{
  "best_metric": 0.11224672198295593,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.023998800059997,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.000119994000299985,
      "grad_norm": 14.69269847869873,
      "learning_rate": 1e-05,
      "loss": 1.0358,
      "step": 1
    },
    {
      "epoch": 0.000119994000299985,
      "eval_loss": 0.8464587330818176,
      "eval_runtime": 1064.4002,
      "eval_samples_per_second": 13.187,
      "eval_steps_per_second": 3.297,
      "step": 1
    },
    {
      "epoch": 0.00023998800059997,
      "grad_norm": 21.460485458374023,
      "learning_rate": 2e-05,
      "loss": 1.3405,
      "step": 2
    },
    {
      "epoch": 0.000359982000899955,
      "grad_norm": 14.620185852050781,
      "learning_rate": 3e-05,
      "loss": 1.0886,
      "step": 3
    },
    {
      "epoch": 0.00047997600119994,
      "grad_norm": 12.862374305725098,
      "learning_rate": 4e-05,
      "loss": 0.7299,
      "step": 4
    },
    {
      "epoch": 0.000599970001499925,
      "grad_norm": 9.000184059143066,
      "learning_rate": 5e-05,
      "loss": 0.5772,
      "step": 5
    },
    {
      "epoch": 0.00071996400179991,
      "grad_norm": 8.684731483459473,
      "learning_rate": 6e-05,
      "loss": 0.4396,
      "step": 6
    },
    {
      "epoch": 0.000839958002099895,
      "grad_norm": 7.313620090484619,
      "learning_rate": 7e-05,
      "loss": 0.4722,
      "step": 7
    },
    {
      "epoch": 0.00095995200239988,
      "grad_norm": 8.14297103881836,
      "learning_rate": 8e-05,
      "loss": 0.4748,
      "step": 8
    },
    {
      "epoch": 0.001079946002699865,
      "grad_norm": 8.349589347839355,
      "learning_rate": 9e-05,
      "loss": 0.4602,
      "step": 9
    },
    {
      "epoch": 0.00119994000299985,
      "grad_norm": 7.073505878448486,
      "learning_rate": 0.0001,
      "loss": 0.5239,
      "step": 10
    },
    {
      "epoch": 0.001319934003299835,
      "grad_norm": 6.517452716827393,
      "learning_rate": 9.999316524962345e-05,
      "loss": 0.3769,
      "step": 11
    },
    {
      "epoch": 0.00143992800359982,
      "grad_norm": 5.647474765777588,
      "learning_rate": 9.997266286704631e-05,
      "loss": 0.3916,
      "step": 12
    },
    {
      "epoch": 0.001559922003899805,
      "grad_norm": 7.816258430480957,
      "learning_rate": 9.993849845741524e-05,
      "loss": 0.4188,
      "step": 13
    },
    {
      "epoch": 0.00167991600419979,
      "grad_norm": 6.69984769821167,
      "learning_rate": 9.989068136093873e-05,
      "loss": 0.3141,
      "step": 14
    },
    {
      "epoch": 0.001799910004499775,
      "grad_norm": 7.248648643493652,
      "learning_rate": 9.98292246503335e-05,
      "loss": 0.3668,
      "step": 15
    },
    {
      "epoch": 0.00191990400479976,
      "grad_norm": 6.345238208770752,
      "learning_rate": 9.975414512725057e-05,
      "loss": 0.4176,
      "step": 16
    },
    {
      "epoch": 0.002039898005099745,
      "grad_norm": 7.475438594818115,
      "learning_rate": 9.966546331768191e-05,
      "loss": 0.3162,
      "step": 17
    },
    {
      "epoch": 0.00215989200539973,
      "grad_norm": 7.604703903198242,
      "learning_rate": 9.956320346634876e-05,
      "loss": 0.4153,
      "step": 18
    },
    {
      "epoch": 0.002279886005699715,
      "grad_norm": 7.029585361480713,
      "learning_rate": 9.944739353007344e-05,
      "loss": 0.3778,
      "step": 19
    },
    {
      "epoch": 0.0023998800059997,
      "grad_norm": 5.61261510848999,
      "learning_rate": 9.931806517013612e-05,
      "loss": 0.3238,
      "step": 20
    },
    {
      "epoch": 0.002519874006299685,
      "grad_norm": 5.305056571960449,
      "learning_rate": 9.917525374361912e-05,
      "loss": 0.443,
      "step": 21
    },
    {
      "epoch": 0.00263986800659967,
      "grad_norm": 4.506830215454102,
      "learning_rate": 9.901899829374047e-05,
      "loss": 0.3533,
      "step": 22
    },
    {
      "epoch": 0.002759862006899655,
      "grad_norm": 7.422048568725586,
      "learning_rate": 9.884934153917997e-05,
      "loss": 0.3858,
      "step": 23
    },
    {
      "epoch": 0.00287985600719964,
      "grad_norm": 5.223104000091553,
      "learning_rate": 9.86663298624003e-05,
      "loss": 0.4165,
      "step": 24
    },
    {
      "epoch": 0.002999850007499625,
      "grad_norm": 6.709174633026123,
