{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 49.986784140969164, "eval_steps": 250, "global_step": 3500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.014096916299559472, "grad_norm": 5357.12548828125, "learning_rate": 1.9047619047619051e-06, "loss": 21.2077, "step": 1 }, { "epoch": 0.028193832599118944, "grad_norm": 7397.6806640625, "learning_rate": 3.8095238095238102e-06, "loss": 22.4147, "step": 2 }, { "epoch": 0.04229074889867841, "grad_norm": 8375.7744140625, "learning_rate": 5.7142857142857145e-06, "loss": 22.8773, "step": 3 }, { "epoch": 0.05638766519823789, "grad_norm": 3280.87890625, "learning_rate": 7.6190476190476205e-06, "loss": 19.5143, "step": 4 }, { "epoch": 0.07048458149779736, "grad_norm": 604.7725830078125, "learning_rate": 9.523809523809523e-06, "loss": 17.9598, "step": 5 }, { "epoch": 0.08458149779735682, "grad_norm": 457.183837890625, "learning_rate": 1.1428571428571429e-05, "loss": 18.2444, "step": 6 }, { "epoch": 0.0986784140969163, "grad_norm": 444.1527099609375, "learning_rate": 1.3333333333333333e-05, "loss": 17.9295, "step": 7 }, { "epoch": 0.11277533039647578, "grad_norm": 164.02256774902344, "learning_rate": 1.5238095238095241e-05, "loss": 15.038, "step": 8 }, { "epoch": 0.12687224669603525, "grad_norm": 159.0693359375, "learning_rate": 1.7142857142857145e-05, "loss": 14.8469, "step": 9 }, { "epoch": 0.14096916299559473, "grad_norm": 121.40792846679688, "learning_rate": 1.9047619047619046e-05, "loss": 14.3252, "step": 10 }, { "epoch": 0.15506607929515417, "grad_norm": 91.15929412841797, "learning_rate": 2.0952380952380954e-05, "loss": 14.0431, "step": 11 }, { "epoch": 0.16916299559471365, "grad_norm": 55.599998474121094, "learning_rate": 2.2857142857142858e-05, "loss": 13.6387, "step": 12 }, { "epoch": 0.18325991189427313, "grad_norm": 47.33129119873047, "learning_rate": 2.4761904761904762e-05, "loss": 13.1092, "step": 13 }, { "epoch": 0.1973568281938326, "grad_norm": 43.76308822631836, "learning_rate": 2.6666666666666667e-05, "loss": 12.7865, "step": 14 }, { "epoch": 0.21145374449339208, "grad_norm": 35.16502380371094, "learning_rate": 2.857142857142857e-05, "loss": 12.6341, "step": 15 }, { "epoch": 0.22555066079295155, "grad_norm": 25.88783836364746, "learning_rate": 3.0476190476190482e-05, "loss": 12.4117, "step": 16 }, { "epoch": 0.239647577092511, "grad_norm": 23.84360122680664, "learning_rate": 3.2380952380952386e-05, "loss": 12.0547, "step": 17 }, { "epoch": 0.2537444933920705, "grad_norm": 23.6061954498291, "learning_rate": 3.428571428571429e-05, "loss": 11.743, "step": 18 }, { "epoch": 0.26784140969162995, "grad_norm": 21.413909912109375, "learning_rate": 3.619047619047619e-05, "loss": 11.5602, "step": 19 }, { "epoch": 0.28193832599118945, "grad_norm": 16.649694442749023, "learning_rate": 3.809523809523809e-05, "loss": 11.4545, "step": 20 }, { "epoch": 0.2960352422907489, "grad_norm": 12.360799789428711, "learning_rate": 4e-05, "loss": 11.4392, "step": 21 }, { "epoch": 0.31013215859030835, "grad_norm": 11.416522979736328, "learning_rate": 4.190476190476191e-05, "loss": 11.0822, "step": 22 }, { "epoch": 0.32422907488986785, "grad_norm": 10.329779624938965, "learning_rate": 4.380952380952381e-05, "loss": 11.2924, "step": 23 }, { "epoch": 0.3383259911894273, "grad_norm": 8.719940185546875, "learning_rate": 4.5714285714285716e-05, "loss": 11.1777, "step": 24 }, { "epoch": 0.3524229074889868, "grad_norm": 9.542926788330078, "learning_rate": 4.761904761904762e-05, "loss": 10.8962, "step": 25 }, { "epoch": 0.36651982378854625, "grad_norm": 7.817182540893555, "learning_rate": 4.9523809523809525e-05, "loss": 11.1445, "step": 26 }, { "epoch": 0.38061674008810575, "grad_norm": 7.721704483032227, "learning_rate": 5.142857142857143e-05, "loss": 10.9657, "step": 27 }, { "epoch": 0.3947136563876652, "grad_norm": 6.033113479614258, "learning_rate": 5.333333333333333e-05, "loss": 11.0756, "step": 28 }, { "epoch": 0.40881057268722465, "grad_norm": 5.228347301483154, "learning_rate": 5.5238095238095244e-05, "loss": 11.0705, "step": 29 }, { "epoch": 0.42290748898678415, "grad_norm": 5.556614398956299, "learning_rate": 5.714285714285714e-05, "loss": 10.9568, "step": 30 }, { "epoch": 0.4370044052863436, "grad_norm": 6.477099418640137, "learning_rate": 5.904761904761905e-05, "loss": 11.1831, "step": 31 }, { "epoch": 0.4511013215859031, "grad_norm": 6.181473255157471, "learning_rate": 6.0952380952380964e-05, "loss": 10.9515, "step": 32 }, { "epoch": 0.46519823788546255, "grad_norm": 4.373332500457764, "learning_rate": 6.285714285714286e-05, "loss": 10.9047, "step": 33 }, { "epoch": 0.479295154185022, "grad_norm": 4.678891658782959, "learning_rate": 6.476190476190477e-05, "loss": 10.9663, "step": 34 }, { "epoch": 0.4933920704845815, "grad_norm": 3.966383934020996, "learning_rate": 6.666666666666667e-05, "loss": 10.8571, "step": 35 }, { "epoch": 0.507488986784141, "grad_norm": 3.549818992614746, "learning_rate": 6.857142857142858e-05, "loss": 10.7054, "step": 36 }, { "epoch": 0.5215859030837005, "grad_norm": 3.516842842102051, "learning_rate": 7.047619047619048e-05, "loss": 10.8657, "step": 37 }, { "epoch": 0.5356828193832599, "grad_norm": 3.4007351398468018, "learning_rate": 7.238095238095238e-05, "loss": 10.8557, "step": 38 }, { "epoch": 0.5497797356828193, "grad_norm": 4.030512809753418, "learning_rate": 7.428571428571429e-05, "loss": 10.799, "step": 39 }, { "epoch": 0.5638766519823789, "grad_norm": 3.3622701168060303, "learning_rate": 7.619047619047618e-05, "loss": 10.8025, "step": 40 }, { "epoch": 0.5779735682819384, "grad_norm": 2.908905029296875, "learning_rate": 7.80952380952381e-05, "loss": 10.8114, "step": 41 }, { "epoch": 0.5920704845814978, "grad_norm": 2.452524423599243, "learning_rate": 8e-05, "loss": 10.7112, "step": 42 }, { "epoch": 0.6061674008810573, "grad_norm": 2.633840322494507, "learning_rate": 8.19047619047619e-05, "loss": 10.6229, "step": 43 }, { "epoch": 0.6202643171806167, "grad_norm": 2.7238128185272217, "learning_rate": 8.380952380952382e-05, "loss": 10.7177, "step": 44 }, { "epoch": 0.6343612334801763, "grad_norm": 3.0034332275390625, "learning_rate": 8.571428571428571e-05, "loss": 10.7488, "step": 45 }, { "epoch": 0.6484581497797357, "grad_norm": 2.262722969055176, "learning_rate": 8.761904761904762e-05, "loss": 10.7086, "step": 46 }, { "epoch": 0.6625550660792952, "grad_norm": 2.6180567741394043, "learning_rate": 8.952380952380953e-05, "loss": 10.6588, "step": 47 }, { "epoch": 0.6766519823788546, "grad_norm": 2.316925287246704, "learning_rate": 9.142857142857143e-05, "loss": 10.859, "step": 48 }, { "epoch": 0.690748898678414, "grad_norm": 2.064025402069092, "learning_rate": 9.333333333333334e-05, "loss": 10.7581, "step": 49 }, { "epoch": 0.7048458149779736, "grad_norm": 2.87385630607605, "learning_rate": 9.523809523809524e-05, "loss": 10.7703, "step": 50 }, { "epoch": 0.718942731277533, "grad_norm": 2.0411784648895264, "learning_rate": 9.714285714285715e-05, "loss": 10.7233, "step": 51 }, { "epoch": 0.7330396475770925, "grad_norm": 2.925504684448242, "learning_rate": 9.904761904761905e-05, "loss": 10.4427, "step": 52 }, { "epoch": 0.747136563876652, "grad_norm": 2.4529292583465576, "learning_rate": 0.00010095238095238096, "loss": 10.4363, "step": 53 }, { "epoch": 0.7612334801762115, "grad_norm": 2.116291046142578, "learning_rate": 0.00010285714285714286, "loss": 10.8726, "step": 54 }, { "epoch": 0.775330396475771, "grad_norm": 5.709266185760498, "learning_rate": 0.00010476190476190477, "loss": 10.7182, "step": 55 }, { "epoch": 0.7894273127753304, "grad_norm": 2.508542776107788, "learning_rate": 0.00010666666666666667, "loss": 10.4941, "step": 56 }, { "epoch": 0.8035242290748899, "grad_norm": 2.073350191116333, "learning_rate": 0.00010857142857142856, "loss": 10.543, "step": 57 }, { "epoch": 0.8176211453744493, "grad_norm": 2.116551399230957, "learning_rate": 0.00011047619047619049, "loss": 10.6302, "step": 58 }, { "epoch": 0.8317180616740089, "grad_norm": 2.0298280715942383, "learning_rate": 0.00011238095238095239, "loss": 10.5102, "step": 59 }, { "epoch": 0.8458149779735683, "grad_norm": 2.2986271381378174, "learning_rate": 0.00011428571428571428, "loss": 10.6954, "step": 60 }, { "epoch": 0.8599118942731278, "grad_norm": 1.6656768321990967, "learning_rate": 0.00011619047619047621, "loss": 10.6927, "step": 61 }, { "epoch": 0.8740088105726872, "grad_norm": 2.3102877140045166, "learning_rate": 0.0001180952380952381, "loss": 10.4144, "step": 62 }, { "epoch": 0.8881057268722466, "grad_norm": 2.5880520343780518, "learning_rate": 0.00012, "loss": 10.2329, "step": 63 }, { "epoch": 0.9022026431718062, "grad_norm": 1.7668675184249878, "learning_rate": 0.00012190476190476193, "loss": 10.4856, "step": 64 }, { "epoch": 0.9162995594713657, "grad_norm": 1.9075130224227905, "learning_rate": 0.0001238095238095238, "loss": 10.6341, "step": 65 }, { "epoch": 0.9303964757709251, "grad_norm": 2.2096028327941895, "learning_rate": 0.00012571428571428572, "loss": 10.4778, "step": 66 }, { "epoch": 0.9444933920704845, "grad_norm": 1.9562084674835205, "learning_rate": 0.0001276190476190476, "loss": 10.4415, "step": 67 }, { "epoch": 0.958590308370044, "grad_norm": 1.7809193134307861, "learning_rate": 0.00012952380952380954, "loss": 10.4699, "step": 68 }, { "epoch": 0.9726872246696036, "grad_norm": 1.9420225620269775, "learning_rate": 0.00013142857142857143, "loss": 10.451, "step": 69 }, { "epoch": 0.986784140969163, "grad_norm": 1.830646276473999, "learning_rate": 0.00013333333333333334, "loss": 10.6183, "step": 70 }, { "epoch": 1.0140969162995594, "grad_norm": 1.699867606163025, "learning_rate": 0.00013523809523809525, "loss": 18.8055, "step": 71 }, { "epoch": 1.028193832599119, "grad_norm": 1.7382451295852661, "learning_rate": 0.00013714285714285716, "loss": 10.5239, "step": 72 }, { "epoch": 1.0422907488986783, "grad_norm": 2.083167314529419, "learning_rate": 0.00013904761904761905, "loss": 10.1895, "step": 73 }, { "epoch": 1.0563876651982378, "grad_norm": 1.9177883863449097, "learning_rate": 0.00014095238095238096, "loss": 10.3891, "step": 74 }, { "epoch": 1.0704845814977975, "grad_norm": 2.1042675971984863, "learning_rate": 0.00014285714285714287, "loss": 10.452, "step": 75 }, { "epoch": 1.084581497797357, "grad_norm": 2.2132911682128906, "learning_rate": 0.00014476190476190475, "loss": 10.3364, "step": 76 }, { "epoch": 1.0986784140969164, "grad_norm": 2.121964931488037, "learning_rate": 0.00014666666666666666, "loss": 10.2949, "step": 77 }, { "epoch": 1.1127753303964758, "grad_norm": 2.082794427871704, "learning_rate": 0.00014857142857142857, "loss": 10.459, "step": 78 }, { "epoch": 1.1268722466960353, "grad_norm": 2.319943904876709, "learning_rate": 0.00015047619047619048, "loss": 10.5161, "step": 79 }, { "epoch": 1.1409691629955947, "grad_norm": 2.3347713947296143, "learning_rate": 0.00015238095238095237, "loss": 10.3928, "step": 80 }, { "epoch": 1.1550660792951541, "grad_norm": 2.7926313877105713, "learning_rate": 0.0001542857142857143, "loss": 10.287, "step": 81 }, { "epoch": 1.1691629955947136, "grad_norm": 2.536135673522949, "learning_rate": 0.0001561904761904762, "loss": 10.4437, "step": 82 }, { "epoch": 1.183259911894273, "grad_norm": 2.6055548191070557, "learning_rate": 0.0001580952380952381, "loss": 10.3372, "step": 83 }, { "epoch": 1.1973568281938327, "grad_norm": 2.6193349361419678, "learning_rate": 0.00016, "loss": 10.4843, "step": 84 }, { "epoch": 1.2114537444933922, "grad_norm": 2.4078187942504883, "learning_rate": 0.00016190476190476192, "loss": 10.3089, "step": 85 }, { "epoch": 1.2255506607929516, "grad_norm": 2.452697515487671, "learning_rate": 0.0001638095238095238, "loss": 10.3286, "step": 86 }, { "epoch": 1.239647577092511, "grad_norm": 2.759438991546631, "learning_rate": 0.00016571428571428575, "loss": 10.3132, "step": 87 }, { "epoch": 1.2537444933920705, "grad_norm": 2.751279592514038, "learning_rate": 0.00016761904761904763, "loss": 10.4765, "step": 88 }, { "epoch": 1.26784140969163, "grad_norm": 3.284719944000244, "learning_rate": 0.00016952380952380954, "loss": 10.3297, "step": 89 }, { "epoch": 1.2819383259911894, "grad_norm": 2.7485530376434326, "learning_rate": 0.00017142857142857143, "loss": 10.3661, "step": 90 }, { "epoch": 1.2960352422907488, "grad_norm": 3.143373966217041, "learning_rate": 0.00017333333333333334, "loss": 10.3279, "step": 91 }, { "epoch": 1.3101321585903083, "grad_norm": 2.9736995697021484, "learning_rate": 0.00017523809523809525, "loss": 10.4474, "step": 92 }, { "epoch": 1.324229074889868, "grad_norm": 2.820429801940918, "learning_rate": 0.00017714285714285713, "loss": 10.1064, "step": 93 }, { "epoch": 1.3383259911894272, "grad_norm": 3.3731212615966797, "learning_rate": 0.00017904761904761907, "loss": 10.2487, "step": 94 }, { "epoch": 1.3524229074889869, "grad_norm": 2.7274646759033203, "learning_rate": 0.00018095238095238095, "loss": 10.5687, "step": 95 }, { "epoch": 1.3665198237885463, "grad_norm": 3.0304393768310547, "learning_rate": 0.00018285714285714286, "loss": 10.3506, "step": 96 }, { "epoch": 1.3806167400881058, "grad_norm": 3.7191991806030273, "learning_rate": 0.00018476190476190478, "loss": 10.2375, "step": 97 }, { "epoch": 1.3947136563876652, "grad_norm": 3.317068099975586, "learning_rate": 0.0001866666666666667, "loss": 10.2413, "step": 98 }, { "epoch": 1.4088105726872246, "grad_norm": 2.423412799835205, "learning_rate": 0.00018857142857142857, "loss": 10.3687, "step": 99 }, { "epoch": 1.422907488986784, "grad_norm": 2.63938045501709, "learning_rate": 0.00019047619047619048, "loss": 10.2288, "step": 100 }, { "epoch": 1.4370044052863435, "grad_norm": 2.648970127105713, "learning_rate": 0.0001923809523809524, "loss": 10.3325, "step": 101 }, { "epoch": 1.4511013215859032, "grad_norm": 3.3892431259155273, "learning_rate": 0.0001942857142857143, "loss": 10.2807, "step": 102 }, { "epoch": 1.4651982378854624, "grad_norm": 3.3631839752197266, "learning_rate": 0.0001961904761904762, "loss": 10.2419, "step": 103 }, { "epoch": 1.479295154185022, "grad_norm": 3.8143887519836426, "learning_rate": 0.0001980952380952381, "loss": 10.2229, "step": 104 }, { "epoch": 1.4933920704845816, "grad_norm": 5.305814266204834, "learning_rate": 0.0002, "loss": 10.3052, "step": 105 }, { "epoch": 1.507488986784141, "grad_norm": 3.1108529567718506, "learning_rate": 0.0001999999571855705, "loss": 10.3291, "step": 106 }, { "epoch": 1.5215859030837005, "grad_norm": 4.598055839538574, "learning_rate": 0.00019999982874231863, "loss": 10.289, "step": 107 }, { "epoch": 1.53568281938326, "grad_norm": 2.774738311767578, "learning_rate": 0.00019999961467035443, "loss": 10.2634, "step": 108 }, { "epoch": 1.5497797356828193, "grad_norm": 2.670076847076416, "learning_rate": 0.00019999931496986114, "loss": 10.1631, "step": 109 }, { "epoch": 1.5638766519823788, "grad_norm": 3.110130548477173, "learning_rate": 0.00019999892964109545, "loss": 10.3533, "step": 110 }, { "epoch": 1.5779735682819385, "grad_norm": 2.909061908721924, "learning_rate": 0.00019999845868438727, "loss": 10.3259, "step": 111 }, { "epoch": 1.5920704845814977, "grad_norm": 2.9035227298736572, "learning_rate": 0.00019999790210013988, "loss": 10.3109, "step": 112 }, { "epoch": 1.6061674008810574, "grad_norm": 3.2428600788116455, "learning_rate": 0.0001999972598888299, "loss": 10.2389, "step": 113 }, { "epoch": 1.6202643171806166, "grad_norm": 2.6514716148376465, "learning_rate": 0.00019999653205100723, "loss": 10.162, "step": 114 }, { "epoch": 1.6343612334801763, "grad_norm": 3.068436861038208, "learning_rate": 0.00019999571858729512, "loss": 10.1654, "step": 115 }, { "epoch": 1.6484581497797357, "grad_norm": 2.98667311668396, "learning_rate": 0.00019999481949839012, "loss": 9.9585, "step": 116 }, { "epoch": 1.6625550660792952, "grad_norm": 3.550619125366211, "learning_rate": 0.0001999938347850621, "loss": 10.2781, "step": 117 }, { "epoch": 1.6766519823788546, "grad_norm": 2.578996419906616, "learning_rate": 0.0001999927644481543, "loss": 10.324, "step": 118 }, { "epoch": 1.690748898678414, "grad_norm": 3.6342506408691406, "learning_rate": 0.0001999916084885832, "loss": 10.1154, "step": 119 }, { "epoch": 1.7048458149779737, "grad_norm": 3.18060040473938, "learning_rate": 0.00019999036690733865, "loss": 10.1144, "step": 120 }, { "epoch": 1.718942731277533, "grad_norm": 4.085129737854004, "learning_rate": 0.00019998903970548377, "loss": 10.1215, "step": 121 }, { "epoch": 1.7330396475770926, "grad_norm": 2.751615285873413, "learning_rate": 0.00019998762688415507, "loss": 10.3077, "step": 122 }, { "epoch": 1.7471365638766518, "grad_norm": 3.2955965995788574, "learning_rate": 0.00019998612844456234, "loss": 10.1335, "step": 123 }, { "epoch": 1.7612334801762115, "grad_norm": 2.6506381034851074, "learning_rate": 0.0001999845443879886, "loss": 10.3805, "step": 124 }, { "epoch": 1.775330396475771, "grad_norm": 3.3389201164245605, "learning_rate": 0.00019998287471579036, "loss": 10.1157, "step": 125 }, { "epoch": 1.7894273127753304, "grad_norm": 3.16748046875, "learning_rate": 0.0001999811194293973, "loss": 9.9452, "step": 126 }, { "epoch": 1.8035242290748899, "grad_norm": 3.0726938247680664, "learning_rate": 0.0001999792785303124, "loss": 10.125, "step": 127 }, { "epoch": 1.8176211453744493, "grad_norm": 3.5298125743865967, "learning_rate": 0.00019997735202011206, "loss": 10.1521, "step": 128 }, { "epoch": 1.831718061674009, "grad_norm": 2.519752025604248, "learning_rate": 0.00019997533990044595, "loss": 10.312, "step": 129 }, { "epoch": 1.8458149779735682, "grad_norm": 4.698988914489746, "learning_rate": 0.00019997324217303696, "loss": 10.1738, "step": 130 }, { "epoch": 1.8599118942731279, "grad_norm": 3.4604554176330566, "learning_rate": 0.00019997105883968138, "loss": 10.268, "step": 131 }, { "epoch": 1.874008810572687, "grad_norm": 2.959848642349243, "learning_rate": 0.0001999687899022488, "loss": 10.1193, "step": 132 }, { "epoch": 1.8881057268722468, "grad_norm": 4.0414509773254395, "learning_rate": 0.00019996643536268204, "loss": 10.043, "step": 133 }, { "epoch": 1.9022026431718062, "grad_norm": 2.5412702560424805, "learning_rate": 0.0001999639952229973, "loss": 9.9677, "step": 134 }, { "epoch": 1.9162995594713657, "grad_norm": 4.8571343421936035, "learning_rate": 0.000199961469485284, "loss": 10.0072, "step": 135 }, { "epoch": 1.930396475770925, "grad_norm": 3.720360040664673, "learning_rate": 0.00019995885815170494, "loss": 9.9401, "step": 136 }, { "epoch": 1.9444933920704845, "grad_norm": 2.62038516998291, "learning_rate": 0.00019995616122449618, "loss": 10.2046, "step": 137 }, { "epoch": 1.958590308370044, "grad_norm": 3.1839091777801514, "learning_rate": 0.00019995337870596703, "loss": 9.9842, "step": 138 }, { "epoch": 1.9726872246696034, "grad_norm": 3.1514859199523926, "learning_rate": 0.00019995051059850018, "loss": 10.0523, "step": 139 }, { "epoch": 1.9867841409691631, "grad_norm": 3.2248549461364746, "learning_rate": 0.00019994755690455152, "loss": 10.1185, "step": 140 }, { "epoch": 2.0140969162995597, "grad_norm": 3.124788284301758, "learning_rate": 0.00019994451762665025, "loss": 18.1426, "step": 141 }, { "epoch": 2.028193832599119, "grad_norm": 3.759366750717163, "learning_rate": 0.0001999413927673989, "loss": 10.0812, "step": 142 }, { "epoch": 2.0422907488986786, "grad_norm": 3.6084985733032227, "learning_rate": 0.00019993818232947325, "loss": 10.2076, "step": 143 }, { "epoch": 2.056387665198238, "grad_norm": 2.93530535697937, "learning_rate": 0.00019993488631562235, "loss": 10.0593, "step": 144 }, { "epoch": 2.0704845814977975, "grad_norm": 3.2223377227783203, "learning_rate": 0.00019993150472866856, "loss": 10.1271, "step": 145 }, { "epoch": 2.0845814977973567, "grad_norm": 2.4636664390563965, "learning_rate": 0.0001999280375715075, "loss": 9.8718, "step": 146 }, { "epoch": 2.0986784140969164, "grad_norm": 2.6860575675964355, "learning_rate": 0.00019992448484710797, "loss": 9.987, "step": 147 }, { "epoch": 2.1127753303964756, "grad_norm": 2.3808062076568604, "learning_rate": 0.00019992084655851225, "loss": 9.9614, "step": 148 }, { "epoch": 2.1268722466960353, "grad_norm": 3.469188928604126, "learning_rate": 0.00019991712270883566, "loss": 10.0988, "step": 149 }, { "epoch": 2.140969162995595, "grad_norm": 2.604292869567871, "learning_rate": 0.00019991331330126698, "loss": 10.0196, "step": 150 }, { "epoch": 2.155066079295154, "grad_norm": 2.6992404460906982, "learning_rate": 0.0001999094183390681, "loss": 10.0096, "step": 151 }, { "epoch": 2.169162995594714, "grad_norm": 2.8032782077789307, "learning_rate": 0.00019990543782557425, "loss": 9.9176, "step": 152 }, { "epoch": 2.183259911894273, "grad_norm": 3.0531551837921143, "learning_rate": 0.0001999013717641939, "loss": 10.0502, "step": 153 }, { "epoch": 2.1973568281938327, "grad_norm": 3.207096576690674, "learning_rate": 0.0001998972201584088, "loss": 9.9995, "step": 154 }, { "epoch": 2.211453744493392, "grad_norm": 2.7035439014434814, "learning_rate": 0.00019989298301177386, "loss": 9.9857, "step": 155 }, { "epoch": 2.2255506607929516, "grad_norm": 3.8166487216949463, "learning_rate": 0.00019988866032791735, "loss": 9.8178, "step": 156 }, { "epoch": 2.239647577092511, "grad_norm": 3.5100674629211426, "learning_rate": 0.00019988425211054072, "loss": 9.9586, "step": 157 }, { "epoch": 2.2537444933920705, "grad_norm": 3.1917805671691895, "learning_rate": 0.00019987975836341867, "loss": 9.9816, "step": 158 }, { "epoch": 2.2678414096916297, "grad_norm": 3.027726411819458, "learning_rate": 0.00019987517909039913, "loss": 9.9242, "step": 159 }, { "epoch": 2.2819383259911894, "grad_norm": 3.144287347793579, "learning_rate": 0.00019987051429540332, "loss": 9.9424, "step": 160 }, { "epoch": 2.296035242290749, "grad_norm": 4.13268518447876, "learning_rate": 0.00019986576398242566, "loss": 10.0038, "step": 161 }, { "epoch": 2.3101321585903083, "grad_norm": 3.1613221168518066, "learning_rate": 0.00019986092815553371, "loss": 10.0293, "step": 162 }, { "epoch": 2.324229074889868, "grad_norm": 4.222150802612305, "learning_rate": 0.00019985600681886843, "loss": 9.8698, "step": 163 }, { "epoch": 2.338325991189427, "grad_norm": 3.2166895866394043, "learning_rate": 0.00019985099997664384, "loss": 10.0137, "step": 164 }, { "epoch": 2.352422907488987, "grad_norm": 2.908047914505005, "learning_rate": 0.00019984590763314722, "loss": 10.0545, "step": 165 }, { "epoch": 2.366519823788546, "grad_norm": 3.1424238681793213, "learning_rate": 0.00019984072979273916, "loss": 9.8177, "step": 166 }, { "epoch": 2.3806167400881058, "grad_norm": 2.9579007625579834, "learning_rate": 0.00019983546645985337, "loss": 10.0976, "step": 167 }, { "epoch": 2.3947136563876654, "grad_norm": 3.5854902267456055, "learning_rate": 0.00019983011763899673, "loss": 9.7573, "step": 168 }, { "epoch": 2.4088105726872246, "grad_norm": 3.6301400661468506, "learning_rate": 0.0001998246833347494, "loss": 10.0005, "step": 169 }, { "epoch": 2.4229074889867843, "grad_norm": 3.2005369663238525, "learning_rate": 0.00019981916355176472, "loss": 9.8448, "step": 170 }, { "epoch": 2.4370044052863435, "grad_norm": 2.8126308917999268, "learning_rate": 0.00019981355829476924, "loss": 9.8087, "step": 171 }, { "epoch": 2.451101321585903, "grad_norm": 3.3285417556762695, "learning_rate": 0.00019980786756856268, "loss": 10.0981, "step": 172 }, { "epoch": 2.4651982378854624, "grad_norm": 2.7718820571899414, "learning_rate": 0.00019980209137801787, "loss": 9.986, "step": 173 }, { "epoch": 2.479295154185022, "grad_norm": 3.042482852935791, "learning_rate": 0.00019979622972808093, "loss": 10.1273, "step": 174 }, { "epoch": 2.4933920704845813, "grad_norm": 3.333012580871582, "learning_rate": 0.00019979028262377118, "loss": 9.8305, "step": 175 }, { "epoch": 2.507488986784141, "grad_norm": 3.3935470581054688, "learning_rate": 0.000199784250070181, "loss": 10.0077, "step": 176 }, { "epoch": 2.5215859030837002, "grad_norm": 3.441991090774536, "learning_rate": 0.00019977813207247599, "loss": 9.8618, "step": 177 }, { "epoch": 2.53568281938326, "grad_norm": 3.071953058242798, "learning_rate": 0.00019977192863589497, "loss": 9.7865, "step": 178 }, { "epoch": 2.5497797356828196, "grad_norm": 3.2733631134033203, "learning_rate": 0.00019976563976574987, "loss": 9.9491, "step": 179 }, { "epoch": 2.563876651982379, "grad_norm": 2.9177439212799072, "learning_rate": 0.00019975926546742572, "loss": 10.009, "step": 180 }, { "epoch": 2.5779735682819385, "grad_norm": 3.273263454437256, "learning_rate": 0.0001997528057463808, "loss": 9.9371, "step": 181 }, { "epoch": 2.5920704845814977, "grad_norm": 2.8411805629730225, "learning_rate": 0.00019974626060814647, "loss": 9.9862, "step": 182 }, { "epoch": 2.6061674008810574, "grad_norm": 3.0905206203460693, "learning_rate": 0.00019973963005832728, "loss": 9.9394, "step": 183 }, { "epoch": 2.6202643171806166, "grad_norm": 3.153876304626465, "learning_rate": 0.00019973291410260092, "loss": 9.9242, "step": 184 }, { "epoch": 2.6343612334801763, "grad_norm": 3.187898874282837, "learning_rate": 0.00019972611274671814, "loss": 9.7442, "step": 185 }, { "epoch": 2.648458149779736, "grad_norm": 3.085737466812134, "learning_rate": 0.00019971922599650285, "loss": 9.7668, "step": 186 }, { "epoch": 2.662555066079295, "grad_norm": 3.2833645343780518, "learning_rate": 0.00019971225385785214, "loss": 9.9399, "step": 187 }, { "epoch": 2.6766519823788544, "grad_norm": 2.986659049987793, "learning_rate": 0.00019970519633673618, "loss": 9.8716, "step": 188 }, { "epoch": 2.690748898678414, "grad_norm": 3.177847146987915, "learning_rate": 0.00019969805343919821, "loss": 10.1159, "step": 189 }, { "epoch": 2.7048458149779737, "grad_norm": 2.9489080905914307, "learning_rate": 0.00019969082517135462, "loss": 9.7865, "step": 190 }, { "epoch": 2.718942731277533, "grad_norm": 3.4162466526031494, "learning_rate": 0.00019968351153939491, "loss": 9.9511, "step": 191 }, { "epoch": 2.7330396475770926, "grad_norm": 3.133100986480713, "learning_rate": 0.00019967611254958164, "loss": 9.848, "step": 192 }, { "epoch": 2.747136563876652, "grad_norm": 3.374758720397949, "learning_rate": 0.00019966862820825046, "loss": 9.8219, "step": 193 }, { "epoch": 2.7612334801762115, "grad_norm": 2.573345899581909, "learning_rate": 0.00019966105852181019, "loss": 9.896, "step": 194 }, { "epoch": 2.7753303964757707, "grad_norm": 3.4740965366363525, "learning_rate": 0.0001996534034967426, "loss": 9.9245, "step": 195 }, { "epoch": 2.7894273127753304, "grad_norm": 3.2498257160186768, "learning_rate": 0.00019964566313960264, "loss": 9.944, "step": 196 }, { "epoch": 2.80352422907489, "grad_norm": 3.3800137042999268, "learning_rate": 0.0001996378374570183, "loss": 9.7894, "step": 197 }, { "epoch": 2.8176211453744493, "grad_norm": 2.661165714263916, "learning_rate": 0.0001996299264556906, "loss": 9.7142, "step": 198 }, { "epoch": 2.831718061674009, "grad_norm": 3.1583774089813232, "learning_rate": 0.00019962193014239361, "loss": 9.9219, "step": 199 }, { "epoch": 2.845814977973568, "grad_norm": 2.9617223739624023, "learning_rate": 0.00019961384852397457, "loss": 10.0413, "step": 200 }, { "epoch": 2.859911894273128, "grad_norm": 3.100464344024658, "learning_rate": 0.0001996056816073536, "loss": 9.9657, "step": 201 }, { "epoch": 2.874008810572687, "grad_norm": 2.5404677391052246, "learning_rate": 0.00019959742939952392, "loss": 10.0007, "step": 202 }, { "epoch": 2.8881057268722468, "grad_norm": 3.0879364013671875, "learning_rate": 0.00019958909190755187, "loss": 9.9011, "step": 203 }, { "epoch": 2.9022026431718064, "grad_norm": 3.4870705604553223, "learning_rate": 0.0001995806691385767, "loss": 9.852, "step": 204 }, { "epoch": 2.9162995594713657, "grad_norm": 3.1029789447784424, "learning_rate": 0.00019957216109981078, "loss": 10.1656, "step": 205 }, { "epoch": 2.930396475770925, "grad_norm": 3.242420196533203, "learning_rate": 0.00019956356779853943, "loss": 9.8272, "step": 206 }, { "epoch": 2.9444933920704845, "grad_norm": 2.764042377471924, "learning_rate": 0.00019955488924212095, "loss": 10.0304, "step": 207 }, { "epoch": 2.958590308370044, "grad_norm": 3.1481311321258545, "learning_rate": 0.00019954612543798673, "loss": 9.9382, "step": 208 }, { "epoch": 2.9726872246696034, "grad_norm": 2.8540408611297607, "learning_rate": 0.00019953727639364109, "loss": 9.9235, "step": 209 }, { "epoch": 2.986784140969163, "grad_norm": 3.19551944732666, "learning_rate": 0.0001995283421166614, "loss": 9.9544, "step": 210 }, { "epoch": 3.0140969162995597, "grad_norm": 2.896125555038452, "learning_rate": 0.000199519322614698, "loss": 17.9678, "step": 211 }, { "epoch": 3.028193832599119, "grad_norm": 2.991377830505371, "learning_rate": 0.00019951021789547412, "loss": 9.7624, "step": 212 }, { "epoch": 3.0422907488986786, "grad_norm": 2.9366540908813477, "learning_rate": 0.00019950102796678602, "loss": 9.8704, "step": 213 }, { "epoch": 3.056387665198238, "grad_norm": 3.179030418395996, "learning_rate": 0.000199491752836503, "loss": 9.7782, "step": 214 }, { "epoch": 3.0704845814977975, "grad_norm": 3.450392007827759, "learning_rate": 0.00019948239251256722, "loss": 9.7195, "step": 215 }, { "epoch": 3.0845814977973567, "grad_norm": 2.9768776893615723, "learning_rate": 0.0001994729470029938, "loss": 9.7533, "step": 216 }, { "epoch": 3.0986784140969164, "grad_norm": 3.192554235458374, "learning_rate": 0.00019946341631587087, "loss": 9.719, "step": 217 }, { "epoch": 3.1127753303964756, "grad_norm": 3.0559046268463135, "learning_rate": 0.0001994538004593594, "loss": 9.7368, "step": 218 }, { "epoch": 3.1268722466960353, "grad_norm": 3.1397745609283447, "learning_rate": 0.00019944409944169333, "loss": 9.7003, "step": 219 }, { "epoch": 3.140969162995595, "grad_norm": 3.4222705364227295, "learning_rate": 0.00019943431327117957, "loss": 9.734, "step": 220 }, { "epoch": 3.155066079295154, "grad_norm": 4.023693561553955, "learning_rate": 0.0001994244419561979, "loss": 9.6702, "step": 221 }, { "epoch": 3.169162995594714, "grad_norm": 2.9380950927734375, "learning_rate": 0.00019941448550520098, "loss": 9.6678, "step": 222 }, { "epoch": 3.183259911894273, "grad_norm": 3.2084286212921143, "learning_rate": 0.00019940444392671446, "loss": 9.7766, "step": 223 }, { "epoch": 3.1973568281938327, "grad_norm": 2.946013927459717, "learning_rate": 0.0001993943172293368, "loss": 9.7599, "step": 224 }, { "epoch": 3.211453744493392, "grad_norm": 3.555432081222534, "learning_rate": 0.00019938410542173936, "loss": 9.8346, "step": 225 }, { "epoch": 3.2255506607929516, "grad_norm": 3.159590244293213, "learning_rate": 0.0001993738085126664, "loss": 9.8569, "step": 226 }, { "epoch": 3.239647577092511, "grad_norm": 4.325509071350098, "learning_rate": 0.00019936342651093506, "loss": 9.6295, "step": 227 }, { "epoch": 3.2537444933920705, "grad_norm": 3.5589067935943604, "learning_rate": 0.00019935295942543534, "loss": 9.7348, "step": 228 }, { "epoch": 3.2678414096916297, "grad_norm": 3.7693979740142822, "learning_rate": 0.00019934240726513004, "loss": 9.7042, "step": 229 }, { "epoch": 3.2819383259911894, "grad_norm": 3.5589675903320312, "learning_rate": 0.00019933177003905492, "loss": 9.6936, "step": 230 }, { "epoch": 3.296035242290749, "grad_norm": 3.078343152999878, "learning_rate": 0.00019932104775631846, "loss": 9.7912, "step": 231 }, { "epoch": 3.3101321585903083, "grad_norm": 3.709376096725464, "learning_rate": 0.00019931024042610206, "loss": 9.788, "step": 232 }, { "epoch": 3.324229074889868, "grad_norm": 4.628760814666748, "learning_rate": 0.00019929934805765987, "loss": 9.7013, "step": 233 }, { "epoch": 3.338325991189427, "grad_norm": 3.359744071960449, "learning_rate": 0.00019928837066031897, "loss": 9.815, "step": 234 }, { "epoch": 3.352422907488987, "grad_norm": 3.485650062561035, "learning_rate": 0.00019927730824347915, "loss": 9.7852, "step": 235 }, { "epoch": 3.366519823788546, "grad_norm": 2.9823124408721924, "learning_rate": 0.000199266160816613, "loss": 9.8472, "step": 236 }, { "epoch": 3.3806167400881058, "grad_norm": 3.4406113624572754, "learning_rate": 0.00019925492838926596, "loss": 9.6218, "step": 237 }, { "epoch": 3.3947136563876654, "grad_norm": 3.544761896133423, "learning_rate": 0.00019924361097105623, "loss": 9.9193, "step": 238 }, { "epoch": 3.4088105726872246, "grad_norm": 3.1024372577667236, "learning_rate": 0.0001992322085716748, "loss": 9.8453, "step": 239 }, { "epoch": 3.4229074889867843, "grad_norm": 3.7376866340637207, "learning_rate": 0.00019922072120088537, "loss": 9.667, "step": 240 }, { "epoch": 3.4370044052863435, "grad_norm": 3.9089548587799072, "learning_rate": 0.00019920914886852452, "loss": 9.7927, "step": 241 }, { "epoch": 3.451101321585903, "grad_norm": 3.4802045822143555, "learning_rate": 0.0001991974915845014, "loss": 9.9255, "step": 242 }, { "epoch": 3.4651982378854624, "grad_norm": 3.8089945316314697, "learning_rate": 0.0001991857493587981, "loss": 9.5364, "step": 243 }, { "epoch": 3.479295154185022, "grad_norm": 3.6338443756103516, "learning_rate": 0.00019917392220146932, "loss": 9.7588, "step": 244 }, { "epoch": 3.4933920704845813, "grad_norm": 3.5499141216278076, "learning_rate": 0.00019916201012264254, "loss": 9.7192, "step": 245 }, { "epoch": 3.507488986784141, "grad_norm": 3.4579482078552246, "learning_rate": 0.00019915001313251788, "loss": 9.5414, "step": 246 }, { "epoch": 3.5215859030837002, "grad_norm": 3.514281749725342, "learning_rate": 0.00019913793124136828, "loss": 9.8118, "step": 247 }, { "epoch": 3.53568281938326, "grad_norm": 3.592318058013916, "learning_rate": 0.00019912576445953932, "loss": 9.9603, "step": 248 }, { "epoch": 3.5497797356828196, "grad_norm": 3.054443597793579, "learning_rate": 0.00019911351279744928, "loss": 9.6886, "step": 249 }, { "epoch": 3.563876651982379, "grad_norm": 3.3153512477874756, "learning_rate": 0.0001991011762655891, "loss": 9.6504, "step": 250 }, { "epoch": 3.563876651982379, "eval_loss": 0.6294392347335815, "eval_runtime": 136.3639, "eval_samples_per_second": 13.259, "eval_steps_per_second": 0.418, "step": 250 }, { "epoch": 3.5779735682819385, "grad_norm": 3.2399954795837402, "learning_rate": 0.0001990887548745224, "loss": 9.8208, "step": 251 }, { "epoch": 3.5920704845814977, "grad_norm": 2.728114128112793, "learning_rate": 0.0001990762486348855, "loss": 9.5707, "step": 252 }, { "epoch": 3.6061674008810574, "grad_norm": 3.2355096340179443, "learning_rate": 0.00019906365755738734, "loss": 9.8348, "step": 253 }, { "epoch": 3.6202643171806166, "grad_norm": 3.731703519821167, "learning_rate": 0.00019905098165280954, "loss": 9.7697, "step": 254 }, { "epoch": 3.6343612334801763, "grad_norm": 2.8202106952667236, "learning_rate": 0.00019903822093200627, "loss": 9.9475, "step": 255 }, { "epoch": 3.648458149779736, "grad_norm": 3.619966745376587, "learning_rate": 0.00019902537540590447, "loss": 10.1378, "step": 256 }, { "epoch": 3.662555066079295, "grad_norm": 3.437009811401367, "learning_rate": 0.00019901244508550354, "loss": 9.767, "step": 257 }, { "epoch": 3.6766519823788544, "grad_norm": 3.1563894748687744, "learning_rate": 0.00019899942998187565, "loss": 9.7646, "step": 258 }, { "epoch": 3.690748898678414, "grad_norm": 3.147744655609131, "learning_rate": 0.00019898633010616542, "loss": 9.896, "step": 259 }, { "epoch": 3.7048458149779737, "grad_norm": 3.9498751163482666, "learning_rate": 0.00019897314546959014, "loss": 9.7497, "step": 260 }, { "epoch": 3.718942731277533, "grad_norm": 3.1575849056243896, "learning_rate": 0.00019895987608343967, "loss": 9.7217, "step": 261 }, { "epoch": 3.7330396475770926, "grad_norm": 4.008087158203125, "learning_rate": 0.00019894652195907644, "loss": 9.7917, "step": 262 }, { "epoch": 3.747136563876652, "grad_norm": 3.300351619720459, "learning_rate": 0.0001989330831079354, "loss": 9.8147, "step": 263 }, { "epoch": 3.7612334801762115, "grad_norm": 4.144752025604248, "learning_rate": 0.00019891955954152413, "loss": 9.7403, "step": 264 }, { "epoch": 3.7753303964757707, "grad_norm": 3.4026801586151123, "learning_rate": 0.0001989059512714227, "loss": 9.7492, "step": 265 }, { "epoch": 3.7894273127753304, "grad_norm": 3.71869158744812, "learning_rate": 0.00019889225830928365, "loss": 9.6506, "step": 266 }, { "epoch": 3.80352422907489, "grad_norm": 3.1379642486572266, "learning_rate": 0.0001988784806668322, "loss": 9.7743, "step": 267 }, { "epoch": 3.8176211453744493, "grad_norm": 3.206462860107422, "learning_rate": 0.00019886461835586593, "loss": 9.7031, "step": 268 }, { "epoch": 3.831718061674009, "grad_norm": 2.9855945110321045, "learning_rate": 0.000198850671388255, "loss": 9.7138, "step": 269 }, { "epoch": 3.845814977973568, "grad_norm": 3.1711626052856445, "learning_rate": 0.00019883663977594208, "loss": 9.7146, "step": 270 }, { "epoch": 3.859911894273128, "grad_norm": 2.8290717601776123, "learning_rate": 0.00019882252353094218, "loss": 9.7228, "step": 271 }, { "epoch": 3.874008810572687, "grad_norm": 3.1588993072509766, "learning_rate": 0.00019880832266534296, "loss": 9.5708, "step": 272 }, { "epoch": 3.8881057268722468, "grad_norm": 3.6332509517669678, "learning_rate": 0.0001987940371913044, "loss": 9.6363, "step": 273 }, { "epoch": 3.9022026431718064, "grad_norm": 3.3781731128692627, "learning_rate": 0.00019877966712105905, "loss": 9.7069, "step": 274 }, { "epoch": 3.9162995594713657, "grad_norm": 3.4325976371765137, "learning_rate": 0.00019876521246691182, "loss": 9.7542, "step": 275 }, { "epoch": 3.930396475770925, "grad_norm": 3.2247672080993652, "learning_rate": 0.00019875067324124003, "loss": 9.6976, "step": 276 }, { "epoch": 3.9444933920704845, "grad_norm": 3.553872585296631, "learning_rate": 0.0001987360494564935, "loss": 9.6936, "step": 277 }, { "epoch": 3.958590308370044, "grad_norm": 3.7139675617218018, "learning_rate": 0.00019872134112519437, "loss": 9.7387, "step": 278 }, { "epoch": 3.9726872246696034, "grad_norm": 4.011850833892822, "learning_rate": 0.00019870654825993725, "loss": 9.7724, "step": 279 }, { "epoch": 3.986784140969163, "grad_norm": 3.060114622116089, "learning_rate": 0.00019869167087338907, "loss": 9.7491, "step": 280 }, { "epoch": 4.014096916299559, "grad_norm": 3.596938133239746, "learning_rate": 0.0001986767089782892, "loss": 17.4964, "step": 281 }, { "epoch": 4.028193832599119, "grad_norm": 3.9812638759613037, "learning_rate": 0.0001986616625874493, "loss": 9.6387, "step": 282 }, { "epoch": 4.042290748898679, "grad_norm": 3.273632287979126, "learning_rate": 0.00019864653171375348, "loss": 9.3872, "step": 283 }, { "epoch": 4.056387665198238, "grad_norm": 3.593200445175171, "learning_rate": 0.00019863131637015807, "loss": 9.7626, "step": 284 }, { "epoch": 4.070484581497797, "grad_norm": 4.051459312438965, "learning_rate": 0.00019861601656969184, "loss": 9.434, "step": 285 }, { "epoch": 4.084581497797357, "grad_norm": 3.843174457550049, "learning_rate": 0.0001986006323254558, "loss": 9.5845, "step": 286 }, { "epoch": 4.098678414096916, "grad_norm": 3.5394649505615234, "learning_rate": 0.00019858516365062334, "loss": 9.4598, "step": 287 }, { "epoch": 4.112775330396476, "grad_norm": 3.6975784301757812, "learning_rate": 0.00019856961055844007, "loss": 9.6501, "step": 288 }, { "epoch": 4.126872246696035, "grad_norm": 3.8248095512390137, "learning_rate": 0.000198553973062224, "loss": 9.5208, "step": 289 }, { "epoch": 4.140969162995595, "grad_norm": 4.02276611328125, "learning_rate": 0.00019853825117536523, "loss": 9.5449, "step": 290 }, { "epoch": 4.155066079295154, "grad_norm": 4.094560623168945, "learning_rate": 0.0001985224449113263, "loss": 9.4726, "step": 291 }, { "epoch": 4.169162995594713, "grad_norm": 4.295591354370117, "learning_rate": 0.00019850655428364192, "loss": 9.6427, "step": 292 }, { "epoch": 4.1832599118942735, "grad_norm": 3.3288402557373047, "learning_rate": 0.00019849057930591908, "loss": 9.5837, "step": 293 }, { "epoch": 4.197356828193833, "grad_norm": 3.523481845855713, "learning_rate": 0.00019847451999183694, "loss": 9.485, "step": 294 }, { "epoch": 4.211453744493392, "grad_norm": 3.4792914390563965, "learning_rate": 0.0001984583763551469, "loss": 9.333, "step": 295 }, { "epoch": 4.225550660792951, "grad_norm": 3.8460693359375, "learning_rate": 0.00019844214840967258, "loss": 9.5288, "step": 296 }, { "epoch": 4.239647577092511, "grad_norm": 3.6774895191192627, "learning_rate": 0.00019842583616930983, "loss": 9.5471, "step": 297 }, { "epoch": 4.2537444933920705, "grad_norm": 3.3646087646484375, "learning_rate": 0.00019840943964802657, "loss": 9.5561, "step": 298 }, { "epoch": 4.26784140969163, "grad_norm": 4.037041187286377, "learning_rate": 0.00019839295885986296, "loss": 9.6223, "step": 299 }, { "epoch": 4.28193832599119, "grad_norm": 3.609894275665283, "learning_rate": 0.00019837639381893135, "loss": 9.641, "step": 300 }, { "epoch": 4.296035242290749, "grad_norm": 3.5577826499938965, "learning_rate": 0.0001983597445394162, "loss": 9.5058, "step": 301 }, { "epoch": 4.310132158590308, "grad_norm": 3.534014940261841, "learning_rate": 0.00019834301103557403, "loss": 9.4143, "step": 302 }, { "epoch": 4.3242290748898675, "grad_norm": 3.9143476486206055, "learning_rate": 0.00019832619332173365, "loss": 9.5105, "step": 303 }, { "epoch": 4.338325991189428, "grad_norm": 4.287281513214111, "learning_rate": 0.0001983092914122958, "loss": 9.6153, "step": 304 }, { "epoch": 4.352422907488987, "grad_norm": 3.427222728729248, "learning_rate": 0.00019829230532173337, "loss": 9.392, "step": 305 }, { "epoch": 4.366519823788546, "grad_norm": 3.778425455093384, "learning_rate": 0.00019827523506459142, "loss": 9.5968, "step": 306 }, { "epoch": 4.380616740088106, "grad_norm": 4.092582702636719, "learning_rate": 0.00019825808065548702, "loss": 9.5811, "step": 307 }, { "epoch": 4.394713656387665, "grad_norm": 4.414474010467529, "learning_rate": 0.00019824084210910925, "loss": 9.4194, "step": 308 }, { "epoch": 4.408810572687225, "grad_norm": 4.16365909576416, "learning_rate": 0.00019822351944021927, "loss": 9.528, "step": 309 }, { "epoch": 4.422907488986784, "grad_norm": 4.132477283477783, "learning_rate": 0.00019820611266365036, "loss": 9.5025, "step": 310 }, { "epoch": 4.437004405286344, "grad_norm": 3.404025077819824, "learning_rate": 0.0001981886217943077, "loss": 9.6416, "step": 311 }, { "epoch": 4.451101321585903, "grad_norm": 4.32616662979126, "learning_rate": 0.00019817104684716848, "loss": 9.5586, "step": 312 }, { "epoch": 4.465198237885462, "grad_norm": 4.17893648147583, "learning_rate": 0.000198153387837282, "loss": 9.4776, "step": 313 }, { "epoch": 4.479295154185022, "grad_norm": 3.971237897872925, "learning_rate": 0.0001981356447797694, "loss": 9.4781, "step": 314 }, { "epoch": 4.493392070484582, "grad_norm": 3.895318031311035, "learning_rate": 0.0001981178176898239, "loss": 9.6297, "step": 315 }, { "epoch": 4.507488986784141, "grad_norm": 3.7573413848876953, "learning_rate": 0.00019809990658271065, "loss": 9.7176, "step": 316 }, { "epoch": 4.5215859030837, "grad_norm": 3.3673951625823975, "learning_rate": 0.00019808191147376668, "loss": 9.6484, "step": 317 }, { "epoch": 4.5356828193832595, "grad_norm": 4.173061847686768, "learning_rate": 0.00019806383237840102, "loss": 9.3998, "step": 318 }, { "epoch": 4.54977973568282, "grad_norm": 4.219361305236816, "learning_rate": 0.00019804566931209462, "loss": 9.5614, "step": 319 }, { "epoch": 4.563876651982379, "grad_norm": 3.43196964263916, "learning_rate": 0.00019802742229040025, "loss": 9.4799, "step": 320 }, { "epoch": 4.577973568281938, "grad_norm": 4.038547992706299, "learning_rate": 0.00019800909132894264, "loss": 9.5319, "step": 321 }, { "epoch": 4.592070484581498, "grad_norm": 3.8661065101623535, "learning_rate": 0.00019799067644341844, "loss": 9.4393, "step": 322 }, { "epoch": 4.606167400881057, "grad_norm": 4.153717994689941, "learning_rate": 0.00019797217764959605, "loss": 9.5786, "step": 323 }, { "epoch": 4.620264317180617, "grad_norm": 4.384703636169434, "learning_rate": 0.0001979535949633158, "loss": 9.6981, "step": 324 }, { "epoch": 4.634361233480176, "grad_norm": 3.8012378215789795, "learning_rate": 0.0001979349284004898, "loss": 9.7701, "step": 325 }, { "epoch": 4.648458149779736, "grad_norm": 4.516049861907959, "learning_rate": 0.00019791617797710202, "loss": 9.4679, "step": 326 }, { "epoch": 4.662555066079295, "grad_norm": 3.421915054321289, "learning_rate": 0.0001978973437092083, "loss": 9.6008, "step": 327 }, { "epoch": 4.676651982378854, "grad_norm": 3.7902607917785645, "learning_rate": 0.00019787842561293614, "loss": 9.5609, "step": 328 }, { "epoch": 4.6907488986784145, "grad_norm": 4.446627616882324, "learning_rate": 0.0001978594237044849, "loss": 9.5333, "step": 329 }, { "epoch": 4.704845814977974, "grad_norm": 4.046638488769531, "learning_rate": 0.00019784033800012568, "loss": 9.5721, "step": 330 }, { "epoch": 4.718942731277533, "grad_norm": 3.7969677448272705, "learning_rate": 0.00019782116851620142, "loss": 9.5951, "step": 331 }, { "epoch": 4.733039647577092, "grad_norm": 2.9969778060913086, "learning_rate": 0.0001978019152691267, "loss": 9.722, "step": 332 }, { "epoch": 4.747136563876652, "grad_norm": 3.6474952697753906, "learning_rate": 0.00019778257827538783, "loss": 9.8127, "step": 333 }, { "epoch": 4.7612334801762115, "grad_norm": 3.5534746646881104, "learning_rate": 0.00019776315755154285, "loss": 9.451, "step": 334 }, { "epoch": 4.775330396475771, "grad_norm": 4.5194411277771, "learning_rate": 0.00019774365311422154, "loss": 9.399, "step": 335 }, { "epoch": 4.789427312775331, "grad_norm": 4.236517429351807, "learning_rate": 0.0001977240649801253, "loss": 9.541, "step": 336 }, { "epoch": 4.80352422907489, "grad_norm": 3.756326913833618, "learning_rate": 0.00019770439316602726, "loss": 9.5448, "step": 337 }, { "epoch": 4.817621145374449, "grad_norm": 3.9417104721069336, "learning_rate": 0.00019768463768877215, "loss": 9.5251, "step": 338 }, { "epoch": 4.8317180616740085, "grad_norm": 3.283564805984497, "learning_rate": 0.00019766479856527633, "loss": 9.7803, "step": 339 }, { "epoch": 4.845814977973569, "grad_norm": 3.9878556728363037, "learning_rate": 0.00019764487581252785, "loss": 9.7086, "step": 340 }, { "epoch": 4.859911894273128, "grad_norm": 3.75227952003479, "learning_rate": 0.00019762486944758636, "loss": 9.5728, "step": 341 }, { "epoch": 4.874008810572687, "grad_norm": 3.8873488903045654, "learning_rate": 0.00019760477948758303, "loss": 9.5961, "step": 342 }, { "epoch": 4.888105726872246, "grad_norm": 3.8353986740112305, "learning_rate": 0.00019758460594972068, "loss": 9.5232, "step": 343 }, { "epoch": 4.902202643171806, "grad_norm": 4.353938102722168, "learning_rate": 0.0001975643488512737, "loss": 9.6592, "step": 344 }, { "epoch": 4.916299559471366, "grad_norm": 3.4037740230560303, "learning_rate": 0.00019754400820958798, "loss": 9.4198, "step": 345 }, { "epoch": 4.930396475770925, "grad_norm": 3.627922773361206, "learning_rate": 0.00019752358404208101, "loss": 9.6106, "step": 346 }, { "epoch": 4.944493392070484, "grad_norm": 3.431326389312744, "learning_rate": 0.00019750307636624175, "loss": 9.3368, "step": 347 }, { "epoch": 4.958590308370044, "grad_norm": 4.202719688415527, "learning_rate": 0.00019748248519963068, "loss": 9.8325, "step": 348 }, { "epoch": 4.9726872246696034, "grad_norm": 3.510605573654175, "learning_rate": 0.00019746181055987983, "loss": 9.6965, "step": 349 }, { "epoch": 4.986784140969163, "grad_norm": 3.208909273147583, "learning_rate": 0.00019744105246469263, "loss": 9.6412, "step": 350 }, { "epoch": 5.014096916299559, "grad_norm": 3.790468215942383, "learning_rate": 0.00019742021093184398, "loss": 17.3577, "step": 351 }, { "epoch": 5.028193832599119, "grad_norm": 3.9231679439544678, "learning_rate": 0.00019739928597918024, "loss": 9.215, "step": 352 }, { "epoch": 5.042290748898679, "grad_norm": 4.515699863433838, "learning_rate": 0.00019737827762461926, "loss": 9.3544, "step": 353 }, { "epoch": 5.056387665198238, "grad_norm": 4.324073314666748, "learning_rate": 0.00019735718588615022, "loss": 9.3575, "step": 354 }, { "epoch": 5.070484581497797, "grad_norm": 4.908707618713379, "learning_rate": 0.00019733601078183375, "loss": 9.2496, "step": 355 }, { "epoch": 5.084581497797357, "grad_norm": 4.255324840545654, "learning_rate": 0.00019731475232980184, "loss": 9.3025, "step": 356 }, { "epoch": 5.098678414096916, "grad_norm": 4.209259033203125, "learning_rate": 0.00019729341054825782, "loss": 9.2327, "step": 357 }, { "epoch": 5.112775330396476, "grad_norm": 4.9791178703308105, "learning_rate": 0.0001972719854554765, "loss": 9.3443, "step": 358 }, { "epoch": 5.126872246696035, "grad_norm": 4.760982036590576, "learning_rate": 0.00019725047706980385, "loss": 9.301, "step": 359 }, { "epoch": 5.140969162995595, "grad_norm": 5.219407081604004, "learning_rate": 0.00019722888540965735, "loss": 9.457, "step": 360 }, { "epoch": 5.155066079295154, "grad_norm": 4.912387371063232, "learning_rate": 0.00019720721049352563, "loss": 9.3489, "step": 361 }, { "epoch": 5.169162995594713, "grad_norm": 5.138525485992432, "learning_rate": 0.0001971854523399687, "loss": 9.1823, "step": 362 }, { "epoch": 5.1832599118942735, "grad_norm": 4.226041316986084, "learning_rate": 0.0001971636109676178, "loss": 9.3478, "step": 363 }, { "epoch": 5.197356828193833, "grad_norm": 5.109675407409668, "learning_rate": 0.00019714168639517544, "loss": 9.1651, "step": 364 }, { "epoch": 5.211453744493392, "grad_norm": 4.776071071624756, "learning_rate": 0.0001971196786414154, "loss": 9.2932, "step": 365 }, { "epoch": 5.225550660792951, "grad_norm": 4.76246452331543, "learning_rate": 0.00019709758772518268, "loss": 9.2431, "step": 366 }, { "epoch": 5.239647577092511, "grad_norm": 4.798648834228516, "learning_rate": 0.00019707541366539345, "loss": 9.2855, "step": 367 }, { "epoch": 5.2537444933920705, "grad_norm": 4.231101036071777, "learning_rate": 0.00019705315648103512, "loss": 9.5153, "step": 368 }, { "epoch": 5.26784140969163, "grad_norm": 4.560917854309082, "learning_rate": 0.00019703081619116625, "loss": 9.4113, "step": 369 }, { "epoch": 5.28193832599119, "grad_norm": 4.471493244171143, "learning_rate": 0.00019700839281491662, "loss": 9.1603, "step": 370 }, { "epoch": 5.296035242290749, "grad_norm": 4.2874755859375, "learning_rate": 0.00019698588637148703, "loss": 9.2231, "step": 371 }, { "epoch": 5.310132158590308, "grad_norm": 4.241594314575195, "learning_rate": 0.0001969632968801496, "loss": 9.2733, "step": 372 }, { "epoch": 5.3242290748898675, "grad_norm": 4.267725467681885, "learning_rate": 0.0001969406243602473, "loss": 9.3382, "step": 373 }, { "epoch": 5.338325991189428, "grad_norm": 3.8417723178863525, "learning_rate": 0.00019691786883119453, "loss": 9.2337, "step": 374 }, { "epoch": 5.352422907488987, "grad_norm": 4.037524223327637, "learning_rate": 0.00019689503031247648, "loss": 9.3856, "step": 375 }, { "epoch": 5.366519823788546, "grad_norm": 4.668492317199707, "learning_rate": 0.0001968721088236495, "loss": 9.4558, "step": 376 }, { "epoch": 5.380616740088106, "grad_norm": 4.520495891571045, "learning_rate": 0.00019684910438434107, "loss": 9.1981, "step": 377 }, { "epoch": 5.394713656387665, "grad_norm": 4.189375400543213, "learning_rate": 0.0001968260170142496, "loss": 9.4403, "step": 378 }, { "epoch": 5.408810572687225, "grad_norm": 5.176163673400879, "learning_rate": 0.0001968028467331445, "loss": 9.1582, "step": 379 }, { "epoch": 5.422907488986784, "grad_norm": 4.846078395843506, "learning_rate": 0.0001967795935608663, "loss": 9.3225, "step": 380 }, { "epoch": 5.437004405286344, "grad_norm": 4.609081745147705, "learning_rate": 0.00019675625751732634, "loss": 9.1775, "step": 381 }, { "epoch": 5.451101321585903, "grad_norm": 4.308326244354248, "learning_rate": 0.00019673283862250705, "loss": 9.2942, "step": 382 }, { "epoch": 5.465198237885462, "grad_norm": 4.469168663024902, "learning_rate": 0.00019670933689646175, "loss": 9.3127, "step": 383 }, { "epoch": 5.479295154185022, "grad_norm": 4.466946125030518, "learning_rate": 0.00019668575235931473, "loss": 9.1185, "step": 384 }, { "epoch": 5.493392070484582, "grad_norm": 3.7987148761749268, "learning_rate": 0.00019666208503126112, "loss": 9.4744, "step": 385 }, { "epoch": 5.507488986784141, "grad_norm": 4.043029308319092, "learning_rate": 0.00019663833493256702, "loss": 9.4672, "step": 386 }, { "epoch": 5.5215859030837, "grad_norm": 3.8083503246307373, "learning_rate": 0.00019661450208356934, "loss": 9.3856, "step": 387 }, { "epoch": 5.5356828193832595, "grad_norm": 4.444696426391602, "learning_rate": 0.0001965905865046759, "loss": 9.5406, "step": 388 }, { "epoch": 5.54977973568282, "grad_norm": 4.390862941741943, "learning_rate": 0.0001965665882163653, "loss": 9.4209, "step": 389 }, { "epoch": 5.563876651982379, "grad_norm": 4.164951324462891, "learning_rate": 0.00019654250723918706, "loss": 9.2149, "step": 390 }, { "epoch": 5.577973568281938, "grad_norm": 4.073455810546875, "learning_rate": 0.00019651834359376137, "loss": 9.2729, "step": 391 }, { "epoch": 5.592070484581498, "grad_norm": 4.087716102600098, "learning_rate": 0.00019649409730077935, "loss": 9.2756, "step": 392 }, { "epoch": 5.606167400881057, "grad_norm": 4.7291364669799805, "learning_rate": 0.0001964697683810028, "loss": 9.3459, "step": 393 }, { "epoch": 5.620264317180617, "grad_norm": 4.607141971588135, "learning_rate": 0.00019644535685526426, "loss": 9.2914, "step": 394 }, { "epoch": 5.634361233480176, "grad_norm": 4.735684394836426, "learning_rate": 0.00019642086274446711, "loss": 9.45, "step": 395 }, { "epoch": 5.648458149779736, "grad_norm": 4.646516799926758, "learning_rate": 0.00019639628606958533, "loss": 9.2433, "step": 396 }, { "epoch": 5.662555066079295, "grad_norm": 4.174551010131836, "learning_rate": 0.00019637162685166368, "loss": 9.4785, "step": 397 }, { "epoch": 5.676651982378854, "grad_norm": 4.495853424072266, "learning_rate": 0.00019634688511181755, "loss": 9.4806, "step": 398 }, { "epoch": 5.6907488986784145, "grad_norm": 4.218163967132568, "learning_rate": 0.00019632206087123296, "loss": 9.16, "step": 399 }, { "epoch": 5.704845814977974, "grad_norm": 3.6901121139526367, "learning_rate": 0.0001962971541511667, "loss": 9.2583, "step": 400 }, { "epoch": 5.718942731277533, "grad_norm": 4.171170711517334, "learning_rate": 0.00019627216497294608, "loss": 9.4106, "step": 401 }, { "epoch": 5.733039647577092, "grad_norm": 4.007437229156494, "learning_rate": 0.00019624709335796904, "loss": 9.3917, "step": 402 }, { "epoch": 5.747136563876652, "grad_norm": 3.8369319438934326, "learning_rate": 0.00019622193932770414, "loss": 9.3099, "step": 403 }, { "epoch": 5.7612334801762115, "grad_norm": 4.527161598205566, "learning_rate": 0.00019619670290369046, "loss": 9.2704, "step": 404 }, { "epoch": 5.775330396475771, "grad_norm": 4.15477991104126, "learning_rate": 0.0001961713841075377, "loss": 9.4659, "step": 405 }, { "epoch": 5.789427312775331, "grad_norm": 6.129367828369141, "learning_rate": 0.000196145982960926, "loss": 9.2748, "step": 406 }, { "epoch": 5.80352422907489, "grad_norm": 4.138612747192383, "learning_rate": 0.00019612049948560613, "loss": 9.4526, "step": 407 }, { "epoch": 5.817621145374449, "grad_norm": 5.097776889801025, "learning_rate": 0.00019609493370339928, "loss": 9.4793, "step": 408 }, { "epoch": 5.8317180616740085, "grad_norm": 3.9001829624176025, "learning_rate": 0.00019606928563619712, "loss": 9.0722, "step": 409 }, { "epoch": 5.845814977973569, "grad_norm": 4.396252632141113, "learning_rate": 0.00019604355530596182, "loss": 9.4252, "step": 410 }, { "epoch": 5.859911894273128, "grad_norm": 3.9753623008728027, "learning_rate": 0.00019601774273472593, "loss": 9.3426, "step": 411 }, { "epoch": 5.874008810572687, "grad_norm": 4.214145183563232, "learning_rate": 0.0001959918479445925, "loss": 9.4373, "step": 412 }, { "epoch": 5.888105726872246, "grad_norm": 4.264339923858643, "learning_rate": 0.00019596587095773495, "loss": 9.5536, "step": 413 }, { "epoch": 5.902202643171806, "grad_norm": 3.9184038639068604, "learning_rate": 0.00019593981179639704, "loss": 9.5995, "step": 414 }, { "epoch": 5.916299559471366, "grad_norm": 3.6481645107269287, "learning_rate": 0.00019591367048289298, "loss": 9.4865, "step": 415 }, { "epoch": 5.930396475770925, "grad_norm": 4.168152332305908, "learning_rate": 0.0001958874470396072, "loss": 9.4023, "step": 416 }, { "epoch": 5.944493392070484, "grad_norm": 3.8270440101623535, "learning_rate": 0.00019586114148899465, "loss": 9.275, "step": 417 }, { "epoch": 5.958590308370044, "grad_norm": 3.4930975437164307, "learning_rate": 0.00019583475385358034, "loss": 9.1589, "step": 418 }, { "epoch": 5.9726872246696034, "grad_norm": 4.085812091827393, "learning_rate": 0.0001958082841559598, "loss": 9.4292, "step": 419 }, { "epoch": 5.986784140969163, "grad_norm": 3.630558729171753, "learning_rate": 0.00019578173241879872, "loss": 9.3545, "step": 420 }, { "epoch": 6.014096916299559, "grad_norm": 4.775903701782227, "learning_rate": 0.00019575509866483302, "loss": 16.6256, "step": 421 }, { "epoch": 6.028193832599119, "grad_norm": 5.301346778869629, "learning_rate": 0.0001957283829168689, "loss": 8.8308, "step": 422 }, { "epoch": 6.042290748898679, "grad_norm": 5.441682815551758, "learning_rate": 0.00019570158519778268, "loss": 8.8001, "step": 423 }, { "epoch": 6.056387665198238, "grad_norm": 6.34150505065918, "learning_rate": 0.00019567470553052103, "loss": 9.1165, "step": 424 }, { "epoch": 6.070484581497797, "grad_norm": 5.672751426696777, "learning_rate": 0.00019564774393810068, "loss": 8.9124, "step": 425 }, { "epoch": 6.084581497797357, "grad_norm": 5.714375019073486, "learning_rate": 0.00019562070044360855, "loss": 8.7891, "step": 426 }, { "epoch": 6.098678414096916, "grad_norm": 6.767154693603516, "learning_rate": 0.00019559357507020162, "loss": 8.7541, "step": 427 }, { "epoch": 6.112775330396476, "grad_norm": 5.126511573791504, "learning_rate": 0.00019556636784110707, "loss": 8.9408, "step": 428 }, { "epoch": 6.126872246696035, "grad_norm": 6.165284156799316, "learning_rate": 0.00019553907877962215, "loss": 9.0022, "step": 429 }, { "epoch": 6.140969162995595, "grad_norm": 5.180137634277344, "learning_rate": 0.00019551170790911416, "loss": 8.8983, "step": 430 }, { "epoch": 6.155066079295154, "grad_norm": 6.014778137207031, "learning_rate": 0.00019548425525302046, "loss": 8.9733, "step": 431 }, { "epoch": 6.169162995594713, "grad_norm": 5.966652870178223, "learning_rate": 0.00019545672083484846, "loss": 9.057, "step": 432 }, { "epoch": 6.1832599118942735, "grad_norm": 5.590867519378662, "learning_rate": 0.00019542910467817555, "loss": 8.6929, "step": 433 }, { "epoch": 6.197356828193833, "grad_norm": 5.276394367218018, "learning_rate": 0.00019540140680664913, "loss": 8.9276, "step": 434 }, { "epoch": 6.211453744493392, "grad_norm": 5.994762897491455, "learning_rate": 0.0001953736272439866, "loss": 9.0754, "step": 435 }, { "epoch": 6.225550660792951, "grad_norm": 6.280161380767822, "learning_rate": 0.00019534576601397524, "loss": 8.9559, "step": 436 }, { "epoch": 6.239647577092511, "grad_norm": 4.739211559295654, "learning_rate": 0.00019531782314047237, "loss": 9.0214, "step": 437 }, { "epoch": 6.2537444933920705, "grad_norm": 4.995026111602783, "learning_rate": 0.0001952897986474051, "loss": 9.1628, "step": 438 }, { "epoch": 6.26784140969163, "grad_norm": 5.560036659240723, "learning_rate": 0.00019526169255877046, "loss": 8.6848, "step": 439 }, { "epoch": 6.28193832599119, "grad_norm": 5.154997825622559, "learning_rate": 0.00019523350489863544, "loss": 9.0454, "step": 440 }, { "epoch": 6.296035242290749, "grad_norm": 5.66609001159668, "learning_rate": 0.00019520523569113677, "loss": 9.0721, "step": 441 }, { "epoch": 6.310132158590308, "grad_norm": 4.960750102996826, "learning_rate": 