{
  "best_metric": 1.3403037786483765,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.19356399709654004,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0009678199854827002,
      "grad_norm": 2.568981885910034,
      "learning_rate": 1e-05,
      "loss": 5.1665,
      "step": 1
    },
    {
      "epoch": 0.0009678199854827002,
      "eval_loss": 1.9932328462600708,
      "eval_runtime": 35.1741,
      "eval_samples_per_second": 49.468,
      "eval_steps_per_second": 12.367,
      "step": 1
    },
    {
      "epoch": 0.0019356399709654004,
      "grad_norm": 2.8290305137634277,
      "learning_rate": 2e-05,
      "loss": 5.1986,
      "step": 2
    },
    {
      "epoch": 0.0029034599564481006,
      "grad_norm": 2.9309866428375244,
      "learning_rate": 3e-05,
      "loss": 5.6663,
      "step": 3
    },
    {
      "epoch": 0.0038712799419308008,
      "grad_norm": 2.9830682277679443,
      "learning_rate": 4e-05,
      "loss": 6.5639,
      "step": 4
    },
    {
      "epoch": 0.004839099927413501,
      "grad_norm": 2.6141579151153564,
      "learning_rate": 5e-05,
      "loss": 5.4302,
      "step": 5
    },
    {
      "epoch": 0.005806919912896201,
      "grad_norm": 2.7034385204315186,
      "learning_rate": 6e-05,
      "loss": 5.9713,
      "step": 6
    },
    {
      "epoch": 0.006774739898378902,
      "grad_norm": 3.039058208465576,
      "learning_rate": 7e-05,
      "loss": 6.224,
      "step": 7
    },
    {
      "epoch": 0.0077425598838616015,
      "grad_norm": 3.1538257598876953,
      "learning_rate": 8e-05,
      "loss": 6.6107,
      "step": 8
    },
    {
      "epoch": 0.008710379869344302,
      "grad_norm": 3.0472660064697266,
      "learning_rate": 9e-05,
      "loss": 6.3953,
      "step": 9
    },
    {
      "epoch": 0.009678199854827003,
      "grad_norm": 3.3916268348693848,
      "learning_rate": 0.0001,
      "loss": 6.4126,
      "step": 10
    },
    {
      "epoch": 0.010646019840309702,
      "grad_norm": 3.6630942821502686,
      "learning_rate": 9.999316524962345e-05,
      "loss": 6.6071,
      "step": 11
    },
    {
      "epoch": 0.011613839825792402,
      "grad_norm": 3.2626712322235107,
      "learning_rate": 9.997266286704631e-05,
      "loss": 6.3308,
      "step": 12
    },
    {
      "epoch": 0.012581659811275103,
      "grad_norm": 3.4051663875579834,
      "learning_rate": 9.993849845741524e-05,
      "loss": 6.5894,
      "step": 13
    },
    {
      "epoch": 0.013549479796757804,
      "grad_norm": 3.654900312423706,
      "learning_rate": 9.989068136093873e-05,
      "loss": 6.5473,
      "step": 14
    },
    {
      "epoch": 0.014517299782240502,
      "grad_norm": 3.4287214279174805,
      "learning_rate": 9.98292246503335e-05,
      "loss": 6.3596,
      "step": 15
    },
    {
      "epoch": 0.015485119767723203,
      "grad_norm": 3.3961222171783447,
      "learning_rate": 9.975414512725057e-05,
      "loss": 6.3359,
      "step": 16
    },
    {
      "epoch": 0.016452939753205904,
      "grad_norm": 3.2967514991760254,
      "learning_rate": 9.966546331768191e-05,
      "loss": 6.2213,
      "step": 17
    },
    {
      "epoch": 0.017420759738688604,
      "grad_norm": 3.308589458465576,
      "learning_rate": 9.956320346634876e-05,
      "loss": 6.4394,
      "step": 18
    },
    {
      "epoch": 0.018388579724171305,
      "grad_norm": 3.3042125701904297,
      "learning_rate": 9.944739353007344e-05,
      "loss": 6.464,
      "step": 19
    },
    {
      "epoch": 0.019356399709654006,
      "grad_norm": 3.4895613193511963,
      "learning_rate": 9.931806517013612e-05,
      "loss": 6.185,
      "step": 20
    },
    {
      "epoch": 0.020324219695136706,
      "grad_norm": 3.320256471633911,
      "learning_rate": 9.917525374361912e-05,
      "loss": 6.5983,
      "step": 21
    },
    {
      "epoch": 0.021292039680619403,
      "grad_norm": 3.424269199371338,
      "learning_rate": 9.901899829374047e-05,
      "loss": 6.882,
      "step": 22
    },
    {
      "epoch": 0.022259859666102104,
      "grad_norm": 3.3450891971588135,
      "learning_rate": 9.884934153917997e-05,
      "loss": 6.2008,
      "step": 23
    },
    {
      "epoch": 0.023227679651584805,
      "grad_norm": 3.732424259185791,
      "learning_rate": 9.86663298624003e-05,
      "loss": 6.4244,
      "step": 24
    },
    {
      "epoch": 0.024195499637067505,
      "grad_norm": 3.5285675525665283,