      "learning_rate": 9.847001329696653e-05,
      "loss": 0.4608,
      "step": 25
    },
    {
      "epoch": 0.00311984400779961,
      "grad_norm": 6.936193943023682,
      "learning_rate": 9.826044551386744e-05,
      "loss": 0.4049,
      "step": 26
    },
    {
      "epoch": 0.003239838008099595,
      "grad_norm": 6.530513286590576,
      "learning_rate": 9.803768380684242e-05,
      "loss": 0.4991,
      "step": 27
    },
    {
      "epoch": 0.00335983200839958,
      "grad_norm": 7.247352123260498,
      "learning_rate": 9.780178907671789e-05,
      "loss": 0.6098,
      "step": 28
    },
    {
      "epoch": 0.003479826008699565,
      "grad_norm": 6.562068462371826,
      "learning_rate": 9.755282581475769e-05,
      "loss": 0.4434,
      "step": 29
    },
    {
      "epoch": 0.00359982000899955,
      "grad_norm": 8.098822593688965,
      "learning_rate": 9.729086208503174e-05,
      "loss": 0.6854,
      "step": 30
    },
    {
      "epoch": 0.003719814009299535,
      "grad_norm": 6.618435859680176,
      "learning_rate": 9.701596950580806e-05,
      "loss": 0.4845,
      "step": 31
    },
    {
      "epoch": 0.00383980800959952,
      "grad_norm": 6.610074996948242,
      "learning_rate": 9.672822322997305e-05,
      "loss": 0.526,
      "step": 32
    },
    {
      "epoch": 0.003959802009899505,
      "grad_norm": 6.292226314544678,
      "learning_rate": 9.642770192448536e-05,
      "loss": 0.4326,
      "step": 33
    },
    {
      "epoch": 0.00407979601019949,
      "grad_norm": 6.810724258422852,
      "learning_rate": 9.611448774886924e-05,
      "loss": 0.563,
      "step": 34
    },
    {
      "epoch": 0.004199790010499475,
      "grad_norm": 5.849240779876709,
      "learning_rate": 9.578866633275288e-05,
      "loss": 0.5525,
      "step": 35
    },
    {
      "epoch": 0.00431978401079946,
      "grad_norm": 6.772653579711914,
      "learning_rate": 9.545032675245813e-05,
      "loss": 0.6247,
      "step": 36
    },
    {
      "epoch": 0.004439778011099445,
      "grad_norm": 6.520594120025635,
      "learning_rate": 9.509956150664796e-05,
      "loss": 0.4617,
      "step": 37
    },
    {
      "epoch": 0.00455977201139943,
      "grad_norm": 7.436050891876221,
      "learning_rate": 9.473646649103818e-05,
      "loss": 0.4841,
      "step": 38
    },
    {
      "epoch": 0.004679766011699415,
      "grad_norm": 6.818819999694824,
      "learning_rate": 9.43611409721806e-05,
      "loss": 0.5131,
      "step": 39
    },
    {
      "epoch": 0.0047997600119994,
      "grad_norm": 10.731443405151367,
      "learning_rate": 9.397368756032445e-05,
      "loss": 0.9989,
      "step": 40
    },
    {
      "epoch": 0.004919754012299385,
      "grad_norm": 10.628727912902832,
      "learning_rate": 9.357421218136386e-05,
      "loss": 0.9415,
      "step": 41
    },
    {
      "epoch": 0.00503974801259937,
      "grad_norm": 11.381721496582031,
      "learning_rate": 9.316282404787871e-05,
      "loss": 1.2441,
      "step": 42
    },
    {
      "epoch": 0.005159742012899355,
      "grad_norm": 13.326120376586914,
      "learning_rate": 9.273963562927695e-05,
      "loss": 1.3748,
      "step": 43
    },
    {
      "epoch": 0.00527973601319934,
      "grad_norm": 13.786096572875977,
      "learning_rate": 9.230476262104677e-05,
      "loss": 1.5034,
      "step": 44
    },
    {
      "epoch": 0.005399730013499325,
      "grad_norm": 11.289826393127441,
      "learning_rate": 9.185832391312644e-05,
      "loss": 1.3799,
      "step": 45
    },
    {
      "epoch": 0.00551972401379931,
      "grad_norm": 11.817275047302246,
      "learning_rate": 9.140044155740101e-05,
      "loss": 1.2506,
      "step": 46
    },
    {
      "epoch": 0.005639718014099295,
      "grad_norm": 13.674202919006348,
      "learning_rate": 9.093124073433463e-05,
      "loss": 1.5622,
      "step": 47
    },
    {
      "epoch": 0.00575971201439928,
      "grad_norm": 13.452319145202637,
      "learning_rate": 9.045084971874738e-05,
      "loss": 1.5993,
      "step": 48
    },
    {
      "epoch": 0.005879706014699265,
      "grad_norm": 16.60169219970703,
      "learning_rate": 8.995939984474624e-05,
      "loss": 2.2226,
      "step": 49
    },
    {
      "epoch": 0.00599970001499925,
      "grad_norm": 20.74740219116211,
      "learning_rate": 8.945702546981969e-05,
      "loss": 2.9004,
      "step": 50
    },
    {
      "epoch": 0.00599970001499925,
