| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 19419, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.007724393635099644, | |
| "grad_norm": 15.9375, | |
| "learning_rate": 2.5746652935118437e-06, | |
| "loss": 2.0265, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.015448787270199289, | |
| "grad_norm": 11.0, | |
| "learning_rate": 5.149330587023687e-06, | |
| "loss": 1.6081, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.023173180905298935, | |
| "grad_norm": 10.375, | |
| "learning_rate": 7.723995880535532e-06, | |
| "loss": 0.9443, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.030897574540398577, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 1.0298661174047375e-05, | |
| "loss": 0.6997, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03862196817549823, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.287332646755922e-05, | |
| "loss": 0.6121, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.04634636181059787, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 1.5447991761071064e-05, | |
| "loss": 0.6053, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.05407075544569751, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 1.8022657054582907e-05, | |
| "loss": 0.4967, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.061795149080797154, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 2.059732234809475e-05, | |
| "loss": 0.466, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.0695195427158968, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 2.3171987641606592e-05, | |
| "loss": 0.4914, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.07724393635099645, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 2.574665293511844e-05, | |
| "loss": 0.4543, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.08496832998609609, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.8321318228630278e-05, | |
| "loss": 0.4135, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.09269272362119574, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 3.089598352214213e-05, | |
| "loss": 0.4241, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.10041711725629537, | |
| "grad_norm": 2.671875, | |
| "learning_rate": 3.3470648815653964e-05, | |
| "loss": 0.4221, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.10814151089139502, | |
| "grad_norm": 3.953125, | |
| "learning_rate": 3.604531410916581e-05, | |
| "loss": 0.4454, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.11586590452649467, | |
| "grad_norm": 3.4375, | |
| "learning_rate": 3.8619979402677656e-05, | |
| "loss": 0.4157, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.12359029816159431, | |
| "grad_norm": 2.125, | |
| "learning_rate": 4.11946446961895e-05, | |
| "loss": 0.4039, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.13131469179669397, | |
| "grad_norm": 2.375, | |
| "learning_rate": 4.376930998970134e-05, | |
| "loss": 0.3939, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.1390390854317936, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 4.6343975283213184e-05, | |
| "loss": 0.4059, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.14676347906689324, | |
| "grad_norm": 1.75, | |
| "learning_rate": 4.891864057672503e-05, | |
| "loss": 0.3746, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.1544878727019929, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 5.149330587023688e-05, | |
| "loss": 0.3894, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.16221226633709254, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 5.406797116374872e-05, | |
| "loss": 0.3776, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.16993665997219218, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 5.6642636457260556e-05, | |
| "loss": 0.3761, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.17766105360729184, | |
| "grad_norm": 3.046875, | |
| "learning_rate": 5.9217301750772405e-05, | |
| "loss": 0.3685, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.18538544724239148, | |
| "grad_norm": 2.921875, | |
| "learning_rate": 6.179196704428426e-05, | |
| "loss": 0.3576, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.1931098408774911, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 6.436663233779608e-05, | |
| "loss": 0.3846, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.20083423451259075, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 6.694129763130793e-05, | |
| "loss": 0.3429, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.2085586281476904, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 6.951596292481978e-05, | |
| "loss": 0.337, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.21628302178279005, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 7.209062821833163e-05, | |
| "loss": 0.385, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.22400741541788968, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 7.466529351184346e-05, | |
| "loss": 0.3183, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.23173180905298935, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 7.723995880535531e-05, | |
| "loss": 0.3743, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.23945620268808898, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 7.981462409886715e-05, | |
| "loss": 0.3434, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.24718059632318862, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 8.2389289392379e-05, | |
| "loss": 0.311, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.25490498995828825, | |
| "grad_norm": 3.765625, | |
| "learning_rate": 8.496395468589084e-05, | |
| "loss": 0.3344, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.26262938359338794, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 8.753861997940268e-05, | |
| "loss": 0.3135, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.2703537772284876, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 9.011328527291453e-05, | |
| "loss": 0.3097, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.2780781708635872, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 9.268795056642637e-05, | |
| "loss": 0.3259, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.28580256449868685, | |
| "grad_norm": 2.671875, | |
| "learning_rate": 9.526261585993821e-05, | |
| "loss": 0.3601, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.2935269581337865, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 9.783728115345005e-05, | |
| "loss": 0.308, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.3012513517688861, | |
| "grad_norm": 2.671875, | |
| "learning_rate": 0.0001004119464469619, | |
| "loss": 0.3259, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.3089757454039858, | |
| "grad_norm": 1.125, | |
| "learning_rate": 0.00010298661174047375, | |
| "loss": 0.2879, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.31670013903908545, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 0.0001055612770339856, | |
| "loss": 0.3252, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.3244245326741851, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 0.00010813594232749744, | |
| "loss": 0.3302, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.3321489263092847, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 0.00011071060762100927, | |
| "loss": 0.313, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.33987331994438436, | |
| "grad_norm": 2.125, | |
| "learning_rate": 0.00011328527291452111, | |
| "loss": 0.3088, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.347597713579484, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 0.00011585993820803295, | |
| "loss": 0.2967, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.3553221072145837, | |
| "grad_norm": 2.375, | |
| "learning_rate": 0.00011843460350154481, | |
| "loss": 0.2821, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.3630465008496833, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 0.00012100926879505665, | |
| "loss": 0.3559, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.37077089448478295, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 0.0001235839340885685, | |
| "loss": 0.295, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.3784952881198826, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 0.00012615859938208034, | |
