{
  "best_metric": 2.3479816913604736,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.6650041562759768,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0033250207813798837,
      "grad_norm": 0.10728424787521362,
      "learning_rate": 1e-05,
      "loss": 1.8197,
      "step": 1
    },
    {
      "epoch": 0.0033250207813798837,
      "eval_loss": 3.2351901531219482,
      "eval_runtime": 36.1822,
      "eval_samples_per_second": 14.012,
      "eval_steps_per_second": 3.51,
      "step": 1
    },
    {
      "epoch": 0.006650041562759767,
      "grad_norm": 0.1415768712759018,
      "learning_rate": 2e-05,
      "loss": 1.7003,
      "step": 2
    },
    {
      "epoch": 0.00997506234413965,
      "grad_norm": 0.17956018447875977,
      "learning_rate": 3e-05,
      "loss": 1.6362,
      "step": 3
    },
    {
      "epoch": 0.013300083125519535,
      "grad_norm": 0.22796227037906647,
      "learning_rate": 4e-05,
      "loss": 1.2192,
      "step": 4
    },
    {
      "epoch": 0.01662510390689942,
      "grad_norm": 0.20629073679447174,
      "learning_rate": 5e-05,
      "loss": 1.2319,
      "step": 5
    },
    {
      "epoch": 0.0199501246882793,
      "grad_norm": 0.26256290078163147,
      "learning_rate": 6e-05,
      "loss": 1.2894,
      "step": 6
    },
    {
      "epoch": 0.023275145469659187,
      "grad_norm": 0.25810831785202026,
      "learning_rate": 7e-05,
      "loss": 1.5151,
      "step": 7
    },
    {
      "epoch": 0.02660016625103907,
      "grad_norm": 0.25480204820632935,
      "learning_rate": 8e-05,
      "loss": 1.5695,
      "step": 8
    },
    {
      "epoch": 0.029925187032418952,
      "grad_norm": 0.2788578271865845,
      "learning_rate": 9e-05,
      "loss": 1.6727,
      "step": 9
    },
    {
      "epoch": 0.03325020781379884,
      "grad_norm": 0.271418958902359,
      "learning_rate": 0.0001,
      "loss": 1.6935,
      "step": 10
    },
    {
      "epoch": 0.03657522859517872,
      "grad_norm": 0.33443188667297363,
      "learning_rate": 9.999316524962345e-05,
      "loss": 1.9383,
      "step": 11
    },
    {
      "epoch": 0.0399002493765586,
      "grad_norm": 0.38337311148643494,
      "learning_rate": 9.997266286704631e-05,
      "loss": 1.5992,
      "step": 12
    },
    {
      "epoch": 0.043225270157938485,
      "grad_norm": 0.45374158024787903,
      "learning_rate": 9.993849845741524e-05,
      "loss": 1.8608,
      "step": 13
    },
    {
      "epoch": 0.046550290939318374,
      "grad_norm": 0.4392660856246948,
      "learning_rate": 9.989068136093873e-05,
      "loss": 1.8438,
      "step": 14
    },
    {
      "epoch": 0.04987531172069826,
      "grad_norm": 0.43151795864105225,
      "learning_rate": 9.98292246503335e-05,
      "loss": 1.7688,
      "step": 15
    },
    {
      "epoch": 0.05320033250207814,
      "grad_norm": 0.4266657829284668,
      "learning_rate": 9.975414512725057e-05,
      "loss": 1.8316,
      "step": 16
    },
    {
      "epoch": 0.05652535328345802,
      "grad_norm": 0.5022886991500854,
      "learning_rate": 9.966546331768191e-05,
      "loss": 1.8342,
      "step": 17
    },
    {
      "epoch": 0.059850374064837904,
      "grad_norm": 0.45997345447540283,
      "learning_rate": 9.956320346634876e-05,
      "loss": 1.8418,
      "step": 18
    },
    {
      "epoch": 0.06317539484621779,
      "grad_norm": 0.4764719605445862,
      "learning_rate": 9.944739353007344e-05,
      "loss": 2.2292,
      "step": 19
    },
    {
      "epoch": 0.06650041562759768,
      "grad_norm": 0.4928871691226959,
      "learning_rate": 9.931806517013612e-05,
      "loss": 1.9597,
      "step": 20
    },
    {
      "epoch": 0.06982543640897755,
      "grad_norm": 0.5077341198921204,
      "learning_rate": 9.917525374361912e-05,
      "loss": 2.2398,
      "step": 21
    },
    {
      "epoch": 0.07315045719035744,
      "grad_norm": 0.5359517335891724,
      "learning_rate": 9.901899829374047e-05,
      "loss": 2.198,
      "step": 22
    },
    {
      "epoch": 0.07647547797173733,
      "grad_norm": 0.5895940661430359,
      "learning_rate": 9.884934153917997e-05,
      "loss": 2.4338,
      "step": 23
    },
    {
      "epoch": 0.0798004987531172,
      "grad_norm": 0.565660297870636,
      "learning_rate": 9.86663298624003e-05,
      "loss": 2.2223,
      "step": 24
    },
    {
      "epoch": 0.0831255195344971,
      "grad_norm": 0.5687828660011292,