0.00019517688496048107, "loss": 8.8652, "step": 442 }, { "epoch": 6.3242290748898675, "grad_norm": 5.229446887969971, "learning_rate": 0.00019514845273094472, "loss": 9.0545, "step": 443 }, { "epoch": 6.338325991189428, "grad_norm": 5.101485252380371, "learning_rate": 0.00019511993902687392, "loss": 9.05, "step": 444 }, { "epoch": 6.352422907488987, "grad_norm": 4.888042449951172, "learning_rate": 0.00019509134387268463, "loss": 8.9118, "step": 445 }, { "epoch": 6.366519823788546, "grad_norm": 5.915888786315918, "learning_rate": 0.0001950626672928626, "loss": 8.8331, "step": 446 }, { "epoch": 6.380616740088106, "grad_norm": 4.6122727394104, "learning_rate": 0.0001950339093119632, "loss": 8.9114, "step": 447 }, { "epoch": 6.394713656387665, "grad_norm": 5.83273458480835, "learning_rate": 0.0001950050699546116, "loss": 9.0487, "step": 448 }, { "epoch": 6.408810572687225, "grad_norm": 5.392518043518066, "learning_rate": 0.0001949761492455026, "loss": 9.0269, "step": 449 }, { "epoch": 6.422907488986784, "grad_norm": 5.434336185455322, "learning_rate": 0.00019494714720940065, "loss": 8.9877, "step": 450 }, { "epoch": 6.437004405286344, "grad_norm": 5.196749210357666, "learning_rate": 0.00019491806387113986, "loss": 8.8804, "step": 451 }, { "epoch": 6.451101321585903, "grad_norm": 5.615591526031494, "learning_rate": 0.000194888899255624, "loss": 9.0704, "step": 452 }, { "epoch": 6.465198237885462, "grad_norm": 5.763280391693115, "learning_rate": 0.0001948596533878264, "loss": 9.02, "step": 453 }, { "epoch": 6.479295154185022, "grad_norm": 5.787596702575684, "learning_rate": 0.0001948303262927899, "loss": 9.1023, "step": 454 }, { "epoch": 6.493392070484582, "grad_norm": 5.57631778717041, "learning_rate": 0.00019480091799562704, "loss": 8.9234, "step": 455 }, { "epoch": 6.507488986784141, "grad_norm": 5.745879650115967, "learning_rate": 0.00019477142852151978, "loss": 9.1502, "step": 456 }, { "epoch": 6.5215859030837, "grad_norm": 5.161474227905273, "learning_rate": 0.00019474185789571956, "loss": 8.8646, "step": 457 }, { "epoch": 6.5356828193832595, "grad_norm": 4.835080623626709, "learning_rate": 0.00019471220614354745, "loss": 8.9387, "step": 458 }, { "epoch": 6.54977973568282, "grad_norm": 4.563788414001465, "learning_rate": 0.0001946824732903939, "loss": 8.8162, "step": 459 }, { "epoch": 6.563876651982379, "grad_norm": 4.487639427185059, "learning_rate": 0.00019465265936171873, "loss": 8.9375, "step": 460 }, { "epoch": 6.577973568281938, "grad_norm": 4.7265625, "learning_rate": 0.00019462276438305133, "loss": 9.0045, "step": 461 }, { "epoch": 6.592070484581498, "grad_norm": 4.520985126495361, "learning_rate": 0.00019459278837999046, "loss": 8.9477, "step": 462 }, { "epoch": 6.606167400881057, "grad_norm": 4.460409641265869, "learning_rate": 0.0001945627313782042, "loss": 8.9606, "step": 463 }, { "epoch": 6.620264317180617, "grad_norm": 4.703360080718994, "learning_rate": 0.00019453259340342996, "loss": 9.2178, "step": 464 }, { "epoch": 6.634361233480176, "grad_norm": 4.667878150939941, "learning_rate": 0.00019450237448147463, "loss": 8.8775, "step": 465 }, { "epoch": 6.648458149779736, "grad_norm": 4.605958461761475, "learning_rate": 0.00019447207463821431, "loss": 9.1204, "step": 466 }, { "epoch": 6.662555066079295, "grad_norm": 4.3512725830078125, "learning_rate": 0.00019444169389959436, "loss": 9.0527, "step": 467 }, { "epoch": 6.676651982378854, "grad_norm": 4.3367600440979, "learning_rate": 0.0001944112322916295, "loss": 8.9816, "step": 468 }, { "epoch": 6.6907488986784145, "grad_norm": 4.650542736053467, "learning_rate": 0.00019438068984040365, "loss": 9.166, "step": 469 }, { "epoch": 6.704845814977974, "grad_norm": 4.754465579986572, "learning_rate": 0.00019435006657206997, "loss": 9.0638, "step": 470 }, { "epoch": 6.718942731277533, "grad_norm": 5.208463668823242, "learning_rate": 0.0001943193625128508, "loss": 9.0653, "step": 471 }, { "epoch": 6.733039647577092, "grad_norm": 4.631998062133789, "learning_rate": 0.0001942885776890377, "loss": 9.0416, "step": 472 }, { "epoch": 6.747136563876652, "grad_norm": 6.289946556091309, "learning_rate": 0.00019425771212699133, "loss": 9.0383, "step": 473 }, { "epoch": 6.7612334801762115, "grad_norm": 4.989789962768555, "learning_rate": 0.00019422676585314155, "loss": 9.0609, "step": 474 }, { "epoch": 6.775330396475771, "grad_norm": 5.379055976867676, "learning_rate": 0.0001941957388939873, "loss": 8.9956, "step": 475 }, { "epoch": 6.789427312775331, "grad_norm": 5.480366230010986, "learning_rate": 0.00019416463127609656, "loss": 9.0921, "step": 476 }, { "epoch": 6.80352422907489, "grad_norm": 5.164433479309082, "learning_rate": 0.00019413344302610647, "loss": 9.122, "step": 477 }, { "epoch": 6.817621145374449, "grad_norm": 4.417089462280273, "learning_rate": 0.00019410217417072318, "loss": 9.0872, "step": 478 }, { "epoch": 6.8317180616740085, "grad_norm": 6.328742980957031, "learning_rate": 0.00019407082473672186, "loss": 9.0613, "step": 479 }, { "epoch": 6.845814977973569, "grad_norm": 5.197303295135498, "learning_rate": 0.00019403939475094663, "loss": 9.1049, "step": 480 }, { "epoch": 6.859911894273128, "grad_norm": 5.625271320343018, "learning_rate": 0.00019400788424031067, "loss": 9.0794, "step": 481 }, { "epoch": 6.874008810572687, "grad_norm": 5.322677135467529, "learning_rate": 0.00019397629323179608, "loss": 9.1214, "step": 482 }, { "epoch": 6.888105726872246, "grad_norm": 4.81593656539917, "learning_rate": 0.00019394462175245381, "loss": 8.9775, "step": 483 }, { "epoch": 6.902202643171806, "grad_norm": 4.784323215484619, "learning_rate": 0.00019391286982940384, "loss": 8.9392, "step": 484 }, { "epoch": 6.916299559471366, "grad_norm": 4.977121829986572, "learning_rate": 0.00019388103748983496, "loss": 9.1445, "step": 485 }, { "epoch": 6.930396475770925, "grad_norm": 4.749843597412109, "learning_rate": 0.0001938491247610049, "loss": 9.1428, "step": 486 }, { "epoch": 6.944493392070484, "grad_norm": 4.57659387588501, "learning_rate": 0.00019381713167024005, "loss": 8.9751, "step": 487 }, { "epoch": 6.958590308370044, "grad_norm": 4.320965766906738, "learning_rate": 0.0001937850582449358, "loss": 8.9299, "step": 488 }, { "epoch": 6.9726872246696034, "grad_norm": 4.675748825073242, "learning_rate": 0.00019375290451255632, "loss": 8.8346, "step": 489 }, { "epoch": 6.986784140969163, "grad_norm": 6.707590579986572, "learning_rate": 0.00019372067050063438, "loss": 9.3676, "step": 490 }, { "epoch": 7.014096916299559, "grad_norm": 7.20290470123291, "learning_rate": 0.00019368835623677163, "loss": 15.8926, "step": 491 }, { "epoch": 7.028193832599119, "grad_norm": 6.21053409576416, "learning_rate": 0.00019365596174863837, "loss": 8.4507, "step": 492 }, { "epoch": 7.042290748898679, "grad_norm": 11.533875465393066, "learning_rate": 0.00019362348706397373, "loss": 8.178, "step": 493 }, { "epoch": 7.056387665198238, "grad_norm": 8.650775909423828, "learning_rate": 0.00019359093221058533, "loss": 8.2813, "step": 494 }, { "epoch": 7.070484581497797, "grad_norm": 7.2661590576171875, "learning_rate": 0.00019355829721634952, "loss": 8.6063, "step": 495 }, { "epoch": 7.084581497797357, "grad_norm": 6.770198345184326, "learning_rate": 0.0001935255821092113, "loss": 8.3977, "step": 496 }, { "epoch": 7.098678414096916, "grad_norm": 6.8975019454956055, "learning_rate": 0.00019349278691718427, "loss": 8.2608, "step": 497 }, { "epoch": 7.112775330396476, "grad_norm": 6.970608234405518, "learning_rate": 0.0001934599116683505, "loss": 8.3164, "step": 498 }, { "epoch": 7.126872246696035, "grad_norm": 6.482853412628174, "learning_rate": 0.00019342695639086078, "loss": 8.194, "step": 499 }, { "epoch": 7.140969162995595, "grad_norm": 7.051389694213867, "learning_rate": 0.00019339392111293425, "loss": 8.2495, "step": 500 }, { "epoch": 7.140969162995595, "eval_loss": 0.6698388457298279, "eval_runtime": 132.5427, "eval_samples_per_second": 13.641, "eval_steps_per_second": 0.43, "step": 500 }, { "epoch": 7.155066079295154, "grad_norm": 6.759793758392334, "learning_rate": 0.0001933608058628587, "loss": 8.0665, "step": 501 }, { "epoch": 7.169162995594713, "grad_norm": 6.561203956604004, "learning_rate": 0.00019332761066899034, "loss": 8.2089, "step": 502 }, { "epoch": 7.1832599118942735, "grad_norm": 7.009693145751953, "learning_rate": 0.00019329433555975383, "loss": 8.4274, "step": 503 }, { "epoch": 7.197356828193833, "grad_norm": 6.348790645599365, "learning_rate": 0.00019326098056364222, "loss": 8.2828, "step": 504 }, { "epoch": 7.211453744493392, "grad_norm": 5.958973407745361, "learning_rate": 0.0001932275457092171, "loss": 8.1071, "step": 505 }, { "epoch": 7.225550660792951, "grad_norm": 6.497834205627441, "learning_rate": 0.00019319403102510826, "loss": 8.2702, "step": 506 }, { "epoch": 7.239647577092511, "grad_norm": 6.634936809539795, "learning_rate": 0.00019316043654001402, "loss": 8.5432, "step": 507 }, { "epoch": 7.2537444933920705, "grad_norm": 6.98307991027832, "learning_rate": 0.0001931267622827009, "loss": 8.3659, "step": 508 }, { "epoch": 7.26784140969163, "grad_norm": 6.675987720489502, "learning_rate": 0.0001930930082820038, "loss": 8.5366, "step": 509 }, { "epoch": 7.28193832599119, "grad_norm": 6.399505138397217, "learning_rate": 0.00019305917456682593, "loss": 8.2875, "step": 510 }, { "epoch": 7.296035242290749, "grad_norm": 6.860027313232422, "learning_rate": 0.00019302526116613864, "loss": 8.3318, "step": 511 }, { "epoch": 7.310132158590308, "grad_norm": 7.725433349609375, "learning_rate": 0.0001929912681089816, "loss": 8.3992, "step": 512 }, { "epoch": 7.3242290748898675, "grad_norm": 6.647876262664795, "learning_rate": 0.00019295719542446272, "loss": 8.6141, "step": 513 }, { "epoch": 7.338325991189428, "grad_norm": 6.3409199714660645, "learning_rate": 0.00019292304314175805, "loss": 8.5061, "step": 514 }, { "epoch": 7.352422907488987, "grad_norm": 6.377959728240967, "learning_rate": 0.00019288881129011175, "loss": 8.3775, "step": 515 }, { "epoch": 7.366519823788546, "grad_norm": 7.711176872253418, "learning_rate": 0.0001928544998988362, "loss": 8.4247, "step": 516 }, { "epoch": 7.380616740088106, "grad_norm": 6.442246913909912, "learning_rate": 0.00019282010899731187, "loss": 8.5574, "step": 517 }, { "epoch": 7.394713656387665, "grad_norm": 6.478960990905762, "learning_rate": 0.00019278563861498723, "loss": 8.7681, "step": 518 }, { "epoch": 7.408810572687225, "grad_norm": 6.030850410461426, "learning_rate": 0.00019275108878137896, "loss": 8.515, "step": 519 }, { "epoch": 7.422907488986784, "grad_norm": 6.225876808166504, "learning_rate": 0.00019271645952607163, "loss": 8.3889, "step": 520 }, { "epoch": 7.437004405286344, "grad_norm": 6.075868129730225, "learning_rate": 0.0001926817508787179, "loss": 8.516, "step": 521 }, { "epoch": 7.451101321585903, "grad_norm": 7.068554878234863, "learning_rate": 0.00019264696286903837, "loss": 8.6087, "step": 522 }, { "epoch": 7.465198237885462, "grad_norm": 6.704993724822998, "learning_rate": 0.00019261209552682163, "loss": 8.5274, "step": 523 }, { "epoch": 7.479295154185022, "grad_norm": 6.38553524017334, "learning_rate": 0.0001925771488819242, "loss": 8.5123, "step": 524 }, { "epoch": 7.493392070484582, "grad_norm": 6.526176452636719, "learning_rate": 0.00019254212296427044, "loss": 8.5231, "step": 525 }, { "epoch": 7.507488986784141, "grad_norm": 6.2339043617248535, "learning_rate": 0.0001925070178038527, "loss": 8.4596, "step": 526 }, { "epoch": 7.5215859030837, "grad_norm": 6.202362537384033, "learning_rate": 0.0001924718334307311, "loss": 8.5845, "step": 527 }, { "epoch": 7.5356828193832595, "grad_norm": 6.216184616088867, "learning_rate": 0.00019243656987503362, "loss": 8.4416, "step": 528 }, { "epoch": 7.54977973568282, "grad_norm": 5.715577125549316, "learning_rate": 0.00019240122716695607, "loss": 8.5523, "step": 529 }, { "epoch": 7.563876651982379, "grad_norm": 7.881626129150391, "learning_rate": 0.000192365805336762, "loss": 8.5166, "step": 530 }, { "epoch": 7.577973568281938, "grad_norm": 6.664647102355957, "learning_rate": 0.00019233030441478268, "loss": 8.3535, "step": 531 }, { "epoch": 7.592070484581498, "grad_norm": 6.0582098960876465, "learning_rate": 0.0001922947244314172, "loss": 8.546, "step": 532 }, { "epoch": 7.606167400881057, "grad_norm": 5.891558647155762, "learning_rate": 0.0001922590654171322, "loss": 8.8031, "step": 533 }, { "epoch": 7.620264317180617, "grad_norm": 5.382675647735596, "learning_rate": 0.00019222332740246218, "loss": 8.7072, "step": 534 }, { "epoch": 7.634361233480176, "grad_norm": 6.37630033493042, "learning_rate": 0.00019218751041800918, "loss": 8.5527, "step": 535 }, { "epoch": 7.648458149779736, "grad_norm": 6.1714887619018555, "learning_rate": 0.00019215161449444282, "loss": 8.4662, "step": 536 }, { "epoch": 7.662555066079295, "grad_norm": 5.279675006866455, "learning_rate": 0.00019211563966250043, "loss": 8.5282, "step": 537 }, { "epoch": 7.676651982378854, "grad_norm": 6.5880866050720215, "learning_rate": 0.00019207958595298684, "loss": 8.3799, "step": 538 }, { "epoch": 7.6907488986784145, "grad_norm": 5.3567423820495605, "learning_rate": 0.00019204345339677442, "loss": 8.3968, "step": 539 }, { "epoch": 7.704845814977974, "grad_norm": 7.184589385986328, "learning_rate": 0.00019200724202480304, "loss": 8.3406, "step": 540 }, { "epoch": 7.718942731277533, "grad_norm": 6.289735317230225, "learning_rate": 0.00019197095186808012, "loss": 8.6371, "step": 541 }, { "epoch": 7.733039647577092, "grad_norm": 5.729371547698975, "learning_rate": 0.0001919345829576805, "loss": 8.6288, "step": 542 }, { "epoch": 7.747136563876652, "grad_norm": 6.52308988571167, "learning_rate": 0.00019189813532474644, "loss": 8.7177, "step": 543 }, { "epoch": 7.7612334801762115, "grad_norm": 5.42264986038208, "learning_rate": 0.00019186160900048764, "loss": 8.654, "step": 544 }, { "epoch": 7.775330396475771, "grad_norm": 5.727489471435547, "learning_rate": 0.0001918250040161812, "loss": 8.526, "step": 545 }, { "epoch": 7.789427312775331, "grad_norm": 5.675527095794678, "learning_rate": 0.00019178832040317155, "loss": 8.5179, "step": 546 }, { "epoch": 7.80352422907489, "grad_norm": 5.594354629516602, "learning_rate": 0.00019175155819287039, "loss": 8.6529, "step": 547 }, { "epoch": 7.817621145374449, "grad_norm": 6.8743462562561035, "learning_rate": 0.00019171471741675683, "loss": 8.406, "step": 548 }, { "epoch": 7.8317180616740085, "grad_norm": 4.944474220275879, "learning_rate": 0.00019167779810637725, "loss": 8.6532, "step": 549 }, { "epoch": 7.845814977973569, "grad_norm": 5.712888240814209, "learning_rate": 0.00019164080029334516, "loss": 8.5152, "step": 550 }, { "epoch": 7.859911894273128, "grad_norm": 5.654259204864502, "learning_rate": 0.0001916037240093414, "loss": 8.5988, "step": 551 }, { "epoch": 7.874008810572687, "grad_norm": 5.709700107574463, "learning_rate": 0.00019156656928611394, "loss": 8.3999, "step": 552 }, { "epoch": 7.888105726872246, "grad_norm": 4.978238582611084, "learning_rate": 0.00019152933615547798, "loss": 8.4067, "step": 553 }, { "epoch": 7.902202643171806, "grad_norm": 6.009710311889648, "learning_rate": 0.00019149202464931577, "loss": 8.7723, "step": 554 }, { "epoch": 7.916299559471366, "grad_norm": 5.66488790512085, "learning_rate": 0.00019145463479957678, "loss": 8.4366, "step": 555 }, { "epoch": 7.930396475770925, "grad_norm": 6.605832576751709, "learning_rate": 0.00019141716663827755, "loss": 8.7254, "step": 556 }, { "epoch": 7.944493392070484, "grad_norm": 5.16806173324585, "learning_rate": 0.00019137962019750155, "loss": 8.8735, "step": 557 }, { "epoch": 7.958590308370044, "grad_norm": 5.620360374450684, "learning_rate": 0.0001913419955093994, "loss": 8.4956, "step": 558 }, { "epoch": 7.9726872246696034, "grad_norm": 5.7274322509765625, "learning_rate": 0.0001913042926061887, "loss": 8.6561, "step": 559 }, { "epoch": 7.986784140969163, "grad_norm": 5.550268173217773, "learning_rate": 0.00019126651152015403, "loss": 8.8168, "step": 560 }, { "epoch": 8.01409691629956, "grad_norm": 8.440166473388672, "learning_rate": 0.00019122865228364687, "loss": 14.6162, "step": 561 }, { "epoch": 8.028193832599118, "grad_norm": 8.103872299194336, "learning_rate": 0.0001911907149290857, "loss": 7.8333, "step": 562 }, { "epoch": 8.042290748898678, "grad_norm": 29.317668914794922, "learning_rate": 0.0001911526994889558, "loss": 7.8205, "step": 563 }, { "epoch": 8.056387665198239, "grad_norm": 18.910913467407227, "learning_rate": 0.00019111460599580932, "loss": 7.9295, "step": 564 }, { "epoch": 8.070484581497798, "grad_norm": 11.96757984161377, "learning_rate": 0.00019107643448226536, "loss": 7.7322, "step": 565 }, { "epoch": 8.084581497797357, "grad_norm": 10.561816215515137, "learning_rate": 0.00019103818498100977, "loss": 7.8902, "step": 566 }, { "epoch": 8.098678414096916, "grad_norm": 11.354853630065918, "learning_rate": 0.00019099985752479506, "loss": 7.7111, "step": 567 }, { "epoch": 8.112775330396476, "grad_norm": 10.522665023803711, "learning_rate": 0.00019096145214644067, "loss": 7.7833, "step": 568 }, { "epoch": 8.126872246696035, "grad_norm": 9.704154014587402, "learning_rate": 0.00019092296887883269, "loss": 7.606, "step": 569 }, { "epoch": 8.140969162995594, "grad_norm": 10.988235473632812, "learning_rate": 0.00019088440775492384, "loss": 7.5961, "step": 570 }, { "epoch": 8.155066079295155, "grad_norm": 8.485323905944824, "learning_rate": 0.00019084576880773363, "loss": 7.5116, "step": 571 }, { "epoch": 8.169162995594714, "grad_norm": 8.54230785369873, "learning_rate": 0.00019080705207034812, "loss": 7.6994, "step": 572 }, { "epoch": 8.183259911894273, "grad_norm": 11.886754035949707, "learning_rate": 0.00019076825757592003, "loss": 7.657, "step": 573 }, { "epoch": 8.197356828193833, "grad_norm": 6.951314926147461, "learning_rate": 0.00019072938535766865, "loss": 7.5714, "step": 574 }, { "epoch": 8.211453744493392, "grad_norm": 8.915898323059082, "learning_rate": 0.00019069043544887976, "loss": 7.3979, "step": 575 }, { "epoch": 8.225550660792951, "grad_norm": 9.225069046020508, "learning_rate": 0.0001906514078829058, "loss": 7.609, "step": 576 }, { "epoch": 8.23964757709251, "grad_norm": 8.531152725219727, "learning_rate": 0.00019061230269316556, "loss": 7.4573, "step": 577 }, { "epoch": 8.25374449339207, "grad_norm": 10.009847640991211, "learning_rate": 0.00019057311991314444, "loss": 7.6174, "step": 578 }, { "epoch": 8.26784140969163, "grad_norm": 8.413101196289062, "learning_rate": 0.00019053385957639414, "loss": 7.4454, "step": 579 }, { "epoch": 8.28193832599119, "grad_norm": 8.147055625915527, "learning_rate": 0.00019049452171653289, "loss": 7.6014, "step": 580 }, { "epoch": 8.296035242290749, "grad_norm": 8.17681884765625, "learning_rate": 0.0001904551063672452, "loss": 7.6878, "step": 581 }, { "epoch": 8.310132158590308, "grad_norm": 7.655210494995117, "learning_rate": 0.00019041561356228202, "loss": 7.6179, "step": 582 }, { "epoch": 8.324229074889868, "grad_norm": 9.147272109985352, "learning_rate": 0.00019037604333546057, "loss": 7.5907, "step": 583 }, { "epoch": 8.338325991189427, "grad_norm": 7.824910640716553, "learning_rate": 0.0001903363957206644, "loss": 7.8489, "step": 584 }, { "epoch": 8.352422907488986, "grad_norm": 7.852499961853027, "learning_rate": 0.00019029667075184332, "loss": 7.5135, "step": 585 }, { "epoch": 8.366519823788547, "grad_norm": 9.403107643127441, "learning_rate": 0.0001902568684630133, "loss": 7.7309, "step": 586 }, { "epoch": 8.380616740088106, "grad_norm": 7.15643835067749, "learning_rate": 0.0001902169888882567, "loss": 7.7212, "step": 587 }, { "epoch": 8.394713656387665, "grad_norm": 9.276999473571777, "learning_rate": 0.00019017703206172185, "loss": 7.6787, "step": 588 }, { "epoch": 8.408810572687225, "grad_norm": 7.227217197418213, "learning_rate": 0.00019013699801762336, "loss": 7.8809, "step": 589 }, { "epoch": 8.422907488986784, "grad_norm": 7.285519599914551, "learning_rate": 0.0001900968867902419, "loss": 7.7442, "step": 590 }, { "epoch": 8.437004405286343, "grad_norm": 7.022513389587402, "learning_rate": 0.0001900566984139243, "loss": 7.5512, "step": 591 }, { "epoch": 8.451101321585902, "grad_norm": 7.100527763366699, "learning_rate": 0.0001900164329230834, "loss": 7.7194, "step": 592 }, { "epoch": 8.465198237885463, "grad_norm": 7.4344587326049805, "learning_rate": 0.00018997609035219807, "loss": 7.6741, "step": 593 }, { "epoch": 8.479295154185023, "grad_norm": 6.638708591461182, "learning_rate": 0.00018993567073581313, "loss": 7.5402, "step": 594 }, { "epoch": 8.493392070484582, "grad_norm": 6.357802391052246, "learning_rate": 0.00018989517410853955, "loss": 7.6824, "step": 595 }, { "epoch": 8.507488986784141, "grad_norm": 6.390590190887451, "learning_rate": 0.00018985460050505406, "loss": 7.866, "step": 596 }, { "epoch": 8.5215859030837, "grad_norm": 6.874635219573975, "learning_rate": 0.00018981394996009935, "loss": 8.0733, "step": 597 }, { "epoch": 8.53568281938326, "grad_norm": 6.748425483703613, "learning_rate": 0.00018977322250848412, "loss": 7.8582, "step": 598 }, { "epoch": 8.549779735682819, "grad_norm": 7.191603183746338, "learning_rate": 0.00018973241818508272, "loss": 7.6925, "step": 599 }, { "epoch": 8.56387665198238, "grad_norm": 6.489673614501953, "learning_rate": 0.00018969153702483546, "loss": 7.794, "step": 600 }, { "epoch": 8.577973568281939, "grad_norm": 7.164043426513672, "learning_rate": 0.00018965057906274844, "loss": 7.7001, "step": 601 }, { "epoch": 8.592070484581498, "grad_norm": 8.128289222717285, "learning_rate": 0.00018960954433389345, "loss": 7.5772, "step": 602 }, { "epoch": 8.606167400881057, "grad_norm": 7.621349334716797, "learning_rate": 0.0001895684328734081, "loss": 7.6807, "step": 603 }, { "epoch": 8.620264317180617, "grad_norm": 6.644317150115967, "learning_rate": 0.00018952724471649565, "loss": 7.6566, "step": 604 }, { "epoch": 8.634361233480176, "grad_norm": 7.822019100189209, "learning_rate": 0.00018948597989842502, "loss": 7.9032, "step": 605 }, { "epoch": 8.648458149779735, "grad_norm": 6.843960762023926, "learning_rate": 0.00018944463845453087, "loss": 7.8378, "step": 606 }, { "epoch": 8.662555066079296, "grad_norm": 7.351415634155273, "learning_rate": 0.00018940322042021332, "loss": 7.7109, "step": 607 }, { "epoch": 8.676651982378855, "grad_norm": 6.609140872955322, "learning_rate": 0.00018936172583093822, "loss": 7.8385, "step": 608 }, { "epoch": 8.690748898678414, "grad_norm": 6.721433639526367, "learning_rate": 0.00018932015472223693, "loss": 7.925, "step": 609 }, { "epoch": 8.704845814977974, "grad_norm": 6.695221424102783, "learning_rate": 0.00018927850712970626, "loss": 7.8322, "step": 610 }, { "epoch": 8.718942731277533, "grad_norm": 7.858670711517334, "learning_rate": 0.00018923678308900858, "loss": 7.8941, "step": 611 }, { "epoch": 8.733039647577092, "grad_norm": 6.464662551879883, "learning_rate": 0.00018919498263587174, "loss": 7.6963, "step": 612 }, { "epoch": 8.747136563876651, "grad_norm": 7.081120491027832, "learning_rate": 0.00018915310580608897, "loss": 7.8398, "step": 613 }, { "epoch": 8.761233480176212, "grad_norm": 6.571445941925049, "learning_rate": 0.00018911115263551892, "loss": 7.8016, "step": 614 }, { "epoch": 8.775330396475772, "grad_norm": 6.400954723358154, "learning_rate": 0.0001890691231600856, "loss": 7.7911, "step": 615 }, { "epoch": 8.78942731277533, "grad_norm": 6.9759345054626465, "learning_rate": 0.0001890270174157784, "loss": 7.7435, "step": 616 }, { "epoch": 8.80352422907489, "grad_norm": 7.637103080749512, "learning_rate": 0.000188984835438652, "loss": 7.785, "step": 617 }, { "epoch": 8.81762114537445, "grad_norm": 6.8414306640625, "learning_rate": 0.0001889425772648263, "loss": 7.8851, "step": 618 }, { "epoch": 8.831718061674009, "grad_norm": 9.036650657653809, "learning_rate": 0.0001889002429304865, "loss": 7.9944, "step": 619 }, { "epoch": 8.845814977973568, "grad_norm": 6.741250514984131, "learning_rate": 0.00018885783247188305, "loss": 7.6192, "step": 620 }, { "epoch": 8.859911894273127, "grad_norm": 9.368353843688965, "learning_rate": 0.0001888153459253315, "loss": 7.8314, "step": 621 }, { "epoch": 8.874008810572688, "grad_norm": 7.046999454498291, "learning_rate": 0.0001887727833272126, "loss": 7.8828, "step": 622 }, { "epoch": 8.888105726872247, "grad_norm": 7.138089656829834, "learning_rate": 0.00018873014471397224, "loss": 7.9338, "step": 623 }, { "epoch": 8.902202643171806, "grad_norm": 6.616199970245361, "learning_rate": 0.00018868743012212137, "loss": 7.9068, "step": 624 }, { "epoch": 8.916299559471366, "grad_norm": 6.500641822814941, "learning_rate": 0.00018864463958823596, "loss": 7.696, "step": 625 }, { "epoch": 8.930396475770925, "grad_norm": 7.378822326660156, "learning_rate": 0.0001886017731489571, "loss": 7.8103, "step": 626 }, { "epoch": 8.944493392070484, "grad_norm": 6.141440391540527, "learning_rate": 0.00018855883084099086, "loss": 7.7781, "step": 627 }, { "epoch": 8.958590308370043, "grad_norm": 6.784660339355469, "learning_rate": 0.0001885158127011082, "loss": 7.9164, "step": 628 }, { "epoch": 8.972687224669604, "grad_norm": 6.786692142486572, "learning_rate": 0.0001884727187661451, "loss": 7.7007, "step": 629 }, { "epoch": 8.986784140969164, "grad_norm": 6.843912124633789, "learning_rate": 0.00018842954907300236, "loss": 8.1184, "step": 630 }, { "epoch": 9.01409691629956, "grad_norm": 7.992321491241455, "learning_rate": 0.00018838630365864572, "loss": 13.1954, "step": 631 }, { "epoch": 9.028193832599118, "grad_norm": 10.668301582336426, "learning_rate": 0.00018834298256010577, "loss": 6.5371, "step": 632 }, { "epoch": 9.042290748898678, "grad_norm": 18.920608520507812, "learning_rate": 0.00018829958581447783, "loss": 6.504, "step": 633 }, { "epoch": 9.056387665198239, "grad_norm": 19.76584243774414, "learning_rate": 0.00018825611345892203, "loss": 6.7145, "step": 634 }, { "epoch": 9.070484581497798, "grad_norm": 11.000935554504395, "learning_rate": 0.00018821256553066325, "loss": 6.4896, "step": 635 }, { "epoch": 9.084581497797357, "grad_norm": 12.650483131408691, "learning_rate": 0.00018816894206699113, "loss": 6.3591, "step": 636 }, { "epoch": 9.098678414096916, "grad_norm": 12.837759971618652, "learning_rate": 0.0001881252431052599, "loss": 6.297, "step": 637 }, { "epoch": 9.112775330396476, "grad_norm": 11.872344970703125, "learning_rate": 0.0001880814686828885, "loss": 6.5, "step": 638 }, { "epoch": 9.126872246696035, "grad_norm": 12.101139068603516, "learning_rate": 0.00018803761883736045, "loss": 6.6353, "step": 639 }, { "epoch": 9.140969162995594, "grad_norm": 9.592211723327637, "learning_rate": 0.00018799369360622393, "loss": 6.594, "step": 640 }, { "epoch": 9.155066079295155, "grad_norm": 11.978349685668945, "learning_rate": 0.00018794969302709155, "loss": 6.4848, "step": 641 }, { "epoch": 9.169162995594714, "grad_norm": 8.122666358947754, "learning_rate": 0.00018790561713764054, "loss": 6.384, "step": 642 }, { "epoch": 9.183259911894273, "grad_norm": 12.07115650177002, "learning_rate": 0.00018786146597561253, "loss": 6.597, "step": 643 }, { "epoch": 9.197356828193833, "grad_norm": 12.047696113586426, "learning_rate": 0.00018781723957881372, "loss": 6.4816, "step": 644 }, { "epoch": 9.211453744493392, "grad_norm": 9.823554992675781, "learning_rate": 0.00018777293798511466, "loss": 6.2848, "step": 645 }, { "epoch": 9.225550660792951, "grad_norm": 9.610517501831055, "learning_rate": 0.00018772856123245026, "loss": 6.4509, "step": 646 }, { "epoch": 9.23964757709251, "grad_norm": 11.417203903198242, "learning_rate": 0.00018768410935881985, "loss": 6.4307, "step": 647 }, { "epoch": 9.25374449339207, "grad_norm": 8.578927040100098, "learning_rate": 0.0001876395824022871, "loss": 6.4612, "step": 648 }, { "epoch": 9.26784140969163, "grad_norm": 10.273001670837402, "learning_rate": 0.00018759498040097985, "loss": 6.2463, "step": 649 }, { "epoch": 9.28193832599119, "grad_norm": 10.263187408447266, "learning_rate": 0.00018755030339309036, "loss": 6.6936, "step": 650 }, { "epoch": 9.296035242290749, "grad_norm": 9.82539176940918, "learning_rate": 0.000187505551416875, "loss": 6.4668, "step": 651 }, { "epoch": 9.310132158590308, "grad_norm": 8.87192440032959, "learning_rate": 0.00018746072451065438, "loss": 6.5723, "step": 652 }, { "epoch": 9.324229074889868, "grad_norm": 9.323250770568848, "learning_rate": 0.00018741582271281332, "loss": 6.6591, "step": 653 }, { "epoch": 9.338325991189427, "grad_norm": 9.699490547180176, "learning_rate": 0.00018737084606180063, "loss": 6.5957, "step": 654 }, { "epoch": 9.352422907488986, "grad_norm": 9.499716758728027, "learning_rate": 0.00018732579459612936, "loss": 6.6375, "step": 655 }, { "epoch": 9.366519823788547, "grad_norm": 8.673836708068848, "learning_rate": 0.0001872806683543766, "loss": 6.4891, "step": 656 }, { "epoch": 9.380616740088106, "grad_norm": 8.535860061645508, "learning_rate": 0.00018723546737518332, "loss": 6.6232, "step": 657 }, { "epoch": 9.394713656387665, "grad_norm": 8.973447799682617, "learning_rate": 0.00018719019169725472, "loss": 6.6999, "step": 658 }, { "epoch": 9.408810572687225, "grad_norm": 7.773440837860107, "learning_rate": 0.00018714484135935977, "loss": 6.6246, "step": 659 }, { "epoch": 9.422907488986784, "grad_norm": 8.653839111328125, "learning_rate": 0.0001870994164003315, "loss": 6.8538, "step": 660 }, { "epoch": 9.437004405286343, "grad_norm": 7.681654930114746, "learning_rate": 0.00018705391685906676, "loss": 6.7284, "step": 661 }, { "epoch": 9.451101321585902, "grad_norm": 8.993474960327148, "learning_rate": 0.00018700834277452627, "loss": 6.7348, "step": 662 }, { "epoch": 9.465198237885463, "grad_norm": 8.730484008789062, "learning_rate": 0.00018696269418573466, "loss": 6.7023, "step": 663 }, { "epoch": 9.479295154185023, "grad_norm": 8.807551383972168, "learning_rate": 0.00018691697113178021, "loss": 6.6025, "step": 664 }, { "epoch": 9.493392070484582, "grad_norm": 11.31397533416748, "learning_rate": 0.00018687117365181512, "loss": 6.6738, "step": 665 }, { "epoch": 9.507488986784141, "grad_norm": 8.353428840637207, "learning_rate": 0.0001868253017850552, "loss": 6.6696, "step": 666 }, { "epoch": 9.5215859030837, "grad_norm": 8.344447135925293, "learning_rate": 0.00018677935557078005, "loss": 6.547, "step": 667 }, { "epoch": 9.53568281938326, "grad_norm": 8.368324279785156, "learning_rate": 0.00018673333504833287, "loss": 6.6792, "step": 668 }, { "epoch": 9.549779735682819, "grad_norm": 8.927423477172852, "learning_rate": 0.00018668724025712048, "loss": 6.9085, "step": 669 }, { "epoch": 9.56387665198238, "grad_norm": 8.537492752075195, "learning_rate": 0.00018664107123661337, "loss": 6.8251, "step": 670 }, { "epoch": 9.577973568281939, "grad_norm": 8.56866455078125, "learning_rate": 0.0001865948280263455, "loss": 6.6761, "step": 671 }, { "epoch": 9.592070484581498, "grad_norm": 9.264017105102539, "learning_rate": 0.00018654851066591448, "loss": 6.7442, "step": 672 }, { "epoch": 9.606167400881057, "grad_norm": 8.229893684387207, "learning_rate": 0.00018650211919498123, "loss": 6.9496, "step": 673 }, { "epoch": 9.620264317180617, "grad_norm": 8.891019821166992, "learning_rate": 0.00018645565365327032, "loss": 6.6148, "step": 674 }, { "epoch": 9.634361233480176, "grad_norm": 8.519316673278809, "learning_rate": 0.00018640911408056963, "loss": 6.7075, "step": 675 }, { "epoch": 9.648458149779735, "grad_norm": 7.906443119049072, "learning_rate": 0.00018636250051673047, "loss": 6.8622, "step": 676 }, { "epoch": 9.662555066079296, "grad_norm": 9.559021949768066, "learning_rate": 0.00018631581300166753, "loss": 6.8145, "step": 677 }, { "epoch": 9.676651982378855, "grad_norm": 8.695327758789062, "learning_rate": 0.00018626905157535873, "loss": 6.9044, "step": 678 }, { "epoch": 9.690748898678414, "grad_norm": 9.234659194946289, "learning_rate": 0.0001862222162778454, "loss": 6.8296, "step": 679 }, { "epoch": 9.704845814977974, "grad_norm": 10.944517135620117, "learning_rate": 0.00018617530714923209, "loss": 6.9705, "step": 680 }, { "epoch": 9.718942731277533, "grad_norm": 7.4156928062438965, "learning_rate": 0.00018612832422968648, "loss": 6.8498, "step": 681 }, { "epoch": 9.733039647577092, "grad_norm": 7.775762557983398, "learning_rate": 0.00018608126755943955, "loss": 6.8248, "step": 682 }, { "epoch": 9.747136563876651, "grad_norm": 8.92798137664795, "learning_rate": 0.00018603413717878538, "loss": 6.833, "step": 683 }, { "epoch": 9.761233480176212, "grad_norm": 7.774735450744629, "learning_rate": 0.00018598693312808122, "loss": 6.9177, "step": 684 }, { "epoch": 9.775330396475772, "grad_norm": 8.935710906982422, "learning_rate": 0.0001859396554477473, "loss": 6.8834, "step": 685 }, { "epoch": 9.78942731277533, "grad_norm": 7.868921279907227, "learning_rate": 0.00018589230417826697, "loss": 6.8683, "step": 686 }, { "epoch": 9.80352422907489, "grad_norm": 7.218008041381836, "learning_rate": 0.00018584487936018661, "loss": 6.688, "step": 687 }, { "epoch": 9.81762114537445, "grad_norm": 7.584420680999756, "learning_rate": 0.00018579738103411555, "loss": 6.9735, "step": 688 }, { "epoch": 9.831718061674009, "grad_norm": 8.057254791259766, "learning_rate": 0.00018574980924072602, "loss": 6.9532, "step": 689 }, { "epoch": 9.845814977973568, "grad_norm": 7.268715858459473, "learning_rate": 0.00018570216402075324, "loss": 6.9811, "step": 690 }, { "epoch": 9.859911894273127, "grad_norm": 10.273457527160645, "learning_rate": 0.00018565444541499527, "loss": 6.8798, "step": 691 }, { "epoch": 9.874008810572688, "grad_norm": 7.8910231590271, "learning_rate": 0.000185606653464313, "loss": 6.9929, "step": 692 }, { "epoch": 9.888105726872247, "grad_norm": 9.55091381072998, "learning_rate": 0.00018555878820963013, "loss": 6.861, "step": 693 }, { "epoch": 9.902202643171806, "grad_norm": 8.149909019470215, "learning_rate": 0.00018551084969193314, "loss": 6.932, "step": 694 }, { "epoch": 9.916299559471366, "grad_norm": 7.278769493103027, "learning_rate": 0.0001854628379522712, "loss": 6.9344, "step": 695 }, { "epoch": 9.930396475770925, "grad_norm": 7.51647424697876, "learning_rate": 0.00018541475303175627, "loss": 6.9045, "step": 696 }, { "epoch": 9.944493392070484, "grad_norm": 8.219710350036621, "learning_rate": 0.0001853665949715629, "loss": 6.8568, "step": 697 }, { "epoch": 9.958590308370043, "grad_norm": 8.047307014465332, "learning_rate": 0.00018531836381292823, "loss": 6.9502, "step": 698 }, { "epoch": 9.972687224669604, "grad_norm": 8.35663890838623, "learning_rate": 0.00018527005959715216, "loss": 6.9515, "step": 699 }, { "epoch": 9.986784140969164, "grad_norm": 7.606381893157959, "learning_rate": 0.00018522168236559695, "loss": 6.8012, "step": 700 }, { "epoch": 10.01409691629956, "grad_norm": 11.636252403259277, "learning_rate": 0.00018517323215968748, "loss": 11.0579, "step": 701 }, { "epoch": 10.028193832599118, "grad_norm": 12.995759010314941, "learning_rate": 0.00018512470902091118, "loss": 5.2763, "step": 702 }, { "epoch": 10.042290748898678, "grad_norm": 14.07719612121582, "learning_rate": 0.00018507611299081774, "loss": 5.1206, "step": 703 }, { "epoch": 10.056387665198239, "grad_norm": 13.267398834228516, "learning_rate": 0.00018502744411101951, "loss": 5.1737, "step": 704 }, { "epoch": 10.070484581497798, "grad_norm": 11.196743965148926, "learning_rate": 0.000184978702423191, "loss": 5.0534, "step": 705 }, { "epoch": 10.084581497797357, "grad_norm": 9.91689395904541, "learning_rate": 0.00018492988796906925, "loss": 5.0594, "step": 706 }, { "epoch": 10.098678414096916, "grad_norm": 11.554791450500488, "learning_rate": 0.00018488100079045344, "loss": 5.1463, "step": 707 }, { "epoch": 10.112775330396476, "grad_norm": 12.985812187194824, "learning_rate": 0.00018483204092920516, "loss": 5.2185, "step": 708 }, { "epoch": 10.126872246696035, "grad_norm": 12.008038520812988, "learning_rate": 0.00018478300842724815, "loss": 5.2886, "step": 709 }, { "epoch": 10.140969162995594, "grad_norm": 13.068727493286133, "learning_rate": 0.00018473390332656838, "loss": 5.1794, "step": 710 }, { "epoch": 10.155066079295155, "grad_norm": 11.546491622924805, "learning_rate": 0.000184684725669214, "loss": 5.041, "step": 711 }, { "epoch": 10.169162995594714, "grad_norm": 11.315566062927246, "learning_rate": 0.00018463547549729526, "loss": 5.232, "step": 712 }, { "epoch": 10.183259911894273, "grad_norm": 13.46137523651123, "learning_rate": 0.00018458615285298454, "loss": 5.1746, "step": 713 }, { "epoch": 10.197356828193833, "grad_norm": 11.167923927307129, "learning_rate": 0.00018453675777851627, "loss": 5.3127, "step": 714 }, { "epoch": 10.211453744493392, "grad_norm": 12.877705574035645, "learning_rate": 0.00018448729031618686, "loss": 5.2685, "step": 715 }, { "epoch": 10.225550660792951, "grad_norm": 12.246356964111328, "learning_rate": 0.00018443775050835473, "loss": 5.2724, "step": 716 }, { "epoch": 10.23964757709251, "grad_norm": 13.94930648803711, "learning_rate": 0.00018438813839744027, "loss": 5.1963, "step": 717 }, { "epoch": 10.25374449339207, "grad_norm": 11.567874908447266, "learning_rate": 0.00018433845402592578, "loss": 5.2272, "step": 718 }, { "epoch": 10.26784140969163, "grad_norm": 12.254955291748047, "learning_rate": 0.00018428869743635537, "loss": 5.4054, "step": 719 }, { "epoch": 10.28193832599119, "grad_norm": 10.23955249786377, "learning_rate": 0.0001842388686713351, "loss": 5.149, "step": 720 }, { "epoch": 10.296035242290749, "grad_norm": 11.207096099853516, "learning_rate": 0.0001841889677735327, "loss": 5.2722, "step": 721 }, { "epoch": 10.310132158590308, "grad_norm": 10.385015487670898, "learning_rate": 0.0001841389947856778, "loss": 5.2371, "step": 722 }, { "epoch": 10.324229074889868, "grad_norm": 9.437115669250488, "learning_rate": 0.0001840889497505617, "loss": 5.3353, "step": 723 }, { "epoch": 10.338325991189427, "grad_norm": 10.258122444152832, "learning_rate": 0.00018403883271103736, "loss": 5.3817, "step": 724 }, { "epoch": 10.352422907488986, "grad_norm": 10.67017936706543, "learning_rate": 0.00018398864371001945, "loss": 5.364, "step": 725 }, { "epoch": 10.366519823788547, "grad_norm": 11.283514976501465, "learning_rate": 0.00018393838279048423, "loss": 5.4898, "step": 726 }, { "epoch": 10.380616740088106, "grad_norm": 10.786419868469238, "learning_rate": 0.00018388804999546957, "loss": 5.5199, "step": 727 }, { "epoch": 10.394713656387665, "grad_norm": 10.378274917602539, "learning_rate": 0.00018383764536807485, "loss": 5.5138, "step": 728 }, { "epoch": 10.408810572687225, "grad_norm": 11.432807922363281, "learning_rate": 0.00018378716895146093, "loss": 5.3531, "step": 729 }, { "epoch": 10.422907488986784, "grad_norm": 10.095375061035156, "learning_rate": 0.00018373662078885027, "loss": 5.3092, "step": 730 }, { "epoch": 10.437004405286343, "grad_norm": 11.438772201538086, "learning_rate": 0.00018368600092352665, "loss": 5.3462, "step": 731 }, { "epoch": 10.451101321585902, "grad_norm": 10.749035835266113, "learning_rate": 0.0001836353093988353, "loss": 5.4529, "step": 732 }, { "epoch": 10.465198237885463, "grad_norm": 10.72800350189209, "learning_rate": 0.0001835845462581828, "loss": 5.6419, "step": 733 }, { "epoch": 10.479295154185023, "grad_norm": 8.919745445251465, "learning_rate": 0.000183533711545037, "loss": 5.3953, "step": 734 }, { "epoch": 10.493392070484582, "grad_norm": 11.762939453125, "learning_rate": 0.00018348280530292713, "loss": 5.4906, "step": 735 }, { "epoch": 10.507488986784141, "grad_norm": 13.187941551208496, "learning_rate": 0.0001834318275754436, "loss": 5.3617, "step": 736 }, { "epoch": 10.5215859030837, "grad_norm": 9.713920593261719, "learning_rate": 0.00018338077840623806, "loss": 5.3088, "step": 737 }, { "epoch": 10.53568281938326, "grad_norm": 13.538158416748047, "learning_rate": 0.00018332965783902333, "loss": 5.4684, "step": 738 }, { "epoch": 10.549779735682819, "grad_norm": 9.097997665405273, "learning_rate": 0.00018327846591757342, "loss": 5.4784, "step": 739 }, { "epoch": 10.56387665198238, "grad_norm": 10.790201187133789, "learning_rate": 0.00018322720268572332, "loss": 5.651, "step": 740 }, { "epoch": 10.577973568281939, "grad_norm": 9.631773948669434, "learning_rate": 0.00018317586818736913, "loss": 5.4481, "step": 741 }, { "epoch": 10.592070484581498, "grad_norm": 9.400154113769531, "learning_rate": 0.0001831244624664681, "loss": 5.5112, "step": 742 }, { "epoch": 10.606167400881057, "grad_norm": 10.330177307128906, "learning_rate": 0.00018307298556703824, "loss": 5.4811, "step": 743 }, { "epoch": 10.620264317180617, "grad_norm": 10.055923461914062, "learning_rate": 0.00018302143753315875, "loss": 5.7626, "step": 744 }, { "epoch": 10.634361233480176, "grad_norm": 8.91289234161377, "learning_rate": 0.00018296981840896953, "loss": 5.4967, "step": 745 }, { "epoch": 10.648458149779735, "grad_norm": 9.808186531066895, "learning_rate": 0.00018291812823867149, "loss": 5.5469, "step": 746 }, { "epoch": 10.662555066079296, "grad_norm": 9.234919548034668, "learning_rate": 0.00018286636706652632, "loss": 5.6144, "step": 747 }, { "epoch": 10.676651982378855, "grad_norm": 10.226089477539062, "learning_rate": 0.00018281453493685653, "loss": 5.4882, "step": 748 }, { "epoch": 10.690748898678414, "grad_norm": 10.571527481079102, "learning_rate": 0.0001827626318940454, "loss": 5.6819, "step": 749 }, { "epoch": 10.704845814977974, "grad_norm": 10.174403190612793, "learning_rate": 0.00018271065798253688, "loss": 5.6614, "step": 750 }, { "epoch": 10.704845814977974, "eval_loss": 0.8026774525642395, "eval_runtime": 132.5485, "eval_samples_per_second": 13.64, "eval_steps_per_second": 0.43, "step": 750 }, { "epoch": 10.718942731277533, "grad_norm": 11.086686134338379, "learning_rate": 0.00018265861324683568, "loss": 5.5722, "step": 751 }, { "epoch": 10.733039647577092, "grad_norm": 10.220298767089844, "learning_rate": 0.00018260649773150703, "loss": 5.7278, "step": 752 }, { "epoch": 10.747136563876651, "grad_norm": 8.675567626953125, "learning_rate": 0.00018255431148117693, "loss": 5.6682, "step": 753 }, { "epoch": 10.761233480176212, "grad_norm": 10.735408782958984, "learning_rate": 0.0001825020545405318, "loss": 5.6196, "step": 754 }, { "epoch": 10.775330396475772, "grad_norm": 8.775152206420898, "learning_rate": 0.00018244972695431874, "loss": 5.6845, "step": 755 }, { "epoch": 10.78942731277533, "grad_norm": 13.366692543029785, "learning_rate": 0.00018239732876734527, "loss": 5.8475, "step": 756 }, { "epoch": 10.80352422907489, "grad_norm": 9.162715911865234, "learning_rate": 0.00018234486002447926, "loss": 5.7817, "step": 757 }, { "epoch": 10.81762114537445, "grad_norm": 12.618876457214355, "learning_rate": 0.0001822923207706492, "loss": 5.7005, "step": 758 }, { "epoch": 10.831718061674009, "grad_norm": 9.270393371582031, "learning_rate": 0.00018223971105084378, "loss": 5.8126, "step": 759 }, { "epoch": 10.845814977973568, "grad_norm": 12.381141662597656, "learning_rate": 0.00018218703091011214, "loss": 5.7977, "step": 760 }, { "epoch": 10.859911894273127, "grad_norm": 12.110838890075684, "learning_rate": 0.00018213428039356367, "loss": 5.7361, "step": 761 }, { "epoch": 10.874008810572688, "grad_norm": 9.083556175231934, "learning_rate": 0.00018208145954636806, "loss": 5.7425, "step": 762 }, { "epoch": 10.888105726872247, "grad_norm": 9.857342720031738, "learning_rate": 0.00018202856841375518, "loss": 5.7752, "step": 763 }, { "epoch": 10.902202643171806, "grad_norm": 8.534427642822266, "learning_rate": 0.0001819756070410151, "loss": 5.6942, "step": 764 }, { "epoch": 10.916299559471366, "grad_norm": 10.399006843566895, "learning_rate": 0.00018192257547349803, "loss": 5.7795, "step": 765 }, { "epoch": 10.930396475770925, "grad_norm": 9.068225860595703, "learning_rate": 0.00018186947375661435, "loss": 5.8657, "step": 766 }, { "epoch": 10.944493392070484, "grad_norm": 8.998698234558105, "learning_rate": 0.00018181630193583436, "loss": 5.7921, "step": 767 }, { "epoch": 10.958590308370043, "grad_norm": 9.429346084594727, "learning_rate": 0.0001817630600566886, "loss": 5.8171, "step": 768 }, { "epoch": 10.972687224669604, "grad_norm": 9.46871566772461, "learning_rate": 0.00018170974816476737, "loss": 5.7138, "step": 769 }, { "epoch": 10.986784140969164, "grad_norm": 11.119671821594238, "learning_rate": 0.0001816563663057211, "loss": 5.7725, "step": 770 }, { "epoch": 11.01409691629956, "grad_norm": 9.374451637268066, "learning_rate": 0.00018160291452526007, "loss": 8.98, "step": 771 }, { "epoch": 11.028193832599118, "grad_norm": 17.0820369720459, "learning_rate": 0.00018154939286915437, "loss": 3.9763, "step": 772 }, { "epoch": 11.042290748898678, "grad_norm": 18.95514488220215, "learning_rate": 0.00018149580138323405, "loss": 3.7597, "step": 773 }, { "epoch": 11.056387665198239, "grad_norm": 13.926958084106445, "learning_rate": 0.00018144214011338886, "loss": 3.8597, "step": 774 }, { "epoch": 11.070484581497798, "grad_norm": 21.842239379882812, "learning_rate": 0.00018138840910556837, "loss": 3.8147, "step": 775 }, { "epoch": 11.084581497797357, "grad_norm": 13.904328346252441, "learning_rate": 0.00018133460840578176, "loss": 3.9071, "step": 776 }, { "epoch": 11.098678414096916, "grad_norm": 15.541295051574707, "learning_rate": 0.000181280738060098, "loss": 3.8043, "step": 777 }, { "epoch": 11.112775330396476, "grad_norm": 12.68041706085205, "learning_rate": 0.00018122679811464565, "loss": 3.7763, "step": 778 }, { "epoch": 11.126872246696035, "grad_norm": 12.804862976074219, "learning_rate": 0.00018117278861561284, "loss": 3.7786, "step": 779 }, { "epoch": 11.140969162995594, "grad_norm": 13.689950942993164, "learning_rate": 0.00018111870960924734, "loss": 3.7247, "step": 780 }, { "epoch": 11.155066079295155, "grad_norm": 12.94391918182373, "learning_rate": 0.00018106456114185636, "loss": 3.62, "step": 781 }, { "epoch": 11.169162995594714, "grad_norm": 12.375044822692871, "learning_rate": 0.00018101034325980657, "loss": 4.0365, "step": 782 }, { "epoch": 11.183259911894273, "grad_norm": 16.17354965209961, "learning_rate": 0.0001809560560095242, "loss": 3.9452, "step": 783 }, { "epoch": 11.197356828193833, "grad_norm": 12.736469268798828, "learning_rate": 0.00018090169943749476, "loss": 3.7761, "step": 784 }, { "epoch": 11.211453744493392, "grad_norm": 12.152167320251465, "learning_rate": 0.00018084727359026317, "loss": 3.8387, "step": 785 }, { "epoch": 11.225550660792951, "grad_norm": 10.252900123596191, "learning_rate": 0.00018079277851443365, "loss": 3.8257, "step": 786 }, { "epoch": 11.23964757709251, "grad_norm": 10.996660232543945, "learning_rate": 0.00018073821425666976, "loss": 4.0892, "step": 787 }, { "epoch": 11.25374449339207, "grad_norm": 11.822881698608398, "learning_rate": 0.0001806835808636942, "loss": 3.8811, "step": 788 }, { "epoch": 11.26784140969163, "grad_norm": 10.44597053527832, "learning_rate": 0.00018062887838228892, "loss": 3.9496, "step": 789 }, { "epoch": 11.28193832599119, "grad_norm": 10.75739574432373, "learning_rate": 0.00018057410685929506, "loss": 3.9241, "step": 790 }, { "epoch": 11.296035242290749, "grad_norm": 11.990365982055664, "learning_rate": 0.00018051926634161282, "loss": 4.0601, "step": 791 }, { "epoch": 11.310132158590308, "grad_norm": 13.239373207092285, "learning_rate": 0.00018046435687620155, "loss": 4.0108, "step": 792 }, { "epoch": 11.324229074889868, "grad_norm": 10.038420677185059, "learning_rate": 0.0001804093785100796, "loss": 4.0867, "step": 793 }, { "epoch": 11.338325991189427, "grad_norm": 13.48196792602539, "learning_rate": 0.00018035433129032425, "loss": 4.0324, "step": 794 }, { "epoch": 11.352422907488986, "grad_norm": 10.531044006347656, "learning_rate": 0.00018029921526407184, "loss": 4.0146, "step": 795 }, { "epoch": 11.366519823788547, "grad_norm": 9.566649436950684, "learning_rate": 0.00018024403047851766, "loss": 4.1372, "step": 796 }, { "epoch": 11.380616740088106, "grad_norm": 11.025485038757324, "learning_rate": 0.0001801887769809157, "loss": 4.1774, "step": 797 }, { "epoch": 11.394713656387665, "grad_norm": 15.027955055236816, "learning_rate": 0.00018013345481857903, "loss": 3.8722, "step": 798 }, { "epoch": 11.408810572687225, "grad_norm": 11.156898498535156, "learning_rate": 0.00018007806403887925, "loss": 4.2031, "step": 799 }, { "epoch": 11.422907488986784, "grad_norm": 17.00517463684082, "learning_rate": 0.00018002260468924695, "loss": 3.9817, "step": 800 }, { "epoch": 11.437004405286343, "grad_norm": 10.233874320983887, "learning_rate": 0.00017996707681717128, "loss": 4.2145, "step": 801 }, { "epoch": 11.451101321585902, "grad_norm": 13.95673942565918, "learning_rate": 0.00017991148047020013, "loss": 3.9962, "step": 802 }, { "epoch": 11.465198237885463, "grad_norm": 9.811971664428711, "learning_rate": 0.00017985581569594008, "loss": 4.2276, "step": 803 }, { "epoch": 11.479295154185023, "grad_norm": 10.874019622802734, "learning_rate": 0.00017980008254205618, "loss": 4.1385, "step": 804 }, { "epoch": 11.493392070484582, "grad_norm": 11.191951751708984, "learning_rate": 0.00017974428105627208, "loss": 4.0461, "step": 805 }, { "epoch": 11.507488986784141, "grad_norm": 9.515632629394531, "learning_rate": 0.00017968841128637, "loss": 4.2187, "step": 806 }, { "epoch": 11.5215859030837, "grad_norm": 11.006308555603027, "learning_rate": 0.0001796324732801906, "loss": 3.9904, "step": 807 }, { "epoch": 11.53568281938326, "grad_norm": 9.76673698425293, "learning_rate": 0.0001795764670856329, "loss": 4.2502, "step": 808 }, { "epoch": 11.549779735682819, "grad_norm": 12.065739631652832, "learning_rate": 0.0001795203927506544, "loss": 4.2288, "step": 809 }, { "epoch": 11.56387665198238, "grad_norm": 10.794450759887695, "learning_rate": 0.0001794642503232709, "loss": 4.1335, "step": 810 }, { "epoch": 11.577973568281939, "grad_norm": 9.835041046142578, "learning_rate": 0.00017940803985155656, "loss": 4.2096, "step": 811 }, { "epoch": 11.592070484581498, "grad_norm": 10.451004981994629, "learning_rate": 0.0001793517613836437, "loss": 4.1884, "step": 812 }, { "epoch": 11.606167400881057, "grad_norm": 12.816031455993652, "learning_rate": 0.00017929541496772297, "loss": 4.2662, "step": 813 }, { "epoch": 11.620264317180617, "grad_norm": 11.205449104309082, "learning_rate": 0.00017923900065204315, "loss": 4.4736, "step": 814 }, { "epoch": 11.634361233480176, "grad_norm": 10.91614055633545, "learning_rate": 0.00017918251848491118, "loss": 4.3472, "step": 815 }, { "epoch": 11.648458149779735, "grad_norm": 10.425148010253906, "learning_rate": 0.0001791259685146921, "loss": 4.5028, "step": 816 }, { "epoch": 11.662555066079296, "grad_norm": 11.742585182189941, "learning_rate": 0.000179069350789809, "loss": 4.2725, "step": 817 }, { "epoch": 11.676651982378855, "grad_norm": 9.74598217010498, "learning_rate": 0.000179012665358743, "loss": 4.3908, "step": 818 }, { "epoch": 11.690748898678414, "grad_norm": 10.737768173217773, "learning_rate": 0.00017895591227003315, "loss": 4.293, "step": 819 }, { "epoch": 11.704845814977974, "grad_norm": 11.701116561889648, "learning_rate": 0.00017889909157227653, "loss": 4.3973, "step": 820 }, { "epoch": 11.718942731277533, "grad_norm": 9.319437980651855, "learning_rate": 0.00017884220331412803, "loss": 4.4642, "step": 821 }, { "epoch": 11.733039647577092, "grad_norm": 10.521557807922363, "learning_rate": 0.00017878524754430037, "loss": 4.4798, "step": 822 }, { "epoch": 11.747136563876651, "grad_norm": 10.320682525634766, "learning_rate": 0.0001787282243115642, "loss": 4.3847, "step": 823 }, { "epoch": 11.761233480176212, "grad_norm": 11.241861343383789, "learning_rate": 0.00017867113366474785, "loss": 4.2454, "step": 824 }, { "epoch": 11.775330396475772, "grad_norm": 12.095932006835938, "learning_rate": 0.00017861397565273733, "loss": 4.2926, "step": 825 }, { "epoch": 11.78942731277533, "grad_norm": 9.937694549560547, "learning_rate": 0.00017855675032447648, "loss": 4.4459, "step": 826 }, { "epoch": 11.80352422907489, "grad_norm": 13.1368408203125, "learning_rate": 0.00017849945772896664, "loss": 4.5773, "step": 827 }, { "epoch": 11.81762114537445, "grad_norm": 9.696205139160156, "learning_rate": 0.00017844209791526678, "loss": 4.3975, "step": 828 }, { "epoch": 11.831718061674009, "grad_norm": 11.21163272857666, "learning_rate": 0.00017838467093249353, "loss": 4.4402, "step": 829 }, { "epoch": 11.845814977973568, "grad_norm": 11.215200424194336, "learning_rate": 0.00017832717682982093, "loss": 4.3281, "step": 830 }, { "epoch": 11.859911894273127, "grad_norm": 9.86231517791748, "learning_rate": 0.0001782696156564805, "loss": 4.4493, "step": 831 }, { "epoch": 11.874008810572688, "grad_norm": 13.157506942749023, "learning_rate": 0.0001782119874617612, "loss": 4.4811, "step": 832 }, { "epoch": 11.888105726872247, "grad_norm": 11.400369644165039, "learning_rate": 0.00017815429229500946, "loss": 4.5496, "step": 833 }, { "epoch": 11.902202643171806, "grad_norm": 10.786680221557617, "learning_rate": 0.00017809653020562895, "loss": 4.3484, "step": 834 }, { "epoch": 11.916299559471366, "grad_norm": 9.901410102844238, "learning_rate": 0.00017803870124308068, "loss": 4.3103, "step": 835 }, { "epoch": 11.930396475770925, "grad_norm": 10.447687149047852, "learning_rate": 0.00017798080545688297, "loss": 4.4517, "step": 836 }, { "epoch": 11.944493392070484, "grad_norm": 11.485855102539062, "learning_rate": 0.0001779228428966113, "loss": 4.3722, "step": 837 }, { "epoch": 11.958590308370043, "grad_norm": 11.360231399536133, "learning_rate": 0.00017786481361189833, "loss": 4.6119, "step": 838 }, { "epoch": 11.972687224669604, "grad_norm": 10.929000854492188, "learning_rate": 0.00017780671765243392, "loss": 4.5387, "step": 839 }, { "epoch": 11.986784140969164, "grad_norm": 11.069295883178711, "learning_rate": 0.00017774855506796496, "loss": 4.2835, "step": 840 }, { "epoch": 12.01409691629956, "grad_norm": 9.764337539672852, "learning_rate": 0.00017769032590829537, "loss": 6.5648, "step": 841 }, { "epoch": 12.028193832599118, "grad_norm": 11.357198715209961, "learning_rate": 0.00017763203022328614, "loss": 2.6885, "step": 842 }, { "epoch": 12.042290748898678, "grad_norm": 13.090156555175781, "learning_rate": 0.00017757366806285521, "loss": 2.56, "step": 843 }, { "epoch": 12.056387665198239, "grad_norm": 12.069717407226562, "learning_rate": 0.00017751523947697742, "loss": 2.6416, "step": 844 }, { "epoch": 12.070484581497798, "grad_norm": 10.410787582397461, "learning_rate": 0.00017745674451568453, "loss": 2.694, "step": 845 }, { "epoch": 12.084581497797357, "grad_norm": 11.40599536895752, "learning_rate": 0.00017739818322906508, "loss": 2.6254, "step": 846 }, { "epoch": 12.098678414096916, "grad_norm": 10.755266189575195, "learning_rate": 0.0001773395556672644, "loss": 2.5836, "step": 847 }, { "epoch": 12.112775330396476, "grad_norm": 11.161855697631836, "learning_rate": 0.00017728086188048466, "loss": 2.5769, "step": 848 }, { "epoch": 12.126872246696035, "grad_norm": 9.945941925048828, "learning_rate": 0.00017722210191898466, "loss": 2.7951, "step": 849 }, { "epoch": 12.140969162995594, "grad_norm": 10.561120986938477, "learning_rate": 0.00017716327583307989, "loss": 2.5078, "step": 850 }, { "epoch": 12.155066079295155, "grad_norm": 13.099159240722656, "learning_rate": 0.00017710438367314244, "loss": 2.7622, "step": 851 }, { "epoch": 12.169162995594714, "grad_norm": 11.372950553894043, "learning_rate": 0.000177045425489601, "loss": 2.6325, "step": 852 }, { "epoch": 12.183259911894273, "grad_norm": 9.826996803283691, "learning_rate": 0.0001769864013329408, "loss": 2.7513, "step": 853 }, { "epoch": 12.197356828193833, "grad_norm": 12.40358829498291, "learning_rate": 0.00017692731125370354, "loss": 2.5696, "step": 854 }, { "epoch": 12.211453744493392, "grad_norm": 13.00760269165039, "learning_rate": 0.0001768681553024874, "loss": 2.6739, "step": 855 }, { "epoch": 12.225550660792951, "grad_norm": 9.857587814331055, "learning_rate": 0.0001768089335299469, "loss": 2.6648, "step": 856 }, { "epoch": 12.23964757709251, "grad_norm": 11.487194061279297, "learning_rate": 0.00017674964598679303, "loss": 2.7409, "step": 857 }, { "epoch": 12.25374449339207, "grad_norm": 10.142484664916992, "learning_rate": 0.000176690292723793, "loss": 2.698, "step": 858 }, { "epoch": 12.26784140969163, "grad_norm": 10.905588150024414, "learning_rate": 0.00017663087379177032, "loss": 2.5782, "step": 859 }, { "epoch": 12.28193832599119, "grad_norm": 10.300945281982422, "learning_rate": 0.00017657138924160476, "loss": 2.6647, "step": 860 }, { "epoch": 12.296035242290749, "grad_norm": 11.517670631408691, "learning_rate": 0.00017651183912423228, "loss": 2.5534, "step": 861 }, { "epoch": 12.310132158590308, "grad_norm": 9.71640396118164, "learning_rate": 0.00017645222349064495, "loss": 2.7742, "step": 862 }, { "epoch": 12.324229074889868, "grad_norm": 11.237198829650879, "learning_rate": 0.00017639254239189096, "loss": 2.7907, "step": 863 }, { "epoch": 12.338325991189427, "grad_norm": 10.487508773803711, "learning_rate": 0.00017633279587907455, "loss": 2.7791, "step": 864 }, { "epoch": 12.352422907488986, "grad_norm": 11.083252906799316, "learning_rate": 0.00017627298400335598, "loss": 2.6738, "step": 865 }, { "epoch": 12.366519823788547, "grad_norm": 11.600593566894531, "learning_rate": 0.00017621310681595147, "loss": 2.7613, "step": 866 }, { "epoch": 12.380616740088106, "grad_norm": 10.35483455657959, "learning_rate": 0.00017615316436813316, "loss": 2.7244, "step": 867 }, { "epoch": 12.394713656387665, "grad_norm": 12.630843162536621, "learning_rate": 0.0001760931567112291, "loss": 2.7782, "step": 868 }, { "epoch": 12.408810572687225, "grad_norm": 10.300714492797852, "learning_rate": 0.00017603308389662318, "loss": 2.8412, "step": 869 }, { "epoch": 12.422907488986784, "grad_norm": 14.418756484985352, "learning_rate": 0.00017597294597575507, "loss": 2.7401, "step": 870 }, { "epoch": 12.437004405286343, "grad_norm": 11.772397994995117, "learning_rate": 0.00017591274300012013, "loss": 2.9312, "step": 871 }, { "epoch": 12.451101321585902, "grad_norm": 12.26634407043457, "learning_rate": 0.00017585247502126955, "loss": 2.738, "step": 872 }, { "epoch": 12.465198237885463, "grad_norm": 11.834174156188965, "learning_rate": 0.00017579214209081004, "loss": 2.6609, "step": 873 }, { "epoch": 12.479295154185023, "grad_norm": 10.61044979095459, "learning_rate": 0.00017573174426040405, "loss": 2.9515, "step": 874 }, { "epoch": 12.493392070484582, "grad_norm": 11.539770126342773, "learning_rate": 0.00017567128158176953, "loss": 2.8626, "step": 875 }, { "epoch": 12.507488986784141, "grad_norm": 11.261491775512695, "learning_rate": 0.00017561075410668, "loss": 2.898, "step": 876 }, { "epoch": 12.5215859030837, "grad_norm": 13.079240798950195, "learning_rate": 0.00017555016188696446, "loss": 2.9892, "step": 877 }, { "epoch": 12.53568281938326, "grad_norm": 10.053821563720703, "learning_rate": 0.00017548950497450728, "loss": 2.8486, "step": 878 }, { "epoch": 12.549779735682819, "grad_norm": 12.232101440429688, "learning_rate": 0.00017542878342124833, "loss": 3.0397, "step": 879 }, { "epoch": 12.56387665198238, "grad_norm": 10.556097984313965, "learning_rate": 0.0001753679972791828, "loss": 2.9426, "step": 880 }, { "epoch": 12.577973568281939, "grad_norm": 13.17647933959961, "learning_rate": 0.00017530714660036112, "loss": 3.0736, "step": 881 }, { "epoch": 12.592070484581498, "grad_norm": 12.86817455291748, "learning_rate": 0.00017524623143688902, "loss": 2.8816, "step": 882 }, { "epoch": 12.606167400881057, "grad_norm": 11.691473007202148, "learning_rate": 0.00017518525184092754, "loss": 2.9101, "step": 883 }, { "epoch": 12.620264317180617, "grad_norm": 13.372227668762207, "learning_rate": 