{
  "best_metric": 1.3939571380615234,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 1.019108280254777,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005095541401273885,
      "grad_norm": 1.6865113973617554,
      "learning_rate": 1e-05,
      "loss": 1.7303,
      "step": 1
    },
    {
      "epoch": 0.005095541401273885,
      "eval_loss": 3.259193181991577,
      "eval_runtime": 24.0921,
      "eval_samples_per_second": 13.739,
      "eval_steps_per_second": 3.445,
      "step": 1
    },
    {
      "epoch": 0.01019108280254777,
      "grad_norm": 1.5365358591079712,
      "learning_rate": 2e-05,
      "loss": 1.9211,
      "step": 2
    },
    {
      "epoch": 0.015286624203821656,
      "grad_norm": 1.6096848249435425,
      "learning_rate": 3e-05,
      "loss": 1.8868,
      "step": 3
    },
    {
      "epoch": 0.02038216560509554,
      "grad_norm": 1.564287781715393,
      "learning_rate": 4e-05,
      "loss": 2.0186,
      "step": 4
    },
    {
      "epoch": 0.025477707006369428,
      "grad_norm": 1.5923267602920532,
      "learning_rate": 5e-05,
      "loss": 2.1118,
      "step": 5
    },
    {
      "epoch": 0.030573248407643312,
      "grad_norm": 1.6752326488494873,
      "learning_rate": 6e-05,
      "loss": 2.0342,
      "step": 6
    },
    {
      "epoch": 0.035668789808917196,
      "grad_norm": 1.8032734394073486,
      "learning_rate": 7e-05,
      "loss": 2.1314,
      "step": 7
    },
    {
      "epoch": 0.04076433121019108,
      "grad_norm": 1.8827780485153198,
      "learning_rate": 8e-05,
      "loss": 2.0909,
      "step": 8
    },
    {
      "epoch": 0.045859872611464965,
      "grad_norm": 1.659525990486145,
      "learning_rate": 9e-05,
      "loss": 1.9336,
      "step": 9
    },
    {
      "epoch": 0.050955414012738856,
      "grad_norm": 2.1119885444641113,
      "learning_rate": 0.0001,
      "loss": 1.8145,
      "step": 10
    },
    {
      "epoch": 0.05605095541401274,
      "grad_norm": 1.7493677139282227,
      "learning_rate": 9.999316524962345e-05,
      "loss": 1.625,
      "step": 11
    },
    {
      "epoch": 0.061146496815286625,
      "grad_norm": 2.1680750846862793,
      "learning_rate": 9.997266286704631e-05,
      "loss": 1.5844,
      "step": 12
    },
    {
      "epoch": 0.06624203821656051,
      "grad_norm": 1.9492312669754028,
      "learning_rate": 9.993849845741524e-05,
      "loss": 1.5817,
      "step": 13
    },
    {
      "epoch": 0.07133757961783439,
      "grad_norm": 1.6355711221694946,
      "learning_rate": 9.989068136093873e-05,
      "loss": 1.5197,
      "step": 14
    },
    {
      "epoch": 0.07643312101910828,
      "grad_norm": 1.529719591140747,
      "learning_rate": 9.98292246503335e-05,
      "loss": 1.5543,
      "step": 15
    },
    {
      "epoch": 0.08152866242038216,
      "grad_norm": 1.396162509918213,
      "learning_rate": 9.975414512725057e-05,
      "loss": 1.5315,
      "step": 16
    },
    {
      "epoch": 0.08662420382165605,
      "grad_norm": 1.388858675956726,
      "learning_rate": 9.966546331768191e-05,
      "loss": 1.5826,
      "step": 17
    },
    {
      "epoch": 0.09171974522292993,
      "grad_norm": 1.3095834255218506,
      "learning_rate": 9.956320346634876e-05,
      "loss": 1.518,
      "step": 18
    },
    {
      "epoch": 0.09681528662420383,
      "grad_norm": 1.4588730335235596,
      "learning_rate": 9.944739353007344e-05,
      "loss": 1.5464,
      "step": 19
    },
    {
      "epoch": 0.10191082802547771,
      "grad_norm": 1.393801212310791,
      "learning_rate": 9.931806517013612e-05,
      "loss": 1.5825,
      "step": 20
    },
    {
      "epoch": 0.1070063694267516,
      "grad_norm": 1.3014535903930664,
      "learning_rate": 9.917525374361912e-05,
      "loss": 1.4673,
      "step": 21
    },
    {
      "epoch": 0.11210191082802548,
      "grad_norm": 1.2713675498962402,
      "learning_rate": 9.901899829374047e-05,
      "loss": 1.4179,
      "step": 22
    },
    {
      "epoch": 0.11719745222929936,
      "grad_norm": 1.3339929580688477,
      "learning_rate": 9.884934153917997e-05,
      "loss": 1.4382,
      "step": 23
    },
    {
      "epoch": 0.12229299363057325,
      "grad_norm": 1.496106505393982,
      "learning_rate": 9.86663298624003e-05,
      "loss": 1.4899,
      "step": 24
    },
    {
      "epoch": 0.12738853503184713,
      "grad_norm": 1.4896008968353271,
