{
  "best_metric": 2.7577829360961914,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 1.1611030478955007,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005805515239477504,
      "grad_norm": 2.573275327682495,
      "learning_rate": 1e-05,
      "loss": 9.0962,
      "step": 1
    },
    {
      "epoch": 0.005805515239477504,
      "eval_loss": 5.559586048126221,
      "eval_runtime": 21.6268,
      "eval_samples_per_second": 13.409,
      "eval_steps_per_second": 3.375,
      "step": 1
    },
    {
      "epoch": 0.011611030478955007,
      "grad_norm": 2.911691665649414,
      "learning_rate": 2e-05,
      "loss": 9.379,
      "step": 2
    },
    {
      "epoch": 0.01741654571843251,
      "grad_norm": 2.7299020290374756,
      "learning_rate": 3e-05,
      "loss": 8.8716,
      "step": 3
    },
    {
      "epoch": 0.023222060957910014,
      "grad_norm": 3.025408983230591,
      "learning_rate": 4e-05,
      "loss": 9.649,
      "step": 4
    },
    {
      "epoch": 0.02902757619738752,
      "grad_norm": 2.4141650199890137,
      "learning_rate": 5e-05,
      "loss": 9.0521,
      "step": 5
    },
    {
      "epoch": 0.03483309143686502,
      "grad_norm": 2.2509329319000244,
      "learning_rate": 6e-05,
      "loss": 8.7478,
      "step": 6
    },
    {
      "epoch": 0.040638606676342524,
      "grad_norm": 1.8296687602996826,
      "learning_rate": 7e-05,
      "loss": 9.7416,
      "step": 7
    },
    {
      "epoch": 0.04644412191582003,
      "grad_norm": 2.0277085304260254,
      "learning_rate": 8e-05,
      "loss": 9.512,
      "step": 8
    },
    {
      "epoch": 0.05224963715529753,
      "grad_norm": 2.173041820526123,
      "learning_rate": 9e-05,
      "loss": 8.6656,
      "step": 9
    },
    {
      "epoch": 0.05805515239477504,
      "grad_norm": 2.508038282394409,
      "learning_rate": 0.0001,
      "loss": 8.4684,
      "step": 10
    },
    {
      "epoch": 0.06386066763425254,
      "grad_norm": 1.5355852842330933,
      "learning_rate": 9.999316524962345e-05,
      "loss": 8.273,
      "step": 11
    },
    {
      "epoch": 0.06966618287373004,
      "grad_norm": 1.3360497951507568,
      "learning_rate": 9.997266286704631e-05,
      "loss": 8.8818,
      "step": 12
    },
    {
      "epoch": 0.07547169811320754,
      "grad_norm": 1.1594655513763428,
      "learning_rate": 9.993849845741524e-05,
      "loss": 7.2362,
      "step": 13
    },
    {
      "epoch": 0.08127721335268505,
      "grad_norm": 1.2919799089431763,
      "learning_rate": 9.989068136093873e-05,
      "loss": 7.842,
      "step": 14
    },
    {
      "epoch": 0.08708272859216255,
      "grad_norm": 4.759025573730469,
      "learning_rate": 9.98292246503335e-05,
      "loss": 7.4901,
      "step": 15
    },
    {
      "epoch": 0.09288824383164006,
      "grad_norm": 5.793337821960449,
      "learning_rate": 9.975414512725057e-05,
      "loss": 7.2166,
      "step": 16
    },
    {
      "epoch": 0.09869375907111756,
      "grad_norm": 3.9417288303375244,
      "learning_rate": 9.966546331768191e-05,
      "loss": 6.9083,
      "step": 17
    },
    {
      "epoch": 0.10449927431059507,
      "grad_norm": 1.789911150932312,
      "learning_rate": 9.956320346634876e-05,
      "loss": 4.3346,
      "step": 18
    },
    {
      "epoch": 0.11030478955007257,
      "grad_norm": 1.993194341659546,
      "learning_rate": 9.944739353007344e-05,
      "loss": 5.7995,
      "step": 19
    },
    {
      "epoch": 0.11611030478955008,
      "grad_norm": 1.6666117906570435,
      "learning_rate": 9.931806517013612e-05,
      "loss": 4.8374,
      "step": 20
    },
    {
      "epoch": 0.12191582002902758,
      "grad_norm": 1.682180643081665,
      "learning_rate": 9.917525374361912e-05,
      "loss": 4.4451,
      "step": 21
    },
    {
      "epoch": 0.12772133526850507,
      "grad_norm": 1.6778733730316162,
      "learning_rate": 9.901899829374047e-05,
      "loss": 3.5964,
      "step": 22
    },
    {
      "epoch": 0.13352685050798258,
      "grad_norm": 1.814108967781067,
      "learning_rate": 9.884934153917997e-05,
      "loss": 3.3797,
      "step": 23
    },
    {
      "epoch": 0.13933236574746008,
      "grad_norm": 1.3642491102218628,
      "learning_rate": 9.86663298624003e-05,
      "loss": 3.1198,
      "step": 24
    },
    {
      "epoch": 0.14513788098693758,
      "grad_norm": 1.3092386722564697,
