{
  "best_metric": 4.293779373168945,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 1.6877637130801688,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008438818565400843,
      "grad_norm": 0.342307984828949,
      "learning_rate": 1e-05,
      "loss": 4.775,
      "step": 1
    },
    {
      "epoch": 0.008438818565400843,
      "eval_loss": 5.433469295501709,
      "eval_runtime": 3.3605,
      "eval_samples_per_second": 59.514,
      "eval_steps_per_second": 14.879,
      "step": 1
    },
    {
      "epoch": 0.016877637130801686,
      "grad_norm": 0.3679039180278778,
      "learning_rate": 2e-05,
      "loss": 4.7829,
      "step": 2
    },
    {
      "epoch": 0.02531645569620253,
      "grad_norm": 0.4522739052772522,
      "learning_rate": 3e-05,
      "loss": 4.8823,
      "step": 3
    },
    {
      "epoch": 0.03375527426160337,
      "grad_norm": 0.41035130620002747,
      "learning_rate": 4e-05,
      "loss": 4.9389,
      "step": 4
    },
    {
      "epoch": 0.04219409282700422,
      "grad_norm": 0.4490402638912201,
      "learning_rate": 5e-05,
      "loss": 4.9465,
      "step": 5
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 0.45817092061042786,
      "learning_rate": 6e-05,
      "loss": 5.0842,
      "step": 6
    },
    {
      "epoch": 0.05907172995780591,
      "grad_norm": 0.4189537763595581,
      "learning_rate": 7e-05,
      "loss": 4.9442,
      "step": 7
    },
    {
      "epoch": 0.06751054852320675,
      "grad_norm": 0.4117410480976105,
      "learning_rate": 8e-05,
      "loss": 4.9375,
      "step": 8
    },
    {
      "epoch": 0.0759493670886076,
      "grad_norm": 0.4152357280254364,
      "learning_rate": 9e-05,
      "loss": 4.9969,
      "step": 9
    },
    {
      "epoch": 0.08438818565400844,
      "grad_norm": 0.4271791875362396,
      "learning_rate": 0.0001,
      "loss": 5.0544,
      "step": 10
    },
    {
      "epoch": 0.09282700421940929,
      "grad_norm": 0.40930861234664917,
      "learning_rate": 9.999316524962345e-05,
      "loss": 5.0555,
      "step": 11
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.4359319806098938,
      "learning_rate": 9.997266286704631e-05,
      "loss": 5.0076,
      "step": 12
    },
    {
      "epoch": 0.10970464135021098,
      "grad_norm": 0.43278539180755615,
      "learning_rate": 9.993849845741524e-05,
      "loss": 4.9623,
      "step": 13
    },
    {
      "epoch": 0.11814345991561181,
      "grad_norm": 0.41361337900161743,
      "learning_rate": 9.989068136093873e-05,
      "loss": 4.9514,
      "step": 14
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 0.4340251088142395,
      "learning_rate": 9.98292246503335e-05,
      "loss": 4.9231,
      "step": 15
    },
    {
      "epoch": 0.1350210970464135,
      "grad_norm": 0.3976637125015259,
      "learning_rate": 9.975414512725057e-05,
      "loss": 5.0262,
      "step": 16
    },
    {
      "epoch": 0.14345991561181434,
      "grad_norm": 0.44714269042015076,
      "learning_rate": 9.966546331768191e-05,
      "loss": 4.8843,
      "step": 17
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 0.4278963506221771,
      "learning_rate": 9.956320346634876e-05,
      "loss": 4.9152,
      "step": 18
    },
    {
      "epoch": 0.16033755274261605,
      "grad_norm": 0.46977096796035767,
      "learning_rate": 9.944739353007344e-05,
      "loss": 5.1004,
      "step": 19
    },
    {
      "epoch": 0.16877637130801687,
      "grad_norm": 0.5219294428825378,
      "learning_rate": 9.931806517013612e-05,
      "loss": 5.0981,
      "step": 20
    },
    {
      "epoch": 0.17721518987341772,
      "grad_norm": 0.5068072080612183,
      "learning_rate": 9.917525374361912e-05,
      "loss": 5.0617,
      "step": 21
    },
    {
      "epoch": 0.18565400843881857,
      "grad_norm": 0.5007924437522888,
      "learning_rate": 9.901899829374047e-05,
      "loss": 4.9164,
      "step": 22
    },
    {
      "epoch": 0.1940928270042194,
      "grad_norm": 0.6181425452232361,
      "learning_rate": 9.884934153917997e-05,
      "loss": 4.9819,
      "step": 23
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.5986412763595581,
      "learning_rate": 9.86663298624003e-05,
      "loss": 5.0348,
      "step": 24
    },
    {
      "epoch": 0.2109704641350211,
      "grad_norm": 0.6753811836242676,