      "eval_loss": 0.22461692988872528,
      "eval_runtime": 1070.6554,
      "eval_samples_per_second": 13.11,
      "eval_steps_per_second": 3.277,
      "step": 50
    },
    {
      "epoch": 0.006119694015299235,
      "grad_norm": 9.093111991882324,
      "learning_rate": 8.894386393810563e-05,
      "loss": 0.6062,
      "step": 51
    },
    {
      "epoch": 0.00623968801559922,
      "grad_norm": 5.328884601593018,
      "learning_rate": 8.842005554284296e-05,
      "loss": 0.388,
      "step": 52
    },
    {
      "epoch": 0.006359682015899205,
      "grad_norm": 2.9744913578033447,
      "learning_rate": 8.788574348801675e-05,
      "loss": 0.2727,
      "step": 53
    },
    {
      "epoch": 0.00647967601619919,
      "grad_norm": 3.164220094680786,
      "learning_rate": 8.73410738492077e-05,
      "loss": 0.272,
      "step": 54
    },
    {
      "epoch": 0.006599670016499175,
      "grad_norm": 4.041276454925537,
      "learning_rate": 8.678619553365659e-05,
      "loss": 0.3748,
      "step": 55
    },
    {
      "epoch": 0.00671966401679916,
      "grad_norm": 3.127711296081543,
      "learning_rate": 8.622126023955446e-05,
      "loss": 0.2968,
      "step": 56
    },
    {
      "epoch": 0.006839658017099145,
      "grad_norm": 3.5590827465057373,
      "learning_rate": 8.564642241456986e-05,
      "loss": 0.3505,
      "step": 57
    },
    {
      "epoch": 0.00695965201739913,
      "grad_norm": 3.513152837753296,
      "learning_rate": 8.506183921362443e-05,
      "loss": 0.3191,
      "step": 58
    },
    {
      "epoch": 0.007079646017699115,
      "grad_norm": 3.8481297492980957,
      "learning_rate": 8.44676704559283e-05,
      "loss": 0.3204,
      "step": 59
    },
    {
      "epoch": 0.0071996400179991,
      "grad_norm": 4.410993576049805,
      "learning_rate": 8.386407858128706e-05,
      "loss": 0.3128,
      "step": 60
    },
    {
      "epoch": 0.007319634018299085,
      "grad_norm": 3.5673422813415527,
      "learning_rate": 8.32512286056924e-05,
      "loss": 0.2641,
      "step": 61
    },
    {
      "epoch": 0.00743962801859907,
      "grad_norm": 4.116729736328125,
      "learning_rate": 8.262928807620843e-05,
      "loss": 0.3518,
      "step": 62
    },
    {
      "epoch": 0.007559622018899055,
      "grad_norm": 4.223738670349121,
      "learning_rate": 8.199842702516583e-05,
      "loss": 0.3054,
      "step": 63
    },
    {
      "epoch": 0.00767961601919904,
      "grad_norm": 4.879368782043457,
      "learning_rate": 8.135881792367686e-05,
      "loss": 0.3329,
      "step": 64
    },
    {
      "epoch": 0.007799610019499025,
      "grad_norm": 3.267726421356201,
      "learning_rate": 8.07106356344834e-05,
      "loss": 0.2509,
      "step": 65
    },
    {
      "epoch": 0.00791960401979901,
      "grad_norm": 5.153939723968506,
      "learning_rate": 8.005405736415126e-05,
      "loss": 0.3346,
      "step": 66
    },
    {
      "epoch": 0.008039598020098996,
      "grad_norm": 4.897646903991699,
      "learning_rate": 7.938926261462366e-05,
      "loss": 0.3928,
      "step": 67
    },
    {
      "epoch": 0.00815959202039898,
      "grad_norm": 2.7879748344421387,
      "learning_rate": 7.871643313414718e-05,
      "loss": 0.2094,
      "step": 68
    },
    {
      "epoch": 0.008279586020698965,
      "grad_norm": 5.264479160308838,
      "learning_rate": 7.803575286758364e-05,
      "loss": 0.4125,
      "step": 69
    },
    {
      "epoch": 0.00839958002099895,
      "grad_norm": 5.170749187469482,
      "learning_rate": 7.734740790612136e-05,
      "loss": 0.4631,
      "step": 70
    },
    {
      "epoch": 0.008519574021298935,
      "grad_norm": 3.2926321029663086,
      "learning_rate": 7.66515864363997e-05,
      "loss": 0.21,
      "step": 71
    },
    {
      "epoch": 0.00863956802159892,
      "grad_norm": 3.446126937866211,
      "learning_rate": 7.594847868906076e-05,
      "loss": 0.3863,
      "step": 72
    },
    {
      "epoch": 0.008759562021898905,
      "grad_norm": 3.908745050430298,
      "learning_rate": 7.52382768867422e-05,
      "loss": 0.3701,
      "step": 73
    },
    {
      "epoch": 0.00887955602219889,
      "grad_norm": 4.224360466003418,
      "learning_rate": 7.452117519152542e-05,
      "loss": 0.3327,
      "step": 74
    },
    {
      "epoch": 0.008999550022498875,
      "grad_norm": 4.923535346984863,
      "learning_rate": 7.379736965185368e-05,
