{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 9324,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 4.222036361694336,
      "learning_rate": 2.9996782496782497e-05,
      "loss": 6.0377,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 3.7563281059265137,
      "learning_rate": 2.9993564993564994e-05,
      "loss": 5.8547,
      "step": 2
    },
    {
      "epoch": 0.0,
      "grad_norm": 4.197787761688232,
      "learning_rate": 2.999034749034749e-05,
      "loss": 5.6588,
      "step": 3
    },
    {
      "epoch": 0.0,
      "grad_norm": 4.9091081619262695,
      "learning_rate": 2.998712998712999e-05,
      "loss": 5.5094,
      "step": 4
    },
    {
      "epoch": 0.0,
      "grad_norm": 4.44374942779541,
      "learning_rate": 2.9983912483912486e-05,
      "loss": 5.3325,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.424834251403809,
      "learning_rate": 2.9980694980694983e-05,
      "loss": 5.1224,
      "step": 6
    },
    {
      "epoch": 0.0,
      "grad_norm": 4.825772762298584,
      "learning_rate": 2.997747747747748e-05,
      "loss": 4.9516,
      "step": 7
    },
    {
      "epoch": 0.0,
      "grad_norm": 6.119178771972656,
      "learning_rate": 2.9974259974259976e-05,
      "loss": 4.6992,
      "step": 8
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.859974384307861,
      "learning_rate": 2.997104247104247e-05,
      "loss": 4.5098,
      "step": 9
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.673393249511719,
      "learning_rate": 2.996782496782497e-05,
      "loss": 4.3641,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 5.682456970214844,
      "learning_rate": 2.9964607464607465e-05,
      "loss": 4.0987,
      "step": 11
    },
    {
      "epoch": 0.0,
      "grad_norm": 6.79993200302124,
      "learning_rate": 2.996138996138996e-05,
      "loss": 4.0507,
      "step": 12
    },
    {
      "epoch": 0.0,
      "grad_norm": 6.858672618865967,
      "learning_rate": 2.9958172458172458e-05,
      "loss": 4.0148,
      "step": 13
    },
    {
      "epoch": 0.0,
      "grad_norm": 8.730584144592285,
      "learning_rate": 2.9954954954954954e-05,
      "loss": 4.206,
      "step": 14
    },
    {
      "epoch": 0.0,
      "grad_norm": 6.24912691116333,
      "learning_rate": 2.995173745173745e-05,
      "loss": 4.0476,
      "step": 15
    },
    {
      "epoch": 0.0,
      "grad_norm": 7.954470157623291,
      "learning_rate": 2.9948519948519947e-05,
      "loss": 3.7035,
      "step": 16
    },
    {
      "epoch": 0.0,
      "grad_norm": 9.172163009643555,
      "learning_rate": 2.9945302445302447e-05,
      "loss": 3.6857,
      "step": 17
    },
    {
      "epoch": 0.0,
      "grad_norm": 9.562233924865723,
      "learning_rate": 2.9942084942084944e-05,
      "loss": 3.4166,
      "step": 18
    },
    {
      "epoch": 0.0,
      "grad_norm": 7.576280117034912,
      "learning_rate": 2.993886743886744e-05,
      "loss": 3.4729,
      "step": 19
    },
    {
      "epoch": 0.0,
      "grad_norm": 13.531746864318848,
      "learning_rate": 2.9935649935649937e-05,
      "loss": 3.2574,
      "step": 20
    },
    {
      "epoch": 0.0,
      "grad_norm": 15.042695999145508,
      "learning_rate": 2.9932432432432433e-05,
      "loss": 2.9611,
      "step": 21
    },
    {
      "epoch": 0.0,
      "grad_norm": 9.060174942016602,
      "learning_rate": 2.992921492921493e-05,
      "loss": 2.7617,
      "step": 22
    },
    {
      "epoch": 0.0,
      "grad_norm": 9.861416816711426,
      "learning_rate": 2.992599742599743e-05,
      "loss": 2.8034,
      "step": 23
    },
    {
      "epoch": 0.01,
      "grad_norm": 24.35716438293457,
      "learning_rate": 2.9922779922779926e-05,
      "loss": 2.7918,
      "step": 24
    },
    {
      "epoch": 0.01,
      "grad_norm": 20.110836029052734,
      "learning_rate": 2.9919562419562422e-05,
      "loss": 2.6288,
      "step": 25
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.263017654418945,
      "learning_rate": 2.991634491634492e-05,
      "loss": 2.776,
      "step": 26
    },
    {
      "epoch": 0.01,
      "grad_norm": 17.68182373046875,
      "learning_rate": 2.9913127413127412e-05,
      "loss": 3.0038,
      "step": 27
    },
    {
      "epoch": 0.01,
      "grad_norm": 14.46109676361084,
      "learning_rate": 2.9909909909909908e-05,
      "loss": 2.703,
      "step": 28
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.268301010131836,
      "learning_rate": 2.9906692406692408e-05,
      "loss": 2.5529,
      "step": 29
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.463258743286133,
      "learning_rate": 2.9903474903474904e-05,
      "loss": 2.7534,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 13.538990020751953,
      "learning_rate": 2.99002574002574e-05,
      "loss": 2.7378,
      "step": 31
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.298742294311523,
      "learning_rate": 2.9897039897039897e-05,
      "loss": 2.8942,
      "step": 32
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.29842758178711,
      "learning_rate": 2.9893822393822394e-05,
      "loss": 2.5627,
      "step": 33
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.457476615905762,
      "learning_rate": 2.989060489060489e-05,
      "loss": 2.2993,
      "step": 34
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.243216514587402,
      "learning_rate": 2.9887387387387387e-05,
      "loss": 2.5958,
      "step": 35
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.63615894317627,
      "learning_rate": 2.9884169884169887e-05,
      "loss": 2.5011,
      "step": 36
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.972916603088379,
      "learning_rate": 2.9880952380952383e-05,
      "loss": 2.3741,
      "step": 37
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.151965141296387,
      "learning_rate": 2.987773487773488e-05,
      "loss": 2.5984,
      "step": 38
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.013257026672363,
      "learning_rate": 2.9874517374517376e-05,
      "loss": 2.7259,
      "step": 39
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.656115531921387,
      "learning_rate": 2.9871299871299872e-05,
      "loss": 2.4663,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.090448379516602,
      "learning_rate": 2.986808236808237e-05,
      "loss": 2.6068,
      "step": 41
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.571541786193848,
      "learning_rate": 2.986486486486487e-05,
      "loss": 2.7506,
      "step": 42
    },
    {
      "epoch": 0.01,
      "grad_norm": 13.002962112426758,
      "learning_rate": 2.9861647361647365e-05,
      "loss": 2.2435,
      "step": 43
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.81763744354248,
      "learning_rate": 2.9858429858429858e-05,
      "loss": 2.5645,
      "step": 44
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.927344799041748,
      "learning_rate": 2.9855212355212355e-05,
      "loss": 2.3772,
      "step": 45
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.249074935913086,
      "learning_rate": 2.985199485199485e-05,
      "loss": 2.2743,
      "step": 46
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.538152694702148,
      "learning_rate": 2.9848777348777348e-05,
      "loss": 2.0459,
      "step": 47
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.560193061828613,
      "learning_rate": 2.9845559845559844e-05,
      "loss": 2.2317,
      "step": 48
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.0293607711792,
      "learning_rate": 2.9842342342342344e-05,
      "loss": 2.4055,
      "step": 49
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.661375999450684,
      "learning_rate": 2.983912483912484e-05,
      "loss": 2.2125,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.431386947631836,
      "learning_rate": 2.9835907335907337e-05,
      "loss": 2.6824,
      "step": 51
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.339584350585938,
      "learning_rate": 2.9832689832689833e-05,
      "loss": 2.6226,
      "step": 52
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.318562507629395,
      "learning_rate": 2.982947232947233e-05,
      "loss": 2.0877,
      "step": 53
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.481164932250977,
      "learning_rate": 2.9826254826254826e-05,
      "loss": 2.1334,
      "step": 54
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.117362976074219,
      "learning_rate": 2.9823037323037326e-05,
      "loss": 2.1992,
      "step": 55
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.039522171020508,
      "learning_rate": 2.9819819819819822e-05,
      "loss": 2.1365,
      "step": 56
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.950950622558594,
      "learning_rate": 2.981660231660232e-05,
      "loss": 2.1184,
      "step": 57
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.990957260131836,
      "learning_rate": 2.9813384813384815e-05,
      "loss": 2.2168,
      "step": 58
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.645991325378418,
      "learning_rate": 2.9810167310167312e-05,
      "loss": 2.2137,
      "step": 59
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.12086296081543,
      "learning_rate": 2.9806949806949808e-05,
      "loss": 1.914,
      "step": 60
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.203763008117676,
      "learning_rate": 2.98037323037323e-05,
      "loss": 2.1189,
      "step": 61
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.492456436157227,
      "learning_rate": 2.98005148005148e-05,
      "loss": 2.2044,
      "step": 62
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.89135456085205,
      "learning_rate": 2.9797297297297298e-05,
      "loss": 2.8807,
      "step": 63
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.258874893188477,
      "learning_rate": 2.9794079794079794e-05,
      "loss": 1.9186,
      "step": 64
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.736390113830566,
      "learning_rate": 2.979086229086229e-05,
      "loss": 1.967,
      "step": 65
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.980557441711426,
      "learning_rate": 2.9787644787644787e-05,
      "loss": 1.8922,
      "step": 66
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.506407737731934,
      "learning_rate": 2.9784427284427283e-05,
      "loss": 2.2008,
      "step": 67
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.90528392791748,
      "learning_rate": 2.9781209781209783e-05,
      "loss": 2.0387,
      "step": 68
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.060418128967285,
      "learning_rate": 2.977799227799228e-05,
      "loss": 1.8205,
      "step": 69
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.243399620056152,
      "learning_rate": 2.9774774774774776e-05,
      "loss": 1.7186,
      "step": 70
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.825020790100098,
      "learning_rate": 2.9771557271557273e-05,
      "loss": 2.1778,
      "step": 71
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.006978034973145,
      "learning_rate": 2.976833976833977e-05,
      "loss": 2.2028,
      "step": 72
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.421529769897461,
      "learning_rate": 2.9765122265122265e-05,
      "loss": 1.9151,
      "step": 73
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.49144172668457,
      "learning_rate": 2.9761904761904762e-05,
      "loss": 2.1674,
      "step": 74
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.524439811706543,
      "learning_rate": 2.9758687258687262e-05,
      "loss": 1.7358,
      "step": 75
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.900327682495117,
      "learning_rate": 2.9755469755469758e-05,
      "loss": 1.9131,
      "step": 76
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.561383247375488,
      "learning_rate": 2.9752252252252255e-05,
      "loss": 2.0451,
      "step": 77
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.260885238647461,
      "learning_rate": 2.974903474903475e-05,
      "loss": 2.2324,
      "step": 78
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.745309829711914,
      "learning_rate": 2.9745817245817244e-05,
      "loss": 1.7487,
      "step": 79
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.257708549499512,
      "learning_rate": 2.974259974259974e-05,
      "loss": 2.1048,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.863325119018555,
      "learning_rate": 2.973938223938224e-05,
      "loss": 2.2193,
      "step": 81
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.887799263000488,
      "learning_rate": 2.9736164736164737e-05,
      "loss": 1.8845,
      "step": 82
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.843587875366211,
      "learning_rate": 2.9732947232947233e-05,
      "loss": 2.0644,
      "step": 83
    },
    {
      "epoch": 0.02,
      "grad_norm": 11.363655090332031,
      "learning_rate": 2.972972972972973e-05,
      "loss": 2.0773,
      "step": 84
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.520673751831055,
      "learning_rate": 2.9726512226512226e-05,
      "loss": 1.8821,
      "step": 85
    },
    {
      "epoch": 0.02,
      "grad_norm": 11.55678939819336,
      "learning_rate": 2.9723294723294723e-05,
      "loss": 2.421,
      "step": 86
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.049636840820312,
      "learning_rate": 2.9720077220077223e-05,
      "loss": 2.2814,
      "step": 87
    },
    {
      "epoch": 0.02,
      "grad_norm": 14.909097671508789,
      "learning_rate": 2.971685971685972e-05,
      "loss": 1.9071,
      "step": 88
    },
    {
      "epoch": 0.02,
      "grad_norm": 11.912261962890625,
      "learning_rate": 2.9713642213642215e-05,
      "loss": 1.9148,
      "step": 89
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.710816383361816,
      "learning_rate": 2.9710424710424712e-05,
      "loss": 1.8639,
      "step": 90
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.137489318847656,
      "learning_rate": 2.970720720720721e-05,
      "loss": 2.0118,
      "step": 91
    },
    {
      "epoch": 0.02,
      "grad_norm": 12.987302780151367,
      "learning_rate": 2.9703989703989705e-05,
      "loss": 1.8338,
      "step": 92
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.273015022277832,
      "learning_rate": 2.97007722007722e-05,
      "loss": 1.7637,
      "step": 93
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.091267585754395,
      "learning_rate": 2.96975546975547e-05,
      "loss": 2.0367,
      "step": 94
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.653928756713867,
      "learning_rate": 2.9694337194337198e-05,
      "loss": 1.8743,
      "step": 95
    },
    {
      "epoch": 0.02,
      "grad_norm": 11.013269424438477,
      "learning_rate": 2.969111969111969e-05,
      "loss": 1.9441,
      "step": 96
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.5699462890625,
      "learning_rate": 2.9687902187902187e-05,
      "loss": 2.0935,
      "step": 97
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.903471946716309,
      "learning_rate": 2.9684684684684683e-05,
      "loss": 2.0907,
      "step": 98
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.190542221069336,
      "learning_rate": 2.968146718146718e-05,
      "loss": 1.8724,
      "step": 99
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.406411170959473,
      "learning_rate": 2.967824967824968e-05,
      "loss": 2.0295,
      "step": 100
    },
    {
      "epoch": 0.02,
      "grad_norm": 15.533895492553711,
      "learning_rate": 2.9675032175032176e-05,
      "loss": 2.1564,
      "step": 101
    },
    {
      "epoch": 0.02,
      "grad_norm": 12.441184997558594,
      "learning_rate": 2.9671814671814673e-05,
      "loss": 2.1813,
      "step": 102
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.165247917175293,
      "learning_rate": 2.966859716859717e-05,
      "loss": 2.0714,
      "step": 103
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.608696937561035,
      "learning_rate": 2.9665379665379666e-05,
      "loss": 1.7906,
      "step": 104
    },
    {
      "epoch": 0.02,
      "grad_norm": 11.659300804138184,
      "learning_rate": 2.9662162162162162e-05,
      "loss": 1.8263,
      "step": 105
    },
    {
      "epoch": 0.02,
      "grad_norm": 13.284748077392578,
      "learning_rate": 2.965894465894466e-05,
      "loss": 1.8195,
      "step": 106
    },
    {
      "epoch": 0.02,
      "grad_norm": 10.918658256530762,
      "learning_rate": 2.965572715572716e-05,
      "loss": 2.1485,
      "step": 107
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.950419425964355,
      "learning_rate": 2.9652509652509655e-05,
      "loss": 2.2802,
      "step": 108
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.8795747756958,
      "learning_rate": 2.964929214929215e-05,
      "loss": 1.4782,
      "step": 109
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.622880935668945,
      "learning_rate": 2.9646074646074648e-05,
      "loss": 2.0353,
      "step": 110
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.232423782348633,
      "learning_rate": 2.9642857142857144e-05,
      "loss": 1.7708,
      "step": 111
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.372254371643066,
      "learning_rate": 2.963963963963964e-05,
      "loss": 2.0746,
      "step": 112
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.188185691833496,
      "learning_rate": 2.963642213642214e-05,
      "loss": 1.6071,
      "step": 113
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.805091857910156,
      "learning_rate": 2.9633204633204633e-05,
      "loss": 1.8581,
      "step": 114
    },
    {
      "epoch": 0.02,
      "grad_norm": 15.246764183044434,
      "learning_rate": 2.962998712998713e-05,
      "loss": 1.7937,
      "step": 115
    },
    {
      "epoch": 0.02,
      "grad_norm": 14.798468589782715,
      "learning_rate": 2.9626769626769626e-05,
      "loss": 2.21,
      "step": 116
    },
    {
      "epoch": 0.03,
      "grad_norm": 12.647953033447266,
      "learning_rate": 2.9623552123552123e-05,
      "loss": 2.1393,
      "step": 117
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.389384269714355,
      "learning_rate": 2.962033462033462e-05,
      "loss": 2.049,
      "step": 118
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.136863708496094,
      "learning_rate": 2.9617117117117116e-05,
      "loss": 1.9286,
      "step": 119
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.246565818786621,
      "learning_rate": 2.9613899613899616e-05,
      "loss": 1.7996,
      "step": 120
    },
    {
      "epoch": 0.03,
      "grad_norm": 12.276095390319824,
      "learning_rate": 2.9610682110682112e-05,
      "loss": 1.9909,
      "step": 121
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.009767532348633,
      "learning_rate": 2.960746460746461e-05,
      "loss": 2.0852,
      "step": 122
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.873170852661133,
      "learning_rate": 2.9604247104247105e-05,
      "loss": 1.6296,
      "step": 123
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.941073417663574,
      "learning_rate": 2.96010296010296e-05,
      "loss": 1.6071,
      "step": 124
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.338553428649902,
      "learning_rate": 2.9597812097812098e-05,
      "loss": 1.5292,
      "step": 125
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.6126070022583,
      "learning_rate": 2.9594594594594598e-05,
      "loss": 1.9089,
      "step": 126
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.378311157226562,
      "learning_rate": 2.9591377091377094e-05,
      "loss": 1.9958,
      "step": 127
    },
    {
      "epoch": 0.03,
      "grad_norm": 11.68854808807373,
      "learning_rate": 2.958815958815959e-05,
      "loss": 1.8321,
      "step": 128
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.570721626281738,
      "learning_rate": 2.9584942084942087e-05,
      "loss": 1.5377,
      "step": 129
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.104324340820312,
      "learning_rate": 2.9581724581724583e-05,
      "loss": 1.7748,
      "step": 130
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.3139009475708,
      "learning_rate": 2.9578507078507077e-05,
      "loss": 1.8756,
      "step": 131
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.6179780960083,
      "learning_rate": 2.9575289575289573e-05,
      "loss": 2.0878,
      "step": 132
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.899459838867188,
      "learning_rate": 2.9572072072072073e-05,
      "loss": 2.3886,
      "step": 133
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.505660057067871,
      "learning_rate": 2.956885456885457e-05,
      "loss": 1.7308,
      "step": 134
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.575037002563477,
      "learning_rate": 2.9565637065637066e-05,
      "loss": 1.5489,
      "step": 135
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.529553413391113,
      "learning_rate": 2.9562419562419562e-05,
      "loss": 1.9336,
      "step": 136
    },
    {
      "epoch": 0.03,
      "grad_norm": 7.9347734451293945,
      "learning_rate": 2.955920205920206e-05,
      "loss": 1.7661,
      "step": 137
    },
    {
      "epoch": 0.03,
      "grad_norm": 7.6012349128723145,
      "learning_rate": 2.9555984555984555e-05,
      "loss": 1.6812,
      "step": 138
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.094590187072754,
      "learning_rate": 2.9552767052767055e-05,
      "loss": 1.6584,
      "step": 139
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.415298461914062,
      "learning_rate": 2.954954954954955e-05,
      "loss": 2.1022,
      "step": 140
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.531627655029297,
      "learning_rate": 2.9546332046332048e-05,
      "loss": 2.0698,
      "step": 141
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.801520347595215,
      "learning_rate": 2.9543114543114544e-05,
      "loss": 1.9003,
      "step": 142
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.831201553344727,
      "learning_rate": 2.953989703989704e-05,
      "loss": 1.5995,
      "step": 143
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.96854019165039,
      "learning_rate": 2.9536679536679537e-05,
      "loss": 2.2821,
      "step": 144
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.958708763122559,
      "learning_rate": 2.9533462033462034e-05,
      "loss": 1.7062,
      "step": 145
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.62470817565918,
      "learning_rate": 2.9530244530244533e-05,
      "loss": 2.0017,
      "step": 146
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.932842254638672,
      "learning_rate": 2.952702702702703e-05,
      "loss": 1.9617,
      "step": 147
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.648544311523438,
      "learning_rate": 2.9523809523809523e-05,
      "loss": 2.006,
      "step": 148
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.422574043273926,
      "learning_rate": 2.952059202059202e-05,
      "loss": 1.7057,
      "step": 149
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.800268173217773,
      "learning_rate": 2.9517374517374516e-05,
      "loss": 1.8033,
      "step": 150
    },
    {
      "epoch": 0.03,
      "grad_norm": 7.836630344390869,
      "learning_rate": 2.9514157014157012e-05,
      "loss": 1.7456,
      "step": 151
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.225390434265137,
      "learning_rate": 2.9510939510939512e-05,
      "loss": 1.9177,
      "step": 152
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.10119915008545,
      "learning_rate": 2.950772200772201e-05,
      "loss": 1.7646,
      "step": 153
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.467876434326172,
      "learning_rate": 2.9504504504504505e-05,
      "loss": 1.5888,
      "step": 154
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.005247116088867,
      "learning_rate": 2.9501287001287e-05,
      "loss": 1.8332,
      "step": 155
    },
    {
      "epoch": 0.03,
      "grad_norm": 9.461679458618164,
      "learning_rate": 2.9498069498069498e-05,
      "loss": 2.0156,
      "step": 156
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.076912879943848,
      "learning_rate": 2.9494851994851994e-05,
      "loss": 1.6595,
      "step": 157
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.08866024017334,
      "learning_rate": 2.9491634491634494e-05,
      "loss": 1.4514,
      "step": 158
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.119258880615234,
      "learning_rate": 2.948841698841699e-05,
      "loss": 2.0006,
      "step": 159
    },
    {
      "epoch": 0.03,
      "grad_norm": 12.805803298950195,
      "learning_rate": 2.9485199485199487e-05,
      "loss": 1.5227,
      "step": 160
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.602291107177734,
      "learning_rate": 2.9481981981981984e-05,
      "loss": 1.3537,
      "step": 161
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.74146842956543,
      "learning_rate": 2.947876447876448e-05,
      "loss": 1.9342,
      "step": 162
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.16552448272705,
      "learning_rate": 2.9475546975546977e-05,
      "loss": 1.7654,
      "step": 163
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.54535961151123,
      "learning_rate": 2.9472329472329473e-05,
      "loss": 1.7898,
      "step": 164
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.517549514770508,
      "learning_rate": 2.9469111969111973e-05,
      "loss": 1.8434,
      "step": 165
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.566190719604492,
      "learning_rate": 2.9465894465894466e-05,
      "loss": 2.0275,
      "step": 166
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.858747482299805,
      "learning_rate": 2.9462676962676962e-05,
      "loss": 1.5693,
      "step": 167
    },
    {
      "epoch": 0.04,
      "grad_norm": 7.875096321105957,
      "learning_rate": 2.945945945945946e-05,
      "loss": 1.3802,
      "step": 168
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.36303997039795,
      "learning_rate": 2.9456241956241955e-05,
      "loss": 1.3452,
      "step": 169
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.93249797821045,
      "learning_rate": 2.9453024453024452e-05,
      "loss": 1.9246,
      "step": 170
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.464676856994629,
      "learning_rate": 2.944980694980695e-05,
      "loss": 1.5675,
      "step": 171
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.116580963134766,
      "learning_rate": 2.9446589446589448e-05,
      "loss": 1.4455,
      "step": 172
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.615945816040039,
      "learning_rate": 2.9443371943371944e-05,
      "loss": 1.7214,
      "step": 173
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.384892463684082,
      "learning_rate": 2.944015444015444e-05,
      "loss": 1.6379,
      "step": 174
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.380600929260254,
      "learning_rate": 2.9436936936936937e-05,
      "loss": 1.8107,
      "step": 175
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.78029727935791,
      "learning_rate": 2.9433719433719434e-05,
      "loss": 1.5841,
      "step": 176
    },
    {
      "epoch": 0.04,
      "grad_norm": 11.236618995666504,
      "learning_rate": 2.943050193050193e-05,
      "loss": 1.8165,
      "step": 177
    },
    {
      "epoch": 0.04,
      "grad_norm": 11.593771934509277,
      "learning_rate": 2.942728442728443e-05,
      "loss": 2.097,
      "step": 178
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.088334083557129,
      "learning_rate": 2.9424066924066927e-05,
      "loss": 1.8917,
      "step": 179
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.924134254455566,
      "learning_rate": 2.9420849420849423e-05,
      "loss": 1.7313,
      "step": 180
    },
    {
      "epoch": 0.04,
      "grad_norm": 15.392868995666504,
      "learning_rate": 2.941763191763192e-05,
      "loss": 1.9306,
      "step": 181
    },
    {
      "epoch": 0.04,
      "grad_norm": 13.010275840759277,
      "learning_rate": 2.9414414414414416e-05,
      "loss": 1.8158,
      "step": 182
    },
    {
      "epoch": 0.04,
      "grad_norm": 12.502853393554688,
      "learning_rate": 2.941119691119691e-05,
      "loss": 1.8865,
      "step": 183
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.34823226928711,
      "learning_rate": 2.940797940797941e-05,
      "loss": 1.6293,
      "step": 184
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.818218231201172,
      "learning_rate": 2.9404761904761905e-05,
      "loss": 1.6928,
      "step": 185
    },
    {
      "epoch": 0.04,
      "grad_norm": 12.046592712402344,
      "learning_rate": 2.9401544401544402e-05,
      "loss": 1.7269,
      "step": 186
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.006397247314453,
      "learning_rate": 2.9398326898326898e-05,
      "loss": 1.4838,
      "step": 187
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.717196464538574,
      "learning_rate": 2.9395109395109395e-05,
      "loss": 2.0447,
      "step": 188
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.830742835998535,
      "learning_rate": 2.939189189189189e-05,
      "loss": 1.5191,
      "step": 189
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.42054271697998,
      "learning_rate": 2.9388674388674388e-05,
      "loss": 1.9991,
      "step": 190
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.7524995803833,
      "learning_rate": 2.9385456885456887e-05,
      "loss": 1.7813,
      "step": 191
    },
    {
      "epoch": 0.04,
      "grad_norm": 13.04479694366455,
      "learning_rate": 2.9382239382239384e-05,
      "loss": 1.8759,
      "step": 192
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.093859672546387,
      "learning_rate": 2.937902187902188e-05,
      "loss": 1.5478,
      "step": 193
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.382877349853516,
      "learning_rate": 2.9375804375804377e-05,
      "loss": 1.3072,
      "step": 194
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.99227523803711,
      "learning_rate": 2.9372586872586873e-05,
      "loss": 1.8962,
      "step": 195
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.74660587310791,
      "learning_rate": 2.936936936936937e-05,
      "loss": 1.4915,
      "step": 196
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.000102043151855,
      "learning_rate": 2.936615186615187e-05,
      "loss": 1.6472,
      "step": 197
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.52940845489502,
      "learning_rate": 2.9362934362934366e-05,
      "loss": 1.7031,
      "step": 198
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.47712516784668,
      "learning_rate": 2.9359716859716862e-05,
      "loss": 1.7298,
      "step": 199
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.509488105773926,
      "learning_rate": 2.935649935649936e-05,
      "loss": 1.9848,
      "step": 200
    },
    {
      "epoch": 0.04,
      "grad_norm": 14.348960876464844,
      "learning_rate": 2.9353281853281852e-05,
      "loss": 1.7919,
      "step": 201
    },
    {
      "epoch": 0.04,
      "grad_norm": 13.042765617370605,
      "learning_rate": 2.935006435006435e-05,
      "loss": 1.6102,
      "step": 202
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.776371955871582,
      "learning_rate": 2.9346846846846848e-05,
      "loss": 1.7545,
      "step": 203
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.236967086791992,
      "learning_rate": 2.9343629343629345e-05,
      "loss": 1.5609,
      "step": 204
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.920976638793945,
      "learning_rate": 2.934041184041184e-05,
      "loss": 2.1136,
      "step": 205
    },
    {
      "epoch": 0.04,
      "grad_norm": 14.1851224899292,
      "learning_rate": 2.9337194337194338e-05,
      "loss": 2.017,
      "step": 206
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.17455005645752,
      "learning_rate": 2.9333976833976834e-05,
      "loss": 1.8087,
      "step": 207
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.93751335144043,
      "learning_rate": 2.933075933075933e-05,
      "loss": 1.8248,
      "step": 208
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.896265983581543,
      "learning_rate": 2.9327541827541827e-05,
      "loss": 1.7536,
      "step": 209
    },
    {
      "epoch": 0.05,
      "grad_norm": 12.59648323059082,
      "learning_rate": 2.9324324324324327e-05,
      "loss": 1.985,
      "step": 210
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.070180892944336,
      "learning_rate": 2.9321106821106823e-05,
      "loss": 1.8123,
      "step": 211
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.1323881149292,
      "learning_rate": 2.931788931788932e-05,
      "loss": 1.284,
      "step": 212
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.741929054260254,
      "learning_rate": 2.9314671814671816e-05,
      "loss": 1.3793,
      "step": 213
    },
    {
      "epoch": 0.05,
      "grad_norm": 11.452217102050781,
      "learning_rate": 2.9311454311454313e-05,
      "loss": 1.5329,
      "step": 214
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.352455139160156,
      "learning_rate": 2.930823680823681e-05,
      "loss": 1.5693,
      "step": 215
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.591645240783691,
      "learning_rate": 2.930501930501931e-05,
      "loss": 1.7363,
      "step": 216
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.718780517578125,
      "learning_rate": 2.9301801801801805e-05,
      "loss": 1.9908,
      "step": 217
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.434111595153809,
      "learning_rate": 2.92985842985843e-05,
      "loss": 1.9244,
      "step": 218
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.281728744506836,
      "learning_rate": 2.9295366795366795e-05,
      "loss": 1.726,
      "step": 219
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.511754989624023,
      "learning_rate": 2.929214929214929e-05,
      "loss": 1.4333,
      "step": 220
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.996186256408691,
      "learning_rate": 2.9288931788931788e-05,
      "loss": 1.8024,
      "step": 221
    },
    {
      "epoch": 0.05,
      "grad_norm": 15.18430233001709,
      "learning_rate": 2.9285714285714284e-05,
      "loss": 1.6886,
      "step": 222
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.822641372680664,
      "learning_rate": 2.9282496782496784e-05,
      "loss": 1.8237,
      "step": 223
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.187376022338867,
      "learning_rate": 2.927927927927928e-05,
      "loss": 1.6152,
      "step": 224
    },
    {
      "epoch": 0.05,
      "grad_norm": 12.130743980407715,
      "learning_rate": 2.9276061776061777e-05,
      "loss": 1.6072,
      "step": 225
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.02490520477295,
      "learning_rate": 2.9272844272844273e-05,
      "loss": 1.6666,
      "step": 226
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.933765411376953,
      "learning_rate": 2.926962676962677e-05,
      "loss": 1.655,
      "step": 227
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.114948272705078,
      "learning_rate": 2.9266409266409266e-05,
      "loss": 1.5711,
      "step": 228
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.082925796508789,
      "learning_rate": 2.9263191763191766e-05,
      "loss": 1.5686,
      "step": 229
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.492475509643555,
      "learning_rate": 2.9259974259974263e-05,
      "loss": 1.6116,
      "step": 230
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.24294376373291,
      "learning_rate": 2.925675675675676e-05,
      "loss": 1.6079,
      "step": 231
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.095669746398926,
      "learning_rate": 2.9253539253539255e-05,
      "loss": 1.7333,
      "step": 232
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.451577186584473,
      "learning_rate": 2.9250321750321752e-05,
      "loss": 1.6967,
      "step": 233
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.744237899780273,
      "learning_rate": 2.924710424710425e-05,
      "loss": 1.7555,
      "step": 234
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.634855270385742,
      "learning_rate": 2.924388674388674e-05,
      "loss": 1.9986,
      "step": 235
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.827519416809082,
      "learning_rate": 2.924066924066924e-05,
      "loss": 1.5114,
      "step": 236
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.777359008789062,
      "learning_rate": 2.9237451737451738e-05,
      "loss": 1.9315,
      "step": 237
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.895609855651855,
      "learning_rate": 2.9234234234234234e-05,
      "loss": 1.749,
      "step": 238
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.99881362915039,
      "learning_rate": 2.923101673101673e-05,
      "loss": 1.9862,
      "step": 239
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.502777099609375,
      "learning_rate": 2.9227799227799227e-05,
      "loss": 1.598,
      "step": 240
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.817286014556885,
      "learning_rate": 2.9224581724581723e-05,
      "loss": 1.2801,
      "step": 241
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.413171768188477,
      "learning_rate": 2.9221364221364223e-05,
      "loss": 1.5415,
      "step": 242
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.041288375854492,
      "learning_rate": 2.921814671814672e-05,
      "loss": 1.5111,
      "step": 243
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.475278854370117,
      "learning_rate": 2.9214929214929216e-05,
      "loss": 1.7544,
      "step": 244
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.40074634552002,
      "learning_rate": 2.9211711711711713e-05,
      "loss": 1.8547,
      "step": 245
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.08721923828125,
      "learning_rate": 2.920849420849421e-05,
      "loss": 1.557,
      "step": 246
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.493610382080078,
      "learning_rate": 2.9205276705276706e-05,
      "loss": 1.413,
      "step": 247
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.332128524780273,
      "learning_rate": 2.9202059202059202e-05,
      "loss": 1.7619,
      "step": 248
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.594747543334961,
      "learning_rate": 2.9198841698841702e-05,
      "loss": 1.6239,
      "step": 249
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.151785850524902,
      "learning_rate": 2.91956241956242e-05,
      "loss": 1.82,
      "step": 250
    },
    {
      "epoch": 0.05,
      "grad_norm": 11.367535591125488,
      "learning_rate": 2.9192406692406695e-05,
      "loss": 1.8197,
      "step": 251
    },
    {
      "epoch": 0.05,
      "grad_norm": 12.060310363769531,
      "learning_rate": 2.918918918918919e-05,
      "loss": 1.5421,
      "step": 252
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.836727142333984,
      "learning_rate": 2.9185971685971684e-05,
      "loss": 1.5994,
      "step": 253
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.861918449401855,
      "learning_rate": 2.918275418275418e-05,
      "loss": 1.6985,
      "step": 254
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.986840724945068,
      "learning_rate": 2.917953667953668e-05,
      "loss": 1.4215,
      "step": 255
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.180085182189941,
      "learning_rate": 2.9176319176319177e-05,
      "loss": 1.9733,
      "step": 256
    },
    {
      "epoch": 0.06,
      "grad_norm": 12.837593078613281,
      "learning_rate": 2.9173101673101673e-05,
      "loss": 2.0426,
      "step": 257
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.791147232055664,
      "learning_rate": 2.916988416988417e-05,
      "loss": 1.7578,
      "step": 258
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.055302619934082,
      "learning_rate": 2.9166666666666666e-05,
      "loss": 1.3842,
      "step": 259
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.913887023925781,
      "learning_rate": 2.9163449163449163e-05,
      "loss": 1.7601,
      "step": 260
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.453872680664062,
      "learning_rate": 2.9160231660231663e-05,
      "loss": 1.6508,
      "step": 261
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.452512741088867,
      "learning_rate": 2.915701415701416e-05,
      "loss": 1.4725,
      "step": 262
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.82564640045166,
      "learning_rate": 2.9153796653796656e-05,
      "loss": 1.6356,
      "step": 263
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.05886173248291,
      "learning_rate": 2.9150579150579152e-05,
      "loss": 1.7094,
      "step": 264
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.537912368774414,
      "learning_rate": 2.914736164736165e-05,
      "loss": 1.7483,
      "step": 265
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.663581848144531,
      "learning_rate": 2.9144144144144145e-05,
      "loss": 1.6812,
      "step": 266
    },
    {
      "epoch": 0.06,
      "grad_norm": 7.709865570068359,
      "learning_rate": 2.914092664092664e-05,
      "loss": 1.357,
      "step": 267
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.307912826538086,
      "learning_rate": 2.913770913770914e-05,
      "loss": 1.8419,
      "step": 268
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.190519332885742,
      "learning_rate": 2.9134491634491638e-05,
      "loss": 1.6695,
      "step": 269
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.801825523376465,
      "learning_rate": 2.913127413127413e-05,
      "loss": 1.4692,
      "step": 270
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.539369583129883,
      "learning_rate": 2.9128056628056627e-05,
      "loss": 1.5779,
      "step": 271
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.949646949768066,
      "learning_rate": 2.9124839124839124e-05,
      "loss": 1.797,
      "step": 272
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.701847076416016,
      "learning_rate": 2.912162162162162e-05,
      "loss": 1.9331,
      "step": 273
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.768564224243164,
      "learning_rate": 2.911840411840412e-05,
      "loss": 1.5281,
      "step": 274
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.565635681152344,
      "learning_rate": 2.9115186615186616e-05,
      "loss": 2.0134,
      "step": 275
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.379804611206055,
      "learning_rate": 2.9111969111969113e-05,
      "loss": 1.7802,
      "step": 276
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.431669235229492,
      "learning_rate": 2.910875160875161e-05,
      "loss": 1.6051,
      "step": 277
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.409300804138184,
      "learning_rate": 2.9105534105534106e-05,
      "loss": 1.8116,
      "step": 278
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.000040054321289,
      "learning_rate": 2.9102316602316602e-05,
      "loss": 1.4914,
      "step": 279
    },
    {
      "epoch": 0.06,
      "grad_norm": 7.463644504547119,
      "learning_rate": 2.90990990990991e-05,
      "loss": 1.5488,
      "step": 280
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.206610679626465,
      "learning_rate": 2.90958815958816e-05,
      "loss": 1.6361,
      "step": 281
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.636147499084473,
      "learning_rate": 2.9092664092664095e-05,
      "loss": 1.9389,
      "step": 282
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.077810287475586,
      "learning_rate": 2.908944658944659e-05,
      "loss": 1.5946,
      "step": 283
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.616569519042969,
      "learning_rate": 2.9086229086229088e-05,
      "loss": 1.9833,
      "step": 284
    },
    {
      "epoch": 0.06,
      "grad_norm": 11.779959678649902,
      "learning_rate": 2.9083011583011584e-05,
      "loss": 1.9427,
      "step": 285
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.272721290588379,
      "learning_rate": 2.907979407979408e-05,
      "loss": 2.0379,
      "step": 286
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.743531227111816,
      "learning_rate": 2.907657657657658e-05,
      "loss": 1.6613,
      "step": 287
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.889345169067383,
      "learning_rate": 2.9073359073359074e-05,
      "loss": 1.8972,
      "step": 288
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.330607414245605,
      "learning_rate": 2.907014157014157e-05,
      "loss": 1.658,
      "step": 289
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.92176342010498,
      "learning_rate": 2.9066924066924067e-05,
      "loss": 1.6632,
      "step": 290
    },
    {
      "epoch": 0.06,
      "grad_norm": 11.867676734924316,
      "learning_rate": 2.9063706563706563e-05,
      "loss": 1.9126,
      "step": 291
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.153913497924805,
      "learning_rate": 2.906048906048906e-05,
      "loss": 1.4336,
      "step": 292
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.372769355773926,
      "learning_rate": 2.9057271557271556e-05,
      "loss": 1.756,
      "step": 293
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.488447189331055,
      "learning_rate": 2.9054054054054056e-05,
      "loss": 1.5957,
      "step": 294
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.782614707946777,
      "learning_rate": 2.9050836550836552e-05,
      "loss": 1.5842,
      "step": 295
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.450554847717285,
      "learning_rate": 2.904761904761905e-05,
      "loss": 1.6659,
      "step": 296
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.804947853088379,
      "learning_rate": 2.9044401544401545e-05,
      "loss": 1.7145,
      "step": 297
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.777469635009766,
      "learning_rate": 2.904118404118404e-05,
      "loss": 1.9364,
      "step": 298
    },
    {
      "epoch": 0.06,
      "grad_norm": 11.7806396484375,
      "learning_rate": 2.9037966537966538e-05,
      "loss": 1.7137,
      "step": 299
    },
    {
      "epoch": 0.06,
      "grad_norm": 7.980794906616211,
      "learning_rate": 2.9034749034749038e-05,
      "loss": 1.2667,
      "step": 300
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.473836898803711,
      "learning_rate": 2.9031531531531534e-05,
      "loss": 1.8796,
      "step": 301
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.058328628540039,
      "learning_rate": 2.902831402831403e-05,
      "loss": 1.8801,
      "step": 302
    },
    {
      "epoch": 0.06,
      "grad_norm": 7.485058784484863,
      "learning_rate": 2.9025096525096527e-05,
      "loss": 1.6769,
      "step": 303
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.915830612182617,
      "learning_rate": 2.9021879021879024e-05,
      "loss": 1.6841,
      "step": 304
    },
    {
      "epoch": 0.07,
      "grad_norm": 10.057333946228027,
      "learning_rate": 2.9018661518661517e-05,
      "loss": 1.6869,
      "step": 305
    },
    {
      "epoch": 0.07,
      "grad_norm": 10.86095142364502,
      "learning_rate": 2.9015444015444013e-05,
      "loss": 2.0631,
      "step": 306
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.916669845581055,
      "learning_rate": 2.9012226512226513e-05,
      "loss": 1.5437,
      "step": 307
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.40907096862793,
      "learning_rate": 2.900900900900901e-05,
      "loss": 1.7072,
      "step": 308
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.254368782043457,
      "learning_rate": 2.9005791505791506e-05,
      "loss": 1.9851,
      "step": 309
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.874781608581543,
      "learning_rate": 2.9002574002574002e-05,
      "loss": 1.9301,
      "step": 310
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.679688453674316,
      "learning_rate": 2.89993564993565e-05,
      "loss": 1.5819,
      "step": 311
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.285836219787598,
      "learning_rate": 2.8996138996138995e-05,
      "loss": 1.8799,
      "step": 312
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.541646003723145,
      "learning_rate": 2.8992921492921495e-05,
      "loss": 1.7016,
      "step": 313
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.922928810119629,
      "learning_rate": 2.898970398970399e-05,
      "loss": 1.684,
      "step": 314
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.38800048828125,
      "learning_rate": 2.8986486486486488e-05,
      "loss": 1.8992,
      "step": 315
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.95584774017334,
      "learning_rate": 2.8983268983268984e-05,
      "loss": 1.5904,
      "step": 316
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.982702255249023,
      "learning_rate": 2.898005148005148e-05,
      "loss": 1.5969,
      "step": 317
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.498199462890625,
      "learning_rate": 2.8976833976833977e-05,
      "loss": 1.5061,
      "step": 318
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.171370506286621,
      "learning_rate": 2.8973616473616477e-05,
      "loss": 1.7546,
      "step": 319
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.849658966064453,
      "learning_rate": 2.8970398970398974e-05,
      "loss": 1.5156,
      "step": 320
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.668301582336426,
      "learning_rate": 2.896718146718147e-05,
      "loss": 1.6019,
      "step": 321
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.597256660461426,
      "learning_rate": 2.8963963963963963e-05,
      "loss": 1.2632,
      "step": 322
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.48445987701416,
      "learning_rate": 2.896074646074646e-05,
      "loss": 1.5813,
      "step": 323
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.847792625427246,
      "learning_rate": 2.8957528957528956e-05,
      "loss": 1.8397,
      "step": 324
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.153615951538086,
      "learning_rate": 2.8954311454311452e-05,
      "loss": 1.6184,
      "step": 325
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.613450050354004,
      "learning_rate": 2.8951093951093952e-05,
      "loss": 1.4939,
      "step": 326
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.086597442626953,
      "learning_rate": 2.894787644787645e-05,
      "loss": 1.8665,
      "step": 327
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.73073959350586,
      "learning_rate": 2.8944658944658945e-05,
      "loss": 1.8261,
      "step": 328
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.485965728759766,
      "learning_rate": 2.894144144144144e-05,
      "loss": 1.7022,
      "step": 329
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.591164588928223,
      "learning_rate": 2.8938223938223938e-05,
      "loss": 1.743,
      "step": 330
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.72520923614502,
      "learning_rate": 2.8935006435006435e-05,
      "loss": 1.7792,
      "step": 331
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.281352996826172,
      "learning_rate": 2.8931788931788934e-05,
      "loss": 1.71,
      "step": 332
    },
    {
      "epoch": 0.07,
      "grad_norm": 7.847713470458984,
      "learning_rate": 2.892857142857143e-05,
      "loss": 1.8267,
      "step": 333
    },
    {
      "epoch": 0.07,
      "grad_norm": 7.802544593811035,
      "learning_rate": 2.8925353925353927e-05,
      "loss": 1.476,
      "step": 334
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.149658203125,
      "learning_rate": 2.8922136422136424e-05,
      "loss": 1.764,
      "step": 335
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.84017276763916,
      "learning_rate": 2.891891891891892e-05,
      "loss": 1.8691,
      "step": 336
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.265738487243652,
      "learning_rate": 2.8915701415701417e-05,
      "loss": 1.4991,
      "step": 337
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.4511079788208,
      "learning_rate": 2.8912483912483913e-05,
      "loss": 1.6088,
      "step": 338
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.426203727722168,
      "learning_rate": 2.8909266409266413e-05,
      "loss": 1.3728,
      "step": 339
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.591914176940918,
      "learning_rate": 2.8906048906048906e-05,
      "loss": 1.2948,
      "step": 340
    },
    {
      "epoch": 0.07,
      "grad_norm": 12.76346206665039,
      "learning_rate": 2.8902831402831402e-05,
      "loss": 1.9353,
      "step": 341
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.622657775878906,
      "learning_rate": 2.88996138996139e-05,
      "loss": 1.6749,
      "step": 342
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.758122444152832,
      "learning_rate": 2.8896396396396395e-05,
      "loss": 1.6257,
      "step": 343
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.336207389831543,
      "learning_rate": 2.8893178893178892e-05,
      "loss": 1.5002,
      "step": 344
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.601122856140137,
      "learning_rate": 2.888996138996139e-05,
      "loss": 1.5545,
      "step": 345
    },
    {
      "epoch": 0.07,
      "grad_norm": 10.955584526062012,
      "learning_rate": 2.8886743886743888e-05,
      "loss": 1.8233,
      "step": 346
    },
    {
      "epoch": 0.07,
      "grad_norm": 7.997803211212158,
      "learning_rate": 2.8883526383526385e-05,
      "loss": 1.4065,
      "step": 347
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.929220199584961,
      "learning_rate": 2.888030888030888e-05,
      "loss": 1.7572,
      "step": 348
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.001924514770508,
      "learning_rate": 2.8877091377091377e-05,
      "loss": 1.4635,
      "step": 349
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.525696754455566,
      "learning_rate": 2.8873873873873874e-05,
      "loss": 1.4214,
      "step": 350
    },
    {
      "epoch": 0.08,
      "grad_norm": 7.768222808837891,
      "learning_rate": 2.887065637065637e-05,
      "loss": 1.3091,
      "step": 351
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.194113731384277,
      "learning_rate": 2.886743886743887e-05,
      "loss": 1.7306,
      "step": 352
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.646299362182617,
      "learning_rate": 2.8864221364221367e-05,
      "loss": 1.1928,
      "step": 353
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.782073974609375,
      "learning_rate": 2.8861003861003863e-05,
      "loss": 1.5565,
      "step": 354
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.89830207824707,
      "learning_rate": 2.885778635778636e-05,
      "loss": 1.7056,
      "step": 355
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.227059364318848,
      "learning_rate": 2.8854568854568856e-05,
      "loss": 1.8796,
      "step": 356
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.322881698608398,
      "learning_rate": 2.885135135135135e-05,
      "loss": 1.3601,
      "step": 357
    },
    {
      "epoch": 0.08,
      "grad_norm": 11.827625274658203,
      "learning_rate": 2.884813384813385e-05,
      "loss": 1.9527,
      "step": 358
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.733790397644043,
      "learning_rate": 2.8844916344916345e-05,
      "loss": 1.4902,
      "step": 359
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.484691619873047,
      "learning_rate": 2.8841698841698842e-05,
      "loss": 1.7971,
      "step": 360
    },
    {
      "epoch": 0.08,
      "grad_norm": 7.481174468994141,
      "learning_rate": 2.8838481338481338e-05,
      "loss": 1.0974,
      "step": 361
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.60807991027832,
      "learning_rate": 2.8835263835263835e-05,
      "loss": 1.6471,
      "step": 362
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.174698829650879,
      "learning_rate": 2.883204633204633e-05,
      "loss": 1.472,
      "step": 363
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.010536193847656,
      "learning_rate": 2.8828828828828828e-05,
      "loss": 1.8161,
      "step": 364
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.38068962097168,
      "learning_rate": 2.8825611325611327e-05,
      "loss": 1.8723,
      "step": 365
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.903082847595215,
      "learning_rate": 2.8822393822393824e-05,
      "loss": 1.5936,
      "step": 366
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.608054161071777,
      "learning_rate": 2.881917631917632e-05,
      "loss": 1.4486,
      "step": 367
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.283194541931152,
      "learning_rate": 2.8815958815958817e-05,
      "loss": 1.5292,
      "step": 368
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.90308666229248,
      "learning_rate": 2.8812741312741313e-05,
      "loss": 1.6058,
      "step": 369
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.989293098449707,
      "learning_rate": 2.880952380952381e-05,
      "loss": 1.7142,
      "step": 370
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.62183952331543,
      "learning_rate": 2.880630630630631e-05,
      "loss": 1.626,
      "step": 371
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.145082473754883,
      "learning_rate": 2.8803088803088806e-05,
      "loss": 1.4683,
      "step": 372
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.083850860595703,
      "learning_rate": 2.8799871299871302e-05,
      "loss": 1.4654,
      "step": 373
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.44560718536377,
      "learning_rate": 2.87966537966538e-05,
      "loss": 1.4096,
      "step": 374
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.780258178710938,
      "learning_rate": 2.8793436293436292e-05,
      "loss": 1.7761,
      "step": 375
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.009560585021973,
      "learning_rate": 2.879021879021879e-05,
      "loss": 1.5948,
      "step": 376
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.756340026855469,
      "learning_rate": 2.8787001287001288e-05,
      "loss": 1.4667,
      "step": 377
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.569910049438477,
      "learning_rate": 2.8783783783783785e-05,
      "loss": 1.5761,
      "step": 378
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.971969604492188,
      "learning_rate": 2.878056628056628e-05,
      "loss": 1.6869,
      "step": 379
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.162534713745117,
      "learning_rate": 2.8777348777348778e-05,
      "loss": 1.5161,
      "step": 380
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.335193634033203,
      "learning_rate": 2.8774131274131274e-05,
      "loss": 1.8033,
      "step": 381
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.140092849731445,
      "learning_rate": 2.877091377091377e-05,
      "loss": 1.6104,
      "step": 382
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.404397964477539,
      "learning_rate": 2.8767696267696267e-05,
      "loss": 1.3751,
      "step": 383
    },
    {
      "epoch": 0.08,
      "grad_norm": 7.868209362030029,
      "learning_rate": 2.8764478764478767e-05,
      "loss": 1.4247,
      "step": 384
    },
    {
      "epoch": 0.08,
      "grad_norm": 7.753672122955322,
      "learning_rate": 2.8761261261261263e-05,
      "loss": 1.4204,
      "step": 385
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.097761154174805,
      "learning_rate": 2.875804375804376e-05,
      "loss": 1.418,
      "step": 386
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.744604110717773,
      "learning_rate": 2.8754826254826256e-05,
      "loss": 1.5857,
      "step": 387
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.471952438354492,
      "learning_rate": 2.8751608751608753e-05,
      "loss": 1.5013,
      "step": 388
    },
    {
      "epoch": 0.08,
      "grad_norm": 7.873376846313477,
      "learning_rate": 2.874839124839125e-05,
      "loss": 1.4834,
      "step": 389
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.227154731750488,
      "learning_rate": 2.874517374517375e-05,
      "loss": 1.3534,
      "step": 390
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.1671781539917,
      "learning_rate": 2.8741956241956245e-05,
      "loss": 1.6937,
      "step": 391
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.339088439941406,
      "learning_rate": 2.873873873873874e-05,
      "loss": 1.6216,
      "step": 392
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.268199920654297,
      "learning_rate": 2.8735521235521235e-05,
      "loss": 1.579,
      "step": 393
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.86851978302002,
      "learning_rate": 2.873230373230373e-05,
      "loss": 1.7057,
      "step": 394
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.860312461853027,
      "learning_rate": 2.8729086229086228e-05,
      "loss": 1.5693,
      "step": 395
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.150809288024902,
      "learning_rate": 2.8725868725868724e-05,
      "loss": 1.7087,
      "step": 396
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.827011108398438,
      "learning_rate": 2.8722651222651224e-05,
      "loss": 1.9053,
      "step": 397
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.569929122924805,
      "learning_rate": 2.871943371943372e-05,
      "loss": 1.534,
      "step": 398
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.509927749633789,
      "learning_rate": 2.8716216216216217e-05,
      "loss": 1.537,
      "step": 399
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.926268577575684,
      "learning_rate": 2.8712998712998713e-05,
      "loss": 1.5972,
      "step": 400
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.0410661697387695,
      "learning_rate": 2.870978120978121e-05,
      "loss": 1.3274,
      "step": 401
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.83040714263916,
      "learning_rate": 2.8706563706563706e-05,
      "loss": 1.5006,
      "step": 402
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.718032836914062,
      "learning_rate": 2.8703346203346206e-05,
      "loss": 1.8471,
      "step": 403
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.6228766441345215,
      "learning_rate": 2.8700128700128703e-05,
      "loss": 1.4329,
      "step": 404
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.065009117126465,
      "learning_rate": 2.86969111969112e-05,
      "loss": 1.4006,
      "step": 405
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.531225204467773,
      "learning_rate": 2.8693693693693696e-05,
      "loss": 1.7577,
      "step": 406
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.86160945892334,
      "learning_rate": 2.8690476190476192e-05,
      "loss": 1.539,
      "step": 407
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.156512260437012,
      "learning_rate": 2.868725868725869e-05,
      "loss": 1.5069,
      "step": 408
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.211043357849121,
      "learning_rate": 2.868404118404118e-05,
      "loss": 1.5222,
      "step": 409
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.6387939453125,
      "learning_rate": 2.868082368082368e-05,
      "loss": 1.6675,
      "step": 410
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.85531234741211,
      "learning_rate": 2.8677606177606178e-05,
      "loss": 1.4694,
      "step": 411
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.482146263122559,
      "learning_rate": 2.8674388674388674e-05,
      "loss": 1.783,
      "step": 412
    },
    {
      "epoch": 0.09,
      "grad_norm": 12.329071044921875,
      "learning_rate": 2.867117117117117e-05,
      "loss": 1.4089,
      "step": 413
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.090802192687988,
      "learning_rate": 2.8667953667953667e-05,
      "loss": 1.295,
      "step": 414
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.04483413696289,
      "learning_rate": 2.8664736164736164e-05,
      "loss": 1.3295,
      "step": 415
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.669087409973145,
      "learning_rate": 2.8661518661518663e-05,
      "loss": 1.8036,
      "step": 416
    },
    {
      "epoch": 0.09,
      "grad_norm": 13.535530090332031,
      "learning_rate": 2.865830115830116e-05,
      "loss": 1.7036,
      "step": 417
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.046473503112793,
      "learning_rate": 2.8655083655083656e-05,
      "loss": 1.493,
      "step": 418
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.917839050292969,
      "learning_rate": 2.8651866151866153e-05,
      "loss": 1.4308,
      "step": 419
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.354515075683594,
      "learning_rate": 2.864864864864865e-05,
      "loss": 1.5685,
      "step": 420
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.415287971496582,
      "learning_rate": 2.8645431145431146e-05,
      "loss": 1.8738,
      "step": 421
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.703280448913574,
      "learning_rate": 2.8642213642213642e-05,
      "loss": 1.5165,
      "step": 422
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.424654960632324,
      "learning_rate": 2.8638996138996142e-05,
      "loss": 1.5706,
      "step": 423
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.973247528076172,
      "learning_rate": 2.863577863577864e-05,
      "loss": 1.7312,
      "step": 424
    },
    {
      "epoch": 0.09,
      "grad_norm": 11.688663482666016,
      "learning_rate": 2.8632561132561135e-05,
      "loss": 1.495,
      "step": 425
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.543071746826172,
      "learning_rate": 2.862934362934363e-05,
      "loss": 1.6672,
      "step": 426
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.372920036315918,
      "learning_rate": 2.8626126126126124e-05,
      "loss": 1.4728,
      "step": 427
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.020102500915527,
      "learning_rate": 2.862290862290862e-05,
      "loss": 1.4035,
      "step": 428
    },
    {
      "epoch": 0.09,
      "grad_norm": 11.362913131713867,
      "learning_rate": 2.861969111969112e-05,
      "loss": 1.6751,
      "step": 429
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.63509464263916,
      "learning_rate": 2.8616473616473617e-05,
      "loss": 1.7556,
      "step": 430
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.921652317047119,
      "learning_rate": 2.8613256113256114e-05,
      "loss": 1.6534,
      "step": 431
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.820009231567383,
      "learning_rate": 2.861003861003861e-05,
      "loss": 1.4621,
      "step": 432
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.026673316955566,
      "learning_rate": 2.8606821106821107e-05,
      "loss": 1.5412,
      "step": 433
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.272088527679443,
      "learning_rate": 2.8603603603603603e-05,
      "loss": 1.3728,
      "step": 434
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.361907958984375,
      "learning_rate": 2.8600386100386103e-05,
      "loss": 1.8042,
      "step": 435
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.566416263580322,
      "learning_rate": 2.85971685971686e-05,
      "loss": 1.6975,
      "step": 436
    },
    {
      "epoch": 0.09,
      "grad_norm": 11.044665336608887,
      "learning_rate": 2.8593951093951096e-05,
      "loss": 1.7214,
      "step": 437
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.738393783569336,
      "learning_rate": 2.8590733590733592e-05,
      "loss": 1.7336,
      "step": 438
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.829809188842773,
      "learning_rate": 2.858751608751609e-05,
      "loss": 1.8688,
      "step": 439
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.359966278076172,
      "learning_rate": 2.8584298584298585e-05,
      "loss": 1.3671,
      "step": 440
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.031973838806152,
      "learning_rate": 2.858108108108108e-05,
      "loss": 1.3024,
      "step": 441
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.532884120941162,
      "learning_rate": 2.857786357786358e-05,
      "loss": 1.2744,
      "step": 442
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.13409423828125,
      "learning_rate": 2.8574646074646078e-05,
      "loss": 1.4091,
      "step": 443
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.146448135375977,
      "learning_rate": 2.857142857142857e-05,
      "loss": 1.5573,
      "step": 444
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.315018653869629,
      "learning_rate": 2.8568211068211067e-05,
      "loss": 1.4514,
      "step": 445
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.48011302947998,
      "learning_rate": 2.8564993564993564e-05,
      "loss": 1.4502,
      "step": 446
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.760313034057617,
      "learning_rate": 2.856177606177606e-05,
      "loss": 1.7463,
      "step": 447
    },
    {
      "epoch": 0.1,
      "grad_norm": 11.578058242797852,
      "learning_rate": 2.855855855855856e-05,
      "loss": 1.9629,
      "step": 448
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.112937927246094,
      "learning_rate": 2.8555341055341057e-05,
      "loss": 1.2898,
      "step": 449
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.03318977355957,
      "learning_rate": 2.8552123552123553e-05,
      "loss": 1.6989,
      "step": 450
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.149540901184082,
      "learning_rate": 2.854890604890605e-05,
      "loss": 1.5371,
      "step": 451
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.531676292419434,
      "learning_rate": 2.8545688545688546e-05,
      "loss": 1.4941,
      "step": 452
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.737814903259277,
      "learning_rate": 2.8542471042471042e-05,
      "loss": 1.2467,
      "step": 453
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.925774574279785,
      "learning_rate": 2.853925353925354e-05,
      "loss": 1.933,
      "step": 454
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.971988677978516,
      "learning_rate": 2.853603603603604e-05,
      "loss": 1.4968,
      "step": 455
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.103311538696289,
      "learning_rate": 2.8532818532818535e-05,
      "loss": 1.28,
      "step": 456
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.4449462890625,
      "learning_rate": 2.852960102960103e-05,
      "loss": 1.365,
      "step": 457
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.075377464294434,
      "learning_rate": 2.8526383526383528e-05,
      "loss": 1.5227,
      "step": 458
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.176989555358887,
      "learning_rate": 2.8523166023166024e-05,
      "loss": 1.5377,
      "step": 459
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.012581825256348,
      "learning_rate": 2.851994851994852e-05,
      "loss": 1.618,
      "step": 460
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.781168937683105,
      "learning_rate": 2.8516731016731017e-05,
      "loss": 1.5577,
      "step": 461
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.60557746887207,
      "learning_rate": 2.8513513513513514e-05,
      "loss": 1.6335,
      "step": 462
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.019732475280762,
      "learning_rate": 2.851029601029601e-05,
      "loss": 1.585,
      "step": 463
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.8321533203125,
      "learning_rate": 2.8507078507078507e-05,
      "loss": 1.3217,
      "step": 464
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.941082000732422,
      "learning_rate": 2.8503861003861003e-05,
      "loss": 1.5203,
      "step": 465
    },
    {
      "epoch": 0.1,
      "grad_norm": 11.907002449035645,
      "learning_rate": 2.85006435006435e-05,
      "loss": 1.7837,
      "step": 466
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.212108612060547,
      "learning_rate": 2.8497425997425996e-05,
      "loss": 1.7492,
      "step": 467
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.431690216064453,
      "learning_rate": 2.8494208494208496e-05,
      "loss": 1.3791,
      "step": 468
    },
    {
      "epoch": 0.1,
      "grad_norm": 12.321235656738281,
      "learning_rate": 2.8490990990990992e-05,
      "loss": 1.6563,
      "step": 469
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.063630104064941,
      "learning_rate": 2.848777348777349e-05,
      "loss": 1.5021,
      "step": 470
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.176536560058594,
      "learning_rate": 2.8484555984555985e-05,
      "loss": 1.2816,
      "step": 471
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.700693130493164,
      "learning_rate": 2.848133848133848e-05,
      "loss": 1.6221,
      "step": 472
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.121115684509277,
      "learning_rate": 2.8478120978120978e-05,
      "loss": 1.6198,
      "step": 473
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.911774635314941,
      "learning_rate": 2.8474903474903478e-05,
      "loss": 1.6621,
      "step": 474
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.903169631958008,
      "learning_rate": 2.8471685971685974e-05,
      "loss": 1.3185,
      "step": 475
    },
    {
      "epoch": 0.1,
      "grad_norm": 7.907713890075684,
      "learning_rate": 2.846846846846847e-05,
      "loss": 1.4916,
      "step": 476
    },
    {
      "epoch": 0.1,
      "grad_norm": 7.987964153289795,
      "learning_rate": 2.8465250965250967e-05,
      "loss": 1.8227,
      "step": 477
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.443188667297363,
      "learning_rate": 2.8462033462033464e-05,
      "loss": 1.3157,
      "step": 478
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.983651161193848,
      "learning_rate": 2.8458815958815957e-05,
      "loss": 1.5749,
      "step": 479
    },
    {
      "epoch": 0.1,
      "grad_norm": 7.793544292449951,
      "learning_rate": 2.8455598455598453e-05,
      "loss": 1.4347,
      "step": 480
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.741055488586426,
      "learning_rate": 2.8452380952380953e-05,
      "loss": 1.6232,
      "step": 481
    },
    {
      "epoch": 0.1,
      "grad_norm": 12.335562705993652,
      "learning_rate": 2.844916344916345e-05,
      "loss": 1.5383,
      "step": 482
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.515409469604492,
      "learning_rate": 2.8445945945945946e-05,
      "loss": 1.5263,
      "step": 483
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.834569931030273,
      "learning_rate": 2.8442728442728442e-05,
      "loss": 1.5996,
      "step": 484
    },
    {
      "epoch": 0.1,
      "grad_norm": 11.758679389953613,
      "learning_rate": 2.843951093951094e-05,
      "loss": 1.8241,
      "step": 485
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.941324234008789,
      "learning_rate": 2.8436293436293435e-05,
      "loss": 1.5296,
      "step": 486
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.728738784790039,
      "learning_rate": 2.8433075933075935e-05,
      "loss": 1.6262,
      "step": 487
    },
    {
      "epoch": 0.1,
      "grad_norm": 7.630702495574951,
      "learning_rate": 2.842985842985843e-05,
      "loss": 1.3919,
      "step": 488
    },
    {
      "epoch": 0.1,
      "grad_norm": 7.374858379364014,
      "learning_rate": 2.8426640926640928e-05,
      "loss": 1.219,
      "step": 489
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.311800003051758,
      "learning_rate": 2.8423423423423425e-05,
      "loss": 1.7162,
      "step": 490
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.917937755584717,
      "learning_rate": 2.842020592020592e-05,
      "loss": 1.2918,
      "step": 491
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.214245796203613,
      "learning_rate": 2.8416988416988417e-05,
      "loss": 1.4018,
      "step": 492
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.303821563720703,
      "learning_rate": 2.8413770913770917e-05,
      "loss": 1.5253,
      "step": 493
    },
    {
      "epoch": 0.11,
      "grad_norm": 11.672669410705566,
      "learning_rate": 2.8410553410553414e-05,
      "loss": 1.7421,
      "step": 494
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.504253387451172,
      "learning_rate": 2.840733590733591e-05,
      "loss": 1.7044,
      "step": 495
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.722330570220947,
      "learning_rate": 2.8404118404118403e-05,
      "loss": 1.4861,
      "step": 496
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.396967887878418,
      "learning_rate": 2.84009009009009e-05,
      "loss": 1.5943,
      "step": 497
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.295594215393066,
      "learning_rate": 2.8397683397683396e-05,
      "loss": 1.404,
      "step": 498
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.767500877380371,
      "learning_rate": 2.8394465894465893e-05,
      "loss": 1.851,
      "step": 499
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.128677368164062,
      "learning_rate": 2.8391248391248392e-05,
      "loss": 1.1262,
      "step": 500
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.29800796508789,
      "learning_rate": 2.838803088803089e-05,
      "loss": 1.4001,
      "step": 501
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.618123054504395,
      "learning_rate": 2.8384813384813385e-05,
      "loss": 1.5072,
      "step": 502
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.187304496765137,
      "learning_rate": 2.8381595881595882e-05,
      "loss": 1.4008,
      "step": 503
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.179542541503906,
      "learning_rate": 2.8378378378378378e-05,
      "loss": 1.6304,
      "step": 504
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.470932960510254,
      "learning_rate": 2.8375160875160875e-05,
      "loss": 1.6314,
      "step": 505
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.709694862365723,
      "learning_rate": 2.8371943371943375e-05,
      "loss": 1.3513,
      "step": 506
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.617488861083984,
      "learning_rate": 2.836872586872587e-05,
      "loss": 1.5829,
      "step": 507
    },
    {
      "epoch": 0.11,
      "grad_norm": 12.400928497314453,
      "learning_rate": 2.8365508365508367e-05,
      "loss": 1.6558,
      "step": 508
    },
    {
      "epoch": 0.11,
      "grad_norm": 12.279901504516602,
      "learning_rate": 2.8362290862290864e-05,
      "loss": 1.8052,
      "step": 509
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.858482360839844,
      "learning_rate": 2.835907335907336e-05,
      "loss": 1.4569,
      "step": 510
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.661469459533691,
      "learning_rate": 2.8355855855855857e-05,
      "loss": 1.3743,
      "step": 511
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.77314567565918,
      "learning_rate": 2.8352638352638353e-05,
      "loss": 1.4962,
      "step": 512
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.824915885925293,
      "learning_rate": 2.8349420849420853e-05,
      "loss": 1.4943,
      "step": 513
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.8538665771484375,
      "learning_rate": 2.8346203346203346e-05,
      "loss": 1.2711,
      "step": 514
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.34766674041748,
      "learning_rate": 2.8342985842985843e-05,
      "loss": 1.4155,
      "step": 515
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.867433547973633,
      "learning_rate": 2.833976833976834e-05,
      "loss": 1.4116,
      "step": 516
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.044109344482422,
      "learning_rate": 2.8336550836550836e-05,
      "loss": 1.4408,
      "step": 517
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.2412691116333,
      "learning_rate": 2.8333333333333332e-05,
      "loss": 1.677,
      "step": 518
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.850975036621094,
      "learning_rate": 2.8330115830115832e-05,
      "loss": 1.4407,
      "step": 519
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.57258129119873,
      "learning_rate": 2.8326898326898328e-05,
      "loss": 1.5908,
      "step": 520
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.022360801696777,
      "learning_rate": 2.8323680823680825e-05,
      "loss": 1.6609,
      "step": 521
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.30443000793457,
      "learning_rate": 2.832046332046332e-05,
      "loss": 0.9983,
      "step": 522
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.00052261352539,
      "learning_rate": 2.8317245817245818e-05,
      "loss": 1.3906,
      "step": 523
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.336234092712402,
      "learning_rate": 2.8314028314028314e-05,
      "loss": 1.3916,
      "step": 524
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.723689556121826,
      "learning_rate": 2.831081081081081e-05,
      "loss": 1.3178,
      "step": 525
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.012369155883789,
      "learning_rate": 2.830759330759331e-05,
      "loss": 1.397,
      "step": 526
    },
    {
      "epoch": 0.11,
      "grad_norm": 11.530648231506348,
      "learning_rate": 2.8304375804375807e-05,
      "loss": 1.8751,
      "step": 527
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.971467018127441,
      "learning_rate": 2.8301158301158303e-05,
      "loss": 1.4203,
      "step": 528
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.046466827392578,
      "learning_rate": 2.82979407979408e-05,
      "loss": 1.5511,
      "step": 529
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.75573444366455,
      "learning_rate": 2.8294723294723296e-05,
      "loss": 1.3978,
      "step": 530
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.166257858276367,
      "learning_rate": 2.829150579150579e-05,
      "loss": 1.4582,
      "step": 531
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.303631782531738,
      "learning_rate": 2.828828828828829e-05,
      "loss": 1.503,
      "step": 532
    },
    {
      "epoch": 0.11,
      "grad_norm": 11.696259498596191,
      "learning_rate": 2.8285070785070786e-05,
      "loss": 1.5636,
      "step": 533
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.09343433380127,
      "learning_rate": 2.8281853281853282e-05,
      "loss": 1.7168,
      "step": 534
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.408019065856934,
      "learning_rate": 2.827863577863578e-05,
      "loss": 1.3302,
      "step": 535
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.63667106628418,
      "learning_rate": 2.8275418275418275e-05,
      "loss": 1.6728,
      "step": 536
    },
    {
      "epoch": 0.12,
      "grad_norm": 7.496615886688232,
      "learning_rate": 2.827220077220077e-05,
      "loss": 1.3307,
      "step": 537
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.884413719177246,
      "learning_rate": 2.8268983268983268e-05,
      "loss": 1.6797,
      "step": 538
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.257469177246094,
      "learning_rate": 2.8265765765765768e-05,
      "loss": 1.6932,
      "step": 539
    },
    {
      "epoch": 0.12,
      "grad_norm": 10.166794776916504,
      "learning_rate": 2.8262548262548264e-05,
      "loss": 1.2707,
      "step": 540
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.012992858886719,
      "learning_rate": 2.825933075933076e-05,
      "loss": 1.4427,
      "step": 541
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.701659202575684,
      "learning_rate": 2.8256113256113257e-05,
      "loss": 1.6735,
      "step": 542
    },
    {
      "epoch": 0.12,
      "grad_norm": 10.042672157287598,
      "learning_rate": 2.8252895752895753e-05,
      "loss": 1.7392,
      "step": 543
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.568726539611816,
      "learning_rate": 2.824967824967825e-05,
      "loss": 1.4627,
      "step": 544
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.704487800598145,
      "learning_rate": 2.824646074646075e-05,
      "loss": 1.4671,
      "step": 545
    },
    {
      "epoch": 0.12,
      "grad_norm": 11.736075401306152,
      "learning_rate": 2.8243243243243246e-05,
      "loss": 1.3183,
      "step": 546
    },
    {
      "epoch": 0.12,
      "grad_norm": 11.734922409057617,
      "learning_rate": 2.8240025740025743e-05,
      "loss": 1.6383,
      "step": 547
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.873263359069824,
      "learning_rate": 2.8236808236808236e-05,
      "loss": 1.6167,
      "step": 548
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.393575668334961,
      "learning_rate": 2.8233590733590732e-05,
      "loss": 1.4461,
      "step": 549
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.794711112976074,
      "learning_rate": 2.823037323037323e-05,
      "loss": 1.3945,
      "step": 550
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.445703506469727,
      "learning_rate": 2.822715572715573e-05,
      "loss": 1.5351,
      "step": 551
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.607747077941895,
      "learning_rate": 2.8223938223938225e-05,
      "loss": 1.767,
      "step": 552
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.415684700012207,
      "learning_rate": 2.822072072072072e-05,
      "loss": 1.6346,
      "step": 553
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.102144241333008,
      "learning_rate": 2.8217503217503218e-05,
      "loss": 1.3684,
      "step": 554
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.341261863708496,
      "learning_rate": 2.8214285714285714e-05,
      "loss": 0.9996,
      "step": 555
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.280244827270508,
      "learning_rate": 2.821106821106821e-05,
      "loss": 1.9282,
      "step": 556
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.245135307312012,
      "learning_rate": 2.8207850707850707e-05,
      "loss": 1.4495,
      "step": 557
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.443021774291992,
      "learning_rate": 2.8204633204633207e-05,
      "loss": 1.1863,
      "step": 558
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.660807609558105,
      "learning_rate": 2.8201415701415703e-05,
      "loss": 1.629,
      "step": 559
    },
    {
      "epoch": 0.12,
      "grad_norm": 10.156190872192383,
      "learning_rate": 2.81981981981982e-05,
      "loss": 1.9366,
      "step": 560
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.634072303771973,
      "learning_rate": 2.8194980694980696e-05,
      "loss": 1.6526,
      "step": 561
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.537891387939453,
      "learning_rate": 2.8191763191763193e-05,
      "loss": 1.548,
      "step": 562
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.821135520935059,
      "learning_rate": 2.818854568854569e-05,
      "loss": 1.6233,
      "step": 563
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.329626083374023,
      "learning_rate": 2.818532818532819e-05,
      "loss": 1.5475,
      "step": 564
    },
    {
      "epoch": 0.12,
      "grad_norm": 10.472304344177246,
      "learning_rate": 2.8182110682110686e-05,
      "loss": 1.6243,
      "step": 565
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.173851013183594,
      "learning_rate": 2.817889317889318e-05,
      "loss": 1.2621,
      "step": 566
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.82541561126709,
      "learning_rate": 2.8175675675675675e-05,
      "loss": 1.6718,
      "step": 567
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.48964786529541,
      "learning_rate": 2.817245817245817e-05,
      "loss": 1.6684,
      "step": 568
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.857165336608887,
      "learning_rate": 2.8169240669240668e-05,
      "loss": 1.458,
      "step": 569
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.145784378051758,
      "learning_rate": 2.8166023166023164e-05,
      "loss": 1.1665,
      "step": 570
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.322288513183594,
      "learning_rate": 2.8162805662805664e-05,
      "loss": 1.5611,
      "step": 571
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.401164054870605,
      "learning_rate": 2.815958815958816e-05,
      "loss": 1.1868,
      "step": 572
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.029150009155273,
      "learning_rate": 2.8156370656370657e-05,
      "loss": 1.2767,
      "step": 573
    },
    {
      "epoch": 0.12,
      "grad_norm": 11.363448143005371,
      "learning_rate": 2.8153153153153154e-05,
      "loss": 1.9192,
      "step": 574
    },
    {
      "epoch": 0.12,
      "grad_norm": 10.484262466430664,
      "learning_rate": 2.814993564993565e-05,
      "loss": 1.6542,
      "step": 575
    },
    {
      "epoch": 0.12,
      "grad_norm": 11.45032024383545,
      "learning_rate": 2.8146718146718146e-05,
      "loss": 1.7833,
      "step": 576
    },
    {
      "epoch": 0.12,
      "grad_norm": 10.225872039794922,
      "learning_rate": 2.8143500643500646e-05,
      "loss": 1.4353,
      "step": 577
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.412259101867676,
      "learning_rate": 2.8140283140283143e-05,
      "loss": 1.2534,
      "step": 578
    },
    {
      "epoch": 0.12,
      "grad_norm": 10.819499969482422,
      "learning_rate": 2.813706563706564e-05,
      "loss": 1.7296,
      "step": 579
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.979551315307617,
      "learning_rate": 2.8133848133848136e-05,
      "loss": 1.5392,
      "step": 580
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.53780460357666,
      "learning_rate": 2.8130630630630632e-05,
      "loss": 1.4133,
      "step": 581
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.801671981811523,
      "learning_rate": 2.812741312741313e-05,
      "loss": 1.3808,
      "step": 582
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.140606880187988,
      "learning_rate": 2.812419562419562e-05,
      "loss": 1.2714,
      "step": 583
    },
    {
      "epoch": 0.13,
      "grad_norm": 11.589751243591309,
      "learning_rate": 2.812097812097812e-05,
      "loss": 1.5622,
      "step": 584
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.625310897827148,
      "learning_rate": 2.8117760617760618e-05,
      "loss": 1.2961,
      "step": 585
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.121614456176758,
      "learning_rate": 2.8114543114543114e-05,
      "loss": 1.7412,
      "step": 586
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.976176261901855,
      "learning_rate": 2.811132561132561e-05,
      "loss": 1.8116,
      "step": 587
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.145041465759277,
      "learning_rate": 2.8108108108108107e-05,
      "loss": 1.6325,
      "step": 588
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.958148002624512,
      "learning_rate": 2.8104890604890604e-05,
      "loss": 1.9701,
      "step": 589
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.826210975646973,
      "learning_rate": 2.8101673101673104e-05,
      "loss": 1.6554,
      "step": 590
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.952293395996094,
      "learning_rate": 2.80984555984556e-05,
      "loss": 1.4829,
      "step": 591
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.050843238830566,
      "learning_rate": 2.8095238095238096e-05,
      "loss": 1.288,
      "step": 592
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.314046859741211,
      "learning_rate": 2.8092020592020593e-05,
      "loss": 1.7996,
      "step": 593
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.6920552253723145,
      "learning_rate": 2.808880308880309e-05,
      "loss": 1.7171,
      "step": 594
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.757037162780762,
      "learning_rate": 2.8085585585585586e-05,
      "loss": 1.6972,
      "step": 595
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.46728515625,
      "learning_rate": 2.8082368082368082e-05,
      "loss": 1.4573,
      "step": 596
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.68526554107666,
      "learning_rate": 2.8079150579150582e-05,
      "loss": 1.6794,
      "step": 597
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.271533012390137,
      "learning_rate": 2.807593307593308e-05,
      "loss": 1.5844,
      "step": 598
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.662840843200684,
      "learning_rate": 2.8072715572715575e-05,
      "loss": 1.7532,
      "step": 599
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.755474090576172,
      "learning_rate": 2.806949806949807e-05,
      "loss": 1.5928,
      "step": 600
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.844522476196289,
      "learning_rate": 2.8066280566280565e-05,
      "loss": 1.399,
      "step": 601
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.06291675567627,
      "learning_rate": 2.806306306306306e-05,
      "loss": 1.4684,
      "step": 602
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.857059478759766,
      "learning_rate": 2.805984555984556e-05,
      "loss": 1.3241,
      "step": 603
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.107574462890625,
      "learning_rate": 2.8056628056628057e-05,
      "loss": 1.4189,
      "step": 604
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.092050552368164,
      "learning_rate": 2.8053410553410554e-05,
      "loss": 1.6754,
      "step": 605
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.167607307434082,
      "learning_rate": 2.805019305019305e-05,
      "loss": 1.4439,
      "step": 606
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.787460327148438,
      "learning_rate": 2.8046975546975547e-05,
      "loss": 1.3436,
      "step": 607
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.592608451843262,
      "learning_rate": 2.8043758043758043e-05,
      "loss": 1.4311,
      "step": 608
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.879838943481445,
      "learning_rate": 2.8040540540540543e-05,
      "loss": 1.9999,
      "step": 609
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.327800750732422,
      "learning_rate": 2.803732303732304e-05,
      "loss": 1.3962,
      "step": 610
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.972732543945312,
      "learning_rate": 2.8034105534105536e-05,
      "loss": 1.651,
      "step": 611
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.581597328186035,
      "learning_rate": 2.8030888030888032e-05,
      "loss": 1.6383,
      "step": 612
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.0392427444458,
      "learning_rate": 2.802767052767053e-05,
      "loss": 1.4925,
      "step": 613
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.917680740356445,
      "learning_rate": 2.8024453024453025e-05,
      "loss": 1.6695,
      "step": 614
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.463658332824707,
      "learning_rate": 2.802123552123552e-05,
      "loss": 1.5277,
      "step": 615
    },
    {
      "epoch": 0.13,
      "grad_norm": 6.949864864349365,
      "learning_rate": 2.801801801801802e-05,
      "loss": 1.2164,
      "step": 616
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.872100830078125,
      "learning_rate": 2.8014800514800518e-05,
      "loss": 1.327,
      "step": 617
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.577195167541504,
      "learning_rate": 2.801158301158301e-05,
      "loss": 1.5494,
      "step": 618
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.389220237731934,
      "learning_rate": 2.8008365508365507e-05,
      "loss": 1.2892,
      "step": 619
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.15446949005127,
      "learning_rate": 2.8005148005148004e-05,
      "loss": 1.4998,
      "step": 620
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.357885360717773,
      "learning_rate": 2.80019305019305e-05,
      "loss": 1.6105,
      "step": 621
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.677713394165039,
      "learning_rate": 2.7998712998713e-05,
      "loss": 1.3427,
      "step": 622
    },
    {
      "epoch": 0.13,
      "grad_norm": 11.13674259185791,
      "learning_rate": 2.7995495495495497e-05,
      "loss": 1.8271,
      "step": 623
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.782164573669434,
      "learning_rate": 2.7992277992277993e-05,
      "loss": 1.4724,
      "step": 624
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.700124740600586,
      "learning_rate": 2.798906048906049e-05,
      "loss": 1.5562,
      "step": 625
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.86253833770752,
      "learning_rate": 2.7985842985842986e-05,
      "loss": 1.4933,
      "step": 626
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.175491333007812,
      "learning_rate": 2.7982625482625482e-05,
      "loss": 1.3469,
      "step": 627
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.145636558532715,
      "learning_rate": 2.797940797940798e-05,
      "loss": 1.6114,
      "step": 628
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.13968276977539,
      "learning_rate": 2.797619047619048e-05,
      "loss": 1.4646,
      "step": 629
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.938583374023438,
      "learning_rate": 2.7972972972972975e-05,
      "loss": 1.2395,
      "step": 630
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.644238471984863,
      "learning_rate": 2.796975546975547e-05,
      "loss": 1.4098,
      "step": 631
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.137377738952637,
      "learning_rate": 2.7966537966537968e-05,
      "loss": 1.3734,
      "step": 632
    },
    {
      "epoch": 0.14,
      "grad_norm": 10.504596710205078,
      "learning_rate": 2.7963320463320465e-05,
      "loss": 1.4543,
      "step": 633
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.67316722869873,
      "learning_rate": 2.796010296010296e-05,
      "loss": 1.4459,
      "step": 634
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.099815368652344,
      "learning_rate": 2.7956885456885457e-05,
      "loss": 1.3916,
      "step": 635
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.864910125732422,
      "learning_rate": 2.7953667953667954e-05,
      "loss": 1.6999,
      "step": 636
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.740618705749512,
      "learning_rate": 2.795045045045045e-05,
      "loss": 1.6399,
      "step": 637
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.605246543884277,
      "learning_rate": 2.7947232947232947e-05,
      "loss": 1.5655,
      "step": 638
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.681963920593262,
      "learning_rate": 2.7944015444015443e-05,
      "loss": 1.4742,
      "step": 639
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.088510513305664,
      "learning_rate": 2.794079794079794e-05,
      "loss": 1.5527,
      "step": 640
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.884695053100586,
      "learning_rate": 2.7937580437580436e-05,
      "loss": 1.3302,
      "step": 641
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.48004150390625,
      "learning_rate": 2.7934362934362936e-05,
      "loss": 1.4265,
      "step": 642
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.021172523498535,
      "learning_rate": 2.7931145431145432e-05,
      "loss": 1.3767,
      "step": 643
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.317270278930664,
      "learning_rate": 2.792792792792793e-05,
      "loss": 1.436,
      "step": 644
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.234036445617676,
      "learning_rate": 2.7924710424710425e-05,
      "loss": 1.3248,
      "step": 645
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.925457954406738,
      "learning_rate": 2.7921492921492922e-05,
      "loss": 1.4976,
      "step": 646
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.194786071777344,
      "learning_rate": 2.7918275418275418e-05,
      "loss": 1.5925,
      "step": 647
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.920014381408691,
      "learning_rate": 2.7915057915057918e-05,
      "loss": 1.7883,
      "step": 648
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.926551818847656,
      "learning_rate": 2.7911840411840415e-05,
      "loss": 1.4319,
      "step": 649
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.449362754821777,
      "learning_rate": 2.790862290862291e-05,
      "loss": 1.5951,
      "step": 650
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.75369644165039,
      "learning_rate": 2.7905405405405407e-05,
      "loss": 1.649,
      "step": 651
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.313851356506348,
      "learning_rate": 2.7902187902187904e-05,
      "loss": 1.5726,
      "step": 652
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.5188627243042,
      "learning_rate": 2.7898970398970397e-05,
      "loss": 1.4618,
      "step": 653
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.154404640197754,
      "learning_rate": 2.7895752895752897e-05,
      "loss": 1.8117,
      "step": 654
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.726032257080078,
      "learning_rate": 2.7892535392535393e-05,
      "loss": 1.396,
      "step": 655
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.242643356323242,
      "learning_rate": 2.788931788931789e-05,
      "loss": 1.3345,
      "step": 656
    },
    {
      "epoch": 0.14,
      "grad_norm": 7.823841571807861,
      "learning_rate": 2.7886100386100386e-05,
      "loss": 1.4311,
      "step": 657
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.599079132080078,
      "learning_rate": 2.7882882882882883e-05,
      "loss": 1.4775,
      "step": 658
    },
    {
      "epoch": 0.14,
      "grad_norm": 7.616494655609131,
      "learning_rate": 2.787966537966538e-05,
      "loss": 1.2921,
      "step": 659
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.896546363830566,
      "learning_rate": 2.7876447876447876e-05,
      "loss": 1.5566,
      "step": 660
    },
    {
      "epoch": 0.14,
      "grad_norm": 7.429052829742432,
      "learning_rate": 2.7873230373230375e-05,
      "loss": 1.331,
      "step": 661
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.286176681518555,
      "learning_rate": 2.7870012870012872e-05,
      "loss": 1.8036,
      "step": 662
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.402307510375977,
      "learning_rate": 2.7866795366795368e-05,
      "loss": 1.5199,
      "step": 663
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.715972900390625,
      "learning_rate": 2.7863577863577865e-05,
      "loss": 1.5739,
      "step": 664
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.694830894470215,
      "learning_rate": 2.786036036036036e-05,
      "loss": 1.2969,
      "step": 665
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.581073760986328,
      "learning_rate": 2.7857142857142858e-05,
      "loss": 1.3532,
      "step": 666
    },
    {
      "epoch": 0.14,
      "grad_norm": 10.60390853881836,
      "learning_rate": 2.7853925353925357e-05,
      "loss": 2.2595,
      "step": 667
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.55583381652832,
      "learning_rate": 2.7850707850707854e-05,
      "loss": 1.5682,
      "step": 668
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.109996795654297,
      "learning_rate": 2.784749034749035e-05,
      "loss": 1.4675,
      "step": 669
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.673164367675781,
      "learning_rate": 2.7844272844272843e-05,
      "loss": 1.6466,
      "step": 670
    },
    {
      "epoch": 0.14,
      "grad_norm": 9.182644844055176,
      "learning_rate": 2.784105534105534e-05,
      "loss": 1.4298,
      "step": 671
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.009683609008789,
      "learning_rate": 2.7837837837837836e-05,
      "loss": 1.2519,
      "step": 672
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.576504707336426,
      "learning_rate": 2.7834620334620333e-05,
      "loss": 1.2513,
      "step": 673
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.572776794433594,
      "learning_rate": 2.7831402831402833e-05,
      "loss": 1.4674,
      "step": 674
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.332525253295898,
      "learning_rate": 2.782818532818533e-05,
      "loss": 1.41,
      "step": 675
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.377999305725098,
      "learning_rate": 2.7824967824967826e-05,
      "loss": 1.6276,
      "step": 676
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.535334587097168,
      "learning_rate": 2.7821750321750322e-05,
      "loss": 1.3022,
      "step": 677
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.462254524230957,
      "learning_rate": 2.781853281853282e-05,
      "loss": 1.3189,
      "step": 678
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.836193084716797,
      "learning_rate": 2.7815315315315315e-05,
      "loss": 1.6843,
      "step": 679
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.202351570129395,
      "learning_rate": 2.7812097812097815e-05,
      "loss": 1.6985,
      "step": 680
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.311161041259766,
      "learning_rate": 2.780888030888031e-05,
      "loss": 1.395,
      "step": 681
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.666946411132812,
      "learning_rate": 2.7805662805662808e-05,
      "loss": 1.1534,
      "step": 682
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.82536792755127,
      "learning_rate": 2.7802445302445304e-05,
      "loss": 1.3774,
      "step": 683
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.099430084228516,
      "learning_rate": 2.77992277992278e-05,
      "loss": 1.3959,
      "step": 684
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.378087997436523,
      "learning_rate": 2.7796010296010297e-05,
      "loss": 1.5988,
      "step": 685
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.572342872619629,
      "learning_rate": 2.7792792792792793e-05,
      "loss": 1.3522,
      "step": 686
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.941052436828613,
      "learning_rate": 2.7789575289575293e-05,
      "loss": 1.5682,
      "step": 687
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.173140525817871,
      "learning_rate": 2.7786357786357786e-05,
      "loss": 1.2968,
      "step": 688
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.245065689086914,
      "learning_rate": 2.7783140283140283e-05,
      "loss": 1.3433,
      "step": 689
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.966949462890625,
      "learning_rate": 2.777992277992278e-05,
      "loss": 1.2641,
      "step": 690
    },
    {
      "epoch": 0.15,
      "grad_norm": 11.538064956665039,
      "learning_rate": 2.7776705276705276e-05,
      "loss": 1.4936,
      "step": 691
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.732046127319336,
      "learning_rate": 2.7773487773487772e-05,
      "loss": 1.1586,
      "step": 692
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.79343318939209,
      "learning_rate": 2.7770270270270272e-05,
      "loss": 1.5392,
      "step": 693
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.990402221679688,
      "learning_rate": 2.776705276705277e-05,
      "loss": 1.6903,
      "step": 694
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.542869567871094,
      "learning_rate": 2.7763835263835265e-05,
      "loss": 1.2517,
      "step": 695
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.0194673538208,
      "learning_rate": 2.776061776061776e-05,
      "loss": 1.2672,
      "step": 696
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.030797958374023,
      "learning_rate": 2.7757400257400258e-05,
      "loss": 1.3912,
      "step": 697
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.990483283996582,
      "learning_rate": 2.7754182754182754e-05,
      "loss": 1.7592,
      "step": 698
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.35617446899414,
      "learning_rate": 2.775096525096525e-05,
      "loss": 1.2677,
      "step": 699
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.946454048156738,
      "learning_rate": 2.774774774774775e-05,
      "loss": 1.5389,
      "step": 700
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.611677169799805,
      "learning_rate": 2.7744530244530247e-05,
      "loss": 1.4919,
      "step": 701
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.203311920166016,
      "learning_rate": 2.7741312741312743e-05,
      "loss": 1.2322,
      "step": 702
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.620719909667969,
      "learning_rate": 2.773809523809524e-05,
      "loss": 1.3085,
      "step": 703
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.017184257507324,
      "learning_rate": 2.7734877734877736e-05,
      "loss": 1.3121,
      "step": 704
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.665204048156738,
      "learning_rate": 2.773166023166023e-05,
      "loss": 1.496,
      "step": 705
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.024779319763184,
      "learning_rate": 2.772844272844273e-05,
      "loss": 1.4432,
      "step": 706
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.491860389709473,
      "learning_rate": 2.7725225225225226e-05,
      "loss": 1.4735,
      "step": 707
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.262229919433594,
      "learning_rate": 2.7722007722007722e-05,
      "loss": 1.4495,
      "step": 708
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.640606880187988,
      "learning_rate": 2.771879021879022e-05,
      "loss": 1.5145,
      "step": 709
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.900903701782227,
      "learning_rate": 2.7715572715572715e-05,
      "loss": 1.3629,
      "step": 710
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.158628463745117,
      "learning_rate": 2.771235521235521e-05,
      "loss": 1.6161,
      "step": 711
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.362253189086914,
      "learning_rate": 2.7709137709137708e-05,
      "loss": 1.0367,
      "step": 712
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.973403930664062,
      "learning_rate": 2.7705920205920208e-05,
      "loss": 1.4437,
      "step": 713
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.65727424621582,
      "learning_rate": 2.7702702702702704e-05,
      "loss": 1.3538,
      "step": 714
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.098772048950195,
      "learning_rate": 2.76994851994852e-05,
      "loss": 1.3895,
      "step": 715
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.1404390335083,
      "learning_rate": 2.7696267696267697e-05,
      "loss": 1.5166,
      "step": 716
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.1705961227417,
      "learning_rate": 2.7693050193050194e-05,
      "loss": 1.4049,
      "step": 717
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.375228881835938,
      "learning_rate": 2.768983268983269e-05,
      "loss": 1.5007,
      "step": 718
    },
    {
      "epoch": 0.15,
      "grad_norm": 9.192866325378418,
      "learning_rate": 2.768661518661519e-05,
      "loss": 1.6184,
      "step": 719
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.870221138000488,
      "learning_rate": 2.7683397683397686e-05,
      "loss": 1.5483,
      "step": 720
    },
    {
      "epoch": 0.15,
      "grad_norm": 7.485607624053955,
      "learning_rate": 2.7680180180180183e-05,
      "loss": 1.1655,
      "step": 721
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.825922966003418,
      "learning_rate": 2.7676962676962676e-05,
      "loss": 1.3303,
      "step": 722
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.918988227844238,
      "learning_rate": 2.7673745173745172e-05,
      "loss": 1.6301,
      "step": 723
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.548290252685547,
      "learning_rate": 2.767052767052767e-05,
      "loss": 1.1208,
      "step": 724
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.198466300964355,
      "learning_rate": 2.766731016731017e-05,
      "loss": 1.4303,
      "step": 725
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.614500045776367,
      "learning_rate": 2.7664092664092665e-05,
      "loss": 1.5648,
      "step": 726
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.375494003295898,
      "learning_rate": 2.766087516087516e-05,
      "loss": 1.2387,
      "step": 727
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.936331748962402,
      "learning_rate": 2.7657657657657658e-05,
      "loss": 1.518,
      "step": 728
    },
    {
      "epoch": 0.16,
      "grad_norm": 10.05263900756836,
      "learning_rate": 2.7654440154440154e-05,
      "loss": 1.4939,
      "step": 729
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.193610191345215,
      "learning_rate": 2.765122265122265e-05,
      "loss": 1.6714,
      "step": 730
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.395607948303223,
      "learning_rate": 2.7648005148005147e-05,
      "loss": 1.5469,
      "step": 731
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.116291999816895,
      "learning_rate": 2.7644787644787647e-05,
      "loss": 1.241,
      "step": 732
    },
    {
      "epoch": 0.16,
      "grad_norm": 10.014991760253906,
      "learning_rate": 2.7641570141570144e-05,
      "loss": 1.4881,
      "step": 733
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.837320327758789,
      "learning_rate": 2.763835263835264e-05,
      "loss": 1.3442,
      "step": 734
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.034908294677734,
      "learning_rate": 2.7635135135135136e-05,
      "loss": 1.1655,
      "step": 735
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.411362648010254,
      "learning_rate": 2.7631917631917633e-05,
      "loss": 1.3116,
      "step": 736
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.757721900939941,
      "learning_rate": 2.762870012870013e-05,
      "loss": 1.5726,
      "step": 737
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.088129043579102,
      "learning_rate": 2.762548262548263e-05,
      "loss": 1.2605,
      "step": 738
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.096224784851074,
      "learning_rate": 2.7622265122265126e-05,
      "loss": 1.2551,
      "step": 739
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.317255973815918,
      "learning_rate": 2.761904761904762e-05,
      "loss": 1.4305,
      "step": 740
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.0960693359375,
      "learning_rate": 2.7615830115830115e-05,
      "loss": 1.3437,
      "step": 741
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.755814075469971,
      "learning_rate": 2.761261261261261e-05,
      "loss": 1.44,
      "step": 742
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.628165245056152,
      "learning_rate": 2.7609395109395108e-05,
      "loss": 1.2891,
      "step": 743
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.555726051330566,
      "learning_rate": 2.7606177606177605e-05,
      "loss": 1.7445,
      "step": 744
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.333622455596924,
      "learning_rate": 2.7602960102960104e-05,
      "loss": 1.1731,
      "step": 745
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.902264595031738,
      "learning_rate": 2.75997425997426e-05,
      "loss": 1.5843,
      "step": 746
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.912365913391113,
      "learning_rate": 2.7596525096525097e-05,
      "loss": 1.6292,
      "step": 747
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.367301940917969,
      "learning_rate": 2.7593307593307594e-05,
      "loss": 1.2877,
      "step": 748
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.941542625427246,
      "learning_rate": 2.759009009009009e-05,
      "loss": 1.4867,
      "step": 749
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.630366325378418,
      "learning_rate": 2.7586872586872587e-05,
      "loss": 1.1417,
      "step": 750
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.438949584960938,
      "learning_rate": 2.7583655083655086e-05,
      "loss": 1.5235,
      "step": 751
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.042325019836426,
      "learning_rate": 2.7580437580437583e-05,
      "loss": 1.4087,
      "step": 752
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.961600303649902,
      "learning_rate": 2.757722007722008e-05,
      "loss": 1.1971,
      "step": 753
    },
    {
      "epoch": 0.16,
      "grad_norm": 10.114777565002441,
      "learning_rate": 2.7574002574002576e-05,
      "loss": 1.3051,
      "step": 754
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.265595436096191,
      "learning_rate": 2.7570785070785072e-05,
      "loss": 1.3087,
      "step": 755
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.878975868225098,
      "learning_rate": 2.756756756756757e-05,
      "loss": 1.3649,
      "step": 756
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.063404083251953,
      "learning_rate": 2.7564350064350062e-05,
      "loss": 1.4595,
      "step": 757
    },
    {
      "epoch": 0.16,
      "grad_norm": 10.23876667022705,
      "learning_rate": 2.756113256113256e-05,
      "loss": 1.6744,
      "step": 758
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.932001113891602,
      "learning_rate": 2.7557915057915058e-05,
      "loss": 1.61,
      "step": 759
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.92740249633789,
      "learning_rate": 2.7554697554697555e-05,
      "loss": 1.4598,
      "step": 760
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.9121856689453125,
      "learning_rate": 2.755148005148005e-05,
      "loss": 1.3122,
      "step": 761
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.730564594268799,
      "learning_rate": 2.7548262548262547e-05,
      "loss": 1.386,
      "step": 762
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.072959899902344,
      "learning_rate": 2.7545045045045044e-05,
      "loss": 1.5224,
      "step": 763
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.730314254760742,
      "learning_rate": 2.7541827541827544e-05,
      "loss": 1.2957,
      "step": 764
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.709595680236816,
      "learning_rate": 2.753861003861004e-05,
      "loss": 1.1059,
      "step": 765
    },
    {
      "epoch": 0.16,
      "grad_norm": 9.916919708251953,
      "learning_rate": 2.7535392535392537e-05,
      "loss": 1.5147,
      "step": 766
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.449118614196777,
      "learning_rate": 2.7532175032175033e-05,
      "loss": 1.3456,
      "step": 767
    },
    {
      "epoch": 0.16,
      "grad_norm": 10.8137845993042,
      "learning_rate": 2.752895752895753e-05,
      "loss": 1.4295,
      "step": 768
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.599611282348633,
      "learning_rate": 2.7525740025740026e-05,
      "loss": 1.2898,
      "step": 769
    },
    {
      "epoch": 0.17,
      "grad_norm": 11.068127632141113,
      "learning_rate": 2.7522522522522522e-05,
      "loss": 1.7504,
      "step": 770
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.114712715148926,
      "learning_rate": 2.7519305019305022e-05,
      "loss": 1.472,
      "step": 771
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.910568237304688,
      "learning_rate": 2.751608751608752e-05,
      "loss": 1.3248,
      "step": 772
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.457756042480469,
      "learning_rate": 2.7512870012870015e-05,
      "loss": 1.3046,
      "step": 773
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.657246589660645,
      "learning_rate": 2.750965250965251e-05,
      "loss": 1.5181,
      "step": 774
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.89584732055664,
      "learning_rate": 2.7506435006435005e-05,
      "loss": 1.4291,
      "step": 775
    },
    {
      "epoch": 0.17,
      "grad_norm": 11.19368839263916,
      "learning_rate": 2.75032175032175e-05,
      "loss": 1.4325,
      "step": 776
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.782533645629883,
      "learning_rate": 2.75e-05,
      "loss": 1.375,
      "step": 777
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.886565208435059,
      "learning_rate": 2.7496782496782497e-05,
      "loss": 1.3044,
      "step": 778
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.605477333068848,
      "learning_rate": 2.7493564993564994e-05,
      "loss": 1.372,
      "step": 779
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.179624557495117,
      "learning_rate": 2.749034749034749e-05,
      "loss": 1.109,
      "step": 780
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.994572639465332,
      "learning_rate": 2.7487129987129987e-05,
      "loss": 1.457,
      "step": 781
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.480780124664307,
      "learning_rate": 2.7483912483912483e-05,
      "loss": 1.0466,
      "step": 782
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.54865550994873,
      "learning_rate": 2.7480694980694983e-05,
      "loss": 1.6613,
      "step": 783
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.986629486083984,
      "learning_rate": 2.747747747747748e-05,
      "loss": 1.137,
      "step": 784
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.90092945098877,
      "learning_rate": 2.7474259974259976e-05,
      "loss": 1.5942,
      "step": 785
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.398429870605469,
      "learning_rate": 2.7471042471042472e-05,
      "loss": 1.3517,
      "step": 786
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.620941162109375,
      "learning_rate": 2.746782496782497e-05,
      "loss": 1.4047,
      "step": 787
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.439732551574707,
      "learning_rate": 2.7464607464607465e-05,
      "loss": 1.3379,
      "step": 788
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.147941589355469,
      "learning_rate": 2.7461389961389962e-05,
      "loss": 1.3031,
      "step": 789
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.833436012268066,
      "learning_rate": 2.745817245817246e-05,
      "loss": 1.3997,
      "step": 790
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.767797470092773,
      "learning_rate": 2.7454954954954958e-05,
      "loss": 1.2328,
      "step": 791
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.939027786254883,
      "learning_rate": 2.745173745173745e-05,
      "loss": 1.5195,
      "step": 792
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.979229927062988,
      "learning_rate": 2.7448519948519948e-05,
      "loss": 1.3571,
      "step": 793
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.640463829040527,
      "learning_rate": 2.7445302445302444e-05,
      "loss": 1.3502,
      "step": 794
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.665351867675781,
      "learning_rate": 2.744208494208494e-05,
      "loss": 1.0534,
      "step": 795
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.050973892211914,
      "learning_rate": 2.743886743886744e-05,
      "loss": 1.4553,
      "step": 796
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.423609733581543,
      "learning_rate": 2.7435649935649937e-05,
      "loss": 1.551,
      "step": 797
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.641240119934082,
      "learning_rate": 2.7432432432432433e-05,
      "loss": 1.4116,
      "step": 798
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.4407877922058105,
      "learning_rate": 2.742921492921493e-05,
      "loss": 1.4299,
      "step": 799
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.341897964477539,
      "learning_rate": 2.7425997425997426e-05,
      "loss": 1.4648,
      "step": 800
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.135894775390625,
      "learning_rate": 2.7422779922779923e-05,
      "loss": 1.6003,
      "step": 801
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.19509220123291,
      "learning_rate": 2.741956241956242e-05,
      "loss": 1.4662,
      "step": 802
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.582186698913574,
      "learning_rate": 2.741634491634492e-05,
      "loss": 1.3915,
      "step": 803
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.143584251403809,
      "learning_rate": 2.7413127413127415e-05,
      "loss": 1.4382,
      "step": 804
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.834012985229492,
      "learning_rate": 2.7409909909909912e-05,
      "loss": 1.7545,
      "step": 805
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.850911140441895,
      "learning_rate": 2.7406692406692408e-05,
      "loss": 1.5036,
      "step": 806
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.822108268737793,
      "learning_rate": 2.7403474903474905e-05,
      "loss": 1.598,
      "step": 807
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.237154960632324,
      "learning_rate": 2.74002574002574e-05,
      "loss": 1.7664,
      "step": 808
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.1543779373168945,
      "learning_rate": 2.7397039897039898e-05,
      "loss": 1.3405,
      "step": 809
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.613035202026367,
      "learning_rate": 2.7393822393822394e-05,
      "loss": 1.3157,
      "step": 810
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.343815803527832,
      "learning_rate": 2.739060489060489e-05,
      "loss": 1.3729,
      "step": 811
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.536295890808105,
      "learning_rate": 2.7387387387387387e-05,
      "loss": 1.3133,
      "step": 812
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.76827335357666,
      "learning_rate": 2.7384169884169883e-05,
      "loss": 1.5468,
      "step": 813
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.866704940795898,
      "learning_rate": 2.738095238095238e-05,
      "loss": 1.4394,
      "step": 814
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.835330963134766,
      "learning_rate": 2.7377734877734876e-05,
      "loss": 1.5675,
      "step": 815
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.480033874511719,
      "learning_rate": 2.7374517374517376e-05,
      "loss": 1.5163,
      "step": 816
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.108931541442871,
      "learning_rate": 2.7371299871299873e-05,
      "loss": 1.4939,
      "step": 817
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.848288536071777,
      "learning_rate": 2.736808236808237e-05,
      "loss": 1.5703,
      "step": 818
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.377382278442383,
      "learning_rate": 2.7364864864864865e-05,
      "loss": 1.3364,
      "step": 819
    },
    {
      "epoch": 0.18,
      "grad_norm": 11.035350799560547,
      "learning_rate": 2.7361647361647362e-05,
      "loss": 1.7514,
      "step": 820
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.274304389953613,
      "learning_rate": 2.735842985842986e-05,
      "loss": 1.8087,
      "step": 821
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.711259841918945,
      "learning_rate": 2.7355212355212358e-05,
      "loss": 1.2565,
      "step": 822
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.636453628540039,
      "learning_rate": 2.7351994851994855e-05,
      "loss": 1.4335,
      "step": 823
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.326488494873047,
      "learning_rate": 2.734877734877735e-05,
      "loss": 1.3423,
      "step": 824
    },
    {
      "epoch": 0.18,
      "grad_norm": 10.165178298950195,
      "learning_rate": 2.7345559845559848e-05,
      "loss": 1.6724,
      "step": 825
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.506902694702148,
      "learning_rate": 2.7342342342342344e-05,
      "loss": 1.4091,
      "step": 826
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.740561485290527,
      "learning_rate": 2.7339124839124837e-05,
      "loss": 1.2076,
      "step": 827
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.022649765014648,
      "learning_rate": 2.7335907335907337e-05,
      "loss": 1.3425,
      "step": 828
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.911575794219971,
      "learning_rate": 2.7332689832689833e-05,
      "loss": 1.412,
      "step": 829
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.705057144165039,
      "learning_rate": 2.732947232947233e-05,
      "loss": 1.672,
      "step": 830
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.765823364257812,
      "learning_rate": 2.7326254826254826e-05,
      "loss": 1.5232,
      "step": 831
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.146833419799805,
      "learning_rate": 2.7323037323037323e-05,
      "loss": 1.3427,
      "step": 832
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.154597282409668,
      "learning_rate": 2.731981981981982e-05,
      "loss": 1.5215,
      "step": 833
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.419775009155273,
      "learning_rate": 2.7316602316602316e-05,
      "loss": 1.3212,
      "step": 834
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.160785675048828,
      "learning_rate": 2.7313384813384815e-05,
      "loss": 1.4196,
      "step": 835
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.176565170288086,
      "learning_rate": 2.7310167310167312e-05,
      "loss": 1.3796,
      "step": 836
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.875576972961426,
      "learning_rate": 2.730694980694981e-05,
      "loss": 1.4216,
      "step": 837
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.222555160522461,
      "learning_rate": 2.7303732303732305e-05,
      "loss": 1.2527,
      "step": 838
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.834124565124512,
      "learning_rate": 2.73005148005148e-05,
      "loss": 1.2745,
      "step": 839
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.038649559020996,
      "learning_rate": 2.7297297297297298e-05,
      "loss": 1.6343,
      "step": 840
    },
    {
      "epoch": 0.18,
      "grad_norm": 10.49461555480957,
      "learning_rate": 2.7294079794079798e-05,
      "loss": 1.7111,
      "step": 841
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.626550674438477,
      "learning_rate": 2.7290862290862294e-05,
      "loss": 1.4872,
      "step": 842
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.187673568725586,
      "learning_rate": 2.728764478764479e-05,
      "loss": 1.6538,
      "step": 843
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.002527236938477,
      "learning_rate": 2.7284427284427284e-05,
      "loss": 1.5443,
      "step": 844
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.01624870300293,
      "learning_rate": 2.728120978120978e-05,
      "loss": 1.5343,
      "step": 845
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.460214138031006,
      "learning_rate": 2.7277992277992276e-05,
      "loss": 1.2905,
      "step": 846
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.3711576461792,
      "learning_rate": 2.7274774774774773e-05,
      "loss": 1.4156,
      "step": 847
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.998973846435547,
      "learning_rate": 2.7271557271557273e-05,
      "loss": 1.5674,
      "step": 848
    },
    {
      "epoch": 0.18,
      "grad_norm": 6.811801433563232,
      "learning_rate": 2.726833976833977e-05,
      "loss": 1.3063,
      "step": 849
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.846255302429199,
      "learning_rate": 2.7265122265122266e-05,
      "loss": 1.2711,
      "step": 850
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.00965404510498,
      "learning_rate": 2.7261904761904762e-05,
      "loss": 1.1989,
      "step": 851
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.7921781539917,
      "learning_rate": 2.725868725868726e-05,
      "loss": 1.4191,
      "step": 852
    },
    {
      "epoch": 0.18,
      "grad_norm": 10.325292587280273,
      "learning_rate": 2.7255469755469755e-05,
      "loss": 1.5873,
      "step": 853
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.56437873840332,
      "learning_rate": 2.7252252252252255e-05,
      "loss": 1.6934,
      "step": 854
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.841498851776123,
      "learning_rate": 2.724903474903475e-05,
      "loss": 1.2905,
      "step": 855
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.687843322753906,
      "learning_rate": 2.7245817245817248e-05,
      "loss": 0.9585,
      "step": 856
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.384859085083008,
      "learning_rate": 2.7242599742599744e-05,
      "loss": 1.3806,
      "step": 857
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.646141052246094,
      "learning_rate": 2.723938223938224e-05,
      "loss": 1.2521,
      "step": 858
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.9562273025512695,
      "learning_rate": 2.7236164736164737e-05,
      "loss": 1.2435,
      "step": 859
    },
    {
      "epoch": 0.18,
      "grad_norm": 10.843585014343262,
      "learning_rate": 2.7232947232947234e-05,
      "loss": 1.5411,
      "step": 860
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.584829330444336,
      "learning_rate": 2.7229729729729733e-05,
      "loss": 1.1892,
      "step": 861
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.575390815734863,
      "learning_rate": 2.7226512226512226e-05,
      "loss": 1.1263,
      "step": 862
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.370863914489746,
      "learning_rate": 2.7223294723294723e-05,
      "loss": 1.8013,
      "step": 863
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.03342342376709,
      "learning_rate": 2.722007722007722e-05,
      "loss": 1.6294,
      "step": 864
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.585229873657227,
      "learning_rate": 2.7216859716859716e-05,
      "loss": 1.319,
      "step": 865
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.46452808380127,
      "learning_rate": 2.7213642213642212e-05,
      "loss": 1.3818,
      "step": 866
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.424798965454102,
      "learning_rate": 2.7210424710424712e-05,
      "loss": 1.4272,
      "step": 867
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.335932731628418,
      "learning_rate": 2.720720720720721e-05,
      "loss": 1.904,
      "step": 868
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.901905536651611,
      "learning_rate": 2.7203989703989705e-05,
      "loss": 1.2046,
      "step": 869
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.60567569732666,
      "learning_rate": 2.72007722007722e-05,
      "loss": 1.3623,
      "step": 870
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.317278861999512,
      "learning_rate": 2.7197554697554698e-05,
      "loss": 1.492,
      "step": 871
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.627752304077148,
      "learning_rate": 2.7194337194337194e-05,
      "loss": 1.3161,
      "step": 872
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.609947204589844,
      "learning_rate": 2.719111969111969e-05,
      "loss": 1.503,
      "step": 873
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.358907699584961,
      "learning_rate": 2.718790218790219e-05,
      "loss": 1.5222,
      "step": 874
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.690942764282227,
      "learning_rate": 2.7184684684684687e-05,
      "loss": 1.3689,
      "step": 875
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.215863227844238,
      "learning_rate": 2.7181467181467184e-05,
      "loss": 1.3244,
      "step": 876
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.843188285827637,
      "learning_rate": 2.717824967824968e-05,
      "loss": 1.5782,
      "step": 877
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.870086669921875,
      "learning_rate": 2.7175032175032176e-05,
      "loss": 1.2891,
      "step": 878
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.675225257873535,
      "learning_rate": 2.717181467181467e-05,
      "loss": 1.4924,
      "step": 879
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.5280985832214355,
      "learning_rate": 2.716859716859717e-05,
      "loss": 1.3157,
      "step": 880
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.13532543182373,
      "learning_rate": 2.7165379665379666e-05,
      "loss": 1.6584,
      "step": 881
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.611435890197754,
      "learning_rate": 2.7162162162162162e-05,
      "loss": 1.8819,
      "step": 882
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.519933700561523,
      "learning_rate": 2.715894465894466e-05,
      "loss": 1.3316,
      "step": 883
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.34742259979248,
      "learning_rate": 2.7155727155727155e-05,
      "loss": 1.0212,
      "step": 884
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.287184715270996,
      "learning_rate": 2.715250965250965e-05,
      "loss": 1.0928,
      "step": 885
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.342183589935303,
      "learning_rate": 2.714929214929215e-05,
      "loss": 1.1576,
      "step": 886
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.05254077911377,
      "learning_rate": 2.7146074646074648e-05,
      "loss": 1.2305,
      "step": 887
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.205321311950684,
      "learning_rate": 2.7142857142857144e-05,
      "loss": 1.2114,
      "step": 888
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.122885704040527,
      "learning_rate": 2.713963963963964e-05,
      "loss": 2.0145,
      "step": 889
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.162055015563965,
      "learning_rate": 2.7136422136422137e-05,
      "loss": 1.2763,
      "step": 890
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.40585708618164,
      "learning_rate": 2.7133204633204634e-05,
      "loss": 1.2186,
      "step": 891
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.060494422912598,
      "learning_rate": 2.712998712998713e-05,
      "loss": 1.7659,
      "step": 892
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.456525802612305,
      "learning_rate": 2.712676962676963e-05,
      "loss": 1.5337,
      "step": 893
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.428579330444336,
      "learning_rate": 2.7123552123552126e-05,
      "loss": 1.4439,
      "step": 894
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.020888328552246,
      "learning_rate": 2.7120334620334623e-05,
      "loss": 1.3209,
      "step": 895
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.671513557434082,
      "learning_rate": 2.7117117117117116e-05,
      "loss": 1.4294,
      "step": 896
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.72792911529541,
      "learning_rate": 2.7113899613899612e-05,
      "loss": 1.2506,
      "step": 897
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.648786544799805,
      "learning_rate": 2.711068211068211e-05,
      "loss": 1.3557,
      "step": 898
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.333364963531494,
      "learning_rate": 2.710746460746461e-05,
      "loss": 1.1467,
      "step": 899
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.244187355041504,
      "learning_rate": 2.7104247104247105e-05,
      "loss": 1.3754,
      "step": 900
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.958316802978516,
      "learning_rate": 2.71010296010296e-05,
      "loss": 1.2357,
      "step": 901
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.992287635803223,
      "learning_rate": 2.7097812097812098e-05,
      "loss": 1.3448,
      "step": 902
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.695870399475098,
      "learning_rate": 2.7094594594594595e-05,
      "loss": 1.5103,
      "step": 903
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.684875965118408,
      "learning_rate": 2.709137709137709e-05,
      "loss": 1.1969,
      "step": 904
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.185920715332031,
      "learning_rate": 2.7088159588159587e-05,
      "loss": 1.3659,
      "step": 905
    },
    {
      "epoch": 0.19,
      "grad_norm": 6.89382791519165,
      "learning_rate": 2.7084942084942087e-05,
      "loss": 1.2689,
      "step": 906
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.726553916931152,
      "learning_rate": 2.7081724581724584e-05,
      "loss": 1.2155,
      "step": 907
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.337228775024414,
      "learning_rate": 2.707850707850708e-05,
      "loss": 1.0078,
      "step": 908
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.89286994934082,
      "learning_rate": 2.7075289575289577e-05,
      "loss": 1.7347,
      "step": 909
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.222264289855957,
      "learning_rate": 2.7072072072072073e-05,
      "loss": 1.8796,
      "step": 910
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.836405277252197,
      "learning_rate": 2.706885456885457e-05,
      "loss": 1.1762,
      "step": 911
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.192138671875,
      "learning_rate": 2.706563706563707e-05,
      "loss": 1.4895,
      "step": 912
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.937642097473145,
      "learning_rate": 2.7062419562419566e-05,
      "loss": 1.3349,
      "step": 913
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.964935779571533,
      "learning_rate": 2.705920205920206e-05,
      "loss": 0.9448,
      "step": 914
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.847426414489746,
      "learning_rate": 2.7055984555984555e-05,
      "loss": 1.3624,
      "step": 915
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.93464469909668,
      "learning_rate": 2.7052767052767052e-05,
      "loss": 1.5305,
      "step": 916
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.909724235534668,
      "learning_rate": 2.7049549549549548e-05,
      "loss": 1.2178,
      "step": 917
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.759422302246094,
      "learning_rate": 2.7046332046332045e-05,
      "loss": 1.4359,
      "step": 918
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.972433090209961,
      "learning_rate": 2.7043114543114545e-05,
      "loss": 1.542,
      "step": 919
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.726749420166016,
      "learning_rate": 2.703989703989704e-05,
      "loss": 1.0891,
      "step": 920
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.910661697387695,
      "learning_rate": 2.7036679536679537e-05,
      "loss": 1.2135,
      "step": 921
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.526871681213379,
      "learning_rate": 2.7033462033462034e-05,
      "loss": 1.0919,
      "step": 922
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.02984619140625,
      "learning_rate": 2.703024453024453e-05,
      "loss": 1.2363,
      "step": 923
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.84571647644043,
      "learning_rate": 2.7027027027027027e-05,
      "loss": 1.3635,
      "step": 924
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.935730934143066,
      "learning_rate": 2.7023809523809527e-05,
      "loss": 1.7121,
      "step": 925
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.438127517700195,
      "learning_rate": 2.7020592020592023e-05,
      "loss": 1.0913,
      "step": 926
    },
    {
      "epoch": 0.2,
      "grad_norm": 11.372014045715332,
      "learning_rate": 2.701737451737452e-05,
      "loss": 1.8667,
      "step": 927
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.00954818725586,
      "learning_rate": 2.7014157014157016e-05,
      "loss": 1.503,
      "step": 928
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.238213539123535,
      "learning_rate": 2.7010939510939512e-05,
      "loss": 1.4751,
      "step": 929
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.527519226074219,
      "learning_rate": 2.700772200772201e-05,
      "loss": 1.4507,
      "step": 930
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.417492866516113,
      "learning_rate": 2.7004504504504502e-05,
      "loss": 1.3717,
      "step": 931
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.830924034118652,
      "learning_rate": 2.7001287001287002e-05,
      "loss": 1.3278,
      "step": 932
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.093011856079102,
      "learning_rate": 2.6998069498069498e-05,
      "loss": 1.3747,
      "step": 933
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.316437721252441,
      "learning_rate": 2.6994851994851995e-05,
      "loss": 1.4292,
      "step": 934
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.032400131225586,
      "learning_rate": 2.699163449163449e-05,
      "loss": 1.3402,
      "step": 935
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.926802158355713,
      "learning_rate": 2.6988416988416988e-05,
      "loss": 1.2203,
      "step": 936
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.22783088684082,
      "learning_rate": 2.6985199485199484e-05,
      "loss": 1.163,
      "step": 937
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.132551193237305,
      "learning_rate": 2.6981981981981984e-05,
      "loss": 1.4845,
      "step": 938
    },
    {
      "epoch": 0.2,
      "grad_norm": 10.276192665100098,
      "learning_rate": 2.697876447876448e-05,
      "loss": 1.3463,
      "step": 939
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.701696395874023,
      "learning_rate": 2.6975546975546977e-05,
      "loss": 1.2125,
      "step": 940
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.201929092407227,
      "learning_rate": 2.6972329472329473e-05,
      "loss": 1.2449,
      "step": 941
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.044903755187988,
      "learning_rate": 2.696911196911197e-05,
      "loss": 1.1854,
      "step": 942
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.17409610748291,
      "learning_rate": 2.6965894465894466e-05,
      "loss": 1.3385,
      "step": 943
    },
    {
      "epoch": 0.2,
      "grad_norm": 10.300969123840332,
      "learning_rate": 2.6962676962676963e-05,
      "loss": 1.5248,
      "step": 944
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.319809913635254,
      "learning_rate": 2.6959459459459462e-05,
      "loss": 1.4086,
      "step": 945
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.613639831542969,
      "learning_rate": 2.695624195624196e-05,
      "loss": 1.213,
      "step": 946
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.313514709472656,
      "learning_rate": 2.6953024453024455e-05,
      "loss": 1.6048,
      "step": 947
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.436203002929688,
      "learning_rate": 2.6949806949806952e-05,
      "loss": 1.2966,
      "step": 948
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.823955535888672,
      "learning_rate": 2.6946589446589445e-05,
      "loss": 1.2222,
      "step": 949
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.307188034057617,
      "learning_rate": 2.694337194337194e-05,
      "loss": 1.2289,
      "step": 950
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.210882186889648,
      "learning_rate": 2.694015444015444e-05,
      "loss": 1.2326,
      "step": 951
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.544060707092285,
      "learning_rate": 2.6936936936936938e-05,
      "loss": 1.4002,
      "step": 952
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.68779182434082,
      "learning_rate": 2.6933719433719434e-05,
      "loss": 1.1882,
      "step": 953
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.28699016571045,
      "learning_rate": 2.693050193050193e-05,
      "loss": 0.9726,
      "step": 954
    },
    {
      "epoch": 0.2,
      "grad_norm": 10.533101081848145,
      "learning_rate": 2.6927284427284427e-05,
      "loss": 1.7144,
      "step": 955
    },
    {
      "epoch": 0.21,
      "grad_norm": 10.668272018432617,
      "learning_rate": 2.6924066924066923e-05,
      "loss": 1.5713,
      "step": 956
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.808518886566162,
      "learning_rate": 2.6920849420849423e-05,
      "loss": 1.2778,
      "step": 957
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.632819175720215,
      "learning_rate": 2.691763191763192e-05,
      "loss": 1.2743,
      "step": 958
    },
    {
      "epoch": 0.21,
      "grad_norm": 10.325730323791504,
      "learning_rate": 2.6914414414414416e-05,
      "loss": 1.4613,
      "step": 959
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.517739295959473,
      "learning_rate": 2.6911196911196913e-05,
      "loss": 1.1422,
      "step": 960
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.782376289367676,
      "learning_rate": 2.690797940797941e-05,
      "loss": 1.3191,
      "step": 961
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.286404609680176,
      "learning_rate": 2.6904761904761905e-05,
      "loss": 1.3892,
      "step": 962
    },
    {
      "epoch": 0.21,
      "grad_norm": 10.527665138244629,
      "learning_rate": 2.6901544401544402e-05,
      "loss": 1.3877,
      "step": 963
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.213756561279297,
      "learning_rate": 2.6898326898326902e-05,
      "loss": 1.2151,
      "step": 964
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.733354568481445,
      "learning_rate": 2.6895109395109398e-05,
      "loss": 1.4137,
      "step": 965
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.912667274475098,
      "learning_rate": 2.689189189189189e-05,
      "loss": 1.4096,
      "step": 966
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.810029029846191,
      "learning_rate": 2.6888674388674388e-05,
      "loss": 1.1789,
      "step": 967
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.9121222496032715,
      "learning_rate": 2.6885456885456884e-05,
      "loss": 1.0947,
      "step": 968
    },
    {
      "epoch": 0.21,
      "grad_norm": 10.009861946105957,
      "learning_rate": 2.688223938223938e-05,
      "loss": 1.4788,
      "step": 969
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.097514152526855,
      "learning_rate": 2.687902187902188e-05,
      "loss": 1.4548,
      "step": 970
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.6453728675842285,
      "learning_rate": 2.6875804375804377e-05,
      "loss": 0.9209,
      "step": 971
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.510148048400879,
      "learning_rate": 2.6872586872586873e-05,
      "loss": 1.292,
      "step": 972
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.9107842445373535,
      "learning_rate": 2.686936936936937e-05,
      "loss": 1.4326,
      "step": 973
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.138959884643555,
      "learning_rate": 2.6866151866151866e-05,
      "loss": 1.2642,
      "step": 974
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.04025650024414,
      "learning_rate": 2.6862934362934363e-05,
      "loss": 0.9955,
      "step": 975
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.769095420837402,
      "learning_rate": 2.685971685971686e-05,
      "loss": 1.2087,
      "step": 976
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.626959800720215,
      "learning_rate": 2.685649935649936e-05,
      "loss": 1.4719,
      "step": 977
    },
    {
      "epoch": 0.21,
      "grad_norm": 10.681221961975098,
      "learning_rate": 2.6853281853281855e-05,
      "loss": 1.2337,
      "step": 978
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.87985897064209,
      "learning_rate": 2.6850064350064352e-05,
      "loss": 1.546,
      "step": 979
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.394567489624023,
      "learning_rate": 2.684684684684685e-05,
      "loss": 1.627,
      "step": 980
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.00055980682373,
      "learning_rate": 2.6843629343629345e-05,
      "loss": 1.1929,
      "step": 981
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.772244453430176,
      "learning_rate": 2.684041184041184e-05,
      "loss": 1.043,
      "step": 982
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.280366897583008,
      "learning_rate": 2.6837194337194338e-05,
      "loss": 1.2762,
      "step": 983
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.418713569641113,
      "learning_rate": 2.6833976833976834e-05,
      "loss": 1.5182,
      "step": 984
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.170751571655273,
      "learning_rate": 2.683075933075933e-05,
      "loss": 1.512,
      "step": 985
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.923095703125,
      "learning_rate": 2.6827541827541827e-05,
      "loss": 1.2777,
      "step": 986
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.677114486694336,
      "learning_rate": 2.6824324324324324e-05,
      "loss": 1.4216,
      "step": 987
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.5696611404418945,
      "learning_rate": 2.682110682110682e-05,
      "loss": 1.3699,
      "step": 988
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.132107734680176,
      "learning_rate": 2.6817889317889316e-05,
      "loss": 1.0521,
      "step": 989
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.6028547286987305,
      "learning_rate": 2.6814671814671816e-05,
      "loss": 0.963,
      "step": 990
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.90681266784668,
      "learning_rate": 2.6811454311454313e-05,
      "loss": 1.3335,
      "step": 991
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.90875768661499,
      "learning_rate": 2.680823680823681e-05,
      "loss": 1.3103,
      "step": 992
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.071565628051758,
      "learning_rate": 2.6805019305019306e-05,
      "loss": 1.1042,
      "step": 993
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.270937442779541,
      "learning_rate": 2.6801801801801802e-05,
      "loss": 1.1097,
      "step": 994
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.903903007507324,
      "learning_rate": 2.67985842985843e-05,
      "loss": 1.5762,
      "step": 995
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.983363628387451,
      "learning_rate": 2.67953667953668e-05,
      "loss": 0.9953,
      "step": 996
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.218119621276855,
      "learning_rate": 2.6792149292149295e-05,
      "loss": 1.2224,
      "step": 997
    },
    {
      "epoch": 0.21,
      "grad_norm": 11.211015701293945,
      "learning_rate": 2.678893178893179e-05,
      "loss": 1.5773,
      "step": 998
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.094499588012695,
      "learning_rate": 2.6785714285714288e-05,
      "loss": 1.5915,
      "step": 999
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.163897514343262,
      "learning_rate": 2.6782496782496784e-05,
      "loss": 0.9963,
      "step": 1000
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.296417236328125,
      "learning_rate": 2.6779279279279277e-05,
      "loss": 1.177,
      "step": 1001
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.79904317855835,
      "learning_rate": 2.6776061776061777e-05,
      "loss": 0.9994,
      "step": 1002
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.970348358154297,
      "learning_rate": 2.6772844272844274e-05,
      "loss": 1.1977,
      "step": 1003
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.68642520904541,
      "learning_rate": 2.676962676962677e-05,
      "loss": 1.3641,
      "step": 1004
    },
    {
      "epoch": 0.22,
      "grad_norm": 10.213547706604004,
      "learning_rate": 2.6766409266409266e-05,
      "loss": 1.6114,
      "step": 1005
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.346342086791992,
      "learning_rate": 2.6763191763191763e-05,
      "loss": 1.1135,
      "step": 1006
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.307868003845215,
      "learning_rate": 2.675997425997426e-05,
      "loss": 1.1048,
      "step": 1007
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.48186206817627,
      "learning_rate": 2.6756756756756756e-05,
      "loss": 1.337,
      "step": 1008
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.853020668029785,
      "learning_rate": 2.6753539253539256e-05,
      "loss": 1.3584,
      "step": 1009
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.044156074523926,
      "learning_rate": 2.6750321750321752e-05,
      "loss": 1.2208,
      "step": 1010
    },
    {
      "epoch": 0.22,
      "grad_norm": 11.071755409240723,
      "learning_rate": 2.674710424710425e-05,
      "loss": 1.1718,
      "step": 1011
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.361820220947266,
      "learning_rate": 2.6743886743886745e-05,
      "loss": 1.2949,
      "step": 1012
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.433032035827637,
      "learning_rate": 2.674066924066924e-05,
      "loss": 1.4146,
      "step": 1013
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.287514686584473,
      "learning_rate": 2.6737451737451738e-05,
      "loss": 1.5002,
      "step": 1014
    },
    {
      "epoch": 0.22,
      "grad_norm": 10.165128707885742,
      "learning_rate": 2.6734234234234238e-05,
      "loss": 1.4242,
      "step": 1015
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.535551071166992,
      "learning_rate": 2.6731016731016734e-05,
      "loss": 1.3203,
      "step": 1016
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.723125457763672,
      "learning_rate": 2.672779922779923e-05,
      "loss": 1.2042,
      "step": 1017
    },
    {
      "epoch": 0.22,
      "grad_norm": 10.163501739501953,
      "learning_rate": 2.6724581724581724e-05,
      "loss": 1.5948,
      "step": 1018
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.887606620788574,
      "learning_rate": 2.672136422136422e-05,
      "loss": 1.2335,
      "step": 1019
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.800739288330078,
      "learning_rate": 2.6718146718146717e-05,
      "loss": 1.0272,
      "step": 1020
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.523853302001953,
      "learning_rate": 2.6714929214929213e-05,
      "loss": 1.4476,
      "step": 1021
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.438216209411621,
      "learning_rate": 2.6711711711711713e-05,
      "loss": 1.7057,
      "step": 1022
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.118048667907715,
      "learning_rate": 2.670849420849421e-05,
      "loss": 1.3541,
      "step": 1023
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.238831520080566,
      "learning_rate": 2.6705276705276706e-05,
      "loss": 1.1536,
      "step": 1024
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.759914875030518,
      "learning_rate": 2.6702059202059202e-05,
      "loss": 0.9827,
      "step": 1025
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.093073844909668,
      "learning_rate": 2.66988416988417e-05,
      "loss": 1.4416,
      "step": 1026
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.883118629455566,
      "learning_rate": 2.6695624195624195e-05,
      "loss": 1.2647,
      "step": 1027
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.615971565246582,
      "learning_rate": 2.6692406692406695e-05,
      "loss": 1.4062,
      "step": 1028
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.5187406539917,
      "learning_rate": 2.668918918918919e-05,
      "loss": 1.2005,
      "step": 1029
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.178502082824707,
      "learning_rate": 2.6685971685971688e-05,
      "loss": 1.3581,
      "step": 1030
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.750381469726562,
      "learning_rate": 2.6682754182754184e-05,
      "loss": 1.5399,
      "step": 1031
    },
    {
      "epoch": 0.22,
      "grad_norm": 10.327668190002441,
      "learning_rate": 2.667953667953668e-05,
      "loss": 1.6001,
      "step": 1032
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.34179973602295,
      "learning_rate": 2.6676319176319177e-05,
      "loss": 1.3243,
      "step": 1033
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.392219543457031,
      "learning_rate": 2.6673101673101674e-05,
      "loss": 1.3734,
      "step": 1034
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.734882354736328,
      "learning_rate": 2.6669884169884174e-05,
      "loss": 1.3896,
      "step": 1035
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.471430778503418,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 1.2785,
      "step": 1036
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.149492263793945,
      "learning_rate": 2.6663449163449163e-05,
      "loss": 1.2184,
      "step": 1037
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.093111038208008,
      "learning_rate": 2.666023166023166e-05,
      "loss": 1.2507,
      "step": 1038
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.219695091247559,
      "learning_rate": 2.6657014157014156e-05,
      "loss": 1.3093,
      "step": 1039
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.4314546585083,
      "learning_rate": 2.6653796653796652e-05,
      "loss": 1.5434,
      "step": 1040
    },
    {
      "epoch": 0.22,
      "grad_norm": 10.203361511230469,
      "learning_rate": 2.6650579150579152e-05,
      "loss": 1.4872,
      "step": 1041
    },
    {
      "epoch": 0.22,
      "grad_norm": 10.272558212280273,
      "learning_rate": 2.664736164736165e-05,
      "loss": 1.288,
      "step": 1042
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.470454692840576,
      "learning_rate": 2.6644144144144145e-05,
      "loss": 1.4537,
      "step": 1043
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.526825904846191,
      "learning_rate": 2.664092664092664e-05,
      "loss": 1.1086,
      "step": 1044
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.229141235351562,
      "learning_rate": 2.6637709137709138e-05,
      "loss": 1.4565,
      "step": 1045
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.04105281829834,
      "learning_rate": 2.6634491634491634e-05,
      "loss": 1.3318,
      "step": 1046
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.247039794921875,
      "learning_rate": 2.663127413127413e-05,
      "loss": 1.353,
      "step": 1047
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.059009552001953,
      "learning_rate": 2.662805662805663e-05,
      "loss": 1.2128,
      "step": 1048
    },
    {
      "epoch": 0.23,
      "grad_norm": 10.065364837646484,
      "learning_rate": 2.6624839124839127e-05,
      "loss": 1.294,
      "step": 1049
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.009159088134766,
      "learning_rate": 2.6621621621621624e-05,
      "loss": 1.4246,
      "step": 1050
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.0078706741333,
      "learning_rate": 2.661840411840412e-05,
      "loss": 1.1257,
      "step": 1051
    },
    {
      "epoch": 0.23,
      "grad_norm": 11.032622337341309,
      "learning_rate": 2.6615186615186617e-05,
      "loss": 1.6817,
      "step": 1052
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.848816871643066,
      "learning_rate": 2.661196911196911e-05,
      "loss": 1.3848,
      "step": 1053
    },
    {
      "epoch": 0.23,
      "grad_norm": 13.389190673828125,
      "learning_rate": 2.660875160875161e-05,
      "loss": 1.7191,
      "step": 1054
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.900801181793213,
      "learning_rate": 2.6605534105534106e-05,
      "loss": 1.0885,
      "step": 1055
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.046304702758789,
      "learning_rate": 2.6602316602316602e-05,
      "loss": 1.2356,
      "step": 1056
    },
    {
      "epoch": 0.23,
      "grad_norm": 10.206510543823242,
      "learning_rate": 2.65990990990991e-05,
      "loss": 1.305,
      "step": 1057
    },
    {
      "epoch": 0.23,
      "grad_norm": 11.328250885009766,
      "learning_rate": 2.6595881595881595e-05,
      "loss": 1.2975,
      "step": 1058
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.773786544799805,
      "learning_rate": 2.6592664092664092e-05,
      "loss": 1.5656,
      "step": 1059
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.97162914276123,
      "learning_rate": 2.658944658944659e-05,
      "loss": 1.2013,
      "step": 1060
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.0463228225708,
      "learning_rate": 2.6586229086229088e-05,
      "loss": 1.4914,
      "step": 1061
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.645218849182129,
      "learning_rate": 2.6583011583011584e-05,
      "loss": 1.2969,
      "step": 1062
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.921472549438477,
      "learning_rate": 2.657979407979408e-05,
      "loss": 1.2222,
      "step": 1063
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.707209587097168,
      "learning_rate": 2.6576576576576577e-05,
      "loss": 1.4329,
      "step": 1064
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.4392571449279785,
      "learning_rate": 2.6573359073359074e-05,
      "loss": 0.9083,
      "step": 1065
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.004874229431152,
      "learning_rate": 2.657014157014157e-05,
      "loss": 1.0682,
      "step": 1066
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.57685661315918,
      "learning_rate": 2.656692406692407e-05,
      "loss": 1.1779,
      "step": 1067
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.606524467468262,
      "learning_rate": 2.6563706563706567e-05,
      "loss": 1.789,
      "step": 1068
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.458276748657227,
      "learning_rate": 2.6560489060489063e-05,
      "loss": 1.5556,
      "step": 1069
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.95613431930542,
      "learning_rate": 2.6557271557271556e-05,
      "loss": 1.0827,
      "step": 1070
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.173651695251465,
      "learning_rate": 2.6554054054054053e-05,
      "loss": 1.3256,
      "step": 1071
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.025677680969238,
      "learning_rate": 2.655083655083655e-05,
      "loss": 1.2592,
      "step": 1072
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.520923614501953,
      "learning_rate": 2.654761904761905e-05,
      "loss": 1.3684,
      "step": 1073
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.092731475830078,
      "learning_rate": 2.6544401544401545e-05,
      "loss": 1.1703,
      "step": 1074
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.970864772796631,
      "learning_rate": 2.6541184041184042e-05,
      "loss": 1.1207,
      "step": 1075
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.329585075378418,
      "learning_rate": 2.6537966537966538e-05,
      "loss": 1.5484,
      "step": 1076
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.526788711547852,
      "learning_rate": 2.6534749034749035e-05,
      "loss": 1.4333,
      "step": 1077
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.537067413330078,
      "learning_rate": 2.653153153153153e-05,
      "loss": 1.5089,
      "step": 1078
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.781491279602051,
      "learning_rate": 2.6528314028314028e-05,
      "loss": 1.0938,
      "step": 1079
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.755642414093018,
      "learning_rate": 2.6525096525096527e-05,
      "loss": 1.2534,
      "step": 1080
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.546951293945312,
      "learning_rate": 2.6521879021879024e-05,
      "loss": 1.2879,
      "step": 1081
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.536312103271484,
      "learning_rate": 2.651866151866152e-05,
      "loss": 1.1072,
      "step": 1082
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.665640830993652,
      "learning_rate": 2.6515444015444017e-05,
      "loss": 1.1929,
      "step": 1083
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.463379859924316,
      "learning_rate": 2.6512226512226513e-05,
      "loss": 1.1957,
      "step": 1084
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.914044380187988,
      "learning_rate": 2.650900900900901e-05,
      "loss": 1.2338,
      "step": 1085
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.525028228759766,
      "learning_rate": 2.650579150579151e-05,
      "loss": 1.3437,
      "step": 1086
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.227401733398438,
      "learning_rate": 2.6502574002574006e-05,
      "loss": 1.4023,
      "step": 1087
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.3428850173950195,
      "learning_rate": 2.64993564993565e-05,
      "loss": 1.1411,
      "step": 1088
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.231085777282715,
      "learning_rate": 2.6496138996138995e-05,
      "loss": 1.3599,
      "step": 1089
    },
    {
      "epoch": 0.23,
      "grad_norm": 10.766886711120605,
      "learning_rate": 2.6492921492921492e-05,
      "loss": 1.3924,
      "step": 1090
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.735498428344727,
      "learning_rate": 2.648970398970399e-05,
      "loss": 1.2116,
      "step": 1091
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.597352981567383,
      "learning_rate": 2.6486486486486485e-05,
      "loss": 1.5279,
      "step": 1092
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.924059867858887,
      "learning_rate": 2.6483268983268985e-05,
      "loss": 1.2261,
      "step": 1093
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.869619369506836,
      "learning_rate": 2.648005148005148e-05,
      "loss": 1.2184,
      "step": 1094
    },
    {
      "epoch": 0.23,
      "grad_norm": 8.857922554016113,
      "learning_rate": 2.6476833976833978e-05,
      "loss": 1.4109,
      "step": 1095
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.121352195739746,
      "learning_rate": 2.6473616473616474e-05,
      "loss": 1.2362,
      "step": 1096
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.027889251708984,
      "learning_rate": 2.647039897039897e-05,
      "loss": 1.5489,
      "step": 1097
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.790159225463867,
      "learning_rate": 2.6467181467181467e-05,
      "loss": 1.3526,
      "step": 1098
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.838161468505859,
      "learning_rate": 2.6463963963963967e-05,
      "loss": 1.3441,
      "step": 1099
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.544195652008057,
      "learning_rate": 2.6460746460746463e-05,
      "loss": 0.9249,
      "step": 1100
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.423432350158691,
      "learning_rate": 2.645752895752896e-05,
      "loss": 1.3488,
      "step": 1101
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.217277526855469,
      "learning_rate": 2.6454311454311456e-05,
      "loss": 1.325,
      "step": 1102
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.931084632873535,
      "learning_rate": 2.6451093951093953e-05,
      "loss": 1.1408,
      "step": 1103
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.002298355102539,
      "learning_rate": 2.644787644787645e-05,
      "loss": 1.4584,
      "step": 1104
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.0680513381958,
      "learning_rate": 2.6444658944658942e-05,
      "loss": 1.12,
      "step": 1105
    },
    {
      "epoch": 0.24,
      "grad_norm": 11.070405006408691,
      "learning_rate": 2.6441441441441442e-05,
      "loss": 2.0573,
      "step": 1106
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.508176803588867,
      "learning_rate": 2.643822393822394e-05,
      "loss": 1.5815,
      "step": 1107
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.331486701965332,
      "learning_rate": 2.6435006435006435e-05,
      "loss": 1.5432,
      "step": 1108
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.48970890045166,
      "learning_rate": 2.643178893178893e-05,
      "loss": 1.5498,
      "step": 1109
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.446732521057129,
      "learning_rate": 2.6428571428571428e-05,
      "loss": 1.524,
      "step": 1110
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.004000663757324,
      "learning_rate": 2.6425353925353924e-05,
      "loss": 1.1836,
      "step": 1111
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.877951622009277,
      "learning_rate": 2.6422136422136424e-05,
      "loss": 1.3849,
      "step": 1112
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.487264633178711,
      "learning_rate": 2.641891891891892e-05,
      "loss": 1.1684,
      "step": 1113
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.139583110809326,
      "learning_rate": 2.6415701415701417e-05,
      "loss": 1.1383,
      "step": 1114
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.832189559936523,
      "learning_rate": 2.6412483912483913e-05,
      "loss": 1.3166,
      "step": 1115
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.387091636657715,
      "learning_rate": 2.640926640926641e-05,
      "loss": 1.1983,
      "step": 1116
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.953089237213135,
      "learning_rate": 2.6406048906048906e-05,
      "loss": 1.151,
      "step": 1117
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.812167167663574,
      "learning_rate": 2.6402831402831403e-05,
      "loss": 1.2357,
      "step": 1118
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.864689826965332,
      "learning_rate": 2.6399613899613903e-05,
      "loss": 1.3526,
      "step": 1119
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.312637329101562,
      "learning_rate": 2.63963963963964e-05,
      "loss": 1.2572,
      "step": 1120
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.614030838012695,
      "learning_rate": 2.6393178893178895e-05,
      "loss": 1.3755,
      "step": 1121
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.797920227050781,
      "learning_rate": 2.6389961389961392e-05,
      "loss": 1.4095,
      "step": 1122
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.662358283996582,
      "learning_rate": 2.6386743886743885e-05,
      "loss": 1.6142,
      "step": 1123
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.250812530517578,
      "learning_rate": 2.638352638352638e-05,
      "loss": 1.282,
      "step": 1124
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.856569290161133,
      "learning_rate": 2.638030888030888e-05,
      "loss": 1.3338,
      "step": 1125
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.443182945251465,
      "learning_rate": 2.6377091377091378e-05,
      "loss": 1.4442,
      "step": 1126
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.735374450683594,
      "learning_rate": 2.6373873873873874e-05,
      "loss": 1.473,
      "step": 1127
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.629547119140625,
      "learning_rate": 2.637065637065637e-05,
      "loss": 1.2654,
      "step": 1128
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.158059120178223,
      "learning_rate": 2.6367438867438867e-05,
      "loss": 1.3782,
      "step": 1129
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.338692665100098,
      "learning_rate": 2.6364221364221364e-05,
      "loss": 1.1353,
      "step": 1130
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.097105979919434,
      "learning_rate": 2.6361003861003863e-05,
      "loss": 1.2348,
      "step": 1131
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.352474212646484,
      "learning_rate": 2.635778635778636e-05,
      "loss": 1.3517,
      "step": 1132
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.425739288330078,
      "learning_rate": 2.6354568854568856e-05,
      "loss": 1.2264,
      "step": 1133
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.505402565002441,
      "learning_rate": 2.6351351351351353e-05,
      "loss": 1.3575,
      "step": 1134
    },
    {
      "epoch": 0.24,
      "grad_norm": 11.18097972869873,
      "learning_rate": 2.634813384813385e-05,
      "loss": 1.1585,
      "step": 1135
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.383801460266113,
      "learning_rate": 2.6344916344916346e-05,
      "loss": 1.4602,
      "step": 1136
    },
    {
      "epoch": 0.24,
      "grad_norm": 9.342517852783203,
      "learning_rate": 2.6341698841698842e-05,
      "loss": 1.4744,
      "step": 1137
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.4425225257873535,
      "learning_rate": 2.6338481338481342e-05,
      "loss": 1.3005,
      "step": 1138
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.62492847442627,
      "learning_rate": 2.633526383526384e-05,
      "loss": 1.2244,
      "step": 1139
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.75383472442627,
      "learning_rate": 2.633204633204633e-05,
      "loss": 1.5252,
      "step": 1140
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.953263282775879,
      "learning_rate": 2.6328828828828828e-05,
      "loss": 1.3204,
      "step": 1141
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.552252769470215,
      "learning_rate": 2.6325611325611324e-05,
      "loss": 1.3497,
      "step": 1142
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.320897102355957,
      "learning_rate": 2.632239382239382e-05,
      "loss": 1.1277,
      "step": 1143
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.7547407150268555,
      "learning_rate": 2.631917631917632e-05,
      "loss": 1.4977,
      "step": 1144
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.075538635253906,
      "learning_rate": 2.6315958815958817e-05,
      "loss": 1.8237,
      "step": 1145
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.26809024810791,
      "learning_rate": 2.6312741312741314e-05,
      "loss": 1.4615,
      "step": 1146
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.586714744567871,
      "learning_rate": 2.630952380952381e-05,
      "loss": 1.4277,
      "step": 1147
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.913267135620117,
      "learning_rate": 2.6306306306306306e-05,
      "loss": 1.5179,
      "step": 1148
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.401288986206055,
      "learning_rate": 2.6303088803088803e-05,
      "loss": 1.1279,
      "step": 1149
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.570776462554932,
      "learning_rate": 2.62998712998713e-05,
      "loss": 0.9574,
      "step": 1150
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.513676643371582,
      "learning_rate": 2.62966537966538e-05,
      "loss": 1.3596,
      "step": 1151
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.385220527648926,
      "learning_rate": 2.6293436293436296e-05,
      "loss": 1.1931,
      "step": 1152
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.01206111907959,
      "learning_rate": 2.6290218790218792e-05,
      "loss": 1.1842,
      "step": 1153
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.580415725708008,
      "learning_rate": 2.628700128700129e-05,
      "loss": 1.1173,
      "step": 1154
    },
    {
      "epoch": 0.25,
      "grad_norm": 11.332547187805176,
      "learning_rate": 2.6283783783783785e-05,
      "loss": 1.0871,
      "step": 1155
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.625669479370117,
      "learning_rate": 2.628056628056628e-05,
      "loss": 1.4341,
      "step": 1156
    },
    {
      "epoch": 0.25,
      "grad_norm": 10.331356048583984,
      "learning_rate": 2.6277348777348778e-05,
      "loss": 1.3352,
      "step": 1157
    },
    {
      "epoch": 0.25,
      "grad_norm": 10.88631534576416,
      "learning_rate": 2.6274131274131274e-05,
      "loss": 1.4785,
      "step": 1158
    },
    {
      "epoch": 0.25,
      "grad_norm": 10.449967384338379,
      "learning_rate": 2.627091377091377e-05,
      "loss": 1.3803,
      "step": 1159
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.10886287689209,
      "learning_rate": 2.6267696267696267e-05,
      "loss": 1.1967,
      "step": 1160
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.872442245483398,
      "learning_rate": 2.6264478764478764e-05,
      "loss": 1.3729,
      "step": 1161
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.16933822631836,
      "learning_rate": 2.626126126126126e-05,
      "loss": 1.3482,
      "step": 1162
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.577134132385254,
      "learning_rate": 2.6258043758043757e-05,
      "loss": 1.2113,
      "step": 1163
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.948328971862793,
      "learning_rate": 2.6254826254826256e-05,
      "loss": 1.2364,
      "step": 1164
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.852004051208496,
      "learning_rate": 2.6251608751608753e-05,
      "loss": 1.1161,
      "step": 1165
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.488945007324219,
      "learning_rate": 2.624839124839125e-05,
      "loss": 1.2805,
      "step": 1166
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.707749366760254,
      "learning_rate": 2.6245173745173746e-05,
      "loss": 1.1616,
      "step": 1167
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.235556602478027,
      "learning_rate": 2.6241956241956242e-05,
      "loss": 1.2634,
      "step": 1168
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.960226058959961,
      "learning_rate": 2.623873873873874e-05,
      "loss": 1.3366,
      "step": 1169
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.923757553100586,
      "learning_rate": 2.623552123552124e-05,
      "loss": 1.2126,
      "step": 1170
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.007299423217773,
      "learning_rate": 2.6232303732303735e-05,
      "loss": 1.2641,
      "step": 1171
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.294140815734863,
      "learning_rate": 2.622908622908623e-05,
      "loss": 1.2246,
      "step": 1172
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.941680908203125,
      "learning_rate": 2.6225868725868728e-05,
      "loss": 1.081,
      "step": 1173
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.710986137390137,
      "learning_rate": 2.6222651222651224e-05,