      "learning_rate": 9.847001329696653e-05,
      "loss": 3.2171,
      "step": 25
    },
    {
      "epoch": 0.1509433962264151,
      "grad_norm": 1.8021727800369263,
      "learning_rate": 9.826044551386744e-05,
      "loss": 3.3156,
      "step": 26
    },
    {
      "epoch": 0.1567489114658926,
      "grad_norm": 1.4289323091506958,
      "learning_rate": 9.803768380684242e-05,
      "loss": 3.4765,
      "step": 27
    },
    {
      "epoch": 0.1625544267053701,
      "grad_norm": 1.4999114274978638,
      "learning_rate": 9.780178907671789e-05,
      "loss": 3.864,
      "step": 28
    },
    {
      "epoch": 0.1683599419448476,
      "grad_norm": 1.3517564535140991,
      "learning_rate": 9.755282581475769e-05,
      "loss": 3.4888,
      "step": 29
    },
    {
      "epoch": 0.1741654571843251,
      "grad_norm": 1.6473299264907837,
      "learning_rate": 9.729086208503174e-05,
      "loss": 3.1538,
      "step": 30
    },
    {
      "epoch": 0.1799709724238026,
      "grad_norm": 1.8896262645721436,
      "learning_rate": 9.701596950580806e-05,
      "loss": 2.9812,
      "step": 31
    },
    {
      "epoch": 0.18577648766328012,
      "grad_norm": 1.6572469472885132,
      "learning_rate": 9.672822322997305e-05,
      "loss": 3.0378,
      "step": 32
    },
    {
      "epoch": 0.19158200290275762,
      "grad_norm": 2.023664951324463,
      "learning_rate": 9.642770192448536e-05,
      "loss": 2.8597,
      "step": 33
    },
    {
      "epoch": 0.19738751814223512,
      "grad_norm": 1.5709890127182007,
      "learning_rate": 9.611448774886924e-05,
      "loss": 2.4736,
      "step": 34
    },
    {
      "epoch": 0.20319303338171263,
      "grad_norm": 1.7750293016433716,
      "learning_rate": 9.578866633275288e-05,
      "loss": 2.5613,
      "step": 35
    },
    {
      "epoch": 0.20899854862119013,
      "grad_norm": 2.6448962688446045,
      "learning_rate": 9.545032675245813e-05,
      "loss": 2.0605,
      "step": 36
    },
    {
      "epoch": 0.21480406386066764,
      "grad_norm": 2.798313856124878,
      "learning_rate": 9.509956150664796e-05,
      "loss": 1.2738,
      "step": 37
    },
    {
      "epoch": 0.22060957910014514,
      "grad_norm": 2.3851444721221924,
      "learning_rate": 9.473646649103818e-05,
      "loss": 1.4869,
      "step": 38
    },
    {
      "epoch": 0.22641509433962265,
      "grad_norm": 3.0625545978546143,
      "learning_rate": 9.43611409721806e-05,
      "loss": 1.1592,
      "step": 39
    },
    {
      "epoch": 0.23222060957910015,
      "grad_norm": 2.394531488418579,
      "learning_rate": 9.397368756032445e-05,
      "loss": 1.1647,
      "step": 40
    },
    {
      "epoch": 0.23802612481857766,
      "grad_norm": 2.437258243560791,
      "learning_rate": 9.357421218136386e-05,
      "loss": 1.2982,
      "step": 41
    },
    {
      "epoch": 0.24383164005805516,
      "grad_norm": 2.3412349224090576,
      "learning_rate": 9.316282404787871e-05,
      "loss": 1.36,
      "step": 42
    },
    {
      "epoch": 0.24963715529753266,
      "grad_norm": 4.212519645690918,
      "learning_rate": 9.273963562927695e-05,
      "loss": 2.0828,
      "step": 43
    },
    {
      "epoch": 0.25544267053701014,
      "grad_norm": 4.577381134033203,
      "learning_rate": 9.230476262104677e-05,
      "loss": 8.3537,
      "step": 44
    },
    {
      "epoch": 0.2612481857764877,
      "grad_norm": 4.020319938659668,
      "learning_rate": 9.185832391312644e-05,
      "loss": 9.0308,
      "step": 45
    },
    {
      "epoch": 0.26705370101596515,
      "grad_norm": 3.1003880500793457,
      "learning_rate": 9.140044155740101e-05,
      "loss": 9.3452,
      "step": 46
    },
    {
      "epoch": 0.2728592162554427,
      "grad_norm": 2.0905094146728516,
      "learning_rate": 9.093124073433463e-05,
      "loss": 8.8027,
      "step": 47
    },
    {
      "epoch": 0.27866473149492016,
      "grad_norm": 1.4608365297317505,
      "learning_rate": 9.045084971874738e-05,
      "loss": 8.1428,
      "step": 48
    },
    {
      "epoch": 0.2844702467343977,
      "grad_norm": 1.6426067352294922,
      "learning_rate": 8.995939984474624e-05,
      "loss": 8.0256,
      "step": 49
    },
    {
      "epoch": 0.29027576197387517,
      "grad_norm": 1.3261559009552002,
      "learning_rate": 8.945702546981969e-05,
      "loss": 8.4734,