0.00017512420786469272, "loss": 3.0887, "step": 884 }, { "epoch": 12.634361233480176, "grad_norm": 10.978402137756348, "learning_rate": 0.0001750630995604559, "loss": 3.0369, "step": 885 }, { "epoch": 12.648458149779735, "grad_norm": 13.259976387023926, "learning_rate": 0.0001750019269805434, "loss": 2.9651, "step": 886 }, { "epoch": 12.662555066079296, "grad_norm": 10.5857515335083, "learning_rate": 0.00017494069017733654, "loss": 3.1343, "step": 887 }, { "epoch": 12.676651982378855, "grad_norm": 10.873541831970215, "learning_rate": 0.0001748793892032718, "loss": 3.0337, "step": 888 }, { "epoch": 12.690748898678414, "grad_norm": 10.409069061279297, "learning_rate": 0.00017481802411084042, "loss": 2.9551, "step": 889 }, { "epoch": 12.704845814977974, "grad_norm": 9.589970588684082, "learning_rate": 0.00017475659495258864, "loss": 2.9223, "step": 890 }, { "epoch": 12.718942731277533, "grad_norm": 10.27968978881836, "learning_rate": 0.0001746951017811176, "loss": 3.0316, "step": 891 }, { "epoch": 12.733039647577092, "grad_norm": 11.248296737670898, "learning_rate": 0.00017463354464908318, "loss": 3.1261, "step": 892 }, { "epoch": 12.747136563876651, "grad_norm": 11.821159362792969, "learning_rate": 0.00017457192360919598, "loss": 3.1636, "step": 893 }, { "epoch": 12.761233480176212, "grad_norm": 11.698129653930664, "learning_rate": 0.0001745102387142215, "loss": 3.2204, "step": 894 }, { "epoch": 12.775330396475772, "grad_norm": 10.93070125579834, "learning_rate": 0.00017444849001697975, "loss": 3.0855, "step": 895 }, { "epoch": 12.78942731277533, "grad_norm": 11.128039360046387, "learning_rate": 0.00017438667757034546, "loss": 3.0331, "step": 896 }, { "epoch": 12.80352422907489, "grad_norm": 10.911291122436523, "learning_rate": 0.00017432480142724787, "loss": 3.2238, "step": 897 }, { "epoch": 12.81762114537445, "grad_norm": 10.06009292602539, "learning_rate": 0.0001742628616406709, "loss": 3.0653, "step": 898 }, { "epoch": 12.831718061674009, "grad_norm": 10.17288589477539, "learning_rate": 0.00017420085826365283, "loss": 3.0459, "step": 899 }, { "epoch": 12.845814977973568, "grad_norm": 11.449868202209473, "learning_rate": 0.00017413879134928643, "loss": 3.0172, "step": 900 }, { "epoch": 12.859911894273127, "grad_norm": 9.854798316955566, "learning_rate": 0.00017407666095071894, "loss": 3.194, "step": 901 }, { "epoch": 12.874008810572688, "grad_norm": 11.030424118041992, "learning_rate": 0.00017401446712115185, "loss": 3.2523, "step": 902 }, { "epoch": 12.888105726872247, "grad_norm": 10.256535530090332, "learning_rate": 0.0001739522099138411, "loss": 3.1981, "step": 903 }, { "epoch": 12.902202643171806, "grad_norm": 11.527730941772461, "learning_rate": 0.00017388988938209677, "loss": 3.1651, "step": 904 }, { "epoch": 12.916299559471366, "grad_norm": 10.498260498046875, "learning_rate": 0.00017382750557928323, "loss": 3.1794, "step": 905 }, { "epoch": 12.930396475770925, "grad_norm": 11.297645568847656, "learning_rate": 0.00017376505855881905, "loss": 3.2531, "step": 906 }, { "epoch": 12.944493392070484, "grad_norm": 12.370041847229004, "learning_rate": 0.00017370254837417685, "loss": 3.0674, "step": 907 }, { "epoch": 12.958590308370043, "grad_norm": 10.075311660766602, "learning_rate": 0.00017363997507888342, "loss": 3.1443, "step": 908 }, { "epoch": 12.972687224669604, "grad_norm": 10.895628929138184, "learning_rate": 0.00017357733872651955, "loss": 3.2253, "step": 909 }, { "epoch": 12.986784140969164, "grad_norm": 10.376541137695312, "learning_rate": 0.00017351463937072004, "loss": 3.1135, "step": 910 }, { "epoch": 13.01409691629956, "grad_norm": 11.433059692382812, "learning_rate": 0.00017345187706517364, "loss": 4.3187, "step": 911 }, { "epoch": 13.028193832599118, "grad_norm": 8.003564834594727, "learning_rate": 0.000173389051863623, "loss": 1.7004, "step": 912 }, { "epoch": 13.042290748898678, "grad_norm": 16.3875732421875, "learning_rate": 0.00017332616381986456, "loss": 1.6745, "step": 913 }, { "epoch": 13.056387665198239, "grad_norm": 10.183499336242676, "learning_rate": 0.00017326321298774875, "loss": 1.6518, "step": 914 }, { "epoch": 13.070484581497798, "grad_norm": 19.106449127197266, "learning_rate": 0.00017320019942117956, "loss": 1.6949, "step": 915 }, { "epoch": 13.084581497797357, "grad_norm": 13.793717384338379, "learning_rate": 0.00017313712317411478, "loss": 1.6319, "step": 916 }, { "epoch": 13.098678414096916, "grad_norm": 11.020490646362305, "learning_rate": 0.00017307398430056593, "loss": 1.6603, "step": 917 }, { "epoch": 13.112775330396476, "grad_norm": 12.881837844848633, "learning_rate": 0.0001730107828545981, "loss": 1.6215, "step": 918 }, { "epoch": 13.126872246696035, "grad_norm": 8.825873374938965, "learning_rate": 0.00017294751889032996, "loss": 1.6333, "step": 919 }, { "epoch": 13.140969162995594, "grad_norm": 9.951656341552734, "learning_rate": 0.00017288419246193373, "loss": 1.7412, "step": 920 }, { "epoch": 13.155066079295155, "grad_norm": 9.380836486816406, "learning_rate": 0.00017282080362363507, "loss": 1.6253, "step": 921 }, { "epoch": 13.169162995594714, "grad_norm": 9.885567665100098, "learning_rate": 0.00017275735242971315, "loss": 1.724, "step": 922 }, { "epoch": 13.183259911894273, "grad_norm": 8.954022407531738, "learning_rate": 0.0001726938389345005, "loss": 1.6774, "step": 923 }, { "epoch": 13.197356828193833, "grad_norm": 10.03066349029541, "learning_rate": 0.00017263026319238301, "loss": 1.7214, "step": 924 }, { "epoch": 13.211453744493392, "grad_norm": 9.706664085388184, "learning_rate": 0.00017256662525779983, "loss": 1.6552, "step": 925 }, { "epoch": 13.225550660792951, "grad_norm": 11.061209678649902, "learning_rate": 0.00017250292518524344, "loss": 1.6767, "step": 926 }, { "epoch": 13.23964757709251, "grad_norm": 9.116509437561035, "learning_rate": 0.00017243916302925945, "loss": 1.5963, "step": 927 }, { "epoch": 13.25374449339207, "grad_norm": 9.825189590454102, "learning_rate": 0.00017237533884444667, "loss": 1.7195, "step": 928 }, { "epoch": 13.26784140969163, "grad_norm": 9.43676471710205, "learning_rate": 0.00017231145268545704, "loss": 1.6948, "step": 929 }, { "epoch": 13.28193832599119, "grad_norm": 11.239097595214844, "learning_rate": 0.00017224750460699556, "loss": 1.7214, "step": 930 }, { "epoch": 13.296035242290749, "grad_norm": 10.715683937072754, "learning_rate": 0.00017218349466382023, "loss": 1.6837, "step": 931 }, { "epoch": 13.310132158590308, "grad_norm": 9.40947151184082, "learning_rate": 0.000172119422910742, "loss": 1.7122, "step": 932 }, { "epoch": 13.324229074889868, "grad_norm": 9.92487621307373, "learning_rate": 0.0001720552894026248, "loss": 1.6781, "step": 933 }, { "epoch": 13.338325991189427, "grad_norm": 10.177139282226562, "learning_rate": 0.00017199109419438545, "loss": 1.8844, "step": 934 }, { "epoch": 13.352422907488986, "grad_norm": 9.973406791687012, "learning_rate": 0.00017192683734099353, "loss": 1.7743, "step": 935 }, { "epoch": 13.366519823788547, "grad_norm": 9.771232604980469, "learning_rate": 0.0001718625188974715, "loss": 1.6261, "step": 936 }, { "epoch": 13.380616740088106, "grad_norm": 9.485361099243164, "learning_rate": 0.00017179813891889445, "loss": 1.8207, "step": 937 }, { "epoch": 13.394713656387665, "grad_norm": 11.314799308776855, "learning_rate": 0.00017173369746039025, "loss": 1.7267, "step": 938 }, { "epoch": 13.408810572687225, "grad_norm": 9.11411190032959, "learning_rate": 0.00017166919457713942, "loss": 1.7376, "step": 939 }, { "epoch": 13.422907488986784, "grad_norm": 9.436162948608398, "learning_rate": 0.000171604630324375, "loss": 1.7857, "step": 940 }, { "epoch": 13.437004405286343, "grad_norm": 11.550241470336914, "learning_rate": 0.0001715400047573826, "loss": 1.8202, "step": 941 }, { "epoch": 13.451101321585902, "grad_norm": 12.73249626159668, "learning_rate": 0.00017147531793150042, "loss": 1.7063, "step": 942 }, { "epoch": 13.465198237885463, "grad_norm": 11.637757301330566, "learning_rate": 0.00017141056990211903, "loss": 1.8169, "step": 943 }, { "epoch": 13.479295154185023, "grad_norm": 15.088281631469727, "learning_rate": 0.00017134576072468143, "loss": 1.7909, "step": 944 }, { "epoch": 13.493392070484582, "grad_norm": 10.587425231933594, "learning_rate": 0.00017128089045468294, "loss": 1.7451, "step": 945 }, { "epoch": 13.507488986784141, "grad_norm": 15.796319961547852, "learning_rate": 0.0001712159591476713, "loss": 1.8881, "step": 946 }, { "epoch": 13.5215859030837, "grad_norm": 10.200616836547852, "learning_rate": 0.00017115096685924635, "loss": 1.8411, "step": 947 }, { "epoch": 13.53568281938326, "grad_norm": 13.758877754211426, "learning_rate": 0.00017108591364506032, "loss": 1.8333, "step": 948 }, { "epoch": 13.549779735682819, "grad_norm": 9.537091255187988, "learning_rate": 0.00017102079956081754, "loss": 1.7495, "step": 949 }, { "epoch": 13.56387665198238, "grad_norm": 11.289345741271973, "learning_rate": 0.0001709556246622744, "loss": 1.7682, "step": 950 }, { "epoch": 13.577973568281939, "grad_norm": 9.43992805480957, "learning_rate": 0.0001708903890052395, "loss": 1.8351, "step": 951 }, { "epoch": 13.592070484581498, "grad_norm": 8.926522254943848, "learning_rate": 0.0001708250926455733, "loss": 1.875, "step": 952 }, { "epoch": 13.606167400881057, "grad_norm": 9.883767127990723, "learning_rate": 0.00017075973563918844, "loss": 1.8526, "step": 953 }, { "epoch": 13.620264317180617, "grad_norm": 10.190685272216797, "learning_rate": 0.00017069431804204928, "loss": 1.808, "step": 954 }, { "epoch": 13.634361233480176, "grad_norm": 9.332940101623535, "learning_rate": 0.00017062883991017218, "loss": 1.9192, "step": 955 }, { "epoch": 13.648458149779735, "grad_norm": 10.973057746887207, "learning_rate": 0.00017056330129962532, "loss": 1.9127, "step": 956 }, { "epoch": 13.662555066079296, "grad_norm": 10.119280815124512, "learning_rate": 0.0001704977022665287, "loss": 1.8831, "step": 957 }, { "epoch": 13.676651982378855, "grad_norm": 10.184469223022461, "learning_rate": 0.00017043204286705402, "loss": 1.9209, "step": 958 }, { "epoch": 13.690748898678414, "grad_norm": 10.132298469543457, "learning_rate": 0.00017036632315742462, "loss": 1.8835, "step": 959 }, { "epoch": 13.704845814977974, "grad_norm": 10.221909523010254, "learning_rate": 0.00017030054319391562, "loss": 1.9792, "step": 960 }, { "epoch": 13.718942731277533, "grad_norm": 9.892573356628418, "learning_rate": 0.00017023470303285357, "loss": 1.9488, "step": 961 }, { "epoch": 13.733039647577092, "grad_norm": 10.380035400390625, "learning_rate": 0.00017016880273061665, "loss": 1.9279, "step": 962 }, { "epoch": 13.747136563876651, "grad_norm": 10.413626670837402, "learning_rate": 0.00017010284234363461, "loss": 2.0199, "step": 963 }, { "epoch": 13.761233480176212, "grad_norm": 10.147306442260742, "learning_rate": 0.00017003682192838853, "loss": 1.8349, "step": 964 }, { "epoch": 13.775330396475772, "grad_norm": 9.8746976852417, "learning_rate": 0.00016997074154141096, "loss": 1.9146, "step": 965 }, { "epoch": 13.78942731277533, "grad_norm": 9.449141502380371, "learning_rate": 0.00016990460123928575, "loss": 1.9277, "step": 966 }, { "epoch": 13.80352422907489, "grad_norm": 10.873295783996582, "learning_rate": 0.0001698384010786481, "loss": 1.8618, "step": 967 }, { "epoch": 13.81762114537445, "grad_norm": 8.70877742767334, "learning_rate": 0.00016977214111618447, "loss": 2.0446, "step": 968 }, { "epoch": 13.831718061674009, "grad_norm": 10.043000221252441, "learning_rate": 0.0001697058214086325, "loss": 1.963, "step": 969 }, { "epoch": 13.845814977973568, "grad_norm": 9.883713722229004, "learning_rate": 0.00016963944201278097, "loss": 1.9188, "step": 970 }, { "epoch": 13.859911894273127, "grad_norm": 9.960838317871094, "learning_rate": 0.00016957300298546982, "loss": 1.9824, "step": 971 }, { "epoch": 13.874008810572688, "grad_norm": 11.725117683410645, "learning_rate": 0.00016950650438359003, "loss": 2.0035, "step": 972 }, { "epoch": 13.888105726872247, "grad_norm": 9.454109191894531, "learning_rate": 0.00016943994626408363, "loss": 1.9866, "step": 973 }, { "epoch": 13.902202643171806, "grad_norm": 11.636567115783691, "learning_rate": 0.00016937332868394354, "loss": 1.9898, "step": 974 }, { "epoch": 13.916299559471366, "grad_norm": 12.874878883361816, "learning_rate": 0.00016930665170021363, "loss": 2.0538, "step": 975 }, { "epoch": 13.930396475770925, "grad_norm": 12.577007293701172, "learning_rate": 0.00016923991536998867, "loss": 2.1394, "step": 976 }, { "epoch": 13.944493392070484, "grad_norm": 10.060060501098633, "learning_rate": 0.00016917311975041418, "loss": 2.1013, "step": 977 }, { "epoch": 13.958590308370043, "grad_norm": 13.227431297302246, "learning_rate": 0.00016910626489868649, "loss": 2.1788, "step": 978 }, { "epoch": 13.972687224669604, "grad_norm": 11.649492263793945, "learning_rate": 0.00016903935087205266, "loss": 2.1408, "step": 979 }, { "epoch": 13.986784140969164, "grad_norm": 9.256147384643555, "learning_rate": 0.00016897237772781044, "loss": 1.9534, "step": 980 }, { "epoch": 14.01409691629956, "grad_norm": 8.927895545959473, "learning_rate": 0.0001689053455233081, "loss": 2.6356, "step": 981 }, { "epoch": 14.028193832599118, "grad_norm": 7.785157203674316, "learning_rate": 0.00016883825431594462, "loss": 0.977, "step": 982 }, { "epoch": 14.042290748898678, "grad_norm": 8.863943099975586, "learning_rate": 0.0001687711041631694, "loss": 0.947, "step": 983 }, { "epoch": 14.056387665198239, "grad_norm": 9.201424598693848, "learning_rate": 0.00016870389512248238, "loss": 1.0001, "step": 984 }, { "epoch": 14.070484581497798, "grad_norm": 10.621182441711426, "learning_rate": 0.00016863662725143382, "loss": 0.9653, "step": 985 }, { "epoch": 14.084581497797357, "grad_norm": 10.133315086364746, "learning_rate": 0.00016856930060762452, "loss": 0.9864, "step": 986 }, { "epoch": 14.098678414096916, "grad_norm": 9.230059623718262, "learning_rate": 0.00016850191524870546, "loss": 0.9618, "step": 987 }, { "epoch": 14.112775330396476, "grad_norm": 10.189534187316895, "learning_rate": 0.00016843447123237798, "loss": 0.9275, "step": 988 }, { "epoch": 14.126872246696035, "grad_norm": 7.3657307624816895, "learning_rate": 0.00016836696861639362, "loss": 0.9503, "step": 989 }, { "epoch": 14.140969162995594, "grad_norm": 8.661985397338867, "learning_rate": 0.00016829940745855408, "loss": 0.9638, "step": 990 }, { "epoch": 14.155066079295155, "grad_norm": 7.641331195831299, "learning_rate": 0.00016823178781671122, "loss": 1.0345, "step": 991 }, { "epoch": 14.169162995594714, "grad_norm": 9.017359733581543, "learning_rate": 0.00016816410974876698, "loss": 1.031, "step": 992 }, { "epoch": 14.183259911894273, "grad_norm": 8.791230201721191, "learning_rate": 0.00016809637331267332, "loss": 0.9446, "step": 993 }, { "epoch": 14.197356828193833, "grad_norm": 8.4191312789917, "learning_rate": 0.00016802857856643215, "loss": 1.0301, "step": 994 }, { "epoch": 14.211453744493392, "grad_norm": 8.902165412902832, "learning_rate": 0.00016796072556809534, "loss": 1.025, "step": 995 }, { "epoch": 14.225550660792951, "grad_norm": 7.930410861968994, "learning_rate": 0.00016789281437576468, "loss": 0.9334, "step": 996 }, { "epoch": 14.23964757709251, "grad_norm": 8.45251750946045, "learning_rate": 0.0001678248450475917, "loss": 0.9822, "step": 997 }, { "epoch": 14.25374449339207, "grad_norm": 10.511399269104004, "learning_rate": 0.00016775681764177777, "loss": 0.9966, "step": 998 }, { "epoch": 14.26784140969163, "grad_norm": 7.462294578552246, "learning_rate": 0.000167688732216574, "loss": 0.9382, "step": 999 }, { "epoch": 14.28193832599119, "grad_norm": 8.451595306396484, "learning_rate": 0.0001676205888302812, "loss": 0.9696, "step": 1000 }, { "epoch": 14.28193832599119, "eval_loss": 1.0723176002502441, "eval_runtime": 132.7015, "eval_samples_per_second": 13.625, "eval_steps_per_second": 0.43, "step": 1000 }, { "epoch": 14.296035242290749, "grad_norm": 8.874797821044922, "learning_rate": 0.00016755238754124965, "loss": 1.0044, "step": 1001 }, { "epoch": 14.310132158590308, "grad_norm": 8.380295753479004, "learning_rate": 0.00016748412840787946, "loss": 0.9833, "step": 1002 }, { "epoch": 14.324229074889868, "grad_norm": 7.869851589202881, "learning_rate": 0.00016741581148862008, "loss": 1.0249, "step": 1003 }, { "epoch": 14.338325991189427, "grad_norm": 10.056136131286621, "learning_rate": 0.00016734743684197054, "loss": 0.9839, "step": 1004 }, { "epoch": 14.352422907488986, "grad_norm": 9.355184555053711, "learning_rate": 0.00016727900452647926, "loss": 0.953, "step": 1005 }, { "epoch": 14.366519823788547, "grad_norm": 10.207636833190918, "learning_rate": 0.00016721051460074404, "loss": 1.0794, "step": 1006 }, { "epoch": 14.380616740088106, "grad_norm": 8.04494571685791, "learning_rate": 0.00016714196712341205, "loss": 1.0004, "step": 1007 }, { "epoch": 14.394713656387665, "grad_norm": 9.550114631652832, "learning_rate": 0.00016707336215317968, "loss": 0.9242, "step": 1008 }, { "epoch": 14.408810572687225, "grad_norm": 7.233541488647461, "learning_rate": 0.0001670046997487926, "loss": 0.9245, "step": 1009 }, { "epoch": 14.422907488986784, "grad_norm": 10.146906852722168, "learning_rate": 0.00016693597996904564, "loss": 0.9939, "step": 1010 }, { "epoch": 14.437004405286343, "grad_norm": 8.330248832702637, "learning_rate": 0.00016686720287278277, "loss": 1.0367, "step": 1011 }, { "epoch": 14.451101321585902, "grad_norm": 8.916499137878418, "learning_rate": 0.00016679836851889702, "loss": 1.0631, "step": 1012 }, { "epoch": 14.465198237885463, "grad_norm": 8.690364837646484, "learning_rate": 0.00016672947696633048, "loss": 1.011, "step": 1013 }, { "epoch": 14.479295154185023, "grad_norm": 8.742980003356934, "learning_rate": 0.00016666052827407418, "loss": 1.0094, "step": 1014 }, { "epoch": 14.493392070484582, "grad_norm": 8.191766738891602, "learning_rate": 0.00016659152250116812, "loss": 1.0736, "step": 1015 }, { "epoch": 14.507488986784141, "grad_norm": 9.328198432922363, "learning_rate": 0.00016652245970670112, "loss": 1.0524, "step": 1016 }, { "epoch": 14.5215859030837, "grad_norm": 7.813931941986084, "learning_rate": 0.00016645333994981095, "loss": 1.0607, "step": 1017 }, { "epoch": 14.53568281938326, "grad_norm": 8.290730476379395, "learning_rate": 0.00016638416328968394, "loss": 1.06, "step": 1018 }, { "epoch": 14.549779735682819, "grad_norm": 8.309978485107422, "learning_rate": 0.00016631492978555536, "loss": 1.022, "step": 1019 }, { "epoch": 14.56387665198238, "grad_norm": 9.718819618225098, "learning_rate": 0.00016624563949670904, "loss": 0.9887, "step": 1020 }, { "epoch": 14.577973568281939, "grad_norm": 8.545339584350586, "learning_rate": 0.0001661762924824775, "loss": 1.0614, "step": 1021 }, { "epoch": 14.592070484581498, "grad_norm": 12.660784721374512, "learning_rate": 0.00016610688880224178, "loss": 1.1334, "step": 1022 }, { "epoch": 14.606167400881057, "grad_norm": 8.326332092285156, "learning_rate": 0.0001660374285154314, "loss": 1.0228, "step": 1023 }, { "epoch": 14.620264317180617, "grad_norm": 9.4974946975708, "learning_rate": 0.0001659679116815245, "loss": 1.0842, "step": 1024 }, { "epoch": 14.634361233480176, "grad_norm": 8.71671199798584, "learning_rate": 0.00016589833836004752, "loss": 1.1496, "step": 1025 }, { "epoch": 14.648458149779735, "grad_norm": 9.390995025634766, "learning_rate": 0.00016582870861057525, "loss": 1.1181, "step": 1026 }, { "epoch": 14.662555066079296, "grad_norm": 8.80832290649414, "learning_rate": 0.00016575902249273093, "loss": 1.036, "step": 1027 }, { "epoch": 14.676651982378855, "grad_norm": 9.354620933532715, "learning_rate": 0.00016568928006618596, "loss": 1.0707, "step": 1028 }, { "epoch": 14.690748898678414, "grad_norm": 8.210136413574219, "learning_rate": 0.00016561948139065996, "loss": 1.1534, "step": 1029 }, { "epoch": 14.704845814977974, "grad_norm": 8.487049102783203, "learning_rate": 0.00016554962652592077, "loss": 1.1326, "step": 1030 }, { "epoch": 14.718942731277533, "grad_norm": 8.902432441711426, "learning_rate": 0.0001654797155317843, "loss": 1.0937, "step": 1031 }, { "epoch": 14.733039647577092, "grad_norm": 8.52372932434082, "learning_rate": 0.00016540974846811451, "loss": 1.1317, "step": 1032 }, { "epoch": 14.747136563876651, "grad_norm": 8.899264335632324, "learning_rate": 0.00016533972539482344, "loss": 1.1023, "step": 1033 }, { "epoch": 14.761233480176212, "grad_norm": 8.590658187866211, "learning_rate": 0.00016526964637187105, "loss": 1.0904, "step": 1034 }, { "epoch": 14.775330396475772, "grad_norm": 8.714057922363281, "learning_rate": 0.00016519951145926515, "loss": 1.1623, "step": 1035 }, { "epoch": 14.78942731277533, "grad_norm": 8.596945762634277, "learning_rate": 0.00016512932071706152, "loss": 1.1164, "step": 1036 }, { "epoch": 14.80352422907489, "grad_norm": 9.523910522460938, "learning_rate": 0.0001650590742053637, "loss": 1.1301, "step": 1037 }, { "epoch": 14.81762114537445, "grad_norm": 9.029563903808594, "learning_rate": 0.00016498877198432296, "loss": 1.1733, "step": 1038 }, { "epoch": 14.831718061674009, "grad_norm": 8.532072067260742, "learning_rate": 0.00016491841411413825, "loss": 1.1513, "step": 1039 }, { "epoch": 14.845814977973568, "grad_norm": 10.287323951721191, "learning_rate": 0.00016484800065505627, "loss": 1.1673, "step": 1040 }, { "epoch": 14.859911894273127, "grad_norm": 9.089445114135742, "learning_rate": 0.00016477753166737125, "loss": 1.1259, "step": 1041 }, { "epoch": 14.874008810572688, "grad_norm": 10.785905838012695, "learning_rate": 0.00016470700721142495, "loss": 1.1531, "step": 1042 }, { "epoch": 14.888105726872247, "grad_norm": 9.060176849365234, "learning_rate": 0.0001646364273476067, "loss": 1.2054, "step": 1043 }, { "epoch": 14.902202643171806, "grad_norm": 9.377079010009766, "learning_rate": 0.00016456579213635318, "loss": 1.1597, "step": 1044 }, { "epoch": 14.916299559471366, "grad_norm": 9.38338851928711, "learning_rate": 0.00016449510163814854, "loss": 1.2775, "step": 1045 }, { "epoch": 14.930396475770925, "grad_norm": 8.978507041931152, "learning_rate": 0.00016442435591352425, "loss": 1.2601, "step": 1046 }, { "epoch": 14.944493392070484, "grad_norm": 9.186592102050781, "learning_rate": 0.00016435355502305906, "loss": 1.1849, "step": 1047 }, { "epoch": 14.958590308370043, "grad_norm": 8.969529151916504, "learning_rate": 0.000164282699027379, "loss": 1.2143, "step": 1048 }, { "epoch": 14.972687224669604, "grad_norm": 9.939716339111328, "learning_rate": 0.0001642117879871572, "loss": 1.2542, "step": 1049 }, { "epoch": 14.986784140969164, "grad_norm": 9.727218627929688, "learning_rate": 0.000164140821963114, "loss": 1.2401, "step": 1050 }, { "epoch": 15.01409691629956, "grad_norm": 7.344968318939209, "learning_rate": 0.00016406980101601679, "loss": 1.5367, "step": 1051 }, { "epoch": 15.028193832599118, "grad_norm": 8.139293670654297, "learning_rate": 0.00016399872520668, "loss": 0.5857, "step": 1052 }, { "epoch": 15.042290748898678, "grad_norm": 7.200758934020996, "learning_rate": 0.00016392759459596504, "loss": 0.5493, "step": 1053 }, { "epoch": 15.056387665198239, "grad_norm": 6.879849910736084, "learning_rate": 0.00016385640924478022, "loss": 0.5945, "step": 1054 }, { "epoch": 15.070484581497798, "grad_norm": 8.077933311462402, "learning_rate": 0.00016378516921408077, "loss": 0.5557, "step": 1055 }, { "epoch": 15.084581497797357, "grad_norm": 7.245030879974365, "learning_rate": 0.0001637138745648687, "loss": 0.5604, "step": 1056 }, { "epoch": 15.098678414096916, "grad_norm": 6.544308185577393, "learning_rate": 0.00016364252535819282, "loss": 0.5731, "step": 1057 }, { "epoch": 15.112775330396476, "grad_norm": 7.236096382141113, "learning_rate": 0.00016357112165514862, "loss": 0.5513, "step": 1058 }, { "epoch": 15.126872246696035, "grad_norm": 7.260135173797607, "learning_rate": 0.0001634996635168783, "loss": 0.5408, "step": 1059 }, { "epoch": 15.140969162995594, "grad_norm": 6.2811970710754395, "learning_rate": 0.00016342815100457063, "loss": 0.5679, "step": 1060 }, { "epoch": 15.155066079295155, "grad_norm": 6.506096363067627, "learning_rate": 0.00016335658417946096, "loss": 0.5498, "step": 1061 }, { "epoch": 15.169162995594714, "grad_norm": 6.293437480926514, "learning_rate": 0.00016328496310283113, "loss": 0.5473, "step": 1062 }, { "epoch": 15.183259911894273, "grad_norm": 6.667970657348633, "learning_rate": 0.0001632132878360095, "loss": 0.5356, "step": 1063 }, { "epoch": 15.197356828193833, "grad_norm": 6.137166500091553, "learning_rate": 0.00016314155844037074, "loss": 0.5584, "step": 1064 }, { "epoch": 15.211453744493392, "grad_norm": 6.05562686920166, "learning_rate": 0.00016306977497733592, "loss": 0.5315, "step": 1065 }, { "epoch": 15.225550660792951, "grad_norm": 6.662344932556152, "learning_rate": 0.00016299793750837238, "loss": 0.5343, "step": 1066 }, { "epoch": 15.23964757709251, "grad_norm": 6.26501989364624, "learning_rate": 0.00016292604609499377, "loss": 0.5343, "step": 1067 }, { "epoch": 15.25374449339207, "grad_norm": 6.768255233764648, "learning_rate": 0.0001628541007987599, "loss": 0.515, "step": 1068 }, { "epoch": 15.26784140969163, "grad_norm": 7.057814598083496, "learning_rate": 0.00016278210168127664, "loss": 0.5708, "step": 1069 }, { "epoch": 15.28193832599119, "grad_norm": 5.88695764541626, "learning_rate": 0.00016271004880419608, "loss": 0.5427, "step": 1070 }, { "epoch": 15.296035242290749, "grad_norm": 6.989599227905273, "learning_rate": 0.0001626379422292162, "loss": 0.5289, "step": 1071 }, { "epoch": 15.310132158590308, "grad_norm": 6.5506272315979, "learning_rate": 0.00016256578201808114, "loss": 0.5197, "step": 1072 }, { "epoch": 15.324229074889868, "grad_norm": 7.221371650695801, "learning_rate": 0.00016249356823258072, "loss": 0.5359, "step": 1073 }, { "epoch": 15.338325991189427, "grad_norm": 6.590622425079346, "learning_rate": 0.00016242130093455096, "loss": 0.5877, "step": 1074 }, { "epoch": 15.352422907488986, "grad_norm": 6.053024768829346, "learning_rate": 0.00016234898018587337, "loss": 0.533, "step": 1075 }, { "epoch": 15.366519823788547, "grad_norm": 7.235744476318359, "learning_rate": 0.00016227660604847545, "loss": 0.607, "step": 1076 }, { "epoch": 15.380616740088106, "grad_norm": 5.981006622314453, "learning_rate": 0.00016220417858433037, "loss": 0.5103, "step": 1077 }, { "epoch": 15.394713656387665, "grad_norm": 5.912636756896973, "learning_rate": 0.0001621316978554569, "loss": 0.5316, "step": 1078 }, { "epoch": 15.408810572687225, "grad_norm": 6.263059616088867, "learning_rate": 0.00016205916392391946, "loss": 0.5378, "step": 1079 }, { "epoch": 15.422907488986784, "grad_norm": 7.0969557762146, "learning_rate": 0.000161986576851828, "loss": 0.5576, "step": 1080 }, { "epoch": 15.437004405286343, "grad_norm": 8.156744956970215, "learning_rate": 0.00016191393670133812, "loss": 0.5447, "step": 1081 }, { "epoch": 15.451101321585902, "grad_norm": 7.438736915588379, "learning_rate": 0.00016184124353465061, "loss": 0.5673, "step": 1082 }, { "epoch": 15.465198237885463, "grad_norm": 9.549761772155762, "learning_rate": 0.0001617684974140119, "loss": 0.5791, "step": 1083 }, { "epoch": 15.479295154185023, "grad_norm": 6.727791786193848, "learning_rate": 0.00016169569840171363, "loss": 0.5728, "step": 1084 }, { "epoch": 15.493392070484582, "grad_norm": 7.009186744689941, "learning_rate": 0.00016162284656009274, "loss": 0.5822, "step": 1085 }, { "epoch": 15.507488986784141, "grad_norm": 8.230440139770508, "learning_rate": 0.00016154994195153146, "loss": 0.5468, "step": 1086 }, { "epoch": 15.5215859030837, "grad_norm": 6.082698345184326, "learning_rate": 0.00016147698463845716, "loss": 0.5913, "step": 1087 }, { "epoch": 15.53568281938326, "grad_norm": 7.906459331512451, "learning_rate": 0.00016140397468334235, "loss": 0.5475, "step": 1088 }, { "epoch": 15.549779735682819, "grad_norm": 6.025688171386719, "learning_rate": 0.00016133091214870463, "loss": 0.5771, "step": 1089 }, { "epoch": 15.56387665198238, "grad_norm": 7.419466972351074, "learning_rate": 0.00016125779709710665, "loss": 0.5952, "step": 1090 }, { "epoch": 15.577973568281939, "grad_norm": 7.1458258628845215, "learning_rate": 0.00016118462959115594, "loss": 0.5649, "step": 1091 }, { "epoch": 15.592070484581498, "grad_norm": 7.558416843414307, "learning_rate": 0.00016111140969350503, "loss": 0.6033, "step": 1092 }, { "epoch": 15.606167400881057, "grad_norm": 7.142691135406494, "learning_rate": 0.00016103813746685126, "loss": 0.6033, "step": 1093 }, { "epoch": 15.620264317180617, "grad_norm": 7.792490005493164, "learning_rate": 0.00016096481297393683, "loss": 0.6081, "step": 1094 }, { "epoch": 15.634361233480176, "grad_norm": 9.550409317016602, "learning_rate": 0.00016089143627754862, "loss": 0.5971, "step": 1095 }, { "epoch": 15.648458149779735, "grad_norm": 6.469656467437744, "learning_rate": 0.00016081800744051833, "loss": 0.5889, "step": 1096 }, { "epoch": 15.662555066079296, "grad_norm": 8.49869155883789, "learning_rate": 0.00016074452652572218, "loss": 0.6253, "step": 1097 }, { "epoch": 15.676651982378855, "grad_norm": 6.780721187591553, "learning_rate": 0.00016067099359608105, "loss": 0.646, "step": 1098 }, { "epoch": 15.690748898678414, "grad_norm": 8.215495109558105, "learning_rate": 0.00016059740871456036, "loss": 0.6197, "step": 1099 }, { "epoch": 15.704845814977974, "grad_norm": 6.773916244506836, "learning_rate": 0.00016052377194417, "loss": 0.6147, "step": 1100 }, { "epoch": 15.718942731277533, "grad_norm": 8.132991790771484, "learning_rate": 0.00016045008334796427, "loss": 0.6286, "step": 1101 }, { "epoch": 15.733039647577092, "grad_norm": 6.990725517272949, "learning_rate": 0.00016037634298904192, "loss": 0.6115, "step": 1102 }, { "epoch": 15.747136563876651, "grad_norm": 7.722681045532227, "learning_rate": 0.00016030255093054595, "loss": 0.6318, "step": 1103 }, { "epoch": 15.761233480176212, "grad_norm": 6.781835556030273, "learning_rate": 0.00016022870723566367, "loss": 0.666, "step": 1104 }, { "epoch": 15.775330396475772, "grad_norm": 9.308338165283203, "learning_rate": 0.0001601548119676266, "loss": 0.6594, "step": 1105 }, { "epoch": 15.78942731277533, "grad_norm": 6.89076042175293, "learning_rate": 0.00016008086518971037, "loss": 0.6606, "step": 1106 }, { "epoch": 15.80352422907489, "grad_norm": 7.977464199066162, "learning_rate": 0.00016000686696523482, "loss": 0.579, "step": 1107 }, { "epoch": 15.81762114537445, "grad_norm": 7.189371585845947, "learning_rate": 0.00015993281735756374, "loss": 0.6372, "step": 1108 }, { "epoch": 15.831718061674009, "grad_norm": 6.517960548400879, "learning_rate": 0.000159858716430105, "loss": 0.6603, "step": 1109 }, { "epoch": 15.845814977973568, "grad_norm": 8.67850112915039, "learning_rate": 0.00015978456424631032, "loss": 0.6379, "step": 1110 }, { "epoch": 15.859911894273127, "grad_norm": 6.9540934562683105, "learning_rate": 0.00015971036086967545, "loss": 0.6374, "step": 1111 }, { "epoch": 15.874008810572688, "grad_norm": 6.850010395050049, "learning_rate": 0.00015963610636373984, "loss": 0.659, "step": 1112 }, { "epoch": 15.888105726872247, "grad_norm": 6.821133136749268, "learning_rate": 0.00015956180079208682, "loss": 0.6685, "step": 1113 }, { "epoch": 15.902202643171806, "grad_norm": 6.479661464691162, "learning_rate": 0.00015948744421834336, "loss": 0.631, "step": 1114 }, { "epoch": 15.916299559471366, "grad_norm": 6.7326507568359375, "learning_rate": 0.00015941303670618018, "loss": 0.6717, "step": 1115 }, { "epoch": 15.930396475770925, "grad_norm": 6.873997211456299, "learning_rate": 0.00015933857831931155, "loss": 0.6812, "step": 1116 }, { "epoch": 15.944493392070484, "grad_norm": 7.113956928253174, "learning_rate": 0.0001592640691214953, "loss": 0.6482, "step": 1117 }, { "epoch": 15.958590308370043, "grad_norm": 7.450249671936035, "learning_rate": 0.00015918950917653292, "loss": 0.7273, "step": 1118 }, { "epoch": 15.972687224669604, "grad_norm": 7.156855583190918, "learning_rate": 0.00015911489854826918, "loss": 0.6729, "step": 1119 }, { "epoch": 15.986784140969164, "grad_norm": 6.532139301300049, "learning_rate": 0.00015904023730059228, "loss": 0.6344, "step": 1120 }, { "epoch": 16.01409691629956, "grad_norm": 4.2600860595703125, "learning_rate": 0.00015896552549743382, "loss": 0.8063, "step": 1121 }, { "epoch": 16.02819383259912, "grad_norm": 4.353399753570557, "learning_rate": 0.0001588907632027687, "loss": 0.2972, "step": 1122 }, { "epoch": 16.04229074889868, "grad_norm": 5.9341278076171875, "learning_rate": 0.00015881595048061495, "loss": 0.2861, "step": 1123 }, { "epoch": 16.056387665198237, "grad_norm": 5.41513204574585, "learning_rate": 0.0001587410873950339, "loss": 0.3302, "step": 1124 }, { "epoch": 16.070484581497798, "grad_norm": 5.276197910308838, "learning_rate": 0.00015866617401012994, "loss": 0.3147, "step": 1125 }, { "epoch": 16.084581497797355, "grad_norm": 5.107941150665283, "learning_rate": 0.00015859121039005057, "loss": 0.3122, "step": 1126 }, { "epoch": 16.098678414096916, "grad_norm": 4.75752067565918, "learning_rate": 0.00015851619659898623, "loss": 0.3044, "step": 1127 }, { "epoch": 16.112775330396477, "grad_norm": 5.227985382080078, "learning_rate": 0.00015844113270117042, "loss": 0.3153, "step": 1128 }, { "epoch": 16.126872246696035, "grad_norm": 4.861165523529053, "learning_rate": 0.0001583660187608795, "loss": 0.3045, "step": 1129 }, { "epoch": 16.140969162995596, "grad_norm": 4.929010391235352, "learning_rate": 0.00015829085484243266, "loss": 0.3052, "step": 1130 }, { "epoch": 16.155066079295153, "grad_norm": 5.2300310134887695, "learning_rate": 0.0001582156410101919, "loss": 0.3083, "step": 1131 }, { "epoch": 16.169162995594714, "grad_norm": 5.222951412200928, "learning_rate": 0.00015814037732856198, "loss": 0.2985, "step": 1132 }, { "epoch": 16.18325991189427, "grad_norm": 4.882615089416504, "learning_rate": 0.0001580650638619903, "loss": 0.3109, "step": 1133 }, { "epoch": 16.197356828193833, "grad_norm": 5.777987003326416, "learning_rate": 0.000157989700674967, "loss": 0.2874, "step": 1134 }, { "epoch": 16.211453744493394, "grad_norm": 4.8152360916137695, "learning_rate": 0.00015791428783202465, "loss": 0.3423, "step": 1135 }, { "epoch": 16.22555066079295, "grad_norm": 4.862696170806885, "learning_rate": 0.00015783882539773845, "loss": 0.3155, "step": 1136 }, { "epoch": 16.239647577092512, "grad_norm": 4.975410461425781, "learning_rate": 0.00015776331343672596, "loss": 0.3048, "step": 1137 }, { "epoch": 16.25374449339207, "grad_norm": 4.358730316162109, "learning_rate": 0.00015768775201364723, "loss": 0.2888, "step": 1138 }, { "epoch": 16.26784140969163, "grad_norm": 4.973907470703125, "learning_rate": 0.00015761214119320468, "loss": 0.305, "step": 1139 }, { "epoch": 16.281938325991188, "grad_norm": 4.944393634796143, "learning_rate": 0.00015753648104014297, "loss": 0.3034, "step": 1140 }, { "epoch": 16.29603524229075, "grad_norm": 4.769799709320068, "learning_rate": 0.00015746077161924905, "loss": 0.2975, "step": 1141 }, { "epoch": 16.31013215859031, "grad_norm": 5.329863548278809, "learning_rate": 0.000157385012995352, "loss": 0.3147, "step": 1142 }, { "epoch": 16.324229074889868, "grad_norm": 5.345793724060059, "learning_rate": 0.0001573092052333231, "loss": 0.3045, "step": 1143 }, { "epoch": 16.33832599118943, "grad_norm": 4.948579788208008, "learning_rate": 0.00015723334839807564, "loss": 0.312, "step": 1144 }, { "epoch": 16.352422907488986, "grad_norm": 6.313098430633545, "learning_rate": 0.00015715744255456496, "loss": 0.3062, "step": 1145 }, { "epoch": 16.366519823788547, "grad_norm": 5.385476112365723, "learning_rate": 0.00015708148776778842, "loss": 0.316, "step": 1146 }, { "epoch": 16.380616740088104, "grad_norm": 4.6305975914001465, "learning_rate": 0.00015700548410278514, "loss": 0.3064, "step": 1147 }, { "epoch": 16.394713656387665, "grad_norm": 4.772847652435303, "learning_rate": 0.00015692943162463628, "loss": 0.3108, "step": 1148 }, { "epoch": 16.408810572687223, "grad_norm": 5.722555160522461, "learning_rate": 0.00015685333039846467, "loss": 0.3108, "step": 1149 }, { "epoch": 16.422907488986784, "grad_norm": 5.0282673835754395, "learning_rate": 0.00015677718048943496, "loss": 0.319, "step": 1150 }, { "epoch": 16.437004405286345, "grad_norm": 6.859221935272217, "learning_rate": 0.00015670098196275337, "loss": 0.3076, "step": 1151 }, { "epoch": 16.451101321585902, "grad_norm": 4.546107292175293, "learning_rate": 0.0001566247348836679, "loss": 0.3012, "step": 1152 }, { "epoch": 16.465198237885463, "grad_norm": 5.649548053741455, "learning_rate": 0.00015654843931746804, "loss": 0.3133, "step": 1153 }, { "epoch": 16.47929515418502, "grad_norm": 6.122790336608887, "learning_rate": 0.0001564720953294848, "loss": 0.3404, "step": 1154 }, { "epoch": 16.493392070484582, "grad_norm": 5.360540390014648, "learning_rate": 0.00015639570298509064, "loss": 0.3059, "step": 1155 }, { "epoch": 16.50748898678414, "grad_norm": 6.478765964508057, "learning_rate": 0.00015631926234969953, "loss": 0.3237, "step": 1156 }, { "epoch": 16.5215859030837, "grad_norm": 4.826329708099365, "learning_rate": 0.00015624277348876665, "loss": 0.3071, "step": 1157 }, { "epoch": 16.53568281938326, "grad_norm": 5.206096172332764, "learning_rate": 0.00015616623646778853, "loss": 0.3244, "step": 1158 }, { "epoch": 16.54977973568282, "grad_norm": 5.528533935546875, "learning_rate": 0.000156089651352303, "loss": 0.3279, "step": 1159 }, { "epoch": 16.56387665198238, "grad_norm": 4.777163505554199, "learning_rate": 0.00015601301820788898, "loss": 0.2974, "step": 1160 }, { "epoch": 16.577973568281937, "grad_norm": 5.815803527832031, "learning_rate": 0.0001559363371001666, "loss": 0.332, "step": 1161 }, { "epoch": 16.592070484581498, "grad_norm": 4.876204490661621, "learning_rate": 0.00015585960809479696, "loss": 0.3106, "step": 1162 }, { "epoch": 16.606167400881056, "grad_norm": 4.991979122161865, "learning_rate": 0.0001557828312574823, "loss": 0.3279, "step": 1163 }, { "epoch": 16.620264317180617, "grad_norm": 5.068574905395508, "learning_rate": 0.00015570600665396563, "loss": 0.3094, "step": 1164 }, { "epoch": 16.634361233480178, "grad_norm": 4.981606960296631, "learning_rate": 0.00015562913435003114, "loss": 0.3456, "step": 1165 }, { "epoch": 16.648458149779735, "grad_norm": 5.0262250900268555, "learning_rate": 0.00015555221441150357, "loss": 0.315, "step": 1166 }, { "epoch": 16.662555066079296, "grad_norm": 5.308091163635254, "learning_rate": 0.00015547524690424867, "loss": 0.3407, "step": 1167 }, { "epoch": 16.676651982378853, "grad_norm": 6.166476726531982, "learning_rate": 0.00015539823189417281, "loss": 0.3459, "step": 1168 }, { "epoch": 16.690748898678414, "grad_norm": 4.898022651672363, "learning_rate": 0.00015532116944722308, "loss": 0.3148, "step": 1169 }, { "epoch": 16.704845814977972, "grad_norm": 4.954848289489746, "learning_rate": 0.00015524405962938714, "loss": 0.3361, "step": 1170 }, { "epoch": 16.718942731277533, "grad_norm": 5.782449245452881, "learning_rate": 0.0001551669025066933, "loss": 0.3498, "step": 1171 }, { "epoch": 16.733039647577094, "grad_norm": 5.731858730316162, "learning_rate": 0.00015508969814521025, "loss": 0.3284, "step": 1172 }, { "epoch": 16.74713656387665, "grad_norm": 4.970213413238525, "learning_rate": 0.00015501244661104727, "loss": 0.3342, "step": 1173 }, { "epoch": 16.761233480176212, "grad_norm": 5.233306407928467, "learning_rate": 0.00015493514797035394, "loss": 0.3138, "step": 1174 }, { "epoch": 16.77533039647577, "grad_norm": 5.503791332244873, "learning_rate": 0.0001548578022893202, "loss": 0.3367, "step": 1175 }, { "epoch": 16.78942731277533, "grad_norm": 6.0833845138549805, "learning_rate": 0.0001547804096341763, "loss": 0.3597, "step": 1176 }, { "epoch": 16.80352422907489, "grad_norm": 5.133409023284912, "learning_rate": 0.00015470297007119264, "loss": 0.3382, "step": 1177 }, { "epoch": 16.81762114537445, "grad_norm": 6.033816814422607, "learning_rate": 0.0001546254836666799, "loss": 0.3379, "step": 1178 }, { "epoch": 16.83171806167401, "grad_norm": 6.4118852615356445, "learning_rate": 0.00015454795048698875, "loss": 0.3438, "step": 1179 }, { "epoch": 16.845814977973568, "grad_norm": 6.046572208404541, "learning_rate": 0.00015447037059851, "loss": 0.3668, "step": 1180 }, { "epoch": 16.85991189427313, "grad_norm": 6.2838568687438965, "learning_rate": 0.00015439274406767436, "loss": 0.3615, "step": 1181 }, { "epoch": 16.874008810572686, "grad_norm": 6.070758819580078, "learning_rate": 0.0001543150709609526, "loss": 0.3671, "step": 1182 }, { "epoch": 16.888105726872247, "grad_norm": 6.129275321960449, "learning_rate": 0.00015423735134485536, "loss": 0.3211, "step": 1183 }, { "epoch": 16.902202643171805, "grad_norm": 5.943122386932373, "learning_rate": 0.00015415958528593296, "loss": 0.3466, "step": 1184 }, { "epoch": 16.916299559471366, "grad_norm": 5.0976409912109375, "learning_rate": 0.00015408177285077565, "loss": 0.3392, "step": 1185 }, { "epoch": 16.930396475770927, "grad_norm": 5.657591342926025, "learning_rate": 0.00015400391410601332, "loss": 0.3833, "step": 1186 }, { "epoch": 16.944493392070484, "grad_norm": 5.64359188079834, "learning_rate": 0.0001539260091183155, "loss": 0.3756, "step": 1187 }, { "epoch": 16.958590308370045, "grad_norm": 5.2195940017700195, "learning_rate": 0.00015384805795439136, "loss": 0.3401, "step": 1188 }, { "epoch": 16.972687224669603, "grad_norm": 5.2666754722595215, "learning_rate": 0.0001537700606809896, "loss": 0.3518, "step": 1189 }, { "epoch": 16.986784140969164, "grad_norm": 5.2971391677856445, "learning_rate": 0.0001536920173648984, "loss": 0.3514, "step": 1190 }, { "epoch": 17.01409691629956, "grad_norm": 5.1455464363098145, "learning_rate": 0.00015361392807294535, "loss": 0.4809, "step": 1191 }, { "epoch": 17.02819383259912, "grad_norm": 3.5332822799682617, "learning_rate": 0.0001535357928719974, "loss": 0.1933, "step": 1192 }, { "epoch": 17.04229074889868, "grad_norm": 4.03357458114624, "learning_rate": 0.00015345761182896083, "loss": 0.1835, "step": 1193 }, { "epoch": 17.056387665198237, "grad_norm": 3.9437761306762695, "learning_rate": 0.00015337938501078127, "loss": 0.1689, "step": 1194 }, { "epoch": 17.070484581497798, "grad_norm": 3.4352126121520996, "learning_rate": 0.0001533011124844433, "loss": 0.1704, "step": 1195 }, { "epoch": 17.084581497797355, "grad_norm": 3.825273275375366, "learning_rate": 0.00015322279431697093, "loss": 0.1966, "step": 1196 }, { "epoch": 17.098678414096916, "grad_norm": 4.808329105377197, "learning_rate": 0.00015314443057542703, "loss": 0.1809, "step": 1197 }, { "epoch": 17.112775330396477, "grad_norm": 3.9495115280151367, "learning_rate": 0.0001530660213269136, "loss": 0.1996, "step": 1198 }, { "epoch": 17.126872246696035, "grad_norm": 4.190450668334961, "learning_rate": 0.0001529875666385716, "loss": 0.1739, "step": 1199 }, { "epoch": 17.140969162995596, "grad_norm": 4.704798698425293, "learning_rate": 0.00015290906657758083, "loss": 0.201, "step": 1200 }, { "epoch": 17.155066079295153, "grad_norm": 4.2649054527282715, "learning_rate": 0.00015283052121116007, "loss": 0.1762, "step": 1201 }, { "epoch": 17.169162995594714, "grad_norm": 4.044339656829834, "learning_rate": 0.00015275193060656678, "loss": 0.1852, "step": 1202 }, { "epoch": 17.18325991189427, "grad_norm": 4.407989501953125, "learning_rate": 0.00015267329483109726, "loss": 0.1909, "step": 1203 }, { "epoch": 17.197356828193833, "grad_norm": 4.484679222106934, "learning_rate": 0.00015259461395208628, "loss": 0.1997, "step": 1204 }, { "epoch": 17.211453744493394, "grad_norm": 4.208234786987305, "learning_rate": 0.00015251588803690753, "loss": 0.1802, "step": 1205 }, { "epoch": 17.22555066079295, "grad_norm": 3.871687889099121, "learning_rate": 0.000152437117152973, "loss": 0.1967, "step": 1206 }, { "epoch": 17.239647577092512, "grad_norm": 4.889731407165527, "learning_rate": 0.0001523583013677334, "loss": 0.1963, "step": 1207 }, { "epoch": 17.25374449339207, "grad_norm": 4.227999687194824, "learning_rate": 0.00015227944074867772, "loss": 0.1765, "step": 1208 }, { "epoch": 17.26784140969163, "grad_norm": 3.904053211212158, "learning_rate": 0.00015220053536333348, "loss": 0.191, "step": 1209 }, { "epoch": 17.281938325991188, "grad_norm": 5.085518836975098, "learning_rate": 0.00015212158527926634, "loss": 0.2017, "step": 1210 }, { "epoch": 17.29603524229075, "grad_norm": 4.48792839050293, "learning_rate": 0.00015204259056408046, "loss": 0.2082, "step": 1211 }, { "epoch": 17.31013215859031, "grad_norm": 4.509613037109375, "learning_rate": 0.0001519635512854181, "loss": 0.1961, "step": 1212 }, { "epoch": 17.324229074889868, "grad_norm": 5.314914703369141, "learning_rate": 0.00015188446751095963, "loss": 0.1767, "step": 1213 }, { "epoch": 17.33832599118943, "grad_norm": 4.620868682861328, "learning_rate": 0.00015180533930842364, "loss": 0.1717, "step": 1214 }, { "epoch": 17.352422907488986, "grad_norm": 3.8745651245117188, "learning_rate": 0.0001517261667455667, "loss": 0.189, "step": 1215 }, { "epoch": 17.366519823788547, "grad_norm": 4.950073719024658, "learning_rate": 0.0001516469498901834, "loss": 0.1736, "step": 1216 }, { "epoch": 17.380616740088104, "grad_norm": 4.198319435119629, "learning_rate": 0.00015156768881010612, "loss": 0.1929, "step": 1217 }, { "epoch": 17.394713656387665, "grad_norm": 4.156249523162842, "learning_rate": 0.00015148838357320537, "loss": 0.1917, "step": 1218 }, { "epoch": 17.408810572687223, "grad_norm": 4.531105041503906, "learning_rate": 0.00015140903424738919, "loss": 0.1916, "step": 1219 }, { "epoch": 17.422907488986784, "grad_norm": 5.966903209686279, "learning_rate": 0.00015132964090060357, "loss": 0.1876, "step": 1220 }, { "epoch": 17.437004405286345, "grad_norm": 4.246871471405029, "learning_rate": 0.00015125020360083212, "loss": 0.176, "step": 1221 }, { "epoch": 17.451101321585902, "grad_norm": 3.9358177185058594, "learning_rate": 0.00015117072241609608, "loss": 0.1925, "step": 1222 }, { "epoch": 17.465198237885463, "grad_norm": 5.734372138977051, "learning_rate": 0.00015109119741445428, "loss": 0.1682, "step": 1223 }, { "epoch": 17.47929515418502, "grad_norm": 4.11866569519043, "learning_rate": 0.0001510116286640031, "loss": 0.1895, "step": 1224 }, { "epoch": 17.493392070484582, "grad_norm": 4.244015693664551, "learning_rate": 0.00015093201623287631, "loss": 0.195, "step": 1225 }, { "epoch": 17.50748898678414, "grad_norm": 4.409208297729492, "learning_rate": 0.00015085236018924517, "loss": 0.1891, "step": 1226 }, { "epoch": 17.5215859030837, "grad_norm": 5.493556499481201, "learning_rate": 0.0001507726606013182, "loss": 0.1821, "step": 1227 }, { "epoch": 17.53568281938326, "grad_norm": 4.082369327545166, "learning_rate": 0.00015069291753734128, "loss": 0.1864, "step": 1228 }, { "epoch": 17.54977973568282, "grad_norm": 5.099764823913574, "learning_rate": 0.00015061313106559746, "loss": 0.1995, "step": 1229 }, { "epoch": 17.56387665198238, "grad_norm": 4.541256427764893, "learning_rate": 0.000150533301254407, "loss": 0.2006, "step": 1230 }, { "epoch": 17.577973568281937, "grad_norm": 4.217094898223877, "learning_rate": 0.00015045342817212727, "loss": 0.1908, "step": 1231 }, { "epoch": 17.592070484581498, "grad_norm": 4.366193771362305, "learning_rate": 0.00015037351188715265, "loss": 0.1844, "step": 1232 }, { "epoch": 17.606167400881056, "grad_norm": 4.3401336669921875, "learning_rate": 0.0001502935524679146, "loss": 0.186, "step": 1233 }, { "epoch": 17.620264317180617, "grad_norm": 3.8250644207000732, "learning_rate": 0.00015021354998288135, "loss": 0.1951, "step": 1234 }, { "epoch": 17.634361233480178, "grad_norm": 4.704011917114258, "learning_rate": 0.0001501335045005582, "loss": 0.199, "step": 1235 }, { "epoch": 17.648458149779735, "grad_norm": 4.064637660980225, "learning_rate": 0.00015005341608948717, "loss": 0.1879, "step": 1236 }, { "epoch": 17.662555066079296, "grad_norm": 4.242600917816162, "learning_rate": 0.00014997328481824708, "loss": 0.1801, "step": 1237 }, { "epoch": 17.676651982378853, "grad_norm": 3.619170665740967, "learning_rate": 0.00014989311075545335, "loss": 0.2029, "step": 1238 }, { "epoch": 17.690748898678414, "grad_norm": 4.5934367179870605, "learning_rate": 0.00014981289396975817, "loss": 0.2059, "step": 1239 }, { "epoch": 17.704845814977972, "grad_norm": 3.8898136615753174, "learning_rate": 0.00014973263452985024, "loss": 0.194, "step": 1240 }, { "epoch": 17.718942731277533, "grad_norm": 3.9668424129486084, "learning_rate": 0.0001496523325044548, "loss": 0.1988, "step": 1241 }, { "epoch": 17.733039647577094, "grad_norm": 3.8632431030273438, "learning_rate": 0.00014957198796233356, "loss": 0.1927, "step": 1242 }, { "epoch": 17.74713656387665, "grad_norm": 4.128298282623291, "learning_rate": 0.00014949160097228465, "loss": 0.2009, "step": 1243 }, { "epoch": 17.761233480176212, "grad_norm": 4.3342695236206055, "learning_rate": 0.0001494111716031425, "loss": 0.2251, "step": 1244 }, { "epoch": 17.77533039647577, "grad_norm": 4.527381420135498, "learning_rate": 0.00014933069992377793, "loss": 0.2167, "step": 1245 }, { "epoch": 17.78942731277533, "grad_norm": 4.217151641845703, "learning_rate": 0.00014925018600309785, "loss": 0.2091, "step": 1246 }, { "epoch": 17.80352422907489, "grad_norm": 3.8254435062408447, "learning_rate": 0.0001491696299100454, "loss": 0.1987, "step": 1247 }, { "epoch": 17.81762114537445, "grad_norm": 3.9555487632751465, "learning_rate": 0.0001490890317135999, "loss": 0.2069, "step": 1248 }, { "epoch": 17.83171806167401, "grad_norm": 4.342312335968018, "learning_rate": 0.00014900839148277665, "loss": 0.2028, "step": 1249 }, { "epoch": 17.845814977973568, "grad_norm": 4.06025505065918, "learning_rate": 0.00014892770928662697, "loss": 0.1985, "step": 1250 }, { "epoch": 17.845814977973568, "eval_loss": 1.2135447263717651, "eval_runtime": 132.5792, "eval_samples_per_second": 13.637, "eval_steps_per_second": 0.43, "step": 1250 }, { "epoch": 17.85991189427313, "grad_norm": 3.848024606704712, "learning_rate": 0.00014884698519423807, "loss": 0.2195, "step": 1251 }, { "epoch": 17.874008810572686, "grad_norm": 4.51121711730957, "learning_rate": 0.00014876621927473307, "loss": 0.1966, "step": 1252 }, { "epoch": 17.888105726872247, "grad_norm": 4.204681873321533, "learning_rate": 0.00014868541159727096, "loss": 0.2005, "step": 1253 }, { "epoch": 17.902202643171805, "grad_norm": 4.231730937957764, "learning_rate": 0.0001486045622310464, "loss": 0.2037, "step": 1254 }, { "epoch": 17.916299559471366, "grad_norm": 4.898785591125488, "learning_rate": 0.00014852367124528975, "loss": 0.2113, "step": 1255 }, { "epoch": 17.930396475770927, "grad_norm": 4.289423942565918, "learning_rate": 0.00014844273870926705, "loss": 0.196, "step": 1256 }, { "epoch": 17.944493392070484, "grad_norm": 4.074591636657715, "learning_rate": 0.00014836176469227994, "loss": 0.2098, "step": 1257 }, { "epoch": 17.958590308370045, "grad_norm": 4.574784755706787, "learning_rate": 0.0001482807492636655, "loss": 0.2131, "step": 1258 }, { "epoch": 17.972687224669603, "grad_norm": 4.786136150360107, "learning_rate": 0.00014819969249279638, "loss": 0.2196, "step": 1259 }, { "epoch": 17.986784140969164, "grad_norm": 5.073024749755859, "learning_rate": 0.00014811859444908052, "loss": 0.2132, "step": 1260 }, { "epoch": 18.01409691629956, "grad_norm": 4.092011451721191, "learning_rate": 0.00014803745520196125, "loss": 0.2987, "step": 1261 }, { "epoch": 18.02819383259912, "grad_norm": 3.562155246734619, "learning_rate": 0.0001479562748209172, "loss": 0.1298, "step": 1262 }, { "epoch": 18.04229074889868, "grad_norm": 3.2159245014190674, "learning_rate": 0.0001478750533754622, "loss": 0.1112, "step": 1263 }, { "epoch": 18.056387665198237, "grad_norm": 2.9857423305511475, "learning_rate": 0.00014779379093514522, "loss": 0.112, "step": 1264 }, { "epoch": 18.070484581497798, "grad_norm": 3.4215831756591797, "learning_rate": 0.00014771248756955042, "loss": 0.1224, "step": 1265 }, { "epoch": 18.084581497797355, "grad_norm": 3.7425010204315186, "learning_rate": 0.00014763114334829687, "loss": 0.1299, "step": 1266 }, { "epoch": 18.098678414096916, "grad_norm": 3.2803685665130615, "learning_rate": 0.00014754975834103877, "loss": 0.1097, "step": 1267 }, { "epoch": 18.112775330396477, "grad_norm": 3.3550188541412354, "learning_rate": 0.0001474683326174651, "loss": 0.122, "step": 1268 }, { "epoch": 18.126872246696035, "grad_norm": 3.553774833679199, "learning_rate": 0.00014738686624729986, "loss": 0.1236, "step": 1269 }, { "epoch": 18.140969162995596, "grad_norm": 3.291675567626953, "learning_rate": 0.00014730535930030172, "loss": 0.1368, "step": 1270 }, { "epoch": 18.155066079295153, "grad_norm": 3.9018638134002686, "learning_rate": 0.00014722381184626416, "loss": 0.1241, "step": 1271 }, { "epoch": 18.169162995594714, "grad_norm": 3.9384210109710693, "learning_rate": 0.00014714222395501535, "loss": 0.1339, "step": 1272 }, { "epoch": 18.18325991189427, "grad_norm": 3.5556485652923584, "learning_rate": 0.00014706059569641805, "loss": 0.1268, "step": 1273 }, { "epoch": 18.197356828193833, "grad_norm": 3.8244125843048096, "learning_rate": 0.00014697892714036958, "loss": 0.1152, "step": 1274 }, { "epoch": 18.211453744493394, "grad_norm": 3.2391648292541504, "learning_rate": 0.00014689721835680182, "loss": 0.1177, "step": 1275 }, { "epoch": 18.22555066079295, "grad_norm": 3.178727865219116, "learning_rate": 0.00014681546941568114, "loss": 0.1168, "step": 1276 }, { "epoch": 18.239647577092512, "grad_norm": 3.8708579540252686, "learning_rate": 0.0001467336803870081, "loss": 0.1184, "step": 1277 }, { "epoch": 18.25374449339207, "grad_norm": 3.591871500015259, "learning_rate": 0.00014665185134081778, "loss": 0.1192, "step": 1278 }, { "epoch": 18.26784140969163, "grad_norm": 3.37579345703125, "learning_rate": 0.00014656998234717944, "loss": 0.1222, "step": 1279 }, { "epoch": 18.281938325991188, "grad_norm": 3.534090518951416, "learning_rate": 0.00014648807347619663, "loss": 0.1269, "step": 1280 }, { "epoch": 18.29603524229075, "grad_norm": 4.108361721038818, "learning_rate": 0.00014640612479800686, "loss": 0.1314, "step": 1281 }, { "epoch": 18.31013215859031, "grad_norm": 3.800574779510498, "learning_rate": 0.00014632413638278194, "loss": 0.1087, "step": 1282 }, { "epoch": 18.324229074889868, "grad_norm": 3.2266008853912354, "learning_rate": 0.00014624210830072758, "loss": 0.1219, "step": 1283 }, { "epoch": 18.33832599118943, "grad_norm": 3.372022867202759, "learning_rate": 0.00014616004062208352, "loss": 0.1225, "step": 1284 }, { "epoch": 18.352422907488986, "grad_norm": 3.6478052139282227, "learning_rate": 0.00014607793341712334, "loss": 0.1189, "step": 1285 }, { "epoch": 18.366519823788547, "grad_norm": 3.485365629196167, "learning_rate": 0.00014599578675615453, "loss": 0.1156, "step": 1286 }, { "epoch": 18.380616740088104, "grad_norm": 3.455592393875122, "learning_rate": 0.00014591360070951833, "loss": 0.1128, "step": 1287 }, { "epoch": 18.394713656387665, "grad_norm": 3.5506317615509033, "learning_rate": 0.00014583137534758967, "loss": 0.126, "step": 1288 }, { "epoch": 18.408810572687223, "grad_norm": 3.3570377826690674, "learning_rate": 0.0001457491107407773, "loss": 0.113, "step": 1289 }, { "epoch": 18.422907488986784, "grad_norm": 3.2512075901031494, "learning_rate": 0.00014566680695952332, "loss": 0.1162, "step": 1290 }, { "epoch": 18.437004405286345, "grad_norm": 3.5002381801605225, "learning_rate": 0.00014558446407430365, "loss": 0.1231, "step": 1291 }, { "epoch": 18.451101321585902, "grad_norm": 3.7197799682617188, "learning_rate": 0.00014550208215562748, "loss": 0.1259, "step": 1292 }, { "epoch": 18.465198237885463, "grad_norm": 3.646031618118286, "learning_rate": 0.00014541966127403757, "loss": 0.1232, "step": 1293 }, { "epoch": 18.47929515418502, "grad_norm": 3.584728717803955, "learning_rate": 0.00014533720150010989, "loss": 0.1261, "step": 1294 }, { "epoch": 18.493392070484582, "grad_norm": 3.6809427738189697, "learning_rate": 0.00014525470290445392, "loss": 0.136, "step": 1295 }, { "epoch": 18.50748898678414, "grad_norm": 3.6057839393615723, "learning_rate": 0.00014517216555771215, "loss": 0.1256, "step": 1296 }, { "epoch": 18.5215859030837, "grad_norm": 3.4861605167388916, "learning_rate": 0.00014508958953056046, "loss": 0.1213, "step": 1297 }, { "epoch": 18.53568281938326, "grad_norm": 4.037405014038086, "learning_rate": 0.00014500697489370769, "loss": 0.1235, "step": 1298 }, { "epoch": 18.54977973568282, "grad_norm": 3.3069121837615967, "learning_rate": 0.00014492432171789584, "loss": 0.1406, "step": 1299 }, { "epoch": 18.56387665198238, "grad_norm": 3.2267117500305176, "learning_rate": 0.0001448416300738999, "loss": 0.1182, "step": 1300 }, { "epoch": 18.577973568281937, "grad_norm": 3.4424149990081787, "learning_rate": 0.00014475890003252777, "loss": 0.1345, "step": 1301 }, { "epoch": 18.592070484581498, "grad_norm": 4.131581783294678, "learning_rate": 0.00014467613166462023, "loss": 0.1201, "step": 1302 }, { "epoch": 18.606167400881056, "grad_norm": 3.2588789463043213, "learning_rate": 0.0001445933250410509, "loss": 0.1311, "step": 1303 }, { "epoch": 18.620264317180617, "grad_norm": 3.458087921142578, "learning_rate": 0.00014451048023272615, "loss": 0.1317, "step": 1304 }, { "epoch": 18.634361233480178, "grad_norm": 4.1016950607299805, "learning_rate": 0.000144427597310585, "loss": 0.142, "step": 1305 }, { "epoch": 18.648458149779735, "grad_norm": 3.9509596824645996, "learning_rate": 0.00014434467634559926, "loss": 0.1316, "step": 1306 }, { "epoch": 18.662555066079296, "grad_norm": 3.733926296234131, "learning_rate": 0.00014426171740877306, "loss": 0.139, "step": 1307 }, { "epoch": 18.676651982378853, "grad_norm": 3.6256072521209717, "learning_rate": 0.0001441787205711433, "loss": 0.1252, "step": 1308 }, { "epoch": 18.690748898678414, "grad_norm": 3.67409610748291, "learning_rate": 0.00014409568590377918, "loss": 0.144, "step": 1309 }, { "epoch": 18.704845814977972, "grad_norm": 3.654256582260132, "learning_rate": 0.00014401261347778233, "loss": 0.1403, "step": 1310 }, { "epoch": 18.718942731277533, "grad_norm": 3.4329934120178223, "learning_rate": 0.00014392950336428677, "loss": 0.1239, "step": 1311 }, { "epoch": 18.733039647577094, "grad_norm": 2.7705938816070557, "learning_rate": 0.0001438463556344587, "loss": 0.1316, "step": 1312 }, { "epoch": 18.74713656387665, "grad_norm": 3.7210757732391357, "learning_rate": 0.0001437631703594966, "loss": 0.1426, "step": 1313 }, { "epoch": 18.761233480176212, "grad_norm": 3.9810402393341064, "learning_rate": 0.000143679947610631, "loss": 0.1504, "step": 1314 }, { "epoch": 18.77533039647577, "grad_norm": 3.9315097332000732, "learning_rate": 0.0001435966874591247, "loss": 0.1311, "step": 1315 }, { "epoch": 18.78942731277533, "grad_norm": 3.3939552307128906, "learning_rate": 0.00014351338997627234, "loss": 0.1443, "step": 1316 }, { "epoch": 18.80352422907489, "grad_norm": 3.816542863845825, "learning_rate": 0.00014343005523340062, "loss": 0.1346, "step": 1317 }, { "epoch": 18.81762114537445, "grad_norm": 3.566352367401123, "learning_rate": 0.00014334668330186814, "loss": 0.1244, "step": 1318 }, { "epoch": 18.83171806167401, "grad_norm": 3.622767448425293, "learning_rate": 0.00014326327425306533, "loss": 0.1372, "step": 1319 }, { "epoch": 18.845814977973568, "grad_norm": 2.93361234664917, "learning_rate": 0.0001431798281584144, "loss": 0.1328, "step": 1320 }, { "epoch": 18.85991189427313, "grad_norm": 3.472195625305176, "learning_rate": 0.00014309634508936933, "loss": 0.129, "step": 1321 }, { "epoch": 18.874008810572686, "grad_norm": 4.127511978149414, "learning_rate": 