{
  "best_metric": 1.4422272443771362,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.6872852233676976,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.003436426116838488,
      "grad_norm": 5.5355377197265625,
      "learning_rate": 1e-05,
      "loss": 7.2406,
      "step": 1
    },
    {
      "epoch": 0.003436426116838488,
      "eval_loss": 2.125288486480713,
      "eval_runtime": 3.0179,
      "eval_samples_per_second": 162.363,
      "eval_steps_per_second": 40.756,
      "step": 1
    },
    {
      "epoch": 0.006872852233676976,
      "grad_norm": 5.726753234863281,
      "learning_rate": 2e-05,
      "loss": 7.1623,
      "step": 2
    },
    {
      "epoch": 0.010309278350515464,
      "grad_norm": 5.353437900543213,
      "learning_rate": 3e-05,
      "loss": 7.2311,
      "step": 3
    },
    {
      "epoch": 0.013745704467353952,
      "grad_norm": 4.229282855987549,
      "learning_rate": 4e-05,
      "loss": 7.1285,
      "step": 4
    },
    {
      "epoch": 0.01718213058419244,
      "grad_norm": 3.9042792320251465,
      "learning_rate": 5e-05,
      "loss": 7.2782,
      "step": 5
    },
    {
      "epoch": 0.020618556701030927,
      "grad_norm": 4.337910175323486,
      "learning_rate": 6e-05,
      "loss": 7.0836,
      "step": 6
    },
    {
      "epoch": 0.024054982817869417,
      "grad_norm": 5.303788661956787,
      "learning_rate": 7e-05,
      "loss": 7.2594,
      "step": 7
    },
    {
      "epoch": 0.027491408934707903,
      "grad_norm": 4.295900821685791,
      "learning_rate": 8e-05,
      "loss": 6.9564,
      "step": 8
    },
    {
      "epoch": 0.030927835051546393,
      "grad_norm": 4.023930072784424,
      "learning_rate": 9e-05,
      "loss": 7.6314,
      "step": 9
    },
    {
      "epoch": 0.03436426116838488,
      "grad_norm": 4.254703998565674,
      "learning_rate": 0.0001,
      "loss": 7.3141,
      "step": 10
    },
    {
      "epoch": 0.037800687285223365,
      "grad_norm": 4.07050895690918,
      "learning_rate": 9.999316524962345e-05,
      "loss": 7.0516,
      "step": 11
    },
    {
      "epoch": 0.041237113402061855,
      "grad_norm": 3.7074577808380127,
      "learning_rate": 9.997266286704631e-05,
      "loss": 7.2291,
      "step": 12
    },
    {
      "epoch": 0.044673539518900345,
      "grad_norm": 3.841063976287842,
      "learning_rate": 9.993849845741524e-05,
      "loss": 7.1282,
      "step": 13
    },
    {
      "epoch": 0.048109965635738834,
      "grad_norm": 3.5177736282348633,
      "learning_rate": 9.989068136093873e-05,
      "loss": 7.4098,
      "step": 14
    },
    {
      "epoch": 0.05154639175257732,
      "grad_norm": 3.4582695960998535,
      "learning_rate": 9.98292246503335e-05,
      "loss": 6.9884,
      "step": 15
    },
    {
      "epoch": 0.054982817869415807,
      "grad_norm": 3.5553205013275146,
      "learning_rate": 9.975414512725057e-05,
      "loss": 7.2802,
      "step": 16
    },
    {
      "epoch": 0.058419243986254296,
      "grad_norm": 3.652503490447998,
      "learning_rate": 9.966546331768191e-05,
      "loss": 6.9503,
      "step": 17
    },
    {
      "epoch": 0.061855670103092786,
      "grad_norm": 3.754910945892334,
      "learning_rate": 9.956320346634876e-05,
      "loss": 7.0223,
      "step": 18
    },
    {
      "epoch": 0.06529209621993128,
      "grad_norm": 3.3444976806640625,
      "learning_rate": 9.944739353007344e-05,
      "loss": 6.919,
      "step": 19
    },
    {
      "epoch": 0.06872852233676977,
      "grad_norm": 3.85040545463562,
      "learning_rate": 9.931806517013612e-05,
      "loss": 7.015,
      "step": 20
    },
    {
      "epoch": 0.07216494845360824,
      "grad_norm": 3.338449001312256,
      "learning_rate": 9.917525374361912e-05,
      "loss": 6.8339,
      "step": 21
    },
    {
      "epoch": 0.07560137457044673,
      "grad_norm": 3.5139732360839844,
      "learning_rate": 9.901899829374047e-05,
      "loss": 7.1054,
      "step": 22
    },
    {
      "epoch": 0.07903780068728522,
      "grad_norm": 3.338864803314209,
      "learning_rate": 9.884934153917997e-05,
      "loss": 7.0334,
      "step": 23
    },
    {
      "epoch": 0.08247422680412371,
      "grad_norm": 3.5280768871307373,
      "learning_rate": 9.86663298624003e-05,
      "loss": 6.9172,
      "step": 24
    },
    {
      "epoch": 0.0859106529209622,
      "grad_norm": 3.7895593643188477,