      "loss": 1.012,
      "step": 1174
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.735367774963379,
      "learning_rate": 2.6219433719433717e-05,
      "loss": 1.2809,
      "step": 1175
    },
    {
      "epoch": 0.25,
      "grad_norm": 10.620098114013672,
      "learning_rate": 2.6216216216216217e-05,
      "loss": 1.7218,
      "step": 1176
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.93567419052124,
      "learning_rate": 2.6212998712998714e-05,
      "loss": 0.9242,
      "step": 1177
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.094076156616211,
      "learning_rate": 2.620978120978121e-05,
      "loss": 1.2813,
      "step": 1178
    },
    {
      "epoch": 0.25,
      "grad_norm": 10.433646202087402,
      "learning_rate": 2.6206563706563707e-05,
      "loss": 1.5642,
      "step": 1179
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.91644287109375,
      "learning_rate": 2.6203346203346203e-05,
      "loss": 1.5291,
      "step": 1180
    },
    {
      "epoch": 0.25,
      "grad_norm": 10.789258003234863,
      "learning_rate": 2.62001287001287e-05,
      "loss": 1.3901,
      "step": 1181
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.958864212036133,
      "learning_rate": 2.6196911196911196e-05,
      "loss": 1.6626,
      "step": 1182
    },
    {
      "epoch": 0.25,
      "grad_norm": 10.471342086791992,
      "learning_rate": 2.6193693693693696e-05,
      "loss": 1.5147,
      "step": 1183
    },
    {
      "epoch": 0.25,
      "grad_norm": 11.17542552947998,
      "learning_rate": 2.6190476190476192e-05,
      "loss": 1.505,
      "step": 1184
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.691872596740723,
      "learning_rate": 2.618725868725869e-05,
      "loss": 1.2491,
      "step": 1185
    },
    {
      "epoch": 0.25,
      "grad_norm": 8.128683090209961,
      "learning_rate": 2.6184041184041185e-05,
      "loss": 1.1716,
      "step": 1186
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.591935157775879,
      "learning_rate": 2.618082368082368e-05,
      "loss": 1.1667,
      "step": 1187
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.920924663543701,
      "learning_rate": 2.6177606177606178e-05,
      "loss": 1.16,
      "step": 1188
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.305174827575684,
      "learning_rate": 2.6174388674388678e-05,
      "loss": 1.3298,
      "step": 1189
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.209298133850098,
      "learning_rate": 2.6171171171171174e-05,
      "loss": 1.1447,
      "step": 1190
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.837162971496582,
      "learning_rate": 2.616795366795367e-05,
      "loss": 1.6497,
      "step": 1191
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.15080738067627,
      "learning_rate": 2.6164736164736164e-05,
      "loss": 1.0422,
      "step": 1192
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.074406623840332,
      "learning_rate": 2.616151866151866e-05,
      "loss": 1.1151,
      "step": 1193
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.424674034118652,
      "learning_rate": 2.6158301158301157e-05,
      "loss": 1.1363,
      "step": 1194
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.299229621887207,
      "learning_rate": 2.6155083655083653e-05,
      "loss": 1.2988,
      "step": 1195
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.211856842041016,
      "learning_rate": 2.6151866151866153e-05,
      "loss": 1.3023,
      "step": 1196
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.194677352905273,
      "learning_rate": 2.614864864864865e-05,
      "loss": 0.9834,
      "step": 1197
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.42780876159668,
      "learning_rate": 2.6145431145431146e-05,
      "loss": 1.2888,
      "step": 1198
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.67248821258545,
      "learning_rate": 2.6142213642213642e-05,
      "loss": 1.3899,
      "step": 1199
    },
    {
      "epoch": 0.26,
      "grad_norm": 12.761173248291016,
      "learning_rate": 2.613899613899614e-05,
      "loss": 1.6939,
      "step": 1200
    },
    {
      "epoch": 0.26,
      "grad_norm": 12.108636856079102,
      "learning_rate": 2.6135778635778635e-05,
      "loss": 1.8606,
      "step": 1201
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.887981414794922,
      "learning_rate": 2.6132561132561135e-05,
      "loss": 1.1253,
      "step": 1202
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.52657413482666,
      "learning_rate": 2.612934362934363e-05,
      "loss": 1.142,
      "step": 1203
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.206074714660645,
      "learning_rate": 2.6126126126126128e-05,
      "loss": 1.5408,
      "step": 1204
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.529194831848145,
      "learning_rate": 2.6122908622908624e-05,
      "loss": 1.0568,
      "step": 1205
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.243480682373047,
      "learning_rate": 2.611969111969112e-05,
      "loss": 1.6113,
      "step": 1206
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.703958988189697,
      "learning_rate": 2.6116473616473617e-05,
      "loss": 0.8855,
      "step": 1207
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.736154556274414,
      "learning_rate": 2.6113256113256114e-05,
      "loss": 1.3865,
      "step": 1208
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.558718681335449,
      "learning_rate": 2.611003861003861e-05,
      "loss": 1.1857,
      "step": 1209
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.214512825012207,
      "learning_rate": 2.6106821106821107e-05,
      "loss": 1.1296,
      "step": 1210
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.896367073059082,
      "learning_rate": 2.6103603603603603e-05,
      "loss": 1.2396,
      "step": 1211
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.70589542388916,
      "learning_rate": 2.61003861003861e-05,
      "loss": 1.3799,
      "step": 1212
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.423079490661621,
      "learning_rate": 2.6097168597168596e-05,
      "loss": 1.2965,
      "step": 1213
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.603525161743164,
      "learning_rate": 2.6093951093951093e-05,
      "loss": 1.2948,
      "step": 1214
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.694169044494629,
      "learning_rate": 2.6090733590733592e-05,
      "loss": 1.3485,
      "step": 1215
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.481608390808105,
      "learning_rate": 2.608751608751609e-05,
      "loss": 1.2884,
      "step": 1216
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.273826599121094,
      "learning_rate": 2.6084298584298585e-05,
      "loss": 1.2601,
      "step": 1217
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.030174255371094,
      "learning_rate": 2.6081081081081082e-05,
      "loss": 1.2106,
      "step": 1218
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.081335067749023,
      "learning_rate": 2.6077863577863578e-05,
      "loss": 1.498,
      "step": 1219
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.981876373291016,
      "learning_rate": 2.6074646074646075e-05,
      "loss": 1.5211,
      "step": 1220
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.152267456054688,
      "learning_rate": 2.607142857142857e-05,
      "loss": 1.2255,
      "step": 1221
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.144830703735352,
      "learning_rate": 2.606821106821107e-05,
      "loss": 1.0676,
      "step": 1222
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.868152618408203,
      "learning_rate": 2.6064993564993567e-05,
      "loss": 1.3594,
      "step": 1223
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.686141014099121,
      "learning_rate": 2.6061776061776064e-05,
      "loss": 1.5379,
      "step": 1224
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.632453441619873,
      "learning_rate": 2.605855855855856e-05,
      "loss": 1.3179,
      "step": 1225
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.581647872924805,
      "learning_rate": 2.6055341055341057e-05,
      "loss": 1.3846,
      "step": 1226
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.593341827392578,
      "learning_rate": 2.605212355212355e-05,
      "loss": 1.1424,
      "step": 1227
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.703335762023926,
      "learning_rate": 2.604890604890605e-05,
      "loss": 1.788,
      "step": 1228
    },
    {
      "epoch": 0.26,
      "grad_norm": 8.304698944091797,
      "learning_rate": 2.6045688545688546e-05,
      "loss": 1.2676,
      "step": 1229
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.783912658691406,
      "learning_rate": 2.6042471042471043e-05,
      "loss": 1.027,
      "step": 1230
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.390741348266602,
      "learning_rate": 2.603925353925354e-05,
      "loss": 1.3866,
      "step": 1231
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.9050421714782715,
      "learning_rate": 2.6036036036036035e-05,
      "loss": 1.5989,
      "step": 1232
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.12626838684082,
      "learning_rate": 2.6032818532818532e-05,
      "loss": 1.2975,
      "step": 1233
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.00925064086914,
      "learning_rate": 2.6029601029601032e-05,
      "loss": 1.3594,
      "step": 1234
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.609435081481934,
      "learning_rate": 2.6026383526383528e-05,
      "loss": 1.4396,
      "step": 1235
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.34754467010498,
      "learning_rate": 2.6023166023166025e-05,
      "loss": 1.6008,
      "step": 1236
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.79605770111084,
      "learning_rate": 2.601994851994852e-05,
      "loss": 1.0919,
      "step": 1237
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.013272285461426,
      "learning_rate": 2.6016731016731018e-05,
      "loss": 1.4217,
      "step": 1238
    },
    {
      "epoch": 0.27,
      "grad_norm": 11.417902946472168,
      "learning_rate": 2.6013513513513514e-05,
      "loss": 1.9768,
      "step": 1239
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.658112525939941,
      "learning_rate": 2.601029601029601e-05,
      "loss": 1.1342,
      "step": 1240
    },
    {
      "epoch": 0.27,
      "grad_norm": 10.448524475097656,
      "learning_rate": 2.600707850707851e-05,
      "loss": 1.3345,
      "step": 1241
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.735007286071777,
      "learning_rate": 2.6003861003861007e-05,
      "loss": 1.2715,
      "step": 1242
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.102211952209473,
      "learning_rate": 2.6000643500643503e-05,
      "loss": 1.1217,
      "step": 1243
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.079392433166504,
      "learning_rate": 2.5997425997425996e-05,
      "loss": 1.3346,
      "step": 1244
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.45957088470459,
      "learning_rate": 2.5994208494208493e-05,
      "loss": 1.2231,
      "step": 1245
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.324480056762695,
      "learning_rate": 2.599099099099099e-05,
      "loss": 1.2464,
      "step": 1246
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.93062686920166,
      "learning_rate": 2.598777348777349e-05,
      "loss": 1.0511,
      "step": 1247
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.953044891357422,
      "learning_rate": 2.5984555984555985e-05,
      "loss": 1.3526,
      "step": 1248
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.740547180175781,
      "learning_rate": 2.5981338481338482e-05,
      "loss": 1.3345,
      "step": 1249
    },
    {
      "epoch": 0.27,
      "grad_norm": 11.244915008544922,
      "learning_rate": 2.597812097812098e-05,
      "loss": 1.4018,
      "step": 1250
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.617172241210938,
      "learning_rate": 2.5974903474903475e-05,
      "loss": 1.187,
      "step": 1251
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.445706844329834,
      "learning_rate": 2.597168597168597e-05,
      "loss": 1.1616,
      "step": 1252
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.871638298034668,
      "learning_rate": 2.5968468468468468e-05,
      "loss": 1.2955,
      "step": 1253
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.971224784851074,
      "learning_rate": 2.5965250965250968e-05,
      "loss": 1.1368,
      "step": 1254
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.834712982177734,
      "learning_rate": 2.5962033462033464e-05,
      "loss": 0.8787,
      "step": 1255
    },
    {
      "epoch": 0.27,
      "grad_norm": 13.396161079406738,
      "learning_rate": 2.595881595881596e-05,
      "loss": 1.6741,
      "step": 1256
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.88947582244873,
      "learning_rate": 2.5955598455598457e-05,
      "loss": 0.9992,
      "step": 1257
    },
    {
      "epoch": 0.27,
      "grad_norm": 10.355356216430664,
      "learning_rate": 2.5952380952380953e-05,
      "loss": 1.6866,
      "step": 1258
    },
    {
      "epoch": 0.27,
      "grad_norm": 10.641451835632324,
      "learning_rate": 2.594916344916345e-05,
      "loss": 1.7178,
      "step": 1259
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.532806396484375,
      "learning_rate": 2.594594594594595e-05,
      "loss": 1.4672,
      "step": 1260
    },
    {
      "epoch": 0.27,
      "grad_norm": 10.537934303283691,
      "learning_rate": 2.5942728442728446e-05,
      "loss": 1.2157,
      "step": 1261
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.812426567077637,
      "learning_rate": 2.593951093951094e-05,
      "loss": 1.6511,
      "step": 1262
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.931818962097168,
      "learning_rate": 2.5936293436293436e-05,
      "loss": 1.2222,
      "step": 1263
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.099395751953125,
      "learning_rate": 2.5933075933075932e-05,
      "loss": 1.2136,
      "step": 1264
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.293469429016113,
      "learning_rate": 2.592985842985843e-05,
      "loss": 1.5233,
      "step": 1265
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.84615421295166,
      "learning_rate": 2.5926640926640925e-05,
      "loss": 1.3407,
      "step": 1266
    },
    {
      "epoch": 0.27,
      "grad_norm": 10.016595840454102,
      "learning_rate": 2.5923423423423425e-05,
      "loss": 1.318,
      "step": 1267
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.280338287353516,
      "learning_rate": 2.592020592020592e-05,
      "loss": 1.3416,
      "step": 1268
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.63392448425293,
      "learning_rate": 2.5916988416988418e-05,
      "loss": 1.0743,
      "step": 1269
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.18840217590332,
      "learning_rate": 2.5913770913770914e-05,
      "loss": 1.2665,
      "step": 1270
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.590978622436523,
      "learning_rate": 2.591055341055341e-05,
      "loss": 1.2881,
      "step": 1271
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.726609230041504,
      "learning_rate": 2.5907335907335907e-05,
      "loss": 1.3539,
      "step": 1272
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.85567045211792,
      "learning_rate": 2.5904118404118407e-05,
      "loss": 1.057,
      "step": 1273
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.952258110046387,
      "learning_rate": 2.5900900900900903e-05,
      "loss": 1.4416,
      "step": 1274
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.753046989440918,
      "learning_rate": 2.58976833976834e-05,
      "loss": 1.2373,
      "step": 1275
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.826437950134277,
      "learning_rate": 2.5894465894465896e-05,
      "loss": 1.2781,
      "step": 1276
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.117496490478516,
      "learning_rate": 2.5891248391248393e-05,
      "loss": 1.0468,
      "step": 1277
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.174921035766602,
      "learning_rate": 2.588803088803089e-05,
      "loss": 1.226,
      "step": 1278
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.223891258239746,
      "learning_rate": 2.5884813384813382e-05,
      "loss": 1.169,
      "step": 1279
    },
    {
      "epoch": 0.27,
      "grad_norm": 10.464613914489746,
      "learning_rate": 2.5881595881595882e-05,
      "loss": 1.5199,
      "step": 1280
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.026021957397461,
      "learning_rate": 2.587837837837838e-05,
      "loss": 1.1397,
      "step": 1281
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.349634170532227,
      "learning_rate": 2.5875160875160875e-05,
      "loss": 1.1405,
      "step": 1282
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.980854988098145,
      "learning_rate": 2.587194337194337e-05,
      "loss": 1.4774,
      "step": 1283
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.198201179504395,
      "learning_rate": 2.5868725868725868e-05,
      "loss": 1.2313,
      "step": 1284
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.738927364349365,
      "learning_rate": 2.5865508365508364e-05,
      "loss": 1.2257,
      "step": 1285
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.897549629211426,
      "learning_rate": 2.5862290862290864e-05,
      "loss": 1.1411,
      "step": 1286
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.81561803817749,
      "learning_rate": 2.585907335907336e-05,
      "loss": 1.2824,
      "step": 1287
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.260760307312012,
      "learning_rate": 2.5855855855855857e-05,
      "loss": 1.2696,
      "step": 1288
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.086288452148438,
      "learning_rate": 2.5852638352638353e-05,
      "loss": 1.4057,
      "step": 1289
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.347599029541016,
      "learning_rate": 2.584942084942085e-05,
      "loss": 1.357,
      "step": 1290
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.628552436828613,
      "learning_rate": 2.5846203346203346e-05,
      "loss": 1.2561,
      "step": 1291
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.281179428100586,
      "learning_rate": 2.5842985842985846e-05,
      "loss": 1.1866,
      "step": 1292
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.917503356933594,
      "learning_rate": 2.5839768339768343e-05,
      "loss": 1.4095,
      "step": 1293
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.127086639404297,
      "learning_rate": 2.583655083655084e-05,
      "loss": 1.2595,
      "step": 1294
    },
    {
      "epoch": 0.28,
      "grad_norm": 11.755021095275879,
      "learning_rate": 2.5833333333333336e-05,
      "loss": 1.502,
      "step": 1295
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.273344039916992,
      "learning_rate": 2.583011583011583e-05,
      "loss": 1.1169,
      "step": 1296
    },
    {
      "epoch": 0.28,
      "grad_norm": 11.19952392578125,
      "learning_rate": 2.5826898326898325e-05,
      "loss": 1.4007,
      "step": 1297
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.057230949401855,
      "learning_rate": 2.582368082368082e-05,
      "loss": 1.3295,
      "step": 1298
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.731072425842285,
      "learning_rate": 2.582046332046332e-05,
      "loss": 1.1857,
      "step": 1299
    },
    {
      "epoch": 0.28,
      "grad_norm": 6.924249649047852,
      "learning_rate": 2.5817245817245818e-05,
      "loss": 0.9916,
      "step": 1300
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.311039924621582,
      "learning_rate": 2.5814028314028314e-05,
      "loss": 1.1718,
      "step": 1301
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.154433250427246,
      "learning_rate": 2.581081081081081e-05,
      "loss": 1.6081,
      "step": 1302
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.554842948913574,
      "learning_rate": 2.5807593307593307e-05,
      "loss": 1.2945,
      "step": 1303
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.78589153289795,
      "learning_rate": 2.5804375804375804e-05,
      "loss": 1.413,
      "step": 1304
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.393954277038574,
      "learning_rate": 2.5801158301158303e-05,
      "loss": 1.1848,
      "step": 1305
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.951700210571289,
      "learning_rate": 2.57979407979408e-05,
      "loss": 1.306,
      "step": 1306
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.515171051025391,
      "learning_rate": 2.5794723294723296e-05,
      "loss": 1.2678,
      "step": 1307
    },
    {
      "epoch": 0.28,
      "grad_norm": 10.391121864318848,
      "learning_rate": 2.5791505791505793e-05,
      "loss": 1.217,
      "step": 1308
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.917411804199219,
      "learning_rate": 2.578828828828829e-05,
      "loss": 1.4641,
      "step": 1309
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.920052528381348,
      "learning_rate": 2.5785070785070786e-05,
      "loss": 1.289,
      "step": 1310
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.116941928863525,
      "learning_rate": 2.5781853281853282e-05,
      "loss": 1.1665,
      "step": 1311
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.313855171203613,
      "learning_rate": 2.5778635778635782e-05,
      "loss": 1.485,
      "step": 1312
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.172476768493652,
      "learning_rate": 2.577541827541828e-05,
      "loss": 1.2107,
      "step": 1313
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.023652076721191,
      "learning_rate": 2.577220077220077e-05,
      "loss": 0.9819,
      "step": 1314
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.958905220031738,
      "learning_rate": 2.5768983268983268e-05,
      "loss": 1.3431,
      "step": 1315
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.58459186553955,
      "learning_rate": 2.5765765765765764e-05,
      "loss": 1.2273,
      "step": 1316
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.152118682861328,
      "learning_rate": 2.576254826254826e-05,
      "loss": 1.3584,
      "step": 1317
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.95303726196289,
      "learning_rate": 2.575933075933076e-05,
      "loss": 1.3915,
      "step": 1318
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.720580101013184,
      "learning_rate": 2.5756113256113257e-05,
      "loss": 0.978,
      "step": 1319
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.10477066040039,
      "learning_rate": 2.5752895752895754e-05,
      "loss": 1.3603,
      "step": 1320
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.169732570648193,
      "learning_rate": 2.574967824967825e-05,
      "loss": 1.1252,
      "step": 1321
    },
    {
      "epoch": 0.28,
      "grad_norm": 10.426152229309082,
      "learning_rate": 2.5746460746460747e-05,
      "loss": 1.2817,
      "step": 1322
    },
    {
      "epoch": 0.28,
      "grad_norm": 11.736039161682129,
      "learning_rate": 2.5743243243243243e-05,
      "loss": 1.635,
      "step": 1323
    },
    {
      "epoch": 0.28,
      "grad_norm": 10.971854209899902,
      "learning_rate": 2.574002574002574e-05,
      "loss": 1.5793,
      "step": 1324
    },
    {
      "epoch": 0.28,
      "grad_norm": 10.724867820739746,
      "learning_rate": 2.573680823680824e-05,
      "loss": 1.6089,
      "step": 1325
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.868919372558594,
      "learning_rate": 2.5733590733590736e-05,
      "loss": 1.1438,
      "step": 1326
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.867087364196777,
      "learning_rate": 2.5730373230373232e-05,
      "loss": 0.8903,
      "step": 1327
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.990559577941895,
      "learning_rate": 2.572715572715573e-05,
      "loss": 1.4275,
      "step": 1328
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.718335151672363,
      "learning_rate": 2.5723938223938225e-05,
      "loss": 1.1193,
      "step": 1329
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.089630126953125,
      "learning_rate": 2.572072072072072e-05,
      "loss": 1.3454,
      "step": 1330
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.83558177947998,
      "learning_rate": 2.5717503217503218e-05,
      "loss": 1.4412,
      "step": 1331
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.914273738861084,
      "learning_rate": 2.5714285714285714e-05,
      "loss": 1.4658,
      "step": 1332
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.517011642456055,
      "learning_rate": 2.571106821106821e-05,
      "loss": 1.2156,
      "step": 1333
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.989364624023438,
      "learning_rate": 2.5707850707850707e-05,
      "loss": 1.3811,
      "step": 1334
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.939057350158691,
      "learning_rate": 2.5704633204633204e-05,
      "loss": 1.4968,
      "step": 1335
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.700156211853027,
      "learning_rate": 2.57014157014157e-05,
      "loss": 1.3077,
      "step": 1336
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.327420234680176,
      "learning_rate": 2.5698198198198197e-05,
      "loss": 1.2938,
      "step": 1337
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.403888702392578,
      "learning_rate": 2.5694980694980697e-05,
      "loss": 1.2723,
      "step": 1338
    },
    {
      "epoch": 0.29,
      "grad_norm": 10.75875473022461,
      "learning_rate": 2.5691763191763193e-05,
      "loss": 1.5472,
      "step": 1339
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.008528709411621,
      "learning_rate": 2.568854568854569e-05,
      "loss": 1.2783,
      "step": 1340
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.096994400024414,
      "learning_rate": 2.5685328185328186e-05,
      "loss": 1.0607,
      "step": 1341
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.67728328704834,
      "learning_rate": 2.5682110682110682e-05,
      "loss": 1.379,
      "step": 1342
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.379941940307617,
      "learning_rate": 2.567889317889318e-05,
      "loss": 1.518,
      "step": 1343
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.261212348937988,
      "learning_rate": 2.567567567567568e-05,
      "loss": 1.3642,
      "step": 1344
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.067544937133789,
      "learning_rate": 2.5672458172458175e-05,
      "loss": 1.2795,
      "step": 1345
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.539948463439941,
      "learning_rate": 2.566924066924067e-05,
      "loss": 1.0745,
      "step": 1346
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.597990989685059,
      "learning_rate": 2.5666023166023168e-05,
      "loss": 1.2837,
      "step": 1347
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.783984184265137,
      "learning_rate": 2.5662805662805664e-05,
      "loss": 1.2356,
      "step": 1348
    },
    {
      "epoch": 0.29,
      "grad_norm": 11.81760025024414,
      "learning_rate": 2.5659588159588158e-05,
      "loss": 1.4647,
      "step": 1349
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.411894798278809,
      "learning_rate": 2.5656370656370657e-05,
      "loss": 1.1542,
      "step": 1350
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.891524314880371,
      "learning_rate": 2.5653153153153154e-05,
      "loss": 1.3931,
      "step": 1351
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.021924018859863,
      "learning_rate": 2.564993564993565e-05,
      "loss": 1.4146,
      "step": 1352
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.604981422424316,
      "learning_rate": 2.5646718146718147e-05,
      "loss": 1.3722,
      "step": 1353
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.60373067855835,
      "learning_rate": 2.5643500643500643e-05,
      "loss": 1.0354,
      "step": 1354
    },
    {
      "epoch": 0.29,
      "grad_norm": 10.07453727722168,
      "learning_rate": 2.564028314028314e-05,
      "loss": 1.074,
      "step": 1355
    },
    {
      "epoch": 0.29,
      "grad_norm": 10.836448669433594,
      "learning_rate": 2.5637065637065636e-05,
      "loss": 1.634,
      "step": 1356
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.898748397827148,
      "learning_rate": 2.5633848133848136e-05,
      "loss": 1.4875,
      "step": 1357
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.539626121520996,
      "learning_rate": 2.5630630630630632e-05,
      "loss": 1.254,
      "step": 1358
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.14703369140625,
      "learning_rate": 2.562741312741313e-05,
      "loss": 1.2374,
      "step": 1359
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.596095085144043,
      "learning_rate": 2.5624195624195625e-05,
      "loss": 1.4397,
      "step": 1360
    },
    {
      "epoch": 0.29,
      "grad_norm": 10.493695259094238,
      "learning_rate": 2.5620978120978122e-05,
      "loss": 1.1982,
      "step": 1361
    },
    {
      "epoch": 0.29,
      "grad_norm": 11.386563301086426,
      "learning_rate": 2.5617760617760618e-05,
      "loss": 1.4324,
      "step": 1362
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.266709327697754,
      "learning_rate": 2.5614543114543118e-05,
      "loss": 0.9307,
      "step": 1363
    },
    {
      "epoch": 0.29,
      "grad_norm": 10.584850311279297,
      "learning_rate": 2.5611325611325614e-05,
      "loss": 1.4496,
      "step": 1364
    },
    {
      "epoch": 0.29,
      "grad_norm": 10.838656425476074,
      "learning_rate": 2.560810810810811e-05,
      "loss": 1.4224,
      "step": 1365
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.145031929016113,
      "learning_rate": 2.5604890604890604e-05,
      "loss": 1.4172,
      "step": 1366
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.71875,
      "learning_rate": 2.56016731016731e-05,
      "loss": 1.2799,
      "step": 1367
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.426546096801758,
      "learning_rate": 2.5598455598455597e-05,
      "loss": 1.3901,
      "step": 1368
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.96825647354126,
      "learning_rate": 2.5595238095238093e-05,
      "loss": 1.1583,
      "step": 1369
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.5622148513793945,
      "learning_rate": 2.5592020592020593e-05,
      "loss": 1.3988,
      "step": 1370
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.90265941619873,
      "learning_rate": 2.558880308880309e-05,
      "loss": 1.3434,
      "step": 1371
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.606028079986572,
      "learning_rate": 2.5585585585585586e-05,
      "loss": 1.518,
      "step": 1372
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.016160488128662,
      "learning_rate": 2.5582368082368083e-05,
      "loss": 0.99,
      "step": 1373
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.410078048706055,
      "learning_rate": 2.557915057915058e-05,
      "loss": 1.6635,
      "step": 1374
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.794859409332275,
      "learning_rate": 2.5575933075933075e-05,
      "loss": 1.2636,
      "step": 1375
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.765759468078613,
      "learning_rate": 2.5572715572715575e-05,
      "loss": 1.2452,
      "step": 1376
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.954371452331543,
      "learning_rate": 2.5569498069498072e-05,
      "loss": 1.2832,
      "step": 1377
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.558645248413086,
      "learning_rate": 2.5566280566280568e-05,
      "loss": 1.5153,
      "step": 1378
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.079626083374023,
      "learning_rate": 2.5563063063063065e-05,
      "loss": 1.2552,
      "step": 1379
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.338003158569336,
      "learning_rate": 2.555984555984556e-05,
      "loss": 1.0555,
      "step": 1380
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.917500495910645,
      "learning_rate": 2.5556628056628058e-05,
      "loss": 1.0714,
      "step": 1381
    },
    {
      "epoch": 0.3,
      "grad_norm": 6.867962837219238,
      "learning_rate": 2.5553410553410554e-05,
      "loss": 1.1081,
      "step": 1382
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.97779369354248,
      "learning_rate": 2.555019305019305e-05,
      "loss": 1.3739,
      "step": 1383
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.261582374572754,
      "learning_rate": 2.5546975546975547e-05,
      "loss": 1.0219,
      "step": 1384
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.76404094696045,
      "learning_rate": 2.5543758043758043e-05,
      "loss": 1.4056,
      "step": 1385
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.735231399536133,
      "learning_rate": 2.554054054054054e-05,
      "loss": 1.1606,
      "step": 1386
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.845941066741943,
      "learning_rate": 2.5537323037323036e-05,
      "loss": 0.9357,
      "step": 1387
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.703399658203125,
      "learning_rate": 2.5534105534105533e-05,
      "loss": 0.9873,
      "step": 1388
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.952198028564453,
      "learning_rate": 2.5530888030888033e-05,
      "loss": 1.1499,
      "step": 1389
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.431999206542969,
      "learning_rate": 2.552767052767053e-05,
      "loss": 1.0859,
      "step": 1390
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.108823776245117,
      "learning_rate": 2.5524453024453025e-05,
      "loss": 1.2826,
      "step": 1391
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.972494125366211,
      "learning_rate": 2.5521235521235522e-05,
      "loss": 0.8786,
      "step": 1392
    },
    {
      "epoch": 0.3,
      "grad_norm": 10.084443092346191,
      "learning_rate": 2.551801801801802e-05,
      "loss": 1.3346,
      "step": 1393
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.856070518493652,
      "learning_rate": 2.5514800514800515e-05,
      "loss": 1.3307,
      "step": 1394
    },
    {
      "epoch": 0.3,
      "grad_norm": 10.50869083404541,
      "learning_rate": 2.551158301158301e-05,
      "loss": 1.1481,
      "step": 1395
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.77332592010498,
      "learning_rate": 2.550836550836551e-05,
      "loss": 1.454,
      "step": 1396
    },
    {
      "epoch": 0.3,
      "grad_norm": 6.979003429412842,
      "learning_rate": 2.5505148005148008e-05,
      "loss": 1.0055,
      "step": 1397
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.949971199035645,
      "learning_rate": 2.5501930501930504e-05,
      "loss": 1.6227,
      "step": 1398
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.566783905029297,
      "learning_rate": 2.5498712998713e-05,
      "loss": 1.575,
      "step": 1399
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.8725152015686035,
      "learning_rate": 2.5495495495495497e-05,
      "loss": 1.2104,
      "step": 1400
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.967185974121094,
      "learning_rate": 2.549227799227799e-05,
      "loss": 1.3816,
      "step": 1401
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.024503707885742,
      "learning_rate": 2.548906048906049e-05,
      "loss": 1.7258,
      "step": 1402
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.411354064941406,
      "learning_rate": 2.5485842985842986e-05,
      "loss": 1.3804,
      "step": 1403
    },
    {
      "epoch": 0.3,
      "grad_norm": 10.43538761138916,
      "learning_rate": 2.5482625482625483e-05,
      "loss": 1.4413,
      "step": 1404
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.069273948669434,
      "learning_rate": 2.547940797940798e-05,
      "loss": 1.3333,
      "step": 1405
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.026012420654297,
      "learning_rate": 2.5476190476190476e-05,
      "loss": 1.368,
      "step": 1406
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.192071914672852,
      "learning_rate": 2.5472972972972972e-05,
      "loss": 1.3167,
      "step": 1407
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.714300155639648,
      "learning_rate": 2.5469755469755472e-05,
      "loss": 0.9992,
      "step": 1408
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.586287498474121,
      "learning_rate": 2.546653796653797e-05,
      "loss": 1.4448,
      "step": 1409
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.5651273727417,
      "learning_rate": 2.5463320463320465e-05,
      "loss": 1.0582,
      "step": 1410
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.098165512084961,
      "learning_rate": 2.546010296010296e-05,
      "loss": 1.1505,
      "step": 1411
    },
    {
      "epoch": 0.3,
      "grad_norm": 10.2263765335083,
      "learning_rate": 2.5456885456885458e-05,
      "loss": 1.3752,
      "step": 1412
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.28394603729248,
      "learning_rate": 2.5453667953667954e-05,
      "loss": 1.2614,
      "step": 1413
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.550552368164062,
      "learning_rate": 2.545045045045045e-05,
      "loss": 1.2217,
      "step": 1414
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.956452369689941,
      "learning_rate": 2.544723294723295e-05,
      "loss": 1.6847,
      "step": 1415
    },
    {
      "epoch": 0.3,
      "grad_norm": 10.333646774291992,
      "learning_rate": 2.5444015444015447e-05,
      "loss": 1.4106,
      "step": 1416
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.321438789367676,
      "learning_rate": 2.5440797940797943e-05,
      "loss": 1.5164,
      "step": 1417
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.984847068786621,
      "learning_rate": 2.5437580437580436e-05,
      "loss": 1.3691,
      "step": 1418
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.357446670532227,
      "learning_rate": 2.5434362934362933e-05,
      "loss": 1.0271,
      "step": 1419
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.903992652893066,
      "learning_rate": 2.543114543114543e-05,
      "loss": 1.3522,
      "step": 1420
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.10715103149414,
      "learning_rate": 2.542792792792793e-05,
      "loss": 1.3782,
      "step": 1421
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.8184380531311035,
      "learning_rate": 2.5424710424710426e-05,
      "loss": 1.2097,
      "step": 1422
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.777304649353027,
      "learning_rate": 2.5421492921492922e-05,
      "loss": 1.4577,
      "step": 1423
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.821316719055176,
      "learning_rate": 2.541827541827542e-05,
      "loss": 1.2562,
      "step": 1424
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.57967472076416,
      "learning_rate": 2.5415057915057915e-05,
      "loss": 1.0676,
      "step": 1425
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.212885856628418,
      "learning_rate": 2.541184041184041e-05,
      "loss": 1.2857,
      "step": 1426
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.7544145584106445,
      "learning_rate": 2.5408622908622908e-05,
      "loss": 1.3207,
      "step": 1427
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.563652992248535,
      "learning_rate": 2.5405405405405408e-05,
      "loss": 1.0678,
      "step": 1428
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.59817123413086,
      "learning_rate": 2.5402187902187904e-05,
      "loss": 1.2542,
      "step": 1429
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.306229591369629,
      "learning_rate": 2.53989703989704e-05,
      "loss": 1.078,
      "step": 1430
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.702378273010254,
      "learning_rate": 2.5395752895752897e-05,
      "loss": 1.0899,
      "step": 1431
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.163904190063477,
      "learning_rate": 2.5392535392535393e-05,
      "loss": 1.0256,
      "step": 1432
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.490063667297363,
      "learning_rate": 2.538931788931789e-05,
      "loss": 1.3183,
      "step": 1433
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.219050407409668,
      "learning_rate": 2.538610038610039e-05,
      "loss": 1.1512,
      "step": 1434
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.361957550048828,
      "learning_rate": 2.5382882882882886e-05,
      "loss": 1.3229,
      "step": 1435
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.13034439086914,
      "learning_rate": 2.537966537966538e-05,
      "loss": 1.1673,
      "step": 1436
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.526092529296875,
      "learning_rate": 2.5376447876447876e-05,
      "loss": 1.0446,
      "step": 1437
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.374307632446289,
      "learning_rate": 2.5373230373230372e-05,
      "loss": 1.1462,
      "step": 1438
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.964253425598145,
      "learning_rate": 2.537001287001287e-05,
      "loss": 1.2614,
      "step": 1439
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.736732482910156,
      "learning_rate": 2.5366795366795365e-05,
      "loss": 1.3504,
      "step": 1440
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.202870845794678,
      "learning_rate": 2.5363577863577865e-05,
      "loss": 1.0159,
      "step": 1441
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.844514846801758,
      "learning_rate": 2.536036036036036e-05,
      "loss": 1.1986,
      "step": 1442
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.007359504699707,
      "learning_rate": 2.5357142857142858e-05,
      "loss": 1.064,
      "step": 1443
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.230769157409668,
      "learning_rate": 2.5353925353925354e-05,
      "loss": 1.4222,
      "step": 1444
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.69124984741211,
      "learning_rate": 2.535070785070785e-05,
      "loss": 1.0652,
      "step": 1445
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.194783210754395,
      "learning_rate": 2.5347490347490347e-05,
      "loss": 1.2174,
      "step": 1446
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.937710762023926,
      "learning_rate": 2.5344272844272847e-05,
      "loss": 1.3902,
      "step": 1447
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.386506080627441,
      "learning_rate": 2.5341055341055343e-05,
      "loss": 1.336,
      "step": 1448
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.57827377319336,
      "learning_rate": 2.533783783783784e-05,
      "loss": 1.2543,
      "step": 1449
    },
    {
      "epoch": 0.31,
      "grad_norm": 10.046238899230957,
      "learning_rate": 2.5334620334620336e-05,
      "loss": 1.2575,
      "step": 1450
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.833523750305176,
      "learning_rate": 2.5331402831402833e-05,
      "loss": 1.3446,
      "step": 1451
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.799053192138672,
      "learning_rate": 2.532818532818533e-05,
      "loss": 1.2252,
      "step": 1452
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.315938949584961,
      "learning_rate": 2.5324967824967826e-05,
      "loss": 1.2863,
      "step": 1453
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.716084957122803,
      "learning_rate": 2.5321750321750322e-05,
      "loss": 1.2204,
      "step": 1454
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.704107761383057,
      "learning_rate": 2.531853281853282e-05,
      "loss": 1.3266,
      "step": 1455
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.016172409057617,
      "learning_rate": 2.5315315315315315e-05,
      "loss": 1.2613,
      "step": 1456
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.395118713378906,
      "learning_rate": 2.531209781209781e-05,
      "loss": 1.5211,
      "step": 1457
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.28683090209961,
      "learning_rate": 2.5308880308880308e-05,
      "loss": 1.4711,
      "step": 1458
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.541897773742676,
      "learning_rate": 2.5305662805662804e-05,
      "loss": 1.64,
      "step": 1459
    },
    {
      "epoch": 0.31,
      "grad_norm": 6.898441314697266,
      "learning_rate": 2.5302445302445304e-05,
      "loss": 0.8801,
      "step": 1460
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.491619110107422,
      "learning_rate": 2.52992277992278e-05,
      "loss": 1.1699,
      "step": 1461
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.828670024871826,
      "learning_rate": 2.5296010296010297e-05,
      "loss": 0.9021,
      "step": 1462
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.885294914245605,
      "learning_rate": 2.5292792792792794e-05,
      "loss": 1.2877,
      "step": 1463
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.978124141693115,
      "learning_rate": 2.528957528957529e-05,
      "loss": 1.1434,
      "step": 1464
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.019301414489746,
      "learning_rate": 2.5286357786357787e-05,
      "loss": 1.2189,
      "step": 1465
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.94127082824707,
      "learning_rate": 2.5283140283140286e-05,
      "loss": 1.2507,
      "step": 1466
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.426460266113281,
      "learning_rate": 2.5279922779922783e-05,
      "loss": 1.1964,
      "step": 1467
    },
    {
      "epoch": 0.31,
      "grad_norm": 10.467484474182129,
      "learning_rate": 2.527670527670528e-05,
      "loss": 1.2419,
      "step": 1468
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.809951782226562,
      "learning_rate": 2.5273487773487776e-05,
      "loss": 1.2797,
      "step": 1469
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.079028129577637,
      "learning_rate": 2.527027027027027e-05,
      "loss": 1.3093,
      "step": 1470
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.897984504699707,
      "learning_rate": 2.5267052767052765e-05,
      "loss": 1.3733,
      "step": 1471
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.164739608764648,
      "learning_rate": 2.526383526383526e-05,
      "loss": 1.4519,
      "step": 1472
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.105911254882812,
      "learning_rate": 2.526061776061776e-05,
      "loss": 1.2108,
      "step": 1473
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.444148063659668,
      "learning_rate": 2.5257400257400258e-05,
      "loss": 1.0761,
      "step": 1474
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.494030952453613,
      "learning_rate": 2.5254182754182754e-05,
      "loss": 1.1393,
      "step": 1475
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.280542373657227,
      "learning_rate": 2.525096525096525e-05,
      "loss": 1.4514,
      "step": 1476
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.897229194641113,
      "learning_rate": 2.5247747747747747e-05,
      "loss": 1.3323,
      "step": 1477
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.985555648803711,
      "learning_rate": 2.5244530244530244e-05,
      "loss": 1.2288,
      "step": 1478
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.255661010742188,
      "learning_rate": 2.5241312741312744e-05,
      "loss": 1.4782,
      "step": 1479
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.386364936828613,
      "learning_rate": 2.523809523809524e-05,
      "loss": 1.4776,
      "step": 1480
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.397079467773438,
      "learning_rate": 2.5234877734877737e-05,
      "loss": 0.9475,
      "step": 1481
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.650309085845947,
      "learning_rate": 2.5231660231660233e-05,
      "loss": 1.0661,
      "step": 1482
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.71941089630127,
      "learning_rate": 2.522844272844273e-05,
      "loss": 1.5318,
      "step": 1483
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.363667488098145,
      "learning_rate": 2.5225225225225226e-05,
      "loss": 1.2348,
      "step": 1484
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.979213714599609,
      "learning_rate": 2.5222007722007722e-05,
      "loss": 1.0889,
      "step": 1485
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.238397598266602,
      "learning_rate": 2.5218790218790222e-05,
      "loss": 1.0419,
      "step": 1486
    },
    {
      "epoch": 0.32,
      "grad_norm": 10.683333396911621,
      "learning_rate": 2.521557271557272e-05,
      "loss": 1.2552,
      "step": 1487
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.612963676452637,
      "learning_rate": 2.521235521235521e-05,
      "loss": 1.1502,
      "step": 1488
    },
    {
      "epoch": 0.32,
      "grad_norm": 12.875786781311035,
      "learning_rate": 2.5209137709137708e-05,
      "loss": 1.7018,
      "step": 1489
    },
    {
      "epoch": 0.32,
      "grad_norm": 10.769491195678711,
      "learning_rate": 2.5205920205920205e-05,
      "loss": 1.0318,
      "step": 1490
    },
    {
      "epoch": 0.32,
      "grad_norm": 10.384236335754395,
      "learning_rate": 2.52027027027027e-05,
      "loss": 1.2346,
      "step": 1491
    },
    {
      "epoch": 0.32,
      "grad_norm": 10.615726470947266,
      "learning_rate": 2.51994851994852e-05,
      "loss": 1.266,
      "step": 1492
    },
    {
      "epoch": 0.32,
      "grad_norm": 11.310644149780273,
      "learning_rate": 2.5196267696267697e-05,
      "loss": 1.4902,
      "step": 1493
    },
    {
      "epoch": 0.32,
      "grad_norm": 10.09049129486084,
      "learning_rate": 2.5193050193050194e-05,
      "loss": 1.0439,
      "step": 1494
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.805563926696777,
      "learning_rate": 2.518983268983269e-05,
      "loss": 1.25,
      "step": 1495
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.329071044921875,
      "learning_rate": 2.5186615186615187e-05,
      "loss": 1.1178,
      "step": 1496
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.315698623657227,
      "learning_rate": 2.5183397683397683e-05,
      "loss": 1.0273,
      "step": 1497
    },
    {
      "epoch": 0.32,
      "grad_norm": 6.728877067565918,
      "learning_rate": 2.518018018018018e-05,
      "loss": 0.8371,
      "step": 1498
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.068680763244629,
      "learning_rate": 2.517696267696268e-05,
      "loss": 1.4194,
      "step": 1499
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.977179050445557,
      "learning_rate": 2.5173745173745176e-05,
      "loss": 1.1444,
      "step": 1500
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.2002081871032715,
      "learning_rate": 2.5170527670527672e-05,
      "loss": 0.913,
      "step": 1501
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.151758193969727,
      "learning_rate": 2.516731016731017e-05,
      "loss": 1.3405,
      "step": 1502
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.010858535766602,
      "learning_rate": 2.5164092664092665e-05,
      "loss": 1.3492,
      "step": 1503
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.884503364562988,
      "learning_rate": 2.516087516087516e-05,
      "loss": 0.9833,
      "step": 1504
    },
    {
      "epoch": 0.32,
      "grad_norm": 10.626091003417969,
      "learning_rate": 2.5157657657657658e-05,
      "loss": 1.5074,
      "step": 1505
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.297991752624512,
      "learning_rate": 2.5154440154440155e-05,
      "loss": 1.4406,
      "step": 1506
    },
    {
      "epoch": 0.32,
      "grad_norm": 11.956210136413574,
      "learning_rate": 2.515122265122265e-05,
      "loss": 1.2528,
      "step": 1507
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.619907379150391,
      "learning_rate": 2.5148005148005147e-05,
      "loss": 1.1399,
      "step": 1508
    },
    {
      "epoch": 0.32,
      "grad_norm": 10.113675117492676,
      "learning_rate": 2.5144787644787644e-05,
      "loss": 1.5202,
      "step": 1509
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.791586875915527,
      "learning_rate": 2.514157014157014e-05,
      "loss": 1.2189,
      "step": 1510
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.695945739746094,
      "learning_rate": 2.5138352638352637e-05,
      "loss": 1.1037,
      "step": 1511
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.924935340881348,
      "learning_rate": 2.5135135135135137e-05,
      "loss": 1.3723,
      "step": 1512
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.714617729187012,
      "learning_rate": 2.5131917631917633e-05,
      "loss": 1.135,
      "step": 1513
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.71010971069336,
      "learning_rate": 2.512870012870013e-05,
      "loss": 1.3558,
      "step": 1514
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.048674583435059,
      "learning_rate": 2.5125482625482626e-05,
      "loss": 1.0667,
      "step": 1515
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.341289520263672,
      "learning_rate": 2.5122265122265122e-05,
      "loss": 1.1656,
      "step": 1516
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.060371398925781,
      "learning_rate": 2.511904761904762e-05,
      "loss": 1.3834,
      "step": 1517
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.571737289428711,
      "learning_rate": 2.511583011583012e-05,
      "loss": 1.2096,
      "step": 1518
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.532161712646484,
      "learning_rate": 2.5112612612612615e-05,
      "loss": 0.9909,
      "step": 1519
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.70201301574707,
      "learning_rate": 2.510939510939511e-05,
      "loss": 1.3245,
      "step": 1520
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.505522727966309,
      "learning_rate": 2.5106177606177608e-05,
      "loss": 1.5566,
      "step": 1521
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.400381088256836,
      "learning_rate": 2.5102960102960105e-05,
      "loss": 1.5161,
      "step": 1522
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.151002883911133,
      "learning_rate": 2.5099742599742598e-05,
      "loss": 1.0214,
      "step": 1523
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.630712985992432,
      "learning_rate": 2.5096525096525097e-05,
      "loss": 0.9645,
      "step": 1524
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.9603166580200195,
      "learning_rate": 2.5093307593307594e-05,
      "loss": 1.0922,
      "step": 1525
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.823659420013428,
      "learning_rate": 2.509009009009009e-05,
      "loss": 1.2279,
      "step": 1526
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.171030044555664,
      "learning_rate": 2.5086872586872587e-05,
      "loss": 1.3912,
      "step": 1527
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.982039451599121,
      "learning_rate": 2.5083655083655083e-05,
      "loss": 1.277,
      "step": 1528
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.358573913574219,
      "learning_rate": 2.508043758043758e-05,
      "loss": 1.1596,
      "step": 1529
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.957079887390137,
      "learning_rate": 2.5077220077220076e-05,
      "loss": 1.1795,
      "step": 1530
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.346745491027832,
      "learning_rate": 2.5074002574002576e-05,
      "loss": 1.258,
      "step": 1531
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.715097904205322,
      "learning_rate": 2.5070785070785072e-05,
      "loss": 1.2297,
      "step": 1532
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.012395858764648,
      "learning_rate": 2.506756756756757e-05,
      "loss": 1.4075,
      "step": 1533
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.470854759216309,
      "learning_rate": 2.5064350064350065e-05,
      "loss": 1.2073,
      "step": 1534
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.235054016113281,
      "learning_rate": 2.5061132561132562e-05,
      "loss": 1.176,
      "step": 1535
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.238664150238037,
      "learning_rate": 2.5057915057915058e-05,
      "loss": 1.2467,
      "step": 1536
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.859672546386719,
      "learning_rate": 2.5054697554697558e-05,
      "loss": 1.5519,
      "step": 1537
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.904205799102783,
      "learning_rate": 2.5051480051480055e-05,
      "loss": 0.9595,
      "step": 1538
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.113250732421875,
      "learning_rate": 2.504826254826255e-05,
      "loss": 1.1314,
      "step": 1539
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.40514850616455,
      "learning_rate": 2.5045045045045044e-05,
      "loss": 1.2292,
      "step": 1540
    },
    {
      "epoch": 0.33,
      "grad_norm": 12.212021827697754,
      "learning_rate": 2.504182754182754e-05,
      "loss": 1.3467,
      "step": 1541
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.174837112426758,
      "learning_rate": 2.5038610038610037e-05,
      "loss": 1.2954,
      "step": 1542
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.745437622070312,
      "learning_rate": 2.5035392535392533e-05,
      "loss": 1.2315,
      "step": 1543
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.410746574401855,
      "learning_rate": 2.5032175032175033e-05,
      "loss": 1.5854,
      "step": 1544
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.953657150268555,
      "learning_rate": 2.502895752895753e-05,
      "loss": 1.1209,
      "step": 1545
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.847073554992676,
      "learning_rate": 2.5025740025740026e-05,
      "loss": 1.2095,
      "step": 1546
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.898786544799805,
      "learning_rate": 2.5022522522522523e-05,
      "loss": 1.3877,
      "step": 1547
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.81867790222168,
      "learning_rate": 2.501930501930502e-05,
      "loss": 1.4024,
      "step": 1548
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.702744483947754,
      "learning_rate": 2.5016087516087516e-05,
      "loss": 0.9613,
      "step": 1549
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.02190113067627,
      "learning_rate": 2.5012870012870015e-05,
      "loss": 1.0731,
      "step": 1550
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.682918548583984,
      "learning_rate": 2.5009652509652512e-05,
      "loss": 1.0643,
      "step": 1551
    },
    {
      "epoch": 0.33,
      "grad_norm": 11.180509567260742,
      "learning_rate": 2.5006435006435008e-05,
      "loss": 1.3789,
      "step": 1552
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.687432289123535,
      "learning_rate": 2.5003217503217505e-05,
      "loss": 1.1637,
      "step": 1553
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.665892601013184,
      "learning_rate": 2.5e-05,
      "loss": 1.1894,
      "step": 1554
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.048742294311523,
      "learning_rate": 2.4996782496782498e-05,
      "loss": 1.114,
      "step": 1555
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.703390121459961,
      "learning_rate": 2.4993564993564994e-05,
      "loss": 1.0082,
      "step": 1556
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.268327713012695,
      "learning_rate": 2.499034749034749e-05,
      "loss": 1.2123,
      "step": 1557
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.961966514587402,
      "learning_rate": 2.4987129987129987e-05,
      "loss": 1.5339,
      "step": 1558
    },
    {
      "epoch": 0.33,
      "grad_norm": 9.237435340881348,
      "learning_rate": 2.4983912483912483e-05,
      "loss": 0.9898,
      "step": 1559
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.729095458984375,
      "learning_rate": 2.498069498069498e-05,
      "loss": 1.4627,
      "step": 1560
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.210552215576172,
      "learning_rate": 2.4977477477477476e-05,
      "loss": 1.064,
      "step": 1561
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.757833480834961,
      "learning_rate": 2.4974259974259973e-05,
      "loss": 1.1106,
      "step": 1562
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.531932830810547,
      "learning_rate": 2.4971042471042473e-05,
      "loss": 1.1043,
      "step": 1563
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.217845916748047,
      "learning_rate": 2.496782496782497e-05,
      "loss": 1.1533,
      "step": 1564
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.889455795288086,
      "learning_rate": 2.4964607464607466e-05,
      "loss": 1.1124,
      "step": 1565
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.5731201171875,
      "learning_rate": 2.4961389961389962e-05,
      "loss": 1.4141,
      "step": 1566
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.432723999023438,
      "learning_rate": 2.495817245817246e-05,
      "loss": 1.1767,
      "step": 1567
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.594404220581055,
      "learning_rate": 2.4954954954954955e-05,
      "loss": 0.99,
      "step": 1568
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.31002426147461,
      "learning_rate": 2.495173745173745e-05,
      "loss": 1.3428,
      "step": 1569
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.791961669921875,
      "learning_rate": 2.494851994851995e-05,
      "loss": 1.359,
      "step": 1570
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.335147857666016,
      "learning_rate": 2.4945302445302448e-05,
      "loss": 1.3154,
      "step": 1571
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.040167808532715,
      "learning_rate": 2.4942084942084944e-05,
      "loss": 1.0288,
      "step": 1572
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.395242691040039,
      "learning_rate": 2.493886743886744e-05,
      "loss": 1.1138,
      "step": 1573
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.125460624694824,
      "learning_rate": 2.4935649935649937e-05,
      "loss": 1.1898,
      "step": 1574
    },
    {
      "epoch": 0.34,
      "grad_norm": 10.339346885681152,
      "learning_rate": 2.493243243243243e-05,
      "loss": 1.4801,
      "step": 1575
    },
    {
      "epoch": 0.34,
      "grad_norm": 10.334222793579102,
      "learning_rate": 2.492921492921493e-05,
      "loss": 1.6649,
      "step": 1576
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.485201835632324,
      "learning_rate": 2.4925997425997426e-05,
      "loss": 1.4573,
      "step": 1577
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.639435291290283,
      "learning_rate": 2.4922779922779923e-05,
      "loss": 1.1411,
      "step": 1578
    },
    {
      "epoch": 0.34,
      "grad_norm": 10.529488563537598,
      "learning_rate": 2.491956241956242e-05,
      "loss": 1.4055,
      "step": 1579
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.201006889343262,
      "learning_rate": 2.4916344916344916e-05,
      "loss": 0.9388,
      "step": 1580
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.768936157226562,
      "learning_rate": 2.4913127413127412e-05,
      "loss": 1.2371,
      "step": 1581
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.436293601989746,
      "learning_rate": 2.4909909909909912e-05,
      "loss": 1.4441,
      "step": 1582
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.164101600646973,
      "learning_rate": 2.490669240669241e-05,
      "loss": 1.2279,
      "step": 1583
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.951783657073975,
      "learning_rate": 2.4903474903474905e-05,
      "loss": 1.2402,
      "step": 1584
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.549193382263184,
      "learning_rate": 2.49002574002574e-05,
      "loss": 1.5837,
      "step": 1585
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.314566612243652,
      "learning_rate": 2.4897039897039898e-05,
      "loss": 0.9568,
      "step": 1586
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.363920211791992,
      "learning_rate": 2.4893822393822394e-05,
      "loss": 1.1728,
      "step": 1587
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.647223472595215,
      "learning_rate": 2.489060489060489e-05,
      "loss": 1.1227,
      "step": 1588
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.30435562133789,
      "learning_rate": 2.488738738738739e-05,
      "loss": 1.3556,
      "step": 1589
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.38513469696045,
      "learning_rate": 2.4884169884169887e-05,
      "loss": 1.0875,
      "step": 1590
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.614442825317383,
      "learning_rate": 2.4880952380952383e-05,
      "loss": 0.9486,
      "step": 1591
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.991761207580566,
      "learning_rate": 2.4877734877734877e-05,
      "loss": 1.2886,
      "step": 1592
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.386474609375,
      "learning_rate": 2.4874517374517373e-05,
      "loss": 1.3013,
      "step": 1593
    },
    {
      "epoch": 0.34,
      "grad_norm": 10.271858215332031,
      "learning_rate": 2.487129987129987e-05,
      "loss": 1.2876,
      "step": 1594
    },
    {
      "epoch": 0.34,
      "grad_norm": 11.410598754882812,
      "learning_rate": 2.486808236808237e-05,
      "loss": 1.3912,
      "step": 1595
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.17082691192627,
      "learning_rate": 2.4864864864864866e-05,
      "loss": 1.0831,
      "step": 1596
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.584736824035645,
      "learning_rate": 2.4861647361647362e-05,
      "loss": 0.976,
      "step": 1597
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.964741706848145,
      "learning_rate": 2.485842985842986e-05,
      "loss": 1.5037,
      "step": 1598
    },
    {
      "epoch": 0.34,
      "grad_norm": 10.689693450927734,
      "learning_rate": 2.4855212355212355e-05,
      "loss": 1.0257,
      "step": 1599
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.301376342773438,
      "learning_rate": 2.485199485199485e-05,
      "loss": 1.4172,
      "step": 1600
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.532317161560059,
      "learning_rate": 2.4848777348777348e-05,
      "loss": 1.2194,
      "step": 1601
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.292110443115234,
      "learning_rate": 2.4845559845559848e-05,
      "loss": 1.2848,
      "step": 1602
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.701294898986816,
      "learning_rate": 2.4842342342342344e-05,
      "loss": 1.444,
      "step": 1603
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.76638412475586,
      "learning_rate": 2.483912483912484e-05,
      "loss": 1.2272,
      "step": 1604
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.884123802185059,
      "learning_rate": 2.4835907335907337e-05,
      "loss": 1.4459,
      "step": 1605
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.056273460388184,
      "learning_rate": 2.4832689832689834e-05,
      "loss": 0.9821,
      "step": 1606
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.941705703735352,
      "learning_rate": 2.482947232947233e-05,
      "loss": 1.4824,
      "step": 1607
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.172957420349121,
      "learning_rate": 2.482625482625483e-05,
      "loss": 1.4437,
      "step": 1608
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.842580318450928,
      "learning_rate": 2.4823037323037326e-05,
      "loss": 1.5776,
      "step": 1609
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.648873329162598,
      "learning_rate": 2.481981981981982e-05,
      "loss": 1.0884,
      "step": 1610
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.652463912963867,
      "learning_rate": 2.4816602316602316e-05,
      "loss": 1.3948,
      "step": 1611
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.177796363830566,
      "learning_rate": 2.4813384813384812e-05,
      "loss": 1.1726,
      "step": 1612
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.1834077835083,
      "learning_rate": 2.481016731016731e-05,
      "loss": 1.3656,
      "step": 1613
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.006274223327637,
      "learning_rate": 2.4806949806949805e-05,
      "loss": 1.4502,
      "step": 1614
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.201849937438965,
      "learning_rate": 2.4803732303732305e-05,
      "loss": 1.1656,
      "step": 1615
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.683061599731445,
      "learning_rate": 2.48005148005148e-05,
      "loss": 1.1291,
      "step": 1616
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.729203224182129,
      "learning_rate": 2.4797297297297298e-05,
      "loss": 1.5961,
      "step": 1617
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.6463093757629395,
      "learning_rate": 2.4794079794079794e-05,
      "loss": 1.0943,
      "step": 1618
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.674798965454102,
      "learning_rate": 2.479086229086229e-05,
      "loss": 1.1168,
      "step": 1619
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.179137229919434,
      "learning_rate": 2.4787644787644787e-05,
      "loss": 1.281,
      "step": 1620
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.595569610595703,
      "learning_rate": 2.4784427284427287e-05,
      "loss": 1.1659,
      "step": 1621
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.10300350189209,
      "learning_rate": 2.4781209781209784e-05,
      "loss": 0.9745,
      "step": 1622
    },
    {
      "epoch": 0.35,
      "grad_norm": 10.197821617126465,
      "learning_rate": 2.477799227799228e-05,
      "loss": 1.5229,
      "step": 1623
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.962986946105957,
      "learning_rate": 2.4774774774774777e-05,
      "loss": 1.3448,
      "step": 1624
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.597389221191406,
      "learning_rate": 2.4771557271557273e-05,
      "loss": 1.1609,
      "step": 1625
    },
    {
      "epoch": 0.35,
      "grad_norm": 10.523836135864258,
      "learning_rate": 2.476833976833977e-05,
      "loss": 1.2772,
      "step": 1626
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.231690406799316,
      "learning_rate": 2.4765122265122266e-05,
      "loss": 1.0767,
      "step": 1627
    },
    {
      "epoch": 0.35,
      "grad_norm": 10.430663108825684,
      "learning_rate": 2.4761904761904762e-05,
      "loss": 1.1732,
      "step": 1628
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.739099502563477,
      "learning_rate": 2.475868725868726e-05,
      "loss": 1.2412,
      "step": 1629
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.291458129882812,
      "learning_rate": 2.4755469755469755e-05,
      "loss": 1.2727,
      "step": 1630
    },
    {
      "epoch": 0.35,
      "grad_norm": 10.464925765991211,
      "learning_rate": 2.475225225225225e-05,
      "loss": 1.5578,
      "step": 1631
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.477471351623535,
      "learning_rate": 2.4749034749034748e-05,
      "loss": 1.0064,
      "step": 1632
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.456615447998047,
      "learning_rate": 2.4745817245817245e-05,
      "loss": 1.5239,
      "step": 1633
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.111398696899414,
      "learning_rate": 2.4742599742599744e-05,
      "loss": 1.3234,
      "step": 1634
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.428000450134277,
      "learning_rate": 2.473938223938224e-05,
      "loss": 1.1953,
      "step": 1635
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.151668548583984,
      "learning_rate": 2.4736164736164737e-05,
      "loss": 1.1613,
      "step": 1636
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.06283187866211,
      "learning_rate": 2.4732947232947234e-05,
      "loss": 1.4246,
      "step": 1637
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.115405082702637,
      "learning_rate": 2.472972972972973e-05,
      "loss": 1.2054,
      "step": 1638
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.657777786254883,
      "learning_rate": 2.4726512226512227e-05,
      "loss": 0.9407,
      "step": 1639
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.391716003417969,
      "learning_rate": 2.4723294723294727e-05,
      "loss": 1.3636,
      "step": 1640
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.747883319854736,
      "learning_rate": 2.4720077220077223e-05,
      "loss": 1.1382,
      "step": 1641
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.893711090087891,
      "learning_rate": 2.471685971685972e-05,
      "loss": 1.1132,
      "step": 1642
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.604870796203613,
      "learning_rate": 2.4713642213642216e-05,
      "loss": 1.4541,
      "step": 1643
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.332478523254395,
      "learning_rate": 2.471042471042471e-05,
      "loss": 1.1718,
      "step": 1644
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.267882347106934,
      "learning_rate": 2.4707207207207205e-05,
      "loss": 1.0584,
      "step": 1645
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.264578819274902,
      "learning_rate": 2.4703989703989702e-05,
      "loss": 1.2212,
      "step": 1646
    },
    {
      "epoch": 0.35,
      "grad_norm": 10.030601501464844,
      "learning_rate": 2.47007722007722e-05,
      "loss": 1.0505,
      "step": 1647
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.853623390197754,
      "learning_rate": 2.4697554697554698e-05,
      "loss": 1.2217,
      "step": 1648
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.676717758178711,
      "learning_rate": 2.4694337194337195e-05,
      "loss": 1.2393,
      "step": 1649
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.960338115692139,
      "learning_rate": 2.469111969111969e-05,
      "loss": 1.1579,
      "step": 1650
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.809491157531738,
      "learning_rate": 2.4687902187902187e-05,
      "loss": 0.9222,
      "step": 1651
    },
    {
      "epoch": 0.35,
      "grad_norm": 10.971148490905762,
      "learning_rate": 2.4684684684684684e-05,
      "loss": 1.5964,
      "step": 1652
    },
    {
      "epoch": 0.35,
      "grad_norm": 10.771039009094238,
      "learning_rate": 2.4681467181467184e-05,
      "loss": 1.6589,
      "step": 1653
    },
    {
      "epoch": 0.35,
      "grad_norm": 10.848199844360352,
      "learning_rate": 2.467824967824968e-05,
      "loss": 1.2067,
      "step": 1654
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.093308448791504,
      "learning_rate": 2.4675032175032177e-05,
      "loss": 1.0168,
      "step": 1655
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.142848014831543,
      "learning_rate": 2.4671814671814673e-05,
      "loss": 1.2213,
      "step": 1656
    },
    {
      "epoch": 0.36,
      "grad_norm": 10.831751823425293,
      "learning_rate": 2.466859716859717e-05,
      "loss": 1.3495,
      "step": 1657
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.290292739868164,
      "learning_rate": 2.4665379665379666e-05,
      "loss": 1.3236,
      "step": 1658
    },
    {
      "epoch": 0.36,
      "grad_norm": 12.370402336120605,
      "learning_rate": 2.4662162162162162e-05,
      "loss": 1.6188,
      "step": 1659
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.849467754364014,
      "learning_rate": 2.4658944658944662e-05,
      "loss": 1.1155,
      "step": 1660
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.958739280700684,
      "learning_rate": 2.465572715572716e-05,
      "loss": 1.424,
      "step": 1661
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.653251647949219,
      "learning_rate": 2.4652509652509652e-05,
      "loss": 0.8615,
      "step": 1662
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.913247108459473,
      "learning_rate": 2.4649292149292148e-05,
      "loss": 0.9876,
      "step": 1663
    },
    {
      "epoch": 0.36,
      "grad_norm": 10.236970901489258,
      "learning_rate": 2.4646074646074645e-05,
      "loss": 1.363,
      "step": 1664
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.09054946899414,
      "learning_rate": 2.464285714285714e-05,
      "loss": 1.1744,
      "step": 1665
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.617445468902588,
      "learning_rate": 2.463963963963964e-05,
      "loss": 1.0689,
      "step": 1666
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.293801307678223,
      "learning_rate": 2.4636422136422137e-05,
      "loss": 1.1118,
      "step": 1667
    },
    {
      "epoch": 0.36,
      "grad_norm": 11.15234088897705,
      "learning_rate": 2.4633204633204634e-05,
      "loss": 1.0139,
      "step": 1668
    },
    {
      "epoch": 0.36,
      "grad_norm": 11.847970008850098,
      "learning_rate": 2.462998712998713e-05,
      "loss": 1.5105,
      "step": 1669
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.610703468322754,
      "learning_rate": 2.4626769626769627e-05,
      "loss": 1.3457,
      "step": 1670
    },
    {
      "epoch": 0.36,
      "grad_norm": 11.224300384521484,
      "learning_rate": 2.4623552123552123e-05,
      "loss": 1.5673,
      "step": 1671
    },
    {
      "epoch": 0.36,
      "grad_norm": 10.160672187805176,
      "learning_rate": 2.462033462033462e-05,
      "loss": 1.3984,
      "step": 1672
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.425350189208984,
      "learning_rate": 2.461711711711712e-05,
      "loss": 1.3497,
      "step": 1673
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.006714820861816,
      "learning_rate": 2.4613899613899616e-05,
      "loss": 1.0185,
      "step": 1674
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.566041469573975,
      "learning_rate": 2.4610682110682112e-05,
      "loss": 0.9837,
      "step": 1675
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.141012191772461,
      "learning_rate": 2.460746460746461e-05,
      "loss": 1.4297,
      "step": 1676
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.998398780822754,
      "learning_rate": 2.4604247104247105e-05,
      "loss": 1.2157,
      "step": 1677
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.1703267097473145,
      "learning_rate": 2.4601029601029602e-05,
      "loss": 0.9207,
      "step": 1678
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.022798538208008,
      "learning_rate": 2.4597812097812098e-05,
      "loss": 0.8862,
      "step": 1679
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.330910682678223,
      "learning_rate": 2.4594594594594595e-05,
      "loss": 1.0389,
      "step": 1680
    },
    {
      "epoch": 0.36,
      "grad_norm": 10.03331184387207,
      "learning_rate": 2.459137709137709e-05,
      "loss": 1.5737,
      "step": 1681
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.03886604309082,
      "learning_rate": 2.4588159588159588e-05,
      "loss": 1.2134,
      "step": 1682
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.359893798828125,
      "learning_rate": 2.4584942084942084e-05,
      "loss": 1.2055,
      "step": 1683
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.442642211914062,
      "learning_rate": 2.458172458172458e-05,
      "loss": 1.1344,
      "step": 1684
    },
    {
      "epoch": 0.36,
      "grad_norm": 10.119388580322266,
      "learning_rate": 2.4578507078507077e-05,
      "loss": 1.3251,
      "step": 1685
    },
    {
      "epoch": 0.36,
      "grad_norm": 11.28286075592041,
      "learning_rate": 2.4575289575289577e-05,
      "loss": 2.0059,
      "step": 1686
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.934686183929443,
      "learning_rate": 2.4572072072072073e-05,
      "loss": 1.2125,
      "step": 1687
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.472248077392578,
      "learning_rate": 2.456885456885457e-05,
      "loss": 1.1425,
      "step": 1688
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.454527854919434,
      "learning_rate": 2.4565637065637066e-05,
      "loss": 1.169,
      "step": 1689
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.800084590911865,
      "learning_rate": 2.4562419562419563e-05,
      "loss": 0.8966,
      "step": 1690
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.721144676208496,
      "learning_rate": 2.455920205920206e-05,
      "loss": 1.3143,
      "step": 1691
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.226163864135742,
      "learning_rate": 2.455598455598456e-05,
      "loss": 1.0188,
      "step": 1692
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.50477409362793,
      "learning_rate": 2.4552767052767055e-05,
      "loss": 1.1885,
      "step": 1693
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.00232219696045,
      "learning_rate": 2.4549549549549552e-05,
      "loss": 0.9778,
      "step": 1694
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.938128471374512,
      "learning_rate": 2.4546332046332048e-05,
      "loss": 1.0372,
      "step": 1695
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.800548553466797,
      "learning_rate": 2.4543114543114545e-05,
      "loss": 1.176,
      "step": 1696
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.787965774536133,
      "learning_rate": 2.4539897039897038e-05,
      "loss": 1.0361,
      "step": 1697
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.133511543273926,
      "learning_rate": 2.4536679536679538e-05,
      "loss": 1.2438,
      "step": 1698
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.142977714538574,
      "learning_rate": 2.4533462033462034e-05,
      "loss": 1.3114,
      "step": 1699
    },
    {
      "epoch": 0.36,
      "grad_norm": 8.897685050964355,
      "learning_rate": 2.453024453024453e-05,
      "loss": 1.1126,
      "step": 1700
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.375411987304688,
      "learning_rate": 2.4527027027027027e-05,
      "loss": 1.1654,
      "step": 1701
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.492084503173828,
      "learning_rate": 2.4523809523809523e-05,
      "loss": 0.9381,
      "step": 1702
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.744837760925293,
      "learning_rate": 2.452059202059202e-05,
      "loss": 1.1545,
      "step": 1703
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.881958961486816,
      "learning_rate": 2.4517374517374516e-05,
      "loss": 1.3903,
      "step": 1704
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.741973876953125,
      "learning_rate": 2.4514157014157016e-05,
      "loss": 1.0067,
      "step": 1705
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.833748817443848,
      "learning_rate": 2.4510939510939513e-05,
      "loss": 1.0942,
      "step": 1706
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.339249610900879,
      "learning_rate": 2.450772200772201e-05,
      "loss": 1.4823,
      "step": 1707
    },
    {
      "epoch": 0.37,
      "grad_norm": 10.026108741760254,
      "learning_rate": 2.4504504504504506e-05,
      "loss": 1.0761,
      "step": 1708
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.269538879394531,
      "learning_rate": 2.4501287001287002e-05,
      "loss": 1.2542,
      "step": 1709
    },
    {
      "epoch": 0.37,
      "grad_norm": 10.254289627075195,
      "learning_rate": 2.44980694980695e-05,
      "loss": 1.623,
      "step": 1710
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.528738975524902,
      "learning_rate": 2.4494851994851998e-05,
      "loss": 1.4674,
      "step": 1711
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.592386245727539,
      "learning_rate": 2.4491634491634495e-05,
      "loss": 1.1641,
      "step": 1712
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.697806358337402,
      "learning_rate": 2.448841698841699e-05,
      "loss": 1.5874,
      "step": 1713
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.751654624938965,
      "learning_rate": 2.4485199485199484e-05,
      "loss": 1.1389,
      "step": 1714
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.874773025512695,
      "learning_rate": 2.448198198198198e-05,
      "loss": 1.4029,
      "step": 1715
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.294597625732422,
      "learning_rate": 2.4478764478764477e-05,
      "loss": 1.2484,
      "step": 1716
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.801019668579102,
      "learning_rate": 2.4475546975546974e-05,
      "loss": 1.3965,
      "step": 1717
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.390128135681152,
      "learning_rate": 2.4472329472329473e-05,
      "loss": 1.0576,
      "step": 1718
    },
    {
      "epoch": 0.37,
      "grad_norm": 6.544334411621094,
      "learning_rate": 2.446911196911197e-05,
      "loss": 0.8125,
      "step": 1719
    },
    {
      "epoch": 0.37,
      "grad_norm": 13.906655311584473,
      "learning_rate": 2.4465894465894466e-05,
      "loss": 1.446,
      "step": 1720
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.655426025390625,
      "learning_rate": 2.4462676962676963e-05,
      "loss": 1.1962,
      "step": 1721
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.29782772064209,
      "learning_rate": 2.445945945945946e-05,
      "loss": 1.1712,
      "step": 1722
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.158010482788086,
      "learning_rate": 2.4456241956241956e-05,
      "loss": 1.1496,
      "step": 1723
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.988526344299316,
      "learning_rate": 2.4453024453024456e-05,
      "loss": 1.1405,
      "step": 1724
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.49443531036377,
      "learning_rate": 2.4449806949806952e-05,
      "loss": 1.3093,
      "step": 1725
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.452977180480957,
      "learning_rate": 2.444658944658945e-05,
      "loss": 1.1248,
      "step": 1726
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.012551307678223,
      "learning_rate": 2.4443371943371945e-05,
      "loss": 1.0796,
      "step": 1727
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.839077949523926,
      "learning_rate": 2.444015444015444e-05,
      "loss": 1.1436,
      "step": 1728
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.372686386108398,
      "learning_rate": 2.4436936936936938e-05,
      "loss": 1.2846,
      "step": 1729
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.94355297088623,
      "learning_rate": 2.4433719433719434e-05,
      "loss": 1.2755,
      "step": 1730
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.052216053009033,
      "learning_rate": 2.443050193050193e-05,
      "loss": 0.7398,
      "step": 1731
    },
    {
      "epoch": 0.37,
      "grad_norm": 11.92138385772705,
      "learning_rate": 2.4427284427284427e-05,
      "loss": 1.5685,
      "step": 1732
    },
    {
      "epoch": 0.37,
      "grad_norm": 10.16118335723877,
      "learning_rate": 2.4424066924066924e-05,
      "loss": 1.223,
      "step": 1733
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.624832153320312,
      "learning_rate": 2.442084942084942e-05,
      "loss": 1.3617,
      "step": 1734
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.792877197265625,
      "learning_rate": 2.4417631917631916e-05,
      "loss": 1.2972,
      "step": 1735
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.865265369415283,
      "learning_rate": 2.4414414414414413e-05,
      "loss": 1.0512,
      "step": 1736
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.778358459472656,
      "learning_rate": 2.4411196911196913e-05,
      "loss": 1.4063,
      "step": 1737
    },
    {
      "epoch": 0.37,
      "grad_norm": 11.232443809509277,
      "learning_rate": 2.440797940797941e-05,
      "loss": 1.1477,
      "step": 1738
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.783764839172363,
      "learning_rate": 2.4404761904761906e-05,
      "loss": 1.3263,
      "step": 1739
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.876152992248535,
      "learning_rate": 2.4401544401544402e-05,
      "loss": 1.1945,
      "step": 1740
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.424886703491211,
      "learning_rate": 2.43983268983269e-05,
      "loss": 1.1426,
      "step": 1741
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.820708274841309,
      "learning_rate": 2.4395109395109395e-05,
      "loss": 1.1126,
      "step": 1742
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.53456974029541,
      "learning_rate": 2.439189189189189e-05,
      "loss": 1.1543,
      "step": 1743
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.578717231750488,
      "learning_rate": 2.438867438867439e-05,
      "loss": 1.3846,
      "step": 1744
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.092554092407227,
      "learning_rate": 2.4385456885456888e-05,
      "loss": 1.1927,
      "step": 1745
    },
    {
      "epoch": 0.37,
      "grad_norm": 10.014244079589844,
      "learning_rate": 2.4382239382239384e-05,
      "loss": 1.313,
      "step": 1746
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.08873462677002,
      "learning_rate": 2.437902187902188e-05,
      "loss": 0.905,
      "step": 1747
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.100590705871582,
      "learning_rate": 2.4375804375804377e-05,
      "loss": 1.3768,
      "step": 1748
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.294144630432129,
      "learning_rate": 2.437258687258687e-05,
      "loss": 1.6366,
      "step": 1749
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.813789367675781,
      "learning_rate": 2.436936936936937e-05,
      "loss": 1.3123,
      "step": 1750
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.725442886352539,
      "learning_rate": 2.4366151866151866e-05,
      "loss": 1.1608,
      "step": 1751
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.670440673828125,
      "learning_rate": 2.4362934362934363e-05,
      "loss": 0.9952,
      "step": 1752
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.391000747680664,
      "learning_rate": 2.435971685971686e-05,
      "loss": 1.3029,
      "step": 1753
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.158426284790039,
      "learning_rate": 2.4356499356499356e-05,
      "loss": 1.4109,
      "step": 1754
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.10976505279541,
      "learning_rate": 2.4353281853281852e-05,
      "loss": 1.0943,
      "step": 1755
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.494565963745117,
      "learning_rate": 2.4350064350064352e-05,
      "loss": 1.1579,
      "step": 1756
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.928674697875977,
      "learning_rate": 2.434684684684685e-05,
      "loss": 1.4367,
      "step": 1757
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.302224159240723,
      "learning_rate": 2.4343629343629345e-05,
      "loss": 1.2515,
      "step": 1758
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.039351463317871,
      "learning_rate": 2.434041184041184e-05,
      "loss": 0.9715,
      "step": 1759
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.987559795379639,
      "learning_rate": 2.4337194337194338e-05,
      "loss": 1.0797,
      "step": 1760
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.610360145568848,
      "learning_rate": 2.4333976833976834e-05,
      "loss": 1.5938,
      "step": 1761
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.392047882080078,
      "learning_rate": 2.433075933075933e-05,
      "loss": 1.2052,
      "step": 1762
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.421436309814453,
      "learning_rate": 2.432754182754183e-05,
      "loss": 1.1971,
      "step": 1763
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.726210117340088,
      "learning_rate": 2.4324324324324327e-05,
      "loss": 0.9142,
      "step": 1764
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.724699974060059,
      "learning_rate": 2.4321106821106824e-05,
      "loss": 1.0518,
      "step": 1765
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.471741676330566,
      "learning_rate": 2.4317889317889317e-05,
      "loss": 1.393,
      "step": 1766
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.459235191345215,
      "learning_rate": 2.4314671814671813e-05,
      "loss": 1.1726,
      "step": 1767
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.473833084106445,
      "learning_rate": 2.431145431145431e-05,
      "loss": 1.13,
      "step": 1768
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.537222862243652,
      "learning_rate": 2.430823680823681e-05,
      "loss": 1.3942,
      "step": 1769
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.114790916442871,
      "learning_rate": 2.4305019305019306e-05,
      "loss": 1.1195,
      "step": 1770
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.39255142211914,
      "learning_rate": 2.4301801801801802e-05,
      "loss": 0.9585,
      "step": 1771
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.15117359161377,
      "learning_rate": 2.42985842985843e-05,
      "loss": 1.3317,
      "step": 1772
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.289314270019531,
      "learning_rate": 2.4295366795366795e-05,
      "loss": 1.122,
      "step": 1773
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.106164932250977,
      "learning_rate": 2.429214929214929e-05,
      "loss": 1.2971,
      "step": 1774
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.002787590026855,
      "learning_rate": 2.4288931788931788e-05,
      "loss": 1.2245,
      "step": 1775
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.54269790649414,
      "learning_rate": 2.4285714285714288e-05,
      "loss": 0.9774,
      "step": 1776
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.226537704467773,
      "learning_rate": 2.4282496782496784e-05,
      "loss": 0.8559,
      "step": 1777
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.987318992614746,
      "learning_rate": 2.427927927927928e-05,
      "loss": 1.0645,
      "step": 1778
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.071639060974121,
      "learning_rate": 2.4276061776061777e-05,
      "loss": 1.0696,
      "step": 1779
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.760765075683594,
      "learning_rate": 2.4272844272844274e-05,
      "loss": 1.3341,
      "step": 1780
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.513437271118164,
      "learning_rate": 2.426962676962677e-05,
      "loss": 1.1226,
      "step": 1781
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.141834259033203,
      "learning_rate": 2.426640926640927e-05,
      "loss": 1.306,
      "step": 1782
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.861509323120117,
      "learning_rate": 2.4263191763191766e-05,
      "loss": 1.3805,
      "step": 1783
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.881958961486816,
      "learning_rate": 2.425997425997426e-05,
      "loss": 1.5435,
      "step": 1784
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.214802742004395,
      "learning_rate": 2.4256756756756756e-05,
      "loss": 1.085,
      "step": 1785
    },
    {
      "epoch": 0.38,
      "grad_norm": 11.746536254882812,
      "learning_rate": 2.4253539253539252e-05,
      "loss": 1.4949,
      "step": 1786
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.463196754455566,
      "learning_rate": 2.425032175032175e-05,
      "loss": 1.3051,
      "step": 1787
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.192131042480469,
      "learning_rate": 2.4247104247104245e-05,
      "loss": 1.1331,
      "step": 1788
    },
    {
      "epoch": 0.38,
      "grad_norm": 12.049812316894531,
      "learning_rate": 2.4243886743886745e-05,
      "loss": 1.2108,
      "step": 1789
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.080999374389648,
      "learning_rate": 2.424066924066924e-05,
      "loss": 1.3328,
      "step": 1790
    },
    {
      "epoch": 0.38,
      "grad_norm": 7.143941402435303,
      "learning_rate": 2.4237451737451738e-05,
      "loss": 1.2238,
      "step": 1791
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.338170051574707,
      "learning_rate": 2.4234234234234235e-05,
      "loss": 1.0822,
      "step": 1792
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.503619194030762,
      "learning_rate": 2.423101673101673e-05,
      "loss": 1.1859,
      "step": 1793
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.489274978637695,
      "learning_rate": 2.4227799227799227e-05,
      "loss": 1.2916,
      "step": 1794
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.180689811706543,
      "learning_rate": 2.4224581724581727e-05,
      "loss": 1.0281,
      "step": 1795
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.749066352844238,
      "learning_rate": 2.4221364221364224e-05,
      "loss": 1.1478,
      "step": 1796
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.885631561279297,
      "learning_rate": 2.421814671814672e-05,
      "loss": 1.5367,
      "step": 1797
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.543378829956055,
      "learning_rate": 2.4214929214929217e-05,
      "loss": 1.0896,
      "step": 1798
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.538227081298828,
      "learning_rate": 2.4211711711711713e-05,
      "loss": 1.2627,
      "step": 1799
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.01951789855957,
      "learning_rate": 2.420849420849421e-05,
      "loss": 1.4125,
      "step": 1800
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.969181060791016,
      "learning_rate": 2.4205276705276706e-05,
      "loss": 1.2871,
      "step": 1801
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.000245094299316,
      "learning_rate": 2.4202059202059202e-05,
      "loss": 1.1259,
      "step": 1802
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.512533187866211,
      "learning_rate": 2.41988416988417e-05,
      "loss": 1.266,
      "step": 1803
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.154868125915527,
      "learning_rate": 2.4195624195624195e-05,
      "loss": 1.1587,
      "step": 1804
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.448160171508789,
      "learning_rate": 2.4192406692406692e-05,
      "loss": 1.2881,
      "step": 1805
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.034968376159668,
      "learning_rate": 2.4189189189189188e-05,
      "loss": 1.1641,
      "step": 1806
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.836882591247559,
      "learning_rate": 2.4185971685971685e-05,
      "loss": 1.227,
      "step": 1807
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.980088233947754,
      "learning_rate": 2.4182754182754185e-05,
      "loss": 1.0424,
      "step": 1808
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.310578346252441,
      "learning_rate": 2.417953667953668e-05,
      "loss": 1.1922,
      "step": 1809
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.821576118469238,
      "learning_rate": 2.4176319176319177e-05,
      "loss": 1.2869,
      "step": 1810
    },
    {
      "epoch": 0.39,
      "grad_norm": 10.617518424987793,
      "learning_rate": 2.4173101673101674e-05,
      "loss": 1.4866,
      "step": 1811
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.358549118041992,
      "learning_rate": 2.416988416988417e-05,
      "loss": 1.4701,
      "step": 1812
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.653206825256348,
      "learning_rate": 2.4166666666666667e-05,
      "loss": 0.994,
      "step": 1813
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.61852741241455,
      "learning_rate": 2.4163449163449167e-05,
      "loss": 1.1491,
      "step": 1814
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.242582321166992,
      "learning_rate": 2.4160231660231663e-05,
      "loss": 1.3414,
      "step": 1815
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.157299995422363,
      "learning_rate": 2.415701415701416e-05,
      "loss": 1.0833,
      "step": 1816
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.66378116607666,
      "learning_rate": 2.4153796653796656e-05,
      "loss": 1.2954,
      "step": 1817
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.520899772644043,
      "learning_rate": 2.415057915057915e-05,
      "loss": 1.0684,
      "step": 1818
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.002314567565918,
      "learning_rate": 2.4147361647361646e-05,
      "loss": 1.0691,
      "step": 1819
    },
    {
      "epoch": 0.39,
      "grad_norm": 10.130359649658203,
      "learning_rate": 2.4144144144144142e-05,
      "loss": 1.1813,
      "step": 1820
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.921329498291016,
      "learning_rate": 2.4140926640926642e-05,
      "loss": 1.3534,
      "step": 1821
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.044438362121582,
      "learning_rate": 2.4137709137709138e-05,
      "loss": 1.0404,
      "step": 1822
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.955984115600586,
      "learning_rate": 2.4134491634491635e-05,
      "loss": 1.4497,
      "step": 1823
    },
    {
      "epoch": 0.39,
      "grad_norm": 10.191411018371582,
      "learning_rate": 2.413127413127413e-05,
      "loss": 1.4137,
      "step": 1824
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.461008071899414,
      "learning_rate": 2.4128056628056628e-05,
      "loss": 0.8684,
      "step": 1825
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.446280479431152,
      "learning_rate": 2.4124839124839124e-05,
      "loss": 1.1403,
      "step": 1826
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.6404948234558105,
      "learning_rate": 2.4121621621621624e-05,
      "loss": 0.9432,
      "step": 1827
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.77010440826416,
      "learning_rate": 2.411840411840412e-05,
      "loss": 1.2461,
      "step": 1828
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.152596473693848,
      "learning_rate": 2.4115186615186617e-05,
      "loss": 1.4641,
      "step": 1829
    },
    {
      "epoch": 0.39,
      "grad_norm": 10.403176307678223,
      "learning_rate": 2.4111969111969113e-05,
      "loss": 1.2575,
      "step": 1830
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.892999649047852,
      "learning_rate": 2.410875160875161e-05,
      "loss": 1.0481,
      "step": 1831
    },
    {
      "epoch": 0.39,
      "grad_norm": 11.27293872833252,
      "learning_rate": 2.4105534105534106e-05,
      "loss": 1.2092,
      "step": 1832
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.498039245605469,
      "learning_rate": 2.4102316602316603e-05,
      "loss": 0.9169,
      "step": 1833
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.313258171081543,
      "learning_rate": 2.4099099099099102e-05,
      "loss": 1.4903,
      "step": 1834
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.124764442443848,
      "learning_rate": 2.40958815958816e-05,
      "loss": 0.8501,
      "step": 1835
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.21234130859375,
      "learning_rate": 2.4092664092664092e-05,
      "loss": 1.1073,
      "step": 1836
    },
    {
      "epoch": 0.39,
      "grad_norm": 10.220108985900879,
      "learning_rate": 2.408944658944659e-05,
      "loss": 1.141,
      "step": 1837
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.966140270233154,
      "learning_rate": 2.4086229086229085e-05,
      "loss": 1.0725,
      "step": 1838
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.898067474365234,
      "learning_rate": 2.408301158301158e-05,
      "loss": 1.1784,
      "step": 1839
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.777458190917969,
      "learning_rate": 2.407979407979408e-05,
      "loss": 1.1009,
      "step": 1840
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.554200172424316,
      "learning_rate": 2.4076576576576578e-05,
      "loss": 1.1073,
      "step": 1841
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.798097133636475,
      "learning_rate": 2.4073359073359074e-05,
      "loss": 0.8411,
      "step": 1842
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.720010757446289,
      "learning_rate": 2.407014157014157e-05,
      "loss": 1.2686,
      "step": 1843
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.29702091217041,
      "learning_rate": 2.4066924066924067e-05,
      "loss": 1.0402,
      "step": 1844
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.316208839416504,
      "learning_rate": 2.4063706563706563e-05,
      "loss": 1.0216,
      "step": 1845
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.755396842956543,
      "learning_rate": 2.406048906048906e-05,
      "loss": 1.0972,
      "step": 1846
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.432315826416016,
      "learning_rate": 2.405727155727156e-05,
      "loss": 1.0695,
      "step": 1847
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.105237007141113,
      "learning_rate": 2.4054054054054056e-05,
      "loss": 1.2996,
      "step": 1848
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.964668273925781,
      "learning_rate": 2.4050836550836553e-05,
      "loss": 1.6012,
      "step": 1849
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.057160377502441,
      "learning_rate": 2.404761904761905e-05,
      "loss": 1.0736,
      "step": 1850
    },
    {
      "epoch": 0.4,
      "grad_norm": 10.118755340576172,
      "learning_rate": 2.4044401544401546e-05,
      "loss": 1.2493,
      "step": 1851
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.009882926940918,
      "learning_rate": 2.4041184041184042e-05,
      "loss": 1.0399,
      "step": 1852
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.846986293792725,
      "learning_rate": 2.403796653796654e-05,
      "loss": 1.1367,
      "step": 1853
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.112967491149902,
      "learning_rate": 2.4034749034749035e-05,
      "loss": 1.1686,
      "step": 1854
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.394120216369629,
      "learning_rate": 2.403153153153153e-05,
      "loss": 1.4521,
      "step": 1855
    },
    {
      "epoch": 0.4,
      "grad_norm": 6.924731731414795,
      "learning_rate": 2.4028314028314028e-05,
      "loss": 1.1404,
      "step": 1856
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.079012870788574,
      "learning_rate": 2.4025096525096524e-05,
      "loss": 1.0426,
      "step": 1857
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.635187149047852,
      "learning_rate": 2.402187902187902e-05,
      "loss": 1.2295,
      "step": 1858
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.826465606689453,
      "learning_rate": 2.401866151866152e-05,
      "loss": 1.1618,
      "step": 1859
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.44950532913208,
      "learning_rate": 2.4015444015444017e-05,
      "loss": 1.1929,
      "step": 1860
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.099780082702637,
      "learning_rate": 2.4012226512226513e-05,
      "loss": 1.0947,
      "step": 1861
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.016019821166992,
      "learning_rate": 2.400900900900901e-05,
      "loss": 1.4019,
      "step": 1862
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.005331993103027,
      "learning_rate": 2.4005791505791506e-05,
      "loss": 1.1807,
      "step": 1863
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.388418197631836,
      "learning_rate": 2.4002574002574003e-05,
      "loss": 1.2045,
      "step": 1864
    },
    {
      "epoch": 0.4,
      "grad_norm": 10.430246353149414,
      "learning_rate": 2.39993564993565e-05,
      "loss": 1.2632,
      "step": 1865
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.170281410217285,
      "learning_rate": 2.3996138996139e-05,
      "loss": 1.4045,
      "step": 1866
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.545838356018066,
      "learning_rate": 2.3992921492921496e-05,
      "loss": 1.1996,
      "step": 1867
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.446737289428711,
      "learning_rate": 2.3989703989703992e-05,
      "loss": 1.3174,
      "step": 1868
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.966959476470947,
      "learning_rate": 2.398648648648649e-05,
      "loss": 1.121,
      "step": 1869
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.607033729553223,
      "learning_rate": 2.3983268983268985e-05,
      "loss": 1.1472,
      "step": 1870
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.239842414855957,
      "learning_rate": 2.3980051480051478e-05,
      "loss": 1.1657,
      "step": 1871
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.711627006530762,
      "learning_rate": 2.3976833976833978e-05,
      "loss": 1.3799,
      "step": 1872
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.447305679321289,
      "learning_rate": 2.3973616473616474e-05,
      "loss": 1.2757,
      "step": 1873
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.443217277526855,
      "learning_rate": 2.397039897039897e-05,
      "loss": 1.1779,
      "step": 1874
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.381601333618164,
      "learning_rate": 2.3967181467181467e-05,
      "loss": 1.0403,
      "step": 1875
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.964205741882324,
      "learning_rate": 2.3963963963963964e-05,
      "loss": 1.4239,
      "step": 1876
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.44223403930664,
      "learning_rate": 2.396074646074646e-05,
      "loss": 1.0415,
      "step": 1877
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.5291109085083,
      "learning_rate": 2.3957528957528956e-05,
      "loss": 0.9388,
      "step": 1878
    },
    {
      "epoch": 0.4,
      "grad_norm": 10.329209327697754,
      "learning_rate": 2.3954311454311456e-05,
      "loss": 1.3785,
      "step": 1879
    },
    {
      "epoch": 0.4,
      "grad_norm": 10.701093673706055,
      "learning_rate": 2.3951093951093953e-05,
      "loss": 1.31,
      "step": 1880
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.848126411437988,
      "learning_rate": 2.394787644787645e-05,
      "loss": 1.073,
      "step": 1881
    },
    {
      "epoch": 0.4,
      "grad_norm": 10.261636734008789,
      "learning_rate": 2.3944658944658946e-05,
      "loss": 1.4646,
      "step": 1882
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.02904987335205,
      "learning_rate": 2.3941441441441442e-05,
      "loss": 1.1426,
      "step": 1883
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.222436904907227,
      "learning_rate": 2.393822393822394e-05,
      "loss": 0.8697,
      "step": 1884
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.982858657836914,
      "learning_rate": 2.393500643500644e-05,
      "loss": 1.1836,
      "step": 1885
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.46439266204834,
      "learning_rate": 2.3931788931788935e-05,
      "loss": 1.0607,
      "step": 1886
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.103071212768555,
      "learning_rate": 2.392857142857143e-05,
      "loss": 0.7063,
      "step": 1887
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.836350440979004,
      "learning_rate": 2.3925353925353924e-05,
      "loss": 1.4661,
      "step": 1888
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.434525489807129,
      "learning_rate": 2.392213642213642e-05,
      "loss": 1.0241,
      "step": 1889
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.985081672668457,
      "learning_rate": 2.3918918918918917e-05,
      "loss": 1.3446,
      "step": 1890
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.051085472106934,
      "learning_rate": 2.3915701415701414e-05,
      "loss": 1.0412,
      "step": 1891
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.49036693572998,
      "learning_rate": 2.3912483912483914e-05,
      "loss": 1.1874,
      "step": 1892
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.821797370910645,
      "learning_rate": 2.390926640926641e-05,
      "loss": 1.4048,
      "step": 1893
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.940635681152344,
      "learning_rate": 2.3906048906048906e-05,
      "loss": 1.3366,
      "step": 1894
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.327083587646484,
      "learning_rate": 2.3902831402831403e-05,
      "loss": 1.186,
      "step": 1895
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.323933601379395,
      "learning_rate": 2.38996138996139e-05,
      "loss": 1.1573,
      "step": 1896
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.26744270324707,
      "learning_rate": 2.3896396396396396e-05,
      "loss": 1.0996,
      "step": 1897
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.952513694763184,
      "learning_rate": 2.3893178893178896e-05,
      "loss": 1.1492,
      "step": 1898
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.697195053100586,
      "learning_rate": 2.3889961389961392e-05,
      "loss": 1.2419,
      "step": 1899
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.233199119567871,
      "learning_rate": 2.388674388674389e-05,
      "loss": 1.3019,
      "step": 1900
    },
    {
      "epoch": 0.41,
      "grad_norm": 10.63820743560791,
      "learning_rate": 2.3883526383526385e-05,
      "loss": 1.0589,
      "step": 1901
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.549599647521973,
      "learning_rate": 2.388030888030888e-05,
      "loss": 1.2764,
      "step": 1902
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.717869758605957,
      "learning_rate": 2.3877091377091378e-05,
      "loss": 0.9668,
      "step": 1903
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.787907123565674,
      "learning_rate": 2.3873873873873874e-05,
      "loss": 1.0019,
      "step": 1904
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.245083808898926,
      "learning_rate": 2.387065637065637e-05,
      "loss": 0.9596,
      "step": 1905
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.853184700012207,
      "learning_rate": 2.3867438867438867e-05,
      "loss": 1.2489,
      "step": 1906
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.085626602172852,
      "learning_rate": 2.3864221364221364e-05,
      "loss": 1.0585,
      "step": 1907
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.154319763183594,
      "learning_rate": 2.386100386100386e-05,
      "loss": 0.9251,
      "step": 1908
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.606108665466309,
      "learning_rate": 2.3857786357786357e-05,
      "loss": 0.9928,
      "step": 1909
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.684764862060547,
      "learning_rate": 2.3854568854568853e-05,
      "loss": 1.2192,
      "step": 1910
    },
    {
      "epoch": 0.41,
      "grad_norm": 10.471686363220215,
      "learning_rate": 2.3851351351351353e-05,
      "loss": 1.1991,
      "step": 1911
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.27008056640625,
      "learning_rate": 2.384813384813385e-05,
      "loss": 1.1143,
      "step": 1912
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.38083553314209,
      "learning_rate": 2.3844916344916346e-05,
      "loss": 1.222,
      "step": 1913
    },
    {
      "epoch": 0.41,
      "grad_norm": 10.738534927368164,
      "learning_rate": 2.3841698841698842e-05,
      "loss": 1.174,
      "step": 1914
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.40157413482666,
      "learning_rate": 2.383848133848134e-05,
      "loss": 0.9647,
      "step": 1915
    },
    {
      "epoch": 0.41,
      "grad_norm": 10.912943840026855,
      "learning_rate": 2.3835263835263835e-05,
      "loss": 1.3584,
      "step": 1916
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.587828636169434,
      "learning_rate": 2.383204633204633e-05,
      "loss": 0.8176,
      "step": 1917
    },
    {
      "epoch": 0.41,
      "grad_norm": 10.198351860046387,
      "learning_rate": 2.382882882882883e-05,
      "loss": 1.1505,
      "step": 1918
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.960567474365234,
      "learning_rate": 2.3825611325611328e-05,
      "loss": 1.2783,
      "step": 1919
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.857816696166992,
      "learning_rate": 2.3822393822393824e-05,
      "loss": 1.1413,
      "step": 1920
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.628227233886719,
      "learning_rate": 2.381917631917632e-05,
      "loss": 1.0131,
      "step": 1921
    },
    {
      "epoch": 0.41,
      "grad_norm": 10.142003059387207,
      "learning_rate": 2.3815958815958817e-05,
      "loss": 1.164,
      "step": 1922
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.914115905761719,
      "learning_rate": 2.381274131274131e-05,
      "loss": 1.3335,
      "step": 1923
    },
    {
      "epoch": 0.41,
      "grad_norm": 10.084278106689453,
      "learning_rate": 2.380952380952381e-05,
      "loss": 1.4698,
      "step": 1924
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.272198677062988,
      "learning_rate": 2.3806306306306307e-05,
      "loss": 1.1357,
      "step": 1925
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.746260166168213,
      "learning_rate": 2.3803088803088803e-05,
      "loss": 0.9911,
      "step": 1926
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.43463134765625,
      "learning_rate": 2.37998712998713e-05,
      "loss": 1.4502,
      "step": 1927
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.834444046020508,
      "learning_rate": 2.3796653796653796e-05,
      "loss": 1.1196,
      "step": 1928
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.766192436218262,
      "learning_rate": 2.3793436293436292e-05,
      "loss": 1.4075,
      "step": 1929
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.245318412780762,
      "learning_rate": 2.3790218790218792e-05,
      "loss": 1.3526,
      "step": 1930
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.638627052307129,
      "learning_rate": 2.378700128700129e-05,
      "loss": 1.2861,
      "step": 1931
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.6592254638671875,
      "learning_rate": 2.3783783783783785e-05,
      "loss": 1.0699,
      "step": 1932
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.267848014831543,
      "learning_rate": 2.378056628056628e-05,
      "loss": 1.1646,
      "step": 1933
    },
    {
      "epoch": 0.41,
      "grad_norm": 11.422981262207031,
      "learning_rate": 2.3777348777348778e-05,
      "loss": 1.142,
      "step": 1934
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.727532386779785,
      "learning_rate": 2.3774131274131275e-05,
      "loss": 1.2778,
      "step": 1935
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.647194862365723,
      "learning_rate": 2.377091377091377e-05,
      "loss": 0.9599,
      "step": 1936
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.447746276855469,
      "learning_rate": 2.376769626769627e-05,
      "loss": 1.1251,
      "step": 1937
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.505449295043945,
      "learning_rate": 2.3764478764478767e-05,
      "loss": 1.1717,
      "step": 1938
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.270505905151367,
      "learning_rate": 2.3761261261261264e-05,
      "loss": 1.2869,
      "step": 1939
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.71439266204834,
      "learning_rate": 2.3758043758043757e-05,
      "loss": 0.9562,
      "step": 1940
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.850190162658691,
      "learning_rate": 2.3754826254826253e-05,
      "loss": 0.9526,
      "step": 1941
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.157587051391602,
      "learning_rate": 2.375160875160875e-05,
      "loss": 0.9196,
      "step": 1942
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.876945495605469,
      "learning_rate": 2.374839124839125e-05,
      "loss": 1.1012,
      "step": 1943
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.269291877746582,
      "learning_rate": 2.3745173745173746e-05,
      "loss": 1.0794,
      "step": 1944
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.739211082458496,
      "learning_rate": 2.3741956241956242e-05,
      "loss": 0.8619,
      "step": 1945
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.510917663574219,
      "learning_rate": 2.373873873873874e-05,
      "loss": 1.1729,
      "step": 1946
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.741676330566406,
      "learning_rate": 2.3735521235521235e-05,
      "loss": 1.308,
      "step": 1947
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.958059310913086,
      "learning_rate": 2.3732303732303732e-05,
      "loss": 1.1223,
      "step": 1948
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.90079402923584,
      "learning_rate": 2.3729086229086228e-05,
      "loss": 1.1672,
      "step": 1949
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.318421363830566,
      "learning_rate": 2.3725868725868728e-05,
      "loss": 0.8196,
      "step": 1950
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.452302932739258,
      "learning_rate": 2.3722651222651225e-05,
      "loss": 1.289,
      "step": 1951
    },
    {
      "epoch": 0.42,
      "grad_norm": 11.043157577514648,
      "learning_rate": 2.371943371943372e-05,
      "loss": 1.1528,
      "step": 1952
    },
    {
      "epoch": 0.42,
      "grad_norm": 10.067887306213379,
      "learning_rate": 2.3716216216216217e-05,
      "loss": 1.1143,
      "step": 1953
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.658963203430176,
      "learning_rate": 2.3712998712998714e-05,
      "loss": 1.1807,
      "step": 1954
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.420164108276367,
      "learning_rate": 2.370978120978121e-05,
      "loss": 0.9549,
      "step": 1955
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.54575252532959,
      "learning_rate": 2.370656370656371e-05,
      "loss": 1.1733,
      "step": 1956
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.091629028320312,
      "learning_rate": 2.3703346203346203e-05,
      "loss": 1.1847,
      "step": 1957
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.894845962524414,
      "learning_rate": 2.37001287001287e-05,
      "loss": 1.1828,
      "step": 1958
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.657892227172852,
      "learning_rate": 2.3696911196911196e-05,
      "loss": 1.1448,
      "step": 1959
    },
    {
      "epoch": 0.42,
      "grad_norm": 10.492029190063477,
      "learning_rate": 2.3693693693693693e-05,
      "loss": 1.4042,
      "step": 1960
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.264330863952637,
      "learning_rate": 2.369047619047619e-05,
      "loss": 1.1127,
      "step": 1961
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.894663333892822,
      "learning_rate": 2.3687258687258685e-05,
      "loss": 1.0215,
      "step": 1962
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.4409356117248535,
      "learning_rate": 2.3684041184041185e-05,
      "loss": 1.0573,
      "step": 1963
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.561649322509766,
      "learning_rate": 2.3680823680823682e-05,
      "loss": 1.2541,
      "step": 1964
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.359405517578125,
      "learning_rate": 2.3677606177606178e-05,
      "loss": 1.34,
      "step": 1965
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.877214431762695,
      "learning_rate": 2.3674388674388675e-05,
      "loss": 1.0981,
      "step": 1966
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.430760383605957,
      "learning_rate": 2.367117117117117e-05,
      "loss": 0.96,
      "step": 1967
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.846302032470703,
      "learning_rate": 2.3667953667953668e-05,
      "loss": 1.168,
      "step": 1968
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.50442123413086,
      "learning_rate": 2.3664736164736167e-05,
      "loss": 1.1447,
      "step": 1969
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.567601203918457,
      "learning_rate": 2.3661518661518664e-05,
      "loss": 0.8315,
      "step": 1970
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.075531959533691,
      "learning_rate": 2.365830115830116e-05,
      "loss": 1.1806,
      "step": 1971
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.416583061218262,
      "learning_rate": 2.3655083655083657e-05,
      "loss": 1.4688,
      "step": 1972
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.402812004089355,
      "learning_rate": 2.3651866151866153e-05,
      "loss": 1.2346,
      "step": 1973
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.505352973937988,
      "learning_rate": 2.364864864864865e-05,
      "loss": 1.1589,
      "step": 1974
    },
    {
      "epoch": 0.42,
      "grad_norm": 10.363417625427246,
      "learning_rate": 2.3645431145431146e-05,
      "loss": 1.258,
      "step": 1975
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.157763481140137,
      "learning_rate": 2.3642213642213643e-05,
      "loss": 0.8421,
      "step": 1976
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.634543418884277,
      "learning_rate": 2.363899613899614e-05,
      "loss": 1.3997,
      "step": 1977
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.300816535949707,
      "learning_rate": 2.3635778635778635e-05,
      "loss": 1.1098,
      "step": 1978
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.161367416381836,
      "learning_rate": 2.3632561132561132e-05,
      "loss": 1.215,
      "step": 1979
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.591649055480957,
      "learning_rate": 2.362934362934363e-05,
      "loss": 1.3698,
      "step": 1980
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.839532852172852,
      "learning_rate": 2.3626126126126125e-05,
      "loss": 1.2965,
      "step": 1981
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.254809379577637,
      "learning_rate": 2.3622908622908625e-05,
      "loss": 1.4365,
      "step": 1982
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.499753952026367,
      "learning_rate": 2.361969111969112e-05,
      "loss": 1.3798,
      "step": 1983
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.479209899902344,
      "learning_rate": 2.3616473616473618e-05,
      "loss": 1.0594,
      "step": 1984
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.223916053771973,
      "learning_rate": 2.3613256113256114e-05,
      "loss": 1.266,
      "step": 1985
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.021401405334473,
      "learning_rate": 2.361003861003861e-05,
      "loss": 1.0216,
      "step": 1986
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.691368103027344,
      "learning_rate": 2.3606821106821107e-05,
      "loss": 1.2902,
      "step": 1987
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.994230270385742,
      "learning_rate": 2.3603603603603607e-05,
      "loss": 1.2745,
      "step": 1988
    },
    {
      "epoch": 0.43,
      "grad_norm": 7.824987888336182,
      "learning_rate": 2.3600386100386103e-05,
      "loss": 1.1039,
      "step": 1989
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.31676959991455,
      "learning_rate": 2.35971685971686e-05,
      "loss": 1.1101,
      "step": 1990
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.266079902648926,
      "learning_rate": 2.3593951093951096e-05,
      "loss": 1.1325,
      "step": 1991
    },
    {
      "epoch": 0.43,
      "grad_norm": 10.203649520874023,
      "learning_rate": 2.359073359073359e-05,
      "loss": 1.2871,
      "step": 1992
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.54558277130127,
      "learning_rate": 2.3587516087516086e-05,
      "loss": 1.2207,
      "step": 1993
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.112969398498535,
      "learning_rate": 2.3584298584298582e-05,
      "loss": 1.2752,
      "step": 1994
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.690529823303223,
      "learning_rate": 2.3581081081081082e-05,
      "loss": 1.04,
      "step": 1995
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.390679359436035,
      "learning_rate": 2.357786357786358e-05,
      "loss": 1.0427,
      "step": 1996
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.233559608459473,
      "learning_rate": 2.3574646074646075e-05,
      "loss": 1.3051,
      "step": 1997
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.014806747436523,
      "learning_rate": 2.357142857142857e-05,
      "loss": 1.2988,
      "step": 1998
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.532296180725098,
      "learning_rate": 2.3568211068211068e-05,
      "loss": 0.9593,
      "step": 1999
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.703890800476074,
      "learning_rate": 2.3564993564993564e-05,
      "loss": 0.9363,
      "step": 2000
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.366865158081055,
      "learning_rate": 2.3561776061776064e-05,
      "loss": 1.1999,
      "step": 2001
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.625,
      "learning_rate": 2.355855855855856e-05,
      "loss": 0.971,
      "step": 2002
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.750784873962402,
      "learning_rate": 2.3555341055341057e-05,
      "loss": 1.1905,
      "step": 2003
    },
    {
      "epoch": 0.43,
      "grad_norm": 10.148306846618652,
      "learning_rate": 2.3552123552123553e-05,
      "loss": 1.2382,
      "step": 2004
    },
    {
      "epoch": 0.43,
      "grad_norm": 7.465278625488281,
      "learning_rate": 2.354890604890605e-05,
      "loss": 0.7074,
      "step": 2005
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.666796684265137,
      "learning_rate": 2.3545688545688546e-05,
      "loss": 1.2006,
      "step": 2006
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.540093421936035,
      "learning_rate": 2.3542471042471043e-05,
      "loss": 1.1005,
      "step": 2007
    },
    {
      "epoch": 0.43,
      "grad_norm": 12.908550262451172,
      "learning_rate": 2.3539253539253543e-05,
      "loss": 1.2596,
      "step": 2008
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.50829792022705,
      "learning_rate": 2.353603603603604e-05,
      "loss": 1.1244,
      "step": 2009
    },
    {
      "epoch": 0.43,
      "grad_norm": 10.139119148254395,
      "learning_rate": 2.3532818532818532e-05,
      "loss": 1.2163,
      "step": 2010
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.853997230529785,
      "learning_rate": 2.352960102960103e-05,
      "loss": 1.2222,
      "step": 2011
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.257988929748535,
      "learning_rate": 2.3526383526383525e-05,
      "loss": 1.226,
      "step": 2012
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.278950691223145,
      "learning_rate": 2.352316602316602e-05,
      "loss": 1.1131,
      "step": 2013
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.143329620361328,
      "learning_rate": 2.351994851994852e-05,
      "loss": 1.1607,
      "step": 2014
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.703496932983398,
      "learning_rate": 2.3516731016731018e-05,
      "loss": 1.0345,
      "step": 2015
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.774778366088867,
      "learning_rate": 2.3513513513513514e-05,
      "loss": 1.3838,
      "step": 2016
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.816611289978027,
      "learning_rate": 2.351029601029601e-05,
      "loss": 1.2172,
      "step": 2017
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.284188270568848,
      "learning_rate": 2.3507078507078507e-05,
      "loss": 1.28,
      "step": 2018
    },
    {
      "epoch": 0.43,
      "grad_norm": 10.197473526000977,
      "learning_rate": 2.3503861003861004e-05,
      "loss": 1.0437,
      "step": 2019
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.228413581848145,
      "learning_rate": 2.35006435006435e-05,
      "loss": 1.2769,
      "step": 2020
    },
    {
      "epoch": 0.43,
      "grad_norm": 7.602457523345947,
      "learning_rate": 2.3497425997426e-05,
      "loss": 1.0007,
      "step": 2021
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.958240509033203,
      "learning_rate": 2.3494208494208496e-05,
      "loss": 1.3141,
      "step": 2022
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.676724433898926,
      "learning_rate": 2.3490990990990993e-05,
      "loss": 1.1266,
      "step": 2023
    },
    {
      "epoch": 0.43,
      "grad_norm": 7.104589939117432,
      "learning_rate": 2.348777348777349e-05,
      "loss": 1.0596,
      "step": 2024
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.447674751281738,
      "learning_rate": 2.3484555984555986e-05,
      "loss": 1.0173,
      "step": 2025
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.29539680480957,
      "learning_rate": 2.3481338481338482e-05,
      "loss": 1.3261,
      "step": 2026
    },
    {
      "epoch": 0.43,
      "grad_norm": 10.198798179626465,
      "learning_rate": 2.347812097812098e-05,
      "loss": 1.498,
      "step": 2027
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.254582405090332,
      "learning_rate": 2.3474903474903475e-05,
      "loss": 1.0057,
      "step": 2028
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.686783790588379,
      "learning_rate": 2.347168597168597e-05,
      "loss": 0.959,
      "step": 2029
    },
    {
      "epoch": 0.44,
      "grad_norm": 7.7781829833984375,
      "learning_rate": 2.3468468468468468e-05,
      "loss": 1.1467,
      "step": 2030
    },
    {
      "epoch": 0.44,
      "grad_norm": 7.765499114990234,
      "learning_rate": 2.3465250965250964e-05,
      "loss": 0.8797,
      "step": 2031
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.630419731140137,
      "learning_rate": 2.346203346203346e-05,
      "loss": 1.1894,
      "step": 2032
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.005719184875488,
      "learning_rate": 2.345881595881596e-05,
      "loss": 0.983,
      "step": 2033
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.476431846618652,
      "learning_rate": 2.3455598455598457e-05,
      "loss": 1.0059,
      "step": 2034
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.6655855178833,
      "learning_rate": 2.3452380952380954e-05,
      "loss": 1.322,
      "step": 2035
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.724005699157715,
      "learning_rate": 2.344916344916345e-05,
      "loss": 1.2993,
      "step": 2036
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.069770812988281,
      "learning_rate": 2.3445945945945946e-05,
      "loss": 0.9476,
      "step": 2037
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.238032341003418,
      "learning_rate": 2.3442728442728443e-05,
      "loss": 1.126,
      "step": 2038
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.885123252868652,
      "learning_rate": 2.343951093951094e-05,
      "loss": 1.4858,
      "step": 2039
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.9751558303833,
      "learning_rate": 2.343629343629344e-05,
      "loss": 1.39,
      "step": 2040
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.268264770507812,
      "learning_rate": 2.3433075933075936e-05,
      "loss": 1.1368,
      "step": 2041
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.425433158874512,
      "learning_rate": 2.3429858429858432e-05,
      "loss": 1.0008,
      "step": 2042
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.727019309997559,
      "learning_rate": 2.342664092664093e-05,
      "loss": 0.8686,
      "step": 2043
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.784852027893066,
      "learning_rate": 2.342342342342342e-05,
      "loss": 1.4122,
      "step": 2044
    },
    {
      "epoch": 0.44,
      "grad_norm": 7.851362228393555,
      "learning_rate": 2.3420205920205918e-05,
      "loss": 1.1032,
      "step": 2045
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.247027397155762,
      "learning_rate": 2.3416988416988418e-05,
      "loss": 1.152,
      "step": 2046
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.461833953857422,
      "learning_rate": 2.3413770913770914e-05,
      "loss": 1.2015,
      "step": 2047
    },
    {
      "epoch": 0.44,
      "grad_norm": 7.990260124206543,
      "learning_rate": 2.341055341055341e-05,
      "loss": 1.0997,
      "step": 2048
    },
    {
      "epoch": 0.44,
      "grad_norm": 7.8678436279296875,
      "learning_rate": 2.3407335907335907e-05,
      "loss": 0.9696,
      "step": 2049
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.650738716125488,
      "learning_rate": 2.3404118404118404e-05,
      "loss": 1.0136,
      "step": 2050
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.997081756591797,
      "learning_rate": 2.34009009009009e-05,
      "loss": 1.2903,
      "step": 2051
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.791197776794434,
      "learning_rate": 2.3397683397683397e-05,
      "loss": 1.2244,
      "step": 2052
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.12528133392334,
      "learning_rate": 2.3394465894465896e-05,
      "loss": 0.9615,
      "step": 2053
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.252815246582031,
      "learning_rate": 2.3391248391248393e-05,
      "loss": 1.2191,
      "step": 2054
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.783193588256836,
      "learning_rate": 2.338803088803089e-05,
      "loss": 1.3059,
      "step": 2055
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.474292755126953,
      "learning_rate": 2.3384813384813386e-05,
      "loss": 1.0195,
      "step": 2056
    },
    {
      "epoch": 0.44,
      "grad_norm": 7.443251132965088,
      "learning_rate": 2.3381595881595882e-05,
      "loss": 1.0728,
      "step": 2057
    },
    {
      "epoch": 0.44,
      "grad_norm": 7.984476089477539,
      "learning_rate": 2.337837837837838e-05,
      "loss": 1.0349,
      "step": 2058
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.6902494430542,
      "learning_rate": 2.337516087516088e-05,
      "loss": 1.3638,
      "step": 2059
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.544766426086426,
      "learning_rate": 2.3371943371943375e-05,
      "loss": 1.1213,
      "step": 2060
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.608477592468262,
      "learning_rate": 2.336872586872587e-05,
      "loss": 1.0159,
      "step": 2061
    },
    {
      "epoch": 0.44,
      "grad_norm": 10.170371055603027,
      "learning_rate": 2.3365508365508365e-05,
      "loss": 1.332,
      "step": 2062
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.555222511291504,
      "learning_rate": 2.336229086229086e-05,
      "loss": 1.4711,
      "step": 2063
    },
    {
      "epoch": 0.44,
      "grad_norm": 10.28255558013916,
      "learning_rate": 2.3359073359073357e-05,
      "loss": 1.1118,
      "step": 2064
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.57934284210205,
      "learning_rate": 2.3355855855855854e-05,
      "loss": 0.9118,
      "step": 2065
    },
    {
      "epoch": 0.44,
      "grad_norm": 7.886806488037109,
      "learning_rate": 2.3352638352638354e-05,
      "loss": 1.1827,
      "step": 2066
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.015555381774902,
      "learning_rate": 2.334942084942085e-05,
      "loss": 1.1059,
      "step": 2067
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.894145011901855,
      "learning_rate": 2.3346203346203347e-05,
      "loss": 1.0078,
      "step": 2068
    },
    {
      "epoch": 0.44,
      "grad_norm": 7.2238993644714355,
      "learning_rate": 2.3342985842985843e-05,
      "loss": 0.9497,
      "step": 2069
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.23896598815918,
      "learning_rate": 2.333976833976834e-05,
      "loss": 1.3213,
      "step": 2070
    },
    {
      "epoch": 0.44,
      "grad_norm": 11.6183500289917,
      "learning_rate": 2.3336550836550836e-05,
      "loss": 0.966,
      "step": 2071
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.670061111450195,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 1.3798,
      "step": 2072
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.361348152160645,
      "learning_rate": 2.3330115830115832e-05,
      "loss": 1.0835,
      "step": 2073
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.417104721069336,
      "learning_rate": 2.332689832689833e-05,
      "loss": 1.482,
      "step": 2074
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.006865501403809,
      "learning_rate": 2.3323680823680825e-05,
      "loss": 1.3127,
      "step": 2075
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.197654724121094,
      "learning_rate": 2.332046332046332e-05,
      "loss": 1.2345,
      "step": 2076
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.640593528747559,
      "learning_rate": 2.3317245817245818e-05,
      "loss": 1.3933,
      "step": 2077
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.206977844238281,
      "learning_rate": 2.3314028314028315e-05,
      "loss": 1.1359,
      "step": 2078
    },
    {
      "epoch": 0.45,
      "grad_norm": 7.280463218688965,
      "learning_rate": 2.331081081081081e-05,
      "loss": 0.9348,
      "step": 2079
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.68053913116455,
      "learning_rate": 2.3307593307593307e-05,
      "loss": 1.1662,
      "step": 2080
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.434091567993164,
      "learning_rate": 2.3304375804375804e-05,
      "loss": 0.9188,
      "step": 2081
    },
    {
      "epoch": 0.45,
      "grad_norm": 7.240901947021484,
      "learning_rate": 2.33011583011583e-05,
      "loss": 0.9153,
      "step": 2082
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.640982627868652,
      "learning_rate": 2.3297940797940797e-05,
      "loss": 1.1924,
      "step": 2083
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.094407081604004,
      "learning_rate": 2.3294723294723293e-05,
      "loss": 1.2584,
      "step": 2084
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.818005561828613,
      "learning_rate": 2.3291505791505793e-05,
      "loss": 1.1814,
      "step": 2085
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.51609992980957,
      "learning_rate": 2.328828828828829e-05,
      "loss": 0.8902,
      "step": 2086
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.912453651428223,
      "learning_rate": 2.3285070785070786e-05,
      "loss": 0.9406,
      "step": 2087
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.295294761657715,
      "learning_rate": 2.3281853281853282e-05,
      "loss": 0.9051,
      "step": 2088
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.691771507263184,
      "learning_rate": 2.327863577863578e-05,
      "loss": 0.9621,
      "step": 2089
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.694781303405762,
      "learning_rate": 2.3275418275418275e-05,
      "loss": 1.1081,
      "step": 2090
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.479084014892578,
      "learning_rate": 2.3272200772200775e-05,
      "loss": 1.4052,
      "step": 2091
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.788583755493164,
      "learning_rate": 2.326898326898327e-05,
      "loss": 1.2136,
      "step": 2092
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.443416595458984,
      "learning_rate": 2.3265765765765768e-05,
      "loss": 0.9704,
      "step": 2093
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.144128799438477,
      "learning_rate": 2.3262548262548265e-05,
      "loss": 1.2686,
      "step": 2094
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.922808647155762,
      "learning_rate": 2.325933075933076e-05,
      "loss": 0.9527,
      "step": 2095
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.42574691772461,
      "learning_rate": 2.3256113256113257e-05,
      "loss": 1.0119,
      "step": 2096
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.023568153381348,
      "learning_rate": 2.325289575289575e-05,
      "loss": 0.8009,
      "step": 2097
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.188454627990723,
      "learning_rate": 2.324967824967825e-05,
      "loss": 0.9412,
      "step": 2098
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.082704544067383,
      "learning_rate": 2.3246460746460747e-05,
      "loss": 0.9739,
      "step": 2099
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.612271308898926,
      "learning_rate": 2.3243243243243243e-05,
      "loss": 1.1708,
      "step": 2100
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.018346786499023,
      "learning_rate": 2.324002574002574e-05,
      "loss": 1.0888,
      "step": 2101
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.10799789428711,
      "learning_rate": 2.3236808236808236e-05,
      "loss": 1.1597,
      "step": 2102
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.869068145751953,
      "learning_rate": 2.3233590733590733e-05,
      "loss": 1.1769,
      "step": 2103
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.562080383300781,
      "learning_rate": 2.3230373230373232e-05,
      "loss": 1.1322,
      "step": 2104
    },
    {
      "epoch": 0.45,
      "grad_norm": 7.963632583618164,
      "learning_rate": 2.322715572715573e-05,
      "loss": 1.055,
      "step": 2105
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.818220138549805,
      "learning_rate": 2.3223938223938225e-05,
      "loss": 1.0374,
      "step": 2106
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.287997245788574,
      "learning_rate": 2.3220720720720722e-05,
      "loss": 1.0558,
      "step": 2107
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.248164176940918,
      "learning_rate": 2.3217503217503218e-05,
      "loss": 1.0454,
      "step": 2108
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.070637702941895,
      "learning_rate": 2.3214285714285715e-05,
      "loss": 1.0888,
      "step": 2109
    },
    {
      "epoch": 0.45,
      "grad_norm": 7.375611305236816,
      "learning_rate": 2.321106821106821e-05,
      "loss": 0.8334,
      "step": 2110
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.023420333862305,
      "learning_rate": 2.320785070785071e-05,
      "loss": 1.0789,
      "step": 2111
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.499464988708496,
      "learning_rate": 2.3204633204633207e-05,
      "loss": 1.0798,
      "step": 2112
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.41149616241455,
      "learning_rate": 2.3201415701415704e-05,
      "loss": 1.3563,
      "step": 2113
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.07431697845459,
      "learning_rate": 2.3198198198198197e-05,
      "loss": 1.1428,
      "step": 2114
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.418291091918945,
      "learning_rate": 2.3194980694980693e-05,
      "loss": 1.0707,
      "step": 2115
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.515170097351074,
      "learning_rate": 2.319176319176319e-05,
      "loss": 1.255,
      "step": 2116
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.06890869140625,
      "learning_rate": 2.318854568854569e-05,
      "loss": 1.0606,
      "step": 2117
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.90317153930664,
      "learning_rate": 2.3185328185328186e-05,
      "loss": 1.3626,
      "step": 2118
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.70065975189209,
      "learning_rate": 2.3182110682110683e-05,
      "loss": 1.2801,
      "step": 2119
    },
    {
      "epoch": 0.45,
      "grad_norm": 7.908420085906982,
      "learning_rate": 2.317889317889318e-05,
      "loss": 1.0526,
      "step": 2120
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.363370895385742,
      "learning_rate": 2.3175675675675675e-05,
      "loss": 1.0961,
      "step": 2121
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.235464096069336,
      "learning_rate": 2.3172458172458172e-05,
      "loss": 1.0702,
      "step": 2122
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.397433280944824,
      "learning_rate": 2.316924066924067e-05,
      "loss": 1.0306,
      "step": 2123
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.00225830078125,
      "learning_rate": 2.3166023166023168e-05,
      "loss": 1.1496,
      "step": 2124
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.060752868652344,
      "learning_rate": 2.3162805662805665e-05,
      "loss": 1.1662,
      "step": 2125
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.02695369720459,
      "learning_rate": 2.315958815958816e-05,
      "loss": 1.1765,
      "step": 2126
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.073675155639648,
      "learning_rate": 2.3156370656370658e-05,
      "loss": 1.2501,
      "step": 2127
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.150376319885254,
      "learning_rate": 2.3153153153153154e-05,
      "loss": 1.1951,
      "step": 2128
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.214688301086426,
      "learning_rate": 2.314993564993565e-05,
      "loss": 1.4866,
      "step": 2129
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.1207275390625,
      "learning_rate": 2.314671814671815e-05,
      "loss": 1.0156,
      "step": 2130
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.972799301147461,
      "learning_rate": 2.3143500643500643e-05,
      "loss": 0.9371,
      "step": 2131
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.436513900756836,
      "learning_rate": 2.314028314028314e-05,
      "loss": 1.1786,
      "step": 2132
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.132317543029785,
      "learning_rate": 2.3137065637065636e-05,
      "loss": 1.1757,
      "step": 2133
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.795341491699219,
      "learning_rate": 2.3133848133848133e-05,
      "loss": 1.5574,
      "step": 2134
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.895837783813477,
      "learning_rate": 2.313063063063063e-05,
      "loss": 1.2501,
      "step": 2135
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.055074691772461,
      "learning_rate": 2.3127413127413126e-05,
      "loss": 1.2691,
      "step": 2136
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.627052307128906,
      "learning_rate": 2.3124195624195625e-05,
      "loss": 1.1788,
      "step": 2137
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.494678497314453,
      "learning_rate": 2.3120978120978122e-05,
      "loss": 1.3345,
      "step": 2138
    },
    {
      "epoch": 0.46,
      "grad_norm": 11.606122016906738,
      "learning_rate": 2.311776061776062e-05,
      "loss": 1.3956,
      "step": 2139
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.524413108825684,
      "learning_rate": 2.3114543114543115e-05,
      "loss": 1.0574,
      "step": 2140
    },
    {
      "epoch": 0.46,
      "grad_norm": 12.123868942260742,
      "learning_rate": 2.311132561132561e-05,
      "loss": 1.4337,
      "step": 2141
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.218252182006836,
      "learning_rate": 2.3108108108108108e-05,
      "loss": 1.0512,
      "step": 2142
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.12955093383789,
      "learning_rate": 2.3104890604890608e-05,
      "loss": 1.1908,
      "step": 2143
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.47974967956543,
      "learning_rate": 2.3101673101673104e-05,
      "loss": 0.9904,
      "step": 2144
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.145479202270508,
      "learning_rate": 2.30984555984556e-05,
      "loss": 1.0297,
      "step": 2145
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.817002296447754,
      "learning_rate": 2.3095238095238097e-05,
      "loss": 1.1566,
      "step": 2146
    },
    {
      "epoch": 0.46,
      "grad_norm": 11.323281288146973,
      "learning_rate": 2.3092020592020593e-05,
      "loss": 1.2688,
      "step": 2147
    },
    {
      "epoch": 0.46,
      "grad_norm": 11.450321197509766,
      "learning_rate": 2.308880308880309e-05,
      "loss": 1.3723,
      "step": 2148
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.838591575622559,
      "learning_rate": 2.3085585585585586e-05,
      "loss": 1.2023,
      "step": 2149
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.694581985473633,
      "learning_rate": 2.3082368082368083e-05,
      "loss": 1.1467,
      "step": 2150
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.987435340881348,
      "learning_rate": 2.307915057915058e-05,
      "loss": 1.3137,
      "step": 2151
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.712952613830566,
      "learning_rate": 2.3075933075933076e-05,
      "loss": 1.3023,
      "step": 2152
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.019118309020996,
      "learning_rate": 2.3072715572715572e-05,
      "loss": 0.9481,
      "step": 2153
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.38272762298584,
      "learning_rate": 2.306949806949807e-05,
      "loss": 1.1535,
      "step": 2154
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.569527626037598,
      "learning_rate": 2.3066280566280565e-05,
      "loss": 0.9142,
      "step": 2155
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.906271934509277,
      "learning_rate": 2.3063063063063065e-05,
      "loss": 1.2523,
      "step": 2156
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.27064323425293,
      "learning_rate": 2.305984555984556e-05,
      "loss": 1.3501,
      "step": 2157
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.297137260437012,
      "learning_rate": 2.3056628056628058e-05,
      "loss": 1.1943,
      "step": 2158
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.9092936515808105,
      "learning_rate": 2.3053410553410554e-05,
      "loss": 0.9223,
      "step": 2159
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.922956466674805,
      "learning_rate": 2.305019305019305e-05,
      "loss": 0.9506,
      "step": 2160
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.95810317993164,
      "learning_rate": 2.3046975546975547e-05,
      "loss": 1.2933,
      "step": 2161
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.038839340209961,
      "learning_rate": 2.3043758043758047e-05,
      "loss": 1.1373,
      "step": 2162
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.429363250732422,
      "learning_rate": 2.3040540540540543e-05,
      "loss": 1.1149,
      "step": 2163
    },
    {
      "epoch": 0.46,
      "grad_norm": 11.094080924987793,
      "learning_rate": 2.303732303732304e-05,
      "loss": 1.4419,
      "step": 2164
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.00954818725586,
      "learning_rate": 2.3034105534105536e-05,
      "loss": 1.2827,
      "step": 2165
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.380525588989258,
      "learning_rate": 2.303088803088803e-05,
      "loss": 1.451,
      "step": 2166
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.347018241882324,
      "learning_rate": 2.3027670527670526e-05,
      "loss": 1.2192,
      "step": 2167
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.168380737304688,
      "learning_rate": 2.3024453024453022e-05,
      "loss": 1.0522,
      "step": 2168
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.851546764373779,
      "learning_rate": 2.3021235521235522e-05,
      "loss": 1.1687,
      "step": 2169
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.136293411254883,
      "learning_rate": 2.301801801801802e-05,
      "loss": 1.1137,
      "step": 2170
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.926255226135254,
      "learning_rate": 2.3014800514800515e-05,
      "loss": 1.2098,
      "step": 2171
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.527799606323242,
      "learning_rate": 2.301158301158301e-05,
      "loss": 1.043,
      "step": 2172
    },
    {
      "epoch": 0.47,
      "grad_norm": 10.596610069274902,
      "learning_rate": 2.3008365508365508e-05,
      "loss": 1.4153,
      "step": 2173
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.496762275695801,
      "learning_rate": 2.3005148005148004e-05,
      "loss": 0.9325,
      "step": 2174
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.681316375732422,
      "learning_rate": 2.3001930501930504e-05,
      "loss": 1.0793,
      "step": 2175
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.082897186279297,
      "learning_rate": 2.2998712998713e-05,
      "loss": 1.1868,
      "step": 2176
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.65053939819336,
      "learning_rate": 2.2995495495495497e-05,
      "loss": 1.0042,
      "step": 2177
    },
    {
      "epoch": 0.47,
      "grad_norm": 10.683305740356445,
      "learning_rate": 2.2992277992277994e-05,
      "loss": 1.1334,
      "step": 2178
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.557844161987305,
      "learning_rate": 2.298906048906049e-05,
      "loss": 1.3916,
      "step": 2179
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.335153579711914,
      "learning_rate": 2.2985842985842986e-05,
      "loss": 1.2917,
      "step": 2180
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.249327182769775,
      "learning_rate": 2.2982625482625483e-05,
      "loss": 0.8964,
      "step": 2181
    },
    {
      "epoch": 0.47,
      "grad_norm": 10.641585350036621,
      "learning_rate": 2.2979407979407983e-05,
      "loss": 1.1886,
      "step": 2182
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.318747520446777,
      "learning_rate": 2.297619047619048e-05,
      "loss": 1.261,
      "step": 2183
    },
    {
      "epoch": 0.47,
      "grad_norm": 10.120965957641602,
      "learning_rate": 2.2972972972972972e-05,
      "loss": 1.0521,
      "step": 2184
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.033830642700195,
      "learning_rate": 2.296975546975547e-05,
      "loss": 1.009,
      "step": 2185
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.813838958740234,
      "learning_rate": 2.2966537966537965e-05,
      "loss": 1.1616,
      "step": 2186
    },
    {
      "epoch": 0.47,
      "grad_norm": 10.738889694213867,
      "learning_rate": 2.296332046332046e-05,
      "loss": 0.8981,
      "step": 2187
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.467135429382324,
      "learning_rate": 2.296010296010296e-05,
      "loss": 1.176,
      "step": 2188
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.13377571105957,
      "learning_rate": 2.2956885456885458e-05,
      "loss": 1.0633,
      "step": 2189
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.785099029541016,
      "learning_rate": 2.2953667953667954e-05,
      "loss": 1.081,
      "step": 2190
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.333328247070312,
      "learning_rate": 2.295045045045045e-05,
      "loss": 1.2005,
      "step": 2191
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.945670127868652,
      "learning_rate": 2.2947232947232947e-05,
      "loss": 1.2383,
      "step": 2192
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.100946426391602,
      "learning_rate": 2.2944015444015444e-05,
      "loss": 0.8548,
      "step": 2193
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.848225593566895,
      "learning_rate": 2.294079794079794e-05,
      "loss": 1.2701,
      "step": 2194
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.367883682250977,
      "learning_rate": 2.293758043758044e-05,
      "loss": 1.3728,
      "step": 2195
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.251631736755371,
      "learning_rate": 2.2934362934362936e-05,
      "loss": 0.8512,
      "step": 2196
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.556583404541016,
      "learning_rate": 2.2931145431145433e-05,
      "loss": 1.1803,
      "step": 2197
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.614463806152344,
      "learning_rate": 2.292792792792793e-05,
      "loss": 1.0619,
      "step": 2198
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.684062957763672,
      "learning_rate": 2.2924710424710426e-05,
      "loss": 1.1992,
      "step": 2199
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.640198707580566,
      "learning_rate": 2.2921492921492922e-05,
      "loss": 1.1942,
      "step": 2200
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.844145774841309,
      "learning_rate": 2.291827541827542e-05,
      "loss": 1.4447,
      "step": 2201
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.62072467803955,
      "learning_rate": 2.2915057915057915e-05,
      "loss": 0.9158,
      "step": 2202
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.574515342712402,
      "learning_rate": 2.291184041184041e-05,
      "loss": 1.1593,
      "step": 2203
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.745797157287598,
      "learning_rate": 2.2908622908622908e-05,
      "loss": 1.0202,
      "step": 2204
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.40080738067627,
      "learning_rate": 2.2905405405405404e-05,
      "loss": 1.2784,
      "step": 2205
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.047880172729492,
      "learning_rate": 2.29021879021879e-05,
      "loss": 0.7727,
      "step": 2206
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.820313453674316,
      "learning_rate": 2.28989703989704e-05,
      "loss": 1.2033,
      "step": 2207
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.547165870666504,
      "learning_rate": 2.2895752895752897e-05,
      "loss": 1.1206,
      "step": 2208
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.131793022155762,
      "learning_rate": 2.2892535392535394e-05,
      "loss": 0.9339,
      "step": 2209
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.79621410369873,
      "learning_rate": 2.288931788931789e-05,
      "loss": 1.1135,
      "step": 2210
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.914697647094727,
      "learning_rate": 2.2886100386100387e-05,
      "loss": 1.156,
      "step": 2211
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.336834907531738,
      "learning_rate": 2.2882882882882883e-05,
      "loss": 0.7323,
      "step": 2212
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.25586986541748,
      "learning_rate": 2.287966537966538e-05,
      "loss": 0.8996,
      "step": 2213
    },
    {
      "epoch": 0.47,
      "grad_norm": 10.740071296691895,
      "learning_rate": 2.287644787644788e-05,
      "loss": 1.1146,
      "step": 2214
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.79356861114502,
      "learning_rate": 2.2873230373230376e-05,
      "loss": 1.0225,
      "step": 2215
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.927678108215332,
      "learning_rate": 2.2870012870012872e-05,
      "loss": 1.348,
      "step": 2216
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.957770347595215,
      "learning_rate": 2.286679536679537e-05,
      "loss": 1.3489,
      "step": 2217
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.516385078430176,
      "learning_rate": 2.2863577863577862e-05,
      "loss": 1.4577,
      "step": 2218
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.18265151977539,
      "learning_rate": 2.2860360360360358e-05,
      "loss": 0.8654,
      "step": 2219
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.291986465454102,
      "learning_rate": 2.2857142857142858e-05,
      "loss": 1.1625,
      "step": 2220
    },
    {
      "epoch": 0.48,
      "grad_norm": 12.016526222229004,
      "learning_rate": 2.2853925353925354e-05,
      "loss": 1.27,
      "step": 2221
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.665726661682129,
      "learning_rate": 2.285070785070785e-05,
      "loss": 1.149,
      "step": 2222
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.282472610473633,
      "learning_rate": 2.2847490347490347e-05,
      "loss": 1.0851,
      "step": 2223
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.253392219543457,
      "learning_rate": 2.2844272844272844e-05,
      "loss": 1.0562,
      "step": 2224
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.6241455078125,
      "learning_rate": 2.284105534105534e-05,
      "loss": 1.2928,
      "step": 2225
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.656253814697266,
      "learning_rate": 2.2837837837837837e-05,
      "loss": 1.2989,
      "step": 2226
    },
    {
      "epoch": 0.48,
      "grad_norm": 7.19988489151001,
      "learning_rate": 2.2834620334620337e-05,
      "loss": 1.0639,
      "step": 2227
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.301018714904785,
      "learning_rate": 2.2831402831402833e-05,
      "loss": 1.0593,
      "step": 2228
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.036376953125,
      "learning_rate": 2.282818532818533e-05,
      "loss": 1.1505,
      "step": 2229
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.601282119750977,
      "learning_rate": 2.2824967824967826e-05,
      "loss": 1.4223,
      "step": 2230
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.742104530334473,
      "learning_rate": 2.2821750321750322e-05,
      "loss": 1.3646,
      "step": 2231
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.471049308776855,
      "learning_rate": 2.281853281853282e-05,
      "loss": 1.2612,
      "step": 2232
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.678384780883789,
      "learning_rate": 2.281531531531532e-05,
      "loss": 0.8059,
      "step": 2233
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.38283634185791,
      "learning_rate": 2.2812097812097815e-05,
      "loss": 1.0674,
      "step": 2234
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.731213569641113,
      "learning_rate": 2.280888030888031e-05,
      "loss": 1.3101,
      "step": 2235
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.654257774353027,
      "learning_rate": 2.2805662805662805e-05,
      "loss": 1.2478,
      "step": 2236
    },
    {
      "epoch": 0.48,
      "grad_norm": 7.633584499359131,
      "learning_rate": 2.28024453024453e-05,
      "loss": 1.083,
      "step": 2237
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.132606506347656,
      "learning_rate": 2.2799227799227798e-05,
      "loss": 1.3252,
      "step": 2238
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.059953689575195,
      "learning_rate": 2.2796010296010294e-05,
      "loss": 0.898,
      "step": 2239
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.161406517028809,
      "learning_rate": 2.2792792792792794e-05,
      "loss": 0.9823,
      "step": 2240
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.630030632019043,
      "learning_rate": 2.278957528957529e-05,
      "loss": 1.0917,
      "step": 2241
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.141992568969727,
      "learning_rate": 2.2786357786357787e-05,
      "loss": 1.3226,
      "step": 2242
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.885354042053223,
      "learning_rate": 2.2783140283140283e-05,
      "loss": 1.097,
      "step": 2243
    },
    {
      "epoch": 0.48,
      "grad_norm": 7.437082290649414,
      "learning_rate": 2.277992277992278e-05,
      "loss": 0.966,
      "step": 2244
    },
    {
      "epoch": 0.48,
      "grad_norm": 7.517196178436279,
      "learning_rate": 2.2776705276705276e-05,
      "loss": 0.9713,
      "step": 2245
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.75912857055664,
      "learning_rate": 2.2773487773487776e-05,
      "loss": 0.8059,
      "step": 2246
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.790953636169434,
      "learning_rate": 2.2770270270270272e-05,
      "loss": 0.9534,
      "step": 2247
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.727465629577637,
      "learning_rate": 2.276705276705277e-05,
      "loss": 1.173,
      "step": 2248
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.239615440368652,
      "learning_rate": 2.2763835263835265e-05,
      "loss": 1.0704,
      "step": 2249
    },
    {
      "epoch": 0.48,
      "grad_norm": 7.884565353393555,
      "learning_rate": 2.2760617760617762e-05,
      "loss": 1.0803,
      "step": 2250
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.517741203308105,
      "learning_rate": 2.2757400257400258e-05,
      "loss": 0.9504,
      "step": 2251
    },
    {
      "epoch": 0.48,
      "grad_norm": 7.158363342285156,
      "learning_rate": 2.2754182754182755e-05,
      "loss": 0.8967,
      "step": 2252
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.53520679473877,
      "learning_rate": 2.275096525096525e-05,
      "loss": 1.1146,
      "step": 2253
    },
    {
      "epoch": 0.48,
      "grad_norm": 12.418694496154785,
      "learning_rate": 2.2747747747747748e-05,
      "loss": 1.019,
      "step": 2254
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.6410551071167,
      "learning_rate": 2.2744530244530244e-05,
      "loss": 1.258,
      "step": 2255
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.178105354309082,
      "learning_rate": 2.274131274131274e-05,
      "loss": 1.3763,
      "step": 2256
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.651556968688965,
      "learning_rate": 2.2738095238095237e-05,
      "loss": 0.8328,
      "step": 2257
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.541257858276367,
      "learning_rate": 2.2734877734877733e-05,
      "loss": 1.1142,
      "step": 2258
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.912108421325684,
      "learning_rate": 2.2731660231660233e-05,
      "loss": 0.6874,
      "step": 2259
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.562232971191406,
      "learning_rate": 2.272844272844273e-05,
      "loss": 1.2628,
      "step": 2260
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.636359214782715,
      "learning_rate": 2.2725225225225226e-05,
      "loss": 1.0334,
      "step": 2261
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.406166076660156,
      "learning_rate": 2.2722007722007723e-05,
      "loss": 1.3435,
      "step": 2262
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.615660667419434,
      "learning_rate": 2.271879021879022e-05,
      "loss": 1.0159,
      "step": 2263
    },
    {
      "epoch": 0.49,
      "grad_norm": 12.085264205932617,
      "learning_rate": 2.2715572715572715e-05,
      "loss": 1.4003,
      "step": 2264
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.8613920211792,
      "learning_rate": 2.2712355212355215e-05,
      "loss": 0.9508,
      "step": 2265
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.33315372467041,
      "learning_rate": 2.2709137709137712e-05,
      "loss": 1.1333,
      "step": 2266
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.277847290039062,
      "learning_rate": 2.2705920205920208e-05,
      "loss": 1.2895,
      "step": 2267
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.310836791992188,
      "learning_rate": 2.2702702702702705e-05,
      "loss": 1.0278,
      "step": 2268
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.543488502502441,
      "learning_rate": 2.26994851994852e-05,
      "loss": 1.354,
      "step": 2269
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.850299835205078,
      "learning_rate": 2.2696267696267698e-05,
      "loss": 1.1966,
      "step": 2270
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.636308670043945,
      "learning_rate": 2.269305019305019e-05,
      "loss": 0.93,
      "step": 2271
    },
    {
      "epoch": 0.49,
      "grad_norm": 7.785768985748291,
      "learning_rate": 2.268983268983269e-05,
      "loss": 0.9375,
      "step": 2272
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.120744705200195,
      "learning_rate": 2.2686615186615187e-05,
      "loss": 1.003,
      "step": 2273
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.894142150878906,
      "learning_rate": 2.2683397683397683e-05,
      "loss": 1.2098,
      "step": 2274
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.879216194152832,
      "learning_rate": 2.268018018018018e-05,
      "loss": 1.1184,
      "step": 2275
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.314091682434082,
      "learning_rate": 2.2676962676962676e-05,
      "loss": 1.1595,
      "step": 2276
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.26358413696289,
      "learning_rate": 2.2673745173745173e-05,
      "loss": 0.9292,
      "step": 2277
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.02707576751709,
      "learning_rate": 2.2670527670527673e-05,
      "loss": 1.1727,
      "step": 2278
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.363289833068848,
      "learning_rate": 2.266731016731017e-05,
      "loss": 0.9106,
      "step": 2279
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.637886047363281,
      "learning_rate": 2.2664092664092665e-05,
      "loss": 1.1773,
      "step": 2280
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.283376693725586,
      "learning_rate": 2.2660875160875162e-05,
      "loss": 1.0024,
      "step": 2281
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.709158897399902,
      "learning_rate": 2.265765765765766e-05,
      "loss": 0.9632,
      "step": 2282
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.624267578125,
      "learning_rate": 2.2654440154440155e-05,
      "loss": 1.5086,
      "step": 2283
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.884819030761719,
      "learning_rate": 2.265122265122265e-05,
      "loss": 1.0972,
      "step": 2284
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.889298439025879,
      "learning_rate": 2.264800514800515e-05,
      "loss": 1.1484,
      "step": 2285
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.898066520690918,
      "learning_rate": 2.2644787644787648e-05,
      "loss": 1.1767,
      "step": 2286
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.842985153198242,
      "learning_rate": 2.2641570141570144e-05,
      "loss": 1.2472,
      "step": 2287
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.888350486755371,
      "learning_rate": 2.2638352638352637e-05,
      "loss": 1.0193,
      "step": 2288
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.994672775268555,
      "learning_rate": 2.2635135135135134e-05,
      "loss": 1.0135,
      "step": 2289
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.035858154296875,
      "learning_rate": 2.263191763191763e-05,
      "loss": 0.9951,
      "step": 2290
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.427449226379395,
      "learning_rate": 2.262870012870013e-05,
      "loss": 1.2855,
      "step": 2291
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.582415580749512,
      "learning_rate": 2.2625482625482626e-05,
      "loss": 1.0034,
      "step": 2292
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.199601173400879,
      "learning_rate": 2.2622265122265123e-05,
      "loss": 1.0136,
      "step": 2293
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.75830078125,
      "learning_rate": 2.261904761904762e-05,
      "loss": 0.9137,
      "step": 2294
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.62547492980957,
      "learning_rate": 2.2615830115830116e-05,
      "loss": 1.0441,
      "step": 2295
    },
    {
      "epoch": 0.49,
      "grad_norm": 7.83886194229126,
      "learning_rate": 2.2612612612612612e-05,
      "loss": 0.7665,
      "step": 2296
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.812451362609863,
      "learning_rate": 2.260939510939511e-05,
      "loss": 1.2525,
      "step": 2297
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.654815673828125,
      "learning_rate": 2.260617760617761e-05,
      "loss": 0.903,
      "step": 2298
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.301655769348145,
      "learning_rate": 2.2602960102960105e-05,
      "loss": 1.1591,
      "step": 2299
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.767720222473145,
      "learning_rate": 2.25997425997426e-05,
      "loss": 1.0035,
      "step": 2300
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.551311492919922,
      "learning_rate": 2.2596525096525098e-05,
      "loss": 0.9777,
      "step": 2301
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.826336860656738,
      "learning_rate": 2.2593307593307594e-05,
      "loss": 1.1562,
      "step": 2302
    },
    {
      "epoch": 0.49,
      "grad_norm": 7.796306133270264,
      "learning_rate": 2.259009009009009e-05,
      "loss": 0.8936,
      "step": 2303
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.543041229248047,
      "learning_rate": 2.258687258687259e-05,
      "loss": 1.0834,
      "step": 2304
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.968381881713867,
      "learning_rate": 2.2583655083655084e-05,
      "loss": 1.258,
      "step": 2305
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.421711921691895,
      "learning_rate": 2.258043758043758e-05,
      "loss": 1.0792,
      "step": 2306
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.02097225189209,
      "learning_rate": 2.2577220077220076e-05,
      "loss": 0.9571,
      "step": 2307
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.765973091125488,
      "learning_rate": 2.2574002574002573e-05,
      "loss": 1.198,
      "step": 2308
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.231783866882324,
      "learning_rate": 2.257078507078507e-05,
      "loss": 0.9928,
      "step": 2309
    },
    {
      "epoch": 0.5,
      "grad_norm": 11.378690719604492,
      "learning_rate": 2.2567567567567566e-05,
      "loss": 1.432,
      "step": 2310
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.286345481872559,
      "learning_rate": 2.2564350064350066e-05,
      "loss": 1.0066,
      "step": 2311
    },
    {
      "epoch": 0.5,
      "grad_norm": 7.765833377838135,
      "learning_rate": 2.2561132561132562e-05,
      "loss": 1.0072,
      "step": 2312
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.298741340637207,
      "learning_rate": 2.255791505791506e-05,
      "loss": 1.0065,
      "step": 2313
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.309855461120605,
      "learning_rate": 2.2554697554697555e-05,
      "loss": 1.1403,
      "step": 2314
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.75036334991455,
      "learning_rate": 2.255148005148005e-05,
      "loss": 1.0556,
      "step": 2315
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.546606063842773,
      "learning_rate": 2.2548262548262548e-05,
      "loss": 0.978,
      "step": 2316
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.154243469238281,
      "learning_rate": 2.2545045045045048e-05,
      "loss": 1.1025,
      "step": 2317
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.302839279174805,
      "learning_rate": 2.2541827541827544e-05,
      "loss": 1.1038,
      "step": 2318
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.760649681091309,
      "learning_rate": 2.253861003861004e-05,
      "loss": 0.8837,
      "step": 2319
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.576130867004395,
      "learning_rate": 2.2535392535392537e-05,
      "loss": 1.1867,
      "step": 2320
    },
    {
      "epoch": 0.5,
      "grad_norm": 11.291117668151855,
      "learning_rate": 2.2532175032175034e-05,
      "loss": 1.1458,
      "step": 2321
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.389486312866211,
      "learning_rate": 2.252895752895753e-05,
      "loss": 1.2359,
      "step": 2322
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.921241760253906,
      "learning_rate": 2.2525740025740026e-05,
      "loss": 1.143,
      "step": 2323
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.554208755493164,
      "learning_rate": 2.2522522522522523e-05,
      "loss": 1.0629,
      "step": 2324
    },
    {
      "epoch": 0.5,
      "grad_norm": 7.755317211151123,
      "learning_rate": 2.251930501930502e-05,
      "loss": 0.8201,
      "step": 2325
    },
    {
      "epoch": 0.5,
      "grad_norm": 7.58151912689209,
      "learning_rate": 2.2516087516087516e-05,
      "loss": 0.6795,
      "step": 2326
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.095394134521484,
      "learning_rate": 2.2512870012870012e-05,
      "loss": 0.9756,
      "step": 2327
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.159586906433105,
      "learning_rate": 2.250965250965251e-05,
      "loss": 0.8998,
      "step": 2328
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.709248542785645,
      "learning_rate": 2.2506435006435005e-05,
      "loss": 1.3006,
      "step": 2329
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.280789375305176,
      "learning_rate": 2.2503217503217505e-05,
      "loss": 0.8021,
      "step": 2330
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.130680084228516,
      "learning_rate": 2.25e-05,
      "loss": 0.9945,
      "step": 2331
    },
    {
      "epoch": 0.5,
      "grad_norm": 11.04887580871582,
      "learning_rate": 2.2496782496782498e-05,
      "loss": 1.0872,
      "step": 2332
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.06617259979248,
      "learning_rate": 2.2493564993564994e-05,
      "loss": 1.1817,
      "step": 2333
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.456418991088867,
      "learning_rate": 2.249034749034749e-05,
      "loss": 0.952,
      "step": 2334
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.622138977050781,
      "learning_rate": 2.2487129987129987e-05,
      "loss": 0.9687,
      "step": 2335
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.40770149230957,
      "learning_rate": 2.2483912483912487e-05,
      "loss": 1.0497,
      "step": 2336
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.096805572509766,
      "learning_rate": 2.2480694980694984e-05,
      "loss": 0.8987,
      "step": 2337
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.060961723327637,
      "learning_rate": 2.247747747747748e-05,
      "loss": 1.192,
      "step": 2338
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.218570709228516,
      "learning_rate": 2.2474259974259976e-05,
      "loss": 1.0742,
      "step": 2339
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.23607063293457,
      "learning_rate": 2.247104247104247e-05,
      "loss": 0.9249,
      "step": 2340