      "learning_rate": 9.847001329696653e-05,
      "loss": 5.2133,
      "step": 25
    },
    {
      "epoch": 0.21940928270042195,
      "grad_norm": 0.7484744787216187,
      "learning_rate": 9.826044551386744e-05,
      "loss": 5.1105,
      "step": 26
    },
    {
      "epoch": 0.22784810126582278,
      "grad_norm": 0.8159131407737732,
      "learning_rate": 9.803768380684242e-05,
      "loss": 5.1731,
      "step": 27
    },
    {
      "epoch": 0.23628691983122363,
      "grad_norm": 1.013075351715088,
      "learning_rate": 9.780178907671789e-05,
      "loss": 5.3069,
      "step": 28
    },
    {
      "epoch": 0.24472573839662448,
      "grad_norm": 1.4735418558120728,
      "learning_rate": 9.755282581475769e-05,
      "loss": 5.3314,
      "step": 29
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 1.3805432319641113,
      "learning_rate": 9.729086208503174e-05,
      "loss": 4.596,
      "step": 30
    },
    {
      "epoch": 0.2616033755274262,
      "grad_norm": 1.5402129888534546,
      "learning_rate": 9.701596950580806e-05,
      "loss": 4.7255,
      "step": 31
    },
    {
      "epoch": 0.270042194092827,
      "grad_norm": 1.283172845840454,
      "learning_rate": 9.672822322997305e-05,
      "loss": 4.6367,
      "step": 32
    },
    {
      "epoch": 0.27848101265822783,
      "grad_norm": 1.0515762567520142,
      "learning_rate": 9.642770192448536e-05,
      "loss": 4.7627,
      "step": 33
    },
    {
      "epoch": 0.2869198312236287,
      "grad_norm": 0.7783506512641907,
      "learning_rate": 9.611448774886924e-05,
      "loss": 4.6866,
      "step": 34
    },
    {
      "epoch": 0.29535864978902954,
      "grad_norm": 0.6623164415359497,
      "learning_rate": 9.578866633275288e-05,
      "loss": 4.5398,
      "step": 35
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.6396310925483704,
      "learning_rate": 9.545032675245813e-05,
      "loss": 4.5623,
      "step": 36
    },
    {
      "epoch": 0.31223628691983124,
      "grad_norm": 0.4664241373538971,
      "learning_rate": 9.509956150664796e-05,
      "loss": 4.6151,
      "step": 37
    },
    {
      "epoch": 0.3206751054852321,
      "grad_norm": 0.48507606983184814,
      "learning_rate": 9.473646649103818e-05,
      "loss": 4.5681,
      "step": 38
    },
    {
      "epoch": 0.3291139240506329,
      "grad_norm": 0.4673691689968109,
      "learning_rate": 9.43611409721806e-05,
      "loss": 4.7461,
      "step": 39
    },
    {
      "epoch": 0.33755274261603374,
      "grad_norm": 0.50004643201828,
      "learning_rate": 9.397368756032445e-05,
      "loss": 4.6914,
      "step": 40
    },
    {
      "epoch": 0.3459915611814346,
      "grad_norm": 0.49235042929649353,
      "learning_rate": 9.357421218136386e-05,
      "loss": 4.6613,
      "step": 41
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 0.4833002984523773,
      "learning_rate": 9.316282404787871e-05,
      "loss": 4.6838,
      "step": 42
    },
    {
      "epoch": 0.3628691983122363,
      "grad_norm": 0.4639793932437897,
      "learning_rate": 9.273963562927695e-05,
      "loss": 4.5547,
      "step": 43
    },
    {
      "epoch": 0.37130801687763715,
      "grad_norm": 0.4577101469039917,
      "learning_rate": 9.230476262104677e-05,
      "loss": 4.5793,
      "step": 44
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 0.5850247740745544,
      "learning_rate": 9.185832391312644e-05,
      "loss": 4.7737,
      "step": 45
    },
    {
      "epoch": 0.3881856540084388,
      "grad_norm": 0.5581880807876587,
      "learning_rate": 9.140044155740101e-05,
      "loss": 4.6576,
      "step": 46
    },
    {
      "epoch": 0.39662447257383965,
      "grad_norm": 0.5836822390556335,
      "learning_rate": 9.093124073433463e-05,
      "loss": 4.7283,
      "step": 47
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 0.5879956483840942,
      "learning_rate": 9.045084971874738e-05,
      "loss": 4.681,
      "step": 48
    },
    {
      "epoch": 0.41350210970464135,
      "grad_norm": 0.5623932480812073,
      "learning_rate": 8.995939984474624e-05,
      "loss": 4.6653,
      "step": 49
    },
    {
      "epoch": 0.4219409282700422,
      "grad_norm": 0.5917908549308777,
      "learning_rate": 8.945702546981969e-05,
      "loss": 4.6608,
      "step": 50
    },
    {
      "epoch": 0.4219409282700422,
