{
  "best_metric": 0.7629001140594482,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.16594067620825556,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008297033810412777,
      "grad_norm": 7.756772518157959,
      "learning_rate": 1e-05,
      "loss": 2.5111,
      "step": 1
    },
    {
      "epoch": 0.0008297033810412777,
      "eval_loss": 1.274738073348999,
      "eval_runtime": 26.0306,
      "eval_samples_per_second": 77.985,
      "eval_steps_per_second": 19.515,
      "step": 1
    },
    {
      "epoch": 0.0016594067620825555,
      "grad_norm": 6.429771900177002,
      "learning_rate": 2e-05,
      "loss": 3.5965,
      "step": 2
    },
    {
      "epoch": 0.002489110143123833,
      "grad_norm": 6.49445104598999,
      "learning_rate": 3e-05,
      "loss": 2.8547,
      "step": 3
    },
    {
      "epoch": 0.003318813524165111,
      "grad_norm": 5.629607200622559,
      "learning_rate": 4e-05,
      "loss": 2.9417,
      "step": 4
    },
    {
      "epoch": 0.004148516905206389,
      "grad_norm": 5.162261486053467,
      "learning_rate": 5e-05,
      "loss": 3.1428,
      "step": 5
    },
    {
      "epoch": 0.004978220286247666,
      "grad_norm": 5.2287492752075195,
      "learning_rate": 6e-05,
      "loss": 3.3361,
      "step": 6
    },
    {
      "epoch": 0.005807923667288944,
      "grad_norm": 6.387019634246826,
      "learning_rate": 7e-05,
      "loss": 3.4478,
      "step": 7
    },
    {
      "epoch": 0.006637627048330222,
      "grad_norm": 4.35686731338501,
      "learning_rate": 8e-05,
      "loss": 3.4745,
      "step": 8
    },
    {
      "epoch": 0.0074673304293715,
      "grad_norm": 4.3740153312683105,
      "learning_rate": 9e-05,
      "loss": 3.7762,
      "step": 9
    },
    {
      "epoch": 0.008297033810412777,
      "grad_norm": 4.73705530166626,
      "learning_rate": 0.0001,
      "loss": 3.5713,
      "step": 10
    },
    {
      "epoch": 0.009126737191454055,
      "grad_norm": 5.937510967254639,
      "learning_rate": 9.999316524962345e-05,
      "loss": 3.1444,
      "step": 11
    },
    {
      "epoch": 0.009956440572495333,
      "grad_norm": 3.565078020095825,
      "learning_rate": 9.997266286704631e-05,
      "loss": 4.0628,
      "step": 12
    },
    {
      "epoch": 0.01078614395353661,
      "grad_norm": 3.097233772277832,
      "learning_rate": 9.993849845741524e-05,
      "loss": 3.4775,
      "step": 13
    },
    {
      "epoch": 0.011615847334577888,
      "grad_norm": 3.4153213500976562,
      "learning_rate": 9.989068136093873e-05,
      "loss": 4.1153,
      "step": 14
    },
    {
      "epoch": 0.012445550715619166,
      "grad_norm": 3.2986154556274414,
      "learning_rate": 9.98292246503335e-05,
      "loss": 3.655,
      "step": 15
    },
    {
      "epoch": 0.013275254096660444,
      "grad_norm": 3.2382259368896484,
      "learning_rate": 9.975414512725057e-05,
      "loss": 3.8608,
      "step": 16
    },
    {
      "epoch": 0.014104957477701721,
      "grad_norm": 2.819169759750366,
      "learning_rate": 9.966546331768191e-05,
      "loss": 3.8011,
      "step": 17
    },
    {
      "epoch": 0.014934660858743,
      "grad_norm": 3.1195507049560547,
      "learning_rate": 9.956320346634876e-05,
      "loss": 3.8963,
      "step": 18
    },
    {
      "epoch": 0.01576436423978428,
      "grad_norm": 2.753371238708496,
      "learning_rate": 9.944739353007344e-05,
      "loss": 3.9488,
      "step": 19
    },
    {
      "epoch": 0.016594067620825555,
      "grad_norm": 2.8049979209899902,
      "learning_rate": 9.931806517013612e-05,
      "loss": 4.1054,
      "step": 20
    },
    {
      "epoch": 0.017423771001866834,
      "grad_norm": 3.020453929901123,
      "learning_rate": 9.917525374361912e-05,
      "loss": 4.68,
      "step": 21
    },
    {
      "epoch": 0.01825347438290811,
      "grad_norm": 2.9152514934539795,
      "learning_rate": 9.901899829374047e-05,
      "loss": 3.933,
      "step": 22
    },
    {
      "epoch": 0.01908317776394939,
      "grad_norm": 2.737259864807129,
      "learning_rate": 9.884934153917997e-05,
      "loss": 4.1059,
      "step": 23
    },
    {
      "epoch": 0.019912881144990666,
      "grad_norm": 2.7363414764404297,
      "learning_rate": 9.86663298624003e-05,
      "loss": 4.0902,
      "step": 24
    },
    {
      "epoch": 0.020742584526031945,
      "grad_norm": 2.649906873703003,