      "learning_rate": 9.847001329696653e-05,
      "loss": 1.5836,
      "step": 25
    },
    {
      "epoch": 0.13248407643312102,
      "grad_norm": 1.593266248703003,
      "learning_rate": 9.826044551386744e-05,
      "loss": 1.5117,
      "step": 26
    },
    {
      "epoch": 0.1375796178343949,
      "grad_norm": 1.4740333557128906,
      "learning_rate": 9.803768380684242e-05,
      "loss": 1.4523,
      "step": 27
    },
    {
      "epoch": 0.14267515923566879,
      "grad_norm": 1.578808069229126,
      "learning_rate": 9.780178907671789e-05,
      "loss": 1.4339,
      "step": 28
    },
    {
      "epoch": 0.14777070063694267,
      "grad_norm": 1.60964035987854,
      "learning_rate": 9.755282581475769e-05,
      "loss": 1.4432,
      "step": 29
    },
    {
      "epoch": 0.15286624203821655,
      "grad_norm": 1.485573649406433,
      "learning_rate": 9.729086208503174e-05,
      "loss": 1.3853,
      "step": 30
    },
    {
      "epoch": 0.15796178343949044,
      "grad_norm": 1.4289450645446777,
      "learning_rate": 9.701596950580806e-05,
      "loss": 1.4828,
      "step": 31
    },
    {
      "epoch": 0.16305732484076432,
      "grad_norm": 1.4461684226989746,
      "learning_rate": 9.672822322997305e-05,
      "loss": 1.5164,
      "step": 32
    },
    {
      "epoch": 0.1681528662420382,
      "grad_norm": 1.5229640007019043,
      "learning_rate": 9.642770192448536e-05,
      "loss": 1.5432,
      "step": 33
    },
    {
      "epoch": 0.1732484076433121,
      "grad_norm": 1.5570660829544067,
      "learning_rate": 9.611448774886924e-05,
      "loss": 1.6238,
      "step": 34
    },
    {
      "epoch": 0.17834394904458598,
      "grad_norm": 1.4717493057250977,
      "learning_rate": 9.578866633275288e-05,
      "loss": 1.457,
      "step": 35
    },
    {
      "epoch": 0.18343949044585986,
      "grad_norm": 1.5823116302490234,
      "learning_rate": 9.545032675245813e-05,
      "loss": 1.5628,
      "step": 36
    },
    {
      "epoch": 0.18853503184713377,
      "grad_norm": 1.481979250907898,
      "learning_rate": 9.509956150664796e-05,
      "loss": 1.4876,
      "step": 37
    },
    {
      "epoch": 0.19363057324840766,
      "grad_norm": 1.7308721542358398,
      "learning_rate": 9.473646649103818e-05,
      "loss": 1.5297,
      "step": 38
    },
    {
      "epoch": 0.19872611464968154,
      "grad_norm": 1.5279890298843384,
      "learning_rate": 9.43611409721806e-05,
      "loss": 1.5549,
      "step": 39
    },
    {
      "epoch": 0.20382165605095542,
      "grad_norm": 1.7055156230926514,
      "learning_rate": 9.397368756032445e-05,
      "loss": 1.5352,
      "step": 40
    },
    {
      "epoch": 0.2089171974522293,
      "grad_norm": 1.6564401388168335,
      "learning_rate": 9.357421218136386e-05,
      "loss": 1.4483,
      "step": 41
    },
    {
      "epoch": 0.2140127388535032,
      "grad_norm": 1.6355057954788208,
      "learning_rate": 9.316282404787871e-05,
      "loss": 1.5618,
      "step": 42
    },
    {
      "epoch": 0.21910828025477708,
      "grad_norm": 2.4172537326812744,
      "learning_rate": 9.273963562927695e-05,
      "loss": 1.5843,
      "step": 43
    },
    {
      "epoch": 0.22420382165605096,
      "grad_norm": 1.8143280744552612,
      "learning_rate": 9.230476262104677e-05,
      "loss": 1.5248,
      "step": 44
    },
    {
      "epoch": 0.22929936305732485,
      "grad_norm": 1.6940934658050537,
      "learning_rate": 9.185832391312644e-05,
      "loss": 1.558,
      "step": 45
    },
    {
      "epoch": 0.23439490445859873,
      "grad_norm": 1.8457539081573486,
      "learning_rate": 9.140044155740101e-05,
      "loss": 1.4784,
      "step": 46
    },
    {
      "epoch": 0.2394904458598726,
      "grad_norm": 1.8089879751205444,
      "learning_rate": 9.093124073433463e-05,
      "loss": 1.568,
      "step": 47
    },
    {
      "epoch": 0.2445859872611465,
      "grad_norm": 2.5819547176361084,
      "learning_rate": 9.045084971874738e-05,
      "loss": 1.5019,
      "step": 48
    },
    {
      "epoch": 0.24968152866242038,
      "grad_norm": 4.068037033081055,
      "learning_rate": 8.995939984474624e-05,
      "loss": 1.5797,
      "step": 49
    },
    {
      "epoch": 0.25477707006369427,
      "grad_norm": 2.636991262435913,
      "learning_rate": 8.945702546981969e-05,
      "loss": 1.6978,
      "step": 50
    },
    {
