{
  "best_metric": 1.3994430303573608,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.0658165364047717,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00032908268202385847,
      "grad_norm": 2.5778968334198,
      "learning_rate": 1e-05,
      "loss": 5.3014,
      "step": 1
    },
    {
      "epoch": 0.00032908268202385847,
      "eval_loss": 1.6017260551452637,
      "eval_runtime": 107.894,
      "eval_samples_per_second": 47.435,
      "eval_steps_per_second": 11.863,
      "step": 1
    },
    {
      "epoch": 0.0006581653640477169,
      "grad_norm": 2.6044929027557373,
      "learning_rate": 2e-05,
      "loss": 5.0754,
      "step": 2
    },
    {
      "epoch": 0.0009872480460715754,
      "grad_norm": 2.145354986190796,
      "learning_rate": 3e-05,
      "loss": 5.2133,
      "step": 3
    },
    {
      "epoch": 0.0013163307280954339,
      "grad_norm": 2.242710828781128,
      "learning_rate": 4e-05,
      "loss": 5.3687,
      "step": 4
    },
    {
      "epoch": 0.0016454134101192926,
      "grad_norm": 1.8690861463546753,
      "learning_rate": 5e-05,
      "loss": 5.069,
      "step": 5
    },
    {
      "epoch": 0.001974496092143151,
      "grad_norm": 1.8592292070388794,
      "learning_rate": 6e-05,
      "loss": 5.2433,
      "step": 6
    },
    {
      "epoch": 0.0023035787741670093,
      "grad_norm": 2.0510525703430176,
      "learning_rate": 7e-05,
      "loss": 5.0986,
      "step": 7
    },
    {
      "epoch": 0.0026326614561908677,
      "grad_norm": 2.3034753799438477,
      "learning_rate": 8e-05,
      "loss": 5.6054,
      "step": 8
    },
    {
      "epoch": 0.0029617441382147267,
      "grad_norm": 2.1648271083831787,
      "learning_rate": 9e-05,
      "loss": 5.5764,
      "step": 9
    },
    {
      "epoch": 0.003290826820238585,
      "grad_norm": 2.0305778980255127,
      "learning_rate": 0.0001,
      "loss": 5.3878,
      "step": 10
    },
    {
      "epoch": 0.0036199095022624436,
      "grad_norm": 2.0604233741760254,
      "learning_rate": 9.999316524962345e-05,
      "loss": 5.5059,
      "step": 11
    },
    {
      "epoch": 0.003948992184286302,
      "grad_norm": 2.3885691165924072,
      "learning_rate": 9.997266286704631e-05,
      "loss": 5.6024,
      "step": 12
    },
    {
      "epoch": 0.0042780748663101605,
      "grad_norm": 2.0734753608703613,
      "learning_rate": 9.993849845741524e-05,
      "loss": 5.3144,
      "step": 13
    },
    {
      "epoch": 0.0046071575483340186,
      "grad_norm": 1.897387146949768,
      "learning_rate": 9.989068136093873e-05,
      "loss": 5.3028,
      "step": 14
    },
    {
      "epoch": 0.0049362402303578775,
      "grad_norm": 2.152360439300537,
      "learning_rate": 9.98292246503335e-05,
      "loss": 5.6148,
      "step": 15
    },
    {
      "epoch": 0.0052653229123817355,
      "grad_norm": 2.202516794204712,
      "learning_rate": 9.975414512725057e-05,
      "loss": 5.5594,
      "step": 16
    },
    {
      "epoch": 0.005594405594405594,
      "grad_norm": 2.2090365886688232,
      "learning_rate": 9.966546331768191e-05,
      "loss": 5.4872,
      "step": 17
    },
    {
      "epoch": 0.005923488276429453,
      "grad_norm": 2.3608784675598145,
      "learning_rate": 9.956320346634876e-05,
      "loss": 5.6626,
      "step": 18
    },
    {
      "epoch": 0.006252570958453311,
      "grad_norm": 2.2158000469207764,
      "learning_rate": 9.944739353007344e-05,
      "loss": 5.294,
      "step": 19
    },
    {
      "epoch": 0.00658165364047717,
      "grad_norm": 2.3356127738952637,
      "learning_rate": 9.931806517013612e-05,
      "loss": 5.577,
      "step": 20
    },
    {
      "epoch": 0.006910736322501028,
      "grad_norm": 2.3093531131744385,
      "learning_rate": 9.917525374361912e-05,
      "loss": 5.556,
      "step": 21
    },
    {
      "epoch": 0.007239819004524887,
      "grad_norm": 2.137331485748291,
      "learning_rate": 9.901899829374047e-05,
      "loss": 5.2592,
      "step": 22
    },
    {
      "epoch": 0.007568901686548745,
      "grad_norm": 2.363743305206299,
      "learning_rate": 9.884934153917997e-05,
      "loss": 5.7913,
      "step": 23
    },
    {
      "epoch": 0.007897984368572603,
      "grad_norm": 2.4328322410583496,
      "learning_rate": 9.86663298624003e-05,
      "loss": 5.6937,
      "step": 24
    },
    {
      "epoch": 0.008227067050596462,
      "grad_norm": 2.290703058242798,
