SKNahin's picture
End of training
2a9ec5f verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.9965635738831615,
"eval_steps": 500,
"global_step": 580,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00859106529209622,
"grad_norm": 2.208123786139425,
"learning_rate": 1.6666666666666667e-06,
"loss": 2.519,
"step": 1
},
{
"epoch": 0.01718213058419244,
"grad_norm": 2.387216471135648,
"learning_rate": 3.3333333333333333e-06,
"loss": 2.5546,
"step": 2
},
{
"epoch": 0.02577319587628866,
"grad_norm": 2.242414180089336,
"learning_rate": 5e-06,
"loss": 2.5143,
"step": 3
},
{
"epoch": 0.03436426116838488,
"grad_norm": 1.3834371622684507,
"learning_rate": 6.666666666666667e-06,
"loss": 2.4234,
"step": 4
},
{
"epoch": 0.0429553264604811,
"grad_norm": 0.8480429507160056,
"learning_rate": 8.333333333333334e-06,
"loss": 2.3533,
"step": 5
},
{
"epoch": 0.05154639175257732,
"grad_norm": 1.0257621263522265,
"learning_rate": 1e-05,
"loss": 2.3478,
"step": 6
},
{
"epoch": 0.06013745704467354,
"grad_norm": 0.9465705617705457,
"learning_rate": 9.999925111542544e-06,
"loss": 2.3559,
"step": 7
},
{
"epoch": 0.06872852233676977,
"grad_norm": 0.8290324702503862,
"learning_rate": 9.999700448413483e-06,
"loss": 2.2821,
"step": 8
},
{
"epoch": 0.07731958762886598,
"grad_norm": 0.7616359146631583,
"learning_rate": 9.999326017342688e-06,
"loss": 2.2554,
"step": 9
},
{
"epoch": 0.0859106529209622,
"grad_norm": 0.7466927981003729,
"learning_rate": 9.998801829546387e-06,
"loss": 2.2472,
"step": 10
},
{
"epoch": 0.09450171821305842,
"grad_norm": 0.4299476068045872,
"learning_rate": 9.998127900726825e-06,
"loss": 2.2512,
"step": 11
},
{
"epoch": 0.10309278350515463,
"grad_norm": 0.5343932519160086,
"learning_rate": 9.997304251071802e-06,
"loss": 2.2265,
"step": 12
},
{
"epoch": 0.11168384879725086,
"grad_norm": 0.4515050667510577,
"learning_rate": 9.99633090525405e-06,
"loss": 2.2205,
"step": 13
},
{
"epoch": 0.12027491408934708,
"grad_norm": 0.3887339848594928,
"learning_rate": 9.995207892430525e-06,
"loss": 2.199,
"step": 14
},
{
"epoch": 0.12886597938144329,
"grad_norm": 0.40393633836826276,
"learning_rate": 9.9939352462415e-06,
"loss": 2.1977,
"step": 15
},
{
"epoch": 0.13745704467353953,
"grad_norm": 0.3676064206819555,
"learning_rate": 9.99251300480958e-06,
"loss": 2.1732,
"step": 16
},
{
"epoch": 0.14604810996563575,
"grad_norm": 0.3396390866366257,
"learning_rate": 9.990941210738553e-06,
"loss": 2.1852,
"step": 17
},
{
"epoch": 0.15463917525773196,
"grad_norm": 0.3343400250980012,
"learning_rate": 9.989219911112114e-06,
"loss": 2.1995,
"step": 18
},
{
"epoch": 0.16323024054982818,
"grad_norm": 0.3395649625426219,
"learning_rate": 9.98734915749245e-06,
"loss": 2.1974,
"step": 19
},
{
"epoch": 0.1718213058419244,
"grad_norm": 0.3168174438284586,
"learning_rate": 9.985329005918702e-06,
"loss": 2.164,
"step": 20
},
{
"epoch": 0.18041237113402062,
"grad_norm": 0.30899087314283036,
"learning_rate": 9.983159516905287e-06,
"loss": 2.1506,
"step": 21
},
{
"epoch": 0.18900343642611683,
"grad_norm": 0.3104914475616672,
"learning_rate": 9.980840755440075e-06,
"loss": 2.1415,
"step": 22
},
{
"epoch": 0.19759450171821305,
"grad_norm": 0.2685339189270518,
"learning_rate": 9.978372790982457e-06,
"loss": 2.1254,
"step": 23
},
{
"epoch": 0.20618556701030927,
"grad_norm": 0.2666593191228905,
"learning_rate": 9.975755697461254e-06,
"loss": 2.1156,
"step": 24
},
{
"epoch": 0.21477663230240548,
"grad_norm": 0.2633683317715755,
"learning_rate": 9.972989553272501e-06,
"loss": 2.1083,
"step": 25
},
{
"epoch": 0.22336769759450173,
"grad_norm": 0.29601191602410226,
"learning_rate": 9.970074441277111e-06,
"loss": 2.1434,
"step": 26
},
{
"epoch": 0.23195876288659795,
"grad_norm": 0.2663773393046616,
"learning_rate": 9.967010448798376e-06,
"loss": 2.1137,
"step": 27
},
{
"epoch": 0.24054982817869416,
"grad_norm": 0.2489436765403495,
"learning_rate": 9.963797667619368e-06,
"loss": 2.105,
"step": 28
},
{
"epoch": 0.24914089347079038,
"grad_norm": 0.2513769484487265,
"learning_rate": 9.960436193980175e-06,
"loss": 2.132,
"step": 29
},
{
"epoch": 0.25773195876288657,
"grad_norm": 0.25792019512975445,
"learning_rate": 9.956926128575026e-06,
"loss": 2.0766,
"step": 30
},
{
"epoch": 0.2663230240549828,
"grad_norm": 0.24059537928347402,
"learning_rate": 9.953267576549279e-06,
"loss": 2.1145,
"step": 31
},
{
"epoch": 0.27491408934707906,
"grad_norm": 0.2413598219033581,
"learning_rate": 9.949460647496258e-06,
"loss": 2.0454,
"step": 32
},
{
"epoch": 0.28350515463917525,
"grad_norm": 0.24689976084131024,
"learning_rate": 9.945505455453983e-06,
"loss": 2.0456,
"step": 33
},
{
"epoch": 0.2920962199312715,
"grad_norm": 0.2625455755837849,
"learning_rate": 9.941402118901743e-06,
"loss": 2.0816,
"step": 34
},
{
"epoch": 0.3006872852233677,
"grad_norm": 0.25617587316052237,
"learning_rate": 9.93715076075656e-06,
"loss": 2.0879,
"step": 35
},
{
"epoch": 0.30927835051546393,
"grad_norm": 0.23182474186259913,
"learning_rate": 9.932751508369492e-06,
"loss": 2.0883,
"step": 36
},
{
"epoch": 0.3178694158075601,
"grad_norm": 0.25298056154312737,
"learning_rate": 9.92820449352183e-06,
"loss": 2.0679,
"step": 37
},
{
"epoch": 0.32646048109965636,
"grad_norm": 0.22793549377250014,
"learning_rate": 9.923509852421144e-06,
"loss": 2.0477,
"step": 38
},
{
"epoch": 0.33505154639175255,
"grad_norm": 0.22304162025326615,
"learning_rate": 9.91866772569721e-06,
"loss": 2.0681,
"step": 39
},
{
"epoch": 0.3436426116838488,
"grad_norm": 0.22391762105061253,
"learning_rate": 9.913678258397785e-06,
"loss": 2.0706,
"step": 40
},
{
"epoch": 0.35223367697594504,
"grad_norm": 0.2465667048609899,
"learning_rate": 9.908541599984276e-06,
"loss": 2.0532,
"step": 41
},
{
"epoch": 0.36082474226804123,
"grad_norm": 0.2225999225572341,
"learning_rate": 9.90325790432725e-06,
"loss": 2.0557,
"step": 42
},
{
"epoch": 0.3694158075601375,
"grad_norm": 0.23145112133504558,
"learning_rate": 9.897827329701834e-06,
"loss": 2.0077,
"step": 43
},
{
"epoch": 0.37800687285223367,
"grad_norm": 0.21978161386987122,
"learning_rate": 9.892250038782972e-06,
"loss": 2.0377,
"step": 44
},
{
"epoch": 0.3865979381443299,
"grad_norm": 0.23457257003146273,
"learning_rate": 9.88652619864055e-06,
"loss": 2.0046,
"step": 45
},
{
"epoch": 0.3951890034364261,
"grad_norm": 0.23366667859372328,
"learning_rate": 9.880655980734391e-06,
"loss": 2.0636,
"step": 46
},
{
"epoch": 0.40378006872852235,
"grad_norm": 0.216406709527668,
"learning_rate": 9.874639560909118e-06,
"loss": 1.9787,
"step": 47
},
{
"epoch": 0.41237113402061853,
"grad_norm": 0.2175332468582654,
"learning_rate": 9.868477119388897e-06,
"loss": 2.0583,
"step": 48
},
{
"epoch": 0.4209621993127148,
"grad_norm": 0.22438652510006393,
"learning_rate": 9.862168840772018e-06,
"loss": 2.0466,
"step": 49
},
{
"epoch": 0.42955326460481097,
"grad_norm": 0.2185523177171903,
"learning_rate": 9.855714914025386e-06,
"loss": 2.0105,
"step": 50
},
{
"epoch": 0.4381443298969072,
"grad_norm": 0.23254935570207558,
"learning_rate": 9.849115532478848e-06,
"loss": 1.9656,
"step": 51
},
{
"epoch": 0.44673539518900346,
"grad_norm": 0.20771583714525066,
"learning_rate": 9.842370893819404e-06,
"loss": 2.0414,
"step": 52
},
{
"epoch": 0.45532646048109965,
"grad_norm": 0.21427380812158775,
"learning_rate": 9.83548120008529e-06,
"loss": 2.0245,
"step": 53
},
{
"epoch": 0.4639175257731959,
"grad_norm": 0.21963377878640172,
"learning_rate": 9.828446657659919e-06,
"loss": 2.0406,
"step": 54
},
{
"epoch": 0.4725085910652921,
"grad_norm": 0.22510704558684724,
"learning_rate": 9.821267477265705e-06,
"loss": 2.001,
"step": 55
},
{
"epoch": 0.48109965635738833,
"grad_norm": 0.22402437197362557,
"learning_rate": 9.813943873957748e-06,
"loss": 1.9952,
"step": 56
},
{
"epoch": 0.4896907216494845,
"grad_norm": 0.21649219084755675,
"learning_rate": 9.806476067117384e-06,
"loss": 1.9775,
"step": 57
},
{
"epoch": 0.49828178694158076,
"grad_norm": 0.23004739414810843,
"learning_rate": 9.798864280445633e-06,
"loss": 2.0617,
"step": 58
},
{
"epoch": 0.506872852233677,
"grad_norm": 0.2234931893353312,
"learning_rate": 9.791108741956476e-06,
"loss": 2.0076,
"step": 59
},
{
"epoch": 0.5154639175257731,
"grad_norm": 0.31632384441079653,
"learning_rate": 9.78320968397004e-06,
"loss": 1.998,
"step": 60
},
{
"epoch": 0.5240549828178694,
"grad_norm": 0.22179973616305915,
"learning_rate": 9.77516734310563e-06,
"loss": 2.0295,
"step": 61
},
{
"epoch": 0.5326460481099656,
"grad_norm": 0.2193332083726788,
"learning_rate": 9.766981960274653e-06,
"loss": 1.973,
"step": 62
},
{
"epoch": 0.5412371134020618,
"grad_norm": 0.22853301555051309,
"learning_rate": 9.758653780673381e-06,
"loss": 2.0529,
"step": 63
},
{
"epoch": 0.5498281786941581,
"grad_norm": 0.2275807035753558,
"learning_rate": 9.750183053775625e-06,
"loss": 1.9989,
"step": 64
},
{
"epoch": 0.5584192439862543,
"grad_norm": 0.23654821220831357,
"learning_rate": 9.741570033325254e-06,
"loss": 1.9916,
"step": 65
},
{
"epoch": 0.5670103092783505,
"grad_norm": 0.2263519719171945,
"learning_rate": 9.732814977328593e-06,
"loss": 2.0294,
"step": 66
},
{
"epoch": 0.5756013745704467,
"grad_norm": 0.21292525210375207,
"learning_rate": 9.723918148046696e-06,
"loss": 2.0083,
"step": 67
},
{
"epoch": 0.584192439862543,
"grad_norm": 0.23686873465932642,
"learning_rate": 9.714879811987496e-06,
"loss": 1.9637,
"step": 68
},
{
"epoch": 0.5927835051546392,
"grad_norm": 0.22309697415764557,
"learning_rate": 9.705700239897809e-06,
"loss": 2.0339,
"step": 69
},
{
"epoch": 0.6013745704467354,
"grad_norm": 0.2363628573649254,
"learning_rate": 9.69637970675523e-06,
"loss": 1.9941,
"step": 70
},
{
"epoch": 0.6099656357388317,
"grad_norm": 0.21970605565754098,
"learning_rate": 9.686918491759904e-06,
"loss": 1.9921,
"step": 71
},
{
"epoch": 0.6185567010309279,
"grad_norm": 0.2399552026616389,
"learning_rate": 9.677316878326144e-06,
"loss": 1.985,
"step": 72
},
{
