model_bcvtgwrf / checkpoint-2848 /trainer_state.json
ugaoo's picture
Upload folder using huggingface_hub
dd03f67 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 2848,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00035112359550561797,
"grad_norm": 2.8951518535614014,
"learning_rate": 5.0000000000000004e-08,
"loss": 1.0953,
"step": 1
},
{
"epoch": 0.0007022471910112359,
"grad_norm": 2.683199167251587,
"learning_rate": 1.0000000000000001e-07,
"loss": 1.0614,
"step": 2
},
{
"epoch": 0.001053370786516854,
"grad_norm": 2.72992205619812,
"learning_rate": 1.5000000000000002e-07,
"loss": 1.0674,
"step": 3
},
{
"epoch": 0.0014044943820224719,
"grad_norm": 2.6386663913726807,
"learning_rate": 2.0000000000000002e-07,
"loss": 1.058,
"step": 4
},
{
"epoch": 0.0017556179775280898,
"grad_norm": 2.672447443008423,
"learning_rate": 2.5000000000000004e-07,
"loss": 1.0523,
"step": 5
},
{
"epoch": 0.002106741573033708,
"grad_norm": 2.5482404232025146,
"learning_rate": 3.0000000000000004e-07,
"loss": 1.0454,
"step": 6
},
{
"epoch": 0.0024578651685393258,
"grad_norm": 2.651028633117676,
"learning_rate": 3.5000000000000004e-07,
"loss": 1.0646,
"step": 7
},
{
"epoch": 0.0028089887640449437,
"grad_norm": 2.76011323928833,
"learning_rate": 4.0000000000000003e-07,
"loss": 1.061,
"step": 8
},
{
"epoch": 0.0031601123595505617,
"grad_norm": 2.4421677589416504,
"learning_rate": 4.5000000000000003e-07,
"loss": 1.0827,
"step": 9
},
{
"epoch": 0.0035112359550561797,
"grad_norm": 2.460603713989258,
"learning_rate": 5.000000000000001e-07,
"loss": 1.0925,
"step": 10
},
{
"epoch": 0.0038623595505617976,
"grad_norm": 2.3482112884521484,
"learning_rate": 5.5e-07,
"loss": 1.0679,
"step": 11
},
{
"epoch": 0.004213483146067416,
"grad_norm": 2.4199249744415283,
"learning_rate": 6.000000000000001e-07,
"loss": 1.0685,
"step": 12
},
{
"epoch": 0.004564606741573034,
"grad_norm": 2.171464443206787,
"learning_rate": 6.5e-07,
"loss": 1.0455,
"step": 13
},
{
"epoch": 0.0049157303370786515,
"grad_norm": 2.212657928466797,
"learning_rate": 7.000000000000001e-07,
"loss": 1.0718,
"step": 14
},
{
"epoch": 0.0052668539325842695,
"grad_norm": 2.0458426475524902,
"learning_rate": 7.5e-07,
"loss": 1.0213,
"step": 15
},
{
"epoch": 0.0056179775280898875,
"grad_norm": 1.8622595071792603,
"learning_rate": 8.000000000000001e-07,
"loss": 1.0266,
"step": 16
},
{
"epoch": 0.0059691011235955055,
"grad_norm": 1.7484514713287354,
"learning_rate": 8.500000000000001e-07,
"loss": 1.0439,
"step": 17
},
{
"epoch": 0.006320224719101123,
"grad_norm": 1.626994252204895,
"learning_rate": 9.000000000000001e-07,
"loss": 1.0274,
"step": 18
},
{
"epoch": 0.006671348314606741,
"grad_norm": 1.462096095085144,
"learning_rate": 9.500000000000001e-07,
"loss": 1.036,
"step": 19
},
{
"epoch": 0.007022471910112359,
"grad_norm": 1.4871413707733154,
"learning_rate": 1.0000000000000002e-06,
"loss": 1.033,
"step": 20
},
{
"epoch": 0.007373595505617977,
"grad_norm": 1.4984582662582397,
"learning_rate": 1.0500000000000001e-06,
"loss": 1.0279,
"step": 21
},
{
"epoch": 0.007724719101123595,
"grad_norm": 1.5213954448699951,
"learning_rate": 1.1e-06,
"loss": 1.0468,
"step": 22
},
{
"epoch": 0.008075842696629214,
"grad_norm": 1.3945717811584473,
"learning_rate": 1.1500000000000002e-06,
"loss": 1.0252,
"step": 23
},
{
"epoch": 0.008426966292134831,
"grad_norm": 1.2632803916931152,
"learning_rate": 1.2000000000000002e-06,
"loss": 0.9637,
"step": 24
},
{
"epoch": 0.00877808988764045,
"grad_norm": 1.2537329196929932,
"learning_rate": 1.25e-06,
"loss": 1.0194,
"step": 25
},
{
"epoch": 0.009129213483146067,
"grad_norm": 1.287577748298645,
"learning_rate": 1.3e-06,
"loss": 0.9491,
"step": 26
},
{
"epoch": 0.009480337078651686,
"grad_norm": 1.1792410612106323,
"learning_rate": 1.3500000000000002e-06,
"loss": 1.0,
"step": 27
},
{
"epoch": 0.009831460674157303,
"grad_norm": 1.1308109760284424,
"learning_rate": 1.4000000000000001e-06,
"loss": 1.002,
"step": 28
},
{
"epoch": 0.010182584269662922,
"grad_norm": 1.1292983293533325,
"learning_rate": 1.45e-06,
"loss": 1.0114,
"step": 29
},
{
"epoch": 0.010533707865168539,
"grad_norm": 1.0696202516555786,
"learning_rate": 1.5e-06,
"loss": 0.9371,
"step": 30
},
{
"epoch": 0.010884831460674158,
"grad_norm": 1.0532927513122559,
"learning_rate": 1.5500000000000002e-06,
"loss": 0.9805,
"step": 31
},
{
"epoch": 0.011235955056179775,
"grad_norm": 1.039662480354309,
"learning_rate": 1.6000000000000001e-06,
"loss": 0.9591,
"step": 32
},
{
"epoch": 0.011587078651685394,
"grad_norm": 0.9709979891777039,
"learning_rate": 1.6500000000000003e-06,
"loss": 0.9599,
"step": 33
},
{
"epoch": 0.011938202247191011,
"grad_norm": 0.9826909303665161,
"learning_rate": 1.7000000000000002e-06,
"loss": 0.9492,
"step": 34
},
{
"epoch": 0.01228932584269663,
"grad_norm": 1.01341712474823,
"learning_rate": 1.75e-06,
"loss": 1.0066,
"step": 35
},
{
"epoch": 0.012640449438202247,
"grad_norm": 1.0003409385681152,
"learning_rate": 1.8000000000000001e-06,
"loss": 0.9641,
"step": 36
},
{
"epoch": 0.012991573033707866,
"grad_norm": 0.9853388667106628,
"learning_rate": 1.85e-06,
"loss": 0.9681,
"step": 37
},
{
"epoch": 0.013342696629213483,
"grad_norm": 0.9508251547813416,
"learning_rate": 1.9000000000000002e-06,
"loss": 0.9242,
"step": 38
},
{
"epoch": 0.013693820224719102,
"grad_norm": 0.9218304753303528,
"learning_rate": 1.9500000000000004e-06,
"loss": 0.8917,
"step": 39
},
{
"epoch": 0.014044943820224719,
"grad_norm": 0.9143016934394836,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.8715,
"step": 40
},
{
"epoch": 0.014396067415730338,
"grad_norm": 0.9491222500801086,
"learning_rate": 2.05e-06,
"loss": 0.9157,
"step": 41
},
{
"epoch": 0.014747191011235955,
"grad_norm": 0.9453691244125366,
"learning_rate": 2.1000000000000002e-06,
"loss": 0.9537,
"step": 42
},
{
"epoch": 0.015098314606741573,
"grad_norm": 0.9372655153274536,
"learning_rate": 2.15e-06,
"loss": 0.9288,
"step": 43
},
{
"epoch": 0.01544943820224719,
"grad_norm": 0.9183003306388855,
"learning_rate": 2.2e-06,
"loss": 0.8841,
"step": 44
},
{
"epoch": 0.01580056179775281,
"grad_norm": 0.924336314201355,
"learning_rate": 2.25e-06,
"loss": 0.9305,
"step": 45
},
{
"epoch": 0.016151685393258428,
"grad_norm": 0.9038196206092834,
"learning_rate": 2.3000000000000004e-06,
"loss": 0.9376,
"step": 46
},
{
"epoch": 0.016502808988764044,
"grad_norm": 0.9091538786888123,
"learning_rate": 2.35e-06,
"loss": 0.9243,
"step": 47
},
{
"epoch": 0.016853932584269662,
"grad_norm": 0.9212149381637573,
"learning_rate": 2.4000000000000003e-06,
"loss": 0.9515,
"step": 48
},
{
"epoch": 0.01720505617977528,
"grad_norm": 0.8974660038948059,
"learning_rate": 2.4500000000000003e-06,
"loss": 0.893,
"step": 49
},
{
"epoch": 0.0175561797752809,
"grad_norm": 0.8716650009155273,
"learning_rate": 2.5e-06,
"loss": 0.9485,
"step": 50
},
{
"epoch": 0.017907303370786515,
"grad_norm": 0.8729946613311768,
"learning_rate": 2.55e-06,
"loss": 0.912,
"step": 51
},
{
"epoch": 0.018258426966292134,
"grad_norm": 0.8486078977584839,
"learning_rate": 2.6e-06,
"loss": 0.9479,
"step": 52
},
{
"epoch": 0.018609550561797753,
"grad_norm": 0.8713150024414062,
"learning_rate": 2.6500000000000005e-06,
"loss": 0.9049,
"step": 53
},
{
"epoch": 0.018960674157303372,
"grad_norm": 0.855265200138092,
"learning_rate": 2.7000000000000004e-06,
"loss": 0.921,
"step": 54
},
{
"epoch": 0.019311797752808987,
"grad_norm": 0.8279822468757629,
"learning_rate": 2.7500000000000004e-06,
"loss": 0.8507,
"step": 55
},
{
"epoch": 0.019662921348314606,
"grad_norm": 0.8624402284622192,
"learning_rate": 2.8000000000000003e-06,
"loss": 0.9292,
"step": 56
},
{
"epoch": 0.020014044943820225,
"grad_norm": 0.8482896089553833,
"learning_rate": 2.85e-06,
"loss": 0.881,
"step": 57
},
{
"epoch": 0.020365168539325844,
"grad_norm": 0.8510951399803162,
"learning_rate": 2.9e-06,
"loss": 0.8702,
"step": 58
},
{
"epoch": 0.02071629213483146,
"grad_norm": 0.8771702647209167,
"learning_rate": 2.95e-06,
"loss": 0.9329,
"step": 59
},
{
"epoch": 0.021067415730337078,
"grad_norm": 0.8551446199417114,
"learning_rate": 3e-06,
"loss": 0.8908,
"step": 60
},
{
"epoch": 0.021418539325842697,
"grad_norm": 0.8884496688842773,
"learning_rate": 3.05e-06,
"loss": 0.8791,
"step": 61
},
{
"epoch": 0.021769662921348316,
"grad_norm": 0.8115338087081909,
"learning_rate": 3.1000000000000004e-06,
"loss": 0.8544,
"step": 62
},
{
"epoch": 0.02212078651685393,
"grad_norm": 0.8549612760543823,
"learning_rate": 3.1500000000000003e-06,
"loss": 0.8963,
"step": 63
},
{
"epoch": 0.02247191011235955,
"grad_norm": 0.8259643316268921,
"learning_rate": 3.2000000000000003e-06,
"loss": 0.8299,
"step": 64
},
{
"epoch": 0.02282303370786517,
"grad_norm": 0.8829634785652161,
"learning_rate": 3.2500000000000002e-06,
"loss": 0.8873,
"step": 65
},
{
"epoch": 0.023174157303370788,
"grad_norm": 0.8536106944084167,
"learning_rate": 3.3000000000000006e-06,
"loss": 0.9009,
"step": 66
},
{
"epoch": 0.023525280898876403,
"grad_norm": 0.8468846082687378,
"learning_rate": 3.3500000000000005e-06,
"loss": 0.8921,
"step": 67
},
{
"epoch": 0.023876404494382022,
"grad_norm": 0.8253934383392334,
"learning_rate": 3.4000000000000005e-06,
"loss": 0.848,
"step": 68
},
{
"epoch": 0.02422752808988764,
"grad_norm": 0.8778877854347229,
"learning_rate": 3.45e-06,
"loss": 0.8887,
"step": 69
},
{
"epoch": 0.02457865168539326,
"grad_norm": 0.8790880441665649,
"learning_rate": 3.5e-06,
"loss": 0.8849,
"step": 70
},
{
"epoch": 0.024929775280898875,
"grad_norm": 0.843599259853363,
"learning_rate": 3.5500000000000003e-06,
"loss": 0.8972,
"step": 71
},
{
"epoch": 0.025280898876404494,
"grad_norm": 0.8428463339805603,
"learning_rate": 3.6000000000000003e-06,
"loss": 0.9027,
"step": 72
},
{
"epoch": 0.025632022471910113,
"grad_norm": 0.8088714480400085,
"learning_rate": 3.65e-06,
"loss": 0.8562,
"step": 73
},
{
"epoch": 0.02598314606741573,
"grad_norm": 0.857280969619751,
"learning_rate": 3.7e-06,
"loss": 0.8725,
"step": 74
},
{
"epoch": 0.026334269662921347,
"grad_norm": 0.8321940898895264,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.8656,
"step": 75
},
{
"epoch": 0.026685393258426966,
"grad_norm": 0.8289700150489807,
"learning_rate": 3.8000000000000005e-06,
"loss": 0.8211,
"step": 76
},
{
"epoch": 0.027036516853932584,
"grad_norm": 0.8454659581184387,
"learning_rate": 3.85e-06,
"loss": 0.9003,
"step": 77
},
{
"epoch": 0.027387640449438203,
"grad_norm": 0.8450126647949219,
"learning_rate": 3.900000000000001e-06,
"loss": 0.8595,
"step": 78
},
{
"epoch": 0.02773876404494382,
"grad_norm": 0.8216248750686646,
"learning_rate": 3.95e-06,
"loss": 0.8503,
"step": 79
},
{
"epoch": 0.028089887640449437,
"grad_norm": 0.8196912407875061,
"learning_rate": 4.000000000000001e-06,
"loss": 0.8965,
"step": 80
},
{
"epoch": 0.028441011235955056,
"grad_norm": 0.8264586329460144,
"learning_rate": 4.05e-06,
"loss": 0.879,
"step": 81
},
{
"epoch": 0.028792134831460675,
"grad_norm": 0.8712279200553894,
"learning_rate": 4.1e-06,
"loss": 0.8825,
"step": 82
},
{
"epoch": 0.02914325842696629,
"grad_norm": 0.8609493970870972,
"learning_rate": 4.15e-06,
"loss": 0.8901,
"step": 83
},
{
"epoch": 0.02949438202247191,
"grad_norm": 0.8273038864135742,
"learning_rate": 4.2000000000000004e-06,
"loss": 0.8589,
"step": 84
},
{
"epoch": 0.029845505617977528,
"grad_norm": 0.8164820075035095,
"learning_rate": 4.25e-06,
"loss": 0.8099,
"step": 85
},
{
"epoch": 0.030196629213483147,
"grad_norm": 0.8666368722915649,
"learning_rate": 4.3e-06,
"loss": 0.8535,
"step": 86
},
{
"epoch": 0.030547752808988762,
"grad_norm": 0.8173773288726807,
"learning_rate": 4.350000000000001e-06,
"loss": 0.806,
"step": 87
},
{
"epoch": 0.03089887640449438,
"grad_norm": 0.8484762907028198,
"learning_rate": 4.4e-06,
"loss": 0.8678,
"step": 88
},
{
"epoch": 0.03125,
"grad_norm": 0.8572366833686829,
"learning_rate": 4.450000000000001e-06,
"loss": 0.8939,
"step": 89
},
{
"epoch": 0.03160112359550562,
"grad_norm": 0.8265636563301086,
"learning_rate": 4.5e-06,
"loss": 0.8373,
"step": 90
},
{
"epoch": 0.03195224719101124,
"grad_norm": 0.8756334185600281,
"learning_rate": 4.5500000000000005e-06,
"loss": 0.8322,
"step": 91
},
{
"epoch": 0.032303370786516857,
"grad_norm": 0.8236197829246521,
"learning_rate": 4.600000000000001e-06,
"loss": 0.8425,
"step": 92
},
{
"epoch": 0.032654494382022475,
"grad_norm": 0.855594277381897,
"learning_rate": 4.65e-06,
"loss": 0.8779,
"step": 93
},
{
"epoch": 0.03300561797752809,
"grad_norm": 0.8004137873649597,
"learning_rate": 4.7e-06,
"loss": 0.8078,
"step": 94
},
{
"epoch": 0.033356741573033706,
"grad_norm": 0.7953411340713501,
"learning_rate": 4.75e-06,
"loss": 0.8363,
"step": 95
},
{
"epoch": 0.033707865168539325,
"grad_norm": 0.8278580904006958,
"learning_rate": 4.800000000000001e-06,
"loss": 0.8301,
"step": 96
},
{
"epoch": 0.034058988764044944,
"grad_norm": 0.8473982810974121,
"learning_rate": 4.85e-06,
"loss": 0.8408,
"step": 97
},
{
"epoch": 0.03441011235955056,
"grad_norm": 0.8142527341842651,
"learning_rate": 4.9000000000000005e-06,
"loss": 0.8295,
"step": 98
},
{
"epoch": 0.03476123595505618,
"grad_norm": 0.8610883355140686,
"learning_rate": 4.95e-06,
"loss": 0.8405,
"step": 99
},
{
"epoch": 0.0351123595505618,
"grad_norm": 0.8903759717941284,
"learning_rate": 5e-06,
"loss": 0.8793,
"step": 100
},
{
"epoch": 0.03546348314606742,
"grad_norm": 0.8775374293327332,
"learning_rate": 4.999999957251071e-06,
"loss": 0.8516,
"step": 101
},
{
"epoch": 0.03581460674157303,
"grad_norm": 0.8515768647193909,
"learning_rate": 4.999999829004281e-06,
"loss": 0.8045,
"step": 102
},
{
"epoch": 0.03616573033707865,
"grad_norm": 1.0148895978927612,
"learning_rate": 4.9999996152596355e-06,
"loss": 0.8247,
"step": 103
},
{
"epoch": 0.03651685393258427,
"grad_norm": 0.829903244972229,
"learning_rate": 4.999999316017144e-06,
"loss": 0.8282,
"step": 104
},
{
"epoch": 0.03686797752808989,
"grad_norm": 0.8467276096343994,
"learning_rate": 4.999998931276815e-06,
"loss": 0.8325,
"step": 105
},
{
"epoch": 0.037219101123595506,
"grad_norm": 0.8355855941772461,
"learning_rate": 4.999998461038661e-06,
"loss": 0.813,
"step": 106
},
{
"epoch": 0.037570224719101125,
"grad_norm": 0.8558890223503113,
"learning_rate": 4.999997905302699e-06,
"loss": 0.8592,
"step": 107
},
{
"epoch": 0.037921348314606744,
"grad_norm": 0.9446994662284851,
"learning_rate": 4.999997264068948e-06,
"loss": 0.8323,
"step": 108
},
{
"epoch": 0.03827247191011236,
"grad_norm": 0.8493000268936157,
"learning_rate": 4.99999653733743e-06,
"loss": 0.8178,
"step": 109
},
{
"epoch": 0.038623595505617975,
"grad_norm": 0.8132368922233582,
"learning_rate": 4.99999572510817e-06,
"loss": 0.7977,
"step": 110
},
{
"epoch": 0.038974719101123594,
"grad_norm": 0.8313632607460022,
"learning_rate": 4.999994827381196e-06,
"loss": 0.8407,
"step": 111
},
{
"epoch": 0.03932584269662921,
"grad_norm": 0.8673725724220276,
"learning_rate": 4.999993844156537e-06,
"loss": 0.862,
"step": 112
},
{
"epoch": 0.03967696629213483,
"grad_norm": 0.8794373273849487,
"learning_rate": 4.999992775434229e-06,
"loss": 0.8599,
"step": 113
},
{
"epoch": 0.04002808988764045,
"grad_norm": 0.8612070083618164,
"learning_rate": 4.9999916212143065e-06,
"loss": 0.8013,
"step": 114
},
{
"epoch": 0.04037921348314607,
"grad_norm": 0.8925568461418152,
"learning_rate": 4.9999903814968095e-06,
"loss": 0.7529,
"step": 115
},
{
"epoch": 0.04073033707865169,
"grad_norm": 0.8791881799697876,
"learning_rate": 4.999989056281781e-06,
"loss": 0.8756,
"step": 116
},
{
"epoch": 0.04108146067415731,
"grad_norm": 0.8983411192893982,
"learning_rate": 4.999987645569267e-06,
"loss": 0.8329,
"step": 117
},
{
"epoch": 0.04143258426966292,
"grad_norm": 0.8358251452445984,
"learning_rate": 4.999986149359314e-06,
"loss": 0.7792,
"step": 118
},
{
"epoch": 0.04178370786516854,
"grad_norm": 0.8848925828933716,
"learning_rate": 4.999984567651974e-06,
"loss": 0.8242,
"step": 119
},
{
"epoch": 0.042134831460674156,
"grad_norm": 0.9365871548652649,
"learning_rate": 4.999982900447301e-06,
"loss": 0.7875,
"step": 120
},
{
"epoch": 0.042485955056179775,
"grad_norm": 0.8330186605453491,
"learning_rate": 4.999981147745352e-06,
"loss": 0.8387,
"step": 121
},
{
"epoch": 0.042837078651685394,
"grad_norm": 0.8438044190406799,
"learning_rate": 4.999979309546187e-06,
"loss": 0.8547,
"step": 122
},
{
"epoch": 0.04318820224719101,
"grad_norm": 0.8812524676322937,
"learning_rate": 4.999977385849869e-06,
"loss": 0.7891,
"step": 123
},
{
"epoch": 0.04353932584269663,
"grad_norm": 0.8498764634132385,
"learning_rate": 4.999975376656464e-06,
"loss": 0.8259,
"step": 124
},
{
"epoch": 0.04389044943820225,
"grad_norm": 0.8646953105926514,
"learning_rate": 4.99997328196604e-06,
"loss": 0.8269,
"step": 125
},
{
"epoch": 0.04424157303370786,
"grad_norm": 0.8175535798072815,
"learning_rate": 4.9999711017786686e-06,
"loss": 0.7943,
"step": 126
},
{
"epoch": 0.04459269662921348,
"grad_norm": 0.8761390447616577,
"learning_rate": 4.999968836094425e-06,
"loss": 0.8668,
"step": 127
},
{
"epoch": 0.0449438202247191,
"grad_norm": 0.8925103545188904,
"learning_rate": 4.999966484913387e-06,
"loss": 0.8562,
"step": 128
},
{
"epoch": 0.04529494382022472,
"grad_norm": 0.8750078678131104,
"learning_rate": 4.999964048235634e-06,
"loss": 0.8058,
"step": 129
},
{
"epoch": 0.04564606741573034,
"grad_norm": 0.8446925282478333,
"learning_rate": 4.999961526061251e-06,
"loss": 0.8197,
"step": 130
},
{
"epoch": 0.045997191011235956,
"grad_norm": 0.8947305679321289,
"learning_rate": 4.999958918390321e-06,
"loss": 0.8347,
"step": 131
},
{
"epoch": 0.046348314606741575,
"grad_norm": 0.8591403961181641,
"learning_rate": 4.999956225222937e-06,
"loss": 0.8117,
"step": 132
},
{
"epoch": 0.046699438202247194,
"grad_norm": 0.8698204755783081,
"learning_rate": 4.999953446559188e-06,
"loss": 0.8378,
"step": 133
},
{
"epoch": 0.047050561797752806,
"grad_norm": 0.8451970219612122,
"learning_rate": 4.999950582399171e-06,
"loss": 0.8324,
"step": 134
},
{
"epoch": 0.047401685393258425,
"grad_norm": 0.8550758957862854,
"learning_rate": 4.999947632742984e-06,
"loss": 0.8291,
"step": 135
},
{
"epoch": 0.047752808988764044,
"grad_norm": 0.8345015645027161,
"learning_rate": 4.999944597590726e-06,
"loss": 0.8315,
"step": 136
},
{
"epoch": 0.04810393258426966,
"grad_norm": 0.9103791117668152,
"learning_rate": 4.999941476942502e-06,
"loss": 0.8009,
"step": 137
},
{
"epoch": 0.04845505617977528,
"grad_norm": 0.9898645877838135,
"learning_rate": 4.99993827079842e-06,
"loss": 0.8574,
"step": 138
},
{
"epoch": 0.0488061797752809,
"grad_norm": 0.8579599261283875,
"learning_rate": 4.9999349791585876e-06,
"loss": 0.7747,
"step": 139
},
{
"epoch": 0.04915730337078652,
"grad_norm": 0.8869126439094543,
"learning_rate": 4.999931602023118e-06,
"loss": 0.8472,
"step": 140
},
{
"epoch": 0.04950842696629214,
"grad_norm": 0.9130380749702454,
"learning_rate": 4.999928139392127e-06,
"loss": 0.8247,
"step": 141
},
{
"epoch": 0.04985955056179775,
"grad_norm": 0.8806836605072021,
"learning_rate": 4.999924591265732e-06,
"loss": 0.8658,
"step": 142
},
{
"epoch": 0.05021067415730337,
"grad_norm": 0.9065203666687012,
"learning_rate": 4.999920957644057e-06,
"loss": 0.8097,
"step": 143
},
{
"epoch": 0.05056179775280899,
"grad_norm": 0.8982106447219849,
"learning_rate": 4.999917238527223e-06,
"loss": 0.8207,
"step": 144
},
{
"epoch": 0.050912921348314606,
"grad_norm": 0.9152344465255737,
"learning_rate": 4.999913433915359e-06,
"loss": 0.845,
"step": 145
},
{
"epoch": 0.051264044943820225,
"grad_norm": 0.8733872771263123,
"learning_rate": 4.999909543808595e-06,
"loss": 0.8337,
"step": 146
},
{
"epoch": 0.051615168539325844,
"grad_norm": 0.8482681512832642,
"learning_rate": 4.999905568207064e-06,
"loss": 0.8226,
"step": 147
},
{
"epoch": 0.05196629213483146,
"grad_norm": 0.8690842986106873,
"learning_rate": 4.999901507110901e-06,
"loss": 0.8677,
"step": 148
},
{
"epoch": 0.05231741573033708,
"grad_norm": 0.8329693675041199,
"learning_rate": 4.999897360520246e-06,
"loss": 0.7628,
"step": 149
},
{
"epoch": 0.05266853932584269,
"grad_norm": 0.913270890712738,
"learning_rate": 4.999893128435241e-06,
"loss": 0.8374,
"step": 150
},
{
"epoch": 0.05301966292134831,
"grad_norm": 0.843010425567627,
"learning_rate": 4.999888810856029e-06,
"loss": 0.8343,
"step": 151
},
{
"epoch": 0.05337078651685393,
"grad_norm": 0.8601850867271423,
"learning_rate": 4.99988440778276e-06,
"loss": 0.8254,
"step": 152
},
{
"epoch": 0.05372191011235955,
"grad_norm": 0.8458811640739441,
"learning_rate": 4.999879919215583e-06,
"loss": 0.8238,
"step": 153
},
{
"epoch": 0.05407303370786517,
"grad_norm": 0.8649472594261169,
"learning_rate": 4.999875345154652e-06,
"loss": 0.8545,
"step": 154
},
{
"epoch": 0.05442415730337079,
"grad_norm": 0.8692646622657776,
"learning_rate": 4.999870685600123e-06,
"loss": 0.769,
"step": 155
},
{
"epoch": 0.054775280898876406,
"grad_norm": 0.8382206559181213,
"learning_rate": 4.999865940552157e-06,
"loss": 0.8076,
"step": 156
},
{
"epoch": 0.055126404494382025,
"grad_norm": 0.8909122943878174,
"learning_rate": 4.999861110010914e-06,
"loss": 0.8118,
"step": 157
},
{
"epoch": 0.05547752808988764,
"grad_norm": 0.8760329484939575,
"learning_rate": 4.99985619397656e-06,
"loss": 0.8625,
"step": 158
},
{
"epoch": 0.055828651685393256,
"grad_norm": 1.0767908096313477,
"learning_rate": 4.999851192449263e-06,
"loss": 0.8038,
"step": 159
},
{
"epoch": 0.056179775280898875,
"grad_norm": 0.8571433424949646,
"learning_rate": 4.999846105429196e-06,
"loss": 0.8123,
"step": 160
},
{
"epoch": 0.056530898876404494,
"grad_norm": 0.8204314112663269,
"learning_rate": 4.99984093291653e-06,
"loss": 0.7713,
"step": 161
},
{
"epoch": 0.05688202247191011,
"grad_norm": 0.8622918128967285,
"learning_rate": 4.9998356749114434e-06,
"loss": 0.7816,
"step": 162
},
{
"epoch": 0.05723314606741573,
"grad_norm": 0.991470456123352,
"learning_rate": 4.999830331414116e-06,
"loss": 0.8171,
"step": 163
},
{
"epoch": 0.05758426966292135,
"grad_norm": 0.9167494177818298,
"learning_rate": 4.99982490242473e-06,
"loss": 0.8096,
"step": 164
},
{
"epoch": 0.05793539325842697,
"grad_norm": 0.8794699311256409,
"learning_rate": 4.999819387943472e-06,
"loss": 0.8284,
"step": 165
},
{
"epoch": 0.05828651685393258,
"grad_norm": 0.9374570250511169,
"learning_rate": 4.99981378797053e-06,
"loss": 0.7887,
"step": 166
},
{
"epoch": 0.0586376404494382,
"grad_norm": 0.8803961277008057,
"learning_rate": 4.999808102506095e-06,
"loss": 0.8331,
"step": 167
},
{
"epoch": 0.05898876404494382,
"grad_norm": 0.903806209564209,
"learning_rate": 4.999802331550363e-06,
"loss": 0.8002,
"step": 168
},
{
"epoch": 0.05933988764044944,
"grad_norm": 0.8599410057067871,
"learning_rate": 4.99979647510353e-06,
"loss": 0.8342,
"step": 169
},
{
"epoch": 0.059691011235955056,
"grad_norm": 0.9048354625701904,
"learning_rate": 4.999790533165797e-06,
"loss": 0.8197,
"step": 170
},
{
"epoch": 0.060042134831460675,
"grad_norm": 0.8594435453414917,
"learning_rate": 4.999784505737366e-06,
"loss": 0.7994,
"step": 171
},
{
"epoch": 0.060393258426966294,
"grad_norm": 0.8999492526054382,
"learning_rate": 4.999778392818444e-06,
"loss": 0.8062,
"step": 172
},
{
"epoch": 0.06074438202247191,
"grad_norm": 0.8730378150939941,
"learning_rate": 4.999772194409241e-06,
"loss": 0.8348,
"step": 173
},
{
"epoch": 0.061095505617977525,
"grad_norm": 0.8853862881660461,
"learning_rate": 4.999765910509967e-06,
"loss": 0.8235,
"step": 174
},
{
"epoch": 0.061446629213483143,
"grad_norm": 0.8512648344039917,
"learning_rate": 4.999759541120839e-06,
"loss": 0.8254,
"step": 175
},
{
"epoch": 0.06179775280898876,
"grad_norm": 0.925890326499939,
"learning_rate": 4.999753086242073e-06,
"loss": 0.8498,
"step": 176
},
{
"epoch": 0.06214887640449438,
"grad_norm": 0.9359338283538818,
"learning_rate": 4.999746545873891e-06,
"loss": 0.8006,
"step": 177
},
{
"epoch": 0.0625,
"grad_norm": 0.8511932492256165,
"learning_rate": 4.999739920016516e-06,
"loss": 0.7843,
"step": 178
},
{
"epoch": 0.06285112359550561,
"grad_norm": 0.920088529586792,
"learning_rate": 4.999733208670174e-06,
"loss": 0.8173,
"step": 179
},
{
"epoch": 0.06320224719101124,
"grad_norm": 0.8991772532463074,
"learning_rate": 4.9997264118350965e-06,
"loss": 0.8227,
"step": 180
},
{
"epoch": 0.06355337078651685,
"grad_norm": 0.9123396873474121,
"learning_rate": 4.9997195295115145e-06,
"loss": 0.7767,
"step": 181
},
{
"epoch": 0.06390449438202248,
"grad_norm": 0.888533353805542,
"learning_rate": 4.9997125616996635e-06,
"loss": 0.8093,
"step": 182
},
{
"epoch": 0.06425561797752809,
"grad_norm": 0.9099261164665222,
"learning_rate": 4.999705508399782e-06,
"loss": 0.8142,
"step": 183
},
{
"epoch": 0.06460674157303371,
"grad_norm": 0.9250622987747192,
"learning_rate": 4.999698369612111e-06,
"loss": 0.7913,
"step": 184
},
{
"epoch": 0.06495786516853932,
"grad_norm": 0.8754624128341675,
"learning_rate": 4.999691145336895e-06,
"loss": 0.7755,
"step": 185
},
{
"epoch": 0.06530898876404495,
"grad_norm": 0.8488731384277344,
"learning_rate": 4.99968383557438e-06,
"loss": 0.8209,
"step": 186
},
{
"epoch": 0.06566011235955056,
"grad_norm": 0.8827488422393799,
"learning_rate": 4.9996764403248175e-06,
"loss": 0.8121,
"step": 187
},
{
"epoch": 0.06601123595505617,
"grad_norm": 0.8604486584663391,
"learning_rate": 4.99966895958846e-06,
"loss": 0.8046,
"step": 188
},
{
"epoch": 0.0663623595505618,
"grad_norm": 0.855309247970581,
"learning_rate": 4.999661393365562e-06,
"loss": 0.7345,
"step": 189
},
{
"epoch": 0.06671348314606741,
"grad_norm": 0.8943816423416138,
"learning_rate": 4.999653741656385e-06,
"loss": 0.7937,
"step": 190
},
{
"epoch": 0.06706460674157304,
"grad_norm": 0.9106631278991699,
"learning_rate": 4.999646004461188e-06,
"loss": 0.8157,
"step": 191
},
{
"epoch": 0.06741573033707865,
"grad_norm": 0.9037092924118042,
"learning_rate": 4.999638181780237e-06,
"loss": 0.8091,
"step": 192
},
{
"epoch": 0.06776685393258428,
"grad_norm": 0.854783833026886,
"learning_rate": 4.9996302736137994e-06,
"loss": 0.7784,
"step": 193
},
{
"epoch": 0.06811797752808989,
"grad_norm": 0.8792631030082703,
"learning_rate": 4.999622279962144e-06,
"loss": 0.7737,
"step": 194
},
{
"epoch": 0.0684691011235955,
"grad_norm": 0.8999345898628235,
"learning_rate": 4.9996142008255465e-06,
"loss": 0.7858,
"step": 195
},
{
"epoch": 0.06882022471910113,
"grad_norm": 0.8954832553863525,
"learning_rate": 4.999606036204282e-06,
"loss": 0.8045,
"step": 196
},
{
"epoch": 0.06917134831460674,
"grad_norm": 0.8854857087135315,
"learning_rate": 4.999597786098631e-06,
"loss": 0.7842,
"step": 197
},
{
"epoch": 0.06952247191011236,
"grad_norm": 0.8617668747901917,
"learning_rate": 4.999589450508874e-06,
"loss": 0.807,
"step": 198
},
{
"epoch": 0.06987359550561797,
"grad_norm": 0.8737844824790955,
"learning_rate": 4.999581029435296e-06,
"loss": 0.8268,
"step": 199
},
{
"epoch": 0.0702247191011236,
"grad_norm": 0.8421310782432556,
"learning_rate": 4.999572522878186e-06,
"loss": 0.7701,
"step": 200
},
{
"epoch": 0.07057584269662921,
"grad_norm": 0.8911473155021667,
"learning_rate": 4.9995639308378365e-06,
"loss": 0.756,
"step": 201
},
{
"epoch": 0.07092696629213484,
"grad_norm": 0.879984974861145,
"learning_rate": 4.999555253314538e-06,
"loss": 0.8023,
"step": 202
},
{
"epoch": 0.07127808988764045,
"grad_norm": 0.9726935625076294,
"learning_rate": 4.9995464903085885e-06,
"loss": 0.8017,
"step": 203
},
{
"epoch": 0.07162921348314606,
"grad_norm": 0.9257283210754395,
"learning_rate": 4.999537641820288e-06,
"loss": 0.8336,
"step": 204
},
{
"epoch": 0.07198033707865169,
"grad_norm": 0.9612401723861694,
"learning_rate": 4.9995287078499385e-06,
"loss": 0.8014,
"step": 205
},
{
"epoch": 0.0723314606741573,
"grad_norm": 0.8790804743766785,
"learning_rate": 4.999519688397847e-06,
"loss": 0.7592,
"step": 206
},
{
"epoch": 0.07268258426966293,
"grad_norm": 0.9160227179527283,
"learning_rate": 4.999510583464321e-06,
"loss": 0.7905,
"step": 207
},
{
"epoch": 0.07303370786516854,
"grad_norm": 0.9160066246986389,
"learning_rate": 4.999501393049671e-06,
"loss": 0.8382,
"step": 208
},
{
"epoch": 0.07338483146067416,
"grad_norm": 0.8971685767173767,
"learning_rate": 4.999492117154214e-06,
"loss": 0.7881,
"step": 209
},
{
"epoch": 0.07373595505617977,
"grad_norm": 0.9171693921089172,
"learning_rate": 4.9994827557782644e-06,
"loss": 0.8101,
"step": 210
},
{
"epoch": 0.07408707865168539,
"grad_norm": 0.8446071743965149,
"learning_rate": 4.999473308922143e-06,
"loss": 0.7749,
"step": 211
},
{
"epoch": 0.07443820224719101,
"grad_norm": 0.8867083787918091,
"learning_rate": 4.999463776586174e-06,
"loss": 0.7656,
"step": 212
},
{
"epoch": 0.07478932584269662,
"grad_norm": 0.8746408820152283,
"learning_rate": 4.9994541587706815e-06,
"loss": 0.7778,
"step": 213
},
{
"epoch": 0.07514044943820225,
"grad_norm": 0.884986162185669,
"learning_rate": 4.999444455475997e-06,
"loss": 0.7975,
"step": 214
},
{
"epoch": 0.07549157303370786,
"grad_norm": 0.8919047713279724,
"learning_rate": 4.9994346667024505e-06,
"loss": 0.7552,
"step": 215
},
{
"epoch": 0.07584269662921349,
"grad_norm": 0.8840435147285461,
"learning_rate": 4.999424792450377e-06,
"loss": 0.7876,
"step": 216
},
{
"epoch": 0.0761938202247191,
"grad_norm": 0.9024021625518799,
"learning_rate": 4.999414832720114e-06,
"loss": 0.7921,
"step": 217
},
{
"epoch": 0.07654494382022473,
"grad_norm": 0.8887689113616943,
"learning_rate": 4.999404787512003e-06,
"loss": 0.7843,
"step": 218
},
{
"epoch": 0.07689606741573034,
"grad_norm": 0.903285026550293,
"learning_rate": 4.9993946568263866e-06,
"loss": 0.8164,
"step": 219
},
{
"epoch": 0.07724719101123595,
"grad_norm": 0.8626529574394226,
"learning_rate": 4.9993844406636124e-06,
"loss": 0.7941,
"step": 220
},
{
"epoch": 0.07759831460674158,
"grad_norm": 0.8919528126716614,
"learning_rate": 4.999374139024028e-06,
"loss": 0.8054,
"step": 221
},
{
"epoch": 0.07794943820224719,
"grad_norm": 0.9159940481185913,
"learning_rate": 4.999363751907988e-06,
"loss": 0.8071,
"step": 222
},
{
"epoch": 0.07830056179775281,
"grad_norm": 0.8792368173599243,
"learning_rate": 4.999353279315846e-06,
"loss": 0.7949,
"step": 223
},
{
"epoch": 0.07865168539325842,
"grad_norm": 0.9184228181838989,
"learning_rate": 4.9993427212479604e-06,
"loss": 0.8013,
"step": 224
},
{
"epoch": 0.07900280898876405,
"grad_norm": 0.8925946354866028,
"learning_rate": 4.999332077704692e-06,
"loss": 0.813,
"step": 225
},
{
"epoch": 0.07935393258426966,
"grad_norm": 0.9273242354393005,
"learning_rate": 4.999321348686406e-06,
"loss": 0.8107,
"step": 226
},
{
"epoch": 0.07970505617977527,
"grad_norm": 0.8914328813552856,
"learning_rate": 4.999310534193468e-06,
"loss": 0.8229,
"step": 227
},
{
"epoch": 0.0800561797752809,
"grad_norm": 0.9217993021011353,
"learning_rate": 4.999299634226249e-06,
"loss": 0.7722,
"step": 228
},
{
"epoch": 0.08040730337078651,
"grad_norm": 0.9615241289138794,
"learning_rate": 4.999288648785121e-06,
"loss": 0.8105,
"step": 229
},
{
"epoch": 0.08075842696629214,
"grad_norm": 0.9098490476608276,
"learning_rate": 4.99927757787046e-06,
"loss": 0.814,
"step": 230
},
{
"epoch": 0.08110955056179775,
"grad_norm": 0.9211421012878418,
"learning_rate": 4.999266421482645e-06,
"loss": 0.7542,
"step": 231
},
{
"epoch": 0.08146067415730338,
"grad_norm": 0.841416597366333,
"learning_rate": 4.999255179622056e-06,
"loss": 0.7961,
"step": 232
},
{
"epoch": 0.08181179775280899,
"grad_norm": 0.8942486643791199,
"learning_rate": 4.9992438522890785e-06,
"loss": 0.8095,
"step": 233
},
{
"epoch": 0.08216292134831461,
"grad_norm": 0.8535089492797852,
"learning_rate": 4.9992324394841005e-06,
"loss": 0.801,
"step": 234
},
{
"epoch": 0.08251404494382023,
"grad_norm": 1.02827787399292,
"learning_rate": 4.999220941207511e-06,
"loss": 0.8057,
"step": 235
},
{
"epoch": 0.08286516853932584,
"grad_norm": 0.9189361929893494,
"learning_rate": 4.9992093574597046e-06,
"loss": 0.7682,
"step": 236
},
{
"epoch": 0.08321629213483146,
"grad_norm": 0.9160380959510803,
"learning_rate": 4.999197688241076e-06,
"loss": 0.8218,
"step": 237
},
{
"epoch": 0.08356741573033707,
"grad_norm": 0.8860989809036255,
"learning_rate": 4.999185933552027e-06,
"loss": 0.7907,
"step": 238
},
{
"epoch": 0.0839185393258427,
"grad_norm": 0.8413179516792297,
"learning_rate": 4.999174093392956e-06,
"loss": 0.7711,
"step": 239
},
{
"epoch": 0.08426966292134831,
"grad_norm": 0.8843936324119568,
"learning_rate": 4.99916216776427e-06,
"loss": 0.7356,
"step": 240
},
{
"epoch": 0.08462078651685394,
"grad_norm": 0.9339604377746582,
"learning_rate": 4.999150156666376e-06,
"loss": 0.8017,
"step": 241
},
{
"epoch": 0.08497191011235955,
"grad_norm": 0.8932393789291382,
"learning_rate": 4.999138060099685e-06,
"loss": 0.7773,
"step": 242
},
{
"epoch": 0.08532303370786516,
"grad_norm": 0.9117473363876343,
"learning_rate": 4.999125878064611e-06,
"loss": 0.7828,
"step": 243
},
{
"epoch": 0.08567415730337079,
"grad_norm": 0.8696773052215576,
"learning_rate": 4.999113610561571e-06,
"loss": 0.7681,
"step": 244
},
{
"epoch": 0.0860252808988764,
"grad_norm": 0.9062797427177429,
"learning_rate": 4.999101257590985e-06,
"loss": 0.7766,
"step": 245
},
{
"epoch": 0.08637640449438203,
"grad_norm": 0.893018364906311,
"learning_rate": 4.999088819153273e-06,
"loss": 0.735,
"step": 246
},
{
"epoch": 0.08672752808988764,
"grad_norm": 0.9077386856079102,
"learning_rate": 4.999076295248863e-06,
"loss": 0.8092,
"step": 247
},
{
"epoch": 0.08707865168539326,
"grad_norm": 0.9474624395370483,
"learning_rate": 4.9990636858781805e-06,
"loss": 0.7841,
"step": 248
},
{
"epoch": 0.08742977528089887,
"grad_norm": 0.8823204040527344,
"learning_rate": 4.9990509910416595e-06,
"loss": 0.7827,
"step": 249
},
{
"epoch": 0.0877808988764045,
"grad_norm": 0.9361265897750854,
"learning_rate": 4.999038210739733e-06,
"loss": 0.8274,
"step": 250
},
{
"epoch": 0.08813202247191011,
"grad_norm": 0.9128723740577698,
"learning_rate": 4.999025344972838e-06,
"loss": 0.7681,
"step": 251
},
{
"epoch": 0.08848314606741572,
"grad_norm": 0.8730833530426025,
"learning_rate": 4.999012393741415e-06,
"loss": 0.8086,
"step": 252
},
{
"epoch": 0.08883426966292135,
"grad_norm": 0.8662939071655273,
"learning_rate": 4.998999357045906e-06,
"loss": 0.7587,
"step": 253
},
{
"epoch": 0.08918539325842696,
"grad_norm": 0.8748407959938049,
"learning_rate": 4.998986234886758e-06,
"loss": 0.7937,
"step": 254
},
{
"epoch": 0.08953651685393259,
"grad_norm": 0.9770292639732361,
"learning_rate": 4.998973027264419e-06,
"loss": 0.7815,
"step": 255
},
{
"epoch": 0.0898876404494382,
"grad_norm": 0.8764484524726868,
"learning_rate": 4.99895973417934e-06,
"loss": 0.766,
"step": 256
},
{
"epoch": 0.09023876404494383,
"grad_norm": 0.8940339684486389,
"learning_rate": 4.998946355631978e-06,
"loss": 0.8166,
"step": 257
},
{
"epoch": 0.09058988764044944,
"grad_norm": 0.903728723526001,
"learning_rate": 4.998932891622788e-06,
"loss": 0.7801,
"step": 258
},
{
"epoch": 0.09094101123595505,
"grad_norm": 0.9147090315818787,
"learning_rate": 4.998919342152232e-06,
"loss": 0.814,
"step": 259
},
{
"epoch": 0.09129213483146068,
"grad_norm": 0.9121170043945312,
"learning_rate": 4.9989057072207725e-06,
"loss": 0.7718,
"step": 260
},
{
"epoch": 0.09164325842696629,
"grad_norm": 0.8610017895698547,
"learning_rate": 4.998891986828877e-06,
"loss": 0.7539,
"step": 261
},
{
"epoch": 0.09199438202247191,
"grad_norm": 0.9325037598609924,
"learning_rate": 4.9988781809770124e-06,
"loss": 0.7699,
"step": 262
},
{
"epoch": 0.09234550561797752,
"grad_norm": 0.92354416847229,
"learning_rate": 4.998864289665654e-06,
"loss": 0.8224,
"step": 263
},
{
"epoch": 0.09269662921348315,
"grad_norm": 0.9121785759925842,
"learning_rate": 4.998850312895274e-06,
"loss": 0.7881,
"step": 264
},
{
"epoch": 0.09304775280898876,
"grad_norm": 0.8718385696411133,
"learning_rate": 4.998836250666352e-06,
"loss": 0.7865,
"step": 265
},
{
"epoch": 0.09339887640449439,
"grad_norm": 0.8944951295852661,
"learning_rate": 4.9988221029793685e-06,
"loss": 0.8315,
"step": 266
},
{
"epoch": 0.09375,
"grad_norm": 0.9362013339996338,
"learning_rate": 4.998807869834807e-06,
"loss": 0.7572,
"step": 267
},
{
"epoch": 0.09410112359550561,
"grad_norm": 0.8949217796325684,
"learning_rate": 4.998793551233155e-06,
"loss": 0.7614,
"step": 268
},
{
"epoch": 0.09445224719101124,
"grad_norm": 0.9898912310600281,
"learning_rate": 4.9987791471749015e-06,
"loss": 0.8111,
"step": 269
},
{
"epoch": 0.09480337078651685,
"grad_norm": 0.8814307451248169,
"learning_rate": 4.9987646576605395e-06,
"loss": 0.7774,
"step": 270
},
{
"epoch": 0.09515449438202248,
"grad_norm": 0.9214469790458679,
"learning_rate": 4.998750082690564e-06,
"loss": 0.7944,
"step": 271
},
{
"epoch": 0.09550561797752809,
"grad_norm": 0.8990268707275391,
"learning_rate": 4.998735422265475e-06,
"loss": 0.7661,
"step": 272
},
{
"epoch": 0.09585674157303371,
"grad_norm": 0.8504504561424255,
"learning_rate": 4.998720676385772e-06,
"loss": 0.7643,
"step": 273
},
{
"epoch": 0.09620786516853932,
"grad_norm": 0.9197403788566589,
"learning_rate": 4.99870584505196e-06,
"loss": 0.7763,
"step": 274
},
{
"epoch": 0.09655898876404495,
"grad_norm": 0.9017580151557922,
"learning_rate": 4.998690928264547e-06,
"loss": 0.8137,
"step": 275
},
{
"epoch": 0.09691011235955056,
"grad_norm": 0.880264937877655,
"learning_rate": 4.9986759260240416e-06,
"loss": 0.8088,
"step": 276
},
{
"epoch": 0.09726123595505617,
"grad_norm": 0.8751420378684998,
"learning_rate": 4.998660838330958e-06,
"loss": 0.7802,
"step": 277
},
{
"epoch": 0.0976123595505618,
"grad_norm": 0.8619885444641113,
"learning_rate": 4.998645665185812e-06,
"loss": 0.7584,
"step": 278
},
{
"epoch": 0.09796348314606741,
"grad_norm": 0.8772746324539185,
"learning_rate": 4.998630406589122e-06,
"loss": 0.7739,
"step": 279
},
{
"epoch": 0.09831460674157304,
"grad_norm": 0.9019191265106201,
"learning_rate": 4.998615062541411e-06,
"loss": 0.806,
"step": 280
},
{
"epoch": 0.09866573033707865,
"grad_norm": 0.9600672721862793,
"learning_rate": 4.998599633043202e-06,
"loss": 0.7902,
"step": 281
},
{
"epoch": 0.09901685393258428,
"grad_norm": 0.9067400097846985,
"learning_rate": 4.9985841180950245e-06,
"loss": 0.7779,
"step": 282
},
{
"epoch": 0.09936797752808989,
"grad_norm": 0.8911014795303345,
"learning_rate": 4.998568517697408e-06,
"loss": 0.7605,
"step": 283
},
{
"epoch": 0.0997191011235955,
"grad_norm": 0.8919551968574524,
"learning_rate": 4.998552831850886e-06,
"loss": 0.7416,
"step": 284
},
{
"epoch": 0.10007022471910113,
"grad_norm": 0.8880231380462646,
"learning_rate": 4.998537060555995e-06,
"loss": 0.7757,
"step": 285
},
{
"epoch": 0.10042134831460674,
"grad_norm": 0.9274663329124451,
"learning_rate": 4.998521203813275e-06,
"loss": 0.7925,
"step": 286
},
{
"epoch": 0.10077247191011236,
"grad_norm": 0.9023240804672241,
"learning_rate": 4.998505261623266e-06,
"loss": 0.8081,
"step": 287
},
{
"epoch": 0.10112359550561797,
"grad_norm": 0.8825272917747498,
"learning_rate": 4.998489233986518e-06,
"loss": 0.7864,
"step": 288
},
{
"epoch": 0.1014747191011236,
"grad_norm": 0.885826826095581,
"learning_rate": 4.9984731209035745e-06,
"loss": 0.7759,
"step": 289
},
{
"epoch": 0.10182584269662921,
"grad_norm": 0.8802645802497864,
"learning_rate": 4.998456922374988e-06,
"loss": 0.769,
"step": 290
},
{
"epoch": 0.10217696629213484,
"grad_norm": 0.9090613126754761,
"learning_rate": 4.998440638401314e-06,
"loss": 0.7931,
"step": 291
},
{
"epoch": 0.10252808988764045,
"grad_norm": 0.9373996257781982,
"learning_rate": 4.998424268983107e-06,
"loss": 0.7443,
"step": 292
},
{
"epoch": 0.10287921348314606,
"grad_norm": 0.9003362655639648,
"learning_rate": 4.998407814120928e-06,
"loss": 0.7983,
"step": 293
},
{
"epoch": 0.10323033707865169,
"grad_norm": 0.9281851053237915,
"learning_rate": 4.998391273815341e-06,
"loss": 0.8026,
"step": 294
},
{
"epoch": 0.1035814606741573,
"grad_norm": 0.8824150562286377,
"learning_rate": 4.998374648066909e-06,
"loss": 0.7435,
"step": 295
},
{
"epoch": 0.10393258426966293,
"grad_norm": 0.9006941318511963,
"learning_rate": 4.998357936876202e-06,
"loss": 0.7567,
"step": 296
},
{
"epoch": 0.10428370786516854,
"grad_norm": 0.8622567057609558,
"learning_rate": 4.998341140243792e-06,
"loss": 0.7532,
"step": 297
},
{
"epoch": 0.10463483146067416,
"grad_norm": 0.8983448147773743,
"learning_rate": 4.9983242581702525e-06,
"loss": 0.7686,
"step": 298
},
{
"epoch": 0.10498595505617977,
"grad_norm": 0.8893060684204102,
"learning_rate": 4.998307290656162e-06,
"loss": 0.7697,
"step": 299
},
{
"epoch": 0.10533707865168539,
"grad_norm": 0.8906247615814209,
"learning_rate": 4.998290237702098e-06,
"loss": 0.7628,
"step": 300
},
{
"epoch": 0.10568820224719101,
"grad_norm": 0.8763146996498108,
"learning_rate": 4.998273099308647e-06,
"loss": 0.7742,
"step": 301
},
{
"epoch": 0.10603932584269662,
"grad_norm": 0.8954668045043945,
"learning_rate": 4.998255875476394e-06,
"loss": 0.7666,
"step": 302
},
{
"epoch": 0.10639044943820225,
"grad_norm": 0.9178535342216492,
"learning_rate": 4.9982385662059275e-06,
"loss": 0.7679,
"step": 303
},
{
"epoch": 0.10674157303370786,
"grad_norm": 0.9487611651420593,
"learning_rate": 4.99822117149784e-06,
"loss": 0.7744,
"step": 304
},
{
"epoch": 0.10709269662921349,
"grad_norm": 0.9466339945793152,
"learning_rate": 4.998203691352726e-06,
"loss": 0.7608,
"step": 305
},
{
"epoch": 0.1074438202247191,
"grad_norm": 0.9768984913825989,
"learning_rate": 4.998186125771184e-06,
"loss": 0.8028,
"step": 306
},
{
"epoch": 0.10779494382022473,
"grad_norm": 0.9045122861862183,
"learning_rate": 4.998168474753814e-06,
"loss": 0.7365,
"step": 307
},
{
"epoch": 0.10814606741573034,
"grad_norm": 0.9124319553375244,
"learning_rate": 4.99815073830122e-06,
"loss": 0.7686,
"step": 308
},
{
"epoch": 0.10849719101123595,
"grad_norm": 0.9172065854072571,
"learning_rate": 4.998132916414008e-06,
"loss": 0.778,
"step": 309
},
{
"epoch": 0.10884831460674158,
"grad_norm": 0.9291918873786926,
"learning_rate": 4.9981150090927876e-06,
"loss": 0.7828,
"step": 310
},
{
"epoch": 0.10919943820224719,
"grad_norm": 0.9258780479431152,
"learning_rate": 4.998097016338172e-06,
"loss": 0.7822,
"step": 311
},
{
"epoch": 0.10955056179775281,
"grad_norm": 0.9169499278068542,
"learning_rate": 4.998078938150776e-06,
"loss": 0.7735,
"step": 312
},
{
"epoch": 0.10990168539325842,
"grad_norm": 0.8983737826347351,
"learning_rate": 4.998060774531218e-06,
"loss": 0.7729,
"step": 313
},
{
"epoch": 0.11025280898876405,
"grad_norm": 0.949192225933075,
"learning_rate": 4.99804252548012e-06,
"loss": 0.8153,
"step": 314
},
{
"epoch": 0.11060393258426966,
"grad_norm": 0.9175098538398743,
"learning_rate": 4.998024190998104e-06,
"loss": 0.7784,
"step": 315
},
{
"epoch": 0.11095505617977527,
"grad_norm": 0.9033719897270203,
"learning_rate": 4.998005771085799e-06,
"loss": 0.748,
"step": 316
},
{
"epoch": 0.1113061797752809,
"grad_norm": 0.9692884683609009,
"learning_rate": 4.997987265743834e-06,
"loss": 0.7795,
"step": 317
},
{
"epoch": 0.11165730337078651,
"grad_norm": 0.8999868035316467,
"learning_rate": 4.997968674972842e-06,
"loss": 0.7564,
"step": 318
},
{
"epoch": 0.11200842696629214,
"grad_norm": 0.9252747893333435,
"learning_rate": 4.997949998773458e-06,
"loss": 0.8054,
"step": 319
},
{
"epoch": 0.11235955056179775,
"grad_norm": 0.8910780549049377,
"learning_rate": 4.997931237146323e-06,
"loss": 0.7671,
"step": 320
},
{
"epoch": 0.11271067415730338,
"grad_norm": 0.9301937818527222,
"learning_rate": 4.997912390092077e-06,
"loss": 0.7906,
"step": 321
},
{
"epoch": 0.11306179775280899,
"grad_norm": 0.9235373735427856,
"learning_rate": 4.997893457611364e-06,
"loss": 0.7626,
"step": 322
},
{
"epoch": 0.11341292134831461,
"grad_norm": 0.9015495777130127,
"learning_rate": 4.997874439704833e-06,
"loss": 0.7634,
"step": 323
},
{
"epoch": 0.11376404494382023,
"grad_norm": 0.8859757781028748,
"learning_rate": 4.997855336373133e-06,
"loss": 0.7435,
"step": 324
},
{
"epoch": 0.11411516853932584,
"grad_norm": 0.9129098057746887,
"learning_rate": 4.9978361476169176e-06,
"loss": 0.7505,
"step": 325
},
{
"epoch": 0.11446629213483146,
"grad_norm": 0.9024680256843567,
"learning_rate": 4.997816873436845e-06,
"loss": 0.7402,
"step": 326
},
{
"epoch": 0.11481741573033707,
"grad_norm": 0.9123002290725708,
"learning_rate": 4.997797513833571e-06,
"loss": 0.7751,
"step": 327
},
{
"epoch": 0.1151685393258427,
"grad_norm": 0.8788150548934937,
"learning_rate": 4.9977780688077606e-06,
"loss": 0.7882,
"step": 328
},
{
"epoch": 0.11551966292134831,
"grad_norm": 0.9502024054527283,
"learning_rate": 4.997758538360077e-06,
"loss": 0.7777,
"step": 329
},
{
"epoch": 0.11587078651685394,
"grad_norm": 0.9458324909210205,
"learning_rate": 4.9977389224911884e-06,
"loss": 0.7666,
"step": 330
},
{
"epoch": 0.11622191011235955,
"grad_norm": 0.9191945195198059,
"learning_rate": 4.997719221201766e-06,
"loss": 0.7678,
"step": 331
},
{
"epoch": 0.11657303370786516,
"grad_norm": 0.9368025660514832,
"learning_rate": 4.997699434492485e-06,
"loss": 0.7689,
"step": 332
},
{
"epoch": 0.11692415730337079,
"grad_norm": 0.9120011925697327,
"learning_rate": 4.99767956236402e-06,
"loss": 0.7489,
"step": 333
},
{
"epoch": 0.1172752808988764,
"grad_norm": 0.9374188780784607,
"learning_rate": 4.997659604817051e-06,
"loss": 0.7358,
"step": 334
},
{
"epoch": 0.11762640449438203,
"grad_norm": 0.9129819273948669,
"learning_rate": 4.997639561852261e-06,
"loss": 0.7591,
"step": 335
},
{
"epoch": 0.11797752808988764,
"grad_norm": 0.9249037504196167,
"learning_rate": 4.997619433470334e-06,
"loss": 0.782,
"step": 336
},
{
"epoch": 0.11832865168539326,
"grad_norm": 0.8872626423835754,
"learning_rate": 4.9975992196719614e-06,
"loss": 0.7477,
"step": 337
},
{
"epoch": 0.11867977528089887,
"grad_norm": 0.9343289732933044,
"learning_rate": 4.9975789204578315e-06,
"loss": 0.7702,
"step": 338
},
{
"epoch": 0.1190308988764045,
"grad_norm": 0.9379141926765442,
"learning_rate": 4.997558535828641e-06,
"loss": 0.7847,
"step": 339
},
{
"epoch": 0.11938202247191011,
"grad_norm": 0.9160316586494446,
"learning_rate": 4.997538065785085e-06,
"loss": 0.7658,
"step": 340
},
{
"epoch": 0.11973314606741572,
"grad_norm": 0.9021292328834534,
"learning_rate": 4.997517510327864e-06,
"loss": 0.8114,
"step": 341
},
{
"epoch": 0.12008426966292135,
"grad_norm": 0.9155046343803406,
"learning_rate": 4.997496869457681e-06,
"loss": 0.7579,
"step": 342
},
{
"epoch": 0.12043539325842696,
"grad_norm": 0.8777235746383667,
"learning_rate": 4.997476143175243e-06,
"loss": 0.7832,
"step": 343
},
{
"epoch": 0.12078651685393259,
"grad_norm": 0.8728359341621399,
"learning_rate": 4.997455331481258e-06,
"loss": 0.7647,
"step": 344
},
{
"epoch": 0.1211376404494382,
"grad_norm": 0.9153422713279724,
"learning_rate": 4.997434434376437e-06,
"loss": 0.7396,
"step": 345
},
{
"epoch": 0.12148876404494383,
"grad_norm": 0.900204598903656,
"learning_rate": 4.9974134518614964e-06,
"loss": 0.7619,
"step": 346
},
{
"epoch": 0.12183988764044944,
"grad_norm": 0.9725254774093628,
"learning_rate": 4.997392383937153e-06,
"loss": 0.7588,
"step": 347
},
{
"epoch": 0.12219101123595505,
"grad_norm": 0.9041215777397156,
"learning_rate": 4.997371230604126e-06,
"loss": 0.7712,
"step": 348
},
{
"epoch": 0.12254213483146068,
"grad_norm": 0.9517892599105835,
"learning_rate": 4.997349991863141e-06,
"loss": 0.7758,
"step": 349
},
{
"epoch": 0.12289325842696629,
"grad_norm": 0.9502705335617065,
"learning_rate": 4.997328667714922e-06,
"loss": 0.8003,
"step": 350
},
{
"epoch": 0.12324438202247191,
"grad_norm": 0.9965865612030029,
"learning_rate": 4.9973072581602005e-06,
"loss": 0.751,
"step": 351
},
{
"epoch": 0.12359550561797752,
"grad_norm": 0.9221112132072449,
"learning_rate": 4.997285763199707e-06,
"loss": 0.7679,
"step": 352
},
{
"epoch": 0.12394662921348315,
"grad_norm": 0.940747857093811,
"learning_rate": 4.997264182834179e-06,
"loss": 0.7863,
"step": 353
},
{
"epoch": 0.12429775280898876,
"grad_norm": 0.925940215587616,
"learning_rate": 4.997242517064351e-06,
"loss": 0.7711,
"step": 354
},
{
"epoch": 0.12464887640449439,
"grad_norm": 0.8775484561920166,
"learning_rate": 4.997220765890967e-06,
"loss": 0.7504,
"step": 355
},
{
"epoch": 0.125,
"grad_norm": 0.9700306057929993,
"learning_rate": 4.997198929314769e-06,
"loss": 0.7385,
"step": 356
},
{
"epoch": 0.12535112359550563,
"grad_norm": 0.9070924520492554,
"learning_rate": 4.997177007336505e-06,
"loss": 0.7538,
"step": 357
},
{
"epoch": 0.12570224719101122,
"grad_norm": 0.8817124962806702,
"learning_rate": 4.997154999956924e-06,
"loss": 0.7549,
"step": 358
},
{