      "loss": 0.5206,
      "step": 75
    },
    {
      "epoch": 0.00911954402279886,
      "grad_norm": 5.599193096160889,
      "learning_rate": 7.30670581489344e-05,
      "loss": 0.4075,
      "step": 76
    },
    {
      "epoch": 0.009239538023098844,
      "grad_norm": 3.947330951690674,
      "learning_rate": 7.233044034264034e-05,
      "loss": 0.4062,
      "step": 77
    },
    {
      "epoch": 0.00935953202339883,
      "grad_norm": 4.35064697265625,
      "learning_rate": 7.158771761692464e-05,
      "loss": 0.312,
      "step": 78
    },
    {
      "epoch": 0.009479526023698816,
      "grad_norm": 5.7611846923828125,
      "learning_rate": 7.083909302476453e-05,
      "loss": 0.4996,
      "step": 79
    },
    {
      "epoch": 0.0095995200239988,
      "grad_norm": 4.726752758026123,
      "learning_rate": 7.008477123264848e-05,
      "loss": 0.3406,
      "step": 80
    },
    {
      "epoch": 0.009719514024298785,
      "grad_norm": 4.820328235626221,
      "learning_rate": 6.932495846462261e-05,
      "loss": 0.3615,
      "step": 81
    },
    {
      "epoch": 0.00983950802459877,
      "grad_norm": 4.702832221984863,
      "learning_rate": 6.855986244591104e-05,
      "loss": 0.4232,
      "step": 82
    },
    {
      "epoch": 0.009959502024898755,
      "grad_norm": 6.913530349731445,
      "learning_rate": 6.778969234612584e-05,
      "loss": 0.5485,
      "step": 83
    },
    {
      "epoch": 0.01007949602519874,
      "grad_norm": 4.074461460113525,
      "learning_rate": 6.701465872208216e-05,
      "loss": 0.3024,
      "step": 84
    },
    {
      "epoch": 0.010199490025498725,
      "grad_norm": 5.071674823760986,
      "learning_rate": 6.623497346023418e-05,
      "loss": 0.4288,
      "step": 85
    },
    {
      "epoch": 0.01031948402579871,
      "grad_norm": 6.0075364112854,
      "learning_rate": 6.545084971874738e-05,
      "loss": 0.5281,
      "step": 86
    },
    {
      "epoch": 0.010439478026098695,
      "grad_norm": 6.791200637817383,
      "learning_rate": 6.466250186922325e-05,
      "loss": 0.63,
      "step": 87
    },
    {
      "epoch": 0.01055947202639868,
      "grad_norm": 8.129353523254395,
      "learning_rate": 6.387014543809223e-05,
      "loss": 0.8785,
      "step": 88
    },
    {
      "epoch": 0.010679466026698664,
      "grad_norm": 9.556791305541992,
      "learning_rate": 6.307399704769099e-05,
      "loss": 1.0798,
      "step": 89
    },
    {
      "epoch": 0.01079946002699865,
      "grad_norm": 9.632142066955566,
      "learning_rate": 6.227427435703997e-05,
      "loss": 0.9272,
      "step": 90
    },
    {
      "epoch": 0.010919454027298636,
      "grad_norm": 9.414872169494629,
      "learning_rate": 6.147119600233758e-05,
      "loss": 1.06,
      "step": 91
    },
    {
      "epoch": 0.01103944802759862,
      "grad_norm": 9.823585510253906,
      "learning_rate": 6.066498153718735e-05,
      "loss": 1.2071,
      "step": 92
    },
    {
      "epoch": 0.011159442027898606,
      "grad_norm": 7.364634037017822,
      "learning_rate": 5.985585137257401e-05,
      "loss": 0.8487,
      "step": 93
    },
    {
      "epoch": 0.01127943602819859,
      "grad_norm": 14.220163345336914,
      "learning_rate": 5.90440267166055e-05,
      "loss": 1.1695,
      "step": 94
    },
    {
      "epoch": 0.011399430028498575,
      "grad_norm": 9.03218936920166,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 1.0572,
      "step": 95
    },
    {
      "epoch": 0.01151942402879856,
      "grad_norm": 10.732368469238281,
      "learning_rate": 5.74131823855921e-05,
      "loss": 1.3187,
      "step": 96
    },
    {
      "epoch": 0.011639418029098545,
      "grad_norm": 11.756142616271973,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 1.474,
      "step": 97
    },
    {
      "epoch": 0.01175941202939853,
      "grad_norm": 12.340117454528809,
      "learning_rate": 5.577423184847932e-05,
      "loss": 1.3705,
      "step": 98
    },
    {
      "epoch": 0.011879406029698515,
      "grad_norm": 10.294230461120605,
      "learning_rate": 5.495227651252315e-05,
      "loss": 0.9746,
      "step": 99
    },
    {
      "epoch": 0.0119994000299985,
      "grad_norm": 18.80799102783203,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 2.3955,
      "step": 100
    },
    {
      "epoch": 0.0119994000299985,
      "eval_loss": 0.1388600617647171,