      "step": 50
    },
    {
      "epoch": 0.29027576197387517,
      "eval_loss": 4.877832889556885,
      "eval_runtime": 21.9769,
      "eval_samples_per_second": 13.196,
      "eval_steps_per_second": 3.322,
      "step": 50
    },
    {
      "epoch": 0.2960812772133527,
      "grad_norm": 1.336865782737732,
      "learning_rate": 8.894386393810563e-05,
      "loss": 8.785,
      "step": 51
    },
    {
      "epoch": 0.3018867924528302,
      "grad_norm": 1.4489983320236206,
      "learning_rate": 8.842005554284296e-05,
      "loss": 8.1106,
      "step": 52
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 1.4393694400787354,
      "learning_rate": 8.788574348801675e-05,
      "loss": 7.9837,
      "step": 53
    },
    {
      "epoch": 0.3134978229317852,
      "grad_norm": 1.2898362874984741,
      "learning_rate": 8.73410738492077e-05,
      "loss": 7.9824,
      "step": 54
    },
    {
      "epoch": 0.3193033381712627,
      "grad_norm": 1.4332189559936523,
      "learning_rate": 8.678619553365659e-05,
      "loss": 7.7971,
      "step": 55
    },
    {
      "epoch": 0.3251088534107402,
      "grad_norm": 1.6546247005462646,
      "learning_rate": 8.622126023955446e-05,
      "loss": 6.7689,
      "step": 56
    },
    {
      "epoch": 0.3309143686502177,
      "grad_norm": 1.7686975002288818,
      "learning_rate": 8.564642241456986e-05,
      "loss": 6.8831,
      "step": 57
    },
    {
      "epoch": 0.3367198838896952,
      "grad_norm": 2.798206090927124,
      "learning_rate": 8.506183921362443e-05,
      "loss": 6.5294,
      "step": 58
    },
    {
      "epoch": 0.34252539912917274,
      "grad_norm": 3.7765955924987793,
      "learning_rate": 8.44676704559283e-05,
      "loss": 6.794,
      "step": 59
    },
    {
      "epoch": 0.3483309143686502,
      "grad_norm": 3.310652017593384,
      "learning_rate": 8.386407858128706e-05,
      "loss": 7.3537,
      "step": 60
    },
    {
      "epoch": 0.35413642960812775,
      "grad_norm": 3.096815824508667,
      "learning_rate": 8.32512286056924e-05,
      "loss": 5.2614,
      "step": 61
    },
    {
      "epoch": 0.3599419448476052,
      "grad_norm": 2.3964264392852783,
      "learning_rate": 8.262928807620843e-05,
      "loss": 3.8162,
      "step": 62
    },
    {
      "epoch": 0.36574746008708275,
      "grad_norm": 2.4574034214019775,
      "learning_rate": 8.199842702516583e-05,
      "loss": 4.4919,
      "step": 63
    },
    {
      "epoch": 0.37155297532656023,
      "grad_norm": 2.3561408519744873,
      "learning_rate": 8.135881792367686e-05,
      "loss": 4.2271,
      "step": 64
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 1.9778844118118286,
      "learning_rate": 8.07106356344834e-05,
      "loss": 3.9258,
      "step": 65
    },
    {
      "epoch": 0.38316400580551524,
      "grad_norm": 1.6830021142959595,
      "learning_rate": 8.005405736415126e-05,
      "loss": 2.9788,
      "step": 66
    },
    {
      "epoch": 0.3889695210449927,
      "grad_norm": 2.028007745742798,
      "learning_rate": 7.938926261462366e-05,
      "loss": 2.7798,
      "step": 67
    },
    {
      "epoch": 0.39477503628447025,
      "grad_norm": 1.6069254875183105,
      "learning_rate": 7.871643313414718e-05,
      "loss": 2.5249,
      "step": 68
    },
    {
      "epoch": 0.4005805515239477,
      "grad_norm": 1.7473154067993164,
      "learning_rate": 7.803575286758364e-05,
      "loss": 2.1422,
      "step": 69
    },
    {
      "epoch": 0.40638606676342526,
      "grad_norm": 1.9494911432266235,
      "learning_rate": 7.734740790612136e-05,
      "loss": 2.4944,
      "step": 70
    },
    {
      "epoch": 0.41219158200290273,
      "grad_norm": 2.2357215881347656,
      "learning_rate": 7.66515864363997e-05,
      "loss": 2.3646,
      "step": 71
    },
    {
      "epoch": 0.41799709724238027,
      "grad_norm": 2.8632986545562744,
      "learning_rate": 7.594847868906076e-05,
      "loss": 3.2673,
      "step": 72
    },
    {
      "epoch": 0.42380261248185774,
      "grad_norm": 2.5669875144958496,
      "learning_rate": 7.52382768867422e-05,
      "loss": 2.6435,
      "step": 73
    },
    {
      "epoch": 0.4296081277213353,
      "grad_norm": 2.3477303981781006,
      "learning_rate": 7.452117519152542e-05,
      "loss": 2.2752,
      "step": 74
    },
    {
      "epoch": 0.43541364296081275,
