{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9983642311886587,
  "eval_steps": 500,
  "global_step": 229,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008724100327153763,
      "grad_norm": 6.035512972360855,
      "learning_rate": 1.4035087719298246e-06,
      "loss": 1.1029,
      "step": 1
    },
    {
      "epoch": 0.017448200654307525,
      "grad_norm": 6.042841036030158,
      "learning_rate": 2.8070175438596493e-06,
      "loss": 1.1024,
      "step": 2
    },
    {
      "epoch": 0.026172300981461286,
      "grad_norm": 5.910306635602092,
      "learning_rate": 4.210526315789474e-06,
      "loss": 1.0986,
      "step": 3
    },
    {
      "epoch": 0.03489640130861505,
      "grad_norm": 4.536951510086868,
      "learning_rate": 5.6140350877192985e-06,
      "loss": 1.0538,
      "step": 4
    },
    {
      "epoch": 0.04362050163576881,
      "grad_norm": 2.3773466328483686,
      "learning_rate": 7.017543859649123e-06,
      "loss": 0.9941,
      "step": 5
    },
    {
      "epoch": 0.05234460196292257,
      "grad_norm": 1.999261332126332,
      "learning_rate": 8.421052631578948e-06,
      "loss": 0.999,
      "step": 6
    },
    {
      "epoch": 0.061068702290076333,
      "grad_norm": 4.369629342726399,
      "learning_rate": 9.824561403508772e-06,
      "loss": 0.9889,
      "step": 7
    },
    {
      "epoch": 0.0697928026172301,
      "grad_norm": 4.899527942424933,
      "learning_rate": 1.1228070175438597e-05,
      "loss": 0.9631,
      "step": 8
    },
    {
      "epoch": 0.07851690294438386,
      "grad_norm": 5.748165018609449,
      "learning_rate": 1.263157894736842e-05,
      "loss": 0.9524,
      "step": 9
    },
    {
      "epoch": 0.08724100327153762,
      "grad_norm": 4.763126150981281,
      "learning_rate": 1.4035087719298246e-05,
      "loss": 0.952,
      "step": 10
    },
    {
      "epoch": 0.09596510359869138,
      "grad_norm": 3.167454561583767,
      "learning_rate": 1.543859649122807e-05,
      "loss": 0.9095,
      "step": 11
    },
    {
      "epoch": 0.10468920392584515,
      "grad_norm": 3.3676041736516433,
      "learning_rate": 1.6842105263157896e-05,
      "loss": 0.89,
      "step": 12
    },
    {
      "epoch": 0.1134133042529989,
      "grad_norm": 2.8110117483164676,
      "learning_rate": 1.824561403508772e-05,
      "loss": 0.8538,
      "step": 13
    },
    {
      "epoch": 0.12213740458015267,
      "grad_norm": 2.0519859086879286,
      "learning_rate": 1.9649122807017544e-05,
      "loss": 0.85,
      "step": 14
    },
    {
      "epoch": 0.13086150490730644,
      "grad_norm": 1.8373157533967603,
      "learning_rate": 2.105263157894737e-05,
      "loss": 0.8349,
      "step": 15
    },
    {
      "epoch": 0.1395856052344602,
      "grad_norm": 1.6156997409278058,
      "learning_rate": 2.2456140350877194e-05,
      "loss": 0.829,
      "step": 16
    },
    {
      "epoch": 0.14830970556161396,
      "grad_norm": 1.2400567097625959,
      "learning_rate": 2.385964912280702e-05,
      "loss": 0.8108,
      "step": 17
    },
    {
      "epoch": 0.15703380588876772,
      "grad_norm": 1.1069155525681142,
      "learning_rate": 2.526315789473684e-05,
      "loss": 0.8029,
      "step": 18
    },
    {
      "epoch": 0.16575790621592149,
      "grad_norm": 1.0769626887248518,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 0.7858,
      "step": 19
    },
    {
      "epoch": 0.17448200654307525,
      "grad_norm": 0.7016586049675165,
      "learning_rate": 2.8070175438596492e-05,
      "loss": 0.7816,
      "step": 20
    },
    {
      "epoch": 0.183206106870229,
      "grad_norm": 0.8460028260574683,
      "learning_rate": 2.9473684210526317e-05,
      "loss": 0.7811,
      "step": 21
    },
    {
      "epoch": 0.19193020719738277,
      "grad_norm": 0.6598083455923888,
      "learning_rate": 3.087719298245614e-05,
      "loss": 0.7643,
      "step": 22
    },
    {
      "epoch": 0.20065430752453653,
      "grad_norm": 0.7215762522174539,
      "learning_rate": 3.228070175438597e-05,
      "loss": 0.7617,
      "step": 23
    },
    {
      "epoch": 0.2093784078516903,
      "grad_norm": 0.5554406103693584,
      "learning_rate": 3.368421052631579e-05,
      "loss": 0.7576,
      "step": 24
    },
    {
      "epoch": 0.21810250817884405,
      "grad_norm": 0.6258543895787715,
      "learning_rate": 3.508771929824562e-05,
      "loss": 0.7629,
      "step": 25
    },
    {
      "epoch": 0.2268266085059978,
      "grad_norm": 0.7676778822568288,
      "learning_rate": 3.649122807017544e-05,
      "loss": 0.7521,
      "step": 26
    },
    {
      "epoch": 0.23555070883315157,
      "grad_norm": 1.0032530689824866,
      "learning_rate": 3.789473684210526e-05,
      "loss": 0.7384,
      "step": 27
    },
    {
      "epoch": 0.24427480916030533,
      "grad_norm": 1.2535316239757839,
      "learning_rate": 3.929824561403509e-05,
      "loss": 0.7474,
      "step": 28
    },
    {
