{
  "best_metric": 2.647024154663086,
  "best_model_checkpoint": "miner_id_24/checkpoint-100",
  "epoch": 2.4024024024024024,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.012012012012012012,
      "grad_norm": 0.2368987798690796,
      "learning_rate": 1e-05,
      "loss": 2.6271,
      "step": 1
    },
    {
      "epoch": 0.012012012012012012,
      "eval_loss": 2.8036038875579834,
      "eval_runtime": 9.9077,
      "eval_samples_per_second": 14.13,
      "eval_steps_per_second": 3.533,
      "step": 1
    },
    {
      "epoch": 0.024024024024024024,
      "grad_norm": 0.24696952104568481,
      "learning_rate": 2e-05,
      "loss": 2.6605,
      "step": 2
    },
    {
      "epoch": 0.036036036036036036,
      "grad_norm": 0.2612786293029785,
      "learning_rate": 3e-05,
      "loss": 2.7005,
      "step": 3
    },
    {
      "epoch": 0.04804804804804805,
      "grad_norm": 0.2701415717601776,
      "learning_rate": 4e-05,
      "loss": 2.6186,
      "step": 4
    },
    {
      "epoch": 0.06006006006006006,
      "grad_norm": 0.2381216585636139,
      "learning_rate": 5e-05,
      "loss": 2.5193,
      "step": 5
    },
    {
      "epoch": 0.07207207207207207,
      "grad_norm": 0.2288311868906021,
      "learning_rate": 6e-05,
      "loss": 2.7264,
      "step": 6
    },
    {
      "epoch": 0.08408408408408409,
      "grad_norm": 0.24674589931964874,
      "learning_rate": 7e-05,
      "loss": 2.6344,
      "step": 7
    },
    {
      "epoch": 0.0960960960960961,
      "grad_norm": 0.26552438735961914,
      "learning_rate": 8e-05,
      "loss": 2.6916,
      "step": 8
    },
    {
      "epoch": 0.10810810810810811,
      "grad_norm": 0.26411452889442444,
      "learning_rate": 9e-05,
      "loss": 2.7281,
      "step": 9
    },
    {
      "epoch": 0.12012012012012012,
      "grad_norm": 0.2651025056838989,
      "learning_rate": 0.0001,
      "loss": 2.7556,
      "step": 10
    },
    {
      "epoch": 0.13213213213213212,
      "grad_norm": 0.28291529417037964,
      "learning_rate": 9.999316524962345e-05,
      "loss": 2.7109,
      "step": 11
    },
    {
      "epoch": 0.14414414414414414,
      "grad_norm": 0.29600831866264343,
      "learning_rate": 9.997266286704631e-05,
      "loss": 2.621,
      "step": 12
    },
    {
      "epoch": 0.15615615615615616,
      "grad_norm": 0.2944232225418091,
      "learning_rate": 9.993849845741524e-05,
      "loss": 2.703,
      "step": 13
    },
    {
      "epoch": 0.16816816816816818,
      "grad_norm": 0.30435818433761597,
      "learning_rate": 9.989068136093873e-05,
      "loss": 2.7051,
      "step": 14
    },
    {
      "epoch": 0.18018018018018017,
      "grad_norm": 0.3093186020851135,
      "learning_rate": 9.98292246503335e-05,
      "loss": 2.6282,
      "step": 15
    },
    {
      "epoch": 0.1921921921921922,
      "grad_norm": 0.38756853342056274,
      "learning_rate": 9.975414512725057e-05,
      "loss": 2.6861,
      "step": 16
    },
    {
      "epoch": 0.2042042042042042,
      "grad_norm": 0.3695060908794403,
      "learning_rate": 9.966546331768191e-05,
      "loss": 2.7373,
      "step": 17
    },
    {
      "epoch": 0.21621621621621623,
      "grad_norm": 0.3841871917247772,
      "learning_rate": 9.956320346634876e-05,
      "loss": 2.6895,
      "step": 18
    },
    {
      "epoch": 0.22822822822822822,
      "grad_norm": 0.4169112741947174,
      "learning_rate": 9.944739353007344e-05,
      "loss": 2.6329,
      "step": 19
    },
    {
      "epoch": 0.24024024024024024,
      "grad_norm": 0.4456412196159363,
      "learning_rate": 9.931806517013612e-05,
      "loss": 2.8288,
      "step": 20
    },
    {
      "epoch": 0.25225225225225223,
      "grad_norm": 0.23516860604286194,
      "learning_rate": 9.917525374361912e-05,
      "loss": 2.5588,
      "step": 21
    },
    {
      "epoch": 0.26426426426426425,
      "grad_norm": 0.22118473052978516,
      "learning_rate": 9.901899829374047e-05,
      "loss": 2.6182,
      "step": 22
    },
    {
      "epoch": 0.27627627627627627,
      "grad_norm": 0.21510648727416992,
      "learning_rate": 9.884934153917997e-05,
      "loss": 2.4664,
      "step": 23
    },
    {
      "epoch": 0.2882882882882883,
      "grad_norm": 0.19684556126594543,
      "learning_rate": 9.86663298624003e-05,
      "loss": 2.5491,
      "step": 24
    },
    {
      "epoch": 0.3003003003003003,
      "grad_norm": 0.1943308711051941,