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.014078140258789,
      "learning_rate": 2.2467824967824966e-05,
      "loss": 1.3123,
      "step": 2341
    },
    {
      "epoch": 0.5,
      "grad_norm": 7.679869651794434,
      "learning_rate": 2.2464607464607462e-05,
      "loss": 1.0265,
      "step": 2342
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.790985107421875,
      "learning_rate": 2.2461389961389962e-05,
      "loss": 1.257,
      "step": 2343
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.424041748046875,
      "learning_rate": 2.245817245817246e-05,
      "loss": 0.7346,
      "step": 2344
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.129029273986816,
      "learning_rate": 2.2454954954954955e-05,
      "loss": 1.0488,
      "step": 2345
    },
    {
      "epoch": 0.5,
      "grad_norm": 7.899497032165527,
      "learning_rate": 2.245173745173745e-05,
      "loss": 1.1505,
      "step": 2346
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.188064575195312,
      "learning_rate": 2.2448519948519948e-05,
      "loss": 0.9278,
      "step": 2347
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.718648910522461,
      "learning_rate": 2.2445302445302444e-05,
      "loss": 0.8284,
      "step": 2348
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.071867942810059,
      "learning_rate": 2.2442084942084944e-05,
      "loss": 1.0896,
      "step": 2349
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.71068286895752,
      "learning_rate": 2.243886743886744e-05,
      "loss": 1.1001,
      "step": 2350
    },
    {
      "epoch": 0.5,
      "grad_norm": 7.0475993156433105,
      "learning_rate": 2.2435649935649937e-05,
      "loss": 0.7049,
      "step": 2351
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.396679878234863,
      "learning_rate": 2.2432432432432434e-05,
      "loss": 0.9166,
      "step": 2352
    },
    {
      "epoch": 0.5,
      "grad_norm": 11.895612716674805,
      "learning_rate": 2.242921492921493e-05,
      "loss": 1.1321,
      "step": 2353
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.013526916503906,
      "learning_rate": 2.2425997425997427e-05,
      "loss": 1.2088,
      "step": 2354
    },
    {
      "epoch": 0.51,
      "grad_norm": 11.452239036560059,
      "learning_rate": 2.2422779922779923e-05,
      "loss": 1.0977,
      "step": 2355
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.287177085876465,
      "learning_rate": 2.2419562419562423e-05,
      "loss": 1.0531,
      "step": 2356
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.550564765930176,
      "learning_rate": 2.241634491634492e-05,
      "loss": 1.0624,
      "step": 2357
    },
    {
      "epoch": 0.51,
      "grad_norm": 11.385383605957031,
      "learning_rate": 2.2413127413127412e-05,
      "loss": 1.1775,
      "step": 2358
    },
    {
      "epoch": 0.51,
      "grad_norm": 11.758357048034668,
      "learning_rate": 2.240990990990991e-05,
      "loss": 1.0853,
      "step": 2359
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.326055526733398,
      "learning_rate": 2.2406692406692405e-05,
      "loss": 1.1762,
      "step": 2360
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.264159202575684,
      "learning_rate": 2.2403474903474902e-05,
      "loss": 1.1712,
      "step": 2361
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.621282577514648,
      "learning_rate": 2.24002574002574e-05,
      "loss": 1.1103,
      "step": 2362
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.427933692932129,
      "learning_rate": 2.2397039897039898e-05,
      "loss": 1.0571,
      "step": 2363
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.066282272338867,
      "learning_rate": 2.2393822393822394e-05,
      "loss": 1.146,
      "step": 2364
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.012704849243164,
      "learning_rate": 2.239060489060489e-05,
      "loss": 1.196,
      "step": 2365
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.921835899353027,
      "learning_rate": 2.2387387387387387e-05,
      "loss": 1.1075,
      "step": 2366
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.302213668823242,
      "learning_rate": 2.2384169884169884e-05,
      "loss": 1.2673,
      "step": 2367
    },
    {
      "epoch": 0.51,
      "grad_norm": 7.7320380210876465,
      "learning_rate": 2.238095238095238e-05,
      "loss": 0.9184,
      "step": 2368
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.156057357788086,
      "learning_rate": 2.237773487773488e-05,
      "loss": 1.0957,
      "step": 2369
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.731452941894531,
      "learning_rate": 2.2374517374517377e-05,
      "loss": 1.2302,
      "step": 2370
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.397522926330566,
      "learning_rate": 2.2371299871299873e-05,
      "loss": 1.1314,
      "step": 2371
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.330928802490234,
      "learning_rate": 2.236808236808237e-05,
      "loss": 1.0123,
      "step": 2372
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.937274932861328,
      "learning_rate": 2.2364864864864866e-05,
      "loss": 1.23,
      "step": 2373
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.542058944702148,
      "learning_rate": 2.2361647361647362e-05,
      "loss": 1.0609,
      "step": 2374
    },
    {
      "epoch": 0.51,
      "grad_norm": 11.043033599853516,
      "learning_rate": 2.235842985842986e-05,
      "loss": 1.4934,
      "step": 2375
    },
    {
      "epoch": 0.51,
      "grad_norm": 11.247213363647461,
      "learning_rate": 2.2355212355212355e-05,
      "loss": 1.311,
      "step": 2376
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.013314247131348,
      "learning_rate": 2.2351994851994852e-05,
      "loss": 1.0526,
      "step": 2377
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.010601997375488,
      "learning_rate": 2.2348777348777348e-05,
      "loss": 1.2013,
      "step": 2378
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.59426212310791,
      "learning_rate": 2.2345559845559845e-05,
      "loss": 1.0147,
      "step": 2379
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.227002143859863,
      "learning_rate": 2.234234234234234e-05,
      "loss": 1.0276,
      "step": 2380
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.369050979614258,
      "learning_rate": 2.233912483912484e-05,
      "loss": 1.0738,
      "step": 2381
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.821013450622559,
      "learning_rate": 2.2335907335907337e-05,
      "loss": 1.0814,
      "step": 2382
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.771223068237305,
      "learning_rate": 2.2332689832689834e-05,
      "loss": 1.1541,
      "step": 2383
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.893915176391602,
      "learning_rate": 2.232947232947233e-05,
      "loss": 1.1545,
      "step": 2384
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.617655754089355,
      "learning_rate": 2.2326254826254827e-05,
      "loss": 1.302,
      "step": 2385
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.39786148071289,
      "learning_rate": 2.2323037323037323e-05,
      "loss": 1.3356,
      "step": 2386
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.11204719543457,
      "learning_rate": 2.231981981981982e-05,
      "loss": 1.5279,
      "step": 2387
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.303118705749512,
      "learning_rate": 2.231660231660232e-05,
      "loss": 1.1943,
      "step": 2388
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.193735122680664,
      "learning_rate": 2.2313384813384816e-05,
      "loss": 1.113,
      "step": 2389
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.237663269042969,
      "learning_rate": 2.2310167310167312e-05,
      "loss": 1.1394,
      "step": 2390
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.402194023132324,
      "learning_rate": 2.230694980694981e-05,
      "loss": 1.2007,
      "step": 2391
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.45577621459961,
      "learning_rate": 2.2303732303732302e-05,
      "loss": 1.3902,
      "step": 2392
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.043103218078613,
      "learning_rate": 2.23005148005148e-05,
      "loss": 0.7884,
      "step": 2393
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.358236312866211,
      "learning_rate": 2.2297297297297298e-05,
      "loss": 0.9695,
      "step": 2394
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.398865699768066,
      "learning_rate": 2.2294079794079795e-05,
      "loss": 1.0629,
      "step": 2395
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.130887031555176,
      "learning_rate": 2.229086229086229e-05,
      "loss": 1.4363,
      "step": 2396
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.300829887390137,
      "learning_rate": 2.2287644787644788e-05,
      "loss": 1.4548,
      "step": 2397
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.367837905883789,
      "learning_rate": 2.2284427284427284e-05,
      "loss": 1.1361,
      "step": 2398
    },
    {
      "epoch": 0.51,
      "grad_norm": 7.915188789367676,
      "learning_rate": 2.228120978120978e-05,
      "loss": 0.9586,
      "step": 2399
    },
    {
      "epoch": 0.51,
      "grad_norm": 9.490331649780273,
      "learning_rate": 2.2277992277992277e-05,
      "loss": 1.2891,
      "step": 2400
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.499619483947754,
      "learning_rate": 2.2274774774774777e-05,
      "loss": 0.9094,
      "step": 2401
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.49809455871582,
      "learning_rate": 2.2271557271557273e-05,
      "loss": 1.0662,
      "step": 2402
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.72047233581543,
      "learning_rate": 2.226833976833977e-05,
      "loss": 1.4364,
      "step": 2403
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.997987747192383,
      "learning_rate": 2.2265122265122266e-05,
      "loss": 1.0717,
      "step": 2404
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.086843490600586,
      "learning_rate": 2.2261904761904763e-05,
      "loss": 0.9164,
      "step": 2405
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.14774227142334,
      "learning_rate": 2.225868725868726e-05,
      "loss": 1.2422,
      "step": 2406
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.84349250793457,
      "learning_rate": 2.225546975546976e-05,
      "loss": 1.2406,
      "step": 2407
    },
    {
      "epoch": 0.52,
      "grad_norm": 7.713539123535156,
      "learning_rate": 2.2252252252252255e-05,
      "loss": 0.8716,
      "step": 2408
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.099452018737793,
      "learning_rate": 2.2249034749034752e-05,
      "loss": 0.9839,
      "step": 2409
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.65322208404541,
      "learning_rate": 2.2245817245817245e-05,
      "loss": 1.1191,
      "step": 2410
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.390517234802246,
      "learning_rate": 2.224259974259974e-05,
      "loss": 0.9891,
      "step": 2411
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.446367263793945,
      "learning_rate": 2.2239382239382238e-05,
      "loss": 1.2867,
      "step": 2412
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.689008712768555,
      "learning_rate": 2.2236164736164734e-05,
      "loss": 0.9219,
      "step": 2413
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.48943042755127,
      "learning_rate": 2.2232947232947234e-05,
      "loss": 1.2358,
      "step": 2414
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.575973510742188,
      "learning_rate": 2.222972972972973e-05,
      "loss": 1.3663,
      "step": 2415
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.329336166381836,
      "learning_rate": 2.2226512226512227e-05,
      "loss": 1.0272,
      "step": 2416
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.927145004272461,
      "learning_rate": 2.2223294723294723e-05,
      "loss": 1.1769,
      "step": 2417
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.258356094360352,
      "learning_rate": 2.222007722007722e-05,
      "loss": 1.2515,
      "step": 2418
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.386484146118164,
      "learning_rate": 2.2216859716859716e-05,
      "loss": 0.9754,
      "step": 2419
    },
    {
      "epoch": 0.52,
      "grad_norm": 7.547216892242432,
      "learning_rate": 2.2213642213642216e-05,
      "loss": 0.9152,
      "step": 2420
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.024325370788574,
      "learning_rate": 2.2210424710424713e-05,
      "loss": 1.0123,
      "step": 2421
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.294087409973145,
      "learning_rate": 2.220720720720721e-05,
      "loss": 1.0813,
      "step": 2422
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.471619606018066,
      "learning_rate": 2.2203989703989705e-05,
      "loss": 1.069,
      "step": 2423
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.810153007507324,
      "learning_rate": 2.2200772200772202e-05,
      "loss": 1.341,
      "step": 2424
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.921043395996094,
      "learning_rate": 2.21975546975547e-05,
      "loss": 0.9252,
      "step": 2425
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.76163387298584,
      "learning_rate": 2.2194337194337195e-05,
      "loss": 1.1104,
      "step": 2426
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.502260208129883,
      "learning_rate": 2.219111969111969e-05,
      "loss": 1.1512,
      "step": 2427
    },
    {
      "epoch": 0.52,
      "grad_norm": 7.594200611114502,
      "learning_rate": 2.2187902187902188e-05,
      "loss": 0.8499,
      "step": 2428
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.695162773132324,
      "learning_rate": 2.2184684684684684e-05,
      "loss": 1.2079,
      "step": 2429
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.569151878356934,
      "learning_rate": 2.218146718146718e-05,
      "loss": 1.1211,
      "step": 2430
    },
    {
      "epoch": 0.52,
      "grad_norm": 11.18136215209961,
      "learning_rate": 2.2178249678249677e-05,
      "loss": 1.206,
      "step": 2431
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.093317985534668,
      "learning_rate": 2.2175032175032173e-05,
      "loss": 1.0169,
      "step": 2432
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.390506744384766,
      "learning_rate": 2.2171814671814673e-05,
      "loss": 1.1101,
      "step": 2433
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.954838752746582,
      "learning_rate": 2.216859716859717e-05,
      "loss": 0.7289,
      "step": 2434
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.448768615722656,
      "learning_rate": 2.2165379665379666e-05,
      "loss": 0.9822,
      "step": 2435
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.293468475341797,
      "learning_rate": 2.2162162162162163e-05,
      "loss": 1.0551,
      "step": 2436
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.023652076721191,
      "learning_rate": 2.215894465894466e-05,
      "loss": 1.138,
      "step": 2437
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.036847114562988,
      "learning_rate": 2.2155727155727156e-05,
      "loss": 0.8921,
      "step": 2438
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.3295259475708,
      "learning_rate": 2.2152509652509655e-05,
      "loss": 1.0727,
      "step": 2439
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.21536636352539,
      "learning_rate": 2.2149292149292152e-05,
      "loss": 1.2142,
      "step": 2440
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.903542518615723,
      "learning_rate": 2.214607464607465e-05,
      "loss": 1.3801,
      "step": 2441
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.728826522827148,
      "learning_rate": 2.2142857142857145e-05,
      "loss": 1.1135,
      "step": 2442
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.535191535949707,
      "learning_rate": 2.213963963963964e-05,
      "loss": 1.3249,
      "step": 2443
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.645769119262695,
      "learning_rate": 2.2136422136422138e-05,
      "loss": 1.4417,
      "step": 2444
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.057226181030273,
      "learning_rate": 2.213320463320463e-05,
      "loss": 1.0135,
      "step": 2445
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.284515380859375,
      "learning_rate": 2.212998712998713e-05,
      "loss": 1.1224,
      "step": 2446
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.432469367980957,
      "learning_rate": 2.2126769626769627e-05,
      "loss": 1.3828,
      "step": 2447
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.907944679260254,
      "learning_rate": 2.2123552123552123e-05,
      "loss": 1.1064,
      "step": 2448
    },
    {
      "epoch": 0.53,
      "grad_norm": 10.027575492858887,
      "learning_rate": 2.212033462033462e-05,
      "loss": 0.9911,
      "step": 2449
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.457405090332031,
      "learning_rate": 2.2117117117117116e-05,
      "loss": 0.8128,
      "step": 2450
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.663128852844238,
      "learning_rate": 2.2113899613899613e-05,
      "loss": 1.1074,
      "step": 2451
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.049463272094727,
      "learning_rate": 2.2110682110682113e-05,
      "loss": 1.2246,
      "step": 2452
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.79466438293457,
      "learning_rate": 2.210746460746461e-05,
      "loss": 1.2148,
      "step": 2453
    },
    {
      "epoch": 0.53,
      "grad_norm": 6.7344770431518555,
      "learning_rate": 2.2104247104247106e-05,
      "loss": 0.8021,
      "step": 2454
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.942646980285645,
      "learning_rate": 2.2101029601029602e-05,
      "loss": 1.2937,
      "step": 2455
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.65300989151001,
      "learning_rate": 2.20978120978121e-05,
      "loss": 0.8773,
      "step": 2456
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.888759136199951,
      "learning_rate": 2.2094594594594595e-05,
      "loss": 0.8986,
      "step": 2457
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.444100379943848,
      "learning_rate": 2.209137709137709e-05,
      "loss": 1.0042,
      "step": 2458
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.358449935913086,
      "learning_rate": 2.208815958815959e-05,
      "loss": 0.9173,
      "step": 2459
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.722016334533691,
      "learning_rate": 2.2084942084942088e-05,
      "loss": 1.0833,
      "step": 2460
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.362723350524902,
      "learning_rate": 2.2081724581724584e-05,
      "loss": 1.0388,
      "step": 2461
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.457125663757324,
      "learning_rate": 2.2078507078507077e-05,
      "loss": 1.0051,
      "step": 2462
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.94926929473877,
      "learning_rate": 2.2075289575289574e-05,
      "loss": 1.1754,
      "step": 2463
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.544604301452637,
      "learning_rate": 2.207207207207207e-05,
      "loss": 1.1187,
      "step": 2464
    },
    {
      "epoch": 0.53,
      "grad_norm": 10.662936210632324,
      "learning_rate": 2.206885456885457e-05,
      "loss": 1.2995,
      "step": 2465
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.614227294921875,
      "learning_rate": 2.2065637065637066e-05,
      "loss": 1.035,
      "step": 2466
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.43929386138916,
      "learning_rate": 2.2062419562419563e-05,
      "loss": 1.1956,
      "step": 2467
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.31195068359375,
      "learning_rate": 2.205920205920206e-05,
      "loss": 1.1935,
      "step": 2468
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.113081932067871,
      "learning_rate": 2.2055984555984556e-05,
      "loss": 0.9517,
      "step": 2469
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.65079402923584,
      "learning_rate": 2.2052767052767052e-05,
      "loss": 0.9893,
      "step": 2470
    },
    {
      "epoch": 0.53,
      "grad_norm": 10.363582611083984,
      "learning_rate": 2.204954954954955e-05,
      "loss": 1.3017,
      "step": 2471
    },
    {
      "epoch": 0.53,
      "grad_norm": 10.209056854248047,
      "learning_rate": 2.204633204633205e-05,
      "loss": 0.9581,
      "step": 2472
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.315794944763184,
      "learning_rate": 2.2043114543114545e-05,
      "loss": 0.9158,
      "step": 2473
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.994805335998535,
      "learning_rate": 2.203989703989704e-05,
      "loss": 1.1616,
      "step": 2474
    },
    {
      "epoch": 0.53,
      "grad_norm": 10.063371658325195,
      "learning_rate": 2.2036679536679538e-05,
      "loss": 1.0871,
      "step": 2475
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.30704116821289,
      "learning_rate": 2.2033462033462034e-05,
      "loss": 1.005,
      "step": 2476
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.88941764831543,
      "learning_rate": 2.203024453024453e-05,
      "loss": 1.2659,
      "step": 2477
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.256254196166992,
      "learning_rate": 2.202702702702703e-05,
      "loss": 0.9812,
      "step": 2478
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.216120719909668,
      "learning_rate": 2.2023809523809524e-05,
      "loss": 0.8904,
      "step": 2479
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.88805103302002,
      "learning_rate": 2.202059202059202e-05,
      "loss": 0.984,
      "step": 2480
    },
    {
      "epoch": 0.53,
      "grad_norm": 11.07582950592041,
      "learning_rate": 2.2017374517374517e-05,
      "loss": 1.5315,
      "step": 2481
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.766288757324219,
      "learning_rate": 2.2014157014157013e-05,
      "loss": 0.9998,
      "step": 2482
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.347466468811035,
      "learning_rate": 2.201093951093951e-05,
      "loss": 1.0409,
      "step": 2483
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.2208251953125,
      "learning_rate": 2.2007722007722006e-05,
      "loss": 1.033,
      "step": 2484
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.955559730529785,
      "learning_rate": 2.2004504504504506e-05,
      "loss": 0.9537,
      "step": 2485
    },
    {
      "epoch": 0.53,
      "grad_norm": 10.399721145629883,
      "learning_rate": 2.2001287001287002e-05,
      "loss": 1.0298,
      "step": 2486
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.564810752868652,
      "learning_rate": 2.19980694980695e-05,
      "loss": 0.8263,
      "step": 2487
    },
    {
      "epoch": 0.53,
      "grad_norm": 10.080057144165039,
      "learning_rate": 2.1994851994851995e-05,
      "loss": 1.2141,
      "step": 2488
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.99401569366455,
      "learning_rate": 2.199163449163449e-05,
      "loss": 1.2188,
      "step": 2489
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.777819633483887,
      "learning_rate": 2.1988416988416988e-05,
      "loss": 1.0533,
      "step": 2490
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.694416046142578,
      "learning_rate": 2.1985199485199488e-05,
      "loss": 1.0281,
      "step": 2491
    },
    {
      "epoch": 0.53,
      "grad_norm": 11.139851570129395,
      "learning_rate": 2.1981981981981984e-05,
      "loss": 1.2539,
      "step": 2492
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.868833541870117,
      "learning_rate": 2.197876447876448e-05,
      "loss": 1.0052,
      "step": 2493
    },
    {
      "epoch": 0.53,
      "grad_norm": 10.068967819213867,
      "learning_rate": 2.1975546975546977e-05,
      "loss": 1.0497,
      "step": 2494
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.497453689575195,
      "learning_rate": 2.1972329472329474e-05,
      "loss": 0.9113,
      "step": 2495
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.543828964233398,
      "learning_rate": 2.196911196911197e-05,
      "loss": 0.9471,
      "step": 2496
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.283011436462402,
      "learning_rate": 2.1965894465894467e-05,
      "loss": 0.9892,
      "step": 2497
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.690478324890137,
      "learning_rate": 2.1962676962676963e-05,
      "loss": 0.919,
      "step": 2498
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.318765640258789,
      "learning_rate": 2.195945945945946e-05,
      "loss": 1.1862,
      "step": 2499
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.871881484985352,
      "learning_rate": 2.1956241956241956e-05,
      "loss": 1.265,
      "step": 2500
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.637395858764648,
      "learning_rate": 2.1953024453024452e-05,
      "loss": 1.2272,
      "step": 2501
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.681711196899414,
      "learning_rate": 2.194980694980695e-05,
      "loss": 1.0613,
      "step": 2502
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.737967491149902,
      "learning_rate": 2.1946589446589445e-05,
      "loss": 1.3597,
      "step": 2503
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.517800331115723,
      "learning_rate": 2.1943371943371945e-05,
      "loss": 0.9722,
      "step": 2504
    },
    {
      "epoch": 0.54,
      "grad_norm": 11.018777847290039,
      "learning_rate": 2.194015444015444e-05,
      "loss": 1.2199,
      "step": 2505
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.482754707336426,
      "learning_rate": 2.1936936936936938e-05,
      "loss": 1.1183,
      "step": 2506
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.274056434631348,
      "learning_rate": 2.1933719433719434e-05,
      "loss": 1.013,
      "step": 2507
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.978628158569336,
      "learning_rate": 2.193050193050193e-05,
      "loss": 1.1027,
      "step": 2508
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.468233108520508,
      "learning_rate": 2.1927284427284427e-05,
      "loss": 1.1659,
      "step": 2509
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.510199546813965,
      "learning_rate": 2.1924066924066927e-05,
      "loss": 1.3075,
      "step": 2510
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.793375015258789,
      "learning_rate": 2.1920849420849424e-05,
      "loss": 0.9314,
      "step": 2511
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.423321723937988,
      "learning_rate": 2.191763191763192e-05,
      "loss": 1.1458,
      "step": 2512
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.340312957763672,
      "learning_rate": 2.1914414414414417e-05,
      "loss": 1.0493,
      "step": 2513
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.485490798950195,
      "learning_rate": 2.191119691119691e-05,
      "loss": 0.975,
      "step": 2514
    },
    {
      "epoch": 0.54,
      "grad_norm": 11.058269500732422,
      "learning_rate": 2.1907979407979406e-05,
      "loss": 1.2428,
      "step": 2515
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.366105079650879,
      "learning_rate": 2.1904761904761903e-05,
      "loss": 0.8478,
      "step": 2516
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.010530471801758,
      "learning_rate": 2.1901544401544402e-05,
      "loss": 1.1343,
      "step": 2517
    },
    {
      "epoch": 0.54,
      "grad_norm": 11.221162796020508,
      "learning_rate": 2.18983268983269e-05,
      "loss": 1.1912,
      "step": 2518
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.397418022155762,
      "learning_rate": 2.1895109395109395e-05,
      "loss": 0.7918,
      "step": 2519
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.284348487854004,
      "learning_rate": 2.1891891891891892e-05,
      "loss": 1.1156,
      "step": 2520
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.33513069152832,
      "learning_rate": 2.1888674388674388e-05,
      "loss": 1.015,
      "step": 2521
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.289148330688477,
      "learning_rate": 2.1885456885456885e-05,
      "loss": 1.3202,
      "step": 2522
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.014851570129395,
      "learning_rate": 2.1882239382239384e-05,
      "loss": 0.8484,
      "step": 2523
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.838643074035645,
      "learning_rate": 2.187902187902188e-05,
      "loss": 1.1514,
      "step": 2524
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.29642391204834,
      "learning_rate": 2.1875804375804377e-05,
      "loss": 0.8366,
      "step": 2525
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.782570838928223,
      "learning_rate": 2.1872586872586874e-05,
      "loss": 1.0354,
      "step": 2526
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.203195571899414,
      "learning_rate": 2.186936936936937e-05,
      "loss": 0.8362,
      "step": 2527
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.855531692504883,
      "learning_rate": 2.1866151866151867e-05,
      "loss": 1.4902,
      "step": 2528
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.922639846801758,
      "learning_rate": 2.1862934362934363e-05,
      "loss": 0.9311,
      "step": 2529
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.307760238647461,
      "learning_rate": 2.1859716859716863e-05,
      "loss": 1.0554,
      "step": 2530
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.702019691467285,
      "learning_rate": 2.185649935649936e-05,
      "loss": 1.1819,
      "step": 2531
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.427835464477539,
      "learning_rate": 2.1853281853281853e-05,
      "loss": 0.9763,
      "step": 2532
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.741948127746582,
      "learning_rate": 2.185006435006435e-05,
      "loss": 1.1831,
      "step": 2533
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.796297550201416,
      "learning_rate": 2.1846846846846845e-05,
      "loss": 0.8541,
      "step": 2534
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.719788551330566,
      "learning_rate": 2.1843629343629342e-05,
      "loss": 1.1663,
      "step": 2535
    },
    {
      "epoch": 0.54,
      "grad_norm": 11.362380981445312,
      "learning_rate": 2.1840411840411842e-05,
      "loss": 1.0657,
      "step": 2536
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.904614448547363,
      "learning_rate": 2.1837194337194338e-05,
      "loss": 1.1121,
      "step": 2537
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.736067771911621,
      "learning_rate": 2.1833976833976835e-05,
      "loss": 1.1733,
      "step": 2538
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.383532524108887,
      "learning_rate": 2.183075933075933e-05,
      "loss": 0.8444,
      "step": 2539
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.854815483093262,
      "learning_rate": 2.1827541827541828e-05,
      "loss": 1.1896,
      "step": 2540
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.048881530761719,
      "learning_rate": 2.1824324324324324e-05,
      "loss": 1.0717,
      "step": 2541
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.955305099487305,
      "learning_rate": 2.182110682110682e-05,
      "loss": 1.1752,
      "step": 2542
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.785613059997559,
      "learning_rate": 2.181788931788932e-05,
      "loss": 1.0021,
      "step": 2543
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.274267196655273,
      "learning_rate": 2.1814671814671817e-05,
      "loss": 1.0287,
      "step": 2544
    },
    {
      "epoch": 0.55,
      "grad_norm": 10.489089012145996,
      "learning_rate": 2.1811454311454313e-05,
      "loss": 1.0263,
      "step": 2545
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.454707145690918,
      "learning_rate": 2.180823680823681e-05,
      "loss": 0.8827,
      "step": 2546
    },
    {
      "epoch": 0.55,
      "grad_norm": 10.005447387695312,
      "learning_rate": 2.1805019305019306e-05,
      "loss": 1.2056,
      "step": 2547
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.844840049743652,
      "learning_rate": 2.1801801801801803e-05,
      "loss": 1.1013,
      "step": 2548
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.25471019744873,
      "learning_rate": 2.17985842985843e-05,
      "loss": 0.9068,
      "step": 2549
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.475262641906738,
      "learning_rate": 2.1795366795366795e-05,
      "loss": 1.0283,
      "step": 2550
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.894894123077393,
      "learning_rate": 2.1792149292149292e-05,
      "loss": 0.8628,
      "step": 2551
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.638346672058105,
      "learning_rate": 2.178893178893179e-05,
      "loss": 1.0748,
      "step": 2552
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.91783332824707,
      "learning_rate": 2.1785714285714285e-05,
      "loss": 0.8574,
      "step": 2553
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.670327186584473,
      "learning_rate": 2.178249678249678e-05,
      "loss": 0.9662,
      "step": 2554
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.51383113861084,
      "learning_rate": 2.177927927927928e-05,
      "loss": 1.1017,
      "step": 2555
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.92487907409668,
      "learning_rate": 2.1776061776061778e-05,
      "loss": 0.82,
      "step": 2556
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.85591459274292,
      "learning_rate": 2.1772844272844274e-05,
      "loss": 0.8474,
      "step": 2557
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.337370872497559,
      "learning_rate": 2.176962676962677e-05,
      "loss": 1.2339,
      "step": 2558
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.413569450378418,
      "learning_rate": 2.1766409266409267e-05,
      "loss": 0.9116,
      "step": 2559
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.391502380371094,
      "learning_rate": 2.1763191763191763e-05,
      "loss": 1.1715,
      "step": 2560
    },
    {
      "epoch": 0.55,
      "grad_norm": 11.58267593383789,
      "learning_rate": 2.175997425997426e-05,
      "loss": 1.2131,
      "step": 2561
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.172009468078613,
      "learning_rate": 2.175675675675676e-05,
      "loss": 0.8875,
      "step": 2562
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.119043350219727,
      "learning_rate": 2.1753539253539256e-05,
      "loss": 1.1193,
      "step": 2563
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.720065116882324,
      "learning_rate": 2.1750321750321753e-05,
      "loss": 0.9022,
      "step": 2564
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.317041397094727,
      "learning_rate": 2.174710424710425e-05,
      "loss": 0.909,
      "step": 2565
    },
    {
      "epoch": 0.55,
      "grad_norm": 10.217742919921875,
      "learning_rate": 2.1743886743886742e-05,
      "loss": 1.004,
      "step": 2566
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.066926956176758,
      "learning_rate": 2.174066924066924e-05,
      "loss": 0.9885,
      "step": 2567
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.116436004638672,
      "learning_rate": 2.173745173745174e-05,
      "loss": 0.8618,
      "step": 2568
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.84638500213623,
      "learning_rate": 2.1734234234234235e-05,
      "loss": 0.9185,
      "step": 2569
    },
    {
      "epoch": 0.55,
      "grad_norm": 11.97932243347168,
      "learning_rate": 2.173101673101673e-05,
      "loss": 1.2089,
      "step": 2570
    },
    {
      "epoch": 0.55,
      "grad_norm": 10.833267211914062,
      "learning_rate": 2.1727799227799228e-05,
      "loss": 1.3294,
      "step": 2571
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.18907642364502,
      "learning_rate": 2.1724581724581724e-05,
      "loss": 0.9619,
      "step": 2572
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.423666000366211,
      "learning_rate": 2.172136422136422e-05,
      "loss": 0.9758,
      "step": 2573
    },
    {
      "epoch": 0.55,
      "grad_norm": 10.215032577514648,
      "learning_rate": 2.1718146718146717e-05,
      "loss": 1.2089,
      "step": 2574
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.773152351379395,
      "learning_rate": 2.1714929214929217e-05,
      "loss": 0.947,
      "step": 2575
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.226311683654785,
      "learning_rate": 2.1711711711711713e-05,
      "loss": 1.1304,
      "step": 2576
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.289237022399902,
      "learning_rate": 2.170849420849421e-05,
      "loss": 0.9646,
      "step": 2577
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.013460159301758,
      "learning_rate": 2.1705276705276706e-05,
      "loss": 1.03,
      "step": 2578
    },
    {
      "epoch": 0.55,
      "grad_norm": 10.118247985839844,
      "learning_rate": 2.1702059202059203e-05,
      "loss": 1.1024,
      "step": 2579
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.690688610076904,
      "learning_rate": 2.16988416988417e-05,
      "loss": 1.1702,
      "step": 2580
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.431748390197754,
      "learning_rate": 2.16956241956242e-05,
      "loss": 1.2586,
      "step": 2581
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.551826477050781,
      "learning_rate": 2.1692406692406695e-05,
      "loss": 0.9756,
      "step": 2582
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.418062210083008,
      "learning_rate": 2.1689189189189192e-05,
      "loss": 1.0719,
      "step": 2583
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.771937370300293,
      "learning_rate": 2.1685971685971685e-05,
      "loss": 1.0654,
      "step": 2584
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.414965629577637,
      "learning_rate": 2.168275418275418e-05,
      "loss": 1.1778,
      "step": 2585
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.615158557891846,
      "learning_rate": 2.1679536679536678e-05,
      "loss": 1.0851,
      "step": 2586
    },
    {
      "epoch": 0.55,
      "grad_norm": 8.467525482177734,
      "learning_rate": 2.1676319176319174e-05,
      "loss": 1.1008,
      "step": 2587
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.170727729797363,
      "learning_rate": 2.1673101673101674e-05,
      "loss": 1.138,
      "step": 2588
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.178673267364502,
      "learning_rate": 2.166988416988417e-05,
      "loss": 0.8537,
      "step": 2589
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.746742248535156,
      "learning_rate": 2.1666666666666667e-05,
      "loss": 1.1593,
      "step": 2590
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.701788902282715,
      "learning_rate": 2.1663449163449163e-05,
      "loss": 0.9366,
      "step": 2591
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.706136226654053,
      "learning_rate": 2.166023166023166e-05,
      "loss": 1.058,
      "step": 2592
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.7190580368042,
      "learning_rate": 2.1657014157014156e-05,
      "loss": 1.1812,
      "step": 2593
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.81433629989624,
      "learning_rate": 2.1653796653796656e-05,
      "loss": 1.2131,
      "step": 2594
    },
    {
      "epoch": 0.56,
      "grad_norm": 11.918575286865234,
      "learning_rate": 2.1650579150579153e-05,
      "loss": 1.1283,
      "step": 2595
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.166764259338379,
      "learning_rate": 2.164736164736165e-05,
      "loss": 1.0987,
      "step": 2596
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.862667560577393,
      "learning_rate": 2.1644144144144146e-05,
      "loss": 0.9507,
      "step": 2597
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.640422821044922,
      "learning_rate": 2.1640926640926642e-05,
      "loss": 1.0499,
      "step": 2598
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.50528335571289,
      "learning_rate": 2.163770913770914e-05,
      "loss": 0.9382,
      "step": 2599
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.90034008026123,
      "learning_rate": 2.1634491634491635e-05,
      "loss": 1.3984,
      "step": 2600
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.648384094238281,
      "learning_rate": 2.163127413127413e-05,
      "loss": 1.1395,
      "step": 2601
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.93450927734375,
      "learning_rate": 2.1628056628056628e-05,
      "loss": 1.2669,
      "step": 2602
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.699728012084961,
      "learning_rate": 2.1624839124839124e-05,
      "loss": 1.092,
      "step": 2603
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.176445960998535,
      "learning_rate": 2.162162162162162e-05,
      "loss": 1.1818,
      "step": 2604
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.528993129730225,
      "learning_rate": 2.1618404118404117e-05,
      "loss": 0.871,
      "step": 2605
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.338607788085938,
      "learning_rate": 2.1615186615186614e-05,
      "loss": 1.0172,
      "step": 2606
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.249470710754395,
      "learning_rate": 2.1611969111969113e-05,
      "loss": 0.9201,
      "step": 2607
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.913167953491211,
      "learning_rate": 2.160875160875161e-05,
      "loss": 1.2647,
      "step": 2608
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.983447074890137,
      "learning_rate": 2.1605534105534106e-05,
      "loss": 0.9222,
      "step": 2609
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.022592544555664,
      "learning_rate": 2.1602316602316603e-05,
      "loss": 0.7401,
      "step": 2610
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.294769287109375,
      "learning_rate": 2.15990990990991e-05,
      "loss": 1.2333,
      "step": 2611
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.294540405273438,
      "learning_rate": 2.1595881595881596e-05,
      "loss": 1.2439,
      "step": 2612
    },
    {
      "epoch": 0.56,
      "grad_norm": 11.439111709594727,
      "learning_rate": 2.1592664092664096e-05,
      "loss": 1.0886,
      "step": 2613
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.911421298980713,
      "learning_rate": 2.1589446589446592e-05,
      "loss": 0.5951,
      "step": 2614
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.7986421585083,
      "learning_rate": 2.158622908622909e-05,
      "loss": 1.0617,
      "step": 2615
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.12961483001709,
      "learning_rate": 2.1583011583011585e-05,
      "loss": 1.1738,
      "step": 2616
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.602049827575684,
      "learning_rate": 2.157979407979408e-05,
      "loss": 1.1761,
      "step": 2617
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.289383888244629,
      "learning_rate": 2.1576576576576578e-05,
      "loss": 1.1133,
      "step": 2618
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.575774192810059,
      "learning_rate": 2.157335907335907e-05,
      "loss": 0.8044,
      "step": 2619
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.478899002075195,
      "learning_rate": 2.157014157014157e-05,
      "loss": 0.9018,
      "step": 2620
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.291579246520996,
      "learning_rate": 2.1566924066924067e-05,
      "loss": 1.2903,
      "step": 2621
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.44237995147705,
      "learning_rate": 2.1563706563706564e-05,
      "loss": 0.9448,
      "step": 2622
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.272027015686035,
      "learning_rate": 2.156048906048906e-05,
      "loss": 1.2175,
      "step": 2623
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.869856834411621,
      "learning_rate": 2.1557271557271557e-05,
      "loss": 1.2965,
      "step": 2624
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.409807205200195,
      "learning_rate": 2.1554054054054053e-05,
      "loss": 1.1586,
      "step": 2625
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.592605590820312,
      "learning_rate": 2.1550836550836553e-05,
      "loss": 1.0495,
      "step": 2626
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.069169044494629,
      "learning_rate": 2.154761904761905e-05,
      "loss": 0.9985,
      "step": 2627
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.01742172241211,
      "learning_rate": 2.1544401544401546e-05,
      "loss": 0.911,
      "step": 2628
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.663751602172852,
      "learning_rate": 2.1541184041184042e-05,
      "loss": 1.2441,
      "step": 2629
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.823939323425293,
      "learning_rate": 2.153796653796654e-05,
      "loss": 0.8325,
      "step": 2630
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.189085960388184,
      "learning_rate": 2.1534749034749035e-05,
      "loss": 0.8591,
      "step": 2631
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.88523006439209,
      "learning_rate": 2.153153153153153e-05,
      "loss": 1.1802,
      "step": 2632
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.999014854431152,
      "learning_rate": 2.152831402831403e-05,
      "loss": 1.2948,
      "step": 2633
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.353347778320312,
      "learning_rate": 2.1525096525096528e-05,
      "loss": 1.3313,
      "step": 2634
    },
    {
      "epoch": 0.57,
      "grad_norm": 7.3266682624816895,
      "learning_rate": 2.1521879021879024e-05,
      "loss": 0.755,
      "step": 2635
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.238184928894043,
      "learning_rate": 2.1518661518661517e-05,
      "loss": 1.0248,
      "step": 2636
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.267760276794434,
      "learning_rate": 2.1515444015444014e-05,
      "loss": 0.9018,
      "step": 2637
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.519793510437012,
      "learning_rate": 2.151222651222651e-05,
      "loss": 1.1158,
      "step": 2638
    },
    {
      "epoch": 0.57,
      "grad_norm": 7.663290977478027,
      "learning_rate": 2.150900900900901e-05,
      "loss": 0.9923,
      "step": 2639
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.93054485321045,
      "learning_rate": 2.1505791505791507e-05,
      "loss": 1.144,
      "step": 2640
    },
    {
      "epoch": 0.57,
      "grad_norm": 7.3633317947387695,
      "learning_rate": 2.1502574002574003e-05,
      "loss": 0.8599,
      "step": 2641
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.771753311157227,
      "learning_rate": 2.14993564993565e-05,
      "loss": 1.1914,
      "step": 2642
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.014059066772461,
      "learning_rate": 2.1496138996138996e-05,
      "loss": 0.8676,
      "step": 2643
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.412409782409668,
      "learning_rate": 2.1492921492921492e-05,
      "loss": 0.8733,
      "step": 2644
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.906844139099121,
      "learning_rate": 2.148970398970399e-05,
      "loss": 1.0782,
      "step": 2645
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.19869327545166,
      "learning_rate": 2.148648648648649e-05,
      "loss": 0.8639,
      "step": 2646
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.648829460144043,
      "learning_rate": 2.1483268983268985e-05,
      "loss": 0.919,
      "step": 2647
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.963582992553711,
      "learning_rate": 2.148005148005148e-05,
      "loss": 1.0634,
      "step": 2648
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.67484188079834,
      "learning_rate": 2.1476833976833978e-05,
      "loss": 0.8767,
      "step": 2649
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.72409439086914,
      "learning_rate": 2.1473616473616474e-05,
      "loss": 0.9523,
      "step": 2650
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.606922149658203,
      "learning_rate": 2.147039897039897e-05,
      "loss": 1.0077,
      "step": 2651
    },
    {
      "epoch": 0.57,
      "grad_norm": 12.258495330810547,
      "learning_rate": 2.146718146718147e-05,
      "loss": 1.2565,
      "step": 2652
    },
    {
      "epoch": 0.57,
      "grad_norm": 7.643850803375244,
      "learning_rate": 2.1463963963963964e-05,
      "loss": 0.9366,
      "step": 2653
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.763330459594727,
      "learning_rate": 2.146074646074646e-05,
      "loss": 1.2778,
      "step": 2654
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.131698608398438,
      "learning_rate": 2.1457528957528957e-05,
      "loss": 1.2984,
      "step": 2655
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.494336128234863,
      "learning_rate": 2.1454311454311453e-05,
      "loss": 0.9113,
      "step": 2656
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.561063766479492,
      "learning_rate": 2.145109395109395e-05,
      "loss": 1.2463,
      "step": 2657
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.44370174407959,
      "learning_rate": 2.144787644787645e-05,
      "loss": 1.0203,
      "step": 2658
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.808847427368164,
      "learning_rate": 2.1444658944658946e-05,
      "loss": 1.0465,
      "step": 2659
    },
    {
      "epoch": 0.57,
      "grad_norm": 7.673030853271484,
      "learning_rate": 2.1441441441441442e-05,
      "loss": 0.7597,
      "step": 2660
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.304078102111816,
      "learning_rate": 2.143822393822394e-05,
      "loss": 1.0644,
      "step": 2661
    },
    {
      "epoch": 0.57,
      "grad_norm": 11.623624801635742,
      "learning_rate": 2.1435006435006435e-05,
      "loss": 1.3314,
      "step": 2662
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.453004837036133,
      "learning_rate": 2.1431788931788932e-05,
      "loss": 1.2861,
      "step": 2663
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.069978713989258,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 0.9834,
      "step": 2664
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.571484565734863,
      "learning_rate": 2.1425353925353928e-05,
      "loss": 1.0912,
      "step": 2665
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.20410442352295,
      "learning_rate": 2.1422136422136424e-05,
      "loss": 1.1205,
      "step": 2666
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.286768913269043,
      "learning_rate": 2.141891891891892e-05,
      "loss": 0.9391,
      "step": 2667
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.36376953125,
      "learning_rate": 2.1415701415701417e-05,
      "loss": 1.0544,
      "step": 2668
    },
    {
      "epoch": 0.57,
      "grad_norm": 11.269440650939941,
      "learning_rate": 2.1412483912483914e-05,
      "loss": 1.0824,
      "step": 2669
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.053434371948242,
      "learning_rate": 2.140926640926641e-05,
      "loss": 0.9564,
      "step": 2670
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.12969970703125,
      "learning_rate": 2.1406048906048907e-05,
      "loss": 1.1133,
      "step": 2671
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.638909339904785,
      "learning_rate": 2.1402831402831403e-05,
      "loss": 0.8398,
      "step": 2672
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.065963745117188,
      "learning_rate": 2.13996138996139e-05,
      "loss": 1.0664,
      "step": 2673
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.001046180725098,
      "learning_rate": 2.1396396396396396e-05,
      "loss": 0.6901,
      "step": 2674
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.412848472595215,
      "learning_rate": 2.1393178893178892e-05,
      "loss": 1.1418,
      "step": 2675
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.573115348815918,
      "learning_rate": 2.138996138996139e-05,
      "loss": 0.8975,
      "step": 2676
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.195610046386719,
      "learning_rate": 2.1386743886743885e-05,
      "loss": 1.119,
      "step": 2677
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.22775936126709,
      "learning_rate": 2.1383526383526385e-05,
      "loss": 0.9173,
      "step": 2678
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.775042533874512,
      "learning_rate": 2.138030888030888e-05,
      "loss": 0.936,
      "step": 2679
    },
    {
      "epoch": 0.57,
      "grad_norm": 8.619747161865234,
      "learning_rate": 2.1377091377091378e-05,
      "loss": 1.0732,
      "step": 2680
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.044519424438477,
      "learning_rate": 2.1373873873873875e-05,
      "loss": 0.8433,
      "step": 2681
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.9891939163208,
      "learning_rate": 2.137065637065637e-05,
      "loss": 1.436,
      "step": 2682
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.138111114501953,
      "learning_rate": 2.1367438867438867e-05,
      "loss": 1.0871,
      "step": 2683
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.932372093200684,
      "learning_rate": 2.1364221364221367e-05,
      "loss": 0.8937,
      "step": 2684
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.501051902770996,
      "learning_rate": 2.1361003861003864e-05,
      "loss": 1.1084,
      "step": 2685
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.541765213012695,
      "learning_rate": 2.135778635778636e-05,
      "loss": 0.9189,
      "step": 2686
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.365545272827148,
      "learning_rate": 2.1354568854568857e-05,
      "loss": 1.078,
      "step": 2687
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.642423629760742,
      "learning_rate": 2.135135135135135e-05,
      "loss": 1.077,
      "step": 2688
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.753179550170898,
      "learning_rate": 2.1348133848133846e-05,
      "loss": 1.1374,
      "step": 2689
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.697073936462402,
      "learning_rate": 2.1344916344916343e-05,
      "loss": 1.1113,
      "step": 2690
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.635124206542969,
      "learning_rate": 2.1341698841698842e-05,
      "loss": 1.2922,
      "step": 2691
    },
    {
      "epoch": 0.58,
      "grad_norm": 7.857353210449219,
      "learning_rate": 2.133848133848134e-05,
      "loss": 0.7946,
      "step": 2692
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.527445793151855,
      "learning_rate": 2.1335263835263835e-05,
      "loss": 0.9418,
      "step": 2693
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.299134254455566,
      "learning_rate": 2.1332046332046332e-05,
      "loss": 1.0808,
      "step": 2694
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.126667022705078,
      "learning_rate": 2.1328828828828828e-05,
      "loss": 0.9341,
      "step": 2695
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.34541130065918,
      "learning_rate": 2.1325611325611325e-05,
      "loss": 1.257,
      "step": 2696
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.83531379699707,
      "learning_rate": 2.1322393822393825e-05,
      "loss": 0.8889,
      "step": 2697
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.806289672851562,
      "learning_rate": 2.131917631917632e-05,
      "loss": 1.0288,
      "step": 2698
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.658234596252441,
      "learning_rate": 2.1315958815958817e-05,
      "loss": 1.2988,
      "step": 2699
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.863208770751953,
      "learning_rate": 2.1312741312741314e-05,
      "loss": 1.0283,
      "step": 2700
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.33183765411377,
      "learning_rate": 2.130952380952381e-05,
      "loss": 1.0846,
      "step": 2701
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.665968894958496,
      "learning_rate": 2.1306306306306307e-05,
      "loss": 0.807,
      "step": 2702
    },
    {
      "epoch": 0.58,
      "grad_norm": 11.188556671142578,
      "learning_rate": 2.1303088803088803e-05,
      "loss": 1.138,
      "step": 2703
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.068321228027344,
      "learning_rate": 2.1299871299871303e-05,
      "loss": 1.0137,
      "step": 2704
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.92458724975586,
      "learning_rate": 2.12966537966538e-05,
      "loss": 1.1338,
      "step": 2705
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.1302490234375,
      "learning_rate": 2.1293436293436293e-05,
      "loss": 0.9218,
      "step": 2706
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.044108390808105,
      "learning_rate": 2.129021879021879e-05,
      "loss": 0.9255,
      "step": 2707
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.52872371673584,
      "learning_rate": 2.1287001287001286e-05,
      "loss": 1.1123,
      "step": 2708
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.374285697937012,
      "learning_rate": 2.1283783783783782e-05,
      "loss": 0.9981,
      "step": 2709
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.792009353637695,
      "learning_rate": 2.1280566280566282e-05,
      "loss": 1.0165,
      "step": 2710
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.348731994628906,
      "learning_rate": 2.1277348777348778e-05,
      "loss": 1.0984,
      "step": 2711
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.679716110229492,
      "learning_rate": 2.1274131274131275e-05,
      "loss": 1.0594,
      "step": 2712
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.376962661743164,
      "learning_rate": 2.127091377091377e-05,
      "loss": 1.2184,
      "step": 2713
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.196380615234375,
      "learning_rate": 2.1267696267696268e-05,
      "loss": 0.9413,
      "step": 2714
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.80842113494873,
      "learning_rate": 2.1264478764478764e-05,
      "loss": 1.1067,
      "step": 2715
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.204976081848145,
      "learning_rate": 2.126126126126126e-05,
      "loss": 1.2285,
      "step": 2716
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.215396881103516,
      "learning_rate": 2.125804375804376e-05,
      "loss": 1.0028,
      "step": 2717
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.528359413146973,
      "learning_rate": 2.1254826254826257e-05,
      "loss": 1.2152,
      "step": 2718
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.92618465423584,
      "learning_rate": 2.1251608751608753e-05,
      "loss": 1.0599,
      "step": 2719
    },
    {
      "epoch": 0.58,
      "grad_norm": 11.368350982666016,
      "learning_rate": 2.124839124839125e-05,
      "loss": 1.2349,
      "step": 2720
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.953558921813965,
      "learning_rate": 2.1245173745173746e-05,
      "loss": 0.9488,
      "step": 2721
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.31515121459961,
      "learning_rate": 2.1241956241956243e-05,
      "loss": 1.036,
      "step": 2722
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.91295051574707,
      "learning_rate": 2.123873873873874e-05,
      "loss": 0.9875,
      "step": 2723
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.043198585510254,
      "learning_rate": 2.1235521235521236e-05,
      "loss": 1.144,
      "step": 2724
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.018251419067383,
      "learning_rate": 2.1232303732303732e-05,
      "loss": 1.4276,
      "step": 2725
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.761055946350098,
      "learning_rate": 2.122908622908623e-05,
      "loss": 1.1708,
      "step": 2726
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.224836349487305,
      "learning_rate": 2.1225868725868725e-05,
      "loss": 1.0551,
      "step": 2727
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.01115894317627,
      "learning_rate": 2.122265122265122e-05,
      "loss": 0.9214,
      "step": 2728
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.036235809326172,
      "learning_rate": 2.121943371943372e-05,
      "loss": 1.0865,
      "step": 2729
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.548211097717285,
      "learning_rate": 2.1216216216216218e-05,
      "loss": 0.8986,
      "step": 2730
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.559911727905273,
      "learning_rate": 2.1212998712998714e-05,
      "loss": 1.0446,
      "step": 2731
    },
    {
      "epoch": 0.59,
      "grad_norm": 11.290431022644043,
      "learning_rate": 2.120978120978121e-05,
      "loss": 1.3296,
      "step": 2732
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.935272216796875,
      "learning_rate": 2.1206563706563707e-05,
      "loss": 0.766,
      "step": 2733
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.781343936920166,
      "learning_rate": 2.1203346203346203e-05,
      "loss": 0.8526,
      "step": 2734
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.589937210083008,
      "learning_rate": 2.12001287001287e-05,
      "loss": 0.7761,
      "step": 2735
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.25551986694336,
      "learning_rate": 2.11969111969112e-05,
      "loss": 0.9609,
      "step": 2736
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.14185905456543,
      "learning_rate": 2.1193693693693696e-05,
      "loss": 0.9752,
      "step": 2737
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.51018238067627,
      "learning_rate": 2.1190476190476193e-05,
      "loss": 0.8848,
      "step": 2738
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.766529083251953,
      "learning_rate": 2.118725868725869e-05,
      "loss": 1.063,
      "step": 2739
    },
    {
      "epoch": 0.59,
      "grad_norm": 11.641081809997559,
      "learning_rate": 2.1184041184041182e-05,
      "loss": 1.13,
      "step": 2740
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.503905773162842,
      "learning_rate": 2.118082368082368e-05,
      "loss": 0.7835,
      "step": 2741
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.246943473815918,
      "learning_rate": 2.117760617760618e-05,
      "loss": 1.2911,
      "step": 2742
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.935830116271973,
      "learning_rate": 2.1174388674388675e-05,
      "loss": 1.085,
      "step": 2743
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.006217956542969,
      "learning_rate": 2.117117117117117e-05,
      "loss": 0.6622,
      "step": 2744
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.042974472045898,
      "learning_rate": 2.1167953667953668e-05,
      "loss": 0.9432,
      "step": 2745
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.866181373596191,
      "learning_rate": 2.1164736164736164e-05,
      "loss": 0.7795,
      "step": 2746
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.355125427246094,
      "learning_rate": 2.116151866151866e-05,
      "loss": 1.2135,
      "step": 2747
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.36242389678955,
      "learning_rate": 2.1158301158301157e-05,
      "loss": 1.1789,
      "step": 2748
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.376669883728027,
      "learning_rate": 2.1155083655083657e-05,
      "loss": 1.1837,
      "step": 2749
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.81640625,
      "learning_rate": 2.1151866151866153e-05,
      "loss": 0.868,
      "step": 2750
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.229677200317383,
      "learning_rate": 2.114864864864865e-05,
      "loss": 1.1307,
      "step": 2751
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.952521324157715,
      "learning_rate": 2.1145431145431146e-05,
      "loss": 1.0333,
      "step": 2752
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.131302833557129,
      "learning_rate": 2.1142213642213643e-05,
      "loss": 1.2445,
      "step": 2753
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.519216537475586,
      "learning_rate": 2.113899613899614e-05,
      "loss": 0.9375,
      "step": 2754
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.698931694030762,
      "learning_rate": 2.113577863577864e-05,
      "loss": 1.1677,
      "step": 2755
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.321867942810059,
      "learning_rate": 2.1132561132561136e-05,
      "loss": 0.9683,
      "step": 2756
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.645792961120605,
      "learning_rate": 2.1129343629343632e-05,
      "loss": 0.9612,
      "step": 2757
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.029123306274414,
      "learning_rate": 2.1126126126126125e-05,
      "loss": 0.861,
      "step": 2758
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.331148147583008,
      "learning_rate": 2.112290862290862e-05,
      "loss": 1.1327,
      "step": 2759
    },
    {
      "epoch": 0.59,
      "grad_norm": 11.252057075500488,
      "learning_rate": 2.1119691119691118e-05,
      "loss": 1.163,
      "step": 2760
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.156002044677734,
      "learning_rate": 2.1116473616473614e-05,
      "loss": 0.7168,
      "step": 2761
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.27966594696045,
      "learning_rate": 2.1113256113256114e-05,
      "loss": 0.828,
      "step": 2762
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.871383666992188,
      "learning_rate": 2.111003861003861e-05,
      "loss": 0.8043,
      "step": 2763
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.867463111877441,
      "learning_rate": 2.1106821106821107e-05,
      "loss": 0.8903,
      "step": 2764
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.727655410766602,
      "learning_rate": 2.1103603603603604e-05,
      "loss": 0.939,
      "step": 2765
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.762966156005859,
      "learning_rate": 2.11003861003861e-05,
      "loss": 0.9229,
      "step": 2766
    },
    {
      "epoch": 0.59,
      "grad_norm": 6.91019868850708,
      "learning_rate": 2.1097168597168597e-05,
      "loss": 0.7882,
      "step": 2767
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.986659049987793,
      "learning_rate": 2.1093951093951096e-05,
      "loss": 0.8726,
      "step": 2768
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.143381118774414,
      "learning_rate": 2.1090733590733593e-05,
      "loss": 1.1696,
      "step": 2769
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.073074340820312,
      "learning_rate": 2.108751608751609e-05,
      "loss": 1.1354,
      "step": 2770
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.876468658447266,
      "learning_rate": 2.1084298584298586e-05,
      "loss": 0.7387,
      "step": 2771
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.30185317993164,
      "learning_rate": 2.1081081081081082e-05,
      "loss": 0.9798,
      "step": 2772
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.006104469299316,
      "learning_rate": 2.107786357786358e-05,
      "loss": 1.0047,
      "step": 2773
    },
    {
      "epoch": 0.6,
      "grad_norm": 11.379852294921875,
      "learning_rate": 2.1074646074646075e-05,
      "loss": 1.6872,
      "step": 2774
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.685883522033691,
      "learning_rate": 2.107142857142857e-05,
      "loss": 0.7895,
      "step": 2775
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.728376388549805,
      "learning_rate": 2.1068211068211068e-05,
      "loss": 1.0252,
      "step": 2776
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.01990795135498,
      "learning_rate": 2.1064993564993564e-05,
      "loss": 0.9423,
      "step": 2777
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.767251014709473,
      "learning_rate": 2.106177606177606e-05,
      "loss": 0.9922,
      "step": 2778
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.025846481323242,
      "learning_rate": 2.1058558558558557e-05,
      "loss": 1.1614,
      "step": 2779
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.91926383972168,
      "learning_rate": 2.1055341055341054e-05,
      "loss": 0.957,
      "step": 2780
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.825723648071289,
      "learning_rate": 2.1052123552123554e-05,
      "loss": 1.0642,
      "step": 2781
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.941263198852539,
      "learning_rate": 2.104890604890605e-05,
      "loss": 1.0173,
      "step": 2782
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.793000221252441,
      "learning_rate": 2.1045688545688547e-05,
      "loss": 0.7084,
      "step": 2783
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.00562572479248,
      "learning_rate": 2.1042471042471043e-05,
      "loss": 1.014,
      "step": 2784
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.289148330688477,
      "learning_rate": 2.103925353925354e-05,
      "loss": 0.9305,
      "step": 2785
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.703402519226074,
      "learning_rate": 2.1036036036036036e-05,
      "loss": 1.0488,
      "step": 2786
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.79773235321045,
      "learning_rate": 2.1032818532818536e-05,
      "loss": 1.0066,
      "step": 2787
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.23941421508789,
      "learning_rate": 2.1029601029601032e-05,
      "loss": 1.0626,
      "step": 2788
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.365097045898438,
      "learning_rate": 2.102638352638353e-05,
      "loss": 1.0688,
      "step": 2789
    },
    {
      "epoch": 0.6,
      "grad_norm": 7.793513298034668,
      "learning_rate": 2.1023166023166025e-05,
      "loss": 0.997,
      "step": 2790
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.026046752929688,
      "learning_rate": 2.101994851994852e-05,
      "loss": 1.1313,
      "step": 2791
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.093482971191406,
      "learning_rate": 2.1016731016731015e-05,
      "loss": 0.8597,
      "step": 2792
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.320773124694824,
      "learning_rate": 2.101351351351351e-05,
      "loss": 1.1895,
      "step": 2793
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.871332168579102,
      "learning_rate": 2.101029601029601e-05,
      "loss": 1.1999,
      "step": 2794
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.995553016662598,
      "learning_rate": 2.1007078507078507e-05,
      "loss": 1.1547,
      "step": 2795
    },
    {
      "epoch": 0.6,
      "grad_norm": 7.951255798339844,
      "learning_rate": 2.1003861003861004e-05,
      "loss": 1.0862,
      "step": 2796
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.467206001281738,
      "learning_rate": 2.10006435006435e-05,
      "loss": 0.9043,
      "step": 2797
    },
    {
      "epoch": 0.6,
      "grad_norm": 11.796592712402344,
      "learning_rate": 2.0997425997425997e-05,
      "loss": 1.2535,
      "step": 2798
    },
    {
      "epoch": 0.6,
      "grad_norm": 11.160659790039062,
      "learning_rate": 2.0994208494208493e-05,
      "loss": 1.3867,
      "step": 2799
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.853734016418457,
      "learning_rate": 2.0990990990990993e-05,
      "loss": 1.1718,
      "step": 2800
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.570667266845703,
      "learning_rate": 2.098777348777349e-05,
      "loss": 1.0739,
      "step": 2801
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.438458442687988,
      "learning_rate": 2.0984555984555986e-05,
      "loss": 0.9132,
      "step": 2802
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.110907554626465,
      "learning_rate": 2.0981338481338482e-05,
      "loss": 0.8515,
      "step": 2803
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.953829765319824,
      "learning_rate": 2.097812097812098e-05,
      "loss": 0.9936,
      "step": 2804
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.393338203430176,
      "learning_rate": 2.0974903474903475e-05,
      "loss": 1.0052,
      "step": 2805
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.042749404907227,
      "learning_rate": 2.097168597168597e-05,
      "loss": 0.9362,
      "step": 2806
    },
    {
      "epoch": 0.6,
      "grad_norm": 11.333725929260254,
      "learning_rate": 2.096846846846847e-05,
      "loss": 1.2118,
      "step": 2807
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.852368354797363,
      "learning_rate": 2.0965250965250968e-05,
      "loss": 1.2012,
      "step": 2808
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.990324020385742,
      "learning_rate": 2.0962033462033464e-05,
      "loss": 0.9594,
      "step": 2809
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.096481323242188,
      "learning_rate": 2.0958815958815957e-05,
      "loss": 0.9516,
      "step": 2810
    },
    {
      "epoch": 0.6,
      "grad_norm": 11.100303649902344,
      "learning_rate": 2.0955598455598454e-05,
      "loss": 1.1726,
      "step": 2811
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.895795822143555,
      "learning_rate": 2.095238095238095e-05,
      "loss": 1.2785,
      "step": 2812
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.789200782775879,
      "learning_rate": 2.094916344916345e-05,
      "loss": 1.0463,
      "step": 2813
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.932344436645508,
      "learning_rate": 2.0945945945945947e-05,
      "loss": 0.871,
      "step": 2814
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.22294807434082,
      "learning_rate": 2.0942728442728443e-05,
      "loss": 0.886,
      "step": 2815
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.548498153686523,
      "learning_rate": 2.093951093951094e-05,
      "loss": 1.091,
      "step": 2816
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.865466117858887,
      "learning_rate": 2.0936293436293436e-05,
      "loss": 1.0254,
      "step": 2817
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.689888954162598,
      "learning_rate": 2.0933075933075932e-05,
      "loss": 0.9701,
      "step": 2818
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.203476905822754,
      "learning_rate": 2.092985842985843e-05,
      "loss": 0.9902,
      "step": 2819
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.612614631652832,
      "learning_rate": 2.092664092664093e-05,
      "loss": 0.9034,
      "step": 2820
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.763526916503906,
      "learning_rate": 2.0923423423423425e-05,
      "loss": 1.0936,
      "step": 2821
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.159196853637695,
      "learning_rate": 2.092020592020592e-05,
      "loss": 0.9897,
      "step": 2822
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.557538986206055,
      "learning_rate": 2.0916988416988418e-05,
      "loss": 0.9649,
      "step": 2823
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.814242362976074,
      "learning_rate": 2.0913770913770915e-05,
      "loss": 1.0098,
      "step": 2824
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.728209495544434,
      "learning_rate": 2.091055341055341e-05,
      "loss": 0.8788,
      "step": 2825
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.61665153503418,
      "learning_rate": 2.090733590733591e-05,
      "loss": 1.0899,
      "step": 2826
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.486169815063477,
      "learning_rate": 2.0904118404118404e-05,
      "loss": 1.0981,
      "step": 2827
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.799489974975586,
      "learning_rate": 2.09009009009009e-05,
      "loss": 0.8275,
      "step": 2828
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.992798805236816,
      "learning_rate": 2.0897683397683397e-05,
      "loss": 1.0309,
      "step": 2829
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.237815856933594,
      "learning_rate": 2.0894465894465893e-05,
      "loss": 1.153,
      "step": 2830
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.129735946655273,
      "learning_rate": 2.089124839124839e-05,
      "loss": 0.9838,
      "step": 2831
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.392934799194336,
      "learning_rate": 2.088803088803089e-05,
      "loss": 0.7634,
      "step": 2832
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.84146785736084,
      "learning_rate": 2.0884813384813386e-05,
      "loss": 0.8491,
      "step": 2833
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.821837425231934,
      "learning_rate": 2.0881595881595882e-05,
      "loss": 0.8678,
      "step": 2834
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.389864921569824,
      "learning_rate": 2.087837837837838e-05,
      "loss": 1.292,
      "step": 2835
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.579080581665039,
      "learning_rate": 2.0875160875160875e-05,
      "loss": 0.9762,
      "step": 2836
    },
    {
      "epoch": 0.61,
      "grad_norm": 11.983601570129395,
      "learning_rate": 2.0871943371943372e-05,
      "loss": 0.9683,
      "step": 2837
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.586101531982422,
      "learning_rate": 2.0868725868725868e-05,
      "loss": 0.9862,
      "step": 2838
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.276551246643066,
      "learning_rate": 2.0865508365508368e-05,
      "loss": 0.9716,
      "step": 2839
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.825735092163086,
      "learning_rate": 2.0862290862290865e-05,
      "loss": 0.7717,
      "step": 2840
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.01533317565918,
      "learning_rate": 2.085907335907336e-05,
      "loss": 0.8249,
      "step": 2841
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.143072128295898,
      "learning_rate": 2.0855855855855857e-05,
      "loss": 0.891,
      "step": 2842
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.57685661315918,
      "learning_rate": 2.0852638352638354e-05,
      "loss": 1.077,
      "step": 2843
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.60960865020752,
      "learning_rate": 2.084942084942085e-05,
      "loss": 1.0637,
      "step": 2844
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.907783508300781,
      "learning_rate": 2.0846203346203347e-05,
      "loss": 0.8779,
      "step": 2845
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.834234237670898,
      "learning_rate": 2.0842985842985843e-05,
      "loss": 0.6931,
      "step": 2846
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.47803020477295,
      "learning_rate": 2.083976833976834e-05,
      "loss": 1.0983,
      "step": 2847
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.337279319763184,
      "learning_rate": 2.0836550836550836e-05,
      "loss": 0.9585,
      "step": 2848
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.789628982543945,
      "learning_rate": 2.0833333333333333e-05,
      "loss": 1.09,
      "step": 2849
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.334077835083008,
      "learning_rate": 2.083011583011583e-05,
      "loss": 1.1289,
      "step": 2850
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.820341110229492,
      "learning_rate": 2.0826898326898326e-05,
      "loss": 1.1193,
      "step": 2851
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.186652183532715,
      "learning_rate": 2.0823680823680825e-05,
      "loss": 1.0275,
      "step": 2852
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.707109451293945,
      "learning_rate": 2.0820463320463322e-05,
      "loss": 0.9854,
      "step": 2853
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.378288269042969,
      "learning_rate": 2.0817245817245818e-05,
      "loss": 1.0021,
      "step": 2854
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.319125175476074,
      "learning_rate": 2.0814028314028315e-05,
      "loss": 1.2867,
      "step": 2855
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.348020553588867,
      "learning_rate": 2.081081081081081e-05,
      "loss": 1.1411,
      "step": 2856
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.638346672058105,
      "learning_rate": 2.0807593307593308e-05,
      "loss": 1.1337,
      "step": 2857
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.666051387786865,
      "learning_rate": 2.0804375804375807e-05,
      "loss": 0.9734,
      "step": 2858
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.839030265808105,
      "learning_rate": 2.0801158301158304e-05,
      "loss": 0.9729,
      "step": 2859
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.480268955230713,
      "learning_rate": 2.07979407979408e-05,
      "loss": 0.7938,
      "step": 2860
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.518192291259766,
      "learning_rate": 2.0794723294723297e-05,
      "loss": 1.0861,
      "step": 2861
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.149147033691406,
      "learning_rate": 2.079150579150579e-05,
      "loss": 0.9599,
      "step": 2862
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.796531677246094,
      "learning_rate": 2.0788288288288286e-05,
      "loss": 1.259,
      "step": 2863
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.009387969970703,
      "learning_rate": 2.0785070785070783e-05,
      "loss": 0.9353,
      "step": 2864
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.9170355796813965,
      "learning_rate": 2.0781853281853283e-05,
      "loss": 0.8964,
      "step": 2865
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.48659896850586,
      "learning_rate": 2.077863577863578e-05,
      "loss": 0.948,
      "step": 2866
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.355876922607422,
      "learning_rate": 2.0775418275418276e-05,
      "loss": 1.1059,
      "step": 2867
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.813748359680176,
      "learning_rate": 2.0772200772200772e-05,
      "loss": 0.954,
      "step": 2868
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.593411445617676,
      "learning_rate": 2.076898326898327e-05,
      "loss": 1.1891,
      "step": 2869
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.729681968688965,
      "learning_rate": 2.0765765765765765e-05,
      "loss": 1.0692,
      "step": 2870
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.42676067352295,
      "learning_rate": 2.0762548262548265e-05,
      "loss": 0.9442,
      "step": 2871
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.473650932312012,
      "learning_rate": 2.075933075933076e-05,
      "loss": 0.8634,
      "step": 2872
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.209397315979004,
      "learning_rate": 2.0756113256113258e-05,
      "loss": 1.3776,
      "step": 2873
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.305933952331543,
      "learning_rate": 2.0752895752895754e-05,
      "loss": 0.8982,
      "step": 2874
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.120150566101074,
      "learning_rate": 2.074967824967825e-05,
      "loss": 0.8903,
      "step": 2875
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.782783508300781,
      "learning_rate": 2.0746460746460747e-05,
      "loss": 1.0416,
      "step": 2876
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.606042861938477,
      "learning_rate": 2.0743243243243243e-05,
      "loss": 1.0934,
      "step": 2877
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.637984275817871,
      "learning_rate": 2.0740025740025743e-05,
      "loss": 0.9603,
      "step": 2878
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.931666374206543,
      "learning_rate": 2.0736808236808236e-05,
      "loss": 0.8418,
      "step": 2879
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.084803581237793,
      "learning_rate": 2.0733590733590733e-05,
      "loss": 0.8455,
      "step": 2880
    },
    {
      "epoch": 0.62,
      "grad_norm": 7.2396559715271,
      "learning_rate": 2.073037323037323e-05,
      "loss": 0.6755,
      "step": 2881
    },
    {
      "epoch": 0.62,
      "grad_norm": 11.697168350219727,
      "learning_rate": 2.0727155727155726e-05,
      "loss": 1.252,
      "step": 2882
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.119372367858887,
      "learning_rate": 2.0723938223938222e-05,
      "loss": 0.9697,
      "step": 2883
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.729304313659668,
      "learning_rate": 2.0720720720720722e-05,
      "loss": 1.0312,
      "step": 2884
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.687933921813965,
      "learning_rate": 2.071750321750322e-05,
      "loss": 1.0565,
      "step": 2885
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.31495475769043,
      "learning_rate": 2.0714285714285715e-05,
      "loss": 1.0423,
      "step": 2886
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.60663890838623,
      "learning_rate": 2.071106821106821e-05,
      "loss": 0.8755,
      "step": 2887
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.770864486694336,
      "learning_rate": 2.0707850707850708e-05,
      "loss": 0.9535,
      "step": 2888
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.478055000305176,
      "learning_rate": 2.0704633204633204e-05,
      "loss": 1.36,
      "step": 2889
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.90584659576416,
      "learning_rate": 2.07014157014157e-05,
      "loss": 1.2249,
      "step": 2890
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.35338306427002,
      "learning_rate": 2.06981981981982e-05,
      "loss": 1.0369,
      "step": 2891
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.702576637268066,
      "learning_rate": 2.0694980694980697e-05,
      "loss": 1.0237,
      "step": 2892
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.908188819885254,
      "learning_rate": 2.0691763191763193e-05,
      "loss": 1.0094,
      "step": 2893
    },
    {
      "epoch": 0.62,
      "grad_norm": 7.421445846557617,
      "learning_rate": 2.068854568854569e-05,
      "loss": 0.8716,
      "step": 2894
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.35153865814209,
      "learning_rate": 2.0685328185328186e-05,
      "loss": 0.9944,
      "step": 2895
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.820887565612793,
      "learning_rate": 2.0682110682110683e-05,
      "loss": 1.0244,
      "step": 2896
    },
    {
      "epoch": 0.62,
      "grad_norm": 7.951667785644531,
      "learning_rate": 2.067889317889318e-05,
      "loss": 0.8943,
      "step": 2897
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.195419311523438,
      "learning_rate": 2.0675675675675676e-05,
      "loss": 0.7828,
      "step": 2898
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.433738708496094,
      "learning_rate": 2.0672458172458172e-05,
      "loss": 1.1866,
      "step": 2899
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.210555076599121,
      "learning_rate": 2.066924066924067e-05,
      "loss": 1.2584,
      "step": 2900
    },
    {
      "epoch": 0.62,
      "grad_norm": 11.122232437133789,
      "learning_rate": 2.0666023166023165e-05,
      "loss": 1.1295,
      "step": 2901
    },
    {
      "epoch": 0.62,
      "grad_norm": 11.257704734802246,
      "learning_rate": 2.066280566280566e-05,
      "loss": 0.9352,
      "step": 2902
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.423574447631836,
      "learning_rate": 2.065958815958816e-05,
      "loss": 0.8269,
      "step": 2903
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.775230407714844,
      "learning_rate": 2.0656370656370658e-05,
      "loss": 1.1005,
      "step": 2904
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.966784477233887,
      "learning_rate": 2.0653153153153154e-05,
      "loss": 0.9312,
      "step": 2905
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.278223991394043,
      "learning_rate": 2.064993564993565e-05,
      "loss": 0.9964,
      "step": 2906
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.855120658874512,
      "learning_rate": 2.0646718146718147e-05,
      "loss": 0.9698,
      "step": 2907
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.015430450439453,
      "learning_rate": 2.0643500643500644e-05,
      "loss": 0.8809,
      "step": 2908
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.973637580871582,
      "learning_rate": 2.064028314028314e-05,
      "loss": 1.2957,
      "step": 2909
    },
    {
      "epoch": 0.62,
      "grad_norm": 11.171722412109375,
      "learning_rate": 2.063706563706564e-05,
      "loss": 1.4145,
      "step": 2910
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.660077095031738,
      "learning_rate": 2.0633848133848136e-05,
      "loss": 0.9685,
      "step": 2911
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.585137367248535,
      "learning_rate": 2.0630630630630633e-05,
      "loss": 1.1007,
      "step": 2912
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.188427925109863,
      "learning_rate": 2.062741312741313e-05,
      "loss": 1.0642,
      "step": 2913
    },
    {
      "epoch": 0.63,
      "grad_norm": 11.072477340698242,
      "learning_rate": 2.0624195624195622e-05,
      "loss": 1.0366,
      "step": 2914
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.086532592773438,
      "learning_rate": 2.062097812097812e-05,
      "loss": 0.8435,
      "step": 2915
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.772308349609375,
      "learning_rate": 2.061776061776062e-05,
      "loss": 0.9945,
      "step": 2916
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.08843994140625,
      "learning_rate": 2.0614543114543115e-05,
      "loss": 0.978,
      "step": 2917
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.956644058227539,
      "learning_rate": 2.061132561132561e-05,
      "loss": 0.7595,
      "step": 2918
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.848326683044434,
      "learning_rate": 2.0608108108108108e-05,
      "loss": 1.127,
      "step": 2919
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.869176864624023,
      "learning_rate": 2.0604890604890604e-05,
      "loss": 0.9702,
      "step": 2920
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.538840293884277,
      "learning_rate": 2.06016731016731e-05,
      "loss": 1.0066,
      "step": 2921
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.90531587600708,
      "learning_rate": 2.0598455598455597e-05,
      "loss": 0.9553,
      "step": 2922
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.631621360778809,
      "learning_rate": 2.0595238095238097e-05,
      "loss": 1.0359,
      "step": 2923
    },
    {
      "epoch": 0.63,
      "grad_norm": 10.626422882080078,
      "learning_rate": 2.0592020592020594e-05,
      "loss": 1.0592,
      "step": 2924
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.15100383758545,
      "learning_rate": 2.058880308880309e-05,
      "loss": 0.9446,
      "step": 2925
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.092101097106934,
      "learning_rate": 2.0585585585585586e-05,
      "loss": 1.0069,
      "step": 2926
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.848550796508789,
      "learning_rate": 2.0582368082368083e-05,
      "loss": 0.8523,
      "step": 2927
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.158607482910156,
      "learning_rate": 2.057915057915058e-05,
      "loss": 1.0696,
      "step": 2928
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.685415267944336,
      "learning_rate": 2.057593307593308e-05,
      "loss": 1.562,
      "step": 2929
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.920199394226074,
      "learning_rate": 2.0572715572715576e-05,
      "loss": 0.9975,
      "step": 2930
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.30274772644043,
      "learning_rate": 2.0569498069498072e-05,
      "loss": 0.9812,
      "step": 2931
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.033167839050293,
      "learning_rate": 2.0566280566280565e-05,
      "loss": 0.8365,
      "step": 2932
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.075596809387207,
      "learning_rate": 2.056306306306306e-05,
      "loss": 0.7999,
      "step": 2933
    },
    {
      "epoch": 0.63,
      "grad_norm": 10.575961112976074,
      "learning_rate": 2.0559845559845558e-05,
      "loss": 1.2242,
      "step": 2934
    },
    {
      "epoch": 0.63,
      "grad_norm": 6.733513832092285,
      "learning_rate": 2.0556628056628055e-05,
      "loss": 0.8469,
      "step": 2935
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.757370948791504,
      "learning_rate": 2.0553410553410554e-05,
      "loss": 1.0754,
      "step": 2936
    },
    {
      "epoch": 0.63,
      "grad_norm": 11.059587478637695,
      "learning_rate": 2.055019305019305e-05,
      "loss": 1.25,
      "step": 2937
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.753438472747803,
      "learning_rate": 2.0546975546975547e-05,
      "loss": 0.6649,
      "step": 2938
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.71579360961914,
      "learning_rate": 2.0543758043758044e-05,
      "loss": 1.0132,
      "step": 2939
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.584855079650879,
      "learning_rate": 2.054054054054054e-05,
      "loss": 1.0235,
      "step": 2940
    },
    {
      "epoch": 0.63,
      "grad_norm": 10.276100158691406,
      "learning_rate": 2.0537323037323037e-05,
      "loss": 1.2333,
      "step": 2941
    },
    {
      "epoch": 0.63,
      "grad_norm": 10.499653816223145,
      "learning_rate": 2.0534105534105536e-05,
      "loss": 1.0126,
      "step": 2942
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.410548210144043,
      "learning_rate": 2.0530888030888033e-05,
      "loss": 1.0749,
      "step": 2943
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.905138969421387,
      "learning_rate": 2.052767052767053e-05,
      "loss": 0.7621,
      "step": 2944
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.144352912902832,
      "learning_rate": 2.0524453024453026e-05,
      "loss": 1.0414,
      "step": 2945
    },
    {
      "epoch": 0.63,
      "grad_norm": 10.541153907775879,
      "learning_rate": 2.0521235521235522e-05,
      "loss": 1.0061,
      "step": 2946
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.817233085632324,
      "learning_rate": 2.051801801801802e-05,
      "loss": 0.8675,
      "step": 2947
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.084356307983398,
      "learning_rate": 2.0514800514800515e-05,
      "loss": 0.99,
      "step": 2948
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.101424217224121,
      "learning_rate": 2.051158301158301e-05,
      "loss": 0.9234,
      "step": 2949
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.76906681060791,
      "learning_rate": 2.0508365508365508e-05,
      "loss": 1.0328,
      "step": 2950
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.490628242492676,
      "learning_rate": 2.0505148005148005e-05,
      "loss": 1.0067,
      "step": 2951
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.907504081726074,
      "learning_rate": 2.05019305019305e-05,
      "loss": 0.735,
      "step": 2952
    },
    {
      "epoch": 0.63,
      "grad_norm": 12.834376335144043,
      "learning_rate": 2.0498712998712997e-05,
      "loss": 1.1653,
      "step": 2953
    },
    {
      "epoch": 0.63,
      "grad_norm": 10.760860443115234,
      "learning_rate": 2.0495495495495494e-05,
      "loss": 1.1205,
      "step": 2954
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.402727127075195,
      "learning_rate": 2.0492277992277994e-05,
      "loss": 0.9364,
      "step": 2955
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.85688591003418,
      "learning_rate": 2.048906048906049e-05,
      "loss": 0.7816,
      "step": 2956
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.426513671875,
      "learning_rate": 2.0485842985842987e-05,
      "loss": 0.9131,
      "step": 2957
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.050203323364258,
      "learning_rate": 2.0482625482625483e-05,
      "loss": 1.0416,
      "step": 2958
    },
    {
      "epoch": 0.63,
      "grad_norm": 10.818263053894043,
      "learning_rate": 2.047940797940798e-05,
      "loss": 1.0972,
      "step": 2959
    },
    {
      "epoch": 0.63,
      "grad_norm": 10.530221939086914,
      "learning_rate": 2.0476190476190476e-05,
      "loss": 1.1388,
      "step": 2960
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.159333229064941,
      "learning_rate": 2.0472972972972976e-05,
      "loss": 0.901,
      "step": 2961
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.928603172302246,
      "learning_rate": 2.0469755469755472e-05,
      "loss": 1.1079,
      "step": 2962
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.791288375854492,
      "learning_rate": 2.046653796653797e-05,
      "loss": 0.8878,
      "step": 2963
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.410244941711426,
      "learning_rate": 2.0463320463320465e-05,
      "loss": 1.0986,
      "step": 2964
    },
    {
      "epoch": 0.64,
      "grad_norm": 12.333078384399414,
      "learning_rate": 2.046010296010296e-05,
      "loss": 1.1217,
      "step": 2965
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.582412719726562,
      "learning_rate": 2.0456885456885455e-05,
      "loss": 0.9811,
      "step": 2966
    },
    {
      "epoch": 0.64,
      "grad_norm": 13.719084739685059,
      "learning_rate": 2.045366795366795e-05,
      "loss": 1.1136,
      "step": 2967
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.171674728393555,
      "learning_rate": 2.045045045045045e-05,
      "loss": 1.0124,
      "step": 2968
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.54813003540039,
      "learning_rate": 2.0447232947232947e-05,
      "loss": 1.1447,
      "step": 2969
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.963827133178711,
      "learning_rate": 2.0444015444015444e-05,
      "loss": 0.9889,
      "step": 2970
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.631826400756836,
      "learning_rate": 2.044079794079794e-05,
      "loss": 1.0417,
      "step": 2971
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.657465934753418,
      "learning_rate": 2.0437580437580437e-05,
      "loss": 0.957,
      "step": 2972
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.272834777832031,
      "learning_rate": 2.0434362934362933e-05,
      "loss": 0.9047,
      "step": 2973
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.44965934753418,
      "learning_rate": 2.0431145431145433e-05,
      "loss": 1.0465,
      "step": 2974
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.849074363708496,
      "learning_rate": 2.042792792792793e-05,
      "loss": 1.0642,
      "step": 2975
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.327103614807129,
      "learning_rate": 2.0424710424710426e-05,
      "loss": 1.15,
      "step": 2976
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.627284049987793,
      "learning_rate": 2.0421492921492922e-05,
      "loss": 0.936,
      "step": 2977
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.692927360534668,
      "learning_rate": 2.041827541827542e-05,
      "loss": 0.9536,
      "step": 2978
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.791589736938477,
      "learning_rate": 2.0415057915057915e-05,
      "loss": 0.8214,
      "step": 2979
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.170125961303711,
      "learning_rate": 2.0411840411840412e-05,
      "loss": 0.9695,
      "step": 2980
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.460793495178223,
      "learning_rate": 2.040862290862291e-05,
      "loss": 1.0429,
      "step": 2981
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.987873077392578,
      "learning_rate": 2.0405405405405408e-05,
      "loss": 0.8304,
      "step": 2982
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.267940521240234,
      "learning_rate": 2.0402187902187905e-05,
      "loss": 0.831,
      "step": 2983
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.157464981079102,
      "learning_rate": 2.0398970398970398e-05,
      "loss": 1.0298,
      "step": 2984
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.977104663848877,
      "learning_rate": 2.0395752895752894e-05,
      "loss": 0.8061,
      "step": 2985
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.934287071228027,
      "learning_rate": 2.039253539253539e-05,
      "loss": 0.9811,
      "step": 2986
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.365497589111328,
      "learning_rate": 2.038931788931789e-05,
      "loss": 1.0978,
      "step": 2987
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.441845893859863,
      "learning_rate": 2.0386100386100387e-05,
      "loss": 0.9124,
      "step": 2988
    },
    {
      "epoch": 0.64,
      "grad_norm": 15.090591430664062,
      "learning_rate": 2.0382882882882883e-05,
      "loss": 1.1382,
      "step": 2989
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.475778579711914,
      "learning_rate": 2.037966537966538e-05,
      "loss": 0.9009,
      "step": 2990
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.578153610229492,
      "learning_rate": 2.0376447876447876e-05,
      "loss": 0.97,
      "step": 2991
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.236559867858887,
      "learning_rate": 2.0373230373230373e-05,
      "loss": 0.7982,
      "step": 2992
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.407381057739258,
      "learning_rate": 2.037001287001287e-05,
      "loss": 1.2198,
      "step": 2993
    },
    {
      "epoch": 0.64,
      "grad_norm": 11.018396377563477,
      "learning_rate": 2.036679536679537e-05,
      "loss": 1.2168,
      "step": 2994
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.408422470092773,
      "learning_rate": 2.0363577863577865e-05,
      "loss": 1.032,
      "step": 2995
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.490292549133301,
      "learning_rate": 2.0360360360360362e-05,
      "loss": 0.7806,
      "step": 2996
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.604016304016113,
      "learning_rate": 2.0357142857142858e-05,
      "loss": 1.1239,
      "step": 2997
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.627461433410645,
      "learning_rate": 2.0353925353925355e-05,
      "loss": 1.0737,
      "step": 2998
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.465473175048828,
      "learning_rate": 2.035070785070785e-05,
      "loss": 0.8044,
      "step": 2999
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.27293872833252,
      "learning_rate": 2.034749034749035e-05,
      "loss": 0.9318,
      "step": 3000
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.146968841552734,
      "learning_rate": 2.0344272844272844e-05,
      "loss": 0.9434,
      "step": 3001
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.95290756225586,
      "learning_rate": 2.034105534105534e-05,
      "loss": 1.0074,
      "step": 3002
    },
    {
      "epoch": 0.64,
      "grad_norm": 11.392744064331055,
      "learning_rate": 2.0337837837837837e-05,
      "loss": 0.975,
      "step": 3003
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.864361763000488,
      "learning_rate": 2.0334620334620333e-05,
      "loss": 1.1192,
      "step": 3004
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.046892166137695,
      "learning_rate": 2.033140283140283e-05,
      "loss": 1.2165,
      "step": 3005
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.453876495361328,
      "learning_rate": 2.032818532818533e-05,
      "loss": 0.8592,
      "step": 3006
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.19019603729248,
      "learning_rate": 2.0324967824967826e-05,
      "loss": 1.1225,
      "step": 3007
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.287096977233887,
      "learning_rate": 2.0321750321750323e-05,
      "loss": 1.0128,
      "step": 3008
    },
    {
      "epoch": 0.65,
      "grad_norm": 7.486690521240234,
      "learning_rate": 2.031853281853282e-05,
      "loss": 0.6426,
      "step": 3009
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.672645568847656,
      "learning_rate": 2.0315315315315316e-05,
      "loss": 1.0658,
      "step": 3010
    },
    {
      "epoch": 0.65,
      "grad_norm": 7.129611015319824,
      "learning_rate": 2.0312097812097812e-05,
      "loss": 0.5388,
      "step": 3011
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.224874496459961,
      "learning_rate": 2.030888030888031e-05,
      "loss": 0.9542,
      "step": 3012
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.807430267333984,
      "learning_rate": 2.0305662805662808e-05,
      "loss": 0.8772,
      "step": 3013
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.746261596679688,
      "learning_rate": 2.0302445302445305e-05,
      "loss": 0.9916,
      "step": 3014
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.631171226501465,
      "learning_rate": 2.02992277992278e-05,
      "loss": 0.8719,
      "step": 3015
    },
    {
      "epoch": 0.65,
      "grad_norm": 11.879642486572266,
      "learning_rate": 2.0296010296010298e-05,
      "loss": 1.0894,
      "step": 3016
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.935650825500488,
      "learning_rate": 2.0292792792792794e-05,
      "loss": 1.1185,
      "step": 3017
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.90463638305664,
      "learning_rate": 2.028957528957529e-05,
      "loss": 1.1141,
      "step": 3018
    },
    {
      "epoch": 0.65,
      "grad_norm": 13.760469436645508,
      "learning_rate": 2.0286357786357787e-05,
      "loss": 1.3506,
      "step": 3019
    },
    {
      "epoch": 0.65,
      "grad_norm": 11.373562812805176,
      "learning_rate": 2.0283140283140283e-05,
      "loss": 0.9862,
      "step": 3020
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.004410743713379,
      "learning_rate": 2.027992277992278e-05,
      "loss": 1.0375,
      "step": 3021
    },
    {
      "epoch": 0.65,
      "grad_norm": 11.489104270935059,
      "learning_rate": 2.0276705276705276e-05,
      "loss": 1.0411,
      "step": 3022
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.760124206542969,
      "learning_rate": 2.0273487773487773e-05,
      "loss": 0.9374,
      "step": 3023
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.120923042297363,
      "learning_rate": 2.027027027027027e-05,
      "loss": 0.751,
      "step": 3024
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.616927146911621,
      "learning_rate": 2.0267052767052766e-05,
      "loss": 0.9709,
      "step": 3025
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.413202285766602,
      "learning_rate": 2.0263835263835266e-05,
      "loss": 0.6986,
      "step": 3026
    },
    {
      "epoch": 0.65,
      "grad_norm": 11.465656280517578,
      "learning_rate": 2.0260617760617762e-05,
      "loss": 1.3649,
      "step": 3027
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.966650009155273,
      "learning_rate": 2.025740025740026e-05,
      "loss": 0.9699,
      "step": 3028
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.835503578186035,
      "learning_rate": 2.0254182754182755e-05,
      "loss": 0.841,
      "step": 3029
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.442048072814941,
      "learning_rate": 2.025096525096525e-05,
      "loss": 0.9787,
      "step": 3030
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.465666770935059,
      "learning_rate": 2.0247747747747748e-05,
      "loss": 1.0348,
      "step": 3031
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.046353340148926,
      "learning_rate": 2.0244530244530248e-05,
      "loss": 1.1417,
      "step": 3032
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.00559139251709,
      "learning_rate": 2.0241312741312744e-05,
      "loss": 0.8902,
      "step": 3033
    },
    {
      "epoch": 0.65,
      "grad_norm": 7.848540306091309,
      "learning_rate": 2.023809523809524e-05,
      "loss": 0.8448,
      "step": 3034
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.137102127075195,
      "learning_rate": 2.0234877734877737e-05,
      "loss": 1.1653,
      "step": 3035
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.891792297363281,
      "learning_rate": 2.023166023166023e-05,
      "loss": 0.9628,
      "step": 3036
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.02392292022705,
      "learning_rate": 2.0228442728442726e-05,
      "loss": 1.1359,
      "step": 3037
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.746793746948242,
      "learning_rate": 2.0225225225225223e-05,
      "loss": 0.9492,
      "step": 3038
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.60445499420166,
      "learning_rate": 2.0222007722007723e-05,
      "loss": 1.1643,
      "step": 3039
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.006393432617188,
      "learning_rate": 2.021879021879022e-05,
      "loss": 0.8232,
      "step": 3040
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.693599700927734,
      "learning_rate": 2.0215572715572716e-05,
      "loss": 1.0047,
      "step": 3041
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.287642478942871,
      "learning_rate": 2.0212355212355212e-05,
      "loss": 0.9783,
      "step": 3042
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.05869197845459,
      "learning_rate": 2.020913770913771e-05,
      "loss": 0.9087,
      "step": 3043
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.4083833694458,
      "learning_rate": 2.0205920205920205e-05,
      "loss": 0.8565,
      "step": 3044
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.283716201782227,
      "learning_rate": 2.0202702702702705e-05,
      "loss": 0.9957,
      "step": 3045
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.631052017211914,
      "learning_rate": 2.01994851994852e-05,
      "loss": 1.2021,
      "step": 3046
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.059320449829102,
      "learning_rate": 2.0196267696267698e-05,
      "loss": 1.3898,
      "step": 3047
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.895965576171875,
      "learning_rate": 2.0193050193050194e-05,
      "loss": 0.939,
      "step": 3048
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.785161018371582,
      "learning_rate": 2.018983268983269e-05,
      "loss": 0.8765,
      "step": 3049
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.563051223754883,
      "learning_rate": 2.0186615186615187e-05,
      "loss": 0.9069,
      "step": 3050
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.40951156616211,
      "learning_rate": 2.0183397683397684e-05,
      "loss": 1.0787,
      "step": 3051
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.405909538269043,
      "learning_rate": 2.0180180180180183e-05,
      "loss": 0.9609,
      "step": 3052
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.029111862182617,
      "learning_rate": 2.0176962676962676e-05,
      "loss": 1.1812,
      "step": 3053
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.761919975280762,
      "learning_rate": 2.0173745173745173e-05,
      "loss": 1.1908,
      "step": 3054
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.109148025512695,
      "learning_rate": 2.017052767052767e-05,
      "loss": 1.3243,
      "step": 3055
    },
    {
      "epoch": 0.66,
      "grad_norm": 11.031946182250977,
      "learning_rate": 2.0167310167310166e-05,
      "loss": 0.7776,
      "step": 3056
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.940176963806152,
      "learning_rate": 2.0164092664092662e-05,
      "loss": 1.0397,
      "step": 3057
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.997865676879883,
      "learning_rate": 2.0160875160875162e-05,
      "loss": 1.1609,
      "step": 3058
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.765340805053711,
      "learning_rate": 2.015765765765766e-05,
      "loss": 0.6896,
      "step": 3059
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.141935348510742,
      "learning_rate": 2.0154440154440155e-05,
      "loss": 1.1893,
      "step": 3060
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.375442504882812,
      "learning_rate": 2.015122265122265e-05,
      "loss": 1.0659,
      "step": 3061
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.438247680664062,
      "learning_rate": 2.0148005148005148e-05,
      "loss": 1.0957,
      "step": 3062
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.025784492492676,
      "learning_rate": 2.0144787644787644e-05,
      "loss": 1.0651,
      "step": 3063
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.195194244384766,
      "learning_rate": 2.0141570141570144e-05,
      "loss": 1.0222,
      "step": 3064
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.983967781066895,
      "learning_rate": 2.013835263835264e-05,
      "loss": 1.1139,
      "step": 3065
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.095212936401367,
      "learning_rate": 2.0135135135135137e-05,
      "loss": 0.9107,
      "step": 3066
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.610756874084473,
      "learning_rate": 2.0131917631917634e-05,
      "loss": 1.4094,
      "step": 3067
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.07381534576416,
      "learning_rate": 2.012870012870013e-05,
      "loss": 1.2221,
      "step": 3068
    },
    {
      "epoch": 0.66,
      "grad_norm": 7.637281894683838,
      "learning_rate": 2.0125482625482626e-05,
      "loss": 0.704,
      "step": 3069
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.732463836669922,
      "learning_rate": 2.0122265122265123e-05,
      "loss": 1.015,
      "step": 3070
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.274285316467285,
      "learning_rate": 2.011904761904762e-05,
      "loss": 0.9993,
      "step": 3071
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.192408561706543,
      "learning_rate": 2.0115830115830116e-05,
      "loss": 0.8409,
      "step": 3072
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.003390312194824,
      "learning_rate": 2.0112612612612612e-05,
      "loss": 0.8911,
      "step": 3073
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.562407493591309,
      "learning_rate": 2.010939510939511e-05,
      "loss": 0.8256,
      "step": 3074
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.17171573638916,
      "learning_rate": 2.0106177606177605e-05,
      "loss": 0.829,
      "step": 3075
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.512508392333984,
      "learning_rate": 2.01029601029601e-05,
      "loss": 0.93,
      "step": 3076
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.933503150939941,
      "learning_rate": 2.00997425997426e-05,
      "loss": 0.9672,
      "step": 3077
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.5822172164917,
      "learning_rate": 2.0096525096525098e-05,
      "loss": 0.9837,
      "step": 3078
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.562739372253418,
      "learning_rate": 2.0093307593307594e-05,
      "loss": 0.7617,
      "step": 3079
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.46097183227539,
      "learning_rate": 2.009009009009009e-05,
      "loss": 0.7055,
      "step": 3080
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.761731147766113,
      "learning_rate": 2.0086872586872587e-05,
      "loss": 1.0154,
      "step": 3081
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.961593627929688,
      "learning_rate": 2.0083655083655084e-05,
      "loss": 0.877,
      "step": 3082
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.744884490966797,
      "learning_rate": 2.008043758043758e-05,
      "loss": 0.7519,
      "step": 3083
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.975309371948242,
      "learning_rate": 2.007722007722008e-05,
      "loss": 1.2485,
      "step": 3084
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.771092414855957,
      "learning_rate": 2.0074002574002576e-05,
      "loss": 0.8538,
      "step": 3085
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.634799003601074,
      "learning_rate": 2.0070785070785073e-05,
      "loss": 0.9697,
      "step": 3086
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.009596824645996,
      "learning_rate": 2.006756756756757e-05,
      "loss": 0.8264,
      "step": 3087
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.744412422180176,
      "learning_rate": 2.0064350064350062e-05,
      "loss": 1.0404,
      "step": 3088
    },
    {
      "epoch": 0.66,
      "grad_norm": 7.81592321395874,
      "learning_rate": 2.006113256113256e-05,
      "loss": 0.804,
      "step": 3089
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.75333023071289,
      "learning_rate": 2.005791505791506e-05,
      "loss": 1.124,
      "step": 3090
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.215246200561523,
      "learning_rate": 2.0054697554697555e-05,
      "loss": 0.8004,
      "step": 3091
    },
    {
      "epoch": 0.66,
      "grad_norm": 7.430184841156006,
      "learning_rate": 2.005148005148005e-05,
      "loss": 0.7239,
      "step": 3092
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.872352600097656,
      "learning_rate": 2.0048262548262548e-05,
      "loss": 0.9149,
      "step": 3093
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.910978317260742,
      "learning_rate": 2.0045045045045045e-05,
      "loss": 0.9934,
      "step": 3094
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.597336769104004,
      "learning_rate": 2.004182754182754e-05,
      "loss": 1.1244,
      "step": 3095
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.203001976013184,
      "learning_rate": 2.0038610038610037e-05,
      "loss": 0.9008,
      "step": 3096
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.121859550476074,
      "learning_rate": 2.0035392535392537e-05,
      "loss": 1.16,
      "step": 3097
    },
    {
      "epoch": 0.66,
      "grad_norm": 8.32854175567627,
      "learning_rate": 2.0032175032175034e-05,
      "loss": 0.891,
      "step": 3098
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.596735000610352,
      "learning_rate": 2.002895752895753e-05,
      "loss": 1.2562,
      "step": 3099
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.910425186157227,
      "learning_rate": 2.0025740025740027e-05,
      "loss": 1.3787,
      "step": 3100
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.994372367858887,
      "learning_rate": 2.0022522522522523e-05,
      "loss": 0.9925,
      "step": 3101
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.66880989074707,
      "learning_rate": 2.001930501930502e-05,
      "loss": 1.1229,
      "step": 3102
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.854484558105469,
      "learning_rate": 2.001608751608752e-05,
      "loss": 1.0415,
      "step": 3103
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.804615020751953,
      "learning_rate": 2.0012870012870016e-05,
      "loss": 0.8162,
      "step": 3104
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.284963607788086,
      "learning_rate": 2.0009652509652512e-05,
      "loss": 1.0361,
      "step": 3105
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.644410133361816,
      "learning_rate": 2.0006435006435005e-05,
      "loss": 1.0758,
      "step": 3106
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.120142936706543,
      "learning_rate": 2.0003217503217502e-05,
      "loss": 0.9061,
      "step": 3107
    },
    {
      "epoch": 0.67,
      "grad_norm": 12.13682746887207,
      "learning_rate": 1.9999999999999998e-05,
      "loss": 1.3735,
      "step": 3108
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.016490936279297,
      "learning_rate": 1.9996782496782495e-05,
      "loss": 1.1895,
      "step": 3109
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.808987617492676,
      "learning_rate": 1.9993564993564995e-05,
      "loss": 1.0934,
      "step": 3110
    },
    {
      "epoch": 0.67,
      "grad_norm": 7.310636043548584,
      "learning_rate": 1.999034749034749e-05,
      "loss": 0.8172,
      "step": 3111
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.841938018798828,
      "learning_rate": 1.9987129987129987e-05,
      "loss": 1.202,
      "step": 3112
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.368939399719238,
      "learning_rate": 1.9983912483912484e-05,
      "loss": 0.9535,
      "step": 3113
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.79665470123291,
      "learning_rate": 1.998069498069498e-05,
      "loss": 1.4235,
      "step": 3114
    },
    {
      "epoch": 0.67,
      "grad_norm": 13.004173278808594,
      "learning_rate": 1.9977477477477477e-05,
      "loss": 1.0288,
      "step": 3115
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.70864486694336,
      "learning_rate": 1.9974259974259977e-05,
      "loss": 0.9248,
      "step": 3116
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.656638145446777,
      "learning_rate": 1.9971042471042473e-05,
      "loss": 1.0698,
      "step": 3117
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.101217269897461,
      "learning_rate": 1.996782496782497e-05,
      "loss": 1.0454,
      "step": 3118
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.191349983215332,
      "learning_rate": 1.9964607464607466e-05,
      "loss": 1.005,
      "step": 3119
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.051136016845703,
      "learning_rate": 1.9961389961389962e-05,
      "loss": 0.8553,
      "step": 3120
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.33238410949707,
      "learning_rate": 1.995817245817246e-05,
      "loss": 0.983,
      "step": 3121
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.522184371948242,
      "learning_rate": 1.9954954954954955e-05,
      "loss": 0.9479,
      "step": 3122
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.056928634643555,
      "learning_rate": 1.9951737451737452e-05,
      "loss": 1.4338,
      "step": 3123
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.498551368713379,
      "learning_rate": 1.9948519948519948e-05,
      "loss": 1.1422,
      "step": 3124
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.761452674865723,
      "learning_rate": 1.9945302445302445e-05,
      "loss": 1.0439,
      "step": 3125
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.47391414642334,
      "learning_rate": 1.994208494208494e-05,
      "loss": 0.9677,
      "step": 3126
    },
    {
      "epoch": 0.67,
      "grad_norm": 7.887888431549072,
      "learning_rate": 1.9938867438867438e-05,
      "loss": 0.7939,
      "step": 3127
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.95065689086914,
      "learning_rate": 1.9935649935649934e-05,
      "loss": 0.9183,
      "step": 3128
    },
    {
      "epoch": 0.67,
      "grad_norm": 7.409521102905273,
      "learning_rate": 1.9932432432432434e-05,
      "loss": 0.8603,
      "step": 3129
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.305754661560059,
      "learning_rate": 1.992921492921493e-05,
      "loss": 1.0128,
      "step": 3130
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.933366775512695,
      "learning_rate": 1.9925997425997427e-05,
      "loss": 1.0226,
      "step": 3131
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.65665054321289,
      "learning_rate": 1.9922779922779923e-05,
      "loss": 1.1139,
      "step": 3132
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.194319725036621,
      "learning_rate": 1.991956241956242e-05,
      "loss": 1.0466,
      "step": 3133
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.373336791992188,
      "learning_rate": 1.9916344916344916e-05,
      "loss": 1.0594,
      "step": 3134
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.003292083740234,
      "learning_rate": 1.9913127413127416e-05,
      "loss": 0.7271,
      "step": 3135
    },
    {
      "epoch": 0.67,
      "grad_norm": 7.349147796630859,
      "learning_rate": 1.9909909909909912e-05,
      "loss": 0.7887,
      "step": 3136
    },
    {
      "epoch": 0.67,
      "grad_norm": 7.513944625854492,
      "learning_rate": 1.990669240669241e-05,
      "loss": 0.7357,
      "step": 3137
    },
    {
      "epoch": 0.67,
      "grad_norm": 11.722733497619629,
      "learning_rate": 1.9903474903474905e-05,
      "loss": 1.1763,
      "step": 3138
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.037771224975586,
      "learning_rate": 1.9900257400257402e-05,
      "loss": 0.8889,
      "step": 3139
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.958525657653809,
      "learning_rate": 1.9897039897039895e-05,
      "loss": 0.9505,
      "step": 3140
    },
    {
      "epoch": 0.67,
      "grad_norm": 11.509056091308594,
      "learning_rate": 1.989382239382239e-05,
      "loss": 0.9693,
      "step": 3141
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.301621437072754,
      "learning_rate": 1.989060489060489e-05,
      "loss": 1.1414,
      "step": 3142
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.417391777038574,
      "learning_rate": 1.9887387387387388e-05,
      "loss": 0.9712,
      "step": 3143
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.169720649719238,
      "learning_rate": 1.9884169884169884e-05,
      "loss": 1.1169,
      "step": 3144
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.683401107788086,
      "learning_rate": 1.988095238095238e-05,
      "loss": 0.9874,
      "step": 3145
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.952269554138184,
      "learning_rate": 1.9877734877734877e-05,
      "loss": 0.9152,
      "step": 3146
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.53797435760498,
      "learning_rate": 1.9874517374517373e-05,
      "loss": 1.1421,
      "step": 3147
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.667241096496582,
      "learning_rate": 1.9871299871299873e-05,
      "loss": 0.7364,
      "step": 3148
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.466986656188965,
      "learning_rate": 1.986808236808237e-05,
      "loss": 0.7437,
      "step": 3149
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.912762641906738,
      "learning_rate": 1.9864864864864866e-05,
      "loss": 0.9732,
      "step": 3150
    },
    {
      "epoch": 0.68,
      "grad_norm": 11.310691833496094,
      "learning_rate": 1.9861647361647363e-05,
      "loss": 1.4722,
      "step": 3151
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.272686958312988,
      "learning_rate": 1.985842985842986e-05,
      "loss": 0.7822,
      "step": 3152
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.907430171966553,
      "learning_rate": 1.9855212355212356e-05,
      "loss": 0.8523,
      "step": 3153
    },
    {
      "epoch": 0.68,
      "grad_norm": 17.170616149902344,
      "learning_rate": 1.9851994851994852e-05,
      "loss": 0.997,
      "step": 3154
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.28730583190918,
      "learning_rate": 1.9848777348777352e-05,
      "loss": 0.7876,
      "step": 3155
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.671127319335938,
      "learning_rate": 1.9845559845559848e-05,
      "loss": 1.1148,
      "step": 3156
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.907724380493164,
      "learning_rate": 1.9842342342342345e-05,
      "loss": 1.1275,
      "step": 3157
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.773312568664551,
      "learning_rate": 1.9839124839124838e-05,
      "loss": 0.9339,
      "step": 3158
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.838888168334961,
      "learning_rate": 1.9835907335907334e-05,
      "loss": 1.065,
      "step": 3159
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.312332153320312,
      "learning_rate": 1.983268983268983e-05,
      "loss": 0.9391,
      "step": 3160
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.215987205505371,
      "learning_rate": 1.982947232947233e-05,
      "loss": 1.1545,
      "step": 3161
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.0523681640625,
      "learning_rate": 1.9826254826254827e-05,
      "loss": 1.2684,
      "step": 3162
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.228703498840332,
      "learning_rate": 1.9823037323037323e-05,
      "loss": 1.058,
      "step": 3163
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.110681533813477,
      "learning_rate": 1.981981981981982e-05,
      "loss": 1.0023,
      "step": 3164
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.720197677612305,
      "learning_rate": 1.9816602316602316e-05,
      "loss": 1.436,
      "step": 3165
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.977775573730469,
      "learning_rate": 1.9813384813384813e-05,
      "loss": 0.9033,
      "step": 3166
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.385703086853027,
      "learning_rate": 1.981016731016731e-05,
      "loss": 0.9788,
      "step": 3167
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.832379341125488,
      "learning_rate": 1.980694980694981e-05,
      "loss": 0.9242,
      "step": 3168
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.01524543762207,
      "learning_rate": 1.9803732303732305e-05,
      "loss": 0.9169,
      "step": 3169
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.378458976745605,
      "learning_rate": 1.9800514800514802e-05,
      "loss": 0.961,
      "step": 3170
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.59385347366333,
      "learning_rate": 1.97972972972973e-05,
      "loss": 0.9192,
      "step": 3171
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.29855728149414,
      "learning_rate": 1.9794079794079795e-05,
      "loss": 1.1854,
      "step": 3172
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.325169563293457,
      "learning_rate": 1.979086229086229e-05,
      "loss": 1.0269,
      "step": 3173
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.250462532043457,
      "learning_rate": 1.978764478764479e-05,
      "loss": 0.882,
      "step": 3174
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.178813934326172,
      "learning_rate": 1.9784427284427284e-05,
      "loss": 0.855,
      "step": 3175
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.035606384277344,
      "learning_rate": 1.978120978120978e-05,
      "loss": 0.9541,
      "step": 3176
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.434504508972168,
      "learning_rate": 1.9777992277992277e-05,
      "loss": 0.9686,
      "step": 3177
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.744836807250977,
      "learning_rate": 1.9774774774774774e-05,
      "loss": 1.2033,
      "step": 3178
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.943106651306152,
      "learning_rate": 1.977155727155727e-05,
      "loss": 0.941,
      "step": 3179
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.295019149780273,
      "learning_rate": 1.976833976833977e-05,
      "loss": 0.8937,
      "step": 3180
    },
    {
      "epoch": 0.68,
      "grad_norm": 11.076874732971191,
      "learning_rate": 1.9765122265122266e-05,
      "loss": 1.1813,
      "step": 3181
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.775910377502441,
      "learning_rate": 1.9761904761904763e-05,
      "loss": 0.8667,
      "step": 3182
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.46155071258545,
      "learning_rate": 1.975868725868726e-05,
      "loss": 1.1207,
      "step": 3183
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.943379402160645,
      "learning_rate": 1.9755469755469756e-05,
      "loss": 1.0733,
      "step": 3184
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.766715049743652,
      "learning_rate": 1.9752252252252252e-05,
      "loss": 0.9054,
      "step": 3185
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.229316711425781,
      "learning_rate": 1.974903474903475e-05,
      "loss": 1.0276,
      "step": 3186
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.35690975189209,
      "learning_rate": 1.974581724581725e-05,
      "loss": 0.8665,
      "step": 3187
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.639402389526367,
      "learning_rate": 1.9742599742599745e-05,
      "loss": 0.9562,
      "step": 3188
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.720255851745605,
      "learning_rate": 1.973938223938224e-05,
      "loss": 0.7729,
      "step": 3189
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.536179542541504,
      "learning_rate": 1.9736164736164738e-05,
      "loss": 1.1533,
      "step": 3190
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.247875213623047,
      "learning_rate": 1.9732947232947234e-05,
      "loss": 0.9998,
      "step": 3191
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.085362434387207,
      "learning_rate": 1.972972972972973e-05,
      "loss": 0.8417,
      "step": 3192
    },
    {
      "epoch": 0.68,
      "grad_norm": 9.198404312133789,
      "learning_rate": 1.9726512226512227e-05,
      "loss": 0.7539,
      "step": 3193
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.792221069335938,
      "learning_rate": 1.9723294723294724e-05,
      "loss": 0.8528,
      "step": 3194
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.15305233001709,
      "learning_rate": 1.972007722007722e-05,
      "loss": 0.6504,
      "step": 3195
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.159163475036621,
      "learning_rate": 1.9716859716859716e-05,
      "loss": 0.9015,
      "step": 3196
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.53315544128418,
      "learning_rate": 1.9713642213642213e-05,
      "loss": 1.1184,
      "step": 3197
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.77278995513916,
      "learning_rate": 1.971042471042471e-05,
      "loss": 1.1145,
      "step": 3198
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.919008255004883,
      "learning_rate": 1.9707207207207206e-05,
      "loss": 1.414,
      "step": 3199
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.639752388000488,
      "learning_rate": 1.9703989703989706e-05,
      "loss": 0.9474,
      "step": 3200
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.274861335754395,
      "learning_rate": 1.9700772200772202e-05,
      "loss": 1.089,
      "step": 3201
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.90084457397461,
      "learning_rate": 1.96975546975547e-05,
      "loss": 0.857,
      "step": 3202
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.505424499511719,
      "learning_rate": 1.9694337194337195e-05,
      "loss": 1.1631,
      "step": 3203
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.827397346496582,
      "learning_rate": 1.969111969111969e-05,
      "loss": 0.9137,
      "step": 3204
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.99448299407959,
      "learning_rate": 1.9687902187902188e-05,
      "loss": 1.1718,
      "step": 3205
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.786697387695312,
      "learning_rate": 1.9684684684684688e-05,
      "loss": 0.6567,
      "step": 3206
    },
    {
      "epoch": 0.69,
      "grad_norm": 12.15822696685791,
      "learning_rate": 1.9681467181467184e-05,
      "loss": 0.9577,
      "step": 3207
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.422820091247559,
      "learning_rate": 1.967824967824968e-05,
      "loss": 1.0011,
      "step": 3208
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.029374122619629,
      "learning_rate": 1.9675032175032177e-05,
      "loss": 0.8396,
      "step": 3209
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.16115951538086,
      "learning_rate": 1.967181467181467e-05,
      "loss": 0.882,
      "step": 3210
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.344889640808105,
      "learning_rate": 1.9668597168597167e-05,
      "loss": 1.0102,
      "step": 3211
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.690756797790527,
      "learning_rate": 1.9665379665379663e-05,
      "loss": 0.9259,
      "step": 3212
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.16116714477539,
      "learning_rate": 1.9662162162162163e-05,
      "loss": 0.7854,
      "step": 3213
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.431041717529297,
      "learning_rate": 1.965894465894466e-05,
      "loss": 1.1404,
      "step": 3214
    },
    {
      "epoch": 0.69,
      "grad_norm": 11.118450164794922,
      "learning_rate": 1.9655727155727156e-05,
      "loss": 1.4277,
      "step": 3215
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.733738899230957,
      "learning_rate": 1.9652509652509652e-05,
      "loss": 0.9634,
      "step": 3216
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.140009880065918,
      "learning_rate": 1.964929214929215e-05,
      "loss": 0.9018,
      "step": 3217
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.581513404846191,
      "learning_rate": 1.9646074646074645e-05,
      "loss": 0.7746,
      "step": 3218
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.982474327087402,
      "learning_rate": 1.9642857142857145e-05,
      "loss": 1.1801,
      "step": 3219
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.261369705200195,
      "learning_rate": 1.963963963963964e-05,
      "loss": 0.8226,
      "step": 3220
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.47103500366211,
      "learning_rate": 1.9636422136422138e-05,
      "loss": 0.9053,
      "step": 3221
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.092398643493652,
      "learning_rate": 1.9633204633204634e-05,
      "loss": 1.0082,
      "step": 3222
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.353787422180176,
      "learning_rate": 1.962998712998713e-05,
      "loss": 0.9463,
      "step": 3223
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.30117130279541,
      "learning_rate": 1.9626769626769627e-05,
      "loss": 0.7813,
      "step": 3224
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.115694046020508,
      "learning_rate": 1.9623552123552124e-05,
      "loss": 0.8787,
      "step": 3225
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.563942909240723,
      "learning_rate": 1.9620334620334624e-05,
      "loss": 0.924,
      "step": 3226
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.856307029724121,
      "learning_rate": 1.9617117117117117e-05,
      "loss": 0.9687,
      "step": 3227
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.870360374450684,
      "learning_rate": 1.9613899613899613e-05,
      "loss": 1.2539,
      "step": 3228
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.771047592163086,
      "learning_rate": 1.961068211068211e-05,
      "loss": 1.1221,
      "step": 3229
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.370138168334961,
      "learning_rate": 1.9607464607464606e-05,
      "loss": 0.7595,
      "step": 3230
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.960370063781738,
      "learning_rate": 1.9604247104247102e-05,
      "loss": 1.0502,
      "step": 3231
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.431014060974121,
      "learning_rate": 1.9601029601029602e-05,
      "loss": 0.8785,
      "step": 3232
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.83760929107666,
      "learning_rate": 1.95978120978121e-05,
      "loss": 0.8273,
      "step": 3233
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.255045890808105,
      "learning_rate": 1.9594594594594595e-05,
      "loss": 0.8352,
      "step": 3234
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.446429252624512,
      "learning_rate": 1.959137709137709e-05,
      "loss": 0.9105,
      "step": 3235
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.176387786865234,
      "learning_rate": 1.9588159588159588e-05,
      "loss": 1.154,
      "step": 3236
    },
    {
      "epoch": 0.69,
      "grad_norm": 11.447836875915527,
      "learning_rate": 1.9584942084942085e-05,
      "loss": 0.8815,
      "step": 3237
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.604933738708496,
      "learning_rate": 1.9581724581724584e-05,
      "loss": 1.0479,
      "step": 3238
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.426666259765625,
      "learning_rate": 1.957850707850708e-05,
      "loss": 0.8235,
      "step": 3239
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.557924270629883,
      "learning_rate": 1.9575289575289577e-05,
      "loss": 0.8976,
      "step": 3240
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.460597038269043,
      "learning_rate": 1.9572072072072074e-05,
      "loss": 1.0718,
      "step": 3241
    },
    {
      "epoch": 0.7,
      "grad_norm": 12.405274391174316,
      "learning_rate": 1.956885456885457e-05,
      "loss": 1.2147,
      "step": 3242
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.241536140441895,
      "learning_rate": 1.9565637065637067e-05,
      "loss": 0.8598,
      "step": 3243
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.290122032165527,
      "learning_rate": 1.9562419562419563e-05,
      "loss": 0.6893,
      "step": 3244
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.540177345275879,
      "learning_rate": 1.955920205920206e-05,
      "loss": 0.9332,
      "step": 3245
    },
    {
      "epoch": 0.7,
      "grad_norm": 7.9142961502075195,
      "learning_rate": 1.9555984555984556e-05,
      "loss": 0.7628,
      "step": 3246
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.506497383117676,
      "learning_rate": 1.9552767052767052e-05,
      "loss": 0.8789,
      "step": 3247
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.977517127990723,
      "learning_rate": 1.954954954954955e-05,
      "loss": 1.0547,
      "step": 3248
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.16476821899414,
      "learning_rate": 1.9546332046332045e-05,
      "loss": 0.9186,
      "step": 3249
    },
    {
      "epoch": 0.7,
      "grad_norm": 7.684288501739502,
      "learning_rate": 1.9543114543114542e-05,
      "loss": 0.7465,
      "step": 3250
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.106436729431152,
      "learning_rate": 1.953989703989704e-05,
      "loss": 0.7783,
      "step": 3251
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.127667427062988,
      "learning_rate": 1.9536679536679538e-05,
      "loss": 0.7342,
      "step": 3252
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.052237510681152,
      "learning_rate": 1.9533462033462035e-05,
      "loss": 0.7997,
      "step": 3253
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.125075340270996,
      "learning_rate": 1.953024453024453e-05,
      "loss": 0.9445,
      "step": 3254
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.87919807434082,
      "learning_rate": 1.9527027027027027e-05,
      "loss": 1.063,
      "step": 3255
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.820924758911133,
      "learning_rate": 1.9523809523809524e-05,
      "loss": 1.0746,
      "step": 3256
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.683048248291016,
      "learning_rate": 1.952059202059202e-05,
      "loss": 1.1459,
      "step": 3257
    },
    {
      "epoch": 0.7,
      "grad_norm": 11.193612098693848,
      "learning_rate": 1.951737451737452e-05,
      "loss": 1.3002,
      "step": 3258
    },
    {
      "epoch": 0.7,
      "grad_norm": 7.432743072509766,
      "learning_rate": 1.9514157014157017e-05,
      "loss": 0.4905,
      "step": 3259
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.458609580993652,
      "learning_rate": 1.9510939510939513e-05,
      "loss": 0.9299,
      "step": 3260
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.457451820373535,
      "learning_rate": 1.950772200772201e-05,
      "loss": 0.8725,
      "step": 3261
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.841734886169434,
      "learning_rate": 1.9504504504504503e-05,
      "loss": 0.9393,
      "step": 3262
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.43917465209961,
      "learning_rate": 1.9501287001287e-05,
      "loss": 1.0833,
      "step": 3263
    },
    {
      "epoch": 0.7,
      "grad_norm": 12.083383560180664,
      "learning_rate": 1.94980694980695e-05,
      "loss": 1.1384,
      "step": 3264
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.0900297164917,
      "learning_rate": 1.9494851994851995e-05,
      "loss": 0.9263,
      "step": 3265
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.654730796813965,
      "learning_rate": 1.9491634491634492e-05,
      "loss": 0.8536,
      "step": 3266
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.487104415893555,
      "learning_rate": 1.9488416988416988e-05,
      "loss": 0.6031,
      "step": 3267
    },
    {
      "epoch": 0.7,
      "grad_norm": 7.5102739334106445,
      "learning_rate": 1.9485199485199485e-05,
      "loss": 0.6702,
      "step": 3268
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.800542831420898,
      "learning_rate": 1.948198198198198e-05,
      "loss": 1.1264,
      "step": 3269
    },
    {
      "epoch": 0.7,
      "grad_norm": 11.181389808654785,
      "learning_rate": 1.9478764478764478e-05,
      "loss": 1.0749,
      "step": 3270
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.554522514343262,
      "learning_rate": 1.9475546975546977e-05,
      "loss": 0.9719,
      "step": 3271
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.732072830200195,
      "learning_rate": 1.9472329472329474e-05,
      "loss": 1.3334,
      "step": 3272
    },
    {
      "epoch": 0.7,
      "grad_norm": 11.208698272705078,
      "learning_rate": 1.946911196911197e-05,
      "loss": 0.9334,
      "step": 3273
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.090578079223633,
      "learning_rate": 1.9465894465894467e-05,
      "loss": 0.9599,
      "step": 3274
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.197056770324707,
      "learning_rate": 1.9462676962676963e-05,
      "loss": 0.7657,
      "step": 3275
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.67638874053955,
      "learning_rate": 1.945945945945946e-05,
      "loss": 0.9467,
      "step": 3276
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.681506156921387,
      "learning_rate": 1.945624195624196e-05,
      "loss": 1.2135,
      "step": 3277
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.724785804748535,
      "learning_rate": 1.9453024453024456e-05,
      "loss": 0.7812,
      "step": 3278
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.277145385742188,
      "learning_rate": 1.9449806949806952e-05,
      "loss": 1.0459,
      "step": 3279
    },
    {
      "epoch": 0.7,
      "grad_norm": 11.695273399353027,
      "learning_rate": 1.9446589446589445e-05,
      "loss": 1.0084,
      "step": 3280
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.56765079498291,
      "learning_rate": 1.9443371943371942e-05,
      "loss": 0.8521,
      "step": 3281
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.471789360046387,
      "learning_rate": 1.944015444015444e-05,
      "loss": 0.9302,
      "step": 3282
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.449275970458984,
      "learning_rate": 1.9436936936936935e-05,
      "loss": 0.9508,
      "step": 3283
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.558394432067871,
      "learning_rate": 1.9433719433719435e-05,
      "loss": 0.8835,
      "step": 3284
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.922657012939453,
      "learning_rate": 1.943050193050193e-05,
      "loss": 1.1653,
      "step": 3285
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.811367988586426,
      "learning_rate": 1.9427284427284428e-05,
      "loss": 0.9801,
      "step": 3286
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.285799980163574,
      "learning_rate": 1.9424066924066924e-05,
      "loss": 0.781,
      "step": 3287
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.154257774353027,
      "learning_rate": 1.942084942084942e-05,
      "loss": 1.1808,
      "step": 3288
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.609659194946289,
      "learning_rate": 1.9417631917631917e-05,
      "loss": 1.0262,
      "step": 3289
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.30760669708252,
      "learning_rate": 1.9414414414414417e-05,
      "loss": 1.1728,
      "step": 3290
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.774943351745605,
      "learning_rate": 1.9411196911196913e-05,
      "loss": 1.0922,
      "step": 3291
    },
    {
      "epoch": 0.71,
      "grad_norm": 13.315610885620117,
      "learning_rate": 1.940797940797941e-05,
      "loss": 1.025,
      "step": 3292
    },
    {
      "epoch": 0.71,
      "grad_norm": 7.689041614532471,
      "learning_rate": 1.9404761904761906e-05,
      "loss": 0.8505,
      "step": 3293
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.27912425994873,
      "learning_rate": 1.9401544401544403e-05,
      "loss": 0.9552,
      "step": 3294
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.163186073303223,
      "learning_rate": 1.93983268983269e-05,
      "loss": 1.0267,
      "step": 3295
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.013572692871094,
      "learning_rate": 1.93951093951094e-05,
      "loss": 1.1764,
      "step": 3296
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.604583740234375,
      "learning_rate": 1.9391891891891892e-05,
      "loss": 0.9499,
      "step": 3297
    },
    {
      "epoch": 0.71,
      "grad_norm": 7.637176513671875,
      "learning_rate": 1.938867438867439e-05,
      "loss": 0.6817,
      "step": 3298
    },
    {
      "epoch": 0.71,
      "grad_norm": 7.845071792602539,
      "learning_rate": 1.9385456885456885e-05,
      "loss": 0.8127,
      "step": 3299
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.254035949707031,
      "learning_rate": 1.938223938223938e-05,
      "loss": 0.9165,
      "step": 3300
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.835768699645996,
      "learning_rate": 1.9379021879021878e-05,
      "loss": 1.0782,
      "step": 3301
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.083417892456055,
      "learning_rate": 1.9375804375804374e-05,
      "loss": 0.9388,
      "step": 3302
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.385967254638672,
      "learning_rate": 1.9372586872586874e-05,
      "loss": 0.7607,
      "step": 3303
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.000085830688477,
      "learning_rate": 1.936936936936937e-05,
      "loss": 0.795,
      "step": 3304
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.798213958740234,
      "learning_rate": 1.9366151866151867e-05,
      "loss": 1.0788,
      "step": 3305
    },
    {
      "epoch": 0.71,
      "grad_norm": 11.624591827392578,
      "learning_rate": 1.9362934362934363e-05,
      "loss": 1.0634,
      "step": 3306
    },
    {
      "epoch": 0.71,
      "grad_norm": 11.881958961486816,
      "learning_rate": 1.935971685971686e-05,
      "loss": 1.1416,
      "step": 3307
    },
    {
      "epoch": 0.71,
      "grad_norm": 11.798579216003418,
      "learning_rate": 1.9356499356499356e-05,
      "loss": 1.6764,
      "step": 3308
    },
    {
      "epoch": 0.71,
      "grad_norm": 7.850048542022705,
      "learning_rate": 1.9353281853281856e-05,
      "loss": 0.8425,
      "step": 3309
    },
    {
      "epoch": 0.71,
      "grad_norm": 11.379166603088379,
      "learning_rate": 1.9350064350064353e-05,
      "loss": 1.1004,
      "step": 3310
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.471384048461914,
      "learning_rate": 1.934684684684685e-05,
      "loss": 1.1335,
      "step": 3311
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.140440940856934,
      "learning_rate": 1.9343629343629345e-05,
      "loss": 0.7849,
      "step": 3312
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.205530166625977,
      "learning_rate": 1.9340411840411842e-05,
      "loss": 1.1489,
      "step": 3313
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.609013557434082,
      "learning_rate": 1.9337194337194335e-05,
      "loss": 0.9415,
      "step": 3314
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.46311092376709,
      "learning_rate": 1.933397683397683e-05,
      "loss": 1.0977,
      "step": 3315
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.376810073852539,
      "learning_rate": 1.933075933075933e-05,
      "loss": 1.1326,
      "step": 3316
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.959331512451172,
      "learning_rate": 1.9327541827541828e-05,
      "loss": 0.9783,
      "step": 3317
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.351418495178223,
      "learning_rate": 1.9324324324324324e-05,
      "loss": 1.0694,
      "step": 3318
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.211148262023926,
      "learning_rate": 1.932110682110682e-05,
      "loss": 0.7015,
      "step": 3319
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.544819831848145,
      "learning_rate": 1.9317889317889317e-05,
      "loss": 0.8649,
      "step": 3320
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.764226913452148,
      "learning_rate": 1.9314671814671814e-05,
      "loss": 0.7972,
      "step": 3321
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.30016803741455,
      "learning_rate": 1.9311454311454313e-05,
      "loss": 0.7289,
      "step": 3322
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.392220497131348,
      "learning_rate": 1.930823680823681e-05,
      "loss": 1.2456,
      "step": 3323
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.380638122558594,
      "learning_rate": 1.9305019305019306e-05,
      "loss": 0.9564,
      "step": 3324
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.38695240020752,
      "learning_rate": 1.9301801801801803e-05,
      "loss": 0.8463,
      "step": 3325
    },
    {
      "epoch": 0.71,
      "grad_norm": 7.3345441818237305,
      "learning_rate": 1.92985842985843e-05,
      "loss": 0.6341,
      "step": 3326
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.563232421875,
      "learning_rate": 1.9295366795366796e-05,
      "loss": 1.0407,
      "step": 3327
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.424884796142578,
      "learning_rate": 1.9292149292149292e-05,
      "loss": 0.7694,
      "step": 3328
    },
    {
      "epoch": 0.71,
      "grad_norm": 8.699213027954102,
      "learning_rate": 1.9288931788931792e-05,
      "loss": 0.636,
      "step": 3329
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.796092987060547,
      "learning_rate": 1.928571428571429e-05,
      "loss": 0.7871,
      "step": 3330
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.265336036682129,
      "learning_rate": 1.9282496782496785e-05,
      "loss": 0.962,
      "step": 3331
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.553380966186523,
      "learning_rate": 1.9279279279279278e-05,
      "loss": 0.7874,
      "step": 3332
    },
    {
      "epoch": 0.71,
      "grad_norm": 11.979629516601562,
      "learning_rate": 1.9276061776061774e-05,
      "loss": 1.2842,
      "step": 3333
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.430025100708008,
      "learning_rate": 1.927284427284427e-05,
      "loss": 0.8807,
      "step": 3334
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.38812255859375,
      "learning_rate": 1.926962676962677e-05,
      "loss": 0.7355,
      "step": 3335
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.789410591125488,
      "learning_rate": 1.9266409266409267e-05,
      "loss": 0.6994,
      "step": 3336
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.397811889648438,
      "learning_rate": 1.9263191763191764e-05,
      "loss": 0.7706,
      "step": 3337
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.34003734588623,
      "learning_rate": 1.925997425997426e-05,
      "loss": 0.9771,
      "step": 3338
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.898310661315918,
      "learning_rate": 1.9256756756756756e-05,
      "loss": 0.8565,
      "step": 3339
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.177764892578125,
      "learning_rate": 1.9253539253539253e-05,
      "loss": 1.0732,
      "step": 3340
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.715775489807129,
      "learning_rate": 1.925032175032175e-05,
      "loss": 0.9156,
      "step": 3341
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.89067268371582,
      "learning_rate": 1.924710424710425e-05,
      "loss": 0.8631,
      "step": 3342
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.364720344543457,
      "learning_rate": 1.9243886743886746e-05,
      "loss": 1.1657,
      "step": 3343
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.77737808227539,
      "learning_rate": 1.9240669240669242e-05,
      "loss": 1.1226,
      "step": 3344
    },
    {
      "epoch": 0.72,
      "grad_norm": 7.346846103668213,
      "learning_rate": 1.923745173745174e-05,
      "loss": 0.6879,
      "step": 3345
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.80948543548584,
      "learning_rate": 1.9234234234234235e-05,
      "loss": 0.8316,
      "step": 3346
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.495645523071289,
      "learning_rate": 1.923101673101673e-05,
      "loss": 1.0404,
      "step": 3347
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.283645629882812,
      "learning_rate": 1.922779922779923e-05,
      "loss": 1.0764,
      "step": 3348
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.167673110961914,
      "learning_rate": 1.9224581724581724e-05,
      "loss": 1.1244,
      "step": 3349
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.930923461914062,
      "learning_rate": 1.922136422136422e-05,
      "loss": 0.9953,
      "step": 3350
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.25959587097168,
      "learning_rate": 1.9218146718146717e-05,
      "loss": 0.7711,
      "step": 3351
    },
    {
      "epoch": 0.72,
      "grad_norm": 7.201839447021484,
      "learning_rate": 1.9214929214929214e-05,
      "loss": 0.9636,
      "step": 3352
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.362531661987305,
      "learning_rate": 1.921171171171171e-05,
      "loss": 1.0125,
      "step": 3353
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.063499450683594,
      "learning_rate": 1.920849420849421e-05,
      "loss": 0.911,
      "step": 3354
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.614521026611328,
      "learning_rate": 1.9205276705276706e-05,
      "loss": 1.0462,
      "step": 3355
    },
    {
      "epoch": 0.72,
      "grad_norm": 7.776280879974365,
      "learning_rate": 1.9202059202059203e-05,
      "loss": 0.8501,
      "step": 3356
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.710615158081055,
      "learning_rate": 1.91988416988417e-05,
      "loss": 0.8799,
      "step": 3357
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.63703441619873,
      "learning_rate": 1.9195624195624196e-05,
      "loss": 0.9751,
      "step": 3358
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.83048152923584,
      "learning_rate": 1.9192406692406692e-05,
      "loss": 0.8833,
      "step": 3359
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.283101081848145,
      "learning_rate": 1.918918918918919e-05,
      "loss": 0.863,
      "step": 3360
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.146061897277832,
      "learning_rate": 1.918597168597169e-05,
      "loss": 0.9602,
      "step": 3361
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.524723052978516,
      "learning_rate": 1.9182754182754185e-05,
      "loss": 1.2762,
      "step": 3362
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.343758583068848,
      "learning_rate": 1.917953667953668e-05,
      "loss": 1.0542,
      "step": 3363
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.493470191955566,
      "learning_rate": 1.9176319176319178e-05,
      "loss": 0.9316,
      "step": 3364
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.36383056640625,
      "learning_rate": 1.9173101673101674e-05,
      "loss": 0.7674,
      "step": 3365
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.52949333190918,
      "learning_rate": 1.916988416988417e-05,
      "loss": 1.0864,
      "step": 3366
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.97976303100586,
      "learning_rate": 1.9166666666666667e-05,
      "loss": 0.7717,
      "step": 3367
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.194219589233398,
      "learning_rate": 1.9163449163449164e-05,
      "loss": 1.0631,
      "step": 3368
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.4571533203125,
      "learning_rate": 1.916023166023166e-05,
      "loss": 1.1822,
      "step": 3369
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.904973983764648,
      "learning_rate": 1.9157014157014157e-05,
      "loss": 0.7335,
      "step": 3370
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.373472213745117,
      "learning_rate": 1.9153796653796653e-05,
      "loss": 0.9261,
      "step": 3371
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.931585311889648,
      "learning_rate": 1.915057915057915e-05,
      "loss": 0.9842,
      "step": 3372
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.6996431350708,
      "learning_rate": 1.9147361647361646e-05,
      "loss": 1.0099,
      "step": 3373
    },
    {
      "epoch": 0.72,
      "grad_norm": 12.0533447265625,
      "learning_rate": 1.9144144144144146e-05,
      "loss": 1.2982,
      "step": 3374
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.167570114135742,
      "learning_rate": 1.9140926640926642e-05,
      "loss": 0.7776,
      "step": 3375
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.595728874206543,
      "learning_rate": 1.913770913770914e-05,
      "loss": 0.8003,
      "step": 3376
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.579943656921387,
      "learning_rate": 1.9134491634491635e-05,
      "loss": 0.8915,
      "step": 3377
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.491996765136719,
      "learning_rate": 1.913127413127413e-05,
      "loss": 1.2723,
      "step": 3378
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.276615142822266,
      "learning_rate": 1.9128056628056628e-05,
      "loss": 1.0504,
      "step": 3379
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.704904556274414,
      "learning_rate": 1.9124839124839128e-05,
      "loss": 0.9932,
      "step": 3380
    },
    {
      "epoch": 0.73,
      "grad_norm": 11.870092391967773,
      "learning_rate": 1.9121621621621624e-05,
      "loss": 1.1212,
      "step": 3381
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.829062461853027,
      "learning_rate": 1.911840411840412e-05,
      "loss": 1.1993,
      "step": 3382
    },
    {
      "epoch": 0.73,
      "grad_norm": 11.225994110107422,
      "learning_rate": 1.9115186615186617e-05,
      "loss": 1.1731,
      "step": 3383
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.935430526733398,
      "learning_rate": 1.911196911196911e-05,
      "loss": 1.0475,
      "step": 3384
    },
    {
      "epoch": 0.73,
      "grad_norm": 11.107146263122559,
      "learning_rate": 1.9108751608751607e-05,
      "loss": 1.1942,
      "step": 3385
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.774248123168945,
      "learning_rate": 1.9105534105534103e-05,
      "loss": 0.8708,
      "step": 3386
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.94216251373291,
      "learning_rate": 1.9102316602316603e-05,
      "loss": 0.9987,
      "step": 3387
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.826024055480957,
      "learning_rate": 1.90990990990991e-05,
      "loss": 0.8809,
      "step": 3388
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.90854549407959,
      "learning_rate": 1.9095881595881596e-05,
      "loss": 0.9397,
      "step": 3389
    },
    {
      "epoch": 0.73,
      "grad_norm": 7.833549499511719,
      "learning_rate": 1.9092664092664092e-05,
      "loss": 0.7084,
      "step": 3390
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.896903991699219,
      "learning_rate": 1.908944658944659e-05,
      "loss": 0.9665,
      "step": 3391
    },
    {
      "epoch": 0.73,
      "grad_norm": 11.33597183227539,
      "learning_rate": 1.9086229086229085e-05,
      "loss": 0.8562,
      "step": 3392
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.075540542602539,
      "learning_rate": 1.9083011583011585e-05,
      "loss": 0.7613,
      "step": 3393
    },
    {
      "epoch": 0.73,
      "grad_norm": 7.701529026031494,
      "learning_rate": 1.907979407979408e-05,
      "loss": 0.65,
      "step": 3394
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.158354759216309,
      "learning_rate": 1.9076576576576578e-05,
      "loss": 1.051,
      "step": 3395
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.379115104675293,
      "learning_rate": 1.9073359073359075e-05,
      "loss": 0.8897,
      "step": 3396
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.379525184631348,
      "learning_rate": 1.907014157014157e-05,
      "loss": 1.002,
      "step": 3397
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.848416328430176,
      "learning_rate": 1.9066924066924067e-05,
      "loss": 0.91,
      "step": 3398
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.15208625793457,
      "learning_rate": 1.9063706563706564e-05,
      "loss": 0.9611,
      "step": 3399
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.004735946655273,
      "learning_rate": 1.9060489060489064e-05,
      "loss": 0.8011,
      "step": 3400
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.805068016052246,
      "learning_rate": 1.9057271557271557e-05,
      "loss": 0.8838,
      "step": 3401
    },
    {
      "epoch": 0.73,
      "grad_norm": 11.745990753173828,
      "learning_rate": 1.9054054054054053e-05,
      "loss": 0.9087,
      "step": 3402
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.444391250610352,
      "learning_rate": 1.905083655083655e-05,
      "loss": 0.8744,
      "step": 3403
    },
    {
      "epoch": 0.73,
      "grad_norm": 11.813033103942871,
      "learning_rate": 1.9047619047619046e-05,
      "loss": 1.3879,
      "step": 3404
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.155762672424316,
      "learning_rate": 1.9044401544401543e-05,
      "loss": 0.8154,
      "step": 3405
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.75130558013916,
      "learning_rate": 1.9041184041184042e-05,
      "loss": 0.8962,
      "step": 3406
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.992424964904785,
      "learning_rate": 1.903796653796654e-05,
      "loss": 0.9226,
      "step": 3407
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.500812530517578,
      "learning_rate": 1.9034749034749035e-05,
      "loss": 1.0292,
      "step": 3408
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.157790184020996,
      "learning_rate": 1.9031531531531532e-05,
      "loss": 0.6279,
      "step": 3409
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.472347259521484,
      "learning_rate": 1.9028314028314028e-05,
      "loss": 0.8425,
      "step": 3410
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.005607604980469,
      "learning_rate": 1.9025096525096525e-05,
      "loss": 0.7669,
      "step": 3411
    },
    {
      "epoch": 0.73,
      "grad_norm": 12.097182273864746,
      "learning_rate": 1.9021879021879024e-05,
      "loss": 1.1552,
      "step": 3412
    },
    {
      "epoch": 0.73,
      "grad_norm": 6.997997760772705,
      "learning_rate": 1.901866151866152e-05,
      "loss": 0.6667,
      "step": 3413
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.423616409301758,
      "learning_rate": 1.9015444015444017e-05,
      "loss": 0.9862,
      "step": 3414
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.17009449005127,
      "learning_rate": 1.9012226512226514e-05,
      "loss": 0.9672,
      "step": 3415
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.237119674682617,
      "learning_rate": 1.900900900900901e-05,
      "loss": 0.9115,
      "step": 3416
    },
    {
      "epoch": 0.73,
      "grad_norm": 12.11893081665039,
      "learning_rate": 1.9005791505791507e-05,
      "loss": 1.1968,
      "step": 3417
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.531949996948242,
      "learning_rate": 1.9002574002574003e-05,
      "loss": 1.0062,
      "step": 3418
    },
    {
      "epoch": 0.73,
      "grad_norm": 7.833646297454834,
      "learning_rate": 1.89993564993565e-05,
      "loss": 0.7074,
      "step": 3419
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.51701545715332,
      "learning_rate": 1.8996138996138996e-05,
      "loss": 0.9966,
      "step": 3420
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.765752792358398,
      "learning_rate": 1.8992921492921493e-05,
      "loss": 0.7689,
      "step": 3421
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.628771781921387,
      "learning_rate": 1.898970398970399e-05,
      "loss": 1.0715,
      "step": 3422
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.054073333740234,
      "learning_rate": 1.8986486486486485e-05,
      "loss": 0.7487,
      "step": 3423
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.363926887512207,
      "learning_rate": 1.8983268983268982e-05,
      "loss": 1.2013,
      "step": 3424
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.10013198852539,
      "learning_rate": 1.8980051480051482e-05,
      "loss": 0.7346,
      "step": 3425
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.731948852539062,
      "learning_rate": 1.8976833976833978e-05,
      "loss": 1.0194,
      "step": 3426
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.503530502319336,
      "learning_rate": 1.8973616473616475e-05,
      "loss": 1.2515,
      "step": 3427
    },
    {
      "epoch": 0.74,
      "grad_norm": 7.977194786071777,
      "learning_rate": 1.897039897039897e-05,
      "loss": 0.8022,
      "step": 3428
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.568246841430664,
      "learning_rate": 1.8967181467181468e-05,
      "loss": 0.9551,
      "step": 3429
    },
    {
      "epoch": 0.74,
      "grad_norm": 11.966891288757324,
      "learning_rate": 1.8963963963963964e-05,
      "loss": 1.0761,
      "step": 3430
    },
    {
      "epoch": 0.74,
      "grad_norm": 11.299346923828125,
      "learning_rate": 1.896074646074646e-05,
      "loss": 1.0973,
      "step": 3431
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.417217254638672,
      "learning_rate": 1.895752895752896e-05,
      "loss": 0.7348,
      "step": 3432
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.089466094970703,
      "learning_rate": 1.8954311454311457e-05,
      "loss": 1.0535,
      "step": 3433
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.574955940246582,
      "learning_rate": 1.8951093951093953e-05,
      "loss": 0.8454,
      "step": 3434
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.742995262145996,
      "learning_rate": 1.894787644787645e-05,
      "loss": 0.8725,
      "step": 3435
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.525208473205566,
      "learning_rate": 1.8944658944658943e-05,
      "loss": 0.7554,
      "step": 3436
    },
    {
      "epoch": 0.74,
      "grad_norm": 7.299197673797607,
      "learning_rate": 1.894144144144144e-05,
      "loss": 0.6198,
      "step": 3437
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.980572700500488,
      "learning_rate": 1.893822393822394e-05,
      "loss": 1.0154,
      "step": 3438
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.949200630187988,
      "learning_rate": 1.8935006435006435e-05,
      "loss": 0.8467,
      "step": 3439
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.715188026428223,
      "learning_rate": 1.8931788931788932e-05,
      "loss": 0.9582,
      "step": 3440
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.998021125793457,
      "learning_rate": 1.892857142857143e-05,
      "loss": 1.0336,
      "step": 3441
    },
    {
      "epoch": 0.74,
      "grad_norm": 14.137299537658691,
      "learning_rate": 1.8925353925353925e-05,
      "loss": 0.9376,
      "step": 3442
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.802117347717285,
      "learning_rate": 1.892213642213642e-05,
      "loss": 0.8593,
      "step": 3443
    },
    {
      "epoch": 0.74,
      "grad_norm": 11.10088062286377,
      "learning_rate": 1.8918918918918918e-05,
      "loss": 1.1319,
      "step": 3444
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.18649673461914,
      "learning_rate": 1.8915701415701418e-05,
      "loss": 0.8211,
      "step": 3445
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.560383796691895,
      "learning_rate": 1.8912483912483914e-05,
      "loss": 1.0181,
      "step": 3446
    },
    {
      "epoch": 0.74,
      "grad_norm": 13.08166217803955,
      "learning_rate": 1.890926640926641e-05,
      "loss": 1.3712,
      "step": 3447
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.252805709838867,
      "learning_rate": 1.8906048906048907e-05,
      "loss": 0.9473,
      "step": 3448
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.636494636535645,
      "learning_rate": 1.8902831402831403e-05,
      "loss": 0.9383,
      "step": 3449
    },
    {
      "epoch": 0.74,
      "grad_norm": 11.383041381835938,
      "learning_rate": 1.88996138996139e-05,
      "loss": 0.8431,
      "step": 3450
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.106830596923828,
      "learning_rate": 1.88963963963964e-05,
      "loss": 0.699,
      "step": 3451
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.664682388305664,
      "learning_rate": 1.8893178893178896e-05,
      "loss": 0.8664,
      "step": 3452
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.193687438964844,
      "learning_rate": 1.8889961389961393e-05,
      "loss": 0.9784,
      "step": 3453
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.596304893493652,
      "learning_rate": 1.8886743886743886e-05,
      "loss": 1.0041,
      "step": 3454
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.686613082885742,
      "learning_rate": 1.8883526383526382e-05,
      "loss": 0.851,
      "step": 3455
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.209312438964844,
      "learning_rate": 1.888030888030888e-05,
      "loss": 0.7969,
      "step": 3456
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.495260238647461,
      "learning_rate": 1.8877091377091375e-05,
      "loss": 0.6885,
      "step": 3457
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.264253616333008,
      "learning_rate": 1.8873873873873875e-05,
      "loss": 0.8483,
      "step": 3458
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.8281888961792,
      "learning_rate": 1.887065637065637e-05,
      "loss": 0.9209,
      "step": 3459
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.680594444274902,
      "learning_rate": 1.8867438867438868e-05,
      "loss": 1.0353,
      "step": 3460
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.515488624572754,
      "learning_rate": 1.8864221364221364e-05,
      "loss": 0.8618,
      "step": 3461
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.547760963439941,
      "learning_rate": 1.886100386100386e-05,
      "loss": 0.8078,
      "step": 3462
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.826253890991211,
      "learning_rate": 1.8857786357786357e-05,
      "loss": 0.6953,
      "step": 3463
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.48820972442627,
      "learning_rate": 1.8854568854568857e-05,
      "loss": 0.9748,
      "step": 3464
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.030888557434082,
      "learning_rate": 1.8851351351351353e-05,
      "loss": 0.8107,
      "step": 3465
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.87478256225586,
      "learning_rate": 1.884813384813385e-05,
      "loss": 1.1787,
      "step": 3466
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.210260391235352,
      "learning_rate": 1.8844916344916346e-05,
      "loss": 1.0247,
      "step": 3467
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.411542892456055,
      "learning_rate": 1.8841698841698843e-05,
      "loss": 1.3546,
      "step": 3468
    },
    {
      "epoch": 0.74,
      "grad_norm": 8.044662475585938,
      "learning_rate": 1.883848133848134e-05,
      "loss": 0.8367,
      "step": 3469
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.132576942443848,
      "learning_rate": 1.883526383526384e-05,
      "loss": 1.1836,
      "step": 3470
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.670833587646484,
      "learning_rate": 1.8832046332046332e-05,
      "loss": 0.878,
      "step": 3471
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.007091522216797,
      "learning_rate": 1.882882882882883e-05,
      "loss": 0.9669,
      "step": 3472
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.34863567352295,
      "learning_rate": 1.8825611325611325e-05,
      "loss": 0.958,
      "step": 3473
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.569668769836426,
      "learning_rate": 1.882239382239382e-05,
      "loss": 0.8166,
      "step": 3474
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.15341567993164,
      "learning_rate": 1.8819176319176318e-05,
      "loss": 0.9244,
      "step": 3475
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.80940055847168,
      "learning_rate": 1.8815958815958814e-05,
      "loss": 0.8387,
      "step": 3476
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.847710609436035,
      "learning_rate": 1.8812741312741314e-05,
      "loss": 1.1527,
      "step": 3477
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.536323547363281,
      "learning_rate": 1.880952380952381e-05,
      "loss": 0.8108,
      "step": 3478
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.884055137634277,
      "learning_rate": 1.8806306306306307e-05,
      "loss": 0.8959,
      "step": 3479
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.710742950439453,
      "learning_rate": 1.8803088803088804e-05,
      "loss": 0.677,
      "step": 3480
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.815295219421387,
      "learning_rate": 1.87998712998713e-05,
      "loss": 1.2064,
      "step": 3481
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.069957733154297,
      "learning_rate": 1.8796653796653796e-05,
      "loss": 0.8362,
      "step": 3482
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.342684745788574,
      "learning_rate": 1.8793436293436296e-05,
      "loss": 0.9813,
      "step": 3483
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.560602188110352,
      "learning_rate": 1.8790218790218793e-05,
      "loss": 0.7734,
      "step": 3484
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.961226463317871,
      "learning_rate": 1.878700128700129e-05,
      "loss": 0.9439,
      "step": 3485
    },
    {
      "epoch": 0.75,
      "grad_norm": 11.184283256530762,
      "learning_rate": 1.8783783783783786e-05,
      "loss": 1.1163,
      "step": 3486
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.378899574279785,
      "learning_rate": 1.8780566280566282e-05,
      "loss": 0.9908,
      "step": 3487
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.041157722473145,
      "learning_rate": 1.8777348777348775e-05,
      "loss": 0.9562,
      "step": 3488
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.754490852355957,
      "learning_rate": 1.877413127413127e-05,
      "loss": 0.8558,
      "step": 3489
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.527295112609863,
      "learning_rate": 1.877091377091377e-05,
      "loss": 0.9171,
      "step": 3490
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.485262870788574,
      "learning_rate": 1.8767696267696268e-05,
      "loss": 0.8139,
      "step": 3491
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.801746368408203,
      "learning_rate": 1.8764478764478764e-05,
      "loss": 0.9162,
      "step": 3492
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.337080955505371,
      "learning_rate": 1.876126126126126e-05,
      "loss": 0.9097,
      "step": 3493
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.546799659729004,
      "learning_rate": 1.8758043758043757e-05,
      "loss": 0.9436,
      "step": 3494
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.903372764587402,
      "learning_rate": 1.8754826254826254e-05,
      "loss": 0.9782,
      "step": 3495
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.620351791381836,
      "learning_rate": 1.8751608751608754e-05,
      "loss": 0.8828,
      "step": 3496
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.07030963897705,
      "learning_rate": 1.874839124839125e-05,
      "loss": 0.7632,
      "step": 3497
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.314420700073242,
      "learning_rate": 1.8745173745173746e-05,
      "loss": 1.0494,
      "step": 3498
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.440428733825684,
      "learning_rate": 1.8741956241956243e-05,
      "loss": 0.9038,
      "step": 3499
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.554862022399902,
      "learning_rate": 1.873873873873874e-05,
      "loss": 0.8056,
      "step": 3500
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.670082092285156,
      "learning_rate": 1.8735521235521236e-05,
      "loss": 0.8232,
      "step": 3501
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.47192668914795,
      "learning_rate": 1.8732303732303732e-05,
      "loss": 1.0649,
      "step": 3502
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.37265682220459,
      "learning_rate": 1.8729086229086232e-05,
      "loss": 1.0709,
      "step": 3503
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.353382110595703,
      "learning_rate": 1.872586872586873e-05,
      "loss": 0.629,
      "step": 3504
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.041901588439941,
      "learning_rate": 1.8722651222651225e-05,
      "loss": 0.9055,
      "step": 3505
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.166107177734375,
      "learning_rate": 1.8719433719433718e-05,
      "loss": 1.0115,
      "step": 3506
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.715058326721191,
      "learning_rate": 1.8716216216216214e-05,
