paligeema_hindi_vgqa_v1.1 / trainer_state.json
sam2ai's picture
Upload folder using huggingface_hub
9ba5625 verified
raw
history blame contribute delete
No virus
156 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.9834710743801653,
"eval_steps": 500,
"global_step": 900,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0022038567493112946,
"grad_norm": 53.35355576018126,
"learning_rate": 6.000000000000001e-08,
"loss": 3.7938,
"step": 1
},
{
"epoch": 0.004407713498622589,
"grad_norm": 56.29807371103618,
"learning_rate": 1.2000000000000002e-07,
"loss": 3.9718,
"step": 2
},
{
"epoch": 0.006611570247933884,
"grad_norm": 61.23476417502119,
"learning_rate": 1.8e-07,
"loss": 4.0815,
"step": 3
},
{
"epoch": 0.008815426997245178,
"grad_norm": 57.23654746894239,
"learning_rate": 2.4000000000000003e-07,
"loss": 4.0256,
"step": 4
},
{
"epoch": 0.011019283746556474,
"grad_norm": 53.79273733579391,
"learning_rate": 3.0000000000000004e-07,
"loss": 3.9356,
"step": 5
},
{
"epoch": 0.013223140495867768,
"grad_norm": 54.1365703275303,
"learning_rate": 3.6e-07,
"loss": 4.0175,
"step": 6
},
{
"epoch": 0.015426997245179064,
"grad_norm": 59.57834616370979,
"learning_rate": 4.2000000000000006e-07,
"loss": 4.0324,
"step": 7
},
{
"epoch": 0.017630853994490357,
"grad_norm": 52.92978303398463,
"learning_rate": 4.800000000000001e-07,
"loss": 3.9872,
"step": 8
},
{
"epoch": 0.019834710743801654,
"grad_norm": 53.00990339068272,
"learning_rate": 5.4e-07,
"loss": 3.9189,
"step": 9
},
{
"epoch": 0.02203856749311295,
"grad_norm": 49.29917728226077,
"learning_rate": 6.000000000000001e-07,
"loss": 3.8985,
"step": 10
},
{
"epoch": 0.024242424242424242,
"grad_norm": 51.04476681098937,
"learning_rate": 6.6e-07,
"loss": 4.0225,
"step": 11
},
{
"epoch": 0.026446280991735537,
"grad_norm": 48.910852181006604,
"learning_rate": 7.2e-07,
"loss": 3.8215,
"step": 12
},
{
"epoch": 0.02865013774104683,
"grad_norm": 47.57420311673295,
"learning_rate": 7.8e-07,
"loss": 3.8253,
"step": 13
},
{
"epoch": 0.03085399449035813,
"grad_norm": 48.04076110975242,
"learning_rate": 8.400000000000001e-07,
"loss": 4.0329,
"step": 14
},
{
"epoch": 0.03305785123966942,
"grad_norm": 45.10942040856652,
"learning_rate": 9e-07,
"loss": 3.6307,
"step": 15
},
{
"epoch": 0.03526170798898071,
"grad_norm": 43.52414046751455,
"learning_rate": 9.600000000000001e-07,
"loss": 3.5453,
"step": 16
},
{
"epoch": 0.03746556473829201,
"grad_norm": 42.85848752134229,
"learning_rate": 1.0200000000000002e-06,
"loss": 3.6339,
"step": 17
},
{
"epoch": 0.03966942148760331,
"grad_norm": 33.19526904841194,
"learning_rate": 1.08e-06,
"loss": 3.4207,
"step": 18
},
{
"epoch": 0.0418732782369146,
"grad_norm": 31.873243117941165,
"learning_rate": 1.14e-06,
"loss": 3.191,
"step": 19
},
{
"epoch": 0.0440771349862259,
"grad_norm": 29.808709603971984,
"learning_rate": 1.2000000000000002e-06,
"loss": 3.2128,
"step": 20
},
{
"epoch": 0.04628099173553719,
"grad_norm": 31.51951976510752,
"learning_rate": 1.26e-06,
"loss": 3.3406,
"step": 21
},
{
"epoch": 0.048484848484848485,
"grad_norm": 27.823761692798925,
"learning_rate": 1.32e-06,
"loss": 3.2884,
"step": 22
},
{
"epoch": 0.05068870523415978,
"grad_norm": 26.429345221994904,
"learning_rate": 1.3800000000000001e-06,
"loss": 3.0015,
"step": 23
},
{
"epoch": 0.05289256198347107,
"grad_norm": 18.155827199892627,
"learning_rate": 1.44e-06,
"loss": 2.892,
"step": 24
},
{
"epoch": 0.05509641873278237,
"grad_norm": 13.907874127382469,
"learning_rate": 1.5e-06,
"loss": 2.6402,
"step": 25
},
{
"epoch": 0.05730027548209366,
"grad_norm": 13.424593330905749,
"learning_rate": 1.56e-06,
"loss": 2.7094,
"step": 26
},
{
"epoch": 0.05950413223140496,
"grad_norm": 13.854259069645929,
"learning_rate": 1.6200000000000002e-06,
"loss": 2.8064,
"step": 27
},
{
"epoch": 0.06170798898071626,
"grad_norm": 11.350285228526166,
"learning_rate": 1.6800000000000002e-06,
"loss": 2.7105,
"step": 28
},
{
"epoch": 0.06391184573002755,
"grad_norm": 11.129757770703522,
"learning_rate": 1.7399999999999999e-06,
"loss": 2.6137,
"step": 29
},
{
"epoch": 0.06611570247933884,
"grad_norm": 10.863006958008057,
"learning_rate": 1.8e-06,
"loss": 2.6008,
"step": 30
},
{
"epoch": 0.06831955922865014,
"grad_norm": 9.805882809079606,
"learning_rate": 1.86e-06,
"loss": 2.5766,
"step": 31
},
{
"epoch": 0.07052341597796143,
"grad_norm": 11.265746488422517,
"learning_rate": 1.9200000000000003e-06,
"loss": 2.5667,
"step": 32
},
{
"epoch": 0.07272727272727272,
"grad_norm": 9.025293134976716,
"learning_rate": 1.98e-06,
"loss": 2.4655,
"step": 33
},
{
"epoch": 0.07493112947658402,
"grad_norm": 11.70511117627225,
"learning_rate": 2.0400000000000004e-06,
"loss": 2.5067,
"step": 34
},
{
"epoch": 0.07713498622589532,
"grad_norm": 10.81705842291705,
"learning_rate": 2.1e-06,
"loss": 2.2768,
"step": 35
},
{
"epoch": 0.07933884297520662,
"grad_norm": 8.888943224142057,
"learning_rate": 2.16e-06,
"loss": 2.314,
"step": 36
},
{
"epoch": 0.0815426997245179,
"grad_norm": 10.751954789293142,
"learning_rate": 2.22e-06,
"loss": 2.558,
"step": 37
},
{
"epoch": 0.0837465564738292,
"grad_norm": 9.989373040049472,
"learning_rate": 2.28e-06,
"loss": 2.1834,
"step": 38
},
{
"epoch": 0.0859504132231405,
"grad_norm": 11.453837389192644,
"learning_rate": 2.34e-06,
"loss": 2.3084,
"step": 39
},
{
"epoch": 0.0881542699724518,
"grad_norm": 14.679296883424508,
"learning_rate": 2.4000000000000003e-06,
"loss": 2.3127,
"step": 40
},
{
"epoch": 0.09035812672176309,
"grad_norm": 19.06017003733088,
"learning_rate": 2.4599999999999997e-06,
"loss": 2.1885,
"step": 41
},
{
"epoch": 0.09256198347107437,
"grad_norm": 25.283284564533062,
"learning_rate": 2.52e-06,
"loss": 2.1116,
"step": 42
},
{
"epoch": 0.09476584022038567,
"grad_norm": 25.192909931447222,
"learning_rate": 2.58e-06,
"loss": 1.9283,
"step": 43
},
{
"epoch": 0.09696969696969697,
"grad_norm": 17.926763299091263,
"learning_rate": 2.64e-06,
"loss": 1.8125,
"step": 44
},
{
"epoch": 0.09917355371900827,
"grad_norm": 15.969502217016815,
"learning_rate": 2.7e-06,
"loss": 1.8421,
"step": 45
},
{
"epoch": 0.10137741046831956,
"grad_norm": 13.673145143593812,
"learning_rate": 2.7600000000000003e-06,
"loss": 1.7697,
"step": 46
},
{
"epoch": 0.10358126721763085,
"grad_norm": 15.24764944924423,
"learning_rate": 2.82e-06,
"loss": 1.8159,
"step": 47
},
{
"epoch": 0.10578512396694215,
"grad_norm": 15.912109474132139,
"learning_rate": 2.88e-06,
"loss": 1.7813,
"step": 48
},
{
"epoch": 0.10798898071625344,
"grad_norm": 17.737822555539395,
"learning_rate": 2.9400000000000002e-06,
"loss": 1.6972,
"step": 49
},
{
"epoch": 0.11019283746556474,
"grad_norm": 14.271394993258342,
"learning_rate": 3e-06,
"loss": 1.6323,
"step": 50
},
{
"epoch": 0.11239669421487604,
"grad_norm": 17.453281450383354,
"learning_rate": 2.9999996311880685e-06,
"loss": 1.6936,
"step": 51
},
{
"epoch": 0.11460055096418732,
"grad_norm": 18.936110023489377,
"learning_rate": 2.999998524752454e-06,
"loss": 1.5923,
"step": 52
},
{
"epoch": 0.11680440771349862,
"grad_norm": 17.54586534165411,
"learning_rate": 2.9999966806937017e-06,
"loss": 1.4402,
"step": 53
},
{
"epoch": 0.11900826446280992,
"grad_norm": 19.941262856339648,
"learning_rate": 2.9999940990127173e-06,
"loss": 1.6165,
"step": 54
},
{
"epoch": 0.12121212121212122,
"grad_norm": 16.195665957935688,
"learning_rate": 2.9999907797107714e-06,
"loss": 1.6445,
"step": 55
},
{
"epoch": 0.12341597796143251,
"grad_norm": 17.906162449636255,
"learning_rate": 2.9999867227894957e-06,
"loss": 1.4222,
"step": 56
},
{
"epoch": 0.1256198347107438,
"grad_norm": 14.766409204808326,
"learning_rate": 2.9999819282508854e-06,
"loss": 1.5734,
"step": 57
},
{
"epoch": 0.1278236914600551,
"grad_norm": 15.330559102455767,
"learning_rate": 2.999976396097298e-06,
"loss": 1.3955,
"step": 58
},
{
"epoch": 0.13002754820936638,
"grad_norm": 17.671499530692422,
"learning_rate": 2.999970126331454e-06,
"loss": 1.3905,
"step": 59
},
{
"epoch": 0.1322314049586777,
"grad_norm": 16.161094239545125,
"learning_rate": 2.999963118956437e-06,
"loss": 1.4074,
"step": 60
},
{
"epoch": 0.13443526170798897,
"grad_norm": 14.463604561427449,
"learning_rate": 2.999955373975692e-06,
"loss": 1.3949,
"step": 61
},
{
"epoch": 0.13663911845730028,
"grad_norm": 22.599162124810228,
"learning_rate": 2.9999468913930286e-06,
"loss": 1.2523,
"step": 62
},
{
"epoch": 0.13884297520661157,
"grad_norm": 15.721083306967905,
"learning_rate": 2.999937671212617e-06,
"loss": 1.2407,
"step": 63
},
{
"epoch": 0.14104683195592285,
"grad_norm": 22.450444426883404,
"learning_rate": 2.999927713438992e-06,
"loss": 1.31,
"step": 64
},
{
"epoch": 0.14325068870523416,
"grad_norm": 13.111698223582062,
"learning_rate": 2.9999170180770503e-06,
"loss": 1.3233,
"step": 65
},
{
"epoch": 0.14545454545454545,
"grad_norm": 17.736015434280038,
"learning_rate": 2.999905585132051e-06,
"loss": 1.2608,
"step": 66
},
{
"epoch": 0.14765840220385676,
"grad_norm": 13.247075058563684,
"learning_rate": 2.9998934146096163e-06,
"loss": 1.226,
"step": 67
},
{
"epoch": 0.14986225895316804,
"grad_norm": 16.093995265042583,
"learning_rate": 2.9998805065157313e-06,
"loss": 1.1238,
"step": 68
},
{
"epoch": 0.15206611570247933,
"grad_norm": 11.789269283624224,
"learning_rate": 2.9998668608567433e-06,
"loss": 1.2928,
"step": 69
},
{
"epoch": 0.15426997245179064,
"grad_norm": 12.799525613499874,
"learning_rate": 2.9998524776393626e-06,
"loss": 1.1069,
"step": 70
},
{
"epoch": 0.15647382920110192,
"grad_norm": 17.988308354823435,
"learning_rate": 2.9998373568706622e-06,
"loss": 1.0719,
"step": 71
},
{
"epoch": 0.15867768595041323,
"grad_norm": 15.73635047176777,
"learning_rate": 2.9998214985580777e-06,
"loss": 1.2518,
"step": 72
},
{
"epoch": 0.16088154269972452,
"grad_norm": 14.654557879657558,
"learning_rate": 2.9998049027094073e-06,
"loss": 1.0366,
"step": 73
},
{
"epoch": 0.1630853994490358,
"grad_norm": 10.392700397421045,
"learning_rate": 2.9997875693328126e-06,
"loss": 1.1564,
"step": 74
},
{
"epoch": 0.1652892561983471,
"grad_norm": 13.308000530271709,
"learning_rate": 2.9997694984368163e-06,
"loss": 1.2075,
"step": 75
},
{
"epoch": 0.1674931129476584,
"grad_norm": 13.182861834247353,
"learning_rate": 2.999750690030305e-06,
"loss": 1.1836,
"step": 76
},
{
"epoch": 0.1696969696969697,
"grad_norm": 11.679937122230694,
"learning_rate": 2.9997311441225286e-06,
"loss": 0.9576,
"step": 77
},
{
"epoch": 0.171900826446281,
"grad_norm": 11.700036415813958,
"learning_rate": 2.9997108607230975e-06,
"loss": 1.1455,
"step": 78
},
{
"epoch": 0.17410468319559227,
"grad_norm": 21.553282982313565,
"learning_rate": 2.9996898398419867e-06,
"loss": 0.9717,
"step": 79
},
{
"epoch": 0.1763085399449036,
"grad_norm": 12.956436863018121,
"learning_rate": 2.9996680814895334e-06,
"loss": 0.9445,
"step": 80
},
{
"epoch": 0.17851239669421487,
"grad_norm": 11.392114741947795,
"learning_rate": 2.9996455856764373e-06,
"loss": 1.1615,
"step": 81
},
{
"epoch": 0.18071625344352618,
"grad_norm": 12.793644484917351,
"learning_rate": 2.99962235241376e-06,
"loss": 0.9005,
"step": 82
},
{
"epoch": 0.18292011019283747,
"grad_norm": 11.278622938092001,
"learning_rate": 2.9995983817129273e-06,
"loss": 0.8012,
"step": 83
},
{
"epoch": 0.18512396694214875,
"grad_norm": 11.924228852181994,
"learning_rate": 2.9995736735857256e-06,
"loss": 0.8359,
"step": 84
},
{
"epoch": 0.18732782369146006,
"grad_norm": 10.574117414679797,
"learning_rate": 2.9995482280443065e-06,
"loss": 0.8495,
"step": 85
},
{
"epoch": 0.18953168044077134,
"grad_norm": 11.186152581473769,
"learning_rate": 2.999522045101182e-06,
"loss": 0.9247,
"step": 86
},
{
"epoch": 0.19173553719008266,
"grad_norm": 10.654547641676169,
"learning_rate": 2.999495124769228e-06,
"loss": 0.9681,
"step": 87
},
{
"epoch": 0.19393939393939394,
"grad_norm": 8.61313592647621,
"learning_rate": 2.9994674670616814e-06,
"loss": 0.9317,
"step": 88
},
{
"epoch": 0.19614325068870522,
"grad_norm": 18.887157394643115,
"learning_rate": 2.9994390719921445e-06,
"loss": 0.9185,
"step": 89
},
{
"epoch": 0.19834710743801653,
"grad_norm": 16.89569803683573,
"learning_rate": 2.9994099395745795e-06,
"loss": 1.0256,
"step": 90
},
{
"epoch": 0.20055096418732782,
"grad_norm": 8.519494558720615,
"learning_rate": 2.9993800698233127e-06,
"loss": 0.72,
"step": 91
},
{
"epoch": 0.20275482093663913,
"grad_norm": 9.970310282841208,
"learning_rate": 2.999349462753032e-06,
"loss": 0.7108,
"step": 92
},
{
"epoch": 0.2049586776859504,
"grad_norm": 10.966172547876596,
"learning_rate": 2.9993181183787886e-06,
"loss": 0.8898,
"step": 93
},
{
"epoch": 0.2071625344352617,
"grad_norm": 8.457980829767811,
"learning_rate": 2.999286036715997e-06,
"loss": 0.7737,
"step": 94
},
{
"epoch": 0.209366391184573,
"grad_norm": 10.709756234820391,
"learning_rate": 2.999253217780432e-06,
"loss": 0.9179,
"step": 95
},
{
"epoch": 0.2115702479338843,
"grad_norm": 11.769839019646232,
"learning_rate": 2.999219661588233e-06,
"loss": 0.8461,
"step": 96
},
{
"epoch": 0.2137741046831956,
"grad_norm": 8.715508819021125,
"learning_rate": 2.9991853681559014e-06,
"loss": 0.8488,
"step": 97
},
{
"epoch": 0.2159779614325069,
"grad_norm": 11.957028477448837,
"learning_rate": 2.9991503375003e-06,
"loss": 0.7625,
"step": 98
},
{
"epoch": 0.21818181818181817,
"grad_norm": 9.66498196931626,
"learning_rate": 2.999114569638656e-06,
"loss": 1.1074,
"step": 99
},
{
"epoch": 0.22038567493112948,
"grad_norm": 10.269942780716466,
"learning_rate": 2.999078064588558e-06,
"loss": 0.6719,
"step": 100
},
{
"epoch": 0.22258953168044077,
"grad_norm": 10.52874796702632,
"learning_rate": 2.9990408223679575e-06,
"loss": 0.8706,
"step": 101
},
{
"epoch": 0.22479338842975208,
"grad_norm": 11.95952602668835,
"learning_rate": 2.999002842995168e-06,
"loss": 0.6669,
"step": 102
},
{
"epoch": 0.22699724517906336,
"grad_norm": 9.38549802466762,
"learning_rate": 2.998964126488866e-06,
"loss": 0.7366,
"step": 103
},
{
"epoch": 0.22920110192837465,
"grad_norm": 9.976231066334218,
"learning_rate": 2.9989246728680897e-06,
"loss": 0.6534,
"step": 104
},
{
"epoch": 0.23140495867768596,
"grad_norm": 9.102475454497984,
"learning_rate": 2.9988844821522415e-06,
"loss": 0.7173,
"step": 105
},
{
"epoch": 0.23360881542699724,
"grad_norm": 9.491160186253202,
"learning_rate": 2.9988435543610844e-06,
"loss": 0.6074,
"step": 106
},
{
"epoch": 0.23581267217630855,
"grad_norm": 10.513632694996241,
"learning_rate": 2.998801889514745e-06,
"loss": 0.6602,
"step": 107
},
{
"epoch": 0.23801652892561984,
"grad_norm": 9.646787450212365,
"learning_rate": 2.998759487633712e-06,
"loss": 0.6238,
"step": 108
},
{
"epoch": 0.24022038567493112,
"grad_norm": 9.92201571128595,
"learning_rate": 2.9987163487388357e-06,
"loss": 0.7945,
"step": 109
},
{
"epoch": 0.24242424242424243,
"grad_norm": 9.806488011921113,
"learning_rate": 2.9986724728513302e-06,
"loss": 0.9243,
"step": 110
},
{
"epoch": 0.24462809917355371,
"grad_norm": 8.879233481017383,
"learning_rate": 2.9986278599927717e-06,
"loss": 0.6467,
"step": 111
},
{
"epoch": 0.24683195592286503,
"grad_norm": 10.905347276046735,
"learning_rate": 2.9985825101850986e-06,
"loss": 0.8978,
"step": 112
},
{
"epoch": 0.2490358126721763,
"grad_norm": 8.064502908123723,
"learning_rate": 2.9985364234506108e-06,
"loss": 0.6498,