      "learning_rate": 9.847001329696653e-05,
      "loss": 2.4689,
      "step": 25
    },
    {
      "epoch": 0.08645054031587697,
      "grad_norm": 0.6717599034309387,
      "learning_rate": 9.826044551386744e-05,
      "loss": 2.6174,
      "step": 26
    },
    {
      "epoch": 0.08977556109725686,
      "grad_norm": 0.6361708641052246,
      "learning_rate": 9.803768380684242e-05,
      "loss": 2.5065,
      "step": 27
    },
    {
      "epoch": 0.09310058187863675,
      "grad_norm": 0.7013790011405945,
      "learning_rate": 9.780178907671789e-05,
      "loss": 2.5903,
      "step": 28
    },
    {
      "epoch": 0.09642560266001662,
      "grad_norm": 0.8236458897590637,
      "learning_rate": 9.755282581475769e-05,
      "loss": 2.4222,
      "step": 29
    },
    {
      "epoch": 0.09975062344139651,
      "grad_norm": 0.7173617482185364,
      "learning_rate": 9.729086208503174e-05,
      "loss": 2.1556,
      "step": 30
    },
    {
      "epoch": 0.10307564422277639,
      "grad_norm": 0.8104385137557983,
      "learning_rate": 9.701596950580806e-05,
      "loss": 2.6324,
      "step": 31
    },
    {
      "epoch": 0.10640066500415628,
      "grad_norm": 0.7957509160041809,
      "learning_rate": 9.672822322997305e-05,
      "loss": 2.4198,
      "step": 32
    },
    {
      "epoch": 0.10972568578553615,
      "grad_norm": 0.9327613115310669,
      "learning_rate": 9.642770192448536e-05,
      "loss": 2.6106,
      "step": 33
    },
    {
      "epoch": 0.11305070656691604,
      "grad_norm": 1.2241650819778442,
      "learning_rate": 9.611448774886924e-05,
      "loss": 2.6662,
      "step": 34
    },
    {
      "epoch": 0.11637572734829593,
      "grad_norm": 0.9784499406814575,
      "learning_rate": 9.578866633275288e-05,
      "loss": 2.4354,
      "step": 35
    },
    {
      "epoch": 0.11970074812967581,
      "grad_norm": 1.0117332935333252,
      "learning_rate": 9.545032675245813e-05,
      "loss": 2.416,
      "step": 36
    },
    {
      "epoch": 0.1230257689110557,
      "grad_norm": 0.942400336265564,
      "learning_rate": 9.509956150664796e-05,
      "loss": 2.8094,
      "step": 37
    },
    {
      "epoch": 0.12635078969243557,
      "grad_norm": 0.9462390542030334,
      "learning_rate": 9.473646649103818e-05,
      "loss": 2.6323,
      "step": 38
    },
    {
      "epoch": 0.12967581047381546,
      "grad_norm": 1.277146577835083,
      "learning_rate": 9.43611409721806e-05,
      "loss": 2.9293,
      "step": 39
    },
    {
      "epoch": 0.13300083125519535,
      "grad_norm": 1.4511051177978516,
      "learning_rate": 9.397368756032445e-05,
      "loss": 2.6989,
      "step": 40
    },
    {
      "epoch": 0.13632585203657524,
      "grad_norm": 1.2115994691848755,
      "learning_rate": 9.357421218136386e-05,
      "loss": 3.1644,
      "step": 41
    },
    {
      "epoch": 0.1396508728179551,
      "grad_norm": 1.1447230577468872,
      "learning_rate": 9.316282404787871e-05,
      "loss": 2.9856,
      "step": 42
    },
    {
      "epoch": 0.142975893599335,
      "grad_norm": 1.1327815055847168,
      "learning_rate": 9.273963562927695e-05,
      "loss": 2.6263,
      "step": 43
    },
    {
      "epoch": 0.14630091438071488,
      "grad_norm": 1.2452101707458496,
      "learning_rate": 9.230476262104677e-05,
      "loss": 2.962,
      "step": 44
    },
    {
      "epoch": 0.14962593516209477,
      "grad_norm": 1.637641429901123,
      "learning_rate": 9.185832391312644e-05,
      "loss": 3.166,
      "step": 45
    },
    {
      "epoch": 0.15295095594347466,
      "grad_norm": 1.2214964628219604,
      "learning_rate": 9.140044155740101e-05,
      "loss": 2.98,
      "step": 46
    },
    {
      "epoch": 0.15627597672485452,
      "grad_norm": 1.1904252767562866,
      "learning_rate": 9.093124073433463e-05,
      "loss": 3.1473,
      "step": 47
    },
    {
      "epoch": 0.1596009975062344,
      "grad_norm": 1.6513088941574097,
      "learning_rate": 9.045084971874738e-05,
      "loss": 3.1954,
      "step": 48
    },
    {
      "epoch": 0.1629260182876143,
      "grad_norm": 1.9471615552902222,
      "learning_rate": 8.995939984474624e-05,
      "loss": 3.4457,
      "step": 49
    },
    {
      "epoch": 0.1662510390689942,
      "grad_norm": 3.147850751876831,
      "learning_rate": 8.945702546981969e-05,
      "loss": 3.9347,
      "step": 50
    },
    {