      "grad_norm": 2.740025281906128,
      "learning_rate": 7.379736965185368e-05,
      "loss": 2.3249,
      "step": 75
    },
    {
      "epoch": 0.4412191582002903,
      "grad_norm": 2.6720454692840576,
      "learning_rate": 7.30670581489344e-05,
      "loss": 2.1179,
      "step": 76
    },
    {
      "epoch": 0.44702467343976776,
      "grad_norm": 1.9160422086715698,
      "learning_rate": 7.233044034264034e-05,
      "loss": 2.1594,
      "step": 77
    },
    {
      "epoch": 0.4528301886792453,
      "grad_norm": 2.08158802986145,
      "learning_rate": 7.158771761692464e-05,
      "loss": 2.0943,
      "step": 78
    },
    {
      "epoch": 0.45863570391872277,
      "grad_norm": 1.9168369770050049,
      "learning_rate": 7.083909302476453e-05,
      "loss": 2.0263,
      "step": 79
    },
    {
      "epoch": 0.4644412191582003,
      "grad_norm": 2.4574244022369385,
      "learning_rate": 7.008477123264848e-05,
      "loss": 1.6031,
      "step": 80
    },
    {
      "epoch": 0.4702467343976778,
      "grad_norm": 3.0724689960479736,
      "learning_rate": 6.932495846462261e-05,
      "loss": 0.7674,
      "step": 81
    },
    {
      "epoch": 0.4760522496371553,
      "grad_norm": 2.662794351577759,
      "learning_rate": 6.855986244591104e-05,
      "loss": 0.6189,
      "step": 82
    },
    {
      "epoch": 0.4818577648766328,
      "grad_norm": 2.886531352996826,
      "learning_rate": 6.778969234612584e-05,
      "loss": 0.6457,
      "step": 83
    },
    {
      "epoch": 0.4876632801161103,
      "grad_norm": 2.6005308628082275,
      "learning_rate": 6.701465872208216e-05,
      "loss": 0.57,
      "step": 84
    },
    {
      "epoch": 0.4934687953555878,
      "grad_norm": 2.8066611289978027,
      "learning_rate": 6.623497346023418e-05,
      "loss": 0.6914,
      "step": 85
    },
    {
      "epoch": 0.49927431059506533,
      "grad_norm": 3.6789891719818115,
      "learning_rate": 6.545084971874738e-05,
      "loss": 1.2436,
      "step": 86
    },
    {
      "epoch": 0.5050798258345428,
      "grad_norm": 7.597439289093018,
      "learning_rate": 6.466250186922325e-05,
      "loss": 8.5955,
      "step": 87
    },
    {
      "epoch": 0.5108853410740203,
      "grad_norm": 7.289349555969238,
      "learning_rate": 6.387014543809223e-05,
      "loss": 8.9549,
      "step": 88
    },
    {
      "epoch": 0.5166908563134979,
      "grad_norm": 6.553876876831055,
      "learning_rate": 6.307399704769099e-05,
      "loss": 7.6972,
      "step": 89
    },
    {
      "epoch": 0.5224963715529753,
      "grad_norm": 5.219284534454346,
      "learning_rate": 6.227427435703997e-05,
      "loss": 8.9052,
      "step": 90
    },
    {
      "epoch": 0.5283018867924528,
      "grad_norm": 3.263063430786133,
      "learning_rate": 6.147119600233758e-05,
      "loss": 7.0543,
      "step": 91
    },
    {
      "epoch": 0.5341074020319303,
      "grad_norm": 2.956650733947754,
      "learning_rate": 6.066498153718735e-05,
      "loss": 7.6972,
      "step": 92
    },
    {
      "epoch": 0.5399129172714079,
      "grad_norm": 2.9508557319641113,
      "learning_rate": 5.985585137257401e-05,
      "loss": 8.1985,
      "step": 93
    },
    {
      "epoch": 0.5457184325108854,
      "grad_norm": 3.1433417797088623,
      "learning_rate": 5.90440267166055e-05,
      "loss": 7.8085,
      "step": 94
    },
    {
      "epoch": 0.5515239477503628,
      "grad_norm": 2.7679953575134277,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 7.7311,
      "step": 95
    },
    {
      "epoch": 0.5573294629898403,
      "grad_norm": 2.3550305366516113,
      "learning_rate": 5.74131823855921e-05,
      "loss": 7.4659,
      "step": 96
    },
    {
      "epoch": 0.5631349782293179,
      "grad_norm": 2.226775646209717,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 6.9464,
      "step": 97
    },
    {
      "epoch": 0.5689404934687954,
      "grad_norm": 2.7561795711517334,
      "learning_rate": 5.577423184847932e-05,
      "loss": 7.3918,
      "step": 98
    },
    {
      "epoch": 0.5747460087082729,
      "grad_norm": 3.498291254043579,
      "learning_rate": 5.495227651252315e-05,
      "loss": 7.0608,
      "step": 99
    },
    {
      "epoch": 0.5805515239477503,
      "grad_norm": 3.7650697231292725,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 6.3372,
      "step": 100
    },
    {
