{
  "best_metric": 12.420177459716797,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.29750836742283376,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0014875418371141688,
      "grad_norm": 0.01454976573586464,
      "learning_rate": 1e-05,
      "loss": 12.4511,
      "step": 1
    },
    {
      "epoch": 0.0014875418371141688,
      "eval_loss": 12.455506324768066,
      "eval_runtime": 8.6451,
      "eval_samples_per_second": 130.942,
      "eval_steps_per_second": 32.735,
      "step": 1
    },
    {
      "epoch": 0.0029750836742283376,
      "grad_norm": 0.01452404074370861,
      "learning_rate": 2e-05,
      "loss": 12.4595,
      "step": 2
    },
    {
      "epoch": 0.004462625511342506,
      "grad_norm": 0.017401931807398796,
      "learning_rate": 3e-05,
      "loss": 12.4582,
      "step": 3
    },
    {
      "epoch": 0.005950167348456675,
      "grad_norm": 0.014948471449315548,
      "learning_rate": 4e-05,
      "loss": 12.4556,
      "step": 4
    },
    {
      "epoch": 0.007437709185570844,
      "grad_norm": 0.018253479152917862,
      "learning_rate": 5e-05,
      "loss": 12.4583,
      "step": 5
    },
    {
      "epoch": 0.008925251022685012,
      "grad_norm": 0.017257850617170334,
      "learning_rate": 6e-05,
      "loss": 12.4453,
      "step": 6
    },
    {
      "epoch": 0.010412792859799182,
      "grad_norm": 0.01633993722498417,
      "learning_rate": 7e-05,
      "loss": 12.4557,
      "step": 7
    },
    {
      "epoch": 0.01190033469691335,
      "grad_norm": 0.01574438251554966,
      "learning_rate": 8e-05,
      "loss": 12.4524,
      "step": 8
    },
    {
      "epoch": 0.01338787653402752,
      "grad_norm": 0.01728111505508423,
      "learning_rate": 9e-05,
      "loss": 12.4587,
      "step": 9
    },
    {
      "epoch": 0.014875418371141688,
      "grad_norm": 0.020564178004860878,
      "learning_rate": 0.0001,
      "loss": 12.4507,
      "step": 10
    },
    {
      "epoch": 0.016362960208255856,
      "grad_norm": 0.018883228302001953,
      "learning_rate": 9.999316524962345e-05,
      "loss": 12.4585,
      "step": 11
    },
    {
      "epoch": 0.017850502045370024,
      "grad_norm": 0.02268098294734955,
      "learning_rate": 9.997266286704631e-05,
      "loss": 12.4459,
      "step": 12
    },
    {
      "epoch": 0.019338043882484196,
      "grad_norm": 0.022393308579921722,
      "learning_rate": 9.993849845741524e-05,
      "loss": 12.4442,
      "step": 13
    },
    {
      "epoch": 0.020825585719598364,
      "grad_norm": 0.021197356283664703,
      "learning_rate": 9.989068136093873e-05,
      "loss": 12.4558,
      "step": 14
    },
    {
      "epoch": 0.022313127556712532,
      "grad_norm": 0.016823744401335716,
      "learning_rate": 9.98292246503335e-05,
      "loss": 12.4615,
      "step": 15
    },
    {
      "epoch": 0.0238006693938267,
      "grad_norm": 0.022657601162791252,
      "learning_rate": 9.975414512725057e-05,
      "loss": 12.4478,
      "step": 16
    },
    {
      "epoch": 0.02528821123094087,
      "grad_norm": 0.02074330672621727,
      "learning_rate": 9.966546331768191e-05,
      "loss": 12.4626,
      "step": 17
    },
    {
      "epoch": 0.02677575306805504,
      "grad_norm": 0.023210735991597176,
      "learning_rate": 9.956320346634876e-05,
      "loss": 12.4518,
      "step": 18
    },
    {
      "epoch": 0.028263294905169208,
      "grad_norm": 0.023989206179976463,
      "learning_rate": 9.944739353007344e-05,
      "loss": 12.4476,
      "step": 19
    },
    {
      "epoch": 0.029750836742283376,
      "grad_norm": 0.027420252561569214,
      "learning_rate": 9.931806517013612e-05,
      "loss": 12.449,
      "step": 20
    },
    {
      "epoch": 0.031238378579397545,
      "grad_norm": 0.021792909130454063,
      "learning_rate": 9.917525374361912e-05,
      "loss": 12.4512,
      "step": 21
    },
    {
      "epoch": 0.03272592041651171,
      "grad_norm": 0.029891427606344223,
      "learning_rate": 9.901899829374047e-05,
      "loss": 12.4526,
      "step": 22
    },
    {
      "epoch": 0.034213462253625884,
      "grad_norm": 0.028047390282154083,
      "learning_rate": 9.884934153917997e-05,
      "loss": 12.4568,
      "step": 23
    },
    {
      "epoch": 0.03570100409074005,
      "grad_norm": 0.026077531278133392,
      "learning_rate": 9.86663298624003e-05,
      "loss": 12.4436,
      "step": 24
    },
    {
      "epoch": 0.03718854592785422,