| "loss": 0.3016, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.3862196817549822, | |
| "grad_norm": 2.703125, | |
| "learning_rate": 0.00012873326467559217, | |
| "loss": 0.2975, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.39394407539008186, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 0.00013130792996910402, | |
| "loss": 0.3172, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.4016684690251815, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 0.00013388259526261585, | |
| "loss": 0.2923, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.4093928626602812, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 0.0001364572605561277, | |
| "loss": 0.2763, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.4171172562953808, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 0.00013903192584963957, | |
| "loss": 0.3312, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.42484164993048046, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 0.0001416065911431514, | |
| "loss": 0.3032, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.4325660435655801, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 0.00014418125643666325, | |
| "loss": 0.3055, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.44029043720067973, | |
| "grad_norm": 3.28125, | |
| "learning_rate": 0.00014675592173017508, | |
| "loss": 0.3208, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.44801483083577937, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 0.0001493305870236869, | |
| "loss": 0.2917, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.45573922447087906, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 0.00015190525231719877, | |
| "loss": 0.2998, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.4634636181059787, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 0.00015447991761071062, | |
| "loss": 0.2823, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.4711880117410783, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 0.00015705458290422245, | |
| "loss": 0.3304, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.47891240537617796, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 0.0001596292481977343, | |
| "loss": 0.2956, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.4866367990112776, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 0.00016220391349124614, | |
| "loss": 0.2934, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.49436119264637723, | |
| "grad_norm": 0.93359375, | |
| "learning_rate": 0.000164778578784758, | |
| "loss": 0.3031, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.5020855862814769, | |
| "grad_norm": 3.046875, | |
| "learning_rate": 0.00016735324407826982, | |
| "loss": 0.2679, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.5098099799165765, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 0.00016992790937178168, | |
| "loss": 0.3182, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.5175343735516762, | |
| "grad_norm": 0.94921875, | |
| "learning_rate": 0.0001725025746652935, | |
| "loss": 0.2888, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.5252587671867759, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 0.00017507723995880537, | |
| "loss": 0.326, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.5329831608218755, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 0.0001776519052523172, | |
| "loss": 0.3184, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.5407075544569752, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 0.00018022657054582905, | |
| "loss": 0.3011, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.5484319480920747, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 0.0001828012358393409, | |
| "loss": 0.309, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.5561563417271744, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 0.00018537590113285274, | |
| "loss": 0.2876, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.563880735362274, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 0.00018795056642636457, | |
| "loss": 0.3071, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.5716051289973737, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 0.00019052523171987642, | |
| "loss": 0.3237, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.5793295226324734, | |
| "grad_norm": 0.77734375, | |
| "learning_rate": 0.00019309989701338825, | |
| "loss": 0.2923, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.587053916267573, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 0.0001956745623069001, | |
| "loss": 0.296, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.5947783099026727, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 0.00019824922760041197, | |
| "loss": 0.2936, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.6025027035377722, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 0.0001999994765359982, | |
| "loss": 0.3222, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.6102270971728719, | |
| "grad_norm": 4.65625, | |
| "learning_rate": 0.00019999109305729934, | |
| "loss": 0.2641, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.6179514908079716, | |
| "grad_norm": 3.1875, | |
| "learning_rate": 0.0001999724866611328, | |
| "loss": 0.2733, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.6256758844430712, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 0.00019994365924978462, | |
| "loss": 0.2864, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.6334002780781709, | |
| "grad_norm": 3.15625, | |
| "learning_rate": 0.00019990461377051967, | |
| "loss": 0.2958, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.6411246717132705, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 0.0001998553542152807, | |
| "loss": 0.2515, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.6488490653483702, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 0.00019979588562027995, | |
| "loss": 0.3032, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.6565734589834697, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 0.0001997262140654843, | |
| "loss": 0.2912, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.6642978526185694, | |
| "grad_norm": 2.375, | |
| "learning_rate": 0.0001996463466739938, | |
| "loss": 0.2887, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.6720222462536691, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 0.0001995562916113133, | |
| "loss": 0.3176, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.6797466398887687, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 0.0001994560580845175, | |
| "loss": 0.2614, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.6874710335238684, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 0.0001993456563413099, | |
| "loss": 0.2594, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.695195427158968, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 0.00019922509766897502, | |
| "loss": 0.2836, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.7029198207940677, | |
| "grad_norm": 3.125, | |
| "learning_rate": 0.0001990943943932242, | |
| "loss": 0.305, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.7106442144291674, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 0.0001989535598769357, | |
| "loss": 0.2652, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.718368608064267, | |
| "grad_norm": 1.625, | |
| "learning_rate": 0.00019880260851878835, | |
| "loss": 0.2911, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.7260930016993666, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 0.00019864155575178953, | |
| "loss": 0.2663, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.7338173953344662, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 0.0001984704180416973, | |
| "loss": 0.2744, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.7415417889695659, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 0.00019828921288533695, | |
| "loss": 0.2956, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.7492661826046655, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 0.00019809795880881216, | |
| "loss": 0.2574, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.7569905762397652, | |
| "grad_norm": 0.734375, | |