      "epoch": 0.5805515239477503,
      "eval_loss": 4.156585693359375,
      "eval_runtime": 22.0005,
      "eval_samples_per_second": 13.181,
      "eval_steps_per_second": 3.318,
      "step": 100
    },
    {
      "epoch": 0.5863570391872278,
      "grad_norm": 4.079714298248291,
      "learning_rate": 5.330452921628497e-05,
      "loss": 5.9446,
      "step": 101
    },
    {
      "epoch": 0.5921625544267054,
      "grad_norm": 5.205807685852051,
      "learning_rate": 5.247918773366112e-05,
      "loss": 6.0759,
      "step": 102
    },
    {
      "epoch": 0.5979680696661829,
      "grad_norm": 6.294081211090088,
      "learning_rate": 5.165316846586541e-05,
      "loss": 5.9588,
      "step": 103
    },
    {
      "epoch": 0.6037735849056604,
      "grad_norm": 4.99482536315918,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 4.1754,
      "step": 104
    },
    {
      "epoch": 0.6095791001451378,
      "grad_norm": 5.103832721710205,
      "learning_rate": 5e-05,
      "loss": 3.4779,
      "step": 105
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 9.644307136535645,
      "learning_rate": 4.917330276168208e-05,
      "loss": 4.129,
      "step": 106
    },
    {
      "epoch": 0.6211901306240929,
      "grad_norm": 4.1341657638549805,
      "learning_rate": 4.834683153413459e-05,
      "loss": 3.7762,
      "step": 107
    },
    {
      "epoch": 0.6269956458635704,
      "grad_norm": 4.543196678161621,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 3.0218,
      "step": 108
    },
    {
      "epoch": 0.6328011611030478,
      "grad_norm": 2.9996960163116455,
      "learning_rate": 4.669547078371504e-05,
      "loss": 1.9354,
      "step": 109
    },
    {
      "epoch": 0.6386066763425254,
      "grad_norm": 2.6885647773742676,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 2.0386,
      "step": 110
    },
    {
      "epoch": 0.6444121915820029,
      "grad_norm": 2.794745683670044,
      "learning_rate": 4.504772348747687e-05,
      "loss": 1.8451,
      "step": 111
    },
    {
      "epoch": 0.6502177068214804,
      "grad_norm": 2.347285509109497,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 1.6655,
      "step": 112
    },
    {
      "epoch": 0.6560232220609579,
      "grad_norm": 2.82843279838562,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 1.6715,
      "step": 113
    },
    {
      "epoch": 0.6618287373004355,
      "grad_norm": 2.8967514038085938,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 1.6404,
      "step": 114
    },
    {
      "epoch": 0.6676342525399129,
      "grad_norm": 3.7805428504943848,
      "learning_rate": 4.17702704859633e-05,
      "loss": 2.1809,
      "step": 115
    },
    {
      "epoch": 0.6734397677793904,
      "grad_norm": 3.838977575302124,
      "learning_rate": 4.095597328339452e-05,
      "loss": 2.1173,
      "step": 116
    },
    {
      "epoch": 0.6792452830188679,
      "grad_norm": 3.3098905086517334,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 1.524,
      "step": 117
    },
    {
      "epoch": 0.6850507982583455,
      "grad_norm": 4.294846534729004,
      "learning_rate": 3.933501846281267e-05,
      "loss": 1.5189,
      "step": 118
    },
    {
      "epoch": 0.690856313497823,
      "grad_norm": 6.8286871910095215,
      "learning_rate": 3.852880399766243e-05,
      "loss": 1.2596,
      "step": 119
    },
    {
      "epoch": 0.6966618287373004,
      "grad_norm": 4.094878196716309,
      "learning_rate": 3.772572564296005e-05,
      "loss": 1.4111,
      "step": 120
    },
    {
      "epoch": 0.7024673439767779,
      "grad_norm": 3.3070149421691895,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 1.5751,
      "step": 121
    },
    {
      "epoch": 0.7082728592162555,
      "grad_norm": 3.541907787322998,
      "learning_rate": 3.612985456190778e-05,
      "loss": 0.9941,
      "step": 122
    },
    {
      "epoch": 0.714078374455733,
      "grad_norm": 2.8128509521484375,
      "learning_rate": 3.533749813077677e-05,
      "loss": 1.1267,
      "step": 123
    },
    {
      "epoch": 0.7198838896952104,
      "grad_norm": 3.034625768661499,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 0.364,
      "step": 124
    },
    {
      "epoch": 0.7256894049346879,
      "grad_norm": 2.248316764831543,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 0.6896,
      "step": 125