      "learning_rate": 9.847001329696653e-05,
      "loss": 3.8191,
      "step": 25
    },
    {
      "epoch": 0.02157228790707322,
      "grad_norm": 3.1954572200775146,
      "learning_rate": 9.826044551386744e-05,
      "loss": 4.4258,
      "step": 26
    },
    {
      "epoch": 0.0224019912881145,
      "grad_norm": 3.30267596244812,
      "learning_rate": 9.803768380684242e-05,
      "loss": 3.9039,
      "step": 27
    },
    {
      "epoch": 0.023231694669155777,
      "grad_norm": 3.4609975814819336,
      "learning_rate": 9.780178907671789e-05,
      "loss": 4.2572,
      "step": 28
    },
    {
      "epoch": 0.024061398050197056,
      "grad_norm": 3.5859968662261963,
      "learning_rate": 9.755282581475769e-05,
      "loss": 3.9162,
      "step": 29
    },
    {
      "epoch": 0.024891101431238332,
      "grad_norm": 3.430258274078369,
      "learning_rate": 9.729086208503174e-05,
      "loss": 3.6712,
      "step": 30
    },
    {
      "epoch": 0.02572080481227961,
      "grad_norm": 3.080860137939453,
      "learning_rate": 9.701596950580806e-05,
      "loss": 3.415,
      "step": 31
    },
    {
      "epoch": 0.026550508193320888,
      "grad_norm": 3.2382426261901855,
      "learning_rate": 9.672822322997305e-05,
      "loss": 3.4176,
      "step": 32
    },
    {
      "epoch": 0.027380211574362167,
      "grad_norm": 4.229242324829102,
      "learning_rate": 9.642770192448536e-05,
      "loss": 4.5244,
      "step": 33
    },
    {
      "epoch": 0.028209914955403443,
      "grad_norm": 3.9501850605010986,
      "learning_rate": 9.611448774886924e-05,
      "loss": 4.5879,
      "step": 34
    },
    {
      "epoch": 0.029039618336444722,
      "grad_norm": 3.9697861671447754,
      "learning_rate": 9.578866633275288e-05,
      "loss": 4.3162,
      "step": 35
    },
    {
      "epoch": 0.029869321717486,
      "grad_norm": 3.828258991241455,
      "learning_rate": 9.545032675245813e-05,
      "loss": 4.048,
      "step": 36
    },
    {
      "epoch": 0.030699025098527278,
      "grad_norm": 4.1518659591674805,
      "learning_rate": 9.509956150664796e-05,
      "loss": 4.7054,
      "step": 37
    },
    {
      "epoch": 0.03152872847956856,
      "grad_norm": 4.125431537628174,
      "learning_rate": 9.473646649103818e-05,
      "loss": 4.0045,
      "step": 38
    },
    {
      "epoch": 0.03235843186060983,
      "grad_norm": 4.2039103507995605,
      "learning_rate": 9.43611409721806e-05,
      "loss": 3.2643,
      "step": 39
    },
    {
      "epoch": 0.03318813524165111,
      "grad_norm": 3.8554229736328125,
      "learning_rate": 9.397368756032445e-05,
      "loss": 3.591,
      "step": 40
    },
    {
      "epoch": 0.034017838622692385,
      "grad_norm": 4.973513126373291,
      "learning_rate": 9.357421218136386e-05,
      "loss": 4.1479,
      "step": 41
    },
    {
      "epoch": 0.03484754200373367,
      "grad_norm": 4.953542232513428,
      "learning_rate": 9.316282404787871e-05,
      "loss": 4.1728,
      "step": 42
    },
    {
      "epoch": 0.035677245384774944,
      "grad_norm": 5.813049793243408,
      "learning_rate": 9.273963562927695e-05,
      "loss": 4.5657,
      "step": 43
    },
    {
      "epoch": 0.03650694876581622,
      "grad_norm": 5.689128398895264,
      "learning_rate": 9.230476262104677e-05,
      "loss": 3.6231,
      "step": 44
    },
    {
      "epoch": 0.037336652146857496,
      "grad_norm": 5.57553243637085,
      "learning_rate": 9.185832391312644e-05,
      "loss": 4.145,
      "step": 45
    },
    {
      "epoch": 0.03816635552789878,
      "grad_norm": 7.020411491394043,
      "learning_rate": 9.140044155740101e-05,
      "loss": 5.4164,
      "step": 46
    },
    {
      "epoch": 0.038996058908940055,
      "grad_norm": 7.255471706390381,
      "learning_rate": 9.093124073433463e-05,
      "loss": 5.152,
      "step": 47
    },
    {
      "epoch": 0.03982576228998133,
      "grad_norm": 7.882279872894287,
      "learning_rate": 9.045084971874738e-05,
      "loss": 5.3973,
      "step": 48
    },
    {
      "epoch": 0.04065546567102261,
      "grad_norm": 9.58462142944336,
      "learning_rate": 8.995939984474624e-05,
      "loss": 4.9437,
      "step": 49
    },
    {
      "epoch": 0.04148516905206389,
      "grad_norm": 13.189021110534668,
      "learning_rate": 8.945702546981969e-05,
      "loss": 5.5762,
      "step": 50
    },
    {