      "learning_rate": 9.847001329696653e-05,
      "loss": 2.7107,
      "step": 25
    },
    {
      "epoch": 0.3123123123123123,
      "grad_norm": 0.20300710201263428,
      "learning_rate": 9.826044551386744e-05,
      "loss": 2.5691,
      "step": 26
    },
    {
      "epoch": 0.32432432432432434,
      "grad_norm": 0.20375128090381622,
      "learning_rate": 9.803768380684242e-05,
      "loss": 2.591,
      "step": 27
    },
    {
      "epoch": 0.33633633633633636,
      "grad_norm": 0.21982106566429138,
      "learning_rate": 9.780178907671789e-05,
      "loss": 2.6421,
      "step": 28
    },
    {
      "epoch": 0.3483483483483483,
      "grad_norm": 0.22565782070159912,
      "learning_rate": 9.755282581475769e-05,
      "loss": 2.6387,
      "step": 29
    },
    {
      "epoch": 0.36036036036036034,
      "grad_norm": 0.2482357770204544,
      "learning_rate": 9.729086208503174e-05,
      "loss": 2.6121,
      "step": 30
    },
    {
      "epoch": 0.37237237237237236,
      "grad_norm": 0.25244030356407166,
      "learning_rate": 9.701596950580806e-05,
      "loss": 2.7402,
      "step": 31
    },
    {
      "epoch": 0.3843843843843844,
      "grad_norm": 0.24448777735233307,
      "learning_rate": 9.672822322997305e-05,
      "loss": 2.678,
      "step": 32
    },
    {
      "epoch": 0.3963963963963964,
      "grad_norm": 0.25511646270751953,
      "learning_rate": 9.642770192448536e-05,
      "loss": 2.6438,
      "step": 33
    },
    {
      "epoch": 0.4084084084084084,
      "grad_norm": 0.27205294370651245,
      "learning_rate": 9.611448774886924e-05,
      "loss": 2.7213,
      "step": 34
    },
    {
      "epoch": 0.42042042042042044,
      "grad_norm": 0.2772773206233978,
      "learning_rate": 9.578866633275288e-05,
      "loss": 2.6789,
      "step": 35
    },
    {
      "epoch": 0.43243243243243246,
      "grad_norm": 0.29589328169822693,
      "learning_rate": 9.545032675245813e-05,
      "loss": 2.7009,
      "step": 36
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.3177066147327423,
      "learning_rate": 9.509956150664796e-05,
      "loss": 2.769,
      "step": 37
    },
    {
      "epoch": 0.45645645645645644,
      "grad_norm": 0.3396986424922943,
      "learning_rate": 9.473646649103818e-05,
      "loss": 2.7435,
      "step": 38
    },
    {
      "epoch": 0.46846846846846846,
      "grad_norm": 0.37219908833503723,
      "learning_rate": 9.43611409721806e-05,
      "loss": 2.6979,
      "step": 39
    },
    {
      "epoch": 0.4804804804804805,
      "grad_norm": 0.4017895460128784,
      "learning_rate": 9.397368756032445e-05,
      "loss": 2.8455,
      "step": 40
    },
    {
      "epoch": 0.4924924924924925,
      "grad_norm": 0.2517055869102478,
      "learning_rate": 9.357421218136386e-05,
      "loss": 2.6304,
      "step": 41
    },
    {
      "epoch": 0.5045045045045045,
      "grad_norm": 0.20714128017425537,
      "learning_rate": 9.316282404787871e-05,
      "loss": 2.5886,
      "step": 42
    },
    {
      "epoch": 0.5165165165165165,
      "grad_norm": 0.20932313799858093,
      "learning_rate": 9.273963562927695e-05,
      "loss": 2.5929,
      "step": 43
    },
    {
      "epoch": 0.5285285285285285,
      "grad_norm": 0.22508971393108368,
      "learning_rate": 9.230476262104677e-05,
      "loss": 2.5286,
      "step": 44
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 0.21495309472084045,
      "learning_rate": 9.185832391312644e-05,
      "loss": 2.6029,
      "step": 45
    },
    {
      "epoch": 0.5525525525525525,
      "grad_norm": 0.1941525787115097,
      "learning_rate": 9.140044155740101e-05,
      "loss": 2.5473,
      "step": 46
    },
    {
      "epoch": 0.5645645645645646,
      "grad_norm": 0.2017686665058136,
      "learning_rate": 9.093124073433463e-05,
      "loss": 2.543,
      "step": 47
    },
    {
      "epoch": 0.5765765765765766,
      "grad_norm": 0.2098369151353836,
      "learning_rate": 9.045084971874738e-05,
      "loss": 2.5588,
      "step": 48
    },
    {
      "epoch": 0.5885885885885885,
      "grad_norm": 0.21940922737121582,
      "learning_rate": 8.995939984474624e-05,
      "loss": 2.6342,
      "step": 49
    },
    {
      "epoch": 0.6006006006006006,
      "grad_norm": 0.2287360280752182,
      "learning_rate": 8.945702546981969e-05,
      "loss": 2.6876,
      "step": 50
    },
    {
      "epoch": 0.6006006006006006,
      "eval_loss": 2.650160551071167,
