |
| "loss": 0.0607, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 1.3227091633466135, | |
| "grad_norm": 1.405368943781722, | |
| "learning_rate": 2.5726076892739127e-06, | |
| "loss": 0.0448, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 1.3247011952191234, | |
| "grad_norm": 1.7969792652573766, | |
| "learning_rate": 2.5589416415131215e-06, | |
| "loss": 0.0517, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.3266932270916334, | |
| "grad_norm": 1.530151249792159, | |
| "learning_rate": 2.5452994944251962e-06, | |
| "loss": 0.0331, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.3286852589641434, | |
| "grad_norm": 1.4807902414940988, | |
| "learning_rate": 2.531681381581913e-06, | |
| "loss": 0.0378, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 1.3306772908366533, | |
| "grad_norm": 1.7906460874526582, | |
| "learning_rate": 2.5180874363197217e-06, | |
| "loss": 0.0409, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.3326693227091633, | |
| "grad_norm": 1.6166848567176595, | |
| "learning_rate": 2.504517791738449e-06, | |
| "loss": 0.0464, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 1.3346613545816732, | |
| "grad_norm": 1.6676564078427882, | |
| "learning_rate": 2.4909725806999847e-06, | |
| "loss": 0.0466, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.3366533864541832, | |
| "grad_norm": 1.8873962166889124, | |
| "learning_rate": 2.4774519358269932e-06, | |
| "loss": 0.0472, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.3386454183266931, | |
| "grad_norm": 1.498576664515062, | |
| "learning_rate": 2.463955989501607e-06, | |
| "loss": 0.0464, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.340637450199203, | |
| "grad_norm": 1.4936995031717264, | |
| "learning_rate": 2.4504848738641313e-06, | |
| "loss": 0.0319, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 1.342629482071713, | |
| "grad_norm": 1.7166013235605342, | |
| "learning_rate": 2.437038720811752e-06, | |
| "loss": 0.0458, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.3446215139442232, | |
| "grad_norm": 1.9084481577153407, | |
| "learning_rate": 2.4236176619972436e-06, | |
| "loss": 0.0479, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.3466135458167332, | |
| "grad_norm": 1.4070919253542853, | |
| "learning_rate": 2.41022182882768e-06, | |
| "loss": 0.0322, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.3486055776892432, | |
| "grad_norm": 2.0522106017301285, | |
| "learning_rate": 2.3968513524631483e-06, | |
| "loss": 0.0533, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 1.3505976095617531, | |
| "grad_norm": 1.4028748923424417, | |
| "learning_rate": 2.3835063638154636e-06, | |
| "loss": 0.0371, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.352589641434263, | |
| "grad_norm": 1.7132235697547435, | |
| "learning_rate": 2.3701869935468893e-06, | |
| "loss": 0.045, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.354581673306773, | |
| "grad_norm": 1.3468986017248528, | |
| "learning_rate": 2.356893372068855e-06, | |
| "loss": 0.0374, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.356573705179283, | |
| "grad_norm": 1.7023203932118756, | |
| "learning_rate": 2.343625629540681e-06, | |
| "loss": 0.0431, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 1.358565737051793, | |
| "grad_norm": 1.7614144791766995, | |
| "learning_rate": 2.3303838958683077e-06, | |
| "loss": 0.0469, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.360557768924303, | |
| "grad_norm": 1.713466874008201, | |
| "learning_rate": 2.3171683007030117e-06, | |
| "loss": 0.0403, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 1.3625498007968129, | |
| "grad_norm": 1.8072313944179408, | |
| "learning_rate": 2.3039789734401524e-06, | |
| "loss": 0.0514, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.3645418326693228, | |
| "grad_norm": 1.8285876817770546, | |
| "learning_rate": 2.2908160432178937e-06, | |
| "loss": 0.047, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.3665338645418328, | |
| "grad_norm": 1.835704830271512, | |
| "learning_rate": 2.277679638915945e-06, | |
| "loss": 0.0491, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.3685258964143427, | |
| "grad_norm": 1.809690467670344, | |
| "learning_rate": 2.264569889154295e-06, | |
| "loss": 0.0527, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 1.3705179282868527, | |
| "grad_norm": 1.3476668235326381, | |
| "learning_rate": 2.251486922291957e-06, | |
| "loss": 0.0385, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.3725099601593627, | |
| "grad_norm": 2.0206149912840194, | |
| "learning_rate": 2.23843086642571e-06, | |
| "loss": 0.0537, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 1.3745019920318726, | |
| "grad_norm": 1.5720430933164902, | |
| "learning_rate": 2.225401849388842e-06, | |
| "loss": 0.048, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.3764940239043826, | |
| "grad_norm": 1.499247928447303, | |
| "learning_rate": 2.2123999987499015e-06, | |
| "loss": 0.0422, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.3784860557768925, | |
| "grad_norm": 1.276340108431395, | |
| "learning_rate": 2.1994254418114524e-06, | |
| "loss": 0.0423, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.3804780876494025, | |
| "grad_norm": 1.5216579445173712, | |
| "learning_rate": 2.186478305608819e-06, | |
| "loss": 0.0412, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.3824701195219125, | |
| "grad_norm": 1.9598772663646258, | |
| "learning_rate": 2.1735587169088435e-06, | |
| "loss": 0.0477, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.3844621513944224, | |
| "grad_norm": 1.6029428282921836, | |
| "learning_rate": 2.1606668022086517e-06, | |
| "loss": 0.0481, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.3864541832669324, | |
| "grad_norm": 1.0893162038986566, | |
| "learning_rate": 2.147802687734409e-06, | |
| "loss": 0.031, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.3884462151394423, | |
| "grad_norm": 1.48589149356358, | |
| "learning_rate": 2.1349664994400853e-06, | |
| "loss": 0.0382, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 1.3904382470119523, | |
| "grad_norm": 1.8673421468251405, | |
| "learning_rate": 2.122158363006223e-06, | |
| "loss": 0.0605, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.3924302788844622, | |
| "grad_norm": 1.4982653691530576, | |
| "learning_rate": 2.109378403838705e-06, | |
| "loss": 0.0441, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 1.3944223107569722, | |
| "grad_norm": 1.6272971890188588, | |
| "learning_rate": 2.0966267470675273e-06, | |
| "loss": 0.0387, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.3964143426294822, | |
| "grad_norm": 1.3009436508973116, | |
| "learning_rate": 2.0839035175455748e-06, | |
| "loss": 0.0364, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 1.3984063745019921, | |
| "grad_norm": 1.5070361743640346, | |
| "learning_rate": 2.071208839847397e-06, | |
| "loss": 0.0384, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.400398406374502, | |
| "grad_norm": 1.8403315806429654, | |
| "learning_rate": 2.0585428382679894e-06, | |
| "loss": 0.0566, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.402390438247012, | |
| "grad_norm": 1.9768359787018837, | |
| "learning_rate": 2.0459056368215786e-06, | |
| "loss": 0.049, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.404382470119522, | |
| "grad_norm": 1.1994848396934092, | |
| "learning_rate": 2.0332973592404027e-06, | |
| "loss": 0.0357, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.406374501992032, | |
| "grad_norm": 1.4693306467977794, | |
| "learning_rate": 2.0207181289735073e-06, | |
| "loss": 0.0359, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.408366533864542, | |