      "loss": 0.9603,
      "step": 3507
    },
    {
      "epoch": 0.75,
      "grad_norm": 7.047022819519043,
      "learning_rate": 1.871299871299871e-05,
      "loss": 0.7582,
      "step": 3508
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.334434509277344,
      "learning_rate": 1.870978120978121e-05,
      "loss": 0.6857,
      "step": 3509
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.96423053741455,
      "learning_rate": 1.8706563706563707e-05,
      "loss": 0.9708,
      "step": 3510
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.050304412841797,
      "learning_rate": 1.8703346203346204e-05,
      "loss": 0.9653,
      "step": 3511
    },
    {
      "epoch": 0.75,
      "grad_norm": 11.066999435424805,
      "learning_rate": 1.87001287001287e-05,
      "loss": 0.9543,
      "step": 3512
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.978850364685059,
      "learning_rate": 1.8696911196911197e-05,
      "loss": 0.869,
      "step": 3513
    },
    {
      "epoch": 0.75,
      "grad_norm": 11.436169624328613,
      "learning_rate": 1.8693693693693693e-05,
      "loss": 1.2473,
      "step": 3514
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.827529907226562,
      "learning_rate": 1.869047619047619e-05,
      "loss": 0.9224,
      "step": 3515
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.28687858581543,
      "learning_rate": 1.868725868725869e-05,
      "loss": 0.8813,
      "step": 3516
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.683738708496094,
      "learning_rate": 1.8684041184041186e-05,
      "loss": 0.7928,
      "step": 3517
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.593374252319336,
      "learning_rate": 1.8680823680823682e-05,
      "loss": 1.0323,
      "step": 3518
    },
    {
      "epoch": 0.75,
      "grad_norm": 8.307929992675781,
      "learning_rate": 1.867760617760618e-05,
      "loss": 0.8309,
      "step": 3519
    },
    {
      "epoch": 0.76,
      "grad_norm": 11.371297836303711,
      "learning_rate": 1.8674388674388675e-05,
      "loss": 1.0543,
      "step": 3520
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.804614067077637,
      "learning_rate": 1.867117117117117e-05,
      "loss": 1.2779,
      "step": 3521
    },
    {
      "epoch": 0.76,
      "grad_norm": 7.348623752593994,
      "learning_rate": 1.866795366795367e-05,
      "loss": 0.5823,
      "step": 3522
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.307595252990723,
      "learning_rate": 1.8664736164736164e-05,
      "loss": 0.7521,
      "step": 3523
    },
    {
      "epoch": 0.76,
      "grad_norm": 12.170893669128418,
      "learning_rate": 1.866151866151866e-05,
      "loss": 1.2653,
      "step": 3524
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.289901733398438,
      "learning_rate": 1.8658301158301157e-05,
      "loss": 0.8008,
      "step": 3525
    },
    {
      "epoch": 0.76,
      "grad_norm": 11.151042938232422,
      "learning_rate": 1.8655083655083654e-05,
      "loss": 1.0749,
      "step": 3526
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.78821086883545,
      "learning_rate": 1.865186615186615e-05,
      "loss": 0.9534,
      "step": 3527
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.162336349487305,
      "learning_rate": 1.864864864864865e-05,
      "loss": 1.1213,
      "step": 3528
    },
    {
      "epoch": 0.76,
      "grad_norm": 12.187725067138672,
      "learning_rate": 1.8645431145431147e-05,
      "loss": 1.0103,
      "step": 3529
    },
    {
      "epoch": 0.76,
      "grad_norm": 7.87891149520874,
      "learning_rate": 1.8642213642213643e-05,
      "loss": 0.7665,
      "step": 3530
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.656694412231445,
      "learning_rate": 1.863899613899614e-05,
      "loss": 0.9403,
      "step": 3531
    },
    {
      "epoch": 0.76,
      "grad_norm": 7.928941249847412,
      "learning_rate": 1.8635778635778636e-05,
      "loss": 0.8694,
      "step": 3532
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.229612350463867,
      "learning_rate": 1.8632561132561132e-05,
      "loss": 1.1175,
      "step": 3533
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.619232177734375,
      "learning_rate": 1.862934362934363e-05,
      "loss": 0.6781,
      "step": 3534
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.746423721313477,
      "learning_rate": 1.862612612612613e-05,
      "loss": 0.7693,
      "step": 3535
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.471336364746094,
      "learning_rate": 1.8622908622908625e-05,
      "loss": 0.8801,
      "step": 3536
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.650364875793457,
      "learning_rate": 1.861969111969112e-05,
      "loss": 1.2227,
      "step": 3537
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.097644805908203,
      "learning_rate": 1.8616473616473618e-05,
      "loss": 1.1152,
      "step": 3538
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.789718627929688,
      "learning_rate": 1.8613256113256114e-05,
      "loss": 1.0177,
      "step": 3539
    },
    {
      "epoch": 0.76,
      "grad_norm": 7.823160171508789,
      "learning_rate": 1.8610038610038608e-05,
      "loss": 0.7252,
      "step": 3540
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.229344367980957,
      "learning_rate": 1.8606821106821107e-05,
      "loss": 0.9648,
      "step": 3541
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.850201606750488,
      "learning_rate": 1.8603603603603604e-05,
      "loss": 0.9093,
      "step": 3542
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.326098442077637,
      "learning_rate": 1.86003861003861e-05,
      "loss": 1.0488,
      "step": 3543
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.48873519897461,
      "learning_rate": 1.8597168597168597e-05,
      "loss": 0.7329,
      "step": 3544
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.141472816467285,
      "learning_rate": 1.8593951093951093e-05,
      "loss": 0.8178,
      "step": 3545
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.984779357910156,
      "learning_rate": 1.859073359073359e-05,
      "loss": 0.9464,
      "step": 3546
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.430809020996094,
      "learning_rate": 1.8587516087516086e-05,
      "loss": 0.7911,
      "step": 3547
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.29464340209961,
      "learning_rate": 1.8584298584298586e-05,
      "loss": 0.7028,
      "step": 3548
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.168835639953613,
      "learning_rate": 1.8581081081081082e-05,
      "loss": 0.6795,
      "step": 3549
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.96800708770752,
      "learning_rate": 1.857786357786358e-05,
      "loss": 0.8862,
      "step": 3550
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.186220169067383,
      "learning_rate": 1.8574646074646075e-05,
      "loss": 1.1709,
      "step": 3551
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.76264762878418,
      "learning_rate": 1.8571428571428572e-05,
      "loss": 0.866,
      "step": 3552
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.632277488708496,
      "learning_rate": 1.8568211068211068e-05,
      "loss": 0.8288,
      "step": 3553
    },
    {
      "epoch": 0.76,
      "grad_norm": 11.300680160522461,
      "learning_rate": 1.8564993564993568e-05,
      "loss": 0.9719,
      "step": 3554
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.840197563171387,
      "learning_rate": 1.8561776061776064e-05,
      "loss": 0.9555,
      "step": 3555
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.830004692077637,
      "learning_rate": 1.855855855855856e-05,
      "loss": 1.1195,
      "step": 3556
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.91372299194336,
      "learning_rate": 1.8555341055341057e-05,
      "loss": 0.9478,
      "step": 3557
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.00920295715332,
      "learning_rate": 1.855212355212355e-05,
      "loss": 0.9931,
      "step": 3558
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.889137268066406,
      "learning_rate": 1.8548906048906047e-05,
      "loss": 0.9713,
      "step": 3559
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.733648300170898,
      "learning_rate": 1.8545688545688543e-05,
      "loss": 1.0053,
      "step": 3560
    },
    {
      "epoch": 0.76,
      "grad_norm": 7.265150547027588,
      "learning_rate": 1.8542471042471043e-05,
      "loss": 0.7895,
      "step": 3561
    },
    {
      "epoch": 0.76,
      "grad_norm": 11.891153335571289,
      "learning_rate": 1.853925353925354e-05,
      "loss": 1.5842,
      "step": 3562
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.01278018951416,
      "learning_rate": 1.8536036036036036e-05,
      "loss": 0.8264,
      "step": 3563
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.29102611541748,
      "learning_rate": 1.8532818532818533e-05,
      "loss": 0.728,
      "step": 3564
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.714838027954102,
      "learning_rate": 1.852960102960103e-05,
      "loss": 1.1154,
      "step": 3565
    },
    {
      "epoch": 0.76,
      "grad_norm": 11.235238075256348,
      "learning_rate": 1.8526383526383525e-05,
      "loss": 0.9588,
      "step": 3566
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.01581859588623,
      "learning_rate": 1.8523166023166025e-05,
      "loss": 0.9554,
      "step": 3567
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.234260559082031,
      "learning_rate": 1.8519948519948522e-05,
      "loss": 0.9407,
      "step": 3568
    },
    {
      "epoch": 0.77,
      "grad_norm": 13.20122241973877,
      "learning_rate": 1.8516731016731018e-05,
      "loss": 1.0139,
      "step": 3569
    },
    {
      "epoch": 0.77,
      "grad_norm": 11.28197193145752,
      "learning_rate": 1.8513513513513515e-05,
      "loss": 1.0885,
      "step": 3570
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.078035354614258,
      "learning_rate": 1.851029601029601e-05,
      "loss": 0.6966,
      "step": 3571
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.215325355529785,
      "learning_rate": 1.8507078507078508e-05,
      "loss": 0.713,
      "step": 3572
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.909844398498535,
      "learning_rate": 1.8503861003861004e-05,
      "loss": 0.8086,
      "step": 3573
    },
    {
      "epoch": 0.77,
      "grad_norm": 12.158482551574707,
      "learning_rate": 1.8500643500643504e-05,
      "loss": 1.3198,
      "step": 3574
    },
    {
      "epoch": 0.77,
      "grad_norm": 11.542522430419922,
      "learning_rate": 1.8497425997425997e-05,
      "loss": 0.9209,
      "step": 3575
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.853772163391113,
      "learning_rate": 1.8494208494208493e-05,
      "loss": 0.792,
      "step": 3576
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.262121200561523,
      "learning_rate": 1.849099099099099e-05,
      "loss": 0.7751,
      "step": 3577
    },
    {
      "epoch": 0.77,
      "grad_norm": 11.380941390991211,
      "learning_rate": 1.8487773487773486e-05,
      "loss": 0.9467,
      "step": 3578
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.832927703857422,
      "learning_rate": 1.8484555984555983e-05,
      "loss": 0.8507,
      "step": 3579
    },
    {
      "epoch": 0.77,
      "grad_norm": 11.393523216247559,
      "learning_rate": 1.8481338481338483e-05,
      "loss": 1.1234,
      "step": 3580
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.244715690612793,
      "learning_rate": 1.847812097812098e-05,
      "loss": 1.1747,
      "step": 3581
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.952823638916016,
      "learning_rate": 1.8474903474903475e-05,
      "loss": 0.9463,
      "step": 3582
    },
    {
      "epoch": 0.77,
      "grad_norm": 11.372620582580566,
      "learning_rate": 1.8471685971685972e-05,
      "loss": 0.8413,
      "step": 3583
    },
    {
      "epoch": 0.77,
      "grad_norm": 11.853337287902832,
      "learning_rate": 1.846846846846847e-05,
      "loss": 0.9465,
      "step": 3584
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.388991355895996,
      "learning_rate": 1.8465250965250965e-05,
      "loss": 0.7969,
      "step": 3585
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.567821502685547,
      "learning_rate": 1.8462033462033465e-05,
      "loss": 0.931,
      "step": 3586
    },
    {
      "epoch": 0.77,
      "grad_norm": 12.104631423950195,
      "learning_rate": 1.845881595881596e-05,
      "loss": 1.0012,
      "step": 3587
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.678364753723145,
      "learning_rate": 1.8455598455598458e-05,
      "loss": 1.0585,
      "step": 3588
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.922004699707031,
      "learning_rate": 1.8452380952380954e-05,
      "loss": 0.8466,
      "step": 3589
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.107527732849121,
      "learning_rate": 1.844916344916345e-05,
      "loss": 0.6817,
      "step": 3590
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.91662311553955,
      "learning_rate": 1.8445945945945947e-05,
      "loss": 0.7467,
      "step": 3591
    },
    {
      "epoch": 0.77,
      "grad_norm": 7.597995758056641,
      "learning_rate": 1.8442728442728443e-05,
      "loss": 0.6422,
      "step": 3592
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.514692306518555,
      "learning_rate": 1.843951093951094e-05,
      "loss": 0.8738,
      "step": 3593
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.080394744873047,
      "learning_rate": 1.8436293436293436e-05,
      "loss": 0.8253,
      "step": 3594
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.468294143676758,
      "learning_rate": 1.8433075933075933e-05,
      "loss": 0.8924,
      "step": 3595
    },
    {
      "epoch": 0.77,
      "grad_norm": 11.425183296203613,
      "learning_rate": 1.842985842985843e-05,
      "loss": 0.9279,
      "step": 3596
    },
    {
      "epoch": 0.77,
      "grad_norm": 11.200297355651855,
      "learning_rate": 1.8426640926640926e-05,
      "loss": 0.8712,
      "step": 3597
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.795265197753906,
      "learning_rate": 1.8423423423423422e-05,
      "loss": 0.7907,
      "step": 3598
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.169936180114746,
      "learning_rate": 1.8420205920205922e-05,
      "loss": 0.881,
      "step": 3599
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.465703010559082,
      "learning_rate": 1.841698841698842e-05,
      "loss": 1.2016,
      "step": 3600
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.605437278747559,
      "learning_rate": 1.8413770913770915e-05,
      "loss": 0.9409,
      "step": 3601
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.407742500305176,
      "learning_rate": 1.841055341055341e-05,
      "loss": 0.9301,
      "step": 3602
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.934569358825684,
      "learning_rate": 1.8407335907335908e-05,
      "loss": 1.3001,
      "step": 3603
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.439228057861328,
      "learning_rate": 1.8404118404118404e-05,
      "loss": 0.9664,
      "step": 3604
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.959433555603027,
      "learning_rate": 1.84009009009009e-05,
      "loss": 0.9103,
      "step": 3605
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.116979598999023,
      "learning_rate": 1.83976833976834e-05,
      "loss": 1.0146,
      "step": 3606
    },
    {
      "epoch": 0.77,
      "grad_norm": 11.848618507385254,
      "learning_rate": 1.8394465894465897e-05,
      "loss": 0.9084,
      "step": 3607
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.246710777282715,
      "learning_rate": 1.8391248391248393e-05,
      "loss": 0.7348,
      "step": 3608
    },
    {
      "epoch": 0.77,
      "grad_norm": 12.350113868713379,
      "learning_rate": 1.838803088803089e-05,
      "loss": 0.9423,
      "step": 3609
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.459317207336426,
      "learning_rate": 1.8384813384813383e-05,
      "loss": 0.7113,
      "step": 3610
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.035294532775879,
      "learning_rate": 1.838159588159588e-05,
      "loss": 0.94,
      "step": 3611
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.602874755859375,
      "learning_rate": 1.837837837837838e-05,
      "loss": 0.7377,
      "step": 3612
    },
    {
      "epoch": 0.77,
      "grad_norm": 11.890663146972656,
      "learning_rate": 1.8375160875160876e-05,
      "loss": 1.1969,
      "step": 3613
    },
    {
      "epoch": 0.78,
      "grad_norm": 7.886688232421875,
      "learning_rate": 1.8371943371943372e-05,
      "loss": 0.7283,
      "step": 3614
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.588741302490234,
      "learning_rate": 1.836872586872587e-05,
      "loss": 0.6703,
      "step": 3615
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.718283653259277,
      "learning_rate": 1.8365508365508365e-05,
      "loss": 1.0098,
      "step": 3616
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.522178649902344,
      "learning_rate": 1.836229086229086e-05,
      "loss": 0.7692,
      "step": 3617
    },
    {
      "epoch": 0.78,
      "grad_norm": 7.846689701080322,
      "learning_rate": 1.8359073359073358e-05,
      "loss": 0.649,
      "step": 3618
    },
    {
      "epoch": 0.78,
      "grad_norm": 6.664380073547363,
      "learning_rate": 1.8355855855855858e-05,
      "loss": 0.6899,
      "step": 3619
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.20476245880127,
      "learning_rate": 1.8352638352638354e-05,
      "loss": 1.0283,
      "step": 3620
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.367640495300293,
      "learning_rate": 1.834942084942085e-05,
      "loss": 1.036,
      "step": 3621
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.46591854095459,
      "learning_rate": 1.8346203346203347e-05,
      "loss": 0.9819,
      "step": 3622
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.56679916381836,
      "learning_rate": 1.8342985842985844e-05,
      "loss": 0.9642,
      "step": 3623
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.81556510925293,
      "learning_rate": 1.833976833976834e-05,
      "loss": 1.1355,
      "step": 3624
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.069551467895508,
      "learning_rate": 1.833655083655084e-05,
      "loss": 0.8275,
      "step": 3625
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.280681610107422,
      "learning_rate": 1.8333333333333336e-05,
      "loss": 1.0491,
      "step": 3626
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.702716827392578,
      "learning_rate": 1.833011583011583e-05,
      "loss": 1.021,
      "step": 3627
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.255231857299805,
      "learning_rate": 1.8326898326898326e-05,
      "loss": 0.8082,
      "step": 3628
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.419022560119629,
      "learning_rate": 1.8323680823680822e-05,
      "loss": 1.2167,
      "step": 3629
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.193939208984375,
      "learning_rate": 1.832046332046332e-05,
      "loss": 0.8267,
      "step": 3630
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.746981620788574,
      "learning_rate": 1.831724581724582e-05,
      "loss": 1.1319,
      "step": 3631
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.092768669128418,
      "learning_rate": 1.8314028314028315e-05,
      "loss": 1.1268,
      "step": 3632
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.108267784118652,
      "learning_rate": 1.831081081081081e-05,
      "loss": 1.1543,
      "step": 3633
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.390913009643555,
      "learning_rate": 1.8307593307593308e-05,
      "loss": 0.9739,
      "step": 3634
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.087530136108398,
      "learning_rate": 1.8304375804375804e-05,
      "loss": 1.0474,
      "step": 3635
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.629110336303711,
      "learning_rate": 1.83011583011583e-05,
      "loss": 0.9497,
      "step": 3636
    },
    {
      "epoch": 0.78,
      "grad_norm": 7.613865375518799,
      "learning_rate": 1.8297940797940797e-05,
      "loss": 0.7451,
      "step": 3637
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.212051391601562,
      "learning_rate": 1.8294723294723297e-05,
      "loss": 0.9136,
      "step": 3638
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.30978775024414,
      "learning_rate": 1.8291505791505794e-05,
      "loss": 0.7831,
      "step": 3639
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.911571502685547,
      "learning_rate": 1.828828828828829e-05,
      "loss": 0.8123,
      "step": 3640
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.703094482421875,
      "learning_rate": 1.8285070785070786e-05,
      "loss": 0.8734,
      "step": 3641
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.380736351013184,
      "learning_rate": 1.8281853281853283e-05,
      "loss": 1.0918,
      "step": 3642
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.582650184631348,
      "learning_rate": 1.827863577863578e-05,
      "loss": 1.1228,
      "step": 3643
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.985349655151367,
      "learning_rate": 1.827541827541828e-05,
      "loss": 0.7029,
      "step": 3644
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.501401901245117,
      "learning_rate": 1.8272200772200772e-05,
      "loss": 1.0654,
      "step": 3645
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.697307586669922,
      "learning_rate": 1.826898326898327e-05,
      "loss": 0.9482,
      "step": 3646
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.904666900634766,
      "learning_rate": 1.8265765765765765e-05,
      "loss": 0.7626,
      "step": 3647
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.937832832336426,
      "learning_rate": 1.826254826254826e-05,
      "loss": 0.835,
      "step": 3648
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.726574897766113,
      "learning_rate": 1.8259330759330758e-05,
      "loss": 0.7707,
      "step": 3649
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.393346786499023,
      "learning_rate": 1.8256113256113254e-05,
      "loss": 1.0148,
      "step": 3650
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.488585472106934,
      "learning_rate": 1.8252895752895754e-05,
      "loss": 0.7983,
      "step": 3651
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.768003463745117,
      "learning_rate": 1.824967824967825e-05,
      "loss": 0.8061,
      "step": 3652
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.112157821655273,
      "learning_rate": 1.8246460746460747e-05,
      "loss": 1.114,
      "step": 3653
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.122828483581543,
      "learning_rate": 1.8243243243243244e-05,
      "loss": 0.8359,
      "step": 3654
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.482038497924805,
      "learning_rate": 1.824002574002574e-05,
      "loss": 0.733,
      "step": 3655
    },
    {
      "epoch": 0.78,
      "grad_norm": 12.95645523071289,
      "learning_rate": 1.8236808236808237e-05,
      "loss": 1.2048,
      "step": 3656
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.380399703979492,
      "learning_rate": 1.8233590733590736e-05,
      "loss": 0.9338,
      "step": 3657
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.71340274810791,
      "learning_rate": 1.8230373230373233e-05,
      "loss": 0.7839,
      "step": 3658
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.84469223022461,
      "learning_rate": 1.822715572715573e-05,
      "loss": 1.0808,
      "step": 3659
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.416122436523438,
      "learning_rate": 1.8223938223938226e-05,
      "loss": 1.0528,
      "step": 3660
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.572912216186523,
      "learning_rate": 1.8220720720720722e-05,
      "loss": 0.7921,
      "step": 3661
    },
    {
      "epoch": 0.79,
      "grad_norm": 11.192949295043945,
      "learning_rate": 1.8217503217503215e-05,
      "loss": 0.7517,
      "step": 3662
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.371752738952637,
      "learning_rate": 1.8214285714285712e-05,
      "loss": 0.8124,
      "step": 3663
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.334141731262207,
      "learning_rate": 1.821106821106821e-05,
      "loss": 0.84,
      "step": 3664
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.830801010131836,
      "learning_rate": 1.8207850707850708e-05,
      "loss": 1.0253,
      "step": 3665
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.189152717590332,
      "learning_rate": 1.8204633204633204e-05,
      "loss": 0.745,
      "step": 3666
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.855298042297363,
      "learning_rate": 1.82014157014157e-05,
      "loss": 0.8305,
      "step": 3667
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.276090621948242,
      "learning_rate": 1.8198198198198197e-05,
      "loss": 0.7371,
      "step": 3668
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.873083114624023,
      "learning_rate": 1.8194980694980694e-05,
      "loss": 0.826,
      "step": 3669
    },
    {
      "epoch": 0.79,
      "grad_norm": 12.464422225952148,
      "learning_rate": 1.8191763191763194e-05,
      "loss": 1.129,
      "step": 3670
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.457127571105957,
      "learning_rate": 1.818854568854569e-05,
      "loss": 0.9289,
      "step": 3671
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.411513328552246,
      "learning_rate": 1.8185328185328187e-05,
      "loss": 0.7628,
      "step": 3672
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.08515453338623,
      "learning_rate": 1.8182110682110683e-05,
      "loss": 1.0228,
      "step": 3673
    },
    {
      "epoch": 0.79,
      "grad_norm": 12.293855667114258,
      "learning_rate": 1.817889317889318e-05,
      "loss": 0.8897,
      "step": 3674
    },
    {
      "epoch": 0.79,
      "grad_norm": 11.646760940551758,
      "learning_rate": 1.8175675675675676e-05,
      "loss": 0.9471,
      "step": 3675
    },
    {
      "epoch": 0.79,
      "grad_norm": 14.042768478393555,
      "learning_rate": 1.8172458172458172e-05,
      "loss": 1.3363,
      "step": 3676
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.998631477355957,
      "learning_rate": 1.8169240669240672e-05,
      "loss": 0.7774,
      "step": 3677
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.226027488708496,
      "learning_rate": 1.816602316602317e-05,
      "loss": 0.6078,
      "step": 3678
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.708740234375,
      "learning_rate": 1.8162805662805665e-05,
      "loss": 0.8121,
      "step": 3679
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.310308456420898,
      "learning_rate": 1.8159588159588158e-05,
      "loss": 0.9042,
      "step": 3680
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.297465324401855,
      "learning_rate": 1.8156370656370655e-05,
      "loss": 1.0732,
      "step": 3681
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.331672668457031,
      "learning_rate": 1.815315315315315e-05,
      "loss": 0.8388,
      "step": 3682
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.779255867004395,
      "learning_rate": 1.814993564993565e-05,
      "loss": 0.8131,
      "step": 3683
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.70742130279541,
      "learning_rate": 1.8146718146718147e-05,
      "loss": 0.8922,
      "step": 3684
    },
    {
      "epoch": 0.79,
      "grad_norm": 11.968099594116211,
      "learning_rate": 1.8143500643500644e-05,
      "loss": 1.0561,
      "step": 3685
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.587940216064453,
      "learning_rate": 1.814028314028314e-05,
      "loss": 0.8399,
      "step": 3686
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.952262878417969,
      "learning_rate": 1.8137065637065637e-05,
      "loss": 0.8769,
      "step": 3687
    },
    {
      "epoch": 0.79,
      "grad_norm": 11.184599876403809,
      "learning_rate": 1.8133848133848133e-05,
      "loss": 0.8392,
      "step": 3688
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.722926139831543,
      "learning_rate": 1.813063063063063e-05,
      "loss": 0.851,
      "step": 3689
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.505992889404297,
      "learning_rate": 1.812741312741313e-05,
      "loss": 1.1802,
      "step": 3690
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.381012916564941,
      "learning_rate": 1.8124195624195626e-05,
      "loss": 0.7886,
      "step": 3691
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.860128402709961,
      "learning_rate": 1.8120978120978122e-05,
      "loss": 0.7966,
      "step": 3692
    },
    {
      "epoch": 0.79,
      "grad_norm": 11.095867156982422,
      "learning_rate": 1.811776061776062e-05,
      "loss": 0.8878,
      "step": 3693
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.868552207946777,
      "learning_rate": 1.8114543114543115e-05,
      "loss": 0.9318,
      "step": 3694
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.779757499694824,
      "learning_rate": 1.8111325611325612e-05,
      "loss": 1.0093,
      "step": 3695
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.678497314453125,
      "learning_rate": 1.810810810810811e-05,
      "loss": 0.9083,
      "step": 3696
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.398209571838379,
      "learning_rate": 1.8104890604890605e-05,
      "loss": 0.866,
      "step": 3697
    },
    {
      "epoch": 0.79,
      "grad_norm": 11.108443260192871,
      "learning_rate": 1.81016731016731e-05,
      "loss": 1.2097,
      "step": 3698
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.460412979125977,
      "learning_rate": 1.8098455598455598e-05,
      "loss": 1.0828,
      "step": 3699
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.237556457519531,
      "learning_rate": 1.8095238095238094e-05,
      "loss": 0.6991,
      "step": 3700
    },
    {
      "epoch": 0.79,
      "grad_norm": 11.37375259399414,
      "learning_rate": 1.809202059202059e-05,
      "loss": 1.1164,
      "step": 3701
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.159126281738281,
      "learning_rate": 1.808880308880309e-05,
      "loss": 0.8664,
      "step": 3702
    },
    {
      "epoch": 0.79,
      "grad_norm": 11.902417182922363,
      "learning_rate": 1.8085585585585587e-05,
      "loss": 0.921,
      "step": 3703
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.133347511291504,
      "learning_rate": 1.8082368082368083e-05,
      "loss": 0.9643,
      "step": 3704
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.377933502197266,
      "learning_rate": 1.807915057915058e-05,
      "loss": 0.8802,
      "step": 3705
    },
    {
      "epoch": 0.79,
      "grad_norm": 11.521212577819824,
      "learning_rate": 1.8075933075933076e-05,
      "loss": 1.0532,
      "step": 3706
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.77522087097168,
      "learning_rate": 1.8072715572715573e-05,
      "loss": 0.9592,
      "step": 3707
    },
    {
      "epoch": 0.8,
      "grad_norm": 11.511610984802246,
      "learning_rate": 1.806949806949807e-05,
      "loss": 0.9419,
      "step": 3708
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.219590187072754,
      "learning_rate": 1.806628056628057e-05,
      "loss": 1.2101,
      "step": 3709
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.284479141235352,
      "learning_rate": 1.8063063063063065e-05,
      "loss": 0.9737,
      "step": 3710
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.586729049682617,
      "learning_rate": 1.8059845559845562e-05,
      "loss": 1.0248,
      "step": 3711
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.197408676147461,
      "learning_rate": 1.8056628056628058e-05,
      "loss": 0.6274,
      "step": 3712
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.920995712280273,
      "learning_rate": 1.8053410553410555e-05,
      "loss": 0.9481,
      "step": 3713
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.593631744384766,
      "learning_rate": 1.8050193050193048e-05,
      "loss": 0.8143,
      "step": 3714
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.366047859191895,
      "learning_rate": 1.8046975546975548e-05,
      "loss": 0.6411,
      "step": 3715
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.39852237701416,
      "learning_rate": 1.8043758043758044e-05,
      "loss": 0.7477,
      "step": 3716
    },
    {
      "epoch": 0.8,
      "grad_norm": 11.796876907348633,
      "learning_rate": 1.804054054054054e-05,
      "loss": 1.0034,
      "step": 3717
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.815340042114258,
      "learning_rate": 1.8037323037323037e-05,
      "loss": 0.7156,
      "step": 3718
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.847369194030762,
      "learning_rate": 1.8034105534105533e-05,
      "loss": 1.215,
      "step": 3719
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.232519149780273,
      "learning_rate": 1.803088803088803e-05,
      "loss": 0.6788,
      "step": 3720
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.014573097229004,
      "learning_rate": 1.8027670527670526e-05,
      "loss": 0.7937,
      "step": 3721
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.523371696472168,
      "learning_rate": 1.8024453024453026e-05,
      "loss": 1.0455,
      "step": 3722
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.936100959777832,
      "learning_rate": 1.8021235521235523e-05,
      "loss": 0.9381,
      "step": 3723
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.016639709472656,
      "learning_rate": 1.801801801801802e-05,
      "loss": 1.0007,
      "step": 3724
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.060392379760742,
      "learning_rate": 1.8014800514800515e-05,
      "loss": 0.6934,
      "step": 3725
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.288843154907227,
      "learning_rate": 1.8011583011583012e-05,
      "loss": 0.5875,
      "step": 3726
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.785593032836914,
      "learning_rate": 1.800836550836551e-05,
      "loss": 0.8595,
      "step": 3727
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.38245677947998,
      "learning_rate": 1.8005148005148008e-05,
      "loss": 1.0277,
      "step": 3728
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.265005111694336,
      "learning_rate": 1.8001930501930505e-05,
      "loss": 0.9476,
      "step": 3729
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.374960899353027,
      "learning_rate": 1.7998712998713e-05,
      "loss": 0.999,
      "step": 3730
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.820871353149414,
      "learning_rate": 1.7995495495495498e-05,
      "loss": 0.7648,
      "step": 3731
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.754864692687988,
      "learning_rate": 1.799227799227799e-05,
      "loss": 1.0516,
      "step": 3732
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.683365821838379,
      "learning_rate": 1.7989060489060487e-05,
      "loss": 0.765,
      "step": 3733
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.644550323486328,
      "learning_rate": 1.7985842985842983e-05,
      "loss": 0.9494,
      "step": 3734
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.302775382995605,
      "learning_rate": 1.7982625482625483e-05,
      "loss": 0.8166,
      "step": 3735
    },
    {
      "epoch": 0.8,
      "grad_norm": 11.135762214660645,
      "learning_rate": 1.797940797940798e-05,
      "loss": 1.0334,
      "step": 3736
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.88996410369873,
      "learning_rate": 1.7976190476190476e-05,
      "loss": 0.6234,
      "step": 3737
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.252067565917969,
      "learning_rate": 1.7972972972972973e-05,
      "loss": 0.8052,
      "step": 3738
    },
    {
      "epoch": 0.8,
      "grad_norm": 11.057599067687988,
      "learning_rate": 1.796975546975547e-05,
      "loss": 1.0066,
      "step": 3739
    },
    {
      "epoch": 0.8,
      "grad_norm": 12.161905288696289,
      "learning_rate": 1.7966537966537966e-05,
      "loss": 0.8515,
      "step": 3740
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.285001754760742,
      "learning_rate": 1.7963320463320465e-05,
      "loss": 0.902,
      "step": 3741
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.589702606201172,
      "learning_rate": 1.7960102960102962e-05,
      "loss": 0.9788,
      "step": 3742
    },
    {
      "epoch": 0.8,
      "grad_norm": 11.224845886230469,
      "learning_rate": 1.795688545688546e-05,
      "loss": 0.9766,
      "step": 3743
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.677592277526855,
      "learning_rate": 1.7953667953667955e-05,
      "loss": 0.7503,
      "step": 3744
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.93350601196289,
      "learning_rate": 1.795045045045045e-05,
      "loss": 0.9026,
      "step": 3745
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.298040390014648,
      "learning_rate": 1.7947232947232948e-05,
      "loss": 0.7965,
      "step": 3746
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.743263244628906,
      "learning_rate": 1.7944015444015444e-05,
      "loss": 0.5889,
      "step": 3747
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.303955078125,
      "learning_rate": 1.7940797940797944e-05,
      "loss": 1.0801,
      "step": 3748
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.418158531188965,
      "learning_rate": 1.7937580437580437e-05,
      "loss": 0.6059,
      "step": 3749
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.601689338684082,
      "learning_rate": 1.7934362934362933e-05,
      "loss": 0.8183,
      "step": 3750
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.187912940979004,
      "learning_rate": 1.793114543114543e-05,
      "loss": 0.8652,
      "step": 3751
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.115297317504883,
      "learning_rate": 1.7927927927927926e-05,
      "loss": 0.5512,
      "step": 3752
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.772300720214844,
      "learning_rate": 1.7924710424710423e-05,
      "loss": 0.7107,
      "step": 3753
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.109601020812988,
      "learning_rate": 1.7921492921492923e-05,
      "loss": 0.6611,
      "step": 3754
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.571114540100098,
      "learning_rate": 1.791827541827542e-05,
      "loss": 0.873,
      "step": 3755
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.467879295349121,
      "learning_rate": 1.7915057915057916e-05,
      "loss": 0.9015,
      "step": 3756
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.754532814025879,
      "learning_rate": 1.7911840411840412e-05,
      "loss": 1.0021,
      "step": 3757
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.230743408203125,
      "learning_rate": 1.790862290862291e-05,
      "loss": 1.1343,
      "step": 3758
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.518461227416992,
      "learning_rate": 1.7905405405405405e-05,
      "loss": 0.8334,
      "step": 3759
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.204854965209961,
      "learning_rate": 1.7902187902187905e-05,
      "loss": 0.772,
      "step": 3760
    },
    {
      "epoch": 0.81,
      "grad_norm": 14.374588966369629,
      "learning_rate": 1.78989703989704e-05,
      "loss": 1.1601,
      "step": 3761
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.69198226928711,
      "learning_rate": 1.7895752895752898e-05,
      "loss": 0.7762,
      "step": 3762
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.452862739562988,
      "learning_rate": 1.7892535392535394e-05,
      "loss": 0.8395,
      "step": 3763
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.166424751281738,
      "learning_rate": 1.788931788931789e-05,
      "loss": 0.7587,
      "step": 3764
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.698578834533691,
      "learning_rate": 1.7886100386100387e-05,
      "loss": 0.6489,
      "step": 3765
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.153011322021484,
      "learning_rate": 1.7882882882882883e-05,
      "loss": 0.897,
      "step": 3766
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.648683547973633,
      "learning_rate": 1.787966537966538e-05,
      "loss": 0.7822,
      "step": 3767
    },
    {
      "epoch": 0.81,
      "grad_norm": 7.314774990081787,
      "learning_rate": 1.7876447876447876e-05,
      "loss": 0.764,
      "step": 3768
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.948917388916016,
      "learning_rate": 1.7873230373230373e-05,
      "loss": 0.6782,
      "step": 3769
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.477534294128418,
      "learning_rate": 1.787001287001287e-05,
      "loss": 1.2607,
      "step": 3770
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.90532112121582,
      "learning_rate": 1.7866795366795366e-05,
      "loss": 1.1856,
      "step": 3771
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.646735191345215,
      "learning_rate": 1.7863577863577862e-05,
      "loss": 0.8826,
      "step": 3772
    },
    {
      "epoch": 0.81,
      "grad_norm": 7.0440778732299805,
      "learning_rate": 1.7860360360360362e-05,
      "loss": 0.7196,
      "step": 3773
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.396625518798828,
      "learning_rate": 1.785714285714286e-05,
      "loss": 1.0333,
      "step": 3774
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.162458419799805,
      "learning_rate": 1.7853925353925355e-05,
      "loss": 0.7423,
      "step": 3775
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.506606101989746,
      "learning_rate": 1.785070785070785e-05,
      "loss": 0.978,
      "step": 3776
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.869784355163574,
      "learning_rate": 1.7847490347490348e-05,
      "loss": 0.8215,
      "step": 3777
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.653630256652832,
      "learning_rate": 1.7844272844272844e-05,
      "loss": 0.81,
      "step": 3778
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.166808128356934,
      "learning_rate": 1.784105534105534e-05,
      "loss": 1.1868,
      "step": 3779
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.032111167907715,
      "learning_rate": 1.783783783783784e-05,
      "loss": 0.9676,
      "step": 3780
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.45064640045166,
      "learning_rate": 1.7834620334620337e-05,
      "loss": 1.2097,
      "step": 3781
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.48880386352539,
      "learning_rate": 1.7831402831402833e-05,
      "loss": 0.8129,
      "step": 3782
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.140438079833984,
      "learning_rate": 1.782818532818533e-05,
      "loss": 1.2972,
      "step": 3783
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.248716354370117,
      "learning_rate": 1.7824967824967823e-05,
      "loss": 0.732,
      "step": 3784
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.033077239990234,
      "learning_rate": 1.782175032175032e-05,
      "loss": 1.0804,
      "step": 3785
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.798832893371582,
      "learning_rate": 1.781853281853282e-05,
      "loss": 0.762,
      "step": 3786
    },
    {
      "epoch": 0.81,
      "grad_norm": 7.752621650695801,
      "learning_rate": 1.7815315315315316e-05,
      "loss": 0.6826,
      "step": 3787
    },
    {
      "epoch": 0.81,
      "grad_norm": 7.641056537628174,
      "learning_rate": 1.7812097812097812e-05,
      "loss": 0.6361,
      "step": 3788
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.495126724243164,
      "learning_rate": 1.780888030888031e-05,
      "loss": 0.9461,
      "step": 3789
    },
    {
      "epoch": 0.81,
      "grad_norm": 12.95165729522705,
      "learning_rate": 1.7805662805662805e-05,
      "loss": 1.035,
      "step": 3790
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.600968360900879,
      "learning_rate": 1.78024453024453e-05,
      "loss": 0.7774,
      "step": 3791
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.998416900634766,
      "learning_rate": 1.7799227799227798e-05,
      "loss": 0.8417,
      "step": 3792
    },
    {
      "epoch": 0.81,
      "grad_norm": 7.5494890213012695,
      "learning_rate": 1.7796010296010298e-05,
      "loss": 0.9027,
      "step": 3793
    },
    {
      "epoch": 0.81,
      "grad_norm": 8.763962745666504,
      "learning_rate": 1.7792792792792794e-05,
      "loss": 0.9524,
      "step": 3794
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.736230850219727,
      "learning_rate": 1.778957528957529e-05,
      "loss": 0.9715,
      "step": 3795
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.45862865447998,
      "learning_rate": 1.7786357786357787e-05,
      "loss": 0.9646,
      "step": 3796
    },
    {
      "epoch": 0.81,
      "grad_norm": 7.789699554443359,
      "learning_rate": 1.7783140283140284e-05,
      "loss": 0.8106,
      "step": 3797
    },
    {
      "epoch": 0.81,
      "grad_norm": 12.4290771484375,
      "learning_rate": 1.777992277992278e-05,
      "loss": 1.0099,
      "step": 3798
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.710225105285645,
      "learning_rate": 1.777670527670528e-05,
      "loss": 1.0046,
      "step": 3799
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.647472381591797,
      "learning_rate": 1.7773487773487776e-05,
      "loss": 0.9115,
      "step": 3800
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.76447582244873,
      "learning_rate": 1.777027027027027e-05,
      "loss": 0.8978,
      "step": 3801
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.64074420928955,
      "learning_rate": 1.7767052767052766e-05,
      "loss": 1.0249,
      "step": 3802
    },
    {
      "epoch": 0.82,
      "grad_norm": 12.131990432739258,
      "learning_rate": 1.7763835263835262e-05,
      "loss": 1.3146,
      "step": 3803
    },
    {
      "epoch": 0.82,
      "grad_norm": 7.484724521636963,
      "learning_rate": 1.776061776061776e-05,
      "loss": 0.5802,
      "step": 3804
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.928018569946289,
      "learning_rate": 1.775740025740026e-05,
      "loss": 1.0223,
      "step": 3805
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.18234920501709,
      "learning_rate": 1.7754182754182755e-05,
      "loss": 1.1248,
      "step": 3806
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.471587181091309,
      "learning_rate": 1.775096525096525e-05,
      "loss": 0.7742,
      "step": 3807
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.553948402404785,
      "learning_rate": 1.7747747747747748e-05,
      "loss": 1.02,
      "step": 3808
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.26923656463623,
      "learning_rate": 1.7744530244530244e-05,
      "loss": 1.1696,
      "step": 3809
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.515508651733398,
      "learning_rate": 1.774131274131274e-05,
      "loss": 0.8433,
      "step": 3810
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.616571426391602,
      "learning_rate": 1.7738095238095237e-05,
      "loss": 0.898,
      "step": 3811
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.042841911315918,
      "learning_rate": 1.7734877734877737e-05,
      "loss": 0.8876,
      "step": 3812
    },
    {
      "epoch": 0.82,
      "grad_norm": 7.927381992340088,
      "learning_rate": 1.7731660231660234e-05,
      "loss": 0.7308,
      "step": 3813
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.251923561096191,
      "learning_rate": 1.772844272844273e-05,
      "loss": 0.7826,
      "step": 3814
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.929142951965332,
      "learning_rate": 1.7725225225225227e-05,
      "loss": 0.7268,
      "step": 3815
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.984843254089355,
      "learning_rate": 1.7722007722007723e-05,
      "loss": 0.9229,
      "step": 3816
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.913546562194824,
      "learning_rate": 1.771879021879022e-05,
      "loss": 0.8569,
      "step": 3817
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.454779624938965,
      "learning_rate": 1.771557271557272e-05,
      "loss": 0.953,
      "step": 3818
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.892192840576172,
      "learning_rate": 1.7712355212355212e-05,
      "loss": 0.8875,
      "step": 3819
    },
    {
      "epoch": 0.82,
      "grad_norm": 14.2193603515625,
      "learning_rate": 1.770913770913771e-05,
      "loss": 1.1376,
      "step": 3820
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.751132011413574,
      "learning_rate": 1.7705920205920205e-05,
      "loss": 0.7971,
      "step": 3821
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.797389030456543,
      "learning_rate": 1.7702702702702702e-05,
      "loss": 1.0694,
      "step": 3822
    },
    {
      "epoch": 0.82,
      "grad_norm": 7.938566207885742,
      "learning_rate": 1.7699485199485198e-05,
      "loss": 0.5456,
      "step": 3823
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.469725608825684,
      "learning_rate": 1.7696267696267695e-05,
      "loss": 1.0339,
      "step": 3824
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.991105079650879,
      "learning_rate": 1.7693050193050194e-05,
      "loss": 1.0938,
      "step": 3825
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.540977478027344,
      "learning_rate": 1.768983268983269e-05,
      "loss": 0.8136,
      "step": 3826
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.9214506149292,
      "learning_rate": 1.7686615186615187e-05,
      "loss": 0.8325,
      "step": 3827
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.946560859680176,
      "learning_rate": 1.7683397683397684e-05,
      "loss": 1.2027,
      "step": 3828
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.501944541931152,
      "learning_rate": 1.768018018018018e-05,
      "loss": 0.8993,
      "step": 3829
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.83857250213623,
      "learning_rate": 1.7676962676962677e-05,
      "loss": 0.929,
      "step": 3830
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.647041320800781,
      "learning_rate": 1.7673745173745177e-05,
      "loss": 0.8758,
      "step": 3831
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.72890853881836,
      "learning_rate": 1.7670527670527673e-05,
      "loss": 0.9611,
      "step": 3832
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.891536712646484,
      "learning_rate": 1.766731016731017e-05,
      "loss": 0.8444,
      "step": 3833
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.799530982971191,
      "learning_rate": 1.7664092664092666e-05,
      "loss": 0.6873,
      "step": 3834
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.784976959228516,
      "learning_rate": 1.7660875160875162e-05,
      "loss": 0.9725,
      "step": 3835
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.14620304107666,
      "learning_rate": 1.7657657657657655e-05,
      "loss": 0.7788,
      "step": 3836
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.845622062683105,
      "learning_rate": 1.7654440154440152e-05,
      "loss": 0.934,
      "step": 3837
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.494268417358398,
      "learning_rate": 1.7651222651222652e-05,
      "loss": 0.9592,
      "step": 3838
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.384468078613281,
      "learning_rate": 1.7648005148005148e-05,
      "loss": 0.8393,
      "step": 3839
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.535968780517578,
      "learning_rate": 1.7644787644787645e-05,
      "loss": 1.0615,
      "step": 3840
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.17859935760498,
      "learning_rate": 1.764157014157014e-05,
      "loss": 0.7798,
      "step": 3841
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.145503044128418,
      "learning_rate": 1.7638352638352638e-05,
      "loss": 0.9213,
      "step": 3842
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.333669662475586,
      "learning_rate": 1.7635135135135134e-05,
      "loss": 0.6527,
      "step": 3843
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.207146644592285,
      "learning_rate": 1.7631917631917634e-05,
      "loss": 1.2377,
      "step": 3844
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.079751014709473,
      "learning_rate": 1.762870012870013e-05,
      "loss": 0.7674,
      "step": 3845
    },
    {
      "epoch": 0.82,
      "grad_norm": 12.670841217041016,
      "learning_rate": 1.7625482625482627e-05,
      "loss": 1.0903,
      "step": 3846
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.234728813171387,
      "learning_rate": 1.7622265122265123e-05,
      "loss": 0.854,
      "step": 3847
    },
    {
      "epoch": 0.83,
      "grad_norm": 11.173053741455078,
      "learning_rate": 1.761904761904762e-05,
      "loss": 1.1142,
      "step": 3848
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.292215347290039,
      "learning_rate": 1.7615830115830116e-05,
      "loss": 0.8603,
      "step": 3849
    },
    {
      "epoch": 0.83,
      "grad_norm": 11.364116668701172,
      "learning_rate": 1.7612612612612613e-05,
      "loss": 0.8402,
      "step": 3850
    },
    {
      "epoch": 0.83,
      "grad_norm": 12.04609203338623,
      "learning_rate": 1.7609395109395112e-05,
      "loss": 0.5894,
      "step": 3851
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.760111808776855,
      "learning_rate": 1.760617760617761e-05,
      "loss": 1.0589,
      "step": 3852
    },
    {
      "epoch": 0.83,
      "grad_norm": 11.149422645568848,
      "learning_rate": 1.7602960102960105e-05,
      "loss": 0.8073,
      "step": 3853
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.258275032043457,
      "learning_rate": 1.75997425997426e-05,
      "loss": 1.0767,
      "step": 3854
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.100471496582031,
      "learning_rate": 1.7596525096525095e-05,
      "loss": 0.9706,
      "step": 3855
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.506904602050781,
      "learning_rate": 1.759330759330759e-05,
      "loss": 0.6953,
      "step": 3856
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.042004585266113,
      "learning_rate": 1.759009009009009e-05,
      "loss": 0.6675,
      "step": 3857
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.493775367736816,
      "learning_rate": 1.7586872586872588e-05,
      "loss": 0.7744,
      "step": 3858
    },
    {
      "epoch": 0.83,
      "grad_norm": 11.549118995666504,
      "learning_rate": 1.7583655083655084e-05,
      "loss": 1.1247,
      "step": 3859
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.403538703918457,
      "learning_rate": 1.758043758043758e-05,
      "loss": 0.63,
      "step": 3860
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.88668441772461,
      "learning_rate": 1.7577220077220077e-05,
      "loss": 0.9711,
      "step": 3861
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.7904052734375,
      "learning_rate": 1.7574002574002573e-05,
      "loss": 0.7494,
      "step": 3862
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.066925048828125,
      "learning_rate": 1.7570785070785073e-05,
      "loss": 0.6073,
      "step": 3863
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.783288955688477,
      "learning_rate": 1.756756756756757e-05,
      "loss": 0.7874,
      "step": 3864
    },
    {
      "epoch": 0.83,
      "grad_norm": 11.358901977539062,
      "learning_rate": 1.7564350064350066e-05,
      "loss": 1.086,
      "step": 3865
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.072903633117676,
      "learning_rate": 1.7561132561132563e-05,
      "loss": 0.9022,
      "step": 3866
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.44265365600586,
      "learning_rate": 1.755791505791506e-05,
      "loss": 0.8191,
      "step": 3867
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.708025932312012,
      "learning_rate": 1.7554697554697555e-05,
      "loss": 1.0478,
      "step": 3868
    },
    {
      "epoch": 0.83,
      "grad_norm": 12.284366607666016,
      "learning_rate": 1.7551480051480052e-05,
      "loss": 0.6838,
      "step": 3869
    },
    {
      "epoch": 0.83,
      "grad_norm": 11.034693717956543,
      "learning_rate": 1.7548262548262552e-05,
      "loss": 1.1323,
      "step": 3870
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.15671157836914,
      "learning_rate": 1.7545045045045045e-05,
      "loss": 0.8004,
      "step": 3871
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.778507232666016,
      "learning_rate": 1.754182754182754e-05,
      "loss": 1.1096,
      "step": 3872
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.061559677124023,
      "learning_rate": 1.7538610038610038e-05,
      "loss": 0.7594,
      "step": 3873
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.546187400817871,
      "learning_rate": 1.7535392535392534e-05,
      "loss": 0.7553,
      "step": 3874
    },
    {
      "epoch": 0.83,
      "grad_norm": 11.964271545410156,
      "learning_rate": 1.753217503217503e-05,
      "loss": 0.972,
      "step": 3875
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.85178279876709,
      "learning_rate": 1.752895752895753e-05,
      "loss": 0.9791,
      "step": 3876
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.543663024902344,
      "learning_rate": 1.7525740025740027e-05,
      "loss": 0.8928,
      "step": 3877
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.174692153930664,
      "learning_rate": 1.7522522522522523e-05,
      "loss": 0.8038,
      "step": 3878
    },
    {
      "epoch": 0.83,
      "grad_norm": 7.380842208862305,
      "learning_rate": 1.751930501930502e-05,
      "loss": 0.8892,
      "step": 3879
    },
    {
      "epoch": 0.83,
      "grad_norm": 12.281495094299316,
      "learning_rate": 1.7516087516087516e-05,
      "loss": 0.9744,
      "step": 3880
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.06323528289795,
      "learning_rate": 1.7512870012870013e-05,
      "loss": 0.8156,
      "step": 3881
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.909087181091309,
      "learning_rate": 1.750965250965251e-05,
      "loss": 0.8702,
      "step": 3882
    },
    {
      "epoch": 0.83,
      "grad_norm": 12.204541206359863,
      "learning_rate": 1.750643500643501e-05,
      "loss": 0.9814,
      "step": 3883
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.264025688171387,
      "learning_rate": 1.7503217503217505e-05,
      "loss": 1.0511,
      "step": 3884
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.972024917602539,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 0.9938,
      "step": 3885
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.941984176635742,
      "learning_rate": 1.74967824967825e-05,
      "loss": 0.9523,
      "step": 3886
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.554635047912598,
      "learning_rate": 1.7493564993564995e-05,
      "loss": 0.6801,
      "step": 3887
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.89773941040039,
      "learning_rate": 1.7490347490347488e-05,
      "loss": 1.1946,
      "step": 3888
    },
    {
      "epoch": 0.83,
      "grad_norm": 7.225334644317627,
      "learning_rate": 1.7487129987129988e-05,
      "loss": 0.6807,
      "step": 3889
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.197216987609863,
      "learning_rate": 1.7483912483912484e-05,
      "loss": 0.8143,
      "step": 3890
    },
    {
      "epoch": 0.83,
      "grad_norm": 11.412395477294922,
      "learning_rate": 1.748069498069498e-05,
      "loss": 0.9504,
      "step": 3891
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.091601371765137,
      "learning_rate": 1.7477477477477477e-05,
      "loss": 0.6572,
      "step": 3892
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.219147682189941,
      "learning_rate": 1.7474259974259973e-05,
      "loss": 1.0004,
      "step": 3893
    },
    {
      "epoch": 0.84,
      "grad_norm": 11.47287654876709,
      "learning_rate": 1.747104247104247e-05,
      "loss": 1.164,
      "step": 3894
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.720614433288574,
      "learning_rate": 1.7467824967824966e-05,
      "loss": 0.8748,
      "step": 3895
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.592947959899902,
      "learning_rate": 1.7464607464607466e-05,
      "loss": 0.8234,
      "step": 3896
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.596833229064941,
      "learning_rate": 1.7461389961389963e-05,
      "loss": 0.8132,
      "step": 3897
    },
    {
      "epoch": 0.84,
      "grad_norm": 11.202554702758789,
      "learning_rate": 1.745817245817246e-05,
      "loss": 1.0071,
      "step": 3898
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.201379776000977,
      "learning_rate": 1.7454954954954956e-05,
      "loss": 1.0257,
      "step": 3899
    },
    {
      "epoch": 0.84,
      "grad_norm": 12.101923942565918,
      "learning_rate": 1.7451737451737452e-05,
      "loss": 0.7012,
      "step": 3900
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.175971984863281,
      "learning_rate": 1.744851994851995e-05,
      "loss": 0.7135,
      "step": 3901
    },
    {
      "epoch": 0.84,
      "grad_norm": 11.4605131149292,
      "learning_rate": 1.7445302445302448e-05,
      "loss": 1.0614,
      "step": 3902
    },
    {
      "epoch": 0.84,
      "grad_norm": 7.998053550720215,
      "learning_rate": 1.7442084942084945e-05,
      "loss": 0.7885,
      "step": 3903
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.228028297424316,
      "learning_rate": 1.743886743886744e-05,
      "loss": 0.9708,
      "step": 3904
    },
    {
      "epoch": 0.84,
      "grad_norm": 11.56467056274414,
      "learning_rate": 1.7435649935649938e-05,
      "loss": 1.0991,
      "step": 3905
    },
    {
      "epoch": 0.84,
      "grad_norm": 13.56535530090332,
      "learning_rate": 1.743243243243243e-05,
      "loss": 1.0305,
      "step": 3906
    },
    {
      "epoch": 0.84,
      "grad_norm": 12.211030006408691,
      "learning_rate": 1.7429214929214927e-05,
      "loss": 0.9938,
      "step": 3907
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.309004783630371,
      "learning_rate": 1.7425997425997424e-05,
      "loss": 0.8026,
      "step": 3908
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.214923858642578,
      "learning_rate": 1.7422779922779923e-05,
      "loss": 0.6776,
      "step": 3909
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.346762657165527,
      "learning_rate": 1.741956241956242e-05,
      "loss": 0.919,
      "step": 3910
    },
    {
      "epoch": 0.84,
      "grad_norm": 11.366528511047363,
      "learning_rate": 1.7416344916344916e-05,
      "loss": 1.1553,
      "step": 3911
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.053949356079102,
      "learning_rate": 1.7413127413127413e-05,
      "loss": 0.729,
      "step": 3912
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.575491905212402,
      "learning_rate": 1.740990990990991e-05,
      "loss": 0.7959,
      "step": 3913
    },
    {
      "epoch": 0.84,
      "grad_norm": 11.88708782196045,
      "learning_rate": 1.7406692406692406e-05,
      "loss": 0.8807,
      "step": 3914
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.85537052154541,
      "learning_rate": 1.7403474903474906e-05,
      "loss": 0.8199,
      "step": 3915
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.249074935913086,
      "learning_rate": 1.7400257400257402e-05,
      "loss": 0.7878,
      "step": 3916
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.604769706726074,
      "learning_rate": 1.73970398970399e-05,
      "loss": 1.0935,
      "step": 3917
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.61694622039795,
      "learning_rate": 1.7393822393822395e-05,
      "loss": 0.9533,
      "step": 3918
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.331284523010254,
      "learning_rate": 1.739060489060489e-05,
      "loss": 0.7705,
      "step": 3919
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.362150192260742,
      "learning_rate": 1.7387387387387388e-05,
      "loss": 0.8323,
      "step": 3920
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.639735221862793,
      "learning_rate": 1.7384169884169884e-05,
      "loss": 1.1779,
      "step": 3921
    },
    {
      "epoch": 0.84,
      "grad_norm": 11.126358032226562,
      "learning_rate": 1.7380952380952384e-05,
      "loss": 0.9087,
      "step": 3922
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.46946907043457,
      "learning_rate": 1.7377734877734877e-05,
      "loss": 0.8726,
      "step": 3923
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.2860746383667,
      "learning_rate": 1.7374517374517374e-05,
      "loss": 0.8414,
      "step": 3924
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.514983177185059,
      "learning_rate": 1.737129987129987e-05,
      "loss": 0.7638,
      "step": 3925
    },
    {
      "epoch": 0.84,
      "grad_norm": 12.714587211608887,
      "learning_rate": 1.7368082368082367e-05,
      "loss": 0.8809,
      "step": 3926
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.160524368286133,
      "learning_rate": 1.7364864864864863e-05,
      "loss": 0.9924,
      "step": 3927
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.421298027038574,
      "learning_rate": 1.7361647361647363e-05,
      "loss": 0.7589,
      "step": 3928
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.687131881713867,
      "learning_rate": 1.735842985842986e-05,
      "loss": 0.8534,
      "step": 3929
    },
    {
      "epoch": 0.84,
      "grad_norm": 11.139305114746094,
      "learning_rate": 1.7355212355212356e-05,
      "loss": 0.9532,
      "step": 3930
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.943548202514648,
      "learning_rate": 1.7351994851994852e-05,
      "loss": 1.0914,
      "step": 3931
    },
    {
      "epoch": 0.84,
      "grad_norm": 11.474292755126953,
      "learning_rate": 1.734877734877735e-05,
      "loss": 0.8268,
      "step": 3932
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.892660140991211,
      "learning_rate": 1.7345559845559845e-05,
      "loss": 0.8322,
      "step": 3933
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.874340057373047,
      "learning_rate": 1.7342342342342345e-05,
      "loss": 1.0693,
      "step": 3934
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.917505264282227,
      "learning_rate": 1.733912483912484e-05,
      "loss": 0.9071,
      "step": 3935
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.656656265258789,
      "learning_rate": 1.7335907335907338e-05,
      "loss": 0.9944,
      "step": 3936
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.838008880615234,
      "learning_rate": 1.7332689832689834e-05,
      "loss": 0.7387,
      "step": 3937
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.503743171691895,
      "learning_rate": 1.732947232947233e-05,
      "loss": 0.6438,
      "step": 3938
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.723138809204102,
      "learning_rate": 1.7326254826254827e-05,
      "loss": 0.7218,
      "step": 3939
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.033740997314453,
      "learning_rate": 1.7323037323037324e-05,
      "loss": 0.7771,
      "step": 3940
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.839373588562012,
      "learning_rate": 1.731981981981982e-05,
      "loss": 0.9937,
      "step": 3941
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.311503410339355,
      "learning_rate": 1.7316602316602317e-05,
      "loss": 0.7377,
      "step": 3942
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.873595237731934,
      "learning_rate": 1.7313384813384813e-05,
      "loss": 0.8354,
      "step": 3943
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.484609603881836,
      "learning_rate": 1.731016731016731e-05,
      "loss": 0.8043,
      "step": 3944
    },
    {
      "epoch": 0.85,
      "grad_norm": 13.085393905639648,
      "learning_rate": 1.7306949806949806e-05,
      "loss": 1.0346,
      "step": 3945
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.66528034210205,
      "learning_rate": 1.7303732303732302e-05,
      "loss": 1.0759,
      "step": 3946
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.302352905273438,
      "learning_rate": 1.7300514800514802e-05,
      "loss": 0.9736,
      "step": 3947
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.581368446350098,
      "learning_rate": 1.72972972972973e-05,
      "loss": 1.0023,
      "step": 3948
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.936838150024414,
      "learning_rate": 1.7294079794079795e-05,
      "loss": 0.9168,
      "step": 3949
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.252141952514648,
      "learning_rate": 1.729086229086229e-05,
      "loss": 0.9161,
      "step": 3950
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.891196250915527,
      "learning_rate": 1.7287644787644788e-05,
      "loss": 1.0557,
      "step": 3951
    },
    {
      "epoch": 0.85,
      "grad_norm": 12.166337013244629,
      "learning_rate": 1.7284427284427284e-05,
      "loss": 1.0174,
      "step": 3952
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.6785249710083,
      "learning_rate": 1.728120978120978e-05,
      "loss": 0.7847,
      "step": 3953
    },
    {
      "epoch": 0.85,
      "grad_norm": 7.220531463623047,
      "learning_rate": 1.727799227799228e-05,
      "loss": 0.7398,
      "step": 3954
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.337693214416504,
      "learning_rate": 1.7274774774774777e-05,
      "loss": 0.7766,
      "step": 3955
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.87789249420166,
      "learning_rate": 1.7271557271557274e-05,
      "loss": 0.9842,
      "step": 3956
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.134621620178223,
      "learning_rate": 1.726833976833977e-05,
      "loss": 0.9776,
      "step": 3957
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.974031448364258,
      "learning_rate": 1.7265122265122263e-05,
      "loss": 0.7561,
      "step": 3958
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.328855514526367,
      "learning_rate": 1.726190476190476e-05,
      "loss": 0.8153,
      "step": 3959
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.509795188903809,
      "learning_rate": 1.725868725868726e-05,
      "loss": 1.1305,
      "step": 3960
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.519885063171387,
      "learning_rate": 1.7255469755469756e-05,
      "loss": 1.0311,
      "step": 3961
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.609169006347656,
      "learning_rate": 1.7252252252252252e-05,
      "loss": 0.9347,
      "step": 3962
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.174147605895996,
      "learning_rate": 1.724903474903475e-05,
      "loss": 0.8244,
      "step": 3963
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.243550300598145,
      "learning_rate": 1.7245817245817245e-05,
      "loss": 1.1452,
      "step": 3964
    },
    {
      "epoch": 0.85,
      "grad_norm": 7.236003398895264,
      "learning_rate": 1.724259974259974e-05,
      "loss": 0.7365,
      "step": 3965
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.191320419311523,
      "learning_rate": 1.7239382239382238e-05,
      "loss": 0.8066,
      "step": 3966
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.182435989379883,
      "learning_rate": 1.7236164736164738e-05,
      "loss": 0.7881,
      "step": 3967
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.554177284240723,
      "learning_rate": 1.7232947232947234e-05,
      "loss": 0.9617,
      "step": 3968
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.503286361694336,
      "learning_rate": 1.722972972972973e-05,
      "loss": 0.9195,
      "step": 3969
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.49791145324707,
      "learning_rate": 1.7226512226512227e-05,
      "loss": 0.825,
      "step": 3970
    },
    {
      "epoch": 0.85,
      "grad_norm": 12.933130264282227,
      "learning_rate": 1.7223294723294724e-05,
      "loss": 1.1504,
      "step": 3971
    },
    {
      "epoch": 0.85,
      "grad_norm": 6.482097148895264,
      "learning_rate": 1.722007722007722e-05,
      "loss": 0.4711,
      "step": 3972
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.200620651245117,
      "learning_rate": 1.721685971685972e-05,
      "loss": 0.9584,
      "step": 3973
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.308752059936523,
      "learning_rate": 1.7213642213642217e-05,
      "loss": 0.8778,
      "step": 3974
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.85908031463623,
      "learning_rate": 1.721042471042471e-05,
      "loss": 1.0322,
      "step": 3975
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.854814529418945,
      "learning_rate": 1.7207207207207206e-05,
      "loss": 0.9479,
      "step": 3976
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.111062049865723,
      "learning_rate": 1.7203989703989702e-05,
      "loss": 1.2209,
      "step": 3977
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.130377769470215,
      "learning_rate": 1.72007722007722e-05,
      "loss": 0.8884,
      "step": 3978
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.523655891418457,
      "learning_rate": 1.71975546975547e-05,
      "loss": 0.8697,
      "step": 3979
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.569526672363281,
      "learning_rate": 1.7194337194337195e-05,
      "loss": 1.0615,
      "step": 3980
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.518078804016113,
      "learning_rate": 1.719111969111969e-05,
      "loss": 0.9036,
      "step": 3981
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.57045841217041,
      "learning_rate": 1.7187902187902188e-05,
      "loss": 0.8592,
      "step": 3982
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.3268461227417,
      "learning_rate": 1.7184684684684685e-05,
      "loss": 0.6522,
      "step": 3983
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.538484573364258,
      "learning_rate": 1.718146718146718e-05,
      "loss": 0.8603,
      "step": 3984
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.044841766357422,
      "learning_rate": 1.7178249678249677e-05,
      "loss": 0.8771,
      "step": 3985
    },
    {
      "epoch": 0.85,
      "grad_norm": 10.476764678955078,
      "learning_rate": 1.7175032175032177e-05,
      "loss": 1.2497,
      "step": 3986
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.497349739074707,
      "learning_rate": 1.7171814671814674e-05,
      "loss": 0.7719,
      "step": 3987
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.978211402893066,
      "learning_rate": 1.716859716859717e-05,
      "loss": 0.9246,
      "step": 3988
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.261571884155273,
      "learning_rate": 1.7165379665379667e-05,
      "loss": 0.5931,
      "step": 3989
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.300415992736816,
      "learning_rate": 1.7162162162162163e-05,
      "loss": 0.8706,
      "step": 3990
    },
    {
      "epoch": 0.86,
      "grad_norm": 11.115483283996582,
      "learning_rate": 1.715894465894466e-05,
      "loss": 0.6802,
      "step": 3991
    },
    {
      "epoch": 0.86,
      "grad_norm": 12.282155990600586,
      "learning_rate": 1.715572715572716e-05,
      "loss": 1.1383,
      "step": 3992
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.47299861907959,
      "learning_rate": 1.7152509652509652e-05,
      "loss": 0.7724,
      "step": 3993
    },
    {
      "epoch": 0.86,
      "grad_norm": 14.006762504577637,
      "learning_rate": 1.714929214929215e-05,
      "loss": 1.0553,
      "step": 3994
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.68498706817627,
      "learning_rate": 1.7146074646074645e-05,
      "loss": 0.8779,
      "step": 3995
    },
    {
      "epoch": 0.86,
      "grad_norm": 12.755329132080078,
      "learning_rate": 1.7142857142857142e-05,
      "loss": 1.4252,
      "step": 3996
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.201739311218262,
      "learning_rate": 1.7139639639639638e-05,
      "loss": 0.8517,
      "step": 3997
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.443880081176758,
      "learning_rate": 1.7136422136422135e-05,
      "loss": 0.9969,
      "step": 3998
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.396809577941895,
      "learning_rate": 1.7133204633204635e-05,
      "loss": 0.9428,
      "step": 3999
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.84251880645752,
      "learning_rate": 1.712998712998713e-05,
      "loss": 0.9734,
      "step": 4000
    },
    {
      "epoch": 0.86,
      "grad_norm": 12.357739448547363,
      "learning_rate": 1.7126769626769627e-05,
      "loss": 0.6929,
      "step": 4001
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.3462553024292,
      "learning_rate": 1.7123552123552124e-05,
      "loss": 0.8125,
      "step": 4002
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.43845272064209,
      "learning_rate": 1.712033462033462e-05,
      "loss": 0.8433,
      "step": 4003
    },
    {
      "epoch": 0.86,
      "grad_norm": 11.68918514251709,
      "learning_rate": 1.7117117117117117e-05,
      "loss": 1.2664,
      "step": 4004
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.78393268585205,
      "learning_rate": 1.7113899613899617e-05,
      "loss": 0.7172,
      "step": 4005
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.547024726867676,
      "learning_rate": 1.7110682110682113e-05,
      "loss": 1.0355,
      "step": 4006
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.787654876708984,
      "learning_rate": 1.710746460746461e-05,
      "loss": 0.7383,
      "step": 4007
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.018963813781738,
      "learning_rate": 1.7104247104247106e-05,
      "loss": 0.6296,
      "step": 4008
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.439315795898438,
      "learning_rate": 1.7101029601029602e-05,
      "loss": 1.0218,
      "step": 4009
    },
    {
      "epoch": 0.86,
      "grad_norm": 7.184160232543945,
      "learning_rate": 1.7097812097812096e-05,
      "loss": 0.6718,
      "step": 4010
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.254461288452148,
      "learning_rate": 1.7094594594594592e-05,
      "loss": 0.7838,
      "step": 4011
    },
    {
      "epoch": 0.86,
      "grad_norm": 11.799650192260742,
      "learning_rate": 1.7091377091377092e-05,
      "loss": 0.9727,
      "step": 4012
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.478065490722656,
      "learning_rate": 1.7088159588159588e-05,
      "loss": 0.9973,
      "step": 4013
    },
    {
      "epoch": 0.86,
      "grad_norm": 11.138068199157715,
      "learning_rate": 1.7084942084942085e-05,
      "loss": 1.263,
      "step": 4014
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.72896957397461,
      "learning_rate": 1.708172458172458e-05,
      "loss": 0.9008,
      "step": 4015
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.678324699401855,
      "learning_rate": 1.7078507078507078e-05,
      "loss": 0.8571,
      "step": 4016
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.080976486206055,
      "learning_rate": 1.7075289575289574e-05,
      "loss": 0.7802,
      "step": 4017
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.15636157989502,
      "learning_rate": 1.7072072072072074e-05,
      "loss": 0.8063,
      "step": 4018
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.971996307373047,
      "learning_rate": 1.706885456885457e-05,
      "loss": 0.8543,
      "step": 4019
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.520601272583008,
      "learning_rate": 1.7065637065637067e-05,
      "loss": 0.7742,
      "step": 4020
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.569695472717285,
      "learning_rate": 1.7062419562419563e-05,
      "loss": 1.1284,
      "step": 4021
    },
    {
      "epoch": 0.86,
      "grad_norm": 11.121689796447754,
      "learning_rate": 1.705920205920206e-05,
      "loss": 0.8815,
      "step": 4022
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.496030807495117,
      "learning_rate": 1.7055984555984556e-05,
      "loss": 1.0154,
      "step": 4023
    },
    {
      "epoch": 0.86,
      "grad_norm": 13.066908836364746,
      "learning_rate": 1.7052767052767053e-05,
      "loss": 0.9671,
      "step": 4024
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.154163360595703,
      "learning_rate": 1.7049549549549552e-05,
      "loss": 0.8557,
      "step": 4025
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.947766304016113,
      "learning_rate": 1.704633204633205e-05,
      "loss": 0.7921,
      "step": 4026
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.13921070098877,
      "learning_rate": 1.7043114543114545e-05,
      "loss": 0.8466,
      "step": 4027
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.926196098327637,
      "learning_rate": 1.703989703989704e-05,
      "loss": 0.9991,
      "step": 4028
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.807832717895508,
      "learning_rate": 1.7036679536679535e-05,
      "loss": 1.0176,
      "step": 4029
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.459173202514648,
      "learning_rate": 1.703346203346203e-05,
      "loss": 0.8573,
      "step": 4030
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.362380981445312,
      "learning_rate": 1.703024453024453e-05,
      "loss": 0.7511,
      "step": 4031
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.357294082641602,
      "learning_rate": 1.7027027027027028e-05,
      "loss": 1.0345,
      "step": 4032
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.113051414489746,
      "learning_rate": 1.7023809523809524e-05,
      "loss": 0.7692,
      "step": 4033
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.502828598022461,
      "learning_rate": 1.702059202059202e-05,
      "loss": 0.9028,
      "step": 4034
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.92676830291748,
      "learning_rate": 1.7017374517374517e-05,
      "loss": 0.7088,
      "step": 4035
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.15585994720459,
      "learning_rate": 1.7014157014157013e-05,
      "loss": 0.9317,
      "step": 4036
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.023937225341797,
      "learning_rate": 1.7010939510939513e-05,
      "loss": 0.9671,
      "step": 4037
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.043231010437012,
      "learning_rate": 1.700772200772201e-05,
      "loss": 0.9933,
      "step": 4038
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.25130558013916,
      "learning_rate": 1.7004504504504506e-05,
      "loss": 0.8556,
      "step": 4039
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.119218826293945,
      "learning_rate": 1.7001287001287003e-05,
      "loss": 0.8614,
      "step": 4040
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.853473663330078,
      "learning_rate": 1.69980694980695e-05,
      "loss": 1.0608,
      "step": 4041
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.458065032958984,
      "learning_rate": 1.6994851994851996e-05,
      "loss": 0.7421,
      "step": 4042
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.598430633544922,
      "learning_rate": 1.6991634491634492e-05,
      "loss": 0.926,
      "step": 4043
    },
    {
      "epoch": 0.87,
      "grad_norm": 7.638561725616455,
      "learning_rate": 1.6988416988416992e-05,
      "loss": 0.7052,
      "step": 4044
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.776971817016602,
      "learning_rate": 1.6985199485199485e-05,
      "loss": 0.7599,
      "step": 4045
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.867331504821777,
      "learning_rate": 1.698198198198198e-05,
      "loss": 0.9828,
      "step": 4046
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.191418647766113,
      "learning_rate": 1.6978764478764478e-05,
      "loss": 0.7407,
      "step": 4047
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.529862403869629,
      "learning_rate": 1.6975546975546974e-05,
      "loss": 0.8813,
      "step": 4048
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.897934913635254,
      "learning_rate": 1.697232947232947e-05,
      "loss": 0.9073,
      "step": 4049
    },
    {
      "epoch": 0.87,
      "grad_norm": 11.08859920501709,
      "learning_rate": 1.696911196911197e-05,
      "loss": 1.2724,
      "step": 4050
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.547486305236816,
      "learning_rate": 1.6965894465894467e-05,
      "loss": 0.7307,
      "step": 4051
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.819746971130371,
      "learning_rate": 1.6962676962676963e-05,
      "loss": 0.8624,
      "step": 4052
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.011914253234863,
      "learning_rate": 1.695945945945946e-05,
      "loss": 0.9043,
      "step": 4053
    },
    {
      "epoch": 0.87,
      "grad_norm": 11.099408149719238,
      "learning_rate": 1.6956241956241956e-05,
      "loss": 0.728,
      "step": 4054
    },
    {
      "epoch": 0.87,
      "grad_norm": 7.908910274505615,
      "learning_rate": 1.6953024453024453e-05,
      "loss": 0.7851,
      "step": 4055
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.699737548828125,
      "learning_rate": 1.694980694980695e-05,
      "loss": 0.7823,
      "step": 4056
    },
    {
      "epoch": 0.87,
      "grad_norm": 7.646626949310303,
      "learning_rate": 1.694658944658945e-05,
      "loss": 0.7025,
      "step": 4057
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.160977363586426,
      "learning_rate": 1.6943371943371946e-05,
      "loss": 0.8119,
      "step": 4058
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.087372779846191,
      "learning_rate": 1.6940154440154442e-05,
      "loss": 1.1298,
      "step": 4059
    },
    {
      "epoch": 0.87,
      "grad_norm": 11.324231147766113,
      "learning_rate": 1.693693693693694e-05,
      "loss": 0.5805,
      "step": 4060
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.718546867370605,
      "learning_rate": 1.6933719433719435e-05,
      "loss": 0.8735,
      "step": 4061
    },
    {
      "epoch": 0.87,
      "grad_norm": 7.788898944854736,
      "learning_rate": 1.6930501930501928e-05,
      "loss": 0.644,
      "step": 4062
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.6111478805542,
      "learning_rate": 1.6927284427284428e-05,
      "loss": 0.8714,
      "step": 4063
    },
    {
      "epoch": 0.87,
      "grad_norm": 13.12974739074707,
      "learning_rate": 1.6924066924066924e-05,
      "loss": 0.9954,
      "step": 4064
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.456220626831055,
      "learning_rate": 1.692084942084942e-05,
      "loss": 0.9136,
      "step": 4065
    },
    {
      "epoch": 0.87,
      "grad_norm": 12.79215145111084,
      "learning_rate": 1.6917631917631917e-05,
      "loss": 0.9599,
      "step": 4066
    },
    {
      "epoch": 0.87,
      "grad_norm": 12.050010681152344,
      "learning_rate": 1.6914414414414414e-05,
      "loss": 0.9246,
      "step": 4067
    },
    {
      "epoch": 0.87,
      "grad_norm": 11.886415481567383,
      "learning_rate": 1.691119691119691e-05,
      "loss": 1.0944,
      "step": 4068
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.494606018066406,
      "learning_rate": 1.6907979407979407e-05,
      "loss": 0.9832,
      "step": 4069
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.666691780090332,
      "learning_rate": 1.6904761904761906e-05,
      "loss": 0.9299,
      "step": 4070
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.841564178466797,
      "learning_rate": 1.6901544401544403e-05,
      "loss": 0.864,
      "step": 4071
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.025947570800781,
      "learning_rate": 1.68983268983269e-05,
      "loss": 1.0001,
      "step": 4072
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.194997787475586,
      "learning_rate": 1.6895109395109396e-05,
      "loss": 0.7492,
      "step": 4073
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.159926414489746,
      "learning_rate": 1.6891891891891892e-05,
      "loss": 0.7421,
      "step": 4074
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.873473167419434,
      "learning_rate": 1.688867438867439e-05,
      "loss": 0.7716,
      "step": 4075
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.723116874694824,
      "learning_rate": 1.688545688545689e-05,
      "loss": 1.12,
      "step": 4076
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.918984413146973,
      "learning_rate": 1.6882239382239385e-05,
      "loss": 0.8639,
      "step": 4077
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.307541847229004,
      "learning_rate": 1.687902187902188e-05,
      "loss": 1.0609,
      "step": 4078
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.4723482131958,
      "learning_rate": 1.6875804375804378e-05,
      "loss": 0.8805,
      "step": 4079
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.196856498718262,
      "learning_rate": 1.687258687258687e-05,
      "loss": 1.0089,
      "step": 4080
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.546357154846191,
      "learning_rate": 1.6869369369369367e-05,
      "loss": 1.0323,
      "step": 4081
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.794266700744629,
      "learning_rate": 1.6866151866151864e-05,
      "loss": 0.8379,
      "step": 4082
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.659869194030762,
      "learning_rate": 1.6862934362934364e-05,
      "loss": 0.7964,
      "step": 4083
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.919051170349121,
      "learning_rate": 1.685971685971686e-05,
      "loss": 0.8962,
      "step": 4084
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.43735408782959,
      "learning_rate": 1.6856499356499357e-05,
      "loss": 0.6329,
      "step": 4085
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.19359302520752,
      "learning_rate": 1.6853281853281853e-05,
      "loss": 1.0051,
      "step": 4086
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.420472145080566,
      "learning_rate": 1.685006435006435e-05,
      "loss": 0.7824,
      "step": 4087
    },
    {
      "epoch": 0.88,
      "grad_norm": 6.8004536628723145,
      "learning_rate": 1.6846846846846846e-05,
      "loss": 0.5713,
      "step": 4088
    },
    {
      "epoch": 0.88,
      "grad_norm": 12.142390251159668,
      "learning_rate": 1.6843629343629346e-05,
      "loss": 1.0913,
      "step": 4089
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.28418254852295,
      "learning_rate": 1.6840411840411842e-05,
      "loss": 0.7605,
      "step": 4090
    },
    {
      "epoch": 0.88,
      "grad_norm": 11.164363861083984,
      "learning_rate": 1.683719433719434e-05,
      "loss": 0.6279,
      "step": 4091
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.698633193969727,
      "learning_rate": 1.6833976833976835e-05,
      "loss": 0.8057,
      "step": 4092
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.851678848266602,
      "learning_rate": 1.683075933075933e-05,
      "loss": 0.665,
      "step": 4093
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.807636260986328,
      "learning_rate": 1.6827541827541828e-05,
      "loss": 0.5291,
      "step": 4094
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.262628555297852,
      "learning_rate": 1.6824324324324324e-05,
      "loss": 0.7041,
      "step": 4095
    },
    {
      "epoch": 0.88,
      "grad_norm": 12.593953132629395,
      "learning_rate": 1.6821106821106824e-05,
      "loss": 1.2914,
      "step": 4096
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.131528854370117,
      "learning_rate": 1.6817889317889317e-05,
      "loss": 0.915,
      "step": 4097
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.87954044342041,
      "learning_rate": 1.6814671814671814e-05,
      "loss": 0.8614,
      "step": 4098
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.554874420166016,
      "learning_rate": 1.681145431145431e-05,
      "loss": 0.9885,
      "step": 4099
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.428017616271973,
      "learning_rate": 1.6808236808236807e-05,
      "loss": 0.6216,
      "step": 4100
    },
    {
      "epoch": 0.88,
      "grad_norm": 13.984448432922363,
      "learning_rate": 1.6805019305019303e-05,
      "loss": 1.2237,
      "step": 4101
    },
    {
      "epoch": 0.88,
      "grad_norm": 13.110265731811523,
      "learning_rate": 1.6801801801801803e-05,
      "loss": 1.0687,
      "step": 4102
    },
    {
      "epoch": 0.88,
      "grad_norm": 11.38529109954834,
      "learning_rate": 1.67985842985843e-05,
      "loss": 0.8035,
      "step": 4103
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.699318885803223,
      "learning_rate": 1.6795366795366796e-05,
      "loss": 1.0681,
      "step": 4104
    },
    {
      "epoch": 0.88,
      "grad_norm": 11.140589714050293,
      "learning_rate": 1.6792149292149292e-05,
      "loss": 0.9519,
      "step": 4105
    },
    {
      "epoch": 0.88,
      "grad_norm": 11.549271583557129,
      "learning_rate": 1.678893178893179e-05,
      "loss": 1.3535,
      "step": 4106
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.530289649963379,
      "learning_rate": 1.6785714285714285e-05,
      "loss": 0.8354,
      "step": 4107
    },
    {
      "epoch": 0.88,
      "grad_norm": 11.790043830871582,
      "learning_rate": 1.6782496782496785e-05,
      "loss": 1.1514,
      "step": 4108
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.189008712768555,
      "learning_rate": 1.677927927927928e-05,
      "loss": 0.8457,
      "step": 4109
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.71516227722168,
      "learning_rate": 1.6776061776061778e-05,
      "loss": 0.7073,
      "step": 4110
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.366422653198242,
      "learning_rate": 1.6772844272844274e-05,
      "loss": 0.7257,
      "step": 4111
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.442183494567871,
      "learning_rate": 1.676962676962677e-05,
      "loss": 0.8508,
      "step": 4112
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.990274429321289,
      "learning_rate": 1.6766409266409267e-05,
      "loss": 0.7646,
      "step": 4113
    },
    {
      "epoch": 0.88,
      "grad_norm": 7.286981582641602,
      "learning_rate": 1.6763191763191764e-05,
      "loss": 0.8208,
      "step": 4114
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.519323348999023,
      "learning_rate": 1.675997425997426e-05,
      "loss": 0.9751,
      "step": 4115
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.882506370544434,
      "learning_rate": 1.6756756756756757e-05,
      "loss": 0.5963,
      "step": 4116
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.282934188842773,
      "learning_rate": 1.6753539253539253e-05,
      "loss": 0.8183,
      "step": 4117
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.020594596862793,
      "learning_rate": 1.675032175032175e-05,
      "loss": 0.7808,
      "step": 4118
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.727900505065918,
      "learning_rate": 1.6747104247104246e-05,
      "loss": 0.8422,
      "step": 4119
    },
    {
      "epoch": 0.88,
      "grad_norm": 12.267413139343262,
      "learning_rate": 1.6743886743886742e-05,
      "loss": 0.8649,
      "step": 4120
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.789545059204102,
      "learning_rate": 1.6740669240669242e-05,
      "loss": 0.8118,
      "step": 4121
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.60367202758789,
      "learning_rate": 1.673745173745174e-05,
      "loss": 0.8544,
      "step": 4122
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.77586841583252,
      "learning_rate": 1.6734234234234235e-05,
      "loss": 0.5471,
      "step": 4123
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.642894744873047,
      "learning_rate": 1.673101673101673e-05,
      "loss": 0.9439,
      "step": 4124
    },
    {
      "epoch": 0.88,
      "grad_norm": 11.107227325439453,
      "learning_rate": 1.6727799227799228e-05,
      "loss": 1.0963,
      "step": 4125
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.40048599243164,
      "learning_rate": 1.6724581724581725e-05,
      "loss": 0.7144,
      "step": 4126
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.815553665161133,
      "learning_rate": 1.672136422136422e-05,
      "loss": 0.6811,
      "step": 4127
    },
    {
      "epoch": 0.89,
      "grad_norm": 11.429134368896484,
      "learning_rate": 1.671814671814672e-05,
      "loss": 0.9653,
      "step": 4128
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.48087215423584,
      "learning_rate": 1.6714929214929217e-05,
      "loss": 0.8218,
      "step": 4129
    },
    {
      "epoch": 0.89,
      "grad_norm": 8.964700698852539,
      "learning_rate": 1.6711711711711714e-05,
      "loss": 0.7139,
      "step": 4130
    },
    {
      "epoch": 0.89,
      "grad_norm": 14.484170913696289,
      "learning_rate": 1.670849420849421e-05,
      "loss": 1.1836,
      "step": 4131
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.887022972106934,
      "learning_rate": 1.6705276705276703e-05,
      "loss": 0.8233,
      "step": 4132
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.065550804138184,
      "learning_rate": 1.67020592020592e-05,
      "loss": 0.9443,
      "step": 4133
    },
    {
      "epoch": 0.89,
      "grad_norm": 11.13161849975586,
      "learning_rate": 1.66988416988417e-05,
      "loss": 0.9229,
      "step": 4134
    },
    {
      "epoch": 0.89,
      "grad_norm": 12.273383140563965,
      "learning_rate": 1.6695624195624196e-05,
      "loss": 0.8722,
      "step": 4135
    },
    {
      "epoch": 0.89,
      "grad_norm": 8.824691772460938,
      "learning_rate": 1.6692406692406692e-05,
      "loss": 0.8519,
      "step": 4136
    },
    {
      "epoch": 0.89,
      "grad_norm": 7.503031253814697,
      "learning_rate": 1.668918918918919e-05,
      "loss": 0.5677,
      "step": 4137
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.638054847717285,
      "learning_rate": 1.6685971685971685e-05,
      "loss": 0.7688,
      "step": 4138
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.89691162109375,
      "learning_rate": 1.6682754182754182e-05,
      "loss": 0.6978,
      "step": 4139
    },
    {
      "epoch": 0.89,
      "grad_norm": 14.519096374511719,
      "learning_rate": 1.6679536679536678e-05,
      "loss": 1.1884,
      "step": 4140
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.839998245239258,
      "learning_rate": 1.6676319176319178e-05,
      "loss": 0.9104,
      "step": 4141
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.753812789916992,
      "learning_rate": 1.6673101673101675e-05,
      "loss": 1.1592,
      "step": 4142
    },
    {
      "epoch": 0.89,
      "grad_norm": 11.302083015441895,
      "learning_rate": 1.666988416988417e-05,
      "loss": 0.98,
      "step": 4143
    },
    {
      "epoch": 0.89,
      "grad_norm": 11.104068756103516,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 1.0131,
      "step": 4144
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.238642692565918,
      "learning_rate": 1.6663449163449164e-05,
      "loss": 0.9586,
      "step": 4145
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.600076675415039,
      "learning_rate": 1.666023166023166e-05,
      "loss": 1.0248,
      "step": 4146
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.486574172973633,
      "learning_rate": 1.665701415701416e-05,
      "loss": 0.9215,
      "step": 4147
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.816352844238281,
      "learning_rate": 1.6653796653796657e-05,
      "loss": 1.1936,
      "step": 4148
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.183245658874512,
      "learning_rate": 1.665057915057915e-05,
      "loss": 0.9053,
      "step": 4149
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.518891334533691,
      "learning_rate": 1.6647361647361646e-05,
      "loss": 0.9362,
      "step": 4150
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.407485961914062,
      "learning_rate": 1.6644144144144143e-05,
      "loss": 0.8885,
      "step": 4151
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.806634902954102,
      "learning_rate": 1.664092664092664e-05,
      "loss": 0.6506,
      "step": 4152
    },
    {
      "epoch": 0.89,
      "grad_norm": 8.465340614318848,
      "learning_rate": 1.663770913770914e-05,
      "loss": 0.8099,
      "step": 4153
    },
    {
      "epoch": 0.89,
      "grad_norm": 11.094554901123047,
      "learning_rate": 1.6634491634491635e-05,
      "loss": 0.8506,
      "step": 4154
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.963090896606445,
      "learning_rate": 1.6631274131274132e-05,
      "loss": 0.7677,
      "step": 4155
    },
    {
      "epoch": 0.89,
      "grad_norm": 7.547951698303223,
      "learning_rate": 1.6628056628056628e-05,
      "loss": 0.8186,
      "step": 4156
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.493752479553223,
      "learning_rate": 1.6624839124839125e-05,
      "loss": 0.917,
      "step": 4157
    },
    {
      "epoch": 0.89,
      "grad_norm": 7.766856670379639,
      "learning_rate": 1.662162162162162e-05,
      "loss": 0.7018,
      "step": 4158
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.714666366577148,
      "learning_rate": 1.6618404118404118e-05,
      "loss": 0.6423,
      "step": 4159
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.62034797668457,
      "learning_rate": 1.6615186615186617e-05,
      "loss": 1.0895,
      "step": 4160
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.686004638671875,
      "learning_rate": 1.6611969111969114e-05,
      "loss": 0.7747,
      "step": 4161
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.492143630981445,
      "learning_rate": 1.660875160875161e-05,
      "loss": 0.845,
      "step": 4162
    },
    {
      "epoch": 0.89,
      "grad_norm": 11.872940063476562,
      "learning_rate": 1.6605534105534107e-05,
      "loss": 1.0949,
      "step": 4163
    },
    {
      "epoch": 0.89,
      "grad_norm": 11.505887985229492,
      "learning_rate": 1.6602316602316603e-05,
      "loss": 1.015,
      "step": 4164
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.997215270996094,
      "learning_rate": 1.65990990990991e-05,
      "loss": 1.0556,
      "step": 4165
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.957977294921875,
      "learning_rate": 1.65958815958816e-05,
      "loss": 0.7262,
      "step": 4166
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.839820861816406,
      "learning_rate": 1.6592664092664093e-05,
      "loss": 0.7094,
      "step": 4167
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.641757011413574,
      "learning_rate": 1.658944658944659e-05,
      "loss": 0.826,
      "step": 4168
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.799592971801758,
      "learning_rate": 1.6586229086229086e-05,
      "loss": 1.1694,
      "step": 4169
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.171452522277832,
      "learning_rate": 1.6583011583011582e-05,
      "loss": 0.8425,
      "step": 4170
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.397672653198242,
      "learning_rate": 1.657979407979408e-05,
      "loss": 0.9763,
      "step": 4171
    },
    {
      "epoch": 0.89,
      "grad_norm": 13.059209823608398,
      "learning_rate": 1.6576576576576575e-05,
      "loss": 1.1964,
      "step": 4172
    },
    {
      "epoch": 0.9,
      "grad_norm": 10.51842212677002,
      "learning_rate": 1.6573359073359075e-05,
      "loss": 1.059,
      "step": 4173
    },
    {
      "epoch": 0.9,
      "grad_norm": 12.28489875793457,
      "learning_rate": 1.657014157014157e-05,
      "loss": 1.1155,
      "step": 4174
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.303770065307617,
      "learning_rate": 1.6566924066924068e-05,
      "loss": 0.6857,
      "step": 4175
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.474114418029785,
      "learning_rate": 1.6563706563706564e-05,
      "loss": 0.6362,
      "step": 4176
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.917424201965332,
      "learning_rate": 1.656048906048906e-05,
      "loss": 0.7035,
      "step": 4177
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.797896385192871,
      "learning_rate": 1.6557271557271557e-05,
      "loss": 0.8629,
      "step": 4178
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.923454761505127,
      "learning_rate": 1.6554054054054057e-05,
      "loss": 0.8515,
      "step": 4179
    },
    {
      "epoch": 0.9,
      "grad_norm": 11.690621376037598,
      "learning_rate": 1.6550836550836553e-05,
      "loss": 1.1701,
      "step": 4180
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.351786613464355,
      "learning_rate": 1.654761904761905e-05,
      "loss": 0.8837,
      "step": 4181
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.78341293334961,
      "learning_rate": 1.6544401544401546e-05,
      "loss": 1.0915,
      "step": 4182
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.144619941711426,
      "learning_rate": 1.6541184041184043e-05,
      "loss": 0.7439,
      "step": 4183
    },
    {
      "epoch": 0.9,
      "grad_norm": 12.03865909576416,
      "learning_rate": 1.6537966537966536e-05,
      "loss": 1.1759,
      "step": 4184
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.798328399658203,
      "learning_rate": 1.6534749034749032e-05,
      "loss": 0.6991,
      "step": 4185
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.018044471740723,
      "learning_rate": 1.6531531531531532e-05,
      "loss": 0.6514,
      "step": 4186
    },
    {
      "epoch": 0.9,
      "grad_norm": 11.061477661132812,
      "learning_rate": 1.652831402831403e-05,
      "loss": 0.989,
      "step": 4187
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.342840194702148,
      "learning_rate": 1.6525096525096525e-05,
      "loss": 0.8154,
      "step": 4188
    },
    {
      "epoch": 0.9,
      "grad_norm": 10.143793106079102,
      "learning_rate": 1.652187902187902e-05,
      "loss": 1.058,
      "step": 4189
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.897955894470215,
      "learning_rate": 1.6518661518661518e-05,
      "loss": 0.7559,
      "step": 4190
    },
    {
      "epoch": 0.9,
      "grad_norm": 6.98883581161499,
      "learning_rate": 1.6515444015444014e-05,
      "loss": 0.5691,
      "step": 4191
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.426319122314453,
      "learning_rate": 1.6512226512226514e-05,
      "loss": 0.8406,
      "step": 4192
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.383644104003906,
      "learning_rate": 1.650900900900901e-05,
      "loss": 0.9716,
      "step": 4193
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.764204025268555,
      "learning_rate": 1.6505791505791507e-05,
      "loss": 0.7568,
      "step": 4194
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.993236541748047,
      "learning_rate": 1.6502574002574003e-05,
      "loss": 0.8685,
      "step": 4195
    },
    {
      "epoch": 0.9,
      "grad_norm": 10.925533294677734,
      "learning_rate": 1.64993564993565e-05,
      "loss": 0.9797,
      "step": 4196
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.987920761108398,
      "learning_rate": 1.6496138996138996e-05,
      "loss": 0.8144,
      "step": 4197
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.713748931884766,
      "learning_rate": 1.6492921492921493e-05,
      "loss": 0.6396,
      "step": 4198
    },
    {
      "epoch": 0.9,
      "grad_norm": 10.107630729675293,
      "learning_rate": 1.6489703989703993e-05,
      "loss": 1.0453,
      "step": 4199
    },
    {
      "epoch": 0.9,
      "grad_norm": 10.834092140197754,
      "learning_rate": 1.648648648648649e-05,
      "loss": 0.9848,
      "step": 4200
    },
    {
      "epoch": 0.9,
      "grad_norm": 11.001203536987305,
      "learning_rate": 1.6483268983268986e-05,
      "loss": 1.1112,
      "step": 4201
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.611854553222656,
      "learning_rate": 1.648005148005148e-05,
      "loss": 0.9787,
      "step": 4202
    },
    {
      "epoch": 0.9,
      "grad_norm": 11.68568229675293,
      "learning_rate": 1.6476833976833975e-05,
      "loss": 1.0339,
      "step": 4203
    },
    {
      "epoch": 0.9,
      "grad_norm": 10.54709243774414,
      "learning_rate": 1.647361647361647e-05,
      "loss": 1.1844,
      "step": 4204
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.519516944885254,
      "learning_rate": 1.647039897039897e-05,
      "loss": 0.637,
      "step": 4205
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.589168548583984,
      "learning_rate": 1.6467181467181468e-05,
      "loss": 0.7439,
      "step": 4206
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.81467056274414,
      "learning_rate": 1.6463963963963964e-05,
      "loss": 0.8389,
      "step": 4207
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.912912845611572,
      "learning_rate": 1.646074646074646e-05,
      "loss": 0.794,
      "step": 4208
    },
    {
      "epoch": 0.9,
      "grad_norm": 8.881245613098145,
      "learning_rate": 1.6457528957528957e-05,
      "loss": 0.729,
      "step": 4209
    },
    {
      "epoch": 0.9,
      "grad_norm": 12.194906234741211,
      "learning_rate": 1.6454311454311454e-05,
      "loss": 0.8222,
      "step": 4210
    },
    {
      "epoch": 0.9,
      "grad_norm": 10.660468101501465,
      "learning_rate": 1.6451093951093953e-05,
      "loss": 1.1089,
      "step": 4211
    },
    {
      "epoch": 0.9,
      "grad_norm": 10.176159858703613,
      "learning_rate": 1.644787644787645e-05,
      "loss": 1.0194,
      "step": 4212
    },
    {
      "epoch": 0.9,
      "grad_norm": 10.685643196105957,
      "learning_rate": 1.6444658944658946e-05,
      "loss": 0.8641,
      "step": 4213
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.661497592926025,
      "learning_rate": 1.6441441441441443e-05,
      "loss": 0.5018,
      "step": 4214
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.665007591247559,
      "learning_rate": 1.643822393822394e-05,
      "loss": 0.9412,
      "step": 4215
    },
    {
      "epoch": 0.9,
      "grad_norm": 11.06138801574707,
      "learning_rate": 1.6435006435006436e-05,
      "loss": 0.9369,
      "step": 4216
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.398731231689453,
      "learning_rate": 1.6431788931788932e-05,
      "loss": 0.9707,
      "step": 4217
    },
    {
      "epoch": 0.9,
      "grad_norm": 10.57482624053955,
      "learning_rate": 1.6428571428571432e-05,
      "loss": 0.7987,
      "step": 4218
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.000255584716797,
      "learning_rate": 1.6425353925353925e-05,
      "loss": 0.6242,
      "step": 4219
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.508152961730957,
      "learning_rate": 1.642213642213642e-05,
      "loss": 0.9585,
      "step": 4220
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.391524314880371,
      "learning_rate": 1.6418918918918918e-05,
      "loss": 0.9169,
      "step": 4221
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.891660690307617,
      "learning_rate": 1.6415701415701414e-05,
      "loss": 1.005,
      "step": 4222
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.582242965698242,
      "learning_rate": 1.641248391248391e-05,
      "loss": 0.6907,
      "step": 4223
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.554920196533203,
      "learning_rate": 1.640926640926641e-05,
      "loss": 0.9992,
      "step": 4224
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.171951293945312,
      "learning_rate": 1.6406048906048907e-05,
      "loss": 0.9343,
      "step": 4225
    },
    {
      "epoch": 0.91,
      "grad_norm": 14.242798805236816,
      "learning_rate": 1.6402831402831404e-05,
      "loss": 0.88,
      "step": 4226
    },
    {
      "epoch": 0.91,
      "grad_norm": 7.930697917938232,
      "learning_rate": 1.63996138996139e-05,
      "loss": 0.7272,
      "step": 4227
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.420123100280762,
      "learning_rate": 1.6396396396396396e-05,
      "loss": 0.9673,
      "step": 4228
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.531608581542969,
      "learning_rate": 1.6393178893178893e-05,
      "loss": 0.8128,
      "step": 4229
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.463180541992188,
      "learning_rate": 1.638996138996139e-05,
      "loss": 0.653,
      "step": 4230
    },
    {
      "epoch": 0.91,
      "grad_norm": 11.002796173095703,
      "learning_rate": 1.638674388674389e-05,
      "loss": 0.7592,
      "step": 4231
    },
    {
      "epoch": 0.91,
      "grad_norm": 12.102829933166504,
      "learning_rate": 1.6383526383526386e-05,
      "loss": 1.0282,
      "step": 4232
    },
    {
      "epoch": 0.91,
      "grad_norm": 11.043006896972656,
      "learning_rate": 1.6380308880308882e-05,
      "loss": 1.1301,
      "step": 4233
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.27319622039795,
      "learning_rate": 1.637709137709138e-05,
      "loss": 0.6109,
      "step": 4234
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.581725120544434,
      "learning_rate": 1.6373873873873875e-05,
      "loss": 0.7895,
      "step": 4235
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.971217155456543,
      "learning_rate": 1.6370656370656368e-05,
      "loss": 1.1088,
      "step": 4236
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.969149589538574,
      "learning_rate": 1.6367438867438868e-05,
      "loss": 0.8321,
      "step": 4237
    },
    {
      "epoch": 0.91,
      "grad_norm": 13.77183723449707,
      "learning_rate": 1.6364221364221364e-05,
      "loss": 0.7893,
      "step": 4238
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.814791679382324,
      "learning_rate": 1.636100386100386e-05,
      "loss": 0.634,
      "step": 4239
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.707128524780273,
      "learning_rate": 1.6357786357786357e-05,
      "loss": 0.7804,
      "step": 4240
    },
    {
      "epoch": 0.91,
      "grad_norm": 11.78310489654541,
      "learning_rate": 1.6354568854568854e-05,
      "loss": 0.9493,
      "step": 4241
    },
    {
      "epoch": 0.91,
      "grad_norm": 13.472021102905273,
      "learning_rate": 1.635135135135135e-05,
      "loss": 0.8394,
      "step": 4242
    },
    {
      "epoch": 0.91,
      "grad_norm": 12.39739990234375,
      "learning_rate": 1.6348133848133847e-05,
      "loss": 0.987,
      "step": 4243
    },
    {
      "epoch": 0.91,
      "grad_norm": 13.492223739624023,
      "learning_rate": 1.6344916344916346e-05,
      "loss": 0.9665,
      "step": 4244
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.766014099121094,
      "learning_rate": 1.6341698841698843e-05,
      "loss": 0.8122,
      "step": 4245
    },
    {
      "epoch": 0.91,
      "grad_norm": 11.07912826538086,
      "learning_rate": 1.633848133848134e-05,
      "loss": 1.1686,
      "step": 4246
    },
    {
      "epoch": 0.91,
      "grad_norm": 11.097978591918945,
      "learning_rate": 1.6335263835263836e-05,
      "loss": 0.8661,
      "step": 4247
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.78089427947998,
      "learning_rate": 1.6332046332046332e-05,
      "loss": 0.7381,
      "step": 4248
    },
    {
      "epoch": 0.91,
      "grad_norm": 11.591184616088867,
      "learning_rate": 1.632882882882883e-05,
      "loss": 0.8553,
      "step": 4249
    },
    {
      "epoch": 0.91,
      "grad_norm": 11.314517974853516,
      "learning_rate": 1.632561132561133e-05,
      "loss": 0.9932,
      "step": 4250
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.90932846069336,
      "learning_rate": 1.6322393822393825e-05,
      "loss": 0.9117,
      "step": 4251
    },
    {
      "epoch": 0.91,
      "grad_norm": 11.431105613708496,
      "learning_rate": 1.631917631917632e-05,
      "loss": 1.0848,
      "step": 4252
    },
    {
      "epoch": 0.91,
      "grad_norm": 14.272811889648438,
      "learning_rate": 1.6315958815958818e-05,
      "loss": 0.923,
      "step": 4253
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.24299430847168,
      "learning_rate": 1.631274131274131e-05,
      "loss": 0.5016,
      "step": 4254
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.920005798339844,
      "learning_rate": 1.6309523809523807e-05,
      "loss": 1.1044,
      "step": 4255
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.586575508117676,
      "learning_rate": 1.6306306306306304e-05,
      "loss": 0.7507,
      "step": 4256
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.286280632019043,
      "learning_rate": 1.6303088803088804e-05,
      "loss": 0.657,
      "step": 4257
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.766932487487793,
      "learning_rate": 1.62998712998713e-05,
      "loss": 0.6846,
      "step": 4258
    },
    {
      "epoch": 0.91,
      "grad_norm": 11.871316909790039,
      "learning_rate": 1.6296653796653797e-05,
      "loss": 0.9927,
      "step": 4259
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.601190567016602,
      "learning_rate": 1.6293436293436293e-05,
      "loss": 0.7695,
      "step": 4260
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.14179515838623,
      "learning_rate": 1.629021879021879e-05,
      "loss": 0.82,
      "step": 4261
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.425090789794922,
      "learning_rate": 1.6287001287001286e-05,
      "loss": 0.8225,
      "step": 4262
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.900506973266602,
      "learning_rate": 1.6283783783783786e-05,
      "loss": 0.8458,
      "step": 4263
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.403993606567383,
      "learning_rate": 1.6280566280566282e-05,
      "loss": 0.7434,
      "step": 4264
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.78586196899414,
      "learning_rate": 1.627734877734878e-05,
      "loss": 0.788,
      "step": 4265
    },
    {
      "epoch": 0.92,
      "grad_norm": 7.161960124969482,
      "learning_rate": 1.6274131274131275e-05,
      "loss": 0.6095,
      "step": 4266
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.971805572509766,
      "learning_rate": 1.627091377091377e-05,
      "loss": 0.7894,
      "step": 4267
    },
    {
      "epoch": 0.92,
      "grad_norm": 11.684466361999512,
      "learning_rate": 1.6267696267696268e-05,
      "loss": 0.8365,
      "step": 4268
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.500791549682617,
      "learning_rate": 1.6264478764478768e-05,
      "loss": 1.0218,
      "step": 4269
    },
    {
      "epoch": 0.92,
      "grad_norm": 11.179215431213379,
      "learning_rate": 1.6261261261261264e-05,
      "loss": 1.2099,
      "step": 4270
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.67319393157959,
      "learning_rate": 1.6258043758043757e-05,
      "loss": 1.0322,
      "step": 4271
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.987672805786133,
      "learning_rate": 1.6254826254826254e-05,
      "loss": 0.9253,
      "step": 4272
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.419499397277832,
      "learning_rate": 1.625160875160875e-05,
      "loss": 0.6823,
      "step": 4273
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.973884582519531,
      "learning_rate": 1.6248391248391247e-05,
      "loss": 0.8509,
      "step": 4274
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.063250541687012,
      "learning_rate": 1.6245173745173743e-05,
      "loss": 0.6395,
      "step": 4275
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.593143463134766,
      "learning_rate": 1.6241956241956243e-05,
      "loss": 0.8313,
      "step": 4276
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.635972023010254,
      "learning_rate": 1.623873873873874e-05,
      "loss": 0.6435,
      "step": 4277
    },
    {
      "epoch": 0.92,
      "grad_norm": 7.935305118560791,
      "learning_rate": 1.6235521235521236e-05,
      "loss": 0.7326,
      "step": 4278
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.656042098999023,
      "learning_rate": 1.6232303732303732e-05,
      "loss": 0.8965,
      "step": 4279
    },
    {
      "epoch": 0.92,
      "grad_norm": 12.862712860107422,
      "learning_rate": 1.622908622908623e-05,
      "loss": 0.8383,
      "step": 4280
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.528581619262695,
      "learning_rate": 1.6225868725868725e-05,
      "loss": 0.6499,
      "step": 4281
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.898799896240234,
      "learning_rate": 1.6222651222651225e-05,
      "loss": 0.8217,
      "step": 4282
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.854244232177734,
      "learning_rate": 1.621943371943372e-05,
      "loss": 0.9227,
      "step": 4283
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.805638313293457,
      "learning_rate": 1.6216216216216218e-05,
      "loss": 1.1363,
      "step": 4284
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.760452270507812,
      "learning_rate": 1.6212998712998715e-05,
      "loss": 0.9125,
      "step": 4285
    },
    {
      "epoch": 0.92,
      "grad_norm": 11.192549705505371,
      "learning_rate": 1.620978120978121e-05,
      "loss": 1.0482,
      "step": 4286
    },
    {
      "epoch": 0.92,
      "grad_norm": 11.909960746765137,
      "learning_rate": 1.6206563706563707e-05,
      "loss": 1.0135,
      "step": 4287
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.909111022949219,
      "learning_rate": 1.62033462033462e-05,
      "loss": 1.048,
      "step": 4288
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.29076862335205,
      "learning_rate": 1.62001287001287e-05,
      "loss": 0.7935,
      "step": 4289
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.961067199707031,
      "learning_rate": 1.6196911196911197e-05,
      "loss": 0.889,
      "step": 4290
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.188336372375488,
      "learning_rate": 1.6193693693693693e-05,
      "loss": 0.984,
      "step": 4291
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.043818473815918,
      "learning_rate": 1.619047619047619e-05,
      "loss": 0.9231,
      "step": 4292
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.312540054321289,
      "learning_rate": 1.6187258687258686e-05,
      "loss": 0.8115,
      "step": 4293
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.626649856567383,
      "learning_rate": 1.6184041184041183e-05,
      "loss": 0.666,
      "step": 4294
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.341856002807617,
      "learning_rate": 1.6180823680823682e-05,
      "loss": 0.8121,
      "step": 4295
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.370980262756348,
      "learning_rate": 1.617760617760618e-05,
      "loss": 0.595,
      "step": 4296
    },
    {
      "epoch": 0.92,
      "grad_norm": 13.093385696411133,
      "learning_rate": 1.6174388674388675e-05,
      "loss": 0.9737,
      "step": 4297
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.460545539855957,
      "learning_rate": 1.6171171171171172e-05,
      "loss": 0.6377,
      "step": 4298
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.759344100952148,
      "learning_rate": 1.6167953667953668e-05,
      "loss": 0.9459,
      "step": 4299
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.6841402053833,
      "learning_rate": 1.6164736164736165e-05,
      "loss": 0.8325,
      "step": 4300
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.908366203308105,
      "learning_rate": 1.616151866151866e-05,
      "loss": 0.9038,
      "step": 4301
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.389922142028809,
      "learning_rate": 1.615830115830116e-05,
      "loss": 0.8877,
      "step": 4302
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.196693420410156,
      "learning_rate": 1.6155083655083657e-05,
      "loss": 0.675,
      "step": 4303
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.034409523010254,
      "learning_rate": 1.6151866151866154e-05,
      "loss": 0.6014,
      "step": 4304
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.8970947265625,
      "learning_rate": 1.614864864864865e-05,
      "loss": 0.8785,
      "step": 4305
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.321054458618164,
      "learning_rate": 1.6145431145431143e-05,
      "loss": 0.7952,
      "step": 4306
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.547527313232422,
      "learning_rate": 1.614221364221364e-05,
      "loss": 0.9276,
      "step": 4307
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.80158519744873,
      "learning_rate": 1.613899613899614e-05,
      "loss": 0.7398,
      "step": 4308
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.776884078979492,
      "learning_rate": 1.6135778635778636e-05,
      "loss": 0.8939,
      "step": 4309
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.711041450500488,
      "learning_rate": 1.6132561132561133e-05,
      "loss": 0.8504,
      "step": 4310
    },
    {
      "epoch": 0.92,
      "grad_norm": 11.654593467712402,
      "learning_rate": 1.612934362934363e-05,
      "loss": 0.8929,
      "step": 4311
    },
    {
      "epoch": 0.92,
      "grad_norm": 11.703030586242676,
      "learning_rate": 1.6126126126126126e-05,
      "loss": 0.8392,
      "step": 4312
    },
    {
      "epoch": 0.93,
      "grad_norm": 12.790799140930176,
      "learning_rate": 1.6122908622908622e-05,
      "loss": 0.861,
      "step": 4313
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.382654190063477,
      "learning_rate": 1.611969111969112e-05,
      "loss": 0.81,
      "step": 4314
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.315925598144531,
      "learning_rate": 1.6116473616473618e-05,
      "loss": 0.9844,
      "step": 4315
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.714505195617676,
      "learning_rate": 1.6113256113256115e-05,
      "loss": 0.8235,
      "step": 4316
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.005294799804688,
      "learning_rate": 1.611003861003861e-05,
      "loss": 0.8217,
      "step": 4317
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.43355655670166,
      "learning_rate": 1.6106821106821108e-05,
      "loss": 0.9785,
      "step": 4318
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.052366256713867,
      "learning_rate": 1.6103603603603604e-05,
      "loss": 0.5596,
      "step": 4319
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.761674880981445,
      "learning_rate": 1.61003861003861e-05,
      "loss": 0.7269,
      "step": 4320
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.046555519104004,
      "learning_rate": 1.60971685971686e-05,
      "loss": 0.9781,
      "step": 4321
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.67860221862793,
      "learning_rate": 1.6093951093951097e-05,
      "loss": 0.8305,
      "step": 4322
    },
    {
      "epoch": 0.93,
      "grad_norm": 11.413511276245117,
      "learning_rate": 1.609073359073359e-05,
      "loss": 0.9934,
      "step": 4323
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.343897819519043,
      "learning_rate": 1.6087516087516086e-05,
      "loss": 0.8434,
      "step": 4324
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.067716598510742,
      "learning_rate": 1.6084298584298583e-05,
      "loss": 0.75,
      "step": 4325
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.1865873336792,
      "learning_rate": 1.608108108108108e-05,
      "loss": 0.89,
      "step": 4326
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.815951347351074,
      "learning_rate": 1.607786357786358e-05,
      "loss": 0.8141,
      "step": 4327
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.392227172851562,
      "learning_rate": 1.6074646074646076e-05,
      "loss": 0.7847,
      "step": 4328
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.34949779510498,
      "learning_rate": 1.6071428571428572e-05,
      "loss": 0.9751,
      "step": 4329
    },
    {
      "epoch": 0.93,
      "grad_norm": 11.100339889526367,
      "learning_rate": 1.606821106821107e-05,
      "loss": 1.2091,
      "step": 4330
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.04423713684082,
      "learning_rate": 1.6064993564993565e-05,
      "loss": 0.923,
      "step": 4331
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.111599922180176,
      "learning_rate": 1.606177606177606e-05,
      "loss": 0.6557,
      "step": 4332
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.503087043762207,
      "learning_rate": 1.6058558558558558e-05,
      "loss": 0.9067,
      "step": 4333
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.4240083694458,
      "learning_rate": 1.6055341055341058e-05,
      "loss": 0.605,
      "step": 4334
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.319803237915039,
      "learning_rate": 1.6052123552123554e-05,
      "loss": 0.8925,
      "step": 4335
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.779838562011719,
      "learning_rate": 1.604890604890605e-05,
      "loss": 0.8283,
      "step": 4336
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.48422622680664,
      "learning_rate": 1.6045688545688547e-05,
      "loss": 0.9546,
      "step": 4337
    },
    {
      "epoch": 0.93,
      "grad_norm": 12.330033302307129,
      "learning_rate": 1.6042471042471043e-05,
      "loss": 1.1789,
      "step": 4338
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.101386070251465,
      "learning_rate": 1.603925353925354e-05,
      "loss": 0.9296,
      "step": 4339
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.24388599395752,
      "learning_rate": 1.603603603603604e-05,
      "loss": 0.9069,
      "step": 4340
    },
    {
      "epoch": 0.93,
      "grad_norm": 12.276779174804688,
      "learning_rate": 1.6032818532818533e-05,
      "loss": 1.1371,
      "step": 4341
    },
    {
      "epoch": 0.93,
      "grad_norm": 11.269691467285156,
      "learning_rate": 1.602960102960103e-05,
      "loss": 0.9417,
      "step": 4342
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.50936508178711,
      "learning_rate": 1.6026383526383526e-05,
      "loss": 0.7663,
      "step": 4343
    },
    {
      "epoch": 0.93,
      "grad_norm": 12.500214576721191,
      "learning_rate": 1.6023166023166022e-05,
      "loss": 1.2525,
      "step": 4344
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.11142635345459,
      "learning_rate": 1.601994851994852e-05,
      "loss": 0.9186,
      "step": 4345
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.829530715942383,
      "learning_rate": 1.6016731016731015e-05,
      "loss": 0.6937,
      "step": 4346
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.794332504272461,
      "learning_rate": 1.6013513513513515e-05,
      "loss": 0.9311,
      "step": 4347
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.239508628845215,
      "learning_rate": 1.601029601029601e-05,
      "loss": 0.7776,
      "step": 4348
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.2250394821167,
      "learning_rate": 1.6007078507078508e-05,
      "loss": 0.874,
      "step": 4349
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.475822448730469,
      "learning_rate": 1.6003861003861004e-05,
      "loss": 1.0596,
      "step": 4350
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.5763521194458,
      "learning_rate": 1.60006435006435e-05,
      "loss": 0.8333,
      "step": 4351
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.527573585510254,
      "learning_rate": 1.5997425997425997e-05,
      "loss": 0.9731,
      "step": 4352
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.634811401367188,
      "learning_rate": 1.5994208494208497e-05,
      "loss": 0.7174,
      "step": 4353
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.186737060546875,
      "learning_rate": 1.5990990990990993e-05,
      "loss": 0.7401,
      "step": 4354
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.275839805603027,
      "learning_rate": 1.598777348777349e-05,
      "loss": 0.8308,
      "step": 4355
    },
    {
      "epoch": 0.93,
      "grad_norm": 12.58425521850586,
      "learning_rate": 1.5984555984555986e-05,
      "loss": 0.812,
      "step": 4356
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.421402931213379,
      "learning_rate": 1.5981338481338483e-05,
      "loss": 0.8218,
      "step": 4357
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.017855644226074,
      "learning_rate": 1.5978120978120976e-05,
      "loss": 0.7785,
      "step": 4358
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.57033634185791,
      "learning_rate": 1.5974903474903472e-05,
      "loss": 1.002,
      "step": 4359
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.895001411437988,
      "learning_rate": 1.5971685971685972e-05,
      "loss": 0.7979,
      "step": 4360
    },
    {
      "epoch": 0.94,
      "grad_norm": 8.931422233581543,
      "learning_rate": 1.596846846846847e-05,
      "loss": 0.8182,
      "step": 4361
    },
    {
      "epoch": 0.94,
      "grad_norm": 8.077957153320312,
      "learning_rate": 1.5965250965250965e-05,
      "loss": 0.7161,
      "step": 4362
    },
    {
      "epoch": 0.94,
      "grad_norm": 11.420437812805176,
      "learning_rate": 1.596203346203346e-05,
      "loss": 1.0206,
      "step": 4363
    },
    {
      "epoch": 0.94,
      "grad_norm": 12.171106338500977,
      "learning_rate": 1.5958815958815958e-05,
      "loss": 1.0145,
      "step": 4364
    },
    {
      "epoch": 0.94,
      "grad_norm": 11.085225105285645,
      "learning_rate": 1.5955598455598454e-05,
      "loss": 0.7726,
      "step": 4365
    },
    {
      "epoch": 0.94,
      "grad_norm": 13.224983215332031,
      "learning_rate": 1.5952380952380954e-05,
      "loss": 1.2187,
      "step": 4366
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.269612312316895,
      "learning_rate": 1.594916344916345e-05,
      "loss": 0.7986,
      "step": 4367
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.22840690612793,
      "learning_rate": 1.5945945945945947e-05,
      "loss": 0.7254,
      "step": 4368
    },
    {
      "epoch": 0.94,
      "grad_norm": 8.2008056640625,
      "learning_rate": 1.5942728442728444e-05,
      "loss": 0.7219,
      "step": 4369
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.715388298034668,
      "learning_rate": 1.593951093951094e-05,
      "loss": 0.9882,
      "step": 4370
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.232000350952148,
      "learning_rate": 1.5936293436293436e-05,
      "loss": 0.8233,
      "step": 4371
    },
    {
      "epoch": 0.94,
      "grad_norm": 12.674283981323242,
      "learning_rate": 1.5933075933075933e-05,
      "loss": 0.9041,
      "step": 4372
    },
    {
      "epoch": 0.94,
      "grad_norm": 11.938748359680176,
      "learning_rate": 1.5929858429858433e-05,
      "loss": 0.9291,
      "step": 4373
    },
    {
      "epoch": 0.94,
      "grad_norm": 12.67946720123291,
      "learning_rate": 1.592664092664093e-05,
      "loss": 0.9312,
      "step": 4374
    },
    {
      "epoch": 0.94,
      "grad_norm": 13.040181159973145,
      "learning_rate": 1.5923423423423422e-05,
      "loss": 0.9977,
      "step": 4375
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.280299186706543,
      "learning_rate": 1.592020592020592e-05,
      "loss": 0.7718,
      "step": 4376
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.810498237609863,
      "learning_rate": 1.5916988416988415e-05,
      "loss": 0.8427,
      "step": 4377
    },
    {
      "epoch": 0.94,
      "grad_norm": 8.147063255310059,
      "learning_rate": 1.591377091377091e-05,
      "loss": 0.6358,
      "step": 4378
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.775753021240234,
      "learning_rate": 1.591055341055341e-05,
      "loss": 0.7067,
      "step": 4379
    },
    {
      "epoch": 0.94,
      "grad_norm": 12.275445938110352,
      "learning_rate": 1.5907335907335908e-05,
      "loss": 1.1323,
      "step": 4380
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.89864444732666,
      "learning_rate": 1.5904118404118404e-05,
      "loss": 0.838,
      "step": 4381
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.11995792388916,
      "learning_rate": 1.59009009009009e-05,
      "loss": 0.7461,
      "step": 4382
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.352688789367676,
      "learning_rate": 1.5897683397683397e-05,
      "loss": 0.5916,
      "step": 4383
    },
    {
      "epoch": 0.94,
      "grad_norm": 8.86256217956543,
      "learning_rate": 1.5894465894465894e-05,
      "loss": 0.7077,
      "step": 4384
    },
    {
      "epoch": 0.94,
      "grad_norm": 8.036880493164062,
      "learning_rate": 1.5891248391248394e-05,
      "loss": 0.675,
      "step": 4385
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.390562057495117,
      "learning_rate": 1.588803088803089e-05,
      "loss": 0.8641,
      "step": 4386
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.162092208862305,
      "learning_rate": 1.5884813384813386e-05,
      "loss": 0.9371,
      "step": 4387
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.876238822937012,
      "learning_rate": 1.5881595881595883e-05,
      "loss": 0.9264,
      "step": 4388
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.046712875366211,
      "learning_rate": 1.587837837837838e-05,
      "loss": 0.9829,
      "step": 4389
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.87602710723877,
      "learning_rate": 1.5875160875160876e-05,
      "loss": 0.9731,
      "step": 4390
    },
    {
      "epoch": 0.94,
      "grad_norm": 7.927401542663574,
      "learning_rate": 1.5871943371943372e-05,
      "loss": 0.5756,
      "step": 4391
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.088887214660645,
      "learning_rate": 1.5868725868725872e-05,
      "loss": 0.8226,
      "step": 4392
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.662688255310059,
      "learning_rate": 1.5865508365508365e-05,
      "loss": 0.7505,
      "step": 4393
    },
    {
      "epoch": 0.94,
      "grad_norm": 8.099783897399902,
      "learning_rate": 1.586229086229086e-05,
      "loss": 0.7013,
      "step": 4394
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.709908485412598,
      "learning_rate": 1.5859073359073358e-05,
      "loss": 0.8305,
      "step": 4395
    },
    {
      "epoch": 0.94,
      "grad_norm": 11.83540153503418,
      "learning_rate": 1.5855855855855855e-05,
      "loss": 1.191,
      "step": 4396
    },
    {
      "epoch": 0.94,
      "grad_norm": 12.421614646911621,
      "learning_rate": 1.585263835263835e-05,
      "loss": 0.8495,
      "step": 4397
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.75296688079834,
      "learning_rate": 1.584942084942085e-05,
      "loss": 0.712,
      "step": 4398
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.496373176574707,
      "learning_rate": 1.5846203346203347e-05,
      "loss": 0.8003,
      "step": 4399
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.400574684143066,
      "learning_rate": 1.5842985842985844e-05,
      "loss": 0.767,
      "step": 4400
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.23278522491455,
      "learning_rate": 1.583976833976834e-05,
      "loss": 0.8372,
      "step": 4401
    },
    {
      "epoch": 0.94,
      "grad_norm": 8.930191040039062,
      "learning_rate": 1.5836550836550837e-05,
      "loss": 0.8885,
      "step": 4402
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.652266502380371,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 0.8241,
      "step": 4403
    },
    {
      "epoch": 0.94,
      "grad_norm": 14.243931770324707,
      "learning_rate": 1.583011583011583e-05,
      "loss": 0.9451,
      "step": 4404
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.471721649169922,
      "learning_rate": 1.582689832689833e-05,
      "loss": 0.8673,
      "step": 4405
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.331722259521484,
      "learning_rate": 1.5823680823680826e-05,
      "loss": 0.6597,
      "step": 4406
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.642090797424316,
      "learning_rate": 1.5820463320463322e-05,
      "loss": 0.8091,
      "step": 4407
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.660026550292969,
      "learning_rate": 1.581724581724582e-05,
      "loss": 0.8834,
      "step": 4408
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.168632507324219,
      "learning_rate": 1.5814028314028315e-05,
      "loss": 0.7498,
      "step": 4409
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.843535423278809,
      "learning_rate": 1.5810810810810808e-05,
      "loss": 0.7939,
      "step": 4410
    },
    {
      "epoch": 0.95,
      "grad_norm": 11.14784049987793,
      "learning_rate": 1.5807593307593308e-05,
      "loss": 0.9108,
      "step": 4411
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.525261878967285,
      "learning_rate": 1.5804375804375805e-05,
      "loss": 0.8983,
      "step": 4412
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.499561309814453,
      "learning_rate": 1.58011583011583e-05,
      "loss": 0.9774,
      "step": 4413
    },
    {
      "epoch": 0.95,
      "grad_norm": 7.5778889656066895,
      "learning_rate": 1.5797940797940797e-05,
      "loss": 0.6993,
      "step": 4414
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.131107330322266,
      "learning_rate": 1.5794723294723294e-05,
      "loss": 0.8174,
      "step": 4415
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.970832824707031,
      "learning_rate": 1.579150579150579e-05,
      "loss": 0.7997,
      "step": 4416
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.193933486938477,
      "learning_rate": 1.5788288288288287e-05,
      "loss": 0.7324,
      "step": 4417
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.24184513092041,
      "learning_rate": 1.5785070785070787e-05,
      "loss": 0.9075,
      "step": 4418
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.617568969726562,
      "learning_rate": 1.5781853281853283e-05,
      "loss": 0.7056,
      "step": 4419
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.716185569763184,
      "learning_rate": 1.577863577863578e-05,
      "loss": 0.757,
      "step": 4420
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.661192893981934,
      "learning_rate": 1.5775418275418276e-05,
      "loss": 0.6419,
      "step": 4421
    },
    {
      "epoch": 0.95,
      "grad_norm": 11.558099746704102,
      "learning_rate": 1.5772200772200772e-05,
      "loss": 0.7929,
      "step": 4422
    },
    {
      "epoch": 0.95,
      "grad_norm": 11.752252578735352,
      "learning_rate": 1.576898326898327e-05,
      "loss": 1.051,
      "step": 4423
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.862500190734863,
      "learning_rate": 1.576576576576577e-05,
      "loss": 0.7474,
      "step": 4424
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.678980827331543,
      "learning_rate": 1.5762548262548265e-05,
      "loss": 0.8178,
      "step": 4425
    },
    {
      "epoch": 0.95,
      "grad_norm": 13.630215644836426,
      "learning_rate": 1.575933075933076e-05,
      "loss": 0.945,
      "step": 4426
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.703466415405273,
      "learning_rate": 1.5756113256113258e-05,
      "loss": 0.6354,
      "step": 4427
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.204572677612305,
      "learning_rate": 1.575289575289575e-05,
      "loss": 0.9364,
      "step": 4428
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.60012149810791,
      "learning_rate": 1.5749678249678248e-05,
      "loss": 0.8963,
      "step": 4429
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.367400169372559,
      "learning_rate": 1.5746460746460747e-05,
      "loss": 1.0046,
      "step": 4430
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.629361152648926,
      "learning_rate": 1.5743243243243244e-05,
      "loss": 0.7543,
      "step": 4431
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.969871520996094,
      "learning_rate": 1.574002574002574e-05,
      "loss": 1.0015,
      "step": 4432
    },
    {
      "epoch": 0.95,
      "grad_norm": 11.248541831970215,
      "learning_rate": 1.5736808236808237e-05,
      "loss": 0.9109,
      "step": 4433
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.481059074401855,
      "learning_rate": 1.5733590733590733e-05,
      "loss": 0.8236,
      "step": 4434
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.850608825683594,
      "learning_rate": 1.573037323037323e-05,
      "loss": 0.8677,
      "step": 4435
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.305078506469727,
      "learning_rate": 1.5727155727155726e-05,
      "loss": 0.67,
      "step": 4436
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.933782577514648,
      "learning_rate": 1.5723938223938226e-05,
      "loss": 0.6858,
      "step": 4437
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.775809288024902,
      "learning_rate": 1.5720720720720722e-05,
      "loss": 0.8932,
      "step": 4438
    },
    {
      "epoch": 0.95,
      "grad_norm": 13.979836463928223,
      "learning_rate": 1.571750321750322e-05,
      "loss": 1.1052,
      "step": 4439
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.806882858276367,
      "learning_rate": 1.5714285714285715e-05,
      "loss": 0.6465,
      "step": 4440
    },
    {
      "epoch": 0.95,
      "grad_norm": 11.578160285949707,
      "learning_rate": 1.5711068211068212e-05,
      "loss": 1.0076,
      "step": 4441
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.732454299926758,
      "learning_rate": 1.5707850707850708e-05,
      "loss": 0.9771,
      "step": 4442
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.646568298339844,
      "learning_rate": 1.5704633204633208e-05,
      "loss": 0.7649,
      "step": 4443
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.544550895690918,
      "learning_rate": 1.5701415701415705e-05,
      "loss": 0.6544,
      "step": 4444
    },
    {
      "epoch": 0.95,
      "grad_norm": 11.050060272216797,
      "learning_rate": 1.5698198198198198e-05,
      "loss": 0.8307,
      "step": 4445
    },
    {
      "epoch": 0.95,
      "grad_norm": 11.58207893371582,
      "learning_rate": 1.5694980694980694e-05,
      "loss": 0.9882,
      "step": 4446
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.124963760375977,
      "learning_rate": 1.569176319176319e-05,
      "loss": 0.848,
      "step": 4447
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.418510437011719,
      "learning_rate": 1.5688545688545687e-05,
      "loss": 0.7533,
      "step": 4448
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.875624656677246,
      "learning_rate": 1.5685328185328183e-05,
      "loss": 0.8592,
      "step": 4449
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.777322769165039,
      "learning_rate": 1.5682110682110683e-05,
      "loss": 1.0251,
      "step": 4450
    },
    {
      "epoch": 0.95,
      "grad_norm": 7.607450485229492,
      "learning_rate": 1.567889317889318e-05,
      "loss": 0.5861,
      "step": 4451
    },
    {
      "epoch": 0.95,
      "grad_norm": 11.149787902832031,
      "learning_rate": 1.5675675675675676e-05,
      "loss": 0.6413,
      "step": 4452
    },
    {
      "epoch": 0.96,
      "grad_norm": 6.809457302093506,
      "learning_rate": 1.5672458172458173e-05,
      "loss": 0.6813,
      "step": 4453
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.271689414978027,
      "learning_rate": 1.566924066924067e-05,
      "loss": 0.8404,
      "step": 4454
    },
    {
      "epoch": 0.96,
      "grad_norm": 12.895333290100098,
      "learning_rate": 1.5666023166023165e-05,
      "loss": 0.8766,
      "step": 4455
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.736028671264648,
      "learning_rate": 1.5662805662805665e-05,
      "loss": 0.9249,
      "step": 4456
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.42031478881836,
      "learning_rate": 1.5659588159588162e-05,
      "loss": 0.9981,
      "step": 4457
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.343477249145508,
      "learning_rate": 1.5656370656370658e-05,
      "loss": 0.7377,
      "step": 4458
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.399075508117676,
      "learning_rate": 1.5653153153153155e-05,
      "loss": 0.6314,
      "step": 4459
    },
    {
      "epoch": 0.96,
      "grad_norm": 12.996109962463379,
      "learning_rate": 1.564993564993565e-05,
      "loss": 0.9922,
      "step": 4460
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.720706939697266,
      "learning_rate": 1.5646718146718148e-05,
      "loss": 0.6803,
      "step": 4461
    },
    {
      "epoch": 0.96,
      "grad_norm": 14.507105827331543,
      "learning_rate": 1.564350064350064e-05,
      "loss": 0.9067,
      "step": 4462
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.579784393310547,
      "learning_rate": 1.564028314028314e-05,
      "loss": 0.6877,
      "step": 4463
    },
    {
      "epoch": 0.96,
      "grad_norm": 11.16296100616455,
      "learning_rate": 1.5637065637065637e-05,
      "loss": 0.8401,
      "step": 4464
    },
    {
      "epoch": 0.96,
      "grad_norm": 13.159083366394043,
      "learning_rate": 1.5633848133848133e-05,
      "loss": 1.0322,
      "step": 4465
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.668066024780273,
      "learning_rate": 1.563063063063063e-05,
      "loss": 0.6348,
      "step": 4466
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.244242668151855,
      "learning_rate": 1.5627413127413126e-05,
      "loss": 0.9801,
      "step": 4467
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.637069702148438,
      "learning_rate": 1.5624195624195623e-05,
      "loss": 0.7833,
      "step": 4468
    },
    {
      "epoch": 0.96,
      "grad_norm": 11.884309768676758,
      "learning_rate": 1.5620978120978123e-05,
      "loss": 1.0532,
      "step": 4469
    },
    {
      "epoch": 0.96,
      "grad_norm": 11.806416511535645,
      "learning_rate": 1.561776061776062e-05,
      "loss": 0.9279,
      "step": 4470
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.393179893493652,
      "learning_rate": 1.5614543114543115e-05,
      "loss": 0.7328,
      "step": 4471
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.927849769592285,
      "learning_rate": 1.5611325611325612e-05,
      "loss": 1.0037,
      "step": 4472
    },
    {
      "epoch": 0.96,
      "grad_norm": 11.636940002441406,
      "learning_rate": 1.560810810810811e-05,
      "loss": 0.897,
      "step": 4473
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.853442192077637,
      "learning_rate": 1.5604890604890605e-05,
      "loss": 0.6167,
      "step": 4474
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.018816947937012,
      "learning_rate": 1.56016731016731e-05,
      "loss": 0.7164,
      "step": 4475
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.947822570800781,
      "learning_rate": 1.55984555984556e-05,
      "loss": 0.8285,
      "step": 4476
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.34401798248291,
      "learning_rate": 1.5595238095238098e-05,
      "loss": 0.6429,
      "step": 4477
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.73180103302002,
      "learning_rate": 1.5592020592020594e-05,
      "loss": 0.9026,
      "step": 4478
    },
    {
      "epoch": 0.96,
      "grad_norm": 12.97130012512207,
      "learning_rate": 1.558880308880309e-05,
      "loss": 1.0452,
      "step": 4479
    },
    {
      "epoch": 0.96,
      "grad_norm": 11.05053424835205,
      "learning_rate": 1.5585585585585584e-05,
      "loss": 0.7859,
      "step": 4480
    },
    {
      "epoch": 0.96,
      "grad_norm": 12.28261661529541,
      "learning_rate": 1.558236808236808e-05,
      "loss": 1.2038,
      "step": 4481
    },
    {
      "epoch": 0.96,
      "grad_norm": 12.472611427307129,
      "learning_rate": 1.557915057915058e-05,
      "loss": 0.9761,
      "step": 4482
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.448943138122559,
      "learning_rate": 1.5575933075933076e-05,
      "loss": 0.5725,
      "step": 4483
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.561295509338379,
      "learning_rate": 1.5572715572715573e-05,
      "loss": 1.2749,
      "step": 4484
    },
    {
      "epoch": 0.96,
      "grad_norm": 11.939794540405273,
      "learning_rate": 1.556949806949807e-05,
      "loss": 1.0024,
      "step": 4485
    },
    {
      "epoch": 0.96,
      "grad_norm": 11.647209167480469,
      "learning_rate": 1.5566280566280566e-05,
      "loss": 0.8354,
      "step": 4486
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.954986572265625,
      "learning_rate": 1.5563063063063062e-05,
      "loss": 0.8899,
      "step": 4487
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.880133628845215,
      "learning_rate": 1.555984555984556e-05,
      "loss": 0.8021,
      "step": 4488
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.213432312011719,
      "learning_rate": 1.555662805662806e-05,
      "loss": 0.9069,
      "step": 4489
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.460265159606934,
      "learning_rate": 1.5553410553410555e-05,
      "loss": 0.7001,
      "step": 4490
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.343094825744629,
      "learning_rate": 1.555019305019305e-05,
      "loss": 0.8451,
      "step": 4491
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.260222434997559,
      "learning_rate": 1.5546975546975548e-05,
      "loss": 0.7724,
      "step": 4492
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.498828887939453,
      "learning_rate": 1.5543758043758044e-05,
      "loss": 0.8021,
      "step": 4493
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.768784523010254,
      "learning_rate": 1.554054054054054e-05,
      "loss": 0.6932,
      "step": 4494
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.540255546569824,
      "learning_rate": 1.553732303732304e-05,
      "loss": 0.9034,
      "step": 4495
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.735116004943848,
      "learning_rate": 1.5534105534105537e-05,
      "loss": 1.2457,
      "step": 4496
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.784881591796875,
      "learning_rate": 1.553088803088803e-05,
      "loss": 0.862,
      "step": 4497
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.420746803283691,
      "learning_rate": 1.5527670527670526e-05,
      "loss": 0.8435,
      "step": 4498
    },
    {
      "epoch": 0.97,
      "grad_norm": 8.562751770019531,
      "learning_rate": 1.5524453024453023e-05,
      "loss": 0.765,
      "step": 4499
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.199134826660156,
      "learning_rate": 1.552123552123552e-05,
      "loss": 0.6758,
      "step": 4500
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.778719902038574,
      "learning_rate": 1.551801801801802e-05,
      "loss": 0.9184,
      "step": 4501
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.493520736694336,
      "learning_rate": 1.5514800514800516e-05,
      "loss": 0.7999,
      "step": 4502
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.136487007141113,
      "learning_rate": 1.5511583011583012e-05,
      "loss": 0.6154,
      "step": 4503
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.744824409484863,
      "learning_rate": 1.550836550836551e-05,
      "loss": 0.9781,
      "step": 4504
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.259568214416504,
      "learning_rate": 1.5505148005148005e-05,
      "loss": 1.1091,
      "step": 4505
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.241259574890137,
      "learning_rate": 1.55019305019305e-05,
      "loss": 0.7951,
      "step": 4506
    },
    {
      "epoch": 0.97,
      "grad_norm": 8.71556282043457,
      "learning_rate": 1.5498712998712998e-05,
      "loss": 0.7413,
      "step": 4507
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.826635360717773,
      "learning_rate": 1.5495495495495498e-05,
      "loss": 0.7705,
      "step": 4508
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.068681716918945,
      "learning_rate": 1.5492277992277994e-05,
      "loss": 0.7536,
      "step": 4509
    },
    {
      "epoch": 0.97,
      "grad_norm": 8.661327362060547,
      "learning_rate": 1.548906048906049e-05,
      "loss": 1.0599,
      "step": 4510
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.223919868469238,
      "learning_rate": 1.5485842985842987e-05,
      "loss": 0.973,
      "step": 4511
    },
    {
      "epoch": 0.97,
      "grad_norm": 12.476638793945312,
      "learning_rate": 1.5482625482625484e-05,
      "loss": 0.8515,
      "step": 4512
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.105192184448242,
      "learning_rate": 1.547940797940798e-05,
      "loss": 0.688,
      "step": 4513
    },
    {
      "epoch": 0.97,
      "grad_norm": 8.625908851623535,
      "learning_rate": 1.547619047619048e-05,
      "loss": 0.8073,
      "step": 4514
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.124497413635254,
      "learning_rate": 1.5472972972972973e-05,
      "loss": 0.923,
      "step": 4515
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.486957550048828,
      "learning_rate": 1.546975546975547e-05,
      "loss": 0.6623,
      "step": 4516
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.928980827331543,
      "learning_rate": 1.5466537966537966e-05,
      "loss": 0.8069,
      "step": 4517
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.287321090698242,
      "learning_rate": 1.5463320463320462e-05,
      "loss": 0.6306,
      "step": 4518
    },
    {
      "epoch": 0.97,
      "grad_norm": 8.596321105957031,
      "learning_rate": 1.546010296010296e-05,
      "loss": 0.7478,
      "step": 4519
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.938410758972168,
      "learning_rate": 1.5456885456885455e-05,
      "loss": 0.9626,
      "step": 4520
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.418139457702637,
      "learning_rate": 1.5453667953667955e-05,
      "loss": 1.0542,
      "step": 4521
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.791895866394043,
      "learning_rate": 1.545045045045045e-05,
      "loss": 0.7606,
      "step": 4522
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.670018196105957,
      "learning_rate": 1.5447232947232948e-05,
      "loss": 0.6595,
      "step": 4523
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.990328788757324,
      "learning_rate": 1.5444015444015444e-05,
      "loss": 1.167,
      "step": 4524
    },
    {
      "epoch": 0.97,
      "grad_norm": 12.006296157836914,
      "learning_rate": 1.544079794079794e-05,
      "loss": 0.7282,
      "step": 4525
    },
    {
      "epoch": 0.97,
      "grad_norm": 12.239548683166504,
      "learning_rate": 1.5437580437580437e-05,
      "loss": 1.0778,
      "step": 4526
    },
    {
      "epoch": 0.97,
      "grad_norm": 12.034429550170898,
      "learning_rate": 1.5434362934362937e-05,
      "loss": 0.699,
      "step": 4527
    },
    {
      "epoch": 0.97,
      "grad_norm": 12.444515228271484,
      "learning_rate": 1.5431145431145434e-05,
      "loss": 0.8986,
      "step": 4528
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.467052459716797,
      "learning_rate": 1.542792792792793e-05,
      "loss": 0.8518,
      "step": 4529
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.154678344726562,
      "learning_rate": 1.5424710424710426e-05,
      "loss": 0.7753,
      "step": 4530
    },
    {
      "epoch": 0.97,
      "grad_norm": 8.806854248046875,
      "learning_rate": 1.5421492921492923e-05,
      "loss": 0.639,
      "step": 4531
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.255912780761719,
      "learning_rate": 1.5418275418275416e-05,
      "loss": 0.7439,
      "step": 4532
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.081774711608887,
      "learning_rate": 1.5415057915057912e-05,
      "loss": 0.6568,
      "step": 4533
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.942432403564453,
      "learning_rate": 1.5411840411840412e-05,
      "loss": 0.8121,
      "step": 4534
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.621522903442383,
      "learning_rate": 1.540862290862291e-05,
      "loss": 0.9279,
      "step": 4535
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.845926284790039,
      "learning_rate": 1.5405405405405405e-05,
      "loss": 1.119,
      "step": 4536
    },
    {
      "epoch": 0.97,
      "grad_norm": 7.982652187347412,
      "learning_rate": 1.54021879021879e-05,
      "loss": 0.701,
      "step": 4537
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.743134498596191,
      "learning_rate": 1.5398970398970398e-05,
      "loss": 0.9087,
      "step": 4538
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.237825393676758,
      "learning_rate": 1.5395752895752895e-05,
      "loss": 0.9742,
      "step": 4539
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.347062110900879,
      "learning_rate": 1.5392535392535394e-05,
      "loss": 0.9089,
      "step": 4540
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.356759071350098,
      "learning_rate": 1.538931788931789e-05,
      "loss": 0.8672,
      "step": 4541
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.143314361572266,
      "learning_rate": 1.5386100386100387e-05,
      "loss": 0.7577,
      "step": 4542
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.088578224182129,
      "learning_rate": 1.5382882882882884e-05,
      "loss": 0.7225,
      "step": 4543
    },
    {
      "epoch": 0.97,
      "grad_norm": 7.77928352355957,
      "learning_rate": 1.537966537966538e-05,
      "loss": 0.5352,
      "step": 4544
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.212502479553223,
      "learning_rate": 1.5376447876447877e-05,
      "loss": 0.7227,
      "step": 4545
    },
    {
      "epoch": 0.98,
      "grad_norm": 8.694991111755371,
      "learning_rate": 1.5373230373230373e-05,
      "loss": 0.7844,
      "step": 4546
    },
    {
      "epoch": 0.98,
      "grad_norm": 8.7684326171875,
      "learning_rate": 1.5370012870012873e-05,
      "loss": 0.635,
      "step": 4547
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.282994270324707,
      "learning_rate": 1.536679536679537e-05,
      "loss": 0.8159,
      "step": 4548
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.624751091003418,
      "learning_rate": 1.5363577863577862e-05,
      "loss": 0.8702,
      "step": 4549
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.2596435546875,
      "learning_rate": 1.536036036036036e-05,
      "loss": 0.9271,
      "step": 4550
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.931050300598145,
      "learning_rate": 1.5357142857142855e-05,
      "loss": 0.7169,
      "step": 4551
    },
    {
      "epoch": 0.98,
      "grad_norm": 8.615495681762695,
      "learning_rate": 1.5353925353925352e-05,
      "loss": 0.6068,
      "step": 4552
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.506830215454102,
      "learning_rate": 1.535070785070785e-05,
      "loss": 1.0396,
      "step": 4553
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.727452278137207,
      "learning_rate": 1.5347490347490348e-05,
      "loss": 0.955,
      "step": 4554
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.30593490600586,
      "learning_rate": 1.5344272844272845e-05,
      "loss": 0.9756,
      "step": 4555
    },
    {
      "epoch": 0.98,
      "grad_norm": 8.952961921691895,
      "learning_rate": 1.534105534105534e-05,
      "loss": 0.8494,
      "step": 4556
    },
    {
      "epoch": 0.98,
      "grad_norm": 14.423856735229492,
      "learning_rate": 1.5337837837837837e-05,
      "loss": 1.0695,
      "step": 4557
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.011528968811035,
      "learning_rate": 1.5334620334620334e-05,
      "loss": 0.7458,
      "step": 4558
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.094159126281738,
      "learning_rate": 1.5331402831402834e-05,
      "loss": 0.8694,
      "step": 4559
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.472038269042969,
      "learning_rate": 1.532818532818533e-05,
      "loss": 0.6347,
      "step": 4560
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.977422714233398,
      "learning_rate": 1.5324967824967827e-05,
      "loss": 0.8483,
      "step": 4561
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.841103553771973,
      "learning_rate": 1.5321750321750323e-05,
      "loss": 0.6845,
      "step": 4562
    },
    {
      "epoch": 0.98,
      "grad_norm": 8.964155197143555,
      "learning_rate": 1.531853281853282e-05,
      "loss": 0.6791,
      "step": 4563
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.876921653747559,
      "learning_rate": 1.5315315315315316e-05,
      "loss": 0.5781,
      "step": 4564
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.305371284484863,
      "learning_rate": 1.5312097812097812e-05,
      "loss": 0.7606,
      "step": 4565
    },
    {
      "epoch": 0.98,
      "grad_norm": 12.159748077392578,
      "learning_rate": 1.5308880308880312e-05,
      "loss": 0.9505,
      "step": 4566
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.183246612548828,
      "learning_rate": 1.5305662805662805e-05,
      "loss": 0.911,
      "step": 4567
    },
    {
      "epoch": 0.98,
      "grad_norm": 8.487006187438965,
      "learning_rate": 1.5302445302445302e-05,
      "loss": 0.523,
      "step": 4568
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.753308296203613,
      "learning_rate": 1.5299227799227798e-05,
      "loss": 0.7827,
      "step": 4569
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.495948791503906,
      "learning_rate": 1.5296010296010295e-05,
      "loss": 0.8996,
      "step": 4570
    },
    {
      "epoch": 0.98,
      "grad_norm": 14.304458618164062,
      "learning_rate": 1.529279279279279e-05,
      "loss": 1.0244,
      "step": 4571
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.940980911254883,
      "learning_rate": 1.528957528957529e-05,
      "loss": 1.0376,
      "step": 4572
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.21595287322998,
      "learning_rate": 1.5286357786357787e-05,
      "loss": 0.7282,
      "step": 4573
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.40418815612793,
      "learning_rate": 1.5283140283140284e-05,
      "loss": 0.9635,
      "step": 4574
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.843350410461426,
      "learning_rate": 1.527992277992278e-05,
      "loss": 0.9112,
      "step": 4575
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.990267753601074,
      "learning_rate": 1.5276705276705277e-05,
      "loss": 0.7144,
      "step": 4576
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.120817184448242,
      "learning_rate": 1.5273487773487773e-05,
      "loss": 0.8286,
      "step": 4577
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.640682220458984,
      "learning_rate": 1.527027027027027e-05,
      "loss": 0.7567,
      "step": 4578
    },
    {
      "epoch": 0.98,
      "grad_norm": 8.99928092956543,
      "learning_rate": 1.526705276705277e-05,
      "loss": 0.6976,
      "step": 4579
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.802035331726074,
      "learning_rate": 1.5263835263835266e-05,
      "loss": 0.9902,
      "step": 4580
    },
    {
      "epoch": 0.98,
      "grad_norm": 8.157488822937012,
      "learning_rate": 1.5260617760617762e-05,
      "loss": 0.701,
      "step": 4581
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.756223678588867,
      "learning_rate": 1.5257400257400259e-05,
      "loss": 0.8469,
      "step": 4582
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.54246711730957,
      "learning_rate": 1.5254182754182755e-05,
      "loss": 0.8309,
      "step": 4583
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.370196342468262,
      "learning_rate": 1.525096525096525e-05,
      "loss": 0.7425,
      "step": 4584
    },
    {
      "epoch": 0.98,
      "grad_norm": 8.355864524841309,
      "learning_rate": 1.5247747747747747e-05,
      "loss": 0.7754,
      "step": 4585
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.053770065307617,
      "learning_rate": 1.5244530244530243e-05,
      "loss": 0.9238,
      "step": 4586
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.623785972595215,
      "learning_rate": 1.5241312741312741e-05,
      "loss": 0.9259,
      "step": 4587
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.378945350646973,
      "learning_rate": 1.5238095238095238e-05,
      "loss": 0.5103,
      "step": 4588
    },
    {
      "epoch": 0.98,
      "grad_norm": 15.359393119812012,
      "learning_rate": 1.5234877734877734e-05,
      "loss": 0.8156,
      "step": 4589
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.042508125305176,
      "learning_rate": 1.5231660231660232e-05,
      "loss": 0.8555,
      "step": 4590
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.041690826416016,
      "learning_rate": 1.5228442728442729e-05,
      "loss": 0.8616,
      "step": 4591
    },
    {
      "epoch": 0.98,
      "grad_norm": 8.38598346710205,
      "learning_rate": 1.5225225225225225e-05,
      "loss": 0.6441,
      "step": 4592
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.726493835449219,
      "learning_rate": 1.5222007722007723e-05,
      "loss": 0.8387,
      "step": 4593
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.388486862182617,
      "learning_rate": 1.521879021879022e-05,
      "loss": 0.8484,
      "step": 4594
    },
    {
      "epoch": 0.99,
      "grad_norm": 11.334742546081543,
      "learning_rate": 1.5215572715572716e-05,
      "loss": 0.7204,
      "step": 4595
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.227750778198242,
      "learning_rate": 1.5212355212355214e-05,
      "loss": 0.6786,
      "step": 4596
    },
    {
      "epoch": 0.99,
      "grad_norm": 7.517525672912598,
      "learning_rate": 1.520913770913771e-05,
      "loss": 0.8459,
      "step": 4597
    },
    {
      "epoch": 0.99,
      "grad_norm": 7.945837497711182,
      "learning_rate": 1.5205920205920207e-05,
      "loss": 0.6942,
      "step": 4598
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.193075180053711,
      "learning_rate": 1.5202702702702704e-05,
      "loss": 0.6605,
      "step": 4599
    },
    {
      "epoch": 0.99,
      "grad_norm": 11.549256324768066,
      "learning_rate": 1.5199485199485202e-05,
      "loss": 0.6625,
      "step": 4600
    },
    {
      "epoch": 0.99,
      "grad_norm": 7.735066890716553,
      "learning_rate": 1.5196267696267698e-05,
      "loss": 0.4461,
      "step": 4601
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.751119613647461,
      "learning_rate": 1.5193050193050191e-05,
      "loss": 0.842,
      "step": 4602
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.2233304977417,
      "learning_rate": 1.518983268983269e-05,
      "loss": 0.7075,
      "step": 4603
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.531262397766113,
      "learning_rate": 1.5186615186615186e-05,
      "loss": 0.7075,
      "step": 4604
    },
    {
      "epoch": 0.99,
      "grad_norm": 13.726881980895996,
      "learning_rate": 1.5183397683397682e-05,
      "loss": 0.8253,
      "step": 4605
    },
    {
      "epoch": 0.99,
      "grad_norm": 12.653054237365723,
      "learning_rate": 1.518018018018018e-05,
      "loss": 0.9776,
      "step": 4606
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.954476356506348,
      "learning_rate": 1.5176962676962677e-05,
      "loss": 1.1351,
      "step": 4607
    },
    {
      "epoch": 0.99,
      "grad_norm": 12.114684104919434,
      "learning_rate": 1.5173745173745173e-05,
      "loss": 0.8225,
      "step": 4608
    },
    {
      "epoch": 0.99,
      "grad_norm": 13.809980392456055,
      "learning_rate": 1.5170527670527672e-05,
      "loss": 0.7802,
      "step": 4609
    },
    {
      "epoch": 0.99,
      "grad_norm": 12.368420600891113,
      "learning_rate": 1.5167310167310168e-05,
      "loss": 0.762,
      "step": 4610
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.277315139770508,
      "learning_rate": 1.5164092664092664e-05,
      "loss": 0.598,
      "step": 4611
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.773187637329102,
      "learning_rate": 1.5160875160875161e-05,
      "loss": 0.6732,
      "step": 4612
    },
    {
      "epoch": 0.99,
      "grad_norm": 11.457979202270508,
      "learning_rate": 1.5157657657657659e-05,
      "loss": 0.9145,
      "step": 4613
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.715402603149414,
      "learning_rate": 1.5154440154440155e-05,
      "loss": 0.7335,
      "step": 4614
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.615744590759277,
      "learning_rate": 1.5151222651222652e-05,
      "loss": 0.9455,
      "step": 4615
    },
    {
      "epoch": 0.99,
      "grad_norm": 12.1511869430542,
      "learning_rate": 1.514800514800515e-05,
      "loss": 0.8851,
      "step": 4616
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.486997604370117,
      "learning_rate": 1.5144787644787647e-05,
      "loss": 0.9558,
      "step": 4617
    },
    {
      "epoch": 0.99,
      "grad_norm": 12.39685344696045,
      "learning_rate": 1.5141570141570143e-05,
      "loss": 1.0803,
      "step": 4618
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.909741401672363,
      "learning_rate": 1.5138352638352638e-05,
      "loss": 0.8078,
      "step": 4619
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.109848022460938,
      "learning_rate": 1.5135135135135134e-05,
      "loss": 0.7164,
      "step": 4620
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.317578315734863,
      "learning_rate": 1.513191763191763e-05,
      "loss": 0.9583,
      "step": 4621
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.97852897644043,
      "learning_rate": 1.5128700128700129e-05,
      "loss": 0.7145,
      "step": 4622
    },
    {
      "epoch": 0.99,
      "grad_norm": 7.967428207397461,
      "learning_rate": 1.5125482625482625e-05,
      "loss": 0.6236,
      "step": 4623
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.584554672241211,
      "learning_rate": 1.5122265122265122e-05,
      "loss": 0.6224,
      "step": 4624
    },
    {
      "epoch": 0.99,
      "grad_norm": 7.940814971923828,
      "learning_rate": 1.511904761904762e-05,
      "loss": 0.6348,
      "step": 4625
    },
    {
      "epoch": 0.99,
      "grad_norm": 11.008105278015137,
      "learning_rate": 1.5115830115830116e-05,
      "loss": 0.7533,
      "step": 4626
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.322983741760254,
      "learning_rate": 1.5112612612612613e-05,
      "loss": 0.9026,
      "step": 4627
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.198866844177246,
      "learning_rate": 1.510939510939511e-05,
      "loss": 0.7937,
      "step": 4628
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.278392791748047,
      "learning_rate": 1.5106177606177607e-05,
      "loss": 0.6902,
      "step": 4629
    },
    {
      "epoch": 0.99,
      "grad_norm": 11.884210586547852,
      "learning_rate": 1.5102960102960104e-05,
      "loss": 0.9897,
      "step": 4630
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.875935554504395,
      "learning_rate": 1.50997425997426e-05,
      "loss": 0.667,
      "step": 4631
    },
    {
      "epoch": 0.99,
      "grad_norm": 11.710469245910645,
      "learning_rate": 1.5096525096525098e-05,
      "loss": 0.6592,
      "step": 4632
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.962102890014648,
      "learning_rate": 1.5093307593307595e-05,
      "loss": 0.9204,
      "step": 4633
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.535521507263184,
      "learning_rate": 1.5090090090090091e-05,
      "loss": 0.9106,
      "step": 4634
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.215255737304688,
      "learning_rate": 1.508687258687259e-05,
      "loss": 0.9284,
      "step": 4635
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.304078102111816,
      "learning_rate": 1.5083655083655082e-05,
      "loss": 0.6622,
      "step": 4636
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.383277893066406,
      "learning_rate": 1.5080437580437579e-05,
      "loss": 0.6322,
      "step": 4637
    },
    {
      "epoch": 0.99,
      "grad_norm": 9.367839813232422,
      "learning_rate": 1.5077220077220077e-05,
      "loss": 0.7365,
      "step": 4638
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.6441650390625,
      "learning_rate": 1.5074002574002574e-05,
      "loss": 0.7143,
      "step": 4639
    },
    {
      "epoch": 1.0,
      "grad_norm": 11.604363441467285,
      "learning_rate": 1.507078507078507e-05,
      "loss": 0.9441,
      "step": 4640
    },
    {
      "epoch": 1.0,
      "grad_norm": 13.644936561584473,
      "learning_rate": 1.5067567567567566e-05,
      "loss": 0.8305,
      "step": 4641
    },
    {
      "epoch": 1.0,
      "grad_norm": 11.659443855285645,
      "learning_rate": 1.5064350064350065e-05,
      "loss": 0.928,
      "step": 4642
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.539076805114746,
      "learning_rate": 1.5061132561132561e-05,
      "loss": 0.9348,
      "step": 4643
    },
    {
      "epoch": 1.0,
      "grad_norm": 11.907176971435547,
      "learning_rate": 1.5057915057915057e-05,
      "loss": 1.1072,
      "step": 4644
    },
    {
      "epoch": 1.0,
      "grad_norm": 11.226696968078613,
      "learning_rate": 1.5054697554697556e-05,
      "loss": 0.7882,
      "step": 4645
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.8637113571167,
      "learning_rate": 1.5051480051480052e-05,
      "loss": 0.9259,
      "step": 4646
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.976282119750977,
      "learning_rate": 1.5048262548262549e-05,
      "loss": 0.9981,
      "step": 4647
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.535440444946289,
      "learning_rate": 1.5045045045045047e-05,
      "loss": 0.6162,
      "step": 4648
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.12368392944336,
      "learning_rate": 1.5041827541827543e-05,
      "loss": 0.6249,
      "step": 4649
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.82941722869873,
      "learning_rate": 1.503861003861004e-05,
      "loss": 0.7033,
      "step": 4650
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.263267517089844,
      "learning_rate": 1.5035392535392538e-05,
      "loss": 0.695,
      "step": 4651
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.784920692443848,
      "learning_rate": 1.5032175032175034e-05,
      "loss": 0.8521,
      "step": 4652
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.96573543548584,
      "learning_rate": 1.502895752895753e-05,
      "loss": 0.8361,
      "step": 4653
    },
    {
      "epoch": 1.0,
      "grad_norm": 12.743813514709473,
      "learning_rate": 1.5025740025740025e-05,
      "loss": 0.9845,
      "step": 4654
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.883023262023926,
      "learning_rate": 1.5022522522522522e-05,
      "loss": 0.8067,
      "step": 4655
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.079431533813477,
      "learning_rate": 1.5019305019305018e-05,
      "loss": 0.8842,
      "step": 4656
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.548842430114746,
      "learning_rate": 1.5016087516087515e-05,
      "loss": 0.8682,
      "step": 4657
    },
    {
      "epoch": 1.0,
      "grad_norm": 11.690617561340332,
      "learning_rate": 1.5012870012870013e-05,
      "loss": 0.8522,
      "step": 4658
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.459872245788574,
      "learning_rate": 1.500965250965251e-05,
      "loss": 0.6153,
      "step": 4659
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.57301139831543,
      "learning_rate": 1.5006435006435006e-05,
      "loss": 0.7509,
      "step": 4660
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.227286338806152,
      "learning_rate": 1.5003217503217504e-05,
      "loss": 0.7588,
      "step": 4661
    },
    {
      "epoch": 1.0,
      "grad_norm": 44.546260833740234,
      "learning_rate": 1.5e-05,
      "loss": 1.0859,
      "step": 4662
    },
    {
      "epoch": 1.0,
      "grad_norm": 12.150619506835938,
      "learning_rate": 1.4996782496782497e-05,
      "loss": 0.8902,
      "step": 4663
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.00283432006836,
      "learning_rate": 1.4993564993564995e-05,
      "loss": 0.7513,
      "step": 4664
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.41715145111084,
      "learning_rate": 1.4990347490347491e-05,
      "loss": 0.6658,
      "step": 4665
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.713772773742676,
      "learning_rate": 1.4987129987129988e-05,
      "loss": 0.7865,
      "step": 4666
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.669293403625488,
      "learning_rate": 1.4983912483912484e-05,
      "loss": 0.7658,
      "step": 4667
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.673441886901855,
      "learning_rate": 1.498069498069498e-05,
      "loss": 0.4892,
      "step": 4668
    },
    {
      "epoch": 1.0,
      "grad_norm": 7.516715049743652,
      "learning_rate": 1.4977477477477477e-05,
      "loss": 0.5594,
      "step": 4669
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.426481246948242,
      "learning_rate": 1.4974259974259974e-05,
      "loss": 0.6373,
      "step": 4670
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.879096984863281,
      "learning_rate": 1.4971042471042472e-05,
      "loss": 0.7695,
      "step": 4671
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.610248565673828,
