{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 59,
  "global_step": 296,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.013513513513513514,
      "grad_norm": 0.25159120559692383,
      "learning_rate": 0.0,
      "loss": 1.6708984375,
      "step": 1
    },
    {
      "epoch": 0.02702702702702703,
      "grad_norm": 0.27036958932876587,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.7001953125,
      "step": 2
    },
    {
      "epoch": 0.04054054054054054,
      "grad_norm": 0.2660757899284363,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 1.701171875,
      "step": 3
    },
    {
      "epoch": 0.05405405405405406,
      "grad_norm": 0.21695369482040405,
      "learning_rate": 2e-05,
      "loss": 1.4892578125,
      "step": 4
    },
    {
      "epoch": 0.06756756756756757,
      "grad_norm": 0.23506470024585724,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 1.6328125,
      "step": 5
    },
    {
      "epoch": 0.08108108108108109,
      "grad_norm": 0.18404515087604523,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 1.4560546875,
      "step": 6
    },
    {
      "epoch": 0.0945945945945946,
      "grad_norm": 0.16316190361976624,
      "learning_rate": 4e-05,
      "loss": 1.53515625,
      "step": 7
    },
    {
      "epoch": 0.10810810810810811,
      "grad_norm": 0.09196184575557709,
      "learning_rate": 4.666666666666667e-05,
      "loss": 1.353515625,
      "step": 8
    },
    {
      "epoch": 0.12162162162162163,
      "grad_norm": 0.2868729531764984,
      "learning_rate": 5.333333333333333e-05,
      "loss": 1.30859375,
      "step": 9
    },
    {
      "epoch": 0.13513513513513514,
      "grad_norm": 0.07717549800872803,
      "learning_rate": 6e-05,
      "loss": 1.3876953125,
      "step": 10
    },
    {
      "epoch": 0.14864864864864866,
      "grad_norm": 0.1163763627409935,
      "learning_rate": 6.666666666666667e-05,
      "loss": 1.3916015625,
      "step": 11
    },
    {
      "epoch": 0.16216216216216217,
      "grad_norm": 0.09600003808736801,
      "learning_rate": 7.333333333333333e-05,
      "loss": 1.3173828125,
      "step": 12
    },
    {
      "epoch": 0.17567567567567569,
      "grad_norm": 0.1387435346841812,
      "learning_rate": 8e-05,
      "loss": 1.3359375,
      "step": 13
    },
    {
      "epoch": 0.1891891891891892,
      "grad_norm": 0.10331880301237106,
      "learning_rate": 8.666666666666667e-05,
      "loss": 1.400390625,
      "step": 14
    },
    {
      "epoch": 0.20270270270270271,
      "grad_norm": 0.09152160584926605,
      "learning_rate": 9.333333333333334e-05,
      "loss": 1.3564453125,
      "step": 15
    },
    {
      "epoch": 0.21621621621621623,
      "grad_norm": 0.05813554674386978,
      "learning_rate": 0.0001,
      "loss": 1.341796875,
      "step": 16
    },
    {
      "epoch": 0.22972972972972974,
      "grad_norm": 0.06166822463274002,
      "learning_rate": 9.999718767763875e-05,
      "loss": 1.31103515625,
      "step": 17
    },
    {
      "epoch": 0.24324324324324326,
      "grad_norm": 0.06363847851753235,
      "learning_rate": 9.998875106207304e-05,
      "loss": 1.3427734375,
      "step": 18
    },
    {
      "epoch": 0.25675675675675674,
      "grad_norm": 0.06172942742705345,
      "learning_rate": 9.997469120781326e-05,
      "loss": 1.3505859375,
      "step": 19
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 0.05632505565881729,
      "learning_rate": 9.995500987223015e-05,
      "loss": 1.2841796875,
      "step": 20
    },
    {
      "epoch": 0.28378378378378377,
      "grad_norm": 0.054518990218639374,
      "learning_rate": 9.992970951533529e-05,
      "loss": 1.09326171875,
      "step": 21
    },
    {
      "epoch": 0.2972972972972973,
      "grad_norm": 0.05610249936580658,
      "learning_rate": 9.989879329947355e-05,
      "loss": 1.25439453125,
      "step": 22
    },
    {
      "epoch": 0.3108108108108108,
      "grad_norm": 0.04417848587036133,
      "learning_rate": 9.98622650889278e-05,
      "loss": 1.3037109375,
      "step": 23
    },
    {
      "epoch": 0.32432432432432434,
      "grad_norm": 0.044958099722862244,
      "learning_rate": 9.982012944943602e-05,
      "loss": 1.248046875,
      "step": 24
    },
    {
      "epoch": 0.33783783783783783,
      "grad_norm": 0.03528568893671036,
      "learning_rate": 9.977239164762041e-05,
      "loss": 1.29296875,
      "step": 25
    },
    {
      "epoch": 0.35135135135135137,
      "grad_norm": 0.038594186305999756,
      "learning_rate": 9.971905765032936e-05,
      "loss": 1.287109375,
      "step": 26
    },
    {
      "epoch": 0.36486486486486486,
      "grad_norm": 0.03776711970567703,
      "learning_rate": 9.966013412389144e-05,
      "loss": 1.265625,
      "step": 27
    },
    {
      "epoch": 0.3783783783783784,
      "grad_norm": 0.03557330369949341,
      "learning_rate": 9.959562843328224e-05,
      "loss": 1.296875,
      "step": 28
    },
    {
      "epoch": 0.3918918918918919,
      "grad_norm": 0.03854518011212349,
      "learning_rate": 9.952554864120383e-05,
      "loss": 1.24755859375,
      "step": 29
    },
    {
      "epoch": 0.40540540540540543,
      "grad_norm": 0.03412532061338425,
      "learning_rate": 9.944990350707692e-05,
      "loss": 1.2314453125,
      "step": 30
    },
    {
      "epoch": 0.4189189189189189,
      "grad_norm": 0.0344543382525444,
      "learning_rate": 9.936870248594607e-05,
      "loss": 1.26171875,
      "step": 31
    },
    {
      "epoch": 0.43243243243243246,
      "grad_norm": 0.02789556235074997,
      "learning_rate": 9.928195572729781e-05,
      "loss": 1.2763671875,
      "step": 32
    },
    {
      "epoch": 0.44594594594594594,
      "grad_norm": 0.037205733358860016,
      "learning_rate": 9.918967407379212e-05,
      "loss": 1.2890625,
      "step": 33
    },
    {
      "epoch": 0.4594594594594595,
      "grad_norm": 0.030634207651019096,
      "learning_rate": 9.909186905990712e-05,
      "loss": 1.26171875,
      "step": 34
    },
    {
      "epoch": 0.47297297297297297,
      "grad_norm": 0.029325243085622787,
      "learning_rate": 9.898855291049736e-05,
      "loss": 1.248046875,
      "step": 35
    },
    {
      "epoch": 0.4864864864864865,
      "grad_norm": 0.027585148811340332,
      "learning_rate": 9.887973853926582e-05,
      "loss": 1.0802001953125,
      "step": 36
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.03394173085689545,
      "learning_rate": 9.876543954714982e-05,
      "loss": 1.2412109375,
      "step": 37
    },
    {
      "epoch": 0.5135135135135135,
      "grad_norm": 0.032742470502853394,
      "learning_rate": 9.8645670220621e-05,
      "loss": 1.21435546875,
      "step": 38
    },
    {