      "epoch": 0.04148516905206389,
      "eval_loss": 1.1816635131835938,
      "eval_runtime": 25.9384,
      "eval_samples_per_second": 78.262,
      "eval_steps_per_second": 19.585,
      "step": 50
    },
    {
      "epoch": 0.042314872433105166,
      "grad_norm": 22.59925651550293,
      "learning_rate": 8.894386393810563e-05,
      "loss": 3.9388,
      "step": 51
    },
    {
      "epoch": 0.04314457581414644,
      "grad_norm": 25.521390914916992,
      "learning_rate": 8.842005554284296e-05,
      "loss": 4.2848,
      "step": 52
    },
    {
      "epoch": 0.04397427919518772,
      "grad_norm": 17.39426612854004,
      "learning_rate": 8.788574348801675e-05,
      "loss": 4.1229,
      "step": 53
    },
    {
      "epoch": 0.044803982576229,
      "grad_norm": 9.67247486114502,
      "learning_rate": 8.73410738492077e-05,
      "loss": 3.9802,
      "step": 54
    },
    {
      "epoch": 0.04563368595727028,
      "grad_norm": 5.470470905303955,
      "learning_rate": 8.678619553365659e-05,
      "loss": 3.3911,
      "step": 55
    },
    {
      "epoch": 0.04646338933831155,
      "grad_norm": 4.268524169921875,
      "learning_rate": 8.622126023955446e-05,
      "loss": 3.5475,
      "step": 56
    },
    {
      "epoch": 0.04729309271935283,
      "grad_norm": 3.6320784091949463,
      "learning_rate": 8.564642241456986e-05,
      "loss": 2.74,
      "step": 57
    },
    {
      "epoch": 0.04812279610039411,
      "grad_norm": 4.06900691986084,
      "learning_rate": 8.506183921362443e-05,
      "loss": 3.2958,
      "step": 58
    },
    {
      "epoch": 0.04895249948143539,
      "grad_norm": 4.261375904083252,
      "learning_rate": 8.44676704559283e-05,
      "loss": 3.2473,
      "step": 59
    },
    {
      "epoch": 0.049782202862476664,
      "grad_norm": 3.505117177963257,
      "learning_rate": 8.386407858128706e-05,
      "loss": 3.3424,
      "step": 60
    },
    {
      "epoch": 0.05061190624351794,
      "grad_norm": 2.6978774070739746,
      "learning_rate": 8.32512286056924e-05,
      "loss": 2.6566,
      "step": 61
    },
    {
      "epoch": 0.05144160962455922,
      "grad_norm": 2.9829816818237305,
      "learning_rate": 8.262928807620843e-05,
      "loss": 3.7595,
      "step": 62
    },
    {
      "epoch": 0.0522713130056005,
      "grad_norm": 3.109691619873047,
      "learning_rate": 8.199842702516583e-05,
      "loss": 3.3401,
      "step": 63
    },
    {
      "epoch": 0.053101016386641775,
      "grad_norm": 3.12528657913208,
      "learning_rate": 8.135881792367686e-05,
      "loss": 3.2267,
      "step": 64
    },
    {
      "epoch": 0.05393071976768305,
      "grad_norm": 2.9566073417663574,
      "learning_rate": 8.07106356344834e-05,
      "loss": 3.0207,
      "step": 65
    },
    {
      "epoch": 0.054760423148724334,
      "grad_norm": 2.640789747238159,
      "learning_rate": 8.005405736415126e-05,
      "loss": 2.7983,
      "step": 66
    },
    {
      "epoch": 0.05559012652976561,
      "grad_norm": 2.4670517444610596,
      "learning_rate": 7.938926261462366e-05,
      "loss": 2.9725,
      "step": 67
    },
    {
      "epoch": 0.056419829910806886,
      "grad_norm": 2.2544631958007812,
      "learning_rate": 7.871643313414718e-05,
      "loss": 2.8911,
      "step": 68
    },
    {
      "epoch": 0.05724953329184816,
      "grad_norm": 3.448986291885376,
      "learning_rate": 7.803575286758364e-05,
      "loss": 4.0082,
      "step": 69
    },
    {
      "epoch": 0.058079236672889445,
      "grad_norm": 2.7839677333831787,
      "learning_rate": 7.734740790612136e-05,
      "loss": 3.1216,
      "step": 70
    },
    {
      "epoch": 0.05890894005393072,
      "grad_norm": 3.0205578804016113,
      "learning_rate": 7.66515864363997e-05,
      "loss": 4.4662,
      "step": 71
    },
    {
      "epoch": 0.059738643434972,
      "grad_norm": 2.8805060386657715,
      "learning_rate": 7.594847868906076e-05,
      "loss": 3.2907,
      "step": 72
    },
    {
      "epoch": 0.06056834681601327,
      "grad_norm": 2.6322007179260254,
      "learning_rate": 7.52382768867422e-05,
      "loss": 2.8668,
      "step": 73
    },
    {
      "epoch": 0.061398050197054556,
      "grad_norm": 2.8242907524108887,
      "learning_rate": 7.452117519152542e-05,
      "loss": 3.985,
      "step": 74
    },
    {
      "epoch": 0.06222775357809583,
      "grad_norm": 2.9598379135131836,