      "learning_rate": 9.847001329696653e-05,
      "loss": 6.0386,
      "step": 25
    },
    {
      "epoch": 0.025163319622550206,
      "grad_norm": 3.930633783340454,
      "learning_rate": 9.826044551386744e-05,
      "loss": 5.9879,
      "step": 26
    },
    {
      "epoch": 0.026131139608032906,
      "grad_norm": 3.7356884479522705,
      "learning_rate": 9.803768380684242e-05,
      "loss": 6.9487,
      "step": 27
    },
    {
      "epoch": 0.027098959593515607,
      "grad_norm": 4.129487037658691,
      "learning_rate": 9.780178907671789e-05,
      "loss": 6.1843,
      "step": 28
    },
    {
      "epoch": 0.028066779578998308,
      "grad_norm": 3.8816277980804443,
      "learning_rate": 9.755282581475769e-05,
      "loss": 6.7426,
      "step": 29
    },
    {
      "epoch": 0.029034599564481005,
      "grad_norm": 4.013293266296387,
      "learning_rate": 9.729086208503174e-05,
      "loss": 5.9813,
      "step": 30
    },
    {
      "epoch": 0.030002419549963705,
      "grad_norm": 4.5263471603393555,
      "learning_rate": 9.701596950580806e-05,
      "loss": 6.8728,
      "step": 31
    },
    {
      "epoch": 0.030970239535446406,
      "grad_norm": 4.742683410644531,
      "learning_rate": 9.672822322997305e-05,
      "loss": 6.7176,
      "step": 32
    },
    {
      "epoch": 0.03193805952092911,
      "grad_norm": 4.742998123168945,
      "learning_rate": 9.642770192448536e-05,
      "loss": 6.867,
      "step": 33
    },
    {
      "epoch": 0.03290587950641181,
      "grad_norm": 4.465890884399414,
      "learning_rate": 9.611448774886924e-05,
      "loss": 6.9411,
      "step": 34
    },
    {
      "epoch": 0.033873699491894504,
      "grad_norm": 4.534784317016602,
      "learning_rate": 9.578866633275288e-05,
      "loss": 6.1718,
      "step": 35
    },
    {
      "epoch": 0.03484151947737721,
      "grad_norm": 5.096901893615723,
      "learning_rate": 9.545032675245813e-05,
      "loss": 6.4728,
      "step": 36
    },
    {
      "epoch": 0.035809339462859906,
      "grad_norm": 5.102327823638916,
      "learning_rate": 9.509956150664796e-05,
      "loss": 7.3079,
      "step": 37
    },
    {
      "epoch": 0.03677715944834261,
      "grad_norm": 5.680893898010254,
      "learning_rate": 9.473646649103818e-05,
      "loss": 6.9684,
      "step": 38
    },
    {
      "epoch": 0.03774497943382531,
      "grad_norm": 5.745683670043945,
      "learning_rate": 9.43611409721806e-05,
      "loss": 7.5348,
      "step": 39
    },
    {
      "epoch": 0.03871279941930801,
      "grad_norm": 4.869477272033691,
      "learning_rate": 9.397368756032445e-05,
      "loss": 6.6506,
      "step": 40
    },
    {
      "epoch": 0.03968061940479071,
      "grad_norm": 6.065983772277832,
      "learning_rate": 9.357421218136386e-05,
      "loss": 6.3111,
      "step": 41
    },
    {
      "epoch": 0.04064843939027341,
      "grad_norm": 6.838200092315674,
      "learning_rate": 9.316282404787871e-05,
      "loss": 6.9735,
      "step": 42
    },
    {
      "epoch": 0.04161625937575611,
      "grad_norm": 6.886222839355469,
      "learning_rate": 9.273963562927695e-05,
      "loss": 6.7475,
      "step": 43
    },
    {
      "epoch": 0.04258407936123881,
      "grad_norm": 6.245997428894043,
      "learning_rate": 9.230476262104677e-05,
      "loss": 7.1447,
      "step": 44
    },
    {
      "epoch": 0.04355189934672151,
      "grad_norm": 5.931451320648193,
      "learning_rate": 9.185832391312644e-05,
      "loss": 6.5921,
      "step": 45
    },
    {
      "epoch": 0.04451971933220421,
      "grad_norm": 6.378926753997803,
      "learning_rate": 9.140044155740101e-05,
      "loss": 6.4475,
      "step": 46
    },
    {
      "epoch": 0.04548753931768691,
      "grad_norm": 7.441506385803223,
      "learning_rate": 9.093124073433463e-05,
      "loss": 7.1372,
      "step": 47
    },
    {
      "epoch": 0.04645535930316961,
      "grad_norm": 8.055678367614746,
      "learning_rate": 9.045084971874738e-05,
      "loss": 6.3702,
      "step": 48
    },
    {
      "epoch": 0.04742317928865231,
      "grad_norm": 8.375804901123047,
      "learning_rate": 8.995939984474624e-05,
      "loss": 6.1476,
      "step": 49
    },
    {
      "epoch": 0.04839099927413501,
      "grad_norm": 10.582734107971191,
      "learning_rate": 8.945702546981969e-05,
      "loss": 6.706,
      "step": 50
    },