| "grad_norm": 1.5649069322977223, | |
| "learning_rate": 2.008168069185525e-06, | |
| "loss": 0.042, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.4103585657370519, | |
| "grad_norm": 1.3393789831970144, | |
| "learning_rate": 1.9956473027554846e-06, | |
| "loss": 0.0383, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.4123505976095618, | |
| "grad_norm": 1.7830258693931829, | |
| "learning_rate": 1.9831559522755976e-06, | |
| "loss": 0.0469, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 1.4143426294820718, | |
| "grad_norm": 1.6147368592510511, | |
| "learning_rate": 1.97069414005006e-06, | |
| "loss": 0.0439, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.4163346613545817, | |
| "grad_norm": 1.3999929614334758, | |
| "learning_rate": 1.9582619880938565e-06, | |
| "loss": 0.0463, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 1.4183266932270917, | |
| "grad_norm": 1.836089297310526, | |
| "learning_rate": 1.9458596181315643e-06, | |
| "loss": 0.0573, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.4203187250996017, | |
| "grad_norm": 1.5850159120530725, | |
| "learning_rate": 1.9334871515961616e-06, | |
| "loss": 0.0378, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 1.4223107569721116, | |
| "grad_norm": 1.3946476121954683, | |
| "learning_rate": 1.9211447096278403e-06, | |
| "loss": 0.0405, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.4243027888446216, | |
| "grad_norm": 1.3008967894406476, | |
| "learning_rate": 1.9088324130728164e-06, | |
| "loss": 0.0326, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.4262948207171315, | |
| "grad_norm": 1.3349769350489, | |
| "learning_rate": 1.8965503824821496e-06, | |
| "loss": 0.0429, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.4282868525896415, | |
| "grad_norm": 1.6914951673768632, | |
| "learning_rate": 1.8842987381105626e-06, | |
| "loss": 0.0367, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 1.4302788844621515, | |
| "grad_norm": 1.3515680548078475, | |
| "learning_rate": 1.872077599915263e-06, | |
| "loss": 0.0379, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.4322709163346614, | |
| "grad_norm": 1.3377996405235246, | |
| "learning_rate": 1.8598870875547691e-06, | |
| "loss": 0.0402, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 1.4342629482071714, | |
| "grad_norm": 1.264515354052566, | |
| "learning_rate": 1.84772732038774e-06, | |
| "loss": 0.0323, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.4362549800796813, | |
| "grad_norm": 1.555389355360336, | |
| "learning_rate": 1.8355984174717994e-06, | |
| "loss": 0.0382, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 1.4382470119521913, | |
| "grad_norm": 1.3182929471725358, | |
| "learning_rate": 1.8235004975623816e-06, | |
| "loss": 0.0307, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 1.4402390438247012, | |
| "grad_norm": 1.4783822554476882, | |
| "learning_rate": 1.811433679111561e-06, | |
| "loss": 0.0391, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 1.4422310756972112, | |
| "grad_norm": 1.7708760927848832, | |
| "learning_rate": 1.7993980802668947e-06, | |
| "loss": 0.0391, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.4442231075697212, | |
| "grad_norm": 1.6784698019995254, | |
| "learning_rate": 1.787393818870264e-06, | |
| "loss": 0.0487, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.4462151394422311, | |
| "grad_norm": 1.2911428050631453, | |
| "learning_rate": 1.7754210124567216e-06, | |
| "loss": 0.0325, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 1.448207171314741, | |
| "grad_norm": 1.6758556031327247, | |
| "learning_rate": 1.7634797782533436e-06, | |
| "loss": 0.0378, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 1.450199203187251, | |
| "grad_norm": 1.319567821261113, | |
| "learning_rate": 1.7515702331780753e-06, | |
| "loss": 0.0368, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.452191235059761, | |
| "grad_norm": 1.2921652740567358, | |
| "learning_rate": 1.7396924938385933e-06, | |
| "loss": 0.0331, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 1.454183266932271, | |
| "grad_norm": 1.1429340732722788, | |
| "learning_rate": 1.7278466765311597e-06, | |
| "loss": 0.0336, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.456175298804781, | |
| "grad_norm": 1.8196882924447233, | |
| "learning_rate": 1.7160328972394835e-06, | |
| "loss": 0.0418, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 1.4581673306772909, | |
| "grad_norm": 1.6752103021159248, | |
| "learning_rate": 1.7042512716335873e-06, | |
| "loss": 0.0397, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.4601593625498008, | |
| "grad_norm": 1.797041780271012, | |
| "learning_rate": 1.6925019150686744e-06, | |
| "loss": 0.055, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 1.4621513944223108, | |
| "grad_norm": 1.4651037298048166, | |
| "learning_rate": 1.6807849425839933e-06, | |
| "loss": 0.0361, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 1.4641434262948207, | |
| "grad_norm": 1.1462478664298237, | |
| "learning_rate": 1.669100468901722e-06, | |
| "loss": 0.0308, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.4661354581673307, | |
| "grad_norm": 1.7311028494874292, | |
| "learning_rate": 1.6574486084258369e-06, | |
| "loss": 0.0447, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.4681274900398407, | |
| "grad_norm": 1.7747156927182413, | |
| "learning_rate": 1.6458294752409943e-06, | |
| "loss": 0.0423, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 1.4701195219123506, | |
| "grad_norm": 1.765615529033668, | |
| "learning_rate": 1.6342431831114153e-06, | |
| "loss": 0.0419, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.4721115537848606, | |
| "grad_norm": 1.588016885038511, | |
| "learning_rate": 1.6226898454797697e-06, | |
| "loss": 0.0437, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.4741035856573705, | |
| "grad_norm": 2.3682981845889954, | |
| "learning_rate": 1.6111695754660667e-06, | |
| "loss": 0.0548, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.4760956175298805, | |
| "grad_norm": 2.196815477790404, | |
| "learning_rate": 1.599682485866546e-06, | |
| "loss": 0.0471, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 1.4780876494023905, | |
| "grad_norm": 1.3771418350242575, | |
| "learning_rate": 1.5882286891525755e-06, | |
| "loss": 0.0428, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.4800796812749004, | |
| "grad_norm": 1.4903254185233181, | |
| "learning_rate": 1.5768082974695476e-06, | |
| "loss": 0.0404, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 1.4820717131474104, | |
| "grad_norm": 1.4985158317383274, | |
| "learning_rate": 1.5654214226357822e-06, | |
| "loss": 0.0329, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.4840637450199203, | |
| "grad_norm": 1.7223869396300333, | |
| "learning_rate": 1.5540681761414327e-06, | |
| "loss": 0.047, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.4860557768924303, | |
| "grad_norm": 1.570813419108133, | |
| "learning_rate": 1.5427486691473942e-06, | |
| "loss": 0.0383, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.4880478087649402, | |
| "grad_norm": 1.69930983396748, | |
| "learning_rate": 1.5314630124842144e-06, | |
| "loss": 0.0399, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 1.4900398406374502, | |
| "grad_norm": 1.7421371319863044, | |
| "learning_rate": 1.5202113166510058e-06, | |
| "loss": 0.0481, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.4920318725099602, | |
| "grad_norm": 1.5181018716448114, | |
| "learning_rate": 1.5089936918143705e-06, | |
| "loss": 0.0349, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.4940239043824701, | |