      "epoch": 0.25477707006369427,
      "eval_loss": 1.781501293182373,
      "eval_runtime": 24.7562,
      "eval_samples_per_second": 13.37,
      "eval_steps_per_second": 3.353,
      "step": 50
    },
    {
      "epoch": 0.25987261146496815,
      "grad_norm": 2.454014778137207,
      "learning_rate": 8.894386393810563e-05,
      "loss": 1.6555,
      "step": 51
    },
    {
      "epoch": 0.26496815286624203,
      "grad_norm": 1.820595145225525,
      "learning_rate": 8.842005554284296e-05,
      "loss": 1.4752,
      "step": 52
    },
    {
      "epoch": 0.2700636942675159,
      "grad_norm": 1.2135746479034424,
      "learning_rate": 8.788574348801675e-05,
      "loss": 1.3949,
      "step": 53
    },
    {
      "epoch": 0.2751592356687898,
      "grad_norm": 1.3661576509475708,
      "learning_rate": 8.73410738492077e-05,
      "loss": 1.5373,
      "step": 54
    },
    {
      "epoch": 0.2802547770700637,
      "grad_norm": 1.3132468461990356,
      "learning_rate": 8.678619553365659e-05,
      "loss": 1.395,
      "step": 55
    },
    {
      "epoch": 0.28535031847133757,
      "grad_norm": 1.3208423852920532,
      "learning_rate": 8.622126023955446e-05,
      "loss": 1.4504,
      "step": 56
    },
    {
      "epoch": 0.29044585987261146,
      "grad_norm": 1.2913376092910767,
      "learning_rate": 8.564642241456986e-05,
      "loss": 1.4432,
      "step": 57
    },
    {
      "epoch": 0.29554140127388534,
      "grad_norm": 1.208329439163208,
      "learning_rate": 8.506183921362443e-05,
      "loss": 1.471,
      "step": 58
    },
    {
      "epoch": 0.3006369426751592,
      "grad_norm": 1.0311087369918823,
      "learning_rate": 8.44676704559283e-05,
      "loss": 1.4476,
      "step": 59
    },
    {
      "epoch": 0.3057324840764331,
      "grad_norm": 1.010650396347046,
      "learning_rate": 8.386407858128706e-05,
      "loss": 1.3946,
      "step": 60
    },
    {
      "epoch": 0.310828025477707,
      "grad_norm": 1.0798571109771729,
      "learning_rate": 8.32512286056924e-05,
      "loss": 1.5203,
      "step": 61
    },
    {
      "epoch": 0.3159235668789809,
      "grad_norm": 1.0454195737838745,
      "learning_rate": 8.262928807620843e-05,
      "loss": 1.3841,
      "step": 62
    },
    {
      "epoch": 0.32101910828025476,
      "grad_norm": 1.061657428741455,
      "learning_rate": 8.199842702516583e-05,
      "loss": 1.2917,
      "step": 63
    },
    {
      "epoch": 0.32611464968152865,
      "grad_norm": 1.2131896018981934,
      "learning_rate": 8.135881792367686e-05,
      "loss": 1.4005,
      "step": 64
    },
    {
      "epoch": 0.33121019108280253,
      "grad_norm": 1.232550024986267,
      "learning_rate": 8.07106356344834e-05,
      "loss": 1.4532,
      "step": 65
    },
    {
      "epoch": 0.3363057324840764,
      "grad_norm": 1.2182544469833374,
      "learning_rate": 8.005405736415126e-05,
      "loss": 1.4274,
      "step": 66
    },
    {
      "epoch": 0.3414012738853503,
      "grad_norm": 1.3798710107803345,
      "learning_rate": 7.938926261462366e-05,
      "loss": 1.5591,
      "step": 67
    },
    {
      "epoch": 0.3464968152866242,
      "grad_norm": 1.170486330986023,
      "learning_rate": 7.871643313414718e-05,
      "loss": 1.3739,
      "step": 68
    },
    {
      "epoch": 0.35159235668789807,
      "grad_norm": 1.1986887454986572,
      "learning_rate": 7.803575286758364e-05,
      "loss": 1.4782,
      "step": 69
    },
    {
      "epoch": 0.35668789808917195,
      "grad_norm": 1.1253483295440674,
      "learning_rate": 7.734740790612136e-05,
      "loss": 1.4429,
      "step": 70
    },
    {
      "epoch": 0.36178343949044584,
      "grad_norm": 1.1329927444458008,
      "learning_rate": 7.66515864363997e-05,
      "loss": 1.4225,
      "step": 71
    },
    {
      "epoch": 0.3668789808917197,
      "grad_norm": 1.0653384923934937,
      "learning_rate": 7.594847868906076e-05,
      "loss": 1.3503,
      "step": 72
    },
    {
      "epoch": 0.3719745222929936,
      "grad_norm": 1.1689294576644897,
      "learning_rate": 7.52382768867422e-05,
      "loss": 1.3501,
      "step": 73
    },
    {
      "epoch": 0.37707006369426754,
      "grad_norm": 1.146478295326233,
      "learning_rate": 7.452117519152542e-05,
      "loss": 1.448,
      "step": 74
    },
    {
      "epoch": 0.3821656050955414,
      "grad_norm": 1.3308639526367188,
