{
  "best_metric": 3.3893699645996094,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.2074688796680498,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001037344398340249,
      "grad_norm": 3.6560280323028564,
      "learning_rate": 1e-05,
      "loss": 3.7424,
      "step": 1
    },
    {
      "epoch": 0.001037344398340249,
      "eval_loss": 5.154352188110352,
      "eval_runtime": 26.5688,
      "eval_samples_per_second": 61.124,
      "eval_steps_per_second": 15.281,
      "step": 1
    },
    {
      "epoch": 0.002074688796680498,
      "grad_norm": 4.519001007080078,
      "learning_rate": 2e-05,
      "loss": 4.0009,
      "step": 2
    },
    {
      "epoch": 0.0031120331950207467,
      "grad_norm": 4.717950820922852,
      "learning_rate": 3e-05,
      "loss": 3.8411,
      "step": 3
    },
    {
      "epoch": 0.004149377593360996,
      "grad_norm": 4.098848819732666,
      "learning_rate": 4e-05,
      "loss": 4.1416,
      "step": 4
    },
    {
      "epoch": 0.005186721991701245,
      "grad_norm": 3.9933223724365234,
      "learning_rate": 5e-05,
      "loss": 4.3214,
      "step": 5
    },
    {
      "epoch": 0.006224066390041493,
      "grad_norm": 3.811678171157837,
      "learning_rate": 6e-05,
      "loss": 4.0918,
      "step": 6
    },
    {
      "epoch": 0.007261410788381743,
      "grad_norm": 3.8618717193603516,
      "learning_rate": 7e-05,
      "loss": 4.1885,
      "step": 7
    },
    {
      "epoch": 0.008298755186721992,
      "grad_norm": 3.741856813430786,
      "learning_rate": 8e-05,
      "loss": 4.0593,
      "step": 8
    },
    {
      "epoch": 0.00933609958506224,
      "grad_norm": 3.6605193614959717,
      "learning_rate": 9e-05,
      "loss": 3.7919,
      "step": 9
    },
    {
      "epoch": 0.01037344398340249,
      "grad_norm": 4.069720268249512,
      "learning_rate": 0.0001,
      "loss": 4.0892,
      "step": 10
    },
    {
      "epoch": 0.011410788381742738,
      "grad_norm": 4.561852931976318,
      "learning_rate": 9.999316524962345e-05,
      "loss": 3.7796,
      "step": 11
    },
    {
      "epoch": 0.012448132780082987,
      "grad_norm": 4.809189796447754,
      "learning_rate": 9.997266286704631e-05,
      "loss": 3.9954,
      "step": 12
    },
    {
      "epoch": 0.013485477178423237,
      "grad_norm": 3.8352808952331543,
      "learning_rate": 9.993849845741524e-05,
      "loss": 3.9553,
      "step": 13
    },
    {
      "epoch": 0.014522821576763486,
      "grad_norm": 3.582754611968994,
      "learning_rate": 9.989068136093873e-05,
      "loss": 3.9918,
      "step": 14
    },
    {
      "epoch": 0.015560165975103735,
      "grad_norm": 3.632657766342163,
      "learning_rate": 9.98292246503335e-05,
      "loss": 3.5872,
      "step": 15
    },
    {
      "epoch": 0.016597510373443983,
      "grad_norm": 3.7454187870025635,
      "learning_rate": 9.975414512725057e-05,
      "loss": 4.036,
      "step": 16
    },
    {
      "epoch": 0.017634854771784232,
      "grad_norm": 3.546095132827759,
      "learning_rate": 9.966546331768191e-05,
      "loss": 3.297,
      "step": 17
    },
    {
      "epoch": 0.01867219917012448,
      "grad_norm": 3.9690802097320557,
      "learning_rate": 9.956320346634876e-05,
      "loss": 3.7145,
      "step": 18
    },
    {
      "epoch": 0.01970954356846473,
      "grad_norm": 3.569537878036499,
      "learning_rate": 9.944739353007344e-05,
      "loss": 4.0681,
      "step": 19
    },
    {
      "epoch": 0.02074688796680498,
      "grad_norm": 3.1518118381500244,
      "learning_rate": 9.931806517013612e-05,
      "loss": 3.6876,
      "step": 20
    },
    {
      "epoch": 0.021784232365145227,
      "grad_norm": 3.6860296726226807,
      "learning_rate": 9.917525374361912e-05,
      "loss": 3.7749,
      "step": 21
    },
    {
      "epoch": 0.022821576763485476,
      "grad_norm": 3.2175776958465576,
      "learning_rate": 9.901899829374047e-05,
      "loss": 3.6225,
      "step": 22
    },
    {
      "epoch": 0.023858921161825725,
      "grad_norm": 3.4466984272003174,
      "learning_rate": 9.884934153917997e-05,
      "loss": 3.7914,
      "step": 23
    },
    {
      "epoch": 0.024896265560165973,
      "grad_norm": 3.4754061698913574,
      "learning_rate": 9.86663298624003e-05,
      "loss": 3.7014,
      "step": 24
    },
    {
      "epoch": 0.025933609958506226,
      "grad_norm": 3.414550542831421,