"epoch": 0.12605337078651685,
"grad_norm": 0.9093615412712097,
"learning_rate": 4.997132907176779e-06,
"loss": 0.7621,
"step": 359
},
{
"epoch": 0.12640449438202248,
"grad_norm": 0.8783479332923889,
"learning_rate": 4.9971107289968256e-06,
"loss": 0.7369,
"step": 360
},
{
"epoch": 0.1267556179775281,
"grad_norm": 0.9039416313171387,
"learning_rate": 4.997088465417822e-06,
"loss": 0.7664,
"step": 361
},
{
"epoch": 0.1271067415730337,
"grad_norm": 0.9135423302650452,
"learning_rate": 4.997066116440529e-06,
"loss": 0.7887,
"step": 362
},
{
"epoch": 0.12745786516853932,
"grad_norm": 0.8956524133682251,
"learning_rate": 4.997043682065712e-06,
"loss": 0.7303,
"step": 363
},
{
"epoch": 0.12780898876404495,
"grad_norm": 0.9221348166465759,
"learning_rate": 4.997021162294138e-06,
"loss": 0.7707,
"step": 364
},
{
"epoch": 0.12816011235955055,
"grad_norm": 0.920876681804657,
"learning_rate": 4.996998557126577e-06,
"loss": 0.7825,
"step": 365
},
{
"epoch": 0.12851123595505617,
"grad_norm": 0.9243044853210449,
"learning_rate": 4.996975866563802e-06,
"loss": 0.7362,
"step": 366
},
{
"epoch": 0.1288623595505618,
"grad_norm": 0.89571613073349,
"learning_rate": 4.996953090606589e-06,
"loss": 0.7645,
"step": 367
},
{
"epoch": 0.12921348314606743,
"grad_norm": 0.9237456917762756,
"learning_rate": 4.9969302292557165e-06,
"loss": 0.766,
"step": 368
},
{
"epoch": 0.12956460674157302,
"grad_norm": 0.9664248824119568,
"learning_rate": 4.996907282511968e-06,
"loss": 0.7801,
"step": 369
},
{
"epoch": 0.12991573033707865,
"grad_norm": 0.9590004682540894,
"learning_rate": 4.9968842503761265e-06,
"loss": 0.7544,
"step": 370
},
{
"epoch": 0.13026685393258428,
"grad_norm": 0.9026381969451904,
"learning_rate": 4.99686113284898e-06,
"loss": 0.7756,
"step": 371
},
{
"epoch": 0.1306179775280899,
"grad_norm": 0.9528935551643372,
"learning_rate": 4.9968379299313195e-06,
"loss": 0.7673,
"step": 372
},
{
"epoch": 0.1309691011235955,
"grad_norm": 0.9700410962104797,
"learning_rate": 4.996814641623937e-06,
"loss": 0.7746,
"step": 373
},
{
"epoch": 0.13132022471910113,
"grad_norm": 0.931694746017456,
"learning_rate": 4.996791267927632e-06,
"loss": 0.8078,
"step": 374
},
{
"epoch": 0.13167134831460675,
"grad_norm": 0.9812837839126587,
"learning_rate": 4.996767808843203e-06,
"loss": 0.7662,
"step": 375
},
{
"epoch": 0.13202247191011235,
"grad_norm": 0.9266939163208008,
"learning_rate": 4.996744264371449e-06,
"loss": 0.7468,
"step": 376
},
{
"epoch": 0.13237359550561797,
"grad_norm": 0.8732329607009888,
"learning_rate": 4.996720634513179e-06,
"loss": 0.7774,
"step": 377
},
{
"epoch": 0.1327247191011236,
"grad_norm": 0.9104029536247253,
"learning_rate": 4.9966969192692e-06,
"loss": 0.7423,
"step": 378
},
{
"epoch": 0.13307584269662923,
"grad_norm": 0.8799782991409302,
"learning_rate": 4.996673118640323e-06,
"loss": 0.7086,
"step": 379
},
{
"epoch": 0.13342696629213482,
"grad_norm": 0.9318313598632812,
"learning_rate": 4.9966492326273605e-06,
"loss": 0.7492,
"step": 380
},
{
"epoch": 0.13377808988764045,
"grad_norm": 0.9577075839042664,
"learning_rate": 4.996625261231131e-06,
"loss": 0.7935,
"step": 381
},
{
"epoch": 0.13412921348314608,
"grad_norm": 0.9912057518959045,
"learning_rate": 4.996601204452455e-06,
"loss": 0.7515,
"step": 382
},
{
"epoch": 0.13448033707865167,
"grad_norm": 0.9166970252990723,
"learning_rate": 4.996577062292154e-06,
"loss": 0.7364,
"step": 383
},
{
"epoch": 0.1348314606741573,
"grad_norm": 0.9245953559875488,
"learning_rate": 4.996552834751053e-06,
"loss": 0.7895,
"step": 384
},
{
"epoch": 0.13518258426966293,
"grad_norm": 0.9169378876686096,
"learning_rate": 4.996528521829982e-06,
"loss": 0.7632,
"step": 385
},
{
"epoch": 0.13553370786516855,
"grad_norm": 0.8847085237503052,
"learning_rate": 4.996504123529772e-06,
"loss": 0.7607,
"step": 386
},
{
"epoch": 0.13588483146067415,
"grad_norm": 0.8744500875473022,
"learning_rate": 4.996479639851256e-06,
"loss": 0.7146,
"step": 387
},
{
"epoch": 0.13623595505617977,
"grad_norm": 0.9195184707641602,
"learning_rate": 4.996455070795274e-06,
"loss": 0.771,
"step": 388
},
{
"epoch": 0.1365870786516854,
"grad_norm": 0.8896746039390564,
"learning_rate": 4.996430416362664e-06,
"loss": 0.7239,
"step": 389
},
{
"epoch": 0.136938202247191,
"grad_norm": 0.9265782237052917,
"learning_rate": 4.99640567655427e-06,
"loss": 0.7607,
"step": 390
},
{
"epoch": 0.13728932584269662,
"grad_norm": 0.9531838893890381,
"learning_rate": 4.996380851370939e-06,
"loss": 0.7487,
"step": 391
},
{
"epoch": 0.13764044943820225,
"grad_norm": 0.9503312706947327,
"learning_rate": 4.996355940813518e-06,
"loss": 0.7889,
"step": 392
},
{
"epoch": 0.13799157303370788,
"grad_norm": 0.8666608929634094,
"learning_rate": 4.99633094488286e-06,
"loss": 0.7539,
"step": 393
},
{
"epoch": 0.13834269662921347,
"grad_norm": 0.9313355684280396,
"learning_rate": 4.996305863579821e-06,
"loss": 0.7674,
"step": 394
},
{
"epoch": 0.1386938202247191,
"grad_norm": 0.9345743656158447,
"learning_rate": 4.9962806969052565e-06,
"loss": 0.8057,
"step": 395
},
{
"epoch": 0.13904494382022473,
"grad_norm": 0.922758936882019,
"learning_rate": 4.996255444860029e-06,
"loss": 0.7404,
"step": 396
},
{
"epoch": 0.13939606741573032,
"grad_norm": 0.8888351321220398,
"learning_rate": 4.996230107445001e-06,
"loss": 0.753,
"step": 397
},
{
"epoch": 0.13974719101123595,
"grad_norm": 0.9118607640266418,
"learning_rate": 4.996204684661039e-06,
"loss": 0.7562,
"step": 398
},
{
"epoch": 0.14009831460674158,
"grad_norm": 0.9427488446235657,
"learning_rate": 4.996179176509013e-06,
"loss": 0.7085,
"step": 399
},
{
"epoch": 0.1404494382022472,
"grad_norm": 0.9284588098526001,
"learning_rate": 4.996153582989795e-06,
"loss": 0.7674,
"step": 400
},
{
"epoch": 0.1408005617977528,
"grad_norm": 0.8856985569000244,
"learning_rate": 4.996127904104261e-06,
"loss": 0.7357,
"step": 401
},
{
"epoch": 0.14115168539325842,
"grad_norm": 0.9342722296714783,
"learning_rate": 4.996102139853289e-06,
"loss": 0.7387,
"step": 402
},
{
"epoch": 0.14150280898876405,
"grad_norm": 0.9672777056694031,
"learning_rate": 4.996076290237759e-06,
"loss": 0.7681,
"step": 403
},
{
"epoch": 0.14185393258426968,
"grad_norm": 0.929104745388031,
"learning_rate": 4.996050355258556e-06,
"loss": 0.7266,
"step": 404
},
{
"epoch": 0.14220505617977527,
"grad_norm": 0.9371368288993835,
"learning_rate": 4.996024334916567e-06,
"loss": 0.7841,
"step": 405
},
{
"epoch": 0.1425561797752809,
"grad_norm": 0.8848541975021362,
"learning_rate": 4.995998229212681e-06,
"loss": 0.7659,
"step": 406
},
{
"epoch": 0.14290730337078653,
"grad_norm": 0.9632852673530579,
"learning_rate": 4.995972038147792e-06,
"loss": 0.7871,
"step": 407
},
{
"epoch": 0.14325842696629212,
"grad_norm": 0.9157865047454834,
"learning_rate": 4.9959457617227946e-06,
"loss": 0.7529,
"step": 408
},
{
"epoch": 0.14360955056179775,
"grad_norm": 0.9337014555931091,
"learning_rate": 4.995919399938588e-06,
"loss": 0.7427,
"step": 409
},
{
"epoch": 0.14396067415730338,
"grad_norm": 0.9515252113342285,
"learning_rate": 4.995892952796074e-06,
"loss": 0.7577,
"step": 410
},
{
"epoch": 0.144311797752809,
"grad_norm": 0.9301478266716003,
"learning_rate": 4.995866420296157e-06,
"loss": 0.7825,
"step": 411
},
{
"epoch": 0.1446629213483146,
"grad_norm": 1.4020230770111084,
"learning_rate": 4.995839802439745e-06,
"loss": 0.7434,
"step": 412
},
{
"epoch": 0.14501404494382023,
"grad_norm": 0.9150766730308533,
"learning_rate": 4.995813099227745e-06,
"loss": 0.7038,
"step": 413
},
{
"epoch": 0.14536516853932585,
"grad_norm": 0.9467854499816895,
"learning_rate": 4.995786310661075e-06,
"loss": 0.7721,
"step": 414
},
{
"epoch": 0.14571629213483145,
"grad_norm": 0.8949883580207825,
"learning_rate": 4.995759436740648e-06,
"loss": 0.7174,
"step": 415
},
{
"epoch": 0.14606741573033707,
"grad_norm": 0.9632487893104553,
"learning_rate": 4.995732477467383e-06,
"loss": 0.8002,
"step": 416
},
{
"epoch": 0.1464185393258427,
"grad_norm": 0.8924590945243835,
"learning_rate": 4.995705432842204e-06,
"loss": 0.7575,
"step": 417
},
{
"epoch": 0.14676966292134833,
"grad_norm": 0.8902977705001831,
"learning_rate": 4.995678302866035e-06,
"loss": 0.73,
"step": 418
},
{
"epoch": 0.14712078651685392,
"grad_norm": 0.9031338095664978,
"learning_rate": 4.9956510875398035e-06,
"loss": 0.7697,
"step": 419
},
{
"epoch": 0.14747191011235955,
"grad_norm": 0.9241436719894409,
"learning_rate": 4.99562378686444e-06,
"loss": 0.7594,
"step": 420
},
{
"epoch": 0.14782303370786518,
"grad_norm": 1.0044499635696411,
"learning_rate": 4.995596400840879e-06,
"loss": 0.7938,
"step": 421
},
{
"epoch": 0.14817415730337077,
"grad_norm": 0.9393342733383179,
"learning_rate": 4.995568929470055e-06,
"loss": 0.8142,
"step": 422
},
{
"epoch": 0.1485252808988764,
"grad_norm": 0.9527561664581299,
"learning_rate": 4.99554137275291e-06,
"loss": 0.7825,
"step": 423
},
{
"epoch": 0.14887640449438203,
"grad_norm": 0.9195716381072998,
"learning_rate": 4.995513730690386e-06,
"loss": 0.7577,
"step": 424
},
{
"epoch": 0.14922752808988765,
"grad_norm": 0.9349954128265381,
"learning_rate": 4.995486003283428e-06,
"loss": 0.7159,
"step": 425
},
{
"epoch": 0.14957865168539325,
"grad_norm": 0.922205924987793,
"learning_rate": 4.9954581905329845e-06,
"loss": 0.8017,
"step": 426
},
{
"epoch": 0.14992977528089887,
"grad_norm": 0.955886721611023,
"learning_rate": 4.995430292440005e-06,
"loss": 0.7813,
"step": 427
},
{
"epoch": 0.1502808988764045,
"grad_norm": 0.9729349613189697,
"learning_rate": 4.995402309005445e-06,
"loss": 0.7744,
"step": 428
},
{
"epoch": 0.1506320224719101,
"grad_norm": 0.8943726420402527,
"learning_rate": 4.995374240230262e-06,
"loss": 0.775,
"step": 429
},
{
"epoch": 0.15098314606741572,
"grad_norm": 0.9235265254974365,
"learning_rate": 4.995346086115416e-06,
"loss": 0.759,
"step": 430
},
{
"epoch": 0.15133426966292135,
"grad_norm": 0.9518915414810181,
"learning_rate": 4.995317846661868e-06,
"loss": 0.7563,
"step": 431
},
{
"epoch": 0.15168539325842698,
"grad_norm": 0.9186070561408997,
"learning_rate": 4.9952895218705845e-06,
"loss": 0.7551,
"step": 432
},
{
"epoch": 0.15203651685393257,
"grad_norm": 0.9712167978286743,
"learning_rate": 4.995261111742536e-06,
"loss": 0.7765,
"step": 433
},
{
"epoch": 0.1523876404494382,
"grad_norm": 0.9043035507202148,
"learning_rate": 4.995232616278691e-06,
"loss": 0.7224,
"step": 434
},
{
"epoch": 0.15273876404494383,
"grad_norm": 1.357176661491394,
"learning_rate": 4.995204035480027e-06,
"loss": 0.7538,
"step": 435
},
{
"epoch": 0.15308988764044945,
"grad_norm": 0.9199062585830688,
"learning_rate": 4.995175369347521e-06,
"loss": 0.7396,
"step": 436
},
{
"epoch": 0.15344101123595505,
"grad_norm": 0.9276919960975647,
"learning_rate": 4.995146617882151e-06,
"loss": 0.7626,
"step": 437
},
{
"epoch": 0.15379213483146068,
"grad_norm": 0.8831456303596497,
"learning_rate": 4.995117781084901e-06,
"loss": 0.7692,
"step": 438
},
{
"epoch": 0.1541432584269663,
"grad_norm": 0.9020641446113586,
"learning_rate": 4.99508885895676e-06,
"loss": 0.7578,
"step": 439
},
{
"epoch": 0.1544943820224719,
"grad_norm": 0.9140202403068542,
"learning_rate": 4.995059851498714e-06,
"loss": 0.7728,
"step": 440
},
{
"epoch": 0.15484550561797752,
"grad_norm": 0.9247411489486694,
"learning_rate": 4.995030758711756e-06,
"loss": 0.7397,
"step": 441
},
{
"epoch": 0.15519662921348315,
"grad_norm": 0.8848091959953308,
"learning_rate": 4.995001580596882e-06,
"loss": 0.6819,
"step": 442
},
{
"epoch": 0.15554775280898878,
"grad_norm": 1.0068150758743286,
"learning_rate": 4.994972317155088e-06,
"loss": 0.7437,
"step": 443
},
{
"epoch": 0.15589887640449437,
"grad_norm": 0.9932318329811096,
"learning_rate": 4.994942968387376e-06,
"loss": 0.7693,
"step": 444
},
{
"epoch": 0.15625,
"grad_norm": 0.920421838760376,
"learning_rate": 4.99491353429475e-06,
"loss": 0.7649,
"step": 445
},
{
"epoch": 0.15660112359550563,
"grad_norm": 0.9176994562149048,
"learning_rate": 4.9948840148782165e-06,
"loss": 0.7391,
"step": 446
},
{
"epoch": 0.15695224719101122,
"grad_norm": 0.9588678479194641,
"learning_rate": 4.994854410138783e-06,
"loss": 0.7615,
"step": 447
},
{
"epoch": 0.15730337078651685,
"grad_norm": 0.9149226546287537,
"learning_rate": 4.9948247200774645e-06,
"loss": 0.75,
"step": 448
},
{
"epoch": 0.15765449438202248,
"grad_norm": 0.8968827724456787,
"learning_rate": 4.994794944695275e-06,
"loss": 0.7582,
"step": 449
},
{
"epoch": 0.1580056179775281,
"grad_norm": 0.8770960569381714,
"learning_rate": 4.994765083993234e-06,
"loss": 0.7121,
"step": 450
},
{
"epoch": 0.1583567415730337,
"grad_norm": 0.9019699692726135,
"learning_rate": 4.9947351379723605e-06,
"loss": 0.7131,
"step": 451
},
{
"epoch": 0.15870786516853932,
"grad_norm": 0.8884302973747253,
"learning_rate": 4.994705106633682e-06,
"loss": 0.7292,
"step": 452
},
{
"epoch": 0.15905898876404495,
"grad_norm": 0.9590429663658142,
"learning_rate": 4.994674989978222e-06,
"loss": 0.7454,
"step": 453
},
{
"epoch": 0.15941011235955055,
"grad_norm": 0.8948072195053101,
"learning_rate": 4.9946447880070124e-06,
"loss": 0.7667,
"step": 454
},
{
"epoch": 0.15976123595505617,
"grad_norm": 0.9360845685005188,
"learning_rate": 4.994614500721086e-06,
"loss": 0.7633,
"step": 455
},
{
"epoch": 0.1601123595505618,
"grad_norm": 0.9061880111694336,
"learning_rate": 4.9945841281214785e-06,
"loss": 0.7026,
"step": 456
},
{
"epoch": 0.16046348314606743,
"grad_norm": 0.918328583240509,
"learning_rate": 4.994553670209228e-06,
"loss": 0.7367,
"step": 457
},
{
"epoch": 0.16081460674157302,
"grad_norm": 0.9201732277870178,
"learning_rate": 4.994523126985377e-06,
"loss": 0.7326,
"step": 458
},
{
"epoch": 0.16116573033707865,
"grad_norm": 0.9703920483589172,
"learning_rate": 4.9944924984509695e-06,
"loss": 0.7638,
"step": 459
},
{
"epoch": 0.16151685393258428,
"grad_norm": 0.8904596567153931,
"learning_rate": 4.994461784607053e-06,
"loss": 0.7717,
"step": 460
},
{
"epoch": 0.1618679775280899,
"grad_norm": 0.9822039604187012,
"learning_rate": 4.994430985454678e-06,
"loss": 0.7674,
"step": 461
},
{
"epoch": 0.1622191011235955,
"grad_norm": 0.9268808364868164,
"learning_rate": 4.9944001009948986e-06,
"loss": 0.7725,
"step": 462
},
{
"epoch": 0.16257022471910113,
"grad_norm": 0.8941001296043396,
"learning_rate": 4.994369131228769e-06,
"loss": 0.7361,
"step": 463
},
{
"epoch": 0.16292134831460675,
"grad_norm": 0.8854597210884094,
"learning_rate": 4.99433807615735e-06,
"loss": 0.7426,
"step": 464
},
{
"epoch": 0.16327247191011235,
"grad_norm": 1.0173115730285645,
"learning_rate": 4.994306935781704e-06,
"loss": 0.7724,
"step": 465
},
{
"epoch": 0.16362359550561797,
"grad_norm": 0.9098212122917175,
"learning_rate": 4.994275710102894e-06,
"loss": 0.6989,
"step": 466
},
{
"epoch": 0.1639747191011236,
"grad_norm": 0.9020840525627136,
"learning_rate": 4.99424439912199e-06,
"loss": 0.7312,
"step": 467
},
{
"epoch": 0.16432584269662923,
"grad_norm": 0.9145157337188721,
"learning_rate": 4.994213002840062e-06,
"loss": 0.7334,
"step": 468
},
{
"epoch": 0.16467696629213482,
"grad_norm": 0.9347027540206909,
"learning_rate": 4.994181521258183e-06,
"loss": 0.7508,
"step": 469
},
{
"epoch": 0.16502808988764045,
"grad_norm": 0.9202942252159119,
"learning_rate": 4.994149954377431e-06,
"loss": 0.7351,
"step": 470
},
{
"epoch": 0.16537921348314608,
"grad_norm": 0.9013196229934692,
"learning_rate": 4.994118302198884e-06,
"loss": 0.7618,
"step": 471
},
{
"epoch": 0.16573033707865167,
"grad_norm": 0.9281770586967468,
"learning_rate": 4.9940865647236255e-06,
"loss": 0.7661,
"step": 472
},
{
"epoch": 0.1660814606741573,
"grad_norm": 0.9201061129570007,
"learning_rate": 4.99405474195274e-06,
"loss": 0.7203,
"step": 473
},
{
"epoch": 0.16643258426966293,
"grad_norm": 0.9211746454238892,
"learning_rate": 4.994022833887318e-06,
"loss": 0.724,
"step": 474
},
{
"epoch": 0.16678370786516855,
"grad_norm": 0.9289520978927612,
"learning_rate": 4.993990840528448e-06,
"loss": 0.7907,
"step": 475
},
{
"epoch": 0.16713483146067415,
"grad_norm": 0.9343749284744263,
"learning_rate": 4.993958761877227e-06,
"loss": 0.7707,
"step": 476
},
{
"epoch": 0.16748595505617977,
"grad_norm": 0.9140953421592712,
"learning_rate": 4.9939265979347495e-06,
"loss": 0.7558,
"step": 477
},
{
"epoch": 0.1678370786516854,
"grad_norm": 0.8889614939689636,
"learning_rate": 4.993894348702117e-06,
"loss": 0.7615,
"step": 478
},
{
"epoch": 0.168188202247191,
"grad_norm": 0.8894888162612915,
"learning_rate": 4.993862014180431e-06,
"loss": 0.7536,
"step": 479
},
{
"epoch": 0.16853932584269662,
"grad_norm": 0.8904688954353333,
"learning_rate": 4.993829594370798e-06,
"loss": 0.7471,
"step": 480
},
{
"epoch": 0.16889044943820225,
"grad_norm": 0.9064297080039978,
"learning_rate": 4.993797089274327e-06,
"loss": 0.7524,
"step": 481
},
{
"epoch": 0.16924157303370788,
"grad_norm": 0.8890877962112427,
"learning_rate": 4.99376449889213e-06,
"loss": 0.7727,
"step": 482
},
{
"epoch": 0.16959269662921347,
"grad_norm": 0.9089619517326355,
"learning_rate": 4.993731823225321e-06,
"loss": 0.7661,
"step": 483
},
{
"epoch": 0.1699438202247191,
"grad_norm": 0.9284274578094482,
"learning_rate": 4.993699062275017e-06,
"loss": 0.7337,
"step": 484
},
{
"epoch": 0.17029494382022473,
"grad_norm": 0.9275715947151184,
"learning_rate": 4.99366621604234e-06,
"loss": 0.7417,
"step": 485
},
{
"epoch": 0.17064606741573032,
"grad_norm": 0.9078076481819153,
"learning_rate": 4.9936332845284105e-06,
"loss": 0.7366,
"step": 486
},
{
"epoch": 0.17099719101123595,
"grad_norm": 0.9106430411338806,
"learning_rate": 4.993600267734359e-06,
"loss": 0.7738,
"step": 487
},
{
"epoch": 0.17134831460674158,
"grad_norm": 1.0632219314575195,
"learning_rate": 4.99356716566131e-06,
"loss": 0.724,
"step": 488
},
{
"epoch": 0.1716994382022472,
"grad_norm": 0.9347037076950073,
"learning_rate": 4.9935339783103985e-06,
"loss": 0.7273,
"step": 489
},
{
"epoch": 0.1720505617977528,
"grad_norm": 0.9145859479904175,
"learning_rate": 4.9935007056827586e-06,
"loss": 0.7603,
"step": 490
},
{
"epoch": 0.17240168539325842,
"grad_norm": 0.924089252948761,
"learning_rate": 4.993467347779529e-06,
"loss": 0.7289,
"step": 491
},
{
"epoch": 0.17275280898876405,
"grad_norm": 0.9254995584487915,
"learning_rate": 4.993433904601849e-06,
"loss": 0.7614,
"step": 492
},
{
"epoch": 0.17310393258426968,
"grad_norm": 0.9493343234062195,
"learning_rate": 4.993400376150863e-06,
"loss": 0.7889,
"step": 493
},
{
"epoch": 0.17345505617977527,
"grad_norm": 0.8984596729278564,
"learning_rate": 4.9933667624277184e-06,
"loss": 0.7304,
"step": 494
},
{
"epoch": 0.1738061797752809,
"grad_norm": 0.9199086427688599,
"learning_rate": 4.9933330634335634e-06,
"loss": 0.7386,
"step": 495
},
{
"epoch": 0.17415730337078653,
"grad_norm": 0.9047686457633972,
"learning_rate": 4.993299279169552e-06,
"loss": 0.7469,
"step": 496
},
{
"epoch": 0.17450842696629212,
"grad_norm": 0.9360693097114563,
"learning_rate": 4.993265409636839e-06,
"loss": 0.7595,
"step": 497
},
{
"epoch": 0.17485955056179775,
"grad_norm": 0.9289848804473877,
"learning_rate": 4.993231454836581e-06,
"loss": 0.7701,
"step": 498
},
{
"epoch": 0.17521067415730338,
"grad_norm": 0.8970044255256653,
"learning_rate": 4.993197414769942e-06,
"loss": 0.7191,
"step": 499
},
{
"epoch": 0.175561797752809,
"grad_norm": 0.92835932970047,
"learning_rate": 4.993163289438085e-06,
"loss": 0.7395,
"step": 500
},
{
"epoch": 0.1759129213483146,
"grad_norm": 0.9345782995223999,
"learning_rate": 4.993129078842176e-06,
"loss": 0.8146,
"step": 501
},
{
"epoch": 0.17626404494382023,
"grad_norm": 0.9650789499282837,
"learning_rate": 4.993094782983386e-06,
"loss": 0.7634,
"step": 502
},
{
"epoch": 0.17661516853932585,
"grad_norm": 0.9201986193656921,
"learning_rate": 4.993060401862888e-06,
"loss": 0.7476,
"step": 503
},
{
"epoch": 0.17696629213483145,
"grad_norm": 0.964146077632904,
"learning_rate": 4.993025935481858e-06,
"loss": 0.7822,
"step": 504
},
{
"epoch": 0.17731741573033707,
"grad_norm": 0.8705427646636963,
"learning_rate": 4.992991383841475e-06,
"loss": 0.6773,
"step": 505
},
{
"epoch": 0.1776685393258427,
"grad_norm": 0.9599820971488953,
"learning_rate": 4.992956746942919e-06,
"loss": 0.7516,
"step": 506
},
{
"epoch": 0.17801966292134833,
"grad_norm": 0.9984298944473267,
"learning_rate": 4.992922024787374e-06,
"loss": 0.7714,
"step": 507
},
{
"epoch": 0.17837078651685392,
"grad_norm": 0.9363481402397156,
"learning_rate": 4.992887217376032e-06,
"loss": 0.784,
"step": 508
},
{
"epoch": 0.17872191011235955,
"grad_norm": 0.8900370597839355,
"learning_rate": 4.992852324710078e-06,
"loss": 0.7749,
"step": 509
},
{
"epoch": 0.17907303370786518,
"grad_norm": 0.9664030075073242,
"learning_rate": 4.9928173467907085e-06,
"loss": 0.7635,
"step": 510
},
{
"epoch": 0.17942415730337077,
"grad_norm": 0.9362422227859497,
"learning_rate": 4.9927822836191185e-06,
"loss": 0.7497,
"step": 511
},
{
"epoch": 0.1797752808988764,
"grad_norm": 0.8915205001831055,
"learning_rate": 4.992747135196508e-06,
"loss": 0.7257,
"step": 512
},
{
"epoch": 0.18012640449438203,
"grad_norm": 0.9136397838592529,
"learning_rate": 4.992711901524079e-06,
"loss": 0.733,
"step": 513
},
{
"epoch": 0.18047752808988765,
"grad_norm": 0.9412914514541626,
"learning_rate": 4.992676582603035e-06,
"loss": 0.7757,
"step": 514
},
{
"epoch": 0.18082865168539325,
"grad_norm": 0.8885975480079651,
"learning_rate": 4.992641178434586e-06,
"loss": 0.7219,
"step": 515
},
{
"epoch": 0.18117977528089887,
"grad_norm": 0.950423538684845,
"learning_rate": 4.992605689019941e-06,
"loss": 0.7596,
"step": 516
},
{
"epoch": 0.1815308988764045,
"grad_norm": 0.9234071969985962,
"learning_rate": 4.992570114360314e-06,
"loss": 0.7624,
"step": 517
},
{
"epoch": 0.1818820224719101,
"grad_norm": 0.9259780049324036,
"learning_rate": 4.9925344544569225e-06,
"loss": 0.7755,
"step": 518
},
{
"epoch": 0.18223314606741572,
"grad_norm": 0.9503250122070312,
"learning_rate": 4.992498709310986e-06,
"loss": 0.7051,
"step": 519
},
{
"epoch": 0.18258426966292135,
"grad_norm": 0.9408488273620605,
"learning_rate": 4.992462878923725e-06,
"loss": 0.7281,
"step": 520
},
{
"epoch": 0.18293539325842698,
"grad_norm": 0.9426605105400085,
"learning_rate": 4.992426963296369e-06,
"loss": 0.7666,
"step": 521
},
{
"epoch": 0.18328651685393257,
"grad_norm": 0.9386591911315918,
"learning_rate": 4.992390962430142e-06,
"loss": 0.7352,
"step": 522
},
{
"epoch": 0.1836376404494382,
"grad_norm": 0.9154810309410095,
"learning_rate": 4.992354876326276e-06,
"loss": 0.7597,
"step": 523
},
{
"epoch": 0.18398876404494383,
"grad_norm": 0.936019241809845,
"learning_rate": 4.992318704986007e-06,
"loss": 0.7497,
"step": 524
},
{
"epoch": 0.18433988764044945,
"grad_norm": 0.9355201721191406,
"learning_rate": 4.99228244841057e-06,
"loss": 0.7868,
"step": 525
},
{
"epoch": 0.18469101123595505,
"grad_norm": 0.9210653305053711,
"learning_rate": 4.9922461066012075e-06,
"loss": 0.6998,
"step": 526
},
{
"epoch": 0.18504213483146068,
"grad_norm": 0.9298537969589233,
"learning_rate": 4.99220967955916e-06,
"loss": 0.7253,
"step": 527
},
{
"epoch": 0.1853932584269663,
"grad_norm": 0.9049412608146667,
"learning_rate": 4.9921731672856746e-06,
"loss": 0.7621,
"step": 528
},
{
"epoch": 0.1857443820224719,
"grad_norm": 0.9317775964736938,
"learning_rate": 4.992136569781999e-06,
"loss": 0.7393,
"step": 529
},
{
"epoch": 0.18609550561797752,
"grad_norm": 0.923850417137146,
"learning_rate": 4.9920998870493856e-06,
"loss": 0.7384,
"step": 530
},
{
"epoch": 0.18644662921348315,
"grad_norm": 0.9435550570487976,
"learning_rate": 4.992063119089088e-06,
"loss": 0.7811,
"step": 531
},
{
"epoch": 0.18679775280898878,
"grad_norm": 0.9249152541160583,
"learning_rate": 4.992026265902364e-06,
"loss": 0.7451,
"step": 532
},
{
"epoch": 0.18714887640449437,
"grad_norm": 0.9371122717857361,
"learning_rate": 4.991989327490475e-06,
"loss": 0.7507,
"step": 533
},
{
"epoch": 0.1875,
"grad_norm": 0.8902071714401245,
"learning_rate": 4.991952303854683e-06,
"loss": 0.7445,
"step": 534
},
{
"epoch": 0.18785112359550563,
"grad_norm": 0.8958204984664917,
"learning_rate": 4.991915194996254e-06,
"loss": 0.7029,
"step": 535
},
{
"epoch": 0.18820224719101122,
"grad_norm": 0.9265782237052917,
"learning_rate": 4.991878000916459e-06,
"loss": 0.7421,
"step": 536
},
{
"epoch": 0.18855337078651685,
"grad_norm": 0.8842232823371887,
"learning_rate": 4.991840721616568e-06,
"loss": 0.7678,
"step": 537
},
{
"epoch": 0.18890449438202248,
"grad_norm": 0.9105944037437439,
"learning_rate": 4.9918033570978565e-06,
"loss": 0.7257,
"step": 538
},
{
"epoch": 0.1892556179775281,
"grad_norm": 0.875582754611969,
"learning_rate": 4.991765907361603e-06,
"loss": 0.739,
"step": 539
},
{
"epoch": 0.1896067415730337,
"grad_norm": 0.90901780128479,
"learning_rate": 4.991728372409087e-06,
"loss": 0.7467,
"step": 540
},
{
"epoch": 0.18995786516853932,
"grad_norm": 0.9725927710533142,
"learning_rate": 4.991690752241594e-06,
"loss": 0.7779,
"step": 541
},
{
"epoch": 0.19030898876404495,
"grad_norm": 0.9324540495872498,
"learning_rate": 4.991653046860408e-06,
"loss": 0.7337,
"step": 542
},
{
"epoch": 0.19066011235955055,
"grad_norm": 0.929425060749054,
"learning_rate": 4.9916152562668205e-06,
"loss": 0.7432,
"step": 543
},
{
"epoch": 0.19101123595505617,
"grad_norm": 0.9117162227630615,
"learning_rate": 4.991577380462124e-06,
"loss": 0.7556,
"step": 544
},
{
"epoch": 0.1913623595505618,
"grad_norm": 0.9356319904327393,
"learning_rate": 4.9915394194476115e-06,
"loss": 0.783,
"step": 545
},
{
"epoch": 0.19171348314606743,
"grad_norm": 0.9007960557937622,
"learning_rate": 4.991501373224584e-06,
"loss": 0.7485,
"step": 546
},
{
"epoch": 0.19206460674157302,
"grad_norm": 0.9083287715911865,
"learning_rate": 4.991463241794342e-06,
"loss": 0.7497,
"step": 547
},
{
"epoch": 0.19241573033707865,
"grad_norm": 0.9226256608963013,
"learning_rate": 4.991425025158188e-06,
"loss": 0.7687,
"step": 548
},
{
"epoch": 0.19276685393258428,
"grad_norm": 0.8791014552116394,
"learning_rate": 4.991386723317432e-06,
"loss": 0.7542,
"step": 549
},
{
"epoch": 0.1931179775280899,
"grad_norm": 0.8970210552215576,
"learning_rate": 4.991348336273381e-06,
"loss": 0.7169,
"step": 550
},
{
"epoch": 0.1934691011235955,
"grad_norm": 0.9191828966140747,
"learning_rate": 4.9913098640273475e-06,
"loss": 0.7261,
"step": 551
},
{
"epoch": 0.19382022471910113,
"grad_norm": 0.9275765419006348,
"learning_rate": 4.991271306580649e-06,
"loss": 0.7469,
"step": 552
},
{
"epoch": 0.19417134831460675,
"grad_norm": 0.9179903864860535,
"learning_rate": 4.9912326639346045e-06,
"loss": 0.7385,
"step": 553
},
{
"epoch": 0.19452247191011235,
"grad_norm": 0.9225823283195496,
"learning_rate": 4.991193936090535e-06,
"loss": 0.6986,
"step": 554
},
{
"epoch": 0.19487359550561797,
"grad_norm": 0.9241051077842712,
"learning_rate": 4.991155123049764e-06,
"loss": 0.783,
"step": 555
},
{
"epoch": 0.1952247191011236,
"grad_norm": 0.9368206858634949,
"learning_rate": 4.991116224813619e-06,
"loss": 0.7712,
"step": 556
},
{
"epoch": 0.19557584269662923,
"grad_norm": 0.9148329496383667,
"learning_rate": 4.991077241383431e-06,
"loss": 0.8179,
"step": 557
},
{
"epoch": 0.19592696629213482,
"grad_norm": 0.8920343518257141,
"learning_rate": 4.991038172760533e-06,
"loss": 0.7271,
"step": 558
},
{
"epoch": 0.19627808988764045,
"grad_norm": 0.9275470972061157,
"learning_rate": 4.990999018946262e-06,
"loss": 0.7704,
"step": 559
},
{
"epoch": 0.19662921348314608,
"grad_norm": 0.9231950640678406,
"learning_rate": 4.990959779941955e-06,
"loss": 0.6871,
"step": 560
},
{
"epoch": 0.19698033707865167,
"grad_norm": 0.9179824590682983,
"learning_rate": 4.990920455748955e-06,
"loss": 0.7573,
"step": 561
},
{
"epoch": 0.1973314606741573,
"grad_norm": 0.9076899290084839,
"learning_rate": 4.9908810463686076e-06,
"loss": 0.7463,
"step": 562
},
{
"epoch": 0.19768258426966293,
"grad_norm": 0.9685149788856506,
"learning_rate": 4.99084155180226e-06,
"loss": 0.7207,
"step": 563
},
{
"epoch": 0.19803370786516855,
"grad_norm": 1.3696738481521606,
"learning_rate": 4.990801972051262e-06,
"loss": 0.768,
"step": 564
},
{
"epoch": 0.19838483146067415,
"grad_norm": 0.9174782633781433,
"learning_rate": 4.990762307116969e-06,
"loss": 0.7711,
"step": 565
},
{
"epoch": 0.19873595505617977,
"grad_norm": 0.9449765086174011,
"learning_rate": 4.990722557000736e-06,
"loss": 0.7267,
"step": 566
},
{
"epoch": 0.1990870786516854,
"grad_norm": 0.9291553497314453,
"learning_rate": 4.990682721703922e-06,
"loss": 0.7017,
"step": 567
},
{
"epoch": 0.199438202247191,
"grad_norm": 0.920346736907959,
"learning_rate": 4.9906428012278915e-06,
"loss": 0.7486,
"step": 568
},
{
"epoch": 0.19978932584269662,
"grad_norm": 0.9206670522689819,
"learning_rate": 4.990602795574007e-06,
"loss": 0.7833,
"step": 569
},
{
"epoch": 0.20014044943820225,
"grad_norm": 0.9318204522132874,
"learning_rate": 4.99056270474364e-06,
"loss": 0.7467,
"step": 570
},
{
"epoch": 0.20049157303370788,
"grad_norm": 0.9259383082389832,
"learning_rate": 4.990522528738159e-06,
"loss": 0.7196,
"step": 571
},
{
"epoch": 0.20084269662921347,
"grad_norm": 0.9336618185043335,
"learning_rate": 4.9904822675589385e-06,
"loss": 0.7498,
"step": 572
},
{
"epoch": 0.2011938202247191,
"grad_norm": 0.9120561480522156,
"learning_rate": 4.990441921207356e-06,
"loss": 0.7621,
"step": 573
},
{
"epoch": 0.20154494382022473,
"grad_norm": 0.9047495126724243,
"learning_rate": 4.99040148968479e-06,
"loss": 0.7103,
"step": 574
},
{
"epoch": 0.20189606741573032,
"grad_norm": 0.9809283018112183,
"learning_rate": 4.990360972992625e-06,
"loss": 0.7341,
"step": 575
},
{
"epoch": 0.20224719101123595,
"grad_norm": 1.0163681507110596,
"learning_rate": 4.990320371132245e-06,
"loss": 0.7219,
"step": 576
},
{
"epoch": 0.20259831460674158,
"grad_norm": 0.8804516792297363,
"learning_rate": 4.99027968410504e-06,
"loss": 0.718,
"step": 577
},
{
"epoch": 0.2029494382022472,
"grad_norm": 0.9172428846359253,
"learning_rate": 4.9902389119124005e-06,
"loss": 0.7361,
"step": 578
},
{
"epoch": 0.2033005617977528,
"grad_norm": 0.9390194416046143,
"learning_rate": 4.990198054555721e-06,
"loss": 0.7195,
"step": 579
},
{
"epoch": 0.20365168539325842,
"grad_norm": 0.9496874213218689,
"learning_rate": 4.990157112036399e-06,
"loss": 0.7305,
"step": 580
},
{
"epoch": 0.20400280898876405,
"grad_norm": 0.9417531490325928,
"learning_rate": 4.990116084355835e-06,
"loss": 0.7647,
"step": 581
},
{
"epoch": 0.20435393258426968,
"grad_norm": 0.8786060810089111,
"learning_rate": 4.99007497151543e-06,
"loss": 0.7654,
"step": 582
},
{
"epoch": 0.20470505617977527,
"grad_norm": 0.8952715992927551,
"learning_rate": 4.990033773516594e-06,
"loss": 0.7149,
"step": 583
},
{
"epoch": 0.2050561797752809,
"grad_norm": 0.9071550369262695,
"learning_rate": 4.989992490360734e-06,
"loss": 0.7368,
"step": 584
},
{
"epoch": 0.20540730337078653,
"grad_norm": 0.8937875628471375,
"learning_rate": 4.989951122049259e-06,
"loss": 0.7469,
"step": 585
},
{
"epoch": 0.20575842696629212,
"grad_norm": 0.9116646647453308,
"learning_rate": 4.989909668583588e-06,
"loss": 0.7348,
"step": 586
},
{
"epoch": 0.20610955056179775,
"grad_norm": 0.9020063877105713,
"learning_rate": 4.989868129965137e-06,
"loss": 0.6871,
"step": 587
},
{
"epoch": 0.20646067415730338,
"grad_norm": 0.8991970419883728,
"learning_rate": 4.989826506195326e-06,
"loss": 0.7136,
"step": 588
},
{
"epoch": 0.206811797752809,
"grad_norm": 0.9156064987182617,
"learning_rate": 4.9897847972755795e-06,
"loss": 0.7082,
"step": 589
},
{
"epoch": 0.2071629213483146,
"grad_norm": 0.9545648097991943,
"learning_rate": 4.989743003207323e-06,
"loss": 0.7628,
"step": 590
},
{
"epoch": 0.20751404494382023,
"grad_norm": 0.9077872633934021,
"learning_rate": 4.989701123991987e-06,
"loss": 0.7595,
"step": 591
},
{
"epoch": 0.20786516853932585,
"grad_norm": 0.9414958953857422,
"learning_rate": 4.989659159631003e-06,
"loss": 0.7604,
"step": 592
},
{
"epoch": 0.20821629213483145,
"grad_norm": 0.9313225150108337,
"learning_rate": 4.989617110125805e-06,
"loss": 0.7175,
"step": 593
},
{
"epoch": 0.20856741573033707,
"grad_norm": 0.9247497320175171,
"learning_rate": 4.989574975477833e-06,
"loss": 0.7497,
"step": 594
},
{
"epoch": 0.2089185393258427,
"grad_norm": 0.8851250410079956,
"learning_rate": 4.989532755688527e-06,
"loss": 0.665,
"step": 595
},
{
"epoch": 0.20926966292134833,
"grad_norm": 0.9276209473609924,
"learning_rate": 4.989490450759331e-06,
"loss": 0.6919,
"step": 596
},
{
"epoch": 0.20962078651685392,
"grad_norm": 0.9292905330657959,
"learning_rate": 4.9894480606916925e-06,
"loss": 0.7545,
"step": 597
},
{
"epoch": 0.20997191011235955,
"grad_norm": 0.9364456534385681,
"learning_rate": 4.98940558548706e-06,
"loss": 0.8026,
"step": 598
},
{
"epoch": 0.21032303370786518,
"grad_norm": 0.9156911969184875,
"learning_rate": 4.9893630251468874e-06,
"loss": 0.7279,
"step": 599
},
{
"epoch": 0.21067415730337077,
"grad_norm": 0.9316357374191284,
"learning_rate": 4.989320379672629e-06,
"loss": 0.7664,
"step": 600
},
{
"epoch": 0.2110252808988764,
"grad_norm": 0.945326566696167,
"learning_rate": 4.989277649065744e-06,
"loss": 0.7456,
"step": 601
},
{
"epoch": 0.21137640449438203,
"grad_norm": 0.9577007293701172,
"learning_rate": 4.989234833327693e-06,
"loss": 0.7097,
"step": 602
},
{
"epoch": 0.21172752808988765,
"grad_norm": 0.898679792881012,
"learning_rate": 4.989191932459941e-06,
"loss": 0.7085,
"step": 603
},
{
"epoch": 0.21207865168539325,
"grad_norm": 0.9334394931793213,
"learning_rate": 4.989148946463955e-06,
"loss": 0.7441,
"step": 604
},
{
"epoch": 0.21242977528089887,
"grad_norm": 0.9490911364555359,
"learning_rate": 4.989105875341206e-06,
"loss": 0.7571,
"step": 605
},
{
"epoch": 0.2127808988764045,
"grad_norm": 0.9073877930641174,
"learning_rate": 4.989062719093164e-06,
"loss": 0.748,
"step": 606
},
{
"epoch": 0.2131320224719101,
"grad_norm": 0.9007843136787415,
"learning_rate": 4.989019477721309e-06,
"loss": 0.7099,
"step": 607
},
{
"epoch": 0.21348314606741572,
"grad_norm": 0.9324343204498291,
"learning_rate": 4.988976151227116e-06,
"loss": 0.7688,
"step": 608
},
{
"epoch": 0.21383426966292135,
"grad_norm": 0.9213019013404846,
"learning_rate": 4.988932739612069e-06,
"loss": 0.7302,
"step": 609
},
{
"epoch": 0.21418539325842698,
"grad_norm": 0.9618836641311646,
"learning_rate": 4.988889242877653e-06,
"loss": 0.7484,
"step": 610
},
{
"epoch": 0.21453651685393257,
"grad_norm": 0.9354860186576843,
"learning_rate": 4.988845661025353e-06,
"loss": 0.7192,
"step": 611
},
{
"epoch": 0.2148876404494382,
"grad_norm": 0.9579604268074036,
"learning_rate": 4.988801994056663e-06,
"loss": 0.7727,
"step": 612
},
{
"epoch": 0.21523876404494383,
"grad_norm": 0.8972249627113342,
"learning_rate": 4.988758241973073e-06,
"loss": 0.7816,
"step": 613
},
{
"epoch": 0.21558988764044945,
"grad_norm": 0.9025394916534424,
"learning_rate": 4.988714404776082e-06,
"loss": 0.6994,
"step": 614
},
{
"epoch": 0.21594101123595505,
"grad_norm": 0.9159374833106995,
"learning_rate": 4.9886704824671874e-06,
"loss": 0.7319,
"step": 615
},
{
"epoch": 0.21629213483146068,
"grad_norm": 0.9703385829925537,
"learning_rate": 4.988626475047892e-06,
"loss": 0.7739,
"step": 616
},
{
"epoch": 0.2166432584269663,
"grad_norm": 0.9332022070884705,
"learning_rate": 4.9885823825197e-06,
"loss": 0.7209,
"step": 617
},
{
"epoch": 0.2169943820224719,
"grad_norm": 0.9919439554214478,
"learning_rate": 4.988538204884121e-06,
"loss": 0.7631,
"step": 618
},
{
"epoch": 0.21734550561797752,
"grad_norm": 0.9095845818519592,
"learning_rate": 4.988493942142664e-06,
"loss": 0.7311,
"step": 619
},
{
"epoch": 0.21769662921348315,
"grad_norm": 0.9250304698944092,
"learning_rate": 4.988449594296845e-06,
"loss": 0.7616,
"step": 620
},
{
"epoch": 0.21804775280898878,
"grad_norm": 0.9156744480133057,
"learning_rate": 4.988405161348178e-06,
"loss": 0.7446,
"step": 621
},
{
"epoch": 0.21839887640449437,
"grad_norm": 0.9540274143218994,
"learning_rate": 4.988360643298184e-06,
"loss": 0.745,
"step": 622
},
{
"epoch": 0.21875,
"grad_norm": 0.9059514999389648,
"learning_rate": 4.988316040148386e-06,
"loss": 0.7169,
"step": 623
},
{
"epoch": 0.21910112359550563,
"grad_norm": 0.8970401287078857,
"learning_rate": 4.988271351900308e-06,
"loss": 0.7551,
"step": 624
},
{
"epoch": 0.21945224719101122,
"grad_norm": 0.8862085938453674,
"learning_rate": 4.988226578555479e-06,
"loss": 0.7141,
"step": 625
},
{
"epoch": 0.21980337078651685,
"grad_norm": 0.9373227953910828,
"learning_rate": 4.98818172011543e-06,
"loss": 0.7637,
"step": 626
},
{
"epoch": 0.22015449438202248,
"grad_norm": 0.9211307168006897,
"learning_rate": 4.988136776581697e-06,
"loss": 0.7428,
"step": 627
},
{
"epoch": 0.2205056179775281,
"grad_norm": 0.9092309474945068,
"learning_rate": 4.9880917479558135e-06,
"loss": 0.7341,
"step": 628
},
{
"epoch": 0.2208567415730337,
"grad_norm": 0.9342793822288513,
"learning_rate": 4.988046634239322e-06,
"loss": 0.7397,
"step": 629
},
{
"epoch": 0.22120786516853932,
"grad_norm": 0.9277600646018982,
"learning_rate": 4.9880014354337645e-06,
"loss": 0.718,
"step": 630
},
{
"epoch": 0.22155898876404495,
"grad_norm": 1.1010304689407349,
"learning_rate": 4.987956151540687e-06,
"loss": 0.7199,
"step": 631
},
{
"epoch": 0.22191011235955055,
"grad_norm": 0.9733719229698181,
"learning_rate": 4.9879107825616375e-06,
"loss": 0.7518,
"step": 632
},
{
"epoch": 0.22226123595505617,
"grad_norm": 0.9306866526603699,
"learning_rate": 4.9878653284981694e-06,
"loss": 0.7388,
"step": 633
},
{
"epoch": 0.2226123595505618,
"grad_norm": 0.9351629614830017,
"learning_rate": 4.987819789351835e-06,
"loss": 0.7326,
"step": 634
},
{
"epoch": 0.22296348314606743,
"grad_norm": 0.9454215168952942,
"learning_rate": 4.987774165124194e-06,
"loss": 0.7929,
"step": 635
},
{
"epoch": 0.22331460674157302,
"grad_norm": 0.8752790093421936,
"learning_rate": 4.987728455816804e-06,
"loss": 0.7282,
"step": 636
},
{
"epoch": 0.22366573033707865,
"grad_norm": 0.9457923173904419,
"learning_rate": 4.98768266143123e-06,
"loss": 0.7297,
"step": 637
},
{
"epoch": 0.22401685393258428,
"grad_norm": 0.9447070360183716,
"learning_rate": 4.987636781969038e-06,
"loss": 0.7502,
"step": 638
},
{
"epoch": 0.2243679775280899,
"grad_norm": 1.4757858514785767,
"learning_rate": 4.987590817431797e-06,
"loss": 0.754,
"step": 639
},
{
"epoch": 0.2247191011235955,
"grad_norm": 0.9355869889259338,
"learning_rate": 4.987544767821078e-06,
"loss": 0.7468,
"step": 640
},
{
"epoch": 0.22507022471910113,
"grad_norm": 0.9021878242492676,
"learning_rate": 4.9874986331384566e-06,
"loss": 0.6914,
"step": 641
},
{
"epoch": 0.22542134831460675,
"grad_norm": 0.9627546072006226,
"learning_rate": 4.98745241338551e-06,
"loss": 0.6998,
"step": 642
},
{
"epoch": 0.22577247191011235,
"grad_norm": 0.9815264940261841,
"learning_rate": 4.98740610856382e-06,
"loss": 0.7392,
"step": 643
},
{
"epoch": 0.22612359550561797,
"grad_norm": 0.9428890943527222,
"learning_rate": 4.98735971867497e-06,
"loss": 0.737,
"step": 644
},
{
"epoch": 0.2264747191011236,
"grad_norm": 0.9130154252052307,
"learning_rate": 4.987313243720546e-06,
"loss": 0.7363,
"step": 645
},
{
"epoch": 0.22682584269662923,
"grad_norm": 0.9495055079460144,
"learning_rate": 4.987266683702137e-06,
"loss": 0.7104,
"step": 646
},
{
"epoch": 0.22717696629213482,
"grad_norm": 0.9111932516098022,
"learning_rate": 4.987220038621336e-06,
"loss": 0.7107,
"step": 647
},
{
"epoch": 0.22752808988764045,
"grad_norm": 0.9328299760818481,
"learning_rate": 4.987173308479738e-06,
"loss": 0.7294,
"step": 648
},
{
"epoch": 0.22787921348314608,
"grad_norm": 0.9401968717575073,
"learning_rate": 4.987126493278942e-06,
"loss": 0.7089,
"step": 649
},
{
"epoch": 0.22823033707865167,
"grad_norm": 0.9041705131530762,
"learning_rate": 4.987079593020547e-06,
"loss": 0.7249,
"step": 650
},
{
"epoch": 0.2285814606741573,
"grad_norm": 0.8919897675514221,
"learning_rate": 4.987032607706158e-06,
"loss": 0.6662,
"step": 651
},
{
"epoch": 0.22893258426966293,
"grad_norm": 0.9634642004966736,
"learning_rate": 4.986985537337384e-06,
"loss": 0.7547,
"step": 652
},
{
"epoch": 0.22928370786516855,
"grad_norm": 0.9561633467674255,
"learning_rate": 4.9869383819158305e-06,
"loss": 0.7448,
"step": 653
},
{
"epoch": 0.22963483146067415,
"grad_norm": 0.9113203287124634,
"learning_rate": 4.986891141443113e-06,
"loss": 0.7064,
"step": 654
},
{
"epoch": 0.22998595505617977,
"grad_norm": 0.8830083012580872,
"learning_rate": 4.986843815920847e-06,
"loss": 0.6909,
"step": 655
},
{
"epoch": 0.2303370786516854,
"grad_norm": 0.9515541791915894,
"learning_rate": 4.98679640535065e-06,
"loss": 0.7262,
"step": 656
},
{
"epoch": 0.230688202247191,
"grad_norm": 0.9469254612922668,
"learning_rate": 4.986748909734145e-06,
"loss": 0.7177,
"step": 657
},
{
"epoch": 0.23103932584269662,
"grad_norm": 0.9044115543365479,
"learning_rate": 4.986701329072954e-06,
"loss": 0.7371,
"step": 658
},
{
"epoch": 0.23139044943820225,
"grad_norm": 0.9210987687110901,
"learning_rate": 4.986653663368706e-06,
"loss": 0.7327,
"step": 659
},
{
"epoch": 0.23174157303370788,
"grad_norm": 0.8826014995574951,
"learning_rate": 4.98660591262303e-06,
"loss": 0.7455,
"step": 660
},
{
"epoch": 0.23209269662921347,
"grad_norm": 0.8898245096206665,
"learning_rate": 4.986558076837561e-06,
"loss": 0.7242,
"step": 661
},
{
"epoch": 0.2324438202247191,
"grad_norm": 0.9125939607620239,
"learning_rate": 4.9865101560139324e-06,
"loss": 0.7587,
"step": 662
},
{
"epoch": 0.23279494382022473,
"grad_norm": 0.9783604741096497,
"learning_rate": 4.9864621501537845e-06,
"loss": 0.7356,
"step": 663
},
{
"epoch": 0.23314606741573032,
"grad_norm": 0.9296882152557373,
"learning_rate": 4.986414059258758e-06,
"loss": 0.7059,
"step": 664
},
{
"epoch": 0.23349719101123595,
"grad_norm": 0.9341956973075867,
"learning_rate": 4.9863658833305e-06,
"loss": 0.7342,
"step": 665
},
{
"epoch": 0.23384831460674158,
"grad_norm": 0.9274594783782959,
"learning_rate": 4.986317622370655e-06,
"loss": 0.7254,
"step": 666
},
{
"epoch": 0.2341994382022472,
"grad_norm": 0.8959493637084961,
"learning_rate": 4.986269276380875e-06,
"loss": 0.7138,
"step": 667
},
{
"epoch": 0.2345505617977528,
"grad_norm": 0.9621686935424805,
"learning_rate": 4.986220845362814e-06,
"loss": 0.7301,
"step": 668
},
{
"epoch": 0.23490168539325842,
"grad_norm": 0.9666121602058411,
"learning_rate": 4.9861723293181275e-06,
"loss": 0.754,
"step": 669
},
{
"epoch": 0.23525280898876405,
"grad_norm": 0.9423872828483582,
"learning_rate": 4.986123728248474e-06,
"loss": 0.7464,
"step": 670
},
{
"epoch": 0.23560393258426968,
"grad_norm": 0.9208926558494568,
"learning_rate": 4.986075042155517e-06,
"loss": 0.7092,
"step": 671
},
{
"epoch": 0.23595505617977527,
"grad_norm": 0.880066454410553,
"learning_rate": 4.986026271040921e-06,
"loss": 0.6935,
"step": 672
},
{
"epoch": 0.2363061797752809,
"grad_norm": 0.9050423502922058,
"learning_rate": 4.985977414906353e-06,
"loss": 0.742,
"step": 673
},
{
"epoch": 0.23665730337078653,
"grad_norm": 0.9089376330375671,
"learning_rate": 4.985928473753486e-06,
"loss": 0.7366,
"step": 674
},
{
"epoch": 0.23700842696629212,
"grad_norm": 0.9702712893486023,
"learning_rate": 4.985879447583992e-06,
"loss": 0.7314,
"step": 675
},
{
"epoch": 0.23735955056179775,
"grad_norm": 0.9272444248199463,
"learning_rate": 4.985830336399548e-06,
"loss": 0.7156,
"step": 676
},
{
"epoch": 0.23771067415730338,
"grad_norm": 0.9623541831970215,
"learning_rate": 4.985781140201833e-06,
"loss": 0.705,
"step": 677
},
{
"epoch": 0.238061797752809,
"grad_norm": 0.9228547215461731,
"learning_rate": 4.985731858992531e-06,
"loss": 0.7254,
"step": 678
},
{
"epoch": 0.2384129213483146,
"grad_norm": 0.8912785649299622,
"learning_rate": 4.985682492773326e-06,
"loss": 0.7254,
"step": 679
},
{
"epoch": 0.23876404494382023,
"grad_norm": 0.9913116693496704,
"learning_rate": 4.985633041545906e-06,
"loss": 0.7172,
"step": 680
},
{
"epoch": 0.23911516853932585,
"grad_norm": 0.9851749539375305,
"learning_rate": 4.985583505311965e-06,
"loss": 0.7711,
"step": 681
},
{
"epoch": 0.23946629213483145,
"grad_norm": 0.9252983331680298,
"learning_rate": 4.9855338840731934e-06,
"loss": 0.7356,
"step": 682
},
{
"epoch": 0.23981741573033707,
"grad_norm": 0.9541698694229126,
"learning_rate": 4.9854841778312905e-06,
"loss": 0.7111,
"step": 683
},
{
"epoch": 0.2401685393258427,
"grad_norm": 0.9459705948829651,
"learning_rate": 4.985434386587956e-06,
"loss": 0.7254,
"step": 684
},
{
"epoch": 0.24051966292134833,
"grad_norm": 1.0094040632247925,
"learning_rate": 4.985384510344892e-06,
"loss": 0.7638,
"step": 685
},
{
"epoch": 0.24087078651685392,
"grad_norm": 0.9937487244606018,
"learning_rate": 4.985334549103804e-06,
"loss": 0.7373,
"step": 686
},
{
"epoch": 0.24122191011235955,
"grad_norm": 0.9358577132225037,
"learning_rate": 4.985284502866401e-06,
"loss": 0.697,
"step": 687
},
{
"epoch": 0.24157303370786518,
"grad_norm": 0.9751777052879333,
"learning_rate": 4.985234371634395e-06,
"loss": 0.7626,
"step": 688
},
{
"epoch": 0.24192415730337077,
"grad_norm": 0.9775291085243225,
"learning_rate": 4.9851841554095e-06,
"loss": 0.7031,
"step": 689
},
{
"epoch": 0.2422752808988764,
"grad_norm": 0.9366273880004883,
"learning_rate": 4.9851338541934345e-06,
"loss": 0.7268,
"step": 690
},
{
"epoch": 0.24262640449438203,
"grad_norm": 0.8989854454994202,
"learning_rate": 4.985083467987917e-06,
"loss": 0.7102,
"step": 691
},
{
"epoch": 0.24297752808988765,
"grad_norm": 0.9326997399330139,
"learning_rate": 4.9850329967946715e-06,
"loss": 0.6963,
"step": 692
},
{
"epoch": 0.24332865168539325,
"grad_norm": 0.9004776477813721,
"learning_rate": 4.984982440615423e-06,
"loss": 0.7133,
"step": 693
},
{
"epoch": 0.24367977528089887,
"grad_norm": 0.915814220905304,
"learning_rate": 4.984931799451903e-06,
"loss": 0.7197,
"step": 694
},
{
"epoch": 0.2440308988764045,
"grad_norm": 0.9562146067619324,
"learning_rate": 4.984881073305841e-06,
"loss": 0.7175,
"step": 695
},
{
"epoch": 0.2443820224719101,
"grad_norm": 0.9829107522964478,
"learning_rate": 4.984830262178974e-06,
"loss": 0.7343,
"step": 696
},
{
"epoch": 0.24473314606741572,
"grad_norm": 0.9095830917358398,
"learning_rate": 4.984779366073036e-06,
"loss": 0.7474,
"step": 697
},
{
"epoch": 0.24508426966292135,
"grad_norm": 0.9279952049255371,
"learning_rate": 4.984728384989772e-06,
"loss": 0.7574,
"step": 698
},
{
"epoch": 0.24543539325842698,
"grad_norm": 1.07504141330719,
"learning_rate": 4.984677318930922e-06,
"loss": 0.7309,
"step": 699
},
{
"epoch": 0.24578651685393257,
"grad_norm": 0.9406894445419312,
"learning_rate": 4.984626167898235e-06,
"loss": 0.7187,
"step": 700
},
{
"epoch": 0.2461376404494382,
"grad_norm": 0.9229064583778381,
"learning_rate": 4.9845749318934575e-06,
"loss": 0.7379,
"step": 701
},
{
"epoch": 0.24648876404494383,
"grad_norm": 0.9469361901283264,
"learning_rate": 4.984523610918345e-06,
"loss": 0.7585,
"step": 702
},
{
"epoch": 0.24683988764044945,
"grad_norm": 0.9020796418190002,
"learning_rate": 4.984472204974651e-06,
"loss": 0.714,
"step": 703
},
{
"epoch": 0.24719101123595505,
"grad_norm": 0.9876411557197571,
"learning_rate": 4.9844207140641336e-06,
"loss": 0.7013,
"step": 704
},
{
"epoch": 0.24754213483146068,
"grad_norm": 0.9139506816864014,
"learning_rate": 4.984369138188553e-06,
"loss": 0.7252,
"step": 705
},
{
"epoch": 0.2478932584269663,
"grad_norm": 0.9223290681838989,
"learning_rate": 4.9843174773496735e-06,
"loss": 0.7603,
"step": 706
},
{
"epoch": 0.2482443820224719,
"grad_norm": 0.9541411995887756,
"learning_rate": 4.984265731549263e-06,
"loss": 0.6997,
"step": 707
},
{
"epoch": 0.24859550561797752,
"grad_norm": 0.9302672147750854,
"learning_rate": 4.98421390078909e-06,
"loss": 0.7252,
"step": 708
},
{
"epoch": 0.24894662921348315,
"grad_norm": 0.9554900527000427,
"learning_rate": 4.984161985070927e-06,
"loss": 0.7305,
"step": 709
},
{
"epoch": 0.24929775280898878,
"grad_norm": 0.9729623794555664,
"learning_rate": 4.98410998439655e-06,
"loss": 0.7318,
"step": 710
},
{
"epoch": 0.24964887640449437,
"grad_norm": 0.9298332929611206,
"learning_rate": 4.984057898767738e-06,
"loss": 0.7348,
"step": 711
},
{
"epoch": 0.25,
"grad_norm": 0.9653823375701904,
"learning_rate": 4.984005728186271e-06,
"loss": 0.7566,
"step": 712
},
{
"epoch": 0.2503511235955056,
"grad_norm": 0.9283461570739746,
"learning_rate": 4.983953472653933e-06,
"loss": 0.6999,
"step": 713
},
{
"epoch": 0.25070224719101125,
"grad_norm": 0.9600170254707336,
"learning_rate": 4.983901132172513e-06,
"loss": 0.7378,
"step": 714
},
{
"epoch": 0.25105337078651685,
"grad_norm": 0.9261863827705383,