      "eval_loss": 4.637380123138428,
      "eval_runtime": 3.3499,
      "eval_samples_per_second": 59.704,
      "eval_steps_per_second": 14.926,
      "step": 50
    },
    {
      "epoch": 0.43037974683544306,
      "grad_norm": 0.6299328207969666,
      "learning_rate": 8.894386393810563e-05,
      "loss": 4.6544,
      "step": 51
    },
    {
      "epoch": 0.4388185654008439,
      "grad_norm": 0.6917715668678284,
      "learning_rate": 8.842005554284296e-05,
      "loss": 4.6308,
      "step": 52
    },
    {
      "epoch": 0.4472573839662447,
      "grad_norm": 0.7258516550064087,
      "learning_rate": 8.788574348801675e-05,
      "loss": 4.7299,
      "step": 53
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 0.8476598262786865,
      "learning_rate": 8.73410738492077e-05,
      "loss": 4.7551,
      "step": 54
    },
    {
      "epoch": 0.4641350210970464,
      "grad_norm": 0.8885517120361328,
      "learning_rate": 8.678619553365659e-05,
      "loss": 4.6168,
      "step": 55
    },
    {
      "epoch": 0.47257383966244726,
      "grad_norm": 1.0589102506637573,
      "learning_rate": 8.622126023955446e-05,
      "loss": 4.8045,
      "step": 56
    },
    {
      "epoch": 0.4810126582278481,
      "grad_norm": 1.3015607595443726,
      "learning_rate": 8.564642241456986e-05,
      "loss": 5.1099,
      "step": 57
    },
    {
      "epoch": 0.48945147679324896,
      "grad_norm": 1.7195920944213867,
      "learning_rate": 8.506183921362443e-05,
      "loss": 5.1336,
      "step": 58
    },
    {
      "epoch": 0.4978902953586498,
      "grad_norm": 0.5516311526298523,
      "learning_rate": 8.44676704559283e-05,
      "loss": 4.3196,
      "step": 59
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.5670045018196106,
      "learning_rate": 8.386407858128706e-05,
      "loss": 4.4345,
      "step": 60
    },
    {
      "epoch": 0.5147679324894515,
      "grad_norm": 0.6434422135353088,
      "learning_rate": 8.32512286056924e-05,
      "loss": 4.3451,
      "step": 61
    },
    {
      "epoch": 0.5232067510548524,
      "grad_norm": 0.5841376185417175,
      "learning_rate": 8.262928807620843e-05,
      "loss": 4.2532,
      "step": 62
    },
    {
      "epoch": 0.5316455696202531,
      "grad_norm": 0.5705608129501343,
      "learning_rate": 8.199842702516583e-05,
      "loss": 4.2861,
      "step": 63
    },
    {
      "epoch": 0.540084388185654,
      "grad_norm": 0.5736099481582642,
      "learning_rate": 8.135881792367686e-05,
      "loss": 4.4719,
      "step": 64
    },
    {
      "epoch": 0.5485232067510548,
      "grad_norm": 0.6140097975730896,
      "learning_rate": 8.07106356344834e-05,
      "loss": 4.4136,
      "step": 65
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 0.6242948174476624,
      "learning_rate": 8.005405736415126e-05,
      "loss": 4.3542,
      "step": 66
    },
    {
      "epoch": 0.5654008438818565,
      "grad_norm": 0.65937340259552,
      "learning_rate": 7.938926261462366e-05,
      "loss": 4.3753,
      "step": 67
    },
    {
      "epoch": 0.5738396624472574,
      "grad_norm": 0.6030390858650208,
      "learning_rate": 7.871643313414718e-05,
      "loss": 4.4532,
      "step": 68
    },
    {
      "epoch": 0.5822784810126582,
      "grad_norm": 0.5688068270683289,
      "learning_rate": 7.803575286758364e-05,
      "loss": 4.4137,
      "step": 69
    },
    {
      "epoch": 0.5907172995780591,
      "grad_norm": 0.5468143820762634,
      "learning_rate": 7.734740790612136e-05,
      "loss": 4.4313,
      "step": 70
    },
    {
      "epoch": 0.5991561181434599,
      "grad_norm": 0.5390614867210388,
      "learning_rate": 7.66515864363997e-05,
      "loss": 4.4213,
      "step": 71
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.5791484713554382,
      "learning_rate": 7.594847868906076e-05,
      "loss": 4.4295,
      "step": 72
    },
    {
      "epoch": 0.6160337552742616,
      "grad_norm": 0.6058916449546814,
      "learning_rate": 7.52382768867422e-05,
      "loss": 4.4847,
      "step": 73
    },
    {
      "epoch": 0.6244725738396625,
      "grad_norm": 0.6058773398399353,
      "learning_rate": 7.452117519152542e-05,
      "loss": 4.4318,
      "step": 74
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.5961236953735352,
      "learning_rate": 7.379736965185368e-05,
      "loss": 4.4236,