"epoch": 0.627147766323024,
"grad_norm": 0.22714040927280302,
"learning_rate": 9.667575154073962e-06,
"loss": 2.0016,
"step": 73
},
{
"epoch": 0.6357388316151202,
"grad_norm": 0.2259659476545292,
"learning_rate": 9.657693610820437e-06,
"loss": 2.0028,
"step": 74
},
{
"epoch": 0.6443298969072165,
"grad_norm": 0.22115483479500747,
"learning_rate": 9.647672544570981e-06,
"loss": 2.0005,
"step": 75
},
{
"epoch": 0.6529209621993127,
"grad_norm": 0.21943556518764748,
"learning_rate": 9.637512255510475e-06,
"loss": 1.9557,
"step": 76
},
{
"epoch": 0.6615120274914089,
"grad_norm": 0.22614114951805067,
"learning_rate": 9.627213047994265e-06,
"loss": 1.9751,
"step": 77
},
{
"epoch": 0.6701030927835051,
"grad_norm": 0.2229779123164752,
"learning_rate": 9.616775230539057e-06,
"loss": 1.9733,
"step": 78
},
{
"epoch": 0.6786941580756014,
"grad_norm": 0.28977693220933487,
"learning_rate": 9.606199115813672e-06,
"loss": 1.9538,
"step": 79
},
{
"epoch": 0.6872852233676976,
"grad_norm": 0.2398515392127217,
"learning_rate": 9.595485020629676e-06,
"loss": 1.9758,
"step": 80
},
{
"epoch": 0.6958762886597938,
"grad_norm": 0.22970549932486983,
"learning_rate": 9.584633265931894e-06,
"loss": 1.9752,
"step": 81
},
{
"epoch": 0.7044673539518901,
"grad_norm": 0.23574425808957022,
"learning_rate": 9.573644176788795e-06,
"loss": 1.9772,
"step": 82
},
{
"epoch": 0.7130584192439863,
"grad_norm": 0.24111810130645636,
"learning_rate": 9.562518082382751e-06,
"loss": 1.9843,
"step": 83
},
{
"epoch": 0.7216494845360825,
"grad_norm": 0.2353238194122449,
"learning_rate": 9.551255316000183e-06,
"loss": 1.9879,
"step": 84
},
{
"epoch": 0.7302405498281787,
"grad_norm": 0.21267148073016073,
"learning_rate": 9.539856215021568e-06,
"loss": 2.0112,
"step": 85
},
{
"epoch": 0.738831615120275,
"grad_norm": 0.24750268562290642,
"learning_rate": 9.528321120911345e-06,
"loss": 1.9686,
"step": 86
},
{
"epoch": 0.7474226804123711,
"grad_norm": 0.22624235631563774,
"learning_rate": 9.516650379207677e-06,
"loss": 1.9531,
"step": 87
},
{
"epoch": 0.7560137457044673,
"grad_norm": 0.2254562873218311,
"learning_rate": 9.504844339512096e-06,
"loss": 1.9336,
"step": 88
},
{
"epoch": 0.7646048109965635,
"grad_norm": 0.24114646864174968,
"learning_rate": 9.492903355479047e-06,
"loss": 1.9649,
"step": 89
},
{
"epoch": 0.7731958762886598,
"grad_norm": 0.2215844220698404,
"learning_rate": 9.480827784805278e-06,
"loss": 1.9919,
"step": 90
},
{
"epoch": 0.781786941580756,
"grad_norm": 0.22673007789344823,
"learning_rate": 9.468617989219136e-06,
"loss": 1.9671,
"step": 91
},
{
"epoch": 0.7903780068728522,
"grad_norm": 0.23381648472171454,
"learning_rate": 9.45627433446972e-06,
"loss": 1.9857,
"step": 92
},
{
"epoch": 0.7989690721649485,
"grad_norm": 0.22439137602734674,
"learning_rate": 9.443797190315938e-06,
"loss": 1.9887,
"step": 93
},
{
"epoch": 0.8075601374570447,
"grad_norm": 0.22502583802282547,
"learning_rate": 9.431186930515419e-06,
"loss": 1.9471,
"step": 94
},
{
"epoch": 0.8161512027491409,
"grad_norm": 0.22566119014941471,
"learning_rate": 9.418443932813328e-06,
"loss": 2.0001,
"step": 95
},
{
"epoch": 0.8247422680412371,
"grad_norm": 0.21651729773145714,
"learning_rate": 9.405568578931042e-06,
"loss": 1.9708,
"step": 96
},
{
"epoch": 0.8333333333333334,
"grad_norm": 0.22510924701029755,
"learning_rate": 9.392561254554712e-06,
"loss": 2.0141,
"step": 97
},
{
"epoch": 0.8419243986254296,
"grad_norm": 0.2239826811443597,
"learning_rate": 9.379422349323728e-06,
"loss": 1.9911,
"step": 98
},
{
"epoch": 0.8505154639175257,
"grad_norm": 0.23622343708573781,
"learning_rate": 9.366152256819025e-06,
"loss": 1.9265,
"step": 99
},
{
"epoch": 0.8591065292096219,
"grad_norm": 0.2330309560243661,
"learning_rate": 9.352751374551305e-06,
"loss": 1.9471,
"step": 100
},
{
"epoch": 0.8676975945017182,
"grad_norm": 0.24392986076413478,
"learning_rate": 9.339220103949132e-06,
"loss": 1.8885,
"step": 101
},
{
"epoch": 0.8762886597938144,
"grad_norm": 0.2283050690340691,
"learning_rate": 9.325558850346897e-06,
"loss": 1.9737,
"step": 102
},
{
"epoch": 0.8848797250859106,
"grad_norm": 0.23216642874875693,
"learning_rate": 9.311768022972682e-06,
"loss": 1.9265,
"step": 103
},
{
"epoch": 0.8934707903780069,
"grad_norm": 0.23718946052920817,
"learning_rate": 9.297848034936007e-06,
"loss": 1.9113,
"step": 104
},
{
"epoch": 0.9020618556701031,
"grad_norm": 0.26062401247343786,
"learning_rate": 9.283799303215442e-06,
"loss": 1.9109,
"step": 105
},
{
"epoch": 0.9106529209621993,
"grad_norm": 0.22346908351749736,
"learning_rate": 9.269622248646124e-06,
"loss": 1.9607,
"step": 106
},
{
"epoch": 0.9192439862542955,
"grad_norm": 0.24711103570338047,
"learning_rate": 9.255317295907158e-06,
"loss": 1.964,
"step": 107
},
{
"epoch": 0.9278350515463918,
"grad_norm": 0.25323775068277576,
"learning_rate": 9.240884873508876e-06,
"loss": 1.9167,
"step": 108
},
{
"epoch": 0.936426116838488,
"grad_norm": 0.25123992738826834,
"learning_rate": 9.226325413780021e-06,
"loss": 1.9516,
"step": 109
},
{
"epoch": 0.9450171821305842,
"grad_norm": 0.2548307883088889,
"learning_rate": 9.211639352854786e-06,
"loss": 1.972,
"step": 110
},
{
"epoch": 0.9536082474226805,
"grad_norm": 0.24129670124461444,
"learning_rate": 9.196827130659752e-06,
"loss": 1.9694,
"step": 111
},
{
"epoch": 0.9621993127147767,
"grad_norm": 0.25262012893284974,
"learning_rate": 9.181889190900702e-06,
"loss": 1.9389,
"step": 112
},
{
"epoch": 0.9707903780068728,
"grad_norm": 0.24800856374400806,
"learning_rate": 9.166825981049345e-06,
"loss": 1.9609,
"step": 113
},
{
"epoch": 0.979381443298969,
"grad_norm": 0.23930062604165217,
"learning_rate": 9.151637952329903e-06,
"loss": 1.9169,
"step": 114
},
{
"epoch": 0.9879725085910653,
"grad_norm": 0.23888942710116004,
"learning_rate": 9.136325559705593e-06,
"loss": 1.9409,
"step": 115
},
{
"epoch": 0.9965635738831615,
"grad_norm": 0.2434327307011672,
"learning_rate": 9.120889261864999e-06,
"loss": 1.9549,
"step": 116
},
{
"epoch": 1.0085910652920962,
"grad_norm": 0.5368696897762311,
"learning_rate": 9.105329521208334e-06,
"loss": 3.9516,
"step": 117
},
{
"epoch": 1.0171821305841924,
"grad_norm": 0.31841091070667155,
"learning_rate": 9.089646803833589e-06,
"loss": 1.8892,
"step": 118
},
{
"epoch": 1.0257731958762886,
"grad_norm": 0.2742869080672636,
"learning_rate": 9.073841579522571e-06,
"loss": 1.8438,
"step": 119
},
{
"epoch": 1.034364261168385,
"grad_norm": 0.3098173382025614,
"learning_rate": 9.057914321726824e-06,
"loss": 1.8524,
"step": 120
},
{
"epoch": 1.0429553264604812,
"grad_norm": 0.35914000861625267,
"learning_rate": 9.041865507553458e-06,
"loss": 1.8578,
"step": 121
},
{
"epoch": 1.0515463917525774,
"grad_norm": 0.24782705778836578,
"learning_rate": 9.025695617750848e-06,
"loss": 1.8968,
"step": 122
},
{
"epoch": 1.0601374570446735,
"grad_norm": 0.29479978245416644,
"learning_rate": 9.009405136694234e-06,
"loss": 1.8527,
"step": 123
},
{
"epoch": 1.0687285223367697,
"grad_norm": 0.267590029709473,
"learning_rate": 8.992994552371217e-06,
"loss": 1.8557,
"step": 124
},
{
"epoch": 1.077319587628866,
"grad_norm": 0.2770218237542545,
"learning_rate": 8.976464356367133e-06,
"loss": 1.88,
"step": 125
},
{
"epoch": 1.0859106529209621,
"grad_norm": 0.24930357718028165,
"learning_rate": 8.959815043850336e-06,
"loss": 1.8914,
"step": 126
},
{
"epoch": 1.0945017182130585,
"grad_norm": 0.2577049899528488,
"learning_rate": 8.943047113557358e-06,
"loss": 1.8551,
"step": 127
},
{
"epoch": 1.1030927835051547,
"grad_norm": 0.2869232186158467,
"learning_rate": 8.926161067777973e-06,
"loss": 1.8314,
"step": 128
},
{
"epoch": 1.111683848797251,
"grad_norm": 0.257531034868115,
"learning_rate": 8.90915741234015e-06,
"loss": 1.8931,
"step": 129
},
{
"epoch": 1.120274914089347,
"grad_norm": 0.2510603923186068,
"learning_rate": 8.892036656594898e-06,
"loss": 1.8864,
"step": 130
},
{
"epoch": 1.1288659793814433,
"grad_norm": 0.22785031834695588,
"learning_rate": 8.874799313401014e-06,
"loss": 1.8966,
"step": 131
},
{
"epoch": 1.1374570446735395,
"grad_norm": 0.2625000755681014,
"learning_rate": 8.857445899109716e-06,
"loss": 1.8894,
"step": 132
},
{
"epoch": 1.1460481099656357,
"grad_norm": 0.25962578916321755,
"learning_rate": 8.839976933549173e-06,
"loss": 1.8373,
"step": 133
},
{
"epoch": 1.1546391752577319,
"grad_norm": 0.2535508034573568,
"learning_rate": 8.822392940008937e-06,
"loss": 1.8603,
"step": 134
},
{
"epoch": 1.1632302405498283,
"grad_norm": 0.2483718305265392,
"learning_rate": 8.804694445224274e-06,
"loss": 1.7806,
"step": 135
},
{
"epoch": 1.1718213058419245,
"grad_norm": 0.23143602082187498,
"learning_rate": 8.786881979360368e-06,
"loss": 1.8827,
"step": 136
},
{
"epoch": 1.1804123711340206,
"grad_norm": 0.26722488115137877,
"learning_rate": 8.76895607599646e-06,
"loss": 1.8687,
"step": 137
},
{
"epoch": 1.1890034364261168,
"grad_norm": 0.23165280409335437,
"learning_rate": 8.750917272109849e-06,
"loss": 1.8204,
"step": 138
},
{
"epoch": 1.197594501718213,
"grad_norm": 0.24114870798141225,
"learning_rate": 8.732766108059814e-06,
"loss": 1.8257,
"step": 139
},
{
"epoch": 1.2061855670103092,
"grad_norm": 0.24450933667975003,
"learning_rate": 8.714503127571425e-06,
"loss": 1.8655,
"step": 140
},
{
"epoch": 1.2147766323024054,
"grad_norm": 0.24886954586593113,
"learning_rate": 8.696128877719258e-06,
"loss": 1.8635,
"step": 141
},
{
"epoch": 1.2233676975945018,
"grad_norm": 0.22947391665498382,
"learning_rate": 8.677643908911007e-06,
"loss": 1.8683,
"step": 142
},
{
"epoch": 1.231958762886598,
"grad_norm": 0.24875679422747346,
"learning_rate": 8.659048774870986e-06,
"loss": 1.8897,
"step": 143
},
{
"epoch": 1.2405498281786942,
"grad_norm": 0.21948813585313118,
"learning_rate": 8.64034403262356e-06,
"loss": 1.8661,
"step": 144
},
{
"epoch": 1.2491408934707904,
"grad_norm": 0.25682738261464727,
"learning_rate": 8.621530242476446e-06,
"loss": 1.8135,
"step": 145
},
{
"epoch": 1.2577319587628866,