"learning_rate": 4.983848706743799e-06,
"loss": 0.7245,
"step": 715
},
{
"epoch": 0.25140449438202245,
"grad_norm": 0.9267969727516174,
"learning_rate": 4.983796196369585e-06,
"loss": 0.6921,
"step": 716
},
{
"epoch": 0.2517556179775281,
"grad_norm": 0.9277015328407288,
"learning_rate": 4.983743601051666e-06,
"loss": 0.715,
"step": 717
},
{
"epoch": 0.2521067415730337,
"grad_norm": 0.9029396176338196,
"learning_rate": 4.9836909207918415e-06,
"loss": 0.7739,
"step": 718
},
{
"epoch": 0.25245786516853935,
"grad_norm": 0.9498785138130188,
"learning_rate": 4.983638155591913e-06,
"loss": 0.7071,
"step": 719
},
{
"epoch": 0.25280898876404495,
"grad_norm": 0.934562087059021,
"learning_rate": 4.983585305453685e-06,
"loss": 0.7183,
"step": 720
},
{
"epoch": 0.25316011235955055,
"grad_norm": 0.9572330713272095,
"learning_rate": 4.983532370378964e-06,
"loss": 0.7535,
"step": 721
},
{
"epoch": 0.2535112359550562,
"grad_norm": 0.9443022608757019,
"learning_rate": 4.983479350369562e-06,
"loss": 0.7127,
"step": 722
},
{
"epoch": 0.2538623595505618,
"grad_norm": 0.9258742928504944,
"learning_rate": 4.983426245427291e-06,
"loss": 0.7002,
"step": 723
},
{
"epoch": 0.2542134831460674,
"grad_norm": 0.939900279045105,
"learning_rate": 4.983373055553968e-06,
"loss": 0.7234,
"step": 724
},
{
"epoch": 0.25456460674157305,
"grad_norm": 0.9580402374267578,
"learning_rate": 4.983319780751411e-06,
"loss": 0.7519,
"step": 725
},
{
"epoch": 0.25491573033707865,
"grad_norm": 0.9556792974472046,
"learning_rate": 4.983266421021442e-06,
"loss": 0.7701,
"step": 726
},
{
"epoch": 0.25526685393258425,
"grad_norm": 0.9831913709640503,
"learning_rate": 4.983212976365887e-06,
"loss": 0.7189,
"step": 727
},
{
"epoch": 0.2556179775280899,
"grad_norm": 0.9363849759101868,
"learning_rate": 4.983159446786573e-06,
"loss": 0.7061,
"step": 728
},
{
"epoch": 0.2559691011235955,
"grad_norm": 0.9291892647743225,
"learning_rate": 4.983105832285331e-06,
"loss": 0.6942,
"step": 729
},
{
"epoch": 0.2563202247191011,
"grad_norm": 0.9837853908538818,
"learning_rate": 4.983052132863995e-06,
"loss": 0.7251,
"step": 730
},
{
"epoch": 0.25667134831460675,
"grad_norm": 0.9630160331726074,
"learning_rate": 4.9829983485244e-06,
"loss": 0.694,
"step": 731
},
{
"epoch": 0.25702247191011235,
"grad_norm": 0.972981333732605,
"learning_rate": 4.982944479268386e-06,
"loss": 0.7232,
"step": 732
},
{
"epoch": 0.257373595505618,
"grad_norm": 0.9112554788589478,
"learning_rate": 4.9828905250977955e-06,
"loss": 0.7553,
"step": 733
},
{
"epoch": 0.2577247191011236,
"grad_norm": 0.9407148361206055,
"learning_rate": 4.982836486014474e-06,
"loss": 0.7931,
"step": 734
},
{
"epoch": 0.2580758426966292,
"grad_norm": 0.9683741927146912,
"learning_rate": 4.98278236202027e-06,
"loss": 0.76,
"step": 735
},
{
"epoch": 0.25842696629213485,
"grad_norm": 0.9538892507553101,
"learning_rate": 4.9827281531170325e-06,
"loss": 0.7491,
"step": 736
},
{
"epoch": 0.25877808988764045,
"grad_norm": 0.9792529344558716,
"learning_rate": 4.982673859306617e-06,
"loss": 0.752,
"step": 737
},
{
"epoch": 0.25912921348314605,
"grad_norm": 0.9384421110153198,
"learning_rate": 4.98261948059088e-06,
"loss": 0.7587,
"step": 738
},
{
"epoch": 0.2594803370786517,
"grad_norm": 0.9048983454704285,
"learning_rate": 4.982565016971682e-06,
"loss": 0.7429,
"step": 739
},
{
"epoch": 0.2598314606741573,
"grad_norm": 0.893500804901123,
"learning_rate": 4.982510468450884e-06,
"loss": 0.6985,
"step": 740
},
{
"epoch": 0.2601825842696629,
"grad_norm": 0.9900831580162048,
"learning_rate": 4.982455835030352e-06,
"loss": 0.7336,
"step": 741
},
{
"epoch": 0.26053370786516855,
"grad_norm": 0.9558880925178528,
"learning_rate": 4.982401116711955e-06,
"loss": 0.7561,
"step": 742
},
{
"epoch": 0.26088483146067415,
"grad_norm": 0.8868340849876404,
"learning_rate": 4.982346313497564e-06,
"loss": 0.6967,
"step": 743
},
{
"epoch": 0.2612359550561798,
"grad_norm": 0.9625622034072876,
"learning_rate": 4.9822914253890535e-06,
"loss": 0.7382,
"step": 744
},
{
"epoch": 0.2615870786516854,
"grad_norm": 0.9891008138656616,
"learning_rate": 4.9822364523883e-06,
"loss": 0.7316,
"step": 745
},
{
"epoch": 0.261938202247191,
"grad_norm": 0.9161337018013,
"learning_rate": 4.982181394497184e-06,
"loss": 0.7225,
"step": 746
},
{
"epoch": 0.26228932584269665,
"grad_norm": 0.9532581567764282,
"learning_rate": 4.982126251717589e-06,
"loss": 0.7412,
"step": 747
},
{
"epoch": 0.26264044943820225,
"grad_norm": 0.90705406665802,
"learning_rate": 4.982071024051399e-06,
"loss": 0.6938,
"step": 748
},
{
"epoch": 0.26299157303370785,
"grad_norm": 0.9263889789581299,
"learning_rate": 4.982015711500505e-06,
"loss": 0.7575,
"step": 749
},
{
"epoch": 0.2633426966292135,
"grad_norm": 0.9247738718986511,
"learning_rate": 4.981960314066797e-06,
"loss": 0.7184,
"step": 750
},
{
"epoch": 0.2636938202247191,
"grad_norm": 0.9485296607017517,
"learning_rate": 4.98190483175217e-06,
"loss": 0.7297,
"step": 751
},
{
"epoch": 0.2640449438202247,
"grad_norm": 0.9109437465667725,
"learning_rate": 4.981849264558523e-06,
"loss": 0.7422,
"step": 752
},
{
"epoch": 0.26439606741573035,
"grad_norm": 0.9430893063545227,
"learning_rate": 4.981793612487753e-06,
"loss": 0.7634,
"step": 753
},
{
"epoch": 0.26474719101123595,
"grad_norm": 0.9232211709022522,
"learning_rate": 4.981737875541765e-06,
"loss": 0.7587,
"step": 754
},
{
"epoch": 0.26509831460674155,
"grad_norm": 0.8920214176177979,
"learning_rate": 4.9816820537224675e-06,
"loss": 0.7222,
"step": 755
},
{
"epoch": 0.2654494382022472,
"grad_norm": 0.9339509010314941,
"learning_rate": 4.981626147031765e-06,
"loss": 0.7253,
"step": 756
},
{
"epoch": 0.2658005617977528,
"grad_norm": 0.890581488609314,
"learning_rate": 4.981570155471573e-06,
"loss": 0.7166,
"step": 757
},
{
"epoch": 0.26615168539325845,
"grad_norm": 0.9297827482223511,
"learning_rate": 4.981514079043806e-06,
"loss": 0.713,
"step": 758
},
{
"epoch": 0.26650280898876405,
"grad_norm": 0.9213199019432068,
"learning_rate": 4.98145791775038e-06,
"loss": 0.7312,
"step": 759
},
{
"epoch": 0.26685393258426965,
"grad_norm": 0.9185107946395874,
"learning_rate": 4.981401671593216e-06,
"loss": 0.7215,
"step": 760
},
{
"epoch": 0.2672050561797753,
"grad_norm": 0.8874849081039429,
"learning_rate": 4.981345340574239e-06,
"loss": 0.7082,
"step": 761
},
{
"epoch": 0.2675561797752809,
"grad_norm": 0.9179455637931824,
"learning_rate": 4.981288924695374e-06,
"loss": 0.7243,
"step": 762
},
{
"epoch": 0.2679073033707865,
"grad_norm": 0.8956063985824585,
"learning_rate": 4.981232423958552e-06,
"loss": 0.704,
"step": 763
},
{
"epoch": 0.26825842696629215,
"grad_norm": 0.9451838731765747,
"learning_rate": 4.981175838365704e-06,
"loss": 0.7518,
"step": 764
},
{
"epoch": 0.26860955056179775,
"grad_norm": 0.9265857338905334,
"learning_rate": 4.981119167918765e-06,
"loss": 0.7143,
"step": 765
},
{
"epoch": 0.26896067415730335,
"grad_norm": 0.9493303298950195,
"learning_rate": 4.981062412619674e-06,
"loss": 0.7172,
"step": 766
},
{
"epoch": 0.269311797752809,
"grad_norm": 0.9223463535308838,
"learning_rate": 4.981005572470372e-06,
"loss": 0.7383,
"step": 767
},
{
"epoch": 0.2696629213483146,
"grad_norm": 0.9502999782562256,
"learning_rate": 4.980948647472802e-06,
"loss": 0.7137,
"step": 768
},
{
"epoch": 0.2700140449438202,
"grad_norm": 0.9198105335235596,
"learning_rate": 4.980891637628911e-06,
"loss": 0.706,
"step": 769
},
{
"epoch": 0.27036516853932585,
"grad_norm": 0.9435210227966309,
"learning_rate": 4.980834542940649e-06,
"loss": 0.6921,
"step": 770
},
{
"epoch": 0.27071629213483145,
"grad_norm": 0.9600005745887756,
"learning_rate": 4.980777363409969e-06,
"loss": 0.72,
"step": 771
},
{
"epoch": 0.2710674157303371,
"grad_norm": 0.9591335654258728,
"learning_rate": 4.980720099038825e-06,
"loss": 0.7386,
"step": 772
},
{
"epoch": 0.2714185393258427,
"grad_norm": 0.9055672883987427,
"learning_rate": 4.980662749829177e-06,
"loss": 0.7568,
"step": 773
},
{
"epoch": 0.2717696629213483,
"grad_norm": 0.9077644348144531,
"learning_rate": 4.9806053157829855e-06,
"loss": 0.7122,
"step": 774
},
{
"epoch": 0.27212078651685395,
"grad_norm": 0.9248702526092529,
"learning_rate": 4.980547796902216e-06,
"loss": 0.6866,
"step": 775
},
{
"epoch": 0.27247191011235955,
"grad_norm": 0.9374347925186157,
"learning_rate": 4.980490193188834e-06,
"loss": 0.7283,
"step": 776
},
{
"epoch": 0.27282303370786515,
"grad_norm": 0.9280509352684021,
"learning_rate": 4.980432504644809e-06,
"loss": 0.6896,
"step": 777
},
{
"epoch": 0.2731741573033708,
"grad_norm": 0.9158201813697815,
"learning_rate": 4.980374731272115e-06,
"loss": 0.7222,
"step": 778
},
{
"epoch": 0.2735252808988764,
"grad_norm": 0.918114960193634,
"learning_rate": 4.980316873072729e-06,
"loss": 0.7528,
"step": 779
},
{
"epoch": 0.273876404494382,
"grad_norm": 0.891121506690979,
"learning_rate": 4.980258930048627e-06,
"loss": 0.7265,
"step": 780
},
{
"epoch": 0.27422752808988765,
"grad_norm": 0.9570785760879517,
"learning_rate": 4.980200902201792e-06,
"loss": 0.6957,
"step": 781
},
{
"epoch": 0.27457865168539325,
"grad_norm": 0.8888586759567261,
"learning_rate": 4.980142789534209e-06,
"loss": 0.7258,
"step": 782
},
{
"epoch": 0.2749297752808989,
"grad_norm": 0.9602726101875305,
"learning_rate": 4.980084592047866e-06,
"loss": 0.741,
"step": 783
},
{
"epoch": 0.2752808988764045,
"grad_norm": 0.9402101635932922,
"learning_rate": 4.9800263097447505e-06,
"loss": 0.7789,
"step": 784
},
{
"epoch": 0.2756320224719101,
"grad_norm": 0.916023850440979,
"learning_rate": 4.9799679426268575e-06,
"loss": 0.6746,
"step": 785
},
{
"epoch": 0.27598314606741575,
"grad_norm": 1.0306038856506348,
"learning_rate": 4.979909490696184e-06,
"loss": 0.7396,
"step": 786
},
{
"epoch": 0.27633426966292135,
"grad_norm": 0.9112997651100159,
"learning_rate": 4.979850953954727e-06,
"loss": 0.7225,
"step": 787
},
{
"epoch": 0.27668539325842695,
"grad_norm": 0.9124306440353394,
"learning_rate": 4.97979233240449e-06,
"loss": 0.7244,
"step": 788
},
{
"epoch": 0.2770365168539326,
"grad_norm": 0.8682541847229004,
"learning_rate": 4.979733626047477e-06,
"loss": 0.6983,
"step": 789
},
{
"epoch": 0.2773876404494382,
"grad_norm": 0.9099353551864624,
"learning_rate": 4.979674834885696e-06,
"loss": 0.723,
"step": 790
},
{
"epoch": 0.2777387640449438,
"grad_norm": 0.9057777523994446,
"learning_rate": 4.979615958921158e-06,
"loss": 0.7063,
"step": 791
},
{
"epoch": 0.27808988764044945,
"grad_norm": 0.9777807593345642,
"learning_rate": 4.979556998155876e-06,
"loss": 0.7047,
"step": 792
},
{
"epoch": 0.27844101123595505,
"grad_norm": 0.952314019203186,
"learning_rate": 4.979497952591865e-06,
"loss": 0.7589,
"step": 793
},
{
"epoch": 0.27879213483146065,
"grad_norm": 0.8948432207107544,
"learning_rate": 4.979438822231147e-06,
"loss": 0.6893,
"step": 794
},
{
"epoch": 0.2791432584269663,
"grad_norm": 0.9998204708099365,
"learning_rate": 4.979379607075743e-06,
"loss": 0.7176,
"step": 795
},
{
"epoch": 0.2794943820224719,
"grad_norm": 0.9392699003219604,
"learning_rate": 4.979320307127678e-06,
"loss": 0.7607,
"step": 796
},
{
"epoch": 0.27984550561797755,
"grad_norm": 0.916340708732605,
"learning_rate": 4.97926092238898e-06,
"loss": 0.6797,
"step": 797
},
{
"epoch": 0.28019662921348315,
"grad_norm": 0.9007791876792908,
"learning_rate": 4.979201452861679e-06,
"loss": 0.7015,
"step": 798
},
{
"epoch": 0.28054775280898875,
"grad_norm": 0.9042453765869141,
"learning_rate": 4.979141898547811e-06,
"loss": 0.7113,
"step": 799
},
{
"epoch": 0.2808988764044944,
"grad_norm": 0.9455578327178955,
"learning_rate": 4.9790822594494116e-06,
"loss": 0.7031,
"step": 800
},
{
"epoch": 0.28125,
"grad_norm": 0.9787691235542297,
"learning_rate": 4.979022535568519e-06,
"loss": 0.7328,
"step": 801
},
{
"epoch": 0.2816011235955056,
"grad_norm": 0.9504345059394836,
"learning_rate": 4.978962726907179e-06,
"loss": 0.7301,
"step": 802
},
{
"epoch": 0.28195224719101125,
"grad_norm": 0.9304360747337341,
"learning_rate": 4.978902833467434e-06,
"loss": 0.7195,
"step": 803
},
{
"epoch": 0.28230337078651685,
"grad_norm": 0.9509701132774353,
"learning_rate": 4.978842855251333e-06,
"loss": 0.7532,
"step": 804
},
{
"epoch": 0.28265449438202245,
"grad_norm": 0.9462277889251709,
"learning_rate": 4.9787827922609276e-06,
"loss": 0.708,
"step": 805
},
{
"epoch": 0.2830056179775281,
"grad_norm": 0.9095754027366638,
"learning_rate": 4.978722644498272e-06,
"loss": 0.7484,
"step": 806
},
{
"epoch": 0.2833567415730337,
"grad_norm": 0.8992451429367065,
"learning_rate": 4.978662411965424e-06,
"loss": 0.7033,
"step": 807
},
{
"epoch": 0.28370786516853935,
"grad_norm": 0.913910984992981,
"learning_rate": 4.978602094664441e-06,
"loss": 0.7333,
"step": 808
},
{
"epoch": 0.28405898876404495,
"grad_norm": 0.9041484594345093,
"learning_rate": 4.978541692597388e-06,
"loss": 0.7196,
"step": 809
},
{
"epoch": 0.28441011235955055,
"grad_norm": 0.9570571184158325,
"learning_rate": 4.97848120576633e-06,
"loss": 0.7199,
"step": 810
},
{
"epoch": 0.2847612359550562,
"grad_norm": 0.9184895753860474,
"learning_rate": 4.978420634173336e-06,
"loss": 0.6982,
"step": 811
},
{
"epoch": 0.2851123595505618,
"grad_norm": 0.9278954267501831,
"learning_rate": 4.978359977820476e-06,
"loss": 0.7475,
"step": 812
},
{
"epoch": 0.2854634831460674,
"grad_norm": 0.9236921668052673,
"learning_rate": 4.978299236709826e-06,
"loss": 0.7343,
"step": 813
},
{
"epoch": 0.28581460674157305,
"grad_norm": 0.9183304309844971,
"learning_rate": 4.978238410843464e-06,
"loss": 0.7111,
"step": 814
},
{
"epoch": 0.28616573033707865,
"grad_norm": 0.895050585269928,
"learning_rate": 4.978177500223468e-06,
"loss": 0.6746,
"step": 815
},
{
"epoch": 0.28651685393258425,
"grad_norm": 0.9193557500839233,
"learning_rate": 4.9781165048519216e-06,
"loss": 0.7025,
"step": 816
},
{
"epoch": 0.2868679775280899,
"grad_norm": 0.9046199917793274,
"learning_rate": 4.978055424730912e-06,
"loss": 0.701,
"step": 817
},
{
"epoch": 0.2872191011235955,
"grad_norm": 0.9200829267501831,
"learning_rate": 4.977994259862527e-06,
"loss": 0.755,
"step": 818
},
{
"epoch": 0.2875702247191011,
"grad_norm": 0.929964542388916,
"learning_rate": 4.977933010248859e-06,
"loss": 0.7274,
"step": 819
},
{
"epoch": 0.28792134831460675,
"grad_norm": 0.9332923293113708,
"learning_rate": 4.977871675892003e-06,
"loss": 0.7503,
"step": 820
},
{
"epoch": 0.28827247191011235,
"grad_norm": 0.8993156552314758,
"learning_rate": 4.977810256794055e-06,
"loss": 0.6954,
"step": 821
},
{
"epoch": 0.288623595505618,
"grad_norm": 0.8985475301742554,
"learning_rate": 4.977748752957116e-06,
"loss": 0.7287,
"step": 822
},
{
"epoch": 0.2889747191011236,
"grad_norm": 0.9083667397499084,
"learning_rate": 4.977687164383291e-06,
"loss": 0.7352,
"step": 823
},
{
"epoch": 0.2893258426966292,
"grad_norm": 0.9311912059783936,
"learning_rate": 4.977625491074686e-06,
"loss": 0.7199,
"step": 824
},
{
"epoch": 0.28967696629213485,
"grad_norm": 0.8959439396858215,
"learning_rate": 4.977563733033408e-06,
"loss": 0.6992,
"step": 825
},
{
"epoch": 0.29002808988764045,
"grad_norm": 1.018567681312561,
"learning_rate": 4.9775018902615715e-06,
"loss": 0.7278,
"step": 826
},
{
"epoch": 0.29037921348314605,
"grad_norm": 0.9925201535224915,
"learning_rate": 4.977439962761289e-06,
"loss": 0.7423,
"step": 827
},
{
"epoch": 0.2907303370786517,
"grad_norm": 1.0001310110092163,
"learning_rate": 4.977377950534681e-06,
"loss": 0.7395,
"step": 828
},
{
"epoch": 0.2910814606741573,
"grad_norm": 0.9196740984916687,
"learning_rate": 4.977315853583866e-06,
"loss": 0.7588,
"step": 829
},
{
"epoch": 0.2914325842696629,
"grad_norm": 0.9672794342041016,
"learning_rate": 4.977253671910969e-06,
"loss": 0.7378,
"step": 830
},
{
"epoch": 0.29178370786516855,
"grad_norm": 1.0301952362060547,
"learning_rate": 4.977191405518116e-06,
"loss": 0.7757,
"step": 831
},
{
"epoch": 0.29213483146067415,
"grad_norm": 0.9453630447387695,
"learning_rate": 4.977129054407437e-06,
"loss": 0.7335,
"step": 832
},
{
"epoch": 0.2924859550561798,
"grad_norm": 0.9876764416694641,
"learning_rate": 4.977066618581065e-06,
"loss": 0.7437,
"step": 833
},
{
"epoch": 0.2928370786516854,
"grad_norm": 0.9211217164993286,
"learning_rate": 4.977004098041133e-06,
"loss": 0.7333,
"step": 834
},
{
"epoch": 0.293188202247191,
"grad_norm": 0.9148357510566711,
"learning_rate": 4.976941492789781e-06,
"loss": 0.677,
"step": 835
},
{
"epoch": 0.29353932584269665,
"grad_norm": 0.9743471741676331,
"learning_rate": 4.976878802829149e-06,
"loss": 0.7586,
"step": 836
},
{
"epoch": 0.29389044943820225,
"grad_norm": 0.9516198039054871,
"learning_rate": 4.976816028161382e-06,
"loss": 0.7364,
"step": 837
},
{
"epoch": 0.29424157303370785,
"grad_norm": 0.8972299098968506,
"learning_rate": 4.9767531687886265e-06,
"loss": 0.6912,
"step": 838
},
{
"epoch": 0.2945926966292135,
"grad_norm": 0.8950793147087097,
"learning_rate": 4.976690224713031e-06,
"loss": 0.6969,
"step": 839
},
{
"epoch": 0.2949438202247191,
"grad_norm": 0.9539402723312378,
"learning_rate": 4.976627195936749e-06,
"loss": 0.7348,
"step": 840
},
{
"epoch": 0.2952949438202247,
"grad_norm": 1.32026207447052,
"learning_rate": 4.976564082461938e-06,
"loss": 0.7192,
"step": 841
},
{
"epoch": 0.29564606741573035,
"grad_norm": 0.9319935441017151,
"learning_rate": 4.976500884290753e-06,
"loss": 0.6957,
"step": 842
},
{
"epoch": 0.29599719101123595,
"grad_norm": 0.9559399485588074,
"learning_rate": 4.976437601425358e-06,
"loss": 0.7394,
"step": 843
},
{
"epoch": 0.29634831460674155,
"grad_norm": 0.9115216135978699,
"learning_rate": 4.976374233867915e-06,
"loss": 0.6955,
"step": 844
},
{
"epoch": 0.2966994382022472,
"grad_norm": 0.8987361788749695,
"learning_rate": 4.976310781620592e-06,
"loss": 0.7057,
"step": 845
},
{
"epoch": 0.2970505617977528,
"grad_norm": 0.9885380268096924,
"learning_rate": 4.976247244685561e-06,
"loss": 0.7542,
"step": 846
},
{
"epoch": 0.29740168539325845,
"grad_norm": 0.9176011681556702,
"learning_rate": 4.9761836230649914e-06,
"loss": 0.6836,
"step": 847
},
{
"epoch": 0.29775280898876405,
"grad_norm": 0.9235329031944275,
"learning_rate": 4.976119916761063e-06,
"loss": 0.7541,
"step": 848
},
{
"epoch": 0.29810393258426965,
"grad_norm": 0.9174650311470032,
"learning_rate": 4.9760561257759504e-06,
"loss": 0.7149,
"step": 849
},
{
"epoch": 0.2984550561797753,
"grad_norm": 0.9441576600074768,
"learning_rate": 4.975992250111837e-06,
"loss": 0.7222,
"step": 850
},
{
"epoch": 0.2988061797752809,
"grad_norm": 0.9702540636062622,
"learning_rate": 4.975928289770909e-06,
"loss": 0.7193,
"step": 851
},
{
"epoch": 0.2991573033707865,
"grad_norm": 0.9554623961448669,
"learning_rate": 4.9758642447553505e-06,
"loss": 0.7537,
"step": 852
},
{
"epoch": 0.29950842696629215,
"grad_norm": 0.9210932850837708,
"learning_rate": 4.975800115067355e-06,
"loss": 0.7159,
"step": 853
},
{
"epoch": 0.29985955056179775,
"grad_norm": 0.902009904384613,
"learning_rate": 4.975735900709113e-06,
"loss": 0.6952,
"step": 854
},
{
"epoch": 0.30021067415730335,
"grad_norm": 0.9207450151443481,
"learning_rate": 4.975671601682822e-06,
"loss": 0.6705,
"step": 855
},
{
"epoch": 0.300561797752809,
"grad_norm": 0.9092718362808228,
"learning_rate": 4.975607217990681e-06,
"loss": 0.723,
"step": 856
},
{
"epoch": 0.3009129213483146,
"grad_norm": 0.955410897731781,
"learning_rate": 4.975542749634891e-06,
"loss": 0.728,
"step": 857
},
{
"epoch": 0.3012640449438202,
"grad_norm": 0.9404045343399048,
"learning_rate": 4.975478196617658e-06,
"loss": 0.7381,
"step": 858
},
{
"epoch": 0.30161516853932585,
"grad_norm": 0.9133435487747192,
"learning_rate": 4.9754135589411875e-06,
"loss": 0.7306,
"step": 859
},
{
"epoch": 0.30196629213483145,
"grad_norm": 0.9013397693634033,
"learning_rate": 4.975348836607693e-06,
"loss": 0.7189,
"step": 860
},
{
"epoch": 0.3023174157303371,
"grad_norm": 0.9037486910820007,
"learning_rate": 4.975284029619386e-06,
"loss": 0.7152,
"step": 861
},
{
"epoch": 0.3026685393258427,
"grad_norm": 0.9141895771026611,
"learning_rate": 4.975219137978484e-06,
"loss": 0.71,
"step": 862
},
{
"epoch": 0.3030196629213483,
"grad_norm": 0.942010223865509,
"learning_rate": 4.975154161687205e-06,
"loss": 0.7255,
"step": 863
},
{
"epoch": 0.30337078651685395,
"grad_norm": 0.9464311599731445,
"learning_rate": 4.9750891007477716e-06,
"loss": 0.7012,
"step": 864
},
{
"epoch": 0.30372191011235955,
"grad_norm": 0.9062106609344482,
"learning_rate": 4.975023955162409e-06,
"loss": 0.6909,
"step": 865
},
{
"epoch": 0.30407303370786515,
"grad_norm": 0.921981155872345,
"learning_rate": 4.974958724933344e-06,
"loss": 0.7312,
"step": 866
},
{
"epoch": 0.3044241573033708,
"grad_norm": 0.902531087398529,
"learning_rate": 4.97489341006281e-06,
"loss": 0.674,
"step": 867
},
{
"epoch": 0.3047752808988764,
"grad_norm": 0.9487978219985962,
"learning_rate": 4.974828010553039e-06,
"loss": 0.7054,
"step": 868
},
{
"epoch": 0.305126404494382,
"grad_norm": 0.9971505999565125,
"learning_rate": 4.9747625264062675e-06,
"loss": 0.7676,
"step": 869
},
{
"epoch": 0.30547752808988765,
"grad_norm": 0.9108356237411499,
"learning_rate": 4.974696957624735e-06,
"loss": 0.7199,
"step": 870
},
{
"epoch": 0.30582865168539325,
"grad_norm": 0.9270719885826111,
"learning_rate": 4.974631304210684e-06,
"loss": 0.748,
"step": 871
},
{
"epoch": 0.3061797752808989,
"grad_norm": 0.9216623306274414,
"learning_rate": 4.97456556616636e-06,
"loss": 0.7333,
"step": 872
},
{
"epoch": 0.3065308988764045,
"grad_norm": 0.980196475982666,
"learning_rate": 4.9744997434940125e-06,
"loss": 0.7175,
"step": 873
},
{
"epoch": 0.3068820224719101,
"grad_norm": 0.8955034613609314,
"learning_rate": 4.97443383619589e-06,
"loss": 0.7133,
"step": 874
},
{
"epoch": 0.30723314606741575,
"grad_norm": 0.9280632734298706,
"learning_rate": 4.974367844274248e-06,
"loss": 0.6953,
"step": 875
},
{
"epoch": 0.30758426966292135,
"grad_norm": 0.9301517605781555,
"learning_rate": 4.974301767731343e-06,
"loss": 0.7057,
"step": 876
},
{
"epoch": 0.30793539325842695,
"grad_norm": 0.9015063047409058,
"learning_rate": 4.974235606569434e-06,
"loss": 0.699,
"step": 877
},
{
"epoch": 0.3082865168539326,
"grad_norm": 0.9064421057701111,
"learning_rate": 4.974169360790786e-06,
"loss": 0.6937,
"step": 878
},
{
"epoch": 0.3086376404494382,
"grad_norm": 0.9030662775039673,
"learning_rate": 4.974103030397662e-06,
"loss": 0.7455,
"step": 879
},
{
"epoch": 0.3089887640449438,
"grad_norm": 0.9049381613731384,
"learning_rate": 4.974036615392332e-06,
"loss": 0.7203,
"step": 880
},
{
"epoch": 0.30933988764044945,
"grad_norm": 0.9038983583450317,
"learning_rate": 4.973970115777067e-06,
"loss": 0.7122,
"step": 881
},
{
"epoch": 0.30969101123595505,
"grad_norm": 0.96694415807724,
"learning_rate": 4.973903531554141e-06,
"loss": 0.7292,
"step": 882
},
{
"epoch": 0.31004213483146065,
"grad_norm": 0.9090932011604309,
"learning_rate": 4.973836862725831e-06,
"loss": 0.7463,
"step": 883
},
{
"epoch": 0.3103932584269663,
"grad_norm": 0.9024604558944702,
"learning_rate": 4.973770109294416e-06,
"loss": 0.7488,
"step": 884
},
{
"epoch": 0.3107443820224719,
"grad_norm": 0.9796269536018372,
"learning_rate": 4.973703271262181e-06,
"loss": 0.7398,
"step": 885
},
{
"epoch": 0.31109550561797755,
"grad_norm": 0.9187832474708557,
"learning_rate": 4.973636348631412e-06,
"loss": 0.7162,
"step": 886
},
{
"epoch": 0.31144662921348315,
"grad_norm": 0.8963474631309509,
"learning_rate": 4.973569341404395e-06,
"loss": 0.6812,
"step": 887
},
{
"epoch": 0.31179775280898875,
"grad_norm": 0.934047281742096,
"learning_rate": 4.973502249583424e-06,
"loss": 0.7176,
"step": 888
},
{
"epoch": 0.3121488764044944,
"grad_norm": 0.9109435677528381,
"learning_rate": 4.9734350731707925e-06,
"loss": 0.684,
"step": 889
},
{
"epoch": 0.3125,
"grad_norm": 0.9629603028297424,
"learning_rate": 4.973367812168798e-06,
"loss": 0.756,
"step": 890
},
{
"epoch": 0.3128511235955056,
"grad_norm": 0.8999819159507751,
"learning_rate": 4.973300466579741e-06,
"loss": 0.7211,
"step": 891
},
{
"epoch": 0.31320224719101125,
"grad_norm": 0.9093006253242493,
"learning_rate": 4.9732330364059245e-06,
"loss": 0.6893,
"step": 892
},
{
"epoch": 0.31355337078651685,
"grad_norm": 0.925269603729248,
"learning_rate": 4.973165521649655e-06,
"loss": 0.6631,
"step": 893
},
{
"epoch": 0.31390449438202245,
"grad_norm": 0.9515416622161865,
"learning_rate": 4.973097922313239e-06,
"loss": 0.7578,
"step": 894
},
{
"epoch": 0.3142556179775281,
"grad_norm": 0.9192783236503601,
"learning_rate": 4.973030238398992e-06,
"loss": 0.6884,
"step": 895
},
{
"epoch": 0.3146067415730337,
"grad_norm": 0.905805230140686,
"learning_rate": 4.972962469909227e-06,
"loss": 0.7037,
"step": 896
},
{
"epoch": 0.31495786516853935,
"grad_norm": 0.9037303328514099,
"learning_rate": 4.972894616846261e-06,
"loss": 0.6989,
"step": 897
},
{
"epoch": 0.31530898876404495,
"grad_norm": 0.8707857728004456,
"learning_rate": 4.9728266792124155e-06,
"loss": 0.6865,
"step": 898
},
{
"epoch": 0.31566011235955055,
"grad_norm": 0.9623267650604248,
"learning_rate": 4.972758657010014e-06,
"loss": 0.7028,
"step": 899
},
{
"epoch": 0.3160112359550562,
"grad_norm": 0.9062806367874146,
"learning_rate": 4.972690550241383e-06,
"loss": 0.6803,
"step": 900
},
{
"epoch": 0.3163623595505618,
"grad_norm": 0.9363368153572083,
"learning_rate": 4.97262235890885e-06,
"loss": 0.7359,
"step": 901
},
{
"epoch": 0.3167134831460674,
"grad_norm": 0.9362224340438843,
"learning_rate": 4.972554083014748e-06,
"loss": 0.7432,
"step": 902
},
{
"epoch": 0.31706460674157305,
"grad_norm": 0.9579148292541504,
"learning_rate": 4.972485722561412e-06,
"loss": 0.7348,
"step": 903
},
{
"epoch": 0.31741573033707865,
"grad_norm": 0.9559216499328613,
"learning_rate": 4.97241727755118e-06,
"loss": 0.7095,
"step": 904
},
{
"epoch": 0.31776685393258425,
"grad_norm": 0.9043166637420654,
"learning_rate": 4.972348747986394e-06,
"loss": 0.7289,
"step": 905
},
{
"epoch": 0.3181179775280899,
"grad_norm": 0.925491988658905,
"learning_rate": 4.972280133869396e-06,
"loss": 0.6921,
"step": 906
},
{
"epoch": 0.3184691011235955,
"grad_norm": 0.9442777037620544,
"learning_rate": 4.972211435202533e-06,
"loss": 0.7104,
"step": 907
},
{
"epoch": 0.3188202247191011,
"grad_norm": 0.943748950958252,
"learning_rate": 4.972142651988153e-06,
"loss": 0.7688,
"step": 908
},
{
"epoch": 0.31917134831460675,
"grad_norm": 0.8944805860519409,
"learning_rate": 4.972073784228612e-06,
"loss": 0.7321,
"step": 909
},
{
"epoch": 0.31952247191011235,
"grad_norm": 0.9088547825813293,
"learning_rate": 4.972004831926261e-06,
"loss": 0.6973,
"step": 910
},
{
"epoch": 0.319873595505618,
"grad_norm": 0.9270226359367371,
"learning_rate": 4.971935795083461e-06,
"loss": 0.6956,
"step": 911
},
{
"epoch": 0.3202247191011236,
"grad_norm": 0.910492479801178,
"learning_rate": 4.971866673702573e-06,
"loss": 0.681,
"step": 912
},
{
"epoch": 0.3205758426966292,
"grad_norm": 0.9591485857963562,
"learning_rate": 4.971797467785958e-06,
"loss": 0.7763,
"step": 913
},
{
"epoch": 0.32092696629213485,
"grad_norm": 0.9132441878318787,
"learning_rate": 4.971728177335986e-06,
"loss": 0.7294,
"step": 914
},
{
"epoch": 0.32127808988764045,
"grad_norm": 0.9387941956520081,
"learning_rate": 4.971658802355025e-06,
"loss": 0.755,
"step": 915
},
{
"epoch": 0.32162921348314605,
"grad_norm": 0.930505096912384,
"learning_rate": 4.971589342845449e-06,
"loss": 0.7337,
"step": 916
},
{
"epoch": 0.3219803370786517,
"grad_norm": 0.8818848133087158,
"learning_rate": 4.971519798809631e-06,
"loss": 0.7114,
"step": 917
},
{
"epoch": 0.3223314606741573,
"grad_norm": 0.9763118028640747,
"learning_rate": 4.971450170249952e-06,
"loss": 0.73,
"step": 918
},
{
"epoch": 0.3226825842696629,
"grad_norm": 0.865503191947937,
"learning_rate": 4.971380457168791e-06,
"loss": 0.6733,
"step": 919
},
{
"epoch": 0.32303370786516855,
"grad_norm": 0.9227094650268555,
"learning_rate": 4.9713106595685336e-06,
"loss": 0.6975,
"step": 920
},
{
"epoch": 0.32338483146067415,
"grad_norm": 0.9164808988571167,
"learning_rate": 4.971240777451566e-06,
"loss": 0.7266,
"step": 921
},
{
"epoch": 0.3237359550561798,
"grad_norm": 0.9115430116653442,
"learning_rate": 4.971170810820279e-06,
"loss": 0.6775,
"step": 922
},
{
"epoch": 0.3240870786516854,
"grad_norm": 0.9133177399635315,
"learning_rate": 4.971100759677064e-06,
"loss": 0.7163,
"step": 923
},
{
"epoch": 0.324438202247191,
"grad_norm": 0.9289098381996155,
"learning_rate": 4.971030624024319e-06,
"loss": 0.7448,
"step": 924
},
{
"epoch": 0.32478932584269665,
"grad_norm": 0.872006356716156,
"learning_rate": 4.97096040386444e-06,
"loss": 0.6492,
"step": 925
},
{
"epoch": 0.32514044943820225,
"grad_norm": 0.9258365035057068,
"learning_rate": 4.97089009919983e-06,
"loss": 0.6931,
"step": 926
},
{
"epoch": 0.32549157303370785,
"grad_norm": 0.9044914841651917,
"learning_rate": 4.970819710032893e-06,
"loss": 0.7002,
"step": 927
},
{
"epoch": 0.3258426966292135,
"grad_norm": 0.9293612837791443,
"learning_rate": 4.970749236366037e-06,
"loss": 0.7471,
"step": 928
},
{
"epoch": 0.3261938202247191,
"grad_norm": 0.8999515771865845,
"learning_rate": 4.9706786782016706e-06,
"loss": 0.6992,
"step": 929
},
{
"epoch": 0.3265449438202247,
"grad_norm": 0.9285963177680969,
"learning_rate": 4.970608035542207e-06,
"loss": 0.7229,
"step": 930
},
{
"epoch": 0.32689606741573035,
"grad_norm": 0.9883245229721069,
"learning_rate": 4.970537308390063e-06,
"loss": 0.7343,
"step": 931
},
{
"epoch": 0.32724719101123595,
"grad_norm": 0.9650030732154846,
"learning_rate": 4.970466496747658e-06,
"loss": 0.7206,
"step": 932
},
{
"epoch": 0.32759831460674155,
"grad_norm": 0.9341951012611389,
"learning_rate": 4.9703956006174125e-06,
"loss": 0.7029,
"step": 933
},
{
"epoch": 0.3279494382022472,
"grad_norm": 0.9105697274208069,
"learning_rate": 4.9703246200017504e-06,
"loss": 0.7127,
"step": 934
},
{
"epoch": 0.3283005617977528,
"grad_norm": 0.8891416192054749,
"learning_rate": 4.970253554903102e-06,
"loss": 0.6952,
"step": 935
},
{
"epoch": 0.32865168539325845,
"grad_norm": 0.9152025580406189,
"learning_rate": 4.970182405323894e-06,
"loss": 0.7126,
"step": 936
},
{
"epoch": 0.32900280898876405,
"grad_norm": 0.9279428720474243,
"learning_rate": 4.970111171266563e-06,
"loss": 0.7066,
"step": 937
},
{
"epoch": 0.32935393258426965,
"grad_norm": 0.9324163198471069,
"learning_rate": 4.970039852733542e-06,
"loss": 0.6796,
"step": 938
},
{
"epoch": 0.3297050561797753,
"grad_norm": 0.9290011525154114,
"learning_rate": 4.9699684497272736e-06,
"loss": 0.7192,
"step": 939
},
{
"epoch": 0.3300561797752809,
"grad_norm": 0.8959920406341553,
"learning_rate": 4.969896962250196e-06,
"loss": 0.7132,
"step": 940
},
{
"epoch": 0.3304073033707865,
"grad_norm": 0.8849387168884277,
"learning_rate": 4.969825390304757e-06,
"loss": 0.6618,
"step": 941
},
{
"epoch": 0.33075842696629215,
"grad_norm": 0.8804863691329956,
"learning_rate": 4.969753733893403e-06,
"loss": 0.7021,
"step": 942
},
{
"epoch": 0.33110955056179775,
"grad_norm": 0.9022026062011719,
"learning_rate": 4.969681993018584e-06,
"loss": 0.6842,
"step": 943
},
{
"epoch": 0.33146067415730335,
"grad_norm": 0.9250811338424683,
"learning_rate": 4.969610167682754e-06,
"loss": 0.6977,
"step": 944
},
{
"epoch": 0.331811797752809,
"grad_norm": 0.9264232516288757,
"learning_rate": 4.969538257888371e-06,
"loss": 0.7135,
"step": 945
},
{
"epoch": 0.3321629213483146,
"grad_norm": 0.9282289147377014,
"learning_rate": 4.969466263637892e-06,
"loss": 0.7304,
"step": 946
},
{
"epoch": 0.3325140449438202,
"grad_norm": 0.8798340559005737,
"learning_rate": 4.9693941849337804e-06,
"loss": 0.6996,
"step": 947
},
{
"epoch": 0.33286516853932585,
"grad_norm": 0.9778987169265747,
"learning_rate": 4.9693220217785e-06,
"loss": 0.7132,
"step": 948
},
{
"epoch": 0.33321629213483145,
"grad_norm": 0.9029282331466675,
"learning_rate": 4.969249774174519e-06,
"loss": 0.6951,
"step": 949
},
{
"epoch": 0.3335674157303371,
"grad_norm": 0.9430015087127686,
"learning_rate": 4.96917744212431e-06,
"loss": 0.6943,
"step": 950
},
{
"epoch": 0.3339185393258427,
"grad_norm": 0.9794940948486328,
"learning_rate": 4.9691050256303445e-06,
"loss": 0.7054,
"step": 951
},
{
"epoch": 0.3342696629213483,
"grad_norm": 0.9288370013237,
"learning_rate": 4.969032524695101e-06,
"loss": 0.7122,
"step": 952
},
{
"epoch": 0.33462078651685395,
"grad_norm": 0.9557552933692932,
"learning_rate": 4.968959939321057e-06,
"loss": 0.7598,
"step": 953
},
{
"epoch": 0.33497191011235955,
"grad_norm": 0.9375066161155701,
"learning_rate": 4.968887269510696e-06,
"loss": 0.6924,
"step": 954
},
{
"epoch": 0.33532303370786515,
"grad_norm": 0.9338163733482361,
"learning_rate": 4.968814515266503e-06,
"loss": 0.7267,
"step": 955
},
{
"epoch": 0.3356741573033708,
"grad_norm": 0.9154375791549683,
"learning_rate": 4.968741676590966e-06,
"loss": 0.7321,
"step": 956
},
{
"epoch": 0.3360252808988764,
"grad_norm": 0.9233065843582153,
"learning_rate": 4.968668753486576e-06,
"loss": 0.7245,
"step": 957
},
{
"epoch": 0.336376404494382,
"grad_norm": 0.9419113993644714,
"learning_rate": 4.968595745955828e-06,
"loss": 0.7007,
"step": 958
},
{
"epoch": 0.33672752808988765,
"grad_norm": 0.961121678352356,
"learning_rate": 4.968522654001217e-06,
"loss": 0.7308,
"step": 959
},
{
"epoch": 0.33707865168539325,
"grad_norm": 0.934877336025238,
"learning_rate": 4.9684494776252445e-06,
"loss": 0.7388,
"step": 960
},
{
"epoch": 0.3374297752808989,
"grad_norm": 0.912398636341095,
"learning_rate": 4.9683762168304115e-06,
"loss": 0.7088,
"step": 961
},
{
"epoch": 0.3377808988764045,
"grad_norm": 0.948833167552948,
"learning_rate": 4.9683028716192246e-06,
"loss": 0.7278,
"step": 962
},
{
"epoch": 0.3381320224719101,
"grad_norm": 0.9257057905197144,
"learning_rate": 4.968229441994191e-06,
"loss": 0.7338,
"step": 963
},
{
"epoch": 0.33848314606741575,
"grad_norm": 0.9129422903060913,
"learning_rate": 4.968155927957823e-06,
"loss": 0.7058,
"step": 964
},
{
"epoch": 0.33883426966292135,
"grad_norm": 0.9381750822067261,
"learning_rate": 4.968082329512634e-06,
"loss": 0.7326,
"step": 965
},
{
"epoch": 0.33918539325842695,
"grad_norm": 0.9108783602714539,
"learning_rate": 4.968008646661141e-06,
"loss": 0.66,
"step": 966
},
{
"epoch": 0.3395365168539326,
"grad_norm": 0.9329748153686523,
"learning_rate": 4.9679348794058655e-06,
"loss": 0.7117,
"step": 967
},
{
"epoch": 0.3398876404494382,
"grad_norm": 0.9643568992614746,
"learning_rate": 4.967861027749327e-06,
"loss": 0.7555,
"step": 968
},
{
"epoch": 0.3402387640449438,
"grad_norm": 0.9551818370819092,
"learning_rate": 4.967787091694055e-06,
"loss": 0.7474,
"step": 969
},
{
"epoch": 0.34058988764044945,
"grad_norm": 0.9381354451179504,
"learning_rate": 4.967713071242576e-06,
"loss": 0.7155,
"step": 970
},
{
"epoch": 0.34094101123595505,
"grad_norm": 0.9513474702835083,
"learning_rate": 4.967638966397421e-06,
"loss": 0.765,
"step": 971
},
{
"epoch": 0.34129213483146065,
"grad_norm": 0.9263571500778198,
"learning_rate": 4.967564777161124e-06,
"loss": 0.7112,
"step": 972
},
{
"epoch": 0.3416432584269663,
"grad_norm": 0.9278532862663269,
"learning_rate": 4.967490503536225e-06,
"loss": 0.7183,
"step": 973
},
{
"epoch": 0.3419943820224719,
"grad_norm": 0.9487691521644592,
"learning_rate": 4.967416145525261e-06,
"loss": 0.7312,
"step": 974
},
{
"epoch": 0.34234550561797755,
"grad_norm": 0.8866797089576721,
"learning_rate": 4.967341703130777e-06,
"loss": 0.704,
"step": 975
},
{
"epoch": 0.34269662921348315,
"grad_norm": 0.9202519655227661,
"learning_rate": 4.967267176355318e-06,
"loss": 0.7434,
"step": 976
},
{
"epoch": 0.34304775280898875,
"grad_norm": 0.9155771732330322,
"learning_rate": 4.967192565201432e-06,
"loss": 0.7166,
"step": 977
},
{
"epoch": 0.3433988764044944,
"grad_norm": 0.9582788944244385,
"learning_rate": 4.967117869671672e-06,
"loss": 0.7536,
"step": 978
},
{
"epoch": 0.34375,
"grad_norm": 0.9524344801902771,
"learning_rate": 4.967043089768592e-06,
"loss": 0.7739,
"step": 979
},
{
"epoch": 0.3441011235955056,
"grad_norm": 0.9230315685272217,
"learning_rate": 4.96696822549475e-06,
"loss": 0.7181,
"step": 980
},
{
"epoch": 0.34445224719101125,
"grad_norm": 0.9080458283424377,
"learning_rate": 4.966893276852704e-06,
"loss": 0.7107,
"step": 981
},
{
"epoch": 0.34480337078651685,
"grad_norm": 0.9102587699890137,
"learning_rate": 4.966818243845021e-06,
"loss": 0.7051,
"step": 982
},
{
"epoch": 0.34515449438202245,
"grad_norm": 0.8925254940986633,
"learning_rate": 4.966743126474264e-06,
"loss": 0.7106,
"step": 983
},
{
"epoch": 0.3455056179775281,
"grad_norm": 0.9102403521537781,
"learning_rate": 4.966667924743002e-06,
"loss": 0.6849,
"step": 984
},
{
"epoch": 0.3458567415730337,
"grad_norm": 0.9164739847183228,
"learning_rate": 4.966592638653808e-06,
"loss": 0.7027,
"step": 985
},
{
"epoch": 0.34620786516853935,
"grad_norm": 0.89588463306427,
"learning_rate": 4.966517268209258e-06,
"loss": 0.7086,
"step": 986
},
{
"epoch": 0.34655898876404495,
"grad_norm": 0.8812805414199829,
"learning_rate": 4.966441813411928e-06,
"loss": 0.7073,
"step": 987
},
{
"epoch": 0.34691011235955055,
"grad_norm": 0.8935489654541016,
"learning_rate": 4.966366274264397e-06,
"loss": 0.6749,
"step": 988
},
{
"epoch": 0.3472612359550562,
"grad_norm": 0.935444712638855,
"learning_rate": 4.966290650769251e-06,
"loss": 0.7283,
"step": 989
},
{
"epoch": 0.3476123595505618,
"grad_norm": 0.9046710729598999,
"learning_rate": 4.966214942929074e-06,
"loss": 0.7138,
"step": 990
},
{
"epoch": 0.3479634831460674,
"grad_norm": 0.9328753352165222,
"learning_rate": 4.966139150746458e-06,
"loss": 0.7022,
"step": 991
},
{
"epoch": 0.34831460674157305,
"grad_norm": 0.9144490361213684,
"learning_rate": 4.966063274223992e-06,
"loss": 0.738,
"step": 992
},
{
"epoch": 0.34866573033707865,
"grad_norm": 0.9606062173843384,
"learning_rate": 4.965987313364272e-06,
"loss": 0.6937,
"step": 993
},
{
"epoch": 0.34901685393258425,
"grad_norm": 0.94340980052948,
"learning_rate": 4.965911268169897e-06,
"loss": 0.6811,
"step": 994
},
{
"epoch": 0.3493679775280899,
"grad_norm": 0.8755514025688171,
"learning_rate": 4.9658351386434665e-06,
"loss": 0.6707,
"step": 995
},
{
"epoch": 0.3497191011235955,
"grad_norm": 0.8640532493591309,
"learning_rate": 4.965758924787584e-06,
"loss": 0.6378,
"step": 996
},
{
"epoch": 0.3500702247191011,
"grad_norm": 0.9261877536773682,
"learning_rate": 4.965682626604856e-06,
"loss": 0.6953,
"step": 997
},
{
"epoch": 0.35042134831460675,
"grad_norm": 0.8998399376869202,
"learning_rate": 4.965606244097894e-06,
"loss": 0.7154,
"step": 998
},
{
"epoch": 0.35077247191011235,
"grad_norm": 0.9440179467201233,
"learning_rate": 4.9655297772693066e-06,
"loss": 0.6992,
"step": 999
},
{
"epoch": 0.351123595505618,
"grad_norm": 0.9105573892593384,
"learning_rate": 4.965453226121711e-06,
"loss": 0.7033,
"step": 1000
},
{
"epoch": 0.3514747191011236,
"grad_norm": 0.9109785556793213,
"learning_rate": 4.965376590657724e-06,
"loss": 0.7057,
"step": 1001
},
{
"epoch": 0.3518258426966292,
"grad_norm": 0.9462383985519409,
"learning_rate": 4.9652998708799684e-06,
"loss": 0.6979,
"step": 1002
},
{
"epoch": 0.35217696629213485,
"grad_norm": 0.9546105861663818,
"learning_rate": 4.965223066791066e-06,
"loss": 0.7227,
"step": 1003
},
{
"epoch": 0.35252808988764045,
"grad_norm": 1.0831300020217896,
"learning_rate": 4.965146178393645e-06,
"loss": 0.6849,
"step": 1004
},
{
"epoch": 0.35287921348314605,
"grad_norm": 0.9323068261146545,
"learning_rate": 4.965069205690334e-06,
"loss": 0.739,
"step": 1005
},
{
"epoch": 0.3532303370786517,
"grad_norm": 0.9407396912574768,
"learning_rate": 4.964992148683765e-06,
"loss": 0.7386,
"step": 1006
},
{
"epoch": 0.3535814606741573,
"grad_norm": 0.9513671398162842,
"learning_rate": 4.964915007376574e-06,
"loss": 0.666,
"step": 1007
},
{
"epoch": 0.3539325842696629,
"grad_norm": 0.9194923043251038,
"learning_rate": 4.964837781771399e-06,
"loss": 0.7601,
"step": 1008
},
{
"epoch": 0.35428370786516855,
"grad_norm": 0.9639756083488464,
"learning_rate": 4.964760471870882e-06,
"loss": 0.7124,
"step": 1009
},
{
"epoch": 0.35463483146067415,
"grad_norm": 0.9628319144248962,
"learning_rate": 4.964683077677664e-06,
"loss": 0.7003,
"step": 1010
},
{
"epoch": 0.3549859550561798,
"grad_norm": 0.9003061652183533,
"learning_rate": 4.964605599194395e-06,
"loss": 0.7199,
"step": 1011
},
{
"epoch": 0.3553370786516854,
"grad_norm": 0.9365134835243225,
"learning_rate": 4.964528036423723e-06,
"loss": 0.6651,
"step": 1012
},
{
"epoch": 0.355688202247191,
"grad_norm": 0.9309297204017639,
"learning_rate": 4.964450389368303e-06,
"loss": 0.7354,
"step": 1013
},
{
"epoch": 0.35603932584269665,
"grad_norm": 0.9624804258346558,
"learning_rate": 4.964372658030786e-06,
"loss": 0.6955,
"step": 1014
},
{
"epoch": 0.35639044943820225,
"grad_norm": 0.9497263431549072,
"learning_rate": 4.964294842413835e-06,
"loss": 0.7374,
"step": 1015
},
{
"epoch": 0.35674157303370785,
"grad_norm": 0.9235740900039673,
"learning_rate": 4.964216942520108e-06,
"loss": 0.6906,
"step": 1016
},
{
"epoch": 0.3570926966292135,
"grad_norm": 0.9985111951828003,
"learning_rate": 4.9641389583522705e-06,
"loss": 0.7131,
"step": 1017
},
{
"epoch": 0.3574438202247191,
"grad_norm": 0.9399976134300232,
"learning_rate": 4.964060889912988e-06,
"loss": 0.6935,
"step": 1018
},
{
"epoch": 0.3577949438202247,
"grad_norm": 0.9676353931427002,
"learning_rate": 4.9639827372049335e-06,
"loss": 0.7513,
"step": 1019
},
{
"epoch": 0.35814606741573035,
"grad_norm": 0.9015031456947327,
"learning_rate": 4.9639045002307765e-06,
"loss": 0.687,
"step": 1020
},
{
"epoch": 0.35849719101123595,
"grad_norm": 0.9401846528053284,
"learning_rate": 4.963826178993195e-06,
"loss": 0.6963,
"step": 1021
},
{
"epoch": 0.35884831460674155,
"grad_norm": 1.00020170211792,
"learning_rate": 4.9637477734948654e-06,
"loss": 0.6747,
"step": 1022
},
{
"epoch": 0.3591994382022472,
"grad_norm": 0.8997200727462769,
"learning_rate": 4.963669283738471e-06,
"loss": 0.7253,
"step": 1023
},
{
"epoch": 0.3595505617977528,
"grad_norm": 0.8761312365531921,
"learning_rate": 4.963590709726695e-06,
"loss": 0.6832,
"step": 1024
},
{
"epoch": 0.35990168539325845,
"grad_norm": 0.9415677189826965,
"learning_rate": 4.963512051462226e-06,
"loss": 0.6998,
"step": 1025
},
{
"epoch": 0.36025280898876405,
"grad_norm": 0.9139187932014465,
"learning_rate": 4.963433308947751e-06,
"loss": 0.7292,
"step": 1026
},
{
"epoch": 0.36060393258426965,
"grad_norm": 0.9207375049591064,
"learning_rate": 4.963354482185966e-06,
"loss": 0.6779,
"step": 1027
},
{
"epoch": 0.3609550561797753,
"grad_norm": 0.9517902135848999,
"learning_rate": 4.963275571179565e-06,
"loss": 0.7525,
"step": 1028
},
{
"epoch": 0.3613061797752809,
"grad_norm": 0.9088241457939148,
"learning_rate": 4.963196575931248e-06,
"loss": 0.7196,
"step": 1029
},
{
"epoch": 0.3616573033707865,
"grad_norm": 0.8898507356643677,
"learning_rate": 4.963117496443715e-06,
"loss": 0.7205,
"step": 1030
},
{
"epoch": 0.36200842696629215,
"grad_norm": 0.9045488834381104,
"learning_rate": 4.963038332719672e-06,
"loss": 0.6937,
"step": 1031
},
{
"epoch": 0.36235955056179775,
"grad_norm": 0.9505866169929504,
"learning_rate": 4.962959084761826e-06,
"loss": 0.7006,
"step": 1032
},
{
"epoch": 0.36271067415730335,
"grad_norm": 0.8908640742301941,
"learning_rate": 4.962879752572885e-06,
"loss": 0.7191,
"step": 1033
},
{
"epoch": 0.363061797752809,
"grad_norm": 0.925959050655365,
"learning_rate": 4.962800336155567e-06,
"loss": 0.6963,
"step": 1034
},
{
"epoch": 0.3634129213483146,
"grad_norm": 0.9082480072975159,
"learning_rate": 4.962720835512583e-06,
"loss": 0.7084,
"step": 1035
},
{
"epoch": 0.3637640449438202,
"grad_norm": 0.9313839077949524,
"learning_rate": 4.962641250646654e-06,
"loss": 0.7425,
"step": 1036
},
{
"epoch": 0.36411516853932585,
"grad_norm": 0.9484413266181946,
"learning_rate": 4.962561581560501e-06,
"loss": 0.6814,
"step": 1037
},
{
"epoch": 0.36446629213483145,
"grad_norm": 0.9550908207893372,
"learning_rate": 4.96248182825685e-06,
"loss": 0.6747,
"step": 1038
},
{
"epoch": 0.3648174157303371,
"grad_norm": 0.9393624663352966,
"learning_rate": 4.962401990738428e-06,
"loss": 0.7172,
"step": 1039
},
{
"epoch": 0.3651685393258427,
"grad_norm": 0.961174726486206,
"learning_rate": 4.962322069007964e-06,
"loss": 0.6967,
"step": 1040
},
{
"epoch": 0.3655196629213483,
"grad_norm": 0.9074354767799377,
"learning_rate": 4.962242063068193e-06,
"loss": 0.7019,
"step": 1041
},
{
"epoch": 0.36587078651685395,
"grad_norm": 0.9572650194168091,
"learning_rate": 4.962161972921851e-06,
"loss": 0.717,
"step": 1042
},
{
"epoch": 0.36622191011235955,
"grad_norm": 0.9262009263038635,
"learning_rate": 4.962081798571675e-06,
"loss": 0.7332,
"step": 1043
},
{
"epoch": 0.36657303370786515,
"grad_norm": 0.8846391439437866,
"learning_rate": 4.9620015400204094e-06,
"loss": 0.7359,
"step": 1044
},
{
"epoch": 0.3669241573033708,
"grad_norm": 0.9351332187652588,
"learning_rate": 4.961921197270797e-06,
"loss": 0.7809,
"step": 1045
},
{
"epoch": 0.3672752808988764,
"grad_norm": 0.9081857204437256,
"learning_rate": 4.961840770325587e-06,
"loss": 0.7169,
"step": 1046
},
{
"epoch": 0.367626404494382,
"grad_norm": 0.9414200782775879,
"learning_rate": 4.961760259187529e-06,
"loss": 0.7154,
"step": 1047
},
{
"epoch": 0.36797752808988765,
"grad_norm": 0.9009916186332703,
"learning_rate": 4.961679663859377e-06,
"loss": 0.7013,
"step": 1048
},
{
"epoch": 0.36832865168539325,
"grad_norm": 0.9196215271949768,
"learning_rate": 4.961598984343886e-06,
"loss": 0.706,
"step": 1049
},
{
"epoch": 0.3686797752808989,
"grad_norm": 0.947223424911499,
"learning_rate": 4.961518220643817e-06,
"loss": 0.712,
"step": 1050
},
{
"epoch": 0.3690308988764045,
"grad_norm": 0.8798839449882507,
"learning_rate": 4.961437372761932e-06,
"loss": 0.6821,
"step": 1051
},
{
"epoch": 0.3693820224719101,
"grad_norm": 0.9458513259887695,
"learning_rate": 4.961356440700994e-06,
"loss": 0.7503,
"step": 1052
},
{
"epoch": 0.36973314606741575,
"grad_norm": 0.9466217160224915,
"learning_rate": 4.961275424463772e-06,
"loss": 0.7086,
"step": 1053
},
{
"epoch": 0.37008426966292135,
"grad_norm": 0.9832736849784851,
"learning_rate": 4.961194324053037e-06,
"loss": 0.7332,
"step": 1054
},
{
"epoch": 0.37043539325842695,
"grad_norm": 0.9869272708892822,
"learning_rate": 4.961113139471562e-06,
"loss": 0.7595,
"step": 1055
},
{
"epoch": 0.3707865168539326,
"grad_norm": 0.9590395092964172,
"learning_rate": 4.961031870722123e-06,
"loss": 0.7471,
"step": 1056
},
{
"epoch": 0.3711376404494382,
"grad_norm": 0.9194419980049133,
"learning_rate": 4.9609505178075015e-06,
"loss": 0.7147,
"step": 1057
},
{
"epoch": 0.3714887640449438,
"grad_norm": 0.8812661170959473,
"learning_rate": 4.960869080730477e-06,
"loss": 0.6723,
"step": 1058
},
{
"epoch": 0.37183988764044945,
"grad_norm": 0.9303286671638489,
"learning_rate": 4.960787559493836e-06,
"loss": 0.7253,
"step": 1059
},
{
"epoch": 0.37219101123595505,
"grad_norm": 0.9536207318305969,
"learning_rate": 4.960705954100366e-06,
"loss": 0.7325,
"step": 1060
},
{
"epoch": 0.37254213483146065,
"grad_norm": 0.9165772795677185,
"learning_rate": 4.960624264552858e-06,
"loss": 0.6949,
"step": 1061
},
{
"epoch": 0.3728932584269663,
"grad_norm": 0.9144523739814758,
"learning_rate": 4.9605424908541065e-06,
"loss": 0.7355,
"step": 1062
},
{
"epoch": 0.3732443820224719,
"grad_norm": 0.8873782753944397,
"learning_rate": 4.960460633006906e-06,
"loss": 0.6819,
"step": 1063
},
{
"epoch": 0.37359550561797755,
"grad_norm": 0.9444409012794495,
"learning_rate": 4.960378691014058e-06,
"loss": 0.6851,
"step": 1064
},
{
"epoch": 0.37394662921348315,
"grad_norm": 0.959851861000061,
"learning_rate": 4.960296664878364e-06,
"loss": 0.7127,
"step": 1065
},
{
"epoch": 0.37429775280898875,
"grad_norm": 0.9353641867637634,
"learning_rate": 4.9602145546026305e-06,
"loss": 0.7208,
"step": 1066
},
{
"epoch": 0.3746488764044944,
"grad_norm": 0.9105174541473389,
"learning_rate": 4.960132360189663e-06,
"loss": 0.6717,
"step": 1067
},
{
"epoch": 0.375,
"grad_norm": 0.9150325655937195,
"learning_rate": 4.960050081642274e-06,
"loss": 0.7233,
"step": 1068
},
{
"epoch": 0.3753511235955056,
"grad_norm": 0.9387450814247131,
"learning_rate": 4.959967718963279e-06,
"loss": 0.7104,
"step": 1069
},
{
"epoch": 0.37570224719101125,
"grad_norm": 0.9268593788146973,
"learning_rate": 4.959885272155491e-06,
"loss": 0.7209,
"step": 1070
},
{
"epoch": 0.37605337078651685,
"grad_norm": 0.9124265909194946,
"learning_rate": 4.959802741221733e-06,
"loss": 0.6982,
"step": 1071
},
{
"epoch": 0.37640449438202245,