0.00014301282511741564, "loss": 0.1386, "step": 1322 }, { "epoch": 18.888105726872247, "grad_norm": 4.117965221405029, "learning_rate": 0.00014292926831407061, "loss": 0.1381, "step": 1323 }, { "epoch": 18.902202643171805, "grad_norm": 3.7157928943634033, "learning_rate": 0.0001428456747508829, "loss": 0.1343, "step": 1324 }, { "epoch": 18.916299559471366, "grad_norm": 3.397097110748291, "learning_rate": 0.0001427620444994328, "loss": 0.1403, "step": 1325 }, { "epoch": 18.930396475770927, "grad_norm": 4.120760440826416, "learning_rate": 0.00014267837763133186, "loss": 0.147, "step": 1326 }, { "epoch": 18.944493392070484, "grad_norm": 3.6232621669769287, "learning_rate": 0.00014259467421822317, "loss": 0.1324, "step": 1327 }, { "epoch": 18.958590308370045, "grad_norm": 3.428833246231079, "learning_rate": 0.0001425109343317809, "loss": 0.1569, "step": 1328 }, { "epoch": 18.972687224669603, "grad_norm": 3.7129247188568115, "learning_rate": 0.00014242715804371064, "loss": 0.133, "step": 1329 }, { "epoch": 18.986784140969164, "grad_norm": 3.470876455307007, "learning_rate": 0.00014234334542574906, "loss": 0.1292, "step": 1330 }, { "epoch": 19.01409691629956, "grad_norm": 3.072256088256836, "learning_rate": 0.00014225949654966388, "loss": 0.1906, "step": 1331 }, { "epoch": 19.02819383259912, "grad_norm": 2.4466896057128906, "learning_rate": 0.00014217561148725398, "loss": 0.0814, "step": 1332 }, { "epoch": 19.04229074889868, "grad_norm": 2.7160122394561768, "learning_rate": 0.00014209169031034917, "loss": 0.0908, "step": 1333 }, { "epoch": 19.056387665198237, "grad_norm": 2.824671745300293, "learning_rate": 0.00014200773309081024, "loss": 0.0881, "step": 1334 }, { "epoch": 19.070484581497798, "grad_norm": 2.6180126667022705, "learning_rate": 0.00014192373990052877, "loss": 0.0721, "step": 1335 }, { "epoch": 19.084581497797355, "grad_norm": 2.628190755844116, "learning_rate": 0.00014183971081142716, "loss": 0.078, "step": 1336 }, { "epoch": 19.098678414096916, "grad_norm": 3.014003038406372, "learning_rate": 0.00014175564589545854, "loss": 0.0841, "step": 1337 }, { "epoch": 19.112775330396477, "grad_norm": 3.8305418491363525, "learning_rate": 0.00014167154522460682, "loss": 0.0936, "step": 1338 }, { "epoch": 19.126872246696035, "grad_norm": 2.7769086360931396, "learning_rate": 0.00014158740887088635, "loss": 0.0762, "step": 1339 }, { "epoch": 19.140969162995596, "grad_norm": 3.0749573707580566, "learning_rate": 0.0001415032369063422, "loss": 0.0781, "step": 1340 }, { "epoch": 19.155066079295153, "grad_norm": 2.8535444736480713, "learning_rate": 0.0001414190294030498, "loss": 0.0831, "step": 1341 }, { "epoch": 19.169162995594714, "grad_norm": 2.7177515029907227, "learning_rate": 0.00014133478643311512, "loss": 0.0835, "step": 1342 }, { "epoch": 19.18325991189427, "grad_norm": 2.933584690093994, "learning_rate": 0.00014125050806867447, "loss": 0.0888, "step": 1343 }, { "epoch": 19.197356828193833, "grad_norm": 3.431065797805786, "learning_rate": 0.0001411661943818944, "loss": 0.0791, "step": 1344 }, { "epoch": 19.211453744493394, "grad_norm": 2.6180272102355957, "learning_rate": 0.00014108184544497178, "loss": 0.0808, "step": 1345 }, { "epoch": 19.22555066079295, "grad_norm": 3.1151609420776367, "learning_rate": 0.00014099746133013367, "loss": 0.0817, "step": 1346 }, { "epoch": 19.239647577092512, "grad_norm": 2.9984402656555176, "learning_rate": 0.00014091304210963717, "loss": 0.0815, "step": 1347 }, { "epoch": 19.25374449339207, "grad_norm": 2.921313762664795, "learning_rate": 0.00014082858785576953, "loss": 0.0769, "step": 1348 }, { "epoch": 19.26784140969163, "grad_norm": 3.945882797241211, "learning_rate": 0.00014074409864084798, "loss": 0.0807, "step": 1349 }, { "epoch": 19.281938325991188, "grad_norm": 3.150618314743042, "learning_rate": 0.00014065957453721962, "loss": 0.0801, "step": 1350 }, { "epoch": 19.29603524229075, "grad_norm": 2.9976282119750977, "learning_rate": 0.00014057501561726157, "loss": 0.0841, "step": 1351 }, { "epoch": 19.31013215859031, "grad_norm": 3.1403846740722656, "learning_rate": 0.00014049042195338053, "loss": 0.0868, "step": 1352 }, { "epoch": 19.324229074889868, "grad_norm": 2.9880218505859375, "learning_rate": 0.00014040579361801315, "loss": 0.0809, "step": 1353 }, { "epoch": 19.33832599118943, "grad_norm": 3.420642614364624, "learning_rate": 0.00014032113068362576, "loss": 0.084, "step": 1354 }, { "epoch": 19.352422907488986, "grad_norm": 2.6853954792022705, "learning_rate": 0.00014023643322271426, "loss": 0.0944, "step": 1355 }, { "epoch": 19.366519823788547, "grad_norm": 3.0444562435150146, "learning_rate": 0.00014015170130780405, "loss": 0.0786, "step": 1356 }, { "epoch": 19.380616740088104, "grad_norm": 3.548752784729004, "learning_rate": 0.00014006693501145012, "loss": 0.0928, "step": 1357 }, { "epoch": 19.394713656387665, "grad_norm": 3.083965301513672, "learning_rate": 0.0001399821344062369, "loss": 0.1018, "step": 1358 }, { "epoch": 19.408810572687223, "grad_norm": 2.956439256668091, "learning_rate": 0.0001398972995647782, "loss": 0.0866, "step": 1359 }, { "epoch": 19.422907488986784, "grad_norm": 3.227774143218994, "learning_rate": 0.00013981243055971712, "loss": 0.0788, "step": 1360 }, { "epoch": 19.437004405286345, "grad_norm": 3.2048323154449463, "learning_rate": 0.00013972752746372596, "loss": 0.09, "step": 1361 }, { "epoch": 19.451101321585902, "grad_norm": 2.737722635269165, "learning_rate": 0.00013964259034950636, "loss": 0.0839, "step": 1362 }, { "epoch": 19.465198237885463, "grad_norm": 3.0307116508483887, "learning_rate": 0.00013955761928978894, "loss": 0.0864, "step": 1363 }, { "epoch": 19.47929515418502, "grad_norm": 2.808877468109131, "learning_rate": 0.0001394726143573335, "loss": 0.081, "step": 1364 }, { "epoch": 19.493392070484582, "grad_norm": 3.129826068878174, "learning_rate": 0.00013938757562492873, "loss": 0.0857, "step": 1365 }, { "epoch": 19.50748898678414, "grad_norm": 2.748124599456787, "learning_rate": 0.00013930250316539238, "loss": 0.0862, "step": 1366 }, { "epoch": 19.5215859030837, "grad_norm": 3.7344605922698975, "learning_rate": 0.00013921739705157098, "loss": 0.0973, "step": 1367 }, { "epoch": 19.53568281938326, "grad_norm": 3.0380783081054688, "learning_rate": 0.00013913225735633997, "loss": 0.0979, "step": 1368 }, { "epoch": 19.54977973568282, "grad_norm": 3.062678813934326, "learning_rate": 0.00013904708415260348, "loss": 0.0801, "step": 1369 }, { "epoch": 19.56387665198238, "grad_norm": 2.795175313949585, "learning_rate": 0.00013896187751329437, "loss": 0.0801, "step": 1370 }, { "epoch": 19.577973568281937, "grad_norm": 2.895685911178589, "learning_rate": 0.00013887663751137406, "loss": 0.0933, "step": 1371 }, { "epoch": 19.592070484581498, "grad_norm": 3.089449882507324, "learning_rate": 0.00013879136421983266, "loss": 0.0897, "step": 1372 }, { "epoch": 19.606167400881056, "grad_norm": 3.14072847366333, "learning_rate": 0.00013870605771168867, "loss": 0.0936, "step": 1373 }, { "epoch": 19.620264317180617, "grad_norm": 3.0422348976135254, "learning_rate": 0.0001386207180599891, "loss": 0.091, "step": 1374 }, { "epoch": 19.634361233480178, "grad_norm": 3.588981866836548, "learning_rate": 0.0001385353453378093, "loss": 0.1045, "step": 1375 }, { "epoch": 19.648458149779735, "grad_norm": 3.4622042179107666, "learning_rate": 0.00013844993961825298, "loss": 0.0911, "step": 1376 }, { "epoch": 19.662555066079296, "grad_norm": 3.607774257659912, "learning_rate": 0.00013836450097445209, "loss": 0.0941, "step": 1377 }, { "epoch": 19.676651982378853, "grad_norm": 4.0004377365112305, "learning_rate": 0.00013827902947956673, "loss": 0.0917, "step": 1378 }, { "epoch": 19.690748898678414, "grad_norm": 3.0106618404388428, "learning_rate": 0.0001381935252067852, "loss": 0.0993, "step": 1379 }, { "epoch": 19.704845814977972, "grad_norm": 3.235206127166748, "learning_rate": 0.00013810798822932378, "loss": 0.0843, "step": 1380 }, { "epoch": 19.718942731277533, "grad_norm": 3.3085057735443115, "learning_rate": 0.00013802241862042688, "loss": 0.0986, "step": 1381 }, { "epoch": 19.733039647577094, "grad_norm": 3.3103058338165283, "learning_rate": 0.00013793681645336674, "loss": 0.0895, "step": 1382 }, { "epoch": 19.74713656387665, "grad_norm": 3.5305607318878174, "learning_rate": 0.00013785118180144353, "loss": 0.0898, "step": 1383 }, { "epoch": 19.761233480176212, "grad_norm": 3.447258472442627, "learning_rate": 0.00013776551473798523, "loss": 0.0969, "step": 1384 }, { "epoch": 19.77533039647577, "grad_norm": 2.838228940963745, "learning_rate": 0.00013767981533634754, "loss": 0.0913, "step": 1385 }, { "epoch": 19.78942731277533, "grad_norm": 2.8133833408355713, "learning_rate": 0.0001375940836699139, "loss": 0.0908, "step": 1386 }, { "epoch": 19.80352422907489, "grad_norm": 3.0196175575256348, "learning_rate": 0.0001375083198120954, "loss": 0.0968, "step": 1387 }, { "epoch": 19.81762114537445, "grad_norm": 2.985166549682617, "learning_rate": 0.00013742252383633057, "loss": 0.0839, "step": 1388 }, { "epoch": 19.83171806167401, "grad_norm": 2.993781328201294, "learning_rate": 0.00013733669581608558, "loss": 0.0959, "step": 1389 }, { "epoch": 19.845814977973568, "grad_norm": 3.602381706237793, "learning_rate": 0.00013725083582485397, "loss": 0.1006, "step": 1390 }, { "epoch": 19.85991189427313, "grad_norm": 2.9341304302215576, "learning_rate": 0.00013716494393615665, "loss": 0.098, "step": 1391 }, { "epoch": 19.874008810572686, "grad_norm": 3.8942208290100098, "learning_rate": 0.0001370790202235419, "loss": 0.0889, "step": 1392 }, { "epoch": 19.888105726872247, "grad_norm": 3.528104305267334, "learning_rate": 0.0001369930647605852, "loss": 0.0989, "step": 1393 }, { "epoch": 19.902202643171805, "grad_norm": 3.2068607807159424, "learning_rate": 0.00013690707762088926, "loss": 0.1026, "step": 1394 }, { "epoch": 19.916299559471366, "grad_norm": 3.066638946533203, "learning_rate": 0.0001368210588780838, "loss": 0.0977, "step": 1395 }, { "epoch": 19.930396475770927, "grad_norm": 3.2156033515930176, "learning_rate": 0.0001367350086058258, "loss": 0.093, "step": 1396 }, { "epoch": 19.944493392070484, "grad_norm": 3.182859420776367, "learning_rate": 0.00013664892687779905, "loss": 0.1036, "step": 1397 }, { "epoch": 19.958590308370045, "grad_norm": 3.3807435035705566, "learning_rate": 0.00013656281376771439, "loss": 0.0999, "step": 1398 }, { "epoch": 19.972687224669603, "grad_norm": 2.6448657512664795, "learning_rate": 0.00013647666934930944, "loss": 0.0897, "step": 1399 }, { "epoch": 19.986784140969164, "grad_norm": 2.8451836109161377, "learning_rate": 0.00013639049369634876, "loss": 0.0987, "step": 1400 }, { "epoch": 20.01409691629956, "grad_norm": 2.6123745441436768, "learning_rate": 0.00013630428688262352, "loss": 0.1385, "step": 1401 }, { "epoch": 20.02819383259912, "grad_norm": 1.9358397722244263, "learning_rate": 0.00013621804898195168, "loss": 0.0539, "step": 1402 }, { "epoch": 20.04229074889868, "grad_norm": 2.462815523147583, "learning_rate": 0.0001361317800681777, "loss": 0.0575, "step": 1403 }, { "epoch": 20.056387665198237, "grad_norm": 2.1757452487945557, "learning_rate": 0.00013604548021517275, "loss": 0.0539, "step": 1404 }, { "epoch": 20.070484581497798, "grad_norm": 2.5211713314056396, "learning_rate": 0.00013595914949683432, "loss": 0.0554, "step": 1405 }, { "epoch": 20.084581497797355, "grad_norm": 2.8848798274993896, "learning_rate": 0.00013587278798708646, "loss": 0.057, "step": 1406 }, { "epoch": 20.098678414096916, "grad_norm": 2.42427659034729, "learning_rate": 0.00013578639575987958, "loss": 0.062, "step": 1407 }, { "epoch": 20.112775330396477, "grad_norm": 3.4246327877044678, "learning_rate": 0.00013569997288919033, "loss": 0.0569, "step": 1408 }, { "epoch": 20.126872246696035, "grad_norm": 2.2734670639038086, "learning_rate": 0.00013561351944902164, "loss": 0.058, "step": 1409 }, { "epoch": 20.140969162995596, "grad_norm": 2.62172794342041, "learning_rate": 0.00013552703551340258, "loss": 0.0585, "step": 1410 }, { "epoch": 20.155066079295153, "grad_norm": 2.1830058097839355, "learning_rate": 0.00013544052115638838, "loss": 0.0565, "step": 1411 }, { "epoch": 20.169162995594714, "grad_norm": 2.343419075012207, "learning_rate": 0.00013535397645206024, "loss": 0.0614, "step": 1412 }, { "epoch": 20.18325991189427, "grad_norm": 2.535973072052002, "learning_rate": 0.00013526740147452553, "loss": 0.0591, "step": 1413 }, { "epoch": 20.197356828193833, "grad_norm": 2.6030547618865967, "learning_rate": 0.00013518079629791724, "loss": 0.0746, "step": 1414 }, { "epoch": 20.211453744493394, "grad_norm": 2.129439115524292, "learning_rate": 0.00013509416099639457, "loss": 0.0561, "step": 1415 }, { "epoch": 20.22555066079295, "grad_norm": 2.547295093536377, "learning_rate": 0.00013500749564414224, "loss": 0.069, "step": 1416 }, { "epoch": 20.239647577092512, "grad_norm": 2.147305488586426, "learning_rate": 0.0001349208003153708, "loss": 0.0633, "step": 1417 }, { "epoch": 20.25374449339207, "grad_norm": 2.8901984691619873, "learning_rate": 0.0001348340750843165, "loss": 0.0567, "step": 1418 }, { "epoch": 20.26784140969163, "grad_norm": 2.069591522216797, "learning_rate": 0.00013474732002524115, "loss": 0.0613, "step": 1419 }, { "epoch": 20.281938325991188, "grad_norm": 2.0721447467803955, "learning_rate": 0.00013466053521243214, "loss": 0.0559, "step": 1420 }, { "epoch": 20.29603524229075, "grad_norm": 2.351008415222168, "learning_rate": 0.0001345737207202023, "loss": 0.0555, "step": 1421 }, { "epoch": 20.31013215859031, "grad_norm": 2.1646671295166016, "learning_rate": 0.00013448687662288988, "loss": 0.0543, "step": 1422 }, { "epoch": 20.324229074889868, "grad_norm": 2.2358577251434326, "learning_rate": 0.0001344000029948585, "loss": 0.0684, "step": 1423 }, { "epoch": 20.33832599118943, "grad_norm": 2.5909204483032227, "learning_rate": 0.00013431309991049706, "loss": 0.0636, "step": 1424 }, { "epoch": 20.352422907488986, "grad_norm": 2.594774007797241, "learning_rate": 0.00013422616744421966, "loss": 0.061, "step": 1425 }, { "epoch": 20.366519823788547, "grad_norm": 1.8119016885757446, "learning_rate": 0.00013413920567046563, "loss": 0.0609, "step": 1426 }, { "epoch": 20.380616740088104, "grad_norm": 3.1045916080474854, "learning_rate": 0.0001340522146636993, "loss": 0.0618, "step": 1427 }, { "epoch": 20.394713656387665, "grad_norm": 2.4877853393554688, "learning_rate": 0.00013396519449841005, "loss": 0.0543, "step": 1428 }, { "epoch": 20.408810572687223, "grad_norm": 2.713595390319824, "learning_rate": 0.0001338781452491123, "loss": 0.0667, "step": 1429 }, { "epoch": 20.422907488986784, "grad_norm": 2.5121889114379883, "learning_rate": 0.00013379106699034537, "loss": 0.0584, "step": 1430 }, { "epoch": 20.437004405286345, "grad_norm": 2.477665424346924, "learning_rate": 0.00013370395979667328, "loss": 0.0659, "step": 1431 }, { "epoch": 20.451101321585902, "grad_norm": 3.100282907485962, "learning_rate": 0.000133616823742685, "loss": 0.0647, "step": 1432 }, { "epoch": 20.465198237885463, "grad_norm": 2.2814414501190186, "learning_rate": 0.00013352965890299408, "loss": 0.0651, "step": 1433 }, { "epoch": 20.47929515418502, "grad_norm": 2.822373151779175, "learning_rate": 0.00013344246535223884, "loss": 0.0615, "step": 1434 }, { "epoch": 20.493392070484582, "grad_norm": 2.387166738510132, "learning_rate": 0.00013335524316508208, "loss": 0.0618, "step": 1435 }, { "epoch": 20.50748898678414, "grad_norm": 2.2945895195007324, "learning_rate": 0.00013326799241621118, "loss": 0.0621, "step": 1436 }, { "epoch": 20.5215859030837, "grad_norm": 2.5794119834899902, "learning_rate": 0.00013318071318033797, "loss": 0.0707, "step": 1437 }, { "epoch": 20.53568281938326, "grad_norm": 2.262784719467163, "learning_rate": 0.00013309340553219862, "loss": 0.0536, "step": 1438 }, { "epoch": 20.54977973568282, "grad_norm": 2.4234402179718018, "learning_rate": 0.00013300606954655373, "loss": 0.0591, "step": 1439 }, { "epoch": 20.56387665198238, "grad_norm": 2.4119112491607666, "learning_rate": 0.0001329187052981881, "loss": 0.0594, "step": 1440 }, { "epoch": 20.577973568281937, "grad_norm": 2.5160562992095947, "learning_rate": 0.00013283131286191067, "loss": 0.0603, "step": 1441 }, { "epoch": 20.592070484581498, "grad_norm": 2.4712822437286377, "learning_rate": 0.00013274389231255466, "loss": 0.0581, "step": 1442 }, { "epoch": 20.606167400881056, "grad_norm": 2.2953784465789795, "learning_rate": 0.0001326564437249773, "loss": 0.0585, "step": 1443 }, { "epoch": 20.620264317180617, "grad_norm": 2.7347171306610107, "learning_rate": 0.00013256896717405976, "loss": 0.0774, "step": 1444 }, { "epoch": 20.634361233480178, "grad_norm": 2.7829556465148926, "learning_rate": 0.00013248146273470726, "loss": 0.0683, "step": 1445 }, { "epoch": 20.648458149779735, "grad_norm": 2.78861403465271, "learning_rate": 0.0001323939304818488, "loss": 0.0746, "step": 1446 }, { "epoch": 20.662555066079296, "grad_norm": 3.0113515853881836, "learning_rate": 0.00013230637049043727, "loss": 0.0734, "step": 1447 }, { "epoch": 20.676651982378853, "grad_norm": 2.562645673751831, "learning_rate": 0.00013221878283544932, "loss": 0.0639, "step": 1448 }, { "epoch": 20.690748898678414, "grad_norm": 2.275923490524292, "learning_rate": 0.00013213116759188523, "loss": 0.0681, "step": 1449 }, { "epoch": 20.704845814977972, "grad_norm": 2.4684221744537354, "learning_rate": 0.00013204352483476897, "loss": 0.0678, "step": 1450 }, { "epoch": 20.718942731277533, "grad_norm": 2.2798893451690674, "learning_rate": 0.00013195585463914798, "loss": 0.0596, "step": 1451 }, { "epoch": 20.733039647577094, "grad_norm": 2.6210591793060303, "learning_rate": 0.0001318681570800933, "loss": 0.0649, "step": 1452 }, { "epoch": 20.74713656387665, "grad_norm": 2.5580661296844482, "learning_rate": 0.0001317804322326993, "loss": 0.0672, "step": 1453 }, { "epoch": 20.761233480176212, "grad_norm": 2.9215903282165527, "learning_rate": 0.00013169268017208382, "loss": 0.0642, "step": 1454 }, { "epoch": 20.77533039647577, "grad_norm": 2.669267416000366, "learning_rate": 0.0001316049009733879, "loss": 0.0664, "step": 1455 }, { "epoch": 20.78942731277533, "grad_norm": 2.8457789421081543, "learning_rate": 0.00013151709471177588, "loss": 0.0637, "step": 1456 }, { "epoch": 20.80352422907489, "grad_norm": 2.6282598972320557, "learning_rate": 0.00013142926146243527, "loss": 0.062, "step": 1457 }, { "epoch": 20.81762114537445, "grad_norm": 2.5472371578216553, "learning_rate": 0.00013134140130057666, "loss": 0.0631, "step": 1458 }, { "epoch": 20.83171806167401, "grad_norm": 2.3707315921783447, "learning_rate": 0.00013125351430143373, "loss": 0.0584, "step": 1459 }, { "epoch": 20.845814977973568, "grad_norm": 2.995197296142578, "learning_rate": 0.0001311656005402631, "loss": 0.0665, "step": 1460 }, { "epoch": 20.85991189427313, "grad_norm": 2.456331491470337, "learning_rate": 0.00013107766009234434, "loss": 0.0619, "step": 1461 }, { "epoch": 20.874008810572686, "grad_norm": 2.8263378143310547, "learning_rate": 0.0001309896930329798, "loss": 0.069, "step": 1462 }, { "epoch": 20.888105726872247, "grad_norm": 2.771138906478882, "learning_rate": 0.00013090169943749476, "loss": 0.0663, "step": 1463 }, { "epoch": 20.902202643171805, "grad_norm": 2.888237476348877, "learning_rate": 0.00013081367938123703, "loss": 0.0756, "step": 1464 }, { "epoch": 20.916299559471366, "grad_norm": 2.9230668544769287, "learning_rate": 0.00013072563293957723, "loss": 0.0603, "step": 1465 }, { "epoch": 20.930396475770927, "grad_norm": 2.3225624561309814, "learning_rate": 0.00013063756018790853, "loss": 0.0611, "step": 1466 }, { "epoch": 20.944493392070484, "grad_norm": 2.5413095951080322, "learning_rate": 0.00013054946120164663, "loss": 0.0776, "step": 1467 }, { "epoch": 20.958590308370045, "grad_norm": 2.8262338638305664, "learning_rate": 0.00013046133605622965, "loss": 0.0796, "step": 1468 }, { "epoch": 20.972687224669603, "grad_norm": 2.9958410263061523, "learning_rate": 0.0001303731848271182, "loss": 0.07, "step": 1469 }, { "epoch": 20.986784140969164, "grad_norm": 2.5129499435424805, "learning_rate": 0.00013028500758979506, "loss": 0.0765, "step": 1470 }, { "epoch": 21.01409691629956, "grad_norm": 2.4091079235076904, "learning_rate": 0.00013019680441976554, "loss": 0.0994, "step": 1471 }, { "epoch": 21.02819383259912, "grad_norm": 2.093433141708374, "learning_rate": 0.0001301085753925569, "loss": 0.0424, "step": 1472 }, { "epoch": 21.04229074889868, "grad_norm": 2.207012414932251, "learning_rate": 0.0001300203205837187, "loss": 0.0399, "step": 1473 }, { "epoch": 21.056387665198237, "grad_norm": 2.23032808303833, "learning_rate": 0.00012993204006882253, "loss": 0.0519, "step": 1474 }, { "epoch": 21.070484581497798, "grad_norm": 2.0519959926605225, "learning_rate": 0.00012984373392346194, "loss": 0.0436, "step": 1475 }, { "epoch": 21.084581497797355, "grad_norm": 2.0391275882720947, "learning_rate": 0.00012975540222325254, "loss": 0.0424, "step": 1476 }, { "epoch": 21.098678414096916, "grad_norm": 1.9109488725662231, "learning_rate": 0.00012966704504383168, "loss": 0.0425, "step": 1477 }, { "epoch": 21.112775330396477, "grad_norm": 2.4588606357574463, "learning_rate": 0.00012957866246085867, "loss": 0.0459, "step": 1478 }, { "epoch": 21.126872246696035, "grad_norm": 2.1022377014160156, "learning_rate": 0.00012949025455001448, "loss": 0.0465, "step": 1479 }, { "epoch": 21.140969162995596, "grad_norm": 2.5386996269226074, "learning_rate": 0.0001294018213870018, "loss": 0.053, "step": 1480 }, { "epoch": 21.155066079295153, "grad_norm": 2.514021873474121, "learning_rate": 0.00012931336304754492, "loss": 0.0442, "step": 1481 }, { "epoch": 21.169162995594714, "grad_norm": 2.7732040882110596, "learning_rate": 0.00012922487960738977, "loss": 0.0435, "step": 1482 }, { "epoch": 21.18325991189427, "grad_norm": 2.365037441253662, "learning_rate": 0.00012913637114230365, "loss": 0.0547, "step": 1483 }, { "epoch": 21.197356828193833, "grad_norm": 2.2920572757720947, "learning_rate": 0.00012904783772807533, "loss": 0.0463, "step": 1484 }, { "epoch": 21.211453744493394, "grad_norm": 2.581615686416626, "learning_rate": 0.00012895927944051502, "loss": 0.0527, "step": 1485 }, { "epoch": 21.22555066079295, "grad_norm": 2.303504705429077, "learning_rate": 0.00012887069635545415, "loss": 0.0553, "step": 1486 }, { "epoch": 21.239647577092512, "grad_norm": 2.2851970195770264, "learning_rate": 0.0001287820885487454, "loss": 0.0471, "step": 1487 }, { "epoch": 21.25374449339207, "grad_norm": 2.3485944271087646, "learning_rate": 0.00012869345609626263, "loss": 0.0551, "step": 1488 }, { "epoch": 21.26784140969163, "grad_norm": 2.057971954345703, "learning_rate": 0.00012860479907390076, "loss": 0.0445, "step": 1489 }, { "epoch": 21.281938325991188, "grad_norm": 2.3579273223876953, "learning_rate": 0.00012851611755757586, "loss": 0.0403, "step": 1490 }, { "epoch": 21.29603524229075, "grad_norm": 3.0866751670837402, "learning_rate": 0.00012842741162322487, "loss": 0.0478, "step": 1491 }, { "epoch": 21.31013215859031, "grad_norm": 2.077207565307617, "learning_rate": 0.0001283386813468056, "loss": 0.0469, "step": 1492 }, { "epoch": 21.324229074889868, "grad_norm": 2.0973665714263916, "learning_rate": 0.00012824992680429692, "loss": 0.0443, "step": 1493 }, { "epoch": 21.33832599118943, "grad_norm": 2.97033953666687, "learning_rate": 0.0001281611480716982, "loss": 0.0509, "step": 1494 }, { "epoch": 21.352422907488986, "grad_norm": 2.5459916591644287, "learning_rate": 0.00012807234522502968, "loss": 0.0495, "step": 1495 }, { "epoch": 21.366519823788547, "grad_norm": 2.366929292678833, "learning_rate": 0.00012798351834033227, "loss": 0.0465, "step": 1496 }, { "epoch": 21.380616740088104, "grad_norm": 2.3903558254241943, "learning_rate": 0.00012789466749366741, "loss": 0.047, "step": 1497 }, { "epoch": 21.394713656387665, "grad_norm": 1.9591631889343262, "learning_rate": 0.00012780579276111702, "loss": 0.0408, "step": 1498 }, { "epoch": 21.408810572687223, "grad_norm": 2.818485975265503, "learning_rate": 0.00012771689421878357, "loss": 0.0526, "step": 1499 }, { "epoch": 21.422907488986784, "grad_norm": 2.8614702224731445, "learning_rate": 0.00012762797194278984, "loss": 0.0494, "step": 1500 }, { "epoch": 21.422907488986784, "eval_loss": 1.3164186477661133, "eval_runtime": 132.2466, "eval_samples_per_second": 13.671, "eval_steps_per_second": 0.431, "step": 1500 }, { "epoch": 21.437004405286345, "grad_norm": 2.5506937503814697, "learning_rate": 0.00012753902600927896, "loss": 0.0602, "step": 1501 }, { "epoch": 21.451101321585902, "grad_norm": 3.752436399459839, "learning_rate": 0.00012745005649441432, "loss": 0.0564, "step": 1502 }, { "epoch": 21.465198237885463, "grad_norm": 2.432732105255127, "learning_rate": 0.00012736106347437955, "loss": 0.0439, "step": 1503 }, { "epoch": 21.47929515418502, "grad_norm": 2.0695109367370605, "learning_rate": 0.00012727204702537824, "loss": 0.0453, "step": 1504 }, { "epoch": 21.493392070484582, "grad_norm": 2.1871156692504883, "learning_rate": 0.0001271830072236343, "loss": 0.048, "step": 1505 }, { "epoch": 21.50748898678414, "grad_norm": 2.2698910236358643, "learning_rate": 0.0001270939441453914, "loss": 0.0435, "step": 1506 }, { "epoch": 21.5215859030837, "grad_norm": 2.3664047718048096, "learning_rate": 0.00012700485786691328, "loss": 0.0505, "step": 1507 }, { "epoch": 21.53568281938326, "grad_norm": 2.564295768737793, "learning_rate": 0.0001269157484644835, "loss": 0.0543, "step": 1508 }, { "epoch": 21.54977973568282, "grad_norm": 2.712817430496216, "learning_rate": 0.0001268266160144054, "loss": 0.0489, "step": 1509 }, { "epoch": 21.56387665198238, "grad_norm": 2.7658562660217285, "learning_rate": 0.00012673746059300208, "loss": 0.0587, "step": 1510 }, { "epoch": 21.577973568281937, "grad_norm": 3.4990315437316895, "learning_rate": 0.00012664828227661632, "loss": 0.052, "step": 1511 }, { "epoch": 21.592070484581498, "grad_norm": 2.571505308151245, "learning_rate": 0.0001265590811416105, "loss": 0.0553, "step": 1512 }, { "epoch": 21.606167400881056, "grad_norm": 2.530256509780884, "learning_rate": 0.00012646985726436655, "loss": 0.0582, "step": 1513 }, { "epoch": 21.620264317180617, "grad_norm": 2.467700481414795, "learning_rate": 0.00012638061072128587, "loss": 0.0466, "step": 1514 }, { "epoch": 21.634361233480178, "grad_norm": 2.641097068786621, "learning_rate": 0.00012629134158878918, "loss": 0.0539, "step": 1515 }, { "epoch": 21.648458149779735, "grad_norm": 2.1798627376556396, "learning_rate": 0.00012620204994331672, "loss": 0.0505, "step": 1516 }, { "epoch": 21.662555066079296, "grad_norm": 2.31363582611084, "learning_rate": 0.00012611273586132785, "loss": 0.0597, "step": 1517 }, { "epoch": 21.676651982378853, "grad_norm": 2.7151615619659424, "learning_rate": 0.00012602339941930116, "loss": 0.0585, "step": 1518 }, { "epoch": 21.690748898678414, "grad_norm": 2.542882204055786, "learning_rate": 0.0001259340406937345, "loss": 0.054, "step": 1519 }, { "epoch": 21.704845814977972, "grad_norm": 1.8885825872421265, "learning_rate": 0.0001258446597611447, "loss": 0.0479, "step": 1520 }, { "epoch": 21.718942731277533, "grad_norm": 2.4095349311828613, "learning_rate": 0.00012575525669806766, "loss": 0.0492, "step": 1521 }, { "epoch": 21.733039647577094, "grad_norm": 2.7167749404907227, "learning_rate": 0.00012566583158105812, "loss": 0.0551, "step": 1522 }, { "epoch": 21.74713656387665, "grad_norm": 2.290691614151001, "learning_rate": 0.00012557638448668987, "loss": 0.0468, "step": 1523 }, { "epoch": 21.761233480176212, "grad_norm": 2.5932741165161133, "learning_rate": 0.0001254869154915554, "loss": 0.0553, "step": 1524 }, { "epoch": 21.77533039647577, "grad_norm": 2.835549831390381, "learning_rate": 0.00012539742467226601, "loss": 0.0584, "step": 1525 }, { "epoch": 21.78942731277533, "grad_norm": 1.9100204706192017, "learning_rate": 0.00012530791210545162, "loss": 0.0461, "step": 1526 }, { "epoch": 21.80352422907489, "grad_norm": 2.127720594406128, "learning_rate": 0.0001252183778677609, "loss": 0.046, "step": 1527 }, { "epoch": 21.81762114537445, "grad_norm": 2.1143150329589844, "learning_rate": 0.00012512882203586092, "loss": 0.049, "step": 1528 }, { "epoch": 21.83171806167401, "grad_norm": 3.0230872631073, "learning_rate": 0.00012503924468643736, "loss": 0.0645, "step": 1529 }, { "epoch": 21.845814977973568, "grad_norm": 2.6014294624328613, "learning_rate": 0.00012494964589619423, "loss": 0.0631, "step": 1530 }, { "epoch": 21.85991189427313, "grad_norm": 2.2868106365203857, "learning_rate": 0.00012486002574185406, "loss": 0.0536, "step": 1531 }, { "epoch": 21.874008810572686, "grad_norm": 2.7254929542541504, "learning_rate": 0.00012477038430015747, "loss": 0.0623, "step": 1532 }, { "epoch": 21.888105726872247, "grad_norm": 2.4429802894592285, "learning_rate": 0.0001246807216478634, "loss": 0.0535, "step": 1533 }, { "epoch": 21.902202643171805, "grad_norm": 2.6781322956085205, "learning_rate": 0.00012459103786174907, "loss": 0.0572, "step": 1534 }, { "epoch": 21.916299559471366, "grad_norm": 2.195347785949707, "learning_rate": 0.00012450133301860952, "loss": 0.0599, "step": 1535 }, { "epoch": 21.930396475770927, "grad_norm": 2.177300453186035, "learning_rate": 0.0001244116071952581, "loss": 0.0581, "step": 1536 }, { "epoch": 21.944493392070484, "grad_norm": 2.4666707515716553, "learning_rate": 0.00012432186046852593, "loss": 0.0516, "step": 1537 }, { "epoch": 21.958590308370045, "grad_norm": 2.9001920223236084, "learning_rate": 0.0001242320929152622, "loss": 0.0592, "step": 1538 }, { "epoch": 21.972687224669603, "grad_norm": 2.1380136013031006, "learning_rate": 0.00012414230461233378, "loss": 0.0446, "step": 1539 }, { "epoch": 21.986784140969164, "grad_norm": 2.084136724472046, "learning_rate": 0.00012405249563662537, "loss": 0.0517, "step": 1540 }, { "epoch": 22.01409691629956, "grad_norm": 2.3445885181427, "learning_rate": 0.0001239626660650394, "loss": 0.0799, "step": 1541 }, { "epoch": 22.02819383259912, "grad_norm": 2.1887481212615967, "learning_rate": 0.00012387281597449587, "loss": 0.0389, "step": 1542 }, { "epoch": 22.04229074889868, "grad_norm": 1.8650740385055542, "learning_rate": 0.0001237829454419324, "loss": 0.0399, "step": 1543 }, { "epoch": 22.056387665198237, "grad_norm": 2.2115941047668457, "learning_rate": 0.00012369305454430413, "loss": 0.0362, "step": 1544 }, { "epoch": 22.070484581497798, "grad_norm": 2.064824342727661, "learning_rate": 0.0001236031433585836, "loss": 0.034, "step": 1545 }, { "epoch": 22.084581497797355, "grad_norm": 1.8030201196670532, "learning_rate": 0.00012351321196176068, "loss": 0.0353, "step": 1546 }, { "epoch": 22.098678414096916, "grad_norm": 1.7186152935028076, "learning_rate": 0.00012342326043084266, "loss": 0.0351, "step": 1547 }, { "epoch": 22.112775330396477, "grad_norm": 1.5780328512191772, "learning_rate": 0.00012333328884285403, "loss": 0.039, "step": 1548 }, { "epoch": 22.126872246696035, "grad_norm": 1.7256238460540771, "learning_rate": 0.00012324329727483635, "loss": 0.038, "step": 1549 }, { "epoch": 22.140969162995596, "grad_norm": 2.2245707511901855, "learning_rate": 0.00012315328580384842, "loss": 0.0318, "step": 1550 }, { "epoch": 22.155066079295153, "grad_norm": 1.8424384593963623, "learning_rate": 0.00012306325450696606, "loss": 0.0421, "step": 1551 }, { "epoch": 22.169162995594714, "grad_norm": 2.1893115043640137, "learning_rate": 0.00012297320346128196, "loss": 0.0424, "step": 1552 }, { "epoch": 22.18325991189427, "grad_norm": 2.1859424114227295, "learning_rate": 0.0001228831327439059, "loss": 0.0376, "step": 1553 }, { "epoch": 22.197356828193833, "grad_norm": 2.0701942443847656, "learning_rate": 0.00012279304243196436, "loss": 0.0443, "step": 1554 }, { "epoch": 22.211453744493394, "grad_norm": 1.8999332189559937, "learning_rate": 0.00012270293260260067, "loss": 0.044, "step": 1555 }, { "epoch": 22.22555066079295, "grad_norm": 1.6519430875778198, "learning_rate": 0.0001226128033329748, "loss": 0.0409, "step": 1556 }, { "epoch": 22.239647577092512, "grad_norm": 1.7961369752883911, "learning_rate": 0.00012252265470026345, "loss": 0.0359, "step": 1557 }, { "epoch": 22.25374449339207, "grad_norm": 2.0399866104125977, "learning_rate": 0.00012243248678165983, "loss": 0.0347, "step": 1558 }, { "epoch": 22.26784140969163, "grad_norm": 1.9464128017425537, "learning_rate": 0.0001223422996543738, "loss": 0.0419, "step": 1559 }, { "epoch": 22.281938325991188, "grad_norm": 2.417759895324707, "learning_rate": 0.00012225209339563145, "loss": 0.0386, "step": 1560 }, { "epoch": 22.29603524229075, "grad_norm": 1.5946882963180542, "learning_rate": 0.00012216186808267546, "loss": 0.0384, "step": 1561 }, { "epoch": 22.31013215859031, "grad_norm": 1.6346818208694458, "learning_rate": 0.00012207162379276464, "loss": 0.0348, "step": 1562 }, { "epoch": 22.324229074889868, "grad_norm": 2.093527317047119, "learning_rate": 0.00012198136060317424, "loss": 0.0342, "step": 1563 }, { "epoch": 22.33832599118943, "grad_norm": 2.54099178314209, "learning_rate": 0.00012189107859119555, "loss": 0.0392, "step": 1564 }, { "epoch": 22.352422907488986, "grad_norm": 2.2723031044006348, "learning_rate": 0.00012180077783413601, "loss": 0.0405, "step": 1565 }, { "epoch": 22.366519823788547, "grad_norm": 2.1723499298095703, "learning_rate": 0.00012171045840931921, "loss": 0.0368, "step": 1566 }, { "epoch": 22.380616740088104, "grad_norm": 2.1933364868164062, "learning_rate": 0.00012162012039408454, "loss": 0.038, "step": 1567 }, { "epoch": 22.394713656387665, "grad_norm": 2.1852850914001465, "learning_rate": 0.0001215297638657875, "loss": 0.0418, "step": 1568 }, { "epoch": 22.408810572687223, "grad_norm": 1.9718141555786133, "learning_rate": 0.00012143938890179929, "loss": 0.0374, "step": 1569 }, { "epoch": 22.422907488986784, "grad_norm": 1.9354143142700195, "learning_rate": 0.00012134899557950698, "loss": 0.0419, "step": 1570 }, { "epoch": 22.437004405286345, "grad_norm": 1.9772416353225708, "learning_rate": 0.00012125858397631336, "loss": 0.0434, "step": 1571 }, { "epoch": 22.451101321585902, "grad_norm": 2.2714321613311768, "learning_rate": 0.00012116815416963684, "loss": 0.0406, "step": 1572 }, { "epoch": 22.465198237885463, "grad_norm": 1.9988983869552612, "learning_rate": 0.00012107770623691142, "loss": 0.0383, "step": 1573 }, { "epoch": 22.47929515418502, "grad_norm": 1.9692227840423584, "learning_rate": 0.00012098724025558665, "loss": 0.0358, "step": 1574 }, { "epoch": 22.493392070484582, "grad_norm": 1.934794545173645, "learning_rate": 0.00012089675630312754, "loss": 0.0399, "step": 1575 }, { "epoch": 22.50748898678414, "grad_norm": 1.7173333168029785, "learning_rate": 0.00012080625445701439, "loss": 0.0386, "step": 1576 }, { "epoch": 22.5215859030837, "grad_norm": 2.1280455589294434, "learning_rate": 0.00012071573479474297, "loss": 0.0348, "step": 1577 }, { "epoch": 22.53568281938326, "grad_norm": 2.2500810623168945, "learning_rate": 0.00012062519739382419, "loss": 0.0507, "step": 1578 }, { "epoch": 22.54977973568282, "grad_norm": 2.257249116897583, "learning_rate": 0.00012053464233178422, "loss": 0.045, "step": 1579 }, { "epoch": 22.56387665198238, "grad_norm": 2.430025339126587, "learning_rate": 0.00012044406968616432, "loss": 0.0375, "step": 1580 }, { "epoch": 22.577973568281937, "grad_norm": 1.6770100593566895, "learning_rate": 0.00012035347953452083, "loss": 0.0363, "step": 1581 }, { "epoch": 22.592070484581498, "grad_norm": 1.9835031032562256, "learning_rate": 0.00012026287195442503, "loss": 0.0424, "step": 1582 }, { "epoch": 22.606167400881056, "grad_norm": 2.3108484745025635, "learning_rate": 0.00012017224702346317, "loss": 0.0369, "step": 1583 }, { "epoch": 22.620264317180617, "grad_norm": 1.8941009044647217, "learning_rate": 0.00012008160481923637, "loss": 0.0463, "step": 1584 }, { "epoch": 22.634361233480178, "grad_norm": 2.192331314086914, "learning_rate": 0.00011999094541936047, "loss": 0.0424, "step": 1585 }, { "epoch": 22.648458149779735, "grad_norm": 2.423029899597168, "learning_rate": 0.00011990026890146609, "loss": 0.0457, "step": 1586 }, { "epoch": 22.662555066079296, "grad_norm": 1.6149996519088745, "learning_rate": 0.00011980957534319852, "loss": 0.0377, "step": 1587 }, { "epoch": 22.676651982378853, "grad_norm": 1.7003978490829468, "learning_rate": 0.00011971886482221762, "loss": 0.0335, "step": 1588 }, { "epoch": 22.690748898678414, "grad_norm": 1.6965943574905396, "learning_rate": 0.00011962813741619777, "loss": 0.0394, "step": 1589 }, { "epoch": 22.704845814977972, "grad_norm": 2.634645462036133, "learning_rate": 0.00011953739320282778, "loss": 0.0421, "step": 1590 }, { "epoch": 22.718942731277533, "grad_norm": 2.2505340576171875, "learning_rate": 0.00011944663225981092, "loss": 0.0438, "step": 1591 }, { "epoch": 22.733039647577094, "grad_norm": 2.3389534950256348, "learning_rate": 0.00011935585466486476, "loss": 0.0418, "step": 1592 }, { "epoch": 22.74713656387665, "grad_norm": 2.019679069519043, "learning_rate": 0.00011926506049572105, "loss": 0.0373, "step": 1593 }, { "epoch": 22.761233480176212, "grad_norm": 1.8082716464996338, "learning_rate": 0.0001191742498301259, "loss": 0.0336, "step": 1594 }, { "epoch": 22.77533039647577, "grad_norm": 1.9082636833190918, "learning_rate": 0.00011908342274583936, "loss": 0.0403, "step": 1595 }, { "epoch": 22.78942731277533, "grad_norm": 1.878173828125, "learning_rate": 0.0001189925793206357, "loss": 0.0388, "step": 1596 }, { "epoch": 22.80352422907489, "grad_norm": 2.6715784072875977, "learning_rate": 0.00011890171963230305, "loss": 0.0439, "step": 1597 }, { "epoch": 22.81762114537445, "grad_norm": 2.5559370517730713, "learning_rate": 0.00011881084375864356, "loss": 0.0386, "step": 1598 }, { "epoch": 22.83171806167401, "grad_norm": 2.592202663421631, "learning_rate": 0.00011871995177747315, "loss": 0.0395, "step": 1599 }, { "epoch": 22.845814977973568, "grad_norm": 2.358149528503418, "learning_rate": 0.00011862904376662167, "loss": 0.041, "step": 1600 }, { "epoch": 22.85991189427313, "grad_norm": 2.278386354446411, "learning_rate": 0.00011853811980393254, "loss": 0.0452, "step": 1601 }, { "epoch": 22.874008810572686, "grad_norm": 2.2522060871124268, "learning_rate": 0.00011844717996726296, "loss": 0.0391, "step": 1602 }, { "epoch": 22.888105726872247, "grad_norm": 1.938319206237793, "learning_rate": 0.00011835622433448361, "loss": 0.0464, "step": 1603 }, { "epoch": 22.902202643171805, "grad_norm": 2.5560243129730225, "learning_rate": 0.00011826525298347886, "loss": 0.0395, "step": 1604 }, { "epoch": 22.916299559471366, "grad_norm": 2.5375092029571533, "learning_rate": 0.00011817426599214636, "loss": 0.0444, "step": 1605 }, { "epoch": 22.930396475770927, "grad_norm": 1.974470853805542, "learning_rate": 0.00011808326343839726, "loss": 0.0416, "step": 1606 }, { "epoch": 22.944493392070484, "grad_norm": 2.2278354167938232, "learning_rate": 0.00011799224540015598, "loss": 0.0388, "step": 1607 }, { "epoch": 22.958590308370045, "grad_norm": 2.1463491916656494, "learning_rate": 0.00011790121195536031, "loss": 0.0404, "step": 1608 }, { "epoch": 22.972687224669603, "grad_norm": 1.9302736520767212, "learning_rate": 0.00011781016318196108, "loss": 0.0386, "step": 1609 }, { "epoch": 22.986784140969164, "grad_norm": 2.09002423286438, "learning_rate": 0.0001177190991579223, "loss": 0.0431, "step": 1610 }, { "epoch": 23.01409691629956, "grad_norm": 2.2779126167297363, "learning_rate": 0.0001176280199612211, "loss": 0.0664, "step": 1611 }, { "epoch": 23.02819383259912, "grad_norm": 1.6150352954864502, "learning_rate": 0.00011753692566984753, "loss": 0.0256, "step": 1612 }, { "epoch": 23.04229074889868, "grad_norm": 1.4255620241165161, "learning_rate": 0.00011744581636180461, "loss": 0.0266, "step": 1613 }, { "epoch": 23.056387665198237, "grad_norm": 1.4316341876983643, "learning_rate": 0.00011735469211510819, "loss": 0.0294, "step": 1614 }, { "epoch": 23.070484581497798, "grad_norm": 1.5519899129867554, "learning_rate": 0.00011726355300778693, "loss": 0.029, "step": 1615 }, { "epoch": 23.084581497797355, "grad_norm": 1.476962685585022, "learning_rate": 0.0001171723991178822, "loss": 0.0253, "step": 1616 }, { "epoch": 23.098678414096916, "grad_norm": 1.3584754467010498, "learning_rate": 0.00011708123052344804, "loss": 0.0271, "step": 1617 }, { "epoch": 23.112775330396477, "grad_norm": 1.9940953254699707, "learning_rate": 0.0001169900473025511, "loss": 0.0308, "step": 1618 }, { "epoch": 23.126872246696035, "grad_norm": 1.2870641946792603, "learning_rate": 0.00011689884953327048, "loss": 0.0299, "step": 1619 }, { "epoch": 23.140969162995596, "grad_norm": 1.695754051208496, "learning_rate": 0.00011680763729369783, "loss": 0.0297, "step": 1620 }, { "epoch": 23.155066079295153, "grad_norm": 1.9906692504882812, "learning_rate": 0.00011671641066193713, "loss": 0.0302, "step": 1621 }, { "epoch": 23.169162995594714, "grad_norm": 1.7047563791275024, "learning_rate": 0.00011662516971610474, "loss": 0.0315, "step": 1622 }, { "epoch": 23.18325991189427, "grad_norm": 1.5961353778839111, "learning_rate": 0.00011653391453432918, "loss": 0.0295, "step": 1623 }, { "epoch": 23.197356828193833, "grad_norm": 1.6458202600479126, "learning_rate": 0.0001164426451947513, "loss": 0.031, "step": 1624 }, { "epoch": 23.211453744493394, "grad_norm": 1.4621548652648926, "learning_rate": 0.00011635136177552391, "loss": 0.028, "step": 1625 }, { "epoch": 23.22555066079295, "grad_norm": 1.8135064840316772, "learning_rate": 0.00011626006435481201, "loss": 0.0321, "step": 1626 }, { "epoch": 23.239647577092512, "grad_norm": 1.7533410787582397, "learning_rate": 0.0001161687530107925, "loss": 0.0299, "step": 1627 }, { "epoch": 23.25374449339207, "grad_norm": 2.0491995811462402, "learning_rate": 0.00011607742782165432, "loss": 0.0254, "step": 1628 }, { "epoch": 23.26784140969163, "grad_norm": 1.8689510822296143, "learning_rate": 0.00011598608886559809, "loss": 0.0321, "step": 1629 }, { "epoch": 23.281938325991188, "grad_norm": 1.6571661233901978, "learning_rate": 0.00011589473622083642, "loss": 0.0327, "step": 1630 }, { "epoch": 23.29603524229075, "grad_norm": 1.7101166248321533, "learning_rate": 0.00011580336996559343, "loss": 0.0246, "step": 1631 }, { "epoch": 23.31013215859031, "grad_norm": 1.2370638847351074, "learning_rate": 0.00011571199017810508, "loss": 0.0247, "step": 1632 }, { "epoch": 23.324229074889868, "grad_norm": 1.6692310571670532, "learning_rate": 0.0001156205969366188, "loss": 0.0301, "step": 1633 }, { "epoch": 23.33832599118943, "grad_norm": 1.594124674797058, "learning_rate": 0.00011552919031939362, "loss": 0.0268, "step": 1634 }, { "epoch": 23.352422907488986, "grad_norm": 1.3996381759643555, "learning_rate": 0.00011543777040469994, "loss": 0.0261, "step": 1635 }, { "epoch": 23.366519823788547, "grad_norm": 1.911834955215454, "learning_rate": 0.00011534633727081958, "loss": 0.0318, "step": 1636 }, { "epoch": 23.380616740088104, "grad_norm": 1.447340965270996, "learning_rate": 0.00011525489099604576, "loss": 0.0298, "step": 1637 }, { "epoch": 23.394713656387665, "grad_norm": 2.2587225437164307, "learning_rate": 0.00011516343165868279, "loss": 0.0308, "step": 1638 }, { "epoch": 23.408810572687223, "grad_norm": 2.0169622898101807, "learning_rate": 0.00011507195933704635, "loss": 0.0287, "step": 1639 }, { "epoch": 23.422907488986784, "grad_norm": 1.4195468425750732, "learning_rate": 0.00011498047410946306, "loss": 0.0278, "step": 1640 }, { "epoch": 23.437004405286345, "grad_norm": 1.444063425064087, "learning_rate": 0.00011488897605427073, "loss": 0.0314, "step": 1641 }, { "epoch": 23.451101321585902, "grad_norm": 1.4500327110290527, "learning_rate": 0.00011479746524981809, "loss": 0.0329, "step": 1642 }, { "epoch": 23.465198237885463, "grad_norm": 1.8200743198394775, "learning_rate": 0.00011470594177446479, "loss": 0.031, "step": 1643 }, { "epoch": 23.47929515418502, "grad_norm": 1.2579234838485718, "learning_rate": 0.00011461440570658132, "loss": 0.0293, "step": 1644 }, { "epoch": 23.493392070484582, "grad_norm": 2.077089309692383, "learning_rate": 0.00011452285712454904, "loss": 0.037, "step": 1645 }, { "epoch": 23.50748898678414, "grad_norm": 2.1869184970855713, "learning_rate": 0.00011443129610675987, "loss": 0.0284, "step": 1646 }, { "epoch": 23.5215859030837, "grad_norm": 1.866665005683899, "learning_rate": 0.00011433972273161653, "loss": 0.0294, "step": 1647 }, { "epoch": 23.53568281938326, "grad_norm": 1.3028141260147095, "learning_rate": 0.00011424813707753221, "loss": 0.0297, "step": 1648 }, { "epoch": 23.54977973568282, "grad_norm": 1.4321597814559937, "learning_rate": 0.00011415653922293071, "loss": 0.0283, "step": 1649 }, { "epoch": 23.56387665198238, "grad_norm": 1.4410364627838135, "learning_rate": 0.00011406492924624614, "loss": 0.0262, "step": 1650 }, { "epoch": 23.577973568281937, "grad_norm": 1.4666935205459595, "learning_rate": 0.00011397330722592318, "loss": 0.0293, "step": 1651 }, { "epoch": 23.592070484581498, "grad_norm": 1.8705646991729736, "learning_rate": 0.00011388167324041669, "loss": 0.0263, "step": 1652 }, { "epoch": 23.606167400881056, "grad_norm": 1.4852664470672607, "learning_rate": 0.00011379002736819176, "loss": 0.0335, "step": 1653 }, { "epoch": 23.620264317180617, "grad_norm": 1.8836369514465332, "learning_rate": 0.00011369836968772376, "loss": 0.0292, "step": 1654 }, { "epoch": 23.634361233480178, "grad_norm": 1.9474687576293945, "learning_rate": 0.00011360670027749807, "loss": 0.0369, "step": 1655 }, { "epoch": 23.648458149779735, "grad_norm": 2.2492499351501465, "learning_rate": 0.0001135150192160102, "loss": 0.0261, "step": 1656 }, { "epoch": 23.662555066079296, "grad_norm": 1.6066397428512573, "learning_rate": 0.00011342332658176555, "loss": 0.0308, "step": 1657 }, { "epoch": 23.676651982378853, "grad_norm": 1.8964306116104126, "learning_rate": 0.00011333162245327956, "loss": 0.0325, "step": 1658 }, { "epoch": 23.690748898678414, "grad_norm": 1.711844801902771, "learning_rate": 0.00011323990690907733, "loss": 0.0278, "step": 1659 }, { "epoch": 23.704845814977972, "grad_norm": 1.4864206314086914, "learning_rate": 0.00011314818002769389, "loss": 0.0292, "step": 1660 }, { "epoch": 23.718942731277533, "grad_norm": 1.726086139678955, "learning_rate": 0.00011305644188767392, "loss": 0.0345, "step": 1661 }, { "epoch": 23.733039647577094, "grad_norm": 2.0668962001800537, "learning_rate": 0.00011296469256757174, "loss": 0.0326, "step": 1662 }, { "epoch": 23.74713656387665, "grad_norm": 2.111243486404419, "learning_rate": 0.00011287293214595123, "loss": 0.0294, "step": 1663 }, { "epoch": 23.761233480176212, "grad_norm": 1.6953628063201904, "learning_rate": 0.0001127811607013858, "loss": 0.0393, "step": 1664 }, { "epoch": 23.77533039647577, "grad_norm": 2.4963598251342773, "learning_rate": 0.0001126893783124583, "loss": 0.0417, "step": 1665 }, { "epoch": 23.78942731277533, "grad_norm": 2.0012173652648926, "learning_rate": 0.00011259758505776092, "loss": 0.037, "step": 1666 }, { "epoch": 23.80352422907489, "grad_norm": 2.34435772895813, "learning_rate": 0.00011250578101589522, "loss": 0.0396, "step": 1667 }, { "epoch": 23.81762114537445, "grad_norm": 2.079598903656006, "learning_rate": 0.00011241396626547189, "loss": 0.0326, "step": 1668 }, { "epoch": 23.83171806167401, "grad_norm": 1.5442795753479004, "learning_rate": 0.00011232214088511092, "loss": 0.0334, "step": 1669 }, { "epoch": 23.845814977973568, "grad_norm": 1.7094478607177734, "learning_rate": 0.00011223030495344127, "loss": 0.0322, "step": 1670 }, { "epoch": 23.85991189427313, "grad_norm": 1.8473354578018188, "learning_rate": 0.00011213845854910108, "loss": 0.037, "step": 1671 }, { "epoch": 23.874008810572686, "grad_norm": 1.922701358795166, "learning_rate": 0.0001120466017507373, "loss": 0.0376, "step": 1672 }, { "epoch": 23.888105726872247, "grad_norm": 1.7453746795654297, "learning_rate": 0.0001119547346370059, "loss": 0.0311, "step": 1673 }, { "epoch": 23.902202643171805, "grad_norm": 1.7518891096115112, "learning_rate": 0.00011186285728657163, "loss": 0.0377, "step": 1674 }, { "epoch": 23.916299559471366, "grad_norm": 1.9885293245315552, "learning_rate": 0.00011177096977810803, "loss": 0.035, "step": 1675 }, { "epoch": 23.930396475770927, "grad_norm": 1.7983494997024536, "learning_rate": 0.00011167907219029729, "loss": 0.0337, "step": 1676 }, { "epoch": 23.944493392070484, "grad_norm": 1.797046184539795, "learning_rate": 0.00011158716460183032, "loss": 0.0336, "step": 1677 }, { "epoch": 23.958590308370045, "grad_norm": 1.6970208883285522, "learning_rate": 0.0001114952470914065, "loss": 0.0277, "step": 1678 }, { "epoch": 23.972687224669603, "grad_norm": 1.6562148332595825, "learning_rate": 0.00011140331973773377, "loss": 0.0327, "step": 1679 }, { "epoch": 23.986784140969164, "grad_norm": 1.8289382457733154, "learning_rate": 0.00011131138261952845, "loss": 0.0373, "step": 1680 }, { "epoch": 24.01409691629956, "grad_norm": 1.8561886548995972, "learning_rate": 0.00011121943581551523, "loss": 0.05, "step": 1681 }, { "epoch": 24.02819383259912, "grad_norm": 1.4400639533996582, "learning_rate": 0.00011112747940442716, "loss": 0.0257, "step": 1682 }, { "epoch": 24.04229074889868, "grad_norm": 1.4834771156311035, "learning_rate": 0.00011103551346500543, "loss": 0.0281, "step": 1683 }, { "epoch": 24.056387665198237, "grad_norm": 1.9006198644638062, "learning_rate": 0.00011094353807599944, "loss": 0.0263, "step": 1684 }, { "epoch": 24.070484581497798, "grad_norm": 1.3086055517196655, "learning_rate": 0.00011085155331616663, "loss": 0.0223, "step": 1685 }, { "epoch": 24.084581497797355, "grad_norm": 1.4496370553970337, "learning_rate": 0.00011075955926427258, "loss": 0.0213, "step": 1686 }, { "epoch": 24.098678414096916, "grad_norm": 1.2559378147125244, "learning_rate": 0.00011066755599909064, "loss": 0.0219, "step": 1687 }, { "epoch": 24.112775330396477, "grad_norm": 1.1244754791259766, "learning_rate": 0.00011057554359940227, "loss": 0.0244, "step": 1688 }, { "epoch": 24.126872246696035, "grad_norm": 1.6085612773895264, "learning_rate": 0.00011048352214399656, "loss": 0.0221, "step": 1689 }, { "epoch": 24.140969162995596, "grad_norm": 1.708907961845398, "learning_rate": 0.00011039149171167045, "loss": 0.0255, "step": 1690 }, { "epoch": 24.155066079295153, "grad_norm": 1.3958662748336792, "learning_rate": 0.00011029945238122854, "loss": 0.0247, "step": 1691 }, { "epoch": 24.169162995594714, "grad_norm": 1.5633028745651245, "learning_rate": 0.00011020740423148311, "loss": 0.0229, "step": 1692 }, { "epoch": 24.18325991189427, "grad_norm": 1.0774860382080078, "learning_rate": 0.00011011534734125389, "loss": 0.0215, "step": 1693 }, { "epoch": 24.197356828193833, "grad_norm": 1.7391856908798218, "learning_rate": 0.00011002328178936811, "loss": 0.0255, "step": 1694 }, { "epoch": 24.211453744493394, "grad_norm": 1.2506052255630493, "learning_rate": 0.00010993120765466056, "loss": 0.0235, "step": 1695 }, { "epoch": 24.22555066079295, "grad_norm": 1.1021692752838135, "learning_rate": 0.00010983912501597315, "loss": 0.0209, "step": 1696 }, { "epoch": 24.239647577092512, "grad_norm": 1.6219544410705566, "learning_rate": 0.00010974703395215531, "loss": 0.0303, "step": 1697 }, { "epoch": 24.25374449339207, "grad_norm": 1.8847260475158691, "learning_rate": 0.00010965493454206348, "loss": 0.022, "step": 1698 }, { "epoch": 24.26784140969163, "grad_norm": 1.593544363975525, "learning_rate": 0.00010956282686456136, "loss": 0.0246, "step": 1699 }, { "epoch": 24.281938325991188, "grad_norm": 1.275105595588684, "learning_rate": 0.00010947071099851971, "loss": 0.0228, "step": 1700 }, { "epoch": 24.29603524229075, "grad_norm": 1.0795419216156006, "learning_rate": 0.00010937858702281631, "loss": 0.0274, "step": 1701 }, { "epoch": 24.31013215859031, "grad_norm": 1.6396112442016602, "learning_rate": 0.00010928645501633583, "loss": 0.0267, "step": 1702 }, { "epoch": 24.324229074889868, "grad_norm": 1.1881648302078247, "learning_rate": 0.00010919431505796989, "loss": 0.0227, "step": 1703 }, { "epoch": 24.33832599118943, "grad_norm": 0.8596903085708618, "learning_rate": 0.00010910216722661686, "loss": 0.0236, "step": 1704 }, { "epoch": 24.352422907488986, "grad_norm": 1.56818425655365, "learning_rate": 0.00010901001160118189, "loss": 0.0256, "step": 1705 }, { "epoch": 24.366519823788547, "grad_norm": 1.3395980596542358, "learning_rate": 0.00010891784826057678, "loss": 0.0239, "step": 1706 }, { "epoch": 24.380616740088104, "grad_norm": 2.6568093299865723, "learning_rate": 0.00010882567728372, "loss": 0.0274, "step": 1707 }, { "epoch": 24.394713656387665, "grad_norm": 1.571498990058899, "learning_rate": 0.0001087334987495364, "loss": 0.0262, "step": 1708 }, { "epoch": 24.408810572687223, "grad_norm": 1.2011593580245972, "learning_rate": 0.00010864131273695752, "loss": 0.0243, "step": 1709 }, { "epoch": 24.422907488986784, "grad_norm": 1.3287863731384277, "learning_rate": 0.00010854911932492114, "loss": 0.0226, "step": 1710 }, { "epoch": 24.437004405286345, "grad_norm": 1.3242104053497314, "learning_rate": 0.00010845691859237139, "loss": 0.0268, "step": 1711 }, { "epoch": 24.451101321585902, "grad_norm": 1.5761470794677734, "learning_rate": 0.00010836471061825877, "loss": 0.0245, "step": 1712 }, { "epoch": 24.465198237885463, "grad_norm": 1.8471205234527588, "learning_rate": 0.00010827249548153985, "loss": 0.0205, "step": 1713 }, { "epoch": 24.47929515418502, "grad_norm": 0.9364733695983887, "learning_rate": 0.00010818027326117748, "loss": 0.0217, "step": 1714 }, { "epoch": 24.493392070484582, "grad_norm": 1.9574388265609741, "learning_rate": 0.00010808804403614043, "loss": 0.0263, "step": 1715 }, { "epoch": 24.50748898678414, "grad_norm": 1.2778065204620361, "learning_rate": 0.00010799580788540358, "loss": 0.023, "step": 1716 }, { "epoch": 24.5215859030837, "grad_norm": 1.7862614393234253, "learning_rate": 0.00010790356488794764, "loss": 0.0306, "step": 1717 }, { "epoch": 24.53568281938326, "grad_norm": 1.3274680376052856, "learning_rate": 0.00010781131512275928, "loss": 0.0263, "step": 1718 }, { "epoch": 24.54977973568282, "grad_norm": 1.3887394666671753, "learning_rate": 0.00010771905866883089, "loss": 0.0251, "step": 1719 }, { "epoch": 24.56387665198238, "grad_norm": 1.6450140476226807, "learning_rate": 0.00010762679560516067, "loss": 0.0281, "step": 1720 }, { "epoch": 24.577973568281937, "grad_norm": 1.3083906173706055, "learning_rate": 0.00010753452601075236, "loss": 0.0294, "step": 1721 }, { "epoch": 24.592070484581498, "grad_norm": 1.5771195888519287, "learning_rate": 0.0001074422499646154, "loss": 0.0257, "step": 1722 }, { "epoch": 24.606167400881056, "grad_norm": 1.9086509943008423, "learning_rate": 0.00010734996754576472, "loss": 0.0273, "step": 1723 }, { "epoch": 24.620264317180617, "grad_norm": 1.4535863399505615, "learning_rate": 0.00010725767883322067, "loss": 0.0236, "step": 1724 }, { "epoch": 24.634361233480178, "grad_norm": 1.5410856008529663, "learning_rate": 0.00010716538390600908, "loss": 0.0288, "step": 1725 }, { "epoch": 24.648458149779735, "grad_norm": 1.5222675800323486, "learning_rate": 0.00010707308284316097, "loss": 0.0275, "step": 1726 }, { "epoch": 24.662555066079296, "grad_norm": 1.776249885559082, "learning_rate": 0.00010698077572371277, "loss": 0.0209, "step": 1727 }, { "epoch": 24.676651982378853, "grad_norm": 1.2653194665908813, "learning_rate": 0.00010688846262670593, "loss": 0.032, "step": 1728 }, { "epoch": 24.690748898678414, "grad_norm": 1.3469547033309937, "learning_rate": 0.00010679614363118717, "loss": 0.0272, "step": 1729 }, { "epoch": 24.704845814977972, "grad_norm": 1.243300437927246, "learning_rate": 0.00010670381881620814, "loss": 0.0241, "step": 1730 }, { "epoch": 24.718942731277533, "grad_norm": 1.2493889331817627, "learning_rate": 0.00010661148826082557, "loss": 0.029, "step": 1731 }, { "epoch": 24.733039647577094, "grad_norm": 1.9413652420043945, "learning_rate": 0.00010651915204410101, "loss": 0.0283, "step": 1732 }, { "epoch": 24.74713656387665, "grad_norm": 1.6140892505645752, "learning_rate": 0.00010642681024510095, "loss": 0.0262, "step": 1733 }, { "epoch": 24.761233480176212, "grad_norm": 1.142042636871338, "learning_rate": 0.0001063344629428966, "loss": 0.0265, "step": 1734 }, { "epoch": 24.77533039647577, "grad_norm": 1.4035106897354126, "learning_rate": 0.00010624211021656392, "loss": 0.0255, "step": 1735 }, { "epoch": 24.78942731277533, "grad_norm": 1.8479640483856201, "learning_rate": 0.0001061497521451835, "loss": 0.0263, "step": 1736 }, { "epoch": 24.80352422907489, "grad_norm": 1.5739433765411377, "learning_rate": 0.00010605738880784045, "loss": 0.0328, "step": 1737 }, { "epoch": 24.81762114537445, "grad_norm": 1.5439226627349854, "learning_rate": 0.00010596502028362449, "loss": 0.0305, "step": 1738 }, { "epoch": 24.83171806167401, "grad_norm": 1.5010600090026855, "learning_rate": 0.00010587264665162972, "loss": 0.0242, "step": 1739 }, { "epoch": 24.845814977973568, "grad_norm": 2.217738389968872, "learning_rate": 0.00010578026799095464, "loss": 0.0328, "step": 1740 }, { "epoch": 24.85991189427313, "grad_norm": 2.014444351196289, "learning_rate": 0.000105687884380702, "loss": 0.0315, "step": 1741 }, { "epoch": 24.874008810572686, "grad_norm": 2.057560682296753, "learning_rate": 0.0001055954958999789, "loss": 0.0218, "step": 1742 }, { "epoch": 24.888105726872247, "grad_norm": 1.4610050916671753, "learning_rate": 0.00010550310262789649, "loss": 0.0306, "step": 1743 }, { "epoch": 24.902202643171805, "grad_norm": 1.370073676109314, "learning_rate": 0.0001054107046435701, "loss": 0.0264, "step": 1744 }, { "epoch": 24.916299559471366, "grad_norm": 1.849617838859558, "learning_rate": 0.00010531830202611904, "loss": 0.027, "step": 1745 }, { "epoch": 24.930396475770927, "grad_norm": 1.3307219743728638, "learning_rate": 0.00010522589485466665, "loss": 0.0242, "step": 1746 }, { "epoch": 24.944493392070484, "grad_norm": 1.3296968936920166, "learning_rate": 0.0001051334832083401, "loss": 0.0238, "step": 1747 }, { "epoch": 24.958590308370045, "grad_norm": 1.213840365409851, "learning_rate": 0.0001050410671662705, "loss": 0.0245, "step": 1748 }, { "epoch": 24.972687224669603, "grad_norm": 1.7029780149459839, "learning_rate": 0.00010494864680759255, "loss": 0.0281, "step": 1749 }, { "epoch": 24.986784140969164, "grad_norm": 1.6823182106018066, "learning_rate": 0.00010485622221144484, "loss": 0.0285, "step": 1750 }, { "epoch": 24.986784140969164, "eval_loss": 1.3581749200820923, "eval_runtime": 132.8831, "eval_samples_per_second": 13.606, "eval_steps_per_second": 0.429, "step": 1750 }, { "epoch": 25.01409691629956, "grad_norm": 1.4125819206237793, "learning_rate": 0.00010476379345696943, "loss": 0.0476, "step": 1751 }, { "epoch": 25.02819383259912, "grad_norm": 1.456520915031433, "learning_rate": 0.00010467136062331207, "loss": 0.0208, "step": 1752 }, { "epoch": 25.04229074889868, "grad_norm": 1.2040554285049438, "learning_rate": 0.00010457892378962188, "loss": 0.018, "step": 1753 }, { "epoch": 25.056387665198237, "grad_norm": 1.2145628929138184, "learning_rate": 0.00010448648303505151, "loss": 0.0182, "step": 1754 }, { "epoch": 25.070484581497798, "grad_norm": 1.1142774820327759, "learning_rate": 0.0001043940384387569, "loss": 0.0206, "step": 1755 }, { "epoch": 25.084581497797355, "grad_norm": 1.7596776485443115, "learning_rate": 0.0001043015900798973, "loss": 0.0219, "step": 1756 }, { "epoch": 25.098678414096916, "grad_norm": 1.41169273853302, "learning_rate": 0.00010420913803763521, "loss": 0.0221, "step": 1757 }, { "epoch": 25.112775330396477, "grad_norm": 1.4117463827133179, "learning_rate": 