"grad_norm": 0.46219652770176195,
"learning_rate": 8.602607968003935e-06,
"loss": 1.8172,
"step": 146
},
{
"epoch": 1.2663230240549828,
"grad_norm": 0.2274158864545253,
"learning_rate": 8.583577776030005e-06,
"loss": 1.8351,
"step": 147
},
{
"epoch": 1.274914089347079,
"grad_norm": 0.25012506751574154,
"learning_rate": 8.564440236611344e-06,
"loss": 1.8969,
"step": 148
},
{
"epoch": 1.2835051546391751,
"grad_norm": 0.2769019876124278,
"learning_rate": 8.545195923020273e-06,
"loss": 1.8762,
"step": 149
},
{
"epoch": 1.2920962199312716,
"grad_norm": 0.26472276342375745,
"learning_rate": 8.525845411727581e-06,
"loss": 1.8824,
"step": 150
},
{
"epoch": 1.3006872852233677,
"grad_norm": 0.24426788274377817,
"learning_rate": 8.506389282385242e-06,
"loss": 1.8752,
"step": 151
},
{
"epoch": 1.309278350515464,
"grad_norm": 0.2850090172193898,
"learning_rate": 8.486828117809057e-06,
"loss": 1.8499,
"step": 152
},
{
"epoch": 1.3178694158075601,
"grad_norm": 0.28133366851138764,
"learning_rate": 8.467162503961209e-06,
"loss": 1.8705,
"step": 153
},
{
"epoch": 1.3264604810996563,
"grad_norm": 0.2455210250126331,
"learning_rate": 8.447393029932692e-06,
"loss": 1.8423,
"step": 154
},
{
"epoch": 1.3350515463917525,
"grad_norm": 0.3008489303713607,
"learning_rate": 8.427520287925669e-06,
"loss": 1.8963,
"step": 155
},
{
"epoch": 1.343642611683849,
"grad_norm": 0.25506919040783055,
"learning_rate": 8.407544873235736e-06,
"loss": 1.8521,
"step": 156
},
{
"epoch": 1.352233676975945,
"grad_norm": 0.27043313385906653,
"learning_rate": 8.387467384234096e-06,
"loss": 1.8709,
"step": 157
},
{
"epoch": 1.3608247422680413,
"grad_norm": 0.27087203176436514,
"learning_rate": 8.367288422349617e-06,
"loss": 1.8636,
"step": 158
},
{
"epoch": 1.3694158075601375,
"grad_norm": 0.25048232585043473,
"learning_rate": 8.347008592050834e-06,
"loss": 1.8694,
"step": 159
},
{
"epoch": 1.3780068728522337,
"grad_norm": 0.27609000035122433,
"learning_rate": 8.326628500827826e-06,
"loss": 1.8344,
"step": 160
},
{
"epoch": 1.3865979381443299,
"grad_norm": 0.27449314945979997,
"learning_rate": 8.306148759174036e-06,
"loss": 1.8706,
"step": 161
},
{
"epoch": 1.395189003436426,
"grad_norm": 0.233840133061648,
"learning_rate": 8.285569980567965e-06,
"loss": 1.8695,
"step": 162
},
{
"epoch": 1.4037800687285222,
"grad_norm": 0.273284165103743,
"learning_rate": 8.264892781454807e-06,
"loss": 1.8518,
"step": 163
},
{
"epoch": 1.4123711340206184,
"grad_norm": 0.26246253700879174,
"learning_rate": 8.244117781227982e-06,
"loss": 1.8214,
"step": 164
},
{
"epoch": 1.4209621993127148,
"grad_norm": 0.23343115890866387,
"learning_rate": 8.22324560221058e-06,
"loss": 1.8531,
"step": 165
},
{
"epoch": 1.429553264604811,
"grad_norm": 0.23604921177468424,
"learning_rate": 8.202276869636713e-06,
"loss": 1.8033,
"step": 166
},
{
"epoch": 1.4381443298969072,
"grad_norm": 0.25286857214079783,
"learning_rate": 8.1812122116328e-06,
"loss": 1.8764,
"step": 167
},
{
"epoch": 1.4467353951890034,
"grad_norm": 0.25559292386906807,
"learning_rate": 8.160052259198737e-06,
"loss": 1.8211,
"step": 168
},
{
"epoch": 1.4553264604810996,
"grad_norm": 0.26293486350593825,
"learning_rate": 8.138797646189e-06,
"loss": 1.8409,
"step": 169
},
{
"epoch": 1.463917525773196,
"grad_norm": 0.2615615587710106,
"learning_rate": 8.117449009293668e-06,
"loss": 1.82,
"step": 170
},
{
"epoch": 1.4725085910652922,
"grad_norm": 0.24029229889760534,
"learning_rate": 8.096006988019331e-06,
"loss": 1.795,
"step": 171
},
{
"epoch": 1.4810996563573884,
"grad_norm": 0.24396907732238718,
"learning_rate": 8.074472224669952e-06,
"loss": 1.8288,
"step": 172
},
{
"epoch": 1.4896907216494846,
"grad_norm": 0.2588430791949496,
"learning_rate": 8.052845364327609e-06,
"loss": 1.8313,
"step": 173
},
{
"epoch": 1.4982817869415808,
"grad_norm": 0.2402748200690545,
"learning_rate": 8.031127054833192e-06,
"loss": 1.8475,
"step": 174
},
{
"epoch": 1.506872852233677,
"grad_norm": 0.26607111639386694,
"learning_rate": 8.009317946766975e-06,
"loss": 1.8318,
"step": 175
},
{
"epoch": 1.5154639175257731,
"grad_norm": 0.2490367029154612,
"learning_rate": 7.987418693429145e-06,
"loss": 1.8713,
"step": 176
},
{
"epoch": 1.5240549828178693,
"grad_norm": 0.24506198481130903,
"learning_rate": 7.965429950820222e-06,
"loss": 1.846,
"step": 177
},
{
"epoch": 1.5326460481099655,
"grad_norm": 0.2510589297764303,
"learning_rate": 7.943352377621414e-06,
"loss": 1.8191,
"step": 178
},
{
"epoch": 1.5412371134020617,
"grad_norm": 0.2463498132205643,
"learning_rate": 7.92118663517488e-06,
"loss": 1.8264,
"step": 179
},
{
"epoch": 1.5498281786941581,
"grad_norm": 0.26856443469145774,
"learning_rate": 7.898933387463924e-06,
"loss": 1.8214,
"step": 180
},
{
"epoch": 1.5584192439862543,
"grad_norm": 0.24408339713981886,
"learning_rate": 7.876593301093104e-06,
"loss": 1.8655,
"step": 181
},
{
"epoch": 1.5670103092783505,
"grad_norm": 0.2824672391033115,
"learning_rate": 7.854167045268265e-06,
"loss": 1.8303,
"step": 182
},
{
"epoch": 1.5756013745704467,
"grad_norm": 0.2652503232543874,
"learning_rate": 7.831655291776484e-06,
"loss": 1.8078,
"step": 183
},
{
"epoch": 1.584192439862543,
"grad_norm": 0.3562892836637005,
"learning_rate": 7.809058714965962e-06,
"loss": 1.8365,
"step": 184
},
{
"epoch": 1.5927835051546393,
"grad_norm": 0.31682382377621443,
"learning_rate": 7.786377991725813e-06,
"loss": 1.8342,
"step": 185
},
{
"epoch": 1.6013745704467355,
"grad_norm": 0.23708795791695556,
"learning_rate": 7.763613801465785e-06,
"loss": 1.8338,
"step": 186
},
{
"epoch": 1.6099656357388317,
"grad_norm": 0.32980665096322415,
"learning_rate": 7.740766826095918e-06,
"loss": 1.8525,
"step": 187
},
{
"epoch": 1.6185567010309279,
"grad_norm": 0.3308585733412994,
"learning_rate": 7.717837750006106e-06,
"loss": 1.8699,
"step": 188
},
{
"epoch": 1.627147766323024,
"grad_norm": 0.23213184583376612,
"learning_rate": 7.694827260045608e-06,
"loss": 1.8346,
"step": 189
},
{
"epoch": 1.6357388316151202,
"grad_norm": 0.36115699508964094,
"learning_rate": 7.671736045502462e-06,
"loss": 1.8296,
"step": 190
},
{
"epoch": 1.6443298969072164,
"grad_norm": 0.3725788178513254,
"learning_rate": 7.648564798082842e-06,
"loss": 1.8621,
"step": 191
},
{
"epoch": 1.6529209621993126,
"grad_norm": 0.244920413507623,
"learning_rate": 7.625314211890342e-06,
"loss": 1.8379,
"step": 192
},
{
"epoch": 1.6615120274914088,
"grad_norm": 0.3003785923390606,
"learning_rate": 7.601984983405173e-06,
"loss": 1.8285,
"step": 193
},
{
"epoch": 1.670103092783505,
"grad_norm": 0.24554481882238194,
"learning_rate": 7.57857781146331e-06,
"loss": 1.8282,
"step": 194
},
{
"epoch": 1.6786941580756014,
"grad_norm": 0.3086444191459068,
"learning_rate": 7.555093397235553e-06,
"loss": 1.8351,
"step": 195
},
{
"epoch": 1.6872852233676976,
"grad_norm": 0.2255666101776277,
"learning_rate": 7.531532444206524e-06,
"loss": 1.8619,
"step": 196
},
{
"epoch": 1.6958762886597938,
"grad_norm": 0.35915951404678703,
"learning_rate": 7.507895658153594e-06,
"loss": 1.8153,
"step": 197
},
{
"epoch": 1.7044673539518902,
"grad_norm": 0.2998884420051895,
"learning_rate": 7.484183747125743e-06,
"loss": 1.8065,
"step": 198
},
{
"epoch": 1.7130584192439864,
"grad_norm": 0.27294046373984265,
"learning_rate": 7.460397421422346e-06,
"loss": 1.8642,
"step": 199
},
{
"epoch": 1.7216494845360826,
"grad_norm": 0.3266219558792723,
"learning_rate": 7.4365373935719e-06,
"loss": 1.8297,
"step": 200
},
{
"epoch": 1.7302405498281788,
"grad_norm": 0.2650164068518729,
"learning_rate": 7.412604378310677e-06,
"loss": 1.8304,
"step": 201
},
{
"epoch": 1.738831615120275,
"grad_norm": 0.27308861174574434,
"learning_rate": 7.388599092561315e-06,
"loss": 1.822,
"step": 202
},
{
"epoch": 1.7474226804123711,
"grad_norm": 0.2463141857916976,
"learning_rate": 7.364522255411342e-06,
"loss": 1.8776,
"step": 203
},
{
"epoch": 1.7560137457044673,
"grad_norm": 0.2599426045804314,
"learning_rate": 7.340374588091638e-06,
"loss": 1.8509,
"step": 204
},
{
"epoch": 1.7646048109965635,
"grad_norm": 0.24005078867376234,
"learning_rate": 7.316156813954821e-06,
"loss": 1.8183,
"step": 205
},
{
"epoch": 1.7731958762886597,
"grad_norm": 0.27120591080632045,
"learning_rate": 7.291869658453594e-06,
"loss": 1.8149,
"step": 206
},
{
"epoch": 1.781786941580756,
"grad_norm": 0.2531942384017391,
"learning_rate": 7.267513849119001e-06,
"loss": 1.8261,
"step": 207
},
{
"epoch": 1.790378006872852,
"grad_norm": 0.2747157434197358,
"learning_rate": 7.243090115538639e-06,
"loss": 1.7975,
"step": 208
},
{
"epoch": 1.7989690721649485,
"grad_norm": 0.24091105341462182,
"learning_rate": 7.218599189334799e-06,
"loss": 1.8353,
"step": 209
},
{
"epoch": 1.8075601374570447,
"grad_norm": 0.24341386656771288,
"learning_rate": 7.194041804142556e-06,
"loss": 1.8859,
"step": 210
},
{
"epoch": 1.8161512027491409,
"grad_norm": 0.22471355270187646,
"learning_rate": 7.169418695587791e-06,
"loss": 1.8431,
"step": 211
},
{
"epoch": 1.824742268041237,
"grad_norm": 0.24506188684019525,
"learning_rate": 7.144730601265148e-06,
"loss": 1.8125,
"step": 212
},
{
"epoch": 1.8333333333333335,
"grad_norm": 0.22810680645845968,
"learning_rate": 7.1199782607159494e-06,
"loss": 1.8194,
"step": 213
},
{
"epoch": 1.8419243986254297,
"grad_norm": 0.2449433442646377,
"learning_rate": 7.095162415406034e-06,
"loss": 1.827,
"step": 214
},
{
"epoch": 1.8505154639175259,
"grad_norm": 0.24559185753817364,
"learning_rate": 7.070283808703553e-06,
"loss": 1.8464,
"step": 215
},
{
"epoch": 1.859106529209622,
"grad_norm": 0.24427706936825086,
"learning_rate": 7.045343185856701e-06,
"loss": 1.843,
"step": 216
},
{
"epoch": 1.8676975945017182,
"grad_norm": 0.24645403275584066,
"learning_rate": 7.020341293971383e-06,
"loss": 1.8306,
"step": 217
},
{
"epoch": 1.8762886597938144,
"grad_norm": 0.23184930559537723,
"learning_rate": 6.995278881988847e-06,
"loss": 1.8015,
"step": 218
},
{
"epoch": 1.8848797250859106,
