diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,17857 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 5.0, + "eval_steps": 500, + "global_step": 2545, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0019646365422396855, + "grad_norm": 5.513031317487342, + "learning_rate": 3.1372549019607843e-07, + "loss": 1.109, + "step": 1 + }, + { + "epoch": 0.003929273084479371, + "grad_norm": 5.506255064928022, + "learning_rate": 6.274509803921569e-07, + "loss": 1.1055, + "step": 2 + }, + { + "epoch": 0.005893909626719057, + "grad_norm": 5.559827388629861, + "learning_rate": 9.411764705882353e-07, + "loss": 1.13, + "step": 3 + }, + { + "epoch": 0.007858546168958742, + "grad_norm": 5.500623835384275, + "learning_rate": 1.2549019607843137e-06, + "loss": 1.1138, + "step": 4 + }, + { + "epoch": 0.009823182711198428, + "grad_norm": 5.391019758036856, + "learning_rate": 1.5686274509803923e-06, + "loss": 1.0912, + "step": 5 + }, + { + "epoch": 0.011787819253438114, + "grad_norm": 5.075312812440821, + "learning_rate": 1.8823529411764707e-06, + "loss": 1.0861, + "step": 6 + }, + { + "epoch": 0.0137524557956778, + "grad_norm": 4.200969757170934, + "learning_rate": 2.196078431372549e-06, + "loss": 1.0666, + "step": 7 + }, + { + "epoch": 0.015717092337917484, + "grad_norm": 3.909850225853453, + "learning_rate": 2.5098039215686274e-06, + "loss": 1.0426, + "step": 8 + }, + { + "epoch": 0.01768172888015717, + "grad_norm": 2.241800473949379, + "learning_rate": 2.8235294117647062e-06, + "loss": 1.0021, + "step": 9 + }, + { + "epoch": 0.019646365422396856, + "grad_norm": 2.115257362756516, + "learning_rate": 3.1372549019607846e-06, + "loss": 1.0005, + "step": 10 + }, + { + "epoch": 0.021611001964636542, + "grad_norm": 1.8828416407298572, + "learning_rate": 3.450980392156863e-06, + "loss": 0.9898, + "step": 11 + }, + { + "epoch": 0.023575638506876228, + "grad_norm": 3.4194380056221667, + "learning_rate": 3.7647058823529414e-06, + "loss": 0.9725, + "step": 12 + }, + { + "epoch": 0.025540275049115914, + "grad_norm": 3.577041048008891, + "learning_rate": 4.07843137254902e-06, + "loss": 0.9764, + "step": 13 + }, + { + "epoch": 0.0275049115913556, + "grad_norm": 3.480551515978332, + "learning_rate": 4.392156862745098e-06, + "loss": 0.9583, + "step": 14 + }, + { + "epoch": 0.029469548133595286, + "grad_norm": 3.3834699637118275, + "learning_rate": 4.705882352941177e-06, + "loss": 0.9515, + "step": 15 + }, + { + "epoch": 0.03143418467583497, + "grad_norm": 2.330021723736843, + "learning_rate": 5.019607843137255e-06, + "loss": 0.909, + "step": 16 + }, + { + "epoch": 0.03339882121807466, + "grad_norm": 2.124442115728281, + "learning_rate": 5.333333333333334e-06, + "loss": 0.9061, + "step": 17 + }, + { + "epoch": 0.03536345776031434, + "grad_norm": 1.8402351084692825, + "learning_rate": 5.6470588235294125e-06, + "loss": 0.8929, + "step": 18 + }, + { + "epoch": 0.03732809430255403, + "grad_norm": 1.5118828364684072, + "learning_rate": 5.96078431372549e-06, + "loss": 0.8655, + "step": 19 + }, + { + "epoch": 0.03929273084479371, + "grad_norm": 1.4117041370780234, + "learning_rate": 6.274509803921569e-06, + "loss": 0.8513, + "step": 20 + }, + { + "epoch": 0.0412573673870334, + "grad_norm": 1.39491499718028, + "learning_rate": 6.588235294117647e-06, + "loss": 0.8439, + "step": 21 + }, + { + "epoch": 0.043222003929273084, + "grad_norm": 1.3518734743810514, + "learning_rate": 6.901960784313726e-06, + "loss": 0.8365, + "step": 22 + }, + { + "epoch": 0.04518664047151277, + "grad_norm": 1.3169741732419362, + "learning_rate": 7.215686274509805e-06, + "loss": 0.83, + "step": 23 + }, + { + "epoch": 0.047151277013752456, + "grad_norm": 1.215187689940333, + "learning_rate": 7.529411764705883e-06, + "loss": 0.8292, + "step": 24 + }, + { + "epoch": 0.04911591355599214, + "grad_norm": 1.2380870948848177, + "learning_rate": 7.84313725490196e-06, + "loss": 0.8084, + "step": 25 + }, + { + "epoch": 0.05108055009823183, + "grad_norm": 1.1739541013219612, + "learning_rate": 8.15686274509804e-06, + "loss": 0.8184, + "step": 26 + }, + { + "epoch": 0.05304518664047151, + "grad_norm": 1.0077046330866353, + "learning_rate": 8.470588235294118e-06, + "loss": 0.796, + "step": 27 + }, + { + "epoch": 0.0550098231827112, + "grad_norm": 0.8842007360532814, + "learning_rate": 8.784313725490196e-06, + "loss": 0.7888, + "step": 28 + }, + { + "epoch": 0.05697445972495088, + "grad_norm": 0.9089060673420012, + "learning_rate": 9.098039215686276e-06, + "loss": 0.8026, + "step": 29 + }, + { + "epoch": 0.05893909626719057, + "grad_norm": 0.8781991231333389, + "learning_rate": 9.411764705882354e-06, + "loss": 0.79, + "step": 30 + }, + { + "epoch": 0.060903732809430254, + "grad_norm": 0.7648100177963718, + "learning_rate": 9.725490196078432e-06, + "loss": 0.7903, + "step": 31 + }, + { + "epoch": 0.06286836935166994, + "grad_norm": 0.6585269564231376, + "learning_rate": 1.003921568627451e-05, + "loss": 0.7749, + "step": 32 + }, + { + "epoch": 0.06483300589390963, + "grad_norm": 0.5967474593510858, + "learning_rate": 1.035294117647059e-05, + "loss": 0.7797, + "step": 33 + }, + { + "epoch": 0.06679764243614932, + "grad_norm": 0.591847426300698, + "learning_rate": 1.0666666666666667e-05, + "loss": 0.7699, + "step": 34 + }, + { + "epoch": 0.068762278978389, + "grad_norm": 0.573612852612207, + "learning_rate": 1.0980392156862747e-05, + "loss": 0.7693, + "step": 35 + }, + { + "epoch": 0.07072691552062868, + "grad_norm": 0.4443524007618329, + "learning_rate": 1.1294117647058825e-05, + "loss": 0.7607, + "step": 36 + }, + { + "epoch": 0.07269155206286837, + "grad_norm": 0.46199017953005933, + "learning_rate": 1.1607843137254905e-05, + "loss": 0.7528, + "step": 37 + }, + { + "epoch": 0.07465618860510806, + "grad_norm": 0.4646582594794358, + "learning_rate": 1.192156862745098e-05, + "loss": 0.7557, + "step": 38 + }, + { + "epoch": 0.07662082514734773, + "grad_norm": 0.409359919743574, + "learning_rate": 1.223529411764706e-05, + "loss": 0.748, + "step": 39 + }, + { + "epoch": 0.07858546168958742, + "grad_norm": 0.3973293810950604, + "learning_rate": 1.2549019607843138e-05, + "loss": 0.744, + "step": 40 + }, + { + "epoch": 0.08055009823182711, + "grad_norm": 0.4225083135987142, + "learning_rate": 1.2862745098039218e-05, + "loss": 0.7614, + "step": 41 + }, + { + "epoch": 0.0825147347740668, + "grad_norm": 0.37677665159245793, + "learning_rate": 1.3176470588235294e-05, + "loss": 0.7414, + "step": 42 + }, + { + "epoch": 0.08447937131630648, + "grad_norm": 0.4089301697055092, + "learning_rate": 1.3490196078431374e-05, + "loss": 0.7484, + "step": 43 + }, + { + "epoch": 0.08644400785854617, + "grad_norm": 0.3551201457048581, + "learning_rate": 1.3803921568627452e-05, + "loss": 0.7302, + "step": 44 + }, + { + "epoch": 0.08840864440078586, + "grad_norm": 0.350849163760616, + "learning_rate": 1.4117647058823532e-05, + "loss": 0.7424, + "step": 45 + }, + { + "epoch": 0.09037328094302555, + "grad_norm": 0.3729061289137647, + "learning_rate": 1.443137254901961e-05, + "loss": 0.7375, + "step": 46 + }, + { + "epoch": 0.09233791748526522, + "grad_norm": 0.2967871748217548, + "learning_rate": 1.474509803921569e-05, + "loss": 0.7248, + "step": 47 + }, + { + "epoch": 0.09430255402750491, + "grad_norm": 0.2695581592464476, + "learning_rate": 1.5058823529411765e-05, + "loss": 0.7189, + "step": 48 + }, + { + "epoch": 0.0962671905697446, + "grad_norm": 0.34055067505364056, + "learning_rate": 1.5372549019607847e-05, + "loss": 0.7237, + "step": 49 + }, + { + "epoch": 0.09823182711198428, + "grad_norm": 0.2635476399612622, + "learning_rate": 1.568627450980392e-05, + "loss": 0.7391, + "step": 50 + }, + { + "epoch": 0.10019646365422397, + "grad_norm": 0.3524295524416384, + "learning_rate": 1.6000000000000003e-05, + "loss": 0.7316, + "step": 51 + }, + { + "epoch": 0.10216110019646366, + "grad_norm": 0.2779987943701492, + "learning_rate": 1.631372549019608e-05, + "loss": 0.7182, + "step": 52 + }, + { + "epoch": 0.10412573673870335, + "grad_norm": 0.2780717046226636, + "learning_rate": 1.662745098039216e-05, + "loss": 0.7173, + "step": 53 + }, + { + "epoch": 0.10609037328094302, + "grad_norm": 0.3234185203688851, + "learning_rate": 1.6941176470588237e-05, + "loss": 0.7271, + "step": 54 + }, + { + "epoch": 0.10805500982318271, + "grad_norm": 0.2799717259961317, + "learning_rate": 1.7254901960784314e-05, + "loss": 0.7352, + "step": 55 + }, + { + "epoch": 0.1100196463654224, + "grad_norm": 0.3532087997733881, + "learning_rate": 1.7568627450980392e-05, + "loss": 0.7187, + "step": 56 + }, + { + "epoch": 0.11198428290766209, + "grad_norm": 0.3674617892079078, + "learning_rate": 1.7882352941176474e-05, + "loss": 0.7111, + "step": 57 + }, + { + "epoch": 0.11394891944990176, + "grad_norm": 0.3430241348388222, + "learning_rate": 1.8196078431372552e-05, + "loss": 0.7151, + "step": 58 + }, + { + "epoch": 0.11591355599214145, + "grad_norm": 0.34469529778261254, + "learning_rate": 1.850980392156863e-05, + "loss": 0.7259, + "step": 59 + }, + { + "epoch": 0.11787819253438114, + "grad_norm": 0.3113445162497525, + "learning_rate": 1.8823529411764708e-05, + "loss": 0.7167, + "step": 60 + }, + { + "epoch": 0.11984282907662082, + "grad_norm": 0.3485091162160714, + "learning_rate": 1.9137254901960786e-05, + "loss": 0.7089, + "step": 61 + }, + { + "epoch": 0.12180746561886051, + "grad_norm": 0.33177946270095493, + "learning_rate": 1.9450980392156864e-05, + "loss": 0.7053, + "step": 62 + }, + { + "epoch": 0.1237721021611002, + "grad_norm": 0.28696459257839224, + "learning_rate": 1.9764705882352945e-05, + "loss": 0.7131, + "step": 63 + }, + { + "epoch": 0.12573673870333987, + "grad_norm": 0.4248576450672093, + "learning_rate": 2.007843137254902e-05, + "loss": 0.7067, + "step": 64 + }, + { + "epoch": 0.12770137524557956, + "grad_norm": 0.5582906087275581, + "learning_rate": 2.0392156862745097e-05, + "loss": 0.7053, + "step": 65 + }, + { + "epoch": 0.12966601178781925, + "grad_norm": 0.6763657391747366, + "learning_rate": 2.070588235294118e-05, + "loss": 0.7054, + "step": 66 + }, + { + "epoch": 0.13163064833005894, + "grad_norm": 1.0027516538627979, + "learning_rate": 2.1019607843137257e-05, + "loss": 0.7044, + "step": 67 + }, + { + "epoch": 0.13359528487229863, + "grad_norm": 1.0547440713213514, + "learning_rate": 2.1333333333333335e-05, + "loss": 0.7125, + "step": 68 + }, + { + "epoch": 0.13555992141453832, + "grad_norm": 0.8450934961660574, + "learning_rate": 2.1647058823529413e-05, + "loss": 0.718, + "step": 69 + }, + { + "epoch": 0.137524557956778, + "grad_norm": 1.858559206505616, + "learning_rate": 2.1960784313725494e-05, + "loss": 0.6989, + "step": 70 + }, + { + "epoch": 0.13948919449901767, + "grad_norm": 0.8345932311981952, + "learning_rate": 2.2274509803921572e-05, + "loss": 0.7033, + "step": 71 + }, + { + "epoch": 0.14145383104125736, + "grad_norm": 1.7022486151530478, + "learning_rate": 2.258823529411765e-05, + "loss": 0.7163, + "step": 72 + }, + { + "epoch": 0.14341846758349705, + "grad_norm": 1.311302503598193, + "learning_rate": 2.2901960784313724e-05, + "loss": 0.7122, + "step": 73 + }, + { + "epoch": 0.14538310412573674, + "grad_norm": 1.545293597474234, + "learning_rate": 2.321568627450981e-05, + "loss": 0.716, + "step": 74 + }, + { + "epoch": 0.14734774066797643, + "grad_norm": 1.31327401729448, + "learning_rate": 2.3529411764705884e-05, + "loss": 0.7067, + "step": 75 + }, + { + "epoch": 0.14931237721021612, + "grad_norm": 1.5764791246661156, + "learning_rate": 2.384313725490196e-05, + "loss": 0.7091, + "step": 76 + }, + { + "epoch": 0.1512770137524558, + "grad_norm": 1.3707317868306823, + "learning_rate": 2.415686274509804e-05, + "loss": 0.7054, + "step": 77 + }, + { + "epoch": 0.15324165029469547, + "grad_norm": 1.3897710220416757, + "learning_rate": 2.447058823529412e-05, + "loss": 0.7026, + "step": 78 + }, + { + "epoch": 0.15520628683693516, + "grad_norm": 1.3073683599987704, + "learning_rate": 2.47843137254902e-05, + "loss": 0.6992, + "step": 79 + }, + { + "epoch": 0.15717092337917485, + "grad_norm": 1.141552207993005, + "learning_rate": 2.5098039215686277e-05, + "loss": 0.6986, + "step": 80 + }, + { + "epoch": 0.15913555992141454, + "grad_norm": 1.5874210056024305, + "learning_rate": 2.5411764705882355e-05, + "loss": 0.687, + "step": 81 + }, + { + "epoch": 0.16110019646365423, + "grad_norm": 1.2842403271145792, + "learning_rate": 2.5725490196078436e-05, + "loss": 0.6907, + "step": 82 + }, + { + "epoch": 0.16306483300589392, + "grad_norm": 1.0288946070098557, + "learning_rate": 2.6039215686274514e-05, + "loss": 0.7104, + "step": 83 + }, + { + "epoch": 0.1650294695481336, + "grad_norm": 1.2509062737294774, + "learning_rate": 2.635294117647059e-05, + "loss": 0.7021, + "step": 84 + }, + { + "epoch": 0.16699410609037327, + "grad_norm": 0.6048284785605917, + "learning_rate": 2.6666666666666667e-05, + "loss": 0.6859, + "step": 85 + }, + { + "epoch": 0.16895874263261296, + "grad_norm": 1.271054301140082, + "learning_rate": 2.6980392156862748e-05, + "loss": 0.6824, + "step": 86 + }, + { + "epoch": 0.17092337917485265, + "grad_norm": 0.8541634882893402, + "learning_rate": 2.7294117647058826e-05, + "loss": 0.6951, + "step": 87 + }, + { + "epoch": 0.17288801571709234, + "grad_norm": 1.3169050911337303, + "learning_rate": 2.7607843137254904e-05, + "loss": 0.6856, + "step": 88 + }, + { + "epoch": 0.17485265225933203, + "grad_norm": 0.7749450714340149, + "learning_rate": 2.7921568627450982e-05, + "loss": 0.6827, + "step": 89 + }, + { + "epoch": 0.17681728880157171, + "grad_norm": 1.4434569911539983, + "learning_rate": 2.8235294117647063e-05, + "loss": 0.6873, + "step": 90 + }, + { + "epoch": 0.1787819253438114, + "grad_norm": 0.8488573268340268, + "learning_rate": 2.854901960784314e-05, + "loss": 0.6848, + "step": 91 + }, + { + "epoch": 0.1807465618860511, + "grad_norm": 1.5783095991394003, + "learning_rate": 2.886274509803922e-05, + "loss": 0.7005, + "step": 92 + }, + { + "epoch": 0.18271119842829076, + "grad_norm": 1.254680592828892, + "learning_rate": 2.9176470588235294e-05, + "loss": 0.6982, + "step": 93 + }, + { + "epoch": 0.18467583497053044, + "grad_norm": 1.3344143504222097, + "learning_rate": 2.949019607843138e-05, + "loss": 0.6834, + "step": 94 + }, + { + "epoch": 0.18664047151277013, + "grad_norm": 1.3519405630178745, + "learning_rate": 2.9803921568627453e-05, + "loss": 0.6918, + "step": 95 + }, + { + "epoch": 0.18860510805500982, + "grad_norm": 0.9387557970637167, + "learning_rate": 3.011764705882353e-05, + "loss": 0.684, + "step": 96 + }, + { + "epoch": 0.1905697445972495, + "grad_norm": 1.2161481030310664, + "learning_rate": 3.043137254901961e-05, + "loss": 0.6812, + "step": 97 + }, + { + "epoch": 0.1925343811394892, + "grad_norm": 0.8114909959812774, + "learning_rate": 3.0745098039215694e-05, + "loss": 0.689, + "step": 98 + }, + { + "epoch": 0.1944990176817289, + "grad_norm": 0.8719705565862638, + "learning_rate": 3.1058823529411765e-05, + "loss": 0.6853, + "step": 99 + }, + { + "epoch": 0.19646365422396855, + "grad_norm": 0.8831102114281192, + "learning_rate": 3.137254901960784e-05, + "loss": 0.6865, + "step": 100 + }, + { + "epoch": 0.19842829076620824, + "grad_norm": 1.1716387819876024, + "learning_rate": 3.168627450980392e-05, + "loss": 0.6824, + "step": 101 + }, + { + "epoch": 0.20039292730844793, + "grad_norm": 1.2796587875334364, + "learning_rate": 3.2000000000000005e-05, + "loss": 0.6832, + "step": 102 + }, + { + "epoch": 0.20235756385068762, + "grad_norm": 0.8240368447983127, + "learning_rate": 3.231372549019608e-05, + "loss": 0.6879, + "step": 103 + }, + { + "epoch": 0.2043222003929273, + "grad_norm": 0.789023602587825, + "learning_rate": 3.262745098039216e-05, + "loss": 0.6845, + "step": 104 + }, + { + "epoch": 0.206286836935167, + "grad_norm": 1.3968317866498383, + "learning_rate": 3.294117647058824e-05, + "loss": 0.6824, + "step": 105 + }, + { + "epoch": 0.2082514734774067, + "grad_norm": 1.015408899330511, + "learning_rate": 3.325490196078432e-05, + "loss": 0.6819, + "step": 106 + }, + { + "epoch": 0.21021611001964635, + "grad_norm": 1.1285066282167666, + "learning_rate": 3.3568627450980395e-05, + "loss": 0.6887, + "step": 107 + }, + { + "epoch": 0.21218074656188604, + "grad_norm": 1.059990561953594, + "learning_rate": 3.388235294117647e-05, + "loss": 0.6869, + "step": 108 + }, + { + "epoch": 0.21414538310412573, + "grad_norm": 1.1782195197526757, + "learning_rate": 3.419607843137255e-05, + "loss": 0.6777, + "step": 109 + }, + { + "epoch": 0.21611001964636542, + "grad_norm": 0.7542214921114484, + "learning_rate": 3.450980392156863e-05, + "loss": 0.6795, + "step": 110 + }, + { + "epoch": 0.2180746561886051, + "grad_norm": 0.649593162526091, + "learning_rate": 3.482352941176471e-05, + "loss": 0.6726, + "step": 111 + }, + { + "epoch": 0.2200392927308448, + "grad_norm": 0.700460681861974, + "learning_rate": 3.5137254901960785e-05, + "loss": 0.6809, + "step": 112 + }, + { + "epoch": 0.2220039292730845, + "grad_norm": 0.7906980099156672, + "learning_rate": 3.545098039215686e-05, + "loss": 0.683, + "step": 113 + }, + { + "epoch": 0.22396856581532418, + "grad_norm": 1.0336340681417664, + "learning_rate": 3.576470588235295e-05, + "loss": 0.682, + "step": 114 + }, + { + "epoch": 0.22593320235756384, + "grad_norm": 1.695571832761537, + "learning_rate": 3.6078431372549025e-05, + "loss": 0.6867, + "step": 115 + }, + { + "epoch": 0.22789783889980353, + "grad_norm": 0.654559949642839, + "learning_rate": 3.6392156862745103e-05, + "loss": 0.6824, + "step": 116 + }, + { + "epoch": 0.22986247544204322, + "grad_norm": 2.1130331776539393, + "learning_rate": 3.6705882352941175e-05, + "loss": 0.6803, + "step": 117 + }, + { + "epoch": 0.2318271119842829, + "grad_norm": 0.8706075318501634, + "learning_rate": 3.701960784313726e-05, + "loss": 0.6737, + "step": 118 + }, + { + "epoch": 0.2337917485265226, + "grad_norm": 2.635446293816418, + "learning_rate": 3.733333333333334e-05, + "loss": 0.6942, + "step": 119 + }, + { + "epoch": 0.2357563850687623, + "grad_norm": 2.340387537162468, + "learning_rate": 3.7647058823529415e-05, + "loss": 0.7044, + "step": 120 + }, + { + "epoch": 0.23772102161100198, + "grad_norm": 1.2580523953666853, + "learning_rate": 3.796078431372549e-05, + "loss": 0.672, + "step": 121 + }, + { + "epoch": 0.23968565815324164, + "grad_norm": 1.9131703681133432, + "learning_rate": 3.827450980392157e-05, + "loss": 0.6886, + "step": 122 + }, + { + "epoch": 0.24165029469548133, + "grad_norm": 1.3751215078942378, + "learning_rate": 3.858823529411765e-05, + "loss": 0.6824, + "step": 123 + }, + { + "epoch": 0.24361493123772102, + "grad_norm": 1.866539386809526, + "learning_rate": 3.890196078431373e-05, + "loss": 0.688, + "step": 124 + }, + { + "epoch": 0.2455795677799607, + "grad_norm": 1.5767508925427318, + "learning_rate": 3.9215686274509805e-05, + "loss": 0.6788, + "step": 125 + }, + { + "epoch": 0.2475442043222004, + "grad_norm": 1.5291169953275543, + "learning_rate": 3.952941176470589e-05, + "loss": 0.6797, + "step": 126 + }, + { + "epoch": 0.24950884086444008, + "grad_norm": 1.388986294678716, + "learning_rate": 3.984313725490197e-05, + "loss": 0.6893, + "step": 127 + }, + { + "epoch": 0.25147347740667975, + "grad_norm": 1.0484457589742635, + "learning_rate": 4.015686274509804e-05, + "loss": 0.6731, + "step": 128 + }, + { + "epoch": 0.25343811394891946, + "grad_norm": 1.56748627633668, + "learning_rate": 4.047058823529412e-05, + "loss": 0.6938, + "step": 129 + }, + { + "epoch": 0.2554027504911591, + "grad_norm": 1.1102110011201525, + "learning_rate": 4.0784313725490195e-05, + "loss": 0.676, + "step": 130 + }, + { + "epoch": 0.25736738703339884, + "grad_norm": 1.2033834376983161, + "learning_rate": 4.109803921568627e-05, + "loss": 0.6743, + "step": 131 + }, + { + "epoch": 0.2593320235756385, + "grad_norm": 1.587875581988948, + "learning_rate": 4.141176470588236e-05, + "loss": 0.6822, + "step": 132 + }, + { + "epoch": 0.26129666011787817, + "grad_norm": 0.9414343148216243, + "learning_rate": 4.1725490196078435e-05, + "loss": 0.6804, + "step": 133 + }, + { + "epoch": 0.2632612966601179, + "grad_norm": 1.36506403534512, + "learning_rate": 4.203921568627451e-05, + "loss": 0.6767, + "step": 134 + }, + { + "epoch": 0.26522593320235754, + "grad_norm": 1.2085054296850877, + "learning_rate": 4.235294117647059e-05, + "loss": 0.6769, + "step": 135 + }, + { + "epoch": 0.26719056974459726, + "grad_norm": 0.9696925439118442, + "learning_rate": 4.266666666666667e-05, + "loss": 0.6761, + "step": 136 + }, + { + "epoch": 0.2691552062868369, + "grad_norm": 0.8865323510730617, + "learning_rate": 4.298039215686275e-05, + "loss": 0.6791, + "step": 137 + }, + { + "epoch": 0.27111984282907664, + "grad_norm": 1.2549162937731584, + "learning_rate": 4.3294117647058825e-05, + "loss": 0.6679, + "step": 138 + }, + { + "epoch": 0.2730844793713163, + "grad_norm": 1.0821006863106035, + "learning_rate": 4.36078431372549e-05, + "loss": 0.6737, + "step": 139 + }, + { + "epoch": 0.275049115913556, + "grad_norm": 0.7982305726536242, + "learning_rate": 4.392156862745099e-05, + "loss": 0.6564, + "step": 140 + }, + { + "epoch": 0.2770137524557957, + "grad_norm": 1.053226719315009, + "learning_rate": 4.4235294117647066e-05, + "loss": 0.6645, + "step": 141 + }, + { + "epoch": 0.27897838899803534, + "grad_norm": 1.6789776340579865, + "learning_rate": 4.4549019607843144e-05, + "loss": 0.6765, + "step": 142 + }, + { + "epoch": 0.28094302554027506, + "grad_norm": 1.1067630174217862, + "learning_rate": 4.486274509803922e-05, + "loss": 0.6802, + "step": 143 + }, + { + "epoch": 0.2829076620825147, + "grad_norm": 1.3334393567629226, + "learning_rate": 4.51764705882353e-05, + "loss": 0.6797, + "step": 144 + }, + { + "epoch": 0.28487229862475444, + "grad_norm": 1.0093968055115998, + "learning_rate": 4.549019607843138e-05, + "loss": 0.6635, + "step": 145 + }, + { + "epoch": 0.2868369351669941, + "grad_norm": 1.658090083086876, + "learning_rate": 4.580392156862745e-05, + "loss": 0.669, + "step": 146 + }, + { + "epoch": 0.2888015717092338, + "grad_norm": 0.9810917756230719, + "learning_rate": 4.611764705882353e-05, + "loss": 0.676, + "step": 147 + }, + { + "epoch": 0.2907662082514735, + "grad_norm": 1.9520593150401058, + "learning_rate": 4.643137254901962e-05, + "loss": 0.6879, + "step": 148 + }, + { + "epoch": 0.29273084479371314, + "grad_norm": 1.528343886605023, + "learning_rate": 4.674509803921569e-05, + "loss": 0.6653, + "step": 149 + }, + { + "epoch": 0.29469548133595286, + "grad_norm": 1.5130928041991847, + "learning_rate": 4.705882352941177e-05, + "loss": 0.6838, + "step": 150 + }, + { + "epoch": 0.2966601178781925, + "grad_norm": 1.6056326906742004, + "learning_rate": 4.7372549019607845e-05, + "loss": 0.664, + "step": 151 + }, + { + "epoch": 0.29862475442043224, + "grad_norm": 1.1629795553361977, + "learning_rate": 4.768627450980392e-05, + "loss": 0.6584, + "step": 152 + }, + { + "epoch": 0.3005893909626719, + "grad_norm": 1.9688588401237872, + "learning_rate": 4.8e-05, + "loss": 0.6675, + "step": 153 + }, + { + "epoch": 0.3025540275049116, + "grad_norm": 1.1655358090375727, + "learning_rate": 4.831372549019608e-05, + "loss": 0.6655, + "step": 154 + }, + { + "epoch": 0.3045186640471513, + "grad_norm": 2.4196436297745674, + "learning_rate": 4.862745098039216e-05, + "loss": 0.6745, + "step": 155 + }, + { + "epoch": 0.30648330058939094, + "grad_norm": 2.1852201298241836, + "learning_rate": 4.894117647058824e-05, + "loss": 0.6711, + "step": 156 + }, + { + "epoch": 0.30844793713163066, + "grad_norm": 1.468557365569968, + "learning_rate": 4.925490196078432e-05, + "loss": 0.6635, + "step": 157 + }, + { + "epoch": 0.3104125736738703, + "grad_norm": 1.326031980953515, + "learning_rate": 4.95686274509804e-05, + "loss": 0.6757, + "step": 158 + }, + { + "epoch": 0.31237721021611004, + "grad_norm": 1.3907348356761473, + "learning_rate": 4.9882352941176476e-05, + "loss": 0.662, + "step": 159 + }, + { + "epoch": 0.3143418467583497, + "grad_norm": 1.1070413827222736, + "learning_rate": 5.0196078431372554e-05, + "loss": 0.6616, + "step": 160 + }, + { + "epoch": 0.3163064833005894, + "grad_norm": 1.5851483097376557, + "learning_rate": 5.050980392156863e-05, + "loss": 0.6766, + "step": 161 + }, + { + "epoch": 0.3182711198428291, + "grad_norm": 1.3104619586195827, + "learning_rate": 5.082352941176471e-05, + "loss": 0.6742, + "step": 162 + }, + { + "epoch": 0.32023575638506874, + "grad_norm": 1.3394831607880449, + "learning_rate": 5.113725490196079e-05, + "loss": 0.6643, + "step": 163 + }, + { + "epoch": 0.32220039292730845, + "grad_norm": 1.5126215240877283, + "learning_rate": 5.145098039215687e-05, + "loss": 0.6755, + "step": 164 + }, + { + "epoch": 0.3241650294695481, + "grad_norm": 1.200559255269103, + "learning_rate": 5.176470588235295e-05, + "loss": 0.662, + "step": 165 + }, + { + "epoch": 0.32612966601178783, + "grad_norm": 1.4359466444458304, + "learning_rate": 5.207843137254903e-05, + "loss": 0.662, + "step": 166 + }, + { + "epoch": 0.3280943025540275, + "grad_norm": 1.1064266293223728, + "learning_rate": 5.23921568627451e-05, + "loss": 0.66, + "step": 167 + }, + { + "epoch": 0.3300589390962672, + "grad_norm": 1.6834622755319286, + "learning_rate": 5.270588235294118e-05, + "loss": 0.6478, + "step": 168 + }, + { + "epoch": 0.3320235756385069, + "grad_norm": 1.1148645493162137, + "learning_rate": 5.3019607843137255e-05, + "loss": 0.6621, + "step": 169 + }, + { + "epoch": 0.33398821218074654, + "grad_norm": 1.6733301128617024, + "learning_rate": 5.333333333333333e-05, + "loss": 0.6471, + "step": 170 + }, + { + "epoch": 0.33595284872298625, + "grad_norm": 1.33511245816292, + "learning_rate": 5.364705882352941e-05, + "loss": 0.6668, + "step": 171 + }, + { + "epoch": 0.3379174852652259, + "grad_norm": 1.3952035224197956, + "learning_rate": 5.3960784313725496e-05, + "loss": 0.6563, + "step": 172 + }, + { + "epoch": 0.33988212180746563, + "grad_norm": 1.082475037782671, + "learning_rate": 5.4274509803921574e-05, + "loss": 0.6509, + "step": 173 + }, + { + "epoch": 0.3418467583497053, + "grad_norm": 1.3094718258445648, + "learning_rate": 5.458823529411765e-05, + "loss": 0.6647, + "step": 174 + }, + { + "epoch": 0.343811394891945, + "grad_norm": 1.081237982103074, + "learning_rate": 5.490196078431373e-05, + "loss": 0.6618, + "step": 175 + }, + { + "epoch": 0.34577603143418467, + "grad_norm": 1.4283088566180793, + "learning_rate": 5.521568627450981e-05, + "loss": 0.6619, + "step": 176 + }, + { + "epoch": 0.3477406679764244, + "grad_norm": 0.9860382809070741, + "learning_rate": 5.5529411764705886e-05, + "loss": 0.6721, + "step": 177 + }, + { + "epoch": 0.34970530451866405, + "grad_norm": 1.2542912979845968, + "learning_rate": 5.5843137254901964e-05, + "loss": 0.6565, + "step": 178 + }, + { + "epoch": 0.3516699410609037, + "grad_norm": 0.876795442006041, + "learning_rate": 5.615686274509804e-05, + "loss": 0.6519, + "step": 179 + }, + { + "epoch": 0.35363457760314343, + "grad_norm": 1.5053270200093063, + "learning_rate": 5.6470588235294126e-05, + "loss": 0.6507, + "step": 180 + }, + { + "epoch": 0.3555992141453831, + "grad_norm": 1.4026768196772805, + "learning_rate": 5.6784313725490204e-05, + "loss": 0.6748, + "step": 181 + }, + { + "epoch": 0.3575638506876228, + "grad_norm": 0.9570680418640009, + "learning_rate": 5.709803921568628e-05, + "loss": 0.6484, + "step": 182 + }, + { + "epoch": 0.35952848722986247, + "grad_norm": 1.8461984560629905, + "learning_rate": 5.741176470588236e-05, + "loss": 0.6561, + "step": 183 + }, + { + "epoch": 0.3614931237721022, + "grad_norm": 0.8386441676181124, + "learning_rate": 5.772549019607844e-05, + "loss": 0.6484, + "step": 184 + }, + { + "epoch": 0.36345776031434185, + "grad_norm": 1.0874441380112334, + "learning_rate": 5.8039215686274516e-05, + "loss": 0.6601, + "step": 185 + }, + { + "epoch": 0.3654223968565815, + "grad_norm": 1.395349196555838, + "learning_rate": 5.835294117647059e-05, + "loss": 0.6608, + "step": 186 + }, + { + "epoch": 0.36738703339882123, + "grad_norm": 1.2499070415897184, + "learning_rate": 5.8666666666666665e-05, + "loss": 0.6619, + "step": 187 + }, + { + "epoch": 0.3693516699410609, + "grad_norm": 1.2641588719329722, + "learning_rate": 5.898039215686276e-05, + "loss": 0.6654, + "step": 188 + }, + { + "epoch": 0.3713163064833006, + "grad_norm": 1.3931031746891431, + "learning_rate": 5.929411764705883e-05, + "loss": 0.647, + "step": 189 + }, + { + "epoch": 0.37328094302554027, + "grad_norm": 1.1342902602148306, + "learning_rate": 5.9607843137254906e-05, + "loss": 0.6541, + "step": 190 + }, + { + "epoch": 0.37524557956778, + "grad_norm": 1.0556357222860804, + "learning_rate": 5.9921568627450984e-05, + "loss": 0.6511, + "step": 191 + }, + { + "epoch": 0.37721021611001965, + "grad_norm": 1.7381759503074103, + "learning_rate": 6.023529411764706e-05, + "loss": 0.6582, + "step": 192 + }, + { + "epoch": 0.3791748526522593, + "grad_norm": 1.0123319134041904, + "learning_rate": 6.054901960784314e-05, + "loss": 0.6565, + "step": 193 + }, + { + "epoch": 0.381139489194499, + "grad_norm": 1.3059861036293752, + "learning_rate": 6.086274509803922e-05, + "loss": 0.6612, + "step": 194 + }, + { + "epoch": 0.3831041257367387, + "grad_norm": 1.0046640180230089, + "learning_rate": 6.117647058823529e-05, + "loss": 0.6508, + "step": 195 + }, + { + "epoch": 0.3850687622789784, + "grad_norm": 1.6557261723219867, + "learning_rate": 6.149019607843139e-05, + "loss": 0.6565, + "step": 196 + }, + { + "epoch": 0.38703339882121807, + "grad_norm": 1.0597075913050573, + "learning_rate": 6.180392156862746e-05, + "loss": 0.65, + "step": 197 + }, + { + "epoch": 0.3889980353634578, + "grad_norm": 1.5579486189461038, + "learning_rate": 6.211764705882353e-05, + "loss": 0.6472, + "step": 198 + }, + { + "epoch": 0.39096267190569745, + "grad_norm": 1.1577128380558905, + "learning_rate": 6.243137254901961e-05, + "loss": 0.6586, + "step": 199 + }, + { + "epoch": 0.3929273084479371, + "grad_norm": 1.4112734351749578, + "learning_rate": 6.274509803921569e-05, + "loss": 0.6547, + "step": 200 + }, + { + "epoch": 0.3948919449901768, + "grad_norm": 1.0046153992504252, + "learning_rate": 6.305882352941177e-05, + "loss": 0.6567, + "step": 201 + }, + { + "epoch": 0.3968565815324165, + "grad_norm": 1.2199161802240173, + "learning_rate": 6.337254901960784e-05, + "loss": 0.6505, + "step": 202 + }, + { + "epoch": 0.3988212180746562, + "grad_norm": 0.8595083156795053, + "learning_rate": 6.368627450980393e-05, + "loss": 0.6294, + "step": 203 + }, + { + "epoch": 0.40078585461689586, + "grad_norm": 0.8671497482956964, + "learning_rate": 6.400000000000001e-05, + "loss": 0.6556, + "step": 204 + }, + { + "epoch": 0.4027504911591356, + "grad_norm": 1.2186747556691497, + "learning_rate": 6.431372549019608e-05, + "loss": 0.6613, + "step": 205 + }, + { + "epoch": 0.40471512770137524, + "grad_norm": 1.1482659435400167, + "learning_rate": 6.462745098039217e-05, + "loss": 0.6589, + "step": 206 + }, + { + "epoch": 0.4066797642436149, + "grad_norm": 1.2226004626093514, + "learning_rate": 6.494117647058824e-05, + "loss": 0.6551, + "step": 207 + }, + { + "epoch": 0.4086444007858546, + "grad_norm": 2.399319737661628, + "learning_rate": 6.525490196078432e-05, + "loss": 0.6527, + "step": 208 + }, + { + "epoch": 0.4106090373280943, + "grad_norm": 1.342168665483398, + "learning_rate": 6.55686274509804e-05, + "loss": 0.6557, + "step": 209 + }, + { + "epoch": 0.412573673870334, + "grad_norm": 3.072369588363727, + "learning_rate": 6.588235294117648e-05, + "loss": 0.6667, + "step": 210 + }, + { + "epoch": 0.41453831041257366, + "grad_norm": 2.9467080602931546, + "learning_rate": 6.619607843137255e-05, + "loss": 0.6673, + "step": 211 + }, + { + "epoch": 0.4165029469548134, + "grad_norm": 1.4232971596700674, + "learning_rate": 6.650980392156863e-05, + "loss": 0.654, + "step": 212 + }, + { + "epoch": 0.41846758349705304, + "grad_norm": 1.8683235713326996, + "learning_rate": 6.682352941176472e-05, + "loss": 0.6635, + "step": 213 + }, + { + "epoch": 0.4204322200392927, + "grad_norm": 1.3145258735954353, + "learning_rate": 6.713725490196079e-05, + "loss": 0.655, + "step": 214 + }, + { + "epoch": 0.4223968565815324, + "grad_norm": 2.15269815596423, + "learning_rate": 6.745098039215687e-05, + "loss": 0.6688, + "step": 215 + }, + { + "epoch": 0.4243614931237721, + "grad_norm": 1.3672790109799209, + "learning_rate": 6.776470588235295e-05, + "loss": 0.6578, + "step": 216 + }, + { + "epoch": 0.4263261296660118, + "grad_norm": 1.9567377674980009, + "learning_rate": 6.807843137254902e-05, + "loss": 0.6662, + "step": 217 + }, + { + "epoch": 0.42829076620825146, + "grad_norm": 1.694907564188827, + "learning_rate": 6.83921568627451e-05, + "loss": 0.6697, + "step": 218 + }, + { + "epoch": 0.4302554027504912, + "grad_norm": 1.3674588484172614, + "learning_rate": 6.870588235294117e-05, + "loss": 0.656, + "step": 219 + }, + { + "epoch": 0.43222003929273084, + "grad_norm": 1.797967980986462, + "learning_rate": 6.901960784313726e-05, + "loss": 0.6644, + "step": 220 + }, + { + "epoch": 0.43418467583497056, + "grad_norm": 1.1066613561580902, + "learning_rate": 6.933333333333334e-05, + "loss": 0.6577, + "step": 221 + }, + { + "epoch": 0.4361493123772102, + "grad_norm": 1.874577327301109, + "learning_rate": 6.964705882352941e-05, + "loss": 0.6592, + "step": 222 + }, + { + "epoch": 0.4381139489194499, + "grad_norm": 1.4534400186686673, + "learning_rate": 6.99607843137255e-05, + "loss": 0.6496, + "step": 223 + }, + { + "epoch": 0.4400785854616896, + "grad_norm": 1.4467298118618857, + "learning_rate": 7.027450980392157e-05, + "loss": 0.6583, + "step": 224 + }, + { + "epoch": 0.44204322200392926, + "grad_norm": 1.537203279348578, + "learning_rate": 7.058823529411765e-05, + "loss": 0.6542, + "step": 225 + }, + { + "epoch": 0.444007858546169, + "grad_norm": 1.0563373909946343, + "learning_rate": 7.090196078431373e-05, + "loss": 0.6601, + "step": 226 + }, + { + "epoch": 0.44597249508840864, + "grad_norm": 1.6427269755028262, + "learning_rate": 7.121568627450981e-05, + "loss": 0.6506, + "step": 227 + }, + { + "epoch": 0.44793713163064836, + "grad_norm": 1.116409595044875, + "learning_rate": 7.15294117647059e-05, + "loss": 0.6471, + "step": 228 + }, + { + "epoch": 0.449901768172888, + "grad_norm": 1.8623999089137084, + "learning_rate": 7.184313725490197e-05, + "loss": 0.6541, + "step": 229 + }, + { + "epoch": 0.4518664047151277, + "grad_norm": 1.288967329125417, + "learning_rate": 7.215686274509805e-05, + "loss": 0.6559, + "step": 230 + }, + { + "epoch": 0.4538310412573674, + "grad_norm": 1.8971055541836412, + "learning_rate": 7.247058823529412e-05, + "loss": 0.6503, + "step": 231 + }, + { + "epoch": 0.45579567779960706, + "grad_norm": 1.2143347300682938, + "learning_rate": 7.278431372549021e-05, + "loss": 0.6613, + "step": 232 + }, + { + "epoch": 0.4577603143418468, + "grad_norm": 2.260415062080299, + "learning_rate": 7.309803921568628e-05, + "loss": 0.6638, + "step": 233 + }, + { + "epoch": 0.45972495088408644, + "grad_norm": 1.9868028526688868, + "learning_rate": 7.341176470588235e-05, + "loss": 0.6682, + "step": 234 + }, + { + "epoch": 0.46168958742632615, + "grad_norm": 1.1933133775380198, + "learning_rate": 7.372549019607843e-05, + "loss": 0.6591, + "step": 235 + }, + { + "epoch": 0.4636542239685658, + "grad_norm": 1.2085981292810568, + "learning_rate": 7.403921568627452e-05, + "loss": 0.6526, + "step": 236 + }, + { + "epoch": 0.4656188605108055, + "grad_norm": 1.4297751160804568, + "learning_rate": 7.435294117647059e-05, + "loss": 0.6416, + "step": 237 + }, + { + "epoch": 0.4675834970530452, + "grad_norm": 0.8916912053095457, + "learning_rate": 7.466666666666667e-05, + "loss": 0.6526, + "step": 238 + }, + { + "epoch": 0.46954813359528486, + "grad_norm": 1.5503879348656397, + "learning_rate": 7.498039215686275e-05, + "loss": 0.6586, + "step": 239 + }, + { + "epoch": 0.4715127701375246, + "grad_norm": 1.2816204759476209, + "learning_rate": 7.529411764705883e-05, + "loss": 0.664, + "step": 240 + }, + { + "epoch": 0.47347740667976423, + "grad_norm": 1.374602808705368, + "learning_rate": 7.56078431372549e-05, + "loss": 0.6445, + "step": 241 + }, + { + "epoch": 0.47544204322200395, + "grad_norm": 1.1003808080440216, + "learning_rate": 7.592156862745099e-05, + "loss": 0.6547, + "step": 242 + }, + { + "epoch": 0.4774066797642436, + "grad_norm": 1.244814104244874, + "learning_rate": 7.623529411764706e-05, + "loss": 0.6506, + "step": 243 + }, + { + "epoch": 0.4793713163064833, + "grad_norm": 1.14299645855317, + "learning_rate": 7.654901960784314e-05, + "loss": 0.658, + "step": 244 + }, + { + "epoch": 0.481335952848723, + "grad_norm": 1.2170312467722815, + "learning_rate": 7.686274509803923e-05, + "loss": 0.6449, + "step": 245 + }, + { + "epoch": 0.48330058939096265, + "grad_norm": 1.4994418857268166, + "learning_rate": 7.71764705882353e-05, + "loss": 0.6513, + "step": 246 + }, + { + "epoch": 0.48526522593320237, + "grad_norm": 0.8814605775173949, + "learning_rate": 7.749019607843138e-05, + "loss": 0.651, + "step": 247 + }, + { + "epoch": 0.48722986247544203, + "grad_norm": 0.8428837776136511, + "learning_rate": 7.780392156862745e-05, + "loss": 0.6407, + "step": 248 + }, + { + "epoch": 0.48919449901768175, + "grad_norm": 1.3141246424700066, + "learning_rate": 7.811764705882354e-05, + "loss": 0.661, + "step": 249 + }, + { + "epoch": 0.4911591355599214, + "grad_norm": 1.2629127529184854, + "learning_rate": 7.843137254901961e-05, + "loss": 0.6579, + "step": 250 + }, + { + "epoch": 0.4931237721021611, + "grad_norm": 1.2357647878393516, + "learning_rate": 7.87450980392157e-05, + "loss": 0.6498, + "step": 251 + }, + { + "epoch": 0.4950884086444008, + "grad_norm": 1.190497756936977, + "learning_rate": 7.905882352941178e-05, + "loss": 0.6547, + "step": 252 + }, + { + "epoch": 0.49705304518664045, + "grad_norm": 1.1282736374674391, + "learning_rate": 7.937254901960785e-05, + "loss": 0.653, + "step": 253 + }, + { + "epoch": 0.49901768172888017, + "grad_norm": 1.1605041411059538, + "learning_rate": 7.968627450980394e-05, + "loss": 0.6447, + "step": 254 + }, + { + "epoch": 0.5009823182711198, + "grad_norm": 1.3361500732361977, + "learning_rate": 8e-05, + "loss": 0.6396, + "step": 255 + }, + { + "epoch": 0.5029469548133595, + "grad_norm": 0.9648889884058455, + "learning_rate": 7.999996235921187e-05, + "loss": 0.6564, + "step": 256 + }, + { + "epoch": 0.5049115913555993, + "grad_norm": 1.1358463516909887, + "learning_rate": 7.999984943691831e-05, + "loss": 0.6365, + "step": 257 + }, + { + "epoch": 0.5068762278978389, + "grad_norm": 1.0517311750723548, + "learning_rate": 7.999966123333183e-05, + "loss": 0.6447, + "step": 258 + }, + { + "epoch": 0.5088408644400786, + "grad_norm": 1.2512133853729694, + "learning_rate": 7.999939774880666e-05, + "loss": 0.6436, + "step": 259 + }, + { + "epoch": 0.5108055009823183, + "grad_norm": 1.0871570031525792, + "learning_rate": 7.999905898383868e-05, + "loss": 0.648, + "step": 260 + }, + { + "epoch": 0.5127701375245579, + "grad_norm": 0.8439337696994582, + "learning_rate": 7.999864493906545e-05, + "loss": 0.6367, + "step": 261 + }, + { + "epoch": 0.5147347740667977, + "grad_norm": 1.1459601182361547, + "learning_rate": 7.999815561526623e-05, + "loss": 0.651, + "step": 262 + }, + { + "epoch": 0.5166994106090373, + "grad_norm": 1.3109656270225871, + "learning_rate": 7.999759101336195e-05, + "loss": 0.6582, + "step": 263 + }, + { + "epoch": 0.518664047151277, + "grad_norm": 0.67850838537405, + "learning_rate": 7.999695113441519e-05, + "loss": 0.6386, + "step": 264 + }, + { + "epoch": 0.5206286836935167, + "grad_norm": 1.1083173045929193, + "learning_rate": 7.999623597963026e-05, + "loss": 0.6454, + "step": 265 + }, + { + "epoch": 0.5225933202357563, + "grad_norm": 1.270195803974325, + "learning_rate": 7.999544555035308e-05, + "loss": 0.6357, + "step": 266 + }, + { + "epoch": 0.5245579567779961, + "grad_norm": 0.9377152326840448, + "learning_rate": 7.999457984807128e-05, + "loss": 0.6438, + "step": 267 + }, + { + "epoch": 0.5265225933202358, + "grad_norm": 1.4397084976070151, + "learning_rate": 7.999363887441416e-05, + "loss": 0.6606, + "step": 268 + }, + { + "epoch": 0.5284872298624754, + "grad_norm": 0.9253891993890428, + "learning_rate": 7.999262263115267e-05, + "loss": 0.6488, + "step": 269 + }, + { + "epoch": 0.5304518664047151, + "grad_norm": 1.3338017909146653, + "learning_rate": 7.999153112019939e-05, + "loss": 0.6528, + "step": 270 + }, + { + "epoch": 0.5324165029469549, + "grad_norm": 1.100317560626675, + "learning_rate": 7.99903643436086e-05, + "loss": 0.6421, + "step": 271 + }, + { + "epoch": 0.5343811394891945, + "grad_norm": 1.05568496704732, + "learning_rate": 7.998912230357623e-05, + "loss": 0.6434, + "step": 272 + }, + { + "epoch": 0.5363457760314342, + "grad_norm": 1.2172336146731881, + "learning_rate": 7.998780500243983e-05, + "loss": 0.642, + "step": 273 + }, + { + "epoch": 0.5383104125736738, + "grad_norm": 1.1207896010801743, + "learning_rate": 7.998641244267863e-05, + "loss": 0.6438, + "step": 274 + }, + { + "epoch": 0.5402750491159135, + "grad_norm": 1.0882943109913454, + "learning_rate": 7.998494462691349e-05, + "loss": 0.6498, + "step": 275 + }, + { + "epoch": 0.5422396856581533, + "grad_norm": 0.7985568817958177, + "learning_rate": 7.998340155790686e-05, + "loss": 0.6374, + "step": 276 + }, + { + "epoch": 0.5442043222003929, + "grad_norm": 1.0276093004510447, + "learning_rate": 7.99817832385629e-05, + "loss": 0.6314, + "step": 277 + }, + { + "epoch": 0.5461689587426326, + "grad_norm": 1.2634877782831329, + "learning_rate": 7.998008967192732e-05, + "loss": 0.6454, + "step": 278 + }, + { + "epoch": 0.5481335952848723, + "grad_norm": 0.8261585765035339, + "learning_rate": 7.997832086118749e-05, + "loss": 0.6301, + "step": 279 + }, + { + "epoch": 0.550098231827112, + "grad_norm": 1.0131895558991106, + "learning_rate": 7.99764768096724e-05, + "loss": 0.6428, + "step": 280 + }, + { + "epoch": 0.5520628683693517, + "grad_norm": 1.1454036381137593, + "learning_rate": 7.997455752085258e-05, + "loss": 0.6452, + "step": 281 + }, + { + "epoch": 0.5540275049115914, + "grad_norm": 0.8421675518494275, + "learning_rate": 7.997256299834025e-05, + "loss": 0.6499, + "step": 282 + }, + { + "epoch": 0.555992141453831, + "grad_norm": 0.9681908048941582, + "learning_rate": 7.997049324588917e-05, + "loss": 0.6401, + "step": 283 + }, + { + "epoch": 0.5579567779960707, + "grad_norm": 1.4927938997447232, + "learning_rate": 7.996834826739469e-05, + "loss": 0.6456, + "step": 284 + }, + { + "epoch": 0.5599214145383105, + "grad_norm": 0.909841522690755, + "learning_rate": 7.996612806689376e-05, + "loss": 0.6344, + "step": 285 + }, + { + "epoch": 0.5618860510805501, + "grad_norm": 0.8928479910950818, + "learning_rate": 7.996383264856485e-05, + "loss": 0.6361, + "step": 286 + }, + { + "epoch": 0.5638506876227898, + "grad_norm": 1.4988422771740129, + "learning_rate": 7.996146201672806e-05, + "loss": 0.6511, + "step": 287 + }, + { + "epoch": 0.5658153241650294, + "grad_norm": 1.0675960123020134, + "learning_rate": 7.995901617584499e-05, + "loss": 0.6433, + "step": 288 + }, + { + "epoch": 0.5677799607072691, + "grad_norm": 1.3403070783571234, + "learning_rate": 7.995649513051882e-05, + "loss": 0.658, + "step": 289 + }, + { + "epoch": 0.5697445972495089, + "grad_norm": 1.1415128312069553, + "learning_rate": 7.995389888549428e-05, + "loss": 0.644, + "step": 290 + }, + { + "epoch": 0.5717092337917485, + "grad_norm": 1.2051673764447095, + "learning_rate": 7.995122744565757e-05, + "loss": 0.6462, + "step": 291 + }, + { + "epoch": 0.5736738703339882, + "grad_norm": 0.9037611530427039, + "learning_rate": 7.994848081603644e-05, + "loss": 0.6494, + "step": 292 + }, + { + "epoch": 0.5756385068762279, + "grad_norm": 1.1795993289678708, + "learning_rate": 7.99456590018002e-05, + "loss": 0.6344, + "step": 293 + }, + { + "epoch": 0.5776031434184676, + "grad_norm": 0.7377458823349726, + "learning_rate": 7.994276200825958e-05, + "loss": 0.6423, + "step": 294 + }, + { + "epoch": 0.5795677799607073, + "grad_norm": 0.7232954877045098, + "learning_rate": 7.993978984086683e-05, + "loss": 0.6293, + "step": 295 + }, + { + "epoch": 0.581532416502947, + "grad_norm": 0.7302993412563716, + "learning_rate": 7.993674250521572e-05, + "loss": 0.6457, + "step": 296 + }, + { + "epoch": 0.5834970530451866, + "grad_norm": 0.6954532779707634, + "learning_rate": 7.993362000704143e-05, + "loss": 0.6348, + "step": 297 + }, + { + "epoch": 0.5854616895874263, + "grad_norm": 0.8113296544138057, + "learning_rate": 7.993042235222062e-05, + "loss": 0.637, + "step": 298 + }, + { + "epoch": 0.587426326129666, + "grad_norm": 0.9359550210213011, + "learning_rate": 7.992714954677143e-05, + "loss": 0.6341, + "step": 299 + }, + { + "epoch": 0.5893909626719057, + "grad_norm": 1.2538683064638312, + "learning_rate": 7.992380159685338e-05, + "loss": 0.6299, + "step": 300 + }, + { + "epoch": 0.5913555992141454, + "grad_norm": 1.057183941130792, + "learning_rate": 7.992037850876745e-05, + "loss": 0.632, + "step": 301 + }, + { + "epoch": 0.593320235756385, + "grad_norm": 0.9220178676269847, + "learning_rate": 7.991688028895605e-05, + "loss": 0.6344, + "step": 302 + }, + { + "epoch": 0.5952848722986247, + "grad_norm": 0.7636051176013996, + "learning_rate": 7.991330694400293e-05, + "loss": 0.6508, + "step": 303 + }, + { + "epoch": 0.5972495088408645, + "grad_norm": 1.1571842297698174, + "learning_rate": 7.99096584806333e-05, + "loss": 0.6338, + "step": 304 + }, + { + "epoch": 0.5992141453831041, + "grad_norm": 0.95020027582896, + "learning_rate": 7.990593490571369e-05, + "loss": 0.6428, + "step": 305 + }, + { + "epoch": 0.6011787819253438, + "grad_norm": 1.036426523094574, + "learning_rate": 7.990213622625202e-05, + "loss": 0.6319, + "step": 306 + }, + { + "epoch": 0.6031434184675835, + "grad_norm": 1.336991704383065, + "learning_rate": 7.989826244939757e-05, + "loss": 0.6335, + "step": 307 + }, + { + "epoch": 0.6051080550098232, + "grad_norm": 0.7377724233482772, + "learning_rate": 7.989431358244093e-05, + "loss": 0.6313, + "step": 308 + }, + { + "epoch": 0.6070726915520629, + "grad_norm": 0.8005072453490583, + "learning_rate": 7.989028963281401e-05, + "loss": 0.6328, + "step": 309 + }, + { + "epoch": 0.6090373280943026, + "grad_norm": 1.1018298585948136, + "learning_rate": 7.988619060809005e-05, + "loss": 0.6306, + "step": 310 + }, + { + "epoch": 0.6110019646365422, + "grad_norm": 1.0084627012780822, + "learning_rate": 7.988201651598359e-05, + "loss": 0.6461, + "step": 311 + }, + { + "epoch": 0.6129666011787819, + "grad_norm": 0.7618980121008478, + "learning_rate": 7.987776736435042e-05, + "loss": 0.6308, + "step": 312 + }, + { + "epoch": 0.6149312377210217, + "grad_norm": 1.0055333619145546, + "learning_rate": 7.987344316118761e-05, + "loss": 0.624, + "step": 313 + }, + { + "epoch": 0.6168958742632613, + "grad_norm": 1.561047172320785, + "learning_rate": 7.986904391463349e-05, + "loss": 0.6298, + "step": 314 + }, + { + "epoch": 0.618860510805501, + "grad_norm": 0.5389232749508904, + "learning_rate": 7.98645696329676e-05, + "loss": 0.6387, + "step": 315 + }, + { + "epoch": 0.6208251473477406, + "grad_norm": 1.48802502991336, + "learning_rate": 7.986002032461074e-05, + "loss": 0.6289, + "step": 316 + }, + { + "epoch": 0.6227897838899804, + "grad_norm": 0.8425455043175256, + "learning_rate": 7.985539599812488e-05, + "loss": 0.6328, + "step": 317 + }, + { + "epoch": 0.6247544204322201, + "grad_norm": 1.1557381280076295, + "learning_rate": 7.985069666221316e-05, + "loss": 0.6255, + "step": 318 + }, + { + "epoch": 0.6267190569744597, + "grad_norm": 0.8219686047849541, + "learning_rate": 7.984592232571994e-05, + "loss": 0.6309, + "step": 319 + }, + { + "epoch": 0.6286836935166994, + "grad_norm": 0.6432457148602014, + "learning_rate": 7.98410729976307e-05, + "loss": 0.632, + "step": 320 + }, + { + "epoch": 0.630648330058939, + "grad_norm": 0.9275855177948447, + "learning_rate": 7.983614868707208e-05, + "loss": 0.6354, + "step": 321 + }, + { + "epoch": 0.6326129666011788, + "grad_norm": 0.9947192887622427, + "learning_rate": 7.983114940331181e-05, + "loss": 0.6406, + "step": 322 + }, + { + "epoch": 0.6345776031434185, + "grad_norm": 1.0317902594282986, + "learning_rate": 7.982607515575874e-05, + "loss": 0.636, + "step": 323 + }, + { + "epoch": 0.6365422396856582, + "grad_norm": 1.430692404919756, + "learning_rate": 7.982092595396282e-05, + "loss": 0.6332, + "step": 324 + }, + { + "epoch": 0.6385068762278978, + "grad_norm": 0.6162550137467003, + "learning_rate": 7.981570180761503e-05, + "loss": 0.6338, + "step": 325 + }, + { + "epoch": 0.6404715127701375, + "grad_norm": 1.1145471210591928, + "learning_rate": 7.981040272654743e-05, + "loss": 0.633, + "step": 326 + }, + { + "epoch": 0.6424361493123772, + "grad_norm": 1.2982968679986528, + "learning_rate": 7.98050287207331e-05, + "loss": 0.641, + "step": 327 + }, + { + "epoch": 0.6444007858546169, + "grad_norm": 0.8640840665662881, + "learning_rate": 7.979957980028612e-05, + "loss": 0.6349, + "step": 328 + }, + { + "epoch": 0.6463654223968566, + "grad_norm": 1.2598551272786302, + "learning_rate": 7.97940559754616e-05, + "loss": 0.6307, + "step": 329 + }, + { + "epoch": 0.6483300589390962, + "grad_norm": 0.668344896857072, + "learning_rate": 7.978845725665557e-05, + "loss": 0.6242, + "step": 330 + }, + { + "epoch": 0.650294695481336, + "grad_norm": 0.8846576109931551, + "learning_rate": 7.978278365440504e-05, + "loss": 0.6338, + "step": 331 + }, + { + "epoch": 0.6522593320235757, + "grad_norm": 0.9435774184990293, + "learning_rate": 7.977703517938797e-05, + "loss": 0.6279, + "step": 332 + }, + { + "epoch": 0.6542239685658153, + "grad_norm": 1.0800351627834228, + "learning_rate": 7.977121184242321e-05, + "loss": 0.6324, + "step": 333 + }, + { + "epoch": 0.656188605108055, + "grad_norm": 0.976409003992104, + "learning_rate": 7.97653136544705e-05, + "loss": 0.6365, + "step": 334 + }, + { + "epoch": 0.6581532416502947, + "grad_norm": 0.9731895769690256, + "learning_rate": 7.975934062663046e-05, + "loss": 0.6379, + "step": 335 + }, + { + "epoch": 0.6601178781925344, + "grad_norm": 1.4320858024218315, + "learning_rate": 7.975329277014459e-05, + "loss": 0.6337, + "step": 336 + }, + { + "epoch": 0.6620825147347741, + "grad_norm": 0.812417558845335, + "learning_rate": 7.974717009639516e-05, + "loss": 0.6283, + "step": 337 + }, + { + "epoch": 0.6640471512770137, + "grad_norm": 0.6981127283097993, + "learning_rate": 7.974097261690531e-05, + "loss": 0.6366, + "step": 338 + }, + { + "epoch": 0.6660117878192534, + "grad_norm": 0.8699957090182442, + "learning_rate": 7.973470034333894e-05, + "loss": 0.6341, + "step": 339 + }, + { + "epoch": 0.6679764243614931, + "grad_norm": 1.3099050731151836, + "learning_rate": 7.97283532875007e-05, + "loss": 0.6288, + "step": 340 + }, + { + "epoch": 0.6699410609037328, + "grad_norm": 0.8766386010499578, + "learning_rate": 7.972193146133599e-05, + "loss": 0.6235, + "step": 341 + }, + { + "epoch": 0.6719056974459725, + "grad_norm": 0.9829486709660343, + "learning_rate": 7.971543487693097e-05, + "loss": 0.6336, + "step": 342 + }, + { + "epoch": 0.6738703339882122, + "grad_norm": 0.9049943944496605, + "learning_rate": 7.970886354651244e-05, + "loss": 0.6206, + "step": 343 + }, + { + "epoch": 0.6758349705304518, + "grad_norm": 0.8551486873721904, + "learning_rate": 7.970221748244793e-05, + "loss": 0.6331, + "step": 344 + }, + { + "epoch": 0.6777996070726916, + "grad_norm": 0.8980417663895452, + "learning_rate": 7.969549669724558e-05, + "loss": 0.6354, + "step": 345 + }, + { + "epoch": 0.6797642436149313, + "grad_norm": 1.069794258551904, + "learning_rate": 7.968870120355418e-05, + "loss": 0.6369, + "step": 346 + }, + { + "epoch": 0.6817288801571709, + "grad_norm": 1.0825942244402398, + "learning_rate": 7.96818310141631e-05, + "loss": 0.6345, + "step": 347 + }, + { + "epoch": 0.6836935166994106, + "grad_norm": 0.8843779951564431, + "learning_rate": 7.967488614200232e-05, + "loss": 0.6233, + "step": 348 + }, + { + "epoch": 0.6856581532416502, + "grad_norm": 1.023417852532195, + "learning_rate": 7.966786660014236e-05, + "loss": 0.6315, + "step": 349 + }, + { + "epoch": 0.68762278978389, + "grad_norm": 1.1642803413660459, + "learning_rate": 7.966077240179428e-05, + "loss": 0.6304, + "step": 350 + }, + { + "epoch": 0.6895874263261297, + "grad_norm": 0.8417790254816612, + "learning_rate": 7.965360356030962e-05, + "loss": 0.6323, + "step": 351 + }, + { + "epoch": 0.6915520628683693, + "grad_norm": 0.842999014395929, + "learning_rate": 7.964636008918045e-05, + "loss": 0.6396, + "step": 352 + }, + { + "epoch": 0.693516699410609, + "grad_norm": 0.8060292553034676, + "learning_rate": 7.963904200203926e-05, + "loss": 0.621, + "step": 353 + }, + { + "epoch": 0.6954813359528488, + "grad_norm": 1.1684327529091523, + "learning_rate": 7.963164931265898e-05, + "loss": 0.6219, + "step": 354 + }, + { + "epoch": 0.6974459724950884, + "grad_norm": 1.0378607100487525, + "learning_rate": 7.962418203495292e-05, + "loss": 0.6196, + "step": 355 + }, + { + "epoch": 0.6994106090373281, + "grad_norm": 0.8434129175093324, + "learning_rate": 7.96166401829748e-05, + "loss": 0.6293, + "step": 356 + }, + { + "epoch": 0.7013752455795678, + "grad_norm": 0.9128348058820919, + "learning_rate": 7.960902377091872e-05, + "loss": 0.6218, + "step": 357 + }, + { + "epoch": 0.7033398821218074, + "grad_norm": 0.7701021478497024, + "learning_rate": 7.960133281311902e-05, + "loss": 0.6253, + "step": 358 + }, + { + "epoch": 0.7053045186640472, + "grad_norm": 0.7611687505046378, + "learning_rate": 7.95935673240504e-05, + "loss": 0.6306, + "step": 359 + }, + { + "epoch": 0.7072691552062869, + "grad_norm": 1.0050602049607942, + "learning_rate": 7.958572731832784e-05, + "loss": 0.6175, + "step": 360 + }, + { + "epoch": 0.7092337917485265, + "grad_norm": 1.0553016255993366, + "learning_rate": 7.95778128107065e-05, + "loss": 0.6192, + "step": 361 + }, + { + "epoch": 0.7111984282907662, + "grad_norm": 0.877933452110947, + "learning_rate": 7.956982381608184e-05, + "loss": 0.6313, + "step": 362 + }, + { + "epoch": 0.7131630648330058, + "grad_norm": 1.053495168837769, + "learning_rate": 7.956176034948942e-05, + "loss": 0.6276, + "step": 363 + }, + { + "epoch": 0.7151277013752456, + "grad_norm": 0.9777598112909676, + "learning_rate": 7.955362242610502e-05, + "loss": 0.6297, + "step": 364 + }, + { + "epoch": 0.7170923379174853, + "grad_norm": 1.2501369988571165, + "learning_rate": 7.954541006124453e-05, + "loss": 0.6188, + "step": 365 + }, + { + "epoch": 0.7190569744597249, + "grad_norm": 0.7748006719268069, + "learning_rate": 7.953712327036396e-05, + "loss": 0.6179, + "step": 366 + }, + { + "epoch": 0.7210216110019646, + "grad_norm": 0.5537578787996011, + "learning_rate": 7.952876206905936e-05, + "loss": 0.6238, + "step": 367 + }, + { + "epoch": 0.7229862475442044, + "grad_norm": 0.8636462666896847, + "learning_rate": 7.952032647306684e-05, + "loss": 0.6179, + "step": 368 + }, + { + "epoch": 0.724950884086444, + "grad_norm": 1.2311995432611944, + "learning_rate": 7.951181649826252e-05, + "loss": 0.6233, + "step": 369 + }, + { + "epoch": 0.7269155206286837, + "grad_norm": 0.7554296553067995, + "learning_rate": 7.950323216066252e-05, + "loss": 0.6282, + "step": 370 + }, + { + "epoch": 0.7288801571709234, + "grad_norm": 0.682725794045951, + "learning_rate": 7.949457347642292e-05, + "loss": 0.6136, + "step": 371 + }, + { + "epoch": 0.730844793713163, + "grad_norm": 0.7601682933086549, + "learning_rate": 7.948584046183967e-05, + "loss": 0.6259, + "step": 372 + }, + { + "epoch": 0.7328094302554028, + "grad_norm": 0.7576124576613958, + "learning_rate": 7.947703313334866e-05, + "loss": 0.6186, + "step": 373 + }, + { + "epoch": 0.7347740667976425, + "grad_norm": 0.8891404954737672, + "learning_rate": 7.94681515075256e-05, + "loss": 0.6334, + "step": 374 + }, + { + "epoch": 0.7367387033398821, + "grad_norm": 1.0478259204461204, + "learning_rate": 7.945919560108612e-05, + "loss": 0.6186, + "step": 375 + }, + { + "epoch": 0.7387033398821218, + "grad_norm": 0.9654324680640894, + "learning_rate": 7.945016543088556e-05, + "loss": 0.6208, + "step": 376 + }, + { + "epoch": 0.7406679764243614, + "grad_norm": 0.9414023492595159, + "learning_rate": 7.944106101391906e-05, + "loss": 0.6219, + "step": 377 + }, + { + "epoch": 0.7426326129666012, + "grad_norm": 0.8078645656794177, + "learning_rate": 7.943188236732147e-05, + "loss": 0.6158, + "step": 378 + }, + { + "epoch": 0.7445972495088409, + "grad_norm": 0.8267976076863125, + "learning_rate": 7.942262950836738e-05, + "loss": 0.6279, + "step": 379 + }, + { + "epoch": 0.7465618860510805, + "grad_norm": 0.9482754087996211, + "learning_rate": 7.941330245447103e-05, + "loss": 0.6238, + "step": 380 + }, + { + "epoch": 0.7485265225933202, + "grad_norm": 1.0420399071014297, + "learning_rate": 7.940390122318633e-05, + "loss": 0.6182, + "step": 381 + }, + { + "epoch": 0.75049115913556, + "grad_norm": 1.0113335720183754, + "learning_rate": 7.939442583220674e-05, + "loss": 0.6243, + "step": 382 + }, + { + "epoch": 0.7524557956777996, + "grad_norm": 1.0393749495318294, + "learning_rate": 7.938487629936532e-05, + "loss": 0.632, + "step": 383 + }, + { + "epoch": 0.7544204322200393, + "grad_norm": 0.8893331707979845, + "learning_rate": 7.937525264263468e-05, + "loss": 0.6315, + "step": 384 + }, + { + "epoch": 0.756385068762279, + "grad_norm": 1.0962134462855055, + "learning_rate": 7.936555488012691e-05, + "loss": 0.6271, + "step": 385 + }, + { + "epoch": 0.7583497053045186, + "grad_norm": 0.7998075588203503, + "learning_rate": 7.935578303009358e-05, + "loss": 0.6099, + "step": 386 + }, + { + "epoch": 0.7603143418467584, + "grad_norm": 0.5965648405483078, + "learning_rate": 7.93459371109257e-05, + "loss": 0.6313, + "step": 387 + }, + { + "epoch": 0.762278978388998, + "grad_norm": 0.7243009589556959, + "learning_rate": 7.933601714115369e-05, + "loss": 0.6181, + "step": 388 + }, + { + "epoch": 0.7642436149312377, + "grad_norm": 1.0528197328033981, + "learning_rate": 7.932602313944731e-05, + "loss": 0.6257, + "step": 389 + }, + { + "epoch": 0.7662082514734774, + "grad_norm": 1.42062209550092, + "learning_rate": 7.931595512461566e-05, + "loss": 0.6186, + "step": 390 + }, + { + "epoch": 0.768172888015717, + "grad_norm": 0.5925779341188574, + "learning_rate": 7.930581311560717e-05, + "loss": 0.6329, + "step": 391 + }, + { + "epoch": 0.7701375245579568, + "grad_norm": 0.9334168395770492, + "learning_rate": 7.929559713150946e-05, + "loss": 0.6175, + "step": 392 + }, + { + "epoch": 0.7721021611001965, + "grad_norm": 1.8586867318271307, + "learning_rate": 7.928530719154944e-05, + "loss": 0.6338, + "step": 393 + }, + { + "epoch": 0.7740667976424361, + "grad_norm": 0.7791325800148028, + "learning_rate": 7.92749433150932e-05, + "loss": 0.6274, + "step": 394 + }, + { + "epoch": 0.7760314341846758, + "grad_norm": 2.7171616572225368, + "learning_rate": 7.926450552164591e-05, + "loss": 0.6212, + "step": 395 + }, + { + "epoch": 0.7779960707269156, + "grad_norm": 2.297776495520809, + "learning_rate": 7.925399383085195e-05, + "loss": 0.6321, + "step": 396 + }, + { + "epoch": 0.7799607072691552, + "grad_norm": 1.4420634560741765, + "learning_rate": 7.924340826249473e-05, + "loss": 0.6167, + "step": 397 + }, + { + "epoch": 0.7819253438113949, + "grad_norm": 1.3041269761490206, + "learning_rate": 7.923274883649671e-05, + "loss": 0.6407, + "step": 398 + }, + { + "epoch": 0.7838899803536346, + "grad_norm": 1.2644421298702428, + "learning_rate": 7.922201557291933e-05, + "loss": 0.6223, + "step": 399 + }, + { + "epoch": 0.7858546168958742, + "grad_norm": 0.7804001242359935, + "learning_rate": 7.921120849196303e-05, + "loss": 0.6203, + "step": 400 + }, + { + "epoch": 0.787819253438114, + "grad_norm": 1.0440055925699654, + "learning_rate": 7.920032761396716e-05, + "loss": 0.6142, + "step": 401 + }, + { + "epoch": 0.7897838899803536, + "grad_norm": 0.7447823988668982, + "learning_rate": 7.918937295940997e-05, + "loss": 0.6219, + "step": 402 + }, + { + "epoch": 0.7917485265225933, + "grad_norm": 0.8081066950920281, + "learning_rate": 7.917834454890854e-05, + "loss": 0.6211, + "step": 403 + }, + { + "epoch": 0.793713163064833, + "grad_norm": 0.7733768005890879, + "learning_rate": 7.916724240321878e-05, + "loss": 0.6294, + "step": 404 + }, + { + "epoch": 0.7956777996070727, + "grad_norm": 0.6497301371545382, + "learning_rate": 7.915606654323534e-05, + "loss": 0.6168, + "step": 405 + }, + { + "epoch": 0.7976424361493124, + "grad_norm": 0.6932571902081731, + "learning_rate": 7.914481698999168e-05, + "loss": 0.6056, + "step": 406 + }, + { + "epoch": 0.7996070726915521, + "grad_norm": 0.6294977634767592, + "learning_rate": 7.913349376465984e-05, + "loss": 0.619, + "step": 407 + }, + { + "epoch": 0.8015717092337917, + "grad_norm": 0.6605719101465373, + "learning_rate": 7.912209688855063e-05, + "loss": 0.6285, + "step": 408 + }, + { + "epoch": 0.8035363457760314, + "grad_norm": 0.8536750357853278, + "learning_rate": 7.91106263831134e-05, + "loss": 0.6055, + "step": 409 + }, + { + "epoch": 0.8055009823182712, + "grad_norm": 0.8902153461284296, + "learning_rate": 7.909908226993609e-05, + "loss": 0.6344, + "step": 410 + }, + { + "epoch": 0.8074656188605108, + "grad_norm": 0.9274860240946536, + "learning_rate": 7.908746457074518e-05, + "loss": 0.6132, + "step": 411 + }, + { + "epoch": 0.8094302554027505, + "grad_norm": 1.0646219340717782, + "learning_rate": 7.907577330740563e-05, + "loss": 0.6268, + "step": 412 + }, + { + "epoch": 0.8113948919449901, + "grad_norm": 1.0827022986135162, + "learning_rate": 7.906400850192087e-05, + "loss": 0.6269, + "step": 413 + }, + { + "epoch": 0.8133595284872298, + "grad_norm": 0.7950222011511615, + "learning_rate": 7.905217017643272e-05, + "loss": 0.6204, + "step": 414 + }, + { + "epoch": 0.8153241650294696, + "grad_norm": 0.6402924516915863, + "learning_rate": 7.904025835322139e-05, + "loss": 0.6137, + "step": 415 + }, + { + "epoch": 0.8172888015717092, + "grad_norm": 0.595096387375667, + "learning_rate": 7.902827305470535e-05, + "loss": 0.6103, + "step": 416 + }, + { + "epoch": 0.8192534381139489, + "grad_norm": 0.7204754142768892, + "learning_rate": 7.901621430344148e-05, + "loss": 0.6044, + "step": 417 + }, + { + "epoch": 0.8212180746561886, + "grad_norm": 0.7760727547971626, + "learning_rate": 7.900408212212478e-05, + "loss": 0.6205, + "step": 418 + }, + { + "epoch": 0.8231827111984283, + "grad_norm": 0.6454939398262192, + "learning_rate": 7.899187653358849e-05, + "loss": 0.6245, + "step": 419 + }, + { + "epoch": 0.825147347740668, + "grad_norm": 0.7402132606166687, + "learning_rate": 7.897959756080402e-05, + "loss": 0.6223, + "step": 420 + }, + { + "epoch": 0.8271119842829077, + "grad_norm": 0.8379521776521917, + "learning_rate": 7.896724522688088e-05, + "loss": 0.6159, + "step": 421 + }, + { + "epoch": 0.8290766208251473, + "grad_norm": 0.9303754877329591, + "learning_rate": 7.895481955506665e-05, + "loss": 0.6068, + "step": 422 + }, + { + "epoch": 0.831041257367387, + "grad_norm": 0.9535916683566573, + "learning_rate": 7.894232056874694e-05, + "loss": 0.6121, + "step": 423 + }, + { + "epoch": 0.8330058939096268, + "grad_norm": 0.916024407183147, + "learning_rate": 7.892974829144532e-05, + "loss": 0.6196, + "step": 424 + }, + { + "epoch": 0.8349705304518664, + "grad_norm": 0.9232685589839004, + "learning_rate": 7.891710274682333e-05, + "loss": 0.6316, + "step": 425 + }, + { + "epoch": 0.8369351669941061, + "grad_norm": 1.01063835952548, + "learning_rate": 7.890438395868035e-05, + "loss": 0.6131, + "step": 426 + }, + { + "epoch": 0.8388998035363457, + "grad_norm": 1.2324366965292792, + "learning_rate": 7.889159195095367e-05, + "loss": 0.6215, + "step": 427 + }, + { + "epoch": 0.8408644400785854, + "grad_norm": 0.9268307155793002, + "learning_rate": 7.887872674771836e-05, + "loss": 0.6167, + "step": 428 + }, + { + "epoch": 0.8428290766208252, + "grad_norm": 0.8318710095493993, + "learning_rate": 7.886578837318722e-05, + "loss": 0.611, + "step": 429 + }, + { + "epoch": 0.8447937131630648, + "grad_norm": 0.7356490541377241, + "learning_rate": 7.885277685171078e-05, + "loss": 0.6138, + "step": 430 + }, + { + "epoch": 0.8467583497053045, + "grad_norm": 0.8807879013262219, + "learning_rate": 7.883969220777724e-05, + "loss": 0.6169, + "step": 431 + }, + { + "epoch": 0.8487229862475442, + "grad_norm": 1.2365031058740072, + "learning_rate": 7.882653446601242e-05, + "loss": 0.6236, + "step": 432 + }, + { + "epoch": 0.8506876227897839, + "grad_norm": 0.5649901032365545, + "learning_rate": 7.88133036511797e-05, + "loss": 0.6173, + "step": 433 + }, + { + "epoch": 0.8526522593320236, + "grad_norm": 0.6030910063688245, + "learning_rate": 7.879999978818001e-05, + "loss": 0.6153, + "step": 434 + }, + { + "epoch": 0.8546168958742633, + "grad_norm": 0.8442191986029831, + "learning_rate": 7.878662290205173e-05, + "loss": 0.6245, + "step": 435 + }, + { + "epoch": 0.8565815324165029, + "grad_norm": 0.9266101824944539, + "learning_rate": 7.877317301797069e-05, + "loss": 0.6229, + "step": 436 + }, + { + "epoch": 0.8585461689587426, + "grad_norm": 1.0267653637711225, + "learning_rate": 7.875965016125011e-05, + "loss": 0.6167, + "step": 437 + }, + { + "epoch": 0.8605108055009824, + "grad_norm": 1.0749942626259819, + "learning_rate": 7.874605435734052e-05, + "loss": 0.6165, + "step": 438 + }, + { + "epoch": 0.862475442043222, + "grad_norm": 0.8612725069470999, + "learning_rate": 7.873238563182979e-05, + "loss": 0.6111, + "step": 439 + }, + { + "epoch": 0.8644400785854617, + "grad_norm": 0.5620157322734657, + "learning_rate": 7.871864401044297e-05, + "loss": 0.6174, + "step": 440 + }, + { + "epoch": 0.8664047151277013, + "grad_norm": 0.42486301299091184, + "learning_rate": 7.870482951904235e-05, + "loss": 0.6143, + "step": 441 + }, + { + "epoch": 0.8683693516699411, + "grad_norm": 0.6465553244536739, + "learning_rate": 7.869094218362735e-05, + "loss": 0.6072, + "step": 442 + }, + { + "epoch": 0.8703339882121808, + "grad_norm": 0.854156937659523, + "learning_rate": 7.867698203033446e-05, + "loss": 0.6132, + "step": 443 + }, + { + "epoch": 0.8722986247544204, + "grad_norm": 0.9864602570619255, + "learning_rate": 7.866294908543726e-05, + "loss": 0.6237, + "step": 444 + }, + { + "epoch": 0.8742632612966601, + "grad_norm": 1.0562705392322964, + "learning_rate": 7.86488433753463e-05, + "loss": 0.6077, + "step": 445 + }, + { + "epoch": 0.8762278978388998, + "grad_norm": 0.8445681828567188, + "learning_rate": 7.863466492660907e-05, + "loss": 0.632, + "step": 446 + }, + { + "epoch": 0.8781925343811395, + "grad_norm": 0.7167766487201322, + "learning_rate": 7.862041376590999e-05, + "loss": 0.603, + "step": 447 + }, + { + "epoch": 0.8801571709233792, + "grad_norm": 0.838838218746552, + "learning_rate": 7.860608992007028e-05, + "loss": 0.6027, + "step": 448 + }, + { + "epoch": 0.8821218074656189, + "grad_norm": 0.91758154414098, + "learning_rate": 7.859169341604802e-05, + "loss": 0.606, + "step": 449 + }, + { + "epoch": 0.8840864440078585, + "grad_norm": 0.818235179902976, + "learning_rate": 7.857722428093797e-05, + "loss": 0.6225, + "step": 450 + }, + { + "epoch": 0.8860510805500982, + "grad_norm": 0.5681204307931715, + "learning_rate": 7.85626825419716e-05, + "loss": 0.6196, + "step": 451 + }, + { + "epoch": 0.888015717092338, + "grad_norm": 0.657402951979804, + "learning_rate": 7.854806822651707e-05, + "loss": 0.6144, + "step": 452 + }, + { + "epoch": 0.8899803536345776, + "grad_norm": 0.8725509275481057, + "learning_rate": 7.853338136207908e-05, + "loss": 0.6087, + "step": 453 + }, + { + "epoch": 0.8919449901768173, + "grad_norm": 0.9521379494589475, + "learning_rate": 7.851862197629888e-05, + "loss": 0.6192, + "step": 454 + }, + { + "epoch": 0.8939096267190569, + "grad_norm": 1.1660101123207185, + "learning_rate": 7.850379009695426e-05, + "loss": 0.6151, + "step": 455 + }, + { + "epoch": 0.8958742632612967, + "grad_norm": 0.745726907354287, + "learning_rate": 7.848888575195933e-05, + "loss": 0.6277, + "step": 456 + }, + { + "epoch": 0.8978388998035364, + "grad_norm": 0.4825938315104395, + "learning_rate": 7.84739089693647e-05, + "loss": 0.6103, + "step": 457 + }, + { + "epoch": 0.899803536345776, + "grad_norm": 0.5573612363402732, + "learning_rate": 7.845885977735726e-05, + "loss": 0.6058, + "step": 458 + }, + { + "epoch": 0.9017681728880157, + "grad_norm": 0.9460311895652972, + "learning_rate": 7.844373820426018e-05, + "loss": 0.6137, + "step": 459 + }, + { + "epoch": 0.9037328094302554, + "grad_norm": 1.1297003428639467, + "learning_rate": 7.842854427853285e-05, + "loss": 0.6146, + "step": 460 + }, + { + "epoch": 0.9056974459724951, + "grad_norm": 0.7707541737768038, + "learning_rate": 7.841327802877085e-05, + "loss": 0.61, + "step": 461 + }, + { + "epoch": 0.9076620825147348, + "grad_norm": 0.7096370565730452, + "learning_rate": 7.839793948370585e-05, + "loss": 0.6129, + "step": 462 + }, + { + "epoch": 0.9096267190569745, + "grad_norm": 0.6922627242103819, + "learning_rate": 7.838252867220559e-05, + "loss": 0.6161, + "step": 463 + }, + { + "epoch": 0.9115913555992141, + "grad_norm": 0.436938136738934, + "learning_rate": 7.836704562327387e-05, + "loss": 0.6191, + "step": 464 + }, + { + "epoch": 0.9135559921414538, + "grad_norm": 0.5445222473053478, + "learning_rate": 7.835149036605033e-05, + "loss": 0.6225, + "step": 465 + }, + { + "epoch": 0.9155206286836935, + "grad_norm": 0.6265908207717932, + "learning_rate": 7.833586292981063e-05, + "loss": 0.6138, + "step": 466 + }, + { + "epoch": 0.9174852652259332, + "grad_norm": 0.5321872775766991, + "learning_rate": 7.832016334396618e-05, + "loss": 0.6135, + "step": 467 + }, + { + "epoch": 0.9194499017681729, + "grad_norm": 0.6644548399548946, + "learning_rate": 7.830439163806427e-05, + "loss": 0.615, + "step": 468 + }, + { + "epoch": 0.9214145383104125, + "grad_norm": 0.9121531542653692, + "learning_rate": 7.828854784178784e-05, + "loss": 0.6136, + "step": 469 + }, + { + "epoch": 0.9233791748526523, + "grad_norm": 1.3202636876428642, + "learning_rate": 7.827263198495554e-05, + "loss": 0.6177, + "step": 470 + }, + { + "epoch": 0.925343811394892, + "grad_norm": 0.6840949805357649, + "learning_rate": 7.825664409752163e-05, + "loss": 0.6179, + "step": 471 + }, + { + "epoch": 0.9273084479371316, + "grad_norm": 0.6386296070915081, + "learning_rate": 7.824058420957598e-05, + "loss": 0.6115, + "step": 472 + }, + { + "epoch": 0.9292730844793713, + "grad_norm": 0.8374237824997119, + "learning_rate": 7.822445235134388e-05, + "loss": 0.621, + "step": 473 + }, + { + "epoch": 0.931237721021611, + "grad_norm": 0.965784167723971, + "learning_rate": 7.820824855318616e-05, + "loss": 0.6106, + "step": 474 + }, + { + "epoch": 0.9332023575638507, + "grad_norm": 1.1942001692143531, + "learning_rate": 7.819197284559901e-05, + "loss": 0.6186, + "step": 475 + }, + { + "epoch": 0.9351669941060904, + "grad_norm": 0.77345992818146, + "learning_rate": 7.817562525921394e-05, + "loss": 0.6151, + "step": 476 + }, + { + "epoch": 0.93713163064833, + "grad_norm": 0.759919952689018, + "learning_rate": 7.815920582479775e-05, + "loss": 0.6062, + "step": 477 + }, + { + "epoch": 0.9390962671905697, + "grad_norm": 0.9836380124974429, + "learning_rate": 7.814271457325247e-05, + "loss": 0.624, + "step": 478 + }, + { + "epoch": 0.9410609037328095, + "grad_norm": 0.8509434349286433, + "learning_rate": 7.812615153561526e-05, + "loss": 0.6069, + "step": 479 + }, + { + "epoch": 0.9430255402750491, + "grad_norm": 0.5632736321304582, + "learning_rate": 7.810951674305845e-05, + "loss": 0.6115, + "step": 480 + }, + { + "epoch": 0.9449901768172888, + "grad_norm": 0.47613649607994324, + "learning_rate": 7.809281022688935e-05, + "loss": 0.6036, + "step": 481 + }, + { + "epoch": 0.9469548133595285, + "grad_norm": 0.5959189452233215, + "learning_rate": 7.80760320185503e-05, + "loss": 0.6277, + "step": 482 + }, + { + "epoch": 0.9489194499017681, + "grad_norm": 0.7840168825659439, + "learning_rate": 7.805918214961851e-05, + "loss": 0.5972, + "step": 483 + }, + { + "epoch": 0.9508840864440079, + "grad_norm": 0.8433934260999976, + "learning_rate": 7.804226065180615e-05, + "loss": 0.6132, + "step": 484 + }, + { + "epoch": 0.9528487229862476, + "grad_norm": 0.7784651306937933, + "learning_rate": 7.802526755696011e-05, + "loss": 0.6143, + "step": 485 + }, + { + "epoch": 0.9548133595284872, + "grad_norm": 0.7742325900541732, + "learning_rate": 7.800820289706207e-05, + "loss": 0.6096, + "step": 486 + }, + { + "epoch": 0.9567779960707269, + "grad_norm": 0.8511370359390014, + "learning_rate": 7.79910667042284e-05, + "loss": 0.6201, + "step": 487 + }, + { + "epoch": 0.9587426326129665, + "grad_norm": 0.9459054810913974, + "learning_rate": 7.797385901071009e-05, + "loss": 0.6141, + "step": 488 + }, + { + "epoch": 0.9607072691552063, + "grad_norm": 0.8693787811219073, + "learning_rate": 7.79565798488927e-05, + "loss": 0.6057, + "step": 489 + }, + { + "epoch": 0.962671905697446, + "grad_norm": 0.7306705431776497, + "learning_rate": 7.793922925129626e-05, + "loss": 0.6096, + "step": 490 + }, + { + "epoch": 0.9646365422396856, + "grad_norm": 0.5798688179401159, + "learning_rate": 7.792180725057533e-05, + "loss": 0.602, + "step": 491 + }, + { + "epoch": 0.9666011787819253, + "grad_norm": 0.5183893253704686, + "learning_rate": 7.790431387951878e-05, + "loss": 0.6079, + "step": 492 + }, + { + "epoch": 0.9685658153241651, + "grad_norm": 0.48092933427786966, + "learning_rate": 7.78867491710498e-05, + "loss": 0.6072, + "step": 493 + }, + { + "epoch": 0.9705304518664047, + "grad_norm": 0.5277067183162334, + "learning_rate": 7.78691131582259e-05, + "loss": 0.5914, + "step": 494 + }, + { + "epoch": 0.9724950884086444, + "grad_norm": 0.6175762658992761, + "learning_rate": 7.785140587423872e-05, + "loss": 0.6081, + "step": 495 + }, + { + "epoch": 0.9744597249508841, + "grad_norm": 0.5258017851850615, + "learning_rate": 7.78336273524141e-05, + "loss": 0.6091, + "step": 496 + }, + { + "epoch": 0.9764243614931237, + "grad_norm": 0.5239164490351178, + "learning_rate": 7.78157776262119e-05, + "loss": 0.607, + "step": 497 + }, + { + "epoch": 0.9783889980353635, + "grad_norm": 0.5557948396583752, + "learning_rate": 7.7797856729226e-05, + "loss": 0.6114, + "step": 498 + }, + { + "epoch": 0.9803536345776032, + "grad_norm": 0.6293568925553193, + "learning_rate": 7.777986469518424e-05, + "loss": 0.6047, + "step": 499 + }, + { + "epoch": 0.9823182711198428, + "grad_norm": 0.7607851705716018, + "learning_rate": 7.776180155794836e-05, + "loss": 0.6133, + "step": 500 + }, + { + "epoch": 0.9842829076620825, + "grad_norm": 0.9928449279027518, + "learning_rate": 7.774366735151386e-05, + "loss": 0.5982, + "step": 501 + }, + { + "epoch": 0.9862475442043221, + "grad_norm": 1.2313282153960308, + "learning_rate": 7.772546211001005e-05, + "loss": 0.6151, + "step": 502 + }, + { + "epoch": 0.9882121807465619, + "grad_norm": 0.6863130776803184, + "learning_rate": 7.770718586769991e-05, + "loss": 0.6202, + "step": 503 + }, + { + "epoch": 0.9901768172888016, + "grad_norm": 0.4930841370730766, + "learning_rate": 7.768883865898005e-05, + "loss": 0.6095, + "step": 504 + }, + { + "epoch": 0.9921414538310412, + "grad_norm": 0.6941145972498602, + "learning_rate": 7.767042051838064e-05, + "loss": 0.6116, + "step": 505 + }, + { + "epoch": 0.9941060903732809, + "grad_norm": 1.0470066627328682, + "learning_rate": 7.765193148056533e-05, + "loss": 0.6162, + "step": 506 + }, + { + "epoch": 0.9960707269155207, + "grad_norm": 1.1322614460513978, + "learning_rate": 7.763337158033125e-05, + "loss": 0.6148, + "step": 507 + }, + { + "epoch": 0.9980353634577603, + "grad_norm": 0.7325017425231686, + "learning_rate": 7.761474085260883e-05, + "loss": 0.6096, + "step": 508 + }, + { + "epoch": 1.0, + "grad_norm": 0.5675484376734906, + "learning_rate": 7.759603933246186e-05, + "loss": 0.6151, + "step": 509 + }, + { + "epoch": 1.0019646365422397, + "grad_norm": 0.5913096853383666, + "learning_rate": 7.757726705508732e-05, + "loss": 0.5859, + "step": 510 + }, + { + "epoch": 1.0039292730844793, + "grad_norm": 0.4376255435797527, + "learning_rate": 7.755842405581538e-05, + "loss": 0.5846, + "step": 511 + }, + { + "epoch": 1.005893909626719, + "grad_norm": 0.44512573335382705, + "learning_rate": 7.753951037010931e-05, + "loss": 0.5801, + "step": 512 + }, + { + "epoch": 1.0078585461689586, + "grad_norm": 0.5442773255579556, + "learning_rate": 7.752052603356542e-05, + "loss": 0.5781, + "step": 513 + }, + { + "epoch": 1.0098231827111985, + "grad_norm": 0.6498291335059194, + "learning_rate": 7.750147108191297e-05, + "loss": 0.5779, + "step": 514 + }, + { + "epoch": 1.0117878192534382, + "grad_norm": 0.9703179220900373, + "learning_rate": 7.748234555101413e-05, + "loss": 0.5883, + "step": 515 + }, + { + "epoch": 1.0137524557956779, + "grad_norm": 1.1656687090252278, + "learning_rate": 7.74631494768639e-05, + "loss": 0.585, + "step": 516 + }, + { + "epoch": 1.0157170923379175, + "grad_norm": 0.6673648308636863, + "learning_rate": 7.744388289559006e-05, + "loss": 0.5718, + "step": 517 + }, + { + "epoch": 1.0176817288801572, + "grad_norm": 0.7882852935410233, + "learning_rate": 7.742454584345306e-05, + "loss": 0.5813, + "step": 518 + }, + { + "epoch": 1.0196463654223968, + "grad_norm": 0.9536092264027282, + "learning_rate": 7.740513835684601e-05, + "loss": 0.5826, + "step": 519 + }, + { + "epoch": 1.0216110019646365, + "grad_norm": 0.9932057302085658, + "learning_rate": 7.738566047229454e-05, + "loss": 0.5807, + "step": 520 + }, + { + "epoch": 1.0235756385068762, + "grad_norm": 0.9725833271622711, + "learning_rate": 7.736611222645682e-05, + "loss": 0.5858, + "step": 521 + }, + { + "epoch": 1.0255402750491158, + "grad_norm": 1.0411646676708426, + "learning_rate": 7.734649365612341e-05, + "loss": 0.5798, + "step": 522 + }, + { + "epoch": 1.0275049115913557, + "grad_norm": 0.8928187987495823, + "learning_rate": 7.732680479821724e-05, + "loss": 0.5816, + "step": 523 + }, + { + "epoch": 1.0294695481335954, + "grad_norm": 0.7197288659337667, + "learning_rate": 7.730704568979351e-05, + "loss": 0.5738, + "step": 524 + }, + { + "epoch": 1.031434184675835, + "grad_norm": 0.6015815276380119, + "learning_rate": 7.728721636803964e-05, + "loss": 0.5877, + "step": 525 + }, + { + "epoch": 1.0333988212180747, + "grad_norm": 0.5892571654809771, + "learning_rate": 7.726731687027518e-05, + "loss": 0.5716, + "step": 526 + }, + { + "epoch": 1.0353634577603144, + "grad_norm": 0.5815536973587995, + "learning_rate": 7.72473472339518e-05, + "loss": 0.5728, + "step": 527 + }, + { + "epoch": 1.037328094302554, + "grad_norm": 0.6195375394142912, + "learning_rate": 7.722730749665312e-05, + "loss": 0.5814, + "step": 528 + }, + { + "epoch": 1.0392927308447937, + "grad_norm": 0.5628755486883786, + "learning_rate": 7.720719769609473e-05, + "loss": 0.5826, + "step": 529 + }, + { + "epoch": 1.0412573673870333, + "grad_norm": 0.4144866738468408, + "learning_rate": 7.718701787012406e-05, + "loss": 0.5888, + "step": 530 + }, + { + "epoch": 1.043222003929273, + "grad_norm": 0.6164219160269979, + "learning_rate": 7.716676805672033e-05, + "loss": 0.5731, + "step": 531 + }, + { + "epoch": 1.0451866404715129, + "grad_norm": 0.6775431310143931, + "learning_rate": 7.714644829399449e-05, + "loss": 0.5773, + "step": 532 + }, + { + "epoch": 1.0471512770137525, + "grad_norm": 0.6214192550660331, + "learning_rate": 7.712605862018914e-05, + "loss": 0.576, + "step": 533 + }, + { + "epoch": 1.0491159135559922, + "grad_norm": 0.6208597083556998, + "learning_rate": 7.710559907367846e-05, + "loss": 0.587, + "step": 534 + }, + { + "epoch": 1.0510805500982319, + "grad_norm": 0.6949790121442408, + "learning_rate": 7.708506969296811e-05, + "loss": 0.5735, + "step": 535 + }, + { + "epoch": 1.0530451866404715, + "grad_norm": 0.7559103184047927, + "learning_rate": 7.706447051669518e-05, + "loss": 0.575, + "step": 536 + }, + { + "epoch": 1.0550098231827112, + "grad_norm": 0.8240742048379832, + "learning_rate": 7.704380158362816e-05, + "loss": 0.5952, + "step": 537 + }, + { + "epoch": 1.0569744597249509, + "grad_norm": 0.8644991869279312, + "learning_rate": 7.702306293266677e-05, + "loss": 0.582, + "step": 538 + }, + { + "epoch": 1.0589390962671905, + "grad_norm": 0.9126273062149377, + "learning_rate": 7.700225460284199e-05, + "loss": 0.5845, + "step": 539 + }, + { + "epoch": 1.0609037328094302, + "grad_norm": 0.9319283937189592, + "learning_rate": 7.69813766333159e-05, + "loss": 0.587, + "step": 540 + }, + { + "epoch": 1.0628683693516698, + "grad_norm": 1.0102030668986082, + "learning_rate": 7.696042906338166e-05, + "loss": 0.5804, + "step": 541 + }, + { + "epoch": 1.0648330058939097, + "grad_norm": 0.9852689313511013, + "learning_rate": 7.693941193246344e-05, + "loss": 0.5782, + "step": 542 + }, + { + "epoch": 1.0667976424361494, + "grad_norm": 0.7787156673996096, + "learning_rate": 7.691832528011632e-05, + "loss": 0.5662, + "step": 543 + }, + { + "epoch": 1.068762278978389, + "grad_norm": 0.5183216597496203, + "learning_rate": 7.689716914602616e-05, + "loss": 0.5748, + "step": 544 + }, + { + "epoch": 1.0707269155206287, + "grad_norm": 0.4490717537412447, + "learning_rate": 7.687594357000968e-05, + "loss": 0.5816, + "step": 545 + }, + { + "epoch": 1.0726915520628684, + "grad_norm": 0.5699314675894777, + "learning_rate": 7.685464859201424e-05, + "loss": 0.5806, + "step": 546 + }, + { + "epoch": 1.074656188605108, + "grad_norm": 0.6509239582180242, + "learning_rate": 7.683328425211783e-05, + "loss": 0.5627, + "step": 547 + }, + { + "epoch": 1.0766208251473477, + "grad_norm": 0.6665586840614245, + "learning_rate": 7.681185059052895e-05, + "loss": 0.5772, + "step": 548 + }, + { + "epoch": 1.0785854616895874, + "grad_norm": 0.7324923057201544, + "learning_rate": 7.679034764758665e-05, + "loss": 0.5776, + "step": 549 + }, + { + "epoch": 1.080550098231827, + "grad_norm": 0.7605854782423378, + "learning_rate": 7.676877546376028e-05, + "loss": 0.584, + "step": 550 + }, + { + "epoch": 1.082514734774067, + "grad_norm": 0.6886752155957028, + "learning_rate": 7.674713407964954e-05, + "loss": 0.5744, + "step": 551 + }, + { + "epoch": 1.0844793713163066, + "grad_norm": 0.4697611675475534, + "learning_rate": 7.672542353598438e-05, + "loss": 0.578, + "step": 552 + }, + { + "epoch": 1.0864440078585462, + "grad_norm": 0.4872959509206045, + "learning_rate": 7.670364387362491e-05, + "loss": 0.5774, + "step": 553 + }, + { + "epoch": 1.0884086444007859, + "grad_norm": 0.4688323897619248, + "learning_rate": 7.668179513356128e-05, + "loss": 0.5673, + "step": 554 + }, + { + "epoch": 1.0903732809430255, + "grad_norm": 0.41141721254521507, + "learning_rate": 7.66598773569137e-05, + "loss": 0.5722, + "step": 555 + }, + { + "epoch": 1.0923379174852652, + "grad_norm": 0.4857015798368415, + "learning_rate": 7.663789058493228e-05, + "loss": 0.572, + "step": 556 + }, + { + "epoch": 1.0943025540275049, + "grad_norm": 0.5596060346606689, + "learning_rate": 7.6615834858997e-05, + "loss": 0.5726, + "step": 557 + }, + { + "epoch": 1.0962671905697445, + "grad_norm": 0.6393543987301373, + "learning_rate": 7.659371022061762e-05, + "loss": 0.5775, + "step": 558 + }, + { + "epoch": 1.0982318271119842, + "grad_norm": 0.7305843402079256, + "learning_rate": 7.657151671143354e-05, + "loss": 0.5815, + "step": 559 + }, + { + "epoch": 1.1001964636542239, + "grad_norm": 0.8329481605831145, + "learning_rate": 7.654925437321386e-05, + "loss": 0.5729, + "step": 560 + }, + { + "epoch": 1.1021611001964637, + "grad_norm": 0.9674028067317119, + "learning_rate": 7.652692324785713e-05, + "loss": 0.5854, + "step": 561 + }, + { + "epoch": 1.1041257367387034, + "grad_norm": 0.9927741127084521, + "learning_rate": 7.650452337739147e-05, + "loss": 0.5691, + "step": 562 + }, + { + "epoch": 1.106090373280943, + "grad_norm": 0.8661979035309698, + "learning_rate": 7.648205480397428e-05, + "loss": 0.575, + "step": 563 + }, + { + "epoch": 1.1080550098231827, + "grad_norm": 0.7605143160666119, + "learning_rate": 7.645951756989231e-05, + "loss": 0.5801, + "step": 564 + }, + { + "epoch": 1.1100196463654224, + "grad_norm": 0.5322726997735636, + "learning_rate": 7.643691171756151e-05, + "loss": 0.58, + "step": 565 + }, + { + "epoch": 1.111984282907662, + "grad_norm": 0.5451413652648641, + "learning_rate": 7.641423728952701e-05, + "loss": 0.5776, + "step": 566 + }, + { + "epoch": 1.1139489194499017, + "grad_norm": 0.8265114647768989, + "learning_rate": 7.639149432846296e-05, + "loss": 0.5787, + "step": 567 + }, + { + "epoch": 1.1159135559921414, + "grad_norm": 0.9926249500480392, + "learning_rate": 7.63686828771725e-05, + "loss": 0.5767, + "step": 568 + }, + { + "epoch": 1.117878192534381, + "grad_norm": 0.8986286574340177, + "learning_rate": 7.63458029785877e-05, + "loss": 0.5709, + "step": 569 + }, + { + "epoch": 1.119842829076621, + "grad_norm": 0.6813804505785781, + "learning_rate": 7.632285467576943e-05, + "loss": 0.5702, + "step": 570 + }, + { + "epoch": 1.1218074656188606, + "grad_norm": 0.8683045028125586, + "learning_rate": 7.629983801190729e-05, + "loss": 0.5732, + "step": 571 + }, + { + "epoch": 1.1237721021611002, + "grad_norm": 1.1419153865664684, + "learning_rate": 7.627675303031954e-05, + "loss": 0.5836, + "step": 572 + }, + { + "epoch": 1.12573673870334, + "grad_norm": 0.6526877630305734, + "learning_rate": 7.625359977445303e-05, + "loss": 0.581, + "step": 573 + }, + { + "epoch": 1.1277013752455796, + "grad_norm": 0.6723743949824018, + "learning_rate": 7.623037828788314e-05, + "loss": 0.5741, + "step": 574 + }, + { + "epoch": 1.1296660117878192, + "grad_norm": 0.8586127926444197, + "learning_rate": 7.620708861431356e-05, + "loss": 0.5765, + "step": 575 + }, + { + "epoch": 1.1316306483300589, + "grad_norm": 0.7904804794967811, + "learning_rate": 7.618373079757641e-05, + "loss": 0.5792, + "step": 576 + }, + { + "epoch": 1.1335952848722985, + "grad_norm": 0.5801119831982163, + "learning_rate": 7.616030488163201e-05, + "loss": 0.581, + "step": 577 + }, + { + "epoch": 1.1355599214145382, + "grad_norm": 0.5639475351337381, + "learning_rate": 7.613681091056887e-05, + "loss": 0.5804, + "step": 578 + }, + { + "epoch": 1.137524557956778, + "grad_norm": 0.6358389211927354, + "learning_rate": 7.611324892860355e-05, + "loss": 0.5855, + "step": 579 + }, + { + "epoch": 1.1394891944990178, + "grad_norm": 0.5888736990392448, + "learning_rate": 7.608961898008066e-05, + "loss": 0.571, + "step": 580 + }, + { + "epoch": 1.1414538310412574, + "grad_norm": 0.5811739919722826, + "learning_rate": 7.606592110947266e-05, + "loss": 0.5694, + "step": 581 + }, + { + "epoch": 1.143418467583497, + "grad_norm": 0.7520700855510468, + "learning_rate": 7.604215536137988e-05, + "loss": 0.5629, + "step": 582 + }, + { + "epoch": 1.1453831041257367, + "grad_norm": 0.89748907173663, + "learning_rate": 7.601832178053042e-05, + "loss": 0.5734, + "step": 583 + }, + { + "epoch": 1.1473477406679764, + "grad_norm": 0.98438425088902, + "learning_rate": 7.599442041177998e-05, + "loss": 0.5792, + "step": 584 + }, + { + "epoch": 1.149312377210216, + "grad_norm": 0.894326618335511, + "learning_rate": 7.597045130011191e-05, + "loss": 0.576, + "step": 585 + }, + { + "epoch": 1.1512770137524557, + "grad_norm": 0.839319171523697, + "learning_rate": 7.594641449063702e-05, + "loss": 0.5833, + "step": 586 + }, + { + "epoch": 1.1532416502946954, + "grad_norm": 0.7847924549499671, + "learning_rate": 7.592231002859352e-05, + "loss": 0.582, + "step": 587 + }, + { + "epoch": 1.1552062868369353, + "grad_norm": 0.6611006090472067, + "learning_rate": 7.589813795934697e-05, + "loss": 0.5709, + "step": 588 + }, + { + "epoch": 1.157170923379175, + "grad_norm": 0.5810886408246105, + "learning_rate": 7.587389832839015e-05, + "loss": 0.5757, + "step": 589 + }, + { + "epoch": 1.1591355599214146, + "grad_norm": 0.4661087583912746, + "learning_rate": 7.5849591181343e-05, + "loss": 0.5739, + "step": 590 + }, + { + "epoch": 1.1611001964636543, + "grad_norm": 0.4001580165819124, + "learning_rate": 7.582521656395253e-05, + "loss": 0.5806, + "step": 591 + }, + { + "epoch": 1.163064833005894, + "grad_norm": 0.37644952874530885, + "learning_rate": 7.580077452209273e-05, + "loss": 0.5807, + "step": 592 + }, + { + "epoch": 1.1650294695481336, + "grad_norm": 0.4732560499275104, + "learning_rate": 7.577626510176448e-05, + "loss": 0.5819, + "step": 593 + }, + { + "epoch": 1.1669941060903732, + "grad_norm": 0.6180725958117006, + "learning_rate": 7.575168834909548e-05, + "loss": 0.585, + "step": 594 + }, + { + "epoch": 1.168958742632613, + "grad_norm": 0.7360194396540714, + "learning_rate": 7.572704431034015e-05, + "loss": 0.5794, + "step": 595 + }, + { + "epoch": 1.1709233791748526, + "grad_norm": 0.7752514386166893, + "learning_rate": 7.570233303187954e-05, + "loss": 0.5746, + "step": 596 + }, + { + "epoch": 1.1728880157170924, + "grad_norm": 0.8022400437016478, + "learning_rate": 7.567755456022124e-05, + "loss": 0.5825, + "step": 597 + }, + { + "epoch": 1.174852652259332, + "grad_norm": 0.7961452421992461, + "learning_rate": 7.565270894199933e-05, + "loss": 0.5734, + "step": 598 + }, + { + "epoch": 1.1768172888015718, + "grad_norm": 0.7058138830339035, + "learning_rate": 7.562779622397424e-05, + "loss": 0.5679, + "step": 599 + }, + { + "epoch": 1.1787819253438114, + "grad_norm": 0.5771330247028719, + "learning_rate": 7.560281645303266e-05, + "loss": 0.5846, + "step": 600 + }, + { + "epoch": 1.180746561886051, + "grad_norm": 0.5441561109762634, + "learning_rate": 7.557776967618754e-05, + "loss": 0.5845, + "step": 601 + }, + { + "epoch": 1.1827111984282908, + "grad_norm": 0.45981125507350523, + "learning_rate": 7.555265594057787e-05, + "loss": 0.5693, + "step": 602 + }, + { + "epoch": 1.1846758349705304, + "grad_norm": 0.44431128682314885, + "learning_rate": 7.552747529346871e-05, + "loss": 0.5605, + "step": 603 + }, + { + "epoch": 1.18664047151277, + "grad_norm": 0.6191126114744904, + "learning_rate": 7.550222778225102e-05, + "loss": 0.5735, + "step": 604 + }, + { + "epoch": 1.1886051080550097, + "grad_norm": 0.7352143396424555, + "learning_rate": 7.547691345444162e-05, + "loss": 0.5644, + "step": 605 + }, + { + "epoch": 1.1905697445972496, + "grad_norm": 0.7790417162272161, + "learning_rate": 7.545153235768306e-05, + "loss": 0.5655, + "step": 606 + }, + { + "epoch": 1.1925343811394893, + "grad_norm": 0.8675246122222922, + "learning_rate": 7.542608453974359e-05, + "loss": 0.5762, + "step": 607 + }, + { + "epoch": 1.194499017681729, + "grad_norm": 0.889774284505018, + "learning_rate": 7.540057004851698e-05, + "loss": 0.5739, + "step": 608 + }, + { + "epoch": 1.1964636542239686, + "grad_norm": 0.8814612616447307, + "learning_rate": 7.53749889320225e-05, + "loss": 0.5827, + "step": 609 + }, + { + "epoch": 1.1984282907662083, + "grad_norm": 0.8182270050245414, + "learning_rate": 7.534934123840486e-05, + "loss": 0.5769, + "step": 610 + }, + { + "epoch": 1.200392927308448, + "grad_norm": 0.6461458728883608, + "learning_rate": 7.532362701593398e-05, + "loss": 0.5712, + "step": 611 + }, + { + "epoch": 1.2023575638506876, + "grad_norm": 0.5451866566265555, + "learning_rate": 7.529784631300508e-05, + "loss": 0.5781, + "step": 612 + }, + { + "epoch": 1.2043222003929273, + "grad_norm": 0.47997602027164377, + "learning_rate": 7.527199917813843e-05, + "loss": 0.578, + "step": 613 + }, + { + "epoch": 1.206286836935167, + "grad_norm": 0.41097898503525754, + "learning_rate": 7.524608565997937e-05, + "loss": 0.5792, + "step": 614 + }, + { + "epoch": 1.2082514734774068, + "grad_norm": 0.4963095060478436, + "learning_rate": 7.522010580729819e-05, + "loss": 0.5682, + "step": 615 + }, + { + "epoch": 1.2102161100196462, + "grad_norm": 0.5398949037541292, + "learning_rate": 7.519405966898995e-05, + "loss": 0.5807, + "step": 616 + }, + { + "epoch": 1.2121807465618861, + "grad_norm": 0.5698321583314754, + "learning_rate": 7.516794729407453e-05, + "loss": 0.5782, + "step": 617 + }, + { + "epoch": 1.2141453831041258, + "grad_norm": 0.5729867481776637, + "learning_rate": 7.514176873169644e-05, + "loss": 0.5709, + "step": 618 + }, + { + "epoch": 1.2161100196463654, + "grad_norm": 0.4761562137415099, + "learning_rate": 7.511552403112477e-05, + "loss": 0.5761, + "step": 619 + }, + { + "epoch": 1.218074656188605, + "grad_norm": 0.6247622661589022, + "learning_rate": 7.50892132417531e-05, + "loss": 0.5749, + "step": 620 + }, + { + "epoch": 1.2200392927308448, + "grad_norm": 0.5828985057658796, + "learning_rate": 7.506283641309935e-05, + "loss": 0.5736, + "step": 621 + }, + { + "epoch": 1.2220039292730844, + "grad_norm": 0.39956594734733475, + "learning_rate": 7.503639359480576e-05, + "loss": 0.5755, + "step": 622 + }, + { + "epoch": 1.223968565815324, + "grad_norm": 0.41129374393479207, + "learning_rate": 7.500988483663873e-05, + "loss": 0.5744, + "step": 623 + }, + { + "epoch": 1.2259332023575638, + "grad_norm": 0.46497111317920176, + "learning_rate": 7.498331018848884e-05, + "loss": 0.5744, + "step": 624 + }, + { + "epoch": 1.2278978388998034, + "grad_norm": 0.5457154436144827, + "learning_rate": 7.495666970037058e-05, + "loss": 0.5706, + "step": 625 + }, + { + "epoch": 1.2298624754420433, + "grad_norm": 0.5392677848358564, + "learning_rate": 7.49299634224224e-05, + "loss": 0.5753, + "step": 626 + }, + { + "epoch": 1.231827111984283, + "grad_norm": 0.4613959152638453, + "learning_rate": 7.490319140490661e-05, + "loss": 0.5846, + "step": 627 + }, + { + "epoch": 1.2337917485265226, + "grad_norm": 0.6624715156862896, + "learning_rate": 7.487635369820916e-05, + "loss": 0.562, + "step": 628 + }, + { + "epoch": 1.2357563850687623, + "grad_norm": 0.8625616494642095, + "learning_rate": 7.484945035283968e-05, + "loss": 0.5787, + "step": 629 + }, + { + "epoch": 1.237721021611002, + "grad_norm": 1.0583451995350774, + "learning_rate": 7.482248141943133e-05, + "loss": 0.5801, + "step": 630 + }, + { + "epoch": 1.2396856581532416, + "grad_norm": 1.0924039276105026, + "learning_rate": 7.479544694874072e-05, + "loss": 0.579, + "step": 631 + }, + { + "epoch": 1.2416502946954813, + "grad_norm": 0.8209574271854506, + "learning_rate": 7.476834699164776e-05, + "loss": 0.5791, + "step": 632 + }, + { + "epoch": 1.243614931237721, + "grad_norm": 0.5826159960086854, + "learning_rate": 7.474118159915566e-05, + "loss": 0.5762, + "step": 633 + }, + { + "epoch": 1.2455795677799606, + "grad_norm": 0.6084552818048826, + "learning_rate": 7.471395082239075e-05, + "loss": 0.5751, + "step": 634 + }, + { + "epoch": 1.2475442043222005, + "grad_norm": 0.6526483977365647, + "learning_rate": 7.468665471260242e-05, + "loss": 0.582, + "step": 635 + }, + { + "epoch": 1.2495088408644401, + "grad_norm": 0.6634921515910455, + "learning_rate": 7.465929332116304e-05, + "loss": 0.5704, + "step": 636 + }, + { + "epoch": 1.2514734774066798, + "grad_norm": 0.6546076551743024, + "learning_rate": 7.463186669956782e-05, + "loss": 0.5758, + "step": 637 + }, + { + "epoch": 1.2534381139489195, + "grad_norm": 0.6276428224386265, + "learning_rate": 7.460437489943473e-05, + "loss": 0.5728, + "step": 638 + }, + { + "epoch": 1.2554027504911591, + "grad_norm": 0.6519224339420223, + "learning_rate": 7.457681797250445e-05, + "loss": 0.569, + "step": 639 + }, + { + "epoch": 1.2573673870333988, + "grad_norm": 0.6433759274463492, + "learning_rate": 7.454919597064017e-05, + "loss": 0.5722, + "step": 640 + }, + { + "epoch": 