      "learning_rate": 9.847001329696653e-05,
      "loss": 7.3838,
      "step": 25
    },
    {
      "epoch": 0.08934707903780069,
      "grad_norm": 4.070635795593262,
      "learning_rate": 9.826044551386744e-05,
      "loss": 7.3445,
      "step": 26
    },
    {
      "epoch": 0.09278350515463918,
      "grad_norm": 3.510694742202759,
      "learning_rate": 9.803768380684242e-05,
      "loss": 7.1226,
      "step": 27
    },
    {
      "epoch": 0.09621993127147767,
      "grad_norm": 3.9243319034576416,
      "learning_rate": 9.780178907671789e-05,
      "loss": 6.834,
      "step": 28
    },
    {
      "epoch": 0.09965635738831616,
      "grad_norm": 4.125380992889404,
      "learning_rate": 9.755282581475769e-05,
      "loss": 7.1731,
      "step": 29
    },
    {
      "epoch": 0.10309278350515463,
      "grad_norm": 3.9437153339385986,
      "learning_rate": 9.729086208503174e-05,
      "loss": 7.0367,
      "step": 30
    },
    {
      "epoch": 0.10652920962199312,
      "grad_norm": 3.753997564315796,
      "learning_rate": 9.701596950580806e-05,
      "loss": 6.7845,
      "step": 31
    },
    {
      "epoch": 0.10996563573883161,
      "grad_norm": 4.580556392669678,
      "learning_rate": 9.672822322997305e-05,
      "loss": 7.0174,
      "step": 32
    },
    {
      "epoch": 0.1134020618556701,
      "grad_norm": 4.476220607757568,
      "learning_rate": 9.642770192448536e-05,
      "loss": 7.0983,
      "step": 33
    },
    {
      "epoch": 0.11683848797250859,
      "grad_norm": 4.243409156799316,
      "learning_rate": 9.611448774886924e-05,
      "loss": 6.7698,
      "step": 34
    },
    {
      "epoch": 0.12027491408934708,
      "grad_norm": 4.007017612457275,
      "learning_rate": 9.578866633275288e-05,
      "loss": 6.918,
      "step": 35
    },
    {
      "epoch": 0.12371134020618557,
      "grad_norm": 4.217367172241211,
      "learning_rate": 9.545032675245813e-05,
      "loss": 6.8922,
      "step": 36
    },
    {
      "epoch": 0.12714776632302405,
      "grad_norm": 4.854759693145752,
      "learning_rate": 9.509956150664796e-05,
      "loss": 7.032,
      "step": 37
    },
    {
      "epoch": 0.13058419243986255,
      "grad_norm": 4.690032958984375,
      "learning_rate": 9.473646649103818e-05,
      "loss": 7.1388,
      "step": 38
    },
    {
      "epoch": 0.13402061855670103,
      "grad_norm": 4.548827648162842,
      "learning_rate": 9.43611409721806e-05,
      "loss": 7.1937,
      "step": 39
    },
    {
      "epoch": 0.13745704467353953,
      "grad_norm": 4.140782833099365,
      "learning_rate": 9.397368756032445e-05,
      "loss": 6.9458,
      "step": 40
    },
    {
      "epoch": 0.140893470790378,
      "grad_norm": 5.5810465812683105,
      "learning_rate": 9.357421218136386e-05,
      "loss": 7.3173,
      "step": 41
    },
    {
      "epoch": 0.14432989690721648,
      "grad_norm": 4.514927387237549,
      "learning_rate": 9.316282404787871e-05,
      "loss": 7.0019,
      "step": 42
    },
    {
      "epoch": 0.14776632302405499,
      "grad_norm": 4.381032943725586,
      "learning_rate": 9.273963562927695e-05,
      "loss": 6.505,
      "step": 43
    },
    {
      "epoch": 0.15120274914089346,
      "grad_norm": 5.786319255828857,
      "learning_rate": 9.230476262104677e-05,
      "loss": 7.4478,
      "step": 44
    },
    {
      "epoch": 0.15463917525773196,
      "grad_norm": 5.37347936630249,
      "learning_rate": 9.185832391312644e-05,
      "loss": 6.7931,
      "step": 45
    },
    {
      "epoch": 0.15807560137457044,
      "grad_norm": 5.375253200531006,
      "learning_rate": 9.140044155740101e-05,
      "loss": 6.9072,
      "step": 46
    },
    {
      "epoch": 0.16151202749140894,
      "grad_norm": 5.934044361114502,
      "learning_rate": 9.093124073433463e-05,
      "loss": 6.4266,
      "step": 47
    },
    {
      "epoch": 0.16494845360824742,
      "grad_norm": 6.617461681365967,
      "learning_rate": 9.045084971874738e-05,
      "loss": 7.1077,
      "step": 48
    },
    {
      "epoch": 0.16838487972508592,
      "grad_norm": 7.583418369293213,
      "learning_rate": 8.995939984474624e-05,
      "loss": 7.0657,
      "step": 49
    },
    {
      "epoch": 0.1718213058419244,
      "grad_norm": 13.168756484985352,
      "learning_rate": 8.945702546981969e-05,
      "loss": 7.0755,
      "step": 50
    },