      "eval_runtime": 1071.7445,
      "eval_samples_per_second": 13.096,
      "eval_steps_per_second": 3.274,
      "step": 100
    },
    {
      "epoch": 0.012119394030298485,
      "grad_norm": 3.0107293128967285,
      "learning_rate": 5.330452921628497e-05,
      "loss": 0.3267,
      "step": 101
    },
    {
      "epoch": 0.01223938803059847,
      "grad_norm": 2.807499885559082,
      "learning_rate": 5.247918773366112e-05,
      "loss": 0.2758,
      "step": 102
    },
    {
      "epoch": 0.012359382030898454,
      "grad_norm": 2.1873130798339844,
      "learning_rate": 5.165316846586541e-05,
      "loss": 0.2336,
      "step": 103
    },
    {
      "epoch": 0.01247937603119844,
      "grad_norm": 2.749326467514038,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 0.1999,
      "step": 104
    },
    {
      "epoch": 0.012599370031498426,
      "grad_norm": 2.766964912414551,
      "learning_rate": 5e-05,
      "loss": 0.2654,
      "step": 105
    },
    {
      "epoch": 0.01271936403179841,
      "grad_norm": 3.0591843128204346,
      "learning_rate": 4.917330276168208e-05,
      "loss": 0.2458,
      "step": 106
    },
    {
      "epoch": 0.012839358032098395,
      "grad_norm": 2.790245532989502,
      "learning_rate": 4.834683153413459e-05,
      "loss": 0.1805,
      "step": 107
    },
    {
      "epoch": 0.01295935203239838,
      "grad_norm": 2.3419551849365234,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 0.1608,
      "step": 108
    },
    {
      "epoch": 0.013079346032698365,
      "grad_norm": 2.3921923637390137,
      "learning_rate": 4.669547078371504e-05,
      "loss": 0.1953,
      "step": 109
    },
    {
      "epoch": 0.01319934003299835,
      "grad_norm": 3.998077154159546,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 0.2784,
      "step": 110
    },
    {
      "epoch": 0.013319334033298335,
      "grad_norm": 4.266413688659668,
      "learning_rate": 4.504772348747687e-05,
      "loss": 0.2673,
      "step": 111
    },
    {
      "epoch": 0.01343932803359832,
      "grad_norm": 2.286405324935913,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 0.1714,
      "step": 112
    },
    {
      "epoch": 0.013559322033898305,
      "grad_norm": 2.546656847000122,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 0.2389,
      "step": 113
    },
    {
      "epoch": 0.01367931603419829,
      "grad_norm": 3.516897439956665,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 0.3157,
      "step": 114
    },
    {
      "epoch": 0.013799310034498274,
      "grad_norm": 3.010638952255249,
      "learning_rate": 4.17702704859633e-05,
      "loss": 0.2736,
      "step": 115
    },
    {
      "epoch": 0.01391930403479826,
      "grad_norm": 2.735196590423584,
      "learning_rate": 4.095597328339452e-05,
      "loss": 0.1766,
      "step": 116
    },
    {
      "epoch": 0.014039298035098246,
      "grad_norm": 4.205049514770508,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 0.3083,
      "step": 117
    },
    {
      "epoch": 0.01415929203539823,
      "grad_norm": 3.531165361404419,
      "learning_rate": 3.933501846281267e-05,
      "loss": 0.2603,
      "step": 118
    },
    {
      "epoch": 0.014279286035698216,
      "grad_norm": 2.789916753768921,
      "learning_rate": 3.852880399766243e-05,
      "loss": 0.2346,
      "step": 119
    },
    {
      "epoch": 0.0143992800359982,
      "grad_norm": 3.2219598293304443,
      "learning_rate": 3.772572564296005e-05,
      "loss": 0.3336,
      "step": 120
    },
    {
      "epoch": 0.014519274036298185,
      "grad_norm": 4.451358795166016,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 0.3649,
      "step": 121
    },
    {
      "epoch": 0.01463926803659817,
      "grad_norm": 2.4760546684265137,
      "learning_rate": 3.612985456190778e-05,
      "loss": 0.1778,
      "step": 122
    },
    {
      "epoch": 0.014759262036898155,
      "grad_norm": 3.4722933769226074,
      "learning_rate": 3.533749813077677e-05,
      "loss": 0.2771,
      "step": 123
    },
    {
      "epoch": 0.01487925603719814,
      "grad_norm": 3.670381784439087,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 0.2508,
      "step": 124
    },
    {
      "epoch": 0.014999250037498125,
      "grad_norm": 4.056758880615234,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 0.2619,
      "step": 125
    },
    {
      "epoch": 0.01511924403779811,