"grad_norm": 0.23719377224785188,
"learning_rate": 6.970156700663244e-06,
"loss": 1.8525,
"step": 219
},
{
"epoch": 1.8934707903780068,
"grad_norm": 0.23400764447147657,
"learning_rate": 6.9449755025391355e-06,
"loss": 1.8119,
"step": 220
},
{
"epoch": 1.902061855670103,
"grad_norm": 0.2599405268669303,
"learning_rate": 6.919736041928956e-06,
"loss": 1.8323,
"step": 221
},
{
"epoch": 1.9106529209621992,
"grad_norm": 0.24632902890049496,
"learning_rate": 6.894439074890413e-06,
"loss": 1.8485,
"step": 222
},
{
"epoch": 1.9192439862542954,
"grad_norm": 0.23367391136495488,
"learning_rate": 6.869085359203844e-06,
"loss": 1.7983,
"step": 223
},
{
"epoch": 1.9278350515463918,
"grad_norm": 0.2510789521553247,
"learning_rate": 6.843675654349513e-06,
"loss": 1.8341,
"step": 224
},
{
"epoch": 1.936426116838488,
"grad_norm": 0.23860360419431745,
"learning_rate": 6.818210721484859e-06,
"loss": 1.8398,
"step": 225
},
{
"epoch": 1.9450171821305842,
"grad_norm": 0.25014975307900233,
"learning_rate": 6.792691323421698e-06,
"loss": 1.8263,
"step": 226
},
{
"epoch": 1.9536082474226806,
"grad_norm": 0.26282781990546983,
"learning_rate": 6.767118224603374e-06,
"loss": 1.8677,
"step": 227
},
{
"epoch": 1.9621993127147768,
"grad_norm": 0.256705179343515,
"learning_rate": 6.741492191081856e-06,
"loss": 1.8442,
"step": 228
},
{
"epoch": 1.970790378006873,
"grad_norm": 0.24544564998960142,
"learning_rate": 6.715813990494793e-06,
"loss": 1.7951,
"step": 229
},
{
"epoch": 1.9793814432989691,
"grad_norm": 0.3053559290033019,
"learning_rate": 6.690084392042514e-06,
"loss": 1.8208,
"step": 230
},
{
"epoch": 1.9879725085910653,
"grad_norm": 0.30531170857183865,
"learning_rate": 6.664304166465e-06,
"loss": 1.8202,
"step": 231
},
{
"epoch": 1.9965635738831615,
"grad_norm": 0.23274931772047064,
"learning_rate": 6.638474086018778e-06,
"loss": 1.8046,
"step": 232
},
{
"epoch": 2.008591065292096,
"grad_norm": 0.5794577169743913,
"learning_rate": 6.612594924453801e-06,
"loss": 3.5412,
"step": 233
},
{
"epoch": 2.0171821305841924,
"grad_norm": 0.343082374090462,
"learning_rate": 6.5866674569902676e-06,
"loss": 1.735,
"step": 234
},
{
"epoch": 2.0257731958762886,
"grad_norm": 0.28141601659964904,
"learning_rate": 6.5606924602953925e-06,
"loss": 1.7572,
"step": 235
},
{
"epoch": 2.0343642611683848,
"grad_norm": 0.34976737295239363,
"learning_rate": 6.534670712460151e-06,
"loss": 1.7368,
"step": 236
},
{
"epoch": 2.042955326460481,
"grad_norm": 0.40040502919537163,
"learning_rate": 6.508602992975963e-06,
"loss": 1.767,
"step": 237
},
{
"epoch": 2.051546391752577,
"grad_norm": 0.2911696775881224,
"learning_rate": 6.4824900827113506e-06,
"loss": 1.7602,
"step": 238
},
{
"epoch": 2.0601374570446733,
"grad_norm": 0.29572445302801437,
"learning_rate": 6.456332763888544e-06,
"loss": 1.7454,
"step": 239
},
{
"epoch": 2.06872852233677,
"grad_norm": 0.26729327030395994,
"learning_rate": 6.430131820060043e-06,
"loss": 1.7436,
"step": 240
},
{
"epoch": 2.077319587628866,
"grad_norm": 0.2992754951710706,
"learning_rate": 6.403888036085155e-06,
"loss": 1.7117,
"step": 241
},
{
"epoch": 2.0859106529209623,
"grad_norm": 0.33077824275444107,
"learning_rate": 6.3776021981064825e-06,
"loss": 1.7511,
"step": 242
},
{
"epoch": 2.0945017182130585,
"grad_norm": 0.3177034331189178,
"learning_rate": 6.3512750935263664e-06,
"loss": 1.7398,
"step": 243
},
{
"epoch": 2.1030927835051547,
"grad_norm": 0.2526942708362804,
"learning_rate": 6.32490751098331e-06,
"loss": 1.7115,
"step": 244
},
{
"epoch": 2.111683848797251,
"grad_norm": 0.28767940216717397,
"learning_rate": 6.298500240328342e-06,
"loss": 1.7521,
"step": 245
},
{
"epoch": 2.120274914089347,
"grad_norm": 0.2814458013423626,
"learning_rate": 6.272054072601374e-06,
"loss": 1.7511,
"step": 246
},
{
"epoch": 2.1288659793814433,
"grad_norm": 0.2605489443356379,
"learning_rate": 6.245569800007484e-06,
"loss": 1.7609,
"step": 247
},
{
"epoch": 2.1374570446735395,
"grad_norm": 0.3285855498307297,
"learning_rate": 6.219048215893204e-06,
"loss": 1.7857,
"step": 248
},
{
"epoch": 2.1460481099656357,
"grad_norm": 0.25393068431559573,
"learning_rate": 6.192490114722741e-06,
"loss": 1.7683,
"step": 249
},
{
"epoch": 2.154639175257732,
"grad_norm": 0.2833431825377502,
"learning_rate": 6.1658962920541875e-06,
"loss": 1.7794,
"step": 250
},
{
"epoch": 2.163230240549828,
"grad_norm": 0.3185646342601711,
"learning_rate": 6.139267544515689e-06,
"loss": 1.7269,
"step": 251
},
{
"epoch": 2.1718213058419242,
"grad_norm": 0.2549426100743674,
"learning_rate": 6.112604669781572e-06,
"loss": 1.7984,
"step": 252
},
{
"epoch": 2.1804123711340204,
"grad_norm": 0.2974211601234815,
"learning_rate": 6.0859084665484645e-06,
"loss": 1.6639,
"step": 253
},
{
"epoch": 2.189003436426117,
"grad_norm": 0.29735313491718546,
"learning_rate": 6.059179734511357e-06,
"loss": 1.7581,
"step": 254
},
{
"epoch": 2.1975945017182132,
"grad_norm": 0.22827203527316411,
"learning_rate": 6.032419274339654e-06,
"loss": 1.768,
"step": 255
},
{
"epoch": 2.2061855670103094,
"grad_norm": 0.2921679486136057,
"learning_rate": 6.005627887653189e-06,
"loss": 1.737,
"step": 256
},
{
"epoch": 2.2147766323024056,
"grad_norm": 0.23174406967073777,
"learning_rate": 5.978806376998209e-06,
"loss": 1.7783,
"step": 257
},
{
"epoch": 2.223367697594502,
"grad_norm": 0.26415721171980494,
"learning_rate": 5.951955545823342e-06,
"loss": 1.7638,
"step": 258
},
{
"epoch": 2.231958762886598,
"grad_norm": 0.3168814647576765,
"learning_rate": 5.925076198455517e-06,
"loss": 1.751,
"step": 259
},
{
"epoch": 2.240549828178694,
"grad_norm": 0.24804457019619985,
"learning_rate": 5.898169140075878e-06,
"loss": 1.7432,
"step": 260
},
{
"epoch": 2.2491408934707904,
"grad_norm": 0.2666809574189019,
"learning_rate": 5.871235176695664e-06,
"loss": 1.7538,
"step": 261
},
{
"epoch": 2.2577319587628866,
"grad_norm": 0.28781492733295405,
"learning_rate": 5.844275115132064e-06,
"loss": 1.7617,
"step": 262
},
{
"epoch": 2.2663230240549828,
"grad_norm": 0.25687309069891207,
"learning_rate": 5.817289762984048e-06,
"loss": 1.7679,
"step": 263
},
{
"epoch": 2.274914089347079,
"grad_norm": 0.31675631158329604,
"learning_rate": 5.790279928608173e-06,
"loss": 1.7867,
"step": 264
},
{
"epoch": 2.283505154639175,
"grad_norm": 0.2357319821112829,
"learning_rate": 5.763246421094373e-06,
"loss": 1.7247,
"step": 265
},
{
"epoch": 2.2920962199312713,
"grad_norm": 0.38591750012011294,
"learning_rate": 5.736190050241719e-06,
"loss": 1.7289,
"step": 266
},
{
"epoch": 2.3006872852233675,
"grad_norm": 0.2948215158582319,
"learning_rate": 5.709111626534161e-06,
"loss": 1.7575,
"step": 267
},
{
"epoch": 2.3092783505154637,
"grad_norm": 0.2885319497537962,
"learning_rate": 5.6820119611162515e-06,
"loss": 1.7103,
"step": 268
},
{
"epoch": 2.31786941580756,
"grad_norm": 0.376037745824099,
"learning_rate": 5.65489186576885e-06,
"loss": 1.746,
"step": 269
},
{
"epoch": 2.3264604810996565,
"grad_norm": 0.2497562431197746,
"learning_rate": 5.627752152884794e-06,
"loss": 1.7603,
"step": 270
},
{
"epoch": 2.3350515463917527,
"grad_norm": 0.2990905369609316,
"learning_rate": 5.600593635444583e-06,
"loss": 1.7515,
"step": 271
},
{
"epoch": 2.343642611683849,
"grad_norm": 0.30117959897506796,
"learning_rate": 5.573417126992004e-06,
"loss": 1.7569,
"step": 272
},
{
"epoch": 2.352233676975945,
"grad_norm": 0.2557741184765267,
"learning_rate": 5.546223441609775e-06,
"loss": 1.7472,
"step": 273
},
{
"epoch": 2.3608247422680413,
"grad_norm": 0.2551208106353442,
"learning_rate": 5.51901339389516e-06,
"loss": 1.7317,
"step": 274
},
{
"epoch": 2.3694158075601375,
"grad_norm": 0.2781777841112468,
"learning_rate": 5.491787798935557e-06,
"loss": 1.6851,
"step": 275
},
{
"epoch": 2.3780068728522337,
"grad_norm": 0.31337500147197084,
"learning_rate": 5.464547472284091e-06,
"loss": 1.7791,
"step": 276
},
{
"epoch": 2.38659793814433,
"grad_norm": 0.23452467271508043,
"learning_rate": 5.437293229935178e-06,
"loss": 1.7582,
"step": 277
},
{
"epoch": 2.395189003436426,
"grad_norm": 0.27327927198344293,
"learning_rate": 5.4100258883000874e-06,
"loss": 1.7133,
"step": 278
},
{
"epoch": 2.4037800687285222,
"grad_norm": 0.27708583184067626,
"learning_rate": 5.38274626418248e-06,
"loss": 1.7163,
"step": 279
},
{
"epoch": 2.4123711340206184,
"grad_norm": 0.23292464029123794,
"learning_rate": 5.355455174753941e-06,
"loss": 1.7298,
"step": 280
},
{
"epoch": 2.4209621993127146,
"grad_norm": 0.2600703750003263,
"learning_rate": 5.328153437529512e-06,
"loss": 1.7573,
"step": 281
},
{
"epoch": 2.429553264604811,
"grad_norm": 0.2523351663963291,
"learning_rate": 5.300841870343183e-06,
"loss": 1.7449,
"step": 282
},
{
"epoch": 2.4381443298969074,
"grad_norm": 0.2670431080125699,
"learning_rate": 5.273521291323411e-06,
"loss": 1.7246,
"step": 283
},
{
"epoch": 2.4467353951890036,
"grad_norm": 0.25254035187471163,
"learning_rate": 5.2461925188686035e-06,
"loss": 1.7451,
"step": 284
},
{
"epoch": 2.4553264604811,
"grad_norm": 0.2508883074086655,
"learning_rate": 5.218856371622605e-06,
"loss": 1.7262,
"step": 285
},
{
"epoch": 2.463917525773196,
"grad_norm": 0.25219229137696275,
"learning_rate": 5.191513668450178e-06,
"loss": 1.737,
"step": 286
},
{
"epoch": 2.472508591065292,
"grad_norm": 0.25320537808898,
"learning_rate": 5.1641652284124645e-06,
"loss": 1.7396,
"step": 287
},
{
"epoch": 2.4810996563573884,
"grad_norm": 0.2806981448693318,
"learning_rate": 5.136811870742462e-06,
"loss": 1.7437,
"step": 288
},
{
"epoch": 2.4896907216494846,
"grad_norm": 0.24173806733280326,
"learning_rate": 5.109454414820475e-06,
"loss": 1.7791,
"step": 289
},
{
"epoch": 2.4982817869415808,
"grad_norm": 0.2707788540745343,
"learning_rate": 5.0820936801495716e-06,
"loss": 1.7543,
"step": 290
},
{
"epoch": 2.506872852233677,
"grad_norm": 0.2599485029275578,
"learning_rate": 5.054730486331041e-06,
"loss": 1.7234,
"step": 291
},
{
"epoch": 2.515463917525773,