      "learning_rate": 9.847001329696653e-05,
      "loss": 5.6763,
      "step": 25
    },
    {
      "epoch": 0.008556149732620321,
      "grad_norm": 2.5753531455993652,
      "learning_rate": 9.826044551386744e-05,
      "loss": 5.9247,
      "step": 26
    },
    {
      "epoch": 0.00888523241464418,
      "grad_norm": 2.5030720233917236,
      "learning_rate": 9.803768380684242e-05,
      "loss": 5.8785,
      "step": 27
    },
    {
      "epoch": 0.009214315096668037,
      "grad_norm": 2.604569435119629,
      "learning_rate": 9.780178907671789e-05,
      "loss": 5.9158,
      "step": 28
    },
    {
      "epoch": 0.009543397778691896,
      "grad_norm": 2.787095546722412,
      "learning_rate": 9.755282581475769e-05,
      "loss": 5.71,
      "step": 29
    },
    {
      "epoch": 0.009872480460715755,
      "grad_norm": 2.845724105834961,
      "learning_rate": 9.729086208503174e-05,
      "loss": 5.4528,
      "step": 30
    },
    {
      "epoch": 0.010201563142739614,
      "grad_norm": 2.8583714962005615,
      "learning_rate": 9.701596950580806e-05,
      "loss": 5.9774,
      "step": 31
    },
    {
      "epoch": 0.010530645824763471,
      "grad_norm": 3.274951219558716,
      "learning_rate": 9.672822322997305e-05,
      "loss": 6.4037,
      "step": 32
    },
    {
      "epoch": 0.01085972850678733,
      "grad_norm": 3.177806854248047,
      "learning_rate": 9.642770192448536e-05,
      "loss": 6.2072,
      "step": 33
    },
    {
      "epoch": 0.011188811188811189,
      "grad_norm": 3.142428398132324,
      "learning_rate": 9.611448774886924e-05,
      "loss": 5.7776,
      "step": 34
    },
    {
      "epoch": 0.011517893870835048,
      "grad_norm": 3.3742871284484863,
      "learning_rate": 9.578866633275288e-05,
      "loss": 6.4645,
      "step": 35
    },
    {
      "epoch": 0.011846976552858907,
      "grad_norm": 3.2504143714904785,
      "learning_rate": 9.545032675245813e-05,
      "loss": 5.7659,
      "step": 36
    },
    {
      "epoch": 0.012176059234882764,
      "grad_norm": 3.792818307876587,
      "learning_rate": 9.509956150664796e-05,
      "loss": 6.3716,
      "step": 37
    },
    {
      "epoch": 0.012505141916906623,
      "grad_norm": 3.7317750453948975,
      "learning_rate": 9.473646649103818e-05,
      "loss": 6.4501,
      "step": 38
    },
    {
      "epoch": 0.012834224598930482,
      "grad_norm": 4.011148929595947,
      "learning_rate": 9.43611409721806e-05,
      "loss": 6.168,
      "step": 39
    },
    {
      "epoch": 0.01316330728095434,
      "grad_norm": 3.899348735809326,
      "learning_rate": 9.397368756032445e-05,
      "loss": 6.1079,
      "step": 40
    },
    {
      "epoch": 0.013492389962978198,
      "grad_norm": 4.315437316894531,
      "learning_rate": 9.357421218136386e-05,
      "loss": 6.3949,
      "step": 41
    },
    {
      "epoch": 0.013821472645002057,
      "grad_norm": 4.627906322479248,
      "learning_rate": 9.316282404787871e-05,
      "loss": 6.7348,
      "step": 42
    },
    {
      "epoch": 0.014150555327025915,
      "grad_norm": 5.08112907409668,
      "learning_rate": 9.273963562927695e-05,
      "loss": 6.2201,
      "step": 43
    },
    {
      "epoch": 0.014479638009049774,
      "grad_norm": 5.626524925231934,
      "learning_rate": 9.230476262104677e-05,
      "loss": 7.4486,
      "step": 44
    },
    {
      "epoch": 0.014808720691073632,
      "grad_norm": 5.002148628234863,
      "learning_rate": 9.185832391312644e-05,
      "loss": 6.3462,
      "step": 45
    },
    {
      "epoch": 0.01513780337309749,
      "grad_norm": 5.748898029327393,
      "learning_rate": 9.140044155740101e-05,
      "loss": 7.1086,
      "step": 46
    },
    {
      "epoch": 0.01546688605512135,
      "grad_norm": 6.279057025909424,
      "learning_rate": 9.093124073433463e-05,
      "loss": 7.5006,
      "step": 47
    },
    {
      "epoch": 0.015795968737145206,
      "grad_norm": 8.499492645263672,
      "learning_rate": 9.045084971874738e-05,
      "loss": 8.2064,
      "step": 48
    },
    {
      "epoch": 0.016125051419169065,
      "grad_norm": 10.74296760559082,
      "learning_rate": 8.995939984474624e-05,
      "loss": 8.5221,
      "step": 49
    },
    {
      "epoch": 0.016454134101192924,
      "grad_norm": 14.588689804077148,
      "learning_rate": 8.945702546981969e-05,
      "loss": 9.7044,
      "step": 50
    },
    {