      "learning_rate": 9.847001329696653e-05,
      "loss": 3.7435,
      "step": 25
    },
    {
      "epoch": 0.026970954356846474,
      "grad_norm": 3.7674219608306885,
      "learning_rate": 9.826044551386744e-05,
      "loss": 4.025,
      "step": 26
    },
    {
      "epoch": 0.028008298755186723,
      "grad_norm": 3.575585126876831,
      "learning_rate": 9.803768380684242e-05,
      "loss": 3.7974,
      "step": 27
    },
    {
      "epoch": 0.029045643153526972,
      "grad_norm": 3.6689274311065674,
      "learning_rate": 9.780178907671789e-05,
      "loss": 3.6685,
      "step": 28
    },
    {
      "epoch": 0.03008298755186722,
      "grad_norm": 3.6703617572784424,
      "learning_rate": 9.755282581475769e-05,
      "loss": 3.9441,
      "step": 29
    },
    {
      "epoch": 0.03112033195020747,
      "grad_norm": 3.904585123062134,
      "learning_rate": 9.729086208503174e-05,
      "loss": 4.0383,
      "step": 30
    },
    {
      "epoch": 0.032157676348547715,
      "grad_norm": 3.659148693084717,
      "learning_rate": 9.701596950580806e-05,
      "loss": 3.9392,
      "step": 31
    },
    {
      "epoch": 0.03319502074688797,
      "grad_norm": 3.6932144165039062,
      "learning_rate": 9.672822322997305e-05,
      "loss": 3.6377,
      "step": 32
    },
    {
      "epoch": 0.03423236514522822,
      "grad_norm": 3.7890701293945312,
      "learning_rate": 9.642770192448536e-05,
      "loss": 3.9012,
      "step": 33
    },
    {
      "epoch": 0.035269709543568464,
      "grad_norm": 4.021544933319092,
      "learning_rate": 9.611448774886924e-05,
      "loss": 3.5325,
      "step": 34
    },
    {
      "epoch": 0.03630705394190872,
      "grad_norm": 3.8901312351226807,
      "learning_rate": 9.578866633275288e-05,
      "loss": 3.8501,
      "step": 35
    },
    {
      "epoch": 0.03734439834024896,
      "grad_norm": 3.625171184539795,
      "learning_rate": 9.545032675245813e-05,
      "loss": 3.59,
      "step": 36
    },
    {
      "epoch": 0.038381742738589214,
      "grad_norm": 3.8629443645477295,
      "learning_rate": 9.509956150664796e-05,
      "loss": 4.0088,
      "step": 37
    },
    {
      "epoch": 0.03941908713692946,
      "grad_norm": 4.15440559387207,
      "learning_rate": 9.473646649103818e-05,
      "loss": 3.7416,
      "step": 38
    },
    {
      "epoch": 0.04045643153526971,
      "grad_norm": 3.913933515548706,
      "learning_rate": 9.43611409721806e-05,
      "loss": 3.9396,
      "step": 39
    },
    {
      "epoch": 0.04149377593360996,
      "grad_norm": 3.9919865131378174,
      "learning_rate": 9.397368756032445e-05,
      "loss": 3.4939,
      "step": 40
    },
    {
      "epoch": 0.04253112033195021,
      "grad_norm": 4.222021102905273,
      "learning_rate": 9.357421218136386e-05,
      "loss": 3.9626,
      "step": 41
    },
    {
      "epoch": 0.043568464730290454,
      "grad_norm": 3.666334867477417,
      "learning_rate": 9.316282404787871e-05,
      "loss": 3.8106,
      "step": 42
    },
    {
      "epoch": 0.044605809128630707,
      "grad_norm": 3.896925210952759,
      "learning_rate": 9.273963562927695e-05,
      "loss": 3.8724,
      "step": 43
    },
    {
      "epoch": 0.04564315352697095,
      "grad_norm": 4.4622578620910645,
      "learning_rate": 9.230476262104677e-05,
      "loss": 3.6229,
      "step": 44
    },
    {
      "epoch": 0.046680497925311204,
      "grad_norm": 4.379663944244385,
      "learning_rate": 9.185832391312644e-05,
      "loss": 3.6872,
      "step": 45
    },
    {
      "epoch": 0.04771784232365145,
      "grad_norm": 4.342223167419434,
      "learning_rate": 9.140044155740101e-05,
      "loss": 3.8005,
      "step": 46
    },
    {
      "epoch": 0.0487551867219917,
      "grad_norm": 5.052137851715088,
      "learning_rate": 9.093124073433463e-05,
      "loss": 3.8598,
      "step": 47
    },
    {
      "epoch": 0.04979253112033195,
      "grad_norm": 4.811618804931641,
      "learning_rate": 9.045084971874738e-05,
      "loss": 3.6151,
      "step": 48
    },
    {
      "epoch": 0.0508298755186722,
      "grad_norm": 5.3153300285339355,
      "learning_rate": 8.995939984474624e-05,
      "loss": 3.9977,
      "step": 49
    },
    {
      "epoch": 0.05186721991701245,
      "grad_norm": 5.884384632110596,
      "learning_rate": 8.945702546981969e-05,
      "loss": 3.5615,
      "step": 50
    },
    {