      "step": 75
    },
    {
      "epoch": 0.6413502109704642,
      "grad_norm": 0.6671427488327026,
      "learning_rate": 7.30670581489344e-05,
      "loss": 4.5977,
      "step": 76
    },
    {
      "epoch": 0.6497890295358649,
      "grad_norm": 0.6705625653266907,
      "learning_rate": 7.233044034264034e-05,
      "loss": 4.4147,
      "step": 77
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 0.6936260461807251,
      "learning_rate": 7.158771761692464e-05,
      "loss": 4.5549,
      "step": 78
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.7389065623283386,
      "learning_rate": 7.083909302476453e-05,
      "loss": 4.5004,
      "step": 79
    },
    {
      "epoch": 0.6751054852320675,
      "grad_norm": 0.7763019800186157,
      "learning_rate": 7.008477123264848e-05,
      "loss": 4.4289,
      "step": 80
    },
    {
      "epoch": 0.6835443037974683,
      "grad_norm": 0.7912735939025879,
      "learning_rate": 6.932495846462261e-05,
      "loss": 4.4703,
      "step": 81
    },
    {
      "epoch": 0.6919831223628692,
      "grad_norm": 0.9331203699111938,
      "learning_rate": 6.855986244591104e-05,
      "loss": 4.4093,
      "step": 82
    },
    {
      "epoch": 0.70042194092827,
      "grad_norm": 1.0904549360275269,
      "learning_rate": 6.778969234612584e-05,
      "loss": 4.5354,
      "step": 83
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 1.3092857599258423,
      "learning_rate": 6.701465872208216e-05,
      "loss": 4.6985,
      "step": 84
    },
    {
      "epoch": 0.7172995780590717,
      "grad_norm": 1.1843961477279663,
      "learning_rate": 6.623497346023418e-05,
      "loss": 4.5391,
      "step": 85
    },
    {
      "epoch": 0.7257383966244726,
      "grad_norm": 1.491106629371643,
      "learning_rate": 6.545084971874738e-05,
      "loss": 4.7882,
      "step": 86
    },
    {
      "epoch": 0.7341772151898734,
      "grad_norm": 1.9464541673660278,
      "learning_rate": 6.466250186922325e-05,
      "loss": 5.1163,
      "step": 87
    },
    {
      "epoch": 0.7426160337552743,
      "grad_norm": 0.6096802949905396,
      "learning_rate": 6.387014543809223e-05,
      "loss": 4.2573,
      "step": 88
    },
    {
      "epoch": 0.7510548523206751,
      "grad_norm": 0.6752976775169373,
      "learning_rate": 6.307399704769099e-05,
      "loss": 4.2524,
      "step": 89
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.6118515729904175,
      "learning_rate": 6.227427435703997e-05,
      "loss": 4.3332,
      "step": 90
    },
    {
      "epoch": 0.7679324894514767,
      "grad_norm": 0.590178370475769,
      "learning_rate": 6.147119600233758e-05,
      "loss": 4.3085,
      "step": 91
    },
    {
      "epoch": 0.7763713080168776,
      "grad_norm": 0.6369143128395081,
      "learning_rate": 6.066498153718735e-05,
      "loss": 4.3197,
      "step": 92
    },
    {
      "epoch": 0.7848101265822784,
      "grad_norm": 0.5484967827796936,
      "learning_rate": 5.985585137257401e-05,
      "loss": 4.274,
      "step": 93
    },
    {
      "epoch": 0.7932489451476793,
      "grad_norm": 0.6025214791297913,
      "learning_rate": 5.90440267166055e-05,
      "loss": 4.2625,
      "step": 94
    },
    {
      "epoch": 0.8016877637130801,
      "grad_norm": 0.5956266522407532,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 4.247,
      "step": 95
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.6351256966590881,
      "learning_rate": 5.74131823855921e-05,
      "loss": 4.3553,
      "step": 96
    },
    {
      "epoch": 0.8185654008438819,
      "grad_norm": 0.5899494886398315,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 4.278,
      "step": 97
    },
    {
      "epoch": 0.8270042194092827,
      "grad_norm": 0.5957287549972534,
      "learning_rate": 5.577423184847932e-05,
      "loss": 4.1869,
      "step": 98
    },
    {
      "epoch": 0.8354430379746836,
      "grad_norm": 0.644184947013855,
      "learning_rate": 5.495227651252315e-05,
      "loss": 4.3236,
      "step": 99
    },
    {
      "epoch": 0.8438818565400844,
      "grad_norm": 0.6234635710716248,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 4.1895,
      "step": 100
    },
    {
      "epoch": 0.8438818565400844,
      "eval_loss": 4.398071765899658,
      "eval_runtime": 3.3484,
      "eval_samples_per_second": 59.73,