      "eval_runtime": 10.1713,
      "eval_samples_per_second": 13.764,
      "eval_steps_per_second": 3.441,
      "step": 50
    },
    {
      "epoch": 0.6126126126126126,
      "grad_norm": 0.22518697381019592,
      "learning_rate": 8.894386393810563e-05,
      "loss": 2.5879,
      "step": 51
    },
    {
      "epoch": 0.6246246246246246,
      "grad_norm": 0.24255499243736267,
      "learning_rate": 8.842005554284296e-05,
      "loss": 2.7205,
      "step": 52
    },
    {
      "epoch": 0.6366366366366366,
      "grad_norm": 0.2652047574520111,
      "learning_rate": 8.788574348801675e-05,
      "loss": 2.5712,
      "step": 53
    },
    {
      "epoch": 0.6486486486486487,
      "grad_norm": 0.2551002502441406,
      "learning_rate": 8.73410738492077e-05,
      "loss": 2.6449,
      "step": 54
    },
    {
      "epoch": 0.6606606606606606,
      "grad_norm": 0.2617751657962799,
      "learning_rate": 8.678619553365659e-05,
      "loss": 2.6196,
      "step": 55
    },
    {
      "epoch": 0.6726726726726727,
      "grad_norm": 0.3021559417247772,
      "learning_rate": 8.622126023955446e-05,
      "loss": 2.7571,
      "step": 56
    },
    {
      "epoch": 0.6846846846846847,
      "grad_norm": 0.3074163496494293,
      "learning_rate": 8.564642241456986e-05,
      "loss": 2.799,
      "step": 57
    },
    {
      "epoch": 0.6966966966966966,
      "grad_norm": 0.3206866681575775,
      "learning_rate": 8.506183921362443e-05,
      "loss": 2.7213,
      "step": 58
    },
    {
      "epoch": 0.7087087087087087,
      "grad_norm": 0.3493919372558594,
      "learning_rate": 8.44676704559283e-05,
      "loss": 2.8387,
      "step": 59
    },
    {
      "epoch": 0.7207207207207207,
      "grad_norm": 0.37002992630004883,
      "learning_rate": 8.386407858128706e-05,
      "loss": 2.9261,
      "step": 60
    },
    {
      "epoch": 0.7327327327327328,
      "grad_norm": 0.1866091787815094,
      "learning_rate": 8.32512286056924e-05,
      "loss": 2.6282,
      "step": 61
    },
    {
      "epoch": 0.7447447447447447,
      "grad_norm": 0.20297600328922272,
      "learning_rate": 8.262928807620843e-05,
      "loss": 2.5916,
      "step": 62
    },
    {
      "epoch": 0.7567567567567568,
      "grad_norm": 0.2202749401330948,
      "learning_rate": 8.199842702516583e-05,
      "loss": 2.5317,
      "step": 63
    },
    {
      "epoch": 0.7687687687687688,
      "grad_norm": 0.1926620751619339,
      "learning_rate": 8.135881792367686e-05,
      "loss": 2.5066,
      "step": 64
    },
    {
      "epoch": 0.7807807807807807,
      "grad_norm": 0.19466735422611237,
      "learning_rate": 8.07106356344834e-05,
      "loss": 2.6228,
      "step": 65
    },
    {
      "epoch": 0.7927927927927928,
      "grad_norm": 0.1981358379125595,
      "learning_rate": 8.005405736415126e-05,
      "loss": 2.6074,
      "step": 66
    },
    {
      "epoch": 0.8048048048048048,
      "grad_norm": 0.1992652863264084,
      "learning_rate": 7.938926261462366e-05,
      "loss": 2.5795,
      "step": 67
    },
    {
      "epoch": 0.8168168168168168,
      "grad_norm": 0.22001273930072784,
      "learning_rate": 7.871643313414718e-05,
      "loss": 2.6496,
      "step": 68
    },
    {
      "epoch": 0.8288288288288288,
      "grad_norm": 0.209934800863266,
      "learning_rate": 7.803575286758364e-05,
      "loss": 2.4917,
      "step": 69
    },
    {
      "epoch": 0.8408408408408409,
      "grad_norm": 0.2247212529182434,
      "learning_rate": 7.734740790612136e-05,
      "loss": 2.6059,
      "step": 70
    },
    {
      "epoch": 0.8528528528528528,
      "grad_norm": 0.22632429003715515,
      "learning_rate": 7.66515864363997e-05,
      "loss": 2.5653,
      "step": 71
    },
    {
      "epoch": 0.8648648648648649,
      "grad_norm": 0.24432453513145447,
      "learning_rate": 7.594847868906076e-05,
      "loss": 2.6121,
      "step": 72
    },
    {
      "epoch": 0.8768768768768769,
      "grad_norm": 0.2481575906276703,
      "learning_rate": 7.52382768867422e-05,
      "loss": 2.6929,
      "step": 73
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.26193660497665405,
      "learning_rate": 7.452117519152542e-05,
      "loss": 2.6311,
      "step": 74
    },
    {
      "epoch": 0.9009009009009009,
      "grad_norm": 0.28893381357192993,
      "learning_rate": 7.379736965185368e-05,
      "loss": 2.5628,
      "step": 75
    },
    {
      "epoch": 0.9129129129129129,