      "epoch": 0.2529989094874591,
      "grad_norm": 0.9953077887554728,
      "learning_rate": 4.070175438596492e-05,
      "loss": 0.7412,
      "step": 29
    },
    {
      "epoch": 0.2617230098146129,
      "grad_norm": 1.3600760898966002,
      "learning_rate": 4.210526315789474e-05,
      "loss": 0.7257,
      "step": 30
    },
    {
      "epoch": 0.27044711014176664,
      "grad_norm": 0.6326945767090465,
      "learning_rate": 4.350877192982457e-05,
      "loss": 0.7247,
      "step": 31
    },
    {
      "epoch": 0.2791712104689204,
      "grad_norm": 1.0488194881380388,
      "learning_rate": 4.491228070175439e-05,
      "loss": 0.7247,
      "step": 32
    },
    {
      "epoch": 0.28789531079607417,
      "grad_norm": 1.1854260719924898,
      "learning_rate": 4.6315789473684214e-05,
      "loss": 0.7265,
      "step": 33
    },
    {
      "epoch": 0.2966194111232279,
      "grad_norm": 1.231800504978088,
      "learning_rate": 4.771929824561404e-05,
      "loss": 0.724,
      "step": 34
    },
    {
      "epoch": 0.3053435114503817,
      "grad_norm": 1.1471910683800164,
      "learning_rate": 4.9122807017543864e-05,
      "loss": 0.7245,
      "step": 35
    },
    {
      "epoch": 0.31406761177753545,
      "grad_norm": 1.2650375204215663,
      "learning_rate": 5.052631578947368e-05,
      "loss": 0.7173,
      "step": 36
    },
    {
      "epoch": 0.3227917121046892,
      "grad_norm": 1.581708517443491,
      "learning_rate": 5.1929824561403515e-05,
      "loss": 0.7146,
      "step": 37
    },
    {
      "epoch": 0.33151581243184297,
      "grad_norm": 0.6899231907687099,
      "learning_rate": 5.333333333333333e-05,
      "loss": 0.7039,
      "step": 38
    },
    {
      "epoch": 0.34023991275899673,
      "grad_norm": 1.9659650918585876,
      "learning_rate": 5.4736842105263165e-05,
      "loss": 0.7129,
      "step": 39
    },
    {
      "epoch": 0.3489640130861505,
      "grad_norm": 0.9602545263207618,
      "learning_rate": 5.6140350877192984e-05,
      "loss": 0.7229,
      "step": 40
    },
    {
      "epoch": 0.35768811341330425,
      "grad_norm": 2.019396688493347,
      "learning_rate": 5.7543859649122816e-05,
      "loss": 0.7173,
      "step": 41
    },
    {
      "epoch": 0.366412213740458,
      "grad_norm": 1.3667216071947406,
      "learning_rate": 5.8947368421052634e-05,
      "loss": 0.7088,
      "step": 42
    },
    {
      "epoch": 0.3751363140676118,
      "grad_norm": 1.6739220492126983,
      "learning_rate": 6.035087719298246e-05,
      "loss": 0.7126,
      "step": 43
    },
    {
      "epoch": 0.38386041439476554,
      "grad_norm": 1.406631076781101,
      "learning_rate": 6.175438596491228e-05,
      "loss": 0.7211,
      "step": 44
    },
    {
      "epoch": 0.3925845147219193,
      "grad_norm": 1.232815581243198,
      "learning_rate": 6.315789473684212e-05,
      "loss": 0.7114,
      "step": 45
    },
    {
      "epoch": 0.40130861504907306,
      "grad_norm": 1.3716611381630202,
      "learning_rate": 6.456140350877194e-05,
      "loss": 0.7174,
      "step": 46
    },
    {
      "epoch": 0.4100327153762268,
      "grad_norm": 1.1415702503019296,
      "learning_rate": 6.596491228070175e-05,
      "loss": 0.7005,
      "step": 47
    },
    {
      "epoch": 0.4187568157033806,
      "grad_norm": 1.2074860027919303,
      "learning_rate": 6.736842105263159e-05,
      "loss": 0.7098,
      "step": 48
    },
    {
      "epoch": 0.42748091603053434,
      "grad_norm": 1.6609425790213797,
      "learning_rate": 6.87719298245614e-05,
      "loss": 0.7098,
      "step": 49
    },
    {
      "epoch": 0.4362050163576881,
      "grad_norm": 0.9769451874458346,
      "learning_rate": 7.017543859649124e-05,
      "loss": 0.7079,
      "step": 50
    },
    {
      "epoch": 0.44492911668484186,
      "grad_norm": 1.7666355593904801,
      "learning_rate": 7.157894736842105e-05,
      "loss": 0.6967,
      "step": 51
    },
    {
      "epoch": 0.4536532170119956,
      "grad_norm": 1.014343915032852,
      "learning_rate": 7.298245614035087e-05,
      "loss": 0.6961,
      "step": 52
    },
    {
      "epoch": 0.4623773173391494,
      "grad_norm": 1.292912081295997,
      "learning_rate": 7.43859649122807e-05,
      "loss": 0.6888,
      "step": 53
    },
    {
      "epoch": 0.47110141766630315,
      "grad_norm": 1.33549845587438,
      "learning_rate": 7.578947368421052e-05,
      "loss": 0.6989,
      "step": 54
    },
    {
      "epoch": 0.4798255179934569,
      "grad_norm": 1.6285853466115705,
      "learning_rate": 7.719298245614036e-05,
      "loss": 0.7016,
      "step": 55
    },
    {
      "epoch": 0.48854961832061067,
      "grad_norm": 0.9433375871219883,
      "learning_rate": 7.859649122807017e-05,
      "loss": 0.6824,
      "step": 56
    },
    {
      "epoch": 0.49727371864776443,
      "grad_norm": 1.2489894514867836,
      "learning_rate": 8e-05,
      "loss": 0.6898,
      "step": 57
    },