      "learning_rate": 7.379736965185368e-05,
      "loss": 3.9807,
      "step": 75
    },
    {
      "epoch": 0.06305745695913711,
      "grad_norm": 2.744391918182373,
      "learning_rate": 7.30670581489344e-05,
      "loss": 3.0852,
      "step": 76
    },
    {
      "epoch": 0.06388716034017838,
      "grad_norm": 3.015636920928955,
      "learning_rate": 7.233044034264034e-05,
      "loss": 3.4085,
      "step": 77
    },
    {
      "epoch": 0.06471686372121967,
      "grad_norm": 3.2922041416168213,
      "learning_rate": 7.158771761692464e-05,
      "loss": 3.4364,
      "step": 78
    },
    {
      "epoch": 0.06554656710226094,
      "grad_norm": 3.2776167392730713,
      "learning_rate": 7.083909302476453e-05,
      "loss": 4.1095,
      "step": 79
    },
    {
      "epoch": 0.06637627048330222,
      "grad_norm": 3.057568311691284,
      "learning_rate": 7.008477123264848e-05,
      "loss": 3.5205,
      "step": 80
    },
    {
      "epoch": 0.0672059738643435,
      "grad_norm": 3.358219623565674,
      "learning_rate": 6.932495846462261e-05,
      "loss": 3.4159,
      "step": 81
    },
    {
      "epoch": 0.06803567724538477,
      "grad_norm": 3.0704445838928223,
      "learning_rate": 6.855986244591104e-05,
      "loss": 3.1004,
      "step": 82
    },
    {
      "epoch": 0.06886538062642605,
      "grad_norm": 3.2741987705230713,
      "learning_rate": 6.778969234612584e-05,
      "loss": 3.449,
      "step": 83
    },
    {
      "epoch": 0.06969508400746734,
      "grad_norm": 3.408635377883911,
      "learning_rate": 6.701465872208216e-05,
      "loss": 3.7203,
      "step": 84
    },
    {
      "epoch": 0.0705247873885086,
      "grad_norm": 3.7510604858398438,
      "learning_rate": 6.623497346023418e-05,
      "loss": 3.9444,
      "step": 85
    },
    {
      "epoch": 0.07135449076954989,
      "grad_norm": 3.24825382232666,
      "learning_rate": 6.545084971874738e-05,
      "loss": 3.0609,
      "step": 86
    },
    {
      "epoch": 0.07218419415059116,
      "grad_norm": 3.389300584793091,
      "learning_rate": 6.466250186922325e-05,
      "loss": 3.411,
      "step": 87
    },
    {
      "epoch": 0.07301389753163244,
      "grad_norm": 3.6074140071868896,
      "learning_rate": 6.387014543809223e-05,
      "loss": 3.5501,
      "step": 88
    },
    {
      "epoch": 0.07384360091267372,
      "grad_norm": 3.6489806175231934,
      "learning_rate": 6.307399704769099e-05,
      "loss": 3.587,
      "step": 89
    },
    {
      "epoch": 0.07467330429371499,
      "grad_norm": 3.770354747772217,
      "learning_rate": 6.227427435703997e-05,
      "loss": 3.6326,
      "step": 90
    },
    {
      "epoch": 0.07550300767475628,
      "grad_norm": 3.604491710662842,
      "learning_rate": 6.147119600233758e-05,
      "loss": 3.3816,
      "step": 91
    },
    {
      "epoch": 0.07633271105579756,
      "grad_norm": 4.239180088043213,
      "learning_rate": 6.066498153718735e-05,
      "loss": 3.7999,
      "step": 92
    },
    {
      "epoch": 0.07716241443683883,
      "grad_norm": 4.287597179412842,
      "learning_rate": 5.985585137257401e-05,
      "loss": 3.8423,
      "step": 93
    },
    {
      "epoch": 0.07799211781788011,
      "grad_norm": 4.97831392288208,
      "learning_rate": 5.90440267166055e-05,
      "loss": 3.4461,
      "step": 94
    },
    {
      "epoch": 0.07882182119892138,
      "grad_norm": 5.106680393218994,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 3.8168,
      "step": 95
    },
    {
      "epoch": 0.07965152457996266,
      "grad_norm": 5.679533004760742,
      "learning_rate": 5.74131823855921e-05,
      "loss": 4.3349,
      "step": 96
    },
    {
      "epoch": 0.08048122796100395,
      "grad_norm": 6.412324905395508,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 4.6441,
      "step": 97
    },
    {
      "epoch": 0.08131093134204521,
      "grad_norm": 6.5507426261901855,
      "learning_rate": 5.577423184847932e-05,
      "loss": 4.3083,
      "step": 98
    },
    {
      "epoch": 0.0821406347230865,
      "grad_norm": 7.7271037101745605,
      "learning_rate": 5.495227651252315e-05,
      "loss": 4.5833,
      "step": 99
    },
    {
      "epoch": 0.08297033810412778,
      "grad_norm": 11.994407653808594,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 5.8545,
      "step": 100
    },
    {
      "epoch": 0.08297033810412778,
