{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0005646527385659,
  "eval_steps": 222,
  "global_step": 443,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002258610954263128,
      "grad_norm": 0.734693706035614,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 2.1698,
      "step": 1
    },
    {
      "epoch": 0.002258610954263128,
      "eval_loss": 2.083223819732666,
      "eval_runtime": 98.3201,
      "eval_samples_per_second": 7.587,
      "eval_steps_per_second": 0.956,
      "step": 1
    },
    {
      "epoch": 0.004517221908526256,
      "grad_norm": 0.7944502830505371,
      "learning_rate": 6.666666666666667e-06,
      "loss": 2.2383,
      "step": 2
    },
    {
      "epoch": 0.006775832862789385,
      "grad_norm": 0.9303544759750366,
      "learning_rate": 1e-05,
      "loss": 1.952,
      "step": 3
    },
    {
      "epoch": 0.009034443817052512,
      "grad_norm": 0.9056912660598755,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 2.306,
      "step": 4
    },
    {
      "epoch": 0.01129305477131564,
      "grad_norm": 1.3058993816375732,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 2.3384,
      "step": 5
    },
    {
      "epoch": 0.01355166572557877,
      "grad_norm": 1.2961996793746948,
      "learning_rate": 2e-05,
      "loss": 2.3395,
      "step": 6
    },
    {
      "epoch": 0.015810276679841896,
      "grad_norm": 0.8225818276405334,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 1.6852,
      "step": 7
    },
    {
      "epoch": 0.018068887634105024,
      "grad_norm": 0.7992662787437439,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 1.7517,
      "step": 8
    },
    {
      "epoch": 0.020327498588368152,
      "grad_norm": 1.9861479997634888,
      "learning_rate": 3e-05,
      "loss": 2.6289,
      "step": 9
    },
    {
      "epoch": 0.02258610954263128,
      "grad_norm": 0.720796525478363,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 2.3053,
      "step": 10
    },
    {
      "epoch": 0.024844720496894408,
      "grad_norm": 0.8933535218238831,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 2.0577,
      "step": 11
    },
    {
      "epoch": 0.02710333145115754,
      "grad_norm": 1.1678780317306519,
      "learning_rate": 4e-05,
      "loss": 2.1599,
      "step": 12
    },
    {
      "epoch": 0.029361942405420668,
      "grad_norm": 1.06686270236969,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 1.9538,
      "step": 13
    },
    {
      "epoch": 0.03162055335968379,
      "grad_norm": 1.0956751108169556,
      "learning_rate": 4.666666666666667e-05,
      "loss": 1.8666,
      "step": 14
    },
    {
      "epoch": 0.03387916431394692,
      "grad_norm": 1.3107072114944458,
      "learning_rate": 5e-05,
      "loss": 2.0312,
      "step": 15
    },
    {
      "epoch": 0.03613777526821005,
      "grad_norm": 1.6173886060714722,
      "learning_rate": 5.333333333333333e-05,
      "loss": 1.9908,
      "step": 16
    },
    {
      "epoch": 0.038396386222473176,
      "grad_norm": 1.1099895238876343,
      "learning_rate": 5.666666666666667e-05,
      "loss": 1.6063,
      "step": 17
    },
    {
      "epoch": 0.040654997176736304,
      "grad_norm": 1.4625017642974854,
      "learning_rate": 6e-05,
      "loss": 1.6362,
      "step": 18
    },
    {
      "epoch": 0.04291360813099943,
      "grad_norm": 1.05028235912323,
      "learning_rate": 6.333333333333333e-05,
      "loss": 1.7419,
      "step": 19
    },
    {
      "epoch": 0.04517221908526256,
      "grad_norm": 0.9315014481544495,
      "learning_rate": 6.666666666666667e-05,
      "loss": 1.6991,
      "step": 20
    },
    {
      "epoch": 0.04743083003952569,
      "grad_norm": 1.2721033096313477,
      "learning_rate": 7e-05,
      "loss": 1.5631,
      "step": 21
    },
    {
      "epoch": 0.049689440993788817,
      "grad_norm": 1.578466534614563,
      "learning_rate": 7.333333333333333e-05,
      "loss": 1.6748,
      "step": 22
    },
    {
      "epoch": 0.05194805194805195,
      "grad_norm": 1.692025065422058,
      "learning_rate": 7.666666666666667e-05,
      "loss": 1.607,
      "step": 23
    },
    {
      "epoch": 0.05420666290231508,
      "grad_norm": 1.251129388809204,
      "learning_rate": 8e-05,
      "loss": 1.8119,
      "step": 24
    },
    {
      "epoch": 0.05646527385657821,
      "grad_norm": 1.340306043624878,
      "learning_rate": 8.333333333333334e-05,
      "loss": 1.6167,
      "step": 25
    },
    {
      "epoch": 0.058723884810841336,
      "grad_norm": 1.1570920944213867,
      "learning_rate": 8.666666666666667e-05,
      "loss": 1.7172,
      "step": 26
    },
    {
      "epoch": 0.060982495765104464,
      "grad_norm": 1.0314528942108154,
      "learning_rate": 9e-05,
      "loss": 1.8385,
      "step": 27
    },
    {
      "epoch": 0.06324110671936758,
      "grad_norm": 1.032861351966858,
      "learning_rate": 9.333333333333334e-05,
      "loss": 1.3839,
      "step": 28
    },
    {
      "epoch": 0.06549971767363072,
      "grad_norm": 0.9600235819816589,
      "learning_rate": 9.666666666666667e-05,
      "loss": 1.4608,
      "step": 29
    },
    {
      "epoch": 0.06775832862789384,
      "grad_norm": 0.9947760105133057,
      "learning_rate": 0.0001,
      "loss": 1.4875,
      "step": 30
    },
    {
      "epoch": 0.07001693958215698,
      "grad_norm": 1.2269097566604614,
      "learning_rate": 9.999855343632036e-05,
      "loss": 1.6437,
      "step": 31
    },
    {
      "epoch": 0.0722755505364201,
      "grad_norm": 0.865011990070343,
      "learning_rate": 9.999421382898329e-05,
      "loss": 1.6969,
      "step": 32
    },
    {
      "epoch": 0.07453416149068323,
      "grad_norm": 0.6729279160499573,
      "learning_rate": 9.998698142908953e-05,
      "loss": 1.2089,
      "step": 33
    },
    {
      "epoch": 0.07679277244494635,
      "grad_norm": 1.4308388233184814,
      "learning_rate": 9.997685665512418e-05,
      "loss": 1.3744,
      "step": 34
    },
    {
      "epoch": 0.07905138339920949,
      "grad_norm": 0.582622230052948,
      "learning_rate": 9.99638400929324e-05,
      "loss": 1.2607,
      "step": 35
    },
    {
      "epoch": 0.08130999435347261,
      "grad_norm": 0.5864912867546082,
      "learning_rate": 9.994793249568569e-05,
      "loss": 1.3634,
      "step": 36
    },
    {
      "epoch": 0.08356860530773574,
      "grad_norm": 0.5478243827819824,
      "learning_rate": 9.99291347838381e-05,
      "loss": 1.5985,
      "step": 37
    },
    {
      "epoch": 0.08582721626199886,
      "grad_norm": 0.7038874626159668,
      "learning_rate": 9.990744804507315e-05,
      "loss": 1.3606,
      "step": 38
    },
    {
      "epoch": 0.088085827216262,
      "grad_norm": 0.5414575934410095,
      "learning_rate": 9.988287353424077e-05,
      "loss": 1.7192,
      "step": 39
    },
    {
      "epoch": 0.09034443817052512,
      "grad_norm": 0.5917218327522278,
      "learning_rate": 9.985541267328477e-05,
      "loss": 1.5298,
      "step": 40
    },
    {
      "epoch": 0.09260304912478826,
      "grad_norm": 0.5787907838821411,
      "learning_rate": 9.98250670511605e-05,
      "loss": 1.9064,
      "step": 41
    },
    {
      "epoch": 0.09486166007905138,
      "grad_norm": 0.8571286201477051,
      "learning_rate": 9.979183842374293e-05,
      "loss": 1.3743,
      "step": 42
    },
    {
      "epoch": 0.09712027103331451,
      "grad_norm": 0.7187583446502686,
      "learning_rate": 9.975572871372513e-05,
      "loss": 1.3655,
      "step": 43
    },
    {
      "epoch": 0.09937888198757763,
      "grad_norm": 1.4509934186935425,
      "learning_rate": 9.971674001050686e-05,
      "loss": 1.4908,
      "step": 44
    },
    {
      "epoch": 0.10163749294184077,
      "grad_norm": 0.7309430837631226,
      "learning_rate": 9.967487457007381e-05,
      "loss": 1.0838,
      "step": 45
    },
    {
      "epoch": 0.1038961038961039,
      "grad_norm": 0.7827832102775574,
      "learning_rate": 9.963013481486703e-05,
      "loss": 1.5432,
      "step": 46
    },
    {
      "epoch": 0.10615471485036702,
      "grad_norm": 0.6648978590965271,
      "learning_rate": 9.958252333364267e-05,
      "loss": 1.1893,
      "step": 47
    },
    {
      "epoch": 0.10841332580463016,
      "grad_norm": 1.085577130317688,
      "learning_rate": 9.953204288132234e-05,
      "loss": 1.3416,
      "step": 48
    },
    {
      "epoch": 0.11067193675889328,
      "grad_norm": 1.0411715507507324,
      "learning_rate": 9.947869637883358e-05,
      "loss": 1.4301,
      "step": 49
    },
    {
      "epoch": 0.11293054771315642,
      "grad_norm": 0.6772856116294861,
      "learning_rate": 9.942248691294093e-05,
      "loss": 1.1494,
      "step": 50
    },
    {
      "epoch": 0.11518915866741954,
      "grad_norm": 1.099043846130371,
      "learning_rate": 9.936341773606723e-05,
      "loss": 1.4349,
      "step": 51
    },
    {
      "epoch": 0.11744776962168267,
      "grad_norm": 0.7465451955795288,
      "learning_rate": 9.930149226610554e-05,
      "loss": 1.4149,
      "step": 52
    },
    {
      "epoch": 0.11970638057594579,
      "grad_norm": 0.6757813096046448,
      "learning_rate": 9.923671408622129e-05,
      "loss": 1.3782,
      "step": 53
    },
    {
      "epoch": 0.12196499153020893,
      "grad_norm": 0.6110934019088745,
      "learning_rate": 9.916908694464492e-05,
      "loss": 1.5505,
      "step": 54
    },
    {
      "epoch": 0.12422360248447205,
      "grad_norm": 0.8363707065582275,
      "learning_rate": 9.909861475445517e-05,
      "loss": 1.6348,
      "step": 55
    },