      "eval_steps_per_second": 14.933,
      "step": 100
    },
    {
      "epoch": 0.8523206751054853,
      "grad_norm": 0.6422309279441833,
      "learning_rate": 5.330452921628497e-05,
      "loss": 4.2532,
      "step": 101
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 0.7042955160140991,
      "learning_rate": 5.247918773366112e-05,
      "loss": 4.304,
      "step": 102
    },
    {
      "epoch": 0.869198312236287,
      "grad_norm": 0.7082456946372986,
      "learning_rate": 5.165316846586541e-05,
      "loss": 4.5844,
      "step": 103
    },
    {
      "epoch": 0.8776371308016878,
      "grad_norm": 0.6789894700050354,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 4.4241,
      "step": 104
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.7082401514053345,
      "learning_rate": 5e-05,
      "loss": 4.4103,
      "step": 105
    },
    {
      "epoch": 0.8945147679324894,
      "grad_norm": 0.7231240272521973,
      "learning_rate": 4.917330276168208e-05,
      "loss": 4.3267,
      "step": 106
    },
    {
      "epoch": 0.9029535864978903,
      "grad_norm": 0.7750891447067261,
      "learning_rate": 4.834683153413459e-05,
      "loss": 4.4647,
      "step": 107
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.852919340133667,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 4.3246,
      "step": 108
    },
    {
      "epoch": 0.919831223628692,
      "grad_norm": 0.872220516204834,
      "learning_rate": 4.669547078371504e-05,
      "loss": 4.3793,
      "step": 109
    },
    {
      "epoch": 0.9282700421940928,
      "grad_norm": 0.9388225078582764,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 4.4762,
      "step": 110
    },
    {
      "epoch": 0.9367088607594937,
      "grad_norm": 1.079044222831726,
      "learning_rate": 4.504772348747687e-05,
      "loss": 4.2097,
      "step": 111
    },
    {
      "epoch": 0.9451476793248945,
      "grad_norm": 1.186145544052124,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 4.5572,
      "step": 112
    },
    {
      "epoch": 0.9535864978902954,
      "grad_norm": 1.2257293462753296,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 4.4764,
      "step": 113
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 1.3871668577194214,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 4.7316,
      "step": 114
    },
    {
      "epoch": 0.9704641350210971,
      "grad_norm": 1.8235058784484863,
      "learning_rate": 4.17702704859633e-05,
      "loss": 4.7155,
      "step": 115
    },
    {
      "epoch": 0.9789029535864979,
      "grad_norm": 2.2903292179107666,
      "learning_rate": 4.095597328339452e-05,
      "loss": 4.9532,
      "step": 116
    },
    {
      "epoch": 0.9873417721518988,
      "grad_norm": 0.6517820358276367,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 4.1947,
      "step": 117
    },
    {
      "epoch": 0.9957805907172996,
      "grad_norm": 0.8299269080162048,
      "learning_rate": 3.933501846281267e-05,
      "loss": 4.325,
      "step": 118
    },
    {
      "epoch": 1.0042194092827004,
      "grad_norm": 2.059263229370117,
      "learning_rate": 3.852880399766243e-05,
      "loss": 6.7985,
      "step": 119
    },
    {
      "epoch": 1.0126582278481013,
      "grad_norm": 0.6434903740882874,
      "learning_rate": 3.772572564296005e-05,
      "loss": 3.9214,
      "step": 120
    },
    {
      "epoch": 1.021097046413502,
      "grad_norm": 0.6045034527778625,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 4.1132,
      "step": 121
    },
    {
      "epoch": 1.029535864978903,
      "grad_norm": 0.6056077480316162,
      "learning_rate": 3.612985456190778e-05,
      "loss": 4.0557,
      "step": 122
    },
    {
      "epoch": 1.0379746835443038,
      "grad_norm": 0.6314725279808044,
      "learning_rate": 3.533749813077677e-05,
      "loss": 4.1441,
      "step": 123
    },
    {
      "epoch": 1.0464135021097047,
      "grad_norm": 0.6316689252853394,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 4.2066,
      "step": 124
    },
    {
      "epoch": 1.0548523206751055,
      "grad_norm": 0.652711033821106,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 4.1347,
      "step": 125
    },
    {
      "epoch": 1.0632911392405062,
      "grad_norm": 0.6759218573570251,
      "learning_rate": 3.298534127791785e-05,