"step": 113
},
{
"epoch": 0.2512396694214876,
"grad_norm": 10.946829774416697,
"learning_rate": 2.998489599811972e-06,
"loss": 0.8144,
"step": 114
},
{
"epoch": 0.2534435261707989,
"grad_norm": 9.718407483089358,
"learning_rate": 2.998442039292208e-06,
"loss": 0.4826,
"step": 115
},
{
"epoch": 0.2556473829201102,
"grad_norm": 7.869198489038032,
"learning_rate": 2.998393741914706e-06,
"loss": 0.5715,
"step": 116
},
{
"epoch": 0.2578512396694215,
"grad_norm": 10.393656330664765,
"learning_rate": 2.998344707703216e-06,
"loss": 0.7082,
"step": 117
},
{
"epoch": 0.26005509641873276,
"grad_norm": 10.247308720424776,
"learning_rate": 2.9982949366818514e-06,
"loss": 0.6026,
"step": 118
},
{
"epoch": 0.2622589531680441,
"grad_norm": 15.762210245537455,
"learning_rate": 2.9982444288750864e-06,
"loss": 0.7132,
"step": 119
},
{
"epoch": 0.2644628099173554,
"grad_norm": 8.438074902652675,
"learning_rate": 2.9981931843077586e-06,
"loss": 0.7253,
"step": 120
},
{
"epoch": 0.26666666666666666,
"grad_norm": 9.95077096598991,
"learning_rate": 2.9981412030050676e-06,
"loss": 0.5475,
"step": 121
},
{
"epoch": 0.26887052341597795,
"grad_norm": 12.749558420196946,
"learning_rate": 2.9980884849925743e-06,
"loss": 0.5953,
"step": 122
},
{
"epoch": 0.27107438016528923,
"grad_norm": 7.913579849257431,
"learning_rate": 2.9980350302962033e-06,
"loss": 0.5146,
"step": 123
},
{
"epoch": 0.27327823691460057,
"grad_norm": 9.250651892900226,
"learning_rate": 2.9979808389422417e-06,
"loss": 0.6412,
"step": 124
},
{
"epoch": 0.27548209366391185,
"grad_norm": 9.158788687798582,
"learning_rate": 2.997925910957337e-06,
"loss": 0.5092,
"step": 125
},
{
"epoch": 0.27768595041322314,
"grad_norm": 10.088855837632275,
"learning_rate": 2.9978702463684995e-06,
"loss": 0.725,
"step": 126
},
{
"epoch": 0.2798898071625344,
"grad_norm": 9.915909343545366,
"learning_rate": 2.9978138452031037e-06,
"loss": 0.5691,
"step": 127
},
{
"epoch": 0.2820936639118457,
"grad_norm": 9.617311338418991,
"learning_rate": 2.997756707488884e-06,
"loss": 0.709,
"step": 128
},
{
"epoch": 0.28429752066115704,
"grad_norm": 8.156885740814575,
"learning_rate": 2.997698833253938e-06,
"loss": 0.4996,
"step": 129
},
{
"epoch": 0.2865013774104683,
"grad_norm": 10.82636109019181,
"learning_rate": 2.997640222526725e-06,
"loss": 0.5769,
"step": 130
},
{
"epoch": 0.2887052341597796,
"grad_norm": 9.323333527183065,
"learning_rate": 2.997580875336067e-06,
"loss": 0.6782,
"step": 131
},
{
"epoch": 0.2909090909090909,
"grad_norm": 8.168393819228521,
"learning_rate": 2.9975207917111483e-06,
"loss": 0.519,
"step": 132
},
{
"epoch": 0.2931129476584022,
"grad_norm": 8.430243817884287,
"learning_rate": 2.9974599716815145e-06,
"loss": 0.5014,
"step": 133
},
{
"epoch": 0.2953168044077135,
"grad_norm": 10.621443566182664,
"learning_rate": 2.997398415277074e-06,
"loss": 0.7323,
"step": 134
},
{
"epoch": 0.2975206611570248,
"grad_norm": 10.903203705940129,
"learning_rate": 2.997336122528097e-06,
"loss": 0.6959,
"step": 135
},
{
"epoch": 0.2997245179063361,
"grad_norm": 13.310218126786243,
"learning_rate": 2.997273093465216e-06,
"loss": 0.6452,
"step": 136
},
{
"epoch": 0.30192837465564737,
"grad_norm": 9.111054312909284,
"learning_rate": 2.997209328119425e-06,
"loss": 0.5339,
"step": 137
},
{
"epoch": 0.30413223140495865,
"grad_norm": 8.5483917916211,
"learning_rate": 2.997144826522082e-06,
"loss": 0.7473,
"step": 138
},
{
"epoch": 0.30633608815427,
"grad_norm": 8.974042959614973,
"learning_rate": 2.9970795887049043e-06,
"loss": 0.6025,
"step": 139
},
{
"epoch": 0.3085399449035813,
"grad_norm": 9.761885242579995,
"learning_rate": 2.9970136146999727e-06,
"loss": 0.5253,
"step": 140
},
{
"epoch": 0.31074380165289256,
"grad_norm": 8.072862771710003,
"learning_rate": 2.9969469045397298e-06,
"loss": 0.6012,
"step": 141
},
{
"epoch": 0.31294765840220384,
"grad_norm": 7.269223969882805,
"learning_rate": 2.996879458256981e-06,
"loss": 0.4309,
"step": 142
},
{
"epoch": 0.3151515151515151,
"grad_norm": 8.584363487659544,
"learning_rate": 2.996811275884892e-06,
"loss": 0.4436,
"step": 143
},
{
"epoch": 0.31735537190082647,
"grad_norm": 8.732413459437103,
"learning_rate": 2.9967423574569928e-06,
"loss": 0.3715,
"step": 144
},
{
"epoch": 0.31955922865013775,
"grad_norm": 11.467423318928054,
"learning_rate": 2.9966727030071725e-06,
"loss": 0.5512,
"step": 145
},
{
"epoch": 0.32176308539944903,
"grad_norm": 10.868202581950138,
"learning_rate": 2.996602312569684e-06,
"loss": 0.4402,
"step": 146
},
{
"epoch": 0.3239669421487603,
"grad_norm": 9.203463518102229,
"learning_rate": 2.9965311861791427e-06,
"loss": 0.4785,
"step": 147
},
{
"epoch": 0.3261707988980716,
"grad_norm": 9.705531063021647,
"learning_rate": 2.9964593238705237e-06,
"loss": 0.5051,
"step": 148
},
{
"epoch": 0.32837465564738294,
"grad_norm": 8.25057132852951,
"learning_rate": 2.996386725679166e-06,
"loss": 0.4323,
"step": 149
},
{
"epoch": 0.3305785123966942,
"grad_norm": 8.259086063115724,
"learning_rate": 2.996313391640769e-06,
"loss": 0.7087,
"step": 150
},
{
"epoch": 0.3327823691460055,
"grad_norm": 9.085318085110034,
"learning_rate": 2.9962393217913956e-06,
"loss": 0.5809,
"step": 151
},
{
"epoch": 0.3349862258953168,
"grad_norm": 11.724016100603865,
"learning_rate": 2.996164516167469e-06,
"loss": 0.4559,
"step": 152
},
{
"epoch": 0.3371900826446281,
"grad_norm": 11.692141310097577,
"learning_rate": 2.996088974805775e-06,
"loss": 0.63,
"step": 153
},
{
"epoch": 0.3393939393939394,
"grad_norm": 14.302015659156769,
"learning_rate": 2.9960126977434607e-06,
"loss": 0.4642,
"step": 154
},
{
"epoch": 0.3415977961432507,
"grad_norm": 10.601023905940114,
"learning_rate": 2.9959356850180356e-06,
"loss": 0.6745,
"step": 155
},
{
"epoch": 0.343801652892562,
"grad_norm": 7.373808999878665,
"learning_rate": 2.9958579366673703e-06,
"loss": 0.5854,
"step": 156
},
{
"epoch": 0.34600550964187327,
"grad_norm": 9.158314159683968,
"learning_rate": 2.995779452729698e-06,
"loss": 0.5532,
"step": 157
},
{
"epoch": 0.34820936639118455,
"grad_norm": 13.128561412436367,
"learning_rate": 2.995700233243613e-06,
"loss": 0.5181,
"step": 158
},
{
"epoch": 0.3504132231404959,
"grad_norm": 9.448628448642765,
"learning_rate": 2.995620278248071e-06,
"loss": 0.4891,
"step": 159
},
{
"epoch": 0.3526170798898072,
"grad_norm": 10.213114577672373,
"learning_rate": 2.995539587782391e-06,
"loss": 0.5077,
"step": 160
},
{
"epoch": 0.35482093663911846,
"grad_norm": 8.38970998579396,
"learning_rate": 2.9954581618862503e-06,
"loss": 0.405,
"step": 161
},
{
"epoch": 0.35702479338842974,
"grad_norm": 13.665397680766324,
"learning_rate": 2.995376000599692e-06,
"loss": 0.5125,
"step": 162
},
{
"epoch": 0.359228650137741,
"grad_norm": 14.179230821803024,
"learning_rate": 2.995293103963118e-06,
"loss": 0.4435,
"step": 163
},
{
"epoch": 0.36143250688705236,
"grad_norm": 8.756398485153353,
"learning_rate": 2.9952094720172932e-06,
"loss": 0.5425,
"step": 164
},
{
"epoch": 0.36363636363636365,
"grad_norm": 11.043239379621582,
"learning_rate": 2.995125104803343e-06,
"loss": 0.3934,
"step": 165
},
{
"epoch": 0.36584022038567493,
"grad_norm": 10.923159516630248,
"learning_rate": 2.9950400023627547e-06,
"loss": 0.505,
"step": 166
},
{
"epoch": 0.3680440771349862,
"grad_norm": 10.628828504738241,
"learning_rate": 2.994954164737378e-06,
"loss": 0.3617,
"step": 167
},
{
"epoch": 0.3702479338842975,
"grad_norm": 15.635865238422193,
"learning_rate": 2.994867591969423e-06,
"loss": 0.6068,
"step": 168
},
{
"epoch": 0.37245179063360884,
"grad_norm": 12.758012717712035,
"learning_rate": 2.994780284101462e-06,
"loss": 0.5216,
"step": 169
},
{
"epoch": 0.3746556473829201,
"grad_norm": 15.578582001690497,
"learning_rate": 2.994692241176429e-06,
"loss": 0.5622,
"step": 170
},
{
"epoch": 0.3768595041322314,
"grad_norm": 7.9750143389034855,
"learning_rate": 2.994603463237618e-06,
"loss": 0.3273,
"step": 171
},
{
"epoch": 0.3790633608815427,
"grad_norm": 12.572717399971113,
"learning_rate": 2.9945139503286868e-06,
"loss": 0.4176,
"step": 172
},
{
"epoch": 0.38126721763085397,
"grad_norm": 20.36436189716221,
"learning_rate": 2.9944237024936524e-06,
"loss": 0.4739,
"step": 173
},
{
"epoch": 0.3834710743801653,
"grad_norm": 10.297340479801056,
"learning_rate": 2.994332719776894e-06,
"loss": 0.4775,
"step": 174
},
{
"epoch": 0.3856749311294766,
"grad_norm": 10.284849583006528,
"learning_rate": 2.994241002223153e-06,
"loss": 0.4431,
"step": 175
},
{
"epoch": 0.3878787878787879,
"grad_norm": 10.928184893741808,
"learning_rate": 2.994148549877531e-06,
"loss": 0.3293,
"step": 176
},
{
"epoch": 0.39008264462809916,
"grad_norm": 13.263047600878323,
"learning_rate": 2.9940553627854913e-06,
"loss": 0.3778,
"step": 177
},
{
"epoch": 0.39228650137741045,
"grad_norm": 7.530956085027773,
"learning_rate": 2.993961440992859e-06,
"loss": 0.4591,
"step": 178
},
{
"epoch": 0.3944903581267218,
"grad_norm": 10.83288966187011,
"learning_rate": 2.993866784545819e-06,
"loss": 0.2411,
"step": 179
},
{
"epoch": 0.39669421487603307,
"grad_norm": 19.45648393646713,
"learning_rate": 2.99377139349092e-06,
"loss": 0.4308,
"step": 180
},
{
"epoch": 0.39889807162534435,
"grad_norm": 11.172989386477646,
"learning_rate": 2.9936752678750693e-06,
"loss": 0.5432,
"step": 181
},
{
"epoch": 0.40110192837465564,
"grad_norm": 8.31782661386756,
"learning_rate": 2.993578407745538e-06,
"loss": 0.3763,
"step": 182
},
{
"epoch": 0.4033057851239669,
"grad_norm": 11.442522440089068,
"learning_rate": 2.993480813149955e-06,
"loss": 0.517,
"step": 183
},
{
"epoch": 0.40550964187327826,
"grad_norm": 8.9587829278875,
"learning_rate": 2.993382484136314e-06,
"loss": 0.3016,
"step": 184
},
{
"epoch": 0.40771349862258954,
"grad_norm": 7.108332306844178,
"learning_rate": 2.9932834207529676e-06,
"loss": 0.2971,
"step": 185
},
{
"epoch": 0.4099173553719008,
"grad_norm": 10.329961838873576,
"learning_rate": 2.9931836230486303e-06,
"loss": 0.5498,
"step": 186
},
{
"epoch": 0.4121212121212121,
"grad_norm": 13.216403474783826,
"learning_rate": 2.9930830910723772e-06,
"loss": 0.2092,
"step": 187
},
{
"epoch": 0.4143250688705234,
"grad_norm": 8.671324083824599,
"learning_rate": 2.9929818248736447e-06,
"loss": 0.2558,
"step": 188
},
{
"epoch": 0.41652892561983473,
"grad_norm": 9.710171011043373,
"learning_rate": 2.992879824502232e-06,
"loss": 0.3006,
"step": 189
},
{
"epoch": 0.418732782369146,
"grad_norm": 21.813616228299125,
"learning_rate": 2.9927770900082955e-06,
"loss": 0.5329,
"step": 190
},
{
"epoch": 0.4209366391184573,
"grad_norm": 13.323515391705687,
"learning_rate": 2.992673621442356e-06,
"loss": 0.4125,
"step": 191
},
{
"epoch": 0.4231404958677686,
"grad_norm": 9.278125853902523,
"learning_rate": 2.992569418855294e-06,
"loss": 0.3332,
"step": 192
},
{
"epoch": 0.42534435261707987,
"grad_norm": 10.791350469404003,
"learning_rate": 2.9924644822983503e-06,
"loss": 0.5235,
"step": 193
},
{
"epoch": 0.4275482093663912,
"grad_norm": 8.898978752132685,
"learning_rate": 2.9923588118231285e-06,
"loss": 0.5324,
"step": 194
},
{
"epoch": 0.4297520661157025,
"grad_norm": 8.533915405225065,
"learning_rate": 2.992252407481591e-06,
"loss": 0.518,
"step": 195
},
{
"epoch": 0.4319559228650138,
"grad_norm": 10.992324651322523,
"learning_rate": 2.9921452693260635e-06,
"loss": 0.5302,
"step": 196
},
{
"epoch": 0.43415977961432506,
"grad_norm": 23.456111817713143,
"learning_rate": 2.9920373974092295e-06,
"loss": 0.4521,
"step": 197
},
{
"epoch": 0.43636363636363634,
"grad_norm": 14.786200878275693,
"learning_rate": 2.9919287917841353e-06,
"loss": 0.5255,
"step": 198
},
{
"epoch": 0.4385674931129477,
"grad_norm": 11.085386121303593,
"learning_rate": 2.9918194525041885e-06,
"loss": 0.2744,
"step": 199
},
{
"epoch": 0.44077134986225897,
"grad_norm": 8.906532048314519,
"learning_rate": 2.9917093796231553e-06,
"loss": 0.5756,
"step": 200
},
{
"epoch": 0.44297520661157025,
"grad_norm": 7.951838632934386,
"learning_rate": 2.9915985731951645e-06,
"loss": 0.5796,
"step": 201
},
{
"epoch": 0.44517906336088153,
"grad_norm": 17.58502984886795,
"learning_rate": 2.9914870332747063e-06,
"loss": 0.489,
"step": 202
},
{
"epoch": 0.4473829201101928,
"grad_norm": 15.295500415608018,
"learning_rate": 2.9913747599166284e-06,
"loss": 0.4597,
"step": 203
},
{
"epoch": 0.44958677685950416,
"grad_norm": 13.049299141244818,
"learning_rate": 2.9912617531761423e-06,
"loss": 0.4749,
"step": 204
},
{
"epoch": 0.45179063360881544,
"grad_norm": 9.931073149347117,
"learning_rate": 2.9911480131088188e-06,
"loss": 0.4997,
"step": 205
},
{
"epoch": 0.4539944903581267,
"grad_norm": 8.566494688187221,
"learning_rate": 2.991033539770589e-06,
"loss": 0.377,
"step": 206
},
{
"epoch": 0.456198347107438,
"grad_norm": 13.85397292070045,
"learning_rate": 2.990918333217746e-06,
"loss": 0.3859,
"step": 207
},
{
"epoch": 0.4584022038567493,
"grad_norm": 11.550854145013796,
"learning_rate": 2.990802393506942e-06,
"loss": 0.5885,
"step": 208
},
{
"epoch": 0.46060606060606063,
"grad_norm": 6.01047647343179,
"learning_rate": 2.99068572069519e-06,
"loss": 0.323,
"step": 209
},
{
"epoch": 0.4628099173553719,
"grad_norm": 8.224271030300471,
"learning_rate": 2.9905683148398643e-06,
"loss": 0.5296,
"step": 210
},
{
"epoch": 0.4650137741046832,
"grad_norm": 15.640561146666892,
"learning_rate": 2.9904501759986982e-06,
"loss": 0.507,
"step": 211
},
{
"epoch": 0.4672176308539945,
"grad_norm": 10.955257491240019,
"learning_rate": 2.9903313042297874e-06,
"loss": 0.3995,
"step": 212
},
{
"epoch": 0.46942148760330576,
"grad_norm": 7.323757447082715,
"learning_rate": 2.9902116995915872e-06,
"loss": 0.3823,
"step": 213
},
{
"epoch": 0.4716253443526171,
"grad_norm": 7.703416812989557,
"learning_rate": 2.9900913621429124e-06,
"loss": 0.2121,
"step": 214
},
{
"epoch": 0.4738292011019284,
"grad_norm": 7.11349178483258,
"learning_rate": 2.9899702919429383e-06,
"loss": 0.4344,
"step": 215
},
{
"epoch": 0.47603305785123967,
"grad_norm": 14.027780027978222,
"learning_rate": 2.989848489051203e-06,
"loss": 0.3547,
"step": 216
},
{
"epoch": 0.47823691460055096,
"grad_norm": 6.760823267329529,
"learning_rate": 2.9897259535276007e-06,
"loss": 0.2463,
"step": 217
},
{
"epoch": 0.48044077134986224,
"grad_norm": 7.913445075265595,
"learning_rate": 2.9896026854323896e-06,
"loss": 0.4024,
"step": 218
},
{
"epoch": 0.4826446280991736,
"grad_norm": 11.412613438060973,
"learning_rate": 2.9894786848261863e-06,
"loss": 0.3003,
"step": 219
},
{
"epoch": 0.48484848484848486,
"grad_norm": 9.62241009305966,
"learning_rate": 2.9893539517699683e-06,
"loss": 0.3287,
"step": 220
},
{
"epoch": 0.48705234159779615,
"grad_norm": 7.753152354810825,
"learning_rate": 2.989228486325072e-06,
"loss": 0.3269,
"step": 221
},
{
"epoch": 0.48925619834710743,
"grad_norm": 7.295729043500175,
"learning_rate": 2.989102288553196e-06,
"loss": 0.3438,
"step": 222
},
{
"epoch": 0.4914600550964187,
"grad_norm": 9.153774556100242,
"learning_rate": 2.9889753585163977e-06,
"loss": 0.3999,
"step": 223
},
{
"epoch": 0.49366391184573005,
"grad_norm": 13.438636520374068,
"learning_rate": 2.988847696277095e-06,
"loss": 0.2573,
"step": 224
},
{
"epoch": 0.49586776859504134,
"grad_norm": 10.228395361152069,
"learning_rate": 2.9887193018980653e-06,
"loss": 0.252,
"step": 225
},
{