      "learning_rate": 1.4967824967824968e-05,
      "loss": 0.5585,
      "step": 4672
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.96116828918457,
      "learning_rate": 1.4964607464607465e-05,
      "loss": 0.8669,
      "step": 4673
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.072827339172363,
      "learning_rate": 1.4961389961389963e-05,
      "loss": 0.669,
      "step": 4674
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.406583786010742,
      "learning_rate": 1.495817245817246e-05,
      "loss": 0.6842,
      "step": 4675
    },
    {
      "epoch": 1.0,
      "grad_norm": 7.710730075836182,
      "learning_rate": 1.4954954954954954e-05,
      "loss": 0.5073,
      "step": 4676
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.696039199829102,
      "learning_rate": 1.4951737451737452e-05,
      "loss": 0.7927,
      "step": 4677
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.69656753540039,
      "learning_rate": 1.4948519948519949e-05,
      "loss": 0.4644,
      "step": 4678
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.784064292907715,
      "learning_rate": 1.4945302445302445e-05,
      "loss": 0.5331,
      "step": 4679
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.846460342407227,
      "learning_rate": 1.4942084942084943e-05,
      "loss": 0.5817,
      "step": 4680
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.597662925720215,
      "learning_rate": 1.493886743886744e-05,
      "loss": 0.6698,
      "step": 4681
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.77672004699707,
      "learning_rate": 1.4935649935649936e-05,
      "loss": 0.5471,
      "step": 4682
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.320650100708008,
      "learning_rate": 1.4932432432432434e-05,
      "loss": 0.8454,
      "step": 4683
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.509571075439453,
      "learning_rate": 1.4929214929214929e-05,
      "loss": 0.6314,
      "step": 4684
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.564868927001953,
      "learning_rate": 1.4925997425997426e-05,
      "loss": 0.8288,
      "step": 4685
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.137802124023438,
      "learning_rate": 1.4922779922779922e-05,
      "loss": 0.6517,
      "step": 4686
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.031932830810547,
      "learning_rate": 1.491956241956242e-05,
      "loss": 0.8522,
      "step": 4687
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.329914093017578,
      "learning_rate": 1.4916344916344917e-05,
      "loss": 0.6802,
      "step": 4688
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.425049781799316,
      "learning_rate": 1.4913127413127413e-05,
      "loss": 0.7708,
      "step": 4689
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.376158714294434,
      "learning_rate": 1.4909909909909911e-05,
      "loss": 0.764,
      "step": 4690
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.419498443603516,
      "learning_rate": 1.4906692406692408e-05,
      "loss": 0.679,
      "step": 4691
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.26028060913086,
      "learning_rate": 1.4903474903474904e-05,
      "loss": 0.6926,
      "step": 4692
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.056951522827148,
      "learning_rate": 1.49002574002574e-05,
      "loss": 0.5787,
      "step": 4693
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.174686431884766,
      "learning_rate": 1.4897039897039897e-05,
      "loss": 0.6445,
      "step": 4694
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.681024551391602,
      "learning_rate": 1.4893822393822393e-05,
      "loss": 0.6868,
      "step": 4695
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.869449615478516,
      "learning_rate": 1.4890604890604892e-05,
      "loss": 0.6217,
      "step": 4696
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.056333541870117,
      "learning_rate": 1.4887387387387388e-05,
      "loss": 1.0372,
      "step": 4697
    },
    {
      "epoch": 1.01,
      "grad_norm": 7.7992472648620605,
      "learning_rate": 1.4884169884169884e-05,
      "loss": 0.5748,
      "step": 4698
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.755010604858398,
      "learning_rate": 1.4880952380952381e-05,
      "loss": 1.0382,
      "step": 4699
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.939626693725586,
      "learning_rate": 1.4877734877734879e-05,
      "loss": 0.7599,
      "step": 4700
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.692713737487793,
      "learning_rate": 1.4874517374517376e-05,
      "loss": 0.7956,
      "step": 4701
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.02245044708252,
      "learning_rate": 1.487129987129987e-05,
      "loss": 0.6034,
      "step": 4702
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.141988754272461,
      "learning_rate": 1.4868082368082368e-05,
      "loss": 0.7196,
      "step": 4703
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.422647476196289,
      "learning_rate": 1.4864864864864865e-05,
      "loss": 0.6632,
      "step": 4704
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.943537712097168,
      "learning_rate": 1.4861647361647361e-05,
      "loss": 0.8637,
      "step": 4705
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.78662395477295,
      "learning_rate": 1.485842985842986e-05,
      "loss": 0.7059,
      "step": 4706
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.6400728225708,
      "learning_rate": 1.4855212355212356e-05,
      "loss": 0.7126,
      "step": 4707
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.079297065734863,
      "learning_rate": 1.4851994851994852e-05,
      "loss": 0.6221,
      "step": 4708
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.663274765014648,
      "learning_rate": 1.484877734877735e-05,
      "loss": 0.6629,
      "step": 4709
    },
    {
      "epoch": 1.01,
      "grad_norm": 12.490964889526367,
      "learning_rate": 1.4845559845559845e-05,
      "loss": 0.6925,
      "step": 4710
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.83564281463623,
      "learning_rate": 1.4842342342342342e-05,
      "loss": 0.4396,
      "step": 4711
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.885788917541504,
      "learning_rate": 1.483912483912484e-05,
      "loss": 0.688,
      "step": 4712
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.71988296508789,
      "learning_rate": 1.4835907335907336e-05,
      "loss": 0.6777,
      "step": 4713
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.91953182220459,
      "learning_rate": 1.4832689832689833e-05,
      "loss": 0.5932,
      "step": 4714
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.220040321350098,
      "learning_rate": 1.482947232947233e-05,
      "loss": 0.8384,
      "step": 4715
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.977184295654297,
      "learning_rate": 1.4826254826254827e-05,
      "loss": 0.6391,
      "step": 4716
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.021781921386719,
      "learning_rate": 1.4823037323037324e-05,
      "loss": 0.4476,
      "step": 4717
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.723058700561523,
      "learning_rate": 1.481981981981982e-05,
      "loss": 0.58,
      "step": 4718
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.991196632385254,
      "learning_rate": 1.4816602316602317e-05,
      "loss": 0.7221,
      "step": 4719
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.717272758483887,
      "learning_rate": 1.4813384813384813e-05,
      "loss": 0.8192,
      "step": 4720
    },
    {
      "epoch": 1.01,
      "grad_norm": 12.475905418395996,
      "learning_rate": 1.481016731016731e-05,
      "loss": 0.8706,
      "step": 4721
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.481074333190918,
      "learning_rate": 1.4806949806949808e-05,
      "loss": 0.6374,
      "step": 4722
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.803999900817871,
      "learning_rate": 1.4803732303732304e-05,
      "loss": 0.5598,
      "step": 4723
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.367037773132324,
      "learning_rate": 1.48005148005148e-05,
      "loss": 0.6529,
      "step": 4724
    },
    {
      "epoch": 1.01,
      "grad_norm": 13.39099407196045,
      "learning_rate": 1.4797297297297299e-05,
      "loss": 0.7756,
      "step": 4725
    },
    {
      "epoch": 1.01,
      "grad_norm": 7.755545616149902,
      "learning_rate": 1.4794079794079795e-05,
      "loss": 0.4312,
      "step": 4726
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.878161430358887,
      "learning_rate": 1.4790862290862292e-05,
      "loss": 0.6487,
      "step": 4727
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.279749870300293,
      "learning_rate": 1.4787644787644786e-05,
      "loss": 0.6693,
      "step": 4728
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.549921035766602,
      "learning_rate": 1.4784427284427285e-05,
      "loss": 0.6297,
      "step": 4729
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.152261734008789,
      "learning_rate": 1.4781209781209781e-05,
      "loss": 0.5173,
      "step": 4730
    },
    {
      "epoch": 1.01,
      "grad_norm": 12.343775749206543,
      "learning_rate": 1.4777992277992278e-05,
      "loss": 0.8931,
      "step": 4731
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.755380630493164,
      "learning_rate": 1.4774774774774776e-05,
      "loss": 0.6555,
      "step": 4732
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.645594596862793,
      "learning_rate": 1.4771557271557272e-05,
      "loss": 0.9466,
      "step": 4733
    },
    {
      "epoch": 1.02,
      "grad_norm": 14.0270357131958,
      "learning_rate": 1.4768339768339769e-05,
      "loss": 0.972,
      "step": 4734
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.457874298095703,
      "learning_rate": 1.4765122265122267e-05,
      "loss": 0.6298,
      "step": 4735
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.7749662399292,
      "learning_rate": 1.4761904761904761e-05,
      "loss": 0.6888,
      "step": 4736
    },
    {
      "epoch": 1.02,
      "grad_norm": 11.944783210754395,
      "learning_rate": 1.4758687258687258e-05,
      "loss": 0.7907,
      "step": 4737
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.144564628601074,
      "learning_rate": 1.4755469755469756e-05,
      "loss": 0.7083,
      "step": 4738
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.822521209716797,
      "learning_rate": 1.4752252252252253e-05,
      "loss": 0.8599,
      "step": 4739
    },
    {
      "epoch": 1.02,
      "grad_norm": 11.439375877380371,
      "learning_rate": 1.4749034749034749e-05,
      "loss": 0.889,
      "step": 4740
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.446714401245117,
      "learning_rate": 1.4745817245817247e-05,
      "loss": 0.6787,
      "step": 4741
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.091769218444824,
      "learning_rate": 1.4742599742599744e-05,
      "loss": 0.7147,
      "step": 4742
    },
    {
      "epoch": 1.02,
      "grad_norm": 11.07342529296875,
      "learning_rate": 1.473938223938224e-05,
      "loss": 0.9358,
      "step": 4743
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.088436126708984,
      "learning_rate": 1.4736164736164736e-05,
      "loss": 0.7626,
      "step": 4744
    },
    {
      "epoch": 1.02,
      "grad_norm": 13.482744216918945,
      "learning_rate": 1.4732947232947233e-05,
      "loss": 0.8244,
      "step": 4745
    },
    {
      "epoch": 1.02,
      "grad_norm": 15.479310989379883,
      "learning_rate": 1.472972972972973e-05,
      "loss": 0.9731,
      "step": 4746
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.742753982543945,
      "learning_rate": 1.4726512226512226e-05,
      "loss": 0.6356,
      "step": 4747
    },
    {
      "epoch": 1.02,
      "grad_norm": 15.460336685180664,
      "learning_rate": 1.4723294723294724e-05,
      "loss": 0.8314,
      "step": 4748
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.093745231628418,
      "learning_rate": 1.472007722007722e-05,
      "loss": 0.7264,
      "step": 4749
    },
    {
      "epoch": 1.02,
      "grad_norm": 7.569711685180664,
      "learning_rate": 1.4716859716859717e-05,
      "loss": 0.5212,
      "step": 4750
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.442863464355469,
      "learning_rate": 1.4713642213642215e-05,
      "loss": 0.6705,
      "step": 4751
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.097798347473145,
      "learning_rate": 1.4710424710424711e-05,
      "loss": 0.7728,
      "step": 4752
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.006537437438965,
      "learning_rate": 1.4707207207207208e-05,
      "loss": 0.5792,
      "step": 4753
    },
    {
      "epoch": 1.02,
      "grad_norm": 11.603912353515625,
      "learning_rate": 1.4703989703989704e-05,
      "loss": 0.5562,
      "step": 4754
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.644207000732422,
      "learning_rate": 1.4700772200772201e-05,
      "loss": 0.7565,
      "step": 4755
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.714015007019043,
      "learning_rate": 1.4697554697554697e-05,
      "loss": 0.5771,
      "step": 4756
    },
    {
      "epoch": 1.02,
      "grad_norm": 11.250645637512207,
      "learning_rate": 1.4694337194337194e-05,
      "loss": 0.7233,
      "step": 4757
    },
    {
      "epoch": 1.02,
      "grad_norm": 7.766149044036865,
      "learning_rate": 1.4691119691119692e-05,
      "loss": 0.5308,
      "step": 4758
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.671820640563965,
      "learning_rate": 1.4687902187902188e-05,
      "loss": 0.7822,
      "step": 4759
    },
    {
      "epoch": 1.02,
      "grad_norm": 11.054652214050293,
      "learning_rate": 1.4684684684684685e-05,
      "loss": 0.7706,
      "step": 4760
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.483529090881348,
      "learning_rate": 1.4681467181467183e-05,
      "loss": 0.7842,
      "step": 4761
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.550127983093262,
      "learning_rate": 1.467824967824968e-05,
      "loss": 0.506,
      "step": 4762
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.417140007019043,
      "learning_rate": 1.4675032175032174e-05,
      "loss": 0.5491,
      "step": 4763
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.180427551269531,
      "learning_rate": 1.4671814671814672e-05,
      "loss": 0.6743,
      "step": 4764
    },
    {
      "epoch": 1.02,
      "grad_norm": 5.907822608947754,
      "learning_rate": 1.4668597168597169e-05,
      "loss": 0.4265,
      "step": 4765
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.555276870727539,
      "learning_rate": 1.4665379665379665e-05,
      "loss": 0.8798,
      "step": 4766
    },
    {
      "epoch": 1.02,
      "grad_norm": 11.507061004638672,
      "learning_rate": 1.4662162162162163e-05,
      "loss": 0.881,
      "step": 4767
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.657910346984863,
      "learning_rate": 1.465894465894466e-05,
      "loss": 0.647,
      "step": 4768
    },
    {
      "epoch": 1.02,
      "grad_norm": 11.044404983520508,
      "learning_rate": 1.4655727155727156e-05,
      "loss": 0.8146,
      "step": 4769
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.847625732421875,
      "learning_rate": 1.4652509652509654e-05,
      "loss": 0.6876,
      "step": 4770
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.593716621398926,
      "learning_rate": 1.464929214929215e-05,
      "loss": 0.6311,
      "step": 4771
    },
    {
      "epoch": 1.02,
      "grad_norm": 11.618921279907227,
      "learning_rate": 1.4646074646074646e-05,
      "loss": 0.803,
      "step": 4772
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.86817455291748,
      "learning_rate": 1.4642857142857142e-05,
      "loss": 0.817,
      "step": 4773
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.9232177734375,
      "learning_rate": 1.463963963963964e-05,
      "loss": 0.7881,
      "step": 4774
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.942276954650879,
      "learning_rate": 1.4636422136422137e-05,
      "loss": 0.5732,
      "step": 4775
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.761350631713867,
      "learning_rate": 1.4633204633204633e-05,
      "loss": 0.6273,
      "step": 4776
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.849753379821777,
      "learning_rate": 1.4629987129987131e-05,
      "loss": 0.6841,
      "step": 4777
    },
    {
      "epoch": 1.02,
      "grad_norm": 10.160075187683105,
      "learning_rate": 1.4626769626769628e-05,
      "loss": 0.6175,
      "step": 4778
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.668465614318848,
      "learning_rate": 1.4623552123552124e-05,
      "loss": 0.4141,
      "step": 4779
    },
    {
      "epoch": 1.03,
      "grad_norm": 11.271156311035156,
      "learning_rate": 1.462033462033462e-05,
      "loss": 0.7747,
      "step": 4780
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.921581268310547,
      "learning_rate": 1.4617117117117117e-05,
      "loss": 0.7536,
      "step": 4781
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.32073974609375,
      "learning_rate": 1.4613899613899614e-05,
      "loss": 0.7374,
      "step": 4782
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.984441757202148,
      "learning_rate": 1.4610682110682112e-05,
      "loss": 0.751,
      "step": 4783
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.801342010498047,
      "learning_rate": 1.4607464607464608e-05,
      "loss": 0.7623,
      "step": 4784
    },
    {
      "epoch": 1.03,
      "grad_norm": 11.57227611541748,
      "learning_rate": 1.4604247104247105e-05,
      "loss": 0.7526,
      "step": 4785
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.27971076965332,
      "learning_rate": 1.4601029601029601e-05,
      "loss": 0.7597,
      "step": 4786
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.26451587677002,
      "learning_rate": 1.45978120978121e-05,
      "loss": 0.5496,
      "step": 4787
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.62536334991455,
      "learning_rate": 1.4594594594594596e-05,
      "loss": 0.6835,
      "step": 4788
    },
    {
      "epoch": 1.03,
      "grad_norm": 12.164536476135254,
      "learning_rate": 1.459137709137709e-05,
      "loss": 0.9419,
      "step": 4789
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.48943042755127,
      "learning_rate": 1.4588159588159589e-05,
      "loss": 0.7651,
      "step": 4790
    },
    {
      "epoch": 1.03,
      "grad_norm": 11.484209060668945,
      "learning_rate": 1.4584942084942085e-05,
      "loss": 0.7237,
      "step": 4791
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.925271987915039,
      "learning_rate": 1.4581724581724581e-05,
      "loss": 0.6943,
      "step": 4792
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.686643600463867,
      "learning_rate": 1.457850707850708e-05,
      "loss": 0.5842,
      "step": 4793
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.495610237121582,
      "learning_rate": 1.4575289575289576e-05,
      "loss": 0.5192,
      "step": 4794
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.90485668182373,
      "learning_rate": 1.4572072072072072e-05,
      "loss": 0.8013,
      "step": 4795
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.406343460083008,
      "learning_rate": 1.456885456885457e-05,
      "loss": 0.6838,
      "step": 4796
    },
    {
      "epoch": 1.03,
      "grad_norm": 11.427634239196777,
      "learning_rate": 1.4565637065637065e-05,
      "loss": 0.7552,
      "step": 4797
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.287079811096191,
      "learning_rate": 1.4562419562419562e-05,
      "loss": 0.5529,
      "step": 4798
    },
    {
      "epoch": 1.03,
      "grad_norm": 7.23927640914917,
      "learning_rate": 1.455920205920206e-05,
      "loss": 0.5609,
      "step": 4799
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.211015701293945,
      "learning_rate": 1.4555984555984556e-05,
      "loss": 0.8323,
      "step": 4800
    },
    {
      "epoch": 1.03,
      "grad_norm": 11.188618659973145,
      "learning_rate": 1.4552767052767053e-05,
      "loss": 0.8159,
      "step": 4801
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.274483680725098,
      "learning_rate": 1.454954954954955e-05,
      "loss": 0.5222,
      "step": 4802
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.914271354675293,
      "learning_rate": 1.4546332046332047e-05,
      "loss": 0.782,
      "step": 4803
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.479706764221191,
      "learning_rate": 1.4543114543114544e-05,
      "loss": 0.5913,
      "step": 4804
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.761517524719238,
      "learning_rate": 1.453989703989704e-05,
      "loss": 0.5429,
      "step": 4805
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.49422550201416,
      "learning_rate": 1.4536679536679537e-05,
      "loss": 0.5359,
      "step": 4806
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.10258960723877,
      "learning_rate": 1.4533462033462033e-05,
      "loss": 0.6842,
      "step": 4807
    },
    {
      "epoch": 1.03,
      "grad_norm": 11.570988655090332,
      "learning_rate": 1.453024453024453e-05,
      "loss": 0.8783,
      "step": 4808
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.506147384643555,
      "learning_rate": 1.4527027027027028e-05,
      "loss": 0.5843,
      "step": 4809
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.967236518859863,
      "learning_rate": 1.4523809523809524e-05,
      "loss": 0.7038,
      "step": 4810
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.368827819824219,
      "learning_rate": 1.452059202059202e-05,
      "loss": 0.7282,
      "step": 4811
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.924996376037598,
      "learning_rate": 1.4517374517374519e-05,
      "loss": 0.7191,
      "step": 4812
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.863205909729004,
      "learning_rate": 1.4514157014157015e-05,
      "loss": 0.8056,
      "step": 4813
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.358541488647461,
      "learning_rate": 1.4510939510939512e-05,
      "loss": 0.5713,
      "step": 4814
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.27963924407959,
      "learning_rate": 1.4507722007722007e-05,
      "loss": 0.6872,
      "step": 4815
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.438884735107422,
      "learning_rate": 1.4504504504504505e-05,
      "loss": 0.5988,
      "step": 4816
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.666583061218262,
      "learning_rate": 1.4501287001287001e-05,
      "loss": 0.5097,
      "step": 4817
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.49399471282959,
      "learning_rate": 1.4498069498069498e-05,
      "loss": 0.588,
      "step": 4818
    },
    {
      "epoch": 1.03,
      "grad_norm": 12.202812194824219,
      "learning_rate": 1.4494851994851996e-05,
      "loss": 0.8633,
      "step": 4819
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.865045547485352,
      "learning_rate": 1.4491634491634492e-05,
      "loss": 0.6285,
      "step": 4820
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.81965446472168,
      "learning_rate": 1.4488416988416989e-05,
      "loss": 0.4827,
      "step": 4821
    },
    {
      "epoch": 1.03,
      "grad_norm": 11.9266996383667,
      "learning_rate": 1.4485199485199487e-05,
      "loss": 0.7316,
      "step": 4822
    },
    {
      "epoch": 1.03,
      "grad_norm": 11.357596397399902,
      "learning_rate": 1.4481981981981982e-05,
      "loss": 0.7119,
      "step": 4823
    },
    {
      "epoch": 1.03,
      "grad_norm": 11.792227745056152,
      "learning_rate": 1.4478764478764478e-05,
      "loss": 0.7085,
      "step": 4824
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.988113403320312,
      "learning_rate": 1.4475546975546976e-05,
      "loss": 0.732,
      "step": 4825
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.054940223693848,
      "learning_rate": 1.4472329472329473e-05,
      "loss": 0.498,
      "step": 4826
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.57999038696289,
      "learning_rate": 1.4469111969111969e-05,
      "loss": 0.7135,
      "step": 4827
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.591034889221191,
      "learning_rate": 1.4465894465894467e-05,
      "loss": 0.5868,
      "step": 4828
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.85687255859375,
      "learning_rate": 1.4462676962676964e-05,
      "loss": 0.7412,
      "step": 4829
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.917573928833008,
      "learning_rate": 1.445945945945946e-05,
      "loss": 0.7101,
      "step": 4830
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.302854537963867,
      "learning_rate": 1.4456241956241957e-05,
      "loss": 0.6245,
      "step": 4831
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.662322044372559,
      "learning_rate": 1.4453024453024453e-05,
      "loss": 0.8322,
      "step": 4832
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.081385612487793,
      "learning_rate": 1.444980694980695e-05,
      "loss": 0.6758,
      "step": 4833
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.841209411621094,
      "learning_rate": 1.4446589446589446e-05,
      "loss": 0.6086,
      "step": 4834
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.51294231414795,
      "learning_rate": 1.4443371943371944e-05,
      "loss": 0.7636,
      "step": 4835
    },
    {
      "epoch": 1.04,
      "grad_norm": 11.389134407043457,
      "learning_rate": 1.444015444015444e-05,
      "loss": 0.7014,
      "step": 4836
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.109190940856934,
      "learning_rate": 1.4436936936936937e-05,
      "loss": 0.8592,
      "step": 4837
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.503807067871094,
      "learning_rate": 1.4433719433719435e-05,
      "loss": 0.691,
      "step": 4838
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.43567943572998,
      "learning_rate": 1.4430501930501932e-05,
      "loss": 0.7147,
      "step": 4839
    },
    {
      "epoch": 1.04,
      "grad_norm": 11.02943229675293,
      "learning_rate": 1.4427284427284428e-05,
      "loss": 0.9466,
      "step": 4840
    },
    {
      "epoch": 1.04,
      "grad_norm": 12.006986618041992,
      "learning_rate": 1.4424066924066924e-05,
      "loss": 0.8969,
      "step": 4841
    },
    {
      "epoch": 1.04,
      "grad_norm": 11.019563674926758,
      "learning_rate": 1.4420849420849421e-05,
      "loss": 0.8868,
      "step": 4842
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.753246307373047,
      "learning_rate": 1.4417631917631917e-05,
      "loss": 0.6188,
      "step": 4843
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.308995246887207,
      "learning_rate": 1.4414414414414414e-05,
      "loss": 0.6611,
      "step": 4844
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.791749954223633,
      "learning_rate": 1.4411196911196912e-05,
      "loss": 0.8207,
      "step": 4845
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.694498062133789,
      "learning_rate": 1.4407979407979408e-05,
      "loss": 0.7726,
      "step": 4846
    },
    {
      "epoch": 1.04,
      "grad_norm": 7.392005920410156,
      "learning_rate": 1.4404761904761905e-05,
      "loss": 0.4099,
      "step": 4847
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.844500541687012,
      "learning_rate": 1.4401544401544403e-05,
      "loss": 0.6297,
      "step": 4848
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.209831237792969,
      "learning_rate": 1.43983268983269e-05,
      "loss": 0.8249,
      "step": 4849
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.137030601501465,
      "learning_rate": 1.4395109395109394e-05,
      "loss": 0.6912,
      "step": 4850
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.180392265319824,
      "learning_rate": 1.4391891891891892e-05,
      "loss": 0.766,
      "step": 4851
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.757245063781738,
      "learning_rate": 1.4388674388674389e-05,
      "loss": 0.7356,
      "step": 4852
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.564484596252441,
      "learning_rate": 1.4385456885456885e-05,
      "loss": 0.6893,
      "step": 4853
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.487906455993652,
      "learning_rate": 1.4382239382239383e-05,
      "loss": 0.684,
      "step": 4854
    },
    {
      "epoch": 1.04,
      "grad_norm": 8.565838813781738,
      "learning_rate": 1.437902187902188e-05,
      "loss": 0.5156,
      "step": 4855
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.72862720489502,
      "learning_rate": 1.4375804375804376e-05,
      "loss": 0.7854,
      "step": 4856
    },
    {
      "epoch": 1.04,
      "grad_norm": 11.907210350036621,
      "learning_rate": 1.4372586872586874e-05,
      "loss": 0.6466,
      "step": 4857
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.57166862487793,
      "learning_rate": 1.436936936936937e-05,
      "loss": 0.8874,
      "step": 4858
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.6239013671875,
      "learning_rate": 1.4366151866151866e-05,
      "loss": 0.8036,
      "step": 4859
    },
    {
      "epoch": 1.04,
      "grad_norm": 13.00149917602539,
      "learning_rate": 1.4362934362934362e-05,
      "loss": 0.8796,
      "step": 4860
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.657496452331543,
      "learning_rate": 1.435971685971686e-05,
      "loss": 0.7394,
      "step": 4861
    },
    {
      "epoch": 1.04,
      "grad_norm": 12.726150512695312,
      "learning_rate": 1.4356499356499357e-05,
      "loss": 0.5584,
      "step": 4862
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.900553703308105,
      "learning_rate": 1.4353281853281853e-05,
      "loss": 0.8077,
      "step": 4863
    },
    {
      "epoch": 1.04,
      "grad_norm": 11.09499740600586,
      "learning_rate": 1.4350064350064351e-05,
      "loss": 0.9281,
      "step": 4864
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.387673377990723,
      "learning_rate": 1.4346846846846848e-05,
      "loss": 0.6825,
      "step": 4865
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.602574348449707,
      "learning_rate": 1.4343629343629344e-05,
      "loss": 0.4486,
      "step": 4866
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.100741386413574,
      "learning_rate": 1.434041184041184e-05,
      "loss": 0.4893,
      "step": 4867
    },
    {
      "epoch": 1.04,
      "grad_norm": 15.443438529968262,
      "learning_rate": 1.4337194337194337e-05,
      "loss": 0.9782,
      "step": 4868
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.495681762695312,
      "learning_rate": 1.4333976833976834e-05,
      "loss": 0.7021,
      "step": 4869
    },
    {
      "epoch": 1.04,
      "grad_norm": 11.85594654083252,
      "learning_rate": 1.4330759330759332e-05,
      "loss": 0.7701,
      "step": 4870
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.651052474975586,
      "learning_rate": 1.4327541827541828e-05,
      "loss": 0.6392,
      "step": 4871
    },
    {
      "epoch": 1.05,
      "grad_norm": 12.919624328613281,
      "learning_rate": 1.4324324324324325e-05,
      "loss": 0.6844,
      "step": 4872
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.932621002197266,
      "learning_rate": 1.4321106821106821e-05,
      "loss": 0.6651,
      "step": 4873
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.32441234588623,
      "learning_rate": 1.431788931788932e-05,
      "loss": 0.5675,
      "step": 4874
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.072693824768066,
      "learning_rate": 1.4314671814671816e-05,
      "loss": 0.8741,
      "step": 4875
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.32227897644043,
      "learning_rate": 1.431145431145431e-05,
      "loss": 0.6847,
      "step": 4876
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.043230056762695,
      "learning_rate": 1.4308236808236809e-05,
      "loss": 0.7295,
      "step": 4877
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.177183151245117,
      "learning_rate": 1.4305019305019305e-05,
      "loss": 0.669,
      "step": 4878
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.41627025604248,
      "learning_rate": 1.4301801801801801e-05,
      "loss": 0.6019,
      "step": 4879
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.188755989074707,
      "learning_rate": 1.42985842985843e-05,
      "loss": 0.7494,
      "step": 4880
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.786432266235352,
      "learning_rate": 1.4295366795366796e-05,
      "loss": 0.7879,
      "step": 4881
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.305196762084961,
      "learning_rate": 1.4292149292149293e-05,
      "loss": 0.7458,
      "step": 4882
    },
    {
      "epoch": 1.05,
      "grad_norm": 7.951949596405029,
      "learning_rate": 1.428893178893179e-05,
      "loss": 0.5274,
      "step": 4883
    },
    {
      "epoch": 1.05,
      "grad_norm": 7.783168315887451,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.5121,
      "step": 4884
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.776348114013672,
      "learning_rate": 1.4282496782496782e-05,
      "loss": 0.8681,
      "step": 4885
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.878802299499512,
      "learning_rate": 1.427927927927928e-05,
      "loss": 0.5128,
      "step": 4886
    },
    {
      "epoch": 1.05,
      "grad_norm": 13.968427658081055,
      "learning_rate": 1.4276061776061776e-05,
      "loss": 0.6415,
      "step": 4887
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.23302936553955,
      "learning_rate": 1.4272844272844273e-05,
      "loss": 0.5316,
      "step": 4888
    },
    {
      "epoch": 1.05,
      "grad_norm": 7.842362403869629,
      "learning_rate": 1.426962676962677e-05,
      "loss": 0.6589,
      "step": 4889
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.485936164855957,
      "learning_rate": 1.4266409266409268e-05,
      "loss": 0.6884,
      "step": 4890
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.247602462768555,
      "learning_rate": 1.4263191763191764e-05,
      "loss": 0.6332,
      "step": 4891
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.441335678100586,
      "learning_rate": 1.425997425997426e-05,
      "loss": 0.6789,
      "step": 4892
    },
    {
      "epoch": 1.05,
      "grad_norm": 7.525876998901367,
      "learning_rate": 1.4256756756756757e-05,
      "loss": 0.6253,
      "step": 4893
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.542802810668945,
      "learning_rate": 1.4253539253539253e-05,
      "loss": 0.8182,
      "step": 4894
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.33840560913086,
      "learning_rate": 1.425032175032175e-05,
      "loss": 0.6941,
      "step": 4895
    },
    {
      "epoch": 1.05,
      "grad_norm": 12.288352012634277,
      "learning_rate": 1.4247104247104248e-05,
      "loss": 0.6128,
      "step": 4896
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.370649337768555,
      "learning_rate": 1.4243886743886744e-05,
      "loss": 0.59,
      "step": 4897
    },
    {
      "epoch": 1.05,
      "grad_norm": 13.36973762512207,
      "learning_rate": 1.424066924066924e-05,
      "loss": 0.8872,
      "step": 4898
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.485958099365234,
      "learning_rate": 1.4237451737451739e-05,
      "loss": 0.5296,
      "step": 4899
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.646842956542969,
      "learning_rate": 1.4234234234234235e-05,
      "loss": 0.9569,
      "step": 4900
    },
    {
      "epoch": 1.05,
      "grad_norm": 8.823404312133789,
      "learning_rate": 1.4231016731016732e-05,
      "loss": 0.6192,
      "step": 4901
    },
    {
      "epoch": 1.05,
      "grad_norm": 8.684821128845215,
      "learning_rate": 1.4227799227799227e-05,
      "loss": 0.7896,
      "step": 4902
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.065362930297852,
      "learning_rate": 1.4224581724581725e-05,
      "loss": 0.7277,
      "step": 4903
    },
    {
      "epoch": 1.05,
      "grad_norm": 8.292929649353027,
      "learning_rate": 1.4221364221364221e-05,
      "loss": 0.6122,
      "step": 4904
    },
    {
      "epoch": 1.05,
      "grad_norm": 12.161736488342285,
      "learning_rate": 1.4218146718146718e-05,
      "loss": 0.8347,
      "step": 4905
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.578341484069824,
      "learning_rate": 1.4214929214929216e-05,
      "loss": 0.5376,
      "step": 4906
    },
    {
      "epoch": 1.05,
      "grad_norm": 7.703756332397461,
      "learning_rate": 1.4211711711711712e-05,
      "loss": 0.47,
      "step": 4907
    },
    {
      "epoch": 1.05,
      "grad_norm": 12.895602226257324,
      "learning_rate": 1.4208494208494209e-05,
      "loss": 0.783,
      "step": 4908
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.833910942077637,
      "learning_rate": 1.4205276705276707e-05,
      "loss": 0.5315,
      "step": 4909
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.190449714660645,
      "learning_rate": 1.4202059202059202e-05,
      "loss": 0.8452,
      "step": 4910
    },
    {
      "epoch": 1.05,
      "grad_norm": 8.813928604125977,
      "learning_rate": 1.4198841698841698e-05,
      "loss": 0.6518,
      "step": 4911
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.758683204650879,
      "learning_rate": 1.4195624195624196e-05,
      "loss": 0.6863,
      "step": 4912
    },
    {
      "epoch": 1.05,
      "grad_norm": 12.50782299041748,
      "learning_rate": 1.4192406692406693e-05,
      "loss": 0.8713,
      "step": 4913
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.348894119262695,
      "learning_rate": 1.4189189189189189e-05,
      "loss": 0.6482,
      "step": 4914
    },
    {
      "epoch": 1.05,
      "grad_norm": 8.223034858703613,
      "learning_rate": 1.4185971685971687e-05,
      "loss": 0.6072,
      "step": 4915
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.42926025390625,
      "learning_rate": 1.4182754182754184e-05,
      "loss": 0.5854,
      "step": 4916
    },
    {
      "epoch": 1.05,
      "grad_norm": 9.191483497619629,
      "learning_rate": 1.417953667953668e-05,
      "loss": 0.5517,
      "step": 4917
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.379210472106934,
      "learning_rate": 1.4176319176319177e-05,
      "loss": 0.6247,
      "step": 4918
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.152480125427246,
      "learning_rate": 1.4173101673101673e-05,
      "loss": 0.4344,
      "step": 4919
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.333388328552246,
      "learning_rate": 1.416988416988417e-05,
      "loss": 0.7313,
      "step": 4920
    },
    {
      "epoch": 1.06,
      "grad_norm": 11.089681625366211,
      "learning_rate": 1.4166666666666666e-05,
      "loss": 0.696,
      "step": 4921
    },
    {
      "epoch": 1.06,
      "grad_norm": 14.574381828308105,
      "learning_rate": 1.4163449163449164e-05,
      "loss": 0.6828,
      "step": 4922
    },
    {
      "epoch": 1.06,
      "grad_norm": 13.769959449768066,
      "learning_rate": 1.416023166023166e-05,
      "loss": 0.7915,
      "step": 4923
    },
    {
      "epoch": 1.06,
      "grad_norm": 7.11357307434082,
      "learning_rate": 1.4157014157014157e-05,
      "loss": 0.3678,
      "step": 4924
    },
    {
      "epoch": 1.06,
      "grad_norm": 8.52049732208252,
      "learning_rate": 1.4153796653796655e-05,
      "loss": 0.4732,
      "step": 4925
    },
    {
      "epoch": 1.06,
      "grad_norm": 11.24434757232666,
      "learning_rate": 1.4150579150579152e-05,
      "loss": 0.7238,
      "step": 4926
    },
    {
      "epoch": 1.06,
      "grad_norm": 11.81529712677002,
      "learning_rate": 1.4147361647361648e-05,
      "loss": 0.749,
      "step": 4927
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.751593589782715,
      "learning_rate": 1.4144144144144145e-05,
      "loss": 0.6004,
      "step": 4928
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.795741081237793,
      "learning_rate": 1.4140926640926641e-05,
      "loss": 0.7554,
      "step": 4929
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.283608436584473,
      "learning_rate": 1.4137709137709137e-05,
      "loss": 0.6693,
      "step": 4930
    },
    {
      "epoch": 1.06,
      "grad_norm": 8.939718246459961,
      "learning_rate": 1.4134491634491634e-05,
      "loss": 0.5409,
      "step": 4931
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.089298248291016,
      "learning_rate": 1.4131274131274132e-05,
      "loss": 0.7153,
      "step": 4932
    },
    {
      "epoch": 1.06,
      "grad_norm": 11.158517837524414,
      "learning_rate": 1.4128056628056628e-05,
      "loss": 0.6429,
      "step": 4933
    },
    {
      "epoch": 1.06,
      "grad_norm": 12.89643669128418,
      "learning_rate": 1.4124839124839125e-05,
      "loss": 0.804,
      "step": 4934
    },
    {
      "epoch": 1.06,
      "grad_norm": 11.373899459838867,
      "learning_rate": 1.4121621621621623e-05,
      "loss": 0.5518,
      "step": 4935
    },
    {
      "epoch": 1.06,
      "grad_norm": 11.332212448120117,
      "learning_rate": 1.4118404118404118e-05,
      "loss": 0.8632,
      "step": 4936
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.651050567626953,
      "learning_rate": 1.4115186615186614e-05,
      "loss": 0.9276,
      "step": 4937
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.026448249816895,
      "learning_rate": 1.4111969111969112e-05,
      "loss": 0.7221,
      "step": 4938
    },
    {
      "epoch": 1.06,
      "grad_norm": 8.382022857666016,
      "learning_rate": 1.4108751608751609e-05,
      "loss": 0.6852,
      "step": 4939
    },
    {
      "epoch": 1.06,
      "grad_norm": 11.195173263549805,
      "learning_rate": 1.4105534105534105e-05,
      "loss": 0.9126,
      "step": 4940
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.009915351867676,
      "learning_rate": 1.4102316602316603e-05,
      "loss": 0.7329,
      "step": 4941
    },
    {
      "epoch": 1.06,
      "grad_norm": 11.954058647155762,
      "learning_rate": 1.40990990990991e-05,
      "loss": 0.7801,
      "step": 4942
    },
    {
      "epoch": 1.06,
      "grad_norm": 11.854996681213379,
      "learning_rate": 1.4095881595881596e-05,
      "loss": 0.7825,
      "step": 4943
    },
    {
      "epoch": 1.06,
      "grad_norm": 8.126480102539062,
      "learning_rate": 1.4092664092664095e-05,
      "loss": 0.6766,
      "step": 4944
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.061216354370117,
      "learning_rate": 1.408944658944659e-05,
      "loss": 0.6392,
      "step": 4945
    },
    {
      "epoch": 1.06,
      "grad_norm": 8.503010749816895,
      "learning_rate": 1.4086229086229086e-05,
      "loss": 0.6816,
      "step": 4946
    },
    {
      "epoch": 1.06,
      "grad_norm": 7.507368564605713,
      "learning_rate": 1.4083011583011582e-05,
      "loss": 0.6148,
      "step": 4947
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.7296781539917,
      "learning_rate": 1.407979407979408e-05,
      "loss": 0.6901,
      "step": 4948
    },
    {
      "epoch": 1.06,
      "grad_norm": 12.329980850219727,
      "learning_rate": 1.4076576576576577e-05,
      "loss": 0.6955,
      "step": 4949
    },
    {
      "epoch": 1.06,
      "grad_norm": 12.306546211242676,
      "learning_rate": 1.4073359073359073e-05,
      "loss": 0.8039,
      "step": 4950
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.99827766418457,
      "learning_rate": 1.4070141570141571e-05,
      "loss": 0.7424,
      "step": 4951
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.929429054260254,
      "learning_rate": 1.4066924066924068e-05,
      "loss": 0.7295,
      "step": 4952
    },
    {
      "epoch": 1.06,
      "grad_norm": 12.076838493347168,
      "learning_rate": 1.4063706563706564e-05,
      "loss": 0.5402,
      "step": 4953
    },
    {
      "epoch": 1.06,
      "grad_norm": 6.9655985832214355,
      "learning_rate": 1.406048906048906e-05,
      "loss": 0.4514,
      "step": 4954
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.360453605651855,
      "learning_rate": 1.4057271557271557e-05,
      "loss": 0.6854,
      "step": 4955
    },
    {
      "epoch": 1.06,
      "grad_norm": 11.037487983703613,
      "learning_rate": 1.4054054054054054e-05,
      "loss": 0.6056,
      "step": 4956
    },
    {
      "epoch": 1.06,
      "grad_norm": 7.423559188842773,
      "learning_rate": 1.4050836550836552e-05,
      "loss": 0.4315,
      "step": 4957
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.807165145874023,
      "learning_rate": 1.4047619047619048e-05,
      "loss": 0.6468,
      "step": 4958
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.741260528564453,
      "learning_rate": 1.4044401544401545e-05,
      "loss": 0.545,
      "step": 4959
    },
    {
      "epoch": 1.06,
      "grad_norm": 12.015979766845703,
      "learning_rate": 1.4041184041184041e-05,
      "loss": 0.7671,
      "step": 4960
    },
    {
      "epoch": 1.06,
      "grad_norm": 14.493820190429688,
      "learning_rate": 1.403796653796654e-05,
      "loss": 0.6867,
      "step": 4961
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.167725563049316,
      "learning_rate": 1.4034749034749036e-05,
      "loss": 0.7252,
      "step": 4962
    },
    {
      "epoch": 1.06,
      "grad_norm": 8.975651741027832,
      "learning_rate": 1.403153153153153e-05,
      "loss": 0.6737,
      "step": 4963
    },
    {
      "epoch": 1.06,
      "grad_norm": 8.056878089904785,
      "learning_rate": 1.4028314028314029e-05,
      "loss": 0.6671,
      "step": 4964
    },
    {
      "epoch": 1.06,
      "grad_norm": 8.944724082946777,
      "learning_rate": 1.4025096525096525e-05,
      "loss": 0.5751,
      "step": 4965
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.687226295471191,
      "learning_rate": 1.4021879021879022e-05,
      "loss": 0.579,
      "step": 4966
    },
    {
      "epoch": 1.07,
      "grad_norm": 14.148222923278809,
      "learning_rate": 1.401866151866152e-05,
      "loss": 1.0082,
      "step": 4967
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.588907241821289,
      "learning_rate": 1.4015444015444016e-05,
      "loss": 0.7138,
      "step": 4968
    },
    {
      "epoch": 1.07,
      "grad_norm": 14.088830947875977,
      "learning_rate": 1.4012226512226513e-05,
      "loss": 0.6223,
      "step": 4969
    },
    {
      "epoch": 1.07,
      "grad_norm": 11.467409133911133,
      "learning_rate": 1.400900900900901e-05,
      "loss": 0.6751,
      "step": 4970
    },
    {
      "epoch": 1.07,
      "grad_norm": 11.57931900024414,
      "learning_rate": 1.4005791505791505e-05,
      "loss": 0.6169,
      "step": 4971
    },
    {
      "epoch": 1.07,
      "grad_norm": 11.864960670471191,
      "learning_rate": 1.4002574002574002e-05,
      "loss": 0.6179,
      "step": 4972
    },
    {
      "epoch": 1.07,
      "grad_norm": 12.380422592163086,
      "learning_rate": 1.39993564993565e-05,
      "loss": 0.7052,
      "step": 4973
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.706289291381836,
      "learning_rate": 1.3996138996138997e-05,
      "loss": 0.7239,
      "step": 4974
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.403722763061523,
      "learning_rate": 1.3992921492921493e-05,
      "loss": 0.7435,
      "step": 4975
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.76307487487793,
      "learning_rate": 1.398970398970399e-05,
      "loss": 0.6579,
      "step": 4976
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.332951545715332,
      "learning_rate": 1.3986486486486488e-05,
      "loss": 0.5858,
      "step": 4977
    },
    {
      "epoch": 1.07,
      "grad_norm": 8.694962501525879,
      "learning_rate": 1.3983268983268984e-05,
      "loss": 0.6914,
      "step": 4978
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.714617729187012,
      "learning_rate": 1.398005148005148e-05,
      "loss": 0.6533,
      "step": 4979
    },
    {
      "epoch": 1.07,
      "grad_norm": 11.973474502563477,
      "learning_rate": 1.3976833976833977e-05,
      "loss": 0.6675,
      "step": 4980
    },
    {
      "epoch": 1.07,
      "grad_norm": 14.446425437927246,
      "learning_rate": 1.3973616473616473e-05,
      "loss": 0.9402,
      "step": 4981
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.806333541870117,
      "learning_rate": 1.397039897039897e-05,
      "loss": 0.8074,
      "step": 4982
    },
    {
      "epoch": 1.07,
      "grad_norm": 11.503890991210938,
      "learning_rate": 1.3967181467181468e-05,
      "loss": 0.8134,
      "step": 4983
    },
    {
      "epoch": 1.07,
      "grad_norm": 12.610923767089844,
      "learning_rate": 1.3963963963963964e-05,
      "loss": 0.7449,
      "step": 4984
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.935331344604492,
      "learning_rate": 1.3960746460746461e-05,
      "loss": 0.664,
      "step": 4985
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.138945579528809,
      "learning_rate": 1.3957528957528959e-05,
      "loss": 0.708,
      "step": 4986
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.55648422241211,
      "learning_rate": 1.3954311454311455e-05,
      "loss": 0.7914,
      "step": 4987
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.387012481689453,
      "learning_rate": 1.3951093951093952e-05,
      "loss": 0.667,
      "step": 4988
    },
    {
      "epoch": 1.07,
      "grad_norm": 7.888586521148682,
      "learning_rate": 1.3947876447876448e-05,
      "loss": 0.6572,
      "step": 4989
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.32381534576416,
      "learning_rate": 1.3944658944658945e-05,
      "loss": 0.5568,
      "step": 4990
    },
    {
      "epoch": 1.07,
      "grad_norm": 8.682762145996094,
      "learning_rate": 1.3941441441441441e-05,
      "loss": 0.6772,
      "step": 4991
    },
    {
      "epoch": 1.07,
      "grad_norm": 11.252490997314453,
      "learning_rate": 1.3938223938223938e-05,
      "loss": 0.6888,
      "step": 4992
    },
    {
      "epoch": 1.07,
      "grad_norm": 11.032356262207031,
      "learning_rate": 1.3935006435006436e-05,
      "loss": 0.5253,
      "step": 4993
    },
    {
      "epoch": 1.07,
      "grad_norm": 15.895098686218262,
      "learning_rate": 1.3931788931788932e-05,
      "loss": 0.9138,
      "step": 4994
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.558392524719238,
      "learning_rate": 1.3928571428571429e-05,
      "loss": 0.6208,
      "step": 4995
    },
    {
      "epoch": 1.07,
      "grad_norm": 8.746026992797852,
      "learning_rate": 1.3925353925353927e-05,
      "loss": 0.6167,
      "step": 4996
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.502511024475098,
      "learning_rate": 1.3922136422136422e-05,
      "loss": 0.6793,
      "step": 4997
    },
    {
      "epoch": 1.07,
      "grad_norm": 11.003608703613281,
      "learning_rate": 1.3918918918918918e-05,
      "loss": 0.6446,
      "step": 4998
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.241630554199219,
      "learning_rate": 1.3915701415701416e-05,
      "loss": 0.7694,
      "step": 4999
    },
    {
      "epoch": 1.07,
      "grad_norm": 12.749432563781738,
      "learning_rate": 1.3912483912483913e-05,
      "loss": 0.9371,
      "step": 5000
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.62473201751709,
      "learning_rate": 1.390926640926641e-05,
      "loss": 0.4224,
      "step": 5001
    },
    {
      "epoch": 1.07,
      "grad_norm": 12.675978660583496,
      "learning_rate": 1.3906048906048907e-05,
      "loss": 0.8215,
      "step": 5002
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.370100021362305,
      "learning_rate": 1.3902831402831404e-05,
      "loss": 0.6629,
      "step": 5003
    },
    {
      "epoch": 1.07,
      "grad_norm": 8.843768119812012,
      "learning_rate": 1.38996138996139e-05,
      "loss": 0.6842,
      "step": 5004
    },
    {
      "epoch": 1.07,
      "grad_norm": 11.988468170166016,
      "learning_rate": 1.3896396396396397e-05,
      "loss": 0.9396,
      "step": 5005
    },
    {
      "epoch": 1.07,
      "grad_norm": 9.683004379272461,
      "learning_rate": 1.3893178893178893e-05,
      "loss": 0.7738,
      "step": 5006
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.2590970993042,
      "learning_rate": 1.388996138996139e-05,
      "loss": 0.7374,
      "step": 5007
    },
    {
      "epoch": 1.07,
      "grad_norm": 13.260856628417969,
      "learning_rate": 1.3886743886743886e-05,
      "loss": 0.8034,
      "step": 5008
    },
    {
      "epoch": 1.07,
      "grad_norm": 12.020692825317383,
      "learning_rate": 1.3883526383526384e-05,
      "loss": 0.7908,
      "step": 5009
    },
    {
      "epoch": 1.07,
      "grad_norm": 12.957403182983398,
      "learning_rate": 1.388030888030888e-05,
      "loss": 0.8177,
      "step": 5010
    },
    {
      "epoch": 1.07,
      "grad_norm": 8.870211601257324,
      "learning_rate": 1.3877091377091377e-05,
      "loss": 0.67,
      "step": 5011
    },
    {
      "epoch": 1.08,
      "grad_norm": 11.2058744430542,
      "learning_rate": 1.3873873873873875e-05,
      "loss": 0.5391,
      "step": 5012
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.54045581817627,
      "learning_rate": 1.3870656370656372e-05,
      "loss": 0.5438,
      "step": 5013
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.25883674621582,
      "learning_rate": 1.3867438867438868e-05,
      "loss": 0.544,
      "step": 5014
    },
    {
      "epoch": 1.08,
      "grad_norm": 11.02603530883789,
      "learning_rate": 1.3864221364221365e-05,
      "loss": 0.6433,
      "step": 5015
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.50733470916748,
      "learning_rate": 1.3861003861003861e-05,
      "loss": 0.6207,
      "step": 5016
    },
    {
      "epoch": 1.08,
      "grad_norm": 7.817746639251709,
      "learning_rate": 1.3857786357786358e-05,
      "loss": 0.5743,
      "step": 5017
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.551469802856445,
      "learning_rate": 1.3854568854568854e-05,
      "loss": 0.5631,
      "step": 5018
    },
    {
      "epoch": 1.08,
      "grad_norm": 6.3197197914123535,
      "learning_rate": 1.3851351351351352e-05,
      "loss": 0.4924,
      "step": 5019
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.764299392700195,
      "learning_rate": 1.3848133848133849e-05,
      "loss": 0.532,
      "step": 5020
    },
    {
      "epoch": 1.08,
      "grad_norm": 11.059623718261719,
      "learning_rate": 1.3844916344916345e-05,
      "loss": 0.7429,
      "step": 5021
    },
    {
      "epoch": 1.08,
      "grad_norm": 12.426314353942871,
      "learning_rate": 1.3841698841698843e-05,
      "loss": 0.7223,
      "step": 5022
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.975492477416992,
      "learning_rate": 1.3838481338481338e-05,
      "loss": 0.6731,
      "step": 5023
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.357464790344238,
      "learning_rate": 1.3835263835263834e-05,
      "loss": 0.7194,
      "step": 5024
    },
    {
      "epoch": 1.08,
      "grad_norm": 12.704051971435547,
      "learning_rate": 1.3832046332046333e-05,
      "loss": 0.6381,
      "step": 5025
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.198379516601562,
      "learning_rate": 1.3828828828828829e-05,
      "loss": 0.539,
      "step": 5026
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.13636589050293,
      "learning_rate": 1.3825611325611325e-05,
      "loss": 0.5857,
      "step": 5027
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.709383010864258,
      "learning_rate": 1.3822393822393824e-05,
      "loss": 0.523,
      "step": 5028
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.811192512512207,
      "learning_rate": 1.381917631917632e-05,
      "loss": 0.5181,
      "step": 5029
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.423086166381836,
      "learning_rate": 1.3815958815958816e-05,
      "loss": 0.8319,
      "step": 5030
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.80160140991211,
      "learning_rate": 1.3812741312741315e-05,
      "loss": 0.7034,
      "step": 5031
    },
    {
      "epoch": 1.08,
      "grad_norm": 11.764652252197266,
      "learning_rate": 1.380952380952381e-05,
      "loss": 0.6986,
      "step": 5032
    },
    {
      "epoch": 1.08,
      "grad_norm": 11.815657615661621,
      "learning_rate": 1.3806306306306306e-05,
      "loss": 0.6795,
      "step": 5033
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.5960111618042,
      "learning_rate": 1.3803088803088802e-05,
      "loss": 0.6463,
      "step": 5034
    },
    {
      "epoch": 1.08,
      "grad_norm": 13.327552795410156,
      "learning_rate": 1.37998712998713e-05,
      "loss": 0.8933,
      "step": 5035
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.87982177734375,
      "learning_rate": 1.3796653796653797e-05,
      "loss": 0.6679,
      "step": 5036
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.538434982299805,
      "learning_rate": 1.3793436293436293e-05,
      "loss": 0.5874,
      "step": 5037
    },
    {
      "epoch": 1.08,
      "grad_norm": 7.436265468597412,
      "learning_rate": 1.3790218790218791e-05,
      "loss": 0.4251,
      "step": 5038
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.91930866241455,
      "learning_rate": 1.3787001287001288e-05,
      "loss": 0.5348,
      "step": 5039
    },
    {
      "epoch": 1.08,
      "grad_norm": 7.661568641662598,
      "learning_rate": 1.3783783783783784e-05,
      "loss": 0.6189,
      "step": 5040
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.077386856079102,
      "learning_rate": 1.378056628056628e-05,
      "loss": 0.6448,
      "step": 5041
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.739809036254883,
      "learning_rate": 1.3777348777348777e-05,
      "loss": 0.6652,
      "step": 5042
    },
    {
      "epoch": 1.08,
      "grad_norm": 12.116921424865723,
      "learning_rate": 1.3774131274131274e-05,
      "loss": 0.6774,
      "step": 5043
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.681513786315918,
      "learning_rate": 1.3770913770913772e-05,
      "loss": 0.8192,
      "step": 5044
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.478577613830566,
      "learning_rate": 1.3767696267696268e-05,
      "loss": 0.7622,
      "step": 5045
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.531167030334473,
      "learning_rate": 1.3764478764478765e-05,
      "loss": 0.913,
      "step": 5046
    },
    {
      "epoch": 1.08,
      "grad_norm": 11.333176612854004,
      "learning_rate": 1.3761261261261261e-05,
      "loss": 0.7437,
      "step": 5047
    },
    {
      "epoch": 1.08,
      "grad_norm": 11.75042724609375,
      "learning_rate": 1.375804375804376e-05,
      "loss": 0.7565,
      "step": 5048
    },
    {
      "epoch": 1.08,
      "grad_norm": 11.893110275268555,
      "learning_rate": 1.3754826254826256e-05,
      "loss": 0.8135,
      "step": 5049
    },
    {
      "epoch": 1.08,
      "grad_norm": 12.348820686340332,
      "learning_rate": 1.375160875160875e-05,
      "loss": 0.7717,
      "step": 5050
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.39014720916748,
      "learning_rate": 1.3748391248391249e-05,
      "loss": 0.6362,
      "step": 5051
    },
    {
      "epoch": 1.08,
      "grad_norm": 7.657466888427734,
      "learning_rate": 1.3745173745173745e-05,
      "loss": 0.5611,
      "step": 5052
    },
    {
      "epoch": 1.08,
      "grad_norm": 11.158803939819336,
      "learning_rate": 1.3741956241956242e-05,
      "loss": 0.8324,
      "step": 5053
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.040008544921875,
      "learning_rate": 1.373873873873874e-05,
      "loss": 0.8095,
      "step": 5054
    },
    {
      "epoch": 1.08,
      "grad_norm": 11.445072174072266,
      "learning_rate": 1.3735521235521236e-05,
      "loss": 0.8442,
      "step": 5055
    },
    {
      "epoch": 1.08,
      "grad_norm": 12.570770263671875,
      "learning_rate": 1.3732303732303733e-05,
      "loss": 0.7591,
      "step": 5056
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.32651138305664,
      "learning_rate": 1.372908622908623e-05,
      "loss": 0.5776,
      "step": 5057
    },
    {
      "epoch": 1.08,
      "grad_norm": 11.249754905700684,
      "learning_rate": 1.3725868725868726e-05,
      "loss": 0.7808,
      "step": 5058
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.959761619567871,
      "learning_rate": 1.3722651222651222e-05,
      "loss": 0.7221,
      "step": 5059
    },
    {
      "epoch": 1.09,
      "grad_norm": 10.399237632751465,
      "learning_rate": 1.371943371943372e-05,
      "loss": 0.7389,
      "step": 5060
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.614335060119629,
      "learning_rate": 1.3716216216216217e-05,
      "loss": 0.6694,
      "step": 5061
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.054352760314941,
      "learning_rate": 1.3712998712998713e-05,
      "loss": 0.5383,
      "step": 5062
    },
    {
      "epoch": 1.09,
      "grad_norm": 7.207559108734131,
      "learning_rate": 1.370978120978121e-05,
      "loss": 0.3629,
      "step": 5063
    },
    {
      "epoch": 1.09,
      "grad_norm": 10.477249145507812,
      "learning_rate": 1.3706563706563708e-05,
      "loss": 0.7821,
      "step": 5064
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.170259475708008,
      "learning_rate": 1.3703346203346204e-05,
      "loss": 0.7302,
      "step": 5065
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.619468688964844,
      "learning_rate": 1.37001287001287e-05,
      "loss": 0.8805,
      "step": 5066
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.712307929992676,
      "learning_rate": 1.3696911196911197e-05,
      "loss": 0.7347,
      "step": 5067
    },
    {
      "epoch": 1.09,
      "grad_norm": 7.829373359680176,
      "learning_rate": 1.3693693693693693e-05,
      "loss": 0.4732,
      "step": 5068
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.483933448791504,
      "learning_rate": 1.369047619047619e-05,
      "loss": 0.8105,
      "step": 5069
    },
    {
      "epoch": 1.09,
      "grad_norm": 6.7573466300964355,
      "learning_rate": 1.3687258687258688e-05,
      "loss": 0.4799,
      "step": 5070
    },
    {
      "epoch": 1.09,
      "grad_norm": 10.342119216918945,
      "learning_rate": 1.3684041184041185e-05,
      "loss": 0.6874,
      "step": 5071
    },
    {
      "epoch": 1.09,
      "grad_norm": 16.005569458007812,
      "learning_rate": 1.3680823680823681e-05,
      "loss": 0.7444,
      "step": 5072
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.725658416748047,
      "learning_rate": 1.3677606177606179e-05,
      "loss": 0.6248,
      "step": 5073
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.983840942382812,
      "learning_rate": 1.3674388674388676e-05,
      "loss": 0.6506,
      "step": 5074
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.508960723876953,
      "learning_rate": 1.3671171171171172e-05,
      "loss": 0.6987,
      "step": 5075
    },
    {
      "epoch": 1.09,
      "grad_norm": 7.671133995056152,
      "learning_rate": 1.3667953667953668e-05,
      "loss": 0.5527,
      "step": 5076
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.472183227539062,
      "learning_rate": 1.3664736164736165e-05,
      "loss": 0.5796,
      "step": 5077
    },
    {
      "epoch": 1.09,
      "grad_norm": 10.347404479980469,
      "learning_rate": 1.3661518661518661e-05,
      "loss": 0.6134,
      "step": 5078
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.948147773742676,
      "learning_rate": 1.3658301158301158e-05,
      "loss": 0.6051,
      "step": 5079
    },
    {
      "epoch": 1.09,
      "grad_norm": 15.558761596679688,
      "learning_rate": 1.3655083655083656e-05,
      "loss": 0.7238,
      "step": 5080
    },
    {
      "epoch": 1.09,
      "grad_norm": 10.123929977416992,
      "learning_rate": 1.3651866151866152e-05,
      "loss": 0.6191,
      "step": 5081
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.48809814453125,
      "learning_rate": 1.3648648648648649e-05,
      "loss": 0.5826,
      "step": 5082
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.097224235534668,
      "learning_rate": 1.3645431145431147e-05,
      "loss": 0.6631,
      "step": 5083
    },
    {
      "epoch": 1.09,
      "grad_norm": 13.747879028320312,
      "learning_rate": 1.3642213642213642e-05,
      "loss": 0.8466,
      "step": 5084
    },
    {
      "epoch": 1.09,
      "grad_norm": 7.856277942657471,
      "learning_rate": 1.3638996138996138e-05,
      "loss": 0.4372,
      "step": 5085
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.394298553466797,
      "learning_rate": 1.3635778635778636e-05,
      "loss": 0.768,
      "step": 5086
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.823302268981934,
      "learning_rate": 1.3632561132561133e-05,
      "loss": 0.6563,
      "step": 5087
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.077655792236328,
      "learning_rate": 1.362934362934363e-05,
      "loss": 0.6382,
      "step": 5088
    },
    {
      "epoch": 1.09,
      "grad_norm": 14.18767261505127,
      "learning_rate": 1.3626126126126127e-05,
      "loss": 0.786,
      "step": 5089
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.359672546386719,
      "learning_rate": 1.3622908622908624e-05,
      "loss": 0.5115,
      "step": 5090
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.384660720825195,
      "learning_rate": 1.361969111969112e-05,
      "loss": 0.8013,
      "step": 5091
    },
    {
      "epoch": 1.09,
      "grad_norm": 12.68889045715332,
      "learning_rate": 1.3616473616473617e-05,
      "loss": 1.0058,
      "step": 5092
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.737579345703125,
      "learning_rate": 1.3613256113256113e-05,
      "loss": 0.5725,
      "step": 5093
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.90058422088623,
      "learning_rate": 1.361003861003861e-05,
      "loss": 0.6329,
      "step": 5094
    },
    {
      "epoch": 1.09,
      "grad_norm": 10.645302772521973,
      "learning_rate": 1.3606821106821106e-05,
      "loss": 0.6738,
      "step": 5095
    },
    {
      "epoch": 1.09,
      "grad_norm": 10.507955551147461,
      "learning_rate": 1.3603603603603604e-05,
      "loss": 0.9994,
      "step": 5096
    },
    {
      "epoch": 1.09,
      "grad_norm": 10.738907814025879,
      "learning_rate": 1.36003861003861e-05,
      "loss": 0.4392,
      "step": 5097
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.685291290283203,
      "learning_rate": 1.3597168597168597e-05,
      "loss": 0.5055,
      "step": 5098
    },
    {
      "epoch": 1.09,
      "grad_norm": 10.455607414245605,
      "learning_rate": 1.3593951093951095e-05,
      "loss": 0.7955,
      "step": 5099
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.629257202148438,
      "learning_rate": 1.3590733590733592e-05,
      "loss": 0.4476,
      "step": 5100
    },
    {
      "epoch": 1.09,
      "grad_norm": 10.20829963684082,
      "learning_rate": 1.3587516087516088e-05,
      "loss": 0.5883,
      "step": 5101
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.629561424255371,
      "learning_rate": 1.3584298584298585e-05,
      "loss": 0.751,
      "step": 5102
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.629286766052246,
      "learning_rate": 1.3581081081081081e-05,
      "loss": 0.5355,
      "step": 5103
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.401065826416016,
      "learning_rate": 1.3577863577863578e-05,
      "loss": 0.7896,
      "step": 5104
    },
    {
      "epoch": 1.1,
      "grad_norm": 8.534168243408203,
      "learning_rate": 1.3574646074646076e-05,
      "loss": 0.5847,
      "step": 5105
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.652274131774902,
      "learning_rate": 1.3571428571428572e-05,
      "loss": 0.7499,
      "step": 5106
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.412129402160645,
      "learning_rate": 1.3568211068211069e-05,
      "loss": 0.5106,
      "step": 5107
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.082197189331055,
      "learning_rate": 1.3564993564993565e-05,
      "loss": 0.7542,
      "step": 5108
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.644840240478516,
      "learning_rate": 1.3561776061776063e-05,
      "loss": 0.6508,
      "step": 5109
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.522741317749023,
      "learning_rate": 1.3558558558558558e-05,
      "loss": 0.8772,
      "step": 5110
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.815024375915527,
      "learning_rate": 1.3555341055341054e-05,
      "loss": 0.6829,
      "step": 5111
    },
    {
      "epoch": 1.1,
      "grad_norm": 7.874946117401123,
      "learning_rate": 1.3552123552123553e-05,
      "loss": 0.6132,
      "step": 5112
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.161203384399414,
      "learning_rate": 1.3548906048906049e-05,
      "loss": 0.6843,
      "step": 5113
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.085577011108398,
      "learning_rate": 1.3545688545688545e-05,
      "loss": 0.4378,
      "step": 5114
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.381163597106934,
      "learning_rate": 1.3542471042471044e-05,
      "loss": 0.6841,
      "step": 5115
    },
    {
      "epoch": 1.1,
      "grad_norm": 12.210912704467773,
      "learning_rate": 1.353925353925354e-05,
      "loss": 0.662,
      "step": 5116
    },
    {
      "epoch": 1.1,
      "grad_norm": 12.240143775939941,
      "learning_rate": 1.3536036036036037e-05,
      "loss": 0.8208,
      "step": 5117
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.896749496459961,
      "learning_rate": 1.3532818532818535e-05,
      "loss": 0.7136,
      "step": 5118
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.576362609863281,
      "learning_rate": 1.352960102960103e-05,
      "loss": 0.9203,
      "step": 5119
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.646180152893066,
      "learning_rate": 1.3526383526383526e-05,
      "loss": 0.4529,
      "step": 5120
    },
    {
      "epoch": 1.1,
      "grad_norm": 6.686718463897705,
      "learning_rate": 1.3523166023166022e-05,
      "loss": 0.4922,
      "step": 5121
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.867918014526367,
      "learning_rate": 1.351994851994852e-05,
      "loss": 0.4228,
      "step": 5122
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.05196475982666,
      "learning_rate": 1.3516731016731017e-05,
      "loss": 0.6662,
      "step": 5123
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.966168403625488,
      "learning_rate": 1.3513513513513513e-05,
      "loss": 0.8335,
      "step": 5124
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.948294639587402,
      "learning_rate": 1.3510296010296012e-05,
      "loss": 0.6754,
      "step": 5125
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.773146629333496,
      "learning_rate": 1.3507078507078508e-05,
      "loss": 0.6926,
      "step": 5126
    },
    {
      "epoch": 1.1,
      "grad_norm": 8.291097640991211,
      "learning_rate": 1.3503861003861004e-05,
      "loss": 0.6122,
      "step": 5127
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.278820991516113,
      "learning_rate": 1.3500643500643501e-05,
      "loss": 0.6886,
      "step": 5128
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.736601829528809,
      "learning_rate": 1.3497425997425997e-05,
      "loss": 0.5589,
      "step": 5129
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.718415260314941,
      "learning_rate": 1.3494208494208494e-05,
      "loss": 0.7274,
      "step": 5130
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.704108238220215,
      "learning_rate": 1.3490990990990992e-05,
      "loss": 0.525,
      "step": 5131
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.94179630279541,
      "learning_rate": 1.3487773487773488e-05,
      "loss": 0.5905,
      "step": 5132
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.149895668029785,
      "learning_rate": 1.3484555984555985e-05,
      "loss": 0.7234,
      "step": 5133
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.319509506225586,
      "learning_rate": 1.3481338481338481e-05,
      "loss": 0.5484,
      "step": 5134
    },
    {
      "epoch": 1.1,
      "grad_norm": 8.492171287536621,
      "learning_rate": 1.347812097812098e-05,
      "loss": 0.5691,
      "step": 5135
    },
    {
      "epoch": 1.1,
      "grad_norm": 12.252252578735352,
      "learning_rate": 1.3474903474903476e-05,
      "loss": 0.8318,
      "step": 5136
    },
    {
      "epoch": 1.1,
      "grad_norm": 8.735631942749023,
      "learning_rate": 1.347168597168597e-05,
      "loss": 0.5067,
      "step": 5137
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.821855545043945,
      "learning_rate": 1.3468468468468469e-05,
      "loss": 0.5247,
      "step": 5138
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.234143257141113,
      "learning_rate": 1.3465250965250965e-05,
      "loss": 0.4634,
      "step": 5139
    },
    {
      "epoch": 1.1,
      "grad_norm": 13.103446960449219,
      "learning_rate": 1.3462033462033462e-05,
      "loss": 0.5111,
      "step": 5140
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.834999084472656,
      "learning_rate": 1.345881595881596e-05,
      "loss": 0.6181,
      "step": 5141
    },
    {
      "epoch": 1.1,
      "grad_norm": 8.245920181274414,
      "learning_rate": 1.3455598455598456e-05,
      "loss": 0.7405,
      "step": 5142
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.529219627380371,
      "learning_rate": 1.3452380952380953e-05,
      "loss": 0.7396,
      "step": 5143
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.539571762084961,
      "learning_rate": 1.3449163449163451e-05,
      "loss": 0.93,
      "step": 5144
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.546026229858398,
      "learning_rate": 1.3445945945945946e-05,
      "loss": 0.5716,
      "step": 5145
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.42587661743164,
      "learning_rate": 1.3442728442728442e-05,
      "loss": 0.6202,
      "step": 5146
    },
    {
      "epoch": 1.1,
      "grad_norm": 12.145190238952637,
      "learning_rate": 1.343951093951094e-05,
      "loss": 0.5398,
      "step": 5147
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.082707405090332,
      "learning_rate": 1.3436293436293437e-05,
      "loss": 0.6681,
      "step": 5148
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.751995086669922,
      "learning_rate": 1.3433075933075933e-05,
      "loss": 0.721,
      "step": 5149
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.443486213684082,
      "learning_rate": 1.342985842985843e-05,
      "loss": 0.7516,
      "step": 5150
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.067510604858398,
      "learning_rate": 1.3426640926640928e-05,
      "loss": 0.5409,
      "step": 5151
    },
    {
      "epoch": 1.11,
      "grad_norm": 9.84510326385498,
      "learning_rate": 1.3423423423423424e-05,
      "loss": 0.6732,
      "step": 5152
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.240466117858887,
      "learning_rate": 1.342020592020592e-05,
      "loss": 0.6256,
      "step": 5153
    },
    {
      "epoch": 1.11,
      "grad_norm": 13.045954704284668,
      "learning_rate": 1.3416988416988417e-05,
      "loss": 0.947,
      "step": 5154
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.896729469299316,
      "learning_rate": 1.3413770913770914e-05,
      "loss": 0.7424,
      "step": 5155
    },
    {
      "epoch": 1.11,
      "grad_norm": 7.882411003112793,
      "learning_rate": 1.341055341055341e-05,
      "loss": 0.574,
      "step": 5156
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.620877265930176,
      "learning_rate": 1.3407335907335908e-05,
      "loss": 0.607,
      "step": 5157
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.60891056060791,
      "learning_rate": 1.3404118404118405e-05,
      "loss": 0.6181,
      "step": 5158
    },
    {
      "epoch": 1.11,
      "grad_norm": 12.78149700164795,
      "learning_rate": 1.3400900900900901e-05,
      "loss": 0.9663,
      "step": 5159
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.7811861038208,
      "learning_rate": 1.33976833976834e-05,
      "loss": 0.8732,
      "step": 5160
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.114187240600586,
      "learning_rate": 1.3394465894465896e-05,
      "loss": 0.781,
      "step": 5161
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.643036842346191,
      "learning_rate": 1.3391248391248392e-05,
      "loss": 0.6752,
      "step": 5162
    },
    {
      "epoch": 1.11,
      "grad_norm": 9.550812721252441,
      "learning_rate": 1.3388030888030889e-05,
      "loss": 0.6121,
      "step": 5163
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.752054214477539,
      "learning_rate": 1.3384813384813385e-05,
      "loss": 0.7572,
      "step": 5164
    },
    {
      "epoch": 1.11,
      "grad_norm": 15.7797269821167,
      "learning_rate": 1.3381595881595881e-05,
      "loss": 0.8236,
      "step": 5165
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.766641616821289,
      "learning_rate": 1.3378378378378378e-05,
      "loss": 0.7783,
      "step": 5166
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.646174430847168,
      "learning_rate": 1.3375160875160876e-05,
      "loss": 0.8262,
      "step": 5167
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.84013557434082,
      "learning_rate": 1.3371943371943372e-05,
      "loss": 0.6999,
      "step": 5168
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.641590118408203,
      "learning_rate": 1.3368725868725869e-05,
      "loss": 0.827,
      "step": 5169
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.924610137939453,
      "learning_rate": 1.3365508365508367e-05,
      "loss": 0.5625,
      "step": 5170
    },
    {
      "epoch": 1.11,
      "grad_norm": 9.441954612731934,
      "learning_rate": 1.3362290862290862e-05,
      "loss": 0.6005,
      "step": 5171
    },
    {
      "epoch": 1.11,
      "grad_norm": 12.509946823120117,
      "learning_rate": 1.3359073359073358e-05,
      "loss": 0.6235,
      "step": 5172
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.715995788574219,
      "learning_rate": 1.3355855855855856e-05,
      "loss": 0.8225,
      "step": 5173
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.65477180480957,
      "learning_rate": 1.3352638352638353e-05,
      "loss": 0.6163,
      "step": 5174
    },
    {
      "epoch": 1.11,
      "grad_norm": 9.236153602600098,
      "learning_rate": 1.334942084942085e-05,
      "loss": 0.6607,
      "step": 5175
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.202003479003906,
      "learning_rate": 1.3346203346203347e-05,
      "loss": 0.6333,
      "step": 5176
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.952473640441895,
      "learning_rate": 1.3342985842985844e-05,
      "loss": 0.7892,
      "step": 5177
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.120407104492188,
      "learning_rate": 1.333976833976834e-05,
      "loss": 0.8332,
      "step": 5178
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.222458839416504,
      "learning_rate": 1.3336550836550837e-05,
      "loss": 0.5874,
      "step": 5179
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.625001907348633,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.6546,
      "step": 5180
    },
    {
      "epoch": 1.11,
      "grad_norm": 12.138298034667969,
      "learning_rate": 1.333011583011583e-05,
      "loss": 0.6537,
      "step": 5181
    },
    {
      "epoch": 1.11,
      "grad_norm": 7.236739158630371,
      "learning_rate": 1.3326898326898326e-05,
      "loss": 0.515,
      "step": 5182
    },
    {
      "epoch": 1.11,
      "grad_norm": 7.27134895324707,
      "learning_rate": 1.3323680823680824e-05,
      "loss": 0.5134,
      "step": 5183
    },
    {
      "epoch": 1.11,
      "grad_norm": 12.10701847076416,
      "learning_rate": 1.332046332046332e-05,
      "loss": 0.4549,
      "step": 5184
    },
    {
      "epoch": 1.11,
      "grad_norm": 8.688936233520508,
      "learning_rate": 1.3317245817245817e-05,
      "loss": 0.6085,
      "step": 5185
    },
    {
      "epoch": 1.11,
      "grad_norm": 9.491938591003418,
      "learning_rate": 1.3314028314028315e-05,
      "loss": 0.5034,
      "step": 5186
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.960509300231934,
      "learning_rate": 1.3310810810810812e-05,
      "loss": 0.7964,
      "step": 5187
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.786365509033203,
      "learning_rate": 1.3307593307593308e-05,
      "loss": 0.6297,
      "step": 5188
    },
    {
      "epoch": 1.11,
      "grad_norm": 9.755793571472168,
      "learning_rate": 1.3304375804375805e-05,
      "loss": 0.6922,
      "step": 5189
    },
    {
      "epoch": 1.11,
      "grad_norm": 12.163238525390625,
      "learning_rate": 1.3301158301158301e-05,
      "loss": 0.7436,
      "step": 5190
    },
    {
      "epoch": 1.11,
      "grad_norm": 12.559078216552734,
      "learning_rate": 1.3297940797940798e-05,
      "loss": 0.6716,
      "step": 5191
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.49539566040039,
      "learning_rate": 1.3294723294723296e-05,
      "loss": 0.7162,
      "step": 5192
    },
    {
      "epoch": 1.11,
      "grad_norm": 11.086441993713379,
      "learning_rate": 1.3291505791505792e-05,
      "loss": 0.9189,
      "step": 5193
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.989916801452637,
      "learning_rate": 1.3288288288288289e-05,
      "loss": 0.8145,
      "step": 5194
    },
    {
      "epoch": 1.11,
      "grad_norm": 7.985320568084717,
      "learning_rate": 1.3285070785070785e-05,
      "loss": 0.6359,
      "step": 5195
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.952129364013672,
      "learning_rate": 1.3281853281853283e-05,
      "loss": 0.8825,
      "step": 5196
    },
    {
      "epoch": 1.11,
      "grad_norm": 7.7213897705078125,
      "learning_rate": 1.3278635778635778e-05,
      "loss": 0.4687,
      "step": 5197
    },
    {
      "epoch": 1.11,
      "grad_norm": 8.692100524902344,
      "learning_rate": 1.3275418275418275e-05,
      "loss": 0.4492,
      "step": 5198
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.95101261138916,
      "learning_rate": 1.3272200772200773e-05,
      "loss": 0.6254,
      "step": 5199
    },
    {
      "epoch": 1.12,
      "grad_norm": 9.379908561706543,
      "learning_rate": 1.3268983268983269e-05,
      "loss": 0.5808,
      "step": 5200
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.897639274597168,
      "learning_rate": 1.3265765765765766e-05,
      "loss": 0.5166,
      "step": 5201
    },
    {
      "epoch": 1.12,
      "grad_norm": 13.046319007873535,
      "learning_rate": 1.3262548262548264e-05,
      "loss": 0.7754,
      "step": 5202
    },
    {
      "epoch": 1.12,
      "grad_norm": 9.256120681762695,
      "learning_rate": 1.325933075933076e-05,
      "loss": 0.6238,
      "step": 5203
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.512558937072754,
      "learning_rate": 1.3256113256113257e-05,
      "loss": 0.7808,
      "step": 5204
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.789363861083984,
      "learning_rate": 1.3252895752895755e-05,
      "loss": 0.7002,
      "step": 5205
    },
    {
      "epoch": 1.12,
      "grad_norm": 11.962141036987305,
      "learning_rate": 1.324967824967825e-05,
      "loss": 0.6416,
      "step": 5206
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.406078338623047,
      "learning_rate": 1.3246460746460746e-05,
      "loss": 0.6838,
      "step": 5207
    },
    {
      "epoch": 1.12,
      "grad_norm": 9.098557472229004,
      "learning_rate": 1.3243243243243242e-05,
      "loss": 0.5123,
      "step": 5208
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.873542785644531,
      "learning_rate": 1.324002574002574e-05,
      "loss": 0.6519,
      "step": 5209
    },
    {
      "epoch": 1.12,
      "grad_norm": 13.510058403015137,
      "learning_rate": 1.3236808236808237e-05,
      "loss": 0.6576,
      "step": 5210
    },
    {
      "epoch": 1.12,
      "grad_norm": 9.03515911102295,
      "learning_rate": 1.3233590733590733e-05,
      "loss": 0.5314,
      "step": 5211
    },
    {
      "epoch": 1.12,
      "grad_norm": 12.013127326965332,
      "learning_rate": 1.3230373230373232e-05,
      "loss": 0.6228,
      "step": 5212
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.75295639038086,
      "learning_rate": 1.3227155727155728e-05,
      "loss": 0.5023,
      "step": 5213
    },
    {
      "epoch": 1.12,
      "grad_norm": 12.54591178894043,
      "learning_rate": 1.3223938223938224e-05,
      "loss": 0.7592,
      "step": 5214
    },
    {
      "epoch": 1.12,
      "grad_norm": 11.296277046203613,
      "learning_rate": 1.3220720720720721e-05,
      "loss": 0.6598,
      "step": 5215
    },
    {
      "epoch": 1.12,
      "grad_norm": 14.787192344665527,
      "learning_rate": 1.3217503217503217e-05,
      "loss": 0.815,
      "step": 5216
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.921116828918457,
      "learning_rate": 1.3214285714285714e-05,
      "loss": 0.7077,
      "step": 5217
    },
    {
      "epoch": 1.12,
      "grad_norm": 14.022708892822266,
      "learning_rate": 1.3211068211068212e-05,
      "loss": 0.9659,
      "step": 5218
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.345845222473145,
      "learning_rate": 1.3207850707850708e-05,
      "loss": 0.5759,
      "step": 5219
    },
    {
      "epoch": 1.12,
      "grad_norm": 9.731948852539062,
      "learning_rate": 1.3204633204633205e-05,
      "loss": 0.6577,
      "step": 5220
    },
    {
      "epoch": 1.12,
      "grad_norm": 12.511012077331543,
      "learning_rate": 1.3201415701415701e-05,
      "loss": 0.768,
      "step": 5221
    },
    {
      "epoch": 1.12,
      "grad_norm": 11.004755020141602,
      "learning_rate": 1.31981981981982e-05,
      "loss": 0.5803,
      "step": 5222
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.515892028808594,
      "learning_rate": 1.3194980694980696e-05,
      "loss": 0.6679,
      "step": 5223
    },
    {
      "epoch": 1.12,
      "grad_norm": 11.19224739074707,
      "learning_rate": 1.319176319176319e-05,
      "loss": 0.571,
      "step": 5224
    },
    {
      "epoch": 1.12,
      "grad_norm": 14.18136215209961,
      "learning_rate": 1.3188545688545689e-05,
      "loss": 0.7283,
      "step": 5225
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.759422302246094,
      "learning_rate": 1.3185328185328185e-05,
      "loss": 0.8404,
      "step": 5226
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.467818260192871,
      "learning_rate": 1.3182110682110682e-05,
      "loss": 0.5658,
      "step": 5227
    },
    {
      "epoch": 1.12,
      "grad_norm": 12.244832992553711,
      "learning_rate": 1.317889317889318e-05,
      "loss": 0.8576,
      "step": 5228
    },
    {
      "epoch": 1.12,
      "grad_norm": 7.8604736328125,
      "learning_rate": 1.3175675675675676e-05,
      "loss": 0.4683,
      "step": 5229
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.159653663635254,
      "learning_rate": 1.3172458172458173e-05,
      "loss": 0.55,
      "step": 5230
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.516440391540527,
      "learning_rate": 1.3169240669240671e-05,
      "loss": 0.5896,
      "step": 5231
    },
    {
      "epoch": 1.12,
      "grad_norm": 7.529036045074463,
      "learning_rate": 1.3166023166023166e-05,
      "loss": 0.636,
      "step": 5232
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.613761901855469,
      "learning_rate": 1.3162805662805662e-05,
      "loss": 0.4772,
      "step": 5233
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.788703918457031,
      "learning_rate": 1.315958815958816e-05,
      "loss": 0.6301,
      "step": 5234
    },
    {
      "epoch": 1.12,
      "grad_norm": 11.347795486450195,
      "learning_rate": 1.3156370656370657e-05,
      "loss": 0.6908,
      "step": 5235
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.760001182556152,
      "learning_rate": 1.3153153153153153e-05,
      "loss": 0.7177,
      "step": 5236
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.893867492675781,
      "learning_rate": 1.314993564993565e-05,
      "loss": 0.7455,
      "step": 5237
    },
    {
      "epoch": 1.12,
      "grad_norm": 11.191086769104004,
      "learning_rate": 1.3146718146718148e-05,
      "loss": 0.7475,
      "step": 5238
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.238869667053223,
      "learning_rate": 1.3143500643500644e-05,
      "loss": 0.4777,
      "step": 5239
    },
    {
      "epoch": 1.12,
      "grad_norm": 6.667503833770752,
      "learning_rate": 1.314028314028314e-05,
      "loss": 0.3707,
      "step": 5240
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.389366149902344,
      "learning_rate": 1.3137065637065637e-05,
      "loss": 0.7082,
      "step": 5241
    },
    {
      "epoch": 1.12,
      "grad_norm": 9.813602447509766,
      "learning_rate": 1.3133848133848134e-05,
      "loss": 0.6124,
      "step": 5242
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.828285217285156,
      "learning_rate": 1.313063063063063e-05,
      "loss": 0.6911,
      "step": 5243
    },
    {
      "epoch": 1.12,
      "grad_norm": 12.090410232543945,
      "learning_rate": 1.3127413127413128e-05,
      "loss": 0.6611,
      "step": 5244
    },
    {
      "epoch": 1.13,
      "grad_norm": 8.969409942626953,
      "learning_rate": 1.3124195624195625e-05,
      "loss": 0.6279,
      "step": 5245
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.89619255065918,
      "learning_rate": 1.3120978120978121e-05,
      "loss": 0.6415,
      "step": 5246
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.328313827514648,
      "learning_rate": 1.311776061776062e-05,
      "loss": 0.6456,
      "step": 5247
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.061506271362305,
      "learning_rate": 1.3114543114543116e-05,
      "loss": 0.6534,
      "step": 5248
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.98448371887207,
      "learning_rate": 1.3111325611325612e-05,
      "loss": 0.8794,
      "step": 5249
    },
    {
      "epoch": 1.13,
      "grad_norm": 11.367032051086426,
      "learning_rate": 1.3108108108108109e-05,
      "loss": 0.6485,
      "step": 5250
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.456483840942383,
      "learning_rate": 1.3104890604890605e-05,
      "loss": 0.5501,
      "step": 5251
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.0925931930542,
      "learning_rate": 1.3101673101673102e-05,
      "loss": 0.4242,
      "step": 5252
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.21027946472168,
      "learning_rate": 1.3098455598455598e-05,
      "loss": 0.6477,
      "step": 5253
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.546984672546387,
      "learning_rate": 1.3095238095238096e-05,
      "loss": 0.677,
      "step": 5254
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.685811042785645,
      "learning_rate": 1.3092020592020593e-05,
      "loss": 0.7188,
      "step": 5255
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.524791717529297,
      "learning_rate": 1.3088803088803089e-05,
      "loss": 0.7188,
      "step": 5256
    },
    {
      "epoch": 1.13,
      "grad_norm": 8.646069526672363,
      "learning_rate": 1.3085585585585587e-05,
      "loss": 0.5872,
      "step": 5257
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.99842357635498,
      "learning_rate": 1.3082368082368082e-05,
      "loss": 0.7686,
      "step": 5258
    },
    {
      "epoch": 1.13,
      "grad_norm": 8.809260368347168,
      "learning_rate": 1.3079150579150578e-05,
      "loss": 0.361,
      "step": 5259
    },
    {
      "epoch": 1.13,
      "grad_norm": 12.99691390991211,
      "learning_rate": 1.3075933075933077e-05,
      "loss": 0.696,
      "step": 5260
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.662376403808594,
      "learning_rate": 1.3072715572715573e-05,
      "loss": 0.5929,
      "step": 5261
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.55571174621582,
      "learning_rate": 1.306949806949807e-05,
      "loss": 0.7703,
      "step": 5262
    },
    {
      "epoch": 1.13,
      "grad_norm": 12.369951248168945,
      "learning_rate": 1.3066280566280568e-05,
      "loss": 0.7916,
      "step": 5263
    },
    {
      "epoch": 1.13,
      "grad_norm": 8.612549781799316,
      "learning_rate": 1.3063063063063064e-05,
      "loss": 0.4689,
      "step": 5264
    },
    {
      "epoch": 1.13,
      "grad_norm": 11.466002464294434,
      "learning_rate": 1.305984555984556e-05,
      "loss": 0.761,
      "step": 5265
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.519682884216309,
      "learning_rate": 1.3056628056628057e-05,
      "loss": 0.4759,
      "step": 5266
    },
    {
      "epoch": 1.13,
      "grad_norm": 11.720043182373047,
      "learning_rate": 1.3053410553410553e-05,
      "loss": 0.7794,
      "step": 5267
    },
    {
      "epoch": 1.13,
      "grad_norm": 16.23905372619629,
      "learning_rate": 1.305019305019305e-05,
      "loss": 0.8135,
      "step": 5268
    },
    {
      "epoch": 1.13,
      "grad_norm": 27.39961051940918,
      "learning_rate": 1.3046975546975546e-05,
      "loss": 0.6355,
      "step": 5269
    },
    {
      "epoch": 1.13,
      "grad_norm": 7.056489944458008,
      "learning_rate": 1.3043758043758044e-05,
      "loss": 0.4723,
      "step": 5270
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.224895477294922,
      "learning_rate": 1.3040540540540541e-05,
      "loss": 0.6874,
      "step": 5271
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.462418556213379,
      "learning_rate": 1.3037323037323037e-05,
      "loss": 0.6652,
      "step": 5272
    },
    {
      "epoch": 1.13,
      "grad_norm": 11.50607967376709,
      "learning_rate": 1.3034105534105535e-05,
      "loss": 0.7411,
      "step": 5273
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.243553161621094,
      "learning_rate": 1.3030888030888032e-05,
      "loss": 0.4754,
      "step": 5274
    },
    {
      "epoch": 1.13,
      "grad_norm": 12.531919479370117,
      "learning_rate": 1.3027670527670528e-05,
      "loss": 0.6954,
      "step": 5275
    },
    {
      "epoch": 1.13,
      "grad_norm": 11.42068099975586,
      "learning_rate": 1.3024453024453025e-05,
      "loss": 0.7806,
      "step": 5276
    },
    {
      "epoch": 1.13,
      "grad_norm": 8.868173599243164,
      "learning_rate": 1.3021235521235521e-05,
      "loss": 0.5181,
      "step": 5277
    },
    {
      "epoch": 1.13,
      "grad_norm": 12.77989673614502,
      "learning_rate": 1.3018018018018018e-05,
      "loss": 0.5947,
      "step": 5278
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.812150001525879,
      "learning_rate": 1.3014800514800516e-05,
      "loss": 0.7323,
      "step": 5279
    },
    {
      "epoch": 1.13,
      "grad_norm": 13.690271377563477,
      "learning_rate": 1.3011583011583012e-05,
      "loss": 0.7292,
      "step": 5280
    },
    {
      "epoch": 1.13,
      "grad_norm": 12.52534294128418,
      "learning_rate": 1.3008365508365509e-05,
      "loss": 0.6895,
      "step": 5281
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.967046737670898,
      "learning_rate": 1.3005148005148005e-05,
      "loss": 0.5976,
      "step": 5282
    },
    {
      "epoch": 1.13,
      "grad_norm": 8.428667068481445,
      "learning_rate": 1.3001930501930503e-05,
      "loss": 0.5319,
      "step": 5283
    },
    {
      "epoch": 1.13,
      "grad_norm": 8.815603256225586,
      "learning_rate": 1.2998712998712998e-05,
      "loss": 0.6641,
      "step": 5284
    },
    {
      "epoch": 1.13,
      "grad_norm": 13.051180839538574,
      "learning_rate": 1.2995495495495495e-05,
      "loss": 1.0179,
      "step": 5285
    },
    {
      "epoch": 1.13,
      "grad_norm": 10.605595588684082,
      "learning_rate": 1.2992277992277993e-05,
      "loss": 0.8401,
      "step": 5286
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.008448600769043,
      "learning_rate": 1.298906048906049e-05,
      "loss": 0.7037,
      "step": 5287
    },
    {
      "epoch": 1.13,
      "grad_norm": 8.270719528198242,
      "learning_rate": 1.2985842985842986e-05,
      "loss": 0.5746,
      "step": 5288
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.402816772460938,
      "learning_rate": 1.2982625482625484e-05,
      "loss": 0.6063,
      "step": 5289
    },
    {
      "epoch": 1.13,
      "grad_norm": 12.113189697265625,
      "learning_rate": 1.297940797940798e-05,
      "loss": 0.9595,
      "step": 5290
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.947223663330078,
      "learning_rate": 1.2976190476190477e-05,
      "loss": 0.6786,
      "step": 5291
    },
    {
      "epoch": 1.14,
      "grad_norm": 12.224297523498535,
      "learning_rate": 1.2972972972972975e-05,
      "loss": 0.7071,
      "step": 5292
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.801689147949219,
      "learning_rate": 1.296975546975547e-05,
      "loss": 0.568,
      "step": 5293
    },
    {
      "epoch": 1.14,
      "grad_norm": 13.121309280395508,
      "learning_rate": 1.2966537966537966e-05,
      "loss": 0.8594,
      "step": 5294
    },
    {
      "epoch": 1.14,
      "grad_norm": 15.162259101867676,
      "learning_rate": 1.2963320463320462e-05,
      "loss": 0.7544,
      "step": 5295
    },
    {
      "epoch": 1.14,
      "grad_norm": 7.4019246101379395,
      "learning_rate": 1.296010296010296e-05,
      "loss": 0.5739,
      "step": 5296
    },
    {
      "epoch": 1.14,
      "grad_norm": 8.429885864257812,
      "learning_rate": 1.2956885456885457e-05,
      "loss": 0.5538,
      "step": 5297
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.91114330291748,
      "learning_rate": 1.2953667953667954e-05,
      "loss": 0.7715,
      "step": 5298
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.832133293151855,
      "learning_rate": 1.2950450450450452e-05,
      "loss": 0.7239,
      "step": 5299
    },
    {
      "epoch": 1.14,
      "grad_norm": 11.550043106079102,
      "learning_rate": 1.2947232947232948e-05,
      "loss": 0.7505,
      "step": 5300
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.5433349609375,
      "learning_rate": 1.2944015444015445e-05,
      "loss": 0.4919,
      "step": 5301
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.84937572479248,
      "learning_rate": 1.2940797940797941e-05,
      "loss": 0.8086,
      "step": 5302
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.579002380371094,
      "learning_rate": 1.2937580437580437e-05,
      "loss": 0.6444,
      "step": 5303
    },
    {
      "epoch": 1.14,
      "grad_norm": 8.268001556396484,
      "learning_rate": 1.2934362934362934e-05,
      "loss": 0.6541,
      "step": 5304
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.26300048828125,
      "learning_rate": 1.2931145431145432e-05,
      "loss": 0.4921,
      "step": 5305
    },
    {
      "epoch": 1.14,
      "grad_norm": 8.268767356872559,
      "learning_rate": 1.2927927927927929e-05,
      "loss": 0.4822,
      "step": 5306
    },
    {
      "epoch": 1.14,
      "grad_norm": 12.618170738220215,
      "learning_rate": 1.2924710424710425e-05,
      "loss": 1.1172,
      "step": 5307
    },
    {
      "epoch": 1.14,
      "grad_norm": 13.78017520904541,
      "learning_rate": 1.2921492921492923e-05,
      "loss": 0.7491,
      "step": 5308
    },
    {
      "epoch": 1.14,
      "grad_norm": 11.680254936218262,
      "learning_rate": 1.291827541827542e-05,
      "loss": 0.5766,
      "step": 5309
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.730586051940918,
      "learning_rate": 1.2915057915057914e-05,
      "loss": 0.6836,
      "step": 5310
    },
    {
      "epoch": 1.14,
      "grad_norm": 8.6458740234375,
      "learning_rate": 1.291184041184041e-05,
      "loss": 0.4929,
      "step": 5311
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.960963249206543,
      "learning_rate": 1.2908622908622909e-05,
      "loss": 0.6539,
      "step": 5312
    },
    {
      "epoch": 1.14,
      "grad_norm": 7.687562465667725,
      "learning_rate": 1.2905405405405405e-05,
      "loss": 0.5236,
      "step": 5313
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.923574447631836,
      "learning_rate": 1.2902187902187902e-05,
      "loss": 0.5323,
      "step": 5314
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.450752258300781,
      "learning_rate": 1.28989703989704e-05,
      "loss": 0.5481,
      "step": 5315
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.702336311340332,
      "learning_rate": 1.2895752895752896e-05,
      "loss": 0.542,
      "step": 5316
    },
    {
      "epoch": 1.14,
      "grad_norm": 11.01758098602295,
      "learning_rate": 1.2892535392535393e-05,
      "loss": 0.6204,
      "step": 5317
    },
    {
      "epoch": 1.14,
      "grad_norm": 13.025856971740723,
      "learning_rate": 1.2889317889317891e-05,
      "loss": 0.8168,
      "step": 5318
    },
    {
      "epoch": 1.14,
      "grad_norm": 8.07052993774414,
      "learning_rate": 1.2886100386100386e-05,
      "loss": 0.6163,
      "step": 5319
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.907517433166504,
      "learning_rate": 1.2882882882882882e-05,
      "loss": 0.7364,
      "step": 5320
    },
    {
      "epoch": 1.14,
      "grad_norm": 12.775591850280762,
      "learning_rate": 1.287966537966538e-05,
      "loss": 1.058,
      "step": 5321
    },
    {
      "epoch": 1.14,
      "grad_norm": 8.640970230102539,
      "learning_rate": 1.2876447876447877e-05,
      "loss": 0.4664,
      "step": 5322
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.405670166015625,
      "learning_rate": 1.2873230373230373e-05,
      "loss": 0.5583,
      "step": 5323
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.676505088806152,
      "learning_rate": 1.287001287001287e-05,
      "loss": 0.5783,
      "step": 5324
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.093316078186035,
      "learning_rate": 1.2866795366795368e-05,
      "loss": 0.7827,
      "step": 5325
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.957427024841309,
      "learning_rate": 1.2863577863577864e-05,
      "loss": 0.6825,
      "step": 5326
    },
    {
      "epoch": 1.14,
      "grad_norm": 11.984309196472168,
      "learning_rate": 1.286036036036036e-05,
      "loss": 0.5863,
      "step": 5327
    },
    {
      "epoch": 1.14,
      "grad_norm": 8.046833038330078,
      "learning_rate": 1.2857142857142857e-05,
      "loss": 0.4576,
      "step": 5328
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.61512279510498,
      "learning_rate": 1.2853925353925354e-05,
      "loss": 0.6214,
      "step": 5329
    },
    {
      "epoch": 1.14,
      "grad_norm": 11.247774124145508,
      "learning_rate": 1.285070785070785e-05,
      "loss": 0.7871,
      "step": 5330
    },
    {
      "epoch": 1.14,
      "grad_norm": 9.95553207397461,
      "learning_rate": 1.2847490347490348e-05,
      "loss": 0.7612,
      "step": 5331
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.12469482421875,
      "learning_rate": 1.2844272844272845e-05,
      "loss": 0.5352,
      "step": 5332
    },
    {
      "epoch": 1.14,
      "grad_norm": 8.951988220214844,
      "learning_rate": 1.2841055341055341e-05,
      "loss": 0.4325,
      "step": 5333
    },
    {
      "epoch": 1.14,
      "grad_norm": 7.765128135681152,
      "learning_rate": 1.283783783783784e-05,
      "loss": 0.6548,
      "step": 5334
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.32986831665039,
      "learning_rate": 1.2834620334620336e-05,
      "loss": 0.6299,
      "step": 5335
    },
    {
      "epoch": 1.14,
      "grad_norm": 12.910330772399902,
      "learning_rate": 1.2831402831402832e-05,
      "loss": 0.8107,
      "step": 5336
    },
    {
      "epoch": 1.14,
      "grad_norm": 10.518227577209473,
      "learning_rate": 1.2828185328185329e-05,
      "loss": 0.7124,
      "step": 5337
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.881951332092285,
      "learning_rate": 1.2824967824967825e-05,
      "loss": 0.72,
      "step": 5338
    },
    {
      "epoch": 1.15,
      "grad_norm": 8.178609848022461,
      "learning_rate": 1.2821750321750322e-05,
      "loss": 0.4802,
      "step": 5339
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.581007957458496,
      "learning_rate": 1.2818532818532818e-05,
      "loss": 0.516,
      "step": 5340
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.202343940734863,
      "learning_rate": 1.2815315315315316e-05,
      "loss": 0.5147,
      "step": 5341
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.518924713134766,
      "learning_rate": 1.2812097812097813e-05,
      "loss": 0.6717,
      "step": 5342
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.389345169067383,
      "learning_rate": 1.2808880308880309e-05,
      "loss": 0.5324,
      "step": 5343
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.303927421569824,
      "learning_rate": 1.2805662805662807e-05,
      "loss": 0.7139,
      "step": 5344
    },
    {
      "epoch": 1.15,
      "grad_norm": 8.730359077453613,
      "learning_rate": 1.2802445302445302e-05,
      "loss": 0.5142,
      "step": 5345
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.068974494934082,
      "learning_rate": 1.2799227799227798e-05,
      "loss": 0.6992,
      "step": 5346
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.794787406921387,
      "learning_rate": 1.2796010296010297e-05,
      "loss": 0.6864,
      "step": 5347
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.028524398803711,
      "learning_rate": 1.2792792792792793e-05,
      "loss": 0.5716,
      "step": 5348
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.254591941833496,
      "learning_rate": 1.278957528957529e-05,
      "loss": 0.5451,
      "step": 5349
    },
    {
      "epoch": 1.15,
      "grad_norm": 8.966127395629883,
      "learning_rate": 1.2786357786357788e-05,
      "loss": 0.7306,
      "step": 5350
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.514725685119629,
      "learning_rate": 1.2783140283140284e-05,
      "loss": 0.6119,
      "step": 5351
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.036415100097656,
      "learning_rate": 1.277992277992278e-05,
      "loss": 0.5255,
      "step": 5352
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.451563835144043,
      "learning_rate": 1.2776705276705277e-05,
      "loss": 0.5104,
      "step": 5353
    },
    {
      "epoch": 1.15,
      "grad_norm": 8.733838081359863,
      "learning_rate": 1.2773487773487773e-05,
      "loss": 0.5646,
      "step": 5354
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.1891508102417,
      "learning_rate": 1.277027027027027e-05,
      "loss": 0.4967,
      "step": 5355
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.955782890319824,
      "learning_rate": 1.2767052767052766e-05,
      "loss": 1.0368,
      "step": 5356
    },
    {
      "epoch": 1.15,
      "grad_norm": 13.351094245910645,
      "learning_rate": 1.2763835263835264e-05,
      "loss": 0.7218,
      "step": 5357
    },
    {
      "epoch": 1.15,
      "grad_norm": 8.678389549255371,
      "learning_rate": 1.2760617760617761e-05,
      "loss": 0.4349,
      "step": 5358
    },
    {
      "epoch": 1.15,
      "grad_norm": 8.855202674865723,
      "learning_rate": 1.2757400257400257e-05,
      "loss": 0.6075,
      "step": 5359
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.017817497253418,
      "learning_rate": 1.2754182754182756e-05,
      "loss": 0.4929,
      "step": 5360
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.03504467010498,
      "learning_rate": 1.2750965250965252e-05,
      "loss": 0.8487,
      "step": 5361
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.93243408203125,
      "learning_rate": 1.2747747747747748e-05,
      "loss": 0.6272,
      "step": 5362
    },
    {
      "epoch": 1.15,
      "grad_norm": 7.137197017669678,
      "learning_rate": 1.2744530244530245e-05,
      "loss": 0.4328,
      "step": 5363
    },
    {
      "epoch": 1.15,
      "grad_norm": 12.209068298339844,
      "learning_rate": 1.2741312741312741e-05,
      "loss": 0.5459,
      "step": 5364
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.822569847106934,
      "learning_rate": 1.2738095238095238e-05,
      "loss": 0.5678,
      "step": 5365
    },
    {
      "epoch": 1.15,
      "grad_norm": 12.22994613647461,
      "learning_rate": 1.2734877734877736e-05,
      "loss": 1.0187,
      "step": 5366
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.291309356689453,
      "learning_rate": 1.2731660231660232e-05,
      "loss": 0.7662,
      "step": 5367
    },
    {
      "epoch": 1.15,
      "grad_norm": 13.999427795410156,
      "learning_rate": 1.2728442728442729e-05,
      "loss": 0.879,
      "step": 5368
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.544657707214355,
      "learning_rate": 1.2725225225225225e-05,
      "loss": 0.6509,
      "step": 5369
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.062723159790039,
      "learning_rate": 1.2722007722007723e-05,
      "loss": 0.5511,
      "step": 5370
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.22580623626709,
      "learning_rate": 1.2718790218790218e-05,
      "loss": 0.6049,
      "step": 5371
    },
    {
      "epoch": 1.15,
      "grad_norm": 8.776140213012695,
      "learning_rate": 1.2715572715572715e-05,
      "loss": 0.4013,
      "step": 5372
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.098968505859375,
      "learning_rate": 1.2712355212355213e-05,
      "loss": 0.6178,
      "step": 5373
    },
    {
      "epoch": 1.15,
      "grad_norm": 8.412960052490234,
      "learning_rate": 1.270913770913771e-05,
      "loss": 0.4923,
      "step": 5374
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.94229793548584,
      "learning_rate": 1.2705920205920206e-05,
      "loss": 0.7678,
      "step": 5375
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.267694473266602,
      "learning_rate": 1.2702702702702704e-05,
      "loss": 0.6239,
      "step": 5376
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.873290061950684,
      "learning_rate": 1.26994851994852e-05,
      "loss": 0.6971,
      "step": 5377
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.215874671936035,
      "learning_rate": 1.2696267696267697e-05,
      "loss": 0.5169,
      "step": 5378
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.897562026977539,
      "learning_rate": 1.2693050193050195e-05,
      "loss": 0.7005,
      "step": 5379
    },
    {
      "epoch": 1.15,
      "grad_norm": 12.563162803649902,
      "learning_rate": 1.268983268983269e-05,
      "loss": 0.4416,
      "step": 5380
    },
    {
      "epoch": 1.15,
      "grad_norm": 10.362838745117188,
      "learning_rate": 1.2686615186615186e-05,
      "loss": 0.7058,
      "step": 5381
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.54654598236084,
      "learning_rate": 1.2683397683397683e-05,
      "loss": 0.7966,
      "step": 5382
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.389301300048828,
      "learning_rate": 1.268018018018018e-05,
      "loss": 0.6143,
      "step": 5383
    },
    {
      "epoch": 1.15,
      "grad_norm": 8.935440063476562,
      "learning_rate": 1.2676962676962677e-05,
      "loss": 0.4564,
      "step": 5384
    },
    {
      "epoch": 1.16,
      "grad_norm": 12.740096092224121,
      "learning_rate": 1.2673745173745174e-05,
      "loss": 0.7127,
      "step": 5385
    },
    {
      "epoch": 1.16,
      "grad_norm": 11.945547103881836,
      "learning_rate": 1.2670527670527672e-05,
      "loss": 0.4973,
      "step": 5386
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.018773078918457,
      "learning_rate": 1.2667310167310168e-05,
      "loss": 0.6013,
      "step": 5387
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.889655113220215,
      "learning_rate": 1.2664092664092665e-05,
      "loss": 0.4698,
      "step": 5388
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.802403450012207,
      "learning_rate": 1.2660875160875161e-05,
      "loss": 0.7043,
      "step": 5389
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.403632164001465,
      "learning_rate": 1.2657657657657658e-05,
      "loss": 0.6576,
      "step": 5390
    },
    {
      "epoch": 1.16,
      "grad_norm": 11.78947639465332,
      "learning_rate": 1.2654440154440154e-05,
      "loss": 0.7207,
      "step": 5391
    },
    {
      "epoch": 1.16,
      "grad_norm": 12.63164234161377,
      "learning_rate": 1.2651222651222652e-05,
      "loss": 0.6276,
      "step": 5392
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.502755165100098,
      "learning_rate": 1.2648005148005149e-05,
      "loss": 0.5552,
      "step": 5393
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.551308631896973,
      "learning_rate": 1.2644787644787645e-05,
      "loss": 0.6147,
      "step": 5394
    },
    {
      "epoch": 1.16,
      "grad_norm": 11.020331382751465,
      "learning_rate": 1.2641570141570143e-05,
      "loss": 0.7201,
      "step": 5395
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.182271957397461,
      "learning_rate": 1.263835263835264e-05,
      "loss": 0.6902,
      "step": 5396
    },
    {
      "epoch": 1.16,
      "grad_norm": 12.03585147857666,
      "learning_rate": 1.2635135135135134e-05,
      "loss": 0.5737,
      "step": 5397
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.439638137817383,
      "learning_rate": 1.263191763191763e-05,
      "loss": 0.7263,
      "step": 5398
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.69489860534668,
      "learning_rate": 1.2628700128700129e-05,
      "loss": 0.5813,
      "step": 5399
    },
    {
      "epoch": 1.16,
      "grad_norm": 8.138993263244629,
      "learning_rate": 1.2625482625482625e-05,
      "loss": 0.363,
      "step": 5400
    },
    {
      "epoch": 1.16,
      "grad_norm": 12.902817726135254,
      "learning_rate": 1.2622265122265122e-05,
      "loss": 0.799,
      "step": 5401
    },
    {
      "epoch": 1.16,
      "grad_norm": 11.214788436889648,
      "learning_rate": 1.261904761904762e-05,
      "loss": 0.736,
      "step": 5402
    },
    {
      "epoch": 1.16,
      "grad_norm": 11.217241287231445,
      "learning_rate": 1.2615830115830116e-05,
      "loss": 0.7309,
      "step": 5403
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.324155807495117,
      "learning_rate": 1.2612612612612613e-05,
      "loss": 0.5738,
      "step": 5404
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.227046012878418,
      "learning_rate": 1.2609395109395111e-05,
      "loss": 0.7118,
      "step": 5405
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.62351131439209,
      "learning_rate": 1.2606177606177606e-05,
      "loss": 0.6166,
      "step": 5406
    },
    {
      "epoch": 1.16,
      "grad_norm": 8.740362167358398,
      "learning_rate": 1.2602960102960102e-05,
      "loss": 0.6911,
      "step": 5407
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.423748970031738,
      "learning_rate": 1.25997425997426e-05,
      "loss": 0.6258,
      "step": 5408
    },
    {
      "epoch": 1.16,
      "grad_norm": 15.596755027770996,
      "learning_rate": 1.2596525096525097e-05,
      "loss": 0.7105,
      "step": 5409
    },
    {
      "epoch": 1.16,
      "grad_norm": 14.881957054138184,
      "learning_rate": 1.2593307593307593e-05,
      "loss": 0.6107,
      "step": 5410
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.438093185424805,
      "learning_rate": 1.259009009009009e-05,
      "loss": 0.7204,
      "step": 5411
    },
    {
      "epoch": 1.16,
      "grad_norm": 11.619977951049805,
      "learning_rate": 1.2586872586872588e-05,
      "loss": 0.8412,
      "step": 5412
    },
    {
      "epoch": 1.16,
      "grad_norm": 8.772790908813477,
      "learning_rate": 1.2583655083655084e-05,
      "loss": 0.37,
      "step": 5413
    },
    {
      "epoch": 1.16,
      "grad_norm": 11.135663986206055,
      "learning_rate": 1.258043758043758e-05,
      "loss": 0.5207,
      "step": 5414
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.649925231933594,
      "learning_rate": 1.2577220077220077e-05,
      "loss": 0.7436,
      "step": 5415
    },
    {
      "epoch": 1.16,
      "grad_norm": 12.243144989013672,
      "learning_rate": 1.2574002574002574e-05,
      "loss": 0.6333,
      "step": 5416
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.278109550476074,
      "learning_rate": 1.257078507078507e-05,
      "loss": 0.654,
      "step": 5417
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.027420997619629,
      "learning_rate": 1.2567567567567568e-05,
      "loss": 0.6841,
      "step": 5418
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.718703269958496,
      "learning_rate": 1.2564350064350065e-05,
      "loss": 0.6098,
      "step": 5419
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.450200080871582,
      "learning_rate": 1.2561132561132561e-05,
      "loss": 0.7622,
      "step": 5420
    },
    {
      "epoch": 1.16,
      "grad_norm": 8.370036125183105,
      "learning_rate": 1.255791505791506e-05,
      "loss": 0.6266,
      "step": 5421
    },
    {
      "epoch": 1.16,
      "grad_norm": 7.168830871582031,
      "learning_rate": 1.2554697554697556e-05,
      "loss": 0.4663,
      "step": 5422
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.348474502563477,
      "learning_rate": 1.2551480051480052e-05,
      "loss": 0.7299,
      "step": 5423
    },
    {
      "epoch": 1.16,
      "grad_norm": 11.28630256652832,
      "learning_rate": 1.2548262548262549e-05,
      "loss": 0.5842,
      "step": 5424
    },
    {
      "epoch": 1.16,
      "grad_norm": 8.870391845703125,
      "learning_rate": 1.2545045045045045e-05,
      "loss": 0.4519,
      "step": 5425
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.60246467590332,
      "learning_rate": 1.2541827541827542e-05,
      "loss": 0.5002,
      "step": 5426
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.04152774810791,
      "learning_rate": 1.2538610038610038e-05,
      "loss": 0.7855,
      "step": 5427
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.650856018066406,
      "learning_rate": 1.2535392535392536e-05,
      "loss": 0.8136,
      "step": 5428
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.92136287689209,
      "learning_rate": 1.2532175032175033e-05,
      "loss": 0.7801,
      "step": 5429
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.669354438781738,
      "learning_rate": 1.2528957528957529e-05,
      "loss": 0.5692,
      "step": 5430
    },
    {
      "epoch": 1.16,
      "grad_norm": 13.173805236816406,
      "learning_rate": 1.2525740025740027e-05,
      "loss": 0.6344,
      "step": 5431
    },
    {
      "epoch": 1.17,
      "grad_norm": 7.776498794555664,
      "learning_rate": 1.2522522522522522e-05,
      "loss": 0.6255,
      "step": 5432
    },
    {
      "epoch": 1.17,
      "grad_norm": 8.40621280670166,
      "learning_rate": 1.2519305019305019e-05,
      "loss": 0.5003,
      "step": 5433
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.352524757385254,
      "learning_rate": 1.2516087516087517e-05,
      "loss": 0.6666,
      "step": 5434
    },
    {
      "epoch": 1.17,
      "grad_norm": 11.167640686035156,
      "learning_rate": 1.2512870012870013e-05,
      "loss": 0.7996,
      "step": 5435
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.183646202087402,
      "learning_rate": 1.250965250965251e-05,
      "loss": 0.5482,
      "step": 5436
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.107569694519043,
      "learning_rate": 1.2506435006435008e-05,
      "loss": 0.5802,
      "step": 5437
    },
    {
      "epoch": 1.17,
      "grad_norm": 8.434996604919434,
      "learning_rate": 1.2503217503217504e-05,
      "loss": 0.686,
      "step": 5438
    },
    {
      "epoch": 1.17,
      "grad_norm": 12.374134063720703,
      "learning_rate": 1.25e-05,
      "loss": 0.6009,
      "step": 5439
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.358153343200684,
      "learning_rate": 1.2496782496782497e-05,
      "loss": 0.5624,
      "step": 5440
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.377528190612793,
      "learning_rate": 1.2493564993564994e-05,
      "loss": 0.6209,
      "step": 5441
    },
    {
      "epoch": 1.17,
      "grad_norm": 12.229966163635254,
      "learning_rate": 1.249034749034749e-05,
      "loss": 0.7983,
      "step": 5442
    },
    {
      "epoch": 1.17,
      "grad_norm": 8.320786476135254,
      "learning_rate": 1.2487129987129986e-05,
      "loss": 0.4734,
      "step": 5443
    },
    {
      "epoch": 1.17,
      "grad_norm": 8.868268013000488,
      "learning_rate": 1.2483912483912485e-05,
      "loss": 0.4695,
      "step": 5444
    },
    {
      "epoch": 1.17,
      "grad_norm": 8.835224151611328,
      "learning_rate": 1.2480694980694981e-05,
      "loss": 0.5282,
      "step": 5445
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.493093490600586,
      "learning_rate": 1.2477477477477477e-05,
      "loss": 0.672,
      "step": 5446
    },
    {
      "epoch": 1.17,
      "grad_norm": 11.4345121383667,
      "learning_rate": 1.2474259974259976e-05,
      "loss": 0.655,
      "step": 5447
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.05898380279541,
      "learning_rate": 1.2471042471042472e-05,
      "loss": 0.7027,
      "step": 5448
    },
    {
      "epoch": 1.17,
      "grad_norm": 12.58997631072998,
      "learning_rate": 1.2467824967824969e-05,
      "loss": 0.7294,
      "step": 5449
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.142993927001953,
      "learning_rate": 1.2464607464607465e-05,
      "loss": 0.6488,
      "step": 5450
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.430173873901367,
      "learning_rate": 1.2461389961389961e-05,
      "loss": 0.6106,
      "step": 5451
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.883241653442383,
      "learning_rate": 1.2458172458172458e-05,
      "loss": 0.6927,
      "step": 5452
    },
    {
      "epoch": 1.17,
      "grad_norm": 8.436565399169922,
      "learning_rate": 1.2454954954954956e-05,
      "loss": 0.5491,
      "step": 5453
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.658500671386719,
      "learning_rate": 1.2451737451737452e-05,
      "loss": 0.5002,
      "step": 5454
    },
    {
      "epoch": 1.17,
      "grad_norm": 13.940569877624512,
      "learning_rate": 1.2448519948519949e-05,
      "loss": 0.7275,
      "step": 5455
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.676763534545898,
      "learning_rate": 1.2445302445302445e-05,
      "loss": 0.5525,
      "step": 5456
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.052932739257812,
      "learning_rate": 1.2442084942084944e-05,
      "loss": 0.4783,
      "step": 5457
    },
    {
      "epoch": 1.17,
      "grad_norm": 13.222050666809082,
      "learning_rate": 1.2438867438867438e-05,
      "loss": 0.6972,
      "step": 5458
    },
    {
      "epoch": 1.17,
      "grad_norm": 11.811882972717285,
      "learning_rate": 1.2435649935649935e-05,
      "loss": 0.7795,
      "step": 5459
    },
    {
      "epoch": 1.17,
      "grad_norm": 15.42492389678955,
      "learning_rate": 1.2432432432432433e-05,
      "loss": 0.6694,
      "step": 5460
    },
    {
      "epoch": 1.17,
      "grad_norm": 7.39992618560791,
      "learning_rate": 1.242921492921493e-05,
      "loss": 0.4855,
      "step": 5461
    },
    {
      "epoch": 1.17,
      "grad_norm": 14.401289939880371,
      "learning_rate": 1.2425997425997426e-05,
      "loss": 0.8662,
      "step": 5462
    },
    {
      "epoch": 1.17,
      "grad_norm": 12.694990158081055,
      "learning_rate": 1.2422779922779924e-05,
      "loss": 0.9246,
      "step": 5463
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.869559288024902,
      "learning_rate": 1.241956241956242e-05,
      "loss": 0.6041,
      "step": 5464
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.598673820495605,
      "learning_rate": 1.2416344916344917e-05,
      "loss": 0.7365,
      "step": 5465
    },
    {
      "epoch": 1.17,
      "grad_norm": 12.349896430969238,
      "learning_rate": 1.2413127413127415e-05,
      "loss": 0.6281,
      "step": 5466
    },
    {
      "epoch": 1.17,
      "grad_norm": 11.535269737243652,
      "learning_rate": 1.240990990990991e-05,
      "loss": 0.7266,
      "step": 5467
    },
    {
      "epoch": 1.17,
      "grad_norm": 8.90776252746582,
      "learning_rate": 1.2406692406692406e-05,
      "loss": 0.5141,
      "step": 5468
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.572636604309082,
      "learning_rate": 1.2403474903474903e-05,
      "loss": 0.5734,
      "step": 5469
    },
    {
      "epoch": 1.17,
      "grad_norm": 11.72801685333252,
      "learning_rate": 1.24002574002574e-05,
      "loss": 0.8079,
      "step": 5470
    },
    {
      "epoch": 1.17,
      "grad_norm": 13.965339660644531,
      "learning_rate": 1.2397039897039897e-05,
      "loss": 0.8307,
      "step": 5471
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.537467956542969,
      "learning_rate": 1.2393822393822394e-05,
      "loss": 0.5871,
      "step": 5472
    },
    {
      "epoch": 1.17,
      "grad_norm": 13.057709693908691,
      "learning_rate": 1.2390604890604892e-05,
      "loss": 0.6614,
      "step": 5473
    },
    {
      "epoch": 1.17,
      "grad_norm": 12.498886108398438,
      "learning_rate": 1.2387387387387388e-05,
      "loss": 0.7238,
      "step": 5474
    },
    {
      "epoch": 1.17,
      "grad_norm": 12.050716400146484,
      "learning_rate": 1.2384169884169885e-05,
      "loss": 0.6759,
      "step": 5475
    },
    {
      "epoch": 1.17,
      "grad_norm": 10.45885944366455,
      "learning_rate": 1.2380952380952381e-05,
      "loss": 0.5804,
      "step": 5476
    },
    {
      "epoch": 1.17,
      "grad_norm": 13.23482608795166,
      "learning_rate": 1.2377734877734878e-05,
      "loss": 0.7461,
      "step": 5477
    },
    {
      "epoch": 1.18,
      "grad_norm": 11.02562141418457,
      "learning_rate": 1.2374517374517374e-05,
      "loss": 0.6415,
      "step": 5478
    },
    {
      "epoch": 1.18,
      "grad_norm": 8.37654972076416,
      "learning_rate": 1.2371299871299872e-05,
      "loss": 0.3686,
      "step": 5479
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.617210388183594,
      "learning_rate": 1.2368082368082369e-05,
      "loss": 0.7424,
      "step": 5480
    },
    {
      "epoch": 1.18,
      "grad_norm": 11.21973705291748,
      "learning_rate": 1.2364864864864865e-05,
      "loss": 0.716,
      "step": 5481
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.231231689453125,
      "learning_rate": 1.2361647361647363e-05,
      "loss": 0.5576,
      "step": 5482
    },
    {
      "epoch": 1.18,
      "grad_norm": 8.589138984680176,
      "learning_rate": 1.235842985842986e-05,
      "loss": 0.4509,
      "step": 5483
    },
    {
      "epoch": 1.18,
      "grad_norm": 11.37207317352295,
      "learning_rate": 1.2355212355212354e-05,
      "loss": 0.9349,
      "step": 5484
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.55196762084961,
      "learning_rate": 1.2351994851994851e-05,
      "loss": 0.532,
      "step": 5485
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.334421157836914,
      "learning_rate": 1.2348777348777349e-05,
      "loss": 0.6604,
      "step": 5486
    },
    {
      "epoch": 1.18,
      "grad_norm": 14.8372163772583,
      "learning_rate": 1.2345559845559846e-05,
      "loss": 0.8144,
      "step": 5487
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.17016887664795,
      "learning_rate": 1.2342342342342342e-05,
      "loss": 0.6474,
      "step": 5488
    },
    {
      "epoch": 1.18,
      "grad_norm": 11.382461547851562,
      "learning_rate": 1.233912483912484e-05,
      "loss": 0.5687,
      "step": 5489
    },
    {
      "epoch": 1.18,
      "grad_norm": 8.626668930053711,
      "learning_rate": 1.2335907335907337e-05,
      "loss": 0.5898,
      "step": 5490
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.944326400756836,
      "learning_rate": 1.2332689832689833e-05,
      "loss": 0.5991,
      "step": 5491
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.740865707397461,
      "learning_rate": 1.2329472329472331e-05,
      "loss": 0.6975,
      "step": 5492
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.665614128112793,
      "learning_rate": 1.2326254826254826e-05,
      "loss": 0.5915,
      "step": 5493
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.370153427124023,
      "learning_rate": 1.2323037323037322e-05,
      "loss": 0.5333,
      "step": 5494
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.389497756958008,
      "learning_rate": 1.231981981981982e-05,
      "loss": 0.7255,
      "step": 5495
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.343724250793457,
      "learning_rate": 1.2316602316602317e-05,
      "loss": 0.4837,
      "step": 5496
    },
    {
      "epoch": 1.18,
      "grad_norm": 11.491949081420898,
      "learning_rate": 1.2313384813384813e-05,
      "loss": 0.5925,
      "step": 5497
    },
    {
      "epoch": 1.18,
      "grad_norm": 13.011469841003418,
      "learning_rate": 1.231016731016731e-05,
      "loss": 0.5607,
      "step": 5498
    },
    {
      "epoch": 1.18,
      "grad_norm": 12.068733215332031,
      "learning_rate": 1.2306949806949808e-05,
      "loss": 0.5537,
      "step": 5499
    },
    {
      "epoch": 1.18,
      "grad_norm": 13.307693481445312,
      "learning_rate": 1.2303732303732304e-05,
      "loss": 0.6291,
      "step": 5500
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.710280418395996,
      "learning_rate": 1.2300514800514801e-05,
      "loss": 0.7169,
      "step": 5501
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.648934364318848,
      "learning_rate": 1.2297297297297297e-05,
      "loss": 0.5959,
      "step": 5502
    },
    {
      "epoch": 1.18,
      "grad_norm": 14.893375396728516,
      "learning_rate": 1.2294079794079794e-05,
      "loss": 0.7591,
      "step": 5503
    },
    {
      "epoch": 1.18,
      "grad_norm": 12.508611679077148,
      "learning_rate": 1.229086229086229e-05,
      "loss": 0.8191,
      "step": 5504
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.499842643737793,
      "learning_rate": 1.2287644787644788e-05,
      "loss": 0.3489,
      "step": 5505
    },
    {
      "epoch": 1.18,
      "grad_norm": 11.186635971069336,
      "learning_rate": 1.2284427284427285e-05,
      "loss": 0.8743,
      "step": 5506
    },
    {
      "epoch": 1.18,
      "grad_norm": 11.72474193572998,
      "learning_rate": 1.2281209781209781e-05,
      "loss": 0.7471,
      "step": 5507
    },
    {
      "epoch": 1.18,
      "grad_norm": 11.907670974731445,
      "learning_rate": 1.227799227799228e-05,
      "loss": 0.5509,
      "step": 5508
    },
    {
      "epoch": 1.18,
      "grad_norm": 11.251079559326172,
      "learning_rate": 1.2274774774774776e-05,
      "loss": 0.5477,
      "step": 5509
    },
    {
      "epoch": 1.18,
      "grad_norm": 11.33117961883545,
      "learning_rate": 1.2271557271557272e-05,
      "loss": 0.7594,
      "step": 5510
    },
    {
      "epoch": 1.18,
      "grad_norm": 12.775083541870117,
      "learning_rate": 1.2268339768339769e-05,
      "loss": 0.894,
      "step": 5511
    },
    {
      "epoch": 1.18,
      "grad_norm": 8.141056060791016,
      "learning_rate": 1.2265122265122265e-05,
      "loss": 0.3984,
      "step": 5512
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.889702796936035,
      "learning_rate": 1.2261904761904762e-05,
      "loss": 0.6255,
      "step": 5513
    },
    {
      "epoch": 1.18,
      "grad_norm": 11.191764831542969,
      "learning_rate": 1.2258687258687258e-05,
      "loss": 0.8027,
      "step": 5514
    },
    {
      "epoch": 1.18,
      "grad_norm": 8.190975189208984,
      "learning_rate": 1.2255469755469756e-05,
      "loss": 0.482,
      "step": 5515
    },
    {
      "epoch": 1.18,
      "grad_norm": 8.758176803588867,
      "learning_rate": 1.2252252252252253e-05,
      "loss": 0.7135,
      "step": 5516
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.461224555969238,
      "learning_rate": 1.224903474903475e-05,
      "loss": 0.5895,
      "step": 5517
    },
    {
      "epoch": 1.18,
      "grad_norm": 10.198875427246094,
      "learning_rate": 1.2245817245817247e-05,
      "loss": 0.6404,
      "step": 5518
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.533930778503418,
      "learning_rate": 1.2242599742599742e-05,
      "loss": 0.6573,
      "step": 5519
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.935884475708008,
      "learning_rate": 1.2239382239382239e-05,
      "loss": 0.649,
      "step": 5520
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.742406845092773,
      "learning_rate": 1.2236164736164737e-05,
      "loss": 0.5731,
      "step": 5521
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.023987770080566,
      "learning_rate": 1.2232947232947233e-05,
      "loss": 0.664,
      "step": 5522
    },
    {
      "epoch": 1.18,
      "grad_norm": 13.488860130310059,
      "learning_rate": 1.222972972972973e-05,
      "loss": 0.8836,
      "step": 5523
    },
    {
      "epoch": 1.18,
      "grad_norm": 11.284333229064941,
      "learning_rate": 1.2226512226512228e-05,
      "loss": 0.6568,
      "step": 5524
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.396317481994629,
      "learning_rate": 1.2223294723294724e-05,
      "loss": 0.8562,
      "step": 5525
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.77694320678711,
      "learning_rate": 1.222007722007722e-05,
      "loss": 0.9217,
      "step": 5526
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.429715156555176,
      "learning_rate": 1.2216859716859717e-05,
      "loss": 0.4455,
      "step": 5527
    },
    {
      "epoch": 1.19,
      "grad_norm": 12.906596183776855,
      "learning_rate": 1.2213642213642214e-05,
      "loss": 0.7389,
      "step": 5528
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.500326156616211,
      "learning_rate": 1.221042471042471e-05,
      "loss": 0.758,
      "step": 5529
    },
    {
      "epoch": 1.19,
      "grad_norm": 12.45553970336914,
      "learning_rate": 1.2207207207207206e-05,
      "loss": 0.6267,
      "step": 5530
    },
    {
      "epoch": 1.19,
      "grad_norm": 8.645438194274902,
      "learning_rate": 1.2203989703989705e-05,
      "loss": 0.5374,
      "step": 5531
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.416162490844727,
      "learning_rate": 1.2200772200772201e-05,
      "loss": 0.5291,
      "step": 5532
    },
    {
      "epoch": 1.19,
      "grad_norm": 8.737247467041016,
      "learning_rate": 1.2197554697554698e-05,
      "loss": 0.6179,
      "step": 5533
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.355081558227539,
      "learning_rate": 1.2194337194337196e-05,
      "loss": 0.7009,
      "step": 5534
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.161620140075684,
      "learning_rate": 1.2191119691119692e-05,
      "loss": 0.6666,
      "step": 5535
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.889230728149414,
      "learning_rate": 1.2187902187902189e-05,
      "loss": 0.8265,
      "step": 5536
    },
    {
      "epoch": 1.19,
      "grad_norm": 8.170022964477539,
      "learning_rate": 1.2184684684684685e-05,
      "loss": 0.5554,
      "step": 5537
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.555206298828125,
      "learning_rate": 1.2181467181467181e-05,
      "loss": 0.8236,
      "step": 5538
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.533485412597656,
      "learning_rate": 1.2178249678249678e-05,
      "loss": 0.5349,
      "step": 5539
    },
    {
      "epoch": 1.19,
      "grad_norm": 8.43478012084961,
      "learning_rate": 1.2175032175032176e-05,
      "loss": 0.625,
      "step": 5540
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.072870254516602,
      "learning_rate": 1.2171814671814673e-05,
      "loss": 0.7332,
      "step": 5541
    },
    {
      "epoch": 1.19,
      "grad_norm": 12.510449409484863,
      "learning_rate": 1.2168597168597169e-05,
      "loss": 0.946,
      "step": 5542
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.108518600463867,
      "learning_rate": 1.2165379665379665e-05,
      "loss": 0.5196,
      "step": 5543
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.25544548034668,
      "learning_rate": 1.2162162162162164e-05,
      "loss": 0.7714,
      "step": 5544
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.743964195251465,
      "learning_rate": 1.2158944658944658e-05,
      "loss": 0.5079,
      "step": 5545
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.355385780334473,
      "learning_rate": 1.2155727155727155e-05,
      "loss": 0.6888,
      "step": 5546
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.9121675491333,
      "learning_rate": 1.2152509652509653e-05,
      "loss": 0.5545,
      "step": 5547
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.555904388427734,
      "learning_rate": 1.214929214929215e-05,
      "loss": 0.8053,
      "step": 5548
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.164778709411621,
      "learning_rate": 1.2146074646074646e-05,
      "loss": 0.5697,
      "step": 5549
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.678691864013672,
      "learning_rate": 1.2142857142857144e-05,
      "loss": 0.5713,
      "step": 5550
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.550882339477539,
      "learning_rate": 1.213963963963964e-05,
      "loss": 0.7177,
      "step": 5551
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.325257301330566,
      "learning_rate": 1.2136422136422137e-05,
      "loss": 0.6358,
      "step": 5552
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.019326210021973,
      "learning_rate": 1.2133204633204635e-05,
      "loss": 0.6142,
      "step": 5553
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.712057113647461,
      "learning_rate": 1.212998712998713e-05,
      "loss": 0.9501,
      "step": 5554
    },
    {
      "epoch": 1.19,
      "grad_norm": 12.978845596313477,
      "learning_rate": 1.2126769626769626e-05,
      "loss": 0.7679,
      "step": 5555
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.210609436035156,
      "learning_rate": 1.2123552123552123e-05,
      "loss": 0.7109,
      "step": 5556
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.195422172546387,
      "learning_rate": 1.212033462033462e-05,
      "loss": 0.5508,
      "step": 5557
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.972443580627441,
      "learning_rate": 1.2117117117117117e-05,
      "loss": 0.8842,
      "step": 5558
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.013911247253418,
      "learning_rate": 1.2113899613899614e-05,
      "loss": 0.5543,
      "step": 5559
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.903312683105469,
      "learning_rate": 1.2110682110682112e-05,
      "loss": 0.6429,
      "step": 5560
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.143915176391602,
      "learning_rate": 1.2107464607464608e-05,
      "loss": 0.695,
      "step": 5561
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.594697952270508,
      "learning_rate": 1.2104247104247105e-05,
      "loss": 0.9303,
      "step": 5562
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.720758438110352,
      "learning_rate": 1.2101029601029601e-05,
      "loss": 0.5825,
      "step": 5563
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.1403169631958,
      "learning_rate": 1.2097812097812098e-05,
      "loss": 0.52,
      "step": 5564
    },
    {
      "epoch": 1.19,
      "grad_norm": 7.987728595733643,
      "learning_rate": 1.2094594594594594e-05,
      "loss": 0.4668,
      "step": 5565
    },
    {
      "epoch": 1.19,
      "grad_norm": 13.141693115234375,
      "learning_rate": 1.2091377091377092e-05,
      "loss": 0.6228,
      "step": 5566
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.372549057006836,
      "learning_rate": 1.2088159588159589e-05,
      "loss": 0.8102,
      "step": 5567
    },
    {
      "epoch": 1.19,
      "grad_norm": 12.446846008300781,
      "learning_rate": 1.2084942084942085e-05,
      "loss": 0.9005,
      "step": 5568
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.559760093688965,
      "learning_rate": 1.2081724581724583e-05,
      "loss": 0.5238,
      "step": 5569
    },
    {
      "epoch": 1.19,
      "grad_norm": 10.331111907958984,
      "learning_rate": 1.207850707850708e-05,
      "loss": 0.5744,
      "step": 5570
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.716878890991211,
      "learning_rate": 1.2075289575289575e-05,
      "loss": 0.5212,
      "step": 5571
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.495830535888672,
      "learning_rate": 1.2072072072072071e-05,
      "loss": 0.6137,
      "step": 5572
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.90797233581543,
      "learning_rate": 1.2068854568854569e-05,
      "loss": 0.7324,
      "step": 5573
    },
    {
      "epoch": 1.2,
      "grad_norm": 12.441437721252441,
      "learning_rate": 1.2065637065637066e-05,
      "loss": 0.6646,
      "step": 5574
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.980324745178223,
      "learning_rate": 1.2062419562419562e-05,
      "loss": 0.6585,
      "step": 5575
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.779523849487305,
      "learning_rate": 1.205920205920206e-05,
      "loss": 0.7106,
      "step": 5576
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.310246467590332,
      "learning_rate": 1.2055984555984557e-05,
      "loss": 0.5191,
      "step": 5577
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.137803077697754,
      "learning_rate": 1.2052767052767053e-05,
      "loss": 0.5822,
      "step": 5578
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.095663070678711,
      "learning_rate": 1.2049549549549551e-05,
      "loss": 0.6953,
      "step": 5579
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.201428413391113,
      "learning_rate": 1.2046332046332046e-05,
      "loss": 0.5656,
      "step": 5580
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.480624198913574,
      "learning_rate": 1.2043114543114542e-05,
      "loss": 0.6037,
      "step": 5581
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.581961631774902,
      "learning_rate": 1.203989703989704e-05,
      "loss": 0.7131,
      "step": 5582
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.1426362991333,
      "learning_rate": 1.2036679536679537e-05,
      "loss": 0.6924,
      "step": 5583
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.817279815673828,
      "learning_rate": 1.2033462033462033e-05,
      "loss": 0.5272,
      "step": 5584
    },
    {
      "epoch": 1.2,
      "grad_norm": 12.816975593566895,
      "learning_rate": 1.203024453024453e-05,
      "loss": 0.4967,
      "step": 5585
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.168670654296875,
      "learning_rate": 1.2027027027027028e-05,
      "loss": 0.7158,
      "step": 5586
    },
    {
      "epoch": 1.2,
      "grad_norm": 13.036796569824219,
      "learning_rate": 1.2023809523809525e-05,
      "loss": 0.7551,
      "step": 5587
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.80281925201416,
      "learning_rate": 1.2020592020592021e-05,
      "loss": 0.6188,
      "step": 5588
    },
    {
      "epoch": 1.2,
      "grad_norm": 12.46837329864502,
      "learning_rate": 1.2017374517374517e-05,
      "loss": 0.7358,
      "step": 5589
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.860552787780762,
      "learning_rate": 1.2014157014157014e-05,
      "loss": 0.5498,
      "step": 5590
    },
    {
      "epoch": 1.2,
      "grad_norm": 12.03942584991455,
      "learning_rate": 1.201093951093951e-05,
      "loss": 0.7835,
      "step": 5591
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.688024520874023,
      "learning_rate": 1.2007722007722008e-05,
      "loss": 0.5059,
      "step": 5592
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.121278762817383,
      "learning_rate": 1.2004504504504505e-05,
      "loss": 0.6057,
      "step": 5593
    },
    {
      "epoch": 1.2,
      "grad_norm": 12.562775611877441,
      "learning_rate": 1.2001287001287001e-05,
      "loss": 0.804,
      "step": 5594
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.878985404968262,
      "learning_rate": 1.19980694980695e-05,
      "loss": 0.5065,
      "step": 5595
    },
    {
      "epoch": 1.2,
      "grad_norm": 13.144977569580078,
      "learning_rate": 1.1994851994851996e-05,
      "loss": 0.5152,
      "step": 5596
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.78722858428955,
      "learning_rate": 1.1991634491634492e-05,
      "loss": 0.5724,
      "step": 5597
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.038661003112793,
      "learning_rate": 1.1988416988416989e-05,
      "loss": 0.375,
      "step": 5598
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.980592727661133,
      "learning_rate": 1.1985199485199485e-05,
      "loss": 0.5407,
      "step": 5599
    },
    {
      "epoch": 1.2,
      "grad_norm": 12.439742088317871,
      "learning_rate": 1.1981981981981982e-05,
      "loss": 0.7852,
      "step": 5600
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.048242568969727,
      "learning_rate": 1.1978764478764478e-05,
      "loss": 0.4915,
      "step": 5601
    },
    {
      "epoch": 1.2,
      "grad_norm": 16.884252548217773,
      "learning_rate": 1.1975546975546976e-05,
      "loss": 0.7145,
      "step": 5602
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.4080171585083,
      "learning_rate": 1.1972329472329473e-05,
      "loss": 0.4317,
      "step": 5603
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.447467803955078,
      "learning_rate": 1.196911196911197e-05,
      "loss": 0.7478,
      "step": 5604
    },
    {
      "epoch": 1.2,
      "grad_norm": 17.059324264526367,
      "learning_rate": 1.1965894465894467e-05,
      "loss": 0.7763,
      "step": 5605
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.467840194702148,
      "learning_rate": 1.1962676962676962e-05,
      "loss": 0.4857,
      "step": 5606
    },
    {
      "epoch": 1.2,
      "grad_norm": 10.40206527709961,
      "learning_rate": 1.1959459459459459e-05,
      "loss": 0.6124,
      "step": 5607
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.37880802154541,
      "learning_rate": 1.1956241956241957e-05,
      "loss": 0.5648,
      "step": 5608
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.144296646118164,
      "learning_rate": 1.1953024453024453e-05,
      "loss": 0.7354,
      "step": 5609
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.092069625854492,
      "learning_rate": 1.194980694980695e-05,
      "loss": 0.3947,
      "step": 5610
    },
    {
      "epoch": 1.2,
      "grad_norm": 13.492918968200684,
      "learning_rate": 1.1946589446589448e-05,
      "loss": 0.7332,
      "step": 5611
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.21018123626709,
      "learning_rate": 1.1943371943371944e-05,
      "loss": 0.6069,
      "step": 5612
    },
    {
      "epoch": 1.2,
      "grad_norm": 12.484994888305664,
      "learning_rate": 1.194015444015444e-05,
      "loss": 0.4783,
      "step": 5613
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.223959922790527,
      "learning_rate": 1.1936936936936937e-05,
      "loss": 0.5247,
      "step": 5614
    },
    {
      "epoch": 1.2,
      "grad_norm": 6.786130905151367,
      "learning_rate": 1.1933719433719434e-05,
      "loss": 0.3562,
      "step": 5615
    },
    {
      "epoch": 1.2,
      "grad_norm": 13.067726135253906,
      "learning_rate": 1.193050193050193e-05,
      "loss": 0.7021,
      "step": 5616
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.209267616271973,
      "learning_rate": 1.1927284427284427e-05,
      "loss": 0.5296,
      "step": 5617
    },
    {
      "epoch": 1.21,
      "grad_norm": 12.054997444152832,
      "learning_rate": 1.1924066924066925e-05,
      "loss": 0.7417,
      "step": 5618
    },
    {
      "epoch": 1.21,
      "grad_norm": 9.723998069763184,
      "learning_rate": 1.1920849420849421e-05,
      "loss": 0.6698,
      "step": 5619
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.472500801086426,
      "learning_rate": 1.1917631917631918e-05,
      "loss": 0.6603,
      "step": 5620
    },
    {
      "epoch": 1.21,
      "grad_norm": 9.209421157836914,
      "learning_rate": 1.1914414414414416e-05,
      "loss": 0.5881,
      "step": 5621
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.54002571105957,
      "learning_rate": 1.1911196911196912e-05,
      "loss": 0.6021,
      "step": 5622
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.957834243774414,
      "learning_rate": 1.1907979407979409e-05,
      "loss": 0.7859,
      "step": 5623
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.576760292053223,
      "learning_rate": 1.1904761904761905e-05,
      "loss": 0.6843,
      "step": 5624
    },
    {
      "epoch": 1.21,
      "grad_norm": 13.092652320861816,
      "learning_rate": 1.1901544401544402e-05,
      "loss": 1.1333,
      "step": 5625
    },
    {
      "epoch": 1.21,
      "grad_norm": 7.636936187744141,
      "learning_rate": 1.1898326898326898e-05,
      "loss": 0.4561,
      "step": 5626
    },
    {
      "epoch": 1.21,
      "grad_norm": 9.250845909118652,
      "learning_rate": 1.1895109395109396e-05,
      "loss": 0.6386,
      "step": 5627
    },
    {
      "epoch": 1.21,
      "grad_norm": 12.089984893798828,
      "learning_rate": 1.1891891891891893e-05,
      "loss": 0.6095,
      "step": 5628
    },
    {
      "epoch": 1.21,
      "grad_norm": 13.426558494567871,
      "learning_rate": 1.1888674388674389e-05,
      "loss": 0.6624,
      "step": 5629
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.29769229888916,
      "learning_rate": 1.1885456885456885e-05,
      "loss": 0.7976,
      "step": 5630
    },
    {
      "epoch": 1.21,
      "grad_norm": 13.056904792785645,
      "learning_rate": 1.1882239382239384e-05,
      "loss": 0.7347,
      "step": 5631
    },
    {
      "epoch": 1.21,
      "grad_norm": 13.318816184997559,
      "learning_rate": 1.1879021879021878e-05,
      "loss": 0.5557,
      "step": 5632
    },
    {
      "epoch": 1.21,
      "grad_norm": 8.74739933013916,
      "learning_rate": 1.1875804375804375e-05,
      "loss": 0.5687,
      "step": 5633
    },
    {
      "epoch": 1.21,
      "grad_norm": 12.35254955291748,
      "learning_rate": 1.1872586872586873e-05,
      "loss": 0.6878,
      "step": 5634
    },
    {
      "epoch": 1.21,
      "grad_norm": 13.31240177154541,
      "learning_rate": 1.186936936936937e-05,
      "loss": 0.7069,
      "step": 5635
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.842521667480469,
      "learning_rate": 1.1866151866151866e-05,
      "loss": 0.5373,
      "step": 5636
    },
    {
      "epoch": 1.21,
      "grad_norm": 9.006057739257812,
      "learning_rate": 1.1862934362934364e-05,
      "loss": 0.609,
      "step": 5637
    },
    {
      "epoch": 1.21,
      "grad_norm": 8.137203216552734,
      "learning_rate": 1.185971685971686e-05,
      "loss": 0.5844,
      "step": 5638
    },
    {
      "epoch": 1.21,
      "grad_norm": 8.047407150268555,
      "learning_rate": 1.1856499356499357e-05,
      "loss": 0.5595,
      "step": 5639
    },
    {
      "epoch": 1.21,
      "grad_norm": 11.361739158630371,
      "learning_rate": 1.1853281853281855e-05,
      "loss": 0.5855,
      "step": 5640
    },
    {
      "epoch": 1.21,
      "grad_norm": 8.561090469360352,
      "learning_rate": 1.185006435006435e-05,
      "loss": 0.5868,
      "step": 5641
    },
    {
      "epoch": 1.21,
      "grad_norm": 11.958687782287598,
      "learning_rate": 1.1846846846846846e-05,
      "loss": 0.7,
      "step": 5642
    },
    {
      "epoch": 1.21,
      "grad_norm": 11.929163932800293,
      "learning_rate": 1.1843629343629343e-05,
      "loss": 0.744,
      "step": 5643
    },
    {
      "epoch": 1.21,
      "grad_norm": 7.506537437438965,
      "learning_rate": 1.1840411840411841e-05,
      "loss": 0.4749,
      "step": 5644
    },
    {
      "epoch": 1.21,
      "grad_norm": 8.610102653503418,
      "learning_rate": 1.1837194337194337e-05,
      "loss": 0.505,
      "step": 5645
    },
    {
      "epoch": 1.21,
      "grad_norm": 11.720768928527832,
      "learning_rate": 1.1833976833976834e-05,
      "loss": 0.4438,
      "step": 5646
    },
    {
      "epoch": 1.21,
      "grad_norm": 12.115388870239258,
      "learning_rate": 1.1830759330759332e-05,
      "loss": 0.8661,
      "step": 5647
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.539558410644531,
      "learning_rate": 1.1827541827541828e-05,
      "loss": 0.8862,
      "step": 5648
    },
    {
      "epoch": 1.21,
      "grad_norm": 9.631455421447754,
      "learning_rate": 1.1824324324324325e-05,
      "loss": 0.776,
      "step": 5649
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.00440502166748,
      "learning_rate": 1.1821106821106821e-05,
      "loss": 0.7238,
      "step": 5650
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.384056091308594,
      "learning_rate": 1.1817889317889318e-05,
      "loss": 0.722,
      "step": 5651
    },
    {
      "epoch": 1.21,
      "grad_norm": 12.449073791503906,
      "learning_rate": 1.1814671814671814e-05,
      "loss": 0.8203,
      "step": 5652
    },
    {
      "epoch": 1.21,
      "grad_norm": 14.760211944580078,
      "learning_rate": 1.1811454311454312e-05,
      "loss": 1.0322,
      "step": 5653
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.513514518737793,
      "learning_rate": 1.1808236808236809e-05,
      "loss": 0.7052,
      "step": 5654
    },
    {
      "epoch": 1.21,
      "grad_norm": 8.513910293579102,
      "learning_rate": 1.1805019305019305e-05,
      "loss": 0.5487,
      "step": 5655
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.948304176330566,
      "learning_rate": 1.1801801801801803e-05,
      "loss": 0.7767,
      "step": 5656
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.365873336791992,
      "learning_rate": 1.17985842985843e-05,
      "loss": 0.4529,
      "step": 5657
    },
    {
      "epoch": 1.21,
      "grad_norm": 10.12270450592041,
      "learning_rate": 1.1795366795366795e-05,
      "loss": 0.5066,
      "step": 5658
    },
    {
      "epoch": 1.21,
      "grad_norm": 8.533530235290527,
      "learning_rate": 1.1792149292149291e-05,
      "loss": 0.7799,
      "step": 5659
    },
    {
      "epoch": 1.21,
      "grad_norm": 13.324342727661133,
      "learning_rate": 1.178893178893179e-05,
      "loss": 0.7511,
      "step": 5660
    },
    {
      "epoch": 1.21,
      "grad_norm": 9.579089164733887,
      "learning_rate": 1.1785714285714286e-05,
      "loss": 0.5354,
      "step": 5661
    },
    {
      "epoch": 1.21,
      "grad_norm": 9.246356964111328,
      "learning_rate": 1.1782496782496782e-05,
      "loss": 0.4413,
      "step": 5662
    },
    {
      "epoch": 1.21,
      "grad_norm": 9.824932098388672,
      "learning_rate": 1.177927927927928e-05,
      "loss": 0.5515,
      "step": 5663
    },
    {
      "epoch": 1.21,
      "grad_norm": 11.456934928894043,
      "learning_rate": 1.1776061776061777e-05,
      "loss": 0.7873,
      "step": 5664
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.706826210021973,
      "learning_rate": 1.1772844272844273e-05,
      "loss": 0.4828,
      "step": 5665
    },
    {
      "epoch": 1.22,
      "grad_norm": 11.460789680480957,
      "learning_rate": 1.1769626769626771e-05,
      "loss": 0.4662,
      "step": 5666
    },
    {
      "epoch": 1.22,
      "grad_norm": 13.272821426391602,
      "learning_rate": 1.1766409266409266e-05,
      "loss": 0.6641,
      "step": 5667
    },
    {
      "epoch": 1.22,
      "grad_norm": 12.775477409362793,
      "learning_rate": 1.1763191763191763e-05,
      "loss": 0.8572,
      "step": 5668
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.879241943359375,
      "learning_rate": 1.175997425997426e-05,
      "loss": 0.4416,
      "step": 5669
    },
    {
      "epoch": 1.22,
      "grad_norm": 11.552146911621094,
      "learning_rate": 1.1756756756756757e-05,
      "loss": 0.955,
      "step": 5670
    },
    {
      "epoch": 1.22,
      "grad_norm": 11.094269752502441,
      "learning_rate": 1.1753539253539254e-05,
      "loss": 0.6954,
      "step": 5671
    },
    {
      "epoch": 1.22,
      "grad_norm": 13.423489570617676,
      "learning_rate": 1.175032175032175e-05,
      "loss": 0.7696,
      "step": 5672
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.062320709228516,
      "learning_rate": 1.1747104247104248e-05,
      "loss": 0.6327,
      "step": 5673
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.109745979309082,
      "learning_rate": 1.1743886743886745e-05,
      "loss": 0.5685,
      "step": 5674
    },
    {
      "epoch": 1.22,
      "grad_norm": 8.946268081665039,
      "learning_rate": 1.1740669240669241e-05,
      "loss": 0.5998,
      "step": 5675
    },
    {
      "epoch": 1.22,
      "grad_norm": 15.679356575012207,
      "learning_rate": 1.1737451737451738e-05,
      "loss": 0.9343,
      "step": 5676
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.517240524291992,
      "learning_rate": 1.1734234234234234e-05,
      "loss": 0.5498,
      "step": 5677
    },
    {
      "epoch": 1.22,
      "grad_norm": 8.095696449279785,
      "learning_rate": 1.173101673101673e-05,
      "loss": 0.6099,
      "step": 5678
    },
    {
      "epoch": 1.22,
      "grad_norm": 8.965607643127441,
      "learning_rate": 1.1727799227799229e-05,
      "loss": 0.5291,
      "step": 5679
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.65093994140625,
      "learning_rate": 1.1724581724581725e-05,
      "loss": 0.6317,
      "step": 5680
    },
    {
      "epoch": 1.22,
      "grad_norm": 7.9654388427734375,
      "learning_rate": 1.1721364221364221e-05,
      "loss": 0.5218,
      "step": 5681
    },
    {
      "epoch": 1.22,
      "grad_norm": 12.487832069396973,
      "learning_rate": 1.171814671814672e-05,
      "loss": 0.775,
      "step": 5682
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.256637573242188,
      "learning_rate": 1.1714929214929216e-05,
      "loss": 0.5517,
      "step": 5683
    },
    {
      "epoch": 1.22,
      "grad_norm": 12.189037322998047,
      "learning_rate": 1.171171171171171e-05,
      "loss": 0.5993,
      "step": 5684
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.066032409667969,
      "learning_rate": 1.1708494208494209e-05,
      "loss": 0.5133,
      "step": 5685
    },
    {
      "epoch": 1.22,
      "grad_norm": 8.325027465820312,
      "learning_rate": 1.1705276705276705e-05,
      "loss": 0.4577,
      "step": 5686
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.237178802490234,
      "learning_rate": 1.1702059202059202e-05,
      "loss": 0.61,
      "step": 5687
    },
    {
      "epoch": 1.22,
      "grad_norm": 11.799751281738281,
      "learning_rate": 1.1698841698841698e-05,
      "loss": 0.6242,
      "step": 5688
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.634851455688477,
      "learning_rate": 1.1695624195624196e-05,
      "loss": 0.5867,
      "step": 5689
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.61033821105957,
      "learning_rate": 1.1692406692406693e-05,
      "loss": 0.546,
      "step": 5690
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.137800216674805,
      "learning_rate": 1.168918918918919e-05,
      "loss": 0.6064,
      "step": 5691
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.949976921081543,
      "learning_rate": 1.1685971685971688e-05,
      "loss": 0.6439,
      "step": 5692
    },
    {
      "epoch": 1.22,
      "grad_norm": 12.623964309692383,
      "learning_rate": 1.1682754182754182e-05,
      "loss": 0.8532,
      "step": 5693
    },
    {
      "epoch": 1.22,
      "grad_norm": 12.316421508789062,
      "learning_rate": 1.1679536679536679e-05,
      "loss": 0.6019,
      "step": 5694
    },
    {
      "epoch": 1.22,
      "grad_norm": 12.036429405212402,
      "learning_rate": 1.1676319176319177e-05,
      "loss": 0.6113,
      "step": 5695
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.449747085571289,
      "learning_rate": 1.1673101673101673e-05,
      "loss": 0.5132,
      "step": 5696
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.901187896728516,
      "learning_rate": 1.166988416988417e-05,
      "loss": 0.7989,
      "step": 5697
    },
    {
      "epoch": 1.22,
      "grad_norm": 15.589920043945312,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.6696,
      "step": 5698
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.279378890991211,
      "learning_rate": 1.1663449163449164e-05,
      "loss": 0.7593,
      "step": 5699
    },
    {
      "epoch": 1.22,
      "grad_norm": 15.204255104064941,
      "learning_rate": 1.166023166023166e-05,
      "loss": 0.84,
      "step": 5700
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.941072463989258,
      "learning_rate": 1.1657014157014157e-05,
      "loss": 0.5019,
      "step": 5701
    },
    {
      "epoch": 1.22,
      "grad_norm": 8.988157272338867,
      "learning_rate": 1.1653796653796654e-05,
      "loss": 0.623,
      "step": 5702
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.87256908416748,
      "learning_rate": 1.165057915057915e-05,
      "loss": 0.6962,
      "step": 5703
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.744824409484863,
      "learning_rate": 1.1647361647361647e-05,
      "loss": 0.6827,
      "step": 5704
    },
    {
      "epoch": 1.22,
      "grad_norm": 8.722306251525879,
      "learning_rate": 1.1644144144144145e-05,
      "loss": 0.6327,
      "step": 5705
    },
    {
      "epoch": 1.22,
      "grad_norm": 12.536724090576172,
      "learning_rate": 1.1640926640926641e-05,
      "loss": 0.6769,
      "step": 5706
    },
    {
      "epoch": 1.22,
      "grad_norm": 11.398816108703613,
      "learning_rate": 1.1637709137709138e-05,
      "loss": 0.9206,
      "step": 5707
    },
    {
      "epoch": 1.22,
      "grad_norm": 11.313015937805176,
      "learning_rate": 1.1634491634491636e-05,
      "loss": 0.6968,
      "step": 5708
    },
    {
      "epoch": 1.22,
      "grad_norm": 12.878440856933594,
      "learning_rate": 1.1631274131274132e-05,
      "loss": 0.635,
      "step": 5709
    },
    {
      "epoch": 1.22,
      "grad_norm": 9.560469627380371,
      "learning_rate": 1.1628056628056629e-05,
      "loss": 0.8918,
      "step": 5710
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.79146671295166,
      "learning_rate": 1.1624839124839125e-05,
      "loss": 0.6876,
      "step": 5711
    },
    {
      "epoch": 1.23,
      "grad_norm": 8.302037239074707,
      "learning_rate": 1.1621621621621622e-05,
      "loss": 0.6135,
      "step": 5712
    },
    {
      "epoch": 1.23,
      "grad_norm": 11.75120735168457,
      "learning_rate": 1.1618404118404118e-05,
      "loss": 0.7988,
      "step": 5713
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.234381675720215,
      "learning_rate": 1.1615186615186616e-05,
      "loss": 0.618,
      "step": 5714
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.207605361938477,
      "learning_rate": 1.1611969111969113e-05,
      "loss": 0.671,
      "step": 5715
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.308612823486328,
      "learning_rate": 1.1608751608751609e-05,
      "loss": 0.5718,
      "step": 5716
    },
    {
      "epoch": 1.23,
      "grad_norm": 11.512102127075195,
      "learning_rate": 1.1605534105534106e-05,
      "loss": 0.8465,
      "step": 5717
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.470547676086426,
      "learning_rate": 1.1602316602316604e-05,
      "loss": 0.8476,
      "step": 5718
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.80661392211914,
      "learning_rate": 1.1599099099099098e-05,
      "loss": 0.6399,
      "step": 5719
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.980842590332031,
      "learning_rate": 1.1595881595881595e-05,
      "loss": 0.7257,
      "step": 5720
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.952272415161133,
      "learning_rate": 1.1592664092664093e-05,
      "loss": 0.6349,
      "step": 5721
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.19614028930664,
      "learning_rate": 1.158944658944659e-05,
      "loss": 0.7503,
      "step": 5722
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.74699878692627,
      "learning_rate": 1.1586229086229086e-05,
      "loss": 0.5474,
      "step": 5723
    },
    {
      "epoch": 1.23,
      "grad_norm": 7.68356466293335,
      "learning_rate": 1.1583011583011584e-05,
      "loss": 0.8408,
      "step": 5724
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.691629409790039,
      "learning_rate": 1.157979407979408e-05,
      "loss": 0.5324,
      "step": 5725
    },
    {
      "epoch": 1.23,
      "grad_norm": 12.425187110900879,
      "learning_rate": 1.1576576576576577e-05,
      "loss": 0.713,
      "step": 5726
    },
    {
      "epoch": 1.23,
      "grad_norm": 11.141385078430176,
      "learning_rate": 1.1573359073359075e-05,
      "loss": 0.7545,
      "step": 5727
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.223686218261719,
      "learning_rate": 1.157014157014157e-05,
      "loss": 0.7908,
      "step": 5728
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.558094024658203,
      "learning_rate": 1.1566924066924066e-05,
      "loss": 0.5841,
      "step": 5729
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.68365478515625,
      "learning_rate": 1.1563706563706563e-05,
      "loss": 0.6197,
      "step": 5730
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.042082786560059,
      "learning_rate": 1.1560489060489061e-05,
      "loss": 0.6628,
      "step": 5731
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.334814071655273,
      "learning_rate": 1.1557271557271557e-05,
      "loss": 0.7426,
      "step": 5732
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.289400100708008,
      "learning_rate": 1.1554054054054054e-05,
      "loss": 0.6655,
      "step": 5733
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.357425689697266,
      "learning_rate": 1.1550836550836552e-05,
      "loss": 0.6318,
      "step": 5734
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.125884056091309,
      "learning_rate": 1.1547619047619048e-05,
      "loss": 0.7254,
      "step": 5735
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.344581604003906,
      "learning_rate": 1.1544401544401545e-05,
      "loss": 0.5111,
      "step": 5736
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.286027908325195,
      "learning_rate": 1.1541184041184041e-05,
      "loss": 0.5016,
      "step": 5737
    },
    {
      "epoch": 1.23,
      "grad_norm": 13.438435554504395,
      "learning_rate": 1.1537966537966538e-05,
      "loss": 1.1259,
      "step": 5738
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.820120811462402,
      "learning_rate": 1.1534749034749034e-05,
      "loss": 0.6303,
      "step": 5739
    },
    {
      "epoch": 1.23,
      "grad_norm": 7.391282081604004,
      "learning_rate": 1.1531531531531532e-05,
      "loss": 0.4927,
      "step": 5740
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.012638092041016,
      "learning_rate": 1.1528314028314029e-05,
      "loss": 0.5931,
      "step": 5741
    },
    {
      "epoch": 1.23,
      "grad_norm": 7.9248881340026855,
      "learning_rate": 1.1525096525096525e-05,
      "loss": 0.4602,
      "step": 5742
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.587740898132324,
      "learning_rate": 1.1521879021879023e-05,
      "loss": 0.507,
      "step": 5743
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.93956470489502,
      "learning_rate": 1.151866151866152e-05,
      "loss": 0.8611,
      "step": 5744
    },
    {
      "epoch": 1.23,
      "grad_norm": 8.677390098571777,
      "learning_rate": 1.1515444015444015e-05,
      "loss": 0.5343,
      "step": 5745
    },
    {
      "epoch": 1.23,
      "grad_norm": 8.547858238220215,
      "learning_rate": 1.1512226512226511e-05,
      "loss": 0.5799,
      "step": 5746
    },
    {
      "epoch": 1.23,
      "grad_norm": 11.190450668334961,
      "learning_rate": 1.150900900900901e-05,
      "loss": 0.7264,
      "step": 5747
    },
    {
      "epoch": 1.23,
      "grad_norm": 12.12058162689209,
      "learning_rate": 1.1505791505791506e-05,
      "loss": 0.8874,
      "step": 5748
    },
    {
      "epoch": 1.23,
      "grad_norm": 11.10948371887207,
      "learning_rate": 1.1502574002574002e-05,
      "loss": 0.6043,
      "step": 5749
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.234842300415039,
      "learning_rate": 1.14993564993565e-05,
      "loss": 0.618,
      "step": 5750
    },
    {
      "epoch": 1.23,
      "grad_norm": 11.574522972106934,
      "learning_rate": 1.1496138996138997e-05,
      "loss": 0.5726,
      "step": 5751
    },
    {
      "epoch": 1.23,
      "grad_norm": 8.594467163085938,
      "learning_rate": 1.1492921492921493e-05,
      "loss": 0.4737,
      "step": 5752
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.486374855041504,
      "learning_rate": 1.1489703989703991e-05,
      "loss": 0.4469,
      "step": 5753
    },
    {
      "epoch": 1.23,
      "grad_norm": 8.36275577545166,
      "learning_rate": 1.1486486486486486e-05,
      "loss": 0.5336,
      "step": 5754
    },
    {
      "epoch": 1.23,
      "grad_norm": 11.812424659729004,
      "learning_rate": 1.1483268983268983e-05,
      "loss": 0.4908,
      "step": 5755
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.33041763305664,
      "learning_rate": 1.148005148005148e-05,
      "loss": 0.5071,
      "step": 5756
    },
    {
      "epoch": 1.23,
      "grad_norm": 9.389996528625488,
      "learning_rate": 1.1476833976833977e-05,
      "loss": 0.8118,
      "step": 5757
    },
    {
      "epoch": 1.24,
      "grad_norm": 8.574983596801758,
      "learning_rate": 1.1473616473616474e-05,
      "loss": 0.6372,
      "step": 5758
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.394654273986816,
      "learning_rate": 1.147039897039897e-05,
      "loss": 0.4996,
      "step": 5759
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.192439079284668,
      "learning_rate": 1.1467181467181468e-05,
      "loss": 0.6348,
      "step": 5760
    },
    {
      "epoch": 1.24,
      "grad_norm": 13.894701957702637,
      "learning_rate": 1.1463963963963965e-05,
      "loss": 0.5682,
      "step": 5761
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.88235092163086,
      "learning_rate": 1.1460746460746461e-05,
      "loss": 0.4393,
      "step": 5762
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.157697677612305,
      "learning_rate": 1.1457528957528958e-05,
      "loss": 0.4799,
      "step": 5763
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.25743579864502,
      "learning_rate": 1.1454311454311454e-05,
      "loss": 0.6137,
      "step": 5764
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.72878360748291,
      "learning_rate": 1.145109395109395e-05,
      "loss": 0.5679,
      "step": 5765
    },
    {
      "epoch": 1.24,
      "grad_norm": 11.360867500305176,
      "learning_rate": 1.1447876447876449e-05,
      "loss": 0.5277,
      "step": 5766
    },
    {
      "epoch": 1.24,
      "grad_norm": 13.596990585327148,
      "learning_rate": 1.1444658944658945e-05,
      "loss": 0.7548,
      "step": 5767
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.135992050170898,
      "learning_rate": 1.1441441441441442e-05,
      "loss": 0.5442,
      "step": 5768
    },
    {
      "epoch": 1.24,
      "grad_norm": 11.342637062072754,
      "learning_rate": 1.143822393822394e-05,
      "loss": 0.7297,
      "step": 5769
    },
    {
      "epoch": 1.24,
      "grad_norm": 11.280220031738281,
      "learning_rate": 1.1435006435006436e-05,
      "loss": 0.5516,
      "step": 5770
    },
    {
      "epoch": 1.24,
      "grad_norm": 12.105536460876465,
      "learning_rate": 1.1431788931788931e-05,
      "loss": 0.656,
      "step": 5771
    },
    {
      "epoch": 1.24,
      "grad_norm": 13.361992835998535,
      "learning_rate": 1.1428571428571429e-05,
      "loss": 0.6652,
      "step": 5772
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.460247039794922,
      "learning_rate": 1.1425353925353925e-05,
      "loss": 0.8315,
      "step": 5773
    },
    {
      "epoch": 1.24,
      "grad_norm": 11.987143516540527,
      "learning_rate": 1.1422136422136422e-05,
      "loss": 0.7799,
      "step": 5774
    },
    {
      "epoch": 1.24,
      "grad_norm": 6.838326454162598,
      "learning_rate": 1.1418918918918918e-05,
      "loss": 0.4949,
      "step": 5775
    },
    {
      "epoch": 1.24,
      "grad_norm": 8.642560005187988,
      "learning_rate": 1.1415701415701417e-05,
      "loss": 0.6434,
      "step": 5776
    },
    {
      "epoch": 1.24,
      "grad_norm": 12.227544784545898,
      "learning_rate": 1.1412483912483913e-05,
      "loss": 0.5102,
      "step": 5777
    },
    {
      "epoch": 1.24,
      "grad_norm": 8.859933853149414,
      "learning_rate": 1.140926640926641e-05,
      "loss": 0.788,
      "step": 5778
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.52625846862793,
      "learning_rate": 1.1406048906048908e-05,
      "loss": 0.6308,
      "step": 5779
    },
    {
      "epoch": 1.24,
      "grad_norm": 8.958970069885254,
      "learning_rate": 1.1402831402831402e-05,
      "loss": 0.5078,
      "step": 5780
    },
    {
      "epoch": 1.24,
      "grad_norm": 8.247206687927246,
      "learning_rate": 1.1399613899613899e-05,
      "loss": 0.5567,
      "step": 5781
    },
    {
      "epoch": 1.24,
      "grad_norm": 11.56781005859375,
      "learning_rate": 1.1396396396396397e-05,
      "loss": 0.7295,
      "step": 5782
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.261177062988281,
      "learning_rate": 1.1393178893178893e-05,
      "loss": 0.8659,
      "step": 5783
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.430974960327148,
      "learning_rate": 1.138996138996139e-05,
      "loss": 0.6618,
      "step": 5784
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.67108154296875,
      "learning_rate": 1.1386743886743888e-05,
      "loss": 0.5428,
      "step": 5785
    },
    {
      "epoch": 1.24,
      "grad_norm": 8.532990455627441,
      "learning_rate": 1.1383526383526384e-05,
      "loss": 0.3941,
      "step": 5786
    },
    {
      "epoch": 1.24,
      "grad_norm": 11.191015243530273,
      "learning_rate": 1.1380308880308881e-05,
      "loss": 0.5394,
      "step": 5787
    },
    {
      "epoch": 1.24,
      "grad_norm": 8.258696556091309,
      "learning_rate": 1.1377091377091377e-05,
      "loss": 0.5467,
      "step": 5788
    },
    {
      "epoch": 1.24,
      "grad_norm": 11.423142433166504,
      "learning_rate": 1.1373873873873874e-05,
      "loss": 0.8677,
      "step": 5789
    },
    {
      "epoch": 1.24,
      "grad_norm": 8.16551399230957,
      "learning_rate": 1.137065637065637e-05,
      "loss": 0.4387,
      "step": 5790
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.890438079833984,
      "learning_rate": 1.1367438867438867e-05,
      "loss": 0.4881,
      "step": 5791
    },
    {
      "epoch": 1.24,
      "grad_norm": 11.194384574890137,
      "learning_rate": 1.1364221364221365e-05,
      "loss": 0.7269,
      "step": 5792
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.965856552124023,
      "learning_rate": 1.1361003861003861e-05,
      "loss": 0.5103,
      "step": 5793
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.460625648498535,
      "learning_rate": 1.1357786357786358e-05,
      "loss": 0.6565,
      "step": 5794
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.09826946258545,
      "learning_rate": 1.1354568854568856e-05,
      "loss": 0.7006,
      "step": 5795
    },
    {
      "epoch": 1.24,
      "grad_norm": 11.521309852600098,
      "learning_rate": 1.1351351351351352e-05,
      "loss": 0.8283,
      "step": 5796
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.73936653137207,
      "learning_rate": 1.1348133848133849e-05,
      "loss": 0.6033,
      "step": 5797
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.50523853302002,
      "learning_rate": 1.1344916344916345e-05,
      "loss": 0.7035,
      "step": 5798
    },
    {
      "epoch": 1.24,
      "grad_norm": 12.121225357055664,
      "learning_rate": 1.1341698841698842e-05,
      "loss": 0.9136,
      "step": 5799
    },
    {
      "epoch": 1.24,
      "grad_norm": 11.747520446777344,
      "learning_rate": 1.1338481338481338e-05,
      "loss": 0.657,
      "step": 5800
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.815031051635742,
      "learning_rate": 1.1335263835263836e-05,
      "loss": 0.48,
      "step": 5801
    },
    {
      "epoch": 1.24,
      "grad_norm": 10.230467796325684,
      "learning_rate": 1.1332046332046333e-05,
      "loss": 0.7886,
      "step": 5802
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.553311347961426,
      "learning_rate": 1.132882882882883e-05,
      "loss": 0.5876,
      "step": 5803
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.914076805114746,
      "learning_rate": 1.1325611325611326e-05,
      "loss": 0.4951,
      "step": 5804
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.573486328125,
      "learning_rate": 1.1322393822393824e-05,
      "loss": 0.7879,
      "step": 5805
    },
    {
      "epoch": 1.25,
      "grad_norm": 8.381064414978027,
      "learning_rate": 1.1319176319176319e-05,
      "loss": 0.4313,
      "step": 5806
    },
    {
      "epoch": 1.25,
      "grad_norm": 9.07919979095459,
      "learning_rate": 1.1315958815958815e-05,
      "loss": 0.4254,
      "step": 5807
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.307326316833496,
      "learning_rate": 1.1312741312741313e-05,
      "loss": 0.6436,
      "step": 5808
    },
    {
      "epoch": 1.25,
      "grad_norm": 17.47904396057129,
      "learning_rate": 1.130952380952381e-05,
      "loss": 1.0393,
      "step": 5809
    },
    {
      "epoch": 1.25,
      "grad_norm": 8.15190315246582,
      "learning_rate": 1.1306306306306306e-05,
      "loss": 0.452,
      "step": 5810
    },
    {
      "epoch": 1.25,
      "grad_norm": 12.115883827209473,
      "learning_rate": 1.1303088803088804e-05,
      "loss": 0.8316,
      "step": 5811
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.718400955200195,
      "learning_rate": 1.12998712998713e-05,
      "loss": 0.4601,
      "step": 5812
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.186429977416992,
      "learning_rate": 1.1296653796653797e-05,
      "loss": 0.5127,
      "step": 5813
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.879590034484863,
      "learning_rate": 1.1293436293436295e-05,
      "loss": 0.5465,
      "step": 5814
    },
    {
      "epoch": 1.25,
      "grad_norm": 12.561304092407227,
      "learning_rate": 1.129021879021879e-05,
      "loss": 0.4655,
      "step": 5815
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.199707984924316,
      "learning_rate": 1.1287001287001286e-05,
      "loss": 0.6705,
      "step": 5816
    },
    {
      "epoch": 1.25,
      "grad_norm": 8.002331733703613,
      "learning_rate": 1.1283783783783783e-05,
      "loss": 0.5984,
      "step": 5817
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.071301460266113,
      "learning_rate": 1.1280566280566281e-05,
      "loss": 0.7008,
      "step": 5818
    },
    {
      "epoch": 1.25,
      "grad_norm": 9.167276382446289,
      "learning_rate": 1.1277348777348777e-05,
      "loss": 0.5462,
      "step": 5819
    },
    {
      "epoch": 1.25,
      "grad_norm": 13.7432279586792,
      "learning_rate": 1.1274131274131274e-05,
      "loss": 0.9316,
      "step": 5820
    },
    {
      "epoch": 1.25,
      "grad_norm": 12.089761734008789,
      "learning_rate": 1.1270913770913772e-05,
      "loss": 0.9293,
      "step": 5821
    },
    {
      "epoch": 1.25,
      "grad_norm": 8.15857982635498,
      "learning_rate": 1.1267696267696269e-05,
      "loss": 0.4512,
      "step": 5822
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.114327430725098,
      "learning_rate": 1.1264478764478765e-05,
      "loss": 0.6513,
      "step": 5823
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.955876350402832,
      "learning_rate": 1.1261261261261261e-05,
      "loss": 0.4726,
      "step": 5824
    },
    {
      "epoch": 1.25,
      "grad_norm": 9.348434448242188,
      "learning_rate": 1.1258043758043758e-05,
      "loss": 0.8784,
      "step": 5825
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.104368209838867,
      "learning_rate": 1.1254826254826254e-05,
      "loss": 0.5871,
      "step": 5826
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.524931907653809,
      "learning_rate": 1.1251608751608752e-05,
      "loss": 0.6089,
      "step": 5827
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.591850280761719,
      "learning_rate": 1.1248391248391249e-05,
      "loss": 0.6507,
      "step": 5828
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.109919548034668,
      "learning_rate": 1.1245173745173745e-05,
      "loss": 0.425,
      "step": 5829
    },
    {
      "epoch": 1.25,
      "grad_norm": 8.450302124023438,
      "learning_rate": 1.1241956241956244e-05,
      "loss": 0.5434,
      "step": 5830
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.30454158782959,
      "learning_rate": 1.123873873873874e-05,
      "loss": 0.6905,
      "step": 5831
    },
    {
      "epoch": 1.25,
      "grad_norm": 9.299742698669434,
      "learning_rate": 1.1235521235521235e-05,
      "loss": 0.6155,
      "step": 5832
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.037215232849121,
      "learning_rate": 1.1232303732303731e-05,
      "loss": 0.5738,
      "step": 5833
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.746889114379883,