0.00010411668239113624, "loss": 0.0239, "step": 1758 }, { "epoch": 25.126872246696035, "grad_norm": 1.0407077074050903, "learning_rate": 0.00010402422321956912, "loss": 0.0196, "step": 1759 }, { "epoch": 25.140969162995596, "grad_norm": 1.5216808319091797, "learning_rate": 0.00010393176060210557, "loss": 0.0218, "step": 1760 }, { "epoch": 25.155066079295153, "grad_norm": 1.7054294347763062, "learning_rate": 0.00010383929461792026, "loss": 0.0246, "step": 1761 }, { "epoch": 25.169162995594714, "grad_norm": 2.0719153881073, "learning_rate": 0.00010374682534619078, "loss": 0.0211, "step": 1762 }, { "epoch": 25.18325991189427, "grad_norm": 1.2967299222946167, "learning_rate": 0.00010365435286609753, "loss": 0.0164, "step": 1763 }, { "epoch": 25.197356828193833, "grad_norm": 0.827609121799469, "learning_rate": 0.00010356187725682359, "loss": 0.0187, "step": 1764 }, { "epoch": 25.211453744493394, "grad_norm": 1.101556658744812, "learning_rate": 0.0001034693985975548, "loss": 0.0213, "step": 1765 }, { "epoch": 25.22555066079295, "grad_norm": 1.4534374475479126, "learning_rate": 0.00010337691696747958, "loss": 0.0201, "step": 1766 }, { "epoch": 25.239647577092512, "grad_norm": 1.2048122882843018, "learning_rate": 0.00010328443244578889, "loss": 0.0176, "step": 1767 }, { "epoch": 25.25374449339207, "grad_norm": 1.064869999885559, "learning_rate": 0.00010319194511167616, "loss": 0.0184, "step": 1768 }, { "epoch": 25.26784140969163, "grad_norm": 1.185474157333374, "learning_rate": 0.00010309945504433729, "loss": 0.0216, "step": 1769 }, { "epoch": 25.281938325991188, "grad_norm": 1.5568983554840088, "learning_rate": 0.0001030069623229704, "loss": 0.0187, "step": 1770 }, { "epoch": 25.29603524229075, "grad_norm": 1.0359594821929932, "learning_rate": 0.00010291446702677599, "loss": 0.0208, "step": 1771 }, { "epoch": 25.31013215859031, "grad_norm": 1.2168452739715576, "learning_rate": 0.00010282196923495672, "loss": 0.0218, "step": 1772 }, { "epoch": 25.324229074889868, "grad_norm": 1.192610263824463, "learning_rate": 0.00010272946902671738, "loss": 0.0188, "step": 1773 }, { "epoch": 25.33832599118943, "grad_norm": 2.0973098278045654, "learning_rate": 0.00010263696648126486, "loss": 0.025, "step": 1774 }, { "epoch": 25.352422907488986, "grad_norm": 0.8135376572608948, "learning_rate": 0.00010254446167780803, "loss": 0.0159, "step": 1775 }, { "epoch": 25.366519823788547, "grad_norm": 1.4995356798171997, "learning_rate": 0.00010245195469555769, "loss": 0.0341, "step": 1776 }, { "epoch": 25.380616740088104, "grad_norm": 1.2023996114730835, "learning_rate": 0.00010235944561372653, "loss": 0.0174, "step": 1777 }, { "epoch": 25.394713656387665, "grad_norm": 0.8008865118026733, "learning_rate": 0.000102266934511529, "loss": 0.0174, "step": 1778 }, { "epoch": 25.408810572687223, "grad_norm": 2.450110673904419, "learning_rate": 0.00010217442146818135, "loss": 0.0183, "step": 1779 }, { "epoch": 25.422907488986784, "grad_norm": 1.0126415491104126, "learning_rate": 0.00010208190656290137, "loss": 0.0195, "step": 1780 }, { "epoch": 25.437004405286345, "grad_norm": 1.055475115776062, "learning_rate": 0.00010198938987490858, "loss": 0.0165, "step": 1781 }, { "epoch": 25.451101321585902, "grad_norm": 0.8367187976837158, "learning_rate": 0.0001018968714834239, "loss": 0.0211, "step": 1782 }, { "epoch": 25.465198237885463, "grad_norm": 1.0434781312942505, "learning_rate": 0.00010180435146766984, "loss": 0.0174, "step": 1783 }, { "epoch": 25.47929515418502, "grad_norm": 1.2656900882720947, "learning_rate": 0.0001017118299068702, "loss": 0.0233, "step": 1784 }, { "epoch": 25.493392070484582, "grad_norm": 1.1865154504776, "learning_rate": 0.00010161930688025017, "loss": 0.0219, "step": 1785 }, { "epoch": 25.50748898678414, "grad_norm": 1.0705287456512451, "learning_rate": 0.00010152678246703609, "loss": 0.019, "step": 1786 }, { "epoch": 25.5215859030837, "grad_norm": 1.4515676498413086, "learning_rate": 0.00010143425674645563, "loss": 0.0214, "step": 1787 }, { "epoch": 25.53568281938326, "grad_norm": 1.0052070617675781, "learning_rate": 0.00010134172979773746, "loss": 0.0206, "step": 1788 }, { "epoch": 25.54977973568282, "grad_norm": 1.2095574140548706, "learning_rate": 0.00010124920170011141, "loss": 0.0183, "step": 1789 }, { "epoch": 25.56387665198238, "grad_norm": 0.8539672493934631, "learning_rate": 0.00010115667253280816, "loss": 0.0194, "step": 1790 }, { "epoch": 25.577973568281937, "grad_norm": 1.318459153175354, "learning_rate": 0.00010106414237505946, "loss": 0.0212, "step": 1791 }, { "epoch": 25.592070484581498, "grad_norm": 1.107632040977478, "learning_rate": 0.00010097161130609773, "loss": 0.0207, "step": 1792 }, { "epoch": 25.606167400881056, "grad_norm": 0.8297216296195984, "learning_rate": 0.00010087907940515637, "loss": 0.0213, "step": 1793 }, { "epoch": 25.620264317180617, "grad_norm": 1.5088599920272827, "learning_rate": 0.00010078654675146933, "loss": 0.0272, "step": 1794 }, { "epoch": 25.634361233480178, "grad_norm": 1.3355053663253784, "learning_rate": 0.00010069401342427129, "loss": 0.0214, "step": 1795 }, { "epoch": 25.648458149779735, "grad_norm": 1.4494032859802246, "learning_rate": 0.00010060147950279747, "loss": 0.022, "step": 1796 }, { "epoch": 25.662555066079296, "grad_norm": 1.2696210145950317, "learning_rate": 0.00010050894506628361, "loss": 0.021, "step": 1797 }, { "epoch": 25.676651982378853, "grad_norm": 1.8374533653259277, "learning_rate": 0.00010041641019396591, "loss": 0.021, "step": 1798 }, { "epoch": 25.690748898678414, "grad_norm": 1.2043551206588745, "learning_rate": 0.00010032387496508089, "loss": 0.0208, "step": 1799 }, { "epoch": 25.704845814977972, "grad_norm": 1.4816652536392212, "learning_rate": 0.00010023133945886545, "loss": 0.0268, "step": 1800 }, { "epoch": 25.718942731277533, "grad_norm": 1.3238766193389893, "learning_rate": 0.00010013880375455664, "loss": 0.0233, "step": 1801 }, { "epoch": 25.733039647577094, "grad_norm": 1.7215882539749146, "learning_rate": 0.0001000462679313918, "loss": 0.026, "step": 1802 }, { "epoch": 25.74713656387665, "grad_norm": 1.428874135017395, "learning_rate": 9.995373206860823e-05, "loss": 0.022, "step": 1803 }, { "epoch": 25.761233480176212, "grad_norm": 1.3733891248703003, "learning_rate": 9.986119624544338e-05, "loss": 0.0215, "step": 1804 }, { "epoch": 25.77533039647577, "grad_norm": 1.148262858390808, "learning_rate": 9.976866054113458e-05, "loss": 0.0212, "step": 1805 }, { "epoch": 25.78942731277533, "grad_norm": 1.9077274799346924, "learning_rate": 9.967612503491914e-05, "loss": 0.0237, "step": 1806 }, { "epoch": 25.80352422907489, "grad_norm": 0.9239325523376465, "learning_rate": 9.958358980603412e-05, "loss": 0.0213, "step": 1807 }, { "epoch": 25.81762114537445, "grad_norm": 1.0274006128311157, "learning_rate": 9.949105493371642e-05, "loss": 0.0211, "step": 1808 }, { "epoch": 25.83171806167401, "grad_norm": 1.6694161891937256, "learning_rate": 9.939852049720253e-05, "loss": 0.0203, "step": 1809 }, { "epoch": 25.845814977973568, "grad_norm": 1.4448310136795044, "learning_rate": 9.930598657572873e-05, "loss": 0.0256, "step": 1810 }, { "epoch": 25.85991189427313, "grad_norm": 1.3596982955932617, "learning_rate": 9.92134532485307e-05, "loss": 0.0175, "step": 1811 }, { "epoch": 25.874008810572686, "grad_norm": 1.0460838079452515, "learning_rate": 9.912092059484364e-05, "loss": 0.022, "step": 1812 }, { "epoch": 25.888105726872247, "grad_norm": 1.266317367553711, "learning_rate": 9.902838869390229e-05, "loss": 0.0245, "step": 1813 }, { "epoch": 25.902202643171805, "grad_norm": 1.6024835109710693, "learning_rate": 9.893585762494056e-05, "loss": 0.0245, "step": 1814 }, { "epoch": 25.916299559471366, "grad_norm": 1.0658347606658936, "learning_rate": 9.884332746719186e-05, "loss": 0.0183, "step": 1815 }, { "epoch": 25.930396475770927, "grad_norm": 1.010819673538208, "learning_rate": 9.875079829988861e-05, "loss": 0.0208, "step": 1816 }, { "epoch": 25.944493392070484, "grad_norm": 1.7563540935516357, "learning_rate": 9.865827020226255e-05, "loss": 0.0213, "step": 1817 }, { "epoch": 25.958590308370045, "grad_norm": 2.1401517391204834, "learning_rate": 9.856574325354438e-05, "loss": 0.021, "step": 1818 }, { "epoch": 25.972687224669603, "grad_norm": 1.2911505699157715, "learning_rate": 9.847321753296394e-05, "loss": 0.0214, "step": 1819 }, { "epoch": 25.986784140969164, "grad_norm": 1.1778661012649536, "learning_rate": 9.838069311974986e-05, "loss": 0.0244, "step": 1820 }, { "epoch": 26.01409691629956, "grad_norm": 0.9284198880195618, "learning_rate": 9.828817009312981e-05, "loss": 0.0312, "step": 1821 }, { "epoch": 26.02819383259912, "grad_norm": 0.7565658092498779, "learning_rate": 9.819564853233016e-05, "loss": 0.0138, "step": 1822 }, { "epoch": 26.04229074889868, "grad_norm": 1.1966131925582886, "learning_rate": 9.810312851657612e-05, "loss": 0.0154, "step": 1823 }, { "epoch": 26.056387665198237, "grad_norm": 0.8782988786697388, "learning_rate": 9.801061012509145e-05, "loss": 0.0139, "step": 1824 }, { "epoch": 26.070484581497798, "grad_norm": 0.8581539988517761, "learning_rate": 9.791809343709865e-05, "loss": 0.0148, "step": 1825 }, { "epoch": 26.084581497797355, "grad_norm": 1.0045098066329956, "learning_rate": 9.782557853181868e-05, "loss": 0.0239, "step": 1826 }, { "epoch": 26.098678414096916, "grad_norm": 0.83587646484375, "learning_rate": 9.7733065488471e-05, "loss": 0.0178, "step": 1827 }, { "epoch": 26.112775330396477, "grad_norm": 1.0540075302124023, "learning_rate": 9.764055438627351e-05, "loss": 0.0139, "step": 1828 }, { "epoch": 26.126872246696035, "grad_norm": 0.638715386390686, "learning_rate": 9.754804530444234e-05, "loss": 0.0142, "step": 1829 }, { "epoch": 26.140969162995596, "grad_norm": 1.2335864305496216, "learning_rate": 9.745553832219202e-05, "loss": 0.022, "step": 1830 }, { "epoch": 26.155066079295153, "grad_norm": 1.454305648803711, "learning_rate": 9.736303351873517e-05, "loss": 0.0156, "step": 1831 }, { "epoch": 26.169162995594714, "grad_norm": 0.8635680675506592, "learning_rate": 9.727053097328267e-05, "loss": 0.0174, "step": 1832 }, { "epoch": 26.18325991189427, "grad_norm": 0.7695879936218262, "learning_rate": 9.717803076504333e-05, "loss": 0.0181, "step": 1833 }, { "epoch": 26.197356828193833, "grad_norm": 0.9188652634620667, "learning_rate": 9.708553297322406e-05, "loss": 0.0162, "step": 1834 }, { "epoch": 26.211453744493394, "grad_norm": 2.0122082233428955, "learning_rate": 9.699303767702963e-05, "loss": 0.0177, "step": 1835 }, { "epoch": 26.22555066079295, "grad_norm": 0.5128809809684753, "learning_rate": 9.690054495566276e-05, "loss": 0.0138, "step": 1836 }, { "epoch": 26.239647577092512, "grad_norm": 1.0685144662857056, "learning_rate": 9.680805488832385e-05, "loss": 0.019, "step": 1837 }, { "epoch": 26.25374449339207, "grad_norm": 1.2214467525482178, "learning_rate": 9.671556755421115e-05, "loss": 0.019, "step": 1838 }, { "epoch": 26.26784140969163, "grad_norm": 1.203458309173584, "learning_rate": 9.662308303252045e-05, "loss": 0.0193, "step": 1839 }, { "epoch": 26.281938325991188, "grad_norm": 0.9580810070037842, "learning_rate": 9.653060140244524e-05, "loss": 0.0166, "step": 1840 }, { "epoch": 26.29603524229075, "grad_norm": 1.3189805746078491, "learning_rate": 9.643812274317644e-05, "loss": 0.0149, "step": 1841 }, { "epoch": 26.31013215859031, "grad_norm": 0.9837426543235779, "learning_rate": 9.634564713390252e-05, "loss": 0.0153, "step": 1842 }, { "epoch": 26.324229074889868, "grad_norm": 1.004402995109558, "learning_rate": 9.625317465380924e-05, "loss": 0.0175, "step": 1843 }, { "epoch": 26.33832599118943, "grad_norm": 1.0111894607543945, "learning_rate": 9.616070538207977e-05, "loss": 0.0168, "step": 1844 }, { "epoch": 26.352422907488986, "grad_norm": 1.4769924879074097, "learning_rate": 9.606823939789448e-05, "loss": 0.029, "step": 1845 }, { "epoch": 26.366519823788547, "grad_norm": 0.9307140111923218, "learning_rate": 9.597577678043093e-05, "loss": 0.0209, "step": 1846 }, { "epoch": 26.380616740088104, "grad_norm": 1.2816414833068848, "learning_rate": 9.58833176088638e-05, "loss": 0.0198, "step": 1847 }, { "epoch": 26.394713656387665, "grad_norm": 1.1267948150634766, "learning_rate": 9.579086196236482e-05, "loss": 0.0156, "step": 1848 }, { "epoch": 26.408810572687223, "grad_norm": 0.9048746228218079, "learning_rate": 9.569840992010274e-05, "loss": 0.0141, "step": 1849 }, { "epoch": 26.422907488986784, "grad_norm": 0.9996411800384521, "learning_rate": 9.560596156124314e-05, "loss": 0.018, "step": 1850 }, { "epoch": 26.437004405286345, "grad_norm": 0.8978670835494995, "learning_rate": 9.551351696494854e-05, "loss": 0.0148, "step": 1851 }, { "epoch": 26.451101321585902, "grad_norm": 0.8456095457077026, "learning_rate": 9.542107621037816e-05, "loss": 0.0165, "step": 1852 }, { "epoch": 26.465198237885463, "grad_norm": 0.9848284125328064, "learning_rate": 9.532863937668798e-05, "loss": 0.0156, "step": 1853 }, { "epoch": 26.47929515418502, "grad_norm": 0.7869148254394531, "learning_rate": 9.52362065430306e-05, "loss": 0.0153, "step": 1854 }, { "epoch": 26.493392070484582, "grad_norm": 0.881644606590271, "learning_rate": 9.514377778855521e-05, "loss": 0.0157, "step": 1855 }, { "epoch": 26.50748898678414, "grad_norm": 0.8253307938575745, "learning_rate": 9.505135319240746e-05, "loss": 0.0188, "step": 1856 }, { "epoch": 26.5215859030837, "grad_norm": 1.7196687459945679, "learning_rate": 9.495893283372951e-05, "loss": 0.0151, "step": 1857 }, { "epoch": 26.53568281938326, "grad_norm": 1.0067554712295532, "learning_rate": 9.48665167916599e-05, "loss": 0.0139, "step": 1858 }, { "epoch": 26.54977973568282, "grad_norm": 1.4028183221817017, "learning_rate": 9.477410514533334e-05, "loss": 0.0214, "step": 1859 }, { "epoch": 26.56387665198238, "grad_norm": 1.0186190605163574, "learning_rate": 9.468169797388097e-05, "loss": 0.0158, "step": 1860 }, { "epoch": 26.577973568281937, "grad_norm": 1.462233066558838, "learning_rate": 9.45892953564299e-05, "loss": 0.0213, "step": 1861 }, { "epoch": 26.592070484581498, "grad_norm": 0.890791654586792, "learning_rate": 9.449689737210352e-05, "loss": 0.0225, "step": 1862 }, { "epoch": 26.606167400881056, "grad_norm": 0.9176374673843384, "learning_rate": 9.440450410002112e-05, "loss": 0.0155, "step": 1863 }, { "epoch": 26.620264317180617, "grad_norm": 1.1918020248413086, "learning_rate": 9.431211561929799e-05, "loss": 0.0222, "step": 1864 }, { "epoch": 26.634361233480178, "grad_norm": 0.7448166012763977, "learning_rate": 9.421973200904538e-05, "loss": 0.0144, "step": 1865 }, { "epoch": 26.648458149779735, "grad_norm": 1.13314688205719, "learning_rate": 9.412735334837029e-05, "loss": 0.0188, "step": 1866 }, { "epoch": 26.662555066079296, "grad_norm": 1.2280124425888062, "learning_rate": 9.403497971637553e-05, "loss": 0.0171, "step": 1867 }, { "epoch": 26.676651982378853, "grad_norm": 1.3342853784561157, "learning_rate": 9.394261119215956e-05, "loss": 0.017, "step": 1868 }, { "epoch": 26.690748898678414, "grad_norm": 1.4968206882476807, "learning_rate": 9.385024785481654e-05, "loss": 0.0183, "step": 1869 }, { "epoch": 26.704845814977972, "grad_norm": 1.1065633296966553, "learning_rate": 9.375788978343608e-05, "loss": 0.0239, "step": 1870 }, { "epoch": 26.718942731277533, "grad_norm": 1.0193146467208862, "learning_rate": 9.36655370571034e-05, "loss": 0.0168, "step": 1871 }, { "epoch": 26.733039647577094, "grad_norm": 0.8451933264732361, "learning_rate": 9.357318975489904e-05, "loss": 0.0184, "step": 1872 }, { "epoch": 26.74713656387665, "grad_norm": 1.6563068628311157, "learning_rate": 9.3480847955899e-05, "loss": 0.0257, "step": 1873 }, { "epoch": 26.761233480176212, "grad_norm": 0.9447290301322937, "learning_rate": 9.338851173917444e-05, "loss": 0.0173, "step": 1874 }, { "epoch": 26.77533039647577, "grad_norm": 0.9668636322021484, "learning_rate": 9.329618118379187e-05, "loss": 0.0176, "step": 1875 }, { "epoch": 26.78942731277533, "grad_norm": 1.9218626022338867, "learning_rate": 9.320385636881283e-05, "loss": 0.0234, "step": 1876 }, { "epoch": 26.80352422907489, "grad_norm": 0.7823576927185059, "learning_rate": 9.311153737329407e-05, "loss": 0.0156, "step": 1877 }, { "epoch": 26.81762114537445, "grad_norm": 0.7970376014709473, "learning_rate": 9.301922427628724e-05, "loss": 0.0164, "step": 1878 }, { "epoch": 26.83171806167401, "grad_norm": 0.9224509000778198, "learning_rate": 9.292691715683901e-05, "loss": 0.0161, "step": 1879 }, { "epoch": 26.845814977973568, "grad_norm": 1.2052466869354248, "learning_rate": 9.283461609399094e-05, "loss": 0.0175, "step": 1880 }, { "epoch": 26.85991189427313, "grad_norm": 1.0670897960662842, "learning_rate": 9.274232116677932e-05, "loss": 0.0199, "step": 1881 }, { "epoch": 26.874008810572686, "grad_norm": 1.2157707214355469, "learning_rate": 9.26500324542353e-05, "loss": 0.0236, "step": 1882 }, { "epoch": 26.888105726872247, "grad_norm": 0.7954749464988708, "learning_rate": 9.255775003538462e-05, "loss": 0.0156, "step": 1883 }, { "epoch": 26.902202643171805, "grad_norm": 0.9992148280143738, "learning_rate": 9.246547398924768e-05, "loss": 0.021, "step": 1884 }, { "epoch": 26.916299559471366, "grad_norm": 0.7961944341659546, "learning_rate": 9.237320439483937e-05, "loss": 0.0187, "step": 1885 }, { "epoch": 26.930396475770927, "grad_norm": 1.18401038646698, "learning_rate": 9.228094133116913e-05, "loss": 0.0174, "step": 1886 }, { "epoch": 26.944493392070484, "grad_norm": 1.5651177167892456, "learning_rate": 9.218868487724074e-05, "loss": 0.0239, "step": 1887 }, { "epoch": 26.958590308370045, "grad_norm": 1.1379789113998413, "learning_rate": 9.20964351120524e-05, "loss": 0.019, "step": 1888 }, { "epoch": 26.972687224669603, "grad_norm": 1.6080833673477173, "learning_rate": 9.200419211459646e-05, "loss": 0.0156, "step": 1889 }, { "epoch": 26.986784140969164, "grad_norm": 1.0382380485534668, "learning_rate": 9.19119559638596e-05, "loss": 0.0235, "step": 1890 }, { "epoch": 27.01409691629956, "grad_norm": 0.7763023376464844, "learning_rate": 9.181972673882255e-05, "loss": 0.0286, "step": 1891 }, { "epoch": 27.02819383259912, "grad_norm": 0.752079427242279, "learning_rate": 9.172750451846016e-05, "loss": 0.0137, "step": 1892 }, { "epoch": 27.04229074889868, "grad_norm": 1.0402010679244995, "learning_rate": 9.163528938174127e-05, "loss": 0.0129, "step": 1893 }, { "epoch": 27.056387665198237, "grad_norm": 0.6439253091812134, "learning_rate": 9.154308140762865e-05, "loss": 0.0116, "step": 1894 }, { "epoch": 27.070484581497798, "grad_norm": 0.5906344652175903, "learning_rate": 9.14508806750789e-05, "loss": 0.0133, "step": 1895 }, { "epoch": 27.084581497797355, "grad_norm": 0.6363478899002075, "learning_rate": 9.135868726304251e-05, "loss": 0.0128, "step": 1896 }, { "epoch": 27.098678414096916, "grad_norm": 0.7940772771835327, "learning_rate": 9.126650125046361e-05, "loss": 0.013, "step": 1897 }, { "epoch": 27.112775330396477, "grad_norm": 0.9968953728675842, "learning_rate": 9.117432271628004e-05, "loss": 0.0127, "step": 1898 }, { "epoch": 27.126872246696035, "grad_norm": 1.0351765155792236, "learning_rate": 9.108215173942324e-05, "loss": 0.0118, "step": 1899 }, { "epoch": 27.140969162995596, "grad_norm": 0.6429433226585388, "learning_rate": 9.098998839881813e-05, "loss": 0.0134, "step": 1900 }, { "epoch": 27.155066079295153, "grad_norm": 0.6452343463897705, "learning_rate": 9.089783277338318e-05, "loss": 0.0129, "step": 1901 }, { "epoch": 27.169162995594714, "grad_norm": 0.6822871565818787, "learning_rate": 9.080568494203013e-05, "loss": 0.0125, "step": 1902 }, { "epoch": 27.18325991189427, "grad_norm": 0.7562715411186218, "learning_rate": 9.07135449836642e-05, "loss": 0.0124, "step": 1903 }, { "epoch": 27.197356828193833, "grad_norm": 0.7658612132072449, "learning_rate": 9.062141297718371e-05, "loss": 0.012, "step": 1904 }, { "epoch": 27.211453744493394, "grad_norm": 0.5673697590827942, "learning_rate": 9.052928900148031e-05, "loss": 0.0123, "step": 1905 }, { "epoch": 27.22555066079295, "grad_norm": 1.302025556564331, "learning_rate": 9.043717313543866e-05, "loss": 0.0157, "step": 1906 }, { "epoch": 27.239647577092512, "grad_norm": 0.7579382658004761, "learning_rate": 9.034506545793655e-05, "loss": 0.0242, "step": 1907 }, { "epoch": 27.25374449339207, "grad_norm": 1.0603113174438477, "learning_rate": 9.025296604784471e-05, "loss": 0.0151, "step": 1908 }, { "epoch": 27.26784140969163, "grad_norm": 1.0289950370788574, "learning_rate": 9.016087498402686e-05, "loss": 0.0139, "step": 1909 }, { "epoch": 27.281938325991188, "grad_norm": 0.7913498878479004, "learning_rate": 9.006879234533946e-05, "loss": 0.0148, "step": 1910 }, { "epoch": 27.29603524229075, "grad_norm": 0.9295881390571594, "learning_rate": 8.997671821063191e-05, "loss": 0.0129, "step": 1911 }, { "epoch": 27.31013215859031, "grad_norm": 0.5933603644371033, "learning_rate": 8.988465265874615e-05, "loss": 0.0137, "step": 1912 }, { "epoch": 27.324229074889868, "grad_norm": 0.9829628467559814, "learning_rate": 8.979259576851693e-05, "loss": 0.017, "step": 1913 }, { "epoch": 27.33832599118943, "grad_norm": 1.1541839838027954, "learning_rate": 8.970054761877148e-05, "loss": 0.0141, "step": 1914 }, { "epoch": 27.352422907488986, "grad_norm": 0.6530998349189758, "learning_rate": 8.960850828832957e-05, "loss": 0.0145, "step": 1915 }, { "epoch": 27.366519823788547, "grad_norm": 1.1363379955291748, "learning_rate": 8.951647785600348e-05, "loss": 0.0131, "step": 1916 }, { "epoch": 27.380616740088104, "grad_norm": 0.8654447793960571, "learning_rate": 8.942445640059775e-05, "loss": 0.0178, "step": 1917 }, { "epoch": 27.394713656387665, "grad_norm": 0.9942761063575745, "learning_rate": 8.933244400090937e-05, "loss": 0.0143, "step": 1918 }, { "epoch": 27.408810572687223, "grad_norm": 0.5920919179916382, "learning_rate": 8.924044073572745e-05, "loss": 0.0134, "step": 1919 }, { "epoch": 27.422907488986784, "grad_norm": 0.9049875140190125, "learning_rate": 8.914844668383339e-05, "loss": 0.0115, "step": 1920 }, { "epoch": 27.437004405286345, "grad_norm": 0.8144906163215637, "learning_rate": 8.905646192400059e-05, "loss": 0.0145, "step": 1921 }, { "epoch": 27.451101321585902, "grad_norm": 0.84299236536026, "learning_rate": 8.896448653499461e-05, "loss": 0.0168, "step": 1922 }, { "epoch": 27.465198237885463, "grad_norm": 0.7382621765136719, "learning_rate": 8.887252059557286e-05, "loss": 0.0147, "step": 1923 }, { "epoch": 27.47929515418502, "grad_norm": 0.7734161019325256, "learning_rate": 8.878056418448479e-05, "loss": 0.0158, "step": 1924 }, { "epoch": 27.493392070484582, "grad_norm": 0.7610594630241394, "learning_rate": 8.868861738047158e-05, "loss": 0.0119, "step": 1925 }, { "epoch": 27.50748898678414, "grad_norm": 0.7646520137786865, "learning_rate": 8.859668026226627e-05, "loss": 0.0147, "step": 1926 }, { "epoch": 27.5215859030837, "grad_norm": 0.8569666743278503, "learning_rate": 8.850475290859351e-05, "loss": 0.015, "step": 1927 }, { "epoch": 27.53568281938326, "grad_norm": 0.8832091689109802, "learning_rate": 8.841283539816969e-05, "loss": 0.013, "step": 1928 }, { "epoch": 27.54977973568282, "grad_norm": 0.789119303226471, "learning_rate": 8.832092780970272e-05, "loss": 0.0123, "step": 1929 }, { "epoch": 27.56387665198238, "grad_norm": 1.1210695505142212, "learning_rate": 8.822903022189199e-05, "loss": 0.0164, "step": 1930 }, { "epoch": 27.577973568281937, "grad_norm": 0.9303926825523376, "learning_rate": 8.81371427134284e-05, "loss": 0.0157, "step": 1931 }, { "epoch": 27.592070484581498, "grad_norm": 0.734791100025177, "learning_rate": 8.804526536299413e-05, "loss": 0.0162, "step": 1932 }, { "epoch": 27.606167400881056, "grad_norm": 0.9312399625778198, "learning_rate": 8.795339824926273e-05, "loss": 0.0154, "step": 1933 }, { "epoch": 27.620264317180617, "grad_norm": 0.8901458978652954, "learning_rate": 8.786154145089896e-05, "loss": 0.0163, "step": 1934 }, { "epoch": 27.634361233480178, "grad_norm": 0.7315226197242737, "learning_rate": 8.776969504655875e-05, "loss": 0.0131, "step": 1935 }, { "epoch": 27.648458149779735, "grad_norm": 0.6000968217849731, "learning_rate": 8.767785911488911e-05, "loss": 0.0129, "step": 1936 }, { "epoch": 27.662555066079296, "grad_norm": 0.9067854285240173, "learning_rate": 8.758603373452814e-05, "loss": 0.0154, "step": 1937 }, { "epoch": 27.676651982378853, "grad_norm": 1.4436746835708618, "learning_rate": 8.749421898410482e-05, "loss": 0.014, "step": 1938 }, { "epoch": 27.690748898678414, "grad_norm": 1.046949863433838, "learning_rate": 8.740241494223911e-05, "loss": 0.0147, "step": 1939 }, { "epoch": 27.704845814977972, "grad_norm": 0.9128537774085999, "learning_rate": 8.731062168754174e-05, "loss": 0.0176, "step": 1940 }, { "epoch": 27.718942731277533, "grad_norm": 0.7700223922729492, "learning_rate": 8.721883929861424e-05, "loss": 0.0181, "step": 1941 }, { "epoch": 27.733039647577094, "grad_norm": 0.9629022479057312, "learning_rate": 8.71270678540488e-05, "loss": 0.0148, "step": 1942 }, { "epoch": 27.74713656387665, "grad_norm": 1.156590223312378, "learning_rate": 8.70353074324283e-05, "loss": 0.0168, "step": 1943 }, { "epoch": 27.761233480176212, "grad_norm": 0.7383662462234497, "learning_rate": 8.69435581123261e-05, "loss": 0.0163, "step": 1944 }, { "epoch": 27.77533039647577, "grad_norm": 1.144924521446228, "learning_rate": 8.685181997230614e-05, "loss": 0.0149, "step": 1945 }, { "epoch": 27.78942731277533, "grad_norm": 1.0982452630996704, "learning_rate": 8.676009309092272e-05, "loss": 0.0155, "step": 1946 }, { "epoch": 27.80352422907489, "grad_norm": 1.1293474435806274, "learning_rate": 8.66683775467205e-05, "loss": 0.0117, "step": 1947 }, { "epoch": 27.81762114537445, "grad_norm": 0.7444452047348022, "learning_rate": 8.657667341823448e-05, "loss": 0.0138, "step": 1948 }, { "epoch": 27.83171806167401, "grad_norm": 1.0646672248840332, "learning_rate": 8.648498078398984e-05, "loss": 0.0125, "step": 1949 }, { "epoch": 27.845814977973568, "grad_norm": 1.0225173234939575, "learning_rate": 8.639329972250198e-05, "loss": 0.0168, "step": 1950 }, { "epoch": 27.85991189427313, "grad_norm": 1.296364188194275, "learning_rate": 8.630163031227629e-05, "loss": 0.0157, "step": 1951 }, { "epoch": 27.874008810572686, "grad_norm": 0.8917664289474487, "learning_rate": 8.620997263180829e-05, "loss": 0.0157, "step": 1952 }, { "epoch": 27.888105726872247, "grad_norm": 1.0903619527816772, "learning_rate": 8.611832675958336e-05, "loss": 0.0132, "step": 1953 }, { "epoch": 27.902202643171805, "grad_norm": 0.5359286069869995, "learning_rate": 8.602669277407685e-05, "loss": 0.015, "step": 1954 }, { "epoch": 27.916299559471366, "grad_norm": 0.7918272018432617, "learning_rate": 8.593507075375387e-05, "loss": 0.0149, "step": 1955 }, { "epoch": 27.930396475770927, "grad_norm": 0.7631111741065979, "learning_rate": 8.584346077706936e-05, "loss": 0.0166, "step": 1956 }, { "epoch": 27.944493392070484, "grad_norm": 1.489310622215271, "learning_rate": 8.575186292246782e-05, "loss": 0.0166, "step": 1957 }, { "epoch": 27.958590308370045, "grad_norm": 0.6005854606628418, "learning_rate": 8.566027726838351e-05, "loss": 0.0161, "step": 1958 }, { "epoch": 27.972687224669603, "grad_norm": 1.2772949934005737, "learning_rate": 8.556870389324014e-05, "loss": 0.0155, "step": 1959 }, { "epoch": 27.986784140969164, "grad_norm": 0.8474998474121094, "learning_rate": 8.5477142875451e-05, "loss": 0.0136, "step": 1960 }, { "epoch": 28.01409691629956, "grad_norm": 0.8547915816307068, "learning_rate": 8.538559429341869e-05, "loss": 0.0256, "step": 1961 }, { "epoch": 28.02819383259912, "grad_norm": 0.5288060903549194, "learning_rate": 8.52940582255352e-05, "loss": 0.0115, "step": 1962 }, { "epoch": 28.04229074889868, "grad_norm": 1.0645703077316284, "learning_rate": 8.520253475018192e-05, "loss": 0.0116, "step": 1963 }, { "epoch": 28.056387665198237, "grad_norm": 0.4976629614830017, "learning_rate": 8.511102394572926e-05, "loss": 0.0113, "step": 1964 }, { "epoch": 28.070484581497798, "grad_norm": 0.8355560898780823, "learning_rate": 8.501952589053695e-05, "loss": 0.0135, "step": 1965 }, { "epoch": 28.084581497797355, "grad_norm": 0.8188231587409973, "learning_rate": 8.492804066295367e-05, "loss": 0.0119, "step": 1966 }, { "epoch": 28.098678414096916, "grad_norm": 0.6970875859260559, "learning_rate": 8.48365683413172e-05, "loss": 0.0113, "step": 1967 }, { "epoch": 28.112775330396477, "grad_norm": 0.35886356234550476, "learning_rate": 8.474510900395428e-05, "loss": 0.0109, "step": 1968 }, { "epoch": 28.126872246696035, "grad_norm": 0.4269399046897888, "learning_rate": 8.46536627291804e-05, "loss": 0.0132, "step": 1969 }, { "epoch": 28.140969162995596, "grad_norm": 0.5846717357635498, "learning_rate": 8.456222959530008e-05, "loss": 0.0153, "step": 1970 }, { "epoch": 28.155066079295153, "grad_norm": 1.0959408283233643, "learning_rate": 8.447080968060638e-05, "loss": 0.0107, "step": 1971 }, { "epoch": 28.169162995594714, "grad_norm": 0.8363583087921143, "learning_rate": 8.437940306338119e-05, "loss": 0.0116, "step": 1972 }, { "epoch": 28.18325991189427, "grad_norm": 0.6454710364341736, "learning_rate": 8.428800982189491e-05, "loss": 0.0128, "step": 1973 }, { "epoch": 28.197356828193833, "grad_norm": 0.8311361074447632, "learning_rate": 8.419663003440657e-05, "loss": 0.0107, "step": 1974 }, { "epoch": 28.211453744493394, "grad_norm": 0.4348561465740204, "learning_rate": 8.410526377916359e-05, "loss": 0.0133, "step": 1975 }, { "epoch": 28.22555066079295, "grad_norm": 0.7578161358833313, "learning_rate": 8.401391113440189e-05, "loss": 0.0113, "step": 1976 }, { "epoch": 28.239647577092512, "grad_norm": 0.7406066656112671, "learning_rate": 8.392257217834568e-05, "loss": 0.0162, "step": 1977 }, { "epoch": 28.25374449339207, "grad_norm": 0.5479652881622314, "learning_rate": 8.383124698920749e-05, "loss": 0.0111, "step": 1978 }, { "epoch": 28.26784140969163, "grad_norm": 0.6164579391479492, "learning_rate": 8.3739935645188e-05, "loss": 0.011, "step": 1979 }, { "epoch": 28.281938325991188, "grad_norm": 0.8219427466392517, "learning_rate": 8.36486382244761e-05, "loss": 0.0108, "step": 1980 }, { "epoch": 28.29603524229075, "grad_norm": 0.4625411331653595, "learning_rate": 8.355735480524874e-05, "loss": 0.0127, "step": 1981 }, { "epoch": 28.31013215859031, "grad_norm": 1.1337915658950806, "learning_rate": 8.346608546567081e-05, "loss": 0.0143, "step": 1982 }, { "epoch": 28.324229074889868, "grad_norm": 0.7474648356437683, "learning_rate": 8.337483028389527e-05, "loss": 0.0125, "step": 1983 }, { "epoch": 28.33832599118943, "grad_norm": 0.9847201108932495, "learning_rate": 8.328358933806286e-05, "loss": 0.0113, "step": 1984 }, { "epoch": 28.352422907488986, "grad_norm": 0.43941929936408997, "learning_rate": 8.319236270630218e-05, "loss": 0.011, "step": 1985 }, { "epoch": 28.366519823788547, "grad_norm": 0.6058652997016907, "learning_rate": 8.310115046672953e-05, "loss": 0.011, "step": 1986 }, { "epoch": 28.380616740088104, "grad_norm": 1.4499359130859375, "learning_rate": 8.300995269744893e-05, "loss": 0.01, "step": 1987 }, { "epoch": 28.394713656387665, "grad_norm": 0.40075552463531494, "learning_rate": 8.291876947655196e-05, "loss": 0.0125, "step": 1988 }, { "epoch": 28.408810572687223, "grad_norm": 0.6904861330986023, "learning_rate": 8.282760088211782e-05, "loss": 0.0114, "step": 1989 }, { "epoch": 28.422907488986784, "grad_norm": 0.7197403907775879, "learning_rate": 8.273644699221308e-05, "loss": 0.0156, "step": 1990 }, { "epoch": 28.437004405286345, "grad_norm": 0.6649377346038818, "learning_rate": 8.264530788489185e-05, "loss": 0.0139, "step": 1991 }, { "epoch": 28.451101321585902, "grad_norm": 0.6681014895439148, "learning_rate": 8.25541836381954e-05, "loss": 0.0124, "step": 1992 }, { "epoch": 28.465198237885463, "grad_norm": 1.2448291778564453, "learning_rate": 8.24630743301525e-05, "loss": 0.0112, "step": 1993 }, { "epoch": 28.47929515418502, "grad_norm": 0.7210597991943359, "learning_rate": 8.237198003877893e-05, "loss": 0.013, "step": 1994 }, { "epoch": 28.493392070484582, "grad_norm": 0.5220016241073608, "learning_rate": 8.228090084207774e-05, "loss": 0.0135, "step": 1995 }, { "epoch": 28.50748898678414, "grad_norm": 0.9907912611961365, "learning_rate": 8.218983681803896e-05, "loss": 0.0163, "step": 1996 }, { "epoch": 28.5215859030837, "grad_norm": 1.4296538829803467, "learning_rate": 8.209878804463972e-05, "loss": 0.0135, "step": 1997 }, { "epoch": 28.53568281938326, "grad_norm": 0.6382324695587158, "learning_rate": 8.200775459984403e-05, "loss": 0.0129, "step": 1998 }, { "epoch": 28.54977973568282, "grad_norm": 0.6878402233123779, "learning_rate": 8.191673656160276e-05, "loss": 0.0116, "step": 1999 }, { "epoch": 28.56387665198238, "grad_norm": 0.6007063388824463, "learning_rate": 8.182573400785367e-05, "loss": 0.013, "step": 2000 }, { "epoch": 28.56387665198238, "eval_loss": 1.397882342338562, "eval_runtime": 132.7065, "eval_samples_per_second": 13.624, "eval_steps_per_second": 0.43, "step": 2000 }, { "epoch": 28.577973568281937, "grad_norm": 1.2889543771743774, "learning_rate": 8.173474701652116e-05, "loss": 0.0103, "step": 2001 }, { "epoch": 28.592070484581498, "grad_norm": 0.6256561279296875, "learning_rate": 8.16437756655164e-05, "loss": 0.0137, "step": 2002 }, { "epoch": 28.606167400881056, "grad_norm": 0.7462039589881897, "learning_rate": 8.155282003273708e-05, "loss": 0.0145, "step": 2003 }, { "epoch": 28.620264317180617, "grad_norm": 0.7119969129562378, "learning_rate": 8.146188019606749e-05, "loss": 0.0105, "step": 2004 }, { "epoch": 28.634361233480178, "grad_norm": 0.5678698420524597, "learning_rate": 8.137095623337836e-05, "loss": 0.0126, "step": 2005 }, { "epoch": 28.648458149779735, "grad_norm": 0.8129593729972839, "learning_rate": 8.128004822252686e-05, "loss": 0.0158, "step": 2006 }, { "epoch": 28.662555066079296, "grad_norm": 1.077824354171753, "learning_rate": 8.118915624135647e-05, "loss": 0.0116, "step": 2007 }, { "epoch": 28.676651982378853, "grad_norm": 0.7011913657188416, "learning_rate": 8.109828036769698e-05, "loss": 0.0157, "step": 2008 }, { "epoch": 28.690748898678414, "grad_norm": 0.7290184497833252, "learning_rate": 8.100742067936431e-05, "loss": 0.0147, "step": 2009 }, { "epoch": 28.704845814977972, "grad_norm": 0.6044295430183411, "learning_rate": 8.091657725416065e-05, "loss": 0.0126, "step": 2010 }, { "epoch": 28.718942731277533, "grad_norm": 1.0199460983276367, "learning_rate": 8.082575016987411e-05, "loss": 0.0113, "step": 2011 }, { "epoch": 28.733039647577094, "grad_norm": 0.5863751173019409, "learning_rate": 8.073493950427896e-05, "loss": 0.0126, "step": 2012 }, { "epoch": 28.74713656387665, "grad_norm": 1.0131657123565674, "learning_rate": 8.064414533513527e-05, "loss": 0.0128, "step": 2013 }, { "epoch": 28.761233480176212, "grad_norm": 0.5188210606575012, "learning_rate": 8.055336774018909e-05, "loss": 0.011, "step": 2014 }, { "epoch": 28.77533039647577, "grad_norm": 0.7827507257461548, "learning_rate": 8.046260679717225e-05, "loss": 0.0115, "step": 2015 }, { "epoch": 28.78942731277533, "grad_norm": 0.8536536693572998, "learning_rate": 8.037186258380226e-05, "loss": 0.0146, "step": 2016 }, { "epoch": 28.80352422907489, "grad_norm": 1.0432523488998413, "learning_rate": 8.02811351777824e-05, "loss": 0.0172, "step": 2017 }, { "epoch": 28.81762114537445, "grad_norm": 0.6204456090927124, "learning_rate": 8.01904246568015e-05, "loss": 0.012, "step": 2018 }, { "epoch": 28.83171806167401, "grad_norm": 1.137407898902893, "learning_rate": 8.009973109853393e-05, "loss": 0.019, "step": 2019 }, { "epoch": 28.845814977973568, "grad_norm": 1.0290502309799194, "learning_rate": 8.000905458063957e-05, "loss": 0.0131, "step": 2020 }, { "epoch": 28.85991189427313, "grad_norm": 0.5879244208335876, "learning_rate": 7.991839518076367e-05, "loss": 0.0137, "step": 2021 }, { "epoch": 28.874008810572686, "grad_norm": 0.8862682580947876, "learning_rate": 7.982775297653684e-05, "loss": 0.0123, "step": 2022 }, { "epoch": 28.888105726872247, "grad_norm": 0.8442701101303101, "learning_rate": 7.973712804557501e-05, "loss": 0.0129, "step": 2023 }, { "epoch": 28.902202643171805, "grad_norm": 0.8656349182128906, "learning_rate": 7.964652046547919e-05, "loss": 0.0138, "step": 2024 }, { "epoch": 28.916299559471366, "grad_norm": 0.936354398727417, "learning_rate": 7.95559303138357e-05, "loss": 0.0169, "step": 2025 }, { "epoch": 28.930396475770927, "grad_norm": 0.8641992211341858, "learning_rate": 7.946535766821579e-05, "loss": 0.0166, "step": 2026 }, { "epoch": 28.944493392070484, "grad_norm": 1.7508525848388672, "learning_rate": 7.937480260617583e-05, "loss": 0.0123, "step": 2027 }, { "epoch": 28.958590308370045, "grad_norm": 0.5103455185890198, "learning_rate": 7.928426520525706e-05, "loss": 0.0131, "step": 2028 }, { "epoch": 28.972687224669603, "grad_norm": 0.9291965961456299, "learning_rate": 7.919374554298564e-05, "loss": 0.0131, "step": 2029 }, { "epoch": 28.986784140969164, "grad_norm": 0.6481388211250305, "learning_rate": 7.91032436968725e-05, "loss": 0.0136, "step": 2030 }, { "epoch": 29.01409691629956, "grad_norm": 0.44885948300361633, "learning_rate": 7.901275974441337e-05, "loss": 0.0187, "step": 2031 }, { "epoch": 29.02819383259912, "grad_norm": 0.3996211588382721, "learning_rate": 7.89222937630886e-05, "loss": 0.0096, "step": 2032 }, { "epoch": 29.04229074889868, "grad_norm": 0.4989689290523529, "learning_rate": 7.883184583036318e-05, "loss": 0.0113, "step": 2033 }, { "epoch": 29.056387665198237, "grad_norm": 0.7303158044815063, "learning_rate": 7.874141602368666e-05, "loss": 0.0095, "step": 2034 }, { "epoch": 29.070484581497798, "grad_norm": 0.8254972696304321, "learning_rate": 7.865100442049303e-05, "loss": 0.0097, "step": 2035 }, { "epoch": 29.084581497797355, "grad_norm": 0.6418335437774658, "learning_rate": 7.856061109820075e-05, "loss": 0.0099, "step": 2036 }, { "epoch": 29.098678414096916, "grad_norm": 0.6365072131156921, "learning_rate": 7.847023613421251e-05, "loss": 0.012, "step": 2037 }, { "epoch": 29.112775330396477, "grad_norm": 0.41604191064834595, "learning_rate": 7.837987960591548e-05, "loss": 0.0103, "step": 2038 }, { "epoch": 29.126872246696035, "grad_norm": 0.6154586672782898, "learning_rate": 7.828954159068081e-05, "loss": 0.012, "step": 2039 }, { "epoch": 29.140969162995596, "grad_norm": 0.518613338470459, "learning_rate": 7.8199222165864e-05, "loss": 0.0106, "step": 2040 }, { "epoch": 29.155066079295153, "grad_norm": 0.43144840002059937, "learning_rate": 7.810892140880448e-05, "loss": 0.0089, "step": 2041 }, { "epoch": 29.169162995594714, "grad_norm": 0.8178680539131165, "learning_rate": 7.80186393968258e-05, "loss": 0.0103, "step": 2042 }, { "epoch": 29.18325991189427, "grad_norm": 0.4618457555770874, "learning_rate": 7.792837620723537e-05, "loss": 0.0114, "step": 2043 }, { "epoch": 29.197356828193833, "grad_norm": 0.45004770159721375, "learning_rate": 7.78381319173246e-05, "loss": 0.0108, "step": 2044 }, { "epoch": 29.211453744493394, "grad_norm": 0.6338578462600708, "learning_rate": 7.774790660436858e-05, "loss": 0.0102, "step": 2045 }, { "epoch": 29.22555066079295, "grad_norm": 0.46576645970344543, "learning_rate": 7.765770034562625e-05, "loss": 0.0084, "step": 2046 }, { "epoch": 29.239647577092512, "grad_norm": 0.427555650472641, "learning_rate": 7.756751321834019e-05, "loss": 0.0085, "step": 2047 }, { "epoch": 29.25374449339207, "grad_norm": 0.39850643277168274, "learning_rate": 7.747734529973659e-05, "loss": 0.0135, "step": 2048 }, { "epoch": 29.26784140969163, "grad_norm": 1.225801706314087, "learning_rate": 7.738719666702526e-05, "loss": 0.0115, "step": 2049 }, { "epoch": 29.281938325991188, "grad_norm": 0.8522629737854004, "learning_rate": 7.729706739739939e-05, "loss": 0.0124, "step": 2050 }, { "epoch": 29.29603524229075, "grad_norm": 0.42482978105545044, "learning_rate": 7.72069575680357e-05, "loss": 0.0101, "step": 2051 }, { "epoch": 29.31013215859031, "grad_norm": 0.7091541886329651, "learning_rate": 7.711686725609413e-05, "loss": 0.0118, "step": 2052 }, { "epoch": 29.324229074889868, "grad_norm": 0.602077841758728, "learning_rate": 7.702679653871807e-05, "loss": 0.0108, "step": 2053 }, { "epoch": 29.33832599118943, "grad_norm": 0.49286144971847534, "learning_rate": 7.6936745493034e-05, "loss": 0.0116, "step": 2054 }, { "epoch": 29.352422907488986, "grad_norm": 0.6383890509605408, "learning_rate": 7.684671419615163e-05, "loss": 0.0087, "step": 2055 }, { "epoch": 29.366519823788547, "grad_norm": 0.5174060463905334, "learning_rate": 7.675670272516369e-05, "loss": 0.0109, "step": 2056 }, { "epoch": 29.380616740088104, "grad_norm": 1.1516528129577637, "learning_rate": 7.666671115714603e-05, "loss": 0.0119, "step": 2057 }, { "epoch": 29.394713656387665, "grad_norm": 1.720984935760498, "learning_rate": 7.657673956915735e-05, "loss": 0.0115, "step": 2058 }, { "epoch": 29.408810572687223, "grad_norm": 0.9535403847694397, "learning_rate": 7.648678803823934e-05, "loss": 0.0148, "step": 2059 }, { "epoch": 29.422907488986784, "grad_norm": 0.8515024781227112, "learning_rate": 7.639685664141643e-05, "loss": 0.0093, "step": 2060 }, { "epoch": 29.437004405286345, "grad_norm": 0.41958749294281006, "learning_rate": 7.63069454556959e-05, "loss": 0.0127, "step": 2061 }, { "epoch": 29.451101321585902, "grad_norm": 0.7432211637496948, "learning_rate": 7.621705455806761e-05, "loss": 0.0111, "step": 2062 }, { "epoch": 29.465198237885463, "grad_norm": 0.5496025085449219, "learning_rate": 7.612718402550417e-05, "loss": 0.01, "step": 2063 }, { "epoch": 29.47929515418502, "grad_norm": 0.6700388789176941, "learning_rate": 7.603733393496066e-05, "loss": 0.0143, "step": 2064 }, { "epoch": 29.493392070484582, "grad_norm": 0.5383269786834717, "learning_rate": 7.594750436337467e-05, "loss": 0.0116, "step": 2065 }, { "epoch": 29.50748898678414, "grad_norm": 0.7984718680381775, "learning_rate": 7.585769538766628e-05, "loss": 0.0115, "step": 2066 }, { "epoch": 29.5215859030837, "grad_norm": 0.5601639747619629, "learning_rate": 7.576790708473783e-05, "loss": 0.0097, "step": 2067 }, { "epoch": 29.53568281938326, "grad_norm": 0.7499014735221863, "learning_rate": 7.567813953147405e-05, "loss": 0.0108, "step": 2068 }, { "epoch": 29.54977973568282, "grad_norm": 0.40570616722106934, "learning_rate": 7.558839280474192e-05, "loss": 0.0103, "step": 2069 }, { "epoch": 29.56387665198238, "grad_norm": 0.6030969023704529, "learning_rate": 7.549866698139048e-05, "loss": 0.0119, "step": 2070 }, { "epoch": 29.577973568281937, "grad_norm": 1.033094048500061, "learning_rate": 7.540896213825097e-05, "loss": 0.0096, "step": 2071 }, { "epoch": 29.592070484581498, "grad_norm": 0.36959248781204224, "learning_rate": 7.531927835213656e-05, "loss": 0.0104, "step": 2072 }, { "epoch": 29.606167400881056, "grad_norm": 0.6346567869186401, "learning_rate": 7.522961569984253e-05, "loss": 0.0127, "step": 2073 }, { "epoch": 29.620264317180617, "grad_norm": 0.7526529431343079, "learning_rate": 7.513997425814595e-05, "loss": 0.0148, "step": 2074 }, { "epoch": 29.634361233480178, "grad_norm": 1.4486616849899292, "learning_rate": 7.505035410380575e-05, "loss": 0.0106, "step": 2075 }, { "epoch": 29.648458149779735, "grad_norm": 0.5640919208526611, "learning_rate": 7.496075531356263e-05, "loss": 0.01, "step": 2076 }, { "epoch": 29.662555066079296, "grad_norm": 0.40498635172843933, "learning_rate": 7.48711779641391e-05, "loss": 0.0123, "step": 2077 }, { "epoch": 29.676651982378853, "grad_norm": 0.5127254724502563, "learning_rate": 7.47816221322391e-05, "loss": 0.0133, "step": 2078 }, { "epoch": 29.690748898678414, "grad_norm": 1.6433721780776978, "learning_rate": 7.469208789454838e-05, "loss": 0.0131, "step": 2079 }, { "epoch": 29.704845814977972, "grad_norm": 0.7100281119346619, "learning_rate": 7.460257532773401e-05, "loss": 0.0105, "step": 2080 }, { "epoch": 29.718942731277533, "grad_norm": 0.4377991557121277, "learning_rate": 7.45130845084446e-05, "loss": 0.0127, "step": 2081 }, { "epoch": 29.733039647577094, "grad_norm": 0.6396180391311646, "learning_rate": 7.442361551331014e-05, "loss": 0.0135, "step": 2082 }, { "epoch": 29.74713656387665, "grad_norm": 0.8097308874130249, "learning_rate": 7.433416841894188e-05, "loss": 0.0107, "step": 2083 }, { "epoch": 29.761233480176212, "grad_norm": 0.487700492143631, "learning_rate": 7.424474330193238e-05, "loss": 0.0111, "step": 2084 }, { "epoch": 29.77533039647577, "grad_norm": 0.5242835879325867, "learning_rate": 7.415534023885529e-05, "loss": 0.0134, "step": 2085 }, { "epoch": 29.78942731277533, "grad_norm": 1.1162769794464111, "learning_rate": 7.40659593062655e-05, "loss": 0.011, "step": 2086 }, { "epoch": 29.80352422907489, "grad_norm": 0.4597700536251068, "learning_rate": 7.397660058069883e-05, "loss": 0.0101, "step": 2087 }, { "epoch": 29.81762114537445, "grad_norm": 0.7440564036369324, "learning_rate": 7.388726413867218e-05, "loss": 0.0124, "step": 2088 }, { "epoch": 29.83171806167401, "grad_norm": 0.794187068939209, "learning_rate": 7.379795005668328e-05, "loss": 0.0126, "step": 2089 }, { "epoch": 29.845814977973568, "grad_norm": 0.9071127772331238, "learning_rate": 7.37086584112108e-05, "loss": 0.0124, "step": 2090 }, { "epoch": 29.85991189427313, "grad_norm": 0.5468192100524902, "learning_rate": 7.361938927871412e-05, "loss": 0.0101, "step": 2091 }, { "epoch": 29.874008810572686, "grad_norm": 0.6489362716674805, "learning_rate": 7.353014273563343e-05, "loss": 0.0102, "step": 2092 }, { "epoch": 29.888105726872247, "grad_norm": 0.5001371502876282, "learning_rate": 7.344091885838948e-05, "loss": 0.0104, "step": 2093 }, { "epoch": 29.902202643171805, "grad_norm": 0.9653443098068237, "learning_rate": 7.335171772338369e-05, "loss": 0.0147, "step": 2094 }, { "epoch": 29.916299559471366, "grad_norm": 0.7406577467918396, "learning_rate": 7.326253940699794e-05, "loss": 0.0113, "step": 2095 }, { "epoch": 29.930396475770927, "grad_norm": 0.516448974609375, "learning_rate": 7.317338398559464e-05, "loss": 0.0107, "step": 2096 }, { "epoch": 29.944493392070484, "grad_norm": 0.9239845871925354, "learning_rate": 7.308425153551653e-05, "loss": 0.0121, "step": 2097 }, { "epoch": 29.958590308370045, "grad_norm": 0.5725674033164978, "learning_rate": 7.299514213308674e-05, "loss": 0.013, "step": 2098 }, { "epoch": 29.972687224669603, "grad_norm": 0.683407187461853, "learning_rate": 7.290605585460863e-05, "loss": 0.0142, "step": 2099 }, { "epoch": 29.986784140969164, "grad_norm": 1.499961256980896, "learning_rate": 7.281699277636572e-05, "loss": 0.0113, "step": 2100 }, { "epoch": 30.01409691629956, "grad_norm": 0.30926793813705444, "learning_rate": 7.272795297462177e-05, "loss": 0.0234, "step": 2101 }, { "epoch": 30.02819383259912, "grad_norm": 0.5773308873176575, "learning_rate": 7.263893652562049e-05, "loss": 0.0108, "step": 2102 }, { "epoch": 30.04229074889868, "grad_norm": 0.9391916990280151, "learning_rate": 7.25499435055857e-05, "loss": 0.0115, "step": 2103 }, { "epoch": 30.056387665198237, "grad_norm": 0.3986680209636688, "learning_rate": 7.246097399072105e-05, "loss": 0.008, "step": 2104 }, { "epoch": 30.070484581497798, "grad_norm": 0.30742761492729187, "learning_rate": 7.237202805721019e-05, "loss": 0.0101, "step": 2105 }, { "epoch": 30.084581497797355, "grad_norm": 0.6178653240203857, "learning_rate": 7.228310578121644e-05, "loss": 0.0106, "step": 2106 }, { "epoch": 30.098678414096916, "grad_norm": 0.4633491635322571, "learning_rate": 7.2194207238883e-05, "loss": 0.0093, "step": 2107 }, { "epoch": 30.112775330396477, "grad_norm": 0.5819514393806458, "learning_rate": 7.210533250633261e-05, "loss": 0.009, "step": 2108 }, { "epoch": 30.126872246696035, "grad_norm": 1.5793830156326294, "learning_rate": 7.201648165966774e-05, "loss": 0.012, "step": 2109 }, { "epoch": 30.140969162995596, "grad_norm": 0.5351549983024597, "learning_rate": 7.192765477497032e-05, "loss": 0.0086, "step": 2110 }, { "epoch": 30.155066079295153, "grad_norm": 0.4899679720401764, "learning_rate": 7.183885192830184e-05, "loss": 0.0092, "step": 2111 }, { "epoch": 30.169162995594714, "grad_norm": 0.751639187335968, "learning_rate": 7.17500731957031e-05, "loss": 0.0163, "step": 2112 }, { "epoch": 30.18325991189427, "grad_norm": 0.5685061812400818, "learning_rate": 7.166131865319438e-05, "loss": 0.0094, "step": 2113 }, { "epoch": 30.197356828193833, "grad_norm": 0.4172273278236389, "learning_rate": 7.157258837677514e-05, "loss": 0.0102, "step": 2114 }, { "epoch": 30.211453744493394, "grad_norm": 0.40955209732055664, "learning_rate": 7.148388244242414e-05, "loss": 0.0097, "step": 2115 }, { "epoch": 30.22555066079295, "grad_norm": 0.494058758020401, "learning_rate": 7.139520092609925e-05, "loss": 0.0092, "step": 2116 }, { "epoch": 30.239647577092512, "grad_norm": 0.40899041295051575, "learning_rate": 7.130654390373738e-05, "loss": 0.01, "step": 2117 }, { "epoch": 30.25374449339207, "grad_norm": 0.8957332372665405, "learning_rate": 7.121791145125463e-05, "loss": 0.01, "step": 2118 }, { "epoch": 30.26784140969163, "grad_norm": 1.2424626350402832, "learning_rate": 7.112930364454586e-05, "loss": 0.0155, "step": 2119 }, { "epoch": 30.281938325991188, "grad_norm": 0.8591352701187134, "learning_rate": 7.1040720559485e-05, "loss": 0.0113, "step": 2120 }, { "epoch": 30.29603524229075, "grad_norm": 0.5112953186035156, "learning_rate": 7.095216227192467e-05, "loss": 0.0125, "step": 2121 }, { "epoch": 30.31013215859031, "grad_norm": 0.7700504660606384, "learning_rate": 7.086362885769639e-05, "loss": 0.0129, "step": 2122 }, { "epoch": 30.324229074889868, "grad_norm": 0.44225749373435974, "learning_rate": 7.077512039261025e-05, "loss": 0.0084, "step": 2123 }, { "epoch": 30.33832599118943, "grad_norm": 0.5659996867179871, "learning_rate": 7.068663695245507e-05, "loss": 0.0117, "step": 2124 }, { "epoch": 30.352422907488986, "grad_norm": 0.39974716305732727, "learning_rate": 7.05981786129982e-05, "loss": 0.0091, "step": 2125 }, { "epoch": 30.366519823788547, "grad_norm": 0.44108834862709045, "learning_rate": 7.050974544998554e-05, "loss": 0.0093, "step": 2126 }, { "epoch": 30.380616740088104, "grad_norm": 0.43166378140449524, "learning_rate": 7.042133753914133e-05, "loss": 0.0109, "step": 2127 }, { "epoch": 30.394713656387665, "grad_norm": 0.7482922077178955, "learning_rate": 7.033295495616834e-05, "loss": 0.0123, "step": 2128 }, { "epoch": 30.408810572687223, "grad_norm": 0.7291294932365417, "learning_rate": 7.024459777674748e-05, "loss": 0.0103, "step": 2129 }, { "epoch": 30.422907488986784, "grad_norm": 0.4640692472457886, "learning_rate": 7.015626607653806e-05, "loss": 0.0087, "step": 2130 }, { "epoch": 30.437004405286345, "grad_norm": 0.385465532541275, "learning_rate": 7.006795993117748e-05, "loss": 0.0096, "step": 2131 }, { "epoch": 30.451101321585902, "grad_norm": 0.3955899775028229, "learning_rate": 6.997967941628131e-05, "loss": 0.0142, "step": 2132 }, { "epoch": 30.465198237885463, "grad_norm": 0.6258742809295654, "learning_rate": 6.989142460744312e-05, "loss": 0.0106, "step": 2133 }, { "epoch": 30.47929515418502, "grad_norm": 0.4871881604194641, "learning_rate": 6.980319558023448e-05, "loss": 0.0089, "step": 2134 }, { "epoch": 30.493392070484582, "grad_norm": 0.269951194524765, "learning_rate": 6.971499241020495e-05, "loss": 0.0092, "step": 2135 }, { "epoch": 30.50748898678414, "grad_norm": 0.9138474464416504, "learning_rate": 6.962681517288185e-05, "loss": 0.0088, "step": 2136 }, { "epoch": 30.5215859030837, "grad_norm": 0.3875309228897095, "learning_rate": 6.953866394377038e-05, "loss": 0.0086, "step": 2137 }, { "epoch": 30.53568281938326, "grad_norm": 0.45838218927383423, "learning_rate": 6.945053879835338e-05, "loss": 0.0086, "step": 2138 }, { "epoch": 30.54977973568282, "grad_norm": 0.6737692952156067, "learning_rate": 6.936243981209147e-05, "loss": 0.0113, "step": 2139 }, { "epoch": 30.56387665198238, "grad_norm": 0.48621734976768494, "learning_rate": 6.927436706042276e-05, "loss": 0.0095, "step": 2140 }, { "epoch": 30.577973568281937, "grad_norm": 0.5678359270095825, "learning_rate": 6.9186320618763e-05, "loss": 0.0108, "step": 2141 }, { "epoch": 30.592070484581498, "grad_norm": 0.52802973985672, "learning_rate": 6.909830056250527e-05, "loss": 0.009, "step": 2142 }, { "epoch": 30.606167400881056, "grad_norm": 0.6237481832504272, "learning_rate": 6.90103069670202e-05, "loss": 0.0106, "step": 2143 }, { "epoch": 30.620264317180617, "grad_norm": 0.5042756199836731, "learning_rate": 6.892233990765568e-05, "loss": 0.0106, "step": 2144 }, { "epoch": 30.634361233480178, "grad_norm": 0.40919509530067444, "learning_rate": 6.883439945973692e-05, "loss": 0.0099, "step": 2145 }, { "epoch": 30.648458149779735, "grad_norm": 0.38325098156929016, "learning_rate": 6.874648569856628e-05, "loss": 0.0085, "step": 2146 }, { "epoch": 30.662555066079296, "grad_norm": 0.21222996711730957, "learning_rate": 6.865859869942338e-05, "loss": 0.0083, "step": 2147 }, { "epoch": 30.676651982378853, "grad_norm": 0.5824845433235168, "learning_rate": 6.857073853756477e-05, "loss": 0.0094, "step": 2148 }, { "epoch": 30.690748898678414, "grad_norm": 0.28489747643470764, "learning_rate": 6.848290528822416e-05, "loss": 0.0088, "step": 2149 }, { "epoch": 30.704845814977972, "grad_norm": 0.3869825303554535, "learning_rate": 6.839509902661216e-05, "loss": 0.0118, "step": 2150 }, { "epoch": 30.718942731277533, "grad_norm": 0.7261682152748108, "learning_rate": 6.830731982791622e-05, "loss": 0.01, "step": 2151 }, { "epoch": 30.733039647577094, "grad_norm": 0.45500192046165466, "learning_rate": 6.821956776730074e-05, "loss": 0.01, "step": 2152 }, { "epoch": 30.74713656387665, "grad_norm": 0.6906630396842957, "learning_rate": 6.813184291990673e-05, "loss": 0.0102, "step": 2153 }, { "epoch": 30.761233480176212, "grad_norm": 0.39501523971557617, "learning_rate": 6.804414536085205e-05, "loss": 0.008, "step": 2154 }, { "epoch": 30.77533039647577, "grad_norm": 0.2770364284515381, "learning_rate": 6.795647516523107e-05, "loss": 0.0119, "step": 2155 }, { "epoch": 30.78942731277533, "grad_norm": 0.706527590751648, "learning_rate": 6.786883240811479e-05, "loss": 0.0094, "step": 2156 }, { "epoch": 30.80352422907489, "grad_norm": 0.9397467970848083, "learning_rate": 6.77812171645507e-05, "loss": 0.0126, "step": 2157 }, { "epoch": 30.81762114537445, "grad_norm": 0.3226694166660309, "learning_rate": 6.769362950956277e-05, "loss": 0.0093, "step": 2158 }, { "epoch": 30.83171806167401, "grad_norm": 0.605153501033783, "learning_rate": 6.760606951815125e-05, "loss": 0.0102, "step": 2159 }, { "epoch": 30.845814977973568, "grad_norm": 0.25541308522224426, "learning_rate": 6.75185372652928e-05, "loss": 0.0117, "step": 2160 }, { "epoch": 30.85991189427313, "grad_norm": 1.0095592737197876, "learning_rate": 6.743103282594027e-05, "loss": 0.009, "step": 2161 }, { "epoch": 30.874008810572686, "grad_norm": 0.3733845055103302, "learning_rate": 6.734355627502273e-05, "loss": 0.0108, "step": 2162 }, { "epoch": 30.888105726872247, "grad_norm": 0.4820424020290375, "learning_rate": 6.725610768744534e-05, "loss": 0.0107, "step": 2163 }, { "epoch": 30.902202643171805, "grad_norm": 0.5326548218727112, "learning_rate": 6.716868713808935e-05, "loss": 0.01, "step": 2164 }, { "epoch": 30.916299559471366, "grad_norm": 0.4886656105518341, "learning_rate": 6.708129470181198e-05, "loss": 0.0092, "step": 2165 }, { "epoch": 30.930396475770927, "grad_norm": 0.5885226130485535, "learning_rate": 6.69939304534463e-05, "loss": 0.012, "step": 2166 }, { "epoch": 30.944493392070484, "grad_norm": 0.3820488154888153, "learning_rate": 6.690659446780141e-05, "loss": 0.0112, "step": 2167 }, { "epoch": 30.958590308370045, "grad_norm": 0.6876059174537659, "learning_rate": 6.681928681966208e-05, "loss": 0.01, "step": 2168 }, { "epoch": 30.972687224669603, "grad_norm": 0.5529156923294067, "learning_rate": 6.673200758378889e-05, "loss": 0.0151, "step": 2169 }, { "epoch": 30.986784140969164, "grad_norm": 0.5554307103157043, "learning_rate": 6.664475683491796e-05, "loss": 0.0107, "step": 2170 }, { "epoch": 31.01409691629956, "grad_norm": 0.7081966400146484, "learning_rate": 6.655753464776122e-05, "loss": 0.0167, "step": 2171 }, { "epoch": 31.02819383259912, "grad_norm": 0.506333589553833, "learning_rate": 6.647034109700597e-05, "loss": 0.0103, "step": 2172 }, { "epoch": 31.04229074889868, "grad_norm": 0.3593117594718933, "learning_rate": 6.638317625731508e-05, "loss": 0.0079, "step": 2173 }, { "epoch": 31.056387665198237, "grad_norm": 0.48380163311958313, "learning_rate": 6.629604020332674e-05, "loss": 0.008, "step": 2174 }, { "epoch": 31.070484581497798, "grad_norm": 0.587170422077179, "learning_rate": 6.620893300965464e-05, "loss": 0.0087, "step": 2175 }, { "epoch": 31.084581497797355, "grad_norm": 0.3809567391872406, "learning_rate": 6.612185475088769e-05, "loss": 0.0081, "step": 2176 }, { "epoch": 31.098678414096916, "grad_norm": 0.338241845369339, "learning_rate": 6.603480550158995e-05, "loss": 0.008, "step": 2177 }, { "epoch": 31.112775330396477, "grad_norm": 0.510821521282196, "learning_rate": 6.594778533630072e-05, "loss": 0.0106, "step": 2178 }, { "epoch": 31.126872246696035, "grad_norm": 0.40219980478286743, "learning_rate": 6.586079432953438e-05, "loss": 0.0092, "step": 2179 }, { "epoch": 31.140969162995596, "grad_norm": 0.3123362362384796, "learning_rate": 6.577383255578033e-05, "loss": 0.0086, "step": 2180 }, { "epoch": 31.155066079295153, "grad_norm": 0.4270247519016266, "learning_rate": 6.568690008950294e-05, "loss": 0.0075, "step": 2181 }, { "epoch": 31.169162995594714, "grad_norm": 0.19385765492916107, "learning_rate": 6.559999700514151e-05, "loss": 0.0094, "step": 2182 }, { "epoch": 31.18325991189427, "grad_norm": 0.8393190503120422, "learning_rate": 6.551312337711012e-05, "loss": 0.0113, "step": 2183 }, { "epoch": 31.197356828193833, "grad_norm": 0.4030812084674835, "learning_rate": 6.542627927979771e-05, "loss": 0.0078, "step": 2184 }, { "epoch": 31.211453744493394, "grad_norm": 0.4769592881202698, "learning_rate": 6.533946478756788e-05, "loss": 0.0087, "step": 2185 }, { "epoch": 31.22555066079295, "grad_norm": 0.27700427174568176, "learning_rate": 6.525267997475886e-05, "loss": 0.0079, "step": 2186 }, { "epoch": 31.239647577092512, "grad_norm": 0.6444690823554993, "learning_rate": 6.516592491568352e-05, "loss": 0.0076, "step": 2187 }, { "epoch": 31.25374449339207, "grad_norm": 0.4964103102684021, "learning_rate": 6.507919968462922e-05, "loss": 0.0087, "step": 2188 }, { "epoch": 31.26784140969163, "grad_norm": 0.2997181713581085, "learning_rate": 6.49925043558578e-05, "loss": 0.007, "step": 2189 }, { "epoch": 31.281938325991188, "grad_norm": 0.3995659351348877, "learning_rate": 6.490583900360543e-05, "loss": 0.0085, "step": 2190 }, { "epoch": 31.29603524229075, "grad_norm": 0.21125581860542297, "learning_rate": 6.481920370208274e-05, "loss": 0.0072, "step": 2191 }, { "epoch": 31.31013215859031, "grad_norm": 0.45320820808410645, "learning_rate": 6.473259852547449e-05, "loss": 0.0084, "step": 2192 }, { "epoch": 31.324229074889868, "grad_norm": 0.15627416968345642, "learning_rate": 6.464602354793974e-05, "loss": 0.0082, "step": 2193 }, { "epoch": 31.33832599118943, "grad_norm": 0.6243215799331665, "learning_rate": 6.455947884361162e-05, "loss": 0.0091, "step": 2194 }, { "epoch": 31.352422907488986, "grad_norm": 0.6323742270469666, "learning_rate": 6.447296448659743e-05, "loss": 0.0077, "step": 2195 }, { "epoch": 31.366519823788547, "grad_norm": 0.39468497037887573, "learning_rate": 6.438648055097835e-05, "loss": 0.0083, "step": 2196 }, { "epoch": 31.380616740088104, "grad_norm": 0.9159161448478699, "learning_rate": 6.430002711080965e-05, "loss": 