      "learning_rate": 7.379736965185368e-05,
      "loss": 1.4878,
      "step": 75
    },
    {
      "epoch": 0.3872611464968153,
      "grad_norm": 1.2801364660263062,
      "learning_rate": 7.30670581489344e-05,
      "loss": 1.5172,
      "step": 76
    },
    {
      "epoch": 0.3923566878980892,
      "grad_norm": 1.2632524967193604,
      "learning_rate": 7.233044034264034e-05,
      "loss": 1.4038,
      "step": 77
    },
    {
      "epoch": 0.3974522292993631,
      "grad_norm": 1.2563971281051636,
      "learning_rate": 7.158771761692464e-05,
      "loss": 1.5025,
      "step": 78
    },
    {
      "epoch": 0.40254777070063696,
      "grad_norm": 1.2936761379241943,
      "learning_rate": 7.083909302476453e-05,
      "loss": 1.4684,
      "step": 79
    },
    {
      "epoch": 0.40764331210191085,
      "grad_norm": 1.2676165103912354,
      "learning_rate": 7.008477123264848e-05,
      "loss": 1.4856,
      "step": 80
    },
    {
      "epoch": 0.41273885350318473,
      "grad_norm": 1.314073920249939,
      "learning_rate": 6.932495846462261e-05,
      "loss": 1.449,
      "step": 81
    },
    {
      "epoch": 0.4178343949044586,
      "grad_norm": 1.2387086153030396,
      "learning_rate": 6.855986244591104e-05,
      "loss": 1.4605,
      "step": 82
    },
    {
      "epoch": 0.4229299363057325,
      "grad_norm": 1.2968636751174927,
      "learning_rate": 6.778969234612584e-05,
      "loss": 1.4965,
      "step": 83
    },
    {
      "epoch": 0.4280254777070064,
      "grad_norm": 1.2759613990783691,
      "learning_rate": 6.701465872208216e-05,
      "loss": 1.4109,
      "step": 84
    },
    {
      "epoch": 0.43312101910828027,
      "grad_norm": 1.3058372735977173,
      "learning_rate": 6.623497346023418e-05,
      "loss": 1.4296,
      "step": 85
    },
    {
      "epoch": 0.43821656050955415,
      "grad_norm": 1.2999426126480103,
      "learning_rate": 6.545084971874738e-05,
      "loss": 1.3434,
      "step": 86
    },
    {
      "epoch": 0.44331210191082804,
      "grad_norm": 1.3399734497070312,
      "learning_rate": 6.466250186922325e-05,
      "loss": 1.4846,
      "step": 87
    },
    {
      "epoch": 0.4484076433121019,
      "grad_norm": 1.4246940612792969,
      "learning_rate": 6.387014543809223e-05,
      "loss": 1.4417,
      "step": 88
    },
    {
      "epoch": 0.4535031847133758,
      "grad_norm": 1.4686065912246704,
      "learning_rate": 6.307399704769099e-05,
      "loss": 1.4917,
      "step": 89
    },
    {
      "epoch": 0.4585987261146497,
      "grad_norm": 1.553601861000061,
      "learning_rate": 6.227427435703997e-05,
      "loss": 1.4788,
      "step": 90
    },
    {
      "epoch": 0.4636942675159236,
      "grad_norm": 1.5551059246063232,
      "learning_rate": 6.147119600233758e-05,
      "loss": 1.6216,
      "step": 91
    },
    {
      "epoch": 0.46878980891719746,
      "grad_norm": 1.3926206827163696,
      "learning_rate": 6.066498153718735e-05,
      "loss": 1.4596,
      "step": 92
    },
    {
      "epoch": 0.47388535031847134,
      "grad_norm": 1.578769564628601,
      "learning_rate": 5.985585137257401e-05,
      "loss": 1.5169,
      "step": 93
    },
    {
      "epoch": 0.4789808917197452,
      "grad_norm": 1.4515339136123657,
      "learning_rate": 5.90440267166055e-05,
      "loss": 1.455,
      "step": 94
    },
    {
      "epoch": 0.4840764331210191,
      "grad_norm": 1.4279038906097412,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 1.5198,
      "step": 95
    },
    {
      "epoch": 0.489171974522293,
      "grad_norm": 1.6243515014648438,
      "learning_rate": 5.74131823855921e-05,
      "loss": 1.4692,
      "step": 96
    },
    {
      "epoch": 0.4942675159235669,
      "grad_norm": 1.7584609985351562,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 1.5059,
      "step": 97
    },
    {
      "epoch": 0.49936305732484076,
      "grad_norm": 1.8533011674880981,
      "learning_rate": 5.577423184847932e-05,
      "loss": 1.4124,
      "step": 98
    },
    {
      "epoch": 0.5044585987261146,
      "grad_norm": 2.320594549179077,
      "learning_rate": 5.495227651252315e-05,
      "loss": 1.6711,
      "step": 99
    },
    {
      "epoch": 0.5095541401273885,
      "grad_norm": 2.2607131004333496,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 1.6153,
      "step": 100
    },
    {
      "epoch": 0.5095541401273885,