    {
      "epoch": 0.5059978189749182,
      "grad_norm": 1.60283780700127,
      "learning_rate": 7.99992499440621e-05,
      "loss": 0.7019,
      "step": 58
    },
    {
      "epoch": 0.514721919302072,
      "grad_norm": 1.349411337094158,
      "learning_rate": 7.999699980437755e-05,
      "loss": 0.692,
      "step": 59
    },
    {
      "epoch": 0.5234460196292258,
      "grad_norm": 1.4027794489135887,
      "learning_rate": 7.999324966533291e-05,
      "loss": 0.6739,
      "step": 60
    },
    {
      "epoch": 0.5321701199563795,
      "grad_norm": 0.8999831714962346,
      "learning_rate": 7.998799966756889e-05,
      "loss": 0.6925,
      "step": 61
    },
    {
      "epoch": 0.5408942202835333,
      "grad_norm": 1.2349556182219792,
      "learning_rate": 7.998125000797506e-05,
      "loss": 0.6811,
      "step": 62
    },
    {
      "epoch": 0.549618320610687,
      "grad_norm": 0.9953482513645127,
      "learning_rate": 7.997300093968255e-05,
      "loss": 0.6919,
      "step": 63
    },
    {
      "epoch": 0.5583424209378408,
      "grad_norm": 1.7631278928848257,
      "learning_rate": 7.99632527720545e-05,
      "loss": 0.7076,
      "step": 64
    },
    {
      "epoch": 0.5670665212649946,
      "grad_norm": 0.9802682486455654,
      "learning_rate": 7.995200587067445e-05,
      "loss": 0.6836,
      "step": 65
    },
    {
      "epoch": 0.5757906215921483,
      "grad_norm": 1.4018686830735987,
      "learning_rate": 7.993926065733265e-05,
      "loss": 0.6998,
      "step": 66
    },
    {
      "epoch": 0.5845147219193021,
      "grad_norm": 1.0081781512305683,
      "learning_rate": 7.992501761001027e-05,
      "loss": 0.6878,
      "step": 67
    },
    {
      "epoch": 0.5932388222464559,
      "grad_norm": 1.2813118135682324,
      "learning_rate": 7.99092772628614e-05,
      "loss": 0.6948,
      "step": 68
    },
    {
      "epoch": 0.6019629225736096,
      "grad_norm": 0.9281112795468938,
      "learning_rate": 7.98920402061931e-05,
      "loss": 0.6969,
      "step": 69
    },
    {
      "epoch": 0.6106870229007634,
      "grad_norm": 1.2375443626494387,
      "learning_rate": 7.987330708644319e-05,
      "loss": 0.6793,
      "step": 70
    },
    {
      "epoch": 0.6194111232279171,
      "grad_norm": 1.0863449033166035,
      "learning_rate": 7.985307860615607e-05,
      "loss": 0.6848,
      "step": 71
    },
    {
      "epoch": 0.6281352235550709,
      "grad_norm": 0.942362107324505,
      "learning_rate": 7.98313555239563e-05,
      "loss": 0.6862,
      "step": 72
    },
    {
      "epoch": 0.6368593238822247,
      "grad_norm": 1.3027161650276327,
      "learning_rate": 7.980813865452026e-05,
      "loss": 0.6735,
      "step": 73
    },
    {
      "epoch": 0.6455834242093784,
      "grad_norm": 0.7977961043410647,
      "learning_rate": 7.978342886854546e-05,
      "loss": 0.6787,
      "step": 74
    },
    {
      "epoch": 0.6543075245365322,
      "grad_norm": 1.3693856212692685,
      "learning_rate": 7.975722709271799e-05,
      "loss": 0.6751,
      "step": 75
    },
    {
      "epoch": 0.6630316248636859,
      "grad_norm": 0.7738139399748106,
      "learning_rate": 7.972953430967773e-05,
      "loss": 0.6726,
      "step": 76
    },
    {
      "epoch": 0.6717557251908397,
      "grad_norm": 0.6453937973753266,
      "learning_rate": 7.97003515579815e-05,
      "loss": 0.6756,
      "step": 77
    },
    {
      "epoch": 0.6804798255179935,
      "grad_norm": 0.7085908664971486,
      "learning_rate": 7.96696799320641e-05,
      "loss": 0.6774,
      "step": 78
    },
    {
      "epoch": 0.6892039258451472,
      "grad_norm": 0.7420105586344222,
      "learning_rate": 7.96375205821973e-05,
      "loss": 0.6647,
      "step": 79
    },
    {
      "epoch": 0.697928026172301,
      "grad_norm": 1.182903610639294,
      "learning_rate": 7.960387471444666e-05,
      "loss": 0.6758,
      "step": 80
    },
    {
      "epoch": 0.7066521264994547,
      "grad_norm": 1.305515943296027,
      "learning_rate": 7.956874359062632e-05,
      "loss": 0.6808,
      "step": 81
    },
    {
      "epoch": 0.7153762268266085,
      "grad_norm": 0.9451990102208602,
      "learning_rate": 7.95321285282517e-05,
      "loss": 0.6697,
      "step": 82
    },
    {
      "epoch": 0.7241003271537623,
      "grad_norm": 0.9432391308082473,
      "learning_rate": 7.949403090049002e-05,
      "loss": 0.6723,
      "step": 83
    },
    {
      "epoch": 0.732824427480916,
      "grad_norm": 1.1371193647756934,
      "learning_rate": 7.94544521361089e-05,
      "loss": 0.6773,
      "step": 84
    },
    {
      "epoch": 0.7415485278080698,
      "grad_norm": 1.1858932832021734,
      "learning_rate": 7.941339371942269e-05,
      "loss": 0.6657,
      "step": 85
    },
    {
      "epoch": 0.7502726281352236,
      "grad_norm": 1.165847606968587,
      "learning_rate": 7.937085719023685e-05,
      "loss": 0.6743,
      "step": 86
    },
    {