0.0134, "step": 2197 }, { "epoch": 31.394713656387665, "grad_norm": 0.6601960062980652, "learning_rate": 6.42136042401204e-05, "loss": 0.0092, "step": 2198 }, { "epoch": 31.408810572687223, "grad_norm": 0.3376065492630005, "learning_rate": 6.412721201291352e-05, "loss": 0.009, "step": 2199 }, { "epoch": 31.422907488986784, "grad_norm": 0.47066497802734375, "learning_rate": 6.404085050316569e-05, "loss": 0.0102, "step": 2200 }, { "epoch": 31.437004405286345, "grad_norm": 0.8670686483383179, "learning_rate": 6.395451978482729e-05, "loss": 0.0093, "step": 2201 }, { "epoch": 31.451101321585902, "grad_norm": 0.5449432134628296, "learning_rate": 6.386821993182233e-05, "loss": 0.0082, "step": 2202 }, { "epoch": 31.465198237885463, "grad_norm": 0.39244958758354187, "learning_rate": 6.378195101804835e-05, "loss": 0.0092, "step": 2203 }, { "epoch": 31.47929515418502, "grad_norm": 0.3765074610710144, "learning_rate": 6.36957131173765e-05, "loss": 0.009, "step": 2204 }, { "epoch": 31.493392070484582, "grad_norm": 0.4850156307220459, "learning_rate": 6.360950630365126e-05, "loss": 0.0079, "step": 2205 }, { "epoch": 31.50748898678414, "grad_norm": 0.6798447370529175, "learning_rate": 6.352333065069057e-05, "loss": 0.0106, "step": 2206 }, { "epoch": 31.5215859030837, "grad_norm": 0.4406050443649292, "learning_rate": 6.343718623228562e-05, "loss": 0.009, "step": 2207 }, { "epoch": 31.53568281938326, "grad_norm": 0.46567195653915405, "learning_rate": 6.335107312220096e-05, "loss": 0.0086, "step": 2208 }, { "epoch": 31.54977973568282, "grad_norm": 0.40771421790122986, "learning_rate": 6.326499139417422e-05, "loss": 0.0097, "step": 2209 }, { "epoch": 31.56387665198238, "grad_norm": 0.4709036946296692, "learning_rate": 6.317894112191621e-05, "loss": 0.0111, "step": 2210 }, { "epoch": 31.577973568281937, "grad_norm": 0.41023072600364685, "learning_rate": 6.309292237911078e-05, "loss": 0.0095, "step": 2211 }, { "epoch": 31.592070484581498, "grad_norm": 0.4231497049331665, "learning_rate": 6.300693523941482e-05, "loss": 0.0078, "step": 2212 }, { "epoch": 31.606167400881056, "grad_norm": 0.39615747332572937, "learning_rate": 6.29209797764581e-05, "loss": 0.0076, "step": 2213 }, { "epoch": 31.620264317180617, "grad_norm": 0.534919261932373, "learning_rate": 6.283505606384336e-05, "loss": 0.0096, "step": 2214 }, { "epoch": 31.634361233480178, "grad_norm": 0.38475582003593445, "learning_rate": 6.274916417514604e-05, "loss": 0.0079, "step": 2215 }, { "epoch": 31.648458149779735, "grad_norm": 0.3158363103866577, "learning_rate": 6.266330418391445e-05, "loss": 0.0088, "step": 2216 }, { "epoch": 31.662555066079296, "grad_norm": 0.20669452846050262, "learning_rate": 6.257747616366947e-05, "loss": 0.0069, "step": 2217 }, { "epoch": 31.676651982378853, "grad_norm": 0.32730910181999207, "learning_rate": 6.249168018790464e-05, "loss": 0.0076, "step": 2218 }, { "epoch": 31.690748898678414, "grad_norm": 0.2967119514942169, "learning_rate": 6.24059163300861e-05, "loss": 0.0094, "step": 2219 }, { "epoch": 31.704845814977972, "grad_norm": 0.5402001738548279, "learning_rate": 6.232018466365247e-05, "loss": 0.008, "step": 2220 }, { "epoch": 31.718942731277533, "grad_norm": 0.5869316458702087, "learning_rate": 6.22344852620148e-05, "loss": 0.0099, "step": 2221 }, { "epoch": 31.733039647577094, "grad_norm": 0.8271963596343994, "learning_rate": 6.214881819855648e-05, "loss": 0.0104, "step": 2222 }, { "epoch": 31.74713656387665, "grad_norm": 0.49950969219207764, "learning_rate": 6.206318354663327e-05, "loss": 0.0123, "step": 2223 }, { "epoch": 31.761233480176212, "grad_norm": 0.3676856458187103, "learning_rate": 6.197758137957313e-05, "loss": 0.0078, "step": 2224 }, { "epoch": 31.77533039647577, "grad_norm": 0.8804897665977478, "learning_rate": 6.189201177067623e-05, "loss": 0.0095, "step": 2225 }, { "epoch": 31.78942731277533, "grad_norm": 0.35559114813804626, "learning_rate": 6.180647479321485e-05, "loss": 0.0086, "step": 2226 }, { "epoch": 31.80352422907489, "grad_norm": 0.5079904198646545, "learning_rate": 6.17209705204333e-05, "loss": 0.009, "step": 2227 }, { "epoch": 31.81762114537445, "grad_norm": 0.3441047668457031, "learning_rate": 6.163549902554794e-05, "loss": 0.009, "step": 2228 }, { "epoch": 31.83171806167401, "grad_norm": 0.27998587489128113, "learning_rate": 6.155006038174704e-05, "loss": 0.0083, "step": 2229 }, { "epoch": 31.845814977973568, "grad_norm": 0.7784952521324158, "learning_rate": 6.146465466219072e-05, "loss": 0.0085, "step": 2230 }, { "epoch": 31.85991189427313, "grad_norm": 0.5978696346282959, "learning_rate": 6.137928194001093e-05, "loss": 0.009, "step": 2231 }, { "epoch": 31.874008810572686, "grad_norm": 0.5129473805427551, "learning_rate": 6.129394228831134e-05, "loss": 0.0107, "step": 2232 }, { "epoch": 31.888105726872247, "grad_norm": 0.6546720862388611, "learning_rate": 6.120863578016735e-05, "loss": 0.0091, "step": 2233 }, { "epoch": 31.902202643171805, "grad_norm": 0.43558427691459656, "learning_rate": 6.112336248862595e-05, "loss": 0.0094, "step": 2234 }, { "epoch": 31.916299559471366, "grad_norm": 0.2918238341808319, "learning_rate": 6.103812248670567e-05, "loss": 0.009, "step": 2235 }, { "epoch": 31.930396475770927, "grad_norm": 0.5220287442207336, "learning_rate": 6.095291584739654e-05, "loss": 0.0099, "step": 2236 }, { "epoch": 31.944493392070484, "grad_norm": 0.3720172941684723, "learning_rate": 6.086774264366004e-05, "loss": 0.0094, "step": 2237 }, { "epoch": 31.958590308370045, "grad_norm": 0.6271647214889526, "learning_rate": 6.0782602948429034e-05, "loss": 0.0151, "step": 2238 }, { "epoch": 31.972687224669603, "grad_norm": 0.8465790748596191, "learning_rate": 6.069749683460765e-05, "loss": 0.008, "step": 2239 }, { "epoch": 31.986784140969164, "grad_norm": 0.5729878544807434, "learning_rate": 6.061242437507131e-05, "loss": 0.0101, "step": 2240 }, { "epoch": 32.01409691629956, "grad_norm": 0.28472837805747986, "learning_rate": 6.0527385642666535e-05, "loss": 0.0145, "step": 2241 }, { "epoch": 32.02819383259912, "grad_norm": 0.31237155199050903, "learning_rate": 6.044238071021108e-05, "loss": 0.0075, "step": 2242 }, { "epoch": 32.04229074889868, "grad_norm": 0.34607720375061035, "learning_rate": 6.035740965049366e-05, "loss": 0.0078, "step": 2243 }, { "epoch": 32.05638766519824, "grad_norm": 0.617678701877594, "learning_rate": 6.027247253627406e-05, "loss": 0.0077, "step": 2244 }, { "epoch": 32.070484581497794, "grad_norm": 0.673148512840271, "learning_rate": 6.018756944028291e-05, "loss": 0.0072, "step": 2245 }, { "epoch": 32.08458149779736, "grad_norm": 0.24490214884281158, "learning_rate": 6.0102700435221815e-05, "loss": 0.0084, "step": 2246 }, { "epoch": 32.098678414096916, "grad_norm": 0.2714008688926697, "learning_rate": 6.00178655937631e-05, "loss": 0.0078, "step": 2247 }, { "epoch": 32.112775330396474, "grad_norm": 0.33235839009284973, "learning_rate": 5.993306498854991e-05, "loss": 0.0088, "step": 2248 }, { "epoch": 32.12687224669604, "grad_norm": 0.7460521459579468, "learning_rate": 5.9848298692195984e-05, "loss": 0.0076, "step": 2249 }, { "epoch": 32.140969162995596, "grad_norm": 0.43327027559280396, "learning_rate": 5.9763566777285786e-05, "loss": 0.0087, "step": 2250 }, { "epoch": 32.140969162995596, "eval_loss": 1.4188281297683716, "eval_runtime": 132.8031, "eval_samples_per_second": 13.614, "eval_steps_per_second": 0.429, "step": 2250 }, { "epoch": 32.15506607929515, "grad_norm": 0.2786136865615845, "learning_rate": 5.967886931637425e-05, "loss": 0.0074, "step": 2251 }, { "epoch": 32.16916299559471, "grad_norm": 0.5824514627456665, "learning_rate": 5.959420638198685e-05, "loss": 0.0089, "step": 2252 }, { "epoch": 32.183259911894275, "grad_norm": 0.32038435339927673, "learning_rate": 5.950957804661952e-05, "loss": 0.0079, "step": 2253 }, { "epoch": 32.19735682819383, "grad_norm": 0.32977545261383057, "learning_rate": 5.942498438273849e-05, "loss": 0.0073, "step": 2254 }, { "epoch": 32.21145374449339, "grad_norm": 0.5877240300178528, "learning_rate": 5.93404254627804e-05, "loss": 0.0073, "step": 2255 }, { "epoch": 32.225550660792955, "grad_norm": 0.3316883146762848, "learning_rate": 5.9255901359152045e-05, "loss": 0.0092, "step": 2256 }, { "epoch": 32.23964757709251, "grad_norm": 0.5813613533973694, "learning_rate": 5.91714121442305e-05, "loss": 0.0104, "step": 2257 }, { "epoch": 32.25374449339207, "grad_norm": 0.47089502215385437, "learning_rate": 5.908695789036285e-05, "loss": 0.008, "step": 2258 }, { "epoch": 32.26784140969163, "grad_norm": 0.35702118277549744, "learning_rate": 5.900253866986638e-05, "loss": 0.0083, "step": 2259 }, { "epoch": 32.28193832599119, "grad_norm": 0.34889763593673706, "learning_rate": 5.891815455502824e-05, "loss": 0.0078, "step": 2260 }, { "epoch": 32.29603524229075, "grad_norm": 0.3289335072040558, "learning_rate": 5.883380561810563e-05, "loss": 0.0078, "step": 2261 }, { "epoch": 32.31013215859031, "grad_norm": 0.6906613707542419, "learning_rate": 5.874949193132556e-05, "loss": 0.0111, "step": 2262 }, { "epoch": 32.32422907488987, "grad_norm": 0.48297712206840515, "learning_rate": 5.86652135668849e-05, "loss": 0.0091, "step": 2263 }, { "epoch": 32.33832599118943, "grad_norm": 0.5515782833099365, "learning_rate": 5.858097059695022e-05, "loss": 0.0084, "step": 2264 }, { "epoch": 32.352422907488986, "grad_norm": 0.34508517384529114, "learning_rate": 5.849676309365786e-05, "loss": 0.007, "step": 2265 }, { "epoch": 32.36651982378854, "grad_norm": 0.32172590494155884, "learning_rate": 5.841259112911368e-05, "loss": 0.0088, "step": 2266 }, { "epoch": 32.38061674008811, "grad_norm": 0.29618191719055176, "learning_rate": 5.832845477539326e-05, "loss": 0.0081, "step": 2267 }, { "epoch": 32.394713656387665, "grad_norm": 0.3051191568374634, "learning_rate": 5.82443541045415e-05, "loss": 0.0081, "step": 2268 }, { "epoch": 32.40881057268722, "grad_norm": 0.8932328820228577, "learning_rate": 5.816028918857288e-05, "loss": 0.0074, "step": 2269 }, { "epoch": 32.42290748898679, "grad_norm": 0.266460120677948, "learning_rate": 5.807626009947127e-05, "loss": 0.0076, "step": 2270 }, { "epoch": 32.437004405286345, "grad_norm": 0.2470874786376953, "learning_rate": 5.799226690918982e-05, "loss": 0.0066, "step": 2271 }, { "epoch": 32.4511013215859, "grad_norm": 0.451489120721817, "learning_rate": 5.7908309689650865e-05, "loss": 0.0078, "step": 2272 }, { "epoch": 32.46519823788546, "grad_norm": 0.5118008852005005, "learning_rate": 5.7824388512746064e-05, "loss": 0.0091, "step": 2273 }, { "epoch": 32.479295154185024, "grad_norm": 0.34958788752555847, "learning_rate": 5.774050345033617e-05, "loss": 0.0081, "step": 2274 }, { "epoch": 32.49339207048458, "grad_norm": 0.36525285243988037, "learning_rate": 5.765665457425102e-05, "loss": 0.009, "step": 2275 }, { "epoch": 32.50748898678414, "grad_norm": 0.4388152062892914, "learning_rate": 5.7572841956289394e-05, "loss": 0.0092, "step": 2276 }, { "epoch": 32.521585903083704, "grad_norm": 0.4884622097015381, "learning_rate": 5.748906566821911e-05, "loss": 0.0087, "step": 2277 }, { "epoch": 32.53568281938326, "grad_norm": 0.2257160097360611, "learning_rate": 5.7405325781776844e-05, "loss": 0.0073, "step": 2278 }, { "epoch": 32.54977973568282, "grad_norm": 0.21137380599975586, "learning_rate": 5.732162236866816e-05, "loss": 0.0068, "step": 2279 }, { "epoch": 32.563876651982376, "grad_norm": 0.3549029231071472, "learning_rate": 5.7237955500567205e-05, "loss": 0.0084, "step": 2280 }, { "epoch": 32.57797356828194, "grad_norm": 0.4107114374637604, "learning_rate": 5.7154325249117077e-05, "loss": 0.0093, "step": 2281 }, { "epoch": 32.5920704845815, "grad_norm": 0.15383535623550415, "learning_rate": 5.707073168592942e-05, "loss": 0.0066, "step": 2282 }, { "epoch": 32.606167400881056, "grad_norm": 0.25406789779663086, "learning_rate": 5.698717488258436e-05, "loss": 0.0066, "step": 2283 }, { "epoch": 32.62026431718062, "grad_norm": 0.7672394514083862, "learning_rate": 5.690365491063068e-05, "loss": 0.0091, "step": 2284 }, { "epoch": 32.63436123348018, "grad_norm": 0.5589924454689026, "learning_rate": 5.682017184158561e-05, "loss": 0.0097, "step": 2285 }, { "epoch": 32.648458149779735, "grad_norm": 0.398880273103714, "learning_rate": 5.673672574693467e-05, "loss": 0.009, "step": 2286 }, { "epoch": 32.66255506607929, "grad_norm": 0.4043208956718445, "learning_rate": 5.665331669813185e-05, "loss": 0.0094, "step": 2287 }, { "epoch": 32.67665198237886, "grad_norm": 0.3847372233867645, "learning_rate": 5.656994476659938e-05, "loss": 0.0089, "step": 2288 }, { "epoch": 32.690748898678414, "grad_norm": 0.45945361256599426, "learning_rate": 5.648661002372768e-05, "loss": 0.0088, "step": 2289 }, { "epoch": 32.70484581497797, "grad_norm": 0.24741578102111816, "learning_rate": 5.6403312540875316e-05, "loss": 0.0078, "step": 2290 }, { "epoch": 32.71894273127754, "grad_norm": 0.3573097586631775, "learning_rate": 5.6320052389368996e-05, "loss": 0.0085, "step": 2291 }, { "epoch": 32.733039647577094, "grad_norm": 0.53316330909729, "learning_rate": 5.623682964050341e-05, "loss": 0.0097, "step": 2292 }, { "epoch": 32.74713656387665, "grad_norm": 0.2028275728225708, "learning_rate": 5.6153644365541315e-05, "loss": 0.0071, "step": 2293 }, { "epoch": 32.76123348017621, "grad_norm": 0.3078833222389221, "learning_rate": 5.607049663571323e-05, "loss": 0.0087, "step": 2294 }, { "epoch": 32.77533039647577, "grad_norm": 0.5933877825737, "learning_rate": 5.5987386522217645e-05, "loss": 0.0112, "step": 2295 }, { "epoch": 32.78942731277533, "grad_norm": 0.43832316994667053, "learning_rate": 5.590431409622081e-05, "loss": 0.0085, "step": 2296 }, { "epoch": 32.80352422907489, "grad_norm": 0.47427651286125183, "learning_rate": 5.582127942885672e-05, "loss": 0.0072, "step": 2297 }, { "epoch": 32.817621145374446, "grad_norm": 0.38293686509132385, "learning_rate": 5.573828259122694e-05, "loss": 0.0112, "step": 2298 }, { "epoch": 32.83171806167401, "grad_norm": 0.5639181137084961, "learning_rate": 5.565532365440076e-05, "loss": 0.0097, "step": 2299 }, { "epoch": 32.84581497797357, "grad_norm": 0.2798483967781067, "learning_rate": 5.557240268941497e-05, "loss": 0.0084, "step": 2300 }, { "epoch": 32.859911894273125, "grad_norm": 0.34482237696647644, "learning_rate": 5.548951976727388e-05, "loss": 0.0097, "step": 2301 }, { "epoch": 32.87400881057269, "grad_norm": 0.7201386094093323, "learning_rate": 5.54066749589491e-05, "loss": 0.0083, "step": 2302 }, { "epoch": 32.88810572687225, "grad_norm": 0.4203905463218689, "learning_rate": 5.532386833537977e-05, "loss": 0.0104, "step": 2303 }, { "epoch": 32.902202643171805, "grad_norm": 0.2698669731616974, "learning_rate": 5.5241099967472256e-05, "loss": 0.0082, "step": 2304 }, { "epoch": 32.91629955947137, "grad_norm": 0.6667635440826416, "learning_rate": 5.5158369926100104e-05, "loss": 0.011, "step": 2305 }, { "epoch": 32.93039647577093, "grad_norm": 0.4234503209590912, "learning_rate": 5.507567828210416e-05, "loss": 0.0097, "step": 2306 }, { "epoch": 32.944493392070484, "grad_norm": 0.45322567224502563, "learning_rate": 5.4993025106292317e-05, "loss": 0.0079, "step": 2307 }, { "epoch": 32.95859030837004, "grad_norm": 0.43554529547691345, "learning_rate": 5.491041046943959e-05, "loss": 0.0084, "step": 2308 }, { "epoch": 32.972687224669606, "grad_norm": 0.4298781454563141, "learning_rate": 5.482783444228787e-05, "loss": 0.0074, "step": 2309 }, { "epoch": 32.986784140969164, "grad_norm": 0.3019631505012512, "learning_rate": 5.474529709554612e-05, "loss": 0.0081, "step": 2310 }, { "epoch": 33.01409691629956, "grad_norm": 0.3561325669288635, "learning_rate": 5.46627984998901e-05, "loss": 0.0166, "step": 2311 }, { "epoch": 33.02819383259912, "grad_norm": 0.31866735219955444, "learning_rate": 5.4580338725962476e-05, "loss": 0.0065, "step": 2312 }, { "epoch": 33.04229074889868, "grad_norm": 0.22215214371681213, "learning_rate": 5.4497917844372524e-05, "loss": 0.0075, "step": 2313 }, { "epoch": 33.05638766519824, "grad_norm": 0.31289801001548767, "learning_rate": 5.441553592569636e-05, "loss": 0.0104, "step": 2314 }, { "epoch": 33.070484581497794, "grad_norm": 0.75337153673172, "learning_rate": 5.433319304047666e-05, "loss": 0.0067, "step": 2315 }, { "epoch": 33.08458149779736, "grad_norm": 0.2799716889858246, "learning_rate": 5.425088925922275e-05, "loss": 0.0074, "step": 2316 }, { "epoch": 33.098678414096916, "grad_norm": 0.38561996817588806, "learning_rate": 5.416862465241033e-05, "loss": 0.0079, "step": 2317 }, { "epoch": 33.112775330396474, "grad_norm": 0.29921817779541016, "learning_rate": 5.408639929048169e-05, "loss": 0.0071, "step": 2318 }, { "epoch": 33.12687224669604, "grad_norm": 0.3070436418056488, "learning_rate": 5.4004213243845514e-05, "loss": 0.0093, "step": 2319 }, { "epoch": 33.140969162995596, "grad_norm": 0.8990926742553711, "learning_rate": 5.392206658287669e-05, "loss": 0.0069, "step": 2320 }, { "epoch": 33.15506607929515, "grad_norm": 0.2669459283351898, "learning_rate": 5.38399593779165e-05, "loss": 0.0066, "step": 2321 }, { "epoch": 33.16916299559471, "grad_norm": 0.28642281889915466, "learning_rate": 5.375789169927241e-05, "loss": 0.0066, "step": 2322 }, { "epoch": 33.183259911894275, "grad_norm": 0.17192259430885315, "learning_rate": 5.36758636172181e-05, "loss": 0.0072, "step": 2323 }, { "epoch": 33.19735682819383, "grad_norm": 0.37343528866767883, "learning_rate": 5.3593875201993174e-05, "loss": 0.0092, "step": 2324 }, { "epoch": 33.21145374449339, "grad_norm": 0.5126631855964661, "learning_rate": 5.3511926523803416e-05, "loss": 0.0068, "step": 2325 }, { "epoch": 33.225550660792955, "grad_norm": 0.1369171142578125, "learning_rate": 5.343001765282054e-05, "loss": 0.0074, "step": 2326 }, { "epoch": 33.23964757709251, "grad_norm": 0.282227486371994, "learning_rate": 5.334814865918225e-05, "loss": 0.0086, "step": 2327 }, { "epoch": 33.25374449339207, "grad_norm": 0.4775007367134094, "learning_rate": 5.326631961299193e-05, "loss": 0.0079, "step": 2328 }, { "epoch": 33.26784140969163, "grad_norm": 0.40840616822242737, "learning_rate": 5.318453058431889e-05, "loss": 0.0067, "step": 2329 }, { "epoch": 33.28193832599119, "grad_norm": 0.24185921251773834, "learning_rate": 5.310278164319815e-05, "loss": 0.0074, "step": 2330 }, { "epoch": 33.29603524229075, "grad_norm": 0.25163736939430237, "learning_rate": 5.302107285963045e-05, "loss": 0.0066, "step": 2331 }, { "epoch": 33.31013215859031, "grad_norm": 0.1917894184589386, "learning_rate": 5.2939404303581996e-05, "loss": 0.0077, "step": 2332 }, { "epoch": 33.32422907488987, "grad_norm": 0.26810863614082336, "learning_rate": 5.285777604498466e-05, "loss": 0.0077, "step": 2333 }, { "epoch": 33.33832599118943, "grad_norm": 0.2958584725856781, "learning_rate": 5.277618815373583e-05, "loss": 0.0064, "step": 2334 }, { "epoch": 33.352422907488986, "grad_norm": 0.15407666563987732, "learning_rate": 5.269464069969829e-05, "loss": 0.0078, "step": 2335 }, { "epoch": 33.36651982378854, "grad_norm": 0.5291561484336853, "learning_rate": 5.261313375270014e-05, "loss": 0.0067, "step": 2336 }, { "epoch": 33.38061674008811, "grad_norm": 0.406332403421402, "learning_rate": 5.253166738253488e-05, "loss": 0.0079, "step": 2337 }, { "epoch": 33.394713656387665, "grad_norm": 0.3468363881111145, "learning_rate": 5.245024165896126e-05, "loss": 0.0069, "step": 2338 }, { "epoch": 33.40881057268722, "grad_norm": 0.29977595806121826, "learning_rate": 5.2368856651703144e-05, "loss": 0.0078, "step": 2339 }, { "epoch": 33.42290748898679, "grad_norm": 0.2530134618282318, "learning_rate": 5.2287512430449604e-05, "loss": 0.0082, "step": 2340 }, { "epoch": 33.437004405286345, "grad_norm": 0.3649113178253174, "learning_rate": 5.2206209064854785e-05, "loss": 0.0067, "step": 2341 }, { "epoch": 33.4511013215859, "grad_norm": 0.25974854826927185, "learning_rate": 5.212494662453784e-05, "loss": 0.0094, "step": 2342 }, { "epoch": 33.46519823788546, "grad_norm": 0.30863693356513977, "learning_rate": 5.204372517908282e-05, "loss": 0.0088, "step": 2343 }, { "epoch": 33.479295154185024, "grad_norm": 0.5945245623588562, "learning_rate": 5.196254479803877e-05, "loss": 0.0078, "step": 2344 }, { "epoch": 33.49339207048458, "grad_norm": 0.4528220295906067, "learning_rate": 5.1881405550919493e-05, "loss": 0.0119, "step": 2345 }, { "epoch": 33.50748898678414, "grad_norm": 0.5403532981872559, "learning_rate": 5.180030750720366e-05, "loss": 0.0071, "step": 2346 }, { "epoch": 33.521585903083704, "grad_norm": 0.24517375230789185, "learning_rate": 5.1719250736334504e-05, "loss": 0.0074, "step": 2347 }, { "epoch": 33.53568281938326, "grad_norm": 0.3140014111995697, "learning_rate": 5.1638235307720075e-05, "loss": 0.0086, "step": 2348 }, { "epoch": 33.54977973568282, "grad_norm": 0.43764039874076843, "learning_rate": 5.155726129073295e-05, "loss": 0.0068, "step": 2349 }, { "epoch": 33.563876651982376, "grad_norm": 0.21823441982269287, "learning_rate": 5.147632875471029e-05, "loss": 0.0081, "step": 2350 }, { "epoch": 33.57797356828194, "grad_norm": 0.2743070721626282, "learning_rate": 5.1395437768953634e-05, "loss": 0.0091, "step": 2351 }, { "epoch": 33.5920704845815, "grad_norm": 0.3084612786769867, "learning_rate": 5.1314588402729044e-05, "loss": 0.0063, "step": 2352 }, { "epoch": 33.606167400881056, "grad_norm": 0.2816929817199707, "learning_rate": 5.123378072526694e-05, "loss": 0.008, "step": 2353 }, { "epoch": 33.62026431718062, "grad_norm": 0.29744067788124084, "learning_rate": 5.115301480576196e-05, "loss": 0.0078, "step": 2354 }, { "epoch": 33.63436123348018, "grad_norm": 0.20973902940750122, "learning_rate": 5.107229071337305e-05, "loss": 0.0082, "step": 2355 }, { "epoch": 33.648458149779735, "grad_norm": 0.28197726607322693, "learning_rate": 5.099160851722333e-05, "loss": 0.0083, "step": 2356 }, { "epoch": 33.66255506607929, "grad_norm": 0.4126361608505249, "learning_rate": 5.0910968286400115e-05, "loss": 0.0083, "step": 2357 }, { "epoch": 33.67665198237886, "grad_norm": 0.23037216067314148, "learning_rate": 5.083037008995461e-05, "loss": 0.0084, "step": 2358 }, { "epoch": 33.690748898678414, "grad_norm": 0.26309987902641296, "learning_rate": 5.074981399690218e-05, "loss": 0.008, "step": 2359 }, { "epoch": 33.70484581497797, "grad_norm": 0.38994014263153076, "learning_rate": 5.0669300076222096e-05, "loss": 0.0084, "step": 2360 }, { "epoch": 33.71894273127754, "grad_norm": 0.32731199264526367, "learning_rate": 5.0588828396857524e-05, "loss": 0.0082, "step": 2361 }, { "epoch": 33.733039647577094, "grad_norm": 0.383971631526947, "learning_rate": 5.050839902771538e-05, "loss": 0.0081, "step": 2362 }, { "epoch": 33.74713656387665, "grad_norm": 0.30301469564437866, "learning_rate": 5.042801203766645e-05, "loss": 0.0099, "step": 2363 }, { "epoch": 33.76123348017621, "grad_norm": 0.4713279902935028, "learning_rate": 5.0347667495545224e-05, "loss": 0.0093, "step": 2364 }, { "epoch": 33.77533039647577, "grad_norm": 0.37209510803222656, "learning_rate": 5.026736547014981e-05, "loss": 0.007, "step": 2365 }, { "epoch": 33.78942731277533, "grad_norm": 0.3752252161502838, "learning_rate": 5.018710603024187e-05, "loss": 0.0077, "step": 2366 }, { "epoch": 33.80352422907489, "grad_norm": 0.2727470099925995, "learning_rate": 5.0106889244546675e-05, "loss": 0.0089, "step": 2367 }, { "epoch": 33.817621145374446, "grad_norm": 0.3495565354824066, "learning_rate": 5.002671518175299e-05, "loss": 0.0081, "step": 2368 }, { "epoch": 33.83171806167401, "grad_norm": 0.6847230195999146, "learning_rate": 4.9946583910512856e-05, "loss": 0.0083, "step": 2369 }, { "epoch": 33.84581497797357, "grad_norm": 0.43738606572151184, "learning_rate": 4.9866495499441824e-05, "loss": 0.0088, "step": 2370 }, { "epoch": 33.859911894273125, "grad_norm": 0.3742659389972687, "learning_rate": 4.978645001711867e-05, "loss": 0.0082, "step": 2371 }, { "epoch": 33.87400881057269, "grad_norm": 0.46508172154426575, "learning_rate": 4.9706447532085475e-05, "loss": 0.0077, "step": 2372 }, { "epoch": 33.88810572687225, "grad_norm": 0.4487709403038025, "learning_rate": 4.962648811284738e-05, "loss": 0.0069, "step": 2373 }, { "epoch": 33.902202643171805, "grad_norm": 0.31101375818252563, "learning_rate": 4.9546571827872755e-05, "loss": 0.0077, "step": 2374 }, { "epoch": 33.91629955947137, "grad_norm": 0.36904633045196533, "learning_rate": 4.946669874559301e-05, "loss": 0.0076, "step": 2375 }, { "epoch": 33.93039647577093, "grad_norm": 0.25134387612342834, "learning_rate": 4.9386868934402585e-05, "loss": 0.0091, "step": 2376 }, { "epoch": 33.944493392070484, "grad_norm": 0.40041449666023254, "learning_rate": 4.930708246265877e-05, "loss": 0.0085, "step": 2377 }, { "epoch": 33.95859030837004, "grad_norm": 0.454822301864624, "learning_rate": 4.9227339398681836e-05, "loss": 0.0075, "step": 2378 }, { "epoch": 33.972687224669606, "grad_norm": 0.282761812210083, "learning_rate": 4.914763981075486e-05, "loss": 0.0081, "step": 2379 }, { "epoch": 33.986784140969164, "grad_norm": 0.30034783482551575, "learning_rate": 4.9067983767123736e-05, "loss": 0.0066, "step": 2380 }, { "epoch": 34.01409691629956, "grad_norm": 0.33784019947052, "learning_rate": 4.8988371335996944e-05, "loss": 0.0133, "step": 2381 }, { "epoch": 34.02819383259912, "grad_norm": 0.3594037890434265, "learning_rate": 4.8908802585545745e-05, "loss": 0.0069, "step": 2382 }, { "epoch": 34.04229074889868, "grad_norm": 0.20853222906589508, "learning_rate": 4.8829277583903945e-05, "loss": 0.0071, "step": 2383 }, { "epoch": 34.05638766519824, "grad_norm": 0.6362195014953613, "learning_rate": 4.874979639916792e-05, "loss": 0.0082, "step": 2384 }, { "epoch": 34.070484581497794, "grad_norm": 0.21263565123081207, "learning_rate": 4.8670359099396466e-05, "loss": 0.0067, "step": 2385 }, { "epoch": 34.08458149779736, "grad_norm": 0.1975855976343155, "learning_rate": 4.859096575261084e-05, "loss": 0.0061, "step": 2386 }, { "epoch": 34.098678414096916, "grad_norm": 0.18415355682373047, "learning_rate": 4.851161642679466e-05, "loss": 0.007, "step": 2387 }, { "epoch": 34.112775330396474, "grad_norm": 0.31137189269065857, "learning_rate": 4.843231118989387e-05, "loss": 0.0071, "step": 2388 }, { "epoch": 34.12687224669604, "grad_norm": 0.2663591504096985, "learning_rate": 4.835305010981662e-05, "loss": 0.0064, "step": 2389 }, { "epoch": 34.140969162995596, "grad_norm": 0.14970038831233978, "learning_rate": 4.827383325443331e-05, "loss": 0.0072, "step": 2390 }, { "epoch": 34.15506607929515, "grad_norm": 0.2836495637893677, "learning_rate": 4.8194660691576365e-05, "loss": 0.0065, "step": 2391 }, { "epoch": 34.16916299559471, "grad_norm": 0.2319047451019287, "learning_rate": 4.811553248904037e-05, "loss": 0.0067, "step": 2392 }, { "epoch": 34.183259911894275, "grad_norm": 0.153662770986557, "learning_rate": 4.803644871458191e-05, "loss": 0.0075, "step": 2393 }, { "epoch": 34.19735682819383, "grad_norm": 0.3402034342288971, "learning_rate": 4.795740943591955e-05, "loss": 0.0072, "step": 2394 }, { "epoch": 34.21145374449339, "grad_norm": 0.33366212248802185, "learning_rate": 4.7878414720733664e-05, "loss": 0.0063, "step": 2395 }, { "epoch": 34.225550660792955, "grad_norm": 0.32169997692108154, "learning_rate": 4.7799464636666545e-05, "loss": 0.0098, "step": 2396 }, { "epoch": 34.23964757709251, "grad_norm": 1.0971580743789673, "learning_rate": 4.772055925132225e-05, "loss": 0.0074, "step": 2397 }, { "epoch": 34.25374449339207, "grad_norm": 0.3202718496322632, "learning_rate": 4.764169863226661e-05, "loss": 0.0068, "step": 2398 }, { "epoch": 34.26784140969163, "grad_norm": 0.2692412734031677, "learning_rate": 4.756288284702699e-05, "loss": 0.0092, "step": 2399 }, { "epoch": 34.28193832599119, "grad_norm": 0.44878625869750977, "learning_rate": 4.748411196309248e-05, "loss": 0.0065, "step": 2400 }, { "epoch": 34.29603524229075, "grad_norm": 0.19396793842315674, "learning_rate": 4.74053860479137e-05, "loss": 0.0065, "step": 2401 }, { "epoch": 34.31013215859031, "grad_norm": 0.2316325604915619, "learning_rate": 4.7326705168902783e-05, "loss": 0.0069, "step": 2402 }, { "epoch": 34.32422907488987, "grad_norm": 0.279545396566391, "learning_rate": 4.7248069393433205e-05, "loss": 0.0074, "step": 2403 }, { "epoch": 34.33832599118943, "grad_norm": 0.24416495859622955, "learning_rate": 4.71694787888399e-05, "loss": 0.0061, "step": 2404 }, { "epoch": 34.352422907488986, "grad_norm": 0.25578126311302185, "learning_rate": 4.709093342241917e-05, "loss": 0.0084, "step": 2405 }, { "epoch": 34.36651982378854, "grad_norm": 0.37564921379089355, "learning_rate": 4.701243336142841e-05, "loss": 0.0068, "step": 2406 }, { "epoch": 34.38061674008811, "grad_norm": 0.30344006419181824, "learning_rate": 4.6933978673086385e-05, "loss": 0.0066, "step": 2407 }, { "epoch": 34.394713656387665, "grad_norm": 0.5353860259056091, "learning_rate": 4.6855569424572955e-05, "loss": 0.0065, "step": 2408 }, { "epoch": 34.40881057268722, "grad_norm": 0.2710483968257904, "learning_rate": 4.6777205683029076e-05, "loss": 0.0072, "step": 2409 }, { "epoch": 34.42290748898679, "grad_norm": 0.2805093228816986, "learning_rate": 4.6698887515556685e-05, "loss": 0.0076, "step": 2410 }, { "epoch": 34.437004405286345, "grad_norm": 0.4146285355091095, "learning_rate": 4.662061498921875e-05, "loss": 0.0091, "step": 2411 }, { "epoch": 34.4511013215859, "grad_norm": 0.5397526621818542, "learning_rate": 4.654238817103913e-05, "loss": 0.0068, "step": 2412 }, { "epoch": 34.46519823788546, "grad_norm": 0.17969098687171936, "learning_rate": 4.6464207128002623e-05, "loss": 0.0067, "step": 2413 }, { "epoch": 34.479295154185024, "grad_norm": 0.2019590586423874, "learning_rate": 4.6386071927054684e-05, "loss": 0.0064, "step": 2414 }, { "epoch": 34.49339207048458, "grad_norm": 0.33136430382728577, "learning_rate": 4.630798263510162e-05, "loss": 0.0063, "step": 2415 }, { "epoch": 34.50748898678414, "grad_norm": 0.15759414434432983, "learning_rate": 4.62299393190104e-05, "loss": 0.0069, "step": 2416 }, { "epoch": 34.521585903083704, "grad_norm": 0.25696730613708496, "learning_rate": 4.6151942045608675e-05, "loss": 0.0066, "step": 2417 }, { "epoch": 34.53568281938326, "grad_norm": 0.24530108273029327, "learning_rate": 4.6073990881684536e-05, "loss": 0.0066, "step": 2418 }, { "epoch": 34.54977973568282, "grad_norm": 0.18404783308506012, "learning_rate": 4.5996085893986705e-05, "loss": 0.0069, "step": 2419 }, { "epoch": 34.563876651982376, "grad_norm": 0.6750898957252502, "learning_rate": 4.59182271492244e-05, "loss": 0.007, "step": 2420 }, { "epoch": 34.57797356828194, "grad_norm": 0.17648589611053467, "learning_rate": 4.584041471406707e-05, "loss": 0.0065, "step": 2421 }, { "epoch": 34.5920704845815, "grad_norm": 0.3857043385505676, "learning_rate": 4.5762648655144666e-05, "loss": 0.006, "step": 2422 }, { "epoch": 34.606167400881056, "grad_norm": 0.40385985374450684, "learning_rate": 4.5684929039047385e-05, "loss": 0.0095, "step": 2423 }, { "epoch": 34.62026431718062, "grad_norm": 0.31481635570526123, "learning_rate": 4.5607255932325665e-05, "loss": 0.0079, "step": 2424 }, { "epoch": 34.63436123348018, "grad_norm": 0.25715172290802, "learning_rate": 4.552962940149005e-05, "loss": 0.0062, "step": 2425 }, { "epoch": 34.648458149779735, "grad_norm": 0.3586297929286957, "learning_rate": 4.545204951301126e-05, "loss": 0.0068, "step": 2426 }, { "epoch": 34.66255506607929, "grad_norm": 0.1372905969619751, "learning_rate": 4.5374516333320104e-05, "loss": 0.0073, "step": 2427 }, { "epoch": 34.67665198237886, "grad_norm": 0.3591713011264801, "learning_rate": 4.529702992880737e-05, "loss": 0.0063, "step": 2428 }, { "epoch": 34.690748898678414, "grad_norm": 0.2339702546596527, "learning_rate": 4.5219590365823714e-05, "loss": 0.0082, "step": 2429 }, { "epoch": 34.70484581497797, "grad_norm": 0.4381974935531616, "learning_rate": 4.51421977106798e-05, "loss": 0.0083, "step": 2430 }, { "epoch": 34.71894273127754, "grad_norm": 0.7837631106376648, "learning_rate": 4.5064852029646056e-05, "loss": 0.0078, "step": 2431 }, { "epoch": 34.733039647577094, "grad_norm": 0.20921994745731354, "learning_rate": 4.498755338895275e-05, "loss": 0.0064, "step": 2432 }, { "epoch": 34.74713656387665, "grad_norm": 0.3005639910697937, "learning_rate": 4.491030185478976e-05, "loss": 0.0084, "step": 2433 }, { "epoch": 34.76123348017621, "grad_norm": 0.267943412065506, "learning_rate": 4.483309749330672e-05, "loss": 0.0075, "step": 2434 }, { "epoch": 34.77533039647577, "grad_norm": 0.2899210751056671, "learning_rate": 4.475594037061285e-05, "loss": 0.0075, "step": 2435 }, { "epoch": 34.78942731277533, "grad_norm": 0.21918809413909912, "learning_rate": 4.467883055277695e-05, "loss": 0.007, "step": 2436 }, { "epoch": 34.80352422907489, "grad_norm": 0.29723089933395386, "learning_rate": 4.46017681058272e-05, "loss": 0.0073, "step": 2437 }, { "epoch": 34.817621145374446, "grad_norm": 0.2601141929626465, "learning_rate": 4.4524753095751324e-05, "loss": 0.0076, "step": 2438 }, { "epoch": 34.83171806167401, "grad_norm": 0.3015618324279785, "learning_rate": 4.444778558849646e-05, "loss": 0.0076, "step": 2439 }, { "epoch": 34.84581497797357, "grad_norm": 0.36538437008857727, "learning_rate": 4.4370865649968906e-05, "loss": 0.0098, "step": 2440 }, { "epoch": 34.859911894273125, "grad_norm": 0.34248438477516174, "learning_rate": 4.429399334603438e-05, "loss": 0.009, "step": 2441 }, { "epoch": 34.87400881057269, "grad_norm": 0.7646989822387695, "learning_rate": 4.421716874251775e-05, "loss": 0.0086, "step": 2442 }, { "epoch": 34.88810572687225, "grad_norm": 0.28248313069343567, "learning_rate": 4.414039190520308e-05, "loss": 0.0068, "step": 2443 }, { "epoch": 34.902202643171805, "grad_norm": 0.27931493520736694, "learning_rate": 4.4063662899833436e-05, "loss": 0.0083, "step": 2444 }, { "epoch": 34.91629955947137, "grad_norm": 0.375320702791214, "learning_rate": 4.3986981792111025e-05, "loss": 0.0077, "step": 2445 }, { "epoch": 34.93039647577093, "grad_norm": 0.3775860369205475, "learning_rate": 4.3910348647696994e-05, "loss": 0.0096, "step": 2446 }, { "epoch": 34.944493392070484, "grad_norm": 0.35871151089668274, "learning_rate": 4.3833763532211495e-05, "loss": 0.0066, "step": 2447 }, { "epoch": 34.95859030837004, "grad_norm": 0.2470172494649887, "learning_rate": 4.375722651123338e-05, "loss": 0.0067, "step": 2448 }, { "epoch": 34.972687224669606, "grad_norm": 0.5143896341323853, "learning_rate": 4.368073765030048e-05, "loss": 0.0072, "step": 2449 }, { "epoch": 34.986784140969164, "grad_norm": 0.25117650628089905, "learning_rate": 4.360429701490934e-05, "loss": 0.0082, "step": 2450 }, { "epoch": 35.01409691629956, "grad_norm": 0.2038593292236328, "learning_rate": 4.352790467051524e-05, "loss": 0.0112, "step": 2451 }, { "epoch": 35.02819383259912, "grad_norm": 0.26186510920524597, "learning_rate": 4.345156068253198e-05, "loss": 0.0066, "step": 2452 }, { "epoch": 35.04229074889868, "grad_norm": 0.2134835422039032, "learning_rate": 4.337526511633211e-05, "loss": 0.006, "step": 2453 }, { "epoch": 35.05638766519824, "grad_norm": 0.13780167698860168, "learning_rate": 4.3299018037246664e-05, "loss": 0.0061, "step": 2454 }, { "epoch": 35.070484581497794, "grad_norm": 0.1615787297487259, "learning_rate": 4.322281951056508e-05, "loss": 0.0058, "step": 2455 }, { "epoch": 35.08458149779736, "grad_norm": 0.17879152297973633, "learning_rate": 4.314666960153534e-05, "loss": 0.0058, "step": 2456 }, { "epoch": 35.098678414096916, "grad_norm": 0.25619542598724365, "learning_rate": 4.307056837536373e-05, "loss": 0.0068, "step": 2457 }, { "epoch": 35.112775330396474, "grad_norm": 0.2199319452047348, "learning_rate": 4.299451589721489e-05, "loss": 0.0056, "step": 2458 }, { "epoch": 35.12687224669604, "grad_norm": 0.20068131387233734, "learning_rate": 4.291851223221163e-05, "loss": 0.006, "step": 2459 }, { "epoch": 35.140969162995596, "grad_norm": 0.32015150785446167, "learning_rate": 4.284255744543505e-05, "loss": 0.0077, "step": 2460 }, { "epoch": 35.15506607929515, "grad_norm": 0.24167773127555847, "learning_rate": 4.2766651601924365e-05, "loss": 0.0066, "step": 2461 }, { "epoch": 35.16916299559471, "grad_norm": 0.2686270773410797, "learning_rate": 4.2690794766676935e-05, "loss": 0.0065, "step": 2462 }, { "epoch": 35.183259911894275, "grad_norm": 0.18451833724975586, "learning_rate": 4.261498700464801e-05, "loss": 0.0068, "step": 2463 }, { "epoch": 35.19735682819383, "grad_norm": 0.2330138087272644, "learning_rate": 4.253922838075095e-05, "loss": 0.0065, "step": 2464 }, { "epoch": 35.21145374449339, "grad_norm": 0.19068285822868347, "learning_rate": 4.246351895985702e-05, "loss": 0.0064, "step": 2465 }, { "epoch": 35.225550660792955, "grad_norm": 0.21536603569984436, "learning_rate": 4.2387858806795345e-05, "loss": 0.0063, "step": 2466 }, { "epoch": 35.23964757709251, "grad_norm": 0.2397756278514862, "learning_rate": 4.231224798635279e-05, "loss": 0.0066, "step": 2467 }, { "epoch": 35.25374449339207, "grad_norm": 0.2479567974805832, "learning_rate": 4.223668656327409e-05, "loss": 0.0067, "step": 2468 }, { "epoch": 35.26784140969163, "grad_norm": 0.22959718108177185, "learning_rate": 4.216117460226159e-05, "loss": 0.0064, "step": 2469 }, { "epoch": 35.28193832599119, "grad_norm": 0.19867554306983948, "learning_rate": 4.208571216797539e-05, "loss": 0.0062, "step": 2470 }, { "epoch": 35.29603524229075, "grad_norm": 0.1561782956123352, "learning_rate": 4.2010299325033034e-05, "loss": 0.0075, "step": 2471 }, { "epoch": 35.31013215859031, "grad_norm": 0.3390534818172455, "learning_rate": 4.1934936138009696e-05, "loss": 0.0069, "step": 2472 }, { "epoch": 35.32422907488987, "grad_norm": 0.36443138122558594, "learning_rate": 4.185962267143807e-05, "loss": 0.0063, "step": 2473 }, { "epoch": 35.33832599118943, "grad_norm": 0.29778555035591125, "learning_rate": 4.178435898980815e-05, "loss": 0.0068, "step": 2474 }, { "epoch": 35.352422907488986, "grad_norm": 0.2625642418861389, "learning_rate": 4.170914515756737e-05, "loss": 0.007, "step": 2475 }, { "epoch": 35.36651982378854, "grad_norm": 0.5740969181060791, "learning_rate": 4.163398123912052e-05, "loss": 0.0073, "step": 2476 }, { "epoch": 35.38061674008811, "grad_norm": 0.3999325931072235, "learning_rate": 4.1558867298829615e-05, "loss": 0.0074, "step": 2477 }, { "epoch": 35.394713656387665, "grad_norm": 0.23500283062458038, "learning_rate": 4.1483803401013796e-05, "loss": 0.0076, "step": 2478 }, { "epoch": 35.40881057268722, "grad_norm": 0.25202012062072754, "learning_rate": 4.140878960994946e-05, "loss": 0.0072, "step": 2479 }, { "epoch": 35.42290748898679, "grad_norm": 0.1877162754535675, "learning_rate": 4.133382598987007e-05, "loss": 0.0067, "step": 2480 }, { "epoch": 35.437004405286345, "grad_norm": 0.3897045850753784, "learning_rate": 4.1258912604966135e-05, "loss": 0.0075, "step": 2481 }, { "epoch": 35.4511013215859, "grad_norm": 0.872191846370697, "learning_rate": 4.1184049519385084e-05, "loss": 0.0075, "step": 2482 }, { "epoch": 35.46519823788546, "grad_norm": 0.34057122468948364, "learning_rate": 4.110923679723134e-05, "loss": 0.0089, "step": 2483 }, { "epoch": 35.479295154185024, "grad_norm": 0.2490684688091278, "learning_rate": 4.1034474502566186e-05, "loss": 0.0062, "step": 2484 }, { "epoch": 35.49339207048458, "grad_norm": 0.3991025984287262, "learning_rate": 4.0959762699407766e-05, "loss": 0.0062, "step": 2485 }, { "epoch": 35.50748898678414, "grad_norm": 0.4681607782840729, "learning_rate": 4.0885101451730865e-05, "loss": 0.0065, "step": 2486 }, { "epoch": 35.521585903083704, "grad_norm": 0.13084010779857635, "learning_rate": 4.081049082346709e-05, "loss": 0.0058, "step": 2487 }, { "epoch": 35.53568281938326, "grad_norm": 0.18938884139060974, "learning_rate": 4.073593087850473e-05, "loss": 0.0066, "step": 2488 }, { "epoch": 35.54977973568282, "grad_norm": 0.24294225871562958, "learning_rate": 4.066142168068852e-05, "loss": 0.0073, "step": 2489 }, { "epoch": 35.563876651982376, "grad_norm": 0.29681339859962463, "learning_rate": 4.058696329381987e-05, "loss": 0.0078, "step": 2490 }, { "epoch": 35.57797356828194, "grad_norm": 0.2685471475124359, "learning_rate": 4.051255578165666e-05, "loss": 0.0063, "step": 2491 }, { "epoch": 35.5920704845815, "grad_norm": 0.2831193208694458, "learning_rate": 4.043819920791322e-05, "loss": 0.0073, "step": 2492 }, { "epoch": 35.606167400881056, "grad_norm": 0.2899085283279419, "learning_rate": 4.0363893636260144e-05, "loss": 0.0061, "step": 2493 }, { "epoch": 35.62026431718062, "grad_norm": 0.19841645658016205, "learning_rate": 4.028963913032453e-05, "loss": 0.0058, "step": 2494 }, { "epoch": 35.63436123348018, "grad_norm": 0.18550072610378265, "learning_rate": 4.0215435753689676e-05, "loss": 0.0065, "step": 2495 }, { "epoch": 35.648458149779735, "grad_norm": 0.6187781691551208, "learning_rate": 4.014128356989503e-05, "loss": 0.0075, "step": 2496 }, { "epoch": 35.66255506607929, "grad_norm": 0.3237227201461792, "learning_rate": 4.006718264243627e-05, "loss": 0.0073, "step": 2497 }, { "epoch": 35.67665198237886, "grad_norm": 0.22121280431747437, "learning_rate": 3.9993133034765185e-05, "loss": 0.006, "step": 2498 }, { "epoch": 35.690748898678414, "grad_norm": 0.19185158610343933, "learning_rate": 3.991913481028965e-05, "loss": 0.0062, "step": 2499 }, { "epoch": 35.70484581497797, "grad_norm": 0.6613407135009766, "learning_rate": 3.9845188032373426e-05, "loss": 0.0074, "step": 2500 }, { "epoch": 35.70484581497797, "eval_loss": 1.4360734224319458, "eval_runtime": 132.7214, "eval_samples_per_second": 13.623, "eval_steps_per_second": 0.429, "step": 2500 }, { "epoch": 35.71894273127754, "grad_norm": 0.23150266706943512, "learning_rate": 3.977129276433633e-05, "loss": 0.006, "step": 2501 }, { "epoch": 35.733039647577094, "grad_norm": 0.2696787416934967, "learning_rate": 3.969744906945404e-05, "loss": 0.0076, "step": 2502 }, { "epoch": 35.74713656387665, "grad_norm": 0.1966117024421692, "learning_rate": 3.962365701095809e-05, "loss": 0.006, "step": 2503 }, { "epoch": 35.76123348017621, "grad_norm": 0.38149067759513855, "learning_rate": 3.954991665203573e-05, "loss": 0.0069, "step": 2504 }, { "epoch": 35.77533039647577, "grad_norm": 0.26715636253356934, "learning_rate": 3.947622805583001e-05, "loss": 0.0076, "step": 2505 }, { "epoch": 35.78942731277533, "grad_norm": 0.35017141699790955, "learning_rate": 3.940259128543967e-05, "loss": 0.0072, "step": 2506 }, { "epoch": 35.80352422907489, "grad_norm": 0.1958409994840622, "learning_rate": 3.932900640391897e-05, "loss": 0.0072, "step": 2507 }, { "epoch": 35.817621145374446, "grad_norm": 0.3187143802642822, "learning_rate": 3.9255473474277827e-05, "loss": 0.0079, "step": 2508 }, { "epoch": 35.83171806167401, "grad_norm": 0.30764374136924744, "learning_rate": 3.918199255948167e-05, "loss": 0.0077, "step": 2509 }, { "epoch": 35.84581497797357, "grad_norm": 0.4027601182460785, "learning_rate": 3.910856372245139e-05, "loss": 0.0071, "step": 2510 }, { "epoch": 35.859911894273125, "grad_norm": 0.3463936150074005, "learning_rate": 3.903518702606319e-05, "loss": 0.0074, "step": 2511 }, { "epoch": 35.87400881057269, "grad_norm": 0.3076491355895996, "learning_rate": 3.896186253314874e-05, "loss": 0.0061, "step": 2512 }, { "epoch": 35.88810572687225, "grad_norm": 0.26995375752449036, "learning_rate": 3.8888590306494974e-05, "loss": 0.0076, "step": 2513 }, { "epoch": 35.902202643171805, "grad_norm": 0.47143593430519104, "learning_rate": 3.881537040884407e-05, "loss": 0.0075, "step": 2514 }, { "epoch": 35.91629955947137, "grad_norm": 0.24302785098552704, "learning_rate": 3.874220290289336e-05, "loss": 0.007, "step": 2515 }, { "epoch": 35.93039647577093, "grad_norm": 0.31675300002098083, "learning_rate": 3.866908785129535e-05, "loss": 0.0074, "step": 2516 }, { "epoch": 35.944493392070484, "grad_norm": 0.347881942987442, "learning_rate": 3.859602531665764e-05, "loss": 0.009, "step": 2517 }, { "epoch": 35.95859030837004, "grad_norm": 0.6523621082305908, "learning_rate": 3.852301536154287e-05, "loss": 0.0072, "step": 2518 }, { "epoch": 35.972687224669606, "grad_norm": 0.4926827847957611, "learning_rate": 3.8450058048468565e-05, "loss": 0.0081, "step": 2519 }, { "epoch": 35.986784140969164, "grad_norm": 0.36400601267814636, "learning_rate": 3.8377153439907266e-05, "loss": 0.008, "step": 2520 }, { "epoch": 36.01409691629956, "grad_norm": 0.3509397506713867, "learning_rate": 3.830430159828641e-05, "loss": 0.0112, "step": 2521 }, { "epoch": 36.02819383259912, "grad_norm": 0.43091192841529846, "learning_rate": 3.823150258598812e-05, "loss": 0.0069, "step": 2522 }, { "epoch": 36.04229074889868, "grad_norm": 0.3713698387145996, "learning_rate": 3.815875646534939e-05, "loss": 0.0066, "step": 2523 }, { "epoch": 36.05638766519824, "grad_norm": 0.24885639548301697, "learning_rate": 3.80860632986619e-05, "loss": 0.0061, "step": 2524 }, { "epoch": 36.070484581497794, "grad_norm": 0.1293696016073227, "learning_rate": 3.801342314817201e-05, "loss": 0.0056, "step": 2525 }, { "epoch": 36.08458149779736, "grad_norm": 0.16193059086799622, "learning_rate": 3.794083607608059e-05, "loss": 0.0067, "step": 2526 }, { "epoch": 36.098678414096916, "grad_norm": 0.16306906938552856, "learning_rate": 3.786830214454315e-05, "loss": 0.0064, "step": 2527 }, { "epoch": 36.112775330396474, "grad_norm": 0.4109697937965393, "learning_rate": 3.7795821415669654e-05, "loss": 0.007, "step": 2528 }, { "epoch": 36.12687224669604, "grad_norm": 0.6125305891036987, "learning_rate": 3.772339395152458e-05, "loss": 0.0059, "step": 2529 }, { "epoch": 36.140969162995596, "grad_norm": 0.1962757557630539, "learning_rate": 3.7651019814126654e-05, "loss": 0.0058, "step": 2530 }, { "epoch": 36.15506607929515, "grad_norm": 0.19840076565742493, "learning_rate": 3.7578699065449074e-05, "loss": 0.0062, "step": 2531 }, { "epoch": 36.16916299559471, "grad_norm": 0.1280612349510193, "learning_rate": 3.750643176741926e-05, "loss": 0.0057, "step": 2532 }, { "epoch": 36.183259911894275, "grad_norm": 0.3248969614505768, "learning_rate": 3.743421798191892e-05, "loss": 0.0072, "step": 2533 }, { "epoch": 36.19735682819383, "grad_norm": 0.19489607214927673, "learning_rate": 3.736205777078381e-05, "loss": 0.0057, "step": 2534 }, { "epoch": 36.21145374449339, "grad_norm": 0.19377432763576508, "learning_rate": 3.7289951195803954e-05, "loss": 0.0071, "step": 2535 }, { "epoch": 36.225550660792955, "grad_norm": 0.3409094214439392, "learning_rate": 3.7217898318723356e-05, "loss": 0.0058, "step": 2536 }, { "epoch": 36.23964757709251, "grad_norm": 0.30412793159484863, "learning_rate": 3.714589920124013e-05, "loss": 0.006, "step": 2537 }, { "epoch": 36.25374449339207, "grad_norm": 0.13733762502670288, "learning_rate": 3.707395390500623e-05, "loss": 0.0058, "step": 2538 }, { "epoch": 36.26784140969163, "grad_norm": 0.179313525557518, "learning_rate": 3.700206249162762e-05, "loss": 0.0064, "step": 2539 }, { "epoch": 36.28193832599119, "grad_norm": 0.26086026430130005, "learning_rate": 3.693022502266413e-05, "loss": 0.0067, "step": 2540 }, { "epoch": 36.29603524229075, "grad_norm": 0.1277211308479309, "learning_rate": 3.6858441559629306e-05, "loss": 0.0063, "step": 2541 }, { "epoch": 36.31013215859031, "grad_norm": 0.24885113537311554, "learning_rate": 3.6786712163990536e-05, "loss": 0.0066, "step": 2542 }, { "epoch": 36.32422907488987, "grad_norm": 0.21123166382312775, "learning_rate": 3.671503689716888e-05, "loss": 0.0068, "step": 2543 }, { "epoch": 36.33832599118943, "grad_norm": 0.15314695239067078, "learning_rate": 3.664341582053908e-05, "loss": 0.0061, "step": 2544 }, { "epoch": 36.352422907488986, "grad_norm": 0.1730436086654663, "learning_rate": 3.6571848995429405e-05, "loss": 0.006, "step": 2545 }, { "epoch": 36.36651982378854, "grad_norm": 0.2311393916606903, "learning_rate": 3.650033648312172e-05, "loss": 0.0065, "step": 2546 }, { "epoch": 36.38061674008811, "grad_norm": 0.228529691696167, "learning_rate": 3.642887834485138e-05, "loss": 0.0067, "step": 2547 }, { "epoch": 36.394713656387665, "grad_norm": 0.19748222827911377, "learning_rate": 3.63574746418072e-05, "loss": 0.0056, "step": 2548 }, { "epoch": 36.40881057268722, "grad_norm": 0.3373155891895294, "learning_rate": 3.62861254351313e-05, "loss": 0.0062, "step": 2549 }, { "epoch": 36.42290748898679, "grad_norm": 0.23062419891357422, "learning_rate": 3.621483078591923e-05, "loss": 0.0058, "step": 2550 }, { "epoch": 36.437004405286345, "grad_norm": 0.19287988543510437, "learning_rate": 3.614359075521977e-05, "loss": 0.006, "step": 2551 }, { "epoch": 36.4511013215859, "grad_norm": 0.251128226518631, "learning_rate": 3.6072405404034984e-05, "loss": 0.0067, "step": 2552 }, { "epoch": 36.46519823788546, "grad_norm": 0.20572823286056519, "learning_rate": 3.600127479332002e-05, "loss": 0.007, "step": 2553 }, { "epoch": 36.479295154185024, "grad_norm": 0.17155574262142181, "learning_rate": 3.593019898398322e-05, "loss": 0.0061, "step": 2554 }, { "epoch": 36.49339207048458, "grad_norm": 0.32488787174224854, "learning_rate": 3.585917803688603e-05, "loss": 0.0078, "step": 2555 }, { "epoch": 36.50748898678414, "grad_norm": 0.18982678651809692, "learning_rate": 3.5788212012842826e-05, "loss": 0.0064, "step": 2556 }, { "epoch": 36.521585903083704, "grad_norm": 0.21168148517608643, "learning_rate": 3.571730097262103e-05, "loss": 0.0069, "step": 2557 }, { "epoch": 36.53568281938326, "grad_norm": 0.31244179606437683, "learning_rate": 3.564644497694094e-05, "loss": 0.0067, "step": 2558 }, { "epoch": 36.54977973568282, "grad_norm": 0.18113215267658234, "learning_rate": 3.557564408647578e-05, "loss": 0.006, "step": 2559 }, { "epoch": 36.563876651982376, "grad_norm": 0.48472627997398376, "learning_rate": 3.550489836185149e-05, "loss": 0.0069, "step": 2560 }, { "epoch": 36.57797356828194, "grad_norm": 0.16598792374134064, "learning_rate": 3.5434207863646854e-05, "loss": 0.0064, "step": 2561 }, { "epoch": 36.5920704845815, "grad_norm": 0.2475729137659073, "learning_rate": 3.5363572652393326e-05, "loss": 0.0058, "step": 2562 }, { "epoch": 36.606167400881056, "grad_norm": 0.23951038718223572, "learning_rate": 3.529299278857508e-05, "loss": 0.0062, "step": 2563 }, { "epoch": 36.62026431718062, "grad_norm": 0.1823936551809311, "learning_rate": 3.5222468332628764e-05, "loss": 0.0064, "step": 2564 }, { "epoch": 36.63436123348018, "grad_norm": 0.388275682926178, "learning_rate": 3.5151999344943735e-05, "loss": 0.0066, "step": 2565 }, { "epoch": 36.648458149779735, "grad_norm": 0.1629100888967514, "learning_rate": 3.5081585885861736e-05, "loss": 0.0058, "step": 2566 }, { "epoch": 36.66255506607929, "grad_norm": 0.15400229394435883, "learning_rate": 3.501122801567708e-05, "loss": 0.0063, "step": 2567 }, { "epoch": 36.67665198237886, "grad_norm": 0.17900967597961426, "learning_rate": 3.4940925794636314e-05, "loss": 0.0059, "step": 2568 }, { "epoch": 36.690748898678414, "grad_norm": 0.21681256592273712, "learning_rate": 3.487067928293848e-05, "loss": 0.0065, "step": 2569 }, { "epoch": 36.70484581497797, "grad_norm": 0.25069376826286316, "learning_rate": 3.4800488540734856e-05, "loss": 0.0074, "step": 2570 }, { "epoch": 36.71894273127754, "grad_norm": 0.6137335300445557, "learning_rate": 3.4730353628129e-05, "loss": 0.0081, "step": 2571 }, { "epoch": 36.733039647577094, "grad_norm": 0.2594534456729889, "learning_rate": 3.466027460517659e-05, "loss": 0.0068, "step": 2572 }, { "epoch": 36.74713656387665, "grad_norm": 0.38757410645484924, "learning_rate": 3.459025153188551e-05, "loss": 0.0084, "step": 2573 }, { "epoch": 36.76123348017621, "grad_norm": 0.20982396602630615, "learning_rate": 3.452028446821577e-05, "loss": 0.0057, "step": 2574 }, { "epoch": 36.77533039647577, "grad_norm": 0.3276127576828003, "learning_rate": 3.4450373474079276e-05, "loss": 0.007, "step": 2575 }, { "epoch": 36.78942731277533, "grad_norm": 0.29381605982780457, "learning_rate": 3.4380518609340076e-05, "loss": 0.0065, "step": 2576 }, { "epoch": 36.80352422907489, "grad_norm": 0.37063631415367126, "learning_rate": 3.431071993381406e-05, "loss": 0.0068, "step": 2577 }, { "epoch": 36.817621145374446, "grad_norm": 0.15015771985054016, "learning_rate": 3.424097750726911e-05, "loss": 0.0056, "step": 2578 }, { "epoch": 36.83171806167401, "grad_norm": 0.3533632755279541, "learning_rate": 3.417129138942478e-05, "loss": 0.0076, "step": 2579 }, { "epoch": 36.84581497797357, "grad_norm": 0.24427290260791779, "learning_rate": 3.4101661639952534e-05, "loss": 0.006, "step": 2580 }, { "epoch": 36.859911894273125, "grad_norm": 0.31687480211257935, "learning_rate": 3.403208831847552e-05, "loss": 0.0069, "step": 2581 }, { "epoch": 36.87400881057269, "grad_norm": 0.32230329513549805, "learning_rate": 3.3962571484568636e-05, "loss": 0.0073, "step": 2582 }, { "epoch": 36.88810572687225, "grad_norm": 0.38379141688346863, "learning_rate": 3.389311119775828e-05, "loss": 0.0072, "step": 2583 }, { "epoch": 36.902202643171805, "grad_norm": 0.31696388125419617, "learning_rate": 3.3823707517522526e-05, "loss": 0.0069, "step": 2584 }, { "epoch": 36.91629955947137, "grad_norm": 0.44513413310050964, "learning_rate": 3.375436050329097e-05, "loss": 0.007, "step": 2585 }, { "epoch": 36.93039647577093, "grad_norm": 0.2985347509384155, "learning_rate": 3.368507021444468e-05, "loss": 0.0078, "step": 2586 }, { "epoch": 36.944493392070484, "grad_norm": 0.2286648154258728, "learning_rate": 3.361583671031609e-05, "loss": 0.0064, "step": 2587 }, { "epoch": 36.95859030837004, "grad_norm": 0.40980517864227295, "learning_rate": 3.354666005018909e-05, "loss": 0.007, "step": 2588 }, { "epoch": 36.972687224669606, "grad_norm": 0.21484334766864777, "learning_rate": 3.34775402932989e-05, "loss": 0.0065, "step": 2589 }, { "epoch": 36.986784140969164, "grad_norm": 0.20655077695846558, "learning_rate": 3.340847749883191e-05, "loss": 0.0067, "step": 2590 }, { "epoch": 37.01409691629956, "grad_norm": 0.18434301018714905, "learning_rate": 3.333947172592584e-05, "loss": 0.0123, "step": 2591 }, { "epoch": 37.02819383259912, "grad_norm": 0.1810382455587387, "learning_rate": 3.327052303366954e-05, "loss": 0.0059, "step": 2592 }, { "epoch": 37.04229074889868, "grad_norm": 0.24454429745674133, "learning_rate": 3.3201631481103014e-05, "loss": 0.0066, "step": 2593 }, { "epoch": 37.05638766519824, "grad_norm": 0.15230068564414978, "learning_rate": 3.313279712721726e-05, "loss": 0.006, "step": 2594 }, { "epoch": 37.070484581497794, "grad_norm": 0.147460475564003, "learning_rate": 3.306402003095438e-05, "loss": 0.0058, "step": 2595 }, { "epoch": 37.08458149779736, "grad_norm": 0.22096075117588043, "learning_rate": 3.299530025120742e-05, "loss": 0.0059, "step": 2596 }, { "epoch": 37.098678414096916, "grad_norm": 0.15228970348834991, "learning_rate": 3.292663784682036e-05, "loss": 0.0056, "step": 2597 }, { "epoch": 37.112775330396474, "grad_norm": 0.1472080945968628, "learning_rate": 3.2858032876588e-05, "loss": 0.0058, "step": 2598 }, { "epoch": 37.12687224669604, "grad_norm": 0.17952805757522583, "learning_rate": 3.2789485399255946e-05, "loss": 0.0059, "step": 2599 }, { "epoch": 37.140969162995596, "grad_norm": 0.1520863175392151, "learning_rate": 3.2720995473520766e-05, "loss": 0.0057, "step": 2600 }, { "epoch": 37.15506607929515, "grad_norm": 0.16592347621917725, "learning_rate": 3.265256315802947e-05, "loss": 0.0058, "step": 2601 }, { "epoch": 37.16916299559471, "grad_norm": 0.1890707015991211, "learning_rate": 3.258418851137992e-05, "loss": 0.0059, "step": 2602 }, { "epoch": 37.183259911894275, "grad_norm": 0.15389230847358704, "learning_rate": 3.251587159212054e-05, "loss": 0.0058, "step": 2603 }, { "epoch": 37.19735682819383, "grad_norm": 0.22566065192222595, "learning_rate": 3.2447612458750365e-05, "loss": 0.006, "step": 2604 }, { "epoch": 37.21145374449339, "grad_norm": 0.2668918967247009, "learning_rate": 3.237941116971884e-05, "loss": 0.0067, "step": 2605 }, { "epoch": 37.225550660792955, "grad_norm": 0.17968614399433136, "learning_rate": 3.231126778342599e-05, "loss": 0.0062, "step": 2606 }, { "epoch": 37.23964757709251, "grad_norm": 0.12095355987548828, "learning_rate": 3.224318235822225e-05, "loss": 0.0054, "step": 2607 }, { "epoch": 37.25374449339207, "grad_norm": 0.20153504610061646, "learning_rate": 3.2175154952408316e-05, "loss": 0.0061, "step": 2608 }, { "epoch": 37.26784140969163, "grad_norm": 0.40175047516822815, "learning_rate": 3.2107185624235325e-05, "loss": 0.0059, "step": 2609 }, { "epoch": 37.28193832599119, "grad_norm": 0.21215341985225677, "learning_rate": 3.203927443190464e-05, "loss": 0.0057, "step": 2610 }, { "epoch": 37.29603524229075, "grad_norm": 0.460979163646698, "learning_rate": 3.197142143356787e-05, "loss": 0.0064, "step": 2611 }, { "epoch": 37.31013215859031, "grad_norm": 0.2577584385871887, "learning_rate": 3.1903626687326696e-05, "loss": 0.0071, "step": 2612 }, { "epoch": 37.32422907488987, "grad_norm": 0.20035308599472046, "learning_rate": 3.1835890251233005e-05, "loss": 0.0058, "step": 2613 }, { "epoch": 37.33832599118943, "grad_norm": 0.5793057680130005, "learning_rate": 3.176821218328876e-05, "loss": 0.0062, "step": 2614 }, { "epoch": 37.352422907488986, "grad_norm": 0.12777845561504364, "learning_rate": 3.170059254144593e-05, "loss": 0.0058, "step": 2615 }, { "epoch": 37.36651982378854, "grad_norm": 0.213422030210495, "learning_rate": 3.163303138360638e-05, "loss": 0.0058, "step": 2616 }, { "epoch": 37.38061674008811, "grad_norm": 0.3228277564048767, "learning_rate": 3.156552876762201e-05, "loss": 0.0059, "step": 2617 }, { "epoch": 37.394713656387665, "grad_norm": 0.17087338864803314, "learning_rate": 3.149808475129452e-05, "loss": 0.0058, "step": 2618 }, { "epoch": 37.40881057268722, "grad_norm": 0.22393253445625305, "learning_rate": 3.14306993923755e-05, "loss": 0.0057, "step": 2619 }, { "epoch": 37.42290748898679, "grad_norm": 0.19596342742443085, "learning_rate": 3.1363372748566176e-05, "loss": 0.006, "step": 2620 }, { "epoch": 37.437004405286345, "grad_norm": 0.11231343448162079, "learning_rate": 3.129610487751763e-05, "loss": 0.0054, "step": 2621 }, { "epoch": 37.4511013215859, "grad_norm": 0.38706645369529724, "learning_rate": 3.1228895836830586e-05, "loss": 0.007, "step": 2622 }, { "epoch": 37.46519823788546, "grad_norm": 0.1904577612876892, "learning_rate": 3.1161745684055385e-05, "loss": 0.006, "step": 2623 }, { "epoch": 37.479295154185024, "grad_norm": 1.4619957208633423, "learning_rate": 3.10946544766919e-05, "loss": 0.0072, "step": 2624 }, { "epoch": 37.49339207048458, "grad_norm": 0.20466011762619019, "learning_rate": 3.102762227218957e-05, "loss": 0.0062, "step": 2625 }, { "epoch": 37.50748898678414, "grad_norm": 0.21006366610527039, "learning_rate": 3.096064912794736e-05, "loss": 0.0063, "step": 2626 }, { "epoch": 37.521585903083704, "grad_norm": 0.22074024379253387, "learning_rate": 3.089373510131354e-05, "loss": 0.0062, "step": 2627 }, { "epoch": 37.53568281938326, "grad_norm": 0.18598806858062744, "learning_rate": 3.0826880249585854e-05, "loss": 0.0057, "step": 2628 }, { "epoch": 37.54977973568282, "grad_norm": 0.25878721475601196, "learning_rate": 3.0760084630011356e-05, "loss": 0.0065, "step": 2629 }, { "epoch": 37.563876651982376, "grad_norm": 0.66555255651474, "learning_rate": 3.0693348299786405e-05, "loss": 0.0064, "step": 2630 }, { "epoch": 37.57797356828194, "grad_norm": 0.22651249170303345, "learning_rate": 3.062667131605649e-05, "loss": 0.0055, "step": 2631 }, { "epoch": 37.5920704845815, "grad_norm": 0.3072776794433594, "learning_rate": 3.056005373591637e-05, "loss": 0.0078, "step": 2632 }, { "epoch": 37.606167400881056, "grad_norm": 0.41494208574295044, "learning_rate": 3.049349561640995e-05, "loss": 0.0071, "step": 2633 }, { "epoch": 37.62026431718062, "grad_norm": 