      "epoch": 0.527027027027027,
      "grad_norm": 0.030860472470521927,
      "learning_rate": 9.852044552989955e-05,
      "loss": 1.2734375,
      "step": 39
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 0.02754375897347927,
      "learning_rate": 9.83897811270832e-05,
      "loss": 1.1982421875,
      "step": 40
    },
    {
      "epoch": 0.5540540540540541,
      "grad_norm": 0.02955382689833641,
      "learning_rate": 9.825369334419065e-05,
      "loss": 1.21875,
      "step": 41
    },
    {
      "epoch": 0.5675675675675675,
      "grad_norm": 0.02716575190424919,
      "learning_rate": 9.811219919112037e-05,
      "loss": 1.2470703125,
      "step": 42
    },
    {
      "epoch": 0.581081081081081,
      "grad_norm": 0.027045253664255142,
      "learning_rate": 9.796531635352438e-05,
      "loss": 0.981689453125,
      "step": 43
    },
    {
      "epoch": 0.5945945945945946,
      "grad_norm": 0.03138129040598869,
      "learning_rate": 9.781306319059777e-05,
      "loss": 1.232421875,
      "step": 44
    },
    {
      "epoch": 0.6081081081081081,
      "grad_norm": 0.026743337512016296,
      "learning_rate": 9.765545873278377e-05,
      "loss": 1.17138671875,
      "step": 45
    },
    {
      "epoch": 0.6216216216216216,
      "grad_norm": 0.02658187970519066,
      "learning_rate": 9.74925226793954e-05,
      "loss": 1.220703125,
      "step": 46
    },
    {
      "epoch": 0.6351351351351351,
      "grad_norm": 0.026990732178092003,
      "learning_rate": 9.73242753961529e-05,
      "loss": 1.2255859375,
      "step": 47
    },
    {
      "epoch": 0.6486486486486487,
      "grad_norm": 0.025836464017629623,
      "learning_rate": 9.715073791263837e-05,
      "loss": 1.201171875,
      "step": 48
    },
    {
      "epoch": 0.6621621621621622,
      "grad_norm": 0.023755870759487152,
      "learning_rate": 9.697193191966714e-05,
      "loss": 1.1962890625,
      "step": 49
    },
    {
      "epoch": 0.6756756756756757,
      "grad_norm": 0.025323517620563507,
      "learning_rate": 9.678787976657668e-05,
      "loss": 1.21484375,
      "step": 50
    },
    {
      "epoch": 0.6891891891891891,
      "grad_norm": 0.025310220196843147,
      "learning_rate": 9.659860445843301e-05,
      "loss": 1.1865234375,
      "step": 51
    },
    {
      "epoch": 0.7027027027027027,
      "grad_norm": 0.024962972849607468,
      "learning_rate": 9.640412965315528e-05,
      "loss": 1.1416015625,
      "step": 52
    },
    {
      "epoch": 0.7162162162162162,
      "grad_norm": 0.026338156312704086,
      "learning_rate": 9.62044796585588e-05,
      "loss": 1.21875,
      "step": 53
    },
    {
      "epoch": 0.7297297297297297,
      "grad_norm": 0.026672903448343277,
      "learning_rate": 9.599967942931664e-05,
      "loss": 1.2021484375,
      "step": 54
    },
    {
      "epoch": 0.7432432432432432,
      "grad_norm": 0.02600967511534691,
      "learning_rate": 9.578975456384056e-05,
      "loss": 1.13671875,
      "step": 55
    },
    {
      "epoch": 0.7567567567567568,
      "grad_norm": 0.026378018781542778,
      "learning_rate": 9.557473130108133e-05,
      "loss": 1.228515625,
      "step": 56
    },
    {
      "epoch": 0.7702702702702703,
      "grad_norm": 0.03431445732712746,
      "learning_rate": 9.53546365172492e-05,
      "loss": 1.03094482421875,
      "step": 57
    },
    {
      "epoch": 0.7837837837837838,
      "grad_norm": 0.026858845725655556,
      "learning_rate": 9.51294977224545e-05,
      "loss": 1.224609375,
      "step": 58
    },
    {
      "epoch": 0.7972972972972973,
      "grad_norm": 0.02474498376250267,
      "learning_rate": 9.489934305726908e-05,
      "loss": 1.1923828125,
      "step": 59
    },
    {
      "epoch": 0.8108108108108109,
      "grad_norm": 0.024861522018909454,
      "learning_rate": 9.4664201289209e-05,
      "loss": 1.15380859375,
      "step": 60
    },
    {
      "epoch": 0.8108108108108109,
      "eval_loss": 1.207275390625,
      "eval_runtime": 5.5894,
      "eval_samples_per_second": 0.716,
      "eval_steps_per_second": 0.179,
      "step": 60
    },
    {
      "epoch": 0.8243243243243243,
      "grad_norm": 0.02610987238585949,
      "learning_rate": 9.442410180913883e-05,
      "loss": 1.1533203125,
      "step": 61
    },
    {
      "epoch": 0.8378378378378378,
      "grad_norm": 0.025056414306163788,
      "learning_rate": 9.417907462759797e-05,
      "loss": 1.19140625,
      "step": 62
    },
    {
      "epoch": 0.8513513513513513,
      "grad_norm": 0.024142485111951828,
      "learning_rate": 9.392915037104961e-05,
      "loss": 1.0244140625,
      "step": 63
    },
    {
      "epoch": 0.8648648648648649,
      "grad_norm": 0.027491150423884392,
      "learning_rate": 9.367436027805267e-05,
      "loss": 1.1982421875,
      "step": 64
    },
    {
      "epoch": 0.8783783783783784,
      "grad_norm": 0.026476306840777397,
      "learning_rate": 9.341473619535714e-05,
      "loss": 1.058837890625,
      "step": 65
    },
    {
      "epoch": 0.8918918918918919,
      "grad_norm": 0.025533553212881088,
      "learning_rate": 9.315031057392366e-05,
      "loss": 1.1708984375,
      "step": 66
    },
    {
      "epoch": 0.9054054054054054,
      "grad_norm": 0.026743927970528603,
      "learning_rate": 9.288111646486724e-05,
      "loss": 1.1669921875,
      "step": 67
    },
    {
      "epoch": 0.918918918918919,
      "grad_norm": 0.02561163157224655,
      "learning_rate": 9.260718751532623e-05,
      "loss": 1.1796875,
      "step": 68
    },
    {
      "epoch": 0.9324324324324325,
      "grad_norm": 0.027323313057422638,
      "learning_rate": 9.232855796425659e-05,
      "loss": 1.2353515625,
      "step": 69
    },
    {
      "epoch": 0.9459459459459459,
      "grad_norm": 0.027578480541706085,
      "learning_rate": 9.204526263815245e-05,
      "loss": 1.1943359375,
      "step": 70
    },
    {
      "epoch": 0.9594594594594594,
      "grad_norm": 0.02868049219250679,
      "learning_rate": 9.175733694669293e-05,
      "loss": 1.2158203125,
      "step": 71
    },
    {
      "epoch": 0.972972972972973,
      "grad_norm": 0.027299581095576286,
      "learning_rate": 9.146481687831627e-05,
      "loss": 1.2080078125,
      "step": 72
    },
    {
      "epoch": 0.9864864864864865,
      "grad_norm": 0.02587740309536457,
      "learning_rate": 9.116773899572155e-05,
      "loss": 1.1787109375,
      "step": 73
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.02506023645401001,
      "learning_rate": 9.086614043129865e-05,
      "loss": 1.12841796875,
      "step": 74
    },
    {
      "epoch": 1.0135135135135136,
      "grad_norm": 0.02797277271747589,
      "learning_rate": 9.0560058882487e-05,
      "loss": 1.2177734375,
      "step": 75
    },