"grad_norm": 0.25054672276405254,
"learning_rate": 5.0273656530398285e-06,
"loss": 1.7352,
"step": 292
},
{
"epoch": 2.5240549828178693,
"grad_norm": 0.2573776336213065,
"learning_rate": 5e-06,
"loss": 1.7514,
"step": 293
},
{
"epoch": 2.5326460481099655,
"grad_norm": 0.23703533832993126,
"learning_rate": 4.972634346960173e-06,
"loss": 1.6359,
"step": 294
},
{
"epoch": 2.5412371134020617,
"grad_norm": 0.2705943259676728,
"learning_rate": 4.945269513668962e-06,
"loss": 1.777,
"step": 295
},
{
"epoch": 2.549828178694158,
"grad_norm": 0.2550242865212212,
"learning_rate": 4.917906319850431e-06,
"loss": 1.7132,
"step": 296
},
{
"epoch": 2.558419243986254,
"grad_norm": 0.26742448602135604,
"learning_rate": 4.890545585179527e-06,
"loss": 1.7566,
"step": 297
},
{
"epoch": 2.5670103092783503,
"grad_norm": 0.266400686306017,
"learning_rate": 4.863188129257539e-06,
"loss": 1.7532,
"step": 298
},
{
"epoch": 2.5756013745704465,
"grad_norm": 0.28864348789360117,
"learning_rate": 4.835834771587537e-06,
"loss": 1.7401,
"step": 299
},
{
"epoch": 2.584192439862543,
"grad_norm": 0.24303483535987125,
"learning_rate": 4.808486331549824e-06,
"loss": 1.7235,
"step": 300
},
{
"epoch": 2.5927835051546393,
"grad_norm": 0.26471152605172477,
"learning_rate": 4.781143628377396e-06,
"loss": 1.7183,
"step": 301
},
{
"epoch": 2.6013745704467355,
"grad_norm": 0.2898803122430285,
"learning_rate": 4.753807481131398e-06,
"loss": 1.7858,
"step": 302
},
{
"epoch": 2.6099656357388317,
"grad_norm": 0.23208150346414513,
"learning_rate": 4.72647870867659e-06,
"loss": 1.7583,
"step": 303
},
{
"epoch": 2.618556701030928,
"grad_norm": 0.258038022151125,
"learning_rate": 4.699158129656818e-06,
"loss": 1.7564,
"step": 304
},
{
"epoch": 2.627147766323024,
"grad_norm": 0.24174383924378637,
"learning_rate": 4.671846562470489e-06,
"loss": 1.7455,
"step": 305
},
{
"epoch": 2.6357388316151202,
"grad_norm": 0.24773074133942538,
"learning_rate": 4.644544825246059e-06,
"loss": 1.7351,
"step": 306
},
{
"epoch": 2.6443298969072164,
"grad_norm": 0.25621933249081513,
"learning_rate": 4.617253735817522e-06,
"loss": 1.7251,
"step": 307
},
{
"epoch": 2.6529209621993126,
"grad_norm": 0.2559480246284349,
"learning_rate": 4.589974111699914e-06,
"loss": 1.7336,
"step": 308
},
{
"epoch": 2.661512027491409,
"grad_norm": 0.25347973386380757,
"learning_rate": 4.562706770064824e-06,
"loss": 1.7625,
"step": 309
},
{
"epoch": 2.670103092783505,
"grad_norm": 0.2572249104558495,
"learning_rate": 4.535452527715911e-06,
"loss": 1.7246,
"step": 310
},
{
"epoch": 2.6786941580756016,
"grad_norm": 0.24310238666111728,
"learning_rate": 4.508212201064446e-06,
"loss": 1.729,
"step": 311
},
{
"epoch": 2.687285223367698,
"grad_norm": 0.2381051713294497,
"learning_rate": 4.480986606104842e-06,
"loss": 1.771,
"step": 312
},
{
"epoch": 2.695876288659794,
"grad_norm": 0.2476231591131004,
"learning_rate": 4.453776558390225e-06,
"loss": 1.762,
"step": 313
},
{
"epoch": 2.70446735395189,
"grad_norm": 0.25432604692808797,
"learning_rate": 4.426582873007999e-06,
"loss": 1.7751,
"step": 314
},
{
"epoch": 2.7130584192439864,
"grad_norm": 0.24148970830649608,
"learning_rate": 4.3994063645554185e-06,
"loss": 1.7629,
"step": 315
},
{
"epoch": 2.7216494845360826,
"grad_norm": 0.2687873203876452,
"learning_rate": 4.3722478471152065e-06,
"loss": 1.6963,
"step": 316
},
{
"epoch": 2.7302405498281788,
"grad_norm": 0.25746700267483097,
"learning_rate": 4.345108134231152e-06,
"loss": 1.7348,
"step": 317
},
{
"epoch": 2.738831615120275,
"grad_norm": 0.27025200631644236,
"learning_rate": 4.31798803888375e-06,
"loss": 1.7875,
"step": 318
},
{
"epoch": 2.747422680412371,
"grad_norm": 0.24968915718366122,
"learning_rate": 4.290888373465841e-06,
"loss": 1.722,
"step": 319
},
{
"epoch": 2.7560137457044673,
"grad_norm": 0.23372356054490234,
"learning_rate": 4.2638099497582835e-06,
"loss": 1.7234,
"step": 320
},
{
"epoch": 2.7646048109965635,
"grad_norm": 0.23476817638617892,
"learning_rate": 4.236753578905627e-06,
"loss": 1.7481,
"step": 321
},
{
"epoch": 2.7731958762886597,
"grad_norm": 0.23530118596904906,
"learning_rate": 4.209720071391827e-06,
"loss": 1.7345,
"step": 322
},
{
"epoch": 2.781786941580756,
"grad_norm": 0.25951314424025645,
"learning_rate": 4.1827102370159526e-06,
"loss": 1.7523,
"step": 323
},
{
"epoch": 2.790378006872852,
"grad_norm": 0.24692340886346503,
"learning_rate": 4.155724884867937e-06,
"loss": 1.7171,
"step": 324
},
{
"epoch": 2.7989690721649483,
"grad_norm": 0.25308660216730366,
"learning_rate": 4.1287648233043366e-06,
"loss": 1.7574,
"step": 325
},
{
"epoch": 2.8075601374570445,
"grad_norm": 0.2567537654878209,
"learning_rate": 4.1018308599241245e-06,
"loss": 1.6823,
"step": 326
},
{
"epoch": 2.8161512027491407,
"grad_norm": 0.25305155829539,
"learning_rate": 4.074923801544485e-06,
"loss": 1.7565,
"step": 327
},
{
"epoch": 2.824742268041237,
"grad_norm": 0.24050906326685917,
"learning_rate": 4.048044454176658e-06,
"loss": 1.7761,
"step": 328
},
{
"epoch": 2.8333333333333335,
"grad_norm": 0.2447388353381909,
"learning_rate": 4.0211936230017915e-06,
"loss": 1.7443,
"step": 329
},
{
"epoch": 2.8419243986254297,
"grad_norm": 0.26240857077065216,
"learning_rate": 3.9943721123468124e-06,
"loss": 1.7509,
"step": 330
},
{
"epoch": 2.850515463917526,
"grad_norm": 0.28811462400375815,
"learning_rate": 3.967580725660348e-06,
"loss": 1.7528,
"step": 331
},
{
"epoch": 2.859106529209622,
"grad_norm": 0.2529330357428984,
"learning_rate": 3.940820265488644e-06,
"loss": 1.7625,
"step": 332
},
{
"epoch": 2.8676975945017182,
"grad_norm": 0.25916932781575436,
"learning_rate": 3.914091533451537e-06,
"loss": 1.7058,
"step": 333
},
{
"epoch": 2.8762886597938144,
"grad_norm": 0.2590407810513385,
"learning_rate": 3.887395330218429e-06,
"loss": 1.7161,
"step": 334
},
{
"epoch": 2.8848797250859106,
"grad_norm": 0.24066882385215688,
"learning_rate": 3.860732455484314e-06,
"loss": 1.7158,
"step": 335
},
{
"epoch": 2.893470790378007,
"grad_norm": 0.27294577615311516,
"learning_rate": 3.8341037079458125e-06,
"loss": 1.7733,
"step": 336
},
{
"epoch": 2.902061855670103,
"grad_norm": 0.2530676827059838,
"learning_rate": 3.8075098852772607e-06,
"loss": 1.7406,
"step": 337
},
{
"epoch": 2.910652920962199,
"grad_norm": 0.2628152164476677,
"learning_rate": 3.7809517841067976e-06,
"loss": 1.7058,
"step": 338
},
{
"epoch": 2.9192439862542954,
"grad_norm": 0.26638727240914895,
"learning_rate": 3.7544301999925176e-06,
"loss": 1.7496,
"step": 339
},
{
"epoch": 2.927835051546392,
"grad_norm": 0.23863083988145045,
"learning_rate": 3.727945927398628e-06,
"loss": 1.7845,
"step": 340
},
{
"epoch": 2.936426116838488,
"grad_norm": 0.24028358323758836,
"learning_rate": 3.7014997596716596e-06,
"loss": 1.745,
"step": 341
},
{
"epoch": 2.9450171821305844,
"grad_norm": 0.2524501050391268,
"learning_rate": 3.675092489016693e-06,
"loss": 1.7389,
"step": 342
},
{
"epoch": 2.9536082474226806,
"grad_norm": 0.29426755660492304,
"learning_rate": 3.6487249064736352e-06,
"loss": 1.7112,
"step": 343
},
{
"epoch": 2.9621993127147768,
"grad_norm": 0.22729315412728335,
"learning_rate": 3.622397801893518e-06,
"loss": 1.7016,
"step": 344
},
{
"epoch": 2.970790378006873,
"grad_norm": 0.25244870327576924,
"learning_rate": 3.5961119639148443e-06,
"loss": 1.7323,
"step": 345
},
{
"epoch": 2.979381443298969,
"grad_norm": 0.24503417973883346,
"learning_rate": 3.569868179939958e-06,
"loss": 1.7169,
"step": 346
},
{
"epoch": 2.9879725085910653,
"grad_norm": 0.2829554992612038,
"learning_rate": 3.543667236111458e-06,
"loss": 1.7268,
"step": 347
},
{
"epoch": 2.9965635738831615,
"grad_norm": 0.24834908746578163,
"learning_rate": 3.5175099172886507e-06,
"loss": 1.746,
"step": 348
},
{
"epoch": 3.008591065292096,
"grad_norm": 0.5922513392483557,
"learning_rate": 3.4913970070240388e-06,
"loss": 3.4119,
"step": 349
},
{
"epoch": 3.0171821305841924,
"grad_norm": 0.4020890443336188,
"learning_rate": 3.4653292875398523e-06,
"loss": 1.6271,
"step": 350
},
{
"epoch": 3.0257731958762886,
"grad_norm": 0.2846472800650302,
"learning_rate": 3.4393075397046105e-06,
"loss": 1.7354,
"step": 351
},
{
"epoch": 3.0343642611683848,
"grad_norm": 0.3696437917954285,
"learning_rate": 3.4133325430097337e-06,
"loss": 1.6708,
"step": 352
},
{
"epoch": 3.042955326460481,
"grad_norm": 0.2451913749329604,
"learning_rate": 3.3874050755461984e-06,
"loss": 1.6865,
"step": 353
},
{
"epoch": 3.051546391752577,
"grad_norm": 0.305602063686548,
"learning_rate": 3.3615259139812227e-06,
"loss": 1.6916,
"step": 354
},
{
"epoch": 3.0601374570446733,
"grad_norm": 0.3028789740936393,
"learning_rate": 3.335695833535001e-06,
"loss": 1.7133,
"step": 355
},
{
"epoch": 3.06872852233677,
"grad_norm": 0.27232200626107655,
"learning_rate": 3.309915607957487e-06,
"loss": 1.7203,
"step": 356
},
{
"epoch": 3.077319587628866,
"grad_norm": 0.24580818891429057,
"learning_rate": 3.2841860095052096e-06,
"loss": 1.6968,
"step": 357
},
{
"epoch": 3.0859106529209623,
"grad_norm": 0.30829456790986676,
"learning_rate": 3.258507808918146e-06,
"loss": 1.6632,
"step": 358
},
{
"epoch": 3.0945017182130585,
"grad_norm": 0.2493311669409693,
"learning_rate": 3.232881775396626e-06,
"loss": 1.6886,
"step": 359
},
{
"epoch": 3.1030927835051547,
"grad_norm": 0.26807381538701364,
"learning_rate": 3.2073086765783023e-06,
"loss": 1.6276,
"step": 360
},
{
"epoch": 3.111683848797251,
"grad_norm": 0.2655770739554699,
"learning_rate": 3.1817892785151426e-06,
"loss": 1.7224,
"step": 361
},
{
"epoch": 3.120274914089347,
"grad_norm": 0.29191249153928056,
"learning_rate": 3.1563243456504877e-06,
"loss": 1.6464,
"step": 362
},
{
"epoch": 3.1288659793814433,
"grad_norm": 0.26142611464602167,
"learning_rate": 3.1309146407961565e-06,
"loss": 1.7042,
"step": 363
},
{
"epoch": 3.1374570446735395,
"grad_norm": 0.25581281444307696,
"learning_rate": 3.1055609251095874e-06,
"loss": 1.684,
"step": 364
},
{
"epoch": 3.1460481099656357,