      "epoch": 0.1662510390689942,
      "eval_loss": 2.619591236114502,
      "eval_runtime": 36.6813,
      "eval_samples_per_second": 13.822,
      "eval_steps_per_second": 3.462,
      "step": 50
    },
    {
      "epoch": 0.16957605985037408,
      "grad_norm": 0.4913792908191681,
      "learning_rate": 8.894386393810563e-05,
      "loss": 1.8187,
      "step": 51
    },
    {
      "epoch": 0.17290108063175394,
      "grad_norm": 0.6059408187866211,
      "learning_rate": 8.842005554284296e-05,
      "loss": 1.9469,
      "step": 52
    },
    {
      "epoch": 0.17622610141313383,
      "grad_norm": 0.7039671540260315,
      "learning_rate": 8.788574348801675e-05,
      "loss": 1.5762,
      "step": 53
    },
    {
      "epoch": 0.17955112219451372,
      "grad_norm": 0.5998632311820984,
      "learning_rate": 8.73410738492077e-05,
      "loss": 1.4354,
      "step": 54
    },
    {
      "epoch": 0.1828761429758936,
      "grad_norm": 0.5073721408843994,
      "learning_rate": 8.678619553365659e-05,
      "loss": 1.3488,
      "step": 55
    },
    {
      "epoch": 0.1862011637572735,
      "grad_norm": 0.39069750905036926,
      "learning_rate": 8.622126023955446e-05,
      "loss": 1.5235,
      "step": 56
    },
    {
      "epoch": 0.18952618453865336,
      "grad_norm": 0.3212682008743286,
      "learning_rate": 8.564642241456986e-05,
      "loss": 1.4037,
      "step": 57
    },
    {
      "epoch": 0.19285120532003325,
      "grad_norm": 0.32433271408081055,
      "learning_rate": 8.506183921362443e-05,
      "loss": 1.4773,
      "step": 58
    },
    {
      "epoch": 0.19617622610141314,
      "grad_norm": 0.32160893082618713,
      "learning_rate": 8.44676704559283e-05,
      "loss": 1.5648,
      "step": 59
    },
    {
      "epoch": 0.19950124688279303,
      "grad_norm": 0.34962618350982666,
      "learning_rate": 8.386407858128706e-05,
      "loss": 1.7763,
      "step": 60
    },
    {
      "epoch": 0.2028262676641729,
      "grad_norm": 0.4209052622318268,
      "learning_rate": 8.32512286056924e-05,
      "loss": 1.7525,
      "step": 61
    },
    {
      "epoch": 0.20615128844555278,
      "grad_norm": 0.43889158964157104,
      "learning_rate": 8.262928807620843e-05,
      "loss": 1.9617,
      "step": 62
    },
    {
      "epoch": 0.20947630922693267,
      "grad_norm": 0.4561992585659027,
      "learning_rate": 8.199842702516583e-05,
      "loss": 1.6328,
      "step": 63
    },
    {
      "epoch": 0.21280133000831256,
      "grad_norm": 0.44138792157173157,
      "learning_rate": 8.135881792367686e-05,
      "loss": 1.6771,
      "step": 64
    },
    {
      "epoch": 0.21612635078969245,
      "grad_norm": 0.5273365378379822,
      "learning_rate": 8.07106356344834e-05,
      "loss": 2.3747,
      "step": 65
    },
    {
      "epoch": 0.2194513715710723,
      "grad_norm": 0.5763283967971802,
      "learning_rate": 8.005405736415126e-05,
      "loss": 2.0103,
      "step": 66
    },
    {
      "epoch": 0.2227763923524522,
      "grad_norm": 0.4764019548892975,
      "learning_rate": 7.938926261462366e-05,
      "loss": 1.9639,
      "step": 67
    },
    {
      "epoch": 0.22610141313383209,
      "grad_norm": 0.5063018202781677,
      "learning_rate": 7.871643313414718e-05,
      "loss": 2.0682,
      "step": 68
    },
    {
      "epoch": 0.22942643391521197,
      "grad_norm": 0.5650584101676941,
      "learning_rate": 7.803575286758364e-05,
      "loss": 2.3073,
      "step": 69
    },
    {
      "epoch": 0.23275145469659186,
      "grad_norm": 0.674048900604248,
      "learning_rate": 7.734740790612136e-05,
      "loss": 2.0572,
      "step": 70
    },
    {
      "epoch": 0.23607647547797173,
      "grad_norm": 0.5179927349090576,
      "learning_rate": 7.66515864363997e-05,
      "loss": 2.2731,
      "step": 71
    },
    {
      "epoch": 0.23940149625935161,
      "grad_norm": 0.6439059972763062,
      "learning_rate": 7.594847868906076e-05,
      "loss": 2.5171,
      "step": 72
    },
    {
      "epoch": 0.2427265170407315,
      "grad_norm": 0.5874778628349304,
      "learning_rate": 7.52382768867422e-05,
      "loss": 2.7064,
      "step": 73
    },
    {
      "epoch": 0.2460515378221114,
      "grad_norm": 0.5716459155082703,
      "learning_rate": 7.452117519152542e-05,
      "loss": 2.472,
      "step": 74
    },
    {
      "epoch": 0.24937655860349128,
      "grad_norm": 0.6581025719642639,