| "learning_rate": 0.0001978966753656109, | |
| "loss": 0.3144, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.7647149698748649, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 0.00019768538313460647, | |
| "loss": 0.2984, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.7724393635099644, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 0.00019746410371795334, | |
| "loss": 0.2347, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.7801637571450641, | |
| "grad_norm": 2.890625, | |
| "learning_rate": 0.00019723285973887865, | |
| "loss": 0.246, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.7878881507801637, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 0.0001969916748393693, | |
| "loss": 0.2651, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.7956125444152634, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 0.00019674057367775487, | |
| "loss": 0.2604, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.803336938050363, | |
| "grad_norm": 3.296875, | |
| "learning_rate": 0.00019647958192618644, | |
| "loss": 0.2768, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.8110613316854627, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 0.00019620872626801203, | |
| "loss": 0.253, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.8187857253205624, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 0.00019592803439504855, | |
| "loss": 0.259, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.826510118955662, | |
| "grad_norm": 4.125, | |
| "learning_rate": 0.00019563753500475054, | |
| "loss": 0.2703, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.8342345125907616, | |
| "grad_norm": 1.25, | |
| "learning_rate": 0.00019533725779727635, | |
| "loss": 0.2941, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.8419589062258612, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 0.00019502723347245143, | |
| "loss": 0.2706, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.8496832998609609, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 0.0001947074937266299, | |
| "loss": 0.3306, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.8574076934960606, | |
| "grad_norm": 3.421875, | |
| "learning_rate": 0.00019437807124945362, | |
| "loss": 0.2738, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.8651320871311602, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 0.00019403899972051045, | |
| "loss": 0.2659, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.8728564807662599, | |
| "grad_norm": 1.5234375, | |
| "learning_rate": 0.0001936903138058906, | |
| "loss": 0.2569, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.8805808744013595, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 0.00019333204915464265, | |
| "loss": 0.2517, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.8883052680364592, | |
| "grad_norm": 2.703125, | |
| "learning_rate": 0.0001929642423951286, | |
| "loss": 0.2738, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.8960296616715587, | |
| "grad_norm": 0.78515625, | |
| "learning_rate": 0.0001925869311312793, | |
| "loss": 0.2498, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.9037540553066584, | |
| "grad_norm": 2.125, | |
| "learning_rate": 0.00019220015393874978, | |
| "loss": 0.231, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.9114784489417581, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 0.0001918039503609753, | |
| "loss": 0.23, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.9192028425768577, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 0.0001913983609051286, | |
| "loss": 0.2203, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 0.9269272362119574, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 0.0001909834270379784, | |
| "loss": 0.2902, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.934651629847057, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 0.00019055919118165002, | |
| "loss": 0.2639, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 0.9423760234821567, | |
| "grad_norm": 2.65625, | |
| "learning_rate": 0.00019012569670928813, | |
| "loss": 0.2503, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.9501004171172563, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 0.00018968298794062242, | |
| "loss": 0.2697, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 0.9578248107523559, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 0.00018923111013743635, | |
| "loss": 0.2265, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.9655492043874556, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 0.00018877010949893975, | |
| "loss": 0.2583, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.9732735980225552, | |
| "grad_norm": 2.703125, | |
| "learning_rate": 0.00018830003315704544, | |
| "loss": 0.24, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.9809979916576549, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 0.00018782092917155047, | |
| "loss": 0.2824, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 0.9887223852927545, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 0.00018733284652522285, | |
| "loss": 0.2911, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.9964467789278542, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 0.0001868358351187932, | |
| "loss": 0.2468, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 1.0041711725629539, | |
| "grad_norm": 1.375, | |
| "learning_rate": 0.00018632994576585343, | |
| "loss": 0.2142, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.0118955661980535, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 0.0001858152301876613, | |
| "loss": 0.2236, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 1.019619959833153, | |
| "grad_norm": 1.0, | |
| "learning_rate": 0.00018529174100785276, | |
| "loss": 0.1858, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 1.0273443534682527, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 0.0001847595317470616, | |
| "loss": 0.1911, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 1.0350687471033524, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 0.00018421865681744786, | |
| "loss": 0.2166, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 1.042793140738452, | |
| "grad_norm": 2.0, | |
| "learning_rate": 0.0001836691715171345, | |
| "loss": 0.2184, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 1.0505175343735518, | |
| "grad_norm": 0.8828125, | |
| "learning_rate": 0.00018311113202455407, | |
| "loss": 0.2082, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 1.0582419280086512, | |
| "grad_norm": 1.125, | |
| "learning_rate": 0.00018254459539270495, | |
| "loss": 0.2122, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 1.065966321643751, | |
| "grad_norm": 2.8125, | |
| "learning_rate": 0.00018196961954331862, | |
| "loss": 0.2062, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 1.0736907152788506, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 0.00018138626326093742, | |
| "loss": 0.203, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 1.0814151089139503, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 0.00018079458618690491, | |
| "loss": 0.1923, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.0891395025490498, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 0.00018019464881326803, | |
| "loss": 0.1923, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 1.0968638961841495, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 0.00017958651247659257, | |
| "loss": 0.1892, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 1.1045882898192492, | |
| "grad_norm": 0.83984375, | |
| "learning_rate": 0.00017897023935169214, | |
| "loss": 0.2116, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 1.1123126834543489, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.0001783458924452717, | |
| "loss": 0.1857, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 1.1200370770894486, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 0.00017771353558948568, | |
| "loss": 0.2102, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 1.1277614707245482, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 0.00017707323343541198, | |
| "loss": 0.2191, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 1.1354858643596477, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 0.00017642505144644212, | |