      "eval_loss": 0.8416540026664734,
      "eval_runtime": 26.046,
      "eval_samples_per_second": 77.939,
      "eval_steps_per_second": 19.504,
      "step": 100
    },
    {
      "epoch": 0.08380004148516905,
      "grad_norm": 3.4828169345855713,
      "learning_rate": 5.330452921628497e-05,
      "loss": 1.9595,
      "step": 101
    },
    {
      "epoch": 0.08462974486621033,
      "grad_norm": 3.2948338985443115,
      "learning_rate": 5.247918773366112e-05,
      "loss": 1.9986,
      "step": 102
    },
    {
      "epoch": 0.0854594482472516,
      "grad_norm": 2.7784478664398193,
      "learning_rate": 5.165316846586541e-05,
      "loss": 3.2244,
      "step": 103
    },
    {
      "epoch": 0.08628915162829288,
      "grad_norm": 3.076371669769287,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 2.4487,
      "step": 104
    },
    {
      "epoch": 0.08711885500933417,
      "grad_norm": 3.1001675128936768,
      "learning_rate": 5e-05,
      "loss": 3.2946,
      "step": 105
    },
    {
      "epoch": 0.08794855839037544,
      "grad_norm": 3.0547237396240234,
      "learning_rate": 4.917330276168208e-05,
      "loss": 2.6136,
      "step": 106
    },
    {
      "epoch": 0.08877826177141672,
      "grad_norm": 2.365811586380005,
      "learning_rate": 4.834683153413459e-05,
      "loss": 2.8773,
      "step": 107
    },
    {
      "epoch": 0.089607965152458,
      "grad_norm": 2.260733127593994,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 2.591,
      "step": 108
    },
    {
      "epoch": 0.09043766853349927,
      "grad_norm": 2.3252551555633545,
      "learning_rate": 4.669547078371504e-05,
      "loss": 3.2155,
      "step": 109
    },
    {
      "epoch": 0.09126737191454055,
      "grad_norm": 2.662757158279419,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 2.4218,
      "step": 110
    },
    {
      "epoch": 0.09209707529558182,
      "grad_norm": 2.295217514038086,
      "learning_rate": 4.504772348747687e-05,
      "loss": 2.8749,
      "step": 111
    },
    {
      "epoch": 0.0929267786766231,
      "grad_norm": 2.226912498474121,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 2.4908,
      "step": 112
    },
    {
      "epoch": 0.09375648205766439,
      "grad_norm": 2.503540515899658,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 2.6177,
      "step": 113
    },
    {
      "epoch": 0.09458618543870566,
      "grad_norm": 2.25520396232605,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 3.3559,
      "step": 114
    },
    {
      "epoch": 0.09541588881974694,
      "grad_norm": 2.3885161876678467,
      "learning_rate": 4.17702704859633e-05,
      "loss": 3.4402,
      "step": 115
    },
    {
      "epoch": 0.09624559220078822,
      "grad_norm": 2.5151796340942383,
      "learning_rate": 4.095597328339452e-05,
      "loss": 3.4725,
      "step": 116
    },
    {
      "epoch": 0.0970752955818295,
      "grad_norm": 2.4680063724517822,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 3.1247,
      "step": 117
    },
    {
      "epoch": 0.09790499896287078,
      "grad_norm": 2.733109951019287,
      "learning_rate": 3.933501846281267e-05,
      "loss": 3.2807,
      "step": 118
    },
    {
      "epoch": 0.09873470234391205,
      "grad_norm": 2.5370259284973145,
      "learning_rate": 3.852880399766243e-05,
      "loss": 2.7042,
      "step": 119
    },
    {
      "epoch": 0.09956440572495333,
      "grad_norm": 2.6658337116241455,
      "learning_rate": 3.772572564296005e-05,
      "loss": 3.3081,
      "step": 120
    },
    {
      "epoch": 0.10039410910599461,
      "grad_norm": 2.469576597213745,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 3.3873,
      "step": 121
    },
    {
      "epoch": 0.10122381248703588,
      "grad_norm": 2.723525285720825,
      "learning_rate": 3.612985456190778e-05,
      "loss": 3.2864,
      "step": 122
    },
    {
      "epoch": 0.10205351586807716,
      "grad_norm": 2.9827985763549805,
      "learning_rate": 3.533749813077677e-05,
      "loss": 3.6777,
      "step": 123
    },
    {
      "epoch": 0.10288321924911845,
      "grad_norm": 2.6939775943756104,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 3.3984,
      "step": 124
    },
    {
      "epoch": 0.10371292263015972,
      "grad_norm": 2.7732110023498535,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 2.6987,
      "step": 125
    },
    {