    },
    {
      "epoch": 0.7314949201741655,
      "grad_norm": 2.26928448677063,
      "learning_rate": 3.298534127791785e-05,
      "loss": 0.4095,
      "step": 126
    },
    {
      "epoch": 0.737300435413643,
      "grad_norm": 1.9625234603881836,
      "learning_rate": 3.221030765387417e-05,
      "loss": 0.3451,
      "step": 127
    },
    {
      "epoch": 0.7431059506531205,
      "grad_norm": 2.8597424030303955,
      "learning_rate": 3.144013755408895e-05,
      "loss": 0.3181,
      "step": 128
    },
    {
      "epoch": 0.7489114658925979,
      "grad_norm": 3.1393826007843018,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 0.7681,
      "step": 129
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 8.472145080566406,
      "learning_rate": 2.991522876735154e-05,
      "loss": 7.3192,
      "step": 130
    },
    {
      "epoch": 0.760522496371553,
      "grad_norm": 8.743332862854004,
      "learning_rate": 2.916090697523549e-05,
      "loss": 8.5486,
      "step": 131
    },
    {
      "epoch": 0.7663280116110305,
      "grad_norm": 7.07462739944458,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 7.6569,
      "step": 132
    },
    {
      "epoch": 0.772133526850508,
      "grad_norm": 5.875594139099121,
      "learning_rate": 2.766955965735968e-05,
      "loss": 7.4553,
      "step": 133
    },
    {
      "epoch": 0.7779390420899854,
      "grad_norm": 5.53272819519043,
      "learning_rate": 2.693294185106562e-05,
      "loss": 6.3956,
      "step": 134
    },
    {
      "epoch": 0.783744557329463,
      "grad_norm": 4.838965892791748,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 6.4606,
      "step": 135
    },
    {
      "epoch": 0.7895500725689405,
      "grad_norm": 5.362940311431885,
      "learning_rate": 2.547882480847461e-05,
      "loss": 7.7062,
      "step": 136
    },
    {
      "epoch": 0.795355587808418,
      "grad_norm": 5.13655948638916,
      "learning_rate": 2.476172311325783e-05,
      "loss": 7.1767,
      "step": 137
    },
    {
      "epoch": 0.8011611030478955,
      "grad_norm": 3.7281370162963867,
      "learning_rate": 2.405152131093926e-05,
      "loss": 6.5861,
      "step": 138
    },
    {
      "epoch": 0.806966618287373,
      "grad_norm": 4.270654678344727,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 6.1048,
      "step": 139
    },
    {
      "epoch": 0.8127721335268505,
      "grad_norm": 3.9250171184539795,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 5.8365,
      "step": 140
    },
    {
      "epoch": 0.818577648766328,
      "grad_norm": 4.774590492248535,
      "learning_rate": 2.196424713241637e-05,
      "loss": 6.3853,
      "step": 141
    },
    {
      "epoch": 0.8243831640058055,
      "grad_norm": 4.177700042724609,
      "learning_rate": 2.128356686585282e-05,
      "loss": 5.512,
      "step": 142
    },
    {
      "epoch": 0.8301886792452831,
      "grad_norm": 5.186211585998535,
      "learning_rate": 2.061073738537635e-05,
      "loss": 5.4402,
      "step": 143
    },
    {
      "epoch": 0.8359941944847605,
      "grad_norm": 7.278317451477051,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 5.8742,
      "step": 144
    },
    {
      "epoch": 0.841799709724238,
      "grad_norm": 6.769549369812012,
      "learning_rate": 1.928936436551661e-05,
      "loss": 5.3561,
      "step": 145
    },
    {
      "epoch": 0.8476052249637155,
      "grad_norm": 7.398763179779053,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 5.0546,
      "step": 146
    },
    {
      "epoch": 0.8534107402031931,
      "grad_norm": 5.987422943115234,
      "learning_rate": 1.800157297483417e-05,
      "loss": 4.6627,
      "step": 147
    },
    {
      "epoch": 0.8592162554426706,
      "grad_norm": 5.008985996246338,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 3.1983,
      "step": 148
    },
    {
      "epoch": 0.865021770682148,
      "grad_norm": 6.092723369598389,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 2.9265,
      "step": 149
    },
    {
      "epoch": 0.8708272859216255,
      "grad_norm": 6.2369842529296875,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 3.2214,
      "step": 150
    },
    {
      "epoch": 0.8708272859216255,
      "eval_loss": 3.171131134033203,
      "eval_runtime": 22.009,
      "eval_samples_per_second": 13.176,
      "eval_steps_per_second": 3.317,