| "loss": 0.203, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 1.1432102579947474, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 0.00017576905589158844, | |
| "loss": 0.2028, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 1.150934651629847, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 0.00017510531383870882, | |
| "loss": 0.2027, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 1.1586590452649468, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 0.00017443389314764977, | |
| "loss": 0.2239, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.1663834389000463, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 0.00017375486246330848, | |
| "loss": 0.1946, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 1.174107832535146, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 0.0001730682912086149, | |
| "loss": 0.2018, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 1.1818322261702456, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 0.0001723742495774339, | |
| "loss": 0.189, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 1.1895566198053453, | |
| "grad_norm": 4.375, | |
| "learning_rate": 0.00017167280852738878, | |
| "loss": 0.2031, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 1.197281013440445, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 0.00017096403977260666, | |
| "loss": 0.2191, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 1.2050054070755445, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 0.0001702480157763867, | |
| "loss": 0.211, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 1.2127298007106442, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00016952480974379136, | |
| "loss": 0.2126, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 1.2204541943457439, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 0.00016879449561416228, | |
| "loss": 0.1914, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 1.2281785879808436, | |
| "grad_norm": 0.8671875, | |
| "learning_rate": 0.00016805714805356056, | |
| "loss": 0.1786, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 1.235902981615943, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 0.00016731284244713335, | |
| "loss": 0.2282, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.2436273752510427, | |
| "grad_norm": 2.953125, | |
| "learning_rate": 0.00016656165489140632, | |
| "loss": 0.1904, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 1.2513517688861424, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 0.0001658036621865039, | |
| "loss": 0.1763, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 1.259076162521242, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 0.0001650389418282973, | |
| "loss": 0.2155, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 1.2668005561563418, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 0.00016426757200048125, | |
| "loss": 0.2114, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 1.2745249497914415, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 0.00016348963156658105, | |
| "loss": 0.1681, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 1.282249343426541, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 0.00016270520006188925, | |
| "loss": 0.1818, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 1.2899737370616406, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 0.00016191435768533457, | |
| "loss": 0.1826, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 1.2976981306967403, | |
| "grad_norm": 0.74609375, | |
| "learning_rate": 0.000161117185291282, | |
| "loss": 0.1847, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 1.30542252433184, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 0.00016031376438126687, | |
| "loss": 0.2101, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 1.3131469179669395, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 0.00015950417709566182, | |
| "loss": 0.193, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.3208713116020392, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 0.0001586885062052793, | |
| "loss": 0.2035, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 1.3285957052371389, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 0.000157866835102909, | |
| "loss": 0.1661, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 1.3363200988722386, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 0.00015703924779479186, | |
| "loss": 0.1882, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 1.3440444925073383, | |
| "grad_norm": 0.7734375, | |
| "learning_rate": 0.00015620582889203175, | |
| "loss": 0.1959, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 1.351768886142438, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 0.00015536666360194454, | |
| "loss": 0.1594, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 1.3594932797775374, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 0.00015452183771934696, | |
| "loss": 0.1855, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 1.3672176734126371, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 0.00015367143761778503, | |
| "loss": 0.1806, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 1.3749420670477368, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 0.00015281555024070323, | |
| "loss": 0.1944, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 1.3826664606828363, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 0.0001519542630925557, | |
| "loss": 0.1872, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 1.390390854317936, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 0.00015108766422985994, | |
| "loss": 0.1587, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.3981152479530357, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 0.00015021584225219394, | |
| "loss": 0.1686, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 1.4058396415881353, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 0.00014933888629313803, | |
| "loss": 0.1766, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 1.413564035223235, | |
| "grad_norm": 2.0, | |
| "learning_rate": 0.00014845688601116194, | |
| "loss": 0.1688, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 1.4212884288583347, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 0.00014756993158045837, | |
| "loss": 0.175, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 1.4290128224934342, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 0.0001466781136817236, | |
| "loss": 0.1549, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 1.436737216128534, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 0.00014578152349288648, | |
| "loss": 0.1776, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 1.4444616097636336, | |
| "grad_norm": 0.953125, | |
| "learning_rate": 0.00014488025267978683, | |
| "loss": 0.1887, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 1.4521860033987333, | |
| "grad_norm": 0.94921875, | |
| "learning_rate": 0.0001439743933868031, | |
| "loss": 0.1854, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 1.4599103970338327, | |
| "grad_norm": 0.79296875, | |
| "learning_rate": 0.0001430640382274323, | |
| "loss": 0.16, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 1.4676347906689324, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 0.000142149280274821, | |
| "loss": 0.1752, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.4753591843040321, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 0.00014123021305224978, | |
| "loss": 0.1845, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 1.4830835779391318, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 0.00014030693052357164, | |
| "loss": 0.1675, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 1.4908079715742315, | |
| "grad_norm": 0.9453125, | |
| "learning_rate": 0.00013937952708360508, | |
| "loss": 0.1677, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 1.4985323652093312, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 0.00013844809754848361, | |
| "loss": 0.1817, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 1.5062567588444307, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 0.0001375127371459618, | |