      "epoch": 0.7589967284623773,
      "grad_norm": 0.5770889497138026,
      "learning_rate": 7.932684414379021e-05,
      "loss": 0.668,
      "step": 87
    },
    {
      "epoch": 0.7677208287895311,
      "grad_norm": 0.925349798073699,
      "learning_rate": 7.928135623069509e-05,
      "loss": 0.6693,
      "step": 88
    },
    {
      "epoch": 0.7764449291166848,
      "grad_norm": 1.2265564072341149,
      "learning_rate": 7.923439515687546e-05,
      "loss": 0.6721,
      "step": 89
    },
    {
      "epoch": 0.7851690294438386,
      "grad_norm": 0.76774338811776,
      "learning_rate": 7.918596268350296e-05,
      "loss": 0.664,
      "step": 90
    },
    {
      "epoch": 0.7938931297709924,
      "grad_norm": 0.811830002272808,
      "learning_rate": 7.913606062693077e-05,
      "loss": 0.6674,
      "step": 91
    },
    {
      "epoch": 0.8026172300981461,
      "grad_norm": 0.7896885933612465,
      "learning_rate": 7.90846908586256e-05,
      "loss": 0.6603,
      "step": 92
    },
    {
      "epoch": 0.8113413304252999,
      "grad_norm": 0.6860009351157326,
      "learning_rate": 7.903185530509743e-05,
      "loss": 0.6541,
      "step": 93
    },
    {
      "epoch": 0.8200654307524536,
      "grad_norm": 0.7887179980130798,
      "learning_rate": 7.89775559478273e-05,
      "loss": 0.6719,
      "step": 94
    },
    {
      "epoch": 0.8287895310796074,
      "grad_norm": 0.6863156036434349,
      "learning_rate": 7.892179482319297e-05,
      "loss": 0.6712,
      "step": 95
    },
    {
      "epoch": 0.8375136314067612,
      "grad_norm": 0.612318445743932,
      "learning_rate": 7.886457402239256e-05,
      "loss": 0.6588,
      "step": 96
    },
    {
      "epoch": 0.8462377317339149,
      "grad_norm": 0.5266659568320967,
      "learning_rate": 7.880589569136616e-05,
      "loss": 0.6477,
      "step": 97
    },
    {
      "epoch": 0.8549618320610687,
      "grad_norm": 0.6815563256417184,
      "learning_rate": 7.874576203071531e-05,
      "loss": 0.6647,
      "step": 98
    },
    {
      "epoch": 0.8636859323882224,
      "grad_norm": 0.7021771362356389,
      "learning_rate": 7.868417529562043e-05,
      "loss": 0.6519,
      "step": 99
    },
    {
      "epoch": 0.8724100327153762,
      "grad_norm": 0.9537663707386643,
      "learning_rate": 7.862113779575638e-05,
      "loss": 0.6655,
      "step": 100
    },
    {
      "epoch": 0.88113413304253,
      "grad_norm": 1.7262458672727332,
      "learning_rate": 7.85566518952057e-05,
      "loss": 0.6617,
      "step": 101
    },
    {
      "epoch": 0.8898582333696837,
      "grad_norm": 0.5924197078630852,
      "learning_rate": 7.849072001237001e-05,
      "loss": 0.6586,
      "step": 102
    },
    {
      "epoch": 0.8985823336968375,
      "grad_norm": 1.784429399577841,
      "learning_rate": 7.842334461987936e-05,
      "loss": 0.6718,
      "step": 103
    },
    {
      "epoch": 0.9073064340239912,
      "grad_norm": 0.8548005705828143,
      "learning_rate": 7.835452824449935e-05,
      "loss": 0.6589,
      "step": 104
    },
    {
      "epoch": 0.916030534351145,
      "grad_norm": 1.5914313223828267,
      "learning_rate": 7.828427346703657e-05,
      "loss": 0.6639,
      "step": 105
    },
    {
      "epoch": 0.9247546346782988,
      "grad_norm": 1.1912921777513241,
      "learning_rate": 7.821258292224166e-05,
      "loss": 0.669,
      "step": 106
    },
    {
      "epoch": 0.9334787350054525,
      "grad_norm": 1.1951669530416693,
      "learning_rate": 7.813945929871056e-05,
      "loss": 0.6497,
      "step": 107
    },
    {
      "epoch": 0.9422028353326063,
      "grad_norm": 1.037062568022807,
      "learning_rate": 7.806490533878368e-05,
      "loss": 0.6632,
      "step": 108
    },
    {
      "epoch": 0.95092693565976,
      "grad_norm": 1.2483557427639183,
      "learning_rate": 7.798892383844303e-05,
      "loss": 0.658,
      "step": 109
    },
    {
      "epoch": 0.9596510359869138,
      "grad_norm": 0.8133626652288988,
      "learning_rate": 7.791151764720737e-05,
      "loss": 0.6571,
      "step": 110
    },
    {
      "epoch": 0.9683751363140676,
      "grad_norm": 1.0059615445006975,
      "learning_rate": 7.783268966802539e-05,
      "loss": 0.6595,
      "step": 111
    },
    {
      "epoch": 0.9770992366412213,
      "grad_norm": 0.7304172503042858,
      "learning_rate": 7.775244285716679e-05,
      "loss": 0.6608,
      "step": 112
    },
    {
      "epoch": 0.9858233369683751,
      "grad_norm": 0.7284490288054312,
      "learning_rate": 7.767078022411139e-05,
      "loss": 0.6492,
      "step": 113
    },
    {
      "epoch": 0.9945474372955289,
      "grad_norm": 0.5901533647047473,
      "learning_rate": 7.758770483143634e-05,
      "loss": 0.6642,
      "step": 114
    },
    {
      "epoch": 1.0038167938931297,
      "grad_norm": 0.8414397249541458,
      "learning_rate": 7.750321979470123e-05,
      "loss": 0.9152,
      "step": 115