      "epoch": 0.05186721991701245,
      "eval_loss": 4.57713508605957,
      "eval_runtime": 26.5411,
      "eval_samples_per_second": 61.188,
      "eval_steps_per_second": 15.297,
      "step": 50
    },
    {
      "epoch": 0.052904564315352696,
      "grad_norm": 24.076704025268555,
      "learning_rate": 8.894386393810563e-05,
      "loss": 5.2672,
      "step": 51
    },
    {
      "epoch": 0.05394190871369295,
      "grad_norm": 19.754859924316406,
      "learning_rate": 8.842005554284296e-05,
      "loss": 4.7904,
      "step": 52
    },
    {
      "epoch": 0.054979253112033194,
      "grad_norm": 16.51922035217285,
      "learning_rate": 8.788574348801675e-05,
      "loss": 4.5515,
      "step": 53
    },
    {
      "epoch": 0.056016597510373446,
      "grad_norm": 13.558797836303711,
      "learning_rate": 8.73410738492077e-05,
      "loss": 4.4899,
      "step": 54
    },
    {
      "epoch": 0.05705394190871369,
      "grad_norm": 7.6642937660217285,
      "learning_rate": 8.678619553365659e-05,
      "loss": 4.2032,
      "step": 55
    },
    {
      "epoch": 0.058091286307053944,
      "grad_norm": 5.949044227600098,
      "learning_rate": 8.622126023955446e-05,
      "loss": 3.695,
      "step": 56
    },
    {
      "epoch": 0.05912863070539419,
      "grad_norm": 4.905073165893555,
      "learning_rate": 8.564642241456986e-05,
      "loss": 3.6473,
      "step": 57
    },
    {
      "epoch": 0.06016597510373444,
      "grad_norm": 4.315762519836426,
      "learning_rate": 8.506183921362443e-05,
      "loss": 3.4442,
      "step": 58
    },
    {
      "epoch": 0.061203319502074686,
      "grad_norm": 3.31467342376709,
      "learning_rate": 8.44676704559283e-05,
      "loss": 3.8328,
      "step": 59
    },
    {
      "epoch": 0.06224066390041494,
      "grad_norm": 2.8772637844085693,
      "learning_rate": 8.386407858128706e-05,
      "loss": 3.6724,
      "step": 60
    },
    {
      "epoch": 0.06327800829875518,
      "grad_norm": 2.7757973670959473,
      "learning_rate": 8.32512286056924e-05,
      "loss": 3.6571,
      "step": 61
    },
    {
      "epoch": 0.06431535269709543,
      "grad_norm": 2.6480956077575684,
      "learning_rate": 8.262928807620843e-05,
      "loss": 3.3738,
      "step": 62
    },
    {
      "epoch": 0.06535269709543569,
      "grad_norm": 2.5773894786834717,
      "learning_rate": 8.199842702516583e-05,
      "loss": 3.4269,
      "step": 63
    },
    {
      "epoch": 0.06639004149377593,
      "grad_norm": 2.5624706745147705,
      "learning_rate": 8.135881792367686e-05,
      "loss": 3.4427,
      "step": 64
    },
    {
      "epoch": 0.06742738589211618,
      "grad_norm": 2.816394329071045,
      "learning_rate": 8.07106356344834e-05,
      "loss": 3.6585,
      "step": 65
    },
    {
      "epoch": 0.06846473029045644,
      "grad_norm": 2.769735813140869,
      "learning_rate": 8.005405736415126e-05,
      "loss": 3.5136,
      "step": 66
    },
    {
      "epoch": 0.06950207468879668,
      "grad_norm": 2.795097827911377,
      "learning_rate": 7.938926261462366e-05,
      "loss": 3.4617,
      "step": 67
    },
    {
      "epoch": 0.07053941908713693,
      "grad_norm": 2.583852529525757,
      "learning_rate": 7.871643313414718e-05,
      "loss": 3.3261,
      "step": 68
    },
    {
      "epoch": 0.07157676348547717,
      "grad_norm": 2.535404920578003,
      "learning_rate": 7.803575286758364e-05,
      "loss": 3.388,
      "step": 69
    },
    {
      "epoch": 0.07261410788381743,
      "grad_norm": 2.6295790672302246,
      "learning_rate": 7.734740790612136e-05,
      "loss": 3.807,
      "step": 70
    },
    {
      "epoch": 0.07365145228215768,
      "grad_norm": 2.7374913692474365,
      "learning_rate": 7.66515864363997e-05,
      "loss": 3.3073,
      "step": 71
    },
    {
      "epoch": 0.07468879668049792,
      "grad_norm": 2.939117908477783,
      "learning_rate": 7.594847868906076e-05,
      "loss": 3.4621,
      "step": 72
    },
    {
      "epoch": 0.07572614107883817,
      "grad_norm": 2.762972831726074,
      "learning_rate": 7.52382768867422e-05,
      "loss": 3.4,
      "step": 73
    },
    {
      "epoch": 0.07676348547717843,
      "grad_norm": 2.9580655097961426,
      "learning_rate": 7.452117519152542e-05,
      "loss": 3.5031,
      "step": 74
    },
    {
      "epoch": 0.07780082987551867,
      "grad_norm": 2.899890661239624,