      "epoch": 0.104542626011201,
      "grad_norm": 3.2264564037323,
      "learning_rate": 3.298534127791785e-05,
      "loss": 3.3814,
      "step": 126
    },
    {
      "epoch": 0.10537232939224227,
      "grad_norm": 3.103226661682129,
      "learning_rate": 3.221030765387417e-05,
      "loss": 4.5089,
      "step": 127
    },
    {
      "epoch": 0.10620203277328355,
      "grad_norm": 3.085115671157837,
      "learning_rate": 3.144013755408895e-05,
      "loss": 3.4284,
      "step": 128
    },
    {
      "epoch": 0.10703173615432483,
      "grad_norm": 3.0896966457366943,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 3.3294,
      "step": 129
    },
    {
      "epoch": 0.1078614395353661,
      "grad_norm": 2.900888442993164,
      "learning_rate": 2.991522876735154e-05,
      "loss": 2.9881,
      "step": 130
    },
    {
      "epoch": 0.10869114291640739,
      "grad_norm": 2.8660571575164795,
      "learning_rate": 2.916090697523549e-05,
      "loss": 2.9304,
      "step": 131
    },
    {
      "epoch": 0.10952084629744867,
      "grad_norm": 3.067079782485962,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 3.3371,
      "step": 132
    },
    {
      "epoch": 0.11035054967848994,
      "grad_norm": 3.177497625350952,
      "learning_rate": 2.766955965735968e-05,
      "loss": 3.3043,
      "step": 133
    },
    {
      "epoch": 0.11118025305953122,
      "grad_norm": 3.1463634967803955,
      "learning_rate": 2.693294185106562e-05,
      "loss": 2.8956,
      "step": 134
    },
    {
      "epoch": 0.11200995644057249,
      "grad_norm": 3.343120813369751,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 3.0637,
      "step": 135
    },
    {
      "epoch": 0.11283965982161377,
      "grad_norm": 3.5586256980895996,
      "learning_rate": 2.547882480847461e-05,
      "loss": 3.1414,
      "step": 136
    },
    {
      "epoch": 0.11366936320265505,
      "grad_norm": 3.7543158531188965,
      "learning_rate": 2.476172311325783e-05,
      "loss": 3.4655,
      "step": 137
    },
    {
      "epoch": 0.11449906658369632,
      "grad_norm": 3.7594501972198486,
      "learning_rate": 2.405152131093926e-05,
      "loss": 3.3674,
      "step": 138
    },
    {
      "epoch": 0.1153287699647376,
      "grad_norm": 3.6149985790252686,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 3.105,
      "step": 139
    },
    {
      "epoch": 0.11615847334577889,
      "grad_norm": 4.05710506439209,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 3.4044,
      "step": 140
    },
    {
      "epoch": 0.11698817672682016,
      "grad_norm": 4.085519313812256,
      "learning_rate": 2.196424713241637e-05,
      "loss": 2.7692,
      "step": 141
    },
    {
      "epoch": 0.11781788010786144,
      "grad_norm": 4.447892665863037,
      "learning_rate": 2.128356686585282e-05,
      "loss": 3.3674,
      "step": 142
    },
    {
      "epoch": 0.11864758348890271,
      "grad_norm": 4.5582122802734375,
      "learning_rate": 2.061073738537635e-05,
      "loss": 3.2484,
      "step": 143
    },
    {
      "epoch": 0.119477286869944,
      "grad_norm": 5.234733581542969,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 4.3494,
      "step": 144
    },
    {
      "epoch": 0.12030699025098528,
      "grad_norm": 5.591521263122559,
      "learning_rate": 1.928936436551661e-05,
      "loss": 3.3993,
      "step": 145
    },
    {
      "epoch": 0.12113669363202655,
      "grad_norm": 5.82487154006958,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 4.1448,
      "step": 146
    },
    {
      "epoch": 0.12196639701306783,
      "grad_norm": 6.688323020935059,
      "learning_rate": 1.800157297483417e-05,
      "loss": 4.4812,
      "step": 147
    },
    {
      "epoch": 0.12279610039410911,
      "grad_norm": 7.02082633972168,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 4.1065,
      "step": 148
    },
    {
      "epoch": 0.12362580377515038,
      "grad_norm": 10.074577331542969,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 4.8377,
      "step": 149
    },
    {
      "epoch": 0.12445550715619166,
      "grad_norm": 12.660624504089355,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 5.0321,
      "step": 150
    },
    {
      "epoch": 0.12445550715619166,
      "eval_loss": 0.7792608141899109,
      "eval_runtime": 25.9227,
      "eval_samples_per_second": 78.31,
      "eval_steps_per_second": 19.597,