| "loss": 0.1588, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 1.5139811524795304, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 0.00013657354150567924, | |
| "loss": 0.1769, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 1.52170554611463, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 0.00013563060664938375, | |
| "loss": 0.1604, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 1.5294299397497295, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 0.00013468402898111407, | |
| "loss": 0.1549, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 1.5371543333848292, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 0.00013373390527734387, | |
| "loss": 0.1789, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 1.544878727019929, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 0.0001327803326770873, | |
| "loss": 0.1651, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.5526031206550286, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 0.0001318234086719678, | |
| "loss": 0.1789, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 1.5603275142901283, | |
| "grad_norm": 2.125, | |
| "learning_rate": 0.00013086323109625072, | |
| "loss": 0.1774, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 1.568051907925228, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 0.0001298998981168408, | |
| "loss": 0.1604, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 1.5757763015603277, | |
| "grad_norm": 2.890625, | |
| "learning_rate": 0.0001289335082232459, | |
| "loss": 0.1755, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 1.5835006951954271, | |
| "grad_norm": 0.7578125, | |
| "learning_rate": 0.0001279641602175075, | |
| "loss": 0.1554, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 1.5912250888305268, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 0.00012699195320409938, | |
| "loss": 0.1518, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 1.5989494824656263, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 0.00012601698657979527, | |
| "loss": 0.1689, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 1.606673876100726, | |
| "grad_norm": 0.6953125, | |
| "learning_rate": 0.00012503936002350699, | |
| "loss": 0.139, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 1.6143982697358257, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 0.00012405917348609296, | |
| "loss": 0.1465, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 1.6221226633709254, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 0.0001230765271801399, | |
| "loss": 0.1446, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.629847057006025, | |
| "grad_norm": 0.9921875, | |
| "learning_rate": 0.00012209152156971692, | |
| "loss": 0.1393, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 1.6375714506411247, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 0.00012110425736010438, | |
| "loss": 0.1418, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 1.6452958442762244, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 0.00012011483548749793, | |
| "loss": 0.1552, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 1.653020237911324, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.00011912335710868905, | |
| "loss": 0.1533, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 1.6607446315464236, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 0.00011812992359072275, | |
| "loss": 0.1625, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 1.6684690251815233, | |
| "grad_norm": 2.6875, | |
| "learning_rate": 0.00011713463650053423, | |
| "loss": 0.1542, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 1.6761934188166228, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 0.00011613759759456462, | |
| "loss": 0.1306, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 1.6839178124517225, | |
| "grad_norm": 0.89453125, | |
| "learning_rate": 0.00011513890880835776, | |
| "loss": 0.1784, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 1.6916422060868221, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 0.00011413867224613843, | |
| "loss": 0.1384, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 1.6993665997219218, | |
| "grad_norm": 2.0, | |
| "learning_rate": 0.00011313699017037327, | |
| "loss": 0.1279, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.7070909933570215, | |
| "grad_norm": 0.90234375, | |
| "learning_rate": 0.00011213396499131581, | |
| "loss": 0.1615, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 1.7148153869921212, | |
| "grad_norm": 1.0, | |
| "learning_rate": 0.00011112969925653618, | |
| "loss": 0.1344, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 1.722539780627221, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00011012429564043667, | |
| "loss": 0.155, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 1.7302641742623204, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00010911785693375478, | |
| "loss": 0.1394, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 1.73798856789742, | |
| "grad_norm": 0.88671875, | |
| "learning_rate": 0.00010811048603305371, | |
| "loss": 0.1499, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 1.7457129615325195, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 0.00010710228593020271, | |
| "loss": 0.1247, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 1.7534373551676192, | |
| "grad_norm": 4.0, | |
| "learning_rate": 0.00010609335970184715, | |
| "loss": 0.1259, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 1.761161748802719, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 0.00010508381049887029, | |
| "loss": 0.1631, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 1.7688861424378186, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 0.00010407374153584716, | |
| "loss": 0.1414, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 1.7766105360729183, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 0.00010306325608049227, | |
| "loss": 0.1276, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.784334929708018, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 0.0001020524574431015, | |
| "loss": 0.1586, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 1.7920593233431177, | |
| "grad_norm": 2.734375, | |
| "learning_rate": 0.00010104144896598997, | |
| "loss": 0.1461, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 1.7997837169782174, | |
| "grad_norm": 0.7734375, | |
| "learning_rate": 0.00010003033401292642, | |
| "loss": 0.1247, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 1.8075081106133168, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 9.901921595856549e-05, | |
| "loss": 0.1482, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 1.8152325042484165, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 9.800819817787906e-05, | |
| "loss": 0.1543, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 1.822956897883516, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 9.699738403558688e-05, | |
| "loss": 0.1474, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 1.8306812915186157, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 9.59868768755894e-05, | |
| "loss": 0.139, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 1.8384056851537154, | |
| "grad_norm": 0.9921875, | |
| "learning_rate": 9.497678001040152e-05, | |
| "loss": 0.1368, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 1.846130078788815, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 9.396719671059039e-05, | |
| "loss": 0.1191, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 1.8538544724239148, | |
| "grad_norm": 1.0859375, | |
| "learning_rate": 9.295823019421707e-05, | |
| "loss": 0.1345, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.8615788660590145, | |
| "grad_norm": 1.0, | |
| "learning_rate": 9.194998361628373e-05, | |
| "loss": 0.1333, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 1.8693032596941141, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 9.094256005818736e-05, | |
| "loss": 0.1419, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 1.8770276533292136, | |
| "grad_norm": 2.765625, | |
| "learning_rate": 8.993606251718072e-05, | |
| "loss": 0.1546, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 1.8847520469643133, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 8.893059389584226e-05, | |