      "grad_norm": 0.2971073389053345,
      "learning_rate": 7.30670581489344e-05,
      "loss": 2.7416,
      "step": 76
    },
    {
      "epoch": 0.924924924924925,
      "grad_norm": 0.3262496590614319,
      "learning_rate": 7.233044034264034e-05,
      "loss": 2.7969,
      "step": 77
    },
    {
      "epoch": 0.9369369369369369,
      "grad_norm": 0.32592859864234924,
      "learning_rate": 7.158771761692464e-05,
      "loss": 2.7319,
      "step": 78
    },
    {
      "epoch": 0.948948948948949,
      "grad_norm": 0.3250863552093506,
      "learning_rate": 7.083909302476453e-05,
      "loss": 2.7505,
      "step": 79
    },
    {
      "epoch": 0.960960960960961,
      "grad_norm": 0.4082293212413788,
      "learning_rate": 7.008477123264848e-05,
      "loss": 2.7908,
      "step": 80
    },
    {
      "epoch": 0.972972972972973,
      "grad_norm": 0.18473713099956512,
      "learning_rate": 6.932495846462261e-05,
      "loss": 2.5129,
      "step": 81
    },
    {
      "epoch": 0.984984984984985,
      "grad_norm": 0.21697285771369934,
      "learning_rate": 6.855986244591104e-05,
      "loss": 2.5306,
      "step": 82
    },
    {
      "epoch": 0.996996996996997,
      "grad_norm": 0.3190633952617645,
      "learning_rate": 6.778969234612584e-05,
      "loss": 2.7935,
      "step": 83
    },
    {
      "epoch": 1.009009009009009,
      "grad_norm": 1.0345549583435059,
      "learning_rate": 6.701465872208216e-05,
      "loss": 4.9466,
      "step": 84
    },
    {
      "epoch": 1.021021021021021,
      "grad_norm": 0.16876251995563507,
      "learning_rate": 6.623497346023418e-05,
      "loss": 2.5027,
      "step": 85
    },
    {
      "epoch": 1.033033033033033,
      "grad_norm": 0.17933110892772675,
      "learning_rate": 6.545084971874738e-05,
      "loss": 2.5302,
      "step": 86
    },
    {
      "epoch": 1.045045045045045,
      "grad_norm": 0.17964915931224823,
      "learning_rate": 6.466250186922325e-05,
      "loss": 2.4274,
      "step": 87
    },
    {
      "epoch": 1.057057057057057,
      "grad_norm": 0.18235374987125397,
      "learning_rate": 6.387014543809223e-05,
      "loss": 2.5331,
      "step": 88
    },
    {
      "epoch": 1.069069069069069,
      "grad_norm": 0.19002363085746765,
      "learning_rate": 6.307399704769099e-05,
      "loss": 2.5192,
      "step": 89
    },
    {
      "epoch": 1.0810810810810811,
      "grad_norm": 0.20002731680870056,
      "learning_rate": 6.227427435703997e-05,
      "loss": 2.4528,
      "step": 90
    },
    {
      "epoch": 1.093093093093093,
      "grad_norm": 0.19832709431648254,
      "learning_rate": 6.147119600233758e-05,
      "loss": 2.4892,
      "step": 91
    },
    {
      "epoch": 1.105105105105105,
      "grad_norm": 0.2067626416683197,
      "learning_rate": 6.066498153718735e-05,
      "loss": 2.5768,
      "step": 92
    },
    {
      "epoch": 1.117117117117117,
      "grad_norm": 0.2166813611984253,
      "learning_rate": 5.985585137257401e-05,
      "loss": 2.4095,
      "step": 93
    },
    {
      "epoch": 1.1291291291291292,
      "grad_norm": 0.23899836838245392,
      "learning_rate": 5.90440267166055e-05,
      "loss": 2.5374,
      "step": 94
    },
    {
      "epoch": 1.1411411411411412,
      "grad_norm": 0.23798096179962158,
      "learning_rate": 5.8229729514036705e-05,
      "loss": 2.5883,
      "step": 95
    },
    {
      "epoch": 1.1531531531531531,
      "grad_norm": 0.24018827080726624,
      "learning_rate": 5.74131823855921e-05,
      "loss": 2.5366,
      "step": 96
    },
    {
      "epoch": 1.165165165165165,
      "grad_norm": 0.25265100598335266,
      "learning_rate": 5.6594608567103456e-05,
      "loss": 2.572,
      "step": 97
    },
    {
      "epoch": 1.1771771771771773,
      "grad_norm": 0.2808419167995453,
      "learning_rate": 5.577423184847932e-05,
      "loss": 2.5192,
      "step": 98
    },
    {
      "epoch": 1.1891891891891893,
      "grad_norm": 0.31896519660949707,
      "learning_rate": 5.495227651252315e-05,
      "loss": 2.5365,
      "step": 99
    },
    {
      "epoch": 1.2012012012012012,
      "grad_norm": 0.33212482929229736,
      "learning_rate": 5.4128967273616625e-05,
      "loss": 2.6007,
      "step": 100
    },
    {
      "epoch": 1.2012012012012012,
      "eval_loss": 2.647024154663086,
      "eval_runtime": 10.1673,
      "eval_samples_per_second": 13.77,
      "eval_steps_per_second": 3.442,
      "step": 100