"grad_norm": 0.9302802681922913,
"learning_rate": 4.9597201261648255e-06,
"loss": 0.7105,
"step": 1072
},
{
"epoch": 0.3767556179775281,
"grad_norm": 0.9560679793357849,
"learning_rate": 4.959637426987595e-06,
"loss": 0.7115,
"step": 1073
},
{
"epoch": 0.3771067415730337,
"grad_norm": 0.9246500730514526,
"learning_rate": 4.9595546436928685e-06,
"loss": 0.7319,
"step": 1074
},
{
"epoch": 0.37745786516853935,
"grad_norm": 0.9188322424888611,
"learning_rate": 4.959471776283479e-06,
"loss": 0.6829,
"step": 1075
},
{
"epoch": 0.37780898876404495,
"grad_norm": 0.9240451455116272,
"learning_rate": 4.959388824762258e-06,
"loss": 0.7183,
"step": 1076
},
{
"epoch": 0.37816011235955055,
"grad_norm": 0.8934212923049927,
"learning_rate": 4.959305789132044e-06,
"loss": 0.6982,
"step": 1077
},
{
"epoch": 0.3785112359550562,
"grad_norm": 0.9386299848556519,
"learning_rate": 4.959222669395677e-06,
"loss": 0.6942,
"step": 1078
},
{
"epoch": 0.3788623595505618,
"grad_norm": 0.9435991644859314,
"learning_rate": 4.959139465555999e-06,
"loss": 0.7382,
"step": 1079
},
{
"epoch": 0.3792134831460674,
"grad_norm": 0.9004935026168823,
"learning_rate": 4.959056177615855e-06,
"loss": 0.6944,
"step": 1080
},
{
"epoch": 0.37956460674157305,
"grad_norm": 0.9269630312919617,
"learning_rate": 4.958972805578095e-06,
"loss": 0.7037,
"step": 1081
},
{
"epoch": 0.37991573033707865,
"grad_norm": 0.9682616591453552,
"learning_rate": 4.9588893494455684e-06,
"loss": 0.6874,
"step": 1082
},
{
"epoch": 0.38026685393258425,
"grad_norm": 0.8845170140266418,
"learning_rate": 4.958805809221131e-06,
"loss": 0.6945,
"step": 1083
},
{
"epoch": 0.3806179775280899,
"grad_norm": 0.940234363079071,
"learning_rate": 4.958722184907638e-06,
"loss": 0.6813,
"step": 1084
},
{
"epoch": 0.3809691011235955,
"grad_norm": 0.9013071656227112,
"learning_rate": 4.95863847650795e-06,
"loss": 0.694,
"step": 1085
},
{
"epoch": 0.3813202247191011,
"grad_norm": 0.8996568322181702,
"learning_rate": 4.958554684024931e-06,
"loss": 0.6558,
"step": 1086
},
{
"epoch": 0.38167134831460675,
"grad_norm": 0.9193875193595886,
"learning_rate": 4.9584708074614455e-06,
"loss": 0.6996,
"step": 1087
},
{
"epoch": 0.38202247191011235,
"grad_norm": 0.9180407524108887,
"learning_rate": 4.958386846820361e-06,
"loss": 0.7147,
"step": 1088
},
{
"epoch": 0.382373595505618,
"grad_norm": 1.0345251560211182,
"learning_rate": 4.958302802104551e-06,
"loss": 0.7393,
"step": 1089
},
{
"epoch": 0.3827247191011236,
"grad_norm": 0.9022161364555359,
"learning_rate": 4.958218673316889e-06,
"loss": 0.6949,
"step": 1090
},
{
"epoch": 0.3830758426966292,
"grad_norm": 0.9014359712600708,
"learning_rate": 4.958134460460251e-06,
"loss": 0.6947,
"step": 1091
},
{
"epoch": 0.38342696629213485,
"grad_norm": 0.9596666693687439,
"learning_rate": 4.958050163537519e-06,
"loss": 0.7763,
"step": 1092
},
{
"epoch": 0.38377808988764045,
"grad_norm": 0.8888493180274963,
"learning_rate": 4.957965782551574e-06,
"loss": 0.7093,
"step": 1093
},
{
"epoch": 0.38412921348314605,
"grad_norm": 0.9068475365638733,
"learning_rate": 4.957881317505303e-06,
"loss": 0.6893,
"step": 1094
},
{
"epoch": 0.3844803370786517,
"grad_norm": 0.8918416500091553,
"learning_rate": 4.957796768401594e-06,
"loss": 0.6926,
"step": 1095
},
{
"epoch": 0.3848314606741573,
"grad_norm": 0.8772871494293213,
"learning_rate": 4.957712135243339e-06,
"loss": 0.6714,
"step": 1096
},
{
"epoch": 0.3851825842696629,
"grad_norm": 0.9466787576675415,
"learning_rate": 4.9576274180334315e-06,
"loss": 0.693,
"step": 1097
},
{
"epoch": 0.38553370786516855,
"grad_norm": 0.8940508961677551,
"learning_rate": 4.957542616774768e-06,
"loss": 0.6727,
"step": 1098
},
{
"epoch": 0.38588483146067415,
"grad_norm": 0.92746901512146,
"learning_rate": 4.957457731470253e-06,
"loss": 0.6832,
"step": 1099
},
{
"epoch": 0.3862359550561798,
"grad_norm": 0.9406936168670654,
"learning_rate": 4.9573727621227845e-06,
"loss": 0.7121,
"step": 1100
},
{
"epoch": 0.3865870786516854,
"grad_norm": 0.9280776381492615,
"learning_rate": 4.957287708735271e-06,
"loss": 0.7149,
"step": 1101
},
{
"epoch": 0.386938202247191,
"grad_norm": 0.9443319439888,
"learning_rate": 4.9572025713106195e-06,
"loss": 0.7205,
"step": 1102
},
{
"epoch": 0.38728932584269665,
"grad_norm": 0.9101063013076782,
"learning_rate": 4.957117349851745e-06,
"loss": 0.703,
"step": 1103
},
{
"epoch": 0.38764044943820225,
"grad_norm": 0.8856722116470337,
"learning_rate": 4.957032044361557e-06,
"loss": 0.7079,
"step": 1104
},
{
"epoch": 0.38799157303370785,
"grad_norm": 0.928821325302124,
"learning_rate": 4.956946654842977e-06,
"loss": 0.6935,
"step": 1105
},
{
"epoch": 0.3883426966292135,
"grad_norm": 0.9314661026000977,
"learning_rate": 4.956861181298924e-06,
"loss": 0.7062,
"step": 1106
},
{
"epoch": 0.3886938202247191,
"grad_norm": 0.9111589193344116,
"learning_rate": 4.956775623732321e-06,
"loss": 0.7343,
"step": 1107
},
{
"epoch": 0.3890449438202247,
"grad_norm": 0.8977031707763672,
"learning_rate": 4.9566899821460935e-06,
"loss": 0.6702,
"step": 1108
},
{
"epoch": 0.38939606741573035,
"grad_norm": 0.9246605634689331,
"learning_rate": 4.956604256543171e-06,
"loss": 0.7062,
"step": 1109
},
{
"epoch": 0.38974719101123595,
"grad_norm": 0.9252159595489502,
"learning_rate": 4.956518446926485e-06,
"loss": 0.6981,
"step": 1110
},
{
"epoch": 0.39009831460674155,
"grad_norm": 0.9128794074058533,
"learning_rate": 4.956432553298971e-06,
"loss": 0.6802,
"step": 1111
},
{
"epoch": 0.3904494382022472,
"grad_norm": 0.9178376197814941,
"learning_rate": 4.956346575663564e-06,
"loss": 0.6753,
"step": 1112
},
{
"epoch": 0.3908005617977528,
"grad_norm": 0.8804501891136169,
"learning_rate": 4.956260514023207e-06,
"loss": 0.7378,
"step": 1113
},
{
"epoch": 0.39115168539325845,
"grad_norm": 0.9001629948616028,
"learning_rate": 4.956174368380842e-06,
"loss": 0.7061,
"step": 1114
},
{
"epoch": 0.39150280898876405,
"grad_norm": 0.8731476664543152,
"learning_rate": 4.956088138739416e-06,
"loss": 0.6749,
"step": 1115
},
{
"epoch": 0.39185393258426965,
"grad_norm": 0.9055764079093933,
"learning_rate": 4.956001825101877e-06,
"loss": 0.7172,
"step": 1116
},
{
"epoch": 0.3922050561797753,
"grad_norm": 0.9398018717765808,
"learning_rate": 4.955915427471178e-06,
"loss": 0.711,
"step": 1117
},
{
"epoch": 0.3925561797752809,
"grad_norm": 0.9070323705673218,
"learning_rate": 4.9558289458502715e-06,
"loss": 0.7467,
"step": 1118
},
{
"epoch": 0.3929073033707865,
"grad_norm": 0.916070818901062,
"learning_rate": 4.9557423802421166e-06,
"loss": 0.6929,
"step": 1119
},
{
"epoch": 0.39325842696629215,
"grad_norm": 0.9212610721588135,
"learning_rate": 4.955655730649674e-06,
"loss": 0.7098,
"step": 1120
},
{
"epoch": 0.39360955056179775,
"grad_norm": 0.8813148736953735,
"learning_rate": 4.955568997075908e-06,
"loss": 0.7241,
"step": 1121
},
{
"epoch": 0.39396067415730335,
"grad_norm": 0.960132360458374,
"learning_rate": 4.955482179523782e-06,
"loss": 0.7531,
"step": 1122
},
{
"epoch": 0.394311797752809,
"grad_norm": 0.8893491625785828,
"learning_rate": 4.955395277996268e-06,
"loss": 0.708,
"step": 1123
},
{
"epoch": 0.3946629213483146,
"grad_norm": 0.9316296577453613,
"learning_rate": 4.955308292496336e-06,
"loss": 0.7129,
"step": 1124
},
{
"epoch": 0.3950140449438202,
"grad_norm": 0.8866797685623169,
"learning_rate": 4.955221223026961e-06,
"loss": 0.6986,
"step": 1125
},
{
"epoch": 0.39536516853932585,
"grad_norm": 0.9073642492294312,
"learning_rate": 4.955134069591121e-06,
"loss": 0.7192,
"step": 1126
},
{
"epoch": 0.39571629213483145,
"grad_norm": 0.8996924757957458,
"learning_rate": 4.9550468321917975e-06,
"loss": 0.6902,
"step": 1127
},
{
"epoch": 0.3960674157303371,
"grad_norm": 0.9362848401069641,
"learning_rate": 4.9549595108319726e-06,
"loss": 0.7331,
"step": 1128
},
{
"epoch": 0.3964185393258427,
"grad_norm": 0.9147123098373413,
"learning_rate": 4.954872105514633e-06,
"loss": 0.731,
"step": 1129
},
{
"epoch": 0.3967696629213483,
"grad_norm": 0.9537502527236938,
"learning_rate": 4.954784616242769e-06,
"loss": 0.7347,
"step": 1130
},
{
"epoch": 0.39712078651685395,
"grad_norm": 0.9200554490089417,
"learning_rate": 4.9546970430193705e-06,
"loss": 0.7205,
"step": 1131
},
{
"epoch": 0.39747191011235955,
"grad_norm": 0.929223358631134,
"learning_rate": 4.954609385847434e-06,
"loss": 0.7108,
"step": 1132
},
{
"epoch": 0.39782303370786515,
"grad_norm": 0.9166401624679565,
"learning_rate": 4.954521644729958e-06,
"loss": 0.732,
"step": 1133
},
{
"epoch": 0.3981741573033708,
"grad_norm": 0.9084526300430298,
"learning_rate": 4.954433819669941e-06,
"loss": 0.7078,
"step": 1134
},
{
"epoch": 0.3985252808988764,
"grad_norm": 0.9210569858551025,
"learning_rate": 4.954345910670388e-06,
"loss": 0.6905,
"step": 1135
},
{
"epoch": 0.398876404494382,
"grad_norm": 0.890738308429718,
"learning_rate": 4.9542579177343054e-06,
"loss": 0.6975,
"step": 1136
},
{
"epoch": 0.39922752808988765,
"grad_norm": 0.8724343776702881,
"learning_rate": 4.954169840864701e-06,
"loss": 0.6323,
"step": 1137
},
{
"epoch": 0.39957865168539325,
"grad_norm": 0.9185892343521118,
"learning_rate": 4.95408168006459e-06,
"loss": 0.7106,
"step": 1138
},
{
"epoch": 0.3999297752808989,
"grad_norm": 0.9605926871299744,
"learning_rate": 4.953993435336983e-06,
"loss": 0.7208,
"step": 1139
},
{
"epoch": 0.4002808988764045,
"grad_norm": 0.9934019446372986,
"learning_rate": 4.953905106684902e-06,
"loss": 0.7394,
"step": 1140
},
{
"epoch": 0.4006320224719101,
"grad_norm": 0.8674790263175964,
"learning_rate": 4.953816694111367e-06,
"loss": 0.6588,
"step": 1141
},
{
"epoch": 0.40098314606741575,
"grad_norm": 0.9420093894004822,
"learning_rate": 4.9537281976194e-06,
"loss": 0.6934,
"step": 1142
},
{
"epoch": 0.40133426966292135,
"grad_norm": 0.8882573246955872,
"learning_rate": 4.953639617212028e-06,
"loss": 0.7244,
"step": 1143
},
{
"epoch": 0.40168539325842695,
"grad_norm": 1.0398857593536377,
"learning_rate": 4.95355095289228e-06,
"loss": 0.7402,
"step": 1144
},
{
"epoch": 0.4020365168539326,
"grad_norm": 0.9207865595817566,
"learning_rate": 4.95346220466319e-06,
"loss": 0.6942,
"step": 1145
},
{
"epoch": 0.4023876404494382,
"grad_norm": 0.9174979329109192,
"learning_rate": 4.9533733725277924e-06,
"loss": 0.7241,
"step": 1146
},
{
"epoch": 0.4027387640449438,
"grad_norm": 0.9054567217826843,
"learning_rate": 4.953284456489124e-06,
"loss": 0.6666,
"step": 1147
},
{
"epoch": 0.40308988764044945,
"grad_norm": 0.9376466870307922,
"learning_rate": 4.953195456550227e-06,
"loss": 0.718,
"step": 1148
},
{
"epoch": 0.40344101123595505,
"grad_norm": 0.9617171883583069,
"learning_rate": 4.953106372714144e-06,
"loss": 0.7226,
"step": 1149
},
{
"epoch": 0.40379213483146065,
"grad_norm": 0.8789179921150208,
"learning_rate": 4.953017204983923e-06,
"loss": 0.6981,
"step": 1150
},
{
"epoch": 0.4041432584269663,
"grad_norm": 0.8860397934913635,
"learning_rate": 4.952927953362612e-06,
"loss": 0.6894,
"step": 1151
},
{
"epoch": 0.4044943820224719,
"grad_norm": 0.9390692114830017,
"learning_rate": 4.952838617853265e-06,
"loss": 0.6788,
"step": 1152
},
{
"epoch": 0.40484550561797755,
"grad_norm": 0.9032955765724182,
"learning_rate": 4.952749198458935e-06,
"loss": 0.7417,
"step": 1153
},
{
"epoch": 0.40519662921348315,
"grad_norm": 0.9163577556610107,
"learning_rate": 4.952659695182682e-06,
"loss": 0.7195,
"step": 1154
},
{
"epoch": 0.40554775280898875,
"grad_norm": 0.8781930804252625,
"learning_rate": 4.952570108027566e-06,
"loss": 0.6691,
"step": 1155
},
{
"epoch": 0.4058988764044944,
"grad_norm": 0.887157678604126,
"learning_rate": 4.9524804369966505e-06,
"loss": 0.6993,
"step": 1156
},
{
"epoch": 0.40625,
"grad_norm": 0.9891164302825928,
"learning_rate": 4.9523906820930035e-06,
"loss": 0.7357,
"step": 1157
},
{
"epoch": 0.4066011235955056,
"grad_norm": 0.9419492483139038,
"learning_rate": 4.952300843319693e-06,
"loss": 0.674,
"step": 1158
},
{
"epoch": 0.40695224719101125,
"grad_norm": 0.9248577356338501,
"learning_rate": 4.952210920679793e-06,
"loss": 0.724,
"step": 1159
},
{
"epoch": 0.40730337078651685,
"grad_norm": 0.9277257919311523,
"learning_rate": 4.9521209141763775e-06,
"loss": 0.7179,
"step": 1160
},
{
"epoch": 0.40765449438202245,
"grad_norm": 0.9654591083526611,
"learning_rate": 4.952030823812525e-06,
"loss": 0.6796,
"step": 1161
},
{
"epoch": 0.4080056179775281,
"grad_norm": 0.9730082154273987,
"learning_rate": 4.951940649591316e-06,
"loss": 0.715,
"step": 1162
},
{
"epoch": 0.4083567415730337,
"grad_norm": 0.9230988621711731,
"learning_rate": 4.9518503915158356e-06,
"loss": 0.7218,
"step": 1163
},
{
"epoch": 0.40870786516853935,
"grad_norm": 0.9274692535400391,
"learning_rate": 4.951760049589169e-06,
"loss": 0.6773,
"step": 1164
},
{
"epoch": 0.40905898876404495,
"grad_norm": 0.894001841545105,
"learning_rate": 4.951669623814408e-06,
"loss": 0.7118,
"step": 1165
},
{
"epoch": 0.40941011235955055,
"grad_norm": 0.924102783203125,
"learning_rate": 4.951579114194643e-06,
"loss": 0.6648,
"step": 1166
},
{
"epoch": 0.4097612359550562,
"grad_norm": 0.9237919449806213,
"learning_rate": 4.95148852073297e-06,
"loss": 0.697,
"step": 1167
},
{
"epoch": 0.4101123595505618,
"grad_norm": 0.9063787460327148,
"learning_rate": 4.951397843432489e-06,
"loss": 0.699,
"step": 1168
},
{
"epoch": 0.4104634831460674,
"grad_norm": 0.8865030407905579,
"learning_rate": 4.951307082296297e-06,
"loss": 0.6768,
"step": 1169
},
{
"epoch": 0.41081460674157305,
"grad_norm": 0.9309532046318054,
"learning_rate": 4.9512162373275015e-06,
"loss": 0.7168,
"step": 1170
},
{
"epoch": 0.41116573033707865,
"grad_norm": 0.9091811776161194,
"learning_rate": 4.951125308529208e-06,
"loss": 0.6962,
"step": 1171
},
{
"epoch": 0.41151685393258425,
"grad_norm": 0.8983654975891113,
"learning_rate": 4.951034295904527e-06,
"loss": 0.7193,
"step": 1172
},
{
"epoch": 0.4118679775280899,
"grad_norm": 0.8983482718467712,
"learning_rate": 4.95094319945657e-06,
"loss": 0.7234,
"step": 1173
},
{
"epoch": 0.4122191011235955,
"grad_norm": 0.9178565740585327,
"learning_rate": 4.950852019188454e-06,
"loss": 0.7133,
"step": 1174
},
{
"epoch": 0.4125702247191011,
"grad_norm": 0.9206750988960266,
"learning_rate": 4.950760755103294e-06,
"loss": 0.7417,
"step": 1175
},
{
"epoch": 0.41292134831460675,
"grad_norm": 0.9141495227813721,
"learning_rate": 4.950669407204215e-06,
"loss": 0.7238,
"step": 1176
},
{
"epoch": 0.41327247191011235,
"grad_norm": 0.9251378178596497,
"learning_rate": 4.9505779754943385e-06,
"loss": 0.6944,
"step": 1177
},
{
"epoch": 0.413623595505618,
"grad_norm": 0.9240232706069946,
"learning_rate": 4.950486459976793e-06,
"loss": 0.7241,
"step": 1178
},
{
"epoch": 0.4139747191011236,
"grad_norm": 0.9050787687301636,
"learning_rate": 4.950394860654707e-06,
"loss": 0.7044,
"step": 1179
},
{
"epoch": 0.4143258426966292,
"grad_norm": 0.9517340064048767,
"learning_rate": 4.950303177531213e-06,
"loss": 0.7363,
"step": 1180
},
{
"epoch": 0.41467696629213485,
"grad_norm": 0.9461115598678589,
"learning_rate": 4.950211410609448e-06,
"loss": 0.6786,
"step": 1181
},
{
"epoch": 0.41502808988764045,
"grad_norm": 0.9148724675178528,
"learning_rate": 4.950119559892549e-06,
"loss": 0.6873,
"step": 1182
},
{
"epoch": 0.41537921348314605,
"grad_norm": 0.9085513353347778,
"learning_rate": 4.950027625383657e-06,
"loss": 0.7209,
"step": 1183
},
{
"epoch": 0.4157303370786517,
"grad_norm": 0.9327560067176819,
"learning_rate": 4.949935607085919e-06,
"loss": 0.7086,
"step": 1184
},
{
"epoch": 0.4160814606741573,
"grad_norm": 0.936128556728363,
"learning_rate": 4.949843505002478e-06,
"loss": 0.6894,
"step": 1185
},
{
"epoch": 0.4164325842696629,
"grad_norm": 0.931978702545166,
"learning_rate": 4.9497513191364855e-06,
"loss": 0.7209,
"step": 1186
},
{
"epoch": 0.41678370786516855,
"grad_norm": 0.8966789245605469,
"learning_rate": 4.949659049491095e-06,
"loss": 0.7211,
"step": 1187
},
{
"epoch": 0.41713483146067415,
"grad_norm": 0.9003615975379944,
"learning_rate": 4.9495666960694596e-06,
"loss": 0.6835,
"step": 1188
},
{
"epoch": 0.4174859550561798,
"grad_norm": 0.933268666267395,
"learning_rate": 4.9494742588747405e-06,
"loss": 0.7335,
"step": 1189
},
{
"epoch": 0.4178370786516854,
"grad_norm": 0.8876200318336487,
"learning_rate": 4.949381737910098e-06,
"loss": 0.7132,
"step": 1190
},
{
"epoch": 0.418188202247191,
"grad_norm": 0.9531140923500061,
"learning_rate": 4.949289133178695e-06,
"loss": 0.746,
"step": 1191
},
{
"epoch": 0.41853932584269665,
"grad_norm": 0.9262682795524597,
"learning_rate": 4.949196444683701e-06,
"loss": 0.7155,
"step": 1192
},
{
"epoch": 0.41889044943820225,
"grad_norm": 0.9318521022796631,
"learning_rate": 4.949103672428282e-06,
"loss": 0.7234,
"step": 1193
},
{
"epoch": 0.41924157303370785,
"grad_norm": 0.922127366065979,
"learning_rate": 4.949010816415616e-06,
"loss": 0.7116,
"step": 1194
},
{
"epoch": 0.4195926966292135,
"grad_norm": 0.9445079565048218,
"learning_rate": 4.948917876648874e-06,
"loss": 0.6993,
"step": 1195
},
{
"epoch": 0.4199438202247191,
"grad_norm": 0.9283545613288879,
"learning_rate": 4.948824853131237e-06,
"loss": 0.7078,
"step": 1196
},
{
"epoch": 0.4202949438202247,
"grad_norm": 0.8923372626304626,
"learning_rate": 4.9487317458658845e-06,
"loss": 0.6818,
"step": 1197
},
{
"epoch": 0.42064606741573035,
"grad_norm": 0.907924234867096,
"learning_rate": 4.948638554856002e-06,
"loss": 0.6799,
"step": 1198
},
{
"epoch": 0.42099719101123595,
"grad_norm": 0.8753466606140137,
"learning_rate": 4.948545280104777e-06,
"loss": 0.6789,
"step": 1199
},
{
"epoch": 0.42134831460674155,
"grad_norm": 0.965340256690979,
"learning_rate": 4.948451921615398e-06,
"loss": 0.6911,
"step": 1200
},
{
"epoch": 0.4216994382022472,
"grad_norm": 0.9519873261451721,
"learning_rate": 4.948358479391059e-06,
"loss": 0.6899,
"step": 1201
},
{
"epoch": 0.4220505617977528,
"grad_norm": 0.907811164855957,
"learning_rate": 4.948264953434955e-06,
"loss": 0.7401,
"step": 1202
},
{
"epoch": 0.42240168539325845,
"grad_norm": 0.9398930072784424,
"learning_rate": 4.948171343750284e-06,
"loss": 0.7036,
"step": 1203
},
{
"epoch": 0.42275280898876405,
"grad_norm": 0.9245709776878357,
"learning_rate": 4.948077650340248e-06,
"loss": 0.7042,
"step": 1204
},
{
"epoch": 0.42310393258426965,
"grad_norm": 0.8742227554321289,
"learning_rate": 4.947983873208052e-06,
"loss": 0.6941,
"step": 1205
},
{
"epoch": 0.4234550561797753,
"grad_norm": 0.9183447957038879,
"learning_rate": 4.947890012356902e-06,
"loss": 0.7432,
"step": 1206
},
{
"epoch": 0.4238061797752809,
"grad_norm": 0.9594230055809021,
"learning_rate": 4.947796067790009e-06,
"loss": 0.7486,
"step": 1207
},
{
"epoch": 0.4241573033707865,
"grad_norm": 0.9002628326416016,
"learning_rate": 4.947702039510583e-06,
"loss": 0.6652,
"step": 1208
},
{
"epoch": 0.42450842696629215,
"grad_norm": 0.9353854060173035,
"learning_rate": 4.947607927521844e-06,
"loss": 0.7501,
"step": 1209
},
{
"epoch": 0.42485955056179775,
"grad_norm": 0.8836169242858887,
"learning_rate": 4.947513731827007e-06,
"loss": 0.6904,
"step": 1210
},
{
"epoch": 0.42521067415730335,
"grad_norm": 0.9366084337234497,
"learning_rate": 4.947419452429295e-06,
"loss": 0.7013,
"step": 1211
},
{
"epoch": 0.425561797752809,
"grad_norm": 0.9252647161483765,
"learning_rate": 4.947325089331932e-06,
"loss": 0.7343,
"step": 1212
},
{
"epoch": 0.4259129213483146,
"grad_norm": 0.9203290343284607,
"learning_rate": 4.947230642538145e-06,
"loss": 0.7129,
"step": 1213
},
{
"epoch": 0.4262640449438202,
"grad_norm": 0.8908094763755798,
"learning_rate": 4.947136112051164e-06,
"loss": 0.6785,
"step": 1214
},
{
"epoch": 0.42661516853932585,
"grad_norm": 0.9249038696289062,
"learning_rate": 4.947041497874223e-06,
"loss": 0.6664,
"step": 1215
},
{
"epoch": 0.42696629213483145,
"grad_norm": 0.9546428918838501,
"learning_rate": 4.946946800010556e-06,
"loss": 0.7276,
"step": 1216
},
{
"epoch": 0.4273174157303371,
"grad_norm": 0.9032455682754517,
"learning_rate": 4.946852018463403e-06,
"loss": 0.6757,
"step": 1217
},
{
"epoch": 0.4276685393258427,
"grad_norm": 0.9056017994880676,
"learning_rate": 4.946757153236003e-06,
"loss": 0.6878,
"step": 1218
},
{
"epoch": 0.4280196629213483,
"grad_norm": 0.9265672564506531,
"learning_rate": 4.946662204331603e-06,
"loss": 0.7043,
"step": 1219
},
{
"epoch": 0.42837078651685395,
"grad_norm": 0.8721897006034851,
"learning_rate": 4.94656717175345e-06,
"loss": 0.627,
"step": 1220
},
{
"epoch": 0.42872191011235955,
"grad_norm": 0.9220138192176819,
"learning_rate": 4.9464720555047935e-06,
"loss": 0.7233,
"step": 1221
},
{
"epoch": 0.42907303370786515,
"grad_norm": 0.9333030581474304,
"learning_rate": 4.9463768555888855e-06,
"loss": 0.7065,
"step": 1222
},
{
"epoch": 0.4294241573033708,
"grad_norm": 0.9611704349517822,
"learning_rate": 4.946281572008982e-06,
"loss": 0.7044,
"step": 1223
},
{
"epoch": 0.4297752808988764,
"grad_norm": 0.9443504810333252,
"learning_rate": 4.946186204768343e-06,
"loss": 0.7322,
"step": 1224
},
{
"epoch": 0.430126404494382,
"grad_norm": 0.9180960059165955,
"learning_rate": 4.946090753870229e-06,
"loss": 0.7435,
"step": 1225
},
{
"epoch": 0.43047752808988765,
"grad_norm": 0.9297470450401306,
"learning_rate": 4.945995219317903e-06,
"loss": 0.7211,
"step": 1226
},
{
"epoch": 0.43082865168539325,
"grad_norm": 0.9243416786193848,
"learning_rate": 4.945899601114634e-06,
"loss": 0.7255,
"step": 1227
},
{
"epoch": 0.4311797752808989,
"grad_norm": 0.9293515086174011,
"learning_rate": 4.945803899263692e-06,
"loss": 0.7466,
"step": 1228
},
{
"epoch": 0.4315308988764045,
"grad_norm": 0.8804196715354919,
"learning_rate": 4.94570811376835e-06,
"loss": 0.6495,
"step": 1229
},
{
"epoch": 0.4318820224719101,
"grad_norm": 0.8709548115730286,
"learning_rate": 4.945612244631884e-06,
"loss": 0.6573,
"step": 1230
},
{
"epoch": 0.43223314606741575,
"grad_norm": 0.9267548322677612,
"learning_rate": 4.94551629185757e-06,
"loss": 0.6949,
"step": 1231
},
{
"epoch": 0.43258426966292135,
"grad_norm": 0.9049427509307861,
"learning_rate": 4.945420255448693e-06,
"loss": 0.6724,
"step": 1232
},
{
"epoch": 0.43293539325842695,
"grad_norm": 0.9540070295333862,
"learning_rate": 4.945324135408535e-06,
"loss": 0.7162,
"step": 1233
},
{
"epoch": 0.4332865168539326,
"grad_norm": 0.9069761633872986,
"learning_rate": 4.945227931740384e-06,
"loss": 0.6851,
"step": 1234
},
{
"epoch": 0.4336376404494382,
"grad_norm": 0.9255860447883606,
"learning_rate": 4.94513164444753e-06,
"loss": 0.7108,
"step": 1235
},
{
"epoch": 0.4339887640449438,
"grad_norm": 0.9611523151397705,
"learning_rate": 4.945035273533267e-06,
"loss": 0.7156,
"step": 1236
},
{
"epoch": 0.43433988764044945,
"grad_norm": 0.9171714782714844,
"learning_rate": 4.9449388190008896e-06,
"loss": 0.6744,
"step": 1237
},
{
"epoch": 0.43469101123595505,
"grad_norm": 0.8625194430351257,
"learning_rate": 4.944842280853696e-06,
"loss": 0.6792,
"step": 1238
},
{
"epoch": 0.43504213483146065,
"grad_norm": 0.945565938949585,
"learning_rate": 4.94474565909499e-06,
"loss": 0.7015,
"step": 1239
},
{
"epoch": 0.4353932584269663,
"grad_norm": 0.8865808248519897,
"learning_rate": 4.944648953728073e-06,
"loss": 0.6664,
"step": 1240
},
{
"epoch": 0.4357443820224719,
"grad_norm": 0.8900501728057861,
"learning_rate": 4.9445521647562545e-06,
"loss": 0.6831,
"step": 1241
},
{
"epoch": 0.43609550561797755,
"grad_norm": 0.9169394373893738,
"learning_rate": 4.944455292182844e-06,
"loss": 0.6377,
"step": 1242
},
{
"epoch": 0.43644662921348315,
"grad_norm": 0.9147849082946777,
"learning_rate": 4.944358336011154e-06,
"loss": 0.735,
"step": 1243
},
{
"epoch": 0.43679775280898875,
"grad_norm": 0.9505800604820251,
"learning_rate": 4.9442612962445e-06,
"loss": 0.7309,
"step": 1244
},
{
"epoch": 0.4371488764044944,
"grad_norm": 0.8858315348625183,
"learning_rate": 4.9441641728862015e-06,
"loss": 0.6904,
"step": 1245
},
{
"epoch": 0.4375,
"grad_norm": 0.9335656762123108,
"learning_rate": 4.94406696593958e-06,
"loss": 0.7219,
"step": 1246
},
{
"epoch": 0.4378511235955056,
"grad_norm": 0.8928138613700867,
"learning_rate": 4.94396967540796e-06,
"loss": 0.6521,
"step": 1247
},
{
"epoch": 0.43820224719101125,
"grad_norm": 0.9080971479415894,
"learning_rate": 4.943872301294668e-06,
"loss": 0.6925,
"step": 1248
},
{
"epoch": 0.43855337078651685,
"grad_norm": 0.9383649826049805,
"learning_rate": 4.943774843603034e-06,
"loss": 0.708,
"step": 1249
},
{
"epoch": 0.43890449438202245,
"grad_norm": 0.9101510047912598,
"learning_rate": 4.943677302336393e-06,
"loss": 0.6901,
"step": 1250
},
{
"epoch": 0.4392556179775281,
"grad_norm": 0.9402358531951904,
"learning_rate": 4.943579677498078e-06,
"loss": 0.7169,
"step": 1251
},
{
"epoch": 0.4396067415730337,
"grad_norm": 0.8998861908912659,
"learning_rate": 4.943481969091429e-06,
"loss": 0.6962,
"step": 1252
},
{
"epoch": 0.43995786516853935,
"grad_norm": 0.9605402946472168,
"learning_rate": 4.943384177119788e-06,
"loss": 0.6783,
"step": 1253
},
{
"epoch": 0.44030898876404495,
"grad_norm": 0.9242511987686157,
"learning_rate": 4.9432863015865e-06,
"loss": 0.71,
"step": 1254
},
{
"epoch": 0.44066011235955055,
"grad_norm": 0.9048755168914795,
"learning_rate": 4.943188342494909e-06,
"loss": 0.7143,
"step": 1255
},
{
"epoch": 0.4410112359550562,
"grad_norm": 0.883234977722168,
"learning_rate": 4.9430902998483685e-06,
"loss": 0.6922,
"step": 1256
},
{
"epoch": 0.4413623595505618,
"grad_norm": 0.9546087980270386,
"learning_rate": 4.942992173650231e-06,
"loss": 0.725,
"step": 1257
},
{
"epoch": 0.4417134831460674,
"grad_norm": 0.9286230206489563,
"learning_rate": 4.942893963903851e-06,
"loss": 0.7263,
"step": 1258
},
{
"epoch": 0.44206460674157305,
"grad_norm": 0.901672899723053,
"learning_rate": 4.942795670612588e-06,
"loss": 0.6826,
"step": 1259
},
{
"epoch": 0.44241573033707865,
"grad_norm": 0.9127205610275269,
"learning_rate": 4.9426972937798035e-06,
"loss": 0.6727,
"step": 1260
},
{
"epoch": 0.44276685393258425,
"grad_norm": 0.9450622797012329,
"learning_rate": 4.942598833408861e-06,
"loss": 0.726,
"step": 1261
},
{
"epoch": 0.4431179775280899,
"grad_norm": 0.9145720601081848,
"learning_rate": 4.94250028950313e-06,
"loss": 0.702,
"step": 1262
},
{
"epoch": 0.4434691011235955,
"grad_norm": 0.9019625782966614,
"learning_rate": 4.9424016620659786e-06,
"loss": 0.6682,
"step": 1263
},
{
"epoch": 0.4438202247191011,
"grad_norm": 0.9257623553276062,
"learning_rate": 4.9423029511007795e-06,
"loss": 0.6981,
"step": 1264
},
{
"epoch": 0.44417134831460675,
"grad_norm": 0.9080988764762878,
"learning_rate": 4.94220415661091e-06,
"loss": 0.708,
"step": 1265
},
{
"epoch": 0.44452247191011235,
"grad_norm": 0.9560781121253967,
"learning_rate": 4.942105278599748e-06,
"loss": 0.7021,
"step": 1266
},
{
"epoch": 0.444873595505618,
"grad_norm": 0.9422893524169922,
"learning_rate": 4.942006317070676e-06,
"loss": 0.6861,
"step": 1267
},
{
"epoch": 0.4452247191011236,
"grad_norm": 0.8909929990768433,
"learning_rate": 4.9419072720270764e-06,
"loss": 0.7143,
"step": 1268
},
{
"epoch": 0.4455758426966292,
"grad_norm": 0.9080603718757629,
"learning_rate": 4.941808143472339e-06,
"loss": 0.7073,
"step": 1269
},
{
"epoch": 0.44592696629213485,
"grad_norm": 0.8691872358322144,
"learning_rate": 4.941708931409851e-06,
"loss": 0.6425,
"step": 1270
},
{
"epoch": 0.44627808988764045,
"grad_norm": 0.9223096966743469,
"learning_rate": 4.9416096358430084e-06,
"loss": 0.6876,
"step": 1271
},
{
"epoch": 0.44662921348314605,
"grad_norm": 0.9030603766441345,
"learning_rate": 4.9415102567752045e-06,
"loss": 0.6893,
"step": 1272
},
{
"epoch": 0.4469803370786517,
"grad_norm": 0.9421838521957397,
"learning_rate": 4.94141079420984e-06,
"loss": 0.7267,
"step": 1273
},
{
"epoch": 0.4473314606741573,
"grad_norm": 0.8724151253700256,
"learning_rate": 4.941311248150314e-06,
"loss": 0.6587,
"step": 1274
},
{
"epoch": 0.4476825842696629,
"grad_norm": 0.8822656869888306,
"learning_rate": 4.9412116186000346e-06,
"loss": 0.6969,
"step": 1275
},
{
"epoch": 0.44803370786516855,
"grad_norm": 0.8846755623817444,
"learning_rate": 4.941111905562405e-06,
"loss": 0.6848,
"step": 1276
},
{
"epoch": 0.44838483146067415,
"grad_norm": 0.8999640941619873,
"learning_rate": 4.941012109040838e-06,
"loss": 0.6465,
"step": 1277
},
{
"epoch": 0.4487359550561798,
"grad_norm": 0.9194796681404114,
"learning_rate": 4.940912229038746e-06,
"loss": 0.6995,
"step": 1278
},
{
"epoch": 0.4490870786516854,
"grad_norm": 0.9296586513519287,
"learning_rate": 4.940812265559543e-06,
"loss": 0.6399,
"step": 1279
},
{
"epoch": 0.449438202247191,
"grad_norm": 0.9012690782546997,
"learning_rate": 4.94071221860665e-06,
"loss": 0.7109,
"step": 1280
},
{
"epoch": 0.44978932584269665,
"grad_norm": 0.9507624506950378,
"learning_rate": 4.940612088183488e-06,
"loss": 0.7498,
"step": 1281
},
{
"epoch": 0.45014044943820225,
"grad_norm": 0.8964364528656006,
"learning_rate": 4.940511874293481e-06,
"loss": 0.674,
"step": 1282
},
{
"epoch": 0.45049157303370785,
"grad_norm": 0.9396047592163086,
"learning_rate": 4.940411576940056e-06,
"loss": 0.699,
"step": 1283
},
{
"epoch": 0.4508426966292135,
"grad_norm": 0.9038118720054626,
"learning_rate": 4.940311196126643e-06,
"loss": 0.6598,
"step": 1284
},
{
"epoch": 0.4511938202247191,
"grad_norm": 0.9001778960227966,
"learning_rate": 4.940210731856675e-06,
"loss": 0.7129,
"step": 1285
},
{
"epoch": 0.4515449438202247,
"grad_norm": 0.9313659071922302,
"learning_rate": 4.94011018413359e-06,
"loss": 0.7138,
"step": 1286
},
{
"epoch": 0.45189606741573035,
"grad_norm": 0.9225174784660339,
"learning_rate": 4.9400095529608226e-06,
"loss": 0.7192,
"step": 1287
},
{
"epoch": 0.45224719101123595,
"grad_norm": 0.9100403189659119,
"learning_rate": 4.939908838341818e-06,
"loss": 0.6593,
"step": 1288
},
{
"epoch": 0.45259831460674155,
"grad_norm": 0.8900116086006165,
"learning_rate": 4.939808040280017e-06,
"loss": 0.6799,
"step": 1289
},
{
"epoch": 0.4529494382022472,
"grad_norm": 0.885908305644989,
"learning_rate": 4.939707158778872e-06,
"loss": 0.6889,
"step": 1290
},
{
"epoch": 0.4533005617977528,
"grad_norm": 0.9219464659690857,
"learning_rate": 4.939606193841827e-06,
"loss": 0.7198,
"step": 1291
},
{
"epoch": 0.45365168539325845,
"grad_norm": 0.9270339012145996,
"learning_rate": 4.939505145472339e-06,
"loss": 0.7121,
"step": 1292
},
{
"epoch": 0.45400280898876405,
"grad_norm": 0.9297556281089783,
"learning_rate": 4.939404013673862e-06,
"loss": 0.7119,
"step": 1293
},
{
"epoch": 0.45435393258426965,
"grad_norm": 0.9266655445098877,
"learning_rate": 4.9393027984498555e-06,
"loss": 0.6994,
"step": 1294
},
{
"epoch": 0.4547050561797753,
"grad_norm": 0.9152787923812866,
"learning_rate": 4.939201499803781e-06,
"loss": 0.6618,
"step": 1295
},
{
"epoch": 0.4550561797752809,
"grad_norm": 0.9240756034851074,
"learning_rate": 4.939100117739102e-06,
"loss": 0.7359,
"step": 1296
},
{
"epoch": 0.4554073033707865,
"grad_norm": 0.9413486123085022,
"learning_rate": 4.9389986522592855e-06,
"loss": 0.7266,
"step": 1297
},
{
"epoch": 0.45575842696629215,
"grad_norm": 0.8976860046386719,
"learning_rate": 4.938897103367803e-06,
"loss": 0.7063,
"step": 1298
},
{
"epoch": 0.45610955056179775,
"grad_norm": 0.9674167633056641,
"learning_rate": 4.938795471068126e-06,
"loss": 0.7191,
"step": 1299
},
{
"epoch": 0.45646067415730335,
"grad_norm": 0.9469913840293884,
"learning_rate": 4.938693755363731e-06,
"loss": 0.7263,
"step": 1300
},
{
"epoch": 0.456811797752809,
"grad_norm": 0.9106466174125671,
"learning_rate": 4.938591956258097e-06,
"loss": 0.7324,
"step": 1301
},
{
"epoch": 0.4571629213483146,
"grad_norm": 0.9906820058822632,
"learning_rate": 4.938490073754705e-06,
"loss": 0.7273,
"step": 1302
},
{
"epoch": 0.4575140449438202,
"grad_norm": 0.9152560234069824,
"learning_rate": 4.938388107857038e-06,
"loss": 0.6648,
"step": 1303
},
{
"epoch": 0.45786516853932585,
"grad_norm": 0.9459682703018188,
"learning_rate": 4.938286058568585e-06,
"loss": 0.6869,
"step": 1304
},
{
"epoch": 0.45821629213483145,
"grad_norm": 0.9387467503547668,
"learning_rate": 4.938183925892835e-06,
"loss": 0.7028,
"step": 1305
},
{
"epoch": 0.4585674157303371,
"grad_norm": 0.9118907451629639,
"learning_rate": 4.938081709833281e-06,
"loss": 0.6875,
"step": 1306
},
{
"epoch": 0.4589185393258427,
"grad_norm": 0.9032610654830933,
"learning_rate": 4.937979410393419e-06,
"loss": 0.7157,
"step": 1307
},
{
"epoch": 0.4592696629213483,
"grad_norm": 0.9495104551315308,
"learning_rate": 4.937877027576748e-06,
"loss": 0.6914,
"step": 1308
},
{
"epoch": 0.45962078651685395,
"grad_norm": 0.8904275894165039,
"learning_rate": 4.937774561386768e-06,
"loss": 0.7351,
"step": 1309
},
{
"epoch": 0.45997191011235955,
"grad_norm": 0.8993387222290039,
"learning_rate": 4.937672011826984e-06,
"loss": 0.7004,
"step": 1310
},
{
"epoch": 0.46032303370786515,
"grad_norm": 0.9243264198303223,
"learning_rate": 4.937569378900903e-06,
"loss": 0.7059,
"step": 1311
},
{
"epoch": 0.4606741573033708,
"grad_norm": 0.8777183294296265,
"learning_rate": 4.937466662612036e-06,
"loss": 0.6795,
"step": 1312
},
{
"epoch": 0.4610252808988764,
"grad_norm": 0.9348961114883423,
"learning_rate": 4.937363862963895e-06,
"loss": 0.6726,
"step": 1313
},
{
"epoch": 0.461376404494382,
"grad_norm": 0.9088345766067505,
"learning_rate": 4.937260979959994e-06,
"loss": 0.7114,
"step": 1314
},
{
"epoch": 0.46172752808988765,
"grad_norm": 0.9143536686897278,
"learning_rate": 4.937158013603854e-06,
"loss": 0.6819,
"step": 1315
},
{
"epoch": 0.46207865168539325,
"grad_norm": 0.9480457901954651,
"learning_rate": 4.937054963898996e-06,
"loss": 0.7335,
"step": 1316
},
{
"epoch": 0.4624297752808989,
"grad_norm": 0.9207438230514526,
"learning_rate": 4.9369518308489425e-06,
"loss": 0.7091,
"step": 1317
},
{
"epoch": 0.4627808988764045,
"grad_norm": 0.9547027349472046,
"learning_rate": 4.936848614457223e-06,
"loss": 0.7388,
"step": 1318
},
{
"epoch": 0.4631320224719101,
"grad_norm": 0.9442094564437866,
"learning_rate": 4.9367453147273645e-06,
"loss": 0.709,
"step": 1319
},
{
"epoch": 0.46348314606741575,
"grad_norm": 0.9056798815727234,
"learning_rate": 4.9366419316629024e-06,
"loss": 0.6919,
"step": 1320
},
{
"epoch": 0.46383426966292135,
"grad_norm": 0.8800448179244995,
"learning_rate": 4.936538465267371e-06,
"loss": 0.7141,
"step": 1321
},
{
"epoch": 0.46418539325842695,
"grad_norm": 0.9331648349761963,
"learning_rate": 4.93643491554431e-06,
"loss": 0.6871,
"step": 1322
},
{
"epoch": 0.4645365168539326,
"grad_norm": 0.9699183702468872,
"learning_rate": 4.9363312824972586e-06,
"loss": 0.729,
"step": 1323
},
{
"epoch": 0.4648876404494382,
"grad_norm": 0.9243112206459045,
"learning_rate": 4.9362275661297634e-06,
"loss": 0.6751,
"step": 1324
},
{
"epoch": 0.4652387640449438,
"grad_norm": 0.9310767650604248,
"learning_rate": 4.936123766445368e-06,
"loss": 0.7263,
"step": 1325
},
{
"epoch": 0.46558988764044945,
"grad_norm": 0.9163933992385864,
"learning_rate": 4.936019883447627e-06,
"loss": 0.715,
"step": 1326
},
{
"epoch": 0.46594101123595505,
"grad_norm": 0.9764875173568726,
"learning_rate": 4.935915917140089e-06,
"loss": 0.7139,
"step": 1327
},
{
"epoch": 0.46629213483146065,
"grad_norm": 0.9456989169120789,
"learning_rate": 4.935811867526311e-06,
"loss": 0.6667,
"step": 1328
},
{
"epoch": 0.4666432584269663,
"grad_norm": 0.9617751836776733,
"learning_rate": 4.935707734609852e-06,
"loss": 0.691,
"step": 1329
},
{
"epoch": 0.4669943820224719,
"grad_norm": 0.8945748209953308,
"learning_rate": 4.935603518394273e-06,
"loss": 0.711,
"step": 1330
},
{
"epoch": 0.46734550561797755,
"grad_norm": 0.9066562652587891,
"learning_rate": 4.935499218883137e-06,
"loss": 0.685,
"step": 1331
},
{
"epoch": 0.46769662921348315,
"grad_norm": 0.8701803684234619,
"learning_rate": 4.935394836080012e-06,
"loss": 0.6735,
"step": 1332
},
{
"epoch": 0.46804775280898875,
"grad_norm": 0.9355669021606445,
"learning_rate": 4.935290369988468e-06,
"loss": 0.7335,
"step": 1333
},
{
"epoch": 0.4683988764044944,
"grad_norm": 0.882905900478363,
"learning_rate": 4.9351858206120774e-06,
"loss": 0.7196,
"step": 1334
},
{
"epoch": 0.46875,
"grad_norm": 0.9127156734466553,
"learning_rate": 4.935081187954415e-06,
"loss": 0.714,
"step": 1335
},
{
"epoch": 0.4691011235955056,
"grad_norm": 1.284045934677124,
"learning_rate": 4.9349764720190595e-06,
"loss": 0.7167,
"step": 1336
},
{
"epoch": 0.46945224719101125,
"grad_norm": 0.919476330280304,
"learning_rate": 4.934871672809594e-06,
"loss": 0.6804,
"step": 1337
},
{
"epoch": 0.46980337078651685,
"grad_norm": 0.9066704511642456,
"learning_rate": 4.934766790329599e-06,
"loss": 0.706,
"step": 1338
},
{
"epoch": 0.47015449438202245,
"grad_norm": 0.9020872116088867,
"learning_rate": 4.934661824582664e-06,
"loss": 0.7277,
"step": 1339
},
{
"epoch": 0.4705056179775281,
"grad_norm": 0.9096556901931763,
"learning_rate": 4.934556775572377e-06,
"loss": 0.7184,
"step": 1340
},
{
"epoch": 0.4708567415730337,
"grad_norm": 0.8852599859237671,
"learning_rate": 4.934451643302332e-06,
"loss": 0.6952,
"step": 1341
},
{
"epoch": 0.47120786516853935,
"grad_norm": 0.9317033290863037,
"learning_rate": 4.934346427776125e-06,
"loss": 0.7087,
"step": 1342
},
{
"epoch": 0.47155898876404495,
"grad_norm": 0.9349064230918884,
"learning_rate": 4.934241128997353e-06,
"loss": 0.7295,
"step": 1343
},
{
"epoch": 0.47191011235955055,
"grad_norm": 0.8978156447410583,
"learning_rate": 4.934135746969617e-06,
"loss": 0.7129,
"step": 1344
},
{
"epoch": 0.4722612359550562,
"grad_norm": 0.8900869488716125,
"learning_rate": 4.934030281696521e-06,
"loss": 0.6737,
"step": 1345
},
{
"epoch": 0.4726123595505618,
"grad_norm": 0.9193541407585144,
"learning_rate": 4.933924733181673e-06,
"loss": 0.7189,
"step": 1346
},
{
"epoch": 0.4729634831460674,
"grad_norm": 0.8839302659034729,
"learning_rate": 4.933819101428681e-06,
"loss": 0.6883,
"step": 1347
},
{
"epoch": 0.47331460674157305,
"grad_norm": 0.8768748641014099,
"learning_rate": 4.933713386441159e-06,
"loss": 0.6825,
"step": 1348
},
{
"epoch": 0.47366573033707865,
"grad_norm": 0.9012611508369446,
"learning_rate": 4.933607588222721e-06,
"loss": 0.7477,
"step": 1349
},
{
"epoch": 0.47401685393258425,
"grad_norm": 0.9348760843276978,
"learning_rate": 4.933501706776986e-06,
"loss": 0.7071,
"step": 1350
},
{
"epoch": 0.4743679775280899,
"grad_norm": 0.9126866459846497,
"learning_rate": 4.933395742107576e-06,
"loss": 0.7259,
"step": 1351
},
{
"epoch": 0.4747191011235955,
"grad_norm": 1.1260316371917725,
"learning_rate": 4.9332896942181135e-06,
"loss": 0.6647,
"step": 1352
},
{
"epoch": 0.4750702247191011,
"grad_norm": 0.9544892907142639,
"learning_rate": 4.933183563112226e-06,
"loss": 0.7228,
"step": 1353
},
{
"epoch": 0.47542134831460675,
"grad_norm": 0.9179894924163818,
"learning_rate": 4.933077348793542e-06,
"loss": 0.7115,
"step": 1354
},
{
"epoch": 0.47577247191011235,
"grad_norm": 0.8785374164581299,
"learning_rate": 4.932971051265695e-06,
"loss": 0.6651,
"step": 1355
},
{
"epoch": 0.476123595505618,
"grad_norm": 0.9006591439247131,
"learning_rate": 4.93286467053232e-06,
"loss": 0.7166,
"step": 1356
},
{
"epoch": 0.4764747191011236,
"grad_norm": 0.9105557799339294,
"learning_rate": 4.932758206597055e-06,
"loss": 0.6876,
"step": 1357
},
{
"epoch": 0.4768258426966292,
"grad_norm": 0.8968350887298584,
"learning_rate": 4.932651659463541e-06,
"loss": 0.6873,
"step": 1358
},
{
"epoch": 0.47717696629213485,
"grad_norm": 0.9052225947380066,
"learning_rate": 4.932545029135422e-06,
"loss": 0.7134,
"step": 1359
},
{
"epoch": 0.47752808988764045,
"grad_norm": 0.946305513381958,
"learning_rate": 4.932438315616345e-06,
"loss": 0.7061,
"step": 1360
},
{
"epoch": 0.47787921348314605,
"grad_norm": 0.9128836393356323,
"learning_rate": 4.9323315189099586e-06,
"loss": 0.7101,
"step": 1361
},
{
"epoch": 0.4782303370786517,
"grad_norm": 0.9199392199516296,
"learning_rate": 4.9322246390199155e-06,
"loss": 0.7073,
"step": 1362
},
{
"epoch": 0.4785814606741573,
"grad_norm": 1.077110767364502,
"learning_rate": 4.932117675949871e-06,
"loss": 0.6828,
"step": 1363
},
{
"epoch": 0.4789325842696629,
"grad_norm": 0.9046890735626221,
"learning_rate": 4.932010629703484e-06,
"loss": 0.6991,
"step": 1364
},
{
"epoch": 0.47928370786516855,
"grad_norm": 0.8711754083633423,
"learning_rate": 4.931903500284414e-06,
"loss": 0.6788,
"step": 1365
},
{
"epoch": 0.47963483146067415,
"grad_norm": 0.8938144445419312,
"learning_rate": 4.931796287696325e-06,
"loss": 0.7219,
"step": 1366
},
{
"epoch": 0.4799859550561798,
"grad_norm": 0.9003678560256958,
"learning_rate": 4.931688991942884e-06,
"loss": 0.6908,
"step": 1367
},
{
"epoch": 0.4803370786516854,
"grad_norm": 0.9107855558395386,
"learning_rate": 4.93158161302776e-06,
"loss": 0.6821,
"step": 1368
},
{
"epoch": 0.480688202247191,
"grad_norm": 0.9014522433280945,
"learning_rate": 4.931474150954626e-06,
"loss": 0.7071,
"step": 1369
},
{
"epoch": 0.48103932584269665,
"grad_norm": 0.9273253083229065,
"learning_rate": 4.931366605727157e-06,
"loss": 0.7027,
"step": 1370
},
{
"epoch": 0.48139044943820225,
"grad_norm": 0.9555942416191101,
"learning_rate": 4.93125897734903e-06,
"loss": 0.6952,
"step": 1371
},
{
"epoch": 0.48174157303370785,
"grad_norm": 0.9457711577415466,
"learning_rate": 4.931151265823927e-06,
"loss": 0.6893,
"step": 1372
},
{
"epoch": 0.4820926966292135,
"grad_norm": 0.9268205761909485,
"learning_rate": 4.931043471155531e-06,
"loss": 0.7169,
"step": 1373
},
{
"epoch": 0.4824438202247191,
"grad_norm": 0.9136787056922913,
"learning_rate": 4.930935593347529e-06,
"loss": 0.7326,
"step": 1374
},
{
"epoch": 0.4827949438202247,
"grad_norm": 0.925116777420044,
"learning_rate": 4.930827632403608e-06,
"loss": 0.7037,
"step": 1375
},
{
"epoch": 0.48314606741573035,
"grad_norm": 0.9161933064460754,
"learning_rate": 4.930719588327464e-06,
"loss": 0.7222,
"step": 1376
},
{
"epoch": 0.48349719101123595,
"grad_norm": 0.9238081574440002,
"learning_rate": 4.93061146112279e-06,
"loss": 0.7186,
"step": 1377
},
{
"epoch": 0.48384831460674155,
"grad_norm": 0.9430643320083618,
"learning_rate": 4.930503250793283e-06,
"loss": 0.7622,
"step": 1378
},
{
"epoch": 0.4841994382022472,
"grad_norm": 0.9295469522476196,
"learning_rate": 4.930394957342645e-06,
"loss": 0.7354,
"step": 1379
},
{
"epoch": 0.4845505617977528,
"grad_norm": 0.9064199924468994,
"learning_rate": 4.9302865807745795e-06,
"loss": 0.6487,
"step": 1380
},
{
"epoch": 0.48490168539325845,
"grad_norm": 0.9455371499061584,
"learning_rate": 4.930178121092792e-06,
"loss": 0.7468,
"step": 1381
},
{
"epoch": 0.48525280898876405,
"grad_norm": 0.9263501763343811,
"learning_rate": 4.930069578300992e-06,
"loss": 0.7099,
"step": 1382
},
{
"epoch": 0.48560393258426965,
"grad_norm": 0.8904842734336853,
"learning_rate": 4.929960952402892e-06,
"loss": 0.6519,
"step": 1383
},
{
"epoch": 0.4859550561797753,
"grad_norm": 0.8946258425712585,
"learning_rate": 4.929852243402208e-06,
"loss": 0.7011,
"step": 1384
},
{
"epoch": 0.4863061797752809,
"grad_norm": 0.9584969282150269,
"learning_rate": 4.929743451302656e-06,
"loss": 0.724,
"step": 1385
},
{
"epoch": 0.4866573033707865,
"grad_norm": 0.9212526082992554,
"learning_rate": 4.929634576107956e-06,
"loss": 0.6821,
"step": 1386
},
{
"epoch": 0.48700842696629215,
"grad_norm": 0.9416231513023376,
"learning_rate": 4.929525617821834e-06,
"loss": 0.7457,
"step": 1387
},
{
"epoch": 0.48735955056179775,
"grad_norm": 0.9377028346061707,
"learning_rate": 4.929416576448014e-06,
"loss": 0.6871,
"step": 1388
},
{
"epoch": 0.48771067415730335,
"grad_norm": 0.9065041542053223,
"learning_rate": 4.929307451990226e-06,
"loss": 0.7261,
"step": 1389
},
{
"epoch": 0.488061797752809,
"grad_norm": 0.844350278377533,
"learning_rate": 4.929198244452202e-06,
"loss": 0.7099,
"step": 1390
},
{
"epoch": 0.4884129213483146,
"grad_norm": 0.9054086804389954,
"learning_rate": 4.9290889538376775e-06,
"loss": 0.669,
"step": 1391
},
{
"epoch": 0.4887640449438202,
"grad_norm": 0.8729206919670105,
"learning_rate": 4.928979580150388e-06,
"loss": 0.6816,
"step": 1392
},
{
"epoch": 0.48911516853932585,
"grad_norm": 0.9076043367385864,
"learning_rate": 4.928870123394077e-06,
"loss": 0.7044,
"step": 1393
},
{
"epoch": 0.48946629213483145,
"grad_norm": 0.9312844276428223,
"learning_rate": 4.928760583572485e-06,
"loss": 0.7107,
"step": 1394
},
{
"epoch": 0.4898174157303371,
"grad_norm": 0.9900015592575073,
"learning_rate": 4.928650960689359e-06,
"loss": 0.703,
"step": 1395
},
{
"epoch": 0.4901685393258427,
"grad_norm": 0.9289678931236267,
"learning_rate": 4.92854125474845e-06,
"loss": 0.714,
"step": 1396
},
{
"epoch": 0.4905196629213483,
"grad_norm": 0.9454885125160217,
"learning_rate": 4.928431465753507e-06,
"loss": 0.7312,
"step": 1397
},
{
"epoch": 0.49087078651685395,
"grad_norm": 0.9117372035980225,
"learning_rate": 4.928321593708286e-06,
"loss": 0.6984,
"step": 1398
},
{
"epoch": 0.49122191011235955,
"grad_norm": 0.9010490775108337,
"learning_rate": 4.928211638616545e-06,
"loss": 0.672,
"step": 1399
},
{
"epoch": 0.49157303370786515,
"grad_norm": 0.9090813994407654,
"learning_rate": 4.928101600482043e-06,
"loss": 0.7132,
"step": 1400
},
{
"epoch": 0.4919241573033708,
"grad_norm": 0.905497133731842,
"learning_rate": 4.927991479308545e-06,
"loss": 0.677,
"step": 1401
},
{
"epoch": 0.4922752808988764,
"grad_norm": 0.902889609336853,
"learning_rate": 4.927881275099815e-06,
"loss": 0.6601,
"step": 1402
},
{
"epoch": 0.492626404494382,
"grad_norm": 0.9104518294334412,
"learning_rate": 4.927770987859624e-06,
"loss": 0.7192,
"step": 1403
},
{
"epoch": 0.49297752808988765,
"grad_norm": 0.8784206509590149,
"learning_rate": 4.9276606175917415e-06,
"loss": 0.6556,
"step": 1404
},
{
"epoch": 0.49332865168539325,
"grad_norm": 0.9406437277793884,
"learning_rate": 4.927550164299944e-06,
"loss": 0.679,
"step": 1405
},
{
"epoch": 0.4936797752808989,
"grad_norm": 0.8956720232963562,
"learning_rate": 4.9274396279880075e-06,
"loss": 0.662,
"step": 1406
},
{
"epoch": 0.4940308988764045,
"grad_norm": 0.9246423840522766,
"learning_rate": 4.927329008659714e-06,
"loss": 0.7155,
"step": 1407
},
{
"epoch": 0.4943820224719101,
"grad_norm": 0.9392451047897339,
"learning_rate": 4.927218306318845e-06,
"loss": 0.6819,
"step": 1408
},
{
"epoch": 0.49473314606741575,
"grad_norm": 0.9309847950935364,
"learning_rate": 4.927107520969187e-06,
"loss": 0.7195,
"step": 1409
},
{
"epoch": 0.49508426966292135,
"grad_norm": 0.890052080154419,
"learning_rate": 4.926996652614529e-06,
"loss": 0.6703,
"step": 1410
},
{
"epoch": 0.49543539325842695,
"grad_norm": 0.9011236429214478,
"learning_rate": 4.926885701258661e-06,
"loss": 0.6822,
"step": 1411
},
{
"epoch": 0.4957865168539326,
"grad_norm": 0.915877103805542,
"learning_rate": 4.92677466690538e-06,
"loss": 0.7156,
"step": 1412
},
{
"epoch": 0.4961376404494382,
"grad_norm": 0.8800505995750427,
"learning_rate": 4.9266635495584814e-06,
"loss": 0.6817,
"step": 1413
},
{
"epoch": 0.4964887640449438,
"grad_norm": 0.9167606234550476,
"learning_rate": 4.926552349221766e-06,
"loss": 0.6982,
"step": 1414
},
{
"epoch": 0.49683988764044945,
"grad_norm": 0.9250441193580627,
"learning_rate": 4.926441065899037e-06,
"loss": 0.6559,
"step": 1415
},
{
"epoch": 0.49719101123595505,
"grad_norm": 0.910899817943573,
"learning_rate": 4.9263296995941004e-06,
"loss": 0.6654,
"step": 1416
},
{
"epoch": 0.49754213483146065,
"grad_norm": 0.8955602049827576,
"learning_rate": 4.926218250310764e-06,
"loss": 0.7022,
"step": 1417
},
{
"epoch": 0.4978932584269663,
"grad_norm": 0.9195224642753601,
"learning_rate": 4.926106718052839e-06,
"loss": 0.7058,
"step": 1418
},
{
"epoch": 0.4982443820224719,
"grad_norm": 0.9715495705604553,
"learning_rate": 4.92599510282414e-06,
"loss": 0.6917,
"step": 1419
},
{
"epoch": 0.49859550561797755,
"grad_norm": 0.8591254353523254,
"learning_rate": 4.925883404628485e-06,
"loss": 0.6415,
"step": 1420
},
{
"epoch": 0.49894662921348315,
"grad_norm": 0.903109610080719,
"learning_rate": 4.925771623469694e-06,
"loss": 0.702,
"step": 1421
},
{
"epoch": 0.49929775280898875,
"grad_norm": 0.9081088900566101,
"learning_rate": 4.9256597593515885e-06,
"loss": 0.7098,
"step": 1422
},
{
"epoch": 0.4996488764044944,
"grad_norm": 0.9099250435829163,
"learning_rate": 4.925547812277994e-06,
"loss": 0.6752,
"step": 1423
},
{
"epoch": 0.5,
"grad_norm": 0.9163993000984192,
"learning_rate": 4.925435782252742e-06,
"loss": 0.7098,
"step": 1424
},
{
"epoch": 0.5003511235955056,
"grad_norm": 0.9282369017601013,
"learning_rate": 4.92532366927966e-06,
"loss": 0.7003,
"step": 1425
},
{
"epoch": 0.5007022471910112,
"grad_norm": 0.9375564455986023,
"learning_rate": 4.925211473362585e-06,