      "step": 150
    },
    {
      "epoch": 0.8766328011611031,
      "grad_norm": 5.054165363311768,
      "learning_rate": 1.553232954407171e-05,
      "loss": 2.3338,
      "step": 151
    },
    {
      "epoch": 0.8824383164005806,
      "grad_norm": 4.2780890464782715,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 1.8224,
      "step": 152
    },
    {
      "epoch": 0.888243831640058,
      "grad_norm": 3.1665282249450684,
      "learning_rate": 1.435357758543015e-05,
      "loss": 1.4727,
      "step": 153
    },
    {
      "epoch": 0.8940493468795355,
      "grad_norm": 3.309196710586548,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 1.0989,
      "step": 154
    },
    {
      "epoch": 0.8998548621190131,
      "grad_norm": 2.938326835632324,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 1.2263,
      "step": 155
    },
    {
      "epoch": 0.9056603773584906,
      "grad_norm": 3.097607374191284,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 1.216,
      "step": 156
    },
    {
      "epoch": 0.9114658925979681,
      "grad_norm": 3.1081392765045166,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 1.2492,
      "step": 157
    },
    {
      "epoch": 0.9172714078374455,
      "grad_norm": 3.6270923614501953,
      "learning_rate": 1.157994445715706e-05,
      "loss": 1.4846,
      "step": 158
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 3.2071192264556885,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 1.1555,
      "step": 159
    },
    {
      "epoch": 0.9288824383164006,
      "grad_norm": 2.7417211532592773,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 1.1196,
      "step": 160
    },
    {
      "epoch": 0.9346879535558781,
      "grad_norm": 2.8154349327087402,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 0.9573,
      "step": 161
    },
    {
      "epoch": 0.9404934687953556,
      "grad_norm": 3.6433558464050293,
      "learning_rate": 9.549150281252633e-06,
      "loss": 1.3039,
      "step": 162
    },
    {
      "epoch": 0.9462989840348331,
      "grad_norm": 2.5722668170928955,
      "learning_rate": 9.068759265665384e-06,
      "loss": 0.9917,
      "step": 163
    },
    {
      "epoch": 0.9521044992743106,
      "grad_norm": 2.8856451511383057,
      "learning_rate": 8.599558442598998e-06,
      "loss": 1.5762,
      "step": 164
    },
    {
      "epoch": 0.9579100145137881,
      "grad_norm": 3.313309669494629,
      "learning_rate": 8.141676086873572e-06,
      "loss": 0.8928,
      "step": 165
    },
    {
      "epoch": 0.9637155297532656,
      "grad_norm": 3.23502516746521,
      "learning_rate": 7.695237378953223e-06,
      "loss": 0.7147,
      "step": 166
    },
    {
      "epoch": 0.969521044992743,
      "grad_norm": 1.9893702268600464,
      "learning_rate": 7.260364370723044e-06,
      "loss": 0.6746,
      "step": 167
    },
    {
      "epoch": 0.9753265602322206,
      "grad_norm": 1.5415370464324951,
      "learning_rate": 6.837175952121306e-06,
      "loss": 0.336,
      "step": 168
    },
    {
      "epoch": 0.9811320754716981,
      "grad_norm": 1.5752980709075928,
      "learning_rate": 6.425787818636131e-06,
      "loss": 0.4111,
      "step": 169
    },
    {
      "epoch": 0.9869375907111756,
      "grad_norm": 1.804923176765442,
      "learning_rate": 6.026312439675552e-06,
      "loss": 0.1679,
      "step": 170
    },
    {
      "epoch": 0.9927431059506531,
      "grad_norm": 2.1963822841644287,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 0.2333,
      "step": 171
    },
    {
      "epoch": 0.9985486211901307,
      "grad_norm": 2.9731996059417725,
      "learning_rate": 5.263533508961827e-06,
      "loss": 0.5603,
      "step": 172
    },
    {
      "epoch": 1.004354136429608,
      "grad_norm": 12.218384742736816,
      "learning_rate": 4.900438493352055e-06,
      "loss": 8.9302,
      "step": 173
    },
    {
      "epoch": 1.0101596516690856,
      "grad_norm": 9.984807014465332,
      "learning_rate": 4.549673247541875e-06,
      "loss": 7.4938,
      "step": 174
    },
    {
      "epoch": 1.0159651669085632,
      "grad_norm": 9.16197681427002,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 6.9905,
      "step": 175
    },
    {
      "epoch": 1.0217706821480406,
      "grad_norm": 9.399922370910645,
      "learning_rate": 3.885512251130763e-06,