"epoch": 0.4980716253443526,
"grad_norm": 10.244058650244439,
"learning_rate": 2.9885901754424465e-06,
"loss": 0.4004,
"step": 226
},
{
"epoch": 0.5002754820936639,
"grad_norm": 6.461802138295602,
"learning_rate": 2.9884603169737363e-06,
"loss": 0.2336,
"step": 227
},
{
"epoch": 0.5024793388429752,
"grad_norm": 11.410239126771371,
"learning_rate": 2.988329726555793e-06,
"loss": 0.3149,
"step": 228
},
{
"epoch": 0.5046831955922865,
"grad_norm": 10.968492435160934,
"learning_rate": 2.988198404252834e-06,
"loss": 0.3722,
"step": 229
},
{
"epoch": 0.5068870523415978,
"grad_norm": 12.497668168814954,
"learning_rate": 2.988066350129437e-06,
"loss": 0.4878,
"step": 230
},
{
"epoch": 0.509090909090909,
"grad_norm": 11.630857036501677,
"learning_rate": 2.9879335642505396e-06,
"loss": 0.2971,
"step": 231
},
{
"epoch": 0.5112947658402204,
"grad_norm": 6.441981218733046,
"learning_rate": 2.9878000466814392e-06,
"loss": 0.3021,
"step": 232
},
{
"epoch": 0.5134986225895317,
"grad_norm": 8.17525488032321,
"learning_rate": 2.987665797487793e-06,
"loss": 0.2385,
"step": 233
},
{
"epoch": 0.515702479338843,
"grad_norm": 25.7467268298053,
"learning_rate": 2.9875308167356174e-06,
"loss": 0.2288,
"step": 234
},
{
"epoch": 0.5179063360881543,
"grad_norm": 10.459874704987152,
"learning_rate": 2.9873951044912893e-06,
"loss": 0.2691,
"step": 235
},
{
"epoch": 0.5201101928374655,
"grad_norm": 7.962423185829474,
"learning_rate": 2.987258660821546e-06,
"loss": 0.3499,
"step": 236
},
{
"epoch": 0.5223140495867769,
"grad_norm": 8.351626000174596,
"learning_rate": 2.9871214857934823e-06,
"loss": 0.2256,
"step": 237
},
{
"epoch": 0.5245179063360882,
"grad_norm": 8.76914626967428,
"learning_rate": 2.9869835794745546e-06,
"loss": 0.3521,
"step": 238
},
{
"epoch": 0.5267217630853994,
"grad_norm": 12.774308842259753,
"learning_rate": 2.986844941932578e-06,
"loss": 0.263,
"step": 239
},
{
"epoch": 0.5289256198347108,
"grad_norm": 8.662606106336478,
"learning_rate": 2.9867055732357275e-06,
"loss": 0.3717,
"step": 240
},
{
"epoch": 0.531129476584022,
"grad_norm": 7.44373651269041,
"learning_rate": 2.986565473452538e-06,
"loss": 0.3603,
"step": 241
},
{
"epoch": 0.5333333333333333,
"grad_norm": 16.113652444436543,
"learning_rate": 2.986424642651902e-06,
"loss": 0.4322,
"step": 242
},
{
"epoch": 0.5355371900826447,
"grad_norm": 7.235187047429284,
"learning_rate": 2.986283080903075e-06,
"loss": 0.2718,
"step": 243
},
{
"epoch": 0.5377410468319559,
"grad_norm": 8.981158375257591,
"learning_rate": 2.986140788275668e-06,
"loss": 0.2909,
"step": 244
},
{
"epoch": 0.5399449035812672,
"grad_norm": 7.728183713731981,
"learning_rate": 2.9859977648396547e-06,
"loss": 0.2688,
"step": 245
},
{
"epoch": 0.5421487603305785,
"grad_norm": 11.385723529297572,
"learning_rate": 2.985854010665366e-06,
"loss": 0.2951,
"step": 246
},
{
"epoch": 0.5443526170798898,
"grad_norm": 22.421774240399724,
"learning_rate": 2.985709525823493e-06,
"loss": 0.2722,
"step": 247
},
{
"epoch": 0.5465564738292011,
"grad_norm": 26.25406750616081,
"learning_rate": 2.9855643103850863e-06,
"loss": 0.2057,
"step": 248
},
{
"epoch": 0.5487603305785124,
"grad_norm": 14.419670718604909,
"learning_rate": 2.985418364421555e-06,
"loss": 0.1859,
"step": 249
},
{
"epoch": 0.5509641873278237,
"grad_norm": 7.18194575744084,
"learning_rate": 2.9852716880046687e-06,
"loss": 0.1823,
"step": 250
},
{
"epoch": 0.5531680440771349,
"grad_norm": 11.041149585270512,
"learning_rate": 2.9851242812065544e-06,
"loss": 0.4397,
"step": 251
},
{
"epoch": 0.5553719008264463,
"grad_norm": 9.800903211627961,
"learning_rate": 2.9849761440996997e-06,
"loss": 0.3035,
"step": 252
},
{
"epoch": 0.5575757575757576,
"grad_norm": 5.42248846810359,
"learning_rate": 2.9848272767569515e-06,
"loss": 0.4126,
"step": 253
},
{
"epoch": 0.5597796143250688,
"grad_norm": 12.064336260743763,
"learning_rate": 2.9846776792515146e-06,
"loss": 0.4134,
"step": 254
},
{
"epoch": 0.5619834710743802,
"grad_norm": 9.069153337986716,
"learning_rate": 2.9845273516569534e-06,
"loss": 0.3347,
"step": 255
},
{
"epoch": 0.5641873278236914,
"grad_norm": 9.708719457668808,
"learning_rate": 2.9843762940471915e-06,
"loss": 0.1482,
"step": 256
},
{
"epoch": 0.5663911845730027,
"grad_norm": 10.658478324774729,
"learning_rate": 2.984224506496512e-06,
"loss": 0.4586,
"step": 257
},
{
"epoch": 0.5685950413223141,
"grad_norm": 7.67634362110084,
"learning_rate": 2.984071989079555e-06,
"loss": 0.3101,
"step": 258
},
{
"epoch": 0.5707988980716253,
"grad_norm": 9.863415592665492,
"learning_rate": 2.983918741871322e-06,
"loss": 0.283,
"step": 259
},
{
"epoch": 0.5730027548209367,
"grad_norm": 15.328770243567073,
"learning_rate": 2.983764764947172e-06,
"loss": 0.3214,
"step": 260
},
{
"epoch": 0.5752066115702479,
"grad_norm": 5.602530886005082,
"learning_rate": 2.983610058382822e-06,
"loss": 0.1458,
"step": 261
},
{
"epoch": 0.5774104683195592,
"grad_norm": 9.188539359413474,
"learning_rate": 2.9834546222543503e-06,
"loss": 0.2688,
"step": 262
},
{
"epoch": 0.5796143250688706,
"grad_norm": 8.913852840097164,
"learning_rate": 2.9832984566381913e-06,
"loss": 0.2904,
"step": 263
},
{
"epoch": 0.5818181818181818,
"grad_norm": 8.408585828146748,
"learning_rate": 2.98314156161114e-06,
"loss": 0.3473,
"step": 264
},
{
"epoch": 0.5840220385674931,
"grad_norm": 7.011982039251787,
"learning_rate": 2.9829839372503496e-06,
"loss": 0.2214,
"step": 265
},
{
"epoch": 0.5862258953168044,
"grad_norm": 12.208817142819923,
"learning_rate": 2.982825583633331e-06,
"loss": 0.228,
"step": 266
},
{
"epoch": 0.5884297520661157,
"grad_norm": 22.300240567856033,
"learning_rate": 2.982666500837955e-06,
"loss": 0.3841,
"step": 267
},
{
"epoch": 0.590633608815427,
"grad_norm": 10.204434125277357,
"learning_rate": 2.9825066889424507e-06,
"loss": 0.4413,
"step": 268
},
{
"epoch": 0.5928374655647383,
"grad_norm": 10.394116435568966,
"learning_rate": 2.9823461480254046e-06,
"loss": 0.2164,
"step": 269
},
{
"epoch": 0.5950413223140496,
"grad_norm": 24.930405170814677,
"learning_rate": 2.982184878165763e-06,
"loss": 0.404,
"step": 270
},
{
"epoch": 0.5972451790633608,
"grad_norm": 16.529549906687567,
"learning_rate": 2.982022879442831e-06,
"loss": 0.1971,
"step": 271
},
{
"epoch": 0.5994490358126722,
"grad_norm": 6.802656347424857,
"learning_rate": 2.98186015193627e-06,
"loss": 0.2763,
"step": 272
},
{
"epoch": 0.6016528925619835,
"grad_norm": 9.967577482684383,
"learning_rate": 2.9816966957261022e-06,
"loss": 0.505,
"step": 273
},
{
"epoch": 0.6038567493112947,
"grad_norm": 5.5700188781618944,
"learning_rate": 2.981532510892707e-06,
"loss": 0.3295,
"step": 274
},
{
"epoch": 0.6060606060606061,
"grad_norm": 13.138835779765238,
"learning_rate": 2.981367597516821e-06,
"loss": 0.2216,
"step": 275
},
{
"epoch": 0.6082644628099173,
"grad_norm": 8.78165681806648,
"learning_rate": 2.9812019556795414e-06,
"loss": 0.4269,
"step": 276
},
{
"epoch": 0.6104683195592286,
"grad_norm": 7.519045608568057,
"learning_rate": 2.981035585462322e-06,
"loss": 0.3028,
"step": 277
},
{
"epoch": 0.61267217630854,
"grad_norm": 12.757117931598634,
"learning_rate": 2.9808684869469756e-06,
"loss": 0.2731,
"step": 278
},
{
"epoch": 0.6148760330578512,
"grad_norm": 14.32345009190218,
"learning_rate": 2.9807006602156723e-06,
"loss": 0.439,
"step": 279
},
{
"epoch": 0.6170798898071626,
"grad_norm": 10.78268372690156,
"learning_rate": 2.9805321053509414e-06,
"loss": 0.3267,
"step": 280
},
{
"epoch": 0.6192837465564738,
"grad_norm": 20.968509670896452,
"learning_rate": 2.9803628224356688e-06,
"loss": 0.1478,
"step": 281
},
{
"epoch": 0.6214876033057851,
"grad_norm": 13.939089444718862,
"learning_rate": 2.9801928115531e-06,
"loss": 0.2811,
"step": 282
},
{
"epoch": 0.6236914600550965,
"grad_norm": 19.46588188306463,
"learning_rate": 2.9800220727868367e-06,
"loss": 0.2355,
"step": 283
},
{
"epoch": 0.6258953168044077,
"grad_norm": 23.31435507775797,
"learning_rate": 2.9798506062208408e-06,
"loss": 0.3445,
"step": 284
},
{
"epoch": 0.628099173553719,
"grad_norm": 7.113401931682151,
"learning_rate": 2.9796784119394296e-06,
"loss": 0.1838,
"step": 285
},
{
"epoch": 0.6303030303030303,
"grad_norm": 9.83090974685671,
"learning_rate": 2.979505490027281e-06,
"loss": 0.3413,
"step": 286
},
{
"epoch": 0.6325068870523416,
"grad_norm": 11.816944932544452,
"learning_rate": 2.979331840569428e-06,
"loss": 0.2624,
"step": 287
},
{
"epoch": 0.6347107438016529,
"grad_norm": 5.505913040295101,
"learning_rate": 2.979157463651263e-06,
"loss": 0.2669,
"step": 288
},
{
"epoch": 0.6369146005509642,
"grad_norm": 7.9610097101417745,
"learning_rate": 2.978982359358536e-06,
"loss": 0.2418,
"step": 289
},
{
"epoch": 0.6391184573002755,
"grad_norm": 14.054084856949025,
"learning_rate": 2.9788065277773537e-06,
"loss": 0.3091,
"step": 290
},
{
"epoch": 0.6413223140495867,
"grad_norm": 6.994997659089347,
"learning_rate": 2.978629968994182e-06,
"loss": 0.3571,
"step": 291
},
{
"epoch": 0.6435261707988981,
"grad_norm": 4.900411467720552,
"learning_rate": 2.978452683095843e-06,
"loss": 0.2389,
"step": 292
},
{
"epoch": 0.6457300275482094,
"grad_norm": 5.986010486219538,
"learning_rate": 2.978274670169517e-06,
"loss": 0.4119,
"step": 293
},
{
"epoch": 0.6479338842975206,
"grad_norm": 9.99633467057808,
"learning_rate": 2.978095930302741e-06,
"loss": 0.1945,
"step": 294
},
{
"epoch": 0.650137741046832,
"grad_norm": 8.88618573335772,
"learning_rate": 2.9779164635834117e-06,
"loss": 0.3021,
"step": 295
},
{
"epoch": 0.6523415977961432,
"grad_norm": 6.1713366384622885,
"learning_rate": 2.9777362700997813e-06,
"loss": 0.3536,
"step": 296
},
{
"epoch": 0.6545454545454545,
"grad_norm": 6.269429712344906,
"learning_rate": 2.9775553499404588e-06,
"loss": 0.2532,
"step": 297
},
{
"epoch": 0.6567493112947659,
"grad_norm": 8.01546204853658,
"learning_rate": 2.9773737031944123e-06,
"loss": 0.2054,
"step": 298
},
{
"epoch": 0.6589531680440771,
"grad_norm": 5.834139768693889,
"learning_rate": 2.9771913299509667e-06,
"loss": 0.3205,
"step": 299
},
{
"epoch": 0.6611570247933884,
"grad_norm": 5.20599030336784,
"learning_rate": 2.977008230299803e-06,
"loss": 0.233,
"step": 300
},
{
"epoch": 0.6633608815426997,
"grad_norm": 7.747852317820463,
"learning_rate": 2.9768244043309612e-06,
"loss": 0.1853,
"step": 301
},
{
"epoch": 0.665564738292011,
"grad_norm": 7.280420219391244,
"learning_rate": 2.9766398521348376e-06,
"loss": 0.2404,
"step": 302
},
{
"epoch": 0.6677685950413224,
"grad_norm": 8.649220755714195,
"learning_rate": 2.9764545738021847e-06,
"loss": 0.2864,
"step": 303
},
{
"epoch": 0.6699724517906336,
"grad_norm": 7.124734041126083,
"learning_rate": 2.976268569424114e-06,
"loss": 0.5867,
"step": 304
},
{
"epoch": 0.6721763085399449,
"grad_norm": 8.708663746697807,
"learning_rate": 2.9760818390920924e-06,
"loss": 0.2924,
"step": 305
},
{
"epoch": 0.6743801652892562,
"grad_norm": 13.18814959982244,
"learning_rate": 2.9758943828979446e-06,
"loss": 0.3782,
"step": 306
},
{
"epoch": 0.6765840220385675,
"grad_norm": 12.243598876193756,
"learning_rate": 2.975706200933852e-06,
"loss": 0.2185,
"step": 307
},
{
"epoch": 0.6787878787878788,
"grad_norm": 5.633700961339941,
"learning_rate": 2.975517293292353e-06,
"loss": 0.3752,
"step": 308
},
{
"epoch": 0.6809917355371901,
"grad_norm": 10.334114912832058,
"learning_rate": 2.9753276600663423e-06,
"loss": 0.3026,
"step": 309
},
{
"epoch": 0.6831955922865014,
"grad_norm": 13.08744426612834,
"learning_rate": 2.9751373013490727e-06,
"loss": 0.3382,
"step": 310
},
{
"epoch": 0.6853994490358126,
"grad_norm": 9.504458358236962,
"learning_rate": 2.9749462172341524e-06,
"loss": 0.5009,
"step": 311
},
{
"epoch": 0.687603305785124,
"grad_norm": 6.3562762612917245,
"learning_rate": 2.9747544078155472e-06,
"loss": 0.3382,
"step": 312
},
{
"epoch": 0.6898071625344353,
"grad_norm": 12.917546406900222,
"learning_rate": 2.974561873187579e-06,
"loss": 0.2846,
"step": 313
},
{
"epoch": 0.6920110192837465,
"grad_norm": 6.944988521051594,
"learning_rate": 2.9743686134449267e-06,
"loss": 0.3612,
"step": 314
},
{
"epoch": 0.6942148760330579,
"grad_norm": 7.791757617860102,
"learning_rate": 2.974174628682626e-06,
"loss": 0.4309,
"step": 315
},
{
"epoch": 0.6964187327823691,
"grad_norm": 6.838689832113007,
"learning_rate": 2.973979918996068e-06,
"loss": 0.3473,
"step": 316
},
{
"epoch": 0.6986225895316804,
"grad_norm": 6.638427266811496,
"learning_rate": 2.973784484481001e-06,
"loss": 0.3482,
"step": 317
},
{
"epoch": 0.7008264462809918,
"grad_norm": 15.019684641662097,
"learning_rate": 2.9735883252335305e-06,
"loss": 0.3447,
"step": 318
},
{
"epoch": 0.703030303030303,
"grad_norm": 7.657611207942621,
"learning_rate": 2.9733914413501172e-06,
"loss": 0.1998,
"step": 319
},
{
"epoch": 0.7052341597796143,
"grad_norm": 12.679008507955187,
"learning_rate": 2.973193832927579e-06,
"loss": 0.2036,
"step": 320
},
{
"epoch": 0.7074380165289256,
"grad_norm": 7.001507137978587,
"learning_rate": 2.9729955000630886e-06,
"loss": 0.4147,
"step": 321
},
{
"epoch": 0.7096418732782369,
"grad_norm": 7.424077459455387,
"learning_rate": 2.972796442854178e-06,
"loss": 0.3658,
"step": 322
},
{
"epoch": 0.7118457300275483,
"grad_norm": 9.447237658007161,
"learning_rate": 2.9725966613987312e-06,
"loss": 0.3984,
"step": 323
},
{
"epoch": 0.7140495867768595,
"grad_norm": 21.361023805823653,
"learning_rate": 2.972396155794992e-06,
"loss": 0.371,
"step": 324
},
{
"epoch": 0.7162534435261708,
"grad_norm": 8.834296167166244,
"learning_rate": 2.9721949261415592e-06,
"loss": 0.3004,
"step": 325
},
{
"epoch": 0.718457300275482,
"grad_norm": 6.742790410332544,
"learning_rate": 2.971992972537386e-06,
"loss": 0.1565,
"step": 326
},
{
"epoch": 0.7206611570247934,
"grad_norm": 11.707104581221218,
"learning_rate": 2.9717902950817833e-06,
"loss": 0.2299,
"step": 327
},
{
"epoch": 0.7228650137741047,
"grad_norm": 5.476783766315915,
"learning_rate": 2.9715868938744182e-06,
"loss": 0.2493,
"step": 328
},
{
"epoch": 0.725068870523416,
"grad_norm": 11.160688397190603,
"learning_rate": 2.971382769015313e-06,
"loss": 0.3175,
"step": 329
},
{
"epoch": 0.7272727272727273,
"grad_norm": 7.043330544504408,
"learning_rate": 2.971177920604846e-06,
"loss": 0.1722,
"step": 330
},
{
"epoch": 0.7294765840220385,
"grad_norm": 5.73739903037248,
"learning_rate": 2.97097234874375e-06,
"loss": 0.2176,
"step": 331
},
{
"epoch": 0.7316804407713499,
"grad_norm": 6.277934895839506,
"learning_rate": 2.9707660535331166e-06,
"loss": 0.2065,
"step": 332
},
{
"epoch": 0.7338842975206612,
"grad_norm": 20.30577221870282,
"learning_rate": 2.9705590350743904e-06,
"loss": 0.1991,
"step": 333
},
{
"epoch": 0.7360881542699724,
"grad_norm": 7.695272096611418,
"learning_rate": 2.970351293469372e-06,
"loss": 0.2131,
"step": 334
},
{
"epoch": 0.7382920110192838,
"grad_norm": 4.916567786531378,
"learning_rate": 2.9701428288202194e-06,
"loss": 0.3827,
"step": 335
},
{
"epoch": 0.740495867768595,
"grad_norm": 7.019231101260278,
"learning_rate": 2.9699336412294444e-06,
"loss": 0.2738,
"step": 336
},
{
"epoch": 0.7426997245179063,
"grad_norm": 7.975195630484479,
"learning_rate": 2.9697237307999146e-06,
"loss": 0.2243,
"step": 337
},
{
"epoch": 0.7449035812672177,
"grad_norm": 7.561687409407015,
"learning_rate": 2.9695130976348534e-06,