1.2593320235756384, + "grad_norm": 0.4754048515351831, + "learning_rate": 7.452150894582761e-05, + "loss": 0.5813, + "step": 641 + }, + { + "epoch": 1.261296660117878, + "grad_norm": 0.41582798244362196, + "learning_rate": 7.449375695017483e-05, + "loss": 0.5731, + "step": 642 + }, + { + "epoch": 1.2632612966601178, + "grad_norm": 0.6212145073849588, + "learning_rate": 7.446594003591216e-05, + "loss": 0.5769, + "step": 643 + }, + { + "epoch": 1.2652259332023577, + "grad_norm": 0.8723951538463921, + "learning_rate": 7.443805825539218e-05, + "loss": 0.5753, + "step": 644 + }, + { + "epoch": 1.2671905697445973, + "grad_norm": 1.1349095433805025, + "learning_rate": 7.441011166108945e-05, + "loss": 0.5913, + "step": 645 + }, + { + "epoch": 1.269155206286837, + "grad_norm": 0.7661507408152195, + "learning_rate": 7.438210030560058e-05, + "loss": 0.5754, + "step": 646 + }, + { + "epoch": 1.2711198428290766, + "grad_norm": 0.446407576937885, + "learning_rate": 7.435402424164405e-05, + "loss": 0.5711, + "step": 647 + }, + { + "epoch": 1.2730844793713163, + "grad_norm": 0.3672308270769251, + "learning_rate": 7.432588352206014e-05, + "loss": 0.5748, + "step": 648 + }, + { + "epoch": 1.275049115913556, + "grad_norm": 0.647914793554022, + "learning_rate": 7.429767819981073e-05, + "loss": 0.5789, + "step": 649 + }, + { + "epoch": 1.2770137524557956, + "grad_norm": 0.7429156395507586, + "learning_rate": 7.426940832797941e-05, + "loss": 0.5711, + "step": 650 + }, + { + "epoch": 1.2789783889980353, + "grad_norm": 0.7693059498944721, + "learning_rate": 7.424107395977116e-05, + "loss": 0.5754, + "step": 651 + }, + { + "epoch": 1.280943025540275, + "grad_norm": 0.7615264051206435, + "learning_rate": 7.42126751485124e-05, + "loss": 0.5796, + "step": 652 + }, + { + "epoch": 1.2829076620825148, + "grad_norm": 0.807133452514194, + "learning_rate": 7.418421194765078e-05, + "loss": 0.5803, + "step": 653 + }, + { + "epoch": 1.2848722986247545, + "grad_norm": 0.8562095428488848, + "learning_rate": 7.41556844107552e-05, + "loss": 0.5667, + "step": 654 + }, + { + "epoch": 1.2868369351669942, + "grad_norm": 0.8061058232885648, + "learning_rate": 7.412709259151558e-05, + "loss": 0.5811, + "step": 655 + }, + { + "epoch": 1.2888015717092338, + "grad_norm": 0.7185506589521004, + "learning_rate": 7.409843654374287e-05, + "loss": 0.575, + "step": 656 + }, + { + "epoch": 1.2907662082514735, + "grad_norm": 0.6932365431245716, + "learning_rate": 7.406971632136888e-05, + "loss": 0.5768, + "step": 657 + }, + { + "epoch": 1.2927308447937131, + "grad_norm": 0.5814069984232206, + "learning_rate": 7.404093197844619e-05, + "loss": 0.5825, + "step": 658 + }, + { + "epoch": 1.2946954813359528, + "grad_norm": 0.581204944936545, + "learning_rate": 7.401208356914806e-05, + "loss": 0.5793, + "step": 659 + }, + { + "epoch": 1.2966601178781925, + "grad_norm": 0.6558343740901095, + "learning_rate": 7.398317114776836e-05, + "loss": 0.5739, + "step": 660 + }, + { + "epoch": 1.2986247544204321, + "grad_norm": 0.8201596092340128, + "learning_rate": 7.395419476872139e-05, + "loss": 0.5818, + "step": 661 + }, + { + "epoch": 1.300589390962672, + "grad_norm": 0.8478949435234185, + "learning_rate": 7.392515448654184e-05, + "loss": 0.5736, + "step": 662 + }, + { + "epoch": 1.3025540275049117, + "grad_norm": 0.7345872252791336, + "learning_rate": 7.389605035588466e-05, + "loss": 0.5735, + "step": 663 + }, + { + "epoch": 1.3045186640471513, + "grad_norm": 0.5866547398180533, + "learning_rate": 7.386688243152499e-05, + "loss": 0.5791, + "step": 664 + }, + { + "epoch": 1.306483300589391, + "grad_norm": 0.6181859661454364, + "learning_rate": 7.383765076835799e-05, + "loss": 0.5677, + "step": 665 + }, + { + "epoch": 1.3084479371316307, + "grad_norm": 0.6566182123791899, + "learning_rate": 7.38083554213988e-05, + "loss": 0.5709, + "step": 666 + }, + { + "epoch": 1.3104125736738703, + "grad_norm": 0.7098127573215737, + "learning_rate": 7.377899644578244e-05, + "loss": 0.5835, + "step": 667 + }, + { + "epoch": 1.31237721021611, + "grad_norm": 0.8397847331220136, + "learning_rate": 7.374957389676364e-05, + "loss": 0.5785, + "step": 668 + }, + { + "epoch": 1.3143418467583496, + "grad_norm": 0.8862970300477165, + "learning_rate": 7.37200878297168e-05, + "loss": 0.5803, + "step": 669 + }, + { + "epoch": 1.3163064833005893, + "grad_norm": 0.7977397693664806, + "learning_rate": 7.369053830013588e-05, + "loss": 0.5638, + "step": 670 + }, + { + "epoch": 1.3182711198428292, + "grad_norm": 0.7169144574978591, + "learning_rate": 7.366092536363423e-05, + "loss": 0.5684, + "step": 671 + }, + { + "epoch": 1.3202357563850686, + "grad_norm": 0.7729930460806103, + "learning_rate": 7.363124907594458e-05, + "loss": 0.5721, + "step": 672 + }, + { + "epoch": 1.3222003929273085, + "grad_norm": 0.8750055810325454, + "learning_rate": 7.360150949291887e-05, + "loss": 0.5805, + "step": 673 + }, + { + "epoch": 1.3241650294695482, + "grad_norm": 0.837440550253823, + "learning_rate": 7.357170667052816e-05, + "loss": 0.5826, + "step": 674 + }, + { + "epoch": 1.3261296660117878, + "grad_norm": 0.5680080655748462, + "learning_rate": 7.354184066486256e-05, + "loss": 0.5714, + "step": 675 + }, + { + "epoch": 1.3280943025540275, + "grad_norm": 0.43522838272200914, + "learning_rate": 7.351191153213103e-05, + "loss": 0.5794, + "step": 676 + }, + { + "epoch": 1.3300589390962672, + "grad_norm": 0.5299616341028329, + "learning_rate": 7.348191932866141e-05, + "loss": 0.5732, + "step": 677 + }, + { + "epoch": 1.3320235756385068, + "grad_norm": 0.5260324328189155, + "learning_rate": 7.34518641109002e-05, + "loss": 0.5755, + "step": 678 + }, + { + "epoch": 1.3339882121807465, + "grad_norm": 0.5575669388697251, + "learning_rate": 7.34217459354125e-05, + "loss": 0.5688, + "step": 679 + }, + { + "epoch": 1.3359528487229864, + "grad_norm": 0.6663257000365891, + "learning_rate": 7.339156485888191e-05, + "loss": 0.5668, + "step": 680 + }, + { + "epoch": 1.3379174852652258, + "grad_norm": 0.7086388852729915, + "learning_rate": 7.33613209381104e-05, + "loss": 0.5745, + "step": 681 + }, + { + "epoch": 1.3398821218074657, + "grad_norm": 0.7526748310224963, + "learning_rate": 7.333101423001821e-05, + "loss": 0.5778, + "step": 682 + }, + { + "epoch": 1.3418467583497053, + "grad_norm": 0.874209447175276, + "learning_rate": 7.330064479164378e-05, + "loss": 0.5809, + "step": 683 + }, + { + "epoch": 1.343811394891945, + "grad_norm": 0.8082893355460667, + "learning_rate": 7.327021268014358e-05, + "loss": 0.5797, + "step": 684 + }, + { + "epoch": 1.3457760314341847, + "grad_norm": 0.6882629621255737, + "learning_rate": 7.323971795279205e-05, + "loss": 0.5772, + "step": 685 + }, + { + "epoch": 1.3477406679764243, + "grad_norm": 0.6887150415847766, + "learning_rate": 7.320916066698145e-05, + "loss": 0.5774, + "step": 686 + }, + { + "epoch": 1.349705304518664, + "grad_norm": 0.6536899614564937, + "learning_rate": 7.317854088022182e-05, + "loss": 0.5736, + "step": 687 + }, + { + "epoch": 1.3516699410609037, + "grad_norm": 0.5746976387549655, + "learning_rate": 7.31478586501408e-05, + "loss": 0.5743, + "step": 688 + }, + { + "epoch": 1.3536345776031435, + "grad_norm": 0.43262305374124715, + "learning_rate": 7.311711403448353e-05, + "loss": 0.57, + "step": 689 + }, + { + "epoch": 1.355599214145383, + "grad_norm": 0.5093012051409963, + "learning_rate": 7.308630709111261e-05, + "loss": 0.5729, + "step": 690 + }, + { + "epoch": 1.3575638506876229, + "grad_norm": 0.5608761163873807, + "learning_rate": 7.305543787800793e-05, + "loss": 0.5608, + "step": 691 + }, + { + "epoch": 1.3595284872298625, + "grad_norm": 0.5534366775584709, + "learning_rate": 7.302450645326654e-05, + "loss": 0.5776, + "step": 692 + }, + { + "epoch": 1.3614931237721022, + "grad_norm": 0.5327887187327054, + "learning_rate": 7.299351287510263e-05, + "loss": 0.5812, + "step": 693 + }, + { + "epoch": 1.3634577603143418, + "grad_norm": 0.5107522450095633, + "learning_rate": 7.296245720184731e-05, + "loss": 0.575, + "step": 694 + }, + { + "epoch": 1.3654223968565815, + "grad_norm": 0.37369375601610927, + "learning_rate": 7.293133949194859e-05, + "loss": 0.5735, + "step": 695 + }, + { + "epoch": 1.3673870333988212, + "grad_norm": 0.3784646346008778, + "learning_rate": 7.290015980397124e-05, + "loss": 0.5735, + "step": 696 + }, + { + "epoch": 1.3693516699410608, + "grad_norm": 0.5663719266854019, + "learning_rate": 7.286891819659664e-05, + "loss": 0.5599, + "step": 697 + }, + { + "epoch": 1.3713163064833007, + "grad_norm": 0.6940622106583807, + "learning_rate": 7.283761472862275e-05, + "loss": 0.571, + "step": 698 + }, + { + "epoch": 1.3732809430255402, + "grad_norm": 0.7816852692672728, + "learning_rate": 7.28062494589639e-05, + "loss": 0.5643, + "step": 699 + }, + { + "epoch": 1.37524557956778, + "grad_norm": 0.6448777922490226, + "learning_rate": 7.277482244665079e-05, + "loss": 0.5747, + "step": 700 + }, + { + "epoch": 1.3772102161100197, + "grad_norm": 0.5175492212626309, + "learning_rate": 7.274333375083027e-05, + "loss": 0.5745, + "step": 701 + }, + { + "epoch": 1.3791748526522594, + "grad_norm": 0.4864448683056957, + "learning_rate": 7.271178343076532e-05, + "loss": 0.5557, + "step": 702 + }, + { + "epoch": 1.381139489194499, + "grad_norm": 0.424048413234014, + "learning_rate": 7.26801715458349e-05, + "loss": 0.5812, + "step": 703 + }, + { + "epoch": 1.3831041257367387, + "grad_norm": 0.3813383027080297, + "learning_rate": 7.26484981555338e-05, + "loss": 0.5814, + "step": 704 + }, + { + "epoch": 1.3850687622789783, + "grad_norm": 0.3946718013778929, + "learning_rate": 7.261676331947261e-05, + "loss": 0.5742, + "step": 705 + }, + { + "epoch": 1.387033398821218, + "grad_norm": 0.41860783877393126, + "learning_rate": 7.258496709737753e-05, + "loss": 0.5762, + "step": 706 + }, + { + "epoch": 1.388998035363458, + "grad_norm": 0.48878017537286156, + "learning_rate": 7.255310954909028e-05, + "loss": 0.5631, + "step": 707 + }, + { + "epoch": 1.3909626719056973, + "grad_norm": 0.543960104371802, + "learning_rate": 7.252119073456808e-05, + "loss": 0.5797, + "step": 708 + }, + { + "epoch": 1.3929273084479372, + "grad_norm": 0.6066582779301862, + "learning_rate": 7.248921071388333e-05, + "loss": 0.5629, + "step": 709 + }, + { + "epoch": 1.3948919449901769, + "grad_norm": 0.5773427443262497, + "learning_rate": 7.245716954722374e-05, + "loss": 0.573, + "step": 710 + }, + { + "epoch": 1.3968565815324165, + "grad_norm": 0.5744916559159152, + "learning_rate": 7.242506729489203e-05, + "loss": 0.5736, + "step": 711 + }, + { + "epoch": 1.3988212180746562, + "grad_norm": 0.5876683049890943, + "learning_rate": 7.23929040173059e-05, + "loss": 0.5769, + "step": 712 + }, + { + "epoch": 1.4007858546168959, + "grad_norm": 0.662260570216204, + "learning_rate": 7.236067977499791e-05, + "loss": 0.5786, + "step": 713 + }, + { + "epoch": 1.4027504911591355, + "grad_norm": 0.7568027192225558, + "learning_rate": 7.232839462861535e-05, + "loss": 0.5672, + "step": 714 + }, + { + "epoch": 1.4047151277013752, + "grad_norm": 0.800607600135809, + "learning_rate": 7.229604863892014e-05, + "loss": 0.5747, + "step": 715 + }, + { + "epoch": 1.4066797642436148, + "grad_norm": 0.8860685369604828, + "learning_rate": 7.226364186678869e-05, + "loss": 0.5814, + "step": 716 + }, + { + "epoch": 1.4086444007858545, + "grad_norm": 1.058588691281162, + "learning_rate": 7.223117437321187e-05, + "loss": 0.5714, + "step": 717 + }, + { + "epoch": 1.4106090373280944, + "grad_norm": 0.9569549620491247, + "learning_rate": 7.219864621929474e-05, + "loss": 0.5824, + "step": 718 + }, + { + "epoch": 1.412573673870334, + "grad_norm": 0.7955066957158038, + "learning_rate": 7.216605746625658e-05, + "loss": 0.5821, + "step": 719 + }, + { + "epoch": 1.4145383104125737, + "grad_norm": 0.6084587871309047, + "learning_rate": 7.21334081754307e-05, + "loss": 0.5769, + "step": 720 + }, + { + "epoch": 1.4165029469548134, + "grad_norm": 0.5758380790178109, + "learning_rate": 7.210069840826437e-05, + "loss": 0.5744, + "step": 721 + }, + { + "epoch": 1.418467583497053, + "grad_norm": 0.5456868990914795, + "learning_rate": 7.206792822631864e-05, + "loss": 0.5735, + "step": 722 + }, + { + "epoch": 1.4204322200392927, + "grad_norm": 0.49760623601812587, + "learning_rate": 7.203509769126831e-05, + "loss": 0.5738, + "step": 723 + }, + { + "epoch": 1.4223968565815324, + "grad_norm": 0.5288191057548994, + "learning_rate": 7.20022068649017e-05, + "loss": 0.5725, + "step": 724 + }, + { + "epoch": 1.424361493123772, + "grad_norm": 0.5052498395309702, + "learning_rate": 7.19692558091207e-05, + "loss": 0.5775, + "step": 725 + }, + { + "epoch": 1.4263261296660117, + "grad_norm": 0.46623755782024745, + "learning_rate": 7.193624458594043e-05, + "loss": 0.5773, + "step": 726 + }, + { + "epoch": 1.4282907662082516, + "grad_norm": 0.3672818610037674, + "learning_rate": 7.190317325748936e-05, + "loss": 0.5725, + "step": 727 + }, + { + "epoch": 1.4302554027504912, + "grad_norm": 0.35959243520731715, + "learning_rate": 7.187004188600901e-05, + "loss": 0.57, + "step": 728 + }, + { + "epoch": 1.432220039292731, + "grad_norm": 0.4532071264537897, + "learning_rate": 7.183685053385393e-05, + "loss": 0.5867, + "step": 729 + }, + { + "epoch": 1.4341846758349706, + "grad_norm": 0.5649675789427756, + "learning_rate": 7.180359926349158e-05, + "loss": 0.5751, + "step": 730 + }, + { + "epoch": 1.4361493123772102, + "grad_norm": 0.6550378812733314, + "learning_rate": 7.177028813750211e-05, + "loss": 0.58, + "step": 731 + }, + { + "epoch": 1.4381139489194499, + "grad_norm": 0.6926538177015795, + "learning_rate": 7.173691721857841e-05, + "loss": 0.5695, + "step": 732 + }, + { + "epoch": 1.4400785854616895, + "grad_norm": 0.7918278199816393, + "learning_rate": 7.170348656952586e-05, + "loss": 0.5733, + "step": 733 + }, + { + "epoch": 1.4420432220039292, + "grad_norm": 0.7970451740087121, + "learning_rate": 7.166999625326225e-05, + "loss": 0.573, + "step": 734 + }, + { + "epoch": 1.4440078585461689, + "grad_norm": 0.8550122770089573, + "learning_rate": 7.163644633281769e-05, + "loss": 0.5632, + "step": 735 + }, + { + "epoch": 1.4459724950884087, + "grad_norm": 0.8011332014855559, + "learning_rate": 7.160283687133442e-05, + "loss": 0.5775, + "step": 736 + }, + { + "epoch": 1.4479371316306484, + "grad_norm": 0.5712023340176959, + "learning_rate": 7.15691679320668e-05, + "loss": 0.5813, + "step": 737 + }, + { + "epoch": 1.449901768172888, + "grad_norm": 0.4415948703800342, + "learning_rate": 7.15354395783811e-05, + "loss": 0.5719, + "step": 738 + }, + { + "epoch": 1.4518664047151277, + "grad_norm": 0.4775828744307346, + "learning_rate": 7.150165187375539e-05, + "loss": 0.5655, + "step": 739 + }, + { + "epoch": 1.4538310412573674, + "grad_norm": 0.45130932219566433, + "learning_rate": 7.146780488177948e-05, + "loss": 0.5718, + "step": 740 + }, + { + "epoch": 1.455795677799607, + "grad_norm": 0.41025558932793627, + "learning_rate": 7.143389866615475e-05, + "loss": 0.5837, + "step": 741 + }, + { + "epoch": 1.4577603143418467, + "grad_norm": 0.44627374423145433, + "learning_rate": 7.139993329069401e-05, + "loss": 0.5667, + "step": 742 + }, + { + "epoch": 1.4597249508840864, + "grad_norm": 0.4989807171081548, + "learning_rate": 7.136590881932144e-05, + "loss": 0.5788, + "step": 743 + }, + { + "epoch": 1.461689587426326, + "grad_norm": 0.49791547401879244, + "learning_rate": 7.133182531607245e-05, + "loss": 0.5768, + "step": 744 + }, + { + "epoch": 1.463654223968566, + "grad_norm": 0.510579673166167, + "learning_rate": 7.129768284509352e-05, + "loss": 0.5688, + "step": 745 + }, + { + "epoch": 1.4656188605108054, + "grad_norm": 0.5124110243995726, + "learning_rate": 7.126348147064214e-05, + "loss": 0.566, + "step": 746 + }, + { + "epoch": 1.4675834970530452, + "grad_norm": 0.4589984366520699, + "learning_rate": 7.122922125708665e-05, + "loss": 0.5861, + "step": 747 + }, + { + "epoch": 1.469548133595285, + "grad_norm": 0.520602567174116, + "learning_rate": 7.119490226890608e-05, + "loss": 0.5744, + "step": 748 + }, + { + "epoch": 1.4715127701375246, + "grad_norm": 0.47242274813846336, + "learning_rate": 7.116052457069018e-05, + "loss": 0.578, + "step": 749 + }, + { + "epoch": 1.4734774066797642, + "grad_norm": 0.5053981749234949, + "learning_rate": 7.112608822713908e-05, + "loss": 0.57, + "step": 750 + }, + { + "epoch": 1.475442043222004, + "grad_norm": 0.5458408802429001, + "learning_rate": 7.109159330306338e-05, + "loss": 0.5734, + "step": 751 + }, + { + "epoch": 1.4774066797642436, + "grad_norm": 0.4711244195283438, + "learning_rate": 7.105703986338384e-05, + "loss": 0.5602, + "step": 752 + }, + { + "epoch": 1.4793713163064832, + "grad_norm": 0.423007079188366, + "learning_rate": 7.102242797313143e-05, + "loss": 0.5901, + "step": 753 + }, + { + "epoch": 1.481335952848723, + "grad_norm": 0.38591907720662005, + "learning_rate": 7.098775769744708e-05, + "loss": 0.5752, + "step": 754 + }, + { + "epoch": 1.4833005893909625, + "grad_norm": 0.4088197276183067, + "learning_rate": 7.095302910158161e-05, + "loss": 0.5616, + "step": 755 + }, + { + "epoch": 1.4852652259332024, + "grad_norm": 0.5170900522472109, + "learning_rate": 7.091824225089563e-05, + "loss": 0.5797, + "step": 756 + }, + { + "epoch": 1.487229862475442, + "grad_norm": 0.5023378076120734, + "learning_rate": 7.088339721085933e-05, + "loss": 0.5744, + "step": 757 + }, + { + "epoch": 1.4891944990176817, + "grad_norm": 0.43995075437076375, + "learning_rate": 7.084849404705245e-05, + "loss": 0.57, + "step": 758 + }, + { + "epoch": 1.4911591355599214, + "grad_norm": 0.444916568283222, + "learning_rate": 7.081353282516416e-05, + "loss": 0.5805, + "step": 759 + }, + { + "epoch": 1.493123772102161, + "grad_norm": 0.468343753265261, + "learning_rate": 7.077851361099281e-05, + "loss": 0.5628, + "step": 760 + }, + { + "epoch": 1.4950884086444007, + "grad_norm": 0.5871929546432679, + "learning_rate": 7.074343647044596e-05, + "loss": 0.5652, + "step": 761 + }, + { + "epoch": 1.4970530451866404, + "grad_norm": 0.8883962582725597, + "learning_rate": 7.070830146954018e-05, + "loss": 0.5772, + "step": 762 + }, + { + "epoch": 1.4990176817288803, + "grad_norm": 1.0476923662882967, + "learning_rate": 7.067310867440093e-05, + "loss": 0.5707, + "step": 763 + }, + { + "epoch": 1.5009823182711197, + "grad_norm": 1.03922658756807, + "learning_rate": 7.063785815126241e-05, + "loss": 0.565, + "step": 764 + }, + { + "epoch": 1.5029469548133596, + "grad_norm": 1.0355915485078073, + "learning_rate": 7.060254996646752e-05, + "loss": 0.5788, + "step": 765 + }, + { + "epoch": 1.5049115913555993, + "grad_norm": 0.8757922891938862, + "learning_rate": 7.056718418646764e-05, + "loss": 0.5639, + "step": 766 + }, + { + "epoch": 1.506876227897839, + "grad_norm": 0.7094232492853936, + "learning_rate": 7.053176087782256e-05, + "loss": 0.5786, + "step": 767 + }, + { + "epoch": 1.5088408644400786, + "grad_norm": 0.6905380739224084, + "learning_rate": 7.049628010720036e-05, + "loss": 0.5684, + "step": 768 + }, + { + "epoch": 1.5108055009823183, + "grad_norm": 0.7791991577965104, + "learning_rate": 7.046074194137723e-05, + "loss": 0.5837, + "step": 769 + }, + { + "epoch": 1.512770137524558, + "grad_norm": 0.8930273173023473, + "learning_rate": 7.042514644723741e-05, + "loss": 0.5752, + "step": 770 + }, + { + "epoch": 1.5147347740667976, + "grad_norm": 0.8640550333281133, + "learning_rate": 7.038949369177303e-05, + "loss": 0.5658, + "step": 771 + }, + { + "epoch": 1.5166994106090375, + "grad_norm": 0.6340139227733929, + "learning_rate": 7.035378374208395e-05, + "loss": 0.5634, + "step": 772 + }, + { + "epoch": 1.518664047151277, + "grad_norm": 0.463138446176254, + "learning_rate": 7.031801666537772e-05, + "loss": 0.5651, + "step": 773 + }, + { + "epoch": 1.5206286836935168, + "grad_norm": 0.5808137260773966, + "learning_rate": 7.02821925289694e-05, + "loss": 0.5752, + "step": 774 + }, + { + "epoch": 1.5225933202357562, + "grad_norm": 0.6810692316479032, + "learning_rate": 7.024631140028142e-05, + "loss": 0.5634, + "step": 775 + }, + { + "epoch": 1.524557956777996, + "grad_norm": 0.6426382028447118, + "learning_rate": 7.021037334684347e-05, + "loss": 0.5762, + "step": 776 + }, + { + "epoch": 1.5265225933202358, + "grad_norm": 0.5081498647902154, + "learning_rate": 7.017437843629239e-05, + "loss": 0.567, + "step": 777 + }, + { + "epoch": 1.5284872298624754, + "grad_norm": 0.3386448679888073, + "learning_rate": 7.013832673637202e-05, + "loss": 0.5748, + "step": 778 + }, + { + "epoch": 1.530451866404715, + "grad_norm": 0.3986613883068482, + "learning_rate": 7.010221831493307e-05, + "loss": 0.5715, + "step": 779 + }, + { + "epoch": 1.5324165029469548, + "grad_norm": 0.6121175747173112, + "learning_rate": 7.006605323993301e-05, + "loss": 0.5721, + "step": 780 + }, + { + "epoch": 1.5343811394891946, + "grad_norm": 0.6278084319213263, + "learning_rate": 7.002983157943596e-05, + "loss": 0.5746, + "step": 781 + }, + { + "epoch": 1.536345776031434, + "grad_norm": 0.5216531752295729, + "learning_rate": 6.99935534016125e-05, + "loss": 0.5692, + "step": 782 + }, + { + "epoch": 1.538310412573674, + "grad_norm": 0.5680510812318424, + "learning_rate": 6.995721877473958e-05, + "loss": 0.5733, + "step": 783 + }, + { + "epoch": 1.5402750491159134, + "grad_norm": 0.5711048279189166, + "learning_rate": 6.992082776720041e-05, + "loss": 0.5761, + "step": 784 + }, + { + "epoch": 1.5422396856581533, + "grad_norm": 0.528853481338425, + "learning_rate": 6.98843804474843e-05, + "loss": 0.574, + "step": 785 + }, + { + "epoch": 1.544204322200393, + "grad_norm": 0.5723205440932766, + "learning_rate": 6.984787688418654e-05, + "loss": 0.5696, + "step": 786 + }, + { + "epoch": 1.5461689587426326, + "grad_norm": 0.6791165793602034, + "learning_rate": 6.981131714600828e-05, + "loss": 0.5651, + "step": 787 + }, + { + "epoch": 1.5481335952848723, + "grad_norm": 0.76440289971057, + "learning_rate": 6.977470130175637e-05, + "loss": 0.5726, + "step": 788 + }, + { + "epoch": 1.550098231827112, + "grad_norm": 0.7705019546582054, + "learning_rate": 6.973802942034329e-05, + "loss": 0.5626, + "step": 789 + }, + { + "epoch": 1.5520628683693518, + "grad_norm": 0.7636752584505546, + "learning_rate": 6.970130157078697e-05, + "loss": 0.5755, + "step": 790 + }, + { + "epoch": 1.5540275049115913, + "grad_norm": 0.6654947497728212, + "learning_rate": 6.966451782221065e-05, + "loss": 0.5805, + "step": 791 + }, + { + "epoch": 1.5559921414538311, + "grad_norm": 0.6378057617691023, + "learning_rate": 6.96276782438428e-05, + "loss": 0.5734, + "step": 792 + }, + { + "epoch": 1.5579567779960706, + "grad_norm": 0.6579869614508671, + "learning_rate": 6.959078290501698e-05, + "loss": 0.5738, + "step": 793 + }, + { + "epoch": 1.5599214145383105, + "grad_norm": 0.6133266155762558, + "learning_rate": 6.955383187517164e-05, + "loss": 0.578, + "step": 794 + }, + { + "epoch": 1.5618860510805501, + "grad_norm": 0.46787783113262355, + "learning_rate": 6.95168252238501e-05, + "loss": 0.5704, + "step": 795 + }, + { + "epoch": 1.5638506876227898, + "grad_norm": 0.5286680452992167, + "learning_rate": 6.947976302070032e-05, + "loss": 0.5697, + "step": 796 + }, + { + "epoch": 1.5658153241650294, + "grad_norm": 0.6100756240101574, + "learning_rate": 6.944264533547484e-05, + "loss": 0.5681, + "step": 797 + }, + { + "epoch": 1.567779960707269, + "grad_norm": 0.45320300922756657, + "learning_rate": 6.940547223803058e-05, + "loss": 0.5658, + "step": 798 + }, + { + "epoch": 1.569744597249509, + "grad_norm": 0.3973953763903243, + "learning_rate": 6.93682437983288e-05, + "loss": 0.5744, + "step": 799 + }, + { + "epoch": 1.5717092337917484, + "grad_norm": 0.4313361694443693, + "learning_rate": 6.933096008643489e-05, + "loss": 0.5739, + "step": 800 + }, + { + "epoch": 1.5736738703339883, + "grad_norm": 0.38291954039436477, + "learning_rate": 6.929362117251826e-05, + "loss": 0.5643, + "step": 801 + }, + { + "epoch": 1.5756385068762278, + "grad_norm": 0.3248455722081132, + "learning_rate": 6.925622712685221e-05, + "loss": 0.5705, + "step": 802 + }, + { + "epoch": 1.5776031434184676, + "grad_norm": 0.2929743816096528, + "learning_rate": 6.921877801981382e-05, + "loss": 0.5676, + "step": 803 + }, + { + "epoch": 1.5795677799607073, + "grad_norm": 0.31937049082410357, + "learning_rate": 6.918127392188377e-05, + "loss": 0.5768, + "step": 804 + }, + { + "epoch": 1.581532416502947, + "grad_norm": 0.43090341301568585, + "learning_rate": 6.914371490364625e-05, + "loss": 0.5626, + "step": 805 + }, + { + "epoch": 1.5834970530451866, + "grad_norm": 0.43870696996058584, + "learning_rate": 6.910610103578884e-05, + "loss": 0.5723, + "step": 806 + }, + { + "epoch": 1.5854616895874263, + "grad_norm": 0.42822672424577085, + "learning_rate": 6.906843238910229e-05, + "loss": 0.5724, + "step": 807 + }, + { + "epoch": 1.5874263261296662, + "grad_norm": 0.4042844142487971, + "learning_rate": 6.90307090344805e-05, + "loss": 0.5747, + "step": 808 + }, + { + "epoch": 1.5893909626719056, + "grad_norm": 0.40045006914468806, + "learning_rate": 6.899293104292028e-05, + "loss": 0.5718, + "step": 809 + }, + { + "epoch": 1.5913555992141455, + "grad_norm": 0.4409447968664627, + "learning_rate": 6.895509848552134e-05, + "loss": 0.5632, + "step": 810 + }, + { + "epoch": 1.593320235756385, + "grad_norm": 0.4954683509442579, + "learning_rate": 6.891721143348601e-05, + "loss": 0.5597, + "step": 811 + }, + { + "epoch": 1.5952848722986248, + "grad_norm": 0.48802830856958906, + "learning_rate": 6.887926995811924e-05, + "loss": 0.5665, + "step": 812 + }, + { + "epoch": 1.5972495088408645, + "grad_norm": 0.5005324410432559, + "learning_rate": 6.884127413082836e-05, + "loss": 0.5672, + "step": 813 + }, + { + "epoch": 1.5992141453831041, + "grad_norm": 0.5579199477903329, + "learning_rate": 6.880322402312302e-05, + "loss": 0.581, + "step": 814 + }, + { + "epoch": 1.6011787819253438, + "grad_norm": 0.7261399503952658, + "learning_rate": 6.876511970661505e-05, + "loss": 0.5781, + "step": 815 + }, + { + "epoch": 1.6031434184675835, + "grad_norm": 0.907532178242556, + "learning_rate": 6.872696125301823e-05, + "loss": 0.5723, + "step": 816 + }, + { + "epoch": 1.6051080550098233, + "grad_norm": 1.0216414433710386, + "learning_rate": 6.86887487341483e-05, + "loss": 0.5776, + "step": 817 + }, + { + "epoch": 1.6070726915520628, + "grad_norm": 0.8555049312968269, + "learning_rate": 6.86504822219227e-05, + "loss": 0.5809, + "step": 818 + }, + { + "epoch": 1.6090373280943027, + "grad_norm": 0.531390712989528, + "learning_rate": 6.861216178836055e-05, + "loss": 0.5784, + "step": 819 + }, + { + "epoch": 1.611001964636542, + "grad_norm": 0.33033391264977857, + "learning_rate": 6.85737875055824e-05, + "loss": 0.568, + "step": 820 + }, + { + "epoch": 1.612966601178782, + "grad_norm": 0.38730391790695634, + "learning_rate": 6.853535944581017e-05, + "loss": 0.5778, + "step": 821 + }, + { + "epoch": 1.6149312377210217, + "grad_norm": 0.5094992872908856, + "learning_rate": 6.849687768136697e-05, + "loss": 0.5728, + "step": 822 + }, + { + "epoch": 1.6168958742632613, + "grad_norm": 0.5107321227016961, + "learning_rate": 6.845834228467702e-05, + "loss": 0.5823, + "step": 823 + }, + { + "epoch": 1.618860510805501, + "grad_norm": 0.40784580633217815, + "learning_rate": 6.841975332826542e-05, + "loss": 0.5753, + "step": 824 + }, + { + "epoch": 1.6208251473477406, + "grad_norm": 0.32466335354968195, + "learning_rate": 6.838111088475812e-05, + "loss": 0.5724, + "step": 825 + }, + { + "epoch": 1.6227897838899805, + "grad_norm": 0.3236152206208413, + "learning_rate": 6.834241502688173e-05, + "loss": 0.5618, + "step": 826 + }, + { + "epoch": 1.62475442043222, + "grad_norm": 0.36444146308318165, + "learning_rate": 6.830366582746338e-05, + "loss": 0.5781, + "step": 827 + }, + { + "epoch": 1.6267190569744598, + "grad_norm": 0.4276736794933113, + "learning_rate": 6.82648633594306e-05, + "loss": 0.5829, + "step": 828 + }, + { + "epoch": 1.6286836935166993, + "grad_norm": 0.45409808486891906, + "learning_rate": 6.822600769581113e-05, + "loss": 0.566, + "step": 829 + }, + { + "epoch": 1.6306483300589392, + "grad_norm": 0.4460958953763306, + "learning_rate": 6.818709890973289e-05, + "loss": 0.5613, + "step": 830 + }, + { + "epoch": 1.6326129666011788, + "grad_norm": 0.46172992612900016, + "learning_rate": 6.814813707442372e-05, + "loss": 0.5722, + "step": 831 + }, + { + "epoch": 1.6345776031434185, + "grad_norm": 0.4445482326795445, + "learning_rate": 6.810912226321136e-05, + "loss": 0.5727, + "step": 832 + }, + { + "epoch": 1.6365422396856582, + "grad_norm": 0.5224031548106275, + "learning_rate": 6.807005454952321e-05, + "loss": 0.5683, + "step": 833 + }, + { + "epoch": 1.6385068762278978, + "grad_norm": 0.618713326636943, + "learning_rate": 6.803093400688626e-05, + "loss": 0.5816, + "step": 834 + }, + { + "epoch": 1.6404715127701375, + "grad_norm": 0.7757376486427527, + "learning_rate": 6.799176070892689e-05, + "loss": 0.5747, + "step": 835 + }, + { + "epoch": 1.6424361493123771, + "grad_norm": 0.9024928436373735, + "learning_rate": 6.79525347293708e-05, + "loss": 0.5706, + "step": 836 + }, + { + "epoch": 1.644400785854617, + "grad_norm": 0.9283122771990955, + "learning_rate": 6.791325614204284e-05, + "loss": 0.5728, + "step": 837 + }, + { + "epoch": 1.6463654223968565, + "grad_norm": 0.8865379940250355, + "learning_rate": 6.787392502086682e-05, + "loss": 0.5644, + "step": 838 + }, + { + "epoch": 1.6483300589390963, + "grad_norm": 0.8566715233696953, + "learning_rate": 6.78345414398655e-05, + "loss": 0.5758, + "step": 839 + }, + { + "epoch": 1.650294695481336, + "grad_norm": 0.923515639502809, + "learning_rate": 6.779510547316034e-05, + "loss": 0.5688, + "step": 840 + }, + { + "epoch": 1.6522593320235757, + "grad_norm": 0.9178986966835758, + "learning_rate": 6.775561719497135e-05, + "loss": 0.5779, + "step": 841 + }, + { + "epoch": 1.6542239685658153, + "grad_norm": 0.8240610754683876, + "learning_rate": 6.771607667961704e-05, + "loss": 0.5746, + "step": 842 + }, + { + "epoch": 1.656188605108055, + "grad_norm": 0.707433326873584, + "learning_rate": 6.767648400151421e-05, + "loss": 0.5603, + "step": 843 + }, + { + "epoch": 1.6581532416502947, + "grad_norm": 0.4978395029914036, + "learning_rate": 6.763683923517785e-05, + "loss": 0.5678, + "step": 844 + }, + { + "epoch": 1.6601178781925343, + "grad_norm": 0.3985153258809859, + "learning_rate": 6.759714245522098e-05, + "loss": 0.5778, + "step": 845 + }, + { + "epoch": 1.6620825147347742, + "grad_norm": 0.5141701248487086, + "learning_rate": 6.755739373635449e-05, + "loss": 0.5698, + "step": 846 + }, + { + "epoch": 1.6640471512770136, + "grad_norm": 0.6510825415364366, + "learning_rate": 6.751759315338704e-05, + "loss": 0.5668, + "step": 847 + }, + { + "epoch": 1.6660117878192535, + "grad_norm": 0.6377104601447621, + "learning_rate": 6.747774078122489e-05, + "loss": 0.5719, + "step": 848 + }, + { + "epoch": 1.667976424361493, + "grad_norm": 0.43796770161671145, + "learning_rate": 6.743783669487178e-05, + "loss": 0.5725, + "step": 849 + }, + { + "epoch": 1.6699410609037328, + "grad_norm": 0.32873247802708633, + "learning_rate": 6.739788096942879e-05, + "loss": 0.5699, + "step": 850 + }, + { + "epoch": 1.6719056974459725, + "grad_norm": 0.4707483153029061, + "learning_rate": 6.735787368009413e-05, + "loss": 0.5627, + "step": 851 + }, + { + "epoch": 1.6738703339882122, + "grad_norm": 0.5644136420209869, + "learning_rate": 6.731781490216313e-05, + "loss": 0.5792, + "step": 852 + }, + { + "epoch": 1.6758349705304518, + "grad_norm": 0.4803740876151933, + "learning_rate": 6.727770471102796e-05, + "loss": 0.5655, + "step": 853 + }, + { + "epoch": 1.6777996070726915, + "grad_norm": 0.4411958411601802, + "learning_rate": 6.723754318217762e-05, + "loss": 0.5759, + "step": 854 + }, + { + "epoch": 1.6797642436149314, + "grad_norm": 0.3513551513047615, + "learning_rate": 6.719733039119765e-05, + "loss": 0.5753, + "step": 855 + }, + { + "epoch": 1.6817288801571708, + "grad_norm": 0.27971674097320015, + "learning_rate": 6.715706641377015e-05, + "loss": 0.5627, + "step": 856 + }, + { + "epoch": 1.6836935166994107, + "grad_norm": 0.40131384062835423, + "learning_rate": 6.711675132567346e-05, + "loss": 0.5672, + "step": 857 + }, + { + "epoch": 1.6856581532416501, + "grad_norm": 0.5160262826053634, + "learning_rate": 6.70763852027822e-05, + "loss": 0.573, + "step": 858 + }, + { + "epoch": 1.68762278978389, + "grad_norm": 0.4828686346555754, + "learning_rate": 6.703596812106698e-05, + "loss": 0.5763, + "step": 859 + }, + { + "epoch": 1.6895874263261297, + "grad_norm": 0.45279624206598074, + "learning_rate": 6.699550015659437e-05, + "loss": 0.5592, + "step": 860 + }, + { + "epoch": 1.6915520628683693, + "grad_norm": 0.4390313239298527, + "learning_rate": 6.695498138552666e-05, + "loss": 0.572, + "step": 861 + }, + { + "epoch": 1.693516699410609, + "grad_norm": 0.4216492472934115, + "learning_rate": 6.691441188412178e-05, + "loss": 0.5661, + "step": 862 + }, + { + "epoch": 1.6954813359528487, + "grad_norm": 0.3920970740691836, + "learning_rate": 6.68737917287331e-05, + "loss": 0.5643, + "step": 863 + }, + { + "epoch": 1.6974459724950886, + "grad_norm": 0.4430109012531991, + "learning_rate": 6.68331209958094e-05, + "loss": 0.5678, + "step": 864 + }, + { + "epoch": 1.699410609037328, + "grad_norm": 0.5500001772923422, + "learning_rate": 6.679239976189457e-05, + "loss": 0.5759, + "step": 865 + }, + { + "epoch": 1.7013752455795679, + "grad_norm": 0.590668581335685, + "learning_rate": 6.675162810362759e-05, + "loss": 0.5717, + "step": 866 + }, + { + "epoch": 1.7033398821218073, + "grad_norm": 0.5254199114316701, + "learning_rate": 6.671080609774231e-05, + "loss": 0.5704, + "step": 867 + }, + { + "epoch": 1.7053045186640472, + "grad_norm": 0.5573040222498371, + "learning_rate": 6.666993382106738e-05, + "loss": 0.5696, + "step": 868 + }, + { + "epoch": 1.7072691552062869, + "grad_norm": 0.6348350010401239, + "learning_rate": 6.662901135052602e-05, + "loss": 0.5681, + "step": 869 + }, + { + "epoch": 1.7092337917485265, + "grad_norm": 0.6424532845377107, + "learning_rate": 6.658803876313592e-05, + "loss": 0.5739, + "step": 870 + }, + { + "epoch": 1.7111984282907662, + "grad_norm": 0.6295667805166777, + "learning_rate": 6.654701613600914e-05, + "loss": 0.5784, + "step": 871 + }, + { + "epoch": 1.7131630648330058, + "grad_norm": 0.6520468285795562, + "learning_rate": 6.650594354635185e-05, + "loss": 0.5684, + "step": 872 + }, + { + "epoch": 1.7151277013752457, + "grad_norm": 0.5536256913104903, + "learning_rate": 6.646482107146428e-05, + "loss": 0.5725, + "step": 873 + }, + { + "epoch": 1.7170923379174852, + "grad_norm": 0.48289127053359165, + "learning_rate": 6.642364878874057e-05, + "loss": 0.5607, + "step": 874 + }, + { + "epoch": 1.719056974459725, + "grad_norm": 0.438002995765159, + "learning_rate": 6.638242677566856e-05, + "loss": 0.5801, + "step": 875 + }, + { + "epoch": 1.7210216110019645, + "grad_norm": 0.42839594298722333, + "learning_rate": 6.634115510982973e-05, + "loss": 0.5709, + "step": 876 + }, + { + "epoch": 1.7229862475442044, + "grad_norm": 0.5541710509309423, + "learning_rate": 6.629983386889896e-05, + "loss": 0.5738, + "step": 877 + }, + { + "epoch": 1.724950884086444, + "grad_norm": 0.7387193805579535, + "learning_rate": 6.625846313064445e-05, + "loss": 0.5627, + "step": 878 + }, + { + "epoch": 1.7269155206286837, + "grad_norm": 0.8628006885661896, + "learning_rate": 6.621704297292755e-05, + "loss": 0.5617, + "step": 879 + }, + { + "epoch": 1.7288801571709234, + "grad_norm": 0.8131901525702913, + "learning_rate": 6.617557347370266e-05, + "loss": 0.5648, + "step": 880 + }, + { + "epoch": 1.730844793713163, + "grad_norm": 0.6453069553687197, + "learning_rate": 6.613405471101701e-05, + "loss": 0.5715, + "step": 881 + }, + { + "epoch": 1.732809430255403, + "grad_norm": 0.4374102250215347, + "learning_rate": 6.609248676301052e-05, + "loss": 0.5725, + "step": 882 + }, + { + "epoch": 1.7347740667976423, + "grad_norm": 0.421710750388363, + "learning_rate": 6.605086970791572e-05, + "loss": 0.5683, + "step": 883 + }, + { + "epoch": 1.7367387033398822, + "grad_norm": 0.4571434256618217, + "learning_rate": 6.600920362405755e-05, + "loss": 0.5808, + "step": 884 + }, + { + "epoch": 1.7387033398821217, + "grad_norm": 0.4054709613385462, + "learning_rate": 6.596748858985323e-05, + "loss": 0.5672, + "step": 885 + }, + { + "epoch": 1.7406679764243616, + "grad_norm": 0.3986314902621721, + "learning_rate": 6.592572468381207e-05, + "loss": 0.5739, + "step": 886 + }, + { + "epoch": 1.7426326129666012, + "grad_norm": 0.5208273656648287, + "learning_rate": 6.588391198453542e-05, + "loss": 0.5703, + "step": 887 + }, + { + "epoch": 1.7445972495088409, + "grad_norm": 0.6013370670648406, + "learning_rate": 6.584205057071641e-05, + "loss": 0.5671, + "step": 888 + }, + { + "epoch": 1.7465618860510805, + "grad_norm": 0.5843268579440807, + "learning_rate": 6.580014052113986e-05, + "loss": 0.5595, + "step": 889 + }, + { + "epoch": 1.7485265225933202, + "grad_norm": 0.5352415968072812, + "learning_rate": 6.575818191468217e-05, + "loss": 0.5581, + "step": 890 + }, + { + "epoch": 1.75049115913556, + "grad_norm": 0.5269041097454885, + "learning_rate": 6.571617483031105e-05, + "loss": 0.5713, + "step": 891 + }, + { + "epoch": 1.7524557956777995, + "grad_norm": 0.476373603399162, + "learning_rate": 6.567411934708553e-05, + "loss": 0.5775, + "step": 892 + }, + { + "epoch": 1.7544204322200394, + "grad_norm": 0.39142981138936667, + "learning_rate": 6.563201554415565e-05, + "loss": 0.5737, + "step": 893 + }, + { + "epoch": 1.7563850687622788, + "grad_norm": 0.3787725523419915, + "learning_rate": 6.558986350076243e-05, + "loss": 0.572, + "step": 894 + }, + { + "epoch": 1.7583497053045187, + "grad_norm": 0.48383350172451994, + "learning_rate": 6.55476632962377e-05, + "loss": 0.5713, + "step": 895 + }, + { + "epoch": 1.7603143418467584, + "grad_norm": 0.6785253087600126, + "learning_rate": 6.550541501000389e-05, + "loss": 0.5524, + "step": 896 + }, + { + "epoch": 1.762278978388998, + "grad_norm": 0.727298295916743, + "learning_rate": 6.546311872157395e-05, + "loss": 0.5662, + "step": 897 + }, + { + "epoch": 1.7642436149312377, + "grad_norm": 0.6644382948590399, + "learning_rate": 6.542077451055116e-05, + "loss": 0.5754, + "step": 898 + }, + { + "epoch": 1.7662082514734774, + "grad_norm": 0.5523032248429363, + "learning_rate": 6.537838245662897e-05, + "loss": 0.5722, + "step": 899 + }, + { + "epoch": 1.768172888015717, + "grad_norm": 0.484981663847217, + "learning_rate": 6.533594263959095e-05, + "loss": 0.5684, + "step": 900 + }, + { + "epoch": 1.7701375245579567, + "grad_norm": 0.4812739465163888, + "learning_rate": 6.529345513931044e-05, + "loss": 0.5666, + "step": 901 + }, + { + "epoch": 1.7721021611001966, + "grad_norm": 0.4466152323274746, + "learning_rate": 6.525092003575064e-05, + "loss": 0.5715, + "step": 902 + }, + { + "epoch": 1.774066797642436, + "grad_norm": 0.38245651912712886, + "learning_rate": 6.520833740896427e-05, + "loss": 0.5719, + "step": 903 + }, + { + "epoch": 1.776031434184676, + "grad_norm": 0.41544755917719484, + "learning_rate": 6.516570733909351e-05, + "loss": 0.5607, + "step": 904 + }, + { + "epoch": 1.7779960707269156, + "grad_norm": 0.3862251963263237, + "learning_rate": 6.512302990636987e-05, + "loss": 0.5687, + "step": 905 + }, + { + "epoch": 1.7799607072691552, + "grad_norm": 0.4024762000694496, + "learning_rate": 6.50803051911139e-05, + "loss": 0.5732, + "step": 906 + }, + { + "epoch": 1.781925343811395, + "grad_norm": 0.38263025612890095, + "learning_rate": 6.503753327373522e-05, + "loss": 0.5734, + "step": 907 + }, + { + "epoch": 1.7838899803536346, + "grad_norm": 0.4102510779499884, + "learning_rate": 6.499471423473228e-05, + "loss": 0.573, + "step": 908 + }, + { + "epoch": 1.7858546168958742, + "grad_norm": 0.568169490056956, + "learning_rate": 6.495184815469219e-05, + "loss": 0.5691, + "step": 909 + }, + { + "epoch": 1.7878192534381139, + "grad_norm": 0.6309280612652918, + "learning_rate": 6.49089351142906e-05, + "loss": 0.5786, + "step": 910 + }, + { + "epoch": 1.7897838899803538, + "grad_norm": 0.4980948229627332, + "learning_rate": 6.486597519429154e-05, + "loss": 0.5716, + "step": 911 + }, + { + "epoch": 1.7917485265225932, + "grad_norm": 0.3798133849178813, + "learning_rate": 6.482296847554729e-05, + "loss": 0.5642, + "step": 912 + }, + { + "epoch": 1.793713163064833, + "grad_norm": 0.3374925842514988, + "learning_rate": 6.477991503899816e-05, + "loss": 0.5699, + "step": 913 + }, + { + "epoch": 1.7956777996070727, + "grad_norm": 0.3767968453637264, + "learning_rate": 6.473681496567243e-05, + "loss": 0.5708, + "step": 914 + }, + { + "epoch": 1.7976424361493124, + "grad_norm": 0.3987975522076679, + "learning_rate": 6.469366833668615e-05, + "loss": 0.5617, + "step": 915 + }, + { + "epoch": 1.799607072691552, + "grad_norm": 0.3816709191954348, + "learning_rate": 6.465047523324295e-05, + "loss": 0.5776, + "step": 916 + }, + { + "epoch": 1.8015717092337917, + "grad_norm": 0.3921925192294356, + "learning_rate": 6.460723573663396e-05, + "loss": 0.5625, + "step": 917 + }, + { + "epoch": 1.8035363457760314, + "grad_norm": 0.44398329177776974, + "learning_rate": 6.456394992823763e-05, + "loss": 0.566, + "step": 918 + }, + { + "epoch": 1.805500982318271, + "grad_norm": 0.4991681719720263, + "learning_rate": 6.452061788951955e-05, + "loss": 0.5626, + "step": 919 + }, + { + "epoch": 1.807465618860511, + "grad_norm": 0.45435670564756186, + "learning_rate": 6.447723970203232e-05, + "loss": 0.5647, + "step": 920 + }, + { + "epoch": 1.8094302554027504, + "grad_norm": 0.4646503355198664, + "learning_rate": 6.44338154474154e-05, + "loss": 0.5644, + "step": 921 + }, + { + "epoch": 1.8113948919449903, + "grad_norm": 0.570767780662275, + "learning_rate": 6.439034520739495e-05, + "loss": 0.5798, + "step": 922 + }, + { + "epoch": 1.8133595284872297, + "grad_norm": 0.7241027885898259, + "learning_rate": 6.434682906378367e-05, + "loss": 0.5684, + "step": 923 + }, + { + "epoch": 1.8153241650294696, + "grad_norm": 0.8389364964431695, + "learning_rate": 6.430326709848067e-05, + "loss": 0.5629, + "step": 924 + }, + { + "epoch": 1.8172888015717092, + "grad_norm": 0.8858578716222357, + "learning_rate": 6.425965939347127e-05, + "loss": 0.5682, + "step": 925 + }, + { + "epoch": 1.819253438113949, + "grad_norm": 0.8169196781004041, + "learning_rate": 6.42160060308269e-05, + "loss": 0.5693, + "step": 926 + }, + { + "epoch": 1.8212180746561886, + "grad_norm": 0.6536227313151313, + "learning_rate": 6.417230709270491e-05, + "loss": 0.5719, + "step": 927 + }, + { + "epoch": 1.8231827111984282, + "grad_norm": 0.45378847847193715, + "learning_rate": 6.41285626613484e-05, + "loss": 0.5731, + "step": 928 + }, + { + "epoch": 1.8251473477406681, + "grad_norm": 0.33193614435344737, + "learning_rate": 6.408477281908616e-05, + "loss": 0.5664, + "step": 929 + }, + { + "epoch": 1.8271119842829076, + "grad_norm": 0.4803487753099259, + "learning_rate": 6.404093764833235e-05, + "loss": 0.5601, + "step": 930 + }, + { + "epoch": 1.8290766208251474, + "grad_norm": 0.6518326549525895, + "learning_rate": 6.399705723158651e-05, + "loss": 0.5666, + "step": 931 + }, + { + "epoch": 1.8310412573673869, + "grad_norm": 0.6690323725233134, + "learning_rate": 6.395313165143333e-05, + "loss": 0.5715, + "step": 932 + }, + { + "epoch": 1.8330058939096268, + "grad_norm": 0.6510472973234042, + "learning_rate": 6.390916099054244e-05, + "loss": 0.5604, + "step": 933 + }, + { + "epoch": 1.8349705304518664, + "grad_norm": 0.5545822322983249, + "learning_rate": 6.386514533166841e-05, + "loss": 0.5641, + "step": 934 + }, + { + "epoch": 1.836935166994106, + "grad_norm": 0.42746590433024045, + "learning_rate": 6.382108475765039e-05, + "loss": 0.5636, + "step": 935 + }, + { + "epoch": 1.8388998035363457, + "grad_norm": 0.49025450767966705, + "learning_rate": 6.377697935141216e-05, + "loss": 0.5696, + "step": 936 + }, + { + "epoch": 1.8408644400785854, + "grad_norm": 0.6271240612482153, + "learning_rate": 6.373282919596182e-05, + "loss": 0.5771, + "step": 937 + }, + { + "epoch": 1.8428290766208253, + "grad_norm": 0.79547766483824, + "learning_rate": 6.368863437439168e-05, + "loss": 0.5667, + "step": 938 + }, + { + "epoch": 1.8447937131630647, + "grad_norm": 0.8306234333071912, + "learning_rate": 6.364439496987818e-05, + "loss": 0.5698, + "step": 939 + }, + { + "epoch": 1.8467583497053046, + "grad_norm": 0.7132338647127635, + "learning_rate": 6.360011106568158e-05, + "loss": 0.5668, + "step": 940 + }, + { + "epoch": 1.848722986247544, + "grad_norm": 0.5806820895599273, + "learning_rate": 6.355578274514595e-05, + "loss": 0.5717, + "step": 941 + }, + { + "epoch": 1.850687622789784, + "grad_norm": 0.5006228018217213, + "learning_rate": 6.351141009169893e-05, + "loss": 0.5627, + "step": 942 + }, + { + "epoch": 1.8526522593320236, + "grad_norm": 0.39517443810708913, + "learning_rate": 6.346699318885161e-05, + "loss": 0.5608, + "step": 943 + }, + { + "epoch": 1.8546168958742633, + "grad_norm": 0.4177259037987791, + "learning_rate": 6.342253212019836e-05, + "loss": 0.5661, + "step": 944 + }, + { + "epoch": 1.856581532416503, + "grad_norm": 0.4580345701063939, + "learning_rate": 6.337802696941665e-05, + "loss": 0.5665, + "step": 945 + }, + { + "epoch": 1.8585461689587426, + "grad_norm": 0.421833833068103, + "learning_rate": 6.333347782026693e-05, + "loss": 0.5628, + "step": 946 + }, + { + "epoch": 1.8605108055009825, + "grad_norm": 0.3308348519422811, + "learning_rate": 6.328888475659245e-05, + "loss": 0.562, + "step": 947 + }, + { + "epoch": 1.862475442043222, + "grad_norm": 0.244670460450005, + "learning_rate": 6.324424786231912e-05, + "loss": 0.5556, + "step": 948 + }, + { + "epoch": 1.8644400785854618, + "grad_norm": 0.31622037872560443, + "learning_rate": 6.319956722145534e-05, + "loss": 0.5543, + "step": 949 + }, + { + "epoch": 1.8664047151277012, + "grad_norm": 0.34727278968348696, + "learning_rate": 6.315484291809181e-05, + "loss": 0.5726, + "step": 950 + }, + { + "epoch": 1.8683693516699411, + "grad_norm": 0.3646264377798013, + "learning_rate": 6.311007503640147e-05, + "loss": 0.5668, + "step": 951 + }, + { + "epoch": 1.8703339882121808, + "grad_norm": 0.41044626280233665, + "learning_rate": 6.306526366063921e-05, + "loss": 0.5641, + "step": 952 + }, + { + "epoch": 1.8722986247544204, + "grad_norm": 0.41669072660758755, + "learning_rate": 6.302040887514182e-05, + "loss": 0.5646, + "step": 953 + }, + { + "epoch": 1.87426326129666, + "grad_norm": 0.4913830279354876, + "learning_rate": 6.297551076432775e-05, + "loss": 0.5714, + "step": 954 + }, + { + "epoch": 1.8762278978388998, + "grad_norm": 0.5696374475728034, + "learning_rate": 6.293056941269703e-05, + "loss": 0.5709, + "step": 955 + }, + { + "epoch": 1.8781925343811396, + "grad_norm": 0.5120186980292625, + "learning_rate": 6.288558490483108e-05, + "loss": 0.5709, + "step": 956 + }, + { + "epoch": 1.880157170923379, + "grad_norm": 0.35217299476654584, + "learning_rate": 6.284055732539246e-05, + "loss": 0.5767, + "step": 957 + }, + { + "epoch": 1.882121807465619, + "grad_norm": 0.3181983393408799, + "learning_rate": 6.279548675912491e-05, + "loss": 0.5662, + "step": 958 + }, + { + "epoch": 1.8840864440078584, + "grad_norm": 0.3304458077544014, + "learning_rate": 6.275037329085296e-05, + "loss": 0.5683, + "step": 959 + }, + { + "epoch": 1.8860510805500983, + "grad_norm": 0.29389041479290523, + "learning_rate": 6.270521700548196e-05, + "loss": 0.5592, + "step": 960 + }, + { + "epoch": 1.888015717092338, + "grad_norm": 0.30432406624614405, + "learning_rate": 6.266001798799781e-05, + "loss": 0.5611, + "step": 961 + }, + { + "epoch": 1.8899803536345776, + "grad_norm": 0.33087895644791454, + "learning_rate": 6.261477632346686e-05, + "loss": 0.5666, + "step": 962 + }, + { + "epoch": 1.8919449901768173, + "grad_norm": 0.3629192037004368, + "learning_rate": 6.256949209703572e-05, + "loss": 0.5653, + "step": 963 + }, + { + "epoch": 1.893909626719057, + "grad_norm": 0.5067541917525441, + "learning_rate": 6.252416539393101e-05, + "loss": 0.5615, + "step": 964 + }, + { + "epoch": 1.8958742632612968, + "grad_norm": 0.6279271818920406, + "learning_rate": 6.247879629945946e-05, + "loss": 0.5573, + "step": 965 + }, + { + "epoch": 1.8978388998035363, + "grad_norm": 0.6230122579138665, + "learning_rate": 6.243338489900745e-05, + "loss": 0.5626, + "step": 966 + }, + { + "epoch": 1.8998035363457761, + "grad_norm": 0.6692484076003836, + "learning_rate": 6.238793127804101e-05, + "loss": 0.5747, + "step": 967 + }, + { + "epoch": 1.9017681728880156, + "grad_norm": 0.7180994091668007, + "learning_rate": 6.234243552210569e-05, + "loss": 0.5745, + "step": 968 + }, + { + "epoch": 1.9037328094302555, + "grad_norm": 0.7630148660719313, + "learning_rate": 6.229689771682626e-05, + "loss": 0.5783, + "step": 969 + }, + { + "epoch": 1.9056974459724951, + "grad_norm": 0.5984079921704016, + "learning_rate": 6.22513179479067e-05, + "loss": 0.5705, + "step": 970 + }, + { + "epoch": 1.9076620825147348, + "grad_norm": 0.4382124890366382, + "learning_rate": 6.220569630112989e-05, + "loss": 0.5682, + "step": 971 + }, + { + "epoch": 1.9096267190569745, + "grad_norm": 0.4048064607754648, + "learning_rate": 6.216003286235758e-05, + "loss": 0.5617, + "step": 972 + }, + { + "epoch": 1.9115913555992141, + "grad_norm": 0.4264218767368254, + "learning_rate": 6.211432771753019e-05, + "loss": 0.5677, + "step": 973 + }, + { + "epoch": 1.9135559921414538, + "grad_norm": 0.39208900193113166, + "learning_rate": 6.206858095266656e-05, + "loss": 0.5746, + "step": 974 + }, + { + "epoch": 1.9155206286836934, + "grad_norm": 0.37860709867962467, + "learning_rate": 6.202279265386394e-05, + "loss": 0.5612, + "step": 975 + }, + { + "epoch": 1.9174852652259333, + "grad_norm": 0.48429082493863873, + "learning_rate": 6.19769629072977e-05, + "loss": 0.5707, + "step": 976 + }, + { + "epoch": 1.9194499017681728, + "grad_norm": 0.5040279723777845, + "learning_rate": 6.19310917992212e-05, + "loss": 0.5573, + "step": 977 + }, + { + "epoch": 1.9214145383104126, + "grad_norm": 0.4648748762883947, + "learning_rate": 6.188517941596573e-05, + "loss": 0.5658, + "step": 978 + }, + { + "epoch": 1.9233791748526523, + "grad_norm": 0.44451566399968334, + "learning_rate": 6.183922584394018e-05, + "loss": 0.5703, + "step": 979 + }, + { + "epoch": 1.925343811394892, + "grad_norm": 0.40795815024839943, + "learning_rate": 6.179323116963095e-05, + "loss": 0.5676, + "step": 980 + }, + { + "epoch": 1.9273084479371316, + "grad_norm": 0.34044886359332155, + "learning_rate": 6.174719547960186e-05, + "loss": 0.5534, + "step": 981 + }, + { + "epoch": 1.9292730844793713, + "grad_norm": 0.34965638138700866, + "learning_rate": 6.17011188604939e-05, + "loss": 0.5599, + "step": 982 + }, + { + "epoch": 1.931237721021611, + "grad_norm": 0.3938978981622104, + "learning_rate": 6.165500139902507e-05, + "loss": 0.5621, + "step": 983 + }, + { + "epoch": 1.9332023575638506, + "grad_norm": 0.42132000304905765, + "learning_rate": 6.160884318199025e-05, + "loss": 0.5675, + "step": 984 + }, + { + "epoch": 1.9351669941060905, + "grad_norm": 0.5070121058048241, + "learning_rate": 6.156264429626103e-05, + "loss": 0.5681, + "step": 985 + }, + { + "epoch": 1.93713163064833, + "grad_norm": 0.6583944808987334, + "learning_rate": 6.151640482878554e-05, + "loss": 0.5662, + "step": 986 + }, + { + "epoch": 1.9390962671905698, + "grad_norm": 0.735575066321496, + "learning_rate": 6.147012486658825e-05, + "loss": 0.5662, + "step": 987 + }, + { + "epoch": 1.9410609037328095, + "grad_norm": 0.813501027354211, + "learning_rate": 6.14238044967699e-05, + "loss": 0.5745, + "step": 988 + }, + { + "epoch": 1.9430255402750491, + "grad_norm": 0.8826804104475269, + "learning_rate": 6.137744380650726e-05, + "loss": 0.5595, + "step": 989 + }, + { + "epoch": 1.9449901768172888, + "grad_norm": 0.8103135325300171, + "learning_rate": 6.133104288305293e-05, + "loss": 0.5607, + "step": 990 + }, + { + "epoch": 1.9469548133595285, + "grad_norm": 0.6332907847188693, + "learning_rate": 6.128460181373533e-05, + "loss": 0.564, + "step": 991 + }, + { + "epoch": 1.9489194499017681, + "grad_norm": 0.4250624550126376, + "learning_rate": 6.123812068595835e-05, + "loss": 0.5719, + "step": 992 + }, + { + "epoch": 1.9508840864440078, + "grad_norm": 0.24970166739934735, + "learning_rate": 6.119159958720132e-05, + "loss": 0.5666, + "step": 993 + }, + { + "epoch": 1.9528487229862477, + "grad_norm": 0.3303188969769002, + "learning_rate": 6.114503860501876e-05, + "loss": 0.5742, + "step": 994 + }, + { + "epoch": 1.9548133595284871, + "grad_norm": 0.40882347155752685, + "learning_rate": 6.109843782704029e-05, + "loss": 0.5546, + "step": 995 + }, + { + "epoch": 1.956777996070727, + "grad_norm": 0.49039641052935895, + "learning_rate": 6.105179734097041e-05, + "loss": 0.5688, + "step": 996 + }, + { + "epoch": 1.9587426326129664, + "grad_norm": 0.4863247519691268, + "learning_rate": 6.100511723458835e-05, + "loss": 0.5673, + "step": 997 + }, + { + "epoch": 1.9607072691552063, + "grad_norm": 0.39857918928461783, + "learning_rate": 6.095839759574791e-05, + "loss": 0.5702, + "step": 998 + }, + { + "epoch": 1.962671905697446, + "grad_norm": 0.3663572679469532, + "learning_rate": 6.0911638512377296e-05, + "loss": 0.5695, + "step": 999 + }, + { + "epoch": 1.9646365422396856, + "grad_norm": 0.3498831663410025, + "learning_rate": 6.086484007247894e-05, + "loss": 0.5772, + "step": 1000 + }, + { + "epoch": 1.9666011787819253, + "grad_norm": 0.40037754568557676, + "learning_rate": 6.081800236412935e-05, + "loss": 0.5643, + "step": 1001 + }, + { + "epoch": 1.968565815324165, + "grad_norm": 0.5057406550004359, + "learning_rate": 6.077112547547894e-05, + "loss": 0.5613, + "step": 1002 + }, + { + "epoch": 1.9705304518664049, + "grad_norm": 0.5966032477353053, + "learning_rate": 6.0724209494751865e-05, + "loss": 0.5561, + "step": 1003 + }, + { + "epoch": 1.9724950884086443, + "grad_norm": 0.6304353673155054, + "learning_rate": 6.067725451024583e-05, + "loss": 0.5619, + "step": 1004 + }, + { + "epoch": 1.9744597249508842, + "grad_norm": 0.597934062804904, + "learning_rate": 6.063026061033201e-05, + "loss": 0.5798, + "step": 1005 + }, + { + "epoch": 1.9764243614931236, + "grad_norm": 0.5720239533486571, + "learning_rate": 6.058322788345473e-05, + "loss": 0.5668, + "step": 1006 + }, + { + "epoch": 1.9783889980353635, + "grad_norm": 0.5212941624082756, + "learning_rate": 6.053615641813146e-05, + "loss": 0.5722, + "step": 1007 + }, + { + "epoch": 1.9803536345776032, + "grad_norm": 0.42855699471623565, + "learning_rate": 6.0489046302952544e-05, + "loss": 0.5613, + "step": 1008 + }, + { + "epoch": 1.9823182711198428, + "grad_norm": 0.3713609025157819, + "learning_rate": 6.044189762658107e-05, + "loss": 0.5681, + "step": 1009 + }, + { + "epoch": 1.9842829076620825, + "grad_norm": 0.4549691732349415, + "learning_rate": 6.039471047775273e-05, + "loss": 0.5658, + "step": 1010 + }, + { + "epoch": 1.9862475442043221, + "grad_norm": 0.5961557350686431, + "learning_rate": 6.034748494527557e-05, + "loss": 0.5715, + "step": 1011 + }, + { + "epoch": 1.988212180746562, + "grad_norm": 0.616214602373178, + "learning_rate": 6.030022111802992e-05, + "loss": 0.5672, + "step": 1012 + }, + { + "epoch": 1.9901768172888015, + "grad_norm": 0.5561650409344118, + "learning_rate": 6.025291908496815e-05, + "loss": 0.5669, + "step": 1013 + }, + { + "epoch": 1.9921414538310414, + "grad_norm": 0.49623958475021923, + "learning_rate": 6.020557893511456e-05, + "loss": 0.564, + "step": 1014 + }, + { + "epoch": 1.9941060903732808, + "grad_norm": 0.40801523864204325, + "learning_rate": 6.01582007575652e-05, + "loss": 0.5625, + "step": 1015 + }, + { + "epoch": 1.9960707269155207, + "grad_norm": 0.34220493406553243, + "learning_rate": 6.011078464148762e-05, + "loss": 0.5756, + "step": 1016 + }, + { + "epoch": 1.9980353634577603, + "grad_norm": 0.3248166199671457, + "learning_rate": 6.006333067612085e-05, + "loss": 0.5603, + "step": 1017 + }, + { + "epoch": 2.0, + "grad_norm": 0.36096544788883017, + "learning_rate": 6.0015838950775116e-05, + "loss": 0.5672, + "step": 1018 + }, + { + "epoch": 2.00196463654224, + "grad_norm": 0.4504458558054362, + "learning_rate": 5.9968309554831705e-05, + "loss": 0.5165, + "step": 1019 + }, + { + "epoch": 2.0039292730844793, + "grad_norm": 0.4208540515396218, + "learning_rate": 5.992074257774285e-05, + "loss": 0.5132, + "step": 1020 + }, + { + "epoch": 2.005893909626719, + "grad_norm": 0.5464193532187787, + "learning_rate": 5.9873138109031435e-05, + "loss": 0.5242, + "step": 1021 + }, + { + "epoch": 2.0078585461689586, + "grad_norm": 0.7646011895251561, + "learning_rate": 5.982549623829096e-05, + "loss": 0.5144, + "step": 1022 + }, + { + "epoch": 2.0098231827111985, + "grad_norm": 0.9085351483627337, + "learning_rate": 5.97778170551853e-05, + "loss": 0.5281, + "step": 1023 + }, + { + "epoch": 2.011787819253438, + "grad_norm": 1.0681277629439256, + "learning_rate": 5.9730100649448586e-05, + "loss": 0.5226, + "step": 1024 + }, + { + "epoch": 2.013752455795678, + "grad_norm": 0.7389329686029219, + "learning_rate": 5.9682347110884925e-05, + "loss": 0.5237, + "step": 1025 + }, + { + "epoch": 2.0157170923379173, + "grad_norm": 0.5642510712153912, + "learning_rate": 5.96345565293684e-05, + "loss": 0.5221, + "step": 1026 + }, + { + "epoch": 2.017681728880157, + "grad_norm": 0.7277780337996136, + "learning_rate": 5.9586728994842744e-05, + "loss": 0.5279, + "step": 1027 + }, + { + "epoch": 2.019646365422397, + "grad_norm": 0.714932646775911, + "learning_rate": 5.9538864597321266e-05, + "loss": 0.5274, + "step": 1028 + }, + { + "epoch": 2.0216110019646365, + "grad_norm": 0.5025906598849955, + "learning_rate": 5.9490963426886664e-05, + "loss": 0.5222, + "step": 1029 + }, + { + "epoch": 2.0235756385068764, + "grad_norm": 0.5366909335293968, + "learning_rate": 5.9443025573690795e-05, + "loss": 0.52, + "step": 1030 + }, + { + "epoch": 2.025540275049116, + "grad_norm": 0.5631151325984085, + "learning_rate": 5.939505112795463e-05, + "loss": 0.5154, + "step": 1031 + }, + { + "epoch": 2.0275049115913557, + "grad_norm": 0.4013151262152183, + "learning_rate": 5.934704017996794e-05, + "loss": 0.5114, + "step": 1032 + }, + { + "epoch": 2.029469548133595, + "grad_norm": 0.4056579776214141, + "learning_rate": 5.929899282008922e-05, + "loss": 0.5122, + "step": 1033 + }, + { + "epoch": 2.031434184675835, + "grad_norm": 0.49872119917729996, + "learning_rate": 5.9250909138745505e-05, + "loss": 0.5166, + "step": 1034 + }, + { + "epoch": 2.0333988212180745, + "grad_norm": 0.39470033498763957, + "learning_rate": 5.920278922643217e-05, + "loss": 0.526, + "step": 1035 + }, + { + "epoch": 2.0353634577603144, + "grad_norm": 0.38470103730734, + "learning_rate": 5.9154633173712805e-05, + "loss": 0.5137, + "step": 1036 + }, + { + "epoch": 2.0373280943025542, + "grad_norm": 0.4911307766240645, + "learning_rate": 5.910644107121897e-05, + "loss": 0.5167, + "step": 1037 + }, + { + "epoch": 2.0392927308447937, + "grad_norm": 0.4014614698220645, + "learning_rate": 5.905821300965012e-05, + "loss": 0.5173, + "step": 1038 + }, + { + "epoch": 2.0412573673870336, + "grad_norm": 0.4249116842526582, + "learning_rate": 5.9009949079773366e-05, + "loss": 0.5204, + "step": 1039 + }, + { + "epoch": 2.043222003929273, + "grad_norm": 0.4281487472132033, + "learning_rate": 5.896164937242331e-05, + "loss": 0.5267, + "step": 1040 + }, + { + "epoch": 2.045186640471513, + "grad_norm": 0.42201421909430137, + "learning_rate": 5.891331397850193e-05, + "loss": 0.5124, + "step": 1041 + }, + { + "epoch": 2.0471512770137523, + "grad_norm": 0.3813557830941727, + "learning_rate": 5.886494298897833e-05, + "loss": 0.5217, + "step": 1042 + }, + { + "epoch": 2.049115913555992, + "grad_norm": 0.3206594290973613, + "learning_rate": 5.881653649488861e-05, + "loss": 0.5164, + "step": 1043 + }, + { + "epoch": 2.0510805500982316, + "grad_norm": 0.48878461311574894, + "learning_rate": 5.8768094587335716e-05, + "loss": 0.5162, + "step": 1044 + }, + { + "epoch": 2.0530451866404715, + "grad_norm": 0.4375510258304667, + "learning_rate": 5.8719617357489204e-05, + "loss": 0.5122, + "step": 1045 + }, + { + "epoch": 2.0550098231827114, + "grad_norm": 0.3137450581031588, + "learning_rate": 5.867110489658516e-05, + "loss": 0.5136, + "step": 1046 + }, + { + "epoch": 2.056974459724951, + "grad_norm": 0.3393555320073471, + "learning_rate": 5.862255729592592e-05, + "loss": 0.5207, + "step": 1047 + }, + { + "epoch": 2.0589390962671907, + "grad_norm": 0.2855561788304838, + "learning_rate": 5.8573974646880005e-05, + "loss": 0.5104, + "step": 1048 + }, + { + "epoch": 2.06090373280943, + "grad_norm": 0.2460896956296693, + "learning_rate": 5.852535704088186e-05, + "loss": 0.5209, + "step": 1049 + }, + { + "epoch": 2.06286836935167, + "grad_norm": 0.3053875094274896, + "learning_rate": 5.847670456943174e-05, + "loss": 0.5221, + "step": 1050 + }, + { + "epoch": 2.0648330058939095, + "grad_norm": 0.32396200351272275, + "learning_rate": 5.842801732409551e-05, + "loss": 0.5116, + "step": 1051 + }, + { + "epoch": 2.0667976424361494, + "grad_norm": 0.3895986199186741, + "learning_rate": 5.837929539650448e-05, + "loss": 0.5152, + "step": 1052 + }, + { + "epoch": 2.068762278978389, + "grad_norm": 0.374011376863708, + "learning_rate": 5.8330538878355254e-05, + "loss": 0.5139, + "step": 1053 + }, + { + "epoch": 2.0707269155206287, + "grad_norm": 0.394053287660596, + "learning_rate": 5.828174786140952e-05, + "loss": 0.5185, + "step": 1054 + }, + { + "epoch": 2.072691552062868, + "grad_norm": 0.39382673443889243, + "learning_rate": 5.8232922437493875e-05, + "loss": 0.521, + "step": 1055 + }, + { + "epoch": 2.074656188605108, + "grad_norm": 0.39792778712417676, + "learning_rate": 5.8184062698499696e-05, + "loss": 0.5086, + "step": 1056 + }, + { + "epoch": 2.076620825147348, + "grad_norm": 0.40427297708014104, + "learning_rate": 5.8135168736382944e-05, + "loss": 0.5186, + "step": 1057 + }, + { + "epoch": 2.0785854616895874, + "grad_norm": 0.40251456668879515, + "learning_rate": 5.808624064316399e-05, + "loss": 0.5173, + "step": 1058 + }, + { + "epoch": 2.0805500982318272, + "grad_norm": 0.36476807674125955, + "learning_rate": 5.803727851092742e-05, + "loss": 0.5177, + "step": 1059 + }, + { + "epoch": 2.0825147347740667, + "grad_norm": 0.3329502111234452, + "learning_rate": 5.798828243182191e-05, + "loss": 0.5157, + "step": 1060 + }, + { + "epoch": 2.0844793713163066, + "grad_norm": 0.2626955289493107, + "learning_rate": 5.7939252498059986e-05, + "loss": 0.5183, + "step": 1061 + }, + { + "epoch": 2.086444007858546, + "grad_norm": 0.26237053652092884, + "learning_rate": 5.7890188801917944e-05, + "loss": 0.5172, + "step": 1062 + }, + { + "epoch": 2.088408644400786, + "grad_norm": 0.3439063992552652, + "learning_rate": 5.784109143573558e-05, + "loss": 0.5208, + "step": 1063 + }, + { + "epoch": 2.0903732809430258, + "grad_norm": 0.4110656743098468, + "learning_rate": 5.779196049191607e-05, + "loss": 0.5174, + "step": 1064 + }, + { + "epoch": 2.092337917485265, + "grad_norm": 0.39865358287743036, + "learning_rate": 5.7742796062925794e-05, + "loss": 0.5171, + "step": 1065 + }, + { + "epoch": 2.094302554027505, + "grad_norm": 0.33886679383107626, + "learning_rate": 5.769359824129414e-05, + "loss": 0.5202, + "step": 1066 + }, + { + "epoch": 2.0962671905697445, + "grad_norm": 0.29615040250294683, + "learning_rate": 5.764436711961334e-05, + "loss": 0.5131, + "step": 1067 + }, + { + "epoch": 2.0982318271119844, + "grad_norm": 0.28224013736210163, + "learning_rate": 5.7595102790538315e-05, + "loss": 0.5156, + "step": 1068 + }, + { + "epoch": 2.100196463654224, + "grad_norm": 0.3522306235329473, + "learning_rate": 5.7545805346786484e-05, + "loss": 0.5199, + "step": 1069 + }, + { + "epoch": 2.1021611001964637, + "grad_norm": 0.3720818535443286, + "learning_rate": 5.749647488113755e-05, + "loss": 0.5281, + "step": 1070 + }, + { + "epoch": 2.104125736738703, + "grad_norm": 0.32550831915375067, + "learning_rate": 5.744711148643342e-05, + "loss": 0.5205, + "step": 1071 + }, + { + "epoch": 2.106090373280943, + "grad_norm": 0.2851759619924045, + "learning_rate": 5.739771525557794e-05, + "loss": 0.5232, + "step": 1072 + }, + { + "epoch": 2.1080550098231825, + "grad_norm": 0.25093584901610516, + "learning_rate": 5.734828628153675e-05, + "loss": 0.5167, + "step": 1073 + }, + { + "epoch": 2.1100196463654224, + "grad_norm": 0.3006980272176056, + "learning_rate": 5.7298824657337154e-05, + "loss": 0.515, + "step": 1074 + }, + { + "epoch": 2.1119842829076623, + "grad_norm": 0.3937222940738968, + "learning_rate": 5.724933047606786e-05, + "loss": 0.5152, + "step": 1075 + }, + { + "epoch": 2.1139489194499017, + "grad_norm": 0.33378831309611945, + "learning_rate": 5.719980383087885e-05, + "loss": 0.516, + "step": 1076 + }, + { + "epoch": 2.1159135559921416, + "grad_norm": 0.25495301987300134, + "learning_rate": 5.715024481498125e-05, + "loss": 0.5325, + "step": 1077 + }, + { + "epoch": 2.117878192534381, + "grad_norm": 0.2961475737010374, + "learning_rate": 5.710065352164709e-05, + "loss": 0.5077, + "step": 1078 + }, + { + "epoch": 2.119842829076621, + "grad_norm": 0.28727776301493396, + "learning_rate": 5.70510300442091e-05, + "loss": 0.5093, + "step": 1079 + }, + { + "epoch": 2.1218074656188604, + "grad_norm": 0.3362206948645093, + "learning_rate": 5.7001374476060646e-05, + "loss": 0.5212, + "step": 1080 + }, + { + "epoch": 2.1237721021611002, + "grad_norm": 0.31974409512633345, + "learning_rate": 5.6951686910655456e-05, + "loss": 0.5137, + "step": 1081 + }, + { + "epoch": 2.1257367387033397, + "grad_norm": 0.3503772840444143, + "learning_rate": 5.690196744150749e-05, + "loss": 0.523, + "step": 1082 + }, + { + "epoch": 2.1277013752455796, + "grad_norm": 0.3675710884228282, + "learning_rate": 5.6852216162190744e-05, + "loss": 0.5132, + "step": 1083 + }, + { + "epoch": 2.1296660117878194, + "grad_norm": 0.39248791362199803, + "learning_rate": 5.68024331663391e-05, + "loss": 0.5219, + "step": 1084 + }, + { + "epoch": 2.131630648330059, + "grad_norm": 0.40145331452078703, + "learning_rate": 5.6752618547646095e-05, + "loss": 0.518, + "step": 1085 + }, + { + "epoch": 2.1335952848722988, + "grad_norm": 0.44545862653002233, + "learning_rate": 5.670277239986481e-05, + "loss": 0.5198, + "step": 1086 + }, + { + "epoch": 2.135559921414538, + "grad_norm": 0.3925762137803909, + "learning_rate": 5.665289481680768e-05, + "loss": 0.5244, + "step": 1087 + }, + { + "epoch": 2.137524557956778, + "grad_norm": 0.3203298298806596, + "learning_rate": 5.660298589234627e-05, + "loss": 0.5264, + "step": 1088 + }, + { + "epoch": 2.1394891944990175, + "grad_norm": 0.3046179148774701, + "learning_rate": 5.655304572041114e-05, + "loss": 0.5163, + "step": 1089 + }, + { + "epoch": 2.1414538310412574, + "grad_norm": 0.35743859919225585, + "learning_rate": 5.650307439499164e-05, + "loss": 0.5133, + "step": 1090 + }, + { + "epoch": 2.143418467583497, + "grad_norm": 0.44495727637317817, + "learning_rate": 5.645307201013581e-05, + "loss": 0.5225, + "step": 1091 + }, + { + "epoch": 2.1453831041257367, + "grad_norm": 0.45354804761979917, + "learning_rate": 5.640303865995011e-05, + "loss": 0.5222, + "step": 1092 + }, + { + "epoch": 2.1473477406679766, + "grad_norm": 0.3105739723842859, + "learning_rate": 5.635297443859924e-05, + "loss": 0.5081, + "step": 1093 + }, + { + "epoch": 2.149312377210216, + "grad_norm": 0.2228706805354587, + "learning_rate": 5.6302879440306064e-05, + "loss": 0.5165, + "step": 1094 + }, + { + "epoch": 2.151277013752456, + "grad_norm": 0.27111902396947674, + "learning_rate": 5.6252753759351325e-05, + "loss": 0.5266, + "step": 1095 + }, + { + "epoch": 2.1532416502946954, + "grad_norm": 0.3031853172902032, + "learning_rate": 5.620259749007356e-05, + "loss": 0.5195, + "step": 1096 + }, + { + "epoch": 2.1552062868369353, + "grad_norm": 0.32709881229630416, + "learning_rate": 5.615241072686883e-05, + "loss": 0.5239, + "step": 1097 + }, + { + "epoch": 2.1571709233791747, + "grad_norm": 0.35332217336483923, + "learning_rate": 5.610219356419058e-05, + "loss": 0.5233, + "step": 1098 + }, + { + "epoch": 2.1591355599214146, + "grad_norm": 0.3233831018208943, + "learning_rate": 5.605194609654952e-05, + "loss": 0.5271, + "step": 1099 + }, + { + "epoch": 2.161100196463654, + "grad_norm": 0.34498663265464036, + "learning_rate": 5.600166841851333e-05, + "loss": 0.5202, + "step": 1100 + }, + { + "epoch": 2.163064833005894, + "grad_norm": 0.41146094852397364, + "learning_rate": 5.5951360624706624e-05, + "loss": 0.5145, + "step": 1101 + }, + { + "epoch": 2.165029469548134, + "grad_norm": 0.4652347348467157, + "learning_rate": 5.5901022809810626e-05, + "loss": 0.5204, + "step": 1102 + }, + { + "epoch": 2.1669941060903732, + "grad_norm": 0.43644847511518026, + "learning_rate": 5.585065506856309e-05, + "loss": 0.5126, + "step": 1103 + }, + { + "epoch": 2.168958742632613, + "grad_norm": 0.41473755481298064, + "learning_rate": 5.580025749575809e-05, + "loss": 0.5148, + "step": 1104 + }, + { + "epoch": 2.1709233791748526, + "grad_norm": 0.3709931914648881, + "learning_rate": 5.574983018624584e-05, + "loss": 0.5265, + "step": 1105 + }, + { + "epoch": 2.1728880157170924, + "grad_norm": 0.3415327741517172, + "learning_rate": 5.569937323493253e-05, + "loss": 0.5209, + "step": 1106 + }, + { + "epoch": 2.174852652259332, + "grad_norm": 0.30928112771585864, + "learning_rate": 5.564888673678014e-05, + "loss": 0.5141, + "step": 1107 + }, + { + "epoch": 2.1768172888015718, + "grad_norm": 0.3953535770878428, + "learning_rate": 5.559837078680623e-05, + "loss": 0.5211, + "step": 1108 + }, + { + "epoch": 2.178781925343811, + "grad_norm": 0.498061582974242, + "learning_rate": 5.5547825480083826e-05, + "loss": 0.523, + "step": 1109 + }, + { + "epoch": 2.180746561886051, + "grad_norm": 0.5118328999253886, + "learning_rate": 5.549725091174117e-05, + "loss": 0.5111, + "step": 1110 + }, + { + "epoch": 2.182711198428291, + "grad_norm": 0.4636568982005801, + "learning_rate": 5.54466471769616e-05, + "loss": 0.5154, + "step": 1111 + }, + { + "epoch": 2.1846758349705304, + "grad_norm": 0.3143722045696957, + "learning_rate": 5.539601437098335e-05, + "loss": 0.5282, + "step": 1112 + }, + { + "epoch": 2.1866404715127703, + "grad_norm": 0.3484399192069343, + "learning_rate": 5.534535258909934e-05, + "loss": 0.5105, + "step": 1113 + }, + { + "epoch": 2.1886051080550097, + "grad_norm": 0.4078918948434889, + "learning_rate": 5.529466192665704e-05, + "loss": 0.5234, + "step": 1114 + }, + { + "epoch": 2.1905697445972496, + "grad_norm": 0.41819760974295056, + "learning_rate": 5.524394247905829e-05, + "loss": 0.523, + "step": 1115 + }, + { + "epoch": 2.192534381139489, + "grad_norm": 0.42998965820179247, + "learning_rate": 5.5193194341759074e-05, + "loss": 0.5208, + "step": 1116 + }, + { + "epoch": 2.194499017681729, + "grad_norm": 0.42811678888637394, + "learning_rate": 5.51424176102694e-05, + "loss": 0.5178, + "step": 1117 + }, + { + "epoch": 2.1964636542239684, + "grad_norm": 0.3057136832353512, + "learning_rate": 5.509161238015306e-05, + "loss": 0.5244, + "step": 1118 + }, + { + "epoch": 2.1984282907662083, + "grad_norm": 0.3082112763734912, + "learning_rate": 5.5040778747027504e-05, + "loss": 0.5153, + "step": 1119 + }, + { + "epoch": 2.2003929273084477, + "grad_norm": 0.2871755481122388, + "learning_rate": 5.4989916806563655e-05, + "loss": 0.5206, + "step": 1120 + }, + { + "epoch": 2.2023575638506876, + "grad_norm": 0.27581777476313096, + "learning_rate": 5.493902665448566e-05, + "loss": 0.5193, + "step": 1121 + }, + { + "epoch": 2.2043222003929275, + "grad_norm": 0.2777910418347305, + "learning_rate": 5.488810838657081e-05, + "loss": 0.5196, + "step": 1122 + }, + { + "epoch": 2.206286836935167, + "grad_norm": 0.2095746731633065, + "learning_rate": 5.483716209864928e-05, + "loss": 0.5175, + "step": 1123 + }, + { + "epoch": 2.208251473477407, + "grad_norm": 0.2497369212146621, + "learning_rate": 5.478618788660398e-05, + "loss": 0.5177, + "step": 1124 + }, + { + "epoch": 2.2102161100196462, + "grad_norm": 0.30316537694955104, + "learning_rate": 5.4735185846370415e-05, + "loss": 0.518, + "step": 1125 + }, + { + "epoch": 2.212180746561886, + "grad_norm": 0.3659038128390413, + "learning_rate": 5.468415607393643e-05, + "loss": 0.5191, + "step": 1126 + }, + { + "epoch": 2.2141453831041256, + "grad_norm": 0.33645381069137126, + "learning_rate": 5.463309866534205e-05, + "loss": 0.5118, + "step": 1127 + }, + { + "epoch": 2.2161100196463654, + "grad_norm": 0.2877438232092255, + "learning_rate": 5.458201371667933e-05, + "loss": 0.52, + "step": 1128 + }, + { + "epoch": 2.2180746561886053, + "grad_norm": 0.28342046851111374, + "learning_rate": 5.453090132409216e-05, + "loss": 0.5175, + "step": 1129 + }, + { + "epoch": 2.2200392927308448, + "grad_norm": 0.2983397453771918, + "learning_rate": 5.447976158377611e-05, + "loss": 0.5206, + "step": 1130 + }, + { + "epoch": 2.2220039292730847, + "grad_norm": 0.25696902289922835, + "learning_rate": 5.442859459197814e-05, + "loss": 0.5201, + "step": 1131 + }, + { + "epoch": 2.223968565815324, + "grad_norm": 0.28200785296195696, + "learning_rate": 5.437740044499656e-05, + "loss": 0.5127, + "step": 1132 + }, + { + "epoch": 2.225933202357564, + "grad_norm": 0.300428969109946, + "learning_rate": 5.432617923918077e-05, + "loss": 0.5162, + "step": 1133 + }, + { + "epoch": 2.2278978388998034, + "grad_norm": 0.36933445521619496, + "learning_rate": 5.427493107093112e-05, + "loss": 0.5099, + "step": 1134 + }, + { + "epoch": 2.2298624754420433, + "grad_norm": 0.3637711412912583, + "learning_rate": 5.422365603669866e-05, + "loss": 0.5196, + "step": 1135 + }, + { + "epoch": 2.2318271119842827, + "grad_norm": 0.35709267458544847, + "learning_rate": 5.417235423298503e-05, + "loss": 0.5281, + "step": 1136 + }, + { + "epoch": 2.2337917485265226, + "grad_norm": 0.3228959737121387, + "learning_rate": 5.412102575634224e-05, + "loss": 0.517, + "step": 1137 + }, + { + "epoch": 2.235756385068762, + "grad_norm": 0.2847481231302459, + "learning_rate": 5.406967070337252e-05, + "loss": 0.5207, + "step": 1138 + }, + { + "epoch": 2.237721021611002, + "grad_norm": 0.30242914098521595, + "learning_rate": 5.401828917072809e-05, + "loss": 0.5153, + "step": 1139 + }, + { + "epoch": 2.239685658153242, + "grad_norm": 0.2731961060540296, + "learning_rate": 5.3966881255111024e-05, + "loss": 0.5216, + "step": 1140 + }, + { + "epoch": 2.2416502946954813, + "grad_norm": 0.2814534534773246, + "learning_rate": 5.391544705327306e-05, + "loss": 0.5109, + "step": 1141 + }, + { + "epoch": 2.243614931237721, + "grad_norm": 0.27713371788941427, + "learning_rate": 5.386398666201536e-05, + "loss": 0.5211, + "step": 1142 + }, + { + "epoch": 2.2455795677799606, + "grad_norm": 0.27742830291773324, + "learning_rate": 5.381250017818843e-05, + "loss": 0.5209, + "step": 1143 + }, + { + "epoch": 2.2475442043222005, + "grad_norm": 0.2919588607608429, + "learning_rate": 5.376098769869187e-05, + "loss": 0.5182, + "step": 1144 + }, + { + "epoch": 2.24950884086444, + "grad_norm": 0.26026500913994954, + "learning_rate": 5.370944932047417e-05, + "loss": 0.5194, + "step": 1145 + }, + { + "epoch": 2.25147347740668, + "grad_norm": 0.27116717902351345, + "learning_rate": 5.365788514053262e-05, + "loss": 0.5176, + "step": 1146 + }, + { + "epoch": 2.2534381139489197, + "grad_norm": 0.2217611484705821, + "learning_rate": 5.360629525591301e-05, + "loss": 0.5182, + "step": 1147 + }, + { + "epoch": 2.255402750491159, + "grad_norm": 0.22882988446863112, + "learning_rate": 5.355467976370955e-05, + "loss": 0.5206, + "step": 1148 + }, + { + "epoch": 2.257367387033399, + "grad_norm": 0.2730234921167444, + "learning_rate": 5.3503038761064636e-05, + "loss": 0.519, + "step": 1149 + }, + { + "epoch": 2.2593320235756384, + "grad_norm": 0.24562754088282718, + "learning_rate": 5.345137234516864e-05, + "loss": 0.5238, + "step": 1150 + }, + { + "epoch": 2.2612966601178783, + "grad_norm": 0.19845273369186014, + "learning_rate": 5.3399680613259845e-05, + "loss": 0.5165, + "step": 1151 + }, + { + "epoch": 2.2632612966601178, + "grad_norm": 0.23286315375459862, + "learning_rate": 5.334796366262408e-05, + "loss": 0.5208, + "step": 1152 + }, + { + "epoch": 2.2652259332023577, + "grad_norm": 0.2902832652389449, + "learning_rate": 5.3296221590594695e-05, + "loss": 0.5133, + "step": 1153 + }, + { + "epoch": 2.267190569744597, + "grad_norm": 0.29420770210554437, + "learning_rate": 5.324445449455233e-05, + "loss": 0.5157, + "step": 1154 + }, + { + "epoch": 2.269155206286837, + "grad_norm": 0.3320621168111837, + "learning_rate": 5.3192662471924684e-05, + "loss": 0.5105, + "step": 1155 + }, + { + "epoch": 2.2711198428290764, + "grad_norm": 0.2910149138067965, + "learning_rate": 5.314084562018638e-05, + "loss": 0.5231, + "step": 1156 + }, + { + "epoch": 2.2730844793713163, + "grad_norm": 0.25692267160447746, + "learning_rate": 5.308900403685878e-05, + "loss": 0.5187, + "step": 1157 + }, + { + "epoch": 2.275049115913556, + "grad_norm": 0.263390283481988, + "learning_rate": 5.3037137819509786e-05, + "loss": 0.527, + "step": 1158 + }, + { + "epoch": 2.2770137524557956, + "grad_norm": 0.29854948931379893, + "learning_rate": 5.298524706575366e-05, + "loss": 0.5149, + "step": 1159 + }, + { + "epoch": 2.2789783889980355, + "grad_norm": 0.34965852786507345, + "learning_rate": 5.293333187325086e-05, + "loss": 0.5172, + "step": 1160 + }, + { + "epoch": 2.280943025540275, + "grad_norm": 0.40611230190283126, + "learning_rate": 5.288139233970781e-05, + "loss": 0.5289, + "step": 1161 + }, + { + "epoch": 2.282907662082515, + "grad_norm": 0.37958565343794637, + "learning_rate": 5.2829428562876754e-05, + "loss": 0.5161, + "step": 1162 + }, + { + "epoch": 2.2848722986247543, + "grad_norm": 0.3338201117928623, + "learning_rate": 5.277744064055558e-05, + "loss": 0.5315, + "step": 1163 + }, + { + "epoch": 2.286836935166994, + "grad_norm": 0.3514328811125468, + "learning_rate": 5.272542867058761e-05, + "loss": 0.5283, + "step": 1164 + }, + { + "epoch": 2.288801571709234, + "grad_norm": 0.3724289159855533, + "learning_rate": 5.267339275086141e-05, + "loss": 0.5161, + "step": 1165 + }, + { + "epoch": 2.2907662082514735, + "grad_norm": 0.43185626525647297, + "learning_rate": 5.262133297931064e-05, + "loss": 0.5165, + "step": 1166 + }, + { + "epoch": 2.292730844793713, + "grad_norm": 0.5557453693631192, + "learning_rate": 5.2569249453913815e-05, + "loss": 0.5162, + "step": 1167 + }, + { + "epoch": 2.294695481335953, + "grad_norm": 0.6169003305189821, + "learning_rate": 5.251714227269423e-05, + "loss": 0.519, + "step": 1168 + }, + { + "epoch": 2.2966601178781927, + "grad_norm": 0.5630541800912817, + "learning_rate": 5.246501153371961e-05, + "loss": 0.5216, + "step": 1169 + }, + { + "epoch": 2.298624754420432, + "grad_norm": 0.4031836951288518, + "learning_rate": 5.241285733510209e-05, + "loss": 0.505, + "step": 1170 + }, + { + "epoch": 2.300589390962672, + "grad_norm": 0.2838793794322749, + "learning_rate": 5.23606797749979e-05, + "loss": 0.5215, + "step": 1171 + }, + { + "epoch": 2.3025540275049114, + "grad_norm": 0.33367703166552265, + "learning_rate": 5.230847895160729e-05, + "loss": 0.5187, + "step": 1172 + }, + { + "epoch": 2.3045186640471513, + "grad_norm": 0.3708154204793441, + "learning_rate": 5.225625496317426e-05, + "loss": 0.5203, + "step": 1173 + }, + { + "epoch": 2.3064833005893908, + "grad_norm": 0.39313469215099084, + "learning_rate": 5.220400790798639e-05, + "loss": 0.5234, + "step": 1174 + }, + { + "epoch": 2.3084479371316307, + "grad_norm": 0.40695601363603545, + "learning_rate": 5.215173788437473e-05, + "loss": 0.5224, + "step": 1175 + }, + { + "epoch": 2.3104125736738705, + "grad_norm": 0.4982129751478107, + "learning_rate": 5.209944499071351e-05, + "loss": 0.5164, + "step": 1176 + }, + { + "epoch": 2.31237721021611, + "grad_norm": 0.4515408526502741, + "learning_rate": 5.2047129325420014e-05, + "loss": 0.5266, + "step": 1177 + }, + { + "epoch": 2.31434184675835, + "grad_norm": 0.2944209530909051, + "learning_rate": 5.199479098695439e-05, + "loss": 0.5236, + "step": 1178 + }, + { + "epoch": 2.3163064833005893, + "grad_norm": 0.3143033052771175, + "learning_rate": 5.194243007381945e-05, + "loss": 0.5216, + "step": 1179 + }, + { + "epoch": 2.318271119842829, + "grad_norm": 0.30494349124568865, + "learning_rate": 5.189004668456051e-05, + "loss": 0.5232, + "step": 1180 + }, + { + "epoch": 2.3202357563850686, + "grad_norm": 0.35922632879055943, + "learning_rate": 5.183764091776515e-05, + "loss": 0.5121, + "step": 1181 + }, + { + "epoch": 2.3222003929273085, + "grad_norm": 0.3455857421566445, + "learning_rate": 5.178521287206309e-05, + "loss": 0.5089, + "step": 1182 + }, + { + "epoch": 2.324165029469548, + "grad_norm": 0.348541856498887, + "learning_rate": 5.1732762646126005e-05, + "loss": 0.5262, + "step": 1183 + }, + { + "epoch": 2.326129666011788, + "grad_norm": 0.32156787348759275, + "learning_rate": 5.1680290338667265e-05, + "loss": 0.5193, + "step": 1184 + }, + { + "epoch": 2.3280943025540273, + "grad_norm": 0.27708493229619446, + "learning_rate": 5.1627796048441824e-05, + "loss": 0.5217, + "step": 1185 + }, + { + "epoch": 2.330058939096267, + "grad_norm": 0.3023478182961343, + "learning_rate": 5.1575279874246005e-05, + "loss": 0.5218, + "step": 1186 + }, + { + "epoch": 2.332023575638507, + "grad_norm": 0.2972904232192658, + "learning_rate": 5.152274191491732e-05, + "loss": 0.5306, + "step": 1187 + }, + { + "epoch": 2.3339882121807465, + "grad_norm": 0.350655771339677, + "learning_rate": 5.147018226933427e-05, + "loss": 0.518, + "step": 1188 + }, + { + "epoch": 2.3359528487229864, + "grad_norm": 0.320271515497672, + "learning_rate": 5.14176010364162e-05, + "loss": 0.5103, + "step": 1189 + }, + { + "epoch": 2.337917485265226, + "grad_norm": 0.29939194861707513, + "learning_rate": 5.1364998315123024e-05, + "loss": 0.5258, + "step": 1190 + }, + { + "epoch": 2.3398821218074657, + "grad_norm": 0.25962321018755774, + "learning_rate": 5.1312374204455175e-05, + "loss": 0.5184, + "step": 1191 + }, + { + "epoch": 2.341846758349705, + "grad_norm": 0.28011514210670607, + "learning_rate": 5.125972880345329e-05, + "loss": 0.5234, + "step": 1192 + }, + { + "epoch": 2.343811394891945, + "grad_norm": 0.3751902436574916, + "learning_rate": 5.120706221119809e-05, + "loss": 0.5285, + "step": 1193 + }, + { + "epoch": 2.345776031434185, + "grad_norm": 0.42074098588577236, + "learning_rate": 5.115437452681018e-05, + "loss": 0.53, + "step": 1194 + }, + { + "epoch": 2.3477406679764243, + "grad_norm": 0.3511624678398766, + "learning_rate": 5.1101665849449834e-05, + "loss": 0.522, + "step": 1195 + }, + { + "epoch": 2.349705304518664, + "grad_norm": 0.27221188674363855, + "learning_rate": 5.104893627831688e-05, + "loss": 0.5222, + "step": 1196 + }, + { + "epoch": 2.3516699410609037, + "grad_norm": 0.34790425636922584, + "learning_rate": 5.099618591265046e-05, + "loss": 0.5219, + "step": 1197 + }, + { + "epoch": 2.3536345776031435, + "grad_norm": 0.3493753275321162, + "learning_rate": 5.094341485172882e-05, + "loss": 0.5246, + "step": 1198 + }, + { + "epoch": 2.355599214145383, + "grad_norm": 0.34718407328605627, + "learning_rate": 5.089062319486919e-05, + "loss": 0.5301, + "step": 1199 + }, + { + "epoch": 2.357563850687623, + "grad_norm": 0.3119267686407314, + "learning_rate": 5.0837811041427534e-05, + "loss": 0.5224, + "step": 1200 + }, + { + "epoch": 2.3595284872298623, + "grad_norm": 0.3138196780734769, + "learning_rate": 5.078497849079842e-05, + "loss": 0.5113, + "step": 1201 + }, + { + "epoch": 2.361493123772102, + "grad_norm": 0.40280282592611577, + "learning_rate": 5.073212564241479e-05, + "loss": 0.5172, + "step": 1202 + }, + { + "epoch": 2.3634577603143416, + "grad_norm": 0.4198556229518336, + "learning_rate": 5.067925259574777e-05, + "loss": 0.5298, + "step": 1203 + }, + { + "epoch": 2.3654223968565815, + "grad_norm": 0.4040963983538494, + "learning_rate": 5.062635945030654e-05, + "loss": 0.5211, + "step": 1204 + }, + { + "epoch": 2.3673870333988214, + "grad_norm": 0.3598550221776955, + "learning_rate": 5.057344630563807e-05, + "loss": 0.5194, + "step": 1205 + }, + { + "epoch": 2.369351669941061, + "grad_norm": 0.3034914260205126, + "learning_rate": 5.0520513261326985e-05, + "loss": 0.5253, + "step": 1206 + }, + { + "epoch": 2.3713163064833007, + "grad_norm": 0.22348880732221432, + "learning_rate": 5.046756041699537e-05, + "loss": 0.5231, + "step": 1207 + }, + { + "epoch": 2.37328094302554, + "grad_norm": 0.21772886050360396, + "learning_rate": 5.041458787230255e-05, + "loss": 0.5092, + "step": 1208 + }, + { + "epoch": 2.37524557956778, + "grad_norm": 0.24524426361184035, + "learning_rate": 5.0361595726944955e-05, + "loss": 0.5179, + "step": 1209 + }, + { + "epoch": 2.3772102161100195, + "grad_norm": 0.32920978262279904, + "learning_rate": 5.030858408065587e-05, + "loss": 0.5307, + "step": 1210 + }, + { + "epoch": 2.3791748526522594, + "grad_norm": 0.31365670018548764, + "learning_rate": 5.025555303320532e-05, + "loss": 0.5201, + "step": 1211 + }, + { + "epoch": 2.3811394891944992, + "grad_norm": 0.2933255231708542, + "learning_rate": 5.020250268439984e-05, + "loss": 0.5229, + "step": 1212 + }, + { + "epoch": 2.3831041257367387, + "grad_norm": 0.2604865427319619, + "learning_rate": 5.014943313408225e-05, + "loss": 0.5243, + "step": 1213 + }, + { + "epoch": 2.3850687622789786, + "grad_norm": 0.18783141094877523, + "learning_rate": 5.009634448213155e-05, + "loss": 0.5261, + "step": 1214 + }, + { + "epoch": 2.387033398821218, + "grad_norm": 0.3194710529072846, + "learning_rate": 5.0043236828462665e-05, + "loss": 0.535, + "step": 1215 + }, + { + "epoch": 2.388998035363458, + "grad_norm": 0.39039536907744726, + "learning_rate": 4.999011027302629e-05, + "loss": 0.5222, + "step": 1216 + }, + { + "epoch": 2.3909626719056973, + "grad_norm": 0.4288323915503875, + "learning_rate": 4.993696491580872e-05, + "loss": 0.5183, + "step": 1217 + }, + { + "epoch": 2.392927308447937, + "grad_norm": 0.37497328282617143, + "learning_rate": 4.988380085683158e-05, + "loss": 0.5267, + "step": 1218 + }, + { + "epoch": 2.3948919449901767, + "grad_norm": 0.3425366395234411, + "learning_rate": 4.983061819615174e-05, + "loss": 0.5356, + "step": 1219 + }, + { + "epoch": 2.3968565815324165, + "grad_norm": 0.3578086926911884, + "learning_rate": 4.977741703386107e-05, + "loss": 0.5269, + "step": 1220 + }, + { + "epoch": 2.398821218074656, + "grad_norm": 0.3125850197507908, + "learning_rate": 4.9724197470086236e-05, + "loss": 0.5131, + "step": 1221 + }, + { + "epoch": 2.400785854616896, + "grad_norm": 0.33173541611268587, + "learning_rate": 4.9670959604988574e-05, + "loss": 0.5213, + "step": 1222 + }, + { + "epoch": 2.4027504911591357, + "grad_norm": 0.4651837337265221, + "learning_rate": 4.961770353876383e-05, + "loss": 0.5174, + "step": 1223 + }, + { + "epoch": 2.404715127701375, + "grad_norm": 0.4969179383448936, + "learning_rate": 4.9564429371642016e-05, + "loss": 0.5185, + "step": 1224 + }, + { + "epoch": 2.406679764243615, + "grad_norm": 0.42072059154953617, + "learning_rate": 4.951113720388723e-05, + "loss": 0.5244, + "step": 1225 + }, + { + "epoch": 2.4086444007858545, + "grad_norm": 0.25638662286022346, + "learning_rate": 4.9457827135797406e-05, + "loss": 0.5188, + "step": 1226 + }, + { + "epoch": 2.4106090373280944, + "grad_norm": 0.2878940895980239, + "learning_rate": 4.9404499267704236e-05, + "loss": 0.5206, + "step": 1227 + }, + { + "epoch": 2.412573673870334, + "grad_norm": 0.3832533783751847, + "learning_rate": 4.9351153699972836e-05, + "loss": 0.5282, + "step": 1228 + }, + { + "epoch": 2.4145383104125737, + "grad_norm": 0.45725279483268727, + "learning_rate": 4.929779053300166e-05, + "loss": 0.5131, + "step": 1229 + }, + { + "epoch": 2.4165029469548136, + "grad_norm": 0.4058965658089161, + "learning_rate": 4.924440986722231e-05, + "loss": 0.5238, + "step": 1230 + }, + { + "epoch": 2.418467583497053, + "grad_norm": 0.32191319084337217, + "learning_rate": 4.919101180309931e-05, + "loss": 0.5138, + "step": 1231 + }, + { + "epoch": 2.4204322200392925, + "grad_norm": 0.302751032659875, + "learning_rate": 4.91375964411299e-05, + "loss": 0.5105, + "step": 1232 + }, + { + "epoch": 2.4223968565815324, + "grad_norm": 0.2957972007864704, + "learning_rate": 4.90841638818439e-05, + "loss": 0.5245, + "step": 1233 + }, + { + "epoch": 2.4243614931237722, + "grad_norm": 0.24135783608419617, + "learning_rate": 4.9030714225803496e-05, + "loss": 0.5193, + "step": 1234 + }, + { + "epoch": 2.4263261296660117, + "grad_norm": 0.22844179039247656, + "learning_rate": 4.8977247573603046e-05, + "loss": 0.5231, + "step": 1235 + }, + { + "epoch": 2.4282907662082516, + "grad_norm": 0.21987219483583326, + "learning_rate": 4.892376402586891e-05, + "loss": 0.5229, + "step": 1236 + }, + { + "epoch": 2.430255402750491, + "grad_norm": 0.2614602309973377, + "learning_rate": 4.887026368325921e-05, + "loss": 0.5293, + "step": 1237 + }, + { + "epoch": 2.432220039292731, + "grad_norm": 0.2660424537907494, + "learning_rate": 4.881674664646371e-05, + "loss": 0.5163, + "step": 1238 + }, + { + "epoch": 2.4341846758349703, + "grad_norm": 0.25356892729724706, + "learning_rate": 4.8763213016203594e-05, + "loss": 0.5125, + "step": 1239 + }, + { + "epoch": 2.43614931237721, + "grad_norm": 0.2722693788324859, + "learning_rate": 4.870966289323123e-05, + "loss": 0.5206, + "step": 1240 + }, + { + "epoch": 2.43811394891945, + "grad_norm": 0.26861376150737215, + "learning_rate": 4.86560963783301e-05, + "loss": 0.5297, + "step": 1241 + }, + { + "epoch": 2.4400785854616895, + "grad_norm": 0.2355384488763813, + "learning_rate": 4.860251357231446e-05, + "loss": 0.5206, + "step": 1242 + }, + { + "epoch": 2.4420432220039294, + "grad_norm": 0.24081050892540198, + "learning_rate": 4.8548914576029306e-05, + "loss": 0.5254, + "step": 1243 + }, + { + "epoch": 2.444007858546169, + "grad_norm": 0.26501560178704436, + "learning_rate": 4.849529949035001e-05, + "loss": 0.5179, + "step": 1244 + }, + { + "epoch": 2.4459724950884087, + "grad_norm": 0.22854433959933895, + "learning_rate": 4.84416684161823e-05, + "loss": 0.5297, + "step": 1245 + }, + { + "epoch": 2.447937131630648, + "grad_norm": 0.22612719987730076, + "learning_rate": 4.838802145446199e-05, + "loss": 0.5223, + "step": 1246 + }, + { + "epoch": 2.449901768172888, + "grad_norm": 0.26406587116450525, + "learning_rate": 4.833435870615474e-05, + "loss": 0.5207, + "step": 1247 + }, + { + "epoch": 2.4518664047151275, + "grad_norm": 0.18949918907380475, + "learning_rate": 4.828068027225599e-05, + "loss": 0.5136, + "step": 1248 + }, + { + "epoch": 2.4538310412573674, + "grad_norm": 0.18825528437402816, + "learning_rate": 4.822698625379063e-05, + "loss": 0.526, + "step": 1249 + }, + { + "epoch": 2.455795677799607, + "grad_norm": 0.2627421106260628, + "learning_rate": 4.817327675181297e-05, + "loss": 0.5263, + "step": 1250 + }, + { + "epoch": 2.4577603143418467, + "grad_norm": 0.3290771030055934, + "learning_rate": 4.811955186740637e-05, + "loss": 0.5215, + "step": 1251 + }, + { + "epoch": 2.4597249508840866, + "grad_norm": 0.2939415056251606, + "learning_rate": 4.806581170168318e-05, + "loss": 0.5189, + "step": 1252 + }, + { + "epoch": 2.461689587426326, + "grad_norm": 0.22542334269014147, + "learning_rate": 4.801205635578453e-05, + "loss": 0.5331, + "step": 1253 + }, + { + "epoch": 2.463654223968566, + "grad_norm": 0.24239948365032304, + "learning_rate": 4.795828593088008e-05, + "loss": 0.5209, + "step": 1254 + }, + { + "epoch": 2.4656188605108054, + "grad_norm": 0.24234857298555373, + "learning_rate": 4.79045005281679e-05, + "loss": 0.51, + "step": 1255 + }, + { + "epoch": 2.4675834970530452, + "grad_norm": 0.23518303728677917, + "learning_rate": 4.7850700248874244e-05, + "loss": 0.5192, + "step": 1256 + }, + { + "epoch": 2.4695481335952847, + "grad_norm": 0.20023881167921215, + "learning_rate": 4.779688519425334e-05, + "loss": 0.5268, + "step": 1257 + }, + { + "epoch": 2.4715127701375246, + "grad_norm": 0.26898366399464796, + "learning_rate": 4.7743055465587256e-05, + "loss": 0.5263, + "step": 1258 + }, + { + "epoch": 2.4734774066797645, + "grad_norm": 0.28498767209944564, + "learning_rate": 4.768921116418565e-05, + "loss": 0.5217, + "step": 1259 + }, + { + "epoch": 2.475442043222004, + "grad_norm": 0.26002114329288895, + "learning_rate": 4.763535239138562e-05, + "loss": 0.5233, + "step": 1260 + }, + { + "epoch": 2.477406679764244, + "grad_norm": 0.2542102589395213, + "learning_rate": 4.758147924855152e-05, + "loss": 0.517, + "step": 1261 + }, + { + "epoch": 2.479371316306483, + "grad_norm": 0.2803656430935204, + "learning_rate": 4.752759183707471e-05, + "loss": 0.5201, + "step": 1262 + }, + { + "epoch": 2.481335952848723, + "grad_norm": 0.27346725141790124, + "learning_rate": 4.747369025837341e-05, + "loss": 0.5149, + "step": 1263 + }, + { + "epoch": 2.4833005893909625, + "grad_norm": 0.22656879345645617, + "learning_rate": 4.741977461389255e-05, + "loss": 0.5164, + "step": 1264 + }, + { + "epoch": 2.4852652259332024, + "grad_norm": 0.22759904451699164, + "learning_rate": 4.736584500510347e-05, + "loss": 0.5258, + "step": 1265 + }, + { + "epoch": 2.487229862475442, + "grad_norm": 0.22286866195097912, + "learning_rate": 4.7311901533503817e-05, + "loss": 0.5154, + "step": 1266 + }, + { + "epoch": 2.4891944990176817, + "grad_norm": 0.2630081929017963, + "learning_rate": 4.725794430061735e-05, + "loss": 0.5273, + "step": 1267 + }, + { + "epoch": 2.491159135559921, + "grad_norm": 0.21847942570079884, + "learning_rate": 4.720397340799369e-05, + "loss": 0.5158, + "step": 1268 + }, + { + "epoch": 2.493123772102161, + "grad_norm": 0.22951117374624064, + "learning_rate": 4.7149988957208184e-05, + "loss": 0.525, + "step": 1269 + }, + { + "epoch": 2.495088408644401, + "grad_norm": 0.2167477754698496, + "learning_rate": 4.709599104986172e-05, + "loss": 0.5146, + "step": 1270 + }, + { + "epoch": 2.4970530451866404, + "grad_norm": 0.22875092220832835, + "learning_rate": 4.704197978758046e-05, + "loss": 0.5272, + "step": 1271 + }, + { + "epoch": 2.4990176817288803, + "grad_norm": 0.2728399601902931, + "learning_rate": 4.698795527201574e-05, + "loss": 0.521, + "step": 1272 + }, + { + "epoch": 2.5009823182711197, + "grad_norm": 0.31422784807178733, + "learning_rate": 4.693391760484382e-05, + "loss": 0.5167, + "step": 1273 + }, + { + "epoch": 2.5029469548133596, + "grad_norm": 0.294737188915029, + "learning_rate": 4.6879866887765723e-05, + "loss": 0.521, + "step": 1274 + }, + { + "epoch": 2.504911591355599, + "grad_norm": 0.2787227328301998, + "learning_rate": 4.682580322250705e-05, + "loss": 0.5132, + "step": 1275 + }, + { + "epoch": 2.506876227897839, + "grad_norm": 0.2739005840040005, + "learning_rate": 4.6771726710817706e-05, + "loss": 0.5111, + "step": 1276 + }, + { + "epoch": 2.508840864440079, + "grad_norm": 0.23924145768311364, + "learning_rate": 4.671763745447186e-05, + "loss": 0.5172, + "step": 1277 + }, + { + "epoch": 2.5108055009823183, + "grad_norm": 0.2676623629031717, + "learning_rate": 4.6663535555267584e-05, + "loss": 0.5225, + "step": 1278 + }, + { + "epoch": 2.5127701375245577, + "grad_norm": 0.22207523048211628, + "learning_rate": 4.660942111502681e-05, + "loss": 0.5197, + "step": 1279 + }, + { + "epoch": 