| "grad_norm": 1.5458296783975771, | |
| "learning_rate": 1.4978102478073165e-06, | |
| "loss": 0.0469, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.49601593625498, | |
| "grad_norm": 1.5770259773948998, | |
| "learning_rate": 1.4866610941281823e-06, | |
| "loss": 0.0325, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 1.49800796812749, | |
| "grad_norm": 1.5341462678859672, | |
| "learning_rate": 1.475546339939568e-06, | |
| "loss": 0.0397, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.2027634510806997, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.028, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 1.50199203187251, | |
| "grad_norm": 1.5111370601251886, | |
| "learning_rate": 1.4534204649991817e-06, | |
| "loss": 0.0386, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 1.50398406374502, | |
| "grad_norm": 1.482657437759512, | |
| "learning_rate": 1.4424095608843036e-06, | |
| "loss": 0.036, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.5059760956175299, | |
| "grad_norm": 1.077070456287382, | |
| "learning_rate": 1.4314334895316095e-06, | |
| "loss": 0.0334, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 1.5079681274900398, | |
| "grad_norm": 1.4629517388254307, | |
| "learning_rate": 1.4204923584090314e-06, | |
| "loss": 0.036, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 1.5099601593625498, | |
| "grad_norm": 1.3805538263025956, | |
| "learning_rate": 1.4095862746423961e-06, | |
| "loss": 0.0356, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 1.5119521912350598, | |
| "grad_norm": 1.8358312305710611, | |
| "learning_rate": 1.3987153450143775e-06, | |
| "loss": 0.0465, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 1.5139442231075697, | |
| "grad_norm": 1.1684136646906567, | |
| "learning_rate": 1.3878796759634544e-06, | |
| "loss": 0.0332, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.5159362549800797, | |
| "grad_norm": 1.2688724525634016, | |
| "learning_rate": 1.3770793735828603e-06, | |
| "loss": 0.0346, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 1.5179282868525896, | |
| "grad_norm": 1.8410552958221407, | |
| "learning_rate": 1.366314543619553e-06, | |
| "loss": 0.0464, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 1.5199203187250996, | |
| "grad_norm": 1.6641830537225055, | |
| "learning_rate": 1.355585291473176e-06, | |
| "loss": 0.0402, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 1.5219123505976095, | |
| "grad_norm": 1.4605320398418256, | |
| "learning_rate": 1.3448917221950264e-06, | |
| "loss": 0.0386, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 1.5239043824701195, | |
| "grad_norm": 1.628381572331683, | |
| "learning_rate": 1.3342339404870253e-06, | |
| "loss": 0.0401, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.5258964143426295, | |
| "grad_norm": 2.3659907711059467, | |
| "learning_rate": 1.3236120507006945e-06, | |
| "loss": 0.0553, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 1.5278884462151394, | |
| "grad_norm": 2.019268389843473, | |
| "learning_rate": 1.3130261568361335e-06, | |
| "loss": 0.0456, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 1.5298804780876494, | |
| "grad_norm": 1.903479551165333, | |
| "learning_rate": 1.3024763625410025e-06, | |
| "loss": 0.0452, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 1.5318725099601593, | |
| "grad_norm": 1.6250609527356852, | |
| "learning_rate": 1.2919627711095068e-06, | |
| "loss": 0.0362, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 1.5338645418326693, | |
| "grad_norm": 1.2317333249828415, | |
| "learning_rate": 1.281485485481384e-06, | |
| "loss": 0.0304, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.5358565737051793, | |
| "grad_norm": 1.8649886760609837, | |
| "learning_rate": 1.2710446082408996e-06, | |
| "loss": 0.0464, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 1.5378486055776892, | |
| "grad_norm": 1.8171004964117743, | |
| "learning_rate": 1.2606402416158391e-06, | |
| "loss": 0.0465, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 1.5398406374501992, | |
| "grad_norm": 1.862322448579373, | |
| "learning_rate": 1.2502724874765087e-06, | |
| "loss": 0.0479, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 1.5418326693227091, | |
| "grad_norm": 1.8583353335497257, | |
| "learning_rate": 1.2399414473347405e-06, | |
| "loss": 0.0424, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 1.543824701195219, | |
| "grad_norm": 1.4852985609369138, | |
| "learning_rate": 1.229647222342889e-06, | |
| "loss": 0.037, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.545816733067729, | |
| "grad_norm": 1.701665909805954, | |
| "learning_rate": 1.2193899132928539e-06, | |
| "loss": 0.0435, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 1.547808764940239, | |
| "grad_norm": 1.5368339751180022, | |
| "learning_rate": 1.2091696206150843e-06, | |
| "loss": 0.0405, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 1.549800796812749, | |
| "grad_norm": 1.4211500123116818, | |
| "learning_rate": 1.1989864443775984e-06, | |
| "loss": 0.0355, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 1.551792828685259, | |
| "grad_norm": 2.0587268481909646, | |
| "learning_rate": 1.1888404842850031e-06, | |
| "loss": 0.0555, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 1.5537848605577689, | |
| "grad_norm": 1.2088357311119127, | |
| "learning_rate": 1.1787318396775188e-06, | |
| "loss": 0.0348, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.5557768924302788, | |
| "grad_norm": 1.4388643186761485, | |
| "learning_rate": 1.1686606095300034e-06, | |
| "loss": 0.0344, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 1.5577689243027888, | |
| "grad_norm": 1.2409791290088654, | |
| "learning_rate": 1.158626892450988e-06, | |
| "loss": 0.0344, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 1.5597609561752988, | |
| "grad_norm": 1.6545764811493544, | |
| "learning_rate": 1.1486307866817082e-06, | |
| "loss": 0.0455, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 1.5617529880478087, | |
| "grad_norm": 1.8070275154016138, | |
| "learning_rate": 1.138672390095143e-06, | |
| "loss": 0.033, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 1.5637450199203187, | |
| "grad_norm": 1.6600852203195762, | |
| "learning_rate": 1.128751800195057e-06, | |
| "loss": 0.0467, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.5657370517928286, | |
| "grad_norm": 1.4417059533847947, | |
| "learning_rate": 1.1188691141150455e-06, | |
| "loss": 0.0383, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 1.5677290836653386, | |
| "grad_norm": 1.2631911364466095, | |
| "learning_rate": 1.1090244286175834e-06, | |
| "loss": 0.0262, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 1.5697211155378485, | |
| "grad_norm": 1.7996066106500415, | |
| "learning_rate": 1.0992178400930753e-06, | |
| "loss": 0.0428, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 1.5717131474103585, | |
| "grad_norm": 1.661450747231149, | |
| "learning_rate": 1.0894494445589171e-06, | |
| "loss": 0.0475, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 1.5737051792828685, | |
| "grad_norm": 1.9431205504964344, | |
| "learning_rate": 1.0797193376585518e-06, | |
| "loss": 0.0451, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.5756972111553784, | |
| "grad_norm": 1.3821996487700146, | |
| "learning_rate": 1.0700276146605349e-06, | |
| "loss": 0.0385, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 1.5776892430278884, | |
| "grad_norm": 1.4713583961609482, | |
| "learning_rate": 1.0603743704575992e-06, | |
| "loss": 0.0402, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 1.5796812749003983, | |
| "grad_norm": 1.396562745448481, | |
| "learning_rate": 1.0507596995657288e-06, | |