      "grad_norm": 0.03540145605802536,
      "learning_rate": 9.847001329696653e-05,
      "loss": 12.448,
      "step": 25
    },
    {
      "epoch": 0.03867608776496839,
      "grad_norm": 0.02705364301800728,
      "learning_rate": 9.826044551386744e-05,
      "loss": 12.4592,
      "step": 26
    },
    {
      "epoch": 0.04016362960208256,
      "grad_norm": 0.030234364792704582,
      "learning_rate": 9.803768380684242e-05,
      "loss": 12.4605,
      "step": 27
    },
    {
      "epoch": 0.04165117143919673,
      "grad_norm": 0.03328441455960274,
      "learning_rate": 9.780178907671789e-05,
      "loss": 12.4572,
      "step": 28
    },
    {
      "epoch": 0.04313871327631089,
      "grad_norm": 0.031843043863773346,
      "learning_rate": 9.755282581475769e-05,
      "loss": 12.4621,
      "step": 29
    },
    {
      "epoch": 0.044626255113425065,
      "grad_norm": 0.04374903067946434,
      "learning_rate": 9.729086208503174e-05,
      "loss": 12.4434,
      "step": 30
    },
    {
      "epoch": 0.046113796950539236,
      "grad_norm": 0.03992782160639763,
      "learning_rate": 9.701596950580806e-05,
      "loss": 12.4547,
      "step": 31
    },
    {
      "epoch": 0.0476013387876534,
      "grad_norm": 0.04382240027189255,
      "learning_rate": 9.672822322997305e-05,
      "loss": 12.4517,
      "step": 32
    },
    {
      "epoch": 0.04908888062476757,
      "grad_norm": 0.04142119735479355,
      "learning_rate": 9.642770192448536e-05,
      "loss": 12.4537,
      "step": 33
    },
    {
      "epoch": 0.05057642246188174,
      "grad_norm": 0.04619595408439636,
      "learning_rate": 9.611448774886924e-05,
      "loss": 12.446,
      "step": 34
    },
    {
      "epoch": 0.05206396429899591,
      "grad_norm": 0.0426214300096035,
      "learning_rate": 9.578866633275288e-05,
      "loss": 12.4604,
      "step": 35
    },
    {
      "epoch": 0.05355150613611008,
      "grad_norm": 0.0436868891119957,
      "learning_rate": 9.545032675245813e-05,
      "loss": 12.4602,
      "step": 36
    },
    {
      "epoch": 0.055039047973224245,
      "grad_norm": 0.05540507659316063,
      "learning_rate": 9.509956150664796e-05,
      "loss": 12.448,
      "step": 37
    },
    {
      "epoch": 0.056526589810338417,
      "grad_norm": 0.06340522319078445,
      "learning_rate": 9.473646649103818e-05,
      "loss": 12.4438,
      "step": 38
    },
    {
      "epoch": 0.05801413164745258,
      "grad_norm": 0.05001519247889519,
      "learning_rate": 9.43611409721806e-05,
      "loss": 12.4505,
      "step": 39
    },
    {
      "epoch": 0.05950167348456675,
      "grad_norm": 0.05245659500360489,
      "learning_rate": 9.397368756032445e-05,
      "loss": 12.4437,
      "step": 40
    },
    {
      "epoch": 0.060989215321680924,
      "grad_norm": 0.07692553102970123,
      "learning_rate": 9.357421218136386e-05,
      "loss": 12.4567,
      "step": 41
    },
    {
      "epoch": 0.06247675715879509,
      "grad_norm": 0.061117518693208694,
      "learning_rate": 9.316282404787871e-05,
      "loss": 12.4636,
      "step": 42
    },
    {
      "epoch": 0.06396429899590926,
      "grad_norm": 0.06437050551176071,
      "learning_rate": 9.273963562927695e-05,
      "loss": 12.4453,
      "step": 43
    },
    {
      "epoch": 0.06545184083302343,
      "grad_norm": 0.07212281972169876,
      "learning_rate": 9.230476262104677e-05,
      "loss": 12.4496,
      "step": 44
    },
    {
      "epoch": 0.0669393826701376,
      "grad_norm": 0.0755159854888916,
      "learning_rate": 9.185832391312644e-05,
      "loss": 12.4475,
      "step": 45
    },
    {
      "epoch": 0.06842692450725177,
      "grad_norm": 0.07953204214572906,
      "learning_rate": 9.140044155740101e-05,
      "loss": 12.466,
      "step": 46
    },
    {
      "epoch": 0.06991446634436593,
      "grad_norm": 0.1227644756436348,
      "learning_rate": 9.093124073433463e-05,
      "loss": 12.4429,
      "step": 47
    },
    {
      "epoch": 0.0714020081814801,
      "grad_norm": 0.11247874051332474,
      "learning_rate": 9.045084971874738e-05,
      "loss": 12.4406,
      "step": 48
    },
    {
      "epoch": 0.07288955001859428,
      "grad_norm": 0.13469244539737701,
      "learning_rate": 8.995939984474624e-05,
      "loss": 12.4453,
      "step": 49
    },
    {
      "epoch": 0.07437709185570844,
      "grad_norm": 0.16619813442230225,
      "learning_rate": 8.945702546981969e-05,
      "loss": 12.4542,
      "step": 50