    },
    {
      "epoch": 1.0125408942202836,
      "grad_norm": 1.0739193248054102,
      "learning_rate": 7.741732828233124e-05,
      "loss": 0.626,
      "step": 116
    },
    {
      "epoch": 1.0212649945474372,
      "grad_norm": 1.1988463490729344,
      "learning_rate": 7.733003351549829e-05,
      "loss": 0.6393,
      "step": 117
    },
    {
      "epoch": 1.0299890948745911,
      "grad_norm": 0.9606655160302263,
      "learning_rate": 7.724133876800031e-05,
      "loss": 0.625,
      "step": 118
    },
    {
      "epoch": 1.0387131952017448,
      "grad_norm": 1.2372012961246242,
      "learning_rate": 7.715124736613839e-05,
      "loss": 0.6278,
      "step": 119
    },
    {
      "epoch": 1.0474372955288986,
      "grad_norm": 0.86001815183547,
      "learning_rate": 7.705976268859207e-05,
      "loss": 0.6187,
      "step": 120
    },
    {
      "epoch": 1.0561613958560523,
      "grad_norm": 1.1413833226380308,
      "learning_rate": 7.696688816629266e-05,
      "loss": 0.63,
      "step": 121
    },
    {
      "epoch": 1.0648854961832062,
      "grad_norm": 1.2327702775009692,
      "learning_rate": 7.687262728229447e-05,
      "loss": 0.6205,
      "step": 122
    },
    {
      "epoch": 1.0736095965103598,
      "grad_norm": 0.7767212251506833,
      "learning_rate": 7.677698357164431e-05,
      "loss": 0.6177,
      "step": 123
    },
    {
      "epoch": 1.0823336968375137,
      "grad_norm": 1.7079661807939999,
      "learning_rate": 7.667996062124884e-05,
      "loss": 0.6242,
      "step": 124
    },
    {
      "epoch": 1.0910577971646673,
      "grad_norm": 1.0068734700150146,
      "learning_rate": 7.658156206974005e-05,
      "loss": 0.6183,
      "step": 125
    },
    {
      "epoch": 1.0997818974918212,
      "grad_norm": 1.8753922872596245,
      "learning_rate": 7.648179160733883e-05,
      "loss": 0.6337,
      "step": 126
    },
    {
      "epoch": 1.1085059978189749,
      "grad_norm": 1.8578444866986128,
      "learning_rate": 7.638065297571657e-05,
      "loss": 0.6171,
      "step": 127
    },
    {
      "epoch": 1.1172300981461287,
      "grad_norm": 0.6478551153484415,
      "learning_rate": 7.627814996785484e-05,
      "loss": 0.6158,
      "step": 128
    },
    {
      "epoch": 1.1259541984732824,
      "grad_norm": 1.4820468569461809,
      "learning_rate": 7.61742864279031e-05,
      "loss": 0.6197,
      "step": 129
    },
    {
      "epoch": 1.1346782988004362,
      "grad_norm": 0.7243487925020168,
      "learning_rate": 7.606906625103464e-05,
      "loss": 0.625,
      "step": 130
    },
    {
      "epoch": 1.14340239912759,
      "grad_norm": 0.9539104928282168,
      "learning_rate": 7.596249338330034e-05,
      "loss": 0.6194,
      "step": 131
    },
    {
      "epoch": 1.1521264994547438,
      "grad_norm": 0.8136844990028549,
      "learning_rate": 7.585457182148081e-05,
      "loss": 0.6172,
      "step": 132
    },
    {
      "epoch": 1.1608505997818974,
      "grad_norm": 0.7125736106844306,
      "learning_rate": 7.57453056129365e-05,
      "loss": 0.6151,
      "step": 133
    },
    {
      "epoch": 1.1695747001090513,
      "grad_norm": 0.5804990189493681,
      "learning_rate": 7.56346988554558e-05,
      "loss": 0.6256,
      "step": 134
    },
    {
      "epoch": 1.178298800436205,
      "grad_norm": 0.6362204084540434,
      "learning_rate": 7.552275569710152e-05,
      "loss": 0.6184,
      "step": 135
    },
    {
      "epoch": 1.1870229007633588,
      "grad_norm": 0.43420077191532125,
      "learning_rate": 7.540948033605513e-05,
      "loss": 0.6105,
      "step": 136
    },
    {
      "epoch": 1.1957470010905125,
      "grad_norm": 0.46232316550703767,
      "learning_rate": 7.529487702045953e-05,
      "loss": 0.6056,
      "step": 137
    },
    {
      "epoch": 1.2044711014176663,
      "grad_norm": 0.4316498267915253,
      "learning_rate": 7.517895004825956e-05,
      "loss": 0.6011,
      "step": 138
    },
    {
      "epoch": 1.21319520174482,
      "grad_norm": 0.38696383075278273,
      "learning_rate": 7.506170376704095e-05,
      "loss": 0.6141,
      "step": 139
    },
    {
      "epoch": 1.2219193020719739,
      "grad_norm": 0.3858939834475368,
      "learning_rate": 7.494314257386715e-05,
      "loss": 0.6065,
      "step": 140
    },
    {
      "epoch": 1.2306434023991275,
      "grad_norm": 0.3738279653817542,
      "learning_rate": 7.48232709151145e-05,
      "loss": 0.6104,
      "step": 141
    },
    {
      "epoch": 1.2393675027262814,
      "grad_norm": 0.34434724315715803,
      "learning_rate": 7.470209328630548e-05,
      "loss": 0.605,
      "step": 142
    },
    {
      "epoch": 1.248091603053435,
      "grad_norm": 0.3291602472493041,
      "learning_rate": 7.457961423194011e-05,
      "loss": 0.6018,
      "step": 143
    },
    {
      "epoch": 1.256815703380589,
      "grad_norm": 0.27837095142325796,
      "learning_rate": 7.445583834532546e-05,