| "loss": 0.0407, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 1.5816733067729083, | |
| "grad_norm": 1.584255914198357, | |
| "learning_rate": 1.0411836961232312e-06, | |
| "loss": 0.0362, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 1.5836653386454183, | |
| "grad_norm": 1.2525048404082542, | |
| "learning_rate": 1.031646453889818e-06, | |
| "loss": 0.0375, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.5856573705179282, | |
| "grad_norm": 1.692525265149603, | |
| "learning_rate": 1.0221480662456845e-06, | |
| "loss": 0.0451, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 1.5876494023904382, | |
| "grad_norm": 1.5918519161268234, | |
| "learning_rate": 1.012688626190596e-06, | |
| "loss": 0.0479, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 1.5896414342629481, | |
| "grad_norm": 1.4243746693979769, | |
| "learning_rate": 1.0032682263429788e-06, | |
| "loss": 0.0331, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 1.591633466135458, | |
| "grad_norm": 1.926086405446464, | |
| "learning_rate": 9.93886958939011e-07, | |
| "loss": 0.0427, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 1.593625498007968, | |
| "grad_norm": 1.4477256907812834, | |
| "learning_rate": 9.845449158317216e-07, | |
| "loss": 0.0359, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.593625498007968, | |
| "eval_loss": 0.10596468299627304, | |
| "eval_runtime": 3.1343, | |
| "eval_samples_per_second": 13.081, | |
| "eval_steps_per_second": 3.51, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.595617529880478, | |
| "grad_norm": 1.421779703061797, | |
| "learning_rate": 9.752421884900915e-07, | |
| "loss": 0.0305, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 1.597609561752988, | |
| "grad_norm": 1.7789603558111702, | |
| "learning_rate": 9.65978867998152e-07, | |
| "loss": 0.0345, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 1.599601593625498, | |
| "grad_norm": 1.9525946533919227, | |
| "learning_rate": 9.567550450541012e-07, | |
| "loss": 0.0536, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 1.6015936254980079, | |
| "grad_norm": 1.4805860459441311, | |
| "learning_rate": 9.475708099694125e-07, | |
| "loss": 0.0425, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 1.6035856573705178, | |
| "grad_norm": 1.6320592012302073, | |
| "learning_rate": 9.384262526679488e-07, | |
| "loss": 0.0385, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.6055776892430278, | |
| "grad_norm": 1.407033676035743, | |
| "learning_rate": 9.293214626850838e-07, | |
| "loss": 0.0398, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 1.6075697211155378, | |
| "grad_norm": 1.1535555756252225, | |
| "learning_rate": 9.202565291668253e-07, | |
| "loss": 0.0314, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 1.6095617529880477, | |
| "grad_norm": 1.4977437784858982, | |
| "learning_rate": 9.112315408689415e-07, | |
| "loss": 0.0374, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 1.6115537848605577, | |
| "grad_norm": 1.732572156525876, | |
| "learning_rate": 9.022465861560931e-07, | |
| "loss": 0.0446, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 1.6135458167330676, | |
| "grad_norm": 1.6891276330650864, | |
| "learning_rate": 8.933017530009669e-07, | |
| "loss": 0.0383, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.6155378486055776, | |
| "grad_norm": 1.897852297890652, | |
| "learning_rate": 8.843971289834157e-07, | |
| "loss": 0.0441, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 1.6175298804780875, | |
| "grad_norm": 1.5594431631506627, | |
| "learning_rate": 8.755328012896002e-07, | |
| "loss": 0.0405, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 1.6195219123505975, | |
| "grad_norm": 1.436757522414069, | |
| "learning_rate": 8.667088567111348e-07, | |
| "loss": 0.0396, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 1.6215139442231075, | |
| "grad_norm": 1.9484255207208512, | |
| "learning_rate": 8.579253816442401e-07, | |
| "loss": 0.0516, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 1.6235059760956174, | |
| "grad_norm": 1.380709884926527, | |
| "learning_rate": 8.491824620888906e-07, | |
| "loss": 0.0356, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.6254980079681274, | |
| "grad_norm": 1.7133470278748053, | |
| "learning_rate": 8.404801836479809e-07, | |
| "loss": 0.0434, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 1.6274900398406373, | |
| "grad_norm": 1.583809883914204, | |
| "learning_rate": 8.318186315264859e-07, | |
| "loss": 0.0394, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 1.6294820717131473, | |
| "grad_norm": 1.468174216651124, | |
| "learning_rate": 8.231978905306204e-07, | |
| "loss": 0.0268, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 1.6314741035856573, | |
| "grad_norm": 1.4408279947153464, | |
| "learning_rate": 8.146180450670155e-07, | |
| "loss": 0.0313, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 1.6334661354581672, | |
| "grad_norm": 1.5748697433294543, | |
| "learning_rate": 8.060791791418887e-07, | |
| "loss": 0.0363, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.6354581673306772, | |
| "grad_norm": 1.6393304512543407, | |
| "learning_rate": 7.975813763602219e-07, | |
| "loss": 0.0325, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 1.6374501992031871, | |
| "grad_norm": 2.1412321953511935, | |
| "learning_rate": 7.891247199249441e-07, | |
| "loss": 0.0532, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 1.639442231075697, | |
| "grad_norm": 1.3138050863993294, | |
| "learning_rate": 7.807092926361154e-07, | |
| "loss": 0.0331, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 1.641434262948207, | |
| "grad_norm": 1.5541240912252534, | |
| "learning_rate": 7.723351768901172e-07, | |
| "loss": 0.0426, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 1.643426294820717, | |
| "grad_norm": 1.406024578499524, | |
| "learning_rate": 7.640024546788449e-07, | |
| "loss": 0.0342, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.645418326693227, | |
| "grad_norm": 1.2539556880610443, | |
| "learning_rate": 7.557112075889034e-07, | |
| "loss": 0.026, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 1.647410358565737, | |
| "grad_norm": 1.394792639284168, | |
| "learning_rate": 7.474615168008126e-07, | |
| "loss": 0.0437, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 1.6494023904382469, | |
| "grad_norm": 1.2380166612525048, | |
| "learning_rate": 7.392534630882092e-07, | |
| "loss": 0.0308, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 1.6513944223107568, | |
| "grad_norm": 1.6208908227480252, | |
| "learning_rate": 7.310871268170566e-07, | |
| "loss": 0.0391, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 1.6533864541832668, | |
| "grad_norm": 1.3790210574012849, | |
| "learning_rate": 7.229625879448577e-07, | |
| "loss": 0.0341, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.6553784860557768, | |
| "grad_norm": 1.3562415289883447, | |
| "learning_rate": 7.148799260198736e-07, | |
| "loss": 0.0323, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 1.6573705179282867, | |
| "grad_norm": 1.4745348978618424, | |
| "learning_rate": 7.06839220180342e-07, | |
| "loss": 0.0394, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 1.6593625498007967, | |
| "grad_norm": 1.8763390835724343, | |
| "learning_rate": 6.988405491537054e-07, | |
| "loss": 0.0403, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 1.6613545816733066, | |
| "grad_norm": 1.5395306915776477, | |
| "learning_rate": 6.908839912558374e-07, | |
| "loss": 0.0375, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 1.6633466135458166, | |
| "grad_norm": 1.6645400590228017, | |
| "learning_rate": 6.829696243902784e-07, | |