"loss": 0.6981,
"step": 1426
},
{
"epoch": 0.5010533707865169,
"grad_norm": 0.9100632071495056,
"learning_rate": 4.925099194505352e-06,
"loss": 0.6794,
"step": 1427
},
{
"epoch": 0.5014044943820225,
"grad_norm": 0.8592130541801453,
"learning_rate": 4.924986832711802e-06,
"loss": 0.6547,
"step": 1428
},
{
"epoch": 0.5017556179775281,
"grad_norm": 0.9379070401191711,
"learning_rate": 4.924874387985777e-06,
"loss": 0.6829,
"step": 1429
},
{
"epoch": 0.5021067415730337,
"grad_norm": 0.9097589254379272,
"learning_rate": 4.924761860331125e-06,
"loss": 0.682,
"step": 1430
},
{
"epoch": 0.5024578651685393,
"grad_norm": 0.925594687461853,
"learning_rate": 4.92464924975169e-06,
"loss": 0.7312,
"step": 1431
},
{
"epoch": 0.5028089887640449,
"grad_norm": 0.8913887739181519,
"learning_rate": 4.924536556251326e-06,
"loss": 0.6613,
"step": 1432
},
{
"epoch": 0.5031601123595506,
"grad_norm": 0.8830445408821106,
"learning_rate": 4.924423779833887e-06,
"loss": 0.689,
"step": 1433
},
{
"epoch": 0.5035112359550562,
"grad_norm": 0.9521420001983643,
"learning_rate": 4.924310920503229e-06,
"loss": 0.678,
"step": 1434
},
{
"epoch": 0.5038623595505618,
"grad_norm": 0.918493390083313,
"learning_rate": 4.924197978263212e-06,
"loss": 0.7188,
"step": 1435
},
{
"epoch": 0.5042134831460674,
"grad_norm": 0.9049700498580933,
"learning_rate": 4.9240849531176986e-06,
"loss": 0.6846,
"step": 1436
},
{
"epoch": 0.504564606741573,
"grad_norm": 0.8993126153945923,
"learning_rate": 4.9239718450705545e-06,
"loss": 0.7298,
"step": 1437
},
{
"epoch": 0.5049157303370787,
"grad_norm": 0.9134989380836487,
"learning_rate": 4.923858654125647e-06,
"loss": 0.6945,
"step": 1438
},
{
"epoch": 0.5052668539325843,
"grad_norm": 0.9160113334655762,
"learning_rate": 4.9237453802868475e-06,
"loss": 0.7074,
"step": 1439
},
{
"epoch": 0.5056179775280899,
"grad_norm": 0.9044978022575378,
"learning_rate": 4.923632023558031e-06,
"loss": 0.6814,
"step": 1440
},
{
"epoch": 0.5059691011235955,
"grad_norm": 0.9281947612762451,
"learning_rate": 4.923518583943072e-06,
"loss": 0.6806,
"step": 1441
},
{
"epoch": 0.5063202247191011,
"grad_norm": 0.8679924607276917,
"learning_rate": 4.923405061445852e-06,
"loss": 0.6654,
"step": 1442
},
{
"epoch": 0.5066713483146067,
"grad_norm": 0.905013918876648,
"learning_rate": 4.923291456070252e-06,
"loss": 0.6938,
"step": 1443
},
{
"epoch": 0.5070224719101124,
"grad_norm": 0.8952536582946777,
"learning_rate": 4.923177767820158e-06,
"loss": 0.6802,
"step": 1444
},
{
"epoch": 0.507373595505618,
"grad_norm": 0.9175031185150146,
"learning_rate": 4.923063996699458e-06,
"loss": 0.6984,
"step": 1445
},
{
"epoch": 0.5077247191011236,
"grad_norm": 0.8895412087440491,
"learning_rate": 4.922950142712044e-06,
"loss": 0.7255,
"step": 1446
},
{
"epoch": 0.5080758426966292,
"grad_norm": 0.874991238117218,
"learning_rate": 4.922836205861807e-06,
"loss": 0.6709,
"step": 1447
},
{
"epoch": 0.5084269662921348,
"grad_norm": 0.8750325441360474,
"learning_rate": 4.922722186152646e-06,
"loss": 0.6414,
"step": 1448
},
{
"epoch": 0.5087780898876404,
"grad_norm": 0.8989003300666809,
"learning_rate": 4.922608083588459e-06,
"loss": 0.6928,
"step": 1449
},
{
"epoch": 0.5091292134831461,
"grad_norm": 0.9451534152030945,
"learning_rate": 4.922493898173148e-06,
"loss": 0.7143,
"step": 1450
},
{
"epoch": 0.5094803370786517,
"grad_norm": 0.9093509912490845,
"learning_rate": 4.922379629910619e-06,
"loss": 0.7027,
"step": 1451
},
{
"epoch": 0.5098314606741573,
"grad_norm": 0.932784914970398,
"learning_rate": 4.9222652788047795e-06,
"loss": 0.7024,
"step": 1452
},
{
"epoch": 0.5101825842696629,
"grad_norm": 0.9492952227592468,
"learning_rate": 4.92215084485954e-06,
"loss": 0.6856,
"step": 1453
},
{
"epoch": 0.5105337078651685,
"grad_norm": 0.8805875182151794,
"learning_rate": 4.922036328078815e-06,
"loss": 0.7114,
"step": 1454
},
{
"epoch": 0.5108848314606742,
"grad_norm": 0.8962169289588928,
"learning_rate": 4.92192172846652e-06,
"loss": 0.6962,
"step": 1455
},
{
"epoch": 0.5112359550561798,
"grad_norm": 0.923744797706604,
"learning_rate": 4.921807046026574e-06,
"loss": 0.7325,
"step": 1456
},
{
"epoch": 0.5115870786516854,
"grad_norm": 0.9285497069358826,
"learning_rate": 4.9216922807628995e-06,
"loss": 0.7208,
"step": 1457
},
{
"epoch": 0.511938202247191,
"grad_norm": 0.879386842250824,
"learning_rate": 4.921577432679421e-06,
"loss": 0.6404,
"step": 1458
},
{
"epoch": 0.5122893258426966,
"grad_norm": 0.9378132820129395,
"learning_rate": 4.921462501780067e-06,
"loss": 0.7286,
"step": 1459
},
{
"epoch": 0.5126404494382022,
"grad_norm": 0.9210785031318665,
"learning_rate": 4.921347488068767e-06,
"loss": 0.7344,
"step": 1460
},
{
"epoch": 0.5129915730337079,
"grad_norm": 0.9228889346122742,
"learning_rate": 4.921232391549456e-06,
"loss": 0.7108,
"step": 1461
},
{
"epoch": 0.5133426966292135,
"grad_norm": 0.8554604649543762,
"learning_rate": 4.9211172122260675e-06,
"loss": 0.6272,
"step": 1462
},
{
"epoch": 0.5136938202247191,
"grad_norm": 0.8758804202079773,
"learning_rate": 4.921001950102543e-06,
"loss": 0.6463,
"step": 1463
},
{
"epoch": 0.5140449438202247,
"grad_norm": 0.9541190266609192,
"learning_rate": 4.920886605182823e-06,
"loss": 0.7274,
"step": 1464
},
{
"epoch": 0.5143960674157303,
"grad_norm": 0.9359861016273499,
"learning_rate": 4.920771177470853e-06,
"loss": 0.6961,
"step": 1465
},
{
"epoch": 0.514747191011236,
"grad_norm": 0.8696016073226929,
"learning_rate": 4.92065566697058e-06,
"loss": 0.7263,
"step": 1466
},
{
"epoch": 0.5150983146067416,
"grad_norm": 0.8890820741653442,
"learning_rate": 4.920540073685956e-06,
"loss": 0.7063,
"step": 1467
},
{
"epoch": 0.5154494382022472,
"grad_norm": 0.90772545337677,
"learning_rate": 4.920424397620931e-06,
"loss": 0.6858,
"step": 1468
},
{
"epoch": 0.5158005617977528,
"grad_norm": 0.9410818219184875,
"learning_rate": 4.920308638779463e-06,
"loss": 0.6655,
"step": 1469
},
{
"epoch": 0.5161516853932584,
"grad_norm": 0.8993090987205505,
"learning_rate": 4.920192797165511e-06,
"loss": 0.7215,
"step": 1470
},
{
"epoch": 0.516502808988764,
"grad_norm": 0.8889127969741821,
"learning_rate": 4.920076872783036e-06,
"loss": 0.7214,
"step": 1471
},
{
"epoch": 0.5168539325842697,
"grad_norm": 0.9573481678962708,
"learning_rate": 4.919960865636004e-06,
"loss": 0.6754,
"step": 1472
},
{
"epoch": 0.5172050561797753,
"grad_norm": 0.9702395796775818,
"learning_rate": 4.91984477572838e-06,
"loss": 0.6986,
"step": 1473
},
{
"epoch": 0.5175561797752809,
"grad_norm": 0.9364273548126221,
"learning_rate": 4.919728603064136e-06,
"loss": 0.7098,
"step": 1474
},
{
"epoch": 0.5179073033707865,
"grad_norm": 1.0006465911865234,
"learning_rate": 4.919612347647244e-06,
"loss": 0.7143,
"step": 1475
},
{
"epoch": 0.5182584269662921,
"grad_norm": 0.9335517287254333,
"learning_rate": 4.91949600948168e-06,
"loss": 0.6837,
"step": 1476
},
{
"epoch": 0.5186095505617978,
"grad_norm": 0.9131624698638916,
"learning_rate": 4.919379588571424e-06,
"loss": 0.7254,
"step": 1477
},
{
"epoch": 0.5189606741573034,
"grad_norm": 0.9354121685028076,
"learning_rate": 4.919263084920455e-06,
"loss": 0.707,
"step": 1478
},
{
"epoch": 0.519311797752809,
"grad_norm": 0.8909806609153748,
"learning_rate": 4.919146498532758e-06,
"loss": 0.6655,
"step": 1479
},
{
"epoch": 0.5196629213483146,
"grad_norm": 0.9220510721206665,
"learning_rate": 4.919029829412323e-06,
"loss": 0.6859,
"step": 1480
},
{
"epoch": 0.5200140449438202,
"grad_norm": 0.9528585076332092,
"learning_rate": 4.918913077563137e-06,
"loss": 0.694,
"step": 1481
},
{
"epoch": 0.5203651685393258,
"grad_norm": 0.9287347793579102,
"learning_rate": 4.918796242989193e-06,
"loss": 0.6775,
"step": 1482
},
{
"epoch": 0.5207162921348315,
"grad_norm": 0.9118848443031311,
"learning_rate": 4.9186793256944885e-06,
"loss": 0.6869,
"step": 1483
},
{
"epoch": 0.5210674157303371,
"grad_norm": 0.8987846374511719,
"learning_rate": 4.91856232568302e-06,
"loss": 0.6883,
"step": 1484
},
{
"epoch": 0.5214185393258427,
"grad_norm": 0.8864469528198242,
"learning_rate": 4.918445242958789e-06,
"loss": 0.7144,
"step": 1485
},
{
"epoch": 0.5217696629213483,
"grad_norm": 0.8598546385765076,
"learning_rate": 4.918328077525802e-06,
"loss": 0.6752,
"step": 1486
},
{
"epoch": 0.5221207865168539,
"grad_norm": 0.8941078782081604,
"learning_rate": 4.918210829388061e-06,
"loss": 0.6733,
"step": 1487
},
{
"epoch": 0.5224719101123596,
"grad_norm": 0.914069414138794,
"learning_rate": 4.918093498549582e-06,
"loss": 0.7032,
"step": 1488
},
{
"epoch": 0.5228230337078652,
"grad_norm": 0.9472395181655884,
"learning_rate": 4.917976085014373e-06,
"loss": 0.6666,
"step": 1489
},
{
"epoch": 0.5231741573033708,
"grad_norm": 0.9576021432876587,
"learning_rate": 4.9178585887864515e-06,
"loss": 0.7056,
"step": 1490
},
{
"epoch": 0.5235252808988764,
"grad_norm": 0.9019284248352051,
"learning_rate": 4.9177410098698354e-06,
"loss": 0.6841,
"step": 1491
},
{
"epoch": 0.523876404494382,
"grad_norm": 0.8970656394958496,
"learning_rate": 4.917623348268546e-06,
"loss": 0.7169,
"step": 1492
},
{
"epoch": 0.5242275280898876,
"grad_norm": 0.8948590159416199,
"learning_rate": 4.917505603986605e-06,
"loss": 0.6666,
"step": 1493
},
{
"epoch": 0.5245786516853933,
"grad_norm": 0.9220554828643799,
"learning_rate": 4.9173877770280425e-06,
"loss": 0.7249,
"step": 1494
},
{
"epoch": 0.5249297752808989,
"grad_norm": 0.9420086741447449,
"learning_rate": 4.917269867396887e-06,
"loss": 0.7205,
"step": 1495
},
{
"epoch": 0.5252808988764045,
"grad_norm": 0.9278038144111633,
"learning_rate": 4.917151875097169e-06,
"loss": 0.7032,
"step": 1496
},
{
"epoch": 0.5256320224719101,
"grad_norm": 0.8682548999786377,
"learning_rate": 4.917033800132927e-06,
"loss": 0.7,
"step": 1497
},
{
"epoch": 0.5259831460674157,
"grad_norm": 0.8908578753471375,
"learning_rate": 4.916915642508196e-06,
"loss": 0.6891,
"step": 1498
},
{
"epoch": 0.5263342696629213,
"grad_norm": 0.8928547501564026,
"learning_rate": 4.916797402227019e-06,
"loss": 0.7247,
"step": 1499
},
{
"epoch": 0.526685393258427,
"grad_norm": 0.9188054203987122,
"learning_rate": 4.916679079293438e-06,
"loss": 0.7001,
"step": 1500
},
{
"epoch": 0.5270365168539326,
"grad_norm": 0.8950523138046265,
"learning_rate": 4.916560673711502e-06,
"loss": 0.6473,
"step": 1501
},
{
"epoch": 0.5273876404494382,
"grad_norm": 0.9325989484786987,
"learning_rate": 4.916442185485258e-06,
"loss": 0.7096,
"step": 1502
},
{
"epoch": 0.5277387640449438,
"grad_norm": 0.9268350005149841,
"learning_rate": 4.9163236146187586e-06,
"loss": 0.7089,
"step": 1503
},
{
"epoch": 0.5280898876404494,
"grad_norm": 1.009692907333374,
"learning_rate": 4.91620496111606e-06,
"loss": 0.6721,
"step": 1504
},
{
"epoch": 0.5284410112359551,
"grad_norm": 0.8952979445457458,
"learning_rate": 4.916086224981219e-06,
"loss": 0.7074,
"step": 1505
},
{
"epoch": 0.5287921348314607,
"grad_norm": 0.896232008934021,
"learning_rate": 4.915967406218296e-06,
"loss": 0.6765,
"step": 1506
},
{
"epoch": 0.5291432584269663,
"grad_norm": 0.9074482917785645,
"learning_rate": 4.915848504831356e-06,
"loss": 0.7296,
"step": 1507
},
{
"epoch": 0.5294943820224719,
"grad_norm": 0.8722943067550659,
"learning_rate": 4.9157295208244635e-06,
"loss": 0.636,
"step": 1508
},
{
"epoch": 0.5298455056179775,
"grad_norm": 0.917375922203064,
"learning_rate": 4.915610454201689e-06,
"loss": 0.6844,
"step": 1509
},
{
"epoch": 0.5301966292134831,
"grad_norm": 0.8942883610725403,
"learning_rate": 4.9154913049671045e-06,
"loss": 0.6589,
"step": 1510
},
{
"epoch": 0.5305477528089888,
"grad_norm": 0.894601583480835,
"learning_rate": 4.915372073124784e-06,
"loss": 0.6919,
"step": 1511
},
{
"epoch": 0.5308988764044944,
"grad_norm": 1.0230928659439087,
"learning_rate": 4.9152527586788056e-06,
"loss": 0.6949,
"step": 1512
},
{
"epoch": 0.53125,
"grad_norm": 1.0016642808914185,
"learning_rate": 4.915133361633249e-06,
"loss": 0.6792,
"step": 1513
},
{
"epoch": 0.5316011235955056,
"grad_norm": 0.8965452909469604,
"learning_rate": 4.9150138819921985e-06,
"loss": 0.6993,
"step": 1514
},
{
"epoch": 0.5319522471910112,
"grad_norm": 0.9045464396476746,
"learning_rate": 4.91489431975974e-06,
"loss": 0.6905,
"step": 1515
},
{
"epoch": 0.5323033707865169,
"grad_norm": 0.9282147288322449,
"learning_rate": 4.914774674939962e-06,
"loss": 0.7121,
"step": 1516
},
{
"epoch": 0.5326544943820225,
"grad_norm": 0.9581758379936218,
"learning_rate": 4.914654947536956e-06,
"loss": 0.7156,
"step": 1517
},
{
"epoch": 0.5330056179775281,
"grad_norm": 0.8986783623695374,
"learning_rate": 4.9145351375548184e-06,
"loss": 0.6827,
"step": 1518
},
{
"epoch": 0.5333567415730337,
"grad_norm": 1.0093053579330444,
"learning_rate": 4.9144152449976444e-06,
"loss": 0.6642,
"step": 1519
},
{
"epoch": 0.5337078651685393,
"grad_norm": 1.0274914503097534,
"learning_rate": 4.914295269869535e-06,
"loss": 0.6667,
"step": 1520
},
{
"epoch": 0.5340589887640449,
"grad_norm": 0.9218011498451233,
"learning_rate": 4.914175212174594e-06,
"loss": 0.6902,
"step": 1521
},
{
"epoch": 0.5344101123595506,
"grad_norm": 0.902390718460083,
"learning_rate": 4.914055071916925e-06,
"loss": 0.7009,
"step": 1522
},
{
"epoch": 0.5347612359550562,
"grad_norm": 0.9239642024040222,
"learning_rate": 4.9139348491006396e-06,
"loss": 0.7038,
"step": 1523
},
{
"epoch": 0.5351123595505618,
"grad_norm": 0.8971140384674072,
"learning_rate": 4.913814543729847e-06,
"loss": 0.7187,
"step": 1524
},
{
"epoch": 0.5354634831460674,
"grad_norm": 0.9525965452194214,
"learning_rate": 4.913694155808663e-06,
"loss": 0.6694,
"step": 1525
},
{
"epoch": 0.535814606741573,
"grad_norm": 0.915665864944458,
"learning_rate": 4.913573685341205e-06,
"loss": 0.7181,
"step": 1526
},
{
"epoch": 0.5361657303370787,
"grad_norm": 0.913693368434906,
"learning_rate": 4.913453132331592e-06,
"loss": 0.7122,
"step": 1527
},
{
"epoch": 0.5365168539325843,
"grad_norm": 0.91572505235672,
"learning_rate": 4.913332496783947e-06,
"loss": 0.7143,
"step": 1528
},
{
"epoch": 0.5368679775280899,
"grad_norm": 0.947823166847229,
"learning_rate": 4.913211778702395e-06,
"loss": 0.7239,
"step": 1529
},
{
"epoch": 0.5372191011235955,
"grad_norm": 0.9384491443634033,
"learning_rate": 4.913090978091066e-06,
"loss": 0.7147,
"step": 1530
},
{
"epoch": 0.5375702247191011,
"grad_norm": 0.882896363735199,
"learning_rate": 4.912970094954091e-06,
"loss": 0.6548,
"step": 1531
},
{
"epoch": 0.5379213483146067,
"grad_norm": 0.8977818489074707,
"learning_rate": 4.912849129295601e-06,
"loss": 0.6898,
"step": 1532
},
{
"epoch": 0.5382724719101124,
"grad_norm": 0.8661186099052429,
"learning_rate": 4.912728081119738e-06,
"loss": 0.6605,
"step": 1533
},
{
"epoch": 0.538623595505618,
"grad_norm": 0.8856679797172546,
"learning_rate": 4.9126069504306385e-06,
"loss": 0.6891,
"step": 1534
},
{
"epoch": 0.5389747191011236,
"grad_norm": 0.9184430241584778,
"learning_rate": 4.9124857372324455e-06,
"loss": 0.7221,
"step": 1535
},
{
"epoch": 0.5393258426966292,
"grad_norm": 0.9198493957519531,
"learning_rate": 4.912364441529306e-06,
"loss": 0.7011,
"step": 1536
},
{
"epoch": 0.5396769662921348,
"grad_norm": 0.9443540573120117,
"learning_rate": 4.912243063325365e-06,
"loss": 0.6523,
"step": 1537
},
{
"epoch": 0.5400280898876404,
"grad_norm": 0.9004515409469604,
"learning_rate": 4.912121602624777e-06,
"loss": 0.6674,
"step": 1538
},
{
"epoch": 0.5403792134831461,
"grad_norm": 0.9147352576255798,
"learning_rate": 4.912000059431693e-06,
"loss": 0.684,
"step": 1539
},
{
"epoch": 0.5407303370786517,
"grad_norm": 0.9233749508857727,
"learning_rate": 4.911878433750271e-06,
"loss": 0.702,
"step": 1540
},
{
"epoch": 0.5410814606741573,
"grad_norm": 0.8945138454437256,
"learning_rate": 4.911756725584672e-06,
"loss": 0.6694,
"step": 1541
},
{
"epoch": 0.5414325842696629,
"grad_norm": 0.9332498908042908,
"learning_rate": 4.911634934939055e-06,
"loss": 0.703,
"step": 1542
},
{
"epoch": 0.5417837078651685,
"grad_norm": 0.8984339237213135,
"learning_rate": 4.911513061817588e-06,
"loss": 0.6607,
"step": 1543
},
{
"epoch": 0.5421348314606742,
"grad_norm": 0.9096390604972839,
"learning_rate": 4.9113911062244364e-06,
"loss": 0.6786,
"step": 1544
},
{
"epoch": 0.5424859550561798,
"grad_norm": 0.8916593790054321,
"learning_rate": 4.911269068163775e-06,
"loss": 0.6644,
"step": 1545
},
{
"epoch": 0.5428370786516854,
"grad_norm": 0.9222617149353027,
"learning_rate": 4.911146947639772e-06,
"loss": 0.6983,
"step": 1546
},
{
"epoch": 0.543188202247191,
"grad_norm": 0.8953000903129578,
"learning_rate": 4.911024744656608e-06,
"loss": 0.7026,
"step": 1547
},
{
"epoch": 0.5435393258426966,
"grad_norm": 0.9042890071868896,
"learning_rate": 4.910902459218461e-06,
"loss": 0.7196,
"step": 1548
},
{
"epoch": 0.5438904494382022,
"grad_norm": 0.9504362940788269,
"learning_rate": 4.910780091329513e-06,
"loss": 0.6707,
"step": 1549
},
{
"epoch": 0.5442415730337079,
"grad_norm": 0.9307934641838074,
"learning_rate": 4.910657640993948e-06,
"loss": 0.683,
"step": 1550
},
{
"epoch": 0.5445926966292135,
"grad_norm": 0.9294170141220093,
"learning_rate": 4.910535108215955e-06,
"loss": 0.6969,
"step": 1551
},
{
"epoch": 0.5449438202247191,
"grad_norm": 0.9138461351394653,
"learning_rate": 4.910412492999725e-06,
"loss": 0.7347,
"step": 1552
},
{
"epoch": 0.5452949438202247,
"grad_norm": 0.8926948308944702,
"learning_rate": 4.910289795349449e-06,
"loss": 0.7217,
"step": 1553
},
{
"epoch": 0.5456460674157303,
"grad_norm": 0.8452489376068115,
"learning_rate": 4.910167015269325e-06,
"loss": 0.6478,
"step": 1554
},
{
"epoch": 0.545997191011236,
"grad_norm": 0.8928716778755188,
"learning_rate": 4.9100441527635516e-06,
"loss": 0.6246,
"step": 1555
},
{
"epoch": 0.5463483146067416,
"grad_norm": 0.8877882361412048,
"learning_rate": 4.90992120783633e-06,
"loss": 0.6771,
"step": 1556
},
{
"epoch": 0.5466994382022472,
"grad_norm": 0.8971658945083618,
"learning_rate": 4.909798180491865e-06,
"loss": 0.6664,
"step": 1557
},
{
"epoch": 0.5470505617977528,
"grad_norm": 0.9224132895469666,
"learning_rate": 4.909675070734365e-06,
"loss": 0.6958,
"step": 1558
},
{
"epoch": 0.5474016853932584,
"grad_norm": 0.9234408140182495,
"learning_rate": 4.9095518785680395e-06,
"loss": 0.6944,
"step": 1559
},
{
"epoch": 0.547752808988764,
"grad_norm": 0.9604536294937134,
"learning_rate": 4.909428603997101e-06,
"loss": 0.6597,
"step": 1560
},
{
"epoch": 0.5481039325842697,
"grad_norm": 0.9268457889556885,
"learning_rate": 4.909305247025767e-06,
"loss": 0.6942,
"step": 1561
},
{
"epoch": 0.5484550561797753,
"grad_norm": 0.9222556352615356,
"learning_rate": 4.909181807658256e-06,
"loss": 0.6747,
"step": 1562
},
{
"epoch": 0.5488061797752809,
"grad_norm": 0.8885440826416016,
"learning_rate": 4.909058285898787e-06,
"loss": 0.6555,
"step": 1563
},
{
"epoch": 0.5491573033707865,
"grad_norm": 0.9378191232681274,
"learning_rate": 4.908934681751586e-06,
"loss": 0.7272,
"step": 1564
},
{
"epoch": 0.5495084269662921,
"grad_norm": 0.9337999820709229,
"learning_rate": 4.908810995220882e-06,
"loss": 0.6721,
"step": 1565
},
{
"epoch": 0.5498595505617978,
"grad_norm": 0.8756289482116699,
"learning_rate": 4.908687226310902e-06,
"loss": 0.6733,
"step": 1566
},
{
"epoch": 0.5502106741573034,
"grad_norm": 0.8794893026351929,
"learning_rate": 4.90856337502588e-06,
"loss": 0.6225,
"step": 1567
},
{
"epoch": 0.550561797752809,
"grad_norm": 0.8720929026603699,
"learning_rate": 4.9084394413700515e-06,
"loss": 0.7074,
"step": 1568
},
{
"epoch": 0.5509129213483146,
"grad_norm": 0.8621073365211487,
"learning_rate": 4.908315425347655e-06,
"loss": 0.6439,
"step": 1569
},
{
"epoch": 0.5512640449438202,
"grad_norm": 0.937415361404419,
"learning_rate": 4.908191326962932e-06,
"loss": 0.7121,
"step": 1570
},
{
"epoch": 0.5516151685393258,
"grad_norm": 0.908561646938324,
"learning_rate": 4.908067146220127e-06,
"loss": 0.699,
"step": 1571
},
{
"epoch": 0.5519662921348315,
"grad_norm": 0.8949597477912903,
"learning_rate": 4.907942883123485e-06,
"loss": 0.7148,
"step": 1572
},
{
"epoch": 0.5523174157303371,
"grad_norm": 0.9320929646492004,
"learning_rate": 4.907818537677258e-06,
"loss": 0.6894,
"step": 1573
},
{
"epoch": 0.5526685393258427,
"grad_norm": 0.8810732960700989,
"learning_rate": 4.907694109885697e-06,
"loss": 0.6814,
"step": 1574
},
{
"epoch": 0.5530196629213483,
"grad_norm": 0.9485206007957458,
"learning_rate": 4.907569599753058e-06,
"loss": 0.7125,
"step": 1575
},
{
"epoch": 0.5533707865168539,
"grad_norm": 0.8823009729385376,
"learning_rate": 4.907445007283598e-06,
"loss": 0.7239,
"step": 1576
},
{
"epoch": 0.5537219101123596,
"grad_norm": 0.9091922640800476,
"learning_rate": 4.907320332481579e-06,
"loss": 0.6949,
"step": 1577
},
{
"epoch": 0.5540730337078652,
"grad_norm": 0.9078152775764465,
"learning_rate": 4.907195575351265e-06,
"loss": 0.701,
"step": 1578
},
{
"epoch": 0.5544241573033708,
"grad_norm": 0.9011525511741638,
"learning_rate": 4.907070735896923e-06,
"loss": 0.7039,
"step": 1579
},
{
"epoch": 0.5547752808988764,
"grad_norm": 0.8794327974319458,
"learning_rate": 4.90694581412282e-06,
"loss": 0.6543,
"step": 1580
},
{
"epoch": 0.555126404494382,
"grad_norm": 0.921090841293335,
"learning_rate": 4.90682081003323e-06,
"loss": 0.7031,
"step": 1581
},
{
"epoch": 0.5554775280898876,
"grad_norm": 0.8931931853294373,
"learning_rate": 4.906695723632429e-06,
"loss": 0.6938,
"step": 1582
},
{
"epoch": 0.5558286516853933,
"grad_norm": 0.9168527722358704,
"learning_rate": 4.9065705549246935e-06,
"loss": 0.7472,
"step": 1583
},
{
"epoch": 0.5561797752808989,
"grad_norm": 0.9369787573814392,
"learning_rate": 4.906445303914303e-06,
"loss": 0.6945,
"step": 1584
},
{
"epoch": 0.5565308988764045,
"grad_norm": 0.9574870467185974,
"learning_rate": 4.906319970605544e-06,
"loss": 0.7117,
"step": 1585
},
{
"epoch": 0.5568820224719101,
"grad_norm": 0.9056572318077087,
"learning_rate": 4.9061945550027e-06,
"loss": 0.7117,
"step": 1586
},
{
"epoch": 0.5572331460674157,
"grad_norm": 0.9152799844741821,
"learning_rate": 4.906069057110062e-06,
"loss": 0.7262,
"step": 1587
},
{
"epoch": 0.5575842696629213,
"grad_norm": 0.9061770439147949,
"learning_rate": 4.9059434769319205e-06,
"loss": 0.6746,
"step": 1588
},
{
"epoch": 0.557935393258427,
"grad_norm": 0.8756861090660095,
"learning_rate": 4.905817814472572e-06,
"loss": 0.6547,
"step": 1589
},
{
"epoch": 0.5582865168539326,
"grad_norm": 0.8801287412643433,
"learning_rate": 4.905692069736312e-06,
"loss": 0.7026,
"step": 1590
},
{
"epoch": 0.5586376404494382,
"grad_norm": 0.871894359588623,
"learning_rate": 4.9055662427274415e-06,
"loss": 0.6967,
"step": 1591
},
{
"epoch": 0.5589887640449438,
"grad_norm": 0.8951716423034668,
"learning_rate": 4.905440333450264e-06,
"loss": 0.6769,
"step": 1592
},
{
"epoch": 0.5593398876404494,
"grad_norm": 0.8958552479743958,
"learning_rate": 4.905314341909086e-06,
"loss": 0.6462,
"step": 1593
},
{
"epoch": 0.5596910112359551,
"grad_norm": 0.8950044512748718,
"learning_rate": 4.9051882681082155e-06,
"loss": 0.6943,
"step": 1594
},
{
"epoch": 0.5600421348314607,
"grad_norm": 0.8931154608726501,
"learning_rate": 4.905062112051965e-06,
"loss": 0.6712,
"step": 1595
},
{
"epoch": 0.5603932584269663,
"grad_norm": 0.9284601211547852,
"learning_rate": 4.904935873744647e-06,
"loss": 0.7057,
"step": 1596
},
{
"epoch": 0.5607443820224719,
"grad_norm": 0.8948965072631836,
"learning_rate": 4.904809553190582e-06,
"loss": 0.6923,
"step": 1597
},
{
"epoch": 0.5610955056179775,
"grad_norm": 0.9073783755302429,
"learning_rate": 4.904683150394086e-06,
"loss": 0.7183,
"step": 1598
},
{
"epoch": 0.5614466292134831,
"grad_norm": 0.9169145226478577,
"learning_rate": 4.904556665359485e-06,
"loss": 0.71,
"step": 1599
},
{
"epoch": 0.5617977528089888,
"grad_norm": 1.0053675174713135,
"learning_rate": 4.904430098091104e-06,
"loss": 0.7078,
"step": 1600
},
{
"epoch": 0.5621488764044944,
"grad_norm": 0.893157422542572,
"learning_rate": 4.904303448593271e-06,
"loss": 0.6814,
"step": 1601
},
{
"epoch": 0.5625,
"grad_norm": 0.8943068981170654,
"learning_rate": 4.904176716870317e-06,
"loss": 0.6915,
"step": 1602
},
{
"epoch": 0.5628511235955056,
"grad_norm": 0.8920650482177734,
"learning_rate": 4.904049902926576e-06,
"loss": 0.6953,
"step": 1603
},
{
"epoch": 0.5632022471910112,
"grad_norm": 0.8903055787086487,
"learning_rate": 4.9039230067663865e-06,
"loss": 0.6708,
"step": 1604
},
{
"epoch": 0.5635533707865169,
"grad_norm": 0.9348129034042358,
"learning_rate": 4.903796028394086e-06,
"loss": 0.6983,
"step": 1605
},
{
"epoch": 0.5639044943820225,
"grad_norm": 0.946652889251709,
"learning_rate": 4.903668967814021e-06,
"loss": 0.6997,
"step": 1606
},
{
"epoch": 0.5642556179775281,
"grad_norm": 0.9114198684692383,
"learning_rate": 4.9035418250305314e-06,
"loss": 0.6741,
"step": 1607
},
{
"epoch": 0.5646067415730337,
"grad_norm": 0.8928046226501465,
"learning_rate": 4.903414600047969e-06,
"loss": 0.6866,
"step": 1608
},
{
"epoch": 0.5649578651685393,
"grad_norm": 0.9179702997207642,
"learning_rate": 4.903287292870684e-06,
"loss": 0.696,
"step": 1609
},
{
"epoch": 0.5653089887640449,
"grad_norm": 0.9323627948760986,
"learning_rate": 4.903159903503031e-06,
"loss": 0.6939,
"step": 1610
},
{
"epoch": 0.5656601123595506,
"grad_norm": 0.8928480744361877,
"learning_rate": 4.903032431949365e-06,
"loss": 0.6819,
"step": 1611
},
{
"epoch": 0.5660112359550562,
"grad_norm": 0.8628221154212952,
"learning_rate": 4.9029048782140454e-06,
"loss": 0.687,
"step": 1612
},
{
"epoch": 0.5663623595505618,
"grad_norm": 0.8917481303215027,
"learning_rate": 4.9027772423014354e-06,
"loss": 0.7087,
"step": 1613
},
{
"epoch": 0.5667134831460674,
"grad_norm": 0.8859778642654419,
"learning_rate": 4.9026495242159e-06,
"loss": 0.6619,
"step": 1614
},
{
"epoch": 0.567064606741573,
"grad_norm": 0.9722530841827393,
"learning_rate": 4.902521723961807e-06,
"loss": 0.7133,
"step": 1615
},
{
"epoch": 0.5674157303370787,
"grad_norm": 0.8985425233840942,
"learning_rate": 4.9023938415435265e-06,
"loss": 0.6599,
"step": 1616
},
{
"epoch": 0.5677668539325843,
"grad_norm": 0.9105679392814636,
"learning_rate": 4.902265876965434e-06,
"loss": 0.6985,
"step": 1617
},
{
"epoch": 0.5681179775280899,
"grad_norm": 0.8673763871192932,
"learning_rate": 4.902137830231903e-06,
"loss": 0.692,
"step": 1618
},
{
"epoch": 0.5684691011235955,
"grad_norm": 0.8669525384902954,
"learning_rate": 4.902009701347313e-06,
"loss": 0.6515,
"step": 1619
},
{
"epoch": 0.5688202247191011,
"grad_norm": 0.9088237881660461,
"learning_rate": 4.9018814903160475e-06,
"loss": 0.6847,
"step": 1620
},
{
"epoch": 0.5691713483146067,
"grad_norm": 0.9107149243354797,
"learning_rate": 4.90175319714249e-06,
"loss": 0.6769,
"step": 1621
},
{
"epoch": 0.5695224719101124,
"grad_norm": 0.8820178508758545,
"learning_rate": 4.901624821831028e-06,
"loss": 0.7169,
"step": 1622
},
{
"epoch": 0.569873595505618,
"grad_norm": 0.8633939623832703,
"learning_rate": 4.901496364386053e-06,
"loss": 0.6395,
"step": 1623
},
{
"epoch": 0.5702247191011236,
"grad_norm": 0.9044683575630188,
"learning_rate": 4.901367824811958e-06,
"loss": 0.7011,
"step": 1624
},
{
"epoch": 0.5705758426966292,
"grad_norm": 0.9092311263084412,
"learning_rate": 4.901239203113137e-06,
"loss": 0.6643,
"step": 1625
},
{
"epoch": 0.5709269662921348,
"grad_norm": 0.9932267665863037,
"learning_rate": 4.90111049929399e-06,
"loss": 0.7203,
"step": 1626
},
{
"epoch": 0.5712780898876404,
"grad_norm": 0.8947046995162964,
"learning_rate": 4.9009817133589186e-06,
"loss": 0.683,
"step": 1627
},
{
"epoch": 0.5716292134831461,
"grad_norm": 0.9035671353340149,
"learning_rate": 4.900852845312328e-06,
"loss": 0.6761,
"step": 1628
},
{
"epoch": 0.5719803370786517,
"grad_norm": 0.905048668384552,
"learning_rate": 4.900723895158624e-06,
"loss": 0.7018,
"step": 1629
},
{
"epoch": 0.5723314606741573,
"grad_norm": 0.9121686220169067,
"learning_rate": 4.9005948629022175e-06,
"loss": 0.7506,
"step": 1630
},
{
"epoch": 0.5726825842696629,
"grad_norm": 0.9041679501533508,
"learning_rate": 4.90046574854752e-06,
"loss": 0.7058,
"step": 1631
},
{
"epoch": 0.5730337078651685,
"grad_norm": 0.8760393857955933,
"learning_rate": 4.900336552098949e-06,
"loss": 0.7067,
"step": 1632
},
{
"epoch": 0.5733848314606742,
"grad_norm": 0.9228975772857666,
"learning_rate": 4.900207273560921e-06,
"loss": 0.6959,
"step": 1633
},
{
"epoch": 0.5737359550561798,
"grad_norm": 0.8803319334983826,
"learning_rate": 4.900077912937858e-06,
"loss": 0.634,
"step": 1634
},
{
"epoch": 0.5740870786516854,
"grad_norm": 0.8787641525268555,
"learning_rate": 4.899948470234185e-06,
"loss": 0.7098,
"step": 1635
},
{
"epoch": 0.574438202247191,
"grad_norm": 0.8598030209541321,
"learning_rate": 4.899818945454328e-06,
"loss": 0.7062,
"step": 1636
},
{
"epoch": 0.5747893258426966,
"grad_norm": 0.920392632484436,
"learning_rate": 4.899689338602716e-06,
"loss": 0.6514,
"step": 1637
},
{
"epoch": 0.5751404494382022,
"grad_norm": 0.9178962111473083,
"learning_rate": 4.899559649683783e-06,
"loss": 0.6626,
"step": 1638
},
{
"epoch": 0.5754915730337079,
"grad_norm": 0.8671183586120605,
"learning_rate": 4.899429878701963e-06,
"loss": 0.704,
"step": 1639
},
{
"epoch": 0.5758426966292135,
"grad_norm": 0.8965294361114502,
"learning_rate": 4.899300025661694e-06,
"loss": 0.7058,
"step": 1640
},
{
"epoch": 0.5761938202247191,
"grad_norm": 0.8801050186157227,
"learning_rate": 4.8991700905674165e-06,
"loss": 0.695,
"step": 1641
},
{
"epoch": 0.5765449438202247,
"grad_norm": 0.9138525724411011,
"learning_rate": 4.899040073423577e-06,
"loss": 0.7061,
"step": 1642
},
{
"epoch": 0.5768960674157303,
"grad_norm": 0.8814200162887573,
"learning_rate": 4.898909974234618e-06,
"loss": 0.6659,
"step": 1643
},
{
"epoch": 0.577247191011236,
"grad_norm": 0.927841305732727,
"learning_rate": 4.898779793004991e-06,
"loss": 0.6864,
"step": 1644
},
{
"epoch": 0.5775983146067416,
"grad_norm": 0.9304884076118469,
"learning_rate": 4.898649529739148e-06,
"loss": 0.66,
"step": 1645
},
{
"epoch": 0.5779494382022472,
"grad_norm": 0.8991161584854126,
"learning_rate": 4.898519184441544e-06,
"loss": 0.6901,
"step": 1646
},
{
"epoch": 0.5783005617977528,
"grad_norm": 0.8874373435974121,
"learning_rate": 4.898388757116636e-06,
"loss": 0.6866,
"step": 1647
},
{
"epoch": 0.5786516853932584,
"grad_norm": 0.8752896785736084,
"learning_rate": 4.898258247768885e-06,
"loss": 0.6966,
"step": 1648
},
{
"epoch": 0.579002808988764,
"grad_norm": 0.9005473852157593,
"learning_rate": 4.898127656402753e-06,
"loss": 0.7192,
"step": 1649
},
{
"epoch": 0.5793539325842697,
"grad_norm": 0.8968993425369263,
"learning_rate": 4.897996983022709e-06,
"loss": 0.7156,
"step": 1650
},
{
"epoch": 0.5797050561797753,
"grad_norm": 0.881846010684967,
"learning_rate": 4.897866227633219e-06,
"loss": 0.6848,
"step": 1651
},
{
"epoch": 0.5800561797752809,
"grad_norm": 0.9290207028388977,
"learning_rate": 4.897735390238756e-06,
"loss": 0.7263,
"step": 1652
},
{
"epoch": 0.5804073033707865,
"grad_norm": 0.8702242374420166,
"learning_rate": 4.897604470843795e-06,
"loss": 0.6796,
"step": 1653
},
{
"epoch": 0.5807584269662921,
"grad_norm": 0.9286425709724426,
"learning_rate": 4.897473469452812e-06,
"loss": 0.7203,
"step": 1654
},
{
"epoch": 0.5811095505617978,
"grad_norm": 0.9703604578971863,
"learning_rate": 4.897342386070289e-06,
"loss": 0.6574,
"step": 1655
},
{
"epoch": 0.5814606741573034,
"grad_norm": 0.9450727701187134,
"learning_rate": 4.897211220700707e-06,
"loss": 0.7004,
"step": 1656
},
{
"epoch": 0.581811797752809,
"grad_norm": 0.8609772324562073,
"learning_rate": 4.897079973348553e-06,
"loss": 0.658,
"step": 1657
},
{
"epoch": 0.5821629213483146,
"grad_norm": 0.8902559876441956,
"learning_rate": 4.896948644018315e-06,
"loss": 0.6995,
"step": 1658
},
{
"epoch": 0.5825140449438202,
"grad_norm": 0.8805756568908691,
"learning_rate": 4.896817232714484e-06,
"loss": 0.6919,
"step": 1659
},
{
"epoch": 0.5828651685393258,
"grad_norm": 0.8647357225418091,
"learning_rate": 4.896685739441555e-06,
"loss": 0.7053,
"step": 1660
},
{
"epoch": 0.5832162921348315,
"grad_norm": 0.8912749290466309,
"learning_rate": 4.896554164204025e-06,
"loss": 0.6848,
"step": 1661
},
{
"epoch": 0.5835674157303371,
"grad_norm": 0.8912277221679688,
"learning_rate": 4.896422507006392e-06,
"loss": 0.687,
"step": 1662
},
{
"epoch": 0.5839185393258427,
"grad_norm": 0.8887133598327637,
"learning_rate": 4.896290767853161e-06,
"loss": 0.6703,
"step": 1663
},
{
"epoch": 0.5842696629213483,
"grad_norm": 0.9464722871780396,
"learning_rate": 4.896158946748837e-06,
"loss": 0.6834,
"step": 1664
},
{
"epoch": 0.5846207865168539,
"grad_norm": 0.8864126801490784,
"learning_rate": 4.8960270436979265e-06,
"loss": 0.6983,
"step": 1665
},
{
"epoch": 0.5849719101123596,
"grad_norm": 0.9276545643806458,
"learning_rate": 4.895895058704941e-06,
"loss": 0.6989,
"step": 1666
},
{
"epoch": 0.5853230337078652,
"grad_norm": 0.9041825532913208,
"learning_rate": 4.8957629917743955e-06,
"loss": 0.7188,
"step": 1667
},
{
"epoch": 0.5856741573033708,
"grad_norm": 0.9835352301597595,
"learning_rate": 4.895630842910805e-06,
"loss": 0.6981,
"step": 1668
},
{
"epoch": 0.5860252808988764,
"grad_norm": 0.8841162323951721,
"learning_rate": 4.89549861211869e-06,
"loss": 0.6686,
"step": 1669
},
{
"epoch": 0.586376404494382,
"grad_norm": 0.9026679992675781,
"learning_rate": 4.895366299402572e-06,
"loss": 0.6782,
"step": 1670
},
{
"epoch": 0.5867275280898876,
"grad_norm": 0.8805919885635376,
"learning_rate": 4.895233904766978e-06,
"loss": 0.7267,
"step": 1671
},
{
"epoch": 0.5870786516853933,
"grad_norm": 0.8851740956306458,
"learning_rate": 4.895101428216432e-06,
"loss": 0.6841,
"step": 1672
},
{
"epoch": 0.5874297752808989,
"grad_norm": 0.911161482334137,
"learning_rate": 4.894968869755467e-06,
"loss": 0.6938,
"step": 1673
},
{
"epoch": 0.5877808988764045,
"grad_norm": 0.900994598865509,
"learning_rate": 4.894836229388617e-06,
"loss": 0.6809,
"step": 1674
},
{
"epoch": 0.5881320224719101,
"grad_norm": 0.9121220111846924,
"learning_rate": 4.894703507120417e-06,
"loss": 0.6941,
"step": 1675
},
{
"epoch": 0.5884831460674157,
"grad_norm": 0.8991403579711914,
"learning_rate": 4.894570702955406e-06,
"loss": 0.6925,
"step": 1676
},
{
"epoch": 0.5888342696629213,
"grad_norm": 0.8850685358047485,
"learning_rate": 4.894437816898125e-06,
"loss": 0.6904,
"step": 1677
},
{
"epoch": 0.589185393258427,
"grad_norm": 0.8999640345573425,
"learning_rate": 4.894304848953122e-06,
"loss": 0.6959,
"step": 1678
},
{
"epoch": 0.5895365168539326,
"grad_norm": 0.9228804707527161,
"learning_rate": 4.89417179912494e-06,
"loss": 0.6663,
"step": 1679
},
{
"epoch": 0.5898876404494382,
"grad_norm": 0.8978350162506104,
"learning_rate": 4.894038667418132e-06,
"loss": 0.6948,
"step": 1680
},
{
"epoch": 0.5902387640449438,
"grad_norm": 0.9258869290351868,
"learning_rate": 4.89390545383725e-06,
"loss": 0.7268,
"step": 1681
},
{
"epoch": 0.5905898876404494,
"grad_norm": 0.9010106325149536,
"learning_rate": 4.893772158386851e-06,
"loss": 0.6819,
"step": 1682
},
{
"epoch": 0.5909410112359551,
"grad_norm": 0.9146981835365295,
"learning_rate": 4.8936387810714904e-06,
"loss": 0.6942,
"step": 1683
},
{
"epoch": 0.5912921348314607,
"grad_norm": 0.959784746170044,
"learning_rate": 4.893505321895733e-06,
"loss": 0.7312,
"step": 1684
},
{
"epoch": 0.5916432584269663,
"grad_norm": 0.9104639887809753,
"learning_rate": 4.8933717808641415e-06,
"loss": 0.6842,
"step": 1685
},
{
"epoch": 0.5919943820224719,
"grad_norm": 0.8836262822151184,
"learning_rate": 4.893238157981283e-06,
"loss": 0.6613,
"step": 1686
},
{
"epoch": 0.5923455056179775,
"grad_norm": 0.903002142906189,
"learning_rate": 4.893104453251728e-06,
"loss": 0.6558,
"step": 1687
},
{
"epoch": 0.5926966292134831,
"grad_norm": 0.9008224010467529,
"learning_rate": 4.892970666680048e-06,
"loss": 0.7283,
"step": 1688
},
{
"epoch": 0.5930477528089888,
"grad_norm": 0.9016857743263245,
"learning_rate": 4.892836798270819e-06,
"loss": 0.704,
"step": 1689
},
{
"epoch": 0.5933988764044944,
"grad_norm": 0.9839923977851868,
"learning_rate": 4.8927028480286195e-06,
"loss": 0.6804,
"step": 1690
},
{
"epoch": 0.59375,
"grad_norm": 0.9112361669540405,
"learning_rate": 4.8925688159580294e-06,
"loss": 0.6885,
"step": 1691
},
{
"epoch": 0.5941011235955056,
"grad_norm": 0.9427493214607239,
"learning_rate": 4.892434702063634e-06,
"loss": 0.6816,
"step": 1692
},
{
"epoch": 0.5944522471910112,
"grad_norm": 0.9189691543579102,
"learning_rate": 4.892300506350018e-06,
"loss": 0.7258,
"step": 1693
},
{
"epoch": 0.5948033707865169,
"grad_norm": 0.9497784972190857,
"learning_rate": 4.892166228821772e-06,
"loss": 0.6805,
"step": 1694
},
{
"epoch": 0.5951544943820225,
"grad_norm": 0.867682933807373,
"learning_rate": 4.892031869483489e-06,
"loss": 0.6952,
"step": 1695
},
{
"epoch": 0.5955056179775281,
"grad_norm": 0.9247297644615173,
"learning_rate": 4.891897428339763e-06,
"loss": 0.6892,
"step": 1696
},
{
"epoch": 0.5958567415730337,
"grad_norm": 0.9473358392715454,
"learning_rate": 4.89176290539519e-06,
"loss": 0.704,
"step": 1697
},
{
"epoch": 0.5962078651685393,
"grad_norm": 0.9231151938438416,
"learning_rate": 4.891628300654373e-06,
"loss": 0.725,
"step": 1698
},
{
"epoch": 0.5965589887640449,
"grad_norm": 0.9092977046966553,
"learning_rate": 4.891493614121915e-06,
"loss": 0.7209,
"step": 1699
},
{
"epoch": 0.5969101123595506,
"grad_norm": 0.9357640147209167,
"learning_rate": 4.891358845802422e-06,
"loss": 0.6577,
"step": 1700
},
{
"epoch": 0.5972612359550562,
"grad_norm": 0.9201985001564026,
"learning_rate": 4.891223995700503e-06,
"loss": 0.7037,
"step": 1701
},
{
"epoch": 0.5976123595505618,
"grad_norm": 0.8873964548110962,
"learning_rate": 4.8910890638207685e-06,
"loss": 0.6655,
"step": 1702
},
{
"epoch": 0.5979634831460674,
"grad_norm": 0.8884997367858887,
"learning_rate": 4.890954050167835e-06,
"loss": 0.6927,
"step": 1703
},
{
"epoch": 0.598314606741573,
"grad_norm": 0.9006896018981934,
"learning_rate": 4.890818954746318e-06,
"loss": 0.6879,
"step": 1704
},
{
"epoch": 0.5986657303370787,
"grad_norm": 0.8921645283699036,
"learning_rate": 4.8906837775608395e-06,
"loss": 0.7091,
"step": 1705
},
{
"epoch": 0.5990168539325843,
"grad_norm": 0.8784016370773315,
"learning_rate": 4.890548518616022e-06,
"loss": 0.6611,
"step": 1706
},
{
"epoch": 0.5993679775280899,
"grad_norm": 0.8995918035507202,
"learning_rate": 4.89041317791649e-06,
"loss": 0.699,
"step": 1707
},
{
"epoch": 0.5997191011235955,
"grad_norm": 0.9034433364868164,
"learning_rate": 4.890277755466872e-06,
"loss": 0.6861,
"step": 1708
},
{
"epoch": 0.6000702247191011,
"grad_norm": 0.896384596824646,
"learning_rate": 4.890142251271801e-06,
"loss": 0.6658,
"step": 1709
},
{
"epoch": 0.6004213483146067,
"grad_norm": 0.8753630518913269,
"learning_rate": 4.890006665335911e-06,
"loss": 0.6886,
"step": 1710
},
{
"epoch": 0.6007724719101124,
"grad_norm": 0.9117717146873474,
"learning_rate": 4.889870997663837e-06,
"loss": 0.6885,
"step": 1711
},
{
"epoch": 0.601123595505618,
"grad_norm": 0.9002839922904968,
"learning_rate": 4.889735248260221e-06,
"loss": 0.6707,
"step": 1712
},
{
"epoch": 0.6014747191011236,
"grad_norm": 0.849831223487854,
"learning_rate": 4.889599417129703e-06,
"loss": 0.6458,
"step": 1713
},
{
"epoch": 0.6018258426966292,
"grad_norm": 0.9035419225692749,
"learning_rate": 4.889463504276931e-06,
"loss": 0.6465,
"step": 1714
},
{
"epoch": 0.6021769662921348,
"grad_norm": 0.8928185701370239,
"learning_rate": 4.889327509706552e-06,
"loss": 0.6596,
"step": 1715
},
{
"epoch": 0.6025280898876404,
"grad_norm": 0.9307925701141357,
"learning_rate": 4.889191433423216e-06,
"loss": 0.6855,
"step": 1716
},
{
"epoch": 0.6028792134831461,
"grad_norm": 0.8939083814620972,
"learning_rate": 4.889055275431578e-06,
"loss": 0.7251,
"step": 1717
},
{
"epoch": 0.6032303370786517,
"grad_norm": 0.8944793343544006,
"learning_rate": 4.888919035736293e-06,
"loss": 0.6952,
"step": 1718
},
{
"epoch": 0.6035814606741573,
"grad_norm": 0.92487633228302,
"learning_rate": 4.888782714342022e-06,
"loss": 0.685,
"step": 1719
},
{
"epoch": 0.6039325842696629,
"grad_norm": 0.9212489724159241,
"learning_rate": 4.888646311253426e-06,
"loss": 0.6487,
"step": 1720
},
{
"epoch": 0.6042837078651685,
"grad_norm": 0.9022998213768005,
"learning_rate": 4.888509826475171e-06,
"loss": 0.6699,
"step": 1721
},
{
"epoch": 0.6046348314606742,
"grad_norm": 0.868482232093811,
"learning_rate": 4.8883732600119225e-06,
"loss": 0.6771,
"step": 1722
},
{
"epoch": 0.6049859550561798,
"grad_norm": 0.8934330344200134,
"learning_rate": 4.888236611868353e-06,
"loss": 0.7027,
"step": 1723
},
{
"epoch": 0.6053370786516854,
"grad_norm": 0.9520792961120605,
"learning_rate": 4.888099882049135e-06,
"loss": 0.701,
"step": 1724
},
{
"epoch": 0.605688202247191,
"grad_norm": 0.921665370464325,
"learning_rate": 4.887963070558944e-06,
"loss": 0.7058,
"step": 1725
},
{
"epoch": 0.6060393258426966,
"grad_norm": 0.8855586647987366,
"learning_rate": 4.88782617740246e-06,
"loss": 0.6789,
"step": 1726
},
{
"epoch": 0.6063904494382022,
"grad_norm": 0.9030165672302246,
"learning_rate": 4.887689202584364e-06,
"loss": 0.6883,
"step": 1727
},
{
"epoch": 0.6067415730337079,
"grad_norm": 0.8876542448997498,
"learning_rate": 4.88755214610934e-06,
"loss": 0.713,
"step": 1728
},
{
"epoch": 0.6070926966292135,
"grad_norm": 0.9762006998062134,
"learning_rate": 4.887415007982076e-06,
"loss": 0.7003,
"step": 1729
},
{
"epoch": 0.6074438202247191,
"grad_norm": 0.9035171270370483,
"learning_rate": 4.887277788207261e-06,
"loss": 0.7089,
"step": 1730
},
{
"epoch": 0.6077949438202247,
"grad_norm": 0.8891435265541077,
"learning_rate": 4.887140486789589e-06,
"loss": 0.6846,
"step": 1731
},
{
"epoch": 0.6081460674157303,
"grad_norm": 0.899974524974823,
"learning_rate": 4.887003103733755e-06,
"loss": 0.6662,
"step": 1732
},
{
"epoch": 0.608497191011236,
"grad_norm": 0.958362340927124,
"learning_rate": 4.886865639044458e-06,
"loss": 0.6732,
"step": 1733
},
{
"epoch": 0.6088483146067416,
"grad_norm": 0.9114941954612732,
"learning_rate": 4.8867280927263985e-06,
"loss": 0.6872,
"step": 1734
},
{
"epoch": 0.6091994382022472,
"grad_norm": 0.9042050242424011,
"learning_rate": 4.88659046478428e-06,
"loss": 0.6654,
"step": 1735
},
{
"epoch": 0.6095505617977528,
"grad_norm": 0.8774390816688538,
"learning_rate": 4.88645275522281e-06,
"loss": 0.666,
"step": 1736
},
{
"epoch": 0.6099016853932584,
"grad_norm": 0.8930482268333435,
"learning_rate": 4.8863149640466975e-06,
"loss": 0.6796,
"step": 1737
},
{
"epoch": 0.610252808988764,
"grad_norm": 0.9068736433982849,
"learning_rate": 4.886177091260656e-06,
"loss": 0.6599,
"step": 1738
},
{
"epoch": 0.6106039325842697,
"grad_norm": 0.8848494291305542,
"learning_rate": 4.8860391368694e-06,
"loss": 0.7131,
"step": 1739
},
{
"epoch": 0.6109550561797753,
"grad_norm": 0.9204060435295105,
"learning_rate": 4.885901100877647e-06,
"loss": 0.7236,
"step": 1740
},
{
"epoch": 0.6113061797752809,
"grad_norm": 0.912101149559021,
"learning_rate": 4.885762983290117e-06,
"loss": 0.7401,
"step": 1741
},
{
"epoch": 0.6116573033707865,
"grad_norm": 0.8925957083702087,
"learning_rate": 4.885624784111536e-06,
"loss": 0.6837,
"step": 1742
},
{
"epoch": 0.6120084269662921,
"grad_norm": 0.9438378214836121,
"learning_rate": 4.885486503346628e-06,
"loss": 0.6398,
"step": 1743
},
{
"epoch": 0.6123595505617978,
"grad_norm": 0.885036826133728,
"learning_rate": 4.8853481410001225e-06,
"loss": 0.6341,
"step": 1744
},
{
"epoch": 0.6127106741573034,
"grad_norm": 0.8819115161895752,
"learning_rate": 4.885209697076752e-06,
"loss": 0.673,
"step": 1745
},
{
"epoch": 0.613061797752809,
"grad_norm": 0.9269865155220032,
"learning_rate": 4.8850711715812515e-06,
"loss": 0.7188,
"step": 1746
},
{
"epoch": 0.6134129213483146,
"grad_norm": 0.9204302430152893,
"learning_rate": 4.8849325645183575e-06,
"loss": 0.7151,
"step": 1747
},
{
"epoch": 0.6137640449438202,
"grad_norm": 0.9573183655738831,
"learning_rate": 4.88479387589281e-06,
"loss": 0.7048,
"step": 1748
},
{
"epoch": 0.6141151685393258,
"grad_norm": 1.0055536031723022,
"learning_rate": 4.884655105709354e-06,
"loss": 0.6865,
"step": 1749
},
{
"epoch": 0.6144662921348315,
"grad_norm": 0.9006555080413818,
"learning_rate": 4.8845162539727335e-06,
"loss": 0.6931,
"step": 1750
},
{
"epoch": 0.6148174157303371,
"grad_norm": 0.8895266056060791,
"learning_rate": 4.884377320687697e-06,
"loss": 0.6788,
"step": 1751
},
{
"epoch": 0.6151685393258427,
"grad_norm": 0.8907827138900757,
"learning_rate": 4.884238305858997e-06,
"loss": 0.6559,
"step": 1752
},
{
"epoch": 0.6155196629213483,
"grad_norm": 0.8671404719352722,
"learning_rate": 4.884099209491386e-06,
"loss": 0.692,
"step": 1753
},
{
"epoch": 0.6158707865168539,
"grad_norm": 0.8965876698493958,
"learning_rate": 4.883960031589624e-06,
"loss": 0.6662,
"step": 1754
},
{
"epoch": 0.6162219101123596,
"grad_norm": 0.8783154487609863,
"learning_rate": 4.883820772158467e-06,
"loss": 0.6802,
"step": 1755
},
{
"epoch": 0.6165730337078652,
"grad_norm": 0.9085580706596375,
"learning_rate": 4.883681431202682e-06,
"loss": 0.7184,
"step": 1756
},
{
"epoch": 0.6169241573033708,
"grad_norm": 0.9406449794769287,
"learning_rate": 4.883542008727029e-06,
"loss": 0.6998,
"step": 1757
},
{
"epoch": 0.6172752808988764,
"grad_norm": 0.9234215021133423,
"learning_rate": 4.88340250473628e-06,
"loss": 0.7069,
"step": 1758
},
{
"epoch": 0.617626404494382,
"grad_norm": 0.8770806789398193,
"learning_rate": 4.883262919235205e-06,
"loss": 0.738,
"step": 1759
},
{
"epoch": 0.6179775280898876,
"grad_norm": 0.9186333417892456,
"learning_rate": 4.8831232522285786e-06,
"loss": 0.7019,
"step": 1760
},
{
"epoch": 0.6183286516853933,
"grad_norm": 0.9067515730857849,
"learning_rate": 4.8829835037211745e-06,
"loss": 0.6724,
"step": 1761
},
{
"epoch": 0.6186797752808989,
"grad_norm": 0.962090790271759,
"learning_rate": 4.8828436737177744e-06,
"loss": 0.6521,
"step": 1762
},
{
"epoch": 0.6190308988764045,
"grad_norm": 0.8877134323120117,
"learning_rate": 4.88270376222316e-06,
"loss": 0.7029,
"step": 1763
},
{
"epoch": 0.6193820224719101,
"grad_norm": 0.9146241545677185,
"learning_rate": 4.882563769242116e-06,
"loss": 0.6999,
"step": 1764
},
{
"epoch": 0.6197331460674157,
"grad_norm": 0.8948286175727844,
"learning_rate": 4.882423694779429e-06,
"loss": 0.6677,
"step": 1765
},
{
"epoch": 0.6200842696629213,
"grad_norm": 0.958892285823822,
"learning_rate": 4.88228353883989e-06,
"loss": 0.7261,
"step": 1766
},
{
"epoch": 0.620435393258427,
"grad_norm": 0.906204104423523,
"learning_rate": 4.882143301428293e-06,
"loss": 0.671,
"step": 1767
},
{
"epoch": 0.6207865168539326,
"grad_norm": 0.9508899450302124,
"learning_rate": 4.882002982549435e-06,
"loss": 0.6718,
"step": 1768
},
{
"epoch": 0.6211376404494382,
"grad_norm": 0.8806086778640747,
"learning_rate": 4.8818625822081116e-06,
"loss": 0.6835,
"step": 1769
},
{
"epoch": 0.6214887640449438,
"grad_norm": 0.9053204655647278,
"learning_rate": 4.8817221004091266e-06,
"loss": 0.7018,
"step": 1770
},
{
"epoch": 0.6218398876404494,
"grad_norm": 0.8768627047538757,
"learning_rate": 4.881581537157284e-06,
"loss": 0.6819,
"step": 1771
},
{
"epoch": 0.6221910112359551,
"grad_norm": 0.8701360821723938,
"learning_rate": 4.881440892457391e-06,
"loss": 0.6766,
"step": 1772
},
{
"epoch": 0.6225421348314607,
"grad_norm": 0.9027068614959717,
"learning_rate": 4.881300166314257e-06,
"loss": 0.6819,
"step": 1773
},
{
"epoch": 0.6228932584269663,
"grad_norm": 0.9427441954612732,
"learning_rate": 4.881159358732695e-06,
"loss": 0.6904,
"step": 1774
},
{
"epoch": 0.6232443820224719,
"grad_norm": 0.9292178750038147,
"learning_rate": 4.88101846971752e-06,
"loss": 0.7045,
"step": 1775
},
{
"epoch": 0.6235955056179775,
"grad_norm": 0.864090085029602,
"learning_rate": 4.880877499273551e-06,
"loss": 0.6496,
"step": 1776
},
{
"epoch": 0.6239466292134831,
"grad_norm": 0.9168904423713684,
"learning_rate": 4.880736447405609e-06,
"loss": 0.6805,
"step": 1777
},
{
"epoch": 0.6242977528089888,
"grad_norm": 0.9072659015655518,
"learning_rate": 4.880595314118518e-06,
"loss": 0.6699,
"step": 1778
},
{
"epoch": 0.6246488764044944,
"grad_norm": 0.9015219807624817,
"learning_rate": 4.880454099417104e-06,
"loss": 0.6781,
"step": 1779
},
{
"epoch": 0.625,
"grad_norm": 0.9149530529975891,
"learning_rate": 4.880312803306197e-06,
"loss": 0.7069,
"step": 1780
},
{
"epoch": 0.6253511235955056,
"grad_norm": 0.9412938952445984,
"learning_rate": 4.880171425790629e-06,
"loss": 0.6851,
"step": 1781
},
{
"epoch": 0.6257022471910112,
"grad_norm": 0.9359396696090698,
"learning_rate": 4.880029966875235e-06,