    {
      "epoch": 0.1718213058419244,
      "eval_loss": 1.6890923976898193,
      "eval_runtime": 3.0745,
      "eval_samples_per_second": 159.375,
      "eval_steps_per_second": 40.006,
      "step": 50
    },
    {
      "epoch": 0.17525773195876287,
      "grad_norm": 26.317058563232422,
      "learning_rate": 8.894386393810563e-05,
      "loss": 7.1542,
      "step": 51
    },
    {
      "epoch": 0.17869415807560138,
      "grad_norm": 20.181238174438477,
      "learning_rate": 8.842005554284296e-05,
      "loss": 7.062,
      "step": 52
    },
    {
      "epoch": 0.18213058419243985,
      "grad_norm": 13.063090324401855,
      "learning_rate": 8.788574348801675e-05,
      "loss": 6.633,
      "step": 53
    },
    {
      "epoch": 0.18556701030927836,
      "grad_norm": 7.912332534790039,
      "learning_rate": 8.73410738492077e-05,
      "loss": 6.7137,
      "step": 54
    },
    {
      "epoch": 0.18900343642611683,
      "grad_norm": 6.21930456161499,
      "learning_rate": 8.678619553365659e-05,
      "loss": 6.346,
      "step": 55
    },
    {
      "epoch": 0.19243986254295534,
      "grad_norm": 6.126475811004639,
      "learning_rate": 8.622126023955446e-05,
      "loss": 6.2751,
      "step": 56
    },
    {
      "epoch": 0.1958762886597938,
      "grad_norm": 5.813158988952637,
      "learning_rate": 8.564642241456986e-05,
      "loss": 6.3434,
      "step": 57
    },
    {
      "epoch": 0.19931271477663232,
      "grad_norm": 5.243718147277832,
      "learning_rate": 8.506183921362443e-05,
      "loss": 6.5097,
      "step": 58
    },
    {
      "epoch": 0.2027491408934708,
      "grad_norm": 4.483036041259766,
      "learning_rate": 8.44676704559283e-05,
      "loss": 6.5788,
      "step": 59
    },
    {
      "epoch": 0.20618556701030927,
      "grad_norm": 3.864474296569824,
      "learning_rate": 8.386407858128706e-05,
      "loss": 6.0119,
      "step": 60
    },
    {
      "epoch": 0.20962199312714777,
      "grad_norm": 4.41072940826416,
      "learning_rate": 8.32512286056924e-05,
      "loss": 6.2114,
      "step": 61
    },
    {
      "epoch": 0.21305841924398625,
      "grad_norm": 4.159519195556641,
      "learning_rate": 8.262928807620843e-05,
      "loss": 6.3664,
      "step": 62
    },
    {
      "epoch": 0.21649484536082475,
      "grad_norm": 4.05870246887207,
      "learning_rate": 8.199842702516583e-05,
      "loss": 6.4329,
      "step": 63
    },
    {
      "epoch": 0.21993127147766323,
      "grad_norm": 4.380086898803711,
      "learning_rate": 8.135881792367686e-05,
      "loss": 6.2169,
      "step": 64
    },
    {
      "epoch": 0.22336769759450173,
      "grad_norm": 4.428918361663818,
      "learning_rate": 8.07106356344834e-05,
      "loss": 6.4895,
      "step": 65
    },
    {
      "epoch": 0.2268041237113402,
      "grad_norm": 3.6982340812683105,
      "learning_rate": 8.005405736415126e-05,
      "loss": 6.2108,
      "step": 66
    },
    {
      "epoch": 0.23024054982817868,
      "grad_norm": 3.8365907669067383,
      "learning_rate": 7.938926261462366e-05,
      "loss": 6.3245,
      "step": 67
    },
    {
      "epoch": 0.23367697594501718,
      "grad_norm": 3.6086442470550537,
      "learning_rate": 7.871643313414718e-05,
      "loss": 6.3426,
      "step": 68
    },
    {
      "epoch": 0.23711340206185566,
      "grad_norm": 3.661900758743286,
      "learning_rate": 7.803575286758364e-05,
      "loss": 5.9977,
      "step": 69
    },
    {
      "epoch": 0.24054982817869416,
      "grad_norm": 3.839407444000244,
      "learning_rate": 7.734740790612136e-05,
      "loss": 5.99,
      "step": 70
    },
    {
      "epoch": 0.24398625429553264,
      "grad_norm": 3.3978495597839355,
      "learning_rate": 7.66515864363997e-05,
      "loss": 6.201,
      "step": 71
    },
    {
      "epoch": 0.24742268041237114,
      "grad_norm": 3.2046682834625244,
      "learning_rate": 7.594847868906076e-05,
      "loss": 6.2305,
      "step": 72
    },
    {
      "epoch": 0.2508591065292096,
      "grad_norm": 3.567070484161377,
      "learning_rate": 7.52382768867422e-05,
      "loss": 6.4727,
      "step": 73
    },
    {
      "epoch": 0.2542955326460481,
      "grad_norm": 3.344513177871704,
      "learning_rate": 7.452117519152542e-05,
      "loss": 6.4742,
      "step": 74
    },
    {
      "epoch": 0.25773195876288657,
      "grad_norm": 3.4348652362823486,