    {
      "epoch": 0.12648221343873517,
      "grad_norm": 0.5607156753540039,
      "learning_rate": 9.902530159335243e-05,
      "loss": 1.6425,
      "step": 56
    },
    {
      "epoch": 0.1287408243929983,
      "grad_norm": 0.6306595802307129,
      "learning_rate": 9.894915170342295e-05,
      "loss": 1.2802,
      "step": 57
    },
    {
      "epoch": 0.13099943534726144,
      "grad_norm": 0.8412261009216309,
      "learning_rate": 9.887016949089333e-05,
      "loss": 1.5162,
      "step": 58
    },
    {
      "epoch": 0.13325804630152457,
      "grad_norm": 1.7483497858047485,
      "learning_rate": 9.878835952587559e-05,
      "loss": 1.4394,
      "step": 59
    },
    {
      "epoch": 0.13551665725578768,
      "grad_norm": 0.6815210580825806,
      "learning_rate": 9.870372654210265e-05,
      "loss": 1.4947,
      "step": 60
    },
    {
      "epoch": 0.13777526821005082,
      "grad_norm": 0.819712221622467,
      "learning_rate": 9.861627543665456e-05,
      "loss": 1.8876,
      "step": 61
    },
    {
      "epoch": 0.14003387916431395,
      "grad_norm": 0.6188887357711792,
      "learning_rate": 9.852601126967502e-05,
      "loss": 1.4265,
      "step": 62
    },
    {
      "epoch": 0.1422924901185771,
      "grad_norm": 0.5934920310974121,
      "learning_rate": 9.843293926407866e-05,
      "loss": 1.4272,
      "step": 63
    },
    {
      "epoch": 0.1445511010728402,
      "grad_norm": 0.554636538028717,
      "learning_rate": 9.833706480524878e-05,
      "loss": 1.6095,
      "step": 64
    },
    {
      "epoch": 0.14680971202710333,
      "grad_norm": 0.545122504234314,
      "learning_rate": 9.82383934407258e-05,
      "loss": 1.2429,
      "step": 65
    },
    {
      "epoch": 0.14906832298136646,
      "grad_norm": 0.674778401851654,
      "learning_rate": 9.81369308798862e-05,
      "loss": 1.3112,
      "step": 66
    },
    {
      "epoch": 0.1513269339356296,
      "grad_norm": 0.4785314202308655,
      "learning_rate": 9.803268299361217e-05,
      "loss": 1.2792,
      "step": 67
    },
    {
      "epoch": 0.1535855448898927,
      "grad_norm": 0.7849199175834656,
      "learning_rate": 9.7925655813952e-05,
      "loss": 1.557,
      "step": 68
    },
    {
      "epoch": 0.15584415584415584,
      "grad_norm": 0.7094368934631348,
      "learning_rate": 9.781585553377085e-05,
      "loss": 1.3404,
      "step": 69
    },
    {
      "epoch": 0.15810276679841898,
      "grad_norm": 0.5156518220901489,
      "learning_rate": 9.770328850639268e-05,
      "loss": 1.2939,
      "step": 70
    },
    {
      "epoch": 0.1603613777526821,
      "grad_norm": 0.6321349143981934,
      "learning_rate": 9.758796124523239e-05,
      "loss": 1.684,
      "step": 71
    },
    {
      "epoch": 0.16261998870694522,
      "grad_norm": 0.8690780401229858,
      "learning_rate": 9.746988042341906e-05,
      "loss": 1.2346,
      "step": 72
    },
    {
      "epoch": 0.16487859966120835,
      "grad_norm": 0.5986551642417908,
      "learning_rate": 9.734905287340985e-05,
      "loss": 1.4559,
      "step": 73
    },
    {
      "epoch": 0.1671372106154715,
      "grad_norm": 0.5338417887687683,
      "learning_rate": 9.722548558659457e-05,
      "loss": 1.6537,
      "step": 74
    },
    {
      "epoch": 0.16939582156973462,
      "grad_norm": 0.7366315126419067,
      "learning_rate": 9.709918571289114e-05,
      "loss": 1.3941,
      "step": 75
    },
    {
      "epoch": 0.17165443252399773,
      "grad_norm": 0.7297102212905884,
      "learning_rate": 9.697016056033201e-05,
      "loss": 1.5824,
      "step": 76
    },
    {
      "epoch": 0.17391304347826086,
      "grad_norm": 0.5532839894294739,
      "learning_rate": 9.683841759464113e-05,
      "loss": 1.4119,
      "step": 77
    },
    {
      "epoch": 0.176171654432524,
      "grad_norm": 0.5008260011672974,
      "learning_rate": 9.670396443880208e-05,
      "loss": 1.5714,
      "step": 78
    },
    {
      "epoch": 0.17843026538678713,
      "grad_norm": 0.7788993716239929,
      "learning_rate": 9.656680887261693e-05,
      "loss": 1.4765,
      "step": 79
    },
    {
      "epoch": 0.18068887634105024,
      "grad_norm": 0.7315067648887634,
      "learning_rate": 9.64269588322561e-05,
      "loss": 1.6218,
      "step": 80
    },
    {
      "epoch": 0.18294748729531338,
      "grad_norm": 0.7663245797157288,
      "learning_rate": 9.628442240979916e-05,
      "loss": 1.381,
      "step": 81
    },
    {
      "epoch": 0.1852060982495765,
      "grad_norm": 0.5851924419403076,
      "learning_rate": 9.613920785276656e-05,
      "loss": 1.6193,
      "step": 82
    },
    {
      "epoch": 0.18746470920383965,
      "grad_norm": 0.8395928144454956,
      "learning_rate": 9.599132356364247e-05,
      "loss": 1.3371,
      "step": 83
    },
    {
      "epoch": 0.18972332015810275,
      "grad_norm": 0.9145395159721375,
      "learning_rate": 9.584077809938855e-05,
      "loss": 1.3984,
      "step": 84
    },
    {
      "epoch": 0.1919819311123659,
      "grad_norm": 0.6220190525054932,
      "learning_rate": 9.568758017094883e-05,
      "loss": 1.2248,
      "step": 85
    },
    {
      "epoch": 0.19424054206662902,
      "grad_norm": 0.5492339730262756,
      "learning_rate": 9.553173864274567e-05,
      "loss": 1.4326,
      "step": 86
    },
    {
      "epoch": 0.19649915302089216,
      "grad_norm": 0.47741296887397766,
      "learning_rate": 9.537326253216685e-05,
      "loss": 1.4092,
      "step": 87
    },
    {
      "epoch": 0.19875776397515527,
      "grad_norm": 0.5220446586608887,
      "learning_rate": 9.521216100904378e-05,
      "loss": 1.5837,
      "step": 88
    },
    {
      "epoch": 0.2010163749294184,
      "grad_norm": 0.5066771507263184,
      "learning_rate": 9.504844339512095e-05,
      "loss": 1.4179,
      "step": 89
    },
    {
      "epoch": 0.20327498588368154,
      "grad_norm": 0.6398409605026245,
      "learning_rate": 9.488211916351656e-05,
      "loss": 1.3896,
      "step": 90
    },
    {
      "epoch": 0.20553359683794467,
      "grad_norm": 0.7091066241264343,
      "learning_rate": 9.471319793817426e-05,
      "loss": 1.6861,
      "step": 91
    },
    {
      "epoch": 0.2077922077922078,
      "grad_norm": 1.067958116531372,
      "learning_rate": 9.454168949330645e-05,
      "loss": 1.417,
      "step": 92
    },
    {
      "epoch": 0.2100508187464709,
      "grad_norm": 0.5348410606384277,
      "learning_rate": 9.436760375282859e-05,
      "loss": 1.4303,
      "step": 93
    },
    {
      "epoch": 0.21230942970073405,
      "grad_norm": 0.5271281599998474,
      "learning_rate": 9.419095078978506e-05,
      "loss": 1.2863,
      "step": 94
    },
    {
      "epoch": 0.21456804065499718,
      "grad_norm": 0.6372822523117065,
      "learning_rate": 9.40117408257663e-05,
      "loss": 1.4036,
      "step": 95
    },
    {
      "epoch": 0.21682665160926032,
      "grad_norm": 0.5258705615997314,
      "learning_rate": 9.382998423031727e-05,
      "loss": 1.7694,
      "step": 96
    },
    {
      "epoch": 0.21908526256352343,
      "grad_norm": 0.5894024968147278,
      "learning_rate": 9.364569152033756e-05,
      "loss": 1.7329,
      "step": 97
    },
    {
      "epoch": 0.22134387351778656,
      "grad_norm": 0.7126629948616028,
      "learning_rate": 9.345887335947281e-05,
      "loss": 1.6303,
      "step": 98
    },
    {
      "epoch": 0.2236024844720497,
      "grad_norm": 0.5408257842063904,
      "learning_rate": 9.326954055749767e-05,
      "loss": 1.6237,
      "step": 99
    },
    {
      "epoch": 0.22586109542631283,
      "grad_norm": 0.673343300819397,
      "learning_rate": 9.30777040696903e-05,
      "loss": 1.1694,
      "step": 100
    },
    {
      "epoch": 0.22811970638057594,
      "grad_norm": 0.6288356781005859,
      "learning_rate": 9.288337499619857e-05,
      "loss": 1.3256,
      "step": 101
    },
    {
      "epoch": 0.23037831733483907,
      "grad_norm": 0.5580915212631226,
      "learning_rate": 9.268656458139762e-05,
      "loss": 1.4626,
      "step": 102
    },
    {
      "epoch": 0.2326369282891022,
      "grad_norm": 0.5487126708030701,
      "learning_rate": 9.248728421323941e-05,
      "loss": 1.5227,
      "step": 103
    },
    {
      "epoch": 0.23489553924336534,
      "grad_norm": 0.712181031703949,
      "learning_rate": 9.22855454225936e-05,
      "loss": 1.082,
      "step": 104
    },
    {
      "epoch": 0.23715415019762845,
      "grad_norm": 0.7502368092536926,
      "learning_rate": 9.208135988258051e-05,
      "loss": 1.4804,
      "step": 105
    },
    {
      "epoch": 0.23941276115189158,
      "grad_norm": 0.9380141496658325,
      "learning_rate": 9.187473940789557e-05,
      "loss": 1.4093,
      "step": 106
    },
    {
      "epoch": 0.24167137210615472,
      "grad_norm": 0.7033969759941101,
      "learning_rate": 9.166569595412575e-05,
      "loss": 1.3279,
      "step": 107
    },
    {
      "epoch": 0.24392998306041785,
      "grad_norm": 0.9007672667503357,
      "learning_rate": 9.145424161705776e-05,
      "loss": 1.3571,
      "step": 108
    },
    {
      "epoch": 0.24618859401468096,
      "grad_norm": 0.555966854095459,
      "learning_rate": 9.124038863197818e-05,
      "loss": 1.475,
      "step": 109
    },
    {
      "epoch": 0.2484472049689441,
      "grad_norm": 0.5234211683273315,
      "learning_rate": 9.10241493729654e-05,
      "loss": 1.5519,
      "step": 110
    },
    {
      "epoch": 0.25070581592320723,
      "grad_norm": 0.5980601906776428,
      "learning_rate": 9.08055363521738e-05,
      "loss": 1.5623,