      "loss": 0.5993,
      "step": 144
    },
    {
      "epoch": 1.2655398037077425,
      "grad_norm": 0.31225187743500077,
      "learning_rate": 7.433077026840346e-05,
      "loss": 0.5984,
      "step": 145
    },
    {
      "epoch": 1.2742639040348964,
      "grad_norm": 0.27063932853508044,
      "learning_rate": 7.420441469157684e-05,
      "loss": 0.6048,
      "step": 146
    },
    {
      "epoch": 1.2829880043620503,
      "grad_norm": 0.28328810560800655,
      "learning_rate": 7.407677635353308e-05,
      "loss": 0.6061,
      "step": 147
    },
    {
      "epoch": 1.291712104689204,
      "grad_norm": 0.3335236260280515,
      "learning_rate": 7.39478600410669e-05,
      "loss": 0.6055,
      "step": 148
    },
    {
      "epoch": 1.3004362050163576,
      "grad_norm": 0.34277641090811606,
      "learning_rate": 7.381767058890056e-05,
      "loss": 0.5956,
      "step": 149
    },
    {
      "epoch": 1.3091603053435115,
      "grad_norm": 0.3944860323036086,
      "learning_rate": 7.368621287950264e-05,
      "loss": 0.5971,
      "step": 150
    },
    {
      "epoch": 1.3178844056706653,
      "grad_norm": 0.390412169412763,
      "learning_rate": 7.355349184290491e-05,
      "loss": 0.5991,
      "step": 151
    },
    {
      "epoch": 1.326608505997819,
      "grad_norm": 0.44167362977698205,
      "learning_rate": 7.341951245651747e-05,
      "loss": 0.5905,
      "step": 152
    },
    {
      "epoch": 1.3353326063249726,
      "grad_norm": 0.4873733423530836,
      "learning_rate": 7.328427974494201e-05,
      "loss": 0.6061,
      "step": 153
    },
    {
      "epoch": 1.3440567066521265,
      "grad_norm": 0.529005701508295,
      "learning_rate": 7.314779877978346e-05,
      "loss": 0.6038,
      "step": 154
    },
    {
      "epoch": 1.3527808069792804,
      "grad_norm": 0.5776151078666283,
      "learning_rate": 7.301007467945974e-05,
      "loss": 0.6063,
      "step": 155
    },
    {
      "epoch": 1.361504907306434,
      "grad_norm": 0.6712943884189805,
      "learning_rate": 7.28711126090098e-05,
      "loss": 0.609,
      "step": 156
    },
    {
      "epoch": 1.3702290076335877,
      "grad_norm": 0.6974077084236574,
      "learning_rate": 7.273091777989997e-05,
      "loss": 0.617,
      "step": 157
    },
    {
      "epoch": 1.3789531079607416,
      "grad_norm": 0.622800075821039,
      "learning_rate": 7.258949544982843e-05,
      "loss": 0.6012,
      "step": 158
    },
    {
      "epoch": 1.3876772082878954,
      "grad_norm": 0.6652165160394861,
      "learning_rate": 7.24468509225281e-05,
      "loss": 0.6213,
      "step": 159
    },
    {
      "epoch": 1.396401308615049,
      "grad_norm": 0.7796997637986155,
      "learning_rate": 7.230298954756772e-05,
      "loss": 0.6116,
      "step": 160
    },
    {
      "epoch": 1.4051254089422027,
      "grad_norm": 0.8050974173338423,
      "learning_rate": 7.215791672015121e-05,
      "loss": 0.6102,
      "step": 161
    },
    {
      "epoch": 1.4138495092693566,
      "grad_norm": 0.6245666904233175,
      "learning_rate": 7.201163788091536e-05,
      "loss": 0.6051,
      "step": 162
    },
    {
      "epoch": 1.4225736095965105,
      "grad_norm": 0.2791815826660419,
      "learning_rate": 7.186415851572579e-05,
      "loss": 0.5978,
      "step": 163
    },
    {
      "epoch": 1.4312977099236641,
      "grad_norm": 0.5658363527146323,
      "learning_rate": 7.171548415547114e-05,
      "loss": 0.6064,
      "step": 164
    },
    {
      "epoch": 1.4400218102508178,
      "grad_norm": 0.6010202995307374,
      "learning_rate": 7.156562037585576e-05,
      "loss": 0.6117,
      "step": 165
    },
    {
      "epoch": 1.4487459105779716,
      "grad_norm": 0.45545843396243296,
      "learning_rate": 7.141457279719053e-05,
      "loss": 0.6079,
      "step": 166
    },
    {
      "epoch": 1.4574700109051255,
      "grad_norm": 0.37879238626475686,
      "learning_rate": 7.126234708418214e-05,
      "loss": 0.5965,
      "step": 167
    },
    {
      "epoch": 1.4661941112322792,
      "grad_norm": 0.4580319723211172,
      "learning_rate": 7.110894894572056e-05,
      "loss": 0.6106,
      "step": 168
    },
    {
      "epoch": 1.4749182115594328,
      "grad_norm": 0.43310826738894753,
      "learning_rate": 7.095438413466503e-05,
      "loss": 0.6152,
      "step": 169
    },
    {
      "epoch": 1.4836423118865867,
      "grad_norm": 0.42673749198290256,
      "learning_rate": 7.079865844762829e-05,
      "loss": 0.6058,
      "step": 170
    },
    {
      "epoch": 1.4923664122137406,
      "grad_norm": 0.5356882769390818,
      "learning_rate": 7.064177772475912e-05,
      "loss": 0.6011,
      "step": 171
    },
    {
      "epoch": 1.5010905125408942,
      "grad_norm": 0.48453744479417066,
      "learning_rate": 7.048374784952343e-05,
      "loss": 0.6014,
      "step": 172
    },
    {
      "epoch": 1.5098146128680479,
      "grad_norm": 0.40573397034428954,