| "loss": 0.0448, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.6653386454183265, | |
| "grad_norm": 1.8788295421219408, | |
| "learning_rate": 6.750975260474718e-07, | |
| "loss": 0.0425, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 1.6673306772908365, | |
| "grad_norm": 1.9902864466850982, | |
| "learning_rate": 6.67267773304004e-07, | |
| "loss": 0.048, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 1.6693227091633465, | |
| "grad_norm": 1.0800153432067903, | |
| "learning_rate": 6.594804428218527e-07, | |
| "loss": 0.0327, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 1.6713147410358564, | |
| "grad_norm": 1.500062093936282, | |
| "learning_rate": 6.517356108476314e-07, | |
| "loss": 0.0366, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 1.6733067729083664, | |
| "grad_norm": 1.9099325765908919, | |
| "learning_rate": 6.440333532118503e-07, | |
| "loss": 0.0432, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.6752988047808763, | |
| "grad_norm": 1.16559644327572, | |
| "learning_rate": 6.36373745328166e-07, | |
| "loss": 0.0258, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 1.6772908366533863, | |
| "grad_norm": 1.3024121385631138, | |
| "learning_rate": 6.287568621926482e-07, | |
| "loss": 0.0329, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 1.6792828685258963, | |
| "grad_norm": 1.7905458621338421, | |
| "learning_rate": 6.211827783830443e-07, | |
| "loss": 0.0388, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 1.6812749003984062, | |
| "grad_norm": 1.507626826133329, | |
| "learning_rate": 6.136515680580479e-07, | |
| "loss": 0.0365, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 1.6832669322709162, | |
| "grad_norm": 1.5818672866222618, | |
| "learning_rate": 6.061633049565735e-07, | |
| "loss": 0.0427, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.6852589641434261, | |
| "grad_norm": 1.7500474110140714, | |
| "learning_rate": 5.987180623970351e-07, | |
| "loss": 0.0419, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 1.687250996015936, | |
| "grad_norm": 1.6005728580669865, | |
| "learning_rate": 5.913159132766272e-07, | |
| "loss": 0.0386, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 1.6892430278884463, | |
| "grad_norm": 1.4765492940539946, | |
| "learning_rate": 5.839569300706127e-07, | |
| "loss": 0.035, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 1.6912350597609562, | |
| "grad_norm": 1.7073641068426852, | |
| "learning_rate": 5.766411848316111e-07, | |
| "loss": 0.0453, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 1.6932270916334662, | |
| "grad_norm": 1.3266813195045528, | |
| "learning_rate": 5.693687491888944e-07, | |
| "loss": 0.0269, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.6952191235059761, | |
| "grad_norm": 1.7375537306550803, | |
| "learning_rate": 5.621396943476865e-07, | |
| "loss": 0.0371, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 1.697211155378486, | |
| "grad_norm": 1.16829582732014, | |
| "learning_rate": 5.549540910884649e-07, | |
| "loss": 0.0292, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 1.699203187250996, | |
| "grad_norm": 1.2524818853083746, | |
| "learning_rate": 5.478120097662654e-07, | |
| "loss": 0.0379, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 1.701195219123506, | |
| "grad_norm": 1.6095889326273223, | |
| "learning_rate": 5.407135203099984e-07, | |
| "loss": 0.0348, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 1.703187250996016, | |
| "grad_norm": 1.7855589434412364, | |
| "learning_rate": 5.336586922217607e-07, | |
| "loss": 0.0426, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.705179282868526, | |
| "grad_norm": 1.3350483084942124, | |
| "learning_rate": 5.266475945761562e-07, | |
| "loss": 0.0324, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 1.707171314741036, | |
| "grad_norm": 1.9650070108762112, | |
| "learning_rate": 5.19680296019619e-07, | |
| "loss": 0.0422, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 1.7091633466135459, | |
| "grad_norm": 1.5395346984385803, | |
| "learning_rate": 5.127568647697407e-07, | |
| "loss": 0.0389, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 1.7111553784860558, | |
| "grad_norm": 1.4407452070494002, | |
| "learning_rate": 5.05877368614604e-07, | |
| "loss": 0.0336, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 1.7131474103585658, | |
| "grad_norm": 1.7638225778002696, | |
| "learning_rate": 4.990418749121179e-07, | |
| "loss": 0.0386, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.7151394422310757, | |
| "grad_norm": 1.3281767084541467, | |
| "learning_rate": 4.922504505893583e-07, | |
| "loss": 0.0292, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 1.7171314741035857, | |
| "grad_norm": 1.626746762497501, | |
| "learning_rate": 4.855031621419143e-07, | |
| "loss": 0.033, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 1.7191235059760956, | |
| "grad_norm": 1.432244514359786, | |
| "learning_rate": 4.788000756332339e-07, | |
| "loss": 0.0342, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 1.7211155378486056, | |
| "grad_norm": 1.355048293376289, | |
| "learning_rate": 4.721412566939804e-07, | |
| "loss": 0.0339, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 1.7231075697211156, | |
| "grad_norm": 1.8150166104991199, | |
| "learning_rate": 4.655267705213884e-07, | |
| "loss": 0.0332, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.7250996015936255, | |
| "grad_norm": 1.3905664903519097, | |
| "learning_rate": 4.5895668187862283e-07, | |
| "loss": 0.0355, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 1.7270916334661355, | |
| "grad_norm": 1.4892701437662028, | |
| "learning_rate": 4.524310550941513e-07, | |
| "loss": 0.0389, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 1.7290836653386454, | |
| "grad_norm": 1.4991400329744464, | |
| "learning_rate": 4.4594995406110785e-07, | |
| "loss": 0.0288, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 1.7310756972111554, | |
| "grad_norm": 2.1763244133944717, | |
| "learning_rate": 4.395134422366715e-07, | |
| "loss": 0.0597, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 1.7330677290836654, | |
| "grad_norm": 1.4364589154779435, | |
| "learning_rate": 4.33121582641442e-07, | |
| "loss": 0.029, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.7350597609561753, | |
| "grad_norm": 1.4528953350538367, | |
| "learning_rate": 4.2677443785882566e-07, | |
| "loss": 0.0299, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 1.7370517928286853, | |
| "grad_norm": 1.9612964941834787, | |
| "learning_rate": 4.2047207003442003e-07, | |
| "loss": 0.0395, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 1.7390438247011952, | |
| "grad_norm": 2.167595941204298, | |
| "learning_rate": 4.142145408754061e-07, | |
| "loss": 0.0443, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 1.7410358565737052, | |
| "grad_norm": 1.91106994896413, | |
| "learning_rate": 4.0800191164994675e-07, | |
| "loss": 0.0389, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 1.7430278884462151, | |
| "grad_norm": 1.4525532274137143, | |
| "learning_rate": 4.018342431865818e-07, | |
| "loss": 0.032, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.745019920318725, | |
| "grad_norm": 1.6060381855232735, | |
| "learning_rate": 3.957115958736374e-07, | |
| "loss": 0.0425, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 1.747011952191235, | |
| "grad_norm": 1.3887626940467759, | |
| "learning_rate": 3.8963402965863094e-07, | |
| "loss": 0.0397, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 1.749003984063745, | |
| "grad_norm": 1.372684240605827, | |
| "learning_rate": 3.8360160404768755e-07, | |
| "loss": 0.0391, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 1.750996015936255, | |