    {
      "epoch": 0.04839099927413501,
      "eval_loss": 1.6609218120574951,
      "eval_runtime": 35.0112,
      "eval_samples_per_second": 49.698,
      "eval_steps_per_second": 12.425,
      "step": 50
    },
    {
      "epoch": 0.049358819259617714,
      "grad_norm": 17.193357467651367,
      "learning_rate": 8.894386393810563e-05,
      "loss": 5.9318,
      "step": 51
    },
    {
      "epoch": 0.05032663924510041,
      "grad_norm": 13.682241439819336,
      "learning_rate": 8.842005554284296e-05,
      "loss": 5.6831,
      "step": 52
    },
    {
      "epoch": 0.05129445923058311,
      "grad_norm": 8.937365531921387,
      "learning_rate": 8.788574348801675e-05,
      "loss": 6.2483,
      "step": 53
    },
    {
      "epoch": 0.05226227921606581,
      "grad_norm": 5.6092143058776855,
      "learning_rate": 8.73410738492077e-05,
      "loss": 5.5013,
      "step": 54
    },
    {
      "epoch": 0.05323009920154851,
      "grad_norm": 4.564438343048096,
      "learning_rate": 8.678619553365659e-05,
      "loss": 5.0638,
      "step": 55
    },
    {
      "epoch": 0.054197919187031214,
      "grad_norm": 4.319308757781982,
      "learning_rate": 8.622126023955446e-05,
      "loss": 5.9632,
      "step": 56
    },
    {
      "epoch": 0.05516573917251391,
      "grad_norm": 3.6995363235473633,
      "learning_rate": 8.564642241456986e-05,
      "loss": 5.7502,
      "step": 57
    },
    {
      "epoch": 0.056133559157996615,
      "grad_norm": 3.658045768737793,
      "learning_rate": 8.506183921362443e-05,
      "loss": 5.7649,
      "step": 58
    },
    {
      "epoch": 0.05710137914347931,
      "grad_norm": 3.488203287124634,
      "learning_rate": 8.44676704559283e-05,
      "loss": 5.7465,
      "step": 59
    },
    {
      "epoch": 0.05806919912896201,
      "grad_norm": 3.6048405170440674,
      "learning_rate": 8.386407858128706e-05,
      "loss": 6.0193,
      "step": 60
    },
    {
      "epoch": 0.059037019114444714,
      "grad_norm": 3.1448967456817627,
      "learning_rate": 8.32512286056924e-05,
      "loss": 5.6209,
      "step": 61
    },
    {
      "epoch": 0.06000483909992741,
      "grad_norm": 3.4668450355529785,
      "learning_rate": 8.262928807620843e-05,
      "loss": 5.8335,
      "step": 62
    },
    {
      "epoch": 0.060972659085410115,
      "grad_norm": 3.147052764892578,
      "learning_rate": 8.199842702516583e-05,
      "loss": 5.5159,
      "step": 63
    },
    {
      "epoch": 0.06194047907089281,
      "grad_norm": 2.807248115539551,
      "learning_rate": 8.135881792367686e-05,
      "loss": 5.7114,
      "step": 64
    },
    {
      "epoch": 0.06290829905637552,
      "grad_norm": 2.7647271156311035,
      "learning_rate": 8.07106356344834e-05,
      "loss": 5.4396,
      "step": 65
    },
    {
      "epoch": 0.06387611904185822,
      "grad_norm": 3.149815797805786,
      "learning_rate": 8.005405736415126e-05,
      "loss": 5.3517,
      "step": 66
    },
    {
      "epoch": 0.06484393902734091,
      "grad_norm": 3.1891109943389893,
      "learning_rate": 7.938926261462366e-05,
      "loss": 6.1686,
      "step": 67
    },
    {
      "epoch": 0.06581175901282361,
      "grad_norm": 2.856987714767456,
      "learning_rate": 7.871643313414718e-05,
      "loss": 5.7709,
      "step": 68
    },
    {
      "epoch": 0.06677957899830632,
      "grad_norm": 3.087233066558838,
      "learning_rate": 7.803575286758364e-05,
      "loss": 5.9863,
      "step": 69
    },
    {
      "epoch": 0.06774739898378901,
      "grad_norm": 3.097813367843628,
      "learning_rate": 7.734740790612136e-05,
      "loss": 5.8892,
      "step": 70
    },
    {
      "epoch": 0.06871521896927171,
      "grad_norm": 3.0293822288513184,
      "learning_rate": 7.66515864363997e-05,
      "loss": 6.4676,
      "step": 71
    },
    {
      "epoch": 0.06968303895475442,
      "grad_norm": 3.2445662021636963,
      "learning_rate": 7.594847868906076e-05,
      "loss": 5.7144,
      "step": 72
    },
    {
      "epoch": 0.07065085894023712,
      "grad_norm": 2.8740124702453613,
      "learning_rate": 7.52382768867422e-05,
      "loss": 5.504,
      "step": 73
    },
    {
      "epoch": 0.07161867892571981,
      "grad_norm": 3.131711721420288,
      "learning_rate": 7.452117519152542e-05,
      "loss": 5.934,
      "step": 74
    },
    {
      "epoch": 0.07258649891120252,
      "grad_norm": 3.09757137298584,