2.5147347740667976, + "grad_norm": 0.2293704129240573, + "learning_rate": 4.6555294235595047e-05, + "loss": 0.5182, + "step": 1280 + }, + { + "epoch": 2.5166994106090375, + "grad_norm": 0.2129707568443277, + "learning_rate": 4.650115501884121e-05, + "loss": 0.5178, + "step": 1281 + }, + { + "epoch": 2.518664047151277, + "grad_norm": 0.23263717689491611, + "learning_rate": 4.644700356665744e-05, + "loss": 0.5275, + "step": 1282 + }, + { + "epoch": 2.520628683693517, + "grad_norm": 0.2646229072911609, + "learning_rate": 4.6392839980958886e-05, + "loss": 0.5283, + "step": 1283 + }, + { + "epoch": 2.522593320235756, + "grad_norm": 0.3061285163686524, + "learning_rate": 4.633866436368358e-05, + "loss": 0.521, + "step": 1284 + }, + { + "epoch": 2.524557956777996, + "grad_norm": 0.2329803508439659, + "learning_rate": 4.628447681679215e-05, + "loss": 0.5247, + "step": 1285 + }, + { + "epoch": 2.5265225933202355, + "grad_norm": 0.21373280132034286, + "learning_rate": 4.623027744226771e-05, + "loss": 0.5337, + "step": 1286 + }, + { + "epoch": 2.5284872298624754, + "grad_norm": 0.24515628078706708, + "learning_rate": 4.61760663421156e-05, + "loss": 0.5251, + "step": 1287 + }, + { + "epoch": 2.5304518664047153, + "grad_norm": 0.2100951375789621, + "learning_rate": 4.612184361836325e-05, + "loss": 0.5192, + "step": 1288 + }, + { + "epoch": 2.5324165029469548, + "grad_norm": 0.19545834874501095, + "learning_rate": 4.606760937305997e-05, + "loss": 0.5223, + "step": 1289 + }, + { + "epoch": 2.5343811394891946, + "grad_norm": 0.22770466382912533, + "learning_rate": 4.6013363708276754e-05, + "loss": 0.5217, + "step": 1290 + }, + { + "epoch": 2.536345776031434, + "grad_norm": 0.2191253173653033, + "learning_rate": 4.595910672610607e-05, + "loss": 0.522, + "step": 1291 + }, + { + "epoch": 2.538310412573674, + "grad_norm": 0.25291365059841875, + "learning_rate": 4.590483852866168e-05, + "loss": 0.5185, + "step": 1292 + }, + { + "epoch": 2.5402750491159134, + "grad_norm": 0.2797101923913896, + "learning_rate": 4.585055921807851e-05, + "loss": 0.526, + "step": 1293 + }, + { + "epoch": 2.5422396856581533, + "grad_norm": 0.25701240661178854, + "learning_rate": 4.579626889651233e-05, + "loss": 0.5313, + "step": 1294 + }, + { + "epoch": 2.544204322200393, + "grad_norm": 0.23778288294527172, + "learning_rate": 4.5741967666139676e-05, + "loss": 0.5157, + "step": 1295 + }, + { + "epoch": 2.5461689587426326, + "grad_norm": 0.324607809035794, + "learning_rate": 4.568765562915761e-05, + "loss": 0.5175, + "step": 1296 + }, + { + "epoch": 2.548133595284872, + "grad_norm": 0.35809692737350896, + "learning_rate": 4.56333328877835e-05, + "loss": 0.5214, + "step": 1297 + }, + { + "epoch": 2.550098231827112, + "grad_norm": 0.2598496968905382, + "learning_rate": 4.557899954425492e-05, + "loss": 0.5122, + "step": 1298 + }, + { + "epoch": 2.552062868369352, + "grad_norm": 0.22630825680763897, + "learning_rate": 4.5524655700829346e-05, + "loss": 0.5323, + "step": 1299 + }, + { + "epoch": 2.5540275049115913, + "grad_norm": 0.23882302446197487, + "learning_rate": 4.5470301459784026e-05, + "loss": 0.5327, + "step": 1300 + }, + { + "epoch": 2.555992141453831, + "grad_norm": 0.261961348410343, + "learning_rate": 4.5415936923415795e-05, + "loss": 0.5227, + "step": 1301 + }, + { + "epoch": 2.5579567779960706, + "grad_norm": 0.23441764282223249, + "learning_rate": 4.536156219404084e-05, + "loss": 0.5136, + "step": 1302 + }, + { + "epoch": 2.5599214145383105, + "grad_norm": 0.20563602504887254, + "learning_rate": 4.5307177373994566e-05, + "loss": 0.5243, + "step": 1303 + }, + { + "epoch": 2.56188605108055, + "grad_norm": 0.2267823493172774, + "learning_rate": 4.525278256563134e-05, + "loss": 0.5181, + "step": 1304 + }, + { + "epoch": 2.56385068762279, + "grad_norm": 0.28309713249973456, + "learning_rate": 4.5198377871324315e-05, + "loss": 0.5279, + "step": 1305 + }, + { + "epoch": 2.5658153241650297, + "grad_norm": 0.21364816105047454, + "learning_rate": 4.51439633934653e-05, + "loss": 0.5241, + "step": 1306 + }, + { + "epoch": 2.567779960707269, + "grad_norm": 0.21218819280323253, + "learning_rate": 4.508953923446445e-05, + "loss": 0.5232, + "step": 1307 + }, + { + "epoch": 2.569744597249509, + "grad_norm": 0.2409679906097701, + "learning_rate": 4.5035105496750216e-05, + "loss": 0.5207, + "step": 1308 + }, + { + "epoch": 2.5717092337917484, + "grad_norm": 0.2612138227652981, + "learning_rate": 4.4980662282769024e-05, + "loss": 0.5136, + "step": 1309 + }, + { + "epoch": 2.5736738703339883, + "grad_norm": 0.2704137513844624, + "learning_rate": 4.4926209694985134e-05, + "loss": 0.5333, + "step": 1310 + }, + { + "epoch": 2.5756385068762278, + "grad_norm": 0.28200487469275926, + "learning_rate": 4.4871747835880466e-05, + "loss": 0.5206, + "step": 1311 + }, + { + "epoch": 2.5776031434184676, + "grad_norm": 0.24262162651090807, + "learning_rate": 4.481727680795441e-05, + "loss": 0.5289, + "step": 1312 + }, + { + "epoch": 2.5795677799607075, + "grad_norm": 0.20644928721791156, + "learning_rate": 4.4762796713723555e-05, + "loss": 0.5224, + "step": 1313 + }, + { + "epoch": 2.581532416502947, + "grad_norm": 0.24435226223282838, + "learning_rate": 4.4708307655721604e-05, + "loss": 0.5141, + "step": 1314 + }, + { + "epoch": 2.5834970530451864, + "grad_norm": 0.38625441604541144, + "learning_rate": 4.465380973649909e-05, + "loss": 0.5279, + "step": 1315 + }, + { + "epoch": 2.5854616895874263, + "grad_norm": 0.37000625438186896, + "learning_rate": 4.459930305862327e-05, + "loss": 0.5269, + "step": 1316 + }, + { + "epoch": 2.587426326129666, + "grad_norm": 0.27635067596950225, + "learning_rate": 4.454478772467786e-05, + "loss": 0.5283, + "step": 1317 + }, + { + "epoch": 2.5893909626719056, + "grad_norm": 0.25422770322343335, + "learning_rate": 4.449026383726284e-05, + "loss": 0.5251, + "step": 1318 + }, + { + "epoch": 2.5913555992141455, + "grad_norm": 0.26280096378800966, + "learning_rate": 4.443573149899434e-05, + "loss": 0.5188, + "step": 1319 + }, + { + "epoch": 2.593320235756385, + "grad_norm": 0.28612652073421474, + "learning_rate": 4.4381190812504375e-05, + "loss": 0.5207, + "step": 1320 + }, + { + "epoch": 2.595284872298625, + "grad_norm": 0.2559967603139363, + "learning_rate": 4.432664188044061e-05, + "loss": 0.5192, + "step": 1321 + }, + { + "epoch": 2.5972495088408643, + "grad_norm": 0.3038990700816129, + "learning_rate": 4.4272084805466366e-05, + "loss": 0.527, + "step": 1322 + }, + { + "epoch": 2.599214145383104, + "grad_norm": 0.2699702257301053, + "learning_rate": 4.421751969026016e-05, + "loss": 0.521, + "step": 1323 + }, + { + "epoch": 2.601178781925344, + "grad_norm": 0.24668170360583927, + "learning_rate": 4.416294663751569e-05, + "loss": 0.5259, + "step": 1324 + }, + { + "epoch": 2.6031434184675835, + "grad_norm": 0.26819257437647714, + "learning_rate": 4.4108365749941616e-05, + "loss": 0.524, + "step": 1325 + }, + { + "epoch": 2.6051080550098233, + "grad_norm": 0.23325981392269732, + "learning_rate": 4.405377713026128e-05, + "loss": 0.5138, + "step": 1326 + }, + { + "epoch": 2.607072691552063, + "grad_norm": 0.25774911447701937, + "learning_rate": 4.399918088121266e-05, + "loss": 0.5226, + "step": 1327 + }, + { + "epoch": 2.6090373280943027, + "grad_norm": 0.2492006008288357, + "learning_rate": 4.394457710554801e-05, + "loss": 0.5259, + "step": 1328 + }, + { + "epoch": 2.611001964636542, + "grad_norm": 0.25981470449627336, + "learning_rate": 4.388996590603381e-05, + "loss": 0.5203, + "step": 1329 + }, + { + "epoch": 2.612966601178782, + "grad_norm": 0.21598020642923996, + "learning_rate": 4.3835347385450487e-05, + "loss": 0.525, + "step": 1330 + }, + { + "epoch": 2.614931237721022, + "grad_norm": 0.28945143397976597, + "learning_rate": 4.378072164659224e-05, + "loss": 0.508, + "step": 1331 + }, + { + "epoch": 2.6168958742632613, + "grad_norm": 0.3416596875235803, + "learning_rate": 4.372608879226688e-05, + "loss": 0.5193, + "step": 1332 + }, + { + "epoch": 2.6188605108055008, + "grad_norm": 0.27009282123283607, + "learning_rate": 4.367144892529558e-05, + "loss": 0.5175, + "step": 1333 + }, + { + "epoch": 2.6208251473477406, + "grad_norm": 0.21661716341384998, + "learning_rate": 4.3616802148512703e-05, + "loss": 0.5198, + "step": 1334 + }, + { + "epoch": 2.6227897838899805, + "grad_norm": 0.29020844521690053, + "learning_rate": 4.356214856476567e-05, + "loss": 0.5157, + "step": 1335 + }, + { + "epoch": 2.62475442043222, + "grad_norm": 0.2962753668865032, + "learning_rate": 4.350748827691465e-05, + "loss": 0.5252, + "step": 1336 + }, + { + "epoch": 2.62671905697446, + "grad_norm": 0.2851361563216777, + "learning_rate": 4.3452821387832496e-05, + "loss": 0.5235, + "step": 1337 + }, + { + "epoch": 2.6286836935166993, + "grad_norm": 0.3470347777192365, + "learning_rate": 4.339814800040442e-05, + "loss": 0.5128, + "step": 1338 + }, + { + "epoch": 2.630648330058939, + "grad_norm": 0.3065767806082005, + "learning_rate": 4.334346821752789e-05, + "loss": 0.5146, + "step": 1339 + }, + { + "epoch": 2.6326129666011786, + "grad_norm": 0.25008516467272474, + "learning_rate": 4.3288782142112425e-05, + "loss": 0.5208, + "step": 1340 + }, + { + "epoch": 2.6345776031434185, + "grad_norm": 0.24283401917757463, + "learning_rate": 4.323408987707937e-05, + "loss": 0.5221, + "step": 1341 + }, + { + "epoch": 2.6365422396856584, + "grad_norm": 0.26610349261153804, + "learning_rate": 4.317939152536171e-05, + "loss": 0.5209, + "step": 1342 + }, + { + "epoch": 2.638506876227898, + "grad_norm": 0.2169278403001787, + "learning_rate": 4.312468718990392e-05, + "loss": 0.5248, + "step": 1343 + }, + { + "epoch": 2.6404715127701373, + "grad_norm": 0.2173649225241763, + "learning_rate": 4.3069976973661697e-05, + "loss": 0.5242, + "step": 1344 + }, + { + "epoch": 2.642436149312377, + "grad_norm": 0.24732127593991116, + "learning_rate": 4.3015260979601835e-05, + "loss": 0.5108, + "step": 1345 + }, + { + "epoch": 2.644400785854617, + "grad_norm": 0.2512413940917044, + "learning_rate": 4.296053931070199e-05, + "loss": 0.5136, + "step": 1346 + }, + { + "epoch": 2.6463654223968565, + "grad_norm": 0.20543283854751387, + "learning_rate": 4.290581206995049e-05, + "loss": 0.5189, + "step": 1347 + }, + { + "epoch": 2.6483300589390963, + "grad_norm": 0.2558779150231826, + "learning_rate": 4.285107936034618e-05, + "loss": 0.5167, + "step": 1348 + }, + { + "epoch": 2.6502946954813362, + "grad_norm": 0.2320074012554696, + "learning_rate": 4.2796341284898144e-05, + "loss": 0.5226, + "step": 1349 + }, + { + "epoch": 2.6522593320235757, + "grad_norm": 0.24148125872376586, + "learning_rate": 4.274159794662563e-05, + "loss": 0.5176, + "step": 1350 + }, + { + "epoch": 2.654223968565815, + "grad_norm": 0.29335561766997686, + "learning_rate": 4.2686849448557736e-05, + "loss": 0.522, + "step": 1351 + }, + { + "epoch": 2.656188605108055, + "grad_norm": 0.331651025146379, + "learning_rate": 4.26320958937333e-05, + "loss": 0.5189, + "step": 1352 + }, + { + "epoch": 2.658153241650295, + "grad_norm": 0.284763896513611, + "learning_rate": 4.2577337385200665e-05, + "loss": 0.527, + "step": 1353 + }, + { + "epoch": 2.6601178781925343, + "grad_norm": 0.2345207108550884, + "learning_rate": 4.252257402601752e-05, + "loss": 0.5226, + "step": 1354 + }, + { + "epoch": 2.662082514734774, + "grad_norm": 0.2748637808818834, + "learning_rate": 4.246780591925063e-05, + "loss": 0.5173, + "step": 1355 + }, + { + "epoch": 2.6640471512770136, + "grad_norm": 0.30299246126424806, + "learning_rate": 4.2413033167975774e-05, + "loss": 0.5177, + "step": 1356 + }, + { + "epoch": 2.6660117878192535, + "grad_norm": 0.3235394094312818, + "learning_rate": 4.235825587527739e-05, + "loss": 0.522, + "step": 1357 + }, + { + "epoch": 2.667976424361493, + "grad_norm": 0.3434670261475503, + "learning_rate": 4.230347414424853e-05, + "loss": 0.5217, + "step": 1358 + }, + { + "epoch": 2.669941060903733, + "grad_norm": 0.2665713374075066, + "learning_rate": 4.2248688077990546e-05, + "loss": 0.5199, + "step": 1359 + }, + { + "epoch": 2.6719056974459727, + "grad_norm": 0.2602199491761626, + "learning_rate": 4.219389777961299e-05, + "loss": 0.5157, + "step": 1360 + }, + { + "epoch": 2.673870333988212, + "grad_norm": 0.30305452132006006, + "learning_rate": 4.2139103352233365e-05, + "loss": 0.5235, + "step": 1361 + }, + { + "epoch": 2.6758349705304516, + "grad_norm": 0.32202841146222816, + "learning_rate": 4.2084304898976935e-05, + "loss": 0.5217, + "step": 1362 + }, + { + "epoch": 2.6777996070726915, + "grad_norm": 0.35179132187779805, + "learning_rate": 4.202950252297654e-05, + "loss": 0.5173, + "step": 1363 + }, + { + "epoch": 2.6797642436149314, + "grad_norm": 0.2889177135873846, + "learning_rate": 4.1974696327372426e-05, + "loss": 0.5186, + "step": 1364 + }, + { + "epoch": 2.681728880157171, + "grad_norm": 0.26762397729246523, + "learning_rate": 4.1919886415312004e-05, + "loss": 0.5178, + "step": 1365 + }, + { + "epoch": 2.6836935166994107, + "grad_norm": 0.21518990648234254, + "learning_rate": 4.186507288994969e-05, + "loss": 0.5333, + "step": 1366 + }, + { + "epoch": 2.68565815324165, + "grad_norm": 0.2630456263477585, + "learning_rate": 4.181025585444672e-05, + "loss": 0.5252, + "step": 1367 + }, + { + "epoch": 2.68762278978389, + "grad_norm": 0.2243062547691368, + "learning_rate": 4.175543541197088e-05, + "loss": 0.521, + "step": 1368 + }, + { + "epoch": 2.6895874263261295, + "grad_norm": 0.25078505763565373, + "learning_rate": 4.170061166569641e-05, + "loss": 0.5202, + "step": 1369 + }, + { + "epoch": 2.6915520628683693, + "grad_norm": 0.2692196814658006, + "learning_rate": 4.164578471880378e-05, + "loss": 0.5223, + "step": 1370 + }, + { + "epoch": 2.6935166994106092, + "grad_norm": 0.24347540066783038, + "learning_rate": 4.159095467447945e-05, + "loss": 0.5139, + "step": 1371 + }, + { + "epoch": 2.6954813359528487, + "grad_norm": 0.2188314347988433, + "learning_rate": 4.153612163591574e-05, + "loss": 0.5142, + "step": 1372 + }, + { + "epoch": 2.6974459724950886, + "grad_norm": 0.24014195890047607, + "learning_rate": 4.1481285706310564e-05, + "loss": 0.5234, + "step": 1373 + }, + { + "epoch": 2.699410609037328, + "grad_norm": 0.2234377959437328, + "learning_rate": 4.142644698886732e-05, + "loss": 0.5224, + "step": 1374 + }, + { + "epoch": 2.701375245579568, + "grad_norm": 0.21331776330039912, + "learning_rate": 4.137160558679463e-05, + "loss": 0.5173, + "step": 1375 + }, + { + "epoch": 2.7033398821218073, + "grad_norm": 0.20551808394481486, + "learning_rate": 4.131676160330618e-05, + "loss": 0.5183, + "step": 1376 + }, + { + "epoch": 2.705304518664047, + "grad_norm": 0.22823097370935677, + "learning_rate": 4.126191514162051e-05, + "loss": 0.5205, + "step": 1377 + }, + { + "epoch": 2.707269155206287, + "grad_norm": 0.29094981706004336, + "learning_rate": 4.1207066304960797e-05, + "loss": 0.5204, + "step": 1378 + }, + { + "epoch": 2.7092337917485265, + "grad_norm": 0.25177696995986226, + "learning_rate": 4.115221519655474e-05, + "loss": 0.5245, + "step": 1379 + }, + { + "epoch": 2.711198428290766, + "grad_norm": 0.1904537069262316, + "learning_rate": 4.109736191963427e-05, + "loss": 0.5251, + "step": 1380 + }, + { + "epoch": 2.713163064833006, + "grad_norm": 0.240161539646952, + "learning_rate": 4.104250657743544e-05, + "loss": 0.5126, + "step": 1381 + }, + { + "epoch": 2.7151277013752457, + "grad_norm": 0.21860188786678716, + "learning_rate": 4.0987649273198145e-05, + "loss": 0.5185, + "step": 1382 + }, + { + "epoch": 2.717092337917485, + "grad_norm": 0.23402048721378071, + "learning_rate": 4.093279011016599e-05, + "loss": 0.5288, + "step": 1383 + }, + { + "epoch": 2.719056974459725, + "grad_norm": 0.20429494761855074, + "learning_rate": 4.087792919158608e-05, + "loss": 0.5152, + "step": 1384 + }, + { + "epoch": 2.7210216110019645, + "grad_norm": 0.19958313514866508, + "learning_rate": 4.082306662070887e-05, + "loss": 0.5131, + "step": 1385 + }, + { + "epoch": 2.7229862475442044, + "grad_norm": 0.2508712728836047, + "learning_rate": 4.0768202500787806e-05, + "loss": 0.5248, + "step": 1386 + }, + { + "epoch": 2.724950884086444, + "grad_norm": 0.2528312213587209, + "learning_rate": 4.0713336935079393e-05, + "loss": 0.5277, + "step": 1387 + }, + { + "epoch": 2.7269155206286837, + "grad_norm": 0.21024465894391983, + "learning_rate": 4.065847002684274e-05, + "loss": 0.5279, + "step": 1388 + }, + { + "epoch": 2.7288801571709236, + "grad_norm": 0.22892296955305103, + "learning_rate": 4.060360187933955e-05, + "loss": 0.5204, + "step": 1389 + }, + { + "epoch": 2.730844793713163, + "grad_norm": 0.2489587591523959, + "learning_rate": 4.054873259583385e-05, + "loss": 0.5147, + "step": 1390 + }, + { + "epoch": 2.732809430255403, + "grad_norm": 0.2033484751923835, + "learning_rate": 4.0493862279591764e-05, + "loss": 0.5084, + "step": 1391 + }, + { + "epoch": 2.7347740667976423, + "grad_norm": 0.2219153113836122, + "learning_rate": 4.0438991033881425e-05, + "loss": 0.51, + "step": 1392 + }, + { + "epoch": 2.7367387033398822, + "grad_norm": 0.23664567877342615, + "learning_rate": 4.038411896197264e-05, + "loss": 0.5181, + "step": 1393 + }, + { + "epoch": 2.7387033398821217, + "grad_norm": 0.1970166229055411, + "learning_rate": 4.0329246167136845e-05, + "loss": 0.5183, + "step": 1394 + }, + { + "epoch": 2.7406679764243616, + "grad_norm": 0.2954930557407708, + "learning_rate": 4.027437275264679e-05, + "loss": 0.5274, + "step": 1395 + }, + { + "epoch": 2.7426326129666014, + "grad_norm": 0.3202887517946904, + "learning_rate": 4.02194988217764e-05, + "loss": 0.5222, + "step": 1396 + }, + { + "epoch": 2.744597249508841, + "grad_norm": 0.20207242543470996, + "learning_rate": 4.0164624477800576e-05, + "loss": 0.5096, + "step": 1397 + }, + { + "epoch": 2.7465618860510803, + "grad_norm": 0.2272141686354254, + "learning_rate": 4.010974982399499e-05, + "loss": 0.5298, + "step": 1398 + }, + { + "epoch": 2.74852652259332, + "grad_norm": 0.24159400572653697, + "learning_rate": 4.005487496363592e-05, + "loss": 0.5239, + "step": 1399 + }, + { + "epoch": 2.75049115913556, + "grad_norm": 0.2432954683124188, + "learning_rate": 4e-05, + "loss": 0.5194, + "step": 1400 + }, + { + "epoch": 2.7524557956777995, + "grad_norm": 0.3120857612306095, + "learning_rate": 3.9945125036364086e-05, + "loss": 0.5116, + "step": 1401 + }, + { + "epoch": 2.7544204322200394, + "grad_norm": 0.28429321222188536, + "learning_rate": 3.989025017600502e-05, + "loss": 0.5134, + "step": 1402 + }, + { + "epoch": 2.756385068762279, + "grad_norm": 0.2573662165124547, + "learning_rate": 3.983537552219944e-05, + "loss": 0.5242, + "step": 1403 + }, + { + "epoch": 2.7583497053045187, + "grad_norm": 0.17004366406368399, + "learning_rate": 3.9780501178223605e-05, + "loss": 0.5228, + "step": 1404 + }, + { + "epoch": 2.760314341846758, + "grad_norm": 0.21324350321050267, + "learning_rate": 3.972562724735323e-05, + "loss": 0.5199, + "step": 1405 + }, + { + "epoch": 2.762278978388998, + "grad_norm": 0.2439026432264047, + "learning_rate": 3.967075383286317e-05, + "loss": 0.5169, + "step": 1406 + }, + { + "epoch": 2.764243614931238, + "grad_norm": 0.25586463131101034, + "learning_rate": 3.961588103802737e-05, + "loss": 0.5113, + "step": 1407 + }, + { + "epoch": 2.7662082514734774, + "grad_norm": 0.22177700828152713, + "learning_rate": 3.956100896611859e-05, + "loss": 0.5133, + "step": 1408 + }, + { + "epoch": 2.768172888015717, + "grad_norm": 0.1611747554870743, + "learning_rate": 3.950613772040825e-05, + "loss": 0.519, + "step": 1409 + }, + { + "epoch": 2.7701375245579567, + "grad_norm": 0.2070982616830753, + "learning_rate": 3.9451267404166164e-05, + "loss": 0.5166, + "step": 1410 + }, + { + "epoch": 2.7721021611001966, + "grad_norm": 0.23832322693108965, + "learning_rate": 3.939639812066045e-05, + "loss": 0.5182, + "step": 1411 + }, + { + "epoch": 2.774066797642436, + "grad_norm": 0.21480412776782165, + "learning_rate": 3.9341529973157266e-05, + "loss": 0.5146, + "step": 1412 + }, + { + "epoch": 2.776031434184676, + "grad_norm": 0.20600327939923183, + "learning_rate": 3.928666306492062e-05, + "loss": 0.5244, + "step": 1413 + }, + { + "epoch": 2.777996070726916, + "grad_norm": 0.18911209629758077, + "learning_rate": 3.923179749921219e-05, + "loss": 0.509, + "step": 1414 + }, + { + "epoch": 2.7799607072691552, + "grad_norm": 0.17683157153186396, + "learning_rate": 3.917693337929116e-05, + "loss": 0.525, + "step": 1415 + }, + { + "epoch": 2.7819253438113947, + "grad_norm": 0.21683348540374792, + "learning_rate": 3.9122070808413924e-05, + "loss": 0.513, + "step": 1416 + }, + { + "epoch": 2.7838899803536346, + "grad_norm": 0.23356116597170726, + "learning_rate": 3.906720988983402e-05, + "loss": 0.5261, + "step": 1417 + }, + { + "epoch": 2.7858546168958744, + "grad_norm": 0.2375651797025771, + "learning_rate": 3.9012350726801875e-05, + "loss": 0.5181, + "step": 1418 + }, + { + "epoch": 2.787819253438114, + "grad_norm": 0.2021064250345991, + "learning_rate": 3.895749342256457e-05, + "loss": 0.5181, + "step": 1419 + }, + { + "epoch": 2.7897838899803538, + "grad_norm": 0.23420337620468304, + "learning_rate": 3.8902638080365734e-05, + "loss": 0.5124, + "step": 1420 + }, + { + "epoch": 2.791748526522593, + "grad_norm": 0.2104787905743406, + "learning_rate": 3.884778480344527e-05, + "loss": 0.5248, + "step": 1421 + }, + { + "epoch": 2.793713163064833, + "grad_norm": 0.19898701843470504, + "learning_rate": 3.879293369503922e-05, + "loss": 0.5184, + "step": 1422 + }, + { + "epoch": 2.7956777996070725, + "grad_norm": 0.2624405818158757, + "learning_rate": 3.873808485837951e-05, + "loss": 0.5202, + "step": 1423 + }, + { + "epoch": 2.7976424361493124, + "grad_norm": 0.2273615901000724, + "learning_rate": 3.868323839669383e-05, + "loss": 0.5254, + "step": 1424 + }, + { + "epoch": 2.7996070726915523, + "grad_norm": 0.28660066001579143, + "learning_rate": 3.862839441320539e-05, + "loss": 0.5284, + "step": 1425 + }, + { + "epoch": 2.8015717092337917, + "grad_norm": 0.2846745228421202, + "learning_rate": 3.8573553011132694e-05, + "loss": 0.5282, + "step": 1426 + }, + { + "epoch": 2.803536345776031, + "grad_norm": 0.2654532945635471, + "learning_rate": 3.851871429368945e-05, + "loss": 0.5276, + "step": 1427 + }, + { + "epoch": 2.805500982318271, + "grad_norm": 0.25274409718507423, + "learning_rate": 3.846387836408428e-05, + "loss": 0.5275, + "step": 1428 + }, + { + "epoch": 2.807465618860511, + "grad_norm": 0.24732148316397304, + "learning_rate": 3.8409045325520556e-05, + "loss": 0.5279, + "step": 1429 + }, + { + "epoch": 2.8094302554027504, + "grad_norm": 0.2873711089954797, + "learning_rate": 3.835421528119623e-05, + "loss": 0.5205, + "step": 1430 + }, + { + "epoch": 2.8113948919449903, + "grad_norm": 0.23385455065178443, + "learning_rate": 3.8299388334303604e-05, + "loss": 0.5184, + "step": 1431 + }, + { + "epoch": 2.8133595284872297, + "grad_norm": 0.21630915700028197, + "learning_rate": 3.8244564588029135e-05, + "loss": 0.5174, + "step": 1432 + }, + { + "epoch": 2.8153241650294696, + "grad_norm": 0.26592137474365163, + "learning_rate": 3.818974414555329e-05, + "loss": 0.5209, + "step": 1433 + }, + { + "epoch": 2.817288801571709, + "grad_norm": 0.27759111246016444, + "learning_rate": 3.813492711005033e-05, + "loss": 0.5138, + "step": 1434 + }, + { + "epoch": 2.819253438113949, + "grad_norm": 0.2801015827712734, + "learning_rate": 3.808011358468801e-05, + "loss": 0.5204, + "step": 1435 + }, + { + "epoch": 2.821218074656189, + "grad_norm": 0.2082133019320925, + "learning_rate": 3.802530367262759e-05, + "loss": 0.5193, + "step": 1436 + }, + { + "epoch": 2.8231827111984282, + "grad_norm": 0.23103597453892327, + "learning_rate": 3.797049747702346e-05, + "loss": 0.5163, + "step": 1437 + }, + { + "epoch": 2.825147347740668, + "grad_norm": 0.26385017130088395, + "learning_rate": 3.7915695101023085e-05, + "loss": 0.5127, + "step": 1438 + }, + { + "epoch": 2.8271119842829076, + "grad_norm": 0.23563820270498295, + "learning_rate": 3.786089664776665e-05, + "loss": 0.5174, + "step": 1439 + }, + { + "epoch": 2.8290766208251474, + "grad_norm": 0.23801068266155692, + "learning_rate": 3.780610222038701e-05, + "loss": 0.5144, + "step": 1440 + }, + { + "epoch": 2.831041257367387, + "grad_norm": 0.17831316037636386, + "learning_rate": 3.775131192200947e-05, + "loss": 0.5196, + "step": 1441 + }, + { + "epoch": 2.8330058939096268, + "grad_norm": 0.28275732404313997, + "learning_rate": 3.769652585575148e-05, + "loss": 0.5189, + "step": 1442 + }, + { + "epoch": 2.8349705304518666, + "grad_norm": 0.24643919915902246, + "learning_rate": 3.764174412472261e-05, + "loss": 0.5243, + "step": 1443 + }, + { + "epoch": 2.836935166994106, + "grad_norm": 0.23032376855674405, + "learning_rate": 3.7586966832024247e-05, + "loss": 0.5159, + "step": 1444 + }, + { + "epoch": 2.8388998035363455, + "grad_norm": 0.23921426896453843, + "learning_rate": 3.7532194080749375e-05, + "loss": 0.5226, + "step": 1445 + }, + { + "epoch": 2.8408644400785854, + "grad_norm": 0.22049587102564033, + "learning_rate": 3.74774259739825e-05, + "loss": 0.5157, + "step": 1446 + }, + { + "epoch": 2.8428290766208253, + "grad_norm": 0.25184785314264246, + "learning_rate": 3.742266261479935e-05, + "loss": 0.5169, + "step": 1447 + }, + { + "epoch": 2.8447937131630647, + "grad_norm": 0.19589468287302142, + "learning_rate": 3.736790410626672e-05, + "loss": 0.5136, + "step": 1448 + }, + { + "epoch": 2.8467583497053046, + "grad_norm": 0.2433787447399285, + "learning_rate": 3.731315055144228e-05, + "loss": 0.5156, + "step": 1449 + }, + { + "epoch": 2.848722986247544, + "grad_norm": 0.26576939107396186, + "learning_rate": 3.725840205337438e-05, + "loss": 0.5175, + "step": 1450 + }, + { + "epoch": 2.850687622789784, + "grad_norm": 0.2216562277560749, + "learning_rate": 3.720365871510186e-05, + "loss": 0.5245, + "step": 1451 + }, + { + "epoch": 2.8526522593320234, + "grad_norm": 0.19649556074356492, + "learning_rate": 3.714892063965383e-05, + "loss": 0.5175, + "step": 1452 + }, + { + "epoch": 2.8546168958742633, + "grad_norm": 0.1908846546071574, + "learning_rate": 3.709418793004951e-05, + "loss": 0.522, + "step": 1453 + }, + { + "epoch": 2.856581532416503, + "grad_norm": 0.21003821649540635, + "learning_rate": 3.7039460689298025e-05, + "loss": 0.5212, + "step": 1454 + }, + { + "epoch": 2.8585461689587426, + "grad_norm": 0.22883824457313462, + "learning_rate": 3.698473902039818e-05, + "loss": 0.5176, + "step": 1455 + }, + { + "epoch": 2.8605108055009825, + "grad_norm": 0.23486761845433068, + "learning_rate": 3.69300230263383e-05, + "loss": 0.5195, + "step": 1456 + }, + { + "epoch": 2.862475442043222, + "grad_norm": 0.23749123946965864, + "learning_rate": 3.68753128100961e-05, + "loss": 0.5234, + "step": 1457 + }, + { + "epoch": 2.864440078585462, + "grad_norm": 0.24457627508245938, + "learning_rate": 3.6820608474638307e-05, + "loss": 0.5214, + "step": 1458 + }, + { + "epoch": 2.8664047151277012, + "grad_norm": 0.23632811687073008, + "learning_rate": 3.676591012292065e-05, + "loss": 0.5192, + "step": 1459 + }, + { + "epoch": 2.868369351669941, + "grad_norm": 0.2157825490521872, + "learning_rate": 3.671121785788759e-05, + "loss": 0.5294, + "step": 1460 + }, + { + "epoch": 2.870333988212181, + "grad_norm": 0.249841152946354, + "learning_rate": 3.665653178247212e-05, + "loss": 0.5352, + "step": 1461 + }, + { + "epoch": 2.8722986247544204, + "grad_norm": 0.29431335336046893, + "learning_rate": 3.660185199959559e-05, + "loss": 0.5178, + "step": 1462 + }, + { + "epoch": 2.87426326129666, + "grad_norm": 0.28255276365754717, + "learning_rate": 3.6547178612167504e-05, + "loss": 0.5234, + "step": 1463 + }, + { + "epoch": 2.8762278978388998, + "grad_norm": 0.2509703939617441, + "learning_rate": 3.6492511723085356e-05, + "loss": 0.5189, + "step": 1464 + }, + { + "epoch": 2.8781925343811396, + "grad_norm": 0.22414281597914645, + "learning_rate": 3.6437851435234346e-05, + "loss": 0.5222, + "step": 1465 + }, + { + "epoch": 2.880157170923379, + "grad_norm": 0.24908449376481143, + "learning_rate": 3.6383197851487296e-05, + "loss": 0.5189, + "step": 1466 + }, + { + "epoch": 2.882121807465619, + "grad_norm": 0.26760074264902806, + "learning_rate": 3.6328551074704446e-05, + "loss": 0.5285, + "step": 1467 + }, + { + "epoch": 2.8840864440078584, + "grad_norm": 0.24386259678670572, + "learning_rate": 3.627391120773313e-05, + "loss": 0.5124, + "step": 1468 + }, + { + "epoch": 2.8860510805500983, + "grad_norm": 0.22130855111208383, + "learning_rate": 3.621927835340776e-05, + "loss": 0.5183, + "step": 1469 + }, + { + "epoch": 2.8880157170923377, + "grad_norm": 0.34088834238528726, + "learning_rate": 3.616465261454953e-05, + "loss": 0.5275, + "step": 1470 + }, + { + "epoch": 2.8899803536345776, + "grad_norm": 0.3222790158532804, + "learning_rate": 3.6110034093966194e-05, + "loss": 0.5198, + "step": 1471 + }, + { + "epoch": 2.8919449901768175, + "grad_norm": 0.17932062890649597, + "learning_rate": 3.6055422894451994e-05, + "loss": 0.5151, + "step": 1472 + }, + { + "epoch": 2.893909626719057, + "grad_norm": 0.2586498685468305, + "learning_rate": 3.600081911878736e-05, + "loss": 0.5208, + "step": 1473 + }, + { + "epoch": 2.895874263261297, + "grad_norm": 0.20751894241911067, + "learning_rate": 3.5946222869738734e-05, + "loss": 0.5233, + "step": 1474 + }, + { + "epoch": 2.8978388998035363, + "grad_norm": 0.19944126658683192, + "learning_rate": 3.58916342500584e-05, + "loss": 0.5206, + "step": 1475 + }, + { + "epoch": 2.899803536345776, + "grad_norm": 0.22604072197788827, + "learning_rate": 3.58370533624843e-05, + "loss": 0.5142, + "step": 1476 + }, + { + "epoch": 2.9017681728880156, + "grad_norm": 0.2044814149017366, + "learning_rate": 3.5782480309739854e-05, + "loss": 0.5151, + "step": 1477 + }, + { + "epoch": 2.9037328094302555, + "grad_norm": 0.190794936925196, + "learning_rate": 3.572791519453364e-05, + "loss": 0.527, + "step": 1478 + }, + { + "epoch": 2.9056974459724954, + "grad_norm": 0.192761987501033, + "learning_rate": 3.567335811955938e-05, + "loss": 0.519, + "step": 1479 + }, + { + "epoch": 2.907662082514735, + "grad_norm": 0.18318605711067887, + "learning_rate": 3.561880918749565e-05, + "loss": 0.519, + "step": 1480 + }, + { + "epoch": 2.9096267190569742, + "grad_norm": 0.24224492320429705, + "learning_rate": 3.556426850100567e-05, + "loss": 0.5148, + "step": 1481 + }, + { + "epoch": 2.911591355599214, + "grad_norm": 0.2200657308261038, + "learning_rate": 3.550973616273716e-05, + "loss": 0.5131, + "step": 1482 + }, + { + "epoch": 2.913555992141454, + "grad_norm": 0.1881364155770987, + "learning_rate": 3.545521227532216e-05, + "loss": 0.5206, + "step": 1483 + }, + { + "epoch": 2.9155206286836934, + "grad_norm": 0.2441865019107066, + "learning_rate": 3.540069694137674e-05, + "loss": 0.5314, + "step": 1484 + }, + { + "epoch": 2.9174852652259333, + "grad_norm": 0.24833051847600104, + "learning_rate": 3.534619026350091e-05, + "loss": 0.5169, + "step": 1485 + }, + { + "epoch": 2.9194499017681728, + "grad_norm": 0.2374538830187009, + "learning_rate": 3.529169234427842e-05, + "loss": 0.5222, + "step": 1486 + }, + { + "epoch": 2.9214145383104126, + "grad_norm": 0.17969794896655836, + "learning_rate": 3.5237203286276465e-05, + "loss": 0.5179, + "step": 1487 + }, + { + "epoch": 2.923379174852652, + "grad_norm": 0.2123279445851223, + "learning_rate": 3.5182723192045606e-05, + "loss": 0.5209, + "step": 1488 + }, + { + "epoch": 2.925343811394892, + "grad_norm": 0.18283452235591588, + "learning_rate": 3.512825216411954e-05, + "loss": 0.5321, + "step": 1489 + }, + { + "epoch": 2.927308447937132, + "grad_norm": 0.19126558497545132, + "learning_rate": 3.507379030501488e-05, + "loss": 0.5206, + "step": 1490 + }, + { + "epoch": 2.9292730844793713, + "grad_norm": 0.1904634966856656, + "learning_rate": 3.501933771723099e-05, + "loss": 0.5186, + "step": 1491 + }, + { + "epoch": 2.9312377210216107, + "grad_norm": 0.18632577679747092, + "learning_rate": 3.496489450324978e-05, + "loss": 0.5155, + "step": 1492 + }, + { + "epoch": 2.9332023575638506, + "grad_norm": 0.2133217624965516, + "learning_rate": 3.491046076553556e-05, + "loss": 0.5217, + "step": 1493 + }, + { + "epoch": 2.9351669941060905, + "grad_norm": 0.21219511132293523, + "learning_rate": 3.485603660653471e-05, + "loss": 0.5291, + "step": 1494 + }, + { + "epoch": 2.93713163064833, + "grad_norm": 0.1914771742938305, + "learning_rate": 3.4801622128675685e-05, + "loss": 0.5182, + "step": 1495 + }, + { + "epoch": 2.93909626719057, + "grad_norm": 0.2121466215617487, + "learning_rate": 3.474721743436868e-05, + "loss": 0.5068, + "step": 1496 + }, + { + "epoch": 2.9410609037328097, + "grad_norm": 0.2313152484177056, + "learning_rate": 3.469282262600544e-05, + "loss": 0.5168, + "step": 1497 + }, + { + "epoch": 2.943025540275049, + "grad_norm": 0.20117612352581646, + "learning_rate": 3.463843780595916e-05, + "loss": 0.5171, + "step": 1498 + }, + { + "epoch": 2.9449901768172886, + "grad_norm": 0.16943662460161132, + "learning_rate": 3.458406307658422e-05, + "loss": 0.5269, + "step": 1499 + }, + { + "epoch": 2.9469548133595285, + "grad_norm": 0.2379492375463512, + "learning_rate": 3.452969854021599e-05, + "loss": 0.5283, + "step": 1500 + }, + { + "epoch": 2.9489194499017684, + "grad_norm": 0.17409239259164497, + "learning_rate": 3.447534429917067e-05, + "loss": 0.5189, + "step": 1501 + }, + { + "epoch": 2.950884086444008, + "grad_norm": 0.1977178323164717, + "learning_rate": 3.4421000455745096e-05, + "loss": 0.5172, + "step": 1502 + }, + { + "epoch": 2.9528487229862477, + "grad_norm": 0.16115275130618845, + "learning_rate": 3.436666711221651e-05, + "loss": 0.515, + "step": 1503 + }, + { + "epoch": 2.954813359528487, + "grad_norm": 0.16426027505232277, + "learning_rate": 3.4312344370842395e-05, + "loss": 0.5211, + "step": 1504 + }, + { + "epoch": 2.956777996070727, + "grad_norm": 0.17474914640105105, + "learning_rate": 3.425803233386032e-05, + "loss": 0.5179, + "step": 1505 + }, + { + "epoch": 2.9587426326129664, + "grad_norm": 0.22733372470002308, + "learning_rate": 3.4203731103487684e-05, + "loss": 0.5109, + "step": 1506 + }, + { + "epoch": 2.9607072691552063, + "grad_norm": 0.22745117201883108, + "learning_rate": 3.41494407819215e-05, + "loss": 0.5112, + "step": 1507 + }, + { + "epoch": 2.962671905697446, + "grad_norm": 0.16116494531331244, + "learning_rate": 3.409516147133832e-05, + "loss": 0.5228, + "step": 1508 + }, + { + "epoch": 2.9646365422396856, + "grad_norm": 0.23871081463565927, + "learning_rate": 3.404089327389395e-05, + "loss": 0.512, + "step": 1509 + }, + { + "epoch": 2.966601178781925, + "grad_norm": 0.22566446655237118, + "learning_rate": 3.398663629172326e-05, + "loss": 0.5196, + "step": 1510 + }, + { + "epoch": 2.968565815324165, + "grad_norm": 0.18463018690213173, + "learning_rate": 3.393239062694003e-05, + "loss": 0.5188, + "step": 1511 + }, + { + "epoch": 2.970530451866405, + "grad_norm": 0.2337480082927675, + "learning_rate": 3.387815638163677e-05, + "loss": 0.5198, + "step": 1512 + }, + { + "epoch": 2.9724950884086443, + "grad_norm": 0.18381664821825586, + "learning_rate": 3.382393365788441e-05, + "loss": 0.5222, + "step": 1513 + }, + { + "epoch": 2.974459724950884, + "grad_norm": 0.2088446868103254, + "learning_rate": 3.3769722557732296e-05, + "loss": 0.5204, + "step": 1514 + }, + { + "epoch": 2.9764243614931236, + "grad_norm": 0.23402265626763996, + "learning_rate": 3.371552318320786e-05, + "loss": 0.5241, + "step": 1515 + }, + { + "epoch": 2.9783889980353635, + "grad_norm": 0.20068839113197512, + "learning_rate": 3.366133563631644e-05, + "loss": 0.5225, + "step": 1516 + }, + { + "epoch": 2.980353634577603, + "grad_norm": 0.21357081319474794, + "learning_rate": 3.360716001904112e-05, + "loss": 0.5177, + "step": 1517 + }, + { + "epoch": 2.982318271119843, + "grad_norm": 0.26226249996175033, + "learning_rate": 3.3552996433342575e-05, + "loss": 0.5234, + "step": 1518 + }, + { + "epoch": 2.9842829076620827, + "grad_norm": 0.20940885280361035, + "learning_rate": 3.3498844981158804e-05, + "loss": 0.5173, + "step": 1519 + }, + { + "epoch": 2.986247544204322, + "grad_norm": 0.18849012465313067, + "learning_rate": 3.344470576440496e-05, + "loss": 0.5186, + "step": 1520 + }, + { + "epoch": 2.988212180746562, + "grad_norm": 0.22430140669469553, + "learning_rate": 3.339057888497319e-05, + "loss": 0.5182, + "step": 1521 + }, + { + "epoch": 2.9901768172888015, + "grad_norm": 0.20105313528112984, + "learning_rate": 3.3336464444732436e-05, + "loss": 0.5096, + "step": 1522 + }, + { + "epoch": 2.9921414538310414, + "grad_norm": 0.23146074274467593, + "learning_rate": 3.3282362545528156e-05, + "loss": 0.5159, + "step": 1523 + }, + { + "epoch": 2.994106090373281, + "grad_norm": 0.27457049379530746, + "learning_rate": 3.3228273289182294e-05, + "loss": 0.5202, + "step": 1524 + }, + { + "epoch": 2.9960707269155207, + "grad_norm": 0.22953100803826634, + "learning_rate": 3.317419677749297e-05, + "loss": 0.519, + "step": 1525 + }, + { + "epoch": 2.9980353634577606, + "grad_norm": 0.2654670692810497, + "learning_rate": 3.312013311223428e-05, + "loss": 0.5217, + "step": 1526 + }, + { + "epoch": 3.0, + "grad_norm": 0.31704720328317004, + "learning_rate": 3.306608239515619e-05, + "loss": 0.5289, + "step": 1527 + }, + { + "epoch": 3.00196463654224, + "grad_norm": 0.28635856662999104, + "learning_rate": 3.301204472798428e-05, + "loss": 0.4695, + "step": 1528 + }, + { + "epoch": 3.0039292730844793, + "grad_norm": 0.26791401889664496, + "learning_rate": 3.295802021241955e-05, + "loss": 0.4772, + "step": 1529 + }, + { + "epoch": 3.005893909626719, + "grad_norm": 0.22961080700496303, + "learning_rate": 3.2904008950138294e-05, + "loss": 0.4772, + "step": 1530 + }, + { + "epoch": 3.0078585461689586, + "grad_norm": 0.2904007784774515, + "learning_rate": 3.285001104279183e-05, + "loss": 0.465, + "step": 1531 + }, + { + "epoch": 3.0098231827111985, + "grad_norm": 0.2364797020619144, + "learning_rate": 3.279602659200632e-05, + "loss": 0.4723, + "step": 1532 + }, + { + "epoch": 3.011787819253438, + "grad_norm": 0.23625264153835762, + "learning_rate": 3.274205569938266e-05, + "loss": 0.474, + "step": 1533 + }, + { + "epoch": 3.013752455795678, + "grad_norm": 0.24273702634013566, + "learning_rate": 3.268809846649619e-05, + "loss": 0.4731, + "step": 1534 + }, + { + "epoch": 3.0157170923379173, + "grad_norm": 0.2421780095955358, + "learning_rate": 3.2634154994896556e-05, + "loss": 0.4793, + "step": 1535 + }, + { + "epoch": 3.017681728880157, + "grad_norm": 0.2309654070890268, + "learning_rate": 3.258022538610747e-05, + "loss": 0.4731, + "step": 1536 + }, + { + "epoch": 3.019646365422397, + "grad_norm": 0.24378169763541285, + "learning_rate": 3.2526309741626594e-05, + "loss": 0.4714, + "step": 1537 + }, + { + "epoch": 3.0216110019646365, + "grad_norm": 0.2209607957475229, + "learning_rate": 3.247240816292531e-05, + "loss": 0.4719, + "step": 1538 + }, + { + "epoch": 3.0235756385068764, + "grad_norm": 0.23277853487474376, + "learning_rate": 3.241852075144849e-05, + "loss": 0.4675, + "step": 1539 + }, + { + "epoch": 3.025540275049116, + "grad_norm": 0.22823586870669224, + "learning_rate": 3.2364647608614376e-05, + "loss": 0.4735, + "step": 1540 + }, + { + "epoch": 3.0275049115913557, + "grad_norm": 0.22440641444849332, + "learning_rate": 3.231078883581437e-05, + "loss": 0.4694, + "step": 1541 + }, + { + "epoch": 3.029469548133595, + "grad_norm": 0.2323499138540528, + "learning_rate": 3.225694453441276e-05, + "loss": 0.471, + "step": 1542 + }, + { + "epoch": 3.031434184675835, + "grad_norm": 0.1918398801873856, + "learning_rate": 3.2203114805746666e-05, + "loss": 0.4749, + "step": 1543 + }, + { + "epoch": 3.0333988212180745, + "grad_norm": 0.23277266617971765, + "learning_rate": 3.2149299751125776e-05, + "loss": 0.4727, + "step": 1544 + }, + { + "epoch": 3.0353634577603144, + "grad_norm": 0.1548374749537498, + "learning_rate": 3.209549947183211e-05, + "loss": 0.4753, + "step": 1545 + }, + { + "epoch": 3.0373280943025542, + "grad_norm": 0.24600433104488023, + "learning_rate": 3.204171406911993e-05, + "loss": 0.4781, + "step": 1546 + }, + { + "epoch": 3.0392927308447937, + "grad_norm": 0.19552511871090675, + "learning_rate": 3.1987943644215486e-05, + "loss": 0.4762, + "step": 1547 + }, + { + "epoch": 3.0412573673870336, + "grad_norm": 0.21888689282370596, + "learning_rate": 3.1934188298316836e-05, + "loss": 0.4771, + "step": 1548 + }, + { + "epoch": 3.043222003929273, + "grad_norm": 0.22552833107845682, + "learning_rate": 3.188044813259365e-05, + "loss": 0.4644, + "step": 1549 + }, + { + "epoch": 3.045186640471513, + "grad_norm": 0.18692831456433406, + "learning_rate": 3.182672324818704e-05, + "loss": 0.47, + "step": 1550 + }, + { + "epoch": 3.0471512770137523, + "grad_norm": 0.2790069486501621, + "learning_rate": 3.177301374620937e-05, + "loss": 0.4664, + "step": 1551 + }, + { + "epoch": 3.049115913555992, + "grad_norm": 0.18747253827882637, + "learning_rate": 3.171931972774402e-05, + "loss": 0.463, + "step": 1552 + }, + { + "epoch": 3.0510805500982316, + "grad_norm": 0.22204325232253408, + "learning_rate": 3.166564129384526e-05, + "loss": 0.4699, + "step": 1553 + }, + { + "epoch": 3.0530451866404715, + "grad_norm": 0.2450285221913732, + "learning_rate": 3.161197854553803e-05, + "loss": 0.4737, + "step": 1554 + }, + { + "epoch": 3.0550098231827114, + "grad_norm": 0.1896272511166455, + "learning_rate": 3.155833158381771e-05, + "loss": 0.4738, + "step": 1555 + }, + { + "epoch": 3.056974459724951, + "grad_norm": 0.21127372580840453, + "learning_rate": 3.1504700509650004e-05, + "loss": 0.4754, + "step": 1556 + }, + { + "epoch": 3.0589390962671907, + "grad_norm": 0.24775898843725852, + "learning_rate": 3.145108542397072e-05, + "loss": 0.472, + "step": 1557 + }, + { + "epoch": 3.06090373280943, + "grad_norm": 0.22904801918141096, + "learning_rate": 3.139748642768554e-05, + "loss": 0.4719, + "step": 1558 + }, + { + "epoch": 3.06286836935167, + "grad_norm": 0.22603217081361407, + "learning_rate": 3.134390362166991e-05, + "loss": 0.4743, + "step": 1559 + }, + { + "epoch": 3.0648330058939095, + "grad_norm": 0.21347174288573212, + "learning_rate": 3.129033710676877e-05, + "loss": 0.47, + "step": 1560 + }, + { + "epoch": 3.0667976424361494, + "grad_norm": 0.2025037617818563, + "learning_rate": 3.1236786983796426e-05, + "loss": 0.4686, + "step": 1561 + }, + { + "epoch": 3.068762278978389, + "grad_norm": 0.19447415096782808, + "learning_rate": 3.1183253353536294e-05, + "loss": 0.4724, + "step": 1562 + }, + { + "epoch": 3.0707269155206287, + "grad_norm": 0.19771742873132833, + "learning_rate": 3.1129736316740794e-05, + "loss": 0.4684, + "step": 1563 + }, + { + "epoch": 3.072691552062868, + "grad_norm": 0.2047493666877048, + "learning_rate": 3.107623597413111e-05, + "loss": 0.4708, + "step": 1564 + }, + { + "epoch": 3.074656188605108, + "grad_norm": 0.17378682944056167, + "learning_rate": 3.102275242639696e-05, + "loss": 0.4693, + "step": 1565 + }, + { + "epoch": 3.076620825147348, + "grad_norm": 0.1838771332715366, + "learning_rate": 3.096928577419651e-05, + "loss": 0.4711, + "step": 1566 + }, + { + "epoch": 3.0785854616895874, + "grad_norm": 0.19472505014418884, + "learning_rate": 3.091583611815612e-05, + "loss": 0.4687, + "step": 1567 + }, + { + "epoch": 3.0805500982318272, + "grad_norm": 0.2230701885682001, + "learning_rate": 3.0862403558870115e-05, + "loss": 0.4839, + "step": 1568 + }, + { + "epoch": 3.0825147347740667, + "grad_norm": 0.21001435093138307, + "learning_rate": 3.08089881969007e-05, + "loss": 0.4695, + "step": 1569 + }, + { + "epoch": 3.0844793713163066, + "grad_norm": 0.21925721192944642, + "learning_rate": 3.075559013277769e-05, + "loss": 0.4656, + "step": 1570 + }, + { + "epoch": 3.086444007858546, + "grad_norm": 0.2557774677139696, + "learning_rate": 3.0702209466998344e-05, + "loss": 0.4773, + "step": 1571 + }, + { + "epoch": 3.088408644400786, + "grad_norm": 0.3365371230839213, + "learning_rate": 3.064884630002718e-05, + "loss": 0.4743, + "step": 1572 + }, + { + "epoch": 3.0903732809430258, + "grad_norm": 0.2433990371310504, + "learning_rate": 3.0595500732295764e-05, + "loss": 0.4706, + "step": 1573 + }, + { + "epoch": 3.092337917485265, + "grad_norm": 0.2960835443019684, + "learning_rate": 3.05421728642026e-05, + "loss": 0.478, + "step": 1574 + }, + { + "epoch": 3.094302554027505, + "grad_norm": 0.31107399716937056, + "learning_rate": 3.0488862796112784e-05, + "loss": 0.4665, + "step": 1575 + }, + { + "epoch": 3.0962671905697445, + "grad_norm": 0.2792422710891195, + "learning_rate": 3.043557062835799e-05, + "loss": 0.4792, + "step": 1576 + }, + { + "epoch": 3.0982318271119844, + "grad_norm": 0.26126740735801995, + "learning_rate": 3.0382296461236187e-05, + "loss": 0.4729, + "step": 1577 + }, + { + "epoch": 3.100196463654224, + "grad_norm": 0.17793606551133861, + "learning_rate": 3.0329040395011436e-05, + "loss": 0.4633, + "step": 1578 + }, + { + "epoch": 3.1021611001964637, + "grad_norm": 0.21858033452585177, + "learning_rate": 3.027580252991377e-05, + "loss": 0.4679, + "step": 1579 + }, + { + "epoch": 3.104125736738703, + "grad_norm": 0.1749479697791413, + "learning_rate": 3.0222582966138944e-05, + "loss": 0.4712, + "step": 1580 + }, + { + "epoch": 3.106090373280943, + "grad_norm": 0.20728447665467475, + "learning_rate": 3.0169381803848264e-05, + "loss": 0.4826, + "step": 1581 + }, + { + "epoch": 3.1080550098231825, + "grad_norm": 0.2298556118465606, + "learning_rate": 3.0116199143168425e-05, + "loss": 0.48, + "step": 1582 + }, + { + "epoch": 3.1100196463654224, + "grad_norm": 0.17860394281329198, + "learning_rate": 3.00630350841913e-05, + "loss": 0.4702, + "step": 1583 + }, + { + "epoch": 3.1119842829076623, + "grad_norm": 0.3019791705611821, + "learning_rate": 3.0009889726973723e-05, + "loss": 0.4809, + "step": 1584 + }, + { + "epoch": 3.1139489194499017, + "grad_norm": 0.28414240989244527, + "learning_rate": 2.995676317153735e-05, + "loss": 0.47, + "step": 1585 + }, + { + "epoch": 3.1159135559921416, + "grad_norm": 0.2114205979243882, + "learning_rate": 2.990365551786847e-05, + "loss": 0.4669, + "step": 1586 + }, + { + "epoch": 3.117878192534381, + "grad_norm": 0.3120025735765655, + "learning_rate": 2.985056686591776e-05, + "loss": 0.4806, + "step": 1587 + }, + { + "epoch": 3.119842829076621, + "grad_norm": 0.28615850835930695, + "learning_rate": 2.979749731560017e-05, + "loss": 0.4795, + "step": 1588 + }, + { + "epoch": 3.1218074656188604, + "grad_norm": 0.25863490273718953, + "learning_rate": 2.9744446966794673e-05, + "loss": 0.4679, + "step": 1589 + }, + { + "epoch": 3.1237721021611002, + "grad_norm": 0.1932097700281833, + "learning_rate": 2.969141591934414e-05, + "loss": 0.4823, + "step": 1590 + }, + { + "epoch": 3.1257367387033397, + "grad_norm": 0.25794864973985515, + "learning_rate": 2.9638404273055062e-05, + "loss": 0.478, + "step": 1591 + }, + { + "epoch": 3.1277013752455796, + "grad_norm": 0.23838571994699298, + "learning_rate": 2.9585412127697456e-05, + "loss": 0.4663, + "step": 1592 + }, + { + "epoch": 3.1296660117878194, + "grad_norm": 0.1587670544293921, + "learning_rate": 2.953243958300465e-05, + "loss": 0.4677, + "step": 1593 + }, + { + "epoch": 3.131630648330059, + "grad_norm": 0.21160995063649057, + "learning_rate": 2.9479486738673028e-05, + "loss": 0.4655, + "step": 1594 + }, + { + "epoch": 3.1335952848722988, + "grad_norm": 0.1886506707354373, + "learning_rate": 2.9426553694361942e-05, + "loss": 0.4777, + "step": 1595 + }, + { + "epoch": 3.135559921414538, + "grad_norm": 0.17847697687227665, + "learning_rate": 2.9373640549693475e-05, + "loss": 0.4866, + "step": 1596 + }, + { + "epoch": 3.137524557956778, + "grad_norm": 0.1962441598674679, + "learning_rate": 