| "grad_norm": 1.3181667404064525, | |
| "learning_rate": 3.7761437810495517e-07, | |
| "loss": 0.0335, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 1.752988047808765, | |
| "grad_norm": 1.6789671311971048, | |
| "learning_rate": 3.7167241045202474e-07, | |
| "loss": 0.0407, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.754980079681275, | |
| "grad_norm": 1.3469532000945257, | |
| "learning_rate": 3.657757592673611e-07, | |
| "loss": 0.0274, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 1.7569721115537849, | |
| "grad_norm": 1.4431142582457726, | |
| "learning_rate": 3.5992448228572895e-07, | |
| "loss": 0.0276, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 1.7589641434262948, | |
| "grad_norm": 1.2459499701396597, | |
| "learning_rate": 3.5411863679762956e-07, | |
| "loss": 0.0313, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 1.7609561752988048, | |
| "grad_norm": 1.8366872404033463, | |
| "learning_rate": 3.483582796487395e-07, | |
| "loss": 0.0434, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 1.7629482071713147, | |
| "grad_norm": 1.6841292031312827, | |
| "learning_rate": 3.426434672393542e-07, | |
| "loss": 0.0375, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.7649402390438247, | |
| "grad_norm": 1.696643515004684, | |
| "learning_rate": 3.3697425552383536e-07, | |
| "loss": 0.0421, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 1.7669322709163346, | |
| "grad_norm": 1.316459868698009, | |
| "learning_rate": 3.3135070001006186e-07, | |
| "loss": 0.033, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 1.7689243027888446, | |
| "grad_norm": 1.9166835989066684, | |
| "learning_rate": 3.257728557588902e-07, | |
| "loss": 0.0438, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 1.7709163346613546, | |
| "grad_norm": 1.6334683496335292, | |
| "learning_rate": 3.202407773836108e-07, | |
| "loss": 0.0356, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 1.7729083665338645, | |
| "grad_norm": 1.7333830974632702, | |
| "learning_rate": 3.1475451904941613e-07, | |
| "loss": 0.0392, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.7749003984063745, | |
| "grad_norm": 1.6486004340046463, | |
| "learning_rate": 3.093141344728695e-07, | |
| "loss": 0.0429, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 1.7768924302788844, | |
| "grad_norm": 1.7192233169908389, | |
| "learning_rate": 3.039196769213787e-07, | |
| "loss": 0.0334, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.7788844621513944, | |
| "grad_norm": 1.4986249160852925, | |
| "learning_rate": 2.985711992126772e-07, | |
| "loss": 0.0332, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 1.7808764940239044, | |
| "grad_norm": 1.391955958968789, | |
| "learning_rate": 2.932687537143003e-07, | |
| "loss": 0.0395, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 1.7828685258964143, | |
| "grad_norm": 1.5022523308748341, | |
| "learning_rate": 2.8801239234308e-07, | |
| "loss": 0.0328, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.7848605577689243, | |
| "grad_norm": 1.5896216675986345, | |
| "learning_rate": 2.828021665646341e-07, | |
| "loss": 0.0395, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 1.7868525896414342, | |
| "grad_norm": 1.5943964105881356, | |
| "learning_rate": 2.776381273928597e-07, | |
| "loss": 0.0335, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 1.7888446215139442, | |
| "grad_norm": 1.4573130385904005, | |
| "learning_rate": 2.725203253894365e-07, | |
| "loss": 0.0328, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 1.7908366533864541, | |
| "grad_norm": 1.5957847319376306, | |
| "learning_rate": 2.6744881066333104e-07, | |
| "loss": 0.0302, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 1.792828685258964, | |
| "grad_norm": 1.5938499333127716, | |
| "learning_rate": 2.6242363287030617e-07, | |
| "loss": 0.0334, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.794820717131474, | |
| "grad_norm": 1.493471769426448, | |
| "learning_rate": 2.5744484121243416e-07, | |
| "loss": 0.0376, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 1.796812749003984, | |
| "grad_norm": 1.4679887246269543, | |
| "learning_rate": 2.5251248443761644e-07, | |
| "loss": 0.0308, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 1.798804780876494, | |
| "grad_norm": 1.4811334241506124, | |
| "learning_rate": 2.47626610839104e-07, | |
| "loss": 0.0461, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 1.800796812749004, | |
| "grad_norm": 1.4378347029531395, | |
| "learning_rate": 2.4278726825502696e-07, | |
| "loss": 0.0308, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 1.802788844621514, | |
| "grad_norm": 1.3874168126207334, | |
| "learning_rate": 2.3799450406792435e-07, | |
| "loss": 0.0383, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.8047808764940239, | |
| "grad_norm": 1.309299013582529, | |
| "learning_rate": 2.3324836520428275e-07, | |
| "loss": 0.0313, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 1.8067729083665338, | |
| "grad_norm": 1.724858228945609, | |
| "learning_rate": 2.285488981340722e-07, | |
| "loss": 0.0355, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 1.8087649402390438, | |
| "grad_norm": 1.696250952718899, | |
| "learning_rate": 2.2389614887029564e-07, | |
| "loss": 0.0314, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 1.8107569721115537, | |
| "grad_norm": 1.4712298958031025, | |
| "learning_rate": 2.1929016296853679e-07, | |
| "loss": 0.0371, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 1.812749003984064, | |
| "grad_norm": 1.2608615587029985, | |
| "learning_rate": 2.147309855265145e-07, | |
| "loss": 0.0313, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.8147410358565739, | |
| "grad_norm": 1.6486609933665899, | |
| "learning_rate": 2.1021866118363987e-07, | |
| "loss": 0.0372, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 1.8167330677290838, | |
| "grad_norm": 1.456648149760954, | |
| "learning_rate": 2.0575323412058036e-07, | |
| "loss": 0.036, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 1.8187250996015938, | |
| "grad_norm": 1.6643795113487712, | |
| "learning_rate": 2.0133474805882735e-07, | |
| "loss": 0.0357, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 1.8207171314741037, | |
| "grad_norm": 1.4592297656846365, | |
| "learning_rate": 1.9696324626026774e-07, | |
| "loss": 0.0343, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 1.8227091633466137, | |
| "grad_norm": 1.5024913962281268, | |
| "learning_rate": 1.926387715267597e-07, | |
| "loss": 0.0332, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.8247011952191237, | |
| "grad_norm": 1.5222331132128035, | |
| "learning_rate": 1.8836136619971468e-07, | |
| "loss": 0.0274, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 1.8266932270916336, | |
| "grad_norm": 1.470688430167571, | |
| "learning_rate": 1.8413107215968174e-07, | |
| "loss": 0.0358, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 1.8286852589641436, | |
| "grad_norm": 1.2370854006861078, | |
| "learning_rate": 1.7994793082593942e-07, | |
| "loss": 0.0294, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 1.8306772908366535, | |
| "grad_norm": 1.479773183899037, | |
| "learning_rate": 1.7581198315608727e-07, | |
| "loss": 0.0404, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 1.8326693227091635, | |
| "grad_norm": 1.9167324471302913, | |
| "learning_rate": 1.7172326964564777e-07, | |
| "loss": 0.0444, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.8346613545816735, | |
| "grad_norm": 1.4329590983432605, | |
| "learning_rate": 1.6768183032766728e-07, | |
| "loss": 0.0372, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 1.8366533864541834, | |