    },
    {
      "epoch": 0.07437709185570844,
      "eval_loss": 12.450380325317383,
      "eval_runtime": 8.6549,
      "eval_samples_per_second": 130.793,
      "eval_steps_per_second": 32.698,
      "step": 50
    },
    {
      "epoch": 0.0758646336928226,
      "grad_norm": 0.03530978038907051,
      "learning_rate": 8.894386393810563e-05,
      "loss": 12.4599,
      "step": 51
    },
    {
      "epoch": 0.07735217552993678,
      "grad_norm": 0.04758322983980179,
      "learning_rate": 8.842005554284296e-05,
      "loss": 12.4417,
      "step": 52
    },
    {
      "epoch": 0.07883971736705095,
      "grad_norm": 0.06841134279966354,
      "learning_rate": 8.788574348801675e-05,
      "loss": 12.4416,
      "step": 53
    },
    {
      "epoch": 0.08032725920416511,
      "grad_norm": 0.050425589084625244,
      "learning_rate": 8.73410738492077e-05,
      "loss": 12.4525,
      "step": 54
    },
    {
      "epoch": 0.08181480104127929,
      "grad_norm": 0.05450756102800369,
      "learning_rate": 8.678619553365659e-05,
      "loss": 12.4444,
      "step": 55
    },
    {
      "epoch": 0.08330234287839346,
      "grad_norm": 0.07208478450775146,
      "learning_rate": 8.622126023955446e-05,
      "loss": 12.4435,
      "step": 56
    },
    {
      "epoch": 0.08478988471550762,
      "grad_norm": 0.08101324737071991,
      "learning_rate": 8.564642241456986e-05,
      "loss": 12.441,
      "step": 57
    },
    {
      "epoch": 0.08627742655262179,
      "grad_norm": 0.06988723576068878,
      "learning_rate": 8.506183921362443e-05,
      "loss": 12.4466,
      "step": 58
    },
    {
      "epoch": 0.08776496838973596,
      "grad_norm": 0.09461096674203873,
      "learning_rate": 8.44676704559283e-05,
      "loss": 12.4479,
      "step": 59
    },
    {
      "epoch": 0.08925251022685013,
      "grad_norm": 0.10347741097211838,
      "learning_rate": 8.386407858128706e-05,
      "loss": 12.4409,
      "step": 60
    },
    {
      "epoch": 0.0907400520639643,
      "grad_norm": 0.09078667312860489,
      "learning_rate": 8.32512286056924e-05,
      "loss": 12.4345,
      "step": 61
    },
    {
      "epoch": 0.09222759390107847,
      "grad_norm": 0.1028178334236145,
      "learning_rate": 8.262928807620843e-05,
      "loss": 12.4462,
      "step": 62
    },
    {
      "epoch": 0.09371513573819264,
      "grad_norm": 0.09257465600967407,
      "learning_rate": 8.199842702516583e-05,
      "loss": 12.4489,
      "step": 63
    },
    {
      "epoch": 0.0952026775753068,
      "grad_norm": 0.1130981594324112,
      "learning_rate": 8.135881792367686e-05,
      "loss": 12.4384,
      "step": 64
    },
    {
      "epoch": 0.09669021941242098,
      "grad_norm": 0.11545998603105545,
      "learning_rate": 8.07106356344834e-05,
      "loss": 12.4417,
      "step": 65
    },
    {
      "epoch": 0.09817776124953514,
      "grad_norm": 0.10885924845933914,
      "learning_rate": 8.005405736415126e-05,
      "loss": 12.449,
      "step": 66
    },
    {
      "epoch": 0.09966530308664931,
      "grad_norm": 0.12517067790031433,
      "learning_rate": 7.938926261462366e-05,
      "loss": 12.4404,
      "step": 67
    },
    {
      "epoch": 0.10115284492376347,
      "grad_norm": 0.12030695378780365,
      "learning_rate": 7.871643313414718e-05,
      "loss": 12.4366,
      "step": 68
    },
    {
      "epoch": 0.10264038676087765,
      "grad_norm": 0.1519274115562439,
      "learning_rate": 7.803575286758364e-05,
      "loss": 12.4372,
      "step": 69
    },
    {
      "epoch": 0.10412792859799182,
      "grad_norm": 0.11230452358722687,
      "learning_rate": 7.734740790612136e-05,
      "loss": 12.4434,
      "step": 70
    },
    {
      "epoch": 0.10561547043510598,
      "grad_norm": 0.14612998068332672,
      "learning_rate": 7.66515864363997e-05,
      "loss": 12.4299,
      "step": 71
    },
    {
      "epoch": 0.10710301227222016,
      "grad_norm": 0.1108112782239914,
      "learning_rate": 7.594847868906076e-05,
      "loss": 12.4462,
      "step": 72
    },
    {
      "epoch": 0.10859055410933433,
      "grad_norm": 0.10579392313957214,
      "learning_rate": 7.52382768867422e-05,
      "loss": 12.4446,
      "step": 73
    },
    {
      "epoch": 0.11007809594644849,
      "grad_norm": 0.14374765753746033,
      "learning_rate": 7.452117519152542e-05,
      "loss": 12.4273,
      "step": 74
    },
    {
      "epoch": 0.11156563778356267,
      "grad_norm": 0.13092581927776337,