      "step": 150
    },
    {
      "epoch": 0.12528521053723293,
      "grad_norm": 5.037601947784424,
      "learning_rate": 1.553232954407171e-05,
      "loss": 1.9302,
      "step": 151
    },
    {
      "epoch": 0.12611491391827423,
      "grad_norm": 2.907715082168579,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 2.4961,
      "step": 152
    },
    {
      "epoch": 0.1269446172993155,
      "grad_norm": 3.101891279220581,
      "learning_rate": 1.435357758543015e-05,
      "loss": 2.7582,
      "step": 153
    },
    {
      "epoch": 0.12777432068035677,
      "grad_norm": 2.8044328689575195,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 2.8822,
      "step": 154
    },
    {
      "epoch": 0.12860402406139804,
      "grad_norm": 2.718334913253784,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 2.6734,
      "step": 155
    },
    {
      "epoch": 0.12943372744243933,
      "grad_norm": 2.3871145248413086,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 2.907,
      "step": 156
    },
    {
      "epoch": 0.1302634308234806,
      "grad_norm": 2.574359655380249,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 3.1625,
      "step": 157
    },
    {
      "epoch": 0.13109313420452187,
      "grad_norm": 2.4399795532226562,
      "learning_rate": 1.157994445715706e-05,
      "loss": 2.5406,
      "step": 158
    },
    {
      "epoch": 0.13192283758556317,
      "grad_norm": 2.3223538398742676,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 2.9261,
      "step": 159
    },
    {
      "epoch": 0.13275254096660444,
      "grad_norm": 2.3015785217285156,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 2.3804,
      "step": 160
    },
    {
      "epoch": 0.1335822443476457,
      "grad_norm": 2.754807949066162,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 2.9268,
      "step": 161
    },
    {
      "epoch": 0.134411947728687,
      "grad_norm": 2.7829599380493164,
      "learning_rate": 9.549150281252633e-06,
      "loss": 4.0291,
      "step": 162
    },
    {
      "epoch": 0.13524165110972827,
      "grad_norm": 2.3959383964538574,
      "learning_rate": 9.068759265665384e-06,
      "loss": 2.8059,
      "step": 163
    },
    {
      "epoch": 0.13607135449076954,
      "grad_norm": 2.4433817863464355,
      "learning_rate": 8.599558442598998e-06,
      "loss": 2.848,
      "step": 164
    },
    {
      "epoch": 0.13690105787181084,
      "grad_norm": 2.4209845066070557,
      "learning_rate": 8.141676086873572e-06,
      "loss": 3.2109,
      "step": 165
    },
    {
      "epoch": 0.1377307612528521,
      "grad_norm": 2.7219181060791016,
      "learning_rate": 7.695237378953223e-06,
      "loss": 3.0729,
      "step": 166
    },
    {
      "epoch": 0.13856046463389338,
      "grad_norm": 2.4126503467559814,
      "learning_rate": 7.260364370723044e-06,
      "loss": 3.1755,
      "step": 167
    },
    {
      "epoch": 0.13939016801493467,
      "grad_norm": 2.526926040649414,
      "learning_rate": 6.837175952121306e-06,
      "loss": 2.5644,
      "step": 168
    },
    {
      "epoch": 0.14021987139597594,
      "grad_norm": 2.8227434158325195,
      "learning_rate": 6.425787818636131e-06,
      "loss": 2.8808,
      "step": 169
    },
    {
      "epoch": 0.1410495747770172,
      "grad_norm": 2.812786102294922,
      "learning_rate": 6.026312439675552e-06,
      "loss": 2.7292,
      "step": 170
    },
    {
      "epoch": 0.14187927815805848,
      "grad_norm": 3.001596450805664,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 3.5321,
      "step": 171
    },
    {
      "epoch": 0.14270898153909978,
      "grad_norm": 2.64874529838562,
      "learning_rate": 5.263533508961827e-06,
      "loss": 2.6967,
      "step": 172
    },
    {
      "epoch": 0.14353868492014105,
      "grad_norm": 3.001011848449707,
      "learning_rate": 4.900438493352055e-06,
      "loss": 3.2194,
      "step": 173
    },
    {
      "epoch": 0.14436838830118232,
      "grad_norm": 2.783543586730957,
      "learning_rate": 4.549673247541875e-06,
      "loss": 3.2433,
      "step": 174
    },
    {
      "epoch": 0.1451980916822236,
      "grad_norm": 2.780012845993042,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 2.8927,
      "step": 175
    },
    {
      "epoch": 0.14602779506326488,
      "grad_norm": 3.1829144954681396,
      "learning_rate": 3.885512251130763e-06,
