SAWM / trainer_state.json
xue-26's picture
Upload 14 files
53798ca verified
Raw
History Blame Contribute Delete
220 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1268,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0007887990534411359,
"grad_norm": 23.821498367793616,
"learning_rate": 0.0,
"loss": 1.6363,
"step": 1
},
{
"epoch": 0.0015775981068822718,
"grad_norm": 23.9527566663704,
"learning_rate": 2.564102564102564e-08,
"loss": 1.6186,
"step": 2
},
{
"epoch": 0.0023663971603234074,
"grad_norm": 22.43554095167205,
"learning_rate": 5.128205128205128e-08,
"loss": 1.5698,
"step": 3
},
{
"epoch": 0.0031551962137645437,
"grad_norm": 22.121374266633797,
"learning_rate": 7.692307692307692e-08,
"loss": 1.5886,
"step": 4
},
{
"epoch": 0.003943995267205679,
"grad_norm": 26.838317494760666,
"learning_rate": 1.0256410256410256e-07,
"loss": 1.5974,
"step": 5
},
{
"epoch": 0.004732794320646815,
"grad_norm": 21.124805258198943,
"learning_rate": 1.2820512820512818e-07,
"loss": 1.5198,
"step": 6
},
{
"epoch": 0.005521593374087951,
"grad_norm": 21.579364005505965,
"learning_rate": 1.5384615384615385e-07,
"loss": 1.5174,
"step": 7
},
{
"epoch": 0.006310392427529087,
"grad_norm": 22.56736957610717,
"learning_rate": 1.7948717948717948e-07,
"loss": 1.5575,
"step": 8
},
{
"epoch": 0.007099191480970223,
"grad_norm": 22.728823094971418,
"learning_rate": 2.0512820512820512e-07,
"loss": 1.5802,
"step": 9
},
{
"epoch": 0.007887990534411358,
"grad_norm": 23.858162996871872,
"learning_rate": 2.3076923076923078e-07,
"loss": 1.63,
"step": 10
},
{
"epoch": 0.008676789587852495,
"grad_norm": 21.75503789864118,
"learning_rate": 2.5641025641025636e-07,
"loss": 1.5617,
"step": 11
},
{
"epoch": 0.00946558864129363,
"grad_norm": 23.83038646872311,
"learning_rate": 2.8205128205128203e-07,
"loss": 1.6325,
"step": 12
},
{
"epoch": 0.010254387694734766,
"grad_norm": 23.678900166762446,
"learning_rate": 3.076923076923077e-07,
"loss": 1.6027,
"step": 13
},
{
"epoch": 0.011043186748175901,
"grad_norm": 21.78825141828257,
"learning_rate": 3.333333333333333e-07,
"loss": 1.556,
"step": 14
},
{
"epoch": 0.011831985801617038,
"grad_norm": 21.081908380396943,
"learning_rate": 3.5897435897435896e-07,
"loss": 1.5294,
"step": 15
},
{
"epoch": 0.012620784855058175,
"grad_norm": 21.810402166235413,
"learning_rate": 3.8461538461538463e-07,
"loss": 1.6072,
"step": 16
},
{
"epoch": 0.01340958390849931,
"grad_norm": 21.425470700554122,
"learning_rate": 4.1025641025641024e-07,
"loss": 1.5201,
"step": 17
},
{
"epoch": 0.014198382961940446,
"grad_norm": 22.168697447769855,
"learning_rate": 4.358974358974359e-07,
"loss": 1.5509,
"step": 18
},
{
"epoch": 0.014987182015381581,
"grad_norm": 21.11162095506983,
"learning_rate": 4.6153846153846156e-07,
"loss": 1.4628,
"step": 19
},
{
"epoch": 0.015775981068822716,
"grad_norm": 21.317719377330565,
"learning_rate": 4.871794871794871e-07,
"loss": 1.4695,
"step": 20
},
{
"epoch": 0.016564780122263853,
"grad_norm": 18.999593991748867,
"learning_rate": 5.128205128205127e-07,
"loss": 1.372,
"step": 21
},
{
"epoch": 0.01735357917570499,
"grad_norm": 19.842618420654098,
"learning_rate": 5.384615384615384e-07,
"loss": 1.4403,
"step": 22
},
{
"epoch": 0.018142378229146126,
"grad_norm": 20.861166562452222,
"learning_rate": 5.641025641025641e-07,
"loss": 1.4226,
"step": 23
},
{
"epoch": 0.01893117728258726,
"grad_norm": 17.992887570574684,
"learning_rate": 5.897435897435898e-07,
"loss": 1.37,
"step": 24
},
{
"epoch": 0.019719976336028396,
"grad_norm": 19.59244514586342,
"learning_rate": 6.153846153846154e-07,
"loss": 1.46,
"step": 25
},
{
"epoch": 0.020508775389469533,
"grad_norm": 18.423342932899235,
"learning_rate": 6.410256410256411e-07,
"loss": 1.3541,
"step": 26
},
{
"epoch": 0.02129757444291067,
"grad_norm": 15.680095179583612,
"learning_rate": 6.666666666666666e-07,
"loss": 1.1953,
"step": 27
},
{
"epoch": 0.022086373496351803,
"grad_norm": 13.396337783252072,
"learning_rate": 6.923076923076922e-07,
"loss": 1.1452,
"step": 28
},
{
"epoch": 0.02287517254979294,
"grad_norm": 11.902502345243457,
"learning_rate": 7.179487179487179e-07,
"loss": 1.1671,
"step": 29
},
{
"epoch": 0.023663971603234076,
"grad_norm": 13.636560243332156,
"learning_rate": 7.435897435897435e-07,
"loss": 1.1456,
"step": 30
},
{
"epoch": 0.024452770656675213,
"grad_norm": 14.891215734551755,
"learning_rate": 7.692307692307693e-07,
"loss": 1.1441,
"step": 31
},
{
"epoch": 0.02524156971011635,
"grad_norm": 12.235113883103503,
"learning_rate": 7.948717948717948e-07,
"loss": 1.0817,
"step": 32
},
{
"epoch": 0.026030368763557483,
"grad_norm": 12.04586402982212,
"learning_rate": 8.205128205128205e-07,
"loss": 1.0814,
"step": 33
},
{
"epoch": 0.02681916781699862,
"grad_norm": 12.138993173935907,
"learning_rate": 8.461538461538461e-07,
"loss": 1.1061,
"step": 34
},
{
"epoch": 0.027607966870439756,
"grad_norm": 11.319224480934183,
"learning_rate": 8.717948717948718e-07,
"loss": 1.0379,
"step": 35
},
{
"epoch": 0.028396765923880893,
"grad_norm": 10.071336362250555,
"learning_rate": 8.974358974358974e-07,
"loss": 1.0049,
"step": 36
},
{
"epoch": 0.029185564977322026,
"grad_norm": 9.147251230442183,
"learning_rate": 9.230769230769231e-07,
"loss": 0.979,
"step": 37
},
{
"epoch": 0.029974364030763163,
"grad_norm": 4.725229992119413,
"learning_rate": 9.487179487179486e-07,
"loss": 0.9005,
"step": 38
},
{
"epoch": 0.0307631630842043,
"grad_norm": 3.5180292086307428,
"learning_rate": 9.743589743589742e-07,
"loss": 0.8805,
"step": 39
},
{
"epoch": 0.03155196213764543,
"grad_norm": 3.4971220169070945,
"learning_rate": 1e-06,
"loss": 0.8989,
"step": 40
},
{
"epoch": 0.03234076119108657,
"grad_norm": 3.7007713412958876,
"learning_rate": 9.999983664375128e-07,
"loss": 0.9226,
"step": 41
},
{
"epoch": 0.033129560244527706,
"grad_norm": 3.6906743489881775,
"learning_rate": 9.999934657607252e-07,
"loss": 0.8423,
"step": 42
},
{
"epoch": 0.03391835929796884,
"grad_norm": 3.5761718345768823,
"learning_rate": 9.999852980016597e-07,
"loss": 0.8834,
"step": 43
},
{
"epoch": 0.03470715835140998,
"grad_norm": 3.6906833806846606,
"learning_rate": 9.999738632136864e-07,
"loss": 0.8506,
"step": 44
},
{
"epoch": 0.03549595740485111,
"grad_norm": 3.6987940305158227,
"learning_rate": 9.999591614715228e-07,
"loss": 0.8754,
"step": 45
},
{
"epoch": 0.03628475645829225,
"grad_norm": 3.8650558275749884,
"learning_rate": 9.999411928712342e-07,
"loss": 0.8723,
"step": 46
},
{
"epoch": 0.037073555511733386,
"grad_norm": 3.830392093292231,
"learning_rate": 9.999199575302316e-07,
"loss": 0.8644,
"step": 47
},
{
"epoch": 0.03786235456517452,
"grad_norm": 3.6475156383869303,
"learning_rate": 9.998954555872722e-07,
"loss": 0.894,
"step": 48
},
{
"epoch": 0.03865115361861566,
"grad_norm": 3.666971497191224,
"learning_rate": 9.998676872024575e-07,
"loss": 0.853,
"step": 49
},
{
"epoch": 0.03943995267205679,
"grad_norm": 3.5979938218472705,
"learning_rate": 9.998366525572335e-07,
"loss": 0.8658,
"step": 50
},
{
"epoch": 0.04022875172549793,
"grad_norm": 3.591852924414306,
"learning_rate": 9.99802351854388e-07,
"loss": 0.8884,
"step": 51
},
{
"epoch": 0.041017550778939066,
"grad_norm": 3.6241417263328133,
"learning_rate": 9.997647853180507e-07,
"loss": 0.8674,
"step": 52
},
{
"epoch": 0.0418063498323802,
"grad_norm": 3.1548064042839563,
"learning_rate": 9.997239531936903e-07,
"loss": 0.8787,
"step": 53
},
{
"epoch": 0.04259514888582134,
"grad_norm": 3.4861397835145245,
"learning_rate": 9.996798557481146e-07,
"loss": 0.8298,
"step": 54
},
{
"epoch": 0.04338394793926247,
"grad_norm": 3.2809323710163274,
"learning_rate": 9.996324932694668e-07,
"loss": 0.8245,
"step": 55
},
{
"epoch": 0.044172746992703606,
"grad_norm": 3.227332680300041,
"learning_rate": 9.995818660672254e-07,
"loss": 0.8886,
"step": 56
},
{
"epoch": 0.044961546046144746,
"grad_norm": 2.946824332318705,
"learning_rate": 9.995279744722016e-07,
"loss": 0.8586,
"step": 57
},
{
"epoch": 0.04575034509958588,
"grad_norm": 3.26654804883275,
"learning_rate": 9.99470818836536e-07,
"loss": 0.7976,
"step": 58
},
{
"epoch": 0.04653914415302702,
"grad_norm": 3.052602823683985,
"learning_rate": 9.99410399533698e-07,
"loss": 0.8262,
"step": 59
},
{
"epoch": 0.04732794320646815,
"grad_norm": 3.414233581618168,
"learning_rate": 9.993467169584823e-07,
"loss": 0.8227,
"step": 60
},
{
"epoch": 0.048116742259909286,
"grad_norm": 3.0290264902443345,
"learning_rate": 9.99279771527007e-07,
"loss": 0.8092,
"step": 61
},
{
"epoch": 0.048905541313350426,
"grad_norm": 2.9719868450361715,
"learning_rate": 9.992095636767103e-07,
"loss": 0.8386,
"step": 62
},
{
"epoch": 0.04969434036679156,
"grad_norm": 3.0268031749477986,
"learning_rate": 9.991360938663475e-07,
"loss": 0.7875,
"step": 63
},
{
"epoch": 0.0504831394202327,
"grad_norm": 2.798561649830226,
"learning_rate": 9.99059362575989e-07,
"loss": 0.8406,
"step": 64
},
{
"epoch": 0.05127193847367383,
"grad_norm": 2.909622934231715,
"learning_rate": 9.989793703070162e-07,
"loss": 0.7645,
"step": 65
},
{
"epoch": 0.052060737527114966,
"grad_norm": 2.8784035929166327,
"learning_rate": 9.988961175821184e-07,
"loss": 0.7412,
"step": 66
},
{
"epoch": 0.052849536580556106,
"grad_norm": 2.8330445419230292,
"learning_rate": 9.9880960494529e-07,
"loss": 0.762,
"step": 67
},
{
"epoch": 0.05363833563399724,
"grad_norm": 2.595688924804014,
"learning_rate": 9.98719832961826e-07,
"loss": 0.7503,
"step": 68
},
{
"epoch": 0.05442713468743837,
"grad_norm": 2.672864806085718,
"learning_rate": 9.98626802218319e-07,
"loss": 0.7514,
"step": 69
},
{
"epoch": 0.05521593374087951,
"grad_norm": 2.6286672019522084,
"learning_rate": 9.98530513322655e-07,
"loss": 0.7387,
"step": 70
},
{
"epoch": 0.056004732794320645,
"grad_norm": 2.5062295678916833,
"learning_rate": 9.984309669040102e-07,
"loss": 0.7978,
"step": 71
},
{
"epoch": 0.056793531847761786,
"grad_norm": 2.6849708650725197,
"learning_rate": 9.983281636128453e-07,
"loss": 0.7566,
"step": 72
},
{
"epoch": 0.05758233090120292,
"grad_norm": 2.5848147232622543,
"learning_rate": 9.982221041209027e-07,
"loss": 0.7712,
"step": 73
},
{
"epoch": 0.05837112995464405,
"grad_norm": 2.470299683431367,
"learning_rate": 9.981127891212019e-07,
"loss": 0.7549,
"step": 74
},
{
"epoch": 0.05915992900808519,
"grad_norm": 2.3230706720997816,
"learning_rate": 9.980002193280341e-07,
"loss": 0.7411,
"step": 75
},
{
"epoch": 0.059948728061526325,
"grad_norm": 2.4491831201172958,
"learning_rate": 9.97884395476959e-07,
"loss": 0.7311,
"step": 76
},
{
"epoch": 0.06073752711496746,
"grad_norm": 2.3852788380234795,
"learning_rate": 9.97765318324798e-07,
"loss": 0.7631,
"step": 77
},
{
"epoch": 0.0615263261684086,
"grad_norm": 2.098867238321296,
"learning_rate": 9.976429886496311e-07,
"loss": 0.7664,
"step": 78
},
{
"epoch": 0.06231512522184973,
"grad_norm": 2.2951148713524026,
"learning_rate": 9.975174072507915e-07,
"loss": 0.7242,
"step": 79
},
{
"epoch": 0.06310392427529087,
"grad_norm": 2.4936102415302326,
"learning_rate": 9.973885749488588e-07,
"loss": 0.6974,
"step": 80
},
{
"epoch": 0.06389272332873201,
"grad_norm": 2.2420956391392974,
"learning_rate": 9.972564925856556e-07,
"loss": 0.7074,
"step": 81
},
{
"epoch": 0.06468152238217315,
"grad_norm": 2.291007122423151,
"learning_rate": 9.971211610242414e-07,
"loss": 0.7235,
"step": 82
},
{
"epoch": 0.06547032143561428,
"grad_norm": 2.180802966416447,
"learning_rate": 9.969825811489062e-07,
"loss": 0.7233,
"step": 83
},
{
"epoch": 0.06625912048905541,
"grad_norm": 2.2569170025933585,
"learning_rate": 9.968407538651656e-07,
"loss": 0.7082,
"step": 84
},
{
"epoch": 0.06704791954249655,
"grad_norm": 2.5138802108685403,
"learning_rate": 9.966956800997546e-07,
"loss": 0.702,
"step": 85
},
{
"epoch": 0.06783671859593768,
"grad_norm": 2.3328171388578745,
"learning_rate": 9.965473608006212e-07,
"loss": 0.7189,
"step": 86
},
{
"epoch": 0.06862551764937883,
"grad_norm": 2.4351769062343647,
"learning_rate": 9.96395796936921e-07,
"loss": 0.7261,
"step": 87
},
{
"epoch": 0.06941431670281996,
"grad_norm": 2.4418488296075793,
"learning_rate": 9.9624098949901e-07,
"loss": 0.7096,
"step": 88
},
{
"epoch": 0.07020311575626109,
"grad_norm": 2.2484405624096,
"learning_rate": 9.96082939498439e-07,
"loss": 0.757,
"step": 89
},
{
"epoch": 0.07099191480970223,
"grad_norm": 2.5770142969870435,
"learning_rate": 9.959216479679458e-07,
"loss": 0.749,
"step": 90
},
{
"epoch": 0.07178071386314336,
"grad_norm": 2.614664891624914,
"learning_rate": 9.957571159614498e-07,
"loss": 0.7,
"step": 91
},
{
"epoch": 0.0725695129165845,
"grad_norm": 2.503853320846588,
"learning_rate": 9.955893445540442e-07,
"loss": 0.7116,
"step": 92
},
{
"epoch": 0.07335831197002564,
"grad_norm": 2.3161217480939165,
"learning_rate": 9.954183348419893e-07,
"loss": 0.7233,
"step": 93
},
{
"epoch": 0.07414711102346677,
"grad_norm": 2.198515267378206,
"learning_rate": 9.952440879427055e-07,
"loss": 0.7112,
"step": 94
},
{
"epoch": 0.0749359100769079,
"grad_norm": 2.189972541962156,
"learning_rate": 9.950666049947652e-07,
"loss": 0.7284,
"step": 95
},
{
"epoch": 0.07572470913034904,
"grad_norm": 2.4180696847370666,
"learning_rate": 9.948858871578867e-07,
"loss": 0.7305,
"step": 96
},
{
"epoch": 0.07651350818379019,
"grad_norm": 2.2549919126041047,
"learning_rate": 9.947019356129254e-07,
"loss": 0.7175,
"step": 97
},
{
"epoch": 0.07730230723723132,
"grad_norm": 2.209908478823275,
"learning_rate": 9.945147515618666e-07,
"loss": 0.6826,
"step": 98
},
{
"epoch": 0.07809110629067245,
"grad_norm": 2.1457174671628323,
"learning_rate": 9.94324336227818e-07,
"loss": 0.686,
"step": 99
},
{
"epoch": 0.07887990534411359,
"grad_norm": 2.510204276682512,
"learning_rate": 9.941306908550004e-07,
"loss": 0.6794,
"step": 100
},
{
"epoch": 0.07966870439755472,
"grad_norm": 1.9984316369122155,
"learning_rate": 9.939338167087415e-07,
"loss": 0.6667,
"step": 101
},
{
"epoch": 0.08045750345099587,
"grad_norm": 2.105987888350643,
"learning_rate": 9.93733715075466e-07,
"loss": 0.7042,
"step": 102
},
{
"epoch": 0.081246302504437,
"grad_norm": 2.1449968700072426,
"learning_rate": 9.935303872626881e-07,
"loss": 0.7269,
"step": 103
},
{
"epoch": 0.08203510155787813,
"grad_norm": 2.1399376402602903,
"learning_rate": 9.933238345990026e-07,
"loss": 0.6725,
"step": 104
},
{
"epoch": 0.08282390061131926,
"grad_norm": 2.2336511334127436,
"learning_rate": 9.93114058434076e-07,
"loss": 0.7227,
"step": 105
},
{
"epoch": 0.0836126996647604,
"grad_norm": 2.174804460647939,
"learning_rate": 9.929010601386384e-07,
"loss": 0.711,
"step": 106
},
{
"epoch": 0.08440149871820153,
"grad_norm": 2.077733993111799,
"learning_rate": 9.92684841104474e-07,
"loss": 0.6869,
"step": 107
},
{
"epoch": 0.08519029777164268,
"grad_norm": 2.0325410693379986,
"learning_rate": 9.924654027444116e-07,
"loss": 0.7444,
"step": 108
},
{
"epoch": 0.08597909682508381,
"grad_norm": 2.016982086847133,
"learning_rate": 9.922427464923166e-07,
"loss": 0.7033,
"step": 109
},
{
"epoch": 0.08676789587852494,
"grad_norm": 2.0897827665188906,
"learning_rate": 9.920168738030807e-07,
"loss": 0.6791,
"step": 110
},
{
"epoch": 0.08755669493196608,
"grad_norm": 2.2334397912019,
"learning_rate": 9.917877861526122e-07,
"loss": 0.6569,
"step": 111
},
{
"epoch": 0.08834549398540721,
"grad_norm": 2.278926252437753,
"learning_rate": 9.915554850378272e-07,
"loss": 0.6726,
"step": 112
},
{
"epoch": 0.08913429303884836,
"grad_norm": 2.0304786199673996,
"learning_rate": 9.913199719766392e-07,
"loss": 0.7221,
"step": 113
},
{
"epoch": 0.08992309209228949,
"grad_norm": 2.1226845835490757,
"learning_rate": 9.910812485079496e-07,
"loss": 0.7002,
"step": 114
},
{
"epoch": 0.09071189114573062,
"grad_norm": 1.9967984638638425,
"learning_rate": 9.908393161916374e-07,
"loss": 0.6531,
"step": 115
},
{
"epoch": 0.09150069019917176,
"grad_norm": 2.1122809420729114,
"learning_rate": 9.90594176608548e-07,
"loss": 0.6529,
"step": 116
},
{
"epoch": 0.09228948925261289,
"grad_norm": 2.214611967238534,
"learning_rate": 9.903458313604856e-07,
"loss": 0.6932,
"step": 117
},
{
"epoch": 0.09307828830605404,
"grad_norm": 1.939537278621756,
"learning_rate": 9.900942820701994e-07,
"loss": 0.6516,
"step": 118
},
{
"epoch": 0.09386708735949517,
"grad_norm": 2.152403210434206,
"learning_rate": 9.89839530381376e-07,
"loss": 0.6487,
"step": 119
},
{
"epoch": 0.0946558864129363,
"grad_norm": 2.0200287174218263,
"learning_rate": 9.895815779586262e-07,
"loss": 0.6749,
"step": 120
},
{
"epoch": 0.09544468546637744,
"grad_norm": 2.1024029674856135,
"learning_rate": 9.893204264874757e-07,
"loss": 0.6754,
"step": 121
},
{
"epoch": 0.09623348451981857,
"grad_norm": 2.1950582348744407,
"learning_rate": 9.890560776743532e-07,
"loss": 0.6591,
"step": 122
},
{
"epoch": 0.09702228357325972,
"grad_norm": 2.4193442997240027,
"learning_rate": 9.887885332465806e-07,
"loss": 0.6537,
"step": 123
},
{
"epoch": 0.09781108262670085,
"grad_norm": 1.9090500133740858,
"learning_rate": 9.885177949523595e-07,
"loss": 0.6791,
"step": 124
},
{
"epoch": 0.09859988168014198,
"grad_norm": 2.2655596429804836,
"learning_rate": 9.882438645607619e-07,
"loss": 0.6285,
"step": 125
},
{
"epoch": 0.09938868073358312,
"grad_norm": 2.3955445313932753,
"learning_rate": 9.879667438617172e-07,
"loss": 0.6668,
"step": 126
},
{
"epoch": 0.10017747978702425,
"grad_norm": 2.228073273296558,
"learning_rate": 9.876864346660013e-07,
"loss": 0.6693,
"step": 127
},
{
"epoch": 0.1009662788404654,
"grad_norm": 2.2233434521309445,
"learning_rate": 9.874029388052248e-07,
"loss": 0.6761,
"step": 128
},
{
"epoch": 0.10175507789390653,
"grad_norm": 2.1780763478003538,
"learning_rate": 9.871162581318205e-07,
"loss": 0.6614,
"step": 129
},
{
"epoch": 0.10254387694734766,
"grad_norm": 2.174738822458843,
"learning_rate": 9.868263945190311e-07,
"loss": 0.6418,
"step": 130
},
{
"epoch": 0.1033326760007888,
"grad_norm": 1.9101621674089266,
"learning_rate": 9.865333498608986e-07,
"loss": 0.7165,
"step": 131
},
{
"epoch": 0.10412147505422993,
"grad_norm": 2.1602769420975974,
"learning_rate": 9.862371260722496e-07,
"loss": 0.6601,
"step": 132
},
{
"epoch": 0.10491027410767106,
"grad_norm": 2.0662885243604276,
"learning_rate": 9.859377250886842e-07,
"loss": 0.6295,
"step": 133
},
{
"epoch": 0.10569907316111221,
"grad_norm": 2.167979726250443,
"learning_rate": 9.856351488665638e-07,
"loss": 0.6565,
"step": 134
},
{
"epoch": 0.10648787221455334,
"grad_norm": 2.4135140669881743,
"learning_rate": 9.853293993829967e-07,
"loss": 0.7407,
"step": 135
},
{
"epoch": 0.10727667126799448,
"grad_norm": 2.1348027174931112,
"learning_rate": 9.850204786358266e-07,
"loss": 0.6429,
"step": 136
},
{
"epoch": 0.10806547032143561,
"grad_norm": 2.3158913329772126,
"learning_rate": 9.847083886436188e-07,
"loss": 0.6374,
"step": 137
},
{
"epoch": 0.10885426937487674,
"grad_norm": 2.221761813749553,
"learning_rate": 9.843931314456473e-07,
"loss": 0.6624,
"step": 138
},
{
"epoch": 0.10964306842831789,
"grad_norm": 2.1424484619812554,
"learning_rate": 9.840747091018814e-07,
"loss": 0.6464,
"step": 139
},
{
"epoch": 0.11043186748175902,
"grad_norm": 2.053523672023974,
"learning_rate": 9.837531236929726e-07,
"loss": 0.6145,
"step": 140
},
{
"epoch": 0.11122066653520016,
"grad_norm": 2.1103531295770748,
"learning_rate": 9.8342837732024e-07,
"loss": 0.6454,
"step": 141
},
{
"epoch": 0.11200946558864129,
"grad_norm": 2.2505689024311075,
"learning_rate": 9.831004721056576e-07,
"loss": 0.6695,
"step": 142
},
{
"epoch": 0.11279826464208242,
"grad_norm": 2.137541688443829,
"learning_rate": 9.8276941019184e-07,
"loss": 0.6279,
"step": 143
},
{
"epoch": 0.11358706369552357,
"grad_norm": 1.9325471763337814,
"learning_rate": 9.824351937420286e-07,
"loss": 0.6617,
"step": 144
},
{
"epoch": 0.1143758627489647,
"grad_norm": 2.343364335607791,
"learning_rate": 9.820978249400772e-07,
"loss": 0.6209,
"step": 145
},
{
"epoch": 0.11516466180240584,
"grad_norm": 2.073265445103016,
"learning_rate": 9.81757305990438e-07,
"loss": 0.6544,
"step": 146
},
{
"epoch": 0.11595346085584697,
"grad_norm": 2.1345678706401885,
"learning_rate": 9.814136391181468e-07,
"loss": 0.6392,
"step": 147
},
{
"epoch": 0.1167422599092881,
"grad_norm": 2.045889999752981,
"learning_rate": 9.810668265688088e-07,
"loss": 0.6356,
"step": 148
},
{
"epoch": 0.11753105896272925,
"grad_norm": 2.203598159559126,
"learning_rate": 9.80716870608584e-07,
"loss": 0.6354,
"step": 149
},
{
"epoch": 0.11831985801617038,
"grad_norm": 1.8693440166545279,
"learning_rate": 9.803637735241719e-07,
"loss": 0.6338,
"step": 150
},
{
"epoch": 0.11910865706961152,
"grad_norm": 2.064579307573498,
"learning_rate": 9.800075376227973e-07,
"loss": 0.6689,
"step": 151
},
{
"epoch": 0.11989745612305265,
"grad_norm": 2.080108515704008,
"learning_rate": 9.796481652321946e-07,
"loss": 0.6176,
"step": 152
},
{
"epoch": 0.12068625517649378,
"grad_norm": 2.200233582877887,
"learning_rate": 9.79285658700593e-07,
"loss": 0.6486,
"step": 153
},
{
"epoch": 0.12147505422993492,
"grad_norm": 2.326480348621232,
"learning_rate": 9.789200203967003e-07,
"loss": 0.6172,
"step": 154
},
{
"epoch": 0.12226385328337606,
"grad_norm": 2.053859826133946,
"learning_rate": 9.78551252709689e-07,
"loss": 0.6119,
"step": 155
},
{
"epoch": 0.1230526523368172,
"grad_norm": 2.4125543948599417,
"learning_rate": 9.78179358049179e-07,
"loss": 0.6243,
"step": 156
},
{
"epoch": 0.12384145139025833,
"grad_norm": 2.1531753469530956,
"learning_rate": 9.778043388452233e-07,
"loss": 0.61,
"step": 157
},
{
"epoch": 0.12463025044369946,
"grad_norm": 2.0602091846668364,
"learning_rate": 9.77426197548291e-07,