      "learning_rate": 1.122908622908623e-05,
      "loss": 0.6045,
      "step": 5834
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.47337818145752,
      "learning_rate": 1.1225868725868726e-05,
      "loss": 0.5052,
      "step": 5835
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.970874786376953,
      "learning_rate": 1.1222651222651222e-05,
      "loss": 0.6853,
      "step": 5836
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.749405860900879,
      "learning_rate": 1.121943371943372e-05,
      "loss": 0.6149,
      "step": 5837
    },
    {
      "epoch": 1.25,
      "grad_norm": 13.642724990844727,
      "learning_rate": 1.1216216216216217e-05,
      "loss": 0.8933,
      "step": 5838
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.97664737701416,
      "learning_rate": 1.1212998712998713e-05,
      "loss": 0.6041,
      "step": 5839
    },
    {
      "epoch": 1.25,
      "grad_norm": 9.749834060668945,
      "learning_rate": 1.1209781209781211e-05,
      "loss": 0.467,
      "step": 5840
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.295796394348145,
      "learning_rate": 1.1206563706563706e-05,
      "loss": 0.668,
      "step": 5841
    },
    {
      "epoch": 1.25,
      "grad_norm": 14.619892120361328,
      "learning_rate": 1.1203346203346203e-05,
      "loss": 0.6718,
      "step": 5842
    },
    {
      "epoch": 1.25,
      "grad_norm": 10.068387031555176,
      "learning_rate": 1.12001287001287e-05,
      "loss": 0.5565,
      "step": 5843
    },
    {
      "epoch": 1.25,
      "grad_norm": 8.940329551696777,
      "learning_rate": 1.1196911196911197e-05,
      "loss": 0.5273,
      "step": 5844
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.85120677947998,
      "learning_rate": 1.1193693693693694e-05,
      "loss": 0.6983,
      "step": 5845
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.432445526123047,
      "learning_rate": 1.119047619047619e-05,
      "loss": 0.8487,
      "step": 5846
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.184476852416992,
      "learning_rate": 1.1187258687258688e-05,
      "loss": 0.7493,
      "step": 5847
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.441741943359375,
      "learning_rate": 1.1184041184041185e-05,
      "loss": 0.6973,
      "step": 5848
    },
    {
      "epoch": 1.25,
      "grad_norm": 11.596246719360352,
      "learning_rate": 1.1180823680823681e-05,
      "loss": 0.599,
      "step": 5849
    },
    {
      "epoch": 1.25,
      "grad_norm": 12.042688369750977,
      "learning_rate": 1.1177606177606178e-05,
      "loss": 0.5679,
      "step": 5850
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.174405097961426,
      "learning_rate": 1.1174388674388674e-05,
      "loss": 0.494,
      "step": 5851
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.774840354919434,
      "learning_rate": 1.117117117117117e-05,
      "loss": 0.7465,
      "step": 5852
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.971283912658691,
      "learning_rate": 1.1167953667953669e-05,
      "loss": 0.5467,
      "step": 5853
    },
    {
      "epoch": 1.26,
      "grad_norm": 11.249275207519531,
      "learning_rate": 1.1164736164736165e-05,
      "loss": 0.5451,
      "step": 5854
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.050008773803711,
      "learning_rate": 1.1161518661518662e-05,
      "loss": 0.4863,
      "step": 5855
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.01704216003418,
      "learning_rate": 1.115830115830116e-05,
      "loss": 0.409,
      "step": 5856
    },
    {
      "epoch": 1.26,
      "grad_norm": 8.424596786499023,
      "learning_rate": 1.1155083655083656e-05,
      "loss": 0.4673,
      "step": 5857
    },
    {
      "epoch": 1.26,
      "grad_norm": 13.564428329467773,
      "learning_rate": 1.1151866151866151e-05,
      "loss": 0.5638,
      "step": 5858
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.458187103271484,
      "learning_rate": 1.1148648648648649e-05,
      "loss": 0.5476,
      "step": 5859
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.536824226379395,
      "learning_rate": 1.1145431145431146e-05,
      "loss": 0.5501,
      "step": 5860
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.589512825012207,
      "learning_rate": 1.1142213642213642e-05,
      "loss": 0.4431,
      "step": 5861
    },
    {
      "epoch": 1.26,
      "grad_norm": 7.696417808532715,
      "learning_rate": 1.1138996138996138e-05,
      "loss": 0.436,
      "step": 5862
    },
    {
      "epoch": 1.26,
      "grad_norm": 8.918874740600586,
      "learning_rate": 1.1135778635778637e-05,
      "loss": 0.5492,
      "step": 5863
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.102540969848633,
      "learning_rate": 1.1132561132561133e-05,
      "loss": 0.6291,
      "step": 5864
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.66883373260498,
      "learning_rate": 1.112934362934363e-05,
      "loss": 0.7485,
      "step": 5865
    },
    {
      "epoch": 1.26,
      "grad_norm": 11.421932220458984,
      "learning_rate": 1.1126126126126128e-05,
      "loss": 0.6908,
      "step": 5866
    },
    {
      "epoch": 1.26,
      "grad_norm": 12.41113567352295,
      "learning_rate": 1.1122908622908622e-05,
      "loss": 0.8608,
      "step": 5867
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.761419296264648,
      "learning_rate": 1.1119691119691119e-05,
      "loss": 0.6525,
      "step": 5868
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.606658935546875,
      "learning_rate": 1.1116473616473617e-05,
      "loss": 0.6717,
      "step": 5869
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.157236099243164,
      "learning_rate": 1.1113256113256113e-05,
      "loss": 0.478,
      "step": 5870
    },
    {
      "epoch": 1.26,
      "grad_norm": 7.354647159576416,
      "learning_rate": 1.111003861003861e-05,
      "loss": 0.3541,
      "step": 5871
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.893209457397461,
      "learning_rate": 1.1106821106821108e-05,
      "loss": 0.6881,
      "step": 5872
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.726308822631836,
      "learning_rate": 1.1103603603603604e-05,
      "loss": 0.4895,
      "step": 5873
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.630404472351074,
      "learning_rate": 1.1100386100386101e-05,
      "loss": 0.7158,
      "step": 5874
    },
    {
      "epoch": 1.26,
      "grad_norm": 11.30309772491455,
      "learning_rate": 1.1097168597168597e-05,
      "loss": 0.7623,
      "step": 5875
    },
    {
      "epoch": 1.26,
      "grad_norm": 11.673160552978516,
      "learning_rate": 1.1093951093951094e-05,
      "loss": 0.762,
      "step": 5876
    },
    {
      "epoch": 1.26,
      "grad_norm": 8.051241874694824,
      "learning_rate": 1.109073359073359e-05,
      "loss": 0.4641,
      "step": 5877
    },
    {
      "epoch": 1.26,
      "grad_norm": 11.026601791381836,
      "learning_rate": 1.1087516087516087e-05,
      "loss": 0.5541,
      "step": 5878
    },
    {
      "epoch": 1.26,
      "grad_norm": 13.399486541748047,
      "learning_rate": 1.1084298584298585e-05,
      "loss": 0.4845,
      "step": 5879
    },
    {
      "epoch": 1.26,
      "grad_norm": 12.599555015563965,
      "learning_rate": 1.1081081081081081e-05,
      "loss": 0.9485,
      "step": 5880
    },
    {
      "epoch": 1.26,
      "grad_norm": 12.497499465942383,
      "learning_rate": 1.1077863577863578e-05,
      "loss": 0.5882,
      "step": 5881
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.439507484436035,
      "learning_rate": 1.1074646074646076e-05,
      "loss": 0.5789,
      "step": 5882
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.613693237304688,
      "learning_rate": 1.1071428571428572e-05,
      "loss": 0.7607,
      "step": 5883
    },
    {
      "epoch": 1.26,
      "grad_norm": 8.103721618652344,
      "learning_rate": 1.1068211068211069e-05,
      "loss": 0.3909,
      "step": 5884
    },
    {
      "epoch": 1.26,
      "grad_norm": 11.714251518249512,
      "learning_rate": 1.1064993564993565e-05,
      "loss": 0.6587,
      "step": 5885
    },
    {
      "epoch": 1.26,
      "grad_norm": 14.737536430358887,
      "learning_rate": 1.1061776061776062e-05,
      "loss": 0.7294,
      "step": 5886
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.06994915008545,
      "learning_rate": 1.1058558558558558e-05,
      "loss": 0.5847,
      "step": 5887
    },
    {
      "epoch": 1.26,
      "grad_norm": 11.693260192871094,
      "learning_rate": 1.1055341055341056e-05,
      "loss": 0.6096,
      "step": 5888
    },
    {
      "epoch": 1.26,
      "grad_norm": 11.082918167114258,
      "learning_rate": 1.1052123552123553e-05,
      "loss": 0.571,
      "step": 5889
    },
    {
      "epoch": 1.26,
      "grad_norm": 7.91583251953125,
      "learning_rate": 1.104890604890605e-05,
      "loss": 0.3669,
      "step": 5890
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.57859992980957,
      "learning_rate": 1.1045688545688546e-05,
      "loss": 0.7054,
      "step": 5891
    },
    {
      "epoch": 1.26,
      "grad_norm": 11.135627746582031,
      "learning_rate": 1.1042471042471044e-05,
      "loss": 0.7001,
      "step": 5892
    },
    {
      "epoch": 1.26,
      "grad_norm": 13.18432903289795,
      "learning_rate": 1.1039253539253539e-05,
      "loss": 0.8396,
      "step": 5893
    },
    {
      "epoch": 1.26,
      "grad_norm": 12.696915626525879,
      "learning_rate": 1.1036036036036035e-05,
      "loss": 0.7244,
      "step": 5894
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.59287166595459,
      "learning_rate": 1.1032818532818533e-05,
      "loss": 0.6218,
      "step": 5895
    },
    {
      "epoch": 1.26,
      "grad_norm": 10.89894962310791,
      "learning_rate": 1.102960102960103e-05,
      "loss": 0.7649,
      "step": 5896
    },
    {
      "epoch": 1.26,
      "grad_norm": 12.376434326171875,
      "learning_rate": 1.1026383526383526e-05,
      "loss": 0.803,
      "step": 5897
    },
    {
      "epoch": 1.27,
      "grad_norm": 8.089430809020996,
      "learning_rate": 1.1023166023166024e-05,
      "loss": 0.4945,
      "step": 5898
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.35808277130127,
      "learning_rate": 1.101994851994852e-05,
      "loss": 0.6055,
      "step": 5899
    },
    {
      "epoch": 1.27,
      "grad_norm": 8.589822769165039,
      "learning_rate": 1.1016731016731017e-05,
      "loss": 0.6561,
      "step": 5900
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.171051025390625,
      "learning_rate": 1.1013513513513515e-05,
      "loss": 0.7619,
      "step": 5901
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.473565101623535,
      "learning_rate": 1.101029601029601e-05,
      "loss": 0.5253,
      "step": 5902
    },
    {
      "epoch": 1.27,
      "grad_norm": 8.482610702514648,
      "learning_rate": 1.1007078507078507e-05,
      "loss": 0.5472,
      "step": 5903
    },
    {
      "epoch": 1.27,
      "grad_norm": 12.702632904052734,
      "learning_rate": 1.1003861003861003e-05,
      "loss": 0.6131,
      "step": 5904
    },
    {
      "epoch": 1.27,
      "grad_norm": 11.361857414245605,
      "learning_rate": 1.1000643500643501e-05,
      "loss": 0.6769,
      "step": 5905
    },
    {
      "epoch": 1.27,
      "grad_norm": 12.311027526855469,
      "learning_rate": 1.0997425997425998e-05,
      "loss": 0.8,
      "step": 5906
    },
    {
      "epoch": 1.27,
      "grad_norm": 12.890204429626465,
      "learning_rate": 1.0994208494208494e-05,
      "loss": 0.5336,
      "step": 5907
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.247005462646484,
      "learning_rate": 1.0990990990990992e-05,
      "loss": 0.5916,
      "step": 5908
    },
    {
      "epoch": 1.27,
      "grad_norm": 12.591339111328125,
      "learning_rate": 1.0987773487773489e-05,
      "loss": 0.7753,
      "step": 5909
    },
    {
      "epoch": 1.27,
      "grad_norm": 14.070786476135254,
      "learning_rate": 1.0984555984555985e-05,
      "loss": 0.8174,
      "step": 5910
    },
    {
      "epoch": 1.27,
      "grad_norm": 13.119126319885254,
      "learning_rate": 1.0981338481338482e-05,
      "loss": 0.4889,
      "step": 5911
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.118582725524902,
      "learning_rate": 1.0978120978120978e-05,
      "loss": 0.6332,
      "step": 5912
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.695318222045898,
      "learning_rate": 1.0974903474903474e-05,
      "loss": 0.6832,
      "step": 5913
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.928569793701172,
      "learning_rate": 1.0971685971685973e-05,
      "loss": 0.7092,
      "step": 5914
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.120392799377441,
      "learning_rate": 1.0968468468468469e-05,
      "loss": 0.5645,
      "step": 5915
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.325758934020996,
      "learning_rate": 1.0965250965250965e-05,
      "loss": 0.6194,
      "step": 5916
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.478569984436035,
      "learning_rate": 1.0962033462033464e-05,
      "loss": 0.6868,
      "step": 5917
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.737464904785156,
      "learning_rate": 1.095881595881596e-05,
      "loss": 0.4566,
      "step": 5918
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.399658203125,
      "learning_rate": 1.0955598455598455e-05,
      "loss": 0.6935,
      "step": 5919
    },
    {
      "epoch": 1.27,
      "grad_norm": 8.453495979309082,
      "learning_rate": 1.0952380952380951e-05,
      "loss": 0.6636,
      "step": 5920
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.075772285461426,
      "learning_rate": 1.094916344916345e-05,
      "loss": 0.7071,
      "step": 5921
    },
    {
      "epoch": 1.27,
      "grad_norm": 13.510936737060547,
      "learning_rate": 1.0945945945945946e-05,
      "loss": 0.5738,
      "step": 5922
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.020852088928223,
      "learning_rate": 1.0942728442728442e-05,
      "loss": 0.6801,
      "step": 5923
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.477398872375488,
      "learning_rate": 1.093951093951094e-05,
      "loss": 0.5694,
      "step": 5924
    },
    {
      "epoch": 1.27,
      "grad_norm": 13.101412773132324,
      "learning_rate": 1.0936293436293437e-05,
      "loss": 0.6894,
      "step": 5925
    },
    {
      "epoch": 1.27,
      "grad_norm": 11.218331336975098,
      "learning_rate": 1.0933075933075933e-05,
      "loss": 0.6779,
      "step": 5926
    },
    {
      "epoch": 1.27,
      "grad_norm": 14.069428443908691,
      "learning_rate": 1.0929858429858432e-05,
      "loss": 0.7831,
      "step": 5927
    },
    {
      "epoch": 1.27,
      "grad_norm": 7.745060443878174,
      "learning_rate": 1.0926640926640926e-05,
      "loss": 0.4686,
      "step": 5928
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.850811004638672,
      "learning_rate": 1.0923423423423423e-05,
      "loss": 0.5217,
      "step": 5929
    },
    {
      "epoch": 1.27,
      "grad_norm": 6.895932674407959,
      "learning_rate": 1.0920205920205921e-05,
      "loss": 0.4351,
      "step": 5930
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.553257942199707,
      "learning_rate": 1.0916988416988417e-05,
      "loss": 0.6845,
      "step": 5931
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.380667686462402,
      "learning_rate": 1.0913770913770914e-05,
      "loss": 0.606,
      "step": 5932
    },
    {
      "epoch": 1.27,
      "grad_norm": 16.53299331665039,
      "learning_rate": 1.091055341055341e-05,
      "loss": 0.9151,
      "step": 5933
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.861746788024902,
      "learning_rate": 1.0907335907335908e-05,
      "loss": 0.7246,
      "step": 5934
    },
    {
      "epoch": 1.27,
      "grad_norm": 7.833929061889648,
      "learning_rate": 1.0904118404118405e-05,
      "loss": 0.4957,
      "step": 5935
    },
    {
      "epoch": 1.27,
      "grad_norm": 11.505667686462402,
      "learning_rate": 1.0900900900900901e-05,
      "loss": 0.546,
      "step": 5936
    },
    {
      "epoch": 1.27,
      "grad_norm": 8.868638038635254,
      "learning_rate": 1.0897683397683398e-05,
      "loss": 0.471,
      "step": 5937
    },
    {
      "epoch": 1.27,
      "grad_norm": 11.034613609313965,
      "learning_rate": 1.0894465894465894e-05,
      "loss": 0.5637,
      "step": 5938
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.480748176574707,
      "learning_rate": 1.089124839124839e-05,
      "loss": 0.5453,
      "step": 5939
    },
    {
      "epoch": 1.27,
      "grad_norm": 10.964192390441895,
      "learning_rate": 1.0888030888030889e-05,
      "loss": 0.8582,
      "step": 5940
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.930976867675781,
      "learning_rate": 1.0884813384813385e-05,
      "loss": 0.6454,
      "step": 5941
    },
    {
      "epoch": 1.27,
      "grad_norm": 11.36777114868164,
      "learning_rate": 1.0881595881595882e-05,
      "loss": 0.6808,
      "step": 5942
    },
    {
      "epoch": 1.27,
      "grad_norm": 14.076896667480469,
      "learning_rate": 1.087837837837838e-05,
      "loss": 0.8259,
      "step": 5943
    },
    {
      "epoch": 1.27,
      "grad_norm": 11.03832721710205,
      "learning_rate": 1.0875160875160876e-05,
      "loss": 0.4968,
      "step": 5944
    },
    {
      "epoch": 1.28,
      "grad_norm": 12.334924697875977,
      "learning_rate": 1.0871943371943371e-05,
      "loss": 0.6013,
      "step": 5945
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.006852149963379,
      "learning_rate": 1.086872586872587e-05,
      "loss": 0.8045,
      "step": 5946
    },
    {
      "epoch": 1.28,
      "grad_norm": 12.652599334716797,
      "learning_rate": 1.0865508365508366e-05,
      "loss": 0.9065,
      "step": 5947
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.753368377685547,
      "learning_rate": 1.0862290862290862e-05,
      "loss": 1.0424,
      "step": 5948
    },
    {
      "epoch": 1.28,
      "grad_norm": 9.256382942199707,
      "learning_rate": 1.0859073359073359e-05,
      "loss": 0.7443,
      "step": 5949
    },
    {
      "epoch": 1.28,
      "grad_norm": 9.258459091186523,
      "learning_rate": 1.0855855855855857e-05,
      "loss": 0.5686,
      "step": 5950
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.13781452178955,
      "learning_rate": 1.0852638352638353e-05,
      "loss": 0.6026,
      "step": 5951
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.676227569580078,
      "learning_rate": 1.084942084942085e-05,
      "loss": 0.5765,
      "step": 5952
    },
    {
      "epoch": 1.28,
      "grad_norm": 9.140128135681152,
      "learning_rate": 1.0846203346203348e-05,
      "loss": 0.4844,
      "step": 5953
    },
    {
      "epoch": 1.28,
      "grad_norm": 12.56234359741211,
      "learning_rate": 1.0842985842985842e-05,
      "loss": 0.9145,
      "step": 5954
    },
    {
      "epoch": 1.28,
      "grad_norm": 13.321176528930664,
      "learning_rate": 1.0839768339768339e-05,
      "loss": 0.6672,
      "step": 5955
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.17457389831543,
      "learning_rate": 1.0836550836550837e-05,
      "loss": 0.6798,
      "step": 5956
    },
    {
      "epoch": 1.28,
      "grad_norm": 9.982216835021973,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 0.5809,
      "step": 5957
    },
    {
      "epoch": 1.28,
      "grad_norm": 9.70875358581543,
      "learning_rate": 1.083011583011583e-05,
      "loss": 0.5759,
      "step": 5958
    },
    {
      "epoch": 1.28,
      "grad_norm": 7.038878440856934,
      "learning_rate": 1.0826898326898328e-05,
      "loss": 0.3728,
      "step": 5959
    },
    {
      "epoch": 1.28,
      "grad_norm": 8.707070350646973,
      "learning_rate": 1.0823680823680825e-05,
      "loss": 0.4711,
      "step": 5960
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.917287826538086,
      "learning_rate": 1.0820463320463321e-05,
      "loss": 0.7545,
      "step": 5961
    },
    {
      "epoch": 1.28,
      "grad_norm": 8.214947700500488,
      "learning_rate": 1.0817245817245817e-05,
      "loss": 0.4315,
      "step": 5962
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.604325294494629,
      "learning_rate": 1.0814028314028314e-05,
      "loss": 1.0676,
      "step": 5963
    },
    {
      "epoch": 1.28,
      "grad_norm": 9.657281875610352,
      "learning_rate": 1.081081081081081e-05,
      "loss": 0.6626,
      "step": 5964
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.18236255645752,
      "learning_rate": 1.0807593307593307e-05,
      "loss": 0.637,
      "step": 5965
    },
    {
      "epoch": 1.28,
      "grad_norm": 7.909306526184082,
      "learning_rate": 1.0804375804375805e-05,
      "loss": 0.4674,
      "step": 5966
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.748529434204102,
      "learning_rate": 1.0801158301158301e-05,
      "loss": 0.6929,
      "step": 5967
    },
    {
      "epoch": 1.28,
      "grad_norm": 9.624189376831055,
      "learning_rate": 1.0797940797940798e-05,
      "loss": 0.5787,
      "step": 5968
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.162662506103516,
      "learning_rate": 1.0794723294723296e-05,
      "loss": 0.5184,
      "step": 5969
    },
    {
      "epoch": 1.28,
      "grad_norm": 8.394309997558594,
      "learning_rate": 1.0791505791505792e-05,
      "loss": 0.5289,
      "step": 5970
    },
    {
      "epoch": 1.28,
      "grad_norm": 13.622846603393555,
      "learning_rate": 1.0788288288288289e-05,
      "loss": 0.8719,
      "step": 5971
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.343360900878906,
      "learning_rate": 1.0785070785070785e-05,
      "loss": 0.6947,
      "step": 5972
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.364758491516113,
      "learning_rate": 1.0781853281853282e-05,
      "loss": 0.5746,
      "step": 5973
    },
    {
      "epoch": 1.28,
      "grad_norm": 8.318331718444824,
      "learning_rate": 1.0778635778635778e-05,
      "loss": 0.3789,
      "step": 5974
    },
    {
      "epoch": 1.28,
      "grad_norm": 14.533262252807617,
      "learning_rate": 1.0775418275418276e-05,
      "loss": 0.5872,
      "step": 5975
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.44509220123291,
      "learning_rate": 1.0772200772200773e-05,
      "loss": 0.6508,
      "step": 5976
    },
    {
      "epoch": 1.28,
      "grad_norm": 7.9016032218933105,
      "learning_rate": 1.076898326898327e-05,
      "loss": 0.5199,
      "step": 5977
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.730363845825195,
      "learning_rate": 1.0765765765765766e-05,
      "loss": 0.6162,
      "step": 5978
    },
    {
      "epoch": 1.28,
      "grad_norm": 8.652939796447754,
      "learning_rate": 1.0762548262548264e-05,
      "loss": 0.6279,
      "step": 5979
    },
    {
      "epoch": 1.28,
      "grad_norm": 9.453096389770508,
      "learning_rate": 1.0759330759330759e-05,
      "loss": 0.5438,
      "step": 5980
    },
    {
      "epoch": 1.28,
      "grad_norm": 14.033077239990234,
      "learning_rate": 1.0756113256113255e-05,
      "loss": 0.5599,
      "step": 5981
    },
    {
      "epoch": 1.28,
      "grad_norm": 9.836490631103516,
      "learning_rate": 1.0752895752895753e-05,
      "loss": 0.4865,
      "step": 5982
    },
    {
      "epoch": 1.28,
      "grad_norm": 12.112929344177246,
      "learning_rate": 1.074967824967825e-05,
      "loss": 0.9051,
      "step": 5983
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.043973922729492,
      "learning_rate": 1.0746460746460746e-05,
      "loss": 0.732,
      "step": 5984
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.007863998413086,
      "learning_rate": 1.0743243243243244e-05,
      "loss": 0.4869,
      "step": 5985
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.099566459655762,
      "learning_rate": 1.074002574002574e-05,
      "loss": 0.5505,
      "step": 5986
    },
    {
      "epoch": 1.28,
      "grad_norm": 12.504559516906738,
      "learning_rate": 1.0736808236808237e-05,
      "loss": 0.6067,
      "step": 5987
    },
    {
      "epoch": 1.28,
      "grad_norm": 8.697098731994629,
      "learning_rate": 1.0733590733590735e-05,
      "loss": 0.6251,
      "step": 5988
    },
    {
      "epoch": 1.28,
      "grad_norm": 9.76908016204834,
      "learning_rate": 1.073037323037323e-05,
      "loss": 0.6723,
      "step": 5989
    },
    {
      "epoch": 1.28,
      "grad_norm": 8.333545684814453,
      "learning_rate": 1.0727155727155727e-05,
      "loss": 0.4931,
      "step": 5990
    },
    {
      "epoch": 1.29,
      "grad_norm": 6.834651947021484,
      "learning_rate": 1.0723938223938225e-05,
      "loss": 0.3617,
      "step": 5991
    },
    {
      "epoch": 1.29,
      "grad_norm": 9.492420196533203,
      "learning_rate": 1.0720720720720721e-05,
      "loss": 0.568,
      "step": 5992
    },
    {
      "epoch": 1.29,
      "grad_norm": 8.453141212463379,
      "learning_rate": 1.0717503217503218e-05,
      "loss": 0.5178,
      "step": 5993
    },
    {
      "epoch": 1.29,
      "grad_norm": 7.411289691925049,
      "learning_rate": 1.0714285714285714e-05,
      "loss": 0.4229,
      "step": 5994
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.842796325683594,
      "learning_rate": 1.0711068211068212e-05,
      "loss": 0.5621,
      "step": 5995
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.842342376708984,
      "learning_rate": 1.0707850707850709e-05,
      "loss": 0.5442,
      "step": 5996
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.044440269470215,
      "learning_rate": 1.0704633204633205e-05,
      "loss": 0.5934,
      "step": 5997
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.213736534118652,
      "learning_rate": 1.0701415701415702e-05,
      "loss": 0.7051,
      "step": 5998
    },
    {
      "epoch": 1.29,
      "grad_norm": 12.734705924987793,
      "learning_rate": 1.0698198198198198e-05,
      "loss": 0.6673,
      "step": 5999
    },
    {
      "epoch": 1.29,
      "grad_norm": 12.299291610717773,
      "learning_rate": 1.0694980694980694e-05,
      "loss": 0.6063,
      "step": 6000
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.97014045715332,
      "learning_rate": 1.0691763191763193e-05,
      "loss": 0.699,
      "step": 6001
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.77499008178711,
      "learning_rate": 1.0688545688545689e-05,
      "loss": 0.5816,
      "step": 6002
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.817117691040039,
      "learning_rate": 1.0685328185328186e-05,
      "loss": 0.6563,
      "step": 6003
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.092387199401855,
      "learning_rate": 1.0682110682110684e-05,
      "loss": 0.5437,
      "step": 6004
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.875229835510254,
      "learning_rate": 1.067889317889318e-05,
      "loss": 0.5765,
      "step": 6005
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.791181564331055,
      "learning_rate": 1.0675675675675675e-05,
      "loss": 0.7137,
      "step": 6006
    },
    {
      "epoch": 1.29,
      "grad_norm": 9.551329612731934,
      "learning_rate": 1.0672458172458171e-05,
      "loss": 0.5319,
      "step": 6007
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.691040992736816,
      "learning_rate": 1.066924066924067e-05,
      "loss": 0.802,
      "step": 6008
    },
    {
      "epoch": 1.29,
      "grad_norm": 8.78536319732666,
      "learning_rate": 1.0666023166023166e-05,
      "loss": 0.5614,
      "step": 6009
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.391691207885742,
      "learning_rate": 1.0662805662805662e-05,
      "loss": 0.6083,
      "step": 6010
    },
    {
      "epoch": 1.29,
      "grad_norm": 9.854461669921875,
      "learning_rate": 1.065958815958816e-05,
      "loss": 0.5452,
      "step": 6011
    },
    {
      "epoch": 1.29,
      "grad_norm": 9.183195114135742,
      "learning_rate": 1.0656370656370657e-05,
      "loss": 0.527,
      "step": 6012
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.745869636535645,
      "learning_rate": 1.0653153153153153e-05,
      "loss": 0.6542,
      "step": 6013
    },
    {
      "epoch": 1.29,
      "grad_norm": 9.900657653808594,
      "learning_rate": 1.0649935649935652e-05,
      "loss": 0.5988,
      "step": 6014
    },
    {
      "epoch": 1.29,
      "grad_norm": 13.211163520812988,
      "learning_rate": 1.0646718146718146e-05,
      "loss": 0.7316,
      "step": 6015
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.244028091430664,
      "learning_rate": 1.0643500643500643e-05,
      "loss": 0.6269,
      "step": 6016
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.317548751831055,
      "learning_rate": 1.0640283140283141e-05,
      "loss": 0.4961,
      "step": 6017
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.12273120880127,
      "learning_rate": 1.0637065637065637e-05,
      "loss": 0.7319,
      "step": 6018
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.802428245544434,
      "learning_rate": 1.0633848133848134e-05,
      "loss": 0.6336,
      "step": 6019
    },
    {
      "epoch": 1.29,
      "grad_norm": 8.173173904418945,
      "learning_rate": 1.063063063063063e-05,
      "loss": 0.3557,
      "step": 6020
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.540935516357422,
      "learning_rate": 1.0627413127413128e-05,
      "loss": 0.7698,
      "step": 6021
    },
    {
      "epoch": 1.29,
      "grad_norm": 9.587428092956543,
      "learning_rate": 1.0624195624195625e-05,
      "loss": 0.4545,
      "step": 6022
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.569751739501953,
      "learning_rate": 1.0620978120978121e-05,
      "loss": 0.6332,
      "step": 6023
    },
    {
      "epoch": 1.29,
      "grad_norm": 9.664766311645508,
      "learning_rate": 1.0617760617760618e-05,
      "loss": 0.5787,
      "step": 6024
    },
    {
      "epoch": 1.29,
      "grad_norm": 12.839691162109375,
      "learning_rate": 1.0614543114543114e-05,
      "loss": 0.7893,
      "step": 6025
    },
    {
      "epoch": 1.29,
      "grad_norm": 13.297111511230469,
      "learning_rate": 1.061132561132561e-05,
      "loss": 0.8191,
      "step": 6026
    },
    {
      "epoch": 1.29,
      "grad_norm": 9.782763481140137,
      "learning_rate": 1.0608108108108109e-05,
      "loss": 0.5678,
      "step": 6027
    },
    {
      "epoch": 1.29,
      "grad_norm": 9.565861701965332,
      "learning_rate": 1.0604890604890605e-05,
      "loss": 0.6213,
      "step": 6028
    },
    {
      "epoch": 1.29,
      "grad_norm": 10.24106216430664,
      "learning_rate": 1.0601673101673102e-05,
      "loss": 0.5709,
      "step": 6029
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.785039901733398,
      "learning_rate": 1.05984555984556e-05,
      "loss": 0.774,
      "step": 6030
    },
    {
      "epoch": 1.29,
      "grad_norm": 8.212697982788086,
      "learning_rate": 1.0595238095238096e-05,
      "loss": 0.4829,
      "step": 6031
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.953965187072754,
      "learning_rate": 1.0592020592020591e-05,
      "loss": 0.6826,
      "step": 6032
    },
    {
      "epoch": 1.29,
      "grad_norm": 12.845919609069824,
      "learning_rate": 1.058880308880309e-05,
      "loss": 0.6385,
      "step": 6033
    },
    {
      "epoch": 1.29,
      "grad_norm": 8.196592330932617,
      "learning_rate": 1.0585585585585586e-05,
      "loss": 0.5088,
      "step": 6034
    },
    {
      "epoch": 1.29,
      "grad_norm": 9.382698059082031,
      "learning_rate": 1.0582368082368082e-05,
      "loss": 0.5444,
      "step": 6035
    },
    {
      "epoch": 1.29,
      "grad_norm": 12.325982093811035,
      "learning_rate": 1.0579150579150579e-05,
      "loss": 0.7308,
      "step": 6036
    },
    {
      "epoch": 1.29,
      "grad_norm": 13.197925567626953,
      "learning_rate": 1.0575933075933077e-05,
      "loss": 0.7245,
      "step": 6037
    },
    {
      "epoch": 1.3,
      "grad_norm": 13.197785377502441,
      "learning_rate": 1.0572715572715573e-05,
      "loss": 0.6482,
      "step": 6038
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.996304512023926,
      "learning_rate": 1.056949806949807e-05,
      "loss": 0.5711,
      "step": 6039
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.281262397766113,
      "learning_rate": 1.0566280566280568e-05,
      "loss": 0.6656,
      "step": 6040
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.986393928527832,
      "learning_rate": 1.0563063063063063e-05,
      "loss": 0.6525,
      "step": 6041
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.201112747192383,
      "learning_rate": 1.0559845559845559e-05,
      "loss": 0.5364,
      "step": 6042
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.594178199768066,
      "learning_rate": 1.0556628056628057e-05,
      "loss": 0.7288,
      "step": 6043
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.957505226135254,
      "learning_rate": 1.0553410553410554e-05,
      "loss": 0.665,
      "step": 6044
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.935213088989258,
      "learning_rate": 1.055019305019305e-05,
      "loss": 0.5562,
      "step": 6045
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.387271881103516,
      "learning_rate": 1.0546975546975548e-05,
      "loss": 0.5068,
      "step": 6046
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.32866382598877,
      "learning_rate": 1.0543758043758045e-05,
      "loss": 0.6443,
      "step": 6047
    },
    {
      "epoch": 1.3,
      "grad_norm": 8.966691017150879,
      "learning_rate": 1.0540540540540541e-05,
      "loss": 0.5639,
      "step": 6048
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.871793746948242,
      "learning_rate": 1.0537323037323038e-05,
      "loss": 0.5968,
      "step": 6049
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.961474418640137,
      "learning_rate": 1.0534105534105534e-05,
      "loss": 0.581,
      "step": 6050
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.851400375366211,
      "learning_rate": 1.053088803088803e-05,
      "loss": 0.5645,
      "step": 6051
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.974244117736816,
      "learning_rate": 1.0527670527670527e-05,
      "loss": 0.5051,
      "step": 6052
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.700910568237305,
      "learning_rate": 1.0524453024453025e-05,
      "loss": 0.554,
      "step": 6053
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.845819473266602,
      "learning_rate": 1.0521235521235521e-05,
      "loss": 0.7156,
      "step": 6054
    },
    {
      "epoch": 1.3,
      "grad_norm": 7.905458450317383,
      "learning_rate": 1.0518018018018018e-05,
      "loss": 0.3566,
      "step": 6055
    },
    {
      "epoch": 1.3,
      "grad_norm": 8.090807914733887,
      "learning_rate": 1.0514800514800516e-05,
      "loss": 0.5485,
      "step": 6056
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.99112319946289,
      "learning_rate": 1.0511583011583013e-05,
      "loss": 0.6955,
      "step": 6057
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.629436492919922,
      "learning_rate": 1.0508365508365507e-05,
      "loss": 0.6374,
      "step": 6058
    },
    {
      "epoch": 1.3,
      "grad_norm": 7.553342342376709,
      "learning_rate": 1.0505148005148005e-05,
      "loss": 0.3444,
      "step": 6059
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.495224952697754,
      "learning_rate": 1.0501930501930502e-05,
      "loss": 0.5771,
      "step": 6060
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.720904350280762,
      "learning_rate": 1.0498712998712998e-05,
      "loss": 0.6202,
      "step": 6061
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.901214599609375,
      "learning_rate": 1.0495495495495496e-05,
      "loss": 0.7207,
      "step": 6062
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.528382301330566,
      "learning_rate": 1.0492277992277993e-05,
      "loss": 0.605,
      "step": 6063
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.891141891479492,
      "learning_rate": 1.048906048906049e-05,
      "loss": 0.7679,
      "step": 6064
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.327199935913086,
      "learning_rate": 1.0485842985842986e-05,
      "loss": 0.653,
      "step": 6065
    },
    {
      "epoch": 1.3,
      "grad_norm": 8.58317756652832,
      "learning_rate": 1.0482625482625484e-05,
      "loss": 0.5144,
      "step": 6066
    },
    {
      "epoch": 1.3,
      "grad_norm": 13.801733016967773,
      "learning_rate": 1.0479407979407979e-05,
      "loss": 0.6096,
      "step": 6067
    },
    {
      "epoch": 1.3,
      "grad_norm": 17.42502784729004,
      "learning_rate": 1.0476190476190475e-05,
      "loss": 0.4457,
      "step": 6068
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.246071815490723,
      "learning_rate": 1.0472972972972973e-05,
      "loss": 0.6212,
      "step": 6069
    },
    {
      "epoch": 1.3,
      "grad_norm": 6.774965286254883,
      "learning_rate": 1.046975546975547e-05,
      "loss": 0.52,
      "step": 6070
    },
    {
      "epoch": 1.3,
      "grad_norm": 8.916576385498047,
      "learning_rate": 1.0466537966537966e-05,
      "loss": 0.5078,
      "step": 6071
    },
    {
      "epoch": 1.3,
      "grad_norm": 10.871039390563965,
      "learning_rate": 1.0463320463320464e-05,
      "loss": 0.6021,
      "step": 6072
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.488404273986816,
      "learning_rate": 1.046010296010296e-05,
      "loss": 0.4806,
      "step": 6073
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.945384979248047,
      "learning_rate": 1.0456885456885457e-05,
      "loss": 0.741,
      "step": 6074
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.43740177154541,
      "learning_rate": 1.0453667953667955e-05,
      "loss": 0.5614,
      "step": 6075
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.003300666809082,
      "learning_rate": 1.045045045045045e-05,
      "loss": 0.5637,
      "step": 6076
    },
    {
      "epoch": 1.3,
      "grad_norm": 7.936720848083496,
      "learning_rate": 1.0447232947232947e-05,
      "loss": 0.3217,
      "step": 6077
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.058335304260254,
      "learning_rate": 1.0444015444015445e-05,
      "loss": 0.5969,
      "step": 6078
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.804239273071289,
      "learning_rate": 1.0440797940797941e-05,
      "loss": 0.7102,
      "step": 6079
    },
    {
      "epoch": 1.3,
      "grad_norm": 11.147397994995117,
      "learning_rate": 1.0437580437580438e-05,
      "loss": 0.7399,
      "step": 6080
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.779451370239258,
      "learning_rate": 1.0434362934362934e-05,
      "loss": 0.4633,
      "step": 6081
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.274054527282715,
      "learning_rate": 1.0431145431145432e-05,
      "loss": 0.7185,
      "step": 6082
    },
    {
      "epoch": 1.3,
      "grad_norm": 8.040630340576172,
      "learning_rate": 1.0427927927927929e-05,
      "loss": 0.4596,
      "step": 6083
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.892231941223145,
      "learning_rate": 1.0424710424710425e-05,
      "loss": 0.6932,
      "step": 6084
    },
    {
      "epoch": 1.31,
      "grad_norm": 8.458772659301758,
      "learning_rate": 1.0421492921492922e-05,
      "loss": 0.5338,
      "step": 6085
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.622815132141113,
      "learning_rate": 1.0418275418275418e-05,
      "loss": 0.5095,
      "step": 6086
    },
    {
      "epoch": 1.31,
      "grad_norm": 8.99202823638916,
      "learning_rate": 1.0415057915057915e-05,
      "loss": 0.5569,
      "step": 6087
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.051015853881836,
      "learning_rate": 1.0411840411840413e-05,
      "loss": 0.5878,
      "step": 6088
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.007990837097168,
      "learning_rate": 1.0408622908622909e-05,
      "loss": 0.5894,
      "step": 6089
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.420199394226074,
      "learning_rate": 1.0405405405405406e-05,
      "loss": 0.6184,
      "step": 6090
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.483840942382812,
      "learning_rate": 1.0402187902187904e-05,
      "loss": 0.6515,
      "step": 6091
    },
    {
      "epoch": 1.31,
      "grad_norm": 8.592266082763672,
      "learning_rate": 1.03989703989704e-05,
      "loss": 0.5331,
      "step": 6092
    },
    {
      "epoch": 1.31,
      "grad_norm": 8.767326354980469,
      "learning_rate": 1.0395752895752895e-05,
      "loss": 0.5334,
      "step": 6093
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.569958686828613,
      "learning_rate": 1.0392535392535391e-05,
      "loss": 0.6251,
      "step": 6094
    },
    {
      "epoch": 1.31,
      "grad_norm": 8.763664245605469,
      "learning_rate": 1.038931788931789e-05,
      "loss": 0.528,
      "step": 6095
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.70541000366211,
      "learning_rate": 1.0386100386100386e-05,
      "loss": 0.3863,
      "step": 6096
    },
    {
      "epoch": 1.31,
      "grad_norm": 13.989203453063965,
      "learning_rate": 1.0382882882882882e-05,
      "loss": 0.7319,
      "step": 6097
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.605998992919922,
      "learning_rate": 1.037966537966538e-05,
      "loss": 0.6175,
      "step": 6098
    },
    {
      "epoch": 1.31,
      "grad_norm": 12.375240325927734,
      "learning_rate": 1.0376447876447877e-05,
      "loss": 0.6014,
      "step": 6099
    },
    {
      "epoch": 1.31,
      "grad_norm": 15.91203784942627,
      "learning_rate": 1.0373230373230373e-05,
      "loss": 0.7958,
      "step": 6100
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.767467498779297,
      "learning_rate": 1.0370012870012872e-05,
      "loss": 0.5744,
      "step": 6101
    },
    {
      "epoch": 1.31,
      "grad_norm": 13.160073280334473,
      "learning_rate": 1.0366795366795366e-05,
      "loss": 0.6269,
      "step": 6102
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.98611068725586,
      "learning_rate": 1.0363577863577863e-05,
      "loss": 0.5974,
      "step": 6103
    },
    {
      "epoch": 1.31,
      "grad_norm": 13.801783561706543,
      "learning_rate": 1.0360360360360361e-05,
      "loss": 0.5688,
      "step": 6104
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.349942207336426,
      "learning_rate": 1.0357142857142857e-05,
      "loss": 0.4623,
      "step": 6105
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.193178176879883,
      "learning_rate": 1.0353925353925354e-05,
      "loss": 0.5437,
      "step": 6106
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.873241424560547,
      "learning_rate": 1.035070785070785e-05,
      "loss": 0.6938,
      "step": 6107
    },
    {
      "epoch": 1.31,
      "grad_norm": 13.427983283996582,
      "learning_rate": 1.0347490347490348e-05,
      "loss": 0.9501,
      "step": 6108
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.02061939239502,
      "learning_rate": 1.0344272844272845e-05,
      "loss": 0.5289,
      "step": 6109
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.339675903320312,
      "learning_rate": 1.0341055341055341e-05,
      "loss": 0.3989,
      "step": 6110
    },
    {
      "epoch": 1.31,
      "grad_norm": 12.295023918151855,
      "learning_rate": 1.0337837837837838e-05,
      "loss": 0.638,
      "step": 6111
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.279026985168457,
      "learning_rate": 1.0334620334620334e-05,
      "loss": 0.5803,
      "step": 6112
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.136301040649414,
      "learning_rate": 1.033140283140283e-05,
      "loss": 0.5194,
      "step": 6113
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.651440620422363,
      "learning_rate": 1.0328185328185329e-05,
      "loss": 0.8549,
      "step": 6114
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.841410636901855,
      "learning_rate": 1.0324967824967825e-05,
      "loss": 0.5329,
      "step": 6115
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.316143035888672,
      "learning_rate": 1.0321750321750322e-05,
      "loss": 0.5641,
      "step": 6116
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.013443946838379,
      "learning_rate": 1.031853281853282e-05,
      "loss": 0.4244,
      "step": 6117
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.130349159240723,
      "learning_rate": 1.0315315315315316e-05,
      "loss": 0.7463,
      "step": 6118
    },
    {
      "epoch": 1.31,
      "grad_norm": 12.12076473236084,
      "learning_rate": 1.0312097812097811e-05,
      "loss": 0.5259,
      "step": 6119
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.190781593322754,
      "learning_rate": 1.030888030888031e-05,
      "loss": 0.8349,
      "step": 6120
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.387514114379883,
      "learning_rate": 1.0305662805662806e-05,
      "loss": 0.6136,
      "step": 6121
    },
    {
      "epoch": 1.31,
      "grad_norm": 18.16321563720703,
      "learning_rate": 1.0302445302445302e-05,
      "loss": 0.7216,
      "step": 6122
    },
    {
      "epoch": 1.31,
      "grad_norm": 12.616767883300781,
      "learning_rate": 1.0299227799227799e-05,
      "loss": 0.7898,
      "step": 6123
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.561367988586426,
      "learning_rate": 1.0296010296010297e-05,
      "loss": 0.6377,
      "step": 6124
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.260821342468262,
      "learning_rate": 1.0292792792792793e-05,
      "loss": 0.8423,
      "step": 6125
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.591438293457031,
      "learning_rate": 1.028957528957529e-05,
      "loss": 0.6371,
      "step": 6126
    },
    {
      "epoch": 1.31,
      "grad_norm": 9.785163879394531,
      "learning_rate": 1.0286357786357788e-05,
      "loss": 0.5042,
      "step": 6127
    },
    {
      "epoch": 1.31,
      "grad_norm": 11.726027488708496,
      "learning_rate": 1.0283140283140283e-05,
      "loss": 0.6192,
      "step": 6128
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.436243057250977,
      "learning_rate": 1.0279922779922779e-05,
      "loss": 0.58,
      "step": 6129
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.660564422607422,
      "learning_rate": 1.0276705276705277e-05,
      "loss": 0.7263,
      "step": 6130
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.52206039428711,
      "learning_rate": 1.0273487773487774e-05,
      "loss": 0.5806,
      "step": 6131
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.536266326904297,
      "learning_rate": 1.027027027027027e-05,
      "loss": 0.6661,
      "step": 6132
    },
    {
      "epoch": 1.32,
      "grad_norm": 12.693780899047852,
      "learning_rate": 1.0267052767052768e-05,
      "loss": 0.5464,
      "step": 6133
    },
    {
      "epoch": 1.32,
      "grad_norm": 12.499454498291016,
      "learning_rate": 1.0263835263835265e-05,
      "loss": 0.528,
      "step": 6134
    },
    {
      "epoch": 1.32,
      "grad_norm": 11.853631019592285,
      "learning_rate": 1.0260617760617761e-05,
      "loss": 0.5664,
      "step": 6135
    },
    {
      "epoch": 1.32,
      "grad_norm": 8.068376541137695,
      "learning_rate": 1.0257400257400258e-05,
      "loss": 0.4201,
      "step": 6136
    },
    {
      "epoch": 1.32,
      "grad_norm": 11.710026741027832,
      "learning_rate": 1.0254182754182754e-05,
      "loss": 0.6027,
      "step": 6137
    },
    {
      "epoch": 1.32,
      "grad_norm": 13.470458984375,
      "learning_rate": 1.025096525096525e-05,
      "loss": 0.6705,
      "step": 6138
    },
    {
      "epoch": 1.32,
      "grad_norm": 11.141728401184082,
      "learning_rate": 1.0247747747747747e-05,
      "loss": 0.8012,
      "step": 6139
    },
    {
      "epoch": 1.32,
      "grad_norm": 14.306894302368164,
      "learning_rate": 1.0244530244530245e-05,
      "loss": 0.7711,
      "step": 6140
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.827816009521484,
      "learning_rate": 1.0241312741312742e-05,
      "loss": 0.51,
      "step": 6141
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.22120189666748,
      "learning_rate": 1.0238095238095238e-05,
      "loss": 0.5853,
      "step": 6142
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.537413597106934,
      "learning_rate": 1.0234877734877736e-05,
      "loss": 0.6292,
      "step": 6143
    },
    {
      "epoch": 1.32,
      "grad_norm": 13.543390274047852,
      "learning_rate": 1.0231660231660233e-05,
      "loss": 0.4772,
      "step": 6144
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.440423965454102,
      "learning_rate": 1.0228442728442727e-05,
      "loss": 0.6869,
      "step": 6145
    },
    {
      "epoch": 1.32,
      "grad_norm": 11.886975288391113,
      "learning_rate": 1.0225225225225226e-05,
      "loss": 0.5559,
      "step": 6146
    },
    {
      "epoch": 1.32,
      "grad_norm": 8.344897270202637,
      "learning_rate": 1.0222007722007722e-05,
      "loss": 0.4972,
      "step": 6147
    },
    {
      "epoch": 1.32,
      "grad_norm": 8.476862907409668,
      "learning_rate": 1.0218790218790218e-05,
      "loss": 0.6206,
      "step": 6148
    },
    {
      "epoch": 1.32,
      "grad_norm": 8.85732650756836,
      "learning_rate": 1.0215572715572717e-05,
      "loss": 0.5548,
      "step": 6149
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.374959945678711,
      "learning_rate": 1.0212355212355213e-05,
      "loss": 0.766,
      "step": 6150
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.766547203063965,
      "learning_rate": 1.020913770913771e-05,
      "loss": 0.5941,
      "step": 6151
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.926553726196289,
      "learning_rate": 1.0205920205920206e-05,
      "loss": 0.7162,
      "step": 6152
    },
    {
      "epoch": 1.32,
      "grad_norm": 7.994480133056641,
      "learning_rate": 1.0202702702702704e-05,
      "loss": 0.7897,
      "step": 6153
    },
    {
      "epoch": 1.32,
      "grad_norm": 15.259339332580566,
      "learning_rate": 1.0199485199485199e-05,
      "loss": 0.9003,
      "step": 6154
    },
    {
      "epoch": 1.32,
      "grad_norm": 7.898843288421631,
      "learning_rate": 1.0196267696267695e-05,
      "loss": 0.4825,
      "step": 6155
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.414298057556152,
      "learning_rate": 1.0193050193050193e-05,
      "loss": 0.5677,
      "step": 6156
    },
    {
      "epoch": 1.32,
      "grad_norm": 12.564446449279785,
      "learning_rate": 1.018983268983269e-05,
      "loss": 0.706,
      "step": 6157
    },
    {
      "epoch": 1.32,
      "grad_norm": 7.988409042358398,
      "learning_rate": 1.0186615186615186e-05,
      "loss": 0.566,
      "step": 6158
    },
    {
      "epoch": 1.32,
      "grad_norm": 12.350157737731934,
      "learning_rate": 1.0183397683397684e-05,
      "loss": 0.7241,
      "step": 6159
    },
    {
      "epoch": 1.32,
      "grad_norm": 7.283298015594482,
      "learning_rate": 1.0180180180180181e-05,
      "loss": 0.5264,
      "step": 6160
    },
    {
      "epoch": 1.32,
      "grad_norm": 11.783570289611816,
      "learning_rate": 1.0176962676962677e-05,
      "loss": 0.4172,
      "step": 6161
    },
    {
      "epoch": 1.32,
      "grad_norm": 8.625517845153809,
      "learning_rate": 1.0173745173745176e-05,
      "loss": 0.4688,
      "step": 6162
    },
    {
      "epoch": 1.32,
      "grad_norm": 13.186787605285645,
      "learning_rate": 1.017052767052767e-05,
      "loss": 0.6737,
      "step": 6163
    },
    {
      "epoch": 1.32,
      "grad_norm": 12.384242057800293,
      "learning_rate": 1.0167310167310167e-05,
      "loss": 0.6282,
      "step": 6164
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.522583961486816,
      "learning_rate": 1.0164092664092665e-05,
      "loss": 0.7135,
      "step": 6165
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.851083755493164,
      "learning_rate": 1.0160875160875161e-05,
      "loss": 0.681,
      "step": 6166
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.224080085754395,
      "learning_rate": 1.0157657657657658e-05,
      "loss": 0.5172,
      "step": 6167
    },
    {
      "epoch": 1.32,
      "grad_norm": 12.075817108154297,
      "learning_rate": 1.0154440154440154e-05,
      "loss": 0.6521,
      "step": 6168
    },
    {
      "epoch": 1.32,
      "grad_norm": 8.07058334350586,
      "learning_rate": 1.0151222651222652e-05,
      "loss": 0.3619,
      "step": 6169
    },
    {
      "epoch": 1.32,
      "grad_norm": 11.827540397644043,
      "learning_rate": 1.0148005148005149e-05,
      "loss": 0.7506,
      "step": 6170
    },
    {
      "epoch": 1.32,
      "grad_norm": 7.974662780761719,
      "learning_rate": 1.0144787644787645e-05,
      "loss": 0.4645,
      "step": 6171
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.14547061920166,
      "learning_rate": 1.0141570141570142e-05,
      "loss": 0.5831,
      "step": 6172
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.3300142288208,
      "learning_rate": 1.0138352638352638e-05,
      "loss": 0.6984,
      "step": 6173
    },
    {
      "epoch": 1.32,
      "grad_norm": 12.861353874206543,
      "learning_rate": 1.0135135135135135e-05,
      "loss": 0.7784,
      "step": 6174
    },
    {
      "epoch": 1.32,
      "grad_norm": 11.371753692626953,
      "learning_rate": 1.0131917631917633e-05,
      "loss": 0.5525,
      "step": 6175
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.054768562316895,
      "learning_rate": 1.012870012870013e-05,
      "loss": 0.6727,
      "step": 6176
    },
    {
      "epoch": 1.32,
      "grad_norm": 10.945122718811035,
      "learning_rate": 1.0125482625482626e-05,
      "loss": 0.5631,
      "step": 6177
    },
    {
      "epoch": 1.33,
      "grad_norm": 14.186384201049805,
      "learning_rate": 1.0122265122265124e-05,
      "loss": 0.8172,
      "step": 6178
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.259407043457031,
      "learning_rate": 1.011904761904762e-05,
      "loss": 0.6955,
      "step": 6179
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.533349990844727,
      "learning_rate": 1.0115830115830115e-05,
      "loss": 0.6697,
      "step": 6180
    },
    {
      "epoch": 1.33,
      "grad_norm": 13.734477043151855,
      "learning_rate": 1.0112612612612611e-05,
      "loss": 0.8134,
      "step": 6181
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.275354385375977,
      "learning_rate": 1.010939510939511e-05,
      "loss": 0.6627,
      "step": 6182
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.085680961608887,
      "learning_rate": 1.0106177606177606e-05,
      "loss": 0.7308,
      "step": 6183
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.884145736694336,
      "learning_rate": 1.0102960102960103e-05,
      "loss": 0.6147,
      "step": 6184
    },
    {
      "epoch": 1.33,
      "grad_norm": 13.290209770202637,
      "learning_rate": 1.00997425997426e-05,
      "loss": 0.564,
      "step": 6185
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.678720474243164,
      "learning_rate": 1.0096525096525097e-05,
      "loss": 0.4282,
      "step": 6186
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.984843254089355,
      "learning_rate": 1.0093307593307594e-05,
      "loss": 0.7985,
      "step": 6187
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.258391380310059,
      "learning_rate": 1.0090090090090092e-05,
      "loss": 0.6467,
      "step": 6188
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.12476921081543,
      "learning_rate": 1.0086872586872586e-05,
      "loss": 0.574,
      "step": 6189
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.117472648620605,
      "learning_rate": 1.0083655083655083e-05,
      "loss": 0.7901,
      "step": 6190
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.655936241149902,
      "learning_rate": 1.0080437580437581e-05,
      "loss": 0.8352,
      "step": 6191
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.79865550994873,
      "learning_rate": 1.0077220077220078e-05,
      "loss": 0.6477,
      "step": 6192
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.376649856567383,
      "learning_rate": 1.0074002574002574e-05,
      "loss": 0.5184,
      "step": 6193
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.18793773651123,
      "learning_rate": 1.0070785070785072e-05,
      "loss": 0.4964,
      "step": 6194
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.875333786010742,
      "learning_rate": 1.0067567567567569e-05,
      "loss": 0.5969,
      "step": 6195
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.576823234558105,
      "learning_rate": 1.0064350064350065e-05,
      "loss": 0.488,
      "step": 6196
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.38052749633789,
      "learning_rate": 1.0061132561132561e-05,
      "loss": 0.596,
      "step": 6197
    },
    {
      "epoch": 1.33,
      "grad_norm": 6.698428630828857,
      "learning_rate": 1.0057915057915058e-05,
      "loss": 0.3541,
      "step": 6198
    },
    {
      "epoch": 1.33,
      "grad_norm": 7.068019866943359,
      "learning_rate": 1.0054697554697554e-05,
      "loss": 0.5449,
      "step": 6199
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.131220817565918,
      "learning_rate": 1.005148005148005e-05,
      "loss": 0.4255,
      "step": 6200
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.750980377197266,
      "learning_rate": 1.0048262548262549e-05,
      "loss": 0.7438,
      "step": 6201
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.38341999053955,
      "learning_rate": 1.0045045045045045e-05,
      "loss": 0.4273,
      "step": 6202
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.905077934265137,
      "learning_rate": 1.0041827541827542e-05,
      "loss": 0.6663,
      "step": 6203
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.396012306213379,
      "learning_rate": 1.003861003861004e-05,
      "loss": 0.5385,
      "step": 6204
    },
    {
      "epoch": 1.33,
      "grad_norm": 13.908981323242188,
      "learning_rate": 1.0035392535392536e-05,
      "loss": 0.7389,
      "step": 6205
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.053441047668457,
      "learning_rate": 1.0032175032175031e-05,
      "loss": 0.4411,
      "step": 6206
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.567312240600586,
      "learning_rate": 1.002895752895753e-05,
      "loss": 0.6267,
      "step": 6207
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.125778198242188,
      "learning_rate": 1.0025740025740026e-05,
      "loss": 0.5536,
      "step": 6208
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.718879699707031,
      "learning_rate": 1.0022522522522522e-05,
      "loss": 0.5522,
      "step": 6209
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.63663387298584,
      "learning_rate": 1.0019305019305019e-05,
      "loss": 0.8472,
      "step": 6210
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.94503402709961,
      "learning_rate": 1.0016087516087517e-05,
      "loss": 0.6866,
      "step": 6211
    },
    {
      "epoch": 1.33,
      "grad_norm": 7.917267322540283,
      "learning_rate": 1.0012870012870013e-05,
      "loss": 0.3014,
      "step": 6212
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.901139259338379,
      "learning_rate": 1.000965250965251e-05,
      "loss": 0.6842,
      "step": 6213
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.09284782409668,
      "learning_rate": 1.0006435006435008e-05,
      "loss": 0.4696,
      "step": 6214
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.121978759765625,
      "learning_rate": 1.0003217503217503e-05,
      "loss": 0.5137,
      "step": 6215
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.976791381835938,
      "learning_rate": 9.999999999999999e-06,
      "loss": 0.7435,
      "step": 6216
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.419821739196777,
      "learning_rate": 9.996782496782497e-06,
      "loss": 0.8376,
      "step": 6217
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.531576156616211,
      "learning_rate": 9.993564993564994e-06,
      "loss": 0.6106,
      "step": 6218
    },
    {
      "epoch": 1.33,
      "grad_norm": 10.894550323486328,
      "learning_rate": 9.99034749034749e-06,
      "loss": 0.7382,
      "step": 6219
    },
    {
      "epoch": 1.33,
      "grad_norm": 12.510377883911133,
      "learning_rate": 9.987129987129988e-06,
      "loss": 0.9416,
      "step": 6220
    },
    {
      "epoch": 1.33,
      "grad_norm": 7.991020679473877,
      "learning_rate": 9.983912483912485e-06,
      "loss": 0.5623,
      "step": 6221
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.779341697692871,
      "learning_rate": 9.980694980694981e-06,
      "loss": 0.6822,
      "step": 6222
    },
    {
      "epoch": 1.33,
      "grad_norm": 11.058351516723633,
      "learning_rate": 9.977477477477478e-06,
      "loss": 0.5115,
      "step": 6223
    },
    {
      "epoch": 1.34,
      "grad_norm": 17.25874137878418,
      "learning_rate": 9.974259974259974e-06,
      "loss": 0.8668,
      "step": 6224
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.152116775512695,
      "learning_rate": 9.97104247104247e-06,
      "loss": 0.6525,
      "step": 6225
    },
    {
      "epoch": 1.34,
      "grad_norm": 14.336889266967773,
      "learning_rate": 9.967824967824967e-06,
      "loss": 0.7537,
      "step": 6226
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.627884864807129,
      "learning_rate": 9.964607464607465e-06,
      "loss": 0.7469,
      "step": 6227
    },
    {
      "epoch": 1.34,
      "grad_norm": 9.10770034790039,
      "learning_rate": 9.961389961389962e-06,
      "loss": 0.472,
      "step": 6228
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.329954147338867,
      "learning_rate": 9.958172458172458e-06,
      "loss": 0.7833,
      "step": 6229
    },
    {
      "epoch": 1.34,
      "grad_norm": 13.394063949584961,
      "learning_rate": 9.954954954954956e-06,
      "loss": 0.906,
      "step": 6230
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.337431907653809,
      "learning_rate": 9.951737451737453e-06,
      "loss": 0.5838,
      "step": 6231
    },
    {
      "epoch": 1.34,
      "grad_norm": 12.701481819152832,
      "learning_rate": 9.948519948519947e-06,
      "loss": 0.7926,
      "step": 6232
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.134413719177246,
      "learning_rate": 9.945302445302446e-06,
      "loss": 0.4502,
      "step": 6233
    },
    {
      "epoch": 1.34,
      "grad_norm": 9.006406784057617,
      "learning_rate": 9.942084942084942e-06,
      "loss": 0.3941,
      "step": 6234
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.525694847106934,
      "learning_rate": 9.938867438867438e-06,
      "loss": 0.5408,
      "step": 6235
    },
    {
      "epoch": 1.34,
      "grad_norm": 9.190122604370117,
      "learning_rate": 9.935649935649937e-06,
      "loss": 0.4905,
      "step": 6236
    },
    {
      "epoch": 1.34,
      "grad_norm": 12.768417358398438,
      "learning_rate": 9.932432432432433e-06,
      "loss": 0.665,
      "step": 6237
    },
    {
      "epoch": 1.34,
      "grad_norm": 9.440169334411621,
      "learning_rate": 9.92921492921493e-06,
      "loss": 0.6688,
      "step": 6238
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.635671615600586,
      "learning_rate": 9.925997425997426e-06,
      "loss": 0.7326,
      "step": 6239
    },
    {
      "epoch": 1.34,
      "grad_norm": 8.3223295211792,
      "learning_rate": 9.922779922779924e-06,
      "loss": 0.5177,
      "step": 6240
    },
    {
      "epoch": 1.34,
      "grad_norm": 12.427668571472168,
      "learning_rate": 9.919562419562419e-06,
      "loss": 0.7061,
      "step": 6241
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.454998016357422,
      "learning_rate": 9.916344916344915e-06,
      "loss": 0.7009,
      "step": 6242
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.957289695739746,
      "learning_rate": 9.913127413127413e-06,
      "loss": 0.6104,
      "step": 6243
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.54112434387207,
      "learning_rate": 9.90990990990991e-06,
      "loss": 0.7258,
      "step": 6244
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.875158309936523,
      "learning_rate": 9.906692406692406e-06,
      "loss": 0.4952,
      "step": 6245
    },
    {
      "epoch": 1.34,
      "grad_norm": 13.482582092285156,
      "learning_rate": 9.903474903474905e-06,
      "loss": 0.6199,
      "step": 6246
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.514887809753418,
      "learning_rate": 9.900257400257401e-06,
      "loss": 0.5445,
      "step": 6247
    },
    {
      "epoch": 1.34,
      "grad_norm": 12.321446418762207,
      "learning_rate": 9.897039897039897e-06,
      "loss": 0.703,
      "step": 6248
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.22139835357666,
      "learning_rate": 9.893822393822396e-06,
      "loss": 0.6959,
      "step": 6249
    },
    {
      "epoch": 1.34,
      "grad_norm": 12.604560852050781,
      "learning_rate": 9.89060489060489e-06,
      "loss": 0.6172,
      "step": 6250
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.349967002868652,
      "learning_rate": 9.887387387387387e-06,
      "loss": 0.6707,
      "step": 6251
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.037968635559082,
      "learning_rate": 9.884169884169885e-06,
      "loss": 0.6865,
      "step": 6252
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.827404975891113,
      "learning_rate": 9.880952380952381e-06,
      "loss": 0.5843,
      "step": 6253
    },
    {
      "epoch": 1.34,
      "grad_norm": 7.996497631072998,
      "learning_rate": 9.877734877734878e-06,
      "loss": 0.3442,
      "step": 6254
    },
    {
      "epoch": 1.34,
      "grad_norm": 8.253238677978516,
      "learning_rate": 9.874517374517374e-06,
      "loss": 0.5206,
      "step": 6255
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.644440650939941,
      "learning_rate": 9.871299871299872e-06,
      "loss": 0.7329,
      "step": 6256
    },
    {
      "epoch": 1.34,
      "grad_norm": 9.47573184967041,
      "learning_rate": 9.868082368082369e-06,
      "loss": 0.5552,
      "step": 6257
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.569482803344727,
      "learning_rate": 9.864864864864865e-06,
      "loss": 0.6582,
      "step": 6258
    },
    {
      "epoch": 1.34,
      "grad_norm": 9.239815711975098,
      "learning_rate": 9.861647361647362e-06,
      "loss": 0.4568,
      "step": 6259
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.536721229553223,
      "learning_rate": 9.858429858429858e-06,
      "loss": 0.5727,
      "step": 6260
    },
    {
      "epoch": 1.34,
      "grad_norm": 8.786112785339355,
      "learning_rate": 9.855212355212355e-06,
      "loss": 0.6097,
      "step": 6261
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.292032241821289,
      "learning_rate": 9.851994851994853e-06,
      "loss": 0.663,
      "step": 6262
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.117236137390137,
      "learning_rate": 9.84877734877735e-06,
      "loss": 0.7588,
      "step": 6263
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.459589004516602,
      "learning_rate": 9.845559845559846e-06,
      "loss": 0.7108,
      "step": 6264
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.522000312805176,
      "learning_rate": 9.842342342342344e-06,
      "loss": 0.4424,
      "step": 6265
    },
    {
      "epoch": 1.34,
      "grad_norm": 9.437715530395508,
      "learning_rate": 9.83912483912484e-06,
      "loss": 0.4781,
      "step": 6266
    },
    {
      "epoch": 1.34,
      "grad_norm": 10.31361198425293,
      "learning_rate": 9.835907335907335e-06,
      "loss": 0.7141,
      "step": 6267
    },
    {
      "epoch": 1.34,
      "grad_norm": 9.576571464538574,
      "learning_rate": 9.832689832689832e-06,
      "loss": 0.589,
      "step": 6268
    },
    {
      "epoch": 1.34,
      "grad_norm": 9.66054916381836,
      "learning_rate": 9.82947232947233e-06,
      "loss": 0.5552,
      "step": 6269
    },
    {
      "epoch": 1.34,
      "grad_norm": 14.411666870117188,
      "learning_rate": 9.826254826254826e-06,
      "loss": 0.4866,
      "step": 6270
    },
    {
      "epoch": 1.35,
      "grad_norm": 8.607586860656738,
      "learning_rate": 9.823037323037323e-06,
      "loss": 0.4806,
      "step": 6271
    },
    {
      "epoch": 1.35,
      "grad_norm": 13.127479553222656,
      "learning_rate": 9.81981981981982e-06,
      "loss": 0.7703,
      "step": 6272
    },
    {
      "epoch": 1.35,
      "grad_norm": 11.45085620880127,
      "learning_rate": 9.816602316602317e-06,
      "loss": 0.5675,
      "step": 6273
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.240673065185547,
      "learning_rate": 9.813384813384814e-06,
      "loss": 0.5673,
      "step": 6274
    },
    {
      "epoch": 1.35,
      "grad_norm": 8.893425941467285,
      "learning_rate": 9.810167310167312e-06,
      "loss": 0.4326,
      "step": 6275
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.040985107421875,
      "learning_rate": 9.806949806949807e-06,
      "loss": 0.676,
      "step": 6276
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.64947509765625,
      "learning_rate": 9.803732303732303e-06,
      "loss": 0.5246,
      "step": 6277
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.312335968017578,
      "learning_rate": 9.800514800514801e-06,
      "loss": 0.5508,
      "step": 6278
    },
    {
      "epoch": 1.35,
      "grad_norm": 11.202442169189453,
      "learning_rate": 9.797297297297298e-06,
      "loss": 0.558,
      "step": 6279
    },
    {
      "epoch": 1.35,
      "grad_norm": 12.091005325317383,
      "learning_rate": 9.794079794079794e-06,
      "loss": 0.8223,
      "step": 6280
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.36121654510498,
      "learning_rate": 9.790862290862292e-06,
      "loss": 0.5227,
      "step": 6281
    },
    {
      "epoch": 1.35,
      "grad_norm": 17.2668514251709,
      "learning_rate": 9.787644787644789e-06,
      "loss": 0.9207,
      "step": 6282
    },
    {
      "epoch": 1.35,
      "grad_norm": 11.289665222167969,
      "learning_rate": 9.784427284427285e-06,
      "loss": 0.7646,
      "step": 6283
    },
    {
      "epoch": 1.35,
      "grad_norm": 14.278360366821289,
      "learning_rate": 9.781209781209782e-06,
      "loss": 0.8,
      "step": 6284
    },
    {
      "epoch": 1.35,
      "grad_norm": 7.772865295410156,
      "learning_rate": 9.777992277992278e-06,
      "loss": 0.5095,
      "step": 6285
    },
    {
      "epoch": 1.35,
      "grad_norm": 12.457809448242188,
      "learning_rate": 9.774774774774774e-06,
      "loss": 0.5814,
      "step": 6286
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.080742835998535,
      "learning_rate": 9.771557271557271e-06,
      "loss": 0.6233,
      "step": 6287
    },
    {
      "epoch": 1.35,
      "grad_norm": 12.103531837463379,
      "learning_rate": 9.768339768339769e-06,
      "loss": 0.6787,
      "step": 6288
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.865301132202148,
      "learning_rate": 9.765122265122265e-06,
      "loss": 0.6664,
      "step": 6289
    },
    {
      "epoch": 1.35,
      "grad_norm": 8.920418739318848,
      "learning_rate": 9.761904761904762e-06,
      "loss": 0.4423,
      "step": 6290
    },
    {
      "epoch": 1.35,
      "grad_norm": 8.64992904663086,
      "learning_rate": 9.75868725868726e-06,
      "loss": 0.5635,
      "step": 6291
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.334039688110352,
      "learning_rate": 9.755469755469757e-06,
      "loss": 0.6989,
      "step": 6292
    },
    {
      "epoch": 1.35,
      "grad_norm": 13.123568534851074,
      "learning_rate": 9.752252252252251e-06,
      "loss": 0.6764,
      "step": 6293
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.68307876586914,
      "learning_rate": 9.74903474903475e-06,
      "loss": 0.5705,
      "step": 6294
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.867582321166992,
      "learning_rate": 9.745817245817246e-06,
      "loss": 0.5482,
      "step": 6295
    },
    {
      "epoch": 1.35,
      "grad_norm": 12.077391624450684,
      "learning_rate": 9.742599742599742e-06,
      "loss": 0.6515,
      "step": 6296
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.593498229980469,
      "learning_rate": 9.739382239382239e-06,
      "loss": 0.4431,
      "step": 6297
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.039839744567871,
      "learning_rate": 9.736164736164737e-06,
      "loss": 0.5506,
      "step": 6298
    },
    {
      "epoch": 1.35,
      "grad_norm": 8.161016464233398,
      "learning_rate": 9.732947232947233e-06,
      "loss": 0.3692,
      "step": 6299
    },
    {
      "epoch": 1.35,
      "grad_norm": 11.640802383422852,
      "learning_rate": 9.72972972972973e-06,
      "loss": 0.576,
      "step": 6300
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.561071395874023,
      "learning_rate": 9.726512226512228e-06,
      "loss": 0.6385,
      "step": 6301
    },
    {
      "epoch": 1.35,
      "grad_norm": 11.688395500183105,
      "learning_rate": 9.723294723294723e-06,
      "loss": 0.5349,
      "step": 6302
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.592639923095703,
      "learning_rate": 9.72007722007722e-06,
      "loss": 0.6573,
      "step": 6303
    },
    {
      "epoch": 1.35,
      "grad_norm": 11.858994483947754,
      "learning_rate": 9.716859716859717e-06,
      "loss": 0.7285,
      "step": 6304
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.936379432678223,
      "learning_rate": 9.713642213642214e-06,
      "loss": 0.4908,
      "step": 6305
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.589293479919434,
      "learning_rate": 9.71042471042471e-06,
      "loss": 0.5217,
      "step": 6306
    },
    {
      "epoch": 1.35,
      "grad_norm": 11.175455093383789,
      "learning_rate": 9.707207207207208e-06,
      "loss": 0.5726,
      "step": 6307
    },
    {
      "epoch": 1.35,
      "grad_norm": 7.498505115509033,
      "learning_rate": 9.703989703989705e-06,
      "loss": 0.4855,
      "step": 6308
    },
    {
      "epoch": 1.35,
      "grad_norm": 13.98232364654541,
      "learning_rate": 9.700772200772201e-06,
      "loss": 0.5993,
      "step": 6309
    },
    {
      "epoch": 1.35,
      "grad_norm": 13.459490776062012,
      "learning_rate": 9.6975546975547e-06,
      "loss": 0.5867,
      "step": 6310
    },
    {
      "epoch": 1.35,
      "grad_norm": 8.586645126342773,
      "learning_rate": 9.694337194337194e-06,
      "loss": 0.5986,
      "step": 6311
    },
    {
      "epoch": 1.35,
      "grad_norm": 12.639885902404785,
      "learning_rate": 9.69111969111969e-06,
      "loss": 0.6598,
      "step": 6312
    },
    {
      "epoch": 1.35,
      "grad_norm": 12.58632755279541,
      "learning_rate": 9.687902187902187e-06,
      "loss": 0.7319,
      "step": 6313
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.46733283996582,
      "learning_rate": 9.684684684684685e-06,
      "loss": 0.4918,
      "step": 6314
    },
    {
      "epoch": 1.35,
      "grad_norm": 10.108266830444336,
      "learning_rate": 9.681467181467182e-06,
      "loss": 0.7203,
      "step": 6315
    },
    {
      "epoch": 1.35,
      "grad_norm": 15.481941223144531,
      "learning_rate": 9.678249678249678e-06,
      "loss": 0.6869,
      "step": 6316
    },
    {
      "epoch": 1.35,
      "grad_norm": 8.503536224365234,
      "learning_rate": 9.675032175032176e-06,
      "loss": 0.5489,
      "step": 6317
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.138465881347656,
      "learning_rate": 9.671814671814673e-06,
      "loss": 0.7303,
      "step": 6318
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.510187149047852,
      "learning_rate": 9.668597168597167e-06,
      "loss": 0.5927,
      "step": 6319
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.089920043945312,
      "learning_rate": 9.665379665379666e-06,
      "loss": 0.5607,
      "step": 6320
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.684447288513184,
      "learning_rate": 9.662162162162162e-06,
      "loss": 0.6348,
      "step": 6321
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.82172966003418,
      "learning_rate": 9.658944658944659e-06,
      "loss": 0.5057,
      "step": 6322
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.428365707397461,
      "learning_rate": 9.655727155727157e-06,
      "loss": 0.5721,
      "step": 6323
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.209098815917969,
      "learning_rate": 9.652509652509653e-06,
      "loss": 0.7344,
      "step": 6324
    },
    {
      "epoch": 1.36,
      "grad_norm": 14.635553359985352,
      "learning_rate": 9.64929214929215e-06,
      "loss": 0.7746,
      "step": 6325
    },
    {
      "epoch": 1.36,
      "grad_norm": 8.544328689575195,
      "learning_rate": 9.646074646074646e-06,
      "loss": 0.431,
      "step": 6326
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.365184783935547,
      "learning_rate": 9.642857142857144e-06,
      "loss": 0.9691,
      "step": 6327
    },
    {
      "epoch": 1.36,
      "grad_norm": 6.531259536743164,
      "learning_rate": 9.639639639639639e-06,
      "loss": 0.284,
      "step": 6328
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.908346176147461,
      "learning_rate": 9.636422136422135e-06,
      "loss": 0.623,
      "step": 6329
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.077668190002441,
      "learning_rate": 9.633204633204634e-06,
      "loss": 0.5634,
      "step": 6330
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.138564109802246,
      "learning_rate": 9.62998712998713e-06,
      "loss": 0.6712,
      "step": 6331
    },
    {
      "epoch": 1.36,
      "grad_norm": 12.968515396118164,
      "learning_rate": 9.626769626769626e-06,
      "loss": 0.5633,
      "step": 6332
    },
    {
      "epoch": 1.36,
      "grad_norm": 15.11110782623291,
      "learning_rate": 9.623552123552125e-06,
      "loss": 0.4486,
      "step": 6333
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.274319648742676,
      "learning_rate": 9.620334620334621e-06,
      "loss": 0.762,
      "step": 6334
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.995645523071289,
      "learning_rate": 9.617117117117117e-06,
      "loss": 0.7017,
      "step": 6335
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.860328674316406,
      "learning_rate": 9.613899613899616e-06,
      "loss": 0.5359,
      "step": 6336
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.070393562316895,
      "learning_rate": 9.61068211068211e-06,
      "loss": 0.4721,
      "step": 6337
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.573576927185059,
      "learning_rate": 9.607464607464607e-06,
      "loss": 0.5705,
      "step": 6338
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.052062034606934,
      "learning_rate": 9.604247104247105e-06,
      "loss": 0.596,
      "step": 6339
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.401593208312988,
      "learning_rate": 9.601029601029601e-06,
      "loss": 0.6562,
      "step": 6340
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.785155296325684,
      "learning_rate": 9.597812097812098e-06,
      "loss": 0.6882,
      "step": 6341
    },
    {
      "epoch": 1.36,
      "grad_norm": 8.47840404510498,
      "learning_rate": 9.594594594594594e-06,
      "loss": 0.4673,
      "step": 6342
    },
    {
      "epoch": 1.36,
      "grad_norm": 8.507781028747559,
      "learning_rate": 9.591377091377092e-06,
      "loss": 0.5388,
      "step": 6343
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.707995414733887,
      "learning_rate": 9.588159588159589e-06,
      "loss": 0.48,
      "step": 6344
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.027190208435059,
      "learning_rate": 9.584942084942085e-06,
      "loss": 0.4796,
      "step": 6345
    },
    {
      "epoch": 1.36,
      "grad_norm": 14.852235794067383,
      "learning_rate": 9.581724581724582e-06,
      "loss": 0.9682,
      "step": 6346
    },
    {
      "epoch": 1.36,
      "grad_norm": 12.676399230957031,
      "learning_rate": 9.578507078507078e-06,
      "loss": 0.6136,
      "step": 6347
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.092205047607422,
      "learning_rate": 9.575289575289575e-06,
      "loss": 0.5722,
      "step": 6348
    },
    {
      "epoch": 1.36,
      "grad_norm": 12.690547943115234,
      "learning_rate": 9.572072072072073e-06,
      "loss": 0.8512,
      "step": 6349
    },
    {
      "epoch": 1.36,
      "grad_norm": 12.12379264831543,
      "learning_rate": 9.56885456885457e-06,
      "loss": 0.4887,
      "step": 6350
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.64571475982666,
      "learning_rate": 9.565637065637066e-06,
      "loss": 0.5295,
      "step": 6351
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.03640365600586,
      "learning_rate": 9.562419562419564e-06,
      "loss": 0.5616,
      "step": 6352
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.951726913452148,
      "learning_rate": 9.55920205920206e-06,
      "loss": 0.7495,
      "step": 6353
    },
    {
      "epoch": 1.36,
      "grad_norm": 12.93911075592041,
      "learning_rate": 9.555984555984555e-06,
      "loss": 0.7871,
      "step": 6354
    },
    {
      "epoch": 1.36,
      "grad_norm": 8.416556358337402,
      "learning_rate": 9.552767052767052e-06,
      "loss": 0.5753,
      "step": 6355
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.441847801208496,
      "learning_rate": 9.54954954954955e-06,
      "loss": 0.522,
      "step": 6356
    },
    {
      "epoch": 1.36,
      "grad_norm": 12.516658782958984,
      "learning_rate": 9.546332046332046e-06,
      "loss": 0.5341,
      "step": 6357
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.212501525878906,
      "learning_rate": 9.543114543114543e-06,
      "loss": 0.4997,
      "step": 6358
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.508016586303711,
      "learning_rate": 9.53989703989704e-06,
      "loss": 0.5772,
      "step": 6359
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.262557983398438,
      "learning_rate": 9.536679536679537e-06,
      "loss": 0.6994,
      "step": 6360
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.296520233154297,
      "learning_rate": 9.533462033462034e-06,
      "loss": 0.8022,
      "step": 6361
    },
    {
      "epoch": 1.36,
      "grad_norm": 13.190725326538086,
      "learning_rate": 9.530244530244532e-06,
      "loss": 0.6155,
      "step": 6362
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.927077293395996,
      "learning_rate": 9.527027027027027e-06,
      "loss": 0.709,
      "step": 6363
    },
    {
      "epoch": 1.37,
      "grad_norm": 8.531389236450195,
      "learning_rate": 9.523809523809523e-06,
      "loss": 0.4431,
      "step": 6364
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.284053802490234,
      "learning_rate": 9.520592020592021e-06,
      "loss": 0.5113,
      "step": 6365
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.87370777130127,
      "learning_rate": 9.517374517374518e-06,
      "loss": 0.6123,
      "step": 6366
    },
    {
      "epoch": 1.37,
      "grad_norm": 11.86104965209961,
      "learning_rate": 9.514157014157014e-06,
      "loss": 0.6037,
      "step": 6367
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.094735145568848,
      "learning_rate": 9.510939510939512e-06,
      "loss": 0.4225,
      "step": 6368
    },
    {
      "epoch": 1.37,
      "grad_norm": 13.12128734588623,
      "learning_rate": 9.507722007722009e-06,
      "loss": 0.5584,
      "step": 6369
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.021773338317871,
      "learning_rate": 9.504504504504505e-06,
      "loss": 0.5172,
      "step": 6370
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.897940635681152,
      "learning_rate": 9.501287001287002e-06,
      "loss": 0.6051,
      "step": 6371
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.31609058380127,
      "learning_rate": 9.498069498069498e-06,
      "loss": 0.5259,
      "step": 6372
    },
    {
      "epoch": 1.37,
      "grad_norm": 12.768012046813965,
      "learning_rate": 9.494851994851995e-06,
      "loss": 0.5508,
      "step": 6373
    },
    {
      "epoch": 1.37,
      "grad_norm": 12.07618522644043,
      "learning_rate": 9.491634491634491e-06,
      "loss": 0.6461,
      "step": 6374
    },
    {
      "epoch": 1.37,
      "grad_norm": 12.123510360717773,
      "learning_rate": 9.488416988416989e-06,
      "loss": 0.7066,
      "step": 6375
    },
    {
      "epoch": 1.37,
      "grad_norm": 11.574881553649902,
      "learning_rate": 9.485199485199486e-06,
      "loss": 0.5775,
      "step": 6376
    },
    {
      "epoch": 1.37,
      "grad_norm": 12.006049156188965,
      "learning_rate": 9.481981981981982e-06,
      "loss": 0.6847,
      "step": 6377
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.658989906311035,
      "learning_rate": 9.47876447876448e-06,
      "loss": 0.4751,
      "step": 6378
    },
    {
      "epoch": 1.37,
      "grad_norm": 8.835227966308594,
      "learning_rate": 9.475546975546977e-06,
      "loss": 0.4986,
      "step": 6379
    },
    {
      "epoch": 1.37,
      "grad_norm": 11.779412269592285,
      "learning_rate": 9.472329472329471e-06,
      "loss": 0.6348,
      "step": 6380
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.415992736816406,
      "learning_rate": 9.46911196911197e-06,
      "loss": 0.4955,
      "step": 6381
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.960650444030762,
      "learning_rate": 9.465894465894466e-06,
      "loss": 0.5255,
      "step": 6382
    },
    {
      "epoch": 1.37,
      "grad_norm": 11.514086723327637,
      "learning_rate": 9.462676962676962e-06,
      "loss": 0.5585,
      "step": 6383
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.554118156433105,
      "learning_rate": 9.459459459459459e-06,
      "loss": 0.4912,
      "step": 6384
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.904060363769531,
      "learning_rate": 9.456241956241957e-06,
      "loss": 0.5804,
      "step": 6385
    },
    {
      "epoch": 1.37,
      "grad_norm": 11.75741195678711,
      "learning_rate": 9.453024453024453e-06,
      "loss": 0.6434,
      "step": 6386
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.314659118652344,
      "learning_rate": 9.44980694980695e-06,
      "loss": 0.595,
      "step": 6387
    },
    {
      "epoch": 1.37,
      "grad_norm": 12.397468566894531,
      "learning_rate": 9.446589446589448e-06,
      "loss": 0.5321,
      "step": 6388
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.314020156860352,
      "learning_rate": 9.443371943371943e-06,
      "loss": 0.5537,
      "step": 6389
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.953603744506836,
      "learning_rate": 9.44015444015444e-06,
      "loss": 0.6834,
      "step": 6390
    },
    {
      "epoch": 1.37,
      "grad_norm": 14.874187469482422,
      "learning_rate": 9.436936936936937e-06,
      "loss": 0.6873,
      "step": 6391
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.681465148925781,
      "learning_rate": 9.433719433719434e-06,
      "loss": 0.4241,
      "step": 6392
    },
    {
      "epoch": 1.37,
      "grad_norm": 11.829019546508789,
      "learning_rate": 9.43050193050193e-06,
      "loss": 0.6374,
      "step": 6393
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.27204418182373,
      "learning_rate": 9.427284427284428e-06,
      "loss": 0.4681,
      "step": 6394
    },
    {
      "epoch": 1.37,
      "grad_norm": 11.477272987365723,
      "learning_rate": 9.424066924066925e-06,
      "loss": 0.5624,
      "step": 6395
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.59615707397461,
      "learning_rate": 9.420849420849421e-06,
      "loss": 0.5815,
      "step": 6396
    },
    {
      "epoch": 1.37,
      "grad_norm": 12.131771087646484,
      "learning_rate": 9.41763191763192e-06,
      "loss": 0.7186,
      "step": 6397
    },
    {
      "epoch": 1.37,
      "grad_norm": 8.017773628234863,
      "learning_rate": 9.414414414414414e-06,
      "loss": 0.599,
      "step": 6398
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.824178695678711,
      "learning_rate": 9.41119691119691e-06,
      "loss": 0.6391,
      "step": 6399
    },
    {
      "epoch": 1.37,
      "grad_norm": 8.475503921508789,
      "learning_rate": 9.407979407979407e-06,
      "loss": 0.5332,
      "step": 6400
    },
    {
      "epoch": 1.37,
      "grad_norm": 7.279738426208496,
      "learning_rate": 9.404761904761905e-06,
      "loss": 0.4503,
      "step": 6401
    },
    {
      "epoch": 1.37,
      "grad_norm": 11.955554962158203,
      "learning_rate": 9.401544401544402e-06,
      "loss": 0.7593,
      "step": 6402
    },
    {
      "epoch": 1.37,
      "grad_norm": 11.502176284790039,
      "learning_rate": 9.398326898326898e-06,
      "loss": 0.6409,
      "step": 6403
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.912076950073242,
      "learning_rate": 9.395109395109396e-06,
      "loss": 0.6127,
      "step": 6404
    },
    {
      "epoch": 1.37,
      "grad_norm": 11.908203125,
      "learning_rate": 9.391891891891893e-06,
      "loss": 1.0341,
      "step": 6405
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.691505432128906,
      "learning_rate": 9.388674388674388e-06,
      "loss": 0.539,
      "step": 6406
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.635562896728516,
      "learning_rate": 9.385456885456886e-06,
      "loss": 0.6577,
      "step": 6407
    },
    {
      "epoch": 1.37,
      "grad_norm": 8.825695991516113,
      "learning_rate": 9.382239382239382e-06,
      "loss": 0.494,
      "step": 6408
    },
    {
      "epoch": 1.37,
      "grad_norm": 8.492881774902344,
      "learning_rate": 9.379021879021879e-06,
      "loss": 0.5878,
      "step": 6409
    },
    {
      "epoch": 1.37,
      "grad_norm": 10.75671100616455,
      "learning_rate": 9.375804375804377e-06,
      "loss": 0.4966,
      "step": 6410
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.489709854125977,
      "learning_rate": 9.372586872586873e-06,
      "loss": 0.5006,
      "step": 6411
    },
    {
      "epoch": 1.38,
      "grad_norm": 12.027740478515625,
      "learning_rate": 9.36936936936937e-06,
      "loss": 0.6788,
      "step": 6412
    },
    {
      "epoch": 1.38,
      "grad_norm": 9.894270896911621,
      "learning_rate": 9.366151866151866e-06,
      "loss": 0.4244,
      "step": 6413
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.282017707824707,
      "learning_rate": 9.362934362934364e-06,
      "loss": 0.5558,
      "step": 6414
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.138701438903809,
      "learning_rate": 9.359716859716859e-06,
      "loss": 0.623,
      "step": 6415
    },
    {
      "epoch": 1.38,
      "grad_norm": 13.887653350830078,
      "learning_rate": 9.356499356499355e-06,
      "loss": 0.5428,
      "step": 6416
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.635730743408203,
      "learning_rate": 9.353281853281854e-06,
      "loss": 0.8778,
      "step": 6417
    },
    {
      "epoch": 1.38,
      "grad_norm": 8.127756118774414,
      "learning_rate": 9.35006435006435e-06,
      "loss": 0.3256,
      "step": 6418
    },
    {
      "epoch": 1.38,
      "grad_norm": 12.250808715820312,
      "learning_rate": 9.346846846846847e-06,
      "loss": 0.6547,
      "step": 6419
    },
    {
      "epoch": 1.38,
      "grad_norm": 13.711136817932129,
      "learning_rate": 9.343629343629345e-06,
      "loss": 0.6641,
      "step": 6420
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.196484565734863,
      "learning_rate": 9.340411840411841e-06,
      "loss": 0.5487,
      "step": 6421
    },
    {
      "epoch": 1.38,
      "grad_norm": 8.86767292022705,
      "learning_rate": 9.337194337194338e-06,
      "loss": 0.4985,
      "step": 6422
    },
    {
      "epoch": 1.38,
      "grad_norm": 12.479515075683594,
      "learning_rate": 9.333976833976836e-06,
      "loss": 0.6343,
      "step": 6423
    },
    {
      "epoch": 1.38,
      "grad_norm": 13.572689056396484,
      "learning_rate": 9.33075933075933e-06,
      "loss": 0.6289,
      "step": 6424
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.617677688598633,
      "learning_rate": 9.327541827541827e-06,
      "loss": 0.6029,
      "step": 6425
    },
    {
      "epoch": 1.38,
      "grad_norm": 12.116973876953125,
      "learning_rate": 9.324324324324325e-06,
      "loss": 0.8034,
      "step": 6426
    },
    {
      "epoch": 1.38,
      "grad_norm": 7.016791820526123,
      "learning_rate": 9.321106821106822e-06,
      "loss": 0.3424,
      "step": 6427
    },
    {
      "epoch": 1.38,
      "grad_norm": 13.211710929870605,
      "learning_rate": 9.317889317889318e-06,
      "loss": 0.4834,
      "step": 6428
    },
    {
      "epoch": 1.38,
      "grad_norm": 12.742944717407227,
      "learning_rate": 9.314671814671814e-06,
      "loss": 0.6231,
      "step": 6429
    },
    {
      "epoch": 1.38,
      "grad_norm": 8.081807136535645,
      "learning_rate": 9.311454311454313e-06,
      "loss": 0.3998,
      "step": 6430
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.154936790466309,
      "learning_rate": 9.308236808236809e-06,
      "loss": 0.6111,
      "step": 6431
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.559601783752441,
      "learning_rate": 9.305019305019304e-06,
      "loss": 0.6922,
      "step": 6432
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.512642860412598,
      "learning_rate": 9.301801801801802e-06,
      "loss": 0.7771,
      "step": 6433
    },
    {
      "epoch": 1.38,
      "grad_norm": 9.074368476867676,
      "learning_rate": 9.298584298584298e-06,
      "loss": 0.6829,
      "step": 6434
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.569233894348145,
      "learning_rate": 9.295366795366795e-06,
      "loss": 0.6159,
      "step": 6435
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.081929206848145,
      "learning_rate": 9.292149292149293e-06,
      "loss": 0.4939,
      "step": 6436
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.277430534362793,
      "learning_rate": 9.28893178893179e-06,
      "loss": 0.719,
      "step": 6437
    },
    {
      "epoch": 1.38,
      "grad_norm": 12.352429389953613,
      "learning_rate": 9.285714285714286e-06,
      "loss": 0.6156,
      "step": 6438
    },
    {
      "epoch": 1.38,
      "grad_norm": 12.952256202697754,
      "learning_rate": 9.282496782496784e-06,
      "loss": 0.7185,
      "step": 6439
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.767824172973633,
      "learning_rate": 9.27927927927928e-06,
      "loss": 0.6026,
      "step": 6440
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.187636375427246,
      "learning_rate": 9.276061776061775e-06,
      "loss": 0.6817,
      "step": 6441
    },
    {
      "epoch": 1.38,
      "grad_norm": 9.47856616973877,
      "learning_rate": 9.272844272844272e-06,
      "loss": 0.6936,
      "step": 6442
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.43702507019043,
      "learning_rate": 9.26962676962677e-06,
      "loss": 0.7787,
      "step": 6443
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.534799575805664,
      "learning_rate": 9.266409266409266e-06,
      "loss": 0.4948,
      "step": 6444
    },
    {
      "epoch": 1.38,
      "grad_norm": 14.976583480834961,
      "learning_rate": 9.263191763191763e-06,
      "loss": 0.533,
      "step": 6445
    },
    {
      "epoch": 1.38,
      "grad_norm": 8.871424674987793,
      "learning_rate": 9.259974259974261e-06,
      "loss": 0.4788,
      "step": 6446
    },
    {
      "epoch": 1.38,
      "grad_norm": 13.490510940551758,
      "learning_rate": 9.256756756756757e-06,
      "loss": 0.5992,
      "step": 6447
    },
    {
      "epoch": 1.38,
      "grad_norm": 7.482876300811768,
      "learning_rate": 9.253539253539254e-06,
      "loss": 0.422,
      "step": 6448
    },
    {
      "epoch": 1.38,
      "grad_norm": 9.15281867980957,
      "learning_rate": 9.250321750321752e-06,
      "loss": 0.5002,
      "step": 6449
    },
    {
      "epoch": 1.38,
      "grad_norm": 12.217500686645508,
      "learning_rate": 9.247104247104247e-06,
      "loss": 0.4564,
      "step": 6450
    },
    {
      "epoch": 1.38,
      "grad_norm": 10.279004096984863,
      "learning_rate": 9.243886743886743e-06,
      "loss": 0.6032,
      "step": 6451
    },
    {
      "epoch": 1.38,
      "grad_norm": 11.745721817016602,
      "learning_rate": 9.240669240669241e-06,
      "loss": 0.6308,
      "step": 6452
    },
    {
      "epoch": 1.38,
      "grad_norm": 15.207335472106934,
      "learning_rate": 9.237451737451738e-06,
      "loss": 0.79,
      "step": 6453
    },
    {
      "epoch": 1.38,
      "grad_norm": 9.39341926574707,
      "learning_rate": 9.234234234234234e-06,
      "loss": 0.5595,
      "step": 6454
    },
    {
      "epoch": 1.38,
      "grad_norm": 9.337964057922363,
      "learning_rate": 9.231016731016732e-06,
      "loss": 0.5721,
      "step": 6455
    },
    {
      "epoch": 1.38,
      "grad_norm": 9.46469497680664,
      "learning_rate": 9.227799227799229e-06,
      "loss": 0.6732,
      "step": 6456
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.29026985168457,
      "learning_rate": 9.224581724581725e-06,
      "loss": 0.68,
      "step": 6457
    },
    {
      "epoch": 1.39,
      "grad_norm": 8.86500358581543,
      "learning_rate": 9.221364221364222e-06,
      "loss": 0.4422,
      "step": 6458
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.259688377380371,
      "learning_rate": 9.218146718146718e-06,
      "loss": 0.5263,
      "step": 6459
    },
    {
      "epoch": 1.39,
      "grad_norm": 12.074053764343262,
      "learning_rate": 9.214929214929215e-06,
      "loss": 0.7107,
      "step": 6460
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.409311294555664,
      "learning_rate": 9.211711711711711e-06,
      "loss": 0.6238,
      "step": 6461
    },
    {
      "epoch": 1.39,
      "grad_norm": 11.175851821899414,
      "learning_rate": 9.20849420849421e-06,
      "loss": 0.598,
      "step": 6462
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.461065292358398,
      "learning_rate": 9.205276705276706e-06,
      "loss": 0.689,
      "step": 6463
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.883188247680664,
      "learning_rate": 9.202059202059202e-06,
      "loss": 0.5125,
      "step": 6464
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.514507293701172,
      "learning_rate": 9.1988416988417e-06,
      "loss": 0.6884,
      "step": 6465
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.956501007080078,
      "learning_rate": 9.195624195624197e-06,
      "loss": 0.7342,
      "step": 6466
    },
    {
      "epoch": 1.39,
      "grad_norm": 11.09099292755127,
      "learning_rate": 9.192406692406691e-06,
      "loss": 0.4357,
      "step": 6467
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.46692180633545,
      "learning_rate": 9.18918918918919e-06,
      "loss": 0.5254,
      "step": 6468
    },
    {
      "epoch": 1.39,
      "grad_norm": 14.202156066894531,
      "learning_rate": 9.185971685971686e-06,
      "loss": 0.5385,
      "step": 6469
    },
    {
      "epoch": 1.39,
      "grad_norm": 11.150973320007324,
      "learning_rate": 9.182754182754182e-06,
      "loss": 0.743,
      "step": 6470
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.587093353271484,
      "learning_rate": 9.179536679536679e-06,
      "loss": 0.6256,
      "step": 6471
    },
    {
      "epoch": 1.39,
      "grad_norm": 8.538211822509766,
      "learning_rate": 9.176319176319177e-06,
      "loss": 0.5704,
      "step": 6472
    },
    {
      "epoch": 1.39,
      "grad_norm": 12.228775978088379,
      "learning_rate": 9.173101673101674e-06,
      "loss": 0.8958,
      "step": 6473
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.722318649291992,
      "learning_rate": 9.16988416988417e-06,
      "loss": 0.5996,
      "step": 6474
    },
    {
      "epoch": 1.39,
      "grad_norm": 8.768101692199707,
      "learning_rate": 9.166666666666668e-06,
      "loss": 0.4652,
      "step": 6475
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.686479568481445,
      "learning_rate": 9.163449163449163e-06,
      "loss": 0.5362,
      "step": 6476
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.9950532913208,
      "learning_rate": 9.16023166023166e-06,
      "loss": 0.5641,
      "step": 6477
    },
    {
      "epoch": 1.39,
      "grad_norm": 12.271047592163086,
      "learning_rate": 9.157014157014157e-06,
      "loss": 0.6383,
      "step": 6478
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.369548797607422,
      "learning_rate": 9.153796653796654e-06,
      "loss": 0.4665,
      "step": 6479
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.47260570526123,
      "learning_rate": 9.15057915057915e-06,
      "loss": 0.5718,
      "step": 6480
    },
    {
      "epoch": 1.39,
      "grad_norm": 14.381143569946289,
      "learning_rate": 9.147361647361649e-06,
      "loss": 0.7215,
      "step": 6481
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.007087707519531,
      "learning_rate": 9.144144144144145e-06,
      "loss": 0.5563,
      "step": 6482
    },
    {
      "epoch": 1.39,
      "grad_norm": 12.416828155517578,
      "learning_rate": 9.140926640926641e-06,
      "loss": 0.5439,
      "step": 6483
    },
    {
      "epoch": 1.39,
      "grad_norm": 11.393545150756836,
      "learning_rate": 9.13770913770914e-06,
      "loss": 0.6205,
      "step": 6484
    },
    {
      "epoch": 1.39,
      "grad_norm": 11.552721977233887,
      "learning_rate": 9.134491634491634e-06,
      "loss": 0.6506,
      "step": 6485
    },
    {
      "epoch": 1.39,
      "grad_norm": 13.734131813049316,
      "learning_rate": 9.13127413127413e-06,
      "loss": 0.6458,
      "step": 6486
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.531832695007324,
      "learning_rate": 9.128056628056627e-06,
      "loss": 0.584,
      "step": 6487
    },
    {
      "epoch": 1.39,
      "grad_norm": 11.85261344909668,
      "learning_rate": 9.124839124839125e-06,
      "loss": 0.4595,
      "step": 6488
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.807887077331543,
      "learning_rate": 9.121621621621622e-06,
      "loss": 0.3788,
      "step": 6489
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.286813735961914,
      "learning_rate": 9.118404118404118e-06,
      "loss": 0.5677,
      "step": 6490
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.624704360961914,
      "learning_rate": 9.115186615186616e-06,
      "loss": 0.4853,
      "step": 6491
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.765459060668945,
      "learning_rate": 9.111969111969113e-06,
      "loss": 0.4948,
      "step": 6492
    },
    {
      "epoch": 1.39,
      "grad_norm": 11.912609100341797,
      "learning_rate": 9.108751608751608e-06,
      "loss": 0.8472,
      "step": 6493
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.754467010498047,
      "learning_rate": 9.105534105534106e-06,
      "loss": 0.3731,
      "step": 6494
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.52951431274414,
      "learning_rate": 9.102316602316602e-06,
      "loss": 0.6048,
      "step": 6495
    },
    {
      "epoch": 1.39,
      "grad_norm": 11.391777038574219,
      "learning_rate": 9.099099099099099e-06,
      "loss": 0.4824,
      "step": 6496
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.053079605102539,
      "learning_rate": 9.095881595881597e-06,
      "loss": 0.4737,
      "step": 6497
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.017467498779297,
      "learning_rate": 9.092664092664093e-06,
      "loss": 0.6301,
      "step": 6498
    },
    {
      "epoch": 1.39,
      "grad_norm": 11.4700288772583,
      "learning_rate": 9.08944658944659e-06,
      "loss": 0.8097,
      "step": 6499
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.48448371887207,
      "learning_rate": 9.086229086229086e-06,
      "loss": 0.4915,
      "step": 6500
    },
    {
      "epoch": 1.39,
      "grad_norm": 10.550519943237305,
      "learning_rate": 9.083011583011584e-06,
      "loss": 0.702,
      "step": 6501
    },
    {
      "epoch": 1.39,
      "grad_norm": 8.249382019042969,
      "learning_rate": 9.079794079794079e-06,
      "loss": 0.5789,
      "step": 6502
    },
    {
      "epoch": 1.39,
      "grad_norm": 8.857006072998047,
      "learning_rate": 9.076576576576576e-06,
      "loss": 0.5369,
      "step": 6503
    },
    {
      "epoch": 1.4,
      "grad_norm": 7.767335414886475,
      "learning_rate": 9.073359073359074e-06,
      "loss": 0.5394,
      "step": 6504
    },
    {
      "epoch": 1.4,
      "grad_norm": 12.514131546020508,
      "learning_rate": 9.07014157014157e-06,
      "loss": 0.8759,
      "step": 6505
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.004819869995117,
      "learning_rate": 9.066924066924067e-06,
      "loss": 0.6165,
      "step": 6506
    },
    {
      "epoch": 1.4,
      "grad_norm": 13.962427139282227,
      "learning_rate": 9.063706563706565e-06,
      "loss": 0.6846,
      "step": 6507
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.682488441467285,
      "learning_rate": 9.060489060489061e-06,
      "loss": 0.5509,
      "step": 6508
    },
    {
      "epoch": 1.4,
      "grad_norm": 12.768360137939453,
      "learning_rate": 9.057271557271558e-06,
      "loss": 0.8333,
      "step": 6509
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.786404609680176,
      "learning_rate": 9.054054054054056e-06,
      "loss": 0.5944,
      "step": 6510
    },
    {
      "epoch": 1.4,
      "grad_norm": 8.671216011047363,
      "learning_rate": 9.05083655083655e-06,
      "loss": 0.6271,
      "step": 6511
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.537859916687012,
      "learning_rate": 9.047619047619047e-06,
      "loss": 0.7471,
      "step": 6512
    },
    {
      "epoch": 1.4,
      "grad_norm": 12.282724380493164,
      "learning_rate": 9.044401544401545e-06,
      "loss": 0.8881,
      "step": 6513
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.823728561401367,
      "learning_rate": 9.041184041184042e-06,
      "loss": 0.5538,
      "step": 6514
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.98121452331543,
      "learning_rate": 9.037966537966538e-06,
      "loss": 0.8023,
      "step": 6515
    },
    {
      "epoch": 1.4,
      "grad_norm": 8.989139556884766,
      "learning_rate": 9.034749034749034e-06,
      "loss": 0.5021,
      "step": 6516
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.618865966796875,
      "learning_rate": 9.031531531531533e-06,
      "loss": 0.6006,
      "step": 6517
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.634018898010254,
      "learning_rate": 9.028314028314029e-06,
      "loss": 0.6468,
      "step": 6518
    },
    {
      "epoch": 1.4,
      "grad_norm": 7.323061943054199,
      "learning_rate": 9.025096525096524e-06,
      "loss": 0.5692,
      "step": 6519
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.409192085266113,
      "learning_rate": 9.021879021879022e-06,
      "loss": 0.5318,
      "step": 6520
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.307624816894531,
      "learning_rate": 9.018661518661518e-06,
      "loss": 0.6345,
      "step": 6521
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.65458869934082,
      "learning_rate": 9.015444015444015e-06,
      "loss": 0.7139,
      "step": 6522
    },
    {
      "epoch": 1.4,
      "grad_norm": 8.841409683227539,
      "learning_rate": 9.012226512226513e-06,
      "loss": 0.5203,
      "step": 6523
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.729172706604004,
      "learning_rate": 9.00900900900901e-06,
      "loss": 0.5727,
      "step": 6524
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.362311363220215,
      "learning_rate": 9.005791505791506e-06,
      "loss": 0.4076,
      "step": 6525
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.920736312866211,
      "learning_rate": 9.002574002574004e-06,
      "loss": 0.563,
      "step": 6526
    },
    {
      "epoch": 1.4,
      "grad_norm": 12.78987979888916,
      "learning_rate": 8.9993564993565e-06,
      "loss": 0.7938,
      "step": 6527
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.188661575317383,
      "learning_rate": 8.996138996138995e-06,
      "loss": 0.6344,
      "step": 6528
    },
    {
      "epoch": 1.4,
      "grad_norm": 8.3733549118042,
      "learning_rate": 8.992921492921492e-06,
      "loss": 0.4761,
      "step": 6529
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.893993377685547,
      "learning_rate": 8.98970398970399e-06,
      "loss": 0.4497,
      "step": 6530
    },
    {
      "epoch": 1.4,
      "grad_norm": 8.055575370788574,
      "learning_rate": 8.986486486486486e-06,
      "loss": 0.3987,
      "step": 6531
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.97335433959961,
      "learning_rate": 8.983268983268983e-06,
      "loss": 0.5319,
      "step": 6532
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.684967041015625,
      "learning_rate": 8.980051480051481e-06,
      "loss": 0.6534,
      "step": 6533
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.602205276489258,
      "learning_rate": 8.976833976833977e-06,
      "loss": 0.4648,
      "step": 6534
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.467331886291504,
      "learning_rate": 8.973616473616474e-06,
      "loss": 0.6515,
      "step": 6535
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.047870635986328,
      "learning_rate": 8.970398970398972e-06,
      "loss": 0.4961,
      "step": 6536
    },
    {
      "epoch": 1.4,
      "grad_norm": 15.32080364227295,
      "learning_rate": 8.967181467181467e-06,
      "loss": 0.5656,
      "step": 6537
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.356947898864746,
      "learning_rate": 8.963963963963963e-06,
      "loss": 0.7422,
      "step": 6538
    },
    {
      "epoch": 1.4,
      "grad_norm": 6.509359836578369,
      "learning_rate": 8.960746460746461e-06,
      "loss": 0.3201,
      "step": 6539
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.349611282348633,
      "learning_rate": 8.957528957528958e-06,
      "loss": 0.3886,
      "step": 6540
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.653144836425781,
      "learning_rate": 8.954311454311454e-06,
      "loss": 0.5598,
      "step": 6541
    },
    {
      "epoch": 1.4,
      "grad_norm": 12.033361434936523,
      "learning_rate": 8.951093951093952e-06,
      "loss": 0.7139,
      "step": 6542
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.02225112915039,
      "learning_rate": 8.947876447876449e-06,
      "loss": 0.5958,
      "step": 6543
    },
    {
      "epoch": 1.4,
      "grad_norm": 12.187891006469727,
      "learning_rate": 8.944658944658945e-06,
      "loss": 0.4888,
      "step": 6544
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.46639347076416,
      "learning_rate": 8.941441441441442e-06,
      "loss": 0.6367,
      "step": 6545
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.324055671691895,
      "learning_rate": 8.938223938223938e-06,
      "loss": 0.6134,
      "step": 6546
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.484247207641602,
      "learning_rate": 8.935006435006435e-06,
      "loss": 0.5343,
      "step": 6547
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.688741683959961,
      "learning_rate": 8.931788931788931e-06,
      "loss": 0.784,
      "step": 6548
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.802038192749023,
      "learning_rate": 8.92857142857143e-06,
      "loss": 0.5649,
      "step": 6549
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.229686737060547,
      "learning_rate": 8.925353925353926e-06,
      "loss": 0.563,
      "step": 6550
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.221271514892578,
      "learning_rate": 8.922136422136422e-06,
      "loss": 0.5268,
      "step": 6551
    },
    {
      "epoch": 1.41,
      "grad_norm": 14.058231353759766,
      "learning_rate": 8.91891891891892e-06,
      "loss": 0.9082,
      "step": 6552
    },
    {
      "epoch": 1.41,
      "grad_norm": 9.538208961486816,
      "learning_rate": 8.915701415701417e-06,
      "loss": 0.6293,
      "step": 6553
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.304450035095215,
      "learning_rate": 8.912483912483911e-06,
      "loss": 0.6976,
      "step": 6554
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.63551139831543,
      "learning_rate": 8.90926640926641e-06,
      "loss": 0.5405,
      "step": 6555
    },
    {
      "epoch": 1.41,
      "grad_norm": 8.83787727355957,
      "learning_rate": 8.906048906048906e-06,
      "loss": 0.4576,
      "step": 6556
    },
    {
      "epoch": 1.41,
      "grad_norm": 8.115928649902344,
      "learning_rate": 8.902831402831403e-06,
      "loss": 0.4752,
      "step": 6557
    },
    {
      "epoch": 1.41,
      "grad_norm": 12.3264799118042,
      "learning_rate": 8.899613899613899e-06,
      "loss": 0.5801,
      "step": 6558
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.676340103149414,
      "learning_rate": 8.896396396396397e-06,
      "loss": 0.5439,
      "step": 6559
    },
    {
      "epoch": 1.41,
      "grad_norm": 13.128365516662598,
      "learning_rate": 8.893178893178894e-06,
      "loss": 0.7761,
      "step": 6560
    },
    {
      "epoch": 1.41,
      "grad_norm": 9.491532325744629,
      "learning_rate": 8.88996138996139e-06,
      "loss": 0.4271,
      "step": 6561
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.818324089050293,
      "learning_rate": 8.886743886743888e-06,
      "loss": 0.5702,
      "step": 6562
    },
    {
      "epoch": 1.41,
      "grad_norm": 8.786713600158691,
      "learning_rate": 8.883526383526383e-06,
      "loss": 0.4284,
      "step": 6563
    },
    {
      "epoch": 1.41,
      "grad_norm": 12.376542091369629,
      "learning_rate": 8.88030888030888e-06,
      "loss": 0.7501,
      "step": 6564
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.746297836303711,
      "learning_rate": 8.877091377091378e-06,
      "loss": 0.7183,
      "step": 6565
    },
    {
      "epoch": 1.41,
      "grad_norm": 8.69247055053711,
      "learning_rate": 8.873873873873874e-06,
      "loss": 0.6717,
      "step": 6566
    },
    {
      "epoch": 1.41,
      "grad_norm": 8.867156982421875,
      "learning_rate": 8.87065637065637e-06,
      "loss": 0.6247,
      "step": 6567
    },
    {
      "epoch": 1.41,
      "grad_norm": 12.236666679382324,
      "learning_rate": 8.867438867438869e-06,
      "loss": 0.5955,
      "step": 6568
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.205656051635742,
      "learning_rate": 8.864221364221365e-06,
      "loss": 0.4722,
      "step": 6569
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.617082595825195,
      "learning_rate": 8.861003861003861e-06,
      "loss": 0.5638,
      "step": 6570
    },
    {
      "epoch": 1.41,
      "grad_norm": 9.342180252075195,
      "learning_rate": 8.85778635778636e-06,
      "loss": 0.5938,
      "step": 6571
    },
    {
      "epoch": 1.41,
      "grad_norm": 12.075029373168945,
      "learning_rate": 8.854568854568854e-06,
      "loss": 0.7043,
      "step": 6572
    },
    {
      "epoch": 1.41,
      "grad_norm": 12.66827392578125,
      "learning_rate": 8.851351351351351e-06,
      "loss": 0.6495,
      "step": 6573
    },
    {
      "epoch": 1.41,
      "grad_norm": 7.898078918457031,
      "learning_rate": 8.848133848133847e-06,
      "loss": 0.4009,
      "step": 6574
    },
    {
      "epoch": 1.41,
      "grad_norm": 9.519030570983887,
      "learning_rate": 8.844916344916345e-06,
      "loss": 0.6113,
      "step": 6575
    },
    {
      "epoch": 1.41,
      "grad_norm": 15.306770324707031,
      "learning_rate": 8.841698841698842e-06,
      "loss": 0.5806,
      "step": 6576
    },
    {
      "epoch": 1.41,
      "grad_norm": 9.21624755859375,
      "learning_rate": 8.838481338481338e-06,
      "loss": 0.6007,
      "step": 6577
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.99435806274414,
      "learning_rate": 8.835263835263836e-06,
      "loss": 0.7672,
      "step": 6578
    },
    {
      "epoch": 1.41,
      "grad_norm": 8.403539657592773,
      "learning_rate": 8.832046332046333e-06,
      "loss": 0.452,
      "step": 6579
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.801168441772461,
      "learning_rate": 8.828828828828828e-06,
      "loss": 0.59,
      "step": 6580
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.1153564453125,
      "learning_rate": 8.825611325611326e-06,
      "loss": 0.5253,
      "step": 6581
    },
    {
      "epoch": 1.41,
      "grad_norm": 12.473978996276855,
      "learning_rate": 8.822393822393822e-06,
      "loss": 0.6415,
      "step": 6582
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.503206253051758,
      "learning_rate": 8.819176319176319e-06,
      "loss": 0.4349,
      "step": 6583
    },
    {
      "epoch": 1.41,
      "grad_norm": 8.336518287658691,
      "learning_rate": 8.815958815958817e-06,
      "loss": 0.4795,
      "step": 6584
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.920021057128906,
      "learning_rate": 8.812741312741313e-06,
      "loss": 0.6659,
      "step": 6585
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.579536437988281,
      "learning_rate": 8.80952380952381e-06,
      "loss": 0.6866,
      "step": 6586
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.876070976257324,
      "learning_rate": 8.806306306306306e-06,
      "loss": 0.5236,
      "step": 6587
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.060585021972656,
      "learning_rate": 8.803088803088804e-06,
      "loss": 0.4974,
      "step": 6588
    },
    {
      "epoch": 1.41,
      "grad_norm": 12.298812866210938,
      "learning_rate": 8.7998712998713e-06,
      "loss": 0.8081,
      "step": 6589
    },
    {
      "epoch": 1.41,
      "grad_norm": 7.280674934387207,
      "learning_rate": 8.796653796653796e-06,
      "loss": 0.3369,
      "step": 6590
    },
    {
      "epoch": 1.41,
      "grad_norm": 12.907698631286621,
      "learning_rate": 8.793436293436294e-06,
      "loss": 0.83,
      "step": 6591
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.143620491027832,
      "learning_rate": 8.79021879021879e-06,
      "loss": 0.5569,
      "step": 6592
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.306180000305176,
      "learning_rate": 8.787001287001287e-06,
      "loss": 0.5861,
      "step": 6593
    },
    {
      "epoch": 1.41,
      "grad_norm": 15.623419761657715,
      "learning_rate": 8.783783783783785e-06,
      "loss": 0.4377,
      "step": 6594
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.279592514038086,
      "learning_rate": 8.780566280566281e-06,
      "loss": 0.6718,
      "step": 6595
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.098777770996094,
      "learning_rate": 8.777348777348778e-06,
      "loss": 0.5233,
      "step": 6596
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.187514305114746,
      "learning_rate": 8.774131274131276e-06,
      "loss": 0.3542,
      "step": 6597
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.139504432678223,
      "learning_rate": 8.77091377091377e-06,
      "loss": 0.5869,
      "step": 6598
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.758601188659668,
      "learning_rate": 8.767696267696267e-06,
      "loss": 0.847,
      "step": 6599
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.941631317138672,
      "learning_rate": 8.764478764478765e-06,
      "loss": 0.7915,
      "step": 6600
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.021463394165039,
      "learning_rate": 8.761261261261262e-06,
      "loss": 0.6121,
      "step": 6601
    },
    {
      "epoch": 1.42,
      "grad_norm": 12.525238990783691,
      "learning_rate": 8.758043758043758e-06,
      "loss": 0.7849,
      "step": 6602
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.830269813537598,
      "learning_rate": 8.754826254826255e-06,
      "loss": 0.7003,
      "step": 6603
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.836006164550781,
      "learning_rate": 8.751608751608753e-06,
      "loss": 0.7395,
      "step": 6604
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.866992950439453,
      "learning_rate": 8.74839124839125e-06,
      "loss": 0.6346,
      "step": 6605
    },
    {
      "epoch": 1.42,
      "grad_norm": 13.439878463745117,
      "learning_rate": 8.745173745173744e-06,
      "loss": 0.5304,
      "step": 6606
    },
    {
      "epoch": 1.42,
      "grad_norm": 12.552800178527832,
      "learning_rate": 8.741956241956242e-06,
      "loss": 0.8603,
      "step": 6607
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.948568344116211,
      "learning_rate": 8.738738738738739e-06,
      "loss": 0.5308,
      "step": 6608
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.737764358520508,
      "learning_rate": 8.735521235521235e-06,
      "loss": 0.4758,
      "step": 6609
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.838640213012695,
      "learning_rate": 8.732303732303733e-06,
      "loss": 0.7568,
      "step": 6610
    },
    {
      "epoch": 1.42,
      "grad_norm": 8.462503433227539,
      "learning_rate": 8.72908622908623e-06,
      "loss": 0.4244,
      "step": 6611
    },
    {
      "epoch": 1.42,
      "grad_norm": 12.55815601348877,
      "learning_rate": 8.725868725868726e-06,
      "loss": 0.9972,
      "step": 6612
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.692358016967773,
      "learning_rate": 8.722651222651224e-06,
      "loss": 0.6644,
      "step": 6613
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.859495162963867,
      "learning_rate": 8.71943371943372e-06,
      "loss": 0.6014,
      "step": 6614
    },
    {
      "epoch": 1.42,
      "grad_norm": 12.784205436706543,
      "learning_rate": 8.716216216216215e-06,
      "loss": 0.7183,
      "step": 6615
    },
    {
      "epoch": 1.42,
      "grad_norm": 18.283708572387695,
      "learning_rate": 8.712998712998712e-06,
      "loss": 0.4873,
      "step": 6616
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.813141822814941,
      "learning_rate": 8.70978120978121e-06,
      "loss": 0.5946,
      "step": 6617
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.801372528076172,
      "learning_rate": 8.706563706563706e-06,
      "loss": 0.799,
      "step": 6618
    },
    {
      "epoch": 1.42,
      "grad_norm": 8.541825294494629,
      "learning_rate": 8.703346203346203e-06,
      "loss": 0.5664,
      "step": 6619
    },
    {
      "epoch": 1.42,
      "grad_norm": 8.239958763122559,
      "learning_rate": 8.700128700128701e-06,
      "loss": 0.4401,
      "step": 6620
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.021510124206543,
      "learning_rate": 8.696911196911197e-06,
      "loss": 0.5268,
      "step": 6621
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.647285461425781,
      "learning_rate": 8.693693693693694e-06,
      "loss": 0.657,
      "step": 6622
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.692704200744629,
      "learning_rate": 8.690476190476192e-06,
      "loss": 0.566,
      "step": 6623
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.752915382385254,
      "learning_rate": 8.687258687258687e-06,
      "loss": 0.5906,
      "step": 6624
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.006624221801758,
      "learning_rate": 8.684041184041183e-06,
      "loss": 0.7129,
      "step": 6625
    },
    {
      "epoch": 1.42,
      "grad_norm": 14.229732513427734,
      "learning_rate": 8.680823680823681e-06,
      "loss": 0.6659,
      "step": 6626
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.281617164611816,
      "learning_rate": 8.677606177606178e-06,
      "loss": 0.6168,
      "step": 6627
    },
    {
      "epoch": 1.42,
      "grad_norm": 7.381348609924316,
      "learning_rate": 8.674388674388674e-06,
      "loss": 0.3314,
      "step": 6628
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.83889102935791,
      "learning_rate": 8.671171171171172e-06,
      "loss": 0.5781,
      "step": 6629
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.788817405700684,
      "learning_rate": 8.667953667953669e-06,
      "loss": 0.5906,
      "step": 6630
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.06147289276123,
      "learning_rate": 8.664736164736165e-06,
      "loss": 0.677,
      "step": 6631
    },
    {
      "epoch": 1.42,
      "grad_norm": 12.213753700256348,
      "learning_rate": 8.661518661518662e-06,
      "loss": 0.7375,
      "step": 6632
    },
    {
      "epoch": 1.42,
      "grad_norm": 12.420419692993164,
      "learning_rate": 8.658301158301158e-06,
      "loss": 0.724,
      "step": 6633
    },
    {
      "epoch": 1.42,
      "grad_norm": 8.222399711608887,
      "learning_rate": 8.655083655083655e-06,
      "loss": 0.506,
      "step": 6634
    },
    {
      "epoch": 1.42,
      "grad_norm": 12.03365707397461,
      "learning_rate": 8.651866151866151e-06,
      "loss": 0.563,
      "step": 6635
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.734453201293945,
      "learning_rate": 8.64864864864865e-06,
      "loss": 0.7249,
      "step": 6636
    },
    {
      "epoch": 1.42,
      "grad_norm": 12.228503227233887,
      "learning_rate": 8.645431145431146e-06,
      "loss": 0.4997,
      "step": 6637
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.432348251342773,
      "learning_rate": 8.642213642213642e-06,
      "loss": 0.5162,
      "step": 6638
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.171280860900879,
      "learning_rate": 8.63899613899614e-06,
      "loss": 0.5985,
      "step": 6639
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.336857795715332,
      "learning_rate": 8.635778635778637e-06,
      "loss": 0.4421,
      "step": 6640
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.442643165588379,
      "learning_rate": 8.632561132561132e-06,
      "loss": 0.4281,
      "step": 6641
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.633660316467285,
      "learning_rate": 8.62934362934363e-06,
      "loss": 0.5391,
      "step": 6642
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.07372760772705,
      "learning_rate": 8.626126126126126e-06,
      "loss": 0.7377,
      "step": 6643
    },
    {
      "epoch": 1.43,
      "grad_norm": 13.131924629211426,
      "learning_rate": 8.622908622908623e-06,
      "loss": 0.5339,
      "step": 6644
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.751506805419922,
      "learning_rate": 8.619691119691119e-06,
      "loss": 0.5334,
      "step": 6645
    },
    {
      "epoch": 1.43,
      "grad_norm": 12.358760833740234,
      "learning_rate": 8.616473616473617e-06,
      "loss": 0.581,
      "step": 6646
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.512619972229004,
      "learning_rate": 8.613256113256114e-06,
      "loss": 0.591,
      "step": 6647
    },
    {
      "epoch": 1.43,
      "grad_norm": 12.212662696838379,
      "learning_rate": 8.61003861003861e-06,
      "loss": 0.627,
      "step": 6648
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.229598045349121,
      "learning_rate": 8.606821106821108e-06,
      "loss": 0.4944,
      "step": 6649
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.945271492004395,
      "learning_rate": 8.603603603603603e-06,
      "loss": 0.4822,
      "step": 6650
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.181884765625,
      "learning_rate": 8.6003861003861e-06,
      "loss": 0.6101,
      "step": 6651
    },
    {
      "epoch": 1.43,
      "grad_norm": 16.833810806274414,
      "learning_rate": 8.597168597168598e-06,
      "loss": 0.7157,
      "step": 6652
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.880590438842773,
      "learning_rate": 8.593951093951094e-06,
      "loss": 0.7795,
      "step": 6653
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.841279983520508,
      "learning_rate": 8.59073359073359e-06,
      "loss": 0.4584,
      "step": 6654
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.484471321105957,
      "learning_rate": 8.587516087516089e-06,
      "loss": 0.5785,
      "step": 6655
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.411406517028809,
      "learning_rate": 8.584298584298585e-06,
      "loss": 0.4289,
      "step": 6656
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.260231971740723,
      "learning_rate": 8.581081081081082e-06,
      "loss": 0.4551,
      "step": 6657
    },
    {
      "epoch": 1.43,
      "grad_norm": 12.471264839172363,
      "learning_rate": 8.57786357786358e-06,
      "loss": 0.518,
      "step": 6658
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.938065528869629,
      "learning_rate": 8.574646074646074e-06,
      "loss": 0.6266,
      "step": 6659
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.330185890197754,
      "learning_rate": 8.571428571428571e-06,
      "loss": 0.5441,
      "step": 6660
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.527573585510254,
      "learning_rate": 8.568211068211067e-06,
      "loss": 0.6905,
      "step": 6661
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.557828903198242,
      "learning_rate": 8.564993564993566e-06,
      "loss": 0.5673,
      "step": 6662
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.631921768188477,
      "learning_rate": 8.561776061776062e-06,
      "loss": 0.4077,
      "step": 6663
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.401269912719727,
      "learning_rate": 8.558558558558558e-06,
      "loss": 0.4144,
      "step": 6664
    },
    {
      "epoch": 1.43,
      "grad_norm": 14.212172508239746,
      "learning_rate": 8.555341055341057e-06,
      "loss": 0.7965,
      "step": 6665
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.933451652526855,
      "learning_rate": 8.552123552123553e-06,
      "loss": 0.5076,
      "step": 6666
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.142087936401367,
      "learning_rate": 8.548906048906048e-06,
      "loss": 0.5936,
      "step": 6667
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.352984428405762,
      "learning_rate": 8.545688545688546e-06,
      "loss": 0.5285,
      "step": 6668
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.770153999328613,
      "learning_rate": 8.542471042471042e-06,
      "loss": 0.6335,
      "step": 6669
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.247640609741211,
      "learning_rate": 8.539253539253539e-06,
      "loss": 0.4756,
      "step": 6670
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.74081802368164,
      "learning_rate": 8.536036036036037e-06,
      "loss": 0.5218,
      "step": 6671
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.318914413452148,
      "learning_rate": 8.532818532818533e-06,
      "loss": 0.5184,
      "step": 6672
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.607584953308105,
      "learning_rate": 8.52960102960103e-06,
      "loss": 0.5588,
      "step": 6673
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.236626625061035,
      "learning_rate": 8.526383526383526e-06,
      "loss": 0.5346,
      "step": 6674
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.803278923034668,
      "learning_rate": 8.523166023166024e-06,
      "loss": 0.4397,
      "step": 6675
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.24655818939209,
      "learning_rate": 8.51994851994852e-06,
      "loss": 0.6628,
      "step": 6676
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.945767402648926,
      "learning_rate": 8.516731016731016e-06,
      "loss": 0.5898,
      "step": 6677
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.536576271057129,
      "learning_rate": 8.513513513513514e-06,
      "loss": 0.5883,
      "step": 6678
    },
    {
      "epoch": 1.43,
      "grad_norm": 12.866883277893066,
      "learning_rate": 8.51029601029601e-06,
      "loss": 0.8725,
      "step": 6679
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.995257377624512,
      "learning_rate": 8.507078507078507e-06,
      "loss": 0.4855,
      "step": 6680
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.929353713989258,
      "learning_rate": 8.503861003861005e-06,
      "loss": 0.5014,
      "step": 6681
    },
    {
      "epoch": 1.43,
      "grad_norm": 12.593605041503906,
      "learning_rate": 8.500643500643501e-06,
      "loss": 0.5475,
      "step": 6682
    },
    {
      "epoch": 1.43,
      "grad_norm": 13.836206436157227,
      "learning_rate": 8.497425997425998e-06,
      "loss": 0.624,
      "step": 6683
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.22904109954834,
      "learning_rate": 8.494208494208496e-06,
      "loss": 0.6121,
      "step": 6684
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.626303672790527,
      "learning_rate": 8.49099099099099e-06,
      "loss": 0.3846,
      "step": 6685
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.89222240447998,
      "learning_rate": 8.487773487773487e-06,
      "loss": 0.6926,
      "step": 6686
    },
    {
      "epoch": 1.43,
      "grad_norm": 10.74537181854248,
      "learning_rate": 8.484555984555985e-06,
      "loss": 0.7048,
      "step": 6687
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.060372352600098,
      "learning_rate": 8.481338481338482e-06,
      "loss": 0.3565,
      "step": 6688
    },
    {
      "epoch": 1.43,
      "grad_norm": 9.940937995910645,
      "learning_rate": 8.478120978120978e-06,
      "loss": 0.6082,
      "step": 6689
    },
    {
      "epoch": 1.44,
      "grad_norm": 11.442227363586426,
      "learning_rate": 8.474903474903475e-06,
      "loss": 0.6472,
      "step": 6690
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.06784725189209,
      "learning_rate": 8.471685971685973e-06,
      "loss": 0.6289,
      "step": 6691
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.656331062316895,
      "learning_rate": 8.46846846846847e-06,
      "loss": 0.6524,
      "step": 6692
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.610892295837402,
      "learning_rate": 8.465250965250964e-06,
      "loss": 0.4078,
      "step": 6693
    },
    {
      "epoch": 1.44,
      "grad_norm": 8.263588905334473,
      "learning_rate": 8.462033462033462e-06,
      "loss": 0.3933,
      "step": 6694
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.411392211914062,
      "learning_rate": 8.458815958815959e-06,
      "loss": 0.7108,
      "step": 6695
    },
    {
      "epoch": 1.44,
      "grad_norm": 8.446610450744629,
      "learning_rate": 8.455598455598455e-06,
      "loss": 0.6237,
      "step": 6696
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.887808799743652,
      "learning_rate": 8.452380952380953e-06,
      "loss": 0.6547,
      "step": 6697
    },
    {
      "epoch": 1.44,
      "grad_norm": 8.593314170837402,
      "learning_rate": 8.44916344916345e-06,
      "loss": 0.4838,
      "step": 6698
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.750635147094727,
      "learning_rate": 8.445945945945946e-06,
      "loss": 0.4378,
      "step": 6699
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.62331485748291,
      "learning_rate": 8.442728442728444e-06,
      "loss": 0.5869,
      "step": 6700
    },
    {
      "epoch": 1.44,
      "grad_norm": 11.99150276184082,
      "learning_rate": 8.43951093951094e-06,
      "loss": 0.5332,
      "step": 6701
    },
    {
      "epoch": 1.44,
      "grad_norm": 13.650615692138672,
      "learning_rate": 8.436293436293435e-06,
      "loss": 0.4593,
      "step": 6702
    },
    {
      "epoch": 1.44,
      "grad_norm": 14.590835571289062,
      "learning_rate": 8.433075933075932e-06,
      "loss": 0.787,
      "step": 6703
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.24910831451416,
      "learning_rate": 8.42985842985843e-06,
      "loss": 0.5718,
      "step": 6704
    },
    {
      "epoch": 1.44,
      "grad_norm": 14.167826652526855,
      "learning_rate": 8.426640926640926e-06,
      "loss": 0.8284,
      "step": 6705
    },
    {
      "epoch": 1.44,
      "grad_norm": 7.141318321228027,
      "learning_rate": 8.423423423423423e-06,
      "loss": 0.3266,
      "step": 6706
    },
    {
      "epoch": 1.44,
      "grad_norm": 12.403409004211426,
      "learning_rate": 8.420205920205921e-06,
      "loss": 0.6318,
      "step": 6707
    },
    {
      "epoch": 1.44,
      "grad_norm": 8.813912391662598,
      "learning_rate": 8.416988416988418e-06,
      "loss": 0.5866,
      "step": 6708
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.234713554382324,
      "learning_rate": 8.413770913770914e-06,
      "loss": 0.4665,
      "step": 6709
    },
    {
      "epoch": 1.44,
      "grad_norm": 7.89685583114624,
      "learning_rate": 8.410553410553412e-06,
      "loss": 0.2966,
      "step": 6710
    },
    {
      "epoch": 1.44,
      "grad_norm": 11.209405899047852,
      "learning_rate": 8.407335907335907e-06,
      "loss": 0.6248,
      "step": 6711
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.580241203308105,
      "learning_rate": 8.404118404118403e-06,
      "loss": 0.7042,
      "step": 6712
    },
    {
      "epoch": 1.44,
      "grad_norm": 12.063618659973145,
      "learning_rate": 8.400900900900901e-06,
      "loss": 0.6003,
      "step": 6713
    },
    {
      "epoch": 1.44,
      "grad_norm": 11.52169132232666,
      "learning_rate": 8.397683397683398e-06,
      "loss": 0.5424,
      "step": 6714
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.640144348144531,
      "learning_rate": 8.394465894465894e-06,
      "loss": 0.5033,
      "step": 6715
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.407087326049805,
      "learning_rate": 8.391248391248393e-06,
      "loss": 0.5611,
      "step": 6716
    },
    {
      "epoch": 1.44,
      "grad_norm": 14.784146308898926,
      "learning_rate": 8.388030888030889e-06,
      "loss": 0.6464,
      "step": 6717
    },
    {
      "epoch": 1.44,
      "grad_norm": 8.298789024353027,
      "learning_rate": 8.384813384813385e-06,
      "loss": 0.5314,
      "step": 6718
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.762864112854004,
      "learning_rate": 8.381595881595882e-06,
      "loss": 0.5477,
      "step": 6719
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.326119422912598,
      "learning_rate": 8.378378378378378e-06,
      "loss": 0.482,
      "step": 6720
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.83774471282959,
      "learning_rate": 8.375160875160875e-06,
      "loss": 0.6666,
      "step": 6721
    },
    {
      "epoch": 1.44,
      "grad_norm": 11.274816513061523,
      "learning_rate": 8.371943371943371e-06,
      "loss": 0.4751,
      "step": 6722
    },
    {
      "epoch": 1.44,
      "grad_norm": 11.375434875488281,
      "learning_rate": 8.36872586872587e-06,
      "loss": 0.6113,
      "step": 6723
    },
    {
      "epoch": 1.44,
      "grad_norm": 8.160884857177734,
      "learning_rate": 8.365508365508366e-06,
      "loss": 0.4764,
      "step": 6724
    },
    {
      "epoch": 1.44,
      "grad_norm": 11.202731132507324,
      "learning_rate": 8.362290862290862e-06,
      "loss": 0.5756,
      "step": 6725
    },
    {
      "epoch": 1.44,
      "grad_norm": 13.1454439163208,
      "learning_rate": 8.35907335907336e-06,
      "loss": 0.715,
      "step": 6726
    },
    {
      "epoch": 1.44,
      "grad_norm": 7.987917900085449,
      "learning_rate": 8.355855855855857e-06,
      "loss": 0.5293,
      "step": 6727
    },
    {
      "epoch": 1.44,
      "grad_norm": 11.423751831054688,
      "learning_rate": 8.352638352638352e-06,
      "loss": 0.4841,
      "step": 6728
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.667183876037598,
      "learning_rate": 8.34942084942085e-06,
      "loss": 0.5784,
      "step": 6729
    },
    {
      "epoch": 1.44,
      "grad_norm": 10.240058898925781,
      "learning_rate": 8.346203346203346e-06,
      "loss": 0.4382,
      "step": 6730
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.83134651184082,
      "learning_rate": 8.342985842985843e-06,
      "loss": 0.4286,
      "step": 6731
    },
    {
      "epoch": 1.44,
      "grad_norm": 8.83073616027832,
      "learning_rate": 8.339768339768339e-06,
      "loss": 0.3481,
      "step": 6732
    },
    {
      "epoch": 1.44,
      "grad_norm": 12.548013687133789,
      "learning_rate": 8.336550836550837e-06,
      "loss": 0.6128,
      "step": 6733
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.500943183898926,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.5116,
      "step": 6734
    },
    {
      "epoch": 1.44,
      "grad_norm": 9.659764289855957,
      "learning_rate": 8.33011583011583e-06,
      "loss": 0.608,
      "step": 6735
    },
    {
      "epoch": 1.44,
      "grad_norm": 15.482324600219727,
      "learning_rate": 8.326898326898328e-06,
      "loss": 0.7747,
      "step": 6736
    },
    {
      "epoch": 1.45,
      "grad_norm": 11.039031028747559,
      "learning_rate": 8.323680823680823e-06,
      "loss": 0.5919,
      "step": 6737
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.49527359008789,
      "learning_rate": 8.32046332046332e-06,
      "loss": 0.5052,
      "step": 6738
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.239442825317383,
      "learning_rate": 8.317245817245818e-06,
      "loss": 0.6198,
      "step": 6739
    },
    {
      "epoch": 1.45,
      "grad_norm": 11.63898754119873,
      "learning_rate": 8.314028314028314e-06,
      "loss": 0.689,
      "step": 6740
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.24828052520752,
      "learning_rate": 8.31081081081081e-06,
      "loss": 0.5613,
      "step": 6741
    },
    {
      "epoch": 1.45,
      "grad_norm": 13.263040542602539,
      "learning_rate": 8.307593307593309e-06,
      "loss": 0.7083,
      "step": 6742
    },
    {
      "epoch": 1.45,
      "grad_norm": 11.243728637695312,
      "learning_rate": 8.304375804375805e-06,
      "loss": 0.5016,
      "step": 6743
    },
    {
      "epoch": 1.45,
      "grad_norm": 8.653467178344727,
      "learning_rate": 8.301158301158302e-06,
      "loss": 0.632,
      "step": 6744
    },
    {
      "epoch": 1.45,
      "grad_norm": 6.664738178253174,
      "learning_rate": 8.2979407979408e-06,
      "loss": 0.4092,
      "step": 6745
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.588616371154785,
      "learning_rate": 8.294723294723295e-06,
      "loss": 0.5554,
      "step": 6746
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.56556510925293,
      "learning_rate": 8.291505791505791e-06,
      "loss": 0.436,
      "step": 6747
    },
    {
      "epoch": 1.45,
      "grad_norm": 7.996731758117676,
      "learning_rate": 8.288288288288287e-06,
      "loss": 0.4457,
      "step": 6748
    },
    {
      "epoch": 1.45,
      "grad_norm": 12.052947998046875,
      "learning_rate": 8.285070785070786e-06,
      "loss": 0.7277,
      "step": 6749
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.53419017791748,
      "learning_rate": 8.281853281853282e-06,
      "loss": 0.6252,
      "step": 6750
    },
    {
      "epoch": 1.45,
      "grad_norm": 16.764240264892578,
      "learning_rate": 8.278635778635778e-06,
      "loss": 0.5844,
      "step": 6751
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.244491577148438,
      "learning_rate": 8.275418275418277e-06,
      "loss": 0.4233,
      "step": 6752
    },
    {
      "epoch": 1.45,
      "grad_norm": 8.90921688079834,
      "learning_rate": 8.272200772200773e-06,
      "loss": 0.4832,
      "step": 6753
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.665393829345703,
      "learning_rate": 8.268983268983268e-06,
      "loss": 0.5542,
      "step": 6754
    },
    {
      "epoch": 1.45,
      "grad_norm": 8.490131378173828,
      "learning_rate": 8.265765765765766e-06,
      "loss": 0.476,
      "step": 6755
    },
    {
      "epoch": 1.45,
      "grad_norm": 13.463126182556152,
      "learning_rate": 8.262548262548262e-06,
      "loss": 0.7712,
      "step": 6756
    },
    {
      "epoch": 1.45,
      "grad_norm": 13.167001724243164,
      "learning_rate": 8.259330759330759e-06,
      "loss": 0.5869,
      "step": 6757
    },
    {
      "epoch": 1.45,
      "grad_norm": 14.090361595153809,
      "learning_rate": 8.256113256113257e-06,
      "loss": 0.5394,
      "step": 6758
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.64168930053711,
      "learning_rate": 8.252895752895753e-06,
      "loss": 0.7078,
      "step": 6759
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.721607208251953,
      "learning_rate": 8.24967824967825e-06,
      "loss": 0.6086,
      "step": 6760
    },
    {
      "epoch": 1.45,
      "grad_norm": 13.73509407043457,
      "learning_rate": 8.246460746460746e-06,
      "loss": 0.5232,
      "step": 6761
    },
    {
      "epoch": 1.45,
      "grad_norm": 13.082063674926758,
      "learning_rate": 8.243243243243245e-06,
      "loss": 0.6371,
      "step": 6762
    },
    {
      "epoch": 1.45,
      "grad_norm": 8.90468692779541,
      "learning_rate": 8.24002574002574e-06,
      "loss": 0.5765,
      "step": 6763
    },
    {
      "epoch": 1.45,
      "grad_norm": 11.600364685058594,
      "learning_rate": 8.236808236808236e-06,
      "loss": 0.7712,
      "step": 6764
    },
    {
      "epoch": 1.45,
      "grad_norm": 11.30432415008545,
      "learning_rate": 8.233590733590734e-06,
      "loss": 0.816,
      "step": 6765
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.051785469055176,
      "learning_rate": 8.23037323037323e-06,
      "loss": 0.5682,
      "step": 6766
    },
    {
      "epoch": 1.45,
      "grad_norm": 12.868947982788086,
      "learning_rate": 8.227155727155727e-06,
      "loss": 0.5363,
      "step": 6767
    },
    {
      "epoch": 1.45,
      "grad_norm": 13.137343406677246,
      "learning_rate": 8.223938223938225e-06,
      "loss": 0.7849,
      "step": 6768
    },
    {
      "epoch": 1.45,
      "grad_norm": 12.02407455444336,
      "learning_rate": 8.220720720720721e-06,
      "loss": 0.5175,
      "step": 6769
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.254332542419434,
      "learning_rate": 8.217503217503218e-06,
      "loss": 0.5441,
      "step": 6770
    },
    {
      "epoch": 1.45,
      "grad_norm": 11.034767150878906,
      "learning_rate": 8.214285714285716e-06,
      "loss": 0.8881,
      "step": 6771
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.060040473937988,
      "learning_rate": 8.21106821106821e-06,
      "loss": 0.4959,
      "step": 6772
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.863129615783691,
      "learning_rate": 8.207850707850707e-06,
      "loss": 0.7512,
      "step": 6773
    },
    {
      "epoch": 1.45,
      "grad_norm": 13.199822425842285,
      "learning_rate": 8.204633204633205e-06,
      "loss": 0.7214,
      "step": 6774
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.731306076049805,
      "learning_rate": 8.201415701415702e-06,
      "loss": 0.5736,
      "step": 6775
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.641324996948242,
      "learning_rate": 8.198198198198198e-06,
      "loss": 0.6082,
      "step": 6776
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.486063957214355,
      "learning_rate": 8.194980694980695e-06,
      "loss": 0.484,
      "step": 6777
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.477164268493652,
      "learning_rate": 8.191763191763193e-06,
      "loss": 0.5738,
      "step": 6778
    },
    {
      "epoch": 1.45,
      "grad_norm": 12.284964561462402,
      "learning_rate": 8.18854568854569e-06,
      "loss": 0.6413,
      "step": 6779
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.517580032348633,
      "learning_rate": 8.185328185328184e-06,
      "loss": 0.499,
      "step": 6780
    },
    {
      "epoch": 1.45,
      "grad_norm": 9.328600883483887,
      "learning_rate": 8.182110682110682e-06,
      "loss": 0.5275,
      "step": 6781
    },
    {
      "epoch": 1.45,
      "grad_norm": 13.838427543640137,
      "learning_rate": 8.178893178893179e-06,
      "loss": 0.6752,
      "step": 6782
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.038884162902832,
      "learning_rate": 8.175675675675675e-06,
      "loss": 0.5017,
      "step": 6783
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.341883659362793,
      "learning_rate": 8.172458172458173e-06,
      "loss": 0.5229,
      "step": 6784
    },
    {
      "epoch": 1.46,
      "grad_norm": 11.887608528137207,
      "learning_rate": 8.16924066924067e-06,
      "loss": 0.544,
      "step": 6785
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.433422088623047,
      "learning_rate": 8.166023166023166e-06,
      "loss": 0.709,
      "step": 6786
    },
    {
      "epoch": 1.46,
      "grad_norm": 11.887358665466309,
      "learning_rate": 8.162805662805664e-06,
      "loss": 0.6806,
      "step": 6787
    },
    {
      "epoch": 1.46,
      "grad_norm": 9.030277252197266,
      "learning_rate": 8.15958815958816e-06,
      "loss": 0.5451,
      "step": 6788
    },
    {
      "epoch": 1.46,
      "grad_norm": 8.036552429199219,
      "learning_rate": 8.156370656370655e-06,
      "loss": 0.4514,
      "step": 6789
    },
    {
      "epoch": 1.46,
      "grad_norm": 7.645504474639893,
      "learning_rate": 8.153153153153152e-06,
      "loss": 0.373,
      "step": 6790
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.637724876403809,
      "learning_rate": 8.14993564993565e-06,
      "loss": 0.5781,
      "step": 6791
    },
    {
      "epoch": 1.46,
      "grad_norm": 9.683504104614258,
      "learning_rate": 8.146718146718147e-06,
      "loss": 0.6454,
      "step": 6792
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.307318687438965,
      "learning_rate": 8.143500643500643e-06,
      "loss": 0.514,
      "step": 6793
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.330479621887207,
      "learning_rate": 8.140283140283141e-06,
      "loss": 0.4661,
      "step": 6794
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.330486297607422,
      "learning_rate": 8.137065637065638e-06,
      "loss": 0.7728,
      "step": 6795
    },
    {
      "epoch": 1.46,
      "grad_norm": 11.875329971313477,
      "learning_rate": 8.133848133848134e-06,
      "loss": 0.7569,
      "step": 6796
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.152606964111328,
      "learning_rate": 8.130630630630632e-06,
      "loss": 0.4689,
      "step": 6797
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.704450607299805,
      "learning_rate": 8.127413127413127e-06,
      "loss": 0.5863,
      "step": 6798
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.283672332763672,
      "learning_rate": 8.124195624195623e-06,
      "loss": 0.4826,
      "step": 6799
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.142906188964844,
      "learning_rate": 8.120978120978122e-06,
      "loss": 0.4442,
      "step": 6800
    },
    {
      "epoch": 1.46,
      "grad_norm": 8.656546592712402,
      "learning_rate": 8.117760617760618e-06,
      "loss": 0.4498,
      "step": 6801
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.191787719726562,
      "learning_rate": 8.114543114543114e-06,
      "loss": 0.5571,
      "step": 6802
    },
    {
      "epoch": 1.46,
      "grad_norm": 14.556913375854492,
      "learning_rate": 8.111325611325613e-06,
      "loss": 0.6972,
      "step": 6803
    },
    {
      "epoch": 1.46,
      "grad_norm": 9.765174865722656,
      "learning_rate": 8.108108108108109e-06,
      "loss": 0.5621,
      "step": 6804
    },
    {
      "epoch": 1.46,
      "grad_norm": 11.675789833068848,
      "learning_rate": 8.104890604890605e-06,
      "loss": 0.4605,
      "step": 6805
    },
    {
      "epoch": 1.46,
      "grad_norm": 11.380899429321289,
      "learning_rate": 8.1016731016731e-06,
      "loss": 0.708,
      "step": 6806
    },
    {
      "epoch": 1.46,
      "grad_norm": 9.422585487365723,
      "learning_rate": 8.098455598455598e-06,
      "loss": 0.7835,
      "step": 6807
    },
    {
      "epoch": 1.46,
      "grad_norm": 9.110347747802734,
      "learning_rate": 8.095238095238095e-06,
      "loss": 0.5573,
      "step": 6808
    },
    {
      "epoch": 1.46,
      "grad_norm": 9.297036170959473,
      "learning_rate": 8.092020592020591e-06,
      "loss": 0.5131,
      "step": 6809
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.280660629272461,
      "learning_rate": 8.08880308880309e-06,
      "loss": 0.7145,
      "step": 6810
    },
    {
      "epoch": 1.46,
      "grad_norm": 8.923484802246094,
      "learning_rate": 8.085585585585586e-06,
      "loss": 0.454,
      "step": 6811
    },
    {
      "epoch": 1.46,
      "grad_norm": 13.994488716125488,
      "learning_rate": 8.082368082368082e-06,
      "loss": 0.6655,
      "step": 6812
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.388103485107422,
      "learning_rate": 8.07915057915058e-06,
      "loss": 0.4334,
      "step": 6813
    },
    {
      "epoch": 1.46,
      "grad_norm": 11.410175323486328,
      "learning_rate": 8.075933075933077e-06,
      "loss": 0.5702,
      "step": 6814
    },
    {
      "epoch": 1.46,
      "grad_norm": 11.489496231079102,
      "learning_rate": 8.072715572715572e-06,
      "loss": 0.6015,
      "step": 6815
    },
    {
      "epoch": 1.46,
      "grad_norm": 8.92212963104248,
      "learning_rate": 8.06949806949807e-06,
      "loss": 0.4231,
      "step": 6816
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.123559951782227,
      "learning_rate": 8.066280566280566e-06,
      "loss": 0.4632,
      "step": 6817
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.43405818939209,
      "learning_rate": 8.063063063063063e-06,
      "loss": 0.8034,
      "step": 6818
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.45401668548584,
      "learning_rate": 8.05984555984556e-06,
      "loss": 0.8494,
      "step": 6819
    },
    {
      "epoch": 1.46,
      "grad_norm": 13.984105110168457,
      "learning_rate": 8.056628056628057e-06,
      "loss": 0.7365,
      "step": 6820
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.490537643432617,
      "learning_rate": 8.053410553410554e-06,
      "loss": 0.6064,
      "step": 6821
    },
    {
      "epoch": 1.46,
      "grad_norm": 8.794524192810059,
      "learning_rate": 8.05019305019305e-06,
      "loss": 0.4195,
      "step": 6822
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.538308143615723,
      "learning_rate": 8.046975546975548e-06,
      "loss": 0.6106,
      "step": 6823
    },
    {
      "epoch": 1.46,
      "grad_norm": 7.556087493896484,
      "learning_rate": 8.043758043758043e-06,
      "loss": 0.4374,
      "step": 6824
    },
    {
      "epoch": 1.46,
      "grad_norm": 8.194212913513184,
      "learning_rate": 8.04054054054054e-06,
      "loss": 0.45,
      "step": 6825
    },
    {
      "epoch": 1.46,
      "grad_norm": 7.717074394226074,
      "learning_rate": 8.037323037323038e-06,
      "loss": 0.3751,
      "step": 6826
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.80569839477539,
      "learning_rate": 8.034105534105534e-06,
      "loss": 0.6348,
      "step": 6827
    },
    {
      "epoch": 1.46,
      "grad_norm": 15.823558807373047,
      "learning_rate": 8.03088803088803e-06,
      "loss": 0.6644,
      "step": 6828
    },
    {
      "epoch": 1.46,
      "grad_norm": 10.757474899291992,
      "learning_rate": 8.027670527670529e-06,
      "loss": 0.6885,
      "step": 6829
    },
    {
      "epoch": 1.47,
      "grad_norm": 13.441444396972656,
      "learning_rate": 8.024453024453025e-06,
      "loss": 0.5354,
      "step": 6830
    },
    {
      "epoch": 1.47,
      "grad_norm": 11.649055480957031,
      "learning_rate": 8.021235521235522e-06,
      "loss": 0.631,
      "step": 6831
    },
    {
      "epoch": 1.47,
      "grad_norm": 11.033647537231445,
      "learning_rate": 8.01801801801802e-06,
      "loss": 0.569,
      "step": 6832
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.403671264648438,
      "learning_rate": 8.014800514800515e-06,
      "loss": 0.5143,
      "step": 6833
    },
    {
      "epoch": 1.47,
      "grad_norm": 11.9056396484375,
      "learning_rate": 8.011583011583011e-06,
      "loss": 0.6763,
      "step": 6834
    },
    {
      "epoch": 1.47,
      "grad_norm": 13.944920539855957,
      "learning_rate": 8.008365508365508e-06,
      "loss": 0.5357,
      "step": 6835
    },
    {
      "epoch": 1.47,
      "grad_norm": 12.307910919189453,
      "learning_rate": 8.005148005148006e-06,
      "loss": 0.6368,
      "step": 6836
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.818414688110352,
      "learning_rate": 8.001930501930502e-06,
      "loss": 0.5153,
      "step": 6837
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.089217185974121,
      "learning_rate": 7.998712998712999e-06,
      "loss": 0.4742,
      "step": 6838
    },
    {
      "epoch": 1.47,
      "grad_norm": 11.051301956176758,
      "learning_rate": 7.995495495495497e-06,
      "loss": 0.6371,
      "step": 6839
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.48458480834961,
      "learning_rate": 7.992277992277993e-06,
      "loss": 0.5048,
      "step": 6840
    },
    {
      "epoch": 1.47,
      "grad_norm": 11.353043556213379,
      "learning_rate": 7.989060489060488e-06,
      "loss": 0.5824,
      "step": 6841
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.971410751342773,
      "learning_rate": 7.985842985842986e-06,
      "loss": 0.5238,
      "step": 6842
    },
    {
      "epoch": 1.47,
      "grad_norm": 14.55504322052002,
      "learning_rate": 7.982625482625483e-06,
      "loss": 0.6223,
      "step": 6843
    },
    {
      "epoch": 1.47,
      "grad_norm": 14.220067024230957,
      "learning_rate": 7.979407979407979e-06,
      "loss": 0.701,
      "step": 6844
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.291683197021484,
      "learning_rate": 7.976190476190477e-06,
      "loss": 0.424,
      "step": 6845
    },
    {
      "epoch": 1.47,
      "grad_norm": 11.72364616394043,
      "learning_rate": 7.972972972972974e-06,
      "loss": 0.4824,
      "step": 6846
    },
    {
      "epoch": 1.47,
      "grad_norm": 8.061595916748047,
      "learning_rate": 7.96975546975547e-06,
      "loss": 0.5269,
      "step": 6847
    },
    {
      "epoch": 1.47,
      "grad_norm": 12.325730323791504,
      "learning_rate": 7.966537966537966e-06,
      "loss": 0.8123,
      "step": 6848
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.150586128234863,
      "learning_rate": 7.963320463320465e-06,
      "loss": 0.4227,
      "step": 6849
    },
    {
      "epoch": 1.47,
      "grad_norm": 13.354206085205078,
      "learning_rate": 7.96010296010296e-06,
      "loss": 0.6886,
      "step": 6850
    },
    {
      "epoch": 1.47,
      "grad_norm": 12.864202499389648,
      "learning_rate": 7.956885456885456e-06,
      "loss": 0.5676,
      "step": 6851
    },
    {
      "epoch": 1.47,
      "grad_norm": 11.620115280151367,
      "learning_rate": 7.953667953667954e-06,
      "loss": 0.726,
      "step": 6852
    },
    {
      "epoch": 1.47,
      "grad_norm": 13.591086387634277,
      "learning_rate": 7.95045045045045e-06,
      "loss": 0.5295,
      "step": 6853
    },
    {
      "epoch": 1.47,
      "grad_norm": 11.81326961517334,
      "learning_rate": 7.947232947232947e-06,
      "loss": 0.5312,
      "step": 6854
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.831013679504395,
      "learning_rate": 7.944015444015445e-06,
      "loss": 0.5767,
      "step": 6855
    },
    {
      "epoch": 1.47,
      "grad_norm": 8.98352336883545,
      "learning_rate": 7.940797940797941e-06,
      "loss": 0.6275,
      "step": 6856
    },
    {
      "epoch": 1.47,
      "grad_norm": 11.748963356018066,
      "learning_rate": 7.937580437580438e-06,
      "loss": 0.4386,
      "step": 6857
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.451261520385742,
      "learning_rate": 7.934362934362936e-06,
      "loss": 0.5214,
      "step": 6858
    },
    {
      "epoch": 1.47,
      "grad_norm": 14.245312690734863,
      "learning_rate": 7.93114543114543e-06,
      "loss": 0.7577,
      "step": 6859
    },
    {
      "epoch": 1.47,
      "grad_norm": 12.221346855163574,
      "learning_rate": 7.927927927927927e-06,
      "loss": 0.599,
      "step": 6860
    },
    {
      "epoch": 1.47,
      "grad_norm": 8.5068359375,
      "learning_rate": 7.924710424710425e-06,
      "loss": 0.492,
      "step": 6861
    },
    {
      "epoch": 1.47,
      "grad_norm": 8.56993293762207,
      "learning_rate": 7.921492921492922e-06,
      "loss": 0.5274,
      "step": 6862
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.27197265625,
      "learning_rate": 7.918275418275418e-06,
      "loss": 0.481,
      "step": 6863
    },
    {
      "epoch": 1.47,
      "grad_norm": 12.968250274658203,
      "learning_rate": 7.915057915057915e-06,
      "loss": 0.5907,
      "step": 6864
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.683408737182617,
      "learning_rate": 7.911840411840413e-06,
      "loss": 0.6471,
      "step": 6865
    },
    {
      "epoch": 1.47,
      "grad_norm": 14.026290893554688,
      "learning_rate": 7.90862290862291e-06,
      "loss": 0.648,
      "step": 6866
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.180994033813477,
      "learning_rate": 7.905405405405404e-06,
      "loss": 0.661,
      "step": 6867
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.798213005065918,
      "learning_rate": 7.902187902187902e-06,
      "loss": 0.3696,
      "step": 6868
    },
    {
      "epoch": 1.47,
      "grad_norm": 10.369603157043457,
      "learning_rate": 7.898970398970399e-06,
      "loss": 0.5646,
      "step": 6869
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.491230964660645,
      "learning_rate": 7.895752895752895e-06,
      "loss": 0.4966,
      "step": 6870
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.415303230285645,
      "learning_rate": 7.892535392535393e-06,
      "loss": 0.5957,
      "step": 6871
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.23666000366211,
      "learning_rate": 7.88931788931789e-06,
      "loss": 0.7113,
      "step": 6872
    },
    {
      "epoch": 1.47,
      "grad_norm": 8.468653678894043,
      "learning_rate": 7.886100386100386e-06,
      "loss": 0.3539,
      "step": 6873
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.726777076721191,
      "learning_rate": 7.882882882882884e-06,
      "loss": 0.5406,
      "step": 6874
    },
    {
      "epoch": 1.47,
      "grad_norm": 8.367538452148438,
      "learning_rate": 7.87966537966538e-06,
      "loss": 0.5132,
      "step": 6875
    },
    {
      "epoch": 1.47,
      "grad_norm": 9.422309875488281,
      "learning_rate": 7.876447876447876e-06,
      "loss": 0.5064,
      "step": 6876
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.900593757629395,
      "learning_rate": 7.873230373230374e-06,
      "loss": 0.7944,
      "step": 6877
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.17431640625,
      "learning_rate": 7.87001287001287e-06,
      "loss": 0.5971,
      "step": 6878
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.7775239944458,
      "learning_rate": 7.866795366795367e-06,
      "loss": 0.6104,
      "step": 6879
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.277740478515625,
      "learning_rate": 7.863577863577863e-06,
      "loss": 0.5284,
      "step": 6880
    },
    {
      "epoch": 1.48,
      "grad_norm": 12.208267211914062,
      "learning_rate": 7.860360360360361e-06,
      "loss": 0.6303,
      "step": 6881
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.075607299804688,
      "learning_rate": 7.857142857142858e-06,
      "loss": 0.4955,
      "step": 6882
    },
    {
      "epoch": 1.48,
      "grad_norm": 7.319411277770996,
      "learning_rate": 7.853925353925354e-06,
      "loss": 0.3475,
      "step": 6883
    },
    {
      "epoch": 1.48,
      "grad_norm": 13.001516342163086,
      "learning_rate": 7.850707850707852e-06,
      "loss": 0.6165,
      "step": 6884
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.798291206359863,
      "learning_rate": 7.847490347490347e-06,
      "loss": 0.5936,
      "step": 6885
    },
    {
      "epoch": 1.48,
      "grad_norm": 7.717795372009277,
      "learning_rate": 7.844272844272843e-06,
      "loss": 0.3763,
      "step": 6886
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.698529243469238,
      "learning_rate": 7.841055341055342e-06,
      "loss": 0.5838,
      "step": 6887
    },
    {
      "epoch": 1.48,
      "grad_norm": 12.113303184509277,
      "learning_rate": 7.837837837837838e-06,
      "loss": 0.8702,
      "step": 6888
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.678671836853027,
      "learning_rate": 7.834620334620335e-06,
      "loss": 0.5842,
      "step": 6889
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.847827911376953,
      "learning_rate": 7.831402831402833e-06,
      "loss": 0.5512,
      "step": 6890
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.741593360900879,
      "learning_rate": 7.828185328185329e-06,
      "loss": 0.6285,
      "step": 6891
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.75109577178955,
      "learning_rate": 7.824967824967826e-06,
      "loss": 0.5509,
      "step": 6892
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.911723136901855,
      "learning_rate": 7.82175032175032e-06,
      "loss": 0.8982,
      "step": 6893
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.589959144592285,
      "learning_rate": 7.818532818532818e-06,
      "loss": 0.6227,
      "step": 6894
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.841647148132324,
      "learning_rate": 7.815315315315315e-06,
      "loss": 0.5528,
      "step": 6895
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.229022026062012,
      "learning_rate": 7.812097812097811e-06,
      "loss": 0.6051,
      "step": 6896
    },
    {
      "epoch": 1.48,
      "grad_norm": 12.1382474899292,
      "learning_rate": 7.80888030888031e-06,
      "loss": 0.576,
      "step": 6897
    },
    {
      "epoch": 1.48,
      "grad_norm": 7.796920299530029,
      "learning_rate": 7.805662805662806e-06,
      "loss": 0.358,
      "step": 6898
    },
    {
      "epoch": 1.48,
      "grad_norm": 7.452349662780762,
      "learning_rate": 7.802445302445302e-06,
      "loss": 0.4501,
      "step": 6899
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.826790809631348,
      "learning_rate": 7.7992277992278e-06,
      "loss": 0.6186,
      "step": 6900
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.974510192871094,
      "learning_rate": 7.796010296010297e-06,
      "loss": 0.8283,
      "step": 6901
    },
    {
      "epoch": 1.48,
      "grad_norm": 12.059252738952637,
      "learning_rate": 7.792792792792792e-06,
      "loss": 0.7428,
      "step": 6902
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.93693733215332,
      "learning_rate": 7.78957528957529e-06,
      "loss": 0.645,
      "step": 6903
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.725439071655273,
      "learning_rate": 7.786357786357786e-06,
      "loss": 0.5479,
      "step": 6904
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.007417678833008,
      "learning_rate": 7.783140283140283e-06,
      "loss": 0.4445,
      "step": 6905
    },
    {
      "epoch": 1.48,
      "grad_norm": 8.1943998336792,
      "learning_rate": 7.77992277992278e-06,
      "loss": 0.4706,
      "step": 6906
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.062432289123535,
      "learning_rate": 7.776705276705277e-06,
      "loss": 0.5225,
      "step": 6907
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.50165843963623,
      "learning_rate": 7.773487773487774e-06,
      "loss": 0.59,
      "step": 6908
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.662631034851074,
      "learning_rate": 7.77027027027027e-06,
      "loss": 0.5996,
      "step": 6909
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.393293380737305,
      "learning_rate": 7.767052767052768e-06,
      "loss": 0.4595,
      "step": 6910
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.084253311157227,
      "learning_rate": 7.763835263835263e-06,
      "loss": 0.5241,
      "step": 6911
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.954118728637695,
      "learning_rate": 7.76061776061776e-06,
      "loss": 0.6501,
      "step": 6912
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.86538314819336,
      "learning_rate": 7.757400257400258e-06,
      "loss": 0.5639,
      "step": 6913
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.557572364807129,
      "learning_rate": 7.754182754182754e-06,
      "loss": 0.5389,
      "step": 6914
    },
    {
      "epoch": 1.48,
      "grad_norm": 12.73644733428955,
      "learning_rate": 7.75096525096525e-06,
      "loss": 0.6615,
      "step": 6915
    },
    {
      "epoch": 1.48,
      "grad_norm": 16.785037994384766,
      "learning_rate": 7.747747747747749e-06,
      "loss": 0.8799,
      "step": 6916
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.80457878112793,
      "learning_rate": 7.744530244530245e-06,
      "loss": 0.5471,
      "step": 6917
    },
    {
      "epoch": 1.48,
      "grad_norm": 13.181736946105957,
      "learning_rate": 7.741312741312742e-06,
      "loss": 0.7311,
      "step": 6918
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.477176666259766,
      "learning_rate": 7.73809523809524e-06,
      "loss": 0.6106,
      "step": 6919
    },
    {
      "epoch": 1.48,
      "grad_norm": 8.879020690917969,
      "learning_rate": 7.734877734877735e-06,
      "loss": 0.4479,
      "step": 6920
    },
    {
      "epoch": 1.48,
      "grad_norm": 10.566399574279785,
      "learning_rate": 7.731660231660231e-06,
      "loss": 0.6889,
      "step": 6921
    },
    {
      "epoch": 1.48,
      "grad_norm": 9.272586822509766,
      "learning_rate": 7.728442728442728e-06,
      "loss": 0.3877,
      "step": 6922
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.098237037658691,
      "learning_rate": 7.725225225225226e-06,
      "loss": 0.6492,
      "step": 6923
    },
    {
      "epoch": 1.49,
      "grad_norm": 11.273629188537598,
      "learning_rate": 7.722007722007722e-06,
      "loss": 0.6193,
      "step": 6924
    },
    {
      "epoch": 1.49,
      "grad_norm": 9.297410011291504,
      "learning_rate": 7.718790218790219e-06,
      "loss": 0.5876,
      "step": 6925
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.253323554992676,
      "learning_rate": 7.715572715572717e-06,
      "loss": 0.7196,
      "step": 6926
    },
    {
      "epoch": 1.49,
      "grad_norm": 12.07715129852295,
      "learning_rate": 7.712355212355213e-06,
      "loss": 0.5576,
      "step": 6927
    },
    {
      "epoch": 1.49,
      "grad_norm": 11.285507202148438,
      "learning_rate": 7.709137709137708e-06,
      "loss": 0.4592,
      "step": 6928
    },
    {
      "epoch": 1.49,
      "grad_norm": 11.078490257263184,
      "learning_rate": 7.705920205920206e-06,
      "loss": 0.4605,
      "step": 6929
    },
    {
      "epoch": 1.49,
      "grad_norm": 11.200446128845215,
      "learning_rate": 7.702702702702703e-06,
      "loss": 0.815,
      "step": 6930
    },
    {
      "epoch": 1.49,
      "grad_norm": 11.357345581054688,
      "learning_rate": 7.699485199485199e-06,
      "loss": 0.8032,
      "step": 6931
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.913156509399414,
      "learning_rate": 7.696267696267697e-06,
      "loss": 0.6497,
      "step": 6932
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.874488830566406,
      "learning_rate": 7.693050193050194e-06,
      "loss": 0.481,
      "step": 6933
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.576268196105957,
      "learning_rate": 7.68983268983269e-06,
      "loss": 0.4301,
      "step": 6934
    },
    {
      "epoch": 1.49,
      "grad_norm": 8.68488597869873,
      "learning_rate": 7.686615186615187e-06,
      "loss": 0.6291,
      "step": 6935
    },
    {
      "epoch": 1.49,
      "grad_norm": 9.163390159606934,
      "learning_rate": 7.683397683397685e-06,
      "loss": 0.5106,
      "step": 6936
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.154144287109375,
      "learning_rate": 7.68018018018018e-06,
      "loss": 0.5966,
      "step": 6937
    },
    {
      "epoch": 1.49,
      "grad_norm": 13.304281234741211,
      "learning_rate": 7.676962676962676e-06,
      "loss": 0.7394,
      "step": 6938
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.221250534057617,
      "learning_rate": 7.673745173745174e-06,
      "loss": 0.5616,
      "step": 6939
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.821033477783203,
      "learning_rate": 7.67052767052767e-06,
      "loss": 0.5445,
      "step": 6940
    },
    {
      "epoch": 1.49,
      "grad_norm": 11.402318954467773,
      "learning_rate": 7.667310167310167e-06,
      "loss": 0.6571,
      "step": 6941
    },
    {
      "epoch": 1.49,
      "grad_norm": 9.25400447845459,
      "learning_rate": 7.664092664092665e-06,
      "loss": 0.5058,
      "step": 6942
    },
    {
      "epoch": 1.49,
      "grad_norm": 9.784470558166504,
      "learning_rate": 7.660875160875162e-06,
      "loss": 0.6038,
      "step": 6943
    },
    {
      "epoch": 1.49,
      "grad_norm": 12.57559871673584,
      "learning_rate": 7.657657657657658e-06,
      "loss": 0.4277,
      "step": 6944
    },
    {
      "epoch": 1.49,
      "grad_norm": 11.062097549438477,
      "learning_rate": 7.654440154440156e-06,
      "loss": 0.6871,
      "step": 6945
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.62734603881836,
      "learning_rate": 7.651222651222651e-06,
      "loss": 0.9346,
      "step": 6946
    },
    {
      "epoch": 1.49,
      "grad_norm": 9.060091018676758,
      "learning_rate": 7.648005148005147e-06,
      "loss": 0.4856,
      "step": 6947
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.113350868225098,
      "learning_rate": 7.644787644787645e-06,
      "loss": 0.5816,
      "step": 6948
    },
    {
      "epoch": 1.49,
      "grad_norm": 9.894210815429688,
      "learning_rate": 7.641570141570142e-06,
      "loss": 0.6125,
      "step": 6949
    },
    {
      "epoch": 1.49,
      "grad_norm": 9.568633079528809,
      "learning_rate": 7.638352638352638e-06,
      "loss": 0.5673,
      "step": 6950
    },
    {
      "epoch": 1.49,
      "grad_norm": 7.4964213371276855,
      "learning_rate": 7.635135135135135e-06,
      "loss": 0.4515,
      "step": 6951
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.526020050048828,
      "learning_rate": 7.631917631917633e-06,
      "loss": 0.5488,
      "step": 6952
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.766468048095703,
      "learning_rate": 7.6287001287001294e-06,
      "loss": 0.7754,
      "step": 6953
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.99411392211914,
      "learning_rate": 7.625482625482625e-06,
      "loss": 0.9704,
      "step": 6954
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.062816619873047,
      "learning_rate": 7.6222651222651215e-06,
      "loss": 0.6171,
      "step": 6955
    },
    {
      "epoch": 1.49,
      "grad_norm": 10.107940673828125,
      "learning_rate": 7.619047619047619e-06,
      "loss": 0.5632,
      "step": 6956
    },
    {
      "epoch": 1.49,
      "grad_norm": 11.633281707763672,
      "learning_rate": 7.615830115830116e-06,
      "loss": 0.8027,
      "step": 6957
    },
    {
      "epoch": 1.49,
      "grad_norm": 8.867230415344238,
      "learning_rate": 7.6126126126126125e-06,
      "loss": 0.514,
      "step": 6958
    },
    {
      "epoch": 1.49,
      "grad_norm": 12.545758247375488,
      "learning_rate": 7.60939510939511e-06,
      "loss": 0.5934,
      "step": 6959
    },
    {
      "epoch": 1.49,
      "grad_norm": 13.618131637573242,
      "learning_rate": 7.606177606177607e-06,
      "loss": 0.673,
      "step": 6960
    },
    {
      "epoch": 1.49,
      "grad_norm": 7.920790195465088,
      "learning_rate": 7.602960102960104e-06,
      "loss": 0.4994,
      "step": 6961
    },
    {
      "epoch": 1.49,
      "grad_norm": 9.460268020629883,
      "learning_rate": 7.599742599742601e-06,
      "loss": 0.3005,
      "step": 6962
    },
    {
      "epoch": 1.49,
      "grad_norm": 9.116876602172852,
      "learning_rate": 7.596525096525096e-06,
      "loss": 0.3972,
      "step": 6963
    },
    {
      "epoch": 1.49,
      "grad_norm": 13.716431617736816,
      "learning_rate": 7.593307593307593e-06,
      "loss": 0.5057,
      "step": 6964
    },
    {
      "epoch": 1.49,
      "grad_norm": 9.570103645324707,
      "learning_rate": 7.59009009009009e-06,
      "loss": 0.5627,
      "step": 6965
    },
    {
      "epoch": 1.49,
      "grad_norm": 9.955166816711426,
      "learning_rate": 7.586872586872587e-06,
      "loss": 0.5277,
      "step": 6966
    },
    {
      "epoch": 1.49,
      "grad_norm": 13.144523620605469,
      "learning_rate": 7.583655083655084e-06,
      "loss": 0.8768,
      "step": 6967
    },
    {
      "epoch": 1.49,
      "grad_norm": 11.620397567749023,
      "learning_rate": 7.5804375804375804e-06,
      "loss": 0.5058,
      "step": 6968
    },
    {
      "epoch": 1.49,
      "grad_norm": 12.498621940612793,
      "learning_rate": 7.577220077220078e-06,
      "loss": 0.7397,
      "step": 6969
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.28661060333252,
      "learning_rate": 7.574002574002575e-06,
      "loss": 0.6025,
      "step": 6970
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.099294662475586,
      "learning_rate": 7.5707850707850715e-06,
      "loss": 0.4812,
      "step": 6971
    },
    {
      "epoch": 1.5,
      "grad_norm": 12.169379234313965,
      "learning_rate": 7.567567567567567e-06,
      "loss": 0.602,
      "step": 6972
    },
    {
      "epoch": 1.5,
      "grad_norm": 12.821435928344727,
      "learning_rate": 7.564350064350064e-06,
      "loss": 0.6391,
      "step": 6973
    },
    {
      "epoch": 1.5,
      "grad_norm": 8.836165428161621,
      "learning_rate": 7.561132561132561e-06,
      "loss": 0.4079,
      "step": 6974
    },
    {
      "epoch": 1.5,
      "grad_norm": 8.260855674743652,
      "learning_rate": 7.557915057915058e-06,
      "loss": 0.4258,
      "step": 6975
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.667478561401367,
      "learning_rate": 7.554697554697555e-06,
      "loss": 0.4545,
      "step": 6976
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.896492958068848,
      "learning_rate": 7.551480051480052e-06,
      "loss": 0.6434,
      "step": 6977
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.01639461517334,
      "learning_rate": 7.548262548262549e-06,
      "loss": 0.4397,
      "step": 6978
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.46435832977295,
      "learning_rate": 7.545045045045046e-06,
      "loss": 0.6404,
      "step": 6979
    },
    {
      "epoch": 1.5,
      "grad_norm": 15.241106033325195,
      "learning_rate": 7.541827541827541e-06,
      "loss": 0.6603,
      "step": 6980
    },
    {
      "epoch": 1.5,
      "grad_norm": 8.794514656066895,
      "learning_rate": 7.5386100386100385e-06,
      "loss": 0.4481,
      "step": 6981
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.407390594482422,
      "learning_rate": 7.535392535392535e-06,
      "loss": 0.514,
      "step": 6982
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.588397026062012,
      "learning_rate": 7.532175032175032e-06,
      "loss": 0.6978,
      "step": 6983
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.742705345153809,
      "learning_rate": 7.528957528957529e-06,
      "loss": 0.4812,
      "step": 6984
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.654431343078613,
      "learning_rate": 7.525740025740026e-06,
      "loss": 0.5808,
      "step": 6985
    },
    {
      "epoch": 1.5,
      "grad_norm": 12.71933650970459,
      "learning_rate": 7.522522522522523e-06,
      "loss": 0.6509,
      "step": 6986
    },
    {
      "epoch": 1.5,
      "grad_norm": 8.531550407409668,
      "learning_rate": 7.51930501930502e-06,
      "loss": 0.3535,
      "step": 6987
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.117488861083984,
      "learning_rate": 7.516087516087517e-06,
      "loss": 0.5823,
      "step": 6988
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.056893348693848,
      "learning_rate": 7.512870012870013e-06,
      "loss": 0.6267,
      "step": 6989
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.435637474060059,
      "learning_rate": 7.509652509652509e-06,
      "loss": 0.6588,
      "step": 6990
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.870551109313965,
      "learning_rate": 7.5064350064350064e-06,
      "loss": 0.5722,
      "step": 6991
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.117162704467773,
      "learning_rate": 7.503217503217503e-06,
      "loss": 0.589,
      "step": 6992
    },
    {
      "epoch": 1.5,
      "grad_norm": 12.664082527160645,
      "learning_rate": 7.5e-06,
      "loss": 0.7869,
      "step": 6993
    },
    {
      "epoch": 1.5,
      "grad_norm": 12.525108337402344,
      "learning_rate": 7.4967824967824975e-06,
      "loss": 0.6361,
      "step": 6994
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.875874519348145,
      "learning_rate": 7.493564993564994e-06,
      "loss": 0.7608,
      "step": 6995
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.837019920349121,
      "learning_rate": 7.49034749034749e-06,
      "loss": 0.6703,
      "step": 6996
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.144347190856934,
      "learning_rate": 7.487129987129987e-06,
      "loss": 0.4527,
      "step": 6997
    },
    {
      "epoch": 1.5,
      "grad_norm": 13.88058090209961,
      "learning_rate": 7.483912483912484e-06,
      "loss": 0.4462,
      "step": 6998
    },
    {
      "epoch": 1.5,
      "grad_norm": 8.329233169555664,
      "learning_rate": 7.4806949806949814e-06,
      "loss": 0.4762,
      "step": 6999
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.080964088439941,