      "step": 111
    },
    {
      "epoch": 0.25296442687747034,
      "grad_norm": 0.6727840900421143,
      "learning_rate": 9.058456221910956e-05,
      "loss": 1.4898,
      "step": 112
    },
    {
      "epoch": 0.2552230378317335,
      "grad_norm": 0.6955089569091797,
      "learning_rate": 9.036123975989892e-05,
      "loss": 1.4752,
      "step": 113
    },
    {
      "epoch": 0.2574816487859966,
      "grad_norm": 0.6179929971694946,
      "learning_rate": 9.013558189654819e-05,
      "loss": 1.4823,
      "step": 114
    },
    {
      "epoch": 0.2597402597402597,
      "grad_norm": 0.6038258671760559,
      "learning_rate": 8.990760168619615e-05,
      "loss": 1.2696,
      "step": 115
    },
    {
      "epoch": 0.2619988706945229,
      "grad_norm": 0.6415106058120728,
      "learning_rate": 8.967731232035847e-05,
      "loss": 1.8538,
      "step": 116
    },
    {
      "epoch": 0.264257481648786,
      "grad_norm": 0.4856817424297333,
      "learning_rate": 8.944472712416447e-05,
      "loss": 1.4395,
      "step": 117
    },
    {
      "epoch": 0.26651609260304915,
      "grad_norm": 0.5568715929985046,
      "learning_rate": 8.9209859555586e-05,
      "loss": 1.2157,
      "step": 118
    },
    {
      "epoch": 0.26877470355731226,
      "grad_norm": 0.5347578525543213,
      "learning_rate": 8.897272320465887e-05,
      "loss": 1.4817,
      "step": 119
    },
    {
      "epoch": 0.27103331451157536,
      "grad_norm": 0.5750213861465454,
      "learning_rate": 8.873333179269635e-05,
      "loss": 1.477,
      "step": 120
    },
    {
      "epoch": 0.2732919254658385,
      "grad_norm": 0.588683009147644,
      "learning_rate": 8.849169917149531e-05,
      "loss": 1.6019,
      "step": 121
    },
    {
      "epoch": 0.27555053642010163,
      "grad_norm": 0.511512279510498,
      "learning_rate": 8.82478393225347e-05,
      "loss": 1.6251,
      "step": 122
    },
    {
      "epoch": 0.27780914737436474,
      "grad_norm": 0.6532254815101624,
      "learning_rate": 8.800176635616657e-05,
      "loss": 1.7525,
      "step": 123
    },
    {
      "epoch": 0.2800677583286279,
      "grad_norm": 0.6906523108482361,
      "learning_rate": 8.775349451079948e-05,
      "loss": 1.3276,
      "step": 124
    },
    {
      "epoch": 0.282326369282891,
      "grad_norm": 0.7139832973480225,
      "learning_rate": 8.750303815207486e-05,
      "loss": 1.6493,
      "step": 125
    },
    {
      "epoch": 0.2845849802371542,
      "grad_norm": 0.6658667325973511,
      "learning_rate": 8.725041177203554e-05,
      "loss": 1.2206,
      "step": 126
    },
    {
      "epoch": 0.2868435911914173,
      "grad_norm": 0.6301794648170471,
      "learning_rate": 8.699562998828738e-05,
      "loss": 1.4198,
      "step": 127
    },
    {
      "epoch": 0.2891022021456804,
      "grad_norm": 0.5198056697845459,
      "learning_rate": 8.673870754315336e-05,
      "loss": 1.3772,
      "step": 128
    },
    {
      "epoch": 0.29136081309994355,
      "grad_norm": 0.6238622069358826,
      "learning_rate": 8.647965930282059e-05,
      "loss": 1.4059,
      "step": 129
    },
    {
      "epoch": 0.29361942405420666,
      "grad_norm": 0.7875000834465027,
      "learning_rate": 8.621850025648009e-05,
      "loss": 1.2813,
      "step": 130
    },
    {
      "epoch": 0.29587803500846976,
      "grad_norm": 0.8984452486038208,
      "learning_rate": 8.59552455154595e-05,
      "loss": 1.7444,
      "step": 131
    },
    {
      "epoch": 0.2981366459627329,
      "grad_norm": 0.520828902721405,
      "learning_rate": 8.56899103123487e-05,
      "loss": 1.3697,
      "step": 132
    },
    {
      "epoch": 0.30039525691699603,
      "grad_norm": 0.5849810242652893,
      "learning_rate": 8.54225100001184e-05,
      "loss": 1.2721,
      "step": 133
    },
    {
      "epoch": 0.3026538678712592,
      "grad_norm": 0.5514599680900574,
      "learning_rate": 8.51530600512318e-05,
      "loss": 1.4966,
      "step": 134
    },
    {
      "epoch": 0.3049124788255223,
      "grad_norm": 1.624732255935669,
      "learning_rate": 8.488157605674925e-05,
      "loss": 1.2903,
      "step": 135
    },
    {
      "epoch": 0.3071710897797854,
      "grad_norm": 0.5830861926078796,
      "learning_rate": 8.460807372542618e-05,
      "loss": 1.2969,
      "step": 136
    },
    {
      "epoch": 0.3094297007340486,
      "grad_norm": 0.8161193132400513,
      "learning_rate": 8.43325688828042e-05,
      "loss": 1.1874,
      "step": 137
    },
    {
      "epoch": 0.3116883116883117,
      "grad_norm": 0.47100022435188293,
      "learning_rate": 8.405507747029523e-05,
      "loss": 1.3229,
      "step": 138
    },
    {
      "epoch": 0.31394692264257484,
      "grad_norm": 0.4437452554702759,
      "learning_rate": 8.377561554425922e-05,
      "loss": 1.1945,
      "step": 139
    },
    {
      "epoch": 0.31620553359683795,
      "grad_norm": 0.566750705242157,
      "learning_rate": 8.349419927507505e-05,
      "loss": 1.4707,
      "step": 140
    },
    {
      "epoch": 0.31846414455110106,
      "grad_norm": 0.5701163411140442,
      "learning_rate": 8.321084494620488e-05,
      "loss": 1.5681,
      "step": 141
    },
    {
      "epoch": 0.3207227555053642,
      "grad_norm": 0.5442900657653809,
      "learning_rate": 8.292556895325194e-05,
      "loss": 1.2291,
      "step": 142
    },
    {
      "epoch": 0.32298136645962733,
      "grad_norm": 0.5684903860092163,
      "learning_rate": 8.263838780301182e-05,
      "loss": 1.3748,
      "step": 143
    },
    {
      "epoch": 0.32523997741389044,
      "grad_norm": 0.7141373157501221,
      "learning_rate": 8.234931811251739e-05,
      "loss": 1.1836,
      "step": 144
    },
    {
      "epoch": 0.3274985883681536,
      "grad_norm": 0.7630109190940857,
      "learning_rate": 8.205837660807725e-05,
      "loss": 1.602,
      "step": 145
    },
    {
      "epoch": 0.3297571993224167,
      "grad_norm": 0.6189699172973633,
      "learning_rate": 8.176558012430791e-05,
      "loss": 1.273,
      "step": 146
    },
    {
      "epoch": 0.33201581027667987,
      "grad_norm": 0.7228075861930847,
      "learning_rate": 8.147094560315977e-05,
      "loss": 1.1563,
      "step": 147
    },
    {
      "epoch": 0.334274421230943,
      "grad_norm": 0.5667940378189087,
      "learning_rate": 8.117449009293668e-05,
      "loss": 1.3525,
      "step": 148
    },
    {
      "epoch": 0.3365330321852061,
      "grad_norm": 0.4601518213748932,
      "learning_rate": 8.08762307473096e-05,
      "loss": 1.2889,
      "step": 149
    },
    {
      "epoch": 0.33879164313946925,
      "grad_norm": 0.5403670072555542,
      "learning_rate": 8.057618482432399e-05,
      "loss": 1.4927,
      "step": 150
    },
    {
      "epoch": 0.34105025409373235,
      "grad_norm": 0.5279664993286133,
      "learning_rate": 8.027436968540123e-05,
      "loss": 1.4293,
      "step": 151
    },
    {
      "epoch": 0.34330886504799546,
      "grad_norm": 0.517234206199646,
      "learning_rate": 7.997080279433402e-05,
      "loss": 1.2646,
      "step": 152
    },
    {
      "epoch": 0.3455674760022586,
      "grad_norm": 0.6740924119949341,
      "learning_rate": 7.966550171627592e-05,
      "loss": 1.3801,
      "step": 153
    },
    {
      "epoch": 0.34782608695652173,
      "grad_norm": 0.47068697214126587,
      "learning_rate": 7.9358484116725e-05,
      "loss": 1.3094,
      "step": 154
    },
    {
      "epoch": 0.3500846979107849,
      "grad_norm": 0.5425541996955872,
      "learning_rate": 7.904976776050156e-05,
      "loss": 1.5867,
      "step": 155
    },
    {
      "epoch": 0.352343308865048,
      "grad_norm": 0.41730475425720215,
      "learning_rate": 7.873937051072035e-05,
      "loss": 1.1469,
      "step": 156
    },
    {
      "epoch": 0.3546019198193111,
      "grad_norm": 0.632966160774231,
      "learning_rate": 7.842731032775687e-05,
      "loss": 1.4293,
      "step": 157
    },
    {
      "epoch": 0.35686053077357427,
      "grad_norm": 0.6313096880912781,
      "learning_rate": 7.81136052682082e-05,
      "loss": 1.2995,
      "step": 158
    },
    {
      "epoch": 0.3591191417278374,
      "grad_norm": 1.015628457069397,
      "learning_rate": 7.779827348384813e-05,
      "loss": 1.6397,
      "step": 159
    },
    {
      "epoch": 0.3613777526821005,
      "grad_norm": 0.6771583557128906,
      "learning_rate": 7.748133322057693e-05,
      "loss": 1.5359,
      "step": 160
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 0.6180778741836548,
      "learning_rate": 7.716280281736551e-05,
      "loss": 1.6158,
      "step": 161
    },
    {
      "epoch": 0.36589497459062675,
      "grad_norm": 0.46877238154411316,
      "learning_rate": 7.68427007051944e-05,
      "loss": 1.3506,
      "step": 162
    },
    {
      "epoch": 0.3681535855448899,
      "grad_norm": 0.5611956119537354,
      "learning_rate": 7.652104540598712e-05,
      "loss": 1.2563,
      "step": 163
    },
    {
      "epoch": 0.370412196499153,
      "grad_norm": 0.5148677229881287,
      "learning_rate": 7.619785553153864e-05,
      "loss": 1.4294,
      "step": 164
    },
    {
      "epoch": 0.37267080745341613,
      "grad_norm": 0.49939846992492676,
      "learning_rate": 7.58731497824383e-05,
      "loss": 1.2108,
      "step": 165
    },
    {
      "epoch": 0.3749294184076793,
      "grad_norm": 0.6522055864334106,
      "learning_rate": 7.554694694698784e-05,
      "loss": 1.3305,
      "step": 166
    },
    {
      "epoch": 0.3771880293619424,
      "grad_norm": 0.5979933738708496,
      "learning_rate": 7.521926590011418e-05,