"loss": 0.6738,
"step": 1782
},
{
"epoch": 0.6260533707865169,
"grad_norm": 0.9232779741287231,
"learning_rate": 4.879888426564853e-06,
"loss": 0.6688,
"step": 1783
},
{
"epoch": 0.6264044943820225,
"grad_norm": 0.9085204005241394,
"learning_rate": 4.879746804864323e-06,
"loss": 0.6738,
"step": 1784
},
{
"epoch": 0.6267556179775281,
"grad_norm": 0.916195809841156,
"learning_rate": 4.879605101778488e-06,
"loss": 0.6691,
"step": 1785
},
{
"epoch": 0.6271067415730337,
"grad_norm": 0.891173243522644,
"learning_rate": 4.8794633173121954e-06,
"loss": 0.6941,
"step": 1786
},
{
"epoch": 0.6274578651685393,
"grad_norm": 0.8935718536376953,
"learning_rate": 4.879321451470294e-06,
"loss": 0.65,
"step": 1787
},
{
"epoch": 0.6278089887640449,
"grad_norm": 0.9326649308204651,
"learning_rate": 4.879179504257634e-06,
"loss": 0.7116,
"step": 1788
},
{
"epoch": 0.6281601123595506,
"grad_norm": 0.9544984102249146,
"learning_rate": 4.879037475679071e-06,
"loss": 0.6531,
"step": 1789
},
{
"epoch": 0.6285112359550562,
"grad_norm": 0.8751811981201172,
"learning_rate": 4.878895365739463e-06,
"loss": 0.7009,
"step": 1790
},
{
"epoch": 0.6288623595505618,
"grad_norm": 0.860333263874054,
"learning_rate": 4.878753174443669e-06,
"loss": 0.7003,
"step": 1791
},
{
"epoch": 0.6292134831460674,
"grad_norm": 0.9002732038497925,
"learning_rate": 4.878610901796551e-06,
"loss": 0.6768,
"step": 1792
},
{
"epoch": 0.629564606741573,
"grad_norm": 0.8901023864746094,
"learning_rate": 4.8784685478029765e-06,
"loss": 0.6572,
"step": 1793
},
{
"epoch": 0.6299157303370787,
"grad_norm": 0.8898550271987915,
"learning_rate": 4.878326112467813e-06,
"loss": 0.7141,
"step": 1794
},
{
"epoch": 0.6302668539325843,
"grad_norm": 0.9085826277732849,
"learning_rate": 4.878183595795931e-06,
"loss": 0.7033,
"step": 1795
},
{
"epoch": 0.6306179775280899,
"grad_norm": 0.9423730969429016,
"learning_rate": 4.878040997792205e-06,
"loss": 0.6763,
"step": 1796
},
{
"epoch": 0.6309691011235955,
"grad_norm": 0.9046456217765808,
"learning_rate": 4.877898318461512e-06,
"loss": 0.6438,
"step": 1797
},
{
"epoch": 0.6313202247191011,
"grad_norm": 0.9388726353645325,
"learning_rate": 4.877755557808731e-06,
"loss": 0.6921,
"step": 1798
},
{
"epoch": 0.6316713483146067,
"grad_norm": 0.8939900398254395,
"learning_rate": 4.877612715838746e-06,
"loss": 0.6729,
"step": 1799
},
{
"epoch": 0.6320224719101124,
"grad_norm": 0.919826328754425,
"learning_rate": 4.877469792556439e-06,
"loss": 0.7064,
"step": 1800
},
{
"epoch": 0.632373595505618,
"grad_norm": 0.8846567273139954,
"learning_rate": 4.8773267879667e-06,
"loss": 0.6463,
"step": 1801
},
{
"epoch": 0.6327247191011236,
"grad_norm": 0.9120215177536011,
"learning_rate": 4.8771837020744195e-06,
"loss": 0.7055,
"step": 1802
},
{
"epoch": 0.6330758426966292,
"grad_norm": 0.9156368970870972,
"learning_rate": 4.87704053488449e-06,
"loss": 0.7379,
"step": 1803
},
{
"epoch": 0.6334269662921348,
"grad_norm": 0.9091489315032959,
"learning_rate": 4.876897286401808e-06,
"loss": 0.6618,
"step": 1804
},
{
"epoch": 0.6337780898876404,
"grad_norm": 0.8682206273078918,
"learning_rate": 4.876753956631274e-06,
"loss": 0.6773,
"step": 1805
},
{
"epoch": 0.6341292134831461,
"grad_norm": 0.9142380952835083,
"learning_rate": 4.876610545577788e-06,
"loss": 0.7102,
"step": 1806
},
{
"epoch": 0.6344803370786517,
"grad_norm": 0.938529908657074,
"learning_rate": 4.876467053246255e-06,
"loss": 0.6967,
"step": 1807
},
{
"epoch": 0.6348314606741573,
"grad_norm": 0.9012477397918701,
"learning_rate": 4.876323479641582e-06,
"loss": 0.6548,
"step": 1808
},
{
"epoch": 0.6351825842696629,
"grad_norm": 0.8724120855331421,
"learning_rate": 4.8761798247686795e-06,
"loss": 0.6205,
"step": 1809
},
{
"epoch": 0.6355337078651685,
"grad_norm": 0.9118915796279907,
"learning_rate": 4.876036088632461e-06,
"loss": 0.6931,
"step": 1810
},
{
"epoch": 0.6358848314606742,
"grad_norm": 0.8679929375648499,
"learning_rate": 4.875892271237841e-06,
"loss": 0.6808,
"step": 1811
},
{
"epoch": 0.6362359550561798,
"grad_norm": 0.8951993584632874,
"learning_rate": 4.875748372589738e-06,
"loss": 0.6665,
"step": 1812
},
{
"epoch": 0.6365870786516854,
"grad_norm": 0.9559075832366943,
"learning_rate": 4.875604392693074e-06,
"loss": 0.6634,
"step": 1813
},
{
"epoch": 0.636938202247191,
"grad_norm": 0.8831387758255005,
"learning_rate": 4.875460331552772e-06,
"loss": 0.7054,
"step": 1814
},
{
"epoch": 0.6372893258426966,
"grad_norm": 0.8540201783180237,
"learning_rate": 4.875316189173761e-06,
"loss": 0.6676,
"step": 1815
},
{
"epoch": 0.6376404494382022,
"grad_norm": 0.8935166001319885,
"learning_rate": 4.875171965560967e-06,
"loss": 0.6789,
"step": 1816
},
{
"epoch": 0.6379915730337079,
"grad_norm": 0.9121273159980774,
"learning_rate": 4.875027660719326e-06,
"loss": 0.6645,
"step": 1817
},
{
"epoch": 0.6383426966292135,
"grad_norm": 0.9133923053741455,
"learning_rate": 4.87488327465377e-06,
"loss": 0.7053,
"step": 1818
},
{
"epoch": 0.6386938202247191,
"grad_norm": 0.9285811185836792,
"learning_rate": 4.874738807369239e-06,
"loss": 0.7172,
"step": 1819
},
{
"epoch": 0.6390449438202247,
"grad_norm": 0.9026995301246643,
"learning_rate": 4.874594258870673e-06,
"loss": 0.6848,
"step": 1820
},
{
"epoch": 0.6393960674157303,
"grad_norm": 0.9294858574867249,
"learning_rate": 4.874449629163015e-06,
"loss": 0.7073,
"step": 1821
},
{
"epoch": 0.639747191011236,
"grad_norm": 0.9029496908187866,
"learning_rate": 4.874304918251212e-06,
"loss": 0.7049,
"step": 1822
},
{
"epoch": 0.6400983146067416,
"grad_norm": 0.8920079469680786,
"learning_rate": 4.874160126140211e-06,
"loss": 0.6649,
"step": 1823
},
{
"epoch": 0.6404494382022472,
"grad_norm": 0.8667442798614502,
"learning_rate": 4.874015252834966e-06,
"loss": 0.6327,
"step": 1824
},
{
"epoch": 0.6408005617977528,
"grad_norm": 0.8910001516342163,
"learning_rate": 4.8738702983404315e-06,
"loss": 0.6964,
"step": 1825
},
{
"epoch": 0.6411516853932584,
"grad_norm": 0.8773640394210815,
"learning_rate": 4.873725262661564e-06,
"loss": 0.6763,
"step": 1826
},
{
"epoch": 0.641502808988764,
"grad_norm": 0.9394330978393555,
"learning_rate": 4.873580145803323e-06,
"loss": 0.7796,
"step": 1827
},
{
"epoch": 0.6418539325842697,
"grad_norm": 0.8596976399421692,
"learning_rate": 4.8734349477706725e-06,
"loss": 0.6229,
"step": 1828
},
{
"epoch": 0.6422050561797753,
"grad_norm": 0.9220896363258362,
"learning_rate": 4.873289668568578e-06,
"loss": 0.6801,
"step": 1829
},
{
"epoch": 0.6425561797752809,
"grad_norm": 0.8582680821418762,
"learning_rate": 4.873144308202008e-06,
"loss": 0.6475,
"step": 1830
},
{
"epoch": 0.6429073033707865,
"grad_norm": 0.916015088558197,
"learning_rate": 4.872998866675933e-06,
"loss": 0.6296,
"step": 1831
},
{
"epoch": 0.6432584269662921,
"grad_norm": 0.8732319474220276,
"learning_rate": 4.872853343995327e-06,
"loss": 0.6294,
"step": 1832
},
{
"epoch": 0.6436095505617978,
"grad_norm": 0.8985139727592468,
"learning_rate": 4.872707740165168e-06,
"loss": 0.6647,
"step": 1833
},
{
"epoch": 0.6439606741573034,
"grad_norm": 0.9100089073181152,
"learning_rate": 4.872562055190433e-06,
"loss": 0.66,
"step": 1834
},
{
"epoch": 0.644311797752809,
"grad_norm": 0.8807663917541504,
"learning_rate": 4.872416289076107e-06,
"loss": 0.714,
"step": 1835
},
{
"epoch": 0.6446629213483146,
"grad_norm": 0.924670934677124,
"learning_rate": 4.872270441827174e-06,
"loss": 0.6851,
"step": 1836
},
{
"epoch": 0.6450140449438202,
"grad_norm": 0.9182237982749939,
"learning_rate": 4.872124513448622e-06,
"loss": 0.6728,
"step": 1837
},
{
"epoch": 0.6453651685393258,
"grad_norm": 0.8587349057197571,
"learning_rate": 4.871978503945442e-06,
"loss": 0.6632,
"step": 1838
},
{
"epoch": 0.6457162921348315,
"grad_norm": 0.8839679956436157,
"learning_rate": 4.871832413322626e-06,
"loss": 0.6531,
"step": 1839
},
{
"epoch": 0.6460674157303371,
"grad_norm": 0.9729541540145874,
"learning_rate": 4.871686241585171e-06,
"loss": 0.6855,
"step": 1840
},
{
"epoch": 0.6464185393258427,
"grad_norm": 0.895285964012146,
"learning_rate": 4.871539988738076e-06,
"loss": 0.6738,
"step": 1841
},
{
"epoch": 0.6467696629213483,
"grad_norm": 0.8678845167160034,
"learning_rate": 4.871393654786344e-06,
"loss": 0.6494,
"step": 1842
},
{
"epoch": 0.6471207865168539,
"grad_norm": 0.8896480798721313,
"learning_rate": 4.871247239734977e-06,
"loss": 0.6823,
"step": 1843
},
{
"epoch": 0.6474719101123596,
"grad_norm": 0.8339593410491943,
"learning_rate": 4.871100743588983e-06,
"loss": 0.6193,
"step": 1844
},
{
"epoch": 0.6478230337078652,
"grad_norm": 0.8970949053764343,
"learning_rate": 4.870954166353374e-06,
"loss": 0.6944,
"step": 1845
},
{
"epoch": 0.6481741573033708,
"grad_norm": 0.8810001015663147,
"learning_rate": 4.870807508033161e-06,
"loss": 0.7222,
"step": 1846
},
{
"epoch": 0.6485252808988764,
"grad_norm": 0.8821990489959717,
"learning_rate": 4.870660768633359e-06,
"loss": 0.6881,
"step": 1847
},
{
"epoch": 0.648876404494382,
"grad_norm": 0.9234027862548828,
"learning_rate": 4.8705139481589885e-06,
"loss": 0.7074,
"step": 1848
},
{
"epoch": 0.6492275280898876,
"grad_norm": 0.9003174901008606,
"learning_rate": 4.870367046615069e-06,
"loss": 0.6754,
"step": 1849
},
{
"epoch": 0.6495786516853933,
"grad_norm": 0.9838910698890686,
"learning_rate": 4.870220064006625e-06,
"loss": 0.716,
"step": 1850
},
{
"epoch": 0.6499297752808989,
"grad_norm": 0.9417749047279358,
"learning_rate": 4.870073000338683e-06,
"loss": 0.6826,
"step": 1851
},
{
"epoch": 0.6502808988764045,
"grad_norm": 0.9156447052955627,
"learning_rate": 4.869925855616272e-06,
"loss": 0.6472,
"step": 1852
},
{
"epoch": 0.6506320224719101,
"grad_norm": 0.8980277180671692,
"learning_rate": 4.869778629844425e-06,
"loss": 0.6536,
"step": 1853
},
{
"epoch": 0.6509831460674157,
"grad_norm": 0.9398206472396851,
"learning_rate": 4.8696313230281776e-06,
"loss": 0.6618,
"step": 1854
},
{
"epoch": 0.6513342696629213,
"grad_norm": 0.9095361232757568,
"learning_rate": 4.869483935172566e-06,
"loss": 0.6758,
"step": 1855
},
{
"epoch": 0.651685393258427,
"grad_norm": 0.9065535664558411,
"learning_rate": 4.86933646628263e-06,
"loss": 0.6785,
"step": 1856
},
{
"epoch": 0.6520365168539326,
"grad_norm": 0.8876499533653259,
"learning_rate": 4.869188916363417e-06,
"loss": 0.642,
"step": 1857
},
{
"epoch": 0.6523876404494382,
"grad_norm": 0.8867465853691101,
"learning_rate": 4.869041285419968e-06,
"loss": 0.6926,
"step": 1858
},
{
"epoch": 0.6527387640449438,
"grad_norm": 0.9016279578208923,
"learning_rate": 4.868893573457335e-06,
"loss": 0.7007,
"step": 1859
},
{
"epoch": 0.6530898876404494,
"grad_norm": 0.9355380535125732,
"learning_rate": 4.868745780480569e-06,
"loss": 0.7025,
"step": 1860
},
{
"epoch": 0.6534410112359551,
"grad_norm": 0.8738915324211121,
"learning_rate": 4.868597906494725e-06,
"loss": 0.6633,
"step": 1861
},
{
"epoch": 0.6537921348314607,
"grad_norm": 0.9167248010635376,
"learning_rate": 4.868449951504859e-06,
"loss": 0.6669,
"step": 1862
},
{
"epoch": 0.6541432584269663,
"grad_norm": 0.9220420122146606,
"learning_rate": 4.868301915516031e-06,
"loss": 0.6419,
"step": 1863
},
{
"epoch": 0.6544943820224719,
"grad_norm": 0.9482148885726929,
"learning_rate": 4.868153798533304e-06,
"loss": 0.6934,
"step": 1864
},
{
"epoch": 0.6548455056179775,
"grad_norm": 0.928871214389801,
"learning_rate": 4.8680056005617436e-06,
"loss": 0.6506,
"step": 1865
},
{
"epoch": 0.6551966292134831,
"grad_norm": 0.9240955114364624,
"learning_rate": 4.867857321606418e-06,
"loss": 0.6745,
"step": 1866
},
{
"epoch": 0.6555477528089888,
"grad_norm": 0.8898888230323792,
"learning_rate": 4.867708961672399e-06,
"loss": 0.6925,
"step": 1867
},
{
"epoch": 0.6558988764044944,
"grad_norm": 0.9390647411346436,
"learning_rate": 4.8675605207647595e-06,
"loss": 0.7212,
"step": 1868
},
{
"epoch": 0.65625,
"grad_norm": 0.9547630548477173,
"learning_rate": 4.8674119988885756e-06,
"loss": 0.7168,
"step": 1869
},
{
"epoch": 0.6566011235955056,
"grad_norm": 0.8770663738250732,
"learning_rate": 4.867263396048927e-06,
"loss": 0.6373,
"step": 1870
},
{
"epoch": 0.6569522471910112,
"grad_norm": 0.899808406829834,
"learning_rate": 4.867114712250897e-06,
"loss": 0.708,
"step": 1871
},
{
"epoch": 0.6573033707865169,
"grad_norm": 0.9099377989768982,
"learning_rate": 4.866965947499568e-06,
"loss": 0.6888,
"step": 1872
},
{
"epoch": 0.6576544943820225,
"grad_norm": 0.896460235118866,
"learning_rate": 4.866817101800029e-06,
"loss": 0.6995,
"step": 1873
},
{
"epoch": 0.6580056179775281,
"grad_norm": 0.8833721280097961,
"learning_rate": 4.866668175157372e-06,
"loss": 0.659,
"step": 1874
},
{
"epoch": 0.6583567415730337,
"grad_norm": 0.830697774887085,
"learning_rate": 4.866519167576687e-06,
"loss": 0.6501,
"step": 1875
},
{
"epoch": 0.6587078651685393,
"grad_norm": 0.8897443413734436,
"learning_rate": 4.866370079063073e-06,
"loss": 0.6769,
"step": 1876
},
{
"epoch": 0.6590589887640449,
"grad_norm": 0.8772610425949097,
"learning_rate": 4.866220909621625e-06,
"loss": 0.6474,
"step": 1877
},
{
"epoch": 0.6594101123595506,
"grad_norm": 0.9042930006980896,
"learning_rate": 4.8660716592574486e-06,
"loss": 0.6827,
"step": 1878
},
{
"epoch": 0.6597612359550562,
"grad_norm": 0.9102811217308044,
"learning_rate": 4.865922327975646e-06,
"loss": 0.7126,
"step": 1879
},
{
"epoch": 0.6601123595505618,
"grad_norm": 0.9031997323036194,
"learning_rate": 4.865772915781324e-06,
"loss": 0.6963,
"step": 1880
},
{
"epoch": 0.6604634831460674,
"grad_norm": 0.89446622133255,
"learning_rate": 4.865623422679593e-06,
"loss": 0.6718,
"step": 1881
},
{
"epoch": 0.660814606741573,
"grad_norm": 0.879215657711029,
"learning_rate": 4.865473848675565e-06,
"loss": 0.6711,
"step": 1882
},
{
"epoch": 0.6611657303370787,
"grad_norm": 0.916566789150238,
"learning_rate": 4.865324193774355e-06,
"loss": 0.66,
"step": 1883
},
{
"epoch": 0.6615168539325843,
"grad_norm": 0.8872631192207336,
"learning_rate": 4.865174457981082e-06,
"loss": 0.6994,
"step": 1884
},
{
"epoch": 0.6618679775280899,
"grad_norm": 0.9287483096122742,
"learning_rate": 4.865024641300866e-06,
"loss": 0.7071,
"step": 1885
},
{
"epoch": 0.6622191011235955,
"grad_norm": 0.8984839916229248,
"learning_rate": 4.864874743738832e-06,
"loss": 0.6512,
"step": 1886
},
{
"epoch": 0.6625702247191011,
"grad_norm": 0.8624014258384705,
"learning_rate": 4.864724765300105e-06,
"loss": 0.6649,
"step": 1887
},
{
"epoch": 0.6629213483146067,
"grad_norm": 0.8853806257247925,
"learning_rate": 4.864574705989814e-06,
"loss": 0.6562,
"step": 1888
},
{
"epoch": 0.6632724719101124,
"grad_norm": 0.9202688336372375,
"learning_rate": 4.864424565813092e-06,
"loss": 0.7005,
"step": 1889
},
{
"epoch": 0.663623595505618,
"grad_norm": 0.9426562786102295,
"learning_rate": 4.8642743447750735e-06,
"loss": 0.6802,
"step": 1890
},
{
"epoch": 0.6639747191011236,
"grad_norm": 0.9027799367904663,
"learning_rate": 4.864124042880895e-06,
"loss": 0.6718,
"step": 1891
},
{
"epoch": 0.6643258426966292,
"grad_norm": 0.9138725399971008,
"learning_rate": 4.863973660135697e-06,
"loss": 0.7225,
"step": 1892
},
{
"epoch": 0.6646769662921348,
"grad_norm": 0.9003073573112488,
"learning_rate": 4.863823196544623e-06,
"loss": 0.6603,
"step": 1893
},
{
"epoch": 0.6650280898876404,
"grad_norm": 0.9345358610153198,
"learning_rate": 4.863672652112819e-06,
"loss": 0.6648,
"step": 1894
},
{
"epoch": 0.6653792134831461,
"grad_norm": 0.9087908864021301,
"learning_rate": 4.863522026845432e-06,
"loss": 0.701,
"step": 1895
},
{
"epoch": 0.6657303370786517,
"grad_norm": 0.889224112033844,
"learning_rate": 4.863371320747615e-06,
"loss": 0.7157,
"step": 1896
},
{
"epoch": 0.6660814606741573,
"grad_norm": 0.8854907751083374,
"learning_rate": 4.863220533824521e-06,
"loss": 0.6965,
"step": 1897
},
{
"epoch": 0.6664325842696629,
"grad_norm": 0.9223704934120178,
"learning_rate": 4.863069666081307e-06,
"loss": 0.6962,
"step": 1898
},
{
"epoch": 0.6667837078651685,
"grad_norm": 0.9045940637588501,
"learning_rate": 4.8629187175231314e-06,
"loss": 0.6394,
"step": 1899
},
{
"epoch": 0.6671348314606742,
"grad_norm": 0.8898254036903381,
"learning_rate": 4.862767688155159e-06,
"loss": 0.6781,
"step": 1900
},
{
"epoch": 0.6674859550561798,
"grad_norm": 0.882418692111969,
"learning_rate": 4.862616577982553e-06,
"loss": 0.6654,
"step": 1901
},
{
"epoch": 0.6678370786516854,
"grad_norm": 0.8898456692695618,
"learning_rate": 4.862465387010482e-06,
"loss": 0.6733,
"step": 1902
},
{
"epoch": 0.668188202247191,
"grad_norm": 0.9240460395812988,
"learning_rate": 4.862314115244116e-06,
"loss": 0.6661,
"step": 1903
},
{
"epoch": 0.6685393258426966,
"grad_norm": 0.8646409511566162,
"learning_rate": 4.862162762688629e-06,
"loss": 0.6751,
"step": 1904
},
{
"epoch": 0.6688904494382022,
"grad_norm": 0.9161674976348877,
"learning_rate": 4.862011329349196e-06,
"loss": 0.6762,
"step": 1905
},
{
"epoch": 0.6692415730337079,
"grad_norm": 0.8989123702049255,
"learning_rate": 4.861859815230998e-06,
"loss": 0.6948,
"step": 1906
},
{
"epoch": 0.6695926966292135,
"grad_norm": 0.9007300138473511,
"learning_rate": 4.861708220339215e-06,
"loss": 0.6954,
"step": 1907
},
{
"epoch": 0.6699438202247191,
"grad_norm": 0.9009419679641724,
"learning_rate": 4.861556544679032e-06,
"loss": 0.6968,
"step": 1908
},
{
"epoch": 0.6702949438202247,
"grad_norm": 0.8684524297714233,
"learning_rate": 4.861404788255636e-06,
"loss": 0.6548,
"step": 1909
},
{
"epoch": 0.6706460674157303,
"grad_norm": 0.8866965174674988,
"learning_rate": 4.861252951074216e-06,
"loss": 0.6867,
"step": 1910
},
{
"epoch": 0.670997191011236,
"grad_norm": 0.8514602184295654,
"learning_rate": 4.861101033139967e-06,
"loss": 0.6677,
"step": 1911
},
{
"epoch": 0.6713483146067416,
"grad_norm": 0.9115993976593018,
"learning_rate": 4.8609490344580815e-06,
"loss": 0.7163,
"step": 1912
},
{
"epoch": 0.6716994382022472,
"grad_norm": 0.9031485319137573,
"learning_rate": 4.86079695503376e-06,
"loss": 0.6557,
"step": 1913
},
{
"epoch": 0.6720505617977528,
"grad_norm": 0.9129601716995239,
"learning_rate": 4.8606447948722036e-06,
"loss": 0.6857,
"step": 1914
},
{
"epoch": 0.6724016853932584,
"grad_norm": 0.8930435180664062,
"learning_rate": 4.860492553978614e-06,
"loss": 0.6888,
"step": 1915
},
{
"epoch": 0.672752808988764,
"grad_norm": 0.8881438374519348,
"learning_rate": 4.860340232358199e-06,
"loss": 0.6979,
"step": 1916
},
{
"epoch": 0.6731039325842697,
"grad_norm": 0.8837933540344238,
"learning_rate": 4.860187830016169e-06,
"loss": 0.6729,
"step": 1917
},
{
"epoch": 0.6734550561797753,
"grad_norm": 0.8822264671325684,
"learning_rate": 4.860035346957734e-06,
"loss": 0.6883,
"step": 1918
},
{
"epoch": 0.6738061797752809,
"grad_norm": 0.8592543005943298,
"learning_rate": 4.8598827831881105e-06,
"loss": 0.6579,
"step": 1919
},
{
"epoch": 0.6741573033707865,
"grad_norm": 0.8520956039428711,
"learning_rate": 4.859730138712514e-06,
"loss": 0.6547,
"step": 1920
},
{
"epoch": 0.6745084269662921,
"grad_norm": 0.8868786096572876,
"learning_rate": 4.859577413536167e-06,
"loss": 0.6834,
"step": 1921
},
{
"epoch": 0.6748595505617978,
"grad_norm": 0.894353449344635,
"learning_rate": 4.859424607664291e-06,
"loss": 0.6622,
"step": 1922
},
{
"epoch": 0.6752106741573034,
"grad_norm": 0.8527143597602844,
"learning_rate": 4.859271721102112e-06,
"loss": 0.6557,
"step": 1923
},
{
"epoch": 0.675561797752809,
"grad_norm": 0.840531587600708,
"learning_rate": 4.85911875385486e-06,
"loss": 0.6392,
"step": 1924
},
{
"epoch": 0.6759129213483146,
"grad_norm": 0.8790932297706604,
"learning_rate": 4.858965705927765e-06,
"loss": 0.7031,
"step": 1925
},
{
"epoch": 0.6762640449438202,
"grad_norm": 0.8507084250450134,
"learning_rate": 4.85881257732606e-06,
"loss": 0.6459,
"step": 1926
},
{
"epoch": 0.6766151685393258,
"grad_norm": 0.894951581954956,
"learning_rate": 4.858659368054986e-06,
"loss": 0.6885,
"step": 1927
},
{
"epoch": 0.6769662921348315,
"grad_norm": 0.9011364579200745,
"learning_rate": 4.858506078119778e-06,
"loss": 0.7135,
"step": 1928
},
{
"epoch": 0.6773174157303371,
"grad_norm": 0.8751960396766663,
"learning_rate": 4.858352707525681e-06,
"loss": 0.6964,
"step": 1929
},
{
"epoch": 0.6776685393258427,
"grad_norm": 0.8583717942237854,
"learning_rate": 4.85819925627794e-06,
"loss": 0.6495,
"step": 1930
},
{
"epoch": 0.6780196629213483,
"grad_norm": 0.9073567390441895,
"learning_rate": 4.858045724381801e-06,
"loss": 0.6693,
"step": 1931
},
{
"epoch": 0.6783707865168539,
"grad_norm": 0.875085711479187,
"learning_rate": 4.857892111842517e-06,
"loss": 0.6552,
"step": 1932
},
{
"epoch": 0.6787219101123596,
"grad_norm": 0.890121340751648,
"learning_rate": 4.8577384186653405e-06,
"loss": 0.6816,
"step": 1933
},
{
"epoch": 0.6790730337078652,
"grad_norm": 0.9079762697219849,
"learning_rate": 4.8575846448555275e-06,
"loss": 0.7012,
"step": 1934
},
{
"epoch": 0.6794241573033708,
"grad_norm": 0.9067320823669434,
"learning_rate": 4.857430790418337e-06,
"loss": 0.6707,
"step": 1935
},
{
"epoch": 0.6797752808988764,
"grad_norm": 0.863487184047699,
"learning_rate": 4.857276855359031e-06,
"loss": 0.6532,
"step": 1936
},
{
"epoch": 0.680126404494382,
"grad_norm": 0.887187659740448,
"learning_rate": 4.857122839682873e-06,
"loss": 0.68,
"step": 1937
},
{
"epoch": 0.6804775280898876,
"grad_norm": 0.9105770587921143,
"learning_rate": 4.856968743395132e-06,
"loss": 0.7263,
"step": 1938
},
{
"epoch": 0.6808286516853933,
"grad_norm": 0.859266459941864,
"learning_rate": 4.856814566501075e-06,
"loss": 0.6612,
"step": 1939
},
{
"epoch": 0.6811797752808989,
"grad_norm": 0.8715479969978333,
"learning_rate": 4.856660309005979e-06,
"loss": 0.7078,
"step": 1940
},
{
"epoch": 0.6815308988764045,
"grad_norm": 0.8754439949989319,
"learning_rate": 4.856505970915115e-06,
"loss": 0.7088,
"step": 1941
},
{
"epoch": 0.6818820224719101,
"grad_norm": 0.8877387642860413,
"learning_rate": 4.856351552233764e-06,
"loss": 0.6944,
"step": 1942
},
{
"epoch": 0.6822331460674157,
"grad_norm": 0.9069793224334717,
"learning_rate": 4.856197052967207e-06,
"loss": 0.6567,
"step": 1943
},
{
"epoch": 0.6825842696629213,
"grad_norm": 0.9112107157707214,
"learning_rate": 4.856042473120726e-06,
"loss": 0.6617,
"step": 1944
},
{
"epoch": 0.682935393258427,
"grad_norm": 0.9331689476966858,
"learning_rate": 4.855887812699609e-06,
"loss": 0.7051,
"step": 1945
},
{
"epoch": 0.6832865168539326,
"grad_norm": 0.909070611000061,
"learning_rate": 4.855733071709145e-06,
"loss": 0.6827,
"step": 1946
},
{
"epoch": 0.6836376404494382,
"grad_norm": 0.9068970084190369,
"learning_rate": 4.855578250154624e-06,
"loss": 0.658,
"step": 1947
},
{
"epoch": 0.6839887640449438,
"grad_norm": 0.9052385687828064,
"learning_rate": 4.855423348041345e-06,
"loss": 0.6991,
"step": 1948
},
{
"epoch": 0.6843398876404494,
"grad_norm": 0.8857588768005371,
"learning_rate": 4.855268365374601e-06,
"loss": 0.6604,
"step": 1949
},
{
"epoch": 0.6846910112359551,
"grad_norm": 0.8484088778495789,
"learning_rate": 4.855113302159697e-06,
"loss": 0.6591,
"step": 1950
},
{
"epoch": 0.6850421348314607,
"grad_norm": 0.9859545826911926,
"learning_rate": 4.854958158401931e-06,
"loss": 0.6652,
"step": 1951
},
{
"epoch": 0.6853932584269663,
"grad_norm": 0.8704702854156494,
"learning_rate": 4.854802934106611e-06,
"loss": 0.6792,
"step": 1952
},
{
"epoch": 0.6857443820224719,
"grad_norm": 0.8603319525718689,
"learning_rate": 4.854647629279047e-06,
"loss": 0.6768,
"step": 1953
},
{
"epoch": 0.6860955056179775,
"grad_norm": 0.9137224555015564,
"learning_rate": 4.854492243924549e-06,
"loss": 0.6865,
"step": 1954
},
{
"epoch": 0.6864466292134831,
"grad_norm": 0.8924287557601929,
"learning_rate": 4.8543367780484304e-06,
"loss": 0.685,
"step": 1955
},
{
"epoch": 0.6867977528089888,
"grad_norm": 0.8736245632171631,
"learning_rate": 4.854181231656009e-06,
"loss": 0.6747,
"step": 1956
},
{
"epoch": 0.6871488764044944,
"grad_norm": 0.8849700093269348,
"learning_rate": 4.854025604752604e-06,
"loss": 0.679,
"step": 1957
},
{
"epoch": 0.6875,
"grad_norm": 0.9219119548797607,
"learning_rate": 4.853869897343537e-06,
"loss": 0.7069,
"step": 1958
},
{
"epoch": 0.6878511235955056,
"grad_norm": 0.8864405751228333,
"learning_rate": 4.853714109434134e-06,
"loss": 0.648,
"step": 1959
},
{
"epoch": 0.6882022471910112,
"grad_norm": 0.9157871603965759,
"learning_rate": 4.853558241029723e-06,
"loss": 0.6838,
"step": 1960
},
{
"epoch": 0.6885533707865169,
"grad_norm": 0.8845330476760864,
"learning_rate": 4.853402292135634e-06,
"loss": 0.67,
"step": 1961
},
{
"epoch": 0.6889044943820225,
"grad_norm": 0.8602401614189148,
"learning_rate": 4.8532462627572e-06,
"loss": 0.6661,
"step": 1962
},
{
"epoch": 0.6892556179775281,
"grad_norm": 0.9007517695426941,
"learning_rate": 4.853090152899758e-06,
"loss": 0.6717,
"step": 1963
},
{
"epoch": 0.6896067415730337,
"grad_norm": 0.8941755890846252,
"learning_rate": 4.852933962568647e-06,
"loss": 0.6824,
"step": 1964
},
{
"epoch": 0.6899578651685393,
"grad_norm": 0.9077978134155273,
"learning_rate": 4.852777691769207e-06,
"loss": 0.695,
"step": 1965
},
{
"epoch": 0.6903089887640449,
"grad_norm": 0.9053329229354858,
"learning_rate": 4.852621340506782e-06,
"loss": 0.6781,
"step": 1966
},
{
"epoch": 0.6906601123595506,
"grad_norm": 0.885201096534729,
"learning_rate": 4.852464908786722e-06,
"loss": 0.6307,
"step": 1967
},
{
"epoch": 0.6910112359550562,
"grad_norm": 0.9176447987556458,
"learning_rate": 4.8523083966143755e-06,
"loss": 0.6797,
"step": 1968
},
{
"epoch": 0.6913623595505618,
"grad_norm": 0.8907782435417175,
"learning_rate": 4.852151803995094e-06,
"loss": 0.661,
"step": 1969
},
{
"epoch": 0.6917134831460674,
"grad_norm": 0.9267505407333374,
"learning_rate": 4.8519951309342335e-06,
"loss": 0.6711,
"step": 1970
},
{
"epoch": 0.692064606741573,
"grad_norm": 0.9260626435279846,
"learning_rate": 4.8518383774371525e-06,
"loss": 0.6667,
"step": 1971
},
{
"epoch": 0.6924157303370787,
"grad_norm": 0.947289228439331,
"learning_rate": 4.851681543509211e-06,
"loss": 0.6802,
"step": 1972
},
{
"epoch": 0.6927668539325843,
"grad_norm": 0.9016532897949219,
"learning_rate": 4.851524629155773e-06,
"loss": 0.6788,
"step": 1973
},
{
"epoch": 0.6931179775280899,
"grad_norm": 0.926960825920105,
"learning_rate": 4.851367634382206e-06,
"loss": 0.6868,
"step": 1974
},
{
"epoch": 0.6934691011235955,
"grad_norm": 0.886313259601593,
"learning_rate": 4.851210559193877e-06,
"loss": 0.664,
"step": 1975
},
{
"epoch": 0.6938202247191011,
"grad_norm": 0.8867488503456116,
"learning_rate": 4.851053403596158e-06,
"loss": 0.6903,
"step": 1976
},
{
"epoch": 0.6941713483146067,
"grad_norm": 0.8759608864784241,
"learning_rate": 4.850896167594426e-06,
"loss": 0.6542,
"step": 1977
},
{
"epoch": 0.6945224719101124,
"grad_norm": 0.8738287091255188,
"learning_rate": 4.8507388511940565e-06,
"loss": 0.6844,
"step": 1978
},
{
"epoch": 0.694873595505618,
"grad_norm": 0.875167191028595,
"learning_rate": 4.85058145440043e-06,
"loss": 0.6515,
"step": 1979
},
{
"epoch": 0.6952247191011236,
"grad_norm": 0.8686906099319458,
"learning_rate": 4.850423977218929e-06,
"loss": 0.6558,
"step": 1980
},
{
"epoch": 0.6955758426966292,
"grad_norm": 0.8697443604469299,
"learning_rate": 4.850266419654939e-06,
"loss": 0.6431,
"step": 1981
},
{
"epoch": 0.6959269662921348,
"grad_norm": 0.9007163047790527,
"learning_rate": 4.850108781713848e-06,
"loss": 0.6996,
"step": 1982
},
{
"epoch": 0.6962780898876404,
"grad_norm": 0.9064307808876038,
"learning_rate": 4.849951063401049e-06,
"loss": 0.6625,
"step": 1983
},
{
"epoch": 0.6966292134831461,
"grad_norm": 0.9589636921882629,
"learning_rate": 4.849793264721934e-06,
"loss": 0.7276,
"step": 1984
},
{
"epoch": 0.6969803370786517,
"grad_norm": 0.9150680899620056,
"learning_rate": 4.8496353856819e-06,
"loss": 0.6808,
"step": 1985
},
{
"epoch": 0.6973314606741573,
"grad_norm": 0.9379653334617615,
"learning_rate": 4.849477426286346e-06,
"loss": 0.7148,
"step": 1986
},
{
"epoch": 0.6976825842696629,
"grad_norm": 0.8566982746124268,
"learning_rate": 4.849319386540675e-06,
"loss": 0.676,
"step": 1987
},
{
"epoch": 0.6980337078651685,
"grad_norm": 0.9016757607460022,
"learning_rate": 4.84916126645029e-06,
"loss": 0.6672,
"step": 1988
},
{
"epoch": 0.6983848314606742,
"grad_norm": 0.9286409020423889,
"learning_rate": 4.849003066020602e-06,
"loss": 0.7028,
"step": 1989
},
{
"epoch": 0.6987359550561798,
"grad_norm": 0.8700134754180908,
"learning_rate": 4.848844785257019e-06,
"loss": 0.6919,
"step": 1990
},
{
"epoch": 0.6990870786516854,
"grad_norm": 0.8518657088279724,
"learning_rate": 4.848686424164953e-06,
"loss": 0.663,
"step": 1991
},
{
"epoch": 0.699438202247191,
"grad_norm": 0.8883944749832153,
"learning_rate": 4.848527982749822e-06,
"loss": 0.6249,
"step": 1992
},
{
"epoch": 0.6997893258426966,
"grad_norm": 0.9609950184822083,
"learning_rate": 4.848369461017043e-06,
"loss": 0.7114,
"step": 1993
},
{
"epoch": 0.7001404494382022,
"grad_norm": 0.8920437097549438,
"learning_rate": 4.848210858972039e-06,
"loss": 0.7093,
"step": 1994
},
{
"epoch": 0.7004915730337079,
"grad_norm": 0.9256541132926941,
"learning_rate": 4.848052176620232e-06,
"loss": 0.6713,
"step": 1995
},
{
"epoch": 0.7008426966292135,
"grad_norm": 0.8984493613243103,
"learning_rate": 4.847893413967049e-06,
"loss": 0.6921,
"step": 1996
},
{
"epoch": 0.7011938202247191,
"grad_norm": 0.889815628528595,
"learning_rate": 4.847734571017923e-06,
"loss": 0.6539,
"step": 1997
},
{
"epoch": 0.7015449438202247,
"grad_norm": 0.856476366519928,
"learning_rate": 4.847575647778281e-06,
"loss": 0.6538,
"step": 1998
},
{
"epoch": 0.7018960674157303,
"grad_norm": 0.9042733311653137,
"learning_rate": 4.847416644253562e-06,
"loss": 0.672,
"step": 1999
},
{
"epoch": 0.702247191011236,
"grad_norm": 0.847193717956543,
"learning_rate": 4.847257560449202e-06,
"loss": 0.6296,
"step": 2000
},
{
"epoch": 0.7025983146067416,
"grad_norm": 0.922281801700592,
"learning_rate": 4.847098396370643e-06,
"loss": 0.6306,
"step": 2001
},
{
"epoch": 0.7029494382022472,
"grad_norm": 0.9229936599731445,
"learning_rate": 4.846939152023326e-06,
"loss": 0.7127,
"step": 2002
},
{
"epoch": 0.7033005617977528,
"grad_norm": 0.8789034485816956,
"learning_rate": 4.8467798274126995e-06,
"loss": 0.6477,
"step": 2003
},
{
"epoch": 0.7036516853932584,
"grad_norm": 0.87799471616745,
"learning_rate": 4.84662042254421e-06,
"loss": 0.646,
"step": 2004
},
{
"epoch": 0.704002808988764,
"grad_norm": 0.8978516459465027,
"learning_rate": 4.846460937423311e-06,
"loss": 0.6327,
"step": 2005
},
{
"epoch": 0.7043539325842697,
"grad_norm": 0.9151999354362488,
"learning_rate": 4.846301372055455e-06,
"loss": 0.6726,
"step": 2006
},
{
"epoch": 0.7047050561797753,
"grad_norm": 0.9256817698478699,
"learning_rate": 4.846141726446101e-06,
"loss": 0.6842,
"step": 2007
},
{
"epoch": 0.7050561797752809,
"grad_norm": 0.89638352394104,
"learning_rate": 4.845982000600706e-06,
"loss": 0.6792,
"step": 2008
},
{
"epoch": 0.7054073033707865,
"grad_norm": 0.8957139849662781,
"learning_rate": 4.845822194524735e-06,
"loss": 0.7032,
"step": 2009
},
{
"epoch": 0.7057584269662921,
"grad_norm": 0.9059911370277405,
"learning_rate": 4.845662308223652e-06,
"loss": 0.7272,
"step": 2010
},
{
"epoch": 0.7061095505617978,
"grad_norm": 0.8813321590423584,
"learning_rate": 4.845502341702926e-06,
"loss": 0.6918,
"step": 2011
},
{
"epoch": 0.7064606741573034,
"grad_norm": 0.9412849545478821,
"learning_rate": 4.845342294968026e-06,
"loss": 0.7301,
"step": 2012
},
{
"epoch": 0.706811797752809,
"grad_norm": 0.9104598164558411,
"learning_rate": 4.845182168024427e-06,
"loss": 0.6655,
"step": 2013
},
{
"epoch": 0.7071629213483146,
"grad_norm": 0.8984184861183167,
"learning_rate": 4.845021960877605e-06,
"loss": 0.6816,
"step": 2014
},
{
"epoch": 0.7075140449438202,
"grad_norm": 0.9031736850738525,
"learning_rate": 4.844861673533038e-06,
"loss": 0.6941,
"step": 2015
},
{
"epoch": 0.7078651685393258,
"grad_norm": 0.9205414652824402,
"learning_rate": 4.8447013059962085e-06,
"loss": 0.6976,
"step": 2016
},
{
"epoch": 0.7082162921348315,
"grad_norm": 0.9202782511711121,
"learning_rate": 4.8445408582726e-06,
"loss": 0.6859,
"step": 2017
},
{
"epoch": 0.7085674157303371,
"grad_norm": 0.9281151294708252,
"learning_rate": 4.844380330367701e-06,
"loss": 0.6827,
"step": 2018
},
{
"epoch": 0.7089185393258427,
"grad_norm": 0.8505613207817078,
"learning_rate": 4.844219722287001e-06,
"loss": 0.6198,
"step": 2019
},
{
"epoch": 0.7092696629213483,
"grad_norm": 0.8852872252464294,
"learning_rate": 4.844059034035993e-06,
"loss": 0.7026,
"step": 2020
},
{
"epoch": 0.7096207865168539,
"grad_norm": 0.8684709072113037,
"learning_rate": 4.843898265620171e-06,
"loss": 0.6602,
"step": 2021
},
{
"epoch": 0.7099719101123596,
"grad_norm": 0.9983372688293457,
"learning_rate": 4.843737417045035e-06,
"loss": 0.6646,
"step": 2022
},
{
"epoch": 0.7103230337078652,
"grad_norm": 0.8986789584159851,
"learning_rate": 4.843576488316084e-06,
"loss": 0.7246,
"step": 2023
},
{
"epoch": 0.7106741573033708,
"grad_norm": 0.898418664932251,
"learning_rate": 4.843415479438822e-06,
"loss": 0.7011,
"step": 2024
},
{
"epoch": 0.7110252808988764,
"grad_norm": 0.9410746097564697,
"learning_rate": 4.843254390418757e-06,
"loss": 0.6753,
"step": 2025
},
{
"epoch": 0.711376404494382,
"grad_norm": 0.9196873307228088,
"learning_rate": 4.843093221261396e-06,
"loss": 0.7028,
"step": 2026
},
{
"epoch": 0.7117275280898876,
"grad_norm": 0.8739240169525146,
"learning_rate": 4.842931971972252e-06,
"loss": 0.6477,
"step": 2027
},
{
"epoch": 0.7120786516853933,
"grad_norm": 0.863125205039978,
"learning_rate": 4.842770642556841e-06,
"loss": 0.6855,
"step": 2028
},
{
"epoch": 0.7124297752808989,
"grad_norm": 0.8746199607849121,
"learning_rate": 4.842609233020677e-06,
"loss": 0.6368,
"step": 2029
},
{
"epoch": 0.7127808988764045,
"grad_norm": 0.9002125859260559,
"learning_rate": 4.842447743369283e-06,
"loss": 0.666,
"step": 2030
},
{
"epoch": 0.7131320224719101,
"grad_norm": 0.8976905345916748,
"learning_rate": 4.84228617360818e-06,
"loss": 0.6807,
"step": 2031
},
{
"epoch": 0.7134831460674157,
"grad_norm": 0.9237125515937805,
"learning_rate": 4.8421245237428935e-06,
"loss": 0.6749,
"step": 2032
},
{
"epoch": 0.7138342696629213,
"grad_norm": 0.8990357518196106,
"learning_rate": 4.841962793778954e-06,
"loss": 0.7136,
"step": 2033
},
{
"epoch": 0.714185393258427,
"grad_norm": 0.9369171857833862,
"learning_rate": 4.84180098372189e-06,
"loss": 0.7157,
"step": 2034
},
{
"epoch": 0.7145365168539326,
"grad_norm": 0.9113399386405945,
"learning_rate": 4.841639093577236e-06,
"loss": 0.6733,
"step": 2035
},
{
"epoch": 0.7148876404494382,
"grad_norm": 0.9038696885108948,
"learning_rate": 4.84147712335053e-06,
"loss": 0.7065,
"step": 2036
},
{
"epoch": 0.7152387640449438,
"grad_norm": 0.8818309903144836,
"learning_rate": 4.841315073047309e-06,
"loss": 0.6633,
"step": 2037
},
{
"epoch": 0.7155898876404494,
"grad_norm": 0.8951486349105835,
"learning_rate": 4.841152942673116e-06,
"loss": 0.6965,
"step": 2038
},
{
"epoch": 0.7159410112359551,
"grad_norm": 0.8958115577697754,
"learning_rate": 4.840990732233496e-06,
"loss": 0.6536,
"step": 2039
},
{
"epoch": 0.7162921348314607,
"grad_norm": 0.8991239070892334,
"learning_rate": 4.840828441733996e-06,
"loss": 0.7002,
"step": 2040
},
{
"epoch": 0.7166432584269663,
"grad_norm": 0.8847957849502563,
"learning_rate": 4.840666071180167e-06,
"loss": 0.7011,
"step": 2041
},
{
"epoch": 0.7169943820224719,
"grad_norm": 0.8697375655174255,
"learning_rate": 4.840503620577561e-06,
"loss": 0.6869,
"step": 2042
},
{
"epoch": 0.7173455056179775,
"grad_norm": 0.8789458274841309,
"learning_rate": 4.840341089931734e-06,
"loss": 0.6624,
"step": 2043
},
{
"epoch": 0.7176966292134831,
"grad_norm": 0.868335485458374,
"learning_rate": 4.8401784792482436e-06,
"loss": 0.6624,
"step": 2044
},
{
"epoch": 0.7180477528089888,
"grad_norm": 0.8949484825134277,
"learning_rate": 4.840015788532653e-06,
"loss": 0.6621,
"step": 2045
},
{
"epoch": 0.7183988764044944,
"grad_norm": 0.8744524717330933,
"learning_rate": 4.839853017790523e-06,
"loss": 0.6055,
"step": 2046
},
{
"epoch": 0.71875,
"grad_norm": 0.8852532505989075,
"learning_rate": 4.8396901670274236e-06,
"loss": 0.7299,
"step": 2047
},
{
"epoch": 0.7191011235955056,
"grad_norm": 0.8787648677825928,
"learning_rate": 4.839527236248922e-06,
"loss": 0.6687,
"step": 2048
},
{
"epoch": 0.7194522471910112,
"grad_norm": 0.8882533311843872,
"learning_rate": 4.839364225460591e-06,
"loss": 0.6488,
"step": 2049
},
{
"epoch": 0.7198033707865169,
"grad_norm": 0.8676881194114685,
"learning_rate": 4.839201134668006e-06,
"loss": 0.7084,
"step": 2050
},
{
"epoch": 0.7201544943820225,
"grad_norm": 0.9025247097015381,
"learning_rate": 4.839037963876743e-06,
"loss": 0.6805,
"step": 2051
},
{
"epoch": 0.7205056179775281,
"grad_norm": 0.8956056833267212,
"learning_rate": 4.838874713092383e-06,
"loss": 0.6905,
"step": 2052
},
{
"epoch": 0.7208567415730337,
"grad_norm": 0.8418490290641785,
"learning_rate": 4.8387113823205104e-06,
"loss": 0.6224,
"step": 2053
},
{
"epoch": 0.7212078651685393,
"grad_norm": 0.897898256778717,
"learning_rate": 4.838547971566708e-06,
"loss": 0.6505,
"step": 2054
},
{
"epoch": 0.7215589887640449,
"grad_norm": 0.9430592060089111,
"learning_rate": 4.838384480836568e-06,
"loss": 0.7031,
"step": 2055
},
{
"epoch": 0.7219101123595506,
"grad_norm": 0.9099498987197876,
"learning_rate": 4.838220910135678e-06,
"loss": 0.7372,
"step": 2056
},
{
"epoch": 0.7222612359550562,
"grad_norm": 0.8936906456947327,
"learning_rate": 4.838057259469635e-06,
"loss": 0.6782,
"step": 2057
},
{
"epoch": 0.7226123595505618,
"grad_norm": 0.8931390643119812,
"learning_rate": 4.8378935288440346e-06,
"loss": 0.6552,
"step": 2058
},
{
"epoch": 0.7229634831460674,
"grad_norm": 0.893640398979187,
"learning_rate": 4.837729718264475e-06,
"loss": 0.6971,
"step": 2059
},
{
"epoch": 0.723314606741573,
"grad_norm": 0.9098536968231201,
"learning_rate": 4.83756582773656e-06,
"loss": 0.6816,
"step": 2060
},
{
"epoch": 0.7236657303370787,
"grad_norm": 0.8403578400611877,
"learning_rate": 4.837401857265894e-06,
"loss": 0.6233,
"step": 2061
},
{
"epoch": 0.7240168539325843,
"grad_norm": 0.894605815410614,
"learning_rate": 4.8372378068580835e-06,
"loss": 0.6747,
"step": 2062
},
{
"epoch": 0.7243679775280899,
"grad_norm": 0.8893696665763855,
"learning_rate": 4.837073676518742e-06,
"loss": 0.6769,
"step": 2063
},
{
"epoch": 0.7247191011235955,
"grad_norm": 0.9210422039031982,
"learning_rate": 4.836909466253479e-06,
"loss": 0.7177,
"step": 2064
},
{
"epoch": 0.7250702247191011,
"grad_norm": 0.9131370782852173,
"learning_rate": 4.836745176067913e-06,
"loss": 0.6899,
"step": 2065
},
{
"epoch": 0.7254213483146067,
"grad_norm": 0.8986336588859558,
"learning_rate": 4.836580805967661e-06,
"loss": 0.68,
"step": 2066
},
{
"epoch": 0.7257724719101124,
"grad_norm": 0.8927890658378601,
"learning_rate": 4.836416355958345e-06,
"loss": 0.6971,
"step": 2067
},
{
"epoch": 0.726123595505618,
"grad_norm": 0.9166001677513123,
"learning_rate": 4.836251826045589e-06,
"loss": 0.7008,
"step": 2068
},
{
"epoch": 0.7264747191011236,
"grad_norm": 0.8931081891059875,
"learning_rate": 4.836087216235021e-06,
"loss": 0.6816,
"step": 2069
},
{
"epoch": 0.7268258426966292,
"grad_norm": 0.8896619081497192,
"learning_rate": 4.835922526532267e-06,
"loss": 0.7011,
"step": 2070
},
{
"epoch": 0.7271769662921348,
"grad_norm": 0.8900657296180725,
"learning_rate": 4.835757756942963e-06,
"loss": 0.6754,
"step": 2071
},
{
"epoch": 0.7275280898876404,
"grad_norm": 0.860923707485199,
"learning_rate": 4.8355929074727415e-06,
"loss": 0.6407,
"step": 2072
},
{
"epoch": 0.7278792134831461,
"grad_norm": 0.8783923983573914,
"learning_rate": 4.835427978127241e-06,
"loss": 0.7052,
"step": 2073
},
{
"epoch": 0.7282303370786517,
"grad_norm": 0.9070163369178772,
"learning_rate": 4.835262968912103e-06,
"loss": 0.7003,
"step": 2074
},
{
"epoch": 0.7285814606741573,
"grad_norm": 1.025547742843628,
"learning_rate": 4.83509787983297e-06,
"loss": 0.6829,
"step": 2075
},
{
"epoch": 0.7289325842696629,
"grad_norm": 0.9013466835021973,
"learning_rate": 4.834932710895486e-06,
"loss": 0.6475,
"step": 2076
},
{
"epoch": 0.7292837078651685,
"grad_norm": 0.894381582736969,
"learning_rate": 4.834767462105303e-06,
"loss": 0.6414,
"step": 2077
},
{
"epoch": 0.7296348314606742,
"grad_norm": 0.8694182634353638,
"learning_rate": 4.83460213346807e-06,
"loss": 0.6442,
"step": 2078
},
{
"epoch": 0.7299859550561798,
"grad_norm": 0.8485740423202515,
"learning_rate": 4.834436724989442e-06,
"loss": 0.6105,
"step": 2079
},
{
"epoch": 0.7303370786516854,
"grad_norm": 0.8988609910011292,
"learning_rate": 4.834271236675076e-06,
"loss": 0.6615,
"step": 2080
},
{
"epoch": 0.730688202247191,
"grad_norm": 0.9217643141746521,
"learning_rate": 4.8341056685306295e-06,
"loss": 0.7194,
"step": 2081
},
{
"epoch": 0.7310393258426966,
"grad_norm": 0.8651581406593323,
"learning_rate": 4.833940020561768e-06,
"loss": 0.6831,
"step": 2082
},
{
"epoch": 0.7313904494382022,
"grad_norm": 0.9073237776756287,
"learning_rate": 4.833774292774154e-06,
"loss": 0.7034,
"step": 2083
},
{
"epoch": 0.7317415730337079,
"grad_norm": 0.9331315159797668,
"learning_rate": 4.833608485173458e-06,
"loss": 0.7328,
"step": 2084
},
{
"epoch": 0.7320926966292135,
"grad_norm": 0.9014981985092163,
"learning_rate": 4.833442597765346e-06,
"loss": 0.6776,
"step": 2085
},
{
"epoch": 0.7324438202247191,
"grad_norm": 0.8719879388809204,
"learning_rate": 4.833276630555496e-06,
"loss": 0.6672,
"step": 2086
},
{
"epoch": 0.7327949438202247,
"grad_norm": 0.8750829100608826,
"learning_rate": 4.8331105835495816e-06,
"loss": 0.6344,
"step": 2087
},
{
"epoch": 0.7331460674157303,
"grad_norm": 0.8622833490371704,
"learning_rate": 4.832944456753281e-06,
"loss": 0.6453,
"step": 2088
},
{
"epoch": 0.733497191011236,
"grad_norm": 0.8901452422142029,
"learning_rate": 4.8327782501722766e-06,
"loss": 0.6809,
"step": 2089
},
{
"epoch": 0.7338483146067416,
"grad_norm": 0.9039089679718018,
"learning_rate": 4.8326119638122525e-06,
"loss": 0.7185,
"step": 2090
},
{
"epoch": 0.7341994382022472,
"grad_norm": 0.9443939328193665,
"learning_rate": 4.832445597678895e-06,
"loss": 0.6647,
"step": 2091
},
{
"epoch": 0.7345505617977528,
"grad_norm": 0.9074206352233887,
"learning_rate": 4.8322791517778945e-06,
"loss": 0.6696,
"step": 2092
},
{
"epoch": 0.7349016853932584,
"grad_norm": 0.9349908232688904,
"learning_rate": 4.832112626114942e-06,
"loss": 0.6735,
"step": 2093
},
{
"epoch": 0.735252808988764,
"grad_norm": 0.8699564933776855,
"learning_rate": 4.831946020695733e-06,
"loss": 0.7143,
"step": 2094
},
{
"epoch": 0.7356039325842697,
"grad_norm": 0.8945705890655518,
"learning_rate": 4.831779335525967e-06,
"loss": 0.6748,
"step": 2095
},
{
"epoch": 0.7359550561797753,
"grad_norm": 0.8274276256561279,
"learning_rate": 4.831612570611343e-06,
"loss": 0.647,
"step": 2096
},
{
"epoch": 0.7363061797752809,
"grad_norm": 0.8812181949615479,
"learning_rate": 4.831445725957563e-06,
"loss": 0.6644,
"step": 2097
},
{
"epoch": 0.7366573033707865,
"grad_norm": 0.8417778015136719,
"learning_rate": 4.831278801570335e-06,
"loss": 0.613,
"step": 2098
},
{
"epoch": 0.7370084269662921,
"grad_norm": 0.862417459487915,
"learning_rate": 4.831111797455366e-06,
"loss": 0.6592,
"step": 2099
},
{
"epoch": 0.7373595505617978,
"grad_norm": 0.8994525074958801,
"learning_rate": 4.830944713618369e-06,
"loss": 0.7038,
"step": 2100
},
{
"epoch": 0.7377106741573034,
"grad_norm": 0.8670651316642761,
"learning_rate": 4.830777550065058e-06,
"loss": 0.6344,
"step": 2101
},
{
"epoch": 0.738061797752809,
"grad_norm": 0.9172879457473755,
"learning_rate": 4.830610306801148e-06,
"loss": 0.657,
"step": 2102
},
{
"epoch": 0.7384129213483146,
"grad_norm": 0.8903508186340332,
"learning_rate": 4.830442983832361e-06,
"loss": 0.6419,
"step": 2103
},
{
"epoch": 0.7387640449438202,
"grad_norm": 0.913711428642273,
"learning_rate": 4.830275581164417e-06,
"loss": 0.6551,
"step": 2104
},
{
"epoch": 0.7391151685393258,
"grad_norm": 0.8964325189590454,
"learning_rate": 4.830108098803043e-06,
"loss": 0.6839,
"step": 2105
},
{
"epoch": 0.7394662921348315,
"grad_norm": 0.8451694846153259,
"learning_rate": 4.829940536753965e-06,
"loss": 0.6846,
"step": 2106
},
{
"epoch": 0.7398174157303371,
"grad_norm": 0.8537585735321045,
"learning_rate": 4.829772895022916e-06,
"loss": 0.6632,
"step": 2107
},
{
"epoch": 0.7401685393258427,
"grad_norm": 0.882422924041748,
"learning_rate": 4.829605173615627e-06,
"loss": 0.6827,
"step": 2108
},
{
"epoch": 0.7405196629213483,
"grad_norm": 0.9608381986618042,
"learning_rate": 4.829437372537834e-06,
"loss": 0.6698,
"step": 2109
},
{
"epoch": 0.7408707865168539,
"grad_norm": 0.8602509498596191,
"learning_rate": 4.829269491795277e-06,
"loss": 0.6842,
"step": 2110
},
{
"epoch": 0.7412219101123596,
"grad_norm": 0.9101216197013855,
"learning_rate": 4.829101531393695e-06,
"loss": 0.6909,
"step": 2111
},
{
"epoch": 0.7415730337078652,
"grad_norm": 0.9232078194618225,
"learning_rate": 4.8289334913388344e-06,
"loss": 0.6544,
"step": 2112
},
{
"epoch": 0.7419241573033708,
"grad_norm": 0.9027913212776184,
"learning_rate": 4.828765371636442e-06,
"loss": 0.7055,
"step": 2113
},
{
"epoch": 0.7422752808988764,
"grad_norm": 0.8847595453262329,
"learning_rate": 4.828597172292266e-06,
"loss": 0.6585,
"step": 2114
},
{
"epoch": 0.742626404494382,
"grad_norm": 0.9057716131210327,
"learning_rate": 4.8284288933120595e-06,
"loss": 0.6815,
"step": 2115
},
{
"epoch": 0.7429775280898876,
"grad_norm": 0.833390474319458,
"learning_rate": 4.828260534701577e-06,
"loss": 0.6113,
"step": 2116
},
{
"epoch": 0.7433286516853933,
"grad_norm": 0.9086408615112305,
"learning_rate": 4.828092096466578e-06,
"loss": 0.6537,
"step": 2117
},
{
"epoch": 0.7436797752808989,
"grad_norm": 0.9064264297485352,
"learning_rate": 4.82792357861282e-06,
"loss": 0.71,
"step": 2118
},
{
"epoch": 0.7440308988764045,
"grad_norm": 0.9102289080619812,
"learning_rate": 4.827754981146068e-06,
"loss": 0.6711,
"step": 2119
},
{
"epoch": 0.7443820224719101,
"grad_norm": 0.9276719093322754,
"learning_rate": 4.827586304072087e-06,
"loss": 0.6667,
"step": 2120
},
{
"epoch": 0.7447331460674157,
"grad_norm": 0.9276329278945923,
"learning_rate": 4.827417547396647e-06,
"loss": 0.7072,
"step": 2121
},
{
"epoch": 0.7450842696629213,
"grad_norm": 0.8715528249740601,
"learning_rate": 4.827248711125518e-06,
"loss": 0.6921,
"step": 2122
},
{
"epoch": 0.745435393258427,
"grad_norm": 0.8835285902023315,
"learning_rate": 4.827079795264475e-06,
"loss": 0.6557,
"step": 2123
},
{
"epoch": 0.7457865168539326,
"grad_norm": 0.9126173853874207,
"learning_rate": 4.826910799819293e-06,
"loss": 0.6965,
"step": 2124
},
{
"epoch": 0.7461376404494382,
"grad_norm": 0.8943221569061279,
"learning_rate": 4.8267417247957545e-06,
"loss": 0.6532,
"step": 2125
},
{
"epoch": 0.7464887640449438,
"grad_norm": 0.8731197714805603,
"learning_rate": 4.826572570199639e-06,
"loss": 0.6462,
"step": 2126
},
{
"epoch": 0.7468398876404494,
"grad_norm": 0.8800104856491089,
"learning_rate": 4.826403336036733e-06,
"loss": 0.6862,
"step": 2127
},
{
"epoch": 0.7471910112359551,
"grad_norm": 0.8647268414497375,
"learning_rate": 4.826234022312823e-06,
"loss": 0.6901,
"step": 2128
},
{
"epoch": 0.7475421348314607,
"grad_norm": 0.8699238896369934,
"learning_rate": 4.826064629033701e-06,
"loss": 0.6848,
"step": 2129
},
{
"epoch": 0.7478932584269663,
"grad_norm": 0.881797730922699,
"learning_rate": 4.825895156205158e-06,
"loss": 0.7022,
"step": 2130
},
{
"epoch": 0.7482443820224719,
"grad_norm": 0.8925601840019226,
"learning_rate": 4.825725603832992e-06,
"loss": 0.6668,
"step": 2131
},
{
"epoch": 0.7485955056179775,
"grad_norm": 0.874530553817749,
"learning_rate": 4.825555971923e-06,
"loss": 0.6914,
"step": 2132
},
{
"epoch": 0.7489466292134831,
"grad_norm": 0.8640453815460205,
"learning_rate": 4.8253862604809836e-06,
"loss": 0.6301,
"step": 2133
},
{
"epoch": 0.7492977528089888,
"grad_norm": 0.9012243151664734,
"learning_rate": 4.8252164695127475e-06,
"loss": 0.6876,
"step": 2134
},
{
"epoch": 0.7496488764044944,
"grad_norm": 0.8837961554527283,
"learning_rate": 4.825046599024098e-06,
"loss": 0.6898,
"step": 2135
},
{
"epoch": 0.75,
"grad_norm": 0.8629834651947021,
"learning_rate": 4.8248766490208434e-06,
"loss": 0.6709,
"step": 2136
},
{
"epoch": 0.7503511235955056,
"grad_norm": 0.8934418559074402,
"learning_rate": 4.824706619508797e-06,
"loss": 0.6406,
"step": 2137
},
{
"epoch": 0.7507022471910112,
"grad_norm": 0.8818619251251221,
"learning_rate": 4.8245365104937734e-06,