      "eval_loss": 1.601906180381775,
      "eval_runtime": 24.7253,
      "eval_samples_per_second": 13.387,
      "eval_steps_per_second": 3.357,
      "step": 100
    },
    {
      "epoch": 0.5146496815286624,
      "grad_norm": 1.9724311828613281,
      "learning_rate": 5.330452921628497e-05,
      "loss": 1.5492,
      "step": 101
    },
    {
      "epoch": 0.5197452229299363,
      "grad_norm": 1.667399287223816,
      "learning_rate": 5.247918773366112e-05,
      "loss": 1.5409,
      "step": 102
    },
    {
      "epoch": 0.5248407643312102,
      "grad_norm": 1.2964222431182861,
      "learning_rate": 5.165316846586541e-05,
      "loss": 1.5121,
      "step": 103
    },
    {
      "epoch": 0.5299363057324841,
      "grad_norm": 1.0274351835250854,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 1.3781,
      "step": 104
    },
    {
      "epoch": 0.535031847133758,
      "grad_norm": 0.9273486733436584,
      "learning_rate": 5e-05,
      "loss": 1.3144,
      "step": 105
    },
    {
      "epoch": 0.5401273885350318,
      "grad_norm": 0.929964542388916,
      "learning_rate": 4.917330276168208e-05,
      "loss": 1.2833,
      "step": 106
    },
    {
      "epoch": 0.5452229299363057,
      "grad_norm": 0.9989023804664612,
      "learning_rate": 4.834683153413459e-05,
      "loss": 1.3783,
      "step": 107
    },
    {
      "epoch": 0.5503184713375796,
      "grad_norm": 1.0930886268615723,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 1.3683,
      "step": 108
    },
    {
      "epoch": 0.5554140127388535,
      "grad_norm": 1.1951261758804321,
      "learning_rate": 4.669547078371504e-05,
      "loss": 1.4119,
      "step": 109
    },
    {
      "epoch": 0.5605095541401274,
      "grad_norm": 1.1240007877349854,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 1.4123,
      "step": 110
    },
    {
      "epoch": 0.5656050955414013,
      "grad_norm": 1.1993601322174072,
      "learning_rate": 4.504772348747687e-05,
      "loss": 1.4275,
      "step": 111
    },
    {
      "epoch": 0.5707006369426751,
      "grad_norm": 1.1673270463943481,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 1.418,
      "step": 112
    },
    {
      "epoch": 0.575796178343949,
      "grad_norm": 1.0964654684066772,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 1.3707,
      "step": 113
    },
    {
      "epoch": 0.5808917197452229,
      "grad_norm": 1.1067087650299072,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 1.407,
      "step": 114
    },
    {
      "epoch": 0.5859872611464968,
      "grad_norm": 1.1007089614868164,
      "learning_rate": 4.17702704859633e-05,
      "loss": 1.4612,
      "step": 115
    },
    {
      "epoch": 0.5910828025477707,
      "grad_norm": 1.1179075241088867,
      "learning_rate": 4.095597328339452e-05,
      "loss": 1.4064,
      "step": 116
    },
    {
      "epoch": 0.5961783439490446,
      "grad_norm": 1.0929796695709229,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 1.4272,
      "step": 117
    },
    {
      "epoch": 0.6012738853503184,
      "grad_norm": 1.0987190008163452,
      "learning_rate": 3.933501846281267e-05,
      "loss": 1.4979,
      "step": 118
    },
    {
      "epoch": 0.6063694267515923,
      "grad_norm": 1.0411033630371094,
      "learning_rate": 3.852880399766243e-05,
      "loss": 1.3349,
      "step": 119
    },
    {
      "epoch": 0.6114649681528662,
      "grad_norm": 1.1143288612365723,
      "learning_rate": 3.772572564296005e-05,
      "loss": 1.4471,
      "step": 120
    },
    {
      "epoch": 0.6165605095541401,
      "grad_norm": 1.1308385133743286,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 1.4265,
      "step": 121
    },
    {
      "epoch": 0.621656050955414,
      "grad_norm": 1.1293975114822388,
      "learning_rate": 3.612985456190778e-05,
      "loss": 1.5129,
      "step": 122
    },
    {
      "epoch": 0.6267515923566879,
      "grad_norm": 1.2257919311523438,
      "learning_rate": 3.533749813077677e-05,
      "loss": 1.4914,
      "step": 123
    },
    {
      "epoch": 0.6318471337579618,
      "grad_norm": 1.155097246170044,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 1.4501,
      "step": 124
    },
    {
      "epoch": 0.6369426751592356,
      "grad_norm": 1.1370073556900024,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 1.3939,
      "step": 125
    },
    {
      "epoch": 0.6420382165605095,