    {
      "epoch": 1.027027027027027,
      "grad_norm": 0.029189402237534523,
      "learning_rate": 9.024953260706366e-05,
      "loss": 1.181640625,
      "step": 76
    },
    {
      "epoch": 1.0405405405405406,
      "grad_norm": 0.02751336619257927,
      "learning_rate": 8.993460041836143e-05,
      "loss": 1.19921875,
      "step": 77
    },
    {
      "epoch": 1.054054054054054,
      "grad_norm": 0.02728724479675293,
      "learning_rate": 8.96153016804175e-05,
      "loss": 1.125,
      "step": 78
    },
    {
      "epoch": 1.0675675675675675,
      "grad_norm": 0.027950149029493332,
      "learning_rate": 8.929167630305323e-05,
      "loss": 1.181640625,
      "step": 79
    },
    {
      "epoch": 1.0810810810810811,
      "grad_norm": 0.027291124686598778,
      "learning_rate": 8.896376473688574e-05,
      "loss": 1.115234375,
      "step": 80
    },
    {
      "epoch": 1.0945945945945945,
      "grad_norm": 0.02762507274746895,
      "learning_rate": 8.863160796827193e-05,
      "loss": 1.181640625,
      "step": 81
    },
    {
      "epoch": 1.1081081081081081,
      "grad_norm": 0.028306778520345688,
      "learning_rate": 8.829524751418549e-05,
      "loss": 1.10302734375,
      "step": 82
    },
    {
      "epoch": 1.1216216216216217,
      "grad_norm": 0.13466431200504303,
      "learning_rate": 8.79547254170276e-05,
      "loss": 1.069580078125,
      "step": 83
    },
    {
      "epoch": 1.135135135135135,
      "grad_norm": 0.03159444034099579,
      "learning_rate": 8.761008423937193e-05,
      "loss": 1.158203125,
      "step": 84
    },
    {
      "epoch": 1.1486486486486487,
      "grad_norm": 0.027644814923405647,
      "learning_rate": 8.726136705864477e-05,
      "loss": 1.1767578125,
      "step": 85
    },
    {
      "epoch": 1.1621621621621623,
      "grad_norm": 0.029982680454850197,
      "learning_rate": 8.690861746174053e-05,
      "loss": 1.1318359375,
      "step": 86
    },
    {
      "epoch": 1.1756756756756757,
      "grad_norm": 0.03274378925561905,
      "learning_rate": 8.655187953957385e-05,
      "loss": 1.13330078125,
      "step": 87
    },
    {
      "epoch": 1.1891891891891893,
      "grad_norm": 0.03125821053981781,
      "learning_rate": 8.619119788156857e-05,
      "loss": 1.201171875,
      "step": 88
    },
    {
      "epoch": 1.2027027027027026,
      "grad_norm": 0.03011920116841793,
      "learning_rate": 8.58266175700843e-05,
      "loss": 1.1572265625,
      "step": 89
    },
    {
      "epoch": 1.2162162162162162,
      "grad_norm": 0.028389113023877144,
      "learning_rate": 8.545818417478163e-05,
      "loss": 1.1669921875,
      "step": 90
    },
    {
      "epoch": 1.2297297297297298,
      "grad_norm": 0.031037267297506332,
      "learning_rate": 8.508594374692615e-05,
      "loss": 1.130859375,
      "step": 91
    },
    {
      "epoch": 1.2432432432432432,
      "grad_norm": 0.028420383110642433,
      "learning_rate": 8.470994281363247e-05,
      "loss": 1.17578125,
      "step": 92
    },
    {
      "epoch": 1.2567567567567568,
      "grad_norm": 0.03020644746720791,
      "learning_rate": 8.433022837204873e-05,
      "loss": 1.193359375,
      "step": 93
    },
    {
      "epoch": 1.2702702702702702,
      "grad_norm": 0.029988085851073265,
      "learning_rate": 8.394684788348225e-05,
      "loss": 1.14111328125,
      "step": 94
    },
    {
      "epoch": 1.2837837837837838,
      "grad_norm": 0.04022596776485443,
      "learning_rate": 8.355984926746731e-05,
      "loss": 0.974365234375,
      "step": 95
    },
    {
      "epoch": 1.2972972972972974,
      "grad_norm": 0.02990506775677204,
      "learning_rate": 8.316928089577547e-05,
      "loss": 1.115234375,
      "step": 96
    },
    {
      "epoch": 1.3108108108108107,
      "grad_norm": 0.029547693207859993,
      "learning_rate": 8.277519158636971e-05,
      "loss": 1.16796875,
      "step": 97
    },
    {
      "epoch": 1.3243243243243243,
      "grad_norm": 0.029393065720796585,
      "learning_rate": 8.237763059730232e-05,
      "loss": 1.11328125,
      "step": 98
    },
    {
      "epoch": 1.3378378378378377,
      "grad_norm": 0.027526037767529488,
      "learning_rate": 8.197664762055816e-05,
      "loss": 1.162109375,
      "step": 99
    },
    {
      "epoch": 1.3513513513513513,
      "grad_norm": 0.03041381575167179,
      "learning_rate": 8.157229277584356e-05,
      "loss": 1.162109375,
      "step": 100
    },
    {
      "epoch": 1.364864864864865,
      "grad_norm": 0.027837971225380898,
      "learning_rate": 8.116461660432168e-05,
      "loss": 1.1455078125,
      "step": 101
    },
    {
      "epoch": 1.3783783783783785,
      "grad_norm": 0.030750012025237083,
      "learning_rate": 8.075367006229537e-05,
      "loss": 1.169921875,
      "step": 102
    },
    {
      "epoch": 1.3918918918918919,
      "grad_norm": 0.02921544760465622,
      "learning_rate": 8.033950451483788e-05,
      "loss": 1.126953125,
      "step": 103
    },
    {
      "epoch": 1.4054054054054055,
      "grad_norm": 0.027677197009325027,
      "learning_rate": 7.992217172937283e-05,
      "loss": 1.1142578125,
      "step": 104
    },
    {
      "epoch": 1.4189189189189189,
      "grad_norm": 0.029318837448954582,
      "learning_rate": 7.950172386920353e-05,
      "loss": 1.146484375,
      "step": 105
    },
    {
      "epoch": 1.4324324324324325,
      "grad_norm": 0.0294735599309206,
      "learning_rate": 7.907821348699303e-05,
      "loss": 1.16357421875,
      "step": 106
    },
    {
      "epoch": 1.445945945945946,
      "grad_norm": 0.03245671093463898,
      "learning_rate": 7.86516935181955e-05,
      "loss": 1.1728515625,
      "step": 107
    },
    {
      "epoch": 1.4594594594594594,
      "grad_norm": 0.02917313016951084,
      "learning_rate": 7.82222172744396e-05,
      "loss": 1.15625,
      "step": 108
    },
    {
      "epoch": 1.472972972972973,
      "grad_norm": 0.03336160629987717,
      "learning_rate": 7.778983843686507e-05,
      "loss": 1.142578125,
      "step": 109
    },
    {
      "epoch": 1.4864864864864864,
      "grad_norm": 0.028797032311558723,
      "learning_rate": 7.735461104941298e-05,
      "loss": 0.987060546875,
      "step": 110
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.034539833664894104,
      "learning_rate": 7.691658951207057e-05,
      "loss": 1.1357421875,
      "step": 111
    },
    {
      "epoch": 1.5135135135135136,
      "grad_norm": 0.031068462878465652,
      "learning_rate": 7.647582857407185e-05,
      "loss": 1.10888671875,
      "step": 112
    },
    {
      "epoch": 1.527027027027027,
      "grad_norm": 0.03172864392399788,
      "learning_rate": 7.60323833270542e-05,
      "loss": 1.169921875,
      "step": 113