      "learning_rate": 7.03245747484835e-05,
      "loss": 0.6021,
      "step": 173
    },
    {
      "epoch": 1.5185387131952017,
      "grad_norm": 0.5052592973873418,
      "learning_rate": 7.016426439107586e-05,
      "loss": 0.5976,
      "step": 174
    },
    {
      "epoch": 1.5272628135223556,
      "grad_norm": 0.5640845358828827,
      "learning_rate": 7.000282278938724e-05,
      "loss": 0.6032,
      "step": 175
    },
    {
      "epoch": 1.5359869138495092,
      "grad_norm": 0.5716127749488464,
      "learning_rate": 6.984025599792926e-05,
      "loss": 0.6069,
      "step": 176
    },
    {
      "epoch": 1.544711014176663,
      "grad_norm": 0.6147245538888289,
      "learning_rate": 6.967657011341126e-05,
      "loss": 0.6017,
      "step": 177
    },
    {
      "epoch": 1.5534351145038168,
      "grad_norm": 0.5972264181543786,
      "learning_rate": 6.951177127451177e-05,
      "loss": 0.6039,
      "step": 178
    },
    {
      "epoch": 1.5621592148309706,
      "grad_norm": 0.4511378630106592,
      "learning_rate": 6.934586566164811e-05,
      "loss": 0.609,
      "step": 179
    },
    {
      "epoch": 1.5708833151581243,
      "grad_norm": 0.3324255798317227,
      "learning_rate": 6.917885949674483e-05,
      "loss": 0.6046,
      "step": 180
    },
    {
      "epoch": 1.579607415485278,
      "grad_norm": 0.3318537793653575,
      "learning_rate": 6.901075904300021e-05,
      "loss": 0.5984,
      "step": 181
    },
    {
      "epoch": 1.5883315158124318,
      "grad_norm": 0.4025671963472955,
      "learning_rate": 6.88415706046514e-05,
      "loss": 0.5991,
      "step": 182
    },
    {
      "epoch": 1.5970556161395857,
      "grad_norm": 0.4039375325984613,
      "learning_rate": 6.867130052673806e-05,
      "loss": 0.6132,
      "step": 183
    },
    {
      "epoch": 1.6057797164667393,
      "grad_norm": 0.42633324449837107,
      "learning_rate": 6.849995519486434e-05,
      "loss": 0.6112,
      "step": 184
    },
    {
      "epoch": 1.614503816793893,
      "grad_norm": 0.4031771610559183,
      "learning_rate": 6.832754103495939e-05,
      "loss": 0.5951,
      "step": 185
    },
    {
      "epoch": 1.6232279171210469,
      "grad_norm": 0.3101345676851429,
      "learning_rate": 6.815406451303647e-05,
      "loss": 0.5977,
      "step": 186
    },
    {
      "epoch": 1.6319520174482007,
      "grad_norm": 0.28180350855331504,
      "learning_rate": 6.797953213495033e-05,
      "loss": 0.6081,
      "step": 187
    },
    {
      "epoch": 1.6406761177753544,
      "grad_norm": 0.3450910788047459,
      "learning_rate": 6.780395044615329e-05,
      "loss": 0.606,
      "step": 188
    },
    {
      "epoch": 1.649400218102508,
      "grad_norm": 0.3941732240418325,
      "learning_rate": 6.762732603144978e-05,
      "loss": 0.6079,
      "step": 189
    },
    {
      "epoch": 1.658124318429662,
      "grad_norm": 0.3241938707733741,
      "learning_rate": 6.744966551474936e-05,
      "loss": 0.604,
      "step": 190
    },
    {
      "epoch": 1.6668484187568158,
      "grad_norm": 0.27804934781744717,
      "learning_rate": 6.727097555881826e-05,
      "loss": 0.595,
      "step": 191
    },
    {
      "epoch": 1.6755725190839694,
      "grad_norm": 0.25336159667743025,
      "learning_rate": 6.709126286502965e-05,
      "loss": 0.6025,
      "step": 192
    },
    {
      "epoch": 1.684296619411123,
      "grad_norm": 0.29341065316587794,
      "learning_rate": 6.691053417311216e-05,
      "loss": 0.6023,
      "step": 193
    },
    {
      "epoch": 1.693020719738277,
      "grad_norm": 0.2878985409168238,
      "learning_rate": 6.672879626089723e-05,
      "loss": 0.5949,
      "step": 194
    },
    {
      "epoch": 1.7017448200654308,
      "grad_norm": 0.26927072720134837,
      "learning_rate": 6.654605594406486e-05,
      "loss": 0.5979,
      "step": 195
    },
    {
      "epoch": 1.7104689203925845,
      "grad_norm": 0.3155329601805801,
      "learning_rate": 6.636232007588805e-05,
      "loss": 0.587,
      "step": 196
    },
    {
      "epoch": 1.7191930207197381,
      "grad_norm": 0.34847815991034525,
      "learning_rate": 6.617759554697573e-05,
      "loss": 0.5978,
      "step": 197
    },
    {
      "epoch": 1.727917121046892,
      "grad_norm": 0.28896074500408486,
      "learning_rate": 6.59918892850144e-05,
      "loss": 0.6069,
      "step": 198
    },
    {
      "epoch": 1.7366412213740459,
      "grad_norm": 0.2981626524269195,
      "learning_rate": 6.580520825450827e-05,
      "loss": 0.5956,
      "step": 199
    },
    {
      "epoch": 1.7453653217011995,
      "grad_norm": 0.31680273274222387,
      "learning_rate": 6.561755945651813e-05,
      "loss": 0.5907,
      "step": 200
    },
    {
      "epoch": 1.7540894220283532,
      "grad_norm": 0.2788023167998957,
      "learning_rate": 6.542894992839873e-05,
      "loss": 0.5988,
      "step": 201
    },
    {
      "epoch": 1.762813522355507,