"loss": 0.6143,
"step": 158
},
{
"epoch": 0.1254190494971406,
"grad_norm": 2.3040349476060418,
"learning_rate": 9.77044936629252e-07,
"loss": 0.6417,
"step": 159
},
{
"epoch": 0.12620784855058173,
"grad_norm": 2.136747041605826,
"learning_rate": 9.7666055857936e-07,
"loss": 0.6333,
"step": 160
},
{
"epoch": 0.12699664760402288,
"grad_norm": 2.0849191714265363,
"learning_rate": 9.762730659102376e-07,
"loss": 0.6339,
"step": 161
},
{
"epoch": 0.12778544665746402,
"grad_norm": 1.9460159578188034,
"learning_rate": 9.758824611538585e-07,
"loss": 0.6637,
"step": 162
},
{
"epoch": 0.12857424571090514,
"grad_norm": 2.1764142009289267,
"learning_rate": 9.754887468625322e-07,
"loss": 0.5856,
"step": 163
},
{
"epoch": 0.1293630447643463,
"grad_norm": 2.0161165289961747,
"learning_rate": 9.75091925608886e-07,
"loss": 0.633,
"step": 164
},
{
"epoch": 0.1301518438177874,
"grad_norm": 1.9549304626583646,
"learning_rate": 9.746919999858492e-07,
"loss": 0.6293,
"step": 165
},
{
"epoch": 0.13094064287122856,
"grad_norm": 2.0059985272816525,
"learning_rate": 9.742889726066358e-07,
"loss": 0.6361,
"step": 166
},
{
"epoch": 0.13172944192466968,
"grad_norm": 2.108068432019833,
"learning_rate": 9.738828461047275e-07,
"loss": 0.6179,
"step": 167
},
{
"epoch": 0.13251824097811082,
"grad_norm": 2.276091751963883,
"learning_rate": 9.734736231338562e-07,
"loss": 0.6155,
"step": 168
},
{
"epoch": 0.13330704003155197,
"grad_norm": 2.218148255025018,
"learning_rate": 9.730613063679874e-07,
"loss": 0.6324,
"step": 169
},
{
"epoch": 0.1340958390849931,
"grad_norm": 2.225906331496186,
"learning_rate": 9.726458985013017e-07,
"loss": 0.6042,
"step": 170
},
{
"epoch": 0.13488463813843424,
"grad_norm": 2.3092785891839145,
"learning_rate": 9.722274022481778e-07,
"loss": 0.6582,
"step": 171
},
{
"epoch": 0.13567343719187536,
"grad_norm": 1.9357661240497763,
"learning_rate": 9.718058203431749e-07,
"loss": 0.606,
"step": 172
},
{
"epoch": 0.1364622362453165,
"grad_norm": 2.1911258272030407,
"learning_rate": 9.713811555410146e-07,
"loss": 0.6218,
"step": 173
},
{
"epoch": 0.13725103529875765,
"grad_norm": 2.0961500104898336,
"learning_rate": 9.70953410616563e-07,
"loss": 0.6056,
"step": 174
},
{
"epoch": 0.13803983435219877,
"grad_norm": 2.012463262930148,
"learning_rate": 9.70522588364812e-07,
"loss": 0.6225,
"step": 175
},
{
"epoch": 0.13882863340563992,
"grad_norm": 2.071026669357088,
"learning_rate": 9.70088691600862e-07,
"loss": 0.566,
"step": 176
},
{
"epoch": 0.13961743245908104,
"grad_norm": 2.0696818714622305,
"learning_rate": 9.696517231599032e-07,
"loss": 0.6042,
"step": 177
},
{
"epoch": 0.14040623151252218,
"grad_norm": 2.0453998756275897,
"learning_rate": 9.692116858971965e-07,
"loss": 0.6138,
"step": 178
},
{
"epoch": 0.14119503056596333,
"grad_norm": 1.9031178086136131,
"learning_rate": 9.687685826880551e-07,
"loss": 0.6159,
"step": 179
},
{
"epoch": 0.14198382961940445,
"grad_norm": 1.839058059178829,
"learning_rate": 9.683224164278263e-07,
"loss": 0.5902,
"step": 180
},
{
"epoch": 0.1427726286728456,
"grad_norm": 1.8634457748819728,
"learning_rate": 9.67873190031872e-07,
"loss": 0.6406,
"step": 181
},
{
"epoch": 0.14356142772628672,
"grad_norm": 2.3423429752571647,
"learning_rate": 9.674209064355497e-07,
"loss": 0.6197,
"step": 182
},
{
"epoch": 0.14435022677972786,
"grad_norm": 2.2209903723096964,
"learning_rate": 9.669655685941932e-07,
"loss": 0.6262,
"step": 183
},
{
"epoch": 0.145139025833169,
"grad_norm": 1.979200832158828,
"learning_rate": 9.665071794830943e-07,
"loss": 0.6283,
"step": 184
},
{
"epoch": 0.14592782488661013,
"grad_norm": 2.0913291773802762,
"learning_rate": 9.660457420974817e-07,
"loss": 0.5824,
"step": 185
},
{
"epoch": 0.14671662394005128,
"grad_norm": 1.9785722976236968,
"learning_rate": 9.655812594525028e-07,
"loss": 0.6433,
"step": 186
},
{
"epoch": 0.1475054229934924,
"grad_norm": 2.0901254186127933,
"learning_rate": 9.65113734583203e-07,
"loss": 0.596,
"step": 187
},
{
"epoch": 0.14829422204693354,
"grad_norm": 2.1133703700117867,
"learning_rate": 9.646431705445068e-07,
"loss": 0.6049,
"step": 188
},
{
"epoch": 0.1490830211003747,
"grad_norm": 2.011797123245168,
"learning_rate": 9.641695704111974e-07,
"loss": 0.5874,
"step": 189
},
{
"epoch": 0.1498718201538158,
"grad_norm": 2.050447388943759,
"learning_rate": 9.636929372778961e-07,
"loss": 0.6301,
"step": 190
},
{
"epoch": 0.15066061920725696,
"grad_norm": 1.9721198253109251,
"learning_rate": 9.632132742590435e-07,
"loss": 0.568,
"step": 191
},
{
"epoch": 0.15144941826069808,
"grad_norm": 2.0051708806154656,
"learning_rate": 9.62730584488877e-07,
"loss": 0.5872,
"step": 192
},
{
"epoch": 0.15223821731413922,
"grad_norm": 2.001494879776667,
"learning_rate": 9.622448711214125e-07,
"loss": 0.5979,
"step": 193
},
{
"epoch": 0.15302701636758037,
"grad_norm": 2.0369341853504377,
"learning_rate": 9.617561373304228e-07,
"loss": 0.5947,
"step": 194
},
{
"epoch": 0.1538158154210215,
"grad_norm": 2.1189387904824186,
"learning_rate": 9.612643863094163e-07,
"loss": 0.5747,
"step": 195
},
{
"epoch": 0.15460461447446264,
"grad_norm": 2.0341452382533496,
"learning_rate": 9.60769621271617e-07,
"loss": 0.6284,
"step": 196
},
{
"epoch": 0.15539341352790376,
"grad_norm": 2.076946897291527,
"learning_rate": 9.602718454499439e-07,
"loss": 0.5979,
"step": 197
},
{
"epoch": 0.1561822125813449,
"grad_norm": 2.1713436984821444,
"learning_rate": 9.59771062096988e-07,
"loss": 0.5825,
"step": 198
},
{
"epoch": 0.15697101163478605,
"grad_norm": 2.0416085105909185,
"learning_rate": 9.59267274484993e-07,
"loss": 0.5789,
"step": 199
},
{
"epoch": 0.15775981068822717,
"grad_norm": 2.2020227481283037,
"learning_rate": 9.587604859058333e-07,
"loss": 0.6093,
"step": 200
},
{
"epoch": 0.15854860974166832,
"grad_norm": 1.8620107460314095,
"learning_rate": 9.582506996709922e-07,
"loss": 0.6052,
"step": 201
},
{
"epoch": 0.15933740879510944,
"grad_norm": 2.1199908991427265,
"learning_rate": 9.5773791911154e-07,
"loss": 0.5797,
"step": 202
},
{
"epoch": 0.16012620784855058,
"grad_norm": 2.2041152162666386,
"learning_rate": 9.572221475781136e-07,
"loss": 0.5836,
"step": 203
},
{
"epoch": 0.16091500690199173,
"grad_norm": 2.066417536876525,
"learning_rate": 9.567033884408926e-07,
"loss": 0.6058,
"step": 204
},
{
"epoch": 0.16170380595543285,
"grad_norm": 2.09224656811423,
"learning_rate": 9.561816450895793e-07,
"loss": 0.5529,
"step": 205
},
{
"epoch": 0.162492605008874,
"grad_norm": 2.2665061926766237,
"learning_rate": 9.556569209333746e-07,
"loss": 0.6086,
"step": 206
},
{
"epoch": 0.16328140406231512,
"grad_norm": 2.135447318507951,
"learning_rate": 9.551292194009578e-07,
"loss": 0.5878,
"step": 207
},
{
"epoch": 0.16407020311575626,
"grad_norm": 2.0380476480250893,
"learning_rate": 9.545985439404624e-07,
"loss": 0.5801,
"step": 208
},
{
"epoch": 0.1648590021691974,
"grad_norm": 1.8888667270876551,
"learning_rate": 9.540648980194547e-07,
"loss": 0.6513,
"step": 209
},
{
"epoch": 0.16564780122263853,
"grad_norm": 2.0898190110034176,
"learning_rate": 9.535282851249102e-07,
"loss": 0.599,
"step": 210
},
{
"epoch": 0.16643660027607968,
"grad_norm": 1.9475094517211673,
"learning_rate": 9.529887087631921e-07,
"loss": 0.5584,
"step": 211
},
{
"epoch": 0.1672253993295208,
"grad_norm": 2.1978457276532457,
"learning_rate": 9.524461724600267e-07,
"loss": 0.6047,
"step": 212
},
{
"epoch": 0.16801419838296194,
"grad_norm": 1.8386164989261797,
"learning_rate": 9.519006797604822e-07,
"loss": 0.5844,
"step": 213
},
{
"epoch": 0.16880299743640306,
"grad_norm": 2.0055899446868954,
"learning_rate": 9.513522342289442e-07,
"loss": 0.6342,
"step": 214
},
{
"epoch": 0.1695917964898442,
"grad_norm": 2.1407367144122404,
"learning_rate": 9.508008394490925e-07,
"loss": 0.5645,
"step": 215
},
{
"epoch": 0.17038059554328536,
"grad_norm": 2.421295934532421,
"learning_rate": 9.50246499023879e-07,
"loss": 0.6355,
"step": 216
},
{
"epoch": 0.17116939459672648,
"grad_norm": 2.081286585057276,
"learning_rate": 9.496892165755023e-07,
"loss": 0.6047,
"step": 217
},
{
"epoch": 0.17195819365016762,
"grad_norm": 2.1273950875255814,
"learning_rate": 9.491289957453852e-07,
"loss": 0.5984,
"step": 218
},
{
"epoch": 0.17274699270360874,
"grad_norm": 2.019906679301822,
"learning_rate": 9.485658401941505e-07,
"loss": 0.625,
"step": 219
},
{
"epoch": 0.1735357917570499,
"grad_norm": 2.162787736997155,
"learning_rate": 9.479997536015977e-07,
"loss": 0.54,
"step": 220
},
{
"epoch": 0.17432459081049104,
"grad_norm": 2.0124485157388863,
"learning_rate": 9.474307396666778e-07,
"loss": 0.6159,
"step": 221
},
{
"epoch": 0.17511338986393216,
"grad_norm": 2.1605811090535116,
"learning_rate": 9.4685880210747e-07,
"loss": 0.6195,
"step": 222
},
{
"epoch": 0.1759021889173733,
"grad_norm": 1.9678985366781254,
"learning_rate": 9.462839446611576e-07,
"loss": 0.5915,
"step": 223
},
{
"epoch": 0.17669098797081442,
"grad_norm": 2.0586937345380796,
"learning_rate": 9.457061710840026e-07,
"loss": 0.6185,
"step": 224
},
{
"epoch": 0.17747978702425557,
"grad_norm": 2.1592735957630995,
"learning_rate": 9.451254851513221e-07,
"loss": 0.5602,
"step": 225
},
{
"epoch": 0.17826858607769672,
"grad_norm": 2.2033520089530465,
"learning_rate": 9.445418906574629e-07,
"loss": 0.5882,
"step": 226
},
{
"epoch": 0.17905738513113784,
"grad_norm": 2.2573660763819285,
"learning_rate": 9.439553914157776e-07,
"loss": 0.5592,
"step": 227
},
{
"epoch": 0.17984618418457898,
"grad_norm": 2.2085388925200036,
"learning_rate": 9.433659912585985e-07,
"loss": 0.5819,
"step": 228
},
{
"epoch": 0.1806349832380201,
"grad_norm": 1.9817363710744105,
"learning_rate": 9.427736940372139e-07,
"loss": 0.5807,
"step": 229
},
{
"epoch": 0.18142378229146125,
"grad_norm": 2.2676602550012213,
"learning_rate": 9.421785036218417e-07,
"loss": 0.6152,
"step": 230
},
{
"epoch": 0.1822125813449024,
"grad_norm": 1.9720782617824928,
"learning_rate": 9.415804239016048e-07,
"loss": 0.5854,
"step": 231
},
{
"epoch": 0.18300138039834352,
"grad_norm": 2.1368660188566166,
"learning_rate": 9.409794587845057e-07,
"loss": 0.609,
"step": 232
},
{
"epoch": 0.18379017945178466,
"grad_norm": 1.9839106576459613,
"learning_rate": 9.403756121974005e-07,
"loss": 0.5725,
"step": 233
},
{
"epoch": 0.18457897850522578,
"grad_norm": 1.9218960570299226,
"learning_rate": 9.397688880859739e-07,
"loss": 0.553,
"step": 234
},
{
"epoch": 0.18536777755866693,
"grad_norm": 2.2952033773587797,
"learning_rate": 9.39159290414713e-07,
"loss": 0.593,
"step": 235
},
{
"epoch": 0.18615657661210808,
"grad_norm": 2.240772612128687,
"learning_rate": 9.38546823166881e-07,
"loss": 0.5718,
"step": 236
},
{
"epoch": 0.1869453756655492,
"grad_norm": 1.8742465867882945,
"learning_rate": 9.379314903444923e-07,
"loss": 0.5556,
"step": 237
},
{
"epoch": 0.18773417471899034,
"grad_norm": 2.0380610417295664,
"learning_rate": 9.373132959682852e-07,
"loss": 0.5709,
"step": 238
},
{
"epoch": 0.18852297377243146,
"grad_norm": 2.0187775917117285,
"learning_rate": 9.366922440776963e-07,
"loss": 0.5503,
"step": 239
},
{
"epoch": 0.1893117728258726,
"grad_norm": 2.0864752168958653,
"learning_rate": 9.360683387308339e-07,
"loss": 0.5821,
"step": 240
},
{
"epoch": 0.19010057187931376,
"grad_norm": 2.275474842760395,
"learning_rate": 9.354415840044516e-07,
"loss": 0.5466,
"step": 241
},
{
"epoch": 0.19088937093275488,
"grad_norm": 2.112984861051366,
"learning_rate": 9.348119839939213e-07,
"loss": 0.5175,
"step": 242
},
{
"epoch": 0.19167816998619602,
"grad_norm": 2.0724999164336904,
"learning_rate": 9.341795428132067e-07,
"loss": 0.5765,
"step": 243
},
{
"epoch": 0.19246696903963714,
"grad_norm": 1.9479549486132637,
"learning_rate": 9.335442645948368e-07,
"loss": 0.5997,
"step": 244
},
{
"epoch": 0.1932557680930783,
"grad_norm": 1.8966629223438753,
"learning_rate": 9.329061534898782e-07,
"loss": 0.6365,
"step": 245
},
{
"epoch": 0.19404456714651944,
"grad_norm": 2.162908805351454,
"learning_rate": 9.322652136679083e-07,
"loss": 0.5674,
"step": 246
},
{
"epoch": 0.19483336619996056,
"grad_norm": 2.0700268976725997,
"learning_rate": 9.316214493169883e-07,
"loss": 0.543,
"step": 247
},
{
"epoch": 0.1956221652534017,
"grad_norm": 2.2169701057415647,
"learning_rate": 9.309748646436351e-07,
"loss": 0.5733,
"step": 248
},
{
"epoch": 0.19641096430684282,
"grad_norm": 1.9739674469349247,
"learning_rate": 9.303254638727947e-07,
"loss": 0.6047,
"step": 249
},
{
"epoch": 0.19719976336028397,
"grad_norm": 2.1000703648812533,
"learning_rate": 9.296732512478138e-07,
"loss": 0.5622,
"step": 250
},
{
"epoch": 0.19798856241372512,
"grad_norm": 2.16881839362488,
"learning_rate": 9.290182310304131e-07,
"loss": 0.5565,
"step": 251
},
{
"epoch": 0.19877736146716624,
"grad_norm": 1.9966901442546032,
"learning_rate": 9.283604075006583e-07,
"loss": 0.576,
"step": 252
},
{
"epoch": 0.19956616052060738,
"grad_norm": 1.9879167946673124,
"learning_rate": 9.276997849569328e-07,
"loss": 0.5673,
"step": 253
},
{
"epoch": 0.2003549595740485,
"grad_norm": 1.9955103115001214,
"learning_rate": 9.270363677159093e-07,
"loss": 0.5627,
"step": 254
},
{
"epoch": 0.20114375862748965,
"grad_norm": 2.1282616301393245,
"learning_rate": 9.263701601125218e-07,
"loss": 0.5448,
"step": 255
},
{
"epoch": 0.2019325576809308,
"grad_norm": 2.087995505337521,
"learning_rate": 9.257011664999376e-07,
"loss": 0.5913,
"step": 256
},
{
"epoch": 0.20272135673437192,
"grad_norm": 2.169710438721766,
"learning_rate": 9.25029391249528e-07,
"loss": 0.608,
"step": 257
},
{
"epoch": 0.20351015578781306,
"grad_norm": 1.9883185864553177,
"learning_rate": 9.243548387508401e-07,
"loss": 0.5677,
"step": 258
},
{
"epoch": 0.20429895484125418,
"grad_norm": 2.2152613767820415,
"learning_rate": 9.236775134115691e-07,
"loss": 0.5059,
"step": 259
},
{
"epoch": 0.20508775389469533,
"grad_norm": 2.021093406588923,
"learning_rate": 9.229974196575279e-07,
"loss": 0.5606,
"step": 260
},
{
"epoch": 0.20587655294813645,
"grad_norm": 1.9667948025951736,
"learning_rate": 9.223145619326187e-07,
"loss": 0.5594,
"step": 261
},
{
"epoch": 0.2066653520015776,
"grad_norm": 2.05579373863808,
"learning_rate": 9.216289446988049e-07,
"loss": 0.5971,
"step": 262
},
{
"epoch": 0.20745415105501874,
"grad_norm": 2.238835948033033,
"learning_rate": 9.209405724360808e-07,
"loss": 0.5432,
"step": 263
},
{
"epoch": 0.20824295010845986,
"grad_norm": 2.026380742640508,
"learning_rate": 9.202494496424428e-07,
"loss": 0.5695,
"step": 264
},
{
"epoch": 0.209031749161901,
"grad_norm": 2.0852327303250444,
"learning_rate": 9.195555808338602e-07,
"loss": 0.5683,
"step": 265
},
{
"epoch": 0.20982054821534213,
"grad_norm": 2.170259217681539,
"learning_rate": 9.188589705442449e-07,
"loss": 0.5719,
"step": 266
},
{
"epoch": 0.21060934726878328,
"grad_norm": 2.1583648146609975,
"learning_rate": 9.181596233254228e-07,
"loss": 0.5809,
"step": 267
},
{
"epoch": 0.21139814632222442,
"grad_norm": 1.9450779880151452,
"learning_rate": 9.174575437471032e-07,
"loss": 0.5888,
"step": 268
},
{
"epoch": 0.21218694537566554,
"grad_norm": 2.358571439161948,
"learning_rate": 9.1675273639685e-07,
"loss": 0.5724,
"step": 269
},
{
"epoch": 0.2129757444291067,
"grad_norm": 1.8722381634859504,
"learning_rate": 9.160452058800503e-07,
"loss": 0.5616,
"step": 270
},
{
"epoch": 0.2137645434825478,
"grad_norm": 1.9167573026603648,
"learning_rate": 9.153349568198853e-07,
"loss": 0.6202,
"step": 271
},
{
"epoch": 0.21455334253598896,
"grad_norm": 1.811701544893456,
"learning_rate": 9.146219938573001e-07,
"loss": 0.5928,
"step": 272
},
{
"epoch": 0.2153421415894301,
"grad_norm": 2.263992733654506,
"learning_rate": 9.139063216509726e-07,
"loss": 0.5622,
"step": 273
},
{
"epoch": 0.21613094064287122,
"grad_norm": 2.1549415505031275,
"learning_rate": 9.131879448772841e-07,
"loss": 0.5482,
"step": 274
},
{
"epoch": 0.21691973969631237,
"grad_norm": 2.135599620753438,
"learning_rate": 9.124668682302881e-07,
"loss": 0.5398,
"step": 275
},
{
"epoch": 0.2177085387497535,
"grad_norm": 2.132871222551695,
"learning_rate": 9.117430964216794e-07,
"loss": 0.5868,
"step": 276
},
{
"epoch": 0.21849733780319464,
"grad_norm": 2.510292056404663,
"learning_rate": 9.11016634180764e-07,
"loss": 0.5629,
"step": 277
},
{
"epoch": 0.21928613685663578,
"grad_norm": 2.37027410248542,
"learning_rate": 9.102874862544276e-07,
"loss": 0.5468,
"step": 278
},
{
"epoch": 0.2200749359100769,
"grad_norm": 1.995125601866521,
"learning_rate": 9.095556574071054e-07,
"loss": 0.5513,
"step": 279
},
{
"epoch": 0.22086373496351805,
"grad_norm": 1.9697880861704435,
"learning_rate": 9.088211524207496e-07,
"loss": 0.5366,
"step": 280
},
{
"epoch": 0.22165253401695917,
"grad_norm": 2.2020676382640407,
"learning_rate": 9.080839760947996e-07,
"loss": 0.5917,
"step": 281
},
{
"epoch": 0.22244133307040032,
"grad_norm": 2.3253880108174707,
"learning_rate": 9.073441332461498e-07,
"loss": 0.6054,
"step": 282
},
{
"epoch": 0.22323013212384146,
"grad_norm": 1.964461829667778,
"learning_rate": 9.066016287091182e-07,
"loss": 0.553,
"step": 283
},
{
"epoch": 0.22401893117728258,
"grad_norm": 2.0015650257376003,
"learning_rate": 9.058564673354149e-07,
"loss": 0.5491,
"step": 284
},
{
"epoch": 0.22480773023072373,
"grad_norm": 2.0150857445191277,
"learning_rate": 9.051086539941108e-07,
"loss": 0.6178,
"step": 285
},
{
"epoch": 0.22559652928416485,
"grad_norm": 2.324842566561679,
"learning_rate": 9.04358193571605e-07,
"loss": 0.6029,
"step": 286
},
{
"epoch": 0.226385328337606,
"grad_norm": 1.9921598865569665,
"learning_rate": 9.036050909715935e-07,
"loss": 0.5282,
"step": 287
},
{
"epoch": 0.22717412739104714,
"grad_norm": 2.0975976042395192,
"learning_rate": 9.028493511150373e-07,
"loss": 0.5143,
"step": 288
},
{
"epoch": 0.22796292644448826,
"grad_norm": 2.2483468437430676,
"learning_rate": 9.02090978940129e-07,
"loss": 0.5545,
"step": 289
},
{
"epoch": 0.2287517254979294,
"grad_norm": 2.0996302946685304,
"learning_rate": 9.01329979402262e-07,
"loss": 0.5517,
"step": 290
},
{
"epoch": 0.22954052455137053,
"grad_norm": 2.237405783761653,
"learning_rate": 9.005663574739979e-07,
"loss": 0.5535,
"step": 291
},
{
"epoch": 0.23032932360481168,
"grad_norm": 2.1351044599590407,
"learning_rate": 8.99800118145033e-07,
"loss": 0.5912,
"step": 292
},
{
"epoch": 0.23111812265825282,
"grad_norm": 2.1385978871433564,
"learning_rate": 8.990312664221664e-07,
"loss": 0.571,
"step": 293
},
{
"epoch": 0.23190692171169394,
"grad_norm": 2.124644380641703,
"learning_rate": 8.982598073292678e-07,
"loss": 0.533,
"step": 294
},
{
"epoch": 0.2326957207651351,
"grad_norm": 1.9509847211579,
"learning_rate": 8.974857459072434e-07,
"loss": 0.515,
"step": 295
},
{
"epoch": 0.2334845198185762,
"grad_norm": 2.3447003487272435,
"learning_rate": 8.967090872140042e-07,
"loss": 0.5368,
"step": 296
},
{
"epoch": 0.23427331887201736,
"grad_norm": 2.089543552431374,
"learning_rate": 8.959298363244323e-07,
"loss": 0.5694,
"step": 297
},
{
"epoch": 0.2350621179254585,
"grad_norm": 2.161254634235908,
"learning_rate": 8.951479983303476e-07,
"loss": 0.5358,
"step": 298
},
{
"epoch": 0.23585091697889962,
"grad_norm": 2.0718558121787067,
"learning_rate": 8.94363578340475e-07,
"loss": 0.5535,
"step": 299
},
{
"epoch": 0.23663971603234077,
"grad_norm": 1.9931339921010534,
"learning_rate": 8.93576581480411e-07,
"loss": 0.5707,
"step": 300
},
{
"epoch": 0.2374285150857819,
"grad_norm": 2.127655983023035,
"learning_rate": 8.927870128925894e-07,
"loss": 0.5476,
"step": 301
},
{
"epoch": 0.23821731413922304,
"grad_norm": 2.3887597037144257,
"learning_rate": 8.919948777362491e-07,
"loss": 0.5286,
"step": 302
},
{
"epoch": 0.23900611319266418,
"grad_norm": 2.0677142024423616,
"learning_rate": 8.91200181187399e-07,
"loss": 0.5313,
"step": 303
},
{
"epoch": 0.2397949122461053,
"grad_norm": 2.1476513653574316,
"learning_rate": 8.90402928438785e-07,
"loss": 0.5622,
"step": 304
},
{
"epoch": 0.24058371129954645,
"grad_norm": 2.100290685464606,
"learning_rate": 8.896031246998558e-07,
"loss": 0.5804,
"step": 305
},
{
"epoch": 0.24137251035298757,
"grad_norm": 2.3605551818435004,
"learning_rate": 8.88800775196729e-07,
"loss": 0.5666,
"step": 306
},
{
"epoch": 0.24216130940642872,
"grad_norm": 2.3547469827457608,
"learning_rate": 8.879958851721567e-07,
"loss": 0.5479,
"step": 307
},
{
"epoch": 0.24295010845986983,
"grad_norm": 2.409010607379717,
"learning_rate": 8.871884598854918e-07,
"loss": 0.567,
"step": 308
},
{
"epoch": 0.24373890751331098,
"grad_norm": 2.0250115427595015,
"learning_rate": 8.863785046126526e-07,
"loss": 0.5459,
"step": 309
},
{
"epoch": 0.24452770656675213,
"grad_norm": 2.064856701439233,
"learning_rate": 8.855660246460895e-07,
"loss": 0.5383,
"step": 310
},
{
"epoch": 0.24531650562019325,
"grad_norm": 2.2720249602537357,
"learning_rate": 8.847510252947495e-07,
"loss": 0.602,
"step": 311
},
{
"epoch": 0.2461053046736344,
"grad_norm": 2.1676155884577923,
"learning_rate": 8.839335118840422e-07,
"loss": 0.5719,
"step": 312
},
{
"epoch": 0.24689410372707551,
"grad_norm": 2.0541670265949885,
"learning_rate": 8.831134897558046e-07,
"loss": 0.5659,
"step": 313
},
{
"epoch": 0.24768290278051666,
"grad_norm": 2.1379939152412666,
"learning_rate": 8.822909642682661e-07,
"loss": 0.5231,
"step": 314
},
{
"epoch": 0.2484717018339578,
"grad_norm": 1.947062284975293,
"learning_rate": 8.81465940796014e-07,
"loss": 0.5865,
"step": 315
},
{
"epoch": 0.24926050088739893,
"grad_norm": 2.160808065778857,
"learning_rate": 8.806384247299577e-07,
"loss": 0.5734,
"step": 316
},
{
"epoch": 0.2500492999408401,
"grad_norm": 2.0530656719584823,
"learning_rate": 8.798084214772941e-07,
"loss": 0.556,
"step": 317
},
{
"epoch": 0.2508380989942812,
"grad_norm": 1.826990861156769,
"learning_rate": 8.789759364614719e-07,
"loss": 0.5757,
"step": 318
},
{
"epoch": 0.25162689804772237,
"grad_norm": 2.202812516463427,
"learning_rate": 8.781409751221564e-07,
"loss": 0.5385,
"step": 319
},
{
"epoch": 0.25241569710116346,
"grad_norm": 2.056249012032673,
"learning_rate": 8.773035429151936e-07,
"loss": 0.5474,
"step": 320
},
{
"epoch": 0.2532044961546046,
"grad_norm": 2.037573756285196,
"learning_rate": 8.764636453125748e-07,
"loss": 0.5668,
"step": 321
},
{
"epoch": 0.25399329520804576,
"grad_norm": 2.1371621596440584,
"learning_rate": 8.756212878024008e-07,
"loss": 0.5765,
"step": 322
},
{
"epoch": 0.2547820942614869,
"grad_norm": 2.15937091982221,
"learning_rate": 8.747764758888464e-07,
"loss": 0.5509,
"step": 323
},
{
"epoch": 0.25557089331492805,
"grad_norm": 2.2092422475411198,
"learning_rate": 8.739292150921235e-07,
"loss": 0.5059,
"step": 324
},
{
"epoch": 0.25635969236836914,
"grad_norm": 2.1025838723377053,
"learning_rate": 8.73079510948446e-07,
"loss": 0.5742,
"step": 325
},
{
"epoch": 0.2571484914218103,
"grad_norm": 2.1055903896358563,
"learning_rate": 8.722273690099932e-07,
"loss": 0.5364,
"step": 326
},
{
"epoch": 0.25793729047525144,
"grad_norm": 2.0481410401242366,
"learning_rate": 8.713727948448736e-07,
"loss": 0.5429,
"step": 327
},
{
"epoch": 0.2587260895286926,
"grad_norm": 1.9427903905614936,
"learning_rate": 8.705157940370882e-07,
"loss": 0.5398,
"step": 328
},
{
"epoch": 0.2595148885821337,
"grad_norm": 2.304804087786779,
"learning_rate": 8.696563721864947e-07,
"loss": 0.5073,
"step": 329
},
{
"epoch": 0.2603036876355748,
"grad_norm": 2.668665013533607,
"learning_rate": 8.687945349087702e-07,
"loss": 0.5118,
"step": 330
},
{
"epoch": 0.26109248668901597,
"grad_norm": 2.07533269721081,
"learning_rate": 8.679302878353746e-07,
"loss": 0.5216,
"step": 331
},
{
"epoch": 0.2618812857424571,
"grad_norm": 2.1090862307014144,
"learning_rate": 8.670636366135149e-07,
"loss": 0.5317,
"step": 332
},
{
"epoch": 0.26267008479589826,
"grad_norm": 2.3943630928389537,
"learning_rate": 8.661945869061062e-07,
"loss": 0.5412,
"step": 333
},
{
"epoch": 0.26345888384933935,
"grad_norm": 2.136786456391686,
"learning_rate": 8.653231443917371e-07,
"loss": 0.5792,
"step": 334
},
{
"epoch": 0.2642476829027805,
"grad_norm": 2.457546212144828,
"learning_rate": 8.644493147646301e-07,
"loss": 0.5408,
"step": 335
},
{
"epoch": 0.26503648195622165,
"grad_norm": 2.3470099739661774,
"learning_rate": 8.635731037346069e-07,
"loss": 0.5076,
"step": 336
},
{
"epoch": 0.2658252810096628,
"grad_norm": 2.310372885466472,
"learning_rate": 8.626945170270492e-07,
"loss": 0.5657,
"step": 337
},
{
"epoch": 0.26661408006310394,
"grad_norm": 2.0829343745562556,
"learning_rate": 8.618135603828623e-07,
"loss": 0.5633,
"step": 338
},
{
"epoch": 0.26740287911654503,
"grad_norm": 2.00979036757978,
"learning_rate": 8.609302395584371e-07,
"loss": 0.5405,
"step": 339
},
{
"epoch": 0.2681916781699862,
"grad_norm": 2.1647165770802466,
"learning_rate": 8.600445603256123e-07,
"loss": 0.531,
"step": 340
},
{
"epoch": 0.26898047722342733,
"grad_norm": 2.0388503064283925,
"learning_rate": 8.591565284716377e-07,
"loss": 0.5359,
"step": 341
},
{
"epoch": 0.2697692762768685,
"grad_norm": 1.9907100487584009,
"learning_rate": 8.582661497991354e-07,
"loss": 0.5596,
"step": 342
},
{
"epoch": 0.2705580753303096,
"grad_norm": 2.295888796482841,
"learning_rate": 8.57373430126062e-07,
"loss": 0.5447,
"step": 343
},
{
"epoch": 0.2713468743837507,
"grad_norm": 2.2248786701833874,
"learning_rate": 8.564783752856713e-07,
"loss": 0.4776,
"step": 344
},
{
"epoch": 0.27213567343719186,
"grad_norm": 2.048251941558661,
"learning_rate": 8.555809911264749e-07,
"loss": 0.5314,
"step": 345
},
{
"epoch": 0.272924472490633,
"grad_norm": 2.1357310529837505,
"learning_rate": 8.546812835122057e-07,
"loss": 0.5158,
"step": 346
},
{
"epoch": 0.27371327154407415,
"grad_norm": 2.057119587264759,
"learning_rate": 8.537792583217778e-07,
"loss": 0.5282,
"step": 347
},
{
"epoch": 0.2745020705975153,
"grad_norm": 2.122280266292899,
"learning_rate": 8.528749214492492e-07,
"loss": 0.5487,
"step": 348
},
{
"epoch": 0.2752908696509564,
"grad_norm": 2.0552777980933024,
"learning_rate": 8.519682788037834e-07,
"loss": 0.5919,
"step": 349
},
{
"epoch": 0.27607966870439754,
"grad_norm": 1.9639783201943508,
"learning_rate": 8.510593363096097e-07,
"loss": 0.5579,
"step": 350
},
{
"epoch": 0.2768684677578387,
"grad_norm": 1.902315684387984,
"learning_rate": 8.501480999059858e-07,
"loss": 0.5273,
"step": 351
},
{
"epoch": 0.27765726681127983,
"grad_norm": 2.109601696468413,
"learning_rate": 8.492345755471579e-07,
"loss": 0.4997,
"step": 352
},
{
"epoch": 0.278446065864721,
"grad_norm": 2.2089075280462915,
"learning_rate": 8.483187692023228e-07,
"loss": 0.497,
"step": 353
},
{
"epoch": 0.2792348649181621,
"grad_norm": 2.077303795278966,
"learning_rate": 8.474006868555877e-07,
"loss": 0.5675,
"step": 354
},
{
"epoch": 0.2800236639716032,
"grad_norm": 2.01729266276797,
"learning_rate": 8.464803345059323e-07,
"loss": 0.5269,
"step": 355
},
{
"epoch": 0.28081246302504437,
"grad_norm": 2.223738976686647,
"learning_rate": 8.455577181671692e-07,
"loss": 0.5057,
"step": 356
},
{
"epoch": 0.2816012620784855,
"grad_norm": 2.1279940703470492,
"learning_rate": 8.446328438679036e-07,
"loss": 0.5345,
"step": 357
},
{
"epoch": 0.28239006113192666,
"grad_norm": 2.076484930820017,
"learning_rate": 8.437057176514957e-07,
"loss": 0.535,
"step": 358
},
{
"epoch": 0.28317886018536775,
"grad_norm": 2.3150840455699333,
"learning_rate": 8.427763455760197e-07,
"loss": 0.5443,
"step": 359
},
{
"epoch": 0.2839676592388089,
"grad_norm": 2.1962037443030926,
"learning_rate": 8.418447337142253e-07,
"loss": 0.573,
"step": 360
},
{
"epoch": 0.28475645829225005,
"grad_norm": 1.912221543514237,
"learning_rate": 8.409108881534972e-07,
"loss": 0.5099,
"step": 361
},
{
"epoch": 0.2855452573456912,
"grad_norm": 2.1094711301958275,
"learning_rate": 8.399748149958155e-07,
"loss": 0.5375,
"step": 362
},
{
"epoch": 0.28633405639913234,
"grad_norm": 2.285925242807296,
"learning_rate": 8.390365203577164e-07,
"loss": 0.5208,
"step": 363
},
{
"epoch": 0.28712285545257343,
"grad_norm": 2.0948745637451776,
"learning_rate": 8.380960103702514e-07,
"loss": 0.5171,
"step": 364
},
{
"epoch": 0.2879116545060146,
"grad_norm": 2.123594228180867,
"learning_rate": 8.371532911789482e-07,
"loss": 0.5209,
"step": 365
},
{
"epoch": 0.2887004535594557,
"grad_norm": 2.075250983548259,
"learning_rate": 8.362083689437692e-07,
"loss": 0.5672,
"step": 366
},
{
"epoch": 0.2894892526128969,
"grad_norm": 2.1044379496552903,
"learning_rate": 8.352612498390728e-07,
"loss": 0.5523,
"step": 367
},
{
"epoch": 0.290278051666338,
"grad_norm": 1.8013039482820488,
"learning_rate": 8.343119400535717e-07,
"loss": 0.53,
"step": 368
},
{
"epoch": 0.2910668507197791,
"grad_norm": 2.1603803493749916,
"learning_rate": 8.333604457902934e-07,
"loss": 0.498,
"step": 369
},
{
"epoch": 0.29185564977322026,
"grad_norm": 2.122631826422624,
"learning_rate": 8.324067732665393e-07,
"loss": 0.537,
"step": 370
},
{
"epoch": 0.2926444488266614,
"grad_norm": 2.095252848387726,
"learning_rate": 8.314509287138441e-07,
"loss": 0.5526,
"step": 371
},
{
"epoch": 0.29343324788010255,
"grad_norm": 2.3713422485294586,
"learning_rate": 8.304929183779347e-07,
"loss": 0.5121,
"step": 372
},
{
"epoch": 0.2942220469335437,
"grad_norm": 2.297191371277777,
"learning_rate": 8.295327485186905e-07,
"loss": 0.5193,
"step": 373
},
{
"epoch": 0.2950108459869848,
"grad_norm": 1.9959100405942893,
"learning_rate": 8.285704254101011e-07,
"loss": 0.5371,
"step": 374
},
{
"epoch": 0.29579964504042594,
"grad_norm": 2.235088533110939,
"learning_rate": 8.276059553402264e-07,
"loss": 0.5398,
"step": 375
},
{
"epoch": 0.2965884440938671,
"grad_norm": 2.1357393409626435,
"learning_rate": 8.266393446111548e-07,
"loss": 0.5732,
"step": 376
},
{
"epoch": 0.29737724314730823,
"grad_norm": 2.151811513808695,
"learning_rate": 8.256705995389624e-07,
"loss": 0.5551,
"step": 377
},
{
"epoch": 0.2981660422007494,
"grad_norm": 1.9920864543158614,
"learning_rate": 8.246997264536715e-07,
"loss": 0.524,
"step": 378
},
{
"epoch": 0.2989548412541905,
"grad_norm": 2.4406812109725067,
"learning_rate": 8.237267316992097e-07,
"loss": 0.4834,
"step": 379
},
{
"epoch": 0.2997436403076316,
"grad_norm": 2.047950653206123,
"learning_rate": 8.227516216333679e-07,
"loss": 0.5599,
"step": 380
},
{
"epoch": 0.30053243936107277,
"grad_norm": 2.124189487236291,
"learning_rate": 8.21774402627759e-07,
"loss": 0.5403,
"step": 381
},
{
"epoch": 0.3013212384145139,
"grad_norm": 2.047743689518608,
"learning_rate": 8.207950810677762e-07,
"loss": 0.51,
"step": 382
},
{
"epoch": 0.30211003746795506,
"grad_norm": 2.0478210193968036,
"learning_rate": 8.198136633525514e-07,
"loss": 0.5112,
"step": 383
},
{
"epoch": 0.30289883652139615,
"grad_norm": 2.0725914220995443,
"learning_rate": 8.188301558949131e-07,
"loss": 0.4808,
"step": 384
},
{
"epoch": 0.3036876355748373,
"grad_norm": 2.1482745384571817,
"learning_rate": 8.17844565121345e-07,
"loss": 0.5262,
"step": 385
},
{
"epoch": 0.30447643462827845,
"grad_norm": 2.1374710836395185,
"learning_rate": 8.168568974719437e-07,
"loss": 0.561,
"step": 386
},
{
"epoch": 0.3052652336817196,
"grad_norm": 1.8745913187359984,
"learning_rate": 8.158671594003763e-07,
"loss": 0.5666,
"step": 387
},
{
"epoch": 0.30605403273516074,
"grad_norm": 1.9003479278431439,
"learning_rate": 8.148753573738385e-07,
"loss": 0.5939,
"step": 388
},
{
"epoch": 0.30684283178860183,
"grad_norm": 2.05301160238648,
"learning_rate": 8.138814978730132e-07,
"loss": 0.5302,
"step": 389
},
{
"epoch": 0.307631630842043,
"grad_norm": 2.0460535958618453,
"learning_rate": 8.128855873920264e-07,
"loss": 0.5522,
"step": 390
},
{
"epoch": 0.3084204298954841,
"grad_norm": 2.102260007807583,
"learning_rate": 8.118876324384061e-07,
"loss": 0.498,
"step": 391
},
{
"epoch": 0.3092092289489253,
"grad_norm": 2.2662860626090624,
"learning_rate": 8.108876395330397e-07,
"loss": 0.5238,
"step": 392
},
{
"epoch": 0.3099980280023664,
"grad_norm": 2.8406157126509983,
"learning_rate": 8.098856152101305e-07,
"loss": 0.5111,
"step": 393
},
{
"epoch": 0.3107868270558075,
"grad_norm": 1.900764837171523,
"learning_rate": 8.08881566017156e-07,
"loss": 0.4849,
"step": 394
},
{
"epoch": 0.31157562610924866,
"grad_norm": 2.043698100076419,
"learning_rate": 8.078754985148246e-07,
"loss": 0.5431,
"step": 395
},
{
"epoch": 0.3123644251626898,
"grad_norm": 2.0227657064156017,
"learning_rate": 8.068674192770327e-07,
"loss": 0.5721,
"step": 396
},
{
"epoch": 0.31315322421613095,
"grad_norm": 1.8504539055096028,
"learning_rate": 8.058573348908222e-07,
"loss": 0.5422,
"step": 397
},
{
"epoch": 0.3139420232695721,
"grad_norm": 2.231937390888154,
"learning_rate": 8.048452519563369e-07,
"loss": 0.4899,
"step": 398
},
{
"epoch": 0.3147308223230132,
"grad_norm": 2.0154275363669045,
"learning_rate": 8.038311770867796e-07,
"loss": 0.499,
"step": 399
},
{
"epoch": 0.31551962137645434,
"grad_norm": 2.122553058724412,
"learning_rate": 8.02815116908369e-07,
"loss": 0.5414,
"step": 400
},
{
"epoch": 0.3163084204298955,
"grad_norm": 2.168046685805534,
"learning_rate": 8.01797078060296e-07,
"loss": 0.5223,
"step": 401
},
{
"epoch": 0.31709721948333663,
"grad_norm": 2.1690223536048703,
"learning_rate": 8.007770671946815e-07,
"loss": 0.5675,
"step": 402
},
{
"epoch": 0.3178860185367778,
"grad_norm": 2.3425772050171103,
"learning_rate": 7.997550909765309e-07,
"loss": 0.53,
"step": 403
},
{
"epoch": 0.3186748175902189,
"grad_norm": 1.9348424236437924,
"learning_rate": 7.987311560836925e-07,
"loss": 0.5553,
"step": 404
},
{
"epoch": 0.31946361664366,
"grad_norm": 2.1124495468558067,
"learning_rate": 7.977052692068127e-07,
"loss": 0.5261,
"step": 405
},
{
"epoch": 0.32025241569710117,
"grad_norm": 2.2173706478893624,
"learning_rate": 7.966774370492928e-07,
"loss": 0.5731,
"step": 406
},
{
"epoch": 0.3210412147505423,
"grad_norm": 2.227913713108743,
"learning_rate": 7.956476663272452e-07,
"loss": 0.5238,
"step": 407
},
{
"epoch": 0.32183001380398346,
"grad_norm": 2.088898054939403,
"learning_rate": 7.946159637694489e-07,
"loss": 0.561,
"step": 408
},
{
"epoch": 0.32261881285742455,
"grad_norm": 2.26819835430573,
"learning_rate": 7.935823361173066e-07,
"loss": 0.5381,
"step": 409
},
{
"epoch": 0.3234076119108657,
"grad_norm": 2.1924999963355676,
"learning_rate": 7.925467901247995e-07,
"loss": 0.5025,
"step": 410
},
{
"epoch": 0.32419641096430685,
"grad_norm": 2.0026195508439066,
"learning_rate": 7.915093325584439e-07,
"loss": 0.5542,
"step": 411
},
{
"epoch": 0.324985210017748,
"grad_norm": 2.157471910614722,
"learning_rate": 7.904699701972471e-07,
"loss": 0.522,
"step": 412
},
{
"epoch": 0.32577400907118914,
"grad_norm": 2.216340475876602,
"learning_rate": 7.894287098326623e-07,
"loss": 0.5711,
"step": 413
},
{
"epoch": 0.32656280812463023,
"grad_norm": 1.9369315246390006,
"learning_rate": 7.883855582685453e-07,
"loss": 0.5044,
"step": 414
},
{
"epoch": 0.3273516071780714,
"grad_norm": 2.1405821631761284,
"learning_rate": 7.873405223211087e-07,
"loss": 0.5426,
"step": 415
},
{
"epoch": 0.3281404062315125,
"grad_norm": 2.1892744208489914,
"learning_rate": 7.862936088188789e-07,
"loss": 0.5198,
"step": 416
},
{
"epoch": 0.3289292052849537,
"grad_norm": 2.24088380662381,
"learning_rate": 7.852448246026504e-07,
"loss": 0.5146,
"step": 417
},
{
"epoch": 0.3297180043383948,
"grad_norm": 2.0898914021762742,
"learning_rate": 7.841941765254414e-07,
"loss": 0.5036,
"step": 418
},
{
"epoch": 0.3305068033918359,
"grad_norm": 2.123501549648856,
"learning_rate": 7.831416714524489e-07,
"loss": 0.5718,
"step": 419
},
{
"epoch": 0.33129560244527706,
"grad_norm": 2.057860826653185,
"learning_rate": 7.820873162610044e-07,
"loss": 0.5432,
"step": 420
},
{
"epoch": 0.3320844014987182,
"grad_norm": 2.4230755774304265,
"learning_rate": 7.810311178405278e-07,
"loss": 0.4946,
"step": 421
},
{
"epoch": 0.33287320055215935,
"grad_norm": 2.1913869862349507,
"learning_rate": 7.79973083092484e-07,
"loss": 0.5421,
"step": 422
},
{
"epoch": 0.33366199960560045,
"grad_norm": 2.1897520765261924,
"learning_rate": 7.789132189303363e-07,
"loss": 0.5004,
"step": 423
},
{
"epoch": 0.3344507986590416,
"grad_norm": 2.00358067747818,
"learning_rate": 7.778515322795021e-07,
"loss": 0.5345,
"step": 424
},
{
"epoch": 0.33523959771248274,
"grad_norm": 1.9349640558251588,
"learning_rate": 7.767880300773073e-07,
"loss": 0.531,
"step": 425
},
{
"epoch": 0.3360283967659239,
"grad_norm": 2.039149588987288,
"learning_rate": 7.757227192729412e-07,
"loss": 0.5548,
"step": 426
},
{
"epoch": 0.33681719581936503,
"grad_norm": 2.02241947690995,
"learning_rate": 7.746556068274109e-07,
"loss": 0.5257,
"step": 427
},
{
"epoch": 0.3376059948728061,
"grad_norm": 2.1321251597690103,
"learning_rate": 7.735866997134957e-07,
"loss": 0.5101,
"step": 428
},
{
"epoch": 0.3383947939262473,
"grad_norm": 2.308071107409399,
"learning_rate": 7.725160049157017e-07,
"loss": 0.5325,
"step": 429
},
{
"epoch": 0.3391835929796884,
"grad_norm": 2.4088202220871255,
"learning_rate": 7.714435294302167e-07,
"loss": 0.5131,
"step": 430
},
{
"epoch": 0.33997239203312957,
"grad_norm": 1.9870249492523793,
"learning_rate": 7.703692802648635e-07,
"loss": 0.54,
"step": 431
},
{
"epoch": 0.3407611910865707,
"grad_norm": 2.1534632225703003,
"learning_rate": 7.692932644390546e-07,
"loss": 0.4884,
"step": 432
},
{
"epoch": 0.3415499901400118,
"grad_norm": 2.352875679116166,
"learning_rate": 7.682154889837463e-07,
"loss": 0.5166,
"step": 433
},
{
"epoch": 0.34233878919345295,
"grad_norm": 2.1192339616164255,
"learning_rate": 7.671359609413926e-07,
"loss": 0.5582,
"step": 434
},
{
"epoch": 0.3431275882468941,
"grad_norm": 2.2232407791132465,
"learning_rate": 7.660546873658999e-07,
"loss": 0.5576,
"step": 435
},
{
"epoch": 0.34391638730033525,
"grad_norm": 2.1380400140284186,
"learning_rate": 7.649716753225802e-07,
"loss": 0.5254,
"step": 436
},
{
"epoch": 0.3447051863537764,
"grad_norm": 2.2000279799990508,
"learning_rate": 7.638869318881042e-07,
"loss": 0.5244,
"step": 437
},
{
"epoch": 0.3454939854072175,
"grad_norm": 2.1606458172053538,
"learning_rate": 7.628004641504573e-07,
"loss": 0.5287,
"step": 438
},
{
"epoch": 0.34628278446065863,
"grad_norm": 1.9670000617278367,
"learning_rate": 7.617122792088908e-07,
"loss": 0.5906,
"step": 439
},
{
"epoch": 0.3470715835140998,
"grad_norm": 2.1000056057786614,
"learning_rate": 7.606223841738775e-07,
"loss": 0.5335,
"step": 440
},
{
"epoch": 0.3478603825675409,
"grad_norm": 2.3026964176816542,
"learning_rate": 7.595307861670637e-07,
"loss": 0.5629,
"step": 441
},
{
"epoch": 0.3486491816209821,
"grad_norm": 2.01784399539267,
"learning_rate": 7.584374923212235e-07,
"loss": 0.5452,
"step": 442
},
{
"epoch": 0.34943798067442317,
"grad_norm": 2.3382330228698165,
"learning_rate": 7.573425097802126e-07,
"loss": 0.5047,
"step": 443
},
{
"epoch": 0.3502267797278643,
"grad_norm": 2.065103279357655,
"learning_rate": 7.562458456989202e-07,
"loss": 0.5568,
"step": 444
},
{
"epoch": 0.35101557878130546,
"grad_norm": 2.1310896070029113,
"learning_rate": 7.551475072432236e-07,
"loss": 0.5622,
"step": 445
},
{
"epoch": 0.3518043778347466,
"grad_norm": 2.1909720621256232,
"learning_rate": 7.54047501589941e-07,
"loss": 0.5565,
"step": 446
},
{
"epoch": 0.35259317688818775,
"grad_norm": 1.9597026067018604,
"learning_rate": 7.52945835926784e-07,
"loss": 0.4942,
"step": 447
},
{
"epoch": 0.35338197594162885,
"grad_norm": 2.0131446081156747,
"learning_rate": 7.518425174523118e-07,
"loss": 0.5745,
"step": 448
},
{
"epoch": 0.35417077499507,
"grad_norm": 2.1465297789608857,
"learning_rate": 7.507375533758827e-07,
"loss": 0.5563,
"step": 449
},
{
"epoch": 0.35495957404851114,
"grad_norm": 2.3211964849398217,
"learning_rate": 7.496309509176082e-07,
"loss": 0.4636,
"step": 450
},
{
"epoch": 0.3557483731019523,
"grad_norm": 2.1043019289103193,
"learning_rate": 7.485227173083055e-07,
"loss": 0.519,
"step": 451
},
{
"epoch": 0.35653717215539343,
"grad_norm": 2.0114595757108797,
"learning_rate": 7.4741285978945e-07,
"loss": 0.5314,
"step": 452
},
{
"epoch": 0.3573259712088345,
"grad_norm": 2.0661086545879117,
"learning_rate": 7.463013856131282e-07,
"loss": 0.5073,
"step": 453
},
{
"epoch": 0.3581147702622757,
"grad_norm": 2.0900068029434387,
"learning_rate": 7.451883020419899e-07,
"loss": 0.5059,
"step": 454
},
{
"epoch": 0.3589035693157168,
"grad_norm": 2.103966500969913,
"learning_rate": 7.440736163492016e-07,
"loss": 0.5213,
"step": 455
},
{
"epoch": 0.35969236836915797,
"grad_norm": 2.0245281002973834,
"learning_rate": 7.429573358183983e-07,
"loss": 0.4941,
"step": 456
},
{
"epoch": 0.3604811674225991,
"grad_norm": 2.2557774839292937,
"learning_rate": 7.418394677436357e-07,
"loss": 0.5179,
"step": 457
},
{
"epoch": 0.3612699664760402,
"grad_norm": 1.9376024457260288,
"learning_rate": 7.407200194293435e-07,
"loss": 0.5025,
"step": 458
},
{
"epoch": 0.36205876552948135,
"grad_norm": 2.148282681240793,
"learning_rate": 7.395989981902765e-07,
"loss": 0.5039,
"step": 459
},
{
"epoch": 0.3628475645829225,
"grad_norm": 2.420727882030519,
"learning_rate": 7.384764113514679e-07,
"loss": 0.5243,
"step": 460
},
{
"epoch": 0.36363636363636365,
"grad_norm": 2.3031173638923366,
"learning_rate": 7.373522662481808e-07,
"loss": 0.5379,
"step": 461
},
{
"epoch": 0.3644251626898048,
"grad_norm": 2.275238316861926,
"learning_rate": 7.3622657022586e-07,
"loss": 0.5399,
"step": 462
},
{
"epoch": 0.3652139617432459,
"grad_norm": 2.051350190050254,
"learning_rate": 7.350993306400849e-07,
"loss": 0.5417,
"step": 463
},
{
"epoch": 0.36600276079668703,
"grad_norm": 2.2250004861949573,
"learning_rate": 7.339705548565205e-07,
"loss": 0.4643,
"step": 464
},
{
"epoch": 0.3667915598501282,
"grad_norm": 1.9741309712433746,
"learning_rate": 7.328402502508701e-07,
"loss": 0.4953,
"step": 465
},
{
"epoch": 0.3675803589035693,
"grad_norm": 2.2500308780144644,
"learning_rate": 7.317084242088263e-07,
"loss": 0.5196,
"step": 466
},
{
"epoch": 0.3683691579570105,
"grad_norm": 2.106653826954423,
"learning_rate": 7.305750841260235e-07,
"loss": 0.5337,
"step": 467
},
{
"epoch": 0.36915795701045157,
"grad_norm": 2.255057858697615,
"learning_rate": 7.294402374079891e-07,
"loss": 0.5224,
"step": 468
},
{
"epoch": 0.3699467560638927,
"grad_norm": 2.191220603944603,
"learning_rate": 7.283038914700951e-07,
"loss": 0.4985,
"step": 469
},
{
"epoch": 0.37073555511733386,
"grad_norm": 2.1076779916683916,
"learning_rate": 7.271660537375102e-07,
"loss": 0.5273,
"step": 470
},
{
"epoch": 0.371524354170775,
"grad_norm": 1.8972119336416169,
"learning_rate": 7.2602673164515e-07,
"loss": 0.5244,
"step": 471
},
{
"epoch": 0.37231315322421615,
"grad_norm": 2.2406182593711774,
"learning_rate": 7.248859326376304e-07,
"loss": 0.5068,
"step": 472
},
{
"epoch": 0.37310195227765725,
"grad_norm": 2.179472150786481,
"learning_rate": 7.237436641692168e-07,
"loss": 0.5,
"step": 473
},
{
"epoch": 0.3738907513310984,
"grad_norm": 2.0430623755065924,
"learning_rate": 7.225999337037771e-07,
"loss": 0.5,
"step": 474
},
{
"epoch": 0.37467955038453954,
"grad_norm": 2.248823383997148,
"learning_rate": 7.21454748714732e-07,
"loss": 0.5163,
"step": 475
},
{
"epoch": 0.3754683494379807,
"grad_norm": 2.0554770933041264,
"learning_rate": 7.203081166850065e-07,
"loss": 0.5045,
"step": 476
},
{
"epoch": 0.37625714849142183,
"grad_norm": 2.0373980848549436,
"learning_rate": 7.191600451069807e-07,
"loss": 0.5213,
"step": 477
},
{
"epoch": 0.3770459475448629,
"grad_norm": 2.2332165399845563,
"learning_rate": 7.180105414824416e-07,
"loss": 0.5649,
"step": 478
},
{
"epoch": 0.37783474659830407,
"grad_norm": 2.0619586679434576,
"learning_rate": 7.168596133225328e-07,
"loss": 0.5112,
"step": 479
},
{
"epoch": 0.3786235456517452,
"grad_norm": 2.3673578822047303,
"learning_rate": 7.157072681477068e-07,
"loss": 0.4604,
"step": 480
},
{
"epoch": 0.37941234470518637,
"grad_norm": 2.0696488666186985,
"learning_rate": 7.14553513487675e-07,
"loss": 0.5753,
"step": 481
},
{
"epoch": 0.3802011437586275,
"grad_norm": 2.326499372685507,
"learning_rate": 7.133983568813586e-07,
"loss": 0.5128,
"step": 482
},
{
"epoch": 0.3809899428120686,
"grad_norm": 2.1483773403215576,
"learning_rate": 7.122418058768398e-07,
"loss": 0.4797,
"step": 483
},
{
"epoch": 0.38177874186550975,
"grad_norm": 2.132756493149541,
"learning_rate": 7.110838680313118e-07,
"loss": 0.5012,
"step": 484
},
{
"epoch": 0.3825675409189509,
"grad_norm": 2.164264128755106,
"learning_rate": 7.099245509110298e-07,
"loss": 0.4745,
"step": 485
},
{
"epoch": 0.38335633997239205,
"grad_norm": 2.283691680195981,
"learning_rate": 7.08763862091262e-07,
"loss": 0.5281,
"step": 486
},
{
"epoch": 0.3841451390258332,
"grad_norm": 2.110858300152156,
"learning_rate": 7.076018091562388e-07,
"loss": 0.511,
"step": 487
},
{
"epoch": 0.3849339380792743,
"grad_norm": 2.18571029843053,
"learning_rate": 7.064383996991048e-07,
"loss": 0.4831,
"step": 488
},
{
"epoch": 0.38572273713271543,
"grad_norm": 2.178831073574741,
"learning_rate": 7.052736413218682e-07,
"loss": 0.4974,
"step": 489
},
{
"epoch": 0.3865115361861566,
"grad_norm": 2.10176296853287,
"learning_rate": 7.041075416353513e-07,
"loss": 0.5647,
"step": 490
},
{
"epoch": 0.3873003352395977,
"grad_norm": 2.168889861754043,
"learning_rate": 7.029401082591408e-07,
"loss": 0.544,
"step": 491
},
{
"epoch": 0.3880891342930389,
"grad_norm": 2.429159916070289,
"learning_rate": 7.017713488215385e-07,
"loss": 0.4432,
"step": 492
},
{
"epoch": 0.38887793334647996,
"grad_norm": 2.0223962175979877,
"learning_rate": 7.006012709595104e-07,
"loss": 0.5208,
"step": 493
},
{
"epoch": 0.3896667323999211,
"grad_norm": 2.4243441791261415,
"learning_rate": 6.994298823186378e-07,
"loss": 0.5121,
"step": 494
},
{
"epoch": 0.39045553145336226,
"grad_norm": 1.8915847860005757,
"learning_rate": 6.982571905530669e-07,
"loss": 0.5294,
"step": 495
},
{
"epoch": 0.3912443305068034,
"grad_norm": 2.23982872653249,
"learning_rate": 6.970832033254588e-07,
"loss": 0.5094,
"step": 496
},
{
"epoch": 0.39203312956024455,
"grad_norm": 2.0569505174202316,
"learning_rate": 6.959079283069394e-07,
"loss": 0.5242,
"step": 497
},
{
"epoch": 0.39282192861368564,
"grad_norm": 2.1589240271344656,
"learning_rate": 6.947313731770495e-07,
"loss": 0.5272,
"step": 498
},
{
"epoch": 0.3936107276671268,
"grad_norm": 2.103084067358058,
"learning_rate": 6.935535456236945e-07,
"loss": 0.4901,
"step": 499
},
{
"epoch": 0.39439952672056794,
"grad_norm": 2.2659931781303,
"learning_rate": 6.923744533430938e-07,
"loss": 0.5176,
"step": 500
},
{
"epoch": 0.3951883257740091,
"grad_norm": 2.2046253054770664,
"learning_rate": 6.911941040397312e-07,
"loss": 0.5253,
"step": 501
},
{
"epoch": 0.39597712482745023,
"grad_norm": 2.2589388443487826,
"learning_rate": 6.900125054263041e-07,
"loss": 0.4772,
"step": 502
},
{
"epoch": 0.3967659238808913,
"grad_norm": 2.3451142698282816,
"learning_rate": 6.888296652236731e-07,
"loss": 0.4789,
"step": 503
},
{
"epoch": 0.39755472293433247,
"grad_norm": 2.018370821680395,
"learning_rate": 6.876455911608118e-07,
"loss": 0.4773,
"step": 504
},
{
"epoch": 0.3983435219877736,
"grad_norm": 2.457567781792435,
"learning_rate": 6.864602909747563e-07,
"loss": 0.5375,
"step": 505
},
{
"epoch": 0.39913232104121477,
"grad_norm": 2.00239260813027,
"learning_rate": 6.852737724105538e-07,
"loss": 0.5249,
"step": 506
},
{
"epoch": 0.3999211200946559,
"grad_norm": 2.2377516872753427,
"learning_rate": 6.840860432212135e-07,
"loss": 0.5335,
"step": 507
},
{
"epoch": 0.400709919148097,
"grad_norm": 1.9235377664216216,
"learning_rate": 6.828971111676546e-07,
"loss": 0.5196,
"step": 508
},
{
"epoch": 0.40149871820153815,
"grad_norm": 2.1776496338320435,
"learning_rate": 6.817069840186565e-07,
"loss": 0.491,
"step": 509
},
{
"epoch": 0.4022875172549793,
"grad_norm": 2.0900696625367194,
"learning_rate": 6.805156695508074e-07,
"loss": 0.5298,
"step": 510
},
{
"epoch": 0.40307631630842045,
"grad_norm": 2.0637407028355246,
"learning_rate": 6.793231755484538e-07,
"loss": 0.5329,
"step": 511
},
{
"epoch": 0.4038651153618616,
"grad_norm": 2.2065814707709257,
"learning_rate": 6.781295098036496e-07,
"loss": 0.5129,
"step": 512
},
{
"epoch": 0.4046539144153027,
"grad_norm": 1.9448645371893218,
"learning_rate": 6.76934680116105e-07,
"loss": 0.5169,
"step": 513
},
{
"epoch": 0.40544271346874383,
"grad_norm": 2.177071136566607,
"learning_rate": 6.757386942931359e-07,
"loss": 0.4706,
"step": 514
},
{
"epoch": 0.406231512522185,
"grad_norm": 2.2070355178486136,
"learning_rate": 6.745415601496126e-07,
"loss": 0.4931,
"step": 515
},
{
"epoch": 0.4070203115756261,
"grad_norm": 2.146060319715439,
"learning_rate": 6.73343285507909e-07,
"loss": 0.5213,
"step": 516
},
{
"epoch": 0.4078091106290672,
"grad_norm": 2.170440228171354,
"learning_rate": 6.721438781978508e-07,
"loss": 0.5477,
"step": 517
},
{
"epoch": 0.40859790968250836,
"grad_norm": 2.1807452801276965,
"learning_rate": 6.709433460566654e-07,
"loss": 0.5101,
"step": 518
},
{
"epoch": 0.4093867087359495,
"grad_norm": 2.0913861606653676,
"learning_rate": 6.697416969289297e-07,
"loss": 0.4745,
"step": 519
},
{
"epoch": 0.41017550778939066,
"grad_norm": 2.25295818058408,
"learning_rate": 6.685389386665196e-07,
"loss": 0.4965,
"step": 520
},
{
"epoch": 0.4109643068428318,
"grad_norm": 2.0153334573129036,
"learning_rate": 6.67335079128558e-07,
"loss": 0.5122,
"step": 521
},
{
"epoch": 0.4117531058962729,
"grad_norm": 2.132034179113491,
"learning_rate": 6.661301261813643e-07,
"loss": 0.5069,
"step": 522
},
{
"epoch": 0.41254190494971404,
"grad_norm": 2.2010318633116857,
"learning_rate": 6.649240876984022e-07,
"loss": 0.5242,
"step": 523
},
{
"epoch": 0.4133307040031552,
"grad_norm": 2.112130583250756,
"learning_rate": 6.637169715602285e-07,
"loss": 0.505,
"step": 524
},
{
"epoch": 0.41411950305659634,
"grad_norm": 2.076784653241366,
"learning_rate": 6.625087856544416e-07,
"loss": 0.4912,
"step": 525
},
{
"epoch": 0.4149083021100375,
"grad_norm": 2.27780776988912,
"learning_rate": 6.612995378756304e-07,
"loss": 0.496,
"step": 526
},
{
"epoch": 0.4156971011634786,
"grad_norm": 2.1976716162586554,
"learning_rate": 6.600892361253222e-07,
"loss": 0.4895,
"step": 527
},
{
"epoch": 0.4164859002169197,
"grad_norm": 2.868408168751329,
"learning_rate": 6.588778883119309e-07,
"loss": 0.5043,
"step": 528
},
{
"epoch": 0.41727469927036087,
"grad_norm": 1.924756886144101,
"learning_rate": 6.576655023507061e-07,
"loss": 0.4787,
"step": 529
},
{
"epoch": 0.418063498323802,
"grad_norm": 1.9984329102702139,
"learning_rate": 6.564520861636807e-07,
"loss": 0.5356,
"step": 530
},
{
"epoch": 0.41885229737724317,
"grad_norm": 2.148407961243213,
"learning_rate": 6.552376476796193e-07,
"loss": 0.4752,
"step": 531
},
{
"epoch": 0.41964109643068426,
"grad_norm": 2.1259941384728296,
"learning_rate": 6.540221948339665e-07,
"loss": 0.4617,
"step": 532
},
{
"epoch": 0.4204298954841254,
"grad_norm": 1.9357007353302174,
"learning_rate": 6.528057355687948e-07,
"loss": 0.4883,
"step": 533
},
{
"epoch": 0.42121869453756655,
"grad_norm": 2.0117500966387176,
"learning_rate": 6.515882778327535e-07,
"loss": 0.5344,
"step": 534
},
{
"epoch": 0.4220074935910077,
"grad_norm": 1.9069311353247935,
"learning_rate": 6.503698295810153e-07,
"loss": 0.527,
"step": 535
},
{
"epoch": 0.42279629264444885,
"grad_norm": 2.0198590134464176,
"learning_rate": 6.491503987752263e-07,
"loss": 0.5107,
"step": 536
},
{
"epoch": 0.42358509169788994,
"grad_norm": 2.388365519847473,
"learning_rate": 6.479299933834514e-07,
"loss": 0.4854,
"step": 537
},
{
"epoch": 0.4243738907513311,
"grad_norm": 2.1322654188674655,
"learning_rate": 6.467086213801248e-07,
"loss": 0.5285,
"step": 538
},
{
"epoch": 0.42516268980477223,
"grad_norm": 2.127975785844204,
"learning_rate": 6.454862907459966e-07,
"loss": 0.4721,
"step": 539
},
{
"epoch": 0.4259514888582134,
"grad_norm": 2.181072579499217,
"learning_rate": 6.442630094680804e-07,
"loss": 0.521,
"step": 540
},
{
"epoch": 0.4267402879116545,
"grad_norm": 2.1206552390036255,
"learning_rate": 6.430387855396022e-07,
"loss": 0.4781,
"step": 541
},
{
"epoch": 0.4275290869650956,
"grad_norm": 2.079779704585345,
"learning_rate": 6.418136269599468e-07,
"loss": 0.496,
"step": 542
},
{
"epoch": 0.42831788601853676,
"grad_norm": 2.2306801877612967,
"learning_rate": 6.405875417346067e-07,
"loss": 0.5159,
"step": 543
},
{
"epoch": 0.4291066850719779,
"grad_norm": 2.063597041593193,
"learning_rate": 6.393605378751292e-07,
"loss": 0.5045,
"step": 544
},
{
"epoch": 0.42989548412541906,
"grad_norm": 2.0134729877949447,
"learning_rate": 6.381326233990643e-07,
"loss": 0.5072,
"step": 545
},
{
"epoch": 0.4306842831788602,
"grad_norm": 2.040551249934524,
"learning_rate": 6.369038063299118e-07,
"loss": 0.4667,
"step": 546
},
{
"epoch": 0.4314730822323013,
"grad_norm": 2.2727384988604595,
"learning_rate": 6.356740946970701e-07,
"loss": 0.5175,
"step": 547
},
{
"epoch": 0.43226188128574244,
"grad_norm": 2.034529773904068,
"learning_rate": 6.344434965357818e-07,
"loss": 0.4833,
"step": 548
},
{
"epoch": 0.4330506803391836,
"grad_norm": 2.370005501189888,
"learning_rate": 6.332120198870834e-07,
"loss": 0.4647,
"step": 549
},
{
"epoch": 0.43383947939262474,
"grad_norm": 2.0770065043495762,
"learning_rate": 6.319796727977508e-07,
"loss": 0.5259,
"step": 550
},
{
"epoch": 0.4346282784460659,
"grad_norm": 2.1036456937958374,
"learning_rate": 6.307464633202478e-07,
"loss": 0.5109,
"step": 551
},
{
"epoch": 0.435417077499507,
"grad_norm": 2.6983174125851406,
"learning_rate": 6.295123995126737e-07,
"loss": 0.5581,
"step": 552
},
{
"epoch": 0.4362058765529481,
"grad_norm": 2.294316329636296,
"learning_rate": 6.282774894387099e-07,
"loss": 0.5493,
"step": 553
},
{
"epoch": 0.43699467560638927,
"grad_norm": 2.0913227290352623,
"learning_rate": 6.27041741167567e-07,
"loss": 0.5171,
"step": 554
},
{
"epoch": 0.4377834746598304,
"grad_norm": 2.292995694573123,
"learning_rate": 6.258051627739336e-07,
"loss": 0.4962,
"step": 555
},
{
"epoch": 0.43857227371327157,
"grad_norm": 1.9579051004007573,
"learning_rate": 6.245677623379216e-07,
"loss": 0.5003,
"step": 556
},
{
"epoch": 0.43936107276671266,
"grad_norm": 2.297306072949353,
"learning_rate": 6.233295479450151e-07,
"loss": 0.4998,
"step": 557
},
{
"epoch": 0.4401498718201538,
"grad_norm": 2.249711424186936,
"learning_rate": 6.220905276860161e-07,
"loss": 0.5279,
"step": 558
},
{
"epoch": 0.44093867087359495,
"grad_norm": 2.1241366839450477,
"learning_rate": 6.208507096569929e-07,
"loss": 0.5109,
"step": 559
},
{
"epoch": 0.4417274699270361,
"grad_norm": 2.108294067641242,
"learning_rate": 6.196101019592263e-07,
"loss": 0.4735,
"step": 560
},
{
"epoch": 0.44251626898047725,
"grad_norm": 2.140285766397059,
"learning_rate": 6.183687126991572e-07,
"loss": 0.4822,
"step": 561
},
{
"epoch": 0.44330506803391834,
"grad_norm": 2.515982308138581,
"learning_rate": 6.171265499883331e-07,
"loss": 0.512,
"step": 562
},
{
"epoch": 0.4440938670873595,
"grad_norm": 2.035321022213538,
"learning_rate": 6.158836219433556e-07,
"loss": 0.5046,
"step": 563
},
{
"epoch": 0.44488266614080063,
"grad_norm": 2.0199791922731083,
"learning_rate": 6.146399366858276e-07,
"loss": 0.4782,
"step": 564
},
{
"epoch": 0.4456714651942418,
"grad_norm": 2.349407665179725,
"learning_rate": 6.13395502342299e-07,
"loss": 0.4614,
"step": 565
},
{
"epoch": 0.4464602642476829,
"grad_norm": 2.40716885639539,
"learning_rate": 6.121503270442152e-07,
"loss": 0.4535,
"step": 566
},
{
"epoch": 0.447249063301124,
"grad_norm": 1.8980007571067041,
"learning_rate": 6.109044189278626e-07,
"loss": 0.5237,
"step": 567
},
{
"epoch": 0.44803786235456516,
"grad_norm": 2.365607392658387,
"learning_rate": 6.096577861343162e-07,
"loss": 0.5268,
"step": 568
},
{
"epoch": 0.4488266614080063,
"grad_norm": 2.2297453603544373,
"learning_rate": 6.084104368093865e-07,
"loss": 0.4914,
"step": 569
},
{
"epoch": 0.44961546046144746,
"grad_norm": 2.2442058753632628,
"learning_rate": 6.071623791035657e-07,
"loss": 0.5302,
"step": 570
},
{
"epoch": 0.4504042595148886,
"grad_norm": 2.247446505429321,
"learning_rate": 6.059136211719744e-07,
"loss": 0.4785,
"step": 571
},
{
"epoch": 0.4511930585683297,
"grad_norm": 2.1610235617883853,
"learning_rate": 6.046641711743097e-07,
"loss": 0.4921,
"step": 572
},
{
"epoch": 0.45198185762177084,
"grad_norm": 2.1507050779008674,
"learning_rate": 6.034140372747898e-07,
"loss": 0.5107,
"step": 573
},
{
"epoch": 0.452770656675212,
"grad_norm": 2.271088563006631,
"learning_rate": 6.02163227642102e-07,
"loss": 0.4959,
"step": 574
},
{
"epoch": 0.45355945572865314,
"grad_norm": 2.146959387271005,
"learning_rate": 6.009117504493493e-07,
"loss": 0.4717,
"step": 575
},
{
"epoch": 0.4543482547820943,
"grad_norm": 2.190827548804745,
"learning_rate": 5.996596138739963e-07,
"loss": 0.5637,
"step": 576
},
{
"epoch": 0.4551370538355354,
"grad_norm": 2.220073172142695,
"learning_rate": 5.984068260978166e-07,
"loss": 0.4889,
"step": 577
},
{
"epoch": 0.4559258528889765,
"grad_norm": 1.9418743500279265,
"learning_rate": 5.971533953068385e-07,
"loss": 0.5214,
"step": 578
},
{
"epoch": 0.45671465194241767,
"grad_norm": 2.1995657886716544,
"learning_rate": 5.95899329691292e-07,
"loss": 0.4965,
"step": 579
},
{
"epoch": 0.4575034509958588,
"grad_norm": 2.487117975097244,
"learning_rate": 5.946446374455554e-07,
"loss": 0.496,
"step": 580
},
{
"epoch": 0.45829225004929997,
"grad_norm": 2.246914412191084,
"learning_rate": 5.933893267681014e-07,
"loss": 0.5042,
"step": 581
},
{
"epoch": 0.45908104910274106,
"grad_norm": 2.442236454193494,
"learning_rate": 5.921334058614439e-07,
"loss": 0.5531,
"step": 582
},
{
"epoch": 0.4598698481561822,
"grad_norm": 2.361991668365852,
"learning_rate": 5.908768829320839e-07,
"loss": 0.5244,
"step": 583
},
{
"epoch": 0.46065864720962335,
"grad_norm": 2.123142316411905,
"learning_rate": 5.896197661904559e-07,
"loss": 0.4996,
"step": 584
},
{
"epoch": 0.4614474462630645,
"grad_norm": 2.2772071846352673,
"learning_rate": 5.883620638508755e-07,
"loss": 0.5043,
"step": 585
},
{
"epoch": 0.46223624531650565,
"grad_norm": 2.2247150419859163,
"learning_rate": 5.87103784131484e-07,
"loss": 0.5129,
"step": 586
},
{
"epoch": 0.46302504436994674,
"grad_norm": 2.080488861402785,
"learning_rate": 5.858449352541953e-07,
"loss": 0.5059,
"step": 587
},
{
"epoch": 0.4638138434233879,
"grad_norm": 2.0535063144008485,
"learning_rate": 5.845855254446429e-07,
"loss": 0.4965,
"step": 588
},
{
"epoch": 0.46460264247682903,
"grad_norm": 2.0609964394882163,
"learning_rate": 5.833255629321252e-07,
"loss": 0.5138,
"step": 589
},
{
"epoch": 0.4653914415302702,
"grad_norm": 2.221045475915457,
"learning_rate": 5.820650559495523e-07,
"loss": 0.4862,
"step": 590
},
{
"epoch": 0.4661802405837113,
"grad_norm": 2.060326533076467,
"learning_rate": 5.808040127333916e-07,
"loss": 0.512,
"step": 591
},
{
"epoch": 0.4669690396371524,
"grad_norm": 1.981325078934691,
"learning_rate": 5.795424415236148e-07,
"loss": 0.5062,
"step": 592
},
{
"epoch": 0.46775783869059356,
"grad_norm": 2.0012054280862914,
"learning_rate": 5.782803505636437e-07,
"loss": 0.5068,
"step": 593
},
{
"epoch": 0.4685466377440347,
"grad_norm": 2.0834441157848462,
"learning_rate": 5.770177481002958e-07,
"loss": 0.5178,
"step": 594
},
{
"epoch": 0.46933543679747586,
"grad_norm": 2.132740183791119,
"learning_rate": 5.757546423837314e-07,
"loss": 0.4875,
"step": 595
},
{
"epoch": 0.470124235850917,
"grad_norm": 2.1014430577342673,
"learning_rate": 5.744910416673988e-07,
"loss": 0.4776,
"step": 596
},
{
"epoch": 0.4709130349043581,
"grad_norm": 2.0329388238004698,
"learning_rate": 5.732269542079811e-07,
"loss": 0.5117,
"step": 597
},
{
"epoch": 0.47170183395779924,
"grad_norm": 2.4174377693831537,
"learning_rate": 5.719623882653417e-07,
"loss": 0.5135,
"step": 598
},
{
"epoch": 0.4724906330112404,
"grad_norm": 2.2649519220019525,
"learning_rate": 5.706973521024703e-07,
"loss": 0.5205,
"step": 599
},
{
"epoch": 0.47327943206468154,
"grad_norm": 2.41424368892111,
"learning_rate": 5.694318539854296e-07,
"loss": 0.4975,
"step": 600
},
{
"epoch": 0.4740682311181227,
"grad_norm": 2.3217296493809374,
"learning_rate": 5.681659021833005e-07,
"loss": 0.5125,
"step": 601
},
{
"epoch": 0.4748570301715638,
"grad_norm": 2.0775856298653257,
"learning_rate": 5.668995049681285e-07,
"loss": 0.4795,
"step": 602
},
{
"epoch": 0.4756458292250049,
"grad_norm": 2.254097185838289,
"learning_rate": 5.656326706148699e-07,
"loss": 0.5195,
"step": 603
},
{
"epoch": 0.47643462827844607,
"grad_norm": 1.9948214882179016,
"learning_rate": 5.643654074013363e-07,
"loss": 0.5198,
"step": 604
},
{
"epoch": 0.4772234273318872,
"grad_norm": 2.048235873908938,
"learning_rate": 5.630977236081429e-07,
"loss": 0.4913,
"step": 605
},
{
"epoch": 0.47801222638532836,
"grad_norm": 2.2336870988275823,
"learning_rate": 5.618296275186522e-07,
"loss": 0.4767,
"step": 606
},
{
"epoch": 0.47880102543876946,
"grad_norm": 2.243933354610756,
"learning_rate": 5.60561127418921e-07,
"loss": 0.4874,
"step": 607
},
{
"epoch": 0.4795898244922106,
"grad_norm": 2.254190317268258,
"learning_rate": 5.592922315976463e-07,
"loss": 0.575,
"step": 608
},
{
"epoch": 0.48037862354565175,
"grad_norm": 2.0558251991763488,
"learning_rate": 5.580229483461101e-07,
"loss": 0.5083,
"step": 609
},
{
"epoch": 0.4811674225990929,
"grad_norm": 2.2527330910137,
"learning_rate": 5.567532859581267e-07,
"loss": 0.5058,
"step": 610
},
{
"epoch": 0.481956221652534,
"grad_norm": 2.185212933393249,
"learning_rate": 5.554832527299874e-07,
"loss": 0.5211,
"step": 611
},
{
"epoch": 0.48274502070597514,
"grad_norm": 2.2873780364797285,
"learning_rate": 5.542128569604068e-07,
"loss": 0.5282,
"step": 612
},
{
"epoch": 0.4835338197594163,
"grad_norm": 2.1128456615428006,
"learning_rate": 5.529421069504684e-07,
"loss": 0.5062,
"step": 613
},
{
"epoch": 0.48432261881285743,
"grad_norm": 2.445213691846106,
"learning_rate": 5.516710110035703e-07,
"loss": 0.4695,
"step": 614
},
{
"epoch": 0.4851114178662986,
"grad_norm": 2.3447554948983274,
"learning_rate": 5.503995774253714e-07,
"loss": 0.4925,
"step": 615
},
{
"epoch": 0.48590021691973967,
"grad_norm": 2.1622996870450315,
"learning_rate": 5.491278145237361e-07,
"loss": 0.5226,
"step": 616
},
{
"epoch": 0.4866890159731808,
"grad_norm": 2.0746908604411405,
"learning_rate": 5.478557306086813e-07,
"loss": 0.4825,
"step": 617
},
{
"epoch": 0.48747781502662196,
"grad_norm": 2.35310002860545,
"learning_rate": 5.465833339923212e-07,
"loss": 0.4979,
"step": 618
},
{
"epoch": 0.4882666140800631,
"grad_norm": 2.006711639131911,
"learning_rate": 5.453106329888135e-07,
"loss": 0.5111,
"step": 619
},
{
"epoch": 0.48905541313350426,
"grad_norm": 2.103462051330557,
"learning_rate": 5.440376359143042e-07,
"loss": 0.4698,
"step": 620
},
{
"epoch": 0.48984421218694535,
"grad_norm": 2.081280512092479,
"learning_rate": 5.427643510868751e-07,
"loss": 0.494,
"step": 621
},
{
"epoch": 0.4906330112403865,
"grad_norm": 2.0281990795500606,
"learning_rate": 5.414907868264868e-07,
"loss": 0.4933,
"step": 622
},
{
"epoch": 0.49142181029382764,
"grad_norm": 1.866655364652089,
"learning_rate": 5.402169514549271e-07,
"loss": 0.5156,
"step": 623
},
{
"epoch": 0.4922106093472688,
"grad_norm": 2.1349539483424302,
"learning_rate": 5.389428532957541e-07,
"loss": 0.5311,
"step": 624
},
{
"epoch": 0.49299940840070994,
"grad_norm": 2.0876184638471567,
"learning_rate": 5.376685006742441e-07,
"loss": 0.4786,
"step": 625
},
{
"epoch": 0.49378820745415103,
"grad_norm": 2.4914185641914925,
"learning_rate": 5.363939019173355e-07,
"loss": 0.4977,
"step": 626
},
{
"epoch": 0.4945770065075922,
"grad_norm": 2.233311934897169,
"learning_rate": 5.351190653535749e-07,
"loss": 0.5191,
"step": 627
},
{
"epoch": 0.4953658055610333,
"grad_norm": 2.152955212138101,
"learning_rate": 5.338439993130637e-07,
"loss": 0.5315,
"step": 628
},
{
"epoch": 0.49615460461447447,
"grad_norm": 1.947932046219484,
"learning_rate": 5.325687121274016e-07,
"loss": 0.4539,
"step": 629
},
{
"epoch": 0.4969434036679156,
"grad_norm": 2.0555163710857274,
"learning_rate": 5.312932121296339e-07,
"loss": 0.5317,
"step": 630
},
{
"epoch": 0.4977322027213567,
"grad_norm": 2.3990319617722276,
"learning_rate": 5.300175076541964e-07,
"loss": 0.4641,
"step": 631
},
{
"epoch": 0.49852100177479786,
"grad_norm": 2.5686914552493856,
"learning_rate": 5.28741607036861e-07,
"loss": 0.5227,
"step": 632
},
{
"epoch": 0.499309800828239,
"grad_norm": 2.288446348903931,
"learning_rate": 5.274655186146814e-07,
"loss": 0.4517,
"step": 633
},
{
"epoch": 0.5000985998816802,
"grad_norm": 2.0510321195651726,
"learning_rate": 5.261892507259383e-07,
"loss": 0.4908,
"step": 634
},
{
"epoch": 0.5008873989351212,
"grad_norm": 2.151882186552225,
"learning_rate": 5.249128117100847e-07,
"loss": 0.4438,
"step": 635
},
{
"epoch": 0.5016761979885624,
"grad_norm": 2.2949744231412876,
"learning_rate": 5.236362099076928e-07,
"loss": 0.5021,
"step": 636
},
{
"epoch": 0.5024649970420035,
"grad_norm": 2.4837568871848585,
"learning_rate": 5.223594536603972e-07,
"loss": 0.4808,
"step": 637
},
{
"epoch": 0.5032537960954447,
"grad_norm": 2.06442493205716,
"learning_rate": 5.210825513108429e-07,
"loss": 0.4866,
"step": 638
},
{
"epoch": 0.5040425951488858,
"grad_norm": 2.216896222718085,
"learning_rate": 5.198055112026288e-07,
"loss": 0.4879,
"step": 639
},
{
"epoch": 0.5048313942023269,
"grad_norm": 2.344315849313264,
"learning_rate": 5.185283416802538e-07,
"loss": 0.4985,
"step": 640
},
{
"epoch": 0.5056201932557681,
"grad_norm": 2.0347114541468057,
"learning_rate": 5.172510510890634e-07,
"loss": 0.507,
"step": 641
},
{
"epoch": 0.5064089923092092,
"grad_norm": 2.0500121468117523,
"learning_rate": 5.159736477751933e-07,
"loss": 0.4968,
"step": 642
},
{
"epoch": 0.5071977913626504,
"grad_norm": 2.269982710071607,
"learning_rate": 5.14696140085516e-07,
"loss": 0.5184,
"step": 643
},
{
"epoch": 0.5079865904160915,
"grad_norm": 2.2756666323317365,
"learning_rate": 5.134185363675862e-07,
"loss": 0.4782,
"step": 644
},
{
"epoch": 0.5087753894695326,
"grad_norm": 2.2501670916838603,
"learning_rate": 5.121408449695856e-07,
"loss": 0.4772,
"step": 645
},
{
"epoch": 0.5095641885229738,
"grad_norm": 2.2996194338211784,
"learning_rate": 5.108630742402695e-07,
"loss": 0.4783,
"step": 646
},
{
"epoch": 0.5103529875764149,
"grad_norm": 2.172209742899572,
"learning_rate": 5.09585232528911e-07,
"loss": 0.5221,
"step": 647
},
{
"epoch": 0.5111417866298561,
"grad_norm": 2.5114347599301783,
"learning_rate": 5.083073281852476e-07,
"loss": 0.4742,
"step": 648
},
{
"epoch": 0.5119305856832972,
"grad_norm": 2.236564807981726,
"learning_rate": 5.070293695594253e-07,
"loss": 0.4581,
"step": 649
},
{
"epoch": 0.5127193847367383,
"grad_norm": 2.4260803567229803,
"learning_rate": 5.057513650019452e-07,
"loss": 0.5139,
"step": 650
},
{
"epoch": 0.5135081837901795,
"grad_norm": 2.233812317045757,
"learning_rate": 5.044733228636087e-07,
"loss": 0.4705,
"step": 651
},
{
"epoch": 0.5142969828436206,
"grad_norm": 2.3736477576440205,
"learning_rate": 5.031952514954626e-07,
"loss": 0.4782,
"step": 652
},
{
"epoch": 0.5150857818970618,
"grad_norm": 2.024685109132759,
"learning_rate": 5.019171592487445e-07,
"loss": 0.4676,
"step": 653
},
{
"epoch": 0.5158745809505029,
"grad_norm": 2.4380432452960235,
"learning_rate": 5.006390544748287e-07,
"loss": 0.4849,
"step": 654
},
{
"epoch": 0.516663380003944,
"grad_norm": 1.9733926549751586,
"learning_rate": 4.993609455251713e-07,
"loss": 0.5258,
"step": 655
},
{
"epoch": 0.5174521790573852,
"grad_norm": 2.2159871492127414,
"learning_rate": 4.980828407512555e-07,
"loss": 0.4611,
"step": 656
},
{
"epoch": 0.5182409781108263,
"grad_norm": 2.141892827424442,
"learning_rate": 4.968047485045373e-07,
"loss": 0.4737,
"step": 657
},
{
"epoch": 0.5190297771642673,
"grad_norm": 2.118748250850451,
"learning_rate": 4.955266771363912e-07,
"loss": 0.4902,
"step": 658
},
{
"epoch": 0.5198185762177086,
"grad_norm": 2.1094830143706265,
"learning_rate": 4.942486349980548e-07,
"loss": 0.5406,
"step": 659
},
{
"epoch": 0.5206073752711496,
"grad_norm": 2.0922130758110753,
"learning_rate": 4.929706304405747e-07,
"loss": 0.5032,
"step": 660
},
{
"epoch": 0.5213961743245908,
"grad_norm": 2.1795332060337147,
"learning_rate": 4.916926718147523e-07,
"loss": 0.4749,
"step": 661
},
{
"epoch": 0.5221849733780319,
"grad_norm": 2.2777362806394827,
"learning_rate": 4.904147674710889e-07,
"loss": 0.4901,
"step": 662
},
{
"epoch": 0.522973772431473,
"grad_norm": 2.2205543429268637,
"learning_rate": 4.891369257597304e-07,
"loss": 0.4709,
"step": 663
},
{
"epoch": 0.5237625714849142,
"grad_norm": 1.9809376025902217,
"learning_rate": 4.878591550304143e-07,
"loss": 0.5032,
"step": 664
},
{
"epoch": 0.5245513705383553,
"grad_norm": 2.1672049794048647,
"learning_rate": 4.865814636324141e-07,
"loss": 0.4632,
"step": 665
},
{
"epoch": 0.5253401695917965,
"grad_norm": 1.9360984730429385,
"learning_rate": 4.85303859914484e-07,
"loss": 0.4888,
"step": 666
},
{
"epoch": 0.5261289686452376,
"grad_norm": 2.169628074481318,
"learning_rate": 4.840263522248067e-07,
"loss": 0.5146,
"step": 667
},
{
"epoch": 0.5269177676986787,
"grad_norm": 2.052231504962686,
"learning_rate": 4.827489489109367e-07,
"loss": 0.4972,
"step": 668
},
{
"epoch": 0.5277065667521199,
"grad_norm": 2.096109396084802,
"learning_rate": 4.814716583197462e-07,
"loss": 0.5114,
"step": 669
},
{
"epoch": 0.528495365805561,
"grad_norm": 2.019844933993389,
"learning_rate": 4.801944887973714e-07,
"loss": 0.4643,
"step": 670
},
{
"epoch": 0.5292841648590022,
"grad_norm": 2.176179065926723,
"learning_rate": 4.789174486891573e-07,
"loss": 0.4481,
"step": 671
},
{
"epoch": 0.5300729639124433,
"grad_norm": 2.3555707826584436,
"learning_rate": 4.776405463396028e-07,
"loss": 0.5262,
"step": 672
},
{
"epoch": 0.5308617629658844,
"grad_norm": 1.9381703662021565,
"learning_rate": 4.763637900923073e-07,
"loss": 0.4548,
"step": 673
},
{
"epoch": 0.5316505620193256,
"grad_norm": 2.0253011565795083,
"learning_rate": 4.7508718828991515e-07,
"loss": 0.498,
"step": 674
},
{
"epoch": 0.5324393610727667,
"grad_norm": 2.073767506839587,
"learning_rate": 4.738107492740619e-07,
"loss": 0.4656,
"step": 675
},
{
"epoch": 0.5332281601262079,
"grad_norm": 2.1886731625904448,
"learning_rate": 4.725344813853186e-07,
"loss": 0.4781,
"step": 676
},
{
"epoch": 0.534016959179649,
"grad_norm": 2.1490532547222254,
"learning_rate": 4.712583929631389e-07,
"loss": 0.5011,
"step": 677
},
{
"epoch": 0.5348057582330901,
"grad_norm": 2.5288054933402244,
"learning_rate": 4.699824923458037e-07,
"loss": 0.4897,
"step": 678
},
{
"epoch": 0.5355945572865313,
"grad_norm": 2.149769020383694,
"learning_rate": 4.6870678787036623e-07,
"loss": 0.4741,
"step": 679
},
{
"epoch": 0.5363833563399724,
"grad_norm": 2.4000488591538938,
"learning_rate": 4.674312878725984e-07,
"loss": 0.4775,
"step": 680
},
{
"epoch": 0.5371721553934136,
"grad_norm": 2.4435610982076654,
"learning_rate": 4.6615600068693645e-07,
"loss": 0.4679,
"step": 681
},
{
"epoch": 0.5379609544468547,
"grad_norm": 2.0830730619207967,
"learning_rate": 4.64880934646425e-07,
"loss": 0.456,
"step": 682
},
{
"epoch": 0.5387497535002957,
"grad_norm": 1.9628230251486485,
"learning_rate": 4.6360609808266453e-07,
"loss": 0.5176,
"step": 683
},
{
"epoch": 0.539538552553737,
"grad_norm": 2.3671900121707754,
"learning_rate": 4.623314993257559e-07,
"loss": 0.5033,
"step": 684
},
{
"epoch": 0.540327351607178,
"grad_norm": 1.997959545102225,
"learning_rate": 4.6105714670424595e-07,
"loss": 0.467,
"step": 685
},
{
"epoch": 0.5411161506606192,
"grad_norm": 2.143381964410816,
"learning_rate": 4.59783048545073e-07,
"loss": 0.5165,
"step": 686
},
{
"epoch": 0.5419049497140603,
"grad_norm": 2.210179705116558,
"learning_rate": 4.585092131735131e-07,
"loss": 0.5208,
"step": 687
},
{
"epoch": 0.5426937487675014,
"grad_norm": 2.3011727007542038,
"learning_rate": 4.5723564891312494e-07,
"loss": 0.5027,
"step": 688
},
{
"epoch": 0.5434825478209426,
"grad_norm": 2.2050119620734656,
"learning_rate": 4.5596236408569566e-07,
"loss": 0.4907,
"step": 689
},
{
"epoch": 0.5442713468743837,
"grad_norm": 1.9932121404423098,
"learning_rate": 4.5468936701118654e-07,
"loss": 0.4663,
"step": 690
},
{
"epoch": 0.5450601459278249,
"grad_norm": 2.239098467100536,
"learning_rate": 4.5341666600767875e-07,
"loss": 0.4191,
"step": 691
},
{
"epoch": 0.545848944981266,
"grad_norm": 1.932453389920233,
"learning_rate": 4.5214426939131874e-07,
"loss": 0.5017,
"step": 692
},
{
"epoch": 0.5466377440347071,
"grad_norm": 2.203107061774875,
"learning_rate": 4.5087218547626386e-07,
"loss": 0.4774,
"step": 693
},
{
"epoch": 0.5474265430881483,
"grad_norm": 2.330734535010093,
"learning_rate": 4.496004225746286e-07,
"loss": 0.4848,
"step": 694
},
{
"epoch": 0.5482153421415894,
"grad_norm": 2.2897132850121165,
"learning_rate": 4.483289889964297e-07,
"loss": 0.5239,
"step": 695
},
{
"epoch": 0.5490041411950306,
"grad_norm": 2.156815955821161,
"learning_rate": 4.470578930495316e-07,
"loss": 0.475,
"step": 696
},
{
"epoch": 0.5497929402484717,
"grad_norm": 2.0607330034972775,
"learning_rate": 4.4578714303959316e-07,
"loss": 0.521,
"step": 697
},
{
"epoch": 0.5505817393019128,
"grad_norm": 2.101290421673513,
"learning_rate": 4.445167472700127e-07,
"loss": 0.4988,
"step": 698
},
{
"epoch": 0.551370538355354,
"grad_norm": 2.294264904561355,
"learning_rate": 4.432467140418734e-07,
"loss": 0.4497,
"step": 699
},
{
"epoch": 0.5521593374087951,
"grad_norm": 2.266187870629658,
"learning_rate": 4.419770516538899e-07,
"loss": 0.5219,
"step": 700
},
{
"epoch": 0.5529481364622363,
"grad_norm": 2.0414790068346593,
"learning_rate": 4.407077684023539e-07,
"loss": 0.4823,
"step": 701
},
{
"epoch": 0.5537369355156774,
"grad_norm": 2.588701920660314,
"learning_rate": 4.394388725810789e-07,
"loss": 0.4481,
"step": 702
},
{
"epoch": 0.5545257345691185,
"grad_norm": 2.3664198854694214,
"learning_rate": 4.3817037248134774e-07,
"loss": 0.464,
"step": 703
},
{
"epoch": 0.5553145336225597,
"grad_norm": 2.326106741037307,
"learning_rate": 4.369022763918572e-07,
"loss": 0.4698,
"step": 704
},
{
"epoch": 0.5561033326760008,
"grad_norm": 2.237029648099948,
"learning_rate": 4.356345925986637e-07,
"loss": 0.4869,
"step": 705
},
{
"epoch": 0.556892131729442,
"grad_norm": 2.150855053027392,
"learning_rate": 4.343673293851302e-07,
"loss": 0.5109,
"step": 706
},
{
"epoch": 0.557680930782883,
"grad_norm": 2.3165345016529475,
"learning_rate": 4.331004950318713e-07,
"loss": 0.5149,
"step": 707
},
{
"epoch": 0.5584697298363241,
"grad_norm": 2.0053640092076552,
"learning_rate": 4.3183409781669954e-07,
"loss": 0.488,
"step": 708
},
{
"epoch": 0.5592585288897653,
"grad_norm": 2.0907676274989626,
"learning_rate": 4.3056814601457047e-07,
"loss": 0.4838,
"step": 709
},
{
"epoch": 0.5600473279432064,
"grad_norm": 2.3402176955361322,
"learning_rate": 4.293026478975297e-07,
"loss": 0.5099,
"step": 710
},
{
"epoch": 0.5608361269966476,
"grad_norm": 1.9529607287176258,
"learning_rate": 4.280376117346585e-07,
"loss": 0.4773,
"step": 711
},
{
"epoch": 0.5616249260500887,
"grad_norm": 1.944305764783215,
"learning_rate": 4.267730457920189e-07,
"loss": 0.4852,
"step": 712
},
{
"epoch": 0.5624137251035298,
"grad_norm": 1.9149430559028453,
"learning_rate": 4.255089583326011e-07,
"loss": 0.4837,
"step": 713
},
{
"epoch": 0.563202524156971,
"grad_norm": 2.3361958735004404,
"learning_rate": 4.2424535761626866e-07,
"loss": 0.4982,
"step": 714
},
{
"epoch": 0.5639913232104121,
"grad_norm": 2.0417984955931723,
"learning_rate": 4.229822518997042e-07,
"loss": 0.5028,
"step": 715
},
{
"epoch": 0.5647801222638533,
"grad_norm": 1.9342378154742026,
"learning_rate": 4.217196494363563e-07,
"loss": 0.4878,
"step": 716
},
{
"epoch": 0.5655689213172944,
"grad_norm": 2.086109610827381,
"learning_rate": 4.204575584763852e-07,
"loss": 0.5042,
"step": 717
},
{
"epoch": 0.5663577203707355,
"grad_norm": 2.2267517675976896,
"learning_rate": 4.191959872666085e-07,
"loss": 0.4976,
"step": 718
},
{
"epoch": 0.5671465194241767,
"grad_norm": 2.186668449180831,
"learning_rate": 4.1793494405044777e-07,
"loss": 0.5078,
"step": 719
},
{
"epoch": 0.5679353184776178,
"grad_norm": 2.1335695201044627,
"learning_rate": 4.1667443706787475e-07,
"loss": 0.5006,
"step": 720
},
{
"epoch": 0.568724117531059,
"grad_norm": 2.373112364869574,
"learning_rate": 4.154144745553571e-07,
"loss": 0.4937,
"step": 721
},
{
"epoch": 0.5695129165845001,
"grad_norm": 2.1293714432679667,
"learning_rate": 4.1415506474580477e-07,
"loss": 0.4324,
"step": 722
},
{
"epoch": 0.5703017156379412,
"grad_norm": 2.0753846048286517,
"learning_rate": 4.1289621586851606e-07,
"loss": 0.4673,
"step": 723
},
{
"epoch": 0.5710905146913824,
"grad_norm": 2.224648127912629,
"learning_rate": 4.1163793614912447e-07,
"loss": 0.4604,
"step": 724
},
{
"epoch": 0.5718793137448235,
"grad_norm": 2.4764881585420766,
"learning_rate": 4.103802338095441e-07,
"loss": 0.4938,
"step": 725
},
{
"epoch": 0.5726681127982647,
"grad_norm": 2.237576220001247,
"learning_rate": 4.0912311706791624e-07,
"loss": 0.4994,
"step": 726
},
{
"epoch": 0.5734569118517058,
"grad_norm": 2.111747698595159,
"learning_rate": 4.0786659413855616e-07,
"loss": 0.4804,
"step": 727
},
{
"epoch": 0.5742457109051469,
"grad_norm": 1.984998758841751,
"learning_rate": 4.066106732318986e-07,
"loss": 0.4481,
"step": 728
},
{
"epoch": 0.5750345099585881,
"grad_norm": 2.137027509941071,
"learning_rate": 4.0535536255444456e-07,
"loss": 0.4969,
"step": 729
},
{
"epoch": 0.5758233090120292,
"grad_norm": 2.1010836408062987,
"learning_rate": 4.041006703087079e-07,
"loss": 0.4614,
"step": 730
},
{
"epoch": 0.5766121080654704,
"grad_norm": 2.2375711143666073,
"learning_rate": 4.0284660469316163e-07,
"loss": 0.5086,
"step": 731
},
{
"epoch": 0.5774009071189115,
"grad_norm": 2.4092184606984075,
"learning_rate": 4.015931739021834e-07,
"loss": 0.4904,
"step": 732
},
{
"epoch": 0.5781897061723525,
"grad_norm": 2.3466393140740007,
"learning_rate": 4.003403861260036e-07,
"loss": 0.4366,
"step": 733
},
{
"epoch": 0.5789785052257937,
"grad_norm": 2.1454637477784964,
"learning_rate": 3.990882495506508e-07,
"loss": 0.4812,
"step": 734
},
{
"epoch": 0.5797673042792348,
"grad_norm": 2.0705477563294177,
"learning_rate": 3.97836772357898e-07,
"loss": 0.4709,
"step": 735
},
{
"epoch": 0.580556103332676,
"grad_norm": 2.0680145780527055,
"learning_rate": 3.965859627252103e-07,
"loss": 0.4502,
"step": 736
},
{
"epoch": 0.5813449023861171,
"grad_norm": 2.2490305551401786,
"learning_rate": 3.953358288256905e-07,
"loss": 0.4882,
"step": 737
},
{
"epoch": 0.5821337014395582,
"grad_norm": 2.280477900384498,
"learning_rate": 3.940863788280256e-07,
"loss": 0.4585,
"step": 738
},
{
"epoch": 0.5829225004929994,
"grad_norm": 2.0728466785387116,
"learning_rate": 3.928376208964344e-07,
"loss": 0.518,
"step": 739
},
{
"epoch": 0.5837112995464405,
"grad_norm": 2.3919902854460777,
"learning_rate": 3.915895631906134e-07,
"loss": 0.4962,
"step": 740
},
{
"epoch": 0.5845000985998817,
"grad_norm": 2.2915017645804765,
"learning_rate": 3.9034221386568384e-07,
"loss": 0.4624,
"step": 741
},
{
"epoch": 0.5852888976533228,
"grad_norm": 2.2959570177979383,
"learning_rate": 3.890955810721374e-07,
"loss": 0.4711,
"step": 742
},
{
"epoch": 0.5860776967067639,
"grad_norm": 2.272399374837785,
"learning_rate": 3.878496729557847e-07,
"loss": 0.4863,
"step": 743
},
{
"epoch": 0.5868664957602051,
"grad_norm": 2.783770313690681,
"learning_rate": 3.8660449765770095e-07,
"loss": 0.4941,
"step": 744
},
{
"epoch": 0.5876552948136462,
"grad_norm": 2.321494468971465,
"learning_rate": 3.8536006331417243e-07,
"loss": 0.4628,
"step": 745
},
{
"epoch": 0.5884440938670874,
"grad_norm": 2.169892048155267,
"learning_rate": 3.841163780566443e-07,
"loss": 0.496,
"step": 746
},
{
"epoch": 0.5892328929205285,
"grad_norm": 1.8981664878810656,
"learning_rate": 3.8287345001166704e-07,
"loss": 0.494,
"step": 747
},
{
"epoch": 0.5900216919739696,
"grad_norm": 2.0403134859400787,
"learning_rate": 3.8163128730084294e-07,
"loss": 0.4822,
"step": 748
},
{
"epoch": 0.5908104910274108,
"grad_norm": 2.197113256493988,
"learning_rate": 3.8038989804077366e-07,
"loss": 0.5308,
"step": 749
},
{
"epoch": 0.5915992900808519,
"grad_norm": 2.23173209600613,
"learning_rate": 3.7914929034300707e-07,
"loss": 0.5053,
"step": 750
},
{
"epoch": 0.5923880891342931,
"grad_norm": 2.1103764135282645,
"learning_rate": 3.779094723139839e-07,
"loss": 0.4954,
"step": 751
},
{
"epoch": 0.5931768881877342,
"grad_norm": 1.9293362680538588,
"learning_rate": 3.7667045205498493e-07,
"loss": 0.4975,
"step": 752
},
{
"epoch": 0.5939656872411753,
"grad_norm": 2.0773223178629543,
"learning_rate": 3.7543223766207836e-07,
"loss": 0.4739,
"step": 753
},
{
"epoch": 0.5947544862946165,
"grad_norm": 2.289131398319478,
"learning_rate": 3.741948372260665e-07,
"loss": 0.4642,
"step": 754
},
{
"epoch": 0.5955432853480576,
"grad_norm": 2.277375723379447,
"learning_rate": 3.72958258832433e-07,
"loss": 0.4572,
"step": 755
},
{
"epoch": 0.5963320844014988,
"grad_norm": 1.9461398440012956,
"learning_rate": 3.7172251056129024e-07,
"loss": 0.4615,
"step": 756
},
{
"epoch": 0.5971208834549399,
"grad_norm": 2.095990827768853,
"learning_rate": 3.704876004873262e-07,
"loss": 0.4706,
"step": 757
},
{
"epoch": 0.597909682508381,
"grad_norm": 2.01172057840265,
"learning_rate": 3.6925353667975216e-07,
"loss": 0.4698,
"step": 758
},
{
"epoch": 0.5986984815618221,
"grad_norm": 2.1574381277155372,
"learning_rate": 3.680203272022493e-07,
"loss": 0.4756,
"step": 759
},
{
"epoch": 0.5994872806152632,
"grad_norm": 2.280359614341495,
"learning_rate": 3.667879801129167e-07,
"loss": 0.4652,
"step": 760
},
{
"epoch": 0.6002760796687044,
"grad_norm": 2.157115994127388,
"learning_rate": 3.6555650346421817e-07,
"loss": 0.4732,
"step": 761
},
{
"epoch": 0.6010648787221455,
"grad_norm": 2.240622674181541,
"learning_rate": 3.6432590530292995e-07,
"loss": 0.4952,
"step": 762
},
{
"epoch": 0.6018536777755866,
"grad_norm": 2.4133052486025184,
"learning_rate": 3.6309619367008805e-07,
"loss": 0.4438,
"step": 763
},
{
"epoch": 0.6026424768290278,
"grad_norm": 1.9774169250299347,
"learning_rate": 3.618673766009359e-07,
"loss": 0.4899,
"step": 764
},
{
"epoch": 0.6034312758824689,
"grad_norm": 2.2086961317412133,
"learning_rate": 3.6063946212487085e-07,
"loss": 0.5033,
"step": 765
},
{
"epoch": 0.6042200749359101,
"grad_norm": 1.9737632269332888,
"learning_rate": 3.594124582653933e-07,
"loss": 0.481,
"step": 766
},
{
"epoch": 0.6050088739893512,
"grad_norm": 2.2821066762407916,
"learning_rate": 3.5818637304005326e-07,
"loss": 0.54,
"step": 767
},
{
"epoch": 0.6057976730427923,
"grad_norm": 2.190217246258578,
"learning_rate": 3.569612144603978e-07,
"loss": 0.4902,
"step": 768
},
{
"epoch": 0.6065864720962335,
"grad_norm": 2.0486729469185345,
"learning_rate": 3.557369905319194e-07,
"loss": 0.464,
"step": 769
},
{
"epoch": 0.6073752711496746,
"grad_norm": 2.4941882932215567,
"learning_rate": 3.545137092540035e-07,
"loss": 0.4576,
"step": 770
},
{
"epoch": 0.6081640702031158,
"grad_norm": 2.4823606272679215,
"learning_rate": 3.532913786198752e-07,
"loss": 0.4665,
"step": 771
},
{
"epoch": 0.6089528692565569,
"grad_norm": 2.3332619798624066,
"learning_rate": 3.5207000661654864e-07,
"loss": 0.4861,
"step": 772
},
{
"epoch": 0.609741668309998,
"grad_norm": 2.1453139151049,
"learning_rate": 3.5084960122477367e-07,
"loss": 0.5055,
"step": 773
},
{
"epoch": 0.6105304673634392,
"grad_norm": 2.0389292264091434,
"learning_rate": 3.4963017041898455e-07,
"loss": 0.4591,
"step": 774
},
{
"epoch": 0.6113192664168803,
"grad_norm": 2.0682537033876662,
"learning_rate": 3.4841172216724653e-07,
"loss": 0.4905,
"step": 775
},
{
"epoch": 0.6121080654703215,
"grad_norm": 2.1285891547482234,
"learning_rate": 3.471942644312051e-07,
"loss": 0.4936,
"step": 776
},
{
"epoch": 0.6128968645237626,
"grad_norm": 2.162574539659009,
"learning_rate": 3.459778051660337e-07,
"loss": 0.4973,
"step": 777
},
{
"epoch": 0.6136856635772037,
"grad_norm": 2.068931705681872,
"learning_rate": 3.4476235232038077e-07,
"loss": 0.4709,
"step": 778
},
{
"epoch": 0.6144744626306449,
"grad_norm": 2.2395459394193002,
"learning_rate": 3.435479138363193e-07,
"loss": 0.4944,
"step": 779
},
{
"epoch": 0.615263261684086,
"grad_norm": 1.9697728266212382,
"learning_rate": 3.4233449764929393e-07,
"loss": 0.464,
"step": 780
},
{
"epoch": 0.6160520607375272,
"grad_norm": 2.1647313521627503,
"learning_rate": 3.411221116880691e-07,
"loss": 0.5129,
"step": 781
},
{
"epoch": 0.6168408597909683,
"grad_norm": 2.2180582230650803,
"learning_rate": 3.399107638746778e-07,
"loss": 0.4611,
"step": 782
},
{
"epoch": 0.6176296588444093,
"grad_norm": 2.192703270944367,
"learning_rate": 3.387004621243696e-07,
"loss": 0.4869,
"step": 783
},
{
"epoch": 0.6184184578978505,
"grad_norm": 1.921941096770739,
"learning_rate": 3.3749121434555847e-07,
"loss": 0.5235,
"step": 784
},
{
"epoch": 0.6192072569512916,
"grad_norm": 2.105224565679266,
"learning_rate": 3.362830284397716e-07,
"loss": 0.5143,
"step": 785
},
{
"epoch": 0.6199960560047328,
"grad_norm": 2.20760686350849,
"learning_rate": 3.3507591230159783e-07,
"loss": 0.5069,
"step": 786
},
{
"epoch": 0.6207848550581739,
"grad_norm": 2.2021780573248413,
"learning_rate": 3.338698738186356e-07,
"loss": 0.4936,
"step": 787
},
{
"epoch": 0.621573654111615,
"grad_norm": 2.316966960695151,
"learning_rate": 3.3266492087144197e-07,
"loss": 0.4775,
"step": 788
},
{
"epoch": 0.6223624531650562,
"grad_norm": 2.309250096944103,
"learning_rate": 3.314610613334804e-07,
"loss": 0.5206,
"step": 789
},
{
"epoch": 0.6231512522184973,
"grad_norm": 2.262806585245431,
"learning_rate": 3.3025830307107033e-07,
"loss": 0.4465,
"step": 790
},
{
"epoch": 0.6239400512719385,
"grad_norm": 2.0958281323459422,
"learning_rate": 3.290566539433347e-07,
"loss": 0.4622,
"step": 791
},
{
"epoch": 0.6247288503253796,
"grad_norm": 2.0021200656592812,
"learning_rate": 3.278561218021492e-07,
"loss": 0.51,
"step": 792
},
{
"epoch": 0.6255176493788207,
"grad_norm": 2.200808184368059,
"learning_rate": 3.266567144920911e-07,
"loss": 0.4682,
"step": 793
},
{
"epoch": 0.6263064484322619,
"grad_norm": 2.2098028132021983,
"learning_rate": 3.2545843985038736e-07,
"loss": 0.4353,
"step": 794
},
{
"epoch": 0.627095247485703,
"grad_norm": 2.063660823792188,
"learning_rate": 3.2426130570686407e-07,
"loss": 0.4759,
"step": 795
},
{
"epoch": 0.6278840465391442,
"grad_norm": 2.239406542428139,
"learning_rate": 3.23065319883895e-07,
"loss": 0.5136,
"step": 796
},
{
"epoch": 0.6286728455925853,
"grad_norm": 2.161750030580745,
"learning_rate": 3.218704901963505e-07,
"loss": 0.477,
"step": 797
},
{
"epoch": 0.6294616446460264,
"grad_norm": 2.1795280689676733,
"learning_rate": 3.2067682445154623e-07,
"loss": 0.5041,
"step": 798
},
{
"epoch": 0.6302504436994676,
"grad_norm": 2.198956221701037,
"learning_rate": 3.1948433044919257e-07,
"loss": 0.4734,
"step": 799
},
{
"epoch": 0.6310392427529087,
"grad_norm": 2.1098107101478605,
"learning_rate": 3.182930159813435e-07,
"loss": 0.4707,
"step": 800
},
{
"epoch": 0.6318280418063499,
"grad_norm": 1.8341365868983686,
"learning_rate": 3.171028888323454e-07,
"loss": 0.5045,
"step": 801
},
{
"epoch": 0.632616840859791,
"grad_norm": 2.174738778296802,
"learning_rate": 3.1591395677878653e-07,
"loss": 0.5067,
"step": 802
},
{
"epoch": 0.6334056399132321,
"grad_norm": 2.026376825474118,
"learning_rate": 3.1472622758944635e-07,
"loss": 0.5089,
"step": 803
},
{
"epoch": 0.6341944389666733,
"grad_norm": 2.2131278914724013,
"learning_rate": 3.1353970902524385e-07,
"loss": 0.4786,
"step": 804
},
{
"epoch": 0.6349832380201144,
"grad_norm": 2.357212012340458,
"learning_rate": 3.1235440883918806e-07,
"loss": 0.5104,
"step": 805
},
{
"epoch": 0.6357720370735556,
"grad_norm": 2.3591976827393615,
"learning_rate": 3.1117033477632676e-07,
"loss": 0.4188,
"step": 806
},
{
"epoch": 0.6365608361269967,
"grad_norm": 2.038748583907236,
"learning_rate": 3.09987494573696e-07,
"loss": 0.4671,
"step": 807
},
{
"epoch": 0.6373496351804377,
"grad_norm": 2.097590633191528,
"learning_rate": 3.0880589596026887e-07,
"loss": 0.4587,
"step": 808
},
{
"epoch": 0.638138434233879,
"grad_norm": 2.0236518331713116,
"learning_rate": 3.076255466569062e-07,
"loss": 0.5152,
"step": 809
},
{
"epoch": 0.63892723328732,
"grad_norm": 2.230718226437944,
"learning_rate": 3.0644645437630566e-07,
"loss": 0.4979,
"step": 810
},
{
"epoch": 0.6397160323407612,
"grad_norm": 2.231310523111879,
"learning_rate": 3.052686268229505e-07,
"loss": 0.5111,
"step": 811
},
{
"epoch": 0.6405048313942023,
"grad_norm": 2.182967357329115,
"learning_rate": 3.0409207169306064e-07,
"loss": 0.426,
"step": 812
},
{
"epoch": 0.6412936304476434,
"grad_norm": 2.4038634053361405,
"learning_rate": 3.0291679667454134e-07,
"loss": 0.4704,
"step": 813
},
{
"epoch": 0.6420824295010846,
"grad_norm": 1.9690412999491156,
"learning_rate": 3.0174280944693315e-07,
"loss": 0.4968,
"step": 814
},
{
"epoch": 0.6428712285545257,
"grad_norm": 2.1189053168702223,
"learning_rate": 3.005701176813622e-07,
"loss": 0.4803,
"step": 815
},
{
"epoch": 0.6436600276079669,
"grad_norm": 2.0342387311212535,
"learning_rate": 2.993987290404897e-07,
"loss": 0.4983,
"step": 816
},
{
"epoch": 0.644448826661408,
"grad_norm": 2.20081222413284,
"learning_rate": 2.982286511784615e-07,
"loss": 0.4912,
"step": 817
},
{
"epoch": 0.6452376257148491,
"grad_norm": 1.8983850283065198,
"learning_rate": 2.970598917408591e-07,
"loss": 0.5105,
"step": 818
},
{
"epoch": 0.6460264247682903,
"grad_norm": 2.4075158292027456,
"learning_rate": 2.958924583646487e-07,
"loss": 0.4828,
"step": 819
},
{
"epoch": 0.6468152238217314,
"grad_norm": 1.9533586249499895,
"learning_rate": 2.9472635867813184e-07,
"loss": 0.4891,
"step": 820
},
{
"epoch": 0.6476040228751726,
"grad_norm": 2.254163317267339,
"learning_rate": 2.935616003008953e-07,
"loss": 0.4757,
"step": 821
},
{
"epoch": 0.6483928219286137,
"grad_norm": 2.1729334172132244,
"learning_rate": 2.923981908437612e-07,
"loss": 0.4983,
"step": 822
},
{
"epoch": 0.6491816209820548,
"grad_norm": 1.9650080179600111,
"learning_rate": 2.9123613790873815e-07,
"loss": 0.4864,
"step": 823
},
{
"epoch": 0.649970420035496,
"grad_norm": 2.4719651193385332,
"learning_rate": 2.900754490889702e-07,
"loss": 0.48,
"step": 824
},
{
"epoch": 0.6507592190889371,
"grad_norm": 2.168775496986839,
"learning_rate": 2.889161319686881e-07,
"loss": 0.4598,
"step": 825
},
{
"epoch": 0.6515480181423783,
"grad_norm": 2.268611827918302,
"learning_rate": 2.8775819412316033e-07,
"loss": 0.4514,
"step": 826
},
{
"epoch": 0.6523368171958194,
"grad_norm": 2.2925150943741013,
"learning_rate": 2.8660164311864133e-07,
"loss": 0.4821,
"step": 827
},
{
"epoch": 0.6531256162492605,
"grad_norm": 2.067769766829148,
"learning_rate": 2.85446486512325e-07,
"loss": 0.4492,
"step": 828
},
{
"epoch": 0.6539144153027017,
"grad_norm": 2.20598797466938,
"learning_rate": 2.8429273185229304e-07,
"loss": 0.4538,
"step": 829
},
{
"epoch": 0.6547032143561428,
"grad_norm": 2.2026185028627334,
"learning_rate": 2.8314038667746733e-07,
"loss": 0.4639,
"step": 830
},
{
"epoch": 0.655492013409584,
"grad_norm": 2.3013781163509455,
"learning_rate": 2.819894585175584e-07,
"loss": 0.5017,
"step": 831
},
{
"epoch": 0.656280812463025,
"grad_norm": 2.2040011491554465,
"learning_rate": 2.808399548930192e-07,
"loss": 0.4979,
"step": 832
},
{
"epoch": 0.6570696115164661,
"grad_norm": 2.1387483968938596,
"learning_rate": 2.796918833149935e-07,
"loss": 0.4892,
"step": 833
},
{
"epoch": 0.6578584105699073,
"grad_norm": 2.1372273434476194,
"learning_rate": 2.7854525128526804e-07,
"loss": 0.4658,
"step": 834
},
{
"epoch": 0.6586472096233484,
"grad_norm": 2.2017808020811755,
"learning_rate": 2.77400066296223e-07,
"loss": 0.4886,
"step": 835
},
{
"epoch": 0.6594360086767896,
"grad_norm": 2.073896623186491,
"learning_rate": 2.762563358307833e-07,
"loss": 0.4783,
"step": 836
},
{
"epoch": 0.6602248077302307,
"grad_norm": 2.162730773723987,
"learning_rate": 2.751140673623696e-07,
"loss": 0.4781,
"step": 837
},
{
"epoch": 0.6610136067836718,
"grad_norm": 2.468700392107243,
"learning_rate": 2.739732683548499e-07,
"loss": 0.4396,
"step": 838
},
{
"epoch": 0.661802405837113,
"grad_norm": 2.345085257931882,
"learning_rate": 2.728339462624899e-07,
"loss": 0.5001,
"step": 839
},
{
"epoch": 0.6625912048905541,
"grad_norm": 2.0785472389777335,
"learning_rate": 2.7169610852990487e-07,
"loss": 0.5118,
"step": 840
},
{
"epoch": 0.6633800039439953,
"grad_norm": 2.068116554324134,
"learning_rate": 2.7055976259201096e-07,
"loss": 0.5059,
"step": 841
},
{
"epoch": 0.6641688029974364,
"grad_norm": 2.6622950688870928,
"learning_rate": 2.694249158739764e-07,
"loss": 0.4655,
"step": 842
},
{
"epoch": 0.6649576020508775,
"grad_norm": 2.038920593218503,
"learning_rate": 2.682915757911739e-07,
"loss": 0.5047,
"step": 843
},
{
"epoch": 0.6657464011043187,
"grad_norm": 1.9686666407311277,
"learning_rate": 2.6715974974913003e-07,
"loss": 0.5189,
"step": 844
},
{
"epoch": 0.6665352001577598,
"grad_norm": 2.0525912974714378,
"learning_rate": 2.660294451434796e-07,
"loss": 0.5339,
"step": 845
},
{
"epoch": 0.6673239992112009,
"grad_norm": 2.1926107324242805,
"learning_rate": 2.649006693599153e-07,
"loss": 0.4821,
"step": 846
},
{
"epoch": 0.6681127982646421,
"grad_norm": 2.358738451676426,
"learning_rate": 2.637734297741401e-07,
"loss": 0.4717,
"step": 847
},
{
"epoch": 0.6689015973180832,
"grad_norm": 2.1626623107487255,
"learning_rate": 2.6264773375181916e-07,
"loss": 0.4687,
"step": 848
},
{
"epoch": 0.6696903963715244,
"grad_norm": 2.094271429296511,
"learning_rate": 2.6152358864853217e-07,
"loss": 0.5117,
"step": 849
},
{
"epoch": 0.6704791954249655,
"grad_norm": 2.2848201352623376,
"learning_rate": 2.6040100180972354e-07,
"loss": 0.4769,
"step": 850
},
{
"epoch": 0.6712679944784066,
"grad_norm": 2.1081320461643553,
"learning_rate": 2.5927998057065664e-07,
"loss": 0.5109,
"step": 851
},
{
"epoch": 0.6720567935318478,
"grad_norm": 2.4747570421182035,
"learning_rate": 2.5816053225636426e-07,
"loss": 0.4331,
"step": 852
},
{
"epoch": 0.6728455925852889,
"grad_norm": 2.3541785223967673,
"learning_rate": 2.570426641816019e-07,
"loss": 0.4919,
"step": 853
},
{
"epoch": 0.6736343916387301,
"grad_norm": 2.1895405717712566,
"learning_rate": 2.5592638365079833e-07,
"loss": 0.4599,
"step": 854
},
{
"epoch": 0.6744231906921712,
"grad_norm": 2.2698187092528976,
"learning_rate": 2.5481169795801004e-07,
"loss": 0.4637,
"step": 855
},
{
"epoch": 0.6752119897456123,
"grad_norm": 2.113503079539194,
"learning_rate": 2.536986143868719e-07,
"loss": 0.4705,
"step": 856
},
{
"epoch": 0.6760007887990535,
"grad_norm": 2.273946319228632,
"learning_rate": 2.5258714021055e-07,
"loss": 0.4596,
"step": 857
},
{
"epoch": 0.6767895878524945,
"grad_norm": 2.161326926247605,
"learning_rate": 2.514772826916943e-07,
"loss": 0.4865,
"step": 858
},
{
"epoch": 0.6775783869059357,
"grad_norm": 2.121940322048116,
"learning_rate": 2.503690490823919e-07,
"loss": 0.5054,
"step": 859
},
{
"epoch": 0.6783671859593768,
"grad_norm": 2.3324267398134424,
"learning_rate": 2.492624466241173e-07,
"loss": 0.4647,
"step": 860
},
{
"epoch": 0.6791559850128179,
"grad_norm": 2.3711520985096652,
"learning_rate": 2.4815748254768827e-07,
"loss": 0.5004,
"step": 861
},
{
"epoch": 0.6799447840662591,
"grad_norm": 2.0453114226822384,
"learning_rate": 2.470541640732157e-07,
"loss": 0.4883,
"step": 862
},
{
"epoch": 0.6807335831197002,
"grad_norm": 2.0204677517701737,
"learning_rate": 2.4595249841005904e-07,
"loss": 0.5089,
"step": 863
},
{
"epoch": 0.6815223821731414,
"grad_norm": 2.118831940491388,
"learning_rate": 2.4485249275677625e-07,
"loss": 0.4928,
"step": 864
},
{
"epoch": 0.6823111812265825,
"grad_norm": 2.0166780797250534,
"learning_rate": 2.4375415430107974e-07,
"loss": 0.4848,
"step": 865
},
{
"epoch": 0.6830999802800236,
"grad_norm": 2.084911705632224,
"learning_rate": 2.4265749021978745e-07,
"loss": 0.5423,
"step": 866
},
{
"epoch": 0.6838887793334648,
"grad_norm": 2.221760619361746,
"learning_rate": 2.4156250767877646e-07,
"loss": 0.5013,
"step": 867
},
{
"epoch": 0.6846775783869059,
"grad_norm": 2.2871304542092856,
"learning_rate": 2.404692138329364e-07,
"loss": 0.482,
"step": 868
},
{
"epoch": 0.6854663774403471,
"grad_norm": 2.179441020761214,
"learning_rate": 2.393776158261226e-07,
"loss": 0.5107,
"step": 869
},
{
"epoch": 0.6862551764937882,
"grad_norm": 2.2090570861102816,
"learning_rate": 2.3828772079110904e-07,
"loss": 0.4814,
"step": 870
},
{
"epoch": 0.6870439755472293,
"grad_norm": 2.430748670368225,
"learning_rate": 2.3719953584954267e-07,
"loss": 0.4888,
"step": 871
},
{
"epoch": 0.6878327746006705,
"grad_norm": 2.344805646743284,
"learning_rate": 2.3611306811189575e-07,
"loss": 0.4266,
"step": 872
},
{
"epoch": 0.6886215736541116,
"grad_norm": 1.949062154623978,
"learning_rate": 2.3502832467741996e-07,
"loss": 0.5055,
"step": 873
},
{
"epoch": 0.6894103727075528,
"grad_norm": 2.335138402244164,
"learning_rate": 2.3394531263410005e-07,
"loss": 0.5246,
"step": 874
},
{
"epoch": 0.6901991717609939,
"grad_norm": 2.0675164974432767,
"learning_rate": 2.328640390586073e-07,
"loss": 0.4637,
"step": 875
},
{
"epoch": 0.690987970814435,
"grad_norm": 2.2388330879358977,
"learning_rate": 2.3178451101625401e-07,
"loss": 0.4531,
"step": 876
},
{
"epoch": 0.6917767698678762,
"grad_norm": 2.037836572040061,
"learning_rate": 2.307067355609455e-07,
"loss": 0.4935,
"step": 877
},
{
"epoch": 0.6925655689213173,
"grad_norm": 2.091235658051316,
"learning_rate": 2.2963071973513655e-07,
"loss": 0.4939,
"step": 878
},
{
"epoch": 0.6933543679747585,
"grad_norm": 2.0426324812937198,
"learning_rate": 2.2855647056978332e-07,
"loss": 0.49,
"step": 879
},
{
"epoch": 0.6941431670281996,
"grad_norm": 2.187972211654667,
"learning_rate": 2.2748399508429834e-07,
"loss": 0.4613,
"step": 880
},
{
"epoch": 0.6949319660816407,
"grad_norm": 2.2374147444124852,
"learning_rate": 2.2641330028650435e-07,
"loss": 0.5121,
"step": 881
},
{
"epoch": 0.6957207651350819,
"grad_norm": 2.187731364813485,
"learning_rate": 2.2534439317258935e-07,
"loss": 0.4766,
"step": 882
},
{
"epoch": 0.696509564188523,
"grad_norm": 2.2144736507098473,
"learning_rate": 2.2427728072705875e-07,
"loss": 0.5494,
"step": 883
},
{
"epoch": 0.6972983632419641,
"grad_norm": 2.255042738197164,
"learning_rate": 2.232119699226927e-07,
"loss": 0.5233,
"step": 884
},
{
"epoch": 0.6980871622954052,
"grad_norm": 2.2903621401614416,
"learning_rate": 2.2214846772049778e-07,
"loss": 0.4664,
"step": 885
},
{
"epoch": 0.6988759613488463,
"grad_norm": 2.1008409460344883,
"learning_rate": 2.2108678106966383e-07,
"loss": 0.4547,
"step": 886
},
{
"epoch": 0.6996647604022875,
"grad_norm": 1.9749273117309982,
"learning_rate": 2.2002691690751595e-07,
"loss": 0.471,
"step": 887
},
{
"epoch": 0.7004535594557286,
"grad_norm": 2.1450015364340396,
"learning_rate": 2.189688821594722e-07,
"loss": 0.4437,
"step": 888
},
{
"epoch": 0.7012423585091698,
"grad_norm": 2.588825760122245,
"learning_rate": 2.1791268373899574e-07,
"loss": 0.4512,
"step": 889
},
{
"epoch": 0.7020311575626109,
"grad_norm": 2.1465049655136346,
"learning_rate": 2.1685832854755115e-07,
"loss": 0.4642,
"step": 890
},
{
"epoch": 0.702819956616052,
"grad_norm": 2.2790933198088066,
"learning_rate": 2.158058234745585e-07,
"loss": 0.5002,
"step": 891
},
{
"epoch": 0.7036087556694932,
"grad_norm": 1.9990935432873584,
"learning_rate": 2.1475517539734966e-07,
"loss": 0.4926,
"step": 892
},
{
"epoch": 0.7043975547229343,
"grad_norm": 2.2681308395280673,
"learning_rate": 2.1370639118112099e-07,
"loss": 0.4513,
"step": 893
},
{
"epoch": 0.7051863537763755,
"grad_norm": 2.1539243352928334,
"learning_rate": 2.126594776788913e-07,
"loss": 0.4843,
"step": 894
},
{
"epoch": 0.7059751528298166,
"grad_norm": 2.172895250910165,
"learning_rate": 2.1161444173145466e-07,
"loss": 0.4409,
"step": 895
},
{
"epoch": 0.7067639518832577,
"grad_norm": 2.3745704139966923,
"learning_rate": 2.1057129016733777e-07,
"loss": 0.4333,
"step": 896
},
{
"epoch": 0.7075527509366989,
"grad_norm": 2.1970964300895015,
"learning_rate": 2.0953002980275287e-07,
"loss": 0.421,
"step": 897
},
{
"epoch": 0.70834154999014,
"grad_norm": 2.2194862487373648,
"learning_rate": 2.0849066744155602e-07,
"loss": 0.4616,
"step": 898
},
{
"epoch": 0.7091303490435812,
"grad_norm": 2.1198860518298623,
"learning_rate": 2.0745320987520054e-07,
"loss": 0.4996,
"step": 899
},
{
"epoch": 0.7099191480970223,
"grad_norm": 2.2210862414262893,
"learning_rate": 2.0641766388269344e-07,
"loss": 0.4793,
"step": 900
},
{
"epoch": 0.7107079471504634,
"grad_norm": 2.404125752325992,
"learning_rate": 2.0538403623055107e-07,
"loss": 0.4954,
"step": 901
},
{
"epoch": 0.7114967462039046,
"grad_norm": 2.0120708506775054,
"learning_rate": 2.0435233367275485e-07,
"loss": 0.4679,
"step": 902
},
{
"epoch": 0.7122855452573457,
"grad_norm": 2.0764920481538756,
"learning_rate": 2.0332256295070705e-07,
"loss": 0.4659,
"step": 903
},
{
"epoch": 0.7130743443107869,
"grad_norm": 2.384743996660934,
"learning_rate": 2.0229473079318727e-07,
"loss": 0.459,
"step": 904
},
{
"epoch": 0.713863143364228,
"grad_norm": 2.305735217088676,
"learning_rate": 2.012688439163075e-07,
"loss": 0.4994,
"step": 905
},
{
"epoch": 0.714651942417669,
"grad_norm": 2.1810441255341577,
"learning_rate": 2.0024490902346913e-07,
"loss": 0.5251,
"step": 906
},
{
"epoch": 0.7154407414711103,
"grad_norm": 2.1607193085261,
"learning_rate": 1.992229328053186e-07,
"loss": 0.5418,
"step": 907
},
{
"epoch": 0.7162295405245513,
"grad_norm": 2.351526899272169,
"learning_rate": 1.9820292193970377e-07,
"loss": 0.4537,
"step": 908
},
{
"epoch": 0.7170183395779925,
"grad_norm": 2.4768845018810874,
"learning_rate": 1.9718488309163118e-07,
"loss": 0.4828,
"step": 909
},
{
"epoch": 0.7178071386314336,
"grad_norm": 2.0379256499508314,
"learning_rate": 1.961688229132204e-07,
"loss": 0.4842,
"step": 910
},
{
"epoch": 0.7185959376848747,
"grad_norm": 2.3407594669381036,
"learning_rate": 1.951547480436631e-07,
"loss": 0.4333,
"step": 911
},
{
"epoch": 0.7193847367383159,
"grad_norm": 2.0617820524384562,
"learning_rate": 1.9414266510917775e-07,
"loss": 0.4748,
"step": 912
},
{
"epoch": 0.720173535791757,
"grad_norm": 2.115891529970973,
"learning_rate": 1.931325807229673e-07,
"loss": 0.4663,
"step": 913
},
{
"epoch": 0.7209623348451982,
"grad_norm": 2.2934922213559865,
"learning_rate": 1.9212450148517533e-07,
"loss": 0.49,
"step": 914
},
{
"epoch": 0.7217511338986393,
"grad_norm": 2.290805211676587,
"learning_rate": 1.911184339828441e-07,
"loss": 0.4601,
"step": 915
},
{
"epoch": 0.7225399329520804,
"grad_norm": 2.1421688291902217,
"learning_rate": 1.9011438478986947e-07,
"loss": 0.4654,
"step": 916
},
{
"epoch": 0.7233287320055216,
"grad_norm": 1.868870606857226,
"learning_rate": 1.8911236046696033e-07,
"loss": 0.4851,
"step": 917
},
{
"epoch": 0.7241175310589627,
"grad_norm": 2.130661447027876,
"learning_rate": 1.8811236756159371e-07,
"loss": 0.4391,
"step": 918
},
{
"epoch": 0.7249063301124039,
"grad_norm": 2.0177688361223303,
"learning_rate": 1.8711441260797373e-07,
"loss": 0.4619,
"step": 919
},
{
"epoch": 0.725695129165845,
"grad_norm": 2.115176990261203,
"learning_rate": 1.8611850212698678e-07,
"loss": 0.4608,
"step": 920
},
{
"epoch": 0.7264839282192861,
"grad_norm": 2.160010086887992,
"learning_rate": 1.851246426261614e-07,
"loss": 0.5085,
"step": 921
},
{
"epoch": 0.7272727272727273,
"grad_norm": 2.1692503809554258,
"learning_rate": 1.841328405996238e-07,
"loss": 0.4462,
"step": 922
},
{
"epoch": 0.7280615263261684,
"grad_norm": 2.209778178159611,
"learning_rate": 1.8314310252805637e-07,
"loss": 0.5031,
"step": 923
},
{
"epoch": 0.7288503253796096,
"grad_norm": 2.2180264612852567,
"learning_rate": 1.8215543487865482e-07,
"loss": 0.5118,
"step": 924
},
{
"epoch": 0.7296391244330507,
"grad_norm": 1.99329623787221,
"learning_rate": 1.8116984410508695e-07,
"loss": 0.5005,
"step": 925
},
{
"epoch": 0.7304279234864918,
"grad_norm": 2.204447908557016,
"learning_rate": 1.8018633664744858e-07,
"loss": 0.4668,
"step": 926
},
{
"epoch": 0.731216722539933,
"grad_norm": 2.211531696986049,
"learning_rate": 1.7920491893222378e-07,
"loss": 0.485,
"step": 927
},
{
"epoch": 0.7320055215933741,
"grad_norm": 2.1209420536302455,
"learning_rate": 1.7822559737224103e-07,
"loss": 0.4987,
"step": 928
},
{
"epoch": 0.7327943206468153,
"grad_norm": 2.348373350787603,
"learning_rate": 1.7724837836663214e-07,
"loss": 0.4879,
"step": 929
},
{
"epoch": 0.7335831197002564,
"grad_norm": 2.0531285978501703,
"learning_rate": 1.762732683007902e-07,
"loss": 0.4725,
"step": 930
},
{
"epoch": 0.7343719187536975,
"grad_norm": 2.338622277648635,
"learning_rate": 1.7530027354632847e-07,
"loss": 0.4977,
"step": 931
},
{
"epoch": 0.7351607178071387,
"grad_norm": 2.2905369822143125,
"learning_rate": 1.7432940046103767e-07,
"loss": 0.4798,
"step": 932
},
{
"epoch": 0.7359495168605797,
"grad_norm": 2.2037787906036566,
"learning_rate": 1.7336065538884525e-07,
"loss": 0.4935,
"step": 933
},
{
"epoch": 0.736738315914021,
"grad_norm": 2.2629700437027482,
"learning_rate": 1.7239404465977363e-07,
"loss": 0.4719,
"step": 934
},
{
"epoch": 0.737527114967462,
"grad_norm": 2.0324517402467475,
"learning_rate": 1.7142957458989887e-07,
"loss": 0.4821,
"step": 935
},
{
"epoch": 0.7383159140209031,
"grad_norm": 2.274394745078279,
"learning_rate": 1.7046725148130942e-07,
"loss": 0.5023,
"step": 936
},
{
"epoch": 0.7391047130743443,
"grad_norm": 1.969712004777129,
"learning_rate": 1.6950708162206517e-07,
"loss": 0.4715,
"step": 937
},
{
"epoch": 0.7398935121277854,
"grad_norm": 2.1273371031029096,
"learning_rate": 1.6854907128615597e-07,
"loss": 0.4691,
"step": 938
},
{
"epoch": 0.7406823111812266,
"grad_norm": 2.121460588491523,
"learning_rate": 1.6759322673346071e-07,
"loss": 0.4898,
"step": 939
},
{
"epoch": 0.7414711102346677,
"grad_norm": 2.5778558595622507,
"learning_rate": 1.6663955420970666e-07,
"loss": 0.4394,
"step": 940
},
{
"epoch": 0.7422599092881088,
"grad_norm": 1.9648582696661259,
"learning_rate": 1.6568805994642826e-07,
"loss": 0.4907,
"step": 941
},
{
"epoch": 0.74304870834155,
"grad_norm": 2.1868084825481016,
"learning_rate": 1.6473875016092736e-07,
"loss": 0.4373,
"step": 942
},
{
"epoch": 0.7438375073949911,
"grad_norm": 2.326297823684635,
"learning_rate": 1.6379163105623073e-07,
"loss": 0.5026,
"step": 943
},
{
"epoch": 0.7446263064484323,
"grad_norm": 2.2117118198660988,
"learning_rate": 1.6284670882105183e-07,
"loss": 0.4594,
"step": 944
},
{
"epoch": 0.7454151055018734,
"grad_norm": 2.201725058473058,
"learning_rate": 1.6190398962974854e-07,
"loss": 0.4671,
"step": 945
},
{
"epoch": 0.7462039045553145,
"grad_norm": 2.0532730880757333,
"learning_rate": 1.609634796422837e-07,
"loss": 0.4769,
"step": 946
},
{
"epoch": 0.7469927036087557,
"grad_norm": 2.3182153994303647,
"learning_rate": 1.6002518500418444e-07,
"loss": 0.5265,
"step": 947
},
{
"epoch": 0.7477815026621968,
"grad_norm": 2.204825815946726,
"learning_rate": 1.5908911184650298e-07,
"loss": 0.5452,
"step": 948
},
{
"epoch": 0.748570301715638,
"grad_norm": 2.283360876683165,
"learning_rate": 1.581552662857747e-07,
"loss": 0.5059,
"step": 949
},
{
"epoch": 0.7493591007690791,
"grad_norm": 2.2082883901553942,
"learning_rate": 1.5722365442398027e-07,
"loss": 0.475,
"step": 950
},
{
"epoch": 0.7501478998225202,
"grad_norm": 2.218919679918722,
"learning_rate": 1.5629428234850427e-07,
"loss": 0.4739,
"step": 951
},
{
"epoch": 0.7509366988759614,
"grad_norm": 2.3417274805802712,
"learning_rate": 1.553671561320965e-07,
"loss": 0.4576,
"step": 952
},
{
"epoch": 0.7517254979294025,
"grad_norm": 2.173793216518148,
"learning_rate": 1.544422818328308e-07,
"loss": 0.4905,
"step": 953
},
{
"epoch": 0.7525142969828437,
"grad_norm": 2.140880113783564,
"learning_rate": 1.5351966549406758e-07,
"loss": 0.4791,