      "loss": 1.4971,
      "step": 167
    },
    {
      "epoch": 0.3794466403162055,
      "grad_norm": 2.253812551498413,
      "learning_rate": 7.489012560227742e-05,
      "loss": 1.1929,
      "step": 168
    },
    {
      "epoch": 0.38170525127046867,
      "grad_norm": 0.5582684874534607,
      "learning_rate": 7.455954509837352e-05,
      "loss": 1.4669,
      "step": 169
    },
    {
      "epoch": 0.3839638622247318,
      "grad_norm": 0.5619791150093079,
      "learning_rate": 7.422754351663252e-05,
      "loss": 1.4651,
      "step": 170
    },
    {
      "epoch": 0.38622247317899494,
      "grad_norm": 0.4569830000400543,
      "learning_rate": 7.389414006751158e-05,
      "loss": 1.4677,
      "step": 171
    },
    {
      "epoch": 0.38848108413325805,
      "grad_norm": 1.0484251976013184,
      "learning_rate": 7.355935404258354e-05,
      "loss": 1.4736,
      "step": 172
    },
    {
      "epoch": 0.39073969508752115,
      "grad_norm": 0.5541074872016907,
      "learning_rate": 7.322320481342054e-05,
      "loss": 1.4604,
      "step": 173
    },
    {
      "epoch": 0.3929983060417843,
      "grad_norm": 0.6257642507553101,
      "learning_rate": 7.288571183047322e-05,
      "loss": 1.1331,
      "step": 174
    },
    {
      "epoch": 0.3952569169960474,
      "grad_norm": 0.5899243354797363,
      "learning_rate": 7.254689462194522e-05,
      "loss": 1.1387,
      "step": 175
    },
    {
      "epoch": 0.39751552795031053,
      "grad_norm": 0.5541165471076965,
      "learning_rate": 7.220677279266327e-05,
      "loss": 1.408,
      "step": 176
    },
    {
      "epoch": 0.3997741389045737,
      "grad_norm": 0.5860551595687866,
      "learning_rate": 7.186536602294278e-05,
      "loss": 1.512,
      "step": 177
    },
    {
      "epoch": 0.4020327498588368,
      "grad_norm": 0.5203275084495544,
      "learning_rate": 7.152269406744903e-05,
      "loss": 1.8094,
      "step": 178
    },
    {
      "epoch": 0.40429136081309996,
      "grad_norm": 0.6692151427268982,
      "learning_rate": 7.117877675405427e-05,
      "loss": 1.4363,
      "step": 179
    },
    {
      "epoch": 0.40654997176736307,
      "grad_norm": 0.624856173992157,
      "learning_rate": 7.083363398269022e-05,
      "loss": 1.4632,
      "step": 180
    },
    {
      "epoch": 0.4088085827216262,
      "grad_norm": 0.5100191235542297,
      "learning_rate": 7.04872857241968e-05,
      "loss": 1.3967,
      "step": 181
    },
    {
      "epoch": 0.41106719367588934,
      "grad_norm": 0.5183005332946777,
      "learning_rate": 7.013975201916648e-05,
      "loss": 1.7088,
      "step": 182
    },
    {
      "epoch": 0.41332580463015245,
      "grad_norm": 0.7876030802726746,
      "learning_rate": 6.979105297678462e-05,
      "loss": 1.4677,
      "step": 183
    },
    {
      "epoch": 0.4155844155844156,
      "grad_norm": 0.5176962614059448,
      "learning_rate": 6.944120877366604e-05,
      "loss": 1.5604,
      "step": 184
    },
    {
      "epoch": 0.4178430265386787,
      "grad_norm": 0.6162248849868774,
      "learning_rate": 6.909023965268746e-05,
      "loss": 1.307,
      "step": 185
    },
    {
      "epoch": 0.4201016374929418,
      "grad_norm": 0.8558477759361267,
      "learning_rate": 6.873816592181617e-05,
      "loss": 1.1848,
      "step": 186
    },
    {
      "epoch": 0.422360248447205,
      "grad_norm": 0.5814364552497864,
      "learning_rate": 6.838500795293505e-05,
      "loss": 1.3717,
      "step": 187
    },
    {
      "epoch": 0.4246188594014681,
      "grad_norm": 0.5346269011497498,
      "learning_rate": 6.803078618066378e-05,
      "loss": 1.444,
      "step": 188
    },
    {
      "epoch": 0.4268774703557312,
      "grad_norm": 0.5764583945274353,
      "learning_rate": 6.767552110117631e-05,
      "loss": 1.4341,
      "step": 189
    },
    {
      "epoch": 0.42913608130999437,
      "grad_norm": 0.4884260892868042,
      "learning_rate": 6.73192332710151e-05,
      "loss": 1.3498,
      "step": 190
    },
    {
      "epoch": 0.4313946922642575,
      "grad_norm": 0.6064755320549011,
      "learning_rate": 6.696194330590151e-05,
      "loss": 1.6995,
      "step": 191
    },
    {
      "epoch": 0.43365330321852064,
      "grad_norm": 0.5664028525352478,
      "learning_rate": 6.660367187954304e-05,
      "loss": 1.418,
      "step": 192
    },
    {
      "epoch": 0.43591191417278374,
      "grad_norm": 0.6558746695518494,
      "learning_rate": 6.624443972243698e-05,
      "loss": 1.2759,
      "step": 193
    },
    {
      "epoch": 0.43817052512704685,
      "grad_norm": 1.469589352607727,
      "learning_rate": 6.5884267620671e-05,
      "loss": 1.2259,
      "step": 194
    },
    {
      "epoch": 0.44042913608131,
      "grad_norm": 0.9882861375808716,
      "learning_rate": 6.552317641472026e-05,
      "loss": 1.4997,
      "step": 195
    },
    {
      "epoch": 0.4426877470355731,
      "grad_norm": 0.4861055910587311,
      "learning_rate": 6.516118699824178e-05,
      "loss": 1.2735,
      "step": 196
    },
    {
      "epoch": 0.4449463579898362,
      "grad_norm": 0.5669353008270264,
      "learning_rate": 6.479832031686521e-05,
      "loss": 1.4849,
      "step": 197
    },
    {
      "epoch": 0.4472049689440994,
      "grad_norm": 0.5639188885688782,
      "learning_rate": 6.443459736698105e-05,
      "loss": 1.6053,
      "step": 198
    },
    {
      "epoch": 0.4494635798983625,
      "grad_norm": 0.5520173907279968,
      "learning_rate": 6.407003919452564e-05,
      "loss": 1.2882,
      "step": 199
    },
    {
      "epoch": 0.45172219085262566,
      "grad_norm": 0.7025582790374756,
      "learning_rate": 6.370466689376342e-05,
      "loss": 1.4892,
      "step": 200
    },
    {
      "epoch": 0.45398080180688877,
      "grad_norm": 0.5788022875785828,
      "learning_rate": 6.33385016060664e-05,
      "loss": 1.7469,
      "step": 201
    },
    {
      "epoch": 0.4562394127611519,
      "grad_norm": 0.5949933528900146,
      "learning_rate": 6.297156451869082e-05,
      "loss": 1.405,
      "step": 202
    },
    {
      "epoch": 0.45849802371541504,
      "grad_norm": 0.5209388732910156,
      "learning_rate": 6.260387686355121e-05,
      "loss": 1.3265,
      "step": 203
    },
    {
      "epoch": 0.46075663466967814,
      "grad_norm": 0.4528113603591919,
      "learning_rate": 6.223545991599184e-05,
      "loss": 1.4738,
      "step": 204
    },
    {
      "epoch": 0.46301524562394125,
      "grad_norm": 0.5056377053260803,
      "learning_rate": 6.186633499355576e-05,
      "loss": 1.6497,
      "step": 205
    },
    {
      "epoch": 0.4652738565782044,
      "grad_norm": 0.5312528014183044,
      "learning_rate": 6.149652345475118e-05,
      "loss": 1.2971,
      "step": 206
    },
    {
      "epoch": 0.4675324675324675,
      "grad_norm": 0.8774858117103577,
      "learning_rate": 6.112604669781572e-05,
      "loss": 1.4465,
      "step": 207
    },
    {
      "epoch": 0.4697910784867307,
      "grad_norm": 0.6204832196235657,
      "learning_rate": 6.075492615947823e-05,
      "loss": 1.3416,
      "step": 208
    },
    {
      "epoch": 0.4720496894409938,
      "grad_norm": 0.4414108097553253,
      "learning_rate": 6.038318331371836e-05,
      "loss": 1.5334,
      "step": 209
    },
    {
      "epoch": 0.4743083003952569,
      "grad_norm": 0.6953609585762024,
      "learning_rate": 6.001083967052408e-05,
      "loss": 1.6932,
      "step": 210
    },
    {
      "epoch": 0.47656691134952006,
      "grad_norm": 0.47844645380973816,
      "learning_rate": 5.963791677464696e-05,
      "loss": 1.7326,
      "step": 211
    },
    {
      "epoch": 0.47882552230378317,
      "grad_norm": 0.5514799952507019,
      "learning_rate": 5.9264436204355724e-05,
      "loss": 1.2659,
      "step": 212
    },
    {
      "epoch": 0.4810841332580463,
      "grad_norm": 0.7047640681266785,
      "learning_rate": 5.889041957018745e-05,
      "loss": 1.2528,
      "step": 213
    },
    {
      "epoch": 0.48334274421230944,
      "grad_norm": 0.5648500919342041,
      "learning_rate": 5.85158885136973e-05,
      "loss": 1.1812,
      "step": 214
    },
    {
      "epoch": 0.48560135516657255,
      "grad_norm": 0.6360241174697876,
      "learning_rate": 5.81408647062062e-05,
      "loss": 1.464,
      "step": 215
    },
    {
      "epoch": 0.4878599661208357,
      "grad_norm": 0.693928599357605,
      "learning_rate": 5.7765369847546916e-05,
      "loss": 1.5254,
      "step": 216
    },
    {
      "epoch": 0.4901185770750988,
      "grad_norm": 0.8636585474014282,
      "learning_rate": 5.7389425664808396e-05,
      "loss": 1.7108,
      "step": 217
    },
    {
      "epoch": 0.4923771880293619,
      "grad_norm": 0.5646620392799377,
      "learning_rate": 5.7013053911078677e-05,
      "loss": 1.4662,
      "step": 218
    },
    {
      "epoch": 0.4946357989836251,
      "grad_norm": 0.48793742060661316,
      "learning_rate": 5.6636276364186105e-05,
      "loss": 1.475,
      "step": 219
    },
    {
      "epoch": 0.4968944099378882,
      "grad_norm": 0.4991232454776764,
      "learning_rate": 5.6259114825439275e-05,
      "loss": 1.5324,
      "step": 220
    },
    {
      "epoch": 0.4991530208921513,
      "grad_norm": 0.6191245317459106,
      "learning_rate": 5.588159111836553e-05,
      "loss": 1.3583,
      "step": 221
    },
    {
      "epoch": 0.5014116318464145,
      "grad_norm": 0.5580266714096069,
      "learning_rate": 5.550372708744815e-05,
      "loss": 1.4188,
      "step": 222
    },
    {
      "epoch": 0.5014116318464145,