      "epoch": 0.016454134101192924,
      "eval_loss": 1.5887035131454468,
      "eval_runtime": 108.4856,
      "eval_samples_per_second": 47.177,
      "eval_steps_per_second": 11.799,
      "step": 50
    },
    {
      "epoch": 0.016783216783216783,
      "grad_norm": 14.235264778137207,
      "learning_rate": 8.894386393810563e-05,
      "loss": 6.0633,
      "step": 51
    },
    {
      "epoch": 0.017112299465240642,
      "grad_norm": 11.91590404510498,
      "learning_rate": 8.842005554284296e-05,
      "loss": 5.8601,
      "step": 52
    },
    {
      "epoch": 0.0174413821472645,
      "grad_norm": 8.146944046020508,
      "learning_rate": 8.788574348801675e-05,
      "loss": 5.6869,
      "step": 53
    },
    {
      "epoch": 0.01777046482928836,
      "grad_norm": 5.188728332519531,
      "learning_rate": 8.73410738492077e-05,
      "loss": 5.7059,
      "step": 54
    },
    {
      "epoch": 0.01809954751131222,
      "grad_norm": 3.3103089332580566,
      "learning_rate": 8.678619553365659e-05,
      "loss": 5.2213,
      "step": 55
    },
    {
      "epoch": 0.018428630193336074,
      "grad_norm": 2.4259917736053467,
      "learning_rate": 8.622126023955446e-05,
      "loss": 5.4405,
      "step": 56
    },
    {
      "epoch": 0.018757712875359933,
      "grad_norm": 1.9551986455917358,
      "learning_rate": 8.564642241456986e-05,
      "loss": 5.1664,
      "step": 57
    },
    {
      "epoch": 0.019086795557383792,
      "grad_norm": 1.8865058422088623,
      "learning_rate": 8.506183921362443e-05,
      "loss": 5.1533,
      "step": 58
    },
    {
      "epoch": 0.01941587823940765,
      "grad_norm": 1.9420619010925293,
      "learning_rate": 8.44676704559283e-05,
      "loss": 5.3082,
      "step": 59
    },
    {
      "epoch": 0.01974496092143151,
      "grad_norm": 1.8333953619003296,
      "learning_rate": 8.386407858128706e-05,
      "loss": 5.2478,
      "step": 60
    },
    {
      "epoch": 0.02007404360345537,
      "grad_norm": 1.6116563081741333,
      "learning_rate": 8.32512286056924e-05,
      "loss": 5.3659,
      "step": 61
    },
    {
      "epoch": 0.020403126285479228,
      "grad_norm": 1.5631070137023926,
      "learning_rate": 8.262928807620843e-05,
      "loss": 5.4314,
      "step": 62
    },
    {
      "epoch": 0.020732208967503087,
      "grad_norm": 1.504607915878296,
      "learning_rate": 8.199842702516583e-05,
      "loss": 5.0626,
      "step": 63
    },
    {
      "epoch": 0.021061291649526942,
      "grad_norm": 1.5705896615982056,
      "learning_rate": 8.135881792367686e-05,
      "loss": 5.5714,
      "step": 64
    },
    {
      "epoch": 0.0213903743315508,
      "grad_norm": 1.606120228767395,
      "learning_rate": 8.07106356344834e-05,
      "loss": 5.4141,
      "step": 65
    },
    {
      "epoch": 0.02171945701357466,
      "grad_norm": 1.5568580627441406,
      "learning_rate": 8.005405736415126e-05,
      "loss": 5.4226,
      "step": 66
    },
    {
      "epoch": 0.02204853969559852,
      "grad_norm": 1.5692131519317627,
      "learning_rate": 7.938926261462366e-05,
      "loss": 5.1188,
      "step": 67
    },
    {
      "epoch": 0.022377622377622378,
      "grad_norm": 1.63429856300354,
      "learning_rate": 7.871643313414718e-05,
      "loss": 5.4083,
      "step": 68
    },
    {
      "epoch": 0.022706705059646237,
      "grad_norm": 1.6308414936065674,
      "learning_rate": 7.803575286758364e-05,
      "loss": 5.4212,
      "step": 69
    },
    {
      "epoch": 0.023035787741670095,
      "grad_norm": 1.7185771465301514,
      "learning_rate": 7.734740790612136e-05,
      "loss": 5.658,
      "step": 70
    },
    {
      "epoch": 0.023364870423693954,
      "grad_norm": 1.7298237085342407,
      "learning_rate": 7.66515864363997e-05,
      "loss": 5.8036,
      "step": 71
    },
    {
      "epoch": 0.023693953105717813,
      "grad_norm": 1.6765174865722656,
      "learning_rate": 7.594847868906076e-05,
      "loss": 5.4251,
      "step": 72
    },
    {
      "epoch": 0.02402303578774167,
      "grad_norm": 1.7852168083190918,
      "learning_rate": 7.52382768867422e-05,
      "loss": 5.6781,
      "step": 73
    },
    {
      "epoch": 0.024352118469765528,
      "grad_norm": 1.8785332441329956,
      "learning_rate": 7.452117519152542e-05,
      "loss": 5.6657,
      "step": 74
    },
    {
      "epoch": 0.024681201151789386,
      "grad_norm": 1.9445163011550903,