      "learning_rate": 7.379736965185368e-05,
      "loss": 5.5717,
      "step": 75
    },
    {
      "epoch": 0.07355431889668522,
      "grad_norm": 3.3649322986602783,
      "learning_rate": 7.30670581489344e-05,
      "loss": 5.9404,
      "step": 76
    },
    {
      "epoch": 0.07452213888216792,
      "grad_norm": 3.4723610877990723,
      "learning_rate": 7.233044034264034e-05,
      "loss": 6.4689,
      "step": 77
    },
    {
      "epoch": 0.07548995886765061,
      "grad_norm": 3.6143648624420166,
      "learning_rate": 7.158771761692464e-05,
      "loss": 6.1636,
      "step": 78
    },
    {
      "epoch": 0.07645777885313332,
      "grad_norm": 3.409687042236328,
      "learning_rate": 7.083909302476453e-05,
      "loss": 6.0006,
      "step": 79
    },
    {
      "epoch": 0.07742559883861602,
      "grad_norm": 3.4327709674835205,
      "learning_rate": 7.008477123264848e-05,
      "loss": 6.3141,
      "step": 80
    },
    {
      "epoch": 0.07839341882409871,
      "grad_norm": 3.7295753955841064,
      "learning_rate": 6.932495846462261e-05,
      "loss": 6.3148,
      "step": 81
    },
    {
      "epoch": 0.07936123880958142,
      "grad_norm": 3.917672634124756,
      "learning_rate": 6.855986244591104e-05,
      "loss": 6.1807,
      "step": 82
    },
    {
      "epoch": 0.08032905879506412,
      "grad_norm": 3.7444727420806885,
      "learning_rate": 6.778969234612584e-05,
      "loss": 5.9584,
      "step": 83
    },
    {
      "epoch": 0.08129687878054682,
      "grad_norm": 4.129123210906982,
      "learning_rate": 6.701465872208216e-05,
      "loss": 6.014,
      "step": 84
    },
    {
      "epoch": 0.08226469876602951,
      "grad_norm": 3.9458887577056885,
      "learning_rate": 6.623497346023418e-05,
      "loss": 6.2226,
      "step": 85
    },
    {
      "epoch": 0.08323251875151222,
      "grad_norm": 4.261864185333252,
      "learning_rate": 6.545084971874738e-05,
      "loss": 5.4426,
      "step": 86
    },
    {
      "epoch": 0.08420033873699492,
      "grad_norm": 4.15214729309082,
      "learning_rate": 6.466250186922325e-05,
      "loss": 6.1595,
      "step": 87
    },
    {
      "epoch": 0.08516815872247761,
      "grad_norm": 4.719447612762451,
      "learning_rate": 6.387014543809223e-05,
      "loss": 6.2272,
      "step": 88
    },
    {
      "epoch": 0.08613597870796032,
      "grad_norm": 4.579586029052734,
      "learning_rate": 6.307399704769099e-05,
      "loss": 6.297,
      "step": 89
    },
    {
      "epoch": 0.08710379869344302,
      "grad_norm": 4.810436725616455,
      "learning_rate": 6.227427435703997e-05,
      "loss": 6.6752,
      "step": 90
    },
    {
      "epoch": 0.08807161867892573,
      "grad_norm": 5.3561811447143555,
      "learning_rate": 6.147119600233758e-05,
      "loss": 6.5366,
      "step": 91
    },
    {
      "epoch": 0.08903943866440842,
      "grad_norm": 4.887711524963379,
      "learning_rate": 6.066498153718735e-05,
      "loss": 5.9314,
      "step": 92
    },
    {
      "epoch": 0.09000725864989112,
      "grad_norm": 4.919368267059326,
      "learning_rate": 5.985585137257401e-05,
      "loss": 6.1798,
      "step": 93
    },
    {
      "epoch": 0.09097507863537382,
      "grad_norm": 5.564383029937744,
      "learning_rate": 5.90440267166055e-05,
      "loss": 6.02,
      "step": 94
    },
    {
      "epoch": 0.09194289862085651,
      "grad_norm": 5.642486572265625,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 6.5244,
      "step": 95
    },
    {
      "epoch": 0.09291071860633922,
      "grad_norm": 6.08746337890625,
      "learning_rate": 5.74131823855921e-05,
      "loss": 5.7704,
      "step": 96
    },
    {
      "epoch": 0.09387853859182192,
      "grad_norm": 6.154653072357178,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 6.4006,
      "step": 97
    },
    {
      "epoch": 0.09484635857730463,
      "grad_norm": 6.223016738891602,
      "learning_rate": 5.577423184847932e-05,
      "loss": 6.0589,
      "step": 98
    },
    {
      "epoch": 0.09581417856278732,
      "grad_norm": 6.987191200256348,
      "learning_rate": 5.495227651252315e-05,
      "loss": 6.4134,
      "step": 99
    },
    {
      "epoch": 0.09678199854827002,
      "grad_norm": 7.913987159729004,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 6.0659,
      "step": 100
    },
    {
      "epoch": 0.09678199854827002,