      "learning_rate": 7.477477477477477e-06,
      "loss": 0.4628,
      "step": 7000
    },
    {
      "epoch": 1.5,
      "grad_norm": 12.538115501403809,
      "learning_rate": 7.474259974259974e-06,
      "loss": 0.7216,
      "step": 7001
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.888996124267578,
      "learning_rate": 7.471042471042472e-06,
      "loss": 0.5314,
      "step": 7002
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.66618537902832,
      "learning_rate": 7.467824967824968e-06,
      "loss": 0.5957,
      "step": 7003
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.139415740966797,
      "learning_rate": 7.4646074646074645e-06,
      "loss": 0.4558,
      "step": 7004
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.341646194458008,
      "learning_rate": 7.461389961389961e-06,
      "loss": 0.6797,
      "step": 7005
    },
    {
      "epoch": 1.5,
      "grad_norm": 13.430399894714355,
      "learning_rate": 7.458172458172458e-06,
      "loss": 0.5529,
      "step": 7006
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.933709144592285,
      "learning_rate": 7.454954954954956e-06,
      "loss": 0.8529,
      "step": 7007
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.52640151977539,
      "learning_rate": 7.451737451737452e-06,
      "loss": 0.4471,
      "step": 7008
    },
    {
      "epoch": 1.5,
      "grad_norm": 8.797947883605957,
      "learning_rate": 7.4485199485199485e-06,
      "loss": 0.4241,
      "step": 7009
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.191524505615234,
      "learning_rate": 7.445302445302446e-06,
      "loss": 0.6858,
      "step": 7010
    },
    {
      "epoch": 1.5,
      "grad_norm": 8.067354202270508,
      "learning_rate": 7.442084942084942e-06,
      "loss": 0.362,
      "step": 7011
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.870619773864746,
      "learning_rate": 7.4388674388674395e-06,
      "loss": 0.484,
      "step": 7012
    },
    {
      "epoch": 1.5,
      "grad_norm": 12.337185859680176,
      "learning_rate": 7.435649935649935e-06,
      "loss": 0.5715,
      "step": 7013
    },
    {
      "epoch": 1.5,
      "grad_norm": 9.344934463500977,
      "learning_rate": 7.4324324324324324e-06,
      "loss": 0.5879,
      "step": 7014
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.235832214355469,
      "learning_rate": 7.42921492921493e-06,
      "loss": 0.5941,
      "step": 7015
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.429084777832031,
      "learning_rate": 7.425997425997426e-06,
      "loss": 0.4605,
      "step": 7016
    },
    {
      "epoch": 1.51,
      "grad_norm": 7.913129806518555,
      "learning_rate": 7.422779922779923e-06,
      "loss": 0.5028,
      "step": 7017
    },
    {
      "epoch": 1.51,
      "grad_norm": 11.670220375061035,
      "learning_rate": 7.41956241956242e-06,
      "loss": 0.6974,
      "step": 7018
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.760784149169922,
      "learning_rate": 7.416344916344916e-06,
      "loss": 0.7515,
      "step": 7019
    },
    {
      "epoch": 1.51,
      "grad_norm": 8.468367576599121,
      "learning_rate": 7.413127413127414e-06,
      "loss": 0.5013,
      "step": 7020
    },
    {
      "epoch": 1.51,
      "grad_norm": 8.927423477172852,
      "learning_rate": 7.40990990990991e-06,
      "loss": 0.4349,
      "step": 7021
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.732887268066406,
      "learning_rate": 7.406692406692407e-06,
      "loss": 0.7117,
      "step": 7022
    },
    {
      "epoch": 1.51,
      "grad_norm": 12.882892608642578,
      "learning_rate": 7.403474903474904e-06,
      "loss": 0.5479,
      "step": 7023
    },
    {
      "epoch": 1.51,
      "grad_norm": 12.581196784973145,
      "learning_rate": 7.4002574002574e-06,
      "loss": 0.4968,
      "step": 7024
    },
    {
      "epoch": 1.51,
      "grad_norm": 16.12222671508789,
      "learning_rate": 7.397039897039898e-06,
      "loss": 0.7927,
      "step": 7025
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.246081352233887,
      "learning_rate": 7.393822393822393e-06,
      "loss": 0.6518,
      "step": 7026
    },
    {
      "epoch": 1.51,
      "grad_norm": 7.984472751617432,
      "learning_rate": 7.3906048906048905e-06,
      "loss": 0.4259,
      "step": 7027
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.776894569396973,
      "learning_rate": 7.387387387387388e-06,
      "loss": 0.5496,
      "step": 7028
    },
    {
      "epoch": 1.51,
      "grad_norm": 11.07243824005127,
      "learning_rate": 7.384169884169884e-06,
      "loss": 0.4958,
      "step": 7029
    },
    {
      "epoch": 1.51,
      "grad_norm": 9.07424545288086,
      "learning_rate": 7.380952380952381e-06,
      "loss": 0.6499,
      "step": 7030
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.479846000671387,
      "learning_rate": 7.377734877734878e-06,
      "loss": 0.6908,
      "step": 7031
    },
    {
      "epoch": 1.51,
      "grad_norm": 11.610418319702148,
      "learning_rate": 7.3745173745173745e-06,
      "loss": 0.6775,
      "step": 7032
    },
    {
      "epoch": 1.51,
      "grad_norm": 13.8910493850708,
      "learning_rate": 7.371299871299872e-06,
      "loss": 0.635,
      "step": 7033
    },
    {
      "epoch": 1.51,
      "grad_norm": 12.724730491638184,
      "learning_rate": 7.368082368082368e-06,
      "loss": 0.6098,
      "step": 7034
    },
    {
      "epoch": 1.51,
      "grad_norm": 9.039070129394531,
      "learning_rate": 7.364864864864865e-06,
      "loss": 0.6399,
      "step": 7035
    },
    {
      "epoch": 1.51,
      "grad_norm": 9.770770072937012,
      "learning_rate": 7.361647361647362e-06,
      "loss": 0.4789,
      "step": 7036
    },
    {
      "epoch": 1.51,
      "grad_norm": 15.279160499572754,
      "learning_rate": 7.3584298584298585e-06,
      "loss": 0.6598,
      "step": 7037
    },
    {
      "epoch": 1.51,
      "grad_norm": 8.43989086151123,
      "learning_rate": 7.355212355212356e-06,
      "loss": 0.4951,
      "step": 7038
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.170294761657715,
      "learning_rate": 7.351994851994852e-06,
      "loss": 0.5015,
      "step": 7039
    },
    {
      "epoch": 1.51,
      "grad_norm": 8.55246639251709,
      "learning_rate": 7.348777348777349e-06,
      "loss": 0.3393,
      "step": 7040
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.05068588256836,
      "learning_rate": 7.345559845559846e-06,
      "loss": 0.4996,
      "step": 7041
    },
    {
      "epoch": 1.51,
      "grad_norm": 8.605865478515625,
      "learning_rate": 7.342342342342342e-06,
      "loss": 0.4502,
      "step": 7042
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.062267303466797,
      "learning_rate": 7.33912483912484e-06,
      "loss": 0.3734,
      "step": 7043
    },
    {
      "epoch": 1.51,
      "grad_norm": 11.12800407409668,
      "learning_rate": 7.335907335907336e-06,
      "loss": 0.4513,
      "step": 7044
    },
    {
      "epoch": 1.51,
      "grad_norm": 12.389505386352539,
      "learning_rate": 7.332689832689833e-06,
      "loss": 0.7233,
      "step": 7045
    },
    {
      "epoch": 1.51,
      "grad_norm": 12.237783432006836,
      "learning_rate": 7.32947232947233e-06,
      "loss": 0.5211,
      "step": 7046
    },
    {
      "epoch": 1.51,
      "grad_norm": 9.361289024353027,
      "learning_rate": 7.326254826254827e-06,
      "loss": 0.6546,
      "step": 7047
    },
    {
      "epoch": 1.51,
      "grad_norm": 14.001768112182617,
      "learning_rate": 7.323037323037323e-06,
      "loss": 0.6684,
      "step": 7048
    },
    {
      "epoch": 1.51,
      "grad_norm": 11.024197578430176,
      "learning_rate": 7.31981981981982e-06,
      "loss": 0.6239,
      "step": 7049
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.571304321289062,
      "learning_rate": 7.3166023166023166e-06,
      "loss": 0.5438,
      "step": 7050
    },
    {
      "epoch": 1.51,
      "grad_norm": 13.273632049560547,
      "learning_rate": 7.313384813384814e-06,
      "loss": 0.5791,
      "step": 7051
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.998658180236816,
      "learning_rate": 7.31016731016731e-06,
      "loss": 0.5003,
      "step": 7052
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.871478080749512,
      "learning_rate": 7.306949806949807e-06,
      "loss": 0.5011,
      "step": 7053
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.934552192687988,
      "learning_rate": 7.303732303732304e-06,
      "loss": 0.676,
      "step": 7054
    },
    {
      "epoch": 1.51,
      "grad_norm": 13.929145812988281,
      "learning_rate": 7.3005148005148005e-06,
      "loss": 0.5529,
      "step": 7055
    },
    {
      "epoch": 1.51,
      "grad_norm": 8.577910423278809,
      "learning_rate": 7.297297297297298e-06,
      "loss": 0.4411,
      "step": 7056
    },
    {
      "epoch": 1.51,
      "grad_norm": 13.220200538635254,
      "learning_rate": 7.294079794079794e-06,
      "loss": 0.6305,
      "step": 7057
    },
    {
      "epoch": 1.51,
      "grad_norm": 12.245271682739258,
      "learning_rate": 7.290862290862291e-06,
      "loss": 0.6387,
      "step": 7058
    },
    {
      "epoch": 1.51,
      "grad_norm": 9.829793930053711,
      "learning_rate": 7.287644787644788e-06,
      "loss": 0.6212,
      "step": 7059
    },
    {
      "epoch": 1.51,
      "grad_norm": 13.752132415771484,
      "learning_rate": 7.284427284427285e-06,
      "loss": 0.7881,
      "step": 7060
    },
    {
      "epoch": 1.51,
      "grad_norm": 11.673895835876465,
      "learning_rate": 7.281209781209781e-06,
      "loss": 0.5675,
      "step": 7061
    },
    {
      "epoch": 1.51,
      "grad_norm": 11.235307693481445,
      "learning_rate": 7.277992277992278e-06,
      "loss": 0.4619,
      "step": 7062
    },
    {
      "epoch": 1.52,
      "grad_norm": 11.794342994689941,
      "learning_rate": 7.274774774774775e-06,
      "loss": 0.5255,
      "step": 7063
    },
    {
      "epoch": 1.52,
      "grad_norm": 8.805100440979004,
      "learning_rate": 7.271557271557272e-06,
      "loss": 0.5351,
      "step": 7064
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.951614379882812,
      "learning_rate": 7.268339768339768e-06,
      "loss": 0.6485,
      "step": 7065
    },
    {
      "epoch": 1.52,
      "grad_norm": 7.57703161239624,
      "learning_rate": 7.265122265122265e-06,
      "loss": 0.4569,
      "step": 7066
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.013513565063477,
      "learning_rate": 7.261904761904762e-06,
      "loss": 0.5789,
      "step": 7067
    },
    {
      "epoch": 1.52,
      "grad_norm": 7.973456382751465,
      "learning_rate": 7.2586872586872595e-06,
      "loss": 0.4576,
      "step": 7068
    },
    {
      "epoch": 1.52,
      "grad_norm": 12.957731246948242,
      "learning_rate": 7.255469755469756e-06,
      "loss": 0.7944,
      "step": 7069
    },
    {
      "epoch": 1.52,
      "grad_norm": 11.039209365844727,
      "learning_rate": 7.252252252252252e-06,
      "loss": 0.6483,
      "step": 7070
    },
    {
      "epoch": 1.52,
      "grad_norm": 12.07095718383789,
      "learning_rate": 7.249034749034749e-06,
      "loss": 0.61,
      "step": 7071
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.425020217895508,
      "learning_rate": 7.245817245817246e-06,
      "loss": 0.5327,
      "step": 7072
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.719114303588867,
      "learning_rate": 7.242599742599743e-06,
      "loss": 0.6659,
      "step": 7073
    },
    {
      "epoch": 1.52,
      "grad_norm": 13.267812728881836,
      "learning_rate": 7.239382239382239e-06,
      "loss": 0.637,
      "step": 7074
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.267725944519043,
      "learning_rate": 7.236164736164736e-06,
      "loss": 0.5036,
      "step": 7075
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.288797378540039,
      "learning_rate": 7.232947232947234e-06,
      "loss": 0.5837,
      "step": 7076
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.264446258544922,
      "learning_rate": 7.22972972972973e-06,
      "loss": 0.4076,
      "step": 7077
    },
    {
      "epoch": 1.52,
      "grad_norm": 11.221858024597168,
      "learning_rate": 7.2265122265122265e-06,
      "loss": 0.5647,
      "step": 7078
    },
    {
      "epoch": 1.52,
      "grad_norm": 11.681816101074219,
      "learning_rate": 7.223294723294723e-06,
      "loss": 0.5237,
      "step": 7079
    },
    {
      "epoch": 1.52,
      "grad_norm": 12.585213661193848,
      "learning_rate": 7.22007722007722e-06,
      "loss": 0.5868,
      "step": 7080
    },
    {
      "epoch": 1.52,
      "grad_norm": 7.296591281890869,
      "learning_rate": 7.2168597168597176e-06,
      "loss": 0.2746,
      "step": 7081
    },
    {
      "epoch": 1.52,
      "grad_norm": 8.968282699584961,
      "learning_rate": 7.213642213642214e-06,
      "loss": 0.5596,
      "step": 7082
    },
    {
      "epoch": 1.52,
      "grad_norm": 13.452356338500977,
      "learning_rate": 7.2104247104247105e-06,
      "loss": 0.5695,
      "step": 7083
    },
    {
      "epoch": 1.52,
      "grad_norm": 15.237629890441895,
      "learning_rate": 7.207207207207207e-06,
      "loss": 0.6737,
      "step": 7084
    },
    {
      "epoch": 1.52,
      "grad_norm": 7.725273609161377,
      "learning_rate": 7.203989703989704e-06,
      "loss": 0.4981,
      "step": 7085
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.747424125671387,
      "learning_rate": 7.2007722007722015e-06,
      "loss": 0.3759,
      "step": 7086
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.736578941345215,
      "learning_rate": 7.197554697554697e-06,
      "loss": 0.6889,
      "step": 7087
    },
    {
      "epoch": 1.52,
      "grad_norm": 12.580384254455566,
      "learning_rate": 7.194337194337194e-06,
      "loss": 0.7664,
      "step": 7088
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.614937782287598,
      "learning_rate": 7.191119691119692e-06,
      "loss": 0.644,
      "step": 7089
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.005681037902832,
      "learning_rate": 7.187902187902188e-06,
      "loss": 0.5859,
      "step": 7090
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.314537048339844,
      "learning_rate": 7.184684684684685e-06,
      "loss": 0.4931,
      "step": 7091
    },
    {
      "epoch": 1.52,
      "grad_norm": 8.853194236755371,
      "learning_rate": 7.181467181467181e-06,
      "loss": 0.4074,
      "step": 7092
    },
    {
      "epoch": 1.52,
      "grad_norm": 13.071059226989746,
      "learning_rate": 7.178249678249678e-06,
      "loss": 0.6516,
      "step": 7093
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.169093132019043,
      "learning_rate": 7.175032175032176e-06,
      "loss": 0.4224,
      "step": 7094
    },
    {
      "epoch": 1.52,
      "grad_norm": 11.331669807434082,
      "learning_rate": 7.171814671814672e-06,
      "loss": 0.7848,
      "step": 7095
    },
    {
      "epoch": 1.52,
      "grad_norm": 13.73763656616211,
      "learning_rate": 7.1685971685971686e-06,
      "loss": 0.9275,
      "step": 7096
    },
    {
      "epoch": 1.52,
      "grad_norm": 13.986745834350586,
      "learning_rate": 7.165379665379666e-06,
      "loss": 0.7697,
      "step": 7097
    },
    {
      "epoch": 1.52,
      "grad_norm": 7.543327331542969,
      "learning_rate": 7.162162162162162e-06,
      "loss": 0.3656,
      "step": 7098
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.207103729248047,
      "learning_rate": 7.15894465894466e-06,
      "loss": 0.5104,
      "step": 7099
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.644760131835938,
      "learning_rate": 7.155727155727155e-06,
      "loss": 0.505,
      "step": 7100
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.813685417175293,
      "learning_rate": 7.1525096525096525e-06,
      "loss": 0.7565,
      "step": 7101
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.060440063476562,
      "learning_rate": 7.14929214929215e-06,
      "loss": 0.6404,
      "step": 7102
    },
    {
      "epoch": 1.52,
      "grad_norm": 11.567472457885742,
      "learning_rate": 7.146074646074646e-06,
      "loss": 0.6813,
      "step": 7103
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.265435218811035,
      "learning_rate": 7.142857142857143e-06,
      "loss": 0.5057,
      "step": 7104
    },
    {
      "epoch": 1.52,
      "grad_norm": 9.681018829345703,
      "learning_rate": 7.13963963963964e-06,
      "loss": 0.4133,
      "step": 7105
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.314081192016602,
      "learning_rate": 7.1364221364221365e-06,
      "loss": 0.4036,
      "step": 7106
    },
    {
      "epoch": 1.52,
      "grad_norm": 8.642311096191406,
      "learning_rate": 7.133204633204634e-06,
      "loss": 0.4367,
      "step": 7107
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.789618492126465,
      "learning_rate": 7.12998712998713e-06,
      "loss": 0.4834,
      "step": 7108
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.866251945495605,
      "learning_rate": 7.126769626769627e-06,
      "loss": 0.6836,
      "step": 7109
    },
    {
      "epoch": 1.53,
      "grad_norm": 9.314248085021973,
      "learning_rate": 7.123552123552124e-06,
      "loss": 0.492,
      "step": 7110
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.041568756103516,
      "learning_rate": 7.12033462033462e-06,
      "loss": 0.3345,
      "step": 7111
    },
    {
      "epoch": 1.53,
      "grad_norm": 13.126811981201172,
      "learning_rate": 7.117117117117118e-06,
      "loss": 0.6821,
      "step": 7112
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.854290962219238,
      "learning_rate": 7.113899613899613e-06,
      "loss": 0.5511,
      "step": 7113
    },
    {
      "epoch": 1.53,
      "grad_norm": 9.829558372497559,
      "learning_rate": 7.110682110682111e-06,
      "loss": 0.5448,
      "step": 7114
    },
    {
      "epoch": 1.53,
      "grad_norm": 8.151616096496582,
      "learning_rate": 7.107464607464608e-06,
      "loss": 0.3066,
      "step": 7115
    },
    {
      "epoch": 1.53,
      "grad_norm": 12.626976013183594,
      "learning_rate": 7.104247104247104e-06,
      "loss": 0.7277,
      "step": 7116
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.578887939453125,
      "learning_rate": 7.101029601029601e-06,
      "loss": 0.6647,
      "step": 7117
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.615313529968262,
      "learning_rate": 7.097812097812098e-06,
      "loss": 0.5556,
      "step": 7118
    },
    {
      "epoch": 1.53,
      "grad_norm": 9.83862590789795,
      "learning_rate": 7.0945945945945946e-06,
      "loss": 0.5016,
      "step": 7119
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.592687606811523,
      "learning_rate": 7.091377091377092e-06,
      "loss": 0.5115,
      "step": 7120
    },
    {
      "epoch": 1.53,
      "grad_norm": 9.616715431213379,
      "learning_rate": 7.088159588159588e-06,
      "loss": 0.534,
      "step": 7121
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.533947944641113,
      "learning_rate": 7.084942084942085e-06,
      "loss": 0.4856,
      "step": 7122
    },
    {
      "epoch": 1.53,
      "grad_norm": 14.073720932006836,
      "learning_rate": 7.081724581724582e-06,
      "loss": 0.7889,
      "step": 7123
    },
    {
      "epoch": 1.53,
      "grad_norm": 7.991637229919434,
      "learning_rate": 7.0785070785070785e-06,
      "loss": 0.4809,
      "step": 7124
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.474357604980469,
      "learning_rate": 7.075289575289576e-06,
      "loss": 0.4852,
      "step": 7125
    },
    {
      "epoch": 1.53,
      "grad_norm": 13.552656173706055,
      "learning_rate": 7.072072072072072e-06,
      "loss": 0.7692,
      "step": 7126
    },
    {
      "epoch": 1.53,
      "grad_norm": 8.804022789001465,
      "learning_rate": 7.068854568854569e-06,
      "loss": 0.4032,
      "step": 7127
    },
    {
      "epoch": 1.53,
      "grad_norm": 12.246576309204102,
      "learning_rate": 7.065637065637066e-06,
      "loss": 0.758,
      "step": 7128
    },
    {
      "epoch": 1.53,
      "grad_norm": 13.068318367004395,
      "learning_rate": 7.0624195624195625e-06,
      "loss": 0.6807,
      "step": 7129
    },
    {
      "epoch": 1.53,
      "grad_norm": 9.851566314697266,
      "learning_rate": 7.059202059202059e-06,
      "loss": 0.6832,
      "step": 7130
    },
    {
      "epoch": 1.53,
      "grad_norm": 8.188239097595215,
      "learning_rate": 7.055984555984556e-06,
      "loss": 0.3913,
      "step": 7131
    },
    {
      "epoch": 1.53,
      "grad_norm": 13.179548263549805,
      "learning_rate": 7.052767052767053e-06,
      "loss": 0.5464,
      "step": 7132
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.18707275390625,
      "learning_rate": 7.04954954954955e-06,
      "loss": 0.7386,
      "step": 7133
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.965587615966797,
      "learning_rate": 7.046332046332047e-06,
      "loss": 0.4181,
      "step": 7134
    },
    {
      "epoch": 1.53,
      "grad_norm": 12.487892150878906,
      "learning_rate": 7.043114543114543e-06,
      "loss": 0.594,
      "step": 7135
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.461325645446777,
      "learning_rate": 7.03989703989704e-06,
      "loss": 0.5822,
      "step": 7136
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.570404052734375,
      "learning_rate": 7.036679536679537e-06,
      "loss": 0.5794,
      "step": 7137
    },
    {
      "epoch": 1.53,
      "grad_norm": 9.543643951416016,
      "learning_rate": 7.033462033462034e-06,
      "loss": 0.5843,
      "step": 7138
    },
    {
      "epoch": 1.53,
      "grad_norm": 9.46340274810791,
      "learning_rate": 7.03024453024453e-06,
      "loss": 0.5389,
      "step": 7139
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.207396507263184,
      "learning_rate": 7.027027027027027e-06,
      "loss": 0.4708,
      "step": 7140
    },
    {
      "epoch": 1.53,
      "grad_norm": 9.840047836303711,
      "learning_rate": 7.023809523809524e-06,
      "loss": 0.503,
      "step": 7141
    },
    {
      "epoch": 1.53,
      "grad_norm": 7.380924701690674,
      "learning_rate": 7.0205920205920206e-06,
      "loss": 0.4641,
      "step": 7142
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.888279914855957,
      "learning_rate": 7.017374517374518e-06,
      "loss": 0.5201,
      "step": 7143
    },
    {
      "epoch": 1.53,
      "grad_norm": 8.878870964050293,
      "learning_rate": 7.014157014157014e-06,
      "loss": 0.6435,
      "step": 7144
    },
    {
      "epoch": 1.53,
      "grad_norm": 12.412318229675293,
      "learning_rate": 7.010939510939511e-06,
      "loss": 0.5656,
      "step": 7145
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.513644218444824,
      "learning_rate": 7.007722007722008e-06,
      "loss": 0.7347,
      "step": 7146
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.99889087677002,
      "learning_rate": 7.004504504504505e-06,
      "loss": 0.7106,
      "step": 7147
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.947152137756348,
      "learning_rate": 7.001287001287001e-06,
      "loss": 0.4887,
      "step": 7148
    },
    {
      "epoch": 1.53,
      "grad_norm": 8.823015213012695,
      "learning_rate": 6.998069498069498e-06,
      "loss": 0.3268,
      "step": 7149
    },
    {
      "epoch": 1.53,
      "grad_norm": 10.957659721374512,
      "learning_rate": 6.994851994851995e-06,
      "loss": 0.4947,
      "step": 7150
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.234158515930176,
      "learning_rate": 6.991634491634492e-06,
      "loss": 0.4512,
      "step": 7151
    },
    {
      "epoch": 1.53,
      "grad_norm": 13.211038589477539,
      "learning_rate": 6.9884169884169885e-06,
      "loss": 0.7296,
      "step": 7152
    },
    {
      "epoch": 1.53,
      "grad_norm": 13.080586433410645,
      "learning_rate": 6.985199485199485e-06,
      "loss": 0.6068,
      "step": 7153
    },
    {
      "epoch": 1.53,
      "grad_norm": 12.908069610595703,
      "learning_rate": 6.981981981981982e-06,
      "loss": 0.5396,
      "step": 7154
    },
    {
      "epoch": 1.53,
      "grad_norm": 13.429213523864746,
      "learning_rate": 6.9787644787644795e-06,
      "loss": 0.686,
      "step": 7155
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.850992202758789,
      "learning_rate": 6.975546975546976e-06,
      "loss": 0.6877,
      "step": 7156
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.598437309265137,
      "learning_rate": 6.9723294723294724e-06,
      "loss": 0.5022,
      "step": 7157
    },
    {
      "epoch": 1.54,
      "grad_norm": 11.852668762207031,
      "learning_rate": 6.969111969111969e-06,
      "loss": 0.6406,
      "step": 7158
    },
    {
      "epoch": 1.54,
      "grad_norm": 9.898855209350586,
      "learning_rate": 6.965894465894466e-06,
      "loss": 0.5015,
      "step": 7159
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.697954177856445,
      "learning_rate": 6.9626769626769635e-06,
      "loss": 0.5981,
      "step": 7160
    },
    {
      "epoch": 1.54,
      "grad_norm": 12.952840805053711,
      "learning_rate": 6.959459459459459e-06,
      "loss": 0.5364,
      "step": 7161
    },
    {
      "epoch": 1.54,
      "grad_norm": 9.48649787902832,
      "learning_rate": 6.956241956241956e-06,
      "loss": 0.6119,
      "step": 7162
    },
    {
      "epoch": 1.54,
      "grad_norm": 8.919635772705078,
      "learning_rate": 6.953024453024454e-06,
      "loss": 0.4763,
      "step": 7163
    },
    {
      "epoch": 1.54,
      "grad_norm": 12.902894973754883,
      "learning_rate": 6.94980694980695e-06,
      "loss": 0.7654,
      "step": 7164
    },
    {
      "epoch": 1.54,
      "grad_norm": 7.820339679718018,
      "learning_rate": 6.946589446589447e-06,
      "loss": 0.3807,
      "step": 7165
    },
    {
      "epoch": 1.54,
      "grad_norm": 15.20125675201416,
      "learning_rate": 6.943371943371943e-06,
      "loss": 0.4996,
      "step": 7166
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.027134895324707,
      "learning_rate": 6.94015444015444e-06,
      "loss": 0.5009,
      "step": 7167
    },
    {
      "epoch": 1.54,
      "grad_norm": 9.634027481079102,
      "learning_rate": 6.936936936936938e-06,
      "loss": 0.5482,
      "step": 7168
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.385451316833496,
      "learning_rate": 6.933719433719434e-06,
      "loss": 0.7325,
      "step": 7169
    },
    {
      "epoch": 1.54,
      "grad_norm": 8.698718070983887,
      "learning_rate": 6.9305019305019305e-06,
      "loss": 0.4892,
      "step": 7170
    },
    {
      "epoch": 1.54,
      "grad_norm": 11.996838569641113,
      "learning_rate": 6.927284427284427e-06,
      "loss": 0.6074,
      "step": 7171
    },
    {
      "epoch": 1.54,
      "grad_norm": 13.715780258178711,
      "learning_rate": 6.924066924066924e-06,
      "loss": 0.4541,
      "step": 7172
    },
    {
      "epoch": 1.54,
      "grad_norm": 11.190897941589355,
      "learning_rate": 6.920849420849422e-06,
      "loss": 0.5345,
      "step": 7173
    },
    {
      "epoch": 1.54,
      "grad_norm": 14.424599647521973,
      "learning_rate": 6.917631917631917e-06,
      "loss": 0.5085,
      "step": 7174
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.904134750366211,
      "learning_rate": 6.9144144144144145e-06,
      "loss": 0.6266,
      "step": 7175
    },
    {
      "epoch": 1.54,
      "grad_norm": 11.049324989318848,
      "learning_rate": 6.911196911196912e-06,
      "loss": 0.6651,
      "step": 7176
    },
    {
      "epoch": 1.54,
      "grad_norm": 12.375825881958008,
      "learning_rate": 6.907979407979408e-06,
      "loss": 0.5639,
      "step": 7177
    },
    {
      "epoch": 1.54,
      "grad_norm": 9.237290382385254,
      "learning_rate": 6.904761904761905e-06,
      "loss": 0.4635,
      "step": 7178
    },
    {
      "epoch": 1.54,
      "grad_norm": 11.365610122680664,
      "learning_rate": 6.901544401544401e-06,
      "loss": 0.7517,
      "step": 7179
    },
    {
      "epoch": 1.54,
      "grad_norm": 12.359386444091797,
      "learning_rate": 6.8983268983268984e-06,
      "loss": 0.4252,
      "step": 7180
    },
    {
      "epoch": 1.54,
      "grad_norm": 8.237447738647461,
      "learning_rate": 6.895109395109396e-06,
      "loss": 0.3534,
      "step": 7181
    },
    {
      "epoch": 1.54,
      "grad_norm": 9.06650447845459,
      "learning_rate": 6.891891891891892e-06,
      "loss": 0.6617,
      "step": 7182
    },
    {
      "epoch": 1.54,
      "grad_norm": 12.271733283996582,
      "learning_rate": 6.888674388674389e-06,
      "loss": 0.6857,
      "step": 7183
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.32021713256836,
      "learning_rate": 6.885456885456886e-06,
      "loss": 0.5474,
      "step": 7184
    },
    {
      "epoch": 1.54,
      "grad_norm": 12.492837905883789,
      "learning_rate": 6.882239382239382e-06,
      "loss": 0.5982,
      "step": 7185
    },
    {
      "epoch": 1.54,
      "grad_norm": 7.356375694274902,
      "learning_rate": 6.87902187902188e-06,
      "loss": 0.3596,
      "step": 7186
    },
    {
      "epoch": 1.54,
      "grad_norm": 8.76328182220459,
      "learning_rate": 6.875804375804375e-06,
      "loss": 0.543,
      "step": 7187
    },
    {
      "epoch": 1.54,
      "grad_norm": 13.43427562713623,
      "learning_rate": 6.872586872586873e-06,
      "loss": 0.584,
      "step": 7188
    },
    {
      "epoch": 1.54,
      "grad_norm": 11.002180099487305,
      "learning_rate": 6.86936936936937e-06,
      "loss": 0.4373,
      "step": 7189
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.474088668823242,
      "learning_rate": 6.866151866151866e-06,
      "loss": 0.5059,
      "step": 7190
    },
    {
      "epoch": 1.54,
      "grad_norm": 14.423788070678711,
      "learning_rate": 6.862934362934363e-06,
      "loss": 0.7597,
      "step": 7191
    },
    {
      "epoch": 1.54,
      "grad_norm": 8.846627235412598,
      "learning_rate": 6.85971685971686e-06,
      "loss": 0.4638,
      "step": 7192
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.152060508728027,
      "learning_rate": 6.8564993564993565e-06,
      "loss": 0.4704,
      "step": 7193
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.294210433959961,
      "learning_rate": 6.853281853281854e-06,
      "loss": 0.3915,
      "step": 7194
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.08164119720459,
      "learning_rate": 6.85006435006435e-06,
      "loss": 0.531,
      "step": 7195
    },
    {
      "epoch": 1.54,
      "grad_norm": 6.900068283081055,
      "learning_rate": 6.846846846846847e-06,
      "loss": 0.3478,
      "step": 7196
    },
    {
      "epoch": 1.54,
      "grad_norm": 9.73946475982666,
      "learning_rate": 6.843629343629344e-06,
      "loss": 0.6316,
      "step": 7197
    },
    {
      "epoch": 1.54,
      "grad_norm": 9.475144386291504,
      "learning_rate": 6.8404118404118405e-06,
      "loss": 0.4981,
      "step": 7198
    },
    {
      "epoch": 1.54,
      "grad_norm": 10.029437065124512,
      "learning_rate": 6.837194337194338e-06,
      "loss": 0.3554,
      "step": 7199
    },
    {
      "epoch": 1.54,
      "grad_norm": 8.640899658203125,
      "learning_rate": 6.833976833976834e-06,
      "loss": 0.5226,
      "step": 7200
    },
    {
      "epoch": 1.54,
      "grad_norm": 11.375310897827148,
      "learning_rate": 6.830759330759331e-06,
      "loss": 0.7174,
      "step": 7201
    },
    {
      "epoch": 1.54,
      "grad_norm": 9.841044425964355,
      "learning_rate": 6.827541827541828e-06,
      "loss": 0.4315,
      "step": 7202
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.524249076843262,
      "learning_rate": 6.8243243243243244e-06,
      "loss": 0.516,
      "step": 7203
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.926637649536133,
      "learning_rate": 6.821106821106821e-06,
      "loss": 0.4752,
      "step": 7204
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.540305137634277,
      "learning_rate": 6.817889317889318e-06,
      "loss": 0.5488,
      "step": 7205
    },
    {
      "epoch": 1.55,
      "grad_norm": 12.991191864013672,
      "learning_rate": 6.814671814671815e-06,
      "loss": 0.6721,
      "step": 7206
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.927919387817383,
      "learning_rate": 6.811454311454312e-06,
      "loss": 0.6726,
      "step": 7207
    },
    {
      "epoch": 1.55,
      "grad_norm": 13.049697875976562,
      "learning_rate": 6.808236808236808e-06,
      "loss": 0.8279,
      "step": 7208
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.636462211608887,
      "learning_rate": 6.805019305019305e-06,
      "loss": 0.5979,
      "step": 7209
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.610857963562012,
      "learning_rate": 6.801801801801802e-06,
      "loss": 0.6333,
      "step": 7210
    },
    {
      "epoch": 1.55,
      "grad_norm": 8.419028282165527,
      "learning_rate": 6.798584298584299e-06,
      "loss": 0.5034,
      "step": 7211
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.264290809631348,
      "learning_rate": 6.795366795366796e-06,
      "loss": 0.519,
      "step": 7212
    },
    {
      "epoch": 1.55,
      "grad_norm": 12.879317283630371,
      "learning_rate": 6.792149292149292e-06,
      "loss": 0.6829,
      "step": 7213
    },
    {
      "epoch": 1.55,
      "grad_norm": 15.379496574401855,
      "learning_rate": 6.788931788931789e-06,
      "loss": 0.6058,
      "step": 7214
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.877914428710938,
      "learning_rate": 6.785714285714286e-06,
      "loss": 0.4878,
      "step": 7215
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.343688011169434,
      "learning_rate": 6.7824967824967825e-06,
      "loss": 0.4535,
      "step": 7216
    },
    {
      "epoch": 1.55,
      "grad_norm": 7.907857894897461,
      "learning_rate": 6.779279279279279e-06,
      "loss": 0.4683,
      "step": 7217
    },
    {
      "epoch": 1.55,
      "grad_norm": 14.420211791992188,
      "learning_rate": 6.776061776061776e-06,
      "loss": 0.5406,
      "step": 7218
    },
    {
      "epoch": 1.55,
      "grad_norm": 11.689910888671875,
      "learning_rate": 6.772844272844273e-06,
      "loss": 0.5928,
      "step": 7219
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.696907043457031,
      "learning_rate": 6.76962676962677e-06,
      "loss": 0.4499,
      "step": 7220
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.836286544799805,
      "learning_rate": 6.766409266409267e-06,
      "loss": 0.5565,
      "step": 7221
    },
    {
      "epoch": 1.55,
      "grad_norm": 8.246121406555176,
      "learning_rate": 6.763191763191763e-06,
      "loss": 0.3562,
      "step": 7222
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.925443649291992,
      "learning_rate": 6.75997425997426e-06,
      "loss": 0.5067,
      "step": 7223
    },
    {
      "epoch": 1.55,
      "grad_norm": 11.279953956604004,
      "learning_rate": 6.756756756756757e-06,
      "loss": 0.7218,
      "step": 7224
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.191397666931152,
      "learning_rate": 6.753539253539254e-06,
      "loss": 0.5316,
      "step": 7225
    },
    {
      "epoch": 1.55,
      "grad_norm": 11.039493560791016,
      "learning_rate": 6.7503217503217504e-06,
      "loss": 0.4837,
      "step": 7226
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.395283699035645,
      "learning_rate": 6.747104247104247e-06,
      "loss": 0.5679,
      "step": 7227
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.825567245483398,
      "learning_rate": 6.743886743886744e-06,
      "loss": 0.4866,
      "step": 7228
    },
    {
      "epoch": 1.55,
      "grad_norm": 7.156013011932373,
      "learning_rate": 6.740669240669241e-06,
      "loss": 0.3226,
      "step": 7229
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.87321949005127,
      "learning_rate": 6.737451737451738e-06,
      "loss": 0.6248,
      "step": 7230
    },
    {
      "epoch": 1.55,
      "grad_norm": 8.174570083618164,
      "learning_rate": 6.734234234234234e-06,
      "loss": 0.435,
      "step": 7231
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.197868347167969,
      "learning_rate": 6.731016731016731e-06,
      "loss": 0.4782,
      "step": 7232
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.05507755279541,
      "learning_rate": 6.727799227799228e-06,
      "loss": 0.3702,
      "step": 7233
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.6591157913208,
      "learning_rate": 6.7245817245817254e-06,
      "loss": 0.6018,
      "step": 7234
    },
    {
      "epoch": 1.55,
      "grad_norm": 8.035320281982422,
      "learning_rate": 6.721364221364221e-06,
      "loss": 0.4615,
      "step": 7235
    },
    {
      "epoch": 1.55,
      "grad_norm": 16.050737380981445,
      "learning_rate": 6.718146718146718e-06,
      "loss": 0.7608,
      "step": 7236
    },
    {
      "epoch": 1.55,
      "grad_norm": 13.698060035705566,
      "learning_rate": 6.714929214929215e-06,
      "loss": 0.7141,
      "step": 7237
    },
    {
      "epoch": 1.55,
      "grad_norm": 8.747855186462402,
      "learning_rate": 6.711711711711712e-06,
      "loss": 0.5471,
      "step": 7238
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.963765144348145,
      "learning_rate": 6.7084942084942085e-06,
      "loss": 0.7205,
      "step": 7239
    },
    {
      "epoch": 1.55,
      "grad_norm": 15.180716514587402,
      "learning_rate": 6.705276705276705e-06,
      "loss": 0.6213,
      "step": 7240
    },
    {
      "epoch": 1.55,
      "grad_norm": 12.375876426696777,
      "learning_rate": 6.702059202059202e-06,
      "loss": 0.6065,
      "step": 7241
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.439138412475586,
      "learning_rate": 6.6988416988417e-06,
      "loss": 0.5009,
      "step": 7242
    },
    {
      "epoch": 1.55,
      "grad_norm": 12.490211486816406,
      "learning_rate": 6.695624195624196e-06,
      "loss": 0.7077,
      "step": 7243
    },
    {
      "epoch": 1.55,
      "grad_norm": 12.894441604614258,
      "learning_rate": 6.6924066924066925e-06,
      "loss": 0.4612,
      "step": 7244
    },
    {
      "epoch": 1.55,
      "grad_norm": 13.660070419311523,
      "learning_rate": 6.689189189189189e-06,
      "loss": 0.6453,
      "step": 7245
    },
    {
      "epoch": 1.55,
      "grad_norm": 12.779606819152832,
      "learning_rate": 6.685971685971686e-06,
      "loss": 0.6116,
      "step": 7246
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.746508598327637,
      "learning_rate": 6.6827541827541835e-06,
      "loss": 0.5024,
      "step": 7247
    },
    {
      "epoch": 1.55,
      "grad_norm": 13.675291061401367,
      "learning_rate": 6.679536679536679e-06,
      "loss": 0.6713,
      "step": 7248
    },
    {
      "epoch": 1.55,
      "grad_norm": 10.06087589263916,
      "learning_rate": 6.6763191763191764e-06,
      "loss": 0.3726,
      "step": 7249
    },
    {
      "epoch": 1.56,
      "grad_norm": 11.931218147277832,
      "learning_rate": 6.673101673101674e-06,
      "loss": 0.5297,
      "step": 7250
    },
    {
      "epoch": 1.56,
      "grad_norm": 8.954782485961914,
      "learning_rate": 6.66988416988417e-06,
      "loss": 0.4568,
      "step": 7251
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.695687294006348,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.5382,
      "step": 7252
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.757903099060059,
      "learning_rate": 6.663449163449163e-06,
      "loss": 0.5118,
      "step": 7253
    },
    {
      "epoch": 1.56,
      "grad_norm": 12.293940544128418,
      "learning_rate": 6.66023166023166e-06,
      "loss": 0.5201,
      "step": 7254
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.52849006652832,
      "learning_rate": 6.657014157014158e-06,
      "loss": 0.5614,
      "step": 7255
    },
    {
      "epoch": 1.56,
      "grad_norm": 12.278661727905273,
      "learning_rate": 6.653796653796654e-06,
      "loss": 0.5821,
      "step": 7256
    },
    {
      "epoch": 1.56,
      "grad_norm": 15.321968078613281,
      "learning_rate": 6.650579150579151e-06,
      "loss": 0.4663,
      "step": 7257
    },
    {
      "epoch": 1.56,
      "grad_norm": 7.250400066375732,
      "learning_rate": 6.647361647361648e-06,
      "loss": 0.3013,
      "step": 7258
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.419771194458008,
      "learning_rate": 6.644144144144144e-06,
      "loss": 0.3958,
      "step": 7259
    },
    {
      "epoch": 1.56,
      "grad_norm": 8.976139068603516,
      "learning_rate": 6.640926640926642e-06,
      "loss": 0.492,
      "step": 7260
    },
    {
      "epoch": 1.56,
      "grad_norm": 12.851411819458008,
      "learning_rate": 6.637709137709137e-06,
      "loss": 0.4413,
      "step": 7261
    },
    {
      "epoch": 1.56,
      "grad_norm": 11.370124816894531,
      "learning_rate": 6.6344916344916345e-06,
      "loss": 0.5385,
      "step": 7262
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.141386985778809,
      "learning_rate": 6.631274131274132e-06,
      "loss": 0.4431,
      "step": 7263
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.621561050415039,
      "learning_rate": 6.628056628056628e-06,
      "loss": 0.4272,
      "step": 7264
    },
    {
      "epoch": 1.56,
      "grad_norm": 11.602530479431152,
      "learning_rate": 6.624839124839125e-06,
      "loss": 0.7535,
      "step": 7265
    },
    {
      "epoch": 1.56,
      "grad_norm": 13.718002319335938,
      "learning_rate": 6.621621621621621e-06,
      "loss": 0.555,
      "step": 7266
    },
    {
      "epoch": 1.56,
      "grad_norm": 13.189579963684082,
      "learning_rate": 6.6184041184041185e-06,
      "loss": 0.6257,
      "step": 7267
    },
    {
      "epoch": 1.56,
      "grad_norm": 13.495157241821289,
      "learning_rate": 6.615186615186616e-06,
      "loss": 0.6475,
      "step": 7268
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.497366905212402,
      "learning_rate": 6.611969111969112e-06,
      "loss": 0.4645,
      "step": 7269
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.816563606262207,
      "learning_rate": 6.608751608751609e-06,
      "loss": 0.5019,
      "step": 7270
    },
    {
      "epoch": 1.56,
      "grad_norm": 13.356399536132812,
      "learning_rate": 6.605534105534106e-06,
      "loss": 0.9083,
      "step": 7271
    },
    {
      "epoch": 1.56,
      "grad_norm": 13.631193161010742,
      "learning_rate": 6.6023166023166025e-06,
      "loss": 0.6018,
      "step": 7272
    },
    {
      "epoch": 1.56,
      "grad_norm": 12.365489959716797,
      "learning_rate": 6.5990990990991e-06,
      "loss": 0.6207,
      "step": 7273
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.156846046447754,
      "learning_rate": 6.595881595881595e-06,
      "loss": 0.4209,
      "step": 7274
    },
    {
      "epoch": 1.56,
      "grad_norm": 13.053959846496582,
      "learning_rate": 6.592664092664093e-06,
      "loss": 0.5692,
      "step": 7275
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.389660835266113,
      "learning_rate": 6.58944658944659e-06,
      "loss": 0.575,
      "step": 7276
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.534908294677734,
      "learning_rate": 6.586229086229086e-06,
      "loss": 0.5154,
      "step": 7277
    },
    {
      "epoch": 1.56,
      "grad_norm": 13.222909927368164,
      "learning_rate": 6.583011583011583e-06,
      "loss": 0.6573,
      "step": 7278
    },
    {
      "epoch": 1.56,
      "grad_norm": 13.159906387329102,
      "learning_rate": 6.57979407979408e-06,
      "loss": 0.6405,
      "step": 7279
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.719954490661621,
      "learning_rate": 6.576576576576577e-06,
      "loss": 0.6178,
      "step": 7280
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.995505332946777,
      "learning_rate": 6.573359073359074e-06,
      "loss": 0.457,
      "step": 7281
    },
    {
      "epoch": 1.56,
      "grad_norm": 11.741860389709473,
      "learning_rate": 6.57014157014157e-06,
      "loss": 0.7914,
      "step": 7282
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.550664901733398,
      "learning_rate": 6.566924066924067e-06,
      "loss": 0.5195,
      "step": 7283
    },
    {
      "epoch": 1.56,
      "grad_norm": 12.460522651672363,
      "learning_rate": 6.563706563706564e-06,
      "loss": 0.7744,
      "step": 7284
    },
    {
      "epoch": 1.56,
      "grad_norm": 8.929073333740234,
      "learning_rate": 6.5604890604890606e-06,
      "loss": 0.3475,
      "step": 7285
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.080497741699219,
      "learning_rate": 6.557271557271558e-06,
      "loss": 0.3983,
      "step": 7286
    },
    {
      "epoch": 1.56,
      "grad_norm": 8.692151069641113,
      "learning_rate": 6.554054054054054e-06,
      "loss": 0.3977,
      "step": 7287
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.337675094604492,
      "learning_rate": 6.550836550836551e-06,
      "loss": 0.7457,
      "step": 7288
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.707769393920898,
      "learning_rate": 6.547619047619048e-06,
      "loss": 0.667,
      "step": 7289
    },
    {
      "epoch": 1.56,
      "grad_norm": 8.832732200622559,
      "learning_rate": 6.5444015444015445e-06,
      "loss": 0.5391,
      "step": 7290
    },
    {
      "epoch": 1.56,
      "grad_norm": 13.13879680633545,
      "learning_rate": 6.541184041184041e-06,
      "loss": 0.5323,
      "step": 7291
    },
    {
      "epoch": 1.56,
      "grad_norm": 12.308133125305176,
      "learning_rate": 6.537966537966538e-06,
      "loss": 0.7143,
      "step": 7292
    },
    {
      "epoch": 1.56,
      "grad_norm": 10.3101806640625,
      "learning_rate": 6.534749034749035e-06,
      "loss": 0.6183,
      "step": 7293
    },
    {
      "epoch": 1.56,
      "grad_norm": 14.275890350341797,
      "learning_rate": 6.531531531531532e-06,
      "loss": 0.5956,
      "step": 7294
    },
    {
      "epoch": 1.56,
      "grad_norm": 9.974764823913574,
      "learning_rate": 6.5283140283140285e-06,
      "loss": 0.5673,
      "step": 7295
    },
    {
      "epoch": 1.56,
      "grad_norm": 8.124897003173828,
      "learning_rate": 6.525096525096525e-06,
      "loss": 0.405,
      "step": 7296
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.329668998718262,
      "learning_rate": 6.521879021879022e-06,
      "loss": 0.5677,
      "step": 7297
    },
    {
      "epoch": 1.57,
      "grad_norm": 14.408679962158203,
      "learning_rate": 6.518661518661519e-06,
      "loss": 0.7865,
      "step": 7298
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.049324035644531,
      "learning_rate": 6.515444015444016e-06,
      "loss": 0.5135,
      "step": 7299
    },
    {
      "epoch": 1.57,
      "grad_norm": 11.489585876464844,
      "learning_rate": 6.512226512226512e-06,
      "loss": 0.4476,
      "step": 7300
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.836689949035645,
      "learning_rate": 6.509009009009009e-06,
      "loss": 0.5318,
      "step": 7301
    },
    {
      "epoch": 1.57,
      "grad_norm": 11.65380859375,
      "learning_rate": 6.505791505791506e-06,
      "loss": 0.6492,
      "step": 7302
    },
    {
      "epoch": 1.57,
      "grad_norm": 11.842164993286133,
      "learning_rate": 6.502574002574003e-06,
      "loss": 0.5741,
      "step": 7303
    },
    {
      "epoch": 1.57,
      "grad_norm": 13.615407943725586,
      "learning_rate": 6.499356499356499e-06,
      "loss": 0.757,
      "step": 7304
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.905867576599121,
      "learning_rate": 6.496138996138996e-06,
      "loss": 0.6015,
      "step": 7305
    },
    {
      "epoch": 1.57,
      "grad_norm": 8.675511360168457,
      "learning_rate": 6.492921492921493e-06,
      "loss": 0.4704,
      "step": 7306
    },
    {
      "epoch": 1.57,
      "grad_norm": 8.126457214355469,
      "learning_rate": 6.48970398970399e-06,
      "loss": 0.4295,
      "step": 7307
    },
    {
      "epoch": 1.57,
      "grad_norm": 11.957507133483887,
      "learning_rate": 6.486486486486487e-06,
      "loss": 0.6141,
      "step": 7308
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.87868881225586,
      "learning_rate": 6.483268983268983e-06,
      "loss": 0.4795,
      "step": 7309
    },
    {
      "epoch": 1.57,
      "grad_norm": 8.9391508102417,
      "learning_rate": 6.48005148005148e-06,
      "loss": 0.5137,
      "step": 7310
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.533799171447754,
      "learning_rate": 6.476833976833977e-06,
      "loss": 0.394,
      "step": 7311
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.453771591186523,
      "learning_rate": 6.473616473616474e-06,
      "loss": 0.5287,
      "step": 7312
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.449750900268555,
      "learning_rate": 6.4703989703989705e-06,
      "loss": 0.6557,
      "step": 7313
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.661185264587402,
      "learning_rate": 6.467181467181467e-06,
      "loss": 0.4985,
      "step": 7314
    },
    {
      "epoch": 1.57,
      "grad_norm": 11.886271476745605,
      "learning_rate": 6.463963963963964e-06,
      "loss": 0.4963,
      "step": 7315
    },
    {
      "epoch": 1.57,
      "grad_norm": 15.162205696105957,
      "learning_rate": 6.4607464607464616e-06,
      "loss": 0.6439,
      "step": 7316
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.824834823608398,
      "learning_rate": 6.457528957528957e-06,
      "loss": 0.5013,
      "step": 7317
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.55231761932373,
      "learning_rate": 6.4543114543114545e-06,
      "loss": 0.6985,
      "step": 7318
    },
    {
      "epoch": 1.57,
      "grad_norm": 13.637072563171387,
      "learning_rate": 6.451093951093951e-06,
      "loss": 0.6978,
      "step": 7319
    },
    {
      "epoch": 1.57,
      "grad_norm": 13.030315399169922,
      "learning_rate": 6.447876447876448e-06,
      "loss": 0.5214,
      "step": 7320
    },
    {
      "epoch": 1.57,
      "grad_norm": 11.615524291992188,
      "learning_rate": 6.4446589446589455e-06,
      "loss": 0.4276,
      "step": 7321
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.505613327026367,
      "learning_rate": 6.441441441441441e-06,
      "loss": 0.5393,
      "step": 7322
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.591684341430664,
      "learning_rate": 6.438223938223938e-06,
      "loss": 0.4508,
      "step": 7323
    },
    {
      "epoch": 1.57,
      "grad_norm": 8.7576265335083,
      "learning_rate": 6.435006435006435e-06,
      "loss": 0.6163,
      "step": 7324
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.486371994018555,
      "learning_rate": 6.431788931788932e-06,
      "loss": 0.4135,
      "step": 7325
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.693131446838379,
      "learning_rate": 6.428571428571429e-06,
      "loss": 0.6765,
      "step": 7326
    },
    {
      "epoch": 1.57,
      "grad_norm": 11.126968383789062,
      "learning_rate": 6.425353925353925e-06,
      "loss": 0.5209,
      "step": 7327
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.199732780456543,
      "learning_rate": 6.422136422136422e-06,
      "loss": 0.6588,
      "step": 7328
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.679689407348633,
      "learning_rate": 6.41891891891892e-06,
      "loss": 0.491,
      "step": 7329
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.155954360961914,
      "learning_rate": 6.415701415701416e-06,
      "loss": 0.5599,
      "step": 7330
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.450560569763184,
      "learning_rate": 6.4124839124839126e-06,
      "loss": 0.6071,
      "step": 7331
    },
    {
      "epoch": 1.57,
      "grad_norm": 11.962911605834961,
      "learning_rate": 6.409266409266409e-06,
      "loss": 0.4702,
      "step": 7332
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.279892921447754,
      "learning_rate": 6.406048906048906e-06,
      "loss": 0.5731,
      "step": 7333
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.43614387512207,
      "learning_rate": 6.402831402831404e-06,
      "loss": 0.4973,
      "step": 7334
    },
    {
      "epoch": 1.57,
      "grad_norm": 13.35091781616211,
      "learning_rate": 6.399613899613899e-06,
      "loss": 0.5388,
      "step": 7335
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.073987007141113,
      "learning_rate": 6.3963963963963965e-06,
      "loss": 0.6652,
      "step": 7336
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.960795402526855,
      "learning_rate": 6.393178893178894e-06,
      "loss": 0.555,
      "step": 7337
    },
    {
      "epoch": 1.57,
      "grad_norm": 12.3216552734375,
      "learning_rate": 6.38996138996139e-06,
      "loss": 0.5948,
      "step": 7338
    },
    {
      "epoch": 1.57,
      "grad_norm": 13.108514785766602,
      "learning_rate": 6.386743886743887e-06,
      "loss": 0.5556,
      "step": 7339
    },
    {
      "epoch": 1.57,
      "grad_norm": 13.270125389099121,
      "learning_rate": 6.383526383526383e-06,
      "loss": 0.7276,
      "step": 7340
    },
    {
      "epoch": 1.57,
      "grad_norm": 8.10543155670166,
      "learning_rate": 6.3803088803088805e-06,
      "loss": 0.4648,
      "step": 7341
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.708698272705078,
      "learning_rate": 6.377091377091378e-06,
      "loss": 0.6497,
      "step": 7342
    },
    {
      "epoch": 1.58,
      "grad_norm": 12.085322380065918,
      "learning_rate": 6.373873873873874e-06,
      "loss": 0.6512,
      "step": 7343
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.248286247253418,
      "learning_rate": 6.370656370656371e-06,
      "loss": 0.5742,
      "step": 7344
    },
    {
      "epoch": 1.58,
      "grad_norm": 6.896910667419434,
      "learning_rate": 6.367438867438868e-06,
      "loss": 0.3712,
      "step": 7345
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.2427396774292,
      "learning_rate": 6.364221364221364e-06,
      "loss": 0.5048,
      "step": 7346
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.884270668029785,
      "learning_rate": 6.361003861003862e-06,
      "loss": 0.5862,
      "step": 7347
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.859156608581543,
      "learning_rate": 6.357786357786357e-06,
      "loss": 0.7508,
      "step": 7348
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.413548469543457,
      "learning_rate": 6.354568854568855e-06,
      "loss": 0.4236,
      "step": 7349
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.977384567260742,
      "learning_rate": 6.351351351351352e-06,
      "loss": 0.5819,
      "step": 7350
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.17349910736084,
      "learning_rate": 6.348133848133848e-06,
      "loss": 0.3556,
      "step": 7351
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.57199764251709,
      "learning_rate": 6.344916344916345e-06,
      "loss": 0.5602,
      "step": 7352
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.594538688659668,
      "learning_rate": 6.341698841698841e-06,
      "loss": 0.6083,
      "step": 7353
    },
    {
      "epoch": 1.58,
      "grad_norm": 15.826937675476074,
      "learning_rate": 6.3384813384813386e-06,
      "loss": 0.6773,
      "step": 7354
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.638530731201172,
      "learning_rate": 6.335263835263836e-06,
      "loss": 0.4934,
      "step": 7355
    },
    {
      "epoch": 1.58,
      "grad_norm": 8.757355690002441,
      "learning_rate": 6.332046332046332e-06,
      "loss": 0.635,
      "step": 7356
    },
    {
      "epoch": 1.58,
      "grad_norm": 8.482718467712402,
      "learning_rate": 6.328828828828829e-06,
      "loss": 0.3939,
      "step": 7357
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.506867408752441,
      "learning_rate": 6.325611325611326e-06,
      "loss": 0.6796,
      "step": 7358
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.177810668945312,
      "learning_rate": 6.3223938223938225e-06,
      "loss": 0.4362,
      "step": 7359
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.783798217773438,
      "learning_rate": 6.31917631917632e-06,
      "loss": 0.6459,
      "step": 7360
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.234658241271973,
      "learning_rate": 6.315958815958815e-06,
      "loss": 0.5804,
      "step": 7361
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.336376190185547,
      "learning_rate": 6.312741312741313e-06,
      "loss": 0.6059,
      "step": 7362
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.711417198181152,
      "learning_rate": 6.30952380952381e-06,
      "loss": 0.8068,
      "step": 7363
    },
    {
      "epoch": 1.58,
      "grad_norm": 13.082701683044434,
      "learning_rate": 6.3063063063063065e-06,
      "loss": 0.5623,
      "step": 7364
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.669142723083496,
      "learning_rate": 6.303088803088803e-06,
      "loss": 0.6287,
      "step": 7365
    },
    {
      "epoch": 1.58,
      "grad_norm": 12.541853904724121,
      "learning_rate": 6.2998712998713e-06,
      "loss": 0.7342,
      "step": 7366
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.69823932647705,
      "learning_rate": 6.296653796653797e-06,
      "loss": 0.598,
      "step": 7367
    },
    {
      "epoch": 1.58,
      "grad_norm": 12.300025939941406,
      "learning_rate": 6.293436293436294e-06,
      "loss": 0.3843,
      "step": 7368
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.258854866027832,
      "learning_rate": 6.29021879021879e-06,
      "loss": 0.4375,
      "step": 7369
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.154510498046875,
      "learning_rate": 6.287001287001287e-06,
      "loss": 0.6516,
      "step": 7370
    },
    {
      "epoch": 1.58,
      "grad_norm": 12.55341911315918,
      "learning_rate": 6.283783783783784e-06,
      "loss": 0.6584,
      "step": 7371
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.300711631774902,
      "learning_rate": 6.280566280566281e-06,
      "loss": 0.5239,
      "step": 7372
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.707707405090332,
      "learning_rate": 6.277348777348778e-06,
      "loss": 0.5582,
      "step": 7373
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.825343132019043,
      "learning_rate": 6.274131274131274e-06,
      "loss": 0.6723,
      "step": 7374
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.037373542785645,
      "learning_rate": 6.270913770913771e-06,
      "loss": 0.6665,
      "step": 7375
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.647271156311035,
      "learning_rate": 6.267696267696268e-06,
      "loss": 0.3463,
      "step": 7376
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.032618522644043,
      "learning_rate": 6.2644787644787646e-06,
      "loss": 0.3111,
      "step": 7377
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.671671867370605,
      "learning_rate": 6.261261261261261e-06,
      "loss": 0.5218,
      "step": 7378
    },
    {
      "epoch": 1.58,
      "grad_norm": 12.361199378967285,
      "learning_rate": 6.258043758043758e-06,
      "loss": 0.4711,
      "step": 7379
    },
    {
      "epoch": 1.58,
      "grad_norm": 13.670161247253418,
      "learning_rate": 6.254826254826255e-06,
      "loss": 0.5942,
      "step": 7380
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.046781539916992,
      "learning_rate": 6.251608751608752e-06,
      "loss": 0.5081,
      "step": 7381
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.514110565185547,
      "learning_rate": 6.2483912483912485e-06,
      "loss": 0.6264,
      "step": 7382
    },
    {
      "epoch": 1.58,
      "grad_norm": 10.528548240661621,
      "learning_rate": 6.245173745173745e-06,
      "loss": 0.6579,
      "step": 7383
    },
    {
      "epoch": 1.58,
      "grad_norm": 13.192953109741211,
      "learning_rate": 6.241956241956242e-06,
      "loss": 0.5721,
      "step": 7384
    },
    {
      "epoch": 1.58,
      "grad_norm": 11.29466724395752,
      "learning_rate": 6.238738738738739e-06,
      "loss": 0.6061,
      "step": 7385
    },
    {
      "epoch": 1.58,
      "grad_norm": 13.574173927307129,
      "learning_rate": 6.235521235521236e-06,
      "loss": 0.6774,
      "step": 7386
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.528806686401367,
      "learning_rate": 6.2323037323037325e-06,
      "loss": 0.6561,
      "step": 7387
    },
    {
      "epoch": 1.58,
      "grad_norm": 8.321700096130371,
      "learning_rate": 6.229086229086229e-06,
      "loss": 0.5705,
      "step": 7388
    },
    {
      "epoch": 1.58,
      "grad_norm": 8.278960227966309,
      "learning_rate": 6.225868725868726e-06,
      "loss": 0.4293,
      "step": 7389
    },
    {
      "epoch": 1.59,
      "grad_norm": 7.418886661529541,
      "learning_rate": 6.222651222651223e-06,
      "loss": 0.4523,
      "step": 7390
    },
    {
      "epoch": 1.59,
      "grad_norm": 8.340412139892578,
      "learning_rate": 6.219433719433719e-06,
      "loss": 0.5423,
      "step": 7391
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.380019187927246,
      "learning_rate": 6.2162162162162164e-06,
      "loss": 0.3558,
      "step": 7392
    },
    {
      "epoch": 1.59,
      "grad_norm": 7.448834419250488,
      "learning_rate": 6.212998712998713e-06,
      "loss": 0.4159,
      "step": 7393
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.765447616577148,
      "learning_rate": 6.20978120978121e-06,
      "loss": 0.4289,
      "step": 7394
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.587309837341309,
      "learning_rate": 6.2065637065637075e-06,
      "loss": 0.6493,
      "step": 7395
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.705138206481934,
      "learning_rate": 6.203346203346203e-06,
      "loss": 0.3282,
      "step": 7396
    },
    {
      "epoch": 1.59,
      "grad_norm": 7.459446430206299,
      "learning_rate": 6.2001287001287e-06,
      "loss": 0.3905,
      "step": 7397
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.320328712463379,
      "learning_rate": 6.196911196911197e-06,
      "loss": 0.4471,
      "step": 7398
    },
    {
      "epoch": 1.59,
      "grad_norm": 7.423252582550049,
      "learning_rate": 6.193693693693694e-06,
      "loss": 0.3245,
      "step": 7399
    },
    {
      "epoch": 1.59,
      "grad_norm": 13.1840181350708,
      "learning_rate": 6.190476190476191e-06,
      "loss": 0.4769,
      "step": 7400
    },
    {
      "epoch": 1.59,
      "grad_norm": 12.564894676208496,
      "learning_rate": 6.187258687258687e-06,
      "loss": 0.5931,
      "step": 7401
    },
    {
      "epoch": 1.59,
      "grad_norm": 13.731856346130371,
      "learning_rate": 6.184041184041184e-06,
      "loss": 0.7451,
      "step": 7402
    },
    {
      "epoch": 1.59,
      "grad_norm": 13.559938430786133,
      "learning_rate": 6.180823680823682e-06,
      "loss": 0.6988,
      "step": 7403
    },
    {
      "epoch": 1.59,
      "grad_norm": 8.829400062561035,
      "learning_rate": 6.177606177606177e-06,
      "loss": 0.4745,
      "step": 7404
    },
    {
      "epoch": 1.59,
      "grad_norm": 6.857423305511475,
      "learning_rate": 6.1743886743886745e-06,
      "loss": 0.3652,
      "step": 7405
    },
    {
      "epoch": 1.59,
      "grad_norm": 7.792933940887451,
      "learning_rate": 6.171171171171171e-06,
      "loss": 0.3818,
      "step": 7406
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.959026336669922,
      "learning_rate": 6.167953667953668e-06,
      "loss": 0.53,
      "step": 7407
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.239564895629883,
      "learning_rate": 6.164736164736166e-06,
      "loss": 0.413,
      "step": 7408
    },
    {
      "epoch": 1.59,
      "grad_norm": 8.931323051452637,
      "learning_rate": 6.161518661518661e-06,
      "loss": 0.5542,
      "step": 7409
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.477128028869629,
      "learning_rate": 6.1583011583011585e-06,
      "loss": 0.6657,
      "step": 7410
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.489099502563477,
      "learning_rate": 6.155083655083655e-06,
      "loss": 0.5177,
      "step": 7411
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.941244125366211,
      "learning_rate": 6.151866151866152e-06,
      "loss": 0.6407,
      "step": 7412
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.669001579284668,
      "learning_rate": 6.148648648648649e-06,
      "loss": 0.5067,
      "step": 7413
    },
    {
      "epoch": 1.59,
      "grad_norm": 13.394919395446777,
      "learning_rate": 6.145431145431145e-06,
      "loss": 0.5493,
      "step": 7414
    },
    {
      "epoch": 1.59,
      "grad_norm": 7.435996055603027,
      "learning_rate": 6.1422136422136424e-06,
      "loss": 0.527,
      "step": 7415
    },
    {
      "epoch": 1.59,
      "grad_norm": 13.3074951171875,
      "learning_rate": 6.13899613899614e-06,
      "loss": 0.5716,
      "step": 7416
    },
    {
      "epoch": 1.59,
      "grad_norm": 12.326931953430176,
      "learning_rate": 6.135778635778636e-06,
      "loss": 0.769,
      "step": 7417
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.512979507446289,
      "learning_rate": 6.132561132561133e-06,
      "loss": 0.5857,
      "step": 7418
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.762713432312012,
      "learning_rate": 6.129343629343629e-06,
      "loss": 0.5728,
      "step": 7419
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.118667602539062,
      "learning_rate": 6.126126126126126e-06,
      "loss": 0.3259,
      "step": 7420
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.258407592773438,
      "learning_rate": 6.122908622908624e-06,
      "loss": 0.5588,
      "step": 7421
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.851798057556152,
      "learning_rate": 6.119691119691119e-06,
      "loss": 0.6338,
      "step": 7422
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.937736511230469,
      "learning_rate": 6.116473616473617e-06,
      "loss": 0.5246,
      "step": 7423
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.564108848571777,
      "learning_rate": 6.113256113256114e-06,
      "loss": 0.6033,
      "step": 7424
    },
    {
      "epoch": 1.59,
      "grad_norm": 14.615561485290527,
      "learning_rate": 6.11003861003861e-06,
      "loss": 0.6497,
      "step": 7425
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.915349960327148,
      "learning_rate": 6.106821106821107e-06,
      "loss": 0.5497,
      "step": 7426
    },
    {
      "epoch": 1.59,
      "grad_norm": 8.473146438598633,
      "learning_rate": 6.103603603603603e-06,
      "loss": 0.3376,
      "step": 7427
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.030640602111816,
      "learning_rate": 6.1003861003861005e-06,
      "loss": 0.483,
      "step": 7428
    },
    {
      "epoch": 1.59,
      "grad_norm": 8.68409538269043,
      "learning_rate": 6.097168597168598e-06,
      "loss": 0.4239,
      "step": 7429
    },
    {
      "epoch": 1.59,
      "grad_norm": 7.794712066650391,
      "learning_rate": 6.093951093951094e-06,
      "loss": 0.3777,
      "step": 7430
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.683650016784668,
      "learning_rate": 6.090733590733591e-06,
      "loss": 0.5672,
      "step": 7431
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.493599891662598,
      "learning_rate": 6.087516087516088e-06,
      "loss": 0.4334,
      "step": 7432
    },
    {
      "epoch": 1.59,
      "grad_norm": 10.26705265045166,
      "learning_rate": 6.0842985842985845e-06,
      "loss": 0.5263,
      "step": 7433
    },
    {
      "epoch": 1.59,
      "grad_norm": 9.836563110351562,
      "learning_rate": 6.081081081081082e-06,
      "loss": 0.4728,
      "step": 7434
    },
    {
      "epoch": 1.59,
      "grad_norm": 15.572848320007324,
      "learning_rate": 6.077863577863577e-06,
      "loss": 0.4874,
      "step": 7435
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.56423282623291,
      "learning_rate": 6.074646074646075e-06,
      "loss": 0.4277,
      "step": 7436
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.00981330871582,
      "learning_rate": 6.071428571428572e-06,
      "loss": 0.5666,
      "step": 7437
    },
    {
      "epoch": 1.6,
      "grad_norm": 8.099461555480957,
      "learning_rate": 6.0682110682110684e-06,
      "loss": 0.4164,
      "step": 7438
    },
    {
      "epoch": 1.6,
      "grad_norm": 8.45649242401123,
      "learning_rate": 6.064993564993565e-06,
      "loss": 0.4782,
      "step": 7439
    },
    {
      "epoch": 1.6,
      "grad_norm": 14.110757827758789,
      "learning_rate": 6.061776061776061e-06,
      "loss": 0.6577,
      "step": 7440
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.945794105529785,
      "learning_rate": 6.058558558558559e-06,
      "loss": 0.5377,
      "step": 7441
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.109940528869629,
      "learning_rate": 6.055341055341056e-06,
      "loss": 0.3707,
      "step": 7442
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.337101936340332,
      "learning_rate": 6.052123552123552e-06,
      "loss": 0.4575,
      "step": 7443
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.409967422485352,
      "learning_rate": 6.048906048906049e-06,
      "loss": 0.5913,
      "step": 7444
    },
    {
      "epoch": 1.6,
      "grad_norm": 11.722919464111328,
      "learning_rate": 6.045688545688546e-06,
      "loss": 0.6086,
      "step": 7445
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.280647277832031,
      "learning_rate": 6.042471042471043e-06,
      "loss": 0.4721,
      "step": 7446
    },
    {
      "epoch": 1.6,
      "grad_norm": 13.77098274230957,
      "learning_rate": 6.03925353925354e-06,
      "loss": 0.3804,
      "step": 7447
    },
    {
      "epoch": 1.6,
      "grad_norm": 8.65255355834961,
      "learning_rate": 6.0360360360360355e-06,
      "loss": 0.4126,
      "step": 7448
    },
    {
      "epoch": 1.6,
      "grad_norm": 7.810187339782715,
      "learning_rate": 6.032818532818533e-06,
      "loss": 0.604,
      "step": 7449
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.619845390319824,
      "learning_rate": 6.02960102960103e-06,
      "loss": 0.615,
      "step": 7450
    },
    {
      "epoch": 1.6,
      "grad_norm": 12.550820350646973,
      "learning_rate": 6.0263835263835265e-06,
      "loss": 0.4151,
      "step": 7451
    },
    {
      "epoch": 1.6,
      "grad_norm": 7.4556989669799805,
      "learning_rate": 6.023166023166023e-06,
      "loss": 0.4999,
      "step": 7452
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.218299865722656,
      "learning_rate": 6.01994851994852e-06,
      "loss": 0.5638,
      "step": 7453
    },
    {
      "epoch": 1.6,
      "grad_norm": 8.863658905029297,
      "learning_rate": 6.016731016731017e-06,
      "loss": 0.529,
      "step": 7454
    },
    {
      "epoch": 1.6,
      "grad_norm": 11.697098731994629,
      "learning_rate": 6.013513513513514e-06,
      "loss": 0.4174,
      "step": 7455
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.789990425109863,
      "learning_rate": 6.0102960102960105e-06,
      "loss": 0.4866,
      "step": 7456
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.378292083740234,
      "learning_rate": 6.007078507078507e-06,
      "loss": 0.5392,
      "step": 7457
    },
    {
      "epoch": 1.6,
      "grad_norm": 15.517570495605469,
      "learning_rate": 6.003861003861004e-06,
      "loss": 0.5431,
      "step": 7458
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.157588005065918,
      "learning_rate": 6.000643500643501e-06,
      "loss": 0.5159,
      "step": 7459
    },
    {
      "epoch": 1.6,
      "grad_norm": 13.072303771972656,
      "learning_rate": 5.997425997425998e-06,
      "loss": 0.466,
      "step": 7460
    },
    {
      "epoch": 1.6,
      "grad_norm": 12.150947570800781,
      "learning_rate": 5.9942084942084944e-06,
      "loss": 0.489,
      "step": 7461
    },
    {
      "epoch": 1.6,
      "grad_norm": 14.599639892578125,
      "learning_rate": 5.990990990990991e-06,
      "loss": 0.7807,
      "step": 7462
    },
    {
      "epoch": 1.6,
      "grad_norm": 8.072911262512207,
      "learning_rate": 5.987773487773488e-06,
      "loss": 0.575,
      "step": 7463
    },
    {
      "epoch": 1.6,
      "grad_norm": 13.210504531860352,
      "learning_rate": 5.984555984555985e-06,
      "loss": 0.6117,
      "step": 7464
    },
    {
      "epoch": 1.6,
      "grad_norm": 12.371790885925293,
      "learning_rate": 5.981338481338481e-06,
      "loss": 0.534,
      "step": 7465
    },
    {
      "epoch": 1.6,
      "grad_norm": 5.985628128051758,
      "learning_rate": 5.978120978120978e-06,
      "loss": 0.3464,
      "step": 7466
    },
    {
      "epoch": 1.6,
      "grad_norm": 7.800355434417725,
      "learning_rate": 5.974903474903475e-06,
      "loss": 0.3636,
      "step": 7467
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.081582069396973,
      "learning_rate": 5.971685971685972e-06,
      "loss": 0.4955,
      "step": 7468
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.061272621154785,
      "learning_rate": 5.968468468468469e-06,
      "loss": 0.4661,
      "step": 7469
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.833864212036133,
      "learning_rate": 5.965250965250965e-06,
      "loss": 0.6263,
      "step": 7470
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.23360824584961,
      "learning_rate": 5.962033462033462e-06,
      "loss": 0.5553,
      "step": 7471
    },
    {
      "epoch": 1.6,
      "grad_norm": 11.683576583862305,
      "learning_rate": 5.958815958815959e-06,
      "loss": 0.4867,
      "step": 7472
    },
    {
      "epoch": 1.6,
      "grad_norm": 11.678071022033691,
      "learning_rate": 5.955598455598456e-06,
      "loss": 0.4415,
      "step": 7473
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.603312492370605,
      "learning_rate": 5.9523809523809525e-06,
      "loss": 0.3179,
      "step": 7474
    },
    {
      "epoch": 1.6,
      "grad_norm": 8.123296737670898,
      "learning_rate": 5.949163449163449e-06,
      "loss": 0.4371,
      "step": 7475
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.143104553222656,
      "learning_rate": 5.945945945945946e-06,
      "loss": 0.4607,
      "step": 7476
    },
    {
      "epoch": 1.6,
      "grad_norm": 9.771835327148438,
      "learning_rate": 5.942728442728443e-06,
      "loss": 0.4943,
      "step": 7477
    },
    {
      "epoch": 1.6,
      "grad_norm": 11.92110824584961,
      "learning_rate": 5.939510939510939e-06,
      "loss": 0.6475,
      "step": 7478
    },
    {
      "epoch": 1.6,
      "grad_norm": 8.67192554473877,
      "learning_rate": 5.9362934362934365e-06,
      "loss": 0.5376,
      "step": 7479
    },
    {
      "epoch": 1.6,
      "grad_norm": 13.387099266052246,
      "learning_rate": 5.933075933075933e-06,
      "loss": 0.6666,
      "step": 7480
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.291061401367188,
      "learning_rate": 5.92985842985843e-06,
      "loss": 0.4807,
      "step": 7481
    },
    {
      "epoch": 1.6,
      "grad_norm": 15.047369956970215,
      "learning_rate": 5.9266409266409275e-06,
      "loss": 0.8157,
      "step": 7482
    },
    {
      "epoch": 1.61,
      "grad_norm": 9.281285285949707,
      "learning_rate": 5.923423423423423e-06,
      "loss": 0.4646,
      "step": 7483
    },
    {
      "epoch": 1.61,
      "grad_norm": 14.86556625366211,
      "learning_rate": 5.9202059202059204e-06,
      "loss": 0.711,
      "step": 7484
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.859213829040527,
      "learning_rate": 5.916988416988417e-06,
      "loss": 0.566,
      "step": 7485
    },
    {
      "epoch": 1.61,
      "grad_norm": 12.34170913696289,
      "learning_rate": 5.913770913770914e-06,
      "loss": 0.6484,
      "step": 7486
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.395462989807129,
      "learning_rate": 5.910553410553411e-06,
      "loss": 0.5588,
      "step": 7487
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.402356147766113,
      "learning_rate": 5.907335907335907e-06,
      "loss": 0.378,
      "step": 7488
    },
    {
      "epoch": 1.61,
      "grad_norm": 12.840353012084961,
      "learning_rate": 5.904118404118404e-06,
      "loss": 0.6772,
      "step": 7489
    },
    {
      "epoch": 1.61,
      "grad_norm": 10.869420051574707,
      "learning_rate": 5.900900900900902e-06,
      "loss": 0.4362,
      "step": 7490
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.475628852844238,
      "learning_rate": 5.897683397683397e-06,
      "loss": 0.5811,
      "step": 7491
    },
    {
      "epoch": 1.61,
      "grad_norm": 10.867398262023926,
      "learning_rate": 5.894465894465895e-06,
      "loss": 0.6476,
      "step": 7492
    },
    {
      "epoch": 1.61,
      "grad_norm": 6.339871406555176,
      "learning_rate": 5.891248391248391e-06,
      "loss": 0.3369,
      "step": 7493
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.493051528930664,
      "learning_rate": 5.888030888030888e-06,
      "loss": 0.6546,
      "step": 7494
    },
    {
      "epoch": 1.61,
      "grad_norm": 14.184857368469238,
      "learning_rate": 5.884813384813386e-06,
      "loss": 0.7449,
      "step": 7495
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.508328437805176,
      "learning_rate": 5.881595881595881e-06,
      "loss": 0.6463,
      "step": 7496
    },
    {
      "epoch": 1.61,
      "grad_norm": 8.001158714294434,
      "learning_rate": 5.8783783783783786e-06,
      "loss": 0.5577,
      "step": 7497
    },
    {
      "epoch": 1.61,
      "grad_norm": 10.93775463104248,
      "learning_rate": 5.875160875160875e-06,
      "loss": 0.5898,
      "step": 7498
    },
    {
      "epoch": 1.61,
      "grad_norm": 10.602176666259766,
      "learning_rate": 5.871943371943372e-06,
      "loss": 0.7066,
      "step": 7499
    },
    {
      "epoch": 1.61,
      "grad_norm": 9.580327987670898,
      "learning_rate": 5.868725868725869e-06,
      "loss": 0.5184,
      "step": 7500
    },
    {
      "epoch": 1.61,
      "grad_norm": 9.384793281555176,
      "learning_rate": 5.865508365508365e-06,
      "loss": 0.4851,
      "step": 7501
    },
    {
      "epoch": 1.61,
      "grad_norm": 12.241350173950195,
      "learning_rate": 5.8622908622908625e-06,
      "loss": 0.4516,
      "step": 7502
    },
    {
      "epoch": 1.61,
      "grad_norm": 6.593988418579102,
      "learning_rate": 5.85907335907336e-06,
      "loss": 0.3068,
      "step": 7503
    },
    {
      "epoch": 1.61,
      "grad_norm": 9.654346466064453,
      "learning_rate": 5.855855855855855e-06,
      "loss": 0.4086,
      "step": 7504
    },
    {
      "epoch": 1.61,
      "grad_norm": 10.747211456298828,
      "learning_rate": 5.852638352638353e-06,
      "loss": 0.526,
      "step": 7505
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.87922191619873,
      "learning_rate": 5.849420849420849e-06,
      "loss": 0.6678,
      "step": 7506
    },
    {
      "epoch": 1.61,
      "grad_norm": 14.893392562866211,
      "learning_rate": 5.8462033462033465e-06,
      "loss": 0.8743,
      "step": 7507
    },
    {
      "epoch": 1.61,
      "grad_norm": 7.733294486999512,
      "learning_rate": 5.842985842985844e-06,
      "loss": 0.4169,
      "step": 7508
    },
    {
      "epoch": 1.61,
      "grad_norm": 10.895306587219238,
      "learning_rate": 5.839768339768339e-06,
      "loss": 0.5845,
      "step": 7509
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.37563419342041,
      "learning_rate": 5.836550836550837e-06,
      "loss": 0.5675,
      "step": 7510
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.78275203704834,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.6202,
      "step": 7511
    },
    {
      "epoch": 1.61,
      "grad_norm": 10.28819751739502,
      "learning_rate": 5.83011583011583e-06,
      "loss": 0.391,
      "step": 7512
    },
    {
      "epoch": 1.61,
      "grad_norm": 14.595450401306152,
      "learning_rate": 5.826898326898327e-06,
      "loss": 0.7452,
      "step": 7513
    },
    {
      "epoch": 1.61,
      "grad_norm": 7.706846714019775,
      "learning_rate": 5.823680823680823e-06,
      "loss": 0.356,
      "step": 7514
    },
    {
      "epoch": 1.61,
      "grad_norm": 8.46711254119873,
      "learning_rate": 5.820463320463321e-06,
      "loss": 0.4431,
      "step": 7515
    },
    {
      "epoch": 1.61,
      "grad_norm": 8.08811092376709,
      "learning_rate": 5.817245817245818e-06,
      "loss": 0.413,
      "step": 7516
    },
    {
      "epoch": 1.61,
      "grad_norm": 12.064764022827148,
      "learning_rate": 5.814028314028314e-06,
      "loss": 0.6639,
      "step": 7517
    },
    {
      "epoch": 1.61,
      "grad_norm": 9.76872730255127,
      "learning_rate": 5.810810810810811e-06,
      "loss": 0.4378,
      "step": 7518
    },
    {
      "epoch": 1.61,
      "grad_norm": 11.584485054016113,
      "learning_rate": 5.807593307593308e-06,
      "loss": 0.5758,
      "step": 7519
    },
    {
      "epoch": 1.61,
      "grad_norm": 9.781500816345215,
      "learning_rate": 5.8043758043758046e-06,
      "loss": 0.4827,
      "step": 7520
    },
    {
      "epoch": 1.61,
      "grad_norm": 10.96768856048584,
      "learning_rate": 5.801158301158302e-06,
      "loss": 0.5643,
      "step": 7521
    },
    {
      "epoch": 1.61,
      "grad_norm": 9.52723217010498,
      "learning_rate": 5.7979407979407975e-06,
      "loss": 0.4501,
      "step": 7522
    },
    {
      "epoch": 1.61,
      "grad_norm": 8.55528736114502,
      "learning_rate": 5.794723294723295e-06,
      "loss": 0.3982,
      "step": 7523
    },
    {
      "epoch": 1.61,
      "grad_norm": 10.55843448638916,
      "learning_rate": 5.791505791505792e-06,
      "loss": 0.5984,
      "step": 7524
    },
    {
      "epoch": 1.61,
      "grad_norm": 14.208714485168457,
      "learning_rate": 5.7882882882882885e-06,
      "loss": 0.5793,
      "step": 7525
    },
    {
      "epoch": 1.61,
      "grad_norm": 12.7923583984375,
      "learning_rate": 5.785070785070785e-06,
      "loss": 0.6654,
      "step": 7526
    },
    {
      "epoch": 1.61,
      "grad_norm": 13.019758224487305,
      "learning_rate": 5.781853281853281e-06,
      "loss": 0.5688,
      "step": 7527
    },
    {
      "epoch": 1.61,
      "grad_norm": 7.5465850830078125,
      "learning_rate": 5.778635778635779e-06,
      "loss": 0.3898,
      "step": 7528
    },
    {
      "epoch": 1.61,
      "grad_norm": 10.054975509643555,
      "learning_rate": 5.775418275418276e-06,
      "loss": 0.5082,
      "step": 7529
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.472352981567383,
      "learning_rate": 5.7722007722007725e-06,
      "loss": 0.466,
      "step": 7530
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.45793628692627,
      "learning_rate": 5.768983268983269e-06,
      "loss": 0.7032,
      "step": 7531
    },
    {
      "epoch": 1.62,
      "grad_norm": 12.540285110473633,
      "learning_rate": 5.765765765765766e-06,
      "loss": 0.3546,
      "step": 7532
    },
    {
      "epoch": 1.62,
      "grad_norm": 13.033102989196777,
      "learning_rate": 5.762548262548263e-06,
      "loss": 0.5545,
      "step": 7533
    },
    {
      "epoch": 1.62,
      "grad_norm": 7.4204559326171875,
      "learning_rate": 5.75933075933076e-06,
      "loss": 0.4135,
      "step": 7534
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.419729232788086,
      "learning_rate": 5.7561132561132556e-06,
      "loss": 0.5397,
      "step": 7535
    },
    {
      "epoch": 1.62,
      "grad_norm": 8.861800193786621,
      "learning_rate": 5.752895752895753e-06,
      "loss": 0.4413,
      "step": 7536
    },
    {
      "epoch": 1.62,
      "grad_norm": 14.170783042907715,
      "learning_rate": 5.74967824967825e-06,
      "loss": 0.7226,
      "step": 7537
    },
    {
      "epoch": 1.62,
      "grad_norm": 11.50087833404541,
      "learning_rate": 5.746460746460747e-06,
      "loss": 0.4803,
      "step": 7538
    },
    {
      "epoch": 1.62,
      "grad_norm": 12.471067428588867,
      "learning_rate": 5.743243243243243e-06,
      "loss": 0.4945,
      "step": 7539
    },
    {
      "epoch": 1.62,
      "grad_norm": 12.553688049316406,
      "learning_rate": 5.74002574002574e-06,
      "loss": 0.6056,
      "step": 7540
    },
    {
      "epoch": 1.62,
      "grad_norm": 13.070096969604492,
      "learning_rate": 5.736808236808237e-06,
      "loss": 0.5389,
      "step": 7541
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.688863754272461,
      "learning_rate": 5.733590733590734e-06,
      "loss": 0.5269,
      "step": 7542
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.479377746582031,
      "learning_rate": 5.7303732303732306e-06,
      "loss": 0.439,
      "step": 7543
    },
    {
      "epoch": 1.62,
      "grad_norm": 7.076297283172607,
      "learning_rate": 5.727155727155727e-06,
      "loss": 0.5342,
      "step": 7544
    },
    {
      "epoch": 1.62,
      "grad_norm": 12.241185188293457,
      "learning_rate": 5.723938223938224e-06,
      "loss": 0.8039,
      "step": 7545
    },
    {
      "epoch": 1.62,
      "grad_norm": 11.518341064453125,
      "learning_rate": 5.720720720720721e-06,
      "loss": 0.6565,
      "step": 7546
    },
    {
      "epoch": 1.62,
      "grad_norm": 13.990534782409668,
      "learning_rate": 5.717503217503218e-06,
      "loss": 0.8256,
      "step": 7547
    },
    {
      "epoch": 1.62,
      "grad_norm": 11.215410232543945,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 0.5587,
      "step": 7548
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.857245445251465,
      "learning_rate": 5.711068211068211e-06,
      "loss": 0.3313,
      "step": 7549
    },
    {
      "epoch": 1.62,
      "grad_norm": 13.490482330322266,
      "learning_rate": 5.707850707850708e-06,
      "loss": 0.724,
      "step": 7550
    },
    {
      "epoch": 1.62,
      "grad_norm": 13.277083396911621,
      "learning_rate": 5.704633204633205e-06,
      "loss": 0.5842,
      "step": 7551
    },
    {
      "epoch": 1.62,
      "grad_norm": 12.327010154724121,
      "learning_rate": 5.701415701415701e-06,
      "loss": 0.4079,
      "step": 7552
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.476134300231934,
      "learning_rate": 5.6981981981981985e-06,
      "loss": 0.5069,
      "step": 7553
    },
    {
      "epoch": 1.62,
      "grad_norm": 13.042057037353516,
      "learning_rate": 5.694980694980695e-06,
      "loss": 0.4934,
      "step": 7554
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.317533493041992,
      "learning_rate": 5.691763191763192e-06,
      "loss": 0.5207,
      "step": 7555
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.666854858398438,
      "learning_rate": 5.688545688545689e-06,
      "loss": 0.5685,
      "step": 7556
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.168272972106934,
      "learning_rate": 5.685328185328185e-06,
      "loss": 0.4106,
      "step": 7557
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.378591537475586,
      "learning_rate": 5.682110682110682e-06,
      "loss": 0.4909,
      "step": 7558
    },
    {
      "epoch": 1.62,
      "grad_norm": 11.129934310913086,
      "learning_rate": 5.678893178893179e-06,
      "loss": 0.5057,
      "step": 7559
    },
    {
      "epoch": 1.62,
      "grad_norm": 8.29308795928955,
      "learning_rate": 5.675675675675676e-06,
      "loss": 0.4637,
      "step": 7560
    },
    {
      "epoch": 1.62,
      "grad_norm": 15.873396873474121,
      "learning_rate": 5.672458172458173e-06,
      "loss": 0.572,
      "step": 7561
    },
    {
      "epoch": 1.62,
      "grad_norm": 12.625581741333008,
      "learning_rate": 5.669240669240669e-06,
      "loss": 0.505,
      "step": 7562
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.748132705688477,
      "learning_rate": 5.666023166023166e-06,
      "loss": 0.5063,
      "step": 7563
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.206901550292969,
      "learning_rate": 5.662805662805663e-06,
      "loss": 0.5128,
      "step": 7564
    },
    {
      "epoch": 1.62,
      "grad_norm": 11.388949394226074,
      "learning_rate": 5.659588159588159e-06,
      "loss": 0.5379,
      "step": 7565
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.788782119750977,
      "learning_rate": 5.6563706563706566e-06,
      "loss": 0.4437,
      "step": 7566
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.769548416137695,
      "learning_rate": 5.653153153153153e-06,
      "loss": 0.3465,
      "step": 7567
    },
    {
      "epoch": 1.62,
      "grad_norm": 7.18800163269043,
      "learning_rate": 5.64993564993565e-06,
      "loss": 0.4098,
      "step": 7568
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.274568557739258,
      "learning_rate": 5.646718146718148e-06,
      "loss": 0.6044,
      "step": 7569
    },
    {
      "epoch": 1.62,
      "grad_norm": 10.811291694641113,
      "learning_rate": 5.643500643500643e-06,
      "loss": 0.4682,
      "step": 7570
    },
    {
      "epoch": 1.62,
      "grad_norm": 12.99849796295166,
      "learning_rate": 5.6402831402831405e-06,
      "loss": 0.6131,
      "step": 7571
    },
    {
      "epoch": 1.62,
      "grad_norm": 7.474158763885498,
      "learning_rate": 5.637065637065637e-06,
      "loss": 0.5615,
      "step": 7572
    },
    {
      "epoch": 1.62,
      "grad_norm": 9.103140830993652,
      "learning_rate": 5.633848133848134e-06,
      "loss": 0.5036,
      "step": 7573
    },
    {
      "epoch": 1.62,
      "grad_norm": 14.123886108398438,
      "learning_rate": 5.630630630630631e-06,
      "loss": 0.5693,
      "step": 7574
    },
    {
      "epoch": 1.62,
      "grad_norm": 11.27999210357666,
      "learning_rate": 5.627413127413127e-06,
      "loss": 0.5862,
      "step": 7575
    },
    {
      "epoch": 1.63,
      "grad_norm": 9.489529609680176,
      "learning_rate": 5.6241956241956245e-06,
      "loss": 0.4495,
      "step": 7576
    },
    {
      "epoch": 1.63,
      "grad_norm": 8.17496395111084,
      "learning_rate": 5.620978120978122e-06,
      "loss": 0.4986,
      "step": 7577
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.661971092224121,
      "learning_rate": 5.617760617760617e-06,
      "loss": 0.5578,
      "step": 7578
    },
    {
      "epoch": 1.63,
      "grad_norm": 12.683906555175781,
      "learning_rate": 5.614543114543115e-06,
      "loss": 0.7792,
      "step": 7579
    },
    {
      "epoch": 1.63,
      "grad_norm": 13.454289436340332,
      "learning_rate": 5.611325611325611e-06,
      "loss": 0.6431,
      "step": 7580
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.212675094604492,
      "learning_rate": 5.608108108108108e-06,
      "loss": 0.6204,
      "step": 7581
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.36351203918457,
      "learning_rate": 5.604890604890606e-06,
      "loss": 0.6317,
      "step": 7582
    },
    {
      "epoch": 1.63,
      "grad_norm": 8.874673843383789,
      "learning_rate": 5.601673101673101e-06,
      "loss": 0.4765,
      "step": 7583
    },
    {
      "epoch": 1.63,
      "grad_norm": 11.827559471130371,
      "learning_rate": 5.598455598455599e-06,
      "loss": 0.4767,
      "step": 7584
    },
    {
      "epoch": 1.63,
      "grad_norm": 11.941055297851562,
      "learning_rate": 5.595238095238095e-06,
      "loss": 0.6544,
      "step": 7585
    },
    {
      "epoch": 1.63,
      "grad_norm": 9.843469619750977,
      "learning_rate": 5.592020592020592e-06,
      "loss": 0.5879,
      "step": 7586
    },
    {
      "epoch": 1.63,
      "grad_norm": 12.66722583770752,
      "learning_rate": 5.588803088803089e-06,
      "loss": 0.55,
      "step": 7587
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.641082763671875,
      "learning_rate": 5.585585585585585e-06,
      "loss": 0.5613,
      "step": 7588
    },
    {
      "epoch": 1.63,
      "grad_norm": 13.48292350769043,
      "learning_rate": 5.5823680823680826e-06,
      "loss": 0.7999,
      "step": 7589
    },
    {
      "epoch": 1.63,
      "grad_norm": 8.781396865844727,
      "learning_rate": 5.57915057915058e-06,
      "loss": 0.5094,
      "step": 7590
    },
    {
      "epoch": 1.63,
      "grad_norm": 12.4613037109375,
      "learning_rate": 5.5759330759330755e-06,
      "loss": 0.556,
      "step": 7591
    },
    {
      "epoch": 1.63,
      "grad_norm": 14.623103141784668,
      "learning_rate": 5.572715572715573e-06,
      "loss": 0.6053,
      "step": 7592
    },
    {
      "epoch": 1.63,
      "grad_norm": 12.396018981933594,
      "learning_rate": 5.569498069498069e-06,
      "loss": 0.8016,
      "step": 7593
    },
    {
      "epoch": 1.63,
      "grad_norm": 11.595718383789062,
      "learning_rate": 5.5662805662805665e-06,
      "loss": 0.3245,
      "step": 7594
    },
    {
      "epoch": 1.63,
      "grad_norm": 9.723567962646484,
      "learning_rate": 5.563063063063064e-06,
      "loss": 0.5002,
      "step": 7595
    },
    {
      "epoch": 1.63,
      "grad_norm": 13.323812484741211,
      "learning_rate": 5.5598455598455594e-06,
      "loss": 0.6968,
      "step": 7596
    },
    {
      "epoch": 1.63,
      "grad_norm": 11.113154411315918,
      "learning_rate": 5.556628056628057e-06,
      "loss": 0.5004,
      "step": 7597
    },
    {
      "epoch": 1.63,
      "grad_norm": 8.954119682312012,
      "learning_rate": 5.553410553410554e-06,
      "loss": 0.4765,
      "step": 7598
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.086511611938477,
      "learning_rate": 5.5501930501930505e-06,
      "loss": 0.4417,
      "step": 7599
    },
    {
      "epoch": 1.63,
      "grad_norm": 11.098613739013672,
      "learning_rate": 5.546975546975547e-06,
      "loss": 0.5127,
      "step": 7600
    },
    {
      "epoch": 1.63,
      "grad_norm": 9.333904266357422,
      "learning_rate": 5.543758043758043e-06,
      "loss": 0.5146,
      "step": 7601
    },
    {
      "epoch": 1.63,
      "grad_norm": 8.47717571258545,
      "learning_rate": 5.540540540540541e-06,
      "loss": 0.4621,
      "step": 7602
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.759828567504883,
      "learning_rate": 5.537323037323038e-06,
      "loss": 0.5917,
      "step": 7603
    },
    {
      "epoch": 1.63,
      "grad_norm": 12.130353927612305,
      "learning_rate": 5.534105534105534e-06,
      "loss": 0.6129,
      "step": 7604
    },
    {
      "epoch": 1.63,
      "grad_norm": 9.006576538085938,
      "learning_rate": 5.530888030888031e-06,
      "loss": 0.3589,
      "step": 7605
    },
    {
      "epoch": 1.63,
      "grad_norm": 11.844138145446777,
      "learning_rate": 5.527670527670528e-06,
      "loss": 0.7329,
      "step": 7606
    },
    {
      "epoch": 1.63,
      "grad_norm": 8.119832992553711,
      "learning_rate": 5.524453024453025e-06,
      "loss": 0.3899,
      "step": 7607
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.797677993774414,
      "learning_rate": 5.521235521235522e-06,
      "loss": 0.5063,
      "step": 7608
    },
    {
      "epoch": 1.63,
      "grad_norm": 12.49677848815918,
      "learning_rate": 5.5180180180180175e-06,
      "loss": 0.4899,
      "step": 7609
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.346790313720703,
      "learning_rate": 5.514800514800515e-06,
      "loss": 0.6413,
      "step": 7610
    },
    {
      "epoch": 1.63,
      "grad_norm": 9.233711242675781,
      "learning_rate": 5.511583011583012e-06,
      "loss": 0.633,
      "step": 7611
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.127501487731934,
      "learning_rate": 5.5083655083655086e-06,
      "loss": 0.4493,
      "step": 7612
    },
    {
      "epoch": 1.63,
      "grad_norm": 9.857946395874023,
      "learning_rate": 5.505148005148005e-06,
      "loss": 0.5569,
      "step": 7613
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.932930946350098,
      "learning_rate": 5.5019305019305015e-06,
      "loss": 0.5297,
      "step": 7614
    },
    {
      "epoch": 1.63,
      "grad_norm": 14.557961463928223,
      "learning_rate": 5.498712998712999e-06,
      "loss": 0.6683,
      "step": 7615
    },
    {
      "epoch": 1.63,
      "grad_norm": 13.941275596618652,
      "learning_rate": 5.495495495495496e-06,
      "loss": 0.5074,
      "step": 7616
    },
    {
      "epoch": 1.63,
      "grad_norm": 8.12596321105957,
      "learning_rate": 5.4922779922779925e-06,
      "loss": 0.4645,
      "step": 7617
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.699536323547363,
      "learning_rate": 5.489060489060489e-06,
      "loss": 0.4186,
      "step": 7618
    },
    {
      "epoch": 1.63,
      "grad_norm": 9.039584159851074,
      "learning_rate": 5.485842985842986e-06,
      "loss": 0.4355,
      "step": 7619
    },
    {
      "epoch": 1.63,
      "grad_norm": 11.777717590332031,
      "learning_rate": 5.482625482625483e-06,
      "loss": 0.7095,
      "step": 7620
    },
    {
      "epoch": 1.63,
      "grad_norm": 10.142265319824219,
      "learning_rate": 5.47940797940798e-06,
      "loss": 0.4614,
      "step": 7621
    },
    {
      "epoch": 1.63,
      "grad_norm": 11.101910591125488,
      "learning_rate": 5.476190476190476e-06,
      "loss": 0.5961,
      "step": 7622
    },
    {
      "epoch": 1.64,
      "grad_norm": 8.442414283752441,
      "learning_rate": 5.472972972972973e-06,
      "loss": 0.4962,
      "step": 7623
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.420427322387695,
      "learning_rate": 5.46975546975547e-06,
      "loss": 0.6416,
      "step": 7624
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.591654777526855,
      "learning_rate": 5.466537966537967e-06,
      "loss": 0.4502,
      "step": 7625
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.627406120300293,
      "learning_rate": 5.463320463320463e-06,
      "loss": 0.422,
      "step": 7626
    },
    {
      "epoch": 1.64,
      "grad_norm": 13.62757396697998,
      "learning_rate": 5.4601029601029604e-06,
      "loss": 0.7744,
      "step": 7627
    },
    {
      "epoch": 1.64,
      "grad_norm": 7.587872505187988,
      "learning_rate": 5.456885456885457e-06,
      "loss": 0.4354,
      "step": 7628
    },
    {
      "epoch": 1.64,
      "grad_norm": 12.879377365112305,
      "learning_rate": 5.453667953667954e-06,
      "loss": 0.6863,
      "step": 7629
    },
    {
      "epoch": 1.64,
      "grad_norm": 12.885239601135254,
      "learning_rate": 5.450450450450451e-06,
      "loss": 0.667,
      "step": 7630
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.981551170349121,
      "learning_rate": 5.447232947232947e-06,
      "loss": 0.4481,
      "step": 7631
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.114948272705078,
      "learning_rate": 5.444015444015444e-06,
      "loss": 0.519,
      "step": 7632
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.525074005126953,
      "learning_rate": 5.440797940797941e-06,
      "loss": 0.6118,
      "step": 7633
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.783196449279785,
      "learning_rate": 5.437580437580438e-06,
      "loss": 0.4854,
      "step": 7634
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.530328750610352,
      "learning_rate": 5.434362934362935e-06,
      "loss": 0.7706,
      "step": 7635
    },
    {
      "epoch": 1.64,
      "grad_norm": 13.095335006713867,
      "learning_rate": 5.431145431145431e-06,
      "loss": 0.4137,
      "step": 7636
    },
    {
      "epoch": 1.64,
      "grad_norm": 8.301831245422363,
      "learning_rate": 5.427927927927928e-06,
      "loss": 0.3752,
      "step": 7637
    },
    {
      "epoch": 1.64,
      "grad_norm": 12.09936809539795,
      "learning_rate": 5.424710424710425e-06,
      "loss": 0.4076,
      "step": 7638
    },
    {
      "epoch": 1.64,
      "grad_norm": 15.610919952392578,
      "learning_rate": 5.421492921492921e-06,
      "loss": 0.4539,
      "step": 7639
    },
    {
      "epoch": 1.64,
      "grad_norm": 12.123153686523438,
      "learning_rate": 5.4182754182754185e-06,
      "loss": 0.4616,
      "step": 7640
    },
    {
      "epoch": 1.64,
      "grad_norm": 12.641314506530762,
      "learning_rate": 5.415057915057915e-06,
      "loss": 0.4595,
      "step": 7641
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.182628631591797,
      "learning_rate": 5.411840411840412e-06,
      "loss": 0.5796,
      "step": 7642
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.557555198669434,
      "learning_rate": 5.408622908622909e-06,
      "loss": 0.6233,
      "step": 7643
    },
    {
      "epoch": 1.64,
      "grad_norm": 9.914264678955078,
      "learning_rate": 5.405405405405405e-06,
      "loss": 0.6336,
      "step": 7644
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.175463676452637,
      "learning_rate": 5.4021879021879025e-06,
      "loss": 0.5205,
      "step": 7645
    },
    {
      "epoch": 1.64,
      "grad_norm": 9.41447925567627,
      "learning_rate": 5.398970398970399e-06,
      "loss": 0.5071,
      "step": 7646
    },
    {
      "epoch": 1.64,
      "grad_norm": 9.502260208129883,
      "learning_rate": 5.395752895752896e-06,
      "loss": 0.3906,
      "step": 7647
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.547467231750488,
      "learning_rate": 5.392535392535393e-06,
      "loss": 0.4659,
      "step": 7648
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.644227981567383,
      "learning_rate": 5.389317889317889e-06,
      "loss": 0.5754,
      "step": 7649
    },
    {
      "epoch": 1.64,
      "grad_norm": 13.215782165527344,
      "learning_rate": 5.3861003861003864e-06,
      "loss": 0.5519,
      "step": 7650
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.85572624206543,
      "learning_rate": 5.382882882882883e-06,
      "loss": 0.5468,
      "step": 7651
    },
    {
      "epoch": 1.64,
      "grad_norm": 9.61147403717041,
      "learning_rate": 5.379665379665379e-06,
      "loss": 0.5795,
      "step": 7652
    },
    {
      "epoch": 1.64,
      "grad_norm": 13.758307456970215,
      "learning_rate": 5.376447876447877e-06,
      "loss": 0.7153,
      "step": 7653
    },
    {
      "epoch": 1.64,
      "grad_norm": 12.10712718963623,
      "learning_rate": 5.373230373230373e-06,
      "loss": 0.5754,
      "step": 7654
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.648677825927734,
      "learning_rate": 5.37001287001287e-06,
      "loss": 0.5703,
      "step": 7655
    },
    {
      "epoch": 1.64,
      "grad_norm": 13.947694778442383,
      "learning_rate": 5.366795366795368e-06,
      "loss": 0.6519,
      "step": 7656
    },
    {
      "epoch": 1.64,
      "grad_norm": 12.11249828338623,
      "learning_rate": 5.363577863577863e-06,
      "loss": 0.5904,
      "step": 7657
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.010812759399414,
      "learning_rate": 5.360360360360361e-06,
      "loss": 0.4502,
      "step": 7658
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.203165054321289,
      "learning_rate": 5.357142857142857e-06,
      "loss": 0.4436,
      "step": 7659
    },
    {
      "epoch": 1.64,
      "grad_norm": 12.11374568939209,
      "learning_rate": 5.353925353925354e-06,
      "loss": 0.5669,
      "step": 7660
    },
    {
      "epoch": 1.64,
      "grad_norm": 12.726213455200195,
      "learning_rate": 5.350707850707851e-06,
      "loss": 0.5132,
      "step": 7661
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.45439338684082,
      "learning_rate": 5.347490347490347e-06,
      "loss": 0.6118,
      "step": 7662
    },
    {
      "epoch": 1.64,
      "grad_norm": 8.68755054473877,
      "learning_rate": 5.3442728442728445e-06,
      "loss": 0.6127,
      "step": 7663
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.052059173583984,
      "learning_rate": 5.341055341055342e-06,
      "loss": 0.4496,
      "step": 7664
    },
    {
      "epoch": 1.64,
      "grad_norm": 11.230579376220703,
      "learning_rate": 5.3378378378378374e-06,
      "loss": 0.6082,
      "step": 7665
    },
    {
      "epoch": 1.64,
      "grad_norm": 9.314970970153809,
      "learning_rate": 5.334620334620335e-06,
      "loss": 0.6959,
      "step": 7666
    },
    {
      "epoch": 1.64,
      "grad_norm": 8.767603874206543,
      "learning_rate": 5.331402831402831e-06,
      "loss": 0.5385,
      "step": 7667
    },
    {
      "epoch": 1.64,
      "grad_norm": 10.71431827545166,
      "learning_rate": 5.3281853281853285e-06,
      "loss": 0.5647,
      "step": 7668
    },
    {
      "epoch": 1.65,
      "grad_norm": 8.895282745361328,
      "learning_rate": 5.324967824967826e-06,
      "loss": 0.4673,
      "step": 7669
    },
    {
      "epoch": 1.65,
      "grad_norm": 8.895647048950195,
      "learning_rate": 5.321750321750321e-06,
      "loss": 0.2992,
      "step": 7670
    },
    {
      "epoch": 1.65,
      "grad_norm": 8.027562141418457,
      "learning_rate": 5.318532818532819e-06,
      "loss": 0.4518,
      "step": 7671
    },
    {
      "epoch": 1.65,
      "grad_norm": 8.248534202575684,
      "learning_rate": 5.315315315315315e-06,
      "loss": 0.3731,
      "step": 7672
    },
    {
      "epoch": 1.65,
      "grad_norm": 11.435273170471191,
      "learning_rate": 5.3120978120978124e-06,
      "loss": 0.5291,
      "step": 7673
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.735199928283691,
      "learning_rate": 5.308880308880309e-06,
      "loss": 0.5546,
      "step": 7674
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.340341567993164,
      "learning_rate": 5.305662805662805e-06,
      "loss": 0.5072,
      "step": 7675
    },
    {
      "epoch": 1.65,
      "grad_norm": 13.640751838684082,
      "learning_rate": 5.302445302445303e-06,
      "loss": 0.7588,
      "step": 7676
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.905435562133789,
      "learning_rate": 5.2992277992278e-06,
      "loss": 0.6596,
      "step": 7677
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.083625793457031,
      "learning_rate": 5.2960102960102955e-06,
      "loss": 0.6018,
      "step": 7678
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.381341934204102,
      "learning_rate": 5.292792792792793e-06,
      "loss": 0.4386,
      "step": 7679
    },
    {
      "epoch": 1.65,
      "grad_norm": 11.449508666992188,
      "learning_rate": 5.289575289575289e-06,
      "loss": 0.748,
      "step": 7680
    },
    {
      "epoch": 1.65,
      "grad_norm": 8.28355598449707,
      "learning_rate": 5.286357786357787e-06,
      "loss": 0.4067,
      "step": 7681
    },
    {
      "epoch": 1.65,
      "grad_norm": 7.194711685180664,
      "learning_rate": 5.283140283140284e-06,
      "loss": 0.4244,
      "step": 7682
    },
    {
      "epoch": 1.65,
      "grad_norm": 11.95019817352295,
      "learning_rate": 5.2799227799227795e-06,
      "loss": 0.5394,
      "step": 7683
    },
    {
      "epoch": 1.65,
      "grad_norm": 14.591204643249512,
      "learning_rate": 5.276705276705277e-06,
      "loss": 0.5675,
      "step": 7684
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.12584114074707,
      "learning_rate": 5.273487773487774e-06,
      "loss": 0.4611,
      "step": 7685
    },
    {
      "epoch": 1.65,
      "grad_norm": 11.153300285339355,
      "learning_rate": 5.2702702702702705e-06,
      "loss": 0.5611,
      "step": 7686
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.552461624145508,
      "learning_rate": 5.267052767052767e-06,
      "loss": 0.6063,
      "step": 7687
    },
    {
      "epoch": 1.65,
      "grad_norm": 11.077914237976074,
      "learning_rate": 5.2638352638352634e-06,
      "loss": 0.5749,
      "step": 7688
    },
    {
      "epoch": 1.65,
      "grad_norm": 14.582127571105957,
      "learning_rate": 5.260617760617761e-06,
      "loss": 0.8093,
      "step": 7689
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.28226375579834,
      "learning_rate": 5.257400257400258e-06,
      "loss": 0.4958,
      "step": 7690
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.696765899658203,
      "learning_rate": 5.254182754182754e-06,
      "loss": 0.5981,
      "step": 7691
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.153651237487793,
      "learning_rate": 5.250965250965251e-06,
      "loss": 0.4593,
      "step": 7692
    },
    {
      "epoch": 1.65,
      "grad_norm": 14.616866111755371,
      "learning_rate": 5.247747747747748e-06,
      "loss": 0.782,
      "step": 7693
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.438899040222168,
      "learning_rate": 5.244530244530245e-06,
      "loss": 0.4201,
      "step": 7694
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.077936172485352,
      "learning_rate": 5.241312741312742e-06,
      "loss": 0.6004,
      "step": 7695
    },
    {
      "epoch": 1.65,
      "grad_norm": 8.566909790039062,
      "learning_rate": 5.238095238095238e-06,
      "loss": 0.3747,
      "step": 7696
    },
    {
      "epoch": 1.65,
      "grad_norm": 14.725369453430176,
      "learning_rate": 5.234877734877735e-06,
      "loss": 0.6347,
      "step": 7697
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.814542770385742,
      "learning_rate": 5.231660231660232e-06,
      "loss": 0.3837,
      "step": 7698
    },
    {
      "epoch": 1.65,
      "grad_norm": 8.928094863891602,
      "learning_rate": 5.228442728442729e-06,
      "loss": 0.3867,
      "step": 7699
    },
    {
      "epoch": 1.65,
      "grad_norm": 7.965097427368164,
      "learning_rate": 5.225225225225225e-06,
      "loss": 0.3625,
      "step": 7700
    },
    {
      "epoch": 1.65,
      "grad_norm": 10.15672492980957,
      "learning_rate": 5.222007722007722e-06,
      "loss": 0.4758,
      "step": 7701
    },
    {
      "epoch": 1.65,
      "grad_norm": 12.06130599975586,
      "learning_rate": 5.218790218790219e-06,
      "loss": 0.4843,
      "step": 7702
    },
    {
      "epoch": 1.65,
      "grad_norm": 15.185957908630371,
      "learning_rate": 5.215572715572716e-06,
      "loss": 0.6212,
      "step": 7703
    },
    {
      "epoch": 1.65,
      "grad_norm": 7.67489767074585,
      "learning_rate": 5.212355212355213e-06,
      "loss": 0.4326,
      "step": 7704
    },
    {
      "epoch": 1.65,
      "grad_norm": 12.618034362792969,
      "learning_rate": 5.209137709137709e-06,
      "loss": 0.6391,
      "step": 7705
    },
    {
      "epoch": 1.65,
      "grad_norm": 8.683451652526855,
      "learning_rate": 5.205920205920206e-06,
      "loss": 0.4916,
      "step": 7706
    },
    {
      "epoch": 1.65,
      "grad_norm": 11.201866149902344,
      "learning_rate": 5.202702702702703e-06,
      "loss": 0.5902,
      "step": 7707
    },
    {
      "epoch": 1.65,
      "grad_norm": 8.647167205810547,
      "learning_rate": 5.1994851994852e-06,
      "loss": 0.447,
      "step": 7708
    },
    {
      "epoch": 1.65,
      "grad_norm": 12.42973804473877,
      "learning_rate": 5.196267696267696e-06,
      "loss": 0.5892,
      "step": 7709
    },
    {
      "epoch": 1.65,
      "grad_norm": 11.631484031677246,
      "learning_rate": 5.193050193050193e-06,
      "loss": 0.6814,
      "step": 7710
    },
    {
      "epoch": 1.65,
      "grad_norm": 14.931436538696289,
      "learning_rate": 5.18983268983269e-06,
      "loss": 0.6511,
      "step": 7711
    },
    {
      "epoch": 1.65,
      "grad_norm": 14.757052421569824,
      "learning_rate": 5.186615186615187e-06,
      "loss": 0.6494,
      "step": 7712
    },
    {
      "epoch": 1.65,
      "grad_norm": 8.506975173950195,
      "learning_rate": 5.183397683397683e-06,
      "loss": 0.3882,
      "step": 7713
    },
    {
      "epoch": 1.65,
      "grad_norm": 13.242058753967285,
      "learning_rate": 5.1801801801801805e-06,
      "loss": 0.4033,
      "step": 7714
    },
    {
      "epoch": 1.65,
      "grad_norm": 8.220222473144531,
      "learning_rate": 5.176962676962677e-06,
      "loss": 0.5047,
      "step": 7715
    },
    {
      "epoch": 1.66,
      "grad_norm": 12.925904273986816,
      "learning_rate": 5.173745173745174e-06,
      "loss": 0.6217,
      "step": 7716
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.895838737487793,
      "learning_rate": 5.170527670527671e-06,
      "loss": 0.5124,
      "step": 7717
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.768260955810547,
      "learning_rate": 5.167310167310167e-06,
      "loss": 0.6092,
      "step": 7718
    },
    {
      "epoch": 1.66,
      "grad_norm": 11.60086441040039,
      "learning_rate": 5.1640926640926644e-06,
      "loss": 0.4894,
      "step": 7719
    },
    {
      "epoch": 1.66,
      "grad_norm": 19.33696174621582,
      "learning_rate": 5.160875160875161e-06,
      "loss": 0.5534,
      "step": 7720
    },
    {
      "epoch": 1.66,
      "grad_norm": 12.830318450927734,
      "learning_rate": 5.157657657657658e-06,
      "loss": 0.6417,
      "step": 7721
    },
    {
      "epoch": 1.66,
      "grad_norm": 13.210615158081055,
      "learning_rate": 5.154440154440155e-06,
      "loss": 0.5946,
      "step": 7722
    },
    {
      "epoch": 1.66,
      "grad_norm": 12.06258487701416,
      "learning_rate": 5.151222651222651e-06,
      "loss": 0.4706,
      "step": 7723
    },
    {
      "epoch": 1.66,
      "grad_norm": 18.508852005004883,
      "learning_rate": 5.148005148005148e-06,
      "loss": 0.4379,
      "step": 7724
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.868378639221191,
      "learning_rate": 5.144787644787645e-06,
      "loss": 0.4918,
      "step": 7725
    },
    {
      "epoch": 1.66,
      "grad_norm": 15.270796775817871,
      "learning_rate": 5.141570141570141e-06,
      "loss": 0.6905,
      "step": 7726
    },
    {
      "epoch": 1.66,
      "grad_norm": 8.62180233001709,
      "learning_rate": 5.138352638352639e-06,
      "loss": 0.3248,
      "step": 7727
    },
    {
      "epoch": 1.66,
      "grad_norm": 14.018756866455078,
      "learning_rate": 5.135135135135135e-06,
      "loss": 0.5838,
      "step": 7728
    },
    {
      "epoch": 1.66,
      "grad_norm": 11.485774040222168,
      "learning_rate": 5.131917631917632e-06,
      "loss": 0.7048,
      "step": 7729
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.531338691711426,
      "learning_rate": 5.128700128700129e-06,
      "loss": 0.4896,
      "step": 7730
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.844805717468262,
      "learning_rate": 5.125482625482625e-06,
      "loss": 0.5002,
      "step": 7731
    },
    {
      "epoch": 1.66,
      "grad_norm": 6.842074871063232,
      "learning_rate": 5.1222651222651226e-06,
      "loss": 0.321,
      "step": 7732
    },
    {
      "epoch": 1.66,
      "grad_norm": 13.706470489501953,
      "learning_rate": 5.119047619047619e-06,
      "loss": 0.6039,
      "step": 7733
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.705791473388672,
      "learning_rate": 5.115830115830116e-06,
      "loss": 0.6836,
      "step": 7734
    },
    {
      "epoch": 1.66,
      "grad_norm": 11.659279823303223,
      "learning_rate": 5.112612612612613e-06,
      "loss": 0.6054,
      "step": 7735
    },
    {
      "epoch": 1.66,
      "grad_norm": 8.686728477478027,
      "learning_rate": 5.109395109395109e-06,
      "loss": 0.4268,
      "step": 7736
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.185722351074219,
      "learning_rate": 5.1061776061776065e-06,
      "loss": 0.578,
      "step": 7737
    },
    {
      "epoch": 1.66,
      "grad_norm": 8.772978782653809,
      "learning_rate": 5.102960102960103e-06,
      "loss": 0.469,
      "step": 7738
    },
    {
      "epoch": 1.66,
      "grad_norm": 12.228527069091797,
      "learning_rate": 5.099742599742599e-06,
      "loss": 0.5451,
      "step": 7739
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.962804794311523,
      "learning_rate": 5.096525096525097e-06,
      "loss": 0.4851,
      "step": 7740
    },
    {
      "epoch": 1.66,
      "grad_norm": 11.257271766662598,
      "learning_rate": 5.093307593307593e-06,
      "loss": 0.5852,
      "step": 7741
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.049186706542969,
      "learning_rate": 5.0900900900900905e-06,
      "loss": 0.431,
      "step": 7742
    },
    {
      "epoch": 1.66,
      "grad_norm": 7.079954147338867,
      "learning_rate": 5.086872586872588e-06,
      "loss": 0.3249,
      "step": 7743
    },
    {
      "epoch": 1.66,
      "grad_norm": 8.840414047241211,
      "learning_rate": 5.083655083655083e-06,
      "loss": 0.4122,
      "step": 7744
    },
    {
      "epoch": 1.66,
      "grad_norm": 12.006795883178711,
      "learning_rate": 5.080437580437581e-06,
      "loss": 0.6587,
      "step": 7745
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.641678810119629,
      "learning_rate": 5.077220077220077e-06,
      "loss": 0.5045,
      "step": 7746
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.106423377990723,
      "learning_rate": 5.074002574002574e-06,
      "loss": 0.5478,
      "step": 7747
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.549559593200684,
      "learning_rate": 5.070785070785071e-06,
      "loss": 0.6538,
      "step": 7748
    },
    {
      "epoch": 1.66,
      "grad_norm": 11.774654388427734,
      "learning_rate": 5.067567567567567e-06,
      "loss": 0.4394,
      "step": 7749
    },
    {
      "epoch": 1.66,
      "grad_norm": 7.628617286682129,
      "learning_rate": 5.064350064350065e-06,
      "loss": 0.5286,
      "step": 7750
    },
    {
      "epoch": 1.66,
      "grad_norm": 8.85586929321289,
      "learning_rate": 5.061132561132562e-06,
      "loss": 0.4795,
      "step": 7751
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.891240119934082,
      "learning_rate": 5.0579150579150575e-06,
      "loss": 0.5023,
      "step": 7752
    },
    {
      "epoch": 1.66,
      "grad_norm": 12.330960273742676,
      "learning_rate": 5.054697554697555e-06,
      "loss": 0.5101,
      "step": 7753
    },
    {
      "epoch": 1.66,
      "grad_norm": 12.114712715148926,
      "learning_rate": 5.051480051480051e-06,
      "loss": 0.556,
      "step": 7754
    },
    {
      "epoch": 1.66,
      "grad_norm": 12.243078231811523,
      "learning_rate": 5.0482625482625486e-06,
      "loss": 0.5798,
      "step": 7755
    },
    {
      "epoch": 1.66,
      "grad_norm": 7.011043071746826,
      "learning_rate": 5.045045045045046e-06,
      "loss": 0.3816,
      "step": 7756
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.886971473693848,
      "learning_rate": 5.0418275418275415e-06,
      "loss": 0.3885,
      "step": 7757
    },
    {
      "epoch": 1.66,
      "grad_norm": 11.02071475982666,
      "learning_rate": 5.038610038610039e-06,
      "loss": 0.4586,
      "step": 7758
    },
    {
      "epoch": 1.66,
      "grad_norm": 11.336213111877441,
      "learning_rate": 5.035392535392536e-06,
      "loss": 0.528,
      "step": 7759
    },
    {
      "epoch": 1.66,
      "grad_norm": 13.308032035827637,
      "learning_rate": 5.0321750321750325e-06,
      "loss": 0.5786,
      "step": 7760
    },
    {
      "epoch": 1.66,
      "grad_norm": 9.933829307556152,
      "learning_rate": 5.028957528957529e-06,
      "loss": 0.4798,
      "step": 7761
    },
    {
      "epoch": 1.66,
      "grad_norm": 11.615636825561523,
      "learning_rate": 5.025740025740025e-06,
      "loss": 0.5323,
      "step": 7762
    },
    {
      "epoch": 1.67,
      "grad_norm": 7.788389682769775,
      "learning_rate": 5.022522522522523e-06,
      "loss": 0.3862,
      "step": 7763
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.29697322845459,
      "learning_rate": 5.01930501930502e-06,
      "loss": 0.3469,
      "step": 7764
    },
    {
      "epoch": 1.67,
      "grad_norm": 8.637171745300293,
      "learning_rate": 5.016087516087516e-06,
      "loss": 0.4394,
      "step": 7765
    },
    {
      "epoch": 1.67,
      "grad_norm": 11.105676651000977,
      "learning_rate": 5.012870012870013e-06,
      "loss": 0.6436,
      "step": 7766
    },
    {
      "epoch": 1.67,
      "grad_norm": 11.47517204284668,
      "learning_rate": 5.009652509652509e-06,
      "loss": 0.5402,
      "step": 7767
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.53274917602539,
      "learning_rate": 5.006435006435007e-06,
      "loss": 0.4258,
      "step": 7768
    },
    {
      "epoch": 1.67,
      "grad_norm": 8.412343978881836,
      "learning_rate": 5.003217503217504e-06,
      "loss": 0.4969,
      "step": 7769
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.790732383728027,
      "learning_rate": 4.9999999999999996e-06,
      "loss": 0.6112,
      "step": 7770
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.828782081604004,
      "learning_rate": 4.996782496782497e-06,
      "loss": 0.398,
      "step": 7771
    },
    {
      "epoch": 1.67,
      "grad_norm": 12.306676864624023,
      "learning_rate": 4.993564993564994e-06,
      "loss": 0.6286,
      "step": 7772
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.729743957519531,
      "learning_rate": 4.990347490347491e-06,
      "loss": 0.3989,
      "step": 7773
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.735662460327148,
      "learning_rate": 4.987129987129987e-06,
      "loss": 0.4287,
      "step": 7774
    },
    {
      "epoch": 1.67,
      "grad_norm": 7.97251558303833,
      "learning_rate": 4.9839124839124835e-06,
      "loss": 0.3087,
      "step": 7775
    },
    {
      "epoch": 1.67,
      "grad_norm": 11.26356029510498,
      "learning_rate": 4.980694980694981e-06,
      "loss": 0.4479,
      "step": 7776
    },
    {
      "epoch": 1.67,
      "grad_norm": 7.460742473602295,
      "learning_rate": 4.977477477477478e-06,
      "loss": 0.4317,
      "step": 7777
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.572537422180176,
      "learning_rate": 4.974259974259974e-06,
      "loss": 0.4053,
      "step": 7778
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.7667818069458,
      "learning_rate": 4.971042471042471e-06,
      "loss": 0.4571,
      "step": 7779
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.811210632324219,
      "learning_rate": 4.967824967824968e-06,
      "loss": 0.5643,
      "step": 7780
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.102707862854004,
      "learning_rate": 4.964607464607465e-06,
      "loss": 0.5711,
      "step": 7781
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.062295913696289,
      "learning_rate": 4.961389961389962e-06,
      "loss": 0.3645,
      "step": 7782
    },
    {
      "epoch": 1.67,
      "grad_norm": 13.197494506835938,
      "learning_rate": 4.958172458172458e-06,
      "loss": 0.8305,
      "step": 7783
    },
    {
      "epoch": 1.67,
      "grad_norm": 12.211474418640137,
      "learning_rate": 4.954954954954955e-06,
      "loss": 0.5612,
      "step": 7784
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.3173828125,
      "learning_rate": 4.951737451737452e-06,
      "loss": 0.3856,
      "step": 7785
    },
    {
      "epoch": 1.67,
      "grad_norm": 12.075206756591797,
      "learning_rate": 4.948519948519949e-06,
      "loss": 0.7305,
      "step": 7786
    },
    {
      "epoch": 1.67,
      "grad_norm": 12.01963996887207,
      "learning_rate": 4.945302445302445e-06,
      "loss": 0.5913,
      "step": 7787
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.054078102111816,
      "learning_rate": 4.9420849420849425e-06,
      "loss": 0.4558,
      "step": 7788
    },
    {
      "epoch": 1.67,
      "grad_norm": 12.916918754577637,
      "learning_rate": 4.938867438867439e-06,
      "loss": 0.6883,
      "step": 7789
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.252179145812988,
      "learning_rate": 4.935649935649936e-06,
      "loss": 0.392,
      "step": 7790
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.952083587646484,
      "learning_rate": 4.932432432432433e-06,
      "loss": 0.5895,
      "step": 7791
    },
    {
      "epoch": 1.67,
      "grad_norm": 11.604815483093262,
      "learning_rate": 4.929214929214929e-06,
      "loss": 0.4528,
      "step": 7792
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.346896171569824,
      "learning_rate": 4.925997425997426e-06,
      "loss": 0.4792,
      "step": 7793
    },
    {
      "epoch": 1.67,
      "grad_norm": 9.928857803344727,
      "learning_rate": 4.922779922779923e-06,
      "loss": 0.4884,
      "step": 7794
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.74377155303955,
      "learning_rate": 4.91956241956242e-06,
      "loss": 0.4839,
      "step": 7795
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.1010103225708,
      "learning_rate": 4.916344916344916e-06,
      "loss": 0.4583,
      "step": 7796
    },
    {
      "epoch": 1.67,
      "grad_norm": 11.408744812011719,
      "learning_rate": 4.913127413127413e-06,
      "loss": 0.571,
      "step": 7797
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.72629451751709,
      "learning_rate": 4.90990990990991e-06,
      "loss": 0.5232,
      "step": 7798
    },
    {
      "epoch": 1.67,
      "grad_norm": 12.712874412536621,
      "learning_rate": 4.906692406692407e-06,
      "loss": 0.5583,
      "step": 7799
    },
    {
      "epoch": 1.67,
      "grad_norm": 13.546791076660156,
      "learning_rate": 4.903474903474903e-06,
      "loss": 0.5033,
      "step": 7800
    },
    {
      "epoch": 1.67,
      "grad_norm": 11.116850852966309,
      "learning_rate": 4.9002574002574006e-06,
      "loss": 0.6598,
      "step": 7801
    },
    {
      "epoch": 1.67,
      "grad_norm": 11.462044715881348,
      "learning_rate": 4.897039897039897e-06,
      "loss": 0.478,
      "step": 7802
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.369306564331055,
      "learning_rate": 4.893822393822394e-06,
      "loss": 0.4435,
      "step": 7803
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.992887496948242,
      "learning_rate": 4.890604890604891e-06,
      "loss": 0.7551,
      "step": 7804
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.72228717803955,
      "learning_rate": 4.887387387387387e-06,
      "loss": 0.5372,
      "step": 7805
    },
    {
      "epoch": 1.67,
      "grad_norm": 13.684765815734863,
      "learning_rate": 4.8841698841698845e-06,
      "loss": 0.6406,
      "step": 7806
    },
    {
      "epoch": 1.67,
      "grad_norm": 11.559279441833496,
      "learning_rate": 4.880952380952381e-06,
      "loss": 0.4138,
      "step": 7807
    },
    {
      "epoch": 1.67,
      "grad_norm": 10.191723823547363,
      "learning_rate": 4.877734877734878e-06,
      "loss": 0.4533,
      "step": 7808
    },
    {
      "epoch": 1.68,
      "grad_norm": 8.670591354370117,
      "learning_rate": 4.874517374517375e-06,
      "loss": 0.5033,
      "step": 7809
    },
    {
      "epoch": 1.68,
      "grad_norm": 11.201313018798828,
      "learning_rate": 4.871299871299871e-06,
      "loss": 0.4123,
      "step": 7810
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.253277778625488,
      "learning_rate": 4.8680823680823685e-06,
      "loss": 0.3719,
      "step": 7811
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.899375915527344,
      "learning_rate": 4.864864864864865e-06,
      "loss": 0.4076,
      "step": 7812
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.848249435424805,
      "learning_rate": 4.861647361647361e-06,
      "loss": 0.6243,
      "step": 7813
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.12615966796875,
      "learning_rate": 4.858429858429859e-06,
      "loss": 0.5375,
      "step": 7814
    },
    {
      "epoch": 1.68,
      "grad_norm": 13.125625610351562,
      "learning_rate": 4.855212355212355e-06,
      "loss": 0.6125,
      "step": 7815
    },
    {
      "epoch": 1.68,
      "grad_norm": 8.408369064331055,
      "learning_rate": 4.851994851994852e-06,
      "loss": 0.4005,
      "step": 7816
    },
    {
      "epoch": 1.68,
      "grad_norm": 12.586851119995117,
      "learning_rate": 4.84877734877735e-06,
      "loss": 0.7753,
      "step": 7817
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.8889799118042,
      "learning_rate": 4.845559845559845e-06,
      "loss": 0.4097,
      "step": 7818
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.33205795288086,
      "learning_rate": 4.842342342342343e-06,
      "loss": 0.5212,
      "step": 7819
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.202783584594727,
      "learning_rate": 4.839124839124839e-06,
      "loss": 0.3899,
      "step": 7820
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.355387687683105,
      "learning_rate": 4.835907335907336e-06,
      "loss": 0.5398,
      "step": 7821
    },
    {
      "epoch": 1.68,
      "grad_norm": 8.441760063171387,
      "learning_rate": 4.832689832689833e-06,
      "loss": 0.4134,
      "step": 7822
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.980335235595703,
      "learning_rate": 4.829472329472329e-06,
      "loss": 0.409,
      "step": 7823
    },
    {
      "epoch": 1.68,
      "grad_norm": 12.027289390563965,
      "learning_rate": 4.8262548262548266e-06,
      "loss": 0.4861,
      "step": 7824
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.561500549316406,
      "learning_rate": 4.823037323037323e-06,
      "loss": 0.4186,
      "step": 7825
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.249710083007812,
      "learning_rate": 4.8198198198198195e-06,
      "loss": 0.4714,
      "step": 7826
    },
    {
      "epoch": 1.68,
      "grad_norm": 11.079460144042969,
      "learning_rate": 4.816602316602317e-06,
      "loss": 0.4474,
      "step": 7827
    },
    {
      "epoch": 1.68,
      "grad_norm": 8.007254600524902,
      "learning_rate": 4.813384813384813e-06,
      "loss": 0.4414,
      "step": 7828
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.220001220703125,
      "learning_rate": 4.8101673101673105e-06,
      "loss": 0.7111,
      "step": 7829
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.22982120513916,
      "learning_rate": 4.806949806949808e-06,
      "loss": 0.4275,
      "step": 7830
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.596960067749023,
      "learning_rate": 4.8037323037323034e-06,
      "loss": 0.5756,
      "step": 7831
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.060981750488281,
      "learning_rate": 4.800514800514801e-06,
      "loss": 0.4786,
      "step": 7832
    },
    {
      "epoch": 1.68,
      "grad_norm": 8.184794425964355,
      "learning_rate": 4.797297297297297e-06,
      "loss": 0.319,
      "step": 7833
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.534828186035156,
      "learning_rate": 4.7940797940797945e-06,
      "loss": 0.414,
      "step": 7834
    },
    {
      "epoch": 1.68,
      "grad_norm": 8.590819358825684,
      "learning_rate": 4.790862290862291e-06,
      "loss": 0.4126,
      "step": 7835
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.358038902282715,
      "learning_rate": 4.787644787644787e-06,
      "loss": 0.5579,
      "step": 7836
    },
    {
      "epoch": 1.68,
      "grad_norm": 11.997620582580566,
      "learning_rate": 4.784427284427285e-06,
      "loss": 0.6018,
      "step": 7837
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.430391311645508,
      "learning_rate": 4.781209781209782e-06,
      "loss": 0.6124,
      "step": 7838
    },
    {
      "epoch": 1.68,
      "grad_norm": 11.286005973815918,
      "learning_rate": 4.777992277992278e-06,
      "loss": 0.4208,
      "step": 7839
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.99931526184082,
      "learning_rate": 4.774774774774775e-06,
      "loss": 0.4498,
      "step": 7840
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.714166641235352,
      "learning_rate": 4.771557271557271e-06,
      "loss": 0.5635,
      "step": 7841
    },
    {
      "epoch": 1.68,
      "grad_norm": 12.34352970123291,
      "learning_rate": 4.768339768339769e-06,
      "loss": 0.479,
      "step": 7842
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.910189628601074,
      "learning_rate": 4.765122265122266e-06,
      "loss": 0.6617,
      "step": 7843
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.359907150268555,
      "learning_rate": 4.7619047619047615e-06,
      "loss": 0.5038,
      "step": 7844
    },
    {
      "epoch": 1.68,
      "grad_norm": 15.159612655639648,
      "learning_rate": 4.758687258687259e-06,
      "loss": 0.9404,
      "step": 7845
    },
    {
      "epoch": 1.68,
      "grad_norm": 7.590959548950195,
      "learning_rate": 4.755469755469756e-06,
      "loss": 0.4261,
      "step": 7846
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.814590454101562,
      "learning_rate": 4.7522522522522526e-06,
      "loss": 0.4372,
      "step": 7847
    },
    {
      "epoch": 1.68,
      "grad_norm": 13.096534729003906,
      "learning_rate": 4.749034749034749e-06,
      "loss": 0.5507,
      "step": 7848
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.32949161529541,
      "learning_rate": 4.7458172458172455e-06,
      "loss": 0.4072,
      "step": 7849
    },
    {
      "epoch": 1.68,
      "grad_norm": 12.215228080749512,
      "learning_rate": 4.742599742599743e-06,
      "loss": 0.5215,
      "step": 7850
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.123655319213867,
      "learning_rate": 4.73938223938224e-06,
      "loss": 0.6061,
      "step": 7851
    },
    {
      "epoch": 1.68,
      "grad_norm": 11.02606201171875,
      "learning_rate": 4.736164736164736e-06,
      "loss": 0.6286,
      "step": 7852
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.442645072937012,
      "learning_rate": 4.732947232947233e-06,
      "loss": 0.5149,
      "step": 7853
    },
    {
      "epoch": 1.68,
      "grad_norm": 11.508666038513184,
      "learning_rate": 4.7297297297297294e-06,
      "loss": 0.5702,
      "step": 7854
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.028362274169922,
      "learning_rate": 4.726512226512227e-06,
      "loss": 0.507,
      "step": 7855
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.701101303100586,
      "learning_rate": 4.723294723294724e-06,
      "loss": 0.4816,
      "step": 7856
    },
    {
      "epoch": 1.69,
      "grad_norm": 7.958484649658203,
      "learning_rate": 4.72007722007722e-06,
      "loss": 0.2885,
      "step": 7857
    },
    {
      "epoch": 1.69,
      "grad_norm": 14.690412521362305,
      "learning_rate": 4.716859716859717e-06,
      "loss": 0.7468,
      "step": 7858
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.530288696289062,
      "learning_rate": 4.713642213642214e-06,
      "loss": 0.5442,
      "step": 7859
    },
    {
      "epoch": 1.69,
      "grad_norm": 12.021073341369629,
      "learning_rate": 4.710424710424711e-06,
      "loss": 0.5248,
      "step": 7860
    },
    {
      "epoch": 1.69,
      "grad_norm": 12.678600311279297,
      "learning_rate": 4.707207207207207e-06,
      "loss": 0.5041,
      "step": 7861
    },
    {
      "epoch": 1.69,
      "grad_norm": 11.16596508026123,
      "learning_rate": 4.703989703989704e-06,
      "loss": 0.4791,
      "step": 7862
    },
    {
      "epoch": 1.69,
      "grad_norm": 12.083756446838379,
      "learning_rate": 4.700772200772201e-06,
      "loss": 0.8151,
      "step": 7863
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.493157386779785,
      "learning_rate": 4.697554697554698e-06,
      "loss": 0.4595,
      "step": 7864
    },
    {
      "epoch": 1.69,
      "grad_norm": 10.18600082397461,
      "learning_rate": 4.694337194337194e-06,
      "loss": 0.4453,
      "step": 7865
    },
    {
      "epoch": 1.69,
      "grad_norm": 11.869142532348633,
      "learning_rate": 4.691119691119691e-06,
      "loss": 0.6064,
      "step": 7866
    },
    {
      "epoch": 1.69,
      "grad_norm": 12.03445053100586,
      "learning_rate": 4.687902187902188e-06,
      "loss": 0.603,
      "step": 7867
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.977259635925293,
      "learning_rate": 4.684684684684685e-06,
      "loss": 0.6437,
      "step": 7868
    },
    {
      "epoch": 1.69,
      "grad_norm": 10.089524269104004,
      "learning_rate": 4.681467181467182e-06,
      "loss": 0.347,
      "step": 7869
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.830704689025879,
      "learning_rate": 4.678249678249678e-06,
      "loss": 0.2726,
      "step": 7870
    },
    {
      "epoch": 1.69,
      "grad_norm": 12.397848129272461,
      "learning_rate": 4.675032175032175e-06,
      "loss": 0.4714,
      "step": 7871
    },
    {
      "epoch": 1.69,
      "grad_norm": 10.536880493164062,
      "learning_rate": 4.671814671814672e-06,
      "loss": 0.4494,
      "step": 7872
    },
    {
      "epoch": 1.69,
      "grad_norm": 13.98963737487793,
      "learning_rate": 4.668597168597169e-06,
      "loss": 0.7372,
      "step": 7873
    },
    {
      "epoch": 1.69,
      "grad_norm": 11.828879356384277,
      "learning_rate": 4.665379665379665e-06,
      "loss": 0.5765,
      "step": 7874
    },
    {
      "epoch": 1.69,
      "grad_norm": 7.6093950271606445,
      "learning_rate": 4.6621621621621625e-06,
      "loss": 0.3982,
      "step": 7875
    },
    {
      "epoch": 1.69,
      "grad_norm": 11.35811710357666,
      "learning_rate": 4.658944658944659e-06,
      "loss": 0.4993,
      "step": 7876
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.134431838989258,
      "learning_rate": 4.655727155727156e-06,
      "loss": 0.2972,
      "step": 7877
    },
    {
      "epoch": 1.69,
      "grad_norm": 10.790802001953125,
      "learning_rate": 4.652509652509652e-06,
      "loss": 0.663,
      "step": 7878
    },
    {
      "epoch": 1.69,
      "grad_norm": 13.620170593261719,
      "learning_rate": 4.649292149292149e-06,
      "loss": 0.6028,
      "step": 7879
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.571343421936035,
      "learning_rate": 4.6460746460746465e-06,
      "loss": 0.4004,
      "step": 7880
    },
    {
      "epoch": 1.69,
      "grad_norm": 13.333313941955566,
      "learning_rate": 4.642857142857143e-06,
      "loss": 0.4969,
      "step": 7881
    },
    {
      "epoch": 1.69,
      "grad_norm": 8.897014617919922,
      "learning_rate": 4.63963963963964e-06,
      "loss": 0.5899,
      "step": 7882
    },
    {
      "epoch": 1.69,
      "grad_norm": 10.499094009399414,
      "learning_rate": 4.636422136422136e-06,
      "loss": 0.5255,
      "step": 7883
    },
    {
      "epoch": 1.69,
      "grad_norm": 11.326708793640137,
      "learning_rate": 4.633204633204633e-06,
      "loss": 0.5887,
      "step": 7884
    },
    {
      "epoch": 1.69,
      "grad_norm": 10.909588813781738,
      "learning_rate": 4.6299871299871304e-06,
      "loss": 0.5427,
      "step": 7885
    },
    {
      "epoch": 1.69,
      "grad_norm": 9.736601829528809,
      "learning_rate": 4.626769626769627e-06,
      "loss": 0.3857,
      "step": 7886
    },
    {
      "epoch": 1.69,
      "grad_norm": 15.461023330688477,
      "learning_rate": 4.623552123552123e-06,
      "loss": 0.6027,
      "step": 7887
    },
    {
      "epoch": 1.69,
      "grad_norm": 13.634675025939941,
      "learning_rate": 4.620334620334621e-06,
      "loss": 0.6114,
      "step": 7888
    },
    {
      "epoch": 1.69,
      "grad_norm": 11.028043746948242,
      "learning_rate": 4.617117117117117e-06,
      "loss": 0.5276,
      "step": 7889
    },
    {
      "epoch": 1.69,
      "grad_norm": 11.046555519104004,
      "learning_rate": 4.613899613899614e-06,
      "loss": 0.4209,
      "step": 7890
    },
    {
      "epoch": 1.69,
      "grad_norm": 12.004850387573242,
      "learning_rate": 4.610682110682111e-06,
      "loss": 0.5904,
      "step": 7891
    },
    {
      "epoch": 1.69,
      "grad_norm": 12.721579551696777,
      "learning_rate": 4.607464607464607e-06,
      "loss": 0.5076,
      "step": 7892
    },
    {
      "epoch": 1.69,
      "grad_norm": 12.706603050231934,
      "learning_rate": 4.604247104247105e-06,
      "loss": 0.5235,
      "step": 7893
    },
    {
      "epoch": 1.69,
      "grad_norm": 11.799979209899902,
      "learning_rate": 4.601029601029601e-06,
      "loss": 0.6407,
      "step": 7894
    },
    {
      "epoch": 1.69,
      "grad_norm": 11.00562858581543,
      "learning_rate": 4.597812097812098e-06,
      "loss": 0.4919,
      "step": 7895
    },
    {
      "epoch": 1.69,
      "grad_norm": 13.48463249206543,
      "learning_rate": 4.594594594594595e-06,
      "loss": 0.651,
      "step": 7896
    },
    {
      "epoch": 1.69,
      "grad_norm": 10.091043472290039,
      "learning_rate": 4.591377091377091e-06,
      "loss": 0.4439,
      "step": 7897
    },
    {
      "epoch": 1.69,
      "grad_norm": 12.743638038635254,
      "learning_rate": 4.5881595881595885e-06,
      "loss": 1.0217,
      "step": 7898
    },
    {
      "epoch": 1.69,
      "grad_norm": 8.080218315124512,
      "learning_rate": 4.584942084942085e-06,
      "loss": 0.4476,
      "step": 7899
    },
    {
      "epoch": 1.69,
      "grad_norm": 12.065183639526367,
      "learning_rate": 4.5817245817245814e-06,
      "loss": 0.4407,
      "step": 7900
    },
    {
      "epoch": 1.69,
      "grad_norm": 7.391098976135254,
      "learning_rate": 4.578507078507079e-06,
      "loss": 0.3828,
      "step": 7901
    },
    {
      "epoch": 1.69,
      "grad_norm": 10.252204895019531,
      "learning_rate": 4.575289575289575e-06,
      "loss": 0.4266,
      "step": 7902
    },
    {
      "epoch": 1.7,
      "grad_norm": 8.375783920288086,
      "learning_rate": 4.5720720720720725e-06,
      "loss": 0.3338,
      "step": 7903
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.463545799255371,
      "learning_rate": 4.56885456885457e-06,
      "loss": 0.7001,
      "step": 7904
    },
    {
      "epoch": 1.7,
      "grad_norm": 10.92947769165039,
      "learning_rate": 4.565637065637065e-06,
      "loss": 0.6201,
      "step": 7905
    },
    {
      "epoch": 1.7,
      "grad_norm": 12.881324768066406,
      "learning_rate": 4.562419562419563e-06,
      "loss": 0.5429,
      "step": 7906
    },
    {
      "epoch": 1.7,
      "grad_norm": 10.414605140686035,
      "learning_rate": 4.559202059202059e-06,
      "loss": 0.639,
      "step": 7907
    },
    {
      "epoch": 1.7,
      "grad_norm": 10.41418743133545,
      "learning_rate": 4.5559845559845564e-06,
      "loss": 0.424,
      "step": 7908
    },
    {
      "epoch": 1.7,
      "grad_norm": 8.459362983703613,
      "learning_rate": 4.552767052767053e-06,
      "loss": 0.3506,
      "step": 7909
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.778884887695312,
      "learning_rate": 4.549549549549549e-06,
      "loss": 0.5249,
      "step": 7910
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.808089256286621,
      "learning_rate": 4.546332046332047e-06,
      "loss": 0.4913,
      "step": 7911
    },
    {
      "epoch": 1.7,
      "grad_norm": 10.93488597869873,
      "learning_rate": 4.543114543114543e-06,
      "loss": 0.4806,
      "step": 7912
    },
    {
      "epoch": 1.7,
      "grad_norm": 10.405084609985352,
      "learning_rate": 4.5398970398970395e-06,
      "loss": 0.5663,
      "step": 7913
    },
    {
      "epoch": 1.7,
      "grad_norm": 12.317994117736816,
      "learning_rate": 4.536679536679537e-06,
      "loss": 0.3907,
      "step": 7914
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.188652992248535,
      "learning_rate": 4.533462033462033e-06,
      "loss": 0.6551,
      "step": 7915
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.948357582092285,
      "learning_rate": 4.530244530244531e-06,
      "loss": 0.5075,
      "step": 7916
    },
    {
      "epoch": 1.7,
      "grad_norm": 16.720840454101562,
      "learning_rate": 4.527027027027028e-06,
      "loss": 0.9516,
      "step": 7917
    },
    {
      "epoch": 1.7,
      "grad_norm": 12.183446884155273,
      "learning_rate": 4.5238095238095235e-06,
      "loss": 0.5266,
      "step": 7918
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.2553129196167,
      "learning_rate": 4.520592020592021e-06,
      "loss": 0.53,
      "step": 7919
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.175683975219727,
      "learning_rate": 4.517374517374517e-06,
      "loss": 0.487,
      "step": 7920
    },
    {
      "epoch": 1.7,
      "grad_norm": 13.725394248962402,
      "learning_rate": 4.5141570141570145e-06,
      "loss": 0.614,
      "step": 7921
    },
    {
      "epoch": 1.7,
      "grad_norm": 10.82953929901123,
      "learning_rate": 4.510939510939511e-06,
      "loss": 0.5848,
      "step": 7922
    },
    {
      "epoch": 1.7,
      "grad_norm": 8.095760345458984,
      "learning_rate": 4.5077220077220074e-06,
      "loss": 0.2954,
      "step": 7923
    },
    {
      "epoch": 1.7,
      "grad_norm": 10.95742130279541,
      "learning_rate": 4.504504504504505e-06,
      "loss": 0.4688,
      "step": 7924
    },
    {
      "epoch": 1.7,
      "grad_norm": 10.410033226013184,
      "learning_rate": 4.501287001287002e-06,
      "loss": 0.5381,
      "step": 7925
    },
    {
      "epoch": 1.7,
      "grad_norm": 7.752319812774658,
      "learning_rate": 4.498069498069498e-06,
      "loss": 0.2968,
      "step": 7926
    },
    {
      "epoch": 1.7,
      "grad_norm": 13.403180122375488,
      "learning_rate": 4.494851994851995e-06,
      "loss": 0.781,
      "step": 7927
    },
    {
      "epoch": 1.7,
      "grad_norm": 10.361571311950684,
      "learning_rate": 4.491634491634491e-06,
      "loss": 0.598,
      "step": 7928
    },
    {
      "epoch": 1.7,
      "grad_norm": 7.958426475524902,
      "learning_rate": 4.488416988416989e-06,
      "loss": 0.4175,
      "step": 7929
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.038212776184082,
      "learning_rate": 4.485199485199486e-06,
      "loss": 0.6412,
      "step": 7930
    },
    {
      "epoch": 1.7,
      "grad_norm": 10.128908157348633,
      "learning_rate": 4.481981981981982e-06,
      "loss": 0.4893,
      "step": 7931
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.83525562286377,
      "learning_rate": 4.478764478764479e-06,
      "loss": 0.4093,
      "step": 7932
    },
    {
      "epoch": 1.7,
      "grad_norm": 12.602128982543945,
      "learning_rate": 4.475546975546976e-06,
      "loss": 0.5359,
      "step": 7933
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.09355354309082,
      "learning_rate": 4.472329472329473e-06,
      "loss": 0.6774,
      "step": 7934
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.25125503540039,
      "learning_rate": 4.469111969111969e-06,
      "loss": 0.5319,
      "step": 7935
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.477646827697754,
      "learning_rate": 4.4658944658944655e-06,
      "loss": 0.4322,
      "step": 7936
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.465322494506836,
      "learning_rate": 4.462676962676963e-06,
      "loss": 0.4736,
      "step": 7937
    },
    {
      "epoch": 1.7,
      "grad_norm": 12.584925651550293,
      "learning_rate": 4.45945945945946e-06,
      "loss": 0.5454,
      "step": 7938
    },
    {
      "epoch": 1.7,
      "grad_norm": 12.839491844177246,
      "learning_rate": 4.456241956241956e-06,
      "loss": 0.5373,
      "step": 7939
    },
    {
      "epoch": 1.7,
      "grad_norm": 14.499930381774902,
      "learning_rate": 4.453024453024453e-06,
      "loss": 0.5435,
      "step": 7940
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.663313865661621,
      "learning_rate": 4.4498069498069495e-06,
      "loss": 0.4701,
      "step": 7941
    },
    {
      "epoch": 1.7,
      "grad_norm": 12.175488471984863,
      "learning_rate": 4.446589446589447e-06,
      "loss": 0.6989,
      "step": 7942
    },
    {
      "epoch": 1.7,
      "grad_norm": 7.956993579864502,
      "learning_rate": 4.443371943371944e-06,
      "loss": 0.4309,
      "step": 7943
    },
    {
      "epoch": 1.7,
      "grad_norm": 8.541441917419434,
      "learning_rate": 4.44015444015444e-06,
      "loss": 0.4799,
      "step": 7944
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.7501802444458,
      "learning_rate": 4.436936936936937e-06,
      "loss": 0.5748,
      "step": 7945
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.698433876037598,
      "learning_rate": 4.433719433719434e-06,
      "loss": 0.3966,
      "step": 7946
    },
    {
      "epoch": 1.7,
      "grad_norm": 6.9036431312561035,
      "learning_rate": 4.430501930501931e-06,
      "loss": 0.3711,
      "step": 7947
    },
    {
      "epoch": 1.7,
      "grad_norm": 10.328205108642578,
      "learning_rate": 4.427284427284427e-06,
      "loss": 0.5875,
      "step": 7948
    },
    {
      "epoch": 1.71,
      "grad_norm": 9.850481986999512,
      "learning_rate": 4.424066924066924e-06,
      "loss": 0.5817,
      "step": 7949
    },
    {
      "epoch": 1.71,
      "grad_norm": 8.780780792236328,
      "learning_rate": 4.420849420849421e-06,
      "loss": 0.4226,
      "step": 7950
    },
    {
      "epoch": 1.71,
      "grad_norm": 9.21139907836914,
      "learning_rate": 4.417631917631918e-06,
      "loss": 0.4147,
      "step": 7951
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.199443817138672,
      "learning_rate": 4.414414414414414e-06,
      "loss": 0.6893,
      "step": 7952
    },
    {
      "epoch": 1.71,
      "grad_norm": 14.605401992797852,
      "learning_rate": 4.411196911196911e-06,
      "loss": 0.5822,
      "step": 7953
    },
    {
      "epoch": 1.71,
      "grad_norm": 7.77404260635376,
      "learning_rate": 4.4079794079794084e-06,
      "loss": 0.4151,
      "step": 7954
    },
    {
      "epoch": 1.71,
      "grad_norm": 14.345096588134766,
      "learning_rate": 4.404761904761905e-06,
      "loss": 0.3721,
      "step": 7955
    },
    {
      "epoch": 1.71,
      "grad_norm": 8.09056282043457,
      "learning_rate": 4.401544401544402e-06,
      "loss": 0.3367,
      "step": 7956
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.791346549987793,
      "learning_rate": 4.398326898326898e-06,
      "loss": 0.5372,
      "step": 7957
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.725730895996094,
      "learning_rate": 4.395109395109395e-06,
      "loss": 0.5933,
      "step": 7958
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.421220779418945,
      "learning_rate": 4.391891891891892e-06,
      "loss": 0.4731,
      "step": 7959
    },
    {
      "epoch": 1.71,
      "grad_norm": 9.294200897216797,
      "learning_rate": 4.388674388674389e-06,
      "loss": 0.3977,
      "step": 7960
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.32169246673584,
      "learning_rate": 4.385456885456885e-06,
      "loss": 0.5977,
      "step": 7961
    },
    {
      "epoch": 1.71,
      "grad_norm": 14.4395170211792,
      "learning_rate": 4.382239382239383e-06,
      "loss": 0.6356,
      "step": 7962
    },
    {
      "epoch": 1.71,
      "grad_norm": 17.062450408935547,
      "learning_rate": 4.379021879021879e-06,
      "loss": 0.5478,
      "step": 7963
    },
    {
      "epoch": 1.71,
      "grad_norm": 9.68635368347168,
      "learning_rate": 4.375804375804376e-06,
      "loss": 0.4467,
      "step": 7964
    },
    {
      "epoch": 1.71,
      "grad_norm": 9.893367767333984,
      "learning_rate": 4.372586872586872e-06,
      "loss": 0.4278,
      "step": 7965
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.98845100402832,
      "learning_rate": 4.369369369369369e-06,
      "loss": 0.7142,
      "step": 7966
    },
    {
      "epoch": 1.71,
      "grad_norm": 12.310770034790039,
      "learning_rate": 4.3661518661518666e-06,
      "loss": 0.5303,
      "step": 7967
    },
    {
      "epoch": 1.71,
      "grad_norm": 13.629460334777832,
      "learning_rate": 4.362934362934363e-06,
      "loss": 0.6848,
      "step": 7968
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.603241920471191,
      "learning_rate": 4.35971685971686e-06,
      "loss": 0.5329,
      "step": 7969
    },
    {
      "epoch": 1.71,
      "grad_norm": 13.905945777893066,
      "learning_rate": 4.356499356499356e-06,
      "loss": 0.4024,
      "step": 7970
    },
    {
      "epoch": 1.71,
      "grad_norm": 13.00112247467041,
      "learning_rate": 4.353281853281853e-06,
      "loss": 0.5049,
      "step": 7971
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.014930725097656,
      "learning_rate": 4.3500643500643505e-06,
      "loss": 0.5285,
      "step": 7972
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.66718578338623,
      "learning_rate": 4.346846846846847e-06,
      "loss": 0.5149,
      "step": 7973
    },
    {
      "epoch": 1.71,
      "grad_norm": 7.2298126220703125,
      "learning_rate": 4.343629343629343e-06,
      "loss": 0.2883,
      "step": 7974
    },
    {
      "epoch": 1.71,
      "grad_norm": 9.513364791870117,
      "learning_rate": 4.340411840411841e-06,
      "loss": 0.694,
      "step": 7975
    },
    {
      "epoch": 1.71,
      "grad_norm": 9.288551330566406,
      "learning_rate": 4.337194337194337e-06,
      "loss": 0.5282,
      "step": 7976
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.337677001953125,
      "learning_rate": 4.3339768339768345e-06,
      "loss": 0.5171,
      "step": 7977
    },
    {
      "epoch": 1.71,
      "grad_norm": 7.494719505310059,
      "learning_rate": 4.330759330759331e-06,
      "loss": 0.3454,
      "step": 7978
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.856880187988281,
      "learning_rate": 4.327541827541827e-06,
      "loss": 0.4384,
      "step": 7979
    },
    {
      "epoch": 1.71,
      "grad_norm": 8.77983570098877,
      "learning_rate": 4.324324324324325e-06,
      "loss": 0.3967,
      "step": 7980
    },
    {
      "epoch": 1.71,
      "grad_norm": 8.162309646606445,
      "learning_rate": 4.321106821106821e-06,
      "loss": 0.4676,
      "step": 7981
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.435277938842773,
      "learning_rate": 4.317889317889318e-06,
      "loss": 0.505,
      "step": 7982
    },
    {
      "epoch": 1.71,
      "grad_norm": 12.022156715393066,
      "learning_rate": 4.314671814671815e-06,
      "loss": 0.4257,
      "step": 7983
    },
    {
      "epoch": 1.71,
      "grad_norm": 13.25869083404541,
      "learning_rate": 4.311454311454311e-06,
      "loss": 0.748,
      "step": 7984
    },
    {
      "epoch": 1.71,
      "grad_norm": 8.975008010864258,
      "learning_rate": 4.308236808236809e-06,
      "loss": 0.5871,
      "step": 7985
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.220196723937988,
      "learning_rate": 4.305019305019305e-06,
      "loss": 0.4127,
      "step": 7986
    },
    {
      "epoch": 1.71,
      "grad_norm": 12.385642051696777,
      "learning_rate": 4.3018018018018015e-06,
      "loss": 0.6097,
      "step": 7987
    },
    {
      "epoch": 1.71,
      "grad_norm": 7.703951358795166,
      "learning_rate": 4.298584298584299e-06,
      "loss": 0.4053,
      "step": 7988
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.466328620910645,
      "learning_rate": 4.295366795366795e-06,
      "loss": 0.5304,
      "step": 7989
    },
    {
      "epoch": 1.71,
      "grad_norm": 7.815202713012695,
      "learning_rate": 4.2921492921492926e-06,
      "loss": 0.3759,
      "step": 7990
    },
    {
      "epoch": 1.71,
      "grad_norm": 8.785209655761719,
      "learning_rate": 4.28893178893179e-06,
      "loss": 0.3584,
      "step": 7991
    },
    {
      "epoch": 1.71,
      "grad_norm": 8.882625579833984,
      "learning_rate": 4.2857142857142855e-06,
      "loss": 0.4691,
      "step": 7992
    },
    {
      "epoch": 1.71,
      "grad_norm": 11.968940734863281,
      "learning_rate": 4.282496782496783e-06,
      "loss": 0.3475,
      "step": 7993
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.40860652923584,
      "learning_rate": 4.279279279279279e-06,
      "loss": 0.4308,
      "step": 7994
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.716511726379395,
      "learning_rate": 4.2760617760617765e-06,
      "loss": 0.5279,
      "step": 7995
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.041241645812988,
      "learning_rate": 4.272844272844273e-06,
      "loss": 0.4283,
      "step": 7996
    },
    {
      "epoch": 1.72,
      "grad_norm": 13.495401382446289,
      "learning_rate": 4.269626769626769e-06,
      "loss": 0.5069,
      "step": 7997
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.971473693847656,
      "learning_rate": 4.266409266409267e-06,
      "loss": 0.6342,
      "step": 7998
    },
    {
      "epoch": 1.72,
      "grad_norm": 10.32748031616211,
      "learning_rate": 4.263191763191763e-06,
      "loss": 0.4829,
      "step": 7999
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.938142776489258,
      "learning_rate": 4.25997425997426e-06,
      "loss": 0.5965,
      "step": 8000
    },
    {
      "epoch": 1.72,
      "grad_norm": 10.850075721740723,
      "learning_rate": 4.256756756756757e-06,
      "loss": 0.3358,
      "step": 8001
    },
    {
      "epoch": 1.72,
      "grad_norm": 8.410441398620605,
      "learning_rate": 4.253539253539253e-06,
      "loss": 0.3361,
      "step": 8002
    },
    {
      "epoch": 1.72,
      "grad_norm": 12.861051559448242,
      "learning_rate": 4.250321750321751e-06,
      "loss": 0.6523,
      "step": 8003
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.9890775680542,
      "learning_rate": 4.247104247104248e-06,
      "loss": 0.8132,
      "step": 8004
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.336822509765625,
      "learning_rate": 4.2438867438867436e-06,
      "loss": 0.4138,
      "step": 8005
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.903658866882324,
      "learning_rate": 4.240669240669241e-06,
      "loss": 0.6127,
      "step": 8006
    },
    {
      "epoch": 1.72,
      "grad_norm": 8.314718246459961,
      "learning_rate": 4.237451737451737e-06,
      "loss": 0.4914,
      "step": 8007
    },
    {
      "epoch": 1.72,
      "grad_norm": 12.33823013305664,
      "learning_rate": 4.234234234234235e-06,
      "loss": 0.6488,
      "step": 8008
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.622666358947754,
      "learning_rate": 4.231016731016731e-06,
      "loss": 0.5816,
      "step": 8009
    },
    {
      "epoch": 1.72,
      "grad_norm": 6.654306888580322,
      "learning_rate": 4.2277992277992275e-06,
      "loss": 0.3012,
      "step": 8010
    },
    {
      "epoch": 1.72,
      "grad_norm": 10.136810302734375,
      "learning_rate": 4.224581724581725e-06,
      "loss": 0.4894,
      "step": 8011
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.900901794433594,
      "learning_rate": 4.221364221364222e-06,
      "loss": 0.5805,
      "step": 8012
    },
    {
      "epoch": 1.72,
      "grad_norm": 8.862235069274902,
      "learning_rate": 4.218146718146718e-06,
      "loss": 0.4903,
      "step": 8013
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.431512832641602,
      "learning_rate": 4.214929214929215e-06,
      "loss": 0.4185,
      "step": 8014
    },
    {
      "epoch": 1.72,
      "grad_norm": 7.079101085662842,
      "learning_rate": 4.2117117117117115e-06,
      "loss": 0.4236,
      "step": 8015
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.49357795715332,
      "learning_rate": 4.208494208494209e-06,
      "loss": 0.5132,
      "step": 8016
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.392556190490723,
      "learning_rate": 4.205276705276706e-06,
      "loss": 0.7497,
      "step": 8017
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.948382377624512,
      "learning_rate": 4.202059202059202e-06,
      "loss": 0.4983,
      "step": 8018
    },
    {
      "epoch": 1.72,
      "grad_norm": 8.523537635803223,
      "learning_rate": 4.198841698841699e-06,
      "loss": 0.4822,
      "step": 8019
    },
    {
      "epoch": 1.72,
      "grad_norm": 12.073881149291992,
      "learning_rate": 4.195624195624196e-06,
      "loss": 0.6477,
      "step": 8020
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.389998435974121,
      "learning_rate": 4.192406692406693e-06,
      "loss": 0.4602,
      "step": 8021
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.58743667602539,
      "learning_rate": 4.189189189189189e-06,
      "loss": 0.6298,
      "step": 8022
    },
    {
      "epoch": 1.72,
      "grad_norm": 14.465950012207031,
      "learning_rate": 4.185971685971686e-06,
      "loss": 0.7491,
      "step": 8023
    },
    {
      "epoch": 1.72,
      "grad_norm": 12.178671836853027,
      "learning_rate": 4.182754182754183e-06,
      "loss": 0.7141,
      "step": 8024
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.032687187194824,
      "learning_rate": 4.17953667953668e-06,
      "loss": 0.5621,
      "step": 8025
    },
    {
      "epoch": 1.72,
      "grad_norm": 8.058003425598145,
      "learning_rate": 4.176319176319176e-06,
      "loss": 0.4883,
      "step": 8026
    },
    {
      "epoch": 1.72,
      "grad_norm": 10.171767234802246,
      "learning_rate": 4.173101673101673e-06,
      "loss": 0.6189,
      "step": 8027
    },
    {
      "epoch": 1.72,
      "grad_norm": 15.620866775512695,
      "learning_rate": 4.1698841698841696e-06,
      "loss": 0.6275,
      "step": 8028
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.988981246948242,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.5425,
      "step": 8029
    },
    {
      "epoch": 1.72,
      "grad_norm": 7.5714263916015625,
      "learning_rate": 4.163449163449164e-06,
      "loss": 0.4198,
      "step": 8030
    },
    {
      "epoch": 1.72,
      "grad_norm": 8.608209609985352,
      "learning_rate": 4.16023166023166e-06,
      "loss": 0.4117,
      "step": 8031
    },
    {
      "epoch": 1.72,
      "grad_norm": 12.086560249328613,
      "learning_rate": 4.157014157014157e-06,
      "loss": 0.6178,
      "step": 8032
    },
    {
      "epoch": 1.72,
      "grad_norm": 11.18484115600586,
      "learning_rate": 4.153796653796654e-06,
      "loss": 0.6107,
      "step": 8033
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.109747886657715,
      "learning_rate": 4.150579150579151e-06,
      "loss": 0.385,
      "step": 8034
    },
    {
      "epoch": 1.72,
      "grad_norm": 13.753311157226562,
      "learning_rate": 4.147361647361647e-06,
      "loss": 0.5456,
      "step": 8035
    },
    {
      "epoch": 1.72,
      "grad_norm": 14.823214530944824,
      "learning_rate": 4.144144144144144e-06,
      "loss": 0.5137,
      "step": 8036
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.00567626953125,
      "learning_rate": 4.140926640926641e-06,
      "loss": 0.5585,
      "step": 8037
    },
    {
      "epoch": 1.72,
      "grad_norm": 10.799083709716797,
      "learning_rate": 4.137709137709138e-06,
      "loss": 0.4249,
      "step": 8038
    },
    {
      "epoch": 1.72,
      "grad_norm": 13.49710464477539,
      "learning_rate": 4.134491634491634e-06,
      "loss": 0.6268,
      "step": 8039
    },
    {
      "epoch": 1.72,
      "grad_norm": 7.619645595550537,
      "learning_rate": 4.131274131274131e-06,
      "loss": 0.3204,
      "step": 8040
    },
    {
      "epoch": 1.72,
      "grad_norm": 13.287687301635742,
      "learning_rate": 4.1280566280566285e-06,
      "loss": 0.5829,
      "step": 8041
    },
    {
      "epoch": 1.73,
      "grad_norm": 11.000640869140625,
      "learning_rate": 4.124839124839125e-06,
      "loss": 0.6146,
      "step": 8042
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.801176071166992,
      "learning_rate": 4.121621621621622e-06,
      "loss": 0.381,
      "step": 8043
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.29078197479248,
      "learning_rate": 4.118404118404118e-06,
      "loss": 0.3459,
      "step": 8044
    },
    {
      "epoch": 1.73,
      "grad_norm": 13.89846420288086,
      "learning_rate": 4.115186615186615e-06,
      "loss": 0.3952,
      "step": 8045
    },
    {
      "epoch": 1.73,
      "grad_norm": 8.173035621643066,
      "learning_rate": 4.1119691119691125e-06,
      "loss": 0.3148,
      "step": 8046
    },
    {
      "epoch": 1.73,
      "grad_norm": 11.110279083251953,
      "learning_rate": 4.108751608751609e-06,
      "loss": 0.6142,
      "step": 8047
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.56863784790039,
      "learning_rate": 4.105534105534105e-06,
      "loss": 0.4695,
      "step": 8048
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.531314849853516,
      "learning_rate": 4.102316602316603e-06,
      "loss": 0.5813,
      "step": 8049
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.622945785522461,
      "learning_rate": 4.099099099099099e-06,
      "loss": 0.5132,
      "step": 8050
    },
    {
      "epoch": 1.73,
      "grad_norm": 8.074546813964844,
      "learning_rate": 4.095881595881596e-06,
      "loss": 0.3602,
      "step": 8051
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.626513481140137,
      "learning_rate": 4.092664092664092e-06,
      "loss": 0.4413,
      "step": 8052
    },
    {
      "epoch": 1.73,
      "grad_norm": 8.07033920288086,
      "learning_rate": 4.089446589446589e-06,
      "loss": 0.2649,
      "step": 8053
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.709245681762695,
      "learning_rate": 4.086229086229087e-06,
      "loss": 0.4491,
      "step": 8054
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.012483596801758,
      "learning_rate": 4.083011583011583e-06,
      "loss": 0.3524,
      "step": 8055
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.469477653503418,
      "learning_rate": 4.07979407979408e-06,
      "loss": 0.3638,
      "step": 8056
    },
    {
      "epoch": 1.73,
      "grad_norm": 11.33372688293457,
      "learning_rate": 4.076576576576576e-06,
      "loss": 0.5259,
      "step": 8057
    },
    {
      "epoch": 1.73,
      "grad_norm": 12.162857055664062,
      "learning_rate": 4.073359073359073e-06,
      "loss": 0.722,
      "step": 8058
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.252849578857422,
      "learning_rate": 4.0701415701415706e-06,
      "loss": 0.4961,
      "step": 8059
    },
    {
      "epoch": 1.73,
      "grad_norm": 12.254288673400879,
      "learning_rate": 4.066924066924067e-06,
      "loss": 0.5677,
      "step": 8060
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.855608940124512,
      "learning_rate": 4.0637065637065635e-06,
      "loss": 0.4634,
      "step": 8061
    },
    {
      "epoch": 1.73,
      "grad_norm": 7.556596279144287,
      "learning_rate": 4.060489060489061e-06,
      "loss": 0.4891,
      "step": 8062
    },
    {
      "epoch": 1.73,
      "grad_norm": 7.868830680847168,
      "learning_rate": 4.057271557271557e-06,
      "loss": 0.2731,
      "step": 8063
    },
    {
      "epoch": 1.73,
      "grad_norm": 11.826692581176758,
      "learning_rate": 4.0540540540540545e-06,
      "loss": 0.7742,
      "step": 8064
    },
    {
      "epoch": 1.73,
      "grad_norm": 12.199031829833984,
      "learning_rate": 4.05083655083655e-06,
      "loss": 0.5504,
      "step": 8065
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.35749340057373,
      "learning_rate": 4.0476190476190474e-06,
      "loss": 0.6364,
      "step": 8066
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.561362266540527,
      "learning_rate": 4.044401544401545e-06,
      "loss": 0.3031,
      "step": 8067
    },
    {
      "epoch": 1.73,
      "grad_norm": 11.11534309387207,
      "learning_rate": 4.041184041184041e-06,
      "loss": 0.368,
      "step": 8068
    },
    {
      "epoch": 1.73,
      "grad_norm": 12.908550262451172,
      "learning_rate": 4.0379665379665385e-06,
      "loss": 0.6579,
      "step": 8069
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.897577285766602,
      "learning_rate": 4.034749034749035e-06,
      "loss": 0.6078,
      "step": 8070
    },
    {
      "epoch": 1.73,
      "grad_norm": 11.587203979492188,
      "learning_rate": 4.031531531531531e-06,
      "loss": 0.5422,
      "step": 8071
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.441731452941895,
      "learning_rate": 4.028314028314029e-06,
      "loss": 0.6066,
      "step": 8072
    },
    {
      "epoch": 1.73,
      "grad_norm": 8.39737319946289,
      "learning_rate": 4.025096525096525e-06,
      "loss": 0.3873,
      "step": 8073
    },
    {
      "epoch": 1.73,
      "grad_norm": 8.36155891418457,
      "learning_rate": 4.021879021879022e-06,
      "loss": 0.3516,
      "step": 8074
    },
    {
      "epoch": 1.73,
      "grad_norm": 8.545485496520996,
      "learning_rate": 4.018661518661519e-06,
      "loss": 0.3812,
      "step": 8075
    },
    {
      "epoch": 1.73,
      "grad_norm": 12.520915031433105,
      "learning_rate": 4.015444015444015e-06,
      "loss": 0.6185,
      "step": 8076
    },
    {
      "epoch": 1.73,
      "grad_norm": 12.727779388427734,
      "learning_rate": 4.012226512226513e-06,
      "loss": 0.5081,
      "step": 8077
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.078700065612793,
      "learning_rate": 4.00900900900901e-06,
      "loss": 0.4957,
      "step": 8078
    },
    {
      "epoch": 1.73,
      "grad_norm": 8.31401538848877,
      "learning_rate": 4.0057915057915055e-06,
      "loss": 0.3955,
      "step": 8079
    },
    {
      "epoch": 1.73,
      "grad_norm": 10.509418487548828,
      "learning_rate": 4.002574002574003e-06,
      "loss": 0.3363,
      "step": 8080
    },
    {
      "epoch": 1.73,
      "grad_norm": 8.97143840789795,
      "learning_rate": 3.999356499356499e-06,
      "loss": 0.4542,
      "step": 8081
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.528990745544434,
      "learning_rate": 3.996138996138997e-06,
      "loss": 0.4681,
      "step": 8082
    },
    {
      "epoch": 1.73,
      "grad_norm": 11.785696983337402,
      "learning_rate": 3.992921492921493e-06,
      "loss": 0.4892,
      "step": 8083
    },
    {
      "epoch": 1.73,
      "grad_norm": 12.058844566345215,
      "learning_rate": 3.9897039897039895e-06,
      "loss": 0.6315,
      "step": 8084
    },
    {
      "epoch": 1.73,
      "grad_norm": 11.374852180480957,
      "learning_rate": 3.986486486486487e-06,
      "loss": 0.403,
      "step": 8085
    },
    {
      "epoch": 1.73,
      "grad_norm": 17.35267448425293,
      "learning_rate": 3.983268983268983e-06,
      "loss": 0.4096,
      "step": 8086
    },
    {
      "epoch": 1.73,
      "grad_norm": 12.00430965423584,
      "learning_rate": 3.98005148005148e-06,
      "loss": 0.5646,
      "step": 8087
    },
    {
      "epoch": 1.73,
      "grad_norm": 15.13976001739502,
      "learning_rate": 3.976833976833977e-06,
      "loss": 0.6286,
      "step": 8088
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.055277824401855,
      "learning_rate": 3.9736164736164734e-06,
      "loss": 0.5505,
      "step": 8089
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.017813682556152,
      "learning_rate": 3.970398970398971e-06,
      "loss": 0.4439,
      "step": 8090
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.395609855651855,
      "learning_rate": 3.967181467181468e-06,
      "loss": 0.4753,
      "step": 8091
    },
    {
      "epoch": 1.74,
      "grad_norm": 13.937660217285156,
      "learning_rate": 3.963963963963964e-06,
      "loss": 0.603,
      "step": 8092
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.821581840515137,
      "learning_rate": 3.960746460746461e-06,
      "loss": 0.4963,
      "step": 8093
    },
    {
      "epoch": 1.74,
      "grad_norm": 9.988286972045898,
      "learning_rate": 3.957528957528957e-06,
      "loss": 0.4568,
      "step": 8094
    },
    {
      "epoch": 1.74,
      "grad_norm": 8.807777404785156,
      "learning_rate": 3.954311454311455e-06,
      "loss": 0.4718,
      "step": 8095
    },
    {
      "epoch": 1.74,
      "grad_norm": 8.263309478759766,
      "learning_rate": 3.951093951093951e-06,
      "loss": 0.4303,
      "step": 8096
    },
    {
      "epoch": 1.74,
      "grad_norm": 9.849143981933594,
      "learning_rate": 3.947876447876448e-06,
      "loss": 0.5334,
      "step": 8097
    },
    {
      "epoch": 1.74,
      "grad_norm": 9.521907806396484,
      "learning_rate": 3.944658944658945e-06,
      "loss": 0.4604,
      "step": 8098
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.634574890136719,
      "learning_rate": 3.941441441441442e-06,
      "loss": 0.56,
      "step": 8099
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.032444953918457,
      "learning_rate": 3.938223938223938e-06,
      "loss": 0.469,
      "step": 8100
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.1465425491333,
      "learning_rate": 3.935006435006435e-06,
      "loss": 0.5152,
      "step": 8101
    },
    {
      "epoch": 1.74,
      "grad_norm": 9.536698341369629,
      "learning_rate": 3.9317889317889315e-06,
      "loss": 0.339,
      "step": 8102
    },
    {
      "epoch": 1.74,
      "grad_norm": 9.553712844848633,
      "learning_rate": 3.928571428571429e-06,
      "loss": 0.4442,
      "step": 8103
    },
    {
      "epoch": 1.74,
      "grad_norm": 12.123459815979004,
      "learning_rate": 3.925353925353926e-06,
      "loss": 0.61,
      "step": 8104
    },
    {
      "epoch": 1.74,
      "grad_norm": 12.532249450683594,
      "learning_rate": 3.922136422136422e-06,
      "loss": 0.5975,
      "step": 8105
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.528851509094238,
      "learning_rate": 3.918918918918919e-06,
      "loss": 0.6194,
      "step": 8106
    },
    {
      "epoch": 1.74,
      "grad_norm": 12.215927124023438,
      "learning_rate": 3.915701415701416e-06,
      "loss": 0.5276,
      "step": 8107
    },
    {
      "epoch": 1.74,
      "grad_norm": 12.2376070022583,
      "learning_rate": 3.912483912483913e-06,
      "loss": 0.5681,
      "step": 8108
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.084220886230469,
      "learning_rate": 3.909266409266409e-06,
      "loss": 0.4934,
      "step": 8109
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.441758155822754,
      "learning_rate": 3.906048906048906e-06,
      "loss": 0.4342,
      "step": 8110
    },
    {
      "epoch": 1.74,
      "grad_norm": 12.053140640258789,
      "learning_rate": 3.902831402831403e-06,
      "loss": 0.613,
      "step": 8111
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.233604431152344,
      "learning_rate": 3.8996138996139e-06,
      "loss": 0.4161,
      "step": 8112
    },
    {
      "epoch": 1.74,
      "grad_norm": 14.33393383026123,
      "learning_rate": 3.896396396396396e-06,
      "loss": 0.58,
      "step": 8113
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.465336799621582,
      "learning_rate": 3.893178893178893e-06,
      "loss": 0.3657,
      "step": 8114
    },
    {
      "epoch": 1.74,
      "grad_norm": 12.296460151672363,
      "learning_rate": 3.88996138996139e-06,
      "loss": 0.5612,
      "step": 8115
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.632782936096191,
      "learning_rate": 3.886743886743887e-06,
      "loss": 0.5269,
      "step": 8116
    },
    {
      "epoch": 1.74,
      "grad_norm": 8.830583572387695,
      "learning_rate": 3.883526383526384e-06,
      "loss": 0.4272,
      "step": 8117
    },
    {
      "epoch": 1.74,
      "grad_norm": 14.818686485290527,
      "learning_rate": 3.88030888030888e-06,
      "loss": 0.4855,
      "step": 8118
    },
    {
      "epoch": 1.74,
      "grad_norm": 8.371703147888184,
      "learning_rate": 3.877091377091377e-06,
      "loss": 0.3853,
      "step": 8119
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.066208839416504,
      "learning_rate": 3.8738738738738744e-06,
      "loss": 0.53,
      "step": 8120
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.925811767578125,
      "learning_rate": 3.870656370656371e-06,
      "loss": 0.5465,
      "step": 8121
    },
    {
      "epoch": 1.74,
      "grad_norm": 12.046127319335938,
      "learning_rate": 3.867438867438867e-06,
      "loss": 0.4227,
      "step": 8122
    },
    {
      "epoch": 1.74,
      "grad_norm": 10.19869613647461,
      "learning_rate": 3.864221364221364e-06,
      "loss": 0.6123,
      "step": 8123
    },
    {
      "epoch": 1.74,
      "grad_norm": 13.105587005615234,
      "learning_rate": 3.861003861003861e-06,
      "loss": 0.7002,
      "step": 8124
    },
    {
      "epoch": 1.74,
      "grad_norm": 9.40406322479248,
      "learning_rate": 3.857786357786358e-06,
      "loss": 0.5978,
      "step": 8125
    },
    {
      "epoch": 1.74,
      "grad_norm": 14.057782173156738,
      "learning_rate": 3.854568854568854e-06,
      "loss": 0.6004,
      "step": 8126
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.548563957214355,
      "learning_rate": 3.851351351351351e-06,
      "loss": 0.388,
      "step": 8127
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.130871772766113,
      "learning_rate": 3.848133848133849e-06,
      "loss": 0.4071,
      "step": 8128
    },
    {
      "epoch": 1.74,
      "grad_norm": 8.807221412658691,
      "learning_rate": 3.844916344916345e-06,
      "loss": 0.4865,
      "step": 8129
    },
    {
      "epoch": 1.74,
      "grad_norm": 9.292908668518066,
      "learning_rate": 3.841698841698842e-06,
      "loss": 0.3933,
      "step": 8130
    },
    {
      "epoch": 1.74,
      "grad_norm": 12.496246337890625,
      "learning_rate": 3.838481338481338e-06,
      "loss": 0.57,
      "step": 8131
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.290009498596191,
      "learning_rate": 3.835263835263835e-06,
      "loss": 0.4495,
      "step": 8132
    },
    {
      "epoch": 1.74,
      "grad_norm": 12.159529685974121,
      "learning_rate": 3.8320463320463325e-06,
      "loss": 0.5684,
      "step": 8133
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.15982723236084,
      "learning_rate": 3.828828828828829e-06,
      "loss": 0.5413,
      "step": 8134
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.333213806152344,
      "learning_rate": 3.8256113256113254e-06,
      "loss": 0.5122,
      "step": 8135
    },
    {
      "epoch": 1.75,
      "grad_norm": 8.212392807006836,
      "learning_rate": 3.822393822393823e-06,
      "loss": 0.403,
      "step": 8136
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.966144561767578,
      "learning_rate": 3.819176319176319e-06,
      "loss": 0.4495,
      "step": 8137
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.802821159362793,
      "learning_rate": 3.8159588159588165e-06,
      "loss": 0.4851,
      "step": 8138
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.790999412536621,
      "learning_rate": 3.8127413127413125e-06,
      "loss": 0.4594,
      "step": 8139
    },
    {
      "epoch": 1.75,
      "grad_norm": 11.44027042388916,
      "learning_rate": 3.8095238095238094e-06,
      "loss": 0.3966,
      "step": 8140
    },
    {
      "epoch": 1.75,
      "grad_norm": 11.55323600769043,
      "learning_rate": 3.8063063063063063e-06,
      "loss": 0.6689,
      "step": 8141
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.920072555541992,
      "learning_rate": 3.8030888030888036e-06,
      "loss": 0.596,
      "step": 8142
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.965496063232422,
      "learning_rate": 3.7998712998713004e-06,
      "loss": 0.4503,
      "step": 8143
    },
    {
      "epoch": 1.75,
      "grad_norm": 7.518810749053955,
      "learning_rate": 3.7966537966537965e-06,
      "loss": 0.3884,
      "step": 8144
    },
    {
      "epoch": 1.75,
      "grad_norm": 11.11453914642334,
      "learning_rate": 3.7934362934362933e-06,
      "loss": 0.5724,
      "step": 8145
    },
    {
      "epoch": 1.75,
      "grad_norm": 11.487732887268066,
      "learning_rate": 3.7902187902187902e-06,
      "loss": 0.5699,
      "step": 8146
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.716800689697266,
      "learning_rate": 3.7870012870012875e-06,
      "loss": 0.3941,
      "step": 8147
    },
    {
      "epoch": 1.75,
      "grad_norm": 12.183723449707031,
      "learning_rate": 3.7837837837837835e-06,
      "loss": 0.6225,
      "step": 8148
    },
    {
      "epoch": 1.75,
      "grad_norm": 13.126376152038574,
      "learning_rate": 3.7805662805662804e-06,
      "loss": 0.6551,
      "step": 8149
    },
    {
      "epoch": 1.75,
      "grad_norm": 14.146993637084961,
      "learning_rate": 3.7773487773487773e-06,
      "loss": 0.5492,
      "step": 8150
    },
    {
      "epoch": 1.75,
      "grad_norm": 11.025483131408691,
      "learning_rate": 3.7741312741312746e-06,
      "loss": 0.5855,
      "step": 8151
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.06520938873291,
      "learning_rate": 3.7709137709137706e-06,
      "loss": 0.5367,
      "step": 8152
    },
    {
      "epoch": 1.75,
      "grad_norm": 8.597920417785645,
      "learning_rate": 3.7676962676962675e-06,
      "loss": 0.3791,
      "step": 8153
    },
    {
      "epoch": 1.75,
      "grad_norm": 13.48943042755127,
      "learning_rate": 3.7644787644787644e-06,
      "loss": 0.5189,
      "step": 8154
    },
    {
      "epoch": 1.75,
      "grad_norm": 8.240830421447754,
      "learning_rate": 3.7612612612612617e-06,
      "loss": 0.4304,
      "step": 8155
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.948019027709961,
      "learning_rate": 3.7580437580437585e-06,
      "loss": 0.4871,
      "step": 8156
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.189432144165039,
      "learning_rate": 3.7548262548262546e-06,
      "loss": 0.5234,
      "step": 8157
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.513077735900879,
      "learning_rate": 3.7516087516087514e-06,
      "loss": 0.4701,
      "step": 8158
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.405010223388672,
      "learning_rate": 3.7483912483912487e-06,
      "loss": 0.5709,
      "step": 8159
    },
    {
      "epoch": 1.75,
      "grad_norm": 11.162299156188965,
      "learning_rate": 3.745173745173745e-06,
      "loss": 0.4091,
      "step": 8160
    },
    {
      "epoch": 1.75,
      "grad_norm": 11.338433265686035,
      "learning_rate": 3.741956241956242e-06,
      "loss": 0.6035,
      "step": 8161
    },
    {
      "epoch": 1.75,
      "grad_norm": 13.874722480773926,
      "learning_rate": 3.7387387387387385e-06,
      "loss": 0.6923,
      "step": 8162
    },
    {
      "epoch": 1.75,
      "grad_norm": 11.683465957641602,
      "learning_rate": 3.735521235521236e-06,
      "loss": 0.5996,
      "step": 8163
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.19886302947998,
      "learning_rate": 3.7323037323037323e-06,
      "loss": 0.5219,
      "step": 8164
    },
    {
      "epoch": 1.75,
      "grad_norm": 6.393564224243164,