      "grad_norm": 1.1457079648971558,
      "learning_rate": 3.298534127791785e-05,
      "loss": 1.4761,
      "step": 126
    },
    {
      "epoch": 0.6471337579617834,
      "grad_norm": 1.1135650873184204,
      "learning_rate": 3.221030765387417e-05,
      "loss": 1.4424,
      "step": 127
    },
    {
      "epoch": 0.6522292993630573,
      "grad_norm": 1.2130106687545776,
      "learning_rate": 3.144013755408895e-05,
      "loss": 1.4474,
      "step": 128
    },
    {
      "epoch": 0.6573248407643312,
      "grad_norm": 1.2500935792922974,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 1.4642,
      "step": 129
    },
    {
      "epoch": 0.6624203821656051,
      "grad_norm": 1.2229430675506592,
      "learning_rate": 2.991522876735154e-05,
      "loss": 1.4752,
      "step": 130
    },
    {
      "epoch": 0.6675159235668789,
      "grad_norm": 1.2784786224365234,
      "learning_rate": 2.916090697523549e-05,
      "loss": 1.3137,
      "step": 131
    },
    {
      "epoch": 0.6726114649681528,
      "grad_norm": 1.2230734825134277,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 1.4229,
      "step": 132
    },
    {
      "epoch": 0.6777070063694267,
      "grad_norm": 1.2815632820129395,
      "learning_rate": 2.766955965735968e-05,
      "loss": 1.4003,
      "step": 133
    },
    {
      "epoch": 0.6828025477707006,
      "grad_norm": 1.3112797737121582,
      "learning_rate": 2.693294185106562e-05,
      "loss": 1.4663,
      "step": 134
    },
    {
      "epoch": 0.6878980891719745,
      "grad_norm": 1.333423376083374,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 1.4113,
      "step": 135
    },
    {
      "epoch": 0.6929936305732484,
      "grad_norm": 1.305199384689331,
      "learning_rate": 2.547882480847461e-05,
      "loss": 1.4611,
      "step": 136
    },
    {
      "epoch": 0.6980891719745222,
      "grad_norm": 1.3084430694580078,
      "learning_rate": 2.476172311325783e-05,
      "loss": 1.4663,
      "step": 137
    },
    {
      "epoch": 0.7031847133757961,
      "grad_norm": 1.4007364511489868,
      "learning_rate": 2.405152131093926e-05,
      "loss": 1.5499,
      "step": 138
    },
    {
      "epoch": 0.70828025477707,
      "grad_norm": 1.4143140316009521,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 1.5543,
      "step": 139
    },
    {
      "epoch": 0.7133757961783439,
      "grad_norm": 1.407924771308899,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 1.4847,
      "step": 140
    },
    {
      "epoch": 0.7184713375796178,
      "grad_norm": 1.3992478847503662,
      "learning_rate": 2.196424713241637e-05,
      "loss": 1.4271,
      "step": 141
    },
    {
      "epoch": 0.7235668789808917,
      "grad_norm": 1.3795225620269775,
      "learning_rate": 2.128356686585282e-05,
      "loss": 1.509,
      "step": 142
    },
    {
      "epoch": 0.7286624203821656,
      "grad_norm": 1.5027143955230713,
      "learning_rate": 2.061073738537635e-05,
      "loss": 1.4519,
      "step": 143
    },
    {
      "epoch": 0.7337579617834394,
      "grad_norm": 1.4326565265655518,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 1.5076,
      "step": 144
    },
    {
      "epoch": 0.7388535031847133,
      "grad_norm": 1.4619704484939575,
      "learning_rate": 1.928936436551661e-05,
      "loss": 1.5012,
      "step": 145
    },
    {
      "epoch": 0.7439490445859872,
      "grad_norm": 1.4852865934371948,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 1.4779,
      "step": 146
    },
    {
      "epoch": 0.7490445859872611,
      "grad_norm": 1.7760199308395386,
      "learning_rate": 1.800157297483417e-05,
      "loss": 1.4853,
      "step": 147
    },
    {
      "epoch": 0.7541401273885351,
      "grad_norm": 1.7638636827468872,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 1.463,
      "step": 148
    },
    {
      "epoch": 0.759235668789809,
      "grad_norm": 1.7473770380020142,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 1.5419,
      "step": 149
    },
    {
      "epoch": 0.7643312101910829,
      "grad_norm": 1.614508032798767,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 1.5332,
      "step": 150
    },
    {
      "epoch": 0.7643312101910829,
      "eval_loss": 1.471155047416687,
      "eval_runtime": 24.7373,
      "eval_samples_per_second": 13.381,
      "eval_steps_per_second": 3.355,