      "loss": 3.5419,
      "step": 176
    },
    {
      "epoch": 0.14685749844430615,
      "grad_norm": 2.676326036453247,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.6604,
      "step": 177
    },
    {
      "epoch": 0.14768720182534745,
      "grad_norm": 3.134683132171631,
      "learning_rate": 3.271776770026963e-06,
      "loss": 3.2348,
      "step": 178
    },
    {
      "epoch": 0.14851690520638872,
      "grad_norm": 2.8262722492218018,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 2.937,
      "step": 179
    },
    {
      "epoch": 0.14934660858742999,
      "grad_norm": 3.1728761196136475,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 3.5081,
      "step": 180
    },
    {
      "epoch": 0.15017631196847128,
      "grad_norm": 3.4726243019104004,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 3.4762,
      "step": 181
    },
    {
      "epoch": 0.15100601534951255,
      "grad_norm": 3.304121494293213,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 3.012,
      "step": 182
    },
    {
      "epoch": 0.15183571873055382,
      "grad_norm": 3.222151756286621,
      "learning_rate": 1.962316193157593e-06,
      "loss": 2.6993,
      "step": 183
    },
    {
      "epoch": 0.15266542211159512,
      "grad_norm": 3.5558347702026367,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 3.1719,
      "step": 184
    },
    {
      "epoch": 0.1534951254926364,
      "grad_norm": 3.4521775245666504,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 3.002,
      "step": 185
    },
    {
      "epoch": 0.15432482887367766,
      "grad_norm": 4.042253494262695,
      "learning_rate": 1.333670137599713e-06,
      "loss": 3.6709,
      "step": 186
    },
    {
      "epoch": 0.15515453225471892,
      "grad_norm": 3.715176820755005,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 3.2547,
      "step": 187
    },
    {
      "epoch": 0.15598423563576022,
      "grad_norm": 4.24824333190918,
      "learning_rate": 9.810017062595322e-07,
      "loss": 3.4294,
      "step": 188
    },
    {
      "epoch": 0.1568139390168015,
      "grad_norm": 4.170515537261963,
      "learning_rate": 8.247462563808817e-07,
      "loss": 3.3346,
      "step": 189
    },
    {
      "epoch": 0.15764364239784276,
      "grad_norm": 4.353022575378418,
      "learning_rate": 6.819348298638839e-07,
      "loss": 3.9641,
      "step": 190
    },
    {
      "epoch": 0.15847334577888406,
      "grad_norm": 3.7188031673431396,
      "learning_rate": 5.526064699265753e-07,
      "loss": 2.6656,
      "step": 191
    },
    {
      "epoch": 0.15930304915992533,
      "grad_norm": 5.262839317321777,
      "learning_rate": 4.367965336512403e-07,
      "loss": 3.8125,
      "step": 192
    },
    {
      "epoch": 0.1601327525409666,
      "grad_norm": 4.676213264465332,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 3.2992,
      "step": 193
    },
    {
      "epoch": 0.1609624559220079,
      "grad_norm": 5.197406768798828,
      "learning_rate": 2.458548727494292e-07,
      "loss": 3.9291,
      "step": 194
    },
    {
      "epoch": 0.16179215930304916,
      "grad_norm": 5.474231243133545,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 3.8101,
      "step": 195
    },
    {
      "epoch": 0.16262186268409043,
      "grad_norm": 6.884095191955566,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 4.7059,
      "step": 196
    },
    {
      "epoch": 0.16345156606513173,
      "grad_norm": 6.367083549499512,
      "learning_rate": 6.150154258476315e-08,
      "loss": 4.0424,
      "step": 197
    },
    {
      "epoch": 0.164281269446173,
      "grad_norm": 8.017609596252441,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 4.0856,
      "step": 198
    },
    {
      "epoch": 0.16511097282721426,
      "grad_norm": 7.671690940856934,
      "learning_rate": 6.834750376549792e-09,
      "loss": 3.711,
      "step": 199
    },
    {
      "epoch": 0.16594067620825556,
      "grad_norm": 14.535627365112305,
      "learning_rate": 0.0,
      "loss": 4.6014,
      "step": 200
    },
    {
      "epoch": 0.16594067620825556,
      "eval_loss": 0.7629001140594482,
      "eval_runtime": 25.9355,
      "eval_samples_per_second": 78.271,
      "eval_steps_per_second": 19.587,
      "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": 1.4778566122143744e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}