"grad_norm": 0.25645637858338083,
"learning_rate": 3.0802639580710465e-06,
"loss": 1.6748,
"step": 365
},
{
"epoch": 3.154639175257732,
"grad_norm": 0.2537099488131362,
"learning_rate": 3.0550244974608675e-06,
"loss": 1.6751,
"step": 366
},
{
"epoch": 3.163230240549828,
"grad_norm": 0.2622401902439414,
"learning_rate": 3.0298432993367577e-06,
"loss": 1.6762,
"step": 367
},
{
"epoch": 3.1718213058419242,
"grad_norm": 0.26777154057402036,
"learning_rate": 3.0047211180111537e-06,
"loss": 1.6637,
"step": 368
},
{
"epoch": 3.1804123711340204,
"grad_norm": 0.2697670915131976,
"learning_rate": 2.979658706028619e-06,
"loss": 1.6409,
"step": 369
},
{
"epoch": 3.189003436426117,
"grad_norm": 0.25302462808285775,
"learning_rate": 2.9546568141433007e-06,
"loss": 1.6811,
"step": 370
},
{
"epoch": 3.1975945017182132,
"grad_norm": 0.2926705457518936,
"learning_rate": 2.9297161912964476e-06,
"loss": 1.6811,
"step": 371
},
{
"epoch": 3.2061855670103094,
"grad_norm": 0.24258346567546993,
"learning_rate": 2.904837584593968e-06,
"loss": 1.6987,
"step": 372
},
{
"epoch": 3.2147766323024056,
"grad_norm": 0.27821484417248077,
"learning_rate": 2.880021739284053e-06,
"loss": 1.7002,
"step": 373
},
{
"epoch": 3.223367697594502,
"grad_norm": 0.2583260720453906,
"learning_rate": 2.8552693987348533e-06,
"loss": 1.6581,
"step": 374
},
{
"epoch": 3.231958762886598,
"grad_norm": 0.2995718829517528,
"learning_rate": 2.83058130441221e-06,
"loss": 1.698,
"step": 375
},
{
"epoch": 3.240549828178694,
"grad_norm": 0.23209310903531505,
"learning_rate": 2.8059581958574434e-06,
"loss": 1.6639,
"step": 376
},
{
"epoch": 3.2491408934707904,
"grad_norm": 0.3141583986931671,
"learning_rate": 2.781400810665201e-06,
"loss": 1.6306,
"step": 377
},
{
"epoch": 3.2577319587628866,
"grad_norm": 0.26043046141639037,
"learning_rate": 2.7569098844613616e-06,
"loss": 1.6428,
"step": 378
},
{
"epoch": 3.2663230240549828,
"grad_norm": 0.2682410210651436,
"learning_rate": 2.7324861508810007e-06,
"loss": 1.7031,
"step": 379
},
{
"epoch": 3.274914089347079,
"grad_norm": 0.23627466269523362,
"learning_rate": 2.708130341546407e-06,
"loss": 1.6912,
"step": 380
},
{
"epoch": 3.283505154639175,
"grad_norm": 0.25452308447859706,
"learning_rate": 2.6838431860451797e-06,
"loss": 1.6972,
"step": 381
},
{
"epoch": 3.2920962199312713,
"grad_norm": 0.2684481414005352,
"learning_rate": 2.659625411908366e-06,
"loss": 1.6126,
"step": 382
},
{
"epoch": 3.3006872852233675,
"grad_norm": 0.28632630512865187,
"learning_rate": 2.635477744588658e-06,
"loss": 1.7118,
"step": 383
},
{
"epoch": 3.3092783505154637,
"grad_norm": 0.2480954277626528,
"learning_rate": 2.611400907438685e-06,
"loss": 1.678,
"step": 384
},
{
"epoch": 3.31786941580756,
"grad_norm": 0.27009672556246167,
"learning_rate": 2.587395621689325e-06,
"loss": 1.6571,
"step": 385
},
{
"epoch": 3.3264604810996565,
"grad_norm": 0.2488236290385475,
"learning_rate": 2.563462606428101e-06,
"loss": 1.698,
"step": 386
},
{
"epoch": 3.3350515463917527,
"grad_norm": 0.265013925369111,
"learning_rate": 2.5396025785776545e-06,
"loss": 1.6874,
"step": 387
},
{
"epoch": 3.343642611683849,
"grad_norm": 0.2501632529845309,
"learning_rate": 2.515816252874258e-06,
"loss": 1.6518,
"step": 388
},
{
"epoch": 3.352233676975945,
"grad_norm": 0.26963209575817787,
"learning_rate": 2.4921043418464085e-06,
"loss": 1.723,
"step": 389
},
{
"epoch": 3.3608247422680413,
"grad_norm": 0.2439138058109664,
"learning_rate": 2.4684675557934766e-06,
"loss": 1.6514,
"step": 390
},
{
"epoch": 3.3694158075601375,
"grad_norm": 0.25131424095425053,
"learning_rate": 2.4449066027644473e-06,
"loss": 1.6717,
"step": 391
},
{
"epoch": 3.3780068728522337,
"grad_norm": 0.23423918940451338,
"learning_rate": 2.4214221885366918e-06,
"loss": 1.6818,
"step": 392
},
{
"epoch": 3.38659793814433,
"grad_norm": 0.23504278302633322,
"learning_rate": 2.398015016594828e-06,
"loss": 1.6851,
"step": 393
},
{
"epoch": 3.395189003436426,
"grad_norm": 0.23108912440397839,
"learning_rate": 2.3746857881096586e-06,
"loss": 1.6573,
"step": 394
},
{
"epoch": 3.4037800687285222,
"grad_norm": 0.2630972598480657,
"learning_rate": 2.351435201917159e-06,
"loss": 1.6736,
"step": 395
},
{
"epoch": 3.4123711340206184,
"grad_norm": 0.2565037203189053,
"learning_rate": 2.32826395449754e-06,
"loss": 1.6683,
"step": 396
},
{
"epoch": 3.4209621993127146,
"grad_norm": 0.27931135209944863,
"learning_rate": 2.3051727399543934e-06,
"loss": 1.6755,
"step": 397
},
{
"epoch": 3.429553264604811,
"grad_norm": 0.29212242360552815,
"learning_rate": 2.282162249993895e-06,
"loss": 1.6334,
"step": 398
},
{
"epoch": 3.4381443298969074,
"grad_norm": 0.26981834777022456,
"learning_rate": 2.259233173904084e-06,
"loss": 1.708,
"step": 399
},
{
"epoch": 3.4467353951890036,
"grad_norm": 0.27252952116031537,
"learning_rate": 2.2363861985342156e-06,
"loss": 1.7145,
"step": 400
},
{
"epoch": 3.4553264604811,
"grad_norm": 0.26856882317170744,
"learning_rate": 2.2136220082741876e-06,
"loss": 1.6999,
"step": 401
},
{
"epoch": 3.463917525773196,
"grad_norm": 0.26539546923935187,
"learning_rate": 2.1909412850340395e-06,
"loss": 1.6981,
"step": 402
},
{
"epoch": 3.472508591065292,
"grad_norm": 0.24609298736654442,
"learning_rate": 2.168344708223519e-06,
"loss": 1.6949,
"step": 403
},
{
"epoch": 3.4810996563573884,
"grad_norm": 0.23740060447946054,
"learning_rate": 2.1458329547317384e-06,
"loss": 1.7044,
"step": 404
},
{
"epoch": 3.4896907216494846,
"grad_norm": 0.26722945895883293,
"learning_rate": 2.1234066989068972e-06,
"loss": 1.6929,
"step": 405
},
{
"epoch": 3.4982817869415808,
"grad_norm": 0.23155932481833713,
"learning_rate": 2.1010666125360767e-06,
"loss": 1.6896,
"step": 406
},
{
"epoch": 3.506872852233677,
"grad_norm": 0.2711065659946207,
"learning_rate": 2.0788133648251207e-06,
"loss": 1.6726,
"step": 407
},
{
"epoch": 3.515463917525773,
"grad_norm": 0.23502373531128723,
"learning_rate": 2.0566476223785857e-06,
"loss": 1.6584,
"step": 408
},
{
"epoch": 3.5240549828178693,
"grad_norm": 0.23647872007839724,
"learning_rate": 2.0345700491797786e-06,
"loss": 1.6842,
"step": 409
},
{
"epoch": 3.5326460481099655,
"grad_norm": 0.25531191284936877,
"learning_rate": 2.0125813065708568e-06,
"loss": 1.7291,
"step": 410
},
{
"epoch": 3.5412371134020617,
"grad_norm": 0.2338734563419503,
"learning_rate": 1.9906820532330262e-06,
"loss": 1.6603,
"step": 411
},
{
"epoch": 3.549828178694158,
"grad_norm": 0.24952190919610998,
"learning_rate": 1.9688729451668116e-06,
"loss": 1.6538,
"step": 412
},
{
"epoch": 3.558419243986254,
"grad_norm": 0.24615091256665883,
"learning_rate": 1.947154635672393e-06,
"loss": 1.6823,
"step": 413
},
{
"epoch": 3.5670103092783503,
"grad_norm": 0.24544513932589554,
"learning_rate": 1.925527775330049e-06,
"loss": 1.6561,
"step": 414
},
{
"epoch": 3.5756013745704465,
"grad_norm": 0.2547223286101074,
"learning_rate": 1.9039930119806698e-06,
"loss": 1.7017,
"step": 415
},
{
"epoch": 3.584192439862543,
"grad_norm": 0.24474946887671897,
"learning_rate": 1.8825509907063328e-06,
"loss": 1.6966,
"step": 416
},
{
"epoch": 3.5927835051546393,
"grad_norm": 0.23579767821458228,
"learning_rate": 1.8612023538109998e-06,
"loss": 1.6644,
"step": 417
},
{
"epoch": 3.6013745704467355,
"grad_norm": 0.236127895866837,
"learning_rate": 1.8399477408012645e-06,
"loss": 1.6935,
"step": 418
},
{
"epoch": 3.6099656357388317,
"grad_norm": 0.2663423480735363,
"learning_rate": 1.8187877883672024e-06,
"loss": 1.6871,
"step": 419
},
{
"epoch": 3.618556701030928,
"grad_norm": 0.23972519408323606,
"learning_rate": 1.797723130363288e-06,
"loss": 1.6964,
"step": 420
},
{
"epoch": 3.627147766323024,
"grad_norm": 0.2378828947434377,
"learning_rate": 1.7767543977894198e-06,
"loss": 1.664,
"step": 421
},
{
"epoch": 3.6357388316151202,
"grad_norm": 0.22727687711471692,
"learning_rate": 1.755882218772018e-06,
"loss": 1.6827,
"step": 422
},
{
"epoch": 3.6443298969072164,
"grad_norm": 0.2255094104159532,
"learning_rate": 1.7351072185451934e-06,
"loss": 1.6893,
"step": 423
},
{
"epoch": 3.6529209621993126,
"grad_norm": 0.24809139447614564,
"learning_rate": 1.7144300194320357e-06,
"loss": 1.6816,
"step": 424
},
{
"epoch": 3.661512027491409,
"grad_norm": 0.25186582278678304,
"learning_rate": 1.6938512408259655e-06,
"loss": 1.6653,
"step": 425
},
{
"epoch": 3.670103092783505,
"grad_norm": 0.2559208529812737,
"learning_rate": 1.6733714991721738e-06,
"loss": 1.6664,
"step": 426
},
{
"epoch": 3.6786941580756016,
"grad_norm": 0.25827292208337704,
"learning_rate": 1.652991407949167e-06,
"loss": 1.6452,
"step": 427
},
{
"epoch": 3.687285223367698,
"grad_norm": 0.24886719673164634,
"learning_rate": 1.6327115776503833e-06,
"loss": 1.7094,
"step": 428
},
{
"epoch": 3.695876288659794,
"grad_norm": 0.2967568140556356,
"learning_rate": 1.6125326157659048e-06,
"loss": 1.6708,
"step": 429
},
{
"epoch": 3.70446735395189,
"grad_norm": 0.24610789821189943,
"learning_rate": 1.5924551267642641e-06,
"loss": 1.7097,
"step": 430
},
{
"epoch": 3.7130584192439864,
"grad_norm": 0.26064599580011155,
"learning_rate": 1.572479712074333e-06,
"loss": 1.6399,
"step": 431
},
{
"epoch": 3.7216494845360826,
"grad_norm": 0.25823550620207686,
"learning_rate": 1.5526069700673108e-06,
"loss": 1.6418,
"step": 432
},
{
"epoch": 3.7302405498281788,
"grad_norm": 0.2509972834327566,
"learning_rate": 1.532837496038792e-06,
"loss": 1.6782,
"step": 433
},
{
"epoch": 3.738831615120275,
"grad_norm": 0.25327011530278,
"learning_rate": 1.5131718821909435e-06,
"loss": 1.705,
"step": 434
},
{
"epoch": 3.747422680412371,
"grad_norm": 0.24464619937684007,
"learning_rate": 1.4936107176147606e-06,
"loss": 1.7135,
"step": 435
},
{
"epoch": 3.7560137457044673,
"grad_norm": 0.25124166991508706,
"learning_rate": 1.4741545882724213e-06,
"loss": 1.7038,
"step": 436
},
{
"epoch": 3.7646048109965635,
"grad_norm": 0.24424448019564532,
"learning_rate": 1.4548040769797255e-06,
"loss": 1.6424,
"step": 437