    },
    {
      "epoch": 1.2132132132132132,
      "grad_norm": 0.3516274690628052,
      "learning_rate": 5.330452921628497e-05,
      "loss": 2.5115,
      "step": 101
    },
    {
      "epoch": 1.2252252252252251,
      "grad_norm": 0.38279256224632263,
      "learning_rate": 5.247918773366112e-05,
      "loss": 2.5937,
      "step": 102
    },
    {
      "epoch": 1.2372372372372373,
      "grad_norm": 0.43585026264190674,
      "learning_rate": 5.165316846586541e-05,
      "loss": 2.751,
      "step": 103
    },
    {
      "epoch": 1.2492492492492493,
      "grad_norm": 0.2676953673362732,
      "learning_rate": 5.0826697238317935e-05,
      "loss": 2.4865,
      "step": 104
    },
    {
      "epoch": 1.2612612612612613,
      "grad_norm": 0.21236161887645721,
      "learning_rate": 5e-05,
      "loss": 2.516,
      "step": 105
    },
    {
      "epoch": 1.2732732732732732,
      "grad_norm": 0.23394270241260529,
      "learning_rate": 4.917330276168208e-05,
      "loss": 2.4999,
      "step": 106
    },
    {
      "epoch": 1.2852852852852852,
      "grad_norm": 0.24831236898899078,
      "learning_rate": 4.834683153413459e-05,
      "loss": 2.3636,
      "step": 107
    },
    {
      "epoch": 1.2972972972972974,
      "grad_norm": 0.2524273693561554,
      "learning_rate": 4.7520812266338885e-05,
      "loss": 2.4915,
      "step": 108
    },
    {
      "epoch": 1.3093093093093093,
      "grad_norm": 0.2599671483039856,
      "learning_rate": 4.669547078371504e-05,
      "loss": 2.5837,
      "step": 109
    },
    {
      "epoch": 1.3213213213213213,
      "grad_norm": 0.24876700341701508,
      "learning_rate": 4.5871032726383386e-05,
      "loss": 2.4708,
      "step": 110
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.25530922412872314,
      "learning_rate": 4.504772348747687e-05,
      "loss": 2.4658,
      "step": 111
    },
    {
      "epoch": 1.3453453453453452,
      "grad_norm": 0.26658859848976135,
      "learning_rate": 4.4225768151520694e-05,
      "loss": 2.5462,
      "step": 112
    },
    {
      "epoch": 1.3573573573573574,
      "grad_norm": 0.2801316976547241,
      "learning_rate": 4.3405391432896555e-05,
      "loss": 2.4544,
      "step": 113
    },
    {
      "epoch": 1.3693693693693694,
      "grad_norm": 0.29828476905822754,
      "learning_rate": 4.2586817614407895e-05,
      "loss": 2.4428,
      "step": 114
    },
    {
      "epoch": 1.3813813813813813,
      "grad_norm": 0.3047056794166565,
      "learning_rate": 4.17702704859633e-05,
      "loss": 2.5501,
      "step": 115
    },
    {
      "epoch": 1.3933933933933935,
      "grad_norm": 0.3029329180717468,
      "learning_rate": 4.095597328339452e-05,
      "loss": 2.4838,
      "step": 116
    },
    {
      "epoch": 1.4054054054054055,
      "grad_norm": 0.3226974904537201,
      "learning_rate": 4.0144148627425993e-05,
      "loss": 2.5677,
      "step": 117
    },
    {
      "epoch": 1.4174174174174174,
      "grad_norm": 0.33047249913215637,
      "learning_rate": 3.933501846281267e-05,
      "loss": 2.4965,
      "step": 118
    },
    {
      "epoch": 1.4294294294294294,
      "grad_norm": 0.3641626834869385,
      "learning_rate": 3.852880399766243e-05,
      "loss": 2.4043,
      "step": 119
    },
    {
      "epoch": 1.4414414414414414,
      "grad_norm": 0.37925711274147034,
      "learning_rate": 3.772572564296005e-05,
      "loss": 2.5555,
      "step": 120
    },
    {
      "epoch": 1.4534534534534536,
      "grad_norm": 0.4225517213344574,
      "learning_rate": 3.6926002952309016e-05,
      "loss": 2.6301,
      "step": 121
    },
    {
      "epoch": 1.4654654654654655,
      "grad_norm": 0.4414302706718445,
      "learning_rate": 3.612985456190778e-05,
      "loss": 2.4514,
      "step": 122
    },
    {
      "epoch": 1.4774774774774775,
      "grad_norm": 0.4988545775413513,
      "learning_rate": 3.533749813077677e-05,
      "loss": 2.3902,
      "step": 123
    },
    {
      "epoch": 1.4894894894894894,
      "grad_norm": 0.3567710816860199,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 2.4857,
      "step": 124
    },
    {
      "epoch": 1.5015015015015014,
      "grad_norm": 0.2748575806617737,
      "learning_rate": 3.3765026539765834e-05,
      "loss": 2.5461,
      "step": 125
    },
    {
      "epoch": 1.5135135135135136,
      "grad_norm": 0.2784069776535034,
      "learning_rate": 3.298534127791785e-05,
      "loss": 2.5181,