      "step": 150
    },
    {
      "epoch": 0.7694267515923567,
      "grad_norm": 1.460286259651184,
      "learning_rate": 1.553232954407171e-05,
      "loss": 1.5378,
      "step": 151
    },
    {
      "epoch": 0.7745222929936306,
      "grad_norm": 1.4250564575195312,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 1.5691,
      "step": 152
    },
    {
      "epoch": 0.7796178343949045,
      "grad_norm": 1.3611723184585571,
      "learning_rate": 1.435357758543015e-05,
      "loss": 1.5065,
      "step": 153
    },
    {
      "epoch": 0.7847133757961784,
      "grad_norm": 1.238294005393982,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 1.4029,
      "step": 154
    },
    {
      "epoch": 0.7898089171974523,
      "grad_norm": 1.1684417724609375,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 1.5045,
      "step": 155
    },
    {
      "epoch": 0.7949044585987262,
      "grad_norm": 1.1050057411193848,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 1.5068,
      "step": 156
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.0719249248504639,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 1.5307,
      "step": 157
    },
    {
      "epoch": 0.8050955414012739,
      "grad_norm": 1.0645467042922974,
      "learning_rate": 1.157994445715706e-05,
      "loss": 1.4514,
      "step": 158
    },
    {
      "epoch": 0.8101910828025478,
      "grad_norm": 1.016603708267212,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 1.416,
      "step": 159
    },
    {
      "epoch": 0.8152866242038217,
      "grad_norm": 1.0058902502059937,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 1.4427,
      "step": 160
    },
    {
      "epoch": 0.8203821656050956,
      "grad_norm": 0.9901459813117981,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 1.4398,
      "step": 161
    },
    {
      "epoch": 0.8254777070063695,
      "grad_norm": 0.9657883644104004,
      "learning_rate": 9.549150281252633e-06,
      "loss": 1.4077,
      "step": 162
    },
    {
      "epoch": 0.8305732484076433,
      "grad_norm": 0.9688636064529419,
      "learning_rate": 9.068759265665384e-06,
      "loss": 1.3933,
      "step": 163
    },
    {
      "epoch": 0.8356687898089172,
      "grad_norm": 1.029929757118225,
      "learning_rate": 8.599558442598998e-06,
      "loss": 1.4636,
      "step": 164
    },
    {
      "epoch": 0.8407643312101911,
      "grad_norm": 1.0182278156280518,
      "learning_rate": 8.141676086873572e-06,
      "loss": 1.3908,
      "step": 165
    },
    {
      "epoch": 0.845859872611465,
      "grad_norm": 1.0764676332473755,
      "learning_rate": 7.695237378953223e-06,
      "loss": 1.4657,
      "step": 166
    },
    {
      "epoch": 0.8509554140127389,
      "grad_norm": 0.991082489490509,
      "learning_rate": 7.260364370723044e-06,
      "loss": 1.3587,
      "step": 167
    },
    {
      "epoch": 0.8560509554140128,
      "grad_norm": 1.0163828134536743,
      "learning_rate": 6.837175952121306e-06,
      "loss": 1.4112,
      "step": 168
    },
    {
      "epoch": 0.8611464968152867,
      "grad_norm": 1.023792028427124,
      "learning_rate": 6.425787818636131e-06,
      "loss": 1.3898,
      "step": 169
    },
    {
      "epoch": 0.8662420382165605,
      "grad_norm": 1.0270365476608276,
      "learning_rate": 6.026312439675552e-06,
      "loss": 1.3414,
      "step": 170
    },
    {
      "epoch": 0.8713375796178344,
      "grad_norm": 1.0704646110534668,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 1.4055,
      "step": 171
    },
    {
      "epoch": 0.8764331210191083,
      "grad_norm": 1.1000847816467285,
      "learning_rate": 5.263533508961827e-06,
      "loss": 1.4956,
      "step": 172
    },
    {
      "epoch": 0.8815286624203822,
      "grad_norm": 1.0627716779708862,
      "learning_rate": 4.900438493352055e-06,
      "loss": 1.5556,
      "step": 173
    },
    {
      "epoch": 0.8866242038216561,
      "grad_norm": 1.1349250078201294,
      "learning_rate": 4.549673247541875e-06,
      "loss": 1.4317,
      "step": 174
    },
    {
      "epoch": 0.89171974522293,
      "grad_norm": 1.0982369184494019,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 1.3892,
      "step": 175
    },
    {
      "epoch": 0.8968152866242038,
      "grad_norm": 1.117073893547058,
      "learning_rate": 3.885512251130763e-06,
      "loss": 1.438,
      "step": 176