},
{
"epoch": 3.7731958762886597,
"grad_norm": 0.25292186500202485,
"learning_rate": 1.4355597633886576e-06,
"loss": 1.6756,
"step": 438
},
{
"epoch": 3.781786941580756,
"grad_norm": 0.23840691684985385,
"learning_rate": 1.416422223969996e-06,
"loss": 1.6786,
"step": 439
},
{
"epoch": 3.790378006872852,
"grad_norm": 0.25906625423522317,
"learning_rate": 1.3973920319960654e-06,
"loss": 1.6766,
"step": 440
},
{
"epoch": 3.7989690721649483,
"grad_norm": 0.26153465080891225,
"learning_rate": 1.378469757523554e-06,
"loss": 1.6828,
"step": 441
},
{
"epoch": 3.8075601374570445,
"grad_norm": 0.23917358482231746,
"learning_rate": 1.3596559673764421e-06,
"loss": 1.6804,
"step": 442
},
{
"epoch": 3.8161512027491407,
"grad_norm": 0.2396941834079674,
"learning_rate": 1.3409512251290164e-06,
"loss": 1.6668,
"step": 443
},
{
"epoch": 3.824742268041237,
"grad_norm": 0.2629866189082095,
"learning_rate": 1.322356091088996e-06,
"loss": 1.7056,
"step": 444
},
{
"epoch": 3.8333333333333335,
"grad_norm": 0.2607750520114126,
"learning_rate": 1.303871122280742e-06,
"loss": 1.7074,
"step": 445
},
{
"epoch": 3.8419243986254297,
"grad_norm": 0.3057642778361461,
"learning_rate": 1.2854968724285755e-06,
"loss": 1.6656,
"step": 446
},
{
"epoch": 3.850515463917526,
"grad_norm": 0.23584672793493383,
"learning_rate": 1.2672338919401866e-06,
"loss": 1.7042,
"step": 447
},
{
"epoch": 3.859106529209622,
"grad_norm": 0.2327615685413911,
"learning_rate": 1.2490827278901513e-06,
"loss": 1.638,
"step": 448
},
{
"epoch": 3.8676975945017182,
"grad_norm": 0.2538402919887143,
"learning_rate": 1.2310439240035415e-06,
"loss": 1.6697,
"step": 449
},
{
"epoch": 3.8762886597938144,
"grad_norm": 0.24590706784319505,
"learning_rate": 1.2131180206396331e-06,
"loss": 1.6567,
"step": 450
},
{
"epoch": 3.8848797250859106,
"grad_norm": 0.24366334985364868,
"learning_rate": 1.195305554775728e-06,
"loss": 1.6873,
"step": 451
},
{
"epoch": 3.893470790378007,
"grad_norm": 0.24933468260526015,
"learning_rate": 1.177607059991065e-06,
"loss": 1.6768,
"step": 452
},
{
"epoch": 3.902061855670103,
"grad_norm": 0.26229582665690454,
"learning_rate": 1.1600230664508288e-06,
"loss": 1.6861,
"step": 453
},
{
"epoch": 3.910652920962199,
"grad_norm": 0.25037526400185023,
"learning_rate": 1.1425541008902852e-06,
"loss": 1.709,
"step": 454
},
{
"epoch": 3.9192439862542954,
"grad_norm": 0.2436612620158865,
"learning_rate": 1.1252006865989868e-06,
"loss": 1.6697,
"step": 455
},
{
"epoch": 3.927835051546392,
"grad_norm": 0.22958914781932735,
"learning_rate": 1.107963343405103e-06,
"loss": 1.6828,
"step": 456
},
{
"epoch": 3.936426116838488,
"grad_norm": 0.24154992738826303,
"learning_rate": 1.0908425876598512e-06,
"loss": 1.7176,
"step": 457
},
{
"epoch": 3.9450171821305844,
"grad_norm": 0.24344258073527894,
"learning_rate": 1.0738389322220276e-06,
"loss": 1.6401,
"step": 458
},
{
"epoch": 3.9536082474226806,
"grad_norm": 0.2390591834019147,
"learning_rate": 1.0569528864426444e-06,
"loss": 1.6637,
"step": 459
},
{
"epoch": 3.9621993127147768,
"grad_norm": 0.24876356770228733,
"learning_rate": 1.0401849561496647e-06,
"loss": 1.672,
"step": 460
},
{
"epoch": 3.970790378006873,
"grad_norm": 0.24788035367409328,
"learning_rate": 1.0235356436328675e-06,
"loss": 1.6737,
"step": 461
},
{
"epoch": 3.979381443298969,
"grad_norm": 0.22083500615355336,
"learning_rate": 1.007005447628785e-06,
"loss": 1.6832,
"step": 462
},
{
"epoch": 3.9879725085910653,
"grad_norm": 0.23586980332078747,
"learning_rate": 9.905948633057666e-07,
"loss": 1.7014,
"step": 463
},
{
"epoch": 3.9965635738831615,
"grad_norm": 0.23418365998229526,
"learning_rate": 9.743043822491528e-07,
"loss": 1.6619,
"step": 464
},
{
"epoch": 4.008591065292096,
"grad_norm": 0.5867691500106113,
"learning_rate": 9.58134492446543e-07,
"loss": 3.3257,
"step": 465
},
{
"epoch": 4.017182130584192,
"grad_norm": 0.22848089001074176,
"learning_rate": 9.420856782731774e-07,
"loss": 1.6432,
"step": 466
},
{
"epoch": 4.025773195876289,
"grad_norm": 0.26643759471086603,
"learning_rate": 9.26158420477431e-07,
"loss": 1.6255,
"step": 467
},
{
"epoch": 4.034364261168385,
"grad_norm": 0.24279390688421273,
"learning_rate": 9.10353196166412e-07,
"loss": 1.6936,
"step": 468
},
{
"epoch": 4.042955326460481,
"grad_norm": 0.24978730161476037,
"learning_rate": 8.946704787916676e-07,
"loss": 1.6602,
"step": 469
},
{
"epoch": 4.051546391752577,
"grad_norm": 0.24882895961508988,
"learning_rate": 8.791107381350028e-07,
"loss": 1.6557,
"step": 470
},
{
"epoch": 4.060137457044673,
"grad_norm": 0.24038548638994323,
"learning_rate": 8.636744402944075e-07,
"loss": 1.649,
"step": 471
},
{
"epoch": 4.0687285223367695,
"grad_norm": 0.23343234428832116,
"learning_rate": 8.483620476700977e-07,
"loss": 1.6623,
"step": 472
},
{
"epoch": 4.077319587628866,
"grad_norm": 0.2479346870830623,
"learning_rate": 8.33174018950656e-07,
"loss": 1.6911,
"step": 473
},
{
"epoch": 4.085910652920962,
"grad_norm": 0.23925299666712668,
"learning_rate": 8.181108090993001e-07,
"loss": 1.6349,
"step": 474
},
{
"epoch": 4.094501718213058,
"grad_norm": 0.23889326115734966,
"learning_rate": 8.031728693402502e-07,
"loss": 1.6721,
"step": 475
},
{
"epoch": 4.103092783505154,
"grad_norm": 0.2296301969294035,
"learning_rate": 7.883606471452138e-07,
"loss": 1.6814,
"step": 476
},
{
"epoch": 4.1116838487972505,
"grad_norm": 0.23775555355784386,
"learning_rate": 7.736745862199785e-07,
"loss": 1.6394,
"step": 477
},
{
"epoch": 4.120274914089347,
"grad_norm": 0.24046066783607495,
"learning_rate": 7.59115126491124e-07,
"loss": 1.6729,
"step": 478
},
{
"epoch": 4.128865979381443,
"grad_norm": 0.23623768362714573,
"learning_rate": 7.446827040928439e-07,
"loss": 1.6354,
"step": 479
},
{
"epoch": 4.13745704467354,
"grad_norm": 0.2575549346066372,
"learning_rate": 7.303777513538762e-07,
"loss": 1.6333,
"step": 480
},
{
"epoch": 4.146048109965636,
"grad_norm": 0.24860345971574782,
"learning_rate": 7.162006967845602e-07,
"loss": 1.6907,
"step": 481
},
{
"epoch": 4.154639175257732,
"grad_norm": 0.23054251038882875,
"learning_rate": 7.021519650639952e-07,
"loss": 1.6044,
"step": 482
},
{
"epoch": 4.1632302405498285,
"grad_norm": 0.24550821257644512,
"learning_rate": 6.882319770273193e-07,
"loss": 1.6748,
"step": 483
},
{
"epoch": 4.171821305841925,
"grad_norm": 0.23474577539619082,
"learning_rate": 6.744411496531045e-07,
"loss": 1.6451,
"step": 484
},
{
"epoch": 4.180412371134021,
"grad_norm": 0.22708100508998744,
"learning_rate": 6.607798960508693e-07,
"loss": 1.6892,
"step": 485
},
{
"epoch": 4.189003436426117,
"grad_norm": 0.23352825875050276,
"learning_rate": 6.472486254486954e-07,
"loss": 1.632,
"step": 486
},
{
"epoch": 4.197594501718213,
"grad_norm": 0.24648114546627972,
"learning_rate": 6.338477431809764e-07,
"loss": 1.6557,
"step": 487
},
{
"epoch": 4.206185567010309,
"grad_norm": 0.24736170079000772,
"learning_rate": 6.205776506762729e-07,
"loss": 1.6461,
"step": 488
},
{
"epoch": 4.214776632302406,
"grad_norm": 0.2380927846200124,
"learning_rate": 6.074387454452891e-07,
"loss": 1.7005,
"step": 489
},
{
"epoch": 4.223367697594502,
"grad_norm": 0.25104258830397624,
"learning_rate": 5.944314210689611e-07,
"loss": 1.6407,
"step": 490
},
{
"epoch": 4.231958762886598,
"grad_norm": 0.24999941862158437,
"learning_rate": 5.815560671866721e-07,
"loss": 1.6439,
"step": 491
},
{
"epoch": 4.240549828178694,
"grad_norm": 0.22907317626435914,
"learning_rate": 5.688130694845817e-07,
"loss": 1.663,
"step": 492
},
{
"epoch": 4.24914089347079,
"grad_norm": 0.24260790107655203,
"learning_rate": 5.562028096840638e-07,
"loss": 1.6133,
"step": 493
},
{
"epoch": 4.257731958762887,
"grad_norm": 0.2449427663820236,
"learning_rate": 5.437256655302814e-07,
"loss": 1.6833,
"step": 494
},
{
"epoch": 4.266323024054983,
"grad_norm": 0.2452720567127199,
"learning_rate": 5.313820107808665e-07,
"loss": 1.6382,
"step": 495
},
{
"epoch": 4.274914089347079,
"grad_norm": 0.24245668714636706,
"learning_rate": 5.191722151947227e-07,
"loss": 1.6065,
"step": 496
},
{
"epoch": 4.283505154639175,
"grad_norm": 0.24686375253425183,
"learning_rate": 5.07096644520954e-07,
"loss": 1.6028,
"step": 497
},
{
"epoch": 4.292096219931271,
"grad_norm": 0.24325465896590973,
"learning_rate": 4.951556604879049e-07,
"loss": 1.627,
"step": 498
},
{
"epoch": 4.3006872852233675,
"grad_norm": 0.25561506607383994,
"learning_rate": 4.83349620792325e-07,
"loss": 1.6532,
"step": 499
},
{
"epoch": 4.309278350515464,
"grad_norm": 0.24432420895344698,
"learning_rate": 4.716788790886545e-07,
"loss": 1.7054,
"step": 500
},
{
"epoch": 4.31786941580756,
"grad_norm": 0.24652763938837308,
"learning_rate": 4.601437849784318e-07,
"loss": 1.6584,
"step": 501
},
{
"epoch": 4.326460481099656,
"grad_norm": 0.24081112601884344,
"learning_rate": 4.487446839998194e-07,
"loss": 1.6579,
"step": 502
},
{
"epoch": 4.335051546391752,
"grad_norm": 0.24077762120128518,
"learning_rate": 4.374819176172501e-07,
"loss": 1.611,
"step": 503
},
{
"epoch": 4.3436426116838485,
"grad_norm": 0.23551283193929728,
"learning_rate": 4.263558232112064e-07,
"loss": 1.6486,
"step": 504
},
{
"epoch": 4.352233676975945,
"grad_norm": 0.24030554806756774,
"learning_rate": 4.153667340681067e-07,
"loss": 1.5969,
"step": 505
},
{
"epoch": 4.360824742268041,
"grad_norm": 0.23691286610236578,
"learning_rate": 4.045149793703257e-07,
"loss": 1.6753,
"step": 506
},
{
"epoch": 4.369415807560138,
"grad_norm": 0.23903869191328392,
"learning_rate": 3.938008841863289e-07,
"loss": 1.646,
"step": 507
},
{
"epoch": 4.378006872852234,
"grad_norm": 0.2389287897855053,
"learning_rate": 3.832247694609442e-07,
"loss": 1.6505,
"step": 508
},
{
"epoch": 4.38659793814433,
"grad_norm": 0.2384304058058299,
"learning_rate": 3.7278695200573754e-07,
"loss": 1.6516,
"step": 509
},
{
"epoch": 4.3951890034364265,
"grad_norm": 0.23170007570321202,
"learning_rate": 3.6248774448952695e-07,
"loss": 1.6408,
"step": 510
},