      "eval_loss": 1.4097435474395752,
      "eval_runtime": 98.3925,
      "eval_samples_per_second": 7.582,
      "eval_steps_per_second": 0.955,
      "step": 222
    },
    {
      "epoch": 0.5036702428006776,
      "grad_norm": 0.4826272130012512,
      "learning_rate": 5.51255445968625e-05,
      "loss": 1.4089,
      "step": 223
    },
    {
      "epoch": 0.5059288537549407,
      "grad_norm": 0.7236170768737793,
      "learning_rate": 5.4747065529210736e-05,
      "loss": 1.3453,
      "step": 224
    },
    {
      "epoch": 0.5081874647092038,
      "grad_norm": 0.5844079256057739,
      "learning_rate": 5.436831178425582e-05,
      "loss": 1.3397,
      "step": 225
    },
    {
      "epoch": 0.510446075663467,
      "grad_norm": 0.8065000176429749,
      "learning_rate": 5.3989305277654156e-05,
      "loss": 1.4753,
      "step": 226
    },
    {
      "epoch": 0.51270468661773,
      "grad_norm": 0.48151111602783203,
      "learning_rate": 5.361006793968764e-05,
      "loss": 1.2529,
      "step": 227
    },
    {
      "epoch": 0.5149632975719932,
      "grad_norm": 0.6532959342002869,
      "learning_rate": 5.32306217139946e-05,
      "loss": 1.4809,
      "step": 228
    },
    {
      "epoch": 0.5172219085262564,
      "grad_norm": 0.6014647483825684,
      "learning_rate": 5.28509885563002e-05,
      "loss": 1.2814,
      "step": 229
    },
    {
      "epoch": 0.5194805194805194,
      "grad_norm": 0.6728442311286926,
      "learning_rate": 5.247119043314592e-05,
      "loss": 1.4694,
      "step": 230
    },
    {
      "epoch": 0.5217391304347826,
      "grad_norm": 0.5741045475006104,
      "learning_rate": 5.209124932061862e-05,
      "loss": 1.6012,
      "step": 231
    },
    {
      "epoch": 0.5239977413890458,
      "grad_norm": 0.49296098947525024,
      "learning_rate": 5.1711187203078824e-05,
      "loss": 1.4489,
      "step": 232
    },
    {
      "epoch": 0.5262563523433089,
      "grad_norm": 0.6149709820747375,
      "learning_rate": 5.133102607188874e-05,
      "loss": 1.4421,
      "step": 233
    },
    {
      "epoch": 0.528514963297572,
      "grad_norm": 0.6720355749130249,
      "learning_rate": 5.0950787924139764e-05,
      "loss": 1.2675,
      "step": 234
    },
    {
      "epoch": 0.5307735742518351,
      "grad_norm": 0.5325198769569397,
      "learning_rate": 5.057049476137967e-05,
      "loss": 1.1803,
      "step": 235
    },
    {
      "epoch": 0.5330321852060983,
      "grad_norm": 0.5668638944625854,
      "learning_rate": 5.0190168588339536e-05,
      "loss": 1.2712,
      "step": 236
    },
    {
      "epoch": 0.5352907961603613,
      "grad_norm": 0.4216601848602295,
      "learning_rate": 4.9809831411660476e-05,
      "loss": 1.322,
      "step": 237
    },
    {
      "epoch": 0.5375494071146245,
      "grad_norm": 0.608249306678772,
      "learning_rate": 4.942950523862033e-05,
      "loss": 1.3996,
      "step": 238
    },
    {
      "epoch": 0.5398080180688877,
      "grad_norm": 0.9624855518341064,
      "learning_rate": 4.904921207586024e-05,
      "loss": 1.3592,
      "step": 239
    },
    {
      "epoch": 0.5420666290231507,
      "grad_norm": 0.5141565203666687,
      "learning_rate": 4.866897392811126e-05,
      "loss": 1.1777,
      "step": 240
    },
    {
      "epoch": 0.5443252399774139,
      "grad_norm": 0.6783860325813293,
      "learning_rate": 4.828881279692119e-05,
      "loss": 1.6956,
      "step": 241
    },
    {
      "epoch": 0.546583850931677,
      "grad_norm": 0.48693007230758667,
      "learning_rate": 4.7908750679381384e-05,
      "loss": 1.6161,
      "step": 242
    },
    {
      "epoch": 0.5488424618859401,
      "grad_norm": 0.5655871033668518,
      "learning_rate": 4.752880956685407e-05,
      "loss": 1.0778,
      "step": 243
    },
    {
      "epoch": 0.5511010728402033,
      "grad_norm": 0.5297839045524597,
      "learning_rate": 4.7149011443699814e-05,
      "loss": 1.485,
      "step": 244
    },
    {
      "epoch": 0.5533596837944664,
      "grad_norm": 0.7380653023719788,
      "learning_rate": 4.676937828600542e-05,
      "loss": 1.4226,
      "step": 245
    },
    {
      "epoch": 0.5556182947487295,
      "grad_norm": 0.693130373954773,
      "learning_rate": 4.638993206031237e-05,
      "loss": 1.4093,
      "step": 246
    },
    {
      "epoch": 0.5578769057029926,
      "grad_norm": 0.6071158647537231,
      "learning_rate": 4.601069472234584e-05,
      "loss": 1.0761,
      "step": 247
    },
    {
      "epoch": 0.5601355166572558,
      "grad_norm": 0.9863161444664001,
      "learning_rate": 4.56316882157442e-05,
      "loss": 1.3143,
      "step": 248
    },
    {
      "epoch": 0.562394127611519,
      "grad_norm": 0.5352652668952942,
      "learning_rate": 4.525293447078927e-05,
      "loss": 1.7061,
      "step": 249
    },
    {
      "epoch": 0.564652738565782,
      "grad_norm": 0.558076798915863,
      "learning_rate": 4.4874455403137514e-05,
      "loss": 1.7395,
      "step": 250
    },
    {
      "epoch": 0.5669113495200452,
      "grad_norm": 1.1520057916641235,
      "learning_rate": 4.449627291255184e-05,
      "loss": 1.6384,
      "step": 251
    },
    {
      "epoch": 0.5691699604743083,
      "grad_norm": 0.531182587146759,
      "learning_rate": 4.411840888163449e-05,
      "loss": 1.3033,
      "step": 252
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.4696061313152313,
      "learning_rate": 4.3740885174560736e-05,
      "loss": 1.4949,
      "step": 253
    },
    {
      "epoch": 0.5736871823828346,
      "grad_norm": 0.6336054801940918,
      "learning_rate": 4.336372363581391e-05,
      "loss": 1.4374,
      "step": 254
    },
    {
      "epoch": 0.5759457933370977,
      "grad_norm": 0.6364861726760864,
      "learning_rate": 4.298694608892134e-05,
      "loss": 1.7703,
      "step": 255
    },
    {
      "epoch": 0.5782044042913608,
      "grad_norm": 1.0777122974395752,
      "learning_rate": 4.2610574335191615e-05,
      "loss": 1.4255,
      "step": 256
    },
    {
      "epoch": 0.5804630152456239,
      "grad_norm": 0.8023219704627991,
      "learning_rate": 4.2234630152453116e-05,
      "loss": 1.384,
      "step": 257
    },
    {
      "epoch": 0.5827216261998871,
      "grad_norm": 0.5222808718681335,
      "learning_rate": 4.185913529379381e-05,
      "loss": 1.1682,
      "step": 258
    },
    {
      "epoch": 0.5849802371541502,
      "grad_norm": 0.7348831295967102,
      "learning_rate": 4.1484111486302704e-05,
      "loss": 1.2724,
      "step": 259
    },
    {
      "epoch": 0.5872388481084133,
      "grad_norm": 0.5531060695648193,
      "learning_rate": 4.110958042981255e-05,
      "loss": 1.4645,
      "step": 260
    },
    {
      "epoch": 0.5894974590626765,
      "grad_norm": 0.6494237780570984,
      "learning_rate": 4.0735563795644294e-05,
      "loss": 1.5516,
      "step": 261
    },
    {
      "epoch": 0.5917560700169395,
      "grad_norm": 0.5829164981842041,
      "learning_rate": 4.0362083225353046e-05,
      "loss": 1.4303,
      "step": 262
    },
    {
      "epoch": 0.5940146809712027,
      "grad_norm": 0.4754260182380676,
      "learning_rate": 3.998916032947594e-05,
      "loss": 1.3245,
      "step": 263
    },
    {
      "epoch": 0.5962732919254659,
      "grad_norm": 0.7233774065971375,
      "learning_rate": 3.961681668628164e-05,
      "loss": 1.272,
      "step": 264
    },
    {
      "epoch": 0.598531902879729,
      "grad_norm": 0.55941241979599,
      "learning_rate": 3.9245073840521765e-05,
      "loss": 1.4211,
      "step": 265
    },
    {
      "epoch": 0.6007905138339921,
      "grad_norm": 0.5475345253944397,
      "learning_rate": 3.887395330218429e-05,
      "loss": 1.3842,
      "step": 266
    },
    {
      "epoch": 0.6030491247882552,
      "grad_norm": 0.7882475256919861,
      "learning_rate": 3.850347654524883e-05,
      "loss": 1.4087,
      "step": 267
    },
    {
      "epoch": 0.6053077357425184,
      "grad_norm": 0.6707928776741028,
      "learning_rate": 3.8133665006444255e-05,
      "loss": 1.2446,
      "step": 268
    },
    {
      "epoch": 0.6075663466967814,
      "grad_norm": 0.6627002358436584,
      "learning_rate": 3.776454008400816e-05,
      "loss": 1.7243,
      "step": 269
    },
    {
      "epoch": 0.6098249576510446,
      "grad_norm": 0.476368248462677,
      "learning_rate": 3.7396123136448824e-05,
      "loss": 1.5209,
      "step": 270
    },
    {
      "epoch": 0.6120835686053078,
      "grad_norm": 0.6583078503608704,
      "learning_rate": 3.70284354813092e-05,
      "loss": 1.3496,
      "step": 271
    },
    {
      "epoch": 0.6143421795595708,
      "grad_norm": 0.6484099626541138,
      "learning_rate": 3.666149839393361e-05,
      "loss": 1.6792,
      "step": 272
    },
    {
      "epoch": 0.616600790513834,
      "grad_norm": 0.42976608872413635,
      "learning_rate": 3.629533310623658e-05,
      "loss": 1.0712,
      "step": 273
    },
    {
      "epoch": 0.6188594014680971,
      "grad_norm": 0.7299443483352661,
      "learning_rate": 3.592996080547438e-05,
      "loss": 1.7172,
      "step": 274
    },
    {
      "epoch": 0.6211180124223602,
      "grad_norm": 0.6105283498764038,
      "learning_rate": 3.556540263301896e-05,
      "loss": 1.5218,
      "step": 275
    },
    {
      "epoch": 0.6233766233766234,
      "grad_norm": 0.6349045038223267,
      "learning_rate": 3.520167968313479e-05,
      "loss": 1.4029,
      "step": 276
    },
    {
      "epoch": 0.6256352343308865,
      "grad_norm": 0.441521555185318,
      "learning_rate": 3.483881300175823e-05,
      "loss": 1.5902,