| "grad_norm": 1.2780231630551226, | |
| "learning_rate": 1.6368770477232622e-07, | |
| "loss": 0.0328, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 1.8386454183266934, | |
| "grad_norm": 1.4118812529232878, | |
| "learning_rate": 1.597409320865506e-07, | |
| "loss": 0.0415, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 1.8406374501992033, | |
| "grad_norm": 1.836114736553896, | |
| "learning_rate": 1.5584155091362907e-07, | |
| "loss": 0.0489, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 1.8426294820717133, | |
| "grad_norm": 1.3679200385396149, | |
| "learning_rate": 1.5198959943283466e-07, | |
| "loss": 0.0323, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.8446215139442232, | |
| "grad_norm": 1.5872343873759975, | |
| "learning_rate": 1.4818511535905077e-07, | |
| "loss": 0.0395, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 1.8466135458167332, | |
| "grad_norm": 1.4421140886448538, | |
| "learning_rate": 1.444281359424038e-07, | |
| "loss": 0.0315, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 1.8486055776892432, | |
| "grad_norm": 1.7604548086333676, | |
| "learning_rate": 1.4071869796789427e-07, | |
| "loss": 0.05, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 1.8505976095617531, | |
| "grad_norm": 1.4556017520765743, | |
| "learning_rate": 1.3705683775504075e-07, | |
| "loss": 0.0281, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 1.852589641434263, | |
| "grad_norm": 1.3271732449749567, | |
| "learning_rate": 1.3344259115752268e-07, | |
| "loss": 0.0324, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.854581673306773, | |
| "grad_norm": 1.2761857510150398, | |
| "learning_rate": 1.2987599356282853e-07, | |
| "loss": 0.0391, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 1.856573705179283, | |
| "grad_norm": 1.5833152792310408, | |
| "learning_rate": 1.263570798919106e-07, | |
| "loss": 0.0339, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 1.858565737051793, | |
| "grad_norm": 1.2639695829653441, | |
| "learning_rate": 1.2288588459884344e-07, | |
| "loss": 0.0345, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 1.860557768924303, | |
| "grad_norm": 1.3658369812613829, | |
| "learning_rate": 1.1946244167048314e-07, | |
| "loss": 0.0315, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 1.8625498007968129, | |
| "grad_norm": 1.4408384970498218, | |
| "learning_rate": 1.1608678462613987e-07, | |
| "loss": 0.0405, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 1.8645418326693228, | |
| "grad_norm": 1.8837289894616709, | |
| "learning_rate": 1.1275894651724517e-07, | |
| "loss": 0.0432, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 1.8665338645418328, | |
| "grad_norm": 1.5772528501243686, | |
| "learning_rate": 1.0947895992703129e-07, | |
| "loss": 0.034, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 1.8685258964143427, | |
| "grad_norm": 1.498705914254895, | |
| "learning_rate": 1.062468569702102e-07, | |
| "loss": 0.0348, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 1.8705179282868527, | |
| "grad_norm": 1.6273002546701671, | |
| "learning_rate": 1.0306266929265951e-07, | |
| "loss": 0.0444, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 1.8725099601593627, | |
| "grad_norm": 1.8115132687464455, | |
| "learning_rate": 9.992642807111486e-08, | |
| "loss": 0.045, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.8745019920318726, | |
| "grad_norm": 1.2331103677449673, | |
| "learning_rate": 9.683816401286017e-08, | |
| "loss": 0.0384, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 1.8764940239043826, | |
| "grad_norm": 1.719077840733274, | |
| "learning_rate": 9.379790735543182e-08, | |
| "loss": 0.0426, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 1.8784860557768925, | |
| "grad_norm": 1.3832648508966021, | |
| "learning_rate": 9.080568786631939e-08, | |
| "loss": 0.0305, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 1.8804780876494025, | |
| "grad_norm": 1.6655589837620999, | |
| "learning_rate": 8.78615348426759e-08, | |
| "loss": 0.0387, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 1.8824701195219125, | |
| "grad_norm": 1.3436243075109746, | |
| "learning_rate": 8.49654771110292e-08, | |
| "loss": 0.0336, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 1.8844621513944224, | |
| "grad_norm": 1.4941725985898642, | |
| "learning_rate": 8.211754302700159e-08, | |
| "loss": 0.035, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 1.8864541832669324, | |
| "grad_norm": 1.2284626460508044, | |
| "learning_rate": 7.93177604750317e-08, | |
| "loss": 0.0285, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 1.8884462151394423, | |
| "grad_norm": 1.225195965424142, | |
| "learning_rate": 7.656615686809976e-08, | |
| "loss": 0.0269, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 1.8904382470119523, | |
| "grad_norm": 1.3770985891500578, | |
| "learning_rate": 7.386275914746222e-08, | |
| "loss": 0.0324, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 1.8924302788844622, | |
| "grad_norm": 1.2826488338620294, | |
| "learning_rate": 7.120759378238585e-08, | |
| "loss": 0.032, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.8944223107569722, | |
| "grad_norm": 1.6006341483416981, | |
| "learning_rate": 6.860068676988907e-08, | |
| "loss": 0.0392, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 1.8964143426294822, | |
| "grad_norm": 1.4187750994999075, | |
| "learning_rate": 6.604206363448662e-08, | |
| "loss": 0.0321, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 1.8984063745019921, | |
| "grad_norm": 1.3893878947453275, | |
| "learning_rate": 6.353174942794138e-08, | |
| "loss": 0.0405, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 1.900398406374502, | |
| "grad_norm": 1.7017032311402343, | |
| "learning_rate": 6.106976872901793e-08, | |
| "loss": 0.0358, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 1.902390438247012, | |
| "grad_norm": 1.5607283128871816, | |
| "learning_rate": 5.865614564324273e-08, | |
| "loss": 0.0366, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 1.904382470119522, | |
| "grad_norm": 1.8745845801550434, | |
| "learning_rate": 5.6290903802665444e-08, | |
| "loss": 0.0493, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 1.906374501992032, | |
| "grad_norm": 1.8941896384592207, | |
| "learning_rate": 5.397406636563296e-08, | |
| "loss": 0.0482, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 1.908366533864542, | |
| "grad_norm": 1.7755712845597584, | |
| "learning_rate": 5.1705656016555196e-08, | |
| "loss": 0.0377, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 1.9103585657370519, | |
| "grad_norm": 1.5081972730474071, | |
| "learning_rate": 4.948569496569078e-08, | |
| "loss": 0.0424, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 1.9123505976095618, | |
| "grad_norm": 1.5670265082331885, | |
| "learning_rate": 4.7314204948923356e-08, | |
| "loss": 0.0352, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.9143426294820718, | |
| "grad_norm": 1.5649598609019262, | |
| "learning_rate": 4.5191207227553437e-08, | |
| "loss": 0.0325, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 1.9163346613545817, | |
| "grad_norm": 1.2941707907708186, | |
| "learning_rate": 4.311672258808575e-08, | |
| "loss": 0.0282, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 1.9183266932270917, | |
| "grad_norm": 1.5926908997623397, | |
| "learning_rate": 4.109077134202999e-08, | |
| "loss": 0.0338, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 1.9203187250996017, | |
| "grad_norm": 1.6503076572985267, | |
| "learning_rate": 3.911337332569876e-08, | |
| "loss": 0.0406, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 1.9223107569721116, | |