| "loss": 0.1493, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 1.892476440599413, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 8.792625699155531e-05, | |
| "loss": 0.1421, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 1.9002008342345125, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 8.692315448599864e-05, | |
| "loss": 0.1322, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 1.9079252278696122, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 8.592138893464792e-05, | |
| "loss": 0.136, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 1.9156496215047119, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 8.492106275629117e-05, | |
| "loss": 0.1334, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 1.9233740151398115, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 8.392227822255723e-05, | |
| "loss": 0.1314, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 1.9310984087749112, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 8.292513744745992e-05, | |
| "loss": 0.1123, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.938822802410011, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 8.192974237695798e-05, | |
| "loss": 0.1132, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 1.9465471960451106, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 8.093619477853233e-05, | |
| "loss": 0.112, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 1.95427158968021, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 7.99445962307815e-05, | |
| "loss": 0.1074, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 1.9619959833153098, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 7.89550481130364e-05, | |
| "loss": 0.1303, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 1.9697203769504092, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 7.796765159499558e-05, | |
| "loss": 0.1272, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 1.977444770585509, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 7.698250762638167e-05, | |
| "loss": 0.1154, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 1.9851691642206086, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 7.599971692662056e-05, | |
| "loss": 0.1245, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 1.9928935578557083, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 7.501937997454392e-05, | |
| "loss": 0.1121, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 2.000617951490808, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 7.404159699811656e-05, | |
| "loss": 0.1051, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 2.0083423451259077, | |
| "grad_norm": 0.87109375, | |
| "learning_rate": 7.306646796418912e-05, | |
| "loss": 0.0843, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.0160667387610074, | |
| "grad_norm": 0.8359375, | |
| "learning_rate": 7.209409256827782e-05, | |
| "loss": 0.0768, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 2.023791132396107, | |
| "grad_norm": 0.98828125, | |
| "learning_rate": 7.112457022437155e-05, | |
| "loss": 0.0752, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 2.0315155260312068, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 7.015800005476818e-05, | |
| "loss": 0.0776, | |
| "step": 13150 | |
| }, | |
| { | |
| "epoch": 2.039239919666306, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 6.919448087994023e-05, | |
| "loss": 0.0857, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 2.0469643133014057, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 6.823411120843188e-05, | |
| "loss": 0.0769, | |
| "step": 13250 | |
| }, | |
| { | |
| "epoch": 2.0546887069365054, | |
| "grad_norm": 0.80078125, | |
| "learning_rate": 6.727698922678745e-05, | |
| "loss": 0.0902, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 2.062413100571605, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 6.632321278951299e-05, | |
| "loss": 0.0841, | |
| "step": 13350 | |
| }, | |
| { | |
| "epoch": 2.070137494206705, | |
| "grad_norm": 1.0546875, | |
| "learning_rate": 6.537287940907194e-05, | |
| "loss": 0.0771, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 2.0778618878418045, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 6.442608624591541e-05, | |
| "loss": 0.0785, | |
| "step": 13450 | |
| }, | |
| { | |
| "epoch": 2.085586281476904, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 6.348293009854896e-05, | |
| "loss": 0.0795, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.093310675112004, | |
| "grad_norm": 3.1875, | |
| "learning_rate": 6.254350739363564e-05, | |
| "loss": 0.0742, | |
| "step": 13550 | |
| }, | |
| { | |
| "epoch": 2.1010350687471036, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 6.160791417613805e-05, | |
| "loss": 0.079, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 2.108759462382203, | |
| "grad_norm": 0.74609375, | |
| "learning_rate": 6.067624609949837e-05, | |
| "loss": 0.0717, | |
| "step": 13650 | |
| }, | |
| { | |
| "epoch": 2.1164838560173025, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 5.974859841585925e-05, | |
| "loss": 0.0743, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 2.124208249652402, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 5.882506596632517e-05, | |
| "loss": 0.0726, | |
| "step": 13750 | |
| }, | |
| { | |
| "epoch": 2.131932643287502, | |
| "grad_norm": 0.94140625, | |
| "learning_rate": 5.790574317126625e-05, | |
| "loss": 0.0848, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 2.1396570369226016, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 5.699072402066468e-05, | |
| "loss": 0.0738, | |
| "step": 13850 | |
| }, | |
| { | |
| "epoch": 2.1473814305577013, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 5.608010206450549e-05, | |
| "loss": 0.0675, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 2.155105824192801, | |
| "grad_norm": 2.5, | |
| "learning_rate": 5.5173970403212174e-05, | |
| "loss": 0.0886, | |
| "step": 13950 | |
| }, | |
| { | |
| "epoch": 2.1628302178279006, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 5.427242167812803e-05, | |
| "loss": 0.0796, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.1705546114630003, | |
| "grad_norm": 0.70703125, | |
| "learning_rate": 5.337554806204497e-05, | |
| "loss": 0.0626, | |
| "step": 14050 | |
| }, | |
| { | |
| "epoch": 2.1782790050980996, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 5.2483441249779755e-05, | |
| "loss": 0.0686, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 2.1860033987331993, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 5.159619244879936e-05, | |
| "loss": 0.07, | |
| "step": 14150 | |
| }, | |
| { | |
| "epoch": 2.193727792368299, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 5.07138923698959e-05, | |
| "loss": 0.0739, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 2.2014521860033986, | |
| "grad_norm": 1.0, | |
| "learning_rate": 4.983663121791289e-05, | |
| "loss": 0.0724, | |
| "step": 14250 | |
| }, | |
| { | |
| "epoch": 2.2091765796384983, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 4.8964498682522374e-05, | |
| "loss": 0.071, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 2.216900973273598, | |
| "grad_norm": 0.9296875, | |
| "learning_rate": 4.80975839290556e-05, | |
| "loss": 0.0729, | |
| "step": 14350 | |
| }, | |
| { | |
| "epoch": 2.2246253669086977, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 4.723597558938672e-05, | |
| "loss": 0.0758, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 2.2323497605437974, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 4.637976175287129e-05, | |
| "loss": 0.0639, | |
| "step": 14450 | |
| }, | |
| { | |
| "epoch": 2.240074154178897, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 4.552902995734003e-05, | |
| "loss": 0.0739, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.247798547813997, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 4.468386718014934e-05, | |
| "loss": 0.0634, | |
| "step": 14550 | |
| }, | |
| { | |
| "epoch": 2.2555229414490965, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 4.384435982928871e-05, | |