      "eval_loss": 1.440259575843811,
      "eval_runtime": 35.0208,
      "eval_samples_per_second": 49.685,
      "eval_steps_per_second": 12.421,
      "step": 100
    },
    {
      "epoch": 0.09774981853375272,
      "grad_norm": 4.391490936279297,
      "learning_rate": 5.330452921628497e-05,
      "loss": 4.3234,
      "step": 101
    },
    {
      "epoch": 0.09871763851923543,
      "grad_norm": 4.863372802734375,
      "learning_rate": 5.247918773366112e-05,
      "loss": 4.5842,
      "step": 102
    },
    {
      "epoch": 0.09968545850471812,
      "grad_norm": 4.380272388458252,
      "learning_rate": 5.165316846586541e-05,
      "loss": 5.2117,
      "step": 103
    },
    {
      "epoch": 0.10065327849020082,
      "grad_norm": 3.587648868560791,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 4.6927,
      "step": 104
    },
    {
      "epoch": 0.10162109847568353,
      "grad_norm": 2.9153084754943848,
      "learning_rate": 5e-05,
      "loss": 5.0881,
      "step": 105
    },
    {
      "epoch": 0.10258891846116622,
      "grad_norm": 2.8398003578186035,
      "learning_rate": 4.917330276168208e-05,
      "loss": 5.2936,
      "step": 106
    },
    {
      "epoch": 0.10355673844664892,
      "grad_norm": 2.9500510692596436,
      "learning_rate": 4.834683153413459e-05,
      "loss": 4.7048,
      "step": 107
    },
    {
      "epoch": 0.10452455843213163,
      "grad_norm": 3.102827548980713,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 5.011,
      "step": 108
    },
    {
      "epoch": 0.10549237841761433,
      "grad_norm": 2.6776440143585205,
      "learning_rate": 4.669547078371504e-05,
      "loss": 5.2105,
      "step": 109
    },
    {
      "epoch": 0.10646019840309702,
      "grad_norm": 2.703883647918701,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 5.2939,
      "step": 110
    },
    {
      "epoch": 0.10742801838857972,
      "grad_norm": 3.025521993637085,
      "learning_rate": 4.504772348747687e-05,
      "loss": 5.4026,
      "step": 111
    },
    {
      "epoch": 0.10839583837406243,
      "grad_norm": 2.7225425243377686,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 4.9054,
      "step": 112
    },
    {
      "epoch": 0.10936365835954512,
      "grad_norm": 3.217029571533203,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 5.2622,
      "step": 113
    },
    {
      "epoch": 0.11033147834502782,
      "grad_norm": 3.098397731781006,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 5.8714,
      "step": 114
    },
    {
      "epoch": 0.11129929833051053,
      "grad_norm": 3.2830111980438232,
      "learning_rate": 4.17702704859633e-05,
      "loss": 5.5096,
      "step": 115
    },
    {
      "epoch": 0.11226711831599323,
      "grad_norm": 2.994551420211792,
      "learning_rate": 4.095597328339452e-05,
      "loss": 5.0694,
      "step": 116
    },
    {
      "epoch": 0.11323493830147592,
      "grad_norm": 3.101209878921509,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 5.1881,
      "step": 117
    },
    {
      "epoch": 0.11420275828695862,
      "grad_norm": 2.929673433303833,
      "learning_rate": 3.933501846281267e-05,
      "loss": 5.6943,
      "step": 118
    },
    {
      "epoch": 0.11517057827244133,
      "grad_norm": 3.033630132675171,
      "learning_rate": 3.852880399766243e-05,
      "loss": 5.4362,
      "step": 119
    },
    {
      "epoch": 0.11613839825792402,
      "grad_norm": 3.13434100151062,
      "learning_rate": 3.772572564296005e-05,
      "loss": 5.5646,
      "step": 120
    },
    {
      "epoch": 0.11710621824340672,
      "grad_norm": 3.506950616836548,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 6.2031,
      "step": 121
    },
    {
      "epoch": 0.11807403822888943,
      "grad_norm": 3.3511886596679688,
      "learning_rate": 3.612985456190778e-05,
      "loss": 5.4874,
      "step": 122
    },
    {
      "epoch": 0.11904185821437213,
      "grad_norm": 3.6331543922424316,
      "learning_rate": 3.533749813077677e-05,
      "loss": 6.1683,
      "step": 123
    },
    {
      "epoch": 0.12000967819985482,
      "grad_norm": 3.1168124675750732,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 5.632,
      "step": 124
    },
    {
      "epoch": 0.12097749818533753,
      "grad_norm": 3.6517815589904785,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 5.8436,
      "step": 125
    },
    {