    },
    {
      "epoch": 1.5405405405405406,
      "grad_norm": 0.030864330008625984,
      "learning_rate": 7.558630919817252e-05,
      "loss": 1.10546875,
      "step": 114
    },
    {
      "epoch": 1.554054054054054,
      "grad_norm": 0.03297742083668709,
      "learning_rate": 7.51376619431711e-05,
      "loss": 1.130859375,
      "step": 115
    },
    {
      "epoch": 1.5675675675675675,
      "grad_norm": 0.03478364273905754,
      "learning_rate": 7.468649763941473e-05,
      "loss": 1.150390625,
      "step": 116
    },
    {
      "epoch": 1.5810810810810811,
      "grad_norm": 0.027595333755016327,
      "learning_rate": 7.423287267887941e-05,
      "loss": 0.90155029296875,
      "step": 117
    },
    {
      "epoch": 1.5945945945945947,
      "grad_norm": 0.041680410504341125,
      "learning_rate": 7.377684376110383e-05,
      "loss": 1.1328125,
      "step": 118
    },
    {
      "epoch": 1.6081081081081081,
      "grad_norm": 0.03135593980550766,
      "learning_rate": 7.331846788610232e-05,
      "loss": 1.0849609375,
      "step": 119
    },
    {
      "epoch": 1.6081081081081081,
      "eval_loss": 1.16015625,
      "eval_runtime": 5.5781,
      "eval_samples_per_second": 0.717,
      "eval_steps_per_second": 0.179,
      "step": 119
    },
    {
      "epoch": 1.6216216216216215,
      "grad_norm": 0.0382126159965992,
      "learning_rate": 7.285780234724036e-05,
      "loss": 1.1318359375,
      "step": 120
    },
    {
      "epoch": 1.635135135135135,
      "grad_norm": 0.03366832807660103,
      "learning_rate": 7.239490472407333e-05,
      "loss": 1.138671875,
      "step": 121
    },
    {
      "epoch": 1.6486486486486487,
      "grad_norm": 0.040457215160131454,
      "learning_rate": 7.19298328751495e-05,
      "loss": 1.123046875,
      "step": 122
    },
    {
      "epoch": 1.6621621621621623,
      "grad_norm": 0.032831333577632904,
      "learning_rate": 7.146264493077827e-05,
      "loss": 1.11328125,
      "step": 123
    },
    {
      "epoch": 1.6756756756756757,
      "grad_norm": 0.037415146827697754,
      "learning_rate": 7.099339928576416e-05,
      "loss": 1.12890625,
      "step": 124
    },
    {
      "epoch": 1.689189189189189,
      "grad_norm": 0.03626786917448044,
      "learning_rate": 7.05221545921081e-05,
      "loss": 1.107421875,
      "step": 125
    },
    {
      "epoch": 1.7027027027027026,
      "grad_norm": 0.03040171228349209,
      "learning_rate": 7.004896975167629e-05,
      "loss": 1.06689453125,
      "step": 126
    },
    {
      "epoch": 1.7162162162162162,
      "grad_norm": 0.04583225026726723,
      "learning_rate": 6.957390390883796e-05,
      "loss": 1.142578125,
      "step": 127
    },
    {
      "epoch": 1.7297297297297298,
      "grad_norm": 0.030556079000234604,
      "learning_rate": 6.909701644307282e-05,
      "loss": 1.12109375,
      "step": 128
    },
    {
      "epoch": 1.7432432432432432,
      "grad_norm": 0.0349075049161911,
      "learning_rate": 6.861836696154901e-05,
      "loss": 1.0615234375,
      "step": 129
    },
    {
      "epoch": 1.7567567567567568,
      "grad_norm": 0.038611482828855515,
      "learning_rate": 6.813801529167276e-05,
      "loss": 1.150390625,
      "step": 130
    },
    {
      "epoch": 1.7702702702702702,
      "grad_norm": 0.028310686349868774,
      "learning_rate": 6.765602147361037e-05,
      "loss": 0.97186279296875,
      "step": 131
    },
    {
      "epoch": 1.7837837837837838,
      "grad_norm": 0.03528667613863945,
      "learning_rate": 6.71724457527838e-05,
      "loss": 1.14453125,
      "step": 132
    },
    {
      "epoch": 1.7972972972972974,
      "grad_norm": 0.030709421262145042,
      "learning_rate": 6.668734857234026e-05,
      "loss": 1.1171875,
      "step": 133
    },
    {
      "epoch": 1.810810810810811,
      "grad_norm": 0.031253691762685776,
      "learning_rate": 6.620079056559744e-05,
      "loss": 1.08203125,
      "step": 134
    },
    {
      "epoch": 1.8243243243243243,
      "grad_norm": 0.03429649397730827,
      "learning_rate": 6.571283254846476e-05,
      "loss": 1.08251953125,
      "step": 135
    },
    {
      "epoch": 1.8378378378378377,
      "grad_norm": 0.032073523849248886,
      "learning_rate": 6.522353551184194e-05,
      "loss": 1.1181640625,
      "step": 136
    },
    {
      "epoch": 1.8513513513513513,
      "grad_norm": 0.02924853004515171,
      "learning_rate": 6.473296061399551e-05,
      "loss": 0.9619140625,
      "step": 137
    },
    {
      "epoch": 1.864864864864865,
      "grad_norm": 0.031025685369968414,
      "learning_rate": 6.424116917291458e-05,
      "loss": 1.1279296875,
      "step": 138
    },
    {
      "epoch": 1.8783783783783785,
      "grad_norm": 0.030699903145432472,
      "learning_rate": 6.374822265864659e-05,
      "loss": 1.002197265625,
      "step": 139
    },
    {
      "epoch": 1.8918918918918919,
      "grad_norm": 0.0317017026245594,
      "learning_rate": 6.325418268561397e-05,
      "loss": 1.10498046875,
      "step": 140
    },
    {
      "epoch": 1.9054054054054053,
      "grad_norm": 0.02943561226129532,
      "learning_rate": 6.275911100491286e-05,
      "loss": 1.09619140625,
      "step": 141
    },
    {
      "epoch": 1.9189189189189189,
      "grad_norm": 0.030847109854221344,
      "learning_rate": 6.226306949659475e-05,
      "loss": 1.109375,
      "step": 142
    },
    {
      "epoch": 1.9324324324324325,
      "grad_norm": 0.035015229135751724,
      "learning_rate": 6.17661201619319e-05,
      "loss": 1.1630859375,
      "step": 143
    },
    {
      "epoch": 1.945945945945946,
      "grad_norm": 0.032277610152959824,
      "learning_rate": 6.12683251156677e-05,
      "loss": 1.13037109375,
      "step": 144
    },
    {
      "epoch": 1.9594594594594594,
      "grad_norm": 0.03431333601474762,
      "learning_rate": 6.0769746578252806e-05,
      "loss": 1.14404296875,
      "step": 145
    },
    {
      "epoch": 1.972972972972973,
      "grad_norm": 0.03228043019771576,
      "learning_rate": 6.0270446868068107e-05,
      "loss": 1.1474609375,
      "step": 146
    },
    {
      "epoch": 1.9864864864864864,
      "grad_norm": 0.032071154564619064,
      "learning_rate": 5.9770488393635374e-05,
      "loss": 1.1123046875,
      "step": 147
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.03132776543498039,
      "learning_rate": 5.926993364581669e-05,
      "loss": 1.06494140625,
      "step": 148
    },
    {
      "epoch": 2.0135135135135136,
      "grad_norm": 0.034268446266651154,
      "learning_rate": 5.876884519000364e-05,
      "loss": 1.14794921875,
      "step": 149
    },
    {
      "epoch": 2.027027027027027,
      "grad_norm": 0.035278040915727615,
      "learning_rate": 5.8267285658297075e-05,
      "loss": 1.1162109375,
      "step": 150
    },
    {
      "epoch": 2.0405405405405403,
      "grad_norm": 0.03631846606731415,
      "learning_rate": 5.776531774167865e-05,
      "loss": 1.134765625,
      "step": 151
    },
    {
      "epoch": 2.054054054054054,
      "grad_norm": 0.03242409974336624,
      "learning_rate": 5.726300418217484e-05,
      "loss": 1.06982421875,
      "step": 152
    },
    {
      "epoch": 2.0675675675675675,
      "grad_norm": 0.035991113632917404,
      "learning_rate": 5.6760407765014786e-05,
      "loss": 1.1220703125,
      "step": 153
    },
    {
      "epoch": 2.081081081081081,
      "grad_norm": 0.03287598490715027,
      "learning_rate": 5.625759131078262e-05,
      "loss": 1.05859375,
      "step": 154
    },
    {
      "epoch": 2.0945945945945947,
      "grad_norm": 0.03510114178061485,
      "learning_rate": 5.5754617667565364e-05,
      "loss": 1.12109375,
      "step": 155
    },
    {
      "epoch": 2.108108108108108,
      "grad_norm": 0.0326821394264698,
      "learning_rate": 5.5251549703097416e-05,
      "loss": 1.04541015625,
      "step": 156
    },
    {
      "epoch": 2.1216216216216215,
      "grad_norm": 0.029910115525126457,
      "learning_rate": 5.4748450296902574e-05,
      "loss": 1.016845703125,
      "step": 157
    },
    {
      "epoch": 2.135135135135135,
      "grad_norm": 0.03820684179663658,
      "learning_rate": 5.424538233243463e-05,
      "loss": 1.10498046875,
      "step": 158
    },
    {
      "epoch": 2.1486486486486487,
      "grad_norm": 0.03222264349460602,
      "learning_rate": 5.3742408689217385e-05,
      "loss": 1.1181640625,
      "step": 159
    },
    {
      "epoch": 2.1621621621621623,
      "grad_norm": 0.033516936004161835,
      "learning_rate": 5.3239592234985224e-05,
      "loss": 1.076171875,
      "step": 160
    },
    {
      "epoch": 2.175675675675676,
      "grad_norm": 0.04023905470967293,
      "learning_rate": 5.2736995817825184e-05,
      "loss": 1.07861328125,
      "step": 161
    },
    {
      "epoch": 2.189189189189189,
      "grad_norm": 0.034855931997299194,
      "learning_rate": 5.2234682258321366e-05,
      "loss": 1.14453125,
      "step": 162
    },
    {
      "epoch": 2.2027027027027026,
      "grad_norm": 0.03789949789643288,
      "learning_rate": 5.173271434170293e-05,
      "loss": 1.0986328125,
      "step": 163
    },
    {
      "epoch": 2.2162162162162162,
      "grad_norm": 0.034364763647317886,
      "learning_rate": 5.1231154809996377e-05,
      "loss": 1.1123046875,
      "step": 164
    },
    {
      "epoch": 2.22972972972973,
      "grad_norm": 0.03597622364759445,
      "learning_rate": 5.073006635418333e-05,
      "loss": 1.076171875,
      "step": 165
    },
    {
      "epoch": 2.2432432432432434,
      "grad_norm": 0.03733183816075325,
      "learning_rate": 5.022951160636466e-05,
      "loss": 1.12109375,
      "step": 166
    },
    {
      "epoch": 2.2567567567567566,
      "grad_norm": 0.03333941474556923,
      "learning_rate": 4.97295531319319e-05,
      "loss": 1.13671875,
      "step": 167
    },
    {
      "epoch": 2.27027027027027,
      "grad_norm": 0.03879544511437416,
      "learning_rate": 4.9230253421747184e-05,
      "loss": 1.0888671875,
      "step": 168
    },
    {
      "epoch": 2.2837837837837838,
      "grad_norm": 0.0365670882165432,
      "learning_rate": 4.873167488433231e-05,
      "loss": 0.932861328125,
      "step": 169
    },
    {
      "epoch": 2.2972972972972974,
      "grad_norm": 0.0319230817258358,
      "learning_rate": 4.823387983806811e-05,
      "loss": 1.064453125,
      "step": 170
    },
    {
      "epoch": 2.310810810810811,
      "grad_norm": 0.03506716713309288,
      "learning_rate": 4.773693050340526e-05,
      "loss": 1.1171875,
      "step": 171
    },
    {
      "epoch": 2.3243243243243246,
      "grad_norm": 0.03766872361302376,
      "learning_rate": 4.724088899508715e-05,
      "loss": 1.0615234375,
      "step": 172
    },
    {
      "epoch": 2.3378378378378377,
      "grad_norm": 0.03584901615977287,
      "learning_rate": 4.674581731438605e-05,
      "loss": 1.1103515625,
      "step": 173
    },
    {
      "epoch": 2.3513513513513513,
      "grad_norm": 0.0347881019115448,
      "learning_rate": 4.625177734135343e-05,
      "loss": 1.11083984375,
      "step": 174
    },
    {
      "epoch": 2.364864864864865,
      "grad_norm": 0.0336078405380249,
      "learning_rate": 4.575883082708543e-05,
      "loss": 1.09375,
      "step": 175
    },
    {
      "epoch": 2.3783783783783785,
      "grad_norm": 0.038653627038002014,
      "learning_rate": 4.5267039386004493e-05,
      "loss": 1.11962890625,
      "step": 176
    },
    {
      "epoch": 2.391891891891892,
      "grad_norm": 0.03559928387403488,
      "learning_rate": 4.4776464488158066e-05,
      "loss": 1.076171875,
      "step": 177
    },
    {
      "epoch": 2.4054054054054053,
      "grad_norm": 0.03515911474823952,
      "learning_rate": 4.428716745153524e-05,
      "loss": 1.06787109375,
      "step": 178
    },
    {
      "epoch": 2.4054054054054053,
      "eval_loss": 1.1416015625,
      "eval_runtime": 5.5819,
      "eval_samples_per_second": 0.717,
      "eval_steps_per_second": 0.179,
      "step": 178
    },
    {
      "epoch": 2.418918918918919,
      "grad_norm": 0.034339118748903275,
      "learning_rate": 4.3799209434402565e-05,
      "loss": 1.099609375,
      "step": 179
    },
    {
      "epoch": 2.4324324324324325,
      "grad_norm": 0.03924519568681717,
      "learning_rate": 4.331265142765975e-05,
      "loss": 1.11328125,
      "step": 180
    },
    {
      "epoch": 2.445945945945946,
      "grad_norm": 0.03959480673074722,
      "learning_rate": 4.2827554247216205e-05,
      "loss": 1.125,
      "step": 181
    },
    {
      "epoch": 2.4594594594594597,
      "grad_norm": 0.03376571461558342,
      "learning_rate": 4.234397852638964e-05,
      "loss": 1.1064453125,
      "step": 182
    },
    {
      "epoch": 2.472972972972973,
      "grad_norm": 0.037850137799978256,
      "learning_rate": 4.186198470832726e-05,
      "loss": 1.0966796875,
      "step": 183
    },
    {
      "epoch": 2.4864864864864864,
      "grad_norm": 0.03709284961223602,
      "learning_rate": 4.138163303845101e-05,
      "loss": 0.94793701171875,
      "step": 184
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.03276029974222183,
      "learning_rate": 4.0902983556927197e-05,
      "loss": 1.0888671875,
      "step": 185
    },
    {
      "epoch": 2.5135135135135136,
      "grad_norm": 0.04109270125627518,
      "learning_rate": 4.042609609116203e-05,
      "loss": 1.064453125,
      "step": 186
    },
    {
      "epoch": 2.527027027027027,