      "learning_rate": 7.379736965185368e-05,
      "loss": 6.3319,
      "step": 75
    },
    {
      "epoch": 0.2611683848797251,
      "grad_norm": 3.7370567321777344,
      "learning_rate": 7.30670581489344e-05,
      "loss": 6.111,
      "step": 76
    },
    {
      "epoch": 0.2646048109965636,
      "grad_norm": 3.6927969455718994,
      "learning_rate": 7.233044034264034e-05,
      "loss": 6.5892,
      "step": 77
    },
    {
      "epoch": 0.26804123711340205,
      "grad_norm": 3.430112838745117,
      "learning_rate": 7.158771761692464e-05,
      "loss": 6.0102,
      "step": 78
    },
    {
      "epoch": 0.27147766323024053,
      "grad_norm": 3.466944932937622,
      "learning_rate": 7.083909302476453e-05,
      "loss": 5.9042,
      "step": 79
    },
    {
      "epoch": 0.27491408934707906,
      "grad_norm": 3.467165946960449,
      "learning_rate": 7.008477123264848e-05,
      "loss": 5.9583,
      "step": 80
    },
    {
      "epoch": 0.27835051546391754,
      "grad_norm": 3.769183397293091,
      "learning_rate": 6.932495846462261e-05,
      "loss": 5.6396,
      "step": 81
    },
    {
      "epoch": 0.281786941580756,
      "grad_norm": 3.625413179397583,
      "learning_rate": 6.855986244591104e-05,
      "loss": 6.2113,
      "step": 82
    },
    {
      "epoch": 0.2852233676975945,
      "grad_norm": 3.946357488632202,
      "learning_rate": 6.778969234612584e-05,
      "loss": 6.4601,
      "step": 83
    },
    {
      "epoch": 0.28865979381443296,
      "grad_norm": 4.32614278793335,
      "learning_rate": 6.701465872208216e-05,
      "loss": 6.4704,
      "step": 84
    },
    {
      "epoch": 0.2920962199312715,
      "grad_norm": 3.818880796432495,
      "learning_rate": 6.623497346023418e-05,
      "loss": 6.4217,
      "step": 85
    },
    {
      "epoch": 0.29553264604810997,
      "grad_norm": 4.1053786277771,
      "learning_rate": 6.545084971874738e-05,
      "loss": 6.3428,
      "step": 86
    },
    {
      "epoch": 0.29896907216494845,
      "grad_norm": 4.126435279846191,
      "learning_rate": 6.466250186922325e-05,
      "loss": 6.4384,
      "step": 87
    },
    {
      "epoch": 0.3024054982817869,
      "grad_norm": 4.1976752281188965,
      "learning_rate": 6.387014543809223e-05,
      "loss": 6.4296,
      "step": 88
    },
    {
      "epoch": 0.30584192439862545,
      "grad_norm": 4.428470611572266,
      "learning_rate": 6.307399704769099e-05,
      "loss": 6.1146,
      "step": 89
    },
    {
      "epoch": 0.30927835051546393,
      "grad_norm": 5.061136245727539,
      "learning_rate": 6.227427435703997e-05,
      "loss": 6.5789,
      "step": 90
    },
    {
      "epoch": 0.3127147766323024,
      "grad_norm": 4.448794841766357,
      "learning_rate": 6.147119600233758e-05,
      "loss": 6.3885,
      "step": 91
    },
    {
      "epoch": 0.3161512027491409,
      "grad_norm": 4.453677654266357,
      "learning_rate": 6.066498153718735e-05,
      "loss": 6.2379,
      "step": 92
    },
    {
      "epoch": 0.31958762886597936,
      "grad_norm": 4.476033687591553,
      "learning_rate": 5.985585137257401e-05,
      "loss": 6.3481,
      "step": 93
    },
    {
      "epoch": 0.3230240549828179,
      "grad_norm": 5.003313064575195,
      "learning_rate": 5.90440267166055e-05,
      "loss": 6.3404,
      "step": 94
    },
    {
      "epoch": 0.32646048109965636,
      "grad_norm": 4.849445343017578,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 6.1549,
      "step": 95
    },
    {
      "epoch": 0.32989690721649484,
      "grad_norm": 5.423452854156494,
      "learning_rate": 5.74131823855921e-05,
      "loss": 5.8565,
      "step": 96
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 5.645740509033203,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 6.2893,
      "step": 97
    },
    {
      "epoch": 0.33676975945017185,
      "grad_norm": 5.822202682495117,
      "learning_rate": 5.577423184847932e-05,
      "loss": 6.5218,
      "step": 98
    },
    {
      "epoch": 0.3402061855670103,
      "grad_norm": 7.405486106872559,
      "learning_rate": 5.495227651252315e-05,
      "loss": 6.8686,
      "step": 99
    },
    {
      "epoch": 0.3436426116838488,
      "grad_norm": 10.680353164672852,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 7.1631,
      "step": 100
    },
    {
      "epoch": 0.3436426116838488,