      "grad_norm": 3.5981719493865967,
      "learning_rate": 3.298534127791785e-05,
      "loss": 0.2916,
      "step": 126
    },
    {
      "epoch": 0.015239238038098095,
      "grad_norm": 4.264205455780029,
      "learning_rate": 3.221030765387417e-05,
      "loss": 0.3233,
      "step": 127
    },
    {
      "epoch": 0.01535923203839808,
      "grad_norm": 3.2355408668518066,
      "learning_rate": 3.144013755408895e-05,
      "loss": 0.2527,
      "step": 128
    },
    {
      "epoch": 0.015479226038698064,
      "grad_norm": 4.891961574554443,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 0.3636,
      "step": 129
    },
    {
      "epoch": 0.01559922003899805,
      "grad_norm": 2.783611297607422,
      "learning_rate": 2.991522876735154e-05,
      "loss": 0.168,
      "step": 130
    },
    {
      "epoch": 0.015719214039298036,
      "grad_norm": 3.5699586868286133,
      "learning_rate": 2.916090697523549e-05,
      "loss": 0.2869,
      "step": 131
    },
    {
      "epoch": 0.01583920803959802,
      "grad_norm": 4.9510064125061035,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 0.3683,
      "step": 132
    },
    {
      "epoch": 0.015959202039898004,
      "grad_norm": 3.5660650730133057,
      "learning_rate": 2.766955965735968e-05,
      "loss": 0.327,
      "step": 133
    },
    {
      "epoch": 0.01607919604019799,
      "grad_norm": 6.181387424468994,
      "learning_rate": 2.693294185106562e-05,
      "loss": 0.3432,
      "step": 134
    },
    {
      "epoch": 0.016199190040497975,
      "grad_norm": 4.752202987670898,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 0.3134,
      "step": 135
    },
    {
      "epoch": 0.01631918404079796,
      "grad_norm": 6.329928398132324,
      "learning_rate": 2.547882480847461e-05,
      "loss": 0.426,
      "step": 136
    },
    {
      "epoch": 0.016439178041097947,
      "grad_norm": 4.913607597351074,
      "learning_rate": 2.476172311325783e-05,
      "loss": 0.3499,
      "step": 137
    },
    {
      "epoch": 0.01655917204139793,
      "grad_norm": 6.277126789093018,
      "learning_rate": 2.405152131093926e-05,
      "loss": 0.4605,
      "step": 138
    },
    {
      "epoch": 0.016679166041697915,
      "grad_norm": 7.791707992553711,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 0.7761,
      "step": 139
    },
    {
      "epoch": 0.0167991600419979,
      "grad_norm": 8.858154296875,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 0.8383,
      "step": 140
    },
    {
      "epoch": 0.016919154042297886,
      "grad_norm": 8.019481658935547,
      "learning_rate": 2.196424713241637e-05,
      "loss": 0.4747,
      "step": 141
    },
    {
      "epoch": 0.01703914804259787,
      "grad_norm": 5.467211723327637,
      "learning_rate": 2.128356686585282e-05,
      "loss": 0.4703,
      "step": 142
    },
    {
      "epoch": 0.017159142042897854,
      "grad_norm": 9.388840675354004,
      "learning_rate": 2.061073738537635e-05,
      "loss": 1.0332,
      "step": 143
    },
    {
      "epoch": 0.01727913604319784,
      "grad_norm": 8.535136222839355,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 0.9216,
      "step": 144
    },
    {
      "epoch": 0.017399130043497826,
      "grad_norm": 13.282774925231934,
      "learning_rate": 1.928936436551661e-05,
      "loss": 0.9477,
      "step": 145
    },
    {
      "epoch": 0.01751912404379781,
      "grad_norm": 10.821352005004883,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 1.4023,
      "step": 146
    },
    {
      "epoch": 0.017639118044097794,
      "grad_norm": 11.919954299926758,
      "learning_rate": 1.800157297483417e-05,
      "loss": 1.6159,
      "step": 147
    },
    {
      "epoch": 0.01775911204439778,
      "grad_norm": 13.018702507019043,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 1.7414,
      "step": 148
    },
    {
      "epoch": 0.017879106044697765,
      "grad_norm": 16.701339721679688,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 1.7914,
      "step": 149
    },
    {
      "epoch": 0.01799910004499775,
      "grad_norm": 16.10375213623047,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 2.5372,
      "step": 150
    },
    {
      "epoch": 0.01799910004499775,
      "eval_loss": 0.12078536301851273,
      "eval_runtime": 1070.3655,
      "eval_samples_per_second": 13.113,
      "eval_steps_per_second": 3.278,