"loss": 0.2396,
"step": 338
},
{
"epoch": 0.7471074380165289,
"grad_norm": 7.638419171445274,
"learning_rate": 2.9693017418378396e-06,
"loss": 0.276,
"step": 339
},
{
"epoch": 0.7493112947658402,
"grad_norm": 12.616136018780299,
"learning_rate": 2.9690896635128073e-06,
"loss": 0.3787,
"step": 340
},
{
"epoch": 0.7515151515151515,
"grad_norm": 6.550803777717444,
"learning_rate": 2.968876862764046e-06,
"loss": 0.2226,
"step": 341
},
{
"epoch": 0.7537190082644628,
"grad_norm": 9.06584766964822,
"learning_rate": 2.9686633396962003e-06,
"loss": 0.369,
"step": 342
},
{
"epoch": 0.7559228650137741,
"grad_norm": 5.386215213674309,
"learning_rate": 2.9684490944142697e-06,
"loss": 0.1736,
"step": 343
},
{
"epoch": 0.7581267217630854,
"grad_norm": 15.906647139869929,
"learning_rate": 2.9682341270236085e-06,
"loss": 0.2781,
"step": 344
},
{
"epoch": 0.7603305785123967,
"grad_norm": 12.1772384094106,
"learning_rate": 2.9680184376299283e-06,
"loss": 0.1544,
"step": 345
},
{
"epoch": 0.7625344352617079,
"grad_norm": 17.88186148600346,
"learning_rate": 2.967802026339293e-06,
"loss": 0.2988,
"step": 346
},
{
"epoch": 0.7647382920110193,
"grad_norm": 21.268773226628287,
"learning_rate": 2.9675848932581235e-06,
"loss": 0.2821,
"step": 347
},
{
"epoch": 0.7669421487603306,
"grad_norm": 11.820036760533492,
"learning_rate": 2.9673670384931945e-06,
"loss": 0.3498,
"step": 348
},
{
"epoch": 0.7691460055096419,
"grad_norm": 7.530415574189705,
"learning_rate": 2.967148462151635e-06,
"loss": 0.1424,
"step": 349
},
{
"epoch": 0.7713498622589532,
"grad_norm": 6.807351681100111,
"learning_rate": 2.9669291643409314e-06,
"loss": 0.2128,
"step": 350
},
{
"epoch": 0.7735537190082644,
"grad_norm": 7.234299741762421,
"learning_rate": 2.9667091451689224e-06,
"loss": 0.2932,
"step": 351
},
{
"epoch": 0.7757575757575758,
"grad_norm": 6.532060796534973,
"learning_rate": 2.9664884047438023e-06,
"loss": 0.2665,
"step": 352
},
{
"epoch": 0.7779614325068871,
"grad_norm": 10.34906346210999,
"learning_rate": 2.96626694317412e-06,
"loss": 0.3598,
"step": 353
},
{
"epoch": 0.7801652892561983,
"grad_norm": 8.210171975505114,
"learning_rate": 2.966044760568779e-06,
"loss": 0.4052,
"step": 354
},
{
"epoch": 0.7823691460055097,
"grad_norm": 4.541296912666131,
"learning_rate": 2.9658218570370374e-06,
"loss": 0.2954,
"step": 355
},
{
"epoch": 0.7845730027548209,
"grad_norm": 5.703338007840205,
"learning_rate": 2.965598232688508e-06,
"loss": 0.3667,
"step": 356
},
{
"epoch": 0.7867768595041322,
"grad_norm": 7.3908500819159135,
"learning_rate": 2.965373887633158e-06,
"loss": 0.0975,
"step": 357
},
{
"epoch": 0.7889807162534436,
"grad_norm": 6.532653722744144,
"learning_rate": 2.9651488219813086e-06,
"loss": 0.2558,
"step": 358
},
{
"epoch": 0.7911845730027548,
"grad_norm": 5.276369110257269,
"learning_rate": 2.964923035843636e-06,
"loss": 0.2056,
"step": 359
},
{
"epoch": 0.7933884297520661,
"grad_norm": 7.121048051130931,
"learning_rate": 2.96469652933117e-06,
"loss": 0.1641,
"step": 360
},
{
"epoch": 0.7955922865013774,
"grad_norm": 11.053322087508501,
"learning_rate": 2.9644693025552957e-06,
"loss": 0.4164,
"step": 361
},
{
"epoch": 0.7977961432506887,
"grad_norm": 6.503259157403907,
"learning_rate": 2.964241355627751e-06,
"loss": 0.2031,
"step": 362
},
{
"epoch": 0.8,
"grad_norm": 7.680655263285343,
"learning_rate": 2.9640126886606287e-06,
"loss": 0.4177,
"step": 363
},
{
"epoch": 0.8022038567493113,
"grad_norm": 9.597872477527922,
"learning_rate": 2.9637833017663757e-06,
"loss": 0.2231,
"step": 364
},
{
"epoch": 0.8044077134986226,
"grad_norm": 5.447713777602937,
"learning_rate": 2.9635531950577927e-06,
"loss": 0.2174,
"step": 365
},
{
"epoch": 0.8066115702479338,
"grad_norm": 7.189031762492828,
"learning_rate": 2.963322368648035e-06,
"loss": 0.2563,
"step": 366
},
{
"epoch": 0.8088154269972452,
"grad_norm": 7.898167623571763,
"learning_rate": 2.963090822650611e-06,
"loss": 0.3355,
"step": 367
},
{
"epoch": 0.8110192837465565,
"grad_norm": 12.577262656676943,
"learning_rate": 2.9628585571793833e-06,
"loss": 0.2235,
"step": 368
},
{
"epoch": 0.8132231404958677,
"grad_norm": 10.849928883515854,
"learning_rate": 2.962625572348568e-06,
"loss": 0.3533,
"step": 369
},
{
"epoch": 0.8154269972451791,
"grad_norm": 7.2964754371110665,
"learning_rate": 2.9623918682727352e-06,
"loss": 0.2871,
"step": 370
},
{
"epoch": 0.8176308539944903,
"grad_norm": 20.660728227860485,
"learning_rate": 2.9621574450668096e-06,
"loss": 0.2393,
"step": 371
},
{
"epoch": 0.8198347107438017,
"grad_norm": 7.220865047189661,
"learning_rate": 2.9619223028460675e-06,
"loss": 0.1926,
"step": 372
},
{
"epoch": 0.822038567493113,
"grad_norm": 4.852195073645113,
"learning_rate": 2.961686441726141e-06,
"loss": 0.2192,
"step": 373
},
{
"epoch": 0.8242424242424242,
"grad_norm": 9.027190273331941,
"learning_rate": 2.9614498618230133e-06,
"loss": 0.1366,
"step": 374
},
{
"epoch": 0.8264462809917356,
"grad_norm": 6.456919976931707,
"learning_rate": 2.961212563253023e-06,
"loss": 0.197,
"step": 375
},
{
"epoch": 0.8286501377410468,
"grad_norm": 5.401198387054753,
"learning_rate": 2.9609745461328625e-06,
"loss": 0.3046,
"step": 376
},
{
"epoch": 0.8308539944903581,
"grad_norm": 7.216796018509394,
"learning_rate": 2.9607358105795746e-06,
"loss": 0.206,
"step": 377
},
{
"epoch": 0.8330578512396695,
"grad_norm": 6.594512431937312,
"learning_rate": 2.9604963567105588e-06,
"loss": 0.1903,
"step": 378
},
{
"epoch": 0.8352617079889807,
"grad_norm": 7.862389931194225,
"learning_rate": 2.960256184643566e-06,
"loss": 0.3653,
"step": 379
},
{
"epoch": 0.837465564738292,
"grad_norm": 7.885741556216934,
"learning_rate": 2.9600152944967004e-06,
"loss": 0.2993,
"step": 380
},
{
"epoch": 0.8396694214876033,
"grad_norm": 8.760183620857365,
"learning_rate": 2.9597736863884197e-06,
"loss": 0.4891,
"step": 381
},
{
"epoch": 0.8418732782369146,
"grad_norm": 5.626515592689077,
"learning_rate": 2.9595313604375345e-06,
"loss": 0.1762,
"step": 382
},
{
"epoch": 0.8440771349862259,
"grad_norm": 14.700970431401748,
"learning_rate": 2.9592883167632086e-06,
"loss": 0.4149,
"step": 383
},
{
"epoch": 0.8462809917355372,
"grad_norm": 31.479222256604665,
"learning_rate": 2.959044555484958e-06,
"loss": 0.4708,
"step": 384
},
{
"epoch": 0.8484848484848485,
"grad_norm": 15.30296170661704,
"learning_rate": 2.9588000767226527e-06,
"loss": 0.4,
"step": 385
},
{
"epoch": 0.8506887052341597,
"grad_norm": 6.22490236971322,
"learning_rate": 2.958554880596515e-06,
"loss": 0.175,
"step": 386
},
{
"epoch": 0.8528925619834711,
"grad_norm": 17.884639044262485,
"learning_rate": 2.9583089672271193e-06,
"loss": 0.1644,
"step": 387
},
{
"epoch": 0.8550964187327824,
"grad_norm": 12.60321691171589,
"learning_rate": 2.9580623367353934e-06,
"loss": 0.3522,
"step": 388
},
{
"epoch": 0.8573002754820936,
"grad_norm": 8.03246045529426,
"learning_rate": 2.9578149892426185e-06,
"loss": 0.2052,
"step": 389
},
{
"epoch": 0.859504132231405,
"grad_norm": 10.469045062705694,
"learning_rate": 2.9575669248704265e-06,
"loss": 0.269,
"step": 390
},
{
"epoch": 0.8617079889807162,
"grad_norm": 5.196595120525184,
"learning_rate": 2.957318143740803e-06,
"loss": 0.3053,
"step": 391
},
{
"epoch": 0.8639118457300275,
"grad_norm": 6.554368800886443,
"learning_rate": 2.957068645976087e-06,
"loss": 0.2728,
"step": 392
},
{
"epoch": 0.8661157024793389,
"grad_norm": 9.262942687282687,
"learning_rate": 2.956818431698968e-06,
"loss": 0.1268,
"step": 393
},
{
"epoch": 0.8683195592286501,
"grad_norm": 4.271103765701704,
"learning_rate": 2.9565675010324883e-06,
"loss": 0.1553,
"step": 394
},
{
"epoch": 0.8705234159779615,
"grad_norm": 8.88374891264536,
"learning_rate": 2.956315854100043e-06,
"loss": 0.191,
"step": 395
},
{
"epoch": 0.8727272727272727,
"grad_norm": 6.6498851365809815,
"learning_rate": 2.9560634910253803e-06,
"loss": 0.2245,
"step": 396
},
{
"epoch": 0.874931129476584,
"grad_norm": 20.390664152678355,
"learning_rate": 2.955810411932599e-06,
"loss": 0.2601,
"step": 397
},
{
"epoch": 0.8771349862258954,
"grad_norm": 10.13737807765569,
"learning_rate": 2.9555566169461497e-06,
"loss": 0.3495,
"step": 398
},
{
"epoch": 0.8793388429752066,
"grad_norm": 11.159270329579886,
"learning_rate": 2.955302106190837e-06,
"loss": 0.2861,
"step": 399
},
{
"epoch": 0.8815426997245179,
"grad_norm": 22.71705363617449,
"learning_rate": 2.955046879791816e-06,
"loss": 0.2585,
"step": 400
},
{
"epoch": 0.8837465564738292,
"grad_norm": 34.11862174617697,
"learning_rate": 2.9547909378745942e-06,
"loss": 0.2624,
"step": 401
},
{
"epoch": 0.8859504132231405,
"grad_norm": 11.799082659394301,
"learning_rate": 2.9545342805650304e-06,
"loss": 0.4528,
"step": 402
},
{
"epoch": 0.8881542699724518,
"grad_norm": 5.250052295042655,
"learning_rate": 2.954276907989336e-06,
"loss": 0.2705,
"step": 403
},
{
"epoch": 0.8903581267217631,
"grad_norm": 5.906133153801479,
"learning_rate": 2.9540188202740737e-06,
"loss": 0.291,
"step": 404
},
{
"epoch": 0.8925619834710744,
"grad_norm": 8.561658386525929,
"learning_rate": 2.953760017546158e-06,
"loss": 0.3078,
"step": 405
},
{
"epoch": 0.8947658402203856,
"grad_norm": 12.819512771737724,
"learning_rate": 2.953500499932855e-06,
"loss": 0.3115,
"step": 406
},
{
"epoch": 0.896969696969697,
"grad_norm": 7.709667951718197,
"learning_rate": 2.9532402675617823e-06,
"loss": 0.2463,
"step": 407
},
{
"epoch": 0.8991735537190083,
"grad_norm": 7.580380677651997,
"learning_rate": 2.9529793205609085e-06,
"loss": 0.3511,
"step": 408
},
{
"epoch": 0.9013774104683195,
"grad_norm": 6.15114721735365,
"learning_rate": 2.952717659058555e-06,
"loss": 0.2768,
"step": 409
},
{
"epoch": 0.9035812672176309,
"grad_norm": 6.10589706130033,
"learning_rate": 2.9524552831833926e-06,
"loss": 0.2892,
"step": 410
},
{
"epoch": 0.9057851239669421,
"grad_norm": 13.28238874201893,
"learning_rate": 2.9521921930644446e-06,
"loss": 0.1546,
"step": 411
},
{
"epoch": 0.9079889807162534,
"grad_norm": 10.451785736319954,
"learning_rate": 2.951928388831086e-06,
"loss": 0.2112,
"step": 412
},
{
"epoch": 0.9101928374655648,
"grad_norm": 8.054370887201316,
"learning_rate": 2.9516638706130427e-06,
"loss": 0.2637,
"step": 413
},
{
"epoch": 0.912396694214876,
"grad_norm": 8.36403593102867,
"learning_rate": 2.95139863854039e-06,
"loss": 0.3171,
"step": 414
},
{
"epoch": 0.9146005509641874,
"grad_norm": 11.138721402799987,
"learning_rate": 2.951132692743556e-06,
"loss": 0.2595,
"step": 415
},
{
"epoch": 0.9168044077134986,
"grad_norm": 10.065171933068807,
"learning_rate": 2.9508660333533202e-06,
"loss": 0.2379,
"step": 416
},
{
"epoch": 0.9190082644628099,
"grad_norm": 9.372860713885858,
"learning_rate": 2.950598660500811e-06,
"loss": 0.4288,
"step": 417
},
{
"epoch": 0.9212121212121213,
"grad_norm": 6.0014294996569255,
"learning_rate": 2.9503305743175096e-06,
"loss": 0.3175,
"step": 418
},
{
"epoch": 0.9234159779614325,
"grad_norm": 6.0687105546706785,
"learning_rate": 2.950061774935247e-06,
"loss": 0.3117,
"step": 419
},
{
"epoch": 0.9256198347107438,
"grad_norm": 11.428281868746616,
"learning_rate": 2.9497922624862047e-06,
"loss": 0.2195,
"step": 420
},
{
"epoch": 0.9278236914600551,
"grad_norm": 10.331982266269929,
"learning_rate": 2.9495220371029163e-06,
"loss": 0.2459,
"step": 421
},
{
"epoch": 0.9300275482093664,
"grad_norm": 8.067422045194437,
"learning_rate": 2.949251098918263e-06,
"loss": 0.3405,
"step": 422
},
{
"epoch": 0.9322314049586777,
"grad_norm": 6.509185070006409,
"learning_rate": 2.9489794480654803e-06,
"loss": 0.2208,
"step": 423
},
{
"epoch": 0.934435261707989,
"grad_norm": 5.996298899230075,
"learning_rate": 2.948707084678152e-06,
"loss": 0.2281,
"step": 424
},
{
"epoch": 0.9366391184573003,
"grad_norm": 5.800748387128115,
"learning_rate": 2.9484340088902114e-06,
"loss": 0.1947,
"step": 425
},
{
"epoch": 0.9388429752066115,
"grad_norm": 9.4543217677755,
"learning_rate": 2.9481602208359446e-06,
"loss": 0.2713,
"step": 426
},
{
"epoch": 0.9410468319559229,
"grad_norm": 5.8179050501506575,
"learning_rate": 2.947885720649986e-06,
"loss": 0.4038,
"step": 427
},
{
"epoch": 0.9432506887052342,
"grad_norm": 5.9968558032817025,
"learning_rate": 2.947610508467321e-06,
"loss": 0.308,
"step": 428
},
{
"epoch": 0.9454545454545454,
"grad_norm": 5.218918783837259,
"learning_rate": 2.947334584423285e-06,
"loss": 0.1904,
"step": 429
},
{
"epoch": 0.9476584022038568,
"grad_norm": 5.893715073575893,
"learning_rate": 2.947057948653564e-06,
"loss": 0.3125,
"step": 430
},
{
"epoch": 0.949862258953168,
"grad_norm": 7.465509921171662,
"learning_rate": 2.946780601294192e-06,
"loss": 0.2722,
"step": 431
},
{
"epoch": 0.9520661157024793,
"grad_norm": 5.375598880031633,
"learning_rate": 2.946502542481556e-06,
"loss": 0.1337,
"step": 432
},
{
"epoch": 0.9542699724517907,
"grad_norm": 6.152314340379761,
"learning_rate": 2.9462237723523904e-06,
"loss": 0.1672,
"step": 433
},
{
"epoch": 0.9564738292011019,
"grad_norm": 7.412726910934929,
"learning_rate": 2.94594429104378e-06,
"loss": 0.2559,
"step": 434
},
{
"epoch": 0.9586776859504132,
"grad_norm": 4.711748751487062,
"learning_rate": 2.94566409869316e-06,
"loss": 0.156,
"step": 435
},
{
"epoch": 0.9608815426997245,
"grad_norm": 7.236972071641592,
"learning_rate": 2.945383195438314e-06,
"loss": 0.3117,
"step": 436
},
{
"epoch": 0.9630853994490358,
"grad_norm": 14.403387281330506,
"learning_rate": 2.9451015814173773e-06,
"loss": 0.3208,
"step": 437
},
{
"epoch": 0.9652892561983472,
"grad_norm": 6.533624513809372,
"learning_rate": 2.9448192567688325e-06,
"loss": 0.1976,
"step": 438
},
{
"epoch": 0.9674931129476584,
"grad_norm": 4.256886100795899,
"learning_rate": 2.944536221631512e-06,
"loss": 0.1006,
"step": 439
},
{
"epoch": 0.9696969696969697,
"grad_norm": 5.130456426054757,
"learning_rate": 2.9442524761445994e-06,
"loss": 0.2601,
"step": 440
},
{
"epoch": 0.971900826446281,
"grad_norm": 7.200199447067768,
"learning_rate": 2.9439680204476253e-06,
"loss": 0.1423,
"step": 441
},
{
"epoch": 0.9741046831955923,
"grad_norm": 9.478705252834285,
"learning_rate": 2.943682854680471e-06,
"loss": 0.2328,
"step": 442
},
{
"epoch": 0.9763085399449036,
"grad_norm": 6.804473740823927,
"learning_rate": 2.9433969789833666e-06,
"loss": 0.1649,
"step": 443
},
{
"epoch": 0.9785123966942149,
"grad_norm": 8.72015955310415,
"learning_rate": 2.9431103934968913e-06,
"loss": 0.1498,
"step": 444
},
{
"epoch": 0.9807162534435262,
"grad_norm": 8.837271411112722,
"learning_rate": 2.942823098361973e-06,
"loss": 0.2892,
"step": 445
},
{
"epoch": 0.9829201101928374,
"grad_norm": 5.118218748161681,
"learning_rate": 2.942535093719889e-06,
"loss": 0.2846,
"step": 446
},
{
"epoch": 0.9851239669421488,
"grad_norm": 7.494790672708267,
"learning_rate": 2.942246379712265e-06,
"loss": 0.2539,
"step": 447
},
{
"epoch": 0.9873278236914601,
"grad_norm": 4.060442461438909,
"learning_rate": 2.9419569564810767e-06,
"loss": 0.3017,
"step": 448
},
{
"epoch": 0.9895316804407713,
"grad_norm": 7.559439545503309,
"learning_rate": 2.9416668241686477e-06,
"loss": 0.3009,
"step": 449
},
{
"epoch": 0.9917355371900827,
"grad_norm": 4.546880168760103,
"learning_rate": 2.9413759829176495e-06,
"loss": 0.2334,
"step": 450
},
{
"epoch": 0.9939393939393939,
"grad_norm": 5.258995040306454,
"learning_rate": 2.9410844328711035e-06,
"loss": 0.1016,
"step": 451
},
{
"epoch": 0.9961432506887052,
"grad_norm": 6.053619389338811,
"learning_rate": 2.9407921741723793e-06,
"loss": 0.2842,
"step": 452
},
{
"epoch": 0.9983471074380166,
"grad_norm": 5.754465135177187,
"learning_rate": 2.9404992069651952e-06,
"loss": 0.3233,
"step": 453
},
{
"epoch": 1.0005509641873278,
"grad_norm": 4.9612019994531975,
"learning_rate": 2.9402055313936167e-06,
"loss": 0.1417,
"step": 454
},
{
"epoch": 1.002754820936639,
"grad_norm": 5.6244102135741825,
"learning_rate": 2.9399111476020595e-06,
"loss": 0.2074,
"step": 455
},
{
"epoch": 1.0049586776859505,
"grad_norm": 3.6159502978670965,
"learning_rate": 2.939616055735286e-06,
"loss": 0.0678,
"step": 456
},
{
"epoch": 1.0071625344352617,
"grad_norm": 4.375326321208629,
"learning_rate": 2.939320255938408e-06,
"loss": 0.1807,
"step": 457
},
{
"epoch": 1.009366391184573,
"grad_norm": 5.632731167384229,
"learning_rate": 2.9390237483568837e-06,
"loss": 0.2764,
"step": 458
},
{
"epoch": 1.0115702479338844,
"grad_norm": 5.570279486677513,
"learning_rate": 2.938726533136522e-06,
"loss": 0.2779,
"step": 459
},
{
"epoch": 1.0137741046831956,
"grad_norm": 6.526157916565178,
"learning_rate": 2.938428610423477e-06,
"loss": 0.2135,
"step": 460
},
{
"epoch": 1.0159779614325068,
"grad_norm": 10.345737600617348,
"learning_rate": 2.9381299803642527e-06,
"loss": 0.1378,
"step": 461
},
{
"epoch": 1.018181818181818,
"grad_norm": 11.833262955463693,
"learning_rate": 2.9378306431057e-06,
"loss": 0.139,
"step": 462
},
{
"epoch": 1.0203856749311295,
"grad_norm": 8.344807812955779,
"learning_rate": 2.9375305987950176e-06,
"loss": 0.0673,
"step": 463
},
{
"epoch": 1.0225895316804408,
"grad_norm": 5.488027829308774,
"learning_rate": 2.9372298475797512e-06,
"loss": 0.1667,
"step": 464
},
{
"epoch": 1.024793388429752,
"grad_norm": 5.683146118159619,
"learning_rate": 2.936928389607797e-06,
"loss": 0.1326,
"step": 465
},
{
"epoch": 1.0269972451790634,
"grad_norm": 8.805650429808152,
"learning_rate": 2.9366262250273957e-06,
"loss": 0.375,
"step": 466
},
{
"epoch": 1.0292011019283747,
"grad_norm": 6.748749150131435,
"learning_rate": 2.9363233539871365e-06,
"loss": 0.201,
"step": 467
},
{
"epoch": 1.031404958677686,
"grad_norm": 5.494562831147715,
"learning_rate": 2.936019776635956e-06,
"loss": 0.1246,
"step": 468
},
{
"epoch": 1.0336088154269973,
"grad_norm": 7.74993668962211,
"learning_rate": 2.9357154931231377e-06,
"loss": 0.3916,
"step": 469
},
{
"epoch": 1.0358126721763086,
"grad_norm": 3.5550318373349805,
"learning_rate": 2.935410503598313e-06,
"loss": 0.1295,
"step": 470
},
{
"epoch": 1.0380165289256198,
"grad_norm": 11.370592776936205,
"learning_rate": 2.9351048082114615e-06,
"loss": 0.2021,
"step": 471
},
{
"epoch": 1.040220385674931,
"grad_norm": 6.916155255829466,
"learning_rate": 2.934798407112907e-06,
"loss": 0.2371,
"step": 472
},
{
"epoch": 1.0424242424242425,
"grad_norm": 4.724194212237889,
"learning_rate": 2.9344913004533225e-06,
"loss": 0.1488,
"step": 473
},
{
"epoch": 1.0446280991735537,
"grad_norm": 5.567099475250736,
"learning_rate": 2.9341834883837276e-06,
"loss": 0.371,
"step": 474
},
{
"epoch": 1.046831955922865,
"grad_norm": 4.549670157011353,
"learning_rate": 2.9338749710554895e-06,
"loss": 0.1426,
"step": 475
},
{
"epoch": 1.0490358126721764,
"grad_norm": 3.664024697925411,
"learning_rate": 2.93356574862032e-06,
"loss": 0.1283,
"step": 476
},
{
"epoch": 1.0512396694214876,
"grad_norm": 5.63620689419913,
"learning_rate": 2.9332558212302795e-06,
"loss": 0.3073,
"step": 477
},
{
"epoch": 1.0534435261707988,
"grad_norm": 5.920190052335488,
"learning_rate": 2.9329451890377745e-06,
"loss": 0.3822,
"step": 478
},
{
"epoch": 1.0556473829201103,
"grad_norm": 4.5184551389608965,
"learning_rate": 2.9326338521955584e-06,
"loss": 0.2088,
"step": 479
},
{
"epoch": 1.0578512396694215,
"grad_norm": 4.849019465438414,
"learning_rate": 2.932321810856731e-06,
"loss": 0.2116,
"step": 480
},
{
"epoch": 1.0600550964187327,
"grad_norm": 9.26096782279171,
"learning_rate": 2.9320090651747376e-06,
"loss": 0.2098,
"step": 481
},
{
"epoch": 1.062258953168044,
"grad_norm": 5.881206475990649,
"learning_rate": 2.9316956153033713e-06,
"loss": 0.3356,
"step": 482
},
{
"epoch": 1.0644628099173554,
"grad_norm": 5.2250601677751325,
"learning_rate": 2.931381461396771e-06,
"loss": 0.1708,
"step": 483
},
{
"epoch": 1.0666666666666667,
"grad_norm": 2.656994585543327,
"learning_rate": 2.9310666036094206e-06,
"loss": 0.1475,
"step": 484
},
{
"epoch": 1.0688705234159779,
"grad_norm": 5.808437762834121,
"learning_rate": 2.9307510420961527e-06,
"loss": 0.1207,
"step": 485
},
{
"epoch": 1.0710743801652893,
"grad_norm": 4.365336973897983,
"learning_rate": 2.9304347770121433e-06,
"loss": 0.1158,
"step": 486
},
{
"epoch": 1.0732782369146006,
"grad_norm": 8.115694826703175,
"learning_rate": 2.9301178085129156e-06,
"loss": 0.3765,
"step": 487
},
{
"epoch": 1.0754820936639118,
"grad_norm": 4.782580532088457,
"learning_rate": 2.929800136754339e-06,
"loss": 0.3289,
"step": 488
},
{
"epoch": 1.0776859504132232,
"grad_norm": 3.916128908970371,
"learning_rate": 2.929481761892629e-06,
"loss": 0.2164,
"step": 489
},
{
"epoch": 1.0798898071625345,
"grad_norm": 4.684165993134075,
"learning_rate": 2.9291626840843446e-06,
"loss": 0.19,
"step": 490
},
{
"epoch": 1.0820936639118457,
"grad_norm": 9.8418516597872,
"learning_rate": 2.9288429034863927e-06,
"loss": 0.1844,
"step": 491
},
{
"epoch": 1.084297520661157,
"grad_norm": 3.8532530548962187,
"learning_rate": 2.928522420256026e-06,
"loss": 0.1772,
"step": 492
},
{
"epoch": 1.0865013774104684,
"grad_norm": 3.98234924660316,
"learning_rate": 2.928201234550841e-06,
"loss": 0.1009,
"step": 493
},
{
"epoch": 1.0887052341597796,
"grad_norm": 4.128780640643151,
"learning_rate": 2.927879346528781e-06,
"loss": 0.2658,
"step": 494
},
{
"epoch": 1.0909090909090908,
"grad_norm": 4.492600919543367,
"learning_rate": 2.9275567563481335e-06,
"loss": 0.1274,
"step": 495
},
{
"epoch": 1.0931129476584023,
"grad_norm": 20.5140436999278,
"learning_rate": 2.927233464167533e-06,
"loss": 0.1561,
"step": 496
},
{
"epoch": 1.0953168044077135,
"grad_norm": 11.77382225835973,
"learning_rate": 2.926909470145957e-06,
"loss": 0.3978,
"step": 497
},
{
"epoch": 1.0975206611570247,
"grad_norm": 5.588583740069422,
"learning_rate": 2.9265847744427307e-06,
"loss": 0.2566,
"step": 498
},
{
"epoch": 1.0997245179063362,
"grad_norm": 3.8060281982906097,
"learning_rate": 2.9262593772175216e-06,
"loss": 0.0789,
"step": 499
},
{
"epoch": 1.1019283746556474,
"grad_norm": 5.73992801608302,
"learning_rate": 2.925933278630344e-06,
"loss": 0.2749,
"step": 500
},
{
"epoch": 1.1041322314049586,
"grad_norm": 5.387804017557412,
"learning_rate": 2.9256064788415577e-06,
"loss": 0.1235,
"step": 501
},
{
"epoch": 1.1063360881542699,
"grad_norm": 4.598141851922886,
"learning_rate": 2.9252789780118643e-06,
"loss": 0.2522,
"step": 502
},
{
"epoch": 1.1085399449035813,
"grad_norm": 5.507025996681415,
"learning_rate": 2.924950776302314e-06,
"loss": 0.1645,
"step": 503
},
{
"epoch": 1.1107438016528925,
"grad_norm": 7.5059518483881735,
"learning_rate": 2.924621873874298e-06,
"loss": 0.1555,
"step": 504
},
{
"epoch": 1.1129476584022038,
"grad_norm": 6.046492745795359,
"learning_rate": 2.924292270889555e-06,
"loss": 0.1493,
"step": 505
},
{
"epoch": 1.1151515151515152,
"grad_norm": 9.450028349535398,
"learning_rate": 2.923961967510167e-06,
"loss": 0.1064,
"step": 506
},
{
"epoch": 1.1173553719008265,
"grad_norm": 10.272796845420118,
"learning_rate": 2.9236309638985597e-06,
"loss": 0.1292,
"step": 507
},
{
"epoch": 1.1195592286501377,
"grad_norm": 4.261466277927439,
"learning_rate": 2.9232992602175044e-06,
"loss": 0.1745,
"step": 508
},
{
"epoch": 1.1217630853994491,
"grad_norm": 3.6583947560721763,
"learning_rate": 2.922966856630116e-06,
"loss": 0.193,
"step": 509
},
{
"epoch": 1.1239669421487604,
"grad_norm": 4.194634352059231,
"learning_rate": 2.9226337532998544e-06,
"loss": 0.156,
"step": 510
},
{
"epoch": 1.1261707988980716,
"grad_norm": 5.271525237798146,
"learning_rate": 2.9222999503905216e-06,
"loss": 0.289,
"step": 511
},
{
"epoch": 1.128374655647383,
"grad_norm": 5.927523030580433,
"learning_rate": 2.9219654480662657e-06,
"loss": 0.2661,
"step": 512
},
{
"epoch": 1.1305785123966943,
"grad_norm": 3.674895915077014,
"learning_rate": 2.921630246491578e-06,
"loss": 0.0736,
"step": 513
},
{
"epoch": 1.1327823691460055,
"grad_norm": 10.052945905533496,
"learning_rate": 2.9212943458312934e-06,
"loss": 0.2025,
"step": 514
},
{
"epoch": 1.1349862258953167,
"grad_norm": 4.6865949140466086,
"learning_rate": 2.9209577462505908e-06,
"loss": 0.2828,
"step": 515
},
{
"epoch": 1.1371900826446282,
"grad_norm": 5.089446996586012,
"learning_rate": 2.9206204479149927e-06,
"loss": 0.2029,
"step": 516
},
{
"epoch": 1.1393939393939394,
"grad_norm": 9.684303751299739,
"learning_rate": 2.9202824509903656e-06,
"loss": 0.1925,
"step": 517
},
{
"epoch": 1.1415977961432506,
"grad_norm": 4.779441705636188,
"learning_rate": 2.919943755642919e-06,
"loss": 0.3132,
"step": 518
},
{
"epoch": 1.143801652892562,
"grad_norm": 3.4207597108395507,
"learning_rate": 2.919604362039207e-06,
"loss": 0.134,
"step": 519
},
{
"epoch": 1.1460055096418733,
"grad_norm": 7.449231612875146,
"learning_rate": 2.9192642703461243e-06,
"loss": 0.2683,
"step": 520
},
{
"epoch": 1.1482093663911845,
"grad_norm": 4.032811008116536,
"learning_rate": 2.918923480730912e-06,
"loss": 0.1896,
"step": 521
},
{
"epoch": 1.1504132231404958,
"grad_norm": 4.905485293464102,
"learning_rate": 2.918581993361153e-06,
"loss": 0.1989,
"step": 522
},
{
"epoch": 1.1526170798898072,
"grad_norm": 7.339812817109276,
"learning_rate": 2.9182398084047736e-06,
"loss": 0.1412,
"step": 523
},
{
"epoch": 1.1548209366391184,
"grad_norm": 3.3709215655039113,
"learning_rate": 2.9178969260300427e-06,
"loss": 0.0843,
"step": 524
},
{
"epoch": 1.1570247933884297,
"grad_norm": 3.9137809188731207,
"learning_rate": 2.917553346405572e-06,
"loss": 0.1604,
"step": 525
},
{
"epoch": 1.1592286501377411,
"grad_norm": 4.547245306347712,
"learning_rate": 2.9172090697003175e-06,
"loss": 0.1775,
"step": 526
},
{
"epoch": 1.1614325068870524,
"grad_norm": 3.8176334567973282,
"learning_rate": 2.916864096083576e-06,
"loss": 0.2646,
"step": 527
},
{
"epoch": 1.1636363636363636,
"grad_norm": 3.534012301229345,
"learning_rate": 2.916518425724989e-06,
"loss": 0.2208,
"step": 528
},
{
"epoch": 1.1658402203856748,
"grad_norm": 5.268005108434388,
"learning_rate": 2.9161720587945387e-06,
"loss": 0.207,
"step": 529
},
{
"epoch": 1.1680440771349863,
"grad_norm": 5.3661996707943,
"learning_rate": 2.9158249954625514e-06,
"loss": 0.0658,
"step": 530
},
{
"epoch": 1.1702479338842975,
"grad_norm": 4.28879923967976,
"learning_rate": 2.915477235899695e-06,
"loss": 0.1175,
"step": 531
},
{
"epoch": 1.172451790633609,
"grad_norm": 5.963529550956904,
"learning_rate": 2.9151287802769803e-06,
"loss": 0.1794,
"step": 532
},
{
"epoch": 1.1746556473829202,
"grad_norm": 5.694876856120757,
"learning_rate": 2.91477962876576e-06,
"loss": 0.0812,
"step": 533
},
{
"epoch": 1.1768595041322314,
"grad_norm": 4.351944044210483,
"learning_rate": 2.9144297815377286e-06,
"loss": 0.3211,
"step": 534
},
{
"epoch": 1.1790633608815426,
"grad_norm": 3.972657146689943,
"learning_rate": 2.9140792387649237e-06,
"loss": 0.1349,
"step": 535
},
{
"epoch": 1.181267217630854,
"grad_norm": 3.7197957379879587,
"learning_rate": 2.9137280006197245e-06,
"loss": 0.2268,
"step": 536
},
{
"epoch": 1.1834710743801653,
"grad_norm": 5.329312317532163,
"learning_rate": 2.9133760672748516e-06,
"loss": 0.2767,
"step": 537
},
{
"epoch": 1.1856749311294765,
"grad_norm": 3.883042008810904,
"learning_rate": 2.9130234389033687e-06,
"loss": 0.1757,
"step": 538
},
{
"epoch": 1.187878787878788,
"grad_norm": 4.5902195667184635,
"learning_rate": 2.9126701156786794e-06,
"loss": 0.241,
"step": 539
},
{
"epoch": 1.1900826446280992,
"grad_norm": 3.3442026394957405,
"learning_rate": 2.912316097774531e-06,
"loss": 0.1603,
"step": 540
},
{
"epoch": 1.1922865013774104,
"grad_norm": 4.162882139124656,
"learning_rate": 2.9119613853650117e-06,
"loss": 0.2736,
"step": 541
},
{
"epoch": 1.1944903581267217,
"grad_norm": 4.783670244425796,
"learning_rate": 2.9116059786245505e-06,
"loss": 0.2848,
"step": 542
},
{
"epoch": 1.1966942148760331,
"grad_norm": 3.644438346409876,
"learning_rate": 2.9112498777279187e-06,
"loss": 0.1325,
"step": 543
},
{
"epoch": 1.1988980716253443,
"grad_norm": 4.2703260216286285,
"learning_rate": 2.9108930828502284e-06,
"loss": 0.2723,
"step": 544
},
{
"epoch": 1.2011019283746556,
"grad_norm": 2.5801703466779036,
"learning_rate": 2.9105355941669337e-06,
"loss": 0.1598,
"step": 545
},
{
"epoch": 1.203305785123967,
"grad_norm": 9.2861271196882,
"learning_rate": 2.9101774118538283e-06,
"loss": 0.254,
"step": 546
},
{
"epoch": 1.2055096418732782,
"grad_norm": 4.017243467065785,
"learning_rate": 2.90981853608705e-06,
"loss": 0.2379,
"step": 547
},
{
"epoch": 1.2077134986225895,
"grad_norm": 3.9003651881550945,
"learning_rate": 2.909458967043074e-06,
"loss": 0.0777,
"step": 548
},
{
"epoch": 1.2099173553719007,
"grad_norm": 8.654932450813039,
"learning_rate": 2.909098704898718e-06,
"loss": 0.2069,
"step": 549
},
{
"epoch": 1.2121212121212122,
"grad_norm": 6.2364635749542625,
"learning_rate": 2.908737749831142e-06,
"loss": 0.2134,
"step": 550
},
{
"epoch": 1.2143250688705234,
"grad_norm": 6.731388982582371,
"learning_rate": 2.9083761020178443e-06,
"loss": 0.2253,
"step": 551
},
{
"epoch": 1.2165289256198348,
"grad_norm": 3.502965933051062,
"learning_rate": 2.908013761636665e-06,
"loss": 0.2066,
"step": 552
},
{
"epoch": 1.218732782369146,
"grad_norm": 14.711406216867847,
"learning_rate": 2.9076507288657855e-06,
"loss": 0.297,
"step": 553
},
{
"epoch": 1.2209366391184573,
"grad_norm": 17.15730611268536,
"learning_rate": 2.9072870038837265e-06,
"loss": 0.3287,
"step": 554
},
{
"epoch": 1.2231404958677685,
"grad_norm": 11.708958106392894,
"learning_rate": 2.906922586869349e-06,
"loss": 0.2429,
"step": 555
},
{
"epoch": 1.22534435261708,
"grad_norm": 5.040925462071028,
"learning_rate": 2.906557478001855e-06,
"loss": 0.3866,
"step": 556
},
{
"epoch": 1.2275482093663912,
"grad_norm": 5.4384090934766105,
"learning_rate": 2.9061916774607868e-06,
"loss": 0.1418,
"step": 557
},
{
"epoch": 1.2297520661157024,
"grad_norm": 4.224206508632855,
"learning_rate": 2.9058251854260266e-06,
"loss": 0.2607,
"step": 558
},
{
"epoch": 1.2319559228650139,
"grad_norm": 6.324360645802438,
"learning_rate": 2.9054580020777965e-06,
"loss": 0.1429,
"step": 559
},
{
"epoch": 1.234159779614325,
"grad_norm": 4.560338619966233,
"learning_rate": 2.9050901275966583e-06,
"loss": 0.2047,
"step": 560
},
{
"epoch": 1.2363636363636363,
"grad_norm": 4.978147123367892,
"learning_rate": 2.9047215621635138e-06,
"loss": 0.1914,
"step": 561
},
{
"epoch": 1.2385674931129476,
"grad_norm": 3.771414086700111,
"learning_rate": 2.904352305959606e-06,
"loss": 0.2456,
"step": 562
},
{
"epoch": 1.240771349862259,
"grad_norm": 5.891321373092408,
"learning_rate": 2.903982359166515e-06,
"loss": 0.1384,
"step": 563
},
{
"epoch": 1.2429752066115702,
"grad_norm": 7.95915988811828,
"learning_rate": 2.9036117219661623e-06,
"loss": 0.1797,
"step": 564
},
{
"epoch": 1.2451790633608815,
"grad_norm": 6.605600633405355,
"learning_rate": 2.903240394540809e-06,
"loss": 0.2491,
"step": 565
},
{
"epoch": 1.247382920110193,
"grad_norm": 5.488205726433599,
"learning_rate": 2.9028683770730546e-06,
"loss": 0.118,
"step": 566
},
{
"epoch": 1.2495867768595041,
"grad_norm": 4.012615367050864,
"learning_rate": 2.9024956697458384e-06,
"loss": 0.2405,
"step": 567
},
{
"epoch": 1.2517906336088154,
"grad_norm": 3.5236670281097413,
"learning_rate": 2.9021222727424384e-06,
"loss": 0.2286,
"step": 568
},
{
"epoch": 1.2539944903581266,
"grad_norm": 9.955313424056465,
"learning_rate": 2.9017481862464737e-06,
"loss": 0.1459,
"step": 569
},
{
"epoch": 1.256198347107438,
"grad_norm": 6.660973468538957,
"learning_rate": 2.9013734104419e-06,
"loss": 0.1938,
"step": 570
},
{
"epoch": 1.2584022038567493,
"grad_norm": 16.963451273767287,
"learning_rate": 2.900997945513013e-06,
"loss": 0.1876,
"step": 571
},
{
"epoch": 1.2606060606060607,
"grad_norm": 5.7870946489445085,
"learning_rate": 2.900621791644448e-06,
"loss": 0.2791,
"step": 572
},
{
"epoch": 1.262809917355372,
"grad_norm": 10.352260259287972,
"learning_rate": 2.9002449490211776e-06,
"loss": 0.1946,
"step": 573
},
{
"epoch": 1.2650137741046832,
"grad_norm": 6.861136772777287,
"learning_rate": 2.899867417828514e-06,
"loss": 0.3179,
"step": 574
},
{
"epoch": 1.2672176308539944,
"grad_norm": 10.960820361107482,
"learning_rate": 2.8994891982521083e-06,
"loss": 0.329,
"step": 575
},
{
"epoch": 1.2694214876033059,
"grad_norm": 21.54723751149437,
"learning_rate": 2.8991102904779495e-06,
"loss": 0.2181,
"step": 576
},
{
"epoch": 1.271625344352617,
"grad_norm": 14.02787145282944,
"learning_rate": 2.898730694692365e-06,
"loss": 0.2812,
"step": 577
},
{
"epoch": 1.2738292011019283,
"grad_norm": 7.0557599098044514,
"learning_rate": 2.8983504110820214e-06,
"loss": 0.353,
"step": 578
},
{
"epoch": 1.2760330578512398,
"grad_norm": 4.770915540206801,
"learning_rate": 2.8979694398339224e-06,
"loss": 0.1426,
"step": 579
},
{
"epoch": 1.278236914600551,
"grad_norm": 4.483304136348754,
"learning_rate": 2.8975877811354097e-06,
"loss": 0.1747,
"step": 580
},
{
"epoch": 1.2804407713498622,
"grad_norm": 6.3811646159524305,
"learning_rate": 2.8972054351741647e-06,
"loss": 0.216,
"step": 581
},
{
"epoch": 1.2826446280991735,
"grad_norm": 8.112493117800566,
"learning_rate": 2.8968224021382054e-06,
"loss": 0.2906,
"step": 582
},
{
"epoch": 1.284848484848485,
"grad_norm": 4.851798637905597,
"learning_rate": 2.8964386822158878e-06,
"loss": 0.2229,
"step": 583
},
{
"epoch": 1.2870523415977961,
"grad_norm": 5.317762416772784,
"learning_rate": 2.8960542755959063e-06,
"loss": 0.1888,
"step": 584
},
{
"epoch": 1.2892561983471074,
"grad_norm": 5.350207917205157,
"learning_rate": 2.895669182467292e-06,
"loss": 0.1425,
"step": 585
},
{
"epoch": 1.2914600550964188,
"grad_norm": 4.3468333293964445,
"learning_rate": 2.8952834030194145e-06,
"loss": 0.1799,
"step": 586
},
{
"epoch": 1.29366391184573,
"grad_norm": 3.3336068941106656,
"learning_rate": 2.8948969374419803e-06,
"loss": 0.2205,
"step": 587
},
{
"epoch": 1.2958677685950413,
"grad_norm": 2.620320493915155,
"learning_rate": 2.894509785925034e-06,
"loss": 0.1478,
"step": 588
},
{
"epoch": 1.2980716253443525,
"grad_norm": 4.3011475238646275,
"learning_rate": 2.894121948658957e-06,
"loss": 0.2253,
"step": 589
},
{
"epoch": 1.300275482093664,
"grad_norm": 2.8022877416603422,
"learning_rate": 2.8937334258344676e-06,
"loss": 0.166,
"step": 590
},
{
"epoch": 1.3024793388429752,
"grad_norm": 3.7618953894768494,
"learning_rate": 2.893344217642621e-06,
"loss": 0.2193,
"step": 591
},
{
"epoch": 1.3046831955922866,
"grad_norm": 4.0491570230427625,
"learning_rate": 2.892954324274812e-06,
"loss": 0.319,
"step": 592
},
{
"epoch": 1.3068870523415979,
"grad_norm": 4.522049092181143,
"learning_rate": 2.892563745922769e-06,
"loss": 0.2692,
"step": 593
},
{
"epoch": 1.309090909090909,
"grad_norm": 3.055036543425317,
"learning_rate": 2.8921724827785583e-06,
"loss": 0.2053,
"step": 594
},
{
"epoch": 1.3112947658402203,
"grad_norm": 5.267763701241814,
"learning_rate": 2.8917805350345846e-06,
"loss": 0.2822,
"step": 595
},
{
"epoch": 1.3134986225895318,
"grad_norm": 4.582217428090828,
"learning_rate": 2.891387902883586e-06,
"loss": 0.289,
"step": 596
},
{
"epoch": 1.315702479338843,
"grad_norm": 4.043911021583197,
"learning_rate": 2.890994586518641e-06,
"loss": 0.1275,
"step": 597
},
{
"epoch": 1.3179063360881542,
"grad_norm": 5.754149214212578,
"learning_rate": 2.8906005861331613e-06,
"loss": 0.2715,
"step": 598
},
{
"epoch": 1.3201101928374657,
"grad_norm": 6.676367069618446,
"learning_rate": 2.8902059019208968e-06,
"loss": 0.267,
"step": 599
},
{
"epoch": 1.322314049586777,
"grad_norm": 3.985915686812139,
"learning_rate": 2.8898105340759333e-06,
"loss": 0.2928,
"step": 600
},
{
"epoch": 1.3245179063360881,
"grad_norm": 3.0780099598439006,
"learning_rate": 2.8894144827926924e-06,
"loss": 0.0972,
"step": 601
},
{
"epoch": 1.3267217630853994,
"grad_norm": 4.750823583413613,
"learning_rate": 2.8890177482659317e-06,
"loss": 0.1178,
"step": 602
},
{
"epoch": 1.3289256198347108,
"grad_norm": 4.26762264184326,
"learning_rate": 2.888620330690746e-06,
"loss": 0.2721,
"step": 603
},
{
"epoch": 1.331129476584022,
"grad_norm": 2.9681851058272506,
"learning_rate": 2.8882222302625646e-06,
"loss": 0.1618,
"step": 604
},
{
"epoch": 1.3333333333333333,
"grad_norm": 3.9008419992283043,
"learning_rate": 2.8878234471771524e-06,
"loss": 0.2589,
"step": 605
},
{
"epoch": 1.3355371900826447,
"grad_norm": 4.32964863227825,
"learning_rate": 2.887423981630612e-06,
"loss": 0.3656,
"step": 606
},
{
"epoch": 1.337741046831956,
"grad_norm": 2.822788224808757,
"learning_rate": 2.8870238338193794e-06,
"loss": 0.1352,
"step": 607
},
{
"epoch": 1.3399449035812672,
"grad_norm": 5.001949178343276,
"learning_rate": 2.8866230039402275e-06,
"loss": 0.2044,
"step": 608
},
{
"epoch": 1.3421487603305784,
"grad_norm": 3.2789633167007586,
"learning_rate": 2.8862214921902627e-06,
"loss": 0.2084,
"step": 609
},
{
"epoch": 1.3443526170798898,
"grad_norm": 3.5145152327020757,
"learning_rate": 2.88581929876693e-06,
"loss": 0.1822,
"step": 610
},
{
"epoch": 1.346556473829201,
"grad_norm": 2.846767948249379,
"learning_rate": 2.885416423868007e-06,
"loss": 0.1922,
"step": 611
},
{
"epoch": 1.3487603305785125,
"grad_norm": 4.320449712641174,
"learning_rate": 2.885012867691607e-06,
"loss": 0.1512,
"step": 612
},
{
"epoch": 1.3509641873278238,
"grad_norm": 5.126925822393263,
"learning_rate": 2.884608630436178e-06,
"loss": 0.342,
"step": 613
},
{
"epoch": 1.353168044077135,
"grad_norm": 5.913929122619571,
"learning_rate": 2.884203712300504e-06,
"loss": 0.2981,
"step": 614
},
{
"epoch": 1.3553719008264462,
"grad_norm": 3.557959909732146,
"learning_rate": 2.8837981134837033e-06,
"loss": 0.2138,
"step": 615
},
{
"epoch": 1.3575757575757577,
"grad_norm": 3.186541176730338,
"learning_rate": 2.8833918341852288e-06,
"loss": 0.0836,
"step": 616
},
{
"epoch": 1.3597796143250689,
"grad_norm": 4.556846309605876,
"learning_rate": 2.882984874604867e-06,
"loss": 0.2688,
"step": 617
},
{
"epoch": 1.3619834710743801,
"grad_norm": 4.406950831041034,
"learning_rate": 2.8825772349427413e-06,
"loss": 0.1912,
"step": 618
},
{
"epoch": 1.3641873278236916,
"grad_norm": 2.398733737100056,
"learning_rate": 2.882168915399307e-06,
"loss": 0.107,
"step": 619
},
{
"epoch": 1.3663911845730028,
"grad_norm": 5.4958527833768285,
"learning_rate": 2.8817599161753557e-06,
"loss": 0.1844,
"step": 620
},
{
"epoch": 1.368595041322314,
"grad_norm": 8.010170031896015,
"learning_rate": 2.8813502374720128e-06,
"loss": 0.1697,
"step": 621
},
{
"epoch": 1.3707988980716252,
"grad_norm": 4.153983084539555,
"learning_rate": 2.8809398794907367e-06,
"loss": 0.1748,
"step": 622
},
{
"epoch": 1.3730027548209367,
"grad_norm": 5.043276780948253,
"learning_rate": 2.880528842433321e-06,
"loss": 0.2369,
"step": 623
},
{
"epoch": 1.375206611570248,
"grad_norm": 5.363109426054472,
"learning_rate": 2.8801171265018927e-06,
"loss": 0.1227,
"step": 624
},
{
"epoch": 1.3774104683195592,
"grad_norm": 5.402978025713253,
"learning_rate": 2.8797047318989125e-06,
"loss": 0.2394,
"step": 625
},
{
"epoch": 1.3796143250688706,
"grad_norm": 4.270847937611633,
"learning_rate": 2.879291658827176e-06,
"loss": 0.2998,
"step": 626
},
{
"epoch": 1.3818181818181818,
"grad_norm": 4.567221237002694,
"learning_rate": 2.878877907489811e-06,
"loss": 0.2392,
"step": 627
},
{
"epoch": 1.384022038567493,
"grad_norm": 4.795273563069606,
"learning_rate": 2.8784634780902797e-06,
"loss": 0.379,
"step": 628
},
{
"epoch": 1.3862258953168043,
"grad_norm": 6.541946363393915,
"learning_rate": 2.878048370832377e-06,
"loss": 0.231,
"step": 629
},
{
"epoch": 1.3884297520661157,
"grad_norm": 6.028831246833223,
"learning_rate": 2.8776325859202315e-06,
"loss": 0.2724,
"step": 630
},
{
"epoch": 1.390633608815427,
"grad_norm": 4.511285043111639,
"learning_rate": 2.877216123558306e-06,
"loss": 0.2531,
"step": 631
},
{
"epoch": 1.3928374655647384,
"grad_norm": 4.6594481678188915,
"learning_rate": 2.8767989839513945e-06,
"loss": 0.2905,
"step": 632
},
{
"epoch": 1.3950413223140496,
"grad_norm": 5.455069765928385,
"learning_rate": 2.8763811673046257e-06,
"loss": 0.1928,
"step": 633
},
{
"epoch": 1.3972451790633609,
"grad_norm": 5.98529944300178,
"learning_rate": 2.875962673823461e-06,
"loss": 0.2185,
"step": 634
},
{
"epoch": 1.399449035812672,
"grad_norm": 4.805666817335156,
"learning_rate": 2.8755435037136932e-06,
"loss": 0.2429,
"step": 635
},
{
"epoch": 1.4016528925619833,
"grad_norm": 4.467346053498231,
"learning_rate": 2.8751236571814497e-06,
"loss": 0.2098,
"step": 636
},
{
"epoch": 1.4038567493112948,
"grad_norm": 3.573352650286319,
"learning_rate": 2.8747031344331895e-06,
"loss": 0.1255,
"step": 637
},
{
"epoch": 1.406060606060606,
"grad_norm": 6.158154604233198,
"learning_rate": 2.8742819356757044e-06,
"loss": 0.197,
"step": 638
},
{
"epoch": 1.4082644628099175,
"grad_norm": 4.281917416552195,
"learning_rate": 2.873860061116118e-06,
"loss": 0.2842,
"step": 639
},
{
"epoch": 1.4104683195592287,
"grad_norm": 11.226892423174888,
"learning_rate": 2.8734375109618874e-06,
"loss": 0.2209,
"step": 640
},
{
"epoch": 1.41267217630854,
"grad_norm": 4.3225627738439645,
"learning_rate": 2.873014285420801e-06,
"loss": 0.2152,
"step": 641
},
{
"epoch": 1.4148760330578511,
"grad_norm": 3.8825529972131867,
"learning_rate": 2.87259038470098e-06,
"loss": 0.2319,
"step": 642
},
{
"epoch": 1.4170798898071626,
"grad_norm": 6.079197790722616,
"learning_rate": 2.872165809010877e-06,
"loss": 0.2112,
"step": 643
},
{
"epoch": 1.4192837465564738,
"grad_norm": 6.1303289009298805,
"learning_rate": 2.8717405585592767e-06,
"loss": 0.2118,
"step": 644
},
{
"epoch": 1.421487603305785,
"grad_norm": 5.12932154131706,
"learning_rate": 2.871314633555296e-06,
"loss": 0.2052,
"step": 645
},
{
"epoch": 1.4236914600550965,
"grad_norm": 4.481220346865912,
"learning_rate": 2.870888034208383e-06,
"loss": 0.0703,
"step": 646
},
{
"epoch": 1.4258953168044077,
"grad_norm": 6.864001797309864,
"learning_rate": 2.8704607607283177e-06,
"loss": 0.0546,
"step": 647
},
{
"epoch": 1.428099173553719,
"grad_norm": 3.8245534470436295,
"learning_rate": 2.8700328133252115e-06,
"loss": 0.2223,
"step": 648
},
{
"epoch": 1.4303030303030302,
"grad_norm": 4.916213275333236,
"learning_rate": 2.869604192209507e-06,
"loss": 0.1604,
"step": 649
},
{
"epoch": 1.4325068870523416,
"grad_norm": 3.2283359865790784,
"learning_rate": 2.869174897591978e-06,
"loss": 0.1728,
"step": 650
},
{
"epoch": 1.4347107438016529,
"grad_norm": 9.18120748323431,
"learning_rate": 2.8687449296837313e-06,
"loss": 0.3458,
"step": 651
},
{
"epoch": 1.4369146005509643,
"grad_norm": 5.808199958957165,
"learning_rate": 2.8683142886962017e-06,
"loss": 0.1804,
"step": 652
},
{
"epoch": 1.4391184573002755,
"grad_norm": 5.665941277081976,
"learning_rate": 2.867882974841157e-06,
"loss": 0.1458,
"step": 653
},
{
"epoch": 1.4413223140495868,
"grad_norm": 3.8288010492231677,
"learning_rate": 2.867450988330696e-06,
"loss": 0.1645,
"step": 654
},
{
"epoch": 1.443526170798898,
"grad_norm": 4.1049680267773425,
"learning_rate": 2.8670183293772466e-06,
"loss": 0.3046,
"step": 655
},
{
"epoch": 1.4457300275482092,
"grad_norm": 4.5412954160400485,
"learning_rate": 2.86658499819357e-06,
"loss": 0.1549,
"step": 656
},
{
"epoch": 1.4479338842975207,
"grad_norm": 4.423878900596039,
"learning_rate": 2.8661509949927553e-06,
"loss": 0.2464,
"step": 657
},
{
"epoch": 1.450137741046832,
"grad_norm": 6.658838123174504,
"learning_rate": 2.865716319988224e-06,
"loss": 0.2806,
"step": 658
},
{
"epoch": 1.4523415977961434,
"grad_norm": 3.4922459804042187,
"learning_rate": 2.8652809733937266e-06,
"loss": 0.1226,
"step": 659
},
{
"epoch": 1.4545454545454546,
"grad_norm": 4.986328878756679,
"learning_rate": 2.8648449554233447e-06,
"loss": 0.3537,
"step": 660
},
{
"epoch": 1.4567493112947658,
"grad_norm": 4.234882868424787,
"learning_rate": 2.8644082662914905e-06,
"loss": 0.1714,
"step": 661
},
{
"epoch": 1.458953168044077,
"grad_norm": 3.9070490424818747,
"learning_rate": 2.8639709062129044e-06,
"loss": 0.1269,
"step": 662
},
{
"epoch": 1.4611570247933885,
"grad_norm": 4.998076466309302,
"learning_rate": 2.863532875402658e-06,
"loss": 0.1779,
"step": 663
},
{
"epoch": 1.4633608815426997,
"grad_norm": 5.106983845887589,
"learning_rate": 2.8630941740761533e-06,
"loss": 0.1186,
"step": 664
},
{
"epoch": 1.465564738292011,
"grad_norm": 3.8551902713378965,
"learning_rate": 2.8626548024491213e-06,
"loss": 0.1354,
"step": 665
},
{
"epoch": 1.4677685950413224,
"grad_norm": 5.2301301811454515,
"learning_rate": 2.862214760737622e-06,
"loss": 0.2037,
"step": 666
},
{
"epoch": 1.4699724517906336,
"grad_norm": 6.0956564948874306,
"learning_rate": 2.861774049158046e-06,
"loss": 0.2184,
"step": 667
},
{
"epoch": 1.4721763085399449,
"grad_norm": 6.291702439669367,
"learning_rate": 2.8613326679271136e-06,
"loss": 0.1466,
"step": 668
},
{
"epoch": 1.474380165289256,
"grad_norm": 3.713580018900493,
"learning_rate": 2.8608906172618723e-06,
"loss": 0.2072,
"step": 669
},
{
"epoch": 1.4765840220385675,
"grad_norm": 5.537260982915714,
"learning_rate": 2.8604478973797005e-06,
"loss": 0.1488,
"step": 670
},
{
"epoch": 1.4787878787878788,
"grad_norm": 5.820569540535823,
"learning_rate": 2.860004508498306e-06,
"loss": 0.223,
"step": 671
},
{
"epoch": 1.4809917355371902,
"grad_norm": 6.1571860143662684,
"learning_rate": 2.8595604508357255e-06,
"loss": 0.4112,
"step": 672
},
{
"epoch": 1.4831955922865014,
"grad_norm": 3.6498194790392082,
"learning_rate": 2.859115724610322e-06,
"loss": 0.1296,
"step": 673
},
{
"epoch": 1.4853994490358127,
"grad_norm": 4.172319048467667,
"learning_rate": 2.858670330040791e-06,
"loss": 0.2823,
"step": 674
},
{
"epoch": 1.487603305785124,
"grad_norm": 3.5422809914301,
"learning_rate": 2.8582242673461545e-06,
"loss": 0.2723,
"step": 675
},
{
"epoch": 1.4898071625344351,
"grad_norm": 3.994421032502254,
"learning_rate": 2.857777536745763e-06,
"loss": 0.2625,
"step": 676
},
{
"epoch": 1.4920110192837466,
"grad_norm": 2.4442328235159136,
"learning_rate": 2.857330138459297e-06,
"loss": 0.0773,
"step": 677
},
{
"epoch": 1.4942148760330578,
"grad_norm": 2.956014467391965,
"learning_rate": 2.856882072706763e-06,
"loss": 0.0263,
"step": 678
},
{
"epoch": 1.4964187327823693,
"grad_norm": 5.063950525118159,
"learning_rate": 2.856433339708498e-06,
"loss": 0.2555,
"step": 679
},
{
"epoch": 1.4986225895316805,
"grad_norm": 5.701498370772839,
"learning_rate": 2.855983939685165e-06,
"loss": 0.2644,
"step": 680
},
{
"epoch": 1.5008264462809917,
"grad_norm": 3.653863039822445,
"learning_rate": 2.8555338728577573e-06,
"loss": 0.1186,
"step": 681
},
{
"epoch": 1.503030303030303,
"grad_norm": 3.9380606238532856,
"learning_rate": 2.8550831394475946e-06,
"loss": 0.2569,
"step": 682
},
{
"epoch": 1.5052341597796142,
"grad_norm": 6.240511119913605,
"learning_rate": 2.854631739676324e-06,
"loss": 0.2301,
"step": 683
},
{
"epoch": 1.5074380165289256,
"grad_norm": 3.969971056794526,
"learning_rate": 2.8541796737659223e-06,
"loss": 0.2162,
"step": 684
},
{
"epoch": 1.509641873278237,
"grad_norm": 4.593297828759556,
"learning_rate": 2.8537269419386913e-06,
"loss": 0.2088,
"step": 685
},
{
"epoch": 1.5118457300275483,
"grad_norm": 3.3975859793614074,
"learning_rate": 2.853273544417262e-06,
"loss": 0.1733,
"step": 686
},
{
"epoch": 1.5140495867768595,
"grad_norm": 4.456601479513547,
"learning_rate": 2.8528194814245925e-06,
"loss": 0.153,
"step": 687
},
{
"epoch": 1.5162534435261708,
"grad_norm": 5.024001056036946,
"learning_rate": 2.852364753183968e-06,
"loss": 0.267,
"step": 688
},
{
"epoch": 1.518457300275482,
"grad_norm": 3.9770565806572504,
"learning_rate": 2.851909359919e-06,
"loss": 0.215,
"step": 689
},
{
"epoch": 1.5206611570247934,
"grad_norm": 4.263990726687718,
"learning_rate": 2.851453301853629e-06,
"loss": 0.1481,
"step": 690
},
{
"epoch": 1.5228650137741047,
"grad_norm": 10.077089034160084,
"learning_rate": 2.8509965792121202e-06,
"loss": 0.1462,
"step": 691
},
{
"epoch": 1.525068870523416,
"grad_norm": 4.929665896210096,
"learning_rate": 2.850539192219067e-06,
"loss": 0.1636,
"step": 692
},
{
"epoch": 1.5272727272727273,
"grad_norm": 5.860454925100347,
"learning_rate": 2.850081141099389e-06,
"loss": 0.3131,
"step": 693
},
{
"epoch": 1.5294765840220386,
"grad_norm": 3.831102793731722,
"learning_rate": 2.8496224260783324e-06,
"loss": 0.2469,
"step": 694
},
{
"epoch": 1.5316804407713498,
"grad_norm": 2.8227649238802925,
"learning_rate": 2.8491630473814703e-06,
"loss": 0.1926,
"step": 695
},
{
"epoch": 1.533884297520661,
"grad_norm": 6.082271321101482,
"learning_rate": 2.8487030052347015e-06,
"loss": 0.3531,
"step": 696
},
{
"epoch": 1.5360881542699725,
"grad_norm": 4.213384468163032,
"learning_rate": 2.8482422998642513e-06,
"loss": 0.1729,
"step": 697
},
{
"epoch": 1.5382920110192837,
"grad_norm": 4.192370379792702,
"learning_rate": 2.8477809314966718e-06,
"loss": 0.1484,
"step": 698
},
{
"epoch": 1.5404958677685952,
"grad_norm": 4.619842000189168,
"learning_rate": 2.84731890035884e-06,
"loss": 0.1681,
"step": 699
},
{
"epoch": 1.5426997245179064,
"grad_norm": 6.240088261713722,
"learning_rate": 2.84685620667796e-06,
"loss": 0.1121,
"step": 700
},
{
"epoch": 1.5449035812672176,
"grad_norm": 7.318489937821377,
"learning_rate": 2.8463928506815605e-06,
"loss": 0.238,
"step": 701
},
{
"epoch": 1.5471074380165288,
"grad_norm": 11.369549225694884,
"learning_rate": 2.8459288325974963e-06,
"loss": 0.221,
"step": 702
},
{
"epoch": 1.54931129476584,
"grad_norm": 3.9222359665127127,
"learning_rate": 2.845464152653948e-06,
"loss": 0.1914,
"step": 703
},
{
"epoch": 1.5515151515151515,
"grad_norm": 3.9561654328009626,
"learning_rate": 2.8449988110794224e-06,
"loss": 0.1936,
"step": 704
},
{
"epoch": 1.553719008264463,
"grad_norm": 5.641732295279923,
"learning_rate": 2.8445328081027502e-06,
"loss": 0.196,
"step": 705
},
{
"epoch": 1.5559228650137742,
"grad_norm": 5.310630607978979,
"learning_rate": 2.844066143953088e-06,
"loss": 0.4102,
"step": 706
},
{
"epoch": 1.5581267217630854,
"grad_norm": 3.7257912432993767,
"learning_rate": 2.8435988188599175e-06,
"loss": 0.1045,
"step": 707
},
{
"epoch": 1.5603305785123966,
"grad_norm": 5.021805378306946,
"learning_rate": 2.8431308330530453e-06,
"loss": 0.2973,
"step": 708
},
{
"epoch": 1.5625344352617079,
"grad_norm": 5.460789289795145,
"learning_rate": 2.8426621867626043e-06,
"loss": 0.177,
"step": 709
},
{
"epoch": 1.5647382920110193,
"grad_norm": 5.497664501546296,
"learning_rate": 2.842192880219049e-06,
"loss": 0.242,
"step": 710
},
{
"epoch": 1.5669421487603306,
"grad_norm": 5.424348953607276,
"learning_rate": 2.841722913653162e-06,
"loss": 0.2714,
"step": 711
},
{
"epoch": 1.569146005509642,
"grad_norm": 5.001059981673565,
"learning_rate": 2.8412522872960486e-06,
"loss": 0.2323,
"step": 712
},
{
"epoch": 1.5713498622589532,
"grad_norm": 3.902252684011861,
"learning_rate": 2.840781001379139e-06,
"loss": 0.0884,
"step": 713
},
{
"epoch": 1.5735537190082645,
"grad_norm": 4.121371795480466,
"learning_rate": 2.8403090561341878e-06,
"loss": 0.2631,
"step": 714
},
{
"epoch": 1.5757575757575757,
"grad_norm": 3.7110214921650324,
"learning_rate": 2.8398364517932727e-06,
"loss": 0.2998,
"step": 715
},
{
"epoch": 1.577961432506887,
"grad_norm": 6.2025699416635,
"learning_rate": 2.8393631885887986e-06,
"loss": 0.1385,
"step": 716
},
{
"epoch": 1.5801652892561984,
"grad_norm": 3.279306066462292,
"learning_rate": 2.8388892667534903e-06,
"loss": 0.1619,
"step": 717
},
{
"epoch": 1.5823691460055096,
"grad_norm": 3.1972358459552996,
"learning_rate": 2.838414686520399e-06,
"loss": 0.2275,
"step": 718
},
{
"epoch": 1.584573002754821,
"grad_norm": 4.729096055141415,
"learning_rate": 2.8379394481229e-06,
"loss": 0.1573,
"step": 719
},
{
"epoch": 1.5867768595041323,
"grad_norm": 4.079885310824538,
"learning_rate": 2.837463551794691e-06,
"loss": 0.3143,
"step": 720
},
{
"epoch": 1.5889807162534435,
"grad_norm": 4.290043095300846,
"learning_rate": 2.8369869977697927e-06,
"loss": 0.1397,
"step": 721
},
{
"epoch": 1.5911845730027547,
"grad_norm": 3.096289822814843,
"learning_rate": 2.8365097862825516e-06,
"loss": 0.0809,
"step": 722
},
{
"epoch": 1.593388429752066,
"grad_norm": 3.9955918707915945,
"learning_rate": 2.836031917567636e-06,
"loss": 0.1987,
"step": 723
},
{
"epoch": 1.5955922865013774,
"grad_norm": 5.319859877967437,
"learning_rate": 2.835553391860036e-06,
"loss": 0.1363,
"step": 724
},
{
"epoch": 1.5977961432506889,
"grad_norm": 5.65396766215741,
"learning_rate": 2.8350742093950677e-06,
"loss": 0.2859,
"step": 725
},
{
"epoch": 1.6,
"grad_norm": 7.765991667810481,
"learning_rate": 2.8345943704083686e-06,
"loss": 0.246,
"step": 726
},
{
"epoch": 1.6022038567493113,
"grad_norm": 2.583152548427552,
"learning_rate": 2.834113875135898e-06,
"loss": 0.1196,
"step": 727
},
{
"epoch": 1.6044077134986225,
"grad_norm": 3.9093087783630867,
"learning_rate": 2.8336327238139407e-06,
"loss": 0.2921,
"step": 728
},
{
"epoch": 1.6066115702479338,
"grad_norm": 3.4398541053621354,
"learning_rate": 2.8331509166791015e-06,
"loss": 0.1223,
"step": 729
},
{
"epoch": 1.6088154269972452,
"grad_norm": 3.8423373441579836,
"learning_rate": 2.832668453968309e-06,
"loss": 0.234,
"step": 730
},
{
"epoch": 1.6110192837465565,
"grad_norm": 4.044067317705418,
"learning_rate": 2.832185335918813e-06,
"loss": 0.1573,
"step": 731
},
{
"epoch": 1.613223140495868,
"grad_norm": 2.8739555688196394,
"learning_rate": 2.8317015627681876e-06,
"loss": 0.1728,
"step": 732
},
{
"epoch": 1.6154269972451791,
"grad_norm": 4.298425879403751,
"learning_rate": 2.8312171347543275e-06,
"loss": 0.1456,
"step": 733
},
{
"epoch": 1.6176308539944904,
"grad_norm": 4.590580736782103,
"learning_rate": 2.83073205211545e-06,
"loss": 0.2604,
"step": 734
},
{
"epoch": 1.6198347107438016,
"grad_norm": 5.5527106677455595,
"learning_rate": 2.830246315090094e-06,
"loss": 0.337,
"step": 735
},
{
"epoch": 1.6220385674931128,
"grad_norm": 4.1351188294540835,
"learning_rate": 2.82975992391712e-06,
"loss": 0.2419,
"step": 736
},
{
"epoch": 1.6242424242424243,
"grad_norm": 4.644664912832784,
"learning_rate": 2.8292728788357105e-06,
"loss": 0.3959,
"step": 737
},
{
"epoch": 1.6264462809917355,
"grad_norm": 3.556852835265641,
"learning_rate": 2.82878518008537e-06,
"loss": 0.1947,
"step": 738
},
{
"epoch": 1.628650137741047,
"grad_norm": 3.630483228818893,
"learning_rate": 2.8282968279059235e-06,
"loss": 0.0405,
"step": 739
},
{
"epoch": 1.6308539944903582,
"grad_norm": 4.745386954667148,
"learning_rate": 2.8278078225375183e-06,
"loss": 0.1604,
"step": 740
},
{
"epoch": 1.6330578512396694,
"grad_norm": 3.53832665010997,
"learning_rate": 2.827318164220622e-06,
"loss": 0.2007,
"step": 741
},
{
"epoch": 1.6352617079889806,
"grad_norm": 5.867754249401183,
"learning_rate": 2.826827853196024e-06,
"loss": 0.1423,
"step": 742
},
{
"epoch": 1.6374655647382919,
"grad_norm": 4.222754274583092,
"learning_rate": 2.8263368897048346e-06,
"loss": 0.1337,
"step": 743
},
{
"epoch": 1.6396694214876033,
"grad_norm": 3.117149882374395,
"learning_rate": 2.825845273988484e-06,
"loss": 0.1453,
"step": 744
},
{
"epoch": 1.6418732782369148,
"grad_norm": 3.4038444714885077,
"learning_rate": 2.8253530062887244e-06,
"loss": 0.0563,
"step": 745
},
{
"epoch": 1.644077134986226,
"grad_norm": 4.623079936470395,
"learning_rate": 2.8248600868476277e-06,
"loss": 0.1519,
"step": 746
},
{
"epoch": 1.6462809917355372,
"grad_norm": 3.313018135102592,
"learning_rate": 2.824366515907587e-06,
"loss": 0.2094,
"step": 747
},
{
"epoch": 1.6484848484848484,
"grad_norm": 5.2839049379385905,
"learning_rate": 2.8238722937113153e-06,
"loss": 0.1547,
"step": 748
},
{
"epoch": 1.6506887052341597,
"grad_norm": 2.9477242559704817,
"learning_rate": 2.823377420501846e-06,
"loss": 0.093,
"step": 749
},
{
"epoch": 1.6528925619834711,
"grad_norm": 3.045171763549146,
"learning_rate": 2.8228818965225326e-06,
"loss": 0.08,
"step": 750
},
{
"epoch": 1.6550964187327823,
"grad_norm": 3.923253731083525,
"learning_rate": 2.8223857220170486e-06,
"loss": 0.1194,
"step": 751
},
{
"epoch": 1.6573002754820938,
"grad_norm": 3.4924421442319726,
"learning_rate": 2.821888897229388e-06,
"loss": 0.1569,
"step": 752
},
{
"epoch": 1.659504132231405,
"grad_norm": 10.507784482223977,
"learning_rate": 2.821391422403863e-06,
"loss": 0.2523,
"step": 753
},
{
"epoch": 1.6617079889807163,
"grad_norm": 6.827016679434054,
"learning_rate": 2.820893297785107e-06,
"loss": 0.1597,
"step": 754
},
{
"epoch": 1.6639118457300275,
"grad_norm": 4.231118216733152,
"learning_rate": 2.8203945236180726e-06,
"loss": 0.2601,
"step": 755
},
{
"epoch": 1.6661157024793387,
"grad_norm": 5.790045882163988,
"learning_rate": 2.8198951001480316e-06,
"loss": 0.1201,
"step": 756
},
{
"epoch": 1.6683195592286502,
"grad_norm": 3.780198445899582,
"learning_rate": 2.8193950276205744e-06,
"loss": 0.3059,
"step": 757
},
{
"epoch": 1.6705234159779614,
"grad_norm": 2.7743504550344724,
"learning_rate": 2.818894306281612e-06,
"loss": 0.0944,
"step": 758
},
{
"epoch": 1.6727272727272728,
"grad_norm": 4.384995400856117,
"learning_rate": 2.818392936377374e-06,
"loss": 0.207,
"step": 759
},
{
"epoch": 1.674931129476584,
"grad_norm": 3.46546366145472,
"learning_rate": 2.817890918154408e-06,
"loss": 0.0987,
"step": 760
},
{
"epoch": 1.6771349862258953,
"grad_norm": 3.944519624313324,
"learning_rate": 2.8173882518595806e-06,
"loss": 0.2076,
"step": 761
},
{
"epoch": 1.6793388429752065,
"grad_norm": 3.0199025674864197,
"learning_rate": 2.816884937740079e-06,
"loss": 0.139,
"step": 762
},
{
"epoch": 1.6815426997245178,
"grad_norm": 4.181192024696917,
"learning_rate": 2.8163809760434065e-06,
"loss": 0.4948,
"step": 763
},
{
"epoch": 1.6837465564738292,
"grad_norm": 5.425368792194891,
"learning_rate": 2.8158763670173866e-06,
"loss": 0.2613,
"step": 764
},
{
"epoch": 1.6859504132231407,
"grad_norm": 2.449214805033337,
"learning_rate": 2.8153711109101596e-06,
"loss": 0.1652,
"step": 765
},
{
"epoch": 1.6881542699724519,
"grad_norm": 3.02375338979952,
"learning_rate": 2.8148652079701852e-06,
"loss": 0.0925,
"step": 766
},
{
"epoch": 1.690358126721763,
"grad_norm": 3.484485189001616,
"learning_rate": 2.814358658446241e-06,
"loss": 0.1745,
"step": 767
},
{
"epoch": 1.6925619834710743,
"grad_norm": 2.666491627882498,
"learning_rate": 2.813851462587422e-06,
"loss": 0.1677,
"step": 768
},
{
"epoch": 1.6947658402203856,
"grad_norm": 3.1016763666636717,
"learning_rate": 2.8133436206431416e-06,
"loss": 0.0979,
"step": 769
},
{
"epoch": 1.696969696969697,
"grad_norm": 2.6268324993637737,
"learning_rate": 2.8128351328631308e-06,
"loss": 0.1548,
"step": 770
},
{
"epoch": 1.6991735537190082,
"grad_norm": 3.5260065593405403,
"learning_rate": 2.8123259994974376e-06,
"loss": 0.2974,
"step": 771
},
{
"epoch": 1.7013774104683197,
"grad_norm": 4.976441509919316,
"learning_rate": 2.811816220796428e-06,
"loss": 0.2308,
"step": 772
},
{
"epoch": 1.703581267217631,
"grad_norm": 5.669812148443133,
"learning_rate": 2.8113057970107864e-06,
"loss": 0.5012,
"step": 773
},
{
"epoch": 1.7057851239669422,
"grad_norm": 5.4074181108538895,
"learning_rate": 2.8107947283915115e-06,
"loss": 0.2426,
"step": 774
},
{
"epoch": 1.7079889807162534,
"grad_norm": 4.925146764134966,
"learning_rate": 2.810283015189922e-06,
"loss": 0.2084,
"step": 775
},
{
"epoch": 1.7101928374655646,
"grad_norm": 4.733811645399949,
"learning_rate": 2.8097706576576524e-06,
"loss": 0.1098,
"step": 776
},
{
"epoch": 1.712396694214876,
"grad_norm": 3.72787024918664,
"learning_rate": 2.8092576560466546e-06,
"loss": 0.2889,
"step": 777
},
{
"epoch": 1.7146005509641873,
"grad_norm": 4.179092902335767,
"learning_rate": 2.808744010609196e-06,
"loss": 0.171,
"step": 778
},
{
"epoch": 1.7168044077134987,
"grad_norm": 3.1698363233858435,
"learning_rate": 2.808229721597861e-06,
"loss": 0.102,
"step": 779
},
{
"epoch": 1.71900826446281,
"grad_norm": 5.330585067718199,
"learning_rate": 2.8077147892655516e-06,
"loss": 0.4345,
"step": 780
},
{
"epoch": 1.7212121212121212,
"grad_norm": 3.9967585194484263,
"learning_rate": 2.8071992138654856e-06,
"loss": 0.1291,
"step": 781
},
{
"epoch": 1.7234159779614324,
"grad_norm": 1.922265729326589,
"learning_rate": 2.806682995651196e-06,
"loss": 0.1489,
"step": 782
},
{
"epoch": 1.7256198347107436,
"grad_norm": 3.133050865277451,
"learning_rate": 2.806166134876533e-06,
"loss": 0.1107,
"step": 783
},
{
"epoch": 1.727823691460055,
"grad_norm": 4.8008500192145735,
"learning_rate": 2.805648631795663e-06,
"loss": 0.1931,
"step": 784
},
{
"epoch": 1.7300275482093666,
"grad_norm": 7.27694610184454,
"learning_rate": 2.8051304866630674e-06,
"loss": 0.1939,
"step": 785
},
{
"epoch": 1.7322314049586778,
"grad_norm": 4.683082050560034,
"learning_rate": 2.8046116997335432e-06,
"loss": 0.1368,
"step": 786
},
{
"epoch": 1.734435261707989,
"grad_norm": 4.192820835260911,
"learning_rate": 2.804092271262204e-06,
"loss": 0.1852,
"step": 787
},
{
"epoch": 1.7366391184573002,
"grad_norm": 2.654337382564104,
"learning_rate": 2.8035722015044784e-06,
"loss": 0.0936,
"step": 788
},
{
"epoch": 1.7388429752066115,
"grad_norm": 4.735087099788159,
"learning_rate": 2.80305149071611e-06,
"loss": 0.1789,
"step": 789
},
{
"epoch": 1.741046831955923,
"grad_norm": 4.54086308882895,
"learning_rate": 2.8025301391531584e-06,
"loss": 0.1054,
"step": 790
},
{
"epoch": 1.7432506887052341,
"grad_norm": 5.877262840590166,
"learning_rate": 2.8020081470719974e-06,
"loss": 0.1144,
"step": 791
},
{
"epoch": 1.7454545454545456,
"grad_norm": 2.9934758004327198,
"learning_rate": 2.8014855147293165e-06,
"loss": 0.1114,
"step": 792
},
{
"epoch": 1.7476584022038568,
"grad_norm": 3.9602949185225427,
"learning_rate": 2.800962242382119e-06,
"loss": 0.1469,
"step": 793
},
{
"epoch": 1.749862258953168,
"grad_norm": 3.4891990751120874,
"learning_rate": 2.8004383302877246e-06,
"loss": 0.2197,
"step": 794
},
{
"epoch": 1.7520661157024793,
"grad_norm": 3.1228880918415314,
"learning_rate": 2.799913778703767e-06,
"loss": 0.0802,
"step": 795
},
{
"epoch": 1.7542699724517905,
"grad_norm": 4.262402531909565,
"learning_rate": 2.7993885878881924e-06,
"loss": 0.1462,
"step": 796
},
{
"epoch": 1.756473829201102,
"grad_norm": 4.464490083715167,
"learning_rate": 2.7988627580992644e-06,
"loss": 0.3734,
"step": 797
},
{
"epoch": 1.7586776859504132,
"grad_norm": 2.9521520450784293,
"learning_rate": 2.7983362895955592e-06,
"loss": 0.0761,
"step": 798
},
{
"epoch": 1.7608815426997246,
"grad_norm": 4.569388820068124,
"learning_rate": 2.797809182635967e-06,
"loss": 0.311,
"step": 799
},
{
"epoch": 1.7630853994490359,
"grad_norm": 2.761531061435357,
"learning_rate": 2.7972814374796923e-06,
"loss": 0.1123,
"step": 800
},
{
"epoch": 1.765289256198347,
"grad_norm": 2.909256889094545,
"learning_rate": 2.7967530543862536e-06,
"loss": 0.1424,
"step": 801
},
{
"epoch": 1.7674931129476583,
"grad_norm": 3.580892106221509,
"learning_rate": 2.7962240336154824e-06,
"loss": 0.1713,
"step": 802
},
{
"epoch": 1.7696969696969695,
"grad_norm": 2.613119863195253,
"learning_rate": 2.795694375427524e-06,
"loss": 0.1453,
"step": 803
},
{
"epoch": 1.771900826446281,
"grad_norm": 3.9385995694287512,
"learning_rate": 2.795164080082839e-06,
"loss": 0.3253,
"step": 804
},
{
"epoch": 1.7741046831955924,
"grad_norm": 2.968556000386392,
"learning_rate": 2.7946331478421985e-06,
"loss": 0.1914,
"step": 805
},
{
"epoch": 1.7763085399449037,
"grad_norm": 4.2982285708776615,
"learning_rate": 2.794101578966688e-06,
"loss": 0.1755,
"step": 806
},
{
"epoch": 1.778512396694215,
"grad_norm": 3.077422083118147,
"learning_rate": 2.793569373717706e-06,
"loss": 0.1018,
"step": 807
},
{
"epoch": 1.7807162534435261,
"grad_norm": 3.2720302498627394,
"learning_rate": 2.7930365323569652e-06,
"loss": 0.0902,
"step": 808
},
{
"epoch": 1.7829201101928374,
"grad_norm": 4.579436283539023,
"learning_rate": 2.7925030551464885e-06,
"loss": 0.1652,
"step": 809
},
{
"epoch": 1.7851239669421488,
"grad_norm": 3.7523711022286372,
"learning_rate": 2.7919689423486137e-06,
"loss": 0.1149,
"step": 810
},
{
"epoch": 1.78732782369146,
"grad_norm": 2.1787023560029017,
"learning_rate": 2.79143419422599e-06,
"loss": 0.1034,
"step": 811
},
{
"epoch": 1.7895316804407715,
"grad_norm": 3.262576182854977,
"learning_rate": 2.7908988110415794e-06,
"loss": 0.1919,
"step": 812
},
{
"epoch": 1.7917355371900827,
"grad_norm": 2.768827717118814,
"learning_rate": 2.7903627930586566e-06,
"loss": 0.0913,
"step": 813
},
{
"epoch": 1.793939393939394,
"grad_norm": 4.020492031322422,
"learning_rate": 2.789826140540808e-06,
"loss": 0.112,
"step": 814
},
{
"epoch": 1.7961432506887052,
"grad_norm": 3.8046638565500457,
"learning_rate": 2.7892888537519313e-06,
"loss": 0.2014,
"step": 815
},
{
"epoch": 1.7983471074380164,
"grad_norm": 3.5265605684044727,
"learning_rate": 2.788750932956238e-06,
"loss": 0.1362,
"step": 816
},
{
"epoch": 1.8005509641873279,
"grad_norm": 4.629808595887051,
"learning_rate": 2.7882123784182493e-06,
"loss": 0.1114,
"step": 817
},
{
"epoch": 1.802754820936639,
"grad_norm": 6.514243778597918,
"learning_rate": 2.7876731904027993e-06,
"loss": 0.1641,
"step": 818
},
{
"epoch": 1.8049586776859505,
"grad_norm": 3.1215686853622335,
"learning_rate": 2.787133369175033e-06,
"loss": 0.2913,
"step": 819
},
{
"epoch": 1.8071625344352618,
"grad_norm": 3.6554868120588466,
"learning_rate": 2.7865929150004082e-06,
"loss": 0.1416,
"step": 820
},
{
"epoch": 1.809366391184573,
"grad_norm": 15.691489907553818,
"learning_rate": 2.7860518281446915e-06,
"loss": 0.1736,
"step": 821
},
{
"epoch": 1.8115702479338842,
"grad_norm": 5.283164839932474,
"learning_rate": 2.7855101088739628e-06,
"loss": 0.2026,
"step": 822
},
{
"epoch": 1.8137741046831954,
"grad_norm": 3.5594046840956057,
"learning_rate": 2.7849677574546113e-06,
"loss": 0.0848,
"step": 823
},
{
"epoch": 1.815977961432507,
"grad_norm": 4.612036787809661,
"learning_rate": 2.7844247741533395e-06,
"loss": 0.179,
"step": 824
},
{
"epoch": 1.8181818181818183,
"grad_norm": 3.65401545072593,
"learning_rate": 2.783881159237157e-06,
"loss": 0.1717,
"step": 825
},
{
"epoch": 1.8203856749311296,
"grad_norm": 3.9012328429874294,
"learning_rate": 2.783336912973387e-06,
"loss": 0.2841,
"step": 826
},
{
"epoch": 1.8225895316804408,
"grad_norm": 5.965848322247876,
"learning_rate": 2.782792035629663e-06,
"loss": 0.3736,
"step": 827
},
{
"epoch": 1.824793388429752,
"grad_norm": 3.271989028222971,
"learning_rate": 2.7822465274739265e-06,
"loss": 0.1078,
"step": 828
},
{
"epoch": 1.8269972451790633,
"grad_norm": 5.339902990752521,
"learning_rate": 2.7817003887744313e-06,
"loss": 0.3349,
"step": 829
},
{
"epoch": 1.8292011019283747,
"grad_norm": 3.0715961642404577,
"learning_rate": 2.781153619799741e-06,
"loss": 0.1475,
"step": 830
},
{
"epoch": 1.831404958677686,
"grad_norm": 5.6237694593788845,
"learning_rate": 2.780606220818729e-06,
"loss": 0.1436,
"step": 831
},
{
"epoch": 1.8336088154269974,
"grad_norm": 5.890345413594224,
"learning_rate": 2.7800581921005778e-06,
"loss": 0.1878,
"step": 832
},
{
"epoch": 1.8358126721763086,
"grad_norm": 4.06712704576657,
"learning_rate": 2.77950953391478e-06,
"loss": 0.2348,
"step": 833
},
{
"epoch": 1.8380165289256198,
"grad_norm": 2.882404649474459,
"learning_rate": 2.7789602465311384e-06,
"loss": 0.1351,
"step": 834
},
{
"epoch": 1.840220385674931,
"grad_norm": 5.473781561460339,
"learning_rate": 2.778410330219764e-06,
"loss": 0.3063,
"step": 835
},
{
"epoch": 1.8424242424242423,
"grad_norm": 3.922211179254337,
"learning_rate": 2.777859785251079e-06,
"loss": 0.1619,
"step": 836
},
{
"epoch": 1.8446280991735537,
"grad_norm": 8.221310063208145,
"learning_rate": 2.777308611895812e-06,
"loss": 0.4785,
"step": 837
},
{
"epoch": 1.846831955922865,
"grad_norm": 4.2268861917345335,
"learning_rate": 2.7767568104250022e-06,
"loss": 0.1154,
"step": 838
},
{
"epoch": 1.8490358126721764,
"grad_norm": 2.7566969326407658,
"learning_rate": 2.7762043811099986e-06,
"loss": 0.0647,
"step": 839
},
{
"epoch": 1.8512396694214877,
"grad_norm": 4.1654096542802925,
"learning_rate": 2.775651324222457e-06,
"loss": 0.2954,
"step": 840
},
{
"epoch": 1.8534435261707989,
"grad_norm": 2.862929107741642,
"learning_rate": 2.775097640034343e-06,
"loss": 0.1234,
"step": 841
},
{
"epoch": 1.8556473829201101,
"grad_norm": 4.035924471619375,
"learning_rate": 2.7745433288179313e-06,
"loss": 0.2158,
"step": 842
},
{
"epoch": 1.8578512396694213,
"grad_norm": 3.9005408283862884,
"learning_rate": 2.773988390845802e-06,
"loss": 0.2078,
"step": 843
},
{
"epoch": 1.8600550964187328,
"grad_norm": 4.0776857291236,
"learning_rate": 2.773432826390847e-06,
"loss": 0.2066,
"step": 844
},
{
"epoch": 1.8622589531680442,
"grad_norm": 4.054788798271851,
"learning_rate": 2.7728766357262645e-06,
"loss": 0.1977,
"step": 845
},
{
"epoch": 1.8644628099173555,
"grad_norm": 4.8154226644199,
"learning_rate": 2.77231981912556e-06,
"loss": 0.2452,
"step": 846
},
{
"epoch": 1.8666666666666667,
"grad_norm": 3.749035576393667,
"learning_rate": 2.7717623768625487e-06,
"loss": 0.1679,
"step": 847
},
{
"epoch": 1.868870523415978,
"grad_norm": 3.0296230257574823,
"learning_rate": 2.7712043092113518e-06,
"loss": 0.1371,
"step": 848
},
{
"epoch": 1.8710743801652892,
"grad_norm": 2.8961077120026397,
"learning_rate": 2.7706456164463987e-06,
"loss": 0.1451,
"step": 849
},
{
"epoch": 1.8732782369146006,
"grad_norm": 4.757014237158706,
"learning_rate": 2.7700862988424264e-06,
"loss": 0.2476,
"step": 850
},
{
"epoch": 1.8754820936639118,
"grad_norm": 3.5605409135922486,
"learning_rate": 2.7695263566744786e-06,
"loss": 0.3223,
"step": 851
},
{
"epoch": 1.8776859504132233,
"grad_norm": 3.673778148979456,
"learning_rate": 2.7689657902179063e-06,
"loss": 0.1496,
"step": 852
},
{
"epoch": 1.8798898071625345,
"grad_norm": 3.9594639471733015,
"learning_rate": 2.7684045997483683e-06,
"loss": 0.1621,
"step": 853
},
{
"epoch": 1.8820936639118457,
"grad_norm": 4.3094578613213015,
"learning_rate": 2.7678427855418293e-06,
"loss": 0.1284,
"step": 854
},
{
"epoch": 1.884297520661157,
"grad_norm": 4.230139062066988,
"learning_rate": 2.767280347874561e-06,
"loss": 0.2194,
"step": 855
},
{
"epoch": 1.8865013774104682,
"grad_norm": 4.483162054895854,
"learning_rate": 2.766717287023141e-06,
"loss": 0.3231,
"step": 856
},
{
"epoch": 1.8887052341597796,
"grad_norm": 3.6045893200147656,
"learning_rate": 2.766153603264455e-06,
"loss": 0.3204,
"step": 857
},
{
"epoch": 1.8909090909090909,
"grad_norm": 2.68825864298539,
"learning_rate": 2.7655892968756935e-06,
"loss": 0.1282,
"step": 858
},
{
"epoch": 1.8931129476584023,
"grad_norm": 3.086935863096946,
"learning_rate": 2.7650243681343543e-06,
"loss": 0.1713,
"step": 859
},
{
"epoch": 1.8953168044077136,
"grad_norm": 4.285553555539189,
"learning_rate": 2.76445881731824e-06,
"loss": 0.1801,
"step": 860
},
{
"epoch": 1.8975206611570248,
"grad_norm": 4.032253713736677,
"learning_rate": 2.76389264470546e-06,
"loss": 0.1143,
"step": 861
},
{
"epoch": 1.899724517906336,
"grad_norm": 3.5305542226645032,
"learning_rate": 2.76332585057443e-06,
"loss": 0.2464,
"step": 862
},
{
"epoch": 1.9019283746556472,
"grad_norm": 4.04266297597818,
"learning_rate": 2.762758435203869e-06,
"loss": 0.1271,
"step": 863
},
{
"epoch": 1.9041322314049587,
"grad_norm": 3.046350178355839,
"learning_rate": 2.762190398872805e-06,
"loss": 0.1708,
"step": 864
},
{
"epoch": 1.9063360881542701,
"grad_norm": 4.39744290609223,
"learning_rate": 2.7616217418605685e-06,
"loss": 0.0867,
"step": 865
},
{
"epoch": 1.9085399449035814,
"grad_norm": 3.9606892297824627,
"learning_rate": 2.7610524644467955e-06,
"loss": 0.1766,
"step": 866
},
{
"epoch": 1.9107438016528926,
"grad_norm": 2.6007521338428163,
"learning_rate": 2.7604825669114284e-06,
"loss": 0.1322,
"step": 867
},
{
"epoch": 1.9129476584022038,
"grad_norm": 4.5370637771574485,
"learning_rate": 2.7599120495347137e-06,
"loss": 0.2271,
"step": 868
},
{
"epoch": 1.915151515151515,
"grad_norm": 4.704518903323181,
"learning_rate": 2.7593409125972037e-06,
"loss": 0.2211,
"step": 869
},
{
"epoch": 1.9173553719008265,
"grad_norm": 3.7308347809758704,
"learning_rate": 2.7587691563797533e-06,
"loss": 0.1474,
"step": 870
},
{
"epoch": 1.9195592286501377,
"grad_norm": 3.531251299278127,
"learning_rate": 2.758196781163524e-06,
"loss": 0.1425,
"step": 871
},
{
"epoch": 1.9217630853994492,
"grad_norm": 3.899278122311197,
"learning_rate": 2.7576237872299805e-06,
"loss": 0.1692,
"step": 872
},
{
"epoch": 1.9239669421487604,
"grad_norm": 3.970267403444949,
"learning_rate": 2.7570501748608922e-06,
"loss": 0.3354,
"step": 873
},
{
"epoch": 1.9261707988980716,
"grad_norm": 5.657505689245682,
"learning_rate": 2.756475944338333e-06,
"loss": 0.2707,
"step": 874
},
{
"epoch": 1.9283746556473829,
"grad_norm": 5.4080580645655125,
"learning_rate": 2.75590109594468e-06,
"loss": 0.2196,
"step": 875
},
{
"epoch": 1.930578512396694,
"grad_norm": 4.072394021014495,
"learning_rate": 2.7553256299626138e-06,
"loss": 0.2698,
"step": 876
},
{
"epoch": 1.9327823691460055,
"grad_norm": 2.667949896246189,
"learning_rate": 2.7547495466751204e-06,
"loss": 0.1677,
"step": 877
},
{
"epoch": 1.9349862258953168,
"grad_norm": 3.638205070981609,
"learning_rate": 2.754172846365488e-06,
"loss": 0.1484,
"step": 878
},
{
"epoch": 1.9371900826446282,
"grad_norm": 3.4775529494976136,
"learning_rate": 2.7535955293173087e-06,
"loss": 0.1624,
"step": 879
},
{
"epoch": 1.9393939393939394,
"grad_norm": 4.173548390099768,
"learning_rate": 2.753017595814477e-06,
"loss": 0.1607,
"step": 880
},
{
"epoch": 1.9415977961432507,
"grad_norm": 4.164962747564953,
"learning_rate": 2.7524390461411916e-06,
"loss": 0.1889,
"step": 881
},
{
"epoch": 1.943801652892562,
"grad_norm": 2.54142311710031,
"learning_rate": 2.751859880581954e-06,
"loss": 0.1741,
"step": 882
},
{
"epoch": 1.9460055096418731,
"grad_norm": 2.178677072386189,
"learning_rate": 2.751280099421569e-06,
"loss": 0.1216,
"step": 883
},
{
"epoch": 1.9482093663911846,
"grad_norm": 3.8814007175409673,
"learning_rate": 2.750699702945142e-06,
"loss": 0.1891,
"step": 884
},
{
"epoch": 1.950413223140496,
"grad_norm": 8.658086942729867,
"learning_rate": 2.7501186914380845e-06,
"loss": 0.1778,
"step": 885
},
{
"epoch": 1.9526170798898073,
"grad_norm": 3.146576269483999,
"learning_rate": 2.749537065186107e-06,
"loss": 0.0524,
"step": 886
},
{
"epoch": 1.9548209366391185,
"grad_norm": 3.2325353182101395,
"learning_rate": 2.748954824475224e-06,
"loss": 0.1623,
"step": 887
},
{
"epoch": 1.9570247933884297,
"grad_norm": 3.1057019071534047,
"learning_rate": 2.748371969591752e-06,
"loss": 0.1908,
"step": 888
},
{
"epoch": 1.959228650137741,
"grad_norm": 2.678469488136285,
"learning_rate": 2.74778850082231e-06,
"loss": 0.1876,
"step": 889
},
{
"epoch": 1.9614325068870524,
"grad_norm": 4.750733810300038,
"learning_rate": 2.747204418453818e-06,
"loss": 0.2796,
"step": 890
},
{
"epoch": 1.9636363636363636,
"grad_norm": 2.4582854680750486,
"learning_rate": 2.7466197227734976e-06,
"loss": 0.0621,
"step": 891
},
{
"epoch": 1.965840220385675,
"grad_norm": 4.052676767921253,
"learning_rate": 2.7460344140688733e-06,
"loss": 0.238,
"step": 892
},
{
"epoch": 1.9680440771349863,
"grad_norm": 3.0237993476789753,
"learning_rate": 2.7454484926277693e-06,
"loss": 0.1915,
"step": 893
},
{
"epoch": 1.9702479338842975,
"grad_norm": 2.989473268610309,
"learning_rate": 2.744861958738313e-06,
"loss": 0.0917,
"step": 894
},
{
"epoch": 1.9724517906336088,
"grad_norm": 2.36311627004035,
"learning_rate": 2.744274812688931e-06,
"loss": 0.1871,
"step": 895
},
{
"epoch": 1.97465564738292,
"grad_norm": 3.4871528972576944,
"learning_rate": 2.743687054768353e-06,
"loss": 0.1615,
"step": 896
},
{
"epoch": 1.9768595041322314,
"grad_norm": 3.7392756691831246,
"learning_rate": 2.743098685265607e-06,
"loss": 0.2817,
"step": 897
},
{
"epoch": 1.9790633608815427,
"grad_norm": 4.617396209400955,
"learning_rate": 2.7425097044700246e-06,
"loss": 0.235,
"step": 898
},
{
"epoch": 1.9812672176308541,
"grad_norm": 3.679289220007701,
"learning_rate": 2.741920112671237e-06,
"loss": 0.1242,
"step": 899
},
{
"epoch": 1.9834710743801653,
"grad_norm": 3.3171681919567293,
"learning_rate": 2.7413299101591738e-06,
"loss": 0.1351,
"step": 900
}
],
"logging_steps": 1,
"max_steps": 4530,
"num_input_tokens_seen": 0,
"num_train_epochs": 10,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 834641013964800.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}