      "eval_loss": 1.5108217000961304,
      "eval_runtime": 3.0069,
      "eval_samples_per_second": 162.96,
      "eval_steps_per_second": 40.906,
      "step": 100
    },
    {
      "epoch": 0.3470790378006873,
      "grad_norm": 6.675581932067871,
      "learning_rate": 5.330452921628497e-05,
      "loss": 6.2955,
      "step": 101
    },
    {
      "epoch": 0.35051546391752575,
      "grad_norm": 5.6114501953125,
      "learning_rate": 5.247918773366112e-05,
      "loss": 6.1979,
      "step": 102
    },
    {
      "epoch": 0.3539518900343643,
      "grad_norm": 5.068195819854736,
      "learning_rate": 5.165316846586541e-05,
      "loss": 5.895,
      "step": 103
    },
    {
      "epoch": 0.35738831615120276,
      "grad_norm": 3.9213099479675293,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 6.0156,
      "step": 104
    },
    {
      "epoch": 0.36082474226804123,
      "grad_norm": 3.551356077194214,
      "learning_rate": 5e-05,
      "loss": 5.896,
      "step": 105
    },
    {
      "epoch": 0.3642611683848797,
      "grad_norm": 3.5604190826416016,
      "learning_rate": 4.917330276168208e-05,
      "loss": 5.9823,
      "step": 106
    },
    {
      "epoch": 0.36769759450171824,
      "grad_norm": 3.683518648147583,
      "learning_rate": 4.834683153413459e-05,
      "loss": 5.895,
      "step": 107
    },
    {
      "epoch": 0.3711340206185567,
      "grad_norm": 3.451303720474243,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 5.9957,
      "step": 108
    },
    {
      "epoch": 0.3745704467353952,
      "grad_norm": 3.6081595420837402,
      "learning_rate": 4.669547078371504e-05,
      "loss": 5.8042,
      "step": 109
    },
    {
      "epoch": 0.37800687285223367,
      "grad_norm": 3.2883875370025635,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 5.8861,
      "step": 110
    },
    {
      "epoch": 0.38144329896907214,
      "grad_norm": 3.2161431312561035,
      "learning_rate": 4.504772348747687e-05,
      "loss": 5.8288,
      "step": 111
    },
    {
      "epoch": 0.3848797250859107,
      "grad_norm": 3.311727523803711,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 6.0077,
      "step": 112
    },
    {
      "epoch": 0.38831615120274915,
      "grad_norm": 3.2296974658966064,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 5.8654,
      "step": 113
    },
    {
      "epoch": 0.3917525773195876,
      "grad_norm": 3.3987650871276855,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 6.1821,
      "step": 114
    },
    {
      "epoch": 0.3951890034364261,
      "grad_norm": 3.1602210998535156,
      "learning_rate": 4.17702704859633e-05,
      "loss": 6.1334,
      "step": 115
    },
    {
      "epoch": 0.39862542955326463,
      "grad_norm": 3.3136990070343018,
      "learning_rate": 4.095597328339452e-05,
      "loss": 5.6617,
      "step": 116
    },
    {
      "epoch": 0.4020618556701031,
      "grad_norm": 3.3616576194763184,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 5.9355,
      "step": 117
    },
    {
      "epoch": 0.4054982817869416,
      "grad_norm": 3.1655445098876953,
      "learning_rate": 3.933501846281267e-05,
      "loss": 6.1473,
      "step": 118
    },
    {
      "epoch": 0.40893470790378006,
      "grad_norm": 3.243842363357544,
      "learning_rate": 3.852880399766243e-05,
      "loss": 6.1438,
      "step": 119
    },
    {
      "epoch": 0.41237113402061853,
      "grad_norm": 3.168661594390869,
      "learning_rate": 3.772572564296005e-05,
      "loss": 5.8414,
      "step": 120
    },
    {
      "epoch": 0.41580756013745707,
      "grad_norm": 3.358571767807007,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 5.964,
      "step": 121
    },
    {
      "epoch": 0.41924398625429554,
      "grad_norm": 3.3115599155426025,
      "learning_rate": 3.612985456190778e-05,
      "loss": 6.2824,
      "step": 122
    },
    {
      "epoch": 0.422680412371134,
      "grad_norm": 3.0641727447509766,
      "learning_rate": 3.533749813077677e-05,
      "loss": 6.0253,
      "step": 123
    },
    {
      "epoch": 0.4261168384879725,
      "grad_norm": 3.083996534347534,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 5.925,
      "step": 124
    },
    {
      "epoch": 0.42955326460481097,
      "grad_norm": 3.3683671951293945,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 5.7022,
      "step": 125
    },