      "step": 126
    },
    {
      "epoch": 1.5255255255255256,
      "grad_norm": 0.2793610990047455,
      "learning_rate": 3.221030765387417e-05,
      "loss": 2.4892,
      "step": 127
    },
    {
      "epoch": 1.5375375375375375,
      "grad_norm": 0.2694374620914459,
      "learning_rate": 3.144013755408895e-05,
      "loss": 2.4716,
      "step": 128
    },
    {
      "epoch": 1.5495495495495497,
      "grad_norm": 0.2895199656486511,
      "learning_rate": 3.0675041535377405e-05,
      "loss": 2.5786,
      "step": 129
    },
    {
      "epoch": 1.5615615615615615,
      "grad_norm": 0.2762201726436615,
      "learning_rate": 2.991522876735154e-05,
      "loss": 2.4547,
      "step": 130
    },
    {
      "epoch": 1.5735735735735736,
      "grad_norm": 0.28506338596343994,
      "learning_rate": 2.916090697523549e-05,
      "loss": 2.436,
      "step": 131
    },
    {
      "epoch": 1.5855855855855856,
      "grad_norm": 0.30929428339004517,
      "learning_rate": 2.8412282383075363e-05,
      "loss": 2.5629,
      "step": 132
    },
    {
      "epoch": 1.5975975975975976,
      "grad_norm": 0.303631454706192,
      "learning_rate": 2.766955965735968e-05,
      "loss": 2.5703,
      "step": 133
    },
    {
      "epoch": 1.6096096096096097,
      "grad_norm": 0.3370870053768158,
      "learning_rate": 2.693294185106562e-05,
      "loss": 2.4663,
      "step": 134
    },
    {
      "epoch": 1.6216216216216215,
      "grad_norm": 0.3156420588493347,
      "learning_rate": 2.6202630348146324e-05,
      "loss": 2.5369,
      "step": 135
    },
    {
      "epoch": 1.6336336336336337,
      "grad_norm": 0.3256956934928894,
      "learning_rate": 2.547882480847461e-05,
      "loss": 2.4589,
      "step": 136
    },
    {
      "epoch": 1.6456456456456456,
      "grad_norm": 0.3577045202255249,
      "learning_rate": 2.476172311325783e-05,
      "loss": 2.5416,
      "step": 137
    },
    {
      "epoch": 1.6576576576576576,
      "grad_norm": 0.3716708719730377,
      "learning_rate": 2.405152131093926e-05,
      "loss": 2.4658,
      "step": 138
    },
    {
      "epoch": 1.6696696696696698,
      "grad_norm": 0.3978820741176605,
      "learning_rate": 2.3348413563600325e-05,
      "loss": 2.4541,
      "step": 139
    },
    {
      "epoch": 1.6816816816816815,
      "grad_norm": 0.3858387768268585,
      "learning_rate": 2.2652592093878666e-05,
      "loss": 2.5498,
      "step": 140
    },
    {
      "epoch": 1.6936936936936937,
      "grad_norm": 0.42268532514572144,
      "learning_rate": 2.196424713241637e-05,
      "loss": 2.6186,
      "step": 141
    },
    {
      "epoch": 1.7057057057057057,
      "grad_norm": 0.4762258231639862,
      "learning_rate": 2.128356686585282e-05,
      "loss": 2.4898,
      "step": 142
    },
    {
      "epoch": 1.7177177177177176,
      "grad_norm": 0.5204850435256958,
      "learning_rate": 2.061073738537635e-05,
      "loss": 2.4133,
      "step": 143
    },
    {
      "epoch": 1.7297297297297298,
      "grad_norm": 0.2928295433521271,
      "learning_rate": 1.9945942635848748e-05,
      "loss": 2.4238,
      "step": 144
    },
    {
      "epoch": 1.7417417417417418,
      "grad_norm": 0.2376270443201065,
      "learning_rate": 1.928936436551661e-05,
      "loss": 2.4666,
      "step": 145
    },
    {
      "epoch": 1.7537537537537538,
      "grad_norm": 0.24902498722076416,
      "learning_rate": 1.8641182076323148e-05,
      "loss": 2.3764,
      "step": 146
    },
    {
      "epoch": 1.7657657657657657,
      "grad_norm": 0.2481139898300171,
      "learning_rate": 1.800157297483417e-05,
      "loss": 2.4266,
      "step": 147
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 0.2662986218929291,
      "learning_rate": 1.7370711923791567e-05,
      "loss": 2.4832,
      "step": 148
    },
    {
      "epoch": 1.7897897897897899,
      "grad_norm": 0.2751111686229706,
      "learning_rate": 1.6748771394307585e-05,
      "loss": 2.4443,
      "step": 149
    },
    {
      "epoch": 1.8018018018018018,
      "grad_norm": 0.29890504479408264,
      "learning_rate": 1.6135921418712956e-05,
      "loss": 2.5217,
      "step": 150
    },
    {
      "epoch": 1.8018018018018018,
      "eval_loss": 2.6516175270080566,
      "eval_runtime": 10.1443,
      "eval_samples_per_second": 13.801,
      "eval_steps_per_second": 3.45,
      "step": 150
    },
    {
      "epoch": 1.8138138138138138,
      "grad_norm": 0.29355207085609436,