| "loss": 0.0658, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 2.2632473350841957, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 4.30105937345466e-05, | |
| "loss": 0.07, | |
| "step": 14650 | |
| }, | |
| { | |
| "epoch": 2.2709717287192954, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 4.218265413873519e-05, | |
| "loss": 0.0671, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 2.278696122354395, | |
| "grad_norm": 0.9765625, | |
| "learning_rate": 4.136062568897556e-05, | |
| "loss": 0.081, | |
| "step": 14750 | |
| }, | |
| { | |
| "epoch": 2.286420515989495, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 4.054459242804338e-05, | |
| "loss": 0.0746, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 2.2941449096245945, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 3.973463778577634e-05, | |
| "loss": 0.0755, | |
| "step": 14850 | |
| }, | |
| { | |
| "epoch": 2.301869303259694, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 3.89308445705449e-05, | |
| "loss": 0.0626, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 2.309593696894794, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 3.813329496078553e-05, | |
| "loss": 0.0772, | |
| "step": 14950 | |
| }, | |
| { | |
| "epoch": 2.3173180905298936, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 3.734207049659937e-05, | |
| "loss": 0.0705, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.3250424841649933, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 3.65572520714155e-05, | |
| "loss": 0.0748, | |
| "step": 15050 | |
| }, | |
| { | |
| "epoch": 2.3327668778000925, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 3.577891992372066e-05, | |
| "loss": 0.0687, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 2.340491271435192, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 3.5007153628855546e-05, | |
| "loss": 0.063, | |
| "step": 15150 | |
| }, | |
| { | |
| "epoch": 2.348215665070292, | |
| "grad_norm": 0.73046875, | |
| "learning_rate": 3.4242032090879536e-05, | |
| "loss": 0.0662, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 2.3559400587053916, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 3.3483633534503413e-05, | |
| "loss": 0.0731, | |
| "step": 15250 | |
| }, | |
| { | |
| "epoch": 2.3636644523404913, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 3.27320354970919e-05, | |
| "loss": 0.0733, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 2.371388845975591, | |
| "grad_norm": 0.77734375, | |
| "learning_rate": 3.1987314820736257e-05, | |
| "loss": 0.0645, | |
| "step": 15350 | |
| }, | |
| { | |
| "epoch": 2.3791132396106907, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 3.12495476443983e-05, | |
| "loss": 0.0822, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 2.3868376332457903, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 3.051880939612592e-05, | |
| "loss": 0.0621, | |
| "step": 15450 | |
| }, | |
| { | |
| "epoch": 2.39456202688089, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 2.9795174785341384e-05, | |
| "loss": 0.0744, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.4022864205159893, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 2.9078717795203482e-05, | |
| "loss": 0.0648, | |
| "step": 15550 | |
| }, | |
| { | |
| "epoch": 2.410010814151089, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 2.836951167504317e-05, | |
| "loss": 0.0699, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 2.4177352077861887, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 2.766762893287509e-05, | |
| "loss": 0.0666, | |
| "step": 15650 | |
| }, | |
| { | |
| "epoch": 2.4254596014212884, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 2.6973141327984232e-05, | |
| "loss": 0.066, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 2.433183995056388, | |
| "grad_norm": 0.671875, | |
| "learning_rate": 2.6286119863589532e-05, | |
| "loss": 0.0709, | |
| "step": 15750 | |
| }, | |
| { | |
| "epoch": 2.4409083886914877, | |
| "grad_norm": 0.8515625, | |
| "learning_rate": 2.5606634779584414e-05, | |
| "loss": 0.0598, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 2.4486327823265874, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 2.4934755545355926e-05, | |
| "loss": 0.0824, | |
| "step": 15850 | |
| }, | |
| { | |
| "epoch": 2.456357175961687, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 2.4270550852681916e-05, | |
| "loss": 0.067, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 2.464081569596787, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 2.361408860870843e-05, | |
| "loss": 0.0591, | |
| "step": 15950 | |
| }, | |
| { | |
| "epoch": 2.471805963231886, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 2.296543592900683e-05, | |
| "loss": 0.0662, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.4795303568669858, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 2.2324659130712e-05, | |
| "loss": 0.0623, | |
| "step": 16050 | |
| }, | |
| { | |
| "epoch": 2.4872547505020854, | |
| "grad_norm": 0.84375, | |
| "learning_rate": 2.169182372574231e-05, | |
| "loss": 0.0692, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 2.494979144137185, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 2.1066994414101703e-05, | |
| "loss": 0.0577, | |
| "step": 16150 | |
| }, | |
| { | |
| "epoch": 2.502703537772285, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.0450235077264955e-05, | |
| "loss": 0.0685, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 2.5104279314073845, | |
| "grad_norm": 0.85546875, | |
| "learning_rate": 1.984160877164637e-05, | |
| "loss": 0.0647, | |
| "step": 16250 | |
| }, | |
| { | |
| "epoch": 2.518152325042484, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 1.9241177722153235e-05, | |
| "loss": 0.0687, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 2.525876718677584, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.8649003315823887e-05, | |
| "loss": 0.068, | |
| "step": 16350 | |
| }, | |
| { | |
| "epoch": 2.5336011123126836, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 1.80651460955517e-05, | |
| "loss": 0.0532, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 2.541325505947783, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 1.7489665753895146e-05, | |
| "loss": 0.0681, | |
| "step": 16450 | |
| }, | |
| { | |
| "epoch": 2.549049899582883, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 1.6922621126975193e-05, | |
| "loss": 0.063, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.556774293217982, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 1.6364070188459646e-05, | |
| "loss": 0.07, | |
| "step": 16550 | |
| }, | |
| { | |
| "epoch": 2.564498686853082, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 1.5814070043636342e-05, | |
| "loss": 0.0799, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 2.5722230804881816, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 1.5272676923574637e-05, | |
| "loss": 0.0697, | |
| "step": 16650 | |
| }, | |
| { | |
| "epoch": 2.5799474741232813, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 1.473994617937643e-05, | |
| "loss": 0.0637, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 2.587671867758381, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 1.4215932276517253e-05, | |
| "loss": 0.0626, | |
| "step": 16750 | |
| }, | |
| { | |
| "epoch": 2.5953962613934807, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 1.3700688789277738e-05, | |
| "loss": 0.0735, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 2.6031206550285804, | |
| "grad_norm": 0.97265625, | |
| "learning_rate": 1.319426839526633e-05, | |
| "loss": 0.062, | |
| "step": 16850 | |
| }, | |
| { | |
| "epoch": 2.61084504866368, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 1.2696722870033529e-05, | |
| "loss": 0.0632, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 2.6185694422987797, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 1.2208103081778555e-05, | |
| "loss": 0.0677, | |
| "step": 16950 | |
| }, | |
| { | |
| "epoch": 2.626293835933879, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 1.1728458986148627e-05, | |
| "loss": 0.0605, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.6340182295689787, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 1.1257839621131584e-05, | |