      "loss": 4.0667,
      "step": 126
    },
    {
      "epoch": 1.0717299578059072,
      "grad_norm": 0.6420354843139648,
      "learning_rate": 3.221030765387417e-05,
      "loss": 4.2425,
      "step": 127
    },
    {
      "epoch": 1.080168776371308,
      "grad_norm": 0.6476247310638428,
      "learning_rate": 3.144013755408895e-05,
      "loss": 4.1874,
      "step": 128
    },
    {
      "epoch": 1.0886075949367089,
      "grad_norm": 0.6662815809249878,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 4.1922,
      "step": 129
    },
    {
      "epoch": 1.0970464135021096,
      "grad_norm": 0.6933587789535522,
      "learning_rate": 2.991522876735154e-05,
      "loss": 4.3272,
      "step": 130
    },
    {
      "epoch": 1.1054852320675106,
      "grad_norm": 0.7019554972648621,
      "learning_rate": 2.916090697523549e-05,
      "loss": 4.0595,
      "step": 131
    },
    {
      "epoch": 1.1139240506329113,
      "grad_norm": 0.7138832807540894,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 4.1291,
      "step": 132
    },
    {
      "epoch": 1.1223628691983123,
      "grad_norm": 0.7018268704414368,
      "learning_rate": 2.766955965735968e-05,
      "loss": 4.2259,
      "step": 133
    },
    {
      "epoch": 1.130801687763713,
      "grad_norm": 0.6958891153335571,
      "learning_rate": 2.693294185106562e-05,
      "loss": 4.2543,
      "step": 134
    },
    {
      "epoch": 1.139240506329114,
      "grad_norm": 0.7048686742782593,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 4.2419,
      "step": 135
    },
    {
      "epoch": 1.1476793248945147,
      "grad_norm": 0.778679370880127,
      "learning_rate": 2.547882480847461e-05,
      "loss": 4.2214,
      "step": 136
    },
    {
      "epoch": 1.1561181434599157,
      "grad_norm": 0.8573774099349976,
      "learning_rate": 2.476172311325783e-05,
      "loss": 4.2996,
      "step": 137
    },
    {
      "epoch": 1.1645569620253164,
      "grad_norm": 0.8210501670837402,
      "learning_rate": 2.405152131093926e-05,
      "loss": 4.3123,
      "step": 138
    },
    {
      "epoch": 1.1729957805907172,
      "grad_norm": 0.8696852326393127,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 4.1909,
      "step": 139
    },
    {
      "epoch": 1.1814345991561181,
      "grad_norm": 0.9292694926261902,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 4.2298,
      "step": 140
    },
    {
      "epoch": 1.189873417721519,
      "grad_norm": 0.8983340859413147,
      "learning_rate": 2.196424713241637e-05,
      "loss": 4.3604,
      "step": 141
    },
    {
      "epoch": 1.1983122362869199,
      "grad_norm": 0.9874733090400696,
      "learning_rate": 2.128356686585282e-05,
      "loss": 4.4573,
      "step": 142
    },
    {
      "epoch": 1.2067510548523206,
      "grad_norm": 1.1241238117218018,
      "learning_rate": 2.061073738537635e-05,
      "loss": 4.31,
      "step": 143
    },
    {
      "epoch": 1.2151898734177216,
      "grad_norm": 1.227468490600586,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 4.1052,
      "step": 144
    },
    {
      "epoch": 1.2236286919831223,
      "grad_norm": 1.3430508375167847,
      "learning_rate": 1.928936436551661e-05,
      "loss": 4.4813,
      "step": 145
    },
    {
      "epoch": 1.2320675105485233,
      "grad_norm": 1.5539218187332153,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 4.6364,
      "step": 146
    },
    {
      "epoch": 1.240506329113924,
      "grad_norm": 2.201293468475342,
      "learning_rate": 1.800157297483417e-05,
      "loss": 4.7293,
      "step": 147
    },
    {
      "epoch": 1.248945147679325,
      "grad_norm": 1.476494550704956,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 4.3854,
      "step": 148
    },
    {
      "epoch": 1.2573839662447257,
      "grad_norm": 0.5694752931594849,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 3.797,
      "step": 149
    },
    {
      "epoch": 1.2658227848101267,
      "grad_norm": 0.6153607368469238,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 3.951,
      "step": 150
    },
    {
      "epoch": 1.2658227848101267,
      "eval_loss": 4.3062286376953125,
      "eval_runtime": 3.3466,
      "eval_samples_per_second": 59.761,
      "eval_steps_per_second": 14.94,
      "step": 150
    },
    {
      "epoch": 1.2742616033755274,