2.9320747404252238e-05, + "loss": 0.4735, + "step": 1597 + }, + { + "epoch": 3.1394891944990175, + "grad_norm": 0.19551183013213663, + "learning_rate": 2.9267874357585218e-05, + "loss": 0.4805, + "step": 1598 + }, + { + "epoch": 3.1414538310412574, + "grad_norm": 0.17457646334594035, + "learning_rate": 2.921502150920159e-05, + "loss": 0.4656, + "step": 1599 + }, + { + "epoch": 3.143418467583497, + "grad_norm": 0.17422709785612414, + "learning_rate": 2.916218895857247e-05, + "loss": 0.4753, + "step": 1600 + }, + { + "epoch": 3.1453831041257367, + "grad_norm": 0.16732983281447208, + "learning_rate": 2.9109376805130817e-05, + "loss": 0.4754, + "step": 1601 + }, + { + "epoch": 3.1473477406679766, + "grad_norm": 0.18581952769663793, + "learning_rate": 2.905658514827118e-05, + "loss": 0.4745, + "step": 1602 + }, + { + "epoch": 3.149312377210216, + "grad_norm": 0.1525016365385959, + "learning_rate": 2.9003814087349558e-05, + "loss": 0.4833, + "step": 1603 + }, + { + "epoch": 3.151277013752456, + "grad_norm": 0.16936168633176155, + "learning_rate": 2.8951063721683127e-05, + "loss": 0.4742, + "step": 1604 + }, + { + "epoch": 3.1532416502946954, + "grad_norm": 0.1757719852236693, + "learning_rate": 2.8898334150550176e-05, + "loss": 0.4791, + "step": 1605 + }, + { + "epoch": 3.1552062868369353, + "grad_norm": 0.13410947634044992, + "learning_rate": 2.8845625473189845e-05, + "loss": 0.4786, + "step": 1606 + }, + { + "epoch": 3.1571709233791747, + "grad_norm": 0.19149882167671878, + "learning_rate": 2.8792937788801924e-05, + "loss": 0.4698, + "step": 1607 + }, + { + "epoch": 3.1591355599214146, + "grad_norm": 0.16157728893134782, + "learning_rate": 2.874027119654671e-05, + "loss": 0.4723, + "step": 1608 + }, + { + "epoch": 3.161100196463654, + "grad_norm": 0.17149743616184762, + "learning_rate": 2.868762579554483e-05, + "loss": 0.4809, + "step": 1609 + }, + { + "epoch": 3.163064833005894, + "grad_norm": 0.13686494698781956, + "learning_rate": 2.863500168487698e-05, + "loss": 0.4716, + "step": 1610 + }, + { + "epoch": 3.165029469548134, + "grad_norm": 0.1630757263409975, + "learning_rate": 2.8582398963583817e-05, + "loss": 0.4653, + "step": 1611 + }, + { + "epoch": 3.1669941060903732, + "grad_norm": 0.1564503851310517, + "learning_rate": 2.852981773066575e-05, + "loss": 0.4663, + "step": 1612 + }, + { + "epoch": 3.168958742632613, + "grad_norm": 0.15492167472737153, + "learning_rate": 2.84772580850827e-05, + "loss": 0.4752, + "step": 1613 + }, + { + "epoch": 3.1709233791748526, + "grad_norm": 0.15862518073075557, + "learning_rate": 2.8424720125754012e-05, + "loss": 0.4709, + "step": 1614 + }, + { + "epoch": 3.1728880157170924, + "grad_norm": 0.15838333572940616, + "learning_rate": 2.837220395155819e-05, + "loss": 0.4731, + "step": 1615 + }, + { + "epoch": 3.174852652259332, + "grad_norm": 0.18937638049515546, + "learning_rate": 2.831970966133275e-05, + "loss": 0.4769, + "step": 1616 + }, + { + "epoch": 3.1768172888015718, + "grad_norm": 0.2231690135576255, + "learning_rate": 2.8267237353874005e-05, + "loss": 0.4745, + "step": 1617 + }, + { + "epoch": 3.178781925343811, + "grad_norm": 0.17017992530374745, + "learning_rate": 2.8214787127936907e-05, + "loss": 0.4749, + "step": 1618 + }, + { + "epoch": 3.180746561886051, + "grad_norm": 0.18537691104335788, + "learning_rate": 2.8162359082234866e-05, + "loss": 0.4746, + "step": 1619 + }, + { + "epoch": 3.182711198428291, + "grad_norm": 0.16457705575696352, + "learning_rate": 2.8109953315439503e-05, + "loss": 0.4724, + "step": 1620 + }, + { + "epoch": 3.1846758349705304, + "grad_norm": 0.15802375066926333, + "learning_rate": 2.8057569926180552e-05, + "loss": 0.4667, + "step": 1621 + }, + { + "epoch": 3.1866404715127703, + "grad_norm": 0.17257837983847132, + "learning_rate": 2.8005209013045624e-05, + "loss": 0.4715, + "step": 1622 + }, + { + "epoch": 3.1886051080550097, + "grad_norm": 0.16931157018297396, + "learning_rate": 2.795287067458e-05, + "loss": 0.472, + "step": 1623 + }, + { + "epoch": 3.1905697445972496, + "grad_norm": 0.1555154471128195, + "learning_rate": 2.7900555009286505e-05, + "loss": 0.4746, + "step": 1624 + }, + { + "epoch": 3.192534381139489, + "grad_norm": 0.17181647959868182, + "learning_rate": 2.7848262115625285e-05, + "loss": 0.4693, + "step": 1625 + }, + { + "epoch": 3.194499017681729, + "grad_norm": 0.20261744327769668, + "learning_rate": 2.779599209201362e-05, + "loss": 0.471, + "step": 1626 + }, + { + "epoch": 3.1964636542239684, + "grad_norm": 0.17325565574256288, + "learning_rate": 2.7743745036825757e-05, + "loss": 0.4677, + "step": 1627 + }, + { + "epoch": 3.1984282907662083, + "grad_norm": 0.21775857152840622, + "learning_rate": 2.769152104839272e-05, + "loss": 0.4751, + "step": 1628 + }, + { + "epoch": 3.2003929273084477, + "grad_norm": 0.26417083404987984, + "learning_rate": 2.7639320225002108e-05, + "loss": 0.4824, + "step": 1629 + }, + { + "epoch": 3.2023575638506876, + "grad_norm": 0.15439282905298962, + "learning_rate": 2.7587142664897926e-05, + "loss": 0.4714, + "step": 1630 + }, + { + "epoch": 3.2043222003929275, + "grad_norm": 0.18994973183448088, + "learning_rate": 2.7534988466280396e-05, + "loss": 0.4706, + "step": 1631 + }, + { + "epoch": 3.206286836935167, + "grad_norm": 0.17751103739786944, + "learning_rate": 2.7482857727305795e-05, + "loss": 0.4762, + "step": 1632 + }, + { + "epoch": 3.208251473477407, + "grad_norm": 0.20178928618046746, + "learning_rate": 2.7430750546086195e-05, + "loss": 0.474, + "step": 1633 + }, + { + "epoch": 3.2102161100196462, + "grad_norm": 0.22368481709506471, + "learning_rate": 2.737866702068938e-05, + "loss": 0.4707, + "step": 1634 + }, + { + "epoch": 3.212180746561886, + "grad_norm": 0.1866997758878776, + "learning_rate": 2.7326607249138607e-05, + "loss": 0.4877, + "step": 1635 + }, + { + "epoch": 3.2141453831041256, + "grad_norm": 0.28773816559033816, + "learning_rate": 2.7274571329412403e-05, + "loss": 0.4753, + "step": 1636 + }, + { + "epoch": 3.2161100196463654, + "grad_norm": 0.2168443402004325, + "learning_rate": 2.7222559359444422e-05, + "loss": 0.4724, + "step": 1637 + }, + { + "epoch": 3.2180746561886053, + "grad_norm": 0.1674494582877347, + "learning_rate": 2.7170571437123253e-05, + "loss": 0.4723, + "step": 1638 + }, + { + "epoch": 3.2200392927308448, + "grad_norm": 0.29994492442155196, + "learning_rate": 2.7118607660292205e-05, + "loss": 0.468, + "step": 1639 + }, + { + "epoch": 3.2220039292730847, + "grad_norm": 0.2705915603113174, + "learning_rate": 2.7066668126749145e-05, + "loss": 0.4766, + "step": 1640 + }, + { + "epoch": 3.223968565815324, + "grad_norm": 0.21408033713837635, + "learning_rate": 2.7014752934246353e-05, + "loss": 0.4768, + "step": 1641 + }, + { + "epoch": 3.225933202357564, + "grad_norm": 0.2840501013018805, + "learning_rate": 2.696286218049023e-05, + "loss": 0.4816, + "step": 1642 + }, + { + "epoch": 3.2278978388998034, + "grad_norm": 0.2152396307203487, + "learning_rate": 2.6910995963141235e-05, + "loss": 0.4739, + "step": 1643 + }, + { + "epoch": 3.2298624754420433, + "grad_norm": 0.3385107521282471, + "learning_rate": 2.6859154379813622e-05, + "loss": 0.4836, + "step": 1644 + }, + { + "epoch": 3.2318271119842827, + "grad_norm": 0.2932151736709931, + "learning_rate": 2.680733752807533e-05, + "loss": 0.4703, + "step": 1645 + }, + { + "epoch": 3.2337917485265226, + "grad_norm": 0.18284069289316868, + "learning_rate": 2.6755545505447678e-05, + "loss": 0.4741, + "step": 1646 + }, + { + "epoch": 3.235756385068762, + "grad_norm": 0.2534661219518587, + "learning_rate": 2.67037784094053e-05, + "loss": 0.4674, + "step": 1647 + }, + { + "epoch": 3.237721021611002, + "grad_norm": 0.2151311459161971, + "learning_rate": 2.6652036337375933e-05, + "loss": 0.4826, + "step": 1648 + }, + { + "epoch": 3.239685658153242, + "grad_norm": 0.21273187795629925, + "learning_rate": 2.660031938674017e-05, + "loss": 0.4797, + "step": 1649 + }, + { + "epoch": 3.2416502946954813, + "grad_norm": 0.2334799948928597, + "learning_rate": 2.6548627654831355e-05, + "loss": 0.475, + "step": 1650 + }, + { + "epoch": 3.243614931237721, + "grad_norm": 0.17987751343649006, + "learning_rate": 2.6496961238935388e-05, + "loss": 0.4714, + "step": 1651 + }, + { + "epoch": 3.2455795677799606, + "grad_norm": 0.2411757813904564, + "learning_rate": 2.6445320236290463e-05, + "loss": 0.4702, + "step": 1652 + }, + { + "epoch": 3.2475442043222005, + "grad_norm": 0.1861035833500122, + "learning_rate": 2.6393704744086998e-05, + "loss": 0.4731, + "step": 1653 + }, + { + "epoch": 3.24950884086444, + "grad_norm": 0.1546145540106127, + "learning_rate": 2.6342114859467396e-05, + "loss": 0.4697, + "step": 1654 + }, + { + "epoch": 3.25147347740668, + "grad_norm": 0.22645606100245402, + "learning_rate": 2.6290550679525834e-05, + "loss": 0.4776, + "step": 1655 + }, + { + "epoch": 3.2534381139489197, + "grad_norm": 0.14345925891038544, + "learning_rate": 2.6239012301308142e-05, + "loss": 0.4672, + "step": 1656 + }, + { + "epoch": 3.255402750491159, + "grad_norm": 0.2632818228233352, + "learning_rate": 2.618749982181157e-05, + "loss": 0.4761, + "step": 1657 + }, + { + "epoch": 3.257367387033399, + "grad_norm": 0.24545595826503033, + "learning_rate": 2.6136013337984646e-05, + "loss": 0.4728, + "step": 1658 + }, + { + "epoch": 3.2593320235756384, + "grad_norm": 0.17668222220523067, + "learning_rate": 2.6084552946726956e-05, + "loss": 0.4765, + "step": 1659 + }, + { + "epoch": 3.2612966601178783, + "grad_norm": 0.2589580026890738, + "learning_rate": 2.6033118744888972e-05, + "loss": 0.4752, + "step": 1660 + }, + { + "epoch": 3.2632612966601178, + "grad_norm": 0.185249120833182, + "learning_rate": 2.5981710829271923e-05, + "loss": 0.4758, + "step": 1661 + }, + { + "epoch": 3.2652259332023577, + "grad_norm": 0.15927138587547982, + "learning_rate": 2.5930329296627497e-05, + "loss": 0.481, + "step": 1662 + }, + { + "epoch": 3.267190569744597, + "grad_norm": 0.1761761937244434, + "learning_rate": 2.5878974243657758e-05, + "loss": 0.4731, + "step": 1663 + }, + { + "epoch": 3.269155206286837, + "grad_norm": 0.1440043009329447, + "learning_rate": 2.5827645767014988e-05, + "loss": 0.4717, + "step": 1664 + }, + { + "epoch": 3.2711198428290764, + "grad_norm": 0.17262939999761354, + "learning_rate": 2.5776343963301352e-05, + "loss": 0.4665, + "step": 1665 + }, + { + "epoch": 3.2730844793713163, + "grad_norm": 0.1754560901583355, + "learning_rate": 2.572506892906889e-05, + "loss": 0.4684, + "step": 1666 + }, + { + "epoch": 3.275049115913556, + "grad_norm": 0.17541586686147742, + "learning_rate": 2.5673820760819236e-05, + "loss": 0.4797, + "step": 1667 + }, + { + "epoch": 3.2770137524557956, + "grad_norm": 0.15434828894712554, + "learning_rate": 2.562259955500345e-05, + "loss": 0.4797, + "step": 1668 + }, + { + "epoch": 3.2789783889980355, + "grad_norm": 0.22869727173898766, + "learning_rate": 2.5571405408021874e-05, + "loss": 0.4726, + "step": 1669 + }, + { + "epoch": 3.280943025540275, + "grad_norm": 0.19172605552914973, + "learning_rate": 2.55202384162239e-05, + "loss": 0.4691, + "step": 1670 + }, + { + "epoch": 3.282907662082515, + "grad_norm": 0.1622440336874402, + "learning_rate": 2.5469098675907848e-05, + "loss": 0.4822, + "step": 1671 + }, + { + "epoch": 3.2848722986247543, + "grad_norm": 0.1562241022408147, + "learning_rate": 2.5417986283320684e-05, + "loss": 0.4714, + "step": 1672 + }, + { + "epoch": 3.286836935166994, + "grad_norm": 0.15179241844082583, + "learning_rate": 2.536690133465796e-05, + "loss": 0.4777, + "step": 1673 + }, + { + "epoch": 3.288801571709234, + "grad_norm": 0.1736829295993475, + "learning_rate": 2.5315843926063588e-05, + "loss": 0.47, + "step": 1674 + }, + { + "epoch": 3.2907662082514735, + "grad_norm": 0.14812935965776836, + "learning_rate": 2.526481415362959e-05, + "loss": 0.4643, + "step": 1675 + }, + { + "epoch": 3.292730844793713, + "grad_norm": 0.2268871447244139, + "learning_rate": 2.5213812113396023e-05, + "loss": 0.4877, + "step": 1676 + }, + { + "epoch": 3.294695481335953, + "grad_norm": 0.2300367103346545, + "learning_rate": 2.516283790135074e-05, + "loss": 0.4773, + "step": 1677 + }, + { + "epoch": 3.2966601178781927, + "grad_norm": 0.15521608786774282, + "learning_rate": 2.5111891613429203e-05, + "loss": 0.4739, + "step": 1678 + }, + { + "epoch": 3.298624754420432, + "grad_norm": 0.28454713218542754, + "learning_rate": 2.5060973345514343e-05, + "loss": 0.4758, + "step": 1679 + }, + { + "epoch": 3.300589390962672, + "grad_norm": 0.24257441060187376, + "learning_rate": 2.501008319343637e-05, + "loss": 0.479, + "step": 1680 + }, + { + "epoch": 3.3025540275049114, + "grad_norm": 0.14705444921064514, + "learning_rate": 2.4959221252972503e-05, + "loss": 0.4733, + "step": 1681 + }, + { + "epoch": 3.3045186640471513, + "grad_norm": 0.2196770243408443, + "learning_rate": 2.4908387619846942e-05, + "loss": 0.47, + "step": 1682 + }, + { + "epoch": 3.3064833005893908, + "grad_norm": 0.20034188124593122, + "learning_rate": 2.4857582389730604e-05, + "loss": 0.4692, + "step": 1683 + }, + { + "epoch": 3.3084479371316307, + "grad_norm": 0.17237218295739737, + "learning_rate": 2.4806805658240936e-05, + "loss": 0.4671, + "step": 1684 + }, + { + "epoch": 3.3104125736738705, + "grad_norm": 0.19908098366459415, + "learning_rate": 2.475605752094172e-05, + "loss": 0.4801, + "step": 1685 + }, + { + "epoch": 3.31237721021611, + "grad_norm": 0.19229998482401855, + "learning_rate": 2.4705338073342963e-05, + "loss": 0.4787, + "step": 1686 + }, + { + "epoch": 3.31434184675835, + "grad_norm": 0.18270332990874583, + "learning_rate": 2.4654647410900675e-05, + "loss": 0.47, + "step": 1687 + }, + { + "epoch": 3.3163064833005893, + "grad_norm": 0.1571752928120313, + "learning_rate": 2.4603985629016664e-05, + "loss": 0.4833, + "step": 1688 + }, + { + "epoch": 3.318271119842829, + "grad_norm": 0.17843613226878258, + "learning_rate": 2.4553352823038404e-05, + "loss": 0.4697, + "step": 1689 + }, + { + "epoch": 3.3202357563850686, + "grad_norm": 0.1426008540451695, + "learning_rate": 2.4502749088258854e-05, + "loss": 0.4688, + "step": 1690 + }, + { + "epoch": 3.3222003929273085, + "grad_norm": 0.21789473581301538, + "learning_rate": 2.4452174519916184e-05, + "loss": 0.4714, + "step": 1691 + }, + { + "epoch": 3.324165029469548, + "grad_norm": 0.20230957191314414, + "learning_rate": 2.4401629213193774e-05, + "loss": 0.4713, + "step": 1692 + }, + { + "epoch": 3.326129666011788, + "grad_norm": 0.16781754413349437, + "learning_rate": 2.435111326321988e-05, + "loss": 0.4767, + "step": 1693 + }, + { + "epoch": 3.3280943025540273, + "grad_norm": 0.2208152353027366, + "learning_rate": 2.430062676506748e-05, + "loss": 0.4737, + "step": 1694 + }, + { + "epoch": 3.330058939096267, + "grad_norm": 0.17636772969911113, + "learning_rate": 2.425016981375417e-05, + "loss": 0.4723, + "step": 1695 + }, + { + "epoch": 3.332023575638507, + "grad_norm": 0.1748016741962013, + "learning_rate": 2.419974250424193e-05, + "loss": 0.4813, + "step": 1696 + }, + { + "epoch": 3.3339882121807465, + "grad_norm": 0.15342329553573453, + "learning_rate": 2.4149344931436927e-05, + "loss": 0.4837, + "step": 1697 + }, + { + "epoch": 3.3359528487229864, + "grad_norm": 0.182043804424447, + "learning_rate": 2.4098977190189388e-05, + "loss": 0.4723, + "step": 1698 + }, + { + "epoch": 3.337917485265226, + "grad_norm": 0.14087208238603463, + "learning_rate": 2.4048639375293382e-05, + "loss": 0.4732, + "step": 1699 + }, + { + "epoch": 3.3398821218074657, + "grad_norm": 0.21664112584103012, + "learning_rate": 2.3998331581486677e-05, + "loss": 0.4758, + "step": 1700 + }, + { + "epoch": 3.341846758349705, + "grad_norm": 0.1824886033883413, + "learning_rate": 2.394805390345049e-05, + "loss": 0.4694, + "step": 1701 + }, + { + "epoch": 3.343811394891945, + "grad_norm": 0.17953423563732143, + "learning_rate": 2.3897806435809425e-05, + "loss": 0.4764, + "step": 1702 + }, + { + "epoch": 3.345776031434185, + "grad_norm": 0.193285993965273, + "learning_rate": 2.3847589273131197e-05, + "loss": 0.481, + "step": 1703 + }, + { + "epoch": 3.3477406679764243, + "grad_norm": 0.159119504245079, + "learning_rate": 2.379740250992645e-05, + "loss": 0.4823, + "step": 1704 + }, + { + "epoch": 3.349705304518664, + "grad_norm": 0.1644529755548064, + "learning_rate": 2.3747246240648675e-05, + "loss": 0.4644, + "step": 1705 + }, + { + "epoch": 3.3516699410609037, + "grad_norm": 0.15273134360438748, + "learning_rate": 2.3697120559693952e-05, + "loss": 0.4787, + "step": 1706 + }, + { + "epoch": 3.3536345776031435, + "grad_norm": 0.1596107206416649, + "learning_rate": 2.3647025561400777e-05, + "loss": 0.4758, + "step": 1707 + }, + { + "epoch": 3.355599214145383, + "grad_norm": 0.14317818733230947, + "learning_rate": 2.359696134004991e-05, + "loss": 0.4713, + "step": 1708 + }, + { + "epoch": 3.357563850687623, + "grad_norm": 0.23092416524669482, + "learning_rate": 2.3546927989864204e-05, + "loss": 0.4684, + "step": 1709 + }, + { + "epoch": 3.3595284872298623, + "grad_norm": 0.18684328189730157, + "learning_rate": 2.3496925605008363e-05, + "loss": 0.4789, + "step": 1710 + }, + { + "epoch": 3.361493123772102, + "grad_norm": 0.1925666009172131, + "learning_rate": 2.3446954279588878e-05, + "loss": 0.4798, + "step": 1711 + }, + { + "epoch": 3.3634577603143416, + "grad_norm": 0.2091065089882808, + "learning_rate": 2.339701410765374e-05, + "loss": 0.4687, + "step": 1712 + }, + { + "epoch": 3.3654223968565815, + "grad_norm": 0.16005682601913626, + "learning_rate": 2.3347105183192324e-05, + "loss": 0.475, + "step": 1713 + }, + { + "epoch": 3.3673870333988214, + "grad_norm": 0.17181503748889307, + "learning_rate": 2.3297227600135184e-05, + "loss": 0.4753, + "step": 1714 + }, + { + "epoch": 3.369351669941061, + "grad_norm": 0.1589706914897711, + "learning_rate": 2.324738145235391e-05, + "loss": 0.4696, + "step": 1715 + }, + { + "epoch": 3.3713163064833007, + "grad_norm": 0.15494990084235905, + "learning_rate": 2.319756683366092e-05, + "loss": 0.4641, + "step": 1716 + }, + { + "epoch": 3.37328094302554, + "grad_norm": 0.19271130912979534, + "learning_rate": 2.314778383780927e-05, + "loss": 0.4818, + "step": 1717 + }, + { + "epoch": 3.37524557956778, + "grad_norm": 0.1791996801837706, + "learning_rate": 2.3098032558492518e-05, + "loss": 0.4625, + "step": 1718 + }, + { + "epoch": 3.3772102161100195, + "grad_norm": 0.17407913208225184, + "learning_rate": 2.304831308934457e-05, + "loss": 0.4728, + "step": 1719 + }, + { + "epoch": 3.3791748526522594, + "grad_norm": 0.17299238186740318, + "learning_rate": 2.2998625523939374e-05, + "loss": 0.4676, + "step": 1720 + }, + { + "epoch": 3.3811394891944992, + "grad_norm": 0.16590980869840838, + "learning_rate": 2.294896995579092e-05, + "loss": 0.4743, + "step": 1721 + }, + { + "epoch": 3.3831041257367387, + "grad_norm": 0.16779961414561204, + "learning_rate": 2.2899346478352928e-05, + "loss": 0.4887, + "step": 1722 + }, + { + "epoch": 3.3850687622789786, + "grad_norm": 0.18979589010406286, + "learning_rate": 2.2849755185018748e-05, + "loss": 0.4791, + "step": 1723 + }, + { + "epoch": 3.387033398821218, + "grad_norm": 0.18979428638620136, + "learning_rate": 2.2800196169121148e-05, + "loss": 0.462, + "step": 1724 + }, + { + "epoch": 3.388998035363458, + "grad_norm": 0.18268063091628753, + "learning_rate": 2.275066952393215e-05, + "loss": 0.4726, + "step": 1725 + }, + { + "epoch": 3.3909626719056973, + "grad_norm": 0.22075042968215566, + "learning_rate": 2.270117534266286e-05, + "loss": 0.482, + "step": 1726 + }, + { + "epoch": 3.392927308447937, + "grad_norm": 0.16031755558190852, + "learning_rate": 2.265171371846325e-05, + "loss": 0.481, + "step": 1727 + }, + { + "epoch": 3.3948919449901767, + "grad_norm": 0.18725371901779914, + "learning_rate": 2.2602284744422063e-05, + "loss": 0.477, + "step": 1728 + }, + { + "epoch": 3.3968565815324165, + "grad_norm": 0.21054982430879923, + "learning_rate": 2.2552888513566592e-05, + "loss": 0.4757, + "step": 1729 + }, + { + "epoch": 3.398821218074656, + "grad_norm": 0.15193693607897366, + "learning_rate": 2.250352511886246e-05, + "loss": 0.48, + "step": 1730 + }, + { + "epoch": 3.400785854616896, + "grad_norm": 0.22584706737936727, + "learning_rate": 2.2454194653213533e-05, + "loss": 0.4748, + "step": 1731 + }, + { + "epoch": 3.4027504911591357, + "grad_norm": 0.20421736523330028, + "learning_rate": 2.2404897209461688e-05, + "loss": 0.4693, + "step": 1732 + }, + { + "epoch": 3.404715127701375, + "grad_norm": 0.15330135299662453, + "learning_rate": 2.2355632880386663e-05, + "loss": 0.4814, + "step": 1733 + }, + { + "epoch": 3.406679764243615, + "grad_norm": 0.21312989331370272, + "learning_rate": 2.230640175870587e-05, + "loss": 0.4768, + "step": 1734 + }, + { + "epoch": 3.4086444007858545, + "grad_norm": 0.1567041871968538, + "learning_rate": 2.2257203937074222e-05, + "loss": 0.4734, + "step": 1735 + }, + { + "epoch": 3.4106090373280944, + "grad_norm": 0.21074341800653573, + "learning_rate": 2.2208039508083946e-05, + "loss": 0.4744, + "step": 1736 + }, + { + "epoch": 3.412573673870334, + "grad_norm": 0.22118957332441924, + "learning_rate": 2.2158908564264436e-05, + "loss": 0.4689, + "step": 1737 + }, + { + "epoch": 3.4145383104125737, + "grad_norm": 0.1488540073628962, + "learning_rate": 2.2109811198082066e-05, + "loss": 0.4755, + "step": 1738 + }, + { + "epoch": 3.4165029469548136, + "grad_norm": 0.2434351887149152, + "learning_rate": 2.206074750194003e-05, + "loss": 0.4788, + "step": 1739 + }, + { + "epoch": 3.418467583497053, + "grad_norm": 0.17191547555575418, + "learning_rate": 2.201171756817811e-05, + "loss": 0.4846, + "step": 1740 + }, + { + "epoch": 3.4204322200392925, + "grad_norm": 0.17054821576058124, + "learning_rate": 2.1962721489072592e-05, + "loss": 0.4738, + "step": 1741 + }, + { + "epoch": 3.4223968565815324, + "grad_norm": 0.22639770232436982, + "learning_rate": 2.191375935683602e-05, + "loss": 0.4746, + "step": 1742 + }, + { + "epoch": 3.4243614931237722, + "grad_norm": 0.14685917165188545, + "learning_rate": 2.1864831263617056e-05, + "loss": 0.4812, + "step": 1743 + }, + { + "epoch": 3.4263261296660117, + "grad_norm": 0.20588350553562218, + "learning_rate": 2.1815937301500307e-05, + "loss": 0.4767, + "step": 1744 + }, + { + "epoch": 3.4282907662082516, + "grad_norm": 0.2019593045861713, + "learning_rate": 2.1767077562506148e-05, + "loss": 0.4734, + "step": 1745 + }, + { + "epoch": 3.430255402750491, + "grad_norm": 0.18494720641064702, + "learning_rate": 2.1718252138590498e-05, + "loss": 0.4697, + "step": 1746 + }, + { + "epoch": 3.432220039292731, + "grad_norm": 0.21616758849073114, + "learning_rate": 2.166946112164475e-05, + "loss": 0.4741, + "step": 1747 + }, + { + "epoch": 3.4341846758349703, + "grad_norm": 0.15418150259990923, + "learning_rate": 2.1620704603495534e-05, + "loss": 0.4752, + "step": 1748 + }, + { + "epoch": 3.43614931237721, + "grad_norm": 0.24207321551216615, + "learning_rate": 2.1571982675904506e-05, + "loss": 0.4745, + "step": 1749 + }, + { + "epoch": 3.43811394891945, + "grad_norm": 0.18980705872017092, + "learning_rate": 2.1523295430568278e-05, + "loss": 0.4759, + "step": 1750 + }, + { + "epoch": 3.4400785854616895, + "grad_norm": 0.20313296967957456, + "learning_rate": 2.147464295911815e-05, + "loss": 0.4821, + "step": 1751 + }, + { + "epoch": 3.4420432220039294, + "grad_norm": 0.2402605516010347, + "learning_rate": 2.1426025353120002e-05, + "loss": 0.4788, + "step": 1752 + }, + { + "epoch": 3.444007858546169, + "grad_norm": 0.19806229860335228, + "learning_rate": 2.137744270407408e-05, + "loss": 0.4751, + "step": 1753 + }, + { + "epoch": 3.4459724950884087, + "grad_norm": 0.1476052052532359, + "learning_rate": 2.1328895103414843e-05, + "loss": 0.4768, + "step": 1754 + }, + { + "epoch": 3.447937131630648, + "grad_norm": 0.2019113584138431, + "learning_rate": 2.1280382642510802e-05, + "loss": 0.4705, + "step": 1755 + }, + { + "epoch": 3.449901768172888, + "grad_norm": 0.13664079435986923, + "learning_rate": 2.1231905412664297e-05, + "loss": 0.4682, + "step": 1756 + }, + { + "epoch": 3.4518664047151275, + "grad_norm": 0.2168297710818654, + "learning_rate": 2.118346350511139e-05, + "loss": 0.4759, + "step": 1757 + }, + { + "epoch": 3.4538310412573674, + "grad_norm": 0.15422013195817297, + "learning_rate": 2.1135057011021694e-05, + "loss": 0.4704, + "step": 1758 + }, + { + "epoch": 3.455795677799607, + "grad_norm": 0.1649919315647033, + "learning_rate": 2.1086686021498083e-05, + "loss": 0.4761, + "step": 1759 + }, + { + "epoch": 3.4577603143418467, + "grad_norm": 0.17407286730275576, + "learning_rate": 2.1038350627576697e-05, + "loss": 0.4812, + "step": 1760 + }, + { + "epoch": 3.4597249508840866, + "grad_norm": 0.1571079323606541, + "learning_rate": 2.0990050920226644e-05, + "loss": 0.4729, + "step": 1761 + }, + { + "epoch": 3.461689587426326, + "grad_norm": 0.1665920387606003, + "learning_rate": 2.0941786990349887e-05, + "loss": 0.4833, + "step": 1762 + }, + { + "epoch": 3.463654223968566, + "grad_norm": 0.16781785611828068, + "learning_rate": 2.0893558928781033e-05, + "loss": 0.4801, + "step": 1763 + }, + { + "epoch": 3.4656188605108054, + "grad_norm": 0.19353694452416167, + "learning_rate": 2.0845366826287215e-05, + "loss": 0.4705, + "step": 1764 + }, + { + "epoch": 3.4675834970530452, + "grad_norm": 0.17720786045858156, + "learning_rate": 2.0797210773567834e-05, + "loss": 0.4694, + "step": 1765 + }, + { + "epoch": 3.4695481335952847, + "grad_norm": 0.17330737986727202, + "learning_rate": 2.07490908612545e-05, + "loss": 0.4653, + "step": 1766 + }, + { + "epoch": 3.4715127701375246, + "grad_norm": 0.21095167655121744, + "learning_rate": 2.0701007179910784e-05, + "loss": 0.4707, + "step": 1767 + }, + { + "epoch": 3.4734774066797645, + "grad_norm": 0.16724811666112605, + "learning_rate": 2.0652959820032083e-05, + "loss": 0.4801, + "step": 1768 + }, + { + "epoch": 3.475442043222004, + "grad_norm": 0.1683131751485588, + "learning_rate": 2.0604948872045388e-05, + "loss": 0.4787, + "step": 1769 + }, + { + "epoch": 3.477406679764244, + "grad_norm": 0.2086738389390189, + "learning_rate": 2.0556974426309208e-05, + "loss": 0.4736, + "step": 1770 + }, + { + "epoch": 3.479371316306483, + "grad_norm": 0.1774644157185238, + "learning_rate": 2.0509036573113353e-05, + "loss": 0.4777, + "step": 1771 + }, + { + "epoch": 3.481335952848723, + "grad_norm": 0.16332180829886403, + "learning_rate": 2.0461135402678737e-05, + "loss": 0.4708, + "step": 1772 + }, + { + "epoch": 3.4833005893909625, + "grad_norm": 0.2080718731036305, + "learning_rate": 2.041327100515726e-05, + "loss": 0.4832, + "step": 1773 + }, + { + "epoch": 3.4852652259332024, + "grad_norm": 0.17412789498013256, + "learning_rate": 2.0365443470631615e-05, + "loss": 0.4778, + "step": 1774 + }, + { + "epoch": 3.487229862475442, + "grad_norm": 0.20232228982494263, + "learning_rate": 2.031765288911508e-05, + "loss": 0.4806, + "step": 1775 + }, + { + "epoch": 3.4891944990176817, + "grad_norm": 0.21544023764320955, + "learning_rate": 2.0269899350551427e-05, + "loss": 0.4701, + "step": 1776 + }, + { + "epoch": 3.491159135559921, + "grad_norm": 0.17593162471759188, + "learning_rate": 2.022218294481471e-05, + "loss": 0.4754, + "step": 1777 + }, + { + "epoch": 3.493123772102161, + "grad_norm": 0.16037791480522542, + "learning_rate": 2.0174503761709058e-05, + "loss": 0.4683, + "step": 1778 + }, + { + "epoch": 3.495088408644401, + "grad_norm": 0.1624606978470816, + "learning_rate": 2.0126861890968582e-05, + "loss": 0.4765, + "step": 1779 + }, + { + "epoch": 3.4970530451866404, + "grad_norm": 0.20193628278600412, + "learning_rate": 2.0079257422257168e-05, + "loss": 0.4712, + "step": 1780 + }, + { + "epoch": 3.4990176817288803, + "grad_norm": 0.18782687631502898, + "learning_rate": 2.0031690445168292e-05, + "loss": 0.4772, + "step": 1781 + }, + { + "epoch": 3.5009823182711197, + "grad_norm": 0.20041087818780895, + "learning_rate": 1.9984161049224894e-05, + "loss": 0.4703, + "step": 1782 + }, + { + "epoch": 3.5029469548133596, + "grad_norm": 0.24444138922697609, + "learning_rate": 1.9936669323879155e-05, + "loss": 0.4813, + "step": 1783 + }, + { + "epoch": 3.504911591355599, + "grad_norm": 0.1906848534133954, + "learning_rate": 1.98892153585124e-05, + "loss": 0.4737, + "step": 1784 + }, + { + "epoch": 3.506876227897839, + "grad_norm": 0.1884439019768591, + "learning_rate": 1.984179924243482e-05, + "loss": 0.4734, + "step": 1785 + }, + { + "epoch": 3.508840864440079, + "grad_norm": 0.17998073510663584, + "learning_rate": 1.979442106488544e-05, + "loss": 0.4757, + "step": 1786 + }, + { + "epoch": 3.5108055009823183, + "grad_norm": 0.15708440637733911, + "learning_rate": 1.9747080915031866e-05, + "loss": 0.4756, + "step": 1787 + }, + { + "epoch": 3.5127701375245577, + "grad_norm": 0.17294882831309152, + "learning_rate": 1.96997788819701e-05, + "loss": 0.4745, + "step": 1788 + }, + { + "epoch": 3.5147347740667976, + "grad_norm": 0.15459277678612302, + "learning_rate": 1.9652515054724444e-05, + "loss": 0.4793, + "step": 1789 + }, + { + "epoch": 3.5166994106090375, + "grad_norm": 0.16353309709637082, + "learning_rate": 1.960528952224728e-05, + "loss": 0.4738, + "step": 1790 + }, + { + "epoch": 3.518664047151277, + "grad_norm": 0.15292723270367078, + "learning_rate": 1.955810237341893e-05, + "loss": 0.4758, + "step": 1791 + }, + { + "epoch": 3.520628683693517, + "grad_norm": 0.15898838534397247, + "learning_rate": 1.9510953697047466e-05, + "loss": 0.4733, + "step": 1792 + }, + { + "epoch": 3.522593320235756, + "grad_norm": 0.1467395925037189, + "learning_rate": 1.946384358186856e-05, + "loss": 0.485, + "step": 1793 + }, + { + "epoch": 3.524557956777996, + "grad_norm": 0.22550986676685614, + "learning_rate": 1.9416772116545286e-05, + "loss": 0.4799, + "step": 1794 + }, + { + "epoch": 3.5265225933202355, + "grad_norm": 0.1845658683841781, + "learning_rate": 1.9369739389668006e-05, + "loss": 0.4761, + "step": 1795 + }, + { + "epoch": 3.5284872298624754, + "grad_norm": 0.1789391141183133, + "learning_rate": 1.932274548975417e-05, + "loss": 0.475, + "step": 1796 + }, + { + "epoch": 3.5304518664047153, + "grad_norm": 0.15172327863667479, + "learning_rate": 1.927579050524816e-05, + "loss": 0.4733, + "step": 1797 + }, + { + "epoch": 3.5324165029469548, + "grad_norm": 0.19085051853611745, + "learning_rate": 1.9228874524521077e-05, + "loss": 0.4735, + "step": 1798 + }, + { + "epoch": 3.5343811394891946, + "grad_norm": 0.17538985202136553, + "learning_rate": 1.918199763587066e-05, + "loss": 0.4759, + "step": 1799 + }, + { + "epoch": 3.536345776031434, + "grad_norm": 0.14862030647006927, + "learning_rate": 1.9135159927521068e-05, + "loss": 0.475, + "step": 1800 + }, + { + "epoch": 3.538310412573674, + "grad_norm": 0.22086488947411295, + "learning_rate": 1.9088361487622707e-05, + "loss": 0.4785, + "step": 1801 + }, + { + "epoch": 3.5402750491159134, + "grad_norm": 0.13521104814411575, + "learning_rate": 1.9041602404252088e-05, + "loss": 0.4721, + "step": 1802 + }, + { + "epoch": 3.5422396856581533, + "grad_norm": 0.25212632988223205, + "learning_rate": 1.899488276541166e-05, + "loss": 0.4729, + "step": 1803 + }, + { + "epoch": 3.544204322200393, + "grad_norm": 0.18013566002270093, + "learning_rate": 1.89482026590296e-05, + "loss": 0.4685, + "step": 1804 + }, + { + "epoch": 3.5461689587426326, + "grad_norm": 0.18320439162934132, + "learning_rate": 1.8901562172959718e-05, + "loss": 0.4704, + "step": 1805 + }, + { + "epoch": 3.548133595284872, + "grad_norm": 0.20565863508837096, + "learning_rate": 1.8854961394981263e-05, + "loss": 0.4783, + "step": 1806 + }, + { + "epoch": 3.550098231827112, + "grad_norm": 0.14310677355411575, + "learning_rate": 1.880840041279871e-05, + "loss": 0.4695, + "step": 1807 + }, + { + "epoch": 3.552062868369352, + "grad_norm": 0.20839475273071603, + "learning_rate": 1.8761879314041666e-05, + "loss": 0.4805, + "step": 1808 + }, + { + "epoch": 3.5540275049115913, + "grad_norm": 0.15283973568752912, + "learning_rate": 1.8715398186264685e-05, + "loss": 0.4807, + "step": 1809 + }, + { + "epoch": 3.555992141453831, + "grad_norm": 0.15929278135615224, + "learning_rate": 1.8668957116947074e-05, + "loss": 0.474, + "step": 1810 + }, + { + "epoch": 3.5579567779960706, + "grad_norm": 0.17016231215713942, + "learning_rate": 1.8622556193492754e-05, + "loss": 0.4676, + "step": 1811 + }, + { + "epoch": 3.5599214145383105, + "grad_norm": 0.15460638686417488, + "learning_rate": 1.8576195503230097e-05, + "loss": 0.4781, + "step": 1812 + }, + { + "epoch": 3.56188605108055, + "grad_norm": 0.22940301315290118, + "learning_rate": 1.852987513341176e-05, + "loss": 0.4784, + "step": 1813 + }, + { + "epoch": 3.56385068762279, + "grad_norm": 0.1466664729193784, + "learning_rate": 1.8483595171214473e-05, + "loss": 0.479, + "step": 1814 + }, + { + "epoch": 3.5658153241650297, + "grad_norm": 0.24374888814259385, + "learning_rate": 1.843735570373897e-05, + "loss": 0.4802, + "step": 1815 + }, + { + "epoch": 3.567779960707269, + "grad_norm": 0.15221881864122272, + "learning_rate": 1.8391156818009764e-05, + "loss": 0.469, + "step": 1816 + }, + { + "epoch": 3.569744597249509, + "grad_norm": 0.15437815246142178, + "learning_rate": 1.8344998600974944e-05, + "loss": 0.4724, + "step": 1817 + }, + { + "epoch": 3.5717092337917484, + "grad_norm": 0.1623319255765196, + "learning_rate": 1.8298881139506113e-05, + "loss": 0.4717, + "step": 1818 + }, + { + "epoch": 3.5736738703339883, + "grad_norm": 0.17718136939862006, + "learning_rate": 1.8252804520398147e-05, + "loss": 0.476, + "step": 1819 + }, + { + "epoch": 3.5756385068762278, + "grad_norm": 0.16695240859332047, + "learning_rate": 1.820676883036906e-05, + "loss": 0.4777, + "step": 1820 + }, + { + "epoch": 3.5776031434184676, + "grad_norm": 0.14624020452896597, + "learning_rate": 1.816077415605984e-05, + "loss": 0.4767, + "step": 1821 + }, + { + "epoch": 3.5795677799607075, + "grad_norm": 0.13799944751513235, + "learning_rate": 1.8114820584034268e-05, + "loss": 0.474, + "step": 1822 + }, + { + "epoch": 3.581532416502947, + "grad_norm": 0.14317867118990785, + "learning_rate": 1.80689082007788e-05, + "loss": 0.4759, + "step": 1823 + }, + { + "epoch": 3.5834970530451864, + "grad_norm": 0.12724058119701237, + "learning_rate": 1.802303709270232e-05, + "loss": 0.4754, + "step": 1824 + }, + { + "epoch": 3.5854616895874263, + "grad_norm": 0.12515013312934709, + "learning_rate": 1.7977207346136064e-05, + "loss": 0.4739, + "step": 1825 + }, + { + "epoch": 3.587426326129666, + "grad_norm": 0.16208910250030928, + "learning_rate": 1.7931419047333452e-05, + "loss": 0.463, + "step": 1826 + }, + { + "epoch": 3.5893909626719056, + "grad_norm": 0.146414996248127, + "learning_rate": 1.788567228246983e-05, + "loss": 0.4742, + "step": 1827 + }, + { + "epoch": 3.5913555992141455, + "grad_norm": 0.1506965254762223, + "learning_rate": 1.7839967137642422e-05, + "loss": 0.4829, + "step": 1828 + }, + { + "epoch": 3.593320235756385, + "grad_norm": 0.17623104119211877, + "learning_rate": 1.779430369887012e-05, + "loss": 0.4778, + "step": 1829 + }, + { + "epoch": 3.595284872298625, + "grad_norm": 0.1845587675784858, + "learning_rate": 1.774868205209331e-05, + "loss": 0.4801, + "step": 1830 + }, + { + "epoch": 3.5972495088408643, + "grad_norm": 0.15201482279010575, + "learning_rate": 1.7703102283173735e-05, + "loss": 0.4799, + "step": 1831 + }, + { + "epoch": 3.599214145383104, + "grad_norm": 0.17538820623384943, + "learning_rate": 1.7657564477894325e-05, + "loss": 0.4724, + "step": 1832 + }, + { + "epoch": 3.601178781925344, + "grad_norm": 0.13900670757719708, + "learning_rate": 1.7612068721958998e-05, + "loss": 0.4738, + "step": 1833 + }, + { + "epoch": 3.6031434184675835, + "grad_norm": 0.14241200942795978, + "learning_rate": 1.756661510099257e-05, + "loss": 0.4789, + "step": 1834 + }, + { + "epoch": 3.6051080550098233, + "grad_norm": 0.14172993973653317, + "learning_rate": 1.7521203700540553e-05, + "loss": 0.4839, + "step": 1835 + }, + { + "epoch": 3.607072691552063, + "grad_norm": 0.13133212054991358, + "learning_rate": 1.7475834606069004e-05, + "loss": 0.4718, + "step": 1836 + }, + { + "epoch": 3.6090373280943027, + "grad_norm": 0.14163696723787256, + "learning_rate": 1.7430507902964313e-05, + "loss": 0.4764, + "step": 1837 + }, + { + "epoch": 3.611001964636542, + "grad_norm": 0.11726842565857101, + "learning_rate": 1.7385223676533144e-05, + "loss": 0.4771, + "step": 1838 + }, + { + "epoch": 3.612966601178782, + "grad_norm": 0.129509682916905, + "learning_rate": 1.733998201200219e-05, + "loss": 0.4719, + "step": 1839 + }, + { + "epoch": 3.614931237721022, + "grad_norm": 0.1238437730151913, + "learning_rate": 1.7294782994518046e-05, + "loss": 0.4711, + "step": 1840 + }, + { + "epoch": 3.6168958742632613, + "grad_norm": 0.1563573994806902, + "learning_rate": 1.724962670914705e-05, + "loss": 0.4806, + "step": 1841 + }, + { + "epoch": 3.6188605108055008, + "grad_norm": 0.13744411793938235, + "learning_rate": 1.720451324087512e-05, + "loss": 0.4739, + "step": 1842 + }, + { + "epoch": 3.6208251473477406, + "grad_norm": 0.14919943191036825, + "learning_rate": 1.7159442674607548e-05, + "loss": 0.4814, + "step": 1843 + }, + { + "epoch": 3.6227897838899805, + "grad_norm": 0.20186524533474318, + "learning_rate": 1.711441509516894e-05, + "loss": 0.4696, + "step": 1844 + }, + { + "epoch": 3.62475442043222, + "grad_norm": 0.13848511334318567, + "learning_rate": 1.7069430587302984e-05, + "loss": 0.47, + "step": 1845 + }, + { + "epoch": 3.62671905697446, + "grad_norm": 0.17710585988717684, + "learning_rate": 1.7024489235672273e-05, + "loss": 0.4702, + "step": 1846 + }, + { + "epoch": 3.6286836935166993, + "grad_norm": 0.1474192595839526, + "learning_rate": 1.6979591124858204e-05, + "loss": 0.4758, + "step": 1847 + }, + { + "epoch": 3.630648330058939, + "grad_norm": 0.21083548501754268, + "learning_rate": 1.6934736339360803e-05, + "loss": 0.4799, + "step": 1848 + }, + { + "epoch": 3.6326129666011786, + "grad_norm": 0.14869728284841327, + "learning_rate": 1.688992496359854e-05, + "loss": 0.4746, + "step": 1849 + }, + { + "epoch": 3.6345776031434185, + "grad_norm": 0.16256174941742454, + "learning_rate": 1.684515708190819e-05, + "loss": 0.4731, + "step": 1850 + }, + { + "epoch": 3.6365422396856584, + "grad_norm": 0.17281583240579498, + "learning_rate": 1.6800432778544665e-05, + "loss": 0.4777, + "step": 1851 + }, + { + "epoch": 3.638506876227898, + "grad_norm": 0.14762972541205815, + "learning_rate": 1.675575213768089e-05, + "loss": 0.4822, + "step": 1852 + }, + { + "epoch": 3.6404715127701373, + "grad_norm": 0.20822643142290867, + "learning_rate": 1.671111524340756e-05, + "loss": 0.484, + "step": 1853 + }, + { + "epoch": 3.642436149312377, + "grad_norm": 0.13875300446600666, + "learning_rate": 1.666652217973308e-05, + "loss": 0.4735, + "step": 1854 + }, + { + "epoch": 3.644400785854617, + "grad_norm": 0.22476250364283945, + "learning_rate": 1.662197303058337e-05, + "loss": 0.4759, + "step": 1855 + }, + { + "epoch": 3.6463654223968565, + "grad_norm": 0.13355537694219133, + "learning_rate": 1.657746787980165e-05, + "loss": 0.4829, + "step": 1856 + }, + { + "epoch": 3.6483300589390963, + "grad_norm": 0.20680091811605122, + "learning_rate": 1.6533006811148395e-05, + "loss": 0.4793, + "step": 1857 + }, + { + "epoch": 3.6502946954813362, + "grad_norm": 0.1410113855245612, + "learning_rate": 1.648858990830108e-05, + "loss": 0.4781, + "step": 1858 + }, + { + "epoch": 3.6522593320235757, + "grad_norm": 0.20058369226222283, + "learning_rate": 1.6444217254854064e-05, + "loss": 0.4729, + "step": 1859 + }, + { + "epoch": 3.654223968565815, + "grad_norm": 0.1298380167628899, + "learning_rate": 1.639988893431843e-05, + "loss": 0.4756, + "step": 1860 + }, + { + "epoch": 3.656188605108055, + "grad_norm": 0.1587724691620548, + "learning_rate": 1.6355605030121844e-05, + "loss": 0.4726, + "step": 1861 + }, + { + "epoch": 3.658153241650295, + "grad_norm": 0.12926562137449166, + "learning_rate": 1.631136562560833e-05, + "loss": 0.4653, + "step": 1862 + }, + { + "epoch": 3.6601178781925343, + "grad_norm": 0.13085378329338532, + "learning_rate": 1.6267170804038196e-05, + "loss": 0.4717, + "step": 1863 + }, + { + "epoch": 3.662082514734774, + "grad_norm": 0.13211976070808318, + "learning_rate": 1.6223020648587848e-05, + "loss": 0.4716, + "step": 1864 + }, + { + "epoch": 3.6640471512770136, + "grad_norm": 0.1216492373514322, + "learning_rate": 1.6178915242349624e-05, + "loss": 0.4764, + "step": 1865 + }, + { + "epoch": 3.6660117878192535, + "grad_norm": 0.13742534074474466, + "learning_rate": 1.6134854668331616e-05, + "loss": 0.4748, + "step": 1866 + }, + { + "epoch": 3.667976424361493, + "grad_norm": 0.14184865546881878, + "learning_rate": 1.609083900945757e-05, + "loss": 0.478, + "step": 1867 + }, + { + "epoch": 3.669941060903733, + "grad_norm": 0.12410748847458082, + "learning_rate": 1.604686834856669e-05, + "loss": 0.4822, + "step": 1868 + }, + { + "epoch": 3.6719056974459727, + "grad_norm": 0.12720457704746918, + "learning_rate": 1.6002942768413494e-05, + "loss": 0.4809, + "step": 1869 + }, + { + "epoch": 3.673870333988212, + "grad_norm": 0.15520287362170423, + "learning_rate": 1.5959062351667656e-05, + "loss": 0.4735, + "step": 1870 + }, + { + "epoch": 3.6758349705304516, + "grad_norm": 0.13140284333687963, + "learning_rate": 1.591522718091386e-05, + "loss": 0.4779, + "step": 1871 + }, + { + "epoch": 3.6777996070726915, + "grad_norm": 0.1291991386641663, + "learning_rate": 1.5871437338651602e-05, + "loss": 0.4748, + "step": 1872 + }, + { + "epoch": 3.6797642436149314, + "grad_norm": 0.14643012785810644, + "learning_rate": 1.5827692907295102e-05, + "loss": 0.4746, + "step": 1873 + }, + { + "epoch": 3.681728880157171, + "grad_norm": 0.1396975444815299, + "learning_rate": 1.5783993969173117e-05, + "loss": 0.4758, + "step": 1874 + }, + { + "epoch": 3.6836935166994107, + "grad_norm": 0.15016007214832672, + "learning_rate": 1.574034060652874e-05, + "loss": 0.4704, + "step": 1875 + }, + { + "epoch": 3.68565815324165, + "grad_norm": 0.12626203885750603, + "learning_rate": 1.5696732901519344e-05, + "loss": 0.4753, + "step": 1876 + }, + { + "epoch": 3.68762278978389, + "grad_norm": 0.13646764793968794, + "learning_rate": 1.5653170936216334e-05, + "loss": 0.4605, + "step": 1877 + }, + { + "epoch": 3.6895874263261295, + "grad_norm": 0.13254415338092723, + "learning_rate": 1.560965479260506e-05, + "loss": 0.473, + "step": 1878 + }, + { + "epoch": 3.6915520628683693, + "grad_norm": 0.14794846875847711, + "learning_rate": 1.5566184552584606e-05, + "loss": 0.4757, + "step": 1879 + }, + { + "epoch": 3.6935166994106092, + "grad_norm": 0.16336139670704272, + "learning_rate": 1.5522760297967682e-05, + "loss": 0.4778, + "step": 1880 + }, + { + "epoch": 3.6954813359528487, + "grad_norm": 0.1372439656272778, + "learning_rate": 1.5479382110480466e-05, + "loss": 0.4719, + "step": 1881 + }, + { + "epoch": 3.6974459724950886, + "grad_norm": 0.20632181254763343, + "learning_rate": 1.543605007176238e-05, + "loss": 0.4669, + "step": 1882 + }, + { + "epoch": 3.699410609037328, + "grad_norm": 0.13975476965772707, + "learning_rate": 1.5392764263366044e-05, + "loss": 0.4716, + "step": 1883 + }, + { + "epoch": 3.701375245579568, + "grad_norm": 0.19077436128998887, + "learning_rate": 1.5349524766757072e-05, + "loss": 0.4785, + "step": 1884 + }, + { + "epoch": 3.7033398821218073, + "grad_norm": 0.1302725555352508, + "learning_rate": 1.5306331663313874e-05, + "loss": 0.4767, + "step": 1885 + }, + { + "epoch": 3.705304518664047, + "grad_norm": 0.16022282570071572, + "learning_rate": 1.526318503432758e-05, + "loss": 0.4729, + "step": 1886 + }, + { + "epoch": 3.707269155206287, + "grad_norm": 0.12844024806569504, + "learning_rate": 1.522008496100185e-05, + "loss": 0.4762, + "step": 1887 + }, + { + "epoch": 3.7092337917485265, + "grad_norm": 0.15181703062438234, + "learning_rate": 1.5177031524452721e-05, + "loss": 0.481, + "step": 1888 + }, + { + "epoch": 3.711198428290766, + "grad_norm": 0.14233746218718704, + "learning_rate": 1.5134024805708461e-05, + "loss": 0.4767, + "step": 1889 + }, + { + "epoch": 3.713163064833006, + "grad_norm": 0.13926941974994173, + "learning_rate": 1.5091064885709398e-05, + "loss": 0.4776, + "step": 1890 + }, + { + "epoch": 3.7151277013752457, + "grad_norm": 0.12812611773888485, + "learning_rate": 1.5048151845307821e-05, + "loss": 0.4825, + "step": 1891 + }, + { + "epoch": 3.717092337917485, + "grad_norm": 0.1351714724812541, + "learning_rate": 1.5005285765267728e-05, + "loss": 0.4712, + "step": 1892 + }, + { + "epoch": 3.719056974459725, + "grad_norm": 0.13357728050444856, + "learning_rate": 1.4962466726264783e-05, + "loss": 0.4773, + "step": 1893 + }, + { + "epoch": 3.7210216110019645, + "grad_norm": 0.1373240554630923, + "learning_rate": 1.4919694808886126e-05, + "loss": 0.4659, + "step": 1894 + }, + { + "epoch": 3.7229862475442044, + "grad_norm": 0.13296036773846914, + "learning_rate": 1.4876970093630156e-05, + "loss": 0.4776, + "step": 1895 + }, + { + "epoch": 3.724950884086444, + "grad_norm": 0.13185846177584348, + "learning_rate": 1.483429266090649e-05, + "loss": 0.4899, + "step": 1896 + }, + { + "epoch": 3.7269155206286837, + "grad_norm": 0.19597404439435678, + "learning_rate": 1.4791662591035735e-05, + "loss": 0.474, + "step": 1897 + }, + { + "epoch": 3.7288801571709236, + "grad_norm": 0.13936994748967788, + "learning_rate": 1.4749079964249364e-05, + "loss": 0.4709, + "step": 1898 + }, + { + "epoch": 3.730844793713163, + "grad_norm": 0.17277851279712303, + "learning_rate": 1.470654486068956e-05, + "loss": 0.473, + "step": 1899 + }, + { + "epoch": 3.732809430255403, + "grad_norm": 0.10653008024412018, + "learning_rate": 1.4664057360409075e-05, + "loss": 0.4725, + "step": 1900 + }, + { + "epoch": 3.7347740667976423, + "grad_norm": 0.2307904751243631, + "learning_rate": 1.4621617543371031e-05, + "loss": 0.4731, + "step": 1901 + }, + { + "epoch": 3.7367387033398822, + "grad_norm": 0.12154190834570353, + "learning_rate": 1.4579225489448852e-05, + "loss": 0.4774, + "step": 1902 + }, + { + "epoch": 3.7387033398821217, + "grad_norm": 0.17478516425162816, + "learning_rate": 1.4536881278426065e-05, + "loss": 0.4751, + "step": 1903 + }, + { + "epoch": 3.7406679764243616, + "grad_norm": 0.15867284888529049, + "learning_rate": 1.4494584989996118e-05, + "loss": 0.4731, + "step": 1904 + }, + { + "epoch": 3.7426326129666014, + "grad_norm": 0.15167371107770328, + "learning_rate": 1.445233670376231e-05, + "loss": 0.4816, + "step": 1905 + }, + { + "epoch": 3.744597249508841, + "grad_norm": 0.15063910271103603, + "learning_rate": 1.4410136499237575e-05, + "loss": 0.4748, + "step": 1906 + }, + { + "epoch": 3.7465618860510803, + "grad_norm": 0.12805381206214836, + "learning_rate": 1.4367984455844365e-05, + "loss": 0.4788, + "step": 1907 + }, + { + "epoch": 3.74852652259332, + "grad_norm": 0.15153467667666354, + "learning_rate": 1.4325880652914483e-05, + "loss": 0.4726, + "step": 1908 + }, + { + "epoch": 3.75049115913556, + "grad_norm": 0.12795788856069606, + "learning_rate": 1.4283825169688945e-05, + "loss": 0.4762, + "step": 1909 + }, + { + "epoch": 3.7524557956777995, + "grad_norm": 0.1495616593190737, + "learning_rate": 1.4241818085317847e-05, + "loss": 0.4847, + "step": 1910 + }, + { + "epoch": 3.7544204322200394, + "grad_norm": 0.14288303123164525, + "learning_rate": 1.4199859478860143e-05, + "loss": 0.4751, + "step": 1911 + }, + { + "epoch": 3.756385068762279, + "grad_norm": 0.15259127844354653, + "learning_rate": 1.41579494292836e-05, + "loss": 0.4806, + "step": 1912 + }, + { + "epoch": 3.7583497053045187, + "grad_norm": 0.14327207435240036, + "learning_rate": 1.4116088015464601e-05, + "loss": 