      "epoch": 0.12194531817082023,
      "grad_norm": 3.773366689682007,
      "learning_rate": 3.298534127791785e-05,
      "loss": 6.1431,
      "step": 126
    },
    {
      "epoch": 0.12291313815630293,
      "grad_norm": 3.5142552852630615,
      "learning_rate": 3.221030765387417e-05,
      "loss": 5.4809,
      "step": 127
    },
    {
      "epoch": 0.12388095814178562,
      "grad_norm": 3.4593284130096436,
      "learning_rate": 3.144013755408895e-05,
      "loss": 6.3835,
      "step": 128
    },
    {
      "epoch": 0.12484877812726833,
      "grad_norm": 3.3765344619750977,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 5.012,
      "step": 129
    },
    {
      "epoch": 0.12581659811275103,
      "grad_norm": 3.6987054347991943,
      "learning_rate": 2.991522876735154e-05,
      "loss": 5.3808,
      "step": 130
    },
    {
      "epoch": 0.12678441809823374,
      "grad_norm": 3.6943674087524414,
      "learning_rate": 2.916090697523549e-05,
      "loss": 5.8744,
      "step": 131
    },
    {
      "epoch": 0.12775223808371644,
      "grad_norm": 3.6965532302856445,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 5.6185,
      "step": 132
    },
    {
      "epoch": 0.12872005806919912,
      "grad_norm": 3.9772191047668457,
      "learning_rate": 2.766955965735968e-05,
      "loss": 6.216,
      "step": 133
    },
    {
      "epoch": 0.12968787805468182,
      "grad_norm": 3.817161798477173,
      "learning_rate": 2.693294185106562e-05,
      "loss": 5.969,
      "step": 134
    },
    {
      "epoch": 0.13065569804016453,
      "grad_norm": 4.254046440124512,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 5.9868,
      "step": 135
    },
    {
      "epoch": 0.13162351802564723,
      "grad_norm": 4.440921783447266,
      "learning_rate": 2.547882480847461e-05,
      "loss": 5.606,
      "step": 136
    },
    {
      "epoch": 0.13259133801112993,
      "grad_norm": 4.226521015167236,
      "learning_rate": 2.476172311325783e-05,
      "loss": 6.016,
      "step": 137
    },
    {
      "epoch": 0.13355915799661264,
      "grad_norm": 4.549649715423584,
      "learning_rate": 2.405152131093926e-05,
      "loss": 5.8866,
      "step": 138
    },
    {
      "epoch": 0.13452697798209534,
      "grad_norm": 4.75260591506958,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 5.8661,
      "step": 139
    },
    {
      "epoch": 0.13549479796757802,
      "grad_norm": 4.542957782745361,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 6.1137,
      "step": 140
    },
    {
      "epoch": 0.13646261795306072,
      "grad_norm": 4.724084377288818,
      "learning_rate": 2.196424713241637e-05,
      "loss": 5.776,
      "step": 141
    },
    {
      "epoch": 0.13743043793854343,
      "grad_norm": 5.398897647857666,
      "learning_rate": 2.128356686585282e-05,
      "loss": 5.752,
      "step": 142
    },
    {
      "epoch": 0.13839825792402613,
      "grad_norm": 5.055912017822266,
      "learning_rate": 2.061073738537635e-05,
      "loss": 5.7374,
      "step": 143
    },
    {
      "epoch": 0.13936607790950883,
      "grad_norm": 5.0658769607543945,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 6.0132,
      "step": 144
    },
    {
      "epoch": 0.14033389789499154,
      "grad_norm": 5.335988521575928,
      "learning_rate": 1.928936436551661e-05,
      "loss": 6.2013,
      "step": 145
    },
    {
      "epoch": 0.14130171788047424,
      "grad_norm": 5.163208484649658,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 6.1269,
      "step": 146
    },
    {
      "epoch": 0.14226953786595692,
      "grad_norm": 6.244584560394287,
      "learning_rate": 1.800157297483417e-05,
      "loss": 6.0443,
      "step": 147
    },
    {
      "epoch": 0.14323735785143962,
      "grad_norm": 6.147749423980713,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 5.9714,
      "step": 148
    },
    {
      "epoch": 0.14420517783692233,
      "grad_norm": 6.914193153381348,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 5.44,
      "step": 149
    },
    {
      "epoch": 0.14517299782240503,
      "grad_norm": 8.9144868850708,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 5.953,
      "step": 150
    },
    {
      "epoch": 0.14517299782240503,
      "eval_loss": 1.3589195013046265,
      "eval_runtime": 34.9514,
      "eval_samples_per_second": 49.783,
      "eval_steps_per_second": 12.446,
      "step": 150
    },
    {
      "epoch": 0.14614081780788774,
      "grad_norm": 2.910027027130127,
      "learning_rate": 1.553232954407171e-05,
      "loss": 4.4007,
      "step": 151
    },
    {
      "epoch": 0.14710863779337044,
      "grad_norm": 2.9558489322662354,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 4.2287,
      "step": 152
    },
    {
      "epoch": 0.14807645777885314,
      "grad_norm": 3.128788948059082,
      "learning_rate": 1.435357758543015e-05,
      "loss": 4.674,
      "step": 153
    },
    {
      "epoch": 0.14904427776433585,
      "grad_norm": 3.1798746585845947,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 5.1158,
      "step": 154
    },
    {
      "epoch": 0.15001209774981852,
      "grad_norm": 3.281080722808838,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 4.8103,
      "step": 155
    },
    {
      "epoch": 0.15097991773530123,
      "grad_norm": 3.0375945568084717,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 4.8034,
      "step": 156
    },
    {
      "epoch": 0.15194773772078393,
      "grad_norm": 2.9213650226593018,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 4.8919,
      "step": 157
    },
    {
      "epoch": 0.15291555770626664,
      "grad_norm": 2.911231279373169,
      "learning_rate": 1.157994445715706e-05,
      "loss": 5.304,
      "step": 158
    },
    {
      "epoch": 0.15388337769174934,
      "grad_norm": 2.8120620250701904,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 4.8839,
      "step": 159
    },
    {
      "epoch": 0.15485119767723204,
      "grad_norm": 3.228389263153076,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 5.1067,
      "step": 160
    },
    {
      "epoch": 0.15581901766271475,
      "grad_norm": 2.934567451477051,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 5.2231,
      "step": 161
    },
    {
      "epoch": 0.15678683764819742,
      "grad_norm": 3.126077175140381,
      "learning_rate": 9.549150281252633e-06,
      "loss": 5.1497,
      "step": 162
    },
    {
      "epoch": 0.15775465763368013,
      "grad_norm": 3.2123360633850098,
      "learning_rate": 9.068759265665384e-06,
      "loss": 5.7045,
      "step": 163
    },
    {
      "epoch": 0.15872247761916283,
      "grad_norm": 3.1366660594940186,
      "learning_rate": 8.599558442598998e-06,
      "loss": 5.3166,
      "step": 164
    },
    {
      "epoch": 0.15969029760464554,
      "grad_norm": 2.939225435256958,
      "learning_rate": 8.141676086873572e-06,
      "loss": 4.6186,
      "step": 165
    },
    {
      "epoch": 0.16065811759012824,
      "grad_norm": 2.8837945461273193,
      "learning_rate": 7.695237378953223e-06,
      "loss": 5.0911,
      "step": 166
    },
    {
      "epoch": 0.16162593757561095,
      "grad_norm": 2.9524667263031006,
      "learning_rate": 7.260364370723044e-06,
      "loss": 5.466,
      "step": 167
    },
    {
      "epoch": 0.16259375756109365,
      "grad_norm": 3.033395767211914,
      "learning_rate": 6.837175952121306e-06,
      "loss": 5.7277,
      "step": 168
    },
    {
      "epoch": 0.16356157754657633,
      "grad_norm": 3.249270439147949,
      "learning_rate": 6.425787818636131e-06,
      "loss": 5.5142,
      "step": 169
    },
    {
      "epoch": 0.16452939753205903,
      "grad_norm": 3.3505072593688965,
      "learning_rate": 6.026312439675552e-06,
      "loss": 5.3579,
      "step": 170
    },
    {
      "epoch": 0.16549721751754173,
      "grad_norm": 3.764362096786499,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 5.4691,
      "step": 171
    },
    {
      "epoch": 0.16646503750302444,
      "grad_norm": 3.434302568435669,
      "learning_rate": 5.263533508961827e-06,
      "loss": 5.2269,
      "step": 172
    },
    {
      "epoch": 0.16743285748850714,
      "grad_norm": 3.264943838119507,
      "learning_rate": 4.900438493352055e-06,
      "loss": 5.1627,
      "step": 173
    },
    {
      "epoch": 0.16840067747398985,
      "grad_norm": 3.4035701751708984,
      "learning_rate": 4.549673247541875e-06,
      "loss": 5.6517,
      "step": 174
    },
    {
      "epoch": 0.16936849745947255,
      "grad_norm": 3.662940740585327,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 5.6504,
      "step": 175
    },
    {
      "epoch": 0.17033631744495523,
      "grad_norm": 3.852566719055176,
      "learning_rate": 3.885512251130763e-06,
      "loss": 5.2973,
      "step": 176
    },
    {
      "epoch": 0.17130413743043793,
      "grad_norm": 3.80283784866333,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 5.9127,
      "step": 177
    },
    {
      "epoch": 0.17227195741592063,
      "grad_norm": 3.7050094604492188,
      "learning_rate": 3.271776770026963e-06,
      "loss": 6.0122,
      "step": 178
    },
    {
      "epoch": 0.17323977740140334,
      "grad_norm": 3.4881134033203125,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 5.6375,
      "step": 179
    },
    {
      "epoch": 0.17420759738688604,
      "grad_norm": 3.6493821144104004,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 5.7916,
      "step": 180
    },
    {
      "epoch": 0.17517541737236875,
      "grad_norm": 3.9559667110443115,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 6.0604,
      "step": 181
    },
    {
      "epoch": 0.17614323735785145,
      "grad_norm": 4.3142290115356445,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 5.232,
      "step": 182
    },
    {
      "epoch": 0.17711105734333413,
      "grad_norm": 3.8944382667541504,
      "learning_rate": 1.962316193157593e-06,
      "loss": 5.8326,
      "step": 183
    },
    {
      "epoch": 0.17807887732881683,
      "grad_norm": 4.235944747924805,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 5.7504,
      "step": 184
    },
    {
      "epoch": 0.17904669731429954,
      "grad_norm": 4.16559362411499,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 5.5269,
      "step": 185
    },
    {
      "epoch": 0.18001451729978224,
      "grad_norm": 4.6381611824035645,
      "learning_rate": 1.333670137599713e-06,
      "loss": 6.1976,
      "step": 186
    },
    {
      "epoch": 0.18098233728526494,
      "grad_norm": 4.65915584564209,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 5.9982,
      "step": 187
    },
    {
      "epoch": 0.18195015727074765,
      "grad_norm": 4.51572322845459,
      "learning_rate": 9.810017062595322e-07,
      "loss": 5.8186,
      "step": 188
    },
    {
      "epoch": 0.18291797725623035,
      "grad_norm": 4.5455427169799805,
      "learning_rate": 8.247462563808817e-07,
      "loss": 5.9113,
      "step": 189
    },
    {
      "epoch": 0.18388579724171303,
      "grad_norm": 4.83941125869751,
      "learning_rate": 6.819348298638839e-07,
      "loss": 5.9133,
      "step": 190
    },
    {
      "epoch": 0.18485361722719573,
      "grad_norm": 5.09579610824585,
      "learning_rate": 5.526064699265753e-07,
      "loss": 5.7195,
      "step": 191
    },
    {
      "epoch": 0.18582143721267844,
      "grad_norm": 5.278320789337158,
      "learning_rate": 4.367965336512403e-07,
      "loss": 5.4237,
      "step": 192
    },
    {
      "epoch": 0.18678925719816114,
      "grad_norm": 5.399392127990723,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 6.1929,
      "step": 193
    },
    {
      "epoch": 0.18775707718364384,
      "grad_norm": 5.9388957023620605,
      "learning_rate": 2.458548727494292e-07,
      "loss": 6.0689,
      "step": 194
    },
    {
      "epoch": 0.18872489716912655,
      "grad_norm": 6.151218891143799,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 5.5454,
      "step": 195
    },
    {
      "epoch": 0.18969271715460925,
      "grad_norm": 6.144274711608887,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 5.8326,
      "step": 196
    },
    {
      "epoch": 0.19066053714009193,
      "grad_norm": 6.6186747550964355,
      "learning_rate": 6.150154258476315e-08,
      "loss": 6.3905,
      "step": 197
    },
    {
      "epoch": 0.19162835712557463,
      "grad_norm": 6.73940372467041,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 5.9856,
      "step": 198
    },
    {
      "epoch": 0.19259617711105734,
      "grad_norm": 8.45466423034668,
      "learning_rate": 6.834750376549792e-09,
      "loss": 5.9808,
      "step": 199
    },
    {
      "epoch": 0.19356399709654004,
      "grad_norm": 10.36650276184082,
      "learning_rate": 0.0,
      "loss": 6.3752,
      "step": 200
    },
    {
      "epoch": 0.19356399709654004,
      "eval_loss": 1.3403037786483765,
      "eval_runtime": 35.0532,
      "eval_samples_per_second": 49.639,
      "eval_steps_per_second": 12.41,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.087009073581261e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}