"loss": 0.6751,
"step": 2138
},
{
"epoch": 0.7510533707865169,
"grad_norm": 0.871660590171814,
"learning_rate": 4.824366321981591e-06,
"loss": 0.6628,
"step": 2139
},
{
"epoch": 0.7514044943820225,
"grad_norm": 0.8862018585205078,
"learning_rate": 4.824196053978069e-06,
"loss": 0.655,
"step": 2140
},
{
"epoch": 0.7517556179775281,
"grad_norm": 0.8964344263076782,
"learning_rate": 4.824025706489031e-06,
"loss": 0.6784,
"step": 2141
},
{
"epoch": 0.7521067415730337,
"grad_norm": 0.8510351181030273,
"learning_rate": 4.8238552795203025e-06,
"loss": 0.6521,
"step": 2142
},
{
"epoch": 0.7524578651685393,
"grad_norm": 0.8770909905433655,
"learning_rate": 4.823684773077712e-06,
"loss": 0.6828,
"step": 2143
},
{
"epoch": 0.7528089887640449,
"grad_norm": 0.9084598422050476,
"learning_rate": 4.823514187167091e-06,
"loss": 0.7081,
"step": 2144
},
{
"epoch": 0.7531601123595506,
"grad_norm": 0.8897272944450378,
"learning_rate": 4.823343521794273e-06,
"loss": 0.6812,
"step": 2145
},
{
"epoch": 0.7535112359550562,
"grad_norm": 0.8769894242286682,
"learning_rate": 4.8231727769650934e-06,
"loss": 0.69,
"step": 2146
},
{
"epoch": 0.7538623595505618,
"grad_norm": 0.8911151885986328,
"learning_rate": 4.823001952685394e-06,
"loss": 0.6573,
"step": 2147
},
{
"epoch": 0.7542134831460674,
"grad_norm": 0.8679448366165161,
"learning_rate": 4.822831048961015e-06,
"loss": 0.7047,
"step": 2148
},
{
"epoch": 0.754564606741573,
"grad_norm": 0.8955689072608948,
"learning_rate": 4.8226600657978024e-06,
"loss": 0.6775,
"step": 2149
},
{
"epoch": 0.7549157303370787,
"grad_norm": 0.8726634383201599,
"learning_rate": 4.822489003201602e-06,
"loss": 0.6744,
"step": 2150
},
{
"epoch": 0.7552668539325843,
"grad_norm": 0.8858023881912231,
"learning_rate": 4.822317861178266e-06,
"loss": 0.6975,
"step": 2151
},
{
"epoch": 0.7556179775280899,
"grad_norm": 0.8668063879013062,
"learning_rate": 4.822146639733647e-06,
"loss": 0.6866,
"step": 2152
},
{
"epoch": 0.7559691011235955,
"grad_norm": 0.8951655626296997,
"learning_rate": 4.821975338873599e-06,
"loss": 0.7165,
"step": 2153
},
{
"epoch": 0.7563202247191011,
"grad_norm": 0.9189617037773132,
"learning_rate": 4.821803958603982e-06,
"loss": 0.6899,
"step": 2154
},
{
"epoch": 0.7566713483146067,
"grad_norm": 0.8782833814620972,
"learning_rate": 4.821632498930656e-06,
"loss": 0.6613,
"step": 2155
},
{
"epoch": 0.7570224719101124,
"grad_norm": 0.882354736328125,
"learning_rate": 4.821460959859485e-06,
"loss": 0.7047,
"step": 2156
},
{
"epoch": 0.757373595505618,
"grad_norm": 0.8961043357849121,
"learning_rate": 4.821289341396336e-06,
"loss": 0.6699,
"step": 2157
},
{
"epoch": 0.7577247191011236,
"grad_norm": 0.9026042819023132,
"learning_rate": 4.8211176435470774e-06,
"loss": 0.6874,
"step": 2158
},
{
"epoch": 0.7580758426966292,
"grad_norm": 0.89781653881073,
"learning_rate": 4.820945866317582e-06,
"loss": 0.6764,
"step": 2159
},
{
"epoch": 0.7584269662921348,
"grad_norm": 0.9029053449630737,
"learning_rate": 4.820774009713725e-06,
"loss": 0.671,
"step": 2160
},
{
"epoch": 0.7587780898876404,
"grad_norm": 0.8790138959884644,
"learning_rate": 4.820602073741381e-06,
"loss": 0.6855,
"step": 2161
},
{
"epoch": 0.7591292134831461,
"grad_norm": 0.8737415671348572,
"learning_rate": 4.820430058406433e-06,
"loss": 0.6746,
"step": 2162
},
{
"epoch": 0.7594803370786517,
"grad_norm": 0.8937375545501709,
"learning_rate": 4.8202579637147615e-06,
"loss": 0.6742,
"step": 2163
},
{
"epoch": 0.7598314606741573,
"grad_norm": 0.8693381547927856,
"learning_rate": 4.820085789672254e-06,
"loss": 0.643,
"step": 2164
},
{
"epoch": 0.7601825842696629,
"grad_norm": 0.8404312133789062,
"learning_rate": 4.819913536284798e-06,
"loss": 0.6247,
"step": 2165
},
{
"epoch": 0.7605337078651685,
"grad_norm": 0.9133322238922119,
"learning_rate": 4.819741203558284e-06,
"loss": 0.6999,
"step": 2166
},
{
"epoch": 0.7608848314606742,
"grad_norm": 0.878828763961792,
"learning_rate": 4.8195687914986055e-06,
"loss": 0.642,
"step": 2167
},
{
"epoch": 0.7612359550561798,
"grad_norm": 0.8653201460838318,
"learning_rate": 4.81939630011166e-06,
"loss": 0.6668,
"step": 2168
},
{
"epoch": 0.7615870786516854,
"grad_norm": 0.8635656833648682,
"learning_rate": 4.819223729403345e-06,
"loss": 0.6749,
"step": 2169
},
{
"epoch": 0.761938202247191,
"grad_norm": 0.9288333654403687,
"learning_rate": 4.819051079379563e-06,
"loss": 0.6687,
"step": 2170
},
{
"epoch": 0.7622893258426966,
"grad_norm": 0.8941252827644348,
"learning_rate": 4.818878350046219e-06,
"loss": 0.6795,
"step": 2171
},
{
"epoch": 0.7626404494382022,
"grad_norm": 0.8719800114631653,
"learning_rate": 4.81870554140922e-06,
"loss": 0.6273,
"step": 2172
},
{
"epoch": 0.7629915730337079,
"grad_norm": 0.9125568866729736,
"learning_rate": 4.818532653474475e-06,
"loss": 0.6992,
"step": 2173
},
{
"epoch": 0.7633426966292135,
"grad_norm": 0.9224469661712646,
"learning_rate": 4.818359686247898e-06,
"loss": 0.6984,
"step": 2174
},
{
"epoch": 0.7636938202247191,
"grad_norm": 0.9007505178451538,
"learning_rate": 4.818186639735403e-06,
"loss": 0.6994,
"step": 2175
},
{
"epoch": 0.7640449438202247,
"grad_norm": 0.9869738817214966,
"learning_rate": 4.818013513942908e-06,
"loss": 0.7248,
"step": 2176
},
{
"epoch": 0.7643960674157303,
"grad_norm": 0.9010109305381775,
"learning_rate": 4.817840308876335e-06,
"loss": 0.6987,
"step": 2177
},
{
"epoch": 0.764747191011236,
"grad_norm": 0.8850472569465637,
"learning_rate": 4.8176670245416076e-06,
"loss": 0.6561,
"step": 2178
},
{
"epoch": 0.7650983146067416,
"grad_norm": 0.9040166139602661,
"learning_rate": 4.81749366094465e-06,
"loss": 0.6789,
"step": 2179
},
{
"epoch": 0.7654494382022472,
"grad_norm": 0.879569411277771,
"learning_rate": 4.817320218091393e-06,
"loss": 0.6728,
"step": 2180
},
{
"epoch": 0.7658005617977528,
"grad_norm": 0.8991327285766602,
"learning_rate": 4.817146695987768e-06,
"loss": 0.6919,
"step": 2181
},
{
"epoch": 0.7661516853932584,
"grad_norm": 0.8922879695892334,
"learning_rate": 4.816973094639708e-06,
"loss": 0.6696,
"step": 2182
},
{
"epoch": 0.766502808988764,
"grad_norm": 0.9196280837059021,
"learning_rate": 4.8167994140531504e-06,
"loss": 0.6928,
"step": 2183
},
{
"epoch": 0.7668539325842697,
"grad_norm": 0.8855655193328857,
"learning_rate": 4.816625654234036e-06,
"loss": 0.67,
"step": 2184
},
{
"epoch": 0.7672050561797753,
"grad_norm": 0.9374614953994751,
"learning_rate": 4.816451815188306e-06,
"loss": 0.6514,
"step": 2185
},
{
"epoch": 0.7675561797752809,
"grad_norm": 0.8807061910629272,
"learning_rate": 4.816277896921907e-06,
"loss": 0.666,
"step": 2186
},
{
"epoch": 0.7679073033707865,
"grad_norm": 0.8804175853729248,
"learning_rate": 4.816103899440785e-06,
"loss": 0.6687,
"step": 2187
},
{
"epoch": 0.7682584269662921,
"grad_norm": 0.8722162842750549,
"learning_rate": 4.815929822750893e-06,
"loss": 0.6924,
"step": 2188
},
{
"epoch": 0.7686095505617978,
"grad_norm": 0.8696702718734741,
"learning_rate": 4.815755666858182e-06,
"loss": 0.6883,
"step": 2189
},
{
"epoch": 0.7689606741573034,
"grad_norm": 0.8709911704063416,
"learning_rate": 4.815581431768609e-06,
"loss": 0.6533,
"step": 2190
},
{
"epoch": 0.769311797752809,
"grad_norm": 0.9215819835662842,
"learning_rate": 4.815407117488132e-06,
"loss": 0.6519,
"step": 2191
},
{
"epoch": 0.7696629213483146,
"grad_norm": 0.9619512557983398,
"learning_rate": 4.815232724022713e-06,
"loss": 0.6523,
"step": 2192
},
{
"epoch": 0.7700140449438202,
"grad_norm": 0.8896926045417786,
"learning_rate": 4.815058251378316e-06,
"loss": 0.6772,
"step": 2193
},
{
"epoch": 0.7703651685393258,
"grad_norm": 0.9150664806365967,
"learning_rate": 4.814883699560908e-06,
"loss": 0.7005,
"step": 2194
},
{
"epoch": 0.7707162921348315,
"grad_norm": 0.8808259963989258,
"learning_rate": 4.8147090685764585e-06,
"loss": 0.683,
"step": 2195
},
{
"epoch": 0.7710674157303371,
"grad_norm": 0.8727423548698425,
"learning_rate": 4.8145343584309404e-06,
"loss": 0.6499,
"step": 2196
},
{
"epoch": 0.7714185393258427,
"grad_norm": 0.8704618811607361,
"learning_rate": 4.814359569130326e-06,
"loss": 0.6678,
"step": 2197
},
{
"epoch": 0.7717696629213483,
"grad_norm": 0.8853801488876343,
"learning_rate": 4.814184700680596e-06,
"loss": 0.6788,
"step": 2198
},
{
"epoch": 0.7721207865168539,
"grad_norm": 0.8947190046310425,
"learning_rate": 4.81400975308773e-06,
"loss": 0.6919,
"step": 2199
},
{
"epoch": 0.7724719101123596,
"grad_norm": 0.8671098351478577,
"learning_rate": 4.813834726357709e-06,
"loss": 0.6344,
"step": 2200
},
{
"epoch": 0.7728230337078652,
"grad_norm": 0.8669817447662354,
"learning_rate": 4.813659620496521e-06,
"loss": 0.6655,
"step": 2201
},
{
"epoch": 0.7731741573033708,
"grad_norm": 0.8825884461402893,
"learning_rate": 4.813484435510154e-06,
"loss": 0.687,
"step": 2202
},
{
"epoch": 0.7735252808988764,
"grad_norm": 0.9186983704566956,
"learning_rate": 4.8133091714045985e-06,
"loss": 0.6807,
"step": 2203
},
{
"epoch": 0.773876404494382,
"grad_norm": 0.8848013877868652,
"learning_rate": 4.81313382818585e-06,
"loss": 0.6974,
"step": 2204
},
{
"epoch": 0.7742275280898876,
"grad_norm": 0.8753053545951843,
"learning_rate": 4.812958405859903e-06,
"loss": 0.7025,
"step": 2205
},
{
"epoch": 0.7745786516853933,
"grad_norm": 0.8476365208625793,
"learning_rate": 4.812782904432758e-06,
"loss": 0.6466,
"step": 2206
},
{
"epoch": 0.7749297752808989,
"grad_norm": 0.8795397281646729,
"learning_rate": 4.812607323910416e-06,
"loss": 0.6669,
"step": 2207
},
{
"epoch": 0.7752808988764045,
"grad_norm": 0.8774227499961853,
"learning_rate": 4.812431664298884e-06,
"loss": 0.6488,
"step": 2208
},
{
"epoch": 0.7756320224719101,
"grad_norm": 0.888466477394104,
"learning_rate": 4.812255925604167e-06,
"loss": 0.6367,
"step": 2209
},
{
"epoch": 0.7759831460674157,
"grad_norm": 0.8345751166343689,
"learning_rate": 4.812080107832275e-06,
"loss": 0.6204,
"step": 2210
},
{
"epoch": 0.7763342696629213,
"grad_norm": 0.8765496611595154,
"learning_rate": 4.811904210989223e-06,
"loss": 0.697,
"step": 2211
},
{
"epoch": 0.776685393258427,
"grad_norm": 0.893101692199707,
"learning_rate": 4.8117282350810255e-06,
"loss": 0.6673,
"step": 2212
},
{
"epoch": 0.7770365168539326,
"grad_norm": 0.909160852432251,
"learning_rate": 4.8115521801137e-06,
"loss": 0.6973,
"step": 2213
},
{
"epoch": 0.7773876404494382,
"grad_norm": 0.8834896087646484,
"learning_rate": 4.811376046093268e-06,
"loss": 0.6878,
"step": 2214
},
{
"epoch": 0.7777387640449438,
"grad_norm": 0.8608538508415222,
"learning_rate": 4.8111998330257535e-06,
"loss": 0.6688,
"step": 2215
},
{
"epoch": 0.7780898876404494,
"grad_norm": 0.9701086282730103,
"learning_rate": 4.811023540917181e-06,
"loss": 0.6721,
"step": 2216
},
{
"epoch": 0.7784410112359551,
"grad_norm": 0.8717615008354187,
"learning_rate": 4.8108471697735826e-06,
"loss": 0.6834,
"step": 2217
},
{
"epoch": 0.7787921348314607,
"grad_norm": 0.9188772439956665,
"learning_rate": 4.810670719600987e-06,
"loss": 0.7135,
"step": 2218
},
{
"epoch": 0.7791432584269663,
"grad_norm": 0.8877284526824951,
"learning_rate": 4.810494190405432e-06,
"loss": 0.7184,
"step": 2219
},
{
"epoch": 0.7794943820224719,
"grad_norm": 0.8861910104751587,
"learning_rate": 4.8103175821929515e-06,
"loss": 0.6545,
"step": 2220
},
{
"epoch": 0.7798455056179775,
"grad_norm": 0.9048117995262146,
"learning_rate": 4.810140894969587e-06,
"loss": 0.6995,
"step": 2221
},
{
"epoch": 0.7801966292134831,
"grad_norm": 0.8994266390800476,
"learning_rate": 4.809964128741381e-06,
"loss": 0.6837,
"step": 2222
},
{
"epoch": 0.7805477528089888,
"grad_norm": 0.9110033512115479,
"learning_rate": 4.8097872835143775e-06,
"loss": 0.6575,
"step": 2223
},
{
"epoch": 0.7808988764044944,
"grad_norm": 0.8434987664222717,
"learning_rate": 4.8096103592946265e-06,
"loss": 0.6535,
"step": 2224
},
{
"epoch": 0.78125,
"grad_norm": 0.91326504945755,
"learning_rate": 4.8094333560881765e-06,
"loss": 0.6714,
"step": 2225
},
{
"epoch": 0.7816011235955056,
"grad_norm": 0.8931390643119812,
"learning_rate": 4.809256273901083e-06,
"loss": 0.667,
"step": 2226
},
{
"epoch": 0.7819522471910112,
"grad_norm": 0.8923385143280029,
"learning_rate": 4.8090791127393995e-06,
"loss": 0.6792,
"step": 2227
},
{
"epoch": 0.7823033707865169,
"grad_norm": 0.8582352995872498,
"learning_rate": 4.808901872609188e-06,
"loss": 0.674,
"step": 2228
},
{
"epoch": 0.7826544943820225,
"grad_norm": 1.004380702972412,
"learning_rate": 4.808724553516507e-06,
"loss": 0.6736,
"step": 2229
},
{
"epoch": 0.7830056179775281,
"grad_norm": 0.9153397083282471,
"learning_rate": 4.808547155467422e-06,
"loss": 0.6842,
"step": 2230
},
{
"epoch": 0.7833567415730337,
"grad_norm": 0.8752025961875916,
"learning_rate": 4.808369678468001e-06,
"loss": 0.702,
"step": 2231
},
{
"epoch": 0.7837078651685393,
"grad_norm": 0.9121509790420532,
"learning_rate": 4.808192122524311e-06,
"loss": 0.6794,
"step": 2232
},
{
"epoch": 0.7840589887640449,
"grad_norm": 0.893732488155365,
"learning_rate": 4.808014487642426e-06,
"loss": 0.6783,
"step": 2233
},
{
"epoch": 0.7844101123595506,
"grad_norm": 0.8836367726325989,
"learning_rate": 4.8078367738284205e-06,
"loss": 0.7095,
"step": 2234
},
{
"epoch": 0.7847612359550562,
"grad_norm": 0.8630282878875732,
"learning_rate": 4.8076589810883725e-06,
"loss": 0.6262,
"step": 2235
},
{
"epoch": 0.7851123595505618,
"grad_norm": 0.8813517093658447,
"learning_rate": 4.807481109428362e-06,
"loss": 0.6728,
"step": 2236
},
{
"epoch": 0.7854634831460674,
"grad_norm": 0.8723793625831604,
"learning_rate": 4.80730315885447e-06,
"loss": 0.6554,
"step": 2237
},
{
"epoch": 0.785814606741573,
"grad_norm": 0.8989778757095337,
"learning_rate": 4.807125129372787e-06,
"loss": 0.6587,
"step": 2238
},
{
"epoch": 0.7861657303370787,
"grad_norm": 0.8957408666610718,
"learning_rate": 4.8069470209893976e-06,
"loss": 0.6604,
"step": 2239
},
{
"epoch": 0.7865168539325843,
"grad_norm": 0.9145181775093079,
"learning_rate": 4.806768833710395e-06,
"loss": 0.6794,
"step": 2240
},
{
"epoch": 0.7868679775280899,
"grad_norm": 0.8940600752830505,
"learning_rate": 4.806590567541871e-06,
"loss": 0.686,
"step": 2241
},
{
"epoch": 0.7872191011235955,
"grad_norm": 0.8759947419166565,
"learning_rate": 4.806412222489925e-06,
"loss": 0.7002,
"step": 2242
},
{
"epoch": 0.7875702247191011,
"grad_norm": 0.8928653001785278,
"learning_rate": 4.8062337985606535e-06,
"loss": 0.6343,
"step": 2243
},
{
"epoch": 0.7879213483146067,
"grad_norm": 0.9009339809417725,
"learning_rate": 4.806055295760159e-06,
"loss": 0.7096,
"step": 2244
},
{
"epoch": 0.7882724719101124,
"grad_norm": 0.8576456308364868,
"learning_rate": 4.8058767140945475e-06,
"loss": 0.6449,
"step": 2245
},
{
"epoch": 0.788623595505618,
"grad_norm": 0.8837461471557617,
"learning_rate": 4.805698053569926e-06,
"loss": 0.6565,
"step": 2246
},
{
"epoch": 0.7889747191011236,
"grad_norm": 1.0404770374298096,
"learning_rate": 4.805519314192403e-06,
"loss": 0.6726,
"step": 2247
},
{
"epoch": 0.7893258426966292,
"grad_norm": 0.9084009528160095,
"learning_rate": 4.805340495968093e-06,
"loss": 0.6518,
"step": 2248
},
{
"epoch": 0.7896769662921348,
"grad_norm": 0.9170441031455994,
"learning_rate": 4.805161598903111e-06,
"loss": 0.6733,
"step": 2249
},
{
"epoch": 0.7900280898876404,
"grad_norm": 0.9394940733909607,
"learning_rate": 4.804982623003574e-06,
"loss": 0.6898,
"step": 2250
},
{
"epoch": 0.7903792134831461,
"grad_norm": 0.9983392953872681,
"learning_rate": 4.8048035682756046e-06,
"loss": 0.6739,
"step": 2251
},
{
"epoch": 0.7907303370786517,
"grad_norm": 0.9523327946662903,
"learning_rate": 4.8046244347253245e-06,
"loss": 0.6999,
"step": 2252
},
{
"epoch": 0.7910814606741573,
"grad_norm": 0.8619253635406494,
"learning_rate": 4.804445222358862e-06,
"loss": 0.6699,
"step": 2253
},
{
"epoch": 0.7914325842696629,
"grad_norm": 0.9346986413002014,
"learning_rate": 4.804265931182343e-06,
"loss": 0.6971,
"step": 2254
},
{
"epoch": 0.7917837078651685,
"grad_norm": 0.912071943283081,
"learning_rate": 4.804086561201903e-06,
"loss": 0.6913,
"step": 2255
},
{
"epoch": 0.7921348314606742,
"grad_norm": 0.9198614358901978,
"learning_rate": 4.8039071124236734e-06,
"loss": 0.6787,
"step": 2256
},
{
"epoch": 0.7924859550561798,
"grad_norm": 0.8504485487937927,
"learning_rate": 4.803727584853792e-06,
"loss": 0.6574,
"step": 2257
},
{
"epoch": 0.7928370786516854,
"grad_norm": 0.860133707523346,
"learning_rate": 4.8035479784983984e-06,
"loss": 0.7023,
"step": 2258
},
{
"epoch": 0.793188202247191,
"grad_norm": 0.8657998442649841,
"learning_rate": 4.803368293363635e-06,
"loss": 0.667,
"step": 2259
},
{
"epoch": 0.7935393258426966,
"grad_norm": 0.8903993368148804,
"learning_rate": 4.803188529455647e-06,
"loss": 0.6504,
"step": 2260
},
{
"epoch": 0.7938904494382022,
"grad_norm": 0.9583306908607483,
"learning_rate": 4.803008686780582e-06,
"loss": 0.6624,
"step": 2261
},
{
"epoch": 0.7942415730337079,
"grad_norm": 0.9079137444496155,
"learning_rate": 4.802828765344592e-06,
"loss": 0.6845,
"step": 2262
},
{
"epoch": 0.7945926966292135,
"grad_norm": 0.8480777740478516,
"learning_rate": 4.802648765153828e-06,
"loss": 0.6416,
"step": 2263
},
{
"epoch": 0.7949438202247191,
"grad_norm": 0.9052068591117859,
"learning_rate": 4.802468686214446e-06,
"loss": 0.676,
"step": 2264
},
{
"epoch": 0.7952949438202247,
"grad_norm": 0.8713710904121399,
"learning_rate": 4.802288528532606e-06,
"loss": 0.6823,
"step": 2265
},
{
"epoch": 0.7956460674157303,
"grad_norm": 0.8888510465621948,
"learning_rate": 4.802108292114468e-06,
"loss": 0.6932,
"step": 2266
},
{
"epoch": 0.795997191011236,
"grad_norm": 0.9380448460578918,
"learning_rate": 4.8019279769661965e-06,
"loss": 0.7137,
"step": 2267
},
{
"epoch": 0.7963483146067416,
"grad_norm": 0.8584917783737183,
"learning_rate": 4.801747583093958e-06,
"loss": 0.66,
"step": 2268
},
{
"epoch": 0.7966994382022472,
"grad_norm": 0.8634847402572632,
"learning_rate": 4.8015671105039205e-06,
"loss": 0.6553,
"step": 2269
},
{
"epoch": 0.7970505617977528,
"grad_norm": 1.1672545671463013,
"learning_rate": 4.801386559202259e-06,
"loss": 0.7002,
"step": 2270
},
{
"epoch": 0.7974016853932584,
"grad_norm": 0.8866782188415527,
"learning_rate": 4.801205929195146e-06,
"loss": 0.6791,
"step": 2271
},
{
"epoch": 0.797752808988764,
"grad_norm": 0.92203289270401,
"learning_rate": 4.801025220488759e-06,
"loss": 0.6586,
"step": 2272
},
{
"epoch": 0.7981039325842697,
"grad_norm": 0.9017732739448547,
"learning_rate": 4.8008444330892785e-06,
"loss": 0.6559,
"step": 2273
},
{
"epoch": 0.7984550561797753,
"grad_norm": 0.9230211973190308,
"learning_rate": 4.800663567002888e-06,
"loss": 0.6968,
"step": 2274
},
{
"epoch": 0.7988061797752809,
"grad_norm": 0.9016463160514832,
"learning_rate": 4.8004826222357715e-06,
"loss": 0.6624,
"step": 2275
},
{
"epoch": 0.7991573033707865,
"grad_norm": 0.851717472076416,
"learning_rate": 4.8003015987941185e-06,
"loss": 0.6454,
"step": 2276
},
{
"epoch": 0.7995084269662921,
"grad_norm": 0.8657079935073853,
"learning_rate": 4.800120496684119e-06,
"loss": 0.6537,
"step": 2277
},
{
"epoch": 0.7998595505617978,
"grad_norm": 0.8756804466247559,
"learning_rate": 4.7999393159119665e-06,
"loss": 0.6777,
"step": 2278
},
{
"epoch": 0.8002106741573034,
"grad_norm": 0.8858351111412048,
"learning_rate": 4.799758056483858e-06,
"loss": 0.673,
"step": 2279
},
{
"epoch": 0.800561797752809,
"grad_norm": 1.0514605045318604,
"learning_rate": 4.799576718405991e-06,
"loss": 0.6417,
"step": 2280
},
{
"epoch": 0.8009129213483146,
"grad_norm": 0.8649700284004211,
"learning_rate": 4.799395301684569e-06,
"loss": 0.6753,
"step": 2281
},
{
"epoch": 0.8012640449438202,
"grad_norm": 0.893130362033844,
"learning_rate": 4.799213806325794e-06,
"loss": 0.7144,
"step": 2282
},
{
"epoch": 0.8016151685393258,
"grad_norm": 0.8683095574378967,
"learning_rate": 4.799032232335875e-06,
"loss": 0.6565,
"step": 2283
},
{
"epoch": 0.8019662921348315,
"grad_norm": 0.8900540471076965,
"learning_rate": 4.798850579721022e-06,
"loss": 0.7065,
"step": 2284
},
{
"epoch": 0.8023174157303371,
"grad_norm": 0.9185386300086975,
"learning_rate": 4.7986688484874446e-06,
"loss": 0.6371,
"step": 2285
},
{
"epoch": 0.8026685393258427,
"grad_norm": 0.8507232666015625,
"learning_rate": 4.79848703864136e-06,
"loss": 0.66,
"step": 2286
},
{
"epoch": 0.8030196629213483,
"grad_norm": 0.8661529421806335,
"learning_rate": 4.798305150188986e-06,
"loss": 0.6643,
"step": 2287
},
{
"epoch": 0.8033707865168539,
"grad_norm": 0.8859586119651794,
"learning_rate": 4.798123183136542e-06,
"loss": 0.652,
"step": 2288
},
{
"epoch": 0.8037219101123596,
"grad_norm": 0.9179227948188782,
"learning_rate": 4.797941137490253e-06,
"loss": 0.7033,
"step": 2289
},
{
"epoch": 0.8040730337078652,
"grad_norm": 0.8715042471885681,
"learning_rate": 4.797759013256343e-06,
"loss": 0.6482,
"step": 2290
},
{
"epoch": 0.8044241573033708,
"grad_norm": 0.8953105211257935,
"learning_rate": 4.797576810441041e-06,
"loss": 0.6986,
"step": 2291
},
{
"epoch": 0.8047752808988764,
"grad_norm": 0.9580156803131104,
"learning_rate": 4.797394529050577e-06,
"loss": 0.6512,
"step": 2292
},
{
"epoch": 0.805126404494382,
"grad_norm": 0.9238266348838806,
"learning_rate": 4.7972121690911865e-06,
"loss": 0.6796,
"step": 2293
},
{
"epoch": 0.8054775280898876,
"grad_norm": 0.8915815949440002,
"learning_rate": 4.797029730569106e-06,
"loss": 0.707,
"step": 2294
},
{
"epoch": 0.8058286516853933,
"grad_norm": 0.8978826999664307,
"learning_rate": 4.796847213490574e-06,
"loss": 0.7162,
"step": 2295
},
{
"epoch": 0.8061797752808989,
"grad_norm": 0.8888537287712097,
"learning_rate": 4.796664617861832e-06,
"loss": 0.6832,
"step": 2296
},
{
"epoch": 0.8065308988764045,
"grad_norm": 0.8834527730941772,
"learning_rate": 4.796481943689127e-06,
"loss": 0.6863,
"step": 2297
},
{
"epoch": 0.8068820224719101,
"grad_norm": 0.8643321394920349,
"learning_rate": 4.796299190978704e-06,
"loss": 0.6739,
"step": 2298
},
{
"epoch": 0.8072331460674157,
"grad_norm": 0.8717115521430969,
"learning_rate": 4.796116359736813e-06,
"loss": 0.6502,
"step": 2299
},
{
"epoch": 0.8075842696629213,
"grad_norm": 0.8730166554450989,
"learning_rate": 4.7959334499697085e-06,
"loss": 0.6604,
"step": 2300
},
{
"epoch": 0.807935393258427,
"grad_norm": 0.8617185950279236,
"learning_rate": 4.795750461683643e-06,
"loss": 0.667,
"step": 2301
},
{
"epoch": 0.8082865168539326,
"grad_norm": 0.826866626739502,
"learning_rate": 4.795567394884878e-06,
"loss": 0.6273,
"step": 2302
},
{
"epoch": 0.8086376404494382,
"grad_norm": 0.9160782098770142,
"learning_rate": 4.795384249579672e-06,
"loss": 0.6841,
"step": 2303
},
{
"epoch": 0.8089887640449438,
"grad_norm": 0.8521285653114319,
"learning_rate": 4.795201025774287e-06,
"loss": 0.6311,
"step": 2304
},
{
"epoch": 0.8093398876404494,
"grad_norm": 0.9138549566268921,
"learning_rate": 4.795017723474993e-06,
"loss": 0.6848,
"step": 2305
},
{
"epoch": 0.8096910112359551,
"grad_norm": 0.8857465386390686,
"learning_rate": 4.794834342688056e-06,
"loss": 0.6457,
"step": 2306
},
{
"epoch": 0.8100421348314607,
"grad_norm": 0.934290885925293,
"learning_rate": 4.7946508834197495e-06,
"loss": 0.6853,
"step": 2307
},
{
"epoch": 0.8103932584269663,
"grad_norm": 0.8928509950637817,
"learning_rate": 4.794467345676345e-06,
"loss": 0.6884,
"step": 2308
},
{
"epoch": 0.8107443820224719,
"grad_norm": 0.9139081835746765,
"learning_rate": 4.79428372946412e-06,
"loss": 0.7047,
"step": 2309
},
{
"epoch": 0.8110955056179775,
"grad_norm": 0.8674109578132629,
"learning_rate": 4.794100034789356e-06,
"loss": 0.6397,
"step": 2310
},
{
"epoch": 0.8114466292134831,
"grad_norm": 0.9009836912155151,
"learning_rate": 4.793916261658333e-06,
"loss": 0.6954,
"step": 2311
},
{
"epoch": 0.8117977528089888,
"grad_norm": 0.8968256115913391,
"learning_rate": 4.7937324100773375e-06,
"loss": 0.6983,
"step": 2312
},
{
"epoch": 0.8121488764044944,
"grad_norm": 0.9040331244468689,
"learning_rate": 4.793548480052656e-06,
"loss": 0.6932,
"step": 2313
},
{
"epoch": 0.8125,
"grad_norm": 0.9160118699073792,
"learning_rate": 4.7933644715905794e-06,
"loss": 0.702,
"step": 2314
},
{
"epoch": 0.8128511235955056,
"grad_norm": 0.8957852721214294,
"learning_rate": 4.7931803846974005e-06,
"loss": 0.6737,
"step": 2315
},
{
"epoch": 0.8132022471910112,
"grad_norm": 0.9111645817756653,
"learning_rate": 4.7929962193794145e-06,
"loss": 0.6716,
"step": 2316
},
{
"epoch": 0.8135533707865169,
"grad_norm": 0.9393349885940552,
"learning_rate": 4.792811975642919e-06,
"loss": 0.7166,
"step": 2317
},
{
"epoch": 0.8139044943820225,
"grad_norm": 0.8632532358169556,
"learning_rate": 4.792627653494217e-06,
"loss": 0.6545,
"step": 2318
},
{
"epoch": 0.8142556179775281,
"grad_norm": 0.8554111123085022,
"learning_rate": 4.792443252939612e-06,
"loss": 0.6421,
"step": 2319
},
{
"epoch": 0.8146067415730337,
"grad_norm": 0.8907461166381836,
"learning_rate": 4.792258773985408e-06,
"loss": 0.6874,
"step": 2320
},
{
"epoch": 0.8149578651685393,
"grad_norm": 0.8739638924598694,
"learning_rate": 4.792074216637916e-06,
"loss": 0.6662,
"step": 2321
},
{
"epoch": 0.8153089887640449,
"grad_norm": 0.8694685101509094,
"learning_rate": 4.791889580903446e-06,
"loss": 0.6536,
"step": 2322
},
{
"epoch": 0.8156601123595506,
"grad_norm": 0.9057086706161499,
"learning_rate": 4.791704866788314e-06,
"loss": 0.6655,
"step": 2323
},
{
"epoch": 0.8160112359550562,
"grad_norm": 0.9522676467895508,
"learning_rate": 4.791520074298837e-06,
"loss": 0.7235,
"step": 2324
},
{
"epoch": 0.8163623595505618,
"grad_norm": 0.8711917996406555,
"learning_rate": 4.791335203441335e-06,
"loss": 0.7015,
"step": 2325
},
{
"epoch": 0.8167134831460674,
"grad_norm": 0.8643499612808228,
"learning_rate": 4.791150254222129e-06,
"loss": 0.6229,
"step": 2326
},
{
"epoch": 0.817064606741573,
"grad_norm": 0.8863927721977234,
"learning_rate": 4.790965226647545e-06,
"loss": 0.7017,
"step": 2327
},
{
"epoch": 0.8174157303370787,
"grad_norm": 0.8969694375991821,
"learning_rate": 4.790780120723911e-06,
"loss": 0.6926,
"step": 2328
},
{
"epoch": 0.8177668539325843,
"grad_norm": 0.923423171043396,
"learning_rate": 4.790594936457556e-06,
"loss": 0.707,
"step": 2329
},
{
"epoch": 0.8181179775280899,
"grad_norm": 0.8579330444335938,
"learning_rate": 4.790409673854815e-06,
"loss": 0.6224,
"step": 2330
},
{
"epoch": 0.8184691011235955,
"grad_norm": 0.8940487504005432,
"learning_rate": 4.790224332922023e-06,
"loss": 0.6859,
"step": 2331
},
{
"epoch": 0.8188202247191011,
"grad_norm": 0.8670837879180908,
"learning_rate": 4.790038913665519e-06,
"loss": 0.6669,
"step": 2332
},
{
"epoch": 0.8191713483146067,
"grad_norm": 0.8588036298751831,
"learning_rate": 4.789853416091643e-06,
"loss": 0.6951,
"step": 2333
},
{
"epoch": 0.8195224719101124,
"grad_norm": 0.8730223774909973,
"learning_rate": 4.78966784020674e-06,
"loss": 0.6751,
"step": 2334
},
{
"epoch": 0.819873595505618,
"grad_norm": 0.8985008001327515,
"learning_rate": 4.789482186017157e-06,
"loss": 0.6927,
"step": 2335
},
{
"epoch": 0.8202247191011236,
"grad_norm": 0.8898709416389465,
"learning_rate": 4.78929645352924e-06,
"loss": 0.6307,
"step": 2336
},
{
"epoch": 0.8205758426966292,
"grad_norm": 0.8730369806289673,
"learning_rate": 4.789110642749346e-06,
"loss": 0.6548,
"step": 2337
},
{
"epoch": 0.8209269662921348,
"grad_norm": 0.8780831694602966,
"learning_rate": 4.788924753683826e-06,
"loss": 0.6451,
"step": 2338
},
{
"epoch": 0.8212780898876404,
"grad_norm": 0.8665900230407715,
"learning_rate": 4.788738786339037e-06,
"loss": 0.6656,
"step": 2339
},
{
"epoch": 0.8216292134831461,
"grad_norm": 0.9199172854423523,
"learning_rate": 4.788552740721341e-06,
"loss": 0.6629,
"step": 2340
},
{
"epoch": 0.8219803370786517,
"grad_norm": 0.8883637189865112,
"learning_rate": 4.7883666168370996e-06,
"loss": 0.6789,
"step": 2341
},
{
"epoch": 0.8223314606741573,
"grad_norm": 0.8635849356651306,
"learning_rate": 4.788180414692678e-06,
"loss": 0.6823,
"step": 2342
},
{
"epoch": 0.8226825842696629,
"grad_norm": 0.9055063128471375,
"learning_rate": 4.787994134294444e-06,
"loss": 0.6667,
"step": 2343
},
{
"epoch": 0.8230337078651685,
"grad_norm": 0.8952557444572449,
"learning_rate": 4.787807775648768e-06,
"loss": 0.6694,
"step": 2344
},
{
"epoch": 0.8233848314606742,
"grad_norm": 0.9171862006187439,
"learning_rate": 4.787621338762025e-06,
"loss": 0.6938,
"step": 2345
},
{
"epoch": 0.8237359550561798,
"grad_norm": 0.9036487936973572,
"learning_rate": 4.787434823640589e-06,
"loss": 0.6565,
"step": 2346
},
{
"epoch": 0.8240870786516854,
"grad_norm": 0.8997727036476135,
"learning_rate": 4.787248230290839e-06,
"loss": 0.6586,
"step": 2347
},
{
"epoch": 0.824438202247191,
"grad_norm": 0.9203174114227295,
"learning_rate": 4.787061558719158e-06,
"loss": 0.6684,
"step": 2348
},
{
"epoch": 0.8247893258426966,
"grad_norm": 0.8566135168075562,
"learning_rate": 4.786874808931928e-06,
"loss": 0.6366,
"step": 2349
},
{
"epoch": 0.8251404494382022,
"grad_norm": 0.8947523832321167,
"learning_rate": 4.786687980935536e-06,
"loss": 0.6954,
"step": 2350
},
{
"epoch": 0.8254915730337079,
"grad_norm": 0.8714900016784668,
"learning_rate": 4.7865010747363725e-06,
"loss": 0.6678,
"step": 2351
},
{
"epoch": 0.8258426966292135,
"grad_norm": 0.8742727041244507,
"learning_rate": 4.786314090340829e-06,
"loss": 0.6356,
"step": 2352
},
{
"epoch": 0.8261938202247191,
"grad_norm": 0.8638887405395508,
"learning_rate": 4.7861270277553e-06,
"loss": 0.6643,
"step": 2353
},
{
"epoch": 0.8265449438202247,
"grad_norm": 0.8582507967948914,
"learning_rate": 4.785939886986182e-06,
"loss": 0.6943,
"step": 2354
},
{
"epoch": 0.8268960674157303,
"grad_norm": 0.8985755443572998,
"learning_rate": 4.785752668039877e-06,
"loss": 0.6899,
"step": 2355
},
{
"epoch": 0.827247191011236,
"grad_norm": 0.889373242855072,
"learning_rate": 4.7855653709227856e-06,
"loss": 0.6687,
"step": 2356
},
{
"epoch": 0.8275983146067416,
"grad_norm": 0.8748716711997986,
"learning_rate": 4.785377995641315e-06,
"loss": 0.6813,
"step": 2357
},
{
"epoch": 0.8279494382022472,
"grad_norm": 0.912598729133606,
"learning_rate": 4.785190542201873e-06,
"loss": 0.7037,
"step": 2358
},
{
"epoch": 0.8283005617977528,
"grad_norm": 0.8579078912734985,
"learning_rate": 4.785003010610869e-06,
"loss": 0.651,
"step": 2359
},
{
"epoch": 0.8286516853932584,
"grad_norm": 0.8841166496276855,
"learning_rate": 4.7848154008747185e-06,
"loss": 0.6527,
"step": 2360
},
{
"epoch": 0.829002808988764,
"grad_norm": 0.9118697047233582,
"learning_rate": 4.784627712999836e-06,
"loss": 0.7196,
"step": 2361
},
{
"epoch": 0.8293539325842697,
"grad_norm": 0.8655892610549927,
"learning_rate": 4.78443994699264e-06,
"loss": 0.6813,
"step": 2362
},
{
"epoch": 0.8297050561797753,
"grad_norm": 0.8933113217353821,
"learning_rate": 4.784252102859553e-06,
"loss": 0.6276,
"step": 2363
},
{
"epoch": 0.8300561797752809,
"grad_norm": 0.8977401852607727,
"learning_rate": 4.784064180606999e-06,
"loss": 0.7107,
"step": 2364
},
{
"epoch": 0.8304073033707865,
"grad_norm": 0.8428375720977783,
"learning_rate": 4.783876180241404e-06,
"loss": 0.6187,
"step": 2365
},
{
"epoch": 0.8307584269662921,
"grad_norm": 0.8943214416503906,
"learning_rate": 4.783688101769199e-06,
"loss": 0.71,
"step": 2366
},
{
"epoch": 0.8311095505617978,
"grad_norm": 0.8977285027503967,
"learning_rate": 4.783499945196814e-06,
"loss": 0.6895,
"step": 2367
},
{
"epoch": 0.8314606741573034,
"grad_norm": 0.8746622800827026,
"learning_rate": 4.783311710530685e-06,
"loss": 0.656,
"step": 2368
},
{
"epoch": 0.831811797752809,
"grad_norm": 0.9107025265693665,
"learning_rate": 4.783123397777249e-06,
"loss": 0.6869,
"step": 2369
},
{
"epoch": 0.8321629213483146,
"grad_norm": 0.8555697202682495,
"learning_rate": 4.782935006942948e-06,
"loss": 0.6607,
"step": 2370
},
{
"epoch": 0.8325140449438202,
"grad_norm": 0.9654397368431091,
"learning_rate": 4.782746538034222e-06,
"loss": 0.7033,
"step": 2371
},
{
"epoch": 0.8328651685393258,
"grad_norm": 0.8786224722862244,
"learning_rate": 4.782557991057519e-06,
"loss": 0.6855,
"step": 2372
},
{
"epoch": 0.8332162921348315,
"grad_norm": 0.8756742477416992,
"learning_rate": 4.782369366019285e-06,
"loss": 0.6665,
"step": 2373
},
{
"epoch": 0.8335674157303371,
"grad_norm": 0.8499640226364136,
"learning_rate": 4.782180662925971e-06,
"loss": 0.6311,
"step": 2374
},
{
"epoch": 0.8339185393258427,
"grad_norm": 0.8776542544364929,
"learning_rate": 4.781991881784033e-06,
"loss": 0.6894,
"step": 2375
},
{
"epoch": 0.8342696629213483,
"grad_norm": 0.8762704730033875,
"learning_rate": 4.7818030225999245e-06,
"loss": 0.6637,
"step": 2376
},
{
"epoch": 0.8346207865168539,
"grad_norm": 0.8762661218643188,
"learning_rate": 4.7816140853801054e-06,
"loss": 0.6672,
"step": 2377
},
{
"epoch": 0.8349719101123596,
"grad_norm": 0.8843420743942261,
"learning_rate": 4.781425070131036e-06,
"loss": 0.6614,
"step": 2378
},
{
"epoch": 0.8353230337078652,
"grad_norm": 0.8788619041442871,
"learning_rate": 4.781235976859183e-06,
"loss": 0.6853,
"step": 2379
},
{
"epoch": 0.8356741573033708,
"grad_norm": 0.9134128093719482,
"learning_rate": 4.78104680557101e-06,
"loss": 0.6904,
"step": 2380
},
{
"epoch": 0.8360252808988764,
"grad_norm": 0.8329752683639526,
"learning_rate": 4.7808575562729894e-06,
"loss": 0.6318,
"step": 2381
},
{
"epoch": 0.836376404494382,
"grad_norm": 0.8506373763084412,
"learning_rate": 4.780668228971592e-06,
"loss": 0.6757,
"step": 2382
},
{
"epoch": 0.8367275280898876,
"grad_norm": 0.8934277892112732,
"learning_rate": 4.780478823673292e-06,
"loss": 0.6961,
"step": 2383
},
{
"epoch": 0.8370786516853933,
"grad_norm": 0.90306556224823,
"learning_rate": 4.780289340384569e-06,
"loss": 0.729,
"step": 2384
},
{
"epoch": 0.8374297752808989,
"grad_norm": 0.8692042231559753,
"learning_rate": 4.780099779111901e-06,
"loss": 0.6695,
"step": 2385
},
{
"epoch": 0.8377808988764045,
"grad_norm": 0.8926328420639038,
"learning_rate": 4.779910139861771e-06,
"loss": 0.6895,
"step": 2386
},
{
"epoch": 0.8381320224719101,
"grad_norm": 0.8823839426040649,
"learning_rate": 4.779720422640666e-06,
"loss": 0.6587,
"step": 2387
},
{
"epoch": 0.8384831460674157,
"grad_norm": 0.8437162637710571,
"learning_rate": 4.7795306274550735e-06,
"loss": 0.6577,
"step": 2388
},
{
"epoch": 0.8388342696629213,
"grad_norm": 0.8754181265830994,
"learning_rate": 4.779340754311483e-06,
"loss": 0.6751,
"step": 2389
},
{
"epoch": 0.839185393258427,
"grad_norm": 0.8797729015350342,
"learning_rate": 4.77915080321639e-06,
"loss": 0.675,
"step": 2390
},
{
"epoch": 0.8395365168539326,
"grad_norm": 0.8657658100128174,
"learning_rate": 4.778960774176291e-06,
"loss": 0.6504,
"step": 2391
},
{
"epoch": 0.8398876404494382,
"grad_norm": 0.8967252969741821,
"learning_rate": 4.778770667197682e-06,
"loss": 0.6623,
"step": 2392
},
{
"epoch": 0.8402387640449438,
"grad_norm": 0.881457507610321,
"learning_rate": 4.778580482287067e-06,
"loss": 0.6514,
"step": 2393
},
{
"epoch": 0.8405898876404494,
"grad_norm": 0.8704044222831726,
"learning_rate": 4.778390219450949e-06,
"loss": 0.6695,
"step": 2394
},
{
"epoch": 0.8409410112359551,
"grad_norm": 0.9062277674674988,
"learning_rate": 4.778199878695835e-06,
"loss": 0.6394,
"step": 2395
},
{
"epoch": 0.8412921348314607,
"grad_norm": 0.873767614364624,
"learning_rate": 4.778009460028235e-06,
"loss": 0.6059,
"step": 2396
},
{
"epoch": 0.8416432584269663,
"grad_norm": 0.9467976689338684,
"learning_rate": 4.77781896345466e-06,
"loss": 0.6834,
"step": 2397
},
{
"epoch": 0.8419943820224719,
"grad_norm": 0.8977733254432678,
"learning_rate": 4.777628388981626e-06,
"loss": 0.703,
"step": 2398
},
{
"epoch": 0.8423455056179775,
"grad_norm": 0.8968080282211304,
"learning_rate": 4.7774377366156505e-06,
"loss": 0.678,
"step": 2399
},
{
"epoch": 0.8426966292134831,
"grad_norm": 0.8892068862915039,
"learning_rate": 4.777247006363253e-06,
"loss": 0.7203,
"step": 2400
},
{
"epoch": 0.8430477528089888,
"grad_norm": 0.8844308853149414,
"learning_rate": 4.777056198230956e-06,
"loss": 0.6694,
"step": 2401
},
{
"epoch": 0.8433988764044944,
"grad_norm": 0.8951604962348938,
"learning_rate": 4.7768653122252865e-06,
"loss": 0.6701,
"step": 2402
},
{
"epoch": 0.84375,
"grad_norm": 0.8704971075057983,
"learning_rate": 4.7766743483527715e-06,
"loss": 0.6616,
"step": 2403
},
{
"epoch": 0.8441011235955056,
"grad_norm": 0.908189058303833,
"learning_rate": 4.776483306619941e-06,
"loss": 0.6668,
"step": 2404
},
{
"epoch": 0.8444522471910112,
"grad_norm": 0.8829559087753296,
"learning_rate": 4.77629218703333e-06,
"loss": 0.6891,
"step": 2405
},
{
"epoch": 0.8448033707865169,
"grad_norm": 0.912239134311676,
"learning_rate": 4.776100989599475e-06,
"loss": 0.7016,
"step": 2406
},
{
"epoch": 0.8451544943820225,
"grad_norm": 0.8914303779602051,
"learning_rate": 4.775909714324912e-06,
"loss": 0.6593,
"step": 2407
},
{
"epoch": 0.8455056179775281,
"grad_norm": 0.9197940826416016,
"learning_rate": 4.775718361216186e-06,
"loss": 0.677,
"step": 2408
},
{
"epoch": 0.8458567415730337,
"grad_norm": 0.9217466115951538,
"learning_rate": 4.775526930279839e-06,
"loss": 0.7026,
"step": 2409
},
{
"epoch": 0.8462078651685393,
"grad_norm": 0.8749791383743286,
"learning_rate": 4.775335421522418e-06,
"loss": 0.6502,
"step": 2410
},
{
"epoch": 0.8465589887640449,
"grad_norm": 0.8961182236671448,
"learning_rate": 4.7751438349504716e-06,
"loss": 0.651,
"step": 2411
},
{
"epoch": 0.8469101123595506,
"grad_norm": 0.8679896593093872,
"learning_rate": 4.7749521705705546e-06,
"loss": 0.6445,
"step": 2412
},
{
"epoch": 0.8472612359550562,
"grad_norm": 0.8874735236167908,
"learning_rate": 4.774760428389219e-06,
"loss": 0.685,
"step": 2413
},
{
"epoch": 0.8476123595505618,
"grad_norm": 0.8967884182929993,
"learning_rate": 4.774568608413024e-06,
"loss": 0.6812,
"step": 2414
},
{
"epoch": 0.8479634831460674,
"grad_norm": 0.8984326720237732,
"learning_rate": 4.774376710648528e-06,
"loss": 0.6847,
"step": 2415
},
{
"epoch": 0.848314606741573,
"grad_norm": 0.861842155456543,
"learning_rate": 4.7741847351022955e-06,
"loss": 0.6643,
"step": 2416
},
{
"epoch": 0.8486657303370787,
"grad_norm": 0.8478919267654419,
"learning_rate": 4.773992681780891e-06,
"loss": 0.6309,
"step": 2417
},
{
"epoch": 0.8490168539325843,
"grad_norm": 0.8753276467323303,
"learning_rate": 4.773800550690882e-06,
"loss": 0.6906,
"step": 2418
},
{
"epoch": 0.8493679775280899,
"grad_norm": 0.9123186469078064,
"learning_rate": 4.7736083418388404e-06,
"loss": 0.6698,
"step": 2419
},
{
"epoch": 0.8497191011235955,
"grad_norm": 0.8448005318641663,
"learning_rate": 4.773416055231339e-06,
"loss": 0.6671,
"step": 2420
},
{
"epoch": 0.8500702247191011,
"grad_norm": 0.886045515537262,
"learning_rate": 4.773223690874954e-06,
"loss": 0.6925,
"step": 2421
},
{
"epoch": 0.8504213483146067,
"grad_norm": 0.8761569857597351,
"learning_rate": 4.773031248776264e-06,
"loss": 0.6988,
"step": 2422
},
{
"epoch": 0.8507724719101124,
"grad_norm": 0.8618139028549194,
"learning_rate": 4.772838728941851e-06,
"loss": 0.6345,
"step": 2423
},
{
"epoch": 0.851123595505618,
"grad_norm": 0.8982943296432495,
"learning_rate": 4.772646131378297e-06,
"loss": 0.6775,
"step": 2424
},
{
"epoch": 0.8514747191011236,
"grad_norm": 0.9190080761909485,
"learning_rate": 4.772453456092191e-06,
"loss": 0.6683,
"step": 2425
},
{
"epoch": 0.8518258426966292,
"grad_norm": 0.8613485097885132,
"learning_rate": 4.772260703090121e-06,
"loss": 0.6378,
"step": 2426
},
{
"epoch": 0.8521769662921348,
"grad_norm": 0.8885965943336487,
"learning_rate": 4.7720678723786796e-06,
"loss": 0.6971,
"step": 2427
},
{
"epoch": 0.8525280898876404,
"grad_norm": 0.8576337099075317,
"learning_rate": 4.771874963964461e-06,
"loss": 0.6417,
"step": 2428
},
{
"epoch": 0.8528792134831461,
"grad_norm": 0.8758781552314758,
"learning_rate": 4.771681977854062e-06,
"loss": 0.668,
"step": 2429
},
{
"epoch": 0.8532303370786517,
"grad_norm": 0.881989061832428,
"learning_rate": 4.771488914054085e-06,
"loss": 0.655,
"step": 2430
},
{
"epoch": 0.8535814606741573,
"grad_norm": 0.8963753581047058,
"learning_rate": 4.77129577257113e-06,
"loss": 0.647,
"step": 2431
},
{
"epoch": 0.8539325842696629,
"grad_norm": 0.8885678648948669,
"learning_rate": 4.771102553411803e-06,
"loss": 0.6832,
"step": 2432
},
{
"epoch": 0.8542837078651685,
"grad_norm": 0.8854667544364929,
"learning_rate": 4.770909256582712e-06,
"loss": 0.6387,
"step": 2433
},
{
"epoch": 0.8546348314606742,
"grad_norm": 0.9203517436981201,
"learning_rate": 4.770715882090468e-06,
"loss": 0.6606,
"step": 2434
},
{
"epoch": 0.8549859550561798,
"grad_norm": 0.9271900653839111,
"learning_rate": 4.770522429941683e-06,
"loss": 0.6787,
"step": 2435
},
{
"epoch": 0.8553370786516854,
"grad_norm": 0.8828155994415283,
"learning_rate": 4.770328900142975e-06,
"loss": 0.6742,
"step": 2436
},
{
"epoch": 0.855688202247191,
"grad_norm": 0.8759010434150696,
"learning_rate": 4.770135292700961e-06,
"loss": 0.6742,
"step": 2437
},
{
"epoch": 0.8560393258426966,
"grad_norm": 0.9363099932670593,
"learning_rate": 4.769941607622262e-06,
"loss": 0.6997,
"step": 2438
},
{
"epoch": 0.8563904494382022,
"grad_norm": 0.89815753698349,
"learning_rate": 4.769747844913503e-06,
"loss": 0.6643,
"step": 2439
},
{
"epoch": 0.8567415730337079,
"grad_norm": 0.8725033402442932,
"learning_rate": 4.76955400458131e-06,
"loss": 0.6433,
"step": 2440
},
{
"epoch": 0.8570926966292135,
"grad_norm": 0.8719140291213989,
"learning_rate": 4.769360086632312e-06,
"loss": 0.6215,
"step": 2441
},
{
"epoch": 0.8574438202247191,
"grad_norm": 0.8736863732337952,
"learning_rate": 4.76916609107314e-06,
"loss": 0.6854,
"step": 2442
},
{
"epoch": 0.8577949438202247,
"grad_norm": 0.8882306814193726,
"learning_rate": 4.768972017910432e-06,
"loss": 0.678,
"step": 2443
},
{
"epoch": 0.8581460674157303,
"grad_norm": 0.8788784146308899,
"learning_rate": 4.76877786715082e-06,
"loss": 0.6413,
"step": 2444
},
{
"epoch": 0.858497191011236,
"grad_norm": 0.8777339458465576,
"learning_rate": 4.768583638800947e-06,
"loss": 0.703,
"step": 2445
},
{
"epoch": 0.8588483146067416,
"grad_norm": 0.8905528783798218,
"learning_rate": 4.7683893328674556e-06,
"loss": 0.6653,
"step": 2446
},
{
"epoch": 0.8591994382022472,
"grad_norm": 0.8809261322021484,
"learning_rate": 4.76819494935699e-06,
"loss": 0.6357,
"step": 2447
},
{
"epoch": 0.8595505617977528,
"grad_norm": 0.9078806638717651,
"learning_rate": 4.768000488276198e-06,
"loss": 0.6924,
"step": 2448
},
{
"epoch": 0.8599016853932584,
"grad_norm": 0.8684207201004028,
"learning_rate": 4.767805949631729e-06,
"loss": 0.6753,
"step": 2449
},
{
"epoch": 0.860252808988764,
"grad_norm": 0.9299638867378235,
"learning_rate": 4.767611333430238e-06,
"loss": 0.6986,
"step": 2450
},
{
"epoch": 0.8606039325842697,
"grad_norm": 0.9025764465332031,
"learning_rate": 4.7674166396783796e-06,
"loss": 0.6832,
"step": 2451
},
{
"epoch": 0.8609550561797753,
"grad_norm": 0.8947761058807373,
"learning_rate": 4.767221868382812e-06,
"loss": 0.6927,
"step": 2452
},
{
"epoch": 0.8613061797752809,
"grad_norm": 0.9044678807258606,
"learning_rate": 4.767027019550198e-06,
"loss": 0.6803,
"step": 2453
},
{
"epoch": 0.8616573033707865,
"grad_norm": 0.8838576078414917,
"learning_rate": 4.766832093187199e-06,
"loss": 0.7069,
"step": 2454
},
{
"epoch": 0.8620084269662921,
"grad_norm": 0.8987936973571777,
"learning_rate": 4.766637089300481e-06,
"loss": 0.6986,
"step": 2455
},
{
"epoch": 0.8623595505617978,
"grad_norm": 0.9236810207366943,
"learning_rate": 4.766442007896716e-06,
"loss": 0.6912,
"step": 2456
},
{
"epoch": 0.8627106741573034,
"grad_norm": 0.9139038920402527,
"learning_rate": 4.766246848982573e-06,
"loss": 0.6647,
"step": 2457
},
{
"epoch": 0.863061797752809,
"grad_norm": 0.8569420576095581,
"learning_rate": 4.766051612564726e-06,
"loss": 0.6092,
"step": 2458
},
{
"epoch": 0.8634129213483146,
"grad_norm": 0.895006537437439,
"learning_rate": 4.765856298649855e-06,
"loss": 0.7071,
"step": 2459
},
{
"epoch": 0.8637640449438202,
"grad_norm": 0.8703930377960205,
"learning_rate": 4.7656609072446355e-06,
"loss": 0.6839,
"step": 2460
},
{
"epoch": 0.8641151685393258,
"grad_norm": 0.8899871706962585,
"learning_rate": 4.765465438355753e-06,
"loss": 0.6543,
"step": 2461
},
{
"epoch": 0.8644662921348315,
"grad_norm": 0.929097592830658,
"learning_rate": 4.765269891989891e-06,
"loss": 0.6848,
"step": 2462
},
{
"epoch": 0.8648174157303371,
"grad_norm": 0.8808541297912598,
"learning_rate": 4.765074268153736e-06,
"loss": 0.6625,
"step": 2463
},
{
"epoch": 0.8651685393258427,
"grad_norm": 0.9321450591087341,
"learning_rate": 4.764878566853981e-06,
"loss": 0.6483,
"step": 2464
},
{
"epoch": 0.8655196629213483,
"grad_norm": 0.8869593143463135,
"learning_rate": 4.764682788097317e-06,
"loss": 0.6546,
"step": 2465
},
{
"epoch": 0.8658707865168539,
"grad_norm": 0.9124717116355896,
"learning_rate": 4.764486931890438e-06,
"loss": 0.6637,
"step": 2466
},
{
"epoch": 0.8662219101123596,
"grad_norm": 0.9037688374519348,
"learning_rate": 4.764290998240044e-06,
"loss": 0.6656,
"step": 2467
},
{
"epoch": 0.8665730337078652,
"grad_norm": 0.9029697179794312,
"learning_rate": 4.7640949871528365e-06,
"loss": 0.6837,
"step": 2468
},
{
"epoch": 0.8669241573033708,
"grad_norm": 0.9010921716690063,
"learning_rate": 4.763898898635517e-06,
"loss": 0.6527,
"step": 2469
},
{
"epoch": 0.8672752808988764,
"grad_norm": 0.8654772043228149,
"learning_rate": 4.763702732694792e-06,
"loss": 0.6983,
"step": 2470
},
{
"epoch": 0.867626404494382,
"grad_norm": 0.8883764147758484,
"learning_rate": 4.763506489337372e-06,
"loss": 0.7035,
"step": 2471
},
{
"epoch": 0.8679775280898876,
"grad_norm": 0.902722179889679,
"learning_rate": 4.763310168569966e-06,
"loss": 0.6928,
"step": 2472
},
{
"epoch": 0.8683286516853933,
"grad_norm": 0.900360643863678,
"learning_rate": 4.76311377039929e-06,
"loss": 0.653,
"step": 2473
},
{
"epoch": 0.8686797752808989,
"grad_norm": 0.8905735015869141,
"learning_rate": 4.762917294832058e-06,
"loss": 0.6644,
"step": 2474
},
{
"epoch": 0.8690308988764045,
"grad_norm": 0.8971735835075378,
"learning_rate": 4.762720741874992e-06,
"loss": 0.6751,
"step": 2475
},
{
"epoch": 0.8693820224719101,
"grad_norm": 0.9062357544898987,
"learning_rate": 4.762524111534813e-06,
"loss": 0.6207,
"step": 2476
},
{
"epoch": 0.8697331460674157,
"grad_norm": 0.8765476942062378,
"learning_rate": 4.762327403818245e-06,
"loss": 0.6972,
"step": 2477
},
{
"epoch": 0.8700842696629213,
"grad_norm": 0.8708880543708801,
"learning_rate": 4.762130618732016e-06,
"loss": 0.6704,
"step": 2478
},
{
"epoch": 0.870435393258427,
"grad_norm": 0.8837103247642517,
"learning_rate": 4.761933756282856e-06,
"loss": 0.6314,
"step": 2479
},
{
"epoch": 0.8707865168539326,
"grad_norm": 0.8385997414588928,
"learning_rate": 4.7617368164774956e-06,
"loss": 0.6511,
"step": 2480
},
{
"epoch": 0.8711376404494382,
"grad_norm": 0.8760436773300171,
"learning_rate": 4.761539799322672e-06,
"loss": 0.6927,
"step": 2481
},
{
"epoch": 0.8714887640449438,
"grad_norm": 0.8821952939033508,
"learning_rate": 4.761342704825124e-06,
"loss": 0.6682,
"step": 2482
},
{
"epoch": 0.8718398876404494,
"grad_norm": 0.8891538381576538,
"learning_rate": 4.761145532991589e-06,
"loss": 0.6796,
"step": 2483
},
{
"epoch": 0.8721910112359551,
"grad_norm": 0.9022207260131836,
"learning_rate": 4.760948283828812e-06,
"loss": 0.7124,
"step": 2484
},
{
"epoch": 0.8725421348314607,
"grad_norm": 0.8623314499855042,
"learning_rate": 4.760750957343538e-06,
"loss": 0.6677,
"step": 2485
},
{
"epoch": 0.8728932584269663,
"grad_norm": 0.8784610033035278,
"learning_rate": 4.760553553542517e-06,
"loss": 0.6731,
"step": 2486
},
{
"epoch": 0.8732443820224719,
"grad_norm": 0.8761183619499207,
"learning_rate": 4.7603560724324985e-06,
"loss": 0.6246,
"step": 2487
},
{
"epoch": 0.8735955056179775,
"grad_norm": 0.8424711227416992,
"learning_rate": 4.7601585140202354e-06,
"loss": 0.6453,
"step": 2488
},
{
"epoch": 0.8739466292134831,
"grad_norm": 0.8858649730682373,
"learning_rate": 4.7599608783124865e-06,
"loss": 0.6869,
"step": 2489
},
{
"epoch": 0.8742977528089888,
"grad_norm": 0.8897294402122498,
"learning_rate": 4.759763165316008e-06,
"loss": 0.673,
"step": 2490
},
{
"epoch": 0.8746488764044944,
"grad_norm": 0.8692971467971802,
"learning_rate": 4.759565375037564e-06,
"loss": 0.6691,
"step": 2491
},
{
"epoch": 0.875,
"grad_norm": 0.9153478741645813,
"learning_rate": 4.759367507483918e-06,
"loss": 0.7012,
"step": 2492
},
{
"epoch": 0.8753511235955056,
"grad_norm": 0.9016615748405457,
"learning_rate": 4.759169562661836e-06,
"loss": 0.6758,
"step": 2493
},
{
"epoch": 0.8757022471910112,
"grad_norm": 0.8769146203994751,
"learning_rate": 4.758971540578088e-06,