      "grad_norm": 0.03824342414736748,
      "learning_rate": 3.995103024832371e-05,
      "loss": 1.1240234375,
      "step": 187
    },
    {
      "epoch": 2.5405405405405403,
      "grad_norm": 0.0359465628862381,
      "learning_rate": 3.9477845407891914e-05,
      "loss": 1.064453125,
      "step": 188
    },
    {
      "epoch": 2.554054054054054,
      "grad_norm": 0.033167410641908646,
      "learning_rate": 3.9006600714235854e-05,
      "loss": 1.08935546875,
      "step": 189
    },
    {
      "epoch": 2.5675675675675675,
      "grad_norm": 0.03669448569417,
      "learning_rate": 3.853735506922174e-05,
      "loss": 1.10693359375,
      "step": 190
    },
    {
      "epoch": 2.581081081081081,
      "grad_norm": 0.03123011440038681,
      "learning_rate": 3.80701671248505e-05,
      "loss": 0.862518310546875,
      "step": 191
    },
    {
      "epoch": 2.5945945945945947,
      "grad_norm": 0.03806449845433235,
      "learning_rate": 3.760509527592669e-05,
      "loss": 1.0869140625,
      "step": 192
    },
    {
      "epoch": 2.608108108108108,
      "grad_norm": 0.03215745836496353,
      "learning_rate": 3.7142197652759656e-05,
      "loss": 1.041015625,
      "step": 193
    },
    {
      "epoch": 2.6216216216216215,
      "grad_norm": 0.03863179311156273,
      "learning_rate": 3.66815321138977e-05,
      "loss": 1.087890625,
      "step": 194
    },
    {
      "epoch": 2.635135135135135,
      "grad_norm": 0.0397212877869606,
      "learning_rate": 3.622315623889619e-05,
      "loss": 1.09619140625,
      "step": 195
    },
    {
      "epoch": 2.6486486486486487,
      "grad_norm": 0.03344908356666565,
      "learning_rate": 3.57671273211206e-05,
      "loss": 1.0771484375,
      "step": 196
    },
    {
      "epoch": 2.6621621621621623,
      "grad_norm": 0.0328289158642292,
      "learning_rate": 3.5313502360585285e-05,
      "loss": 1.07275390625,
      "step": 197
    },
    {
      "epoch": 2.6756756756756754,
      "grad_norm": 0.03553028404712677,
      "learning_rate": 3.486233805682892e-05,
      "loss": 1.087890625,
      "step": 198
    },
    {
      "epoch": 2.689189189189189,
      "grad_norm": 0.035465750843286514,
      "learning_rate": 3.441369080182749e-05,
      "loss": 1.0712890625,
      "step": 199
    },
    {
      "epoch": 2.7027027027027026,
      "grad_norm": 0.03351772576570511,
      "learning_rate": 3.396761667294579e-05,
      "loss": 1.02783203125,
      "step": 200
    },
    {
      "epoch": 2.7162162162162162,
      "grad_norm": 0.034879669547080994,
      "learning_rate": 3.352417142592816e-05,
      "loss": 1.1015625,
      "step": 201
    },
    {
      "epoch": 2.72972972972973,
      "grad_norm": 0.03477325290441513,
      "learning_rate": 3.308341048792944e-05,
      "loss": 1.0830078125,
      "step": 202
    },
    {
      "epoch": 2.743243243243243,
      "grad_norm": 0.04783156141638756,
      "learning_rate": 3.264538895058705e-05,
      "loss": 1.02294921875,
      "step": 203
    },
    {
      "epoch": 2.756756756756757,
      "grad_norm": 0.036511294543743134,
      "learning_rate": 3.221016156313495e-05,
      "loss": 1.107421875,
      "step": 204
    },
    {
      "epoch": 2.77027027027027,
      "grad_norm": 0.03128586709499359,
      "learning_rate": 3.177778272556041e-05,
      "loss": 0.9393310546875,
      "step": 205
    },
    {
      "epoch": 2.7837837837837838,
      "grad_norm": 0.03910524770617485,
      "learning_rate": 3.134830648180451e-05,
      "loss": 1.103515625,
      "step": 206
    },
    {
      "epoch": 2.7972972972972974,
      "grad_norm": 0.03559153527021408,
      "learning_rate": 3.0921786513006976e-05,
      "loss": 1.0791015625,
      "step": 207
    },
    {
      "epoch": 2.810810810810811,
      "grad_norm": 0.03547922149300575,
      "learning_rate": 3.0498276130796476e-05,
      "loss": 1.04345703125,
      "step": 208
    },
    {
      "epoch": 2.8243243243243246,
      "grad_norm": 0.03541857749223709,
      "learning_rate": 3.0077828270627184e-05,
      "loss": 1.0458984375,
      "step": 209
    },
    {
      "epoch": 2.8378378378378377,
      "grad_norm": 0.034462820738554,
      "learning_rate": 2.966049548516212e-05,
      "loss": 1.07861328125,
      "step": 210
    },
    {
      "epoch": 2.8513513513513513,
      "grad_norm": 0.03096124716103077,
      "learning_rate": 2.9246329937704654e-05,
      "loss": 0.931396484375,
      "step": 211
    },
    {
      "epoch": 2.864864864864865,
      "grad_norm": 0.03517831116914749,
      "learning_rate": 2.8835383395678328e-05,
      "loss": 1.0927734375,
      "step": 212
    },
    {
      "epoch": 2.8783783783783785,
      "grad_norm": 0.03210766986012459,
      "learning_rate": 2.8427707224156453e-05,
      "loss": 0.970703125,
      "step": 213
    },
    {
      "epoch": 2.891891891891892,
      "grad_norm": 0.03376515209674835,
      "learning_rate": 2.8023352379441848e-05,
      "loss": 1.06640625,
      "step": 214
    },
    {
      "epoch": 2.9054054054054053,
      "grad_norm": 0.03345406427979469,
      "learning_rate": 2.7622369402697678e-05,
      "loss": 1.0625,
      "step": 215
    },
    {
      "epoch": 2.918918918918919,
      "grad_norm": 0.034529391676187515,
      "learning_rate": 2.7224808413630287e-05,
      "loss": 1.076171875,
      "step": 216
    },
    {
      "epoch": 2.9324324324324325,
      "grad_norm": 0.03561599180102348,
      "learning_rate": 2.6830719104224533e-05,
      "loss": 1.1240234375,
      "step": 217
    },
    {
      "epoch": 2.945945945945946,
      "grad_norm": 0.03500950708985329,
      "learning_rate": 2.644015073253272e-05,
      "loss": 1.09814453125,
      "step": 218
    },
    {
      "epoch": 2.9594594594594597,
      "grad_norm": 0.03659572824835777,
      "learning_rate": 2.605315211651776e-05,
      "loss": 1.1064453125,
      "step": 219
    },
    {
      "epoch": 2.972972972972973,
      "grad_norm": 0.034403424710035324,
      "learning_rate": 2.5669771627951277e-05,
      "loss": 1.1103515625,
      "step": 220
    },
    {
      "epoch": 2.9864864864864864,
      "grad_norm": 0.03362520784139633,
      "learning_rate": 2.5290057186367526e-05,
      "loss": 1.0791015625,
      "step": 221
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.03195793554186821,
      "learning_rate": 2.491405625307386e-05,
      "loss": 1.03076171875,
      "step": 222
    },
    {
      "epoch": 3.0135135135135136,
      "grad_norm": 0.03462374582886696,
      "learning_rate": 2.4541815825218376e-05,
      "loss": 1.1142578125,
      "step": 223
    },
    {
      "epoch": 3.027027027027027,
      "grad_norm": 0.03713052719831467,
      "learning_rate": 2.4173382429915705e-05,
      "loss": 1.080078125,
      "step": 224
    },
    {
      "epoch": 3.0405405405405403,
      "grad_norm": 0.03585672751069069,
      "learning_rate": 2.380880211843143e-05,
      "loss": 1.09912109375,
      "step": 225
    },
    {
      "epoch": 3.054054054054054,
      "grad_norm": 0.03537458926439285,
      "learning_rate": 2.344812046042615e-05,
      "loss": 1.03759765625,
      "step": 226
    },
    {
      "epoch": 3.0675675675675675,
      "grad_norm": 0.03894190490245819,
      "learning_rate": 2.309138253825948e-05,
      "loss": 1.08935546875,
      "step": 227
    },
    {
      "epoch": 3.081081081081081,
      "grad_norm": 0.03595813736319542,
      "learning_rate": 2.273863294135525e-05,
      "loss": 1.02734375,
      "step": 228
    },
    {
      "epoch": 3.0945945945945947,
      "grad_norm": 0.0337008535861969,
      "learning_rate": 2.2389915760628077e-05,
      "loss": 1.08984375,
      "step": 229
    },
    {
      "epoch": 3.108108108108108,
      "grad_norm": 0.0355982780456543,
      "learning_rate": 2.204527458297242e-05,
      "loss": 1.01416015625,
      "step": 230
    },
    {
      "epoch": 3.1216216216216215,
      "grad_norm": 0.03433150053024292,
      "learning_rate": 2.170475248581452e-05,
      "loss": 0.987548828125,
      "step": 231
    },
    {
      "epoch": 3.135135135135135,
      "grad_norm": 0.03790724277496338,
      "learning_rate": 2.1368392031728067e-05,
      "loss": 1.07421875,
      "step": 232
    },
    {
      "epoch": 3.1486486486486487,
      "grad_norm": 0.03570462763309479,
      "learning_rate": 2.1036235263114264e-05,
      "loss": 1.08935546875,
      "step": 233
    },
    {
      "epoch": 3.1621621621621623,
      "grad_norm": 0.03378927335143089,
      "learning_rate": 2.0708323696946785e-05,
      "loss": 1.04833984375,
      "step": 234
    },
    {
      "epoch": 3.175675675675676,
      "grad_norm": 0.03625774011015892,
      "learning_rate": 2.03846983195825e-05,
      "loss": 1.0517578125,
      "step": 235
    },
    {
      "epoch": 3.189189189189189,
      "grad_norm": 0.03646084666252136,
      "learning_rate": 2.0065399581638575e-05,
      "loss": 1.1123046875,
      "step": 236
    },
    {
      "epoch": 3.2027027027027026,
      "grad_norm": 0.04458549618721008,
      "learning_rate": 1.975046739293635e-05,
      "loss": 1.0703125,
      "step": 237
    },
    {
      "epoch": 3.2027027027027026,
      "eval_loss": 1.13330078125,
      "eval_runtime": 5.5685,
      "eval_samples_per_second": 0.718,
      "eval_steps_per_second": 0.18,
      "step": 237
    },
    {
      "epoch": 3.2162162162162162,
      "grad_norm": 0.03523408621549606,
      "learning_rate": 1.943994111751301e-05,
      "loss": 1.0830078125,
      "step": 238
    },
    {
      "epoch": 3.22972972972973,
      "grad_norm": 0.03858271241188049,
      "learning_rate": 1.9133859568701357e-05,
      "loss": 1.0478515625,
      "step": 239
    },
    {
      "epoch": 3.2432432432432434,
      "grad_norm": 0.03811842203140259,
      "learning_rate": 1.883226100427846e-05,
      "loss": 1.08984375,
      "step": 240
    },
    {
      "epoch": 3.2567567567567566,
      "grad_norm": 0.03568769991397858,
      "learning_rate": 1.8535183121683752e-05,
      "loss": 1.10693359375,
      "step": 241
    },
    {
      "epoch": 3.27027027027027,
      "grad_norm": 0.035731736570596695,
      "learning_rate": 1.824266305330709e-05,
      "loss": 1.06103515625,
      "step": 242
    },
    {
      "epoch": 3.2837837837837838,
      "grad_norm": 0.030849639326334,
      "learning_rate": 1.7954737361847566e-05,
      "loss": 0.91015625,
      "step": 243
    },
    {
      "epoch": 3.2972972972972974,
      "grad_norm": 0.033462926745414734,
      "learning_rate": 1.7671442035743415e-05,
      "loss": 1.03955078125,
      "step": 244
    },
    {
      "epoch": 3.310810810810811,
      "grad_norm": 0.03543676435947418,
      "learning_rate": 1.7392812484673794e-05,
      "loss": 1.087890625,
      "step": 245
    },
    {
      "epoch": 3.3243243243243246,
      "grad_norm": 0.03611475229263306,
      "learning_rate": 1.7118883535132756e-05,
      "loss": 1.03515625,
      "step": 246
    },
    {
      "epoch": 3.3378378378378377,
      "grad_norm": 0.034762006253004074,
      "learning_rate": 1.684968942607634e-05,
      "loss": 1.083984375,
      "step": 247
    },
    {
      "epoch": 3.3513513513513513,
      "grad_norm": 0.03383259475231171,
      "learning_rate": 1.6585263804642866e-05,
      "loss": 1.08544921875,
      "step": 248
    },
    {
      "epoch": 3.364864864864865,
      "grad_norm": 0.034660059958696365,
      "learning_rate": 1.6325639721947355e-05,
      "loss": 1.0693359375,
      "step": 249
    },
    {
      "epoch": 3.3783783783783785,
      "grad_norm": 0.03567449375987053,
      "learning_rate": 1.60708496289504e-05,
      "loss": 1.09326171875,
      "step": 250
    },
    {
      "epoch": 3.391891891891892,
      "grad_norm": 0.035272885113954544,
      "learning_rate": 1.582092537240204e-05,
      "loss": 1.04736328125,
      "step": 251
    },
    {
      "epoch": 3.4054054054054053,
      "grad_norm": 0.03724433481693268,
      "learning_rate": 1.557589819086118e-05,
      "loss": 1.04296875,
      "step": 252
    },
    {
      "epoch": 3.418918918918919,
      "grad_norm": 0.039717625826597214,
      "learning_rate": 1.5335798710791012e-05,
      "loss": 1.076171875,
      "step": 253
    },
    {
      "epoch": 3.4324324324324325,
      "grad_norm": 0.035618238151073456,
      "learning_rate": 1.5100656942730934e-05,
      "loss": 1.08642578125,
      "step": 254
    },
    {
      "epoch": 3.445945945945946,
      "grad_norm": 0.03635922446846962,
      "learning_rate": 1.4870502277545517e-05,
      "loss": 1.09765625,
      "step": 255
    },
    {
      "epoch": 3.4594594594594597,
      "grad_norm": 0.036835722625255585,
      "learning_rate": 1.464536348275081e-05,
      "loss": 1.0830078125,
      "step": 256
    },
    {
      "epoch": 3.472972972972973,
      "grad_norm": 0.03637174889445305,
      "learning_rate": 1.4425268698918679e-05,
      "loss": 1.06982421875,
      "step": 257
    },
    {
      "epoch": 3.4864864864864864,
      "grad_norm": 0.0317935049533844,
      "learning_rate": 1.4210245436159455e-05,
      "loss": 0.9285888671875,
      "step": 258
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.03613898158073425,
      "learning_rate": 1.4000320570683356e-05,
      "loss": 1.0625,
      "step": 259
    },
    {
      "epoch": 3.5135135135135136,
      "grad_norm": 0.034447554498910904,
      "learning_rate": 1.3795520341441192e-05,
      "loss": 1.041015625,
      "step": 260
    },
    {
      "epoch": 3.527027027027027,
      "grad_norm": 0.03817327693104744,
      "learning_rate": 1.3595870346844723e-05,
      "loss": 1.1025390625,
      "step": 261
    },
    {
      "epoch": 3.5405405405405403,
      "grad_norm": 0.03468001261353493,
      "learning_rate": 1.3401395541567011e-05,
      "loss": 1.04150390625,
      "step": 262
    },
    {
      "epoch": 3.554054054054054,
      "grad_norm": 0.03400794789195061,
      "learning_rate": 1.3212120233423334e-05,
      "loss": 1.06884765625,
      "step": 263
    },
    {
      "epoch": 3.5675675675675675,
      "grad_norm": 0.03681923449039459,
      "learning_rate": 1.3028068080332856e-05,
      "loss": 1.08447265625,
      "step": 264
    },
    {
      "epoch": 3.581081081081081,
      "grad_norm": 0.03200948238372803,
      "learning_rate": 1.2849262087361638e-05,
      "loss": 0.845703125,
      "step": 265
    },
    {
      "epoch": 3.5945945945945947,
      "grad_norm": 0.03910684958100319,
      "learning_rate": 1.2675724603847103e-05,
      "loss": 1.0654296875,
      "step": 266
    },
    {
      "epoch": 3.608108108108108,
      "grad_norm": 0.03780676797032356,
      "learning_rate": 1.250747732060461e-05,
      "loss": 1.0205078125,
      "step": 267
    },
    {
      "epoch": 3.6216216216216215,
      "grad_norm": 0.03405677527189255,
      "learning_rate": 1.2344541267216227e-05,
      "loss": 1.06640625,
      "step": 268
    },
    {
      "epoch": 3.635135135135135,
      "grad_norm": 0.04388781636953354,
      "learning_rate": 1.218693680940225e-05,
      "loss": 1.07470703125,
      "step": 269
    },
    {
      "epoch": 3.6486486486486487,
      "grad_norm": 0.034076027572155,
      "learning_rate": 1.2034683646475617e-05,
      "loss": 1.05712890625,
      "step": 270
    },
    {
      "epoch": 3.6621621621621623,
      "grad_norm": 0.035192202776670456,
      "learning_rate": 1.1887800808879632e-05,
      "loss": 1.05322265625,
      "step": 271
    },
    {
      "epoch": 3.6756756756756754,
      "grad_norm": 0.03537910059094429,
      "learning_rate": 1.1746306655809353e-05,
      "loss": 1.0693359375,
      "step": 272
    },
    {
      "epoch": 3.689189189189189,
      "grad_norm": 0.03666433319449425,
      "learning_rate": 1.1610218872916807e-05,
      "loss": 1.048828125,
      "step": 273
    },
    {
      "epoch": 3.7027027027027026,
      "grad_norm": 0.03601507470011711,
      "learning_rate": 1.1479554470100447e-05,
      "loss": 1.00634765625,
      "step": 274
    },
    {
      "epoch": 3.7162162162162162,
      "grad_norm": 0.03602784872055054,
      "learning_rate": 1.1354329779379003e-05,
      "loss": 1.08154296875,
      "step": 275
    },
    {
      "epoch": 3.72972972972973,
      "grad_norm": 0.03820501267910004,
      "learning_rate": 1.1234560452850176e-05,
      "loss": 1.06396484375,
      "step": 276
    },
    {
      "epoch": 3.743243243243243,
      "grad_norm": 0.0349762998521328,
      "learning_rate": 1.1120261460734187e-05,
      "loss": 1.00634765625,
      "step": 277
    },
    {
      "epoch": 3.756756756756757,
      "grad_norm": 0.036815810948610306,
      "learning_rate": 1.101144708950266e-05,
      "loss": 1.0888671875,
      "step": 278
    },
    {
      "epoch": 3.77027027027027,
      "grad_norm": 0.03410951793193817,
      "learning_rate": 1.0908130940092891e-05,
      "loss": 0.924041748046875,
      "step": 279
    },
    {
      "epoch": 3.7837837837837838,
      "grad_norm": 0.03487250208854675,
      "learning_rate": 1.0810325926207884e-05,
      "loss": 1.08349609375,
      "step": 280
    },
    {
      "epoch": 3.7972972972972974,
      "grad_norm": 0.03428065776824951,
      "learning_rate": 1.0718044272702193e-05,
      "loss": 1.06103515625,
      "step": 281
    },
    {
      "epoch": 3.810810810810811,
      "grad_norm": 0.032850828021764755,
      "learning_rate": 1.0631297514053941e-05,
      "loss": 1.02783203125,
      "step": 282
    },
    {
      "epoch": 3.8243243243243246,
      "grad_norm": 0.034615907818078995,
      "learning_rate": 1.0550096492923078e-05,
      "loss": 1.0283203125,
      "step": 283
    },
    {
      "epoch": 3.8378378378378377,
      "grad_norm": 0.03582344204187393,
      "learning_rate": 1.0474451358796167e-05,
      "loss": 1.0634765625,
      "step": 284
    },
    {
      "epoch": 3.8513513513513513,
      "grad_norm": 0.03418773040175438,
      "learning_rate": 1.040437156671776e-05,
      "loss": 0.9150390625,
      "step": 285
    },
    {
      "epoch": 3.864864864864865,
      "grad_norm": 0.0336853452026844,
      "learning_rate": 1.0339865876108563e-05,
      "loss": 1.07373046875,
      "step": 286
    },
    {
      "epoch": 3.8783783783783785,
      "grad_norm": 0.032579585909843445,
      "learning_rate": 1.0280942349670642e-05,
      "loss": 0.95703125,
      "step": 287
    },
    {
      "epoch": 3.891891891891892,
      "grad_norm": 0.03456036001443863,
      "learning_rate": 1.0227608352379591e-05,
      "loss": 1.04931640625,
      "step": 288
    },
    {
      "epoch": 3.9054054054054053,
      "grad_norm": 0.03553551062941551,
      "learning_rate": 1.0179870550563996e-05,
      "loss": 1.04443359375,
      "step": 289
    },
    {
      "epoch": 3.918918918918919,
      "grad_norm": 0.03818385303020477,
      "learning_rate": 1.0137734911072196e-05,
      "loss": 1.06103515625,
      "step": 290
    },
    {
      "epoch": 3.9324324324324325,
      "grad_norm": 0.037002529948949814,
      "learning_rate": 1.0101206700526458e-05,
      "loss": 1.1083984375,
      "step": 291
    },
    {
      "epoch": 3.945945945945946,
      "grad_norm": 0.03469846770167351,
      "learning_rate": 1.0070290484664712e-05,
      "loss": 1.08251953125,
      "step": 292
    },
    {
      "epoch": 3.9594594594594597,
      "grad_norm": 0.03774961829185486,
      "learning_rate": 1.0044990127769854e-05,
      "loss": 1.091796875,
      "step": 293
    },
    {
      "epoch": 3.972972972972973,
      "grad_norm": 0.0354500412940979,
      "learning_rate": 1.0025308792186745e-05,
      "loss": 1.095703125,
      "step": 294
    },
    {
      "epoch": 3.9864864864864864,
      "grad_norm": 0.03584533929824829,
      "learning_rate": 1.001124893792696e-05,
      "loss": 1.06298828125,
      "step": 295
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.034482669085264206,
      "learning_rate": 1.0002812322361266e-05,
      "loss": 1.01611328125,
      "step": 296
    },
    {
      "epoch": 4.0,
      "eval_loss": 1.12939453125,
      "eval_runtime": 5.5864,
      "eval_samples_per_second": 0.716,
      "eval_steps_per_second": 0.179,
      "step": 296
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 296,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.57668398362489e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}