0.1654483526945114, "learning_rate": 3.04269970145302e-05, "loss": 0.0056, "step": 2634 }, { "epoch": 37.63436123348018, "grad_norm": 0.23326967656612396, "learning_rate": 3.0360557987219053e-05, "loss": 0.0069, "step": 2635 }, { "epoch": 37.648458149779735, "grad_norm": 1.17448890209198, "learning_rate": 3.029417859136753e-05, "loss": 0.0077, "step": 2636 }, { "epoch": 37.66255506607929, "grad_norm": 0.34484609961509705, "learning_rate": 3.022785888381553e-05, "loss": 0.0082, "step": 2637 }, { "epoch": 37.67665198237886, "grad_norm": 0.2268429696559906, "learning_rate": 3.016159892135192e-05, "loss": 0.0058, "step": 2638 }, { "epoch": 37.690748898678414, "grad_norm": 0.23429714143276215, "learning_rate": 3.0095398760714267e-05, "loss": 0.006, "step": 2639 }, { "epoch": 37.70484581497797, "grad_norm": 0.1626560091972351, "learning_rate": 3.0029258458589048e-05, "loss": 0.0059, "step": 2640 }, { "epoch": 37.71894273127754, "grad_norm": 0.1364603191614151, "learning_rate": 2.9963178071611487e-05, "loss": 0.0055, "step": 2641 }, { "epoch": 37.733039647577094, "grad_norm": 0.14087998867034912, "learning_rate": 2.9897157656365405e-05, "loss": 0.0056, "step": 2642 }, { "epoch": 37.74713656387665, "grad_norm": 0.2093956023454666, "learning_rate": 2.983119726938335e-05, "loss": 0.0063, "step": 2643 }, { "epoch": 37.76123348017621, "grad_norm": 0.23660559952259064, "learning_rate": 2.976529696714645e-05, "loss": 0.0061, "step": 2644 }, { "epoch": 37.77533039647577, "grad_norm": 0.22210091352462769, "learning_rate": 2.9699456806084426e-05, "loss": 0.0059, "step": 2645 }, { "epoch": 37.78942731277533, "grad_norm": 0.22096191346645355, "learning_rate": 2.9633676842575387e-05, "loss": 0.0061, "step": 2646 }, { "epoch": 37.80352422907489, "grad_norm": 0.5013126730918884, "learning_rate": 2.956795713294599e-05, "loss": 0.0083, "step": 2647 }, { "epoch": 37.817621145374446, "grad_norm": 0.23935861885547638, "learning_rate": 2.9502297733471286e-05, "loss": 0.0058, "step": 2648 }, { "epoch": 37.83171806167401, "grad_norm": 0.20824825763702393, "learning_rate": 2.9436698700374688e-05, "loss": 0.0056, "step": 2649 }, { "epoch": 37.84581497797357, "grad_norm": 0.15694352984428406, "learning_rate": 2.9371160089827842e-05, "loss": 0.0057, "step": 2650 }, { "epoch": 37.859911894273125, "grad_norm": 0.3522908687591553, "learning_rate": 2.9305681957950747e-05, "loss": 0.0063, "step": 2651 }, { "epoch": 37.87400881057269, "grad_norm": 0.1927536427974701, "learning_rate": 2.9240264360811577e-05, "loss": 0.0062, "step": 2652 }, { "epoch": 37.88810572687225, "grad_norm": 0.31471553444862366, "learning_rate": 2.9174907354426696e-05, "loss": 0.0074, "step": 2653 }, { "epoch": 37.902202643171805, "grad_norm": 0.4828501343727112, "learning_rate": 2.9109610994760515e-05, "loss": 0.0065, "step": 2654 }, { "epoch": 37.91629955947137, "grad_norm": 0.3238373398780823, "learning_rate": 2.9044375337725593e-05, "loss": 0.007, "step": 2655 }, { "epoch": 37.93039647577093, "grad_norm": 0.24290423095226288, "learning_rate": 2.8979200439182496e-05, "loss": 0.0061, "step": 2656 }, { "epoch": 37.944493392070484, "grad_norm": 0.4313226342201233, "learning_rate": 2.8914086354939707e-05, "loss": 0.0075, "step": 2657 }, { "epoch": 37.95859030837004, "grad_norm": 0.19211484491825104, "learning_rate": 2.884903314075368e-05, "loss": 0.0063, "step": 2658 }, { "epoch": 37.972687224669606, "grad_norm": 0.2725053131580353, "learning_rate": 2.8784040852328743e-05, "loss": 0.0067, "step": 2659 }, { "epoch": 37.986784140969164, "grad_norm": 0.2761980891227722, "learning_rate": 2.8719109545317103e-05, "loss": 0.0068, "step": 2660 }, { "epoch": 38.01409691629956, "grad_norm": 0.17537494003772736, "learning_rate": 2.8654239275318608e-05, "loss": 0.011, "step": 2661 }, { "epoch": 38.02819383259912, "grad_norm": 0.15196725726127625, "learning_rate": 2.8589430097880976e-05, "loss": 0.0057, "step": 2662 }, { "epoch": 38.04229074889868, "grad_norm": 0.17498135566711426, "learning_rate": 2.8524682068499574e-05, "loss": 0.0052, "step": 2663 }, { "epoch": 38.05638766519824, "grad_norm": 0.15377739071846008, "learning_rate": 2.845999524261741e-05, "loss": 0.0055, "step": 2664 }, { "epoch": 38.070484581497794, "grad_norm": 0.13032636046409607, "learning_rate": 2.839536967562504e-05, "loss": 0.0056, "step": 2665 }, { "epoch": 38.08458149779736, "grad_norm": 0.15669189393520355, "learning_rate": 2.8330805422860606e-05, "loss": 0.0058, "step": 2666 }, { "epoch": 38.098678414096916, "grad_norm": 0.16501875221729279, "learning_rate": 2.8266302539609745e-05, "loss": 0.0051, "step": 2667 }, { "epoch": 38.112775330396474, "grad_norm": 0.11612409353256226, "learning_rate": 2.8201861081105585e-05, "loss": 0.0054, "step": 2668 }, { "epoch": 38.12687224669604, "grad_norm": 0.14419996738433838, "learning_rate": 2.813748110252854e-05, "loss": 0.006, "step": 2669 }, { "epoch": 38.140969162995596, "grad_norm": 0.22038216888904572, "learning_rate": 2.807316265900649e-05, "loss": 0.0064, "step": 2670 }, { "epoch": 38.15506607929515, "grad_norm": 0.36697107553482056, "learning_rate": 2.800890580561457e-05, "loss": 0.0054, "step": 2671 }, { "epoch": 38.16916299559471, "grad_norm": 0.10079199075698853, "learning_rate": 2.7944710597375235e-05, "loss": 0.0062, "step": 2672 }, { "epoch": 38.183259911894275, "grad_norm": 0.20903602242469788, "learning_rate": 2.788057708925803e-05, "loss": 0.0053, "step": 2673 }, { "epoch": 38.19735682819383, "grad_norm": 0.10410485416650772, "learning_rate": 2.7816505336179798e-05, "loss": 0.0065, "step": 2674 }, { "epoch": 38.21145374449339, "grad_norm": 0.27186867594718933, "learning_rate": 2.7752495393004475e-05, "loss": 0.006, "step": 2675 }, { "epoch": 38.225550660792955, "grad_norm": 0.25012218952178955, "learning_rate": 2.7688547314542967e-05, "loss": 0.0063, "step": 2676 }, { "epoch": 38.23964757709251, "grad_norm": 0.23925872147083282, "learning_rate": 2.7624661155553345e-05, "loss": 0.0064, "step": 2677 }, { "epoch": 38.25374449339207, "grad_norm": 0.27681171894073486, "learning_rate": 2.7560836970740568e-05, "loss": 0.0064, "step": 2678 }, { "epoch": 38.26784140969163, "grad_norm": 0.27797433733940125, "learning_rate": 2.7497074814756608e-05, "loss": 0.006, "step": 2679 }, { "epoch": 38.28193832599119, "grad_norm": 0.1970745474100113, "learning_rate": 2.7433374742200192e-05, "loss": 0.0057, "step": 2680 }, { "epoch": 38.29603524229075, "grad_norm": 0.15866519510746002, "learning_rate": 2.736973680761702e-05, "loss": 0.0055, "step": 2681 }, { "epoch": 38.31013215859031, "grad_norm": 0.1399049609899521, "learning_rate": 2.7306161065499502e-05, "loss": 0.0057, "step": 2682 }, { "epoch": 38.32422907488987, "grad_norm": 0.17335154116153717, "learning_rate": 2.7242647570286884e-05, "loss": 0.0057, "step": 2683 }, { "epoch": 38.33832599118943, "grad_norm": 0.13008949160575867, "learning_rate": 2.7179196376364956e-05, "loss": 0.0075, "step": 2684 }, { "epoch": 38.352422907488986, "grad_norm": 0.2724975645542145, "learning_rate": 2.7115807538066297e-05, "loss": 0.0057, "step": 2685 }, { "epoch": 38.36651982378854, "grad_norm": 0.15065895020961761, "learning_rate": 2.7052481109670035e-05, "loss": 0.0054, "step": 2686 }, { "epoch": 38.38061674008811, "grad_norm": 0.1554323136806488, "learning_rate": 2.6989217145401912e-05, "loss": 0.0053, "step": 2687 }, { "epoch": 38.394713656387665, "grad_norm": 0.21194282174110413, "learning_rate": 2.6926015699434072e-05, "loss": 0.0057, "step": 2688 }, { "epoch": 38.40881057268722, "grad_norm": 0.18430671095848083, "learning_rate": 2.686287682588522e-05, "loss": 0.0064, "step": 2689 }, { "epoch": 38.42290748898679, "grad_norm": 0.2828986942768097, "learning_rate": 2.679980057882049e-05, "loss": 0.0068, "step": 2690 }, { "epoch": 38.437004405286345, "grad_norm": 0.2901768982410431, "learning_rate": 2.6736787012251287e-05, "loss": 0.0053, "step": 2691 }, { "epoch": 38.4511013215859, "grad_norm": 0.12829385697841644, "learning_rate": 2.6673836180135438e-05, "loss": 0.0055, "step": 2692 }, { "epoch": 38.46519823788546, "grad_norm": 0.2066812664270401, "learning_rate": 2.6610948136377022e-05, "loss": 0.0067, "step": 2693 }, { "epoch": 38.479295154185024, "grad_norm": 0.2616673707962036, "learning_rate": 2.6548122934826392e-05, "loss": 0.0068, "step": 2694 }, { "epoch": 38.49339207048458, "grad_norm": 0.36694610118865967, "learning_rate": 2.6485360629279987e-05, "loss": 0.0056, "step": 2695 }, { "epoch": 38.50748898678414, "grad_norm": 0.16546913981437683, "learning_rate": 2.6422661273480474e-05, "loss": 0.0067, "step": 2696 }, { "epoch": 38.521585903083704, "grad_norm": 0.32250329852104187, "learning_rate": 2.63600249211166e-05, "loss": 0.0055, "step": 2697 }, { "epoch": 38.53568281938326, "grad_norm": 0.13653482496738434, "learning_rate": 2.62974516258232e-05, "loss": 0.006, "step": 2698 }, { "epoch": 38.54977973568282, "grad_norm": 0.22198747098445892, "learning_rate": 2.6234941441181004e-05, "loss": 0.0055, "step": 2699 }, { "epoch": 38.563876651982376, "grad_norm": 0.11950288712978363, "learning_rate": 2.6172494420716797e-05, "loss": 0.0061, "step": 2700 }, { "epoch": 38.57797356828194, "grad_norm": 0.3190992474555969, "learning_rate": 2.611011061790326e-05, "loss": 0.0059, "step": 2701 }, { "epoch": 38.5920704845815, "grad_norm": 0.16060449182987213, "learning_rate": 2.6047790086158952e-05, "loss": 0.0061, "step": 2702 }, { "epoch": 38.606167400881056, "grad_norm": 0.23475761711597443, "learning_rate": 2.5985532878848174e-05, "loss": 0.0068, "step": 2703 }, { "epoch": 38.62026431718062, "grad_norm": 0.35195475816726685, "learning_rate": 2.592333904928109e-05, "loss": 0.0067, "step": 2704 }, { "epoch": 38.63436123348018, "grad_norm": 0.20397743582725525, "learning_rate": 2.5861208650713588e-05, "loss": 0.0056, "step": 2705 }, { "epoch": 38.648458149779735, "grad_norm": 0.2650679349899292, "learning_rate": 2.579914173634719e-05, "loss": 0.007, "step": 2706 }, { "epoch": 38.66255506607929, "grad_norm": 0.3148314952850342, "learning_rate": 2.5737138359329105e-05, "loss": 0.0058, "step": 2707 }, { "epoch": 38.67665198237886, "grad_norm": 0.22373588383197784, "learning_rate": 2.5675198572752134e-05, "loss": 0.0073, "step": 2708 }, { "epoch": 38.690748898678414, "grad_norm": 0.47256478667259216, "learning_rate": 2.5613322429654574e-05, "loss": 0.0068, "step": 2709 }, { "epoch": 38.70484581497797, "grad_norm": 0.2132301777601242, "learning_rate": 2.555150998302026e-05, "loss": 0.0057, "step": 2710 }, { "epoch": 38.71894273127754, "grad_norm": 0.191494420170784, "learning_rate": 2.5489761285778502e-05, "loss": 0.0057, "step": 2711 }, { "epoch": 38.733039647577094, "grad_norm": 0.20574460923671722, "learning_rate": 2.5428076390804023e-05, "loss": 0.0058, "step": 2712 }, { "epoch": 38.74713656387665, "grad_norm": 0.2055557370185852, "learning_rate": 2.536645535091685e-05, "loss": 0.0057, "step": 2713 }, { "epoch": 38.76123348017621, "grad_norm": 0.154335156083107, "learning_rate": 2.53048982188824e-05, "loss": 0.006, "step": 2714 }, { "epoch": 38.77533039647577, "grad_norm": 0.1868479698896408, "learning_rate": 2.524340504741135e-05, "loss": 0.0052, "step": 2715 }, { "epoch": 38.78942731277533, "grad_norm": 0.09551180154085159, "learning_rate": 2.5181975889159615e-05, "loss": 0.0056, "step": 2716 }, { "epoch": 38.80352422907489, "grad_norm": 0.14299239218235016, "learning_rate": 2.512061079672823e-05, "loss": 0.006, "step": 2717 }, { "epoch": 38.817621145374446, "grad_norm": 0.30871105194091797, "learning_rate": 2.5059309822663457e-05, "loss": 0.0064, "step": 2718 }, { "epoch": 38.83171806167401, "grad_norm": 0.23027952015399933, "learning_rate": 2.4998073019456614e-05, "loss": 0.005, "step": 2719 }, { "epoch": 38.84581497797357, "grad_norm": 0.19178281724452972, "learning_rate": 2.4936900439544107e-05, "loss": 0.0058, "step": 2720 }, { "epoch": 38.859911894273125, "grad_norm": 0.20345382392406464, "learning_rate": 2.4875792135307274e-05, "loss": 0.0059, "step": 2721 }, { "epoch": 38.87400881057269, "grad_norm": 0.22741150856018066, "learning_rate": 2.4814748159072475e-05, "loss": 0.0062, "step": 2722 }, { "epoch": 38.88810572687225, "grad_norm": 0.23995071649551392, "learning_rate": 2.475376856311097e-05, "loss": 0.0071, "step": 2723 }, { "epoch": 38.902202643171805, "grad_norm": 0.23105265200138092, "learning_rate": 2.4692853399638917e-05, "loss": 0.0056, "step": 2724 }, { "epoch": 38.91629955947137, "grad_norm": 0.33738479018211365, "learning_rate": 2.4632002720817228e-05, "loss": 0.006, "step": 2725 }, { "epoch": 38.93039647577093, "grad_norm": 0.26705676317214966, "learning_rate": 2.4571216578751666e-05, "loss": 0.006, "step": 2726 }, { "epoch": 38.944493392070484, "grad_norm": 0.2115054726600647, "learning_rate": 2.4510495025492742e-05, "loss": 0.0053, "step": 2727 }, { "epoch": 38.95859030837004, "grad_norm": 0.2250537872314453, "learning_rate": 2.4449838113035573e-05, "loss": 0.008, "step": 2728 }, { "epoch": 38.972687224669606, "grad_norm": 0.33330774307250977, "learning_rate": 2.438924589332e-05, "loss": 0.0061, "step": 2729 }, { "epoch": 38.986784140969164, "grad_norm": 0.37091150879859924, "learning_rate": 2.432871841823047e-05, "loss": 0.0067, "step": 2730 }, { "epoch": 39.01409691629956, "grad_norm": 0.21391142904758453, "learning_rate": 2.426825573959598e-05, "loss": 0.0116, "step": 2731 }, { "epoch": 39.02819383259912, "grad_norm": 0.1810266077518463, "learning_rate": 2.4207857909189978e-05, "loss": 0.0057, "step": 2732 }, { "epoch": 39.04229074889868, "grad_norm": 0.15593752264976501, "learning_rate": 2.4147524978730485e-05, "loss": 0.0056, "step": 2733 }, { "epoch": 39.05638766519824, "grad_norm": 0.21363723278045654, "learning_rate": 2.4087256999879872e-05, "loss": 0.0059, "step": 2734 }, { "epoch": 39.070484581497794, "grad_norm": 0.18474233150482178, "learning_rate": 2.4027054024244956e-05, "loss": 0.006, "step": 2735 }, { "epoch": 39.08458149779736, "grad_norm": 0.15248070657253265, "learning_rate": 2.3966916103376813e-05, "loss": 0.0055, "step": 2736 }, { "epoch": 39.098678414096916, "grad_norm": 0.14090746641159058, "learning_rate": 2.3906843288770886e-05, "loss": 0.0057, "step": 2737 }, { "epoch": 39.112775330396474, "grad_norm": 0.281452476978302, "learning_rate": 2.384683563186684e-05, "loss": 0.0053, "step": 2738 }, { "epoch": 39.12687224669604, "grad_norm": 0.3689461350440979, "learning_rate": 2.3786893184048565e-05, "loss": 0.0053, "step": 2739 }, { "epoch": 39.140969162995596, "grad_norm": 0.17445966601371765, "learning_rate": 2.372701599664404e-05, "loss": 0.0058, "step": 2740 }, { "epoch": 39.15506607929515, "grad_norm": 0.2680707573890686, "learning_rate": 2.3667204120925456e-05, "loss": 0.0064, "step": 2741 }, { "epoch": 39.16916299559471, "grad_norm": 0.17811763286590576, "learning_rate": 2.3607457608109062e-05, "loss": 0.0058, "step": 2742 }, { "epoch": 39.183259911894275, "grad_norm": 0.49128392338752747, "learning_rate": 2.3547776509355056e-05, "loss": 0.0058, "step": 2743 }, { "epoch": 39.19735682819383, "grad_norm": 0.19114474952220917, "learning_rate": 2.3488160875767717e-05, "loss": 0.0054, "step": 2744 }, { "epoch": 39.21145374449339, "grad_norm": 0.17602548003196716, "learning_rate": 2.342861075839523e-05, "loss": 0.0057, "step": 2745 }, { "epoch": 39.225550660792955, "grad_norm": 0.1698213368654251, "learning_rate": 2.336912620822971e-05, "loss": 0.0066, "step": 2746 }, { "epoch": 39.23964757709251, "grad_norm": 0.39317142963409424, "learning_rate": 2.3309707276207036e-05, "loss": 0.0064, "step": 2747 }, { "epoch": 39.25374449339207, "grad_norm": 0.2234458178281784, "learning_rate": 2.3250354013207e-05, "loss": 0.0052, "step": 2748 }, { "epoch": 39.26784140969163, "grad_norm": 0.0966029018163681, "learning_rate": 2.3191066470053103e-05, "loss": 0.0052, "step": 2749 }, { "epoch": 39.28193832599119, "grad_norm": 0.1713956594467163, "learning_rate": 2.313184469751265e-05, "loss": 0.0056, "step": 2750 }, { "epoch": 39.28193832599119, "eval_loss": 1.4457305669784546, "eval_runtime": 137.3756, "eval_samples_per_second": 13.161, "eval_steps_per_second": 0.415, "step": 2750 }, { "epoch": 39.29603524229075, "grad_norm": 0.20765604078769684, "learning_rate": 2.307268874629649e-05, "loss": 0.0063, "step": 2751 }, { "epoch": 39.31013215859031, "grad_norm": 0.21504372358322144, "learning_rate": 2.3013598667059223e-05, "loss": 0.006, "step": 2752 }, { "epoch": 39.32422907488987, "grad_norm": 0.1965988725423813, "learning_rate": 2.295457451039901e-05, "loss": 0.0055, "step": 2753 }, { "epoch": 39.33832599118943, "grad_norm": 0.3008180558681488, "learning_rate": 2.2895616326857595e-05, "loss": 0.0065, "step": 2754 }, { "epoch": 39.352422907488986, "grad_norm": 0.16718116402626038, "learning_rate": 2.283672416692013e-05, "loss": 0.0057, "step": 2755 }, { "epoch": 39.36651982378854, "grad_norm": 0.17545385658740997, "learning_rate": 2.2777898081015346e-05, "loss": 0.0058, "step": 2756 }, { "epoch": 39.38061674008811, "grad_norm": 0.19107350707054138, "learning_rate": 2.271913811951537e-05, "loss": 0.0056, "step": 2757 }, { "epoch": 39.394713656387665, "grad_norm": 0.14596544206142426, "learning_rate": 2.266044433273562e-05, "loss": 0.0055, "step": 2758 }, { "epoch": 39.40881057268722, "grad_norm": 0.21374908089637756, "learning_rate": 2.260181677093496e-05, "loss": 0.0056, "step": 2759 }, { "epoch": 39.42290748898679, "grad_norm": 0.27169856429100037, "learning_rate": 2.2543255484315484e-05, "loss": 0.0057, "step": 2760 }, { "epoch": 39.437004405286345, "grad_norm": 0.2053861916065216, "learning_rate": 2.2484760523022607e-05, "loss": 0.0055, "step": 2761 }, { "epoch": 39.4511013215859, "grad_norm": 0.1934477984905243, "learning_rate": 2.242633193714482e-05, "loss": 0.0059, "step": 2762 }, { "epoch": 39.46519823788546, "grad_norm": 0.3475813865661621, "learning_rate": 2.2367969776713883e-05, "loss": 0.0064, "step": 2763 }, { "epoch": 39.479295154185024, "grad_norm": 0.12988074123859406, "learning_rate": 2.2309674091704645e-05, "loss": 0.0053, "step": 2764 }, { "epoch": 39.49339207048458, "grad_norm": 0.19085194170475006, "learning_rate": 2.2251444932035094e-05, "loss": 0.0056, "step": 2765 }, { "epoch": 39.50748898678414, "grad_norm": 0.22306379675865173, "learning_rate": 2.2193282347566103e-05, "loss": 0.0053, "step": 2766 }, { "epoch": 39.521585903083704, "grad_norm": 0.14915184676647186, "learning_rate": 2.213518638810167e-05, "loss": 0.0054, "step": 2767 }, { "epoch": 39.53568281938326, "grad_norm": 0.14616067707538605, "learning_rate": 2.207715710338871e-05, "loss": 0.0053, "step": 2768 }, { "epoch": 39.54977973568282, "grad_norm": 0.27523353695869446, "learning_rate": 2.201919454311706e-05, "loss": 0.0058, "step": 2769 }, { "epoch": 39.563876651982376, "grad_norm": 0.21638405323028564, "learning_rate": 2.1961298756919335e-05, "loss": 0.0055, "step": 2770 }, { "epoch": 39.57797356828194, "grad_norm": 0.10188217461109161, "learning_rate": 2.190346979437108e-05, "loss": 0.0054, "step": 2771 }, { "epoch": 39.5920704845815, "grad_norm": 0.23783938586711884, "learning_rate": 2.184570770499056e-05, "loss": 0.0061, "step": 2772 }, { "epoch": 39.606167400881056, "grad_norm": 0.19334082305431366, "learning_rate": 2.1788012538238834e-05, "loss": 0.0056, "step": 2773 }, { "epoch": 39.62026431718062, "grad_norm": 0.3164243698120117, "learning_rate": 2.173038434351955e-05, "loss": 0.0065, "step": 2774 }, { "epoch": 39.63436123348018, "grad_norm": 0.2850169539451599, "learning_rate": 2.16728231701791e-05, "loss": 0.0064, "step": 2775 }, { "epoch": 39.648458149779735, "grad_norm": 0.25633805990219116, "learning_rate": 2.1615329067506505e-05, "loss": 0.0057, "step": 2776 }, { "epoch": 39.66255506607929, "grad_norm": 0.190456360578537, "learning_rate": 2.155790208473324e-05, "loss": 0.006, "step": 2777 }, { "epoch": 39.67665198237886, "grad_norm": 0.16425979137420654, "learning_rate": 2.1500542271033396e-05, "loss": 0.0055, "step": 2778 }, { "epoch": 39.690748898678414, "grad_norm": 0.17328476905822754, "learning_rate": 2.1443249675523536e-05, "loss": 0.0059, "step": 2779 }, { "epoch": 39.70484581497797, "grad_norm": 0.18468528985977173, "learning_rate": 2.138602434726269e-05, "loss": 0.0056, "step": 2780 }, { "epoch": 39.71894273127754, "grad_norm": 0.10446669161319733, "learning_rate": 2.1328866335252183e-05, "loss": 0.0051, "step": 2781 }, { "epoch": 39.733039647577094, "grad_norm": 0.17997783422470093, "learning_rate": 2.1271775688435803e-05, "loss": 0.0054, "step": 2782 }, { "epoch": 39.74713656387665, "grad_norm": 0.2174323946237564, "learning_rate": 2.1214752455699627e-05, "loss": 0.0058, "step": 2783 }, { "epoch": 39.76123348017621, "grad_norm": 0.2618938684463501, "learning_rate": 2.1157796685872012e-05, "loss": 0.0067, "step": 2784 }, { "epoch": 39.77533039647577, "grad_norm": 0.09625861048698425, "learning_rate": 2.11009084277235e-05, "loss": 0.0049, "step": 2785 }, { "epoch": 39.78942731277533, "grad_norm": 0.17363327741622925, "learning_rate": 2.1044087729966856e-05, "loss": 0.0056, "step": 2786 }, { "epoch": 39.80352422907489, "grad_norm": 0.22528056800365448, "learning_rate": 2.0987334641257017e-05, "loss": 0.0059, "step": 2787 }, { "epoch": 39.817621145374446, "grad_norm": 0.23196683824062347, "learning_rate": 2.0930649210191033e-05, "loss": 0.006, "step": 2788 }, { "epoch": 39.83171806167401, "grad_norm": 0.13433155417442322, "learning_rate": 2.0874031485307922e-05, "loss": 0.0053, "step": 2789 }, { "epoch": 39.84581497797357, "grad_norm": 0.29440590739250183, "learning_rate": 2.0817481515088833e-05, "loss": 0.007, "step": 2790 }, { "epoch": 39.859911894273125, "grad_norm": 0.23024556040763855, "learning_rate": 2.076099934795689e-05, "loss": 0.0067, "step": 2791 }, { "epoch": 39.87400881057269, "grad_norm": 0.2259541153907776, "learning_rate": 2.070458503227707e-05, "loss": 0.0056, "step": 2792 }, { "epoch": 39.88810572687225, "grad_norm": 0.19678209722042084, "learning_rate": 2.0648238616356332e-05, "loss": 0.0058, "step": 2793 }, { "epoch": 39.902202643171805, "grad_norm": 0.1959572583436966, "learning_rate": 2.059196014844347e-05, "loss": 0.0058, "step": 2794 }, { "epoch": 39.91629955947137, "grad_norm": 0.2963307797908783, "learning_rate": 2.053574967672912e-05, "loss": 0.0061, "step": 2795 }, { "epoch": 39.93039647577093, "grad_norm": 0.17915739119052887, "learning_rate": 2.0479607249345624e-05, "loss": 0.0061, "step": 2796 }, { "epoch": 39.944493392070484, "grad_norm": 0.22679711878299713, "learning_rate": 2.0423532914367116e-05, "loss": 0.0061, "step": 2797 }, { "epoch": 39.95859030837004, "grad_norm": 0.26272404193878174, "learning_rate": 2.0367526719809414e-05, "loss": 0.0065, "step": 2798 }, { "epoch": 39.972687224669606, "grad_norm": 0.20853693783283234, "learning_rate": 2.0311588713630013e-05, "loss": 0.0062, "step": 2799 }, { "epoch": 39.986784140969164, "grad_norm": 0.2960755228996277, "learning_rate": 2.025571894372794e-05, "loss": 0.0058, "step": 2800 }, { "epoch": 40.01409691629956, "grad_norm": 0.32903534173965454, "learning_rate": 2.0199917457943853e-05, "loss": 0.0119, "step": 2801 }, { "epoch": 40.02819383259912, "grad_norm": 0.2345069944858551, "learning_rate": 2.0144184304059945e-05, "loss": 0.0059, "step": 2802 }, { "epoch": 40.04229074889868, "grad_norm": 0.19072334468364716, "learning_rate": 2.008851952979989e-05, "loss": 0.005, "step": 2803 }, { "epoch": 40.05638766519824, "grad_norm": 0.10397754609584808, "learning_rate": 2.0032923182828757e-05, "loss": 0.0053, "step": 2804 }, { "epoch": 40.070484581497794, "grad_norm": 0.16157592833042145, "learning_rate": 1.9977395310753087e-05, "loss": 0.0057, "step": 2805 }, { "epoch": 40.08458149779736, "grad_norm": 0.16149640083312988, "learning_rate": 1.9921935961120762e-05, "loss": 0.0055, "step": 2806 }, { "epoch": 40.098678414096916, "grad_norm": 0.20070068538188934, "learning_rate": 1.9866545181421013e-05, "loss": 0.0053, "step": 2807 }, { "epoch": 40.112775330396474, "grad_norm": 0.08771330863237381, "learning_rate": 1.9811223019084303e-05, "loss": 0.0051, "step": 2808 }, { "epoch": 40.12687224669604, "grad_norm": 0.10338329523801804, "learning_rate": 1.9755969521482353e-05, "loss": 0.0054, "step": 2809 }, { "epoch": 40.140969162995596, "grad_norm": 0.22595001757144928, "learning_rate": 1.970078473592817e-05, "loss": 0.0059, "step": 2810 }, { "epoch": 40.15506607929515, "grad_norm": 0.233780637383461, "learning_rate": 1.9645668709675757e-05, "loss": 0.0055, "step": 2811 }, { "epoch": 40.16916299559471, "grad_norm": 0.16045378148555756, "learning_rate": 1.9590621489920414e-05, "loss": 0.0049, "step": 2812 }, { "epoch": 40.183259911894275, "grad_norm": 0.1296129673719406, "learning_rate": 1.9535643123798454e-05, "loss": 0.005, "step": 2813 }, { "epoch": 40.19735682819383, "grad_norm": 0.11536107212305069, "learning_rate": 1.9480733658387175e-05, "loss": 0.0055, "step": 2814 }, { "epoch": 40.21145374449339, "grad_norm": 0.1822543740272522, "learning_rate": 1.9425893140704943e-05, "loss": 0.0056, "step": 2815 }, { "epoch": 40.225550660792955, "grad_norm": 0.14021390676498413, "learning_rate": 1.9371121617711085e-05, "loss": 0.0054, "step": 2816 }, { "epoch": 40.23964757709251, "grad_norm": 0.19843703508377075, "learning_rate": 1.931641913630583e-05, "loss": 0.0057, "step": 2817 }, { "epoch": 40.25374449339207, "grad_norm": 0.18687653541564941, "learning_rate": 1.9261785743330265e-05, "loss": 0.0059, "step": 2818 }, { "epoch": 40.26784140969163, "grad_norm": 0.1805855929851532, "learning_rate": 1.920722148556634e-05, "loss": 0.006, "step": 2819 }, { "epoch": 40.28193832599119, "grad_norm": 0.20184938609600067, "learning_rate": 1.915272640973683e-05, "loss": 0.0051, "step": 2820 }, { "epoch": 40.29603524229075, "grad_norm": 0.1798013597726822, "learning_rate": 1.9098300562505266e-05, "loss": 0.0057, "step": 2821 }, { "epoch": 40.31013215859031, "grad_norm": 0.19435690343379974, "learning_rate": 1.9043943990475822e-05, "loss": 0.0052, "step": 2822 }, { "epoch": 40.32422907488987, "grad_norm": 0.1823473423719406, "learning_rate": 1.8989656740193427e-05, "loss": 0.0054, "step": 2823 }, { "epoch": 40.33832599118943, "grad_norm": 0.1832914650440216, "learning_rate": 1.893543885814366e-05, "loss": 0.0059, "step": 2824 }, { "epoch": 40.352422907488986, "grad_norm": 0.16824114322662354, "learning_rate": 1.888129039075268e-05, "loss": 0.0052, "step": 2825 }, { "epoch": 40.36651982378854, "grad_norm": 0.17591902613639832, "learning_rate": 1.882721138438717e-05, "loss": 0.0065, "step": 2826 }, { "epoch": 40.38061674008811, "grad_norm": 0.27327805757522583, "learning_rate": 1.8773201885354362e-05, "loss": 0.0057, "step": 2827 }, { "epoch": 40.394713656387665, "grad_norm": 0.20753471553325653, "learning_rate": 1.871926193990202e-05, "loss": 0.0056, "step": 2828 }, { "epoch": 40.40881057268722, "grad_norm": 0.25241586565971375, "learning_rate": 1.8665391594218262e-05, "loss": 0.0055, "step": 2829 }, { "epoch": 40.42290748898679, "grad_norm": 0.3885049521923065, "learning_rate": 1.8611590894431653e-05, "loss": 0.0052, "step": 2830 }, { "epoch": 40.437004405286345, "grad_norm": 0.14994610846042633, "learning_rate": 1.855785988661113e-05, "loss": 0.005, "step": 2831 }, { "epoch": 40.4511013215859, "grad_norm": 0.1499677300453186, "learning_rate": 1.8504198616765956e-05, "loss": 0.0063, "step": 2832 }, { "epoch": 40.46519823788546, "grad_norm": 0.2036386877298355, "learning_rate": 1.8450607130845633e-05, "loss": 0.0062, "step": 2833 }, { "epoch": 40.479295154185024, "grad_norm": 0.21759121119976044, "learning_rate": 1.839708547473995e-05, "loss": 0.0055, "step": 2834 }, { "epoch": 40.49339207048458, "grad_norm": 0.20345331728458405, "learning_rate": 1.8343633694278895e-05, "loss": 0.005, "step": 2835 }, { "epoch": 40.50748898678414, "grad_norm": 0.11861360818147659, "learning_rate": 1.8290251835232653e-05, "loss": 0.0054, "step": 2836 }, { "epoch": 40.521585903083704, "grad_norm": 0.14915995299816132, "learning_rate": 1.8236939943311426e-05, "loss": 0.0054, "step": 2837 }, { "epoch": 40.53568281938326, "grad_norm": 0.167075976729393, "learning_rate": 1.818369806416562e-05, "loss": 0.0052, "step": 2838 }, { "epoch": 40.54977973568282, "grad_norm": 0.1506580412387848, "learning_rate": 1.813052624338566e-05, "loss": 0.0052, "step": 2839 }, { "epoch": 40.563876651982376, "grad_norm": 0.2320716381072998, "learning_rate": 1.807742452650196e-05, "loss": 0.0059, "step": 2840 }, { "epoch": 40.57797356828194, "grad_norm": 0.18333980441093445, "learning_rate": 1.8024392958984903e-05, "loss": 0.0062, "step": 2841 }, { "epoch": 40.5920704845815, "grad_norm": 0.2327832281589508, "learning_rate": 1.7971431586244815e-05, "loss": 0.0061, "step": 2842 }, { "epoch": 40.606167400881056, "grad_norm": 0.2264801412820816, "learning_rate": 1.7918540453631948e-05, "loss": 0.0052, "step": 2843 }, { "epoch": 40.62026431718062, "grad_norm": 0.1389697790145874, "learning_rate": 1.7865719606436328e-05, "loss": 0.0056, "step": 2844 }, { "epoch": 40.63436123348018, "grad_norm": 0.2619597613811493, "learning_rate": 1.7812969089887864e-05, "loss": 0.0052, "step": 2845 }, { "epoch": 40.648458149779735, "grad_norm": 0.13600577414035797, "learning_rate": 1.7760288949156224e-05, "loss": 0.0066, "step": 2846 }, { "epoch": 40.66255506607929, "grad_norm": 0.24489852786064148, "learning_rate": 1.770767922935084e-05, "loss": 0.0057, "step": 2847 }, { "epoch": 40.67665198237886, "grad_norm": 0.19103917479515076, "learning_rate": 1.765513997552075e-05, "loss": 0.0056, "step": 2848 }, { "epoch": 40.690748898678414, "grad_norm": 0.29392340779304504, "learning_rate": 1.7602671232654754e-05, "loss": 0.0064, "step": 2849 }, { "epoch": 40.70484581497797, "grad_norm": 0.2708081603050232, "learning_rate": 1.7550273045681242e-05, "loss": 0.006, "step": 2850 }, { "epoch": 40.71894273127754, "grad_norm": 0.2684490978717804, "learning_rate": 1.749794545946821e-05, "loss": 0.0065, "step": 2851 }, { "epoch": 40.733039647577094, "grad_norm": 0.3061693012714386, "learning_rate": 1.744568851882311e-05, "loss": 0.0059, "step": 2852 }, { "epoch": 40.74713656387665, "grad_norm": 0.2095465064048767, "learning_rate": 1.739350226849299e-05, "loss": 0.0058, "step": 2853 }, { "epoch": 40.76123348017621, "grad_norm": 0.16517797112464905, "learning_rate": 1.7341386753164358e-05, "loss": 0.0052, "step": 2854 }, { "epoch": 40.77533039647577, "grad_norm": 0.15087895095348358, "learning_rate": 1.7289342017463138e-05, "loss": 0.0061, "step": 2855 }, { "epoch": 40.78942731277533, "grad_norm": 0.25667980313301086, "learning_rate": 1.723736810595461e-05, "loss": 0.0056, "step": 2856 }, { "epoch": 40.80352422907489, "grad_norm": 0.18664833903312683, "learning_rate": 1.7185465063143458e-05, "loss": 0.0059, "step": 2857 }, { "epoch": 40.817621145374446, "grad_norm": 0.20639033615589142, "learning_rate": 1.7133632933473675e-05, "loss": 0.0061, "step": 2858 }, { "epoch": 40.83171806167401, "grad_norm": 0.3244227468967438, "learning_rate": 1.7081871761328528e-05, "loss": 0.0052, "step": 2859 }, { "epoch": 40.84581497797357, "grad_norm": 0.18346798419952393, "learning_rate": 1.7030181591030493e-05, "loss": 0.0067, "step": 2860 }, { "epoch": 40.859911894273125, "grad_norm": 0.2464415431022644, "learning_rate": 1.6978562466841274e-05, "loss": 0.0065, "step": 2861 }, { "epoch": 40.87400881057269, "grad_norm": 0.319151908159256, "learning_rate": 1.6927014432961773e-05, "loss": 0.0062, "step": 2862 }, { "epoch": 40.88810572687225, "grad_norm": 0.22158381342887878, "learning_rate": 1.6875537533531948e-05, "loss": 0.0054, "step": 2863 }, { "epoch": 40.902202643171805, "grad_norm": 0.18050429224967957, "learning_rate": 1.6824131812630882e-05, "loss": 0.0053, "step": 2864 }, { "epoch": 40.91629955947137, "grad_norm": 0.24506382644176483, "learning_rate": 1.6772797314276713e-05, "loss": 0.0051, "step": 2865 }, { "epoch": 40.93039647577093, "grad_norm": 0.13488171994686127, "learning_rate": 1.6721534082426628e-05, "loss": 0.0076, "step": 2866 }, { "epoch": 40.944493392070484, "grad_norm": 0.36154961585998535, "learning_rate": 1.667034216097667e-05, "loss": 0.0056, "step": 2867 }, { "epoch": 40.95859030837004, "grad_norm": 0.1861923784017563, "learning_rate": 1.6619221593761947e-05, "loss": 0.0055, "step": 2868 }, { "epoch": 40.972687224669606, "grad_norm": 0.18877923488616943, "learning_rate": 1.6568172424556405e-05, "loss": 0.0071, "step": 2869 }, { "epoch": 40.986784140969164, "grad_norm": 0.4237464666366577, "learning_rate": 1.65171946970729e-05, "loss": 0.0067, "step": 2870 }, { "epoch": 41.01409691629956, "grad_norm": 0.14356425404548645, "learning_rate": 1.6466288454963018e-05, "loss": 0.0093, "step": 2871 }, { "epoch": 41.02819383259912, "grad_norm": 0.13157550990581512, "learning_rate": 1.6415453741817212e-05, "loss": 0.0053, "step": 2872 }, { "epoch": 41.04229074889868, "grad_norm": 0.19610707461833954, "learning_rate": 1.636469060116469e-05, "loss": 0.0051, "step": 2873 }, { "epoch": 41.05638766519824, "grad_norm": 0.16429096460342407, "learning_rate": 1.631399907647335e-05, "loss": 0.0052, "step": 2874 }, { "epoch": 41.070484581497794, "grad_norm": 0.11986181139945984, "learning_rate": 1.6263379211149732e-05, "loss": 0.0052, "step": 2875 }, { "epoch": 41.08458149779736, "grad_norm": 0.17189113795757294, "learning_rate": 1.6212831048539067e-05, "loss": 0.0054, "step": 2876 }, { "epoch": 41.098678414096916, "grad_norm": 0.13554032146930695, "learning_rate": 1.6162354631925204e-05, "loss": 0.0051, "step": 2877 }, { "epoch": 41.112775330396474, "grad_norm": 0.12144771218299866, "learning_rate": 1.611195000453046e-05, "loss": 0.0049, "step": 2878 }, { "epoch": 41.12687224669604, "grad_norm": 0.12028533965349197, "learning_rate": 1.606161720951579e-05, "loss": 0.0049, "step": 2879 }, { "epoch": 41.140969162995596, "grad_norm": 0.1070239320397377, "learning_rate": 1.601135628998056e-05, "loss": 0.0053, "step": 2880 }, { "epoch": 41.15506607929515, "grad_norm": 0.25527846813201904, "learning_rate": 1.5961167288962663e-05, "loss": 0.0057, "step": 2881 }, { "epoch": 41.16916299559471, "grad_norm": 0.19516222178936005, "learning_rate": 1.591105024943832e-05, "loss": 0.0056, "step": 2882 }, { "epoch": 41.183259911894275, "grad_norm": 0.1433981955051422, "learning_rate": 1.58610052143222e-05, "loss": 0.0059, "step": 2883 }, { "epoch": 41.19735682819383, "grad_norm": 0.2030337154865265, "learning_rate": 1.5811032226467305e-05, "loss": 0.0053, "step": 2884 }, { "epoch": 41.21145374449339, "grad_norm": 0.09888993948698044, "learning_rate": 1.5761131328664947e-05, "loss": 0.0051, "step": 2885 }, { "epoch": 41.225550660792955, "grad_norm": 0.16769817471504211, "learning_rate": 1.571130256364466e-05, "loss": 0.0054, "step": 2886 }, { "epoch": 41.23964757709251, "grad_norm": 0.13499030470848083, "learning_rate": 1.5661545974074254e-05, "loss": 0.0055, "step": 2887 }, { "epoch": 41.25374449339207, "grad_norm": 0.2865696847438812, "learning_rate": 1.5611861602559742e-05, "loss": 0.006, "step": 2888 }, { "epoch": 41.26784140969163, "grad_norm": 0.2179524004459381, "learning_rate": 1.5562249491645298e-05, "loss": 0.0055, "step": 2889 }, { "epoch": 41.28193832599119, "grad_norm": 0.13712427020072937, "learning_rate": 1.5512709683813164e-05, "loss": 0.0056, "step": 2890 }, { "epoch": 41.29603524229075, "grad_norm": 0.28491440415382385, "learning_rate": 1.5463242221483743e-05, "loss": 0.005, "step": 2891 }, { "epoch": 41.31013215859031, "grad_norm": 0.12728430330753326, "learning_rate": 1.541384714701548e-05, "loss": 0.0056, "step": 2892 }, { "epoch": 41.32422907488987, "grad_norm": 0.16567613184452057, "learning_rate": 1.5364524502704758e-05, "loss": 0.0055, "step": 2893 }, { "epoch": 41.33832599118943, "grad_norm": 0.2039705365896225, "learning_rate": 1.531527433078602e-05, "loss": 0.0056, "step": 2894 }, { "epoch": 41.352422907488986, "grad_norm": 0.1972440928220749, "learning_rate": 1.526609667343163e-05, "loss": 0.0048, "step": 2895 }, { "epoch": 41.36651982378854, "grad_norm": 0.19511409103870392, "learning_rate": 1.5216991572751882e-05, "loss": 0.0053, "step": 2896 }, { "epoch": 41.38061674008811, "grad_norm": 0.15538185834884644, "learning_rate": 1.5167959070794857e-05, "loss": 0.0055, "step": 2897 }, { "epoch": 41.394713656387665, "grad_norm": 0.14551900327205658, "learning_rate": 1.5118999209546559e-05, "loss": 0.005, "step": 2898 }, { "epoch": 41.40881057268722, "grad_norm": 0.19013819098472595, "learning_rate": 1.5070112030930761e-05, "loss": 0.006, "step": 2899 }, { "epoch": 41.42290748898679, "grad_norm": 0.24738967418670654, "learning_rate": 1.5021297576809001e-05, "loss": 0.005, "step": 2900 }, { "epoch": 41.437004405286345, "grad_norm": 0.13636180758476257, "learning_rate": 1.497255588898051e-05, "loss": 0.0054, "step": 2901 }, { "epoch": 41.4511013215859, "grad_norm": 0.2341054379940033, "learning_rate": 1.4923887009182258e-05, "loss": 0.0056, "step": 2902 }, { "epoch": 41.46519823788546, "grad_norm": 0.1631186157464981, "learning_rate": 1.4875290979088852e-05, "loss": 0.0052, "step": 2903 }, { "epoch": 41.479295154185024, "grad_norm": 0.14380085468292236, "learning_rate": 1.4826767840312539e-05, "loss": 0.0056, "step": 2904 }, { "epoch": 41.49339207048458, "grad_norm": 0.17979268729686737, "learning_rate": 1.4778317634403083e-05, "loss": 0.0059, "step": 2905 }, { "epoch": 41.50748898678414, "grad_norm": 0.1998508870601654, "learning_rate": 1.4729940402847874e-05, "loss": 0.005, "step": 2906 }, { "epoch": 41.521585903083704, "grad_norm": 0.14465568959712982, "learning_rate": 1.4681636187071769e-05, "loss": 0.0053, "step": 2907 }, { "epoch": 41.53568281938326, "grad_norm": 0.235871821641922, "learning_rate": 1.4633405028437152e-05, "loss": 0.0056, "step": 2908 }, { "epoch": 41.54977973568282, "grad_norm": 0.18092119693756104, "learning_rate": 1.4585246968243748e-05, "loss": 0.006, "step": 2909 }, { "epoch": 41.563876651982376, "grad_norm": 0.2576563358306885, "learning_rate": 1.4537162047728804e-05, "loss": 0.0049, "step": 2910 }, { "epoch": 41.57797356828194, "grad_norm": 0.10252278298139572, "learning_rate": 1.4489150308066901e-05, "loss": 0.0059, "step": 2911 }, { "epoch": 41.5920704845815, "grad_norm": 0.20768144726753235, "learning_rate": 1.444121179036989e-05, "loss": 0.0053, "step": 2912 }, { "epoch": 41.606167400881056, "grad_norm": 0.2129778414964676, "learning_rate": 1.4393346535687013e-05, "loss": 0.0055, "step": 2913 }, { "epoch": 41.62026431718062, "grad_norm": 0.2557953894138336, "learning_rate": 1.434555458500474e-05, "loss": 0.0057, "step": 2914 }, { "epoch": 41.63436123348018, "grad_norm": 0.17295946180820465, "learning_rate": 1.4297835979246777e-05, "loss": 0.0056, "step": 2915 }, { "epoch": 41.648458149779735, "grad_norm": 0.24948449432849884, "learning_rate": 1.4250190759273995e-05, "loss": 0.0053, "step": 2916 }, { "epoch": 41.66255506607929, "grad_norm": 0.12463697791099548, "learning_rate": 1.4202618965884451e-05, "loss": 0.0056, "step": 2917 }, { "epoch": 41.67665198237886, "grad_norm": 0.3447701930999756, "learning_rate": 1.415512063981339e-05, "loss": 0.0058, "step": 2918 }, { "epoch": 41.690748898678414, "grad_norm": 0.14278756082057953, "learning_rate": 1.4107695821733025e-05, "loss": 0.0052, "step": 2919 }, { "epoch": 41.70484581497797, "grad_norm": 0.16867083311080933, "learning_rate": 1.4060344552252702e-05, "loss": 0.0048, "step": 2920 }, { "epoch": 41.71894273127754, "grad_norm": 0.15446732938289642, "learning_rate": 1.4013066871918779e-05, "loss": 0.0057, "step": 2921 }, { "epoch": 41.733039647577094, "grad_norm": 0.15919719636440277, "learning_rate": 1.3965862821214615e-05, "loss": 0.006, "step": 2922 }, { "epoch": 41.74713656387665, "grad_norm": 0.18656526505947113, "learning_rate": 1.3918732440560456e-05, "loss": 0.0057, "step": 2923 }, { "epoch": 41.76123348017621, "grad_norm": 0.18486542999744415, "learning_rate": 1.387167577031353e-05, "loss": 0.0066, "step": 2924 }, { "epoch": 41.77533039647577, "grad_norm": 0.2955957055091858, "learning_rate": 1.3824692850767929e-05, "loss": 0.0056, "step": 2925 }, { "epoch": 41.78942731277533, "grad_norm": 0.20293715596199036, "learning_rate": 1.3777783722154603e-05, "loss": 0.0054, "step": 2926 }, { "epoch": 41.80352422907489, "grad_norm": 0.15566140413284302, "learning_rate": 1.3730948424641276e-05, "loss": 0.0058, "step": 2927 }, { "epoch": 41.817621145374446, "grad_norm": 0.18062597513198853, "learning_rate": 1.3684186998332481e-05, "loss": 0.006, "step": 2928 }, { "epoch": 41.83171806167401, "grad_norm": 0.275473415851593, "learning_rate": 1.3637499483269544e-05, "loss": 0.0053, "step": 2929 }, { "epoch": 41.84581497797357, "grad_norm": 0.19605189561843872, "learning_rate": 1.3590885919430384e-05, "loss": 0.0061, "step": 2930 }, { "epoch": 41.859911894273125, "grad_norm": 0.3227013349533081, "learning_rate": 1.3544346346729685e-05, "loss": 0.0067, "step": 2931 }, { "epoch": 41.87400881057269, "grad_norm": 0.27962446212768555, "learning_rate": 1.349788080501877e-05, "loss": 0.0058, "step": 2932 }, { "epoch": 41.88810572687225, "grad_norm": 0.25402286648750305, "learning_rate": 1.3451489334085554e-05, "loss": 0.0053, "step": 2933 }, { "epoch": 41.902202643171805, "grad_norm": 0.13540636003017426, "learning_rate": 1.3405171973654495e-05, "loss": 0.0059, "step": 2934 }, { "epoch": 41.91629955947137, "grad_norm": 0.21643157303333282, "learning_rate": 1.3358928763386646e-05, "loss": 0.0062, "step": 2935 }, { "epoch": 41.93039647577093, "grad_norm": 0.19060176610946655, "learning_rate": 1.3312759742879521e-05, "loss": 0.0054, "step": 2936 }, { "epoch": 41.944493392070484, "grad_norm": 0.18372926115989685, "learning_rate": 1.3266664951667162e-05, "loss": 0.0054, "step": 2937 }, { "epoch": 41.95859030837004, "grad_norm": 0.2094092220067978, "learning_rate": 1.3220644429219964e-05, "loss": 0.0051, "step": 2938 }, { "epoch": 41.972687224669606, "grad_norm": 0.10103478282690048, "learning_rate": 1.3174698214944802e-05, "loss": 0.0057, "step": 2939 }, { "epoch": 41.986784140969164, "grad_norm": 0.21591974794864655, "learning_rate": 1.3128826348184887e-05, "loss": 0.0052, "step": 2940 }, { "epoch": 42.01409691629956, "grad_norm": 0.12953142821788788, "learning_rate": 1.3083028868219804e-05, "loss": 0.0103, "step": 2941 }, { "epoch": 42.02819383259912, "grad_norm": 0.18716180324554443, "learning_rate": 1.3037305814265366e-05, "loss": 0.0053, "step": 2942 }, { "epoch": 42.04229074889868, "grad_norm": 0.13905547559261322, "learning_rate": 1.2991657225473718e-05, "loss": 0.0051, "step": 2943 }, { "epoch": 42.05638766519824, "grad_norm": 0.24612180888652802, "learning_rate": 1.2946083140933263e-05, "loss": 0.0066, "step": 2944 }, { "epoch": 42.070484581497794, "grad_norm": 0.21470221877098083, "learning_rate": 1.2900583599668515e-05, "loss": 0.0047, "step": 2945 }, { "epoch": 42.08458149779736, "grad_norm": 0.1067691296339035, "learning_rate": 1.285515864064023e-05, "loss": 0.0049, "step": 2946 }, { "epoch": 42.098678414096916, "grad_norm": 0.14876466989517212, "learning_rate": 1.2809808302745297e-05, "loss": 0.0053, "step": 2947 }, { "epoch": 42.112775330396474, "grad_norm": 0.16044555604457855, "learning_rate": 1.2764532624816695e-05, "loss": 0.0054, "step": 2948 }, { "epoch": 42.12687224669604, "grad_norm": 0.21802353858947754, "learning_rate": 1.2719331645623444e-05, "loss": 0.0052, "step": 2949 }, { "epoch": 42.140969162995596, "grad_norm": 0.15310291945934296, "learning_rate": 1.2674205403870642e-05, "loss": 0.0051, "step": 2950 }, { "epoch": 42.15506607929515, "grad_norm": 0.19920718669891357, "learning_rate": 1.2629153938199379e-05, "loss": 0.0053, "step": 2951 }, { "epoch": 42.16916299559471, "grad_norm": 0.2299223393201828, "learning_rate": 1.2584177287186717e-05, "loss": 0.0052, "step": 2952 }, { "epoch": 42.183259911894275, "grad_norm": 0.13478247821331024, "learning_rate": 1.2539275489345625e-05, "loss": 0.0053, "step": 2953 }, { "epoch": 42.19735682819383, "grad_norm": 0.18687324225902557, "learning_rate": 1.2494448583125018e-05, "loss": 0.0055, "step": 2954 }, { "epoch": 42.21145374449339, "grad_norm": 0.264803409576416, "learning_rate": 1.2449696606909655e-05, "loss": 0.0057, "step": 2955 }, { "epoch": 42.225550660792955, "grad_norm": 0.21263718605041504, "learning_rate": 1.2405019599020173e-05, "loss": 0.0053, "step": 2956 }, { "epoch": 42.23964757709251, "grad_norm": 0.17493349313735962, "learning_rate": 1.2360417597712925e-05, "loss": 0.0053, "step": 2957 }, { "epoch": 42.25374449339207, "grad_norm": 0.13423672318458557, "learning_rate": 1.231589064118015e-05, "loss": 0.0055, "step": 2958 }, { "epoch": 42.26784140969163, "grad_norm": 0.2147996425628662, "learning_rate": 1.2271438767549737e-05, "loss": 0.0052, "step": 2959 }, { "epoch": 42.28193832599119, "grad_norm": 0.19962966442108154, "learning_rate": 1.222706201488536e-05, "loss": 0.0051, "step": 2960 }, { "epoch": 42.29603524229075, "grad_norm": 0.3405938446521759, "learning_rate": 1.218276042118629e-05, "loss": 0.0056, "step": 2961 }, { "epoch": 42.31013215859031, "grad_norm": 0.329609215259552, "learning_rate": 1.2138534024387471e-05, "loss": 0.0057, "step": 2962 }, { "epoch": 42.32422907488987, "grad_norm": 0.21350157260894775, "learning_rate": 1.209438286235951e-05, "loss": 0.0051, "step": 2963 }, { "epoch": 42.33832599118943, "grad_norm": 0.10745759308338165, "learning_rate": 1.2050306972908476e-05, "loss": 0.0054, "step": 2964 }, { "epoch": 42.352422907488986, "grad_norm": 0.2160874307155609, "learning_rate": 1.200630639377609e-05, "loss": 0.0051, "step": 2965 }, { "epoch": 42.36651982378854, "grad_norm": 0.1808461993932724, "learning_rate": 1.1962381162639547e-05, "loss": 0.0055, "step": 2966 }, { "epoch": 42.38061674008811, "grad_norm": 0.1749766319990158, "learning_rate": 1.1918531317111537e-05, "loss": 0.0064, "step": 2967 }, { "epoch": 42.394713656387665, "grad_norm": 0.19162903726100922, "learning_rate": 1.1874756894740135e-05, "loss": 0.0052, "step": 2968 }, { "epoch": 42.40881057268722, "grad_norm": 0.40385469794273376, "learning_rate": 1.1831057933008904e-05, "loss": 0.0052, "step": 2969 }, { "epoch": 42.42290748898679, "grad_norm": 0.21826773881912231, "learning_rate": 1.1787434469336766e-05, "loss": 0.0052, "step": 2970 }, { "epoch": 42.437004405286345, "grad_norm": 0.12622348964214325, "learning_rate": 1.1743886541078019e-05, "loss": 0.0051, "step": 2971 }, { "epoch": 42.4511013215859, "grad_norm": 0.17522944509983063, "learning_rate": 1.1700414185522214e-05, "loss": 0.0052, "step": 2972 }, { "epoch": 42.46519823788546, "grad_norm": 0.13414254784584045, "learning_rate": 1.1657017439894247e-05, "loss": 0.0055, "step": 2973 }, { "epoch": 42.479295154185024, "grad_norm": 0.1888834685087204, "learning_rate": 1.1613696341354264e-05, "loss": 0.0053, "step": 2974 }, { "epoch": 42.49339207048458, "grad_norm": 0.15492010116577148, "learning_rate": 1.1570450926997655e-05, "loss": 0.0053, "step": 2975 }, { "epoch": 42.50748898678414, "grad_norm": 0.20626381039619446, "learning_rate": 1.1527281233854914e-05, "loss": 0.0053, "step": 2976 }, { "epoch": 42.521585903083704, "grad_norm": 0.15103289484977722, "learning_rate": 1.1484187298891802e-05, "loss": 0.0063, "step": 2977 }, { "epoch": 42.53568281938326, "grad_norm": 0.24713575839996338, "learning_rate": 1.1441169159009157e-05, "loss": 0.0061, "step": 2978 }, { "epoch": 42.54977973568282, "grad_norm": 0.2022518366575241, "learning_rate": 1.1398226851042903e-05, "loss": 0.0053, "step": 2979 }, { "epoch": 42.563876651982376, "grad_norm": 0.24039119482040405, "learning_rate": 1.135536041176406e-05, "loss": 0.0056, "step": 2980 }, { "epoch": 42.57797356828194, "grad_norm": 0.22834563255310059, "learning_rate": 1.1312569877878664e-05, "loss": 0.0051, "step": 2981 }, { "epoch": 42.5920704845815, "grad_norm": 0.1828175038099289, "learning_rate": 1.1269855286027797e-05, "loss": 0.0049, "step": 2982 }, { "epoch": 42.606167400881056, "grad_norm": 0.15398603677749634, "learning_rate": 1.1227216672787421e-05, "loss": 0.0059, "step": 2983 }, { "epoch": 42.62026431718062, "grad_norm": 0.4454115629196167, "learning_rate": 1.1184654074668521e-05, "loss": 0.0065, "step": 2984 }, { "epoch": 42.63436123348018, "grad_norm": 0.24703869223594666, "learning_rate": 1.1142167528116965e-05, "loss": 0.0057, "step": 2985 }, { "epoch": 42.648458149779735, "grad_norm": 0.19456914067268372, "learning_rate": 1.1099757069513516e-05, "loss": 0.0053, "step": 2986 }, { "epoch": 42.66255506607929, "grad_norm": 0.15528170764446259, "learning_rate": 1.1057422735173728e-05, "loss": 0.005, "step": 2987 }, { "epoch": 42.67665198237886, "grad_norm": 0.09974423050880432, "learning_rate": 1.1015164561348023e-05, "loss": 0.0049, "step": 2988 }, { "epoch": 42.690748898678414, "grad_norm": 0.10822991281747818, "learning_rate": 1.0972982584221592e-05, "loss": 0.005, "step": 2989 }, { "epoch": 42.70484581497797, "grad_norm": 0.15058653056621552, "learning_rate": 1.0930876839914416e-05, "loss": 0.0059, "step": 2990 }, { "epoch": 42.71894273127754, "grad_norm": 0.20429091155529022, "learning_rate": 1.0888847364481113e-05, "loss": 0.005, "step": 2991 }, { "epoch": 42.733039647577094, "grad_norm": 0.1830826848745346, "learning_rate": 1.0846894193911061e-05, "loss": 0.0056, "step": 2992 }, { "epoch": 42.74713656387665, "grad_norm": 0.22417870163917542, "learning_rate": 1.08050173641283e-05, "loss": 0.0053, "step": 2993 }, { "epoch": 42.76123348017621, "grad_norm": 0.12988367676734924, "learning_rate": 1.0763216910991447e-05, "loss": 0.0052, "step": 2994 }, { "epoch": 42.77533039647577, "grad_norm": 0.3119809031486511, "learning_rate": 1.072149287029377e-05, "loss": 0.0055, "step": 2995 }, { "epoch": 42.78942731277533, "grad_norm": 0.24795161187648773, "learning_rate": 1.067984527776309e-05, "loss": 0.0048, "step": 2996 }, { "epoch": 42.80352422907489, "grad_norm": 0.2074027806520462, "learning_rate": 1.0638274169061791e-05, "loss": 0.0055, "step": 2997 }, { "epoch": 42.817621145374446, "grad_norm": 0.158954918384552, "learning_rate": 1.0596779579786686e-05, "loss": 0.0056, "step": 2998 }, { "epoch": 42.83171806167401, "grad_norm": 0.2701192796230316, "learning_rate": 1.0555361545469156e-05, "loss": 0.006, "step": 2999 }, { "epoch": 42.84581497797357, "grad_norm": 0.18455933034420013, "learning_rate": 1.0514020101574972e-05, "loss": 0.0054, "step": 3000 }, { "epoch": 42.84581497797357, "eval_loss": 1.4510937929153442, "eval_runtime": 137.2516, "eval_samples_per_second": 13.173, "eval_steps_per_second": 0.415, "step": 3000 }, { "epoch": 42.859911894273125, "grad_norm": 0.19510111212730408, "learning_rate": 1.0472755283504376e-05, "loss": 0.0053, "step": 3001 }, { "epoch": 42.87400881057269, "grad_norm": 0.14811444282531738, "learning_rate": 1.0431567126591913e-05, "loss": 0.0054, "step": 3002 }, { "epoch": 42.88810572687225, "grad_norm": 0.22263136506080627, "learning_rate": 1.0390455666106547e-05, "loss": 0.0054, "step": 3003 }, { "epoch": 42.902202643171805, "grad_norm": 0.18924176692962646, "learning_rate": 1.0349420937251564e-05, "loss": 0.0056, "step": 3004 }, { "epoch": 42.91629955947137, "grad_norm": 0.19353562593460083, "learning_rate": 1.0308462975164545e-05, "loss": 0.0055, "step": 3005 }, { "epoch": 42.93039647577093, "grad_norm": 0.22553567588329315, "learning_rate": 1.0267581814917305e-05, "loss": 0.0052, "step": 3006 }, { "epoch": 42.944493392070484, "grad_norm": 0.17179977893829346, "learning_rate": 1.0226777491515905e-05, "loss": 0.0057, "step": 3007 }, { "epoch": 42.95859030837004, "grad_norm": 0.11725383251905441, "learning_rate": 1.0186050039900641e-05, "loss": 0.005, "step": 3008 }, { "epoch": 42.972687224669606, "grad_norm": 0.2636050879955292, "learning_rate": 1.0145399494945973e-05, "loss": 0.0054, "step": 3009 }, { "epoch": 42.986784140969164, "grad_norm": 0.17816250026226044, "learning_rate": 1.010482589146048e-05, "loss": 0.0063, "step": 3010 }, { "epoch": 43.01409691629956, "grad_norm": 0.3092566430568695, "learning_rate": 1.0064329264186878e-05, "loss": 0.011, "step": 3011 }, { "epoch": 43.02819383259912, "grad_norm": 0.12105906009674072, "learning_rate": 1.0023909647801988e-05, "loss": 0.0049, "step": 3012 }, { "epoch": 43.04229074889868, "grad_norm": 0.12412483245134354, "learning_rate": 9.983567076916623e-06, "loss": 0.0053, "step": 3013 }, { "epoch": 43.05638766519824, "grad_norm": 0.18184040486812592, "learning_rate": 9.943301586075704e-06, "loss": 0.005, "step": 3014 }, { "epoch": 43.070484581497794, "grad_norm": 0.1560290902853012, "learning_rate": 9.903113209758096e-06, "loss": 0.0046, "step": 3015 }, { "epoch": 43.08458149779736, "grad_norm": 0.12057439982891083, "learning_rate": 9.863001982376662e-06, "loss": 0.0051, "step": 3016 }, { "epoch": 43.098678414096916, "grad_norm": 0.1378868818283081, "learning_rate": 9.822967938278171e-06, "loss": 0.0049, "step": 3017 }, { "epoch": 43.112775330396474, "grad_norm": 0.16920062899589539, "learning_rate": 9.783011111743312e-06, "loss": 0.0052, "step": 3018 }, { "epoch": 43.12687224669604, "grad_norm": 0.14837457239627838, "learning_rate": 9.743131536986682e-06, "loss": 0.0058, "step": 3019 }, { "epoch": 43.140969162995596, "grad_norm": 0.1735960841178894, "learning_rate": 9.703329248156711e-06, "loss": 0.0049, "step": 3020 }, { "epoch": 43.15506607929515, "grad_norm": 0.1541978120803833, "learning_rate": 9.663604279335614e-06, "loss": 0.0049, "step": 3021 }, { "epoch": 43.16916299559471, "grad_norm": 0.150648832321167, "learning_rate": 9.623956664539435e-06, "loss": 0.0053, "step": 3022 }, { "epoch": 43.183259911894275, "grad_norm": 0.18879033625125885, "learning_rate": 9.584386437717997e-06, "loss": 0.006, "step": 3023 }, { "epoch": 43.19735682819383, "grad_norm": 0.2154386043548584, "learning_rate": 9.544893632754814e-06, "loss": 0.0053, "step": 3024 }, { "epoch": 43.21145374449339, "grad_norm": 0.11177663505077362, "learning_rate": 9.505478283467128e-06, "loss": 0.0049, "step": 3025 }, { "epoch": 43.225550660792955, "grad_norm": 0.15984982252120972, "learning_rate": 9.466140423605851e-06, "loss": 0.0056, "step": 3026 }, { "epoch": 43.23964757709251, "grad_norm": 0.200510174036026, "learning_rate": 9.426880086855583e-06, "loss": 0.0053, "step": 3027 }, { "epoch": 43.25374449339207, "grad_norm": 0.16160137951374054, "learning_rate": 9.387697306834431e-06, "loss": 0.0052, "step": 3028 }, { "epoch": 43.26784140969163, "grad_norm": 0.20187309384346008, "learning_rate": 9.348592117094213e-06, "loss": 0.0051, "step": 3029 }, { "epoch": 43.28193832599119, "grad_norm": 0.18963110446929932, "learning_rate": 9.309564551120254e-06, "loss": 0.0052, "step": 3030 }, { "epoch": 43.29603524229075, "grad_norm": 0.16687248647212982, "learning_rate": 9.270614642331376e-06, "loss": 0.0052, "step": 3031 }, { "epoch": 43.31013215859031, "grad_norm": 0.14888469874858856, "learning_rate": 9.231742424079981e-06, "loss": 0.0051, "step": 3032 }, { "epoch": 43.32422907488987, "grad_norm": 0.1738998293876648, "learning_rate": 9.192947929651885e-06, "loss": 0.0053, "step": 3033 }, { "epoch": 43.33832599118943, "grad_norm": 0.10821952670812607, "learning_rate": 9.154231192266394e-06, "loss": 0.005, "step": 3034 }, { "epoch": 43.352422907488986, "grad_norm": 0.2462385892868042, "learning_rate": 9.115592245076177e-06, "loss": 0.0055, "step": 3035 }, { "epoch": 43.36651982378854, "grad_norm": 0.13557825982570648, "learning_rate": 9.077031121167346e-06, "loss": 0.0058, "step": 3036 }, { "epoch": 43.38061674008811, "grad_norm": 0.24745112657546997, "learning_rate": 9.038547853559331e-06, "loss": 0.0059, "step": 3037 }, { "epoch": 43.394713656387665, "grad_norm": 0.14951689541339874, "learning_rate": 9.000142475204964e-06, "loss": 0.0053, "step": 3038 }, { "epoch": 43.40881057268722, "grad_norm": 0.11930710077285767, "learning_rate": 8.96181501899027e-06, "loss": 0.0051, "step": 3039 }, { "epoch": 43.42290748898679, "grad_norm": 0.23022764921188354, "learning_rate": 8.923565517734633e-06, "loss": 0.0053, "step": 3040 }, { "epoch": 43.437004405286345, "grad_norm": 0.23991283774375916, "learning_rate": 8.885394004190684e-06, "loss": 0.0053, "step": 3041 }, { "epoch": 43.4511013215859, "grad_norm": 0.2465878427028656, "learning_rate": 8.847300511044255e-06, "loss": 0.005, "step": 3042 }, { "epoch": 43.46519823788546, "grad_norm": 0.11751066893339157, "learning_rate": 8.80928507091433e-06, "loss": 0.0054, "step": 3043 }, { "epoch": 43.479295154185024, "grad_norm": 0.19513565301895142, "learning_rate": 8.77134771635314e-06, "loss": 0.0052, "step": 3044 }, { "epoch": 43.49339207048458, "grad_norm": 0.16935032606124878, "learning_rate": 8.733488479845997e-06, "loss": 0.0052, "step": 3045 }, { "epoch": 43.50748898678414, "grad_norm": 0.2615572512149811, "learning_rate": 8.695707393811314e-06, "loss": 0.0056, "step": 3046 }, { "epoch": 43.521585903083704, "grad_norm": 0.22485290467739105, "learning_rate": 8.658004490600612e-06, "loss": 0.0053, "step": 3047 }, { "epoch": 43.53568281938326, "grad_norm": 0.2470780462026596, "learning_rate": 8.620379802498468e-06, "loss": 0.0053, "step": 3048 }, { "epoch": 43.54977973568282, "grad_norm": 0.13749882578849792, "learning_rate": 8.582833361722487e-06, "loss": 0.0053, "step": 3049 }, { "epoch": 43.563876651982376, "grad_norm": 0.3178144097328186, "learning_rate": 8.545365200423217e-06, "loss": 0.006, "step": 3050 }, { "epoch": 43.57797356828194, "grad_norm": 0.15958961844444275, "learning_rate": 8.50797535068424e-06, "loss": 0.0055, "step": 3051 }, { "epoch": 43.5920704845815, "grad_norm": 0.15888915956020355, "learning_rate": 8.470663844522052e-06, "loss": 0.0052, "step": 3052 }, { "epoch": 43.606167400881056, "grad_norm": 0.20948021113872528, "learning_rate": 8.433430713886092e-06, "loss": 0.005, "step": 3053 }, { "epoch": 43.62026431718062, "grad_norm": 0.24396541714668274, "learning_rate": 8.396275990658632e-06, "loss": 0.005, "step": 3054 }, { "epoch": 43.63436123348018, "grad_norm": 0.1900947391986847, "learning_rate": 8.359199706654853e-06, "loss": 0.0058, "step": 3055 }, { "epoch": 43.648458149779735, "grad_norm": 0.2802414000034332, "learning_rate": 8.32220189362276e-06, "loss": 0.0051, "step": 3056 }, { "epoch": 43.66255506607929, "grad_norm": 0.1280800849199295, "learning_rate": 8.28528258324316e-06, "loss": 0.0048, "step": 3057 }, { "epoch": 43.67665198237886, "grad_norm": 0.10064869374036789, "learning_rate": 8.248441807129614e-06, "loss": 0.0054, "step": 3058 }, { "epoch": 43.690748898678414, "grad_norm": 0.21400988101959229, "learning_rate": 8.21167959682848e-06, "loss": 0.0054, "step": 3059 }, { "epoch": 43.70484581497797, "grad_norm": 0.19073961675167084, "learning_rate": 8.1749959838188e-06, "loss": 0.0058, "step": 3060 }, { "epoch": 43.71894273127754, "grad_norm": 0.19426429271697998, "learning_rate": 8.13839099951238e-06, "loss": 0.005, "step": 3061 }, { "epoch": 43.733039647577094, "grad_norm": 0.1414536088705063, "learning_rate": 8.101864675253589e-06, "loss": 0.0056, "step": 3062 }, { "epoch": 43.74713656387665, "grad_norm": 0.23450967669487, "learning_rate": 8.065417042319523e-06, "loss": 0.0051, "step": 3063 }, { "epoch": 43.76123348017621, "grad_norm": 0.09095324575901031, "learning_rate": 8.029048131919903e-06, "loss": 0.0054, "step": 3064 }, { "epoch": 43.77533039647577, "grad_norm": 0.16620582342147827, "learning_rate": 7.992757975196974e-06, "loss": 0.0048, "step": 3065 }, { "epoch": 43.78942731277533, "grad_norm": 0.14026495814323425, "learning_rate": 7.956546603225601e-06, "loss": 0.0049, "step": 3066 }, { "epoch": 43.80352422907489, "grad_norm": 0.14409665763378143, "learning_rate": 7.920414047013159e-06, "loss": 0.0055, "step": 3067 }, { "epoch": 43.817621145374446, "grad_norm": 0.1226925253868103, "learning_rate": 7.884360337499574e-06, "loss": 0.0056, "step": 