0.4696, + "step": 1913 + }, + { + "epoch": 3.760314341846758, + "grad_norm": 0.13610601806662567, + "learning_rate": 1.4074275316187941e-05, + "loss": 0.4705, + "step": 1914 + }, + { + "epoch": 3.762278978388998, + "grad_norm": 0.14507693996493903, + "learning_rate": 1.4032511410146792e-05, + "loss": 0.4776, + "step": 1915 + }, + { + "epoch": 3.764243614931238, + "grad_norm": 0.13404687254029263, + "learning_rate": 1.3990796375942459e-05, + "loss": 0.4801, + "step": 1916 + }, + { + "epoch": 3.7662082514734774, + "grad_norm": 0.15479667147898524, + "learning_rate": 1.3949130292084286e-05, + "loss": 0.4772, + "step": 1917 + }, + { + "epoch": 3.768172888015717, + "grad_norm": 0.12842947181350367, + "learning_rate": 1.3907513236989486e-05, + "loss": 0.466, + "step": 1918 + }, + { + "epoch": 3.7701375245579567, + "grad_norm": 0.1432979330636418, + "learning_rate": 1.3865945288982997e-05, + "loss": 0.4734, + "step": 1919 + }, + { + "epoch": 3.7721021611001966, + "grad_norm": 0.11932831236343353, + "learning_rate": 1.3824426526297346e-05, + "loss": 0.4701, + "step": 1920 + }, + { + "epoch": 3.774066797642436, + "grad_norm": 0.12338436409398305, + "learning_rate": 1.378295702707245e-05, + "loss": 0.4733, + "step": 1921 + }, + { + "epoch": 3.776031434184676, + "grad_norm": 0.1392897072938261, + "learning_rate": 1.3741536869355563e-05, + "loss": 0.4755, + "step": 1922 + }, + { + "epoch": 3.777996070726916, + "grad_norm": 0.13751315178974577, + "learning_rate": 1.3700166131101064e-05, + "loss": 0.4696, + "step": 1923 + }, + { + "epoch": 3.7799607072691552, + "grad_norm": 0.1285940337605227, + "learning_rate": 1.3658844890170281e-05, + "loss": 0.4856, + "step": 1924 + }, + { + "epoch": 3.7819253438113947, + "grad_norm": 0.1301854466355493, + "learning_rate": 1.3617573224331442e-05, + "loss": 0.4793, + "step": 1925 + }, + { + "epoch": 3.7838899803536346, + "grad_norm": 0.1240280492246372, + "learning_rate": 1.3576351211259438e-05, + "loss": 0.4754, + "step": 1926 + }, + { + "epoch": 3.7858546168958744, + "grad_norm": 0.14091599867102242, + "learning_rate": 1.3535178928535721e-05, + "loss": 0.4641, + "step": 1927 + }, + { + "epoch": 3.787819253438114, + "grad_norm": 0.12726039253834973, + "learning_rate": 1.349405645364816e-05, + "loss": 0.4782, + "step": 1928 + }, + { + "epoch": 3.7897838899803538, + "grad_norm": 0.13532411002739544, + "learning_rate": 1.345298386399088e-05, + "loss": 0.4764, + "step": 1929 + }, + { + "epoch": 3.791748526522593, + "grad_norm": 0.12726704149702292, + "learning_rate": 1.3411961236864088e-05, + "loss": 0.4783, + "step": 1930 + }, + { + "epoch": 3.793713163064833, + "grad_norm": 0.13145266092139746, + "learning_rate": 1.3370988649473996e-05, + "loss": 0.4716, + "step": 1931 + }, + { + "epoch": 3.7956777996070725, + "grad_norm": 0.12665619051259766, + "learning_rate": 1.333006617893263e-05, + "loss": 0.4764, + "step": 1932 + }, + { + "epoch": 3.7976424361493124, + "grad_norm": 0.13563014527363057, + "learning_rate": 1.3289193902257704e-05, + "loss": 0.4684, + "step": 1933 + }, + { + "epoch": 3.7996070726915523, + "grad_norm": 0.12964951232699864, + "learning_rate": 1.3248371896372429e-05, + "loss": 0.4775, + "step": 1934 + }, + { + "epoch": 3.8015717092337917, + "grad_norm": 0.133576469678522, + "learning_rate": 1.3207600238105429e-05, + "loss": 0.479, + "step": 1935 + }, + { + "epoch": 3.803536345776031, + "grad_norm": 0.13268153827142093, + "learning_rate": 1.3166879004190608e-05, + "loss": 0.4798, + "step": 1936 + }, + { + "epoch": 3.805500982318271, + "grad_norm": 0.14448697958897222, + "learning_rate": 1.3126208271266897e-05, + "loss": 0.478, + "step": 1937 + }, + { + "epoch": 3.807465618860511, + "grad_norm": 0.1322967892665231, + "learning_rate": 1.308558811587823e-05, + "loss": 0.4776, + "step": 1938 + }, + { + "epoch": 3.8094302554027504, + "grad_norm": 0.1401358816909406, + "learning_rate": 1.304501861447335e-05, + "loss": 0.4762, + "step": 1939 + }, + { + "epoch": 3.8113948919449903, + "grad_norm": 0.14336362835637262, + "learning_rate": 1.3004499843405633e-05, + "loss": 0.4752, + "step": 1940 + }, + { + "epoch": 3.8133595284872297, + "grad_norm": 0.13419196639706946, + "learning_rate": 1.2964031878933021e-05, + "loss": 0.4754, + "step": 1941 + }, + { + "epoch": 3.8153241650294696, + "grad_norm": 0.12035073070596389, + "learning_rate": 1.2923614797217824e-05, + "loss": 0.4789, + "step": 1942 + }, + { + "epoch": 3.817288801571709, + "grad_norm": 0.13837487939901022, + "learning_rate": 1.2883248674326558e-05, + "loss": 0.4701, + "step": 1943 + }, + { + "epoch": 3.819253438113949, + "grad_norm": 0.13435351408869675, + "learning_rate": 1.2842933586229877e-05, + "loss": 0.4739, + "step": 1944 + }, + { + "epoch": 3.821218074656189, + "grad_norm": 0.12231807780197144, + "learning_rate": 1.2802669608802343e-05, + "loss": 0.471, + "step": 1945 + }, + { + "epoch": 3.8231827111984282, + "grad_norm": 0.14158775402666038, + "learning_rate": 1.276245681782239e-05, + "loss": 0.4765, + "step": 1946 + }, + { + "epoch": 3.825147347740668, + "grad_norm": 0.1395882734382486, + "learning_rate": 1.2722295288972037e-05, + "loss": 0.4744, + "step": 1947 + }, + { + "epoch": 3.8271119842829076, + "grad_norm": 0.13628617249629457, + "learning_rate": 1.2682185097836879e-05, + "loss": 0.4727, + "step": 1948 + }, + { + "epoch": 3.8290766208251474, + "grad_norm": 0.17531963504636036, + "learning_rate": 1.2642126319905886e-05, + "loss": 0.4771, + "step": 1949 + }, + { + "epoch": 3.831041257367387, + "grad_norm": 0.1412407211313014, + "learning_rate": 1.260211903057123e-05, + "loss": 0.4695, + "step": 1950 + }, + { + "epoch": 3.8330058939096268, + "grad_norm": 0.16146436459238514, + "learning_rate": 1.2562163305128224e-05, + "loss": 0.476, + "step": 1951 + }, + { + "epoch": 3.8349705304518666, + "grad_norm": 0.15071570745613413, + "learning_rate": 1.2522259218775129e-05, + "loss": 0.4713, + "step": 1952 + }, + { + "epoch": 3.836935166994106, + "grad_norm": 0.13990910604896165, + "learning_rate": 1.248240684661298e-05, + "loss": 0.4697, + "step": 1953 + }, + { + "epoch": 3.8388998035363455, + "grad_norm": 0.17241079694525463, + "learning_rate": 1.2442606263645512e-05, + "loss": 0.4729, + "step": 1954 + }, + { + "epoch": 3.8408644400785854, + "grad_norm": 0.12622933160934083, + "learning_rate": 1.240285754477903e-05, + "loss": 0.4778, + "step": 1955 + }, + { + "epoch": 3.8428290766208253, + "grad_norm": 0.16406225539713495, + "learning_rate": 1.2363160764822148e-05, + "loss": 0.4764, + "step": 1956 + }, + { + "epoch": 3.8447937131630647, + "grad_norm": 0.12418447377402649, + "learning_rate": 1.2323515998485793e-05, + "loss": 0.4682, + "step": 1957 + }, + { + "epoch": 3.8467583497053046, + "grad_norm": 0.12320377314214123, + "learning_rate": 1.2283923320382978e-05, + "loss": 0.4719, + "step": 1958 + }, + { + "epoch": 3.848722986247544, + "grad_norm": 0.1295292947017915, + "learning_rate": 1.2244382805028665e-05, + "loss": 0.4767, + "step": 1959 + }, + { + "epoch": 3.850687622789784, + "grad_norm": 0.12969788064261878, + "learning_rate": 1.220489452683967e-05, + "loss": 0.4777, + "step": 1960 + }, + { + "epoch": 3.8526522593320234, + "grad_norm": 0.12810181191825365, + "learning_rate": 1.2165458560134495e-05, + "loss": 0.468, + "step": 1961 + }, + { + "epoch": 3.8546168958742633, + "grad_norm": 0.12819386937134983, + "learning_rate": 1.2126074979133193e-05, + "loss": 0.4734, + "step": 1962 + }, + { + "epoch": 3.856581532416503, + "grad_norm": 0.13129993829533743, + "learning_rate": 1.2086743857957189e-05, + "loss": 0.4815, + "step": 1963 + }, + { + "epoch": 3.8585461689587426, + "grad_norm": 0.11659760764235022, + "learning_rate": 1.2047465270629202e-05, + "loss": 0.4721, + "step": 1964 + }, + { + "epoch": 3.8605108055009825, + "grad_norm": 0.14222971637123596, + "learning_rate": 1.2008239291073118e-05, + "loss": 0.4859, + "step": 1965 + }, + { + "epoch": 3.862475442043222, + "grad_norm": 0.11367015153272012, + "learning_rate": 1.1969065993113746e-05, + "loss": 0.474, + "step": 1966 + }, + { + "epoch": 3.864440078585462, + "grad_norm": 0.13605887238570435, + "learning_rate": 1.1929945450476782e-05, + "loss": 0.4716, + "step": 1967 + }, + { + "epoch": 3.8664047151277012, + "grad_norm": 0.13045813015562494, + "learning_rate": 1.1890877736788645e-05, + "loss": 0.4683, + "step": 1968 + }, + { + "epoch": 3.868369351669941, + "grad_norm": 0.11514469635409404, + "learning_rate": 1.1851862925576283e-05, + "loss": 0.479, + "step": 1969 + }, + { + "epoch": 3.870333988212181, + "grad_norm": 0.14665370923757978, + "learning_rate": 1.1812901090267128e-05, + "loss": 0.4775, + "step": 1970 + }, + { + "epoch": 3.8722986247544204, + "grad_norm": 0.13618327742674677, + "learning_rate": 1.1773992304188892e-05, + "loss": 0.4814, + "step": 1971 + }, + { + "epoch": 3.87426326129666, + "grad_norm": 0.13047958469860632, + "learning_rate": 1.1735136640569426e-05, + "loss": 0.4741, + "step": 1972 + }, + { + "epoch": 3.8762278978388998, + "grad_norm": 0.123684693289355, + "learning_rate": 1.1696334172536617e-05, + "loss": 0.4815, + "step": 1973 + }, + { + "epoch": 3.8781925343811396, + "grad_norm": 0.15558849795875423, + "learning_rate": 1.1657584973118263e-05, + "loss": 0.482, + "step": 1974 + }, + { + "epoch": 3.880157170923379, + "grad_norm": 0.12773309346831555, + "learning_rate": 1.1618889115241885e-05, + "loss": 0.477, + "step": 1975 + }, + { + "epoch": 3.882121807465619, + "grad_norm": 0.13207192367344703, + "learning_rate": 1.158024667173459e-05, + "loss": 0.4711, + "step": 1976 + }, + { + "epoch": 3.8840864440078584, + "grad_norm": 0.12601877909486112, + "learning_rate": 1.1541657715322994e-05, + "loss": 0.4773, + "step": 1977 + }, + { + "epoch": 3.8860510805500983, + "grad_norm": 0.15161886662235652, + "learning_rate": 1.1503122318633038e-05, + "loss": 0.4732, + "step": 1978 + }, + { + "epoch": 3.8880157170923377, + "grad_norm": 0.1330684832321734, + "learning_rate": 1.1464640554189837e-05, + "loss": 0.4743, + "step": 1979 + }, + { + "epoch": 3.8899803536345776, + "grad_norm": 0.1395920728580842, + "learning_rate": 1.1426212494417599e-05, + "loss": 0.4725, + "step": 1980 + }, + { + "epoch": 3.8919449901768175, + "grad_norm": 0.12806323395231084, + "learning_rate": 1.1387838211639464e-05, + "loss": 0.4742, + "step": 1981 + }, + { + "epoch": 3.893909626719057, + "grad_norm": 0.13109033594118033, + "learning_rate": 1.13495177780773e-05, + "loss": 0.4781, + "step": 1982 + }, + { + "epoch": 3.895874263261297, + "grad_norm": 0.11226107273194252, + "learning_rate": 1.1311251265851713e-05, + "loss": 0.4715, + "step": 1983 + }, + { + "epoch": 3.8978388998035363, + "grad_norm": 0.1540644728142834, + "learning_rate": 1.1273038746981787e-05, + "loss": 0.486, + "step": 1984 + }, + { + "epoch": 3.899803536345776, + "grad_norm": 0.12056085924639302, + "learning_rate": 1.123488029338497e-05, + "loss": 0.4684, + "step": 1985 + }, + { + "epoch": 3.9017681728880156, + "grad_norm": 0.11914037930963058, + "learning_rate": 1.1196775976876978e-05, + "loss": 0.4803, + "step": 1986 + }, + { + "epoch": 3.9037328094302555, + "grad_norm": 0.13888570247386003, + "learning_rate": 1.1158725869171643e-05, + "loss": 0.4792, + "step": 1987 + }, + { + "epoch": 3.9056974459724954, + "grad_norm": 0.14022300811414434, + "learning_rate": 1.1120730041880772e-05, + "loss": 0.4785, + "step": 1988 + }, + { + "epoch": 3.907662082514735, + "grad_norm": 0.1420932349739875, + "learning_rate": 1.1082788566513991e-05, + "loss": 0.479, + "step": 1989 + }, + { + "epoch": 3.9096267190569742, + "grad_norm": 0.14453946767947715, + "learning_rate": 1.1044901514478665e-05, + "loss": 0.475, + "step": 1990 + }, + { + "epoch": 3.911591355599214, + "grad_norm": 0.12439542431573422, + "learning_rate": 1.100706895707973e-05, + "loss": 0.4708, + "step": 1991 + }, + { + "epoch": 3.913555992141454, + "grad_norm": 0.13414663144931546, + "learning_rate": 1.0969290965519508e-05, + "loss": 0.4696, + "step": 1992 + }, + { + "epoch": 3.9155206286836934, + "grad_norm": 0.12270946880015605, + "learning_rate": 1.0931567610897709e-05, + "loss": 0.4638, + "step": 1993 + }, + { + "epoch": 3.9174852652259333, + "grad_norm": 0.11804906946178509, + "learning_rate": 1.0893898964211172e-05, + "loss": 0.4729, + "step": 1994 + }, + { + "epoch": 3.9194499017681728, + "grad_norm": 0.13677288582670427, + "learning_rate": 1.0856285096353752e-05, + "loss": 0.468, + "step": 1995 + }, + { + "epoch": 3.9214145383104126, + "grad_norm": 0.10792966594515598, + "learning_rate": 1.081872607811624e-05, + "loss": 0.4777, + "step": 1996 + }, + { + "epoch": 3.923379174852652, + "grad_norm": 0.13570760924105785, + "learning_rate": 1.0781221980186198e-05, + "loss": 0.4747, + "step": 1997 + }, + { + "epoch": 3.925343811394892, + "grad_norm": 0.11786410542703092, + "learning_rate": 1.0743772873147802e-05, + "loss": 0.4667, + "step": 1998 + }, + { + "epoch": 3.927308447937132, + "grad_norm": 0.11839846937969158, + "learning_rate": 1.0706378827481748e-05, + "loss": 0.4741, + "step": 1999 + }, + { + "epoch": 3.9292730844793713, + "grad_norm": 0.11057986264676882, + "learning_rate": 1.0669039913565124e-05, + "loss": 0.473, + "step": 2000 + }, + { + "epoch": 3.9312377210216107, + "grad_norm": 0.12582602994360723, + "learning_rate": 1.0631756201671198e-05, + "loss": 0.4697, + "step": 2001 + }, + { + "epoch": 3.9332023575638506, + "grad_norm": 0.11525299372028208, + "learning_rate": 1.0594527761969418e-05, + "loss": 0.4755, + "step": 2002 + }, + { + "epoch": 3.9351669941060905, + "grad_norm": 0.11100272632417427, + "learning_rate": 1.0557354664525166e-05, + "loss": 0.4853, + "step": 2003 + }, + { + "epoch": 3.93713163064833, + "grad_norm": 0.1149582223604783, + "learning_rate": 1.0520236979299688e-05, + "loss": 0.4814, + "step": 2004 + }, + { + "epoch": 3.93909626719057, + "grad_norm": 0.1259306856839234, + "learning_rate": 1.0483174776149907e-05, + "loss": 0.4744, + "step": 2005 + }, + { + "epoch": 3.9410609037328097, + "grad_norm": 0.10975107963479526, + "learning_rate": 1.0446168124828358e-05, + "loss": 0.4734, + "step": 2006 + }, + { + "epoch": 3.943025540275049, + "grad_norm": 0.1093897767468634, + "learning_rate": 1.0409217094983032e-05, + "loss": 0.4809, + "step": 2007 + }, + { + "epoch": 3.9449901768172886, + "grad_norm": 0.11900208354260462, + "learning_rate": 1.0372321756157203e-05, + "loss": 0.4778, + "step": 2008 + }, + { + "epoch": 3.9469548133595285, + "grad_norm": 0.13409619637346837, + "learning_rate": 1.033548217778936e-05, + "loss": 0.4789, + "step": 2009 + }, + { + "epoch": 3.9489194499017684, + "grad_norm": 0.11792013552954979, + "learning_rate": 1.029869842921305e-05, + "loss": 0.4773, + "step": 2010 + }, + { + "epoch": 3.950884086444008, + "grad_norm": 0.1330974215349077, + "learning_rate": 1.0261970579656708e-05, + "loss": 0.4711, + "step": 2011 + }, + { + "epoch": 3.9528487229862477, + "grad_norm": 0.12933969783388383, + "learning_rate": 1.0225298698243629e-05, + "loss": 0.4803, + "step": 2012 + }, + { + "epoch": 3.954813359528487, + "grad_norm": 0.12402128201033763, + "learning_rate": 1.0188682853991737e-05, + "loss": 0.483, + "step": 2013 + }, + { + "epoch": 3.956777996070727, + "grad_norm": 0.12444162736327548, + "learning_rate": 1.0152123115813467e-05, + "loss": 0.4746, + "step": 2014 + }, + { + "epoch": 3.9587426326129664, + "grad_norm": 0.13214248352131322, + "learning_rate": 1.0115619552515702e-05, + "loss": 0.4738, + "step": 2015 + }, + { + "epoch": 3.9607072691552063, + "grad_norm": 0.1384033965915973, + "learning_rate": 1.007917223279959e-05, + "loss": 0.4703, + "step": 2016 + }, + { + "epoch": 3.962671905697446, + "grad_norm": 0.12674407547534627, + "learning_rate": 1.004278122526043e-05, + "loss": 0.4749, + "step": 2017 + }, + { + "epoch": 3.9646365422396856, + "grad_norm": 0.13005806550304355, + "learning_rate": 1.0006446598387512e-05, + "loss": 0.4754, + "step": 2018 + }, + { + "epoch": 3.966601178781925, + "grad_norm": 0.14331257426109825, + "learning_rate": 9.970168420564046e-06, + "loss": 0.4736, + "step": 2019 + }, + { + "epoch": 3.968565815324165, + "grad_norm": 0.1356126132442129, + "learning_rate": 9.933946760066991e-06, + "loss": 0.4737, + "step": 2020 + }, + { + "epoch": 3.970530451866405, + "grad_norm": 0.1426830415362139, + "learning_rate": 9.89778168506694e-06, + "loss": 0.4715, + "step": 2021 + }, + { + "epoch": 3.9724950884086443, + "grad_norm": 0.1275110890912989, + "learning_rate": 9.861673263627987e-06, + "loss": 0.4715, + "step": 2022 + }, + { + "epoch": 3.974459724950884, + "grad_norm": 0.1256584537868224, + "learning_rate": 9.825621563707623e-06, + "loss": 0.4762, + "step": 2023 + }, + { + "epoch": 3.9764243614931236, + "grad_norm": 0.12257490826206864, + "learning_rate": 9.789626653156538e-06, + "loss": 0.4683, + "step": 2024 + }, + { + "epoch": 3.9783889980353635, + "grad_norm": 0.140990547867585, + "learning_rate": 9.753688599718586e-06, + "loss": 0.4713, + "step": 2025 + }, + { + "epoch": 3.980353634577603, + "grad_norm": 0.1136052290487511, + "learning_rate": 9.717807471030612e-06, + "loss": 0.4666, + "step": 2026 + }, + { + "epoch": 3.982318271119843, + "grad_norm": 0.1429135213615646, + "learning_rate": 9.681983334622287e-06, + "loss": 0.4707, + "step": 2027 + }, + { + "epoch": 3.9842829076620827, + "grad_norm": 0.13845261946796622, + "learning_rate": 9.646216257916064e-06, + "loss": 0.4713, + "step": 2028 + }, + { + "epoch": 3.986247544204322, + "grad_norm": 0.11833296581612218, + "learning_rate": 9.610506308226987e-06, + "loss": 0.4808, + "step": 2029 + }, + { + "epoch": 3.988212180746562, + "grad_norm": 0.11177053693232786, + "learning_rate": 9.574853552762594e-06, + "loss": 0.4716, + "step": 2030 + }, + { + "epoch": 3.9901768172888015, + "grad_norm": 0.1463679246953433, + "learning_rate": 9.53925805862277e-06, + "loss": 0.474, + "step": 2031 + }, + { + "epoch": 3.9921414538310414, + "grad_norm": 0.13503457552764342, + "learning_rate": 9.503719892799644e-06, + "loss": 0.4793, + "step": 2032 + }, + { + "epoch": 3.994106090373281, + "grad_norm": 0.13293424080880223, + "learning_rate": 9.468239122177447e-06, + "loss": 0.4705, + "step": 2033 + }, + { + "epoch": 3.9960707269155207, + "grad_norm": 0.12500673516938463, + "learning_rate": 9.43281581353237e-06, + "loss": 0.4732, + "step": 2034 + }, + { + "epoch": 3.9980353634577606, + "grad_norm": 0.11894536068910787, + "learning_rate": 9.397450033532487e-06, + "loss": 0.4747, + "step": 2035 + }, + { + "epoch": 4.0, + "grad_norm": 0.13984017975447918, + "learning_rate": 9.3621418487376e-06, + "loss": 0.4731, + "step": 2036 + }, + { + "epoch": 4.001964636542239, + "grad_norm": 0.19729288791790972, + "learning_rate": 9.326891325599083e-06, + "loss": 0.4536, + "step": 2037 + }, + { + "epoch": 4.00392927308448, + "grad_norm": 0.14055041750892458, + "learning_rate": 9.291698530459824e-06, + "loss": 0.4435, + "step": 2038 + }, + { + "epoch": 4.005893909626719, + "grad_norm": 0.15285604762400465, + "learning_rate": 9.25656352955404e-06, + "loss": 0.4493, + "step": 2039 + }, + { + "epoch": 4.007858546168959, + "grad_norm": 0.15397214708118823, + "learning_rate": 9.221486389007199e-06, + "loss": 0.4486, + "step": 2040 + }, + { + "epoch": 4.009823182711198, + "grad_norm": 0.16132928952334205, + "learning_rate": 9.186467174835854e-06, + "loss": 0.4454, + "step": 2041 + }, + { + "epoch": 4.011787819253438, + "grad_norm": 0.15704455951371174, + "learning_rate": 9.151505952947545e-06, + "loss": 0.4484, + "step": 2042 + }, + { + "epoch": 4.013752455795678, + "grad_norm": 0.14770081346672517, + "learning_rate": 9.116602789140692e-06, + "loss": 0.4495, + "step": 2043 + }, + { + "epoch": 4.015717092337917, + "grad_norm": 0.14189017561046158, + "learning_rate": 9.081757749104389e-06, + "loss": 0.4453, + "step": 2044 + }, + { + "epoch": 4.017681728880157, + "grad_norm": 0.12580587627643433, + "learning_rate": 9.04697089841839e-06, + "loss": 0.4396, + "step": 2045 + }, + { + "epoch": 4.019646365422397, + "grad_norm": 0.12986405388147773, + "learning_rate": 9.012242302552937e-06, + "loss": 0.4434, + "step": 2046 + }, + { + "epoch": 4.0216110019646365, + "grad_norm": 0.13542137711540764, + "learning_rate": 8.97757202686858e-06, + "loss": 0.4441, + "step": 2047 + }, + { + "epoch": 4.023575638506876, + "grad_norm": 0.12261802109966066, + "learning_rate": 8.942960136616165e-06, + "loss": 0.4467, + "step": 2048 + }, + { + "epoch": 4.025540275049116, + "grad_norm": 0.1317511684262517, + "learning_rate": 8.908406696936636e-06, + "loss": 0.4417, + "step": 2049 + }, + { + "epoch": 4.027504911591356, + "grad_norm": 0.12438822797711419, + "learning_rate": 8.873911772860916e-06, + "loss": 0.4557, + "step": 2050 + }, + { + "epoch": 4.029469548133595, + "grad_norm": 0.1392789348801343, + "learning_rate": 8.839475429309826e-06, + "loss": 0.4369, + "step": 2051 + }, + { + "epoch": 4.031434184675835, + "grad_norm": 0.12238192198779436, + "learning_rate": 8.805097731093922e-06, + "loss": 0.4445, + "step": 2052 + }, + { + "epoch": 4.033398821218075, + "grad_norm": 0.1279955657459334, + "learning_rate": 8.770778742913371e-06, + "loss": 0.4354, + "step": 2053 + }, + { + "epoch": 4.035363457760314, + "grad_norm": 0.1257275872059187, + "learning_rate": 8.736518529357866e-06, + "loss": 0.4421, + "step": 2054 + }, + { + "epoch": 4.037328094302554, + "grad_norm": 0.12340375135795781, + "learning_rate": 8.702317154906495e-06, + "loss": 0.4446, + "step": 2055 + }, + { + "epoch": 4.039292730844794, + "grad_norm": 0.11424838754049221, + "learning_rate": 8.668174683927568e-06, + "loss": 0.4362, + "step": 2056 + }, + { + "epoch": 4.041257367387034, + "grad_norm": 0.1686105285264625, + "learning_rate": 8.634091180678571e-06, + "loss": 0.4447, + "step": 2057 + }, + { + "epoch": 4.043222003929273, + "grad_norm": 0.12560955582510888, + "learning_rate": 8.600066709306008e-06, + "loss": 0.4515, + "step": 2058 + }, + { + "epoch": 4.045186640471512, + "grad_norm": 0.11820545912379736, + "learning_rate": 8.56610133384526e-06, + "loss": 0.4343, + "step": 2059 + }, + { + "epoch": 4.047151277013753, + "grad_norm": 0.14092789544749937, + "learning_rate": 8.532195118220516e-06, + "loss": 0.4408, + "step": 2060 + }, + { + "epoch": 4.049115913555992, + "grad_norm": 0.12261016226929441, + "learning_rate": 8.498348126244607e-06, + "loss": 0.4355, + "step": 2061 + }, + { + "epoch": 4.051080550098232, + "grad_norm": 0.12899993644947907, + "learning_rate": 8.464560421618912e-06, + "loss": 0.4374, + "step": 2062 + }, + { + "epoch": 4.053045186640471, + "grad_norm": 0.12431055902526018, + "learning_rate": 8.430832067933203e-06, + "loss": 0.4483, + "step": 2063 + }, + { + "epoch": 4.055009823182711, + "grad_norm": 0.1227957451996647, + "learning_rate": 8.397163128665582e-06, + "loss": 0.4397, + "step": 2064 + }, + { + "epoch": 4.056974459724951, + "grad_norm": 0.13442304087018314, + "learning_rate": 8.363553667182329e-06, + "loss": 0.4529, + "step": 2065 + }, + { + "epoch": 4.05893909626719, + "grad_norm": 0.11928272240662115, + "learning_rate": 8.330003746737758e-06, + "loss": 0.4463, + "step": 2066 + }, + { + "epoch": 4.060903732809431, + "grad_norm": 0.12396182363382274, + "learning_rate": 8.296513430474146e-06, + "loss": 0.4474, + "step": 2067 + }, + { + "epoch": 4.06286836935167, + "grad_norm": 0.12378356208598935, + "learning_rate": 8.263082781421592e-06, + "loss": 0.4419, + "step": 2068 + }, + { + "epoch": 4.0648330058939095, + "grad_norm": 0.12426121309467264, + "learning_rate": 8.229711862497889e-06, + "loss": 0.4392, + "step": 2069 + }, + { + "epoch": 4.066797642436149, + "grad_norm": 0.11419515198466368, + "learning_rate": 8.196400736508429e-06, + "loss": 0.4402, + "step": 2070 + }, + { + "epoch": 4.068762278978389, + "grad_norm": 0.14148108063118428, + "learning_rate": 8.163149466146065e-06, + "loss": 0.4432, + "step": 2071 + }, + { + "epoch": 4.070726915520629, + "grad_norm": 0.13926477214775124, + "learning_rate": 8.129958113990999e-06, + "loss": 0.4472, + "step": 2072 + }, + { + "epoch": 4.072691552062868, + "grad_norm": 0.1191247464908331, + "learning_rate": 8.096826742510653e-06, + "loss": 0.4313, + "step": 2073 + }, + { + "epoch": 4.0746561886051085, + "grad_norm": 0.12061250274756262, + "learning_rate": 8.063755414059575e-06, + "loss": 0.4371, + "step": 2074 + }, + { + "epoch": 4.076620825147348, + "grad_norm": 0.13257185897505241, + "learning_rate": 8.030744190879325e-06, + "loss": 0.4451, + "step": 2075 + }, + { + "epoch": 4.078585461689587, + "grad_norm": 0.11330178202474384, + "learning_rate": 7.997793135098302e-06, + "loss": 0.44, + "step": 2076 + }, + { + "epoch": 4.080550098231827, + "grad_norm": 0.1000619739232309, + "learning_rate": 7.964902308731703e-06, + "loss": 0.4399, + "step": 2077 + }, + { + "epoch": 4.082514734774067, + "grad_norm": 0.11845637195822148, + "learning_rate": 7.932071773681365e-06, + "loss": 0.4405, + "step": 2078 + }, + { + "epoch": 4.084479371316307, + "grad_norm": 0.12448285254983095, + "learning_rate": 7.899301591735638e-06, + "loss": 0.4558, + "step": 2079 + }, + { + "epoch": 4.086444007858546, + "grad_norm": 0.10520886365266217, + "learning_rate": 7.866591824569299e-06, + "loss": 0.435, + "step": 2080 + }, + { + "epoch": 4.088408644400785, + "grad_norm": 0.1398458107638834, + "learning_rate": 7.833942533743437e-06, + "loss": 0.4384, + "step": 2081 + }, + { + "epoch": 4.090373280943026, + "grad_norm": 0.10845158578378299, + "learning_rate": 7.801353780705274e-06, + "loss": 0.4467, + "step": 2082 + }, + { + "epoch": 4.092337917485265, + "grad_norm": 0.1053233737415637, + "learning_rate": 7.76882562678814e-06, + "loss": 0.4401, + "step": 2083 + }, + { + "epoch": 4.094302554027505, + "grad_norm": 0.11747308918933207, + "learning_rate": 7.736358133211306e-06, + "loss": 0.4388, + "step": 2084 + }, + { + "epoch": 4.096267190569745, + "grad_norm": 0.1056137201150823, + "learning_rate": 7.703951361079881e-06, + "loss": 0.44, + "step": 2085 + }, + { + "epoch": 4.098231827111984, + "grad_norm": 0.10798920332922188, + "learning_rate": 7.671605371384667e-06, + "loss": 0.4455, + "step": 2086 + }, + { + "epoch": 4.100196463654224, + "grad_norm": 0.11389024082964656, + "learning_rate": 7.639320225002106e-06, + "loss": 0.442, + "step": 2087 + }, + { + "epoch": 4.102161100196463, + "grad_norm": 0.10886061184826111, + "learning_rate": 7.607095982694108e-06, + "loss": 0.4384, + "step": 2088 + }, + { + "epoch": 4.104125736738704, + "grad_norm": 0.10243544738964576, + "learning_rate": 7.574932705107976e-06, + "loss": 0.4351, + "step": 2089 + }, + { + "epoch": 4.106090373280943, + "grad_norm": 0.1093064712956818, + "learning_rate": 7.542830452776257e-06, + "loss": 0.4455, + "step": 2090 + }, + { + "epoch": 4.1080550098231825, + "grad_norm": 0.11278013588995478, + "learning_rate": 7.510789286116673e-06, + "loss": 0.4352, + "step": 2091 + }, + { + "epoch": 4.110019646365423, + "grad_norm": 0.1016294009701617, + "learning_rate": 7.478809265431932e-06, + "loss": 0.4443, + "step": 2092 + }, + { + "epoch": 4.111984282907662, + "grad_norm": 0.10707289934043657, + "learning_rate": 7.4468904509097164e-06, + "loss": 0.4441, + "step": 2093 + }, + { + "epoch": 4.113948919449902, + "grad_norm": 0.12156781205987957, + "learning_rate": 7.4150329026224924e-06, + "loss": 0.4441, + "step": 2094 + }, + { + "epoch": 4.115913555992141, + "grad_norm": 0.10589954932945264, + "learning_rate": 7.383236680527402e-06, + "loss": 0.4383, + "step": 2095 + }, + { + "epoch": 4.1178781925343815, + "grad_norm": 0.10607315915475365, + "learning_rate": 7.351501844466202e-06, + "loss": 0.4461, + "step": 2096 + }, + { + "epoch": 4.119842829076621, + "grad_norm": 0.10370874865294417, + "learning_rate": 7.319828454165105e-06, + "loss": 0.4346, + "step": 2097 + }, + { + "epoch": 4.12180746561886, + "grad_norm": 0.10604235505669532, + "learning_rate": 7.288216569234681e-06, + "loss": 0.4384, + "step": 2098 + }, + { + "epoch": 4.1237721021611, + "grad_norm": 0.11135036863555303, + "learning_rate": 7.256666249169737e-06, + "loss": 0.442, + "step": 2099 + }, + { + "epoch": 4.12573673870334, + "grad_norm": 0.10396254229286951, + "learning_rate": 7.225177553349221e-06, + "loss": 0.4472, + "step": 2100 + }, + { + "epoch": 4.12770137524558, + "grad_norm": 0.10594483876837803, + "learning_rate": 7.193750541036113e-06, + "loss": 0.4437, + "step": 2101 + }, + { + "epoch": 4.129666011787819, + "grad_norm": 0.10152705443623856, + "learning_rate": 7.162385271377266e-06, + "loss": 0.4339, + "step": 2102 + }, + { + "epoch": 4.131630648330059, + "grad_norm": 0.10299731605808453, + "learning_rate": 7.131081803403362e-06, + "loss": 0.432, + "step": 2103 + }, + { + "epoch": 4.133595284872299, + "grad_norm": 0.10245433437455277, + "learning_rate": 7.099840196028771e-06, + "loss": 0.4362, + "step": 2104 + }, + { + "epoch": 4.135559921414538, + "grad_norm": 0.10400945850088911, + "learning_rate": 7.0686605080514124e-06, + "loss": 0.4495, + "step": 2105 + }, + { + "epoch": 4.137524557956778, + "grad_norm": 0.10534156672813239, + "learning_rate": 7.037542798152697e-06, + "loss": 0.4386, + "step": 2106 + }, + { + "epoch": 4.139489194499018, + "grad_norm": 0.10286117057166225, + "learning_rate": 7.006487124897376e-06, + "loss": 0.4384, + "step": 2107 + }, + { + "epoch": 4.141453831041257, + "grad_norm": 0.11067665592439438, + "learning_rate": 6.975493546733458e-06, + "loss": 0.4326, + "step": 2108 + }, + { + "epoch": 4.143418467583497, + "grad_norm": 0.11335922198040675, + "learning_rate": 6.944562121992078e-06, + "loss": 0.4642, + "step": 2109 + }, + { + "epoch": 4.145383104125736, + "grad_norm": 0.10285339242096596, + "learning_rate": 6.913692908887401e-06, + "loss": 0.4452, + "step": 2110 + }, + { + "epoch": 4.147347740667977, + "grad_norm": 0.12261326727537573, + "learning_rate": 6.882885965516481e-06, + "loss": 0.4473, + "step": 2111 + }, + { + "epoch": 4.149312377210216, + "grad_norm": 0.11210969852918513, + "learning_rate": 6.852141349859218e-06, + "loss": 0.4406, + "step": 2112 + }, + { + "epoch": 4.1512770137524555, + "grad_norm": 0.10572555539351954, + "learning_rate": 6.821459119778185e-06, + "loss": 0.4407, + "step": 2113 + }, + { + "epoch": 4.153241650294696, + "grad_norm": 0.1161906187201173, + "learning_rate": 6.790839333018557e-06, + "loss": 0.4391, + "step": 2114 + }, + { + "epoch": 4.155206286836935, + "grad_norm": 0.11580827020762374, + "learning_rate": 6.76028204720796e-06, + "loss": 0.4382, + "step": 2115 + }, + { + "epoch": 4.157170923379175, + "grad_norm": 0.1155609569470305, + "learning_rate": 6.729787319856425e-06, + "loss": 0.443, + "step": 2116 + }, + { + "epoch": 4.159135559921414, + "grad_norm": 0.1114240404121962, + "learning_rate": 6.699355208356224e-06, + "loss": 0.4331, + "step": 2117 + }, + { + "epoch": 4.1611001964636545, + "grad_norm": 0.11102563054367216, + "learning_rate": 6.668985769981789e-06, + "loss": 0.4451, + "step": 2118 + }, + { + "epoch": 4.163064833005894, + "grad_norm": 0.11421037683139396, + "learning_rate": 6.6386790618896106e-06, + "loss": 0.4371, + "step": 2119 + }, + { + "epoch": 4.165029469548133, + "grad_norm": 0.10808426882893715, + "learning_rate": 6.608435141118104e-06, + "loss": 0.4479, + "step": 2120 + }, + { + "epoch": 4.166994106090374, + "grad_norm": 0.11167569500985938, + "learning_rate": 6.57825406458751e-06, + "loss": 0.442, + "step": 2121 + }, + { + "epoch": 4.168958742632613, + "grad_norm": 0.11532026962906024, + "learning_rate": 6.548135889099807e-06, + "loss": 0.4326, + "step": 2122 + }, + { + "epoch": 4.170923379174853, + "grad_norm": 0.13110757462623995, + "learning_rate": 6.5180806713386025e-06, + "loss": 0.4389, + "step": 2123 + }, + { + "epoch": 4.172888015717092, + "grad_norm": 0.11401025519530562, + "learning_rate": 6.4880884678689825e-06, + "loss": 0.4492, + "step": 2124 + }, + { + "epoch": 4.174852652259332, + "grad_norm": 0.1184432231143975, + "learning_rate": 6.458159335137461e-06, + "loss": 0.4446, + "step": 2125 + }, + { + "epoch": 4.176817288801572, + "grad_norm": 0.1286157915641744, + "learning_rate": 6.428293329471849e-06, + "loss": 0.4438, + "step": 2126 + }, + { + "epoch": 4.178781925343811, + "grad_norm": 0.1119969519107301, + "learning_rate": 6.398490507081141e-06, + "loss": 0.4418, + "step": 2127 + }, + { + "epoch": 4.1807465618860515, + "grad_norm": 0.11106946050619011, + "learning_rate": 6.368750924055427e-06, + "loss": 0.4485, + "step": 2128 + }, + { + "epoch": 4.182711198428291, + "grad_norm": 0.1347738134858084, + "learning_rate": 6.339074636365778e-06, + "loss": 0.448, + "step": 2129 + }, + { + "epoch": 4.18467583497053, + "grad_norm": 0.12758129311516628, + "learning_rate": 6.309461699864137e-06, + "loss": 0.4357, + "step": 2130 + }, + { + "epoch": 4.18664047151277, + "grad_norm": 0.10653656733140503, + "learning_rate": 6.279912170283204e-06, + "loss": 0.4432, + "step": 2131 + }, + { + "epoch": 4.18860510805501, + "grad_norm": 0.13267674000301252, + "learning_rate": 6.250426103236371e-06, + "loss": 0.4401, + "step": 2132 + }, + { + "epoch": 4.19056974459725, + "grad_norm": 0.11719484093098677, + "learning_rate": 6.221003554217579e-06, + "loss": 0.4393, + "step": 2133 + }, + { + "epoch": 4.192534381139489, + "grad_norm": 0.10923477281915751, + "learning_rate": 6.1916445786012145e-06, + "loss": 0.4401, + "step": 2134 + }, + { + "epoch": 4.1944990176817285, + "grad_norm": 0.10482903497684103, + "learning_rate": 6.162349231642033e-06, + "loss": 0.4445, + "step": 2135 + }, + { + "epoch": 4.196463654223969, + "grad_norm": 0.11188880504655846, + "learning_rate": 6.133117568475029e-06, + "loss": 0.4432, + "step": 2136 + }, + { + "epoch": 4.198428290766208, + "grad_norm": 0.11077676984309726, + "learning_rate": 6.1039496441153415e-06, + "loss": 0.443, + "step": 2137 + }, + { + "epoch": 4.200392927308448, + "grad_norm": 0.10820542221956783, + "learning_rate": 6.074845513458165e-06, + "loss": 0.436, + "step": 2138 + }, + { + "epoch": 4.202357563850688, + "grad_norm": 0.11049412298111368, + "learning_rate": 6.04580523127861e-06, + "loss": 0.4388, + "step": 2139 + }, + { + "epoch": 4.2043222003929275, + "grad_norm": 0.09570009710875935, + "learning_rate": 6.016828852231649e-06, + "loss": 0.4468, + "step": 2140 + }, + { + "epoch": 4.206286836935167, + "grad_norm": 0.09949327341826474, + "learning_rate": 5.987916430851944e-06, + "loss": 0.4483, + "step": 2141 + }, + { + "epoch": 4.208251473477406, + "grad_norm": 0.09877757192317552, + "learning_rate": 5.959068021553824e-06, + "loss": 0.4393, + "step": 2142 + }, + { + "epoch": 4.210216110019647, + "grad_norm": 0.09917719318844953, + "learning_rate": 5.930283678631137e-06, + "loss": 0.4377, + "step": 2143 + }, + { + "epoch": 4.212180746561886, + "grad_norm": 0.09724433512390349, + "learning_rate": 5.901563456257142e-06, + "loss": 0.4398, + "step": 2144 + }, + { + "epoch": 4.214145383104126, + "grad_norm": 0.09800182449754223, + "learning_rate": 5.87290740848443e-06, + "loss": 0.4423, + "step": 2145 + }, + { + "epoch": 4.216110019646365, + "grad_norm": 0.09808281635423224, + "learning_rate": 5.8443155892448135e-06, + "loss": 0.4331, + "step": 2146 + }, + { + "epoch": 4.218074656188605, + "grad_norm": 0.10133183127601512, + "learning_rate": 5.815788052349223e-06, + "loss": 0.4416, + "step": 2147 + }, + { + "epoch": 4.220039292730845, + "grad_norm": 0.10147874510521823, + "learning_rate": 5.787324851487612e-06, + "loss": 0.444, + "step": 2148 + }, + { + "epoch": 4.222003929273084, + "grad_norm": 0.09691927253171907, + "learning_rate": 5.758926040228847e-06, + "loss": 0.4418, + "step": 2149 + }, + { + "epoch": 4.2239685658153245, + "grad_norm": 0.10269350219619096, + "learning_rate": 5.7305916720206e-06, + "loss": 0.4353, + "step": 2150 + }, + { + "epoch": 4.225933202357564, + "grad_norm": 0.09855413581360566, + "learning_rate": 5.702321800189272e-06, + "loss": 0.4466, + "step": 2151 + }, + { + "epoch": 4.227897838899803, + "grad_norm": 0.09686964333254855, + "learning_rate": 5.674116477939877e-06, + "loss": 0.4388, + "step": 2152 + }, + { + "epoch": 4.229862475442043, + "grad_norm": 0.10011924116895811, + "learning_rate": 5.645975758355957e-06, + "loss": 0.4324, + "step": 2153 + }, + { + "epoch": 4.231827111984283, + "grad_norm": 0.10056439731338904, + "learning_rate": 5.617899694399427e-06, + "loss": 0.4486, + "step": 2154 + }, + { + "epoch": 4.233791748526523, + "grad_norm": 0.0943129593710493, + "learning_rate": 5.589888338910565e-06, + "loss": 0.4436, + "step": 2155 + }, + { + "epoch": 4.235756385068762, + "grad_norm": 0.09219802569739646, + "learning_rate": 5.5619417446078375e-06, + "loss": 0.4325, + "step": 2156 + }, + { + "epoch": 4.237721021611002, + "grad_norm": 0.10009969626142431, + "learning_rate": 5.534059964087841e-06, + "loss": 0.4454, + "step": 2157 + }, + { + "epoch": 4.239685658153242, + "grad_norm": 0.09909850415481683, + "learning_rate": 5.5062430498251885e-06, + "loss": 0.442, + "step": 2158 + }, + { + "epoch": 4.241650294695481, + "grad_norm": 0.09852209561826329, + "learning_rate": 5.47849105417241e-06, + "loss": 0.4405, + "step": 2159 + }, + { + "epoch": 4.243614931237721, + "grad_norm": 0.09494445147113635, + "learning_rate": 5.450804029359842e-06, + "loss": 0.4355, + "step": 2160 + }, + { + "epoch": 4.245579567779961, + "grad_norm": 0.09996885610392373, + "learning_rate": 5.423182027495562e-06, + "loss": 0.4359, + "step": 2161 + }, + { + "epoch": 4.2475442043222005, + "grad_norm": 0.10032690551771707, + "learning_rate": 5.395625100565278e-06, + "loss": 0.4451, + "step": 2162 + }, + { + "epoch": 4.24950884086444, + "grad_norm": 0.10333372312970512, + "learning_rate": 5.36813330043219e-06, + "loss": 0.4402, + "step": 2163 + }, + { + "epoch": 4.251473477406679, + "grad_norm": 0.10314278931571536, + "learning_rate": 5.340706678836966e-06, + "loss": 0.4429, + "step": 2164 + }, + { + "epoch": 4.25343811394892, + "grad_norm": 0.09614931323007002, + "learning_rate": 5.313345287397585e-06, + "loss": 0.4471, + "step": 2165 + }, + { + "epoch": 4.255402750491159, + "grad_norm": 0.10092035775881958, + "learning_rate": 5.286049177609256e-06, + "loss": 0.4387, + "step": 2166 + }, + { + "epoch": 4.257367387033399, + "grad_norm": 0.1097042440966823, + "learning_rate": 5.2588184008443455e-06, + "loss": 0.4426, + "step": 2167 + }, + { + "epoch": 4.259332023575639, + "grad_norm": 0.09890445549488247, + "learning_rate": 5.231653008352245e-06, + "loss": 0.4461, + "step": 2168 + }, + { + "epoch": 4.261296660117878, + "grad_norm": 0.10224145284229123, + "learning_rate": 5.204553051259296e-06, + "loss": 0.4452, + "step": 2169 + }, + { + "epoch": 4.263261296660118, + "grad_norm": 0.10016904126100445, + "learning_rate": 5.177518580568674e-06, + "loss": 0.4447, + "step": 2170 + }, + { + "epoch": 4.265225933202357, + "grad_norm": 0.09909141749671603, + "learning_rate": 5.15054964716033e-06, + "loss": 0.4406, + "step": 2171 + }, + { + "epoch": 4.2671905697445975, + "grad_norm": 0.0982479467701367, + "learning_rate": 5.123646301790861e-06, + "loss": 0.4495, + "step": 2172 + }, + { + "epoch": 4.269155206286837, + "grad_norm": 0.10171847963019827, + "learning_rate": 5.096808595093406e-06, + "loss": 0.44, + "step": 2173 + }, + { + "epoch": 4.271119842829076, + "grad_norm": 0.09881709600926032, + "learning_rate": 5.070036577577599e-06, + "loss": 0.4359, + "step": 2174 + }, + { + "epoch": 4.273084479371317, + "grad_norm": 0.10382569839388205, + "learning_rate": 5.0433302996294365e-06, + "loss": 0.453, + "step": 2175 + }, + { + "epoch": 4.275049115913556, + "grad_norm": 0.10083836168007418, + "learning_rate": 5.016689811511173e-06, + "loss": 0.4432, + "step": 2176 + }, + { + "epoch": 4.277013752455796, + "grad_norm": 0.10034845893672231, + "learning_rate": 4.990115163361271e-06, + "loss": 0.4404, + "step": 2177 + }, + { + "epoch": 4.278978388998035, + "grad_norm": 0.10265636724524191, + "learning_rate": 4.9636064051942616e-06, + "loss": 0.4438, + "step": 2178 + }, + { + "epoch": 4.280943025540275, + "grad_norm": 0.09687627332381989, + "learning_rate": 4.937163586900662e-06, + "loss": 0.4374, + "step": 2179 + }, + { + "epoch": 4.282907662082515, + "grad_norm": 0.09745689531812134, + "learning_rate": 4.9107867582469085e-06, + "loss": 0.4339, + "step": 2180 + }, + { + "epoch": 4.284872298624754, + "grad_norm": 0.10522251898645671, + "learning_rate": 4.884475968875229e-06, + "loss": 0.4515, + "step": 2181 + }, + { + "epoch": 4.286836935166994, + "grad_norm": 0.09450880142546599, + "learning_rate": 4.858231268303572e-06, + "loss": 0.4293, + "step": 2182 + }, + { + "epoch": 4.288801571709234, + "grad_norm": 0.096661949234695, + "learning_rate": 4.832052705925487e-06, + "loss": 0.4485, + "step": 2183 + }, + { + "epoch": 4.2907662082514735, + "grad_norm": 0.0964886896852349, + "learning_rate": 4.8059403310100635e-06, + "loss": 0.4484, + "step": 2184 + }, + { + "epoch": 4.292730844793713, + "grad_norm": 0.09999729355800414, + "learning_rate": 4.779894192701826e-06, + "loss": 0.448, + "step": 2185 + }, + { + "epoch": 4.294695481335953, + "grad_norm": 0.10340157408907742, + "learning_rate": 4.753914340020625e-06, + "loss": 0.4466, + "step": 2186 + }, + { + "epoch": 4.296660117878193, + "grad_norm": 0.09749860851785701, + "learning_rate": 4.728000821861574e-06, + "loss": 0.4401, + "step": 2187 + }, + { + "epoch": 4.298624754420432, + "grad_norm": 0.09492479211389322, + "learning_rate": 4.702153686994941e-06, + "loss": 0.4353, + "step": 2188 + }, + { + "epoch": 4.300589390962672, + "grad_norm": 0.09817261786156144, + "learning_rate": 4.676372984066033e-06, + "loss": 0.4374, + "step": 2189 + }, + { + "epoch": 4.302554027504912, + "grad_norm": 0.10364902873028085, + "learning_rate": 4.650658761595162e-06, + "loss": 0.4379, + "step": 2190 + }, + { + "epoch": 4.304518664047151, + "grad_norm": 0.09398675486516612, + "learning_rate": 4.625011067977516e-06, + "loss": 0.4514, + "step": 2191 + }, + { + "epoch": 4.306483300589391, + "grad_norm": 0.09563039401883067, + "learning_rate": 4.599429951483041e-06, + "loss": 0.4339, + "step": 2192 + }, + { + "epoch": 4.30844793713163, + "grad_norm": 0.09876649223703932, + "learning_rate": 4.573915460256424e-06, + "loss": 0.4452, + "step": 2193 + }, + { + "epoch": 4.3104125736738705, + "grad_norm": 0.10491894932061191, + "learning_rate": 4.548467642316943e-06, + "loss": 0.4467, + "step": 2194 + }, + { + "epoch": 4.31237721021611, + "grad_norm": 0.0952170837132662, + "learning_rate": 4.523086545558384e-06, + "loss": 0.4493, + "step": 2195 + }, + { + "epoch": 4.314341846758349, + "grad_norm": 0.0949882643938253, + "learning_rate": 4.497772217748981e-06, + "loss": 0.4403, + "step": 2196 + }, + { + "epoch": 4.31630648330059, + "grad_norm": 0.09898286124209162, + "learning_rate": 4.472524706531292e-06, + "loss": 0.4453, + "step": 2197 + }, + { + "epoch": 4.318271119842829, + "grad_norm": 0.09182050800859727, + "learning_rate": 4.4473440594221405e-06, + "loss": 0.4435, + "step": 2198 + }, + { + "epoch": 4.320235756385069, + "grad_norm": 0.10898852889019824, + "learning_rate": 4.422230323812473e-06, + "loss": 0.4414, + "step": 2199 + }, + { + "epoch": 4.322200392927308, + "grad_norm": 0.09447219772414896, + "learning_rate": 4.397183546967343e-06, + "loss": 0.4376, + "step": 2200 + }, + { + "epoch": 4.324165029469548, + "grad_norm": 0.09351405858793416, + "learning_rate": 4.372203776025781e-06, + "loss": 0.4445, + "step": 2201 + }, + { + "epoch": 4.326129666011788, + "grad_norm": 0.09614912300047074, + "learning_rate": 4.347291058000678e-06, + "loss": 0.4426, + "step": 2202 + }, + { + "epoch": 4.328094302554027, + "grad_norm": 0.09370015584081862, + "learning_rate": 4.322445439778764e-06, + "loss": 0.4417, + "step": 2203 + }, + { + "epoch": 4.330058939096268, + "grad_norm": 0.0930531030094998, + "learning_rate": 4.297666968120471e-06, + "loss": 0.4446, + "step": 2204 + }, + { + "epoch": 4.332023575638507, + "grad_norm": 0.09679977391824321, + "learning_rate": 4.272955689659859e-06, + "loss": 0.453, + "step": 2205 + }, + { + "epoch": 4.3339882121807465, + "grad_norm": 0.09301112612616434, + "learning_rate": 4.2483116509045306e-06, + "loss": 0.4404, + "step": 2206 + }, + { + "epoch": 4.335952848722986, + "grad_norm": 0.09422118662590862, + "learning_rate": 4.223734898235536e-06, + "loss": 0.4371, + "step": 2207 + }, + { + "epoch": 4.337917485265226, + "grad_norm": 0.10024020110602878, + "learning_rate": 4.199225477907285e-06, + "loss": 0.4417, + "step": 2208 + }, + { + "epoch": 4.339882121807466, + "grad_norm": 0.09154428867109872, + "learning_rate": 4.174783436047483e-06, + "loss": 0.4464, + "step": 2209 + }, + { + "epoch": 4.341846758349705, + "grad_norm": 0.09288789216074118, + "learning_rate": 4.150408818657008e-06, + "loss": 0.4415, + "step": 2210 + }, + { + "epoch": 4.3438113948919455, + "grad_norm": 0.0966383008923141, + "learning_rate": 4.126101671609859e-06, + "loss": 0.4365, + "step": 2211 + }, + { + "epoch": 4.345776031434185, + "grad_norm": 0.10134056951677198, + "learning_rate": 4.101862040653033e-06, + "loss": 0.4411, + "step": 2212 + }, + { + "epoch": 4.347740667976424, + "grad_norm": 0.09780319882004399, + "learning_rate": 4.077689971406478e-06, + "loss": 0.4399, + "step": 2213 + }, + { + "epoch": 4.349705304518664, + "grad_norm": 0.093298710095767, + "learning_rate": 4.05358550936298e-06, + "loss": 0.4348, + "step": 2214 + }, + { + "epoch": 4.351669941060904, + "grad_norm": 0.09566340223812082, + "learning_rate": 4.0295486998880885e-06, + "loss": 0.4494, + "step": 2215 + }, + { + "epoch": 4.3536345776031435, + "grad_norm": 0.09308399435415046, + "learning_rate": 4.005579588220023e-06, + "loss": 0.4426, + "step": 2216 + }, + { + "epoch": 4.355599214145383, + "grad_norm": 0.1009287820158751, + "learning_rate": 3.981678219469602e-06, + "loss": 0.4451, + "step": 2217 + }, + { + "epoch": 4.357563850687622, + "grad_norm": 0.09292643293561982, + "learning_rate": 3.957844638620132e-06, + "loss": 0.4399, + "step": 2218 + }, + { + "epoch": 4.359528487229863, + "grad_norm": 0.08967035745466466, + "learning_rate": 3.934078890527357e-06, + "loss": 0.4452, + "step": 2219 + }, + { + "epoch": 4.361493123772102, + "grad_norm": 0.10006137217528763, + "learning_rate": 3.9103810199193625e-06, + "loss": 0.4389, + "step": 2220 + }, + { + "epoch": 4.363457760314342, + "grad_norm": 0.09466623450521533, + "learning_rate": 3.886751071396457e-06, + "loss": 0.4415, + "step": 2221 + }, + { + "epoch": 4.365422396856582, + "grad_norm": 0.09518504458814794, + "learning_rate": 3.863189089431143e-06, + "loss": 0.4437, + "step": 2222 + }, + { + "epoch": 4.367387033398821, + "grad_norm": 0.09606883064031116, + "learning_rate": 3.839695118367996e-06, + "loss": 0.4412, + "step": 2223 + }, + { + "epoch": 4.369351669941061, + "grad_norm": 0.09528200994038874, + "learning_rate": 3.8162692024235996e-06, + "loss": 0.4414, + "step": 2224 + }, + { + "epoch": 4.3713163064833, + "grad_norm": 0.09394723295140671, + "learning_rate": 3.7929113856864485e-06, + "loss": 0.4365, + "step": 2225 + }, + { + "epoch": 4.373280943025541, + "grad_norm": 0.08849201152831374, + "learning_rate": 3.7696217121168777e-06, + "loss": 0.4342, + "step": 2226 + }, + { + "epoch": 4.37524557956778, + "grad_norm": 0.09760744732278162, + "learning_rate": 3.7464002255469712e-06, + "loss": 0.4441, + "step": 2227 + }, + { + "epoch": 4.3772102161100195, + "grad_norm": 0.09457240474414914, + "learning_rate": 3.7232469696804718e-06, + "loss": 0.4499, + "step": 2228 + }, + { + "epoch": 4.379174852652259, + "grad_norm": 0.09214654026062598, + "learning_rate": 3.7001619880927276e-06, + "loss": 0.4397, + "step": 2229 + }, + { + "epoch": 4.381139489194499, + "grad_norm": 0.09524429099135905, + "learning_rate": 3.6771453242305886e-06, + "loss": 0.4436, + "step": 2230 + }, + { + "epoch": 4.383104125736739, + "grad_norm": 