      "grad_norm": 0.26041179212976306,
      "learning_rate": 6.52393867435349e-05,
      "loss": 0.5879,
      "step": 202
    },
    {
      "epoch": 1.771537622682661,
      "grad_norm": 0.34950012098507055,
      "learning_rate": 6.504887701107626e-05,
      "loss": 0.5933,
      "step": 203
    },
    {
      "epoch": 1.7802617230098146,
      "grad_norm": 0.5298511066656313,
      "learning_rate": 6.48574278756706e-05,
      "loss": 0.59,
      "step": 204
    },
    {
      "epoch": 1.7889858233369684,
      "grad_norm": 0.6366920336388178,
      "learning_rate": 6.466504651719598e-05,
      "loss": 0.5967,
      "step": 205
    },
    {
      "epoch": 1.7977099236641223,
      "grad_norm": 0.6299010554780248,
      "learning_rate": 6.447174015049139e-05,
      "loss": 0.5974,
      "step": 206
    },
    {
      "epoch": 1.806434023991276,
      "grad_norm": 0.562858852871968,
      "learning_rate": 6.427751602508628e-05,
      "loss": 0.6057,
      "step": 207
    },
    {
      "epoch": 1.8151581243184296,
      "grad_norm": 0.4004595325182879,
      "learning_rate": 6.408238142492855e-05,
      "loss": 0.59,
      "step": 208
    },
    {
      "epoch": 1.8238822246455835,
      "grad_norm": 0.21584901270572757,
      "learning_rate": 6.388634366811146e-05,
      "loss": 0.5943,
      "step": 209
    },
    {
      "epoch": 1.8326063249727373,
      "grad_norm": 0.37265049074019335,
      "learning_rate": 6.368941010659921e-05,
      "loss": 0.6096,
      "step": 210
    },
    {
      "epoch": 1.841330425299891,
      "grad_norm": 0.5267910816953789,
      "learning_rate": 6.349158812595116e-05,
      "loss": 0.6053,
      "step": 211
    },
    {
      "epoch": 1.8500545256270446,
      "grad_norm": 0.47801721985552514,
      "learning_rate": 6.329288514504487e-05,
      "loss": 0.5936,
      "step": 212
    },
    {
      "epoch": 1.8587786259541985,
      "grad_norm": 0.27531874942679646,
      "learning_rate": 6.309330861579786e-05,
      "loss": 0.5997,
      "step": 213
    },
    {
      "epoch": 1.8675027262813524,
      "grad_norm": 0.3657379875213062,
      "learning_rate": 6.28928660228882e-05,
      "loss": 0.5963,
      "step": 214
    },
    {
      "epoch": 1.876226826608506,
      "grad_norm": 0.3633999933893769,
      "learning_rate": 6.269156488347372e-05,
      "loss": 0.5899,
      "step": 215
    },
    {
      "epoch": 1.8849509269356597,
      "grad_norm": 0.2855859194806056,
      "learning_rate": 6.248941274691017e-05,
      "loss": 0.5963,
      "step": 216
    },
    {
      "epoch": 1.8936750272628136,
      "grad_norm": 0.26871535087996673,
      "learning_rate": 6.228641719446808e-05,
      "loss": 0.5832,
      "step": 217
    },
    {
      "epoch": 1.9023991275899674,
      "grad_norm": 0.26413333351828056,
      "learning_rate": 6.208258583904841e-05,
      "loss": 0.5913,
      "step": 218
    },
    {
      "epoch": 1.911123227917121,
      "grad_norm": 0.27861019506862306,
      "learning_rate": 6.18779263248971e-05,
      "loss": 0.6088,
      "step": 219
    },
    {
      "epoch": 1.9198473282442747,
      "grad_norm": 0.26490994643954596,
      "learning_rate": 6.16724463273183e-05,
      "loss": 0.6072,
      "step": 220
    },
    {
      "epoch": 1.9285714285714286,
      "grad_norm": 0.2464264085957813,
      "learning_rate": 6.146615355238668e-05,
      "loss": 0.5998,
      "step": 221
    },
    {
      "epoch": 1.9372955288985825,
      "grad_norm": 0.2060880785957384,
      "learning_rate": 6.125905573665824e-05,
      "loss": 0.5998,
      "step": 222
    },
    {
      "epoch": 1.9460196292257361,
      "grad_norm": 0.26971063557098873,
      "learning_rate": 6.105116064688033e-05,
      "loss": 0.6048,
      "step": 223
    },
    {
      "epoch": 1.9547437295528898,
      "grad_norm": 0.32136759035233387,
      "learning_rate": 6.0842476079700264e-05,
      "loss": 0.6019,
      "step": 224
    },
    {
      "epoch": 1.9634678298800436,
      "grad_norm": 0.23286070304886092,
      "learning_rate": 6.063300986137297e-05,
      "loss": 0.5978,
      "step": 225
    },
    {
      "epoch": 1.9721919302071975,
      "grad_norm": 0.18824642933626035,
      "learning_rate": 6.04227698474675e-05,
      "loss": 0.5853,
      "step": 226
    },
    {
      "epoch": 1.9809160305343512,
      "grad_norm": 0.21674772391599217,
      "learning_rate": 6.02117639225724e-05,
      "loss": 0.6078,
      "step": 227
    },
    {
      "epoch": 1.9896401308615048,
      "grad_norm": 0.2934361150620414,
      "learning_rate": 6.000000000000001e-05,
      "loss": 0.5984,
      "step": 228
    },
    {
      "epoch": 1.9983642311886587,
      "grad_norm": 0.44422300505199064,
      "learning_rate": 5.9787486021489705e-05,
      "loss": 0.8056,
      "step": 229
    }
  ],
  "logging_steps": 1,
  "max_steps": 570,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6.092582874781843e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}