    {
      "epoch": 0.4329896907216495,
      "grad_norm": 3.336850166320801,
      "learning_rate": 3.298534127791785e-05,
      "loss": 6.0412,
      "step": 126
    },
    {
      "epoch": 0.436426116838488,
      "grad_norm": 3.226863145828247,
      "learning_rate": 3.221030765387417e-05,
      "loss": 6.1091,
      "step": 127
    },
    {
      "epoch": 0.43986254295532645,
      "grad_norm": 3.6488137245178223,
      "learning_rate": 3.144013755408895e-05,
      "loss": 6.333,
      "step": 128
    },
    {
      "epoch": 0.44329896907216493,
      "grad_norm": 3.4202322959899902,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 5.9763,
      "step": 129
    },
    {
      "epoch": 0.44673539518900346,
      "grad_norm": 3.4294393062591553,
      "learning_rate": 2.991522876735154e-05,
      "loss": 6.1592,
      "step": 130
    },
    {
      "epoch": 0.45017182130584193,
      "grad_norm": 3.5797719955444336,
      "learning_rate": 2.916090697523549e-05,
      "loss": 5.9032,
      "step": 131
    },
    {
      "epoch": 0.4536082474226804,
      "grad_norm": 3.3395729064941406,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 5.9443,
      "step": 132
    },
    {
      "epoch": 0.4570446735395189,
      "grad_norm": 3.6710522174835205,
      "learning_rate": 2.766955965735968e-05,
      "loss": 6.3763,
      "step": 133
    },
    {
      "epoch": 0.46048109965635736,
      "grad_norm": 3.664593458175659,
      "learning_rate": 2.693294185106562e-05,
      "loss": 6.2203,
      "step": 134
    },
    {
      "epoch": 0.4639175257731959,
      "grad_norm": 4.004647254943848,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 6.0376,
      "step": 135
    },
    {
      "epoch": 0.46735395189003437,
      "grad_norm": 4.148707389831543,
      "learning_rate": 2.547882480847461e-05,
      "loss": 5.9936,
      "step": 136
    },
    {
      "epoch": 0.47079037800687284,
      "grad_norm": 3.9335989952087402,
      "learning_rate": 2.476172311325783e-05,
      "loss": 5.9218,
      "step": 137
    },
    {
      "epoch": 0.4742268041237113,
      "grad_norm": 4.188965797424316,
      "learning_rate": 2.405152131093926e-05,
      "loss": 6.355,
      "step": 138
    },
    {
      "epoch": 0.47766323024054985,
      "grad_norm": 3.9909534454345703,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 5.5585,
      "step": 139
    },
    {
      "epoch": 0.48109965635738833,
      "grad_norm": 4.091626167297363,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 6.1739,
      "step": 140
    },
    {
      "epoch": 0.4845360824742268,
      "grad_norm": 4.292383193969727,
      "learning_rate": 2.196424713241637e-05,
      "loss": 6.1337,
      "step": 141
    },
    {
      "epoch": 0.4879725085910653,
      "grad_norm": 4.317835330963135,
      "learning_rate": 2.128356686585282e-05,
      "loss": 5.9295,
      "step": 142
    },
    {
      "epoch": 0.49140893470790376,
      "grad_norm": 4.203305721282959,
      "learning_rate": 2.061073738537635e-05,
      "loss": 5.7966,
      "step": 143
    },
    {
      "epoch": 0.4948453608247423,
      "grad_norm": 4.6097893714904785,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 6.0458,
      "step": 144
    },
    {
      "epoch": 0.49828178694158076,
      "grad_norm": 5.053816318511963,
      "learning_rate": 1.928936436551661e-05,
      "loss": 6.0613,
      "step": 145
    },
    {
      "epoch": 0.5017182130584192,
      "grad_norm": 4.852388381958008,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 5.8047,
      "step": 146
    },
    {
      "epoch": 0.5051546391752577,
      "grad_norm": 5.086940765380859,
      "learning_rate": 1.800157297483417e-05,
      "loss": 6.3277,
      "step": 147
    },
    {
      "epoch": 0.5085910652920962,
      "grad_norm": 5.872974872589111,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 6.2816,
      "step": 148
    },
    {
      "epoch": 0.5120274914089347,
      "grad_norm": 6.175573348999023,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 6.3846,
      "step": 149
    },
    {
      "epoch": 0.5154639175257731,
      "grad_norm": 9.520306587219238,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 6.511,
      "step": 150
    },
    {
      "epoch": 0.5154639175257731,
      "eval_loss": 1.4620585441589355,
      "eval_runtime": 2.9993,
      "eval_samples_per_second": 163.37,