3068 }, { "epoch": 43.83171806167401, "grad_norm": 0.2062675803899765, "learning_rate": 7.848385505557176e-06, "loss": 0.0054, "step": 3069 }, { "epoch": 43.84581497797357, "grad_norm": 0.15026704967021942, "learning_rate": 7.812489581990844e-06, "loss": 0.0049, "step": 3070 }, { "epoch": 43.859911894273125, "grad_norm": 0.217172309756279, "learning_rate": 7.776672597537827e-06, "loss": 0.0056, "step": 3071 }, { "epoch": 43.87400881057269, "grad_norm": 0.16421233117580414, "learning_rate": 7.74093458286782e-06, "loss": 0.0053, "step": 3072 }, { "epoch": 43.88810572687225, "grad_norm": 0.11700178682804108, "learning_rate": 7.705275568582848e-06, "loss": 0.0051, "step": 3073 }, { "epoch": 43.902202643171805, "grad_norm": 0.27165380120277405, "learning_rate": 7.66969558521734e-06, "loss": 0.0061, "step": 3074 }, { "epoch": 43.91629955947137, "grad_norm": 0.17903371155261993, "learning_rate": 7.634194663238015e-06, "loss": 0.0052, "step": 3075 }, { "epoch": 43.93039647577093, "grad_norm": 0.1227598488330841, "learning_rate": 7.598772833043932e-06, "loss": 0.0049, "step": 3076 }, { "epoch": 43.944493392070484, "grad_norm": 0.2229098230600357, "learning_rate": 7.5634301249663755e-06, "loss": 0.0063, "step": 3077 }, { "epoch": 43.95859030837004, "grad_norm": 0.19699111580848694, "learning_rate": 7.528166569268913e-06, "loss": 0.0051, "step": 3078 }, { "epoch": 43.972687224669606, "grad_norm": 0.14560317993164062, "learning_rate": 7.492982196147336e-06, "loss": 0.0055, "step": 3079 }, { "epoch": 43.986784140969164, "grad_norm": 0.22183571755886078, "learning_rate": 7.457877035729588e-06, "loss": 0.0053, "step": 3080 }, { "epoch": 44.01409691629956, "grad_norm": 0.103655606508255, "learning_rate": 7.422851118075846e-06, "loss": 0.009, "step": 3081 }, { "epoch": 44.02819383259912, "grad_norm": 0.16993626952171326, "learning_rate": 7.3879044731783975e-06, "loss": 0.005, "step": 3082 }, { "epoch": 44.04229074889868, "grad_norm": 0.16448165476322174, "learning_rate": 7.353037130961671e-06, "loss": 0.0054, "step": 3083 }, { "epoch": 44.05638766519824, "grad_norm": 0.09809138625860214, "learning_rate": 7.318249121282128e-06, "loss": 0.0049, "step": 3084 }, { "epoch": 44.070484581497794, "grad_norm": 0.10641877353191376, "learning_rate": 7.283540473928385e-06, "loss": 0.0048, "step": 3085 }, { "epoch": 44.08458149779736, "grad_norm": 0.17998577654361725, "learning_rate": 7.248911218621057e-06, "loss": 0.0054, "step": 3086 }, { "epoch": 44.098678414096916, "grad_norm": 0.15640848875045776, "learning_rate": 7.21436138501278e-06, "loss": 0.0053, "step": 3087 }, { "epoch": 44.112775330396474, "grad_norm": 0.1922016590833664, "learning_rate": 7.179891002688155e-06, "loss": 0.0054, "step": 3088 }, { "epoch": 44.12687224669604, "grad_norm": 0.18591929972171783, "learning_rate": 7.145500101163804e-06, "loss": 0.0053, "step": 3089 }, { "epoch": 44.140969162995596, "grad_norm": 0.13664504885673523, "learning_rate": 7.11118870988825e-06, "loss": 0.0053, "step": 3090 }, { "epoch": 44.15506607929515, "grad_norm": 0.12976562976837158, "learning_rate": 7.076956858241979e-06, "loss": 0.0047, "step": 3091 }, { "epoch": 44.16916299559471, "grad_norm": 0.1415586918592453, "learning_rate": 7.042804575537287e-06, "loss": 0.0052, "step": 3092 }, { "epoch": 44.183259911894275, "grad_norm": 0.176945760846138, "learning_rate": 7.008731891018405e-06, "loss": 0.0053, "step": 3093 }, { "epoch": 44.19735682819383, "grad_norm": 0.23750530183315277, "learning_rate": 6.974738833861383e-06, "loss": 0.0061, "step": 3094 }, { "epoch": 44.21145374449339, "grad_norm": 0.15721862018108368, "learning_rate": 6.940825433174103e-06, "loss": 0.0051, "step": 3095 }, { "epoch": 44.225550660792955, "grad_norm": 0.22597970068454742, "learning_rate": 6.906991717996203e-06, "loss": 0.0053, "step": 3096 }, { "epoch": 44.23964757709251, "grad_norm": 0.1972382664680481, "learning_rate": 6.873237717299108e-06, "loss": 0.0058, "step": 3097 }, { "epoch": 44.25374449339207, "grad_norm": 0.09687539935112, "learning_rate": 6.839563459986009e-06, "loss": 0.0047, "step": 3098 }, { "epoch": 44.26784140969163, "grad_norm": 0.17441712319850922, "learning_rate": 6.805968974891752e-06, "loss": 0.0053, "step": 3099 }, { "epoch": 44.28193832599119, "grad_norm": 0.1672023981809616, "learning_rate": 6.772454290782926e-06, "loss": 0.0052, "step": 3100 }, { "epoch": 44.29603524229075, "grad_norm": 0.16512802243232727, "learning_rate": 6.739019436357774e-06, "loss": 0.0052, "step": 3101 }, { "epoch": 44.31013215859031, "grad_norm": 0.18048758804798126, "learning_rate": 6.705664440246196e-06, "loss": 0.0046, "step": 3102 }, { "epoch": 44.32422907488987, "grad_norm": 0.1892092525959015, "learning_rate": 6.672389331009665e-06, "loss": 0.0056, "step": 3103 }, { "epoch": 44.33832599118943, "grad_norm": 0.1689048856496811, "learning_rate": 6.639194137141291e-06, "loss": 0.005, "step": 3104 }, { "epoch": 44.352422907488986, "grad_norm": 0.17507663369178772, "learning_rate": 6.606078887065748e-06, "loss": 0.005, "step": 3105 }, { "epoch": 44.36651982378854, "grad_norm": 0.13498033583164215, "learning_rate": 6.573043609139251e-06, "loss": 0.0052, "step": 3106 }, { "epoch": 44.38061674008811, "grad_norm": 0.14264729619026184, "learning_rate": 6.540088331649497e-06, "loss": 0.0051, "step": 3107 }, { "epoch": 44.394713656387665, "grad_norm": 0.12114330381155014, "learning_rate": 6.507213082815744e-06, "loss": 0.0051, "step": 3108 }, { "epoch": 44.40881057268722, "grad_norm": 0.18631602823734283, "learning_rate": 6.4744178907886865e-06, "loss": 0.0055, "step": 3109 }, { "epoch": 44.42290748898679, "grad_norm": 0.11437942832708359, "learning_rate": 6.441702783650494e-06, "loss": 0.0049, "step": 3110 }, { "epoch": 44.437004405286345, "grad_norm": 0.12540662288665771, "learning_rate": 6.409067789414691e-06, "loss": 0.0052, "step": 3111 }, { "epoch": 44.4511013215859, "grad_norm": 0.195652037858963, "learning_rate": 6.37651293602628e-06, "loss": 0.0052, "step": 3112 }, { "epoch": 44.46519823788546, "grad_norm": 0.14621053636074066, "learning_rate": 6.344038251361628e-06, "loss": 0.005, "step": 3113 }, { "epoch": 44.479295154185024, "grad_norm": 0.215066596865654, "learning_rate": 6.311643763228403e-06, "loss": 0.0055, "step": 3114 }, { "epoch": 44.49339207048458, "grad_norm": 0.2706260085105896, "learning_rate": 6.2793294993656494e-06, "loss": 0.0061, "step": 3115 }, { "epoch": 44.50748898678414, "grad_norm": 0.23228813707828522, "learning_rate": 6.2470954874437e-06, "loss": 0.0051, "step": 3116 }, { "epoch": 44.521585903083704, "grad_norm": 0.17580263316631317, "learning_rate": 6.214941755064185e-06, "loss": 0.0053, "step": 3117 }, { "epoch": 44.53568281938326, "grad_norm": 0.1480007767677307, "learning_rate": 6.182868329759961e-06, "loss": 0.0049, "step": 3118 }, { "epoch": 44.54977973568282, "grad_norm": 0.09259267896413803, "learning_rate": 6.150875238995135e-06, "loss": 0.0046, "step": 3119 }, { "epoch": 44.563876651982376, "grad_norm": 0.16161957383155823, "learning_rate": 6.118962510165038e-06, "loss": 0.0053, "step": 3120 }, { "epoch": 44.57797356828194, "grad_norm": 0.20151600241661072, "learning_rate": 6.087130170596189e-06, "loss": 0.0051, "step": 3121 }, { "epoch": 44.5920704845815, "grad_norm": 0.2046988606452942, "learning_rate": 6.055378247546218e-06, "loss": 0.0058, "step": 3122 }, { "epoch": 44.606167400881056, "grad_norm": 0.1874491423368454, "learning_rate": 6.023706768203963e-06, "loss": 0.005, "step": 3123 }, { "epoch": 44.62026431718062, "grad_norm": 0.2240460366010666, "learning_rate": 5.992115759689332e-06, "loss": 0.005, "step": 3124 }, { "epoch": 44.63436123348018, "grad_norm": 0.1837216019630432, "learning_rate": 5.960605249053386e-06, "loss": 0.0056, "step": 3125 }, { "epoch": 44.648458149779735, "grad_norm": 0.17020055651664734, "learning_rate": 5.9291752632781636e-06, "loss": 0.0049, "step": 3126 }, { "epoch": 44.66255506607929, "grad_norm": 0.17558841407299042, "learning_rate": 5.8978258292768265e-06, "loss": 0.0054, "step": 3127 }, { "epoch": 44.67665198237886, "grad_norm": 0.23532618582248688, "learning_rate": 5.866556973893545e-06, "loss": 0.0057, "step": 3128 }, { "epoch": 44.690748898678414, "grad_norm": 0.2115897834300995, "learning_rate": 5.835368723903456e-06, "loss": 0.0052, "step": 3129 }, { "epoch": 44.70484581497797, "grad_norm": 0.2307427078485489, "learning_rate": 5.8042611060127315e-06, "loss": 0.0054, "step": 3130 }, { "epoch": 44.71894273127754, "grad_norm": 0.21360504627227783, "learning_rate": 5.773234146858464e-06, "loss": 0.0056, "step": 3131 }, { "epoch": 44.733039647577094, "grad_norm": 0.20878787338733673, "learning_rate": 5.742287873008678e-06, "loss": 0.0053, "step": 3132 }, { "epoch": 44.74713656387665, "grad_norm": 0.166913241147995, "learning_rate": 5.7114223109623115e-06, "loss": 0.0053, "step": 3133 }, { "epoch": 44.76123348017621, "grad_norm": 0.15856032073497772, "learning_rate": 5.6806374871491985e-06, "loss": 0.0053, "step": 3134 }, { "epoch": 44.77533039647577, "grad_norm": 0.28434082865715027, "learning_rate": 5.649933427930043e-06, "loss": 0.006, "step": 3135 }, { "epoch": 44.78942731277533, "grad_norm": 0.152896910905838, "learning_rate": 5.6193101595963585e-06, "loss": 0.0052, "step": 3136 }, { "epoch": 44.80352422907489, "grad_norm": 0.13170908391475677, "learning_rate": 5.5887677083705125e-06, "loss": 0.0047, "step": 3137 }, { "epoch": 44.817621145374446, "grad_norm": 0.09505784511566162, "learning_rate": 5.558306100405653e-06, "loss": 0.0047, "step": 3138 }, { "epoch": 44.83171806167401, "grad_norm": 0.1286177784204483, "learning_rate": 5.527925361785724e-06, "loss": 0.0054, "step": 3139 }, { "epoch": 44.84581497797357, "grad_norm": 0.17008164525032043, "learning_rate": 5.497625518525373e-06, "loss": 0.0061, "step": 3140 }, { "epoch": 44.859911894273125, "grad_norm": 0.23830875754356384, "learning_rate": 5.467406596570035e-06, "loss": 0.0051, "step": 3141 }, { "epoch": 44.87400881057269, "grad_norm": 0.1610742062330246, "learning_rate": 5.437268621795821e-06, "loss": 0.0049, "step": 3142 }, { "epoch": 44.88810572687225, "grad_norm": 0.10069489479064941, "learning_rate": 5.407211620009544e-06, "loss": 0.0048, "step": 3143 }, { "epoch": 44.902202643171805, "grad_norm": 0.14704009890556335, "learning_rate": 5.377235616948661e-06, "loss": 0.0052, "step": 3144 }, { "epoch": 44.91629955947137, "grad_norm": 0.13210012018680573, "learning_rate": 5.347340638281273e-06, "loss": 0.005, "step": 3145 }, { "epoch": 44.93039647577093, "grad_norm": 0.25106000900268555, "learning_rate": 5.317526709606125e-06, "loss": 0.0054, "step": 3146 }, { "epoch": 44.944493392070484, "grad_norm": 0.21174417436122894, "learning_rate": 5.2877938564525495e-06, "loss": 0.0057, "step": 3147 }, { "epoch": 44.95859030837004, "grad_norm": 0.2293211966753006, "learning_rate": 5.258142104280439e-06, "loss": 0.0054, "step": 3148 }, { "epoch": 44.972687224669606, "grad_norm": 0.15222808718681335, "learning_rate": 5.2285714784802375e-06, "loss": 0.0059, "step": 3149 }, { "epoch": 44.986784140969164, "grad_norm": 0.24396970868110657, "learning_rate": 5.199082004372957e-06, "loss": 0.0051, "step": 3150 }, { "epoch": 45.01409691629956, "grad_norm": 0.23132318258285522, "learning_rate": 5.169673707210087e-06, "loss": 0.0099, "step": 3151 }, { "epoch": 45.02819383259912, "grad_norm": 0.1761818826198578, "learning_rate": 5.140346612173608e-06, "loss": 0.0047, "step": 3152 }, { "epoch": 45.04229074889868, "grad_norm": 0.14733724296092987, "learning_rate": 5.1111007443759855e-06, "loss": 0.0054, "step": 3153 }, { "epoch": 45.05638766519824, "grad_norm": 0.19008779525756836, "learning_rate": 5.081936128860143e-06, "loss": 0.0053, "step": 3154 }, { "epoch": 45.070484581497794, "grad_norm": 0.16125279664993286, "learning_rate": 5.052852790599383e-06, "loss": 0.0051, "step": 3155 }, { "epoch": 45.08458149779736, "grad_norm": 0.11107031255960464, "learning_rate": 5.023850754497428e-06, "loss": 0.0048, "step": 3156 }, { "epoch": 45.098678414096916, "grad_norm": 0.16728179156780243, "learning_rate": 4.994930045388413e-06, "loss": 0.0053, "step": 3157 }, { "epoch": 45.112775330396474, "grad_norm": 0.1314859539270401, "learning_rate": 4.966090688036806e-06, "loss": 0.0048, "step": 3158 }, { "epoch": 45.12687224669604, "grad_norm": 0.16542252898216248, "learning_rate": 4.93733270713741e-06, "loss": 0.0053, "step": 3159 }, { "epoch": 45.140969162995596, "grad_norm": 0.15862467885017395, "learning_rate": 4.908656127315359e-06, "loss": 0.005, "step": 3160 }, { "epoch": 45.15506607929515, "grad_norm": 0.23872783780097961, "learning_rate": 4.880060973126088e-06, "loss": 0.0052, "step": 3161 }, { "epoch": 45.16916299559471, "grad_norm": 0.20890763401985168, "learning_rate": 4.851547269055312e-06, "loss": 0.0055, "step": 3162 }, { "epoch": 45.183259911894275, "grad_norm": 0.20229151844978333, "learning_rate": 4.8231150395189666e-06, "loss": 0.005, "step": 3163 }, { "epoch": 45.19735682819383, "grad_norm": 0.12213220447301865, "learning_rate": 4.794764308863242e-06, "loss": 0.0048, "step": 3164 }, { "epoch": 45.21145374449339, "grad_norm": 0.256910502910614, "learning_rate": 4.7664951013645875e-06, "loss": 0.0054, "step": 3165 }, { "epoch": 45.225550660792955, "grad_norm": 0.09108830243349075, "learning_rate": 4.73830744122955e-06, "loss": 0.0047, "step": 3166 }, { "epoch": 45.23964757709251, "grad_norm": 0.14474941790103912, "learning_rate": 4.710201352594923e-06, "loss": 0.0051, "step": 3167 }, { "epoch": 45.25374449339207, "grad_norm": 0.12063156813383102, "learning_rate": 4.682176859527643e-06, "loss": 0.0048, "step": 3168 }, { "epoch": 45.26784140969163, "grad_norm": 0.13912518322467804, "learning_rate": 4.6542339860247584e-06, "loss": 0.0049, "step": 3169 }, { "epoch": 45.28193832599119, "grad_norm": 0.14217424392700195, "learning_rate": 4.626372756013409e-06, "loss": 0.005, "step": 3170 }, { "epoch": 45.29603524229075, "grad_norm": 0.19744612276554108, "learning_rate": 4.5985931933508754e-06, "loss": 0.005, "step": 3171 }, { "epoch": 45.31013215859031, "grad_norm": 0.10002440959215164, "learning_rate": 4.570895321824465e-06, "loss": 0.0049, "step": 3172 }, { "epoch": 45.32422907488987, "grad_norm": 0.1056085154414177, "learning_rate": 4.5432791651515705e-06, "loss": 0.0049, "step": 3173 }, { "epoch": 45.33832599118943, "grad_norm": 0.1546996384859085, "learning_rate": 4.515744746979556e-06, "loss": 0.0051, "step": 3174 }, { "epoch": 45.352422907488986, "grad_norm": 0.1290605217218399, "learning_rate": 4.48829209088586e-06, "loss": 0.005, "step": 3175 }, { "epoch": 45.36651982378854, "grad_norm": 0.211063414812088, "learning_rate": 4.460921220377856e-06, "loss": 0.0055, "step": 3176 }, { "epoch": 45.38061674008811, "grad_norm": 0.14346465468406677, "learning_rate": 4.433632158892942e-06, "loss": 0.005, "step": 3177 }, { "epoch": 45.394713656387665, "grad_norm": 0.20932093262672424, "learning_rate": 4.406424929798403e-06, "loss": 0.0054, "step": 3178 }, { "epoch": 45.40881057268722, "grad_norm": 0.2253430038690567, "learning_rate": 4.379299556391469e-06, "loss": 0.0057, "step": 3179 }, { "epoch": 45.42290748898679, "grad_norm": 0.3143933117389679, "learning_rate": 4.352256061899329e-06, "loss": 0.0061, "step": 3180 }, { "epoch": 45.437004405286345, "grad_norm": 0.08022678643465042, "learning_rate": 4.325294469478968e-06, "loss": 0.0046, "step": 3181 }, { "epoch": 45.4511013215859, "grad_norm": 0.23060749471187592, "learning_rate": 4.298414802217332e-06, "loss": 0.0063, "step": 3182 }, { "epoch": 45.46519823788546, "grad_norm": 0.1789335310459137, "learning_rate": 4.271617083131141e-06, "loss": 0.0056, "step": 3183 }, { "epoch": 45.479295154185024, "grad_norm": 0.12316460907459259, "learning_rate": 4.244901335167006e-06, "loss": 0.0047, "step": 3184 }, { "epoch": 45.49339207048458, "grad_norm": 0.18632884323596954, "learning_rate": 4.2182675812012965e-06, "loss": 0.005, "step": 3185 }, { "epoch": 45.50748898678414, "grad_norm": 0.17624638974666595, "learning_rate": 4.1917158440401895e-06, "loss": 0.0054, "step": 3186 }, { "epoch": 45.521585903083704, "grad_norm": 0.1655586212873459, "learning_rate": 4.165246146419655e-06, "loss": 0.0051, "step": 3187 }, { "epoch": 45.53568281938326, "grad_norm": 0.16250984370708466, "learning_rate": 4.138858511005383e-06, "loss": 0.0054, "step": 3188 }, { "epoch": 45.54977973568282, "grad_norm": 0.24676015973091125, "learning_rate": 4.112552960392801e-06, "loss": 0.0051, "step": 3189 }, { "epoch": 45.563876651982376, "grad_norm": 0.13578608632087708, "learning_rate": 4.086329517107046e-06, "loss": 0.0051, "step": 3190 }, { "epoch": 45.57797356828194, "grad_norm": 0.1320914328098297, "learning_rate": 4.060188203602955e-06, "loss": 0.0051, "step": 3191 }, { "epoch": 45.5920704845815, "grad_norm": 0.18774719536304474, "learning_rate": 4.034129042265066e-06, "loss": 0.0053, "step": 3192 }, { "epoch": 45.606167400881056, "grad_norm": 0.18465903401374817, "learning_rate": 4.008152055407499e-06, "loss": 0.0052, "step": 3193 }, { "epoch": 45.62026431718062, "grad_norm": 0.23765979707241058, "learning_rate": 3.982257265274081e-06, "loss": 0.0056, "step": 3194 }, { "epoch": 45.63436123348018, "grad_norm": 0.13689181208610535, "learning_rate": 3.9564446940382084e-06, "loss": 0.0052, "step": 3195 }, { "epoch": 45.648458149779735, "grad_norm": 0.19309017062187195, "learning_rate": 3.930714363802901e-06, "loss": 0.0054, "step": 3196 }, { "epoch": 45.66255506607929, "grad_norm": 0.16123777627944946, "learning_rate": 3.905066296600745e-06, "loss": 0.0053, "step": 3197 }, { "epoch": 45.67665198237886, "grad_norm": 0.24765531718730927, "learning_rate": 3.879500514393874e-06, "loss": 0.0056, "step": 3198 }, { "epoch": 45.690748898678414, "grad_norm": 0.17967095971107483, "learning_rate": 3.854017039074009e-06, "loss": 0.0055, "step": 3199 }, { "epoch": 45.70484581497797, "grad_norm": 0.18351756036281586, "learning_rate": 3.828615892462328e-06, "loss": 0.0051, "step": 3200 }, { "epoch": 45.71894273127754, "grad_norm": 0.14961770176887512, "learning_rate": 3.803297096309555e-06, "loss": 0.0049, "step": 3201 }, { "epoch": 45.733039647577094, "grad_norm": 0.21736842393875122, "learning_rate": 3.77806067229588e-06, "loss": 0.0054, "step": 3202 }, { "epoch": 45.74713656387665, "grad_norm": 0.14942164719104767, "learning_rate": 3.7529066420309733e-06, "loss": 0.0054, "step": 3203 }, { "epoch": 45.76123348017621, "grad_norm": 0.16754430532455444, "learning_rate": 3.7278350270539384e-06, "loss": 0.0052, "step": 3204 }, { "epoch": 45.77533039647577, "grad_norm": 0.12601113319396973, "learning_rate": 3.7028458488333138e-06, "loss": 0.0049, "step": 3205 }, { "epoch": 45.78942731277533, "grad_norm": 0.21815557777881622, "learning_rate": 3.6779391287670494e-06, "loss": 0.0053, "step": 3206 }, { "epoch": 45.80352422907489, "grad_norm": 0.1526600420475006, "learning_rate": 3.6531148881824962e-06, "loss": 0.0054, "step": 3207 }, { "epoch": 45.817621145374446, "grad_norm": 0.23582018911838531, "learning_rate": 3.6283731483363393e-06, "loss": 0.0054, "step": 3208 }, { "epoch": 45.83171806167401, "grad_norm": 0.18773604929447174, "learning_rate": 3.6037139304146762e-06, "loss": 0.0053, "step": 3209 }, { "epoch": 45.84581497797357, "grad_norm": 0.15210607647895813, "learning_rate": 3.5791372555328937e-06, "loss": 0.0052, "step": 3210 }, { "epoch": 45.859911894273125, "grad_norm": 0.10319545120000839, "learning_rate": 3.5546431447357477e-06, "loss": 0.0047, "step": 3211 }, { "epoch": 45.87400881057269, "grad_norm": 0.2141682356595993, "learning_rate": 3.530231618997226e-06, "loss": 0.0052, "step": 3212 }, { "epoch": 45.88810572687225, "grad_norm": 0.18333624303340912, "learning_rate": 3.5059026992206647e-06, "loss": 0.0053, "step": 3213 }, { "epoch": 45.902202643171805, "grad_norm": 0.23267517983913422, "learning_rate": 3.4816564062386427e-06, "loss": 0.0052, "step": 3214 }, { "epoch": 45.91629955947137, "grad_norm": 0.18764956295490265, "learning_rate": 3.457492760812975e-06, "loss": 0.0053, "step": 3215 }, { "epoch": 45.93039647577093, "grad_norm": 0.1920033097267151, "learning_rate": 3.4334117836347103e-06, "loss": 0.0054, "step": 3216 }, { "epoch": 45.944493392070484, "grad_norm": 0.20086729526519775, "learning_rate": 3.409413495324121e-06, "loss": 0.0054, "step": 3217 }, { "epoch": 45.95859030837004, "grad_norm": 0.12244732677936554, "learning_rate": 3.38549791643068e-06, "loss": 0.0049, "step": 3218 }, { "epoch": 45.972687224669606, "grad_norm": 0.20846551656723022, "learning_rate": 3.361665067432995e-06, "loss": 0.0057, "step": 3219 }, { "epoch": 45.986784140969164, "grad_norm": 0.21844924986362457, "learning_rate": 3.3379149687388867e-06, "loss": 0.0052, "step": 3220 }, { "epoch": 46.01409691629956, "grad_norm": 0.17777897417545319, "learning_rate": 3.314247640685286e-06, "loss": 0.0105, "step": 3221 }, { "epoch": 46.02819383259912, "grad_norm": 0.2437198907136917, "learning_rate": 3.290663103538261e-06, "loss": 0.0049, "step": 3222 }, { "epoch": 46.04229074889868, "grad_norm": 0.1851365715265274, "learning_rate": 3.2671613774929776e-06, "loss": 0.0051, "step": 3223 }, { "epoch": 46.05638766519824, "grad_norm": 0.2532176077365875, "learning_rate": 3.243742482673684e-06, "loss": 0.0051, "step": 3224 }, { "epoch": 46.070484581497794, "grad_norm": 0.11412458121776581, "learning_rate": 3.2204064391337273e-06, "loss": 0.0057, "step": 3225 }, { "epoch": 46.08458149779736, "grad_norm": 0.21863538026809692, "learning_rate": 3.1971532668555128e-06, "loss": 0.0049, "step": 3226 }, { "epoch": 46.098678414096916, "grad_norm": 0.08787374943494797, "learning_rate": 3.1739829857504234e-06, "loss": 0.0051, "step": 3227 }, { "epoch": 46.112775330396474, "grad_norm": 0.2364179491996765, "learning_rate": 3.1508956156589332e-06, "loss": 0.0055, "step": 3228 }, { "epoch": 46.12687224669604, "grad_norm": 0.1849663406610489, "learning_rate": 3.127891176350506e-06, "loss": 0.0057, "step": 3229 }, { "epoch": 46.140969162995596, "grad_norm": 0.19479987025260925, "learning_rate": 3.1049696875235512e-06, "loss": 0.0052, "step": 3230 }, { "epoch": 46.15506607929515, "grad_norm": 0.12292339652776718, "learning_rate": 3.082131168805491e-06, "loss": 0.0048, "step": 3231 }, { "epoch": 46.16916299559471, "grad_norm": 0.10942355543375015, "learning_rate": 3.0593756397526817e-06, "loss": 0.0048, "step": 3232 }, { "epoch": 46.183259911894275, "grad_norm": 0.12949985265731812, "learning_rate": 3.0367031198504482e-06, "loss": 0.0046, "step": 3233 }, { "epoch": 46.19735682819383, "grad_norm": 0.07617957144975662, "learning_rate": 3.014113628512982e-06, "loss": 0.0053, "step": 3234 }, { "epoch": 46.21145374449339, "grad_norm": 0.2256639301776886, "learning_rate": 2.9916071850834005e-06, "loss": 0.0058, "step": 3235 }, { "epoch": 46.225550660792955, "grad_norm": 0.19888536632061005, "learning_rate": 2.969183808833753e-06, "loss": 0.0048, "step": 3236 }, { "epoch": 46.23964757709251, "grad_norm": 0.1473282426595688, "learning_rate": 2.9468435189648925e-06, "loss": 0.005, "step": 3237 }, { "epoch": 46.25374449339207, "grad_norm": 0.12569357454776764, "learning_rate": 2.9245863346065604e-06, "loss": 0.0054, "step": 3238 }, { "epoch": 46.26784140969163, "grad_norm": 0.17000888288021088, "learning_rate": 2.9024122748173345e-06, "loss": 0.0053, "step": 3239 }, { "epoch": 46.28193832599119, "grad_norm": 0.2002563774585724, "learning_rate": 2.8803213585846035e-06, "loss": 0.0048, "step": 3240 }, { "epoch": 46.29603524229075, "grad_norm": 0.15725058317184448, "learning_rate": 2.8583136048245697e-06, "loss": 0.0049, "step": 3241 }, { "epoch": 46.31013215859031, "grad_norm": 0.10260513424873352, "learning_rate": 2.836389032382214e-06, "loss": 0.0052, "step": 3242 }, { "epoch": 46.32422907488987, "grad_norm": 0.1727161705493927, "learning_rate": 2.814547660031308e-06, "loss": 0.0056, "step": 3243 }, { "epoch": 46.33832599118943, "grad_norm": 0.2677318751811981, "learning_rate": 2.7927895064743694e-06, "loss": 0.0055, "step": 3244 }, { "epoch": 46.352422907488986, "grad_norm": 0.22769194841384888, "learning_rate": 2.7711145903426496e-06, "loss": 0.0049, "step": 3245 }, { "epoch": 46.36651982378854, "grad_norm": 0.12293656915426254, "learning_rate": 2.7495229301961356e-06, "loss": 0.0052, "step": 3246 }, { "epoch": 46.38061674008811, "grad_norm": 0.18619374930858612, "learning_rate": 2.728014544523516e-06, "loss": 0.005, "step": 3247 }, { "epoch": 46.394713656387665, "grad_norm": 0.13346119225025177, "learning_rate": 2.706589451742181e-06, "loss": 0.0049, "step": 3248 }, { "epoch": 46.40881057268722, "grad_norm": 0.1351984292268753, "learning_rate": 2.6852476701981876e-06, "loss": 0.0053, "step": 3249 }, { "epoch": 46.42290748898679, "grad_norm": 0.15558120608329773, "learning_rate": 2.663989218166274e-06, "loss": 0.005, "step": 3250 }, { "epoch": 46.42290748898679, "eval_loss": 1.452907919883728, "eval_runtime": 135.3295, "eval_samples_per_second": 13.36, "eval_steps_per_second": 0.421, "step": 3250 }, { "epoch": 46.437004405286345, "grad_norm": 0.15262937545776367, "learning_rate": 2.642814113849801e-06, "loss": 0.005, "step": 3251 }, { "epoch": 46.4511013215859, "grad_norm": 0.17883974313735962, "learning_rate": 2.621722375380753e-06, "loss": 0.005, "step": 3252 }, { "epoch": 46.46519823788546, "grad_norm": 0.11460565775632858, "learning_rate": 2.600714020819761e-06, "loss": 0.0049, "step": 3253 }, { "epoch": 46.479295154185024, "grad_norm": 0.12418580800294876, "learning_rate": 2.5797890681560468e-06, "loss": 0.0053, "step": 3254 }, { "epoch": 46.49339207048458, "grad_norm": 0.24074402451515198, "learning_rate": 2.5589475353073988e-06, "loss": 0.0051, "step": 3255 }, { "epoch": 46.50748898678414, "grad_norm": 0.17416901886463165, "learning_rate": 2.5381894401201757e-06, "loss": 0.0053, "step": 3256 }, { "epoch": 46.521585903083704, "grad_norm": 0.16847871243953705, "learning_rate": 2.517514800369314e-06, "loss": 0.0051, "step": 3257 }, { "epoch": 46.53568281938326, "grad_norm": 0.19676998257637024, "learning_rate": 2.4969236337582637e-06, "loss": 0.0049, "step": 3258 }, { "epoch": 46.54977973568282, "grad_norm": 0.1103222519159317, "learning_rate": 2.4764159579190095e-06, "loss": 0.005, "step": 3259 }, { "epoch": 46.563876651982376, "grad_norm": 0.13004785776138306, "learning_rate": 2.455991790412027e-06, "loss": 0.0047, "step": 3260 }, { "epoch": 46.57797356828194, "grad_norm": 0.0902215763926506, "learning_rate": 2.4356511487263143e-06, "loss": 0.005, "step": 3261 }, { "epoch": 46.5920704845815, "grad_norm": 0.1797294318675995, "learning_rate": 2.415394050279318e-06, "loss": 0.0047, "step": 3262 }, { "epoch": 46.606167400881056, "grad_norm": 0.1330057680606842, "learning_rate": 2.395220512416985e-06, "loss": 0.0049, "step": 3263 }, { "epoch": 46.62026431718062, "grad_norm": 0.11397356539964676, "learning_rate": 2.3751305524136513e-06, "loss": 0.0049, "step": 3264 }, { "epoch": 46.63436123348018, "grad_norm": 0.14937102794647217, "learning_rate": 2.355124187472135e-06, "loss": 0.0049, "step": 3265 }, { "epoch": 46.648458149779735, "grad_norm": 0.17829111218452454, "learning_rate": 2.3352014347236774e-06, "loss": 0.0056, "step": 3266 }, { "epoch": 46.66255506607929, "grad_norm": 0.1921476274728775, "learning_rate": 2.3153623112278757e-06, "loss": 0.0052, "step": 3267 }, { "epoch": 46.67665198237886, "grad_norm": 0.19977952539920807, "learning_rate": 2.295606833972752e-06, "loss": 0.0049, "step": 3268 }, { "epoch": 46.690748898678414, "grad_norm": 0.11170845478773117, "learning_rate": 2.2759350198746976e-06, "loss": 0.0055, "step": 3269 }, { "epoch": 46.70484581497797, "grad_norm": 0.16515663266181946, "learning_rate": 2.2563468857784818e-06, "loss": 0.0055, "step": 3270 }, { "epoch": 46.71894273127754, "grad_norm": 0.19528886675834656, "learning_rate": 2.236842448457166e-06, "loss": 0.0057, "step": 3271 }, { "epoch": 46.733039647577094, "grad_norm": 0.20047548413276672, "learning_rate": 2.217421724612201e-06, "loss": 0.006, "step": 3272 }, { "epoch": 46.74713656387665, "grad_norm": 0.30587443709373474, "learning_rate": 2.1980847308733177e-06, "loss": 0.0054, "step": 3273 }, { "epoch": 46.76123348017621, "grad_norm": 0.26512664556503296, "learning_rate": 2.1788314837985824e-06, "loss": 0.0052, "step": 3274 }, { "epoch": 46.77533039647577, "grad_norm": 0.19528301060199738, "learning_rate": 2.159661999874307e-06, "loss": 0.0048, "step": 3275 }, { "epoch": 46.78942731277533, "grad_norm": 0.11245682090520859, "learning_rate": 2.1405762955151176e-06, "loss": 0.0047, "step": 3276 }, { "epoch": 46.80352422907489, "grad_norm": 0.14431720972061157, "learning_rate": 2.121574387063885e-06, "loss": 0.0048, "step": 3277 }, { "epoch": 46.817621145374446, "grad_norm": 0.12015689164400101, "learning_rate": 2.102656290791727e-06, "loss": 0.0061, "step": 3278 }, { "epoch": 46.83171806167401, "grad_norm": 0.2500021755695343, "learning_rate": 2.0838220228979743e-06, "loss": 0.0047, "step": 3279 }, { "epoch": 46.84581497797357, "grad_norm": 0.129334956407547, "learning_rate": 2.0650715995102267e-06, "loss": 0.0053, "step": 3280 }, { "epoch": 46.859911894273125, "grad_norm": 0.19566504657268524, "learning_rate": 2.04640503668424e-06, "loss": 0.0052, "step": 3281 }, { "epoch": 46.87400881057269, "grad_norm": 0.11147371679544449, "learning_rate": 2.027822350403963e-06, "loss": 0.005, "step": 3282 }, { "epoch": 46.88810572687225, "grad_norm": 0.12369511276483536, "learning_rate": 2.009323556581566e-06, "loss": 0.005, "step": 3283 }, { "epoch": 46.902202643171805, "grad_norm": 0.19443906843662262, "learning_rate": 1.9909086710573455e-06, "loss": 0.0055, "step": 3284 }, { "epoch": 46.91629955947137, "grad_norm": 0.1899430900812149, "learning_rate": 1.9725777095997657e-06, "loss": 0.0053, "step": 3285 }, { "epoch": 46.93039647577093, "grad_norm": 0.16920971870422363, "learning_rate": 1.9543306879054034e-06, "loss": 0.0057, "step": 3286 }, { "epoch": 46.944493392070484, "grad_norm": 0.18749023973941803, "learning_rate": 1.936167621598972e-06, "loss": 0.0053, "step": 3287 }, { "epoch": 46.95859030837004, "grad_norm": 0.19871167838573456, "learning_rate": 1.9180885262333195e-06, "loss": 0.0051, "step": 3288 }, { "epoch": 46.972687224669606, "grad_norm": 0.1639748513698578, "learning_rate": 1.9000934172893615e-06, "loss": 0.0049, "step": 3289 }, { "epoch": 46.986784140969164, "grad_norm": 0.14231142401695251, "learning_rate": 1.882182310176095e-06, "loss": 0.0047, "step": 3290 }, { "epoch": 47.01409691629956, "grad_norm": 0.4539938271045685, "learning_rate": 1.864355220230607e-06, "loss": 0.0091, "step": 3291 }, { "epoch": 47.02819383259912, "grad_norm": 0.15678849816322327, "learning_rate": 1.84661216271802e-06, "loss": 0.0051, "step": 3292 }, { "epoch": 47.04229074889868, "grad_norm": 0.18227137625217438, "learning_rate": 1.8289531528315251e-06, "loss": 0.0053, "step": 3293 }, { "epoch": 47.05638766519824, "grad_norm": 0.17287705838680267, "learning_rate": 1.8113782056923268e-06, "loss": 0.0057, "step": 3294 }, { "epoch": 47.070484581497794, "grad_norm": 0.2381402552127838, "learning_rate": 1.7938873363496422e-06, "loss": 0.0054, "step": 3295 }, { "epoch": 47.08458149779736, "grad_norm": 0.16326670348644257, "learning_rate": 1.7764805597807132e-06, "loss": 0.0055, "step": 3296 }, { "epoch": 47.098678414096916, "grad_norm": 0.14031168818473816, "learning_rate": 1.7591578908907724e-06, "loss": 0.005, "step": 3297 }, { "epoch": 47.112775330396474, "grad_norm": 0.1449560523033142, "learning_rate": 1.7419193445129988e-06, "loss": 0.0053, "step": 3298 }, { "epoch": 47.12687224669604, "grad_norm": 0.16722328960895538, "learning_rate": 1.7247649354085737e-06, "loss": 0.0053, "step": 3299 }, { "epoch": 47.140969162995596, "grad_norm": 0.1646098643541336, "learning_rate": 1.707694678266636e-06, "loss": 0.0053, "step": 3300 }, { "epoch": 47.15506607929515, "grad_norm": 0.17422163486480713, "learning_rate": 1.6907085877042373e-06, "loss": 0.0048, "step": 3301 }, { "epoch": 47.16916299559471, "grad_norm": 0.11328745633363724, "learning_rate": 1.6738066782663652e-06, "loss": 0.005, "step": 3302 }, { "epoch": 47.183259911894275, "grad_norm": 0.1835532933473587, "learning_rate": 1.656988964425954e-06, "loss": 0.0051, "step": 3303 }, { "epoch": 47.19735682819383, "grad_norm": 0.15074199438095093, "learning_rate": 1.6402554605838172e-06, "loss": 0.0054, "step": 3304 }, { "epoch": 47.21145374449339, "grad_norm": 0.16869275271892548, "learning_rate": 1.6236061810686487e-06, "loss": 0.0055, "step": 3305 }, { "epoch": 47.225550660792955, "grad_norm": 0.1851722151041031, "learning_rate": 1.6070411401370334e-06, "loss": 0.0055, "step": 3306 }, { "epoch": 47.23964757709251, "grad_norm": 0.24182730913162231, "learning_rate": 1.5905603519734468e-06, "loss": 0.0052, "step": 3307 }, { "epoch": 47.25374449339207, "grad_norm": 0.3833407759666443, "learning_rate": 1.5741638306901895e-06, "loss": 0.0053, "step": 3308 }, { "epoch": 47.26784140969163, "grad_norm": 0.1854896992444992, "learning_rate": 1.5578515903274194e-06, "loss": 0.005, "step": 3309 }, { "epoch": 47.28193832599119, "grad_norm": 0.22857484221458435, "learning_rate": 1.5416236448531074e-06, "loss": 0.0051, "step": 3310 }, { "epoch": 47.29603524229075, "grad_norm": 0.18825024366378784, "learning_rate": 1.5254800081630826e-06, "loss": 0.0053, "step": 3311 }, { "epoch": 47.31013215859031, "grad_norm": 0.17823229730129242, "learning_rate": 1.5094206940809429e-06, "loss": 0.0053, "step": 3312 }, { "epoch": 47.32422907488987, "grad_norm": 0.17839904129505157, "learning_rate": 1.4934457163580885e-06, "loss": 0.005, "step": 3313 }, { "epoch": 47.33832599118943, "grad_norm": 0.17847523093223572, "learning_rate": 1.4775550886737099e-06, "loss": 0.0053, "step": 3314 }, { "epoch": 47.352422907488986, "grad_norm": 0.16949820518493652, "learning_rate": 1.461748824634801e-06, "loss": 0.005, "step": 3315 }, { "epoch": 47.36651982378854, "grad_norm": 0.14676721394062042, "learning_rate": 1.446026937776046e-06, "loss": 0.0049, "step": 3316 }, { "epoch": 47.38061674008811, "grad_norm": 0.12744539976119995, "learning_rate": 1.4303894415599318e-06, "loss": 0.005, "step": 3317 }, { "epoch": 47.394713656387665, "grad_norm": 0.15543480217456818, "learning_rate": 1.4148363493766802e-06, "loss": 0.0051, "step": 3318 }, { "epoch": 47.40881057268722, "grad_norm": 0.14195838570594788, "learning_rate": 1.3993676745442163e-06, "loss": 0.0055, "step": 3319 }, { "epoch": 47.42290748898679, "grad_norm": 0.14566664397716522, "learning_rate": 1.383983430308189e-06, "loss": 0.0047, "step": 3320 }, { "epoch": 47.437004405286345, "grad_norm": 0.11560948193073273, "learning_rate": 1.3686836298419492e-06, "loss": 0.0051, "step": 3321 }, { "epoch": 47.4511013215859, "grad_norm": 0.216293603181839, "learning_rate": 1.3534682862465397e-06, "loss": 0.0053, "step": 3322 }, { "epoch": 47.46519823788546, "grad_norm": 0.20462720096111298, "learning_rate": 1.3383374125507054e-06, "loss": 0.0052, "step": 3323 }, { "epoch": 47.479295154185024, "grad_norm": 0.1591673195362091, "learning_rate": 1.3232910217108153e-06, "loss": 0.0055, "step": 3324 }, { "epoch": 47.49339207048458, "grad_norm": 0.1939169019460678, "learning_rate": 1.30832912661093e-06, "loss": 0.0056, "step": 3325 }, { "epoch": 47.50748898678414, "grad_norm": 0.23221001029014587, "learning_rate": 1.2934517400627565e-06, "loss": 0.0053, "step": 3326 }, { "epoch": 47.521585903083704, "grad_norm": 0.23326475918293, "learning_rate": 1.2786588748056383e-06, "loss": 0.0047, "step": 3327 }, { "epoch": 47.53568281938326, "grad_norm": 0.17013853788375854, "learning_rate": 1.2639505435065091e-06, "loss": 0.0052, "step": 3328 }, { "epoch": 47.54977973568282, "grad_norm": 0.13661177456378937, "learning_rate": 1.2493267587599723e-06, "loss": 0.0056, "step": 3329 }, { "epoch": 47.563876651982376, "grad_norm": 0.2296970784664154, "learning_rate": 1.2347875330881886e-06, "loss": 0.0048, "step": 3330 }, { "epoch": 47.57797356828194, "grad_norm": 0.1067473366856575, "learning_rate": 1.2203328789409552e-06, "loss": 0.0048, "step": 3331 }, { "epoch": 47.5920704845815, "grad_norm": 0.18074867129325867, "learning_rate": 1.2059628086956044e-06, "loss": 0.0057, "step": 3332 }, { "epoch": 47.606167400881056, "grad_norm": 0.22049853205680847, "learning_rate": 1.1916773346570599e-06, "loss": 0.0049, "step": 3333 }, { "epoch": 47.62026431718062, "grad_norm": 0.12493827193975449, "learning_rate": 1.1774764690578365e-06, "loss": 0.0048, "step": 3334 }, { "epoch": 47.63436123348018, "grad_norm": 0.1554771512746811, "learning_rate": 1.1633602240579522e-06, "loss": 0.0049, "step": 3335 }, { "epoch": 47.648458149779735, "grad_norm": 0.15109919011592865, "learning_rate": 1.149328611744993e-06, "loss": 0.005, "step": 3336 }, { "epoch": 47.66255506607929, "grad_norm": 0.19867581129074097, "learning_rate": 1.1353816441340704e-06, "loss": 0.0051, "step": 3337 }, { "epoch": 47.67665198237886, "grad_norm": 0.18867865204811096, "learning_rate": 1.1215193331678198e-06, "loss": 0.0054, "step": 3338 }, { "epoch": 47.690748898678414, "grad_norm": 0.2238408327102661, "learning_rate": 1.1077416907163574e-06, "loss": 0.0049, "step": 3339 }, { "epoch": 47.70484581497797, "grad_norm": 0.17140847444534302, "learning_rate": 1.0940487285773459e-06, "loss": 0.0053, "step": 3340 }, { "epoch": 47.71894273127754, "grad_norm": 0.17628434300422668, "learning_rate": 1.0804404584758954e-06, "loss": 0.0051, "step": 3341 }, { "epoch": 47.733039647577094, "grad_norm": 0.1180846318602562, "learning_rate": 1.0669168920646178e-06, "loss": 0.005, "step": 3342 }, { "epoch": 47.74713656387665, "grad_norm": 0.18354260921478271, "learning_rate": 1.0534780409235835e-06, "loss": 0.0053, "step": 3343 }, { "epoch": 47.76123348017621, "grad_norm": 0.25055161118507385, "learning_rate": 1.0401239165603427e-06, "loss": 0.0045, "step": 3344 }, { "epoch": 47.77533039647577, "grad_norm": 0.17795011401176453, "learning_rate": 1.026854530409882e-06, "loss": 0.0053, "step": 3345 }, { "epoch": 47.78942731277533, "grad_norm": 0.16970980167388916, "learning_rate": 1.0136698938346011e-06, "loss": 0.0053, "step": 3346 }, { "epoch": 47.80352422907489, "grad_norm": 0.13586905598640442, "learning_rate": 1.0005700181243693e-06, "loss": 0.005, "step": 3347 }, { "epoch": 47.817621145374446, "grad_norm": 0.12493523955345154, "learning_rate": 9.875549144964469e-07, "loss": 0.0052, "step": 3348 }, { "epoch": 47.83171806167401, "grad_norm": 0.19032184779644012, "learning_rate": 9.746245940955523e-07, "loss": 0.0053, "step": 3349 }, { "epoch": 47.84581497797357, "grad_norm": 0.25364425778388977, "learning_rate": 9.61779067993729e-07, "loss": 0.0053, "step": 3350 }, { "epoch": 47.859911894273125, "grad_norm": 0.10078289359807968, "learning_rate": 9.49018347190489e-07, "loss": 0.0048, "step": 3351 }, { "epoch": 47.87400881057269, "grad_norm": 0.11131312698125839, "learning_rate": 9.363424426126699e-07, "loss": 0.0047, "step": 3352 }, { "epoch": 47.88810572687225, "grad_norm": 0.1860755831003189, "learning_rate": 9.237513651145225e-07, "loss": 0.0051, "step": 3353 }, { "epoch": 47.902202643171805, "grad_norm": 0.11872584372758865, "learning_rate": 9.112451254776222e-07, "loss": 0.0047, "step": 3354 }, { "epoch": 47.91629955947137, "grad_norm": 0.1445693075656891, "learning_rate": 8.988237344109251e-07, "loss": 0.005, "step": 3355 }, { "epoch": 47.93039647577093, "grad_norm": 0.1689257025718689, "learning_rate": 8.864872025507342e-07, "loss": 0.0048, "step": 3356 }, { "epoch": 47.944493392070484, "grad_norm": 0.14817039668560028, "learning_rate": 8.742355404606773e-07, "loss": 0.005, "step": 3357 }, { "epoch": 47.95859030837004, "grad_norm": 0.13473553955554962, "learning_rate": 8.620687586317177e-07, "loss": 0.0055, "step": 3358 }, { "epoch": 47.972687224669606, "grad_norm": 0.21030649542808533, "learning_rate": 8.499868674821221e-07, "loss": 0.0048, "step": 3359 }, { "epoch": 47.986784140969164, "grad_norm": 0.09564495086669922, "learning_rate": 8.379898773574924e-07, "loss": 0.0048, "step": 3360 }, { "epoch": 48.01409691629956, "grad_norm": 0.09008097648620605, "learning_rate": 8.260777985306889e-07, "loss": 0.0088, "step": 3361 }, { "epoch": 48.02819383259912, "grad_norm": 0.14406180381774902, "learning_rate": 8.14250641201908e-07, "loss": 0.0054, "step": 3362 }, { "epoch": 48.04229074889868, "grad_norm": 0.185677170753479, "learning_rate": 8.02508415498604e-07, "loss": 0.0048, "step": 3363 }, { "epoch": 48.05638766519824, "grad_norm": 0.12490896135568619, "learning_rate": 7.908511314755118e-07, "loss": 0.0056, "step": 3364 }, { "epoch": 48.070484581497794, "grad_norm": 0.19851799309253693, "learning_rate": 7.792787991146355e-07, "loss": 0.0048, "step": 3365 }, { "epoch": 48.08458149779736, "grad_norm": 0.14821159839630127, "learning_rate": 7.677914283252152e-07, "loss": 0.0053, "step": 3366 }, { "epoch": 48.098678414096916, "grad_norm": 0.2224603146314621, "learning_rate": 7.563890289437825e-07, "loss": 0.0053, "step": 3367 }, { "epoch": 48.112775330396474, "grad_norm": 0.23068180680274963, "learning_rate": 7.450716107340495e-07, "loss": 0.0054, "step": 3368 }, { "epoch": 48.12687224669604, "grad_norm": 0.13255973160266876, "learning_rate": 7.338391833870084e-07, "loss": 0.0052, "step": 3369 }, { "epoch": 48.140969162995596, "grad_norm": 0.1833551675081253, "learning_rate": 7.226917565208657e-07, "loss": 0.0045, "step": 3370 }, { "epoch": 48.15506607929515, "grad_norm": 0.13117021322250366, "learning_rate": 7.116293396810303e-07, "loss": 0.0051, "step": 3371 }, { "epoch": 48.16916299559471, "grad_norm": 0.15292230248451233, "learning_rate": 7.006519423401137e-07, "loss": 0.005, "step": 3372 }, { "epoch": 48.183259911894275, "grad_norm": 0.11757276207208633, "learning_rate": 6.897595738979523e-07, "loss": 0.0049, "step": 3373 }, { "epoch": 48.19735682819383, "grad_norm": 0.12171191722154617, "learning_rate": 6.78952243681541e-07, "loss": 0.005, "step": 3374 }, { "epoch": 48.21145374449339, "grad_norm": 0.23208770155906677, "learning_rate": 6.682299609450993e-07, "loss": 0.0052, "step": 3375 }, { "epoch": 48.225550660792955, "grad_norm": 0.22014695405960083, "learning_rate": 6.575927348699606e-07, "loss": 0.0055, "step": 3376 }, { "epoch": 48.23964757709251, "grad_norm": 0.1388731598854065, "learning_rate": 6.470405745646724e-07, "loss": 0.005, "step": 3377 }, { "epoch": 48.25374449339207, "grad_norm": 0.13486085832118988, "learning_rate": 6.365734890649399e-07, "loss": 0.0054, "step": 3378 }, { "epoch": 48.26784140969163, "grad_norm": 0.20802487432956696, "learning_rate": 6.261914873336161e-07, "loss": 0.0051, "step": 3379 }, { "epoch": 48.28193832599119, "grad_norm": 0.08843111991882324, "learning_rate": 6.158945782606673e-07, "loss": 0.0048, "step": 3380 }, { "epoch": 48.29603524229075, "grad_norm": 0.14346489310264587, "learning_rate": 6.056827706632185e-07, "loss": 0.0054, "step": 3381 }, { "epoch": 48.31013215859031, "grad_norm": 0.17829114198684692, "learning_rate": 5.955560732855526e-07, "loss": 0.0051, "step": 3382 }, { "epoch": 48.32422907488987, "grad_norm": 0.16027168929576874, "learning_rate": 5.855144947990221e-07, "loss": 0.0055, "step": 3383 }, { "epoch": 48.33832599118943, "grad_norm": 0.22895179688930511, "learning_rate": 5.755580438021157e-07, "loss": 0.0059, "step": 3384 }, { "epoch": 48.352422907488986, "grad_norm": 0.22922536730766296, "learning_rate": 5.65686728820436e-07, "loss": 0.0052, "step": 3385 }, { "epoch": 48.36651982378854, "grad_norm": 0.12064827233552933, "learning_rate": 5.559005583066767e-07, "loss": 0.0049, "step": 3386 }, { "epoch": 48.38061674008811, "grad_norm": 0.1648784577846527, "learning_rate": 5.461995406406239e-07, "loss": 0.0063, "step": 3387 }, { "epoch": 48.394713656387665, "grad_norm": 0.23036998510360718, "learning_rate": 5.365836841291438e-07, "loss": 0.0054, "step": 3388 }, { "epoch": 48.40881057268722, "grad_norm": 0.13731110095977783, "learning_rate": 5.270529970061944e-07, "loss": 0.005, "step": 3389 }, { "epoch": 48.42290748898679, "grad_norm": 0.2294774204492569, "learning_rate": 5.176074874327919e-07, "loss": 0.0053, "step": 3390 }, { "epoch": 48.437004405286345, "grad_norm": 0.10320563614368439, "learning_rate": 5.082471634970109e-07, "loss": 0.0056, "step": 3391 }, { "epoch": 48.4511013215859, "grad_norm": 0.227541983127594, "learning_rate": 4.989720332139847e-07, "loss": 0.0049, "step": 3392 }, { "epoch": 48.46519823788546, "grad_norm": 0.17654788494110107, "learning_rate": 4.897821045259154e-07, "loss": 0.005, "step": 3393 }, { "epoch": 48.479295154185024, "grad_norm": 0.19405320286750793, "learning_rate": 4.806773853020419e-07, "loss": 0.0058, "step": 3394 }, { "epoch": 48.49339207048458, "grad_norm": 0.17391347885131836, "learning_rate": 4.7165788333860536e-07, "loss": 0.0047, "step": 3395 }, { "epoch": 48.50748898678414, "grad_norm": 0.1584153175354004, "learning_rate": 4.627236063589169e-07, "loss": 0.0054, "step": 3396 }, { "epoch": 48.521585903083704, "grad_norm": 0.1561795324087143, "learning_rate": 4.5387456201329006e-07, "loss": 0.0051, "step": 3397 }, { "epoch": 48.53568281938326, "grad_norm": 0.15638592839241028, "learning_rate": 4.4511075787907476e-07, "loss": 0.0049, "step": 3398 }, { "epoch": 48.54977973568282, "grad_norm": 0.20842371881008148, "learning_rate": 4.364322014606126e-07, "loss": 0.005, "step": 3399 }, { "epoch": 48.563876651982376, "grad_norm": 0.1745104193687439, "learning_rate": 4.278389001892369e-07, "loss": 0.0054, "step": 3400 }, { "epoch": 48.57797356828194, "grad_norm": 0.28112706542015076, "learning_rate": 4.19330861423306e-07, "loss": 0.0051, "step": 3401 }, { "epoch": 48.5920704845815, "grad_norm": 0.09561113268136978, "learning_rate": 4.1090809244814785e-07, "loss": 0.0046, "step": 3402 }, { "epoch": 48.606167400881056, "grad_norm": 0.13022972643375397, "learning_rate": 4.025706004760932e-07, "loss": 0.0051, "step": 3403 }, { "epoch": 48.62026431718062, "grad_norm": 0.17977289855480194, "learning_rate": 3.943183926464422e-07, "loss": 0.0051, "step": 3404 }, { "epoch": 48.63436123348018, "grad_norm": 0.22141589224338531, "learning_rate": 3.8615147602546473e-07, "loss": 0.0057, "step": 3405 }, { "epoch": 48.648458149779735, "grad_norm": 0.1746680587530136, "learning_rate": 3.7806985760638904e-07, "loss": 0.005, "step": 3406 }, { "epoch": 48.66255506607929, "grad_norm": 0.1466980278491974, "learning_rate": 3.700735443094128e-07, "loss": 0.0051, "step": 3407 }, { "epoch": 48.67665198237886, "grad_norm": 0.12130752950906754, "learning_rate": 3.6216254298170326e-07, "loss": 0.005, "step": 3408 }, { "epoch": 48.690748898678414, "grad_norm": 0.14321866631507874, "learning_rate": 3.543368603973529e-07, "loss": 0.005, "step": 3409 }, { "epoch": 48.70484581497797, "grad_norm": 0.1566079556941986, "learning_rate": 3.4659650325740147e-07, "loss": 0.0045, "step": 3410 }, { "epoch": 48.71894273127754, "grad_norm": 0.09472651779651642, "learning_rate": 3.389414781898248e-07, "loss": 0.0053, "step": 3411 }, { "epoch": 48.733039647577094, "grad_norm": 0.21215948462486267, "learning_rate": 3.3137179174953513e-07, "loss": 0.0053, "step": 3412 }, { "epoch": 48.74713656387665, "grad_norm": 0.1421307772397995, "learning_rate": 3.2388745041838084e-07, "loss": 0.0048, "step": 3413 }, { "epoch": 48.76123348017621, "grad_norm": 0.2200143188238144, "learning_rate": 3.1648846060511327e-07, "loss": 0.0055, "step": 3414 }, { "epoch": 48.77533039647577, "grad_norm": 0.11542090773582458, "learning_rate": 3.091748286453866e-07, "loss": 0.0047, "step": 3415 }, { "epoch": 48.78942731277533, "grad_norm": 0.10758901387453079, "learning_rate": 3.019465608018024e-07, "loss": 0.0048, "step": 3416 }, { "epoch": 48.80352422907489, "grad_norm": 0.15592269599437714, "learning_rate": 2.948036632638318e-07, "loss": 0.005, "step": 3417 }, { "epoch": 48.817621145374446, "grad_norm": 0.1289716213941574, "learning_rate": 2.8774614214786e-07, "loss": 0.0047, "step": 3418 }, { "epoch": 48.83171806167401, "grad_norm": 0.1315491944551468, "learning_rate": 2.8077400349715287e-07, "loss": 0.005, "step": 3419 }, { "epoch": 48.84581497797357, "grad_norm": 0.11354026943445206, "learning_rate": 2.7388725328189036e-07, "loss": 0.0049, "step": 3420 }, { "epoch": 48.859911894273125, "grad_norm": 0.1244683489203453, "learning_rate": 2.6708589739909973e-07, "loss": 0.0049, "step": 3421 }, { "epoch": 48.87400881057269, "grad_norm": 0.14770478010177612, "learning_rate": 2.6036994167272235e-07, "loss": 0.0048, "step": 3422 }, { "epoch": 48.88810572687225, "grad_norm": 0.16919481754302979, "learning_rate": 2.537393918535358e-07, "loss": 0.0051, "step": 3423 }, { "epoch": 48.902202643171805, "grad_norm": 0.17387746274471283, "learning_rate": 2.4719425361923175e-07, "loss": 0.0058, "step": 3424 }, { "epoch": 48.91629955947137, "grad_norm": 0.29527902603149414, "learning_rate": 2.407345325743049e-07, "loss": 0.0056, "step": 3425 }, { "epoch": 48.93039647577093, "grad_norm": 0.15345600247383118, "learning_rate": 2.3436023425015275e-07, "loss": 0.005, "step": 3426 }, { "epoch": 48.944493392070484, "grad_norm": 0.12046130001544952, "learning_rate": 2.2807136410503137e-07, "loss": 0.0052, "step": 3427 }, { "epoch": 48.95859030837004, "grad_norm": 0.21992115676403046, "learning_rate": 2.21867927524011e-07, "loss": 0.0052, "step": 3428 }, { "epoch": 48.972687224669606, "grad_norm": 0.10843037813901901, "learning_rate": 2.157499298190202e-07, "loss": 0.0051, "step": 3429 }, { "epoch": 48.986784140969164, "grad_norm": 0.11199086904525757, "learning_rate": 2.0971737622883515e-07, "loss": 0.0048, "step": 3430 }, { "epoch": 49.01409691629956, "grad_norm": 0.1703552007675171, "learning_rate": 2.0377027191906816e-07, "loss": 0.0097, "step": 3431 }, { "epoch": 49.02819383259912, "grad_norm": 0.24098294973373413, "learning_rate": 1.9790862198215686e-07, "loss": 0.0052, "step": 3432 }, { "epoch": 49.04229074889868, "grad_norm": 0.1591465324163437, "learning_rate": 1.9213243143735293e-07, "loss": 0.005, "step": 3433 }, { "epoch": 49.05638766519824, "grad_norm": 0.22354242205619812, "learning_rate": 1.8644170523075542e-07, "loss": 0.0052, "step": 3434 }, { "epoch": 49.070484581497794, "grad_norm": 0.10176011919975281, "learning_rate": 1.8083644823526647e-07, "loss": 0.0046, "step": 3435 }, { "epoch": 49.08458149779736, "grad_norm": 0.0874878466129303, "learning_rate": 1.753166652506022e-07, "loss": 0.0046, "step": 3436 }, { "epoch": 49.098678414096916, "grad_norm": 0.3064175546169281, "learning_rate": 1.6988236100329292e-07, "loss": 0.0056, "step": 3437 }, { "epoch": 49.112775330396474, "grad_norm": 0.13333846628665924, "learning_rate": 1.6453354014666077e-07, "loss": 0.0047, "step": 3438 }, { "epoch": 49.12687224669604, "grad_norm": 0.13867926597595215, "learning_rate": 1.5927020726084208e-07, "loss": 0.0049, "step": 3439 }, { "epoch": 49.140969162995596, "grad_norm": 0.10166803002357483, "learning_rate": 1.5409236685277605e-07, "loss": 0.0049, "step": 3440 }, { "epoch": 49.15506607929515, "grad_norm": 0.15353748202323914, "learning_rate": 1.490000233561939e-07, "loss": 0.0054, "step": 3441 }, { "epoch": 49.16916299559471, "grad_norm": 0.24546825885772705, "learning_rate": 1.4399318113159642e-07, "loss": 0.0055, "step": 3442 }, { "epoch": 49.183259911894275, "grad_norm": 0.1380389928817749, "learning_rate": 1.3907184446628751e-07, "loss": 0.0048, "step": 3443 }, { "epoch": 49.19735682819383, "grad_norm": 0.1627977341413498, "learning_rate": 1.3423601757436287e-07, "loss": 0.0052, "step": 3444 }, { "epoch": 49.21145374449339, "grad_norm": 0.16835033893585205, "learning_rate": 1.2948570459667686e-07, "loss": 0.0048, "step": 3445 }, { "epoch": 49.225550660792955, "grad_norm": 0.15917377173900604, "learning_rate": 1.2482090960086457e-07, "loss": 0.005, "step": 3446 }, { "epoch": 49.23964757709251, "grad_norm": 0.13692837953567505, "learning_rate": 1.202416365813419e-07, "loss": 0.0047, "step": 3447 }, { "epoch": 49.25374449339207, "grad_norm": 0.20360274612903595, "learning_rate": 1.1574788945929449e-07, "loss": 0.0056, "step": 3448 }, { "epoch": 49.26784140969163, "grad_norm": 0.12393037229776382, "learning_rate": 1.1133967208266649e-07, "loss": 0.0048, "step": 3449 }, { "epoch": 49.28193832599119, "grad_norm": 0.215937077999115, "learning_rate": 1.0701698822614959e-07, "loss": 0.0053, "step": 3450 }, { "epoch": 49.29603524229075, "grad_norm": 0.12980450689792633, "learning_rate": 1.0277984159122733e-07, "loss": 0.0054, "step": 3451 }, { "epoch": 49.31013215859031, "grad_norm": 0.16118280589580536, "learning_rate": 9.862823580609748e-08, "loss": 0.0055, "step": 3452 }, { "epoch": 49.32422907488987, "grad_norm": 0.17360904812812805, "learning_rate": 9.456217442576076e-08, "loss": 0.0052, "step": 3453 }, { "epoch": 49.33832599118943, "grad_norm": 0.16218332946300507, "learning_rate": 9.058166093190989e-08, "loss": 0.005, "step": 3454 }, { "epoch": 49.352422907488986, "grad_norm": 0.1862349808216095, "learning_rate": 8.668669873304058e-08, "loss": 0.0055, "step": 3455 }, { "epoch": 49.36651982378854, "grad_norm": 0.15025226771831512, "learning_rate": 8.28772911643405e-08, "loss": 0.0052, "step": 3456 }, { "epoch": 49.38061674008811, "grad_norm": 0.14368559420108795, "learning_rate": 7.915344148777814e-08, "loss": 0.0049, "step": 3457 }, { "epoch": 49.394713656387665, "grad_norm": 0.12663346529006958, "learning_rate": 7.551515289203615e-08, "loss": 0.0048, "step": 3458 }, { "epoch": 49.40881057268722, "grad_norm": 0.22469303011894226, "learning_rate": 7.196242849253354e-08, "loss": 0.006, "step": 3459 }, { "epoch": 49.42290748898679, "grad_norm": 0.2338666319847107, "learning_rate": 6.849527133144796e-08, "loss": 0.0052, "step": 3460 }, { "epoch": 49.437004405286345, "grad_norm": 0.1263556331396103, "learning_rate": 6.511368437764898e-08, "loss": 0.005, "step": 3461 }, { "epoch": 49.4511013215859, "grad_norm": 0.162134051322937, "learning_rate": 6.18176705267537e-08, "loss": 0.0052, "step": 3462 }, { "epoch": 49.46519823788546, "grad_norm": 0.14820446074008942, "learning_rate": 5.860723260110445e-08, "loss": 0.0051, "step": 3463 }, { "epoch": 49.479295154185024, "grad_norm": 0.1633535474538803, "learning_rate": 5.548237334975781e-08, "loss": 0.0052, "step": 3464 }, { "epoch": 49.49339207048458, "grad_norm": 0.20993120968341827, "learning_rate": 5.2443095448506674e-08, "loss": 0.0054, "step": 3465 }, { "epoch": 49.50748898678414, "grad_norm": 0.1944410353899002, "learning_rate": 4.9489401499847046e-08, "loss": 0.0052, "step": 3466 }, { "epoch": 49.521585903083704, "grad_norm": 0.12426972389221191, "learning_rate": 4.662129403297799e-08, "loss": 0.0047, "step": 3467 }, { "epoch": 49.53568281938326, "grad_norm": 0.15334121882915497, "learning_rate": 4.3838775503834974e-08, "loss": 0.0047, "step": 3468 }, { "epoch": 49.54977973568282, "grad_norm": 0.1623985916376114, "learning_rate": 4.114184829506762e-08, "loss": 0.0051, "step": 3469 }, { "epoch": 49.563876651982376, "grad_norm": 0.19940084218978882, "learning_rate": 3.8530514716017544e-08, "loss": 0.005, "step": 3470 }, { "epoch": 49.57797356828194, "grad_norm": 0.11215222626924515, "learning_rate": 3.600477700272942e-08, "loss": 0.0047, "step": 3471 }, { "epoch": 49.5920704845815, "grad_norm": 0.18384836614131927, "learning_rate": 3.356463731798432e-08, "loss": 0.0053, "step": 3472 }, { "epoch": 49.606167400881056, "grad_norm": 0.16840510070323944, "learning_rate": 3.121009775122197e-08, "loss": 0.0049, "step": 3473 }, { "epoch": 49.62026431718062, "grad_norm": 0.16464023292064667, "learning_rate": 2.894116031862959e-08, "loss": 0.0051, "step": 3474 }, { "epoch": 49.63436123348018, "grad_norm": 0.1754046082496643, "learning_rate": 2.6757826963053066e-08, "loss": 0.0054, "step": 3475 }, { "epoch": 49.648458149779735, "grad_norm": 0.204054594039917, "learning_rate": 2.4660099554074667e-08, "loss": 0.0052, "step": 3476 }, { "epoch": 49.66255506607929, "grad_norm": 0.2882646322250366, "learning_rate": 2.2647979887946424e-08, "loss": 0.0049, "step": 3477 }, { "epoch": 49.67665198237886, "grad_norm": 0.16121457517147064, "learning_rate": 2.0721469687612348e-08, "loss": 0.0055, "step": 3478 }, { "epoch": 49.690748898678414, "grad_norm": 0.20051628351211548, "learning_rate": 1.888057060274173e-08, "loss": 0.0052, "step": 3479 }, { "epoch": 49.70484581497797, "grad_norm": 0.15214277803897858, "learning_rate": 1.712528420966253e-08, "loss": 0.0052, "step": 3480 }, { "epoch": 49.71894273127754, "grad_norm": 0.19046339392662048, "learning_rate": 1.5455612011405773e-08, "loss": 0.0054, "step": 3481 }, { "epoch": 49.733039647577094, "grad_norm": 0.1639482080936432, "learning_rate": 1.3871555437694472e-08, "loss": 0.0049, "step": 3482 }, { "epoch": 49.74713656387665, "grad_norm": 0.1836724579334259, "learning_rate": 1.2373115844943606e-08, "loss": 0.0055, "step": 3483 }, { "epoch": 49.76123348017621, "grad_norm": 0.21310393512248993, "learning_rate": 1.0960294516237924e-08, "loss": 0.0052, "step": 3484 }, { "epoch": 49.77533039647577, "grad_norm": 0.1910141259431839, "learning_rate": 9.633092661376352e-09, "loss": 0.0051, "step": 3485 }, { "epoch": 49.78942731277533, "grad_norm": 0.12836365401744843, "learning_rate": 8.391511416816489e-09, "loss": 0.0048, "step": 3486 }, { "epoch": 49.80352422907489, "grad_norm": 0.16087549924850464, "learning_rate": 7.2355518457079e-09, "loss": 0.0054, "step": 3487 }, { "epoch": 49.817621145374446, "grad_norm": 0.1536581814289093, "learning_rate": 6.165214937892127e-09, "loss": 0.0052, "step": 3488 }, { "epoch": 49.83171806167401, "grad_norm": 0.17593036592006683, "learning_rate": 5.1805016098915856e-09, "loss": 0.0055, "step": 3489 }, { "epoch": 49.84581497797357, "grad_norm": 0.13062648475170135, "learning_rate": 4.281412704887356e-09, "loss": 0.0051, "step": 3490 }, { "epoch": 49.859911894273125, "grad_norm": 0.3169082999229431, "learning_rate": 3.4679489927746943e-09, "loss": 0.0049, "step": 3491 }, { "epoch": 49.87400881057269, "grad_norm": 0.14756940305233002, "learning_rate": 2.740111170107529e-09, "loss": 0.0051, "step": 3492 }, { "epoch": 49.88810572687225, "grad_norm": 0.1629554033279419, "learning_rate": 2.0978998601206556e-09, "loss": 0.0055, "step": 3493 }, { "epoch": 49.902202643171805, "grad_norm": 0.2318698763847351, "learning_rate": 1.5413156127408457e-09, "loss": 0.0049, "step": 3494 }, { "epoch": 49.91629955947137, "grad_norm": 0.2624201476573944, "learning_rate": 1.070358904564639e-09, "loss": 0.0053, "step": 3495 }, { "epoch": 49.93039647577093, "grad_norm": 0.23803582787513733, "learning_rate": 6.850301388583446e-10, "loss": 0.0056, "step": 3496 }, { "epoch": 49.944493392070484, "grad_norm": 0.10023318231105804, "learning_rate": 3.8532964558024577e-10, "loss": 0.0045, "step": 3497 }, { "epoch": 49.95859030837004, "grad_norm": 0.12463587522506714, "learning_rate": 1.7125768135839436e-10, "loss": 0.0051, "step": 3498 }, { "epoch": 49.972687224669606, "grad_norm": 0.16557706892490387, "learning_rate": 4.2814429501714106e-11, "loss": 0.0049, "step": 3499 }, { "epoch": 49.986784140969164, "grad_norm": 0.10480985045433044, "learning_rate": 0.0, "loss": 0.0047, "step": 3500 }, { "epoch": 49.986784140969164, "eval_loss": 1.453356146812439, "eval_runtime": 133.5101, "eval_samples_per_second": 13.542, "eval_steps_per_second": 0.427, "step": 3500 }, { "epoch": 49.986784140969164, "step": 3500, "total_flos": 3.735366265481462e+19, "train_loss": 0.0011541401028100932, "train_runtime": 35303.552, "train_samples_per_second": 25.716, "train_steps_per_second": 0.099 } ], "logging_steps": 1.0, "max_steps": 3500, "num_input_tokens_seen": 0, "num_train_epochs": 50, "save_steps": 250, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3.735366265481462e+19, "train_batch_size": 4, "trial_name": null, "trial_params": null }