0.09746242185335974, + "learning_rate": 3.6541970214123066e-06, + "loss": 0.4456, + "step": 2231 + }, + { + "epoch": 4.385068762278978, + "grad_norm": 0.09304220670442669, + "learning_rate": 3.6313171228275066e-06, + "loss": 0.4442, + "step": 2232 + }, + { + "epoch": 4.3870333988212185, + "grad_norm": 0.09374529865284163, + "learning_rate": 3.6085056715370504e-06, + "loss": 0.4423, + "step": 2233 + }, + { + "epoch": 4.388998035363458, + "grad_norm": 0.09981105953101409, + "learning_rate": 3.585762710472995e-06, + "loss": 0.442, + "step": 2234 + }, + { + "epoch": 4.390962671905697, + "grad_norm": 0.09834973012338746, + "learning_rate": 3.5630882824384893e-06, + "loss": 0.4509, + "step": 2235 + }, + { + "epoch": 4.392927308447937, + "grad_norm": 0.09346097473270883, + "learning_rate": 3.5404824301076946e-06, + "loss": 0.4375, + "step": 2236 + }, + { + "epoch": 4.394891944990177, + "grad_norm": 0.0934957321717044, + "learning_rate": 3.517945196025725e-06, + "loss": 0.4455, + "step": 2237 + }, + { + "epoch": 4.3968565815324165, + "grad_norm": 0.09589813370139916, + "learning_rate": 3.4954766226085355e-06, + "loss": 0.4405, + "step": 2238 + }, + { + "epoch": 4.398821218074656, + "grad_norm": 0.09166308076645979, + "learning_rate": 3.4730767521428654e-06, + "loss": 0.4511, + "step": 2239 + }, + { + "epoch": 4.400785854616895, + "grad_norm": 0.08869240905006665, + "learning_rate": 3.4507456267861605e-06, + "loss": 0.4426, + "step": 2240 + }, + { + "epoch": 4.402750491159136, + "grad_norm": 0.09717104371981981, + "learning_rate": 3.428483288566473e-06, + "loss": 0.4464, + "step": 2241 + }, + { + "epoch": 4.404715127701375, + "grad_norm": 0.09041712851455892, + "learning_rate": 3.406289779382399e-06, + "loss": 0.435, + "step": 2242 + }, + { + "epoch": 4.406679764243615, + "grad_norm": 0.09125388493669863, + "learning_rate": 3.3841651410030063e-06, + "loss": 0.4374, + "step": 2243 + }, + { + "epoch": 4.408644400785855, + "grad_norm": 0.09312894180981898, + "learning_rate": 3.362109415067725e-06, + "loss": 0.4388, + "step": 2244 + }, + { + "epoch": 4.410609037328094, + "grad_norm": 0.09317066333465474, + "learning_rate": 3.3401226430863098e-06, + "loss": 0.4409, + "step": 2245 + }, + { + "epoch": 4.412573673870334, + "grad_norm": 0.09420181671124139, + "learning_rate": 3.3182048664387367e-06, + "loss": 0.4379, + "step": 2246 + }, + { + "epoch": 4.414538310412573, + "grad_norm": 0.09409718456956674, + "learning_rate": 3.2963561263751066e-06, + "loss": 0.4461, + "step": 2247 + }, + { + "epoch": 4.416502946954814, + "grad_norm": 0.09466112847944505, + "learning_rate": 3.27457646401562e-06, + "loss": 0.4499, + "step": 2248 + }, + { + "epoch": 4.418467583497053, + "grad_norm": 0.0949966428046833, + "learning_rate": 3.252865920350461e-06, + "loss": 0.4475, + "step": 2249 + }, + { + "epoch": 4.4204322200392925, + "grad_norm": 0.0965547572140136, + "learning_rate": 3.231224536239732e-06, + "loss": 0.4483, + "step": 2250 + }, + { + "epoch": 4.422396856581533, + "grad_norm": 0.09103391051506604, + "learning_rate": 3.209652352413359e-06, + "loss": 0.4532, + "step": 2251 + }, + { + "epoch": 4.424361493123772, + "grad_norm": 0.09660635192844322, + "learning_rate": 3.1881494094710486e-06, + "loss": 0.4326, + "step": 2252 + }, + { + "epoch": 4.426326129666012, + "grad_norm": 0.09407458747600667, + "learning_rate": 3.1667157478821875e-06, + "loss": 0.4391, + "step": 2253 + }, + { + "epoch": 4.428290766208251, + "grad_norm": 0.08903345751219295, + "learning_rate": 3.1453514079857706e-06, + "loss": 0.4344, + "step": 2254 + }, + { + "epoch": 4.4302554027504915, + "grad_norm": 0.0923279073826575, + "learning_rate": 3.1240564299903277e-06, + "loss": 0.4469, + "step": 2255 + }, + { + "epoch": 4.432220039292731, + "grad_norm": 0.09380173871504252, + "learning_rate": 3.1028308539738526e-06, + "loss": 0.439, + "step": 2256 + }, + { + "epoch": 4.43418467583497, + "grad_norm": 0.09749206003405829, + "learning_rate": 3.0816747198837026e-06, + "loss": 0.4508, + "step": 2257 + }, + { + "epoch": 4.436149312377211, + "grad_norm": 0.09041895837046535, + "learning_rate": 3.0605880675365606e-06, + "loss": 0.4476, + "step": 2258 + }, + { + "epoch": 4.43811394891945, + "grad_norm": 0.09141047988174399, + "learning_rate": 3.0395709366183477e-06, + "loss": 0.4412, + "step": 2259 + }, + { + "epoch": 4.4400785854616895, + "grad_norm": 0.09862339367301928, + "learning_rate": 3.018623366684117e-06, + "loss": 0.4451, + "step": 2260 + }, + { + "epoch": 4.442043222003929, + "grad_norm": 0.0997667919980629, + "learning_rate": 2.9977453971580296e-06, + "loss": 0.439, + "step": 2261 + }, + { + "epoch": 4.444007858546169, + "grad_norm": 0.0975335418464718, + "learning_rate": 2.976937067333241e-06, + "loss": 0.4335, + "step": 2262 + }, + { + "epoch": 4.445972495088409, + "grad_norm": 0.09468640645498533, + "learning_rate": 2.956198416371856e-06, + "loss": 0.4476, + "step": 2263 + }, + { + "epoch": 4.447937131630648, + "grad_norm": 0.09771552792602614, + "learning_rate": 2.9355294833048264e-06, + "loss": 0.4425, + "step": 2264 + }, + { + "epoch": 4.449901768172888, + "grad_norm": 0.09988354915551544, + "learning_rate": 2.9149303070319024e-06, + "loss": 0.436, + "step": 2265 + }, + { + "epoch": 4.451866404715128, + "grad_norm": 0.08916336302325194, + "learning_rate": 2.8944009263215523e-06, + "loss": 0.4409, + "step": 2266 + }, + { + "epoch": 4.453831041257367, + "grad_norm": 0.09474394408999887, + "learning_rate": 2.873941379810865e-06, + "loss": 0.4423, + "step": 2267 + }, + { + "epoch": 4.455795677799607, + "grad_norm": 0.10361586719123533, + "learning_rate": 2.853551706005515e-06, + "loss": 0.4449, + "step": 2268 + }, + { + "epoch": 4.457760314341847, + "grad_norm": 0.09106795914621058, + "learning_rate": 2.8332319432796905e-06, + "loss": 0.4337, + "step": 2269 + }, + { + "epoch": 4.459724950884087, + "grad_norm": 0.09180163697742676, + "learning_rate": 2.8129821298759607e-06, + "loss": 0.442, + "step": 2270 + }, + { + "epoch": 4.461689587426326, + "grad_norm": 0.09242447670784805, + "learning_rate": 2.7928023039052844e-06, + "loss": 0.4384, + "step": 2271 + }, + { + "epoch": 4.4636542239685655, + "grad_norm": 0.09455438300414364, + "learning_rate": 2.77269250334689e-06, + "loss": 0.4425, + "step": 2272 + }, + { + "epoch": 4.465618860510806, + "grad_norm": 0.08833960694905114, + "learning_rate": 2.7526527660482096e-06, + "loss": 0.4455, + "step": 2273 + }, + { + "epoch": 4.467583497053045, + "grad_norm": 0.09789827690072947, + "learning_rate": 2.7326831297248245e-06, + "loss": 0.4486, + "step": 2274 + }, + { + "epoch": 4.469548133595285, + "grad_norm": 0.10096293028302188, + "learning_rate": 2.712783631960383e-06, + "loss": 0.4514, + "step": 2275 + }, + { + "epoch": 4.471512770137524, + "grad_norm": 0.09037264038914417, + "learning_rate": 2.6929543102065083e-06, + "loss": 0.4487, + "step": 2276 + }, + { + "epoch": 4.4734774066797645, + "grad_norm": 0.09412073017988594, + "learning_rate": 2.67319520178277e-06, + "loss": 0.4449, + "step": 2277 + }, + { + "epoch": 4.475442043222004, + "grad_norm": 0.10478799448909454, + "learning_rate": 2.653506343876595e-06, + "loss": 0.4443, + "step": 2278 + }, + { + "epoch": 4.477406679764243, + "grad_norm": 0.0918989414722328, + "learning_rate": 2.633887773543191e-06, + "loss": 0.4492, + "step": 2279 + }, + { + "epoch": 4.479371316306484, + "grad_norm": 0.09155584943859593, + "learning_rate": 2.6143395277054717e-06, + "loss": 0.4401, + "step": 2280 + }, + { + "epoch": 4.481335952848723, + "grad_norm": 0.08898862010844427, + "learning_rate": 2.5948616431540073e-06, + "loss": 0.4409, + "step": 2281 + }, + { + "epoch": 4.4833005893909625, + "grad_norm": 0.11131612113526604, + "learning_rate": 2.5754541565469502e-06, + "loss": 0.4506, + "step": 2282 + }, + { + "epoch": 4.485265225933202, + "grad_norm": 0.09527085142878736, + "learning_rate": 2.5561171044099497e-06, + "loss": 0.4489, + "step": 2283 + }, + { + "epoch": 4.487229862475442, + "grad_norm": 0.08963818670758296, + "learning_rate": 2.536850523136103e-06, + "loss": 0.4396, + "step": 2284 + }, + { + "epoch": 4.489194499017682, + "grad_norm": 0.09596774280508374, + "learning_rate": 2.5176544489858803e-06, + "loss": 0.4578, + "step": 2285 + }, + { + "epoch": 4.491159135559921, + "grad_norm": 0.09529252208011356, + "learning_rate": 2.4985289180870356e-06, + "loss": 0.4425, + "step": 2286 + }, + { + "epoch": 4.493123772102161, + "grad_norm": 0.09193115628195245, + "learning_rate": 2.4794739664345846e-06, + "loss": 0.4388, + "step": 2287 + }, + { + "epoch": 4.495088408644401, + "grad_norm": 0.08830260316426848, + "learning_rate": 2.4604896298906945e-06, + "loss": 0.4422, + "step": 2288 + }, + { + "epoch": 4.49705304518664, + "grad_norm": 0.10613372470986067, + "learning_rate": 2.4415759441846287e-06, + "loss": 0.4438, + "step": 2289 + }, + { + "epoch": 4.49901768172888, + "grad_norm": 0.10564338502374511, + "learning_rate": 2.4227329449126913e-06, + "loss": 0.4462, + "step": 2290 + }, + { + "epoch": 4.50098231827112, + "grad_norm": 0.09138134016829641, + "learning_rate": 2.4039606675381454e-06, + "loss": 0.4422, + "step": 2291 + }, + { + "epoch": 4.50294695481336, + "grad_norm": 0.0876223178935145, + "learning_rate": 2.385259147391179e-06, + "loss": 0.4523, + "step": 2292 + }, + { + "epoch": 4.504911591355599, + "grad_norm": 0.09380336316031762, + "learning_rate": 2.3666284196687615e-06, + "loss": 0.4397, + "step": 2293 + }, + { + "epoch": 4.506876227897839, + "grad_norm": 0.0906468074350511, + "learning_rate": 2.3480685194346717e-06, + "loss": 0.434, + "step": 2294 + }, + { + "epoch": 4.508840864440079, + "grad_norm": 0.0984240732465253, + "learning_rate": 2.3295794816193773e-06, + "loss": 0.4497, + "step": 2295 + }, + { + "epoch": 4.510805500982318, + "grad_norm": 0.09153859104253585, + "learning_rate": 2.3111613410199585e-06, + "loss": 0.4382, + "step": 2296 + }, + { + "epoch": 4.512770137524558, + "grad_norm": 0.08841396408703797, + "learning_rate": 2.2928141323000965e-06, + "loss": 0.4442, + "step": 2297 + }, + { + "epoch": 4.514734774066798, + "grad_norm": 0.09882255733564546, + "learning_rate": 2.2745378899899605e-06, + "loss": 0.4364, + "step": 2298 + }, + { + "epoch": 4.5166994106090375, + "grad_norm": 0.08823243338763993, + "learning_rate": 2.2563326484861526e-06, + "loss": 0.431, + "step": 2299 + }, + { + "epoch": 4.518664047151277, + "grad_norm": 0.09192426803828922, + "learning_rate": 2.2381984420516555e-06, + "loss": 0.4517, + "step": 2300 + }, + { + "epoch": 4.520628683693516, + "grad_norm": 0.09429173941054349, + "learning_rate": 2.2201353048157604e-06, + "loss": 0.4387, + "step": 2301 + }, + { + "epoch": 4.522593320235757, + "grad_norm": 0.09904142614747488, + "learning_rate": 2.2021432707740065e-06, + "loss": 0.4383, + "step": 2302 + }, + { + "epoch": 4.524557956777996, + "grad_norm": 0.10130531789195918, + "learning_rate": 2.1842223737881073e-06, + "loss": 0.4404, + "step": 2303 + }, + { + "epoch": 4.5265225933202355, + "grad_norm": 0.08894964327293883, + "learning_rate": 2.1663726475859017e-06, + "loss": 0.4425, + "step": 2304 + }, + { + "epoch": 4.528487229862476, + "grad_norm": 0.09240581493114322, + "learning_rate": 2.148594125761281e-06, + "loss": 0.4379, + "step": 2305 + }, + { + "epoch": 4.530451866404715, + "grad_norm": 0.0939783158014248, + "learning_rate": 2.1308868417741113e-06, + "loss": 0.4453, + "step": 2306 + }, + { + "epoch": 4.532416502946955, + "grad_norm": 0.09457354982588836, + "learning_rate": 2.1132508289502063e-06, + "loss": 0.4468, + "step": 2307 + }, + { + "epoch": 4.534381139489194, + "grad_norm": 0.09298748078072044, + "learning_rate": 2.095686120481242e-06, + "loss": 0.4396, + "step": 2308 + }, + { + "epoch": 4.5363457760314345, + "grad_norm": 0.09436081996738403, + "learning_rate": 2.0781927494246813e-06, + "loss": 0.4449, + "step": 2309 + }, + { + "epoch": 4.538310412573674, + "grad_norm": 0.09240957534792582, + "learning_rate": 2.06077074870374e-06, + "loss": 0.4381, + "step": 2310 + }, + { + "epoch": 4.540275049115913, + "grad_norm": 0.09167195792702142, + "learning_rate": 2.0434201511073184e-06, + "loss": 0.4474, + "step": 2311 + }, + { + "epoch": 4.542239685658153, + "grad_norm": 0.089769576382118, + "learning_rate": 2.0261409892899177e-06, + "loss": 0.4375, + "step": 2312 + }, + { + "epoch": 4.544204322200393, + "grad_norm": 0.09299121586234627, + "learning_rate": 2.0089332957716044e-06, + "loss": 0.445, + "step": 2313 + }, + { + "epoch": 4.546168958742633, + "grad_norm": 0.09760348410969803, + "learning_rate": 1.9917971029379403e-06, + "loss": 0.4443, + "step": 2314 + }, + { + "epoch": 4.548133595284872, + "grad_norm": 0.093881885865992, + "learning_rate": 1.9747324430398995e-06, + "loss": 0.4458, + "step": 2315 + }, + { + "epoch": 4.550098231827112, + "grad_norm": 0.09569758211943898, + "learning_rate": 1.957739348193859e-06, + "loss": 0.4433, + "step": 2316 + }, + { + "epoch": 4.552062868369352, + "grad_norm": 0.09031705532712961, + "learning_rate": 1.9408178503814933e-06, + "loss": 0.4438, + "step": 2317 + }, + { + "epoch": 4.554027504911591, + "grad_norm": 0.08851908882692641, + "learning_rate": 1.9239679814497194e-06, + "loss": 0.4355, + "step": 2318 + }, + { + "epoch": 4.555992141453831, + "grad_norm": 0.0958534890467064, + "learning_rate": 1.9071897731106536e-06, + "loss": 0.4424, + "step": 2319 + }, + { + "epoch": 4.557956777996071, + "grad_norm": 0.09774375691046018, + "learning_rate": 1.8904832569415487e-06, + "loss": 0.4535, + "step": 2320 + }, + { + "epoch": 4.5599214145383105, + "grad_norm": 0.09128906312053554, + "learning_rate": 1.873848464384742e-06, + "loss": 0.437, + "step": 2321 + }, + { + "epoch": 4.56188605108055, + "grad_norm": 0.0904081520943023, + "learning_rate": 1.857285426747546e-06, + "loss": 0.4518, + "step": 2322 + }, + { + "epoch": 4.563850687622789, + "grad_norm": 0.08947533379446683, + "learning_rate": 1.8407941752022562e-06, + "loss": 0.4469, + "step": 2323 + }, + { + "epoch": 4.56581532416503, + "grad_norm": 0.09013854062438491, + "learning_rate": 1.8243747407860723e-06, + "loss": 0.4389, + "step": 2324 + }, + { + "epoch": 4.567779960707269, + "grad_norm": 0.09364075635107004, + "learning_rate": 1.8080271544009953e-06, + "loss": 0.4459, + "step": 2325 + }, + { + "epoch": 4.5697445972495085, + "grad_norm": 0.0907508245327325, + "learning_rate": 1.7917514468138409e-06, + "loss": 0.4491, + "step": 2326 + }, + { + "epoch": 4.571709233791749, + "grad_norm": 0.09093727249742344, + "learning_rate": 1.7755476486561285e-06, + "loss": 0.4404, + "step": 2327 + }, + { + "epoch": 4.573673870333988, + "grad_norm": 0.09015284611261098, + "learning_rate": 1.7594157904240416e-06, + "loss": 0.4426, + "step": 2328 + }, + { + "epoch": 4.575638506876228, + "grad_norm": 0.09357362922569408, + "learning_rate": 1.7433559024783697e-06, + "loss": 0.442, + "step": 2329 + }, + { + "epoch": 4.577603143418468, + "grad_norm": 0.09209956659300036, + "learning_rate": 1.7273680150444727e-06, + "loss": 0.4407, + "step": 2330 + }, + { + "epoch": 4.5795677799607075, + "grad_norm": 0.08832751364526356, + "learning_rate": 1.7114521582121702e-06, + "loss": 0.4362, + "step": 2331 + }, + { + "epoch": 4.581532416502947, + "grad_norm": 0.09051901712120458, + "learning_rate": 1.6956083619357322e-06, + "loss": 0.4411, + "step": 2332 + }, + { + "epoch": 4.583497053045186, + "grad_norm": 0.0884378736395962, + "learning_rate": 1.6798366560338131e-06, + "loss": 0.4439, + "step": 2333 + }, + { + "epoch": 4.585461689587426, + "grad_norm": 0.09013189863028269, + "learning_rate": 1.6641370701893844e-06, + "loss": 0.4386, + "step": 2334 + }, + { + "epoch": 4.587426326129666, + "grad_norm": 0.09457893760449976, + "learning_rate": 1.6485096339496775e-06, + "loss": 0.437, + "step": 2335 + }, + { + "epoch": 4.589390962671906, + "grad_norm": 0.09319414953085187, + "learning_rate": 1.632954376726148e-06, + "loss": 0.4389, + "step": 2336 + }, + { + "epoch": 4.591355599214145, + "grad_norm": 0.08684214500080513, + "learning_rate": 1.6174713277944132e-06, + "loss": 0.4416, + "step": 2337 + }, + { + "epoch": 4.593320235756385, + "grad_norm": 0.0861927751456855, + "learning_rate": 1.6020605162941637e-06, + "loss": 0.4451, + "step": 2338 + }, + { + "epoch": 4.595284872298625, + "grad_norm": 0.09020739853167395, + "learning_rate": 1.5867219712291592e-06, + "loss": 0.4442, + "step": 2339 + }, + { + "epoch": 4.597249508840864, + "grad_norm": 0.08927095702426895, + "learning_rate": 1.5714557214671566e-06, + "loss": 0.4482, + "step": 2340 + }, + { + "epoch": 4.599214145383105, + "grad_norm": 0.09453118716138621, + "learning_rate": 1.5562617957398264e-06, + "loss": 0.448, + "step": 2341 + }, + { + "epoch": 4.601178781925344, + "grad_norm": 0.08801774908828154, + "learning_rate": 1.541140222642743e-06, + "loss": 0.4426, + "step": 2342 + }, + { + "epoch": 4.6031434184675835, + "grad_norm": 0.090997104196872, + "learning_rate": 1.526091030635306e-06, + "loss": 0.4398, + "step": 2343 + }, + { + "epoch": 4.605108055009823, + "grad_norm": 0.09358233741922772, + "learning_rate": 1.5111142480406816e-06, + "loss": 0.4398, + "step": 2344 + }, + { + "epoch": 4.607072691552063, + "grad_norm": 0.08658161656180609, + "learning_rate": 1.4962099030457622e-06, + "loss": 0.4385, + "step": 2345 + }, + { + "epoch": 4.609037328094303, + "grad_norm": 0.08701074961352408, + "learning_rate": 1.4813780237011143e-06, + "loss": 0.4399, + "step": 2346 + }, + { + "epoch": 4.611001964636542, + "grad_norm": 0.09508699650719465, + "learning_rate": 1.466618637920929e-06, + "loss": 0.4456, + "step": 2347 + }, + { + "epoch": 4.6129666011787815, + "grad_norm": 0.08999161805364124, + "learning_rate": 1.451931773482933e-06, + "loss": 0.4419, + "step": 2348 + }, + { + "epoch": 4.614931237721022, + "grad_norm": 0.09165307974372763, + "learning_rate": 1.437317458028402e-06, + "loss": 0.4486, + "step": 2349 + }, + { + "epoch": 4.616895874263261, + "grad_norm": 0.08666484683501732, + "learning_rate": 1.422775719062046e-06, + "loss": 0.4485, + "step": 2350 + }, + { + "epoch": 4.618860510805501, + "grad_norm": 0.08494237728424962, + "learning_rate": 1.408306583951995e-06, + "loss": 0.437, + "step": 2351 + }, + { + "epoch": 4.620825147347741, + "grad_norm": 0.08726645767879376, + "learning_rate": 1.3939100799297234e-06, + "loss": 0.441, + "step": 2352 + }, + { + "epoch": 4.6227897838899805, + "grad_norm": 0.09137054098606091, + "learning_rate": 1.379586234090029e-06, + "loss": 0.4322, + "step": 2353 + }, + { + "epoch": 4.62475442043222, + "grad_norm": 0.09185807615325223, + "learning_rate": 1.3653350733909431e-06, + "loss": 0.4464, + "step": 2354 + }, + { + "epoch": 4.626719056974459, + "grad_norm": 0.08866220868470145, + "learning_rate": 1.3511566246537179e-06, + "loss": 0.4472, + "step": 2355 + }, + { + "epoch": 4.6286836935167, + "grad_norm": 0.08632871766652946, + "learning_rate": 1.3370509145627542e-06, + "loss": 0.4369, + "step": 2356 + }, + { + "epoch": 4.630648330058939, + "grad_norm": 0.08681028108353864, + "learning_rate": 1.32301796966555e-06, + "loss": 0.4463, + "step": 2357 + }, + { + "epoch": 4.632612966601179, + "grad_norm": 0.08632883193362295, + "learning_rate": 1.3090578163726587e-06, + "loss": 0.4392, + "step": 2358 + }, + { + "epoch": 4.634577603143418, + "grad_norm": 0.08693078872676903, + "learning_rate": 1.2951704809576503e-06, + "loss": 0.4477, + "step": 2359 + }, + { + "epoch": 4.636542239685658, + "grad_norm": 0.08606943330862722, + "learning_rate": 1.2813559895570316e-06, + "loss": 0.4372, + "step": 2360 + }, + { + "epoch": 4.638506876227898, + "grad_norm": 0.08715756063477327, + "learning_rate": 1.2676143681702136e-06, + "loss": 0.4508, + "step": 2361 + }, + { + "epoch": 4.640471512770137, + "grad_norm": 0.0883461169742, + "learning_rate": 1.2539456426594731e-06, + "loss": 0.4483, + "step": 2362 + }, + { + "epoch": 4.642436149312378, + "grad_norm": 0.08753204672422492, + "learning_rate": 1.2403498387498947e-06, + "loss": 0.4467, + "step": 2363 + }, + { + "epoch": 4.644400785854617, + "grad_norm": 0.09018536739688464, + "learning_rate": 1.226826982029312e-06, + "loss": 0.4454, + "step": 2364 + }, + { + "epoch": 4.6463654223968565, + "grad_norm": 0.08639374623022951, + "learning_rate": 1.213377097948274e-06, + "loss": 0.4396, + "step": 2365 + }, + { + "epoch": 4.648330058939096, + "grad_norm": 0.08816474901994686, + "learning_rate": 1.2000002118199982e-06, + "loss": 0.443, + "step": 2366 + }, + { + "epoch": 4.650294695481336, + "grad_norm": 0.08927216532846033, + "learning_rate": 1.1866963488202975e-06, + "loss": 0.4403, + "step": 2367 + }, + { + "epoch": 4.652259332023576, + "grad_norm": 0.087261640710909, + "learning_rate": 1.1734655339875833e-06, + "loss": 0.4351, + "step": 2368 + }, + { + "epoch": 4.654223968565815, + "grad_norm": 0.08753796225257791, + "learning_rate": 1.160307792222768e-06, + "loss": 0.4398, + "step": 2369 + }, + { + "epoch": 4.6561886051080545, + "grad_norm": 0.08550131200490538, + "learning_rate": 1.1472231482892293e-06, + "loss": 0.4447, + "step": 2370 + }, + { + "epoch": 4.658153241650295, + "grad_norm": 0.0873247331974059, + "learning_rate": 1.134211626812789e-06, + "loss": 0.4494, + "step": 2371 + }, + { + "epoch": 4.660117878192534, + "grad_norm": 0.08924864146360152, + "learning_rate": 1.1212732522816494e-06, + "loss": 0.4525, + "step": 2372 + }, + { + "epoch": 4.662082514734774, + "grad_norm": 0.08736782770665524, + "learning_rate": 1.1084080490463278e-06, + "loss": 0.4423, + "step": 2373 + }, + { + "epoch": 4.664047151277014, + "grad_norm": 0.092233053059242, + "learning_rate": 1.0956160413196558e-06, + "loss": 0.439, + "step": 2374 + }, + { + "epoch": 4.6660117878192535, + "grad_norm": 0.08537642841754556, + "learning_rate": 1.0828972531766868e-06, + "loss": 0.439, + "step": 2375 + }, + { + "epoch": 4.667976424361493, + "grad_norm": 0.08967333642956993, + "learning_rate": 1.0702517085546904e-06, + "loss": 0.4367, + "step": 2376 + }, + { + "epoch": 4.669941060903733, + "grad_norm": 0.0887559549216283, + "learning_rate": 1.0576794312530647e-06, + "loss": 0.4499, + "step": 2377 + }, + { + "epoch": 4.671905697445973, + "grad_norm": 0.09075241149928856, + "learning_rate": 1.0451804449333492e-06, + "loss": 0.4462, + "step": 2378 + }, + { + "epoch": 4.673870333988212, + "grad_norm": 0.08938791617150371, + "learning_rate": 1.0327547731191267e-06, + "loss": 0.4421, + "step": 2379 + }, + { + "epoch": 4.675834970530452, + "grad_norm": 0.0841487808064533, + "learning_rate": 1.0204024391959844e-06, + "loss": 0.4388, + "step": 2380 + }, + { + "epoch": 4.677799607072692, + "grad_norm": 0.08801477619909914, + "learning_rate": 1.008123466411517e-06, + "loss": 0.4479, + "step": 2381 + }, + { + "epoch": 4.679764243614931, + "grad_norm": 0.08575368851208556, + "learning_rate": 9.959178778752343e-07, + "loss": 0.4401, + "step": 2382 + }, + { + "epoch": 4.681728880157171, + "grad_norm": 0.08925844718632052, + "learning_rate": 9.837856965585301e-07, + "loss": 0.4465, + "step": 2383 + }, + { + "epoch": 4.68369351669941, + "grad_norm": 0.08487315906481356, + "learning_rate": 9.717269452946465e-07, + "loss": 0.4413, + "step": 2384 + }, + { + "epoch": 4.685658153241651, + "grad_norm": 0.089053017999721, + "learning_rate": 9.597416467786335e-07, + "loss": 0.447, + "step": 2385 + }, + { + "epoch": 4.68762278978389, + "grad_norm": 0.08699048357536948, + "learning_rate": 9.478298235672879e-07, + "loss": 0.4405, + "step": 2386 + }, + { + "epoch": 4.6895874263261295, + "grad_norm": 0.08710704664685442, + "learning_rate": 9.35991498079134e-07, + "loss": 0.4439, + "step": 2387 + }, + { + "epoch": 4.69155206286837, + "grad_norm": 0.08916091434414938, + "learning_rate": 9.242266925943722e-07, + "loss": 0.4349, + "step": 2388 + }, + { + "epoch": 4.693516699410609, + "grad_norm": 0.08508595001330588, + "learning_rate": 9.125354292548283e-07, + "loss": 0.4425, + "step": 2389 + }, + { + "epoch": 4.695481335952849, + "grad_norm": 0.08452580085341715, + "learning_rate": 9.00917730063915e-07, + "loss": 0.4487, + "step": 2390 + }, + { + "epoch": 4.697445972495088, + "grad_norm": 0.08594621552675058, + "learning_rate": 8.89373616886604e-07, + "loss": 0.4509, + "step": 2391 + }, + { + "epoch": 4.699410609037328, + "grad_norm": 0.08672688285253398, + "learning_rate": 8.779031114493741e-07, + "loss": 0.4471, + "step": 2392 + }, + { + "epoch": 4.701375245579568, + "grad_norm": 0.09225046170291723, + "learning_rate": 8.66506235340161e-07, + "loss": 0.4585, + "step": 2393 + }, + { + "epoch": 4.703339882121807, + "grad_norm": 0.08800829125435226, + "learning_rate": 8.551830100083403e-07, + "loss": 0.4457, + "step": 2394 + }, + { + "epoch": 4.705304518664047, + "grad_norm": 0.08388061125207875, + "learning_rate": 8.439334567646651e-07, + "loss": 0.4381, + "step": 2395 + }, + { + "epoch": 4.707269155206287, + "grad_norm": 0.08517129740544042, + "learning_rate": 8.327575967812351e-07, + "loss": 0.4392, + "step": 2396 + }, + { + "epoch": 4.7092337917485265, + "grad_norm": 0.08652506346031441, + "learning_rate": 8.216554510914698e-07, + "loss": 0.439, + "step": 2397 + }, + { + "epoch": 4.711198428290766, + "grad_norm": 0.088548586030388, + "learning_rate": 8.106270405900374e-07, + "loss": 0.4407, + "step": 2398 + }, + { + "epoch": 4.713163064833006, + "grad_norm": 0.08609607559263352, + "learning_rate": 7.996723860328415e-07, + "loss": 0.4428, + "step": 2399 + }, + { + "epoch": 4.715127701375246, + "grad_norm": 0.0881231005267988, + "learning_rate": 7.887915080369723e-07, + "loss": 0.4494, + "step": 2400 + }, + { + "epoch": 4.717092337917485, + "grad_norm": 0.0859683522390467, + "learning_rate": 7.779844270806758e-07, + "loss": 0.4437, + "step": 2401 + }, + { + "epoch": 4.719056974459725, + "grad_norm": 0.08921482552730474, + "learning_rate": 7.672511635033041e-07, + "loss": 0.4484, + "step": 2402 + }, + { + "epoch": 4.721021611001965, + "grad_norm": 0.0858243554331488, + "learning_rate": 7.565917375052723e-07, + "loss": 0.44, + "step": 2403 + }, + { + "epoch": 4.722986247544204, + "grad_norm": 0.09280213177966423, + "learning_rate": 7.460061691480524e-07, + "loss": 0.4424, + "step": 2404 + }, + { + "epoch": 4.724950884086444, + "grad_norm": 0.08637827353467896, + "learning_rate": 7.354944783540951e-07, + "loss": 0.4451, + "step": 2405 + }, + { + "epoch": 4.726915520628683, + "grad_norm": 0.08537215574258815, + "learning_rate": 7.250566849068152e-07, + "loss": 0.4308, + "step": 2406 + }, + { + "epoch": 4.728880157170924, + "grad_norm": 0.08697079252578524, + "learning_rate": 7.146928084505567e-07, + "loss": 0.4386, + "step": 2407 + }, + { + "epoch": 4.730844793713163, + "grad_norm": 0.08862610740650634, + "learning_rate": 7.044028684905435e-07, + "loss": 0.4504, + "step": 2408 + }, + { + "epoch": 4.7328094302554025, + "grad_norm": 0.09217366177147225, + "learning_rate": 6.941868843928401e-07, + "loss": 0.4496, + "step": 2409 + }, + { + "epoch": 4.734774066797643, + "grad_norm": 0.08720497762218045, + "learning_rate": 6.840448753843376e-07, + "loss": 0.449, + "step": 2410 + }, + { + "epoch": 4.736738703339882, + "grad_norm": 0.08977713685087566, + "learning_rate": 6.739768605527008e-07, + "loss": 0.4417, + "step": 2411 + }, + { + "epoch": 4.738703339882122, + "grad_norm": 0.08681823699233308, + "learning_rate": 6.639828588463193e-07, + "loss": 0.4373, + "step": 2412 + }, + { + "epoch": 4.740667976424361, + "grad_norm": 0.08627200871193763, + "learning_rate": 6.540628890743073e-07, + "loss": 0.4413, + "step": 2413 + }, + { + "epoch": 4.742632612966601, + "grad_norm": 0.08545483598168499, + "learning_rate": 6.442169699064327e-07, + "loss": 0.4427, + "step": 2414 + }, + { + "epoch": 4.744597249508841, + "grad_norm": 0.08353702458200256, + "learning_rate": 6.344451198731039e-07, + "loss": 0.4522, + "step": 2415 + }, + { + "epoch": 4.74656188605108, + "grad_norm": 0.08719121933702266, + "learning_rate": 6.24747357365334e-07, + "loss": 0.4453, + "step": 2416 + }, + { + "epoch": 4.74852652259332, + "grad_norm": 0.0866439604670392, + "learning_rate": 6.151237006346878e-07, + "loss": 0.4384, + "step": 2417 + }, + { + "epoch": 4.75049115913556, + "grad_norm": 0.08653296086109108, + "learning_rate": 6.05574167793268e-07, + "loss": 0.4455, + "step": 2418 + }, + { + "epoch": 4.7524557956777995, + "grad_norm": 0.08401419733934166, + "learning_rate": 5.960987768136762e-07, + "loss": 0.4429, + "step": 2419 + }, + { + "epoch": 4.754420432220039, + "grad_norm": 0.0917862860730644, + "learning_rate": 5.866975455289626e-07, + "loss": 0.4468, + "step": 2420 + }, + { + "epoch": 4.756385068762279, + "grad_norm": 0.08826956710573618, + "learning_rate": 5.773704916326273e-07, + "loss": 0.4329, + "step": 2421 + }, + { + "epoch": 4.758349705304519, + "grad_norm": 0.08618525770157885, + "learning_rate": 5.681176326785398e-07, + "loss": 0.4334, + "step": 2422 + }, + { + "epoch": 4.760314341846758, + "grad_norm": 0.08702727952731756, + "learning_rate": 5.589389860809569e-07, + "loss": 0.4386, + "step": 2423 + }, + { + "epoch": 4.7622789783889985, + "grad_norm": 0.0855741538620586, + "learning_rate": 5.498345691144425e-07, + "loss": 0.4396, + "step": 2424 + }, + { + "epoch": 4.764243614931238, + "grad_norm": 0.08487897776589558, + "learning_rate": 5.408043989138767e-07, + "loss": 0.4377, + "step": 2425 + }, + { + "epoch": 4.766208251473477, + "grad_norm": 0.08371666706644029, + "learning_rate": 5.318484924743939e-07, + "loss": 0.4414, + "step": 2426 + }, + { + "epoch": 4.768172888015717, + "grad_norm": 0.08608924286054552, + "learning_rate": 5.229668666513599e-07, + "loss": 0.4407, + "step": 2427 + }, + { + "epoch": 4.770137524557957, + "grad_norm": 0.0860728168029985, + "learning_rate": 5.141595381603459e-07, + "loss": 0.4411, + "step": 2428 + }, + { + "epoch": 4.772102161100197, + "grad_norm": 0.08521361990011199, + "learning_rate": 5.054265235770928e-07, + "loss": 0.4358, + "step": 2429 + }, + { + "epoch": 4.774066797642436, + "grad_norm": 0.08531405800213306, + "learning_rate": 4.967678393374753e-07, + "loss": 0.4374, + "step": 2430 + }, + { + "epoch": 4.7760314341846755, + "grad_norm": 0.08603743535413592, + "learning_rate": 4.881835017374803e-07, + "loss": 0.4428, + "step": 2431 + }, + { + "epoch": 4.777996070726916, + "grad_norm": 0.08701333415057666, + "learning_rate": 4.796735269331709e-07, + "loss": 0.4382, + "step": 2432 + }, + { + "epoch": 4.779960707269155, + "grad_norm": 0.08794223111055562, + "learning_rate": 4.712379309406512e-07, + "loss": 0.4331, + "step": 2433 + }, + { + "epoch": 4.781925343811395, + "grad_norm": 0.08705603962381622, + "learning_rate": 4.628767296360481e-07, + "loss": 0.4441, + "step": 2434 + }, + { + "epoch": 4.783889980353635, + "grad_norm": 0.08649584491197988, + "learning_rate": 4.545899387554675e-07, + "loss": 0.4399, + "step": 2435 + }, + { + "epoch": 4.785854616895874, + "grad_norm": 0.08406642299570218, + "learning_rate": 4.4637757389498495e-07, + "loss": 0.4369, + "step": 2436 + }, + { + "epoch": 4.787819253438114, + "grad_norm": 0.08579565775695955, + "learning_rate": 4.382396505105879e-07, + "loss": 0.4479, + "step": 2437 + }, + { + "epoch": 4.789783889980353, + "grad_norm": 0.08519736598091521, + "learning_rate": 4.301761839181717e-07, + "loss": 0.4424, + "step": 2438 + }, + { + "epoch": 4.791748526522594, + "grad_norm": 0.08584645175151177, + "learning_rate": 4.221871892934992e-07, + "loss": 0.4484, + "step": 2439 + }, + { + "epoch": 4.793713163064833, + "grad_norm": 0.08436666067869182, + "learning_rate": 4.142726816721698e-07, + "loss": 0.4398, + "step": 2440 + }, + { + "epoch": 4.7956777996070725, + "grad_norm": 0.08541942438650811, + "learning_rate": 4.0643267594959734e-07, + "loss": 0.4466, + "step": 2441 + }, + { + "epoch": 4.797642436149312, + "grad_norm": 0.08733991651177281, + "learning_rate": 3.986671868809877e-07, + "loss": 0.4469, + "step": 2442 + }, + { + "epoch": 4.799607072691552, + "grad_norm": 0.08569550395771183, + "learning_rate": 3.909762290812902e-07, + "loss": 0.4382, + "step": 2443 + }, + { + "epoch": 4.801571709233792, + "grad_norm": 0.08672525025537993, + "learning_rate": 3.833598170251973e-07, + "loss": 0.4445, + "step": 2444 + }, + { + "epoch": 4.803536345776031, + "grad_norm": 0.08790385974283632, + "learning_rate": 3.7581796504709166e-07, + "loss": 0.4404, + "step": 2445 + }, + { + "epoch": 4.8055009823182715, + "grad_norm": 0.08643953295509511, + "learning_rate": 3.6835068734103695e-07, + "loss": 0.4425, + "step": 2446 + }, + { + "epoch": 4.807465618860511, + "grad_norm": 0.08694295278052921, + "learning_rate": 3.6095799796075137e-07, + "loss": 0.4455, + "step": 2447 + }, + { + "epoch": 4.80943025540275, + "grad_norm": 0.08429313319560228, + "learning_rate": 3.5363991081955427e-07, + "loss": 0.4386, + "step": 2448 + }, + { + "epoch": 4.81139489194499, + "grad_norm": 0.08542493624503501, + "learning_rate": 3.463964396903841e-07, + "loss": 0.4402, + "step": 2449 + }, + { + "epoch": 4.81335952848723, + "grad_norm": 0.08508684846592472, + "learning_rate": 3.392275982057358e-07, + "loss": 0.4425, + "step": 2450 + }, + { + "epoch": 4.81532416502947, + "grad_norm": 0.0862939267014723, + "learning_rate": 3.3213339985765256e-07, + "loss": 0.4484, + "step": 2451 + }, + { + "epoch": 4.817288801571709, + "grad_norm": 0.08649425831274027, + "learning_rate": 3.251138579976898e-07, + "loss": 0.4454, + "step": 2452 + }, + { + "epoch": 4.8192534381139485, + "grad_norm": 0.08408361263585828, + "learning_rate": 3.181689858369108e-07, + "loss": 0.4477, + "step": 2453 + }, + { + "epoch": 4.821218074656189, + "grad_norm": 0.08795252117991173, + "learning_rate": 3.112987964458292e-07, + "loss": 0.439, + "step": 2454 + }, + { + "epoch": 4.823182711198428, + "grad_norm": 0.08634158190046692, + "learning_rate": 3.04503302754422e-07, + "loss": 0.4444, + "step": 2455 + }, + { + "epoch": 4.825147347740668, + "grad_norm": 0.08532194550786446, + "learning_rate": 2.977825175520677e-07, + "loss": 0.4444, + "step": 2456 + }, + { + "epoch": 4.827111984282908, + "grad_norm": 0.086659868848149, + "learning_rate": 2.9113645348755936e-07, + "loss": 0.4422, + "step": 2457 + }, + { + "epoch": 4.829076620825147, + "grad_norm": 0.08457660269307941, + "learning_rate": 2.8456512306903827e-07, + "loss": 0.4404, + "step": 2458 + }, + { + "epoch": 4.831041257367387, + "grad_norm": 0.08716967424567204, + "learning_rate": 2.7806853866401585e-07, + "loss": 0.4429, + "step": 2459 + }, + { + "epoch": 4.833005893909627, + "grad_norm": 0.08801932332825675, + "learning_rate": 2.7164671249932053e-07, + "loss": 0.4373, + "step": 2460 + }, + { + "epoch": 4.834970530451867, + "grad_norm": 0.08682704158179305, + "learning_rate": 2.652996566610755e-07, + "loss": 0.4426, + "step": 2461 + }, + { + "epoch": 4.836935166994106, + "grad_norm": 0.08689199014198651, + "learning_rate": 2.590273830946943e-07, + "loss": 0.4472, + "step": 2462 + }, + { + "epoch": 4.8388998035363455, + "grad_norm": 0.08423990456203899, + "learning_rate": 2.5282990360484096e-07, + "loss": 0.4443, + "step": 2463 + }, + { + "epoch": 4.840864440078585, + "grad_norm": 0.08841723273989292, + "learning_rate": 2.4670722985541627e-07, + "loss": 0.4388, + "step": 2464 + }, + { + "epoch": 4.842829076620825, + "grad_norm": 0.08435184388092522, + "learning_rate": 2.406593733695406e-07, + "loss": 0.44, + "step": 2465 + }, + { + "epoch": 4.844793713163065, + "grad_norm": 0.08735562450093359, + "learning_rate": 2.3468634552950898e-07, + "loss": 0.45, + "step": 2466 + }, + { + "epoch": 4.846758349705304, + "grad_norm": 0.08531745040610135, + "learning_rate": 2.2878815757680028e-07, + "loss": 0.4439, + "step": 2467 + }, + { + "epoch": 4.8487229862475445, + "grad_norm": 0.0866313887331337, + "learning_rate": 2.2296482061203273e-07, + "loss": 0.449, + "step": 2468 + }, + { + "epoch": 4.850687622789784, + "grad_norm": 0.08493310843148892, + "learning_rate": 2.1721634559496384e-07, + "loss": 0.4432, + "step": 2469 + }, + { + "epoch": 4.852652259332023, + "grad_norm": 0.08252953406303144, + "learning_rate": 2.1154274334443726e-07, + "loss": 0.4371, + "step": 2470 + }, + { + "epoch": 4.854616895874264, + "grad_norm": 0.08781368412509903, + "learning_rate": 2.0594402453840478e-07, + "loss": 0.4432, + "step": 2471 + }, + { + "epoch": 4.856581532416503, + "grad_norm": 0.08558023117596784, + "learning_rate": 2.0042019971387772e-07, + "loss": 0.4395, + "step": 2472 + }, + { + "epoch": 4.858546168958743, + "grad_norm": 0.08696081837158251, + "learning_rate": 1.9497127926690451e-07, + "loss": 0.4354, + "step": 2473 + }, + { + "epoch": 4.860510805500982, + "grad_norm": 0.08540596336407202, + "learning_rate": 1.8959727345257527e-07, + "loss": 0.4493, + "step": 2474 + }, + { + "epoch": 4.862475442043222, + "grad_norm": 0.08380040734064106, + "learning_rate": 1.8429819238497738e-07, + "loss": 0.4381, + "step": 2475 + }, + { + "epoch": 4.864440078585462, + "grad_norm": 0.0863470732413271, + "learning_rate": 1.7907404603719093e-07, + "loss": 0.4426, + "step": 2476 + }, + { + "epoch": 4.866404715127701, + "grad_norm": 0.08200283509372672, + "learning_rate": 1.7392484424126665e-07, + "loss": 0.4418, + "step": 2477 + }, + { + "epoch": 4.868369351669941, + "grad_norm": 0.08600735877294456, + "learning_rate": 1.6885059668819926e-07, + "loss": 0.4384, + "step": 2478 + }, + { + "epoch": 4.870333988212181, + "grad_norm": 0.08331649440514936, + "learning_rate": 1.6385131292793176e-07, + "loss": 0.4361, + "step": 2479 + }, + { + "epoch": 4.87229862475442, + "grad_norm": 0.08273252430120344, + "learning_rate": 1.5892700236930235e-07, + "loss": 0.4403, + "step": 2480 + }, + { + "epoch": 4.87426326129666, + "grad_norm": 0.08723326665107071, + "learning_rate": 1.540776742800665e-07, + "loss": 0.4403, + "step": 2481 + }, + { + "epoch": 4.8762278978389, + "grad_norm": 0.08314707788314188, + "learning_rate": 1.4930333778684803e-07, + "loss": 0.4358, + "step": 2482 + }, + { + "epoch": 4.87819253438114, + "grad_norm": 0.08724999206091466, + "learning_rate": 1.446040018751349e-07, + "loss": 0.4399, + "step": 2483 + }, + { + "epoch": 4.880157170923379, + "grad_norm": 0.08542802641820901, + "learning_rate": 1.3997967538926572e-07, + "loss": 0.4401, + "step": 2484 + }, + { + "epoch": 4.8821218074656185, + "grad_norm": 0.08561614633469654, + "learning_rate": 1.3543036703239864e-07, + "loss": 0.4483, + "step": 2485 + }, + { + "epoch": 4.884086444007859, + "grad_norm": 0.08346642869656568, + "learning_rate": 1.3095608536652038e-07, + "loss": 0.4395, + "step": 2486 + }, + { + "epoch": 4.886051080550098, + "grad_norm": 0.08407992675408951, + "learning_rate": 1.2655683881240167e-07, + "loss": 0.436, + "step": 2487 + }, + { + "epoch": 4.888015717092338, + "grad_norm": 0.0895871609694908, + "learning_rate": 1.2223263564959286e-07, + "loss": 0.4487, + "step": 2488 + }, + { + "epoch": 4.889980353634577, + "grad_norm": 0.08575499626498645, + "learning_rate": 1.1798348401641958e-07, + "loss": 0.4449, + "step": 2489 + }, + { + "epoch": 4.8919449901768175, + "grad_norm": 0.08455932797246347, + "learning_rate": 1.1380939190995587e-07, + "loss": 0.4385, + "step": 2490 + }, + { + "epoch": 4.893909626719057, + "grad_norm": 0.08345983390336496, + "learning_rate": 1.0971036718599781e-07, + "loss": 0.441, + "step": 2491 + }, + { + "epoch": 4.895874263261296, + "grad_norm": 0.08130315974702403, + "learning_rate": 1.0568641755908104e-07, + "loss": 0.4367, + "step": 2492 + }, + { + "epoch": 4.897838899803537, + "grad_norm": 0.08503156843639774, + "learning_rate": 1.0173755060243207e-07, + "loss": 0.4461, + "step": 2493 + }, + { + "epoch": 4.899803536345776, + "grad_norm": 0.08670424280402983, + "learning_rate": 9.786377374797262e-08, + "loss": 0.4465, + "step": 2494 + }, + { + "epoch": 4.901768172888016, + "grad_norm": 0.08869645521380927, + "learning_rate": 9.406509428631083e-08, + "loss": 0.4424, + "step": 2495 + }, + { + "epoch": 4.903732809430255, + "grad_norm": 0.08320340575053557, + "learning_rate": 9.034151936670566e-08, + "loss": 0.4423, + "step": 2496 + }, + { + "epoch": 4.905697445972495, + "grad_norm": 0.08453112319303341, + "learning_rate": 8.669305599707134e-08, + "loss": 0.4539, + "step": 2497 + }, + { + "epoch": 4.907662082514735, + "grad_norm": 0.08487135847751147, + "learning_rate": 8.31197110439641e-08, + "loss": 0.4398, + "step": 2498 + }, + { + "epoch": 4.909626719056974, + "grad_norm": 0.08611505719028877, + "learning_rate": 7.962149123255547e-08, + "loss": 0.4453, + "step": 2499 + }, + { + "epoch": 4.911591355599214, + "grad_norm": 0.08462569687078017, + "learning_rate": 7.619840314663229e-08, + "loss": 0.4463, + "step": 2500 + }, + { + "epoch": 4.913555992141454, + "grad_norm": 0.08434342942587726, + "learning_rate": 7.285045322858342e-08, + "loss": 0.4434, + "step": 2501 + }, + { + "epoch": 4.915520628683693, + "grad_norm": 0.0847196988320346, + "learning_rate": 6.957764777938636e-08, + "loss": 0.4445, + "step": 2502 + }, + { + "epoch": 4.917485265225933, + "grad_norm": 0.08351977152207293, + "learning_rate": 6.637999295858511e-08, + "loss": 0.4389, + "step": 2503 + }, + { + "epoch": 4.919449901768173, + "grad_norm": 0.08438897577848146, + "learning_rate": 6.325749478429011e-08, + "loss": 0.4418, + "step": 2504 + }, + { + "epoch": 4.921414538310413, + "grad_norm": 0.08349963533554272, + "learning_rate": 6.021015913316941e-08, + "loss": 0.4446, + "step": 2505 + }, + { + "epoch": 4.923379174852652, + "grad_norm": 0.08400768954127598, + "learning_rate": 5.723799174043087e-08, + "loss": 0.4412, + "step": 2506 + }, + { + "epoch": 4.925343811394892, + "grad_norm": 0.08431682870308162, + "learning_rate": 5.434099819980887e-08, + "loss": 0.4427, + "step": 2507 + }, + { + "epoch": 4.927308447937132, + "grad_norm": 0.08427587445457796, + "learning_rate": 5.1519183963555374e-08, + "loss": 0.4392, + "step": 2508 + }, + { + "epoch": 4.929273084479371, + "grad_norm": 0.0842033913119077, + "learning_rate": 4.8772554342439995e-08, + "loss": 0.4463, + "step": 2509 + }, + { + "epoch": 4.931237721021611, + "grad_norm": 0.08414686695824386, + "learning_rate": 4.610111450572774e-08, + "loss": 0.4429, + "step": 2510 + }, + { + "epoch": 4.933202357563851, + "grad_norm": 0.08434292104058787, + "learning_rate": 4.350486948117461e-08, + "loss": 0.4461, + "step": 2511 + }, + { + "epoch": 4.9351669941060905, + "grad_norm": 0.0823914509837016, + "learning_rate": 4.098382415500979e-08, + "loss": 0.4419, + "step": 2512 + }, + { + "epoch": 4.93713163064833, + "grad_norm": 0.08334061288356942, + "learning_rate": 3.853798327194902e-08, + "loss": 0.4371, + "step": 2513 + }, + { + "epoch": 4.939096267190569, + "grad_norm": 0.08629697407046513, + "learning_rate": 3.61673514351546e-08, + "loss": 0.4456, + "step": 2514 + }, + { + "epoch": 4.94106090373281, + "grad_norm": 0.08456266879851708, + "learning_rate": 3.387193310625314e-08, + "loss": 0.4462, + "step": 2515 + }, + { + "epoch": 4.943025540275049, + "grad_norm": 0.08595418445288092, + "learning_rate": 3.16517326053134e-08, + "loss": 0.4433, + "step": 2516 + }, + { + "epoch": 4.944990176817289, + "grad_norm": 0.08479669504878344, + "learning_rate": 2.9506754110832926e-08, + "loss": 0.4481, + "step": 2517 + }, + { + "epoch": 4.946954813359529, + "grad_norm": 0.08313548346870327, + "learning_rate": 2.7437001659751385e-08, + "loss": 0.4506, + "step": 2518 + }, + { + "epoch": 4.948919449901768, + "grad_norm": 0.08601603270472373, + "learning_rate": 2.5442479147423925e-08, + "loss": 0.4402, + "step": 2519 + }, + { + "epoch": 4.950884086444008, + "grad_norm": 0.082254821248563, + "learning_rate": 2.352319032761674e-08, + "loss": 0.4423, + "step": 2520 + }, + { + "epoch": 4.952848722986247, + "grad_norm": 0.08536677929816872, + "learning_rate": 2.167913881251593e-08, + "loss": 0.4455, + "step": 2521 + }, + { + "epoch": 4.954813359528488, + "grad_norm": 0.08248578134750814, + "learning_rate": 1.9910328072687556e-08, + "loss": 0.437, + "step": 2522 + }, + { + "epoch": 4.956777996070727, + "grad_norm": 0.08382600029268458, + "learning_rate": 1.821676143711315e-08, + "loss": 0.4482, + "step": 2523 + }, + { + "epoch": 4.958742632612966, + "grad_norm": 0.08645061810896823, + "learning_rate": 1.6598442093145317e-08, + "loss": 0.4543, + "step": 2524 + }, + { + "epoch": 4.960707269155206, + "grad_norm": 0.08522588082403543, + "learning_rate": 1.5055373086525492e-08, + "loss": 0.4418, + "step": 2525 + }, + { + "epoch": 4.962671905697446, + "grad_norm": 0.0879383087650699, + "learning_rate": 1.3587557321375067e-08, + "loss": 0.4456, + "step": 2526 + }, + { + "epoch": 4.964636542239686, + "grad_norm": 0.08344927146027245, + "learning_rate": 1.2194997560177613e-08, + "loss": 0.4368, + "step": 2527 + }, + { + "epoch": 4.966601178781925, + "grad_norm": 0.08489409981380447, + "learning_rate": 1.0877696423787776e-08, + "loss": 0.4393, + "step": 2528 + }, + { + "epoch": 4.968565815324165, + "grad_norm": 0.0839737546378588, + "learning_rate": 9.635656391413505e-09, + "loss": 0.4367, + "step": 2529 + }, + { + "epoch": 4.970530451866405, + "grad_norm": 0.08437171736823382, + "learning_rate": 8.468879800629381e-09, + "loss": 0.4396, + "step": 2530 + }, + { + "epoch": 4.972495088408644, + "grad_norm": 0.0858011871663949, + "learning_rate": 7.377368847345523e-09, + "loss": 0.4372, + "step": 2531 + }, + { + "epoch": 4.974459724950884, + "grad_norm": 0.08625760905454406, + "learning_rate": 6.361125585838679e-09, + "loss": 0.4453, + "step": 2532 + }, + { + "epoch": 4.976424361493124, + "grad_norm": 0.08424861183422554, + "learning_rate": 5.4201519287166994e-09, + "loss": 0.4439, + "step": 2533 + }, + { + "epoch": 4.9783889980353635, + "grad_norm": 0.08254080071895513, + "learning_rate": 4.554449646922976e-09, + "loss": 0.4353, + "step": 2534 + }, + { + "epoch": 4.980353634577603, + "grad_norm": 0.08422468629396888, + "learning_rate": 3.764020369749766e-09, + "loss": 0.4403, + "step": 2535 + }, + { + "epoch": 4.982318271119842, + "grad_norm": 0.08749485932012267, + "learning_rate": 3.0488655848115444e-09, + "loss": 0.4514, + "step": 2536 + }, + { + "epoch": 4.984282907662083, + "grad_norm": 0.08411282070689433, + "learning_rate": 2.4089866380627714e-09, + "loss": 0.4496, + "step": 2537 + }, + { + "epoch": 4.986247544204322, + "grad_norm": 0.09166804907521955, + "learning_rate": 1.8443847337756837e-09, + "loss": 0.4478, + "step": 2538 + }, + { + "epoch": 4.988212180746562, + "grad_norm": 0.08395607119815886, + "learning_rate": 1.3550609345536203e-09, + "loss": 0.4448, + "step": 2539 + }, + { + "epoch": 4.990176817288802, + "grad_norm": 0.08618189345070469, + "learning_rate": 9.410161613310209e-10, + "loss": 0.4417, + "step": 2540 + }, + { + "epoch": 4.992141453831041, + "grad_norm": 0.08816368957871021, + "learning_rate": 6.022511933467811e-10, + "loss": 0.4485, + "step": 2541 + }, + { + "epoch": 4.994106090373281, + "grad_norm": 0.08279148025343501, + "learning_rate": 3.3876666817533874e-10, + "loss": 0.4447, + "step": 2542 + }, + { + "epoch": 4.996070726915521, + "grad_norm": 0.08477068004302633, + "learning_rate": 1.505630817000281e-10, + "loss": 0.4426, + "step": 2543 + }, + { + "epoch": 4.998035363457761, + "grad_norm": 0.08348112694109847, + "learning_rate": 3.764078813528471e-11, + "loss": 0.4403, + "step": 2544 + }, + { + "epoch": 5.0, + "grad_norm": 0.08525863550749231, + "learning_rate": 0.0, + "loss": 0.4402, + "step": 2545 + }, + { + "epoch": 5.0, + "step": 2545, + "total_flos": 4.269528204705792e+16, + "train_loss": 0.5355927676606507, + "train_runtime": 39267.2174, + "train_samples_per_second": 33.179, + "train_steps_per_second": 0.065 + } + ], + "logging_steps": 1, + "max_steps": 2545, + "num_input_tokens_seen": 0, + "num_train_epochs": 5, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 4.269528204705792e+16, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}