      "learning_rate": 1.553232954407171e-05,
      "loss": 2.5488,
      "step": 151
    },
    {
      "epoch": 1.825825825825826,
      "grad_norm": 0.30669480562210083,
      "learning_rate": 1.4938160786375572e-05,
      "loss": 2.523,
      "step": 152
    },
    {
      "epoch": 1.8378378378378377,
      "grad_norm": 0.3132804036140442,
      "learning_rate": 1.435357758543015e-05,
      "loss": 2.5265,
      "step": 153
    },
    {
      "epoch": 1.84984984984985,
      "grad_norm": 0.31633105874061584,
      "learning_rate": 1.3778739760445552e-05,
      "loss": 2.4539,
      "step": 154
    },
    {
      "epoch": 1.8618618618618619,
      "grad_norm": 0.3367692232131958,
      "learning_rate": 1.3213804466343421e-05,
      "loss": 2.6088,
      "step": 155
    },
    {
      "epoch": 1.8738738738738738,
      "grad_norm": 0.36491456627845764,
      "learning_rate": 1.2658926150792322e-05,
      "loss": 2.4123,
      "step": 156
    },
    {
      "epoch": 1.885885885885886,
      "grad_norm": 0.3622000813484192,
      "learning_rate": 1.2114256511983274e-05,
      "loss": 2.5865,
      "step": 157
    },
    {
      "epoch": 1.8978978978978978,
      "grad_norm": 0.38642406463623047,
      "learning_rate": 1.157994445715706e-05,
      "loss": 2.3859,
      "step": 158
    },
    {
      "epoch": 1.90990990990991,
      "grad_norm": 0.4168998599052429,
      "learning_rate": 1.1056136061894384e-05,
      "loss": 2.5424,
      "step": 159
    },
    {
      "epoch": 1.921921921921922,
      "grad_norm": 0.43780872225761414,
      "learning_rate": 1.0542974530180327e-05,
      "loss": 2.4477,
      "step": 160
    },
    {
      "epoch": 1.9339339339339339,
      "grad_norm": 0.4617201089859009,
      "learning_rate": 1.0040600155253765e-05,
      "loss": 2.4271,
      "step": 161
    },
    {
      "epoch": 1.945945945945946,
      "grad_norm": 0.5035412311553955,
      "learning_rate": 9.549150281252633e-06,
      "loss": 2.5789,
      "step": 162
    },
    {
      "epoch": 1.9579579579579578,
      "grad_norm": 0.5353533029556274,
      "learning_rate": 9.068759265665384e-06,
      "loss": 2.5436,
      "step": 163
    },
    {
      "epoch": 1.96996996996997,
      "grad_norm": 0.35868555307388306,
      "learning_rate": 8.599558442598998e-06,
      "loss": 2.6367,
      "step": 164
    },
    {
      "epoch": 1.981981981981982,
      "grad_norm": 0.3028857409954071,
      "learning_rate": 8.141676086873572e-06,
      "loss": 2.5196,
      "step": 165
    },
    {
      "epoch": 1.993993993993994,
      "grad_norm": 0.4011765420436859,
      "learning_rate": 7.695237378953223e-06,
      "loss": 2.5983,
      "step": 166
    },
    {
      "epoch": 2.006006006006006,
      "grad_norm": 1.5913151502609253,
      "learning_rate": 7.260364370723044e-06,
      "loss": 4.1206,
      "step": 167
    },
    {
      "epoch": 2.018018018018018,
      "grad_norm": 0.23217424750328064,
      "learning_rate": 6.837175952121306e-06,
      "loss": 2.4445,
      "step": 168
    },
    {
      "epoch": 2.03003003003003,
      "grad_norm": 0.2567754089832306,
      "learning_rate": 6.425787818636131e-06,
      "loss": 2.4274,
      "step": 169
    },
    {
      "epoch": 2.042042042042042,
      "grad_norm": 0.24294909834861755,
      "learning_rate": 6.026312439675552e-06,
      "loss": 2.3139,
      "step": 170
    },
    {
      "epoch": 2.054054054054054,
      "grad_norm": 0.24930836260318756,
      "learning_rate": 5.6388590278194096e-06,
      "loss": 2.4251,
      "step": 171
    },
    {
      "epoch": 2.066066066066066,
      "grad_norm": 0.26466673612594604,
      "learning_rate": 5.263533508961827e-06,
      "loss": 2.4533,
      "step": 172
    },
    {
      "epoch": 2.078078078078078,
      "grad_norm": 0.27082571387290955,
      "learning_rate": 4.900438493352055e-06,
      "loss": 2.4512,
      "step": 173
    },
    {
      "epoch": 2.09009009009009,
      "grad_norm": 0.2794547975063324,
      "learning_rate": 4.549673247541875e-06,
      "loss": 2.4727,
      "step": 174
    },
    {
      "epoch": 2.1021021021021022,
      "grad_norm": 0.2870752513408661,
      "learning_rate": 4.2113336672471245e-06,
      "loss": 2.4888,
      "step": 175
    },
    {
      "epoch": 2.114114114114114,
      "grad_norm": 0.31091153621673584,
      "learning_rate": 3.885512251130763e-06,
      "loss": 2.4041,
      "step": 176
    },
    {
      "epoch": 2.126126126126126,
