{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.955414012738854,
  "eval_steps": 500,
  "global_step": 364,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01910828025477707,
      "grad_norm": 16.385162993426572,
      "learning_rate": 5.405405405405406e-07,
      "loss": 1.5698,
      "step": 1
    },
    {
      "epoch": 0.03821656050955414,
      "grad_norm": 16.504800604597186,
      "learning_rate": 1.0810810810810812e-06,
      "loss": 1.5972,
      "step": 2
    },
    {
      "epoch": 0.05732484076433121,
      "grad_norm": 15.46635677824654,
      "learning_rate": 1.6216216216216219e-06,
      "loss": 1.5254,
      "step": 3
    },
    {
      "epoch": 0.07643312101910828,
      "grad_norm": 16.32262911321192,
      "learning_rate": 2.1621621621621623e-06,
      "loss": 1.5602,
      "step": 4
    },
    {
      "epoch": 0.09554140127388536,
      "grad_norm": 14.532866256919611,
      "learning_rate": 2.702702702702703e-06,
      "loss": 1.4957,
      "step": 5
    },
    {
      "epoch": 0.11464968152866242,
      "grad_norm": 12.429084270136249,
      "learning_rate": 3.2432432432432437e-06,
      "loss": 1.4258,
      "step": 6
    },
    {
      "epoch": 0.1337579617834395,
      "grad_norm": 9.200276130835531,
      "learning_rate": 3.7837837837837844e-06,
      "loss": 1.2829,
      "step": 7
    },
    {
      "epoch": 0.15286624203821655,
      "grad_norm": 7.518795812144619,
      "learning_rate": 4.324324324324325e-06,
      "loss": 1.265,
      "step": 8
    },
    {
      "epoch": 0.17197452229299362,
      "grad_norm": 6.557943979035445,
      "learning_rate": 4.864864864864866e-06,
      "loss": 1.2011,
      "step": 9
    },
    {
      "epoch": 0.1910828025477707,
      "grad_norm": 7.761386310023502,
      "learning_rate": 5.405405405405406e-06,
      "loss": 1.1285,
      "step": 10
    },
    {
      "epoch": 0.21019108280254778,
      "grad_norm": 5.800801599651337,
      "learning_rate": 5.945945945945947e-06,
      "loss": 1.0849,
      "step": 11
    },
    {
      "epoch": 0.22929936305732485,
      "grad_norm": 4.458529924701665,
      "learning_rate": 6.486486486486487e-06,
      "loss": 1.0295,
      "step": 12
    },
    {
      "epoch": 0.2484076433121019,
      "grad_norm": 4.449464961059666,
      "learning_rate": 7.027027027027028e-06,
      "loss": 1.0054,
      "step": 13
    },
    {
      "epoch": 0.267515923566879,
      "grad_norm": 5.151158650335397,
      "learning_rate": 7.567567567567569e-06,
      "loss": 0.9589,
      "step": 14
    },
    {
      "epoch": 0.28662420382165604,
      "grad_norm": 4.757079257630854,
      "learning_rate": 8.108108108108109e-06,
      "loss": 0.9881,
      "step": 15
    },
    {
      "epoch": 0.3057324840764331,
      "grad_norm": 4.440520703195754,
      "learning_rate": 8.64864864864865e-06,
      "loss": 0.955,
      "step": 16
    },
    {
      "epoch": 0.3248407643312102,
      "grad_norm": 2.8333651824679036,
      "learning_rate": 9.189189189189191e-06,
      "loss": 0.9141,
      "step": 17
    },
    {
      "epoch": 0.34394904458598724,
      "grad_norm": 2.759048068164436,
      "learning_rate": 9.729729729729732e-06,
      "loss": 0.9251,
      "step": 18
    },
    {
      "epoch": 0.3630573248407643,
      "grad_norm": 3.20055476511223,
      "learning_rate": 1.027027027027027e-05,
      "loss": 0.9106,
      "step": 19
    },
    {
      "epoch": 0.3821656050955414,
      "grad_norm": 3.541249127738148,
      "learning_rate": 1.0810810810810812e-05,
      "loss": 0.8812,
      "step": 20
    },
    {
      "epoch": 0.4012738853503185,
      "grad_norm": 2.807031455712289,
      "learning_rate": 1.1351351351351352e-05,
      "loss": 0.8684,
      "step": 21
    },
    {
      "epoch": 0.42038216560509556,
      "grad_norm": 2.042523718413902,
      "learning_rate": 1.1891891891891894e-05,
      "loss": 0.8832,
      "step": 22
    },
    {
      "epoch": 0.4394904458598726,
      "grad_norm": 2.1481689144751694,
      "learning_rate": 1.2432432432432433e-05,
      "loss": 0.878,
      "step": 23
    },
    {
      "epoch": 0.4585987261146497,
      "grad_norm": 2.4022110137127726,
      "learning_rate": 1.2972972972972975e-05,
      "loss": 0.864,
      "step": 24
    },
    {
      "epoch": 0.47770700636942676,
      "grad_norm": 2.1446981991533782,
      "learning_rate": 1.3513513513513515e-05,
      "loss": 0.8365,
      "step": 25
    },
    {
      "epoch": 0.4968152866242038,
      "grad_norm": 2.0893745183886896,
      "learning_rate": 1.4054054054054055e-05,
      "loss": 0.8362,
      "step": 26
    },
    {
      "epoch": 0.5159235668789809,
      "grad_norm": 1.6488707388541979,
      "learning_rate": 1.4594594594594596e-05,
      "loss": 0.8495,
      "step": 27
    },
    {
      "epoch": 0.535031847133758,
      "grad_norm": 1.6921694203114936,
      "learning_rate": 1.5135135135135138e-05,
      "loss": 0.8239,
      "step": 28
    },
    {
      "epoch": 0.554140127388535,
      "grad_norm": 2.0573030839069117,
      "learning_rate": 1.5675675675675676e-05,
      "loss": 0.8448,
      "step": 29
    },
    {
      "epoch": 0.5732484076433121,
      "grad_norm": 1.7248249635999773,
      "learning_rate": 1.6216216216216218e-05,
      "loss": 0.8311,
      "step": 30
    },
    {
      "epoch": 0.5923566878980892,
      "grad_norm": 1.6940388762105907,
      "learning_rate": 1.6756756756756757e-05,
      "loss": 0.8207,
      "step": 31
    },
    {
      "epoch": 0.6114649681528662,
      "grad_norm": 1.6926871774648706,
      "learning_rate": 1.72972972972973e-05,
      "loss": 0.8042,
      "step": 32
    },
    {
      "epoch": 0.6305732484076433,
      "grad_norm": 1.6302594155794337,
      "learning_rate": 1.783783783783784e-05,
      "loss": 0.809,
      "step": 33
    },
    {
      "epoch": 0.6496815286624203,
      "grad_norm": 1.8069821223276006,
      "learning_rate": 1.8378378378378383e-05,
      "loss": 0.8475,
      "step": 34
    },
    {
      "epoch": 0.6687898089171974,
      "grad_norm": 1.6810398894595866,
      "learning_rate": 1.891891891891892e-05,
      "loss": 0.795,
      "step": 35
    },
    {
      "epoch": 0.6878980891719745,
      "grad_norm": 1.6752610079484405,
      "learning_rate": 1.9459459459459463e-05,
      "loss": 0.8243,
      "step": 36
    },
    {
      "epoch": 0.7070063694267515,
      "grad_norm": 1.748395181450843,
      "learning_rate": 2e-05,
      "loss": 0.8132,
      "step": 37
    },
    {
      "epoch": 0.7261146496815286,
      "grad_norm": 1.648537098710576,
      "learning_rate": 1.9999538500851633e-05,
      "loss": 0.8151,
      "step": 38
    },
    {
      "epoch": 0.7452229299363057,
      "grad_norm": 1.5757172038673941,
      "learning_rate": 1.9998154046002822e-05,
      "loss": 0.8062,
      "step": 39
    },
    {
      "epoch": 0.7643312101910829,
      "grad_norm": 1.8845590574584812,
      "learning_rate": 1.9995846763238514e-05,
      "loss": 0.8148,
      "step": 40
    },
    {
      "epoch": 0.7834394904458599,
      "grad_norm": 2.094114813878872,
      "learning_rate": 1.9992616865520515e-05,
      "loss": 0.8219,
      "step": 41
    },
    {
      "epoch": 0.802547770700637,
      "grad_norm": 1.6890917880505276,
      "learning_rate": 1.9988464650967834e-05,
      "loss": 0.8053,
      "step": 42
    },
    {
      "epoch": 0.821656050955414,
      "grad_norm": 1.606154927757665,
      "learning_rate": 1.9983390502829168e-05,
      "loss": 0.8001,
      "step": 43
    },
    {
      "epoch": 0.8407643312101911,
      "grad_norm": 1.4922460443430643,
      "learning_rate": 1.9977394889447526e-05,
      "loss": 0.8023,
      "step": 44
    },
    {
      "epoch": 0.8598726114649682,
      "grad_norm": 1.721473976909461,