      "learning_rate": 7.379736965185368e-05,
      "loss": 2.4448,
      "step": 75
    },
    {
      "epoch": 0.25270157938487114,
      "grad_norm": 0.6672977805137634,
      "learning_rate": 7.30670581489344e-05,
      "loss": 2.5213,
      "step": 76
    },
    {
      "epoch": 0.25602660016625106,
      "grad_norm": 0.8753868937492371,
      "learning_rate": 7.233044034264034e-05,
      "loss": 2.5466,
      "step": 77
    },
    {
      "epoch": 0.2593516209476309,
      "grad_norm": 0.7117350101470947,
      "learning_rate": 7.158771761692464e-05,
      "loss": 2.4761,
      "step": 78
    },
    {
      "epoch": 0.2626766417290108,
      "grad_norm": 1.0922863483428955,
      "learning_rate": 7.083909302476453e-05,
      "loss": 2.549,
      "step": 79
    },
    {
      "epoch": 0.2660016625103907,
      "grad_norm": 0.7558392882347107,
      "learning_rate": 7.008477123264848e-05,
      "loss": 2.5863,
      "step": 80
    },
    {
      "epoch": 0.26932668329177056,
      "grad_norm": 0.7467756271362305,
      "learning_rate": 6.932495846462261e-05,
      "loss": 2.5209,
      "step": 81
    },
    {
      "epoch": 0.2726517040731505,
      "grad_norm": 0.8661202192306519,
      "learning_rate": 6.855986244591104e-05,
      "loss": 2.9134,
      "step": 82
    },
    {
      "epoch": 0.27597672485453034,
      "grad_norm": 0.7943313121795654,
      "learning_rate": 6.778969234612584e-05,
      "loss": 2.4747,
      "step": 83
    },
    {
      "epoch": 0.2793017456359102,
      "grad_norm": 0.8861684799194336,
      "learning_rate": 6.701465872208216e-05,
      "loss": 2.6083,
      "step": 84
    },
    {
      "epoch": 0.2826267664172901,
      "grad_norm": 0.8328458070755005,
      "learning_rate": 6.623497346023418e-05,
      "loss": 2.594,
      "step": 85
    },
    {
      "epoch": 0.28595178719867,
      "grad_norm": 0.7635563015937805,
      "learning_rate": 6.545084971874738e-05,
      "loss": 2.4037,
      "step": 86
    },
    {
      "epoch": 0.2892768079800499,
      "grad_norm": 0.8964987397193909,
      "learning_rate": 6.466250186922325e-05,
      "loss": 2.6744,
      "step": 87
    },
    {
      "epoch": 0.29260182876142976,
      "grad_norm": 0.94998699426651,
      "learning_rate": 6.387014543809223e-05,
      "loss": 2.8877,
      "step": 88
    },
    {
      "epoch": 0.2959268495428096,
      "grad_norm": 0.8938713073730469,
      "learning_rate": 6.307399704769099e-05,
      "loss": 2.5389,
      "step": 89
    },
    {
      "epoch": 0.29925187032418954,
      "grad_norm": 0.8773661851882935,
      "learning_rate": 6.227427435703997e-05,
      "loss": 2.5848,
      "step": 90
    },
    {
      "epoch": 0.3025768911055694,
      "grad_norm": 0.9464321732521057,
      "learning_rate": 6.147119600233758e-05,
      "loss": 2.5551,
      "step": 91
    },
    {
      "epoch": 0.3059019118869493,
      "grad_norm": 1.0121573209762573,
      "learning_rate": 6.066498153718735e-05,
      "loss": 2.8209,
      "step": 92
    },
    {
      "epoch": 0.3092269326683292,
      "grad_norm": 0.9872815608978271,
      "learning_rate": 5.985585137257401e-05,
      "loss": 2.7237,
      "step": 93
    },
    {
      "epoch": 0.31255195344970904,
      "grad_norm": 1.0549014806747437,
      "learning_rate": 5.90440267166055e-05,
      "loss": 2.7921,
      "step": 94
    },
    {
      "epoch": 0.31587697423108896,
      "grad_norm": 1.0411452054977417,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 2.8538,
      "step": 95
    },
    {
      "epoch": 0.3192019950124688,
      "grad_norm": 1.1173065900802612,
      "learning_rate": 5.74131823855921e-05,
      "loss": 2.6827,
      "step": 96
    },
    {
      "epoch": 0.32252701579384874,
      "grad_norm": 1.327447533607483,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 3.2663,
      "step": 97
    },
    {
      "epoch": 0.3258520365752286,
      "grad_norm": 1.235247254371643,
      "learning_rate": 5.577423184847932e-05,
      "loss": 3.1765,
      "step": 98
    },
    {
      "epoch": 0.32917705735660846,
      "grad_norm": 1.459521770477295,
      "learning_rate": 5.495227651252315e-05,
      "loss": 3.1847,
      "step": 99
    },
    {
      "epoch": 0.3325020781379884,
      "grad_norm": 2.0694055557250977,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 3.5904,
      "step": 100
    },
    {
      "epoch": 0.3325020781379884,
      "eval_loss": 2.512159585952759,
      "eval_runtime": 36.9675,
      "eval_samples_per_second": 13.715,
      "eval_steps_per_second": 3.435,
      "step": 100
    },
    {
      "epoch": 0.33582709891936824,
      "grad_norm": 0.1913176029920578,
      "learning_rate": 5.330452921628497e-05,
      "loss": 1.7317,
      "step": 101
    },
    {
      "epoch": 0.33915211970074816,
      "grad_norm": 0.2875644564628601,
      "learning_rate": 5.247918773366112e-05,
      "loss": 1.8991,
      "step": 102
    },
    {
      "epoch": 0.342477140482128,
      "grad_norm": 0.332629531621933,
      "learning_rate": 5.165316846586541e-05,
      "loss": 1.2576,
      "step": 103
    },
    {
      "epoch": 0.3458021612635079,
      "grad_norm": 0.3853623569011688,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 1.4941,
      "step": 104
    },
    {
      "epoch": 0.3491271820448878,
      "grad_norm": 0.3697432577610016,
      "learning_rate": 5e-05,
      "loss": 1.4407,
      "step": 105
    },
    {
      "epoch": 0.35245220282626766,
      "grad_norm": 0.3243190050125122,
      "learning_rate": 4.917330276168208e-05,
      "loss": 1.0466,
      "step": 106
    },
    {
      "epoch": 0.3557772236076476,
      "grad_norm": 0.4234490990638733,
      "learning_rate": 4.834683153413459e-05,
      "loss": 1.65,
      "step": 107
    },
    {
      "epoch": 0.35910224438902744,
      "grad_norm": 0.37311846017837524,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 1.6035,
      "step": 108
    },
    {
      "epoch": 0.3624272651704073,
      "grad_norm": 0.42025139927864075,
      "learning_rate": 4.669547078371504e-05,
      "loss": 1.7678,
      "step": 109
    },
    {
      "epoch": 0.3657522859517872,
      "grad_norm": 0.41857877373695374,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 1.687,
      "step": 110
    },
    {
      "epoch": 0.3690773067331671,
      "grad_norm": 0.49679869413375854,
      "learning_rate": 4.504772348747687e-05,
      "loss": 1.7468,
      "step": 111
    },
    {
      "epoch": 0.372402327514547,
      "grad_norm": 0.6674458980560303,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 1.8978,
      "step": 112
    },
    {
      "epoch": 0.37572734829592686,
      "grad_norm": 0.5891672372817993,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 1.91,
      "step": 113
    },
    {
      "epoch": 0.3790523690773067,
      "grad_norm": 0.5837957859039307,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 1.8848,
      "step": 114
    },
    {
      "epoch": 0.38237738985868663,
      "grad_norm": 0.5408995747566223,
      "learning_rate": 4.17702704859633e-05,
      "loss": 1.954,
      "step": 115
    },
    {
      "epoch": 0.3857024106400665,
      "grad_norm": 0.4996670186519623,
      "learning_rate": 4.095597328339452e-05,
      "loss": 1.8418,
      "step": 116
    },
    {
      "epoch": 0.38902743142144636,
      "grad_norm": 0.46126365661621094,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 2.1418,
      "step": 117
    },
    {
      "epoch": 0.3923524522028263,
      "grad_norm": 0.4791029095649719,
      "learning_rate": 3.933501846281267e-05,
      "loss": 1.9923,
      "step": 118
    },
    {
      "epoch": 0.39567747298420614,
      "grad_norm": 0.48812511563301086,
      "learning_rate": 3.852880399766243e-05,
      "loss": 2.3085,
      "step": 119
    },
    {
      "epoch": 0.39900249376558605,
      "grad_norm": 0.5243691205978394,
      "learning_rate": 3.772572564296005e-05,
      "loss": 2.251,
      "step": 120
    },
    {
      "epoch": 0.4023275145469659,
      "grad_norm": 0.5315554738044739,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 2.2106,
      "step": 121
    },
    {
      "epoch": 0.4056525353283458,
      "grad_norm": 0.6056045293807983,
      "learning_rate": 3.612985456190778e-05,
      "loss": 2.067,
      "step": 122
    },
    {
      "epoch": 0.4089775561097257,
      "grad_norm": 0.6105309128761292,
      "learning_rate": 3.533749813077677e-05,
      "loss": 2.1116,
      "step": 123
    },
    {
      "epoch": 0.41230257689110555,
      "grad_norm": 0.6107189059257507,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 2.3749,
      "step": 124
    },
    {
      "epoch": 0.41562759767248547,
      "grad_norm": 0.6972635388374329,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 2.4589,
      "step": 125
    },
    {
      "epoch": 0.41895261845386533,
      "grad_norm": 0.6330546140670776,
      "learning_rate": 3.298534127791785e-05,
      "loss": 2.5495,
      "step": 126
    },
    {
      "epoch": 0.4222776392352452,
      "grad_norm": 0.6755483746528625,
      "learning_rate": 3.221030765387417e-05,
      "loss": 2.5935,
      "step": 127
    },
    {
      "epoch": 0.4256026600166251,
      "grad_norm": 0.716001570224762,
      "learning_rate": 3.144013755408895e-05,
      "loss": 2.4323,
      "step": 128
    },
    {
      "epoch": 0.428927680798005,
      "grad_norm": 0.739983320236206,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 2.627,
      "step": 129
    },
    {
      "epoch": 0.4322527015793849,
      "grad_norm": 0.6974490284919739,
      "learning_rate": 2.991522876735154e-05,
      "loss": 2.4962,
      "step": 130
    },
    {
      "epoch": 0.43557772236076475,
      "grad_norm": 0.8209483027458191,
      "learning_rate": 2.916090697523549e-05,
      "loss": 2.4155,
      "step": 131
    },
    {
      "epoch": 0.4389027431421446,
      "grad_norm": 0.7780238389968872,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 2.5729,
      "step": 132
    },
    {
      "epoch": 0.44222776392352453,
      "grad_norm": 0.7966702580451965,
      "learning_rate": 2.766955965735968e-05,
      "loss": 2.2295,
      "step": 133
    },
    {
      "epoch": 0.4455527847049044,
      "grad_norm": 0.7544305920600891,
      "learning_rate": 2.693294185106562e-05,
      "loss": 2.3463,
      "step": 134
    },
    {
      "epoch": 0.4488778054862843,
      "grad_norm": 0.874858558177948,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 2.5865,
      "step": 135
    },
    {
      "epoch": 0.45220282626766417,
      "grad_norm": 0.817158043384552,
      "learning_rate": 2.547882480847461e-05,
      "loss": 2.5725,
      "step": 136
    },
    {
      "epoch": 0.45552784704904403,
      "grad_norm": 0.7389349341392517,
      "learning_rate": 2.476172311325783e-05,
      "loss": 2.4777,
      "step": 137
    },
    {
      "epoch": 0.45885286783042395,
      "grad_norm": 0.7815312743186951,
      "learning_rate": 2.405152131093926e-05,
      "loss": 2.6474,
      "step": 138
    },
    {
      "epoch": 0.4621778886118038,
      "grad_norm": 0.8713777661323547,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 2.6075,
      "step": 139
    },
    {
      "epoch": 0.46550290939318373,
      "grad_norm": 0.933516263961792,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 2.8983,
      "step": 140
    },
    {
      "epoch": 0.4688279301745636,
      "grad_norm": 0.8990171551704407,
      "learning_rate": 2.196424713241637e-05,
      "loss": 2.7301,
      "step": 141
    },
    {
      "epoch": 0.47215295095594345,
      "grad_norm": 1.0071741342544556,
      "learning_rate": 2.128356686585282e-05,
      "loss": 3.0113,
      "step": 142
    },
    {
      "epoch": 0.47547797173732337,
      "grad_norm": 1.1403377056121826,
      "learning_rate": 2.061073738537635e-05,
      "loss": 3.0827,
      "step": 143
    },
    {
      "epoch": 0.47880299251870323,
      "grad_norm": 0.9958221912384033,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 2.9719,
      "step": 144
    },
    {
      "epoch": 0.48212801330008315,
      "grad_norm": 1.0008690357208252,
      "learning_rate": 1.928936436551661e-05,
      "loss": 2.7593,
      "step": 145
    },
    {
      "epoch": 0.485453034081463,
      "grad_norm": 1.0845717191696167,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 2.8462,
      "step": 146
    },
    {
      "epoch": 0.48877805486284287,
      "grad_norm": 1.205183744430542,
      "learning_rate": 1.800157297483417e-05,
      "loss": 3.2379,
      "step": 147
    },
    {
      "epoch": 0.4921030756442228,
      "grad_norm": 1.617091178894043,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 3.2233,
      "step": 148
    },
    {
      "epoch": 0.49542809642560265,
      "grad_norm": 1.5926539897918701,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 3.4084,
      "step": 149
    },
    {
      "epoch": 0.49875311720698257,
      "grad_norm": 2.652520179748535,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 3.9244,
      "step": 150
    },
    {
      "epoch": 0.49875311720698257,
      "eval_loss": 2.364495277404785,
      "eval_runtime": 36.7277,
      "eval_samples_per_second": 13.804,
      "eval_steps_per_second": 3.458,
      "step": 150
    },
    {
      "epoch": 0.5020781379883624,
      "grad_norm": 0.13091854751110077,
      "learning_rate": 1.553232954407171e-05,
      "loss": 2.0167,
      "step": 151
    },
    {
      "epoch": 0.5054031587697423,
      "grad_norm": 0.1495131105184555,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 1.491,
      "step": 152
    },
    {
      "epoch": 0.5087281795511222,
      "grad_norm": 0.1672244817018509,
      "learning_rate": 1.435357758543015e-05,
      "loss": 1.297,
      "step": 153
    },
    {
      "epoch": 0.5120532003325021,
      "grad_norm": 0.19511078298091888,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 1.3551,
      "step": 154
    },
    {
      "epoch": 0.515378221113882,
      "grad_norm": 0.19908133149147034,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 1.2202,
      "step": 155
    },
    {
      "epoch": 0.5187032418952618,
      "grad_norm": 0.20949244499206543,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 1.1198,
      "step": 156
    },
    {
      "epoch": 0.5220282626766417,
      "grad_norm": 0.283833771944046,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 1.7081,
      "step": 157
    },
    {
      "epoch": 0.5253532834580216,
      "grad_norm": 0.28999269008636475,
      "learning_rate": 1.157994445715706e-05,
      "loss": 1.5825,
      "step": 158
    },
    {
      "epoch": 0.5286783042394015,
      "grad_norm": 0.33534905314445496,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 1.4001,
      "step": 159
    },
    {
      "epoch": 0.5320033250207814,
      "grad_norm": 0.4424426555633545,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 1.8759,
      "step": 160
    },
    {
      "epoch": 0.5353283458021613,
      "grad_norm": 0.4760247468948364,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 2.0996,
      "step": 161
    },
    {
      "epoch": 0.5386533665835411,
      "grad_norm": 0.5403597354888916,
      "learning_rate": 9.549150281252633e-06,
      "loss": 1.9253,
      "step": 162
    },
    {
      "epoch": 0.541978387364921,
      "grad_norm": 0.6103106141090393,
      "learning_rate": 9.068759265665384e-06,
      "loss": 1.8665,
      "step": 163
    },
    {
      "epoch": 0.545303408146301,
      "grad_norm": 0.6575484871864319,
      "learning_rate": 8.599558442598998e-06,
      "loss": 1.919,
      "step": 164
    },
    {
      "epoch": 0.5486284289276808,
      "grad_norm": 0.6090027689933777,
      "learning_rate": 8.141676086873572e-06,
      "loss": 2.1069,
      "step": 165
    },
    {
      "epoch": 0.5519534497090607,
      "grad_norm": 0.5894596576690674,
      "learning_rate": 7.695237378953223e-06,
      "loss": 2.2157,
      "step": 166
    },
    {
      "epoch": 0.5552784704904405,
      "grad_norm": 0.6697620153427124,
      "learning_rate": 7.260364370723044e-06,
      "loss": 1.8026,
      "step": 167
    },
    {
      "epoch": 0.5586034912718204,
      "grad_norm": 0.5990772247314453,
      "learning_rate": 6.837175952121306e-06,
      "loss": 2.0866,
      "step": 168
    },
    {
      "epoch": 0.5619285120532004,
      "grad_norm": 0.5796670913696289,
      "learning_rate": 6.425787818636131e-06,
      "loss": 2.2148,
      "step": 169
    },
    {
      "epoch": 0.5652535328345802,
      "grad_norm": 0.6188122034072876,
      "learning_rate": 6.026312439675552e-06,
      "loss": 2.2896,
      "step": 170
    },
    {
      "epoch": 0.5685785536159601,
      "grad_norm": 0.583895742893219,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 2.1988,
      "step": 171
    },
    {
      "epoch": 0.57190357439734,
      "grad_norm": 0.6323407292366028,
      "learning_rate": 5.263533508961827e-06,
      "loss": 2.1278,
      "step": 172
    },
    {
      "epoch": 0.5752285951787198,
      "grad_norm": 0.7294795513153076,
      "learning_rate": 4.900438493352055e-06,
      "loss": 2.469,
      "step": 173
    },
    {
      "epoch": 0.5785536159600998,
      "grad_norm": 0.6735561490058899,
      "learning_rate": 4.549673247541875e-06,
      "loss": 2.4645,
      "step": 174
    },
    {
      "epoch": 0.5818786367414797,
      "grad_norm": 0.6964554786682129,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 2.4965,
      "step": 175
    },
    {
      "epoch": 0.5852036575228595,
      "grad_norm": 0.6979163885116577,
      "learning_rate": 3.885512251130763e-06,
      "loss": 2.4399,
      "step": 176
    },
    {
      "epoch": 0.5885286783042394,
      "grad_norm": 0.7108325958251953,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.6086,
      "step": 177
    },
    {
      "epoch": 0.5918536990856192,
      "grad_norm": 0.7156085968017578,
      "learning_rate": 3.271776770026963e-06,
      "loss": 2.5234,
      "step": 178
    },
    {
      "epoch": 0.5951787198669992,
      "grad_norm": 0.7812105417251587,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 2.4373,
      "step": 179
    },
    {
      "epoch": 0.5985037406483791,
      "grad_norm": 0.8563242554664612,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 2.564,
      "step": 180
    },
    {
      "epoch": 0.6018287614297589,
      "grad_norm": 0.7718956470489502,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 2.1678,
      "step": 181
    },
    {
      "epoch": 0.6051537822111388,
      "grad_norm": 0.8392854332923889,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 2.6139,
      "step": 182
    },
    {
      "epoch": 0.6084788029925187,
      "grad_norm": 0.8928911685943604,
      "learning_rate": 1.962316193157593e-06,
      "loss": 2.5417,
      "step": 183
    },
    {
      "epoch": 0.6118038237738986,
      "grad_norm": 0.8103169202804565,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 2.5984,
      "step": 184
    },
    {
      "epoch": 0.6151288445552785,
      "grad_norm": 0.8044213056564331,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 2.5431,
      "step": 185
    },
    {
      "epoch": 0.6184538653366584,
      "grad_norm": 0.8493407964706421,
      "learning_rate": 1.333670137599713e-06,
      "loss": 2.8052,
      "step": 186
    },
    {
      "epoch": 0.6217788861180382,
      "grad_norm": 0.865878164768219,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 2.6322,
      "step": 187
    },
    {
      "epoch": 0.6251039068994181,
      "grad_norm": 0.8963941931724548,
      "learning_rate": 9.810017062595322e-07,
      "loss": 2.7874,
      "step": 188
    },
    {
      "epoch": 0.628428927680798,
      "grad_norm": 0.8851121068000793,
      "learning_rate": 8.247462563808817e-07,
      "loss": 2.8037,
      "step": 189
    },
    {
      "epoch": 0.6317539484621779,
      "grad_norm": 0.9431672692298889,
      "learning_rate": 6.819348298638839e-07,
      "loss": 3.0261,
      "step": 190
    },
    {
      "epoch": 0.6350789692435578,
      "grad_norm": 0.882543683052063,
      "learning_rate": 5.526064699265753e-07,
      "loss": 2.9318,
      "step": 191
    },
    {
      "epoch": 0.6384039900249376,
      "grad_norm": 0.9423844218254089,
      "learning_rate": 4.367965336512403e-07,
      "loss": 3.0992,
      "step": 192
    },
    {
      "epoch": 0.6417290108063175,
      "grad_norm": 1.2414581775665283,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 3.2177,
      "step": 193
    },
    {
      "epoch": 0.6450540315876975,
      "grad_norm": 1.1261076927185059,
      "learning_rate": 2.458548727494292e-07,
      "loss": 3.0623,
      "step": 194
    },
    {
      "epoch": 0.6483790523690773,
      "grad_norm": 1.0396223068237305,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 2.7971,
      "step": 195
    },
    {
      "epoch": 0.6517040731504572,
      "grad_norm": 1.0330443382263184,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 2.9151,
      "step": 196
    },
    {
      "epoch": 0.6550290939318371,
      "grad_norm": 1.1228389739990234,
      "learning_rate": 6.150154258476315e-08,
      "loss": 2.9858,
      "step": 197
    },
    {
      "epoch": 0.6583541147132169,
      "grad_norm": 1.2906904220581055,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 3.1842,
      "step": 198
    },
    {
      "epoch": 0.6616791354945969,
      "grad_norm": 1.5825787782669067,
      "learning_rate": 6.834750376549792e-09,
      "loss": 3.5745,
      "step": 199
    },
    {
      "epoch": 0.6650041562759768,
      "grad_norm": 2.3015079498291016,
      "learning_rate": 0.0,
      "loss": 4.5316,
      "step": 200
    },
    {
      "epoch": 0.6650041562759768,
      "eval_loss": 2.3479816913604736,
      "eval_runtime": 36.7019,
      "eval_samples_per_second": 13.814,
      "eval_steps_per_second": 3.46,
      "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": 2.8899887529590784e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}