      "grad_norm": 0.32006552815437317,
      "learning_rate": 3.5722980755146517e-06,
      "loss": 2.4248,
      "step": 177
    },
    {
      "epoch": 2.138138138138138,
      "grad_norm": 0.3408612012863159,
      "learning_rate": 3.271776770026963e-06,
      "loss": 2.4448,
      "step": 178
    },
    {
      "epoch": 2.15015015015015,
      "grad_norm": 0.33888041973114014,
      "learning_rate": 2.9840304941919415e-06,
      "loss": 2.5826,
      "step": 179
    },
    {
      "epoch": 2.1621621621621623,
      "grad_norm": 0.34623295068740845,
      "learning_rate": 2.7091379149682685e-06,
      "loss": 2.3866,
      "step": 180
    },
    {
      "epoch": 2.174174174174174,
      "grad_norm": 0.37865787744522095,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 2.4662,
      "step": 181
    },
    {
      "epoch": 2.186186186186186,
      "grad_norm": 0.39972391724586487,
      "learning_rate": 2.1982109232821178e-06,
      "loss": 2.3468,
      "step": 182
    },
    {
      "epoch": 2.1981981981981984,
      "grad_norm": 0.4252571165561676,
      "learning_rate": 1.962316193157593e-06,
      "loss": 2.3554,
      "step": 183
    },
    {
      "epoch": 2.21021021021021,
      "grad_norm": 0.46080446243286133,
      "learning_rate": 1.7395544861325718e-06,
      "loss": 2.4643,
      "step": 184
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.4863089621067047,
      "learning_rate": 1.5299867030334814e-06,
      "loss": 2.4737,
      "step": 185
    },
    {
      "epoch": 2.234234234234234,
      "grad_norm": 0.555431604385376,
      "learning_rate": 1.333670137599713e-06,
      "loss": 2.3652,
      "step": 186
    },
    {
      "epoch": 2.2462462462462462,
      "grad_norm": 0.433740496635437,
      "learning_rate": 1.1506584608200367e-06,
      "loss": 2.4827,
      "step": 187
    },
    {
      "epoch": 2.2582582582582584,
      "grad_norm": 0.23512916266918182,
      "learning_rate": 9.810017062595322e-07,
      "loss": 2.465,
      "step": 188
    },
    {
      "epoch": 2.27027027027027,
      "grad_norm": 0.24517232179641724,
      "learning_rate": 8.247462563808817e-07,
      "loss": 2.41,
      "step": 189
    },
    {
      "epoch": 2.2822822822822824,
      "grad_norm": 0.2632448077201843,
      "learning_rate": 6.819348298638839e-07,
      "loss": 2.394,
      "step": 190
    },
    {
      "epoch": 2.294294294294294,
      "grad_norm": 0.26392462849617004,
      "learning_rate": 5.526064699265753e-07,
      "loss": 2.3438,
      "step": 191
    },
    {
      "epoch": 2.3063063063063063,
      "grad_norm": 0.2736310064792633,
      "learning_rate": 4.367965336512403e-07,
      "loss": 2.4787,
      "step": 192
    },
    {
      "epoch": 2.3183183183183185,
      "grad_norm": 0.27571040391921997,
      "learning_rate": 3.3453668231809286e-07,
      "loss": 2.3615,
      "step": 193
    },
    {
      "epoch": 2.33033033033033,
      "grad_norm": 0.29297369718551636,
      "learning_rate": 2.458548727494292e-07,
      "loss": 2.4322,
      "step": 194
    },
    {
      "epoch": 2.3423423423423424,
      "grad_norm": 0.30095139145851135,
      "learning_rate": 1.7077534966650766e-07,
      "loss": 2.4916,
      "step": 195
    },
    {
      "epoch": 2.3543543543543546,
      "grad_norm": 0.32010048627853394,
      "learning_rate": 1.0931863906127327e-07,
      "loss": 2.5235,
      "step": 196
    },
    {
      "epoch": 2.3663663663663663,
      "grad_norm": 0.3276914954185486,
      "learning_rate": 6.150154258476315e-08,
      "loss": 2.3901,
      "step": 197
    },
    {
      "epoch": 2.3783783783783785,
      "grad_norm": 0.3295910358428955,
      "learning_rate": 2.7337132953697554e-08,
      "loss": 2.4143,
      "step": 198
    },
    {
      "epoch": 2.3903903903903903,
      "grad_norm": 0.35997962951660156,
      "learning_rate": 6.834750376549792e-09,
      "loss": 2.4509,
      "step": 199
    },
    {
      "epoch": 2.4024024024024024,
      "grad_norm": 0.37004679441452026,
      "learning_rate": 0.0,
      "loss": 2.4413,
      "step": 200
    },
    {
      "epoch": 2.4024024024024024,
      "eval_loss": 2.653998374938965,
      "eval_runtime": 10.1702,
      "eval_samples_per_second": 13.766,
      "eval_steps_per_second": 3.441,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 2
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.8579962575388672e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}