      "learning_rate": 1.9970478364217e-05,
      "loss": 0.7922,
      "step": 45
    },
    {
      "epoch": 0.8789808917197452,
      "grad_norm": 1.492822163666417,
      "learning_rate": 1.9962641565531694e-05,
      "loss": 0.8043,
      "step": 46
    },
    {
      "epoch": 0.8980891719745223,
      "grad_norm": 1.4028695347925793,
      "learning_rate": 1.9953885216726788e-05,
      "loss": 0.8107,
      "step": 47
    },
    {
      "epoch": 0.9171974522292994,
      "grad_norm": 1.4948166631903215,
      "learning_rate": 1.994421012601179e-05,
      "loss": 0.7702,
      "step": 48
    },
    {
      "epoch": 0.9363057324840764,
      "grad_norm": 1.5163532996354703,
      "learning_rate": 1.9933617186395917e-05,
      "loss": 0.8117,
      "step": 49
    },
    {
      "epoch": 0.9554140127388535,
      "grad_norm": 1.550783453013226,
      "learning_rate": 1.99221073756057e-05,
      "loss": 0.7898,
      "step": 50
    },
    {
      "epoch": 0.9745222929936306,
      "grad_norm": 1.6260454029830684,
      "learning_rate": 1.990968175599471e-05,
      "loss": 0.8044,
      "step": 51
    },
    {
      "epoch": 0.9936305732484076,
      "grad_norm": 1.6316779083657356,
      "learning_rate": 1.9896341474445526e-05,
      "loss": 0.7751,
      "step": 52
    },
    {
      "epoch": 1.0127388535031847,
      "grad_norm": 1.6273455814047662,
      "learning_rate": 1.9882087762263857e-05,
      "loss": 0.7185,
      "step": 53
    },
    {
      "epoch": 1.0318471337579618,
      "grad_norm": 1.3702016927688272,
      "learning_rate": 1.9866921935064907e-05,
      "loss": 0.6317,
      "step": 54
    },
    {
      "epoch": 1.0509554140127388,
      "grad_norm": 1.3702776063715345,
      "learning_rate": 1.985084539265195e-05,
      "loss": 0.5979,
      "step": 55
    },
    {
      "epoch": 1.070063694267516,
      "grad_norm": 1.357835755609063,
      "learning_rate": 1.983385961888711e-05,
      "loss": 0.599,
      "step": 56
    },
    {
      "epoch": 1.089171974522293,
      "grad_norm": 1.5607197079673851,
      "learning_rate": 1.9815966181554412e-05,
      "loss": 0.602,
      "step": 57
    },
    {
      "epoch": 1.10828025477707,
      "grad_norm": 1.5351352622145285,
      "learning_rate": 1.9797166732215078e-05,
      "loss": 0.6138,
      "step": 58
    },
    {
      "epoch": 1.127388535031847,
      "grad_norm": 1.7266095022061323,
      "learning_rate": 1.977746300605507e-05,
      "loss": 0.5711,
      "step": 59
    },
    {
      "epoch": 1.1464968152866242,
      "grad_norm": 1.587015285885074,
      "learning_rate": 1.975685682172497e-05,
      "loss": 0.5927,
      "step": 60
    },
    {
      "epoch": 1.1656050955414012,
      "grad_norm": 1.3602798340924052,
      "learning_rate": 1.973535008117207e-05,
      "loss": 0.5961,
      "step": 61
    },
    {
      "epoch": 1.1847133757961783,
      "grad_norm": 1.5396673283104687,
      "learning_rate": 1.9712944769464864e-05,
      "loss": 0.5776,
      "step": 62
    },
    {
      "epoch": 1.2038216560509554,
      "grad_norm": 1.4151140824773938,
      "learning_rate": 1.9689642954609808e-05,
      "loss": 0.5703,
      "step": 63
    },
    {
      "epoch": 1.2229299363057324,
      "grad_norm": 1.6305530726818593,
      "learning_rate": 1.9665446787360444e-05,
      "loss": 0.6179,
      "step": 64
    },
    {
      "epoch": 1.2420382165605095,
      "grad_norm": 1.5052390637526016,
      "learning_rate": 1.9640358501018885e-05,
      "loss": 0.5533,
      "step": 65
    },
    {
      "epoch": 1.2611464968152866,
      "grad_norm": 1.4443125571283753,
      "learning_rate": 1.9614380411229693e-05,
      "loss": 0.5839,
      "step": 66
    },
    {
      "epoch": 1.2802547770700636,
      "grad_norm": 1.3308482373115775,
      "learning_rate": 1.9587514915766124e-05,
      "loss": 0.5883,
      "step": 67
    },
    {
      "epoch": 1.2993630573248407,
      "grad_norm": 1.42729767228037,
      "learning_rate": 1.9559764494308838e-05,
      "loss": 0.5448,
      "step": 68
    },
    {
      "epoch": 1.3184713375796178,
      "grad_norm": 1.3781970251357278,
      "learning_rate": 1.9531131708217005e-05,
      "loss": 0.5767,
      "step": 69
    },
    {
      "epoch": 1.3375796178343948,
      "grad_norm": 1.3677879447269978,
      "learning_rate": 1.950161920029191e-05,
      "loss": 0.5755,
      "step": 70
    },
    {
      "epoch": 1.356687898089172,
      "grad_norm": 1.453105482367857,
      "learning_rate": 1.9471229694533003e-05,
      "loss": 0.5469,
      "step": 71
    },
    {
      "epoch": 1.3757961783439492,
      "grad_norm": 1.4482246018744436,
      "learning_rate": 1.943996599588649e-05,
      "loss": 0.5784,
      "step": 72
    },
    {
      "epoch": 1.394904458598726,
      "grad_norm": 1.401709151246391,
      "learning_rate": 1.940783098998643e-05,
      "loss": 0.592,
      "step": 73
    },
    {
      "epoch": 1.4140127388535033,
      "grad_norm": 1.445141295836924,
      "learning_rate": 1.93748276428884e-05,
      "loss": 0.5365,
      "step": 74
    },
    {
      "epoch": 1.4331210191082802,
      "grad_norm": 1.3732889276145823,
      "learning_rate": 1.9340959000795707e-05,
      "loss": 0.5772,
      "step": 75
    },
    {
      "epoch": 1.4522292993630574,
      "grad_norm": 1.2182338563233495,
      "learning_rate": 1.9306228189778255e-05,
      "loss": 0.5819,
      "step": 76
    },
    {
      "epoch": 1.4713375796178343,
      "grad_norm": 1.470612047637857,
      "learning_rate": 1.927063841548398e-05,
      "loss": 0.5574,
      "step": 77
    },
    {
      "epoch": 1.4904458598726116,
      "grad_norm": 1.586990211389091,
      "learning_rate": 1.9234192962842996e-05,
      "loss": 0.5665,
      "step": 78
    },
    {
      "epoch": 1.5095541401273884,
      "grad_norm": 1.3622127174538021,
      "learning_rate": 1.9196895195764363e-05,
      "loss": 0.6213,
      "step": 79
    },
    {
      "epoch": 1.5286624203821657,
      "grad_norm": 1.5279213108356509,
      "learning_rate": 1.9158748556825637e-05,
      "loss": 0.5928,
      "step": 80
    },
    {
      "epoch": 1.5477707006369426,
      "grad_norm": 1.381548176474239,
      "learning_rate": 1.9119756566955092e-05,
      "loss": 0.6015,
      "step": 81
    },
    {
      "epoch": 1.5668789808917198,
      "grad_norm": 1.4791457207376197,
      "learning_rate": 1.907992282510675e-05,
      "loss": 0.5937,
      "step": 82
    },
    {
      "epoch": 1.5859872611464967,
      "grad_norm": 1.3848517552915744,
      "learning_rate": 1.90392510079282e-05,
      "loss": 0.5835,
      "step": 83
    },
    {
      "epoch": 1.605095541401274,
      "grad_norm": 1.4796492304425215,
      "learning_rate": 1.8997744869421248e-05,
      "loss": 0.5785,
      "step": 84
    },
    {
      "epoch": 1.6242038216560508,
      "grad_norm": 1.4205681356051292,
      "learning_rate": 1.8955408240595396e-05,
      "loss": 0.5773,
      "step": 85
    },
    {
      "epoch": 1.643312101910828,
      "grad_norm": 1.5299136367520119,
      "learning_rate": 1.891224502911428e-05,
      "loss": 0.6068,
      "step": 86
    },
    {
      "epoch": 1.662420382165605,
      "grad_norm": 1.4517525362811488,
      "learning_rate": 1.886825921893497e-05,
      "loss": 0.5602,
      "step": 87
    },
    {
      "epoch": 1.6815286624203822,
      "grad_norm": 1.3605482820727088,
      "learning_rate": 1.8823454869940243e-05,
      "loss": 0.582,
      "step": 88
    },
    {
      "epoch": 1.700636942675159,
      "grad_norm": 1.462809151995504,
      "learning_rate": 1.8777836117563894e-05,
      "loss": 0.578,
      "step": 89
    },
    {
      "epoch": 1.7197452229299364,
      "grad_norm": 1.309078307464712,
      "learning_rate": 1.873140717240899e-05,
      "loss": 0.6215,
      "step": 90
    },
    {
      "epoch": 1.7388535031847132,
      "grad_norm": 1.3851134513499674,
      "learning_rate": 1.8684172319859258e-05,
      "loss": 0.6236,
      "step": 91
    },
    {
      "epoch": 1.7579617834394905,
      "grad_norm": 1.3737393579288095,
      "learning_rate": 1.863613591968355e-05,
      "loss": 0.5538,
      "step": 92
    },
    {
      "epoch": 1.7770700636942676,
      "grad_norm": 1.2714423428796768,
      "learning_rate": 1.858730240563342e-05,
      "loss": 0.5492,
      "step": 93
    },
    {
      "epoch": 1.7961783439490446,
      "grad_norm": 1.3886107352803891,
      "learning_rate": 1.8537676285033886e-05,
      "loss": 0.6016,
      "step": 94
    },
    {
      "epoch": 1.8152866242038217,
      "grad_norm": 1.3906573234567905,
      "learning_rate": 1.848726213836744e-05,
      "loss": 0.5578,
      "step": 95
    },
    {
      "epoch": 1.8343949044585988,
      "grad_norm": 1.3072957741832658,
      "learning_rate": 1.8436064618851225e-05,
      "loss": 0.5573,
      "step": 96
    },
    {
      "epoch": 1.8535031847133758,
      "grad_norm": 1.3086263321701983,
      "learning_rate": 1.838408845200758e-05,
      "loss": 0.5524,
      "step": 97
    },
    {
      "epoch": 1.872611464968153,
      "grad_norm": 1.6817742508337377,
      "learning_rate": 1.8331338435227838e-05,
      "loss": 0.5807,
      "step": 98
    },
    {
      "epoch": 1.89171974522293,
      "grad_norm": 1.4420032351612682,
      "learning_rate": 1.8277819437329577e-05,
      "loss": 0.5997,
      "step": 99
    },
    {
      "epoch": 1.910828025477707,
      "grad_norm": 1.2716603040915755,
      "learning_rate": 1.8223536398107177e-05,
      "loss": 0.5822,
      "step": 100
    },
    {
      "epoch": 1.929936305732484,
      "grad_norm": 1.255547700459211,
      "learning_rate": 1.8168494327875918e-05,
      "loss": 0.6029,
      "step": 101
    },
    {
      "epoch": 1.9490445859872612,
      "grad_norm": 1.5220214086917268,
      "learning_rate": 1.8112698307009506e-05,
      "loss": 0.5896,
      "step": 102
    },
    {
      "epoch": 1.9681528662420382,
      "grad_norm": 1.3560096197263136,
      "learning_rate": 1.8056153485471167e-05,
      "loss": 0.5807,
      "step": 103
    },
    {
      "epoch": 1.9872611464968153,
      "grad_norm": 1.4780227859635648,
      "learning_rate": 1.799886508233829e-05,
      "loss": 0.5835,
      "step": 104
    },
    {
      "epoch": 2.0063694267515926,
      "grad_norm": 1.2866181457082495,
      "learning_rate": 1.7940838385320732e-05,
      "loss": 0.4879,
      "step": 105
    },
    {
      "epoch": 2.0254777070063694,
      "grad_norm": 1.2198554612027852,
      "learning_rate": 1.788207875027274e-05,
      "loss": 0.3399,
      "step": 106
    },
    {
      "epoch": 2.0445859872611467,
      "grad_norm": 1.1638004148938548,
      "learning_rate": 1.7822591600698632e-05,
      "loss": 0.3263,
      "step": 107
    },
    {
      "epoch": 2.0636942675159236,
      "grad_norm": 1.6947898795763063,
      "learning_rate": 1.776238242725217e-05,
      "loss": 0.3745,
      "step": 108
    },
    {
      "epoch": 2.082802547770701,
      "grad_norm": 1.4731026648636676,
      "learning_rate": 1.7701456787229805e-05,
      "loss": 0.3178,
      "step": 109
    },
    {
      "epoch": 2.1019108280254777,
      "grad_norm": 1.2004912660652227,
      "learning_rate": 1.7639820304057745e-05,
      "loss": 0.3473,
      "step": 110
    },
    {
      "epoch": 2.121019108280255,
      "grad_norm": 1.2476423160721966,
      "learning_rate": 1.7577478666772886e-05,
      "loss": 0.3422,
      "step": 111
    },
    {
      "epoch": 2.140127388535032,
      "grad_norm": 1.193306106849365,
      "learning_rate": 1.751443762949772e-05,
      "loss": 0.3631,
      "step": 112
    },
    {
      "epoch": 2.159235668789809,
      "grad_norm": 1.0898051156699036,
      "learning_rate": 1.7450703010909263e-05,
      "loss": 0.3147,
      "step": 113
    },
    {
      "epoch": 2.178343949044586,
      "grad_norm": 1.3273161284848005,
      "learning_rate": 1.738628069370195e-05,
      "loss": 0.3676,
      "step": 114
    },
    {
      "epoch": 2.1974522292993632,
      "grad_norm": 1.2284704646661384,
      "learning_rate": 1.732117662404469e-05,
      "loss": 0.3665,
      "step": 115
    },
    {
      "epoch": 2.21656050955414,
      "grad_norm": 1.2096350181391013,
      "learning_rate": 1.7255396811032014e-05,
      "loss": 0.3573,
      "step": 116
    },
    {
      "epoch": 2.2356687898089174,
      "grad_norm": 1.2389791773164829,
      "learning_rate": 1.718894732612947e-05,
      "loss": 0.322,
      "step": 117
    },
    {
      "epoch": 2.254777070063694,
      "grad_norm": 1.111622582088061,
      "learning_rate": 1.712183430261319e-05,
      "loss": 0.3427,
      "step": 118
    },
    {
      "epoch": 2.2738853503184715,
      "grad_norm": 1.180404315815012,
      "learning_rate": 1.7054063935003813e-05,
      "loss": 0.3678,
      "step": 119
    },
    {
      "epoch": 2.2929936305732483,
      "grad_norm": 1.1375955561384368,
      "learning_rate": 1.698564247849473e-05,
      "loss": 0.3314,
      "step": 120
    },
    {
      "epoch": 2.3121019108280256,
      "grad_norm": 1.1098720646050049,
      "learning_rate": 1.691657624837472e-05,
      "loss": 0.3298,
      "step": 121
    },
    {
      "epoch": 2.3312101910828025,
      "grad_norm": 1.210067269931286,
      "learning_rate": 1.684687161944506e-05,
      "loss": 0.3599,
      "step": 122
    },
    {
      "epoch": 2.3503184713375798,
      "grad_norm": 1.1374262295996864,
      "learning_rate": 1.677653502543113e-05,
      "loss": 0.359,
      "step": 123
    },
    {
      "epoch": 2.3694267515923566,
      "grad_norm": 1.111840116040498,
      "learning_rate": 1.6705572958388576e-05,
      "loss": 0.3477,
      "step": 124
    },
    {
      "epoch": 2.388535031847134,
      "grad_norm": 1.1183070469663134,
      "learning_rate": 1.6633991968104095e-05,
      "loss": 0.3131,
      "step": 125
    },
    {
      "epoch": 2.4076433121019107,
      "grad_norm": 1.0815893984722618,
      "learning_rate": 1.6561798661490904e-05,
      "loss": 0.3548,
      "step": 126
    },
    {
      "epoch": 2.426751592356688,
      "grad_norm": 1.1159336391630241,
      "learning_rate": 1.6488999701978905e-05,
      "loss": 0.3491,
      "step": 127
    },
    {
      "epoch": 2.445859872611465,
      "grad_norm": 1.0974239822986869,
      "learning_rate": 1.6415601808899658e-05,
      "loss": 0.3177,
      "step": 128
    },
    {
      "epoch": 2.464968152866242,
      "grad_norm": 1.181955191907005,
      "learning_rate": 1.63416117568662e-05,
      "loss": 0.3594,
      "step": 129
    },
    {
      "epoch": 2.484076433121019,
      "grad_norm": 1.1059188589471078,
      "learning_rate": 1.6267036375147728e-05,
      "loss": 0.3424,
      "step": 130
    },
    {
      "epoch": 2.5031847133757963,
      "grad_norm": 1.1747469983595025,
      "learning_rate": 1.619188254703927e-05,
      "loss": 0.3567,
      "step": 131
    },
    {
      "epoch": 2.522292993630573,
      "grad_norm": 1.1490766705292863,
      "learning_rate": 1.6116157209226356e-05,
      "loss": 0.3231,
      "step": 132
    },
    {
      "epoch": 2.5414012738853504,
      "grad_norm": 1.147679134732286,
      "learning_rate": 1.6039867351144778e-05,
      "loss": 0.362,
      "step": 133
    },
    {
      "epoch": 2.5605095541401273,
      "grad_norm": 1.1049665212638309,
      "learning_rate": 1.5963020014335437e-05,
      "loss": 0.3236,
      "step": 134
    },
    {
      "epoch": 2.5796178343949046,
      "grad_norm": 1.2888513595868938,
      "learning_rate": 1.588562229179443e-05,
      "loss": 0.3675,
      "step": 135
    },
    {
      "epoch": 2.5987261146496814,
      "grad_norm": 1.153461765608754,
      "learning_rate": 1.5807681327318372e-05,
      "loss": 0.3488,
      "step": 136
    },
    {
      "epoch": 2.6178343949044587,
      "grad_norm": 1.1111216445665915,
      "learning_rate": 1.5729204314845002e-05,
      "loss": 0.3616,
      "step": 137
    },
    {
      "epoch": 2.6369426751592355,
      "grad_norm": 1.1610643251674635,
      "learning_rate": 1.56501984977892e-05,
      "loss": 0.3413,
      "step": 138
    },
    {
      "epoch": 2.656050955414013,
      "grad_norm": 1.0992867252068856,
      "learning_rate": 1.557067116837444e-05,
      "loss": 0.3049,
      "step": 139
    },
    {
      "epoch": 2.6751592356687897,
      "grad_norm": 1.1231428941054005,
      "learning_rate": 1.5490629666959668e-05,
      "loss": 0.3631,
      "step": 140
    },
    {
      "epoch": 2.694267515923567,
      "grad_norm": 1.1373285108620568,
      "learning_rate": 1.541008138136183e-05,
      "loss": 0.3233,
      "step": 141
    },
    {
      "epoch": 2.713375796178344,
      "grad_norm": 1.0712877988374556,
      "learning_rate": 1.5329033746173975e-05,
      "loss": 0.358,
      "step": 142
    },
    {
      "epoch": 2.732484076433121,
      "grad_norm": 1.0291514447015422,
      "learning_rate": 1.5247494242079024e-05,
      "loss": 0.37,
      "step": 143
    },
    {
      "epoch": 2.7515923566878984,
      "grad_norm": 1.1128200584801455,
      "learning_rate": 1.5165470395159314e-05,
      "loss": 0.3301,
      "step": 144
    },
    {
      "epoch": 2.770700636942675,
      "grad_norm": 1.1323881148146315,
      "learning_rate": 1.5082969776201948e-05,
      "loss": 0.3067,
      "step": 145
    },
    {
      "epoch": 2.789808917197452,
      "grad_norm": 1.0308429646272341,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.3289,
      "step": 146
    },
    {
      "epoch": 2.8089171974522293,
      "grad_norm": 1.0552781992097453,
      "learning_rate": 1.4916568724649688e-05,
      "loss": 0.3226,
      "step": 147
    },
    {
      "epoch": 2.8280254777070066,
      "grad_norm": 1.1483190940288472,
      "learning_rate": 1.483268365084351e-05,
      "loss": 0.341,
      "step": 148
    },
    {
      "epoch": 2.8471337579617835,
      "grad_norm": 1.1983775278717306,
      "learning_rate": 1.4748352521159492e-05,
      "loss": 0.3085,
      "step": 149
    },
    {
      "epoch": 2.8662420382165603,
      "grad_norm": 1.0020507576002027,
      "learning_rate": 1.466358311934654e-05,
      "loss": 0.3429,
      "step": 150
    },
    {
      "epoch": 2.8853503184713376,
      "grad_norm": 1.1972458385216016,
      "learning_rate": 1.4578383269606004e-05,
      "loss": 0.2933,
      "step": 151
    },
    {
      "epoch": 2.904458598726115,
      "grad_norm": 1.1589718554108799,
      "learning_rate": 1.4492760835869504e-05,
      "loss": 0.345,
      "step": 152
    },
    {
      "epoch": 2.9235668789808917,
      "grad_norm": 1.1444814297688288,
      "learning_rate": 1.4406723721073088e-05,
      "loss": 0.3898,
      "step": 153
    },
    {
      "epoch": 2.9426751592356686,
      "grad_norm": 1.1048277109165345,
      "learning_rate": 1.4320279866427798e-05,
      "loss": 0.3674,
      "step": 154
    },
    {
      "epoch": 2.961783439490446,
      "grad_norm": 1.14131770709984,
      "learning_rate": 1.4233437250686695e-05,
      "loss": 0.3271,
      "step": 155
    },
    {
      "epoch": 2.980891719745223,
      "grad_norm": 1.0730828437212399,
      "learning_rate": 1.4146203889408418e-05,
      "loss": 0.3342,
      "step": 156
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.0986481112547453,
      "learning_rate": 1.4058587834217356e-05,
      "loss": 0.3276,
      "step": 157
    },
    {
      "epoch": 3.0191082802547773,
      "grad_norm": 0.9727813760048125,
      "learning_rate": 1.3970597172060482e-05,
      "loss": 0.2086,
      "step": 158
    },
    {
      "epoch": 3.038216560509554,
      "grad_norm": 0.9351469513287544,
      "learning_rate": 1.3882240024460928e-05,
      "loss": 0.2144,
      "step": 159
    },
    {
      "epoch": 3.0573248407643314,
      "grad_norm": 0.9596216840602939,
      "learning_rate": 1.3793524546768358e-05,
      "loss": 0.1854,
      "step": 160
    },
    {
      "epoch": 3.0764331210191083,
      "grad_norm": 1.0070708882490527,
      "learning_rate": 1.3704458927406261e-05,
      "loss": 0.1979,
      "step": 161
    },
    {
      "epoch": 3.0955414012738856,
      "grad_norm": 1.1947399346702603,
      "learning_rate": 1.3615051387116131e-05,
      "loss": 0.2177,
      "step": 162
    },
    {
      "epoch": 3.1146496815286624,
      "grad_norm": 0.9221956509184932,
      "learning_rate": 1.3525310178198707e-05,
      "loss": 0.2099,
      "step": 163
    },
    {
      "epoch": 3.1337579617834397,
      "grad_norm": 0.9410399106958189,
      "learning_rate": 1.3435243583752294e-05,
      "loss": 0.2038,
      "step": 164
    },
    {
      "epoch": 3.1528662420382165,
      "grad_norm": 0.9814965993205671,
      "learning_rate": 1.3344859916908206e-05,
      "loss": 0.2378,
      "step": 165
    },
    {
      "epoch": 3.171974522292994,
      "grad_norm": 0.9231538478983656,
      "learning_rate": 1.325416752006351e-05,
      "loss": 0.1767,
      "step": 166
    },
    {
      "epoch": 3.1910828025477707,
      "grad_norm": 0.8950308599796192,
      "learning_rate": 1.3163174764110985e-05,
      "loss": 0.1753,
      "step": 167
    },
    {
      "epoch": 3.210191082802548,
      "grad_norm": 0.9109948031677837,
      "learning_rate": 1.3071890047666498e-05,
      "loss": 0.2195,
      "step": 168
    },
    {
      "epoch": 3.229299363057325,
      "grad_norm": 0.9602237468554572,
      "learning_rate": 1.2980321796293838e-05,
      "loss": 0.2231,
      "step": 169
    },
    {
      "epoch": 3.248407643312102,
      "grad_norm": 0.9472126025942549,
      "learning_rate": 1.288847846172701e-05,
      "loss": 0.233,
      "step": 170
    },
    {
      "epoch": 3.267515923566879,
      "grad_norm": 0.9448974877228525,
      "learning_rate": 1.2796368521090143e-05,
      "loss": 0.2293,
      "step": 171
    },
    {
      "epoch": 3.286624203821656,
      "grad_norm": 0.8735754301637381,
      "learning_rate": 1.2704000476115079e-05,
      "loss": 0.1923,
      "step": 172
    },
    {
      "epoch": 3.305732484076433,
      "grad_norm": 0.9320466823822048,
      "learning_rate": 1.2611382852356632e-05,
      "loss": 0.241,
      "step": 173
    },
    {
      "epoch": 3.3248407643312103,
      "grad_norm": 0.8692677947193066,
      "learning_rate": 1.2518524198405699e-05,
      "loss": 0.227,
      "step": 174
    },
    {
      "epoch": 3.343949044585987,
      "grad_norm": 0.8801146416841564,
      "learning_rate": 1.2425433085100224e-05,
      "loss": 0.1892,
      "step": 175
    },
    {
      "epoch": 3.3630573248407645,
      "grad_norm": 1.024131338299322,
      "learning_rate": 1.233211810473411e-05,
      "loss": 0.2217,
      "step": 176
    },
    {
      "epoch": 3.3821656050955413,
      "grad_norm": 0.8806375469042244,
      "learning_rate": 1.2238587870264152e-05,
      "loss": 0.2289,
      "step": 177
    },
    {
      "epoch": 3.4012738853503186,
      "grad_norm": 0.8949628237735877,
      "learning_rate": 1.2144851014515055e-05,
      "loss": 0.1952,
      "step": 178
    },
    {
      "epoch": 3.4203821656050954,
      "grad_norm": 0.9317408522256906,
      "learning_rate": 1.2050916189382646e-05,
      "loss": 0.2116,
      "step": 179
    },
    {
      "epoch": 3.4394904458598727,
      "grad_norm": 0.8998850494532862,
      "learning_rate": 1.1956792065035281e-05,
      "loss": 0.1798,
      "step": 180
    },
    {
      "epoch": 3.4585987261146496,
      "grad_norm": 0.854442662864881,
      "learning_rate": 1.1862487329113606e-05,
      "loss": 0.213,
      "step": 181
    },
    {
      "epoch": 3.477707006369427,
      "grad_norm": 0.904951729039575,
      "learning_rate": 1.1768010685928686e-05,
      "loss": 0.1573,
      "step": 182
    },
    {
      "epoch": 3.4968152866242037,
      "grad_norm": 0.8905053917581376,
      "learning_rate": 1.1673370855658592e-05,
      "loss": 0.1893,
      "step": 183
    },
    {
      "epoch": 3.515923566878981,
      "grad_norm": 0.8324598623669239,
      "learning_rate": 1.1578576573543541e-05,
      "loss": 0.2159,
      "step": 184
    },
    {
      "epoch": 3.535031847133758,
      "grad_norm": 1.0077484019543617,
      "learning_rate": 1.1483636589079627e-05,
      "loss": 0.1904,
      "step": 185
    },
    {
      "epoch": 3.554140127388535,
      "grad_norm": 0.8809032468402521,
      "learning_rate": 1.1388559665211241e-05,
      "loss": 0.2147,
      "step": 186
    },
    {
      "epoch": 3.573248407643312,
      "grad_norm": 0.8655097326224989,
      "learning_rate": 1.1293354577522264e-05,
      "loss": 0.2162,
      "step": 187
    },
    {
      "epoch": 3.5923566878980893,
      "grad_norm": 0.8789799920919437,
      "learning_rate": 1.1198030113426074e-05,
      "loss": 0.19,
      "step": 188
    },
    {
      "epoch": 3.611464968152866,
      "grad_norm": 0.903268539210289,
      "learning_rate": 1.1102595071354471e-05,
      "loss": 0.1852,
      "step": 189
    },
    {
      "epoch": 3.6305732484076434,
      "grad_norm": 0.9455765723436995,
      "learning_rate": 1.1007058259945584e-05,
      "loss": 0.2172,
      "step": 190
    },
    {
      "epoch": 3.6496815286624202,
      "grad_norm": 0.854386751908126,
      "learning_rate": 1.0911428497230834e-05,
      "loss": 0.2218,
      "step": 191
    },
    {
      "epoch": 3.6687898089171975,
      "grad_norm": 0.8554909616748702,
      "learning_rate": 1.0815714609821027e-05,
      "loss": 0.2033,
      "step": 192
    },
    {
      "epoch": 3.6878980891719744,
      "grad_norm": 0.9177260635164998,
      "learning_rate": 1.0719925432091671e-05,
      "loss": 0.2166,
      "step": 193
    },
    {
      "epoch": 3.7070063694267517,
      "grad_norm": 0.8574723299072876,
      "learning_rate": 1.0624069805367558e-05,
      "loss": 0.208,
      "step": 194
    },
    {
      "epoch": 3.7261146496815285,
      "grad_norm": 0.8427079553036521,
      "learning_rate": 1.0528156577106703e-05,
      "loss": 0.2083,
      "step": 195
    },
    {
      "epoch": 3.745222929936306,
      "grad_norm": 0.8880834494398775,
      "learning_rate": 1.043219460008374e-05,
      "loss": 0.2112,
      "step": 196
    },
    {
      "epoch": 3.7643312101910826,
      "grad_norm": 0.895077225646693,
      "learning_rate": 1.0336192731572805e-05,
      "loss": 0.2181,
      "step": 197
    },
    {
      "epoch": 3.78343949044586,
      "grad_norm": 0.8535144125727894,
      "learning_rate": 1.0240159832530007e-05,
      "loss": 0.1875,
      "step": 198
    },
    {
      "epoch": 3.802547770700637,
      "grad_norm": 0.8852101248249045,
      "learning_rate": 1.0144104766775574e-05,
      "loss": 0.2243,
      "step": 199
    },
    {
      "epoch": 3.821656050955414,
      "grad_norm": 0.8844111940812327,
      "learning_rate": 1.004803640017571e-05,
      "loss": 0.2107,
      "step": 200
    },
    {
      "epoch": 3.840764331210191,
      "grad_norm": 0.8703098886346262,
      "learning_rate": 9.951963599824294e-06,
      "loss": 0.2522,
      "step": 201
    },
    {
      "epoch": 3.859872611464968,
      "grad_norm": 0.8379951683247892,
      "learning_rate": 9.855895233224431e-06,
      "loss": 0.2271,
      "step": 202
    },
    {
      "epoch": 3.8789808917197455,
      "grad_norm": 0.8485804093654242,
      "learning_rate": 9.759840167469995e-06,
      "loss": 0.2111,
      "step": 203
    },
    {
      "epoch": 3.8980891719745223,
      "grad_norm": 0.9209244983709053,
      "learning_rate": 9.663807268427197e-06,
      "loss": 0.2085,
      "step": 204
    },
    {
      "epoch": 3.917197452229299,
      "grad_norm": 0.8341014031710875,
      "learning_rate": 9.56780539991626e-06,
      "loss": 0.2054,
      "step": 205
    },
    {
      "epoch": 3.9363057324840764,
      "grad_norm": 0.830481999318346,
      "learning_rate": 9.471843422893299e-06,
      "loss": 0.1696,
      "step": 206
    },
    {
      "epoch": 3.9554140127388537,
      "grad_norm": 0.870075031630067,
      "learning_rate": 9.375930194632447e-06,
      "loss": 0.2088,
      "step": 207
    },
    {
      "epoch": 3.9745222929936306,
      "grad_norm": 0.8949036645858451,
      "learning_rate": 9.28007456790833e-06,
      "loss": 0.1856,
      "step": 208
    },
    {
      "epoch": 3.9936305732484074,
      "grad_norm": 0.8409816207558508,
      "learning_rate": 9.184285390178978e-06,
      "loss": 0.1888,
      "step": 209
    },
    {
      "epoch": 4.012738853503185,
      "grad_norm": 0.7770572252155418,
      "learning_rate": 9.08857150276917e-06,
      "loss": 0.1449,
      "step": 210
    },
    {
      "epoch": 4.031847133757962,
      "grad_norm": 0.7207065511670053,
      "learning_rate": 8.992941740054418e-06,
      "loss": 0.1393,
      "step": 211
    },
    {
      "epoch": 4.050955414012739,
      "grad_norm": 0.6573904924432173,
      "learning_rate": 8.897404928645529e-06,
      "loss": 0.1382,
      "step": 212
    },
    {
      "epoch": 4.070063694267516,
      "grad_norm": 0.5998961182192761,
      "learning_rate": 8.80196988657393e-06,
      "loss": 0.1179,
      "step": 213
    },
    {
      "epoch": 4.089171974522293,
      "grad_norm": 0.7353709913341053,
      "learning_rate": 8.706645422477739e-06,
      "loss": 0.1296,
      "step": 214
    },
    {
      "epoch": 4.10828025477707,
      "grad_norm": 0.8195740194511104,
      "learning_rate": 8.611440334788762e-06,
      "loss": 0.1485,
      "step": 215
    },
    {
      "epoch": 4.127388535031847,
      "grad_norm": 0.7112832621893119,
      "learning_rate": 8.516363410920376e-06,
      "loss": 0.1182,
      "step": 216
    },
    {
      "epoch": 4.146496815286624,
      "grad_norm": 0.7215183552128456,
      "learning_rate": 8.42142342645646e-06,
      "loss": 0.1575,
      "step": 217
    },
    {
      "epoch": 4.165605095541402,
      "grad_norm": 0.6361659814580954,
      "learning_rate": 8.326629144341408e-06,
      "loss": 0.1021,
      "step": 218
    },
    {
      "epoch": 4.1847133757961785,
      "grad_norm": 0.6779558705834515,
      "learning_rate": 8.231989314071318e-06,
      "loss": 0.1243,
      "step": 219
    },
    {
      "epoch": 4.203821656050955,
      "grad_norm": 0.6707598534659711,
      "learning_rate": 8.137512670886397e-06,
      "loss": 0.0998,
      "step": 220
    },
    {
      "epoch": 4.222929936305732,
      "grad_norm": 0.6470471778032241,
      "learning_rate": 8.043207934964722e-06,
      "loss": 0.148,
      "step": 221
    },
    {
      "epoch": 4.24203821656051,
      "grad_norm": 0.661244303933043,
      "learning_rate": 7.949083810617358e-06,
      "loss": 0.1246,
      "step": 222
    },
    {
      "epoch": 4.261146496815287,
      "grad_norm": 0.6142982587848405,
      "learning_rate": 7.855148985484946e-06,
      "loss": 0.1198,
      "step": 223
    },
    {
      "epoch": 4.280254777070064,
      "grad_norm": 0.6629685162864739,
      "learning_rate": 7.761412129735853e-06,
      "loss": 0.1212,
      "step": 224
    },
    {
      "epoch": 4.2993630573248405,
      "grad_norm": 0.6952845083203738,
      "learning_rate": 7.667881895265895e-06,
      "loss": 0.1298,
      "step": 225
    },
    {
      "epoch": 4.318471337579618,
      "grad_norm": 0.6333882044772869,
      "learning_rate": 7.574566914899779e-06,
      "loss": 0.1225,
      "step": 226
    },
    {
      "epoch": 4.337579617834395,
      "grad_norm": 0.6333292253877881,
      "learning_rate": 7.481475801594302e-06,
      "loss": 0.0954,
      "step": 227
    },
    {
      "epoch": 4.356687898089172,
      "grad_norm": 0.6771048371022191,
      "learning_rate": 7.388617147643371e-06,
      "loss": 0.1156,
      "step": 228
    },
    {
      "epoch": 4.375796178343949,
      "grad_norm": 0.6522245094480688,
      "learning_rate": 7.295999523884921e-06,
      "loss": 0.1123,
      "step": 229
    },
    {
      "epoch": 4.3949044585987265,
      "grad_norm": 0.6404010369130144,
      "learning_rate": 7.203631478909857e-06,
      "loss": 0.113,
      "step": 230
    },
    {
      "epoch": 4.414012738853503,
      "grad_norm": 0.6917698578909479,
      "learning_rate": 7.111521538272997e-06,
      "loss": 0.1179,
      "step": 231
    },
    {
      "epoch": 4.43312101910828,
      "grad_norm": 0.6432868110985619,
      "learning_rate": 7.019678203706164e-06,
      "loss": 0.1284,
      "step": 232
    },
    {
      "epoch": 4.452229299363057,
      "grad_norm": 0.5862011184671926,
      "learning_rate": 6.928109952333506e-06,
      "loss": 0.1318,
      "step": 233
    },
    {
      "epoch": 4.471337579617835,
      "grad_norm": 0.5983391729216361,
      "learning_rate": 6.83682523588902e-06,
      "loss": 0.1361,
      "step": 234
    },
    {
      "epoch": 4.490445859872612,
      "grad_norm": 0.7384156441048828,
      "learning_rate": 6.745832479936492e-06,
      "loss": 0.1202,
      "step": 235
    },
    {
      "epoch": 4.509554140127388,
      "grad_norm": 0.5801257613679321,
      "learning_rate": 6.655140083091794e-06,
      "loss": 0.1142,
      "step": 236
    },
    {
      "epoch": 4.528662420382165,
      "grad_norm": 0.6644460629119762,
      "learning_rate": 6.564756416247712e-06,
      "loss": 0.1132,
      "step": 237
    },
    {
      "epoch": 4.547770700636943,
      "grad_norm": 0.6338897010453224,
      "learning_rate": 6.474689821801295e-06,
      "loss": 0.1224,
      "step": 238
    },
    {
      "epoch": 4.56687898089172,
      "grad_norm": 0.6265192886702466,
      "learning_rate": 6.384948612883872e-06,
      "loss": 0.1081,
      "step": 239
    },
    {
      "epoch": 4.585987261146497,
      "grad_norm": 0.6375524943746127,
      "learning_rate": 6.2955410725937405e-06,
      "loss": 0.142,
      "step": 240
    },
    {
      "epoch": 4.6050955414012735,
      "grad_norm": 0.6024003247740342,
      "learning_rate": 6.206475453231644e-06,
      "loss": 0.1098,
      "step": 241
    },
    {
      "epoch": 4.624203821656051,
      "grad_norm": 0.5422627616710795,
      "learning_rate": 6.117759975539075e-06,
      "loss": 0.1084,
      "step": 242
    },
    {
      "epoch": 4.643312101910828,
      "grad_norm": 0.6082249236985866,
      "learning_rate": 6.029402827939519e-06,
      "loss": 0.1356,
      "step": 243
    },
    {
      "epoch": 4.662420382165605,
      "grad_norm": 0.6409844864728751,
      "learning_rate": 5.941412165782645e-06,
      "loss": 0.1072,
      "step": 244
    },
    {
      "epoch": 4.681528662420382,
      "grad_norm": 0.6047083213755711,
      "learning_rate": 5.853796110591583e-06,
      "loss": 0.1199,
      "step": 245
    },
    {
      "epoch": 4.7006369426751595,
      "grad_norm": 0.578772450983975,
      "learning_rate": 5.766562749313309e-06,
      "loss": 0.1397,
      "step": 246
    },
    {
      "epoch": 4.719745222929936,
      "grad_norm": 0.651473421215594,
      "learning_rate": 5.6797201335722064e-06,
      "loss": 0.1534,
      "step": 247
    },
    {
      "epoch": 4.738853503184713,
      "grad_norm": 0.5959248362521793,
      "learning_rate": 5.593276278926912e-06,
      "loss": 0.1258,
      "step": 248
    },
    {
      "epoch": 4.757961783439491,
      "grad_norm": 0.5896005713691379,
      "learning_rate": 5.507239164130501e-06,
      "loss": 0.1215,
      "step": 249
    },
    {
      "epoch": 4.777070063694268,
      "grad_norm": 0.6246354502017855,
      "learning_rate": 5.421616730394e-06,
      "loss": 0.1421,
      "step": 250
    },
    {
      "epoch": 4.796178343949045,
      "grad_norm": 0.5802170532932973,
      "learning_rate": 5.336416880653461e-06,
      "loss": 0.1217,
      "step": 251
    },
    {
      "epoch": 4.8152866242038215,
      "grad_norm": 0.6301892377174323,
      "learning_rate": 5.251647478840511e-06,
      "loss": 0.1253,
      "step": 252
    },
    {
      "epoch": 4.834394904458598,
      "grad_norm": 0.6324248660174402,
      "learning_rate": 5.167316349156495e-06,
      "loss": 0.1049,
      "step": 253
    },
    {
      "epoch": 4.853503184713376,
      "grad_norm": 0.6270125086141117,
      "learning_rate": 5.083431275350312e-06,
      "loss": 0.0929,
      "step": 254
    },
    {
      "epoch": 4.872611464968153,
      "grad_norm": 0.589626433719279,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.1183,
      "step": 255
    },
    {
      "epoch": 4.89171974522293,
      "grad_norm": 0.6333892873260404,
      "learning_rate": 4.917030223798057e-06,
      "loss": 0.1385,
      "step": 256
    },
    {
      "epoch": 4.9108280254777075,
      "grad_norm": 0.6049252674697725,
      "learning_rate": 4.834529604840686e-06,
      "loss": 0.106,
      "step": 257
    },
    {
      "epoch": 4.929936305732484,
      "grad_norm": 0.5901985327547369,
      "learning_rate": 4.7525057579209775e-06,
      "loss": 0.096,
      "step": 258
    },
    {
      "epoch": 4.949044585987261,
      "grad_norm": 0.5559455099052016,
      "learning_rate": 4.670966253826027e-06,
      "loss": 0.1009,
      "step": 259
    },
    {
      "epoch": 4.968152866242038,
      "grad_norm": 0.6211677396356956,
      "learning_rate": 4.589918618638173e-06,
      "loss": 0.1286,
      "step": 260
    },
    {
      "epoch": 4.987261146496815,
      "grad_norm": 0.5873071270705379,
      "learning_rate": 4.5093703330403385e-06,
      "loss": 0.1145,
      "step": 261
    },
    {
      "epoch": 5.006369426751593,
      "grad_norm": 0.469843710090882,
      "learning_rate": 4.429328831625565e-06,
      "loss": 0.1196,
      "step": 262
    },
    {
      "epoch": 5.025477707006369,
      "grad_norm": 0.43945316537836715,
      "learning_rate": 4.349801502210801e-06,
      "loss": 0.0876,
      "step": 263
    },
    {
      "epoch": 5.044585987261146,
      "grad_norm": 0.37285253307993454,
      "learning_rate": 4.270795685155001e-06,
      "loss": 0.0992,
      "step": 264
    },
    {
      "epoch": 5.063694267515924,
      "grad_norm": 0.36075563930480936,
      "learning_rate": 4.192318672681631e-06,
      "loss": 0.0811,
      "step": 265
    },
    {
      "epoch": 5.082802547770701,
      "grad_norm": 0.36018265438822283,
      "learning_rate": 4.1143777082055715e-06,
      "loss": 0.0765,
      "step": 266
    },
    {
      "epoch": 5.101910828025478,
      "grad_norm": 0.40524551115007706,
      "learning_rate": 4.036979985664566e-06,
      "loss": 0.0993,
      "step": 267
    },
    {
      "epoch": 5.1210191082802545,
      "grad_norm": 0.3902710741865906,
      "learning_rate": 3.960132648855226e-06,
      "loss": 0.0704,
      "step": 268
    },
    {
      "epoch": 5.140127388535032,
      "grad_norm": 0.4257168548853022,
      "learning_rate": 3.883842790773647e-06,
      "loss": 0.078,
      "step": 269
    },
    {
      "epoch": 5.159235668789809,
      "grad_norm": 0.42709657802643913,
      "learning_rate": 3.8081174529607346e-06,
      "loss": 0.0901,
      "step": 270
    },
    {
      "epoch": 5.178343949044586,
      "grad_norm": 0.4464466835988097,
      "learning_rate": 3.732963624852275e-06,
      "loss": 0.1005,
      "step": 271
    },
    {
      "epoch": 5.197452229299363,
      "grad_norm": 0.43813407531215287,
      "learning_rate": 3.6583882431338047e-06,
      "loss": 0.0794,
      "step": 272
    },
    {
      "epoch": 5.2165605095541405,
      "grad_norm": 0.42534884458138056,
      "learning_rate": 3.584398191100341e-06,
      "loss": 0.0804,
      "step": 273
    },
    {
      "epoch": 5.235668789808917,
      "grad_norm": 0.3904545598805854,
      "learning_rate": 3.511000298021098e-06,
      "loss": 0.0746,
      "step": 274
    },
    {
      "epoch": 5.254777070063694,
      "grad_norm": 0.38122125804588974,
      "learning_rate": 3.4382013385090985e-06,
      "loss": 0.0771,
      "step": 275
    },
    {
      "epoch": 5.273885350318471,
      "grad_norm": 0.3910636832792095,
      "learning_rate": 3.3660080318959043e-06,
      "loss": 0.0763,
      "step": 276
    },
    {
      "epoch": 5.292993630573249,
      "grad_norm": 0.3943252213563477,
      "learning_rate": 3.2944270416114256e-06,
      "loss": 0.0824,
      "step": 277
    },
    {
      "epoch": 5.312101910828026,
      "grad_norm": 0.4153720046814973,
      "learning_rate": 3.223464974568874e-06,
      "loss": 0.0808,
      "step": 278
    },
    {
      "epoch": 5.3312101910828025,
      "grad_norm": 0.37449475995730286,
      "learning_rate": 3.153128380554941e-06,
      "loss": 0.0669,
      "step": 279
    },
    {
      "epoch": 5.350318471337579,
      "grad_norm": 0.3678087641481315,
      "learning_rate": 3.0834237516252817e-06,
      "loss": 0.078,
      "step": 280
    },
    {
      "epoch": 5.369426751592357,
      "grad_norm": 0.39958648443736494,
      "learning_rate": 3.0143575215052732e-06,
      "loss": 0.1098,
      "step": 281
    },
    {
      "epoch": 5.388535031847134,
      "grad_norm": 0.44061456973739005,
      "learning_rate": 2.94593606499619e-06,
      "loss": 0.0736,
      "step": 282
    },
    {
      "epoch": 5.407643312101911,
      "grad_norm": 0.40798693413680315,
      "learning_rate": 2.878165697386812e-06,
      "loss": 0.0743,
      "step": 283
    },
    {
      "epoch": 5.426751592356688,
      "grad_norm": 0.40002583724881186,
      "learning_rate": 2.8110526738705345e-06,
      "loss": 0.0973,
      "step": 284
    },
    {
      "epoch": 5.445859872611465,
      "grad_norm": 0.4027836230760282,
      "learning_rate": 2.7446031889679893e-06,
      "loss": 0.0863,
      "step": 285
    },
    {
      "epoch": 5.464968152866242,
      "grad_norm": 0.4223422247231742,
      "learning_rate": 2.678823375955314e-06,
      "loss": 0.0827,
      "step": 286
    },
    {
      "epoch": 5.484076433121019,
      "grad_norm": 0.3753472338541051,
      "learning_rate": 2.6137193062980506e-06,
      "loss": 0.0927,
      "step": 287
    },
    {
      "epoch": 5.503184713375796,
      "grad_norm": 0.4371880580691285,
      "learning_rate": 2.5492969890907383e-06,
      "loss": 0.0862,
      "step": 288
    },
    {
      "epoch": 5.522292993630574,
      "grad_norm": 0.43678290237390494,
      "learning_rate": 2.485562370502279e-06,
      "loss": 0.1076,
      "step": 289
    },
    {
      "epoch": 5.54140127388535,
      "grad_norm": 0.4686470682614719,
      "learning_rate": 2.4225213332271203e-06,
      "loss": 0.0658,
      "step": 290
    },
    {
      "epoch": 5.560509554140127,
      "grad_norm": 0.399090511069466,
      "learning_rate": 2.3601796959422585e-06,
      "loss": 0.0867,
      "step": 291
    },
    {
      "epoch": 5.579617834394904,
      "grad_norm": 0.38033742520384534,
      "learning_rate": 2.2985432127701945e-06,
      "loss": 0.0939,
      "step": 292
    },
    {
      "epoch": 5.598726114649682,
      "grad_norm": 0.4195573707985727,
      "learning_rate": 2.2376175727478346e-06,
      "loss": 0.0792,
      "step": 293
    },
    {
      "epoch": 5.617834394904459,
      "grad_norm": 0.39410042613894186,
      "learning_rate": 2.1774083993013715e-06,
      "loss": 0.0939,
      "step": 294
    },
    {
      "epoch": 5.6369426751592355,
      "grad_norm": 0.42028801078240485,
      "learning_rate": 2.1179212497272582e-06,
      "loss": 0.0914,
      "step": 295
    },
    {
      "epoch": 5.656050955414012,
      "grad_norm": 0.4373301107215838,
      "learning_rate": 2.0591616146792705e-06,
      "loss": 0.0612,
      "step": 296
    },
    {
      "epoch": 5.67515923566879,
      "grad_norm": 0.42230760844766996,
      "learning_rate": 2.0011349176617133e-06,
      "loss": 0.0745,
      "step": 297
    },
    {
      "epoch": 5.694267515923567,
      "grad_norm": 0.43204660972362535,
      "learning_rate": 1.9438465145288377e-06,
      "loss": 0.0862,
      "step": 298
    },
    {
      "epoch": 5.713375796178344,
      "grad_norm": 0.42621175764084096,
      "learning_rate": 1.8873016929904942e-06,
      "loss": 0.0902,
      "step": 299
    },
    {
      "epoch": 5.732484076433121,
      "grad_norm": 0.3834518525429406,
      "learning_rate": 1.8315056721240831e-06,
      "loss": 0.0651,
      "step": 300
    },
    {
      "epoch": 5.751592356687898,
      "grad_norm": 0.37147796573780106,
      "learning_rate": 1.7764636018928249e-06,
      "loss": 0.0725,
      "step": 301
    },
    {
      "epoch": 5.770700636942675,
      "grad_norm": 0.3537695521438638,
      "learning_rate": 1.722180562670428e-06,
      "loss": 0.0805,
      "step": 302
    },
    {
      "epoch": 5.789808917197452,
      "grad_norm": 0.3894501723809,
      "learning_rate": 1.6686615647721638e-06,
      "loss": 0.0774,
      "step": 303
    },
    {
      "epoch": 5.80891719745223,
      "grad_norm": 0.37961916616617003,
      "learning_rate": 1.6159115479924259e-06,
      "loss": 0.0927,
      "step": 304
    },
    {
      "epoch": 5.828025477707007,
      "grad_norm": 0.3872561941363768,
      "learning_rate": 1.5639353811487744e-06,
      "loss": 0.0741,
      "step": 305
    },
    {
      "epoch": 5.8471337579617835,
      "grad_norm": 0.34833321067453243,
      "learning_rate": 1.5127378616325606e-06,
      "loss": 0.0744,
      "step": 306
    },
    {
      "epoch": 5.86624203821656,
      "grad_norm": 0.3855037708428381,
      "learning_rate": 1.462323714966114e-06,
      "loss": 0.0908,
      "step": 307
    },
    {
      "epoch": 5.885350318471337,
      "grad_norm": 0.40297529412539035,
      "learning_rate": 1.4126975943665844e-06,
      "loss": 0.0998,
      "step": 308
    },
    {
      "epoch": 5.904458598726115,
      "grad_norm": 0.4029126709486481,
      "learning_rate": 1.3638640803164516e-06,
      "loss": 0.0911,
      "step": 309
    },
    {
      "epoch": 5.923566878980892,
      "grad_norm": 0.4258982687557646,
      "learning_rate": 1.3158276801407432e-06,
      "loss": 0.0849,
      "step": 310
    },
    {
      "epoch": 5.942675159235669,
      "grad_norm": 0.3553371859042462,
      "learning_rate": 1.2685928275910142e-06,
      "loss": 0.0676,
      "step": 311
    },
    {
      "epoch": 5.961783439490446,
      "grad_norm": 0.3979066431111897,
      "learning_rate": 1.222163882436107e-06,
      "loss": 0.0509,
      "step": 312
    },
    {
      "epoch": 5.980891719745223,
      "grad_norm": 0.4141884197499124,
      "learning_rate": 1.1765451300597574e-06,
      "loss": 0.089,
      "step": 313
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.3770050896807073,
      "learning_rate": 1.1317407810650372e-06,
      "loss": 0.0544,
      "step": 314
    },
    {
      "epoch": 6.019108280254777,
      "grad_norm": 0.32674830108323855,
      "learning_rate": 1.0877549708857228e-06,
      "loss": 0.0603,
      "step": 315