      "grad_norm": 0.7403513193130493,
      "learning_rate": 1.553232954407171e-05,
      "loss": 4.1448,
      "step": 151
    },
    {
      "epoch": 1.2827004219409281,
      "grad_norm": 0.6847442984580994,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 4.0855,
      "step": 152
    },
    {
      "epoch": 1.2911392405063291,
      "grad_norm": 0.6602143049240112,
      "learning_rate": 1.435357758543015e-05,
      "loss": 4.1842,
      "step": 153
    },
    {
      "epoch": 1.29957805907173,
      "grad_norm": 0.6228960156440735,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 3.9985,
      "step": 154
    },
    {
      "epoch": 1.3080168776371308,
      "grad_norm": 0.641762375831604,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 4.0142,
      "step": 155
    },
    {
      "epoch": 1.3164556962025316,
      "grad_norm": 0.6571816802024841,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 4.0916,
      "step": 156
    },
    {
      "epoch": 1.3248945147679325,
      "grad_norm": 0.6504724621772766,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 4.2335,
      "step": 157
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.6643102765083313,
      "learning_rate": 1.157994445715706e-05,
      "loss": 4.146,
      "step": 158
    },
    {
      "epoch": 1.3417721518987342,
      "grad_norm": 0.6883799433708191,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 4.1033,
      "step": 159
    },
    {
      "epoch": 1.350210970464135,
      "grad_norm": 0.6689712405204773,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 4.2054,
      "step": 160
    },
    {
      "epoch": 1.358649789029536,
      "grad_norm": 0.7081148028373718,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 4.131,
      "step": 161
    },
    {
      "epoch": 1.3670886075949367,
      "grad_norm": 0.7011632323265076,
      "learning_rate": 9.549150281252633e-06,
      "loss": 4.2224,
      "step": 162
    },
    {
      "epoch": 1.3755274261603376,
      "grad_norm": 0.6964609622955322,
      "learning_rate": 9.068759265665384e-06,
      "loss": 4.3058,
      "step": 163
    },
    {
      "epoch": 1.3839662447257384,
      "grad_norm": 0.7627602815628052,
      "learning_rate": 8.599558442598998e-06,
      "loss": 4.2075,
      "step": 164
    },
    {
      "epoch": 1.3924050632911391,
      "grad_norm": 0.7444495558738708,
      "learning_rate": 8.141676086873572e-06,
      "loss": 4.3161,
      "step": 165
    },
    {
      "epoch": 1.40084388185654,
      "grad_norm": 0.7474559545516968,
      "learning_rate": 7.695237378953223e-06,
      "loss": 4.222,
      "step": 166
    },
    {
      "epoch": 1.409282700421941,
      "grad_norm": 0.9042083621025085,
      "learning_rate": 7.260364370723044e-06,
      "loss": 4.2578,
      "step": 167
    },
    {
      "epoch": 1.4177215189873418,
      "grad_norm": 0.8884591460227966,
      "learning_rate": 6.837175952121306e-06,
      "loss": 4.3039,
      "step": 168
    },
    {
      "epoch": 1.4261603375527425,
      "grad_norm": 0.8870271444320679,
      "learning_rate": 6.425787818636131e-06,
      "loss": 4.2751,
      "step": 169
    },
    {
      "epoch": 1.4345991561181435,
      "grad_norm": 1.0532454252243042,
      "learning_rate": 6.026312439675552e-06,
      "loss": 4.1375,
      "step": 170
    },
    {
      "epoch": 1.4430379746835442,
      "grad_norm": 0.9839035868644714,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 4.212,
      "step": 171
    },
    {
      "epoch": 1.4514767932489452,
      "grad_norm": 1.0723378658294678,
      "learning_rate": 5.263533508961827e-06,
      "loss": 4.3268,
      "step": 172
    },
    {
      "epoch": 1.459915611814346,
      "grad_norm": 1.323378324508667,
      "learning_rate": 4.900438493352055e-06,
      "loss": 4.3946,
      "step": 173
    },
    {
      "epoch": 1.4683544303797469,
      "grad_norm": 1.3638732433319092,
      "learning_rate": 4.549673247541875e-06,
      "loss": 4.3259,
      "step": 174
    },
    {
      "epoch": 1.4767932489451476,
      "grad_norm": 1.5440855026245117,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 4.5607,
      "step": 175
    },
    {
      "epoch": 1.4852320675105486,
      "grad_norm": 2.070108413696289,
      "learning_rate": 3.885512251130763e-06,
      "loss": 5.1395,
      "step": 176
    },