      "learning_rate": 7.379736965185368e-05,
      "loss": 5.2849,
      "step": 75
    },
    {
      "epoch": 0.025010283833813245,
      "grad_norm": 2.142787218093872,
      "learning_rate": 7.30670581489344e-05,
      "loss": 5.7116,
      "step": 76
    },
    {
      "epoch": 0.025339366515837104,
      "grad_norm": 2.126593828201294,
      "learning_rate": 7.233044034264034e-05,
      "loss": 5.69,
      "step": 77
    },
    {
      "epoch": 0.025668449197860963,
      "grad_norm": 2.2754554748535156,
      "learning_rate": 7.158771761692464e-05,
      "loss": 5.6095,
      "step": 78
    },
    {
      "epoch": 0.025997531879884822,
      "grad_norm": 2.201345682144165,
      "learning_rate": 7.083909302476453e-05,
      "loss": 5.5129,
      "step": 79
    },
    {
      "epoch": 0.02632661456190868,
      "grad_norm": 2.3233513832092285,
      "learning_rate": 7.008477123264848e-05,
      "loss": 5.7344,
      "step": 80
    },
    {
      "epoch": 0.026655697243932536,
      "grad_norm": 2.4314122200012207,
      "learning_rate": 6.932495846462261e-05,
      "loss": 6.1961,
      "step": 81
    },
    {
      "epoch": 0.026984779925956395,
      "grad_norm": 2.332638740539551,
      "learning_rate": 6.855986244591104e-05,
      "loss": 5.5605,
      "step": 82
    },
    {
      "epoch": 0.027313862607980254,
      "grad_norm": 2.543510675430298,
      "learning_rate": 6.778969234612584e-05,
      "loss": 6.0827,
      "step": 83
    },
    {
      "epoch": 0.027642945290004113,
      "grad_norm": 2.736861228942871,
      "learning_rate": 6.701465872208216e-05,
      "loss": 6.079,
      "step": 84
    },
    {
      "epoch": 0.027972027972027972,
      "grad_norm": 2.8094890117645264,
      "learning_rate": 6.623497346023418e-05,
      "loss": 5.9357,
      "step": 85
    },
    {
      "epoch": 0.02830111065405183,
      "grad_norm": 2.7342352867126465,
      "learning_rate": 6.545084971874738e-05,
      "loss": 6.0413,
      "step": 86
    },
    {
      "epoch": 0.02863019333607569,
      "grad_norm": 2.9445390701293945,
      "learning_rate": 6.466250186922325e-05,
      "loss": 5.9232,
      "step": 87
    },
    {
      "epoch": 0.02895927601809955,
      "grad_norm": 3.0767359733581543,
      "learning_rate": 6.387014543809223e-05,
      "loss": 6.4853,
      "step": 88
    },
    {
      "epoch": 0.029288358700123408,
      "grad_norm": 3.2457170486450195,
      "learning_rate": 6.307399704769099e-05,
      "loss": 6.2963,
      "step": 89
    },
    {
      "epoch": 0.029617441382147263,
      "grad_norm": 3.223606586456299,
      "learning_rate": 6.227427435703997e-05,
      "loss": 6.4778,
      "step": 90
    },
    {
      "epoch": 0.029946524064171122,
      "grad_norm": 3.691720724105835,
      "learning_rate": 6.147119600233758e-05,
      "loss": 6.2681,
      "step": 91
    },
    {
      "epoch": 0.03027560674619498,
      "grad_norm": 4.100921154022217,
      "learning_rate": 6.066498153718735e-05,
      "loss": 6.7137,
      "step": 92
    },
    {
      "epoch": 0.03060468942821884,
      "grad_norm": 4.009209156036377,
      "learning_rate": 5.985585137257401e-05,
      "loss": 6.8803,
      "step": 93
    },
    {
      "epoch": 0.0309337721102427,
      "grad_norm": 5.135426044464111,
      "learning_rate": 5.90440267166055e-05,
      "loss": 7.223,
      "step": 94
    },
    {
      "epoch": 0.03126285479226656,
      "grad_norm": 4.95713996887207,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 6.8503,
      "step": 95
    },
    {
      "epoch": 0.03159193747429041,
      "grad_norm": 5.09060001373291,
      "learning_rate": 5.74131823855921e-05,
      "loss": 6.8141,
      "step": 96
    },
    {
      "epoch": 0.031921020156314275,
      "grad_norm": 4.727902889251709,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 7.0527,
      "step": 97
    },
    {
      "epoch": 0.03225010283833813,
      "grad_norm": 5.564077377319336,
      "learning_rate": 5.577423184847932e-05,
      "loss": 7.2004,
      "step": 98
    },
    {
      "epoch": 0.03257918552036199,
      "grad_norm": 6.635467529296875,
      "learning_rate": 5.495227651252315e-05,
      "loss": 7.9314,
      "step": 99
    },
    {
      "epoch": 0.03290826820238585,
      "grad_norm": 8.974549293518066,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 7.938,
      "step": 100
    },
    {
      "epoch": 0.03290826820238585,