      "step": 150
    },
    {
      "epoch": 0.018119094045297737,
      "grad_norm": 2.608668804168701,
      "learning_rate": 1.553232954407171e-05,
      "loss": 0.3289,
      "step": 151
    },
    {
      "epoch": 0.01823908804559772,
      "grad_norm": 2.31552791595459,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 0.1797,
      "step": 152
    },
    {
      "epoch": 0.018359082045897705,
      "grad_norm": 2.2959024906158447,
      "learning_rate": 1.435357758543015e-05,
      "loss": 0.1791,
      "step": 153
    },
    {
      "epoch": 0.01847907604619769,
      "grad_norm": 2.9551842212677,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 0.2582,
      "step": 154
    },
    {
      "epoch": 0.018599070046497676,
      "grad_norm": 2.9195237159729004,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 0.1873,
      "step": 155
    },
    {
      "epoch": 0.01871906404679766,
      "grad_norm": 2.9310009479522705,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 0.1992,
      "step": 156
    },
    {
      "epoch": 0.018839058047097644,
      "grad_norm": 3.476837158203125,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 0.2575,
      "step": 157
    },
    {
      "epoch": 0.01895905204739763,
      "grad_norm": 2.1316118240356445,
      "learning_rate": 1.157994445715706e-05,
      "loss": 0.156,
      "step": 158
    },
    {
      "epoch": 0.019079046047697616,
      "grad_norm": 2.770718812942505,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 0.2213,
      "step": 159
    },
    {
      "epoch": 0.0191990400479976,
      "grad_norm": 2.258989095687866,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 0.1466,
      "step": 160
    },
    {
      "epoch": 0.019319034048297584,
      "grad_norm": 2.9024837017059326,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 0.248,
      "step": 161
    },
    {
      "epoch": 0.01943902804859757,
      "grad_norm": 2.896731376647949,
      "learning_rate": 9.549150281252633e-06,
      "loss": 0.2342,
      "step": 162
    },
    {
      "epoch": 0.019559022048897555,
      "grad_norm": 2.872271776199341,
      "learning_rate": 9.068759265665384e-06,
      "loss": 0.3128,
      "step": 163
    },
    {
      "epoch": 0.01967901604919754,
      "grad_norm": 2.630967140197754,
      "learning_rate": 8.599558442598998e-06,
      "loss": 0.1874,
      "step": 164
    },
    {
      "epoch": 0.019799010049497526,
      "grad_norm": 3.22659969329834,
      "learning_rate": 8.141676086873572e-06,
      "loss": 0.2708,
      "step": 165
    },
    {
      "epoch": 0.01991900404979751,
      "grad_norm": 2.946460485458374,
      "learning_rate": 7.695237378953223e-06,
      "loss": 0.2269,
      "step": 166
    },
    {
      "epoch": 0.020038998050097494,
      "grad_norm": 2.2356905937194824,
      "learning_rate": 7.260364370723044e-06,
      "loss": 0.1398,
      "step": 167
    },
    {
      "epoch": 0.02015899205039748,
      "grad_norm": 2.713515520095825,
      "learning_rate": 6.837175952121306e-06,
      "loss": 0.2251,
      "step": 168
    },
    {
      "epoch": 0.020278986050697466,
      "grad_norm": 3.488715171813965,
      "learning_rate": 6.425787818636131e-06,
      "loss": 0.2335,
      "step": 169
    },
    {
      "epoch": 0.02039898005099745,
      "grad_norm": 3.062786340713501,
      "learning_rate": 6.026312439675552e-06,
      "loss": 0.2609,
      "step": 170
    },
    {
      "epoch": 0.020518974051297434,
      "grad_norm": 3.8978452682495117,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 0.3301,
      "step": 171
    },
    {
      "epoch": 0.02063896805159742,
      "grad_norm": 4.458298683166504,
      "learning_rate": 5.263533508961827e-06,
      "loss": 0.2873,
      "step": 172
    },
    {
      "epoch": 0.020758962051897405,
      "grad_norm": 3.616582155227661,
      "learning_rate": 4.900438493352055e-06,
      "loss": 0.2961,
      "step": 173
    },
    {
      "epoch": 0.02087895605219739,
      "grad_norm": 4.008535861968994,
      "learning_rate": 4.549673247541875e-06,
      "loss": 0.3048,
      "step": 174
    },
    {
      "epoch": 0.020998950052497377,
      "grad_norm": 3.1957318782806396,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 0.2554,
      "step": 175
    },
    {
      "epoch": 0.02111894405279736,
      "grad_norm": 4.088643550872803,
      "learning_rate": 3.885512251130763e-06,
      "loss": 0.2835,
      "step": 176
    },
    {
      "epoch": 0.021238938053097345,
      "grad_norm": 4.323953151702881,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 0.3932,
      "step": 177
    },
    {
      "epoch": 0.02135893205339733,
      "grad_norm": 4.531554222106934,
      "learning_rate": 3.271776770026963e-06,
      "loss": 0.365,
      "step": 178
    },
    {
      "epoch": 0.021478926053697316,
      "grad_norm": 3.5299766063690186,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 0.3037,
      "step": 179
    },
    {
      "epoch": 0.0215989200539973,
      "grad_norm": 4.123356819152832,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 0.3214,
      "step": 180
    },
    {
      "epoch": 0.021718914054297284,
      "grad_norm": 3.4844515323638916,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 0.2414,
      "step": 181
    },
    {
      "epoch": 0.021838908054597272,
      "grad_norm": 4.899460315704346,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 0.3566,
      "step": 182
    },
    {
      "epoch": 0.021958902054897256,
      "grad_norm": 3.9767019748687744,
      "learning_rate": 1.962316193157593e-06,
      "loss": 0.3021,
      "step": 183
    },
    {
      "epoch": 0.02207889605519724,
      "grad_norm": 5.092921733856201,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 0.4322,
      "step": 184
    },
    {
      "epoch": 0.022198890055497224,
      "grad_norm": 4.441225051879883,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 0.2508,
      "step": 185
    },
    {
      "epoch": 0.02231888405579721,
      "grad_norm": 5.182182312011719,
      "learning_rate": 1.333670137599713e-06,
      "loss": 0.4555,
      "step": 186
    },
    {
      "epoch": 0.022438878056097195,
      "grad_norm": 5.512903690338135,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 0.5701,
      "step": 187
    },
    {
      "epoch": 0.02255887205639718,
      "grad_norm": 5.51069974899292,
      "learning_rate": 9.810017062595322e-07,
      "loss": 0.5275,
      "step": 188
    },
    {
      "epoch": 0.022678866056697167,
      "grad_norm": 5.107576370239258,
      "learning_rate": 8.247462563808817e-07,
      "loss": 0.4779,
      "step": 189
    },
    {
      "epoch": 0.02279886005699715,
      "grad_norm": 6.986170768737793,
      "learning_rate": 6.819348298638839e-07,
      "loss": 0.655,
      "step": 190
    },
    {
      "epoch": 0.022918854057297135,
      "grad_norm": 8.67822265625,
      "learning_rate": 5.526064699265753e-07,
      "loss": 0.7399,
      "step": 191
    },
    {
      "epoch": 0.02303884805759712,
      "grad_norm": 6.858239650726318,
      "learning_rate": 4.367965336512403e-07,
      "loss": 0.5185,
      "step": 192
    },
    {
      "epoch": 0.023158842057897106,
      "grad_norm": 7.37984561920166,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 0.7377,
      "step": 193
    },
    {
      "epoch": 0.02327883605819709,
      "grad_norm": 9.369331359863281,
      "learning_rate": 2.458548727494292e-07,
      "loss": 0.9378,
      "step": 194
    },
    {
      "epoch": 0.023398830058497074,
      "grad_norm": 8.622685432434082,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 1.1502,
      "step": 195
    },
    {
      "epoch": 0.02351882405879706,
      "grad_norm": 12.027359962463379,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 1.2777,
      "step": 196
    },
    {
      "epoch": 0.023638818059097046,
      "grad_norm": 11.856021881103516,
      "learning_rate": 6.150154258476315e-08,
      "loss": 1.4603,
      "step": 197
    },
    {
      "epoch": 0.02375881205939703,
      "grad_norm": 12.14152717590332,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 1.5095,
      "step": 198
    },
    {
      "epoch": 0.023878806059697014,
      "grad_norm": 14.508707046508789,
      "learning_rate": 6.834750376549792e-09,
      "loss": 1.7037,
      "step": 199
    },
    {
      "epoch": 0.023998800059997,
      "grad_norm": 16.844295501708984,
      "learning_rate": 0.0,
      "loss": 2.1869,
      "step": 200
    },
    {
      "epoch": 0.023998800059997,
      "eval_loss": 0.11224672198295593,
      "eval_runtime": 1070.2217,
      "eval_samples_per_second": 13.115,
      "eval_steps_per_second": 3.279,
      "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": 2.861997628588032e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}