      "learning_rate": 7.379736965185368e-05,
      "loss": 12.4381,
      "step": 75
    },
    {
      "epoch": 0.11305317962067683,
      "grad_norm": 0.11139856278896332,
      "learning_rate": 7.30670581489344e-05,
      "loss": 12.4354,
      "step": 76
    },
    {
      "epoch": 0.114540721457791,
      "grad_norm": 0.14735162258148193,
      "learning_rate": 7.233044034264034e-05,
      "loss": 12.4368,
      "step": 77
    },
    {
      "epoch": 0.11602826329490516,
      "grad_norm": 0.1441231667995453,
      "learning_rate": 7.158771761692464e-05,
      "loss": 12.4376,
      "step": 78
    },
    {
      "epoch": 0.11751580513201934,
      "grad_norm": 0.13695688545703888,
      "learning_rate": 7.083909302476453e-05,
      "loss": 12.4539,
      "step": 79
    },
    {
      "epoch": 0.1190033469691335,
      "grad_norm": 0.11217241734266281,
      "learning_rate": 7.008477123264848e-05,
      "loss": 12.4509,
      "step": 80
    },
    {
      "epoch": 0.12049088880624767,
      "grad_norm": 0.13610312342643738,
      "learning_rate": 6.932495846462261e-05,
      "loss": 12.4278,
      "step": 81
    },
    {
      "epoch": 0.12197843064336185,
      "grad_norm": 0.12111765891313553,
      "learning_rate": 6.855986244591104e-05,
      "loss": 12.4254,
      "step": 82
    },
    {
      "epoch": 0.12346597248047601,
      "grad_norm": 0.10756219923496246,
      "learning_rate": 6.778969234612584e-05,
      "loss": 12.4193,
      "step": 83
    },
    {
      "epoch": 0.12495351431759018,
      "grad_norm": 0.13630057871341705,
      "learning_rate": 6.701465872208216e-05,
      "loss": 12.4506,
      "step": 84
    },
    {
      "epoch": 0.12644105615470436,
      "grad_norm": 0.1435575932264328,
      "learning_rate": 6.623497346023418e-05,
      "loss": 12.4363,
      "step": 85
    },
    {
      "epoch": 0.12792859799181852,
      "grad_norm": 0.13258320093154907,
      "learning_rate": 6.545084971874738e-05,
      "loss": 12.4285,
      "step": 86
    },
    {
      "epoch": 0.12941613982893269,
      "grad_norm": 0.12382647395133972,
      "learning_rate": 6.466250186922325e-05,
      "loss": 12.4428,
      "step": 87
    },
    {
      "epoch": 0.13090368166604685,
      "grad_norm": 0.11985477805137634,
      "learning_rate": 6.387014543809223e-05,
      "loss": 12.4337,
      "step": 88
    },
    {
      "epoch": 0.13239122350316102,
      "grad_norm": 0.1534048616886139,
      "learning_rate": 6.307399704769099e-05,
      "loss": 12.4279,
      "step": 89
    },
    {
      "epoch": 0.1338787653402752,
      "grad_norm": 0.11954528838396072,
      "learning_rate": 6.227427435703997e-05,
      "loss": 12.4346,
      "step": 90
    },
    {
      "epoch": 0.13536630717738937,
      "grad_norm": 0.15268510580062866,
      "learning_rate": 6.147119600233758e-05,
      "loss": 12.432,
      "step": 91
    },
    {
      "epoch": 0.13685384901450354,
      "grad_norm": 0.18274381756782532,
      "learning_rate": 6.066498153718735e-05,
      "loss": 12.4227,
      "step": 92
    },
    {
      "epoch": 0.1383413908516177,
      "grad_norm": 0.17264324426651,
      "learning_rate": 5.985585137257401e-05,
      "loss": 12.4284,
      "step": 93
    },
    {
      "epoch": 0.13982893268873187,
      "grad_norm": 0.17840361595153809,
      "learning_rate": 5.90440267166055e-05,
      "loss": 12.4325,
      "step": 94
    },
    {
      "epoch": 0.14131647452584603,
      "grad_norm": 0.20858131349086761,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 12.439,
      "step": 95
    },
    {
      "epoch": 0.1428040163629602,
      "grad_norm": 0.19127607345581055,
      "learning_rate": 5.74131823855921e-05,
      "loss": 12.4291,
      "step": 96
    },
    {
      "epoch": 0.1442915582000744,
      "grad_norm": 0.21284298598766327,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 12.419,
      "step": 97
    },
    {
      "epoch": 0.14577910003718855,
      "grad_norm": 0.22342140972614288,
      "learning_rate": 5.577423184847932e-05,
      "loss": 12.424,
      "step": 98
    },
    {
      "epoch": 0.14726664187430272,
      "grad_norm": 0.2617979943752289,
      "learning_rate": 5.495227651252315e-05,
      "loss": 12.4203,
      "step": 99
    },
    {
      "epoch": 0.14875418371141688,
      "grad_norm": 0.3342357575893402,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 12.4199,
      "step": 100
    },
    {
      "epoch": 0.14875418371141688,
      "eval_loss": 12.430908203125,
      "eval_runtime": 8.6509,
      "eval_samples_per_second": 130.853,
      "eval_steps_per_second": 32.713,
      "step": 100
    },
    {
      "epoch": 0.15024172554853105,
      "grad_norm": 0.07840681076049805,
      "learning_rate": 5.330452921628497e-05,
      "loss": 12.4449,
      "step": 101
    },
    {
      "epoch": 0.1517292673856452,
      "grad_norm": 0.07088601589202881,
      "learning_rate": 5.247918773366112e-05,
      "loss": 12.4335,
      "step": 102
    },
    {
      "epoch": 0.1532168092227594,
      "grad_norm": 0.0869700238108635,
      "learning_rate": 5.165316846586541e-05,
      "loss": 12.4387,
      "step": 103
    },
    {
      "epoch": 0.15470435105987357,
      "grad_norm": 0.1161155253648758,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 12.4312,
      "step": 104
    },
    {
      "epoch": 0.15619189289698773,
      "grad_norm": 0.1100078821182251,
      "learning_rate": 5e-05,
      "loss": 12.4237,
      "step": 105
    },
    {
      "epoch": 0.1576794347341019,
      "grad_norm": 0.10071438550949097,
      "learning_rate": 4.917330276168208e-05,
      "loss": 12.4411,
      "step": 106
    },
    {
      "epoch": 0.15916697657121606,
      "grad_norm": 0.10358981043100357,
      "learning_rate": 4.834683153413459e-05,
      "loss": 12.4226,
      "step": 107
    },
    {
      "epoch": 0.16065451840833023,
      "grad_norm": 0.12287390977144241,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 12.4332,
      "step": 108
    },
    {
      "epoch": 0.1621420602454444,
      "grad_norm": 0.1145147979259491,
      "learning_rate": 4.669547078371504e-05,
      "loss": 12.4308,
      "step": 109
    },
    {
      "epoch": 0.16362960208255858,
      "grad_norm": 0.09826528280973434,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 12.433,
      "step": 110
    },
    {
      "epoch": 0.16511714391967275,
      "grad_norm": 0.09297890216112137,
      "learning_rate": 4.504772348747687e-05,
      "loss": 12.422,
      "step": 111
    },
    {
      "epoch": 0.1666046857567869,
      "grad_norm": 0.12921568751335144,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 12.4256,
      "step": 112
    },
    {
      "epoch": 0.16809222759390108,
      "grad_norm": 0.12267349660396576,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 12.4253,
      "step": 113
    },
    {
      "epoch": 0.16957976943101524,
      "grad_norm": 0.12786097824573517,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 12.4314,
      "step": 114
    },
    {
      "epoch": 0.1710673112681294,
      "grad_norm": 0.11280398070812225,
      "learning_rate": 4.17702704859633e-05,
      "loss": 12.4189,
      "step": 115
    },
    {
      "epoch": 0.17255485310524357,
      "grad_norm": 0.09307178109884262,
      "learning_rate": 4.095597328339452e-05,
      "loss": 12.4297,
      "step": 116
    },
    {
      "epoch": 0.17404239494235776,
      "grad_norm": 0.15064336359500885,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 12.4357,
      "step": 117
    },
    {
      "epoch": 0.17552993677947193,
      "grad_norm": 0.11467739194631577,
      "learning_rate": 3.933501846281267e-05,
      "loss": 12.4266,
      "step": 118
    },
    {
      "epoch": 0.1770174786165861,
      "grad_norm": 0.13051538169384003,
      "learning_rate": 3.852880399766243e-05,
      "loss": 12.4321,
      "step": 119
    },
    {
      "epoch": 0.17850502045370026,
      "grad_norm": 0.1229558140039444,
      "learning_rate": 3.772572564296005e-05,
      "loss": 12.4271,
      "step": 120
    },
    {
      "epoch": 0.17999256229081442,
      "grad_norm": 0.13453353941440582,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 12.4222,
      "step": 121
    },
    {
      "epoch": 0.1814801041279286,
      "grad_norm": 0.1455361247062683,
      "learning_rate": 3.612985456190778e-05,
      "loss": 12.4217,
      "step": 122
    },
    {
      "epoch": 0.18296764596504278,
      "grad_norm": 0.13423196971416473,
      "learning_rate": 3.533749813077677e-05,
      "loss": 12.4333,
      "step": 123
    },
    {
      "epoch": 0.18445518780215694,
      "grad_norm": 0.1520964801311493,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 12.4243,
      "step": 124
    },
    {
      "epoch": 0.1859427296392711,
      "grad_norm": 0.1529669165611267,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 12.4263,
      "step": 125
    },
    {
      "epoch": 0.18743027147638527,
      "grad_norm": 0.14568060636520386,
      "learning_rate": 3.298534127791785e-05,
      "loss": 12.4191,
      "step": 126
    },
    {
      "epoch": 0.18891781331349944,
      "grad_norm": 0.12386560440063477,
      "learning_rate": 3.221030765387417e-05,
      "loss": 12.4353,
      "step": 127
    },
    {
      "epoch": 0.1904053551506136,
      "grad_norm": 0.12791499495506287,
      "learning_rate": 3.144013755408895e-05,
      "loss": 12.4199,
      "step": 128
    },
    {
      "epoch": 0.19189289698772777,
      "grad_norm": 0.14995191991329193,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 12.4305,
      "step": 129
    },
    {
      "epoch": 0.19338043882484196,
      "grad_norm": 0.1866651475429535,
      "learning_rate": 2.991522876735154e-05,
      "loss": 12.4211,
      "step": 130
    },
    {
      "epoch": 0.19486798066195612,
      "grad_norm": 0.12598629295825958,
      "learning_rate": 2.916090697523549e-05,
      "loss": 12.4208,
      "step": 131
    },
    {
      "epoch": 0.1963555224990703,
      "grad_norm": 0.17085856199264526,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 12.4235,
      "step": 132
    },
    {
      "epoch": 0.19784306433618445,
      "grad_norm": 0.1857568919658661,
      "learning_rate": 2.766955965735968e-05,
      "loss": 12.4229,
      "step": 133
    },
    {
      "epoch": 0.19933060617329862,
      "grad_norm": 0.1484638750553131,
      "learning_rate": 2.693294185106562e-05,
      "loss": 12.425,
      "step": 134
    },
    {
      "epoch": 0.20081814801041278,
      "grad_norm": 0.15867190062999725,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 12.4269,
      "step": 135
    },
    {
      "epoch": 0.20230568984752695,
      "grad_norm": 0.12770259380340576,
      "learning_rate": 2.547882480847461e-05,
      "loss": 12.4237,
      "step": 136
    },
    {
      "epoch": 0.20379323168464114,
      "grad_norm": 0.13640864193439484,
      "learning_rate": 2.476172311325783e-05,
      "loss": 12.4295,
      "step": 137
    },
    {
      "epoch": 0.2052807735217553,
      "grad_norm": 0.16590316593647003,
      "learning_rate": 2.405152131093926e-05,
      "loss": 12.4114,
      "step": 138
    },
    {
      "epoch": 0.20676831535886947,
      "grad_norm": 0.17572230100631714,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 12.4247,
      "step": 139
    },
    {
      "epoch": 0.20825585719598363,
      "grad_norm": 0.15877486765384674,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 12.4166,
      "step": 140
    },
    {
      "epoch": 0.2097433990330978,
      "grad_norm": 0.1801818609237671,
      "learning_rate": 2.196424713241637e-05,
      "loss": 12.4224,
      "step": 141
    },
    {
      "epoch": 0.21123094087021196,
      "grad_norm": 0.1973048597574234,
      "learning_rate": 2.128356686585282e-05,
      "loss": 12.4164,
      "step": 142
    },
    {
      "epoch": 0.21271848270732616,
      "grad_norm": 0.1884666532278061,
      "learning_rate": 2.061073738537635e-05,
      "loss": 12.4128,
      "step": 143
    },
    {
      "epoch": 0.21420602454444032,
      "grad_norm": 0.20464059710502625,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 12.4155,
      "step": 144
    },
    {
      "epoch": 0.21569356638155449,
      "grad_norm": 0.21521951258182526,
      "learning_rate": 1.928936436551661e-05,
      "loss": 12.4225,
      "step": 145
    },
    {
      "epoch": 0.21718110821866865,
      "grad_norm": 0.2099846452474594,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 12.4075,
      "step": 146
    },
    {
      "epoch": 0.21866865005578282,
      "grad_norm": 0.27252429723739624,
      "learning_rate": 1.800157297483417e-05,
      "loss": 12.4005,
      "step": 147
    },
    {
      "epoch": 0.22015619189289698,
      "grad_norm": 0.27681055665016174,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 12.411,
      "step": 148
    },
    {
      "epoch": 0.22164373373001114,
      "grad_norm": 0.23941662907600403,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 12.3939,
      "step": 149
    },
    {
      "epoch": 0.22313127556712534,
      "grad_norm": 0.3490579426288605,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 12.3896,
      "step": 150
    },
    {
      "epoch": 0.22313127556712534,
      "eval_loss": 12.421548843383789,
      "eval_runtime": 8.6489,
      "eval_samples_per_second": 130.884,
      "eval_steps_per_second": 32.721,
      "step": 150
    },
    {
      "epoch": 0.2246188174042395,
      "grad_norm": 0.084788978099823,
      "learning_rate": 1.553232954407171e-05,
      "loss": 12.4167,
      "step": 151
    },
    {
      "epoch": 0.22610635924135367,
      "grad_norm": 0.08166802674531937,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 12.44,
      "step": 152
    },
    {
      "epoch": 0.22759390107846783,
      "grad_norm": 0.09580064564943314,
      "learning_rate": 1.435357758543015e-05,
      "loss": 12.4369,
      "step": 153
    },
    {
      "epoch": 0.229081442915582,
      "grad_norm": 0.10094612091779709,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 12.4331,
      "step": 154
    },
    {
      "epoch": 0.23056898475269616,
      "grad_norm": 0.10198447853326797,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 12.4219,
      "step": 155
    },
    {
      "epoch": 0.23205652658981032,
      "grad_norm": 0.09021537750959396,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 12.4344,
      "step": 156
    },
    {
      "epoch": 0.23354406842692452,
      "grad_norm": 0.10131917148828506,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 12.4168,
      "step": 157
    },
    {
      "epoch": 0.23503161026403868,
      "grad_norm": 0.1040782704949379,
      "learning_rate": 1.157994445715706e-05,
      "loss": 12.429,
      "step": 158
    },
    {
      "epoch": 0.23651915210115285,
      "grad_norm": 0.11763695627450943,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 12.4191,
      "step": 159
    },
    {
      "epoch": 0.238006693938267,
      "grad_norm": 0.10440471768379211,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 12.4364,
      "step": 160
    },
    {
      "epoch": 0.23949423577538118,
      "grad_norm": 0.13081613183021545,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 12.4298,
      "step": 161
    },
    {
      "epoch": 0.24098177761249534,
      "grad_norm": 0.13708677887916565,
      "learning_rate": 9.549150281252633e-06,
      "loss": 12.4303,
      "step": 162
    },
    {
      "epoch": 0.24246931944960953,
      "grad_norm": 0.10357038676738739,
      "learning_rate": 9.068759265665384e-06,
      "loss": 12.4248,
      "step": 163
    },
    {
      "epoch": 0.2439568612867237,
      "grad_norm": 0.12095082551240921,
      "learning_rate": 8.599558442598998e-06,
      "loss": 12.4264,
      "step": 164
    },
    {
      "epoch": 0.24544440312383786,
      "grad_norm": 0.1381208449602127,
      "learning_rate": 8.141676086873572e-06,
      "loss": 12.404,
      "step": 165
    },
    {
      "epoch": 0.24693194496095203,
      "grad_norm": 0.1601957529783249,
      "learning_rate": 7.695237378953223e-06,
      "loss": 12.4313,
      "step": 166
    },
    {
      "epoch": 0.2484194867980662,
      "grad_norm": 0.10592945665121078,
      "learning_rate": 7.260364370723044e-06,
      "loss": 12.4195,
      "step": 167
    },
    {
      "epoch": 0.24990702863518036,
      "grad_norm": 0.10930639505386353,
      "learning_rate": 6.837175952121306e-06,
      "loss": 12.4396,
      "step": 168
    },
    {
      "epoch": 0.2513945704722945,
      "grad_norm": 0.12032309174537659,
      "learning_rate": 6.425787818636131e-06,
      "loss": 12.4322,
      "step": 169
    },
    {
      "epoch": 0.2528821123094087,
      "grad_norm": 0.12323746830224991,
      "learning_rate": 6.026312439675552e-06,
      "loss": 12.4158,
      "step": 170
    },
    {
      "epoch": 0.25436965414652285,
      "grad_norm": 0.1220136508345604,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 12.425,
      "step": 171
    },
    {
      "epoch": 0.25585719598363704,
      "grad_norm": 0.12643319368362427,
      "learning_rate": 5.263533508961827e-06,
      "loss": 12.4234,
      "step": 172
    },
    {
      "epoch": 0.25734473782075123,
      "grad_norm": 0.1419905424118042,
      "learning_rate": 4.900438493352055e-06,
      "loss": 12.4119,
      "step": 173
    },
    {
      "epoch": 0.25883227965786537,
      "grad_norm": 0.1195748820900917,
      "learning_rate": 4.549673247541875e-06,
      "loss": 12.4191,
      "step": 174
    },
    {
      "epoch": 0.26031982149497956,
      "grad_norm": 0.17883086204528809,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 12.4234,
      "step": 175
    },
    {
      "epoch": 0.2618073633320937,
      "grad_norm": 0.13239678740501404,
      "learning_rate": 3.885512251130763e-06,
      "loss": 12.4151,
      "step": 176
    },
    {
      "epoch": 0.2632949051692079,
      "grad_norm": 0.1445365846157074,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 12.423,
      "step": 177
    },
    {
      "epoch": 0.26478244700632203,
      "grad_norm": 0.14252984523773193,
      "learning_rate": 3.271776770026963e-06,
      "loss": 12.4203,
      "step": 178
    },
    {
      "epoch": 0.2662699888434362,
      "grad_norm": 0.14436165988445282,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 12.4157,
      "step": 179
    },
    {
      "epoch": 0.2677575306805504,
      "grad_norm": 0.17960700392723083,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 12.4281,
      "step": 180
    },
    {
      "epoch": 0.26924507251766455,
      "grad_norm": 0.1522863209247589,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 12.4301,
      "step": 181
    },
    {
      "epoch": 0.27073261435477874,
      "grad_norm": 0.16103801131248474,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 12.4169,
      "step": 182
    },
    {
      "epoch": 0.2722201561918929,
      "grad_norm": 0.20220495760440826,
      "learning_rate": 1.962316193157593e-06,
      "loss": 12.4228,
      "step": 183
    },
    {
      "epoch": 0.2737076980290071,
      "grad_norm": 0.17075666785240173,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 12.4041,
      "step": 184
    },
    {
      "epoch": 0.2751952398661212,
      "grad_norm": 0.16758950054645538,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 12.4181,
      "step": 185
    },
    {
      "epoch": 0.2766827817032354,
      "grad_norm": 0.17145587503910065,
      "learning_rate": 1.333670137599713e-06,
      "loss": 12.414,
      "step": 186
    },
    {
      "epoch": 0.2781703235403496,
      "grad_norm": 0.1732020080089569,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 12.4177,
      "step": 187
    },
    {
      "epoch": 0.27965786537746373,
      "grad_norm": 0.1677519977092743,
      "learning_rate": 9.810017062595322e-07,
      "loss": 12.4192,
      "step": 188
    },
    {
      "epoch": 0.2811454072145779,
      "grad_norm": 0.17412829399108887,
      "learning_rate": 8.247462563808817e-07,
      "loss": 12.4296,
      "step": 189
    },
    {
      "epoch": 0.28263294905169206,
      "grad_norm": 0.19234704971313477,
      "learning_rate": 6.819348298638839e-07,
      "loss": 12.4085,
      "step": 190
    },
    {
      "epoch": 0.28412049088880625,
      "grad_norm": 0.1887720823287964,
      "learning_rate": 5.526064699265753e-07,
      "loss": 12.4035,
      "step": 191
    },
    {
      "epoch": 0.2856080327259204,
      "grad_norm": 0.17081232368946075,
      "learning_rate": 4.367965336512403e-07,
      "loss": 12.4145,
      "step": 192
    },
    {
      "epoch": 0.2870955745630346,
      "grad_norm": 0.22588427364826202,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 12.4138,
      "step": 193
    },
    {
      "epoch": 0.2885831164001488,
      "grad_norm": 0.1740548014640808,
      "learning_rate": 2.458548727494292e-07,
      "loss": 12.4041,
      "step": 194
    },
    {
      "epoch": 0.2900706582372629,
      "grad_norm": 0.26420214772224426,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 12.402,
      "step": 195
    },
    {
      "epoch": 0.2915582000743771,
      "grad_norm": 0.17956629395484924,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 12.4082,
      "step": 196
    },
    {
      "epoch": 0.29304574191149124,
      "grad_norm": 0.25978899002075195,
      "learning_rate": 6.150154258476315e-08,
      "loss": 12.4012,
      "step": 197
    },
    {
      "epoch": 0.29453328374860543,
      "grad_norm": 0.28849345445632935,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 12.4098,
      "step": 198
    },
    {
      "epoch": 0.29602082558571957,
      "grad_norm": 0.2748938798904419,
      "learning_rate": 6.834750376549792e-09,
      "loss": 12.4056,
      "step": 199
    },
    {
      "epoch": 0.29750836742283376,
      "grad_norm": 0.3981285095214844,
      "learning_rate": 0.0,
      "loss": 12.3966,
      "step": 200
    },
    {
      "epoch": 0.29750836742283376,
      "eval_loss": 12.420177459716797,
      "eval_runtime": 8.6493,
      "eval_samples_per_second": 130.878,
      "eval_steps_per_second": 32.72,
      "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": 771332505600.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}