      "eval_steps_per_second": 41.009,
      "step": 150
    },
    {
      "epoch": 0.5189003436426117,
      "grad_norm": 4.624889373779297,
      "learning_rate": 1.553232954407171e-05,
      "loss": 5.8534,
      "step": 151
    },
    {
      "epoch": 0.5223367697594502,
      "grad_norm": 4.521676540374756,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 5.545,
      "step": 152
    },
    {
      "epoch": 0.5257731958762887,
      "grad_norm": 4.373575687408447,
      "learning_rate": 1.435357758543015e-05,
      "loss": 5.8098,
      "step": 153
    },
    {
      "epoch": 0.5292096219931272,
      "grad_norm": 3.853301763534546,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 5.8018,
      "step": 154
    },
    {
      "epoch": 0.5326460481099656,
      "grad_norm": 3.5780298709869385,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 5.7234,
      "step": 155
    },
    {
      "epoch": 0.5360824742268041,
      "grad_norm": 3.484659433364868,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 5.6648,
      "step": 156
    },
    {
      "epoch": 0.5395189003436426,
      "grad_norm": 3.2986772060394287,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 5.8196,
      "step": 157
    },
    {
      "epoch": 0.5429553264604811,
      "grad_norm": 3.1305108070373535,
      "learning_rate": 1.157994445715706e-05,
      "loss": 5.3952,
      "step": 158
    },
    {
      "epoch": 0.5463917525773195,
      "grad_norm": 3.20410418510437,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 5.5052,
      "step": 159
    },
    {
      "epoch": 0.5498281786941581,
      "grad_norm": 2.9345285892486572,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 5.5852,
      "step": 160
    },
    {
      "epoch": 0.5532646048109966,
      "grad_norm": 3.1026036739349365,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 5.6665,
      "step": 161
    },
    {
      "epoch": 0.5567010309278351,
      "grad_norm": 3.1828982830047607,
      "learning_rate": 9.549150281252633e-06,
      "loss": 5.5818,
      "step": 162
    },
    {
      "epoch": 0.5601374570446735,
      "grad_norm": 3.118027925491333,
      "learning_rate": 9.068759265665384e-06,
      "loss": 6.0083,
      "step": 163
    },
    {
      "epoch": 0.563573883161512,
      "grad_norm": 3.262162208557129,
      "learning_rate": 8.599558442598998e-06,
      "loss": 5.9832,
      "step": 164
    },
    {
      "epoch": 0.5670103092783505,
      "grad_norm": 3.092862129211426,
      "learning_rate": 8.141676086873572e-06,
      "loss": 5.4643,
      "step": 165
    },
    {
      "epoch": 0.570446735395189,
      "grad_norm": 3.293733835220337,
      "learning_rate": 7.695237378953223e-06,
      "loss": 5.8818,
      "step": 166
    },
    {
      "epoch": 0.5738831615120275,
      "grad_norm": 3.2925784587860107,
      "learning_rate": 7.260364370723044e-06,
      "loss": 5.8372,
      "step": 167
    },
    {
      "epoch": 0.5773195876288659,
      "grad_norm": 3.1014089584350586,
      "learning_rate": 6.837175952121306e-06,
      "loss": 5.821,
      "step": 168
    },
    {
      "epoch": 0.5807560137457045,
      "grad_norm": 3.2673656940460205,
      "learning_rate": 6.425787818636131e-06,
      "loss": 5.9251,
      "step": 169
    },
    {
      "epoch": 0.584192439862543,
      "grad_norm": 3.169187307357788,
      "learning_rate": 6.026312439675552e-06,
      "loss": 5.9423,
      "step": 170
    },
    {
      "epoch": 0.5876288659793815,
      "grad_norm": 3.117835760116577,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 5.9387,
      "step": 171
    },
    {
      "epoch": 0.5910652920962199,
      "grad_norm": 3.1760330200195312,
      "learning_rate": 5.263533508961827e-06,
      "loss": 5.8594,
      "step": 172
    },
    {
      "epoch": 0.5945017182130584,
      "grad_norm": 3.24760103225708,
      "learning_rate": 4.900438493352055e-06,
      "loss": 5.7606,
      "step": 173
    },
    {
      "epoch": 0.5979381443298969,
      "grad_norm": 3.240525245666504,
      "learning_rate": 4.549673247541875e-06,
      "loss": 5.4602,
      "step": 174
    },
    {
      "epoch": 0.6013745704467354,
      "grad_norm": 3.4659368991851807,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 5.5783,
      "step": 175
    },
    {
      "epoch": 0.6048109965635738,
      "grad_norm": 3.069030284881592,
      "learning_rate": 3.885512251130763e-06,