| "loss": 0.0606, | |
| "step": 17050 | |
| }, | |
| { | |
| "epoch": 2.6417426232040784, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 1.0796293102042243e-05, | |
| "loss": 0.0641, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 2.649467016839178, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 1.034386661660337e-05, | |
| "loss": 0.0632, | |
| "step": 17150 | |
| }, | |
| { | |
| "epoch": 2.6571914104742778, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 9.900606420121062e-06, | |
| "loss": 0.0671, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 2.6649158041093775, | |
| "grad_norm": 0.81640625, | |
| "learning_rate": 9.466557830755862e-06, | |
| "loss": 0.0646, | |
| "step": 17250 | |
| }, | |
| { | |
| "epoch": 2.672640197744477, | |
| "grad_norm": 0.91015625, | |
| "learning_rate": 9.041765224889399e-06, | |
| "loss": 0.0674, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 2.680364591379577, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 8.62627203258748e-06, | |
| "loss": 0.058, | |
| "step": 17350 | |
| }, | |
| { | |
| "epoch": 2.6880889850146765, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 8.220120733159831e-06, | |
| "loss": 0.0563, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 2.6958133786497758, | |
| "grad_norm": 0.91796875, | |
| "learning_rate": 7.823352850817124e-06, | |
| "loss": 0.0673, | |
| "step": 17450 | |
| }, | |
| { | |
| "epoch": 2.703537772284876, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 7.4360089504256215e-06, | |
| "loss": 0.0664, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.711262165919975, | |
| "grad_norm": 0.78125, | |
| "learning_rate": 7.058128633359806e-06, | |
| "loss": 0.0691, | |
| "step": 17550 | |
| }, | |
| { | |
| "epoch": 2.718986559555075, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 6.689750533453765e-06, | |
| "loss": 0.0606, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 2.7267109531901745, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 6.33091231305114e-06, | |
| "loss": 0.0665, | |
| "step": 17650 | |
| }, | |
| { | |
| "epoch": 2.7344353468252742, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 5.9816506591547475e-06, | |
| "loss": 0.0623, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 2.742159740460374, | |
| "grad_norm": 0.72265625, | |
| "learning_rate": 5.6420012796757015e-06, | |
| "loss": 0.0665, | |
| "step": 17750 | |
| }, | |
| { | |
| "epoch": 2.7498841340954736, | |
| "grad_norm": 0.625, | |
| "learning_rate": 5.311998899782711e-06, | |
| "loss": 0.057, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 2.7576085277305733, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 4.99167725835179e-06, | |
| "loss": 0.0732, | |
| "step": 17850 | |
| }, | |
| { | |
| "epoch": 2.7653329213656725, | |
| "grad_norm": 0.8671875, | |
| "learning_rate": 4.681069104516922e-06, | |
| "loss": 0.0653, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 2.7730573150007727, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 4.380206194321812e-06, | |
| "loss": 0.0636, | |
| "step": 17950 | |
| }, | |
| { | |
| "epoch": 2.780781708635872, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 4.089119287473242e-06, | |
| "loss": 0.0631, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.7885061022709716, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 3.8078381441961407e-06, | |
| "loss": 0.063, | |
| "step": 18050 | |
| }, | |
| { | |
| "epoch": 2.7962304959060713, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 3.536391522191096e-06, | |
| "loss": 0.0797, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 2.803954889541171, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 3.2748071736941164e-06, | |
| "loss": 0.0595, | |
| "step": 18150 | |
| }, | |
| { | |
| "epoch": 2.8116792831762707, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 3.0231118426393035e-06, | |
| "loss": 0.0596, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 2.8194036768113704, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 2.78133126192468e-06, | |
| "loss": 0.0642, | |
| "step": 18250 | |
| }, | |
| { | |
| "epoch": 2.82712807044647, | |
| "grad_norm": 0.96484375, | |
| "learning_rate": 2.5494901507811418e-06, | |
| "loss": 0.0549, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 2.8348524640815693, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 2.327612212245389e-06, | |
| "loss": 0.0636, | |
| "step": 18350 | |
| }, | |
| { | |
| "epoch": 2.8425768577166695, | |
| "grad_norm": 0.89453125, | |
| "learning_rate": 2.1157201307364315e-06, | |
| "loss": 0.0667, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 2.8503012513517687, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 1.913835569736444e-06, | |
| "loss": 0.0613, | |
| "step": 18450 | |
| }, | |
| { | |
| "epoch": 2.8580256449868684, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 1.7219791695758824e-06, | |
| "loss": 0.071, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.865750038621968, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 1.5401705453233051e-06, | |
| "loss": 0.076, | |
| "step": 18550 | |
| }, | |
| { | |
| "epoch": 2.873474432257068, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 1.3684282847799213e-06, | |
| "loss": 0.0667, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 2.8811988258921675, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 1.2067699465792336e-06, | |
| "loss": 0.0651, | |
| "step": 18650 | |
| }, | |
| { | |
| "epoch": 2.888923219527267, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 1.0552120583918412e-06, | |
| "loss": 0.0575, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 2.896647613162367, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 9.137701152356792e-07, | |
| "loss": 0.0564, | |
| "step": 18750 | |
| }, | |
| { | |
| "epoch": 2.9043720067974665, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 7.824585778919091e-07, | |
| "loss": 0.0699, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 2.9120964004325662, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 6.612908714263455e-07, | |
| "loss": 0.0628, | |
| "step": 18850 | |
| }, | |
| { | |
| "epoch": 2.9198207940676655, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 5.502793838170428e-07, | |
| "loss": 0.0642, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 2.927545187702765, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 4.49435464687642e-07, | |
| "loss": 0.0642, | |
| "step": 18950 | |
| }, | |
| { | |
| "epoch": 2.935269581337865, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 3.587694241470874e-07, | |
| "loss": 0.0614, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.9429939749729646, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 2.7829053173550333e-07, | |
| "loss": 0.0658, | |
| "step": 19050 | |
| }, | |
| { | |
| "epoch": 2.9507183686080642, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 2.0800701547651902e-07, | |
| "loss": 0.0675, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 2.958442762243164, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 1.479260610360078e-07, | |
| "loss": 0.0703, | |
| "step": 19150 | |
| }, | |
| { | |
| "epoch": 2.9661671558782636, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 9.805381098746402e-08, | |
| "loss": 0.0703, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 2.9738915495133633, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 5.839536418401625e-08, | |
| "loss": 0.0633, | |
| "step": 19250 | |
| }, | |
| { | |
| "epoch": 2.981615943148463, | |
| "grad_norm": 2.765625, | |
| "learning_rate": 2.8954775237088894e-08, | |
| "loss": 0.0634, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 2.9893403367835623, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 9.735054101878138e-09, | |
| "loss": 0.058, | |
| "step": 19350 | |
| }, | |
| { | |
| "epoch": 2.9970647304186624, | |
| "grad_norm": 0.9296875, | |
| "learning_rate": 7.381657696425848e-10, | |
| "loss": 0.0586, | |
| "step": 19400 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 19419, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 0, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.3806574237646848e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |