Files
OpenR1-Distill-1.5B-ours/trainer_state.json
ModelHub XC f8e7a3d2dc 初始化项目,由ModelHub XC社区提供模型
Model: seele123/OpenR1-Distill-1.5B-ours
Source: Original Platform
2026-05-25 21:10:23 +08:00

15437 lines
373 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 2199,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00136542072025943,
"grad_norm": 27.29060173034668,
"learning_rate": 0.0,
"loss": 11.8282,
"step": 1
},
{
"epoch": 0.00273084144051886,
"grad_norm": 27.101430892944336,
"learning_rate": 6.060606060606061e-07,
"loss": 11.9583,
"step": 2
},
{
"epoch": 0.00409626216077829,
"grad_norm": 27.203603744506836,
"learning_rate": 1.2121212121212122e-06,
"loss": 11.7199,
"step": 3
},
{
"epoch": 0.00546168288103772,
"grad_norm": 27.445079803466797,
"learning_rate": 1.8181818181818183e-06,
"loss": 11.8413,
"step": 4
},
{
"epoch": 0.006827103601297149,
"grad_norm": 26.943281173706055,
"learning_rate": 2.4242424242424244e-06,
"loss": 11.7987,
"step": 5
},
{
"epoch": 0.00819252432155658,
"grad_norm": 26.69304847717285,
"learning_rate": 3.0303030303030305e-06,
"loss": 11.7112,
"step": 6
},
{
"epoch": 0.009557945041816009,
"grad_norm": 26.09612464904785,
"learning_rate": 3.6363636363636366e-06,
"loss": 11.6788,
"step": 7
},
{
"epoch": 0.01092336576207544,
"grad_norm": 25.46772003173828,
"learning_rate": 4.242424242424243e-06,
"loss": 11.6706,
"step": 8
},
{
"epoch": 0.01228878648233487,
"grad_norm": 22.130035400390625,
"learning_rate": 4.848484848484849e-06,
"loss": 11.147,
"step": 9
},
{
"epoch": 0.013654207202594299,
"grad_norm": 22.017324447631836,
"learning_rate": 5.4545454545454545e-06,
"loss": 10.8546,
"step": 10
},
{
"epoch": 0.01501962792285373,
"grad_norm": 21.423377990722656,
"learning_rate": 6.060606060606061e-06,
"loss": 10.8027,
"step": 11
},
{
"epoch": 0.01638504864311316,
"grad_norm": 17.77073860168457,
"learning_rate": 6.666666666666667e-06,
"loss": 9.4719,
"step": 12
},
{
"epoch": 0.01775046936337259,
"grad_norm": 18.365257263183594,
"learning_rate": 7.272727272727273e-06,
"loss": 9.4704,
"step": 13
},
{
"epoch": 0.019115890083632018,
"grad_norm": 17.95745277404785,
"learning_rate": 7.87878787878788e-06,
"loss": 9.1954,
"step": 14
},
{
"epoch": 0.02048131080389145,
"grad_norm": 26.231185913085938,
"learning_rate": 8.484848484848486e-06,
"loss": 8.7755,
"step": 15
},
{
"epoch": 0.02184673152415088,
"grad_norm": 82.10050964355469,
"learning_rate": 9.090909090909091e-06,
"loss": 7.2191,
"step": 16
},
{
"epoch": 0.023212152244410308,
"grad_norm": 22.369043350219727,
"learning_rate": 9.696969696969698e-06,
"loss": 6.9942,
"step": 17
},
{
"epoch": 0.02457757296466974,
"grad_norm": 22.22688865661621,
"learning_rate": 1.0303030303030304e-05,
"loss": 6.4808,
"step": 18
},
{
"epoch": 0.02594299368492917,
"grad_norm": 14.63390827178955,
"learning_rate": 1.0909090909090909e-05,
"loss": 6.0061,
"step": 19
},
{
"epoch": 0.027308414405188598,
"grad_norm": 22.937284469604492,
"learning_rate": 1.1515151515151517e-05,
"loss": 5.5741,
"step": 20
},
{
"epoch": 0.02867383512544803,
"grad_norm": 11.154397964477539,
"learning_rate": 1.2121212121212122e-05,
"loss": 5.3143,
"step": 21
},
{
"epoch": 0.03003925584570746,
"grad_norm": 7.8057379722595215,
"learning_rate": 1.2727272727272728e-05,
"loss": 4.833,
"step": 22
},
{
"epoch": 0.03140467656596689,
"grad_norm": 6.351495265960693,
"learning_rate": 1.3333333333333333e-05,
"loss": 4.6488,
"step": 23
},
{
"epoch": 0.03277009728622632,
"grad_norm": 4.073518753051758,
"learning_rate": 1.3939393939393942e-05,
"loss": 4.6527,
"step": 24
},
{
"epoch": 0.034135518006485746,
"grad_norm": 4.10856819152832,
"learning_rate": 1.4545454545454546e-05,
"loss": 4.6148,
"step": 25
},
{
"epoch": 0.03550093872674518,
"grad_norm": 3.631073474884033,
"learning_rate": 1.5151515151515153e-05,
"loss": 4.523,
"step": 26
},
{
"epoch": 0.03686635944700461,
"grad_norm": 3.0445265769958496,
"learning_rate": 1.575757575757576e-05,
"loss": 4.2615,
"step": 27
},
{
"epoch": 0.038231780167264036,
"grad_norm": 3.002786874771118,
"learning_rate": 1.6363636363636366e-05,
"loss": 4.2207,
"step": 28
},
{
"epoch": 0.03959720088752347,
"grad_norm": 2.9761602878570557,
"learning_rate": 1.6969696969696972e-05,
"loss": 4.2448,
"step": 29
},
{
"epoch": 0.0409626216077829,
"grad_norm": 3.0609021186828613,
"learning_rate": 1.7575757575757576e-05,
"loss": 4.3805,
"step": 30
},
{
"epoch": 0.042328042328042326,
"grad_norm": 2.874307870864868,
"learning_rate": 1.8181818181818182e-05,
"loss": 4.1549,
"step": 31
},
{
"epoch": 0.04369346304830176,
"grad_norm": 2.971336603164673,
"learning_rate": 1.8787878787878792e-05,
"loss": 4.2545,
"step": 32
},
{
"epoch": 0.04505888376856119,
"grad_norm": 2.9191112518310547,
"learning_rate": 1.9393939393939395e-05,
"loss": 4.1873,
"step": 33
},
{
"epoch": 0.046424304488820615,
"grad_norm": 2.93782114982605,
"learning_rate": 2e-05,
"loss": 4.054,
"step": 34
},
{
"epoch": 0.04778972520908005,
"grad_norm": 2.930889844894409,
"learning_rate": 2.0606060606060608e-05,
"loss": 3.9244,
"step": 35
},
{
"epoch": 0.04915514592933948,
"grad_norm": 2.9650650024414062,
"learning_rate": 2.121212121212121e-05,
"loss": 3.9049,
"step": 36
},
{
"epoch": 0.050520566649598905,
"grad_norm": 2.910654067993164,
"learning_rate": 2.1818181818181818e-05,
"loss": 4.0147,
"step": 37
},
{
"epoch": 0.05188598736985834,
"grad_norm": 2.877976179122925,
"learning_rate": 2.2424242424242424e-05,
"loss": 3.9195,
"step": 38
},
{
"epoch": 0.05325140809011777,
"grad_norm": 2.840790271759033,
"learning_rate": 2.3030303030303034e-05,
"loss": 3.8173,
"step": 39
},
{
"epoch": 0.054616828810377195,
"grad_norm": 2.8126590251922607,
"learning_rate": 2.363636363636364e-05,
"loss": 3.8882,
"step": 40
},
{
"epoch": 0.05598224953063663,
"grad_norm": 2.8062613010406494,
"learning_rate": 2.4242424242424244e-05,
"loss": 3.8439,
"step": 41
},
{
"epoch": 0.05734767025089606,
"grad_norm": 3.009477376937866,
"learning_rate": 2.484848484848485e-05,
"loss": 3.7638,
"step": 42
},
{
"epoch": 0.058713090971155485,
"grad_norm": 2.8368887901306152,
"learning_rate": 2.5454545454545457e-05,
"loss": 3.8081,
"step": 43
},
{
"epoch": 0.06007851169141492,
"grad_norm": 2.742190361022949,
"learning_rate": 2.6060606060606063e-05,
"loss": 3.8122,
"step": 44
},
{
"epoch": 0.06144393241167435,
"grad_norm": 2.91174054145813,
"learning_rate": 2.6666666666666667e-05,
"loss": 3.7679,
"step": 45
},
{
"epoch": 0.06280935313193378,
"grad_norm": 2.898702621459961,
"learning_rate": 2.7272727272727273e-05,
"loss": 3.8033,
"step": 46
},
{
"epoch": 0.06417477385219321,
"grad_norm": 2.9040398597717285,
"learning_rate": 2.7878787878787883e-05,
"loss": 3.7048,
"step": 47
},
{
"epoch": 0.06554019457245264,
"grad_norm": 2.885359048843384,
"learning_rate": 2.848484848484849e-05,
"loss": 3.7252,
"step": 48
},
{
"epoch": 0.06690561529271206,
"grad_norm": 2.8904311656951904,
"learning_rate": 2.9090909090909093e-05,
"loss": 3.8476,
"step": 49
},
{
"epoch": 0.06827103601297149,
"grad_norm": 2.7308733463287354,
"learning_rate": 2.96969696969697e-05,
"loss": 3.5271,
"step": 50
},
{
"epoch": 0.06963645673323093,
"grad_norm": 2.8446156978607178,
"learning_rate": 3.0303030303030306e-05,
"loss": 3.8744,
"step": 51
},
{
"epoch": 0.07100187745349036,
"grad_norm": 2.8620967864990234,
"learning_rate": 3.090909090909091e-05,
"loss": 3.6392,
"step": 52
},
{
"epoch": 0.07236729817374979,
"grad_norm": 2.812472105026245,
"learning_rate": 3.151515151515152e-05,
"loss": 3.7459,
"step": 53
},
{
"epoch": 0.07373271889400922,
"grad_norm": 2.9046828746795654,
"learning_rate": 3.212121212121212e-05,
"loss": 3.6442,
"step": 54
},
{
"epoch": 0.07509813961426864,
"grad_norm": 2.8417768478393555,
"learning_rate": 3.272727272727273e-05,
"loss": 3.6122,
"step": 55
},
{
"epoch": 0.07646356033452807,
"grad_norm": 2.7872884273529053,
"learning_rate": 3.3333333333333335e-05,
"loss": 3.5001,
"step": 56
},
{
"epoch": 0.07782898105478751,
"grad_norm": 2.7673394680023193,
"learning_rate": 3.3939393939393945e-05,
"loss": 3.5532,
"step": 57
},
{
"epoch": 0.07919440177504694,
"grad_norm": 2.7456703186035156,
"learning_rate": 3.454545454545455e-05,
"loss": 3.5906,
"step": 58
},
{
"epoch": 0.08055982249530637,
"grad_norm": 2.8259494304656982,
"learning_rate": 3.515151515151515e-05,
"loss": 3.5843,
"step": 59
},
{
"epoch": 0.0819252432155658,
"grad_norm": 2.6943252086639404,
"learning_rate": 3.575757575757576e-05,
"loss": 3.3294,
"step": 60
},
{
"epoch": 0.08329066393582522,
"grad_norm": 2.7992498874664307,
"learning_rate": 3.6363636363636364e-05,
"loss": 3.5269,
"step": 61
},
{
"epoch": 0.08465608465608465,
"grad_norm": 2.731005907058716,
"learning_rate": 3.6969696969696974e-05,
"loss": 3.4626,
"step": 62
},
{
"epoch": 0.08602150537634409,
"grad_norm": 2.6247339248657227,
"learning_rate": 3.7575757575757584e-05,
"loss": 3.5402,
"step": 63
},
{
"epoch": 0.08738692609660352,
"grad_norm": 2.6144378185272217,
"learning_rate": 3.818181818181819e-05,
"loss": 3.4253,
"step": 64
},
{
"epoch": 0.08875234681686295,
"grad_norm": 2.5079870223999023,
"learning_rate": 3.878787878787879e-05,
"loss": 3.3501,
"step": 65
},
{
"epoch": 0.09011776753712238,
"grad_norm": 2.4330976009368896,
"learning_rate": 3.93939393939394e-05,
"loss": 3.4232,
"step": 66
},
{
"epoch": 0.0914831882573818,
"grad_norm": 2.2991628646850586,
"learning_rate": 4e-05,
"loss": 3.3386,
"step": 67
},
{
"epoch": 0.09284860897764123,
"grad_norm": 2.1482725143432617,
"learning_rate": 3.999998047637464e-05,
"loss": 3.3887,
"step": 68
},
{
"epoch": 0.09421402969790067,
"grad_norm": 2.032754898071289,
"learning_rate": 3.99999219055409e-05,
"loss": 3.2551,
"step": 69
},
{
"epoch": 0.0955794504181601,
"grad_norm": 1.8964202404022217,
"learning_rate": 3.9999824287625855e-05,
"loss": 3.4667,
"step": 70
},
{
"epoch": 0.09694487113841953,
"grad_norm": 1.8003062009811401,
"learning_rate": 3.999968762284124e-05,
"loss": 3.39,
"step": 71
},
{
"epoch": 0.09831029185867896,
"grad_norm": 1.703377366065979,
"learning_rate": 3.999951191148356e-05,
"loss": 3.3535,
"step": 72
},
{
"epoch": 0.09967571257893838,
"grad_norm": 1.7009670734405518,
"learning_rate": 3.999929715393395e-05,
"loss": 3.3069,
"step": 73
},
{
"epoch": 0.10104113329919781,
"grad_norm": 1.6436004638671875,
"learning_rate": 3.999904335065829e-05,
"loss": 3.3685,
"step": 74
},
{
"epoch": 0.10240655401945725,
"grad_norm": 1.7554755210876465,
"learning_rate": 3.999875050220715e-05,
"loss": 3.4039,
"step": 75
},
{
"epoch": 0.10377197473971668,
"grad_norm": 1.6258020401000977,
"learning_rate": 3.9998418609215826e-05,
"loss": 3.1369,
"step": 76
},
{
"epoch": 0.10513739545997611,
"grad_norm": 1.6664680242538452,
"learning_rate": 3.9998047672404265e-05,
"loss": 3.1407,
"step": 77
},
{
"epoch": 0.10650281618023553,
"grad_norm": 1.6535282135009766,
"learning_rate": 3.9997637692577146e-05,
"loss": 3.276,
"step": 78
},
{
"epoch": 0.10786823690049496,
"grad_norm": 1.5287853479385376,
"learning_rate": 3.999718867062384e-05,
"loss": 3.1718,
"step": 79
},
{
"epoch": 0.10923365762075439,
"grad_norm": 1.6630401611328125,
"learning_rate": 3.99967006075184e-05,
"loss": 3.5108,
"step": 80
},
{
"epoch": 0.11059907834101383,
"grad_norm": 1.627175211906433,
"learning_rate": 3.999617350431958e-05,
"loss": 3.0597,
"step": 81
},
{
"epoch": 0.11196449906127326,
"grad_norm": 1.8537323474884033,
"learning_rate": 3.9995607362170823e-05,
"loss": 3.0389,
"step": 82
},
{
"epoch": 0.11332991978153269,
"grad_norm": 1.7223312854766846,
"learning_rate": 3.999500218230025e-05,
"loss": 3.2699,
"step": 83
},
{
"epoch": 0.11469534050179211,
"grad_norm": 1.7203669548034668,
"learning_rate": 3.999435796602068e-05,
"loss": 3.1521,
"step": 84
},
{
"epoch": 0.11606076122205154,
"grad_norm": 1.855812430381775,
"learning_rate": 3.99936747147296e-05,
"loss": 3.286,
"step": 85
},
{
"epoch": 0.11742618194231097,
"grad_norm": 1.630742073059082,
"learning_rate": 3.999295242990918e-05,
"loss": 3.1285,
"step": 86
},
{
"epoch": 0.11879160266257041,
"grad_norm": 1.834832787513733,
"learning_rate": 3.999219111312627e-05,
"loss": 3.374,
"step": 87
},
{
"epoch": 0.12015702338282984,
"grad_norm": 1.5942246913909912,
"learning_rate": 3.99913907660324e-05,
"loss": 3.1751,
"step": 88
},
{
"epoch": 0.12152244410308927,
"grad_norm": 1.8591982126235962,
"learning_rate": 3.9990551390363735e-05,
"loss": 3.3035,
"step": 89
},
{
"epoch": 0.1228878648233487,
"grad_norm": 1.6944773197174072,
"learning_rate": 3.998967298794113e-05,
"loss": 3.3072,
"step": 90
},
{
"epoch": 0.12425328554360812,
"grad_norm": 1.7387422323226929,
"learning_rate": 3.998875556067011e-05,
"loss": 3.3605,
"step": 91
},
{
"epoch": 0.12561870626386756,
"grad_norm": 1.5652858018875122,
"learning_rate": 3.9987799110540836e-05,
"loss": 3.2174,
"step": 92
},
{
"epoch": 0.12698412698412698,
"grad_norm": 1.6528273820877075,
"learning_rate": 3.9986803639628124e-05,
"loss": 3.2931,
"step": 93
},
{
"epoch": 0.12834954770438642,
"grad_norm": 1.698073387145996,
"learning_rate": 3.9985769150091434e-05,
"loss": 3.2661,
"step": 94
},
{
"epoch": 0.12971496842464583,
"grad_norm": 1.7020331621170044,
"learning_rate": 3.998469564417489e-05,
"loss": 3.0376,
"step": 95
},
{
"epoch": 0.13108038914490527,
"grad_norm": 1.5728801488876343,
"learning_rate": 3.998358312420723e-05,
"loss": 3.1267,
"step": 96
},
{
"epoch": 0.13244580986516472,
"grad_norm": 1.5581607818603516,
"learning_rate": 3.998243159260184e-05,
"loss": 3.3959,
"step": 97
},
{
"epoch": 0.13381123058542413,
"grad_norm": 1.584854006767273,
"learning_rate": 3.9981241051856734e-05,
"loss": 3.1853,
"step": 98
},
{
"epoch": 0.13517665130568357,
"grad_norm": 1.5182476043701172,
"learning_rate": 3.998001150455452e-05,
"loss": 3.237,
"step": 99
},
{
"epoch": 0.13654207202594298,
"grad_norm": 1.5051459074020386,
"learning_rate": 3.997874295336246e-05,
"loss": 3.1649,
"step": 100
},
{
"epoch": 0.13790749274620243,
"grad_norm": 1.522617220878601,
"learning_rate": 3.9977435401032415e-05,
"loss": 3.0658,
"step": 101
},
{
"epoch": 0.13927291346646187,
"grad_norm": 1.6695613861083984,
"learning_rate": 3.997608885040083e-05,
"loss": 3.2961,
"step": 102
},
{
"epoch": 0.14063833418672128,
"grad_norm": 1.5407582521438599,
"learning_rate": 3.997470330438879e-05,
"loss": 3.276,
"step": 103
},
{
"epoch": 0.14200375490698072,
"grad_norm": 1.5296891927719116,
"learning_rate": 3.997327876600194e-05,
"loss": 3.2126,
"step": 104
},
{
"epoch": 0.14336917562724014,
"grad_norm": 1.6092779636383057,
"learning_rate": 3.997181523833051e-05,
"loss": 3.1453,
"step": 105
},
{
"epoch": 0.14473459634749958,
"grad_norm": 1.5463693141937256,
"learning_rate": 3.9970312724549325e-05,
"loss": 3.1401,
"step": 106
},
{
"epoch": 0.146100017067759,
"grad_norm": 1.5076404809951782,
"learning_rate": 3.996877122791778e-05,
"loss": 3.1713,
"step": 107
},
{
"epoch": 0.14746543778801843,
"grad_norm": 1.5350888967514038,
"learning_rate": 3.9967190751779815e-05,
"loss": 3.3033,
"step": 108
},
{
"epoch": 0.14883085850827787,
"grad_norm": 1.61422598361969,
"learning_rate": 3.9965571299563966e-05,
"loss": 3.2258,
"step": 109
},
{
"epoch": 0.1501962792285373,
"grad_norm": 1.4621989727020264,
"learning_rate": 3.9963912874783276e-05,
"loss": 3.1758,
"step": 110
},
{
"epoch": 0.15156169994879673,
"grad_norm": 1.5220097303390503,
"learning_rate": 3.9962215481035374e-05,
"loss": 3.0574,
"step": 111
},
{
"epoch": 0.15292712066905614,
"grad_norm": 1.5093547105789185,
"learning_rate": 3.9960479122002384e-05,
"loss": 3.0645,
"step": 112
},
{
"epoch": 0.15429254138931559,
"grad_norm": 1.6433111429214478,
"learning_rate": 3.995870380145098e-05,
"loss": 3.2438,
"step": 113
},
{
"epoch": 0.15565796210957503,
"grad_norm": 1.621281385421753,
"learning_rate": 3.995688952323236e-05,
"loss": 3.3748,
"step": 114
},
{
"epoch": 0.15702338282983444,
"grad_norm": 1.5065555572509766,
"learning_rate": 3.99550362912822e-05,
"loss": 3.2262,
"step": 115
},
{
"epoch": 0.15838880355009388,
"grad_norm": 1.5167025327682495,
"learning_rate": 3.995314410962073e-05,
"loss": 3.0886,
"step": 116
},
{
"epoch": 0.1597542242703533,
"grad_norm": 1.4831392765045166,
"learning_rate": 3.995121298235262e-05,
"loss": 3.1595,
"step": 117
},
{
"epoch": 0.16111964499061274,
"grad_norm": 1.4959598779678345,
"learning_rate": 3.9949242913667066e-05,
"loss": 3.23,
"step": 118
},
{
"epoch": 0.16248506571087215,
"grad_norm": 1.431364893913269,
"learning_rate": 3.994723390783771e-05,
"loss": 2.9825,
"step": 119
},
{
"epoch": 0.1638504864311316,
"grad_norm": 1.5354398488998413,
"learning_rate": 3.994518596922267e-05,
"loss": 3.122,
"step": 120
},
{
"epoch": 0.16521590715139103,
"grad_norm": 1.508204460144043,
"learning_rate": 3.994309910226453e-05,
"loss": 2.9463,
"step": 121
},
{
"epoch": 0.16658132787165045,
"grad_norm": 1.4888150691986084,
"learning_rate": 3.994097331149031e-05,
"loss": 3.0929,
"step": 122
},
{
"epoch": 0.1679467485919099,
"grad_norm": 1.4554660320281982,
"learning_rate": 3.9938808601511475e-05,
"loss": 3.1214,
"step": 123
},
{
"epoch": 0.1693121693121693,
"grad_norm": 1.4815845489501953,
"learning_rate": 3.99366049770239e-05,
"loss": 3.203,
"step": 124
},
{
"epoch": 0.17067759003242874,
"grad_norm": 1.4656120538711548,
"learning_rate": 3.9934362442807904e-05,
"loss": 3.1883,
"step": 125
},
{
"epoch": 0.17204301075268819,
"grad_norm": 1.4258489608764648,
"learning_rate": 3.993208100372818e-05,
"loss": 3.0039,
"step": 126
},
{
"epoch": 0.1734084314729476,
"grad_norm": 1.4694154262542725,
"learning_rate": 3.992976066473386e-05,
"loss": 3.221,
"step": 127
},
{
"epoch": 0.17477385219320704,
"grad_norm": 1.4860414266586304,
"learning_rate": 3.992740143085842e-05,
"loss": 3.0728,
"step": 128
},
{
"epoch": 0.17613927291346645,
"grad_norm": 1.6058411598205566,
"learning_rate": 3.992500330721973e-05,
"loss": 3.2692,
"step": 129
},
{
"epoch": 0.1775046936337259,
"grad_norm": 1.4404314756393433,
"learning_rate": 3.9922566299020015e-05,
"loss": 3.1371,
"step": 130
},
{
"epoch": 0.1788701143539853,
"grad_norm": 1.395689606666565,
"learning_rate": 3.9920090411545855e-05,
"loss": 3.1913,
"step": 131
},
{
"epoch": 0.18023553507424475,
"grad_norm": 1.3928699493408203,
"learning_rate": 3.991757565016818e-05,
"loss": 3.2454,
"step": 132
},
{
"epoch": 0.1816009557945042,
"grad_norm": 1.4996368885040283,
"learning_rate": 3.991502202034225e-05,
"loss": 3.1346,
"step": 133
},
{
"epoch": 0.1829663765147636,
"grad_norm": 1.3858879804611206,
"learning_rate": 3.991242952760761e-05,
"loss": 3.0907,
"step": 134
},
{
"epoch": 0.18433179723502305,
"grad_norm": 1.3553260564804077,
"learning_rate": 3.990979817758815e-05,
"loss": 3.0519,
"step": 135
},
{
"epoch": 0.18569721795528246,
"grad_norm": 1.4780244827270508,
"learning_rate": 3.990712797599202e-05,
"loss": 3.1632,
"step": 136
},
{
"epoch": 0.1870626386755419,
"grad_norm": 1.4284120798110962,
"learning_rate": 3.9904418928611686e-05,
"loss": 3.2388,
"step": 137
},
{
"epoch": 0.18842805939580134,
"grad_norm": 1.4712940454483032,
"learning_rate": 3.9901671041323856e-05,
"loss": 3.1466,
"step": 138
},
{
"epoch": 0.18979348011606076,
"grad_norm": 1.5241836309432983,
"learning_rate": 3.989888432008948e-05,
"loss": 3.176,
"step": 139
},
{
"epoch": 0.1911589008363202,
"grad_norm": 1.4123114347457886,
"learning_rate": 3.98960587709538e-05,
"loss": 3.2112,
"step": 140
},
{
"epoch": 0.19252432155657961,
"grad_norm": 1.4714604616165161,
"learning_rate": 3.989319440004623e-05,
"loss": 2.9738,
"step": 141
},
{
"epoch": 0.19388974227683906,
"grad_norm": 1.465427041053772,
"learning_rate": 3.9890291213580444e-05,
"loss": 3.134,
"step": 142
},
{
"epoch": 0.19525516299709847,
"grad_norm": 1.4285204410552979,
"learning_rate": 3.988734921785429e-05,
"loss": 3.1983,
"step": 143
},
{
"epoch": 0.1966205837173579,
"grad_norm": 1.4719756841659546,
"learning_rate": 3.988436841924982e-05,
"loss": 3.0901,
"step": 144
},
{
"epoch": 0.19798600443761735,
"grad_norm": 1.4046169519424438,
"learning_rate": 3.988134882423326e-05,
"loss": 3.1468,
"step": 145
},
{
"epoch": 0.19935142515787677,
"grad_norm": 1.434508204460144,
"learning_rate": 3.987829043935498e-05,
"loss": 3.1504,
"step": 146
},
{
"epoch": 0.2007168458781362,
"grad_norm": 1.4397052526474,
"learning_rate": 3.987519327124951e-05,
"loss": 3.0815,
"step": 147
},
{
"epoch": 0.20208226659839562,
"grad_norm": 1.3975492715835571,
"learning_rate": 3.9872057326635525e-05,
"loss": 3.2318,
"step": 148
},
{
"epoch": 0.20344768731865506,
"grad_norm": 1.4486109018325806,
"learning_rate": 3.986888261231579e-05,
"loss": 3.1083,
"step": 149
},
{
"epoch": 0.2048131080389145,
"grad_norm": 1.4100350141525269,
"learning_rate": 3.98656691351772e-05,
"loss": 3.0157,
"step": 150
},
{
"epoch": 0.20617852875917392,
"grad_norm": 1.4731138944625854,
"learning_rate": 3.9862416902190706e-05,
"loss": 3.1266,
"step": 151
},
{
"epoch": 0.20754394947943336,
"grad_norm": 1.3232550621032715,
"learning_rate": 3.985912592041136e-05,
"loss": 2.9004,
"step": 152
},
{
"epoch": 0.20890937019969277,
"grad_norm": 1.5442419052124023,
"learning_rate": 3.9855796196978265e-05,
"loss": 3.015,
"step": 153
},
{
"epoch": 0.21027479091995221,
"grad_norm": 1.52120041847229,
"learning_rate": 3.985242773911456e-05,
"loss": 3.2035,
"step": 154
},
{
"epoch": 0.21164021164021163,
"grad_norm": 1.4759293794631958,
"learning_rate": 3.9849020554127415e-05,
"loss": 3.0335,
"step": 155
},
{
"epoch": 0.21300563236047107,
"grad_norm": 1.4842160940170288,
"learning_rate": 3.9845574649407996e-05,
"loss": 3.1482,
"step": 156
},
{
"epoch": 0.2143710530807305,
"grad_norm": 1.6772977113723755,
"learning_rate": 3.984209003243149e-05,
"loss": 3.1364,
"step": 157
},
{
"epoch": 0.21573647380098993,
"grad_norm": 1.5249519348144531,
"learning_rate": 3.9838566710757044e-05,
"loss": 3.3697,
"step": 158
},
{
"epoch": 0.21710189452124937,
"grad_norm": 1.5777902603149414,
"learning_rate": 3.983500469202776e-05,
"loss": 3.1309,
"step": 159
},
{
"epoch": 0.21846731524150878,
"grad_norm": 1.5024640560150146,
"learning_rate": 3.983140398397071e-05,
"loss": 3.1293,
"step": 160
},
{
"epoch": 0.21983273596176822,
"grad_norm": 1.5011036396026611,
"learning_rate": 3.982776459439687e-05,
"loss": 3.0945,
"step": 161
},
{
"epoch": 0.22119815668202766,
"grad_norm": 1.5172325372695923,
"learning_rate": 3.982408653120115e-05,
"loss": 3.1207,
"step": 162
},
{
"epoch": 0.22256357740228708,
"grad_norm": 1.501692771911621,
"learning_rate": 3.9820369802362325e-05,
"loss": 3.1473,
"step": 163
},
{
"epoch": 0.22392899812254652,
"grad_norm": 1.4966915845870972,
"learning_rate": 3.981661441594308e-05,
"loss": 3.054,
"step": 164
},
{
"epoch": 0.22529441884280593,
"grad_norm": 1.5072288513183594,
"learning_rate": 3.981282038008993e-05,
"loss": 3.0506,
"step": 165
},
{
"epoch": 0.22665983956306537,
"grad_norm": 1.4281047582626343,
"learning_rate": 3.9808987703033256e-05,
"loss": 2.9523,
"step": 166
},
{
"epoch": 0.2280252602833248,
"grad_norm": 1.466499924659729,
"learning_rate": 3.980511639308724e-05,
"loss": 3.0773,
"step": 167
},
{
"epoch": 0.22939068100358423,
"grad_norm": 1.4500519037246704,
"learning_rate": 3.980120645864991e-05,
"loss": 3.1234,
"step": 168
},
{
"epoch": 0.23075610172384367,
"grad_norm": 1.3653678894042969,
"learning_rate": 3.979725790820302e-05,
"loss": 3.0479,
"step": 169
},
{
"epoch": 0.23212152244410308,
"grad_norm": 1.4903764724731445,
"learning_rate": 3.979327075031215e-05,
"loss": 3.1529,
"step": 170
},
{
"epoch": 0.23348694316436253,
"grad_norm": 1.5020712614059448,
"learning_rate": 3.97892449936266e-05,
"loss": 3.1186,
"step": 171
},
{
"epoch": 0.23485236388462194,
"grad_norm": 1.5594592094421387,
"learning_rate": 3.9785180646879406e-05,
"loss": 3.063,
"step": 172
},
{
"epoch": 0.23621778460488138,
"grad_norm": 1.5814682245254517,
"learning_rate": 3.978107771888733e-05,
"loss": 3.3003,
"step": 173
},
{
"epoch": 0.23758320532514082,
"grad_norm": 1.630428671836853,
"learning_rate": 3.9776936218550825e-05,
"loss": 3.0884,
"step": 174
},
{
"epoch": 0.23894862604540024,
"grad_norm": 1.5162523984909058,
"learning_rate": 3.977275615485401e-05,
"loss": 3.0227,
"step": 175
},
{
"epoch": 0.24031404676565968,
"grad_norm": 1.487644076347351,
"learning_rate": 3.9768537536864646e-05,
"loss": 3.0575,
"step": 176
},
{
"epoch": 0.2416794674859191,
"grad_norm": 1.653088927268982,
"learning_rate": 3.9764280373734165e-05,
"loss": 3.0835,
"step": 177
},
{
"epoch": 0.24304488820617853,
"grad_norm": 1.4124138355255127,
"learning_rate": 3.9759984674697584e-05,
"loss": 3.0413,
"step": 178
},
{
"epoch": 0.24441030892643795,
"grad_norm": 1.5104941129684448,
"learning_rate": 3.975565044907354e-05,
"loss": 3.0228,
"step": 179
},
{
"epoch": 0.2457757296466974,
"grad_norm": 1.4053763151168823,
"learning_rate": 3.975127770626422e-05,
"loss": 3.0466,
"step": 180
},
{
"epoch": 0.24714115036695683,
"grad_norm": 1.4529576301574707,
"learning_rate": 3.9746866455755385e-05,
"loss": 3.1507,
"step": 181
},
{
"epoch": 0.24850657108721624,
"grad_norm": 1.4030247926712036,
"learning_rate": 3.974241670711632e-05,
"loss": 3.0599,
"step": 182
},
{
"epoch": 0.24987199180747569,
"grad_norm": 1.391717791557312,
"learning_rate": 3.973792846999984e-05,
"loss": 3.0897,
"step": 183
},
{
"epoch": 0.2512374125277351,
"grad_norm": 1.4277074337005615,
"learning_rate": 3.973340175414222e-05,
"loss": 3.018,
"step": 184
},
{
"epoch": 0.25260283324799454,
"grad_norm": 1.401803970336914,
"learning_rate": 3.972883656936325e-05,
"loss": 3.0227,
"step": 185
},
{
"epoch": 0.25396825396825395,
"grad_norm": 1.446486234664917,
"learning_rate": 3.972423292556613e-05,
"loss": 3.0875,
"step": 186
},
{
"epoch": 0.2553336746885134,
"grad_norm": 1.5253474712371826,
"learning_rate": 3.9719590832737516e-05,
"loss": 3.1615,
"step": 187
},
{
"epoch": 0.25669909540877284,
"grad_norm": 1.5694295167922974,
"learning_rate": 3.9714910300947454e-05,
"loss": 3.1936,
"step": 188
},
{
"epoch": 0.25806451612903225,
"grad_norm": 1.515258550643921,
"learning_rate": 3.9710191340349385e-05,
"loss": 3.103,
"step": 189
},
{
"epoch": 0.25942993684929166,
"grad_norm": 1.4086757898330688,
"learning_rate": 3.9705433961180115e-05,
"loss": 2.9179,
"step": 190
},
{
"epoch": 0.26079535756955113,
"grad_norm": 1.4511244297027588,
"learning_rate": 3.970063817375979e-05,
"loss": 3.0582,
"step": 191
},
{
"epoch": 0.26216077828981055,
"grad_norm": 1.3760454654693604,
"learning_rate": 3.969580398849186e-05,
"loss": 3.0178,
"step": 192
},
{
"epoch": 0.26352619901006996,
"grad_norm": 1.4180570840835571,
"learning_rate": 3.96909314158631e-05,
"loss": 3.0052,
"step": 193
},
{
"epoch": 0.26489161973032943,
"grad_norm": 1.3801015615463257,
"learning_rate": 3.9686020466443514e-05,
"loss": 3.0294,
"step": 194
},
{
"epoch": 0.26625704045058884,
"grad_norm": 1.451645016670227,
"learning_rate": 3.968107115088641e-05,
"loss": 3.1715,
"step": 195
},
{
"epoch": 0.26762246117084826,
"grad_norm": 1.3865185976028442,
"learning_rate": 3.96760834799283e-05,
"loss": 3.0625,
"step": 196
},
{
"epoch": 0.26898788189110767,
"grad_norm": 1.3763779401779175,
"learning_rate": 3.967105746438887e-05,
"loss": 3.0502,
"step": 197
},
{
"epoch": 0.27035330261136714,
"grad_norm": 1.373224139213562,
"learning_rate": 3.966599311517103e-05,
"loss": 2.9775,
"step": 198
},
{
"epoch": 0.27171872333162655,
"grad_norm": 1.4603142738342285,
"learning_rate": 3.9660890443260836e-05,
"loss": 3.1125,
"step": 199
},
{
"epoch": 0.27308414405188597,
"grad_norm": 1.438384771347046,
"learning_rate": 3.9655749459727465e-05,
"loss": 3.0832,
"step": 200
},
{
"epoch": 0.27444956477214544,
"grad_norm": 1.356074333190918,
"learning_rate": 3.965057017572321e-05,
"loss": 2.8803,
"step": 201
},
{
"epoch": 0.27581498549240485,
"grad_norm": 1.4421138763427734,
"learning_rate": 3.9645352602483456e-05,
"loss": 3.0424,
"step": 202
},
{
"epoch": 0.27718040621266427,
"grad_norm": 1.5339027643203735,
"learning_rate": 3.964009675132662e-05,
"loss": 3.2676,
"step": 203
},
{
"epoch": 0.27854582693292373,
"grad_norm": 1.3659694194793701,
"learning_rate": 3.963480263365421e-05,
"loss": 3.0624,
"step": 204
},
{
"epoch": 0.27991124765318315,
"grad_norm": 1.4103004932403564,
"learning_rate": 3.962947026095069e-05,
"loss": 3.1935,
"step": 205
},
{
"epoch": 0.28127666837344256,
"grad_norm": 1.409977674484253,
"learning_rate": 3.9624099644783526e-05,
"loss": 3.1349,
"step": 206
},
{
"epoch": 0.282642089093702,
"grad_norm": 1.3748815059661865,
"learning_rate": 3.9618690796803155e-05,
"loss": 2.8683,
"step": 207
},
{
"epoch": 0.28400750981396145,
"grad_norm": 1.3962953090667725,
"learning_rate": 3.961324372874296e-05,
"loss": 2.9815,
"step": 208
},
{
"epoch": 0.28537293053422086,
"grad_norm": 1.3806780576705933,
"learning_rate": 3.960775845241921e-05,
"loss": 3.0506,
"step": 209
},
{
"epoch": 0.2867383512544803,
"grad_norm": 1.4572687149047852,
"learning_rate": 3.960223497973108e-05,
"loss": 2.9926,
"step": 210
},
{
"epoch": 0.28810377197473974,
"grad_norm": 1.3185594081878662,
"learning_rate": 3.959667332266057e-05,
"loss": 3.0307,
"step": 211
},
{
"epoch": 0.28946919269499916,
"grad_norm": 1.434801459312439,
"learning_rate": 3.959107349327256e-05,
"loss": 3.0045,
"step": 212
},
{
"epoch": 0.29083461341525857,
"grad_norm": 1.4034205675125122,
"learning_rate": 3.95854355037147e-05,
"loss": 3.2118,
"step": 213
},
{
"epoch": 0.292200034135518,
"grad_norm": 1.4422470331192017,
"learning_rate": 3.9579759366217444e-05,
"loss": 3.1018,
"step": 214
},
{
"epoch": 0.29356545485577745,
"grad_norm": 1.3593902587890625,
"learning_rate": 3.9574045093093983e-05,
"loss": 3.0055,
"step": 215
},
{
"epoch": 0.29493087557603687,
"grad_norm": 1.3980135917663574,
"learning_rate": 3.9568292696740236e-05,
"loss": 3.0399,
"step": 216
},
{
"epoch": 0.2962962962962963,
"grad_norm": 1.4268746376037598,
"learning_rate": 3.9562502189634846e-05,
"loss": 3.0863,
"step": 217
},
{
"epoch": 0.29766171701655575,
"grad_norm": 1.3492878675460815,
"learning_rate": 3.95566735843391e-05,
"loss": 2.9681,
"step": 218
},
{
"epoch": 0.29902713773681516,
"grad_norm": 1.4067364931106567,
"learning_rate": 3.9550806893496944e-05,
"loss": 2.987,
"step": 219
},
{
"epoch": 0.3003925584570746,
"grad_norm": 1.433929204940796,
"learning_rate": 3.9544902129834945e-05,
"loss": 3.1468,
"step": 220
},
{
"epoch": 0.301757979177334,
"grad_norm": 1.3934516906738281,
"learning_rate": 3.953895930616226e-05,
"loss": 3.0444,
"step": 221
},
{
"epoch": 0.30312339989759346,
"grad_norm": 1.468695878982544,
"learning_rate": 3.9532978435370594e-05,
"loss": 3.1784,
"step": 222
},
{
"epoch": 0.3044888206178529,
"grad_norm": 1.452207326889038,
"learning_rate": 3.952695953043422e-05,
"loss": 3.0834,
"step": 223
},
{
"epoch": 0.3058542413381123,
"grad_norm": 1.4073649644851685,
"learning_rate": 3.952090260440989e-05,
"loss": 2.9741,
"step": 224
},
{
"epoch": 0.30721966205837176,
"grad_norm": 1.4219416379928589,
"learning_rate": 3.951480767043683e-05,
"loss": 2.9256,
"step": 225
},
{
"epoch": 0.30858508277863117,
"grad_norm": 1.3768128156661987,
"learning_rate": 3.950867474173675e-05,
"loss": 2.9853,
"step": 226
},
{
"epoch": 0.3099505034988906,
"grad_norm": 1.4296373128890991,
"learning_rate": 3.950250383161375e-05,
"loss": 2.9574,
"step": 227
},
{
"epoch": 0.31131592421915005,
"grad_norm": 1.3591349124908447,
"learning_rate": 3.9496294953454336e-05,
"loss": 2.9602,
"step": 228
},
{
"epoch": 0.31268134493940947,
"grad_norm": 1.4345592260360718,
"learning_rate": 3.949004812072738e-05,
"loss": 3.1352,
"step": 229
},
{
"epoch": 0.3140467656596689,
"grad_norm": 1.4081335067749023,
"learning_rate": 3.948376334698409e-05,
"loss": 2.9472,
"step": 230
},
{
"epoch": 0.3154121863799283,
"grad_norm": 1.4144725799560547,
"learning_rate": 3.947744064585795e-05,
"loss": 2.9073,
"step": 231
},
{
"epoch": 0.31677760710018776,
"grad_norm": 1.3229619264602661,
"learning_rate": 3.9471080031064765e-05,
"loss": 2.9551,
"step": 232
},
{
"epoch": 0.3181430278204472,
"grad_norm": 1.3419454097747803,
"learning_rate": 3.946468151640255e-05,
"loss": 2.917,
"step": 233
},
{
"epoch": 0.3195084485407066,
"grad_norm": 1.3524123430252075,
"learning_rate": 3.945824511575157e-05,
"loss": 2.9219,
"step": 234
},
{
"epoch": 0.32087386926096606,
"grad_norm": 1.387582540512085,
"learning_rate": 3.945177084307423e-05,
"loss": 3.0767,
"step": 235
},
{
"epoch": 0.3222392899812255,
"grad_norm": 1.3753900527954102,
"learning_rate": 3.944525871241514e-05,
"loss": 3.0015,
"step": 236
},
{
"epoch": 0.3236047107014849,
"grad_norm": 1.398117184638977,
"learning_rate": 3.9438708737900986e-05,
"loss": 3.0974,
"step": 237
},
{
"epoch": 0.3249701314217443,
"grad_norm": 1.562734842300415,
"learning_rate": 3.94321209337406e-05,
"loss": 2.9214,
"step": 238
},
{
"epoch": 0.32633555214200377,
"grad_norm": 1.4604074954986572,
"learning_rate": 3.942549531422483e-05,
"loss": 3.0322,
"step": 239
},
{
"epoch": 0.3277009728622632,
"grad_norm": 1.4694637060165405,
"learning_rate": 3.941883189372659e-05,
"loss": 3.0074,
"step": 240
},
{
"epoch": 0.3290663935825226,
"grad_norm": 1.3830381631851196,
"learning_rate": 3.941213068670078e-05,
"loss": 2.9779,
"step": 241
},
{
"epoch": 0.33043181430278207,
"grad_norm": 1.4520971775054932,
"learning_rate": 3.940539170768426e-05,
"loss": 3.1271,
"step": 242
},
{
"epoch": 0.3317972350230415,
"grad_norm": 1.5289911031723022,
"learning_rate": 3.939861497129586e-05,
"loss": 3.3404,
"step": 243
},
{
"epoch": 0.3331626557433009,
"grad_norm": 1.3928587436676025,
"learning_rate": 3.939180049223629e-05,
"loss": 3.0863,
"step": 244
},
{
"epoch": 0.3345280764635603,
"grad_norm": 1.3869800567626953,
"learning_rate": 3.9384948285288145e-05,
"loss": 2.9346,
"step": 245
},
{
"epoch": 0.3358934971838198,
"grad_norm": 1.3455641269683838,
"learning_rate": 3.937805836531585e-05,
"loss": 3.0091,
"step": 246
},
{
"epoch": 0.3372589179040792,
"grad_norm": 1.469413161277771,
"learning_rate": 3.937113074726567e-05,
"loss": 3.1037,
"step": 247
},
{
"epoch": 0.3386243386243386,
"grad_norm": 1.3640011548995972,
"learning_rate": 3.9364165446165615e-05,
"loss": 3.0471,
"step": 248
},
{
"epoch": 0.3399897593445981,
"grad_norm": 1.427222490310669,
"learning_rate": 3.935716247712546e-05,
"loss": 3.0044,
"step": 249
},
{
"epoch": 0.3413551800648575,
"grad_norm": 1.3031620979309082,
"learning_rate": 3.9350121855336694e-05,
"loss": 2.9551,
"step": 250
},
{
"epoch": 0.3427206007851169,
"grad_norm": 1.3024346828460693,
"learning_rate": 3.9343043596072464e-05,
"loss": 2.904,
"step": 251
},
{
"epoch": 0.34408602150537637,
"grad_norm": 1.4278597831726074,
"learning_rate": 3.9335927714687584e-05,
"loss": 2.9646,
"step": 252
},
{
"epoch": 0.3454514422256358,
"grad_norm": 1.3256884813308716,
"learning_rate": 3.9328774226618485e-05,
"loss": 3.0222,
"step": 253
},
{
"epoch": 0.3468168629458952,
"grad_norm": 1.4290896654129028,
"learning_rate": 3.932158314738317e-05,
"loss": 2.9425,
"step": 254
},
{
"epoch": 0.3481822836661546,
"grad_norm": 1.4056618213653564,
"learning_rate": 3.931435449258118e-05,
"loss": 2.9839,
"step": 255
},
{
"epoch": 0.3495477043864141,
"grad_norm": 1.3533239364624023,
"learning_rate": 3.930708827789357e-05,
"loss": 2.8758,
"step": 256
},
{
"epoch": 0.3509131251066735,
"grad_norm": 1.4235435724258423,
"learning_rate": 3.9299784519082894e-05,
"loss": 2.8751,
"step": 257
},
{
"epoch": 0.3522785458269329,
"grad_norm": 1.4764057397842407,
"learning_rate": 3.929244323199313e-05,
"loss": 2.9559,
"step": 258
},
{
"epoch": 0.3536439665471924,
"grad_norm": 1.3680962324142456,
"learning_rate": 3.928506443254967e-05,
"loss": 2.9041,
"step": 259
},
{
"epoch": 0.3550093872674518,
"grad_norm": 1.3964440822601318,
"learning_rate": 3.927764813675928e-05,
"loss": 2.7772,
"step": 260
},
{
"epoch": 0.3563748079877112,
"grad_norm": 1.3815569877624512,
"learning_rate": 3.9270194360710064e-05,
"loss": 2.941,
"step": 261
},
{
"epoch": 0.3577402287079706,
"grad_norm": 1.5464051961898804,
"learning_rate": 3.926270312057145e-05,
"loss": 3.196,
"step": 262
},
{
"epoch": 0.3591056494282301,
"grad_norm": 1.4574122428894043,
"learning_rate": 3.925517443259412e-05,
"loss": 2.9771,
"step": 263
},
{
"epoch": 0.3604710701484895,
"grad_norm": 1.3610405921936035,
"learning_rate": 3.924760831310999e-05,
"loss": 3.0683,
"step": 264
},
{
"epoch": 0.3618364908687489,
"grad_norm": 1.4155353307724,
"learning_rate": 3.924000477853218e-05,
"loss": 2.852,
"step": 265
},
{
"epoch": 0.3632019115890084,
"grad_norm": 1.4231737852096558,
"learning_rate": 3.9232363845354974e-05,
"loss": 2.8902,
"step": 266
},
{
"epoch": 0.3645673323092678,
"grad_norm": 1.3444836139678955,
"learning_rate": 3.922468553015379e-05,
"loss": 2.9308,
"step": 267
},
{
"epoch": 0.3659327530295272,
"grad_norm": 1.4084193706512451,
"learning_rate": 3.9216969849585135e-05,
"loss": 2.994,
"step": 268
},
{
"epoch": 0.3672981737497866,
"grad_norm": 1.3839198350906372,
"learning_rate": 3.920921682038656e-05,
"loss": 2.9382,
"step": 269
},
{
"epoch": 0.3686635944700461,
"grad_norm": 1.4327356815338135,
"learning_rate": 3.9201426459376655e-05,
"loss": 3.0459,
"step": 270
},
{
"epoch": 0.3700290151903055,
"grad_norm": 1.3953254222869873,
"learning_rate": 3.9193598783454984e-05,
"loss": 3.047,
"step": 271
},
{
"epoch": 0.3713944359105649,
"grad_norm": 1.4276584386825562,
"learning_rate": 3.918573380960206e-05,
"loss": 2.9688,
"step": 272
},
{
"epoch": 0.3727598566308244,
"grad_norm": 1.3536823987960815,
"learning_rate": 3.91778315548793e-05,
"loss": 3.0259,
"step": 273
},
{
"epoch": 0.3741252773510838,
"grad_norm": 1.3419265747070312,
"learning_rate": 3.9169892036429014e-05,
"loss": 2.9833,
"step": 274
},
{
"epoch": 0.3754906980713432,
"grad_norm": 1.3257263898849487,
"learning_rate": 3.916191527147431e-05,
"loss": 2.8981,
"step": 275
},
{
"epoch": 0.3768561187916027,
"grad_norm": 1.3887954950332642,
"learning_rate": 3.915390127731914e-05,
"loss": 3.0064,
"step": 276
},
{
"epoch": 0.3782215395118621,
"grad_norm": 1.3079547882080078,
"learning_rate": 3.91458500713482e-05,
"loss": 2.8363,
"step": 277
},
{
"epoch": 0.3795869602321215,
"grad_norm": 1.3726911544799805,
"learning_rate": 3.913776167102688e-05,
"loss": 2.9567,
"step": 278
},
{
"epoch": 0.38095238095238093,
"grad_norm": 1.3575981855392456,
"learning_rate": 3.9129636093901284e-05,
"loss": 2.9757,
"step": 279
},
{
"epoch": 0.3823178016726404,
"grad_norm": 1.4176872968673706,
"learning_rate": 3.9121473357598175e-05,
"loss": 3.1465,
"step": 280
},
{
"epoch": 0.3836832223928998,
"grad_norm": 1.38345468044281,
"learning_rate": 3.91132734798249e-05,
"loss": 2.8905,
"step": 281
},
{
"epoch": 0.38504864311315923,
"grad_norm": 1.4212768077850342,
"learning_rate": 3.910503647836939e-05,
"loss": 3.0615,
"step": 282
},
{
"epoch": 0.3864140638334187,
"grad_norm": 1.3602497577667236,
"learning_rate": 3.9096762371100094e-05,
"loss": 3.0335,
"step": 283
},
{
"epoch": 0.3877794845536781,
"grad_norm": 1.3397915363311768,
"learning_rate": 3.908845117596597e-05,
"loss": 3.0985,
"step": 284
},
{
"epoch": 0.3891449052739375,
"grad_norm": 1.2894964218139648,
"learning_rate": 3.908010291099643e-05,
"loss": 2.8898,
"step": 285
},
{
"epoch": 0.39051032599419694,
"grad_norm": 1.3393354415893555,
"learning_rate": 3.907171759430128e-05,
"loss": 3.0582,
"step": 286
},
{
"epoch": 0.3918757467144564,
"grad_norm": 1.2876499891281128,
"learning_rate": 3.9063295244070743e-05,
"loss": 2.9869,
"step": 287
},
{
"epoch": 0.3932411674347158,
"grad_norm": 1.265411615371704,
"learning_rate": 3.9054835878575336e-05,
"loss": 2.8476,
"step": 288
},
{
"epoch": 0.39460658815497524,
"grad_norm": 1.3758430480957031,
"learning_rate": 3.904633951616589e-05,
"loss": 2.9314,
"step": 289
},
{
"epoch": 0.3959720088752347,
"grad_norm": 1.3400568962097168,
"learning_rate": 3.90378061752735e-05,
"loss": 3.2395,
"step": 290
},
{
"epoch": 0.3973374295954941,
"grad_norm": 1.3283575773239136,
"learning_rate": 3.902923587440947e-05,
"loss": 3.11,
"step": 291
},
{
"epoch": 0.39870285031575353,
"grad_norm": 1.3666977882385254,
"learning_rate": 3.902062863216528e-05,
"loss": 2.9748,
"step": 292
},
{
"epoch": 0.40006827103601295,
"grad_norm": 1.342584252357483,
"learning_rate": 3.901198446721256e-05,
"loss": 2.9218,
"step": 293
},
{
"epoch": 0.4014336917562724,
"grad_norm": 1.3122738599777222,
"learning_rate": 3.900330339830301e-05,
"loss": 2.8833,
"step": 294
},
{
"epoch": 0.40279911247653183,
"grad_norm": 1.3478033542633057,
"learning_rate": 3.899458544426841e-05,
"loss": 3.0038,
"step": 295
},
{
"epoch": 0.40416453319679124,
"grad_norm": 1.312415361404419,
"learning_rate": 3.8985830624020533e-05,
"loss": 2.8595,
"step": 296
},
{
"epoch": 0.4055299539170507,
"grad_norm": 1.4030121564865112,
"learning_rate": 3.897703895655115e-05,
"loss": 3.0907,
"step": 297
},
{
"epoch": 0.4068953746373101,
"grad_norm": 1.2528835535049438,
"learning_rate": 3.8968210460931957e-05,
"loss": 2.7231,
"step": 298
},
{
"epoch": 0.40826079535756954,
"grad_norm": 1.2837860584259033,
"learning_rate": 3.8959345156314524e-05,
"loss": 2.8879,
"step": 299
},
{
"epoch": 0.409626216077829,
"grad_norm": 1.2902891635894775,
"learning_rate": 3.8950443061930295e-05,
"loss": 2.9437,
"step": 300
},
{
"epoch": 0.4109916367980884,
"grad_norm": 1.3145170211791992,
"learning_rate": 3.89415041970905e-05,
"loss": 2.8968,
"step": 301
},
{
"epoch": 0.41235705751834784,
"grad_norm": 1.375896692276001,
"learning_rate": 3.8932528581186145e-05,
"loss": 3.1013,
"step": 302
},
{
"epoch": 0.41372247823860725,
"grad_norm": 1.279364824295044,
"learning_rate": 3.892351623368796e-05,
"loss": 2.9239,
"step": 303
},
{
"epoch": 0.4150878989588667,
"grad_norm": 1.3359878063201904,
"learning_rate": 3.891446717414635e-05,
"loss": 3.0631,
"step": 304
},
{
"epoch": 0.41645331967912613,
"grad_norm": 1.3311201333999634,
"learning_rate": 3.8905381422191385e-05,
"loss": 3.0484,
"step": 305
},
{
"epoch": 0.41781874039938555,
"grad_norm": 1.329620599746704,
"learning_rate": 3.8896258997532685e-05,
"loss": 2.9581,
"step": 306
},
{
"epoch": 0.419184161119645,
"grad_norm": 1.3566920757293701,
"learning_rate": 3.8887099919959475e-05,
"loss": 2.9822,
"step": 307
},
{
"epoch": 0.42054958183990443,
"grad_norm": 1.42951500415802,
"learning_rate": 3.887790420934045e-05,
"loss": 2.9736,
"step": 308
},
{
"epoch": 0.42191500256016384,
"grad_norm": 1.3591212034225464,
"learning_rate": 3.886867188562379e-05,
"loss": 3.0734,
"step": 309
},
{
"epoch": 0.42328042328042326,
"grad_norm": 1.3106906414031982,
"learning_rate": 3.8859402968837106e-05,
"loss": 3.0949,
"step": 310
},
{
"epoch": 0.4246458440006827,
"grad_norm": 1.3801910877227783,
"learning_rate": 3.885009747908737e-05,
"loss": 3.0056,
"step": 311
},
{
"epoch": 0.42601126472094214,
"grad_norm": 1.3937325477600098,
"learning_rate": 3.884075543656092e-05,
"loss": 3.0166,
"step": 312
},
{
"epoch": 0.42737668544120155,
"grad_norm": 1.3688315153121948,
"learning_rate": 3.883137686152336e-05,
"loss": 2.9224,
"step": 313
},
{
"epoch": 0.428742106161461,
"grad_norm": 1.3972077369689941,
"learning_rate": 3.8821961774319555e-05,
"loss": 2.9957,
"step": 314
},
{
"epoch": 0.43010752688172044,
"grad_norm": 1.385298728942871,
"learning_rate": 3.881251019537358e-05,
"loss": 3.0206,
"step": 315
},
{
"epoch": 0.43147294760197985,
"grad_norm": 1.4394325017929077,
"learning_rate": 3.880302214518866e-05,
"loss": 2.9026,
"step": 316
},
{
"epoch": 0.43283836832223926,
"grad_norm": 1.3800427913665771,
"learning_rate": 3.879349764434716e-05,
"loss": 3.1173,
"step": 317
},
{
"epoch": 0.43420378904249873,
"grad_norm": 1.4867631196975708,
"learning_rate": 3.878393671351048e-05,
"loss": 3.0277,
"step": 318
},
{
"epoch": 0.43556920976275815,
"grad_norm": 1.2858531475067139,
"learning_rate": 3.877433937341908e-05,
"loss": 2.8638,
"step": 319
},
{
"epoch": 0.43693463048301756,
"grad_norm": 1.4738233089447021,
"learning_rate": 3.876470564489239e-05,
"loss": 2.7714,
"step": 320
},
{
"epoch": 0.43830005120327703,
"grad_norm": 1.5328595638275146,
"learning_rate": 3.875503554882878e-05,
"loss": 3.1111,
"step": 321
},
{
"epoch": 0.43966547192353644,
"grad_norm": 1.4056901931762695,
"learning_rate": 3.8745329106205486e-05,
"loss": 3.0781,
"step": 322
},
{
"epoch": 0.44103089264379586,
"grad_norm": 1.4081515073776245,
"learning_rate": 3.873558633807865e-05,
"loss": 2.8776,
"step": 323
},
{
"epoch": 0.4423963133640553,
"grad_norm": 1.3121602535247803,
"learning_rate": 3.872580726558315e-05,
"loss": 2.8528,
"step": 324
},
{
"epoch": 0.44376173408431474,
"grad_norm": 1.3597134351730347,
"learning_rate": 3.871599190993266e-05,
"loss": 3.0126,
"step": 325
},
{
"epoch": 0.44512715480457415,
"grad_norm": 1.4099037647247314,
"learning_rate": 3.8706140292419546e-05,
"loss": 2.8402,
"step": 326
},
{
"epoch": 0.44649257552483357,
"grad_norm": 1.4041759967803955,
"learning_rate": 3.869625243441484e-05,
"loss": 3.1083,
"step": 327
},
{
"epoch": 0.44785799624509304,
"grad_norm": 1.355782151222229,
"learning_rate": 3.868632835736819e-05,
"loss": 3.0502,
"step": 328
},
{
"epoch": 0.44922341696535245,
"grad_norm": 1.3535327911376953,
"learning_rate": 3.867636808280781e-05,
"loss": 2.9945,
"step": 329
},
{
"epoch": 0.45058883768561186,
"grad_norm": 1.3494234085083008,
"learning_rate": 3.866637163234045e-05,
"loss": 2.8163,
"step": 330
},
{
"epoch": 0.45195425840587133,
"grad_norm": 1.4025193452835083,
"learning_rate": 3.865633902765133e-05,
"loss": 3.1347,
"step": 331
},
{
"epoch": 0.45331967912613075,
"grad_norm": 1.3067885637283325,
"learning_rate": 3.8646270290504085e-05,
"loss": 2.924,
"step": 332
},
{
"epoch": 0.45468509984639016,
"grad_norm": 1.3395352363586426,
"learning_rate": 3.8636165442740746e-05,
"loss": 2.9047,
"step": 333
},
{
"epoch": 0.4560505205666496,
"grad_norm": 1.3719593286514282,
"learning_rate": 3.862602450628167e-05,
"loss": 3.0079,
"step": 334
},
{
"epoch": 0.45741594128690904,
"grad_norm": 1.4157326221466064,
"learning_rate": 3.861584750312552e-05,
"loss": 3.0759,
"step": 335
},
{
"epoch": 0.45878136200716846,
"grad_norm": 1.3816114664077759,
"learning_rate": 3.860563445534918e-05,
"loss": 2.9723,
"step": 336
},
{
"epoch": 0.46014678272742787,
"grad_norm": 1.3497341871261597,
"learning_rate": 3.859538538510772e-05,
"loss": 3.0346,
"step": 337
},
{
"epoch": 0.46151220344768734,
"grad_norm": 1.350602626800537,
"learning_rate": 3.858510031463438e-05,
"loss": 2.8743,
"step": 338
},
{
"epoch": 0.46287762416794676,
"grad_norm": 1.387095332145691,
"learning_rate": 3.857477926624046e-05,
"loss": 3.0133,
"step": 339
},
{
"epoch": 0.46424304488820617,
"grad_norm": 1.4189172983169556,
"learning_rate": 3.8564422262315336e-05,
"loss": 3.2004,
"step": 340
},
{
"epoch": 0.4656084656084656,
"grad_norm": 1.3419146537780762,
"learning_rate": 3.855402932532637e-05,
"loss": 2.9068,
"step": 341
},
{
"epoch": 0.46697388632872505,
"grad_norm": 1.3219139575958252,
"learning_rate": 3.854360047781886e-05,
"loss": 2.8327,
"step": 342
},
{
"epoch": 0.46833930704898447,
"grad_norm": 1.317773699760437,
"learning_rate": 3.8533135742416045e-05,
"loss": 3.0316,
"step": 343
},
{
"epoch": 0.4697047277692439,
"grad_norm": 1.4251924753189087,
"learning_rate": 3.852263514181897e-05,
"loss": 3.0074,
"step": 344
},
{
"epoch": 0.47107014848950335,
"grad_norm": 1.4333152770996094,
"learning_rate": 3.8512098698806506e-05,
"loss": 3.1109,
"step": 345
},
{
"epoch": 0.47243556920976276,
"grad_norm": 1.3378651142120361,
"learning_rate": 3.850152643623527e-05,
"loss": 2.9431,
"step": 346
},
{
"epoch": 0.4738009899300222,
"grad_norm": 1.4802751541137695,
"learning_rate": 3.8490918377039576e-05,
"loss": 3.091,
"step": 347
},
{
"epoch": 0.47516641065028165,
"grad_norm": 1.3342303037643433,
"learning_rate": 3.8480274544231415e-05,
"loss": 3.0976,
"step": 348
},
{
"epoch": 0.47653183137054106,
"grad_norm": 1.3089995384216309,
"learning_rate": 3.846959496090036e-05,
"loss": 2.9495,
"step": 349
},
{
"epoch": 0.4778972520908005,
"grad_norm": 1.3713091611862183,
"learning_rate": 3.8458879650213534e-05,
"loss": 2.9096,
"step": 350
},
{
"epoch": 0.4792626728110599,
"grad_norm": 1.338836431503296,
"learning_rate": 3.844812863541557e-05,
"loss": 2.9119,
"step": 351
},
{
"epoch": 0.48062809353131936,
"grad_norm": 1.444617509841919,
"learning_rate": 3.843734193982857e-05,
"loss": 2.9674,
"step": 352
},
{
"epoch": 0.48199351425157877,
"grad_norm": 1.434388518333435,
"learning_rate": 3.842651958685202e-05,
"loss": 3.1323,
"step": 353
},
{
"epoch": 0.4833589349718382,
"grad_norm": 1.4443765878677368,
"learning_rate": 3.8415661599962746e-05,
"loss": 3.0445,
"step": 354
},
{
"epoch": 0.48472435569209765,
"grad_norm": 1.519293189048767,
"learning_rate": 3.8404768002714904e-05,
"loss": 2.9302,
"step": 355
},
{
"epoch": 0.48608977641235707,
"grad_norm": 1.3553463220596313,
"learning_rate": 3.8393838818739874e-05,
"loss": 2.8599,
"step": 356
},
{
"epoch": 0.4874551971326165,
"grad_norm": 1.339248538017273,
"learning_rate": 3.838287407174626e-05,
"loss": 2.9773,
"step": 357
},
{
"epoch": 0.4888206178528759,
"grad_norm": 1.3699098825454712,
"learning_rate": 3.8371873785519785e-05,
"loss": 2.9303,
"step": 358
},
{
"epoch": 0.49018603857313536,
"grad_norm": 1.3392761945724487,
"learning_rate": 3.836083798392327e-05,
"loss": 2.9212,
"step": 359
},
{
"epoch": 0.4915514592933948,
"grad_norm": 1.4360921382904053,
"learning_rate": 3.83497666908966e-05,
"loss": 3.0258,
"step": 360
},
{
"epoch": 0.4929168800136542,
"grad_norm": 1.3094574213027954,
"learning_rate": 3.833865993045665e-05,
"loss": 2.9445,
"step": 361
},
{
"epoch": 0.49428230073391366,
"grad_norm": 1.3445249795913696,
"learning_rate": 3.8327517726697194e-05,
"loss": 2.9284,
"step": 362
},
{
"epoch": 0.4956477214541731,
"grad_norm": 1.3780626058578491,
"learning_rate": 3.8316340103788947e-05,
"loss": 3.0074,
"step": 363
},
{
"epoch": 0.4970131421744325,
"grad_norm": 1.3903131484985352,
"learning_rate": 3.830512708597942e-05,
"loss": 2.893,
"step": 364
},
{
"epoch": 0.4983785628946919,
"grad_norm": 1.338554859161377,
"learning_rate": 3.829387869759293e-05,
"loss": 2.9588,
"step": 365
},
{
"epoch": 0.49974398361495137,
"grad_norm": 1.3329784870147705,
"learning_rate": 3.82825949630305e-05,
"loss": 2.8984,
"step": 366
},
{
"epoch": 0.5011094043352108,
"grad_norm": 1.3981587886810303,
"learning_rate": 3.827127590676986e-05,
"loss": 2.8382,
"step": 367
},
{
"epoch": 0.5024748250554703,
"grad_norm": 1.3244234323501587,
"learning_rate": 3.8259921553365324e-05,
"loss": 2.9084,
"step": 368
},
{
"epoch": 0.5038402457757296,
"grad_norm": 1.4119958877563477,
"learning_rate": 3.8248531927447805e-05,
"loss": 3.0194,
"step": 369
},
{
"epoch": 0.5052056664959891,
"grad_norm": 1.46544349193573,
"learning_rate": 3.823710705372473e-05,
"loss": 2.8543,
"step": 370
},
{
"epoch": 0.5065710872162486,
"grad_norm": 1.4029783010482788,
"learning_rate": 3.822564695697998e-05,
"loss": 2.833,
"step": 371
},
{
"epoch": 0.5079365079365079,
"grad_norm": 1.4057574272155762,
"learning_rate": 3.8214151662073854e-05,
"loss": 2.9731,
"step": 372
},
{
"epoch": 0.5093019286567674,
"grad_norm": 1.5014464855194092,
"learning_rate": 3.8202621193942976e-05,
"loss": 3.1724,
"step": 373
},
{
"epoch": 0.5106673493770268,
"grad_norm": 1.3028396368026733,
"learning_rate": 3.819105557760032e-05,
"loss": 2.8206,
"step": 374
},
{
"epoch": 0.5120327700972862,
"grad_norm": 1.4192930459976196,
"learning_rate": 3.8179454838135066e-05,
"loss": 2.9899,
"step": 375
},
{
"epoch": 0.5133981908175457,
"grad_norm": 1.331186056137085,
"learning_rate": 3.8167819000712615e-05,
"loss": 2.9847,
"step": 376
},
{
"epoch": 0.514763611537805,
"grad_norm": 1.3640269041061401,
"learning_rate": 3.815614809057448e-05,
"loss": 2.8577,
"step": 377
},
{
"epoch": 0.5161290322580645,
"grad_norm": 1.3373394012451172,
"learning_rate": 3.814444213303828e-05,
"loss": 2.9442,
"step": 378
},
{
"epoch": 0.517494452978324,
"grad_norm": 1.2525429725646973,
"learning_rate": 3.8132701153497655e-05,
"loss": 2.755,
"step": 379
},
{
"epoch": 0.5188598736985833,
"grad_norm": 1.392865538597107,
"learning_rate": 3.8120925177422196e-05,
"loss": 2.9872,
"step": 380
},
{
"epoch": 0.5202252944188428,
"grad_norm": 1.423351526260376,
"learning_rate": 3.810911423035745e-05,
"loss": 3.0558,
"step": 381
},
{
"epoch": 0.5215907151391023,
"grad_norm": 1.2661659717559814,
"learning_rate": 3.809726833792479e-05,
"loss": 2.8862,
"step": 382
},
{
"epoch": 0.5229561358593616,
"grad_norm": 1.4103831052780151,
"learning_rate": 3.808538752582142e-05,
"loss": 2.9663,
"step": 383
},
{
"epoch": 0.5243215565796211,
"grad_norm": 1.3018399477005005,
"learning_rate": 3.8073471819820294e-05,
"loss": 2.8138,
"step": 384
},
{
"epoch": 0.5256869772998806,
"grad_norm": 1.3350375890731812,
"learning_rate": 3.8061521245770044e-05,
"loss": 2.9331,
"step": 385
},
{
"epoch": 0.5270523980201399,
"grad_norm": 1.2823617458343506,
"learning_rate": 3.8049535829594956e-05,
"loss": 2.8896,
"step": 386
},
{
"epoch": 0.5284178187403994,
"grad_norm": 1.3584403991699219,
"learning_rate": 3.803751559729489e-05,
"loss": 3.0395,
"step": 387
},
{
"epoch": 0.5297832394606589,
"grad_norm": 1.3150285482406616,
"learning_rate": 3.802546057494524e-05,
"loss": 3.0553,
"step": 388
},
{
"epoch": 0.5311486601809182,
"grad_norm": 1.3261231184005737,
"learning_rate": 3.801337078869686e-05,
"loss": 2.9124,
"step": 389
},
{
"epoch": 0.5325140809011777,
"grad_norm": 1.3283610343933105,
"learning_rate": 3.8001246264776025e-05,
"loss": 2.966,
"step": 390
},
{
"epoch": 0.5338795016214372,
"grad_norm": 1.300338625907898,
"learning_rate": 3.798908702948437e-05,
"loss": 2.9111,
"step": 391
},
{
"epoch": 0.5352449223416965,
"grad_norm": 1.3091778755187988,
"learning_rate": 3.797689310919883e-05,
"loss": 2.9385,
"step": 392
},
{
"epoch": 0.536610343061956,
"grad_norm": 1.3251817226409912,
"learning_rate": 3.7964664530371565e-05,
"loss": 2.8677,
"step": 393
},
{
"epoch": 0.5379757637822153,
"grad_norm": 1.3573335409164429,
"learning_rate": 3.795240131952992e-05,
"loss": 2.837,
"step": 394
},
{
"epoch": 0.5393411845024748,
"grad_norm": 1.3734980821609497,
"learning_rate": 3.7940103503276383e-05,
"loss": 3.0215,
"step": 395
},
{
"epoch": 0.5407066052227343,
"grad_norm": 1.4021447896957397,
"learning_rate": 3.792777110828852e-05,
"loss": 3.0128,
"step": 396
},
{
"epoch": 0.5420720259429936,
"grad_norm": 1.340755820274353,
"learning_rate": 3.791540416131887e-05,
"loss": 3.0792,
"step": 397
},
{
"epoch": 0.5434374466632531,
"grad_norm": 1.2638185024261475,
"learning_rate": 3.7903002689194976e-05,
"loss": 2.8853,
"step": 398
},
{
"epoch": 0.5448028673835126,
"grad_norm": 1.3345012664794922,
"learning_rate": 3.7890566718819216e-05,
"loss": 2.9527,
"step": 399
},
{
"epoch": 0.5461682881037719,
"grad_norm": 1.3717132806777954,
"learning_rate": 3.787809627716887e-05,
"loss": 2.8532,
"step": 400
},
{
"epoch": 0.5475337088240314,
"grad_norm": 1.3624272346496582,
"learning_rate": 3.786559139129594e-05,
"loss": 2.9944,
"step": 401
},
{
"epoch": 0.5488991295442909,
"grad_norm": 1.3174376487731934,
"learning_rate": 3.785305208832717e-05,
"loss": 2.9651,
"step": 402
},
{
"epoch": 0.5502645502645502,
"grad_norm": 1.368506669998169,
"learning_rate": 3.784047839546399e-05,
"loss": 2.9833,
"step": 403
},
{
"epoch": 0.5516299709848097,
"grad_norm": 1.386845588684082,
"learning_rate": 3.7827870339982385e-05,
"loss": 2.9086,
"step": 404
},
{
"epoch": 0.5529953917050692,
"grad_norm": 1.3105249404907227,
"learning_rate": 3.781522794923291e-05,
"loss": 2.9186,
"step": 405
},
{
"epoch": 0.5543608124253285,
"grad_norm": 1.2829691171646118,
"learning_rate": 3.780255125064061e-05,
"loss": 2.729,
"step": 406
},
{
"epoch": 0.555726233145588,
"grad_norm": 1.3970478773117065,
"learning_rate": 3.7789840271704925e-05,
"loss": 3.1215,
"step": 407
},
{
"epoch": 0.5570916538658475,
"grad_norm": 1.312456727027893,
"learning_rate": 3.777709503999968e-05,
"loss": 2.8841,
"step": 408
},
{
"epoch": 0.5584570745861068,
"grad_norm": 1.3093911409378052,
"learning_rate": 3.776431558317302e-05,
"loss": 3.0049,
"step": 409
},
{
"epoch": 0.5598224953063663,
"grad_norm": 1.3421498537063599,
"learning_rate": 3.775150192894729e-05,
"loss": 2.9349,
"step": 410
},
{
"epoch": 0.5611879160266257,
"grad_norm": 1.3915818929672241,
"learning_rate": 3.773865410511906e-05,
"loss": 2.9318,
"step": 411
},
{
"epoch": 0.5625533367468851,
"grad_norm": 1.2593127489089966,
"learning_rate": 3.7725772139559e-05,
"loss": 2.7427,
"step": 412
},
{
"epoch": 0.5639187574671446,
"grad_norm": 1.3098043203353882,
"learning_rate": 3.771285606021184e-05,
"loss": 2.902,
"step": 413
},
{
"epoch": 0.565284178187404,
"grad_norm": 1.3974379301071167,
"learning_rate": 3.769990589509635e-05,
"loss": 2.9425,
"step": 414
},
{
"epoch": 0.5666495989076634,
"grad_norm": 1.2861424684524536,
"learning_rate": 3.768692167230519e-05,
"loss": 2.8829,
"step": 415
},
{
"epoch": 0.5680150196279229,
"grad_norm": 1.362608551979065,
"learning_rate": 3.7673903420004946e-05,
"loss": 2.979,
"step": 416
},
{
"epoch": 0.5693804403481822,
"grad_norm": 1.4318872690200806,
"learning_rate": 3.7660851166436e-05,
"loss": 2.9063,
"step": 417
},
{
"epoch": 0.5707458610684417,
"grad_norm": 1.320885419845581,
"learning_rate": 3.764776493991249e-05,
"loss": 2.8517,
"step": 418
},
{
"epoch": 0.5721112817887012,
"grad_norm": 1.353554129600525,
"learning_rate": 3.763464476882226e-05,
"loss": 2.9256,
"step": 419
},
{
"epoch": 0.5734767025089605,
"grad_norm": 1.3187509775161743,
"learning_rate": 3.76214906816268e-05,
"loss": 2.8558,
"step": 420
},
{
"epoch": 0.57484212322922,
"grad_norm": 1.334849238395691,
"learning_rate": 3.7608302706861154e-05,
"loss": 2.9241,
"step": 421
},
{
"epoch": 0.5762075439494795,
"grad_norm": 1.2721538543701172,
"learning_rate": 3.759508087313388e-05,
"loss": 2.9531,
"step": 422
},
{
"epoch": 0.5775729646697388,
"grad_norm": 1.378733515739441,
"learning_rate": 3.7581825209127e-05,
"loss": 2.9066,
"step": 423
},
{
"epoch": 0.5789383853899983,
"grad_norm": 1.4031331539154053,
"learning_rate": 3.756853574359592e-05,
"loss": 3.0673,
"step": 424
},
{
"epoch": 0.5803038061102577,
"grad_norm": 1.2762374877929688,
"learning_rate": 3.7555212505369355e-05,
"loss": 2.8023,
"step": 425
},
{
"epoch": 0.5816692268305171,
"grad_norm": 1.3040027618408203,
"learning_rate": 3.7541855523349306e-05,
"loss": 2.7411,
"step": 426
},
{
"epoch": 0.5830346475507766,
"grad_norm": 1.372672438621521,
"learning_rate": 3.7528464826510965e-05,
"loss": 2.9527,
"step": 427
},
{
"epoch": 0.584400068271036,
"grad_norm": 1.3174269199371338,
"learning_rate": 3.751504044390264e-05,
"loss": 2.8347,
"step": 428
},
{
"epoch": 0.5857654889912954,
"grad_norm": 1.4506326913833618,
"learning_rate": 3.750158240464576e-05,
"loss": 3.0221,
"step": 429
},
{
"epoch": 0.5871309097115549,
"grad_norm": 1.2877005338668823,
"learning_rate": 3.7488090737934714e-05,
"loss": 2.788,
"step": 430
},
{
"epoch": 0.5884963304318143,
"grad_norm": 1.398964524269104,
"learning_rate": 3.7474565473036886e-05,
"loss": 3.111,
"step": 431
},
{
"epoch": 0.5898617511520737,
"grad_norm": 1.3040918111801147,
"learning_rate": 3.74610066392925e-05,
"loss": 2.7763,
"step": 432
},
{
"epoch": 0.5912271718723332,
"grad_norm": 1.3138196468353271,
"learning_rate": 3.744741426611464e-05,
"loss": 2.9258,
"step": 433
},
{
"epoch": 0.5925925925925926,
"grad_norm": 1.2859567403793335,
"learning_rate": 3.7433788382989106e-05,
"loss": 2.683,
"step": 434
},
{
"epoch": 0.593958013312852,
"grad_norm": 1.3327406644821167,
"learning_rate": 3.742012901947443e-05,
"loss": 3.0007,
"step": 435
},
{
"epoch": 0.5953234340331115,
"grad_norm": 1.3187899589538574,
"learning_rate": 3.7406436205201755e-05,
"loss": 2.9108,
"step": 436
},
{
"epoch": 0.5966888547533709,
"grad_norm": 1.3259321451187134,
"learning_rate": 3.7392709969874785e-05,
"loss": 3.0018,
"step": 437
},
{
"epoch": 0.5980542754736303,
"grad_norm": 1.2598587274551392,
"learning_rate": 3.737895034326973e-05,
"loss": 2.7834,
"step": 438
},
{
"epoch": 0.5994196961938898,
"grad_norm": 1.2821201086044312,
"learning_rate": 3.736515735523523e-05,
"loss": 2.8404,
"step": 439
},
{
"epoch": 0.6007851169141492,
"grad_norm": 1.3087409734725952,
"learning_rate": 3.735133103569231e-05,
"loss": 2.7119,
"step": 440
},
{
"epoch": 0.6021505376344086,
"grad_norm": 1.2687911987304688,
"learning_rate": 3.733747141463428e-05,
"loss": 2.7103,
"step": 441
},
{
"epoch": 0.603515958354668,
"grad_norm": 1.2933452129364014,
"learning_rate": 3.7323578522126704e-05,
"loss": 2.8893,
"step": 442
},
{
"epoch": 0.6048813790749274,
"grad_norm": 1.3670916557312012,
"learning_rate": 3.730965238830732e-05,
"loss": 2.9303,
"step": 443
},
{
"epoch": 0.6062467997951869,
"grad_norm": 1.3215402364730835,
"learning_rate": 3.729569304338597e-05,
"loss": 2.8011,
"step": 444
},
{
"epoch": 0.6076122205154463,
"grad_norm": 1.316300868988037,
"learning_rate": 3.7281700517644554e-05,
"loss": 2.9002,
"step": 445
},
{
"epoch": 0.6089776412357057,
"grad_norm": 1.4437733888626099,
"learning_rate": 3.726767484143694e-05,
"loss": 2.9884,
"step": 446
},
{
"epoch": 0.6103430619559652,
"grad_norm": 1.4378395080566406,
"learning_rate": 3.72536160451889e-05,
"loss": 2.8825,
"step": 447
},
{
"epoch": 0.6117084826762246,
"grad_norm": 1.3129379749298096,
"learning_rate": 3.7239524159398076e-05,
"loss": 2.8515,
"step": 448
},
{
"epoch": 0.613073903396484,
"grad_norm": 1.3621375560760498,
"learning_rate": 3.722539921463387e-05,
"loss": 2.8922,
"step": 449
},
{
"epoch": 0.6144393241167435,
"grad_norm": 1.4121601581573486,
"learning_rate": 3.7211241241537414e-05,
"loss": 3.0759,
"step": 450
},
{
"epoch": 0.6158047448370029,
"grad_norm": 1.3335072994232178,
"learning_rate": 3.7197050270821485e-05,
"loss": 2.9079,
"step": 451
},
{
"epoch": 0.6171701655572623,
"grad_norm": 1.4576181173324585,
"learning_rate": 3.7182826333270436e-05,
"loss": 3.1032,
"step": 452
},
{
"epoch": 0.6185355862775218,
"grad_norm": 1.3565021753311157,
"learning_rate": 3.7168569459740126e-05,
"loss": 2.8538,
"step": 453
},
{
"epoch": 0.6199010069977812,
"grad_norm": 1.336393117904663,
"learning_rate": 3.715427968115789e-05,
"loss": 2.8774,
"step": 454
},
{
"epoch": 0.6212664277180406,
"grad_norm": 1.362845540046692,
"learning_rate": 3.713995702852241e-05,
"loss": 2.9532,
"step": 455
},
{
"epoch": 0.6226318484383001,
"grad_norm": 1.335113525390625,
"learning_rate": 3.71256015329037e-05,
"loss": 2.9395,
"step": 456
},
{
"epoch": 0.6239972691585595,
"grad_norm": 1.3559925556182861,
"learning_rate": 3.711121322544302e-05,
"loss": 2.9057,
"step": 457
},
{
"epoch": 0.6253626898788189,
"grad_norm": 1.445767879486084,
"learning_rate": 3.709679213735283e-05,
"loss": 2.888,
"step": 458
},
{
"epoch": 0.6267281105990783,
"grad_norm": 1.4072614908218384,
"learning_rate": 3.708233829991663e-05,
"loss": 2.9206,
"step": 459
},
{
"epoch": 0.6280935313193378,
"grad_norm": 1.2806531190872192,
"learning_rate": 3.706785174448905e-05,
"loss": 2.7264,
"step": 460
},
{
"epoch": 0.6294589520395972,
"grad_norm": 1.3245452642440796,
"learning_rate": 3.7053332502495636e-05,
"loss": 2.8157,
"step": 461
},
{
"epoch": 0.6308243727598566,
"grad_norm": 1.3116190433502197,
"learning_rate": 3.703878060543286e-05,
"loss": 2.8256,
"step": 462
},
{
"epoch": 0.6321897934801161,
"grad_norm": 1.2462058067321777,
"learning_rate": 3.702419608486803e-05,
"loss": 2.8425,
"step": 463
},
{
"epoch": 0.6335552142003755,
"grad_norm": 1.2823293209075928,
"learning_rate": 3.700957897243924e-05,
"loss": 2.7328,
"step": 464
},
{
"epoch": 0.6349206349206349,
"grad_norm": 1.3262630701065063,
"learning_rate": 3.6994929299855254e-05,
"loss": 2.936,
"step": 465
},
{
"epoch": 0.6362860556408944,
"grad_norm": 1.4157698154449463,
"learning_rate": 3.6980247098895485e-05,
"loss": 2.7993,
"step": 466
},
{
"epoch": 0.6376514763611538,
"grad_norm": 1.3853559494018555,
"learning_rate": 3.696553240140993e-05,
"loss": 2.9282,
"step": 467
},
{
"epoch": 0.6390168970814132,
"grad_norm": 1.4059337377548218,
"learning_rate": 3.695078523931903e-05,
"loss": 3.0765,
"step": 468
},
{
"epoch": 0.6403823178016727,
"grad_norm": 1.387112021446228,
"learning_rate": 3.693600564461371e-05,
"loss": 2.8957,
"step": 469
},
{
"epoch": 0.6417477385219321,
"grad_norm": 1.3582907915115356,
"learning_rate": 3.6921193649355206e-05,
"loss": 3.0103,
"step": 470
},
{
"epoch": 0.6431131592421915,
"grad_norm": 1.3183821439743042,
"learning_rate": 3.6906349285675065e-05,
"loss": 2.8717,
"step": 471
},
{
"epoch": 0.644478579962451,
"grad_norm": 1.261896014213562,
"learning_rate": 3.6891472585775034e-05,
"loss": 2.8289,
"step": 472
},
{
"epoch": 0.6458440006827103,
"grad_norm": 1.3002463579177856,
"learning_rate": 3.687656358192702e-05,
"loss": 2.8723,
"step": 473
},
{
"epoch": 0.6472094214029698,
"grad_norm": 1.3367985486984253,
"learning_rate": 3.6861622306473e-05,
"loss": 2.9214,
"step": 474
},
{
"epoch": 0.6485748421232292,
"grad_norm": 1.2751028537750244,
"learning_rate": 3.684664879182496e-05,
"loss": 2.8831,
"step": 475
},
{
"epoch": 0.6499402628434886,
"grad_norm": 1.2681186199188232,
"learning_rate": 3.6831643070464806e-05,
"loss": 2.7936,
"step": 476
},
{
"epoch": 0.6513056835637481,
"grad_norm": 1.314717173576355,
"learning_rate": 3.681660517494436e-05,
"loss": 2.8979,
"step": 477
},
{
"epoch": 0.6526711042840075,
"grad_norm": 1.2387361526489258,
"learning_rate": 3.680153513788517e-05,
"loss": 2.8042,
"step": 478
},
{
"epoch": 0.6540365250042669,
"grad_norm": 1.325661540031433,
"learning_rate": 3.678643299197855e-05,
"loss": 2.9468,
"step": 479
},
{
"epoch": 0.6554019457245264,
"grad_norm": 1.2384459972381592,
"learning_rate": 3.677129876998548e-05,
"loss": 2.7237,
"step": 480
},
{
"epoch": 0.6567673664447858,
"grad_norm": 1.3409217596054077,
"learning_rate": 3.675613250473648e-05,
"loss": 2.9,
"step": 481
},
{
"epoch": 0.6581327871650452,
"grad_norm": 1.2916115522384644,
"learning_rate": 3.6740934229131606e-05,
"loss": 2.8951,
"step": 482
},
{
"epoch": 0.6594982078853047,
"grad_norm": 1.2838852405548096,
"learning_rate": 3.672570397614036e-05,
"loss": 2.8885,
"step": 483
},
{
"epoch": 0.6608636286055641,
"grad_norm": 1.3061528205871582,
"learning_rate": 3.67104417788016e-05,
"loss": 2.8753,
"step": 484
},
{
"epoch": 0.6622290493258235,
"grad_norm": 1.2723528146743774,
"learning_rate": 3.669514767022349e-05,
"loss": 2.8564,
"step": 485
},
{
"epoch": 0.663594470046083,
"grad_norm": 1.3076740503311157,
"learning_rate": 3.6679821683583396e-05,
"loss": 2.8132,
"step": 486
},
{
"epoch": 0.6649598907663424,
"grad_norm": 1.334466576576233,
"learning_rate": 3.666446385212787e-05,
"loss": 2.9244,
"step": 487
},
{
"epoch": 0.6663253114866018,
"grad_norm": 1.3105363845825195,
"learning_rate": 3.664907420917252e-05,
"loss": 2.9365,
"step": 488
},
{
"epoch": 0.6676907322068613,
"grad_norm": 1.322219729423523,
"learning_rate": 3.6633652788101975e-05,
"loss": 2.9781,
"step": 489
},
{
"epoch": 0.6690561529271206,
"grad_norm": 1.2741801738739014,
"learning_rate": 3.66181996223698e-05,
"loss": 2.8713,
"step": 490
},
{
"epoch": 0.6704215736473801,
"grad_norm": 1.3462766408920288,
"learning_rate": 3.66027147454984e-05,
"loss": 2.9106,
"step": 491
},
{
"epoch": 0.6717869943676396,
"grad_norm": 1.3164334297180176,
"learning_rate": 3.6587198191079014e-05,
"loss": 2.804,
"step": 492
},
{
"epoch": 0.6731524150878989,
"grad_norm": 1.349804162979126,
"learning_rate": 3.657164999277156e-05,
"loss": 2.831,
"step": 493
},
{
"epoch": 0.6745178358081584,
"grad_norm": 1.2329461574554443,
"learning_rate": 3.6556070184304624e-05,
"loss": 2.8624,
"step": 494
},
{
"epoch": 0.6758832565284179,
"grad_norm": 1.2478387355804443,
"learning_rate": 3.654045879947535e-05,
"loss": 2.8427,
"step": 495
},
{
"epoch": 0.6772486772486772,
"grad_norm": 1.3101484775543213,
"learning_rate": 3.652481587214939e-05,
"loss": 2.8203,
"step": 496
},
{
"epoch": 0.6786140979689367,
"grad_norm": 1.3023991584777832,
"learning_rate": 3.650914143626081e-05,
"loss": 2.7205,
"step": 497
},
{
"epoch": 0.6799795186891961,
"grad_norm": 1.299358606338501,
"learning_rate": 3.6493435525812034e-05,
"loss": 2.9856,
"step": 498
},
{
"epoch": 0.6813449394094555,
"grad_norm": 1.4544318914413452,
"learning_rate": 3.647769817487377e-05,
"loss": 2.9477,
"step": 499
},
{
"epoch": 0.682710360129715,
"grad_norm": 1.2803057432174683,
"learning_rate": 3.6461929417584915e-05,
"loss": 2.8648,
"step": 500
},
{
"epoch": 0.6840757808499744,
"grad_norm": 1.3588660955429077,
"learning_rate": 3.6446129288152514e-05,
"loss": 2.8725,
"step": 501
},
{
"epoch": 0.6854412015702338,
"grad_norm": 1.2590023279190063,
"learning_rate": 3.643029782085165e-05,
"loss": 2.9037,
"step": 502
},
{
"epoch": 0.6868066222904933,
"grad_norm": 1.249646782875061,
"learning_rate": 3.64144350500254e-05,
"loss": 2.8656,
"step": 503
},
{
"epoch": 0.6881720430107527,
"grad_norm": 1.3663312196731567,
"learning_rate": 3.639854101008472e-05,
"loss": 3.0374,
"step": 504
},
{
"epoch": 0.6895374637310121,
"grad_norm": 1.511388897895813,
"learning_rate": 3.638261573550844e-05,
"loss": 2.8988,
"step": 505
},
{
"epoch": 0.6909028844512716,
"grad_norm": 1.2522783279418945,
"learning_rate": 3.636665926084313e-05,
"loss": 2.7906,
"step": 506
},
{
"epoch": 0.6922683051715309,
"grad_norm": 1.3429816961288452,
"learning_rate": 3.6350671620703025e-05,
"loss": 3.0203,
"step": 507
},
{
"epoch": 0.6936337258917904,
"grad_norm": 1.2900177240371704,
"learning_rate": 3.633465284976998e-05,
"loss": 2.8085,
"step": 508
},
{
"epoch": 0.6949991466120499,
"grad_norm": 1.2798503637313843,
"learning_rate": 3.631860298279338e-05,
"loss": 2.8639,
"step": 509
},
{
"epoch": 0.6963645673323092,
"grad_norm": 1.3201751708984375,
"learning_rate": 3.630252205459009e-05,
"loss": 2.833,
"step": 510
},
{
"epoch": 0.6977299880525687,
"grad_norm": 1.2585594654083252,
"learning_rate": 3.628641010004431e-05,
"loss": 2.7823,
"step": 511
},
{
"epoch": 0.6990954087728282,
"grad_norm": 1.3217718601226807,
"learning_rate": 3.6270267154107576e-05,
"loss": 2.9667,
"step": 512
},
{
"epoch": 0.7004608294930875,
"grad_norm": 1.3991295099258423,
"learning_rate": 3.625409325179865e-05,
"loss": 2.9473,
"step": 513
},
{
"epoch": 0.701826250213347,
"grad_norm": 1.2950828075408936,
"learning_rate": 3.623788842820344e-05,
"loss": 2.9907,
"step": 514
},
{
"epoch": 0.7031916709336065,
"grad_norm": 1.302770972251892,
"learning_rate": 3.622165271847494e-05,
"loss": 2.8413,
"step": 515
},
{
"epoch": 0.7045570916538658,
"grad_norm": 1.384196162223816,
"learning_rate": 3.6205386157833144e-05,
"loss": 3.0346,
"step": 516
},
{
"epoch": 0.7059225123741253,
"grad_norm": 1.262619137763977,
"learning_rate": 3.618908878156496e-05,
"loss": 2.8783,
"step": 517
},
{
"epoch": 0.7072879330943848,
"grad_norm": 1.3378034830093384,
"learning_rate": 3.617276062502415e-05,
"loss": 2.8852,
"step": 518
},
{
"epoch": 0.7086533538146441,
"grad_norm": 1.303502082824707,
"learning_rate": 3.615640172363126e-05,
"loss": 2.8648,
"step": 519
},
{
"epoch": 0.7100187745349036,
"grad_norm": 1.3309911489486694,
"learning_rate": 3.61400121128735e-05,
"loss": 3.0171,
"step": 520
},
{
"epoch": 0.7113841952551629,
"grad_norm": 1.3169946670532227,
"learning_rate": 3.6123591828304726e-05,
"loss": 2.8624,
"step": 521
},
{
"epoch": 0.7127496159754224,
"grad_norm": 1.2952288389205933,
"learning_rate": 3.610714090554533e-05,
"loss": 2.7668,
"step": 522
},
{
"epoch": 0.7141150366956819,
"grad_norm": 1.2812696695327759,
"learning_rate": 3.609065938028216e-05,
"loss": 2.9468,
"step": 523
},
{
"epoch": 0.7154804574159412,
"grad_norm": 1.281525731086731,
"learning_rate": 3.6074147288268456e-05,
"loss": 2.737,
"step": 524
},
{
"epoch": 0.7168458781362007,
"grad_norm": 1.3946808576583862,
"learning_rate": 3.605760466532374e-05,
"loss": 3.1228,
"step": 525
},
{
"epoch": 0.7182112988564602,
"grad_norm": 1.3583076000213623,
"learning_rate": 3.604103154733382e-05,
"loss": 2.9225,
"step": 526
},
{
"epoch": 0.7195767195767195,
"grad_norm": 1.2891432046890259,
"learning_rate": 3.602442797025061e-05,
"loss": 2.7942,
"step": 527
},
{
"epoch": 0.720942140296979,
"grad_norm": 1.3038846254348755,
"learning_rate": 3.6007793970092095e-05,
"loss": 2.9226,
"step": 528
},
{
"epoch": 0.7223075610172385,
"grad_norm": 1.3120319843292236,
"learning_rate": 3.599112958294229e-05,
"loss": 2.9012,
"step": 529
},
{
"epoch": 0.7236729817374978,
"grad_norm": 1.3864185810089111,
"learning_rate": 3.597443484495112e-05,
"loss": 2.9309,
"step": 530
},
{
"epoch": 0.7250384024577573,
"grad_norm": 1.2283259630203247,
"learning_rate": 3.595770979233433e-05,
"loss": 2.8252,
"step": 531
},
{
"epoch": 0.7264038231780168,
"grad_norm": 1.2426667213439941,
"learning_rate": 3.594095446137343e-05,
"loss": 2.8775,
"step": 532
},
{
"epoch": 0.7277692438982761,
"grad_norm": 1.2631727457046509,
"learning_rate": 3.592416888841564e-05,
"loss": 2.8984,
"step": 533
},
{
"epoch": 0.7291346646185356,
"grad_norm": 1.3043859004974365,
"learning_rate": 3.5907353109873745e-05,
"loss": 2.9097,
"step": 534
},
{
"epoch": 0.7305000853387951,
"grad_norm": 1.2307614088058472,
"learning_rate": 3.589050716222609e-05,
"loss": 2.8628,
"step": 535
},
{
"epoch": 0.7318655060590544,
"grad_norm": 1.2963740825653076,
"learning_rate": 3.5873631082016445e-05,
"loss": 2.8542,
"step": 536
},
{
"epoch": 0.7332309267793139,
"grad_norm": 1.3001391887664795,
"learning_rate": 3.585672490585394e-05,
"loss": 2.9331,
"step": 537
},
{
"epoch": 0.7345963474995733,
"grad_norm": 1.2801858186721802,
"learning_rate": 3.583978867041302e-05,
"loss": 2.8885,
"step": 538
},
{
"epoch": 0.7359617682198327,
"grad_norm": 1.2577100992202759,
"learning_rate": 3.5822822412433304e-05,
"loss": 2.8451,
"step": 539
},
{
"epoch": 0.7373271889400922,
"grad_norm": 1.3069607019424438,
"learning_rate": 3.5805826168719564e-05,
"loss": 2.8202,
"step": 540
},
{
"epoch": 0.7386926096603516,
"grad_norm": 1.3131080865859985,
"learning_rate": 3.578879997614161e-05,
"loss": 2.9385,
"step": 541
},
{
"epoch": 0.740058030380611,
"grad_norm": 1.2916386127471924,
"learning_rate": 3.577174387163421e-05,
"loss": 2.9576,
"step": 542
},
{
"epoch": 0.7414234511008705,
"grad_norm": 1.3010426759719849,
"learning_rate": 3.5754657892197056e-05,
"loss": 2.9518,
"step": 543
},
{
"epoch": 0.7427888718211298,
"grad_norm": 1.2421988248825073,
"learning_rate": 3.5737542074894595e-05,
"loss": 2.7443,
"step": 544
},
{
"epoch": 0.7441542925413893,
"grad_norm": 1.293596863746643,
"learning_rate": 3.572039645685603e-05,
"loss": 2.9496,
"step": 545
},
{
"epoch": 0.7455197132616488,
"grad_norm": 1.3112589120864868,
"learning_rate": 3.5703221075275227e-05,
"loss": 3.0441,
"step": 546
},
{
"epoch": 0.7468851339819081,
"grad_norm": 1.2407242059707642,
"learning_rate": 3.5686015967410584e-05,
"loss": 2.7161,
"step": 547
},
{
"epoch": 0.7482505547021676,
"grad_norm": 1.248875379562378,
"learning_rate": 3.5668781170584996e-05,
"loss": 2.7916,
"step": 548
},
{
"epoch": 0.7496159754224271,
"grad_norm": 1.3158230781555176,
"learning_rate": 3.565151672218578e-05,
"loss": 2.8347,
"step": 549
},
{
"epoch": 0.7509813961426864,
"grad_norm": 1.344825029373169,
"learning_rate": 3.563422265966454e-05,
"loss": 3.0494,
"step": 550
},
{
"epoch": 0.7523468168629459,
"grad_norm": 1.3518285751342773,
"learning_rate": 3.5616899020537166e-05,
"loss": 2.7562,
"step": 551
},
{
"epoch": 0.7537122375832054,
"grad_norm": 1.2998853921890259,
"learning_rate": 3.559954584238366e-05,
"loss": 2.7406,
"step": 552
},
{
"epoch": 0.7550776583034647,
"grad_norm": 1.257563829421997,
"learning_rate": 3.558216316284815e-05,
"loss": 2.9462,
"step": 553
},
{
"epoch": 0.7564430790237242,
"grad_norm": 1.2820138931274414,
"learning_rate": 3.556475101963872e-05,
"loss": 2.8143,
"step": 554
},
{
"epoch": 0.7578084997439836,
"grad_norm": 1.3024104833602905,
"learning_rate": 3.5547309450527394e-05,
"loss": 2.7173,
"step": 555
},
{
"epoch": 0.759173920464243,
"grad_norm": 1.2970681190490723,
"learning_rate": 3.5529838493350034e-05,
"loss": 2.8941,
"step": 556
},
{
"epoch": 0.7605393411845025,
"grad_norm": 1.2700526714324951,
"learning_rate": 3.551233818600623e-05,
"loss": 2.6633,
"step": 557
},
{
"epoch": 0.7619047619047619,
"grad_norm": 1.3061867952346802,
"learning_rate": 3.5494808566459267e-05,
"loss": 2.9219,
"step": 558
},
{
"epoch": 0.7632701826250213,
"grad_norm": 1.2967194318771362,
"learning_rate": 3.5477249672735976e-05,
"loss": 2.9349,
"step": 559
},
{
"epoch": 0.7646356033452808,
"grad_norm": 1.313602089881897,
"learning_rate": 3.5459661542926756e-05,
"loss": 2.8445,
"step": 560
},
{
"epoch": 0.7660010240655402,
"grad_norm": 1.219576358795166,
"learning_rate": 3.5442044215185363e-05,
"loss": 2.8255,
"step": 561
},
{
"epoch": 0.7673664447857996,
"grad_norm": 1.30018949508667,
"learning_rate": 3.542439772772893e-05,
"loss": 2.9768,
"step": 562
},
{
"epoch": 0.7687318655060591,
"grad_norm": 1.2077510356903076,
"learning_rate": 3.540672211883783e-05,
"loss": 2.79,
"step": 563
},
{
"epoch": 0.7700972862263185,
"grad_norm": 1.2231390476226807,
"learning_rate": 3.538901742685563e-05,
"loss": 2.7321,
"step": 564
},
{
"epoch": 0.7714627069465779,
"grad_norm": 1.233472466468811,
"learning_rate": 3.537128369018895e-05,
"loss": 2.7442,
"step": 565
},
{
"epoch": 0.7728281276668374,
"grad_norm": 1.2162251472473145,
"learning_rate": 3.535352094730746e-05,
"loss": 2.7649,
"step": 566
},
{
"epoch": 0.7741935483870968,
"grad_norm": 1.3011486530303955,
"learning_rate": 3.533572923674372e-05,
"loss": 3.0022,
"step": 567
},
{
"epoch": 0.7755589691073562,
"grad_norm": 1.2810593843460083,
"learning_rate": 3.531790859709313e-05,
"loss": 2.8362,
"step": 568
},
{
"epoch": 0.7769243898276156,
"grad_norm": 1.3131102323532104,
"learning_rate": 3.5300059067013885e-05,
"loss": 2.8971,
"step": 569
},
{
"epoch": 0.778289810547875,
"grad_norm": 1.291644811630249,
"learning_rate": 3.528218068522679e-05,
"loss": 2.7924,
"step": 570
},
{
"epoch": 0.7796552312681345,
"grad_norm": 1.2757987976074219,
"learning_rate": 3.526427349051528e-05,
"loss": 2.763,
"step": 571
},
{
"epoch": 0.7810206519883939,
"grad_norm": 1.302538275718689,
"learning_rate": 3.52463375217253e-05,
"loss": 2.8442,
"step": 572
},
{
"epoch": 0.7823860727086533,
"grad_norm": 1.339532732963562,
"learning_rate": 3.522837281776519e-05,
"loss": 2.8294,
"step": 573
},
{
"epoch": 0.7837514934289128,
"grad_norm": 1.2920910120010376,
"learning_rate": 3.521037941760562e-05,
"loss": 2.9086,
"step": 574
},
{
"epoch": 0.7851169141491722,
"grad_norm": 1.3217432498931885,
"learning_rate": 3.5192357360279536e-05,
"loss": 2.8346,
"step": 575
},
{
"epoch": 0.7864823348694316,
"grad_norm": 1.3719730377197266,
"learning_rate": 3.517430668488203e-05,
"loss": 2.8723,
"step": 576
},
{
"epoch": 0.7878477555896911,
"grad_norm": 1.2575520277023315,
"learning_rate": 3.5156227430570294e-05,
"loss": 2.8521,
"step": 577
},
{
"epoch": 0.7892131763099505,
"grad_norm": 1.3220593929290771,
"learning_rate": 3.5138119636563495e-05,
"loss": 2.9331,
"step": 578
},
{
"epoch": 0.7905785970302099,
"grad_norm": 1.2344598770141602,
"learning_rate": 3.5119983342142726e-05,
"loss": 2.8002,
"step": 579
},
{
"epoch": 0.7919440177504694,
"grad_norm": 1.2928383350372314,
"learning_rate": 3.51018185866509e-05,
"loss": 2.7776,
"step": 580
},
{
"epoch": 0.7933094384707288,
"grad_norm": 1.3142491579055786,
"learning_rate": 3.508362540949266e-05,
"loss": 2.8554,
"step": 581
},
{
"epoch": 0.7946748591909882,
"grad_norm": 1.287967324256897,
"learning_rate": 3.506540385013433e-05,
"loss": 2.9613,
"step": 582
},
{
"epoch": 0.7960402799112477,
"grad_norm": 1.3606940507888794,
"learning_rate": 3.5047153948103775e-05,
"loss": 2.9892,
"step": 583
},
{
"epoch": 0.7974057006315071,
"grad_norm": 1.281965732574463,
"learning_rate": 3.502887574299036e-05,
"loss": 2.7951,
"step": 584
},
{
"epoch": 0.7987711213517665,
"grad_norm": 1.2725483179092407,
"learning_rate": 3.501056927444485e-05,
"loss": 2.7232,
"step": 585
},
{
"epoch": 0.8001365420720259,
"grad_norm": 1.245137333869934,
"learning_rate": 3.499223458217931e-05,
"loss": 2.6519,
"step": 586
},
{
"epoch": 0.8015019627922854,
"grad_norm": 1.3419994115829468,
"learning_rate": 3.497387170596703e-05,
"loss": 3.0079,
"step": 587
},
{
"epoch": 0.8028673835125448,
"grad_norm": 1.3098769187927246,
"learning_rate": 3.4955480685642454e-05,
"loss": 2.9278,
"step": 588
},
{
"epoch": 0.8042328042328042,
"grad_norm": 1.3644626140594482,
"learning_rate": 3.493706156110107e-05,
"loss": 2.9749,
"step": 589
},
{
"epoch": 0.8055982249530637,
"grad_norm": 1.3489822149276733,
"learning_rate": 3.491861437229933e-05,
"loss": 2.989,
"step": 590
},
{
"epoch": 0.8069636456733231,
"grad_norm": 1.279343605041504,
"learning_rate": 3.490013915925456e-05,
"loss": 2.868,
"step": 591
},
{
"epoch": 0.8083290663935825,
"grad_norm": 1.2475160360336304,
"learning_rate": 3.4881635962044894e-05,
"loss": 2.8691,
"step": 592
},
{
"epoch": 0.809694487113842,
"grad_norm": 1.3186317682266235,
"learning_rate": 3.486310482080917e-05,
"loss": 2.8482,
"step": 593
},
{
"epoch": 0.8110599078341014,
"grad_norm": 1.2843761444091797,
"learning_rate": 3.4844545775746835e-05,
"loss": 2.7708,
"step": 594
},
{
"epoch": 0.8124253285543608,
"grad_norm": 1.2376091480255127,
"learning_rate": 3.482595886711786e-05,
"loss": 2.8785,
"step": 595
},
{
"epoch": 0.8137907492746203,
"grad_norm": 1.2332408428192139,
"learning_rate": 3.480734413524268e-05,
"loss": 2.7759,
"step": 596
},
{
"epoch": 0.8151561699948797,
"grad_norm": 1.2218464612960815,
"learning_rate": 3.478870162050209e-05,
"loss": 2.8464,
"step": 597
},
{
"epoch": 0.8165215907151391,
"grad_norm": 1.2308814525604248,
"learning_rate": 3.4770031363337126e-05,
"loss": 2.6718,
"step": 598
},
{
"epoch": 0.8178870114353985,
"grad_norm": 1.286547064781189,
"learning_rate": 3.4751333404249026e-05,
"loss": 2.7857,
"step": 599
},
{
"epoch": 0.819252432155658,
"grad_norm": 1.261696457862854,
"learning_rate": 3.4732607783799125e-05,
"loss": 2.8666,
"step": 600
},
{
"epoch": 0.8206178528759174,
"grad_norm": 2.709296703338623,
"learning_rate": 3.4713854542608744e-05,
"loss": 3.0213,
"step": 601
},
{
"epoch": 0.8219832735961768,
"grad_norm": 1.297171950340271,
"learning_rate": 3.469507372135914e-05,
"loss": 2.8193,
"step": 602
},
{
"epoch": 0.8233486943164362,
"grad_norm": 1.3143333196640015,
"learning_rate": 3.4676265360791396e-05,
"loss": 2.9438,
"step": 603
},
{
"epoch": 0.8247141150366957,
"grad_norm": 1.2548006772994995,
"learning_rate": 3.465742950170633e-05,
"loss": 2.7751,
"step": 604
},
{
"epoch": 0.8260795357569551,
"grad_norm": 1.2405400276184082,
"learning_rate": 3.4638566184964414e-05,
"loss": 2.7434,
"step": 605
},
{
"epoch": 0.8274449564772145,
"grad_norm": 1.3934334516525269,
"learning_rate": 3.4619675451485686e-05,
"loss": 2.9338,
"step": 606
},
{
"epoch": 0.828810377197474,
"grad_norm": 1.2526410818099976,
"learning_rate": 3.460075734224966e-05,
"loss": 2.7289,
"step": 607
},
{
"epoch": 0.8301757979177334,
"grad_norm": 1.2939361333847046,
"learning_rate": 3.458181189829523e-05,
"loss": 2.7588,
"step": 608
},
{
"epoch": 0.8315412186379928,
"grad_norm": 1.3009428977966309,
"learning_rate": 3.4562839160720586e-05,
"loss": 2.8298,
"step": 609
},
{
"epoch": 0.8329066393582523,
"grad_norm": 1.3296343088150024,
"learning_rate": 3.4543839170683135e-05,
"loss": 2.9854,
"step": 610
},
{
"epoch": 0.8342720600785117,
"grad_norm": 1.346441626548767,
"learning_rate": 3.45248119693994e-05,
"loss": 2.9774,
"step": 611
},
{
"epoch": 0.8356374807987711,
"grad_norm": 1.29673433303833,
"learning_rate": 3.4505757598144926e-05,
"loss": 2.8763,
"step": 612
},
{
"epoch": 0.8370029015190306,
"grad_norm": 1.242966651916504,
"learning_rate": 3.4486676098254206e-05,
"loss": 2.7548,
"step": 613
},
{
"epoch": 0.83836832223929,
"grad_norm": 1.3075823783874512,
"learning_rate": 3.4467567511120577e-05,
"loss": 2.7328,
"step": 614
},
{
"epoch": 0.8397337429595494,
"grad_norm": 1.2685598134994507,
"learning_rate": 3.444843187819614e-05,
"loss": 2.8541,
"step": 615
},
{
"epoch": 0.8410991636798089,
"grad_norm": 1.303835153579712,
"learning_rate": 3.442926924099167e-05,
"loss": 2.7868,
"step": 616
},
{
"epoch": 0.8424645844000682,
"grad_norm": 1.2787021398544312,
"learning_rate": 3.441007964107651e-05,
"loss": 2.8887,
"step": 617
},
{
"epoch": 0.8438300051203277,
"grad_norm": 1.3013572692871094,
"learning_rate": 3.43908631200785e-05,
"loss": 2.9006,
"step": 618
},
{
"epoch": 0.8451954258405872,
"grad_norm": 1.3113733530044556,
"learning_rate": 3.437161971968388e-05,
"loss": 2.7733,
"step": 619
},
{
"epoch": 0.8465608465608465,
"grad_norm": 1.2344701290130615,
"learning_rate": 3.435234948163721e-05,
"loss": 2.7694,
"step": 620
},
{
"epoch": 0.847926267281106,
"grad_norm": 1.2834700345993042,
"learning_rate": 3.433305244774124e-05,
"loss": 2.7795,
"step": 621
},
{
"epoch": 0.8492916880013655,
"grad_norm": 1.2224047183990479,
"learning_rate": 3.4313728659856873e-05,
"loss": 2.7868,
"step": 622
},
{
"epoch": 0.8506571087216248,
"grad_norm": 1.3431744575500488,
"learning_rate": 3.4294378159903044e-05,
"loss": 2.9298,
"step": 623
},
{
"epoch": 0.8520225294418843,
"grad_norm": 1.2859256267547607,
"learning_rate": 3.4275000989856634e-05,
"loss": 2.7768,
"step": 624
},
{
"epoch": 0.8533879501621437,
"grad_norm": 1.2665132284164429,
"learning_rate": 3.4255597191752365e-05,
"loss": 2.7742,
"step": 625
},
{
"epoch": 0.8547533708824031,
"grad_norm": 1.2370717525482178,
"learning_rate": 3.423616680768276e-05,
"loss": 2.9145,
"step": 626
},
{
"epoch": 0.8561187916026626,
"grad_norm": 1.286116600036621,
"learning_rate": 3.421670987979796e-05,
"loss": 2.7915,
"step": 627
},
{
"epoch": 0.857484212322922,
"grad_norm": 1.3869763612747192,
"learning_rate": 3.419722645030574e-05,
"loss": 3.0081,
"step": 628
},
{
"epoch": 0.8588496330431814,
"grad_norm": 1.2512043714523315,
"learning_rate": 3.417771656147133e-05,
"loss": 2.781,
"step": 629
},
{
"epoch": 0.8602150537634409,
"grad_norm": 1.2461572885513306,
"learning_rate": 3.4158180255617365e-05,
"loss": 2.8337,
"step": 630
},
{
"epoch": 0.8615804744837003,
"grad_norm": 1.2329281568527222,
"learning_rate": 3.413861757512381e-05,
"loss": 2.7472,
"step": 631
},
{
"epoch": 0.8629458952039597,
"grad_norm": 1.2923029661178589,
"learning_rate": 3.411902856242782e-05,
"loss": 2.812,
"step": 632
},
{
"epoch": 0.8643113159242192,
"grad_norm": 1.2564419507980347,
"learning_rate": 3.4099413260023654e-05,
"loss": 2.8627,
"step": 633
},
{
"epoch": 0.8656767366444785,
"grad_norm": 1.3133330345153809,
"learning_rate": 3.407977171046267e-05,
"loss": 2.9075,
"step": 634
},
{
"epoch": 0.867042157364738,
"grad_norm": 1.2775532007217407,
"learning_rate": 3.406010395635307e-05,
"loss": 2.8755,
"step": 635
},
{
"epoch": 0.8684075780849975,
"grad_norm": 1.329980492591858,
"learning_rate": 3.404041004035998e-05,
"loss": 2.9359,
"step": 636
},
{
"epoch": 0.8697729988052568,
"grad_norm": 1.2722467184066772,
"learning_rate": 3.4020690005205245e-05,
"loss": 2.9149,
"step": 637
},
{
"epoch": 0.8711384195255163,
"grad_norm": 1.3029361963272095,
"learning_rate": 3.400094389366737e-05,
"loss": 2.9161,
"step": 638
},
{
"epoch": 0.8725038402457758,
"grad_norm": 1.2868540287017822,
"learning_rate": 3.3981171748581434e-05,
"loss": 2.6881,
"step": 639
},
{
"epoch": 0.8738692609660351,
"grad_norm": 1.223644733428955,
"learning_rate": 3.396137361283898e-05,
"loss": 2.6368,
"step": 640
},
{
"epoch": 0.8752346816862946,
"grad_norm": 1.274714708328247,
"learning_rate": 3.3941549529387936e-05,
"loss": 2.8138,
"step": 641
},
{
"epoch": 0.8766001024065541,
"grad_norm": 1.3436501026153564,
"learning_rate": 3.392169954123253e-05,
"loss": 2.7889,
"step": 642
},
{
"epoch": 0.8779655231268134,
"grad_norm": 1.2697542905807495,
"learning_rate": 3.390182369143319e-05,
"loss": 2.7488,
"step": 643
},
{
"epoch": 0.8793309438470729,
"grad_norm": 1.2328263521194458,
"learning_rate": 3.3881922023106414e-05,
"loss": 2.7076,
"step": 644
},
{
"epoch": 0.8806963645673324,
"grad_norm": 1.283242106437683,
"learning_rate": 3.3861994579424725e-05,
"loss": 2.7996,
"step": 645
},
{
"epoch": 0.8820617852875917,
"grad_norm": 1.290000557899475,
"learning_rate": 3.384204140361657e-05,
"loss": 2.9553,
"step": 646
},
{
"epoch": 0.8834272060078512,
"grad_norm": 1.3263002634048462,
"learning_rate": 3.38220625389662e-05,
"loss": 2.8436,
"step": 647
},
{
"epoch": 0.8847926267281107,
"grad_norm": 1.296995759010315,
"learning_rate": 3.380205802881362e-05,
"loss": 2.8812,
"step": 648
},
{
"epoch": 0.88615804744837,
"grad_norm": 1.1882314682006836,
"learning_rate": 3.378202791655441e-05,
"loss": 2.6222,
"step": 649
},
{
"epoch": 0.8875234681686295,
"grad_norm": 1.2342636585235596,
"learning_rate": 3.376197224563977e-05,
"loss": 2.731,
"step": 650
},
{
"epoch": 0.8888888888888888,
"grad_norm": 1.2116607427597046,
"learning_rate": 3.374189105957627e-05,
"loss": 2.7918,
"step": 651
},
{
"epoch": 0.8902543096091483,
"grad_norm": 1.2263665199279785,
"learning_rate": 3.372178440192588e-05,
"loss": 2.9312,
"step": 652
},
{
"epoch": 0.8916197303294078,
"grad_norm": 1.2436758279800415,
"learning_rate": 3.3701652316305784e-05,
"loss": 2.7438,
"step": 653
},
{
"epoch": 0.8929851510496671,
"grad_norm": 1.2518842220306396,
"learning_rate": 3.3681494846388365e-05,
"loss": 2.7772,
"step": 654
},
{
"epoch": 0.8943505717699266,
"grad_norm": 1.3337477445602417,
"learning_rate": 3.366131203590106e-05,
"loss": 2.7328,
"step": 655
},
{
"epoch": 0.8957159924901861,
"grad_norm": 1.3460004329681396,
"learning_rate": 3.364110392862625e-05,
"loss": 2.8666,
"step": 656
},
{
"epoch": 0.8970814132104454,
"grad_norm": 1.2906866073608398,
"learning_rate": 3.362087056840123e-05,
"loss": 2.7562,
"step": 657
},
{
"epoch": 0.8984468339307049,
"grad_norm": 1.26626455783844,
"learning_rate": 3.360061199911807e-05,
"loss": 2.8764,
"step": 658
},
{
"epoch": 0.8998122546509644,
"grad_norm": 1.2349644899368286,
"learning_rate": 3.358032826472349e-05,
"loss": 2.7971,
"step": 659
},
{
"epoch": 0.9011776753712237,
"grad_norm": 1.334713339805603,
"learning_rate": 3.3560019409218845e-05,
"loss": 2.9036,
"step": 660
},
{
"epoch": 0.9025430960914832,
"grad_norm": 1.265499472618103,
"learning_rate": 3.3539685476659964e-05,
"loss": 2.7105,
"step": 661
},
{
"epoch": 0.9039085168117427,
"grad_norm": 1.2752310037612915,
"learning_rate": 3.351932651115708e-05,
"loss": 2.7944,
"step": 662
},
{
"epoch": 0.905273937532002,
"grad_norm": 1.230279803276062,
"learning_rate": 3.3498942556874724e-05,
"loss": 2.836,
"step": 663
},
{
"epoch": 0.9066393582522615,
"grad_norm": 1.2341333627700806,
"learning_rate": 3.3478533658031645e-05,
"loss": 2.5839,
"step": 664
},
{
"epoch": 0.9080047789725209,
"grad_norm": 1.2281657457351685,
"learning_rate": 3.345809985890069e-05,
"loss": 2.7524,
"step": 665
},
{
"epoch": 0.9093701996927803,
"grad_norm": 1.2955389022827148,
"learning_rate": 3.343764120380873e-05,
"loss": 2.8666,
"step": 666
},
{
"epoch": 0.9107356204130398,
"grad_norm": 1.2975099086761475,
"learning_rate": 3.341715773713657e-05,
"loss": 2.8252,
"step": 667
},
{
"epoch": 0.9121010411332992,
"grad_norm": 1.2703649997711182,
"learning_rate": 3.339664950331881e-05,
"loss": 2.8396,
"step": 668
},
{
"epoch": 0.9134664618535586,
"grad_norm": 1.265570878982544,
"learning_rate": 3.3376116546843784e-05,
"loss": 2.7271,
"step": 669
},
{
"epoch": 0.9148318825738181,
"grad_norm": 1.3075271844863892,
"learning_rate": 3.335555891225349e-05,
"loss": 2.9552,
"step": 670
},
{
"epoch": 0.9161973032940774,
"grad_norm": 1.3111639022827148,
"learning_rate": 3.333497664414341e-05,
"loss": 2.806,
"step": 671
},
{
"epoch": 0.9175627240143369,
"grad_norm": 1.2256838083267212,
"learning_rate": 3.33143697871625e-05,
"loss": 2.6456,
"step": 672
},
{
"epoch": 0.9189281447345964,
"grad_norm": 1.2534023523330688,
"learning_rate": 3.3293738386013046e-05,
"loss": 2.5962,
"step": 673
},
{
"epoch": 0.9202935654548557,
"grad_norm": 1.3422966003417969,
"learning_rate": 3.327308248545057e-05,
"loss": 2.9627,
"step": 674
},
{
"epoch": 0.9216589861751152,
"grad_norm": 1.2841782569885254,
"learning_rate": 3.325240213028374e-05,
"loss": 2.9914,
"step": 675
},
{
"epoch": 0.9230244068953747,
"grad_norm": 1.2441655397415161,
"learning_rate": 3.323169736537429e-05,
"loss": 2.776,
"step": 676
},
{
"epoch": 0.924389827615634,
"grad_norm": 1.3254841566085815,
"learning_rate": 3.321096823563689e-05,
"loss": 3.053,
"step": 677
},
{
"epoch": 0.9257552483358935,
"grad_norm": 1.311414122581482,
"learning_rate": 3.3190214786039065e-05,
"loss": 2.9002,
"step": 678
},
{
"epoch": 0.927120669056153,
"grad_norm": 1.2594579458236694,
"learning_rate": 3.316943706160111e-05,
"loss": 2.7641,
"step": 679
},
{
"epoch": 0.9284860897764123,
"grad_norm": 1.251226544380188,
"learning_rate": 3.314863510739597e-05,
"loss": 2.776,
"step": 680
},
{
"epoch": 0.9298515104966718,
"grad_norm": 1.2537087202072144,
"learning_rate": 3.312780896854915e-05,
"loss": 2.7878,
"step": 681
},
{
"epoch": 0.9312169312169312,
"grad_norm": 1.243266224861145,
"learning_rate": 3.310695869023861e-05,
"loss": 2.8274,
"step": 682
},
{
"epoch": 0.9325823519371906,
"grad_norm": 1.206876516342163,
"learning_rate": 3.308608431769469e-05,
"loss": 2.6728,
"step": 683
},
{
"epoch": 0.9339477726574501,
"grad_norm": 1.2700166702270508,
"learning_rate": 3.30651858962e-05,
"loss": 2.7512,
"step": 684
},
{
"epoch": 0.9353131933777095,
"grad_norm": 1.2888001203536987,
"learning_rate": 3.304426347108932e-05,
"loss": 3.0157,
"step": 685
},
{
"epoch": 0.9366786140979689,
"grad_norm": 1.2521600723266602,
"learning_rate": 3.302331708774947e-05,
"loss": 2.6743,
"step": 686
},
{
"epoch": 0.9380440348182284,
"grad_norm": 1.2771605253219604,
"learning_rate": 3.300234679161928e-05,
"loss": 2.9194,
"step": 687
},
{
"epoch": 0.9394094555384878,
"grad_norm": 1.2342579364776611,
"learning_rate": 3.298135262818944e-05,
"loss": 2.7792,
"step": 688
},
{
"epoch": 0.9407748762587472,
"grad_norm": 1.316307544708252,
"learning_rate": 3.296033464300241e-05,
"loss": 2.8363,
"step": 689
},
{
"epoch": 0.9421402969790067,
"grad_norm": 1.2764463424682617,
"learning_rate": 3.293929288165234e-05,
"loss": 2.8647,
"step": 690
},
{
"epoch": 0.943505717699266,
"grad_norm": 1.2586461305618286,
"learning_rate": 3.291822738978494e-05,
"loss": 2.882,
"step": 691
},
{
"epoch": 0.9448711384195255,
"grad_norm": 1.319823980331421,
"learning_rate": 3.289713821309741e-05,
"loss": 2.8941,
"step": 692
},
{
"epoch": 0.946236559139785,
"grad_norm": 1.2705409526824951,
"learning_rate": 3.287602539733832e-05,
"loss": 2.849,
"step": 693
},
{
"epoch": 0.9476019798600444,
"grad_norm": 1.346256136894226,
"learning_rate": 3.2854888988307536e-05,
"loss": 2.9216,
"step": 694
},
{
"epoch": 0.9489674005803038,
"grad_norm": 1.302469253540039,
"learning_rate": 3.2833729031856086e-05,
"loss": 2.824,
"step": 695
},
{
"epoch": 0.9503328213005633,
"grad_norm": 1.229063868522644,
"learning_rate": 3.281254557388609e-05,
"loss": 2.7571,
"step": 696
},
{
"epoch": 0.9516982420208226,
"grad_norm": 1.3340593576431274,
"learning_rate": 3.279133866035065e-05,
"loss": 2.9479,
"step": 697
},
{
"epoch": 0.9530636627410821,
"grad_norm": 1.2611497640609741,
"learning_rate": 3.277010833725375e-05,
"loss": 2.8909,
"step": 698
},
{
"epoch": 0.9544290834613415,
"grad_norm": 1.2591370344161987,
"learning_rate": 3.2748854650650156e-05,
"loss": 2.8929,
"step": 699
},
{
"epoch": 0.955794504181601,
"grad_norm": 1.3075501918792725,
"learning_rate": 3.27275776466453e-05,
"loss": 2.8476,
"step": 700
},
{
"epoch": 0.9571599249018604,
"grad_norm": 1.294305682182312,
"learning_rate": 3.270627737139522e-05,
"loss": 2.8072,
"step": 701
},
{
"epoch": 0.9585253456221198,
"grad_norm": 1.2306734323501587,
"learning_rate": 3.268495387110644e-05,
"loss": 2.8255,
"step": 702
},
{
"epoch": 0.9598907663423792,
"grad_norm": 1.19227933883667,
"learning_rate": 3.266360719203582e-05,
"loss": 2.7115,
"step": 703
},
{
"epoch": 0.9612561870626387,
"grad_norm": 1.3305531740188599,
"learning_rate": 3.264223738049056e-05,
"loss": 2.934,
"step": 704
},
{
"epoch": 0.9626216077828981,
"grad_norm": 1.2004867792129517,
"learning_rate": 3.262084448282801e-05,
"loss": 2.6128,
"step": 705
},
{
"epoch": 0.9639870285031575,
"grad_norm": 1.2114498615264893,
"learning_rate": 3.259942854545559e-05,
"loss": 2.8224,
"step": 706
},
{
"epoch": 0.965352449223417,
"grad_norm": 1.3151301145553589,
"learning_rate": 3.257798961483074e-05,
"loss": 2.9076,
"step": 707
},
{
"epoch": 0.9667178699436764,
"grad_norm": 1.2859467267990112,
"learning_rate": 3.2556527737460744e-05,
"loss": 2.9579,
"step": 708
},
{
"epoch": 0.9680832906639358,
"grad_norm": 1.1956194639205933,
"learning_rate": 3.2535042959902663e-05,
"loss": 2.6498,
"step": 709
},
{
"epoch": 0.9694487113841953,
"grad_norm": 1.255617380142212,
"learning_rate": 3.2513535328763264e-05,
"loss": 2.7343,
"step": 710
},
{
"epoch": 0.9708141321044547,
"grad_norm": 1.3147718906402588,
"learning_rate": 3.249200489069886e-05,
"loss": 2.8163,
"step": 711
},
{
"epoch": 0.9721795528247141,
"grad_norm": 1.2712767124176025,
"learning_rate": 3.247045169241526e-05,
"loss": 2.8664,
"step": 712
},
{
"epoch": 0.9735449735449735,
"grad_norm": 1.2978441715240479,
"learning_rate": 3.2448875780667637e-05,
"loss": 2.8297,
"step": 713
},
{
"epoch": 0.974910394265233,
"grad_norm": 1.3107860088348389,
"learning_rate": 3.242727720226044e-05,
"loss": 2.8116,
"step": 714
},
{
"epoch": 0.9762758149854924,
"grad_norm": 1.313315749168396,
"learning_rate": 3.240565600404727e-05,
"loss": 2.9201,
"step": 715
},
{
"epoch": 0.9776412357057518,
"grad_norm": 1.2412936687469482,
"learning_rate": 3.2384012232930835e-05,
"loss": 2.6672,
"step": 716
},
{
"epoch": 0.9790066564260113,
"grad_norm": 1.2813894748687744,
"learning_rate": 3.236234593586278e-05,
"loss": 2.802,
"step": 717
},
{
"epoch": 0.9803720771462707,
"grad_norm": 1.352304458618164,
"learning_rate": 3.234065715984361e-05,
"loss": 2.9553,
"step": 718
},
{
"epoch": 0.9817374978665301,
"grad_norm": 1.2740548849105835,
"learning_rate": 3.2318945951922626e-05,
"loss": 2.8205,
"step": 719
},
{
"epoch": 0.9831029185867896,
"grad_norm": 1.2704315185546875,
"learning_rate": 3.2297212359197765e-05,
"loss": 2.8346,
"step": 720
},
{
"epoch": 0.984468339307049,
"grad_norm": 1.242711067199707,
"learning_rate": 3.227545642881553e-05,
"loss": 2.8913,
"step": 721
},
{
"epoch": 0.9858337600273084,
"grad_norm": 1.3351891040802002,
"learning_rate": 3.2253678207970874e-05,
"loss": 2.9166,
"step": 722
},
{
"epoch": 0.9871991807475679,
"grad_norm": 1.250417947769165,
"learning_rate": 3.223187774390712e-05,
"loss": 2.7145,
"step": 723
},
{
"epoch": 0.9885646014678273,
"grad_norm": 1.2289533615112305,
"learning_rate": 3.221005508391583e-05,
"loss": 2.8529,
"step": 724
},
{
"epoch": 0.9899300221880867,
"grad_norm": 1.2092610597610474,
"learning_rate": 3.218821027533671e-05,
"loss": 2.7472,
"step": 725
},
{
"epoch": 0.9912954429083461,
"grad_norm": 1.2371174097061157,
"learning_rate": 3.2166343365557534e-05,
"loss": 2.7295,
"step": 726
},
{
"epoch": 0.9926608636286056,
"grad_norm": 1.273874044418335,
"learning_rate": 3.214445440201401e-05,
"loss": 2.892,
"step": 727
},
{
"epoch": 0.994026284348865,
"grad_norm": 1.3234708309173584,
"learning_rate": 3.2122543432189685e-05,
"loss": 2.9901,
"step": 728
},
{
"epoch": 0.9953917050691244,
"grad_norm": 1.2832272052764893,
"learning_rate": 3.210061050361583e-05,
"loss": 2.6648,
"step": 729
},
{
"epoch": 0.9967571257893838,
"grad_norm": 1.3132859468460083,
"learning_rate": 3.207865566387138e-05,
"loss": 2.8767,
"step": 730
},
{
"epoch": 0.9981225465096433,
"grad_norm": 1.2053296566009521,
"learning_rate": 3.2056678960582776e-05,
"loss": 2.777,
"step": 731
},
{
"epoch": 0.9994879672299027,
"grad_norm": 1.1889965534210205,
"learning_rate": 3.203468044142391e-05,
"loss": 2.7235,
"step": 732
},
{
"epoch": 1.0,
"grad_norm": 1.1889965534210205,
"learning_rate": 3.2012660154115974e-05,
"loss": 1.052,
"step": 733
},
{
"epoch": 1.0013654207202594,
"grad_norm": 1.3007735013961792,
"learning_rate": 3.19906181464274e-05,
"loss": 2.366,
"step": 734
},
{
"epoch": 1.002730841440519,
"grad_norm": 1.3528233766555786,
"learning_rate": 3.1968554466173736e-05,
"loss": 2.3713,
"step": 735
},
{
"epoch": 1.0040962621607783,
"grad_norm": 1.1850340366363525,
"learning_rate": 3.1946469161217534e-05,
"loss": 2.3266,
"step": 736
},
{
"epoch": 1.0054616828810377,
"grad_norm": 1.2342414855957031,
"learning_rate": 3.192436227946828e-05,
"loss": 2.4724,
"step": 737
},
{
"epoch": 1.0068271036012972,
"grad_norm": 1.2932716608047485,
"learning_rate": 3.190223386888223e-05,
"loss": 2.3667,
"step": 738
},
{
"epoch": 1.0081925243215566,
"grad_norm": 1.24687922000885,
"learning_rate": 3.188008397746236e-05,
"loss": 2.3615,
"step": 739
},
{
"epoch": 1.009557945041816,
"grad_norm": 1.4638174772262573,
"learning_rate": 3.185791265325826e-05,
"loss": 2.3729,
"step": 740
},
{
"epoch": 1.0109233657620755,
"grad_norm": 1.484650731086731,
"learning_rate": 3.1835719944365994e-05,
"loss": 2.4536,
"step": 741
},
{
"epoch": 1.012288786482335,
"grad_norm": 1.346647024154663,
"learning_rate": 3.1813505898928017e-05,
"loss": 2.3052,
"step": 742
},
{
"epoch": 1.0136542072025942,
"grad_norm": 1.3402996063232422,
"learning_rate": 3.179127056513309e-05,
"loss": 2.5899,
"step": 743
},
{
"epoch": 1.0150196279228538,
"grad_norm": 1.3082900047302246,
"learning_rate": 3.176901399121611e-05,
"loss": 2.4095,
"step": 744
},
{
"epoch": 1.0163850486431132,
"grad_norm": 1.238875150680542,
"learning_rate": 3.1746736225458104e-05,
"loss": 2.2713,
"step": 745
},
{
"epoch": 1.0177504693633725,
"grad_norm": 1.1435068845748901,
"learning_rate": 3.172443731618602e-05,
"loss": 2.3114,
"step": 746
},
{
"epoch": 1.0191158900836321,
"grad_norm": 1.2557450532913208,
"learning_rate": 3.1702117311772725e-05,
"loss": 2.4071,
"step": 747
},
{
"epoch": 1.0204813108038915,
"grad_norm": 1.314193844795227,
"learning_rate": 3.16797762606368e-05,
"loss": 2.4469,
"step": 748
},
{
"epoch": 1.0218467315241508,
"grad_norm": 1.277459740638733,
"learning_rate": 3.165741421124251e-05,
"loss": 2.483,
"step": 749
},
{
"epoch": 1.0232121522444102,
"grad_norm": 1.238173484802246,
"learning_rate": 3.163503121209966e-05,
"loss": 2.4167,
"step": 750
},
{
"epoch": 1.0245775729646698,
"grad_norm": 1.2115769386291504,
"learning_rate": 3.161262731176351e-05,
"loss": 2.4411,
"step": 751
},
{
"epoch": 1.0259429936849291,
"grad_norm": 1.2024219036102295,
"learning_rate": 3.159020255883464e-05,
"loss": 2.2472,
"step": 752
},
{
"epoch": 1.0273084144051885,
"grad_norm": 1.154970407485962,
"learning_rate": 3.15677570019589e-05,
"loss": 2.3995,
"step": 753
},
{
"epoch": 1.028673835125448,
"grad_norm": 1.3096437454223633,
"learning_rate": 3.154529068982724e-05,
"loss": 2.5174,
"step": 754
},
{
"epoch": 1.0300392558457074,
"grad_norm": 1.2959457635879517,
"learning_rate": 3.152280367117565e-05,
"loss": 2.4114,
"step": 755
},
{
"epoch": 1.0314046765659668,
"grad_norm": 1.3084999322891235,
"learning_rate": 3.150029599478502e-05,
"loss": 2.3414,
"step": 756
},
{
"epoch": 1.0327700972862264,
"grad_norm": 1.216001033782959,
"learning_rate": 3.147776770948108e-05,
"loss": 2.4283,
"step": 757
},
{
"epoch": 1.0341355180064857,
"grad_norm": 1.2088611125946045,
"learning_rate": 3.145521886413425e-05,
"loss": 2.4019,
"step": 758
},
{
"epoch": 1.035500938726745,
"grad_norm": 1.2276612520217896,
"learning_rate": 3.143264950765954e-05,
"loss": 2.4979,
"step": 759
},
{
"epoch": 1.0368663594470047,
"grad_norm": 1.186261773109436,
"learning_rate": 3.1410059689016485e-05,
"loss": 2.3569,
"step": 760
},
{
"epoch": 1.038231780167264,
"grad_norm": 1.2320445775985718,
"learning_rate": 3.138744945720899e-05,
"loss": 2.4025,
"step": 761
},
{
"epoch": 1.0395972008875234,
"grad_norm": 1.2397140264511108,
"learning_rate": 3.136481886128523e-05,
"loss": 2.4283,
"step": 762
},
{
"epoch": 1.040962621607783,
"grad_norm": 1.2361187934875488,
"learning_rate": 3.134216795033758e-05,
"loss": 2.3359,
"step": 763
},
{
"epoch": 1.0423280423280423,
"grad_norm": 1.1806995868682861,
"learning_rate": 3.131949677350248e-05,
"loss": 2.2691,
"step": 764
},
{
"epoch": 1.0436934630483017,
"grad_norm": 1.2157903909683228,
"learning_rate": 3.12968053799603e-05,
"loss": 2.4098,
"step": 765
},
{
"epoch": 1.0450588837685613,
"grad_norm": 1.233747124671936,
"learning_rate": 3.127409381893531e-05,
"loss": 2.3853,
"step": 766
},
{
"epoch": 1.0464243044888206,
"grad_norm": 1.181455373764038,
"learning_rate": 3.125136213969552e-05,
"loss": 2.5347,
"step": 767
},
{
"epoch": 1.04778972520908,
"grad_norm": 1.2235243320465088,
"learning_rate": 3.1228610391552556e-05,
"loss": 2.4146,
"step": 768
},
{
"epoch": 1.0491551459293396,
"grad_norm": 1.2277873754501343,
"learning_rate": 3.12058386238616e-05,
"loss": 2.2987,
"step": 769
},
{
"epoch": 1.050520566649599,
"grad_norm": 1.2242408990859985,
"learning_rate": 3.1183046886021266e-05,
"loss": 2.3255,
"step": 770
},
{
"epoch": 1.0518859873698583,
"grad_norm": 1.2719402313232422,
"learning_rate": 3.116023522747348e-05,
"loss": 2.378,
"step": 771
},
{
"epoch": 1.0532514080901179,
"grad_norm": 1.2798495292663574,
"learning_rate": 3.1137403697703376e-05,
"loss": 2.2835,
"step": 772
},
{
"epoch": 1.0546168288103772,
"grad_norm": 1.2535274028778076,
"learning_rate": 3.111455234623921e-05,
"loss": 2.2596,
"step": 773
},
{
"epoch": 1.0559822495306366,
"grad_norm": 1.2075456380844116,
"learning_rate": 3.109168122265222e-05,
"loss": 2.3479,
"step": 774
},
{
"epoch": 1.0573476702508962,
"grad_norm": 1.2067813873291016,
"learning_rate": 3.106879037655656e-05,
"loss": 2.4002,
"step": 775
},
{
"epoch": 1.0587130909711555,
"grad_norm": 1.1607502698898315,
"learning_rate": 3.1045879857609146e-05,
"loss": 2.2174,
"step": 776
},
{
"epoch": 1.0600785116914149,
"grad_norm": 1.2418051958084106,
"learning_rate": 3.102294971550957e-05,
"loss": 2.4897,
"step": 777
},
{
"epoch": 1.0614439324116745,
"grad_norm": 1.3349454402923584,
"learning_rate": 3.1e-05,
"loss": 2.4338,
"step": 778
},
{
"epoch": 1.0628093531319338,
"grad_norm": 1.24398672580719,
"learning_rate": 3.097703076086508e-05,
"loss": 2.3657,
"step": 779
},
{
"epoch": 1.0641747738521932,
"grad_norm": 1.2262773513793945,
"learning_rate": 3.095404204793176e-05,
"loss": 2.4001,
"step": 780
},
{
"epoch": 1.0655401945724527,
"grad_norm": 1.2271606922149658,
"learning_rate": 3.093103391106929e-05,
"loss": 2.3163,
"step": 781
},
{
"epoch": 1.066905615292712,
"grad_norm": 1.2613786458969116,
"learning_rate": 3.090800640018903e-05,
"loss": 2.4153,
"step": 782
},
{
"epoch": 1.0682710360129715,
"grad_norm": 1.257781982421875,
"learning_rate": 3.088495956524435e-05,
"loss": 2.3656,
"step": 783
},
{
"epoch": 1.0696364567332308,
"grad_norm": 1.2626638412475586,
"learning_rate": 3.086189345623058e-05,
"loss": 2.4738,
"step": 784
},
{
"epoch": 1.0710018774534904,
"grad_norm": 1.2129454612731934,
"learning_rate": 3.083880812318482e-05,
"loss": 2.2247,
"step": 785
},
{
"epoch": 1.0723672981737498,
"grad_norm": 1.2723970413208008,
"learning_rate": 3.0815703616185886e-05,
"loss": 2.3489,
"step": 786
},
{
"epoch": 1.0737327188940091,
"grad_norm": 1.2267957925796509,
"learning_rate": 3.079257998535423e-05,
"loss": 2.3608,
"step": 787
},
{
"epoch": 1.0750981396142687,
"grad_norm": 1.266006350517273,
"learning_rate": 3.076943728085171e-05,
"loss": 2.5007,
"step": 788
},
{
"epoch": 1.076463560334528,
"grad_norm": 1.272132158279419,
"learning_rate": 3.074627555288163e-05,
"loss": 2.3828,
"step": 789
},
{
"epoch": 1.0778289810547874,
"grad_norm": 1.2638702392578125,
"learning_rate": 3.072309485168853e-05,
"loss": 2.4909,
"step": 790
},
{
"epoch": 1.079194401775047,
"grad_norm": 1.2475417852401733,
"learning_rate": 3.0699895227558105e-05,
"loss": 2.2922,
"step": 791
},
{
"epoch": 1.0805598224953064,
"grad_norm": 1.2859658002853394,
"learning_rate": 3.067667673081711e-05,
"loss": 2.3449,
"step": 792
},
{
"epoch": 1.0819252432155657,
"grad_norm": 1.2649673223495483,
"learning_rate": 3.065343941183324e-05,
"loss": 2.4047,
"step": 793
},
{
"epoch": 1.0832906639358253,
"grad_norm": 1.2742799520492554,
"learning_rate": 3.063018332101502e-05,
"loss": 2.3943,
"step": 794
},
{
"epoch": 1.0846560846560847,
"grad_norm": 1.2191989421844482,
"learning_rate": 3.060690850881169e-05,
"loss": 2.3651,
"step": 795
},
{
"epoch": 1.086021505376344,
"grad_norm": 1.2377209663391113,
"learning_rate": 3.058361502571312e-05,
"loss": 2.324,
"step": 796
},
{
"epoch": 1.0873869260966036,
"grad_norm": 1.1660215854644775,
"learning_rate": 3.056030292224964e-05,
"loss": 2.4516,
"step": 797
},
{
"epoch": 1.088752346816863,
"grad_norm": 1.2605676651000977,
"learning_rate": 3.053697224899203e-05,
"loss": 2.3711,
"step": 798
},
{
"epoch": 1.0901177675371223,
"grad_norm": 1.2061911821365356,
"learning_rate": 3.0513623056551323e-05,
"loss": 2.2943,
"step": 799
},
{
"epoch": 1.0914831882573819,
"grad_norm": 1.1918188333511353,
"learning_rate": 3.0490255395578713e-05,
"loss": 2.391,
"step": 800
},
{
"epoch": 1.0928486089776412,
"grad_norm": 1.2193245887756348,
"learning_rate": 3.0466869316765487e-05,
"loss": 2.4471,
"step": 801
},
{
"epoch": 1.0942140296979006,
"grad_norm": 1.217197299003601,
"learning_rate": 3.0443464870842873e-05,
"loss": 2.2482,
"step": 802
},
{
"epoch": 1.0955794504181602,
"grad_norm": 1.2516602277755737,
"learning_rate": 3.0420042108581934e-05,
"loss": 2.3523,
"step": 803
},
{
"epoch": 1.0969448711384195,
"grad_norm": 1.2566345930099487,
"learning_rate": 3.039660108079348e-05,
"loss": 2.3851,
"step": 804
},
{
"epoch": 1.098310291858679,
"grad_norm": 1.2261894941329956,
"learning_rate": 3.037314183832793e-05,
"loss": 2.2808,
"step": 805
},
{
"epoch": 1.0996757125789385,
"grad_norm": 1.2877192497253418,
"learning_rate": 3.0349664432075227e-05,
"loss": 2.4582,
"step": 806
},
{
"epoch": 1.1010411332991978,
"grad_norm": 1.2833895683288574,
"learning_rate": 3.0326168912964725e-05,
"loss": 2.342,
"step": 807
},
{
"epoch": 1.1024065540194572,
"grad_norm": 1.2046228647232056,
"learning_rate": 3.0302655331965044e-05,
"loss": 2.3051,
"step": 808
},
{
"epoch": 1.1037719747397168,
"grad_norm": 1.2776175737380981,
"learning_rate": 3.0279123740084005e-05,
"loss": 2.4439,
"step": 809
},
{
"epoch": 1.1051373954599761,
"grad_norm": 1.2844034433364868,
"learning_rate": 3.02555741883685e-05,
"loss": 2.3217,
"step": 810
},
{
"epoch": 1.1065028161802355,
"grad_norm": 1.2023413181304932,
"learning_rate": 3.0232006727904368e-05,
"loss": 2.3179,
"step": 811
},
{
"epoch": 1.107868236900495,
"grad_norm": 1.259875774383545,
"learning_rate": 3.0208421409816303e-05,
"loss": 2.4463,
"step": 812
},
{
"epoch": 1.1092336576207544,
"grad_norm": 1.2018908262252808,
"learning_rate": 3.018481828526776e-05,
"loss": 2.3046,
"step": 813
},
{
"epoch": 1.1105990783410138,
"grad_norm": 1.184941053390503,
"learning_rate": 3.0161197405460774e-05,
"loss": 2.3271,
"step": 814
},
{
"epoch": 1.1119644990612731,
"grad_norm": 1.250959038734436,
"learning_rate": 3.013755882163595e-05,
"loss": 2.4071,
"step": 815
},
{
"epoch": 1.1133299197815327,
"grad_norm": 1.1493854522705078,
"learning_rate": 3.0113902585072252e-05,
"loss": 2.3381,
"step": 816
},
{
"epoch": 1.114695340501792,
"grad_norm": 1.41165292263031,
"learning_rate": 3.0090228747086968e-05,
"loss": 2.4111,
"step": 817
},
{
"epoch": 1.1160607612220514,
"grad_norm": 1.2390671968460083,
"learning_rate": 3.0066537359035554e-05,
"loss": 2.3519,
"step": 818
},
{
"epoch": 1.117426181942311,
"grad_norm": 1.252549409866333,
"learning_rate": 3.0042828472311548e-05,
"loss": 2.5,
"step": 819
},
{
"epoch": 1.1187916026625704,
"grad_norm": 1.2065517902374268,
"learning_rate": 3.0019102138346427e-05,
"loss": 2.3803,
"step": 820
},
{
"epoch": 1.1201570233828297,
"grad_norm": 1.271870493888855,
"learning_rate": 2.9995358408609547e-05,
"loss": 2.4116,
"step": 821
},
{
"epoch": 1.1215224441030893,
"grad_norm": 1.1853049993515015,
"learning_rate": 2.9971597334607977e-05,
"loss": 2.3875,
"step": 822
},
{
"epoch": 1.1228878648233487,
"grad_norm": 1.24370539188385,
"learning_rate": 2.9947818967886416e-05,
"loss": 2.321,
"step": 823
},
{
"epoch": 1.124253285543608,
"grad_norm": 1.2376235723495483,
"learning_rate": 2.9924023360027078e-05,
"loss": 2.4163,
"step": 824
},
{
"epoch": 1.1256187062638676,
"grad_norm": 1.211007833480835,
"learning_rate": 2.9900210562649587e-05,
"loss": 2.2701,
"step": 825
},
{
"epoch": 1.126984126984127,
"grad_norm": 1.1674623489379883,
"learning_rate": 2.9876380627410832e-05,
"loss": 2.3198,
"step": 826
},
{
"epoch": 1.1283495477043863,
"grad_norm": 1.259652018547058,
"learning_rate": 2.98525336060049e-05,
"loss": 2.4805,
"step": 827
},
{
"epoch": 1.129714968424646,
"grad_norm": 1.2471641302108765,
"learning_rate": 2.9828669550162942e-05,
"loss": 2.3068,
"step": 828
},
{
"epoch": 1.1310803891449053,
"grad_norm": 1.1802136898040771,
"learning_rate": 2.9804788511653054e-05,
"loss": 2.2306,
"step": 829
},
{
"epoch": 1.1324458098651646,
"grad_norm": 1.1555675268173218,
"learning_rate": 2.9780890542280173e-05,
"loss": 2.1659,
"step": 830
},
{
"epoch": 1.1338112305854242,
"grad_norm": 1.232371211051941,
"learning_rate": 2.975697569388596e-05,
"loss": 2.2894,
"step": 831
},
{
"epoch": 1.1351766513056836,
"grad_norm": 1.2489995956420898,
"learning_rate": 2.9733044018348707e-05,
"loss": 2.304,
"step": 832
},
{
"epoch": 1.136542072025943,
"grad_norm": 1.194148063659668,
"learning_rate": 2.9709095567583194e-05,
"loss": 2.3264,
"step": 833
},
{
"epoch": 1.1379074927462025,
"grad_norm": 1.2902276515960693,
"learning_rate": 2.96851303935406e-05,
"loss": 2.4445,
"step": 834
},
{
"epoch": 1.1392729134664619,
"grad_norm": 1.2605379819869995,
"learning_rate": 2.966114854820837e-05,
"loss": 2.4595,
"step": 835
},
{
"epoch": 1.1406383341867212,
"grad_norm": 1.3062224388122559,
"learning_rate": 2.9637150083610136e-05,
"loss": 2.3895,
"step": 836
},
{
"epoch": 1.1420037549069808,
"grad_norm": 1.2228200435638428,
"learning_rate": 2.961313505180556e-05,
"loss": 2.431,
"step": 837
},
{
"epoch": 1.1433691756272402,
"grad_norm": 1.2395297288894653,
"learning_rate": 2.9589103504890254e-05,
"loss": 2.3203,
"step": 838
},
{
"epoch": 1.1447345963474995,
"grad_norm": 1.3226234912872314,
"learning_rate": 2.9565055494995645e-05,
"loss": 2.4152,
"step": 839
},
{
"epoch": 1.1461000170677589,
"grad_norm": 1.3246808052062988,
"learning_rate": 2.9540991074288895e-05,
"loss": 2.4196,
"step": 840
},
{
"epoch": 1.1474654377880185,
"grad_norm": 1.3043259382247925,
"learning_rate": 2.9516910294972745e-05,
"loss": 2.3748,
"step": 841
},
{
"epoch": 1.1488308585082778,
"grad_norm": 1.2049700021743774,
"learning_rate": 2.9492813209285428e-05,
"loss": 2.3288,
"step": 842
},
{
"epoch": 1.1501962792285374,
"grad_norm": 1.2596042156219482,
"learning_rate": 2.9468699869500552e-05,
"loss": 2.3893,
"step": 843
},
{
"epoch": 1.1515616999487968,
"grad_norm": 1.253206491470337,
"learning_rate": 2.9444570327926983e-05,
"loss": 2.502,
"step": 844
},
{
"epoch": 1.1529271206690561,
"grad_norm": 1.2741444110870361,
"learning_rate": 2.9420424636908755e-05,
"loss": 2.5141,
"step": 845
},
{
"epoch": 1.1542925413893155,
"grad_norm": 1.2161749601364136,
"learning_rate": 2.9396262848824887e-05,
"loss": 2.3185,
"step": 846
},
{
"epoch": 1.155657962109575,
"grad_norm": 1.2241460084915161,
"learning_rate": 2.937208501608936e-05,
"loss": 2.2743,
"step": 847
},
{
"epoch": 1.1570233828298344,
"grad_norm": 1.2649509906768799,
"learning_rate": 2.934789119115095e-05,
"loss": 2.3243,
"step": 848
},
{
"epoch": 1.158388803550094,
"grad_norm": 1.3003116846084595,
"learning_rate": 2.9323681426493117e-05,
"loss": 2.3508,
"step": 849
},
{
"epoch": 1.1597542242703534,
"grad_norm": 1.245038390159607,
"learning_rate": 2.9299455774633894e-05,
"loss": 2.2541,
"step": 850
},
{
"epoch": 1.1611196449906127,
"grad_norm": 1.2251543998718262,
"learning_rate": 2.9275214288125794e-05,
"loss": 2.4737,
"step": 851
},
{
"epoch": 1.162485065710872,
"grad_norm": 1.3493139743804932,
"learning_rate": 2.9250957019555674e-05,
"loss": 2.3517,
"step": 852
},
{
"epoch": 1.1638504864311316,
"grad_norm": 1.3227975368499756,
"learning_rate": 2.9226684021544627e-05,
"loss": 2.4873,
"step": 853
},
{
"epoch": 1.165215907151391,
"grad_norm": 1.3293849229812622,
"learning_rate": 2.9202395346747865e-05,
"loss": 2.353,
"step": 854
},
{
"epoch": 1.1665813278716504,
"grad_norm": 1.3362562656402588,
"learning_rate": 2.9178091047854603e-05,
"loss": 2.3884,
"step": 855
},
{
"epoch": 1.16794674859191,
"grad_norm": 1.2474746704101562,
"learning_rate": 2.915377117758796e-05,
"loss": 2.3201,
"step": 856
},
{
"epoch": 1.1693121693121693,
"grad_norm": 1.2002493143081665,
"learning_rate": 2.9129435788704827e-05,
"loss": 2.4961,
"step": 857
},
{
"epoch": 1.1706775900324287,
"grad_norm": 1.2316269874572754,
"learning_rate": 2.9105084933995763e-05,
"loss": 2.3991,
"step": 858
},
{
"epoch": 1.1720430107526882,
"grad_norm": 1.2648407220840454,
"learning_rate": 2.908071866628487e-05,
"loss": 2.3619,
"step": 859
},
{
"epoch": 1.1734084314729476,
"grad_norm": 1.21544349193573,
"learning_rate": 2.9056337038429708e-05,
"loss": 2.4144,
"step": 860
},
{
"epoch": 1.174773852193207,
"grad_norm": 1.2424837350845337,
"learning_rate": 2.9031940103321115e-05,
"loss": 2.4836,
"step": 861
},
{
"epoch": 1.1761392729134665,
"grad_norm": 1.1828193664550781,
"learning_rate": 2.9007527913883176e-05,
"loss": 2.3297,
"step": 862
},
{
"epoch": 1.177504693633726,
"grad_norm": 1.2255760431289673,
"learning_rate": 2.8983100523073044e-05,
"loss": 2.3258,
"step": 863
},
{
"epoch": 1.1788701143539853,
"grad_norm": 1.2073782682418823,
"learning_rate": 2.895865798388086e-05,
"loss": 2.3375,
"step": 864
},
{
"epoch": 1.1802355350742448,
"grad_norm": 1.2225022315979004,
"learning_rate": 2.8934200349329623e-05,
"loss": 2.361,
"step": 865
},
{
"epoch": 1.1816009557945042,
"grad_norm": 1.2762993574142456,
"learning_rate": 2.8909727672475072e-05,
"loss": 2.338,
"step": 866
},
{
"epoch": 1.1829663765147636,
"grad_norm": 1.2287622690200806,
"learning_rate": 2.888524000640558e-05,
"loss": 2.3487,
"step": 867
},
{
"epoch": 1.1843317972350231,
"grad_norm": 1.1822468042373657,
"learning_rate": 2.886073740424205e-05,
"loss": 2.4022,
"step": 868
},
{
"epoch": 1.1856972179552825,
"grad_norm": 1.238525390625,
"learning_rate": 2.8836219919137764e-05,
"loss": 2.3097,
"step": 869
},
{
"epoch": 1.1870626386755418,
"grad_norm": 1.2410509586334229,
"learning_rate": 2.8811687604278297e-05,
"loss": 2.3714,
"step": 870
},
{
"epoch": 1.1884280593958014,
"grad_norm": 1.2486552000045776,
"learning_rate": 2.8787140512881398e-05,
"loss": 2.2827,
"step": 871
},
{
"epoch": 1.1897934801160608,
"grad_norm": 1.2500061988830566,
"learning_rate": 2.8762578698196876e-05,
"loss": 2.3645,
"step": 872
},
{
"epoch": 1.1911589008363201,
"grad_norm": 1.1996694803237915,
"learning_rate": 2.8738002213506462e-05,
"loss": 2.3794,
"step": 873
},
{
"epoch": 1.1925243215565797,
"grad_norm": 1.21322500705719,
"learning_rate": 2.8713411112123722e-05,
"loss": 2.3274,
"step": 874
},
{
"epoch": 1.193889742276839,
"grad_norm": 1.2636908292770386,
"learning_rate": 2.868880544739394e-05,
"loss": 2.3805,
"step": 875
},
{
"epoch": 1.1952551629970984,
"grad_norm": 1.1915347576141357,
"learning_rate": 2.866418527269396e-05,
"loss": 2.2621,
"step": 876
},
{
"epoch": 1.1966205837173578,
"grad_norm": 1.2206077575683594,
"learning_rate": 2.8639550641432145e-05,
"loss": 2.3926,
"step": 877
},
{
"epoch": 1.1979860044376174,
"grad_norm": 1.2606523036956787,
"learning_rate": 2.861490160704818e-05,
"loss": 2.3456,
"step": 878
},
{
"epoch": 1.1993514251578767,
"grad_norm": 1.1896344423294067,
"learning_rate": 2.859023822301302e-05,
"loss": 2.2806,
"step": 879
},
{
"epoch": 1.2007168458781363,
"grad_norm": 1.2329208850860596,
"learning_rate": 2.856556054282873e-05,
"loss": 2.4176,
"step": 880
},
{
"epoch": 1.2020822665983957,
"grad_norm": 1.231826901435852,
"learning_rate": 2.854086862002841e-05,
"loss": 2.3699,
"step": 881
},
{
"epoch": 1.203447687318655,
"grad_norm": 1.2495681047439575,
"learning_rate": 2.8516162508176034e-05,
"loss": 2.463,
"step": 882
},
{
"epoch": 1.2048131080389144,
"grad_norm": 1.272783637046814,
"learning_rate": 2.8491442260866372e-05,
"loss": 2.3406,
"step": 883
},
{
"epoch": 1.206178528759174,
"grad_norm": 1.235047459602356,
"learning_rate": 2.8466707931724836e-05,
"loss": 2.2468,
"step": 884
},
{
"epoch": 1.2075439494794333,
"grad_norm": 1.2175613641738892,
"learning_rate": 2.8441959574407415e-05,
"loss": 2.2956,
"step": 885
},
{
"epoch": 1.2089093701996927,
"grad_norm": 1.262965202331543,
"learning_rate": 2.841719724260052e-05,
"loss": 2.5032,
"step": 886
},
{
"epoch": 1.2102747909199523,
"grad_norm": 1.2792342901229858,
"learning_rate": 2.8392420990020864e-05,
"loss": 2.4555,
"step": 887
},
{
"epoch": 1.2116402116402116,
"grad_norm": 1.278132677078247,
"learning_rate": 2.8367630870415364e-05,
"loss": 2.3753,
"step": 888
},
{
"epoch": 1.213005632360471,
"grad_norm": 1.2498081922531128,
"learning_rate": 2.834282693756102e-05,
"loss": 2.4198,
"step": 889
},
{
"epoch": 1.2143710530807306,
"grad_norm": 1.2764676809310913,
"learning_rate": 2.8318009245264814e-05,
"loss": 2.4232,
"step": 890
},
{
"epoch": 1.21573647380099,
"grad_norm": 1.2853691577911377,
"learning_rate": 2.829317784736353e-05,
"loss": 2.3146,
"step": 891
},
{
"epoch": 1.2171018945212493,
"grad_norm": 1.254612922668457,
"learning_rate": 2.826833279772375e-05,
"loss": 2.4378,
"step": 892
},
{
"epoch": 1.2184673152415089,
"grad_norm": 1.2399685382843018,
"learning_rate": 2.8243474150241602e-05,
"loss": 2.4583,
"step": 893
},
{
"epoch": 1.2198327359617682,
"grad_norm": 1.2260444164276123,
"learning_rate": 2.821860195884276e-05,
"loss": 2.3931,
"step": 894
},
{
"epoch": 1.2211981566820276,
"grad_norm": 1.2050138711929321,
"learning_rate": 2.819371627748226e-05,
"loss": 2.2196,
"step": 895
},
{
"epoch": 1.2225635774022872,
"grad_norm": 1.201066255569458,
"learning_rate": 2.816881716014441e-05,
"loss": 2.4628,
"step": 896
},
{
"epoch": 1.2239289981225465,
"grad_norm": 1.2208431959152222,
"learning_rate": 2.814390466084264e-05,
"loss": 2.3979,
"step": 897
},
{
"epoch": 1.2252944188428059,
"grad_norm": 1.2809184789657593,
"learning_rate": 2.8118978833619458e-05,
"loss": 2.4864,
"step": 898
},
{
"epoch": 1.2266598395630655,
"grad_norm": 1.2579782009124756,
"learning_rate": 2.8094039732546225e-05,
"loss": 2.3905,
"step": 899
},
{
"epoch": 1.2280252602833248,
"grad_norm": 1.2523733377456665,
"learning_rate": 2.806908741172314e-05,
"loss": 2.2768,
"step": 900
},
{
"epoch": 1.2293906810035842,
"grad_norm": 1.2336443662643433,
"learning_rate": 2.8044121925279068e-05,
"loss": 2.3323,
"step": 901
},
{
"epoch": 1.2307561017238438,
"grad_norm": 1.2365236282348633,
"learning_rate": 2.801914332737143e-05,
"loss": 2.3098,
"step": 902
},
{
"epoch": 1.2321215224441031,
"grad_norm": 1.2441930770874023,
"learning_rate": 2.799415167218608e-05,
"loss": 2.2595,
"step": 903
},
{
"epoch": 1.2334869431643625,
"grad_norm": 1.248464584350586,
"learning_rate": 2.7969147013937227e-05,
"loss": 2.3735,
"step": 904
},
{
"epoch": 1.234852363884622,
"grad_norm": 1.2634050846099854,
"learning_rate": 2.794412940686725e-05,
"loss": 2.3045,
"step": 905
},
{
"epoch": 1.2362177846048814,
"grad_norm": 1.2350529432296753,
"learning_rate": 2.7919098905246645e-05,
"loss": 2.4167,
"step": 906
},
{
"epoch": 1.2375832053251408,
"grad_norm": 1.2136973142623901,
"learning_rate": 2.7894055563373883e-05,
"loss": 2.3633,
"step": 907
},
{
"epoch": 1.2389486260454001,
"grad_norm": 1.2299065589904785,
"learning_rate": 2.7868999435575256e-05,
"loss": 2.317,
"step": 908
},
{
"epoch": 1.2403140467656597,
"grad_norm": 1.3067283630371094,
"learning_rate": 2.7843930576204833e-05,
"loss": 2.4601,
"step": 909
},
{
"epoch": 1.241679467485919,
"grad_norm": 1.194559097290039,
"learning_rate": 2.7818849039644272e-05,
"loss": 2.3162,
"step": 910
},
{
"epoch": 1.2430448882061786,
"grad_norm": 1.2037074565887451,
"learning_rate": 2.7793754880302747e-05,
"loss": 2.3135,
"step": 911
},
{
"epoch": 1.244410308926438,
"grad_norm": 1.2231028079986572,
"learning_rate": 2.7768648152616802e-05,
"loss": 2.2711,
"step": 912
},
{
"epoch": 1.2457757296466974,
"grad_norm": 1.1659778356552124,
"learning_rate": 2.7743528911050267e-05,
"loss": 2.3021,
"step": 913
},
{
"epoch": 1.2471411503669567,
"grad_norm": 1.2174005508422852,
"learning_rate": 2.7718397210094092e-05,
"loss": 2.3702,
"step": 914
},
{
"epoch": 1.2485065710872163,
"grad_norm": 1.1640428304672241,
"learning_rate": 2.7693253104266276e-05,
"loss": 2.2262,
"step": 915
},
{
"epoch": 1.2498719918074757,
"grad_norm": 1.233407735824585,
"learning_rate": 2.7668096648111714e-05,
"loss": 2.3001,
"step": 916
},
{
"epoch": 1.2512374125277352,
"grad_norm": 1.2429823875427246,
"learning_rate": 2.764292789620209e-05,
"loss": 2.3332,
"step": 917
},
{
"epoch": 1.2526028332479946,
"grad_norm": 1.2613537311553955,
"learning_rate": 2.761774690313579e-05,
"loss": 2.4765,
"step": 918
},
{
"epoch": 1.253968253968254,
"grad_norm": 1.2685474157333374,
"learning_rate": 2.7592553723537718e-05,
"loss": 2.2863,
"step": 919
},
{
"epoch": 1.2553336746885133,
"grad_norm": 1.2619364261627197,
"learning_rate": 2.7567348412059226e-05,
"loss": 2.2419,
"step": 920
},
{
"epoch": 1.256699095408773,
"grad_norm": 1.228325366973877,
"learning_rate": 2.7542131023378004e-05,
"loss": 2.353,
"step": 921
},
{
"epoch": 1.2580645161290323,
"grad_norm": 1.2420254945755005,
"learning_rate": 2.751690161219791e-05,
"loss": 2.3702,
"step": 922
},
{
"epoch": 1.2594299368492916,
"grad_norm": 1.2838985919952393,
"learning_rate": 2.74916602332489e-05,
"loss": 2.4912,
"step": 923
},
{
"epoch": 1.2607953575695512,
"grad_norm": 1.3119853734970093,
"learning_rate": 2.7466406941286886e-05,
"loss": 2.4636,
"step": 924
},
{
"epoch": 1.2621607782898105,
"grad_norm": 1.279038667678833,
"learning_rate": 2.7441141791093627e-05,
"loss": 2.4114,
"step": 925
},
{
"epoch": 1.26352619901007,
"grad_norm": 1.2429629564285278,
"learning_rate": 2.7415864837476614e-05,
"loss": 2.3813,
"step": 926
},
{
"epoch": 1.2648916197303295,
"grad_norm": 1.2583948373794556,
"learning_rate": 2.7390576135268915e-05,
"loss": 2.4365,
"step": 927
},
{
"epoch": 1.2662570404505888,
"grad_norm": 1.205424427986145,
"learning_rate": 2.7365275739329112e-05,
"loss": 2.3744,
"step": 928
},
{
"epoch": 1.2676224611708482,
"grad_norm": 1.2175387144088745,
"learning_rate": 2.7339963704541143e-05,
"loss": 2.3951,
"step": 929
},
{
"epoch": 1.2689878818911078,
"grad_norm": 1.2389016151428223,
"learning_rate": 2.7314640085814196e-05,
"loss": 2.4391,
"step": 930
},
{
"epoch": 1.2703533026113671,
"grad_norm": 1.2554258108139038,
"learning_rate": 2.7289304938082583e-05,
"loss": 2.2553,
"step": 931
},
{
"epoch": 1.2717187233316265,
"grad_norm": 1.2264779806137085,
"learning_rate": 2.7263958316305627e-05,
"loss": 2.3366,
"step": 932
},
{
"epoch": 1.2730841440518859,
"grad_norm": 1.2106153964996338,
"learning_rate": 2.723860027546756e-05,
"loss": 2.2292,
"step": 933
},
{
"epoch": 1.2744495647721454,
"grad_norm": 1.2736351490020752,
"learning_rate": 2.721323087057735e-05,
"loss": 2.3685,
"step": 934
},
{
"epoch": 1.2758149854924048,
"grad_norm": 1.2360061407089233,
"learning_rate": 2.7187850156668648e-05,
"loss": 2.4408,
"step": 935
},
{
"epoch": 1.2771804062126644,
"grad_norm": 1.2158254384994507,
"learning_rate": 2.716245818879963e-05,
"loss": 2.3291,
"step": 936
},
{
"epoch": 1.2785458269329237,
"grad_norm": 1.2642351388931274,
"learning_rate": 2.713705502205287e-05,
"loss": 2.4047,
"step": 937
},
{
"epoch": 1.279911247653183,
"grad_norm": 1.2295008897781372,
"learning_rate": 2.711164071153525e-05,
"loss": 2.3455,
"step": 938
},
{
"epoch": 1.2812766683734425,
"grad_norm": 1.2446925640106201,
"learning_rate": 2.7086215312377836e-05,
"loss": 2.4245,
"step": 939
},
{
"epoch": 1.282642089093702,
"grad_norm": 1.2333112955093384,
"learning_rate": 2.7060778879735714e-05,
"loss": 2.3489,
"step": 940
},
{
"epoch": 1.2840075098139614,
"grad_norm": 1.22257399559021,
"learning_rate": 2.7035331468787945e-05,
"loss": 2.308,
"step": 941
},
{
"epoch": 1.285372930534221,
"grad_norm": 1.2384806871414185,
"learning_rate": 2.7009873134737375e-05,
"loss": 2.3857,
"step": 942
},
{
"epoch": 1.2867383512544803,
"grad_norm": 1.2525445222854614,
"learning_rate": 2.698440393281056e-05,
"loss": 2.345,
"step": 943
},
{
"epoch": 1.2881037719747397,
"grad_norm": 1.2513006925582886,
"learning_rate": 2.6958923918257634e-05,
"loss": 2.3366,
"step": 944
},
{
"epoch": 1.289469192694999,
"grad_norm": 1.2394795417785645,
"learning_rate": 2.693343314635217e-05,
"loss": 2.3443,
"step": 945
},
{
"epoch": 1.2908346134152586,
"grad_norm": 1.2281118631362915,
"learning_rate": 2.6907931672391106e-05,
"loss": 2.333,
"step": 946
},
{
"epoch": 1.292200034135518,
"grad_norm": 1.2325307130813599,
"learning_rate": 2.688241955169456e-05,
"loss": 2.3287,
"step": 947
},
{
"epoch": 1.2935654548557776,
"grad_norm": 1.2684941291809082,
"learning_rate": 2.6856896839605772e-05,
"loss": 2.5691,
"step": 948
},
{
"epoch": 1.294930875576037,
"grad_norm": 1.279487133026123,
"learning_rate": 2.6831363591490945e-05,
"loss": 2.389,
"step": 949
},
{
"epoch": 1.2962962962962963,
"grad_norm": 1.2265145778656006,
"learning_rate": 2.6805819862739146e-05,
"loss": 2.3358,
"step": 950
},
{
"epoch": 1.2976617170165556,
"grad_norm": 1.2082444429397583,
"learning_rate": 2.6780265708762177e-05,
"loss": 2.2491,
"step": 951
},
{
"epoch": 1.2990271377368152,
"grad_norm": 1.2949508428573608,
"learning_rate": 2.6754701184994445e-05,
"loss": 2.3956,
"step": 952
},
{
"epoch": 1.3003925584570746,
"grad_norm": 1.2472524642944336,
"learning_rate": 2.6729126346892863e-05,
"loss": 2.3988,
"step": 953
},
{
"epoch": 1.301757979177334,
"grad_norm": 1.268323540687561,
"learning_rate": 2.670354124993672e-05,
"loss": 2.4286,
"step": 954
},
{
"epoch": 1.3031233998975935,
"grad_norm": 1.2821838855743408,
"learning_rate": 2.6677945949627547e-05,
"loss": 2.4855,
"step": 955
},
{
"epoch": 1.3044888206178529,
"grad_norm": 1.221441626548767,
"learning_rate": 2.6652340501489026e-05,
"loss": 2.2783,
"step": 956
},
{
"epoch": 1.3058542413381122,
"grad_norm": 1.2497761249542236,
"learning_rate": 2.662672496106683e-05,
"loss": 2.4689,
"step": 957
},
{
"epoch": 1.3072196620583718,
"grad_norm": 1.2634248733520508,
"learning_rate": 2.6601099383928545e-05,
"loss": 2.4282,
"step": 958
},
{
"epoch": 1.3085850827786312,
"grad_norm": 1.2188899517059326,
"learning_rate": 2.657546382566353e-05,
"loss": 2.3681,
"step": 959
},
{
"epoch": 1.3099505034988905,
"grad_norm": 1.3508509397506714,
"learning_rate": 2.6549818341882776e-05,
"loss": 2.264,
"step": 960
},
{
"epoch": 1.31131592421915,
"grad_norm": 1.2509104013442993,
"learning_rate": 2.652416298821883e-05,
"loss": 2.3805,
"step": 961
},
{
"epoch": 1.3126813449394095,
"grad_norm": 1.259556770324707,
"learning_rate": 2.6498497820325625e-05,
"loss": 2.4912,
"step": 962
},
{
"epoch": 1.3140467656596688,
"grad_norm": 1.2539644241333008,
"learning_rate": 2.6472822893878415e-05,
"loss": 2.423,
"step": 963
},
{
"epoch": 1.3154121863799282,
"grad_norm": 1.304323673248291,
"learning_rate": 2.644713826457359e-05,
"loss": 2.3,
"step": 964
},
{
"epoch": 1.3167776071001878,
"grad_norm": 1.2002923488616943,
"learning_rate": 2.6421443988128607e-05,
"loss": 2.4383,
"step": 965
},
{
"epoch": 1.3181430278204471,
"grad_norm": 1.2204906940460205,
"learning_rate": 2.639574012028185e-05,
"loss": 2.278,
"step": 966
},
{
"epoch": 1.3195084485407067,
"grad_norm": 1.2220135927200317,
"learning_rate": 2.6370026716792505e-05,
"loss": 2.3305,
"step": 967
},
{
"epoch": 1.320873869260966,
"grad_norm": 1.223619818687439,
"learning_rate": 2.6344303833440443e-05,
"loss": 2.3942,
"step": 968
},
{
"epoch": 1.3222392899812254,
"grad_norm": 1.2592391967773438,
"learning_rate": 2.63185715260261e-05,
"loss": 2.4512,
"step": 969
},
{
"epoch": 1.3236047107014848,
"grad_norm": 1.225287914276123,
"learning_rate": 2.6292829850370355e-05,
"loss": 2.3417,
"step": 970
},
{
"epoch": 1.3249701314217444,
"grad_norm": 1.219190239906311,
"learning_rate": 2.626707886231443e-05,
"loss": 2.4017,
"step": 971
},
{
"epoch": 1.3263355521420037,
"grad_norm": 1.2635302543640137,
"learning_rate": 2.6241318617719703e-05,
"loss": 2.2509,
"step": 972
},
{
"epoch": 1.3277009728622633,
"grad_norm": 1.2044737339019775,
"learning_rate": 2.621554917246768e-05,
"loss": 2.2805,
"step": 973
},
{
"epoch": 1.3290663935825227,
"grad_norm": 1.238142967224121,
"learning_rate": 2.618977058245979e-05,
"loss": 2.3811,
"step": 974
},
{
"epoch": 1.330431814302782,
"grad_norm": 1.242542028427124,
"learning_rate": 2.616398290361733e-05,
"loss": 2.3598,
"step": 975
},
{
"epoch": 1.3317972350230414,
"grad_norm": 1.340445876121521,
"learning_rate": 2.613818619188129e-05,
"loss": 2.4058,
"step": 976
},
{
"epoch": 1.333162655743301,
"grad_norm": 1.2982323169708252,
"learning_rate": 2.6112380503212255e-05,
"loss": 2.4023,
"step": 977
},
{
"epoch": 1.3345280764635603,
"grad_norm": 1.223367691040039,
"learning_rate": 2.6086565893590302e-05,
"loss": 2.2586,
"step": 978
},
{
"epoch": 1.33589349718382,
"grad_norm": 1.21322762966156,
"learning_rate": 2.606074241901484e-05,
"loss": 2.4034,
"step": 979
},
{
"epoch": 1.3372589179040792,
"grad_norm": 1.2169383764266968,
"learning_rate": 2.6034910135504534e-05,
"loss": 2.3915,
"step": 980
},
{
"epoch": 1.3386243386243386,
"grad_norm": 1.2652515172958374,
"learning_rate": 2.6009069099097125e-05,
"loss": 2.4723,
"step": 981
},
{
"epoch": 1.339989759344598,
"grad_norm": 1.2372570037841797,
"learning_rate": 2.5983219365849367e-05,
"loss": 2.3504,
"step": 982
},
{
"epoch": 1.3413551800648575,
"grad_norm": 1.317415475845337,
"learning_rate": 2.595736099183688e-05,
"loss": 2.3791,
"step": 983
},
{
"epoch": 1.342720600785117,
"grad_norm": 1.2165470123291016,
"learning_rate": 2.5931494033154002e-05,
"loss": 2.3239,
"step": 984
},
{
"epoch": 1.3440860215053765,
"grad_norm": 1.1797916889190674,
"learning_rate": 2.5905618545913728e-05,
"loss": 2.3106,
"step": 985
},
{
"epoch": 1.3454514422256358,
"grad_norm": 1.1931771039962769,
"learning_rate": 2.5879734586247534e-05,
"loss": 2.3399,
"step": 986
},
{
"epoch": 1.3468168629458952,
"grad_norm": 1.2794438600540161,
"learning_rate": 2.5853842210305268e-05,
"loss": 2.5,
"step": 987
},
{
"epoch": 1.3481822836661546,
"grad_norm": 1.2815632820129395,
"learning_rate": 2.582794147425506e-05,
"loss": 2.5131,
"step": 988
},
{
"epoch": 1.3495477043864141,
"grad_norm": 1.1562837362289429,
"learning_rate": 2.5802032434283156e-05,
"loss": 2.1855,
"step": 989
},
{
"epoch": 1.3509131251066735,
"grad_norm": 1.2836556434631348,
"learning_rate": 2.5776115146593827e-05,
"loss": 2.3952,
"step": 990
},
{
"epoch": 1.3522785458269329,
"grad_norm": 1.2334662675857544,
"learning_rate": 2.5750189667409222e-05,
"loss": 2.3789,
"step": 991
},
{
"epoch": 1.3536439665471924,
"grad_norm": 1.230539083480835,
"learning_rate": 2.572425605296928e-05,
"loss": 2.2625,
"step": 992
},
{
"epoch": 1.3550093872674518,
"grad_norm": 1.1765425205230713,
"learning_rate": 2.5698314359531557e-05,
"loss": 2.2633,
"step": 993
},
{
"epoch": 1.3563748079877112,
"grad_norm": 1.2777494192123413,
"learning_rate": 2.567236464337117e-05,
"loss": 2.3347,
"step": 994
},
{
"epoch": 1.3577402287079705,
"grad_norm": 1.2749028205871582,
"learning_rate": 2.5646406960780614e-05,
"loss": 2.4116,
"step": 995
},
{
"epoch": 1.35910564942823,
"grad_norm": 1.2012782096862793,
"learning_rate": 2.5620441368069675e-05,
"loss": 2.3296,
"step": 996
},
{
"epoch": 1.3604710701484894,
"grad_norm": 1.2508784532546997,
"learning_rate": 2.5594467921565308e-05,
"loss": 2.4091,
"step": 997
},
{
"epoch": 1.361836490868749,
"grad_norm": 1.2551524639129639,
"learning_rate": 2.5568486677611487e-05,
"loss": 2.2965,
"step": 998
},
{
"epoch": 1.3632019115890084,
"grad_norm": 1.208871603012085,
"learning_rate": 2.5542497692569103e-05,
"loss": 2.3919,
"step": 999
},
{
"epoch": 1.3645673323092677,
"grad_norm": 1.1751126050949097,
"learning_rate": 2.5516501022815857e-05,
"loss": 2.4404,
"step": 1000
},
{
"epoch": 1.365932753029527,
"grad_norm": 1.2942649126052856,
"learning_rate": 2.5490496724746104e-05,
"loss": 2.419,
"step": 1001
},
{
"epoch": 1.3672981737497867,
"grad_norm": 1.2418574094772339,
"learning_rate": 2.5464484854770755e-05,
"loss": 2.2908,
"step": 1002
},
{
"epoch": 1.368663594470046,
"grad_norm": 1.2185275554656982,
"learning_rate": 2.5438465469317142e-05,
"loss": 2.4586,
"step": 1003
},
{
"epoch": 1.3700290151903056,
"grad_norm": 1.2331212759017944,
"learning_rate": 2.54124386248289e-05,
"loss": 2.4888,
"step": 1004
},
{
"epoch": 1.371394435910565,
"grad_norm": 1.292909860610962,
"learning_rate": 2.538640437776585e-05,
"loss": 2.4315,
"step": 1005
},
{
"epoch": 1.3727598566308243,
"grad_norm": 1.2930008172988892,
"learning_rate": 2.5360362784603866e-05,
"loss": 2.3693,
"step": 1006
},
{
"epoch": 1.3741252773510837,
"grad_norm": 1.2821295261383057,
"learning_rate": 2.533431390183476e-05,
"loss": 2.4084,
"step": 1007
},
{
"epoch": 1.3754906980713433,
"grad_norm": 1.230641484260559,
"learning_rate": 2.530825778596616e-05,
"loss": 2.3927,
"step": 1008
},
{
"epoch": 1.3768561187916026,
"grad_norm": 1.2552179098129272,
"learning_rate": 2.528219449352138e-05,
"loss": 2.4241,
"step": 1009
},
{
"epoch": 1.3782215395118622,
"grad_norm": 1.2691830396652222,
"learning_rate": 2.5256124081039307e-05,
"loss": 2.2644,
"step": 1010
},
{
"epoch": 1.3795869602321216,
"grad_norm": 1.2305405139923096,
"learning_rate": 2.5230046605074262e-05,
"loss": 2.4219,
"step": 1011
},
{
"epoch": 1.380952380952381,
"grad_norm": 1.2326072454452515,
"learning_rate": 2.5203962122195915e-05,
"loss": 2.3853,
"step": 1012
},
{
"epoch": 1.3823178016726403,
"grad_norm": 1.2373192310333252,
"learning_rate": 2.517787068898909e-05,
"loss": 2.281,
"step": 1013
},
{
"epoch": 1.3836832223928999,
"grad_norm": 1.2571709156036377,
"learning_rate": 2.5151772362053742e-05,
"loss": 2.4298,
"step": 1014
},
{
"epoch": 1.3850486431131592,
"grad_norm": 1.2474796772003174,
"learning_rate": 2.512566719800475e-05,
"loss": 2.3231,
"step": 1015
},
{
"epoch": 1.3864140638334188,
"grad_norm": 1.1855806112289429,
"learning_rate": 2.5099555253471824e-05,
"loss": 2.3131,
"step": 1016
},
{
"epoch": 1.3877794845536782,
"grad_norm": 1.2695180177688599,
"learning_rate": 2.5073436585099384e-05,
"loss": 2.4002,
"step": 1017
},
{
"epoch": 1.3891449052739375,
"grad_norm": 1.2916597127914429,
"learning_rate": 2.5047311249546464e-05,
"loss": 2.4284,
"step": 1018
},
{
"epoch": 1.3905103259941969,
"grad_norm": 1.284002423286438,
"learning_rate": 2.502117930348651e-05,
"loss": 2.373,
"step": 1019
},
{
"epoch": 1.3918757467144565,
"grad_norm": 1.2568774223327637,
"learning_rate": 2.499504080360735e-05,
"loss": 2.3322,
"step": 1020
},
{
"epoch": 1.3932411674347158,
"grad_norm": 1.2575657367706299,
"learning_rate": 2.4968895806611014e-05,
"loss": 2.4252,
"step": 1021
},
{
"epoch": 1.3946065881549752,
"grad_norm": 1.224495530128479,
"learning_rate": 2.4942744369213626e-05,
"loss": 2.2994,
"step": 1022
},
{
"epoch": 1.3959720088752348,
"grad_norm": 1.316319465637207,
"learning_rate": 2.4916586548145286e-05,
"loss": 2.2959,
"step": 1023
},
{
"epoch": 1.3973374295954941,
"grad_norm": 1.2435154914855957,
"learning_rate": 2.489042240014993e-05,
"loss": 2.4232,
"step": 1024
},
{
"epoch": 1.3987028503157535,
"grad_norm": 1.3580273389816284,
"learning_rate": 2.4864251981985233e-05,
"loss": 2.3562,
"step": 1025
},
{
"epoch": 1.4000682710360128,
"grad_norm": 1.2626869678497314,
"learning_rate": 2.483807535042246e-05,
"loss": 2.4745,
"step": 1026
},
{
"epoch": 1.4014336917562724,
"grad_norm": 1.2748677730560303,
"learning_rate": 2.4811892562246373e-05,
"loss": 2.3718,
"step": 1027
},
{
"epoch": 1.4027991124765318,
"grad_norm": 1.232909083366394,
"learning_rate": 2.4785703674255062e-05,
"loss": 2.341,
"step": 1028
},
{
"epoch": 1.4041645331967914,
"grad_norm": 1.2865220308303833,
"learning_rate": 2.475950874325988e-05,
"loss": 2.4262,
"step": 1029
},
{
"epoch": 1.4055299539170507,
"grad_norm": 1.1778664588928223,
"learning_rate": 2.4733307826085265e-05,
"loss": 2.3194,
"step": 1030
},
{
"epoch": 1.40689537463731,
"grad_norm": 1.2491105794906616,
"learning_rate": 2.4707100979568646e-05,
"loss": 2.4309,
"step": 1031
},
{
"epoch": 1.4082607953575694,
"grad_norm": 1.2606797218322754,
"learning_rate": 2.468088826056032e-05,
"loss": 2.316,
"step": 1032
},
{
"epoch": 1.409626216077829,
"grad_norm": 1.2384246587753296,
"learning_rate": 2.465466972592333e-05,
"loss": 2.3803,
"step": 1033
},
{
"epoch": 1.4109916367980884,
"grad_norm": 1.2098149061203003,
"learning_rate": 2.4628445432533315e-05,
"loss": 2.2819,
"step": 1034
},
{
"epoch": 1.412357057518348,
"grad_norm": 1.276442289352417,
"learning_rate": 2.4602215437278422e-05,
"loss": 2.4833,
"step": 1035
},
{
"epoch": 1.4137224782386073,
"grad_norm": 1.2431776523590088,
"learning_rate": 2.4575979797059165e-05,
"loss": 2.2797,
"step": 1036
},
{
"epoch": 1.4150878989588667,
"grad_norm": 1.2979681491851807,
"learning_rate": 2.4549738568788293e-05,
"loss": 2.3272,
"step": 1037
},
{
"epoch": 1.416453319679126,
"grad_norm": 1.2811650037765503,
"learning_rate": 2.4523491809390686e-05,
"loss": 2.4231,
"step": 1038
},
{
"epoch": 1.4178187403993856,
"grad_norm": 1.225139856338501,
"learning_rate": 2.4497239575803233e-05,
"loss": 2.2473,
"step": 1039
},
{
"epoch": 1.419184161119645,
"grad_norm": 1.2249808311462402,
"learning_rate": 2.447098192497467e-05,
"loss": 2.2835,
"step": 1040
},
{
"epoch": 1.4205495818399045,
"grad_norm": 1.257360577583313,
"learning_rate": 2.4444718913865506e-05,
"loss": 2.4347,
"step": 1041
},
{
"epoch": 1.421915002560164,
"grad_norm": 1.2087372541427612,
"learning_rate": 2.441845059944789e-05,
"loss": 2.318,
"step": 1042
},
{
"epoch": 1.4232804232804233,
"grad_norm": 1.2171880006790161,
"learning_rate": 2.4392177038705432e-05,
"loss": 2.3752,
"step": 1043
},
{
"epoch": 1.4246458440006826,
"grad_norm": 1.2110891342163086,
"learning_rate": 2.436589828863317e-05,
"loss": 2.3574,
"step": 1044
},
{
"epoch": 1.4260112647209422,
"grad_norm": 1.2780026197433472,
"learning_rate": 2.433961440623737e-05,
"loss": 2.3068,
"step": 1045
},
{
"epoch": 1.4273766854412016,
"grad_norm": 1.1810399293899536,
"learning_rate": 2.431332544853544e-05,
"loss": 2.2941,
"step": 1046
},
{
"epoch": 1.4287421061614611,
"grad_norm": 1.2623032331466675,
"learning_rate": 2.4287031472555796e-05,
"loss": 2.3146,
"step": 1047
},
{
"epoch": 1.4301075268817205,
"grad_norm": 1.250718355178833,
"learning_rate": 2.426073253533775e-05,
"loss": 2.3973,
"step": 1048
},
{
"epoch": 1.4314729476019799,
"grad_norm": 1.2748031616210938,
"learning_rate": 2.4234428693931355e-05,
"loss": 2.4401,
"step": 1049
},
{
"epoch": 1.4328383683222392,
"grad_norm": 1.1808706521987915,
"learning_rate": 2.420812000539734e-05,
"loss": 2.1996,
"step": 1050
},
{
"epoch": 1.4342037890424988,
"grad_norm": 1.2356598377227783,
"learning_rate": 2.4181806526806893e-05,
"loss": 2.3672,
"step": 1051
},
{
"epoch": 1.4355692097627581,
"grad_norm": 1.2032191753387451,
"learning_rate": 2.415548831524165e-05,
"loss": 2.2864,
"step": 1052
},
{
"epoch": 1.4369346304830175,
"grad_norm": 1.2007215023040771,
"learning_rate": 2.4129165427793475e-05,
"loss": 2.3323,
"step": 1053
},
{
"epoch": 1.438300051203277,
"grad_norm": 1.2430835962295532,
"learning_rate": 2.41028379215644e-05,
"loss": 2.4303,
"step": 1054
},
{
"epoch": 1.4396654719235364,
"grad_norm": 1.2647690773010254,
"learning_rate": 2.4076505853666458e-05,
"loss": 2.3199,
"step": 1055
},
{
"epoch": 1.4410308926437958,
"grad_norm": 1.1827330589294434,
"learning_rate": 2.405016928122159e-05,
"loss": 2.2713,
"step": 1056
},
{
"epoch": 1.4423963133640554,
"grad_norm": 1.2489955425262451,
"learning_rate": 2.40238282613615e-05,
"loss": 2.4077,
"step": 1057
},
{
"epoch": 1.4437617340843147,
"grad_norm": 1.275048851966858,
"learning_rate": 2.399748285122755e-05,
"loss": 2.3997,
"step": 1058
},
{
"epoch": 1.445127154804574,
"grad_norm": 1.2162705659866333,
"learning_rate": 2.3971133107970617e-05,
"loss": 2.2288,
"step": 1059
},
{
"epoch": 1.4464925755248337,
"grad_norm": 1.2077502012252808,
"learning_rate": 2.3944779088750975e-05,
"loss": 2.3569,
"step": 1060
},
{
"epoch": 1.447857996245093,
"grad_norm": 1.221274733543396,
"learning_rate": 2.391842085073818e-05,
"loss": 2.3245,
"step": 1061
},
{
"epoch": 1.4492234169653524,
"grad_norm": 1.2395799160003662,
"learning_rate": 2.389205845111095e-05,
"loss": 2.3427,
"step": 1062
},
{
"epoch": 1.4505888376856118,
"grad_norm": 1.1973603963851929,
"learning_rate": 2.386569194705701e-05,
"loss": 2.3743,
"step": 1063
},
{
"epoch": 1.4519542584058713,
"grad_norm": 1.2639575004577637,
"learning_rate": 2.383932139577301e-05,
"loss": 2.3444,
"step": 1064
},
{
"epoch": 1.4533196791261307,
"grad_norm": 1.23374342918396,
"learning_rate": 2.381294685446435e-05,
"loss": 2.3924,
"step": 1065
},
{
"epoch": 1.4546850998463903,
"grad_norm": 1.3239967823028564,
"learning_rate": 2.378656838034511e-05,
"loss": 2.3175,
"step": 1066
},
{
"epoch": 1.4560505205666496,
"grad_norm": 1.2364667654037476,
"learning_rate": 2.3760186030637905e-05,
"loss": 2.2509,
"step": 1067
},
{
"epoch": 1.457415941286909,
"grad_norm": 1.2950961589813232,
"learning_rate": 2.3733799862573746e-05,
"loss": 2.5016,
"step": 1068
},
{
"epoch": 1.4587813620071683,
"grad_norm": 1.3054630756378174,
"learning_rate": 2.3707409933391918e-05,
"loss": 2.3983,
"step": 1069
},
{
"epoch": 1.460146782727428,
"grad_norm": 1.2503340244293213,
"learning_rate": 2.3681016300339875e-05,
"loss": 2.364,
"step": 1070
},
{
"epoch": 1.4615122034476873,
"grad_norm": 1.2855103015899658,
"learning_rate": 2.3654619020673124e-05,
"loss": 2.396,
"step": 1071
},
{
"epoch": 1.4628776241679469,
"grad_norm": 1.1567610502243042,
"learning_rate": 2.3628218151655043e-05,
"loss": 2.2414,
"step": 1072
},
{
"epoch": 1.4642430448882062,
"grad_norm": 1.2140599489212036,
"learning_rate": 2.3601813750556834e-05,
"loss": 2.2305,
"step": 1073
},
{
"epoch": 1.4656084656084656,
"grad_norm": 1.2495917081832886,
"learning_rate": 2.3575405874657343e-05,
"loss": 2.3646,
"step": 1074
},
{
"epoch": 1.466973886328725,
"grad_norm": 1.2478123903274536,
"learning_rate": 2.3548994581242948e-05,
"loss": 2.6012,
"step": 1075
},
{
"epoch": 1.4683393070489845,
"grad_norm": 1.2799197435379028,
"learning_rate": 2.352257992760745e-05,
"loss": 2.303,
"step": 1076
},
{
"epoch": 1.4697047277692439,
"grad_norm": 1.2615288496017456,
"learning_rate": 2.349616197105195e-05,
"loss": 2.375,
"step": 1077
},
{
"epoch": 1.4710701484895035,
"grad_norm": 1.1668665409088135,
"learning_rate": 2.346974076888469e-05,
"loss": 2.199,
"step": 1078
},
{
"epoch": 1.4724355692097628,
"grad_norm": 1.1774864196777344,
"learning_rate": 2.3443316378420975e-05,
"loss": 2.3174,
"step": 1079
},
{
"epoch": 1.4738009899300222,
"grad_norm": 1.236282229423523,
"learning_rate": 2.3416888856983003e-05,
"loss": 2.4059,
"step": 1080
},
{
"epoch": 1.4751664106502815,
"grad_norm": 1.2326610088348389,
"learning_rate": 2.3390458261899783e-05,
"loss": 2.4145,
"step": 1081
},
{
"epoch": 1.4765318313705411,
"grad_norm": 1.2931314706802368,
"learning_rate": 2.3364024650506986e-05,
"loss": 2.3436,
"step": 1082
},
{
"epoch": 1.4778972520908005,
"grad_norm": 1.218035340309143,
"learning_rate": 2.333758808014684e-05,
"loss": 2.304,
"step": 1083
},
{
"epoch": 1.4792626728110598,
"grad_norm": 1.196103811264038,
"learning_rate": 2.3311148608167944e-05,
"loss": 2.266,
"step": 1084
},
{
"epoch": 1.4806280935313194,
"grad_norm": 1.24627685546875,
"learning_rate": 2.3284706291925257e-05,
"loss": 2.4143,
"step": 1085
},
{
"epoch": 1.4819935142515788,
"grad_norm": 1.2291640043258667,
"learning_rate": 2.3258261188779863e-05,
"loss": 2.4181,
"step": 1086
},
{
"epoch": 1.4833589349718381,
"grad_norm": 1.2706551551818848,
"learning_rate": 2.323181335609891e-05,
"loss": 2.3843,
"step": 1087
},
{
"epoch": 1.4847243556920977,
"grad_norm": 1.222783088684082,
"learning_rate": 2.3205362851255453e-05,
"loss": 2.219,
"step": 1088
},
{
"epoch": 1.486089776412357,
"grad_norm": 1.2496508359909058,
"learning_rate": 2.3178909731628365e-05,
"loss": 2.1473,
"step": 1089
},
{
"epoch": 1.4874551971326164,
"grad_norm": 1.1811407804489136,
"learning_rate": 2.3152454054602166e-05,
"loss": 2.4481,
"step": 1090
},
{
"epoch": 1.488820617852876,
"grad_norm": 1.3018665313720703,
"learning_rate": 2.312599587756695e-05,
"loss": 2.4836,
"step": 1091
},
{
"epoch": 1.4901860385731354,
"grad_norm": 1.2372182607650757,
"learning_rate": 2.3099535257918212e-05,
"loss": 2.3584,
"step": 1092
},
{
"epoch": 1.4915514592933947,
"grad_norm": 1.2060118913650513,
"learning_rate": 2.307307225305676e-05,
"loss": 2.3097,
"step": 1093
},
{
"epoch": 1.492916880013654,
"grad_norm": 1.256250023841858,
"learning_rate": 2.304660692038857e-05,
"loss": 2.3829,
"step": 1094
},
{
"epoch": 1.4942823007339137,
"grad_norm": 1.220645785331726,
"learning_rate": 2.302013931732466e-05,
"loss": 2.4993,
"step": 1095
},
{
"epoch": 1.495647721454173,
"grad_norm": 1.2205283641815186,
"learning_rate": 2.2993669501280995e-05,
"loss": 2.3144,
"step": 1096
},
{
"epoch": 1.4970131421744326,
"grad_norm": 1.2228596210479736,
"learning_rate": 2.2967197529678323e-05,
"loss": 2.2846,
"step": 1097
},
{
"epoch": 1.498378562894692,
"grad_norm": 1.1909816265106201,
"learning_rate": 2.2940723459942067e-05,
"loss": 2.3701,
"step": 1098
},
{
"epoch": 1.4997439836149513,
"grad_norm": 1.2442682981491089,
"learning_rate": 2.2914247349502206e-05,
"loss": 2.4626,
"step": 1099
},
{
"epoch": 1.5011094043352107,
"grad_norm": 1.228248119354248,
"learning_rate": 2.2887769255793165e-05,
"loss": 2.3222,
"step": 1100
},
{
"epoch": 1.5024748250554703,
"grad_norm": 1.2103149890899658,
"learning_rate": 2.2861289236253627e-05,
"loss": 2.3756,
"step": 1101
},
{
"epoch": 1.5038402457757296,
"grad_norm": 1.2455346584320068,
"learning_rate": 2.2834807348326497e-05,
"loss": 2.1895,
"step": 1102
},
{
"epoch": 1.5052056664959892,
"grad_norm": 1.1884119510650635,
"learning_rate": 2.2808323649458703e-05,
"loss": 2.3963,
"step": 1103
},
{
"epoch": 1.5065710872162486,
"grad_norm": 1.23482346534729,
"learning_rate": 2.278183819710113e-05,
"loss": 2.3652,
"step": 1104
},
{
"epoch": 1.507936507936508,
"grad_norm": 1.2856261730194092,
"learning_rate": 2.2755351048708428e-05,
"loss": 2.4035,
"step": 1105
},
{
"epoch": 1.5093019286567673,
"grad_norm": 1.240214228630066,
"learning_rate": 2.2728862261738972e-05,
"loss": 2.3496,
"step": 1106
},
{
"epoch": 1.5106673493770268,
"grad_norm": 1.198083758354187,
"learning_rate": 2.2702371893654646e-05,
"loss": 2.3861,
"step": 1107
},
{
"epoch": 1.5120327700972862,
"grad_norm": 1.1971734762191772,
"learning_rate": 2.2675880001920806e-05,
"loss": 2.3152,
"step": 1108
},
{
"epoch": 1.5133981908175458,
"grad_norm": 1.1949067115783691,
"learning_rate": 2.264938664400608e-05,
"loss": 2.3054,
"step": 1109
},
{
"epoch": 1.5147636115378051,
"grad_norm": 1.2134844064712524,
"learning_rate": 2.2622891877382285e-05,
"loss": 2.3676,
"step": 1110
},
{
"epoch": 1.5161290322580645,
"grad_norm": 1.1985032558441162,
"learning_rate": 2.2596395759524316e-05,
"loss": 2.2658,
"step": 1111
},
{
"epoch": 1.5174944529783239,
"grad_norm": 1.2136932611465454,
"learning_rate": 2.2569898347909967e-05,
"loss": 2.3402,
"step": 1112
},
{
"epoch": 1.5188598736985832,
"grad_norm": 1.2195227146148682,
"learning_rate": 2.2543399700019858e-05,
"loss": 2.3061,
"step": 1113
},
{
"epoch": 1.5202252944188428,
"grad_norm": 1.241693139076233,
"learning_rate": 2.2516899873337286e-05,
"loss": 2.3131,
"step": 1114
},
{
"epoch": 1.5215907151391024,
"grad_norm": 1.1966034173965454,
"learning_rate": 2.2490398925348108e-05,
"loss": 2.2667,
"step": 1115
},
{
"epoch": 1.5229561358593617,
"grad_norm": 1.2732926607131958,
"learning_rate": 2.2463896913540605e-05,
"loss": 2.3658,
"step": 1116
},
{
"epoch": 1.524321556579621,
"grad_norm": 1.2363969087600708,
"learning_rate": 2.2437393895405377e-05,
"loss": 2.3828,
"step": 1117
},
{
"epoch": 1.5256869772998805,
"grad_norm": 1.2610344886779785,
"learning_rate": 2.24108899284352e-05,
"loss": 2.3947,
"step": 1118
},
{
"epoch": 1.5270523980201398,
"grad_norm": 1.3581113815307617,
"learning_rate": 2.2384385070124907e-05,
"loss": 2.4109,
"step": 1119
},
{
"epoch": 1.5284178187403994,
"grad_norm": 1.242037057876587,
"learning_rate": 2.2357879377971265e-05,
"loss": 2.2902,
"step": 1120
},
{
"epoch": 1.529783239460659,
"grad_norm": 1.235823154449463,
"learning_rate": 2.2331372909472863e-05,
"loss": 2.3601,
"step": 1121
},
{
"epoch": 1.5311486601809183,
"grad_norm": 1.2350409030914307,
"learning_rate": 2.2304865722129948e-05,
"loss": 2.379,
"step": 1122
},
{
"epoch": 1.5325140809011777,
"grad_norm": 1.271335482597351,
"learning_rate": 2.227835787344436e-05,
"loss": 2.3458,
"step": 1123
},
{
"epoch": 1.533879501621437,
"grad_norm": 1.2029156684875488,
"learning_rate": 2.2251849420919342e-05,
"loss": 2.3131,
"step": 1124
},
{
"epoch": 1.5352449223416964,
"grad_norm": 1.2248457670211792,
"learning_rate": 2.2225340422059463e-05,
"loss": 2.4609,
"step": 1125
},
{
"epoch": 1.536610343061956,
"grad_norm": 1.2357947826385498,
"learning_rate": 2.2198830934370474e-05,
"loss": 2.2153,
"step": 1126
},
{
"epoch": 1.5379757637822153,
"grad_norm": 1.2080334424972534,
"learning_rate": 2.2172321015359196e-05,
"loss": 2.3497,
"step": 1127
},
{
"epoch": 1.539341184502475,
"grad_norm": 1.1954673528671265,
"learning_rate": 2.214581072253336e-05,
"loss": 2.1777,
"step": 1128
},
{
"epoch": 1.5407066052227343,
"grad_norm": 1.1885936260223389,
"learning_rate": 2.2119300113401527e-05,
"loss": 2.35,
"step": 1129
},
{
"epoch": 1.5420720259429936,
"grad_norm": 1.2441728115081787,
"learning_rate": 2.209278924547296e-05,
"loss": 2.4242,
"step": 1130
},
{
"epoch": 1.543437446663253,
"grad_norm": 1.2682271003723145,
"learning_rate": 2.206627817625744e-05,
"loss": 2.2407,
"step": 1131
},
{
"epoch": 1.5448028673835126,
"grad_norm": 1.2359076738357544,
"learning_rate": 2.2039766963265237e-05,
"loss": 2.3846,
"step": 1132
},
{
"epoch": 1.546168288103772,
"grad_norm": 1.303910732269287,
"learning_rate": 2.201325566400688e-05,
"loss": 2.3654,
"step": 1133
},
{
"epoch": 1.5475337088240315,
"grad_norm": 1.2179566621780396,
"learning_rate": 2.1986744335993126e-05,
"loss": 2.2462,
"step": 1134
},
{
"epoch": 1.5488991295442909,
"grad_norm": 1.2264856100082397,
"learning_rate": 2.1960233036734766e-05,
"loss": 2.3594,
"step": 1135
},
{
"epoch": 1.5502645502645502,
"grad_norm": 1.3011183738708496,
"learning_rate": 2.1933721823742554e-05,
"loss": 2.368,
"step": 1136
},
{
"epoch": 1.5516299709848096,
"grad_norm": 1.26678466796875,
"learning_rate": 2.190721075452705e-05,
"loss": 2.3207,
"step": 1137
},
{
"epoch": 1.5529953917050692,
"grad_norm": 1.2640334367752075,
"learning_rate": 2.188069988659848e-05,
"loss": 2.4533,
"step": 1138
},
{
"epoch": 1.5543608124253285,
"grad_norm": 1.269268274307251,
"learning_rate": 2.185418927746665e-05,
"loss": 2.3017,
"step": 1139
},
{
"epoch": 1.5557262331455881,
"grad_norm": 1.3183499574661255,
"learning_rate": 2.1827678984640816e-05,
"loss": 2.3446,
"step": 1140
},
{
"epoch": 1.5570916538658475,
"grad_norm": 1.2338331937789917,
"learning_rate": 2.180116906562953e-05,
"loss": 2.3517,
"step": 1141
},
{
"epoch": 1.5584570745861068,
"grad_norm": 1.2574589252471924,
"learning_rate": 2.177465957794054e-05,
"loss": 2.4078,
"step": 1142
},
{
"epoch": 1.5598224953063662,
"grad_norm": 1.220977544784546,
"learning_rate": 2.1748150579080667e-05,
"loss": 2.3536,
"step": 1143
},
{
"epoch": 1.5611879160266255,
"grad_norm": 1.2266207933425903,
"learning_rate": 2.1721642126555647e-05,
"loss": 2.3016,
"step": 1144
},
{
"epoch": 1.5625533367468851,
"grad_norm": 1.2747420072555542,
"learning_rate": 2.1695134277870054e-05,
"loss": 2.4355,
"step": 1145
},
{
"epoch": 1.5639187574671447,
"grad_norm": 1.2303208112716675,
"learning_rate": 2.1668627090527143e-05,
"loss": 2.3895,
"step": 1146
},
{
"epoch": 1.565284178187404,
"grad_norm": 1.2184724807739258,
"learning_rate": 2.1642120622028737e-05,
"loss": 2.2917,
"step": 1147
},
{
"epoch": 1.5666495989076634,
"grad_norm": 1.2458738088607788,
"learning_rate": 2.16156149298751e-05,
"loss": 2.3045,
"step": 1148
},
{
"epoch": 1.5680150196279228,
"grad_norm": 1.2226184606552124,
"learning_rate": 2.158911007156481e-05,
"loss": 2.3442,
"step": 1149
},
{
"epoch": 1.5693804403481821,
"grad_norm": 1.3232007026672363,
"learning_rate": 2.1562606104594633e-05,
"loss": 2.4039,
"step": 1150
},
{
"epoch": 1.5707458610684417,
"grad_norm": 1.2720632553100586,
"learning_rate": 2.15361030864594e-05,
"loss": 2.2755,
"step": 1151
},
{
"epoch": 1.5721112817887013,
"grad_norm": 1.2147183418273926,
"learning_rate": 2.1509601074651898e-05,
"loss": 2.3679,
"step": 1152
},
{
"epoch": 1.5734767025089607,
"grad_norm": 1.285058856010437,
"learning_rate": 2.1483100126662716e-05,
"loss": 2.4596,
"step": 1153
},
{
"epoch": 1.57484212322922,
"grad_norm": 1.282956838607788,
"learning_rate": 2.1456600299980148e-05,
"loss": 2.3848,
"step": 1154
},
{
"epoch": 1.5762075439494794,
"grad_norm": 1.2431665658950806,
"learning_rate": 2.1430101652090042e-05,
"loss": 2.2509,
"step": 1155
},
{
"epoch": 1.5775729646697387,
"grad_norm": 1.2542580366134644,
"learning_rate": 2.140360424047569e-05,
"loss": 2.3691,
"step": 1156
},
{
"epoch": 1.5789383853899983,
"grad_norm": 1.314717411994934,
"learning_rate": 2.1377108122617717e-05,
"loss": 2.4533,
"step": 1157
},
{
"epoch": 1.5803038061102577,
"grad_norm": 1.2104456424713135,
"learning_rate": 2.135061335599393e-05,
"loss": 2.3082,
"step": 1158
},
{
"epoch": 1.5816692268305172,
"grad_norm": 1.2363940477371216,
"learning_rate": 2.1324119998079196e-05,
"loss": 2.4173,
"step": 1159
},
{
"epoch": 1.5830346475507766,
"grad_norm": 1.255435585975647,
"learning_rate": 2.1297628106345357e-05,
"loss": 2.438,
"step": 1160
},
{
"epoch": 1.584400068271036,
"grad_norm": 1.2767750024795532,
"learning_rate": 2.1271137738261033e-05,
"loss": 2.3624,
"step": 1161
},
{
"epoch": 1.5857654889912953,
"grad_norm": 1.2551891803741455,
"learning_rate": 2.1244648951291578e-05,
"loss": 2.3915,
"step": 1162
},
{
"epoch": 1.587130909711555,
"grad_norm": 1.2307158708572388,
"learning_rate": 2.121816180289888e-05,
"loss": 2.3746,
"step": 1163
},
{
"epoch": 1.5884963304318143,
"grad_norm": 1.2425217628479004,
"learning_rate": 2.11916763505413e-05,
"loss": 2.342,
"step": 1164
},
{
"epoch": 1.5898617511520738,
"grad_norm": 1.2080459594726562,
"learning_rate": 2.116519265167351e-05,
"loss": 2.3766,
"step": 1165
},
{
"epoch": 1.5912271718723332,
"grad_norm": 1.23215651512146,
"learning_rate": 2.1138710763746385e-05,
"loss": 2.4228,
"step": 1166
},
{
"epoch": 1.5925925925925926,
"grad_norm": 1.242877721786499,
"learning_rate": 2.1112230744206847e-05,
"loss": 2.3667,
"step": 1167
},
{
"epoch": 1.593958013312852,
"grad_norm": 1.178984522819519,
"learning_rate": 2.1085752650497796e-05,
"loss": 2.3217,
"step": 1168
},
{
"epoch": 1.5953234340331115,
"grad_norm": 1.2545174360275269,
"learning_rate": 2.105927654005794e-05,
"loss": 2.3273,
"step": 1169
},
{
"epoch": 1.5966888547533709,
"grad_norm": 1.2036995887756348,
"learning_rate": 2.1032802470321686e-05,
"loss": 2.3873,
"step": 1170
},
{
"epoch": 1.5980542754736304,
"grad_norm": 1.2009844779968262,
"learning_rate": 2.1006330498719008e-05,
"loss": 2.2696,
"step": 1171
},
{
"epoch": 1.5994196961938898,
"grad_norm": 1.2024636268615723,
"learning_rate": 2.0979860682675344e-05,
"loss": 2.2695,
"step": 1172
},
{
"epoch": 1.6007851169141492,
"grad_norm": 1.2805627584457397,
"learning_rate": 2.095339307961144e-05,
"loss": 2.3947,
"step": 1173
},
{
"epoch": 1.6021505376344085,
"grad_norm": 1.2442028522491455,
"learning_rate": 2.0926927746943243e-05,
"loss": 2.3886,
"step": 1174
},
{
"epoch": 1.6035159583546679,
"grad_norm": 1.1755213737487793,
"learning_rate": 2.0900464742081793e-05,
"loss": 2.3443,
"step": 1175
},
{
"epoch": 1.6048813790749274,
"grad_norm": 1.2607914209365845,
"learning_rate": 2.087400412243305e-05,
"loss": 2.3431,
"step": 1176
},
{
"epoch": 1.606246799795187,
"grad_norm": 1.2009880542755127,
"learning_rate": 2.0847545945397836e-05,
"loss": 2.4301,
"step": 1177
},
{
"epoch": 1.6076122205154464,
"grad_norm": 1.2095117568969727,
"learning_rate": 2.0821090268371648e-05,
"loss": 2.3518,
"step": 1178
},
{
"epoch": 1.6089776412357057,
"grad_norm": 1.2514299154281616,
"learning_rate": 2.0794637148744553e-05,
"loss": 2.4091,
"step": 1179
},
{
"epoch": 1.610343061955965,
"grad_norm": 1.2250244617462158,
"learning_rate": 2.0768186643901094e-05,
"loss": 2.2549,
"step": 1180
},
{
"epoch": 1.6117084826762245,
"grad_norm": 1.2491207122802734,
"learning_rate": 2.0741738811220142e-05,
"loss": 2.3577,
"step": 1181
},
{
"epoch": 1.613073903396484,
"grad_norm": 1.2476170063018799,
"learning_rate": 2.0715293708074745e-05,
"loss": 2.3185,
"step": 1182
},
{
"epoch": 1.6144393241167436,
"grad_norm": 1.289138913154602,
"learning_rate": 2.0688851391832058e-05,
"loss": 2.4487,
"step": 1183
},
{
"epoch": 1.615804744837003,
"grad_norm": 1.196171760559082,
"learning_rate": 2.0662411919853177e-05,
"loss": 2.4455,
"step": 1184
},
{
"epoch": 1.6171701655572623,
"grad_norm": 1.259664535522461,
"learning_rate": 2.063597534949302e-05,
"loss": 2.2797,
"step": 1185
},
{
"epoch": 1.6185355862775217,
"grad_norm": 1.2392253875732422,
"learning_rate": 2.0609541738100223e-05,
"loss": 2.507,
"step": 1186
},
{
"epoch": 1.619901006997781,
"grad_norm": 1.2401288747787476,
"learning_rate": 2.0583111143017007e-05,
"loss": 2.2597,
"step": 1187
},
{
"epoch": 1.6212664277180406,
"grad_norm": 1.212021827697754,
"learning_rate": 2.0556683621579037e-05,
"loss": 2.4304,
"step": 1188
},
{
"epoch": 1.6226318484383002,
"grad_norm": 1.2344893217086792,
"learning_rate": 2.0530259231115308e-05,
"loss": 2.3359,
"step": 1189
},
{
"epoch": 1.6239972691585596,
"grad_norm": 1.2323673963546753,
"learning_rate": 2.0503838028948055e-05,
"loss": 2.352,
"step": 1190
},
{
"epoch": 1.625362689878819,
"grad_norm": 1.2047122716903687,
"learning_rate": 2.0477420072392555e-05,
"loss": 2.2971,
"step": 1191
},
{
"epoch": 1.6267281105990783,
"grad_norm": 1.2283841371536255,
"learning_rate": 2.0451005418757058e-05,
"loss": 2.4028,
"step": 1192
},
{
"epoch": 1.6280935313193377,
"grad_norm": 1.166121244430542,
"learning_rate": 2.0424594125342666e-05,
"loss": 2.2876,
"step": 1193
},
{
"epoch": 1.6294589520395972,
"grad_norm": 1.1811237335205078,
"learning_rate": 2.0398186249443172e-05,
"loss": 2.3199,
"step": 1194
},
{
"epoch": 1.6308243727598566,
"grad_norm": 1.208305835723877,
"learning_rate": 2.037178184834496e-05,
"loss": 2.4115,
"step": 1195
},
{
"epoch": 1.6321897934801162,
"grad_norm": 1.2738384008407593,
"learning_rate": 2.034538097932689e-05,
"loss": 2.4104,
"step": 1196
},
{
"epoch": 1.6335552142003755,
"grad_norm": 1.2632803916931152,
"learning_rate": 2.031898369966013e-05,
"loss": 2.3641,
"step": 1197
},
{
"epoch": 1.6349206349206349,
"grad_norm": 1.211148738861084,
"learning_rate": 2.029259006660809e-05,
"loss": 2.2214,
"step": 1198
},
{
"epoch": 1.6362860556408942,
"grad_norm": 1.2432053089141846,
"learning_rate": 2.0266200137426263e-05,
"loss": 2.2065,
"step": 1199
},
{
"epoch": 1.6376514763611538,
"grad_norm": 1.1672781705856323,
"learning_rate": 2.0239813969362097e-05,
"loss": 2.2865,
"step": 1200
},
{
"epoch": 1.6390168970814132,
"grad_norm": 1.197617769241333,
"learning_rate": 2.0213431619654894e-05,
"loss": 2.2293,
"step": 1201
},
{
"epoch": 1.6403823178016728,
"grad_norm": 1.1703133583068848,
"learning_rate": 2.0187053145535663e-05,
"loss": 2.2546,
"step": 1202
},
{
"epoch": 1.6417477385219321,
"grad_norm": 1.2296996116638184,
"learning_rate": 2.0160678604227e-05,
"loss": 2.3584,
"step": 1203
},
{
"epoch": 1.6431131592421915,
"grad_norm": 1.259677767753601,
"learning_rate": 2.0134308052942995e-05,
"loss": 2.3224,
"step": 1204
},
{
"epoch": 1.6444785799624508,
"grad_norm": 1.17851984500885,
"learning_rate": 2.0107941548889056e-05,
"loss": 2.3071,
"step": 1205
},
{
"epoch": 1.6458440006827102,
"grad_norm": 1.2315696477890015,
"learning_rate": 2.008157914926182e-05,
"loss": 2.3561,
"step": 1206
},
{
"epoch": 1.6472094214029698,
"grad_norm": 1.2477763891220093,
"learning_rate": 2.0055220911249034e-05,
"loss": 2.3604,
"step": 1207
},
{
"epoch": 1.6485748421232294,
"grad_norm": 1.2165106534957886,
"learning_rate": 2.0028866892029395e-05,
"loss": 2.3117,
"step": 1208
},
{
"epoch": 1.6499402628434887,
"grad_norm": 1.245399832725525,
"learning_rate": 2.000251714877246e-05,
"loss": 2.3411,
"step": 1209
},
{
"epoch": 1.651305683563748,
"grad_norm": 1.173814058303833,
"learning_rate": 1.9976171738638504e-05,
"loss": 2.3077,
"step": 1210
},
{
"epoch": 1.6526711042840074,
"grad_norm": 1.1953760385513306,
"learning_rate": 1.9949830718778415e-05,
"loss": 2.3804,
"step": 1211
},
{
"epoch": 1.6540365250042668,
"grad_norm": 1.2615714073181152,
"learning_rate": 1.9923494146333544e-05,
"loss": 2.4287,
"step": 1212
},
{
"epoch": 1.6554019457245264,
"grad_norm": 1.2128006219863892,
"learning_rate": 1.9897162078435613e-05,
"loss": 2.2973,
"step": 1213
},
{
"epoch": 1.656767366444786,
"grad_norm": 1.2068262100219727,
"learning_rate": 1.9870834572206527e-05,
"loss": 2.2809,
"step": 1214
},
{
"epoch": 1.6581327871650453,
"grad_norm": 1.2407810688018799,
"learning_rate": 1.984451168475836e-05,
"loss": 2.4074,
"step": 1215
},
{
"epoch": 1.6594982078853047,
"grad_norm": 1.2685725688934326,
"learning_rate": 1.9818193473193113e-05,
"loss": 2.3072,
"step": 1216
},
{
"epoch": 1.660863628605564,
"grad_norm": 1.2917054891586304,
"learning_rate": 1.9791879994602674e-05,
"loss": 2.396,
"step": 1217
},
{
"epoch": 1.6622290493258234,
"grad_norm": 1.2833771705627441,
"learning_rate": 1.9765571306068647e-05,
"loss": 2.3787,
"step": 1218
},
{
"epoch": 1.663594470046083,
"grad_norm": 1.1846504211425781,
"learning_rate": 1.973926746466226e-05,
"loss": 2.2715,
"step": 1219
},
{
"epoch": 1.6649598907663425,
"grad_norm": 1.2178796529769897,
"learning_rate": 1.9712968527444216e-05,
"loss": 2.3325,
"step": 1220
},
{
"epoch": 1.666325311486602,
"grad_norm": 1.167587399482727,
"learning_rate": 1.9686674551464566e-05,
"loss": 2.4707,
"step": 1221
},
{
"epoch": 1.6676907322068613,
"grad_norm": 1.2548010349273682,
"learning_rate": 1.966038559376264e-05,
"loss": 2.3621,
"step": 1222
},
{
"epoch": 1.6690561529271206,
"grad_norm": 1.2430486679077148,
"learning_rate": 1.9634101711366833e-05,
"loss": 2.3817,
"step": 1223
},
{
"epoch": 1.67042157364738,
"grad_norm": 1.2409595251083374,
"learning_rate": 1.9607822961294567e-05,
"loss": 2.3581,
"step": 1224
},
{
"epoch": 1.6717869943676396,
"grad_norm": 1.1380168199539185,
"learning_rate": 1.958154940055212e-05,
"loss": 2.2757,
"step": 1225
},
{
"epoch": 1.673152415087899,
"grad_norm": 1.2114099264144897,
"learning_rate": 1.9555281086134496e-05,
"loss": 2.4302,
"step": 1226
},
{
"epoch": 1.6745178358081585,
"grad_norm": 1.1802924871444702,
"learning_rate": 1.952901807502533e-05,
"loss": 2.3748,
"step": 1227
},
{
"epoch": 1.6758832565284179,
"grad_norm": 1.2748708724975586,
"learning_rate": 1.9502760424196773e-05,
"loss": 2.3904,
"step": 1228
},
{
"epoch": 1.6772486772486772,
"grad_norm": 1.2163430452346802,
"learning_rate": 1.9476508190609316e-05,
"loss": 2.3279,
"step": 1229
},
{
"epoch": 1.6786140979689366,
"grad_norm": 1.2480100393295288,
"learning_rate": 1.945026143121171e-05,
"loss": 2.3195,
"step": 1230
},
{
"epoch": 1.6799795186891961,
"grad_norm": 1.2400996685028076,
"learning_rate": 1.9424020202940844e-05,
"loss": 2.306,
"step": 1231
},
{
"epoch": 1.6813449394094555,
"grad_norm": 1.2779078483581543,
"learning_rate": 1.939778456272158e-05,
"loss": 2.3093,
"step": 1232
},
{
"epoch": 1.682710360129715,
"grad_norm": 1.1824541091918945,
"learning_rate": 1.9371554567466687e-05,
"loss": 2.2926,
"step": 1233
},
{
"epoch": 1.6840757808499744,
"grad_norm": 1.1961851119995117,
"learning_rate": 1.9345330274076675e-05,
"loss": 2.3872,
"step": 1234
},
{
"epoch": 1.6854412015702338,
"grad_norm": 1.2460107803344727,
"learning_rate": 1.931911173943968e-05,
"loss": 2.5012,
"step": 1235
},
{
"epoch": 1.6868066222904932,
"grad_norm": 1.234752893447876,
"learning_rate": 1.9292899020431356e-05,
"loss": 2.3076,
"step": 1236
},
{
"epoch": 1.6881720430107527,
"grad_norm": 1.21217942237854,
"learning_rate": 1.9266692173914747e-05,
"loss": 2.3378,
"step": 1237
},
{
"epoch": 1.689537463731012,
"grad_norm": 1.1792680025100708,
"learning_rate": 1.9240491256740127e-05,
"loss": 2.2452,
"step": 1238
},
{
"epoch": 1.6909028844512717,
"grad_norm": 1.2648640871047974,
"learning_rate": 1.921429632574494e-05,
"loss": 2.3229,
"step": 1239
},
{
"epoch": 1.692268305171531,
"grad_norm": 1.18714439868927,
"learning_rate": 1.9188107437753633e-05,
"loss": 2.4348,
"step": 1240
},
{
"epoch": 1.6936337258917904,
"grad_norm": 1.2279962301254272,
"learning_rate": 1.9161924649577542e-05,
"loss": 2.2269,
"step": 1241
},
{
"epoch": 1.6949991466120498,
"grad_norm": 1.194790005683899,
"learning_rate": 1.913574801801478e-05,
"loss": 2.314,
"step": 1242
},
{
"epoch": 1.6963645673323091,
"grad_norm": 1.2696503400802612,
"learning_rate": 1.9109577599850083e-05,
"loss": 2.3511,
"step": 1243
},
{
"epoch": 1.6977299880525687,
"grad_norm": 1.282675862312317,
"learning_rate": 1.908341345185472e-05,
"loss": 2.3288,
"step": 1244
},
{
"epoch": 1.6990954087728283,
"grad_norm": 1.2242058515548706,
"learning_rate": 1.9057255630786376e-05,
"loss": 2.386,
"step": 1245
},
{
"epoch": 1.7004608294930876,
"grad_norm": 1.2264022827148438,
"learning_rate": 1.903110419338899e-05,
"loss": 2.4116,
"step": 1246
},
{
"epoch": 1.701826250213347,
"grad_norm": 1.2401572465896606,
"learning_rate": 1.9004959196392652e-05,
"loss": 2.3218,
"step": 1247
},
{
"epoch": 1.7031916709336063,
"grad_norm": 1.2544509172439575,
"learning_rate": 1.89788206965135e-05,
"loss": 2.4009,
"step": 1248
},
{
"epoch": 1.7045570916538657,
"grad_norm": 1.2233247756958008,
"learning_rate": 1.895268875045355e-05,
"loss": 2.2914,
"step": 1249
},
{
"epoch": 1.7059225123741253,
"grad_norm": 1.2209675312042236,
"learning_rate": 1.892656341490062e-05,
"loss": 2.4089,
"step": 1250
},
{
"epoch": 1.7072879330943849,
"grad_norm": 1.2649309635162354,
"learning_rate": 1.890044474652818e-05,
"loss": 2.2918,
"step": 1251
},
{
"epoch": 1.7086533538146442,
"grad_norm": 1.1954387426376343,
"learning_rate": 1.8874332801995258e-05,
"loss": 2.2829,
"step": 1252
},
{
"epoch": 1.7100187745349036,
"grad_norm": 1.1896917819976807,
"learning_rate": 1.8848227637946257e-05,
"loss": 2.3117,
"step": 1253
},
{
"epoch": 1.711384195255163,
"grad_norm": 1.2037822008132935,
"learning_rate": 1.8822129311010918e-05,
"loss": 2.2518,
"step": 1254
},
{
"epoch": 1.7127496159754223,
"grad_norm": 1.1890069246292114,
"learning_rate": 1.8796037877804098e-05,
"loss": 2.3709,
"step": 1255
},
{
"epoch": 1.7141150366956819,
"grad_norm": 1.2203930616378784,
"learning_rate": 1.8769953394925747e-05,
"loss": 2.2648,
"step": 1256
},
{
"epoch": 1.7154804574159412,
"grad_norm": 1.2275177240371704,
"learning_rate": 1.8743875918960702e-05,
"loss": 2.3832,
"step": 1257
},
{
"epoch": 1.7168458781362008,
"grad_norm": 1.264441967010498,
"learning_rate": 1.8717805506478624e-05,
"loss": 2.4046,
"step": 1258
},
{
"epoch": 1.7182112988564602,
"grad_norm": 1.1798632144927979,
"learning_rate": 1.869174221403385e-05,
"loss": 2.3198,
"step": 1259
},
{
"epoch": 1.7195767195767195,
"grad_norm": 1.233816146850586,
"learning_rate": 1.866568609816525e-05,
"loss": 2.398,
"step": 1260
},
{
"epoch": 1.720942140296979,
"grad_norm": 1.2309916019439697,
"learning_rate": 1.863963721539614e-05,
"loss": 2.3004,
"step": 1261
},
{
"epoch": 1.7223075610172385,
"grad_norm": 1.2181870937347412,
"learning_rate": 1.8613595622234158e-05,
"loss": 2.3404,
"step": 1262
},
{
"epoch": 1.7236729817374978,
"grad_norm": 1.2026022672653198,
"learning_rate": 1.8587561375171104e-05,
"loss": 2.2914,
"step": 1263
},
{
"epoch": 1.7250384024577574,
"grad_norm": 1.2416661977767944,
"learning_rate": 1.856153453068286e-05,
"loss": 2.341,
"step": 1264
},
{
"epoch": 1.7264038231780168,
"grad_norm": 1.2317922115325928,
"learning_rate": 1.8535515145229247e-05,
"loss": 2.3215,
"step": 1265
},
{
"epoch": 1.7277692438982761,
"grad_norm": 1.2122260332107544,
"learning_rate": 1.85095032752539e-05,
"loss": 2.3181,
"step": 1266
},
{
"epoch": 1.7291346646185355,
"grad_norm": 1.1933279037475586,
"learning_rate": 1.8483498977184152e-05,
"loss": 2.2199,
"step": 1267
},
{
"epoch": 1.730500085338795,
"grad_norm": 1.2105554342269897,
"learning_rate": 1.84575023074309e-05,
"loss": 2.1619,
"step": 1268
},
{
"epoch": 1.7318655060590544,
"grad_norm": 1.2069892883300781,
"learning_rate": 1.8431513322388522e-05,
"loss": 2.2365,
"step": 1269
},
{
"epoch": 1.733230926779314,
"grad_norm": 1.1760435104370117,
"learning_rate": 1.8405532078434695e-05,
"loss": 2.2783,
"step": 1270
},
{
"epoch": 1.7345963474995734,
"grad_norm": 1.1959654092788696,
"learning_rate": 1.8379558631930324e-05,
"loss": 2.3919,
"step": 1271
},
{
"epoch": 1.7359617682198327,
"grad_norm": 1.231494426727295,
"learning_rate": 1.835359303921939e-05,
"loss": 2.3484,
"step": 1272
},
{
"epoch": 1.737327188940092,
"grad_norm": 1.2006170749664307,
"learning_rate": 1.832763535662884e-05,
"loss": 2.248,
"step": 1273
},
{
"epoch": 1.7386926096603514,
"grad_norm": 1.1837365627288818,
"learning_rate": 1.830168564046845e-05,
"loss": 2.3024,
"step": 1274
},
{
"epoch": 1.740058030380611,
"grad_norm": 1.2078886032104492,
"learning_rate": 1.827574394703073e-05,
"loss": 2.4261,
"step": 1275
},
{
"epoch": 1.7414234511008706,
"grad_norm": 1.298533320426941,
"learning_rate": 1.824981033259078e-05,
"loss": 2.3713,
"step": 1276
},
{
"epoch": 1.74278887182113,
"grad_norm": 1.216442584991455,
"learning_rate": 1.8223884853406172e-05,
"loss": 2.3904,
"step": 1277
},
{
"epoch": 1.7441542925413893,
"grad_norm": 1.219859004020691,
"learning_rate": 1.8197967565716847e-05,
"loss": 2.2529,
"step": 1278
},
{
"epoch": 1.7455197132616487,
"grad_norm": 1.2334622144699097,
"learning_rate": 1.8172058525744948e-05,
"loss": 2.3991,
"step": 1279
},
{
"epoch": 1.746885133981908,
"grad_norm": 1.1952390670776367,
"learning_rate": 1.8146157789694737e-05,
"loss": 2.3859,
"step": 1280
},
{
"epoch": 1.7482505547021676,
"grad_norm": 1.2213901281356812,
"learning_rate": 1.8120265413752475e-05,
"loss": 2.3523,
"step": 1281
},
{
"epoch": 1.7496159754224272,
"grad_norm": 1.1949186325073242,
"learning_rate": 1.8094381454086278e-05,
"loss": 2.2649,
"step": 1282
},
{
"epoch": 1.7509813961426866,
"grad_norm": 1.176440954208374,
"learning_rate": 1.8068505966846007e-05,
"loss": 2.313,
"step": 1283
},
{
"epoch": 1.752346816862946,
"grad_norm": 1.242050051689148,
"learning_rate": 1.8042639008163133e-05,
"loss": 2.4241,
"step": 1284
},
{
"epoch": 1.7537122375832053,
"grad_norm": 1.2013907432556152,
"learning_rate": 1.8016780634150635e-05,
"loss": 2.2546,
"step": 1285
},
{
"epoch": 1.7550776583034646,
"grad_norm": 1.2207469940185547,
"learning_rate": 1.7990930900902884e-05,
"loss": 2.3802,
"step": 1286
},
{
"epoch": 1.7564430790237242,
"grad_norm": 1.2393518686294556,
"learning_rate": 1.7965089864495472e-05,
"loss": 2.2244,
"step": 1287
},
{
"epoch": 1.7578084997439836,
"grad_norm": 1.2343543767929077,
"learning_rate": 1.7939257580985162e-05,
"loss": 2.427,
"step": 1288
},
{
"epoch": 1.7591739204642431,
"grad_norm": 1.2377870082855225,
"learning_rate": 1.791343410640971e-05,
"loss": 2.2635,
"step": 1289
},
{
"epoch": 1.7605393411845025,
"grad_norm": 1.190020203590393,
"learning_rate": 1.7887619496787754e-05,
"loss": 2.3753,
"step": 1290
},
{
"epoch": 1.7619047619047619,
"grad_norm": 1.2214184999465942,
"learning_rate": 1.786181380811872e-05,
"loss": 2.2924,
"step": 1291
},
{
"epoch": 1.7632701826250212,
"grad_norm": 1.2515887022018433,
"learning_rate": 1.7836017096382673e-05,
"loss": 2.4225,
"step": 1292
},
{
"epoch": 1.7646356033452808,
"grad_norm": 1.2050586938858032,
"learning_rate": 1.7810229417540213e-05,
"loss": 2.3559,
"step": 1293
},
{
"epoch": 1.7660010240655402,
"grad_norm": 1.2048786878585815,
"learning_rate": 1.7784450827532323e-05,
"loss": 2.2441,
"step": 1294
},
{
"epoch": 1.7673664447857997,
"grad_norm": 1.2292776107788086,
"learning_rate": 1.7758681382280302e-05,
"loss": 2.4483,
"step": 1295
},
{
"epoch": 1.768731865506059,
"grad_norm": 1.2190500497817993,
"learning_rate": 1.773292113768558e-05,
"loss": 2.3465,
"step": 1296
},
{
"epoch": 1.7700972862263185,
"grad_norm": 1.1827260255813599,
"learning_rate": 1.770717014962965e-05,
"loss": 2.4278,
"step": 1297
},
{
"epoch": 1.7714627069465778,
"grad_norm": 1.2594752311706543,
"learning_rate": 1.7681428473973908e-05,
"loss": 2.4079,
"step": 1298
},
{
"epoch": 1.7728281276668374,
"grad_norm": 1.205233097076416,
"learning_rate": 1.7655696166559563e-05,
"loss": 2.2988,
"step": 1299
},
{
"epoch": 1.7741935483870968,
"grad_norm": 1.1580568552017212,
"learning_rate": 1.76299732832075e-05,
"loss": 2.1781,
"step": 1300
},
{
"epoch": 1.7755589691073563,
"grad_norm": 1.1895431280136108,
"learning_rate": 1.760425987971816e-05,
"loss": 2.4093,
"step": 1301
},
{
"epoch": 1.7769243898276157,
"grad_norm": 1.1872233152389526,
"learning_rate": 1.7578556011871398e-05,
"loss": 2.3125,
"step": 1302
},
{
"epoch": 1.778289810547875,
"grad_norm": 1.2323864698410034,
"learning_rate": 1.7552861735426415e-05,
"loss": 2.2704,
"step": 1303
},
{
"epoch": 1.7796552312681344,
"grad_norm": 1.1709593534469604,
"learning_rate": 1.7527177106121588e-05,
"loss": 2.1519,
"step": 1304
},
{
"epoch": 1.7810206519883938,
"grad_norm": 1.196088194847107,
"learning_rate": 1.7501502179674378e-05,
"loss": 2.3095,
"step": 1305
},
{
"epoch": 1.7823860727086533,
"grad_norm": 1.1769793033599854,
"learning_rate": 1.7475837011781175e-05,
"loss": 2.3466,
"step": 1306
},
{
"epoch": 1.783751493428913,
"grad_norm": 1.1503105163574219,
"learning_rate": 1.745018165811723e-05,
"loss": 2.3492,
"step": 1307
},
{
"epoch": 1.7851169141491723,
"grad_norm": 1.2130194902420044,
"learning_rate": 1.7424536174336478e-05,
"loss": 2.3666,
"step": 1308
},
{
"epoch": 1.7864823348694316,
"grad_norm": 1.2520149946212769,
"learning_rate": 1.7398900616071457e-05,
"loss": 2.3877,
"step": 1309
},
{
"epoch": 1.787847755589691,
"grad_norm": 1.1659858226776123,
"learning_rate": 1.737327503893318e-05,
"loss": 2.2843,
"step": 1310
},
{
"epoch": 1.7892131763099504,
"grad_norm": 1.2109911441802979,
"learning_rate": 1.7347659498510983e-05,
"loss": 2.3669,
"step": 1311
},
{
"epoch": 1.79057859703021,
"grad_norm": 1.2372288703918457,
"learning_rate": 1.7322054050372455e-05,
"loss": 2.4022,
"step": 1312
},
{
"epoch": 1.7919440177504695,
"grad_norm": 1.1807947158813477,
"learning_rate": 1.7296458750063285e-05,
"loss": 2.4867,
"step": 1313
},
{
"epoch": 1.7933094384707289,
"grad_norm": 1.2210131883621216,
"learning_rate": 1.7270873653107143e-05,
"loss": 2.3282,
"step": 1314
},
{
"epoch": 1.7946748591909882,
"grad_norm": 1.2372682094573975,
"learning_rate": 1.724529881500556e-05,
"loss": 2.3697,
"step": 1315
},
{
"epoch": 1.7960402799112476,
"grad_norm": 1.2280601263046265,
"learning_rate": 1.7219734291237832e-05,
"loss": 2.3421,
"step": 1316
},
{
"epoch": 1.797405700631507,
"grad_norm": 1.2569504976272583,
"learning_rate": 1.7194180137260853e-05,
"loss": 2.4402,
"step": 1317
},
{
"epoch": 1.7987711213517665,
"grad_norm": 1.277348279953003,
"learning_rate": 1.7168636408509064e-05,
"loss": 2.2867,
"step": 1318
},
{
"epoch": 1.800136542072026,
"grad_norm": 1.2433885335922241,
"learning_rate": 1.714310316039424e-05,
"loss": 2.3006,
"step": 1319
},
{
"epoch": 1.8015019627922855,
"grad_norm": 1.1478744745254517,
"learning_rate": 1.7117580448305454e-05,
"loss": 2.1821,
"step": 1320
},
{
"epoch": 1.8028673835125448,
"grad_norm": 1.1595771312713623,
"learning_rate": 1.7092068327608903e-05,
"loss": 2.3428,
"step": 1321
},
{
"epoch": 1.8042328042328042,
"grad_norm": 1.2210897207260132,
"learning_rate": 1.706656685364783e-05,
"loss": 2.2515,
"step": 1322
},
{
"epoch": 1.8055982249530635,
"grad_norm": 1.2028534412384033,
"learning_rate": 1.704107608174237e-05,
"loss": 2.4068,
"step": 1323
},
{
"epoch": 1.8069636456733231,
"grad_norm": 1.239604115486145,
"learning_rate": 1.7015596067189444e-05,
"loss": 2.3543,
"step": 1324
},
{
"epoch": 1.8083290663935825,
"grad_norm": 1.2691093683242798,
"learning_rate": 1.6990126865262634e-05,
"loss": 2.2681,
"step": 1325
},
{
"epoch": 1.809694487113842,
"grad_norm": 1.1803323030471802,
"learning_rate": 1.6964668531212064e-05,
"loss": 2.3009,
"step": 1326
},
{
"epoch": 1.8110599078341014,
"grad_norm": 1.2273375988006592,
"learning_rate": 1.693922112026429e-05,
"loss": 2.3473,
"step": 1327
},
{
"epoch": 1.8124253285543608,
"grad_norm": 1.1603344678878784,
"learning_rate": 1.691378468762217e-05,
"loss": 2.1896,
"step": 1328
},
{
"epoch": 1.8137907492746201,
"grad_norm": 1.2181870937347412,
"learning_rate": 1.688835928846475e-05,
"loss": 2.2582,
"step": 1329
},
{
"epoch": 1.8151561699948797,
"grad_norm": 1.1933072805404663,
"learning_rate": 1.686294497794714e-05,
"loss": 2.3481,
"step": 1330
},
{
"epoch": 1.816521590715139,
"grad_norm": 1.2468600273132324,
"learning_rate": 1.6837541811200382e-05,
"loss": 2.2836,
"step": 1331
},
{
"epoch": 1.8178870114353987,
"grad_norm": 1.2400648593902588,
"learning_rate": 1.6812149843331358e-05,
"loss": 2.4269,
"step": 1332
},
{
"epoch": 1.819252432155658,
"grad_norm": 1.1988747119903564,
"learning_rate": 1.678676912942266e-05,
"loss": 2.2971,
"step": 1333
},
{
"epoch": 1.8206178528759174,
"grad_norm": 1.218732237815857,
"learning_rate": 1.6761399724532446e-05,
"loss": 2.3435,
"step": 1334
},
{
"epoch": 1.8219832735961767,
"grad_norm": 1.225645661354065,
"learning_rate": 1.6736041683694372e-05,
"loss": 2.3266,
"step": 1335
},
{
"epoch": 1.823348694316436,
"grad_norm": 1.23872709274292,
"learning_rate": 1.671069506191743e-05,
"loss": 2.3109,
"step": 1336
},
{
"epoch": 1.8247141150366957,
"grad_norm": 1.1791586875915527,
"learning_rate": 1.668535991418582e-05,
"loss": 2.3324,
"step": 1337
},
{
"epoch": 1.8260795357569553,
"grad_norm": 1.193030834197998,
"learning_rate": 1.6660036295458863e-05,
"loss": 2.2953,
"step": 1338
},
{
"epoch": 1.8274449564772146,
"grad_norm": 1.1756483316421509,
"learning_rate": 1.6634724260670894e-05,
"loss": 2.3178,
"step": 1339
},
{
"epoch": 1.828810377197474,
"grad_norm": 1.230737328529358,
"learning_rate": 1.660942386473109e-05,
"loss": 2.3131,
"step": 1340
},
{
"epoch": 1.8301757979177333,
"grad_norm": 1.2487858533859253,
"learning_rate": 1.658413516252339e-05,
"loss": 2.4517,
"step": 1341
},
{
"epoch": 1.8315412186379927,
"grad_norm": 1.2753077745437622,
"learning_rate": 1.6558858208906375e-05,
"loss": 2.4082,
"step": 1342
},
{
"epoch": 1.8329066393582523,
"grad_norm": 1.2872533798217773,
"learning_rate": 1.653359305871312e-05,
"loss": 2.3859,
"step": 1343
},
{
"epoch": 1.8342720600785118,
"grad_norm": 1.277876615524292,
"learning_rate": 1.6508339766751106e-05,
"loss": 2.5175,
"step": 1344
},
{
"epoch": 1.8356374807987712,
"grad_norm": 1.263545274734497,
"learning_rate": 1.6483098387802098e-05,
"loss": 2.3267,
"step": 1345
},
{
"epoch": 1.8370029015190306,
"grad_norm": 1.2160521745681763,
"learning_rate": 1.6457868976622e-05,
"loss": 2.2412,
"step": 1346
},
{
"epoch": 1.83836832223929,
"grad_norm": 1.214532732963562,
"learning_rate": 1.643265158794077e-05,
"loss": 2.4211,
"step": 1347
},
{
"epoch": 1.8397337429595493,
"grad_norm": 1.175498127937317,
"learning_rate": 1.640744627646229e-05,
"loss": 2.3183,
"step": 1348
},
{
"epoch": 1.8410991636798089,
"grad_norm": 1.2451056241989136,
"learning_rate": 1.6382253096864217e-05,
"loss": 2.36,
"step": 1349
},
{
"epoch": 1.8424645844000682,
"grad_norm": 1.2524595260620117,
"learning_rate": 1.635707210379791e-05,
"loss": 2.3537,
"step": 1350
},
{
"epoch": 1.8438300051203278,
"grad_norm": 1.1536561250686646,
"learning_rate": 1.6331903351888295e-05,
"loss": 2.2428,
"step": 1351
},
{
"epoch": 1.8451954258405872,
"grad_norm": 1.1689746379852295,
"learning_rate": 1.630674689573373e-05,
"loss": 2.3641,
"step": 1352
},
{
"epoch": 1.8465608465608465,
"grad_norm": 1.2238224744796753,
"learning_rate": 1.628160278990591e-05,
"loss": 2.3178,
"step": 1353
},
{
"epoch": 1.8479262672811059,
"grad_norm": 1.2244027853012085,
"learning_rate": 1.6256471088949742e-05,
"loss": 2.2811,
"step": 1354
},
{
"epoch": 1.8492916880013655,
"grad_norm": 1.2538769245147705,
"learning_rate": 1.62313518473832e-05,
"loss": 2.3414,
"step": 1355
},
{
"epoch": 1.8506571087216248,
"grad_norm": 1.1916768550872803,
"learning_rate": 1.6206245119697262e-05,
"loss": 2.3556,
"step": 1356
},
{
"epoch": 1.8520225294418844,
"grad_norm": 1.2084734439849854,
"learning_rate": 1.6181150960355734e-05,
"loss": 2.3021,
"step": 1357
},
{
"epoch": 1.8533879501621437,
"grad_norm": 1.206769347190857,
"learning_rate": 1.615606942379517e-05,
"loss": 2.3629,
"step": 1358
},
{
"epoch": 1.854753370882403,
"grad_norm": 1.1988277435302734,
"learning_rate": 1.6131000564424747e-05,
"loss": 2.2833,
"step": 1359
},
{
"epoch": 1.8561187916026625,
"grad_norm": 1.2607680559158325,
"learning_rate": 1.6105944436626123e-05,
"loss": 2.4577,
"step": 1360
},
{
"epoch": 1.857484212322922,
"grad_norm": 1.1977181434631348,
"learning_rate": 1.6080901094753357e-05,
"loss": 2.3069,
"step": 1361
},
{
"epoch": 1.8588496330431814,
"grad_norm": 1.3004677295684814,
"learning_rate": 1.6055870593132756e-05,
"loss": 2.3781,
"step": 1362
},
{
"epoch": 1.860215053763441,
"grad_norm": 1.1765613555908203,
"learning_rate": 1.6030852986062782e-05,
"loss": 2.2055,
"step": 1363
},
{
"epoch": 1.8615804744837003,
"grad_norm": 1.21068274974823,
"learning_rate": 1.6005848327813924e-05,
"loss": 2.2742,
"step": 1364
},
{
"epoch": 1.8629458952039597,
"grad_norm": 1.1883211135864258,
"learning_rate": 1.5980856672628583e-05,
"loss": 2.3113,
"step": 1365
},
{
"epoch": 1.864311315924219,
"grad_norm": 1.2216191291809082,
"learning_rate": 1.5955878074720938e-05,
"loss": 2.3638,
"step": 1366
},
{
"epoch": 1.8656767366444784,
"grad_norm": 1.2610712051391602,
"learning_rate": 1.5930912588276863e-05,
"loss": 2.3281,
"step": 1367
},
{
"epoch": 1.867042157364738,
"grad_norm": 1.2435632944107056,
"learning_rate": 1.590596026745378e-05,
"loss": 2.4589,
"step": 1368
},
{
"epoch": 1.8684075780849976,
"grad_norm": 1.2494769096374512,
"learning_rate": 1.5881021166380547e-05,
"loss": 2.3411,
"step": 1369
},
{
"epoch": 1.869772998805257,
"grad_norm": 1.1486777067184448,
"learning_rate": 1.5856095339157357e-05,
"loss": 2.2008,
"step": 1370
},
{
"epoch": 1.8711384195255163,
"grad_norm": 1.2144020795822144,
"learning_rate": 1.58311828398556e-05,
"loss": 2.2796,
"step": 1371
},
{
"epoch": 1.8725038402457757,
"grad_norm": 1.185417890548706,
"learning_rate": 1.5806283722517747e-05,
"loss": 2.356,
"step": 1372
},
{
"epoch": 1.873869260966035,
"grad_norm": 1.2093297243118286,
"learning_rate": 1.5781398041157245e-05,
"loss": 2.3936,
"step": 1373
},
{
"epoch": 1.8752346816862946,
"grad_norm": 1.263375997543335,
"learning_rate": 1.5756525849758407e-05,
"loss": 2.3605,
"step": 1374
},
{
"epoch": 1.8766001024065542,
"grad_norm": 1.219164252281189,
"learning_rate": 1.573166720227626e-05,
"loss": 2.4565,
"step": 1375
},
{
"epoch": 1.8779655231268135,
"grad_norm": 1.250172734260559,
"learning_rate": 1.5706822152636472e-05,
"loss": 2.4952,
"step": 1376
},
{
"epoch": 1.8793309438470729,
"grad_norm": 1.2893790006637573,
"learning_rate": 1.56819907547352e-05,
"loss": 2.3845,
"step": 1377
},
{
"epoch": 1.8806963645673322,
"grad_norm": 1.2518771886825562,
"learning_rate": 1.5657173062438986e-05,
"loss": 2.273,
"step": 1378
},
{
"epoch": 1.8820617852875916,
"grad_norm": 1.2362000942230225,
"learning_rate": 1.5632369129584645e-05,
"loss": 2.3237,
"step": 1379
},
{
"epoch": 1.8834272060078512,
"grad_norm": 1.200493574142456,
"learning_rate": 1.5607579009979145e-05,
"loss": 2.3248,
"step": 1380
},
{
"epoch": 1.8847926267281108,
"grad_norm": 1.2331230640411377,
"learning_rate": 1.5582802757399485e-05,
"loss": 2.4362,
"step": 1381
},
{
"epoch": 1.8861580474483701,
"grad_norm": 1.2420063018798828,
"learning_rate": 1.5558040425592584e-05,
"loss": 2.2996,
"step": 1382
},
{
"epoch": 1.8875234681686295,
"grad_norm": 1.2456883192062378,
"learning_rate": 1.5533292068275173e-05,
"loss": 2.3489,
"step": 1383
},
{
"epoch": 1.8888888888888888,
"grad_norm": 1.2516740560531616,
"learning_rate": 1.5508557739133647e-05,
"loss": 2.3793,
"step": 1384
},
{
"epoch": 1.8902543096091482,
"grad_norm": 1.1862624883651733,
"learning_rate": 1.5483837491823975e-05,
"loss": 2.2961,
"step": 1385
},
{
"epoch": 1.8916197303294078,
"grad_norm": 1.2264900207519531,
"learning_rate": 1.545913137997159e-05,
"loss": 2.2731,
"step": 1386
},
{
"epoch": 1.8929851510496671,
"grad_norm": 1.2587826251983643,
"learning_rate": 1.5434439457171274e-05,
"loss": 2.3172,
"step": 1387
},
{
"epoch": 1.8943505717699267,
"grad_norm": 1.2544316053390503,
"learning_rate": 1.540976177698698e-05,
"loss": 2.416,
"step": 1388
},
{
"epoch": 1.895715992490186,
"grad_norm": 1.2008434534072876,
"learning_rate": 1.5385098392951826e-05,
"loss": 2.2583,
"step": 1389
},
{
"epoch": 1.8970814132104454,
"grad_norm": 1.2177056074142456,
"learning_rate": 1.5360449358567864e-05,
"loss": 2.2604,
"step": 1390
},
{
"epoch": 1.8984468339307048,
"grad_norm": 1.1929326057434082,
"learning_rate": 1.5335814727306044e-05,
"loss": 2.3991,
"step": 1391
},
{
"epoch": 1.8998122546509644,
"grad_norm": 1.241988182067871,
"learning_rate": 1.5311194552606067e-05,
"loss": 2.3787,
"step": 1392
},
{
"epoch": 1.9011776753712237,
"grad_norm": 1.2393487691879272,
"learning_rate": 1.528658888787628e-05,
"loss": 2.4142,
"step": 1393
},
{
"epoch": 1.9025430960914833,
"grad_norm": 1.2541463375091553,
"learning_rate": 1.5261997786493547e-05,
"loss": 2.3728,
"step": 1394
},
{
"epoch": 1.9039085168117427,
"grad_norm": 1.2482483386993408,
"learning_rate": 1.5237421301803136e-05,
"loss": 2.273,
"step": 1395
},
{
"epoch": 1.905273937532002,
"grad_norm": 1.24033784866333,
"learning_rate": 1.521285948711861e-05,
"loss": 2.2912,
"step": 1396
},
{
"epoch": 1.9066393582522614,
"grad_norm": 1.2042142152786255,
"learning_rate": 1.5188312395721712e-05,
"loss": 2.3003,
"step": 1397
},
{
"epoch": 1.9080047789725207,
"grad_norm": 1.2854644060134888,
"learning_rate": 1.5163780080862247e-05,
"loss": 2.3533,
"step": 1398
},
{
"epoch": 1.9093701996927803,
"grad_norm": 1.2115676403045654,
"learning_rate": 1.5139262595757954e-05,
"loss": 2.2281,
"step": 1399
},
{
"epoch": 1.91073562041304,
"grad_norm": 1.2456870079040527,
"learning_rate": 1.5114759993594419e-05,
"loss": 2.3682,
"step": 1400
},
{
"epoch": 1.9121010411332993,
"grad_norm": 1.21597421169281,
"learning_rate": 1.5090272327524935e-05,
"loss": 2.4307,
"step": 1401
},
{
"epoch": 1.9134664618535586,
"grad_norm": 1.2519721984863281,
"learning_rate": 1.5065799650670383e-05,
"loss": 2.3114,
"step": 1402
},
{
"epoch": 1.914831882573818,
"grad_norm": 1.2159708738327026,
"learning_rate": 1.5041342016119145e-05,
"loss": 2.3308,
"step": 1403
},
{
"epoch": 1.9161973032940773,
"grad_norm": 1.2429728507995605,
"learning_rate": 1.5016899476926962e-05,
"loss": 2.43,
"step": 1404
},
{
"epoch": 1.917562724014337,
"grad_norm": 1.204867959022522,
"learning_rate": 1.4992472086116828e-05,
"loss": 2.3426,
"step": 1405
},
{
"epoch": 1.9189281447345965,
"grad_norm": 1.230167269706726,
"learning_rate": 1.4968059896678898e-05,
"loss": 2.4001,
"step": 1406
},
{
"epoch": 1.9202935654548559,
"grad_norm": 1.263527750968933,
"learning_rate": 1.4943662961570304e-05,
"loss": 2.4552,
"step": 1407
},
{
"epoch": 1.9216589861751152,
"grad_norm": 1.2129266262054443,
"learning_rate": 1.4919281333715136e-05,
"loss": 2.3536,
"step": 1408
},
{
"epoch": 1.9230244068953746,
"grad_norm": 1.2071499824523926,
"learning_rate": 1.489491506600424e-05,
"loss": 2.2174,
"step": 1409
},
{
"epoch": 1.924389827615634,
"grad_norm": 1.2222583293914795,
"learning_rate": 1.487056421129518e-05,
"loss": 2.3559,
"step": 1410
},
{
"epoch": 1.9257552483358935,
"grad_norm": 1.197881817817688,
"learning_rate": 1.4846228822412046e-05,
"loss": 2.2971,
"step": 1411
},
{
"epoch": 1.927120669056153,
"grad_norm": 1.2255703210830688,
"learning_rate": 1.4821908952145409e-05,
"loss": 2.2738,
"step": 1412
},
{
"epoch": 1.9284860897764124,
"grad_norm": 1.229466199874878,
"learning_rate": 1.4797604653252142e-05,
"loss": 2.3474,
"step": 1413
},
{
"epoch": 1.9298515104966718,
"grad_norm": 1.2755507230758667,
"learning_rate": 1.477331597845538e-05,
"loss": 2.1397,
"step": 1414
},
{
"epoch": 1.9312169312169312,
"grad_norm": 1.191508412361145,
"learning_rate": 1.4749042980444332e-05,
"loss": 2.2721,
"step": 1415
},
{
"epoch": 1.9325823519371905,
"grad_norm": 1.2486668825149536,
"learning_rate": 1.4724785711874208e-05,
"loss": 2.4176,
"step": 1416
},
{
"epoch": 1.93394777265745,
"grad_norm": 1.2215373516082764,
"learning_rate": 1.4700544225366114e-05,
"loss": 2.3141,
"step": 1417
},
{
"epoch": 1.9353131933777095,
"grad_norm": 1.2214670181274414,
"learning_rate": 1.4676318573506896e-05,
"loss": 2.3555,
"step": 1418
},
{
"epoch": 1.936678614097969,
"grad_norm": 1.2122772932052612,
"learning_rate": 1.465210880884906e-05,
"loss": 2.2284,
"step": 1419
},
{
"epoch": 1.9380440348182284,
"grad_norm": 1.1815413236618042,
"learning_rate": 1.4627914983910643e-05,
"loss": 2.4631,
"step": 1420
},
{
"epoch": 1.9394094555384878,
"grad_norm": 1.2228288650512695,
"learning_rate": 1.4603737151175122e-05,
"loss": 2.3494,
"step": 1421
},
{
"epoch": 1.9407748762587471,
"grad_norm": 1.2137821912765503,
"learning_rate": 1.4579575363091252e-05,
"loss": 2.2671,
"step": 1422
},
{
"epoch": 1.9421402969790067,
"grad_norm": 1.2307125329971313,
"learning_rate": 1.4555429672073014e-05,
"loss": 2.4963,
"step": 1423
},
{
"epoch": 1.943505717699266,
"grad_norm": 1.272155523300171,
"learning_rate": 1.4531300130499455e-05,
"loss": 2.3605,
"step": 1424
},
{
"epoch": 1.9448711384195256,
"grad_norm": 1.1575047969818115,
"learning_rate": 1.4507186790714581e-05,
"loss": 2.3041,
"step": 1425
},
{
"epoch": 1.946236559139785,
"grad_norm": 1.2062952518463135,
"learning_rate": 1.4483089705027267e-05,
"loss": 2.3556,
"step": 1426
},
{
"epoch": 1.9476019798600444,
"grad_norm": 1.2023831605911255,
"learning_rate": 1.4459008925711104e-05,
"loss": 2.2748,
"step": 1427
},
{
"epoch": 1.9489674005803037,
"grad_norm": 1.2403326034545898,
"learning_rate": 1.4434944505004355e-05,
"loss": 2.4439,
"step": 1428
},
{
"epoch": 1.9503328213005633,
"grad_norm": 1.192940354347229,
"learning_rate": 1.4410896495109752e-05,
"loss": 2.3165,
"step": 1429
},
{
"epoch": 1.9516982420208226,
"grad_norm": 1.2951130867004395,
"learning_rate": 1.4386864948194447e-05,
"loss": 2.422,
"step": 1430
},
{
"epoch": 1.9530636627410822,
"grad_norm": 1.187801480293274,
"learning_rate": 1.4362849916389867e-05,
"loss": 2.2974,
"step": 1431
},
{
"epoch": 1.9544290834613416,
"grad_norm": 1.2528104782104492,
"learning_rate": 1.4338851451791631e-05,
"loss": 2.4501,
"step": 1432
},
{
"epoch": 1.955794504181601,
"grad_norm": 1.199501633644104,
"learning_rate": 1.4314869606459408e-05,
"loss": 2.3958,
"step": 1433
},
{
"epoch": 1.9571599249018603,
"grad_norm": 1.2060744762420654,
"learning_rate": 1.4290904432416813e-05,
"loss": 2.2625,
"step": 1434
},
{
"epoch": 1.9585253456221197,
"grad_norm": 1.2410277128219604,
"learning_rate": 1.42669559816513e-05,
"loss": 2.3214,
"step": 1435
},
{
"epoch": 1.9598907663423792,
"grad_norm": 1.202832818031311,
"learning_rate": 1.4243024306114048e-05,
"loss": 2.3188,
"step": 1436
},
{
"epoch": 1.9612561870626388,
"grad_norm": 1.2308259010314941,
"learning_rate": 1.4219109457719839e-05,
"loss": 2.3961,
"step": 1437
},
{
"epoch": 1.9626216077828982,
"grad_norm": 1.2404173612594604,
"learning_rate": 1.4195211488346953e-05,
"loss": 2.2899,
"step": 1438
},
{
"epoch": 1.9639870285031575,
"grad_norm": 1.1884961128234863,
"learning_rate": 1.4171330449837063e-05,
"loss": 2.373,
"step": 1439
},
{
"epoch": 1.965352449223417,
"grad_norm": 1.2462489604949951,
"learning_rate": 1.4147466393995109e-05,
"loss": 2.3116,
"step": 1440
},
{
"epoch": 1.9667178699436763,
"grad_norm": 1.205567479133606,
"learning_rate": 1.412361937258917e-05,
"loss": 2.2824,
"step": 1441
},
{
"epoch": 1.9680832906639358,
"grad_norm": 1.199938178062439,
"learning_rate": 1.4099789437350424e-05,
"loss": 2.3717,
"step": 1442
},
{
"epoch": 1.9694487113841954,
"grad_norm": 1.3135393857955933,
"learning_rate": 1.4075976639972928e-05,
"loss": 2.2107,
"step": 1443
},
{
"epoch": 1.9708141321044548,
"grad_norm": 1.2465384006500244,
"learning_rate": 1.4052181032113596e-05,
"loss": 2.4146,
"step": 1444
},
{
"epoch": 1.9721795528247141,
"grad_norm": 1.231345772743225,
"learning_rate": 1.4028402665392027e-05,
"loss": 2.2268,
"step": 1445
},
{
"epoch": 1.9735449735449735,
"grad_norm": 1.3131641149520874,
"learning_rate": 1.4004641591390455e-05,
"loss": 2.4308,
"step": 1446
},
{
"epoch": 1.9749103942652328,
"grad_norm": 1.2397103309631348,
"learning_rate": 1.3980897861653575e-05,
"loss": 2.304,
"step": 1447
},
{
"epoch": 1.9762758149854924,
"grad_norm": 1.2151758670806885,
"learning_rate": 1.3957171527688466e-05,
"loss": 2.2748,
"step": 1448
},
{
"epoch": 1.9776412357057518,
"grad_norm": 1.2326351404190063,
"learning_rate": 1.393346264096445e-05,
"loss": 2.4372,
"step": 1449
},
{
"epoch": 1.9790066564260114,
"grad_norm": 1.223924994468689,
"learning_rate": 1.3909771252913038e-05,
"loss": 2.3241,
"step": 1450
},
{
"epoch": 1.9803720771462707,
"grad_norm": 1.2367944717407227,
"learning_rate": 1.3886097414927754e-05,
"loss": 2.4258,
"step": 1451
},
{
"epoch": 1.98173749786653,
"grad_norm": 1.2360970973968506,
"learning_rate": 1.386244117836406e-05,
"loss": 2.3976,
"step": 1452
},
{
"epoch": 1.9831029185867894,
"grad_norm": 1.206300973892212,
"learning_rate": 1.3838802594539228e-05,
"loss": 2.3041,
"step": 1453
},
{
"epoch": 1.984468339307049,
"grad_norm": 1.1823267936706543,
"learning_rate": 1.3815181714732251e-05,
"loss": 2.3594,
"step": 1454
},
{
"epoch": 1.9858337600273084,
"grad_norm": 1.2500476837158203,
"learning_rate": 1.3791578590183701e-05,
"loss": 2.4902,
"step": 1455
},
{
"epoch": 1.987199180747568,
"grad_norm": 1.2190042734146118,
"learning_rate": 1.3767993272095643e-05,
"loss": 2.4029,
"step": 1456
},
{
"epoch": 1.9885646014678273,
"grad_norm": 1.2220433950424194,
"learning_rate": 1.3744425811631512e-05,
"loss": 2.4227,
"step": 1457
},
{
"epoch": 1.9899300221880867,
"grad_norm": 1.18170964717865,
"learning_rate": 1.3720876259915996e-05,
"loss": 2.2376,
"step": 1458
},
{
"epoch": 1.991295442908346,
"grad_norm": 1.2125738859176636,
"learning_rate": 1.3697344668034967e-05,
"loss": 2.4209,
"step": 1459
},
{
"epoch": 1.9926608636286056,
"grad_norm": 1.2430468797683716,
"learning_rate": 1.3673831087035286e-05,
"loss": 2.3765,
"step": 1460
},
{
"epoch": 1.994026284348865,
"grad_norm": 1.2244330644607544,
"learning_rate": 1.3650335567924779e-05,
"loss": 2.2975,
"step": 1461
},
{
"epoch": 1.9953917050691246,
"grad_norm": 1.2216966152191162,
"learning_rate": 1.3626858161672074e-05,
"loss": 2.3473,
"step": 1462
},
{
"epoch": 1.996757125789384,
"grad_norm": 1.2209025621414185,
"learning_rate": 1.3603398919206523e-05,
"loss": 2.3286,
"step": 1463
},
{
"epoch": 1.9981225465096433,
"grad_norm": 1.1662893295288086,
"learning_rate": 1.3579957891418068e-05,
"loss": 2.2821,
"step": 1464
},
{
"epoch": 1.9994879672299026,
"grad_norm": 1.210418939590454,
"learning_rate": 1.3556535129157138e-05,
"loss": 2.2984,
"step": 1465
},
{
"epoch": 2.0,
"grad_norm": 1.210418939590454,
"learning_rate": 1.3533130683234515e-05,
"loss": 0.8701,
"step": 1466
},
{
"epoch": 2.0013654207202594,
"grad_norm": 1.146583080291748,
"learning_rate": 1.3509744604421289e-05,
"loss": 2.0388,
"step": 1467
},
{
"epoch": 2.0027308414405187,
"grad_norm": 1.2640188932418823,
"learning_rate": 1.3486376943448686e-05,
"loss": 1.9709,
"step": 1468
},
{
"epoch": 2.004096262160778,
"grad_norm": 1.2100741863250732,
"learning_rate": 1.3463027751007975e-05,
"loss": 1.9241,
"step": 1469
},
{
"epoch": 2.005461682881038,
"grad_norm": 1.1397905349731445,
"learning_rate": 1.3439697077750367e-05,
"loss": 2.1248,
"step": 1470
},
{
"epoch": 2.0068271036012972,
"grad_norm": 1.1227343082427979,
"learning_rate": 1.3416384974286895e-05,
"loss": 2.0604,
"step": 1471
},
{
"epoch": 2.0081925243215566,
"grad_norm": 1.0937353372573853,
"learning_rate": 1.3393091491188317e-05,
"loss": 1.9704,
"step": 1472
},
{
"epoch": 2.009557945041816,
"grad_norm": 1.1105605363845825,
"learning_rate": 1.3369816678984986e-05,
"loss": 1.9982,
"step": 1473
},
{
"epoch": 2.0109233657620753,
"grad_norm": 1.1038012504577637,
"learning_rate": 1.3346560588166764e-05,
"loss": 1.9851,
"step": 1474
},
{
"epoch": 2.0122887864823347,
"grad_norm": 1.149288296699524,
"learning_rate": 1.3323323269182891e-05,
"loss": 1.8834,
"step": 1475
},
{
"epoch": 2.0136542072025945,
"grad_norm": 1.2271548509597778,
"learning_rate": 1.3300104772441897e-05,
"loss": 2.0603,
"step": 1476
},
{
"epoch": 2.015019627922854,
"grad_norm": 1.2477668523788452,
"learning_rate": 1.3276905148311481e-05,
"loss": 1.953,
"step": 1477
},
{
"epoch": 2.016385048643113,
"grad_norm": 1.2525019645690918,
"learning_rate": 1.325372444711838e-05,
"loss": 1.8891,
"step": 1478
},
{
"epoch": 2.0177504693633725,
"grad_norm": 1.2048678398132324,
"learning_rate": 1.3230562719148292e-05,
"loss": 2.02,
"step": 1479
},
{
"epoch": 2.019115890083632,
"grad_norm": 1.2122950553894043,
"learning_rate": 1.320742001464578e-05,
"loss": 1.94,
"step": 1480
},
{
"epoch": 2.0204813108038913,
"grad_norm": 1.1750199794769287,
"learning_rate": 1.3184296383814114e-05,
"loss": 1.9207,
"step": 1481
},
{
"epoch": 2.021846731524151,
"grad_norm": 1.208469033241272,
"learning_rate": 1.3161191876815194e-05,
"loss": 1.9812,
"step": 1482
},
{
"epoch": 2.0232121522444104,
"grad_norm": 1.1509740352630615,
"learning_rate": 1.3138106543769433e-05,
"loss": 2.0222,
"step": 1483
},
{
"epoch": 2.02457757296467,
"grad_norm": 1.1613163948059082,
"learning_rate": 1.3115040434755656e-05,
"loss": 2.0043,
"step": 1484
},
{
"epoch": 2.025942993684929,
"grad_norm": 1.1147252321243286,
"learning_rate": 1.309199359981098e-05,
"loss": 1.997,
"step": 1485
},
{
"epoch": 2.0273084144051885,
"grad_norm": 1.159620761871338,
"learning_rate": 1.3068966088930713e-05,
"loss": 2.008,
"step": 1486
},
{
"epoch": 2.028673835125448,
"grad_norm": 1.1480010747909546,
"learning_rate": 1.3045957952068242e-05,
"loss": 2.0153,
"step": 1487
},
{
"epoch": 2.0300392558457077,
"grad_norm": 1.1194638013839722,
"learning_rate": 1.3022969239134927e-05,
"loss": 1.9913,
"step": 1488
},
{
"epoch": 2.031404676565967,
"grad_norm": 1.1164032220840454,
"learning_rate": 1.3000000000000006e-05,
"loss": 2.0751,
"step": 1489
},
{
"epoch": 2.0327700972862264,
"grad_norm": 1.1992695331573486,
"learning_rate": 1.297705028449044e-05,
"loss": 2.0849,
"step": 1490
},
{
"epoch": 2.0341355180064857,
"grad_norm": 1.1376081705093384,
"learning_rate": 1.2954120142390863e-05,
"loss": 1.9686,
"step": 1491
},
{
"epoch": 2.035500938726745,
"grad_norm": 1.1503387689590454,
"learning_rate": 1.293120962344344e-05,
"loss": 1.9227,
"step": 1492
},
{
"epoch": 2.0368663594470044,
"grad_norm": 1.1579762697219849,
"learning_rate": 1.2908318777347778e-05,
"loss": 2.0542,
"step": 1493
},
{
"epoch": 2.0382317801672643,
"grad_norm": 1.1448447704315186,
"learning_rate": 1.2885447653760804e-05,
"loss": 2.0296,
"step": 1494
},
{
"epoch": 2.0395972008875236,
"grad_norm": 1.1701680421829224,
"learning_rate": 1.2862596302296635e-05,
"loss": 1.8718,
"step": 1495
},
{
"epoch": 2.040962621607783,
"grad_norm": 1.1812020540237427,
"learning_rate": 1.2839764772526527e-05,
"loss": 1.9208,
"step": 1496
},
{
"epoch": 2.0423280423280423,
"grad_norm": 1.1286402940750122,
"learning_rate": 1.2816953113978738e-05,
"loss": 1.9031,
"step": 1497
},
{
"epoch": 2.0436934630483017,
"grad_norm": 1.1178324222564697,
"learning_rate": 1.2794161376138403e-05,
"loss": 1.9357,
"step": 1498
},
{
"epoch": 2.045058883768561,
"grad_norm": 1.129090428352356,
"learning_rate": 1.2771389608447452e-05,
"loss": 1.9072,
"step": 1499
},
{
"epoch": 2.0464243044888204,
"grad_norm": 1.097334861755371,
"learning_rate": 1.2748637860304486e-05,
"loss": 1.9097,
"step": 1500
},
{
"epoch": 2.04778972520908,
"grad_norm": 1.1420358419418335,
"learning_rate": 1.2725906181064693e-05,
"loss": 1.9709,
"step": 1501
},
{
"epoch": 2.0491551459293396,
"grad_norm": 1.1314576864242554,
"learning_rate": 1.2703194620039706e-05,
"loss": 2.0499,
"step": 1502
},
{
"epoch": 2.050520566649599,
"grad_norm": 1.142789363861084,
"learning_rate": 1.2680503226497533e-05,
"loss": 2.0436,
"step": 1503
},
{
"epoch": 2.0518859873698583,
"grad_norm": 1.1338099241256714,
"learning_rate": 1.2657832049662426e-05,
"loss": 1.9189,
"step": 1504
},
{
"epoch": 2.0532514080901176,
"grad_norm": 1.1203646659851074,
"learning_rate": 1.2635181138714771e-05,
"loss": 1.9004,
"step": 1505
},
{
"epoch": 2.054616828810377,
"grad_norm": 1.1215237379074097,
"learning_rate": 1.261255054279102e-05,
"loss": 1.9765,
"step": 1506
},
{
"epoch": 2.055982249530637,
"grad_norm": 1.1525779962539673,
"learning_rate": 1.2589940310983519e-05,
"loss": 2.0145,
"step": 1507
},
{
"epoch": 2.057347670250896,
"grad_norm": 1.0830928087234497,
"learning_rate": 1.2567350492340466e-05,
"loss": 1.9518,
"step": 1508
},
{
"epoch": 2.0587130909711555,
"grad_norm": 1.0851759910583496,
"learning_rate": 1.2544781135865755e-05,
"loss": 1.9049,
"step": 1509
},
{
"epoch": 2.060078511691415,
"grad_norm": 1.1207575798034668,
"learning_rate": 1.2522232290518924e-05,
"loss": 1.9779,
"step": 1510
},
{
"epoch": 2.0614439324116742,
"grad_norm": 1.1156715154647827,
"learning_rate": 1.2499704005214983e-05,
"loss": 2.1131,
"step": 1511
},
{
"epoch": 2.0628093531319336,
"grad_norm": 1.1843677759170532,
"learning_rate": 1.2477196328824364e-05,
"loss": 2.0288,
"step": 1512
},
{
"epoch": 2.0641747738521934,
"grad_norm": 1.1832064390182495,
"learning_rate": 1.2454709310172763e-05,
"loss": 1.9656,
"step": 1513
},
{
"epoch": 2.0655401945724527,
"grad_norm": 1.150840163230896,
"learning_rate": 1.2432242998041101e-05,
"loss": 1.9899,
"step": 1514
},
{
"epoch": 2.066905615292712,
"grad_norm": 1.155748724937439,
"learning_rate": 1.2409797441165363e-05,
"loss": 1.9656,
"step": 1515
},
{
"epoch": 2.0682710360129715,
"grad_norm": 1.1391583681106567,
"learning_rate": 1.2387372688236499e-05,
"loss": 2.0434,
"step": 1516
},
{
"epoch": 2.069636456733231,
"grad_norm": 1.1313402652740479,
"learning_rate": 1.2364968787900344e-05,
"loss": 1.9569,
"step": 1517
},
{
"epoch": 2.07100187745349,
"grad_norm": 1.1111037731170654,
"learning_rate": 1.2342585788757492e-05,
"loss": 1.8981,
"step": 1518
},
{
"epoch": 2.07236729817375,
"grad_norm": 1.1136736869812012,
"learning_rate": 1.2320223739363204e-05,
"loss": 1.9177,
"step": 1519
},
{
"epoch": 2.0737327188940093,
"grad_norm": 1.122404932975769,
"learning_rate": 1.229788268822728e-05,
"loss": 2.0029,
"step": 1520
},
{
"epoch": 2.0750981396142687,
"grad_norm": 1.1285018920898438,
"learning_rate": 1.2275562683813984e-05,
"loss": 2.0271,
"step": 1521
},
{
"epoch": 2.076463560334528,
"grad_norm": 1.151413917541504,
"learning_rate": 1.2253263774541901e-05,
"loss": 2.0136,
"step": 1522
},
{
"epoch": 2.0778289810547874,
"grad_norm": 1.1226527690887451,
"learning_rate": 1.2230986008783891e-05,
"loss": 1.9649,
"step": 1523
},
{
"epoch": 2.0791944017750468,
"grad_norm": 1.1200917959213257,
"learning_rate": 1.2208729434866923e-05,
"loss": 1.9129,
"step": 1524
},
{
"epoch": 2.0805598224953066,
"grad_norm": 1.1146423816680908,
"learning_rate": 1.218649410107199e-05,
"loss": 2.0204,
"step": 1525
},
{
"epoch": 2.081925243215566,
"grad_norm": 1.138129711151123,
"learning_rate": 1.216428005563401e-05,
"loss": 2.0263,
"step": 1526
},
{
"epoch": 2.0832906639358253,
"grad_norm": 1.1686633825302124,
"learning_rate": 1.2142087346741742e-05,
"loss": 1.9263,
"step": 1527
},
{
"epoch": 2.0846560846560847,
"grad_norm": 1.1207084655761719,
"learning_rate": 1.2119916022537644e-05,
"loss": 2.0608,
"step": 1528
},
{
"epoch": 2.086021505376344,
"grad_norm": 1.1394667625427246,
"learning_rate": 1.2097766131117781e-05,
"loss": 2.0945,
"step": 1529
},
{
"epoch": 2.0873869260966034,
"grad_norm": 1.2030839920043945,
"learning_rate": 1.207563772053173e-05,
"loss": 2.0958,
"step": 1530
},
{
"epoch": 2.0887523468168627,
"grad_norm": 1.1938525438308716,
"learning_rate": 1.2053530838782468e-05,
"loss": 2.0289,
"step": 1531
},
{
"epoch": 2.0901177675371225,
"grad_norm": 1.1356427669525146,
"learning_rate": 1.203144553382627e-05,
"loss": 2.003,
"step": 1532
},
{
"epoch": 2.091483188257382,
"grad_norm": 1.1546865701675415,
"learning_rate": 1.2009381853572608e-05,
"loss": 2.0054,
"step": 1533
},
{
"epoch": 2.0928486089776412,
"grad_norm": 1.140323519706726,
"learning_rate": 1.1987339845884036e-05,
"loss": 2.0093,
"step": 1534
},
{
"epoch": 2.0942140296979006,
"grad_norm": 1.156976580619812,
"learning_rate": 1.1965319558576103e-05,
"loss": 2.0344,
"step": 1535
},
{
"epoch": 2.09557945041816,
"grad_norm": 1.2063413858413696,
"learning_rate": 1.1943321039417234e-05,
"loss": 1.9337,
"step": 1536
},
{
"epoch": 2.0969448711384193,
"grad_norm": 1.1303611993789673,
"learning_rate": 1.1921344336128633e-05,
"loss": 2.0799,
"step": 1537
},
{
"epoch": 2.098310291858679,
"grad_norm": 1.186185359954834,
"learning_rate": 1.189938949638418e-05,
"loss": 1.9715,
"step": 1538
},
{
"epoch": 2.0996757125789385,
"grad_norm": 1.1383761167526245,
"learning_rate": 1.187745656781032e-05,
"loss": 1.9733,
"step": 1539
},
{
"epoch": 2.101041133299198,
"grad_norm": 1.1534337997436523,
"learning_rate": 1.1855545597985989e-05,
"loss": 2.0453,
"step": 1540
},
{
"epoch": 2.102406554019457,
"grad_norm": 1.164361834526062,
"learning_rate": 1.183365663444247e-05,
"loss": 1.9914,
"step": 1541
},
{
"epoch": 2.1037719747397166,
"grad_norm": 1.146592378616333,
"learning_rate": 1.1811789724663298e-05,
"loss": 2.0094,
"step": 1542
},
{
"epoch": 2.105137395459976,
"grad_norm": 1.1418884992599487,
"learning_rate": 1.1789944916084177e-05,
"loss": 1.926,
"step": 1543
},
{
"epoch": 2.1065028161802357,
"grad_norm": 1.1620937585830688,
"learning_rate": 1.1768122256092887e-05,
"loss": 1.9724,
"step": 1544
},
{
"epoch": 2.107868236900495,
"grad_norm": 1.1262433528900146,
"learning_rate": 1.174632179202913e-05,
"loss": 2.0638,
"step": 1545
},
{
"epoch": 2.1092336576207544,
"grad_norm": 1.137506365776062,
"learning_rate": 1.1724543571184476e-05,
"loss": 2.0084,
"step": 1546
},
{
"epoch": 2.110599078341014,
"grad_norm": 1.1238553524017334,
"learning_rate": 1.1702787640802238e-05,
"loss": 1.9462,
"step": 1547
},
{
"epoch": 2.111964499061273,
"grad_norm": 1.2142099142074585,
"learning_rate": 1.1681054048077378e-05,
"loss": 2.1155,
"step": 1548
},
{
"epoch": 2.1133299197815325,
"grad_norm": 1.1988736391067505,
"learning_rate": 1.1659342840156394e-05,
"loss": 1.9976,
"step": 1549
},
{
"epoch": 2.1146953405017923,
"grad_norm": 1.1824698448181152,
"learning_rate": 1.163765406413723e-05,
"loss": 2.0702,
"step": 1550
},
{
"epoch": 2.1160607612220517,
"grad_norm": 1.127030372619629,
"learning_rate": 1.1615987767069173e-05,
"loss": 1.9057,
"step": 1551
},
{
"epoch": 2.117426181942311,
"grad_norm": 1.1208933591842651,
"learning_rate": 1.1594343995952728e-05,
"loss": 2.1173,
"step": 1552
},
{
"epoch": 2.1187916026625704,
"grad_norm": 1.2297381162643433,
"learning_rate": 1.1572722797739572e-05,
"loss": 2.0815,
"step": 1553
},
{
"epoch": 2.1201570233828297,
"grad_norm": 1.1967939138412476,
"learning_rate": 1.1551124219332367e-05,
"loss": 1.954,
"step": 1554
},
{
"epoch": 2.121522444103089,
"grad_norm": 1.1408159732818604,
"learning_rate": 1.1529548307584748e-05,
"loss": 2.0679,
"step": 1555
},
{
"epoch": 2.122887864823349,
"grad_norm": 1.1509073972702026,
"learning_rate": 1.150799510930114e-05,
"loss": 1.9662,
"step": 1556
},
{
"epoch": 2.1242532855436083,
"grad_norm": 1.172161340713501,
"learning_rate": 1.1486464671236738e-05,
"loss": 1.9397,
"step": 1557
},
{
"epoch": 2.1256187062638676,
"grad_norm": 1.1243789196014404,
"learning_rate": 1.1464957040097337e-05,
"loss": 1.9028,
"step": 1558
},
{
"epoch": 2.126984126984127,
"grad_norm": 1.0887666940689087,
"learning_rate": 1.144347226253927e-05,
"loss": 1.936,
"step": 1559
},
{
"epoch": 2.1283495477043863,
"grad_norm": 1.0996894836425781,
"learning_rate": 1.1422010385169262e-05,
"loss": 1.9489,
"step": 1560
},
{
"epoch": 2.1297149684246457,
"grad_norm": 1.105038046836853,
"learning_rate": 1.1400571454544412e-05,
"loss": 2.0866,
"step": 1561
},
{
"epoch": 2.1310803891449055,
"grad_norm": 1.209944725036621,
"learning_rate": 1.1379155517172e-05,
"loss": 1.9113,
"step": 1562
},
{
"epoch": 2.132445809865165,
"grad_norm": 1.1354070901870728,
"learning_rate": 1.1357762619509445e-05,
"loss": 1.9841,
"step": 1563
},
{
"epoch": 2.133811230585424,
"grad_norm": 1.154759407043457,
"learning_rate": 1.1336392807964185e-05,
"loss": 1.8951,
"step": 1564
},
{
"epoch": 2.1351766513056836,
"grad_norm": 1.1262272596359253,
"learning_rate": 1.1315046128893571e-05,
"loss": 2.0162,
"step": 1565
},
{
"epoch": 2.136542072025943,
"grad_norm": 1.1389347314834595,
"learning_rate": 1.1293722628604782e-05,
"loss": 2.0102,
"step": 1566
},
{
"epoch": 2.1379074927462023,
"grad_norm": 1.1754345893859863,
"learning_rate": 1.1272422353354706e-05,
"loss": 2.0082,
"step": 1567
},
{
"epoch": 2.1392729134664616,
"grad_norm": 1.1821562051773071,
"learning_rate": 1.1251145349349855e-05,
"loss": 2.043,
"step": 1568
},
{
"epoch": 2.1406383341867214,
"grad_norm": 1.1311756372451782,
"learning_rate": 1.1229891662746248e-05,
"loss": 1.9931,
"step": 1569
},
{
"epoch": 2.142003754906981,
"grad_norm": 1.1404931545257568,
"learning_rate": 1.1208661339649348e-05,
"loss": 1.9649,
"step": 1570
},
{
"epoch": 2.14336917562724,
"grad_norm": 1.1803452968597412,
"learning_rate": 1.1187454426113917e-05,
"loss": 2.0332,
"step": 1571
},
{
"epoch": 2.1447345963474995,
"grad_norm": 1.1680721044540405,
"learning_rate": 1.1166270968143923e-05,
"loss": 1.9419,
"step": 1572
},
{
"epoch": 2.146100017067759,
"grad_norm": 1.130502700805664,
"learning_rate": 1.1145111011692471e-05,
"loss": 1.9237,
"step": 1573
},
{
"epoch": 2.1474654377880182,
"grad_norm": 1.1211011409759521,
"learning_rate": 1.1123974602661685e-05,
"loss": 2.0883,
"step": 1574
},
{
"epoch": 2.148830858508278,
"grad_norm": 1.2083908319473267,
"learning_rate": 1.1102861786902595e-05,
"loss": 1.9568,
"step": 1575
},
{
"epoch": 2.1501962792285374,
"grad_norm": 1.125117301940918,
"learning_rate": 1.108177261021507e-05,
"loss": 1.9609,
"step": 1576
},
{
"epoch": 2.1515616999487968,
"grad_norm": 1.1487423181533813,
"learning_rate": 1.1060707118347662e-05,
"loss": 2.0161,
"step": 1577
},
{
"epoch": 2.152927120669056,
"grad_norm": 1.1642670631408691,
"learning_rate": 1.103966535699759e-05,
"loss": 1.9868,
"step": 1578
},
{
"epoch": 2.1542925413893155,
"grad_norm": 1.1711370944976807,
"learning_rate": 1.1018647371810562e-05,
"loss": 1.9635,
"step": 1579
},
{
"epoch": 2.155657962109575,
"grad_norm": 1.2143476009368896,
"learning_rate": 1.0997653208380723e-05,
"loss": 2.0522,
"step": 1580
},
{
"epoch": 2.1570233828298346,
"grad_norm": 1.190935492515564,
"learning_rate": 1.0976682912250537e-05,
"loss": 1.8711,
"step": 1581
},
{
"epoch": 2.158388803550094,
"grad_norm": 1.121225118637085,
"learning_rate": 1.0955736528910692e-05,
"loss": 1.9589,
"step": 1582
},
{
"epoch": 2.1597542242703534,
"grad_norm": 1.1459887027740479,
"learning_rate": 1.0934814103800004e-05,
"loss": 2.0435,
"step": 1583
},
{
"epoch": 2.1611196449906127,
"grad_norm": 1.1818052530288696,
"learning_rate": 1.0913915682305319e-05,
"loss": 2.1193,
"step": 1584
},
{
"epoch": 2.162485065710872,
"grad_norm": 1.1633687019348145,
"learning_rate": 1.0893041309761403e-05,
"loss": 1.8864,
"step": 1585
},
{
"epoch": 2.1638504864311314,
"grad_norm": 1.1369997262954712,
"learning_rate": 1.0872191031450861e-05,
"loss": 1.9937,
"step": 1586
},
{
"epoch": 2.1652159071513912,
"grad_norm": 1.1436717510223389,
"learning_rate": 1.0851364892604032e-05,
"loss": 1.9842,
"step": 1587
},
{
"epoch": 2.1665813278716506,
"grad_norm": 1.1580510139465332,
"learning_rate": 1.0830562938398893e-05,
"loss": 1.9535,
"step": 1588
},
{
"epoch": 2.16794674859191,
"grad_norm": 1.1783857345581055,
"learning_rate": 1.080978521396094e-05,
"loss": 2.022,
"step": 1589
},
{
"epoch": 2.1693121693121693,
"grad_norm": 1.2002201080322266,
"learning_rate": 1.0789031764363114e-05,
"loss": 1.9589,
"step": 1590
},
{
"epoch": 2.1706775900324287,
"grad_norm": 1.1553797721862793,
"learning_rate": 1.0768302634625716e-05,
"loss": 2.0492,
"step": 1591
},
{
"epoch": 2.172043010752688,
"grad_norm": 1.138646125793457,
"learning_rate": 1.0747597869716265e-05,
"loss": 1.9955,
"step": 1592
},
{
"epoch": 2.1734084314729474,
"grad_norm": 1.1653780937194824,
"learning_rate": 1.072691751454944e-05,
"loss": 2.0341,
"step": 1593
},
{
"epoch": 2.174773852193207,
"grad_norm": 1.188685655593872,
"learning_rate": 1.0706261613986961e-05,
"loss": 2.0317,
"step": 1594
},
{
"epoch": 2.1761392729134665,
"grad_norm": 1.1902343034744263,
"learning_rate": 1.0685630212837503e-05,
"loss": 2.0087,
"step": 1595
},
{
"epoch": 2.177504693633726,
"grad_norm": 1.166034460067749,
"learning_rate": 1.0665023355856596e-05,
"loss": 2.0042,
"step": 1596
},
{
"epoch": 2.1788701143539853,
"grad_norm": 1.1893069744110107,
"learning_rate": 1.064444108774652e-05,
"loss": 2.0491,
"step": 1597
},
{
"epoch": 2.1802355350742446,
"grad_norm": 1.1973901987075806,
"learning_rate": 1.0623883453156218e-05,
"loss": 2.1741,
"step": 1598
},
{
"epoch": 2.1816009557945044,
"grad_norm": 1.190361499786377,
"learning_rate": 1.0603350496681197e-05,
"loss": 1.9309,
"step": 1599
},
{
"epoch": 2.1829663765147638,
"grad_norm": 1.161087155342102,
"learning_rate": 1.0582842262863442e-05,
"loss": 1.9948,
"step": 1600
},
{
"epoch": 2.184331797235023,
"grad_norm": 1.1410590410232544,
"learning_rate": 1.0562358796191274e-05,
"loss": 1.9834,
"step": 1601
},
{
"epoch": 2.1856972179552825,
"grad_norm": 1.1307111978530884,
"learning_rate": 1.0541900141099319e-05,
"loss": 2.0245,
"step": 1602
},
{
"epoch": 2.187062638675542,
"grad_norm": 1.1609324216842651,
"learning_rate": 1.052146634196836e-05,
"loss": 1.8932,
"step": 1603
},
{
"epoch": 2.188428059395801,
"grad_norm": 1.1601325273513794,
"learning_rate": 1.0501057443125277e-05,
"loss": 1.9055,
"step": 1604
},
{
"epoch": 2.1897934801160606,
"grad_norm": 1.1565388441085815,
"learning_rate": 1.0480673488842923e-05,
"loss": 2.1283,
"step": 1605
},
{
"epoch": 2.1911589008363204,
"grad_norm": 1.1876124143600464,
"learning_rate": 1.0460314523340045e-05,
"loss": 1.8878,
"step": 1606
},
{
"epoch": 2.1925243215565797,
"grad_norm": 1.1314270496368408,
"learning_rate": 1.0439980590781161e-05,
"loss": 1.9657,
"step": 1607
},
{
"epoch": 2.193889742276839,
"grad_norm": 1.1396403312683105,
"learning_rate": 1.0419671735276518e-05,
"loss": 2.0119,
"step": 1608
},
{
"epoch": 2.1952551629970984,
"grad_norm": 1.2485142946243286,
"learning_rate": 1.0399388000881942e-05,
"loss": 2.0265,
"step": 1609
},
{
"epoch": 2.196620583717358,
"grad_norm": 1.2038015127182007,
"learning_rate": 1.0379129431598775e-05,
"loss": 2.0171,
"step": 1610
},
{
"epoch": 2.197986004437617,
"grad_norm": 1.1367573738098145,
"learning_rate": 1.0358896071373758e-05,
"loss": 2.0448,
"step": 1611
},
{
"epoch": 2.199351425157877,
"grad_norm": 1.1328151226043701,
"learning_rate": 1.0338687964098953e-05,
"loss": 2.0014,
"step": 1612
},
{
"epoch": 2.2007168458781363,
"grad_norm": 1.150307297706604,
"learning_rate": 1.0318505153611642e-05,
"loss": 1.9455,
"step": 1613
},
{
"epoch": 2.2020822665983957,
"grad_norm": 1.165669560432434,
"learning_rate": 1.0298347683694224e-05,
"loss": 2.0205,
"step": 1614
},
{
"epoch": 2.203447687318655,
"grad_norm": 1.200374722480774,
"learning_rate": 1.0278215598074133e-05,
"loss": 2.0584,
"step": 1615
},
{
"epoch": 2.2048131080389144,
"grad_norm": 1.1646145582199097,
"learning_rate": 1.025810894042373e-05,
"loss": 1.9527,
"step": 1616
},
{
"epoch": 2.2061785287591738,
"grad_norm": 1.111045002937317,
"learning_rate": 1.0238027754360238e-05,
"loss": 2.0103,
"step": 1617
},
{
"epoch": 2.2075439494794336,
"grad_norm": 1.1974886655807495,
"learning_rate": 1.0217972083445591e-05,
"loss": 2.0685,
"step": 1618
},
{
"epoch": 2.208909370199693,
"grad_norm": 1.237799048423767,
"learning_rate": 1.0197941971186393e-05,
"loss": 2.1097,
"step": 1619
},
{
"epoch": 2.2102747909199523,
"grad_norm": 1.1992545127868652,
"learning_rate": 1.0177937461033798e-05,
"loss": 1.9797,
"step": 1620
},
{
"epoch": 2.2116402116402116,
"grad_norm": 1.2122840881347656,
"learning_rate": 1.0157958596383433e-05,
"loss": 2.051,
"step": 1621
},
{
"epoch": 2.213005632360471,
"grad_norm": 1.1240154504776,
"learning_rate": 1.0138005420575281e-05,
"loss": 1.946,
"step": 1622
},
{
"epoch": 2.2143710530807303,
"grad_norm": 1.1616336107254028,
"learning_rate": 1.01180779768936e-05,
"loss": 2.0729,
"step": 1623
},
{
"epoch": 2.21573647380099,
"grad_norm": 1.1519469022750854,
"learning_rate": 1.0098176308566817e-05,
"loss": 2.101,
"step": 1624
},
{
"epoch": 2.2171018945212495,
"grad_norm": 1.1439200639724731,
"learning_rate": 1.0078300458767467e-05,
"loss": 1.9156,
"step": 1625
},
{
"epoch": 2.218467315241509,
"grad_norm": 1.1231272220611572,
"learning_rate": 1.0058450470612068e-05,
"loss": 1.9486,
"step": 1626
},
{
"epoch": 2.219832735961768,
"grad_norm": 1.1443954706192017,
"learning_rate": 1.003862638716103e-05,
"loss": 1.9831,
"step": 1627
},
{
"epoch": 2.2211981566820276,
"grad_norm": 1.2110002040863037,
"learning_rate": 1.0018828251418576e-05,
"loss": 1.89,
"step": 1628
},
{
"epoch": 2.222563577402287,
"grad_norm": 1.131422519683838,
"learning_rate": 9.999056106332635e-06,
"loss": 2.0307,
"step": 1629
},
{
"epoch": 2.2239289981225463,
"grad_norm": 1.1750086545944214,
"learning_rate": 9.979309994794757e-06,
"loss": 1.9842,
"step": 1630
},
{
"epoch": 2.225294418842806,
"grad_norm": 1.136098027229309,
"learning_rate": 9.959589959640021e-06,
"loss": 1.9184,
"step": 1631
},
{
"epoch": 2.2266598395630655,
"grad_norm": 1.1384086608886719,
"learning_rate": 9.939896043646932e-06,
"loss": 2.1782,
"step": 1632
},
{
"epoch": 2.228025260283325,
"grad_norm": 1.1740875244140625,
"learning_rate": 9.920228289537337e-06,
"loss": 2.1236,
"step": 1633
},
{
"epoch": 2.229390681003584,
"grad_norm": 1.1853106021881104,
"learning_rate": 9.90058673997634e-06,
"loss": 1.961,
"step": 1634
},
{
"epoch": 2.2307561017238435,
"grad_norm": 1.1684694290161133,
"learning_rate": 9.880971437572194e-06,
"loss": 1.9986,
"step": 1635
},
{
"epoch": 2.232121522444103,
"grad_norm": 1.155835509300232,
"learning_rate": 9.861382424876196e-06,
"loss": 1.9934,
"step": 1636
},
{
"epoch": 2.2334869431643627,
"grad_norm": 1.199952483177185,
"learning_rate": 9.841819744382634e-06,
"loss": 1.9918,
"step": 1637
},
{
"epoch": 2.234852363884622,
"grad_norm": 1.1183953285217285,
"learning_rate": 9.822283438528677e-06,
"loss": 2.0141,
"step": 1638
},
{
"epoch": 2.2362177846048814,
"grad_norm": 1.1736717224121094,
"learning_rate": 9.802773549694267e-06,
"loss": 1.9891,
"step": 1639
},
{
"epoch": 2.2375832053251408,
"grad_norm": 1.14573335647583,
"learning_rate": 9.783290120202043e-06,
"loss": 1.9699,
"step": 1640
},
{
"epoch": 2.2389486260454,
"grad_norm": 1.1581631898880005,
"learning_rate": 9.763833192317248e-06,
"loss": 1.9273,
"step": 1641
},
{
"epoch": 2.2403140467656595,
"grad_norm": 1.1063188314437866,
"learning_rate": 9.744402808247635e-06,
"loss": 2.039,
"step": 1642
},
{
"epoch": 2.2416794674859193,
"grad_norm": 1.1506023406982422,
"learning_rate": 9.724999010143372e-06,
"loss": 1.9951,
"step": 1643
},
{
"epoch": 2.2430448882061786,
"grad_norm": 1.1540553569793701,
"learning_rate": 9.70562184009696e-06,
"loss": 2.0294,
"step": 1644
},
{
"epoch": 2.244410308926438,
"grad_norm": 1.1372369527816772,
"learning_rate": 9.686271340143132e-06,
"loss": 1.948,
"step": 1645
},
{
"epoch": 2.2457757296466974,
"grad_norm": 1.1267369985580444,
"learning_rate": 9.666947552258764e-06,
"loss": 2.0442,
"step": 1646
},
{
"epoch": 2.2471411503669567,
"grad_norm": 1.1574680805206299,
"learning_rate": 9.647650518362799e-06,
"loss": 2.0136,
"step": 1647
},
{
"epoch": 2.248506571087216,
"grad_norm": 1.1667635440826416,
"learning_rate": 9.628380280316124e-06,
"loss": 1.9556,
"step": 1648
},
{
"epoch": 2.249871991807476,
"grad_norm": 1.1534496545791626,
"learning_rate": 9.609136879921506e-06,
"loss": 2.0214,
"step": 1649
},
{
"epoch": 2.2512374125277352,
"grad_norm": 1.1363247632980347,
"learning_rate": 9.589920358923492e-06,
"loss": 2.0243,
"step": 1650
},
{
"epoch": 2.2526028332479946,
"grad_norm": 1.200705647468567,
"learning_rate": 9.570730759008331e-06,
"loss": 1.8716,
"step": 1651
},
{
"epoch": 2.253968253968254,
"grad_norm": 1.0914024114608765,
"learning_rate": 9.551568121803859e-06,
"loss": 1.972,
"step": 1652
},
{
"epoch": 2.2553336746885133,
"grad_norm": 1.1444605588912964,
"learning_rate": 9.53243248887943e-06,
"loss": 1.992,
"step": 1653
},
{
"epoch": 2.2566990954087727,
"grad_norm": 1.1734261512756348,
"learning_rate": 9.5133239017458e-06,
"loss": 2.0342,
"step": 1654
},
{
"epoch": 2.258064516129032,
"grad_norm": 1.1729016304016113,
"learning_rate": 9.494242401855081e-06,
"loss": 2.0437,
"step": 1655
},
{
"epoch": 2.259429936849292,
"grad_norm": 1.1631510257720947,
"learning_rate": 9.475188030600608e-06,
"loss": 2.0655,
"step": 1656
},
{
"epoch": 2.260795357569551,
"grad_norm": 1.147375226020813,
"learning_rate": 9.456160829316872e-06,
"loss": 1.889,
"step": 1657
},
{
"epoch": 2.2621607782898105,
"grad_norm": 1.103159785270691,
"learning_rate": 9.437160839279422e-06,
"loss": 2.1335,
"step": 1658
},
{
"epoch": 2.26352619901007,
"grad_norm": 1.187808871269226,
"learning_rate": 9.418188101704779e-06,
"loss": 2.002,
"step": 1659
},
{
"epoch": 2.2648916197303293,
"grad_norm": 1.1828166246414185,
"learning_rate": 9.399242657750346e-06,
"loss": 2.0596,
"step": 1660
},
{
"epoch": 2.266257040450589,
"grad_norm": 1.1728928089141846,
"learning_rate": 9.380324548514316e-06,
"loss": 1.9575,
"step": 1661
},
{
"epoch": 2.2676224611708484,
"grad_norm": 1.1334058046340942,
"learning_rate": 9.361433815035588e-06,
"loss": 1.8906,
"step": 1662
},
{
"epoch": 2.268987881891108,
"grad_norm": 1.1529375314712524,
"learning_rate": 9.342570498293668e-06,
"loss": 2.0344,
"step": 1663
},
{
"epoch": 2.270353302611367,
"grad_norm": 1.133925437927246,
"learning_rate": 9.323734639208607e-06,
"loss": 1.8931,
"step": 1664
},
{
"epoch": 2.2717187233316265,
"grad_norm": 1.13822340965271,
"learning_rate": 9.304926278640865e-06,
"loss": 2.0319,
"step": 1665
},
{
"epoch": 2.273084144051886,
"grad_norm": 1.1660736799240112,
"learning_rate": 9.286145457391265e-06,
"loss": 1.9676,
"step": 1666
},
{
"epoch": 2.274449564772145,
"grad_norm": 1.1501853466033936,
"learning_rate": 9.26739221620088e-06,
"loss": 1.961,
"step": 1667
},
{
"epoch": 2.275814985492405,
"grad_norm": 1.1225939989089966,
"learning_rate": 9.248666595750976e-06,
"loss": 1.9155,
"step": 1668
},
{
"epoch": 2.2771804062126644,
"grad_norm": 1.1339390277862549,
"learning_rate": 9.229968636662878e-06,
"loss": 2.0527,
"step": 1669
},
{
"epoch": 2.2785458269329237,
"grad_norm": 1.166846513748169,
"learning_rate": 9.21129837949792e-06,
"loss": 1.9564,
"step": 1670
},
{
"epoch": 2.279911247653183,
"grad_norm": 1.135148048400879,
"learning_rate": 9.192655864757318e-06,
"loss": 1.9404,
"step": 1671
},
{
"epoch": 2.2812766683734425,
"grad_norm": 1.1577703952789307,
"learning_rate": 9.174041132882142e-06,
"loss": 2.0681,
"step": 1672
},
{
"epoch": 2.282642089093702,
"grad_norm": 1.1706666946411133,
"learning_rate": 9.155454224253172e-06,
"loss": 1.9762,
"step": 1673
},
{
"epoch": 2.2840075098139616,
"grad_norm": 1.1468843221664429,
"learning_rate": 9.136895179190832e-06,
"loss": 2.0729,
"step": 1674
},
{
"epoch": 2.285372930534221,
"grad_norm": 1.1795384883880615,
"learning_rate": 9.118364037955105e-06,
"loss": 2.0004,
"step": 1675
},
{
"epoch": 2.2867383512544803,
"grad_norm": 1.2244058847427368,
"learning_rate": 9.099860840745443e-06,
"loss": 2.0322,
"step": 1676
},
{
"epoch": 2.2881037719747397,
"grad_norm": 1.1860140562057495,
"learning_rate": 9.081385627700679e-06,
"loss": 1.9917,
"step": 1677
},
{
"epoch": 2.289469192694999,
"grad_norm": 1.1088379621505737,
"learning_rate": 9.062938438898936e-06,
"loss": 1.9874,
"step": 1678
},
{
"epoch": 2.2908346134152584,
"grad_norm": 1.2267651557922363,
"learning_rate": 9.04451931435755e-06,
"loss": 1.9497,
"step": 1679
},
{
"epoch": 2.2922000341355178,
"grad_norm": 1.1261852979660034,
"learning_rate": 9.02612829403297e-06,
"loss": 1.9654,
"step": 1680
},
{
"epoch": 2.2935654548557776,
"grad_norm": 1.1619963645935059,
"learning_rate": 9.007765417820693e-06,
"loss": 2.0292,
"step": 1681
},
{
"epoch": 2.294930875576037,
"grad_norm": 1.138156533241272,
"learning_rate": 8.989430725555157e-06,
"loss": 2.0463,
"step": 1682
},
{
"epoch": 2.2962962962962963,
"grad_norm": 1.1535521745681763,
"learning_rate": 8.971124257009644e-06,
"loss": 1.9308,
"step": 1683
},
{
"epoch": 2.2976617170165556,
"grad_norm": 1.1425036191940308,
"learning_rate": 8.952846051896229e-06,
"loss": 1.9825,
"step": 1684
},
{
"epoch": 2.299027137736815,
"grad_norm": 1.1328743696212769,
"learning_rate": 8.934596149865675e-06,
"loss": 1.9403,
"step": 1685
},
{
"epoch": 2.300392558457075,
"grad_norm": 1.1798624992370605,
"learning_rate": 8.916374590507344e-06,
"loss": 1.927,
"step": 1686
},
{
"epoch": 2.301757979177334,
"grad_norm": 1.1388009786605835,
"learning_rate": 8.898181413349107e-06,
"loss": 2.0535,
"step": 1687
},
{
"epoch": 2.3031233998975935,
"grad_norm": 1.1744321584701538,
"learning_rate": 8.880016657857278e-06,
"loss": 1.9411,
"step": 1688
},
{
"epoch": 2.304488820617853,
"grad_norm": 1.1813966035842896,
"learning_rate": 8.861880363436507e-06,
"loss": 2.0849,
"step": 1689
},
{
"epoch": 2.3058542413381122,
"grad_norm": 1.1790677309036255,
"learning_rate": 8.843772569429712e-06,
"loss": 2.059,
"step": 1690
},
{
"epoch": 2.3072196620583716,
"grad_norm": 1.1730358600616455,
"learning_rate": 8.825693315117974e-06,
"loss": 2.0023,
"step": 1691
},
{
"epoch": 2.308585082778631,
"grad_norm": 1.1869069337844849,
"learning_rate": 8.807642639720475e-06,
"loss": 2.0288,
"step": 1692
},
{
"epoch": 2.3099505034988908,
"grad_norm": 1.159970760345459,
"learning_rate": 8.789620582394385e-06,
"loss": 1.9623,
"step": 1693
},
{
"epoch": 2.31131592421915,
"grad_norm": 1.1603195667266846,
"learning_rate": 8.771627182234824e-06,
"loss": 1.9996,
"step": 1694
},
{
"epoch": 2.3126813449394095,
"grad_norm": 1.1524176597595215,
"learning_rate": 8.753662478274704e-06,
"loss": 1.9765,
"step": 1695
},
{
"epoch": 2.314046765659669,
"grad_norm": 1.129731297492981,
"learning_rate": 8.735726509484722e-06,
"loss": 2.0091,
"step": 1696
},
{
"epoch": 2.315412186379928,
"grad_norm": 1.164451241493225,
"learning_rate": 8.717819314773213e-06,
"loss": 1.9598,
"step": 1697
},
{
"epoch": 2.316777607100188,
"grad_norm": 1.1587612628936768,
"learning_rate": 8.699940932986121e-06,
"loss": 1.9889,
"step": 1698
},
{
"epoch": 2.3181430278204473,
"grad_norm": 1.1855641603469849,
"learning_rate": 8.682091402906871e-06,
"loss": 1.9548,
"step": 1699
},
{
"epoch": 2.3195084485407067,
"grad_norm": 1.104287028312683,
"learning_rate": 8.66427076325629e-06,
"loss": 1.8798,
"step": 1700
},
{
"epoch": 2.320873869260966,
"grad_norm": 1.1085776090621948,
"learning_rate": 8.646479052692541e-06,
"loss": 2.001,
"step": 1701
},
{
"epoch": 2.3222392899812254,
"grad_norm": 1.1575961112976074,
"learning_rate": 8.62871630981105e-06,
"loss": 2.0229,
"step": 1702
},
{
"epoch": 2.3236047107014848,
"grad_norm": 1.1817338466644287,
"learning_rate": 8.610982573144376e-06,
"loss": 2.1197,
"step": 1703
},
{
"epoch": 2.324970131421744,
"grad_norm": 1.19670569896698,
"learning_rate": 8.593277881162171e-06,
"loss": 2.0872,
"step": 1704
},
{
"epoch": 2.326335552142004,
"grad_norm": 1.159096121788025,
"learning_rate": 8.575602272271077e-06,
"loss": 1.984,
"step": 1705
},
{
"epoch": 2.3277009728622633,
"grad_norm": 1.1681056022644043,
"learning_rate": 8.557955784814647e-06,
"loss": 2.1071,
"step": 1706
},
{
"epoch": 2.3290663935825227,
"grad_norm": 1.1719386577606201,
"learning_rate": 8.540338457073253e-06,
"loss": 1.9543,
"step": 1707
},
{
"epoch": 2.330431814302782,
"grad_norm": 1.221099615097046,
"learning_rate": 8.522750327264024e-06,
"loss": 1.9552,
"step": 1708
},
{
"epoch": 2.3317972350230414,
"grad_norm": 1.155889868736267,
"learning_rate": 8.505191433540743e-06,
"loss": 1.9871,
"step": 1709
},
{
"epoch": 2.3331626557433007,
"grad_norm": 1.1443977355957031,
"learning_rate": 8.487661813993767e-06,
"loss": 2.0017,
"step": 1710
},
{
"epoch": 2.3345280764635605,
"grad_norm": 1.1238549947738647,
"learning_rate": 8.47016150664997e-06,
"loss": 2.0162,
"step": 1711
},
{
"epoch": 2.33589349718382,
"grad_norm": 1.1937847137451172,
"learning_rate": 8.452690549472608e-06,
"loss": 1.9845,
"step": 1712
},
{
"epoch": 2.3372589179040792,
"grad_norm": 1.1769059896469116,
"learning_rate": 8.43524898036129e-06,
"loss": 1.9769,
"step": 1713
},
{
"epoch": 2.3386243386243386,
"grad_norm": 1.1830600500106812,
"learning_rate": 8.417836837151857e-06,
"loss": 1.9595,
"step": 1714
},
{
"epoch": 2.339989759344598,
"grad_norm": 1.1209391355514526,
"learning_rate": 8.400454157616343e-06,
"loss": 1.9851,
"step": 1715
},
{
"epoch": 2.3413551800648573,
"grad_norm": 1.141635775566101,
"learning_rate": 8.383100979462843e-06,
"loss": 1.9849,
"step": 1716
},
{
"epoch": 2.3427206007851167,
"grad_norm": 1.1458054780960083,
"learning_rate": 8.365777340335466e-06,
"loss": 2.0071,
"step": 1717
},
{
"epoch": 2.3440860215053765,
"grad_norm": 1.139327049255371,
"learning_rate": 8.348483277814228e-06,
"loss": 1.9089,
"step": 1718
},
{
"epoch": 2.345451442225636,
"grad_norm": 1.136906623840332,
"learning_rate": 8.331218829415008e-06,
"loss": 1.9176,
"step": 1719
},
{
"epoch": 2.346816862945895,
"grad_norm": 1.0827677249908447,
"learning_rate": 8.313984032589423e-06,
"loss": 1.9509,
"step": 1720
},
{
"epoch": 2.3481822836661546,
"grad_norm": 1.162163257598877,
"learning_rate": 8.296778924724777e-06,
"loss": 1.936,
"step": 1721
},
{
"epoch": 2.349547704386414,
"grad_norm": 1.1270315647125244,
"learning_rate": 8.27960354314397e-06,
"loss": 1.9451,
"step": 1722
},
{
"epoch": 2.3509131251066737,
"grad_norm": 1.1139661073684692,
"learning_rate": 8.262457925105414e-06,
"loss": 1.8272,
"step": 1723
},
{
"epoch": 2.352278545826933,
"grad_norm": 1.1409908533096313,
"learning_rate": 8.245342107802954e-06,
"loss": 1.9406,
"step": 1724
},
{
"epoch": 2.3536439665471924,
"grad_norm": 1.1486446857452393,
"learning_rate": 8.22825612836579e-06,
"loss": 2.0512,
"step": 1725
},
{
"epoch": 2.355009387267452,
"grad_norm": 1.1808385848999023,
"learning_rate": 8.2112000238584e-06,
"loss": 1.9244,
"step": 1726
},
{
"epoch": 2.356374807987711,
"grad_norm": 1.152113676071167,
"learning_rate": 8.194173831280438e-06,
"loss": 1.9924,
"step": 1727
},
{
"epoch": 2.3577402287079705,
"grad_norm": 1.1574903726577759,
"learning_rate": 8.177177587566698e-06,
"loss": 1.9877,
"step": 1728
},
{
"epoch": 2.35910564942823,
"grad_norm": 1.137571930885315,
"learning_rate": 8.160211329586991e-06,
"loss": 1.9495,
"step": 1729
},
{
"epoch": 2.3604710701484897,
"grad_norm": 1.206718921661377,
"learning_rate": 8.143275094146068e-06,
"loss": 1.9831,
"step": 1730
},
{
"epoch": 2.361836490868749,
"grad_norm": 1.1976664066314697,
"learning_rate": 8.12636891798356e-06,
"loss": 1.9604,
"step": 1731
},
{
"epoch": 2.3632019115890084,
"grad_norm": 1.1441067457199097,
"learning_rate": 8.109492837773912e-06,
"loss": 1.9223,
"step": 1732
},
{
"epoch": 2.3645673323092677,
"grad_norm": 1.121513843536377,
"learning_rate": 8.092646890126255e-06,
"loss": 2.0806,
"step": 1733
},
{
"epoch": 2.365932753029527,
"grad_norm": 1.2334645986557007,
"learning_rate": 8.075831111584364e-06,
"loss": 2.0784,
"step": 1734
},
{
"epoch": 2.3672981737497865,
"grad_norm": 1.1921037435531616,
"learning_rate": 8.05904553862657e-06,
"loss": 2.0306,
"step": 1735
},
{
"epoch": 2.3686635944700463,
"grad_norm": 1.181428074836731,
"learning_rate": 8.042290207665677e-06,
"loss": 1.869,
"step": 1736
},
{
"epoch": 2.3700290151903056,
"grad_norm": 1.1490988731384277,
"learning_rate": 8.025565155048884e-06,
"loss": 2.0124,
"step": 1737
},
{
"epoch": 2.371394435910565,
"grad_norm": 1.1137522459030151,
"learning_rate": 8.008870417057709e-06,
"loss": 2.0171,
"step": 1738
},
{
"epoch": 2.3727598566308243,
"grad_norm": 1.1768802404403687,
"learning_rate": 7.99220602990791e-06,
"loss": 2.0773,
"step": 1739
},
{
"epoch": 2.3741252773510837,
"grad_norm": 1.1892272233963013,
"learning_rate": 7.975572029749402e-06,
"loss": 1.8676,
"step": 1740
},
{
"epoch": 2.375490698071343,
"grad_norm": 1.1409162282943726,
"learning_rate": 7.958968452666184e-06,
"loss": 1.9957,
"step": 1741
},
{
"epoch": 2.376856118791603,
"grad_norm": 1.1904845237731934,
"learning_rate": 7.94239533467626e-06,
"loss": 1.9067,
"step": 1742
},
{
"epoch": 2.378221539511862,
"grad_norm": 1.1122486591339111,
"learning_rate": 7.925852711731557e-06,
"loss": 1.978,
"step": 1743
},
{
"epoch": 2.3795869602321216,
"grad_norm": 1.1298130750656128,
"learning_rate": 7.909340619717848e-06,
"loss": 1.8801,
"step": 1744
},
{
"epoch": 2.380952380952381,
"grad_norm": 1.118901252746582,
"learning_rate": 7.892859094454672e-06,
"loss": 1.9176,
"step": 1745
},
{
"epoch": 2.3823178016726403,
"grad_norm": 1.1540254354476929,
"learning_rate": 7.87640817169528e-06,
"loss": 2.0089,
"step": 1746
},
{
"epoch": 2.3836832223928996,
"grad_norm": 1.1526018381118774,
"learning_rate": 7.859987887126512e-06,
"loss": 1.9921,
"step": 1747
},
{
"epoch": 2.3850486431131595,
"grad_norm": 1.1401500701904297,
"learning_rate": 7.843598276368756e-06,
"loss": 1.9359,
"step": 1748
},
{
"epoch": 2.386414063833419,
"grad_norm": 1.1453697681427002,
"learning_rate": 7.827239374975853e-06,
"loss": 2.0905,
"step": 1749
},
{
"epoch": 2.387779484553678,
"grad_norm": 1.215514898300171,
"learning_rate": 7.810911218435044e-06,
"loss": 1.9321,
"step": 1750
},
{
"epoch": 2.3891449052739375,
"grad_norm": 1.129007339477539,
"learning_rate": 7.794613842166859e-06,
"loss": 1.9894,
"step": 1751
},
{
"epoch": 2.390510325994197,
"grad_norm": 1.1532918214797974,
"learning_rate": 7.778347281525058e-06,
"loss": 2.0324,
"step": 1752
},
{
"epoch": 2.3918757467144562,
"grad_norm": 1.1395232677459717,
"learning_rate": 7.762111571796562e-06,
"loss": 1.8054,
"step": 1753
},
{
"epoch": 2.3932411674347156,
"grad_norm": 1.119819164276123,
"learning_rate": 7.745906748201354e-06,
"loss": 1.9846,
"step": 1754
},
{
"epoch": 2.3946065881549754,
"grad_norm": 1.1947277784347534,
"learning_rate": 7.72973284589243e-06,
"loss": 1.9281,
"step": 1755
},
{
"epoch": 2.3959720088752348,
"grad_norm": 1.1208232641220093,
"learning_rate": 7.713589899955698e-06,
"loss": 2.05,
"step": 1756
},
{
"epoch": 2.397337429595494,
"grad_norm": 1.166419267654419,
"learning_rate": 7.69747794540992e-06,
"loss": 1.9776,
"step": 1757
},
{
"epoch": 2.3987028503157535,
"grad_norm": 1.21503484249115,
"learning_rate": 7.681397017206621e-06,
"loss": 2.0448,
"step": 1758
},
{
"epoch": 2.400068271036013,
"grad_norm": 1.1262987852096558,
"learning_rate": 7.665347150230029e-06,
"loss": 2.0476,
"step": 1759
},
{
"epoch": 2.4014336917562726,
"grad_norm": 1.1786149740219116,
"learning_rate": 7.649328379296984e-06,
"loss": 2.0104,
"step": 1760
},
{
"epoch": 2.402799112476532,
"grad_norm": 1.1558293104171753,
"learning_rate": 7.633340739156879e-06,
"loss": 1.9336,
"step": 1761
},
{
"epoch": 2.4041645331967914,
"grad_norm": 1.1994011402130127,
"learning_rate": 7.617384264491558e-06,
"loss": 1.9279,
"step": 1762
},
{
"epoch": 2.4055299539170507,
"grad_norm": 1.1507861614227295,
"learning_rate": 7.601458989915283e-06,
"loss": 2.0639,
"step": 1763
},
{
"epoch": 2.40689537463731,
"grad_norm": 1.141413927078247,
"learning_rate": 7.585564949974615e-06,
"loss": 2.0175,
"step": 1764
},
{
"epoch": 2.4082607953575694,
"grad_norm": 1.23114013671875,
"learning_rate": 7.569702179148357e-06,
"loss": 1.9706,
"step": 1765
},
{
"epoch": 2.409626216077829,
"grad_norm": 1.201511263847351,
"learning_rate": 7.553870711847489e-06,
"loss": 1.9928,
"step": 1766
},
{
"epoch": 2.4109916367980886,
"grad_norm": 1.1835371255874634,
"learning_rate": 7.538070582415087e-06,
"loss": 1.9937,
"step": 1767
},
{
"epoch": 2.412357057518348,
"grad_norm": 1.15757155418396,
"learning_rate": 7.522301825126235e-06,
"loss": 1.9344,
"step": 1768
},
{
"epoch": 2.4137224782386073,
"grad_norm": 1.1326370239257812,
"learning_rate": 7.506564474187972e-06,
"loss": 1.9968,
"step": 1769
},
{
"epoch": 2.4150878989588667,
"grad_norm": 1.1341724395751953,
"learning_rate": 7.490858563739198e-06,
"loss": 1.9258,
"step": 1770
},
{
"epoch": 2.416453319679126,
"grad_norm": 1.1367473602294922,
"learning_rate": 7.475184127850619e-06,
"loss": 1.948,
"step": 1771
},
{
"epoch": 2.4178187403993854,
"grad_norm": 1.1591957807540894,
"learning_rate": 7.459541200524655e-06,
"loss": 2.0133,
"step": 1772
},
{
"epoch": 2.419184161119645,
"grad_norm": 1.1769999265670776,
"learning_rate": 7.443929815695381e-06,
"loss": 2.0722,
"step": 1773
},
{
"epoch": 2.4205495818399045,
"grad_norm": 1.1746711730957031,
"learning_rate": 7.428350007228442e-06,
"loss": 2.0398,
"step": 1774
},
{
"epoch": 2.421915002560164,
"grad_norm": 1.1451435089111328,
"learning_rate": 7.412801808920987e-06,
"loss": 1.9854,
"step": 1775
},
{
"epoch": 2.4232804232804233,
"grad_norm": 1.1569113731384277,
"learning_rate": 7.397285254501603e-06,
"loss": 2.0005,
"step": 1776
},
{
"epoch": 2.4246458440006826,
"grad_norm": 1.1916972398757935,
"learning_rate": 7.3818003776302105e-06,
"loss": 2.0972,
"step": 1777
},
{
"epoch": 2.426011264720942,
"grad_norm": 1.1936883926391602,
"learning_rate": 7.366347211898031e-06,
"loss": 1.9279,
"step": 1778
},
{
"epoch": 2.4273766854412013,
"grad_norm": 1.1212059259414673,
"learning_rate": 7.350925790827481e-06,
"loss": 1.9314,
"step": 1779
},
{
"epoch": 2.428742106161461,
"grad_norm": 1.1527293920516968,
"learning_rate": 7.335536147872131e-06,
"loss": 2.0102,
"step": 1780
},
{
"epoch": 2.4301075268817205,
"grad_norm": 1.1258546113967896,
"learning_rate": 7.320178316416604e-06,
"loss": 2.005,
"step": 1781
},
{
"epoch": 2.43147294760198,
"grad_norm": 1.1910921335220337,
"learning_rate": 7.30485232977652e-06,
"loss": 2.0559,
"step": 1782
},
{
"epoch": 2.432838368322239,
"grad_norm": 1.2117758989334106,
"learning_rate": 7.289558221198401e-06,
"loss": 2.0011,
"step": 1783
},
{
"epoch": 2.4342037890424986,
"grad_norm": 1.1756689548492432,
"learning_rate": 7.2742960238596415e-06,
"loss": 1.9803,
"step": 1784
},
{
"epoch": 2.4355692097627584,
"grad_norm": 1.1738263368606567,
"learning_rate": 7.259065770868397e-06,
"loss": 1.968,
"step": 1785
},
{
"epoch": 2.4369346304830177,
"grad_norm": 1.096643328666687,
"learning_rate": 7.243867495263527e-06,
"loss": 2.0182,
"step": 1786
},
{
"epoch": 2.438300051203277,
"grad_norm": 1.143823266029358,
"learning_rate": 7.228701230014528e-06,
"loss": 1.9669,
"step": 1787
},
{
"epoch": 2.4396654719235364,
"grad_norm": 1.1696970462799072,
"learning_rate": 7.2135670080214516e-06,
"loss": 2.004,
"step": 1788
},
{
"epoch": 2.441030892643796,
"grad_norm": 1.167273998260498,
"learning_rate": 7.198464862114838e-06,
"loss": 1.9823,
"step": 1789
},
{
"epoch": 2.442396313364055,
"grad_norm": 1.1593377590179443,
"learning_rate": 7.183394825055652e-06,
"loss": 1.9194,
"step": 1790
},
{
"epoch": 2.4437617340843145,
"grad_norm": 1.1354905366897583,
"learning_rate": 7.168356929535196e-06,
"loss": 1.9676,
"step": 1791
},
{
"epoch": 2.4451271548045743,
"grad_norm": 1.1306127309799194,
"learning_rate": 7.153351208175048e-06,
"loss": 2.0386,
"step": 1792
},
{
"epoch": 2.4464925755248337,
"grad_norm": 1.1695749759674072,
"learning_rate": 7.138377693527011e-06,
"loss": 2.0283,
"step": 1793
},
{
"epoch": 2.447857996245093,
"grad_norm": 1.109798789024353,
"learning_rate": 7.123436418072989e-06,
"loss": 1.9902,
"step": 1794
},
{
"epoch": 2.4492234169653524,
"grad_norm": 1.172412633895874,
"learning_rate": 7.108527414224974e-06,
"loss": 1.8803,
"step": 1795
},
{
"epoch": 2.4505888376856118,
"grad_norm": 1.1297450065612793,
"learning_rate": 7.09365071432494e-06,
"loss": 2.0167,
"step": 1796
},
{
"epoch": 2.4519542584058716,
"grad_norm": 1.2103015184402466,
"learning_rate": 7.078806350644796e-06,
"loss": 1.9646,
"step": 1797
},
{
"epoch": 2.453319679126131,
"grad_norm": 1.1565308570861816,
"learning_rate": 7.063994355386293e-06,
"loss": 2.0531,
"step": 1798
},
{
"epoch": 2.4546850998463903,
"grad_norm": 1.1517778635025024,
"learning_rate": 7.049214760680973e-06,
"loss": 2.0818,
"step": 1799
},
{
"epoch": 2.4560505205666496,
"grad_norm": 1.1525768041610718,
"learning_rate": 7.03446759859008e-06,
"loss": 1.9867,
"step": 1800
},
{
"epoch": 2.457415941286909,
"grad_norm": 1.1512937545776367,
"learning_rate": 7.0197529011045165e-06,
"loss": 1.9202,
"step": 1801
},
{
"epoch": 2.4587813620071683,
"grad_norm": 1.143919587135315,
"learning_rate": 7.005070700144754e-06,
"loss": 2.0085,
"step": 1802
},
{
"epoch": 2.4601467827274277,
"grad_norm": 1.1674836874008179,
"learning_rate": 6.990421027560767e-06,
"loss": 1.8961,
"step": 1803
},
{
"epoch": 2.4615122034476875,
"grad_norm": 1.1142269372940063,
"learning_rate": 6.975803915131972e-06,
"loss": 1.9979,
"step": 1804
},
{
"epoch": 2.462877624167947,
"grad_norm": 1.1487048864364624,
"learning_rate": 6.961219394567148e-06,
"loss": 1.9235,
"step": 1805
},
{
"epoch": 2.4642430448882062,
"grad_norm": 1.1297390460968018,
"learning_rate": 6.946667497504371e-06,
"loss": 1.9975,
"step": 1806
},
{
"epoch": 2.4656084656084656,
"grad_norm": 1.198565125465393,
"learning_rate": 6.932148255510955e-06,
"loss": 1.9977,
"step": 1807
},
{
"epoch": 2.466973886328725,
"grad_norm": 1.1676534414291382,
"learning_rate": 6.917661700083371e-06,
"loss": 1.9301,
"step": 1808
},
{
"epoch": 2.4683393070489843,
"grad_norm": 1.1382228136062622,
"learning_rate": 6.9032078626471765e-06,
"loss": 1.9651,
"step": 1809
},
{
"epoch": 2.469704727769244,
"grad_norm": 1.1406251192092896,
"learning_rate": 6.888786774556972e-06,
"loss": 2.0434,
"step": 1810
},
{
"epoch": 2.4710701484895035,
"grad_norm": 1.1619056463241577,
"learning_rate": 6.8743984670963045e-06,
"loss": 1.912,
"step": 1811
},
{
"epoch": 2.472435569209763,
"grad_norm": 1.1315535306930542,
"learning_rate": 6.860042971477601e-06,
"loss": 1.9869,
"step": 1812
},
{
"epoch": 2.473800989930022,
"grad_norm": 1.1548961400985718,
"learning_rate": 6.845720318842117e-06,
"loss": 1.9452,
"step": 1813
},
{
"epoch": 2.4751664106502815,
"grad_norm": 1.1344841718673706,
"learning_rate": 6.831430540259875e-06,
"loss": 1.9741,
"step": 1814
},
{
"epoch": 2.476531831370541,
"grad_norm": 1.1183604001998901,
"learning_rate": 6.8171736667295695e-06,
"loss": 2.0076,
"step": 1815
},
{
"epoch": 2.4778972520908003,
"grad_norm": 1.158690333366394,
"learning_rate": 6.802949729178516e-06,
"loss": 1.9996,
"step": 1816
},
{
"epoch": 2.47926267281106,
"grad_norm": 1.1644699573516846,
"learning_rate": 6.788758758462586e-06,
"loss": 2.0708,
"step": 1817
},
{
"epoch": 2.4806280935313194,
"grad_norm": 1.168336272239685,
"learning_rate": 6.774600785366134e-06,
"loss": 2.0358,
"step": 1818
},
{
"epoch": 2.4819935142515788,
"grad_norm": 1.138435959815979,
"learning_rate": 6.7604758406019325e-06,
"loss": 1.9077,
"step": 1819
},
{
"epoch": 2.483358934971838,
"grad_norm": 1.1225330829620361,
"learning_rate": 6.746383954811109e-06,
"loss": 1.9646,
"step": 1820
},
{
"epoch": 2.4847243556920975,
"grad_norm": 1.1390725374221802,
"learning_rate": 6.732325158563071e-06,
"loss": 2.0892,
"step": 1821
},
{
"epoch": 2.4860897764123573,
"grad_norm": 1.205795168876648,
"learning_rate": 6.718299482355448e-06,
"loss": 1.9228,
"step": 1822
},
{
"epoch": 2.4874551971326166,
"grad_norm": 1.098448395729065,
"learning_rate": 6.704306956614035e-06,
"loss": 1.8744,
"step": 1823
},
{
"epoch": 2.488820617852876,
"grad_norm": 1.1436505317687988,
"learning_rate": 6.690347611692688e-06,
"loss": 2.0349,
"step": 1824
},
{
"epoch": 2.4901860385731354,
"grad_norm": 1.1744968891143799,
"learning_rate": 6.676421477873305e-06,
"loss": 1.8819,
"step": 1825
},
{
"epoch": 2.4915514592933947,
"grad_norm": 1.1274679899215698,
"learning_rate": 6.662528585365724e-06,
"loss": 1.9992,
"step": 1826
},
{
"epoch": 2.492916880013654,
"grad_norm": 1.1404972076416016,
"learning_rate": 6.648668964307694e-06,
"loss": 1.9579,
"step": 1827
},
{
"epoch": 2.4942823007339134,
"grad_norm": 1.1242655515670776,
"learning_rate": 6.6348426447647756e-06,
"loss": 1.9794,
"step": 1828
},
{
"epoch": 2.4956477214541732,
"grad_norm": 1.1378898620605469,
"learning_rate": 6.621049656730278e-06,
"loss": 2.017,
"step": 1829
},
{
"epoch": 2.4970131421744326,
"grad_norm": 1.1961363554000854,
"learning_rate": 6.607290030125219e-06,
"loss": 1.9035,
"step": 1830
},
{
"epoch": 2.498378562894692,
"grad_norm": 1.1399198770523071,
"learning_rate": 6.5935637947982485e-06,
"loss": 1.9177,
"step": 1831
},
{
"epoch": 2.4997439836149513,
"grad_norm": 1.1363757848739624,
"learning_rate": 6.579870980525573e-06,
"loss": 2.0347,
"step": 1832
},
{
"epoch": 2.5011094043352107,
"grad_norm": 1.1126019954681396,
"learning_rate": 6.566211617010899e-06,
"loss": 1.9307,
"step": 1833
},
{
"epoch": 2.5024748250554705,
"grad_norm": 1.1316169500350952,
"learning_rate": 6.552585733885369e-06,
"loss": 1.8813,
"step": 1834
},
{
"epoch": 2.50384024577573,
"grad_norm": 1.1355621814727783,
"learning_rate": 6.538993360707504e-06,
"loss": 2.0635,
"step": 1835
},
{
"epoch": 2.505205666495989,
"grad_norm": 1.2004729509353638,
"learning_rate": 6.525434526963119e-06,
"loss": 2.0879,
"step": 1836
},
{
"epoch": 2.5065710872162486,
"grad_norm": 1.1733347177505493,
"learning_rate": 6.511909262065287e-06,
"loss": 1.9706,
"step": 1837
},
{
"epoch": 2.507936507936508,
"grad_norm": 1.1407383680343628,
"learning_rate": 6.498417595354247e-06,
"loss": 1.9866,
"step": 1838
},
{
"epoch": 2.5093019286567673,
"grad_norm": 1.2119171619415283,
"learning_rate": 6.484959556097359e-06,
"loss": 1.9701,
"step": 1839
},
{
"epoch": 2.5106673493770266,
"grad_norm": 1.1421514749526978,
"learning_rate": 6.4715351734890465e-06,
"loss": 2.0201,
"step": 1840
},
{
"epoch": 2.512032770097286,
"grad_norm": 1.1614763736724854,
"learning_rate": 6.458144476650699e-06,
"loss": 1.9961,
"step": 1841
},
{
"epoch": 2.513398190817546,
"grad_norm": 1.2180473804473877,
"learning_rate": 6.44478749463065e-06,
"loss": 2.0534,
"step": 1842
},
{
"epoch": 2.514763611537805,
"grad_norm": 1.1740957498550415,
"learning_rate": 6.431464256404084e-06,
"loss": 1.9125,
"step": 1843
},
{
"epoch": 2.5161290322580645,
"grad_norm": 1.1058861017227173,
"learning_rate": 6.418174790873002e-06,
"loss": 1.9706,
"step": 1844
},
{
"epoch": 2.517494452978324,
"grad_norm": 1.1636439561843872,
"learning_rate": 6.404919126866124e-06,
"loss": 1.9868,
"step": 1845
},
{
"epoch": 2.518859873698583,
"grad_norm": 1.137733817100525,
"learning_rate": 6.391697293138856e-06,
"loss": 1.9444,
"step": 1846
},
{
"epoch": 2.520225294418843,
"grad_norm": 1.204471468925476,
"learning_rate": 6.378509318373205e-06,
"loss": 2.0361,
"step": 1847
},
{
"epoch": 2.5215907151391024,
"grad_norm": 1.1572145223617554,
"learning_rate": 6.36535523117774e-06,
"loss": 2.0729,
"step": 1848
},
{
"epoch": 2.5229561358593617,
"grad_norm": 1.1491202116012573,
"learning_rate": 6.352235060087516e-06,
"loss": 2.0005,
"step": 1849
},
{
"epoch": 2.524321556579621,
"grad_norm": 1.1602143049240112,
"learning_rate": 6.3391488335640075e-06,
"loss": 1.9631,
"step": 1850
},
{
"epoch": 2.5256869772998805,
"grad_norm": 1.1818360090255737,
"learning_rate": 6.326096579995058e-06,
"loss": 2.0925,
"step": 1851
},
{
"epoch": 2.52705239802014,
"grad_norm": 1.2058542966842651,
"learning_rate": 6.313078327694814e-06,
"loss": 1.9747,
"step": 1852
},
{
"epoch": 2.528417818740399,
"grad_norm": 1.1478102207183838,
"learning_rate": 6.30009410490366e-06,
"loss": 2.1545,
"step": 1853
},
{
"epoch": 2.529783239460659,
"grad_norm": 1.2126845121383667,
"learning_rate": 6.2871439397881664e-06,
"loss": 1.9659,
"step": 1854
},
{
"epoch": 2.5311486601809183,
"grad_norm": 1.1527148485183716,
"learning_rate": 6.2742278604410135e-06,
"loss": 2.1155,
"step": 1855
},
{
"epoch": 2.5325140809011777,
"grad_norm": 1.1743618249893188,
"learning_rate": 6.261345894880947e-06,
"loss": 1.8983,
"step": 1856
},
{
"epoch": 2.533879501621437,
"grad_norm": 1.1428196430206299,
"learning_rate": 6.248498071052713e-06,
"loss": 1.9407,
"step": 1857
},
{
"epoch": 2.5352449223416964,
"grad_norm": 1.1778620481491089,
"learning_rate": 6.235684416826989e-06,
"loss": 2.043,
"step": 1858
},
{
"epoch": 2.536610343061956,
"grad_norm": 1.1823382377624512,
"learning_rate": 6.222904960000322e-06,
"loss": 2.1256,
"step": 1859
},
{
"epoch": 2.5379757637822156,
"grad_norm": 1.1378893852233887,
"learning_rate": 6.210159728295081e-06,
"loss": 1.9857,
"step": 1860
},
{
"epoch": 2.539341184502475,
"grad_norm": 1.1813747882843018,
"learning_rate": 6.197448749359399e-06,
"loss": 2.0536,
"step": 1861
},
{
"epoch": 2.5407066052227343,
"grad_norm": 1.1867306232452393,
"learning_rate": 6.184772050767094e-06,
"loss": 1.9548,
"step": 1862
},
{
"epoch": 2.5420720259429936,
"grad_norm": 1.1601258516311646,
"learning_rate": 6.172129660017622e-06,
"loss": 1.9592,
"step": 1863
},
{
"epoch": 2.543437446663253,
"grad_norm": 1.1704524755477905,
"learning_rate": 6.159521604536017e-06,
"loss": 2.0577,
"step": 1864
},
{
"epoch": 2.5448028673835124,
"grad_norm": 1.1493463516235352,
"learning_rate": 6.1469479116728315e-06,
"loss": 1.9636,
"step": 1865
},
{
"epoch": 2.5461682881037717,
"grad_norm": 1.1779242753982544,
"learning_rate": 6.13440860870407e-06,
"loss": 2.0038,
"step": 1866
},
{
"epoch": 2.5475337088240315,
"grad_norm": 1.1490920782089233,
"learning_rate": 6.121903722831139e-06,
"loss": 1.8697,
"step": 1867
},
{
"epoch": 2.548899129544291,
"grad_norm": 1.1036715507507324,
"learning_rate": 6.109433281180784e-06,
"loss": 1.9055,
"step": 1868
},
{
"epoch": 2.5502645502645502,
"grad_norm": 1.142822027206421,
"learning_rate": 6.096997310805033e-06,
"loss": 2.0474,
"step": 1869
},
{
"epoch": 2.5516299709848096,
"grad_norm": 1.1565924882888794,
"learning_rate": 6.0845958386811285e-06,
"loss": 2.0113,
"step": 1870
},
{
"epoch": 2.5529953917050694,
"grad_norm": 1.176205039024353,
"learning_rate": 6.072228891711487e-06,
"loss": 1.9685,
"step": 1871
},
{
"epoch": 2.5543608124253288,
"grad_norm": 1.1641393899917603,
"learning_rate": 6.05989649672362e-06,
"loss": 1.9868,
"step": 1872
},
{
"epoch": 2.555726233145588,
"grad_norm": 1.1368814706802368,
"learning_rate": 6.047598680470086e-06,
"loss": 2.0248,
"step": 1873
},
{
"epoch": 2.5570916538658475,
"grad_norm": 1.1509157419204712,
"learning_rate": 6.035335469628445e-06,
"loss": 1.9186,
"step": 1874
},
{
"epoch": 2.558457074586107,
"grad_norm": 1.1338989734649658,
"learning_rate": 6.023106890801177e-06,
"loss": 2.0014,
"step": 1875
},
{
"epoch": 2.559822495306366,
"grad_norm": 1.1391099691390991,
"learning_rate": 6.010912970515631e-06,
"loss": 1.9774,
"step": 1876
},
{
"epoch": 2.5611879160266255,
"grad_norm": 1.1549036502838135,
"learning_rate": 5.9987537352239755e-06,
"loss": 1.9657,
"step": 1877
},
{
"epoch": 2.562553336746885,
"grad_norm": 1.1228522062301636,
"learning_rate": 5.986629211303147e-06,
"loss": 1.9256,
"step": 1878
},
{
"epoch": 2.5639187574671447,
"grad_norm": 1.1399633884429932,
"learning_rate": 5.9745394250547695e-06,
"loss": 1.9701,
"step": 1879
},
{
"epoch": 2.565284178187404,
"grad_norm": 1.1724207401275635,
"learning_rate": 5.962484402705118e-06,
"loss": 2.0512,
"step": 1880
},
{
"epoch": 2.5666495989076634,
"grad_norm": 1.1585307121276855,
"learning_rate": 5.950464170405052e-06,
"loss": 2.0084,
"step": 1881
},
{
"epoch": 2.568015019627923,
"grad_norm": 1.175304651260376,
"learning_rate": 5.938478754229964e-06,
"loss": 2.0646,
"step": 1882
},
{
"epoch": 2.569380440348182,
"grad_norm": 1.189307451248169,
"learning_rate": 5.926528180179712e-06,
"loss": 1.9593,
"step": 1883
},
{
"epoch": 2.570745861068442,
"grad_norm": 1.1163198947906494,
"learning_rate": 5.914612474178582e-06,
"loss": 2.0873,
"step": 1884
},
{
"epoch": 2.5721112817887013,
"grad_norm": 1.1498438119888306,
"learning_rate": 5.902731662075215e-06,
"loss": 1.9559,
"step": 1885
},
{
"epoch": 2.5734767025089607,
"grad_norm": 1.10408353805542,
"learning_rate": 5.890885769642553e-06,
"loss": 1.9502,
"step": 1886
},
{
"epoch": 2.57484212322922,
"grad_norm": 1.1636332273483276,
"learning_rate": 5.879074822577806e-06,
"loss": 1.9467,
"step": 1887
},
{
"epoch": 2.5762075439494794,
"grad_norm": 1.1333767175674438,
"learning_rate": 5.8672988465023515e-06,
"loss": 2.0818,
"step": 1888
},
{
"epoch": 2.5775729646697387,
"grad_norm": 1.1534925699234009,
"learning_rate": 5.855557866961721e-06,
"loss": 1.8907,
"step": 1889
},
{
"epoch": 2.578938385389998,
"grad_norm": 1.111303687095642,
"learning_rate": 5.843851909425517e-06,
"loss": 2.002,
"step": 1890
},
{
"epoch": 2.5803038061102574,
"grad_norm": 1.1671916246414185,
"learning_rate": 5.8321809992873865e-06,
"loss": 1.9509,
"step": 1891
},
{
"epoch": 2.5816692268305172,
"grad_norm": 1.1496440172195435,
"learning_rate": 5.820545161864934e-06,
"loss": 1.9139,
"step": 1892
},
{
"epoch": 2.5830346475507766,
"grad_norm": 1.1134886741638184,
"learning_rate": 5.808944422399689e-06,
"loss": 2.0814,
"step": 1893
},
{
"epoch": 2.584400068271036,
"grad_norm": 1.149273157119751,
"learning_rate": 5.797378806057029e-06,
"loss": 1.9617,
"step": 1894
},
{
"epoch": 2.5857654889912953,
"grad_norm": 1.127129316329956,
"learning_rate": 5.785848337926157e-06,
"loss": 1.9473,
"step": 1895
},
{
"epoch": 2.587130909711555,
"grad_norm": 1.1746573448181152,
"learning_rate": 5.774353043020022e-06,
"loss": 1.9775,
"step": 1896
},
{
"epoch": 2.5884963304318145,
"grad_norm": 1.1435160636901855,
"learning_rate": 5.762892946275271e-06,
"loss": 1.9692,
"step": 1897
},
{
"epoch": 2.589861751152074,
"grad_norm": 1.141309142112732,
"learning_rate": 5.751468072552196e-06,
"loss": 1.9656,
"step": 1898
},
{
"epoch": 2.591227171872333,
"grad_norm": 1.1548477411270142,
"learning_rate": 5.740078446634683e-06,
"loss": 2.036,
"step": 1899
},
{
"epoch": 2.5925925925925926,
"grad_norm": 1.1896356344223022,
"learning_rate": 5.72872409323015e-06,
"loss": 1.9262,
"step": 1900
},
{
"epoch": 2.593958013312852,
"grad_norm": 1.0994958877563477,
"learning_rate": 5.717405036969504e-06,
"loss": 1.8957,
"step": 1901
},
{
"epoch": 2.5953234340331113,
"grad_norm": 1.1340858936309814,
"learning_rate": 5.706121302407077e-06,
"loss": 1.9231,
"step": 1902
},
{
"epoch": 2.5966888547533706,
"grad_norm": 1.1649575233459473,
"learning_rate": 5.694872914020581e-06,
"loss": 2.0434,
"step": 1903
},
{
"epoch": 2.5980542754736304,
"grad_norm": 1.1841235160827637,
"learning_rate": 5.683659896211058e-06,
"loss": 1.9368,
"step": 1904
},
{
"epoch": 2.59941969619389,
"grad_norm": 1.1741371154785156,
"learning_rate": 5.672482273302811e-06,
"loss": 2.101,
"step": 1905
},
{
"epoch": 2.600785116914149,
"grad_norm": 1.1330941915512085,
"learning_rate": 5.6613400695433605e-06,
"loss": 2.051,
"step": 1906
},
{
"epoch": 2.6021505376344085,
"grad_norm": 1.164360523223877,
"learning_rate": 5.6502333091033975e-06,
"loss": 1.9978,
"step": 1907
},
{
"epoch": 2.603515958354668,
"grad_norm": 1.2083550691604614,
"learning_rate": 5.6391620160767315e-06,
"loss": 2.0506,
"step": 1908
},
{
"epoch": 2.6048813790749277,
"grad_norm": 1.1650205850601196,
"learning_rate": 5.628126214480222e-06,
"loss": 2.0245,
"step": 1909
},
{
"epoch": 2.606246799795187,
"grad_norm": 1.1456691026687622,
"learning_rate": 5.617125928253745e-06,
"loss": 1.9931,
"step": 1910
},
{
"epoch": 2.6076122205154464,
"grad_norm": 1.1492379903793335,
"learning_rate": 5.606161181260123e-06,
"loss": 1.9363,
"step": 1911
},
{
"epoch": 2.6089776412357057,
"grad_norm": 1.1063522100448608,
"learning_rate": 5.595231997285099e-06,
"loss": 1.9836,
"step": 1912
},
{
"epoch": 2.610343061955965,
"grad_norm": 1.1654490232467651,
"learning_rate": 5.584338400037258e-06,
"loss": 1.9494,
"step": 1913
},
{
"epoch": 2.6117084826762245,
"grad_norm": 1.1163939237594604,
"learning_rate": 5.5734804131479894e-06,
"loss": 1.9548,
"step": 1914
},
{
"epoch": 2.613073903396484,
"grad_norm": 1.1431180238723755,
"learning_rate": 5.562658060171433e-06,
"loss": 1.9837,
"step": 1915
},
{
"epoch": 2.6144393241167436,
"grad_norm": 1.1585724353790283,
"learning_rate": 5.551871364584433e-06,
"loss": 1.9088,
"step": 1916
},
{
"epoch": 2.615804744837003,
"grad_norm": 1.0876420736312866,
"learning_rate": 5.541120349786477e-06,
"loss": 2.0206,
"step": 1917
},
{
"epoch": 2.6171701655572623,
"grad_norm": 1.1713416576385498,
"learning_rate": 5.53040503909965e-06,
"loss": 1.9895,
"step": 1918
},
{
"epoch": 2.6185355862775217,
"grad_norm": 1.161706805229187,
"learning_rate": 5.51972545576859e-06,
"loss": 1.9108,
"step": 1919
},
{
"epoch": 2.619901006997781,
"grad_norm": 1.113403081893921,
"learning_rate": 5.509081622960423e-06,
"loss": 1.9834,
"step": 1920
},
{
"epoch": 2.621266427718041,
"grad_norm": 1.1592470407485962,
"learning_rate": 5.498473563764734e-06,
"loss": 1.9883,
"step": 1921
},
{
"epoch": 2.6226318484383,
"grad_norm": 1.1119205951690674,
"learning_rate": 5.487901301193501e-06,
"loss": 1.958,
"step": 1922
},
{
"epoch": 2.6239972691585596,
"grad_norm": 1.1241275072097778,
"learning_rate": 5.477364858181034e-06,
"loss": 2.0574,
"step": 1923
},
{
"epoch": 2.625362689878819,
"grad_norm": 1.1901804208755493,
"learning_rate": 5.466864257583956e-06,
"loss": 2.0445,
"step": 1924
},
{
"epoch": 2.6267281105990783,
"grad_norm": 1.1756021976470947,
"learning_rate": 5.456399522181137e-06,
"loss": 2.0976,
"step": 1925
},
{
"epoch": 2.6280935313193377,
"grad_norm": 1.1963251829147339,
"learning_rate": 5.445970674673636e-06,
"loss": 2.0187,
"step": 1926
},
{
"epoch": 2.629458952039597,
"grad_norm": 1.1437311172485352,
"learning_rate": 5.435577737684668e-06,
"loss": 2.0543,
"step": 1927
},
{
"epoch": 2.6308243727598564,
"grad_norm": 1.1268452405929565,
"learning_rate": 5.4252207337595445e-06,
"loss": 1.9845,
"step": 1928
},
{
"epoch": 2.632189793480116,
"grad_norm": 1.1659737825393677,
"learning_rate": 5.414899685365627e-06,
"loss": 2.032,
"step": 1929
},
{
"epoch": 2.6335552142003755,
"grad_norm": 1.120768666267395,
"learning_rate": 5.404614614892283e-06,
"loss": 1.8942,
"step": 1930
},
{
"epoch": 2.634920634920635,
"grad_norm": 1.111363410949707,
"learning_rate": 5.394365544650826e-06,
"loss": 1.9973,
"step": 1931
},
{
"epoch": 2.6362860556408942,
"grad_norm": 1.1710448265075684,
"learning_rate": 5.384152496874483e-06,
"loss": 2.1005,
"step": 1932
},
{
"epoch": 2.637651476361154,
"grad_norm": 1.1965410709381104,
"learning_rate": 5.373975493718329e-06,
"loss": 1.9854,
"step": 1933
},
{
"epoch": 2.6390168970814134,
"grad_norm": 1.1734230518341064,
"learning_rate": 5.363834557259262e-06,
"loss": 1.9711,
"step": 1934
},
{
"epoch": 2.6403823178016728,
"grad_norm": 1.1404948234558105,
"learning_rate": 5.3537297094959236e-06,
"loss": 1.9996,
"step": 1935
},
{
"epoch": 2.641747738521932,
"grad_norm": 1.1832468509674072,
"learning_rate": 5.3436609723486765e-06,
"loss": 2.0819,
"step": 1936
},
{
"epoch": 2.6431131592421915,
"grad_norm": 1.1924104690551758,
"learning_rate": 5.333628367659548e-06,
"loss": 1.9471,
"step": 1937
},
{
"epoch": 2.644478579962451,
"grad_norm": 1.1803066730499268,
"learning_rate": 5.32363191719219e-06,
"loss": 2.035,
"step": 1938
},
{
"epoch": 2.64584400068271,
"grad_norm": 1.1220929622650146,
"learning_rate": 5.313671642631817e-06,
"loss": 1.985,
"step": 1939
},
{
"epoch": 2.6472094214029696,
"grad_norm": 1.1628798246383667,
"learning_rate": 5.303747565585169e-06,
"loss": 1.9805,
"step": 1940
},
{
"epoch": 2.6485748421232294,
"grad_norm": 1.1286931037902832,
"learning_rate": 5.2938597075804606e-06,
"loss": 1.8639,
"step": 1941
},
{
"epoch": 2.6499402628434887,
"grad_norm": 1.1496410369873047,
"learning_rate": 5.284008090067348e-06,
"loss": 1.9788,
"step": 1942
},
{
"epoch": 2.651305683563748,
"grad_norm": 1.1481356620788574,
"learning_rate": 5.274192734416857e-06,
"loss": 1.9577,
"step": 1943
},
{
"epoch": 2.6526711042840074,
"grad_norm": 1.1670453548431396,
"learning_rate": 5.2644136619213595e-06,
"loss": 2.0576,
"step": 1944
},
{
"epoch": 2.654036525004267,
"grad_norm": 1.1816532611846924,
"learning_rate": 5.254670893794515e-06,
"loss": 2.0647,
"step": 1945
},
{
"epoch": 2.6554019457245266,
"grad_norm": 1.1702381372451782,
"learning_rate": 5.2449644511712325e-06,
"loss": 1.8714,
"step": 1946
},
{
"epoch": 2.656767366444786,
"grad_norm": 1.127424716949463,
"learning_rate": 5.235294355107616e-06,
"loss": 1.9994,
"step": 1947
},
{
"epoch": 2.6581327871650453,
"grad_norm": 1.1549973487854004,
"learning_rate": 5.225660626580924e-06,
"loss": 2.0124,
"step": 1948
},
{
"epoch": 2.6594982078853047,
"grad_norm": 1.169043779373169,
"learning_rate": 5.216063286489525e-06,
"loss": 1.8905,
"step": 1949
},
{
"epoch": 2.660863628605564,
"grad_norm": 1.1068698167800903,
"learning_rate": 5.206502355652844e-06,
"loss": 1.9398,
"step": 1950
},
{
"epoch": 2.6622290493258234,
"grad_norm": 1.1375091075897217,
"learning_rate": 5.196977854811341e-06,
"loss": 2.0095,
"step": 1951
},
{
"epoch": 2.6635944700460827,
"grad_norm": 1.142715573310852,
"learning_rate": 5.1874898046264255e-06,
"loss": 2.0459,
"step": 1952
},
{
"epoch": 2.6649598907663425,
"grad_norm": 1.2181109189987183,
"learning_rate": 5.178038225680451e-06,
"loss": 2.0845,
"step": 1953
},
{
"epoch": 2.666325311486602,
"grad_norm": 1.161192536354065,
"learning_rate": 5.168623138476645e-06,
"loss": 1.9103,
"step": 1954
},
{
"epoch": 2.6676907322068613,
"grad_norm": 1.176962971687317,
"learning_rate": 5.159244563439085e-06,
"loss": 2.0789,
"step": 1955
},
{
"epoch": 2.6690561529271206,
"grad_norm": 1.1671562194824219,
"learning_rate": 5.149902520912631e-06,
"loss": 1.9887,
"step": 1956
},
{
"epoch": 2.67042157364738,
"grad_norm": 1.1885942220687866,
"learning_rate": 5.140597031162901e-06,
"loss": 2.0267,
"step": 1957
},
{
"epoch": 2.67178699436764,
"grad_norm": 1.159234881401062,
"learning_rate": 5.131328114376213e-06,
"loss": 2.1069,
"step": 1958
},
{
"epoch": 2.673152415087899,
"grad_norm": 1.2258284091949463,
"learning_rate": 5.1220957906595556e-06,
"loss": 1.8931,
"step": 1959
},
{
"epoch": 2.6745178358081585,
"grad_norm": 1.1501591205596924,
"learning_rate": 5.112900080040529e-06,
"loss": 1.9747,
"step": 1960
},
{
"epoch": 2.675883256528418,
"grad_norm": 1.1986619234085083,
"learning_rate": 5.103741002467314e-06,
"loss": 1.9653,
"step": 1961
},
{
"epoch": 2.677248677248677,
"grad_norm": 1.1245447397232056,
"learning_rate": 5.094618577808621e-06,
"loss": 1.9347,
"step": 1962
},
{
"epoch": 2.6786140979689366,
"grad_norm": 1.1675920486450195,
"learning_rate": 5.085532825853651e-06,
"loss": 2.0863,
"step": 1963
},
{
"epoch": 2.679979518689196,
"grad_norm": 1.1118495464324951,
"learning_rate": 5.076483766312049e-06,
"loss": 2.0225,
"step": 1964
},
{
"epoch": 2.6813449394094553,
"grad_norm": 1.1714025735855103,
"learning_rate": 5.0674714188138645e-06,
"loss": 2.0317,
"step": 1965
},
{
"epoch": 2.682710360129715,
"grad_norm": 1.1392548084259033,
"learning_rate": 5.05849580290951e-06,
"loss": 1.9913,
"step": 1966
},
{
"epoch": 2.6840757808499744,
"grad_norm": 1.13945734500885,
"learning_rate": 5.0495569380697085e-06,
"loss": 1.9602,
"step": 1967
},
{
"epoch": 2.685441201570234,
"grad_norm": 1.1369609832763672,
"learning_rate": 5.040654843685476e-06,
"loss": 2.0809,
"step": 1968
},
{
"epoch": 2.686806622290493,
"grad_norm": 1.1467934846878052,
"learning_rate": 5.031789539068047e-06,
"loss": 1.9797,
"step": 1969
},
{
"epoch": 2.688172043010753,
"grad_norm": 1.174498200416565,
"learning_rate": 5.022961043448851e-06,
"loss": 1.9627,
"step": 1970
},
{
"epoch": 2.6895374637310123,
"grad_norm": 1.1654930114746094,
"learning_rate": 5.014169375979469e-06,
"loss": 1.9345,
"step": 1971
},
{
"epoch": 2.6909028844512717,
"grad_norm": 1.1670429706573486,
"learning_rate": 5.005414555731598e-06,
"loss": 2.1003,
"step": 1972
},
{
"epoch": 2.692268305171531,
"grad_norm": 1.1597691774368286,
"learning_rate": 4.996696601696996e-06,
"loss": 1.8764,
"step": 1973
},
{
"epoch": 2.6936337258917904,
"grad_norm": 1.083889126777649,
"learning_rate": 4.9880155327874445e-06,
"loss": 2.0504,
"step": 1974
},
{
"epoch": 2.6949991466120498,
"grad_norm": 1.2172187566757202,
"learning_rate": 4.97937136783472e-06,
"loss": 1.8598,
"step": 1975
},
{
"epoch": 2.696364567332309,
"grad_norm": 1.0814568996429443,
"learning_rate": 4.97076412559053e-06,
"loss": 1.8581,
"step": 1976
},
{
"epoch": 2.6977299880525685,
"grad_norm": 1.1346497535705566,
"learning_rate": 4.9621938247265e-06,
"loss": 2.0118,
"step": 1977
},
{
"epoch": 2.6990954087728283,
"grad_norm": 1.1771613359451294,
"learning_rate": 4.953660483834112e-06,
"loss": 2.0654,
"step": 1978
},
{
"epoch": 2.7004608294930876,
"grad_norm": 1.1661196947097778,
"learning_rate": 4.9451641214246696e-06,
"loss": 1.9035,
"step": 1979
},
{
"epoch": 2.701826250213347,
"grad_norm": 1.1093000173568726,
"learning_rate": 4.936704755929259e-06,
"loss": 2.0255,
"step": 1980
},
{
"epoch": 2.7031916709336063,
"grad_norm": 1.183701515197754,
"learning_rate": 4.928282405698719e-06,
"loss": 1.9673,
"step": 1981
},
{
"epoch": 2.7045570916538657,
"grad_norm": 1.1400455236434937,
"learning_rate": 4.919897089003578e-06,
"loss": 1.9242,
"step": 1982
},
{
"epoch": 2.7059225123741255,
"grad_norm": 1.1725705862045288,
"learning_rate": 4.911548824034037e-06,
"loss": 2.0549,
"step": 1983
},
{
"epoch": 2.707287933094385,
"grad_norm": 1.1636077165603638,
"learning_rate": 4.903237628899912e-06,
"loss": 2.0201,
"step": 1984
},
{
"epoch": 2.7086533538146442,
"grad_norm": 1.2140169143676758,
"learning_rate": 4.8949635216306176e-06,
"loss": 1.962,
"step": 1985
},
{
"epoch": 2.7100187745349036,
"grad_norm": 1.1270188093185425,
"learning_rate": 4.886726520175104e-06,
"loss": 1.9203,
"step": 1986
},
{
"epoch": 2.711384195255163,
"grad_norm": 1.0895811319351196,
"learning_rate": 4.878526642401829e-06,
"loss": 1.9967,
"step": 1987
},
{
"epoch": 2.7127496159754223,
"grad_norm": 1.1982686519622803,
"learning_rate": 4.870363906098718e-06,
"loss": 2.0042,
"step": 1988
},
{
"epoch": 2.7141150366956817,
"grad_norm": 1.1646535396575928,
"learning_rate": 4.862238328973129e-06,
"loss": 1.9609,
"step": 1989
},
{
"epoch": 2.715480457415941,
"grad_norm": 1.159096598625183,
"learning_rate": 4.854149928651809e-06,
"loss": 2.0994,
"step": 1990
},
{
"epoch": 2.716845878136201,
"grad_norm": 1.1897298097610474,
"learning_rate": 4.846098722680857e-06,
"loss": 1.9678,
"step": 1991
},
{
"epoch": 2.71821129885646,
"grad_norm": 1.1401512622833252,
"learning_rate": 4.838084728525688e-06,
"loss": 2.0579,
"step": 1992
},
{
"epoch": 2.7195767195767195,
"grad_norm": 1.1835845708847046,
"learning_rate": 4.830107963570993e-06,
"loss": 2.0894,
"step": 1993
},
{
"epoch": 2.720942140296979,
"grad_norm": 1.1764168739318848,
"learning_rate": 4.822168445120702e-06,
"loss": 1.926,
"step": 1994
},
{
"epoch": 2.7223075610172387,
"grad_norm": 1.1647770404815674,
"learning_rate": 4.814266190397946e-06,
"loss": 2.0599,
"step": 1995
},
{
"epoch": 2.723672981737498,
"grad_norm": 1.181841492652893,
"learning_rate": 4.806401216545022e-06,
"loss": 1.9893,
"step": 1996
},
{
"epoch": 2.7250384024577574,
"grad_norm": 1.1662906408309937,
"learning_rate": 4.798573540623349e-06,
"loss": 1.9609,
"step": 1997
},
{
"epoch": 2.7264038231780168,
"grad_norm": 1.1738672256469727,
"learning_rate": 4.790783179613446e-06,
"loss": 1.8933,
"step": 1998
},
{
"epoch": 2.727769243898276,
"grad_norm": 1.157853126525879,
"learning_rate": 4.783030150414873e-06,
"loss": 2.1409,
"step": 1999
},
{
"epoch": 2.7291346646185355,
"grad_norm": 1.1880639791488647,
"learning_rate": 4.775314469846213e-06,
"loss": 2.0107,
"step": 2000
},
{
"epoch": 2.730500085338795,
"grad_norm": 1.174072265625,
"learning_rate": 4.767636154645027e-06,
"loss": 1.9685,
"step": 2001
},
{
"epoch": 2.731865506059054,
"grad_norm": 1.1626956462860107,
"learning_rate": 4.759995221467824e-06,
"loss": 2.024,
"step": 2002
},
{
"epoch": 2.733230926779314,
"grad_norm": 1.17644464969635,
"learning_rate": 4.752391686890016e-06,
"loss": 1.9916,
"step": 2003
},
{
"epoch": 2.7345963474995734,
"grad_norm": 1.1642102003097534,
"learning_rate": 4.744825567405886e-06,
"loss": 2.0728,
"step": 2004
},
{
"epoch": 2.7359617682198327,
"grad_norm": 1.1925033330917358,
"learning_rate": 4.737296879428551e-06,
"loss": 2.0926,
"step": 2005
},
{
"epoch": 2.737327188940092,
"grad_norm": 1.1904150247573853,
"learning_rate": 4.729805639289938e-06,
"loss": 1.9866,
"step": 2006
},
{
"epoch": 2.7386926096603514,
"grad_norm": 1.1785517930984497,
"learning_rate": 4.7223518632407304e-06,
"loss": 1.9798,
"step": 2007
},
{
"epoch": 2.7400580303806112,
"grad_norm": 1.1711405515670776,
"learning_rate": 4.714935567450339e-06,
"loss": 2.0357,
"step": 2008
},
{
"epoch": 2.7414234511008706,
"grad_norm": 1.15545654296875,
"learning_rate": 4.7075567680068766e-06,
"loss": 1.9012,
"step": 2009
},
{
"epoch": 2.74278887182113,
"grad_norm": 1.1282776594161987,
"learning_rate": 4.700215480917109e-06,
"loss": 1.9274,
"step": 2010
},
{
"epoch": 2.7441542925413893,
"grad_norm": 1.091373085975647,
"learning_rate": 4.692911722106433e-06,
"loss": 1.9082,
"step": 2011
},
{
"epoch": 2.7455197132616487,
"grad_norm": 1.1047842502593994,
"learning_rate": 4.685645507418829e-06,
"loss": 1.8821,
"step": 2012
},
{
"epoch": 2.746885133981908,
"grad_norm": 1.1439313888549805,
"learning_rate": 4.678416852616837e-06,
"loss": 1.9486,
"step": 2013
},
{
"epoch": 2.7482505547021674,
"grad_norm": 1.1958791017532349,
"learning_rate": 4.671225773381514e-06,
"loss": 2.0148,
"step": 2014
},
{
"epoch": 2.749615975422427,
"grad_norm": 1.14925217628479,
"learning_rate": 4.664072285312418e-06,
"loss": 1.9155,
"step": 2015
},
{
"epoch": 2.7509813961426866,
"grad_norm": 1.108605146408081,
"learning_rate": 4.656956403927546e-06,
"loss": 2.0012,
"step": 2016
},
{
"epoch": 2.752346816862946,
"grad_norm": 1.1260026693344116,
"learning_rate": 4.649878144663317e-06,
"loss": 2.0226,
"step": 2017
},
{
"epoch": 2.7537122375832053,
"grad_norm": 1.1353247165679932,
"learning_rate": 4.642837522874542e-06,
"loss": 2.052,
"step": 2018
},
{
"epoch": 2.7550776583034646,
"grad_norm": 1.1638880968093872,
"learning_rate": 4.635834553834389e-06,
"loss": 1.991,
"step": 2019
},
{
"epoch": 2.7564430790237244,
"grad_norm": 1.1422680616378784,
"learning_rate": 4.628869252734334e-06,
"loss": 1.9832,
"step": 2020
},
{
"epoch": 2.757808499743984,
"grad_norm": 1.154308795928955,
"learning_rate": 4.621941634684151e-06,
"loss": 1.9101,
"step": 2021
},
{
"epoch": 2.759173920464243,
"grad_norm": 1.1312297582626343,
"learning_rate": 4.615051714711862e-06,
"loss": 2.0545,
"step": 2022
},
{
"epoch": 2.7605393411845025,
"grad_norm": 1.1639527082443237,
"learning_rate": 4.608199507763714e-06,
"loss": 2.0445,
"step": 2023
},
{
"epoch": 2.761904761904762,
"grad_norm": 1.2108618021011353,
"learning_rate": 4.6013850287041444e-06,
"loss": 2.0027,
"step": 2024
},
{
"epoch": 2.763270182625021,
"grad_norm": 1.1489726305007935,
"learning_rate": 4.594608292315743e-06,
"loss": 1.9939,
"step": 2025
},
{
"epoch": 2.7646356033452806,
"grad_norm": 1.1837903261184692,
"learning_rate": 4.5878693132992285e-06,
"loss": 2.1202,
"step": 2026
},
{
"epoch": 2.76600102406554,
"grad_norm": 1.156774878501892,
"learning_rate": 4.581168106273413e-06,
"loss": 2.0349,
"step": 2027
},
{
"epoch": 2.7673664447857997,
"grad_norm": 1.1649513244628906,
"learning_rate": 4.574504685775173e-06,
"loss": 2.0197,
"step": 2028
},
{
"epoch": 2.768731865506059,
"grad_norm": 1.1609634160995483,
"learning_rate": 4.5678790662594055e-06,
"loss": 2.0405,
"step": 2029
},
{
"epoch": 2.7700972862263185,
"grad_norm": 1.1418075561523438,
"learning_rate": 4.5612912620990145e-06,
"loss": 2.0005,
"step": 2030
},
{
"epoch": 2.771462706946578,
"grad_norm": 1.1754897832870483,
"learning_rate": 4.5547412875848645e-06,
"loss": 2.0248,
"step": 2031
},
{
"epoch": 2.7728281276668376,
"grad_norm": 1.1675280332565308,
"learning_rate": 4.548229156925768e-06,
"loss": 1.929,
"step": 2032
},
{
"epoch": 2.774193548387097,
"grad_norm": 1.1345016956329346,
"learning_rate": 4.541754884248436e-06,
"loss": 2.0706,
"step": 2033
},
{
"epoch": 2.7755589691073563,
"grad_norm": 1.1708017587661743,
"learning_rate": 4.535318483597449e-06,
"loss": 1.9916,
"step": 2034
},
{
"epoch": 2.7769243898276157,
"grad_norm": 1.1320682764053345,
"learning_rate": 4.52891996893524e-06,
"loss": 1.8922,
"step": 2035
},
{
"epoch": 2.778289810547875,
"grad_norm": 1.1360291242599487,
"learning_rate": 4.522559354142054e-06,
"loss": 1.9984,
"step": 2036
},
{
"epoch": 2.7796552312681344,
"grad_norm": 1.156357765197754,
"learning_rate": 4.51623665301592e-06,
"loss": 2.0651,
"step": 2037
},
{
"epoch": 2.7810206519883938,
"grad_norm": 1.1358588933944702,
"learning_rate": 4.50995187927262e-06,
"loss": 1.9121,
"step": 2038
},
{
"epoch": 2.782386072708653,
"grad_norm": 1.1554834842681885,
"learning_rate": 4.503705046545664e-06,
"loss": 2.0025,
"step": 2039
},
{
"epoch": 2.783751493428913,
"grad_norm": 1.154569149017334,
"learning_rate": 4.497496168386252e-06,
"loss": 1.972,
"step": 2040
},
{
"epoch": 2.7851169141491723,
"grad_norm": 1.161489486694336,
"learning_rate": 4.491325258263255e-06,
"loss": 2.0566,
"step": 2041
},
{
"epoch": 2.7864823348694316,
"grad_norm": 1.1591715812683105,
"learning_rate": 4.485192329563172e-06,
"loss": 1.9892,
"step": 2042
},
{
"epoch": 2.787847755589691,
"grad_norm": 1.1330506801605225,
"learning_rate": 4.479097395590118e-06,
"loss": 1.9347,
"step": 2043
},
{
"epoch": 2.7892131763099504,
"grad_norm": 1.129957914352417,
"learning_rate": 4.4730404695657845e-06,
"loss": 1.9819,
"step": 2044
},
{
"epoch": 2.79057859703021,
"grad_norm": 1.1663581132888794,
"learning_rate": 4.467021564629407e-06,
"loss": 1.9284,
"step": 2045
},
{
"epoch": 2.7919440177504695,
"grad_norm": 1.1699559688568115,
"learning_rate": 4.461040693837746e-06,
"loss": 2.0289,
"step": 2046
},
{
"epoch": 2.793309438470729,
"grad_norm": 1.1938339471817017,
"learning_rate": 4.455097870165058e-06,
"loss": 2.069,
"step": 2047
},
{
"epoch": 2.7946748591909882,
"grad_norm": 1.177240252494812,
"learning_rate": 4.449193106503059e-06,
"loss": 1.9412,
"step": 2048
},
{
"epoch": 2.7960402799112476,
"grad_norm": 1.1549038887023926,
"learning_rate": 4.4433264156609035e-06,
"loss": 2.0647,
"step": 2049
},
{
"epoch": 2.797405700631507,
"grad_norm": 1.1525830030441284,
"learning_rate": 4.437497810365158e-06,
"loss": 1.9951,
"step": 2050
},
{
"epoch": 2.7987711213517663,
"grad_norm": 1.1260254383087158,
"learning_rate": 4.431707303259766e-06,
"loss": 1.994,
"step": 2051
},
{
"epoch": 2.8001365420720257,
"grad_norm": 1.1267759799957275,
"learning_rate": 4.425954906906022e-06,
"loss": 1.9897,
"step": 2052
},
{
"epoch": 2.8015019627922855,
"grad_norm": 1.1501511335372925,
"learning_rate": 4.420240633782562e-06,
"loss": 2.1075,
"step": 2053
},
{
"epoch": 2.802867383512545,
"grad_norm": 1.195828914642334,
"learning_rate": 4.414564496285303e-06,
"loss": 2.1901,
"step": 2054
},
{
"epoch": 2.804232804232804,
"grad_norm": 1.1991605758666992,
"learning_rate": 4.408926506727444e-06,
"loss": 1.9727,
"step": 2055
},
{
"epoch": 2.8055982249530635,
"grad_norm": 1.1188796758651733,
"learning_rate": 4.403326677339432e-06,
"loss": 1.885,
"step": 2056
},
{
"epoch": 2.8069636456733233,
"grad_norm": 1.1417156457901,
"learning_rate": 4.397765020268929e-06,
"loss": 1.9688,
"step": 2057
},
{
"epoch": 2.8083290663935827,
"grad_norm": 1.151779055595398,
"learning_rate": 4.392241547580792e-06,
"loss": 2.0562,
"step": 2058
},
{
"epoch": 2.809694487113842,
"grad_norm": 1.1420207023620605,
"learning_rate": 4.386756271257042e-06,
"loss": 1.9388,
"step": 2059
},
{
"epoch": 2.8110599078341014,
"grad_norm": 1.1549649238586426,
"learning_rate": 4.381309203196847e-06,
"loss": 1.9432,
"step": 2060
},
{
"epoch": 2.812425328554361,
"grad_norm": 1.118360996246338,
"learning_rate": 4.3759003552164834e-06,
"loss": 1.8272,
"step": 2061
},
{
"epoch": 2.81379074927462,
"grad_norm": 1.1257715225219727,
"learning_rate": 4.3705297390493205e-06,
"loss": 1.9425,
"step": 2062
},
{
"epoch": 2.8151561699948795,
"grad_norm": 1.1140422821044922,
"learning_rate": 4.365197366345796e-06,
"loss": 1.9146,
"step": 2063
},
{
"epoch": 2.816521590715139,
"grad_norm": 1.1124956607818604,
"learning_rate": 4.359903248673378e-06,
"loss": 1.9796,
"step": 2064
},
{
"epoch": 2.8178870114353987,
"grad_norm": 1.1386439800262451,
"learning_rate": 4.354647397516551e-06,
"loss": 2.0464,
"step": 2065
},
{
"epoch": 2.819252432155658,
"grad_norm": 1.1820223331451416,
"learning_rate": 4.349429824276789e-06,
"loss": 1.9017,
"step": 2066
},
{
"epoch": 2.8206178528759174,
"grad_norm": 1.1546006202697754,
"learning_rate": 4.344250540272537e-06,
"loss": 2.047,
"step": 2067
},
{
"epoch": 2.8219832735961767,
"grad_norm": 1.1624810695648193,
"learning_rate": 4.339109556739164e-06,
"loss": 1.9075,
"step": 2068
},
{
"epoch": 2.823348694316436,
"grad_norm": 1.128680944442749,
"learning_rate": 4.334006884828971e-06,
"loss": 2.1168,
"step": 2069
},
{
"epoch": 2.824714115036696,
"grad_norm": 1.2393641471862793,
"learning_rate": 4.328942535611134e-06,
"loss": 2.1022,
"step": 2070
},
{
"epoch": 2.8260795357569553,
"grad_norm": 1.166141152381897,
"learning_rate": 4.323916520071709e-06,
"loss": 1.9522,
"step": 2071
},
{
"epoch": 2.8274449564772146,
"grad_norm": 1.1621108055114746,
"learning_rate": 4.318928849113587e-06,
"loss": 1.9545,
"step": 2072
},
{
"epoch": 2.828810377197474,
"grad_norm": 1.1608256101608276,
"learning_rate": 4.313979533556485e-06,
"loss": 1.9749,
"step": 2073
},
{
"epoch": 2.8301757979177333,
"grad_norm": 1.126189947128296,
"learning_rate": 4.309068584136909e-06,
"loss": 1.9636,
"step": 2074
},
{
"epoch": 2.8315412186379927,
"grad_norm": 1.138491153717041,
"learning_rate": 4.304196011508142e-06,
"loss": 1.9372,
"step": 2075
},
{
"epoch": 2.832906639358252,
"grad_norm": 1.1437461376190186,
"learning_rate": 4.299361826240214e-06,
"loss": 1.8707,
"step": 2076
},
{
"epoch": 2.834272060078512,
"grad_norm": 1.1284011602401733,
"learning_rate": 4.2945660388198866e-06,
"loss": 2.019,
"step": 2077
},
{
"epoch": 2.835637480798771,
"grad_norm": 1.2004722356796265,
"learning_rate": 4.289808659650617e-06,
"loss": 2.0232,
"step": 2078
},
{
"epoch": 2.8370029015190306,
"grad_norm": 1.1668102741241455,
"learning_rate": 4.285089699052549e-06,
"loss": 2.0075,
"step": 2079
},
{
"epoch": 2.83836832223929,
"grad_norm": 1.1123415231704712,
"learning_rate": 4.28040916726249e-06,
"loss": 2.0328,
"step": 2080
},
{
"epoch": 2.8397337429595493,
"grad_norm": 1.202068567276001,
"learning_rate": 4.275767074433873e-06,
"loss": 1.9768,
"step": 2081
},
{
"epoch": 2.841099163679809,
"grad_norm": 1.1408039331436157,
"learning_rate": 4.271163430636755e-06,
"loss": 2.0391,
"step": 2082
},
{
"epoch": 2.8424645844000684,
"grad_norm": 1.192373514175415,
"learning_rate": 4.2665982458577795e-06,
"loss": 2.0386,
"step": 2083
},
{
"epoch": 2.843830005120328,
"grad_norm": 1.2188798189163208,
"learning_rate": 4.262071530000165e-06,
"loss": 2.0595,
"step": 2084
},
{
"epoch": 2.845195425840587,
"grad_norm": 1.147842288017273,
"learning_rate": 4.2575832928836795e-06,
"loss": 1.8756,
"step": 2085
},
{
"epoch": 2.8465608465608465,
"grad_norm": 1.1598711013793945,
"learning_rate": 4.253133544244619e-06,
"loss": 2.0826,
"step": 2086
},
{
"epoch": 2.847926267281106,
"grad_norm": 1.172425627708435,
"learning_rate": 4.2487222937357844e-06,
"loss": 2.0506,
"step": 2087
},
{
"epoch": 2.8492916880013652,
"grad_norm": 1.1938341856002808,
"learning_rate": 4.2443495509264666e-06,
"loss": 1.9169,
"step": 2088
},
{
"epoch": 2.8506571087216246,
"grad_norm": 1.1228644847869873,
"learning_rate": 4.2400153253024175e-06,
"loss": 2.0153,
"step": 2089
},
{
"epoch": 2.8520225294418844,
"grad_norm": 1.133897066116333,
"learning_rate": 4.235719626265841e-06,
"loss": 1.9922,
"step": 2090
},
{
"epoch": 2.8533879501621437,
"grad_norm": 1.1309428215026855,
"learning_rate": 4.2314624631353575e-06,
"loss": 1.981,
"step": 2091
},
{
"epoch": 2.854753370882403,
"grad_norm": 1.150156021118164,
"learning_rate": 4.227243845146e-06,
"loss": 1.936,
"step": 2092
},
{
"epoch": 2.8561187916026625,
"grad_norm": 1.1496373414993286,
"learning_rate": 4.223063781449177e-06,
"loss": 2.0416,
"step": 2093
},
{
"epoch": 2.8574842123229223,
"grad_norm": 1.187489628791809,
"learning_rate": 4.218922281112668e-06,
"loss": 2.0135,
"step": 2094
},
{
"epoch": 2.8588496330431816,
"grad_norm": 1.1455422639846802,
"learning_rate": 4.214819353120597e-06,
"loss": 2.0438,
"step": 2095
},
{
"epoch": 2.860215053763441,
"grad_norm": 1.1550781726837158,
"learning_rate": 4.210755006373406e-06,
"loss": 2.0642,
"step": 2096
},
{
"epoch": 2.8615804744837003,
"grad_norm": 1.1155083179473877,
"learning_rate": 4.2067292496878556e-06,
"loss": 2.0717,
"step": 2097
},
{
"epoch": 2.8629458952039597,
"grad_norm": 1.180711030960083,
"learning_rate": 4.202742091796985e-06,
"loss": 1.9533,
"step": 2098
},
{
"epoch": 2.864311315924219,
"grad_norm": 1.1659834384918213,
"learning_rate": 4.198793541350099e-06,
"loss": 1.9332,
"step": 2099
},
{
"epoch": 2.8656767366444784,
"grad_norm": 1.160171627998352,
"learning_rate": 4.194883606912757e-06,
"loss": 1.9365,
"step": 2100
},
{
"epoch": 2.8670421573647378,
"grad_norm": 1.144218921661377,
"learning_rate": 4.191012296966747e-06,
"loss": 1.9799,
"step": 2101
},
{
"epoch": 2.8684075780849976,
"grad_norm": 1.1516847610473633,
"learning_rate": 4.187179619910074e-06,
"loss": 2.0546,
"step": 2102
},
{
"epoch": 2.869772998805257,
"grad_norm": 1.1319117546081543,
"learning_rate": 4.183385584056927e-06,
"loss": 1.9454,
"step": 2103
},
{
"epoch": 2.8711384195255163,
"grad_norm": 1.154268503189087,
"learning_rate": 4.1796301976376785e-06,
"loss": 2.0057,
"step": 2104
},
{
"epoch": 2.8725038402457757,
"grad_norm": 1.1424412727355957,
"learning_rate": 4.175913468798858e-06,
"loss": 1.9848,
"step": 2105
},
{
"epoch": 2.873869260966035,
"grad_norm": 1.1737909317016602,
"learning_rate": 4.172235405603132e-06,
"loss": 1.8969,
"step": 2106
},
{
"epoch": 2.875234681686295,
"grad_norm": 1.1386523246765137,
"learning_rate": 4.168596016029295e-06,
"loss": 2.1477,
"step": 2107
},
{
"epoch": 2.876600102406554,
"grad_norm": 1.2063666582107544,
"learning_rate": 4.164995307972244e-06,
"loss": 1.9645,
"step": 2108
},
{
"epoch": 2.8779655231268135,
"grad_norm": 1.162590742111206,
"learning_rate": 4.161433289242963e-06,
"loss": 1.9663,
"step": 2109
},
{
"epoch": 2.879330943847073,
"grad_norm": 1.1674400568008423,
"learning_rate": 4.157909967568514e-06,
"loss": 1.9663,
"step": 2110
},
{
"epoch": 2.8806963645673322,
"grad_norm": 1.1457250118255615,
"learning_rate": 4.154425350592006e-06,
"loss": 1.9584,
"step": 2111
},
{
"epoch": 2.8820617852875916,
"grad_norm": 1.1521270275115967,
"learning_rate": 4.150979445872593e-06,
"loss": 1.9733,
"step": 2112
},
{
"epoch": 2.883427206007851,
"grad_norm": 1.1624807119369507,
"learning_rate": 4.1475722608854405e-06,
"loss": 1.9833,
"step": 2113
},
{
"epoch": 2.8847926267281108,
"grad_norm": 1.1142884492874146,
"learning_rate": 4.144203803021736e-06,
"loss": 1.9405,
"step": 2114
},
{
"epoch": 2.88615804744837,
"grad_norm": 1.1681132316589355,
"learning_rate": 4.140874079588641e-06,
"loss": 1.992,
"step": 2115
},
{
"epoch": 2.8875234681686295,
"grad_norm": 1.1797648668289185,
"learning_rate": 4.137583097809299e-06,
"loss": 2.0313,
"step": 2116
},
{
"epoch": 2.888888888888889,
"grad_norm": 1.16402268409729,
"learning_rate": 4.134330864822808e-06,
"loss": 1.9758,
"step": 2117
},
{
"epoch": 2.890254309609148,
"grad_norm": 1.1705846786499023,
"learning_rate": 4.131117387684213e-06,
"loss": 2.0422,
"step": 2118
},
{
"epoch": 2.891619730329408,
"grad_norm": 1.1587177515029907,
"learning_rate": 4.127942673364479e-06,
"loss": 1.8711,
"step": 2119
},
{
"epoch": 2.8929851510496674,
"grad_norm": 1.1352663040161133,
"learning_rate": 4.124806728750492e-06,
"loss": 1.9578,
"step": 2120
},
{
"epoch": 2.8943505717699267,
"grad_norm": 1.1301008462905884,
"learning_rate": 4.121709560645026e-06,
"loss": 1.9596,
"step": 2121
},
{
"epoch": 2.895715992490186,
"grad_norm": 1.1259797811508179,
"learning_rate": 4.118651175766747e-06,
"loss": 1.9879,
"step": 2122
},
{
"epoch": 2.8970814132104454,
"grad_norm": 1.1493821144104004,
"learning_rate": 4.115631580750181e-06,
"loss": 1.9616,
"step": 2123
},
{
"epoch": 2.898446833930705,
"grad_norm": 1.1439268589019775,
"learning_rate": 4.112650782145711e-06,
"loss": 2.0159,
"step": 2124
},
{
"epoch": 2.899812254650964,
"grad_norm": 1.1418503522872925,
"learning_rate": 4.1097087864195576e-06,
"loss": 1.9343,
"step": 2125
},
{
"epoch": 2.9011776753712235,
"grad_norm": 1.1514840126037598,
"learning_rate": 4.106805599953769e-06,
"loss": 2.0459,
"step": 2126
},
{
"epoch": 2.9025430960914833,
"grad_norm": 1.2038418054580688,
"learning_rate": 4.103941229046205e-06,
"loss": 2.033,
"step": 2127
},
{
"epoch": 2.9039085168117427,
"grad_norm": 1.1307311058044434,
"learning_rate": 4.101115679910519e-06,
"loss": 1.9953,
"step": 2128
},
{
"epoch": 2.905273937532002,
"grad_norm": 1.1434341669082642,
"learning_rate": 4.098328958676152e-06,
"loss": 2.0055,
"step": 2129
},
{
"epoch": 2.9066393582522614,
"grad_norm": 1.1366305351257324,
"learning_rate": 4.095581071388315e-06,
"loss": 1.9633,
"step": 2130
},
{
"epoch": 2.9080047789725207,
"grad_norm": 1.1424757242202759,
"learning_rate": 4.092872024007979e-06,
"loss": 1.9802,
"step": 2131
},
{
"epoch": 2.9093701996927805,
"grad_norm": 1.1709935665130615,
"learning_rate": 4.090201822411856e-06,
"loss": 2.0136,
"step": 2132
},
{
"epoch": 2.91073562041304,
"grad_norm": 1.1127263307571411,
"learning_rate": 4.087570472392394e-06,
"loss": 2.0174,
"step": 2133
},
{
"epoch": 2.9121010411332993,
"grad_norm": 1.1392323970794678,
"learning_rate": 4.084977979657756e-06,
"loss": 1.9298,
"step": 2134
},
{
"epoch": 2.9134664618535586,
"grad_norm": 1.1422020196914673,
"learning_rate": 4.082424349831818e-06,
"loss": 2.0651,
"step": 2135
},
{
"epoch": 2.914831882573818,
"grad_norm": 1.173170804977417,
"learning_rate": 4.079909588454145e-06,
"loss": 2.1054,
"step": 2136
},
{
"epoch": 2.9161973032940773,
"grad_norm": 1.165525197982788,
"learning_rate": 4.077433700979991e-06,
"loss": 2.0495,
"step": 2137
},
{
"epoch": 2.9175627240143367,
"grad_norm": 1.183082938194275,
"learning_rate": 4.074996692780277e-06,
"loss": 2.0761,
"step": 2138
},
{
"epoch": 2.9189281447345965,
"grad_norm": 1.1737467050552368,
"learning_rate": 4.072598569141582e-06,
"loss": 1.9768,
"step": 2139
},
{
"epoch": 2.920293565454856,
"grad_norm": 1.1471754312515259,
"learning_rate": 4.070239335266142e-06,
"loss": 1.893,
"step": 2140
},
{
"epoch": 2.921658986175115,
"grad_norm": 1.1069848537445068,
"learning_rate": 4.067918996271815e-06,
"loss": 1.9328,
"step": 2141
},
{
"epoch": 2.9230244068953746,
"grad_norm": 1.105900526046753,
"learning_rate": 4.0656375571921006e-06,
"loss": 2.0201,
"step": 2142
},
{
"epoch": 2.924389827615634,
"grad_norm": 1.1653119325637817,
"learning_rate": 4.063395022976103e-06,
"loss": 1.9834,
"step": 2143
},
{
"epoch": 2.9257552483358937,
"grad_norm": 1.150187611579895,
"learning_rate": 4.061191398488533e-06,
"loss": 1.9481,
"step": 2144
},
{
"epoch": 2.927120669056153,
"grad_norm": 1.1499505043029785,
"learning_rate": 4.059026688509696e-06,
"loss": 1.9465,
"step": 2145
},
{
"epoch": 2.9284860897764124,
"grad_norm": 1.1801728010177612,
"learning_rate": 4.0569008977354756e-06,
"loss": 2.1298,
"step": 2146
},
{
"epoch": 2.929851510496672,
"grad_norm": 1.14891517162323,
"learning_rate": 4.054814030777336e-06,
"loss": 1.9772,
"step": 2147
},
{
"epoch": 2.931216931216931,
"grad_norm": 1.1447868347167969,
"learning_rate": 4.052766092162299e-06,
"loss": 2.0489,
"step": 2148
},
{
"epoch": 2.9325823519371905,
"grad_norm": 1.1374461650848389,
"learning_rate": 4.050757086332941e-06,
"loss": 1.9688,
"step": 2149
},
{
"epoch": 2.93394777265745,
"grad_norm": 1.1698285341262817,
"learning_rate": 4.048787017647381e-06,
"loss": 2.0696,
"step": 2150
},
{
"epoch": 2.9353131933777092,
"grad_norm": 1.1569222211837769,
"learning_rate": 4.046855890379276e-06,
"loss": 1.959,
"step": 2151
},
{
"epoch": 2.936678614097969,
"grad_norm": 1.1298071146011353,
"learning_rate": 4.0449637087178e-06,
"loss": 2.1045,
"step": 2152
},
{
"epoch": 2.9380440348182284,
"grad_norm": 1.2330018281936646,
"learning_rate": 4.043110476767649e-06,
"loss": 1.9134,
"step": 2153
},
{
"epoch": 2.9394094555384878,
"grad_norm": 1.1497013568878174,
"learning_rate": 4.041296198549024e-06,
"loss": 1.9994,
"step": 2154
},
{
"epoch": 2.940774876258747,
"grad_norm": 1.151652455329895,
"learning_rate": 4.0395208779976204e-06,
"loss": 1.9574,
"step": 2155
},
{
"epoch": 2.942140296979007,
"grad_norm": 1.139811396598816,
"learning_rate": 4.037784518964631e-06,
"loss": 2.054,
"step": 2156
},
{
"epoch": 2.9435057176992663,
"grad_norm": 1.1432186365127563,
"learning_rate": 4.0360871252167235e-06,
"loss": 1.9703,
"step": 2157
},
{
"epoch": 2.9448711384195256,
"grad_norm": 1.1401286125183105,
"learning_rate": 4.03442870043604e-06,
"loss": 2.0027,
"step": 2158
},
{
"epoch": 2.946236559139785,
"grad_norm": 1.173891305923462,
"learning_rate": 4.032809248220187e-06,
"loss": 1.9668,
"step": 2159
},
{
"epoch": 2.9476019798600444,
"grad_norm": 1.1533308029174805,
"learning_rate": 4.031228772082228e-06,
"loss": 1.953,
"step": 2160
},
{
"epoch": 2.9489674005803037,
"grad_norm": 1.1704301834106445,
"learning_rate": 4.0296872754506785e-06,
"loss": 1.9809,
"step": 2161
},
{
"epoch": 2.950332821300563,
"grad_norm": 1.1092525720596313,
"learning_rate": 4.028184761669494e-06,
"loss": 1.9477,
"step": 2162
},
{
"epoch": 2.9516982420208224,
"grad_norm": 1.1252529621124268,
"learning_rate": 4.026721233998066e-06,
"loss": 1.9313,
"step": 2163
},
{
"epoch": 2.9530636627410822,
"grad_norm": 1.1292991638183594,
"learning_rate": 4.02529669561121e-06,
"loss": 1.9885,
"step": 2164
},
{
"epoch": 2.9544290834613416,
"grad_norm": 1.147470235824585,
"learning_rate": 4.023911149599168e-06,
"loss": 1.9989,
"step": 2165
},
{
"epoch": 2.955794504181601,
"grad_norm": 1.142715573310852,
"learning_rate": 4.022564598967591e-06,
"loss": 1.9532,
"step": 2166
},
{
"epoch": 2.9571599249018603,
"grad_norm": 1.1440792083740234,
"learning_rate": 4.021257046637543e-06,
"loss": 2.0536,
"step": 2167
},
{
"epoch": 2.9585253456221197,
"grad_norm": 1.1772955656051636,
"learning_rate": 4.019988495445487e-06,
"loss": 2.0066,
"step": 2168
},
{
"epoch": 2.9598907663423795,
"grad_norm": 1.1505646705627441,
"learning_rate": 4.018758948143276e-06,
"loss": 2.0028,
"step": 2169
},
{
"epoch": 2.961256187062639,
"grad_norm": 1.1476067304611206,
"learning_rate": 4.017568407398161e-06,
"loss": 1.8543,
"step": 2170
},
{
"epoch": 2.962621607782898,
"grad_norm": 1.1357698440551758,
"learning_rate": 4.016416875792769e-06,
"loss": 2.0084,
"step": 2171
},
{
"epoch": 2.9639870285031575,
"grad_norm": 1.1889153718948364,
"learning_rate": 4.015304355825112e-06,
"loss": 1.9818,
"step": 2172
},
{
"epoch": 2.965352449223417,
"grad_norm": 1.1382731199264526,
"learning_rate": 4.014230849908567e-06,
"loss": 1.9222,
"step": 2173
},
{
"epoch": 2.9667178699436763,
"grad_norm": 1.1318186521530151,
"learning_rate": 4.0131963603718825e-06,
"loss": 1.9212,
"step": 2174
},
{
"epoch": 2.9680832906639356,
"grad_norm": 1.1510584354400635,
"learning_rate": 4.012200889459169e-06,
"loss": 2.1015,
"step": 2175
},
{
"epoch": 2.9694487113841954,
"grad_norm": 1.2627224922180176,
"learning_rate": 4.01124443932989e-06,
"loss": 1.9832,
"step": 2176
},
{
"epoch": 2.9708141321044548,
"grad_norm": 1.1568018198013306,
"learning_rate": 4.01032701205887e-06,
"loss": 1.9921,
"step": 2177
},
{
"epoch": 2.972179552824714,
"grad_norm": 1.1396300792694092,
"learning_rate": 4.0094486096362705e-06,
"loss": 1.97,
"step": 2178
},
{
"epoch": 2.9735449735449735,
"grad_norm": 1.1441128253936768,
"learning_rate": 4.008609233967607e-06,
"loss": 2.0013,
"step": 2179
},
{
"epoch": 2.974910394265233,
"grad_norm": 1.1298733949661255,
"learning_rate": 4.007808886873731e-06,
"loss": 1.9432,
"step": 2180
},
{
"epoch": 2.9762758149854927,
"grad_norm": 1.1511354446411133,
"learning_rate": 4.007047570090823e-06,
"loss": 1.9903,
"step": 2181
},
{
"epoch": 2.977641235705752,
"grad_norm": 1.1404117345809937,
"learning_rate": 4.006325285270407e-06,
"loss": 1.9335,
"step": 2182
},
{
"epoch": 2.9790066564260114,
"grad_norm": 1.1191532611846924,
"learning_rate": 4.0056420339793265e-06,
"loss": 1.9786,
"step": 2183
},
{
"epoch": 2.9803720771462707,
"grad_norm": 1.2450544834136963,
"learning_rate": 4.004997817699755e-06,
"loss": 2.0282,
"step": 2184
},
{
"epoch": 2.98173749786653,
"grad_norm": 1.1714141368865967,
"learning_rate": 4.004392637829182e-06,
"loss": 1.9615,
"step": 2185
},
{
"epoch": 2.9831029185867894,
"grad_norm": 1.1336561441421509,
"learning_rate": 4.003826495680423e-06,
"loss": 1.9554,
"step": 2186
},
{
"epoch": 2.984468339307049,
"grad_norm": 1.1253331899642944,
"learning_rate": 4.003299392481602e-06,
"loss": 2.0951,
"step": 2187
},
{
"epoch": 2.985833760027308,
"grad_norm": 1.1735095977783203,
"learning_rate": 4.002811329376164e-06,
"loss": 1.9488,
"step": 2188
},
{
"epoch": 2.987199180747568,
"grad_norm": 1.1404340267181396,
"learning_rate": 4.002362307422857e-06,
"loss": 1.9643,
"step": 2189
},
{
"epoch": 2.9885646014678273,
"grad_norm": 1.1273071765899658,
"learning_rate": 4.00195232759574e-06,
"loss": 1.895,
"step": 2190
},
{
"epoch": 2.9899300221880867,
"grad_norm": 1.1335234642028809,
"learning_rate": 4.00158139078418e-06,
"loss": 2.0343,
"step": 2191
},
{
"epoch": 2.991295442908346,
"grad_norm": 1.1854145526885986,
"learning_rate": 4.001249497792849e-06,
"loss": 2.0295,
"step": 2192
},
{
"epoch": 2.992660863628606,
"grad_norm": 1.1808243989944458,
"learning_rate": 4.000956649341716e-06,
"loss": 1.9341,
"step": 2193
},
{
"epoch": 2.994026284348865,
"grad_norm": 1.11687171459198,
"learning_rate": 4.000702846066058e-06,
"loss": 1.9731,
"step": 2194
},
{
"epoch": 2.9953917050691246,
"grad_norm": 1.1572637557983398,
"learning_rate": 4.000488088516446e-06,
"loss": 1.9625,
"step": 2195
},
{
"epoch": 2.996757125789384,
"grad_norm": 1.1488933563232422,
"learning_rate": 4.000312377158754e-06,
"loss": 1.9827,
"step": 2196
},
{
"epoch": 2.9981225465096433,
"grad_norm": 1.1624726057052612,
"learning_rate": 4.000175712374149e-06,
"loss": 1.9656,
"step": 2197
},
{
"epoch": 2.9994879672299026,
"grad_norm": 1.1880913972854614,
"learning_rate": 4.000078094459098e-06,
"loss": 1.9438,
"step": 2198
},
{
"epoch": 3.0,
"grad_norm": 1.1598238945007324,
"learning_rate": 4.000019523625363e-06,
"loss": 0.7646,
"step": 2199
},
{
"epoch": 3.0,
"step": 2199,
"total_flos": 2.2641188459144806e+18,
"train_loss": 2.5134489446024615,
"train_runtime": 29377.3396,
"train_samples_per_second": 9.572,
"train_steps_per_second": 0.075
}
],
"logging_steps": 1,
"max_steps": 2199,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.2641188459144806e+18,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}