      "learning_rate": 3.729086229086229e-06,
      "loss": 0.344,
      "step": 8165
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.076096534729004,
      "learning_rate": 3.725868725868726e-06,
      "loss": 0.5579,
      "step": 8166
    },
    {
      "epoch": 1.75,
      "grad_norm": 12.921758651733398,
      "learning_rate": 3.722651222651223e-06,
      "loss": 0.605,
      "step": 8167
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.202638626098633,
      "learning_rate": 3.7194337194337198e-06,
      "loss": 0.611,
      "step": 8168
    },
    {
      "epoch": 1.75,
      "grad_norm": 13.379653930664062,
      "learning_rate": 3.7162162162162162e-06,
      "loss": 0.6247,
      "step": 8169
    },
    {
      "epoch": 1.75,
      "grad_norm": 12.139145851135254,
      "learning_rate": 3.712998712998713e-06,
      "loss": 0.6011,
      "step": 8170
    },
    {
      "epoch": 1.75,
      "grad_norm": 12.054837226867676,
      "learning_rate": 3.70978120978121e-06,
      "loss": 0.5049,
      "step": 8171
    },
    {
      "epoch": 1.75,
      "grad_norm": 7.529928684234619,
      "learning_rate": 3.706563706563707e-06,
      "loss": 0.4135,
      "step": 8172
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.047761917114258,
      "learning_rate": 3.7033462033462033e-06,
      "loss": 0.5239,
      "step": 8173
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.384648323059082,
      "learning_rate": 3.7001287001287e-06,
      "loss": 0.5193,
      "step": 8174
    },
    {
      "epoch": 1.75,
      "grad_norm": 12.8467378616333,
      "learning_rate": 3.6969111969111966e-06,
      "loss": 0.4939,
      "step": 8175
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.29322338104248,
      "learning_rate": 3.693693693693694e-06,
      "loss": 0.3469,
      "step": 8176
    },
    {
      "epoch": 1.75,
      "grad_norm": 14.18348217010498,
      "learning_rate": 3.6904761904761904e-06,
      "loss": 0.5351,
      "step": 8177
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.526959419250488,
      "learning_rate": 3.6872586872586872e-06,
      "loss": 0.3623,
      "step": 8178
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.99173355102539,
      "learning_rate": 3.684041184041184e-06,
      "loss": 0.5126,
      "step": 8179
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.3534517288208,
      "learning_rate": 3.680823680823681e-06,
      "loss": 0.4265,
      "step": 8180
    },
    {
      "epoch": 1.75,
      "grad_norm": 16.86173439025879,
      "learning_rate": 3.677606177606178e-06,
      "loss": 0.7683,
      "step": 8181
    },
    {
      "epoch": 1.76,
      "grad_norm": 7.505889892578125,
      "learning_rate": 3.6743886743886743e-06,
      "loss": 0.3683,
      "step": 8182
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.1529541015625,
      "learning_rate": 3.671171171171171e-06,
      "loss": 0.498,
      "step": 8183
    },
    {
      "epoch": 1.76,
      "grad_norm": 13.449538230895996,
      "learning_rate": 3.667953667953668e-06,
      "loss": 0.9558,
      "step": 8184
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.762483596801758,
      "learning_rate": 3.664736164736165e-06,
      "loss": 0.6729,
      "step": 8185
    },
    {
      "epoch": 1.76,
      "grad_norm": 13.589247703552246,
      "learning_rate": 3.6615186615186614e-06,
      "loss": 0.7342,
      "step": 8186
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.97815990447998,
      "learning_rate": 3.6583011583011583e-06,
      "loss": 0.4687,
      "step": 8187
    },
    {
      "epoch": 1.76,
      "grad_norm": 7.821135520935059,
      "learning_rate": 3.655083655083655e-06,
      "loss": 0.3882,
      "step": 8188
    },
    {
      "epoch": 1.76,
      "grad_norm": 8.569605827331543,
      "learning_rate": 3.651866151866152e-06,
      "loss": 0.4311,
      "step": 8189
    },
    {
      "epoch": 1.76,
      "grad_norm": 7.116994380950928,
      "learning_rate": 3.648648648648649e-06,
      "loss": 0.3689,
      "step": 8190
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.895792007446289,
      "learning_rate": 3.6454311454311454e-06,
      "loss": 0.4529,
      "step": 8191
    },
    {
      "epoch": 1.76,
      "grad_norm": 8.312883377075195,
      "learning_rate": 3.6422136422136427e-06,
      "loss": 0.4299,
      "step": 8192
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.650344848632812,
      "learning_rate": 3.638996138996139e-06,
      "loss": 0.3772,
      "step": 8193
    },
    {
      "epoch": 1.76,
      "grad_norm": 13.951356887817383,
      "learning_rate": 3.635778635778636e-06,
      "loss": 0.4446,
      "step": 8194
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.98641586303711,
      "learning_rate": 3.6325611325611324e-06,
      "loss": 0.5268,
      "step": 8195
    },
    {
      "epoch": 1.76,
      "grad_norm": 8.414530754089355,
      "learning_rate": 3.6293436293436297e-06,
      "loss": 0.3444,
      "step": 8196
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.540217399597168,
      "learning_rate": 3.626126126126126e-06,
      "loss": 0.5668,
      "step": 8197
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.602180480957031,
      "learning_rate": 3.622908622908623e-06,
      "loss": 0.3782,
      "step": 8198
    },
    {
      "epoch": 1.76,
      "grad_norm": 12.154678344726562,
      "learning_rate": 3.6196911196911195e-06,
      "loss": 0.6832,
      "step": 8199
    },
    {
      "epoch": 1.76,
      "grad_norm": 12.5065279006958,
      "learning_rate": 3.616473616473617e-06,
      "loss": 0.6987,
      "step": 8200
    },
    {
      "epoch": 1.76,
      "grad_norm": 12.17429256439209,
      "learning_rate": 3.6132561132561133e-06,
      "loss": 0.4666,
      "step": 8201
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.774383544921875,
      "learning_rate": 3.61003861003861e-06,
      "loss": 0.3692,
      "step": 8202
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.123200416564941,
      "learning_rate": 3.606821106821107e-06,
      "loss": 0.4879,
      "step": 8203
    },
    {
      "epoch": 1.76,
      "grad_norm": 12.735798835754395,
      "learning_rate": 3.6036036036036035e-06,
      "loss": 0.5673,
      "step": 8204
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.939860343933105,
      "learning_rate": 3.6003861003861008e-06,
      "loss": 0.6246,
      "step": 8205
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.021197319030762,
      "learning_rate": 3.597168597168597e-06,
      "loss": 0.4706,
      "step": 8206
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.095907211303711,
      "learning_rate": 3.593951093951094e-06,
      "loss": 0.4519,
      "step": 8207
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.869614601135254,
      "learning_rate": 3.5907335907335905e-06,
      "loss": 0.5213,
      "step": 8208
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.859158515930176,
      "learning_rate": 3.587516087516088e-06,
      "loss": 0.6959,
      "step": 8209
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.450126647949219,
      "learning_rate": 3.5842985842985843e-06,
      "loss": 0.3796,
      "step": 8210
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.636703491210938,
      "learning_rate": 3.581081081081081e-06,
      "loss": 0.3977,
      "step": 8211
    },
    {
      "epoch": 1.76,
      "grad_norm": 8.848400115966797,
      "learning_rate": 3.5778635778635776e-06,
      "loss": 0.4553,
      "step": 8212
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.78817081451416,
      "learning_rate": 3.574646074646075e-06,
      "loss": 0.5094,
      "step": 8213
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.529473304748535,
      "learning_rate": 3.5714285714285714e-06,
      "loss": 0.5004,
      "step": 8214
    },
    {
      "epoch": 1.76,
      "grad_norm": 12.831048011779785,
      "learning_rate": 3.5682110682110682e-06,
      "loss": 0.4601,
      "step": 8215
    },
    {
      "epoch": 1.76,
      "grad_norm": 13.39727783203125,
      "learning_rate": 3.564993564993565e-06,
      "loss": 0.7146,
      "step": 8216
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.495378494262695,
      "learning_rate": 3.561776061776062e-06,
      "loss": 0.3716,
      "step": 8217
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.259482383728027,
      "learning_rate": 3.558558558558559e-06,
      "loss": 0.4051,
      "step": 8218
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.762388229370117,
      "learning_rate": 3.5553410553410553e-06,
      "loss": 0.5224,
      "step": 8219
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.193269729614258,
      "learning_rate": 3.552123552123552e-06,
      "loss": 0.5014,
      "step": 8220
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.660438537597656,
      "learning_rate": 3.548906048906049e-06,
      "loss": 0.4227,
      "step": 8221
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.402886390686035,
      "learning_rate": 3.545688545688546e-06,
      "loss": 0.5878,
      "step": 8222
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.500566482543945,
      "learning_rate": 3.5424710424710424e-06,
      "loss": 0.5344,
      "step": 8223
    },
    {
      "epoch": 1.76,
      "grad_norm": 8.738072395324707,
      "learning_rate": 3.5392535392535393e-06,
      "loss": 0.4536,
      "step": 8224
    },
    {
      "epoch": 1.76,
      "grad_norm": 11.968039512634277,
      "learning_rate": 3.536036036036036e-06,
      "loss": 0.4631,
      "step": 8225
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.105412483215332,
      "learning_rate": 3.532818532818533e-06,
      "loss": 0.3435,
      "step": 8226
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.321816444396973,
      "learning_rate": 3.5296010296010295e-06,
      "loss": 0.4637,
      "step": 8227
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.33018684387207,
      "learning_rate": 3.5263835263835263e-06,
      "loss": 0.4493,
      "step": 8228
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.553813934326172,
      "learning_rate": 3.5231660231660236e-06,
      "loss": 0.6658,
      "step": 8229
    },
    {
      "epoch": 1.77,
      "grad_norm": 8.254966735839844,
      "learning_rate": 3.51994851994852e-06,
      "loss": 0.4448,
      "step": 8230
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.87977409362793,
      "learning_rate": 3.516731016731017e-06,
      "loss": 0.5509,
      "step": 8231
    },
    {
      "epoch": 1.77,
      "grad_norm": 8.49821662902832,
      "learning_rate": 3.5135135135135134e-06,
      "loss": 0.3677,
      "step": 8232
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.444320678710938,
      "learning_rate": 3.5102960102960103e-06,
      "loss": 0.4308,
      "step": 8233
    },
    {
      "epoch": 1.77,
      "grad_norm": 13.338600158691406,
      "learning_rate": 3.507078507078507e-06,
      "loss": 0.4598,
      "step": 8234
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.018606185913086,
      "learning_rate": 3.503861003861004e-06,
      "loss": 0.4494,
      "step": 8235
    },
    {
      "epoch": 1.77,
      "grad_norm": 12.159656524658203,
      "learning_rate": 3.5006435006435005e-06,
      "loss": 0.4307,
      "step": 8236
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.39559268951416,
      "learning_rate": 3.4974259974259974e-06,
      "loss": 0.6583,
      "step": 8237
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.229377746582031,
      "learning_rate": 3.4942084942084942e-06,
      "loss": 0.3617,
      "step": 8238
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.09165096282959,
      "learning_rate": 3.490990990990991e-06,
      "loss": 0.6299,
      "step": 8239
    },
    {
      "epoch": 1.77,
      "grad_norm": 14.104784965515137,
      "learning_rate": 3.487773487773488e-06,
      "loss": 0.6659,
      "step": 8240
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.616721153259277,
      "learning_rate": 3.4845559845559844e-06,
      "loss": 0.4577,
      "step": 8241
    },
    {
      "epoch": 1.77,
      "grad_norm": 7.409374237060547,
      "learning_rate": 3.4813384813384817e-06,
      "loss": 0.3079,
      "step": 8242
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.42624568939209,
      "learning_rate": 3.478120978120978e-06,
      "loss": 0.3486,
      "step": 8243
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.154046058654785,
      "learning_rate": 3.474903474903475e-06,
      "loss": 0.3656,
      "step": 8244
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.291521072387695,
      "learning_rate": 3.4716859716859715e-06,
      "loss": 0.5627,
      "step": 8245
    },
    {
      "epoch": 1.77,
      "grad_norm": 8.830683708190918,
      "learning_rate": 3.468468468468469e-06,
      "loss": 0.4665,
      "step": 8246
    },
    {
      "epoch": 1.77,
      "grad_norm": 7.0438971519470215,
      "learning_rate": 3.4652509652509653e-06,
      "loss": 0.2567,
      "step": 8247
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.74195671081543,
      "learning_rate": 3.462033462033462e-06,
      "loss": 0.5203,
      "step": 8248
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.541772842407227,
      "learning_rate": 3.4588159588159586e-06,
      "loss": 0.5629,
      "step": 8249
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.913785934448242,
      "learning_rate": 3.455598455598456e-06,
      "loss": 0.4216,
      "step": 8250
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.634805679321289,
      "learning_rate": 3.4523809523809523e-06,
      "loss": 0.7038,
      "step": 8251
    },
    {
      "epoch": 1.77,
      "grad_norm": 12.749835014343262,
      "learning_rate": 3.4491634491634492e-06,
      "loss": 0.6362,
      "step": 8252
    },
    {
      "epoch": 1.77,
      "grad_norm": 8.660198211669922,
      "learning_rate": 3.445945945945946e-06,
      "loss": 0.3286,
      "step": 8253
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.608009338378906,
      "learning_rate": 3.442728442728443e-06,
      "loss": 0.616,
      "step": 8254
    },
    {
      "epoch": 1.77,
      "grad_norm": 12.631658554077148,
      "learning_rate": 3.43951093951094e-06,
      "loss": 0.7883,
      "step": 8255
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.320611953735352,
      "learning_rate": 3.4362934362934363e-06,
      "loss": 0.3995,
      "step": 8256
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.26423168182373,
      "learning_rate": 3.433075933075933e-06,
      "loss": 0.4485,
      "step": 8257
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.075340270996094,
      "learning_rate": 3.42985842985843e-06,
      "loss": 0.5528,
      "step": 8258
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.501089096069336,
      "learning_rate": 3.426640926640927e-06,
      "loss": 0.7313,
      "step": 8259
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.783011436462402,
      "learning_rate": 3.4234234234234234e-06,
      "loss": 0.379,
      "step": 8260
    },
    {
      "epoch": 1.77,
      "grad_norm": 8.496153831481934,
      "learning_rate": 3.4202059202059202e-06,
      "loss": 0.4471,
      "step": 8261
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.836524963378906,
      "learning_rate": 3.416988416988417e-06,
      "loss": 0.5287,
      "step": 8262
    },
    {
      "epoch": 1.77,
      "grad_norm": 12.604878425598145,
      "learning_rate": 3.413770913770914e-06,
      "loss": 0.4216,
      "step": 8263
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.242387771606445,
      "learning_rate": 3.4105534105534104e-06,
      "loss": 0.5143,
      "step": 8264
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.812342643737793,
      "learning_rate": 3.4073359073359073e-06,
      "loss": 0.5515,
      "step": 8265
    },
    {
      "epoch": 1.77,
      "grad_norm": 9.082261085510254,
      "learning_rate": 3.404118404118404e-06,
      "loss": 0.7362,
      "step": 8266
    },
    {
      "epoch": 1.77,
      "grad_norm": 12.194621086120605,
      "learning_rate": 3.400900900900901e-06,
      "loss": 0.6251,
      "step": 8267
    },
    {
      "epoch": 1.77,
      "grad_norm": 18.926837921142578,
      "learning_rate": 3.397683397683398e-06,
      "loss": 0.4985,
      "step": 8268
    },
    {
      "epoch": 1.77,
      "grad_norm": 12.70247745513916,
      "learning_rate": 3.3944658944658944e-06,
      "loss": 0.5038,
      "step": 8269
    },
    {
      "epoch": 1.77,
      "grad_norm": 12.105090141296387,
      "learning_rate": 3.3912483912483913e-06,
      "loss": 0.5259,
      "step": 8270
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.476213455200195,
      "learning_rate": 3.388030888030888e-06,
      "loss": 0.3504,
      "step": 8271
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.434536933898926,
      "learning_rate": 3.384813384813385e-06,
      "loss": 0.537,
      "step": 8272
    },
    {
      "epoch": 1.77,
      "grad_norm": 12.246953964233398,
      "learning_rate": 3.3815958815958815e-06,
      "loss": 0.6622,
      "step": 8273
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.432859420776367,
      "learning_rate": 3.3783783783783783e-06,
      "loss": 0.6482,
      "step": 8274
    },
    {
      "epoch": 1.77,
      "grad_norm": 11.669441223144531,
      "learning_rate": 3.3751608751608752e-06,
      "loss": 0.6138,
      "step": 8275
    },
    {
      "epoch": 1.78,
      "grad_norm": 12.057372093200684,
      "learning_rate": 3.371943371943372e-06,
      "loss": 0.862,
      "step": 8276
    },
    {
      "epoch": 1.78,
      "grad_norm": 9.672422409057617,
      "learning_rate": 3.368725868725869e-06,
      "loss": 0.4008,
      "step": 8277
    },
    {
      "epoch": 1.78,
      "grad_norm": 9.309264183044434,
      "learning_rate": 3.3655083655083654e-06,
      "loss": 0.4888,
      "step": 8278
    },
    {
      "epoch": 1.78,
      "grad_norm": 8.793511390686035,
      "learning_rate": 3.3622908622908627e-06,
      "loss": 0.429,
      "step": 8279
    },
    {
      "epoch": 1.78,
      "grad_norm": 12.486104011535645,
      "learning_rate": 3.359073359073359e-06,
      "loss": 0.5744,
      "step": 8280
    },
    {
      "epoch": 1.78,
      "grad_norm": 15.681099891662598,
      "learning_rate": 3.355855855855856e-06,
      "loss": 0.8148,
      "step": 8281
    },
    {
      "epoch": 1.78,
      "grad_norm": 12.470253944396973,
      "learning_rate": 3.3526383526383525e-06,
      "loss": 0.5671,
      "step": 8282
    },
    {
      "epoch": 1.78,
      "grad_norm": 14.722732543945312,
      "learning_rate": 3.34942084942085e-06,
      "loss": 0.4554,
      "step": 8283
    },
    {
      "epoch": 1.78,
      "grad_norm": 11.419551849365234,
      "learning_rate": 3.3462033462033462e-06,
      "loss": 0.5802,
      "step": 8284
    },
    {
      "epoch": 1.78,
      "grad_norm": 9.442078590393066,
      "learning_rate": 3.342985842985843e-06,
      "loss": 0.5834,
      "step": 8285
    },
    {
      "epoch": 1.78,
      "grad_norm": 8.4558744430542,
      "learning_rate": 3.3397683397683396e-06,
      "loss": 0.3086,
      "step": 8286
    },
    {
      "epoch": 1.78,
      "grad_norm": 17.89988136291504,
      "learning_rate": 3.336550836550837e-06,
      "loss": 0.5322,
      "step": 8287
    },
    {
      "epoch": 1.78,
      "grad_norm": 11.065286636352539,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.6877,
      "step": 8288
    },
    {
      "epoch": 1.78,
      "grad_norm": 9.102602005004883,
      "learning_rate": 3.33011583011583e-06,
      "loss": 0.5593,
      "step": 8289
    },
    {
      "epoch": 1.78,
      "grad_norm": 11.435242652893066,
      "learning_rate": 3.326898326898327e-06,
      "loss": 0.616,
      "step": 8290
    },
    {
      "epoch": 1.78,
      "grad_norm": 8.778029441833496,
      "learning_rate": 3.323680823680824e-06,
      "loss": 0.4688,
      "step": 8291
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.65153980255127,
      "learning_rate": 3.320463320463321e-06,
      "loss": 0.3267,
      "step": 8292
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.560384750366211,
      "learning_rate": 3.3172458172458173e-06,
      "loss": 0.5665,
      "step": 8293
    },
    {
      "epoch": 1.78,
      "grad_norm": 9.86024284362793,
      "learning_rate": 3.314028314028314e-06,
      "loss": 0.4914,
      "step": 8294
    },
    {
      "epoch": 1.78,
      "grad_norm": 13.702964782714844,
      "learning_rate": 3.3108108108108106e-06,
      "loss": 0.6452,
      "step": 8295
    },
    {
      "epoch": 1.78,
      "grad_norm": 12.044875144958496,
      "learning_rate": 3.307593307593308e-06,
      "loss": 0.5187,
      "step": 8296
    },
    {
      "epoch": 1.78,
      "grad_norm": 14.659886360168457,
      "learning_rate": 3.3043758043758044e-06,
      "loss": 0.7134,
      "step": 8297
    },
    {
      "epoch": 1.78,
      "grad_norm": 8.775678634643555,
      "learning_rate": 3.3011583011583012e-06,
      "loss": 0.492,
      "step": 8298
    },
    {
      "epoch": 1.78,
      "grad_norm": 11.269683837890625,
      "learning_rate": 3.2979407979407977e-06,
      "loss": 0.5952,
      "step": 8299
    },
    {
      "epoch": 1.78,
      "grad_norm": 11.398768424987793,
      "learning_rate": 3.294723294723295e-06,
      "loss": 0.3877,
      "step": 8300
    },
    {
      "epoch": 1.78,
      "grad_norm": 12.01357364654541,
      "learning_rate": 3.2915057915057914e-06,
      "loss": 0.7035,
      "step": 8301
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.588897705078125,
      "learning_rate": 3.2882882882882883e-06,
      "loss": 0.5356,
      "step": 8302
    },
    {
      "epoch": 1.78,
      "grad_norm": 14.476763725280762,
      "learning_rate": 3.285070785070785e-06,
      "loss": 0.7233,
      "step": 8303
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.006732940673828,
      "learning_rate": 3.281853281853282e-06,
      "loss": 0.4313,
      "step": 8304
    },
    {
      "epoch": 1.78,
      "grad_norm": 9.225655555725098,
      "learning_rate": 3.278635778635779e-06,
      "loss": 0.5286,
      "step": 8305
    },
    {
      "epoch": 1.78,
      "grad_norm": 11.415725708007812,
      "learning_rate": 3.2754182754182754e-06,
      "loss": 0.5484,
      "step": 8306
    },
    {
      "epoch": 1.78,
      "grad_norm": 11.0482816696167,
      "learning_rate": 3.2722007722007723e-06,
      "loss": 0.5223,
      "step": 8307
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.278223037719727,
      "learning_rate": 3.268983268983269e-06,
      "loss": 0.3546,
      "step": 8308
    },
    {
      "epoch": 1.78,
      "grad_norm": 12.027695655822754,
      "learning_rate": 3.265765765765766e-06,
      "loss": 0.6671,
      "step": 8309
    },
    {
      "epoch": 1.78,
      "grad_norm": 13.154679298400879,
      "learning_rate": 3.2625482625482625e-06,
      "loss": 0.5301,
      "step": 8310
    },
    {
      "epoch": 1.78,
      "grad_norm": 8.986579895019531,
      "learning_rate": 3.2593307593307593e-06,
      "loss": 0.4772,
      "step": 8311
    },
    {
      "epoch": 1.78,
      "grad_norm": 7.196601867675781,
      "learning_rate": 3.256113256113256e-06,
      "loss": 0.3855,
      "step": 8312
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.765398979187012,
      "learning_rate": 3.252895752895753e-06,
      "loss": 0.5899,
      "step": 8313
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.920270919799805,
      "learning_rate": 3.2496782496782495e-06,
      "loss": 0.6031,
      "step": 8314
    },
    {
      "epoch": 1.78,
      "grad_norm": 11.769896507263184,
      "learning_rate": 3.2464607464607464e-06,
      "loss": 0.5846,
      "step": 8315
    },
    {
      "epoch": 1.78,
      "grad_norm": 13.38318157196045,
      "learning_rate": 3.2432432432432437e-06,
      "loss": 0.5294,
      "step": 8316
    },
    {
      "epoch": 1.78,
      "grad_norm": 11.738441467285156,
      "learning_rate": 3.24002574002574e-06,
      "loss": 0.4627,
      "step": 8317
    },
    {
      "epoch": 1.78,
      "grad_norm": 7.241338729858398,
      "learning_rate": 3.236808236808237e-06,
      "loss": 0.3429,
      "step": 8318
    },
    {
      "epoch": 1.78,
      "grad_norm": 12.934293746948242,
      "learning_rate": 3.2335907335907335e-06,
      "loss": 0.4683,
      "step": 8319
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.170077323913574,
      "learning_rate": 3.2303732303732308e-06,
      "loss": 0.447,
      "step": 8320
    },
    {
      "epoch": 1.78,
      "grad_norm": 9.37217903137207,
      "learning_rate": 3.2271557271557272e-06,
      "loss": 0.4833,
      "step": 8321
    },
    {
      "epoch": 1.79,
      "grad_norm": 12.636224746704102,
      "learning_rate": 3.223938223938224e-06,
      "loss": 0.4928,
      "step": 8322
    },
    {
      "epoch": 1.79,
      "grad_norm": 11.352656364440918,
      "learning_rate": 3.2207207207207206e-06,
      "loss": 0.5154,
      "step": 8323
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.767221450805664,
      "learning_rate": 3.2175032175032174e-06,
      "loss": 0.4719,
      "step": 8324
    },
    {
      "epoch": 1.79,
      "grad_norm": 13.110780715942383,
      "learning_rate": 3.2142857142857143e-06,
      "loss": 0.5262,
      "step": 8325
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.118586540222168,
      "learning_rate": 3.211068211068211e-06,
      "loss": 0.4424,
      "step": 8326
    },
    {
      "epoch": 1.79,
      "grad_norm": 12.527872085571289,
      "learning_rate": 3.207850707850708e-06,
      "loss": 0.4982,
      "step": 8327
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.256003379821777,
      "learning_rate": 3.2046332046332045e-06,
      "loss": 0.5424,
      "step": 8328
    },
    {
      "epoch": 1.79,
      "grad_norm": 12.207365989685059,
      "learning_rate": 3.201415701415702e-06,
      "loss": 0.539,
      "step": 8329
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.062939643859863,
      "learning_rate": 3.1981981981981983e-06,
      "loss": 0.5349,
      "step": 8330
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.929707527160645,
      "learning_rate": 3.194980694980695e-06,
      "loss": 0.5269,
      "step": 8331
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.962636947631836,
      "learning_rate": 3.1917631917631916e-06,
      "loss": 0.4156,
      "step": 8332
    },
    {
      "epoch": 1.79,
      "grad_norm": 12.183863639831543,
      "learning_rate": 3.188545688545689e-06,
      "loss": 0.6291,
      "step": 8333
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.36735725402832,
      "learning_rate": 3.1853281853281853e-06,
      "loss": 0.4798,
      "step": 8334
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.260275840759277,
      "learning_rate": 3.182110682110682e-06,
      "loss": 0.5255,
      "step": 8335
    },
    {
      "epoch": 1.79,
      "grad_norm": 13.68548583984375,
      "learning_rate": 3.1788931788931787e-06,
      "loss": 0.6489,
      "step": 8336
    },
    {
      "epoch": 1.79,
      "grad_norm": 8.29008960723877,
      "learning_rate": 3.175675675675676e-06,
      "loss": 0.4693,
      "step": 8337
    },
    {
      "epoch": 1.79,
      "grad_norm": 13.55160140991211,
      "learning_rate": 3.1724581724581724e-06,
      "loss": 0.6929,
      "step": 8338
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.353275299072266,
      "learning_rate": 3.1692406692406693e-06,
      "loss": 0.4187,
      "step": 8339
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.241353988647461,
      "learning_rate": 3.166023166023166e-06,
      "loss": 0.3512,
      "step": 8340
    },
    {
      "epoch": 1.79,
      "grad_norm": 12.551490783691406,
      "learning_rate": 3.162805662805663e-06,
      "loss": 0.5323,
      "step": 8341
    },
    {
      "epoch": 1.79,
      "grad_norm": 12.637453079223633,
      "learning_rate": 3.15958815958816e-06,
      "loss": 0.7078,
      "step": 8342
    },
    {
      "epoch": 1.79,
      "grad_norm": 11.622503280639648,
      "learning_rate": 3.1563706563706564e-06,
      "loss": 0.3721,
      "step": 8343
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.51010513305664,
      "learning_rate": 3.1531531531531532e-06,
      "loss": 0.5928,
      "step": 8344
    },
    {
      "epoch": 1.79,
      "grad_norm": 11.241978645324707,
      "learning_rate": 3.14993564993565e-06,
      "loss": 0.4454,
      "step": 8345
    },
    {
      "epoch": 1.79,
      "grad_norm": 11.276993751525879,
      "learning_rate": 3.146718146718147e-06,
      "loss": 0.4908,
      "step": 8346
    },
    {
      "epoch": 1.79,
      "grad_norm": 8.993622779846191,
      "learning_rate": 3.1435006435006434e-06,
      "loss": 0.3006,
      "step": 8347
    },
    {
      "epoch": 1.79,
      "grad_norm": 8.52222728729248,
      "learning_rate": 3.1402831402831403e-06,
      "loss": 0.4497,
      "step": 8348
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.101472854614258,
      "learning_rate": 3.137065637065637e-06,
      "loss": 0.5094,
      "step": 8349
    },
    {
      "epoch": 1.79,
      "grad_norm": 8.48009967803955,
      "learning_rate": 3.133848133848134e-06,
      "loss": 0.5751,
      "step": 8350
    },
    {
      "epoch": 1.79,
      "grad_norm": 7.943315505981445,
      "learning_rate": 3.1306306306306305e-06,
      "loss": 0.3467,
      "step": 8351
    },
    {
      "epoch": 1.79,
      "grad_norm": 8.791330337524414,
      "learning_rate": 3.1274131274131274e-06,
      "loss": 0.4956,
      "step": 8352
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.752552032470703,
      "learning_rate": 3.1241956241956243e-06,
      "loss": 0.5338,
      "step": 8353
    },
    {
      "epoch": 1.79,
      "grad_norm": 12.993391990661621,
      "learning_rate": 3.120978120978121e-06,
      "loss": 0.6394,
      "step": 8354
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.17874813079834,
      "learning_rate": 3.117760617760618e-06,
      "loss": 0.4414,
      "step": 8355
    },
    {
      "epoch": 1.79,
      "grad_norm": 12.004009246826172,
      "learning_rate": 3.1145431145431145e-06,
      "loss": 0.3814,
      "step": 8356
    },
    {
      "epoch": 1.79,
      "grad_norm": 12.269937515258789,
      "learning_rate": 3.1113256113256113e-06,
      "loss": 0.8436,
      "step": 8357
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.172966957092285,
      "learning_rate": 3.1081081081081082e-06,
      "loss": 0.4386,
      "step": 8358
    },
    {
      "epoch": 1.79,
      "grad_norm": 7.136044979095459,
      "learning_rate": 3.104890604890605e-06,
      "loss": 0.2946,
      "step": 8359
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.771517753601074,
      "learning_rate": 3.1016731016731015e-06,
      "loss": 0.3203,
      "step": 8360
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.191667556762695,
      "learning_rate": 3.0984555984555984e-06,
      "loss": 0.5698,
      "step": 8361
    },
    {
      "epoch": 1.79,
      "grad_norm": 11.71028995513916,
      "learning_rate": 3.0952380952380953e-06,
      "loss": 0.3992,
      "step": 8362
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.745012283325195,
      "learning_rate": 3.092020592020592e-06,
      "loss": 0.3949,
      "step": 8363
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.437554359436035,
      "learning_rate": 3.0888030888030886e-06,
      "loss": 0.5309,
      "step": 8364
    },
    {
      "epoch": 1.79,
      "grad_norm": 10.543304443359375,
      "learning_rate": 3.0855855855855855e-06,
      "loss": 0.551,
      "step": 8365
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.50113582611084,
      "learning_rate": 3.082368082368083e-06,
      "loss": 0.4173,
      "step": 8366
    },
    {
      "epoch": 1.79,
      "grad_norm": 12.180273056030273,
      "learning_rate": 3.0791505791505792e-06,
      "loss": 0.5594,
      "step": 8367
    },
    {
      "epoch": 1.79,
      "grad_norm": 9.859417915344238,
      "learning_rate": 3.075933075933076e-06,
      "loss": 0.5736,
      "step": 8368
    },
    {
      "epoch": 1.8,
      "grad_norm": 8.43635082244873,
      "learning_rate": 3.0727155727155726e-06,
      "loss": 0.4891,
      "step": 8369
    },
    {
      "epoch": 1.8,
      "grad_norm": 7.967068195343018,
      "learning_rate": 3.06949806949807e-06,
      "loss": 0.3473,
      "step": 8370
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.471643447875977,
      "learning_rate": 3.0662805662805663e-06,
      "loss": 0.5779,
      "step": 8371
    },
    {
      "epoch": 1.8,
      "grad_norm": 11.763907432556152,
      "learning_rate": 3.063063063063063e-06,
      "loss": 0.3774,
      "step": 8372
    },
    {
      "epoch": 1.8,
      "grad_norm": 12.535530090332031,
      "learning_rate": 3.0598455598455596e-06,
      "loss": 0.5544,
      "step": 8373
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.380614280700684,
      "learning_rate": 3.056628056628057e-06,
      "loss": 0.4999,
      "step": 8374
    },
    {
      "epoch": 1.8,
      "grad_norm": 13.656112670898438,
      "learning_rate": 3.0534105534105534e-06,
      "loss": 0.4251,
      "step": 8375
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.737751960754395,
      "learning_rate": 3.0501930501930503e-06,
      "loss": 0.5715,
      "step": 8376
    },
    {
      "epoch": 1.8,
      "grad_norm": 9.858240127563477,
      "learning_rate": 3.046975546975547e-06,
      "loss": 0.5337,
      "step": 8377
    },
    {
      "epoch": 1.8,
      "grad_norm": 11.427053451538086,
      "learning_rate": 3.043758043758044e-06,
      "loss": 0.3335,
      "step": 8378
    },
    {
      "epoch": 1.8,
      "grad_norm": 7.896209239959717,
      "learning_rate": 3.040540540540541e-06,
      "loss": 0.4944,
      "step": 8379
    },
    {
      "epoch": 1.8,
      "grad_norm": 13.850829124450684,
      "learning_rate": 3.0373230373230373e-06,
      "loss": 0.4643,
      "step": 8380
    },
    {
      "epoch": 1.8,
      "grad_norm": 9.149469375610352,
      "learning_rate": 3.0341055341055342e-06,
      "loss": 0.3,
      "step": 8381
    },
    {
      "epoch": 1.8,
      "grad_norm": 15.43426513671875,
      "learning_rate": 3.0308880308880307e-06,
      "loss": 0.8616,
      "step": 8382
    },
    {
      "epoch": 1.8,
      "grad_norm": 12.139071464538574,
      "learning_rate": 3.027670527670528e-06,
      "loss": 0.4358,
      "step": 8383
    },
    {
      "epoch": 1.8,
      "grad_norm": 12.002842903137207,
      "learning_rate": 3.0244530244530244e-06,
      "loss": 0.3416,
      "step": 8384
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.049283981323242,
      "learning_rate": 3.0212355212355213e-06,
      "loss": 0.4963,
      "step": 8385
    },
    {
      "epoch": 1.8,
      "grad_norm": 9.975092887878418,
      "learning_rate": 3.0180180180180177e-06,
      "loss": 0.4303,
      "step": 8386
    },
    {
      "epoch": 1.8,
      "grad_norm": 12.14023494720459,
      "learning_rate": 3.014800514800515e-06,
      "loss": 0.4284,
      "step": 8387
    },
    {
      "epoch": 1.8,
      "grad_norm": 9.450207710266113,
      "learning_rate": 3.0115830115830115e-06,
      "loss": 0.3964,
      "step": 8388
    },
    {
      "epoch": 1.8,
      "grad_norm": 11.021932601928711,
      "learning_rate": 3.0083655083655084e-06,
      "loss": 0.5507,
      "step": 8389
    },
    {
      "epoch": 1.8,
      "grad_norm": 8.422616004943848,
      "learning_rate": 3.0051480051480052e-06,
      "loss": 0.3696,
      "step": 8390
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.810750961303711,
      "learning_rate": 3.001930501930502e-06,
      "loss": 0.4578,
      "step": 8391
    },
    {
      "epoch": 1.8,
      "grad_norm": 11.921943664550781,
      "learning_rate": 2.998712998712999e-06,
      "loss": 0.5305,
      "step": 8392
    },
    {
      "epoch": 1.8,
      "grad_norm": 13.400609016418457,
      "learning_rate": 2.9954954954954954e-06,
      "loss": 0.8958,
      "step": 8393
    },
    {
      "epoch": 1.8,
      "grad_norm": 9.553144454956055,
      "learning_rate": 2.9922779922779923e-06,
      "loss": 0.57,
      "step": 8394
    },
    {
      "epoch": 1.8,
      "grad_norm": 8.312195777893066,
      "learning_rate": 2.989060489060489e-06,
      "loss": 0.377,
      "step": 8395
    },
    {
      "epoch": 1.8,
      "grad_norm": 8.243773460388184,
      "learning_rate": 2.985842985842986e-06,
      "loss": 0.4203,
      "step": 8396
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.59284782409668,
      "learning_rate": 2.9826254826254825e-06,
      "loss": 0.5474,
      "step": 8397
    },
    {
      "epoch": 1.8,
      "grad_norm": 12.711773872375488,
      "learning_rate": 2.9794079794079794e-06,
      "loss": 0.5634,
      "step": 8398
    },
    {
      "epoch": 1.8,
      "grad_norm": 13.965332984924316,
      "learning_rate": 2.9761904761904763e-06,
      "loss": 0.8547,
      "step": 8399
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.098588943481445,
      "learning_rate": 2.972972972972973e-06,
      "loss": 0.4872,
      "step": 8400
    },
    {
      "epoch": 1.8,
      "grad_norm": 11.058145523071289,
      "learning_rate": 2.9697554697554696e-06,
      "loss": 0.4754,
      "step": 8401
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.087059020996094,
      "learning_rate": 2.9665379665379665e-06,
      "loss": 0.329,
      "step": 8402
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.97556209564209,
      "learning_rate": 2.9633204633204638e-06,
      "loss": 0.4582,
      "step": 8403
    },
    {
      "epoch": 1.8,
      "grad_norm": 13.69932746887207,
      "learning_rate": 2.9601029601029602e-06,
      "loss": 0.664,
      "step": 8404
    },
    {
      "epoch": 1.8,
      "grad_norm": 13.488910675048828,
      "learning_rate": 2.956885456885457e-06,
      "loss": 0.5504,
      "step": 8405
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.551952362060547,
      "learning_rate": 2.9536679536679535e-06,
      "loss": 0.4634,
      "step": 8406
    },
    {
      "epoch": 1.8,
      "grad_norm": 11.980945587158203,
      "learning_rate": 2.950450450450451e-06,
      "loss": 0.4445,
      "step": 8407
    },
    {
      "epoch": 1.8,
      "grad_norm": 9.791387557983398,
      "learning_rate": 2.9472329472329473e-06,
      "loss": 0.5267,
      "step": 8408
    },
    {
      "epoch": 1.8,
      "grad_norm": 8.754323959350586,
      "learning_rate": 2.944015444015444e-06,
      "loss": 0.5118,
      "step": 8409
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.984283447265625,
      "learning_rate": 2.9407979407979406e-06,
      "loss": 0.5862,
      "step": 8410
    },
    {
      "epoch": 1.8,
      "grad_norm": 8.921906471252441,
      "learning_rate": 2.9375804375804375e-06,
      "loss": 0.4622,
      "step": 8411
    },
    {
      "epoch": 1.8,
      "grad_norm": 14.080056190490723,
      "learning_rate": 2.9343629343629344e-06,
      "loss": 0.7116,
      "step": 8412
    },
    {
      "epoch": 1.8,
      "grad_norm": 8.610692024230957,
      "learning_rate": 2.9311454311454313e-06,
      "loss": 0.5615,
      "step": 8413
    },
    {
      "epoch": 1.8,
      "grad_norm": 12.260862350463867,
      "learning_rate": 2.9279279279279277e-06,
      "loss": 0.5773,
      "step": 8414
    },
    {
      "epoch": 1.81,
      "grad_norm": 8.038182258605957,
      "learning_rate": 2.9247104247104246e-06,
      "loss": 0.3679,
      "step": 8415
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.078004837036133,
      "learning_rate": 2.921492921492922e-06,
      "loss": 0.494,
      "step": 8416
    },
    {
      "epoch": 1.81,
      "grad_norm": 7.781179428100586,
      "learning_rate": 2.9182754182754183e-06,
      "loss": 0.4031,
      "step": 8417
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.793485641479492,
      "learning_rate": 2.915057915057915e-06,
      "loss": 0.6689,
      "step": 8418
    },
    {
      "epoch": 1.81,
      "grad_norm": 13.308121681213379,
      "learning_rate": 2.9118404118404117e-06,
      "loss": 0.6224,
      "step": 8419
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.418156623840332,
      "learning_rate": 2.908622908622909e-06,
      "loss": 0.4059,
      "step": 8420
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.019391059875488,
      "learning_rate": 2.9054054054054054e-06,
      "loss": 0.5061,
      "step": 8421
    },
    {
      "epoch": 1.81,
      "grad_norm": 7.224006175994873,
      "learning_rate": 2.9021879021879023e-06,
      "loss": 0.4291,
      "step": 8422
    },
    {
      "epoch": 1.81,
      "grad_norm": 7.8459553718566895,
      "learning_rate": 2.8989703989703987e-06,
      "loss": 0.365,
      "step": 8423
    },
    {
      "epoch": 1.81,
      "grad_norm": 14.962118148803711,
      "learning_rate": 2.895752895752896e-06,
      "loss": 0.4554,
      "step": 8424
    },
    {
      "epoch": 1.81,
      "grad_norm": 8.554043769836426,
      "learning_rate": 2.8925353925353925e-06,
      "loss": 0.2746,
      "step": 8425
    },
    {
      "epoch": 1.81,
      "grad_norm": 14.850507736206055,
      "learning_rate": 2.8893178893178894e-06,
      "loss": 0.3522,
      "step": 8426
    },
    {
      "epoch": 1.81,
      "grad_norm": 12.24509334564209,
      "learning_rate": 2.8861003861003862e-06,
      "loss": 0.615,
      "step": 8427
    },
    {
      "epoch": 1.81,
      "grad_norm": 9.413944244384766,
      "learning_rate": 2.882882882882883e-06,
      "loss": 0.517,
      "step": 8428
    },
    {
      "epoch": 1.81,
      "grad_norm": 13.737472534179688,
      "learning_rate": 2.87966537966538e-06,
      "loss": 0.6074,
      "step": 8429
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.83360767364502,
      "learning_rate": 2.8764478764478764e-06,
      "loss": 0.6242,
      "step": 8430
    },
    {
      "epoch": 1.81,
      "grad_norm": 8.743237495422363,
      "learning_rate": 2.8732303732303733e-06,
      "loss": 0.4228,
      "step": 8431
    },
    {
      "epoch": 1.81,
      "grad_norm": 9.438132286071777,
      "learning_rate": 2.87001287001287e-06,
      "loss": 0.5472,
      "step": 8432
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.179725646972656,
      "learning_rate": 2.866795366795367e-06,
      "loss": 0.4863,
      "step": 8433
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.305292129516602,
      "learning_rate": 2.8635778635778635e-06,
      "loss": 0.4762,
      "step": 8434
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.52528190612793,
      "learning_rate": 2.8603603603603604e-06,
      "loss": 0.57,
      "step": 8435
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.196735382080078,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 0.6114,
      "step": 8436
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.830265045166016,
      "learning_rate": 2.853925353925354e-06,
      "loss": 0.4205,
      "step": 8437
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.185787200927734,
      "learning_rate": 2.8507078507078506e-06,
      "loss": 0.4482,
      "step": 8438
    },
    {
      "epoch": 1.81,
      "grad_norm": 9.513829231262207,
      "learning_rate": 2.8474903474903475e-06,
      "loss": 0.4367,
      "step": 8439
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.137819290161133,
      "learning_rate": 2.8442728442728443e-06,
      "loss": 0.4162,
      "step": 8440
    },
    {
      "epoch": 1.81,
      "grad_norm": 13.211014747619629,
      "learning_rate": 2.841055341055341e-06,
      "loss": 0.4694,
      "step": 8441
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.273589134216309,
      "learning_rate": 2.837837837837838e-06,
      "loss": 0.3528,
      "step": 8442
    },
    {
      "epoch": 1.81,
      "grad_norm": 9.97707748413086,
      "learning_rate": 2.8346203346203345e-06,
      "loss": 0.4399,
      "step": 8443
    },
    {
      "epoch": 1.81,
      "grad_norm": 13.797260284423828,
      "learning_rate": 2.8314028314028314e-06,
      "loss": 0.4563,
      "step": 8444
    },
    {
      "epoch": 1.81,
      "grad_norm": 9.628640174865723,
      "learning_rate": 2.8281853281853283e-06,
      "loss": 0.3517,
      "step": 8445
    },
    {
      "epoch": 1.81,
      "grad_norm": 8.660446166992188,
      "learning_rate": 2.824967824967825e-06,
      "loss": 0.5117,
      "step": 8446
    },
    {
      "epoch": 1.81,
      "grad_norm": 9.516331672668457,
      "learning_rate": 2.8217503217503216e-06,
      "loss": 0.4776,
      "step": 8447
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.454891204833984,
      "learning_rate": 2.8185328185328185e-06,
      "loss": 0.5151,
      "step": 8448
    },
    {
      "epoch": 1.81,
      "grad_norm": 9.229538917541504,
      "learning_rate": 2.8153153153153154e-06,
      "loss": 0.4171,
      "step": 8449
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.647714614868164,
      "learning_rate": 2.8120978120978122e-06,
      "loss": 0.6499,
      "step": 8450
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.894794464111328,
      "learning_rate": 2.8088803088803087e-06,
      "loss": 0.4969,
      "step": 8451
    },
    {
      "epoch": 1.81,
      "grad_norm": 12.814407348632812,
      "learning_rate": 2.8056628056628056e-06,
      "loss": 0.5984,
      "step": 8452
    },
    {
      "epoch": 1.81,
      "grad_norm": 15.298995971679688,
      "learning_rate": 2.802445302445303e-06,
      "loss": 0.4347,
      "step": 8453
    },
    {
      "epoch": 1.81,
      "grad_norm": 12.350019454956055,
      "learning_rate": 2.7992277992277993e-06,
      "loss": 0.5653,
      "step": 8454
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.119585037231445,
      "learning_rate": 2.796010296010296e-06,
      "loss": 0.4901,
      "step": 8455
    },
    {
      "epoch": 1.81,
      "grad_norm": 9.959598541259766,
      "learning_rate": 2.7927927927927926e-06,
      "loss": 0.6226,
      "step": 8456
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.891674041748047,
      "learning_rate": 2.78957528957529e-06,
      "loss": 0.621,
      "step": 8457
    },
    {
      "epoch": 1.81,
      "grad_norm": 10.792990684509277,
      "learning_rate": 2.7863577863577864e-06,
      "loss": 0.5348,
      "step": 8458
    },
    {
      "epoch": 1.81,
      "grad_norm": 12.487275123596191,
      "learning_rate": 2.7831402831402833e-06,
      "loss": 0.4868,
      "step": 8459
    },
    {
      "epoch": 1.81,
      "grad_norm": 7.486839294433594,
      "learning_rate": 2.7799227799227797e-06,
      "loss": 0.4353,
      "step": 8460
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.966978073120117,
      "learning_rate": 2.776705276705277e-06,
      "loss": 0.6407,
      "step": 8461
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.67630672454834,
      "learning_rate": 2.7734877734877735e-06,
      "loss": 0.5707,
      "step": 8462
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.708325386047363,
      "learning_rate": 2.7702702702702703e-06,
      "loss": 0.5263,
      "step": 8463
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.485636711120605,
      "learning_rate": 2.767052767052767e-06,
      "loss": 0.3722,
      "step": 8464
    },
    {
      "epoch": 1.82,
      "grad_norm": 7.216131687164307,
      "learning_rate": 2.763835263835264e-06,
      "loss": 0.3933,
      "step": 8465
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.173367500305176,
      "learning_rate": 2.760617760617761e-06,
      "loss": 0.4068,
      "step": 8466
    },
    {
      "epoch": 1.82,
      "grad_norm": 13.407514572143555,
      "learning_rate": 2.7574002574002574e-06,
      "loss": 0.5635,
      "step": 8467
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.060855865478516,
      "learning_rate": 2.7541827541827543e-06,
      "loss": 0.4929,
      "step": 8468
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.73710823059082,
      "learning_rate": 2.7509652509652507e-06,
      "loss": 0.4063,
      "step": 8469
    },
    {
      "epoch": 1.82,
      "grad_norm": 9.727356910705566,
      "learning_rate": 2.747747747747748e-06,
      "loss": 0.5125,
      "step": 8470
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.102143287658691,
      "learning_rate": 2.7445302445302445e-06,
      "loss": 0.4819,
      "step": 8471
    },
    {
      "epoch": 1.82,
      "grad_norm": 9.007627487182617,
      "learning_rate": 2.7413127413127414e-06,
      "loss": 0.5642,
      "step": 8472
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.358109474182129,
      "learning_rate": 2.738095238095238e-06,
      "loss": 0.5391,
      "step": 8473
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.66494083404541,
      "learning_rate": 2.734877734877735e-06,
      "loss": 0.5219,
      "step": 8474
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.559310913085938,
      "learning_rate": 2.7316602316602316e-06,
      "loss": 0.5451,
      "step": 8475
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.686568260192871,
      "learning_rate": 2.7284427284427284e-06,
      "loss": 0.5748,
      "step": 8476
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.229429244995117,
      "learning_rate": 2.7252252252252253e-06,
      "loss": 0.4449,
      "step": 8477
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.031475067138672,
      "learning_rate": 2.722007722007722e-06,
      "loss": 0.503,
      "step": 8478
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.80917739868164,
      "learning_rate": 2.718790218790219e-06,
      "loss": 0.6425,
      "step": 8479
    },
    {
      "epoch": 1.82,
      "grad_norm": 8.431861877441406,
      "learning_rate": 2.7155727155727155e-06,
      "loss": 0.4137,
      "step": 8480
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.539666175842285,
      "learning_rate": 2.7123552123552124e-06,
      "loss": 0.5754,
      "step": 8481
    },
    {
      "epoch": 1.82,
      "grad_norm": 9.547799110412598,
      "learning_rate": 2.7091377091377093e-06,
      "loss": 0.4027,
      "step": 8482
    },
    {
      "epoch": 1.82,
      "grad_norm": 9.956905364990234,
      "learning_rate": 2.705920205920206e-06,
      "loss": 0.5482,
      "step": 8483
    },
    {
      "epoch": 1.82,
      "grad_norm": 12.089627265930176,
      "learning_rate": 2.7027027027027026e-06,
      "loss": 0.5558,
      "step": 8484
    },
    {
      "epoch": 1.82,
      "grad_norm": 14.95789623260498,
      "learning_rate": 2.6994851994851995e-06,
      "loss": 0.5747,
      "step": 8485
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.117624282836914,
      "learning_rate": 2.6962676962676963e-06,
      "loss": 0.5262,
      "step": 8486
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.13149642944336,
      "learning_rate": 2.6930501930501932e-06,
      "loss": 0.4875,
      "step": 8487
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.75609016418457,
      "learning_rate": 2.6898326898326897e-06,
      "loss": 0.6095,
      "step": 8488
    },
    {
      "epoch": 1.82,
      "grad_norm": 8.94497299194336,
      "learning_rate": 2.6866151866151865e-06,
      "loss": 0.4359,
      "step": 8489
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.724812507629395,
      "learning_rate": 2.683397683397684e-06,
      "loss": 0.518,
      "step": 8490
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.444424629211426,
      "learning_rate": 2.6801801801801803e-06,
      "loss": 0.44,
      "step": 8491
    },
    {
      "epoch": 1.82,
      "grad_norm": 12.39590835571289,
      "learning_rate": 2.676962676962677e-06,
      "loss": 0.6638,
      "step": 8492
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.898370742797852,
      "learning_rate": 2.6737451737451736e-06,
      "loss": 0.6154,
      "step": 8493
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.417187690734863,
      "learning_rate": 2.670527670527671e-06,
      "loss": 0.4682,
      "step": 8494
    },
    {
      "epoch": 1.82,
      "grad_norm": 9.944619178771973,
      "learning_rate": 2.6673101673101674e-06,
      "loss": 0.4611,
      "step": 8495
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.627894401550293,
      "learning_rate": 2.6640926640926642e-06,
      "loss": 0.4909,
      "step": 8496
    },
    {
      "epoch": 1.82,
      "grad_norm": 8.631427764892578,
      "learning_rate": 2.6608751608751607e-06,
      "loss": 0.5338,
      "step": 8497
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.069109916687012,
      "learning_rate": 2.6576576576576576e-06,
      "loss": 0.3662,
      "step": 8498
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.463919639587402,
      "learning_rate": 2.6544401544401544e-06,
      "loss": 0.5389,
      "step": 8499
    },
    {
      "epoch": 1.82,
      "grad_norm": 13.933974266052246,
      "learning_rate": 2.6512226512226513e-06,
      "loss": 0.6101,
      "step": 8500
    },
    {
      "epoch": 1.82,
      "grad_norm": 10.174469947814941,
      "learning_rate": 2.6480051480051478e-06,
      "loss": 0.5104,
      "step": 8501
    },
    {
      "epoch": 1.82,
      "grad_norm": 9.542536735534668,
      "learning_rate": 2.6447876447876446e-06,
      "loss": 0.3774,
      "step": 8502
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.36348819732666,
      "learning_rate": 2.641570141570142e-06,
      "loss": 0.6202,
      "step": 8503
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.004154205322266,
      "learning_rate": 2.6383526383526384e-06,
      "loss": 0.3498,
      "step": 8504
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.653096199035645,
      "learning_rate": 2.6351351351351353e-06,
      "loss": 0.6056,
      "step": 8505
    },
    {
      "epoch": 1.82,
      "grad_norm": 7.665951728820801,
      "learning_rate": 2.6319176319176317e-06,
      "loss": 0.3211,
      "step": 8506
    },
    {
      "epoch": 1.82,
      "grad_norm": 8.345185279846191,
      "learning_rate": 2.628700128700129e-06,
      "loss": 0.4259,
      "step": 8507
    },
    {
      "epoch": 1.82,
      "grad_norm": 11.417954444885254,
      "learning_rate": 2.6254826254826255e-06,
      "loss": 0.4532,
      "step": 8508
    },
    {
      "epoch": 1.83,
      "grad_norm": 11.935803413391113,
      "learning_rate": 2.6222651222651223e-06,
      "loss": 0.6544,
      "step": 8509
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.181653022766113,
      "learning_rate": 2.619047619047619e-06,
      "loss": 0.5685,
      "step": 8510
    },
    {
      "epoch": 1.83,
      "grad_norm": 8.63326644897461,
      "learning_rate": 2.615830115830116e-06,
      "loss": 0.4977,
      "step": 8511
    },
    {
      "epoch": 1.83,
      "grad_norm": 9.184489250183105,
      "learning_rate": 2.6126126126126125e-06,
      "loss": 0.4476,
      "step": 8512
    },
    {
      "epoch": 1.83,
      "grad_norm": 8.9111328125,
      "learning_rate": 2.6093951093951094e-06,
      "loss": 0.4289,
      "step": 8513
    },
    {
      "epoch": 1.83,
      "grad_norm": 9.457598686218262,
      "learning_rate": 2.6061776061776063e-06,
      "loss": 0.5549,
      "step": 8514
    },
    {
      "epoch": 1.83,
      "grad_norm": 11.04207706451416,
      "learning_rate": 2.602960102960103e-06,
      "loss": 0.543,
      "step": 8515
    },
    {
      "epoch": 1.83,
      "grad_norm": 11.206500053405762,
      "learning_rate": 2.5997425997426e-06,
      "loss": 0.4772,
      "step": 8516
    },
    {
      "epoch": 1.83,
      "grad_norm": 9.453638076782227,
      "learning_rate": 2.5965250965250965e-06,
      "loss": 0.4358,
      "step": 8517
    },
    {
      "epoch": 1.83,
      "grad_norm": 15.617684364318848,
      "learning_rate": 2.5933075933075934e-06,
      "loss": 0.6603,
      "step": 8518
    },
    {
      "epoch": 1.83,
      "grad_norm": 9.478924751281738,
      "learning_rate": 2.5900900900900902e-06,
      "loss": 0.4667,
      "step": 8519
    },
    {
      "epoch": 1.83,
      "grad_norm": 14.39090347290039,
      "learning_rate": 2.586872586872587e-06,
      "loss": 0.5039,
      "step": 8520
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.419942855834961,
      "learning_rate": 2.5836550836550836e-06,
      "loss": 0.4447,
      "step": 8521
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.024911880493164,
      "learning_rate": 2.5804375804375804e-06,
      "loss": 0.4995,
      "step": 8522
    },
    {
      "epoch": 1.83,
      "grad_norm": 9.974100112915039,
      "learning_rate": 2.5772200772200773e-06,
      "loss": 0.3716,
      "step": 8523
    },
    {
      "epoch": 1.83,
      "grad_norm": 8.422863006591797,
      "learning_rate": 2.574002574002574e-06,
      "loss": 0.6327,
      "step": 8524
    },
    {
      "epoch": 1.83,
      "grad_norm": 7.99601936340332,
      "learning_rate": 2.5707850707850707e-06,
      "loss": 0.3536,
      "step": 8525
    },
    {
      "epoch": 1.83,
      "grad_norm": 9.70492935180664,
      "learning_rate": 2.5675675675675675e-06,
      "loss": 0.3589,
      "step": 8526
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.032659530639648,
      "learning_rate": 2.5643500643500644e-06,
      "loss": 0.468,
      "step": 8527
    },
    {
      "epoch": 1.83,
      "grad_norm": 11.47100830078125,
      "learning_rate": 2.5611325611325613e-06,
      "loss": 0.6244,
      "step": 8528
    },
    {
      "epoch": 1.83,
      "grad_norm": 9.615069389343262,
      "learning_rate": 2.557915057915058e-06,
      "loss": 0.5084,
      "step": 8529
    },
    {
      "epoch": 1.83,
      "grad_norm": 12.180094718933105,
      "learning_rate": 2.5546975546975546e-06,
      "loss": 0.6402,
      "step": 8530
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.15190601348877,
      "learning_rate": 2.5514800514800515e-06,
      "loss": 0.5027,
      "step": 8531
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.339813232421875,
      "learning_rate": 2.5482625482625484e-06,
      "loss": 0.4557,
      "step": 8532
    },
    {
      "epoch": 1.83,
      "grad_norm": 11.700761795043945,
      "learning_rate": 2.5450450450450452e-06,
      "loss": 0.7163,
      "step": 8533
    },
    {
      "epoch": 1.83,
      "grad_norm": 8.71394157409668,
      "learning_rate": 2.5418275418275417e-06,
      "loss": 0.5343,
      "step": 8534
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.674482345581055,
      "learning_rate": 2.5386100386100386e-06,
      "loss": 0.4806,
      "step": 8535
    },
    {
      "epoch": 1.83,
      "grad_norm": 13.501008987426758,
      "learning_rate": 2.5353925353925354e-06,
      "loss": 0.5365,
      "step": 8536
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.733193397521973,
      "learning_rate": 2.5321750321750323e-06,
      "loss": 0.4672,
      "step": 8537
    },
    {
      "epoch": 1.83,
      "grad_norm": 13.354889869689941,
      "learning_rate": 2.5289575289575288e-06,
      "loss": 0.4985,
      "step": 8538
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.790650367736816,
      "learning_rate": 2.5257400257400256e-06,
      "loss": 0.4902,
      "step": 8539
    },
    {
      "epoch": 1.83,
      "grad_norm": 13.452027320861816,
      "learning_rate": 2.522522522522523e-06,
      "loss": 0.4904,
      "step": 8540
    },
    {
      "epoch": 1.83,
      "grad_norm": 12.81370735168457,
      "learning_rate": 2.5193050193050194e-06,
      "loss": 0.4858,
      "step": 8541
    },
    {
      "epoch": 1.83,
      "grad_norm": 8.867133140563965,
      "learning_rate": 2.5160875160875163e-06,
      "loss": 0.3713,
      "step": 8542
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.669609069824219,
      "learning_rate": 2.5128700128700127e-06,
      "loss": 0.4462,
      "step": 8543
    },
    {
      "epoch": 1.83,
      "grad_norm": 13.096739768981934,
      "learning_rate": 2.50965250965251e-06,
      "loss": 0.6018,
      "step": 8544
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.615771293640137,
      "learning_rate": 2.5064350064350065e-06,
      "loss": 0.612,
      "step": 8545
    },
    {
      "epoch": 1.83,
      "grad_norm": 9.693830490112305,
      "learning_rate": 2.5032175032175033e-06,
      "loss": 0.491,
      "step": 8546
    },
    {
      "epoch": 1.83,
      "grad_norm": 12.07260513305664,
      "learning_rate": 2.4999999999999998e-06,
      "loss": 0.6697,
      "step": 8547
    },
    {
      "epoch": 1.83,
      "grad_norm": 11.117100715637207,
      "learning_rate": 2.496782496782497e-06,
      "loss": 0.5065,
      "step": 8548
    },
    {
      "epoch": 1.83,
      "grad_norm": 15.398804664611816,
      "learning_rate": 2.4935649935649935e-06,
      "loss": 0.5714,
      "step": 8549
    },
    {
      "epoch": 1.83,
      "grad_norm": 12.840124130249023,
      "learning_rate": 2.4903474903474904e-06,
      "loss": 0.5101,
      "step": 8550
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.081388473510742,
      "learning_rate": 2.487129987129987e-06,
      "loss": 0.4808,
      "step": 8551
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.496368408203125,
      "learning_rate": 2.483912483912484e-06,
      "loss": 0.5482,
      "step": 8552
    },
    {
      "epoch": 1.83,
      "grad_norm": 9.688492774963379,
      "learning_rate": 2.480694980694981e-06,
      "loss": 0.4789,
      "step": 8553
    },
    {
      "epoch": 1.83,
      "grad_norm": 9.525228500366211,
      "learning_rate": 2.4774774774774775e-06,
      "loss": 0.7313,
      "step": 8554
    },
    {
      "epoch": 1.84,
      "grad_norm": 9.367168426513672,
      "learning_rate": 2.4742599742599744e-06,
      "loss": 0.4471,
      "step": 8555
    },
    {
      "epoch": 1.84,
      "grad_norm": 9.685452461242676,
      "learning_rate": 2.4710424710424712e-06,
      "loss": 0.5097,
      "step": 8556
    },
    {
      "epoch": 1.84,
      "grad_norm": 7.296173572540283,
      "learning_rate": 2.467824967824968e-06,
      "loss": 0.3961,
      "step": 8557
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.291838645935059,
      "learning_rate": 2.4646074646074646e-06,
      "loss": 0.4596,
      "step": 8558
    },
    {
      "epoch": 1.84,
      "grad_norm": 12.55722713470459,
      "learning_rate": 2.4613899613899614e-06,
      "loss": 0.4941,
      "step": 8559
    },
    {
      "epoch": 1.84,
      "grad_norm": 11.937041282653809,
      "learning_rate": 2.458172458172458e-06,
      "loss": 0.487,
      "step": 8560
    },
    {
      "epoch": 1.84,
      "grad_norm": 15.241929054260254,
      "learning_rate": 2.454954954954955e-06,
      "loss": 0.6191,
      "step": 8561
    },
    {
      "epoch": 1.84,
      "grad_norm": 13.655635833740234,
      "learning_rate": 2.4517374517374516e-06,
      "loss": 0.6267,
      "step": 8562
    },
    {
      "epoch": 1.84,
      "grad_norm": 12.22413444519043,
      "learning_rate": 2.4485199485199485e-06,
      "loss": 0.5712,
      "step": 8563
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.06064224243164,
      "learning_rate": 2.4453024453024454e-06,
      "loss": 0.4679,
      "step": 8564
    },
    {
      "epoch": 1.84,
      "grad_norm": 13.629780769348145,
      "learning_rate": 2.4420849420849423e-06,
      "loss": 0.5436,
      "step": 8565
    },
    {
      "epoch": 1.84,
      "grad_norm": 12.941228866577148,
      "learning_rate": 2.438867438867439e-06,
      "loss": 0.4684,
      "step": 8566
    },
    {
      "epoch": 1.84,
      "grad_norm": 8.675895690917969,
      "learning_rate": 2.4356499356499356e-06,
      "loss": 0.5177,
      "step": 8567
    },
    {
      "epoch": 1.84,
      "grad_norm": 9.025931358337402,
      "learning_rate": 2.4324324324324325e-06,
      "loss": 0.5038,
      "step": 8568
    },
    {
      "epoch": 1.84,
      "grad_norm": 11.743083000183105,
      "learning_rate": 2.4292149292149293e-06,
      "loss": 0.6071,
      "step": 8569
    },
    {
      "epoch": 1.84,
      "grad_norm": 8.976631164550781,
      "learning_rate": 2.425997425997426e-06,
      "loss": 0.4133,
      "step": 8570
    },
    {
      "epoch": 1.84,
      "grad_norm": 8.947089195251465,
      "learning_rate": 2.4227799227799227e-06,
      "loss": 0.3534,
      "step": 8571
    },
    {
      "epoch": 1.84,
      "grad_norm": 11.198953628540039,
      "learning_rate": 2.4195624195624195e-06,
      "loss": 0.5557,
      "step": 8572
    },
    {
      "epoch": 1.84,
      "grad_norm": 8.879084587097168,
      "learning_rate": 2.4163449163449164e-06,
      "loss": 0.3145,
      "step": 8573
    },
    {
      "epoch": 1.84,
      "grad_norm": 13.055782318115234,
      "learning_rate": 2.4131274131274133e-06,
      "loss": 0.4989,
      "step": 8574
    },
    {
      "epoch": 1.84,
      "grad_norm": 14.382573127746582,
      "learning_rate": 2.4099099099099097e-06,
      "loss": 0.7507,
      "step": 8575
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.440579414367676,
      "learning_rate": 2.4066924066924066e-06,
      "loss": 0.4015,
      "step": 8576
    },
    {
      "epoch": 1.84,
      "grad_norm": 9.781270027160645,
      "learning_rate": 2.403474903474904e-06,
      "loss": 0.3947,
      "step": 8577
    },
    {
      "epoch": 1.84,
      "grad_norm": 9.502378463745117,
      "learning_rate": 2.4002574002574004e-06,
      "loss": 0.5001,
      "step": 8578
    },
    {
      "epoch": 1.84,
      "grad_norm": 14.704780578613281,
      "learning_rate": 2.3970398970398972e-06,
      "loss": 0.6847,
      "step": 8579
    },
    {
      "epoch": 1.84,
      "grad_norm": 11.436848640441895,
      "learning_rate": 2.3938223938223937e-06,
      "loss": 0.5497,
      "step": 8580
    },
    {
      "epoch": 1.84,
      "grad_norm": 12.497705459594727,
      "learning_rate": 2.390604890604891e-06,
      "loss": 0.4552,
      "step": 8581
    },
    {
      "epoch": 1.84,
      "grad_norm": 9.090459823608398,
      "learning_rate": 2.3873873873873874e-06,
      "loss": 0.3905,
      "step": 8582
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.953229904174805,
      "learning_rate": 2.3841698841698843e-06,
      "loss": 0.3873,
      "step": 8583
    },
    {
      "epoch": 1.84,
      "grad_norm": 12.25234317779541,
      "learning_rate": 2.3809523809523808e-06,
      "loss": 0.4854,
      "step": 8584
    },
    {
      "epoch": 1.84,
      "grad_norm": 12.28246021270752,
      "learning_rate": 2.377734877734878e-06,
      "loss": 0.6199,
      "step": 8585
    },
    {
      "epoch": 1.84,
      "grad_norm": 7.970196723937988,
      "learning_rate": 2.3745173745173745e-06,
      "loss": 0.3945,
      "step": 8586
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.80158519744873,
      "learning_rate": 2.3712998712998714e-06,
      "loss": 0.6583,
      "step": 8587
    },
    {
      "epoch": 1.84,
      "grad_norm": 7.363993167877197,
      "learning_rate": 2.368082368082368e-06,
      "loss": 0.3579,
      "step": 8588
    },
    {
      "epoch": 1.84,
      "grad_norm": 8.542867660522461,
      "learning_rate": 2.3648648648648647e-06,
      "loss": 0.4256,
      "step": 8589
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.446789741516113,
      "learning_rate": 2.361647361647362e-06,
      "loss": 0.5043,
      "step": 8590
    },
    {
      "epoch": 1.84,
      "grad_norm": 7.906985759735107,
      "learning_rate": 2.3584298584298585e-06,
      "loss": 0.3921,
      "step": 8591
    },
    {
      "epoch": 1.84,
      "grad_norm": 12.645427703857422,
      "learning_rate": 2.3552123552123553e-06,
      "loss": 0.4236,
      "step": 8592
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.733208656311035,
      "learning_rate": 2.351994851994852e-06,
      "loss": 0.5315,
      "step": 8593
    },
    {
      "epoch": 1.84,
      "grad_norm": 11.345610618591309,
      "learning_rate": 2.348777348777349e-06,
      "loss": 0.5294,
      "step": 8594
    },
    {
      "epoch": 1.84,
      "grad_norm": 9.725733757019043,
      "learning_rate": 2.3455598455598455e-06,
      "loss": 0.5858,
      "step": 8595
    },
    {
      "epoch": 1.84,
      "grad_norm": 11.782147407531738,
      "learning_rate": 2.3423423423423424e-06,
      "loss": 0.3709,
      "step": 8596
    },
    {
      "epoch": 1.84,
      "grad_norm": 7.966995716094971,
      "learning_rate": 2.339124839124839e-06,
      "loss": 0.4297,
      "step": 8597
    },
    {
      "epoch": 1.84,
      "grad_norm": 12.329697608947754,
      "learning_rate": 2.335907335907336e-06,
      "loss": 0.5429,
      "step": 8598
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.189899444580078,
      "learning_rate": 2.3326898326898326e-06,
      "loss": 0.3875,
      "step": 8599
    },
    {
      "epoch": 1.84,
      "grad_norm": 11.844954490661621,
      "learning_rate": 2.3294723294723295e-06,
      "loss": 0.4484,
      "step": 8600
    },
    {
      "epoch": 1.84,
      "grad_norm": 10.4822998046875,
      "learning_rate": 2.326254826254826e-06,
      "loss": 0.6283,
      "step": 8601
    },
    {
      "epoch": 1.85,
      "grad_norm": 11.392524719238281,
      "learning_rate": 2.3230373230373232e-06,
      "loss": 0.3113,
      "step": 8602
    },
    {
      "epoch": 1.85,
      "grad_norm": 8.507864952087402,
      "learning_rate": 2.31981981981982e-06,
      "loss": 0.4763,
      "step": 8603
    },
    {
      "epoch": 1.85,
      "grad_norm": 8.917369842529297,
      "learning_rate": 2.3166023166023166e-06,
      "loss": 0.5118,
      "step": 8604
    },
    {
      "epoch": 1.85,
      "grad_norm": 11.443848609924316,
      "learning_rate": 2.3133848133848134e-06,
      "loss": 0.6418,
      "step": 8605
    },
    {
      "epoch": 1.85,
      "grad_norm": 11.003026008605957,
      "learning_rate": 2.3101673101673103e-06,
      "loss": 0.5701,
      "step": 8606
    },
    {
      "epoch": 1.85,
      "grad_norm": 8.64135456085205,
      "learning_rate": 2.306949806949807e-06,
      "loss": 0.3041,
      "step": 8607
    },
    {
      "epoch": 1.85,
      "grad_norm": 12.784475326538086,
      "learning_rate": 2.3037323037323036e-06,
      "loss": 0.6166,
      "step": 8608
    },
    {
      "epoch": 1.85,
      "grad_norm": 11.65021800994873,
      "learning_rate": 2.3005148005148005e-06,
      "loss": 0.417,
      "step": 8609
    },
    {
      "epoch": 1.85,
      "grad_norm": 12.144742965698242,
      "learning_rate": 2.2972972972972974e-06,
      "loss": 0.5558,
      "step": 8610
    },
    {
      "epoch": 1.85,
      "grad_norm": 12.416581153869629,
      "learning_rate": 2.2940797940797943e-06,
      "loss": 0.3763,
      "step": 8611
    },
    {
      "epoch": 1.85,
      "grad_norm": 8.729936599731445,
      "learning_rate": 2.2908622908622907e-06,
      "loss": 0.5175,
      "step": 8612
    },
    {
      "epoch": 1.85,
      "grad_norm": 12.81494426727295,
      "learning_rate": 2.2876447876447876e-06,
      "loss": 0.4655,
      "step": 8613
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.934486389160156,
      "learning_rate": 2.284427284427285e-06,
      "loss": 0.6373,
      "step": 8614
    },
    {
      "epoch": 1.85,
      "grad_norm": 7.238990783691406,
      "learning_rate": 2.2812097812097813e-06,
      "loss": 0.3313,
      "step": 8615
    },
    {
      "epoch": 1.85,
      "grad_norm": 8.853034973144531,
      "learning_rate": 2.2779922779922782e-06,
      "loss": 0.5535,
      "step": 8616
    },
    {
      "epoch": 1.85,
      "grad_norm": 13.610222816467285,
      "learning_rate": 2.2747747747747747e-06,
      "loss": 0.5459,
      "step": 8617
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.338613510131836,
      "learning_rate": 2.2715572715572715e-06,
      "loss": 0.5534,
      "step": 8618
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.008832931518555,
      "learning_rate": 2.2683397683397684e-06,
      "loss": 0.4371,
      "step": 8619
    },
    {
      "epoch": 1.85,
      "grad_norm": 11.411930084228516,
      "learning_rate": 2.2651222651222653e-06,
      "loss": 0.5541,
      "step": 8620
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.451348304748535,
      "learning_rate": 2.2619047619047617e-06,
      "loss": 0.4484,
      "step": 8621
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.98710823059082,
      "learning_rate": 2.2586872586872586e-06,
      "loss": 0.3654,
      "step": 8622
    },
    {
      "epoch": 1.85,
      "grad_norm": 7.759233474731445,
      "learning_rate": 2.2554697554697555e-06,
      "loss": 0.3543,
      "step": 8623
    },
    {
      "epoch": 1.85,
      "grad_norm": 7.951656818389893,
      "learning_rate": 2.2522522522522524e-06,
      "loss": 0.5252,
      "step": 8624
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.892328262329102,
      "learning_rate": 2.249034749034749e-06,
      "loss": 0.5259,
      "step": 8625
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.432415008544922,
      "learning_rate": 2.2458172458172457e-06,
      "loss": 0.5801,
      "step": 8626
    },
    {
      "epoch": 1.85,
      "grad_norm": 11.644968032836914,
      "learning_rate": 2.242599742599743e-06,
      "loss": 0.5715,
      "step": 8627
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.234803199768066,
      "learning_rate": 2.2393822393822394e-06,
      "loss": 0.4536,
      "step": 8628
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.154881477355957,
      "learning_rate": 2.2361647361647363e-06,
      "loss": 0.4359,
      "step": 8629
    },
    {
      "epoch": 1.85,
      "grad_norm": 12.454032897949219,
      "learning_rate": 2.2329472329472328e-06,
      "loss": 0.71,
      "step": 8630
    },
    {
      "epoch": 1.85,
      "grad_norm": 12.238669395446777,
      "learning_rate": 2.22972972972973e-06,
      "loss": 0.4357,
      "step": 8631
    },
    {
      "epoch": 1.85,
      "grad_norm": 13.500297546386719,
      "learning_rate": 2.2265122265122265e-06,
      "loss": 0.5937,
      "step": 8632
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.841011047363281,
      "learning_rate": 2.2232947232947234e-06,
      "loss": 0.4898,
      "step": 8633
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.319060325622559,
      "learning_rate": 2.22007722007722e-06,
      "loss": 0.44,
      "step": 8634
    },
    {
      "epoch": 1.85,
      "grad_norm": 7.77271032333374,
      "learning_rate": 2.216859716859717e-06,
      "loss": 0.2292,
      "step": 8635
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.334663391113281,
      "learning_rate": 2.2136422136422136e-06,
      "loss": 0.3481,
      "step": 8636
    },
    {
      "epoch": 1.85,
      "grad_norm": 12.771906852722168,
      "learning_rate": 2.2104247104247105e-06,
      "loss": 0.5637,
      "step": 8637
    },
    {
      "epoch": 1.85,
      "grad_norm": 11.945746421813965,
      "learning_rate": 2.207207207207207e-06,
      "loss": 0.6704,
      "step": 8638
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.74802303314209,
      "learning_rate": 2.2039897039897042e-06,
      "loss": 0.3876,
      "step": 8639
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.515950202941895,
      "learning_rate": 2.200772200772201e-06,
      "loss": 0.5114,
      "step": 8640
    },
    {
      "epoch": 1.85,
      "grad_norm": 11.194597244262695,
      "learning_rate": 2.1975546975546976e-06,
      "loss": 0.5296,
      "step": 8641
    },
    {
      "epoch": 1.85,
      "grad_norm": 16.69384765625,
      "learning_rate": 2.1943371943371944e-06,
      "loss": 0.7667,
      "step": 8642
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.228426933288574,
      "learning_rate": 2.1911196911196913e-06,
      "loss": 0.2458,
      "step": 8643
    },
    {
      "epoch": 1.85,
      "grad_norm": 11.190729141235352,
      "learning_rate": 2.187902187902188e-06,
      "loss": 0.4839,
      "step": 8644
    },
    {
      "epoch": 1.85,
      "grad_norm": 9.958837509155273,
      "learning_rate": 2.1846846846846846e-06,
      "loss": 0.5988,
      "step": 8645
    },
    {
      "epoch": 1.85,
      "grad_norm": 11.685744285583496,
      "learning_rate": 2.1814671814671815e-06,
      "loss": 0.806,
      "step": 8646
    },
    {
      "epoch": 1.85,
      "grad_norm": 10.36163330078125,
      "learning_rate": 2.178249678249678e-06,
      "loss": 0.4312,
      "step": 8647
    },
    {
      "epoch": 1.85,
      "grad_norm": 8.604080200195312,
      "learning_rate": 2.1750321750321753e-06,
      "loss": 0.4488,
      "step": 8648
    },
    {
      "epoch": 1.86,
      "grad_norm": 8.21401309967041,
      "learning_rate": 2.1718146718146717e-06,
      "loss": 0.3032,
      "step": 8649
    },
    {
      "epoch": 1.86,
      "grad_norm": 9.670101165771484,
      "learning_rate": 2.1685971685971686e-06,
      "loss": 0.4953,
      "step": 8650
    },
    {
      "epoch": 1.86,
      "grad_norm": 13.376473426818848,
      "learning_rate": 2.1653796653796655e-06,
      "loss": 0.4933,
      "step": 8651
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.344146728515625,
      "learning_rate": 2.1621621621621623e-06,
      "loss": 0.3981,
      "step": 8652
    },
    {
      "epoch": 1.86,
      "grad_norm": 12.0217924118042,
      "learning_rate": 2.158944658944659e-06,
      "loss": 0.5512,
      "step": 8653
    },
    {
      "epoch": 1.86,
      "grad_norm": 8.453635215759277,
      "learning_rate": 2.1557271557271557e-06,
      "loss": 0.4706,
      "step": 8654
    },
    {
      "epoch": 1.86,
      "grad_norm": 8.056635856628418,
      "learning_rate": 2.1525096525096525e-06,
      "loss": 0.4216,
      "step": 8655
    },
    {
      "epoch": 1.86,
      "grad_norm": 12.460457801818848,
      "learning_rate": 2.1492921492921494e-06,
      "loss": 0.4391,
      "step": 8656
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.917055130004883,
      "learning_rate": 2.1460746460746463e-06,
      "loss": 0.4513,
      "step": 8657
    },
    {
      "epoch": 1.86,
      "grad_norm": 9.72336196899414,
      "learning_rate": 2.1428571428571427e-06,
      "loss": 0.4144,
      "step": 8658
    },
    {
      "epoch": 1.86,
      "grad_norm": 9.578145980834961,
      "learning_rate": 2.1396396396396396e-06,
      "loss": 0.3487,
      "step": 8659
    },
    {
      "epoch": 1.86,
      "grad_norm": 8.814848899841309,
      "learning_rate": 2.1364221364221365e-06,
      "loss": 0.4208,
      "step": 8660
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.83579158782959,
      "learning_rate": 2.1332046332046334e-06,
      "loss": 0.4744,
      "step": 8661
    },
    {
      "epoch": 1.86,
      "grad_norm": 12.560216903686523,
      "learning_rate": 2.12998712998713e-06,
      "loss": 0.5975,
      "step": 8662
    },
    {
      "epoch": 1.86,
      "grad_norm": 8.21786117553711,
      "learning_rate": 2.1267696267696267e-06,
      "loss": 0.3253,
      "step": 8663
    },
    {
      "epoch": 1.86,
      "grad_norm": 12.466236114501953,
      "learning_rate": 2.123552123552124e-06,
      "loss": 0.3614,
      "step": 8664
    },
    {
      "epoch": 1.86,
      "grad_norm": 9.572443008422852,
      "learning_rate": 2.1203346203346204e-06,
      "loss": 0.4139,
      "step": 8665
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.957869529724121,
      "learning_rate": 2.1171171171171173e-06,
      "loss": 0.6385,
      "step": 8666
    },
    {
      "epoch": 1.86,
      "grad_norm": 9.76514720916748,
      "learning_rate": 2.1138996138996138e-06,
      "loss": 0.4545,
      "step": 8667
    },
    {
      "epoch": 1.86,
      "grad_norm": 8.845823287963867,
      "learning_rate": 2.110682110682111e-06,
      "loss": 0.4892,
      "step": 8668
    },
    {
      "epoch": 1.86,
      "grad_norm": 13.730112075805664,
      "learning_rate": 2.1074646074646075e-06,
      "loss": 0.474,
      "step": 8669
    },
    {
      "epoch": 1.86,
      "grad_norm": 12.138509750366211,
      "learning_rate": 2.1042471042471044e-06,
      "loss": 0.5166,
      "step": 8670
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.36009407043457,
      "learning_rate": 2.101029601029601e-06,
      "loss": 0.2385,
      "step": 8671
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.047723770141602,
      "learning_rate": 2.097812097812098e-06,
      "loss": 0.5555,
      "step": 8672
    },
    {
      "epoch": 1.86,
      "grad_norm": 11.661168098449707,
      "learning_rate": 2.0945945945945946e-06,
      "loss": 0.6807,
      "step": 8673
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.436773300170898,
      "learning_rate": 2.0913770913770915e-06,
      "loss": 0.425,
      "step": 8674
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.930255889892578,
      "learning_rate": 2.088159588159588e-06,
      "loss": 0.5557,
      "step": 8675
    },
    {
      "epoch": 1.86,
      "grad_norm": 7.739737033843994,
      "learning_rate": 2.0849420849420848e-06,
      "loss": 0.3966,
      "step": 8676
    },
    {
      "epoch": 1.86,
      "grad_norm": 6.787098407745361,
      "learning_rate": 2.081724581724582e-06,
      "loss": 0.3219,
      "step": 8677
    },
    {
      "epoch": 1.86,
      "grad_norm": 11.267870903015137,
      "learning_rate": 2.0785070785070785e-06,
      "loss": 0.5135,
      "step": 8678
    },
    {
      "epoch": 1.86,
      "grad_norm": 13.425273895263672,
      "learning_rate": 2.0752895752895754e-06,
      "loss": 0.5408,
      "step": 8679
    },
    {
      "epoch": 1.86,
      "grad_norm": 11.525093078613281,
      "learning_rate": 2.072072072072072e-06,
      "loss": 0.554,
      "step": 8680
    },
    {
      "epoch": 1.86,
      "grad_norm": 17.49736213684082,
      "learning_rate": 2.068854568854569e-06,
      "loss": 0.78,
      "step": 8681
    },
    {
      "epoch": 1.86,
      "grad_norm": 8.852110862731934,
      "learning_rate": 2.0656370656370656e-06,
      "loss": 0.3719,
      "step": 8682
    },
    {
      "epoch": 1.86,
      "grad_norm": 9.122016906738281,
      "learning_rate": 2.0624195624195625e-06,
      "loss": 0.3601,
      "step": 8683
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.451210021972656,
      "learning_rate": 2.059202059202059e-06,
      "loss": 0.5895,
      "step": 8684
    },
    {
      "epoch": 1.86,
      "grad_norm": 13.215267181396484,
      "learning_rate": 2.0559845559845562e-06,
      "loss": 0.6827,
      "step": 8685
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.320707321166992,
      "learning_rate": 2.0527670527670527e-06,
      "loss": 0.5617,
      "step": 8686
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.987252235412598,
      "learning_rate": 2.0495495495495496e-06,
      "loss": 0.4238,
      "step": 8687
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.615214347839355,
      "learning_rate": 2.046332046332046e-06,
      "loss": 0.5365,
      "step": 8688
    },
    {
      "epoch": 1.86,
      "grad_norm": 15.04124927520752,
      "learning_rate": 2.0431145431145433e-06,
      "loss": 0.6404,
      "step": 8689
    },
    {
      "epoch": 1.86,
      "grad_norm": 11.388566970825195,
      "learning_rate": 2.03989703989704e-06,
      "loss": 0.4856,
      "step": 8690
    },
    {
      "epoch": 1.86,
      "grad_norm": 12.17077350616455,
      "learning_rate": 2.0366795366795366e-06,
      "loss": 0.4749,
      "step": 8691
    },
    {
      "epoch": 1.86,
      "grad_norm": 10.051279067993164,
      "learning_rate": 2.0334620334620335e-06,
      "loss": 0.3131,
      "step": 8692
    },
    {
      "epoch": 1.86,
      "grad_norm": 9.625734329223633,
      "learning_rate": 2.0302445302445304e-06,
      "loss": 0.413,
      "step": 8693
    },
    {
      "epoch": 1.86,
      "grad_norm": 8.272562980651855,
      "learning_rate": 2.0270270270270273e-06,
      "loss": 0.3827,
      "step": 8694
    },
    {
      "epoch": 1.87,
      "grad_norm": 13.207551956176758,
      "learning_rate": 2.0238095238095237e-06,
      "loss": 0.504,
      "step": 8695
    },
    {
      "epoch": 1.87,
      "grad_norm": 12.204014778137207,
      "learning_rate": 2.0205920205920206e-06,
      "loss": 0.8218,
      "step": 8696
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.049127578735352,
      "learning_rate": 2.0173745173745175e-06,
      "loss": 0.3809,
      "step": 8697
    },
    {
      "epoch": 1.87,
      "grad_norm": 10.963434219360352,
      "learning_rate": 2.0141570141570143e-06,
      "loss": 0.3951,
      "step": 8698
    },
    {
      "epoch": 1.87,
      "grad_norm": 9.768683433532715,
      "learning_rate": 2.010939510939511e-06,
      "loss": 0.4144,
      "step": 8699
    },
    {
      "epoch": 1.87,
      "grad_norm": 9.301487922668457,
      "learning_rate": 2.0077220077220077e-06,
      "loss": 0.3734,
      "step": 8700
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.956038475036621,
      "learning_rate": 2.004504504504505e-06,
      "loss": 0.4516,
      "step": 8701
    },
    {
      "epoch": 1.87,
      "grad_norm": 10.520015716552734,
      "learning_rate": 2.0012870012870014e-06,
      "loss": 0.4294,
      "step": 8702
    },
    {
      "epoch": 1.87,
      "grad_norm": 11.182358741760254,
      "learning_rate": 1.9980694980694983e-06,
      "loss": 0.4407,
      "step": 8703
    },
    {
      "epoch": 1.87,
      "grad_norm": 9.725834846496582,
      "learning_rate": 1.9948519948519947e-06,
      "loss": 0.3612,
      "step": 8704
    },
    {
      "epoch": 1.87,
      "grad_norm": 7.769437313079834,
      "learning_rate": 1.9916344916344916e-06,
      "loss": 0.248,
      "step": 8705
    },
    {
      "epoch": 1.87,
      "grad_norm": 11.128263473510742,
      "learning_rate": 1.9884169884169885e-06,
      "loss": 0.409,
      "step": 8706
    },
    {
      "epoch": 1.87,
      "grad_norm": 9.274724006652832,
      "learning_rate": 1.9851994851994854e-06,
      "loss": 0.5715,
      "step": 8707
    },
    {
      "epoch": 1.87,
      "grad_norm": 12.123099327087402,
      "learning_rate": 1.981981981981982e-06,
      "loss": 0.7314,
      "step": 8708
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.938173294067383,
      "learning_rate": 1.9787644787644787e-06,
      "loss": 0.4618,
      "step": 8709
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.812295913696289,
      "learning_rate": 1.9755469755469756e-06,
      "loss": 0.61,
      "step": 8710
    },
    {
      "epoch": 1.87,
      "grad_norm": 12.832316398620605,
      "learning_rate": 1.9723294723294724e-06,
      "loss": 0.4565,
      "step": 8711
    },
    {
      "epoch": 1.87,
      "grad_norm": 11.022101402282715,
      "learning_rate": 1.969111969111969e-06,
      "loss": 0.5122,
      "step": 8712
    },
    {
      "epoch": 1.87,
      "grad_norm": 10.544683456420898,
      "learning_rate": 1.9658944658944658e-06,
      "loss": 0.522,
      "step": 8713
    },
    {
      "epoch": 1.87,
      "grad_norm": 13.171460151672363,
      "learning_rate": 1.962676962676963e-06,
      "loss": 0.481,
      "step": 8714
    },
    {
      "epoch": 1.87,
      "grad_norm": 11.503893852233887,
      "learning_rate": 1.9594594594594595e-06,
      "loss": 0.5135,
      "step": 8715
    },
    {
      "epoch": 1.87,
      "grad_norm": 12.551074981689453,
      "learning_rate": 1.9562419562419564e-06,
      "loss": 0.6234,
      "step": 8716
    },
    {
      "epoch": 1.87,
      "grad_norm": 7.903550624847412,
      "learning_rate": 1.953024453024453e-06,
      "loss": 0.2749,
      "step": 8717
    },
    {
      "epoch": 1.87,
      "grad_norm": 10.30753231048584,
      "learning_rate": 1.94980694980695e-06,
      "loss": 0.3466,
      "step": 8718
    },
    {
      "epoch": 1.87,
      "grad_norm": 9.995673179626465,
      "learning_rate": 1.9465894465894466e-06,
      "loss": 0.4439,
      "step": 8719
    },
    {
      "epoch": 1.87,
      "grad_norm": 10.974950790405273,
      "learning_rate": 1.9433719433719435e-06,
      "loss": 0.4879,
      "step": 8720
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.718350410461426,
      "learning_rate": 1.94015444015444e-06,
      "loss": 0.3613,
      "step": 8721
    },
    {
      "epoch": 1.87,
      "grad_norm": 11.96205997467041,
      "learning_rate": 1.9369369369369372e-06,
      "loss": 0.6029,
      "step": 8722
    },
    {
      "epoch": 1.87,
      "grad_norm": 15.677772521972656,
      "learning_rate": 1.9337194337194337e-06,
      "loss": 0.8406,
      "step": 8723
    },
    {
      "epoch": 1.87,
      "grad_norm": 10.407610893249512,
      "learning_rate": 1.9305019305019305e-06,
      "loss": 0.4612,
      "step": 8724
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.637035369873047,
      "learning_rate": 1.927284427284427e-06,
      "loss": 0.3559,
      "step": 8725
    },
    {
      "epoch": 1.87,
      "grad_norm": 11.171377182006836,
      "learning_rate": 1.9240669240669243e-06,
      "loss": 0.5593,
      "step": 8726
    },
    {
      "epoch": 1.87,
      "grad_norm": 13.497333526611328,
      "learning_rate": 1.920849420849421e-06,
      "loss": 0.6576,
      "step": 8727
    },
    {
      "epoch": 1.87,
      "grad_norm": 11.230589866638184,
      "learning_rate": 1.9176319176319176e-06,
      "loss": 0.4842,
      "step": 8728
    },
    {
      "epoch": 1.87,
      "grad_norm": 9.065217018127441,
      "learning_rate": 1.9144144144144145e-06,
      "loss": 0.4182,
      "step": 8729
    },
    {
      "epoch": 1.87,
      "grad_norm": 13.99914836883545,
      "learning_rate": 1.9111969111969114e-06,
      "loss": 0.4361,
      "step": 8730
    },
    {
      "epoch": 1.87,
      "grad_norm": 11.689861297607422,
      "learning_rate": 1.9079794079794082e-06,
      "loss": 0.5223,
      "step": 8731
    },
    {
      "epoch": 1.87,
      "grad_norm": 9.824408531188965,
      "learning_rate": 1.9047619047619047e-06,
      "loss": 0.6871,
      "step": 8732
    },
    {
      "epoch": 1.87,
      "grad_norm": 11.531563758850098,
      "learning_rate": 1.9015444015444018e-06,
      "loss": 0.4994,
      "step": 8733
    },
    {
      "epoch": 1.87,
      "grad_norm": 13.007808685302734,
      "learning_rate": 1.8983268983268982e-06,
      "loss": 0.6607,
      "step": 8734
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.837922096252441,
      "learning_rate": 1.8951093951093951e-06,
      "loss": 0.2863,
      "step": 8735
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.7832670211792,
      "learning_rate": 1.8918918918918918e-06,
      "loss": 0.4514,
      "step": 8736
    },
    {
      "epoch": 1.87,
      "grad_norm": 12.424203872680664,
      "learning_rate": 1.8886743886743886e-06,
      "loss": 0.464,
      "step": 8737
    },
    {
      "epoch": 1.87,
      "grad_norm": 12.1986665725708,
      "learning_rate": 1.8854568854568853e-06,
      "loss": 0.4443,
      "step": 8738
    },
    {
      "epoch": 1.87,
      "grad_norm": 10.138498306274414,
      "learning_rate": 1.8822393822393822e-06,
      "loss": 0.478,
      "step": 8739
    },
    {
      "epoch": 1.87,
      "grad_norm": 11.662186622619629,
      "learning_rate": 1.8790218790218793e-06,
      "loss": 0.5941,
      "step": 8740
    },
    {
      "epoch": 1.87,
      "grad_norm": 11.880834579467773,
      "learning_rate": 1.8758043758043757e-06,
      "loss": 0.5455,
      "step": 8741
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.755162239074707,
      "learning_rate": 1.8725868725868726e-06,
      "loss": 0.3858,
      "step": 8742
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.71273422241211,
      "learning_rate": 1.8693693693693693e-06,
      "loss": 0.4617,
      "step": 8743
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.093693733215332,
      "learning_rate": 1.8661518661518661e-06,
      "loss": 0.4655,
      "step": 8744
    },
    {
      "epoch": 1.88,
      "grad_norm": 8.97195816040039,
      "learning_rate": 1.862934362934363e-06,
      "loss": 0.4946,
      "step": 8745
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.00629997253418,
      "learning_rate": 1.8597168597168599e-06,
      "loss": 0.5141,
      "step": 8746
    },
    {
      "epoch": 1.88,
      "grad_norm": 9.843446731567383,
      "learning_rate": 1.8564993564993565e-06,
      "loss": 0.4937,
      "step": 8747
    },
    {
      "epoch": 1.88,
      "grad_norm": 12.18907356262207,
      "learning_rate": 1.8532818532818534e-06,
      "loss": 0.5375,
      "step": 8748
    },
    {
      "epoch": 1.88,
      "grad_norm": 9.061447143554688,
      "learning_rate": 1.85006435006435e-06,
      "loss": 0.3982,
      "step": 8749
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.678226470947266,
      "learning_rate": 1.846846846846847e-06,
      "loss": 0.474,
      "step": 8750
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.334227561950684,
      "learning_rate": 1.8436293436293436e-06,
      "loss": 0.5445,
      "step": 8751
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.778157234191895,
      "learning_rate": 1.8404118404118405e-06,
      "loss": 0.5111,
      "step": 8752
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.743410110473633,
      "learning_rate": 1.8371943371943372e-06,
      "loss": 0.4886,
      "step": 8753
    },
    {
      "epoch": 1.88,
      "grad_norm": 13.35010051727295,
      "learning_rate": 1.833976833976834e-06,
      "loss": 0.6357,
      "step": 8754
    },
    {
      "epoch": 1.88,
      "grad_norm": 8.636313438415527,
      "learning_rate": 1.8307593307593307e-06,
      "loss": 0.3993,
      "step": 8755
    },
    {
      "epoch": 1.88,
      "grad_norm": 12.028886795043945,
      "learning_rate": 1.8275418275418276e-06,
      "loss": 0.4979,
      "step": 8756
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.179948806762695,
      "learning_rate": 1.8243243243243245e-06,
      "loss": 0.3877,
      "step": 8757
    },
    {
      "epoch": 1.88,
      "grad_norm": 12.39285945892334,
      "learning_rate": 1.8211068211068213e-06,
      "loss": 0.4711,
      "step": 8758
    },
    {
      "epoch": 1.88,
      "grad_norm": 14.138014793395996,
      "learning_rate": 1.817889317889318e-06,
      "loss": 0.6146,
      "step": 8759
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.166512489318848,
      "learning_rate": 1.8146718146718149e-06,
      "loss": 0.6267,
      "step": 8760
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.315607070922852,
      "learning_rate": 1.8114543114543115e-06,
      "loss": 0.4769,
      "step": 8761
    },
    {
      "epoch": 1.88,
      "grad_norm": 9.827555656433105,
      "learning_rate": 1.8082368082368084e-06,
      "loss": 0.4718,
      "step": 8762
    },
    {
      "epoch": 1.88,
      "grad_norm": 9.243650436401367,
      "learning_rate": 1.805019305019305e-06,
      "loss": 0.5102,
      "step": 8763
    },
    {
      "epoch": 1.88,
      "grad_norm": 9.871223449707031,
      "learning_rate": 1.8018018018018017e-06,
      "loss": 0.5483,
      "step": 8764
    },
    {
      "epoch": 1.88,
      "grad_norm": 14.537323951721191,
      "learning_rate": 1.7985842985842986e-06,
      "loss": 0.6319,
      "step": 8765
    },
    {
      "epoch": 1.88,
      "grad_norm": 14.46630573272705,
      "learning_rate": 1.7953667953667953e-06,
      "loss": 0.598,
      "step": 8766
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.910676002502441,
      "learning_rate": 1.7921492921492921e-06,
      "loss": 0.4588,
      "step": 8767
    },
    {
      "epoch": 1.88,
      "grad_norm": 18.06475067138672,
      "learning_rate": 1.7889317889317888e-06,
      "loss": 0.534,
      "step": 8768
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.35988712310791,
      "learning_rate": 1.7857142857142857e-06,
      "loss": 0.6103,
      "step": 8769
    },
    {
      "epoch": 1.88,
      "grad_norm": 6.3410844802856445,
      "learning_rate": 1.7824967824967826e-06,
      "loss": 0.2079,
      "step": 8770
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.451010704040527,
      "learning_rate": 1.7792792792792794e-06,
      "loss": 0.5144,
      "step": 8771
    },
    {
      "epoch": 1.88,
      "grad_norm": 8.086788177490234,
      "learning_rate": 1.776061776061776e-06,
      "loss": 0.5057,
      "step": 8772
    },
    {
      "epoch": 1.88,
      "grad_norm": 9.12031078338623,
      "learning_rate": 1.772844272844273e-06,
      "loss": 0.4395,
      "step": 8773
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.519620895385742,
      "learning_rate": 1.7696267696267696e-06,
      "loss": 0.4133,
      "step": 8774
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.539241790771484,
      "learning_rate": 1.7664092664092665e-06,
      "loss": 0.4443,
      "step": 8775
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.952920913696289,
      "learning_rate": 1.7631917631917632e-06,
      "loss": 0.5691,
      "step": 8776
    },
    {
      "epoch": 1.88,
      "grad_norm": 13.767050743103027,
      "learning_rate": 1.75997425997426e-06,
      "loss": 0.7545,
      "step": 8777
    },
    {
      "epoch": 1.88,
      "grad_norm": 8.723671913146973,
      "learning_rate": 1.7567567567567567e-06,
      "loss": 0.4706,
      "step": 8778
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.037862777709961,
      "learning_rate": 1.7535392535392536e-06,
      "loss": 0.4896,
      "step": 8779
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.45051383972168,
      "learning_rate": 1.7503217503217502e-06,
      "loss": 0.5458,
      "step": 8780
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.003470420837402,
      "learning_rate": 1.7471042471042471e-06,
      "loss": 0.4826,
      "step": 8781
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.94933032989502,
      "learning_rate": 1.743886743886744e-06,
      "loss": 0.5658,
      "step": 8782
    },
    {
      "epoch": 1.88,
      "grad_norm": 9.177206993103027,
      "learning_rate": 1.7406692406692409e-06,
      "loss": 0.5459,
      "step": 8783
    },
    {
      "epoch": 1.88,
      "grad_norm": 15.3768310546875,
      "learning_rate": 1.7374517374517375e-06,
      "loss": 0.6103,
      "step": 8784
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.320843696594238,
      "learning_rate": 1.7342342342342344e-06,
      "loss": 0.5894,
      "step": 8785
    },
    {
      "epoch": 1.88,
      "grad_norm": 10.950011253356934,
      "learning_rate": 1.731016731016731e-06,
      "loss": 0.4792,
      "step": 8786
    },
    {
      "epoch": 1.88,
      "grad_norm": 12.636842727661133,
      "learning_rate": 1.727799227799228e-06,
      "loss": 0.7399,
      "step": 8787
    },
    {
      "epoch": 1.89,
      "grad_norm": 8.279215812683105,
      "learning_rate": 1.7245817245817246e-06,
      "loss": 0.3542,
      "step": 8788
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.941113471984863,
      "learning_rate": 1.7213642213642215e-06,
      "loss": 0.5685,
      "step": 8789
    },
    {
      "epoch": 1.89,
      "grad_norm": 8.563420295715332,
      "learning_rate": 1.7181467181467181e-06,
      "loss": 0.3812,
      "step": 8790
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.354174613952637,
      "learning_rate": 1.714929214929215e-06,
      "loss": 0.4251,
      "step": 8791
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.939016342163086,
      "learning_rate": 1.7117117117117117e-06,
      "loss": 0.5666,
      "step": 8792
    },
    {
      "epoch": 1.89,
      "grad_norm": 13.69731330871582,
      "learning_rate": 1.7084942084942086e-06,
      "loss": 0.668,
      "step": 8793
    },
    {
      "epoch": 1.89,
      "grad_norm": 6.587306499481201,
      "learning_rate": 1.7052767052767052e-06,
      "loss": 0.3834,
      "step": 8794
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.712138175964355,
      "learning_rate": 1.702059202059202e-06,
      "loss": 0.5615,
      "step": 8795
    },
    {
      "epoch": 1.89,
      "grad_norm": 11.021648406982422,
      "learning_rate": 1.698841698841699e-06,
      "loss": 0.4962,
      "step": 8796
    },
    {
      "epoch": 1.89,
      "grad_norm": 8.767364501953125,
      "learning_rate": 1.6956241956241956e-06,
      "loss": 0.4831,
      "step": 8797
    },
    {
      "epoch": 1.89,
      "grad_norm": 11.064661979675293,
      "learning_rate": 1.6924066924066925e-06,
      "loss": 0.5959,
      "step": 8798
    },
    {
      "epoch": 1.89,
      "grad_norm": 8.49501895904541,
      "learning_rate": 1.6891891891891892e-06,
      "loss": 0.3653,
      "step": 8799
    },
    {
      "epoch": 1.89,
      "grad_norm": 8.850297927856445,
      "learning_rate": 1.685971685971686e-06,
      "loss": 0.541,
      "step": 8800
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.157611846923828,
      "learning_rate": 1.6827541827541827e-06,
      "loss": 0.4373,
      "step": 8801
    },
    {
      "epoch": 1.89,
      "grad_norm": 11.016095161437988,
      "learning_rate": 1.6795366795366796e-06,
      "loss": 0.4643,
      "step": 8802
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.583808898925781,
      "learning_rate": 1.6763191763191762e-06,
      "loss": 0.4856,
      "step": 8803
    },
    {
      "epoch": 1.89,
      "grad_norm": 11.315680503845215,
      "learning_rate": 1.6731016731016731e-06,
      "loss": 0.6494,
      "step": 8804
    },
    {
      "epoch": 1.89,
      "grad_norm": 11.252931594848633,
      "learning_rate": 1.6698841698841698e-06,
      "loss": 0.5573,
      "step": 8805
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.701258659362793,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.5361,
      "step": 8806
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.674838066101074,
      "learning_rate": 1.6634491634491635e-06,
      "loss": 0.4294,
      "step": 8807
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.659982681274414,
      "learning_rate": 1.6602316602316604e-06,
      "loss": 0.4422,
      "step": 8808
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.4617919921875,
      "learning_rate": 1.657014157014157e-06,
      "loss": 0.593,
      "step": 8809
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.384109497070312,
      "learning_rate": 1.653796653796654e-06,
      "loss": 0.3979,
      "step": 8810
    },
    {
      "epoch": 1.89,
      "grad_norm": 13.411449432373047,
      "learning_rate": 1.6505791505791506e-06,
      "loss": 0.5381,
      "step": 8811
    },
    {
      "epoch": 1.89,
      "grad_norm": 7.1511688232421875,
      "learning_rate": 1.6473616473616475e-06,
      "loss": 0.3455,
      "step": 8812
    },
    {
      "epoch": 1.89,
      "grad_norm": 11.835088729858398,
      "learning_rate": 1.6441441441441442e-06,
      "loss": 0.5728,
      "step": 8813
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.087495803833008,
      "learning_rate": 1.640926640926641e-06,
      "loss": 0.5104,
      "step": 8814
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.985125541687012,
      "learning_rate": 1.6377091377091377e-06,
      "loss": 0.5802,
      "step": 8815
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.150833129882812,
      "learning_rate": 1.6344916344916346e-06,
      "loss": 0.4116,
      "step": 8816
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.126520156860352,
      "learning_rate": 1.6312741312741312e-06,
      "loss": 0.4758,
      "step": 8817
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.050145149230957,
      "learning_rate": 1.628056628056628e-06,
      "loss": 0.368,
      "step": 8818
    },
    {
      "epoch": 1.89,
      "grad_norm": 9.852734565734863,
      "learning_rate": 1.6248391248391248e-06,
      "loss": 0.4917,
      "step": 8819
    },
    {
      "epoch": 1.89,
      "grad_norm": 14.199252128601074,
      "learning_rate": 1.6216216216216219e-06,
      "loss": 0.5315,
      "step": 8820
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.135322570800781,
      "learning_rate": 1.6184041184041185e-06,
      "loss": 0.6191,
      "step": 8821
    },
    {
      "epoch": 1.89,
      "grad_norm": 14.59547233581543,
      "learning_rate": 1.6151866151866154e-06,
      "loss": 0.6287,
      "step": 8822
    },
    {
      "epoch": 1.89,
      "grad_norm": 11.51407241821289,
      "learning_rate": 1.611969111969112e-06,
      "loss": 0.6651,
      "step": 8823
    },
    {
      "epoch": 1.89,
      "grad_norm": 14.238398551940918,
      "learning_rate": 1.6087516087516087e-06,
      "loss": 0.5425,
      "step": 8824
    },
    {
      "epoch": 1.89,
      "grad_norm": 14.270798683166504,
      "learning_rate": 1.6055341055341056e-06,
      "loss": 0.5915,
      "step": 8825
    },
    {
      "epoch": 1.89,
      "grad_norm": 12.864799499511719,
      "learning_rate": 1.6023166023166023e-06,
      "loss": 0.4971,
      "step": 8826
    },
    {
      "epoch": 1.89,
      "grad_norm": 11.121648788452148,
      "learning_rate": 1.5990990990990991e-06,
      "loss": 0.7356,
      "step": 8827
    },
    {
      "epoch": 1.89,
      "grad_norm": 14.520915985107422,
      "learning_rate": 1.5958815958815958e-06,
      "loss": 0.4952,
      "step": 8828
    },
    {
      "epoch": 1.89,
      "grad_norm": 11.3899564743042,
      "learning_rate": 1.5926640926640927e-06,
      "loss": 0.5666,
      "step": 8829
    },
    {
      "epoch": 1.89,
      "grad_norm": 11.361396789550781,
      "learning_rate": 1.5894465894465893e-06,
      "loss": 0.53,
      "step": 8830
    },
    {
      "epoch": 1.89,
      "grad_norm": 11.71109676361084,
      "learning_rate": 1.5862290862290862e-06,
      "loss": 0.5501,
      "step": 8831
    },
    {
      "epoch": 1.89,
      "grad_norm": 8.913434982299805,
      "learning_rate": 1.583011583011583e-06,
      "loss": 0.4581,
      "step": 8832
    },
    {
      "epoch": 1.89,
      "grad_norm": 8.741549491882324,
      "learning_rate": 1.57979407979408e-06,
      "loss": 0.3413,
      "step": 8833
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.094813346862793,
      "learning_rate": 1.5765765765765766e-06,
      "loss": 0.5312,
      "step": 8834
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.925127029418945,
      "learning_rate": 1.5733590733590735e-06,
      "loss": 0.4747,
      "step": 8835
    },
    {
      "epoch": 1.9,
      "grad_norm": 8.638671875,
      "learning_rate": 1.5701415701415702e-06,
      "loss": 0.4906,
      "step": 8836
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.245434761047363,
      "learning_rate": 1.566924066924067e-06,
      "loss": 0.3449,
      "step": 8837
    },
    {
      "epoch": 1.9,
      "grad_norm": 7.406683444976807,
      "learning_rate": 1.5637065637065637e-06,
      "loss": 0.2422,
      "step": 8838
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.339016914367676,
      "learning_rate": 1.5604890604890606e-06,
      "loss": 0.5982,
      "step": 8839
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.862789154052734,
      "learning_rate": 1.5572715572715572e-06,
      "loss": 0.6191,
      "step": 8840
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.918458938598633,
      "learning_rate": 1.5540540540540541e-06,
      "loss": 0.3993,
      "step": 8841
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.74767017364502,
      "learning_rate": 1.5508365508365508e-06,
      "loss": 0.6388,
      "step": 8842
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.395136833190918,
      "learning_rate": 1.5476190476190476e-06,
      "loss": 0.4143,
      "step": 8843
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.458680152893066,
      "learning_rate": 1.5444015444015443e-06,
      "loss": 0.3948,
      "step": 8844
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.367324829101562,
      "learning_rate": 1.5411840411840414e-06,
      "loss": 0.3037,
      "step": 8845
    },
    {
      "epoch": 1.9,
      "grad_norm": 14.672276496887207,
      "learning_rate": 1.537966537966538e-06,
      "loss": 0.4025,
      "step": 8846
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.251364707946777,
      "learning_rate": 1.534749034749035e-06,
      "loss": 0.7138,
      "step": 8847
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.261743545532227,
      "learning_rate": 1.5315315315315316e-06,
      "loss": 0.5157,
      "step": 8848
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.1831693649292,
      "learning_rate": 1.5283140283140285e-06,
      "loss": 0.41,
      "step": 8849
    },
    {
      "epoch": 1.9,
      "grad_norm": 15.280720710754395,
      "learning_rate": 1.5250965250965251e-06,
      "loss": 0.5549,
      "step": 8850
    },
    {
      "epoch": 1.9,
      "grad_norm": 13.395429611206055,
      "learning_rate": 1.521879021879022e-06,
      "loss": 0.718,
      "step": 8851
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.564294815063477,
      "learning_rate": 1.5186615186615187e-06,
      "loss": 0.3685,
      "step": 8852
    },
    {
      "epoch": 1.9,
      "grad_norm": 13.159331321716309,
      "learning_rate": 1.5154440154440153e-06,
      "loss": 0.391,
      "step": 8853
    },
    {
      "epoch": 1.9,
      "grad_norm": 8.117081642150879,
      "learning_rate": 1.5122265122265122e-06,
      "loss": 0.2512,
      "step": 8854
    },
    {
      "epoch": 1.9,
      "grad_norm": 8.90794563293457,
      "learning_rate": 1.5090090090090089e-06,
      "loss": 0.5429,
      "step": 8855
    },
    {
      "epoch": 1.9,
      "grad_norm": 15.016254425048828,
      "learning_rate": 1.5057915057915057e-06,
      "loss": 0.3391,
      "step": 8856
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.035867691040039,
      "learning_rate": 1.5025740025740026e-06,
      "loss": 0.4806,
      "step": 8857
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.666963577270508,
      "learning_rate": 1.4993564993564995e-06,
      "loss": 0.4493,
      "step": 8858
    },
    {
      "epoch": 1.9,
      "grad_norm": 12.852065086364746,
      "learning_rate": 1.4961389961389962e-06,
      "loss": 0.3762,
      "step": 8859
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.016260147094727,
      "learning_rate": 1.492921492921493e-06,
      "loss": 0.506,
      "step": 8860
    },
    {
      "epoch": 1.9,
      "grad_norm": 14.814437866210938,
      "learning_rate": 1.4897039897039897e-06,
      "loss": 0.593,
      "step": 8861
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.013469696044922,
      "learning_rate": 1.4864864864864866e-06,
      "loss": 0.5058,
      "step": 8862
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.26519775390625,
      "learning_rate": 1.4832689832689832e-06,
      "loss": 0.4545,
      "step": 8863
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.64936351776123,
      "learning_rate": 1.4800514800514801e-06,
      "loss": 0.3744,
      "step": 8864
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.863261222839355,
      "learning_rate": 1.4768339768339768e-06,
      "loss": 0.5141,
      "step": 8865
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.025975227355957,
      "learning_rate": 1.4736164736164736e-06,
      "loss": 0.4849,
      "step": 8866
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.83883285522461,
      "learning_rate": 1.4703989703989703e-06,
      "loss": 0.5401,
      "step": 8867
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.694555282592773,
      "learning_rate": 1.4671814671814672e-06,
      "loss": 0.4492,
      "step": 8868
    },
    {
      "epoch": 1.9,
      "grad_norm": 12.25759220123291,
      "learning_rate": 1.4639639639639639e-06,
      "loss": 0.5845,
      "step": 8869
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.448917388916016,
      "learning_rate": 1.460746460746461e-06,
      "loss": 0.5199,
      "step": 8870
    },
    {
      "epoch": 1.9,
      "grad_norm": 8.96464729309082,
      "learning_rate": 1.4575289575289576e-06,
      "loss": 0.4331,
      "step": 8871
    },
    {
      "epoch": 1.9,
      "grad_norm": 13.249844551086426,
      "learning_rate": 1.4543114543114545e-06,
      "loss": 0.4088,
      "step": 8872
    },
    {
      "epoch": 1.9,
      "grad_norm": 10.893659591674805,
      "learning_rate": 1.4510939510939511e-06,
      "loss": 0.5764,
      "step": 8873
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.490081787109375,
      "learning_rate": 1.447876447876448e-06,
      "loss": 0.4867,
      "step": 8874
    },
    {
      "epoch": 1.9,
      "grad_norm": 11.215291976928711,
      "learning_rate": 1.4446589446589447e-06,
      "loss": 0.5404,
      "step": 8875
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.089798927307129,
      "learning_rate": 1.4414414414414416e-06,
      "loss": 0.5419,
      "step": 8876
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.18283748626709,
      "learning_rate": 1.4382239382239382e-06,
      "loss": 0.4029,
      "step": 8877
    },
    {
      "epoch": 1.9,
      "grad_norm": 8.735549926757812,
      "learning_rate": 1.435006435006435e-06,
      "loss": 0.4043,
      "step": 8878
    },
    {
      "epoch": 1.9,
      "grad_norm": 8.729789733886719,
      "learning_rate": 1.4317889317889318e-06,
      "loss": 0.392,
      "step": 8879
    },
    {
      "epoch": 1.9,
      "grad_norm": 12.104376792907715,
      "learning_rate": 1.4285714285714286e-06,
      "loss": 0.7032,
      "step": 8880
    },
    {
      "epoch": 1.9,
      "grad_norm": 14.696111679077148,
      "learning_rate": 1.4253539253539253e-06,
      "loss": 0.5427,
      "step": 8881
    },
    {
      "epoch": 1.91,
      "grad_norm": 11.587226867675781,
      "learning_rate": 1.4221364221364222e-06,
      "loss": 0.6676,
      "step": 8882
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.123538970947266,
      "learning_rate": 1.418918918918919e-06,
      "loss": 0.3429,
      "step": 8883
    },
    {
      "epoch": 1.91,
      "grad_norm": 11.51640510559082,
      "learning_rate": 1.4157014157014157e-06,
      "loss": 0.764,
      "step": 8884
    },
    {
      "epoch": 1.91,
      "grad_norm": 11.977684020996094,
      "learning_rate": 1.4124839124839126e-06,
      "loss": 0.4287,
      "step": 8885
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.352004051208496,
      "learning_rate": 1.4092664092664092e-06,
      "loss": 0.6338,
      "step": 8886
    },
    {
      "epoch": 1.91,
      "grad_norm": 11.866009712219238,
      "learning_rate": 1.4060489060489061e-06,
      "loss": 0.6302,
      "step": 8887
    },
    {
      "epoch": 1.91,
      "grad_norm": 13.097360610961914,
      "learning_rate": 1.4028314028314028e-06,
      "loss": 0.5774,
      "step": 8888
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.122404098510742,
      "learning_rate": 1.3996138996138997e-06,
      "loss": 0.5084,
      "step": 8889
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.877086639404297,
      "learning_rate": 1.3963963963963963e-06,
      "loss": 0.4581,
      "step": 8890
    },
    {
      "epoch": 1.91,
      "grad_norm": 9.20517635345459,
      "learning_rate": 1.3931788931788932e-06,
      "loss": 0.492,
      "step": 8891
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.613807678222656,
      "learning_rate": 1.3899613899613899e-06,
      "loss": 0.4609,
      "step": 8892
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.26860237121582,
      "learning_rate": 1.3867438867438867e-06,
      "loss": 0.5918,
      "step": 8893
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.050334930419922,
      "learning_rate": 1.3835263835263836e-06,
      "loss": 0.5783,
      "step": 8894
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.114643096923828,
      "learning_rate": 1.3803088803088805e-06,
      "loss": 0.4465,
      "step": 8895
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.06496524810791,
      "learning_rate": 1.3770913770913771e-06,
      "loss": 0.2786,
      "step": 8896
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.07131290435791,
      "learning_rate": 1.373873873873874e-06,
      "loss": 0.473,
      "step": 8897
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.856708526611328,
      "learning_rate": 1.3706563706563707e-06,
      "loss": 0.4287,
      "step": 8898
    },
    {
      "epoch": 1.91,
      "grad_norm": 7.141381740570068,
      "learning_rate": 1.3674388674388676e-06,
      "loss": 0.4677,
      "step": 8899
    },
    {
      "epoch": 1.91,
      "grad_norm": 7.92179012298584,
      "learning_rate": 1.3642213642213642e-06,
      "loss": 0.2895,
      "step": 8900
    },
    {
      "epoch": 1.91,
      "grad_norm": 7.5126752853393555,
      "learning_rate": 1.361003861003861e-06,
      "loss": 0.3533,
      "step": 8901
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.704422950744629,
      "learning_rate": 1.3577863577863578e-06,
      "loss": 0.6014,
      "step": 8902
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.30005168914795,
      "learning_rate": 1.3545688545688546e-06,
      "loss": 0.4269,
      "step": 8903
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.627790451049805,
      "learning_rate": 1.3513513513513513e-06,
      "loss": 0.4948,
      "step": 8904
    },
    {
      "epoch": 1.91,
      "grad_norm": 9.020503044128418,
      "learning_rate": 1.3481338481338482e-06,
      "loss": 0.4716,
      "step": 8905
    },
    {
      "epoch": 1.91,
      "grad_norm": 11.656542778015137,
      "learning_rate": 1.3449163449163448e-06,
      "loss": 0.5295,
      "step": 8906
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.956891059875488,
      "learning_rate": 1.341698841698842e-06,
      "loss": 0.5185,
      "step": 8907
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.883187294006348,
      "learning_rate": 1.3384813384813386e-06,
      "loss": 0.4796,
      "step": 8908
    },
    {
      "epoch": 1.91,
      "grad_norm": 13.939093589782715,
      "learning_rate": 1.3352638352638355e-06,
      "loss": 0.4024,
      "step": 8909
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.034184455871582,
      "learning_rate": 1.3320463320463321e-06,
      "loss": 0.3971,
      "step": 8910
    },
    {
      "epoch": 1.91,
      "grad_norm": 12.229240417480469,
      "learning_rate": 1.3288288288288288e-06,
      "loss": 0.4774,
      "step": 8911
    },
    {
      "epoch": 1.91,
      "grad_norm": 11.044054985046387,
      "learning_rate": 1.3256113256113257e-06,
      "loss": 0.3792,
      "step": 8912
    },
    {
      "epoch": 1.91,
      "grad_norm": 9.753396034240723,
      "learning_rate": 1.3223938223938223e-06,
      "loss": 0.6211,
      "step": 8913
    },
    {
      "epoch": 1.91,
      "grad_norm": 11.001375198364258,
      "learning_rate": 1.3191763191763192e-06,
      "loss": 0.4514,
      "step": 8914
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.622613906860352,
      "learning_rate": 1.3159588159588159e-06,
      "loss": 0.5294,
      "step": 8915
    },
    {
      "epoch": 1.91,
      "grad_norm": 17.059457778930664,
      "learning_rate": 1.3127413127413127e-06,
      "loss": 0.5499,
      "step": 8916
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.496405601501465,
      "learning_rate": 1.3095238095238094e-06,
      "loss": 0.4585,
      "step": 8917
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.09658432006836,
      "learning_rate": 1.3063063063063063e-06,
      "loss": 0.4617,
      "step": 8918
    },
    {
      "epoch": 1.91,
      "grad_norm": 13.169740676879883,
      "learning_rate": 1.3030888030888031e-06,
      "loss": 0.4304,
      "step": 8919
    },
    {
      "epoch": 1.91,
      "grad_norm": 9.630132675170898,
      "learning_rate": 1.2998712998713e-06,
      "loss": 0.3631,
      "step": 8920
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.229021072387695,
      "learning_rate": 1.2966537966537967e-06,
      "loss": 0.5493,
      "step": 8921
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.212311744689941,
      "learning_rate": 1.2934362934362936e-06,
      "loss": 0.3725,
      "step": 8922
    },
    {
      "epoch": 1.91,
      "grad_norm": 8.024866104125977,
      "learning_rate": 1.2902187902187902e-06,
      "loss": 0.4693,
      "step": 8923
    },
    {
      "epoch": 1.91,
      "grad_norm": 7.879472255706787,
      "learning_rate": 1.287001287001287e-06,
      "loss": 0.4489,
      "step": 8924
    },
    {
      "epoch": 1.91,
      "grad_norm": 13.355793952941895,
      "learning_rate": 1.2837837837837838e-06,
      "loss": 0.4296,
      "step": 8925
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.764801025390625,
      "learning_rate": 1.2805662805662806e-06,
      "loss": 0.471,
      "step": 8926
    },
    {
      "epoch": 1.91,
      "grad_norm": 11.790544509887695,
      "learning_rate": 1.2773487773487773e-06,
      "loss": 0.5598,
      "step": 8927
    },
    {
      "epoch": 1.92,
      "grad_norm": 11.870955467224121,
      "learning_rate": 1.2741312741312742e-06,
      "loss": 0.5659,
      "step": 8928
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.611364364624023,
      "learning_rate": 1.2709137709137708e-06,
      "loss": 0.6442,
      "step": 8929
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.901581764221191,
      "learning_rate": 1.2676962676962677e-06,
      "loss": 0.43,
      "step": 8930
    },
    {
      "epoch": 1.92,
      "grad_norm": 15.23404598236084,
      "learning_rate": 1.2644787644787644e-06,
      "loss": 0.6137,
      "step": 8931
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.138059616088867,
      "learning_rate": 1.2612612612612615e-06,
      "loss": 0.3484,
      "step": 8932
    },
    {
      "epoch": 1.92,
      "grad_norm": 8.26215934753418,
      "learning_rate": 1.2580437580437581e-06,
      "loss": 0.3199,
      "step": 8933
    },
    {
      "epoch": 1.92,
      "grad_norm": 6.928267002105713,
      "learning_rate": 1.254826254826255e-06,
      "loss": 0.2749,
      "step": 8934
    },
    {
      "epoch": 1.92,
      "grad_norm": 8.448274612426758,
      "learning_rate": 1.2516087516087517e-06,
      "loss": 0.3397,
      "step": 8935
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.780158042907715,
      "learning_rate": 1.2483912483912485e-06,
      "loss": 0.5583,
      "step": 8936
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.927079200744629,
      "learning_rate": 1.2451737451737452e-06,
      "loss": 0.4739,
      "step": 8937
    },
    {
      "epoch": 1.92,
      "grad_norm": 12.367884635925293,
      "learning_rate": 1.241956241956242e-06,
      "loss": 0.5297,
      "step": 8938
    },
    {
      "epoch": 1.92,
      "grad_norm": 12.295692443847656,
      "learning_rate": 1.2387387387387387e-06,
      "loss": 0.4396,
      "step": 8939
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.88158130645752,
      "learning_rate": 1.2355212355212356e-06,
      "loss": 0.4813,
      "step": 8940
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.147455215454102,
      "learning_rate": 1.2323037323037323e-06,
      "loss": 0.3437,
      "step": 8941
    },
    {
      "epoch": 1.92,
      "grad_norm": 13.01839542388916,
      "learning_rate": 1.229086229086229e-06,
      "loss": 0.4406,
      "step": 8942
    },
    {
      "epoch": 1.92,
      "grad_norm": 12.510418891906738,
      "learning_rate": 1.2258687258687258e-06,
      "loss": 0.758,
      "step": 8943
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.697504997253418,
      "learning_rate": 1.2226512226512227e-06,
      "loss": 0.5468,
      "step": 8944
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.11119270324707,
      "learning_rate": 1.2194337194337196e-06,
      "loss": 0.5152,
      "step": 8945
    },
    {
      "epoch": 1.92,
      "grad_norm": 12.258732795715332,
      "learning_rate": 1.2162162162162162e-06,
      "loss": 0.418,
      "step": 8946
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.341889381408691,
      "learning_rate": 1.212998712998713e-06,
      "loss": 0.5602,
      "step": 8947
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.986352920532227,
      "learning_rate": 1.2097812097812098e-06,
      "loss": 0.3173,
      "step": 8948
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.876452445983887,
      "learning_rate": 1.2065637065637066e-06,
      "loss": 0.5216,
      "step": 8949
    },
    {
      "epoch": 1.92,
      "grad_norm": 11.965291023254395,
      "learning_rate": 1.2033462033462033e-06,
      "loss": 0.5837,
      "step": 8950
    },
    {
      "epoch": 1.92,
      "grad_norm": 12.791353225708008,
      "learning_rate": 1.2001287001287002e-06,
      "loss": 0.4251,
      "step": 8951
    },
    {
      "epoch": 1.92,
      "grad_norm": 7.57658576965332,
      "learning_rate": 1.1969111969111968e-06,
      "loss": 0.4712,
      "step": 8952
    },
    {
      "epoch": 1.92,
      "grad_norm": 12.90672779083252,
      "learning_rate": 1.1936936936936937e-06,
      "loss": 0.5128,
      "step": 8953
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.161299705505371,
      "learning_rate": 1.1904761904761904e-06,
      "loss": 0.4354,
      "step": 8954
    },
    {
      "epoch": 1.92,
      "grad_norm": 8.27947998046875,
      "learning_rate": 1.1872586872586873e-06,
      "loss": 0.4772,
      "step": 8955
    },
    {
      "epoch": 1.92,
      "grad_norm": 11.10891342163086,
      "learning_rate": 1.184041184041184e-06,
      "loss": 0.4891,
      "step": 8956
    },
    {
      "epoch": 1.92,
      "grad_norm": 8.893481254577637,
      "learning_rate": 1.180823680823681e-06,
      "loss": 0.3663,
      "step": 8957
    },
    {
      "epoch": 1.92,
      "grad_norm": 8.83199405670166,
      "learning_rate": 1.1776061776061777e-06,
      "loss": 0.4364,
      "step": 8958
    },
    {
      "epoch": 1.92,
      "grad_norm": 13.548731803894043,
      "learning_rate": 1.1743886743886745e-06,
      "loss": 0.4748,
      "step": 8959
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.225077629089355,
      "learning_rate": 1.1711711711711712e-06,
      "loss": 0.6091,
      "step": 8960
    },
    {
      "epoch": 1.92,
      "grad_norm": 12.327251434326172,
      "learning_rate": 1.167953667953668e-06,
      "loss": 0.7886,
      "step": 8961
    },
    {
      "epoch": 1.92,
      "grad_norm": 12.845756530761719,
      "learning_rate": 1.1647361647361647e-06,
      "loss": 0.3877,
      "step": 8962
    },
    {
      "epoch": 1.92,
      "grad_norm": 12.05014705657959,
      "learning_rate": 1.1615186615186616e-06,
      "loss": 0.6459,
      "step": 8963
    },
    {
      "epoch": 1.92,
      "grad_norm": 8.640949249267578,
      "learning_rate": 1.1583011583011583e-06,
      "loss": 0.5924,
      "step": 8964
    },
    {
      "epoch": 1.92,
      "grad_norm": 7.970583438873291,
      "learning_rate": 1.1550836550836552e-06,
      "loss": 0.3313,
      "step": 8965
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.533285140991211,
      "learning_rate": 1.1518661518661518e-06,
      "loss": 0.4199,
      "step": 8966
    },
    {
      "epoch": 1.92,
      "grad_norm": 12.152959823608398,
      "learning_rate": 1.1486486486486487e-06,
      "loss": 0.6169,
      "step": 8967
    },
    {
      "epoch": 1.92,
      "grad_norm": 11.5794095993042,
      "learning_rate": 1.1454311454311454e-06,
      "loss": 0.7533,
      "step": 8968
    },
    {
      "epoch": 1.92,
      "grad_norm": 8.664268493652344,
      "learning_rate": 1.1422136422136424e-06,
      "loss": 0.4575,
      "step": 8969
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.114885330200195,
      "learning_rate": 1.1389961389961391e-06,
      "loss": 0.3507,
      "step": 8970
    },
    {
      "epoch": 1.92,
      "grad_norm": 11.922348976135254,
      "learning_rate": 1.1357786357786358e-06,
      "loss": 0.4426,
      "step": 8971
    },
    {
      "epoch": 1.92,
      "grad_norm": 9.510671615600586,
      "learning_rate": 1.1325611325611326e-06,
      "loss": 0.613,
      "step": 8972
    },
    {
      "epoch": 1.92,
      "grad_norm": 13.184967041015625,
      "learning_rate": 1.1293436293436293e-06,
      "loss": 0.3834,
      "step": 8973
    },
    {
      "epoch": 1.92,
      "grad_norm": 10.290412902832031,
      "learning_rate": 1.1261261261261262e-06,
      "loss": 0.431,
      "step": 8974
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.226539611816406,
      "learning_rate": 1.1229086229086228e-06,
      "loss": 0.4649,
      "step": 8975
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.909472465515137,
      "learning_rate": 1.1196911196911197e-06,
      "loss": 0.5048,
      "step": 8976
    },
    {
      "epoch": 1.93,
      "grad_norm": 7.479516983032227,
      "learning_rate": 1.1164736164736164e-06,
      "loss": 0.3364,
      "step": 8977
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.762482643127441,
      "learning_rate": 1.1132561132561133e-06,
      "loss": 0.6574,
      "step": 8978
    },
    {
      "epoch": 1.93,
      "grad_norm": 15.220121383666992,
      "learning_rate": 1.11003861003861e-06,
      "loss": 0.4557,
      "step": 8979
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.815576553344727,
      "learning_rate": 1.1068211068211068e-06,
      "loss": 0.5492,
      "step": 8980
    },
    {
      "epoch": 1.93,
      "grad_norm": 8.710431098937988,
      "learning_rate": 1.1036036036036035e-06,
      "loss": 0.3734,
      "step": 8981
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.552896499633789,
      "learning_rate": 1.1003861003861005e-06,
      "loss": 0.6419,
      "step": 8982
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.123735427856445,
      "learning_rate": 1.0971685971685972e-06,
      "loss": 0.4987,
      "step": 8983
    },
    {
      "epoch": 1.93,
      "grad_norm": 12.838871002197266,
      "learning_rate": 1.093951093951094e-06,
      "loss": 0.5194,
      "step": 8984
    },
    {
      "epoch": 1.93,
      "grad_norm": 7.249022006988525,
      "learning_rate": 1.0907335907335908e-06,
      "loss": 0.3879,
      "step": 8985
    },
    {
      "epoch": 1.93,
      "grad_norm": 12.430036544799805,
      "learning_rate": 1.0875160875160876e-06,
      "loss": 0.5534,
      "step": 8986
    },
    {
      "epoch": 1.93,
      "grad_norm": 7.386796951293945,
      "learning_rate": 1.0842985842985843e-06,
      "loss": 0.4185,
      "step": 8987
    },
    {
      "epoch": 1.93,
      "grad_norm": 13.520341873168945,
      "learning_rate": 1.0810810810810812e-06,
      "loss": 0.69,
      "step": 8988
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.09800910949707,
      "learning_rate": 1.0778635778635778e-06,
      "loss": 0.4259,
      "step": 8989
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.552093505859375,
      "learning_rate": 1.0746460746460747e-06,
      "loss": 0.4992,
      "step": 8990
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.101153373718262,
      "learning_rate": 1.0714285714285714e-06,
      "loss": 0.353,
      "step": 8991
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.454930305480957,
      "learning_rate": 1.0682110682110682e-06,
      "loss": 0.5085,
      "step": 8992
    },
    {
      "epoch": 1.93,
      "grad_norm": 7.653312683105469,
      "learning_rate": 1.064993564993565e-06,
      "loss": 0.3256,
      "step": 8993
    },
    {
      "epoch": 1.93,
      "grad_norm": 12.767512321472168,
      "learning_rate": 1.061776061776062e-06,
      "loss": 0.5334,
      "step": 8994
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.075899124145508,
      "learning_rate": 1.0585585585585587e-06,
      "loss": 0.4891,
      "step": 8995
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.071955680847168,
      "learning_rate": 1.0553410553410555e-06,
      "loss": 0.3932,
      "step": 8996
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.378616333007812,
      "learning_rate": 1.0521235521235522e-06,
      "loss": 0.4785,
      "step": 8997
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.580912590026855,
      "learning_rate": 1.048906048906049e-06,
      "loss": 0.5206,
      "step": 8998
    },
    {
      "epoch": 1.93,
      "grad_norm": 8.403154373168945,
      "learning_rate": 1.0456885456885457e-06,
      "loss": 0.444,
      "step": 8999
    },
    {
      "epoch": 1.93,
      "grad_norm": 8.297141075134277,
      "learning_rate": 1.0424710424710424e-06,
      "loss": 0.3243,
      "step": 9000
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.128799438476562,
      "learning_rate": 1.0392535392535393e-06,
      "loss": 0.4334,
      "step": 9001
    },
    {
      "epoch": 1.93,
      "grad_norm": 13.127787590026855,
      "learning_rate": 1.036036036036036e-06,
      "loss": 0.7265,
      "step": 9002
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.55935001373291,
      "learning_rate": 1.0328185328185328e-06,
      "loss": 0.6041,
      "step": 9003
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.90408706665039,
      "learning_rate": 1.0296010296010295e-06,
      "loss": 0.4896,
      "step": 9004
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.134827613830566,
      "learning_rate": 1.0263835263835263e-06,
      "loss": 0.5467,
      "step": 9005
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.637383460998535,
      "learning_rate": 1.023166023166023e-06,
      "loss": 0.3343,
      "step": 9006
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.41937255859375,
      "learning_rate": 1.01994851994852e-06,
      "loss": 0.4917,
      "step": 9007
    },
    {
      "epoch": 1.93,
      "grad_norm": 8.74335765838623,
      "learning_rate": 1.0167310167310168e-06,
      "loss": 0.5193,
      "step": 9008
    },
    {
      "epoch": 1.93,
      "grad_norm": 12.257126808166504,
      "learning_rate": 1.0135135135135136e-06,
      "loss": 0.4735,
      "step": 9009
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.956939697265625,
      "learning_rate": 1.0102960102960103e-06,
      "loss": 0.3737,
      "step": 9010
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.201946258544922,
      "learning_rate": 1.0070785070785072e-06,
      "loss": 0.4127,
      "step": 9011
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.572041511535645,
      "learning_rate": 1.0038610038610038e-06,
      "loss": 0.3979,
      "step": 9012
    },
    {
      "epoch": 1.93,
      "grad_norm": 9.096888542175293,
      "learning_rate": 1.0006435006435007e-06,
      "loss": 0.6254,
      "step": 9013
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.864278793334961,
      "learning_rate": 9.974259974259974e-07,
      "loss": 0.4046,
      "step": 9014
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.178313255310059,
      "learning_rate": 9.942084942084942e-07,
      "loss": 0.5322,
      "step": 9015
    },
    {
      "epoch": 1.93,
      "grad_norm": 11.307493209838867,
      "learning_rate": 9.90990990990991e-07,
      "loss": 0.5164,
      "step": 9016
    },
    {
      "epoch": 1.93,
      "grad_norm": 7.593134880065918,
      "learning_rate": 9.877734877734878e-07,
      "loss": 0.3727,
      "step": 9017
    },
    {
      "epoch": 1.93,
      "grad_norm": 13.85462474822998,
      "learning_rate": 9.845559845559844e-07,
      "loss": 0.4331,
      "step": 9018
    },
    {
      "epoch": 1.93,
      "grad_norm": 13.405901908874512,
      "learning_rate": 9.813384813384815e-07,
      "loss": 0.6083,
      "step": 9019
    },
    {
      "epoch": 1.93,
      "grad_norm": 13.498706817626953,
      "learning_rate": 9.781209781209782e-07,
      "loss": 0.6134,
      "step": 9020
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.911770820617676,
      "learning_rate": 9.74903474903475e-07,
      "loss": 0.4518,
      "step": 9021
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.78171443939209,
      "learning_rate": 9.716859716859717e-07,
      "loss": 0.4684,
      "step": 9022
    },
    {
      "epoch": 1.94,
      "grad_norm": 12.510881423950195,
      "learning_rate": 9.684684684684686e-07,
      "loss": 0.5779,
      "step": 9023
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.631181716918945,
      "learning_rate": 9.652509652509653e-07,
      "loss": 0.3795,
      "step": 9024
    },
    {
      "epoch": 1.94,
      "grad_norm": 12.832488059997559,
      "learning_rate": 9.620334620334621e-07,
      "loss": 0.5968,
      "step": 9025
    },
    {
      "epoch": 1.94,
      "grad_norm": 12.402261734008789,
      "learning_rate": 9.588159588159588e-07,
      "loss": 0.3979,
      "step": 9026
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.092130661010742,
      "learning_rate": 9.555984555984557e-07,
      "loss": 0.4722,
      "step": 9027
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.557873725891113,
      "learning_rate": 9.523809523809523e-07,
      "loss": 0.4724,
      "step": 9028
    },
    {
      "epoch": 1.94,
      "grad_norm": 12.109800338745117,
      "learning_rate": 9.491634491634491e-07,
      "loss": 0.5402,
      "step": 9029
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.341702461242676,
      "learning_rate": 9.459459459459459e-07,
      "loss": 0.4524,
      "step": 9030
    },
    {
      "epoch": 1.94,
      "grad_norm": 8.765115737915039,
      "learning_rate": 9.427284427284427e-07,
      "loss": 0.3875,
      "step": 9031
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.371682167053223,
      "learning_rate": 9.395109395109396e-07,
      "loss": 0.4461,
      "step": 9032
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.644906044006348,
      "learning_rate": 9.362934362934363e-07,
      "loss": 0.5326,
      "step": 9033
    },
    {
      "epoch": 1.94,
      "grad_norm": 8.30403995513916,
      "learning_rate": 9.330759330759331e-07,
      "loss": 0.4591,
      "step": 9034
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.832475662231445,
      "learning_rate": 9.298584298584299e-07,
      "loss": 0.3651,
      "step": 9035
    },
    {
      "epoch": 1.94,
      "grad_norm": 12.559409141540527,
      "learning_rate": 9.266409266409267e-07,
      "loss": 0.4361,
      "step": 9036
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.654409408569336,
      "learning_rate": 9.234234234234235e-07,
      "loss": 0.4902,
      "step": 9037
    },
    {
      "epoch": 1.94,
      "grad_norm": 12.208739280700684,
      "learning_rate": 9.202059202059202e-07,
      "loss": 0.4917,
      "step": 9038
    },
    {
      "epoch": 1.94,
      "grad_norm": 13.442873001098633,
      "learning_rate": 9.16988416988417e-07,
      "loss": 0.5827,
      "step": 9039
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.265085220336914,
      "learning_rate": 9.137709137709138e-07,
      "loss": 0.6806,
      "step": 9040
    },
    {
      "epoch": 1.94,
      "grad_norm": 20.528789520263672,
      "learning_rate": 9.105534105534107e-07,
      "loss": 0.54,
      "step": 9041
    },
    {
      "epoch": 1.94,
      "grad_norm": 8.567280769348145,
      "learning_rate": 9.073359073359074e-07,
      "loss": 0.3967,
      "step": 9042
    },
    {
      "epoch": 1.94,
      "grad_norm": 6.959560871124268,
      "learning_rate": 9.041184041184042e-07,
      "loss": 0.2519,
      "step": 9043
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.382619857788086,
      "learning_rate": 9.009009009009009e-07,
      "loss": 0.6425,
      "step": 9044
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.695967674255371,
      "learning_rate": 8.976833976833976e-07,
      "loss": 0.4222,
      "step": 9045
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.051957130432129,
      "learning_rate": 8.944658944658944e-07,
      "loss": 0.3706,
      "step": 9046
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.582085609436035,
      "learning_rate": 8.912483912483913e-07,
      "loss": 0.5488,
      "step": 9047
    },
    {
      "epoch": 1.94,
      "grad_norm": 8.995124816894531,
      "learning_rate": 8.88030888030888e-07,
      "loss": 0.524,
      "step": 9048
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.147608757019043,
      "learning_rate": 8.848133848133848e-07,
      "loss": 0.3825,
      "step": 9049
    },
    {
      "epoch": 1.94,
      "grad_norm": 12.057157516479492,
      "learning_rate": 8.815958815958816e-07,
      "loss": 0.4017,
      "step": 9050
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.750699043273926,
      "learning_rate": 8.783783783783784e-07,
      "loss": 0.4327,
      "step": 9051
    },
    {
      "epoch": 1.94,
      "grad_norm": 8.941617012023926,
      "learning_rate": 8.751608751608751e-07,
      "loss": 0.5566,
      "step": 9052
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.53017807006836,
      "learning_rate": 8.71943371943372e-07,
      "loss": 0.3683,
      "step": 9053
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.105064392089844,
      "learning_rate": 8.687258687258688e-07,
      "loss": 0.6562,
      "step": 9054
    },
    {
      "epoch": 1.94,
      "grad_norm": 16.423673629760742,
      "learning_rate": 8.655083655083655e-07,
      "loss": 0.419,
      "step": 9055
    },
    {
      "epoch": 1.94,
      "grad_norm": 13.871034622192383,
      "learning_rate": 8.622908622908623e-07,
      "loss": 0.447,
      "step": 9056
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.797966003417969,
      "learning_rate": 8.590733590733591e-07,
      "loss": 0.6798,
      "step": 9057
    },
    {
      "epoch": 1.94,
      "grad_norm": 13.221912384033203,
      "learning_rate": 8.558558558558558e-07,
      "loss": 0.5442,
      "step": 9058
    },
    {
      "epoch": 1.94,
      "grad_norm": 16.449817657470703,
      "learning_rate": 8.526383526383526e-07,
      "loss": 0.7096,
      "step": 9059
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.324084281921387,
      "learning_rate": 8.494208494208495e-07,
      "loss": 0.4251,
      "step": 9060
    },
    {
      "epoch": 1.94,
      "grad_norm": 13.407254219055176,
      "learning_rate": 8.462033462033463e-07,
      "loss": 0.5571,
      "step": 9061
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.455618858337402,
      "learning_rate": 8.42985842985843e-07,
      "loss": 0.5234,
      "step": 9062
    },
    {
      "epoch": 1.94,
      "grad_norm": 13.884166717529297,
      "learning_rate": 8.397683397683398e-07,
      "loss": 0.6093,
      "step": 9063
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.535658836364746,
      "learning_rate": 8.365508365508366e-07,
      "loss": 0.7309,
      "step": 9064
    },
    {
      "epoch": 1.94,
      "grad_norm": 11.46945858001709,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.5848,
      "step": 9065
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.83887004852295,
      "learning_rate": 8.301158301158302e-07,
      "loss": 0.488,
      "step": 9066
    },
    {
      "epoch": 1.94,
      "grad_norm": 10.250383377075195,
      "learning_rate": 8.26898326898327e-07,
      "loss": 0.4176,
      "step": 9067
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.67304515838623,
      "learning_rate": 8.236808236808237e-07,
      "loss": 0.5611,
      "step": 9068
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.098649978637695,
      "learning_rate": 8.204633204633205e-07,
      "loss": 0.5832,
      "step": 9069
    },
    {
      "epoch": 1.95,
      "grad_norm": 8.605087280273438,
      "learning_rate": 8.172458172458173e-07,
      "loss": 0.5023,
      "step": 9070
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.253371238708496,
      "learning_rate": 8.14028314028314e-07,
      "loss": 0.504,
      "step": 9071
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.965600967407227,
      "learning_rate": 8.108108108108109e-07,
      "loss": 0.4354,
      "step": 9072
    },
    {
      "epoch": 1.95,
      "grad_norm": 12.565620422363281,
      "learning_rate": 8.075933075933077e-07,
      "loss": 0.5386,
      "step": 9073
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.608099937438965,
      "learning_rate": 8.043758043758044e-07,
      "loss": 0.5392,
      "step": 9074
    },
    {
      "epoch": 1.95,
      "grad_norm": 7.0229034423828125,
      "learning_rate": 8.011583011583011e-07,
      "loss": 0.295,
      "step": 9075
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.902487754821777,
      "learning_rate": 7.979407979407979e-07,
      "loss": 0.5127,
      "step": 9076
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.024736404418945,
      "learning_rate": 7.947232947232947e-07,
      "loss": 0.5005,
      "step": 9077
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.533149719238281,
      "learning_rate": 7.915057915057915e-07,
      "loss": 0.5421,
      "step": 9078
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.745038986206055,
      "learning_rate": 7.882882882882883e-07,
      "loss": 0.478,
      "step": 9079
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.171170234680176,
      "learning_rate": 7.850707850707851e-07,
      "loss": 0.4119,
      "step": 9080
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.218092918395996,
      "learning_rate": 7.818532818532818e-07,
      "loss": 0.4708,
      "step": 9081
    },
    {
      "epoch": 1.95,
      "grad_norm": 7.3062825202941895,
      "learning_rate": 7.786357786357786e-07,
      "loss": 0.417,
      "step": 9082
    },
    {
      "epoch": 1.95,
      "grad_norm": 12.020988464355469,
      "learning_rate": 7.754182754182754e-07,
      "loss": 0.5342,
      "step": 9083
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.923880577087402,
      "learning_rate": 7.722007722007722e-07,
      "loss": 0.6486,
      "step": 9084
    },
    {
      "epoch": 1.95,
      "grad_norm": 14.296586990356445,
      "learning_rate": 7.68983268983269e-07,
      "loss": 0.6247,
      "step": 9085
    },
    {
      "epoch": 1.95,
      "grad_norm": 14.192597389221191,
      "learning_rate": 7.657657657657658e-07,
      "loss": 0.4859,
      "step": 9086
    },
    {
      "epoch": 1.95,
      "grad_norm": 8.645506858825684,
      "learning_rate": 7.625482625482626e-07,
      "loss": 0.4657,
      "step": 9087
    },
    {
      "epoch": 1.95,
      "grad_norm": 8.491231918334961,
      "learning_rate": 7.593307593307593e-07,
      "loss": 0.3797,
      "step": 9088
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.497044563293457,
      "learning_rate": 7.561132561132561e-07,
      "loss": 0.543,
      "step": 9089
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.105480194091797,
      "learning_rate": 7.528957528957529e-07,
      "loss": 0.6612,
      "step": 9090
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.288554191589355,
      "learning_rate": 7.496782496782497e-07,
      "loss": 0.5008,
      "step": 9091
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.33183765411377,
      "learning_rate": 7.464607464607465e-07,
      "loss": 0.6875,
      "step": 9092
    },
    {
      "epoch": 1.95,
      "grad_norm": 12.594658851623535,
      "learning_rate": 7.432432432432433e-07,
      "loss": 0.5596,
      "step": 9093
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.368429183959961,
      "learning_rate": 7.400257400257401e-07,
      "loss": 0.4797,
      "step": 9094
    },
    {
      "epoch": 1.95,
      "grad_norm": 7.865571975708008,
      "learning_rate": 7.368082368082368e-07,
      "loss": 0.4024,
      "step": 9095
    },
    {
      "epoch": 1.95,
      "grad_norm": 12.157693862915039,
      "learning_rate": 7.335907335907336e-07,
      "loss": 0.4691,
      "step": 9096
    },
    {
      "epoch": 1.95,
      "grad_norm": 8.490434646606445,
      "learning_rate": 7.303732303732305e-07,
      "loss": 0.5031,
      "step": 9097
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.117355346679688,
      "learning_rate": 7.271557271557272e-07,
      "loss": 0.4987,
      "step": 9098
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.203729629516602,
      "learning_rate": 7.23938223938224e-07,
      "loss": 0.4318,
      "step": 9099
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.422192573547363,
      "learning_rate": 7.207207207207208e-07,
      "loss": 0.5556,
      "step": 9100
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.445335388183594,
      "learning_rate": 7.175032175032175e-07,
      "loss": 0.3658,
      "step": 9101
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.676146507263184,
      "learning_rate": 7.142857142857143e-07,
      "loss": 0.5456,
      "step": 9102
    },
    {
      "epoch": 1.95,
      "grad_norm": 8.503374099731445,
      "learning_rate": 7.110682110682111e-07,
      "loss": 0.3115,
      "step": 9103
    },
    {
      "epoch": 1.95,
      "grad_norm": 7.362860679626465,
      "learning_rate": 7.078507078507079e-07,
      "loss": 0.3373,
      "step": 9104
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.872715950012207,
      "learning_rate": 7.046332046332046e-07,
      "loss": 0.4759,
      "step": 9105
    },
    {
      "epoch": 1.95,
      "grad_norm": 14.429544448852539,
      "learning_rate": 7.014157014157014e-07,
      "loss": 0.5518,
      "step": 9106
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.56467056274414,
      "learning_rate": 6.981981981981982e-07,
      "loss": 0.4134,
      "step": 9107
    },
    {
      "epoch": 1.95,
      "grad_norm": 15.105201721191406,
      "learning_rate": 6.949806949806949e-07,
      "loss": 0.5994,
      "step": 9108
    },
    {
      "epoch": 1.95,
      "grad_norm": 13.471410751342773,
      "learning_rate": 6.917631917631918e-07,
      "loss": 0.5375,
      "step": 9109
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.9700927734375,
      "learning_rate": 6.885456885456886e-07,
      "loss": 0.5005,
      "step": 9110
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.918020248413086,
      "learning_rate": 6.853281853281853e-07,
      "loss": 0.4783,
      "step": 9111
    },
    {
      "epoch": 1.95,
      "grad_norm": 10.917070388793945,
      "learning_rate": 6.821106821106821e-07,
      "loss": 0.5385,
      "step": 9112
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.00948429107666,
      "learning_rate": 6.788931788931789e-07,
      "loss": 0.3341,
      "step": 9113
    },
    {
      "epoch": 1.95,
      "grad_norm": 8.48723316192627,
      "learning_rate": 6.756756756756756e-07,
      "loss": 0.3676,
      "step": 9114
    },
    {
      "epoch": 1.96,
      "grad_norm": 16.706586837768555,
      "learning_rate": 6.724581724581724e-07,
      "loss": 0.6815,
      "step": 9115
    },
    {
      "epoch": 1.96,
      "grad_norm": 8.439419746398926,
      "learning_rate": 6.692406692406693e-07,
      "loss": 0.4144,
      "step": 9116
    },
    {
      "epoch": 1.96,
      "grad_norm": 8.472970962524414,
      "learning_rate": 6.660231660231661e-07,
      "loss": 0.4674,
      "step": 9117
    },
    {
      "epoch": 1.96,
      "grad_norm": 9.193650245666504,
      "learning_rate": 6.628056628056628e-07,
      "loss": 0.4586,
      "step": 9118
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.770673751831055,
      "learning_rate": 6.595881595881596e-07,
      "loss": 0.634,
      "step": 9119
    },
    {
      "epoch": 1.96,
      "grad_norm": 17.283369064331055,
      "learning_rate": 6.563706563706564e-07,
      "loss": 0.6729,
      "step": 9120
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.421284675598145,
      "learning_rate": 6.531531531531531e-07,
      "loss": 0.5401,
      "step": 9121
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.191167831420898,
      "learning_rate": 6.4993564993565e-07,
      "loss": 0.4912,
      "step": 9122
    },
    {
      "epoch": 1.96,
      "grad_norm": 9.753426551818848,
      "learning_rate": 6.467181467181468e-07,
      "loss": 0.5063,
      "step": 9123
    },
    {
      "epoch": 1.96,
      "grad_norm": 12.001924514770508,
      "learning_rate": 6.435006435006435e-07,
      "loss": 0.7666,
      "step": 9124
    },
    {
      "epoch": 1.96,
      "grad_norm": 13.610767364501953,
      "learning_rate": 6.402831402831403e-07,
      "loss": 0.5173,
      "step": 9125
    },
    {
      "epoch": 1.96,
      "grad_norm": 14.7679443359375,
      "learning_rate": 6.370656370656371e-07,
      "loss": 0.5392,
      "step": 9126
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.42783260345459,
      "learning_rate": 6.338481338481339e-07,
      "loss": 0.4894,
      "step": 9127
    },
    {
      "epoch": 1.96,
      "grad_norm": 13.002822875976562,
      "learning_rate": 6.306306306306307e-07,
      "loss": 0.582,
      "step": 9128
    },
    {
      "epoch": 1.96,
      "grad_norm": 9.86281967163086,
      "learning_rate": 6.274131274131275e-07,
      "loss": 0.395,
      "step": 9129
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.67943000793457,
      "learning_rate": 6.241956241956243e-07,
      "loss": 0.5667,
      "step": 9130
    },
    {
      "epoch": 1.96,
      "grad_norm": 12.030247688293457,
      "learning_rate": 6.20978120978121e-07,
      "loss": 0.4432,
      "step": 9131
    },
    {
      "epoch": 1.96,
      "grad_norm": 8.689725875854492,
      "learning_rate": 6.177606177606178e-07,
      "loss": 0.4324,
      "step": 9132
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.52776050567627,
      "learning_rate": 6.145431145431145e-07,
      "loss": 0.4402,
      "step": 9133
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.938210487365723,
      "learning_rate": 6.113256113256113e-07,
      "loss": 0.4854,
      "step": 9134
    },
    {
      "epoch": 1.96,
      "grad_norm": 9.090615272521973,
      "learning_rate": 6.081081081081081e-07,
      "loss": 0.545,
      "step": 9135
    },
    {
      "epoch": 1.96,
      "grad_norm": 8.943923950195312,
      "learning_rate": 6.048906048906049e-07,
      "loss": 0.4589,
      "step": 9136
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.391565322875977,
      "learning_rate": 6.016731016731017e-07,
      "loss": 0.4897,
      "step": 9137
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.256881713867188,
      "learning_rate": 5.984555984555984e-07,
      "loss": 0.4247,
      "step": 9138
    },
    {
      "epoch": 1.96,
      "grad_norm": 12.004615783691406,
      "learning_rate": 5.952380952380952e-07,
      "loss": 0.7027,
      "step": 9139
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.001754760742188,
      "learning_rate": 5.92020592020592e-07,
      "loss": 0.4941,
      "step": 9140
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.195340156555176,
      "learning_rate": 5.888030888030888e-07,
      "loss": 0.4118,
      "step": 9141
    },
    {
      "epoch": 1.96,
      "grad_norm": 12.183869361877441,
      "learning_rate": 5.855855855855856e-07,
      "loss": 0.6096,
      "step": 9142
    },
    {
      "epoch": 1.96,
      "grad_norm": 10.847394943237305,
      "learning_rate": 5.823680823680824e-07,
      "loss": 0.5315,
      "step": 9143
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.670429229736328,
      "learning_rate": 5.791505791505791e-07,
      "loss": 0.4726,
      "step": 9144
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.423516273498535,
      "learning_rate": 5.759330759330759e-07,
      "loss": 0.4721,
      "step": 9145
    },
    {
      "epoch": 1.96,
      "grad_norm": 7.399167537689209,
      "learning_rate": 5.727155727155727e-07,
      "loss": 0.3492,
      "step": 9146
    },
    {
      "epoch": 1.96,
      "grad_norm": 8.133597373962402,
      "learning_rate": 5.694980694980696e-07,
      "loss": 0.4113,
      "step": 9147
    },
    {
      "epoch": 1.96,
      "grad_norm": 12.114322662353516,
      "learning_rate": 5.662805662805663e-07,
      "loss": 0.4214,
      "step": 9148
    },
    {
      "epoch": 1.96,
      "grad_norm": 12.856621742248535,
      "learning_rate": 5.630630630630631e-07,
      "loss": 0.6078,
      "step": 9149
    },
    {
      "epoch": 1.96,
      "grad_norm": 12.897561073303223,
      "learning_rate": 5.598455598455599e-07,
      "loss": 0.6854,
      "step": 9150
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.322882652282715,
      "learning_rate": 5.566280566280566e-07,
      "loss": 0.6565,
      "step": 9151
    },
    {
      "epoch": 1.96,
      "grad_norm": 9.64810848236084,
      "learning_rate": 5.534105534105534e-07,
      "loss": 0.6007,
      "step": 9152
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.276318550109863,
      "learning_rate": 5.501930501930503e-07,
      "loss": 0.3321,
      "step": 9153
    },
    {
      "epoch": 1.96,
      "grad_norm": 9.07102108001709,
      "learning_rate": 5.46975546975547e-07,
      "loss": 0.4672,
      "step": 9154
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.04989242553711,
      "learning_rate": 5.437580437580438e-07,
      "loss": 0.5967,
      "step": 9155
    },
    {
      "epoch": 1.96,
      "grad_norm": 17.94063949584961,
      "learning_rate": 5.405405405405406e-07,
      "loss": 0.5591,
      "step": 9156
    },
    {
      "epoch": 1.96,
      "grad_norm": 9.476765632629395,
      "learning_rate": 5.373230373230374e-07,
      "loss": 0.5749,
      "step": 9157
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.680791854858398,
      "learning_rate": 5.341055341055341e-07,
      "loss": 0.5144,
      "step": 9158
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.656250953674316,
      "learning_rate": 5.30888030888031e-07,
      "loss": 0.4038,
      "step": 9159
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.58141040802002,
      "learning_rate": 5.276705276705278e-07,
      "loss": 0.3693,
      "step": 9160
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.845081329345703,
      "learning_rate": 5.244530244530245e-07,
      "loss": 0.3968,
      "step": 9161
    },
    {
      "epoch": 1.97,
      "grad_norm": 12.163293838500977,
      "learning_rate": 5.212355212355212e-07,
      "loss": 0.6546,
      "step": 9162
    },
    {
      "epoch": 1.97,
      "grad_norm": 12.129267692565918,
      "learning_rate": 5.18018018018018e-07,
      "loss": 0.6052,
      "step": 9163
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.013593673706055,
      "learning_rate": 5.148005148005147e-07,
      "loss": 0.4346,
      "step": 9164
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.189879417419434,
      "learning_rate": 5.115830115830115e-07,
      "loss": 0.5265,
      "step": 9165
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.22214412689209,
      "learning_rate": 5.083655083655084e-07,
      "loss": 0.4386,
      "step": 9166
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.911391258239746,
      "learning_rate": 5.051480051480051e-07,
      "loss": 0.4741,
      "step": 9167
    },
    {
      "epoch": 1.97,
      "grad_norm": 8.322164535522461,
      "learning_rate": 5.019305019305019e-07,
      "loss": 0.3594,
      "step": 9168
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.929170608520508,
      "learning_rate": 4.987129987129987e-07,
      "loss": 0.5103,
      "step": 9169
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.314948081970215,
      "learning_rate": 4.954954954954955e-07,
      "loss": 0.4613,
      "step": 9170
    },
    {
      "epoch": 1.97,
      "grad_norm": 8.087967872619629,
      "learning_rate": 4.922779922779922e-07,
      "loss": 0.5584,
      "step": 9171
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.479588508605957,
      "learning_rate": 4.890604890604891e-07,
      "loss": 0.566,
      "step": 9172
    },
    {
      "epoch": 1.97,
      "grad_norm": 8.242652893066406,
      "learning_rate": 4.858429858429859e-07,
      "loss": 0.3186,
      "step": 9173
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.662955284118652,
      "learning_rate": 4.826254826254826e-07,
      "loss": 0.3708,
      "step": 9174
    },
    {
      "epoch": 1.97,
      "grad_norm": 12.860875129699707,
      "learning_rate": 4.794079794079794e-07,
      "loss": 0.7411,
      "step": 9175
    },
    {
      "epoch": 1.97,
      "grad_norm": 13.417494773864746,
      "learning_rate": 4.761904761904762e-07,
      "loss": 0.6806,
      "step": 9176
    },
    {
      "epoch": 1.97,
      "grad_norm": 8.66100025177002,
      "learning_rate": 4.7297297297297294e-07,
      "loss": 0.4697,
      "step": 9177
    },
    {
      "epoch": 1.97,
      "grad_norm": 8.346163749694824,
      "learning_rate": 4.697554697554698e-07,
      "loss": 0.3501,
      "step": 9178
    },
    {
      "epoch": 1.97,
      "grad_norm": 8.501439094543457,
      "learning_rate": 4.6653796653796653e-07,
      "loss": 0.4024,
      "step": 9179
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.002878189086914,
      "learning_rate": 4.6332046332046336e-07,
      "loss": 0.397,
      "step": 9180
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.019742965698242,
      "learning_rate": 4.601029601029601e-07,
      "loss": 0.4072,
      "step": 9181
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.788640975952148,
      "learning_rate": 4.568854568854569e-07,
      "loss": 0.3861,
      "step": 9182
    },
    {
      "epoch": 1.97,
      "grad_norm": 12.244176864624023,
      "learning_rate": 4.536679536679537e-07,
      "loss": 0.4929,
      "step": 9183
    },
    {
      "epoch": 1.97,
      "grad_norm": 11.675085067749023,
      "learning_rate": 4.5045045045045043e-07,
      "loss": 0.6337,
      "step": 9184
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.622889518737793,
      "learning_rate": 4.472329472329472e-07,
      "loss": 0.4024,
      "step": 9185
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.039494514465332,
      "learning_rate": 4.44015444015444e-07,
      "loss": 0.4718,
      "step": 9186
    },
    {
      "epoch": 1.97,
      "grad_norm": 6.5915608406066895,
      "learning_rate": 4.407979407979408e-07,
      "loss": 0.2679,
      "step": 9187
    },
    {
      "epoch": 1.97,
      "grad_norm": 11.424247741699219,
      "learning_rate": 4.3758043758043756e-07,
      "loss": 0.5916,
      "step": 9188
    },
    {
      "epoch": 1.97,
      "grad_norm": 17.652902603149414,
      "learning_rate": 4.343629343629344e-07,
      "loss": 0.659,
      "step": 9189
    },
    {
      "epoch": 1.97,
      "grad_norm": 8.92506217956543,
      "learning_rate": 4.3114543114543115e-07,
      "loss": 0.4281,
      "step": 9190
    },
    {
      "epoch": 1.97,
      "grad_norm": 11.62391471862793,
      "learning_rate": 4.279279279279279e-07,
      "loss": 0.4662,
      "step": 9191
    },
    {
      "epoch": 1.97,
      "grad_norm": 13.331753730773926,
      "learning_rate": 4.2471042471042474e-07,
      "loss": 0.641,
      "step": 9192
    },
    {
      "epoch": 1.97,
      "grad_norm": 11.258801460266113,
      "learning_rate": 4.214929214929215e-07,
      "loss": 0.4722,
      "step": 9193
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.07225513458252,
      "learning_rate": 4.182754182754183e-07,
      "loss": 0.4002,
      "step": 9194
    },
    {
      "epoch": 1.97,
      "grad_norm": 12.035822868347168,
      "learning_rate": 4.150579150579151e-07,
      "loss": 0.6343,
      "step": 9195
    },
    {
      "epoch": 1.97,
      "grad_norm": 12.411063194274902,
      "learning_rate": 4.1184041184041187e-07,
      "loss": 0.5179,
      "step": 9196
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.476591110229492,
      "learning_rate": 4.0862290862290864e-07,
      "loss": 0.5437,
      "step": 9197
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.091841697692871,
      "learning_rate": 4.0540540540540546e-07,
      "loss": 0.4535,
      "step": 9198
    },
    {
      "epoch": 1.97,
      "grad_norm": 11.386109352111816,
      "learning_rate": 4.021879021879022e-07,
      "loss": 0.5703,
      "step": 9199
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.878549575805664,
      "learning_rate": 3.9897039897039895e-07,
      "loss": 0.4176,
      "step": 9200
    },
    {
      "epoch": 1.97,
      "grad_norm": 6.558528900146484,
      "learning_rate": 3.9575289575289577e-07,
      "loss": 0.3308,
      "step": 9201
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.05831241607666,
      "learning_rate": 3.9253539253539254e-07,
      "loss": 0.6256,
      "step": 9202
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.619301795959473,
      "learning_rate": 3.893178893178893e-07,
      "loss": 0.4075,
      "step": 9203
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.853023529052734,
      "learning_rate": 3.861003861003861e-07,
      "loss": 0.4165,
      "step": 9204
    },
    {
      "epoch": 1.97,
      "grad_norm": 15.383463859558105,
      "learning_rate": 3.828828828828829e-07,
      "loss": 0.694,
      "step": 9205
    },
    {
      "epoch": 1.97,
      "grad_norm": 10.96263599395752,
      "learning_rate": 3.7966537966537967e-07,
      "loss": 0.3811,
      "step": 9206
    },
    {
      "epoch": 1.97,
      "grad_norm": 9.839561462402344,
      "learning_rate": 3.7644787644787644e-07,
      "loss": 0.5353,
      "step": 9207
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.298849105834961,
      "learning_rate": 3.7323037323037326e-07,
      "loss": 0.4641,
      "step": 9208
    },
    {
      "epoch": 1.98,
      "grad_norm": 11.843244552612305,
      "learning_rate": 3.7001287001287003e-07,
      "loss": 0.5938,
      "step": 9209
    },
    {
      "epoch": 1.98,
      "grad_norm": 11.02099323272705,
      "learning_rate": 3.667953667953668e-07,
      "loss": 0.4582,
      "step": 9210
    },
    {
      "epoch": 1.98,
      "grad_norm": 8.847176551818848,
      "learning_rate": 3.635778635778636e-07,
      "loss": 0.48,
      "step": 9211
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.091920852661133,
      "learning_rate": 3.603603603603604e-07,
      "loss": 0.3816,
      "step": 9212
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.342947006225586,
      "learning_rate": 3.5714285714285716e-07,
      "loss": 0.3984,
      "step": 9213
    },
    {
      "epoch": 1.98,
      "grad_norm": 11.250204086303711,
      "learning_rate": 3.539253539253539e-07,
      "loss": 0.5286,
      "step": 9214
    },
    {
      "epoch": 1.98,
      "grad_norm": 14.391362190246582,
      "learning_rate": 3.507078507078507e-07,
      "loss": 0.6844,
      "step": 9215
    },
    {
      "epoch": 1.98,
      "grad_norm": 12.91857624053955,
      "learning_rate": 3.4749034749034746e-07,
      "loss": 0.6161,
      "step": 9216
    },
    {
      "epoch": 1.98,
      "grad_norm": 9.950221061706543,
      "learning_rate": 3.442728442728443e-07,
      "loss": 0.4611,
      "step": 9217
    },
    {
      "epoch": 1.98,
      "grad_norm": 12.019424438476562,
      "learning_rate": 3.4105534105534105e-07,
      "loss": 0.5263,
      "step": 9218
    },
    {
      "epoch": 1.98,
      "grad_norm": 11.78425407409668,
      "learning_rate": 3.378378378378378e-07,
      "loss": 0.6162,
      "step": 9219
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.160080909729004,
      "learning_rate": 3.3462033462033465e-07,
      "loss": 0.5118,
      "step": 9220
    },
    {
      "epoch": 1.98,
      "grad_norm": 14.636565208435059,
      "learning_rate": 3.314028314028314e-07,
      "loss": 0.6144,
      "step": 9221
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.045703887939453,
      "learning_rate": 3.281853281853282e-07,
      "loss": 0.5168,
      "step": 9222
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.188369750976562,
      "learning_rate": 3.24967824967825e-07,
      "loss": 0.5643,
      "step": 9223
    },
    {
      "epoch": 1.98,
      "grad_norm": 12.58443832397461,
      "learning_rate": 3.217503217503218e-07,
      "loss": 0.5534,
      "step": 9224
    },
    {
      "epoch": 1.98,
      "grad_norm": 11.672258377075195,
      "learning_rate": 3.1853281853281854e-07,
      "loss": 0.7019,
      "step": 9225
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.989978790283203,
      "learning_rate": 3.1531531531531537e-07,
      "loss": 0.3815,
      "step": 9226
    },
    {
      "epoch": 1.98,
      "grad_norm": 14.061473846435547,
      "learning_rate": 3.1209781209781213e-07,
      "loss": 0.6267,
      "step": 9227
    },
    {
      "epoch": 1.98,
      "grad_norm": 7.363963603973389,
      "learning_rate": 3.088803088803089e-07,
      "loss": 0.2789,
      "step": 9228
    },
    {
      "epoch": 1.98,
      "grad_norm": 7.864991188049316,
      "learning_rate": 3.0566280566280567e-07,
      "loss": 0.3807,
      "step": 9229
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.413660049438477,
      "learning_rate": 3.0244530244530244e-07,
      "loss": 0.4565,
      "step": 9230
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.55013656616211,
      "learning_rate": 2.992277992277992e-07,
      "loss": 0.3704,
      "step": 9231
    },
    {
      "epoch": 1.98,
      "grad_norm": 9.826130867004395,
      "learning_rate": 2.96010296010296e-07,
      "loss": 0.4793,
      "step": 9232
    },
    {
      "epoch": 1.98,
      "grad_norm": 8.458977699279785,
      "learning_rate": 2.927927927927928e-07,
      "loss": 0.3451,
      "step": 9233
    },
    {
      "epoch": 1.98,
      "grad_norm": 9.112563133239746,
      "learning_rate": 2.8957528957528957e-07,
      "loss": 0.4286,
      "step": 9234
    },
    {
      "epoch": 1.98,
      "grad_norm": 11.426466941833496,
      "learning_rate": 2.8635778635778634e-07,
      "loss": 0.5013,
      "step": 9235
    },
    {
      "epoch": 1.98,
      "grad_norm": 9.945372581481934,
      "learning_rate": 2.8314028314028316e-07,
      "loss": 0.457,
      "step": 9236
    },
    {
      "epoch": 1.98,
      "grad_norm": 12.284810066223145,
      "learning_rate": 2.7992277992277993e-07,
      "loss": 0.4359,
      "step": 9237
    },
    {
      "epoch": 1.98,
      "grad_norm": 9.691364288330078,
      "learning_rate": 2.767052767052767e-07,
      "loss": 0.4259,
      "step": 9238
    },
    {
      "epoch": 1.98,
      "grad_norm": 7.742102146148682,
      "learning_rate": 2.734877734877735e-07,
      "loss": 0.3588,
      "step": 9239
    },
    {
      "epoch": 1.98,
      "grad_norm": 12.371809959411621,
      "learning_rate": 2.702702702702703e-07,
      "loss": 0.5449,
      "step": 9240
    },
    {
      "epoch": 1.98,
      "grad_norm": 12.343169212341309,
      "learning_rate": 2.6705276705276706e-07,
      "loss": 0.6796,
      "step": 9241
    },
    {
      "epoch": 1.98,
      "grad_norm": 13.034626960754395,
      "learning_rate": 2.638352638352639e-07,
      "loss": 0.5172,
      "step": 9242
    },
    {
      "epoch": 1.98,
      "grad_norm": 7.263753414154053,
      "learning_rate": 2.606177606177606e-07,
      "loss": 0.3083,
      "step": 9243
    },
    {
      "epoch": 1.98,
      "grad_norm": 12.677129745483398,
      "learning_rate": 2.5740025740025737e-07,
      "loss": 0.6035,
      "step": 9244
    },
    {
      "epoch": 1.98,
      "grad_norm": 10.29814338684082,
      "learning_rate": 2.541827541827542e-07,
      "loss": 0.4997,
      "step": 9245
    },
    {
      "epoch": 1.98,
      "grad_norm": 14.055529594421387,
      "learning_rate": 2.5096525096525096e-07,
      "loss": 0.4643,
      "step": 9246
    },
    {
      "epoch": 1.98,
      "grad_norm": 12.726248741149902,
      "learning_rate": 2.4774774774774773e-07,
      "loss": 0.6099,
      "step": 9247
    },
    {
      "epoch": 1.98,
      "grad_norm": 12.946471214294434,
      "learning_rate": 2.4453024453024455e-07,
      "loss": 0.614,
      "step": 9248
    },
    {
      "epoch": 1.98,
      "grad_norm": 14.038660049438477,
      "learning_rate": 2.413127413127413e-07,
      "loss": 0.489,
      "step": 9249
    },
    {
      "epoch": 1.98,
      "grad_norm": 14.418968200683594,
      "learning_rate": 2.380952380952381e-07,
      "loss": 0.7391,
      "step": 9250
    },
    {
      "epoch": 1.98,
      "grad_norm": 9.874394416809082,
      "learning_rate": 2.348777348777349e-07,
      "loss": 0.509,
      "step": 9251
    },
    {
      "epoch": 1.98,
      "grad_norm": 12.100699424743652,
      "learning_rate": 2.3166023166023168e-07,
      "loss": 0.6674,
      "step": 9252
    },
    {
      "epoch": 1.98,
      "grad_norm": 7.990898132324219,
      "learning_rate": 2.2844272844272845e-07,
      "loss": 0.4475,
      "step": 9253
    },
    {
      "epoch": 1.98,
      "grad_norm": 9.355510711669922,
      "learning_rate": 2.2522522522522522e-07,
      "loss": 0.4112,
      "step": 9254
    },
    {
      "epoch": 1.99,
      "grad_norm": 9.640372276306152,
      "learning_rate": 2.22007722007722e-07,
      "loss": 0.4983,
      "step": 9255
    },
    {
      "epoch": 1.99,
      "grad_norm": 11.446083068847656,
      "learning_rate": 2.1879021879021878e-07,
      "loss": 0.6432,
      "step": 9256
    },
    {
      "epoch": 1.99,
      "grad_norm": 13.404082298278809,
      "learning_rate": 2.1557271557271558e-07,
      "loss": 0.4668,
      "step": 9257
    },
    {
      "epoch": 1.99,
      "grad_norm": 13.174975395202637,
      "learning_rate": 2.1235521235521237e-07,
      "loss": 0.617,
      "step": 9258
    },
    {
      "epoch": 1.99,
      "grad_norm": 11.215835571289062,
      "learning_rate": 2.0913770913770914e-07,
      "loss": 0.4362,
      "step": 9259
    },
    {
      "epoch": 1.99,
      "grad_norm": 9.98351001739502,
      "learning_rate": 2.0592020592020594e-07,
      "loss": 0.3143,
      "step": 9260
    },
    {
      "epoch": 1.99,
      "grad_norm": 11.724245071411133,
      "learning_rate": 2.0270270270270273e-07,
      "loss": 0.5514,
      "step": 9261
    },
    {
      "epoch": 1.99,
      "grad_norm": 11.114069938659668,
      "learning_rate": 1.9948519948519947e-07,
      "loss": 0.5298,
      "step": 9262
    },
    {
      "epoch": 1.99,
      "grad_norm": 11.064886093139648,
      "learning_rate": 1.9626769626769627e-07,
      "loss": 0.3949,
      "step": 9263
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.44289493560791,
      "learning_rate": 1.9305019305019304e-07,
      "loss": 0.4248,
      "step": 9264
    },
    {
      "epoch": 1.99,
      "grad_norm": 6.699348449707031,
      "learning_rate": 1.8983268983268983e-07,
      "loss": 0.4597,
      "step": 9265
    },
    {
      "epoch": 1.99,
      "grad_norm": 13.585929870605469,
      "learning_rate": 1.8661518661518663e-07,
      "loss": 0.6293,
      "step": 9266
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.59924030303955,
      "learning_rate": 1.833976833976834e-07,
      "loss": 0.5046,
      "step": 9267
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.678101539611816,
      "learning_rate": 1.801801801801802e-07,
      "loss": 0.4615,
      "step": 9268
    },
    {
      "epoch": 1.99,
      "grad_norm": 12.11205005645752,
      "learning_rate": 1.7696267696267696e-07,
      "loss": 0.5415,
      "step": 9269
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.921358108520508,
      "learning_rate": 1.7374517374517373e-07,
      "loss": 0.4721,
      "step": 9270
    },
    {
      "epoch": 1.99,
      "grad_norm": 14.040493965148926,
      "learning_rate": 1.7052767052767053e-07,
      "loss": 0.5157,
      "step": 9271
    },
    {
      "epoch": 1.99,
      "grad_norm": 8.816844940185547,
      "learning_rate": 1.6731016731016732e-07,
      "loss": 0.448,
      "step": 9272
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.667736053466797,
      "learning_rate": 1.640926640926641e-07,
      "loss": 0.4695,
      "step": 9273
    },
    {
      "epoch": 1.99,
      "grad_norm": 8.192551612854004,
      "learning_rate": 1.608751608751609e-07,
      "loss": 0.5516,
      "step": 9274
    },
    {
      "epoch": 1.99,
      "grad_norm": 9.55359172821045,
      "learning_rate": 1.5765765765765768e-07,
      "loss": 0.3609,
      "step": 9275
    },
    {
      "epoch": 1.99,
      "grad_norm": 8.937917709350586,
      "learning_rate": 1.5444015444015445e-07,
      "loss": 0.4156,
      "step": 9276
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.44440746307373,
      "learning_rate": 1.5122265122265122e-07,
      "loss": 0.4433,
      "step": 9277
    },
    {
      "epoch": 1.99,
      "grad_norm": 12.848092079162598,
      "learning_rate": 1.48005148005148e-07,
      "loss": 0.571,
      "step": 9278
    },
    {
      "epoch": 1.99,
      "grad_norm": 11.78586483001709,
      "learning_rate": 1.4478764478764479e-07,
      "loss": 0.6611,
      "step": 9279
    },
    {
      "epoch": 1.99,
      "grad_norm": 11.6875581741333,
      "learning_rate": 1.4157014157014158e-07,
      "loss": 0.3526,
      "step": 9280
    },
    {
      "epoch": 1.99,
      "grad_norm": 12.536789894104004,
      "learning_rate": 1.3835263835263835e-07,
      "loss": 0.5138,
      "step": 9281
    },
    {
      "epoch": 1.99,
      "grad_norm": 12.965822219848633,
      "learning_rate": 1.3513513513513515e-07,
      "loss": 0.6298,
      "step": 9282
    },
    {
      "epoch": 1.99,
      "grad_norm": 8.667576789855957,
      "learning_rate": 1.3191763191763194e-07,
      "loss": 0.3974,
      "step": 9283
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.580541610717773,
      "learning_rate": 1.2870012870012868e-07,
      "loss": 0.4123,
      "step": 9284
    },
    {
      "epoch": 1.99,
      "grad_norm": 13.122830390930176,
      "learning_rate": 1.2548262548262548e-07,
      "loss": 0.4993,
      "step": 9285
    },
    {
      "epoch": 1.99,
      "grad_norm": 9.653629302978516,
      "learning_rate": 1.2226512226512227e-07,
      "loss": 0.3358,
      "step": 9286
    },
    {
      "epoch": 1.99,
      "grad_norm": 9.031495094299316,
      "learning_rate": 1.1904761904761904e-07,
      "loss": 0.3908,
      "step": 9287
    },
    {
      "epoch": 1.99,
      "grad_norm": 12.925725936889648,
      "learning_rate": 1.1583011583011584e-07,
      "loss": 0.6566,
      "step": 9288
    },
    {
      "epoch": 1.99,
      "grad_norm": 6.892827987670898,
      "learning_rate": 1.1261261261261261e-07,
      "loss": 0.3521,
      "step": 9289
    },
    {
      "epoch": 1.99,
      "grad_norm": 14.049337387084961,
      "learning_rate": 1.0939510939510939e-07,
      "loss": 0.6423,
      "step": 9290
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.999833106994629,
      "learning_rate": 1.0617760617760619e-07,
      "loss": 0.3765,
      "step": 9291
    },
    {
      "epoch": 1.99,
      "grad_norm": 12.20887279510498,
      "learning_rate": 1.0296010296010297e-07,
      "loss": 0.3583,
      "step": 9292
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.349849700927734,
      "learning_rate": 9.974259974259974e-08,
      "loss": 0.5331,
      "step": 9293
    },
    {
      "epoch": 1.99,
      "grad_norm": 8.567801475524902,
      "learning_rate": 9.652509652509652e-08,
      "loss": 0.3698,
      "step": 9294
    },
    {
      "epoch": 1.99,
      "grad_norm": 9.549954414367676,
      "learning_rate": 9.330759330759331e-08,
      "loss": 0.6353,
      "step": 9295
    },
    {
      "epoch": 1.99,
      "grad_norm": 8.741156578063965,
      "learning_rate": 9.00900900900901e-08,
      "loss": 0.3779,
      "step": 9296
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.536890983581543,
      "learning_rate": 8.687258687258687e-08,
      "loss": 0.4197,
      "step": 9297
    },
    {
      "epoch": 1.99,
      "grad_norm": 8.025442123413086,
      "learning_rate": 8.365508365508366e-08,
      "loss": 0.4409,
      "step": 9298
    },
    {
      "epoch": 1.99,
      "grad_norm": 10.90553092956543,
      "learning_rate": 8.043758043758044e-08,
      "loss": 0.4528,
      "step": 9299
    },
    {
      "epoch": 1.99,
      "grad_norm": 9.572848320007324,
      "learning_rate": 7.722007722007723e-08,
      "loss": 0.4622,
      "step": 9300
    },
    {
      "epoch": 2.0,
      "grad_norm": 15.756446838378906,
      "learning_rate": 7.4002574002574e-08,
      "loss": 0.5364,
      "step": 9301
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.535000801086426,
      "learning_rate": 7.078507078507079e-08,
      "loss": 0.3185,
      "step": 9302
    },
    {
      "epoch": 2.0,
      "grad_norm": 12.143465995788574,
      "learning_rate": 6.756756756756757e-08,
      "loss": 0.4019,
      "step": 9303
    },
    {
      "epoch": 2.0,
      "grad_norm": 15.02393913269043,
      "learning_rate": 6.435006435006434e-08,
      "loss": 0.5813,
      "step": 9304
    },
    {
      "epoch": 2.0,
      "grad_norm": 9.504505157470703,
      "learning_rate": 6.113256113256114e-08,
      "loss": 0.5991,
      "step": 9305
    },
    {
      "epoch": 2.0,
      "grad_norm": 11.713574409484863,
      "learning_rate": 5.791505791505792e-08,
      "loss": 0.4514,
      "step": 9306
    },
    {
      "epoch": 2.0,
      "grad_norm": 11.789973258972168,
      "learning_rate": 5.4697554697554695e-08,
      "loss": 0.3444,
      "step": 9307
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.278773307800293,
      "learning_rate": 5.1480051480051484e-08,
      "loss": 0.4485,
      "step": 9308
    },
    {
      "epoch": 2.0,
      "grad_norm": 14.576332092285156,
      "learning_rate": 4.826254826254826e-08,
      "loss": 0.7948,
      "step": 9309
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.298701286315918,
      "learning_rate": 4.504504504504505e-08,
      "loss": 0.4826,
      "step": 9310
    },
    {
      "epoch": 2.0,
      "grad_norm": 11.51756763458252,
      "learning_rate": 4.182754182754183e-08,
      "loss": 0.3991,
      "step": 9311
    },
    {
      "epoch": 2.0,
      "grad_norm": 11.309240341186523,
      "learning_rate": 3.861003861003861e-08,
      "loss": 0.3799,
      "step": 9312
    },
    {
      "epoch": 2.0,
      "grad_norm": 9.807941436767578,
      "learning_rate": 3.5392535392535395e-08,
      "loss": 0.505,
      "step": 9313
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.732566833496094,
      "learning_rate": 3.217503217503217e-08,
      "loss": 0.4802,
      "step": 9314
    },
    {
      "epoch": 2.0,
      "grad_norm": 9.466045379638672,
      "learning_rate": 2.895752895752896e-08,
      "loss": 0.4702,
      "step": 9315
    },
    {
      "epoch": 2.0,
      "grad_norm": 13.318941116333008,
      "learning_rate": 2.5740025740025742e-08,
      "loss": 0.6006,
      "step": 9316
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.055461883544922,
      "learning_rate": 2.2522522522522524e-08,
      "loss": 0.4127,
      "step": 9317
    },
    {
      "epoch": 2.0,
      "grad_norm": 9.802825927734375,
      "learning_rate": 1.9305019305019306e-08,
      "loss": 0.4575,
      "step": 9318
    },
    {
      "epoch": 2.0,
      "grad_norm": 12.75990104675293,
      "learning_rate": 1.6087516087516085e-08,
      "loss": 0.4575,
      "step": 9319
    },
    {
      "epoch": 2.0,
      "grad_norm": 12.870038032531738,
      "learning_rate": 1.2870012870012871e-08,
      "loss": 0.3221,
      "step": 9320
    },
    {
      "epoch": 2.0,
      "grad_norm": 12.106593132019043,
      "learning_rate": 9.652509652509653e-09,
      "loss": 0.6112,
      "step": 9321
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.305700302124023,
      "learning_rate": 6.4350064350064355e-09,
      "loss": 0.5163,
      "step": 9322
    },
    {
      "epoch": 2.0,
      "grad_norm": 11.209067344665527,
      "learning_rate": 3.2175032175032177e-09,
      "loss": 0.6655,
      "step": 9323
    },
    {
      "epoch": 2.0,
      "grad_norm": 16.937227249145508,
      "learning_rate": 0.0,
      "loss": 0.1924,
      "step": 9324
    },
    {
      "epoch": 2.0,
      "step": 9324,
      "total_flos": 1.7538290571471667e+17,
      "train_loss": 0.8744708896402457,
      "train_runtime": 2872.5945,
      "train_samples_per_second": 311.543,
      "train_steps_per_second": 3.246
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 9324,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 1.7538290571471667e+17,
  "train_batch_size": 96,
  "trial_name": null,
  "trial_params": null
}