"step": 954
},
{
"epoch": 0.7533030960362848,
"grad_norm": 2.1337726088446045,
"learning_rate": 1.5259931314441226e-07,
"loss": 0.4766,
"step": 955
},
{
"epoch": 0.7540918950897258,
"grad_norm": 2.1903397883538713,
"learning_rate": 1.516812307976773e-07,
"loss": 0.4729,
"step": 956
},
{
"epoch": 0.754880694143167,
"grad_norm": 2.486597048944914,
"learning_rate": 1.5076542445284196e-07,
"loss": 0.4324,
"step": 957
},
{
"epoch": 0.7556694931966081,
"grad_norm": 2.3022174149340167,
"learning_rate": 1.4985190009401432e-07,
"loss": 0.4528,
"step": 958
},
{
"epoch": 0.7564582922500493,
"grad_norm": 2.4541908943056865,
"learning_rate": 1.4894066369039022e-07,
"loss": 0.4373,
"step": 959
},
{
"epoch": 0.7572470913034904,
"grad_norm": 2.3225184008196686,
"learning_rate": 1.480317211962166e-07,
"loss": 0.4575,
"step": 960
},
{
"epoch": 0.7580358903569315,
"grad_norm": 2.1582873413149177,
"learning_rate": 1.4712507855075073e-07,
"loss": 0.4592,
"step": 961
},
{
"epoch": 0.7588246894103727,
"grad_norm": 1.9217118622686844,
"learning_rate": 1.4622074167822228e-07,
"loss": 0.46,
"step": 962
},
{
"epoch": 0.7596134884638138,
"grad_norm": 2.136316261506256,
"learning_rate": 1.453187164877942e-07,
"loss": 0.4767,
"step": 963
},
{
"epoch": 0.760402287517255,
"grad_norm": 2.1050851943480464,
"learning_rate": 1.4441900887352493e-07,
"loss": 0.4833,
"step": 964
},
{
"epoch": 0.7611910865706961,
"grad_norm": 2.2968728675127807,
"learning_rate": 1.435216247143287e-07,
"loss": 0.5493,
"step": 965
},
{
"epoch": 0.7619798856241372,
"grad_norm": 2.0795143314755413,
"learning_rate": 1.4262656987393796e-07,
"loss": 0.4712,
"step": 966
},
{
"epoch": 0.7627686846775784,
"grad_norm": 2.410113637265569,
"learning_rate": 1.4173385020086465e-07,
"loss": 0.5067,
"step": 967
},
{
"epoch": 0.7635574837310195,
"grad_norm": 2.2218199595845047,
"learning_rate": 1.4084347152836234e-07,
"loss": 0.4898,
"step": 968
},
{
"epoch": 0.7643462827844607,
"grad_norm": 2.1682529099876056,
"learning_rate": 1.3995543967438766e-07,
"loss": 0.4889,
"step": 969
},
{
"epoch": 0.7651350818379018,
"grad_norm": 2.1893824488079785,
"learning_rate": 1.3906976044156298e-07,
"loss": 0.4588,
"step": 970
},
{
"epoch": 0.7659238808913429,
"grad_norm": 2.362590102193667,
"learning_rate": 1.3818643961713767e-07,
"loss": 0.4625,
"step": 971
},
{
"epoch": 0.7667126799447841,
"grad_norm": 2.481535899678208,
"learning_rate": 1.3730548297295075e-07,
"loss": 0.4409,
"step": 972
},
{
"epoch": 0.7675014789982252,
"grad_norm": 2.070929810010827,
"learning_rate": 1.364268962653931e-07,
"loss": 0.4945,
"step": 973
},
{
"epoch": 0.7682902780516664,
"grad_norm": 2.273636212339407,
"learning_rate": 1.3555068523536983e-07,
"loss": 0.477,
"step": 974
},
{
"epoch": 0.7690790771051075,
"grad_norm": 2.233128343344179,
"learning_rate": 1.3467685560826313e-07,
"loss": 0.4874,
"step": 975
},
{
"epoch": 0.7698678761585486,
"grad_norm": 2.2033651595069768,
"learning_rate": 1.338054130938937e-07,
"loss": 0.4915,
"step": 976
},
{
"epoch": 0.7706566752119898,
"grad_norm": 2.347800780563448,
"learning_rate": 1.329363633864851e-07,
"loss": 0.4895,
"step": 977
},
{
"epoch": 0.7714454742654309,
"grad_norm": 2.0580926526740666,
"learning_rate": 1.320697121646253e-07,
"loss": 0.4864,
"step": 978
},
{
"epoch": 0.7722342733188721,
"grad_norm": 2.3811283081058043,
"learning_rate": 1.3120546509122993e-07,
"loss": 0.4367,
"step": 979
},
{
"epoch": 0.7730230723723132,
"grad_norm": 2.3005374680451416,
"learning_rate": 1.3034362781350523e-07,
"loss": 0.4756,
"step": 980
},
{
"epoch": 0.7738118714257542,
"grad_norm": 2.4126245125705204,
"learning_rate": 1.2948420596291182e-07,
"loss": 0.4805,
"step": 981
},
{
"epoch": 0.7746006704791955,
"grad_norm": 1.9346907776034008,
"learning_rate": 1.2862720515512636e-07,
"loss": 0.5043,
"step": 982
},
{
"epoch": 0.7753894695326365,
"grad_norm": 2.1341693928607954,
"learning_rate": 1.2777263099000673e-07,
"loss": 0.4794,
"step": 983
},
{
"epoch": 0.7761782685860777,
"grad_norm": 2.0407987410502195,
"learning_rate": 1.26920489051554e-07,
"loss": 0.44,
"step": 984
},
{
"epoch": 0.7769670676395188,
"grad_norm": 2.10225527576356,
"learning_rate": 1.260707849078766e-07,
"loss": 0.5092,
"step": 985
},
{
"epoch": 0.7777558666929599,
"grad_norm": 2.191391137078327,
"learning_rate": 1.2522352411115362e-07,
"loss": 0.4817,
"step": 986
},
{
"epoch": 0.7785446657464011,
"grad_norm": 2.1538135670617513,
"learning_rate": 1.243787121975991e-07,
"loss": 0.4837,
"step": 987
},
{
"epoch": 0.7793334647998422,
"grad_norm": 2.2748749633222776,
"learning_rate": 1.235363546874252e-07,
"loss": 0.4694,
"step": 988
},
{
"epoch": 0.7801222638532834,
"grad_norm": 2.205284137789626,
"learning_rate": 1.2269645708480641e-07,
"loss": 0.5083,
"step": 989
},
{
"epoch": 0.7809110629067245,
"grad_norm": 2.337515205980958,
"learning_rate": 1.218590248778434e-07,
"loss": 0.4371,
"step": 990
},
{
"epoch": 0.7816998619601656,
"grad_norm": 2.1511768204109423,
"learning_rate": 1.2102406353852807e-07,
"loss": 0.517,
"step": 991
},
{
"epoch": 0.7824886610136068,
"grad_norm": 2.172429149349416,
"learning_rate": 1.2019157852270578e-07,
"loss": 0.4493,
"step": 992
},
{
"epoch": 0.7832774600670479,
"grad_norm": 2.4370059083198354,
"learning_rate": 1.1936157527004226e-07,
"loss": 0.4215,
"step": 993
},
{
"epoch": 0.7840662591204891,
"grad_norm": 2.291716098712647,
"learning_rate": 1.1853405920398607e-07,
"loss": 0.4537,
"step": 994
},
{
"epoch": 0.7848550581739302,
"grad_norm": 2.1501129900599394,
"learning_rate": 1.1770903573173397e-07,
"loss": 0.4726,
"step": 995
},
{
"epoch": 0.7856438572273713,
"grad_norm": 2.551242705986409,
"learning_rate": 1.1688651024419538e-07,
"loss": 0.4631,
"step": 996
},
{
"epoch": 0.7864326562808125,
"grad_norm": 2.1429282269404033,
"learning_rate": 1.1606648811595776e-07,
"loss": 0.4768,
"step": 997
},
{
"epoch": 0.7872214553342536,
"grad_norm": 2.1539849902176935,
"learning_rate": 1.1524897470525052e-07,
"loss": 0.5226,
"step": 998
},
{
"epoch": 0.7880102543876948,
"grad_norm": 2.1765357732242414,
"learning_rate": 1.1443397535391058e-07,
"loss": 0.4443,
"step": 999
},
{
"epoch": 0.7887990534411359,
"grad_norm": 2.1585030534649428,
"learning_rate": 1.1362149538734744e-07,
"loss": 0.4997,
"step": 1000
},
{
"epoch": 0.789587852494577,
"grad_norm": 2.130929654566126,
"learning_rate": 1.1281154011450827e-07,
"loss": 0.4959,
"step": 1001
},
{
"epoch": 0.7903766515480182,
"grad_norm": 1.9866227778758776,
"learning_rate": 1.1200411482784317e-07,
"loss": 0.4523,
"step": 1002
},
{
"epoch": 0.7911654506014593,
"grad_norm": 2.14724068381651,
"learning_rate": 1.1119922480327098e-07,
"loss": 0.4844,
"step": 1003
},
{
"epoch": 0.7919542496549005,
"grad_norm": 2.0267804655029673,
"learning_rate": 1.1039687530014418e-07,
"loss": 0.4706,
"step": 1004
},
{
"epoch": 0.7927430487083416,
"grad_norm": 2.086690515507619,
"learning_rate": 1.0959707156121506e-07,
"loss": 0.4677,
"step": 1005
},
{
"epoch": 0.7935318477617826,
"grad_norm": 2.1548684706389905,
"learning_rate": 1.0879981881260108e-07,
"loss": 0.5205,
"step": 1006
},
{
"epoch": 0.7943206468152239,
"grad_norm": 2.0756670947366684,
"learning_rate": 1.080051222637508e-07,
"loss": 0.4673,
"step": 1007
},
{
"epoch": 0.7951094458686649,
"grad_norm": 2.327985692482114,
"learning_rate": 1.0721298710741061e-07,
"loss": 0.4647,
"step": 1008
},
{
"epoch": 0.7958982449221061,
"grad_norm": 2.248453205926095,
"learning_rate": 1.0642341851958902e-07,
"loss": 0.5074,
"step": 1009
},
{
"epoch": 0.7966870439755472,
"grad_norm": 1.9144416508969329,
"learning_rate": 1.0563642165952497e-07,
"loss": 0.4736,
"step": 1010
},
{
"epoch": 0.7974758430289883,
"grad_norm": 1.9717115920565795,
"learning_rate": 1.0485200166965241e-07,
"loss": 0.4872,
"step": 1011
},
{
"epoch": 0.7982646420824295,
"grad_norm": 2.203286326088002,
"learning_rate": 1.0407016367556775e-07,
"loss": 0.4544,
"step": 1012
},
{
"epoch": 0.7990534411358706,
"grad_norm": 1.9665973292402223,
"learning_rate": 1.0329091278599567e-07,
"loss": 0.4742,
"step": 1013
},
{
"epoch": 0.7998422401893118,
"grad_norm": 2.0835750284062544,
"learning_rate": 1.0251425409275671e-07,
"loss": 0.4955,
"step": 1014
},
{
"epoch": 0.8006310392427529,
"grad_norm": 2.217767952097549,
"learning_rate": 1.0174019267073226e-07,
"loss": 0.5322,
"step": 1015
},
{
"epoch": 0.801419838296194,
"grad_norm": 2.4361561202663933,
"learning_rate": 1.0096873357783359e-07,
"loss": 0.4572,
"step": 1016
},
{
"epoch": 0.8022086373496352,
"grad_norm": 2.203846116279926,
"learning_rate": 1.0019988185496713e-07,
"loss": 0.5056,
"step": 1017
},
{
"epoch": 0.8029974364030763,
"grad_norm": 2.027560600491381,
"learning_rate": 9.943364252600217e-08,
"loss": 0.5316,
"step": 1018
},
{
"epoch": 0.8037862354565175,
"grad_norm": 2.481754200785473,
"learning_rate": 9.867002059773788e-08,
"loss": 0.4483,
"step": 1019
},
{
"epoch": 0.8045750345099586,
"grad_norm": 2.3591317893505424,
"learning_rate": 9.790902105987103e-08,
"loss": 0.4446,
"step": 1020
},
{
"epoch": 0.8053638335633997,
"grad_norm": 2.2567559500932326,
"learning_rate": 9.715064888496277e-08,
"loss": 0.445,
"step": 1021
},
{
"epoch": 0.8061526326168409,
"grad_norm": 2.241401857728176,
"learning_rate": 9.639490902840641e-08,
"loss": 0.4958,
"step": 1022
},
{
"epoch": 0.806941431670282,
"grad_norm": 2.0463981302272107,
"learning_rate": 9.564180642839492e-08,
"loss": 0.4777,
"step": 1023
},
{
"epoch": 0.8077302307237232,
"grad_norm": 2.1606155604104518,
"learning_rate": 9.489134600588932e-08,
"loss": 0.4589,
"step": 1024
},
{
"epoch": 0.8085190297771643,
"grad_norm": 2.344788556938593,
"learning_rate": 9.414353266458508e-08,
"loss": 0.5157,
"step": 1025
},
{
"epoch": 0.8093078288306054,
"grad_norm": 2.675005237255825,
"learning_rate": 9.339837129088184e-08,
"loss": 0.4654,
"step": 1026
},
{
"epoch": 0.8100966278840466,
"grad_norm": 1.9831792529426433,
"learning_rate": 9.265586675385017e-08,
"loss": 0.4561,
"step": 1027
},
{
"epoch": 0.8108854269374877,
"grad_norm": 2.1149057877689725,
"learning_rate": 9.191602390520032e-08,
"loss": 0.4896,
"step": 1028
},
{
"epoch": 0.8116742259909289,
"grad_norm": 1.9507383524203066,
"learning_rate": 9.11788475792502e-08,
"loss": 0.5281,
"step": 1029
},
{
"epoch": 0.81246302504437,
"grad_norm": 2.4432618966506707,
"learning_rate": 9.044434259289452e-08,
"loss": 0.4437,
"step": 1030
},
{
"epoch": 0.813251824097811,
"grad_norm": 2.2559894136250307,
"learning_rate": 8.971251374557225e-08,
"loss": 0.4781,
"step": 1031
},
{
"epoch": 0.8140406231512523,
"grad_norm": 2.1489114392118625,
"learning_rate": 8.898336581923598e-08,
"loss": 0.4733,
"step": 1032
},
{
"epoch": 0.8148294222046933,
"grad_norm": 2.0520080443452673,
"learning_rate": 8.825690357832061e-08,
"loss": 0.4998,
"step": 1033
},
{
"epoch": 0.8156182212581344,
"grad_norm": 2.2684590434506875,
"learning_rate": 8.753313176971195e-08,
"loss": 0.5161,
"step": 1034
},
{
"epoch": 0.8164070203115756,
"grad_norm": 2.148499612350094,
"learning_rate": 8.681205512271572e-08,
"loss": 0.5174,
"step": 1035
},
{
"epoch": 0.8171958193650167,
"grad_norm": 2.4936409901746046,
"learning_rate": 8.609367834902731e-08,
"loss": 0.4471,
"step": 1036
},
{
"epoch": 0.8179846184184579,
"grad_norm": 2.0125129563257,
"learning_rate": 8.537800614269997e-08,
"loss": 0.5432,
"step": 1037
},
{
"epoch": 0.818773417471899,
"grad_norm": 2.2604401927608966,
"learning_rate": 8.46650431801147e-08,
"loss": 0.4729,
"step": 1038
},
{
"epoch": 0.8195622165253401,
"grad_norm": 2.2133660362240657,
"learning_rate": 8.395479411994977e-08,
"loss": 0.4973,
"step": 1039
},
{
"epoch": 0.8203510155787813,
"grad_norm": 2.4376958967799265,
"learning_rate": 8.324726360314993e-08,
"loss": 0.4763,
"step": 1040
},
{
"epoch": 0.8211398146322224,
"grad_norm": 2.0731054660479926,
"learning_rate": 8.25424562528968e-08,
"loss": 0.504,
"step": 1041
},
{
"epoch": 0.8219286136856636,
"grad_norm": 2.0616913530411387,
"learning_rate": 8.184037667457728e-08,
"loss": 0.4627,
"step": 1042
},
{
"epoch": 0.8227174127391047,
"grad_norm": 1.931730636869577,
"learning_rate": 8.114102945575513e-08,
"loss": 0.4697,
"step": 1043
},
{
"epoch": 0.8235062117925458,
"grad_norm": 2.423559053753917,
"learning_rate": 8.044441916613981e-08,
"loss": 0.4605,
"step": 1044
},
{
"epoch": 0.824295010845987,
"grad_norm": 2.183293131898637,
"learning_rate": 7.975055035755712e-08,
"loss": 0.4826,
"step": 1045
},
{
"epoch": 0.8250838098994281,
"grad_norm": 2.1950803556959717,
"learning_rate": 7.905942756391903e-08,
"loss": 0.4828,
"step": 1046
},
{
"epoch": 0.8258726089528693,
"grad_norm": 2.2739586757672403,
"learning_rate": 7.837105530119514e-08,
"loss": 0.4564,
"step": 1047
},
{
"epoch": 0.8266614080063104,
"grad_norm": 2.1836352167837836,
"learning_rate": 7.768543806738126e-08,
"loss": 0.4495,
"step": 1048
},
{
"epoch": 0.8274502070597515,
"grad_norm": 2.204451903446692,
"learning_rate": 7.70025803424722e-08,
"loss": 0.4832,
"step": 1049
},
{
"epoch": 0.8282390061131927,
"grad_norm": 2.4106553517759948,
"learning_rate": 7.632248658843088e-08,
"loss": 0.4519,
"step": 1050
},
{
"epoch": 0.8290278051666338,
"grad_norm": 2.0156630203515356,
"learning_rate": 7.564516124915981e-08,
"loss": 0.4712,
"step": 1051
},
{
"epoch": 0.829816604220075,
"grad_norm": 2.4782990035561396,
"learning_rate": 7.497060875047206e-08,
"loss": 0.4985,
"step": 1052
},
{
"epoch": 0.8306054032735161,
"grad_norm": 2.102270932427116,
"learning_rate": 7.429883350006239e-08,
"loss": 0.4036,
"step": 1053
},
{
"epoch": 0.8313942023269572,
"grad_norm": 2.2003205130366457,
"learning_rate": 7.362983988747818e-08,
"loss": 0.4352,
"step": 1054
},
{
"epoch": 0.8321830013803984,
"grad_norm": 2.13090706833757,
"learning_rate": 7.296363228409075e-08,
"loss": 0.4709,
"step": 1055
},
{
"epoch": 0.8329718004338394,
"grad_norm": 2.509154107178616,
"learning_rate": 7.230021504306716e-08,
"loss": 0.4942,
"step": 1056
},
{
"epoch": 0.8337605994872807,
"grad_norm": 2.2587648320471776,
"learning_rate": 7.16395924993417e-08,
"loss": 0.4447,
"step": 1057
},
{
"epoch": 0.8345493985407217,
"grad_norm": 1.8668872993117411,
"learning_rate": 7.09817689695868e-08,
"loss": 0.5468,
"step": 1058
},
{
"epoch": 0.8353381975941628,
"grad_norm": 1.979177205134116,
"learning_rate": 7.032674875218613e-08,
"loss": 0.4448,
"step": 1059
},
{
"epoch": 0.836126996647604,
"grad_norm": 2.2829706767319236,
"learning_rate": 6.967453612720541e-08,
"loss": 0.4743,
"step": 1060
},
{
"epoch": 0.8369157957010451,
"grad_norm": 2.319812936602923,
"learning_rate": 6.902513535636501e-08,
"loss": 0.4227,
"step": 1061
},
{
"epoch": 0.8377045947544863,
"grad_norm": 2.2275810706188857,
"learning_rate": 6.837855068301168e-08,
"loss": 0.4841,
"step": 1062
},
{
"epoch": 0.8384933938079274,
"grad_norm": 2.2667149451455897,
"learning_rate": 6.773478633209157e-08,
"loss": 0.4494,
"step": 1063
},
{
"epoch": 0.8392821928613685,
"grad_norm": 2.1368859925730135,
"learning_rate": 6.709384651012173e-08,
"loss": 0.5063,
"step": 1064
},
{
"epoch": 0.8400709919148097,
"grad_norm": 2.0559943710137545,
"learning_rate": 6.645573540516314e-08,
"loss": 0.4862,
"step": 1065
},
{
"epoch": 0.8408597909682508,
"grad_norm": 2.223579767989312,
"learning_rate": 6.582045718679324e-08,
"loss": 0.4884,
"step": 1066
},
{
"epoch": 0.841648590021692,
"grad_norm": 2.1584885431970635,
"learning_rate": 6.518801600607882e-08,
"loss": 0.5029,
"step": 1067
},
{
"epoch": 0.8424373890751331,
"grad_norm": 2.072122529540598,
"learning_rate": 6.455841599554834e-08,
"loss": 0.4855,
"step": 1068
},
{
"epoch": 0.8432261881285742,
"grad_norm": 2.2589727363405823,
"learning_rate": 6.3931661269166e-08,
"loss": 0.5056,
"step": 1069
},
{
"epoch": 0.8440149871820154,
"grad_norm": 2.448309486593933,
"learning_rate": 6.330775592230364e-08,
"loss": 0.4552,
"step": 1070
},
{
"epoch": 0.8448037862354565,
"grad_norm": 2.004377188250325,
"learning_rate": 6.268670403171483e-08,
"loss": 0.4878,
"step": 1071
},
{
"epoch": 0.8455925852888977,
"grad_norm": 2.3518127756771503,
"learning_rate": 6.206850965550775e-08,
"loss": 0.4883,
"step": 1072
},
{
"epoch": 0.8463813843423388,
"grad_norm": 2.195656744928062,
"learning_rate": 6.145317683311902e-08,
"loss": 0.4848,
"step": 1073
},
{
"epoch": 0.8471701833957799,
"grad_norm": 2.1683134274629348,
"learning_rate": 6.084070958528714e-08,
"loss": 0.4555,
"step": 1074
},
{
"epoch": 0.8479589824492211,
"grad_norm": 2.2552763020698605,
"learning_rate": 6.023111191402602e-08,
"loss": 0.4455,
"step": 1075
},
{
"epoch": 0.8487477815026622,
"grad_norm": 2.4279988906949517,
"learning_rate": 5.962438780259948e-08,
"loss": 0.4718,
"step": 1076
},
{
"epoch": 0.8495365805561034,
"grad_norm": 2.137142449268899,
"learning_rate": 5.902054121549438e-08,
"loss": 0.4612,
"step": 1077
},
{
"epoch": 0.8503253796095445,
"grad_norm": 2.111854969035769,
"learning_rate": 5.8419576098395265e-08,
"loss": 0.4953,
"step": 1078
},
{
"epoch": 0.8511141786629856,
"grad_norm": 2.1410566907936066,
"learning_rate": 5.7821496378158254e-08,
"loss": 0.4455,
"step": 1079
},
{
"epoch": 0.8519029777164268,
"grad_norm": 2.1521897546900544,
"learning_rate": 5.722630596278616e-08,
"loss": 0.4401,
"step": 1080
},
{
"epoch": 0.8526917767698678,
"grad_norm": 2.1319727518168077,
"learning_rate": 5.663400874140139e-08,
"loss": 0.4631,
"step": 1081
},
{
"epoch": 0.853480575823309,
"grad_norm": 2.1202826735625244,
"learning_rate": 5.6044608584222444e-08,
"loss": 0.4569,
"step": 1082
},
{
"epoch": 0.8542693748767501,
"grad_norm": 2.1616769527767774,
"learning_rate": 5.5458109342537084e-08,
"loss": 0.442,
"step": 1083
},
{
"epoch": 0.8550581739301912,
"grad_norm": 2.217778823981212,
"learning_rate": 5.487451484867794e-08,
"loss": 0.4586,
"step": 1084
},
{
"epoch": 0.8558469729836324,
"grad_norm": 2.181886524953751,
"learning_rate": 5.429382891599732e-08,
"loss": 0.4287,
"step": 1085
},
{
"epoch": 0.8566357720370735,
"grad_norm": 1.9165125995699759,
"learning_rate": 5.3716055338842326e-08,
"loss": 0.5104,
"step": 1086
},
{
"epoch": 0.8574245710905147,
"grad_norm": 2.3293883855157613,
"learning_rate": 5.314119789252991e-08,
"loss": 0.4704,
"step": 1087
},
{
"epoch": 0.8582133701439558,
"grad_norm": 2.034534973029934,
"learning_rate": 5.2569260333322264e-08,
"loss": 0.5057,
"step": 1088
},
{
"epoch": 0.8590021691973969,
"grad_norm": 2.0970042411312466,
"learning_rate": 5.200024639840228e-08,
"loss": 0.489,
"step": 1089
},
{
"epoch": 0.8597909682508381,
"grad_norm": 2.297857542073598,
"learning_rate": 5.1434159805849474e-08,
"loss": 0.4833,
"step": 1090
},
{
"epoch": 0.8605797673042792,
"grad_norm": 2.296749187421612,
"learning_rate": 5.087100425461482e-08,
"loss": 0.4503,
"step": 1091
},
{
"epoch": 0.8613685663577204,
"grad_norm": 2.2460865939862606,
"learning_rate": 5.0310783424497735e-08,
"loss": 0.4331,
"step": 1092
},
{
"epoch": 0.8621573654111615,
"grad_norm": 2.2387967842546765,
"learning_rate": 4.975350097612096e-08,
"loss": 0.4901,
"step": 1093
},
{
"epoch": 0.8629461644646026,
"grad_norm": 2.3578530940565128,
"learning_rate": 4.919916055090739e-08,
"loss": 0.4586,
"step": 1094
},
{
"epoch": 0.8637349635180438,
"grad_norm": 2.1620591909350027,
"learning_rate": 4.864776577105584e-08,
"loss": 0.4506,
"step": 1095
},
{
"epoch": 0.8645237625714849,
"grad_norm": 2.0136467906011073,
"learning_rate": 4.809932023951774e-08,
"loss": 0.527,
"step": 1096
},
{
"epoch": 0.8653125616249261,
"grad_norm": 2.1774290380649117,
"learning_rate": 4.7553827539973225e-08,
"loss": 0.4351,
"step": 1097
},
{
"epoch": 0.8661013606783672,
"grad_norm": 1.9175177715539433,
"learning_rate": 4.701129123680791e-08,
"loss": 0.5016,
"step": 1098
},
{
"epoch": 0.8668901597318083,
"grad_norm": 2.197846907284896,
"learning_rate": 4.6471714875089665e-08,
"loss": 0.4758,
"step": 1099
},
{
"epoch": 0.8676789587852495,
"grad_norm": 1.9618893953588064,
"learning_rate": 4.593510198054529e-08,
"loss": 0.4978,
"step": 1100
},
{
"epoch": 0.8684677578386906,
"grad_norm": 2.13672417930152,
"learning_rate": 4.540145605953738e-08,
"loss": 0.4338,
"step": 1101
},
{
"epoch": 0.8692565568921318,
"grad_norm": 2.5894893901796734,
"learning_rate": 4.487078059904204e-08,
"loss": 0.505,
"step": 1102
},
{
"epoch": 0.8700453559455729,
"grad_norm": 1.8330130649906766,
"learning_rate": 4.4343079066625267e-08,
"loss": 0.4683,
"step": 1103
},
{
"epoch": 0.870834154999014,
"grad_norm": 2.4602987840611537,
"learning_rate": 4.381835491042079e-08,
"loss": 0.4527,
"step": 1104
},
{
"epoch": 0.8716229540524552,
"grad_norm": 2.18013469016652,
"learning_rate": 4.3296611559107356e-08,
"loss": 0.4558,
"step": 1105
},
{
"epoch": 0.8724117531058962,
"grad_norm": 2.3472610894408303,
"learning_rate": 4.277785242188642e-08,
"loss": 0.4656,
"step": 1106
},
{
"epoch": 0.8732005521593375,
"grad_norm": 2.05528624251208,
"learning_rate": 4.2262080888459897e-08,
"loss": 0.519,
"step": 1107
},
{
"epoch": 0.8739893512127785,
"grad_norm": 2.20225622415052,
"learning_rate": 4.174930032900781e-08,
"loss": 0.4998,
"step": 1108
},
{
"epoch": 0.8747781502662196,
"grad_norm": 2.173086068909838,
"learning_rate": 4.123951409416665e-08,
"loss": 0.4571,
"step": 1109
},
{
"epoch": 0.8755669493196608,
"grad_norm": 2.013545206618128,
"learning_rate": 4.0732725515007004e-08,
"loss": 0.4597,
"step": 1110
},
{
"epoch": 0.8763557483731019,
"grad_norm": 2.2398553432849937,
"learning_rate": 4.022893790301218e-08,
"loss": 0.4461,
"step": 1111
},
{
"epoch": 0.8771445474265431,
"grad_norm": 2.112508580719257,
"learning_rate": 3.972815455005618e-08,
"loss": 0.5469,
"step": 1112
},
{
"epoch": 0.8779333464799842,