{
"epoch": 4.403780068728523,
"grad_norm": 0.2451725310139443,
"learning_rate": 3.52327455429019e-07,
"loss": 1.6548,
"step": 511
},
{
"epoch": 4.412371134020619,
"grad_norm": 0.2311001266895989,
"learning_rate": 3.423063891795647e-07,
"loss": 1.674,
"step": 512
},
{
"epoch": 4.420962199312715,
"grad_norm": 0.22939351746207862,
"learning_rate": 3.324248459260393e-07,
"loss": 1.668,
"step": 513
},
{
"epoch": 4.429553264604811,
"grad_norm": 0.24921772990695243,
"learning_rate": 3.2268312167385687e-07,
"loss": 1.6644,
"step": 514
},
{
"epoch": 4.438144329896907,
"grad_norm": 0.24729846809806982,
"learning_rate": 3.1308150824009785e-07,
"loss": 1.67,
"step": 515
},
{
"epoch": 4.446735395189004,
"grad_norm": 0.22640721408797101,
"learning_rate": 3.0362029324477015e-07,
"loss": 1.6431,
"step": 516
},
{
"epoch": 4.4553264604811,
"grad_norm": 0.2514965526696115,
"learning_rate": 2.942997601021924e-07,
"loss": 1.6297,
"step": 517
},
{
"epoch": 4.463917525773196,
"grad_norm": 0.24800570829587237,
"learning_rate": 2.851201880125043e-07,
"loss": 1.5746,
"step": 518
},
{
"epoch": 4.472508591065292,
"grad_norm": 0.23974608400063066,
"learning_rate": 2.760818519533037e-07,
"loss": 1.6556,
"step": 519
},
{
"epoch": 4.481099656357388,
"grad_norm": 0.23413547062949885,
"learning_rate": 2.6718502267140844e-07,
"loss": 1.6916,
"step": 520
},
{
"epoch": 4.489690721649485,
"grad_norm": 0.24104933467023465,
"learning_rate": 2.584299666747475e-07,
"loss": 1.6593,
"step": 521
},
{
"epoch": 4.498281786941581,
"grad_norm": 0.23289250853229676,
"learning_rate": 2.4981694622437546e-07,
"loss": 1.6469,
"step": 522
},
{
"epoch": 4.506872852233677,
"grad_norm": 0.23052100983273746,
"learning_rate": 2.4134621932661916e-07,
"loss": 1.6194,
"step": 523
},
{
"epoch": 4.515463917525773,
"grad_norm": 0.220108887733881,
"learning_rate": 2.330180397253473e-07,
"loss": 1.6437,
"step": 524
},
{
"epoch": 4.524054982817869,
"grad_norm": 0.22944959560694397,
"learning_rate": 2.2483265689436929e-07,
"loss": 1.6934,
"step": 525
},
{
"epoch": 4.5326460481099655,
"grad_norm": 0.2417227052649889,
"learning_rate": 2.167903160299617e-07,
"loss": 1.6367,
"step": 526
},
{
"epoch": 4.541237113402062,
"grad_norm": 0.22985546756470726,
"learning_rate": 2.0889125804352595e-07,
"loss": 1.6216,
"step": 527
},
{
"epoch": 4.549828178694158,
"grad_norm": 0.24066618548462096,
"learning_rate": 2.0113571955436895e-07,
"loss": 1.6804,
"step": 528
},
{
"epoch": 4.558419243986254,
"grad_norm": 0.2416438260307757,
"learning_rate": 1.9352393288261717e-07,
"loss": 1.6393,
"step": 529
},
{
"epoch": 4.56701030927835,
"grad_norm": 0.239698741927309,
"learning_rate": 1.8605612604225388e-07,
"loss": 1.6105,
"step": 530
},
{
"epoch": 4.5756013745704465,
"grad_norm": 0.24413712058845366,
"learning_rate": 1.787325227342951e-07,
"loss": 1.6315,
"step": 531
},
{
"epoch": 4.584192439862543,
"grad_norm": 0.23637696946735323,
"learning_rate": 1.715533423400817e-07,
"loss": 1.6307,
"step": 532
},
{
"epoch": 4.592783505154639,
"grad_norm": 0.23872270203188434,
"learning_rate": 1.6451879991471186e-07,
"loss": 1.6844,
"step": 533
},
{
"epoch": 4.601374570446735,
"grad_norm": 0.22209573936584914,
"learning_rate": 1.576291061805979e-07,
"loss": 1.6774,
"step": 534
},
{
"epoch": 4.609965635738831,
"grad_norm": 0.22781684978261602,
"learning_rate": 1.5088446752115403e-07,
"loss": 1.6365,
"step": 535
},
{
"epoch": 4.618556701030927,
"grad_norm": 0.22231536055946152,
"learning_rate": 1.4428508597461587e-07,
"loss": 1.6518,
"step": 536
},
{
"epoch": 4.627147766323024,
"grad_norm": 0.24597301241764716,
"learning_rate": 1.378311592279835e-07,
"loss": 1.6387,
"step": 537
},
{
"epoch": 4.63573883161512,
"grad_norm": 0.23651255807420338,
"learning_rate": 1.3152288061110518e-07,
"loss": 1.6367,
"step": 538
},
{
"epoch": 4.644329896907217,
"grad_norm": 0.24194404648436793,
"learning_rate": 1.253604390908819e-07,
"loss": 1.6404,
"step": 539
},
{
"epoch": 4.652920962199313,
"grad_norm": 0.24096603427636226,
"learning_rate": 1.193440192656109e-07,
"loss": 1.6487,
"step": 540
},
{
"epoch": 4.661512027491409,
"grad_norm": 0.23073526119633206,
"learning_rate": 1.1347380135945108e-07,
"loss": 1.6366,
"step": 541
},
{
"epoch": 4.670103092783505,
"grad_norm": 0.23367410856487186,
"learning_rate": 1.0774996121702907e-07,
"loss": 1.66,
"step": 542
},
{
"epoch": 4.678694158075602,
"grad_norm": 0.23333593518319265,
"learning_rate": 1.0217267029816736e-07,
"loss": 1.6204,
"step": 543
},
{
"epoch": 4.687285223367698,
"grad_norm": 0.2400651207345224,
"learning_rate": 9.674209567275161e-08,
"loss": 1.6296,
"step": 544
},
{
"epoch": 4.695876288659794,
"grad_norm": 0.22715088472979858,
"learning_rate": 9.145840001572537e-08,
"loss": 1.687,
"step": 545
},
{
"epoch": 4.70446735395189,
"grad_norm": 0.2304516857645299,
"learning_rate": 8.632174160221496e-08,
"loss": 1.6795,
"step": 546
},
{
"epoch": 4.713058419243986,
"grad_norm": 0.22846319291531494,
"learning_rate": 8.133227430279055e-08,
"loss": 1.6327,
"step": 547
},
{
"epoch": 4.721649484536083,
"grad_norm": 0.22995523965104772,
"learning_rate": 7.649014757885597e-08,
"loss": 1.6455,
"step": 548
},
{
"epoch": 4.730240549828179,
"grad_norm": 0.23040086094496828,
"learning_rate": 7.179550647817224e-08,
"loss": 1.6503,
"step": 549
},
{
"epoch": 4.738831615120275,
"grad_norm": 0.22952265926622467,
"learning_rate": 6.724849163050995e-08,
"loss": 1.6438,
"step": 550
},
{
"epoch": 4.747422680412371,
"grad_norm": 0.24926903275682435,
"learning_rate": 6.28492392434421e-08,
"loss": 1.6643,
"step": 551
},
{
"epoch": 4.756013745704467,
"grad_norm": 0.2358707255568968,
"learning_rate": 5.8597881098257924e-08,
"loss": 1.7022,
"step": 552
},
{
"epoch": 4.7646048109965635,
"grad_norm": 0.2232349079812786,
"learning_rate": 5.4494544546018216e-08,
"loss": 1.6886,
"step": 553
},
{
"epoch": 4.77319587628866,
"grad_norm": 0.229127449017454,
"learning_rate": 5.0539352503741756e-08,
"loss": 1.6316,
"step": 554
},
{
"epoch": 4.781786941580756,
"grad_norm": 0.23843128476180983,
"learning_rate": 4.67324234507216e-08,
"loss": 1.6608,
"step": 555
},
{
"epoch": 4.790378006872852,
"grad_norm": 0.22282768820417342,
"learning_rate": 4.307387142497399e-08,
"loss": 1.6164,
"step": 556
},
{
"epoch": 4.798969072164948,
"grad_norm": 0.23128077792110194,
"learning_rate": 3.956380601982668e-08,
"loss": 1.6358,
"step": 557
},
{
"epoch": 4.8075601374570445,
"grad_norm": 0.2362232510882238,
"learning_rate": 3.620233238063375e-08,
"loss": 1.6744,
"step": 558
},
{
"epoch": 4.816151202749141,
"grad_norm": 0.227329534924112,
"learning_rate": 3.2989551201624836e-08,
"loss": 1.6351,
"step": 559
},
{
"epoch": 4.824742268041237,
"grad_norm": 0.2567304697567019,
"learning_rate": 2.992555872289082e-08,
"loss": 1.6755,
"step": 560
},
{
"epoch": 4.833333333333333,
"grad_norm": 0.24133539792181574,
"learning_rate": 2.7010446727498974e-08,
"loss": 1.6486,
"step": 561
},
{
"epoch": 4.841924398625429,
"grad_norm": 0.24504456603848984,
"learning_rate": 2.4244302538746765e-08,
"loss": 1.6242,
"step": 562
},
{
"epoch": 4.850515463917525,
"grad_norm": 0.2415111188070595,
"learning_rate": 2.162720901754234e-08,
"loss": 1.6145,
"step": 563
},
{
"epoch": 4.859106529209622,
"grad_norm": 0.22918242520793192,
"learning_rate": 1.9159244559924795e-08,
"loss": 1.6562,
"step": 564
},
{
"epoch": 4.867697594501719,
"grad_norm": 0.2340766045949212,
"learning_rate": 1.6840483094713867e-08,
"loss": 1.659,
"step": 565
},
{
"epoch": 4.876288659793815,
"grad_norm": 0.22284549945463064,
"learning_rate": 1.4670994081297796e-08,
"loss": 1.6773,
"step": 566
},
{
"epoch": 4.884879725085911,
"grad_norm": 0.2370179289602322,
"learning_rate": 1.2650842507550554e-08,
"loss": 1.618,
"step": 567
},
{
"epoch": 4.893470790378007,
"grad_norm": 0.23457628394636465,
"learning_rate": 1.078008888788673e-08,
"loss": 1.6379,
"step": 568
},
{
"epoch": 4.902061855670103,
"grad_norm": 0.23811030764385882,
"learning_rate": 9.058789261446876e-09,
"loss": 1.6881,
"step": 569
},
{
"epoch": 4.9106529209622,
"grad_norm": 0.2389166558669095,
"learning_rate": 7.48699519042051e-09,
"loss": 1.6499,
"step": 570
},
{
"epoch": 4.919243986254296,
"grad_norm": 0.22731504830513133,
"learning_rate": 6.0647537585017956e-09,
"loss": 1.6758,
"step": 571
},
{
"epoch": 4.927835051546392,
"grad_norm": 0.2389929769641339,
"learning_rate": 4.792107569476789e-09,
"loss": 1.6404,
"step": 572
},
{
"epoch": 4.936426116838488,
"grad_norm": 0.2426297580002159,
"learning_rate": 3.669094745950008e-09,
"loss": 1.6558,
"step": 573
},
{
"epoch": 4.945017182130584,
"grad_norm": 0.24206612450984835,
"learning_rate": 2.6957489281997927e-09,
"loss": 1.6999,
"step": 574
},
{
"epoch": 4.953608247422681,
"grad_norm": 0.24884129420163512,
"learning_rate": 1.8720992731741104e-09,
"loss": 1.6552,
"step": 575
},
{
"epoch": 4.962199312714777,
"grad_norm": 0.23391618293615507,
"learning_rate": 1.1981704536129234e-09,
"loss": 1.6809,
"step": 576
},
{
"epoch": 4.970790378006873,
"grad_norm": 0.22674183113418866,
"learning_rate": 6.739826573121111e-10,
"loss": 1.6519,
"step": 577
},
{
"epoch": 4.979381443298969,
"grad_norm": 0.23743467756025552,
"learning_rate": 2.995515865183984e-10,
"loss": 1.6308,
"step": 578
},
{
"epoch": 4.987972508591065,
"grad_norm": 0.2322779280097835,
"learning_rate": 7.48884574575115e-11,
"loss": 1.591,
"step": 579
},
{
"epoch": 4.9965635738831615,
"grad_norm": 0.24133960284150027,
"learning_rate": 0.0,
"loss": 1.6497,
"step": 580
},
{
"epoch": 4.9965635738831615,
"step": 580,
"total_flos": 96622779432960.0,
"train_loss": 1.8076569943592466,
"train_runtime": 4607.1316,
"train_samples_per_second": 12.617,
"train_steps_per_second": 0.126
}
],
"logging_steps": 1,
"max_steps": 580,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 50,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 96622779432960.0,
"train_batch_size": 5,
"trial_name": null,
"trial_params": null
}