      "eval_loss": 1.44220769405365,
      "eval_runtime": 108.4706,
      "eval_samples_per_second": 47.183,
      "eval_steps_per_second": 11.8,
      "step": 100
    },
    {
      "epoch": 0.03323735088440971,
      "grad_norm": 3.6810882091522217,
      "learning_rate": 5.330452921628497e-05,
      "loss": 5.1349,
      "step": 101
    },
    {
      "epoch": 0.033566433566433566,
      "grad_norm": 3.7210822105407715,
      "learning_rate": 5.247918773366112e-05,
      "loss": 5.1968,
      "step": 102
    },
    {
      "epoch": 0.03389551624845742,
      "grad_norm": 2.9325783252716064,
      "learning_rate": 5.165316846586541e-05,
      "loss": 5.1365,
      "step": 103
    },
    {
      "epoch": 0.034224598930481284,
      "grad_norm": 2.1330795288085938,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 4.9477,
      "step": 104
    },
    {
      "epoch": 0.03455368161250514,
      "grad_norm": 1.6018720865249634,
      "learning_rate": 5e-05,
      "loss": 5.3499,
      "step": 105
    },
    {
      "epoch": 0.034882764294529,
      "grad_norm": 1.413811206817627,
      "learning_rate": 4.917330276168208e-05,
      "loss": 4.9815,
      "step": 106
    },
    {
      "epoch": 0.03521184697655286,
      "grad_norm": 1.3371868133544922,
      "learning_rate": 4.834683153413459e-05,
      "loss": 4.994,
      "step": 107
    },
    {
      "epoch": 0.03554092965857672,
      "grad_norm": 1.3457586765289307,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 4.9829,
      "step": 108
    },
    {
      "epoch": 0.035870012340600575,
      "grad_norm": 1.4395283460617065,
      "learning_rate": 4.669547078371504e-05,
      "loss": 5.1806,
      "step": 109
    },
    {
      "epoch": 0.03619909502262444,
      "grad_norm": 1.3400384187698364,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 4.9449,
      "step": 110
    },
    {
      "epoch": 0.03652817770464829,
      "grad_norm": 1.4371681213378906,
      "learning_rate": 4.504772348747687e-05,
      "loss": 5.1171,
      "step": 111
    },
    {
      "epoch": 0.03685726038667215,
      "grad_norm": 1.4166725873947144,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 5.2312,
      "step": 112
    },
    {
      "epoch": 0.03718634306869601,
      "grad_norm": 1.5122697353363037,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 5.3472,
      "step": 113
    },
    {
      "epoch": 0.037515425750719866,
      "grad_norm": 1.5778615474700928,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 5.508,
      "step": 114
    },
    {
      "epoch": 0.03784450843274373,
      "grad_norm": 1.4268945455551147,
      "learning_rate": 4.17702704859633e-05,
      "loss": 5.2124,
      "step": 115
    },
    {
      "epoch": 0.038173591114767584,
      "grad_norm": 1.5441595315933228,
      "learning_rate": 4.095597328339452e-05,
      "loss": 5.4638,
      "step": 116
    },
    {
      "epoch": 0.038502673796791446,
      "grad_norm": 1.4720258712768555,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 5.2319,
      "step": 117
    },
    {
      "epoch": 0.0388317564788153,
      "grad_norm": 1.5355894565582275,
      "learning_rate": 3.933501846281267e-05,
      "loss": 5.565,
      "step": 118
    },
    {
      "epoch": 0.039160839160839164,
      "grad_norm": 1.7273974418640137,
      "learning_rate": 3.852880399766243e-05,
      "loss": 5.5324,
      "step": 119
    },
    {
      "epoch": 0.03948992184286302,
      "grad_norm": 1.6047154664993286,
      "learning_rate": 3.772572564296005e-05,
      "loss": 5.4994,
      "step": 120
    },
    {
      "epoch": 0.039819004524886875,
      "grad_norm": 1.6618292331695557,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 5.7171,
      "step": 121
    },
    {
      "epoch": 0.04014808720691074,
      "grad_norm": 1.6939467191696167,
      "learning_rate": 3.612985456190778e-05,
      "loss": 5.1011,
      "step": 122
    },
    {
      "epoch": 0.04047716988893459,
      "grad_norm": 1.7045646905899048,
      "learning_rate": 3.533749813077677e-05,
      "loss": 5.2405,
      "step": 123
    },
    {
      "epoch": 0.040806252570958455,
      "grad_norm": 1.8032441139221191,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 5.7272,
      "step": 124
    },
    {
      "epoch": 0.04113533525298231,
      "grad_norm": 1.8825111389160156,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 5.4933,
      "step": 125
    },
    {