      "loss": 5.4955,
      "step": 176
    },
    {
      "epoch": 0.6082474226804123,
      "grad_norm": 3.4746615886688232,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 5.8842,
      "step": 177
    },
    {
      "epoch": 0.6116838487972509,
      "grad_norm": 3.5092272758483887,
      "learning_rate": 3.271776770026963e-06,
      "loss": 5.9466,
      "step": 178
    },
    {
      "epoch": 0.6151202749140894,
      "grad_norm": 3.783738851547241,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 6.1775,
      "step": 179
    },
    {
      "epoch": 0.6185567010309279,
      "grad_norm": 3.741182565689087,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 6.0154,
      "step": 180
    },
    {
      "epoch": 0.6219931271477663,
      "grad_norm": 3.708571195602417,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 6.1505,
      "step": 181
    },
    {
      "epoch": 0.6254295532646048,
      "grad_norm": 3.6742103099823,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 5.63,
      "step": 182
    },
    {
      "epoch": 0.6288659793814433,
      "grad_norm": 3.710308790206909,
      "learning_rate": 1.962316193157593e-06,
      "loss": 5.8234,
      "step": 183
    },
    {
      "epoch": 0.6323024054982818,
      "grad_norm": 3.702098846435547,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 5.8575,
      "step": 184
    },
    {
      "epoch": 0.6357388316151202,
      "grad_norm": 3.6370420455932617,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 5.8737,
      "step": 185
    },
    {
      "epoch": 0.6391752577319587,
      "grad_norm": 4.125608444213867,
      "learning_rate": 1.333670137599713e-06,
      "loss": 6.3107,
      "step": 186
    },
    {
      "epoch": 0.6426116838487973,
      "grad_norm": 3.9875564575195312,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 6.3561,
      "step": 187
    },
    {
      "epoch": 0.6460481099656358,
      "grad_norm": 4.156374931335449,
      "learning_rate": 9.810017062595322e-07,
      "loss": 6.2221,
      "step": 188
    },
    {
      "epoch": 0.6494845360824743,
      "grad_norm": 4.256380081176758,
      "learning_rate": 8.247462563808817e-07,
      "loss": 5.5288,
      "step": 189
    },
    {
      "epoch": 0.6529209621993127,
      "grad_norm": 4.587343692779541,
      "learning_rate": 6.819348298638839e-07,
      "loss": 5.7013,
      "step": 190
    },
    {
      "epoch": 0.6563573883161512,
      "grad_norm": 4.451593399047852,
      "learning_rate": 5.526064699265753e-07,
      "loss": 6.0763,
      "step": 191
    },
    {
      "epoch": 0.6597938144329897,
      "grad_norm": 4.318260192871094,
      "learning_rate": 4.367965336512403e-07,
      "loss": 6.0544,
      "step": 192
    },
    {
      "epoch": 0.6632302405498282,
      "grad_norm": 4.781836032867432,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 5.8835,
      "step": 193
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 4.952905654907227,
      "learning_rate": 2.458548727494292e-07,
      "loss": 5.9851,
      "step": 194
    },
    {
      "epoch": 0.6701030927835051,
      "grad_norm": 5.504292964935303,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 6.2795,
      "step": 195
    },
    {
      "epoch": 0.6735395189003437,
      "grad_norm": 5.493088245391846,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 5.8648,
      "step": 196
    },
    {
      "epoch": 0.6769759450171822,
      "grad_norm": 6.559046268463135,
      "learning_rate": 6.150154258476315e-08,
      "loss": 5.646,
      "step": 197
    },
    {
      "epoch": 0.6804123711340206,
      "grad_norm": 6.614935874938965,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 6.0554,
      "step": 198
    },
    {
      "epoch": 0.6838487972508591,
      "grad_norm": 8.228913307189941,
      "learning_rate": 6.834750376549792e-09,
      "loss": 6.3254,
      "step": 199
    },
    {
      "epoch": 0.6872852233676976,
      "grad_norm": 13.58453369140625,
      "learning_rate": 0.0,
      "loss": 7.2507,
      "step": 200
    },
    {
      "epoch": 0.6872852233676976,
      "eval_loss": 1.4422272443771362,
      "eval_runtime": 2.9891,
      "eval_samples_per_second": 163.928,
      "eval_steps_per_second": 41.149,
      "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": 3936001739194368.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}