      "step": 277
    },
    {
      "epoch": 0.6278938452851497,
      "grad_norm": 0.6537282466888428,
      "learning_rate": 3.447682358527974e-05,
      "loss": 1.2511,
      "step": 278
    },
    {
      "epoch": 0.6301524562394127,
      "grad_norm": 0.64874267578125,
      "learning_rate": 3.411573237932904e-05,
      "loss": 1.4145,
      "step": 279
    },
    {
      "epoch": 0.6324110671936759,
      "grad_norm": 0.6560806035995483,
      "learning_rate": 3.3755560277563023e-05,
      "loss": 1.3575,
      "step": 280
    },
    {
      "epoch": 0.6346696781479391,
      "grad_norm": 0.4486541152000427,
      "learning_rate": 3.339632812045696e-05,
      "loss": 1.4707,
      "step": 281
    },
    {
      "epoch": 0.6369282891022021,
      "grad_norm": 0.5472518801689148,
      "learning_rate": 3.303805669409848e-05,
      "loss": 1.2591,
      "step": 282
    },
    {
      "epoch": 0.6391869000564653,
      "grad_norm": 0.8374768495559692,
      "learning_rate": 3.268076672898492e-05,
      "loss": 1.2883,
      "step": 283
    },
    {
      "epoch": 0.6414455110107284,
      "grad_norm": 0.4761299788951874,
      "learning_rate": 3.2324478898823705e-05,
      "loss": 1.46,
      "step": 284
    },
    {
      "epoch": 0.6437041219649915,
      "grad_norm": 0.58656907081604,
      "learning_rate": 3.196921381933624e-05,
      "loss": 1.366,
      "step": 285
    },
    {
      "epoch": 0.6459627329192547,
      "grad_norm": 0.6311814785003662,
      "learning_rate": 3.1614992047064945e-05,
      "loss": 1.2755,
      "step": 286
    },
    {
      "epoch": 0.6482213438735178,
      "grad_norm": 0.5806014537811279,
      "learning_rate": 3.126183407818384e-05,
      "loss": 1.3936,
      "step": 287
    },
    {
      "epoch": 0.6504799548277809,
      "grad_norm": 0.7558007836341858,
      "learning_rate": 3.090976034731257e-05,
      "loss": 1.3133,
      "step": 288
    },
    {
      "epoch": 0.652738565782044,
      "grad_norm": 0.6248241662979126,
      "learning_rate": 3.055879122633397e-05,
      "loss": 1.3684,
      "step": 289
    },
    {
      "epoch": 0.6549971767363072,
      "grad_norm": 0.6883110404014587,
      "learning_rate": 3.020894702321539e-05,
      "loss": 1.5355,
      "step": 290
    },
    {
      "epoch": 0.6572557876905702,
      "grad_norm": 0.47955361008644104,
      "learning_rate": 2.9860247980833532e-05,
      "loss": 1.2752,
      "step": 291
    },
    {
      "epoch": 0.6595143986448334,
      "grad_norm": 0.6376725435256958,
      "learning_rate": 2.951271427580321e-05,
      "loss": 1.3576,
      "step": 292
    },
    {
      "epoch": 0.6617730095990966,
      "grad_norm": 0.6355816125869751,
      "learning_rate": 2.91663660173098e-05,
      "loss": 1.3411,
      "step": 293
    },
    {
      "epoch": 0.6640316205533597,
      "grad_norm": 0.6872756481170654,
      "learning_rate": 2.882122324594575e-05,
      "loss": 1.5319,
      "step": 294
    },
    {
      "epoch": 0.6662902315076228,
      "grad_norm": 1.0146925449371338,
      "learning_rate": 2.847730593255097e-05,
      "loss": 1.4574,
      "step": 295
    },
    {
      "epoch": 0.668548842461886,
      "grad_norm": 0.46255797147750854,
      "learning_rate": 2.8134633977057235e-05,
      "loss": 1.6074,
      "step": 296
    },
    {
      "epoch": 0.6708074534161491,
      "grad_norm": 0.5414189100265503,
      "learning_rate": 2.779322720733673e-05,
      "loss": 1.6655,
      "step": 297
    },
    {
      "epoch": 0.6730660643704122,
      "grad_norm": 0.4797675609588623,
      "learning_rate": 2.745310537805479e-05,
      "loss": 1.3457,
      "step": 298
    },
    {
      "epoch": 0.6753246753246753,
      "grad_norm": 0.5770863890647888,
      "learning_rate": 2.7114288169526793e-05,
      "loss": 1.5853,
      "step": 299
    },
    {
      "epoch": 0.6775832862789385,
      "grad_norm": 0.4539095163345337,
      "learning_rate": 2.6776795186579468e-05,
      "loss": 1.2473,
      "step": 300
    },
    {
      "epoch": 0.6798418972332015,
      "grad_norm": 0.5590886473655701,
      "learning_rate": 2.6440645957416484e-05,
      "loss": 1.7489,
      "step": 301
    },
    {
      "epoch": 0.6821005081874647,
      "grad_norm": 0.5184858441352844,
      "learning_rate": 2.610585993248843e-05,
      "loss": 1.4334,
      "step": 302
    },
    {
      "epoch": 0.6843591191417279,
      "grad_norm": 0.5701838135719299,
      "learning_rate": 2.5772456483367497e-05,
      "loss": 1.2721,
      "step": 303
    },
    {
      "epoch": 0.6866177300959909,
      "grad_norm": 0.5292041301727295,
      "learning_rate": 2.5440454901626486e-05,
      "loss": 1.4493,
      "step": 304
    },
    {
      "epoch": 0.6888763410502541,
      "grad_norm": 0.6524962782859802,
      "learning_rate": 2.510987439772261e-05,
      "loss": 1.1314,
      "step": 305
    },
    {
      "epoch": 0.6911349520045172,
      "grad_norm": 0.5966671109199524,
      "learning_rate": 2.4780734099885833e-05,
      "loss": 1.9383,
      "step": 306
    },
    {
      "epoch": 0.6933935629587803,
      "grad_norm": 0.506033718585968,
      "learning_rate": 2.4453053053012187e-05,
      "loss": 1.4983,
      "step": 307
    },
    {
      "epoch": 0.6956521739130435,
      "grad_norm": 0.687870979309082,
      "learning_rate": 2.4126850217561698e-05,
      "loss": 1.5229,
      "step": 308
    },
    {
      "epoch": 0.6979107848673066,
      "grad_norm": 0.4640161991119385,
      "learning_rate": 2.3802144468461367e-05,
      "loss": 1.5815,
      "step": 309
    },
    {
      "epoch": 0.7001693958215698,
      "grad_norm": 0.9109538793563843,
      "learning_rate": 2.347895459401288e-05,
      "loss": 1.419,
      "step": 310
    },
    {
      "epoch": 0.7024280067758328,
      "grad_norm": 0.5041788220405579,
      "learning_rate": 2.3157299294805613e-05,
      "loss": 1.3753,
      "step": 311
    },
    {
      "epoch": 0.704686617730096,
      "grad_norm": 0.6865159869194031,
      "learning_rate": 2.2837197182634483e-05,
      "loss": 1.6305,
      "step": 312
    },
    {
      "epoch": 0.7069452286843592,
      "grad_norm": 0.7781485319137573,
      "learning_rate": 2.2518666779423074e-05,
      "loss": 1.4607,
      "step": 313
    },
    {
      "epoch": 0.7092038396386222,
      "grad_norm": 0.5647329092025757,
      "learning_rate": 2.2201726516151882e-05,
      "loss": 1.4964,
      "step": 314
    },
    {
      "epoch": 0.7114624505928854,
      "grad_norm": 0.577785074710846,
      "learning_rate": 2.1886394731791816e-05,
      "loss": 1.5494,
      "step": 315
    },
    {
      "epoch": 0.7137210615471485,
      "grad_norm": 0.5110263228416443,
      "learning_rate": 2.157268967224314e-05,
      "loss": 1.4089,
      "step": 316
    },
    {
      "epoch": 0.7159796725014116,
      "grad_norm": 0.5485273003578186,
      "learning_rate": 2.126062948927966e-05,
      "loss": 1.5507,
      "step": 317
    },
    {
      "epoch": 0.7182382834556748,
      "grad_norm": 0.6336625814437866,
      "learning_rate": 2.0950232239498446e-05,
      "loss": 1.1833,
      "step": 318
    },
    {
      "epoch": 0.7204968944099379,
      "grad_norm": 0.5420766472816467,
      "learning_rate": 2.064151588327501e-05,
      "loss": 1.5497,
      "step": 319
    },
    {
      "epoch": 0.722755505364201,
      "grad_norm": 0.43920066952705383,
      "learning_rate": 2.0334498283724078e-05,
      "loss": 1.7342,
      "step": 320
    },
    {
      "epoch": 0.7250141163184641,
      "grad_norm": 0.4995698928833008,
      "learning_rate": 2.002919720566599e-05,
      "loss": 1.2779,
      "step": 321
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 1.1329647302627563,
      "learning_rate": 1.9725630314598782e-05,
      "loss": 1.3353,
      "step": 322
    },
    {
      "epoch": 0.7295313382269905,
      "grad_norm": 0.6960721015930176,
      "learning_rate": 1.9423815175676025e-05,
      "loss": 1.4193,
      "step": 323
    },
    {
      "epoch": 0.7317899491812535,
      "grad_norm": 0.5718068480491638,
      "learning_rate": 1.912376925269041e-05,
      "loss": 1.2262,
      "step": 324
    },
    {
      "epoch": 0.7340485601355167,
      "grad_norm": 0.633811891078949,
      "learning_rate": 1.8825509907063327e-05,
      "loss": 1.5999,
      "step": 325
    },
    {
      "epoch": 0.7363071710897798,
      "grad_norm": 0.5763436555862427,
      "learning_rate": 1.8529054396840234e-05,
      "loss": 1.5864,
      "step": 326
    },
    {
      "epoch": 0.7385657820440429,
      "grad_norm": 0.49580734968185425,
      "learning_rate": 1.8234419875692105e-05,
      "loss": 1.9146,
      "step": 327
    },
    {
      "epoch": 0.740824392998306,
      "grad_norm": 0.597285270690918,
      "learning_rate": 1.7941623391922772e-05,
      "loss": 1.2687,
      "step": 328
    },
    {
      "epoch": 0.7430830039525692,
      "grad_norm": 0.6545310020446777,
      "learning_rate": 1.7650681887482628e-05,
      "loss": 1.5924,
      "step": 329
    },
    {
      "epoch": 0.7453416149068323,
      "grad_norm": 0.760567307472229,
      "learning_rate": 1.7361612196988174e-05,
      "loss": 1.4892,
      "step": 330
    },
    {
      "epoch": 0.7476002258610954,
      "grad_norm": 0.5573106408119202,
      "learning_rate": 1.7074431046748075e-05,
      "loss": 1.3689,
      "step": 331
    },
    {
      "epoch": 0.7498588368153586,
      "grad_norm": 0.5589758157730103,
      "learning_rate": 1.678915505379513e-05,
      "loss": 1.2265,
      "step": 332
    },
    {
      "epoch": 0.7521174477696216,
      "grad_norm": 0.5898151397705078,