      "epoch": 0.04146441793500617,
      "grad_norm": 1.9822322130203247,
      "learning_rate": 3.298534127791785e-05,
      "loss": 5.8981,
      "step": 126
    },
    {
      "epoch": 0.04179350061703003,
      "grad_norm": 1.9839398860931396,
      "learning_rate": 3.221030765387417e-05,
      "loss": 5.5461,
      "step": 127
    },
    {
      "epoch": 0.042122583299053884,
      "grad_norm": 2.2322514057159424,
      "learning_rate": 3.144013755408895e-05,
      "loss": 6.034,
      "step": 128
    },
    {
      "epoch": 0.042451665981077746,
      "grad_norm": 2.266590118408203,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 5.6987,
      "step": 129
    },
    {
      "epoch": 0.0427807486631016,
      "grad_norm": 2.2711355686187744,
      "learning_rate": 2.991522876735154e-05,
      "loss": 5.8564,
      "step": 130
    },
    {
      "epoch": 0.043109831345125464,
      "grad_norm": 2.230027675628662,
      "learning_rate": 2.916090697523549e-05,
      "loss": 6.0793,
      "step": 131
    },
    {
      "epoch": 0.04343891402714932,
      "grad_norm": 2.445558786392212,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 5.8758,
      "step": 132
    },
    {
      "epoch": 0.04376799670917318,
      "grad_norm": 2.2881009578704834,
      "learning_rate": 2.766955965735968e-05,
      "loss": 5.3625,
      "step": 133
    },
    {
      "epoch": 0.04409707939119704,
      "grad_norm": 2.5429751873016357,
      "learning_rate": 2.693294185106562e-05,
      "loss": 5.9394,
      "step": 134
    },
    {
      "epoch": 0.0444261620732209,
      "grad_norm": 2.66085147857666,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 6.0131,
      "step": 135
    },
    {
      "epoch": 0.044755244755244755,
      "grad_norm": 2.6408989429473877,
      "learning_rate": 2.547882480847461e-05,
      "loss": 6.0871,
      "step": 136
    },
    {
      "epoch": 0.04508432743726861,
      "grad_norm": 2.7951669692993164,
      "learning_rate": 2.476172311325783e-05,
      "loss": 5.7674,
      "step": 137
    },
    {
      "epoch": 0.04541341011929247,
      "grad_norm": 2.793738603591919,
      "learning_rate": 2.405152131093926e-05,
      "loss": 6.2006,
      "step": 138
    },
    {
      "epoch": 0.04574249280131633,
      "grad_norm": 3.1170003414154053,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 6.2622,
      "step": 139
    },
    {
      "epoch": 0.04607157548334019,
      "grad_norm": 3.1824400424957275,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 6.3068,
      "step": 140
    },
    {
      "epoch": 0.046400658165364046,
      "grad_norm": 3.123488187789917,
      "learning_rate": 2.196424713241637e-05,
      "loss": 5.832,
      "step": 141
    },
    {
      "epoch": 0.04672974084738791,
      "grad_norm": 3.353222370147705,
      "learning_rate": 2.128356686585282e-05,
      "loss": 6.3385,
      "step": 142
    },
    {
      "epoch": 0.047058823529411764,
      "grad_norm": 4.413583278656006,
      "learning_rate": 2.061073738537635e-05,
      "loss": 7.4745,
      "step": 143
    },
    {
      "epoch": 0.047387906211435626,
      "grad_norm": 3.9719223976135254,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 6.7669,
      "step": 144
    },
    {
      "epoch": 0.04771698889345948,
      "grad_norm": 4.2340803146362305,
      "learning_rate": 1.928936436551661e-05,
      "loss": 6.6568,
      "step": 145
    },
    {
      "epoch": 0.04804607157548334,
      "grad_norm": 4.857604503631592,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 6.8252,
      "step": 146
    },
    {
      "epoch": 0.0483751542575072,
      "grad_norm": 5.29097843170166,
      "learning_rate": 1.800157297483417e-05,
      "loss": 7.6368,
      "step": 147
    },
    {
      "epoch": 0.048704236939531055,
      "grad_norm": 5.462820053100586,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 6.7752,
      "step": 148
    },
    {
      "epoch": 0.04903331962155492,
      "grad_norm": 5.956571578979492,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 7.2624,
      "step": 149
    },
    {
      "epoch": 0.04936240230357877,
      "grad_norm": 7.7059760093688965,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 8.0265,
      "step": 150
    },
    {
      "epoch": 0.04936240230357877,
      "eval_loss": 1.4010411500930786,
      "eval_runtime": 108.5301,
      "eval_samples_per_second": 47.157,
      "eval_steps_per_second": 11.794,