    {
      "epoch": 1.4936708860759493,
      "grad_norm": 1.3233076333999634,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 3.9364,
      "step": 177
    },
    {
      "epoch": 1.50210970464135,
      "grad_norm": 0.48516422510147095,
      "learning_rate": 3.271776770026963e-06,
      "loss": 3.8936,
      "step": 178
    },
    {
      "epoch": 1.510548523206751,
      "grad_norm": 0.5165314078330994,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 4.0922,
      "step": 179
    },
    {
      "epoch": 1.518987341772152,
      "grad_norm": 0.5385967493057251,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 4.0656,
      "step": 180
    },
    {
      "epoch": 1.5274261603375527,
      "grad_norm": 0.5883888602256775,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 4.0643,
      "step": 181
    },
    {
      "epoch": 1.5358649789029535,
      "grad_norm": 0.5779720544815063,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 4.0979,
      "step": 182
    },
    {
      "epoch": 1.5443037974683544,
      "grad_norm": 0.5666366219520569,
      "learning_rate": 1.962316193157593e-06,
      "loss": 4.1866,
      "step": 183
    },
    {
      "epoch": 1.5527426160337554,
      "grad_norm": 0.5771864056587219,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 4.253,
      "step": 184
    },
    {
      "epoch": 1.5611814345991561,
      "grad_norm": 0.6178004145622253,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 4.1141,
      "step": 185
    },
    {
      "epoch": 1.5696202531645569,
      "grad_norm": 0.6329324841499329,
      "learning_rate": 1.333670137599713e-06,
      "loss": 4.153,
      "step": 186
    },
    {
      "epoch": 1.5780590717299579,
      "grad_norm": 0.6661432385444641,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 4.0629,
      "step": 187
    },
    {
      "epoch": 1.5864978902953588,
      "grad_norm": 0.6318032145500183,
      "learning_rate": 9.810017062595322e-07,
      "loss": 4.189,
      "step": 188
    },
    {
      "epoch": 1.5949367088607596,
      "grad_norm": 0.6761325001716614,
      "learning_rate": 8.247462563808817e-07,
      "loss": 4.1711,
      "step": 189
    },
    {
      "epoch": 1.6033755274261603,
      "grad_norm": 0.6816501617431641,
      "learning_rate": 6.819348298638839e-07,
      "loss": 4.1101,
      "step": 190
    },
    {
      "epoch": 1.611814345991561,
      "grad_norm": 0.7030408382415771,
      "learning_rate": 5.526064699265753e-07,
      "loss": 4.2276,
      "step": 191
    },
    {
      "epoch": 1.620253164556962,
      "grad_norm": 0.7384068369865417,
      "learning_rate": 4.367965336512403e-07,
      "loss": 4.2872,
      "step": 192
    },
    {
      "epoch": 1.628691983122363,
      "grad_norm": 0.7516822218894958,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 4.175,
      "step": 193
    },
    {
      "epoch": 1.6371308016877637,
      "grad_norm": 0.7844093441963196,
      "learning_rate": 2.458548727494292e-07,
      "loss": 4.3874,
      "step": 194
    },
    {
      "epoch": 1.6455696202531644,
      "grad_norm": 0.7780900597572327,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 4.2789,
      "step": 195
    },
    {
      "epoch": 1.6540084388185654,
      "grad_norm": 0.8451712131500244,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 4.2738,
      "step": 196
    },
    {
      "epoch": 1.6624472573839664,
      "grad_norm": 0.8849952220916748,
      "learning_rate": 6.150154258476315e-08,
      "loss": 4.2407,
      "step": 197
    },
    {
      "epoch": 1.6708860759493671,
      "grad_norm": 0.888460099697113,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 4.2191,
      "step": 198
    },
    {
      "epoch": 1.6793248945147679,
      "grad_norm": 1.0812249183654785,
      "learning_rate": 6.834750376549792e-09,
      "loss": 4.1497,
      "step": 199
    },
    {
      "epoch": 1.6877637130801688,
      "grad_norm": 1.0116652250289917,
      "learning_rate": 0.0,
      "loss": 4.2673,
      "step": 200
    },
    {
      "epoch": 1.6877637130801688,
      "eval_loss": 4.293779373168945,
      "eval_runtime": 3.3529,
      "eval_samples_per_second": 59.65,
      "eval_steps_per_second": 14.913,
      "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": 1.372917298102272e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}