      "loss": 7.0781,
      "step": 176
    },
    {
      "epoch": 1.0275761973875182,
      "grad_norm": 8.3427152633667,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 6.8744,
      "step": 177
    },
    {
      "epoch": 1.0333817126269957,
      "grad_norm": 7.790874004364014,
      "learning_rate": 3.271776770026963e-06,
      "loss": 5.7031,
      "step": 178
    },
    {
      "epoch": 1.039187227866473,
      "grad_norm": 7.484927654266357,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 5.9761,
      "step": 179
    },
    {
      "epoch": 1.0449927431059507,
      "grad_norm": 9.040209770202637,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 7.0571,
      "step": 180
    },
    {
      "epoch": 1.050798258345428,
      "grad_norm": 7.728246212005615,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 6.5578,
      "step": 181
    },
    {
      "epoch": 1.0566037735849056,
      "grad_norm": 6.477789402008057,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 5.5775,
      "step": 182
    },
    {
      "epoch": 1.0624092888243832,
      "grad_norm": 6.882674694061279,
      "learning_rate": 1.962316193157593e-06,
      "loss": 5.4078,
      "step": 183
    },
    {
      "epoch": 1.0682148040638606,
      "grad_norm": 6.807581424713135,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 4.7827,
      "step": 184
    },
    {
      "epoch": 1.0740203193033382,
      "grad_norm": 7.545950412750244,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 5.3387,
      "step": 185
    },
    {
      "epoch": 1.0798258345428158,
      "grad_norm": 7.99847412109375,
      "learning_rate": 1.333670137599713e-06,
      "loss": 4.08,
      "step": 186
    },
    {
      "epoch": 1.0856313497822931,
      "grad_norm": 10.168889045715332,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 4.9327,
      "step": 187
    },
    {
      "epoch": 1.0914368650217707,
      "grad_norm": 7.2750959396362305,
      "learning_rate": 9.810017062595322e-07,
      "loss": 3.8219,
      "step": 188
    },
    {
      "epoch": 1.097242380261248,
      "grad_norm": 7.462510108947754,
      "learning_rate": 8.247462563808817e-07,
      "loss": 4.0535,
      "step": 189
    },
    {
      "epoch": 1.1030478955007257,
      "grad_norm": 5.24201774597168,
      "learning_rate": 6.819348298638839e-07,
      "loss": 3.0705,
      "step": 190
    },
    {
      "epoch": 1.1088534107402033,
      "grad_norm": 4.043258190155029,
      "learning_rate": 5.526064699265753e-07,
      "loss": 1.7934,
      "step": 191
    },
    {
      "epoch": 1.1146589259796806,
      "grad_norm": 5.162657737731934,
      "learning_rate": 4.367965336512403e-07,
      "loss": 2.2909,
      "step": 192
    },
    {
      "epoch": 1.1204644412191582,
      "grad_norm": 4.913023948669434,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 2.1575,
      "step": 193
    },
    {
      "epoch": 1.1262699564586356,
      "grad_norm": 5.183408260345459,
      "learning_rate": 2.458548727494292e-07,
      "loss": 1.3281,
      "step": 194
    },
    {
      "epoch": 1.1320754716981132,
      "grad_norm": 3.4359192848205566,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 0.745,
      "step": 195
    },
    {
      "epoch": 1.1378809869375908,
      "grad_norm": 3.918090581893921,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 1.0365,
      "step": 196
    },
    {
      "epoch": 1.1436865021770681,
      "grad_norm": 2.9474949836730957,
      "learning_rate": 6.150154258476315e-08,
      "loss": 0.5556,
      "step": 197
    },
    {
      "epoch": 1.1494920174165457,
      "grad_norm": 2.838902235031128,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 0.8301,
      "step": 198
    },
    {
      "epoch": 1.1552975326560233,
      "grad_norm": 2.094698190689087,
      "learning_rate": 6.834750376549792e-09,
      "loss": 0.8476,
      "step": 199
    },
    {
      "epoch": 1.1611030478955007,
      "grad_norm": 2.4736406803131104,
      "learning_rate": 0.0,
      "loss": 0.7629,
      "step": 200
    },
    {
      "epoch": 1.1611030478955007,
      "eval_loss": 2.7577829360961914,
      "eval_runtime": 22.0635,
      "eval_samples_per_second": 13.144,
      "eval_steps_per_second": 3.309,
      "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": 3.016084336504996e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}