      "step": 150
    },
    {
      "epoch": 0.049691484985602635,
      "grad_norm": 1.2888128757476807,
      "learning_rate": 1.553232954407171e-05,
      "loss": 5.0022,
      "step": 151
    },
    {
      "epoch": 0.05002056766762649,
      "grad_norm": 1.2177847623825073,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 5.0077,
      "step": 152
    },
    {
      "epoch": 0.05034965034965035,
      "grad_norm": 1.2098649740219116,
      "learning_rate": 1.435357758543015e-05,
      "loss": 4.9148,
      "step": 153
    },
    {
      "epoch": 0.05067873303167421,
      "grad_norm": 1.2028580904006958,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 4.9005,
      "step": 154
    },
    {
      "epoch": 0.051007815713698064,
      "grad_norm": 1.2271428108215332,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 4.9403,
      "step": 155
    },
    {
      "epoch": 0.051336898395721926,
      "grad_norm": 1.246469497680664,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 5.1103,
      "step": 156
    },
    {
      "epoch": 0.05166598107774578,
      "grad_norm": 1.1502357721328735,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 5.1616,
      "step": 157
    },
    {
      "epoch": 0.051995063759769644,
      "grad_norm": 1.1639046669006348,
      "learning_rate": 1.157994445715706e-05,
      "loss": 5.0807,
      "step": 158
    },
    {
      "epoch": 0.0523241464417935,
      "grad_norm": 1.190629243850708,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 5.0023,
      "step": 159
    },
    {
      "epoch": 0.05265322912381736,
      "grad_norm": 1.25421142578125,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 5.3119,
      "step": 160
    },
    {
      "epoch": 0.05298231180584122,
      "grad_norm": 1.2661880254745483,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 5.0613,
      "step": 161
    },
    {
      "epoch": 0.05331139448786507,
      "grad_norm": 1.2556160688400269,
      "learning_rate": 9.549150281252633e-06,
      "loss": 5.1723,
      "step": 162
    },
    {
      "epoch": 0.053640477169888935,
      "grad_norm": 1.2914961576461792,
      "learning_rate": 9.068759265665384e-06,
      "loss": 5.3062,
      "step": 163
    },
    {
      "epoch": 0.05396955985191279,
      "grad_norm": 1.3207286596298218,
      "learning_rate": 8.599558442598998e-06,
      "loss": 5.3207,
      "step": 164
    },
    {
      "epoch": 0.05429864253393665,
      "grad_norm": 1.4279741048812866,
      "learning_rate": 8.141676086873572e-06,
      "loss": 5.1589,
      "step": 165
    },
    {
      "epoch": 0.05462772521596051,
      "grad_norm": 1.508889079093933,
      "learning_rate": 7.695237378953223e-06,
      "loss": 5.8427,
      "step": 166
    },
    {
      "epoch": 0.05495680789798437,
      "grad_norm": 1.4998936653137207,
      "learning_rate": 7.260364370723044e-06,
      "loss": 5.4891,
      "step": 167
    },
    {
      "epoch": 0.055285890580008226,
      "grad_norm": 1.5592706203460693,
      "learning_rate": 6.837175952121306e-06,
      "loss": 5.3468,
      "step": 168
    },
    {
      "epoch": 0.05561497326203209,
      "grad_norm": 1.6828714609146118,
      "learning_rate": 6.425787818636131e-06,
      "loss": 5.5534,
      "step": 169
    },
    {
      "epoch": 0.055944055944055944,
      "grad_norm": 1.5871682167053223,
      "learning_rate": 6.026312439675552e-06,
      "loss": 5.1593,
      "step": 170
    },
    {
      "epoch": 0.0562731386260798,
      "grad_norm": 1.7101870775222778,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 5.4082,
      "step": 171
    },
    {
      "epoch": 0.05660222130810366,
      "grad_norm": 1.8228991031646729,
      "learning_rate": 5.263533508961827e-06,
      "loss": 5.757,
      "step": 172
    },
    {
      "epoch": 0.05693130399012752,
      "grad_norm": 1.7279367446899414,
      "learning_rate": 4.900438493352055e-06,
      "loss": 5.2391,
      "step": 173
    },
    {
      "epoch": 0.05726038667215138,
      "grad_norm": 1.9224296808242798,
      "learning_rate": 4.549673247541875e-06,
      "loss": 5.565,
      "step": 174
    },
    {
      "epoch": 0.057589469354175235,
      "grad_norm": 2.0328540802001953,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 5.6517,
      "step": 175
    },
    {
      "epoch": 0.0579185520361991,
      "grad_norm": 1.9473679065704346,
      "learning_rate": 3.885512251130763e-06,