| "grad_norm": 1.7379935919621892, | |
| "learning_rate": 3.718454790001546e-08, | |
| "loss": 0.0402, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 1.9243027888446216, | |
| "grad_norm": 1.2573758118317542, | |
| "learning_rate": 3.530431395032396e-08, | |
| "loss": 0.0308, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 1.9262948207171315, | |
| "grad_norm": 1.8088122825079003, | |
| "learning_rate": 3.347268988620256e-08, | |
| "loss": 0.0416, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 1.9282868525896415, | |
| "grad_norm": 1.5128035830310647, | |
| "learning_rate": 3.168969364128527e-08, | |
| "loss": 0.0314, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 1.9302788844621515, | |
| "grad_norm": 1.5006247617840172, | |
| "learning_rate": 2.995534267308697e-08, | |
| "loss": 0.0336, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 1.9322709163346614, | |
| "grad_norm": 1.4914522988842465, | |
| "learning_rate": 2.8269653962829104e-08, | |
| "loss": 0.0391, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.9342629482071714, | |
| "grad_norm": 1.795362403964496, | |
| "learning_rate": 2.6632644015276987e-08, | |
| "loss": 0.0516, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 1.9362549800796813, | |
| "grad_norm": 1.516180508821745, | |
| "learning_rate": 2.5044328858576105e-08, | |
| "loss": 0.0379, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 1.9382470119521913, | |
| "grad_norm": 2.0020652857017813, | |
| "learning_rate": 2.3504724044097206e-08, | |
| "loss": 0.0469, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 1.9402390438247012, | |
| "grad_norm": 1.5598747529026058, | |
| "learning_rate": 2.2013844646280313e-08, | |
| "loss": 0.0396, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 1.9422310756972112, | |
| "grad_norm": 1.6142472802956638, | |
| "learning_rate": 2.057170526249097e-08, | |
| "loss": 0.0404, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.9442231075697212, | |
| "grad_norm": 1.2019709175359712, | |
| "learning_rate": 1.917832001287645e-08, | |
| "loss": 0.0287, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 1.9462151394422311, | |
| "grad_norm": 1.251311639854653, | |
| "learning_rate": 1.783370254022587e-08, | |
| "loss": 0.0376, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 1.948207171314741, | |
| "grad_norm": 1.6937487382902727, | |
| "learning_rate": 1.6537866009837533e-08, | |
| "loss": 0.0487, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 1.950199203187251, | |
| "grad_norm": 1.9336485273597244, | |
| "learning_rate": 1.5290823109390673e-08, | |
| "loss": 0.0436, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 1.952191235059761, | |
| "grad_norm": 1.2972478534447982, | |
| "learning_rate": 1.4092586048820578e-08, | |
| "loss": 0.0355, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.954183266932271, | |
| "grad_norm": 1.5530751540422953, | |
| "learning_rate": 1.2943166560199228e-08, | |
| "loss": 0.0328, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 1.956175298804781, | |
| "grad_norm": 1.5748922448723512, | |
| "learning_rate": 1.1842575897619835e-08, | |
| "loss": 0.0325, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 1.9581673306772909, | |
| "grad_norm": 1.4683412245601868, | |
| "learning_rate": 1.0790824837088043e-08, | |
| "loss": 0.0305, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 1.9601593625498008, | |
| "grad_norm": 1.704418853645297, | |
| "learning_rate": 9.787923676414235e-09, | |
| "loss": 0.043, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 1.9621513944223108, | |
| "grad_norm": 1.2313086946224918, | |
| "learning_rate": 8.833882235115277e-09, | |
| "loss": 0.041, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 1.9641434262948207, | |
| "grad_norm": 1.5411732148234083, | |
| "learning_rate": 7.928709854316818e-09, | |
| "loss": 0.0319, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 1.9661354581673307, | |
| "grad_norm": 1.803560123914151, | |
| "learning_rate": 7.072415396661703e-09, | |
| "loss": 0.0405, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 1.9681274900398407, | |
| "grad_norm": 1.4510860737042486, | |
| "learning_rate": 6.265007246223365e-09, | |
| "loss": 0.0348, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 1.9701195219123506, | |
| "grad_norm": 1.508510166730464, | |
| "learning_rate": 5.506493308425342e-09, | |
| "loss": 0.0334, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 1.9721115537848606, | |
| "grad_norm": 1.7329289625239936, | |
| "learning_rate": 4.796881009961341e-09, | |
| "loss": 0.0427, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.9741035856573705, | |
| "grad_norm": 1.6711540109669145, | |
| "learning_rate": 4.136177298724176e-09, | |
| "loss": 0.0321, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 1.9760956175298805, | |
| "grad_norm": 1.5519285892825812, | |
| "learning_rate": 3.524388643736387e-09, | |
| "loss": 0.0337, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 1.9780876494023905, | |
| "grad_norm": 1.5062338202236834, | |
| "learning_rate": 2.9615210350891764e-09, | |
| "loss": 0.0342, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 1.9800796812749004, | |
| "grad_norm": 1.4830859961400638, | |
| "learning_rate": 2.447579983881343e-09, | |
| "loss": 0.0316, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 1.9820717131474104, | |
| "grad_norm": 1.723742367527845, | |
| "learning_rate": 1.9825705221665493e-09, | |
| "loss": 0.0375, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 1.9840637450199203, | |
| "grad_norm": 1.41394511694927, | |
| "learning_rate": 1.566497202904471e-09, | |
| "loss": 0.0296, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 1.9860557768924303, | |
| "grad_norm": 1.3088020276395265, | |
| "learning_rate": 1.1993640999147238e-09, | |
| "loss": 0.0267, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 1.9880478087649402, | |
| "grad_norm": 1.6100561146810857, | |
| "learning_rate": 8.811748078385584e-10, | |
| "loss": 0.0358, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 1.9900398406374502, | |
| "grad_norm": 1.535784061618093, | |
| "learning_rate": 6.119324421016704e-10, | |
| "loss": 0.035, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 1.9920318725099602, | |
| "grad_norm": 1.4352594424209704, | |
| "learning_rate": 3.916396388869981e-10, | |
| "loss": 0.0378, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.9920318725099602, | |
| "eval_loss": 0.10359270870685577, | |
| "eval_runtime": 3.1378, | |
| "eval_samples_per_second": 13.067, | |
| "eval_steps_per_second": 3.506, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.9940239043824701, | |
| "grad_norm": 1.4155470820818645, | |
| "learning_rate": 2.2029855510474762e-10, | |
| "loss": 0.0368, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 1.99601593625498, | |
| "grad_norm": 1.7851317595092946, | |
| "learning_rate": 9.791086837573905e-11, | |
| "loss": 0.0397, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 1.99800796812749, | |
| "grad_norm": 1.4719784863811807, | |
| "learning_rate": 2.4477777010312175e-11, | |
| "loss": 0.0351, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.1952136070405424, | |
| "learning_rate": 0.0, | |
| "loss": 0.028, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 1004, | |
| "total_flos": 18718017454080.0, | |
| "train_loss": 0.09493647176267735, | |
| "train_runtime": 1399.7108, | |
| "train_samples_per_second": 5.73, | |
| "train_steps_per_second": 0.717 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1004, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 5000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 18718017454080.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |