Files
Qwen3-1.7B-Base-Openthought…/checkpoint-700/trainer_state.json

4935 lines
128 KiB
JSON
Raw Normal View History

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9375523187677883,
"eval_steps": 500,
"global_step": 700,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0013393604553825547,
"grad_norm": 7.59987480034225,
"learning_rate": 0.0,
"loss": 1.2203794717788696,
"step": 1
},
{
"epoch": 0.0026787209107651095,
"grad_norm": 7.556419191633864,
"learning_rate": 1.066666666666667e-06,
"loss": 1.2077422142028809,
"step": 2
},
{
"epoch": 0.004018081366147665,
"grad_norm": 7.43424441190293,
"learning_rate": 2.133333333333334e-06,
"loss": 1.2097434997558594,
"step": 3
},
{
"epoch": 0.005357441821530219,
"grad_norm": 7.021666134821503,
"learning_rate": 3.2000000000000003e-06,
"loss": 1.1891963481903076,
"step": 4
},
{
"epoch": 0.006696802276912774,
"grad_norm": 5.446482336710434,
"learning_rate": 4.266666666666668e-06,
"loss": 1.1889076232910156,
"step": 5
},
{
"epoch": 0.00803616273229533,
"grad_norm": 3.3854892179134715,
"learning_rate": 5.333333333333334e-06,
"loss": 1.1870930194854736,
"step": 6
},
{
"epoch": 0.009375523187677884,
"grad_norm": 3.8522461652029576,
"learning_rate": 6.4000000000000006e-06,
"loss": 1.166532278060913,
"step": 7
},
{
"epoch": 0.010714883643060438,
"grad_norm": 4.114885499626225,
"learning_rate": 7.4666666666666675e-06,
"loss": 1.1786075830459595,
"step": 8
},
{
"epoch": 0.012054244098442994,
"grad_norm": 3.7848771870154274,
"learning_rate": 8.533333333333335e-06,
"loss": 1.1633087396621704,
"step": 9
},
{
"epoch": 0.013393604553825548,
"grad_norm": 3.480585429874886,
"learning_rate": 9.600000000000001e-06,
"loss": 1.1381144523620605,
"step": 10
},
{
"epoch": 0.014732965009208104,
"grad_norm": 3.0527912777786415,
"learning_rate": 1.0666666666666667e-05,
"loss": 1.130035161972046,
"step": 11
},
{
"epoch": 0.01607232546459066,
"grad_norm": 2.1785228593328774,
"learning_rate": 1.1733333333333335e-05,
"loss": 1.113619327545166,
"step": 12
},
{
"epoch": 0.017411685919973212,
"grad_norm": 2.153276697775204,
"learning_rate": 1.2800000000000001e-05,
"loss": 1.1094508171081543,
"step": 13
},
{
"epoch": 0.018751046375355768,
"grad_norm": 2.1639832023027568,
"learning_rate": 1.3866666666666669e-05,
"loss": 1.1093754768371582,
"step": 14
},
{
"epoch": 0.020090406830738324,
"grad_norm": 1.5819539807956415,
"learning_rate": 1.4933333333333335e-05,
"loss": 1.0970914363861084,
"step": 15
},
{
"epoch": 0.021429767286120876,
"grad_norm": 0.9981596182699652,
"learning_rate": 1.6000000000000003e-05,
"loss": 1.0732706785202026,
"step": 16
},
{
"epoch": 0.02276912774150343,
"grad_norm": 1.150670436099805,
"learning_rate": 1.706666666666667e-05,
"loss": 1.067190408706665,
"step": 17
},
{
"epoch": 0.024108488196885988,
"grad_norm": 1.241573069994366,
"learning_rate": 1.8133333333333335e-05,
"loss": 1.0586261749267578,
"step": 18
},
{
"epoch": 0.025447848652268543,
"grad_norm": 1.0318839224030094,
"learning_rate": 1.9200000000000003e-05,
"loss": 1.065382957458496,
"step": 19
},
{
"epoch": 0.026787209107651096,
"grad_norm": 0.7926548431099205,
"learning_rate": 2.026666666666667e-05,
"loss": 1.0423641204833984,
"step": 20
},
{
"epoch": 0.02812656956303365,
"grad_norm": 0.8651245571596619,
"learning_rate": 2.1333333333333335e-05,
"loss": 1.0538524389266968,
"step": 21
},
{
"epoch": 0.029465930018416207,
"grad_norm": 0.7773713907984426,
"learning_rate": 2.2400000000000002e-05,
"loss": 1.050022840499878,
"step": 22
},
{
"epoch": 0.03080529047379876,
"grad_norm": 0.6131435887086818,
"learning_rate": 2.346666666666667e-05,
"loss": 1.0374431610107422,
"step": 23
},
{
"epoch": 0.03214465092918132,
"grad_norm": 0.6411600233078315,
"learning_rate": 2.4533333333333334e-05,
"loss": 1.0369596481323242,
"step": 24
},
{
"epoch": 0.03348401138456387,
"grad_norm": 0.6105825840662947,
"learning_rate": 2.5600000000000002e-05,
"loss": 1.0248405933380127,
"step": 25
},
{
"epoch": 0.034823371839946424,
"grad_norm": 0.5514039178404373,
"learning_rate": 2.6666666666666667e-05,
"loss": 1.0261945724487305,
"step": 26
},
{
"epoch": 0.03616273229532898,
"grad_norm": 0.47980911346053196,
"learning_rate": 2.7733333333333338e-05,
"loss": 1.0281734466552734,
"step": 27
},
{
"epoch": 0.037502092750711535,
"grad_norm": 0.5015032388738222,
"learning_rate": 2.8800000000000002e-05,
"loss": 1.0181447267532349,
"step": 28
},
{
"epoch": 0.03884145320609409,
"grad_norm": 0.4412925876210946,
"learning_rate": 2.986666666666667e-05,
"loss": 1.0037535429000854,
"step": 29
},
{
"epoch": 0.04018081366147665,
"grad_norm": 0.4791162369947075,
"learning_rate": 3.093333333333334e-05,
"loss": 1.0166124105453491,
"step": 30
},
{
"epoch": 0.0415201741168592,
"grad_norm": 0.38218322408541294,
"learning_rate": 3.2000000000000005e-05,
"loss": 1.015824794769287,
"step": 31
},
{
"epoch": 0.04285953457224175,
"grad_norm": 0.4068277079215438,
"learning_rate": 3.3066666666666666e-05,
"loss": 0.9952138662338257,
"step": 32
},
{
"epoch": 0.04419889502762431,
"grad_norm": 0.41007437644363576,
"learning_rate": 3.413333333333334e-05,
"loss": 0.9999523162841797,
"step": 33
},
{
"epoch": 0.04553825548300686,
"grad_norm": 0.39463041187020714,
"learning_rate": 3.52e-05,
"loss": 1.0010638236999512,
"step": 34
},
{
"epoch": 0.04687761593838942,
"grad_norm": 0.38858670221650043,
"learning_rate": 3.626666666666667e-05,
"loss": 1.0023632049560547,
"step": 35
},
{
"epoch": 0.048216976393771975,
"grad_norm": 0.42457022462001964,
"learning_rate": 3.733333333333334e-05,
"loss": 0.9928240776062012,
"step": 36
},
{
"epoch": 0.04955633684915453,
"grad_norm": 0.4203545219017498,
"learning_rate": 3.8400000000000005e-05,
"loss": 0.9923948049545288,
"step": 37
},
{
"epoch": 0.05089569730453709,
"grad_norm": 0.6018289451317828,
"learning_rate": 3.946666666666667e-05,
"loss": 0.9856561422348022,
"step": 38
},
{
"epoch": 0.052235057759919636,
"grad_norm": 0.9556415125996957,
"learning_rate": 4.053333333333334e-05,
"loss": 0.9986469745635986,
"step": 39
},
{
"epoch": 0.05357441821530219,
"grad_norm": 1.6335316008464362,
"learning_rate": 4.16e-05,
"loss": 0.9917392730712891,
"step": 40
},
{
"epoch": 0.05491377867068475,
"grad_norm": 0.5854953221201857,
"learning_rate": 4.266666666666667e-05,
"loss": 0.992963433265686,
"step": 41
},
{
"epoch": 0.0562531391260673,
"grad_norm": 0.9145316728664418,
"learning_rate": 4.373333333333334e-05,
"loss": 0.9989960789680481,
"step": 42
},
{
"epoch": 0.05759249958144986,
"grad_norm": 1.2192655126783132,
"learning_rate": 4.4800000000000005e-05,
"loss": 0.9921097159385681,
"step": 43
},
{
"epoch": 0.058931860036832415,
"grad_norm": 0.7759793940795199,
"learning_rate": 4.586666666666667e-05,
"loss": 0.9693036675453186,
"step": 44
},
{
"epoch": 0.06027122049221497,
"grad_norm": 1.0242111422485076,
"learning_rate": 4.693333333333334e-05,
"loss": 0.9812385439872742,
"step": 45
},
{
"epoch": 0.06161058094759752,
"grad_norm": 1.1346825970435035,
"learning_rate": 4.8e-05,
"loss": 0.9808453321456909,
"step": 46
},
{
"epoch": 0.06294994140298008,
"grad_norm": 0.8278266210784355,
"learning_rate": 4.906666666666667e-05,
"loss": 0.9826059937477112,
"step": 47
},
{
"epoch": 0.06428930185836264,
"grad_norm": 0.8341908558108033,
"learning_rate": 5.0133333333333343e-05,
"loss": 0.9739001393318176,
"step": 48
},
{
"epoch": 0.06562866231374519,
"grad_norm": 0.8183375480539544,
"learning_rate": 5.1200000000000004e-05,
"loss": 0.9766189455986023,
"step": 49
},
{
"epoch": 0.06696802276912774,
"grad_norm": 1.1789309992406718,
"learning_rate": 5.226666666666667e-05,
"loss": 0.9653265476226807,
"step": 50
},
{
"epoch": 0.0683073832245103,
"grad_norm": 1.200678628689884,
"learning_rate": 5.333333333333333e-05,
"loss": 0.9755718111991882,
"step": 51
},
{
"epoch": 0.06964674367989285,
"grad_norm": 1.2668282300020823,
"learning_rate": 5.440000000000001e-05,
"loss": 0.9740673899650574,
"step": 52
},
{
"epoch": 0.07098610413527541,
"grad_norm": 1.3742322857772529,
"learning_rate": 5.5466666666666675e-05,
"loss": 0.9595431089401245,
"step": 53
},
{
"epoch": 0.07232546459065796,
"grad_norm": 0.7207560308493244,
"learning_rate": 5.6533333333333336e-05,
"loss": 0.9671750068664551,
"step": 54
},
{
"epoch": 0.07366482504604052,
"grad_norm": 1.5441091916589376,
"learning_rate": 5.7600000000000004e-05,
"loss": 0.9589823484420776,
"step": 55
},
{
"epoch": 0.07500418550142307,
"grad_norm": 0.8387565660519588,
"learning_rate": 5.8666666666666665e-05,
"loss": 0.9685125350952148,
"step": 56
},
{
"epoch": 0.07634354595680562,
"grad_norm": 0.963471962771943,
"learning_rate": 5.973333333333334e-05,
"loss": 0.9614203572273254,
"step": 57
},
{
"epoch": 0.07768290641218818,
"grad_norm": 1.2156734129108198,
"learning_rate": 6.080000000000001e-05,
"loss": 0.9662147164344788,
"step": 58
},
{
"epoch": 0.07902226686757073,
"grad_norm": 1.2490419972325981,
"learning_rate": 6.186666666666668e-05,
"loss": 0.9815347194671631,
"step": 59
},
{
"epoch": 0.0803616273229533,
"grad_norm": 0.8182137749498168,
"learning_rate": 6.293333333333334e-05,
"loss": 0.9416379332542419,
"step": 60
},
{
"epoch": 0.08170098777833584,
"grad_norm": 0.9511257567878127,
"learning_rate": 6.400000000000001e-05,
"loss": 0.9608907103538513,
"step": 61
},
{
"epoch": 0.0830403482337184,
"grad_norm": 1.1244950887988947,
"learning_rate": 6.506666666666667e-05,
"loss": 0.9711039662361145,
"step": 62
},
{
"epoch": 0.08437970868910095,
"grad_norm": 0.7241607125581461,
"learning_rate": 6.613333333333333e-05,
"loss": 0.956953763961792,
"step": 63
},
{
"epoch": 0.0857190691444835,
"grad_norm": 1.2263338483627315,
"learning_rate": 6.720000000000001e-05,
"loss": 0.9655635952949524,
"step": 64
},
{
"epoch": 0.08705842959986607,
"grad_norm": 1.2341337964430132,
"learning_rate": 6.826666666666668e-05,
"loss": 0.9627944231033325,
"step": 65
},
{
"epoch": 0.08839779005524862,
"grad_norm": 1.1775448887951347,
"learning_rate": 6.933333333333334e-05,
"loss": 0.9672762155532837,
"step": 66
},
{
"epoch": 0.08973715051063118,
"grad_norm": 0.9411494481784036,
"learning_rate": 7.04e-05,
"loss": 0.9656286835670471,
"step": 67
},
{
"epoch": 0.09107651096601373,
"grad_norm": 1.5058542772932988,
"learning_rate": 7.146666666666666e-05,
"loss": 0.9635396003723145,
"step": 68
},
{
"epoch": 0.09241587142139629,
"grad_norm": 0.8791327927912097,
"learning_rate": 7.253333333333334e-05,
"loss": 0.9501133561134338,
"step": 69
},
{
"epoch": 0.09375523187677884,
"grad_norm": 1.0136457894860957,
"learning_rate": 7.360000000000001e-05,
"loss": 0.9675654172897339,
"step": 70
},
{
"epoch": 0.09509459233216139,
"grad_norm": 0.9418617062383696,
"learning_rate": 7.466666666666667e-05,
"loss": 0.9422785043716431,
"step": 71
},
{
"epoch": 0.09643395278754395,
"grad_norm": 1.7268687993740954,
"learning_rate": 7.573333333333334e-05,
"loss": 0.9585210084915161,
"step": 72
},
{
"epoch": 0.0977733132429265,
"grad_norm": 0.5922280173398358,
"learning_rate": 7.680000000000001e-05,
"loss": 0.9478949308395386,
"step": 73
},
{
"epoch": 0.09911267369830906,
"grad_norm": 1.578711533324252,
"learning_rate": 7.786666666666667e-05,
"loss": 0.9586780071258545,
"step": 74
},
{
"epoch": 0.10045203415369161,
"grad_norm": 0.8717604778689436,
"learning_rate": 7.893333333333335e-05,
"loss": 0.9589528441429138,
"step": 75
},
{
"epoch": 0.10179139460907417,
"grad_norm": 1.0780985444934892,
"learning_rate": 8e-05,
"loss": 0.9439480304718018,
"step": 76
},
{
"epoch": 0.10313075506445672,
"grad_norm": 1.0712670283093053,
"learning_rate": 7.999956289034042e-05,
"loss": 0.9548436999320984,
"step": 77
},
{
"epoch": 0.10447011551983927,
"grad_norm": 1.4418577692497994,
"learning_rate": 7.999825157091493e-05,
"loss": 0.9496671557426453,
"step": 78
},
{
"epoch": 0.10580947597522183,
"grad_norm": 1.2367569549227773,
"learning_rate": 7.999606607038301e-05,
"loss": 0.9423863887786865,
"step": 79
},
{
"epoch": 0.10714883643060438,
"grad_norm": 1.0193656490711345,
"learning_rate": 7.999300643650986e-05,
"loss": 0.9572180509567261,
"step": 80
},
{
"epoch": 0.10848819688598695,
"grad_norm": 1.4112949078071588,
"learning_rate": 7.998907273616527e-05,
"loss": 0.9576992392539978,
"step": 81
},
{
"epoch": 0.1098275573413695,
"grad_norm": 1.0050483615871197,
"learning_rate": 7.998426505532213e-05,
"loss": 0.9516314268112183,
"step": 82
},
{
"epoch": 0.11116691779675206,
"grad_norm": 1.2066376507121481,
"learning_rate": 7.997858349905463e-05,
"loss": 0.9690903425216675,
"step": 83
},
{
"epoch": 0.1125062782521346,
"grad_norm": 0.9035971939740141,
"learning_rate": 7.997202819153595e-05,
"loss": 0.9476237893104553,
"step": 84
},
{
"epoch": 0.11384563870751715,
"grad_norm": 1.1057535061713546,
"learning_rate": 7.996459927603547e-05,
"loss": 0.9481550455093384,
"step": 85
},
{
"epoch": 0.11518499916289972,
"grad_norm": 1.0543897140522707,
"learning_rate": 7.995629691491577e-05,
"loss": 0.9473648071289062,
"step": 86
},
{
"epoch": 0.11652435961828227,
"grad_norm": 1.2215825631690842,
"learning_rate": 7.994712128962892e-05,
"loss": 0.9453766942024231,
"step": 87
},
{
"epoch": 0.11786372007366483,
"grad_norm": 1.0801046215143588,
"learning_rate": 7.993707260071268e-05,
"loss": 0.94734787940979,
"step": 88
},
{
"epoch": 0.11920308052904738,
"grad_norm": 1.1005953739452812,
"learning_rate": 7.992615106778595e-05,
"loss": 0.9533910751342773,
"step": 89
},
{
"epoch": 0.12054244098442994,
"grad_norm": 1.1039131539238454,
"learning_rate": 7.991435692954414e-05,
"loss": 0.9366779327392578,
"step": 90
},
{
"epoch": 0.12188180143981249,
"grad_norm": 1.0826300445910615,
"learning_rate": 7.990169044375385e-05,
"loss": 0.949599027633667,
"step": 91
},
{
"epoch": 0.12322116189519504,
"grad_norm": 1.080804013195341,
"learning_rate": 7.988815188724721e-05,
"loss": 0.9338595867156982,
"step": 92
},
{
"epoch": 0.1245605223505776,
"grad_norm": 0.9661792241983362,
"learning_rate": 7.987374155591594e-05,
"loss": 0.9345501661300659,
"step": 93
},
{
"epoch": 0.12589988280596015,
"grad_norm": 0.9692184575560373,
"learning_rate": 7.985845976470478e-05,
"loss": 0.9287736415863037,
"step": 94
},
{
"epoch": 0.1272392432613427,
"grad_norm": 0.8333110857933443,
"learning_rate": 7.984230684760464e-05,
"loss": 0.9591622352600098,
"step": 95
},
{
"epoch": 0.12857860371672528,
"grad_norm": 0.9273848223858157,
"learning_rate": 7.982528315764535e-05,
"loss": 0.9320292472839355,
"step": 96
},
{
"epoch": 0.12991796417210782,
"grad_norm": 1.2902401930657603,
"learning_rate": 7.980738906688788e-05,
"loss": 0.9414469003677368,
"step": 97
},
{
"epoch": 0.13125732462749037,
"grad_norm": 0.7527994455173133,
"learning_rate": 7.97886249664162e-05,
"loss": 0.9305897355079651,
"step": 98
},
{
"epoch": 0.13259668508287292,
"grad_norm": 0.9576732066553404,
"learning_rate": 7.976899126632883e-05,
"loss": 0.9291049838066101,
"step": 99
},
{
"epoch": 0.13393604553825547,
"grad_norm": 0.8162756897476297,
"learning_rate": 7.974848839572971e-05,
"loss": 0.9590339660644531,
"step": 100
},
{
"epoch": 0.13527540599363805,
"grad_norm": 0.6762070221393078,
"learning_rate": 7.972711680271902e-05,
"loss": 0.9453955888748169,
"step": 101
},
{
"epoch": 0.1366147664490206,
"grad_norm": 0.8308301450039904,
"learning_rate": 7.970487695438324e-05,
"loss": 0.9262537956237793,
"step": 102
},
{
"epoch": 0.13795412690440315,
"grad_norm": 0.6778798009469817,
"learning_rate": 7.968176933678499e-05,
"loss": 0.9314043521881104,
"step": 103
},
{
"epoch": 0.1392934873597857,
"grad_norm": 0.6607037720687131,
"learning_rate": 7.965779445495243e-05,
"loss": 0.9274415969848633,
"step": 104
},
{
"epoch": 0.14063284781516824,
"grad_norm": 0.6622639262728551,
"learning_rate": 7.963295283286816e-05,
"loss": 0.9263443946838379,
"step": 105
},
{
"epoch": 0.14197220827055082,
"grad_norm": 0.7912652055136309,
"learning_rate": 7.960724501345783e-05,
"loss": 0.9300252199172974,
"step": 106
},
{
"epoch": 0.14331156872593337,
"grad_norm": 1.054978432558347,
"learning_rate": 7.958067155857827e-05,
"loss": 0.9237234592437744,
"step": 107
},
{
"epoch": 0.14465092918131592,
"grad_norm": 1.0654154939069729,
"learning_rate": 7.955323304900514e-05,
"loss": 0.9344139695167542,
"step": 108
},
{
"epoch": 0.14599028963669847,
"grad_norm": 0.8524134781957012,
"learning_rate": 7.952493008442036e-05,
"loss": 0.9211926460266113,
"step": 109
},
{
"epoch": 0.14732965009208104,
"grad_norm": 0.7025932587219853,
"learning_rate": 7.949576328339886e-05,
"loss": 0.9348384737968445,
"step": 110
},
{
"epoch": 0.1486690105474636,
"grad_norm": 0.6306347151205624,
"learning_rate": 7.946573328339517e-05,
"loss": 0.9268158078193665,
"step": 111
},
{
"epoch": 0.15000837100284614,
"grad_norm": 0.555961808840402,
"learning_rate": 7.943484074072943e-05,
"loss": 0.9165964126586914,
"step": 112
},
{
"epoch": 0.1513477314582287,
"grad_norm": 0.5538014043761714,
"learning_rate": 7.940308633057311e-05,
"loss": 0.9227532148361206,
"step": 113
},
{
"epoch": 0.15268709191361124,
"grad_norm": 0.616239699596105,
"learning_rate": 7.937047074693418e-05,
"loss": 0.9242917895317078,
"step": 114
},
{
"epoch": 0.15402645236899382,
"grad_norm": 0.6871147186152412,
"learning_rate": 7.933699470264194e-05,
"loss": 0.9295473098754883,
"step": 115
},
{
"epoch": 0.15536581282437636,
"grad_norm": 0.7296739512651451,
"learning_rate": 7.930265892933154e-05,
"loss": 0.9315687417984009,
"step": 116
},
{
"epoch": 0.1567051732797589,
"grad_norm": 0.6490446540604792,
"learning_rate": 7.926746417742787e-05,
"loss": 0.9344536066055298,
"step": 117
},
{
"epoch": 0.15804453373514146,
"grad_norm": 0.5888098073239115,
"learning_rate": 7.923141121612922e-05,
"loss": 0.9065610766410828,
"step": 118
},
{
"epoch": 0.159383894190524,
"grad_norm": 0.48420697795240103,
"learning_rate": 7.919450083339051e-05,
"loss": 0.9160439968109131,
"step": 119
},
{
"epoch": 0.1607232546459066,
"grad_norm": 0.6698312291927234,
"learning_rate": 7.915673383590594e-05,
"loss": 0.918449342250824,
"step": 120
},
{
"epoch": 0.16206261510128914,
"grad_norm": 0.755676653353074,
"learning_rate": 7.911811104909151e-05,
"loss": 0.9130410552024841,
"step": 121
},
{
"epoch": 0.16340197555667169,
"grad_norm": 0.604901772991201,
"learning_rate": 7.907863331706687e-05,
"loss": 0.9236974716186523,
"step": 122
},
{
"epoch": 0.16474133601205423,
"grad_norm": 0.5352954980579561,
"learning_rate": 7.903830150263693e-05,
"loss": 0.9243772625923157,
"step": 123
},
{
"epoch": 0.1660806964674368,
"grad_norm": 0.5519596915483821,
"learning_rate": 7.899711648727294e-05,
"loss": 0.9172723293304443,
"step": 124
},
{
"epoch": 0.16742005692281936,
"grad_norm": 0.5188579724106568,
"learning_rate": 7.895507917109335e-05,
"loss": 0.9457906484603882,
"step": 125
},
{
"epoch": 0.1687594173782019,
"grad_norm": 0.6671320105409932,
"learning_rate": 7.891219047284398e-05,
"loss": 0.9342581629753113,
"step": 126
},
{
"epoch": 0.17009877783358446,
"grad_norm": 0.6755983903544845,
"learning_rate": 7.886845132987804e-05,
"loss": 0.9132858514785767,
"step": 127
},
{
"epoch": 0.171438138288967,
"grad_norm": 0.6948054233230689,
"learning_rate": 7.882386269813564e-05,
"loss": 0.930912971496582,
"step": 128
},
{
"epoch": 0.17277749874434958,
"grad_norm": 1.195409472208412,
"learning_rate": 7.877842555212286e-05,
"loss": 0.9310725927352905,
"step": 129
},
{
"epoch": 0.17411685919973213,
"grad_norm": 1.1692345945453273,
"learning_rate": 7.873214088489047e-05,
"loss": 0.9095422029495239,
"step": 130
},
{
"epoch": 0.17545621965511468,
"grad_norm": 0.6863826376658689,
"learning_rate": 7.868500970801221e-05,
"loss": 0.9226965308189392,
"step": 131
},
{
"epoch": 0.17679558011049723,
"grad_norm": 0.5755734885102781,
"learning_rate": 7.863703305156273e-05,
"loss": 0.9249308109283447,
"step": 132
},
{
"epoch": 0.17813494056587978,
"grad_norm": 0.673888177777299,
"learning_rate": 7.858821196409504e-05,
"loss": 0.906527042388916,
"step": 133
},
{
"epoch": 0.17947430102126236,
"grad_norm": 0.7536373280389009,
"learning_rate": 7.853854751261757e-05,
"loss": 0.9176697731018066,
"step": 134
},
{
"epoch": 0.1808136614766449,
"grad_norm": 0.8145195593383961,
"learning_rate": 7.84880407825709e-05,
"loss": 0.9187479019165039,
"step": 135
},
{
"epoch": 0.18215302193202745,
"grad_norm": 0.8866001504567084,
"learning_rate": 7.843669287780399e-05,
"loss": 0.9249315857887268,
"step": 136
},
{
"epoch": 0.18349238238741,
"grad_norm": 1.0158024079305474,
"learning_rate": 7.838450492055013e-05,
"loss": 0.9169516563415527,
"step": 137
},
{
"epoch": 0.18483174284279258,
"grad_norm": 1.0310885995767975,
"learning_rate": 7.83314780514023e-05,
"loss": 0.9106529951095581,
"step": 138
},
{
"epoch": 0.18617110329817513,
"grad_norm": 0.9123831996966875,
"learning_rate": 7.827761342928836e-05,
"loss": 0.9233460426330566,
"step": 139
},
{
"epoch": 0.18751046375355768,
"grad_norm": 0.8139934469336522,
"learning_rate": 7.822291223144564e-05,
"loss": 0.9139848947525024,
"step": 140
},
{
"epoch": 0.18884982420894023,
"grad_norm": 0.9194655551693208,
"learning_rate": 7.816737565339522e-05,
"loss": 0.9265908002853394,
"step": 141
},
{
"epoch": 0.19018918466432277,
"grad_norm": 0.7853146807972304,
"learning_rate": 7.811100490891586e-05,
"loss": 0.9096084833145142,
"step": 142
},
{
"epoch": 0.19152854511970535,
"grad_norm": 0.45037783087212563,
"learning_rate": 7.805380123001738e-05,
"loss": 0.9067190885543823,
"step": 143
},
{
"epoch": 0.1928679055750879,
"grad_norm": 0.8272701165356915,
"learning_rate": 7.799576586691385e-05,
"loss": 0.9237310886383057,
"step": 144
},
{
"epoch": 0.19420726603047045,
"grad_norm": 0.7914425746519923,
"learning_rate": 7.793690008799613e-05,
"loss": 0.910027027130127,
"step": 145
},
{
"epoch": 0.195546626485853,
"grad_norm": 0.6632510463555492,
"learning_rate": 7.787720517980424e-05,
"loss": 0.8966417908668518,
"step": 146
},
{
"epoch": 0.19688598694123555,
"grad_norm": 0.6447162018999167,
"learning_rate": 7.781668244699925e-05,
"loss": 0.9306788444519043,
"step": 147
},
{
"epoch": 0.19822534739661812,
"grad_norm": 0.5478719041076034,
"learning_rate": 7.775533321233471e-05,
"loss": 0.9183269739151001,
"step": 148
},
{
"epoch": 0.19956470785200067,
"grad_norm": 0.6091679840338063,
"learning_rate": 7.769315881662778e-05,
"loss": 0.9106794595718384,
"step": 149
},
{
"epoch": 0.20090406830738322,
"grad_norm": 0.5062433888598267,
"learning_rate": 7.763016061872989e-05,
"loss": 0.9021029472351074,
"step": 150
},
{
"epoch": 0.20224342876276577,
"grad_norm": 0.5863067489697747,
"learning_rate": 7.756633999549712e-05,
"loss": 0.9056645631790161,
"step": 151
},
{
"epoch": 0.20358278921814835,
"grad_norm": 0.6320856385434267,
"learning_rate": 7.750169834175998e-05,
"loss": 0.9039326310157776,
"step": 152
},
{
"epoch": 0.2049221496735309,
"grad_norm": 0.6560683461251866,
"learning_rate": 7.743623707029304e-05,
"loss": 0.9264083504676819,
"step": 153
},
{
"epoch": 0.20626151012891344,
"grad_norm": 0.6341868974563667,
"learning_rate": 7.736995761178399e-05,
"loss": 0.9121809601783752,
"step": 154
},
{
"epoch": 0.207600870584296,
"grad_norm": 0.5253972386667246,
"learning_rate": 7.730286141480244e-05,
"loss": 0.9089382290840149,
"step": 155
},
{
"epoch": 0.20894023103967854,
"grad_norm": 0.4094715328691596,
"learning_rate": 7.723494994576818e-05,
"loss": 0.9171542525291443,
"step": 156
},
{
"epoch": 0.21027959149506112,
"grad_norm": 0.4055413751407779,
"learning_rate": 7.716622468891914e-05,
"loss": 0.903844952583313,
"step": 157
},
{
"epoch": 0.21161895195044367,
"grad_norm": 0.5240987539662486,
"learning_rate": 7.7096687146279e-05,
"loss": 0.9096250534057617,
"step": 158
},
{
"epoch": 0.21295831240582622,
"grad_norm": 0.5332665630971269,
"learning_rate": 7.702633883762437e-05,
"loss": 0.9141714572906494,
"step": 159
},
{
"epoch": 0.21429767286120877,
"grad_norm": 0.6210995520121828,
"learning_rate": 7.695518130045147e-05,
"loss": 0.902588963508606,
"step": 160
},
{
"epoch": 0.21563703331659131,
"grad_norm": 0.6533753394367477,
"learning_rate": 7.68832160899427e-05,
"loss": 0.8970866799354553,
"step": 161
},
{
"epoch": 0.2169763937719739,
"grad_norm": 0.666843709876833,
"learning_rate": 7.681044477893243e-05,
"loss": 0.9080835580825806,
"step": 162
},
{
"epoch": 0.21831575422735644,
"grad_norm": 0.6227292968864494,
"learning_rate": 7.673686895787285e-05,
"loss": 0.9074655175209045,
"step": 163
},
{
"epoch": 0.219655114682739,
"grad_norm": 0.5347100691461162,
"learning_rate": 7.666249023479905e-05,
"loss": 0.9084531664848328,
"step": 164
},
{
"epoch": 0.22099447513812154,
"grad_norm": 0.5175594119380308,
"learning_rate": 7.658731023529396e-05,
"loss": 0.9043017029762268,
"step": 165
},
{
"epoch": 0.22233383559350411,
"grad_norm": 0.5436688066191951,
"learning_rate": 7.651133060245276e-05,
"loss": 0.9071587324142456,
"step": 166
},
{
"epoch": 0.22367319604888666,
"grad_norm": 0.6055580577896438,
"learning_rate": 7.643455299684704e-05,
"loss": 0.9014830589294434,
"step": 167
},
{
"epoch": 0.2250125565042692,
"grad_norm": 0.6931253151907791,
"learning_rate": 7.635697909648844e-05,
"loss": 0.9084873795509338,
"step": 168
},
{
"epoch": 0.22635191695965176,
"grad_norm": 0.8546025727166501,
"learning_rate": 7.627861059679202e-05,
"loss": 0.9147599339485168,
"step": 169
},
{
"epoch": 0.2276912774150343,
"grad_norm": 1.0516728931472508,
"learning_rate": 7.61994492105392e-05,
"loss": 0.9146803617477417,
"step": 170
},
{
"epoch": 0.2290306378704169,
"grad_norm": 1.1750636161132695,
"learning_rate": 7.61194966678403e-05,
"loss": 0.9208601117134094,
"step": 171
},
{
"epoch": 0.23036999832579944,
"grad_norm": 0.6895169678113093,
"learning_rate": 7.603875471609677e-05,
"loss": 0.9078114032745361,
"step": 172
},
{
"epoch": 0.23170935878118198,
"grad_norm": 0.49807653314280315,
"learning_rate": 7.595722511996295e-05,
"loss": 0.9131247401237488,
"step": 173
},
{
"epoch": 0.23304871923656453,
"grad_norm": 0.7343320798017458,
"learning_rate": 7.587490966130754e-05,
"loss": 0.9078603386878967,
"step": 174
},
{
"epoch": 0.23438807969194708,
"grad_norm": 0.8708473343532465,
"learning_rate": 7.579181013917465e-05,
"loss": 0.9184587001800537,
"step": 175
},
{
"epoch": 0.23572744014732966,
"grad_norm": 0.7054594615675117,
"learning_rate": 7.570792836974447e-05,
"loss": 0.9165666699409485,
"step": 176
},
{
"epoch": 0.2370668006027122,
"grad_norm": 0.4925485684667248,
"learning_rate": 7.562326618629359e-05,
"loss": 0.895115315914154,
"step": 177
},
{
"epoch": 0.23840616105809476,
"grad_norm": 0.6006056466613413,
"learning_rate": 7.55378254391549e-05,
"loss": 0.9025277495384216,
"step": 178
},
{
"epoch": 0.2397455215134773,
"grad_norm": 0.5980928196151747,
"learning_rate": 7.545160799567722e-05,
"loss": 0.9152911901473999,
"step": 179
},
{
"epoch": 0.24108488196885988,
"grad_norm": 0.44318994378693194,
"learning_rate": 7.536461574018439e-05,
"loss": 0.9092339277267456,
"step": 180
},
{
"epoch": 0.24242424242424243,
"grad_norm": 0.4895486082026665,
"learning_rate": 7.52768505739342e-05,
"loss": 0.8995894193649292,
"step": 181
},
{
"epoch": 0.24376360287962498,
"grad_norm": 0.5807322769308096,
"learning_rate": 7.518831441507674e-05,
"loss": 0.8934712409973145,
"step": 182
},
{
"epoch": 0.24510296333500753,
"grad_norm": 0.507338986090725,
"learning_rate": 7.509900919861251e-05,
"loss": 0.9049476385116577,
"step": 183
},
{
"epoch": 0.24644232379039008,
"grad_norm": 0.46447343306662686,
"learning_rate": 7.500893687635015e-05,
"loss": 0.9096876382827759,
"step": 184
},
{
"epoch": 0.24778168424577265,
"grad_norm": 0.43290671210374365,
"learning_rate": 7.491809941686377e-05,
"loss": 0.9005957245826721,
"step": 185
},
{
"epoch": 0.2491210447011552,
"grad_norm": 0.4233601661390767,
"learning_rate": 7.482649880544994e-05,
"loss": 0.8930633068084717,
"step": 186
},
{
"epoch": 0.2504604051565378,
"grad_norm": 0.47483889768660387,
"learning_rate": 7.473413704408423e-05,
"loss": 0.8927392363548279,
"step": 187
},
{
"epoch": 0.2517997656119203,
"grad_norm": 0.41430034014432743,
"learning_rate": 7.464101615137756e-05,
"loss": 0.9012651443481445,
"step": 188
},
{
"epoch": 0.2531391260673029,
"grad_norm": 0.44657266747672564,
"learning_rate": 7.454713816253201e-05,
"loss": 0.9092891812324524,
"step": 189
},
{
"epoch": 0.2544784865226854,
"grad_norm": 0.5758963401160683,
"learning_rate": 7.445250512929637e-05,
"loss": 0.9000071287155151,
"step": 190
},
{
"epoch": 0.255817846978068,
"grad_norm": 0.6348599234700715,
"learning_rate": 7.435711911992128e-05,
"loss": 0.9062003493309021,
"step": 191
},
{
"epoch": 0.25715720743345055,
"grad_norm": 0.7445779300105985,
"learning_rate": 7.426098221911406e-05,
"loss": 0.9109039306640625,
"step": 192
},
{
"epoch": 0.2584965678888331,
"grad_norm": 0.9461198656306916,
"learning_rate": 7.416409652799311e-05,
"loss": 0.9037576913833618,
"step": 193
},
{
"epoch": 0.25983592834421565,
"grad_norm": 1.079965574399244,
"learning_rate": 7.406646416404198e-05,
"loss": 0.9101488590240479,
"step": 194
},
{
"epoch": 0.26117528879959817,
"grad_norm": 0.8952709420552329,
"learning_rate": 7.396808726106316e-05,
"loss": 0.9010953903198242,
"step": 195
},
{
"epoch": 0.26251464925498075,
"grad_norm": 0.7295843220367842,
"learning_rate": 7.386896796913137e-05,
"loss": 0.9150183796882629,
"step": 196
},
{
"epoch": 0.2638540097103633,
"grad_norm": 0.4969681568476923,
"learning_rate": 7.376910845454662e-05,
"loss": 0.8933469653129578,
"step": 197
},
{
"epoch": 0.26519337016574585,
"grad_norm": 0.5892305543752261,
"learning_rate": 7.36685108997868e-05,
"loss": 0.8998183012008667,
"step": 198
},
{
"epoch": 0.2665327306211284,
"grad_norm": 0.7912320036949558,
"learning_rate": 7.35671775034601e-05,
"loss": 0.9029569625854492,
"step": 199
},
{
"epoch": 0.26787209107651094,
"grad_norm": 0.7213642003398647,
"learning_rate": 7.34651104802568e-05,
"loss": 0.901849627494812,
"step": 200
},
{
"epoch": 0.2692114515318935,
"grad_norm": 0.45576846929011733,
"learning_rate": 7.3362312060901e-05,
"loss": 0.9074373245239258,
"step": 201
},
{
"epoch": 0.2705508119872761,
"grad_norm": 0.4442880589446713,
"learning_rate": 7.325878449210182e-05,
"loss": 0.894812822341919,
"step": 202
},
{
"epoch": 0.2718901724426586,
"grad_norm": 0.6350453317740167,
"learning_rate": 7.315453003650425e-05,
"loss": 0.9038816094398499,
"step": 203
},
{
"epoch": 0.2732295328980412,
"grad_norm": 0.5669825177653348,
"learning_rate": 7.30495509726398e-05,
"loss": 0.9047380089759827,
"step": 204
},
{
"epoch": 0.2745688933534237,
"grad_norm": 0.40694618687161566,
"learning_rate": 7.29438495948766e-05,
"loss": 0.907146692276001,
"step": 205
},
{
"epoch": 0.2759082538088063,
"grad_norm": 0.5158298515102957,
"learning_rate": 7.28374282133693e-05,
"loss": 0.9054394364356995,
"step": 206
},
{
"epoch": 0.27724761426418887,
"grad_norm": 0.4674214046015115,
"learning_rate": 7.273028915400862e-05,
"loss": 0.89744633436203,
"step": 207
},
{
"epoch": 0.2785869747195714,
"grad_norm": 0.4010198365112395,
"learning_rate": 7.262243475837041e-05,
"loss": 0.8924587965011597,
"step": 208
},
{
"epoch": 0.27992633517495397,
"grad_norm": 0.5034263045570243,
"learning_rate": 7.251386738366462e-05,
"loss": 0.8863202333450317,
"step": 209
},
{
"epoch": 0.2812656956303365,
"grad_norm": 0.5023444681579953,
"learning_rate": 7.240458940268362e-05,
"loss": 0.8941077589988708,
"step": 210
},
{
"epoch": 0.28260505608571906,
"grad_norm": 0.46891769868263444,
"learning_rate": 7.229460320375051e-05,
"loss": 0.906989574432373,
"step": 211
},
{
"epoch": 0.28394441654110164,
"grad_norm": 0.5647560272622669,
"learning_rate": 7.218391119066674e-05,
"loss": 0.9065754413604736,
"step": 212
},
{
"epoch": 0.28528377699648416,
"grad_norm": 0.6564851196340278,
"learning_rate": 7.207251578265974e-05,
"loss": 0.9066266417503357,
"step": 213
},
{
"epoch": 0.28662313745186674,
"grad_norm": 0.6605526813011159,
"learning_rate": 7.196041941432998e-05,
"loss": 0.8938264846801758,
"step": 214
},
{
"epoch": 0.2879624979072493,
"grad_norm": 0.7004298849905322,
"learning_rate": 7.184762453559767e-05,
"loss": 0.901447057723999,
"step": 215
},
{
"epoch": 0.28930185836263184,
"grad_norm": 0.6723620245213667,
"learning_rate": 7.173413361164941e-05,
"loss": 0.8870468139648438,
"step": 216
},
{
"epoch": 0.2906412188180144,
"grad_norm": 0.5895576054421268,
"learning_rate": 7.161994912288415e-05,
"loss": 0.8934962153434753,
"step": 217
},
{
"epoch": 0.29198057927339693,
"grad_norm": 0.6129669412207072,
"learning_rate": 7.150507356485903e-05,
"loss": 0.8962942957878113,
"step": 218
},
{
"epoch": 0.2933199397287795,
"grad_norm": 0.5802537294296758,
"learning_rate": 7.138950944823485e-05,
"loss": 0.8967691659927368,
"step": 219
},
{
"epoch": 0.2946593001841621,
"grad_norm": 0.4820962308267513,
"learning_rate": 7.12732592987212e-05,
"loss": 0.8924927711486816,
"step": 220
},
{
"epoch": 0.2959986606395446,
"grad_norm": 0.45589506205455754,
"learning_rate": 7.115632565702124e-05,
"loss": 0.894336998462677,
"step": 221
},
{
"epoch": 0.2973380210949272,
"grad_norm": 0.621116741269062,
"learning_rate": 7.10387110787762e-05,
"loss": 0.8900793790817261,
"step": 222
},
{
"epoch": 0.2986773815503097,
"grad_norm": 0.771116132471669,
"learning_rate": 7.092041813450948e-05,
"loss": 0.9094074964523315,
"step": 223
},
{
"epoch": 0.3000167420056923,
"grad_norm": 0.8818936933822646,
"learning_rate": 7.080144940957053e-05,
"loss": 0.8994623422622681,
"step": 224
},
{
"epoch": 0.30135610246107486,
"grad_norm": 0.9044298940788849,
"learning_rate": 7.068180750407825e-05,
"loss": 0.9020993113517761,
"step": 225
},
{
"epoch": 0.3026954629164574,
"grad_norm": 0.8105053927630517,
"learning_rate": 7.05614950328643e-05,
"loss": 0.8913443088531494,
"step": 226
},
{
"epoch": 0.30403482337183996,
"grad_norm": 0.6552999377507921,
"learning_rate": 7.044051462541585e-05,
"loss": 0.8977623581886292,
"step": 227
},
{
"epoch": 0.3053741838272225,
"grad_norm": 0.6276225150513225,
"learning_rate": 7.031886892581813e-05,
"loss": 0.9037375450134277,
"step": 228
},
{
"epoch": 0.30671354428260506,
"grad_norm": 0.47767505306157954,
"learning_rate": 7.019656059269664e-05,
"loss": 0.8967139720916748,
"step": 229
},
{
"epoch": 0.30805290473798763,
"grad_norm": 0.43735915139443376,
"learning_rate": 7.00735922991591e-05,
"loss": 0.887144923210144,
"step": 230
},
{
"epoch": 0.30939226519337015,
"grad_norm": 0.4990369099751219,
"learning_rate": 6.994996673273694e-05,
"loss": 0.8820310831069946,
"step": 231
},
{
"epoch": 0.31073162564875273,
"grad_norm": 0.4708659320973128,
"learning_rate": 6.982568659532663e-05,
"loss": 0.8966023921966553,
"step": 232
},
{
"epoch": 0.31207098610413525,
"grad_norm": 0.5323277193114596,
"learning_rate": 6.970075460313059e-05,
"loss": 0.8898813724517822,
"step": 233
},
{
"epoch": 0.3134103465595178,
"grad_norm": 0.4765758370618773,
"learning_rate": 6.957517348659785e-05,
"loss": 0.9058773517608643,
"step": 234
},
{
"epoch": 0.3147497070149004,
"grad_norm": 0.2922891582098603,
"learning_rate": 6.944894599036439e-05,
"loss": 0.8985657691955566,
"step": 235
},
{
"epoch": 0.3160890674702829,
"grad_norm": 0.4017359965605547,
"learning_rate": 6.932207487319305e-05,
"loss": 0.8944315910339355,
"step": 236
},
{
"epoch": 0.3174284279256655,
"grad_norm": 0.40396755775149173,
"learning_rate": 6.919456290791343e-05,
"loss": 0.9019736051559448,
"step": 237
},
{
"epoch": 0.318767788381048,
"grad_norm": 0.3448754102708148,
"learning_rate": 6.906641288136109e-05,
"loss": 0.8921661376953125,
"step": 238
},
{
"epoch": 0.3201071488364306,
"grad_norm": 0.3563739009758267,
"learning_rate": 6.893762759431675e-05,
"loss": 0.8991590142250061,
"step": 239
},
{
"epoch": 0.3214465092918132,
"grad_norm": 0.3025703943184552,
"learning_rate": 6.880820986144507e-05,
"loss": 0.8966647386550903,
"step": 240
},
{
"epoch": 0.3227858697471957,
"grad_norm": 0.32710433074697187,
"learning_rate": 6.86781625112331e-05,
"loss": 0.8932673335075378,
"step": 241
},
{
"epoch": 0.3241252302025783,
"grad_norm": 0.3159486886720857,
"learning_rate": 6.854748838592849e-05,
"loss": 0.8867123126983643,
"step": 242
},
{
"epoch": 0.32546459065796085,
"grad_norm": 0.33576352782212976,
"learning_rate": 6.841619034147738e-05,
"loss": 0.8945485949516296,
"step": 243
},
{
"epoch": 0.32680395111334337,
"grad_norm": 0.3382299578038365,
"learning_rate": 6.828427124746191e-05,
"loss": 0.8893507122993469,
"step": 244
},
{
"epoch": 0.32814331156872595,
"grad_norm": 0.3145669367280524,
"learning_rate": 6.815173398703762e-05,
"loss": 0.8952306509017944,
"step": 245
},
{
"epoch": 0.32948267202410847,
"grad_norm": 0.3702066361612188,
"learning_rate": 6.801858145687033e-05,
"loss": 0.900354266166687,
"step": 246
},
{
"epoch": 0.33082203247949105,
"grad_norm": 0.3590790337895165,
"learning_rate": 6.788481656707294e-05,
"loss": 0.8962044715881348,
"step": 247
},
{
"epoch": 0.3321613929348736,
"grad_norm": 0.29409387021807054,
"learning_rate": 6.775044224114168e-05,
"loss": 0.8917175531387329,
"step": 248
},
{
"epoch": 0.33350075339025614,
"grad_norm": 0.33139455469582174,
"learning_rate": 6.761546141589235e-05,
"loss": 0.8945397138595581,
"step": 249
},
{
"epoch": 0.3348401138456387,
"grad_norm": 0.3304502768847914,
"learning_rate": 6.747987704139607e-05,
"loss": 0.8865037560462952,
"step": 250
},
{
"epoch": 0.33617947430102124,
"grad_norm": 0.35207067018590943,
"learning_rate": 6.734369208091486e-05,
"loss": 0.8916755318641663,
"step": 251
},
{
"epoch": 0.3375188347564038,
"grad_norm": 0.40919798096964943,
"learning_rate": 6.720690951083678e-05,
"loss": 0.8847212791442871,
"step": 252
},
{
"epoch": 0.3388581952117864,
"grad_norm": 0.3513659416141469,
"learning_rate": 6.706953232061098e-05,
"loss": 0.8843253254890442,
"step": 253
},
{
"epoch": 0.3401975556671689,
"grad_norm": 0.30034711529232383,
"learning_rate": 6.693156351268229e-05,
"loss": 0.8973168730735779,
"step": 254
},
{
"epoch": 0.3415369161225515,
"grad_norm": 0.25812676412176466,
"learning_rate": 6.679300610242564e-05,
"loss": 0.9002690315246582,
"step": 255
},
{
"epoch": 0.342876276577934,
"grad_norm": 0.26476799198584866,
"learning_rate": 6.665386311808017e-05,
"loss": 0.8933571577072144,
"step": 256
},
{
"epoch": 0.3442156370333166,
"grad_norm": 0.3055500682498974,
"learning_rate": 6.651413760068298e-05,
"loss": 0.9000169634819031,
"step": 257
},
{
"epoch": 0.34555499748869917,
"grad_norm": 0.2753301578914753,
"learning_rate": 6.637383260400276e-05,
"loss": 0.8983247876167297,
"step": 258
},
{
"epoch": 0.3468943579440817,
"grad_norm": 0.3283199044322027,
"learning_rate": 6.623295119447297e-05,
"loss": 0.8985056281089783,
"step": 259
},
{
"epoch": 0.34823371839946426,
"grad_norm": 0.3582940804838448,
"learning_rate": 6.609149645112485e-05,
"loss": 0.8870316743850708,
"step": 260
},
{
"epoch": 0.3495730788548468,
"grad_norm": 0.4306838172471171,
"learning_rate": 6.594947146552015e-05,
"loss": 0.8838645815849304,
"step": 261
},
{
"epoch": 0.35091243931022936,
"grad_norm": 0.48455804455635887,
"learning_rate": 6.580687934168352e-05,
"loss": 0.8925988674163818,
"step": 262
},
{
"epoch": 0.35225179976561194,
"grad_norm": 0.5758294275573599,
"learning_rate": 6.566372319603467e-05,
"loss": 0.8979231715202332,
"step": 263
},
{
"epoch": 0.35359116022099446,
"grad_norm": 0.6666731640546475,
"learning_rate": 6.552000615732035e-05,
"loss": 0.8943511843681335,
"step": 264
},
{
"epoch": 0.35493052067637704,
"grad_norm": 0.6604209524822312,
"learning_rate": 6.537573136654582e-05,
"loss": 0.8853351473808289,
"step": 265
},
{
"epoch": 0.35626988113175956,
"grad_norm": 0.4123786870554884,
"learning_rate": 6.523090197690634e-05,
"loss": 0.8918644785881042,
"step": 266
},
{
"epoch": 0.35760924158714213,
"grad_norm": 0.2502223960825913,
"learning_rate": 6.508552115371814e-05,
"loss": 0.893090546131134,
"step": 267
},
{
"epoch": 0.3589486020425247,
"grad_norm": 0.4103290878988621,
"learning_rate": 6.493959207434934e-05,
"loss": 0.8756798505783081,
"step": 268
},
{
"epoch": 0.36028796249790723,
"grad_norm": 0.5340723449895526,
"learning_rate": 6.479311792815047e-05,
"loss": 0.8869861364364624,
"step": 269
},
{
"epoch": 0.3616273229532898,
"grad_norm": 0.5092937192195168,
"learning_rate": 6.464610191638472e-05,
"loss": 0.8688637018203735,
"step": 270
},
{
"epoch": 0.3629666834086724,
"grad_norm": 0.36841736308738604,
"learning_rate": 6.449854725215801e-05,
"loss": 0.8819414973258972,
"step": 271
},
{
"epoch": 0.3643060438640549,
"grad_norm": 0.23108168052962968,
"learning_rate": 6.435045716034883e-05,
"loss": 0.8896312713623047,
"step": 272
},
{
"epoch": 0.3656454043194375,
"grad_norm": 0.3314939187707688,
"learning_rate": 6.420183487753765e-05,
"loss": 0.8876444101333618,
"step": 273
},
{
"epoch": 0.36698476477482,
"grad_norm": 0.44064601491510763,
"learning_rate": 6.405268365193624e-05,
"loss": 0.8837222456932068,
"step": 274
},
{
"epoch": 0.3683241252302026,
"grad_norm": 0.4117399045978558,
"learning_rate": 6.390300674331667e-05,
"loss": 0.8932456970214844,
"step": 275
},
{
"epoch": 0.36966348568558516,
"grad_norm": 0.3501980802705,
"learning_rate": 6.375280742294007e-05,
"loss": 0.896408200263977,
"step": 276
},
{
"epoch": 0.3710028461409677,
"grad_norm": 0.31925464072542226,
"learning_rate": 6.360208897348513e-05,
"loss": 0.890372097492218,
"step": 277
},
{
"epoch": 0.37234220659635026,
"grad_norm": 0.3672953058879004,
"learning_rate": 6.345085468897637e-05,
"loss": 0.8912795782089233,
"step": 278
},
{
"epoch": 0.3736815670517328,
"grad_norm": 0.37585534860719505,
"learning_rate": 6.32991078747121e-05,
"loss": 0.8913096785545349,
"step": 279
},
{
"epoch": 0.37502092750711535,
"grad_norm": 0.3686117104703527,
"learning_rate": 6.314685184719224e-05,
"loss": 0.8860468864440918,
"step": 280
},
{
"epoch": 0.37636028796249793,
"grad_norm": 0.36203560770739496,
"learning_rate": 6.29940899340458e-05,
"loss": 0.8942425847053528,
"step": 281
},
{
"epoch": 0.37769964841788045,
"grad_norm": 0.39656735441651286,
"learning_rate": 6.284082547395822e-05,
"loss": 0.884894609451294,
"step": 282
},
{
"epoch": 0.37903900887326303,
"grad_norm": 0.43021602998882896,
"learning_rate": 6.268706181659823e-05,
"loss": 0.8949446678161621,
"step": 283
},
{
"epoch": 0.38037836932864555,
"grad_norm": 0.4560535418745229,
"learning_rate": 6.253280232254489e-05,
"loss": 0.8839805722236633,
"step": 284
},
{
"epoch": 0.3817177297840281,
"grad_norm": 0.46627492864603226,
"learning_rate": 6.23780503632139e-05,
"loss": 0.8896687626838684,
"step": 285
},
{
"epoch": 0.3830570902394107,
"grad_norm": 0.4331165409015457,
"learning_rate": 6.22228093207841e-05,
"loss": 0.8688606023788452,
"step": 286
},
{
"epoch": 0.3843964506947932,
"grad_norm": 0.32952692176687587,
"learning_rate": 6.206708258812343e-05,
"loss": 0.8844860196113586,
"step": 287
},
{
"epoch": 0.3857358111501758,
"grad_norm": 0.32252625784796196,
"learning_rate": 6.191087356871486e-05,
"loss": 0.8889749646186829,
"step": 288
},
{
"epoch": 0.3870751716055583,
"grad_norm": 0.36745276817661704,
"learning_rate": 6.175418567658196e-05,
"loss": 0.8868741989135742,
"step": 289
},
{
"epoch": 0.3884145320609409,
"grad_norm": 0.43191395512680925,
"learning_rate": 6.159702233621427e-05,
"loss": 0.8811583518981934,
"step": 290
},
{
"epoch": 0.3897538925163235,
"grad_norm": 0.42293907743115355,
"learning_rate": 6.143938698249251e-05,
"loss": 0.8857870697975159,
"step": 291
},
{
"epoch": 0.391093252971706,
"grad_norm": 0.3558451089435877,
"learning_rate": 6.128128306061347e-05,
"loss": 0.8891360759735107,
"step": 292
},
{
"epoch": 0.3924326134270886,
"grad_norm": 0.2878485843617252,
"learning_rate": 6.112271402601472e-05,
"loss": 0.8804606199264526,
"step": 293
},
{
"epoch": 0.3937719738824711,
"grad_norm": 0.27858889793959796,
"learning_rate": 6.096368334429911e-05,
"loss": 0.8926126956939697,
"step": 294
},
{
"epoch": 0.39511133433785367,
"grad_norm": 0.2819024964747486,
"learning_rate": 6.0804194491158987e-05,
"loss": 0.8852767944335938,
"step": 295
},
{
"epoch": 0.39645069479323625,
"grad_norm": 0.2540202943224736,
"learning_rate": 6.064425095230027e-05,
"loss": 0.8917317390441895,
"step": 296
},
{
"epoch": 0.39779005524861877,
"grad_norm": 0.24541270777843874,
"learning_rate": 6.048385622336625e-05,
"loss": 0.8798589706420898,
"step": 297
},
{
"epoch": 0.39912941570400134,
"grad_norm": 0.2536269299289805,
"learning_rate": 6.0323013809861185e-05,
"loss": 0.8902722597122192,
"step": 298
},
{
"epoch": 0.4004687761593839,
"grad_norm": 0.27800051254159697,
"learning_rate": 6.016172722707372e-05,
"loss": 0.9056130647659302,
"step": 299
},
{
"epoch": 0.40180813661476644,
"grad_norm": 0.27633691297170004,
"learning_rate": 6.000000000000001e-05,
"loss": 0.8697539567947388,
"step": 300
},
{
"epoch": 0.403147497070149,
"grad_norm": 0.27434445994827267,
"learning_rate": 5.9837835663266724e-05,
"loss": 0.8804255127906799,
"step": 301
},
{
"epoch": 0.40448685752553154,
"grad_norm": 0.2524623974437759,
"learning_rate": 5.967523776105374e-05,
"loss": 0.8784493803977966,
"step": 302
},
{
"epoch": 0.4058262179809141,
"grad_norm": 0.25420975374572236,
"learning_rate": 5.9512209847016766e-05,
"loss": 0.8766741752624512,
"step": 303
},
{
"epoch": 0.4071655784362967,
"grad_norm": 0.2510235716517718,
"learning_rate": 5.9348755484209597e-05,
"loss": 0.8951706886291504,
"step": 304
},
{
"epoch": 0.4085049388916792,
"grad_norm": 0.28128699677214214,
"learning_rate": 5.9184878245006275e-05,
"loss": 0.8903220891952515,
"step": 305
},
{
"epoch": 0.4098442993470618,
"grad_norm": 0.2553241167540996,
"learning_rate": 5.902058171102302e-05,
"loss": 0.8924455642700195,
"step": 306
},
{
"epoch": 0.4111836598024443,
"grad_norm": 0.27284815580460237,
"learning_rate": 5.8855869473039924e-05,
"loss": 0.8797352313995361,
"step": 307
},
{
"epoch": 0.4125230202578269,
"grad_norm": 0.2961577585360785,
"learning_rate": 5.869074513092249e-05,
"loss": 0.889675498008728,
"step": 308
},
{
"epoch": 0.41386238071320947,
"grad_norm": 0.36149390074479637,
"learning_rate": 5.852521229354298e-05,
"loss": 0.8952120542526245,
"step": 309
},
{
"epoch": 0.415201741168592,
"grad_norm": 0.4056564851356049,
"learning_rate": 5.835927457870151e-05,
"loss": 0.8908874988555908,
"step": 310
},
{
"epoch": 0.41654110162397456,
"grad_norm": 0.45959990311749727,
"learning_rate": 5.8192935613046986e-05,
"loss": 0.8896716833114624,
"step": 311
},
{
"epoch": 0.4178804620793571,
"grad_norm": 0.44501497393699513,
"learning_rate": 5.802619903199782e-05,
"loss": 0.8718587160110474,
"step": 312
},
{
"epoch": 0.41921982253473966,
"grad_norm": 0.4099982344769972,
"learning_rate": 5.785906847966252e-05,
"loss": 0.87396639585495,
"step": 313
},
{
"epoch": 0.42055918299012224,
"grad_norm": 0.38915136386883503,
"learning_rate": 5.7691547608760055e-05,
"loss": 0.8890815377235413,
"step": 314
},
{
"epoch": 0.42189854344550476,
"grad_norm": 0.3916008388754374,
"learning_rate": 5.7523640080539954e-05,
"loss": 0.8961409330368042,
"step": 315
},
{
"epoch": 0.42323790390088734,
"grad_norm": 0.34461259063570354,
"learning_rate": 5.735534956470233e-05,
"loss": 0.8904027938842773,
"step": 316
},
{
"epoch": 0.42457726435626986,
"grad_norm": 0.3351489985095559,
"learning_rate": 5.71866797393177e-05,
"loss": 0.8841983079910278,
"step": 317
},
{
"epoch": 0.42591662481165243,
"grad_norm": 0.3415829415188758,
"learning_rate": 5.701763429074657e-05,
"loss": 0.8725353479385376,
"step": 318
},
{
"epoch": 0.427255985267035,
"grad_norm": 0.32192931188812973,
"learning_rate": 5.6848216913558847e-05,
"loss": 0.8806099891662598,
"step": 319
},
{
"epoch": 0.42859534572241753,
"grad_norm": 0.3215781282881794,
"learning_rate": 5.6678431310453134e-05,
"loss": 0.8860760927200317,
"step": 320
},
{
"epoch": 0.4299347061778001,
"grad_norm": 0.34914156634746246,
"learning_rate": 5.6508281192175794e-05,
"loss": 0.8721968531608582,
"step": 321
},
{
"epoch": 0.43127406663318263,
"grad_norm": 0.2658667412041895,
"learning_rate": 5.6337770277439854e-05,
"loss": 0.8862148523330688,
"step": 322
},
{
"epoch": 0.4326134270885652,
"grad_norm": 0.34695464253952446,
"learning_rate": 5.616690229284368e-05,
"loss": 0.8746384382247925,
"step": 323
},
{
"epoch": 0.4339527875439478,
"grad_norm": 0.4197423684536046,
"learning_rate": 5.5995680972789634e-05,
"loss": 0.8776325583457947,
"step": 324
},
{
"epoch": 0.4352921479993303,
"grad_norm": 0.46568817493418574,
"learning_rate": 5.582411005940236e-05,
"loss": 0.887018084526062,
"step": 325
},
{
"epoch": 0.4366315084547129,
"grad_norm": 0.38017911947951655,
"learning_rate": 5.5652193302447e-05,
"loss": 0.8889976739883423,
"step": 326
},
{
"epoch": 0.43797086891009546,
"grad_norm": 0.3060971777017487,
"learning_rate": 5.547993445924735e-05,
"loss": 0.8834925889968872,
"step": 327
},
{
"epoch": 0.439310229365478,
"grad_norm": 0.3878310131471193,
"learning_rate": 5.5307337294603595e-05,
"loss": 0.8835778832435608,
"step": 328
},
{
"epoch": 0.44064958982086055,
"grad_norm": 0.37272583086257793,
"learning_rate": 5.513440558071015e-05,
"loss": 0.88037109375,
"step": 329
},
{
"epoch": 0.4419889502762431,
"grad_norm": 0.2980108695701489,
"learning_rate": 5.496114309707313e-05,
"loss": 0.8899520635604858,
"step": 330
},
{
"epoch": 0.44332831073162565,
"grad_norm": 0.2942897917948679,
"learning_rate": 5.478755363042781e-05,
"loss": 0.8873578906059265,
"step": 331
},
{
"epoch": 0.44466767118700823,
"grad_norm": 0.35500305163807966,
"learning_rate": 5.461364097465581e-05,
"loss": 0.8870940208435059,
"step": 332
},
{
"epoch": 0.44600703164239075,
"grad_norm": 0.3079369949147375,
"learning_rate": 5.443940893070223e-05,
"loss": 0.8757143616676331,
"step": 333
},
{
"epoch": 0.4473463920977733,
"grad_norm": 0.3007868096995874,
"learning_rate": 5.4264861306492525e-05,
"loss": 0.8642425537109375,
"step": 334
},
{
"epoch": 0.44868575255315585,
"grad_norm": 0.31595038196673936,
"learning_rate": 5.409000191684935e-05,
"loss": 0.8808598518371582,
"step": 335
},
{
"epoch": 0.4500251130085384,
"grad_norm": 0.3146477839645367,
"learning_rate": 5.391483458340909e-05,
"loss": 0.8812928199768066,
"step": 336
},
{
"epoch": 0.451364473463921,
"grad_norm": 0.30376403427053666,
"learning_rate": 5.3739363134538446e-05,
"loss": 0.8947619795799255,
"step": 337
},
{
"epoch": 0.4527038339193035,
"grad_norm": 0.27657117052236885,
"learning_rate": 5.356359140525067e-05,
"loss": 0.8850589990615845,
"step": 338
},
{
"epoch": 0.4540431943746861,
"grad_norm": 0.30680178018188403,
"learning_rate": 5.3387523237121805e-05,
"loss": 0.8739247918128967,
"step": 339
},
{
"epoch": 0.4553825548300686,
"grad_norm": 0.3047542635453717,
"learning_rate": 5.321116247820669e-05,
"loss": 0.8850078582763672,
"step": 340
},
{
"epoch": 0.4567219152854512,
"grad_norm": 0.2923184588762319,
"learning_rate": 5.30345129829549e-05,
"loss": 0.8884513974189758,
"step": 341
},
{
"epoch": 0.4580612757408338,
"grad_norm": 0.306238293604028,
"learning_rate": 5.2857578612126466e-05,
"loss": 0.8790913224220276,
"step": 342
},
{
"epoch": 0.4594006361962163,
"grad_norm": 0.2776234461903938,
"learning_rate": 5.2680363232707533e-05,
"loss": 0.8726402521133423,
"step": 343
},
{
"epoch": 0.46073999665159887,
"grad_norm": 0.23559770485475257,
"learning_rate": 5.250287071782579e-05,
"loss": 0.8780654668807983,
"step": 344
},
{
"epoch": 0.4620793571069814,
"grad_norm": 0.2656621107355478,
"learning_rate": 5.232510494666588e-05,
"loss": 0.8802477717399597,
"step": 345
},
{
"epoch": 0.46341871756236397,
"grad_norm": 0.2654377895352511,
"learning_rate": 5.214706980438459e-05,
"loss": 0.8784705400466919,
"step": 346
},
{
"epoch": 0.46475807801774655,
"grad_norm": 0.25513055691779934,
"learning_rate": 5.1968769182025955e-05,
"loss": 0.8827534317970276,
"step": 347
},
{
"epoch": 0.46609743847312907,
"grad_norm": 0.23931345424644,
"learning_rate": 5.179020697643618e-05,
"loss": 0.8781850934028625,
"step": 348
},
{
"epoch": 0.46743679892851164,
"grad_norm": 0.2662310932045871,
"learning_rate": 5.16113870901785e-05,
"loss": 0.8855897784233093,
"step": 349
},
{
"epoch": 0.46877615938389416,
"grad_norm": 0.2719290408201649,
"learning_rate": 5.1432313431447915e-05,
"loss": 0.8810721039772034,
"step": 350
},
{
"epoch": 0.47011551983927674,
"grad_norm": 0.30347422955032666,
"learning_rate": 5.125298991398571e-05,
"loss": 0.8826359510421753,
"step": 351
},
{
"epoch": 0.4714548802946593,
"grad_norm": 0.23887523910106984,
"learning_rate": 5.107342045699397e-05,
"loss": 0.8747484683990479,
"step": 352
},
{
"epoch": 0.47279424075004184,
"grad_norm": 0.2505774705404714,
"learning_rate": 5.089360898504994e-05,
"loss": 0.8729058504104614,
"step": 353
},
{
"epoch": 0.4741336012054244,
"grad_norm": 0.279473567045265,
"learning_rate": 5.071355942802013e-05,
"loss": 0.8787670135498047,
"step": 354
},
{
"epoch": 0.475472961660807,
"grad_norm": 0.26199289777497575,
"learning_rate": 5.053327572097459e-05,
"loss": 0.8815037608146667,
"step": 355
},
{
"epoch": 0.4768123221161895,
"grad_norm": 0.21510974254409215,
"learning_rate": 5.0352761804100835e-05,
"loss": 0.8883016109466553,
"step": 356
},
{
"epoch": 0.4781516825715721,
"grad_norm": 0.2179594057045395,
"learning_rate": 5.017202162261767e-05,
"loss": 0.8739045858383179,
"step": 357
},
{
"epoch": 0.4794910430269546,
"grad_norm": 0.2583618991462232,
"learning_rate": 4.999105912668908e-05,
"loss": 0.875806987285614,
"step": 358
},
{
"epoch": 0.4808304034823372,
"grad_norm": 0.31852029663882087,
"learning_rate": 4.9809878271337794e-05,
"loss": 0.8874423503875732,
"step": 359
},
{
"epoch": 0.48216976393771976,
"grad_norm": 0.2835760526800237,
"learning_rate": 4.962848301635893e-05,
"loss": 0.8798930048942566,
"step": 360
},
{
"epoch": 0.4835091243931023,
"grad_norm": 0.22973024932768488,
"learning_rate": 4.9446877326233384e-05,
"loss": 0.8873306512832642,
"step": 361
},
{
"epoch": 0.48484848484848486,
"grad_norm": 0.2500936218554607,
"learning_rate": 4.926506517004122e-05,
"loss": 0.871864378452301,
"step": 362
},
{
"epoch": 0.4861878453038674,
"grad_norm": 0.2995047737515285,
"learning_rate": 4.9083050521374936e-05,
"loss": 0.8763599991798401,
"step": 363
},
{
"epoch": 0.48752720575924996,
"grad_norm": 0.32246297930373985,
"learning_rate": 4.890083735825258e-05,
"loss": 0.8670933842658997,
"step": 364
},
{
"epoch": 0.48886656621463254,
"grad_norm": 0.33433005720802467,
"learning_rate": 4.871842966303084e-05,
"loss": 0.8823835849761963,
"step": 365
},
{
"epoch": 0.49020592667001506,
"grad_norm": 0.27949324766081196,
"learning_rate": 4.8535831422318005e-05,
"loss": 0.8845060467720032,
"step": 366
},
{
"epoch": 0.49154528712539763,
"grad_norm": 0.2669463330828984,
"learning_rate": 4.83530466268868e-05,
"loss": 0.8807919025421143,
"step": 367
},
{
"epoch": 0.49288464758078016,
"grad_norm": 0.28921225779452586,
"learning_rate": 4.817007927158722e-05,
"loss": 0.8733174800872803,
"step": 368
},
{
"epoch": 0.49422400803616273,
"grad_norm": 0.3068794189080787,
"learning_rate": 4.798693335525918e-05,
"loss": 0.8903564214706421,
"step": 369
},
{
"epoch": 0.4955633684915453,
"grad_norm": 0.3383936832444029,
"learning_rate": 4.780361288064514e-05,
"loss": 0.8755660057067871,
"step": 370
},
{
"epoch": 0.49690272894692783,
"grad_norm": 0.33403170126533727,
"learning_rate": 4.76201218543026e-05,
"loss": 0.8842828273773193,
"step": 371
},
{
"epoch": 0.4982420894023104,
"grad_norm": 0.2502976358342153,
"learning_rate": 4.743646428651659e-05,
"loss": 0.8743306398391724,
"step": 372
},
{
"epoch": 0.4995814498576929,
"grad_norm": 0.2514415794283494,
"learning_rate": 4.7252644191211944e-05,
"loss": 0.8848379850387573,
"step": 373
},
{
"epoch": 0.5009208103130756,
"grad_norm": 0.3311666521257641,
"learning_rate": 4.7068665585865615e-05,
"loss": 0.8726536631584167,
"step": 374
},
{
"epoch": 0.5022601707684581,
"grad_norm": 0.29622349551027,
"learning_rate": 4.688453249141888e-05,
"loss": 0.8713983297348022,
"step": 375
},
{
"epoch": 0.5035995312238406,
"grad_norm": 0.24591467342470477,
"learning_rate": 4.670024893218946e-05,
"loss": 0.879889965057373,
"step": 376
},
{
"epoch": 0.5049388916792231,
"grad_norm": 0.24907780928618234,
"learning_rate": 4.651581893578356e-05,
"loss": 0.8837929964065552,
"step": 377
},
{
"epoch": 0.5062782521346058,
"grad_norm": 0.29743214535113877,
"learning_rate": 4.6331246533007795e-05,
"loss": 0.880785346031189,
"step": 378
},
{
"epoch": 0.5076176125899883,
"grad_norm": 0.27816072063242153,
"learning_rate": 4.6146535757781206e-05,
"loss": 0.8790816068649292,
"step": 379
},
{
"epoch": 0.5089569730453708,
"grad_norm": 0.258901378722433,
"learning_rate": 4.5961690647046974e-05,
"loss": 0.8754535913467407,
"step": 380
},
{
"epoch": 0.5102963335007534,
"grad_norm": 0.22897322988292967,
"learning_rate": 4.5776715240684296e-05,
"loss": 0.884695291519165,
"step": 381
},
{
"epoch": 0.511635693956136,
"grad_norm": 0.25410900340906356,
"learning_rate": 4.5591613581419984e-05,
"loss": 0.8682093620300293,
"step": 382
},
{
"epoch": 0.5129750544115185,
"grad_norm": 0.2292960696193583,
"learning_rate": 4.540638971474022e-05,
"loss": 0.8801172375679016,
"step": 383
},
{
"epoch": 0.5143144148669011,
"grad_norm": 0.27039953771276,
"learning_rate": 4.522104768880208e-05,
"loss": 0.8841304779052734,
"step": 384
},
{
"epoch": 0.5156537753222836,
"grad_norm": 0.2510162507401578,
"learning_rate": 4.5035591554345024e-05,
"loss": 0.8804336190223694,
"step": 385
},
{
"epoch": 0.5169931357776661,
"grad_norm": 0.19962477334591794,
"learning_rate": 4.485002536460246e-05,
"loss": 0.8744645118713379,
"step": 386
},
{
"epoch": 0.5183324962330487,
"grad_norm": 0.20953576879547658,
"learning_rate": 4.4664353175213106e-05,
"loss": 0.8919574022293091,
"step": 387
},
{
"epoch": 0.5196718566884313,
"grad_norm": 0.24110013426626412,
"learning_rate": 4.4478579044132314e-05,
"loss": 0.8700906038284302,
"step": 388
},
{
"epoch": 0.5210112171438138,
"grad_norm": 0.21570606604265194,
"learning_rate": 4.429270703154346e-05,
"loss": 0.8669141530990601,
"step": 389
},
{
"epoch": 0.5223505775991963,
"grad_norm": 0.22537953028075838,
"learning_rate": 4.4106741199769144e-05,
"loss": 0.893110990524292,
"step": 390
},
{
"epoch": 0.523689938054579,
"grad_norm": 0.20502638429010045,
"learning_rate": 4.392068561318244e-05,
"loss": 0.8845847845077515,
"step": 391
},
{
"epoch": 0.5250292985099615,
"grad_norm": 0.2029877791588085,
"learning_rate": 4.373454433811803e-05,
"loss": 0.8702945709228516,
"step": 392
},
{
"epoch": 0.526368658965344,
"grad_norm": 0.20027359499294622,
"learning_rate": 4.354832144278342e-05,
"loss": 0.876413106918335,
"step": 393
},
{
"epoch": 0.5277080194207266,
"grad_norm": 0.20852532140317728,
"learning_rate": 4.336202099716991e-05,
"loss": 0.880332887172699,
"step": 394
},
{
"epoch": 0.5290473798761092,
"grad_norm": 0.19377163489291163,
"learning_rate": 4.317564707296372e-05,
"loss": 0.8733063340187073,
"step": 395
},
{
"epoch": 0.5303867403314917,
"grad_norm": 0.20359933193356164,
"learning_rate": 4.298920374345698e-05,
"loss": 0.8691454529762268,
"step": 396
},
{
"epoch": 0.5317261007868743,
"grad_norm": 0.2228972219539793,
"learning_rate": 4.2802695083458694e-05,
"loss": 0.8702418208122253,
"step": 397
},
{
"epoch": 0.5330654612422568,
"grad_norm": 0.22204449924352906,
"learning_rate": 4.261612516920573e-05,
"loss": 0.8784664273262024,
"step": 398
},
{
"epoch": 0.5344048216976394,
"grad_norm": 0.1827068322619916,
"learning_rate": 4.242949807827365e-05,
"loss": 0.8793500065803528,
"step": 399
},
{
"epoch": 0.5357441821530219,
"grad_norm": 0.23064253884484145,
"learning_rate": 4.2242817889487676e-05,
"loss": 0.8827623128890991,
"step": 400
},
{
"epoch": 0.5370835426084045,
"grad_norm": 0.2363908476355385,
"learning_rate": 4.2056088682833495e-05,
"loss": 0.8775718212127686,
"step": 401
},
{
"epoch": 0.538422903063787,
"grad_norm": 0.24370104788240307,
"learning_rate": 4.18693145393681e-05,
"loss": 0.8720149993896484,
"step": 402
},
{
"epoch": 0.5397622635191696,
"grad_norm": 0.24206951622443026,
"learning_rate": 4.1682499541130614e-05,
"loss": 0.8702287077903748,
"step": 403
},
{
"epoch": 0.5411016239745522,
"grad_norm": 0.22053262939581833,
"learning_rate": 4.149564777105304e-05,
"loss": 0.88590008020401,
"step": 404
},
{
"epoch": 0.5424409844299347,
"grad_norm": 0.24738370911830754,
"learning_rate": 4.130876331287105e-05,
"loss": 0.8856267333030701,
"step": 405
},
{
"epoch": 0.5437803448853172,
"grad_norm": 0.2742407776361635,
"learning_rate": 4.112185025103476e-05,
"loss": 0.8771058917045593,
"step": 406
},
{
"epoch": 0.5451197053406999,
"grad_norm": 0.18672650383453882,
"learning_rate": 4.09349126706194e-05,
"loss": 0.877077579498291,
"step": 407
},
{
"epoch": 0.5464590657960824,
"grad_norm": 0.24422099387413226,
"learning_rate": 4.074795465723609e-05,
"loss": 0.8644412755966187,
"step": 408
},
{
"epoch": 0.5477984262514649,
"grad_norm": 0.22384826877195602,
"learning_rate": 4.0560980296942496e-05,
"loss": 0.8765589594841003,
"step": 409
},
{
"epoch": 0.5491377867068474,
"grad_norm": 0.18458733932067822,
"learning_rate": 4.0373993676153575e-05,
"loss": 0.8797497749328613,
"step": 410
},
{
"epoch": 0.5504771471622301,
"grad_norm": 0.23047653087943096,
"learning_rate": 4.0186998881552224e-05,
"loss": 0.8682187795639038,
"step": 411
},
{
"epoch": 0.5518165076176126,
"grad_norm": 0.21729249762981923,
"learning_rate": 4e-05,
"loss": 0.8836795091629028,
"step": 412
},
{
"epoch": 0.5531558680729951,
"grad_norm": 0.22721453642149825,
"learning_rate": 3.981300111844779e-05,
"loss": 0.8802202939987183,
"step": 413
},
{
"epoch": 0.5544952285283777,
"grad_norm": 0.20949925349974394,
"learning_rate": 3.962600632384644e-05,
"loss": 0.8641338348388672,
"step": 414
},
{
"epoch": 0.5558345889837603,
"grad_norm": 0.20464298350965635,
"learning_rate": 3.943901970305752e-05,
"loss": 0.8797010183334351,
"step": 415
},
{
"epoch": 0.5571739494391428,
"grad_norm": 0.22110350804377993,
"learning_rate": 3.925204534276392e-05,
"loss": 0.8767843246459961,
"step": 416
},
{
"epoch": 0.5585133098945254,
"grad_norm": 0.2387677867134895,
"learning_rate": 3.906508732938061e-05,
"loss": 0.8820819854736328,
"step": 417
},
{
"epoch": 0.5598526703499079,
"grad_norm": 0.2507989336100575,
"learning_rate": 3.8878149748965245e-05,
"loss": 0.8826998472213745,
"step": 418
},
{
"epoch": 0.5611920308052905,
"grad_norm": 0.20816367910535047,
"learning_rate": 3.869123668712896e-05,
"loss": 0.8565121293067932,
"step": 419
},
{
"epoch": 0.562531391260673,
"grad_norm": 0.2250240614615784,
"learning_rate": 3.850435222894698e-05,
"loss": 0.8700610399246216,
"step": 420
},
{
"epoch": 0.5638707517160556,
"grad_norm": 0.21855522780024444,
"learning_rate": 3.8317500458869407e-05,
"loss": 0.8730891942977905,
"step": 421
},
{
"epoch": 0.5652101121714381,
"grad_norm": 0.2231654916323255,
"learning_rate": 3.8130685460631915e-05,
"loss": 0.8856726288795471,
"step": 422
},
{
"epoch": 0.5665494726268206,
"grad_norm": 0.19953733222134984,
"learning_rate": 3.794391131716651e-05,
"loss": 0.8693549036979675,
"step": 423
},
{
"epoch": 0.5678888330822033,
"grad_norm": 0.17190873502705367,
"learning_rate": 3.775718211051233e-05,
"loss": 0.8690682649612427,
"step": 424
},
{
"epoch": 0.5692281935375858,
"grad_norm": 0.23716881942612886,
"learning_rate": 3.757050192172635e-05,
"loss": 0.8726681470870972,
"step": 425
},
{
"epoch": 0.5705675539929683,
"grad_norm": 0.22290256643263015,
"learning_rate": 3.738387483079428e-05,
"loss": 0.879726767539978,
"step": 426
},
{
"epoch": 0.571906914448351,
"grad_norm": 0.18381395165951167,
"learning_rate": 3.7197304916541306e-05,
"loss": 0.8783985376358032,
"step": 427
},
{
"epoch": 0.5732462749037335,
"grad_norm": 0.22765000500597882,
"learning_rate": 3.7010796256543034e-05,
"loss": 0.878974437713623,
"step": 428
},
{
"epoch": 0.574585635359116,
"grad_norm": 0.22163007061538173,
"learning_rate": 3.682435292703629e-05,
"loss": 0.8713866472244263,
"step": 429
},
{
"epoch": 0.5759249958144986,
"grad_norm": 0.20371172723764758,
"learning_rate": 3.6637979002830106e-05,
"loss": 0.8796266317367554,
"step": 430
},
{
"epoch": 0.5772643562698812,
"grad_norm": 0.1998206764522067,
"learning_rate": 3.6451678557216594e-05,
"loss": 0.8608413934707642,
"step": 431
},
{
"epoch": 0.5786037167252637,
"grad_norm": 0.21874058996853543,
"learning_rate": 3.626545566188198e-05,
"loss": 0.8731131553649902,
"step": 432
},
{
"epoch": 0.5799430771806462,
"grad_norm": 0.20785679907222504,
"learning_rate": 3.6079314386817575e-05,
"loss": 0.8750489354133606,
"step": 433
},
{
"epoch": 0.5812824376360288,
"grad_norm": 0.2139826824510004,
"learning_rate": 3.589325880023087e-05,
"loss": 0.867850661277771,
"step": 434
},
{
"epoch": 0.5826217980914113,
"grad_norm": 0.2058397086852482,
"learning_rate": 3.5707292968456544e-05,
"loss": 0.8824687600135803,
"step": 435
},
{
"epoch": 0.5839611585467939,
"grad_norm": 0.2234799265297003,
"learning_rate": 3.552142095586769e-05,
"loss": 0.8690378665924072,
"step": 436
},
{
"epoch": 0.5853005190021765,
"grad_norm": 0.19760641522762626,
"learning_rate": 3.533564682478691e-05,
"loss": 0.8751097917556763,
"step": 437
},
{
"epoch": 0.586639879457559,
"grad_norm": 0.2334335898022528,
"learning_rate": 3.5149974635397547e-05,
"loss": 0.8742689490318298,
"step": 438
},
{
"epoch": 0.5879792399129415,
"grad_norm": 0.2095792662279493,
"learning_rate": 3.4964408445654996e-05,
"loss": 0.8669417500495911,
"step": 439
},
{
"epoch": 0.5893186003683242,
"grad_norm": 0.21099729245604798,
"learning_rate": 3.477895231119795e-05,
"loss": 0.8878942728042603,
"step": 440
},
{
"epoch": 0.5906579608237067,
"grad_norm": 0.19382668238270218,
"learning_rate": 3.459361028525979e-05,
"loss": 0.8797996044158936,
"step": 441
},
{
"epoch": 0.5919973212790892,
"grad_norm": 0.1843091422043857,
"learning_rate": 3.4408386418580036e-05,
"loss": 0.8693681955337524,
"step": 442
},
{
"epoch": 0.5933366817344717,
"grad_norm": 0.20683570676709698,
"learning_rate": 3.422328475931573e-05,
"loss": 0.8908190727233887,
"step": 443
},
{
"epoch": 0.5946760421898544,
"grad_norm": 0.2057924792370247,
"learning_rate": 3.403830935295302e-05,
"loss": 0.8776031136512756,
"step": 444
},
{
"epoch": 0.5960154026452369,
"grad_norm": 0.17532326111046756,
"learning_rate": 3.3853464242218794e-05,
"loss": 0.8776813745498657,
"step": 445
},
{
"epoch": 0.5973547631006194,
"grad_norm": 0.21443600648578945,
"learning_rate": 3.3668753466992205e-05,
"loss": 0.8645883798599243,
"step": 446
},
{
"epoch": 0.598694123556002,
"grad_norm": 0.2124598468576842,
"learning_rate": 3.3484181064216456e-05,
"loss": 0.8678313493728638,
"step": 447
},
{
"epoch": 0.6000334840113846,
"grad_norm": 0.20402524205810743,
"learning_rate": 3.329975106781055e-05,
"loss": 0.8687726259231567,
"step": 448
},
{
"epoch": 0.6013728444667671,
"grad_norm": 0.18612685972506135,
"learning_rate": 3.311546750858113e-05,
"loss": 0.8742750287055969,
"step": 449
},
{
"epoch": 0.6027122049221497,
"grad_norm": 0.1891108432413768,
"learning_rate": 3.2931334414134405e-05,
"loss": 0.8701868057250977,
"step": 450
},
{
"epoch": 0.6040515653775322,
"grad_norm": 0.16982398331582746,
"learning_rate": 3.2747355808788076e-05,
"loss": 0.8736456632614136,
"step": 451
},
{
"epoch": 0.6053909258329148,
"grad_norm": 0.17659874457896263,
"learning_rate": 3.256353571348342e-05,
"loss": 0.865876317024231,
"step": 452
},
{
"epoch": 0.6067302862882974,
"grad_norm": 0.17274518370972755,
"learning_rate": 3.23798781456974e-05,
"loss": 0.8682874441146851,
"step": 453
},
{
"epoch": 0.6080696467436799,
"grad_norm": 0.18370111962886002,
"learning_rate": 3.219638711935488e-05,
"loss": 0.8759806156158447,
"step": 454
},
{
"epoch": 0.6094090071990624,
"grad_norm": 0.18017141493044456,
"learning_rate": 3.201306664474083e-05,
"loss": 0.8729393482208252,
"step": 455
},
{
"epoch": 0.610748367654445,
"grad_norm": 0.17913777693968835,
"learning_rate": 3.182992072841279e-05,
"loss": 0.8672225475311279,
"step": 456
},
{
"epoch": 0.6120877281098276,
"grad_norm": 0.18273030607069873,
"learning_rate": 3.164695337311321e-05,
"loss": 0.8832433223724365,
"step": 457
},
{
"epoch": 0.6134270885652101,
"grad_norm": 0.16888489983875138,
"learning_rate": 3.146416857768201e-05,
"loss": 0.8711962699890137,
"step": 458
},
{
"epoch": 0.6147664490205926,
"grad_norm": 0.183689134893473,
"learning_rate": 3.1281570336969175e-05,
"loss": 0.8889421224594116,
"step": 459
},
{
"epoch": 0.6161058094759753,
"grad_norm": 0.19643613570894447,
"learning_rate": 3.109916264174743e-05,
"loss": 0.8681093454360962,
"step": 460
},
{
"epoch": 0.6174451699313578,
"grad_norm": 0.19084392537480904,
"learning_rate": 3.091694947862508e-05,
"loss": 0.8883289098739624,
"step": 461
},
{
"epoch": 0.6187845303867403,
"grad_norm": 0.19455102639388797,
"learning_rate": 3.07349348299588e-05,
"loss": 0.8684327602386475,
"step": 462
},
{
"epoch": 0.6201238908421229,
"grad_norm": 0.2110042976412982,
"learning_rate": 3.055312267376664e-05,
"loss": 0.8856212496757507,
"step": 463
},
{
"epoch": 0.6214632512975055,
"grad_norm": 0.20264326773939578,
"learning_rate": 3.0371516983641092e-05,
"loss": 0.8823476433753967,
"step": 464
},
{
"epoch": 0.622802611752888,
"grad_norm": 0.2408395898053557,
"learning_rate": 3.0190121728662202e-05,
"loss": 0.8705859780311584,
"step": 465
},
{
"epoch": 0.6241419722082705,
"grad_norm": 0.16660600272013446,
"learning_rate": 3.000894087331092e-05,
"loss": 0.8687702417373657,
"step": 466
},
{
"epoch": 0.6254813326636531,
"grad_norm": 0.1833218068837441,
"learning_rate": 2.9827978377382326e-05,
"loss": 0.886631965637207,
"step": 467
},
{
"epoch": 0.6268206931190357,
"grad_norm": 0.16224401095054353,
"learning_rate": 2.9647238195899168e-05,
"loss": 0.872834324836731,
"step": 468
},
{
"epoch": 0.6281600535744182,
"grad_norm": 0.19626912635658397,
"learning_rate": 2.946672427902541e-05,
"loss": 0.8802139759063721,
"step": 469
},
{
"epoch": 0.6294994140298008,
"grad_norm": 0.18658582323373188,
"learning_rate": 2.9286440571979883e-05,
"loss": 0.8691961169242859,
"step": 470
},
{
"epoch": 0.6308387744851833,
"grad_norm": 0.17933302895676084,
"learning_rate": 2.910639101495008e-05,
"loss": 0.8831414580345154,
"step": 471
},
{
"epoch": 0.6321781349405659,
"grad_norm": 0.21056145987820624,
"learning_rate": 2.892657954300603e-05,
"loss": 0.8592995405197144,
"step": 472
},
{
"epoch": 0.6335174953959485,
"grad_norm": 0.19348734309236396,
"learning_rate": 2.87470100860143e-05,
"loss": 0.8894145488739014,
"step": 473
},
{
"epoch": 0.634856855851331,
"grad_norm": 0.19617888436938632,
"learning_rate": 2.85676865685521e-05,
"loss": 0.8913651704788208,
"step": 474
},
{
"epoch": 0.6361962163067135,
"grad_norm": 0.17402543550384714,
"learning_rate": 2.8388612909821515e-05,
"loss": 0.8747457265853882,
"step": 475
},
{
"epoch": 0.637535576762096,
"grad_norm": 0.2012596289013423,
"learning_rate": 2.8209793023563833e-05,
"loss": 0.8532263040542603,
"step": 476
},
{
"epoch": 0.6388749372174787,
"grad_norm": 0.16632768987251,
"learning_rate": 2.8031230817974058e-05,
"loss": 0.867146372795105,
"step": 477
},
{
"epoch": 0.6402142976728612,
"grad_norm": 0.17609764997939167,
"learning_rate": 2.7852930195615413e-05,
"loss": 0.873975396156311,
"step": 478
},
{
"epoch": 0.6415536581282437,
"grad_norm": 0.16290767187768426,
"learning_rate": 2.767489505333413e-05,
"loss": 0.8610473871231079,
"step": 479
},
{
"epoch": 0.6428930185836264,
"grad_norm": 0.15305292702967407,
"learning_rate": 2.7497129282174228e-05,
"loss": 0.8679625988006592,
"step": 480
},
{
"epoch": 0.6442323790390089,
"grad_norm": 0.1861499687391075,
"learning_rate": 2.7319636767292483e-05,
"loss": 0.8768002986907959,
"step": 481
},
{
"epoch": 0.6455717394943914,
"grad_norm": 0.1682720928607583,
"learning_rate": 2.7142421387873548e-05,
"loss": 0.8597822189331055,
"step": 482
},
{
"epoch": 0.646911099949774,
"grad_norm": 0.16548537206188335,
"learning_rate": 2.6965487017045114e-05,
"loss": 0.8771868348121643,
"step": 483
},
{
"epoch": 0.6482504604051565,
"grad_norm": 0.1551493965377178,
"learning_rate": 2.678883752179333e-05,
"loss": 0.8683747053146362,
"step": 484
},
{
"epoch": 0.6495898208605391,
"grad_norm": 0.17612866007244074,
"learning_rate": 2.661247676287822e-05,
"loss": 0.8681081533432007,
"step": 485
},
{
"epoch": 0.6509291813159217,
"grad_norm": 0.18465671357420232,
"learning_rate": 2.6436408594749332e-05,
"loss": 0.8595088720321655,
"step": 486
},
{
"epoch": 0.6522685417713042,
"grad_norm": 0.19775818406535858,
"learning_rate": 2.6260636865461554e-05,
"loss": 0.8700419664382935,
"step": 487
},
{
"epoch": 0.6536079022266867,
"grad_norm": 0.18083686062565354,
"learning_rate": 2.608516541659091e-05,
"loss": 0.8755306005477905,
"step": 488
},
{
"epoch": 0.6549472626820693,
"grad_norm": 0.1817072211213672,
"learning_rate": 2.5909998083150662e-05,
"loss": 0.875275731086731,
"step": 489
},
{
"epoch": 0.6562866231374519,
"grad_norm": 0.20261985924363612,
"learning_rate": 2.573513869350748e-05,
"loss": 0.8786300420761108,
"step": 490
},
{
"epoch": 0.6576259835928344,
"grad_norm": 0.17226016407298883,
"learning_rate": 2.556059106929778e-05,
"loss": 0.8597031831741333,
"step": 491
},
{
"epoch": 0.6589653440482169,
"grad_norm": 0.1827246814121758,
"learning_rate": 2.53863590253442e-05,
"loss": 0.8801453113555908,
"step": 492
},
{
"epoch": 0.6603047045035996,
"grad_norm": 0.16642060261390323,
"learning_rate": 2.5212446369572203e-05,
"loss": 0.8641231060028076,
"step": 493
},
{
"epoch": 0.6616440649589821,
"grad_norm": 0.1611151525632456,
"learning_rate": 2.5038856902926874e-05,
"loss": 0.868855357170105,
"step": 494
},
{
"epoch": 0.6629834254143646,
"grad_norm": 0.16125774426215733,
"learning_rate": 2.4865594419289862e-05,
"loss": 0.8626859188079834,
"step": 495
},
{
"epoch": 0.6643227858697472,
"grad_norm": 0.1676315935693186,
"learning_rate": 2.4692662705396412e-05,
"loss": 0.8734741806983948,
"step": 496
},
{
"epoch": 0.6656621463251298,
"grad_norm": 0.15126849376524884,
"learning_rate": 2.4520065540752668e-05,
"loss": 0.8615834712982178,
"step": 497
},
{
"epoch": 0.6670015067805123,
"grad_norm": 0.18347321722525894,
"learning_rate": 2.434780669755301e-05,
"loss": 0.8751839399337769,
"step": 498
},
{
"epoch": 0.6683408672358948,
"grad_norm": 0.1508841019626686,
"learning_rate": 2.4175889940597657e-05,
"loss": 0.884482741355896,
"step": 499
},
{
"epoch": 0.6696802276912774,
"grad_norm": 0.19960327521635535,
"learning_rate": 2.400431902721037e-05,
"loss": 0.8566937446594238,
"step": 500
},
{
"epoch": 0.67101958814666,
"grad_norm": 0.14443772776443664,
"learning_rate": 2.3833097707156324e-05,
"loss": 0.8727964162826538,
"step": 501
},
{
"epoch": 0.6723589486020425,
"grad_norm": 0.16725330298938068,
"learning_rate": 2.366222972256016e-05,
"loss": 0.8809594511985779,
"step": 502
},
{
"epoch": 0.6736983090574251,
"grad_norm": 0.16781357308473263,
"learning_rate": 2.3491718807824216e-05,
"loss": 0.8673738241195679,
"step": 503
},
{
"epoch": 0.6750376695128076,
"grad_norm": 0.17570747562365038,
"learning_rate": 2.3321568689546893e-05,
"loss": 0.8616470098495483,
"step": 504
},
{
"epoch": 0.6763770299681902,
"grad_norm": 0.16051510536372204,
"learning_rate": 2.3151783086441184e-05,
"loss": 0.8740379810333252,
"step": 505
},
{
"epoch": 0.6777163904235728,
"grad_norm": 0.18129021585072444,
"learning_rate": 2.298236570925346e-05,
"loss": 0.8758866786956787,
"step": 506
},
{
"epoch": 0.6790557508789553,
"grad_norm": 0.16295406262498888,
"learning_rate": 2.2813320260682306e-05,
"loss": 0.867128312587738,
"step": 507
},
{
"epoch": 0.6803951113343378,
"grad_norm": 0.16478256891061058,
"learning_rate": 2.264465043529768e-05,
"loss": 0.8716765642166138,
"step": 508
},
{
"epoch": 0.6817344717897204,
"grad_norm": 0.15474659633991988,
"learning_rate": 2.2476359919460063e-05,
"loss": 0.8846920728683472,
"step": 509
},
{
"epoch": 0.683073832245103,
"grad_norm": 0.15329182470431504,
"learning_rate": 2.2308452391239958e-05,
"loss": 0.8774445056915283,
"step": 510
},
{
"epoch": 0.6844131927004855,
"grad_norm": 0.16056212194512698,
"learning_rate": 2.214093152033749e-05,
"loss": 0.8718781471252441,
"step": 511
},
{
"epoch": 0.685752553155868,
"grad_norm": 0.1575919454514403,
"learning_rate": 2.1973800968002184e-05,
"loss": 0.8818187713623047,
"step": 512
},
{
"epoch": 0.6870919136112507,
"grad_norm": 0.17718245777491798,
"learning_rate": 2.1807064386953024e-05,
"loss": 0.865309476852417,
"step": 513
},
{
"epoch": 0.6884312740666332,
"grad_norm": 0.15545488226689105,
"learning_rate": 2.1640725421298487e-05,
"loss": 0.8730646967887878,
"step": 514
},
{
"epoch": 0.6897706345220157,
"grad_norm": 0.15608129600810844,
"learning_rate": 2.1474787706457022e-05,
"loss": 0.8821252584457397,
"step": 515
},
{
"epoch": 0.6911099949773983,
"grad_norm": 0.13882389199813575,
"learning_rate": 2.130925486907752e-05,
"loss": 0.8648711442947388,
"step": 516
},
{
"epoch": 0.6924493554327809,
"grad_norm": 0.15450321733786967,
"learning_rate": 2.1144130526960093e-05,
"loss": 0.8667118549346924,
"step": 517
},
{
"epoch": 0.6937887158881634,
"grad_norm": 0.1374984241876389,
"learning_rate": 2.0979418288976996e-05,
"loss": 0.8739535808563232,
"step": 518
},
{
"epoch": 0.695128076343546,
"grad_norm": 0.14197049787475696,
"learning_rate": 2.0815121754993735e-05,
"loss": 0.8651376366615295,
"step": 519
},
{
"epoch": 0.6964674367989285,
"grad_norm": 0.15028526402034278,
"learning_rate": 2.065124451579041e-05,
"loss": 0.8664205074310303,
"step": 520
},
{
"epoch": 0.697806797254311,
"grad_norm": 0.13549391085379284,
"learning_rate": 2.0487790152983244e-05,
"loss": 0.8679447770118713,
"step": 521
},
{
"epoch": 0.6991461577096936,
"grad_norm": 0.1460017601006368,
"learning_rate": 2.0324762238946275e-05,
"loss": 0.8691471815109253,
"step": 522
},
{
"epoch": 0.7004855181650762,
"grad_norm": 0.1431846851216918,
"learning_rate": 2.0162164336733303e-05,
"loss": 0.8775564432144165,
"step": 523
},
{
"epoch": 0.7018248786204587,
"grad_norm": 0.13257622094406188,
"learning_rate": 2.0000000000000012e-05,
"loss": 0.8781810998916626,
"step": 524
},
{
"epoch": 0.7031642390758412,
"grad_norm": 0.1675549621603434,
"learning_rate": 1.9838272772926302e-05,
"loss": 0.8698958158493042,
"step": 525
},
{
"epoch": 0.7045035995312239,
"grad_norm": 0.13480027126291538,
"learning_rate": 1.9676986190138835e-05,
"loss": 0.8663254976272583,
"step": 526
},
{
"epoch": 0.7058429599866064,
"grad_norm": 0.15294108103460244,
"learning_rate": 1.9516143776633774e-05,
"loss": 0.8629517555236816,
"step": 527
},
{
"epoch": 0.7071823204419889,
"grad_norm": 0.14520176029650325,
"learning_rate": 1.935574904769973e-05,
"loss": 0.8624333143234253,
"step": 528
},
{
"epoch": 0.7085216808973716,
"grad_norm": 0.134713885922547,
"learning_rate": 1.9195805508841013e-05,
"loss": 0.8724282383918762,
"step": 529
},
{
"epoch": 0.7098610413527541,
"grad_norm": 0.15108114436487047,
"learning_rate": 1.9036316655700893e-05,
"loss": 0.8668349981307983,
"step": 530
},
{
"epoch": 0.7112004018081366,
"grad_norm": 0.13671066235890095,
"learning_rate": 1.8877285973985282e-05,
"loss": 0.8796319961547852,
"step": 531
},
{
"epoch": 0.7125397622635191,
"grad_norm": 0.13002954166230035,
"learning_rate": 1.8718716939386543e-05,
"loss": 0.8816193342208862,
"step": 532
},
{
"epoch": 0.7138791227189017,
"grad_norm": 0.149579455889861,
"learning_rate": 1.8560613017507506e-05,
"loss": 0.8821731209754944,
"step": 533
},
{
"epoch": 0.7152184831742843,
"grad_norm": 0.13743335640356336,
"learning_rate": 1.840297766378575e-05,
"loss": 0.8548274040222168,
"step": 534
},
{
"epoch": 0.7165578436296668,
"grad_norm": 0.14190992365830207,
"learning_rate": 1.8245814323418063e-05,
"loss": 0.8705548048019409,
"step": 535
},
{
"epoch": 0.7178972040850494,
"grad_norm": 0.14016507952842097,
"learning_rate": 1.8089126431285152e-05,
"loss": 0.8656442165374756,
"step": 536
},
{
"epoch": 0.7192365645404319,
"grad_norm": 0.1504281446801305,
"learning_rate": 1.7932917411876586e-05,
"loss": 0.8762339353561401,
"step": 537
},
{
"epoch": 0.7205759249958145,
"grad_norm": 0.13435473240658485,
"learning_rate": 1.7777190679215923e-05,
"loss": 0.8614360690116882,
"step": 538
},
{
"epoch": 0.7219152854511971,
"grad_norm": 0.1536602881696007,
"learning_rate": 1.7621949636786104e-05,
"loss": 0.8601592779159546,
"step": 539
},
{
"epoch": 0.7232546459065796,
"grad_norm": 0.1530233513972156,
"learning_rate": 1.7467197677455118e-05,
"loss": 0.864124059677124,
"step": 540
},
{
"epoch": 0.7245940063619621,
"grad_norm": 0.14199812212808016,
"learning_rate": 1.7312938183401765e-05,
"loss": 0.8773407936096191,
"step": 541
},
{
"epoch": 0.7259333668173448,
"grad_norm": 0.15995982411725335,
"learning_rate": 1.7159174526041794e-05,
"loss": 0.8686542510986328,
"step": 542
},
{
"epoch": 0.7272727272727273,
"grad_norm": 0.15354853553335845,
"learning_rate": 1.7005910065954197e-05,
"loss": 0.8662307858467102,
"step": 543
},
{
"epoch": 0.7286120877281098,
"grad_norm": 0.14078496335090188,
"learning_rate": 1.6853148152807774e-05,
"loss": 0.8632434606552124,
"step": 544
},
{
"epoch": 0.7299514481834923,
"grad_norm": 0.15261704641462037,
"learning_rate": 1.670089212528792e-05,
"loss": 0.8669806122779846,
"step": 545
},
{
"epoch": 0.731290808638875,
"grad_norm": 0.1513819134513577,
"learning_rate": 1.654914531102365e-05,
"loss": 0.8775612115859985,
"step": 546
},
{
"epoch": 0.7326301690942575,
"grad_norm": 0.1358534877229751,
"learning_rate": 1.639791102651488e-05,
"loss": 0.8712992072105408,
"step": 547
},
{
"epoch": 0.73396952954964,
"grad_norm": 0.16067628596578593,
"learning_rate": 1.6247192577059943e-05,
"loss": 0.8825572729110718,
"step": 548
},
{
"epoch": 0.7353088900050226,
"grad_norm": 0.13837364745245015,
"learning_rate": 1.6096993256683332e-05,
"loss": 0.868409276008606,
"step": 549
},
{
"epoch": 0.7366482504604052,
"grad_norm": 0.1488210527589111,
"learning_rate": 1.5947316348063764e-05,
"loss": 0.8783029913902283,
"step": 550
},
{
"epoch": 0.7379876109157877,
"grad_norm": 0.15531256043223096,
"learning_rate": 1.5798165122462355e-05,
"loss": 0.866256594657898,
"step": 551
},
{
"epoch": 0.7393269713711703,
"grad_norm": 0.130400681920341,
"learning_rate": 1.5649542839651175e-05,
"loss": 0.8711825609207153,
"step": 552
},
{
"epoch": 0.7406663318265528,
"grad_norm": 0.15097118961428962,
"learning_rate": 1.5501452747841997e-05,
"loss": 0.8813204765319824,
"step": 553
},
{
"epoch": 0.7420056922819354,
"grad_norm": 0.1444971041300422,
"learning_rate": 1.53538980836153e-05,
"loss": 0.875779390335083,
"step": 554
},
{
"epoch": 0.7433450527373179,
"grad_norm": 0.15983906195186567,
"learning_rate": 1.5206882071849545e-05,
"loss": 0.8741669654846191,
"step": 555
},
{
"epoch": 0.7446844131927005,
"grad_norm": 0.1355928911062347,
"learning_rate": 1.5060407925650662e-05,
"loss": 0.8598854541778564,
"step": 556
},
{
"epoch": 0.746023773648083,
"grad_norm": 0.13602562978268276,
"learning_rate": 1.4914478846281872e-05,
"loss": 0.8700109720230103,
"step": 557
},
{
"epoch": 0.7473631341034656,
"grad_norm": 0.12852632137304454,
"learning_rate": 1.4769098023093676e-05,
"loss": 0.8685686588287354,
"step": 558
},
{
"epoch": 0.7487024945588482,
"grad_norm": 0.13847511439052035,
"learning_rate": 1.4624268633454187e-05,
"loss": 0.8838005661964417,
"step": 559
},
{
"epoch": 0.7500418550142307,
"grad_norm": 0.12583676980637623,
"learning_rate": 1.4479993842679663e-05,
"loss": 0.8625704050064087,
"step": 560
},
{
"epoch": 0.7513812154696132,
"grad_norm": 0.128203485289489,
"learning_rate": 1.4336276803965338e-05,
"loss": 0.8662985563278198,
"step": 561
},
{
"epoch": 0.7527205759249959,
"grad_norm": 0.13653789826965723,
"learning_rate": 1.4193120658316506e-05,
"loss": 0.8692458868026733,
"step": 562
},
{
"epoch": 0.7540599363803784,
"grad_norm": 0.11863623059533325,
"learning_rate": 1.405052853447987e-05,
"loss": 0.8643911480903625,
"step": 563
},
{
"epoch": 0.7553992968357609,
"grad_norm": 0.13264464820938757,
"learning_rate": 1.3908503548875167e-05,
"loss": 0.8673012256622314,
"step": 564
},
{
"epoch": 0.7567386572911434,
"grad_norm": 0.12704980324875467,
"learning_rate": 1.3767048805527039e-05,
"loss": 0.862754225730896,
"step": 565
},
{
"epoch": 0.7580780177465261,
"grad_norm": 0.1238471642458571,
"learning_rate": 1.3626167395997247e-05,
"loss": 0.8769301772117615,
"step": 566
},
{
"epoch": 0.7594173782019086,
"grad_norm": 0.13783491088045657,
"learning_rate": 1.3485862399317027e-05,
"loss": 0.8871258497238159,
"step": 567
},
{
"epoch": 0.7607567386572911,
"grad_norm": 0.13031583245387215,
"learning_rate": 1.3346136881919845e-05,
"loss": 0.8605697154998779,
"step": 568
},
{
"epoch": 0.7620960991126737,
"grad_norm": 0.12766262505196324,
"learning_rate": 1.320699389757437e-05,
"loss": 0.8688619136810303,
"step": 569
},
{
"epoch": 0.7634354595680563,
"grad_norm": 0.13303758007462416,
"learning_rate": 1.3068436487317713e-05,
"loss": 0.8767508864402771,
"step": 570
},
{
"epoch": 0.7647748200234388,
"grad_norm": 0.13228607356250646,
"learning_rate": 1.2930467679389022e-05,
"loss": 0.8681412935256958,
"step": 571
},
{
"epoch": 0.7661141804788214,
"grad_norm": 0.12005222020555555,
"learning_rate": 1.2793090489163218e-05,
"loss": 0.8661867380142212,
"step": 572
},
{
"epoch": 0.7674535409342039,
"grad_norm": 0.13195468524460277,
"learning_rate": 1.2656307919085146e-05,
"loss": 0.8680728673934937,
"step": 573
},
{
"epoch": 0.7687929013895864,
"grad_norm": 0.1320165820638193,
"learning_rate": 1.2520122958603933e-05,
"loss": 0.868954062461853,
"step": 574
},
{
"epoch": 0.7701322618449691,
"grad_norm": 0.12717407261567398,
"learning_rate": 1.2384538584107668e-05,
"loss": 0.877269983291626,
"step": 575
},
{
"epoch": 0.7714716223003516,
"grad_norm": 0.13462181233364415,
"learning_rate": 1.2249557758858335e-05,
"loss": 0.8749093413352966,
"step": 576
},
{
"epoch": 0.7728109827557341,
"grad_norm": 0.13228322165397566,
"learning_rate": 1.2115183432927071e-05,
"loss": 0.868626594543457,
"step": 577
},
{
"epoch": 0.7741503432111166,
"grad_norm": 0.11359418193030604,
"learning_rate": 1.1981418543129668e-05,
"loss": 0.8568856716156006,
"step": 578
},
{
"epoch": 0.7754897036664993,
"grad_norm": 0.12638751116359853,
"learning_rate": 1.1848266012962393e-05,
"loss": 0.8607402443885803,
"step": 579
},
{
"epoch": 0.7768290641218818,
"grad_norm": 0.12528883435025034,
"learning_rate": 1.1715728752538103e-05,
"loss": 0.8683662414550781,
"step": 580
},
{
"epoch": 0.7781684245772643,
"grad_norm": 0.11779122344927996,
"learning_rate": 1.1583809658522643e-05,
"loss": 0.8854649066925049,
"step": 581
},
{
"epoch": 0.779507785032647,
"grad_norm": 0.11889377542793812,
"learning_rate": 1.1452511614071518e-05,
"loss": 0.8731423616409302,
"step": 582
},
{
"epoch": 0.7808471454880295,
"grad_norm": 0.12276953420922905,
"learning_rate": 1.1321837488766913e-05,
"loss": 0.8679988384246826,
"step": 583
},
{
"epoch": 0.782186505943412,
"grad_norm": 0.12170266578498971,
"learning_rate": 1.1191790138554945e-05,
"loss": 0.8663763999938965,
"step": 584
},
{
"epoch": 0.7835258663987946,
"grad_norm": 0.11319872062759567,
"learning_rate": 1.1062372405683263e-05,
"loss": 0.8651244044303894,
"step": 585
},
{
"epoch": 0.7848652268541771,
"grad_norm": 0.11810212796138198,
"learning_rate": 1.0933587118638927e-05,
"loss": 0.8747796416282654,
"step": 586
},
{
"epoch": 0.7862045873095597,
"grad_norm": 0.10887190880014788,
"learning_rate": 1.0805437092086586e-05,
"loss": 0.8664879202842712,
"step": 587
},
{
"epoch": 0.7875439477649422,
"grad_norm": 0.1060428677550431,
"learning_rate": 1.0677925126806956e-05,
"loss": 0.8619367480278015,
"step": 588
},
{
"epoch": 0.7888833082203248,
"grad_norm": 0.11316065290832693,
"learning_rate": 1.0551054009635636e-05,
"loss": 0.8755437135696411,
"step": 589
},
{
"epoch": 0.7902226686757073,
"grad_norm": 0.12213570051138428,
"learning_rate": 1.0424826513402162e-05,
"loss": 0.8632907867431641,
"step": 590
},
{
"epoch": 0.7915620291310899,
"grad_norm": 0.11811385521052746,
"learning_rate": 1.0299245396869418e-05,
"loss": 0.8566582202911377,
"step": 591
},
{
"epoch": 0.7929013895864725,
"grad_norm": 0.11870469860749323,
"learning_rate": 1.0174313404673378e-05,
"loss": 0.8652503490447998,
"step": 592
},
{
"epoch": 0.794240750041855,
"grad_norm": 0.11642632041514202,
"learning_rate": 1.0050033267263068e-05,
"loss": 0.8666040897369385,
"step": 593
},
{
"epoch": 0.7955801104972375,
"grad_norm": 0.11271123011165501,
"learning_rate": 9.92640770084091e-06,
"loss": 0.8616808652877808,
"step": 594
},
{
"epoch": 0.7969194709526202,
"grad_norm": 0.11624230367298656,
"learning_rate": 9.803439407303369e-06,
"loss": 0.8712943196296692,
"step": 595
},
{
"epoch": 0.7982588314080027,
"grad_norm": 0.11751534209024293,
"learning_rate": 9.681131074181876e-06,
"loss": 0.8700709342956543,
"step": 596
},
{
"epoch": 0.7995981918633852,
"grad_norm": 0.12175042858094896,
"learning_rate": 9.55948537458415e-06,
"loss": 0.8653758764266968,
"step": 597
},
{
"epoch": 0.8009375523187678,
"grad_norm": 0.11051730011718072,
"learning_rate": 9.438504967135703e-06,
"loss": 0.859759509563446,
"step": 598
},
{
"epoch": 0.8022769127741504,
"grad_norm": 0.11944757063536243,
"learning_rate": 9.318192495921754e-06,
"loss": 0.8628091216087341,
"step": 599
},
{
"epoch": 0.8036162732295329,
"grad_norm": 0.1300501966435939,
"learning_rate": 9.19855059042949e-06,
"loss": 0.8775379657745361,
"step": 600
},
{
"epoch": 0.8049556336849154,
"grad_norm": 0.11067723833940438,
"learning_rate": 9.07958186549052e-06,
"loss": 0.8566068410873413,
"step": 601
},
{
"epoch": 0.806294994140298,
"grad_norm": 0.12435798088723733,
"learning_rate": 8.961288921223805e-06,
"loss": 0.8680790066719055,
"step": 602
},
{
"epoch": 0.8076343545956806,
"grad_norm": 0.11526262371532592,
"learning_rate": 8.84367434297876e-06,
"loss": 0.8559243679046631,
"step": 603
},
{
"epoch": 0.8089737150510631,
"grad_norm": 0.11562289937311675,
"learning_rate": 8.72674070127881e-06,
"loss": 0.8632556796073914,
"step": 604
},
{
"epoch": 0.8103130755064457,
"grad_norm": 0.11157152293753178,
"learning_rate": 8.610490551765162e-06,
"loss": 0.8667713403701782,
"step": 605
},
{
"epoch": 0.8116524359618282,
"grad_norm": 0.11862861011909966,
"learning_rate": 8.494926435140986e-06,
"loss": 0.8680082559585571,
"step": 606
},
{
"epoch": 0.8129917964172108,
"grad_norm": 0.11058276447133751,
"learning_rate": 8.380050877115864e-06,
"loss": 0.8787844777107239,
"step": 607
},
{
"epoch": 0.8143311568725934,
"grad_norm": 0.11035182964574768,
"learning_rate": 8.265866388350598e-06,
"loss": 0.8844963312149048,
"step": 608
},
{
"epoch": 0.8156705173279759,
"grad_norm": 0.10573409759536821,
"learning_rate": 8.152375464402342e-06,
"loss": 0.859270453453064,
"step": 609
},
{
"epoch": 0.8170098777833584,
"grad_norm": 0.11228477530407104,
"learning_rate": 8.039580585670047e-06,
"loss": 0.8675272464752197,
"step": 610
},
{
"epoch": 0.818349238238741,
"grad_norm": 0.11021493931190712,
"learning_rate": 7.927484217340268e-06,
"loss": 0.8565207719802856,
"step": 611
},
{
"epoch": 0.8196885986941236,
"grad_norm": 0.11751800069891458,
"learning_rate": 7.816088809333266e-06,
"loss": 0.8686612844467163,
"step": 612
},
{
"epoch": 0.8210279591495061,
"grad_norm": 0.11099301486626775,
"learning_rate": 7.705396796249505e-06,
"loss": 0.8637645244598389,
"step": 613
},
{
"epoch": 0.8223673196048886,
"grad_norm": 0.10684585744907993,
"learning_rate": 7.595410597316379e-06,
"loss": 0.8606627583503723,
"step": 614
},
{
"epoch": 0.8237066800602713,
"grad_norm": 0.114479487802595,
"learning_rate": 7.486132616335395e-06,
"loss": 0.8715918064117432,
"step": 615
},
{
"epoch": 0.8250460405156538,
"grad_norm": 0.12357407311236206,
"learning_rate": 7.3775652416295936e-06,
"loss": 0.8811098337173462,
"step": 616
},
{
"epoch": 0.8263854009710363,
"grad_norm": 0.1018906001195981,
"learning_rate": 7.2697108459914e-06,
"loss": 0.8510515689849854,
"step": 617
},
{
"epoch": 0.8277247614264189,
"grad_norm": 0.11510793709687688,
"learning_rate": 7.162571786630707e-06,
"loss": 0.8802266120910645,
"step": 618
},
{
"epoch": 0.8290641218818015,
"grad_norm": 0.10985562417753968,
"learning_rate": 7.056150405123414e-06,
"loss": 0.8502565622329712,
"step": 619
},
{
"epoch": 0.830403482337184,
"grad_norm": 0.10600131076359613,
"learning_rate": 6.950449027360213e-06,
"loss": 0.8826883435249329,
"step": 620
},
{
"epoch": 0.8317428427925665,
"grad_norm": 0.11433471498911951,
"learning_rate": 6.8454699634957586e-06,
"loss": 0.8738099336624146,
"step": 621
},
{
"epoch": 0.8330822032479491,
"grad_norm": 0.10950279122420588,
"learning_rate": 6.7412155078981865e-06,
"loss": 0.8639442324638367,
"step": 622
},
{
"epoch": 0.8344215637033316,
"grad_norm": 0.10799080989032715,
"learning_rate": 6.6376879390990025e-06,
"loss": 0.8762117624282837,
"step": 623
},
{
"epoch": 0.8357609241587142,
"grad_norm": 0.10733902528411342,
"learning_rate": 6.53488951974321e-06,
"loss": 0.8581281900405884,
"step": 624
},
{
"epoch": 0.8371002846140968,
"grad_norm": 0.11176168812915092,
"learning_rate": 6.432822496539915e-06,
"loss": 0.8563675880432129,
"step": 625
},
{
"epoch": 0.8384396450694793,
"grad_norm": 0.10221663384828403,
"learning_rate": 6.331489100213204e-06,
"loss": 0.865772008895874,
"step": 626
},
{
"epoch": 0.8397790055248618,
"grad_norm": 0.09837435329440289,
"learning_rate": 6.230891545453395e-06,
"loss": 0.8748994469642639,
"step": 627
},
{
"epoch": 0.8411183659802445,
"grad_norm": 0.10807279392506593,
"learning_rate": 6.1310320308686354e-06,
"loss": 0.8680225014686584,
"step": 628
},
{
"epoch": 0.842457726435627,
"grad_norm": 0.09768064046733724,
"learning_rate": 6.031912738936849e-06,
"loss": 0.8678373098373413,
"step": 629
},
{
"epoch": 0.8437970868910095,
"grad_norm": 0.1015809029022958,
"learning_rate": 5.933535835958029e-06,
"loss": 0.8708404302597046,
"step": 630
},
{
"epoch": 0.8451364473463922,
"grad_norm": 0.10794002356946901,
"learning_rate": 5.835903472006905e-06,
"loss": 0.8744752407073975,
"step": 631
},
{
"epoch": 0.8464758078017747,
"grad_norm": 0.09772198649085652,
"learning_rate": 5.7390177808859475e-06,
"loss": 0.860190212726593,
"step": 632
},
{
"epoch": 0.8478151682571572,
"grad_norm": 0.11209427408989413,
"learning_rate": 5.642880880078721e-06,
"loss": 0.8820698261260986,
"step": 633
},
{
"epoch": 0.8491545287125397,
"grad_norm": 0.10860351131929465,
"learning_rate": 5.547494870703642e-06,
"loss": 0.8616386651992798,
"step": 634
},
{
"epoch": 0.8504938891679223,
"grad_norm": 0.10866684859584659,
"learning_rate": 5.452861837468005e-06,
"loss": 0.8636765480041504,
"step": 635
},
{
"epoch": 0.8518332496233049,
"grad_norm": 0.1006939328389969,
"learning_rate": 5.358983848622452e-06,
"loss": 0.8664864897727966,
"step": 636
},
{
"epoch": 0.8531726100786874,
"grad_norm": 0.10559916320249045,
"learning_rate": 5.2658629559157835e-06,
"loss": 0.8764059543609619,
"step": 637
},
{
"epoch": 0.85451197053407,
"grad_norm": 0.1052114260414204,
"learning_rate": 5.173501194550072e-06,
"loss": 0.8566386699676514,
"step": 638
},
{
"epoch": 0.8558513309894525,
"grad_norm": 0.10455890325177675,
"learning_rate": 5.081900583136228e-06,
"loss": 0.8712558746337891,
"step": 639
},
{
"epoch": 0.8571906914448351,
"grad_norm": 0.10238359327551756,
"learning_rate": 4.991063123649853e-06,
"loss": 0.8549286127090454,
"step": 640
},
{
"epoch": 0.8585300519002177,
"grad_norm": 0.1052929910920113,
"learning_rate": 4.900990801387497e-06,
"loss": 0.8717979788780212,
"step": 641
},
{
"epoch": 0.8598694123556002,
"grad_norm": 0.10156544713233225,
"learning_rate": 4.811685584923269e-06,
"loss": 0.8659064173698425,
"step": 642
},
{
"epoch": 0.8612087728109827,
"grad_norm": 0.09243659771376618,
"learning_rate": 4.723149426065803e-06,
"loss": 0.8679224252700806,
"step": 643
},
{
"epoch": 0.8625481332663653,
"grad_norm": 0.09700721513112183,
"learning_rate": 4.635384259815614e-06,
"loss": 0.8493891954421997,
"step": 644
},
{
"epoch": 0.8638874937217479,
"grad_norm": 0.10574690432772352,
"learning_rate": 4.548392004322799e-06,
"loss": 0.8844394087791443,
"step": 645
},
{
"epoch": 0.8652268541771304,
"grad_norm": 0.10280067073915403,
"learning_rate": 4.462174560845114e-06,
"loss": 0.851788341999054,
"step": 646
},
{
"epoch": 0.8665662146325129,
"grad_norm": 0.0944681029234863,
"learning_rate": 4.376733813706433e-06,
"loss": 0.880761981010437,
"step": 647
},
{
"epoch": 0.8679055750878956,
"grad_norm": 0.10350747616416575,
"learning_rate": 4.292071630255548e-06,
"loss": 0.8677074909210205,
"step": 648
},
{
"epoch": 0.8692449355432781,
"grad_norm": 0.09650967181000819,
"learning_rate": 4.208189860825358e-06,
"loss": 0.8634464740753174,
"step": 649
},
{
"epoch": 0.8705842959986606,
"grad_norm": 0.09916241002259231,
"learning_rate": 4.12509033869247e-06,
"loss": 0.8697370290756226,
"step": 650
},
{
"epoch": 0.8719236564540432,
"grad_norm": 0.10108324879778724,
"learning_rate": 4.042774880037059e-06,
"loss": 0.8599544763565063,
"step": 651
},
{
"epoch": 0.8732630169094258,
"grad_norm": 0.09243394131495927,
"learning_rate": 3.961245283903239e-06,
"loss": 0.8817688226699829,
"step": 652
},
{
"epoch": 0.8746023773648083,
"grad_norm": 0.10331641148619426,
"learning_rate": 3.8805033321597015e-06,
"loss": 0.8564207553863525,
"step": 653
},
{
"epoch": 0.8759417378201909,
"grad_norm": 0.10558580045435674,
"learning_rate": 3.8005507894607997e-06,
"loss": 0.860696017742157,
"step": 654
},
{
"epoch": 0.8772810982755734,
"grad_norm": 0.10444551296183989,
"learning_rate": 3.7213894032079778e-06,
"loss": 0.8789005279541016,
"step": 655
},
{
"epoch": 0.878620458730956,
"grad_norm": 0.0969379244894827,
"learning_rate": 3.643020903511563e-06,
"loss": 0.8776953220367432,
"step": 656
},
{
"epoch": 0.8799598191863385,
"grad_norm": 0.10790624806273608,
"learning_rate": 3.565447003152964e-06,
"loss": 0.8571181297302246,
"step": 657
},
{
"epoch": 0.8812991796417211,
"grad_norm": 0.10477691549480385,
"learning_rate": 3.4886693975472443e-06,
"loss": 0.8560389280319214,
"step": 658
},
{
"epoch": 0.8826385400971036,
"grad_norm": 0.09813559375826432,
"learning_rate": 3.4126897647060474e-06,
"loss": 0.8699684143066406,
"step": 659
},
{
"epoch": 0.8839779005524862,
"grad_norm": 0.09813832554601809,
"learning_rate": 3.3375097652009526e-06,
"loss": 0.8595161437988281,
"step": 660
},
{
"epoch": 0.8853172610078688,
"grad_norm": 0.09587458422790734,
"learning_rate": 3.263131042127157e-06,
"loss": 0.8548429012298584,
"step": 661
},
{
"epoch": 0.8866566214632513,
"grad_norm": 0.09801177294143108,
"learning_rate": 3.189555221067577e-06,
"loss": 0.8632456064224243,
"step": 662
},
{
"epoch": 0.8879959819186338,
"grad_norm": 0.09722647195070884,
"learning_rate": 3.1167839100573194e-06,
"loss": 0.8654548525810242,
"step": 663
},
{
"epoch": 0.8893353423740165,
"grad_norm": 0.1030729011054221,
"learning_rate": 3.0448186995485307e-06,
"loss": 0.865413248538971,
"step": 664
},
{
"epoch": 0.890674702829399,
"grad_norm": 0.09159907373518272,
"learning_rate": 2.973661162375647e-06,
"loss": 0.8626821041107178,
"step": 665
},
{
"epoch": 0.8920140632847815,
"grad_norm": 0.09822649096101539,
"learning_rate": 2.9033128537210077e-06,
"loss": 0.8579151630401611,
"step": 666
},
{
"epoch": 0.893353423740164,
"grad_norm": 0.09554990867644066,
"learning_rate": 2.8337753110808753e-06,
"loss": 0.8664767742156982,
"step": 667
},
{
"epoch": 0.8946927841955467,
"grad_norm": 0.0940544920631781,
"learning_rate": 2.765050054231835e-06,
"loss": 0.8667508363723755,
"step": 668
},
{
"epoch": 0.8960321446509292,
"grad_norm": 0.09505186941767597,
"learning_rate": 2.6971385851975608e-06,
"loss": 0.8664491176605225,
"step": 669
},
{
"epoch": 0.8973715051063117,
"grad_norm": 0.09640496159648018,
"learning_rate": 2.630042388216012e-06,
"loss": 0.8690519332885742,
"step": 670
},
{
"epoch": 0.8987108655616943,
"grad_norm": 0.09338166771434768,
"learning_rate": 2.563762929706979e-06,
"loss": 0.8708868622779846,
"step": 671
},
{
"epoch": 0.9000502260170768,
"grad_norm": 0.09032253422132287,
"learning_rate": 2.4983016582400366e-06,
"loss": 0.8571633100509644,
"step": 672
},
{
"epoch": 0.9013895864724594,
"grad_norm": 0.09985352924424004,
"learning_rate": 2.4336600045028956e-06,
"loss": 0.8728021383285522,
"step": 673
},
{
"epoch": 0.902728946927842,
"grad_norm": 0.09095804473388606,
"learning_rate": 2.3698393812701114e-06,
"loss": 0.8710513114929199,
"step": 674
},
{
"epoch": 0.9040683073832245,
"grad_norm": 0.0907778752241714,
"learning_rate": 2.3068411833722283e-06,
"loss": 0.8682019710540771,
"step": 675
},
{
"epoch": 0.905407667838607,
"grad_norm": 0.08725840048506821,
"learning_rate": 2.244666787665297e-06,
"loss": 0.8690208792686462,
"step": 676
},
{
"epoch": 0.9067470282939896,
"grad_norm": 0.08769527712682709,
"learning_rate": 2.183317553000763e-06,
"loss": 0.8633980751037598,
"step": 677
},
{
"epoch": 0.9080863887493722,
"grad_norm": 0.09024779836807585,
"learning_rate": 2.122794820195777e-06,
"loss": 0.8639845848083496,
"step": 678
},
{
"epoch": 0.9094257492047547,
"grad_norm": 0.09693190682746106,
"learning_rate": 2.063099912003894e-06,
"loss": 0.883812665939331,
"step": 679
},
{
"epoch": 0.9107651096601372,
"grad_norm": 0.0861902022397013,
"learning_rate": 2.0042341330861605e-06,
"loss": 0.8602584600448608,
"step": 680
},
{
"epoch": 0.9121044701155199,
"grad_norm": 0.09371370939823905,
"learning_rate": 1.9461987699826147e-06,
"loss": 0.877825140953064,
"step": 681
},
{
"epoch": 0.9134438305709024,
"grad_norm": 0.09042366335672848,
"learning_rate": 1.888995091084147e-06,
"loss": 0.8823691010475159,
"step": 682
},
{
"epoch": 0.9147831910262849,
"grad_norm": 0.08752488797811156,
"learning_rate": 1.8326243466047833e-06,
"loss": 0.8719531297683716,
"step": 683
},
{
"epoch": 0.9161225514816675,
"grad_norm": 0.09838692247268259,
"learning_rate": 1.7770877685543687e-06,
"loss": 0.8698108792304993,
"step": 684
},
{
"epoch": 0.9174619119370501,
"grad_norm": 0.08608689280980435,
"learning_rate": 1.7223865707116472e-06,
"loss": 0.8581531047821045,
"step": 685
},
{
"epoch": 0.9188012723924326,
"grad_norm": 0.08652624883389393,
"learning_rate": 1.6685219485977056e-06,
"loss": 0.8714857697486877,
"step": 686
},
{
"epoch": 0.9201406328478152,
"grad_norm": 0.08764678535346417,
"learning_rate": 1.615495079449887e-06,
"loss": 0.8779809474945068,
"step": 687
},
{
"epoch": 0.9214799933031977,
"grad_norm": 0.08867186936205497,
"learning_rate": 1.5633071221960205e-06,
"loss": 0.8661234378814697,
"step": 688
},
{
"epoch": 0.9228193537585803,
"grad_norm": 0.0901392673827971,
"learning_rate": 1.5119592174291175e-06,
"loss": 0.8790245056152344,
"step": 689
},
{
"epoch": 0.9241587142139628,
"grad_norm": 0.10070892675692567,
"learning_rate": 1.4614524873824398e-06,
"loss": 0.8705023527145386,
"step": 690
},
{
"epoch": 0.9254980746693454,
"grad_norm": 0.09379245050344401,
"learning_rate": 1.4117880359049684e-06,
"loss": 0.8885618448257446,
"step": 691
},
{
"epoch": 0.9268374351247279,
"grad_norm": 0.09167671555096236,
"learning_rate": 1.3629669484372722e-06,
"loss": 0.8792763948440552,
"step": 692
},
{
"epoch": 0.9281767955801105,
"grad_norm": 0.08805121953326282,
"learning_rate": 1.3149902919877966e-06,
"loss": 0.8643807172775269,
"step": 693
},
{
"epoch": 0.9295161560354931,
"grad_norm": 0.0922229306076492,
"learning_rate": 1.2678591151095466e-06,
"loss": 0.862453281879425,
"step": 694
},
{
"epoch": 0.9308555164908756,
"grad_norm": 0.09307453015713951,
"learning_rate": 1.2215744478771473e-06,
"loss": 0.8628978729248047,
"step": 695
},
{
"epoch": 0.9321948769462581,
"grad_norm": 0.09390847242485158,
"learning_rate": 1.176137301864362e-06,
"loss": 0.8640226721763611,
"step": 696
},
{
"epoch": 0.9335342374016408,
"grad_norm": 0.09047699788223823,
"learning_rate": 1.1315486701219603e-06,
"loss": 0.8615423440933228,
"step": 697
},
{
"epoch": 0.9348735978570233,
"grad_norm": 0.09225121912036802,
"learning_rate": 1.0878095271560274e-06,
"loss": 0.8629521131515503,
"step": 698
},
{
"epoch": 0.9362129583124058,
"grad_norm": 0.09468262537112955,
"learning_rate": 1.0449208289066549e-06,
"loss": 0.8680487275123596,
"step": 699
},
{
"epoch": 0.9375523187677883,
"grad_norm": 0.08864131732671222,
"learning_rate": 1.0028835127270553e-06,
"loss": 0.8672420382499695,
"step": 700
}
],
"logging_steps": 1,
"max_steps": 747,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"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": 5327911649280000.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}