"grad_norm": 2.136736972853266,
"learning_rate": 3.923037872838292e-08,
"loss": 0.4496,
"step": 1113
},
{
"epoch": 0.8787221455334253,
"grad_norm": 2.2419603301220046,
"learning_rate": 3.87356136905837e-08,
"loss": 0.4963,
"step": 1114
},
{
"epoch": 0.8795109445868665,
"grad_norm": 2.1358039835267832,
"learning_rate": 3.8243862669577186e-08,
"loss": 0.4755,
"step": 1115
},
{
"epoch": 0.8802997436403076,
"grad_norm": 2.425241469482105,
"learning_rate": 3.775512887858734e-08,
"loss": 0.4962,
"step": 1116
},
{
"epoch": 0.8810885426937488,
"grad_norm": 2.113826381797499,
"learning_rate": 3.7269415511122945e-08,
"loss": 0.4843,
"step": 1117
},
{
"epoch": 0.8818773417471899,
"grad_norm": 1.9573960813654754,
"learning_rate": 3.678672574095648e-08,
"loss": 0.4726,
"step": 1118
},
{
"epoch": 0.882666140800631,
"grad_norm": 2.13209187056848,
"learning_rate": 3.630706272210371e-08,
"loss": 0.4751,
"step": 1119
},
{
"epoch": 0.8834549398540722,
"grad_norm": 2.246392158952085,
"learning_rate": 3.583042958880256e-08,
"loss": 0.5152,
"step": 1120
},
{
"epoch": 0.8842437389075133,
"grad_norm": 2.104235617411064,
"learning_rate": 3.535682945549312e-08,
"loss": 0.4602,
"step": 1121
},
{
"epoch": 0.8850325379609545,
"grad_norm": 2.2907950317272334,
"learning_rate": 3.48862654167969e-08,
"loss": 0.4709,
"step": 1122
},
{
"epoch": 0.8858213370143956,
"grad_norm": 2.173218470054423,
"learning_rate": 3.441874054749727e-08,
"loss": 0.4509,
"step": 1123
},
{
"epoch": 0.8866101360678367,
"grad_norm": 2.187243954911098,
"learning_rate": 3.395425790251816e-08,
"loss": 0.4331,
"step": 1124
},
{
"epoch": 0.8873989351212779,
"grad_norm": 2.4428193679797414,
"learning_rate": 3.3492820516905605e-08,
"loss": 0.4647,
"step": 1125
},
{
"epoch": 0.888187734174719,
"grad_norm": 2.207307597227354,
"learning_rate": 3.303443140580664e-08,
"loss": 0.4901,
"step": 1126
},
{
"epoch": 0.8889765332281602,
"grad_norm": 2.0438104206109733,
"learning_rate": 3.257909356445038e-08,
"loss": 0.4914,
"step": 1127
},
{
"epoch": 0.8897653322816013,
"grad_norm": 2.232724144060166,
"learning_rate": 3.212680996812794e-08,
"loss": 0.4832,
"step": 1128
},
{
"epoch": 0.8905541313350424,
"grad_norm": 2.1147762909514567,
"learning_rate": 3.1677583572173704e-08,
"loss": 0.4619,
"step": 1129
},
{
"epoch": 0.8913429303884836,
"grad_norm": 2.3899411906539867,
"learning_rate": 3.123141731194484e-08,
"loss": 0.4144,
"step": 1130
},
{
"epoch": 0.8921317294419246,
"grad_norm": 2.060034105546834,
"learning_rate": 3.078831410280347e-08,
"loss": 0.4648,
"step": 1131
},
{
"epoch": 0.8929205284953659,
"grad_norm": 1.9707718960305716,
"learning_rate": 3.0348276840096664e-08,
"loss": 0.476,
"step": 1132
},
{
"epoch": 0.8937093275488069,
"grad_norm": 2.1795697011453274,
"learning_rate": 2.99113083991378e-08,
"loss": 0.5229,
"step": 1133
},
{
"epoch": 0.894498126602248,
"grad_norm": 2.3033006728619894,
"learning_rate": 2.947741163518791e-08,
"loss": 0.4753,
"step": 1134
},
{
"epoch": 0.8952869256556892,
"grad_norm": 2.2367672180444633,
"learning_rate": 2.9046589383437015e-08,
"loss": 0.4575,
"step": 1135
},
{
"epoch": 0.8960757247091303,
"grad_norm": 2.471309149645162,
"learning_rate": 2.8618844458985326e-08,
"loss": 0.4623,
"step": 1136
},
{
"epoch": 0.8968645237625715,
"grad_norm": 2.1556296875214476,
"learning_rate": 2.8194179656825133e-08,
"loss": 0.4784,
"step": 1137
},
{
"epoch": 0.8976533228160126,
"grad_norm": 2.6464417527400657,
"learning_rate": 2.777259775182228e-08,
"loss": 0.4876,
"step": 1138
},
{
"epoch": 0.8984421218694537,
"grad_norm": 1.9954198703063286,
"learning_rate": 2.7354101498698377e-08,
"loss": 0.4837,
"step": 1139
},
{
"epoch": 0.8992309209228949,
"grad_norm": 1.8908313100919454,
"learning_rate": 2.6938693632012567e-08,
"loss": 0.4648,
"step": 1140
},
{
"epoch": 0.900019719976336,
"grad_norm": 2.2250555886701107,
"learning_rate": 2.6526376866143618e-08,
"loss": 0.4464,
"step": 1141
},
{
"epoch": 0.9008085190297772,
"grad_norm": 2.096778313532699,
"learning_rate": 2.6117153895272414e-08,
"loss": 0.4114,
"step": 1142
},
{
"epoch": 0.9015973180832183,
"grad_norm": 2.3832883058090397,
"learning_rate": 2.5711027393364115e-08,
"loss": 0.51,
"step": 1143
},
{
"epoch": 0.9023861171366594,
"grad_norm": 1.8074394378950092,
"learning_rate": 2.5308000014150788e-08,
"loss": 0.5033,
"step": 1144
},
{
"epoch": 0.9031749161901006,
"grad_norm": 2.1265149143582924,
"learning_rate": 2.490807439111392e-08,
"loss": 0.4443,
"step": 1145
},
{
"epoch": 0.9039637152435417,
"grad_norm": 2.0751663077061084,
"learning_rate": 2.4511253137467824e-08,
"loss": 0.4883,
"step": 1146
},
{
"epoch": 0.9047525142969829,
"grad_norm": 2.3687578219805565,
"learning_rate": 2.4117538846141428e-08,
"loss": 0.4426,
"step": 1147
},
{
"epoch": 0.905541313350424,
"grad_norm": 2.181956633318485,
"learning_rate": 2.3726934089762507e-08,
"loss": 0.4915,
"step": 1148
},
{
"epoch": 0.9063301124038651,
"grad_norm": 1.9849351281698502,
"learning_rate": 2.3339441420640095e-08,
"loss": 0.4604,
"step": 1149
},
{
"epoch": 0.9071189114573063,
"grad_norm": 2.2057193267227975,
"learning_rate": 2.295506337074815e-08,
"loss": 0.5014,
"step": 1150
},
{
"epoch": 0.9079077105107474,
"grad_norm": 2.2566363517069528,
"learning_rate": 2.2573802451708924e-08,
"loss": 0.4827,
"step": 1151
},
{
"epoch": 0.9086965095641886,
"grad_norm": 2.2338537195698023,
"learning_rate": 2.2195661154776613e-08,
"loss": 0.4383,
"step": 1152
},
{
"epoch": 0.9094853086176297,
"grad_norm": 2.161988846582854,
"learning_rate": 2.1820641950820896e-08,
"loss": 0.4771,
"step": 1153
},
{
"epoch": 0.9102741076710708,
"grad_norm": 2.0952563227603047,
"learning_rate": 2.1448747290311097e-08,
"loss": 0.4871,
"step": 1154
},
{
"epoch": 0.911062906724512,
"grad_norm": 2.1592299438465785,
"learning_rate": 2.1079979603299657e-08,
"loss": 0.4895,
"step": 1155
},
{
"epoch": 0.911851705777953,
"grad_norm": 2.154241199852767,
"learning_rate": 2.0714341299407134e-08,
"loss": 0.432,
"step": 1156
},
{
"epoch": 0.9126405048313943,
"grad_norm": 2.4383483097350944,
"learning_rate": 2.0351834767805276e-08,
"loss": 0.5105,
"step": 1157
},
{
"epoch": 0.9134293038848353,
"grad_norm": 2.3496734654637845,
"learning_rate": 1.9992462377202647e-08,
"loss": 0.4693,
"step": 1158
},
{
"epoch": 0.9142181029382764,
"grad_norm": 2.042338461017465,
"learning_rate": 1.9636226475828078e-08,
"loss": 0.5039,
"step": 1159
},
{
"epoch": 0.9150069019917176,
"grad_norm": 2.1787394071068857,
"learning_rate": 1.9283129391416075e-08,
"loss": 0.4427,
"step": 1160
},
{
"epoch": 0.9157957010451587,
"grad_norm": 2.3109395501024754,
"learning_rate": 1.893317343119116e-08,
"loss": 0.4284,
"step": 1161
},
{
"epoch": 0.9165845000985999,
"grad_norm": 2.02530434922422,
"learning_rate": 1.8586360881853225e-08,
"loss": 0.4688,
"step": 1162
},
{
"epoch": 0.917373299152041,
"grad_norm": 2.414443579735098,
"learning_rate": 1.824269400956191e-08,
"loss": 0.4853,
"step": 1163
},
{
"epoch": 0.9181620982054821,
"grad_norm": 2.2106321690067827,
"learning_rate": 1.790217505992264e-08,
"loss": 0.4504,
"step": 1164
},
{
"epoch": 0.9189508972589233,
"grad_norm": 2.3637887202907613,
"learning_rate": 1.7564806257971355e-08,
"loss": 0.4,
"step": 1165
},
{
"epoch": 0.9197396963123644,
"grad_norm": 2.02723130462276,
"learning_rate": 1.723058980816e-08,
"loss": 0.4913,
"step": 1166
},
{
"epoch": 0.9205284953658056,
"grad_norm": 2.1125931520549113,
"learning_rate": 1.68995278943424e-08,
"loss": 0.4254,
"step": 1167
},
{
"epoch": 0.9213172944192467,
"grad_norm": 2.4678318285078458,
"learning_rate": 1.6571622679760034e-08,
"loss": 0.4968,
"step": 1168
},
{
"epoch": 0.9221060934726878,
"grad_norm": 1.9835749610347755,
"learning_rate": 1.6246876307027368e-08,
"loss": 0.5,
"step": 1169
},
{
"epoch": 0.922894892526129,
"grad_norm": 2.438541414059369,
"learning_rate": 1.5925290898118393e-08,
"loss": 0.4673,
"step": 1170
},
{
"epoch": 0.9236836915795701,
"grad_norm": 2.0130083847111306,
"learning_rate": 1.5606868554352613e-08,
"loss": 0.4876,
"step": 1171
},
{
"epoch": 0.9244724906330113,
"grad_norm": 2.1558318106131007,
"learning_rate": 1.529161135638113e-08,
"loss": 0.485,
"step": 1172
},
{
"epoch": 0.9252612896864524,
"grad_norm": 1.969535375263106,
"learning_rate": 1.4979521364173354e-08,
"loss": 0.4991,
"step": 1173
},
{
"epoch": 0.9260500887398935,
"grad_norm": 2.3239492140805917,
"learning_rate": 1.4670600617003203e-08,
"loss": 0.4886,
"step": 1174
},
{
"epoch": 0.9268388877933347,
"grad_norm": 2.313842410039279,
"learning_rate": 1.4364851133436163e-08,
"loss": 0.4274,
"step": 1175
},
{
"epoch": 0.9276276868467758,
"grad_norm": 2.262188915945084,
"learning_rate": 1.4062274911315674e-08,
"loss": 0.4758,
"step": 1176
},
{
"epoch": 0.928416485900217,
"grad_norm": 2.014475966755726,
"learning_rate": 1.3762873927750495e-08,
"loss": 0.5084,
"step": 1177
},
{
"epoch": 0.9292052849536581,
"grad_norm": 2.1405330191360044,
"learning_rate": 1.3466650139101422e-08,
"loss": 0.4693,
"step": 1178
},
{
"epoch": 0.9299940840070992,
"grad_norm": 2.1314202076605384,
"learning_rate": 1.3173605480968853e-08,
"loss": 0.4705,
"step": 1179
},
{
"epoch": 0.9307828830605404,
"grad_norm": 2.056640010782438,
"learning_rate": 1.2883741868179588e-08,
"loss": 0.4756,
"step": 1180
},
{
"epoch": 0.9315716821139814,
"grad_norm": 2.1404728994758777,
"learning_rate": 1.259706119477516e-08,
"loss": 0.4925,
"step": 1181
},
{
"epoch": 0.9323604811674227,
"grad_norm": 2.05729487323212,
"learning_rate": 1.2313565333998633e-08,
"loss": 0.45,
"step": 1182
},
{
"epoch": 0.9331492802208637,
"grad_norm": 2.3290532910134516,
"learning_rate": 1.2033256138282876e-08,
"loss": 0.4611,
"step": 1183
},
{
"epoch": 0.9339380792743048,
"grad_norm": 2.2167031119668725,
"learning_rate": 1.1756135439238091e-08,
"loss": 0.4861,
"step": 1184
},
{
"epoch": 0.934726878327746,
"grad_norm": 2.1830130660521063,
"learning_rate": 1.148220504764047e-08,
"loss": 0.4813,
"step": 1185
},
{
"epoch": 0.9355156773811871,
"grad_norm": 2.5827208533768578,
"learning_rate": 1.1211466753419329e-08,
"loss": 0.473,
"step": 1186
},
{
"epoch": 0.9363044764346283,
"grad_norm": 2.079265821307451,
"learning_rate": 1.0943922325646614e-08,
"loss": 0.5016,
"step": 1187
},
{
"epoch": 0.9370932754880694,
"grad_norm": 2.247572891018767,
"learning_rate": 1.0679573512524353e-08,
"loss": 0.4374,
"step": 1188
},
{
"epoch": 0.9378820745415105,
"grad_norm": 1.8993830240447653,
"learning_rate": 1.0418422041373887e-08,
"loss": 0.475,
"step": 1189
},
{
"epoch": 0.9386708735949517,
"grad_norm": 2.035600000528139,
"learning_rate": 1.016046961862399e-08,
"loss": 0.519,
"step": 1190
},
{
"epoch": 0.9394596726483928,
"grad_norm": 2.370219608261723,
"learning_rate": 9.905717929800494e-09,
"loss": 0.4134,
"step": 1191
},
{
"epoch": 0.940248471701834,
"grad_norm": 2.047288957062214,
"learning_rate": 9.65416863951446e-09,
"loss": 0.4541,
"step": 1192
},
{
"epoch": 0.9410372707552751,
"grad_norm": 2.079818417871965,
"learning_rate": 9.405823391451906e-09,
"loss": 0.468,
"step": 1193
},
{
"epoch": 0.9418260698087162,
"grad_norm": 2.1191337795344505,
"learning_rate": 9.160683808362712e-09,
"loss": 0.4569,
"step": 1194
},
{
"epoch": 0.9426148688621574,
"grad_norm": 2.392347468354819,
"learning_rate": 8.918751492050236e-09,
"loss": 0.488,
"step": 1195
},
{
"epoch": 0.9434036679155985,
"grad_norm": 2.0917130104929904,
"learning_rate": 8.680028023360652e-09,
"loss": 0.5122,
"step": 1196
},
{
"epoch": 0.9441924669690397,
"grad_norm": 2.1732646059569403,
"learning_rate": 8.444514962172744e-09,
"loss": 0.499,
"step": 1197
},
{
"epoch": 0.9449812660224808,
"grad_norm": 2.2418116820706038,
"learning_rate": 8.21221384738785e-09,
"loss": 0.4887,
"step": 1198
},
{
"epoch": 0.9457700650759219,
"grad_norm": 2.067401633441307,
"learning_rate": 7.983126196919376e-09,
"loss": 0.5116,
"step": 1199
},
{
"epoch": 0.9465588641293631,
"grad_norm": 1.9928537848964127,
"learning_rate": 7.757253507683303e-09,
"loss": 0.4831,
"step": 1200
},
{
"epoch": 0.9473476631828042,
"grad_norm": 2.059265311156936,
"learning_rate": 7.53459725558836e-09,
"loss": 0.4578,
"step": 1201
},
{
"epoch": 0.9481364622362454,
"grad_norm": 2.1910644540935085,
"learning_rate": 7.315158895526086e-09,
"loss": 0.4687,
"step": 1202
},
{
"epoch": 0.9489252612896865,
"grad_norm": 2.2367336721936657,
"learning_rate": 7.098939861361508e-09,
"loss": 0.4982,
"step": 1203
},
{
"epoch": 0.9497140603431276,
"grad_norm": 1.9741236780555866,
"learning_rate": 6.885941565923925e-09,
"loss": 0.4866,
"step": 1204
},
{
"epoch": 0.9505028593965688,
"grad_norm": 2.118712502970971,
"learning_rate": 6.676165400997413e-09,
"loss": 0.5016,
"step": 1205
},
{
"epoch": 0.9512916584500098,
"grad_norm": 2.3053438052878246,
"learning_rate": 6.469612737311836e-09,
"loss": 0.48,
"step": 1206
},
{
"epoch": 0.952080457503451,
"grad_norm": 2.6879429976405422,
"learning_rate": 6.266284924533904e-09,
"loss": 0.4626,
"step": 1207
},
{
"epoch": 0.9528692565568921,
"grad_norm": 1.9692449943204864,
"learning_rate": 6.06618329125852e-09,
"loss": 0.4607,
"step": 1208
},
{
"epoch": 0.9536580556103332,
"grad_norm": 2.21106916010321,
"learning_rate": 5.8693091449996126e-09,
"loss": 0.4859,
"step": 1209
},
{
"epoch": 0.9544468546637744,
"grad_norm": 2.301378135152648,
"learning_rate": 5.675663772182093e-09,
"loss": 0.4454,
"step": 1210
},
{
"epoch": 0.9552356537172155,
"grad_norm": 2.185039853181121,
"learning_rate": 5.48524843813325e-09,
"loss": 0.4661,
"step": 1211
},
{
"epoch": 0.9560244527706567,
"grad_norm": 2.338507164928051,
"learning_rate": 5.29806438707453e-09,
"loss": 0.4419,
"step": 1212
},
{
"epoch": 0.9568132518240978,
"grad_norm": 1.9490135345970243,
"learning_rate": 5.1141128421132166e-09,
"loss": 0.454,
"step": 1213
},
{
"epoch": 0.9576020508775389,
"grad_norm": 2.0578252754591797,
"learning_rate": 4.933395005234709e-09,
"loss": 0.4425,
"step": 1214
},
{
"epoch": 0.9583908499309801,
"grad_norm": 2.231422102924521,
"learning_rate": 4.755912057294531e-09,
"loss": 0.4782,
"step": 1215
},
{
"epoch": 0.9591796489844212,
"grad_norm": 2.336576060112637,
"learning_rate": 4.581665158010617e-09,
"loss": 0.4326,
"step": 1216
},
{
"epoch": 0.9599684480378624,
"grad_norm": 2.294997475954355,
"learning_rate": 4.410655445955758e-09,
"loss": 0.4397,
"step": 1217
},
{
"epoch": 0.9607572470913035,
"grad_norm": 2.5214719115866626,
"learning_rate": 4.24288403855022e-09,
"loss": 0.4507,
"step": 1218
},
{
"epoch": 0.9615460461447446,
"grad_norm": 2.2746728258030595,
"learning_rate": 4.0783520320542e-09,
"loss": 0.4685,
"step": 1219
},
{
"epoch": 0.9623348451981858,
"grad_norm": 2.4869708999825684,
"learning_rate": 3.917060501561043e-09,
"loss": 0.5067,
"step": 1220
},
{
"epoch": 0.9631236442516269,
"grad_norm": 2.18717863028764,
"learning_rate": 3.759010500989868e-09,
"loss": 0.4572,
"step": 1221
},
{
"epoch": 0.963912443305068,
"grad_norm": 2.3869918692157692,
"learning_rate": 3.604203063078959e-09,
"loss": 0.4405,
"step": 1222
},
{
"epoch": 0.9647012423585092,
"grad_norm": 2.177219563280282,
"learning_rate": 3.4526391993787153e-09,
"loss": 0.4909,
"step": 1223
},
{
"epoch": 0.9654900414119503,
"grad_norm": 2.017378045563056,
"learning_rate": 3.304319900245378e-09,
"loss": 0.4554,
"step": 1224
},
{
"epoch": 0.9662788404653915,
"grad_norm": 2.394087370691006,
"learning_rate": 3.159246134834259e-09,
"loss": 0.4358,
"step": 1225
},
{
"epoch": 0.9670676395188326,
"grad_norm": 2.2192875307106235,
"learning_rate": 3.017418851093634e-09,
"loss": 0.4279,
"step": 1226
},
{
"epoch": 0.9678564385722737,
"grad_norm": 2.080102213112167,
"learning_rate": 2.878838975758413e-09,
"loss": 0.4536,
"step": 1227
},
{
"epoch": 0.9686452376257149,
"grad_norm": 2.004955012395581,
"learning_rate": 2.7435074143442595e-09,
"loss": 0.4711,
"step": 1228
},
{
"epoch": 0.969434036679156,
"grad_norm": 2.047762065634272,
"learning_rate": 2.611425051141203e-09,
"loss": 0.5207,
"step": 1229
},
{
"epoch": 0.9702228357325972,
"grad_norm": 2.2664328343749545,
"learning_rate": 2.482592749208534e-09,
"loss": 0.4548,
"step": 1230
},
{
"epoch": 0.9710116347860382,
"grad_norm": 2.1815456484852147,
"learning_rate": 2.3570113503687537e-09,
"loss": 0.4755,
"step": 1231
},
{
"epoch": 0.9718004338394793,
"grad_norm": 2.040796426399282,
"learning_rate": 2.234681675202077e-09,
"loss": 0.4469,
"step": 1232
},
{
"epoch": 0.9725892328929205,
"grad_norm": 2.0086558410483697,
"learning_rate": 2.1156045230411033e-09,
"loss": 0.4519,
"step": 1233
},
{
"epoch": 0.9733780319463616,
"grad_norm": 2.1562000349978776,
"learning_rate": 1.9997806719658226e-09,
"loss": 0.4543,
"step": 1234
},
{
"epoch": 0.9741668309998028,
"grad_norm": 2.1206020382882422,
"learning_rate": 1.8872108787981735e-09,
"loss": 0.4957,
"step": 1235
},
{
"epoch": 0.9749556300532439,
"grad_norm": 2.0939171623261075,
"learning_rate": 1.7778958790972687e-09,
"loss": 0.5027,
"step": 1236
},
{
"epoch": 0.975744429106685,
"grad_norm": 2.4533745913330955,
"learning_rate": 1.6718363871547902e-09,
"loss": 0.438,
"step": 1237
},
{
"epoch": 0.9765332281601262,
"grad_norm": 2.0128392356567804,
"learning_rate": 1.5690330959898245e-09,
"loss": 0.461,
"step": 1238
},
{
"epoch": 0.9773220272135673,
"grad_norm": 2.0454039424068062,
"learning_rate": 1.4694866773448112e-09,
"loss": 0.5271,
"step": 1239
},
{
"epoch": 0.9781108262670085,
"grad_norm": 2.3256606305815244,
"learning_rate": 1.3731977816809348e-09,
"loss": 0.5093,
"step": 1240
},
{
"epoch": 0.9788996253204496,
"grad_norm": 2.0901288689109925,
"learning_rate": 1.2801670381739072e-09,
"loss": 0.5044,
"step": 1241
},
{
"epoch": 0.9796884243738907,
"grad_norm": 2.108049449466968,
"learning_rate": 1.190395054709914e-09,
"loss": 0.4702,
"step": 1242
},
{
"epoch": 0.9804772234273319,
"grad_norm": 2.167558665053736,
"learning_rate": 1.103882417881452e-09,
"loss": 0.5112,
"step": 1243
},
{
"epoch": 0.981266022480773,
"grad_norm": 2.280985570328114,
"learning_rate": 1.020629692983721e-09,
"loss": 0.4038,
"step": 1244
},
{
"epoch": 0.9820548215342142,
"grad_norm": 2.244284434471773,
"learning_rate": 9.406374240109593e-10,
"loss": 0.4953,
"step": 1245
},
{
"epoch": 0.9828436205876553,
"grad_norm": 2.042103204031043,
"learning_rate": 8.639061336524478e-10,
"loss": 0.4716,
"step": 1246
},
{
"epoch": 0.9836324196410964,
"grad_norm": 2.130472917316118,
"learning_rate": 7.904363232897338e-10,
"loss": 0.4861,
"step": 1247
},
{
"epoch": 0.9844212186945376,
"grad_norm": 2.3285216290260777,
"learning_rate": 7.20228472992912e-10,
"loss": 0.4486,
"step": 1248
},
{
"epoch": 0.9852100177479787,
"grad_norm": 2.039367579112449,
"learning_rate": 6.532830415176272e-10,
"loss": 0.4975,
"step": 1249
},
{
"epoch": 0.9859988168014199,
"grad_norm": 1.9676543675802594,
"learning_rate": 5.896004663020205e-10,
"loss": 0.4693,
"step": 1250
},
{
"epoch": 0.986787615854861,
"grad_norm": 2.312234669455349,
"learning_rate": 5.291811634640652e-10,
"loss": 0.4547,
"step": 1251
},
{
"epoch": 0.9875764149083021,
"grad_norm": 2.1818147095198084,
"learning_rate": 4.720255277984586e-10,
"loss": 0.4538,
"step": 1252
},
{
"epoch": 0.9883652139617433,
"grad_norm": 1.9900427688433966,
"learning_rate": 4.1813393277445597e-10,
"loss": 0.4935,
"step": 1253
},
{
"epoch": 0.9891540130151844,
"grad_norm": 2.206854551535642,
"learning_rate": 3.675067305332069e-10,
"loss": 0.4699,
"step": 1254
},
{
"epoch": 0.9899428120686256,
"grad_norm": 1.9706410648385435,
"learning_rate": 3.201442518854791e-10,
"loss": 0.4598,
"step": 1255
},
{
"epoch": 0.9907316111220666,
"grad_norm": 2.225340731753896,
"learning_rate": 2.7604680630965995e-10,
"loss": 0.4751,
"step": 1256
},
{
"epoch": 0.9915204101755077,
"grad_norm": 2.6647283035533103,
"learning_rate": 2.3521468194931395e-10,
"loss": 0.4837,
"step": 1257
},
{
"epoch": 0.9923092092289489,
"grad_norm": 2.4136306109273247,
"learning_rate": 1.9764814561190613e-10,
"loss": 0.4947,
"step": 1258
},
{
"epoch": 0.99309800828239,
"grad_norm": 2.464718850948773,
"learning_rate": 1.633474427664705e-10,
"loss": 0.4432,
"step": 1259
},
{
"epoch": 0.9938868073358312,
"grad_norm": 2.139109020422269,
"learning_rate": 1.3231279754238878e-10,
"loss": 0.4699,
"step": 1260
},
{
"epoch": 0.9946756063892723,
"grad_norm": 2.065302641920417,
"learning_rate": 1.045444127278361e-10,
"loss": 0.4832,
"step": 1261
},
{
"epoch": 0.9954644054427134,
"grad_norm": 2.066829740976868,
"learning_rate": 8.004246976833772e-11,
"loss": 0.5132,
"step": 1262
},
{
"epoch": 0.9962532044961546,
"grad_norm": 2.161177985103768,
"learning_rate": 5.880712876571437e-11,
"loss": 0.4945,
"step": 1263
},
{
"epoch": 0.9970420035495957,
"grad_norm": 1.921652375827139,
"learning_rate": 4.08385284770274e-11,
"loss": 0.4889,
"step": 1264
},
{
"epoch": 0.9978308026030369,
"grad_norm": 1.9616599390350928,
"learning_rate": 2.613678631357974e-11,
"loss": 0.4887,
"step": 1265
},
{
"epoch": 0.998619601656478,
"grad_norm": 2.020972170869089,
"learning_rate": 1.4701998340194145e-11,
"loss": 0.5187,
"step": 1266
},
{
"epoch": 0.9994084007099191,
"grad_norm": 1.9361662167504259,
"learning_rate": 6.534239274658127e-12,
"loss": 0.4318,
"step": 1267
},
{
"epoch": 1.0,
"grad_norm": 1.9361662167504259,
"learning_rate": 1.6335624871688381e-12,
"loss": 0.4392,
"step": 1268
},
{
"epoch": 1.0,
"step": 1268,
"total_flos": 741274716962816.0,
"train_loss": 0.5522690528198747,
"train_runtime": 8573.9813,
"train_samples_per_second": 17.246,
"train_steps_per_second": 0.148
}
],
"logging_steps": 1.0,
"max_steps": 1268,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 741274716962816.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}