      "learning_rate": 1.650580072492496e-05,
      "loss": 1.3948,
      "step": 333
    },
    {
      "epoch": 0.7543760587238848,
      "grad_norm": 0.6013538241386414,
      "learning_rate": 1.6224384455740788e-05,
      "loss": 1.5207,
      "step": 334
    },
    {
      "epoch": 0.756634669678148,
      "grad_norm": 0.5616132020950317,
      "learning_rate": 1.5944922529704777e-05,
      "loss": 1.3745,
      "step": 335
    },
    {
      "epoch": 0.758893280632411,
      "grad_norm": 0.6084350943565369,
      "learning_rate": 1.5667431117195814e-05,
      "loss": 1.4857,
      "step": 336
    },
    {
      "epoch": 0.7611518915866742,
      "grad_norm": 0.6251053810119629,
      "learning_rate": 1.539192627457382e-05,
      "loss": 1.2732,
      "step": 337
    },
    {
      "epoch": 0.7634105025409373,
      "grad_norm": 0.8703694343566895,
      "learning_rate": 1.5118423943250771e-05,
      "loss": 1.5002,
      "step": 338
    },
    {
      "epoch": 0.7656691134952005,
      "grad_norm": 0.5518003702163696,
      "learning_rate": 1.4846939948768218e-05,
      "loss": 1.8132,
      "step": 339
    },
    {
      "epoch": 0.7679277244494636,
      "grad_norm": 0.5927785038948059,
      "learning_rate": 1.45774899998816e-05,
      "loss": 1.1446,
      "step": 340
    },
    {
      "epoch": 0.7701863354037267,
      "grad_norm": 0.5158970952033997,
      "learning_rate": 1.4310089687651301e-05,
      "loss": 1.2931,
      "step": 341
    },
    {
      "epoch": 0.7724449463579899,
      "grad_norm": 0.7069851160049438,
      "learning_rate": 1.40447544845405e-05,
      "loss": 1.4927,
      "step": 342
    },
    {
      "epoch": 0.7747035573122529,
      "grad_norm": 0.5784515142440796,
      "learning_rate": 1.378149974351991e-05,
      "loss": 1.3116,
      "step": 343
    },
    {
      "epoch": 0.7769621682665161,
      "grad_norm": 0.49697092175483704,
      "learning_rate": 1.3520340697179406e-05,
      "loss": 1.5359,
      "step": 344
    },
    {
      "epoch": 0.7792207792207793,
      "grad_norm": 0.6225888729095459,
      "learning_rate": 1.3261292456846647e-05,
      "loss": 1.4725,
      "step": 345
    },
    {
      "epoch": 0.7814793901750423,
      "grad_norm": 0.9478232860565186,
      "learning_rate": 1.3004370011712624e-05,
      "loss": 1.3835,
      "step": 346
    },
    {
      "epoch": 0.7837380011293055,
      "grad_norm": 0.5357046127319336,
      "learning_rate": 1.2749588227964465e-05,
      "loss": 1.2114,
      "step": 347
    },
    {
      "epoch": 0.7859966120835686,
      "grad_norm": 0.6515849828720093,
      "learning_rate": 1.2496961847925153e-05,
      "loss": 1.4449,
      "step": 348
    },
    {
      "epoch": 0.7882552230378317,
      "grad_norm": 0.7444521188735962,
      "learning_rate": 1.2246505489200532e-05,
      "loss": 1.2324,
      "step": 349
    },
    {
      "epoch": 0.7905138339920948,
      "grad_norm": 0.6521069407463074,
      "learning_rate": 1.1998233643833457e-05,
      "loss": 1.3745,
      "step": 350
    },
    {
      "epoch": 0.792772444946358,
      "grad_norm": 0.5900723934173584,
      "learning_rate": 1.1752160677465286e-05,
      "loss": 1.3835,
      "step": 351
    },
    {
      "epoch": 0.7950310559006211,
      "grad_norm": 0.5521007180213928,
      "learning_rate": 1.150830082850468e-05,
      "loss": 1.4689,
      "step": 352
    },
    {
      "epoch": 0.7972896668548842,
      "grad_norm": 0.43919599056243896,
      "learning_rate": 1.126666820730366e-05,
      "loss": 1.379,
      "step": 353
    },
    {
      "epoch": 0.7995482778091474,
      "grad_norm": 0.4976402223110199,
      "learning_rate": 1.1027276795341135e-05,
      "loss": 1.4056,
      "step": 354
    },
    {
      "epoch": 0.8018068887634106,
      "grad_norm": 0.7465052008628845,
      "learning_rate": 1.0790140444414e-05,
      "loss": 1.1132,
      "step": 355
    },
    {
      "epoch": 0.8040654997176736,
      "grad_norm": 1.412811279296875,
      "learning_rate": 1.0555272875835537e-05,
      "loss": 1.3229,
      "step": 356
    },
    {
      "epoch": 0.8063241106719368,
      "grad_norm": 0.38720905780792236,
      "learning_rate": 1.0322687679641523e-05,
      "loss": 1.3627,
      "step": 357
    },
    {
      "epoch": 0.8085827216261999,
      "grad_norm": 0.6548861265182495,
      "learning_rate": 1.0092398313803863e-05,
      "loss": 1.6172,
      "step": 358
    },
    {
      "epoch": 0.810841332580463,
      "grad_norm": 0.5794533491134644,
      "learning_rate": 9.864418103451828e-06,
      "loss": 1.5726,
      "step": 359
    },
    {
      "epoch": 0.8130999435347261,
      "grad_norm": 0.5520102977752686,
      "learning_rate": 9.638760240101102e-06,
      "loss": 1.294,
      "step": 360
    },
    {
      "epoch": 0.8153585544889893,
      "grad_norm": 0.5520592927932739,
      "learning_rate": 9.415437780890451e-06,
      "loss": 1.3229,
      "step": 361
    },
    {
      "epoch": 0.8176171654432524,
      "grad_norm": 0.644109845161438,
      "learning_rate": 9.194463647826223e-06,
      "loss": 1.6856,
      "step": 362
    },
    {
      "epoch": 0.8198757763975155,
      "grad_norm": 0.6648945808410645,
      "learning_rate": 8.975850627034604e-06,
      "loss": 1.4002,
      "step": 363
    },
    {
      "epoch": 0.8221343873517787,
      "grad_norm": 0.6234399676322937,
      "learning_rate": 8.759611368021831e-06,
      "loss": 1.33,
      "step": 364
    },
    {
      "epoch": 0.8243929983060417,
      "grad_norm": 0.6238119602203369,
      "learning_rate": 8.545758382942232e-06,
      "loss": 1.2347,
      "step": 365
    },
    {
      "epoch": 0.8266516092603049,
      "grad_norm": 0.5199403166770935,
      "learning_rate": 8.334304045874247e-06,
      "loss": 1.5012,
      "step": 366
    },
    {
      "epoch": 0.8289102202145681,
      "grad_norm": 0.43542924523353577,
      "learning_rate": 8.125260592104445e-06,
      "loss": 1.4619,
      "step": 367
    },
    {
      "epoch": 0.8311688311688312,
      "grad_norm": 0.6356791853904724,
      "learning_rate": 7.918640117419507e-06,
      "loss": 1.2857,
      "step": 368
    },
    {
      "epoch": 0.8334274421230943,
      "grad_norm": 0.8213852643966675,
      "learning_rate": 7.71445457740641e-06,
      "loss": 1.3405,
      "step": 369
    },
    {
      "epoch": 0.8356860530773574,
      "grad_norm": 0.6396834254264832,
      "learning_rate": 7.512715786760605e-06,
      "loss": 1.675,
      "step": 370
    },
    {
      "epoch": 0.8379446640316206,
      "grad_norm": 0.5670115947723389,
      "learning_rate": 7.313435418602388e-06,
      "loss": 1.4417,
      "step": 371
    },
    {
      "epoch": 0.8402032749858837,
      "grad_norm": 0.8237860798835754,
      "learning_rate": 7.116625003801436e-06,
      "loss": 1.3525,
      "step": 372
    },
    {
      "epoch": 0.8424618859401468,
      "grad_norm": 0.6882337927818298,
      "learning_rate": 6.922295930309691e-06,
      "loss": 1.5534,
      "step": 373
    },
    {
      "epoch": 0.84472049689441,
      "grad_norm": 0.750078022480011,
      "learning_rate": 6.730459442502329e-06,
      "loss": 1.5733,
      "step": 374
    },
    {
      "epoch": 0.846979107848673,
      "grad_norm": 0.5471389889717102,
      "learning_rate": 6.541126640527195e-06,
      "loss": 1.6169,
      "step": 375
    },
    {
      "epoch": 0.8492377188029362,
      "grad_norm": 0.4878935217857361,
      "learning_rate": 6.354308479662446e-06,
      "loss": 1.5414,
      "step": 376
    },
    {
      "epoch": 0.8514963297571994,
      "grad_norm": 0.5643681883811951,
      "learning_rate": 6.170015769682741e-06,
      "loss": 1.6472,
      "step": 377
    },
    {
      "epoch": 0.8537549407114624,
      "grad_norm": 0.5661868453025818,
      "learning_rate": 5.988259174233713e-06,
      "loss": 1.3295,
      "step": 378
    },
    {
      "epoch": 0.8560135516657256,
      "grad_norm": 0.8035285472869873,
      "learning_rate": 5.80904921021494e-06,
      "loss": 1.4848,
      "step": 379
    },
    {
      "epoch": 0.8582721626199887,
      "grad_norm": 0.48411625623703003,
      "learning_rate": 5.6323962471714286e-06,
      "loss": 1.3458,
      "step": 380
    },
    {
      "epoch": 0.8605307735742518,
      "grad_norm": 0.5900686979293823,
      "learning_rate": 5.458310506693571e-06,
      "loss": 1.5151,
      "step": 381
    },
    {
      "epoch": 0.862789384528515,
      "grad_norm": 0.5503493547439575,
      "learning_rate": 5.286802061825752e-06,
      "loss": 1.4399,
      "step": 382
    },
    {
      "epoch": 0.8650479954827781,
      "grad_norm": 0.4653548002243042,
      "learning_rate": 5.117880836483452e-06,
      "loss": 1.5298,
      "step": 383
    },
    {
      "epoch": 0.8673066064370413,
      "grad_norm": 0.5434414148330688,
      "learning_rate": 4.951556604879048e-06,
      "loss": 1.3049,
      "step": 384
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 0.8023334741592407,
      "learning_rate": 4.7878389909562285e-06,
      "loss": 1.4366,
      "step": 385
    },
    {
      "epoch": 0.8718238283455675,
      "grad_norm": 0.5557988286018372,
      "learning_rate": 4.62673746783317e-06,
      "loss": 1.2165,
      "step": 386
    },
    {
      "epoch": 0.8740824392998306,
      "grad_norm": 0.4488828778266907,
      "learning_rate": 4.468261357254339e-06,
      "loss": 1.1717,
      "step": 387
    },
    {
      "epoch": 0.8763410502540937,
      "grad_norm": 0.5039349794387817,
      "learning_rate": 4.312419829051173e-06,
      "loss": 1.5923,
      "step": 388
    },
    {
      "epoch": 0.8785996612083569,
      "grad_norm": 0.5717617273330688,
      "learning_rate": 4.15922190061146e-06,
      "loss": 1.3688,
      "step": 389
    },
    {
      "epoch": 0.88085827216262,
      "grad_norm": 0.6537705063819885,
      "learning_rate": 4.008676436357539e-06,
      "loss": 1.4746,
      "step": 390
    },
    {
      "epoch": 0.8831168831168831,
      "grad_norm": 0.7428813576698303,
      "learning_rate": 3.86079214723345e-06,
      "loss": 1.6453,
      "step": 391
    },
    {
      "epoch": 0.8853754940711462,
      "grad_norm": 0.6995297074317932,
      "learning_rate": 3.7155775902008526e-06,
      "loss": 1.2413,
      "step": 392
    },
    {
      "epoch": 0.8876341050254094,
      "grad_norm": 0.5746084451675415,
      "learning_rate": 3.5730411677439125e-06,
      "loss": 1.5308,
      "step": 393
    },
    {
      "epoch": 0.8898927159796725,
      "grad_norm": 0.7818319797515869,
      "learning_rate": 3.4331911273830784e-06,
      "loss": 1.2739,
      "step": 394
    },
    {
      "epoch": 0.8921513269339356,
      "grad_norm": 0.4729510545730591,
      "learning_rate": 3.2960355611979245e-06,
      "loss": 1.2642,
      "step": 395
    },
    {
      "epoch": 0.8944099378881988,
      "grad_norm": 0.5390079021453857,
      "learning_rate": 3.161582405358876e-06,
      "loss": 1.4629,
      "step": 396
    },
    {
      "epoch": 0.8966685488424618,
      "grad_norm": 0.5690498352050781,
      "learning_rate": 3.029839439668003e-06,
      "loss": 1.576,
      "step": 397
    },
    {
      "epoch": 0.898927159796725,
      "grad_norm": 0.6022957563400269,
      "learning_rate": 2.9008142871088663e-06,
      "loss": 1.4105,
      "step": 398
    },
    {
      "epoch": 0.9011857707509882,
      "grad_norm": 0.4936705231666565,
      "learning_rate": 2.7745144134054433e-06,
      "loss": 1.3807,
      "step": 399
    },
    {
      "epoch": 0.9034443817052513,
      "grad_norm": 0.5045138001441956,
      "learning_rate": 2.6509471265901477e-06,
      "loss": 1.427,
      "step": 400
    },
    {
      "epoch": 0.9057029926595144,
      "grad_norm": 0.5444932579994202,
      "learning_rate": 2.530119576580936e-06,
      "loss": 1.3433,
      "step": 401
    },
    {
      "epoch": 0.9079616036137775,
      "grad_norm": 0.48583537340164185,
      "learning_rate": 2.412038754767626e-06,
      "loss": 1.2593,
      "step": 402
    },
    {
      "epoch": 0.9102202145680407,
      "grad_norm": 0.7965908050537109,
      "learning_rate": 2.296711493607334e-06,
      "loss": 1.304,
      "step": 403
    },
    {
      "epoch": 0.9124788255223037,
      "grad_norm": 0.6211166381835938,
      "learning_rate": 2.1841444662291543e-06,
      "loss": 1.4832,
      "step": 404
    },
    {
      "epoch": 0.9147374364765669,
      "grad_norm": 0.5326806902885437,
      "learning_rate": 2.074344186048022e-06,
      "loss": 1.239,
      "step": 405
    },
    {
      "epoch": 0.9169960474308301,
      "grad_norm": 0.46467143297195435,
      "learning_rate": 1.967317006387831e-06,
      "loss": 1.5442,
      "step": 406
    },
    {
      "epoch": 0.9192546583850931,
      "grad_norm": 0.6954487562179565,
      "learning_rate": 1.863069120113814e-06,
      "loss": 1.0449,
      "step": 407
    },
    {
      "epoch": 0.9215132693393563,
      "grad_norm": 0.5407376885414124,
      "learning_rate": 1.7616065592742038e-06,
      "loss": 1.368,
      "step": 408
    },
    {
      "epoch": 0.9237718802936195,
      "grad_norm": 0.5914408564567566,
      "learning_rate": 1.6629351947512195e-06,
      "loss": 1.4579,
      "step": 409
    },
    {
      "epoch": 0.9260304912478825,
      "grad_norm": 0.5790956020355225,
      "learning_rate": 1.567060735921344e-06,
      "loss": 1.5103,
      "step": 410
    },
    {
      "epoch": 0.9282891022021457,
      "grad_norm": 0.7009652853012085,
      "learning_rate": 1.4739887303249877e-06,
      "loss": 1.4829,
      "step": 411
    },
    {
      "epoch": 0.9305477131564088,
      "grad_norm": 0.6565191149711609,
      "learning_rate": 1.383724563345451e-06,
      "loss": 1.6771,
      "step": 412
    },
    {
      "epoch": 0.932806324110672,
      "grad_norm": 0.6377475261688232,
      "learning_rate": 1.2962734578973568e-06,
      "loss": 1.4032,
      "step": 413
    },
    {
      "epoch": 0.935064935064935,
      "grad_norm": 0.6361109018325806,
      "learning_rate": 1.2116404741244203e-06,
      "loss": 1.4081,
      "step": 414
    },
    {
      "epoch": 0.9373235460191982,
      "grad_norm": 0.517658531665802,
      "learning_rate": 1.1298305091066664e-06,
      "loss": 1.3421,
      "step": 415
    },
    {
      "epoch": 0.9395821569734614,
      "grad_norm": 0.7061721086502075,
      "learning_rate": 1.0508482965770505e-06,
      "loss": 1.5959,
      "step": 416
    },
    {
      "epoch": 0.9418407679277244,
      "grad_norm": 0.5791245698928833,
      "learning_rate": 9.746984066475729e-07,
      "loss": 1.3128,
      "step": 417
    },
    {
      "epoch": 0.9440993788819876,
      "grad_norm": 0.7557624578475952,
      "learning_rate": 9.013852455448335e-07,
      "loss": 1.2763,
      "step": 418
    },
    {
      "epoch": 0.9463579898362507,
      "grad_norm": 0.9951817989349365,
      "learning_rate": 8.309130553550815e-07,
      "loss": 1.4941,
      "step": 419
    },
    {
      "epoch": 0.9486166007905138,
      "grad_norm": 0.3905503749847412,
      "learning_rate": 7.63285913778733e-07,
      "loss": 1.3464,
      "step": 420
    },
    {
      "epoch": 0.950875211744777,
      "grad_norm": 0.5072855949401855,
      "learning_rate": 6.985077338944657e-07,
      "loss": 1.0404,
      "step": 421
    },
    {
      "epoch": 0.9531338226990401,
      "grad_norm": 0.5355942845344543,
      "learning_rate": 6.365822639327723e-07,
      "loss": 1.2642,
      "step": 422
    },
    {
      "epoch": 0.9553924336533032,
      "grad_norm": 0.5786038041114807,
      "learning_rate": 5.775130870590783e-07,
      "loss": 1.3897,
      "step": 423
    },
    {
      "epoch": 0.9576510446075663,
      "grad_norm": 0.564463198184967,
      "learning_rate": 5.213036211664191e-07,
      "loss": 1.4783,
      "step": 424
    },
    {
      "epoch": 0.9599096555618295,
      "grad_norm": 0.5604745745658875,
      "learning_rate": 4.6795711867766436e-07,
      "loss": 1.6895,
      "step": 425
    },
    {
      "epoch": 0.9621682665160926,
      "grad_norm": 0.7266672849655151,
      "learning_rate": 4.1747666635733597e-07,
      "loss": 1.3143,
      "step": 426
    },
    {
      "epoch": 0.9644268774703557,
      "grad_norm": 0.7507563829421997,
      "learning_rate": 3.698651851329837e-07,
      "loss": 1.4809,
      "step": 427
    },
    {
      "epoch": 0.9666854884246189,
      "grad_norm": 0.5136380195617676,
      "learning_rate": 3.251254299261874e-07,
      "loss": 1.6922,
      "step": 428
    },
    {
      "epoch": 0.968944099378882,
      "grad_norm": 0.7942318916320801,
      "learning_rate": 2.8325998949314536e-07,
      "loss": 1.5874,
      "step": 429
    },
    {
      "epoch": 0.9712027103331451,
      "grad_norm": 0.46300673484802246,
      "learning_rate": 2.442712862748775e-07,
      "loss": 1.7898,
      "step": 430
    },
    {
      "epoch": 0.9734613212874083,
      "grad_norm": 0.6884695291519165,
      "learning_rate": 2.0816157625706545e-07,
      "loss": 1.33,
      "step": 431
    },
    {
      "epoch": 0.9757199322416714,
      "grad_norm": 0.5976503491401672,
      "learning_rate": 1.749329488395124e-07,
      "loss": 1.6437,
      "step": 432
    },
    {
      "epoch": 0.9779785431959345,
      "grad_norm": 0.42137616872787476,
      "learning_rate": 1.4458732671523977e-07,
      "loss": 1.2883,
      "step": 433
    },
    {
      "epoch": 0.9802371541501976,
      "grad_norm": 0.6945174932479858,
      "learning_rate": 1.1712646575922637e-07,
      "loss": 1.3545,
      "step": 434
    },
    {
      "epoch": 0.9824957651044608,
      "grad_norm": 0.5032137036323547,
      "learning_rate": 9.255195492685609e-08,
      "loss": 1.4777,
      "step": 435
    },
    {
      "epoch": 0.9847543760587238,
      "grad_norm": 0.6473844647407532,
      "learning_rate": 7.086521616190279e-08,
      "loss": 1.4482,
      "step": 436
    },
    {
      "epoch": 0.987012987012987,
      "grad_norm": 1.0584694147109985,
      "learning_rate": 5.2067504314323723e-08,
      "loss": 1.2144,
      "step": 437
    },
    {
      "epoch": 0.9892715979672502,
      "grad_norm": 0.4894103407859802,
      "learning_rate": 3.6159907067601085e-08,
      "loss": 1.3691,
      "step": 438
    },
    {
      "epoch": 0.9915302089215132,
      "grad_norm": 0.6497173309326172,
      "learning_rate": 2.3143344875831142e-08,
      "loss": 1.3208,
      "step": 439
    },
    {
      "epoch": 0.9937888198757764,
      "grad_norm": 0.5225093364715576,
      "learning_rate": 1.3018570910466877e-08,
      "loss": 1.1954,
      "step": 440
    },
    {
      "epoch": 0.9960474308300395,
      "grad_norm": 0.574373185634613,
      "learning_rate": 5.786171016708419e-09,
      "loss": 1.598,
      "step": 441
    },
    {
      "epoch": 0.9983060417843026,
      "grad_norm": 0.6945062875747681,
      "learning_rate": 1.446563679641244e-09,
      "loss": 1.4279,
      "step": 442
    },
    {
      "epoch": 1.0005646527385659,
      "grad_norm": 0.6715332865715027,
      "learning_rate": 0.0,
      "loss": 1.5682,
      "step": 443
    }
  ],
  "logging_steps": 1,
  "max_steps": 443,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.630732306700042e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}