      "loss": 5.5727,
      "step": 176
    },
    {
      "epoch": 0.05824763471822295,
      "grad_norm": 2.134977102279663,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 5.5729,
      "step": 177
    },
    {
      "epoch": 0.058576717400246815,
      "grad_norm": 2.233104705810547,
      "learning_rate": 3.271776770026963e-06,
      "loss": 6.1055,
      "step": 178
    },
    {
      "epoch": 0.05890580008227067,
      "grad_norm": 2.272578477859497,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 5.9918,
      "step": 179
    },
    {
      "epoch": 0.059234882764294526,
      "grad_norm": 2.3617653846740723,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 5.9966,
      "step": 180
    },
    {
      "epoch": 0.05956396544631839,
      "grad_norm": 2.2694661617279053,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 5.6238,
      "step": 181
    },
    {
      "epoch": 0.059893048128342244,
      "grad_norm": 2.2927069664001465,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 5.4318,
      "step": 182
    },
    {
      "epoch": 0.060222130810366106,
      "grad_norm": 2.756525754928589,
      "learning_rate": 1.962316193157593e-06,
      "loss": 6.205,
      "step": 183
    },
    {
      "epoch": 0.06055121349238996,
      "grad_norm": 2.776837110519409,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 6.0609,
      "step": 184
    },
    {
      "epoch": 0.060880296174413824,
      "grad_norm": 2.918240547180176,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 6.1536,
      "step": 185
    },
    {
      "epoch": 0.06120937885643768,
      "grad_norm": 2.784555196762085,
      "learning_rate": 1.333670137599713e-06,
      "loss": 5.8498,
      "step": 186
    },
    {
      "epoch": 0.06153846153846154,
      "grad_norm": 2.858046531677246,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 6.1287,
      "step": 187
    },
    {
      "epoch": 0.0618675442204854,
      "grad_norm": 2.8780221939086914,
      "learning_rate": 9.810017062595322e-07,
      "loss": 5.9005,
      "step": 188
    },
    {
      "epoch": 0.06219662690250925,
      "grad_norm": 3.3040332794189453,
      "learning_rate": 8.247462563808817e-07,
      "loss": 6.1179,
      "step": 189
    },
    {
      "epoch": 0.06252570958453312,
      "grad_norm": 3.0962657928466797,
      "learning_rate": 6.819348298638839e-07,
      "loss": 5.9738,
      "step": 190
    },
    {
      "epoch": 0.06285479226655698,
      "grad_norm": 3.6452810764312744,
      "learning_rate": 5.526064699265753e-07,
      "loss": 6.6203,
      "step": 191
    },
    {
      "epoch": 0.06318387494858083,
      "grad_norm": 3.975619077682495,
      "learning_rate": 4.367965336512403e-07,
      "loss": 6.7489,
      "step": 192
    },
    {
      "epoch": 0.06351295763060469,
      "grad_norm": 3.6615569591522217,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 6.276,
      "step": 193
    },
    {
      "epoch": 0.06384204031262855,
      "grad_norm": 4.099253177642822,
      "learning_rate": 2.458548727494292e-07,
      "loss": 6.3504,
      "step": 194
    },
    {
      "epoch": 0.06417112299465241,
      "grad_norm": 4.67241907119751,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 6.3966,
      "step": 195
    },
    {
      "epoch": 0.06450020567667626,
      "grad_norm": 4.869143962860107,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 7.1098,
      "step": 196
    },
    {
      "epoch": 0.06482928835870012,
      "grad_norm": 4.802524566650391,
      "learning_rate": 6.150154258476315e-08,
      "loss": 6.4907,
      "step": 197
    },
    {
      "epoch": 0.06515837104072399,
      "grad_norm": 6.322317123413086,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 8.4675,
      "step": 198
    },
    {
      "epoch": 0.06548745372274783,
      "grad_norm": 7.752302646636963,
      "learning_rate": 6.834750376549792e-09,
      "loss": 8.1981,
      "step": 199
    },
    {
      "epoch": 0.0658165364047717,
      "grad_norm": 10.216737747192383,
      "learning_rate": 0.0,
      "loss": 7.8471,
      "step": 200
    },
    {
      "epoch": 0.0658165364047717,
      "eval_loss": 1.3994430303573608,
      "eval_runtime": 108.6438,
      "eval_samples_per_second": 47.108,
      "eval_steps_per_second": 11.782,
      "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": 5.354419086006682e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}