      "learning_rate": 7.379736965185368e-05,
      "loss": 3.3536,
      "step": 75
    },
    {
      "epoch": 0.07883817427385892,
      "grad_norm": 2.7998228073120117,
      "learning_rate": 7.30670581489344e-05,
      "loss": 3.7074,
      "step": 76
    },
    {
      "epoch": 0.07987551867219916,
      "grad_norm": 3.0808844566345215,
      "learning_rate": 7.233044034264034e-05,
      "loss": 3.3757,
      "step": 77
    },
    {
      "epoch": 0.08091286307053942,
      "grad_norm": 3.0572876930236816,
      "learning_rate": 7.158771761692464e-05,
      "loss": 3.4954,
      "step": 78
    },
    {
      "epoch": 0.08195020746887967,
      "grad_norm": 2.852540969848633,
      "learning_rate": 7.083909302476453e-05,
      "loss": 3.2891,
      "step": 79
    },
    {
      "epoch": 0.08298755186721991,
      "grad_norm": 3.0258772373199463,
      "learning_rate": 7.008477123264848e-05,
      "loss": 3.6323,
      "step": 80
    },
    {
      "epoch": 0.08402489626556017,
      "grad_norm": 3.0334479808807373,
      "learning_rate": 6.932495846462261e-05,
      "loss": 3.6048,
      "step": 81
    },
    {
      "epoch": 0.08506224066390042,
      "grad_norm": 3.7473435401916504,
      "learning_rate": 6.855986244591104e-05,
      "loss": 3.4813,
      "step": 82
    },
    {
      "epoch": 0.08609958506224066,
      "grad_norm": 3.568814277648926,
      "learning_rate": 6.778969234612584e-05,
      "loss": 3.747,
      "step": 83
    },
    {
      "epoch": 0.08713692946058091,
      "grad_norm": 3.2257232666015625,
      "learning_rate": 6.701465872208216e-05,
      "loss": 3.2229,
      "step": 84
    },
    {
      "epoch": 0.08817427385892117,
      "grad_norm": 3.3203718662261963,
      "learning_rate": 6.623497346023418e-05,
      "loss": 3.5216,
      "step": 85
    },
    {
      "epoch": 0.08921161825726141,
      "grad_norm": 3.708003520965576,
      "learning_rate": 6.545084971874738e-05,
      "loss": 3.293,
      "step": 86
    },
    {
      "epoch": 0.09024896265560166,
      "grad_norm": 3.2470216751098633,
      "learning_rate": 6.466250186922325e-05,
      "loss": 3.5909,
      "step": 87
    },
    {
      "epoch": 0.0912863070539419,
      "grad_norm": 3.2357749938964844,
      "learning_rate": 6.387014543809223e-05,
      "loss": 3.3392,
      "step": 88
    },
    {
      "epoch": 0.09232365145228216,
      "grad_norm": 3.4167912006378174,
      "learning_rate": 6.307399704769099e-05,
      "loss": 3.4653,
      "step": 89
    },
    {
      "epoch": 0.09336099585062241,
      "grad_norm": 3.372326135635376,
      "learning_rate": 6.227427435703997e-05,
      "loss": 3.3182,
      "step": 90
    },
    {
      "epoch": 0.09439834024896265,
      "grad_norm": 3.4928457736968994,
      "learning_rate": 6.147119600233758e-05,
      "loss": 3.5979,
      "step": 91
    },
    {
      "epoch": 0.0954356846473029,
      "grad_norm": 3.7030575275421143,
      "learning_rate": 6.066498153718735e-05,
      "loss": 3.8812,
      "step": 92
    },
    {
      "epoch": 0.09647302904564316,
      "grad_norm": 3.790165662765503,
      "learning_rate": 5.985585137257401e-05,
      "loss": 3.8599,
      "step": 93
    },
    {
      "epoch": 0.0975103734439834,
      "grad_norm": 3.600940465927124,
      "learning_rate": 5.90440267166055e-05,
      "loss": 3.5759,
      "step": 94
    },
    {
      "epoch": 0.09854771784232365,
      "grad_norm": 3.944754123687744,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 3.8586,
      "step": 95
    },
    {
      "epoch": 0.0995850622406639,
      "grad_norm": 3.779947280883789,
      "learning_rate": 5.74131823855921e-05,
      "loss": 3.3266,
      "step": 96
    },
    {
      "epoch": 0.10062240663900415,
      "grad_norm": 4.06245756149292,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 3.5335,
      "step": 97
    },
    {
      "epoch": 0.1016597510373444,
      "grad_norm": 4.446831226348877,
      "learning_rate": 5.577423184847932e-05,
      "loss": 3.5554,
      "step": 98
    },
    {
      "epoch": 0.10269709543568464,
      "grad_norm": 4.253654479980469,
      "learning_rate": 5.495227651252315e-05,
      "loss": 3.2417,
      "step": 99
    },
    {
      "epoch": 0.1037344398340249,
      "grad_norm": 5.696987628936768,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 3.9678,
      "step": 100
    },
    {
      "epoch": 0.1037344398340249,
      "eval_loss": 3.7914316654205322,
      "eval_runtime": 26.5428,
      "eval_samples_per_second": 61.184,
      "eval_steps_per_second": 15.296,
      "step": 100
    },
    {
      "epoch": 0.10477178423236515,
      "grad_norm": 8.166366577148438,
      "learning_rate": 5.330452921628497e-05,
      "loss": 3.8268,
      "step": 101
    },
    {
      "epoch": 0.10580912863070539,
      "grad_norm": 6.857130527496338,
      "learning_rate": 5.247918773366112e-05,
      "loss": 3.7574,
      "step": 102
    },
    {
      "epoch": 0.10684647302904564,
      "grad_norm": 6.688436985015869,
      "learning_rate": 5.165316846586541e-05,
      "loss": 3.7533,
      "step": 103
    },
    {
      "epoch": 0.1078838174273859,
      "grad_norm": 6.446659564971924,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 3.9865,
      "step": 104
    },
    {
      "epoch": 0.10892116182572614,
      "grad_norm": 5.080942153930664,
      "learning_rate": 5e-05,
      "loss": 3.4969,
      "step": 105
    },
    {
      "epoch": 0.10995850622406639,
      "grad_norm": 4.44970703125,
      "learning_rate": 4.917330276168208e-05,
      "loss": 3.5119,
      "step": 106
    },
    {
      "epoch": 0.11099585062240663,
      "grad_norm": 3.998990058898926,
      "learning_rate": 4.834683153413459e-05,
      "loss": 3.6256,
      "step": 107
    },
    {
      "epoch": 0.11203319502074689,
      "grad_norm": 3.2981390953063965,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 3.4402,
      "step": 108
    },
    {
      "epoch": 0.11307053941908714,
      "grad_norm": 3.044083833694458,
      "learning_rate": 4.669547078371504e-05,
      "loss": 3.5525,
      "step": 109
    },
    {
      "epoch": 0.11410788381742738,
      "grad_norm": 2.891031503677368,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 3.1974,
      "step": 110
    },
    {
      "epoch": 0.11514522821576763,
      "grad_norm": 2.743988037109375,
      "learning_rate": 4.504772348747687e-05,
      "loss": 3.2082,
      "step": 111
    },
    {
      "epoch": 0.11618257261410789,
      "grad_norm": 2.7417633533477783,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 3.0046,
      "step": 112
    },
    {
      "epoch": 0.11721991701244813,
      "grad_norm": 2.6141929626464844,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 3.1606,
      "step": 113
    },
    {
      "epoch": 0.11825726141078838,
      "grad_norm": 2.5122275352478027,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 3.4908,
      "step": 114
    },
    {
      "epoch": 0.11929460580912864,
      "grad_norm": 2.8065576553344727,
      "learning_rate": 4.17702704859633e-05,
      "loss": 3.5437,
      "step": 115
    },
    {
      "epoch": 0.12033195020746888,
      "grad_norm": 2.443136692047119,
      "learning_rate": 4.095597328339452e-05,
      "loss": 3.4685,
      "step": 116
    },
    {
      "epoch": 0.12136929460580913,
      "grad_norm": 2.5843472480773926,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 3.319,
      "step": 117
    },
    {
      "epoch": 0.12240663900414937,
      "grad_norm": 2.6064820289611816,
      "learning_rate": 3.933501846281267e-05,
      "loss": 3.3902,
      "step": 118
    },
    {
      "epoch": 0.12344398340248963,
      "grad_norm": 2.74259352684021,
      "learning_rate": 3.852880399766243e-05,
      "loss": 3.81,
      "step": 119
    },
    {
      "epoch": 0.12448132780082988,
      "grad_norm": 2.7329461574554443,
      "learning_rate": 3.772572564296005e-05,
      "loss": 3.4747,
      "step": 120
    },
    {
      "epoch": 0.12551867219917012,
      "grad_norm": 2.7343878746032715,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 3.5092,
      "step": 121
    },
    {
      "epoch": 0.12655601659751037,
      "grad_norm": 2.749302625656128,
      "learning_rate": 3.612985456190778e-05,
      "loss": 3.8106,
      "step": 122
    },
    {
      "epoch": 0.1275933609958506,
      "grad_norm": 3.006464719772339,
      "learning_rate": 3.533749813077677e-05,
      "loss": 3.6465,
      "step": 123
    },
    {
      "epoch": 0.12863070539419086,
      "grad_norm": 2.8483219146728516,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 3.3153,
      "step": 124
    },
    {
      "epoch": 0.12966804979253113,
      "grad_norm": 3.621854782104492,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 3.0212,
      "step": 125
    },
    {
      "epoch": 0.13070539419087138,
      "grad_norm": 2.9148917198181152,
      "learning_rate": 3.298534127791785e-05,
      "loss": 3.2572,
      "step": 126
    },
    {
      "epoch": 0.13174273858921162,
      "grad_norm": 2.9206199645996094,
      "learning_rate": 3.221030765387417e-05,
      "loss": 3.4323,
      "step": 127
    },
    {
      "epoch": 0.13278008298755187,
      "grad_norm": 3.293531656265259,
      "learning_rate": 3.144013755408895e-05,
      "loss": 3.4568,
      "step": 128
    },
    {
      "epoch": 0.1338174273858921,
      "grad_norm": 2.9713733196258545,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 3.4724,
      "step": 129
    },
    {
      "epoch": 0.13485477178423236,
      "grad_norm": 3.215953826904297,
      "learning_rate": 2.991522876735154e-05,
      "loss": 3.6332,
      "step": 130
    },
    {
      "epoch": 0.1358921161825726,
      "grad_norm": 3.0868947505950928,
      "learning_rate": 2.916090697523549e-05,
      "loss": 3.4988,
      "step": 131
    },
    {
      "epoch": 0.13692946058091288,
      "grad_norm": 3.1920242309570312,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 3.3852,
      "step": 132
    },
    {
      "epoch": 0.13796680497925312,
      "grad_norm": 3.4271225929260254,
      "learning_rate": 2.766955965735968e-05,
      "loss": 3.8709,
      "step": 133
    },
    {
      "epoch": 0.13900414937759337,
      "grad_norm": 3.4510085582733154,
      "learning_rate": 2.693294185106562e-05,
      "loss": 3.4216,
      "step": 134
    },
    {
      "epoch": 0.1400414937759336,
      "grad_norm": 3.2544167041778564,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 3.5809,
      "step": 135
    },
    {
      "epoch": 0.14107883817427386,
      "grad_norm": 3.22052001953125,
      "learning_rate": 2.547882480847461e-05,
      "loss": 3.2472,
      "step": 136
    },
    {
      "epoch": 0.1421161825726141,
      "grad_norm": 3.308317184448242,
      "learning_rate": 2.476172311325783e-05,
      "loss": 3.5626,
      "step": 137
    },
    {
      "epoch": 0.14315352697095435,
      "grad_norm": 3.3772473335266113,
      "learning_rate": 2.405152131093926e-05,
      "loss": 3.2607,
      "step": 138
    },
    {
      "epoch": 0.1441908713692946,
      "grad_norm": 3.1532044410705566,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 3.5309,
      "step": 139
    },
    {
      "epoch": 0.14522821576763487,
      "grad_norm": 3.440068006515503,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 3.5348,
      "step": 140
    },
    {
      "epoch": 0.1462655601659751,
      "grad_norm": 3.2612829208374023,
      "learning_rate": 2.196424713241637e-05,
      "loss": 3.5397,
      "step": 141
    },
    {
      "epoch": 0.14730290456431536,
      "grad_norm": 3.7282817363739014,
      "learning_rate": 2.128356686585282e-05,
      "loss": 3.7273,
      "step": 142
    },
    {
      "epoch": 0.1483402489626556,
      "grad_norm": 3.4528229236602783,
      "learning_rate": 2.061073738537635e-05,
      "loss": 3.3793,
      "step": 143
    },
    {
      "epoch": 0.14937759336099585,
      "grad_norm": 3.423583507537842,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 3.2933,
      "step": 144
    },
    {
      "epoch": 0.1504149377593361,
      "grad_norm": 3.6867237091064453,
      "learning_rate": 1.928936436551661e-05,
      "loss": 3.8415,
      "step": 145
    },
    {
      "epoch": 0.15145228215767634,
      "grad_norm": 4.240562438964844,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 3.9054,
      "step": 146
    },
    {
      "epoch": 0.1524896265560166,
      "grad_norm": 4.737523078918457,
      "learning_rate": 1.800157297483417e-05,
      "loss": 3.7616,
      "step": 147
    },
    {
      "epoch": 0.15352697095435686,
      "grad_norm": 4.262228965759277,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 3.6418,
      "step": 148
    },
    {
      "epoch": 0.1545643153526971,
      "grad_norm": 4.715056896209717,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 3.796,
      "step": 149
    },
    {
      "epoch": 0.15560165975103735,
      "grad_norm": 6.44474983215332,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 4.0393,
      "step": 150
    },
    {
      "epoch": 0.15560165975103735,
      "eval_loss": 3.443917751312256,
      "eval_runtime": 26.554,
      "eval_samples_per_second": 61.158,
      "eval_steps_per_second": 15.29,
      "step": 150
    },
    {
      "epoch": 0.1566390041493776,
      "grad_norm": 2.9477157592773438,
      "learning_rate": 1.553232954407171e-05,
      "loss": 3.157,
      "step": 151
    },
    {
      "epoch": 0.15767634854771784,
      "grad_norm": 3.1337876319885254,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 3.1213,
      "step": 152
    },
    {
      "epoch": 0.15871369294605808,
      "grad_norm": 3.1577205657958984,
      "learning_rate": 1.435357758543015e-05,
      "loss": 3.4105,
      "step": 153
    },
    {
      "epoch": 0.15975103734439833,
      "grad_norm": 3.00323748588562,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 3.2181,
      "step": 154
    },
    {
      "epoch": 0.1607883817427386,
      "grad_norm": 2.866039514541626,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 3.2601,
      "step": 155
    },
    {
      "epoch": 0.16182572614107885,
      "grad_norm": 3.097904920578003,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 3.1206,
      "step": 156
    },
    {
      "epoch": 0.1628630705394191,
      "grad_norm": 3.0080816745758057,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 3.464,
      "step": 157
    },
    {
      "epoch": 0.16390041493775934,
      "grad_norm": 3.1983859539031982,
      "learning_rate": 1.157994445715706e-05,
      "loss": 3.5615,
      "step": 158
    },
    {
      "epoch": 0.16493775933609958,
      "grad_norm": 2.9502131938934326,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 3.3529,
      "step": 159
    },
    {
      "epoch": 0.16597510373443983,
      "grad_norm": 3.120685577392578,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 3.4895,
      "step": 160
    },
    {
      "epoch": 0.16701244813278007,
      "grad_norm": 2.6355931758880615,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 3.1231,
      "step": 161
    },
    {
      "epoch": 0.16804979253112035,
      "grad_norm": 2.8759355545043945,
      "learning_rate": 9.549150281252633e-06,
      "loss": 3.3021,
      "step": 162
    },
    {
      "epoch": 0.1690871369294606,
      "grad_norm": 2.9746170043945312,
      "learning_rate": 9.068759265665384e-06,
      "loss": 3.3881,
      "step": 163
    },
    {
      "epoch": 0.17012448132780084,
      "grad_norm": 2.8339903354644775,
      "learning_rate": 8.599558442598998e-06,
      "loss": 3.2455,
      "step": 164
    },
    {
      "epoch": 0.17116182572614108,
      "grad_norm": 2.7609705924987793,
      "learning_rate": 8.141676086873572e-06,
      "loss": 3.108,
      "step": 165
    },
    {
      "epoch": 0.17219917012448133,
      "grad_norm": 3.246521472930908,
      "learning_rate": 7.695237378953223e-06,
      "loss": 3.6929,
      "step": 166
    },
    {
      "epoch": 0.17323651452282157,
      "grad_norm": 2.688962459564209,
      "learning_rate": 7.260364370723044e-06,
      "loss": 3.138,
      "step": 167
    },
    {
      "epoch": 0.17427385892116182,
      "grad_norm": 2.9067366123199463,
      "learning_rate": 6.837175952121306e-06,
      "loss": 3.718,
      "step": 168
    },
    {
      "epoch": 0.17531120331950206,
      "grad_norm": 2.7512147426605225,
      "learning_rate": 6.425787818636131e-06,
      "loss": 3.4175,
      "step": 169
    },
    {
      "epoch": 0.17634854771784234,
      "grad_norm": 2.548328399658203,
      "learning_rate": 6.026312439675552e-06,
      "loss": 3.1292,
      "step": 170
    },
    {
      "epoch": 0.17738589211618258,
      "grad_norm": 2.8721537590026855,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 3.5308,
      "step": 171
    },
    {
      "epoch": 0.17842323651452283,
      "grad_norm": 3.4221110343933105,
      "learning_rate": 5.263533508961827e-06,
      "loss": 3.8456,
      "step": 172
    },
    {
      "epoch": 0.17946058091286307,
      "grad_norm": 3.0033349990844727,
      "learning_rate": 4.900438493352055e-06,
      "loss": 3.4229,
      "step": 173
    },
    {
      "epoch": 0.18049792531120332,
      "grad_norm": 2.841123342514038,
      "learning_rate": 4.549673247541875e-06,
      "loss": 3.4928,
      "step": 174
    },
    {
      "epoch": 0.18153526970954356,
      "grad_norm": 2.7495405673980713,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 3.2769,
      "step": 175
    },
    {
      "epoch": 0.1825726141078838,
      "grad_norm": 3.1454813480377197,
      "learning_rate": 3.885512251130763e-06,
      "loss": 3.1546,
      "step": 176
    },
    {
      "epoch": 0.18360995850622408,
      "grad_norm": 2.9140090942382812,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 3.2434,
      "step": 177
    },
    {
      "epoch": 0.18464730290456433,
      "grad_norm": 2.936655044555664,
      "learning_rate": 3.271776770026963e-06,
      "loss": 3.7009,
      "step": 178
    },
    {
      "epoch": 0.18568464730290457,
      "grad_norm": 2.8600358963012695,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 2.9847,
      "step": 179
    },
    {
      "epoch": 0.18672199170124482,
      "grad_norm": 3.1558871269226074,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 3.9366,
      "step": 180
    },
    {
      "epoch": 0.18775933609958506,
      "grad_norm": 2.904458522796631,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 2.9716,
      "step": 181
    },
    {
      "epoch": 0.1887966804979253,
      "grad_norm": 3.277076005935669,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 3.3847,
      "step": 182
    },
    {
      "epoch": 0.18983402489626555,
      "grad_norm": 3.215095281600952,
      "learning_rate": 1.962316193157593e-06,
      "loss": 3.4074,
      "step": 183
    },
    {
      "epoch": 0.1908713692946058,
      "grad_norm": 3.3069441318511963,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 3.3479,
      "step": 184
    },
    {
      "epoch": 0.19190871369294607,
      "grad_norm": 3.70759916305542,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 3.5217,
      "step": 185
    },
    {
      "epoch": 0.19294605809128632,
      "grad_norm": 3.380190849304199,
      "learning_rate": 1.333670137599713e-06,
      "loss": 3.4685,
      "step": 186
    },
    {
      "epoch": 0.19398340248962656,
      "grad_norm": 3.148998498916626,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 3.644,
      "step": 187
    },
    {
      "epoch": 0.1950207468879668,
      "grad_norm": 3.2292802333831787,
      "learning_rate": 9.810017062595322e-07,
      "loss": 3.4821,
      "step": 188
    },
    {
      "epoch": 0.19605809128630705,
      "grad_norm": 3.3793811798095703,
      "learning_rate": 8.247462563808817e-07,
      "loss": 3.3928,
      "step": 189
    },
    {
      "epoch": 0.1970954356846473,
      "grad_norm": 3.4184722900390625,
      "learning_rate": 6.819348298638839e-07,
      "loss": 3.7607,
      "step": 190
    },
    {
      "epoch": 0.19813278008298754,
      "grad_norm": 3.5667853355407715,
      "learning_rate": 5.526064699265753e-07,
      "loss": 3.7305,
      "step": 191
    },
    {
      "epoch": 0.1991701244813278,
      "grad_norm": 3.350822687149048,
      "learning_rate": 4.367965336512403e-07,
      "loss": 3.2065,
      "step": 192
    },
    {
      "epoch": 0.20020746887966806,
      "grad_norm": 3.8035242557525635,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 3.5918,
      "step": 193
    },
    {
      "epoch": 0.2012448132780083,
      "grad_norm": 3.9756436347961426,
      "learning_rate": 2.458548727494292e-07,
      "loss": 3.7081,
      "step": 194
    },
    {
      "epoch": 0.20228215767634855,
      "grad_norm": 4.188309192657471,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 3.7841,
      "step": 195
    },
    {
      "epoch": 0.2033195020746888,
      "grad_norm": 4.394107818603516,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 3.6746,
      "step": 196
    },
    {
      "epoch": 0.20435684647302904,
      "grad_norm": 4.623396396636963,
      "learning_rate": 6.150154258476315e-08,
      "loss": 3.8618,
      "step": 197
    },
    {
      "epoch": 0.2053941908713693,
      "grad_norm": 4.6601762771606445,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 3.3425,
      "step": 198
    },
    {
      "epoch": 0.20643153526970953,
      "grad_norm": 5.034449577331543,
      "learning_rate": 6.834750376549792e-09,
      "loss": 3.9177,
      "step": 199
    },
    {
      "epoch": 0.2074688796680498,
      "grad_norm": 7.078446388244629,
      "learning_rate": 0.0,
      "loss": 4.6096,
      "step": 200
    },
    {
      "epoch": 0.2074688796680498,
      "eval_loss": 3.3893699645996094,
      "eval_runtime": 26.548,
      "eval_samples_per_second": 61.172,
      "eval_steps_per_second": 15.293,
      "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": 1.3317552734208e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}