"loss": 0.6989,
"step": 2494
},
{
"epoch": 0.8760533707865169,
"grad_norm": 0.8640443086624146,
"learning_rate": 4.758773441239447e-06,
"loss": 0.6355,
"step": 2495
},
{
"epoch": 0.8764044943820225,
"grad_norm": 0.9241236448287964,
"learning_rate": 4.758575264652687e-06,
"loss": 0.642,
"step": 2496
},
{
"epoch": 0.8767556179775281,
"grad_norm": 0.8472598791122437,
"learning_rate": 4.758377010824585e-06,
"loss": 0.6229,
"step": 2497
},
{
"epoch": 0.8771067415730337,
"grad_norm": 0.8446848392486572,
"learning_rate": 4.758178679761922e-06,
"loss": 0.6234,
"step": 2498
},
{
"epoch": 0.8774578651685393,
"grad_norm": 0.8826449513435364,
"learning_rate": 4.7579802714714815e-06,
"loss": 0.6799,
"step": 2499
},
{
"epoch": 0.8778089887640449,
"grad_norm": 0.8934155106544495,
"learning_rate": 4.757781785960047e-06,
"loss": 0.655,
"step": 2500
},
{
"epoch": 0.8781601123595506,
"grad_norm": 0.8733129501342773,
"learning_rate": 4.757583223234408e-06,
"loss": 0.6754,
"step": 2501
},
{
"epoch": 0.8785112359550562,
"grad_norm": 0.8651498556137085,
"learning_rate": 4.757384583301353e-06,
"loss": 0.6839,
"step": 2502
},
{
"epoch": 0.8788623595505618,
"grad_norm": 0.9135331511497498,
"learning_rate": 4.757185866167679e-06,
"loss": 0.6593,
"step": 2503
},
{
"epoch": 0.8792134831460674,
"grad_norm": 0.86785888671875,
"learning_rate": 4.756987071840179e-06,
"loss": 0.6568,
"step": 2504
},
{
"epoch": 0.879564606741573,
"grad_norm": 0.9271560311317444,
"learning_rate": 4.756788200325652e-06,
"loss": 0.6364,
"step": 2505
},
{
"epoch": 0.8799157303370787,
"grad_norm": 0.911849319934845,
"learning_rate": 4.7565892516309e-06,
"loss": 0.675,
"step": 2506
},
{
"epoch": 0.8802668539325843,
"grad_norm": 0.8646685481071472,
"learning_rate": 4.756390225762727e-06,
"loss": 0.6056,
"step": 2507
},
{
"epoch": 0.8806179775280899,
"grad_norm": 0.8820457458496094,
"learning_rate": 4.756191122727939e-06,
"loss": 0.6632,
"step": 2508
},
{
"epoch": 0.8809691011235955,
"grad_norm": 0.8895183801651001,
"learning_rate": 4.755991942533344e-06,
"loss": 0.6819,
"step": 2509
},
{
"epoch": 0.8813202247191011,
"grad_norm": 0.8657563924789429,
"learning_rate": 4.755792685185758e-06,
"loss": 0.649,
"step": 2510
},
{
"epoch": 0.8816713483146067,
"grad_norm": 0.9095283150672913,
"learning_rate": 4.755593350691989e-06,
"loss": 0.718,
"step": 2511
},
{
"epoch": 0.8820224719101124,
"grad_norm": 0.8777075409889221,
"learning_rate": 4.75539393905886e-06,
"loss": 0.6915,
"step": 2512
},
{
"epoch": 0.882373595505618,
"grad_norm": 0.8759570121765137,
"learning_rate": 4.755194450293187e-06,
"loss": 0.6749,
"step": 2513
},
{
"epoch": 0.8827247191011236,
"grad_norm": 0.8764015436172485,
"learning_rate": 4.754994884401794e-06,
"loss": 0.6652,
"step": 2514
},
{
"epoch": 0.8830758426966292,
"grad_norm": 0.9175446033477783,
"learning_rate": 4.754795241391505e-06,
"loss": 0.6626,
"step": 2515
},
{
"epoch": 0.8834269662921348,
"grad_norm": 0.871394693851471,
"learning_rate": 4.754595521269149e-06,
"loss": 0.694,
"step": 2516
},
{
"epoch": 0.8837780898876404,
"grad_norm": 0.9084363579750061,
"learning_rate": 4.754395724041556e-06,
"loss": 0.6725,
"step": 2517
},
{
"epoch": 0.8841292134831461,
"grad_norm": 0.8901112079620361,
"learning_rate": 4.754195849715557e-06,
"loss": 0.6614,
"step": 2518
},
{
"epoch": 0.8844803370786517,
"grad_norm": 0.8882237672805786,
"learning_rate": 4.7539958982979894e-06,
"loss": 0.6605,
"step": 2519
},
{
"epoch": 0.8848314606741573,
"grad_norm": 0.899763286113739,
"learning_rate": 4.753795869795691e-06,
"loss": 0.6423,
"step": 2520
},
{
"epoch": 0.8851825842696629,
"grad_norm": 0.9296014904975891,
"learning_rate": 4.753595764215503e-06,
"loss": 0.7166,
"step": 2521
},
{
"epoch": 0.8855337078651685,
"grad_norm": 0.883310854434967,
"learning_rate": 4.753395581564267e-06,
"loss": 0.6404,
"step": 2522
},
{
"epoch": 0.8858848314606742,
"grad_norm": 0.8889510035514832,
"learning_rate": 4.753195321848831e-06,
"loss": 0.649,
"step": 2523
},
{
"epoch": 0.8862359550561798,
"grad_norm": 0.9097694754600525,
"learning_rate": 4.752994985076044e-06,
"loss": 0.6604,
"step": 2524
},
{
"epoch": 0.8865870786516854,
"grad_norm": 0.8716392517089844,
"learning_rate": 4.752794571252755e-06,
"loss": 0.6989,
"step": 2525
},
{
"epoch": 0.886938202247191,
"grad_norm": 0.8778386116027832,
"learning_rate": 4.75259408038582e-06,
"loss": 0.638,
"step": 2526
},
{
"epoch": 0.8872893258426966,
"grad_norm": 0.8789075016975403,
"learning_rate": 4.752393512482095e-06,
"loss": 0.6713,
"step": 2527
},
{
"epoch": 0.8876404494382022,
"grad_norm": 0.8854068517684937,
"learning_rate": 4.752192867548439e-06,
"loss": 0.6534,
"step": 2528
},
{
"epoch": 0.8879915730337079,
"grad_norm": 0.8878395557403564,
"learning_rate": 4.7519921455917145e-06,
"loss": 0.684,
"step": 2529
},
{
"epoch": 0.8883426966292135,
"grad_norm": 0.8584818840026855,
"learning_rate": 4.751791346618786e-06,
"loss": 0.6259,
"step": 2530
},
{
"epoch": 0.8886938202247191,
"grad_norm": 0.8890157341957092,
"learning_rate": 4.75159047063652e-06,
"loss": 0.6607,
"step": 2531
},
{
"epoch": 0.8890449438202247,
"grad_norm": 0.851600706577301,
"learning_rate": 4.751389517651787e-06,
"loss": 0.66,
"step": 2532
},
{
"epoch": 0.8893960674157303,
"grad_norm": 0.8763266801834106,
"learning_rate": 4.751188487671457e-06,
"loss": 0.6427,
"step": 2533
},
{
"epoch": 0.889747191011236,
"grad_norm": 0.9055187702178955,
"learning_rate": 4.75098738070241e-06,
"loss": 0.6804,
"step": 2534
},
{
"epoch": 0.8900983146067416,
"grad_norm": 0.8860245943069458,
"learning_rate": 4.750786196751519e-06,
"loss": 0.6534,
"step": 2535
},
{
"epoch": 0.8904494382022472,
"grad_norm": 0.8951229453086853,
"learning_rate": 4.7505849358256665e-06,
"loss": 0.6637,
"step": 2536
},
{
"epoch": 0.8908005617977528,
"grad_norm": 0.9056205153465271,
"learning_rate": 4.7503835979317345e-06,
"loss": 0.6742,
"step": 2537
},
{
"epoch": 0.8911516853932584,
"grad_norm": 0.8903935551643372,
"learning_rate": 4.75018218307661e-06,
"loss": 0.6889,
"step": 2538
},
{
"epoch": 0.891502808988764,
"grad_norm": 0.9006560444831848,
"learning_rate": 4.74998069126718e-06,
"loss": 0.7042,
"step": 2539
},
{
"epoch": 0.8918539325842697,
"grad_norm": 0.8635199069976807,
"learning_rate": 4.749779122510337e-06,
"loss": 0.6692,
"step": 2540
},
{
"epoch": 0.8922050561797753,
"grad_norm": 0.8788262009620667,
"learning_rate": 4.749577476812971e-06,
"loss": 0.6652,
"step": 2541
},
{
"epoch": 0.8925561797752809,
"grad_norm": 0.8723106384277344,
"learning_rate": 4.749375754181981e-06,
"loss": 0.6373,
"step": 2542
},
{
"epoch": 0.8929073033707865,
"grad_norm": 0.8863840103149414,
"learning_rate": 4.749173954624265e-06,
"loss": 0.6701,
"step": 2543
},
{
"epoch": 0.8932584269662921,
"grad_norm": 0.9265256524085999,
"learning_rate": 4.748972078146725e-06,
"loss": 0.6591,
"step": 2544
},
{
"epoch": 0.8936095505617978,
"grad_norm": 0.9035506844520569,
"learning_rate": 4.748770124756265e-06,
"loss": 0.6585,
"step": 2545
},
{
"epoch": 0.8939606741573034,
"grad_norm": 0.8865588307380676,
"learning_rate": 4.748568094459789e-06,
"loss": 0.6903,
"step": 2546
},
{
"epoch": 0.894311797752809,
"grad_norm": 0.8840475678443909,
"learning_rate": 4.74836598726421e-06,
"loss": 0.6704,
"step": 2547
},
{
"epoch": 0.8946629213483146,
"grad_norm": 0.8937922120094299,
"learning_rate": 4.748163803176437e-06,
"loss": 0.6649,
"step": 2548
},
{
"epoch": 0.8950140449438202,
"grad_norm": 0.897407054901123,
"learning_rate": 4.747961542203387e-06,
"loss": 0.6705,
"step": 2549
},
{
"epoch": 0.8953651685393258,
"grad_norm": 0.8580518364906311,
"learning_rate": 4.747759204351975e-06,
"loss": 0.6217,
"step": 2550
},
{
"epoch": 0.8957162921348315,
"grad_norm": 0.8982558250427246,
"learning_rate": 4.747556789629121e-06,
"loss": 0.6873,
"step": 2551
},
{
"epoch": 0.8960674157303371,
"grad_norm": 0.922828197479248,
"learning_rate": 4.747354298041749e-06,
"loss": 0.6833,
"step": 2552
},
{
"epoch": 0.8964185393258427,
"grad_norm": 0.9047195315361023,
"learning_rate": 4.747151729596783e-06,
"loss": 0.6731,
"step": 2553
},
{
"epoch": 0.8967696629213483,
"grad_norm": 0.9132267236709595,
"learning_rate": 4.7469490843011515e-06,
"loss": 0.6752,
"step": 2554
},
{
"epoch": 0.8971207865168539,
"grad_norm": 0.8471392393112183,
"learning_rate": 4.7467463621617825e-06,
"loss": 0.6552,
"step": 2555
},
{
"epoch": 0.8974719101123596,
"grad_norm": 0.8814915418624878,
"learning_rate": 4.746543563185612e-06,
"loss": 0.676,
"step": 2556
},
{
"epoch": 0.8978230337078652,
"grad_norm": 0.8573102355003357,
"learning_rate": 4.746340687379573e-06,
"loss": 0.6397,
"step": 2557
},
{
"epoch": 0.8981741573033708,
"grad_norm": 0.8620328903198242,
"learning_rate": 4.746137734750604e-06,
"loss": 0.625,
"step": 2558
},
{
"epoch": 0.8985252808988764,
"grad_norm": 0.8743171691894531,
"learning_rate": 4.745934705305648e-06,
"loss": 0.6729,
"step": 2559
},
{
"epoch": 0.898876404494382,
"grad_norm": 0.838996410369873,
"learning_rate": 4.745731599051646e-06,
"loss": 0.6507,
"step": 2560
},
{
"epoch": 0.8992275280898876,
"grad_norm": 0.9150429964065552,
"learning_rate": 4.745528415995546e-06,
"loss": 0.6869,
"step": 2561
},
{
"epoch": 0.8995786516853933,
"grad_norm": 0.8738797903060913,
"learning_rate": 4.745325156144295e-06,
"loss": 0.6843,
"step": 2562
},
{
"epoch": 0.8999297752808989,
"grad_norm": 0.871146559715271,
"learning_rate": 4.745121819504845e-06,
"loss": 0.6773,
"step": 2563
},
{
"epoch": 0.9002808988764045,
"grad_norm": 0.8615120649337769,
"learning_rate": 4.74491840608415e-06,
"loss": 0.6383,
"step": 2564
},
{
"epoch": 0.9006320224719101,
"grad_norm": 0.8572788834571838,
"learning_rate": 4.744714915889167e-06,
"loss": 0.6488,
"step": 2565
},
{
"epoch": 0.9009831460674157,
"grad_norm": 0.8589732050895691,
"learning_rate": 4.744511348926855e-06,
"loss": 0.6464,
"step": 2566
},
{
"epoch": 0.9013342696629213,
"grad_norm": 0.8711311221122742,
"learning_rate": 4.744307705204174e-06,
"loss": 0.6809,
"step": 2567
},
{
"epoch": 0.901685393258427,
"grad_norm": 0.9025472402572632,
"learning_rate": 4.744103984728091e-06,
"loss": 0.6973,
"step": 2568
},
{
"epoch": 0.9020365168539326,
"grad_norm": 0.8648887872695923,
"learning_rate": 4.743900187505572e-06,
"loss": 0.6666,
"step": 2569
},
{
"epoch": 0.9023876404494382,
"grad_norm": 0.8922040462493896,
"learning_rate": 4.7436963135435864e-06,
"loss": 0.6875,
"step": 2570
},
{
"epoch": 0.9027387640449438,
"grad_norm": 0.8693889379501343,
"learning_rate": 4.743492362849107e-06,
"loss": 0.6926,
"step": 2571
},
{
"epoch": 0.9030898876404494,
"grad_norm": 0.8809205889701843,
"learning_rate": 4.743288335429109e-06,
"loss": 0.6841,
"step": 2572
},
{
"epoch": 0.9034410112359551,
"grad_norm": 0.8538480401039124,
"learning_rate": 4.743084231290569e-06,
"loss": 0.6416,
"step": 2573
},
{
"epoch": 0.9037921348314607,
"grad_norm": 0.8628817200660706,
"learning_rate": 4.742880050440468e-06,
"loss": 0.649,
"step": 2574
},
{
"epoch": 0.9041432584269663,
"grad_norm": 0.9055478572845459,
"learning_rate": 4.742675792885788e-06,
"loss": 0.6819,
"step": 2575
},
{
"epoch": 0.9044943820224719,
"grad_norm": 0.8601617217063904,
"learning_rate": 4.742471458633516e-06,
"loss": 0.6675,
"step": 2576
},
{
"epoch": 0.9048455056179775,
"grad_norm": 0.8635382056236267,
"learning_rate": 4.742267047690638e-06,
"loss": 0.6537,
"step": 2577
},
{
"epoch": 0.9051966292134831,
"grad_norm": 0.892162561416626,
"learning_rate": 4.742062560064146e-06,
"loss": 0.6627,
"step": 2578
},
{
"epoch": 0.9055477528089888,
"grad_norm": 0.9016808867454529,
"learning_rate": 4.741857995761033e-06,
"loss": 0.666,
"step": 2579
},
{
"epoch": 0.9058988764044944,
"grad_norm": 0.8793184161186218,
"learning_rate": 4.741653354788295e-06,
"loss": 0.638,
"step": 2580
},
{
"epoch": 0.90625,
"grad_norm": 0.8721537590026855,
"learning_rate": 4.741448637152931e-06,
"loss": 0.6816,
"step": 2581
},
{
"epoch": 0.9066011235955056,
"grad_norm": 0.9141138792037964,
"learning_rate": 4.741243842861941e-06,
"loss": 0.6959,
"step": 2582
},
{
"epoch": 0.9069522471910112,
"grad_norm": 0.8552782535552979,
"learning_rate": 4.741038971922329e-06,
"loss": 0.6868,
"step": 2583
},
{
"epoch": 0.9073033707865169,
"grad_norm": 0.8622416257858276,
"learning_rate": 4.740834024341103e-06,
"loss": 0.6861,
"step": 2584
},
{
"epoch": 0.9076544943820225,
"grad_norm": 0.8589463829994202,
"learning_rate": 4.74062900012527e-06,
"loss": 0.6525,
"step": 2585
},
{
"epoch": 0.9080056179775281,
"grad_norm": 0.9055203795433044,
"learning_rate": 4.740423899281843e-06,
"loss": 0.6724,
"step": 2586
},
{
"epoch": 0.9083567415730337,
"grad_norm": 0.9066650867462158,
"learning_rate": 4.740218721817836e-06,
"loss": 0.6929,
"step": 2587
},
{
"epoch": 0.9087078651685393,
"grad_norm": 0.8654912114143372,
"learning_rate": 4.7400134677402655e-06,
"loss": 0.6704,
"step": 2588
},
{
"epoch": 0.9090589887640449,
"grad_norm": 0.9301373958587646,
"learning_rate": 4.739808137056151e-06,
"loss": 0.6797,
"step": 2589
},
{
"epoch": 0.9094101123595506,
"grad_norm": 0.8802602291107178,
"learning_rate": 4.739602729772515e-06,
"loss": 0.6657,
"step": 2590
},
{
"epoch": 0.9097612359550562,
"grad_norm": 0.873859167098999,
"learning_rate": 4.739397245896382e-06,
"loss": 0.6364,
"step": 2591
},
{
"epoch": 0.9101123595505618,
"grad_norm": 0.8515645861625671,
"learning_rate": 4.739191685434779e-06,
"loss": 0.6472,
"step": 2592
},
{
"epoch": 0.9104634831460674,
"grad_norm": 0.8645114898681641,
"learning_rate": 4.738986048394737e-06,
"loss": 0.6792,
"step": 2593
},
{
"epoch": 0.910814606741573,
"grad_norm": 0.862397313117981,
"learning_rate": 4.738780334783288e-06,
"loss": 0.6522,
"step": 2594
},
{
"epoch": 0.9111657303370787,
"grad_norm": 0.8489444255828857,
"learning_rate": 4.738574544607467e-06,
"loss": 0.6757,
"step": 2595
},
{
"epoch": 0.9115168539325843,
"grad_norm": 0.8827484250068665,
"learning_rate": 4.738368677874313e-06,
"loss": 0.6842,
"step": 2596
},
{
"epoch": 0.9118679775280899,
"grad_norm": 0.8966019749641418,
"learning_rate": 4.738162734590865e-06,
"loss": 0.6658,
"step": 2597
},
{
"epoch": 0.9122191011235955,
"grad_norm": 0.8434582948684692,
"learning_rate": 4.7379567147641665e-06,
"loss": 0.6195,
"step": 2598
},
{
"epoch": 0.9125702247191011,
"grad_norm": 0.9089177250862122,
"learning_rate": 4.737750618401264e-06,
"loss": 0.7168,
"step": 2599
},
{
"epoch": 0.9129213483146067,
"grad_norm": 0.9108870625495911,
"learning_rate": 4.737544445509204e-06,
"loss": 0.6995,
"step": 2600
},
{
"epoch": 0.9132724719101124,
"grad_norm": 0.8936570882797241,
"learning_rate": 4.73733819609504e-06,
"loss": 0.6726,
"step": 2601
},
{
"epoch": 0.913623595505618,
"grad_norm": 0.8761941194534302,
"learning_rate": 4.737131870165824e-06,
"loss": 0.6642,
"step": 2602
},
{
"epoch": 0.9139747191011236,
"grad_norm": 0.870694100856781,
"learning_rate": 4.736925467728612e-06,
"loss": 0.6945,
"step": 2603
},
{
"epoch": 0.9143258426966292,
"grad_norm": 0.8931704759597778,
"learning_rate": 4.736718988790464e-06,
"loss": 0.6777,
"step": 2604
},
{
"epoch": 0.9146769662921348,
"grad_norm": 0.8875551223754883,
"learning_rate": 4.73651243335844e-06,
"loss": 0.6664,
"step": 2605
},
{
"epoch": 0.9150280898876404,
"grad_norm": 0.8898186087608337,
"learning_rate": 4.736305801439603e-06,
"loss": 0.6652,
"step": 2606
},
{
"epoch": 0.9153792134831461,
"grad_norm": 0.8893354535102844,
"learning_rate": 4.736099093041023e-06,
"loss": 0.6587,
"step": 2607
},
{
"epoch": 0.9157303370786517,
"grad_norm": 0.9278960824012756,
"learning_rate": 4.735892308169768e-06,
"loss": 0.6674,
"step": 2608
},
{
"epoch": 0.9160814606741573,
"grad_norm": 0.9112112522125244,
"learning_rate": 4.735685446832908e-06,
"loss": 0.6435,
"step": 2609
},
{
"epoch": 0.9164325842696629,
"grad_norm": 0.902930736541748,
"learning_rate": 4.735478509037519e-06,
"loss": 0.6402,
"step": 2610
},
{
"epoch": 0.9167837078651685,
"grad_norm": 0.8808385729789734,
"learning_rate": 4.7352714947906785e-06,
"loss": 0.707,
"step": 2611
},
{
"epoch": 0.9171348314606742,
"grad_norm": 0.8619163036346436,
"learning_rate": 4.735064404099465e-06,
"loss": 0.6728,
"step": 2612
},
{
"epoch": 0.9174859550561798,
"grad_norm": 0.8795300126075745,
"learning_rate": 4.734857236970962e-06,
"loss": 0.6543,
"step": 2613
},
{
"epoch": 0.9178370786516854,
"grad_norm": 0.9281381964683533,
"learning_rate": 4.734649993412254e-06,
"loss": 0.6594,
"step": 2614
},
{
"epoch": 0.918188202247191,
"grad_norm": 0.8637987971305847,
"learning_rate": 4.734442673430428e-06,
"loss": 0.6664,
"step": 2615
},
{
"epoch": 0.9185393258426966,
"grad_norm": 0.8663271069526672,
"learning_rate": 4.734235277032575e-06,
"loss": 0.6658,
"step": 2616
},
{
"epoch": 0.9188904494382022,
"grad_norm": 0.9028787016868591,
"learning_rate": 4.734027804225787e-06,
"loss": 0.6473,
"step": 2617
},
{
"epoch": 0.9192415730337079,
"grad_norm": 0.8826063275337219,
"learning_rate": 4.733820255017161e-06,
"loss": 0.626,
"step": 2618
},
{
"epoch": 0.9195926966292135,
"grad_norm": 0.8579033613204956,
"learning_rate": 4.7336126294137915e-06,
"loss": 0.6527,
"step": 2619
},
{
"epoch": 0.9199438202247191,
"grad_norm": 0.8992764353752136,
"learning_rate": 4.733404927422783e-06,
"loss": 0.6841,
"step": 2620
},
{
"epoch": 0.9202949438202247,
"grad_norm": 0.9202234148979187,
"learning_rate": 4.733197149051237e-06,
"loss": 0.6871,
"step": 2621
},
{
"epoch": 0.9206460674157303,
"grad_norm": 0.8881893754005432,
"learning_rate": 4.73298929430626e-06,
"loss": 0.6851,
"step": 2622
},
{
"epoch": 0.920997191011236,
"grad_norm": 0.9517401456832886,
"learning_rate": 4.732781363194958e-06,
"loss": 0.6748,
"step": 2623
},
{
"epoch": 0.9213483146067416,
"grad_norm": 0.9222083687782288,
"learning_rate": 4.7325733557244455e-06,
"loss": 0.6939,
"step": 2624
},
{
"epoch": 0.9216994382022472,
"grad_norm": 0.8639568090438843,
"learning_rate": 4.732365271901834e-06,
"loss": 0.664,
"step": 2625
},
{
"epoch": 0.9220505617977528,
"grad_norm": 0.9032568335533142,
"learning_rate": 4.732157111734241e-06,
"loss": 0.6779,
"step": 2626
},
{
"epoch": 0.9224016853932584,
"grad_norm": 0.9278046488761902,
"learning_rate": 4.731948875228784e-06,
"loss": 0.682,
"step": 2627
},
{
"epoch": 0.922752808988764,
"grad_norm": 0.9298899173736572,
"learning_rate": 4.731740562392587e-06,
"loss": 0.687,
"step": 2628
},
{
"epoch": 0.9231039325842697,
"grad_norm": 0.8798907995223999,
"learning_rate": 4.7315321732327715e-06,
"loss": 0.6722,
"step": 2629
},
{
"epoch": 0.9234550561797753,
"grad_norm": 0.8544273376464844,
"learning_rate": 4.731323707756465e-06,
"loss": 0.6801,
"step": 2630
},
{
"epoch": 0.9238061797752809,
"grad_norm": 0.8940860033035278,
"learning_rate": 4.731115165970797e-06,
"loss": 0.6548,
"step": 2631
},
{
"epoch": 0.9241573033707865,
"grad_norm": 0.9600421190261841,
"learning_rate": 4.730906547882901e-06,
"loss": 0.6935,
"step": 2632
},
{
"epoch": 0.9245084269662921,
"grad_norm": 0.9062276482582092,
"learning_rate": 4.730697853499908e-06,
"loss": 0.7043,
"step": 2633
},
{
"epoch": 0.9248595505617978,
"grad_norm": 0.8218638300895691,
"learning_rate": 4.730489082828959e-06,
"loss": 0.5993,
"step": 2634
},
{
"epoch": 0.9252106741573034,
"grad_norm": 0.8894577622413635,
"learning_rate": 4.7302802358771905e-06,
"loss": 0.6411,
"step": 2635
},
{
"epoch": 0.925561797752809,
"grad_norm": 0.881628155708313,
"learning_rate": 4.730071312651747e-06,
"loss": 0.6686,
"step": 2636
},
{
"epoch": 0.9259129213483146,
"grad_norm": 0.925878643989563,
"learning_rate": 4.7298623131597735e-06,
"loss": 0.6674,
"step": 2637
},
{
"epoch": 0.9262640449438202,
"grad_norm": 0.8464183211326599,
"learning_rate": 4.729653237408416e-06,
"loss": 0.6902,
"step": 2638
},
{
"epoch": 0.9266151685393258,
"grad_norm": 0.873918890953064,
"learning_rate": 4.729444085404826e-06,
"loss": 0.6737,
"step": 2639
},
{
"epoch": 0.9269662921348315,
"grad_norm": 0.8666852116584778,
"learning_rate": 4.7292348571561566e-06,
"loss": 0.6601,
"step": 2640
},
{
"epoch": 0.9273174157303371,
"grad_norm": 1.0832674503326416,
"learning_rate": 4.729025552669563e-06,
"loss": 0.654,
"step": 2641
},
{
"epoch": 0.9276685393258427,
"grad_norm": 0.8796826004981995,
"learning_rate": 4.728816171952202e-06,
"loss": 0.6365,
"step": 2642
},
{
"epoch": 0.9280196629213483,
"grad_norm": 0.8894056081771851,
"learning_rate": 4.728606715011235e-06,
"loss": 0.6955,
"step": 2643
},
{
"epoch": 0.9283707865168539,
"grad_norm": 0.9130917191505432,
"learning_rate": 4.728397181853826e-06,
"loss": 0.6807,
"step": 2644
},
{
"epoch": 0.9287219101123596,
"grad_norm": 0.8895679116249084,
"learning_rate": 4.72818757248714e-06,
"loss": 0.71,
"step": 2645
},
{
"epoch": 0.9290730337078652,
"grad_norm": 0.8796029090881348,
"learning_rate": 4.727977886918345e-06,
"loss": 0.666,
"step": 2646
},
{
"epoch": 0.9294241573033708,
"grad_norm": 0.8870053887367249,
"learning_rate": 4.727768125154614e-06,
"loss": 0.6724,
"step": 2647
},
{
"epoch": 0.9297752808988764,
"grad_norm": 0.8785473108291626,
"learning_rate": 4.727558287203119e-06,
"loss": 0.6735,
"step": 2648
},
{
"epoch": 0.930126404494382,
"grad_norm": 0.8722987174987793,
"learning_rate": 4.727348373071037e-06,
"loss": 0.6816,
"step": 2649
},
{
"epoch": 0.9304775280898876,
"grad_norm": 0.8995219469070435,
"learning_rate": 4.7271383827655455e-06,
"loss": 0.6657,
"step": 2650
},
{
"epoch": 0.9308286516853933,
"grad_norm": 0.8835882544517517,
"learning_rate": 4.726928316293828e-06,
"loss": 0.6973,
"step": 2651
},
{
"epoch": 0.9311797752808989,
"grad_norm": 0.8654088973999023,
"learning_rate": 4.726718173663068e-06,
"loss": 0.6083,
"step": 2652
},
{
"epoch": 0.9315308988764045,
"grad_norm": 0.832149088382721,
"learning_rate": 4.726507954880451e-06,
"loss": 0.5915,
"step": 2653
},
{
"epoch": 0.9318820224719101,
"grad_norm": 0.8820677995681763,
"learning_rate": 4.726297659953169e-06,
"loss": 0.641,
"step": 2654
},
{
"epoch": 0.9322331460674157,
"grad_norm": 0.9029566049575806,
"learning_rate": 4.7260872888884105e-06,
"loss": 0.6845,
"step": 2655
},
{
"epoch": 0.9325842696629213,
"grad_norm": 0.8395662307739258,
"learning_rate": 4.725876841693372e-06,
"loss": 0.6566,
"step": 2656
},
{
"epoch": 0.932935393258427,
"grad_norm": 0.894307017326355,
"learning_rate": 4.7256663183752495e-06,
"loss": 0.6985,
"step": 2657
},
{
"epoch": 0.9332865168539326,
"grad_norm": 0.9100427031517029,
"learning_rate": 4.725455718941245e-06,
"loss": 0.6687,
"step": 2658
},
{
"epoch": 0.9336376404494382,
"grad_norm": 0.874563455581665,
"learning_rate": 4.725245043398558e-06,
"loss": 0.6692,
"step": 2659
},
{
"epoch": 0.9339887640449438,
"grad_norm": 0.8920844793319702,
"learning_rate": 4.725034291754394e-06,
"loss": 0.6334,
"step": 2660
},
{
"epoch": 0.9343398876404494,
"grad_norm": 0.8333306908607483,
"learning_rate": 4.724823464015963e-06,
"loss": 0.6322,
"step": 2661
},
{
"epoch": 0.9346910112359551,
"grad_norm": 0.8971813917160034,
"learning_rate": 4.724612560190473e-06,
"loss": 0.6823,
"step": 2662
},
{
"epoch": 0.9350421348314607,
"grad_norm": 0.8798514008522034,
"learning_rate": 4.724401580285136e-06,
"loss": 0.6742,
"step": 2663
},
{
"epoch": 0.9353932584269663,
"grad_norm": 0.827732503414154,
"learning_rate": 4.724190524307169e-06,
"loss": 0.6574,
"step": 2664
},
{
"epoch": 0.9357443820224719,
"grad_norm": 0.8697452545166016,
"learning_rate": 4.7239793922637896e-06,
"loss": 0.637,
"step": 2665
},
{
"epoch": 0.9360955056179775,
"grad_norm": 0.887448251247406,
"learning_rate": 4.723768184162217e-06,
"loss": 0.6762,
"step": 2666
},
{
"epoch": 0.9364466292134831,
"grad_norm": 0.9078186750411987,
"learning_rate": 4.723556900009677e-06,
"loss": 0.6832,
"step": 2667
},
{
"epoch": 0.9367977528089888,
"grad_norm": 0.8617350459098816,
"learning_rate": 4.723345539813393e-06,
"loss": 0.6591,
"step": 2668
},
{
"epoch": 0.9371488764044944,
"grad_norm": 0.8676525354385376,
"learning_rate": 4.723134103580594e-06,
"loss": 0.6941,
"step": 2669
},
{
"epoch": 0.9375,
"grad_norm": 0.8551449179649353,
"learning_rate": 4.722922591318511e-06,
"loss": 0.6142,
"step": 2670
},
{
"epoch": 0.9378511235955056,
"grad_norm": 0.9027611017227173,
"learning_rate": 4.722711003034378e-06,
"loss": 0.6893,
"step": 2671
},
{
"epoch": 0.9382022471910112,
"grad_norm": 0.8602604269981384,
"learning_rate": 4.7224993387354305e-06,
"loss": 0.644,
"step": 2672
},
{
"epoch": 0.9385533707865169,
"grad_norm": 0.9275681376457214,
"learning_rate": 4.722287598428907e-06,
"loss": 0.676,
"step": 2673
},
{
"epoch": 0.9389044943820225,
"grad_norm": 0.880574107170105,
"learning_rate": 4.7220757821220495e-06,
"loss": 0.6068,
"step": 2674
},
{
"epoch": 0.9392556179775281,
"grad_norm": 0.8658181428909302,
"learning_rate": 4.721863889822102e-06,
"loss": 0.6714,
"step": 2675
},
{
"epoch": 0.9396067415730337,
"grad_norm": 0.890250563621521,
"learning_rate": 4.7216519215363095e-06,
"loss": 0.6624,
"step": 2676
},
{
"epoch": 0.9399578651685393,
"grad_norm": 0.870218813419342,
"learning_rate": 4.721439877271924e-06,
"loss": 0.6692,
"step": 2677
},
{
"epoch": 0.9403089887640449,
"grad_norm": 0.8716661334037781,
"learning_rate": 4.721227757036195e-06,
"loss": 0.6513,
"step": 2678
},
{
"epoch": 0.9406601123595506,
"grad_norm": 0.8979821801185608,
"learning_rate": 4.721015560836377e-06,
"loss": 0.6658,
"step": 2679
},
{
"epoch": 0.9410112359550562,
"grad_norm": 1.3192576169967651,
"learning_rate": 4.7208032886797285e-06,
"loss": 0.6726,
"step": 2680
},
{
"epoch": 0.9413623595505618,
"grad_norm": 0.9058758020401001,
"learning_rate": 4.720590940573507e-06,
"loss": 0.6764,
"step": 2681
},
{
"epoch": 0.9417134831460674,
"grad_norm": 0.8641902208328247,
"learning_rate": 4.720378516524976e-06,
"loss": 0.7075,
"step": 2682
},
{
"epoch": 0.942064606741573,
"grad_norm": 0.8593077659606934,
"learning_rate": 4.7201660165413994e-06,
"loss": 0.6575,
"step": 2683
},
{
"epoch": 0.9424157303370787,
"grad_norm": 0.8111205697059631,
"learning_rate": 4.719953440630045e-06,
"loss": 0.5865,
"step": 2684
},
{
"epoch": 0.9427668539325843,
"grad_norm": 0.8845921754837036,
"learning_rate": 4.719740788798182e-06,
"loss": 0.6787,
"step": 2685
},
{
"epoch": 0.9431179775280899,
"grad_norm": 0.893483579158783,
"learning_rate": 4.719528061053084e-06,
"loss": 0.657,
"step": 2686
},
{
"epoch": 0.9434691011235955,
"grad_norm": 0.882802426815033,
"learning_rate": 4.719315257402026e-06,
"loss": 0.6678,
"step": 2687
},
{
"epoch": 0.9438202247191011,
"grad_norm": 0.8567876815795898,
"learning_rate": 4.7191023778522844e-06,
"loss": 0.6167,
"step": 2688
},
{
"epoch": 0.9441713483146067,
"grad_norm": 0.8992242217063904,
"learning_rate": 4.718889422411141e-06,
"loss": 0.6691,
"step": 2689
},
{
"epoch": 0.9445224719101124,
"grad_norm": 0.8735215663909912,
"learning_rate": 4.718676391085878e-06,
"loss": 0.63,
"step": 2690
},
{
"epoch": 0.944873595505618,
"grad_norm": 0.8774729371070862,
"learning_rate": 4.71846328388378e-06,
"loss": 0.6702,
"step": 2691
},
{
"epoch": 0.9452247191011236,
"grad_norm": 0.9470273852348328,
"learning_rate": 4.718250100812138e-06,
"loss": 0.6997,
"step": 2692
},
{
"epoch": 0.9455758426966292,
"grad_norm": 0.9050356149673462,
"learning_rate": 4.71803684187824e-06,
"loss": 0.6451,
"step": 2693
},
{
"epoch": 0.9459269662921348,
"grad_norm": 0.9217947125434875,
"learning_rate": 4.7178235070893796e-06,
"loss": 0.6556,
"step": 2694
},
{
"epoch": 0.9462780898876404,
"grad_norm": 0.909965991973877,
"learning_rate": 4.717610096452852e-06,
"loss": 0.6772,
"step": 2695
},
{
"epoch": 0.9466292134831461,
"grad_norm": 0.8807794451713562,
"learning_rate": 4.717396609975959e-06,
"loss": 0.6805,
"step": 2696
},
{
"epoch": 0.9469803370786517,
"grad_norm": 0.9008699059486389,
"learning_rate": 4.717183047665998e-06,
"loss": 0.6718,
"step": 2697
},
{
"epoch": 0.9473314606741573,
"grad_norm": 0.9289125204086304,
"learning_rate": 4.716969409530274e-06,
"loss": 0.6705,
"step": 2698
},
{
"epoch": 0.9476825842696629,
"grad_norm": 0.9074335098266602,
"learning_rate": 4.716755695576094e-06,
"loss": 0.6897,
"step": 2699
},
{
"epoch": 0.9480337078651685,
"grad_norm": 0.9094159603118896,
"learning_rate": 4.716541905810766e-06,
"loss": 0.6625,
"step": 2700
},
{
"epoch": 0.9483848314606742,
"grad_norm": 0.8951701521873474,
"learning_rate": 4.716328040241601e-06,
"loss": 0.6234,
"step": 2701
},
{
"epoch": 0.9487359550561798,
"grad_norm": 0.8777403831481934,
"learning_rate": 4.716114098875914e-06,
"loss": 0.6418,
"step": 2702
},
{
"epoch": 0.9490870786516854,
"grad_norm": 0.9449620842933655,
"learning_rate": 4.715900081721021e-06,
"loss": 0.6814,
"step": 2703
},
{
"epoch": 0.949438202247191,
"grad_norm": 0.8588343262672424,
"learning_rate": 4.7156859887842416e-06,
"loss": 0.6042,
"step": 2704
},
{
"epoch": 0.9497893258426966,
"grad_norm": 0.8675066232681274,
"learning_rate": 4.715471820072896e-06,
"loss": 0.6591,
"step": 2705
},
{
"epoch": 0.9501404494382022,
"grad_norm": 0.8843618631362915,
"learning_rate": 4.715257575594312e-06,
"loss": 0.6594,
"step": 2706
},
{
"epoch": 0.9504915730337079,
"grad_norm": 0.8740993142127991,
"learning_rate": 4.715043255355814e-06,
"loss": 0.6578,
"step": 2707
},
{
"epoch": 0.9508426966292135,
"grad_norm": 0.8732353448867798,
"learning_rate": 4.7148288593647315e-06,
"loss": 0.6399,
"step": 2708
},
{
"epoch": 0.9511938202247191,
"grad_norm": 0.8841140866279602,
"learning_rate": 4.714614387628398e-06,
"loss": 0.6879,
"step": 2709
},
{
"epoch": 0.9515449438202247,
"grad_norm": 0.9016351103782654,
"learning_rate": 4.714399840154147e-06,
"loss": 0.6618,
"step": 2710
},
{
"epoch": 0.9518960674157303,
"grad_norm": 0.8737030625343323,
"learning_rate": 4.714185216949317e-06,
"loss": 0.6616,
"step": 2711
},
{
"epoch": 0.952247191011236,
"grad_norm": 0.8543287515640259,
"learning_rate": 4.713970518021246e-06,
"loss": 0.627,
"step": 2712
},
{
"epoch": 0.9525983146067416,
"grad_norm": 0.8959774374961853,
"learning_rate": 4.713755743377278e-06,
"loss": 0.6936,
"step": 2713
},
{
"epoch": 0.9529494382022472,
"grad_norm": 0.9039850831031799,
"learning_rate": 4.713540893024759e-06,
"loss": 0.6465,
"step": 2714
},
{
"epoch": 0.9533005617977528,
"grad_norm": 0.8787552118301392,
"learning_rate": 4.713325966971035e-06,
"loss": 0.6127,
"step": 2715
},
{
"epoch": 0.9536516853932584,
"grad_norm": 0.8473005294799805,
"learning_rate": 4.713110965223457e-06,
"loss": 0.6648,
"step": 2716
},
{
"epoch": 0.954002808988764,
"grad_norm": 0.937426745891571,
"learning_rate": 4.712895887789377e-06,
"loss": 0.6594,
"step": 2717
},
{
"epoch": 0.9543539325842697,
"grad_norm": 0.8719070553779602,
"learning_rate": 4.712680734676153e-06,
"loss": 0.6652,
"step": 2718
},
{
"epoch": 0.9547050561797753,
"grad_norm": 0.8872880935668945,
"learning_rate": 4.7124655058911395e-06,
"loss": 0.6987,
"step": 2719
},
{
"epoch": 0.9550561797752809,
"grad_norm": 0.8662359714508057,
"learning_rate": 4.7122502014417e-06,
"loss": 0.6723,
"step": 2720
},
{
"epoch": 0.9554073033707865,
"grad_norm": 0.8790738582611084,
"learning_rate": 4.712034821335196e-06,
"loss": 0.6421,
"step": 2721
},
{
"epoch": 0.9557584269662921,
"grad_norm": 0.8946492671966553,
"learning_rate": 4.711819365578995e-06,
"loss": 0.6439,
"step": 2722
},
{
"epoch": 0.9561095505617978,
"grad_norm": 0.9047174453735352,
"learning_rate": 4.711603834180464e-06,
"loss": 0.6859,
"step": 2723
},
{
"epoch": 0.9564606741573034,
"grad_norm": 0.9128047823905945,
"learning_rate": 4.711388227146974e-06,
"loss": 0.671,
"step": 2724
},
{
"epoch": 0.956811797752809,
"grad_norm": 0.8675251007080078,
"learning_rate": 4.711172544485898e-06,
"loss": 0.672,
"step": 2725
},
{
"epoch": 0.9571629213483146,
"grad_norm": 0.8741472363471985,
"learning_rate": 4.710956786204614e-06,
"loss": 0.6614,
"step": 2726
},
{
"epoch": 0.9575140449438202,
"grad_norm": 0.851613461971283,
"learning_rate": 4.710740952310501e-06,
"loss": 0.645,
"step": 2727
},
{
"epoch": 0.9578651685393258,
"grad_norm": 0.9370160698890686,
"learning_rate": 4.7105250428109375e-06,
"loss": 0.6743,
"step": 2728
},
{
"epoch": 0.9582162921348315,
"grad_norm": 0.9051042199134827,
"learning_rate": 4.71030905771331e-06,
"loss": 0.6166,
"step": 2729
},
{
"epoch": 0.9585674157303371,
"grad_norm": 0.8878385424613953,
"learning_rate": 4.7100929970250035e-06,
"loss": 0.6447,
"step": 2730
},
{
"epoch": 0.9589185393258427,
"grad_norm": 0.857430100440979,
"learning_rate": 4.709876860753409e-06,
"loss": 0.6557,
"step": 2731
},
{
"epoch": 0.9592696629213483,
"grad_norm": 0.8934675455093384,
"learning_rate": 4.709660648905916e-06,
"loss": 0.675,
"step": 2732
},
{
"epoch": 0.9596207865168539,
"grad_norm": 0.8855241537094116,
"learning_rate": 4.709444361489918e-06,
"loss": 0.6705,
"step": 2733
},
{
"epoch": 0.9599719101123596,
"grad_norm": 0.8473418354988098,
"learning_rate": 4.7092279985128165e-06,
"loss": 0.6398,
"step": 2734
},
{
"epoch": 0.9603230337078652,
"grad_norm": 0.9071454405784607,
"learning_rate": 4.709011559982006e-06,
"loss": 0.6824,
"step": 2735
},
{
"epoch": 0.9606741573033708,
"grad_norm": 0.8947198390960693,
"learning_rate": 4.708795045904891e-06,
"loss": 0.6794,
"step": 2736
},
{
"epoch": 0.9610252808988764,
"grad_norm": 0.8824878931045532,
"learning_rate": 4.708578456288875e-06,
"loss": 0.6453,
"step": 2737
},
{
"epoch": 0.961376404494382,
"grad_norm": 0.8897850513458252,
"learning_rate": 4.708361791141367e-06,
"loss": 0.6708,
"step": 2738
},
{
"epoch": 0.9617275280898876,
"grad_norm": 0.8753004670143127,
"learning_rate": 4.708145050469774e-06,
"loss": 0.6574,
"step": 2739
},
{
"epoch": 0.9620786516853933,
"grad_norm": 0.8702830672264099,
"learning_rate": 4.7079282342815105e-06,
"loss": 0.6363,
"step": 2740
},
{
"epoch": 0.9624297752808989,
"grad_norm": 0.8557823896408081,
"learning_rate": 4.707711342583991e-06,
"loss": 0.6607,
"step": 2741
},
{
"epoch": 0.9627808988764045,
"grad_norm": 0.9189360737800598,
"learning_rate": 4.7074943753846324e-06,
"loss": 0.6732,
"step": 2742
},
{
"epoch": 0.9631320224719101,
"grad_norm": 0.8712623715400696,
"learning_rate": 4.707277332690855e-06,
"loss": 0.6962,
"step": 2743
},
{
"epoch": 0.9634831460674157,
"grad_norm": 0.8758993744850159,
"learning_rate": 4.707060214510082e-06,
"loss": 0.688,
"step": 2744
},
{
"epoch": 0.9638342696629213,
"grad_norm": 0.9074702262878418,
"learning_rate": 4.706843020849738e-06,
"loss": 0.6668,
"step": 2745
},
{
"epoch": 0.964185393258427,
"grad_norm": 0.8798408508300781,
"learning_rate": 4.706625751717251e-06,
"loss": 0.6728,
"step": 2746
},
{
"epoch": 0.9645365168539326,
"grad_norm": 0.8691076636314392,
"learning_rate": 4.706408407120053e-06,
"loss": 0.6217,
"step": 2747
},
{
"epoch": 0.9648876404494382,
"grad_norm": 0.8737530708312988,
"learning_rate": 4.7061909870655734e-06,
"loss": 0.6618,
"step": 2748
},
{
"epoch": 0.9652387640449438,
"grad_norm": 0.893994152545929,
"learning_rate": 4.70597349156125e-06,
"loss": 0.6899,
"step": 2749
},
{
"epoch": 0.9655898876404494,
"grad_norm": 0.8533192276954651,
"learning_rate": 4.705755920614522e-06,
"loss": 0.601,
"step": 2750
},
{
"epoch": 0.9659410112359551,
"grad_norm": 0.8612004518508911,
"learning_rate": 4.705538274232829e-06,
"loss": 0.6407,
"step": 2751
},
{
"epoch": 0.9662921348314607,
"grad_norm": 0.8780084848403931,
"learning_rate": 4.7053205524236136e-06,
"loss": 0.6689,
"step": 2752
},
{
"epoch": 0.9666432584269663,
"grad_norm": 0.8959993124008179,
"learning_rate": 4.705102755194323e-06,
"loss": 0.6588,
"step": 2753
},
{
"epoch": 0.9669943820224719,
"grad_norm": 0.9221241474151611,
"learning_rate": 4.704884882552405e-06,
"loss": 0.6454,
"step": 2754
},
{
"epoch": 0.9673455056179775,
"grad_norm": 0.8571012020111084,
"learning_rate": 4.70466693450531e-06,
"loss": 0.6101,
"step": 2755
},
{
"epoch": 0.9676966292134831,
"grad_norm": 0.9270880222320557,
"learning_rate": 4.704448911060493e-06,
"loss": 0.6561,
"step": 2756
},
{
"epoch": 0.9680477528089888,
"grad_norm": 0.8872186541557312,
"learning_rate": 4.70423081222541e-06,
"loss": 0.6704,
"step": 2757
},
{
"epoch": 0.9683988764044944,
"grad_norm": 0.8734565377235413,
"learning_rate": 4.704012638007519e-06,
"loss": 0.6626,
"step": 2758
},
{
"epoch": 0.96875,
"grad_norm": 0.9009286761283875,
"learning_rate": 4.703794388414281e-06,
"loss": 0.6727,
"step": 2759
},
{
"epoch": 0.9691011235955056,
"grad_norm": 0.8499953746795654,
"learning_rate": 4.703576063453162e-06,
"loss": 0.6574,
"step": 2760
},
{
"epoch": 0.9694522471910112,
"grad_norm": 0.8824490308761597,
"learning_rate": 4.703357663131627e-06,
"loss": 0.6618,
"step": 2761
},
{
"epoch": 0.9698033707865169,
"grad_norm": 0.8579809665679932,
"learning_rate": 4.703139187457145e-06,
"loss": 0.6677,
"step": 2762
},
{
"epoch": 0.9701544943820225,
"grad_norm": 0.8739168643951416,
"learning_rate": 4.702920636437188e-06,
"loss": 0.6648,
"step": 2763
},
{
"epoch": 0.9705056179775281,
"grad_norm": 0.8557081818580627,
"learning_rate": 4.702702010079231e-06,
"loss": 0.6588,
"step": 2764
},
{
"epoch": 0.9708567415730337,
"grad_norm": 0.8667842149734497,
"learning_rate": 4.702483308390749e-06,
"loss": 0.6432,
"step": 2765
},
{
"epoch": 0.9712078651685393,
"grad_norm": 0.8367511034011841,
"learning_rate": 4.7022645313792235e-06,
"loss": 0.6428,
"step": 2766
},
{
"epoch": 0.9715589887640449,
"grad_norm": 0.8607786297798157,
"learning_rate": 4.702045679052136e-06,
"loss": 0.628,
"step": 2767
},
{
"epoch": 0.9719101123595506,
"grad_norm": 0.8698323369026184,
"learning_rate": 4.701826751416969e-06,
"loss": 0.6241,
"step": 2768
},
{
"epoch": 0.9722612359550562,
"grad_norm": 0.8620527386665344,
"learning_rate": 4.701607748481213e-06,
"loss": 0.6698,
"step": 2769
},
{
"epoch": 0.9726123595505618,
"grad_norm": 0.8832960724830627,
"learning_rate": 4.701388670252355e-06,
"loss": 0.6332,
"step": 2770
},
{
"epoch": 0.9729634831460674,
"grad_norm": 0.8846637606620789,
"learning_rate": 4.701169516737888e-06,
"loss": 0.6968,
"step": 2771
},
{
"epoch": 0.973314606741573,
"grad_norm": 0.8862726092338562,
"learning_rate": 4.7009502879453076e-06,
"loss": 0.7282,
"step": 2772
},
{
"epoch": 0.9736657303370787,
"grad_norm": 0.8758227229118347,
"learning_rate": 4.70073098388211e-06,
"loss": 0.658,
"step": 2773
},
{
"epoch": 0.9740168539325843,
"grad_norm": 0.8702247738838196,
"learning_rate": 4.700511604555796e-06,
"loss": 0.65,
"step": 2774
},
{
"epoch": 0.9743679775280899,
"grad_norm": 0.8507943153381348,
"learning_rate": 4.700292149973869e-06,
"loss": 0.6625,
"step": 2775
},
{
"epoch": 0.9747191011235955,
"grad_norm": 0.8701155185699463,
"learning_rate": 4.700072620143833e-06,
"loss": 0.6742,
"step": 2776
},
{
"epoch": 0.9750702247191011,
"grad_norm": 0.8407930135726929,
"learning_rate": 4.6998530150731955e-06,
"loss": 0.642,
"step": 2777
},
{
"epoch": 0.9754213483146067,
"grad_norm": 0.858070969581604,
"learning_rate": 4.6996333347694675e-06,
"loss": 0.6752,
"step": 2778
},
{
"epoch": 0.9757724719101124,
"grad_norm": 0.8673107624053955,
"learning_rate": 4.6994135792401615e-06,
"loss": 0.6826,
"step": 2779
},
{
"epoch": 0.976123595505618,
"grad_norm": 0.8591476082801819,
"learning_rate": 4.699193748492794e-06,
"loss": 0.6972,
"step": 2780
},
{
"epoch": 0.9764747191011236,
"grad_norm": 0.8738571405410767,
"learning_rate": 4.698973842534882e-06,
"loss": 0.6559,
"step": 2781
},
{
"epoch": 0.9768258426966292,
"grad_norm": 0.8602177500724792,
"learning_rate": 4.698753861373946e-06,
"loss": 0.6433,
"step": 2782
},
{
"epoch": 0.9771769662921348,
"grad_norm": 0.8604179620742798,
"learning_rate": 4.6985338050175096e-06,
"loss": 0.6499,
"step": 2783
},
{
"epoch": 0.9775280898876404,
"grad_norm": 0.8430880904197693,
"learning_rate": 4.698313673473098e-06,
"loss": 0.6008,
"step": 2784
},
{
"epoch": 0.9778792134831461,
"grad_norm": 0.8904216885566711,
"learning_rate": 4.69809346674824e-06,
"loss": 0.6869,
"step": 2785
},
{
"epoch": 0.9782303370786517,
"grad_norm": 0.8424625396728516,
"learning_rate": 4.6978731848504665e-06,
"loss": 0.6519,
"step": 2786
},
{
"epoch": 0.9785814606741573,
"grad_norm": 0.8511930704116821,
"learning_rate": 4.697652827787311e-06,
"loss": 0.6511,
"step": 2787
},
{
"epoch": 0.9789325842696629,
"grad_norm": 0.8731825351715088,
"learning_rate": 4.697432395566309e-06,
"loss": 0.6756,
"step": 2788
},
{
"epoch": 0.9792837078651685,
"grad_norm": 0.8926572203636169,
"learning_rate": 4.697211888194999e-06,
"loss": 0.6457,
"step": 2789
},
{
"epoch": 0.9796348314606742,
"grad_norm": 0.8796569108963013,
"learning_rate": 4.696991305680923e-06,
"loss": 0.6435,
"step": 2790
},
{
"epoch": 0.9799859550561798,
"grad_norm": 0.89739990234375,
"learning_rate": 4.6967706480316256e-06,
"loss": 0.666,
"step": 2791
},
{
"epoch": 0.9803370786516854,
"grad_norm": 0.8666273951530457,
"learning_rate": 4.69654991525465e-06,
"loss": 0.6658,
"step": 2792
},
{
"epoch": 0.980688202247191,
"grad_norm": 0.8648870587348938,
"learning_rate": 4.696329107357548e-06,
"loss": 0.6541,
"step": 2793
},
{
"epoch": 0.9810393258426966,
"grad_norm": 0.8609452247619629,
"learning_rate": 4.69610822434787e-06,
"loss": 0.6646,
"step": 2794
},
{
"epoch": 0.9813904494382022,
"grad_norm": 0.8994244933128357,
"learning_rate": 4.695887266233169e-06,
"loss": 0.6778,
"step": 2795
},
{
"epoch": 0.9817415730337079,
"grad_norm": 0.8689225316047668,
"learning_rate": 4.6956662330210035e-06,
"loss": 0.6581,
"step": 2796
},
{
"epoch": 0.9820926966292135,
"grad_norm": 0.8597514629364014,
"learning_rate": 4.695445124718932e-06,
"loss": 0.6322,
"step": 2797
},
{
"epoch": 0.9824438202247191,
"grad_norm": 0.8898642063140869,
"learning_rate": 4.695223941334514e-06,
"loss": 0.6866,
"step": 2798
},
{
"epoch": 0.9827949438202247,
"grad_norm": 0.9043296575546265,
"learning_rate": 4.695002682875317e-06,
"loss": 0.686,
"step": 2799
},
{
"epoch": 0.9831460674157303,
"grad_norm": 0.8787916898727417,
"learning_rate": 4.694781349348907e-06,
"loss": 0.6819,
"step": 2800
},
{
"epoch": 0.983497191011236,
"grad_norm": 0.86040198802948,
"learning_rate": 4.694559940762853e-06,
"loss": 0.6591,
"step": 2801
},
{
"epoch": 0.9838483146067416,
"grad_norm": 0.8670945167541504,
"learning_rate": 4.694338457124726e-06,
"loss": 0.6828,
"step": 2802
},
{
"epoch": 0.9841994382022472,
"grad_norm": 0.8818464875221252,
"learning_rate": 4.694116898442102e-06,
"loss": 0.6556,
"step": 2803
},
{
"epoch": 0.9845505617977528,
"grad_norm": 0.850309431552887,
"learning_rate": 4.693895264722557e-06,
"loss": 0.6803,
"step": 2804
},
{
"epoch": 0.9849016853932584,
"grad_norm": 0.8555431365966797,
"learning_rate": 4.6936735559736715e-06,
"loss": 0.6484,
"step": 2805
},
{
"epoch": 0.985252808988764,
"grad_norm": 0.8949291110038757,
"learning_rate": 4.693451772203028e-06,
"loss": 0.6889,
"step": 2806
},
{
"epoch": 0.9856039325842697,
"grad_norm": 0.8468831181526184,
"learning_rate": 4.6932299134182104e-06,
"loss": 0.6398,
"step": 2807
},
{
"epoch": 0.9859550561797753,
"grad_norm": 0.8642711043357849,
"learning_rate": 4.693007979626807e-06,
"loss": 0.6789,
"step": 2808
},
{
"epoch": 0.9863061797752809,
"grad_norm": 0.8652757406234741,
"learning_rate": 4.692785970836407e-06,
"loss": 0.6862,
"step": 2809
},
{
"epoch": 0.9866573033707865,
"grad_norm": 0.8652501106262207,
"learning_rate": 4.692563887054603e-06,
"loss": 0.647,
"step": 2810
},
{
"epoch": 0.9870084269662921,
"grad_norm": 0.8379038572311401,
"learning_rate": 4.69234172828899e-06,
"loss": 0.616,
"step": 2811
},
{
"epoch": 0.9873595505617978,
"grad_norm": 0.8986461758613586,
"learning_rate": 4.692119494547167e-06,
"loss": 0.6746,
"step": 2812
},
{
"epoch": 0.9877106741573034,
"grad_norm": 0.8548451662063599,
"learning_rate": 4.691897185836733e-06,
"loss": 0.6607,
"step": 2813
},
{
"epoch": 0.988061797752809,
"grad_norm": 0.8519273400306702,
"learning_rate": 4.69167480216529e-06,
"loss": 0.6341,
"step": 2814
},
{
"epoch": 0.9884129213483146,
"grad_norm": 0.8934301137924194,
"learning_rate": 4.691452343540445e-06,
"loss": 0.6601,
"step": 2815
},
{
"epoch": 0.9887640449438202,
"grad_norm": 0.8649933338165283,
"learning_rate": 4.6912298099698046e-06,
"loss": 0.6406,
"step": 2816
},
{
"epoch": 0.9891151685393258,
"grad_norm": 0.8976009488105774,
"learning_rate": 4.69100720146098e-06,
"loss": 0.6391,
"step": 2817
},
{
"epoch": 0.9894662921348315,
"grad_norm": 0.8619791865348816,
"learning_rate": 4.690784518021584e-06,
"loss": 0.6932,
"step": 2818
},
{
"epoch": 0.9898174157303371,
"grad_norm": 0.8594460487365723,
"learning_rate": 4.690561759659232e-06,
"loss": 0.6666,
"step": 2819
},
{
"epoch": 0.9901685393258427,
"grad_norm": 0.8755744099617004,
"learning_rate": 4.690338926381542e-06,
"loss": 0.6458,
"step": 2820
},
{
"epoch": 0.9905196629213483,
"grad_norm": 0.8844261765480042,
"learning_rate": 4.690116018196135e-06,
"loss": 0.6344,
"step": 2821
},
{
"epoch": 0.9908707865168539,
"grad_norm": 0.871732234954834,
"learning_rate": 4.689893035110635e-06,
"loss": 0.6316,
"step": 2822
},
{
"epoch": 0.9912219101123596,
"grad_norm": 0.8930003046989441,
"learning_rate": 4.689669977132667e-06,
"loss": 0.6887,
"step": 2823
},
{
"epoch": 0.9915730337078652,
"grad_norm": 0.9592486619949341,
"learning_rate": 4.689446844269859e-06,
"loss": 0.653,
"step": 2824
},
{
"epoch": 0.9919241573033708,
"grad_norm": 0.8865694999694824,
"learning_rate": 4.6892236365298425e-06,
"loss": 0.648,
"step": 2825
},
{
"epoch": 0.9922752808988764,
"grad_norm": 0.8786531686782837,
"learning_rate": 4.689000353920251e-06,
"loss": 0.647,
"step": 2826
},
{
"epoch": 0.992626404494382,
"grad_norm": 0.8980904817581177,
"learning_rate": 4.6887769964487215e-06,
"loss": 0.6556,
"step": 2827
},
{
"epoch": 0.9929775280898876,
"grad_norm": 0.8920478224754333,
"learning_rate": 4.688553564122891e-06,
"loss": 0.6802,
"step": 2828
},
{
"epoch": 0.9933286516853933,
"grad_norm": 0.8804338574409485,
"learning_rate": 4.688330056950401e-06,
"loss": 0.6695,
"step": 2829
},
{
"epoch": 0.9936797752808989,
"grad_norm": 0.8846019506454468,
"learning_rate": 4.688106474938896e-06,
"loss": 0.6484,
"step": 2830
},
{
"epoch": 0.9940308988764045,
"grad_norm": 0.8795493841171265,
"learning_rate": 4.687882818096022e-06,
"loss": 0.6465,
"step": 2831
},
{
"epoch": 0.9943820224719101,
"grad_norm": 0.858067512512207,
"learning_rate": 4.687659086429427e-06,
"loss": 0.6471,
"step": 2832
},
{
"epoch": 0.9947331460674157,
"grad_norm": 0.8714587092399597,
"learning_rate": 4.687435279946765e-06,
"loss": 0.6446,
"step": 2833
},
{
"epoch": 0.9950842696629213,
"grad_norm": 0.8674041628837585,
"learning_rate": 4.687211398655686e-06,
"loss": 0.6419,
"step": 2834
},
{
"epoch": 0.995435393258427,
"grad_norm": 0.8507161736488342,
"learning_rate": 4.6869874425638505e-06,
"loss": 0.6202,
"step": 2835
},
{
"epoch": 0.9957865168539326,
"grad_norm": 0.8718125224113464,
"learning_rate": 4.686763411678915e-06,
"loss": 0.6807,
"step": 2836
},
{
"epoch": 0.9961376404494382,
"grad_norm": 0.8641071319580078,
"learning_rate": 4.686539306008543e-06,
"loss": 0.6492,
"step": 2837
},
{
"epoch": 0.9964887640449438,
"grad_norm": 0.8599060773849487,
"learning_rate": 4.686315125560397e-06,
"loss": 0.6779,
"step": 2838
},
{
"epoch": 0.9968398876404494,
"grad_norm": 0.8444222211837769,
"learning_rate": 4.686090870342145e-06,
"loss": 0.6841,
"step": 2839
},
{
"epoch": 0.9971910112359551,
"grad_norm": 0.86264568567276,
"learning_rate": 4.685866540361456e-06,
"loss": 0.6627,
"step": 2840
},
{
"epoch": 0.9975421348314607,
"grad_norm": 0.9181270599365234,
"learning_rate": 4.685642135626001e-06,
"loss": 0.6537,
"step": 2841
},
{
"epoch": 0.9978932584269663,
"grad_norm": 0.8701741695404053,
"learning_rate": 4.685417656143456e-06,
"loss": 0.6545,
"step": 2842
},
{
"epoch": 0.9982443820224719,
"grad_norm": 0.8788411021232605,
"learning_rate": 4.685193101921497e-06,
"loss": 0.6431,
"step": 2843
},
{
"epoch": 0.9985955056179775,
"grad_norm": 0.8748607635498047,
"learning_rate": 4.684968472967804e-06,
"loss": 0.6698,
"step": 2844
},
{
"epoch": 0.9989466292134831,
"grad_norm": 0.9054082632064819,
"learning_rate": 4.684743769290059e-06,
"loss": 0.6801,
"step": 2845
},
{
"epoch": 0.9992977528089888,
"grad_norm": 0.8799982666969299,
"learning_rate": 4.684518990895945e-06,
"loss": 0.6764,
"step": 2846
},
{
"epoch": 0.9996488764044944,
"grad_norm": 0.9407274127006531,
"learning_rate": 4.684294137793153e-06,
"loss": 0.7098,
"step": 2847
},
{
"epoch": 1.0,
"grad_norm": 0.91474848985672,
"learning_rate": 4.684069209989369e-06,
"loss": 0.6617,
"step": 2848
}
],
"logging_steps": 1,
"max_steps": 17088,
"num_input_tokens_seen": 0,
"num_train_epochs": 6,
"save_steps": 2848,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 7.194152512373391e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}