    },
    {
      "epoch": 0.9019108280254777,
      "grad_norm": 1.0480406284332275,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 1.385,
      "step": 177
    },
    {
      "epoch": 0.9070063694267516,
      "grad_norm": 1.1920479536056519,
      "learning_rate": 3.271776770026963e-06,
      "loss": 1.5013,
      "step": 178
    },
    {
      "epoch": 0.9121019108280255,
      "grad_norm": 1.1584534645080566,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 1.4814,
      "step": 179
    },
    {
      "epoch": 0.9171974522292994,
      "grad_norm": 1.1337345838546753,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 1.3615,
      "step": 180
    },
    {
      "epoch": 0.9222929936305733,
      "grad_norm": 1.2368887662887573,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 1.5315,
      "step": 181
    },
    {
      "epoch": 0.9273885350318471,
      "grad_norm": 1.1957945823669434,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 1.4633,
      "step": 182
    },
    {
      "epoch": 0.932484076433121,
      "grad_norm": 1.201175570487976,
      "learning_rate": 1.962316193157593e-06,
      "loss": 1.4864,
      "step": 183
    },
    {
      "epoch": 0.9375796178343949,
      "grad_norm": 1.2664943933486938,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 1.479,
      "step": 184
    },
    {
      "epoch": 0.9426751592356688,
      "grad_norm": 1.2017543315887451,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 1.4352,
      "step": 185
    },
    {
      "epoch": 0.9477707006369427,
      "grad_norm": 1.2606137990951538,
      "learning_rate": 1.333670137599713e-06,
      "loss": 1.4695,
      "step": 186
    },
    {
      "epoch": 0.9528662420382166,
      "grad_norm": 1.3612840175628662,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 1.5487,
      "step": 187
    },
    {
      "epoch": 0.9579617834394905,
      "grad_norm": 1.4412602186203003,
      "learning_rate": 9.810017062595322e-07,
      "loss": 1.5304,
      "step": 188
    },
    {
      "epoch": 0.9630573248407643,
      "grad_norm": 1.42183256149292,
      "learning_rate": 8.247462563808817e-07,
      "loss": 1.4792,
      "step": 189
    },
    {
      "epoch": 0.9681528662420382,
      "grad_norm": 1.6373732089996338,
      "learning_rate": 6.819348298638839e-07,
      "loss": 1.6268,
      "step": 190
    },
    {
      "epoch": 0.9732484076433121,
      "grad_norm": 1.5159034729003906,
      "learning_rate": 5.526064699265753e-07,
      "loss": 1.527,
      "step": 191
    },
    {
      "epoch": 0.978343949044586,
      "grad_norm": 1.533638834953308,
      "learning_rate": 4.367965336512403e-07,
      "loss": 1.5435,
      "step": 192
    },
    {
      "epoch": 0.9834394904458599,
      "grad_norm": 1.7228260040283203,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 1.4993,
      "step": 193
    },
    {
      "epoch": 0.9885350318471338,
      "grad_norm": 1.753058910369873,
      "learning_rate": 2.458548727494292e-07,
      "loss": 1.6041,
      "step": 194
    },
    {
      "epoch": 0.9936305732484076,
      "grad_norm": 1.9408619403839111,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 1.5683,
      "step": 195
    },
    {
      "epoch": 0.9987261146496815,
      "grad_norm": 2.8912229537963867,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 1.7543,
      "step": 196
    },
    {
      "epoch": 1.0038216560509554,
      "grad_norm": 3.4295921325683594,
      "learning_rate": 6.150154258476315e-08,
      "loss": 2.534,
      "step": 197
    },
    {
      "epoch": 1.0089171974522293,
      "grad_norm": 0.7272393703460693,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 1.0972,
      "step": 198
    },
    {
      "epoch": 1.0140127388535032,
      "grad_norm": 0.7166330218315125,
      "learning_rate": 6.834750376549792e-09,
      "loss": 1.183,
      "step": 199
    },
    {
      "epoch": 1.019108280254777,
      "grad_norm": 0.7291227579116821,
      "learning_rate": 0.0,
      "loss": 1.1375,
      "step": 200
    },
    {
      "epoch": 1.019108280254777,
      "eval_loss": 1.3939571380615234,
      "eval_runtime": 24.8217,
      "eval_samples_per_second": 13.335,
      "eval_steps_per_second": 3.344,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "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": 3.01514151551828e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}