    },
    {
      "epoch": 6.038216560509555,
      "grad_norm": 0.3123835583619939,
      "learning_rate": 1.0445917594046073e-06,
      "loss": 0.0645,
      "step": 316
    },
    {
      "epoch": 6.057324840764331,
      "grad_norm": 0.2533884278161009,
      "learning_rate": 1.0022551305787564e-06,
      "loss": 0.0571,
      "step": 317
    },
    {
      "epoch": 6.076433121019108,
      "grad_norm": 0.25372043981270836,
      "learning_rate": 9.607489920717983e-07,
      "loss": 0.0677,
      "step": 318
    },
    {
      "epoch": 6.095541401273885,
      "grad_norm": 0.28932834883810127,
      "learning_rate": 9.200771748932513e-07,
      "loss": 0.0665,
      "step": 319
    },
    {
      "epoch": 6.114649681528663,
      "grad_norm": 0.29574813210798573,
      "learning_rate": 8.802434330449128e-07,
      "loss": 0.0605,
      "step": 320
    },
    {
      "epoch": 6.13375796178344,
      "grad_norm": 0.32412205688268597,
      "learning_rate": 8.412514431743657e-07,
      "loss": 0.0759,
      "step": 321
    },
    {
      "epoch": 6.1528662420382165,
      "grad_norm": 0.298971869628302,
      "learning_rate": 8.031048042356393e-07,
      "loss": 0.0944,
      "step": 322
    },
    {
      "epoch": 6.171974522292993,
      "grad_norm": 0.28058094255268096,
      "learning_rate": 7.65807037157007e-07,
      "loss": 0.0737,
      "step": 323
    },
    {
      "epoch": 6.191082802547771,
      "grad_norm": 0.26273990851994955,
      "learning_rate": 7.293615845160196e-07,
      "loss": 0.0801,
      "step": 324
    },
    {
      "epoch": 6.210191082802548,
      "grad_norm": 0.26749718281850127,
      "learning_rate": 6.937718102217461e-07,
      "loss": 0.0843,
      "step": 325
    },
    {
      "epoch": 6.229299363057325,
      "grad_norm": 0.27788643017096765,
      "learning_rate": 6.590409992042957e-07,
      "loss": 0.0479,
      "step": 326
    },
    {
      "epoch": 6.248407643312102,
      "grad_norm": 0.32662056891673275,
      "learning_rate": 6.251723571116031e-07,
      "loss": 0.0996,
      "step": 327
    },
    {
      "epoch": 6.267515923566879,
      "grad_norm": 0.34232243605883206,
      "learning_rate": 5.921690100135713e-07,
      "loss": 0.1195,
      "step": 328
    },
    {
      "epoch": 6.286624203821656,
      "grad_norm": 0.26341475326491803,
      "learning_rate": 5.600340041135133e-07,
      "loss": 0.041,
      "step": 329
    },
    {
      "epoch": 6.305732484076433,
      "grad_norm": 0.32091863933742903,
      "learning_rate": 5.287703054670012e-07,
      "loss": 0.0896,
      "step": 330
    },
    {
      "epoch": 6.32484076433121,
      "grad_norm": 0.26324953620812674,
      "learning_rate": 4.983807997080925e-07,
      "loss": 0.0507,
      "step": 331
    },
    {
      "epoch": 6.343949044585988,
      "grad_norm": 0.31884114580503486,
      "learning_rate": 4.6886829178299676e-07,
      "loss": 0.0916,
      "step": 332
    },
    {
      "epoch": 6.3630573248407645,
      "grad_norm": 0.2987800921220235,
      "learning_rate": 4.402355056911656e-07,
      "loss": 0.0787,
      "step": 333
    },
    {
      "epoch": 6.382165605095541,
      "grad_norm": 0.27285623552894217,
      "learning_rate": 4.124850842338779e-07,
      "loss": 0.0898,
      "step": 334
    },
    {
      "epoch": 6.401273885350318,
      "grad_norm": 0.32925345757232627,
      "learning_rate": 3.8561958877030957e-07,
      "loss": 0.082,
      "step": 335
    },
    {
      "epoch": 6.420382165605096,
      "grad_norm": 0.25722868705327834,
      "learning_rate": 3.5964149898111587e-07,
      "loss": 0.0537,
      "step": 336
    },
    {
      "epoch": 6.439490445859873,
      "grad_norm": 0.309845365274313,
      "learning_rate": 3.345532126395579e-07,
      "loss": 0.0579,
      "step": 337
    },
    {
      "epoch": 6.45859872611465,
      "grad_norm": 0.2977940307136604,
      "learning_rate": 3.1035704539019384e-07,
      "loss": 0.0726,
      "step": 338
    },
    {
      "epoch": 6.477707006369426,
      "grad_norm": 0.31025376844846553,
      "learning_rate": 2.870552305351382e-07,
      "loss": 0.0706,
      "step": 339
    },
    {
      "epoch": 6.496815286624204,
      "grad_norm": 0.2554835376087535,
      "learning_rate": 2.646499188279328e-07,
      "loss": 0.0424,
      "step": 340
    },
    {
      "epoch": 6.515923566878981,
      "grad_norm": 0.3180912293312616,
      "learning_rate": 2.4314317827503375e-07,
      "loss": 0.1085,
      "step": 341
    },
    {
      "epoch": 6.535031847133758,
      "grad_norm": 0.2982157435124411,
      "learning_rate": 2.2253699394493066e-07,
      "loss": 0.0576,
      "step": 342
    },
    {
      "epoch": 6.554140127388535,
      "grad_norm": 0.3315729789977526,
      "learning_rate": 2.028332677849254e-07,
      "loss": 0.0573,
      "step": 343
    },
    {
      "epoch": 6.573248407643312,
      "grad_norm": 0.25427048529330454,
      "learning_rate": 1.840338184455881e-07,
      "loss": 0.059,
      "step": 344
    },
    {
      "epoch": 6.592356687898089,
      "grad_norm": 0.2785181993003191,
      "learning_rate": 1.6614038111289034e-07,
      "loss": 0.0611,
      "step": 345
    },
    {
      "epoch": 6.611464968152866,
      "grad_norm": 0.29248938451924733,
      "learning_rate": 1.49154607348051e-07,
      "loss": 0.0793,
      "step": 346
    },
    {
      "epoch": 6.630573248407643,
      "grad_norm": 0.24475616642622175,
      "learning_rate": 1.330780649350938e-07,
      "loss": 0.0613,
      "step": 347
    },
    {
      "epoch": 6.649681528662421,
      "grad_norm": 0.28091544427398124,
      "learning_rate": 1.1791223773614635e-07,
      "loss": 0.0487,
      "step": 348
    },
    {
      "epoch": 6.6687898089171975,
      "grad_norm": 0.30442981684389875,
      "learning_rate": 1.0365852555447642e-07,
      "loss": 0.0642,
      "step": 349
    },
    {
      "epoch": 6.687898089171974,
      "grad_norm": 0.2746202449333878,
      "learning_rate": 9.031824400528854e-08,
      "loss": 0.0756,
      "step": 350
    },
    {
      "epoch": 6.707006369426751,
      "grad_norm": 0.3087108598983567,
      "learning_rate": 7.789262439430012e-08,
      "loss": 0.0848,
      "step": 351
    },
    {
      "epoch": 6.726114649681529,
      "grad_norm": 0.24776098383715156,
      "learning_rate": 6.638281360408339e-08,
      "loss": 0.0532,
      "step": 352
    },
    {
      "epoch": 6.745222929936306,
      "grad_norm": 0.26456060953324073,
      "learning_rate": 5.578987398821345e-08,
      "loss": 0.0662,
      "step": 353
    },
    {
      "epoch": 6.764331210191083,
      "grad_norm": 0.2955614791032562,
      "learning_rate": 4.6114783273213395e-08,
      "loss": 0.0583,
      "step": 354
    },
    {
      "epoch": 6.7834394904458595,
      "grad_norm": 0.30673675317506477,
      "learning_rate": 3.735843446830867e-08,
      "loss": 0.0782,
      "step": 355
    },
    {
      "epoch": 6.802547770700637,
      "grad_norm": 0.28925621808565305,
      "learning_rate": 2.9521635783001932e-08,
      "loss": 0.0671,
      "step": 356
    },
    {
      "epoch": 6.821656050955414,
      "grad_norm": 0.30205551410687065,
      "learning_rate": 2.2605110552477162e-08,
      "loss": 0.0832,
      "step": 357
    },
    {
      "epoch": 6.840764331210191,
      "grad_norm": 0.2919584904581562,
      "learning_rate": 1.6609497170834154e-08,
      "loss": 0.0729,
      "step": 358
    },
    {
      "epoch": 6.859872611464969,
      "grad_norm": 0.24648188888791672,
      "learning_rate": 1.1535349032167908e-08,
      "loss": 0.0611,
      "step": 359
    },
    {
      "epoch": 6.8789808917197455,
      "grad_norm": 0.3133304348225492,
      "learning_rate": 7.3831344794872415e-09,
      "loss": 0.0791,
      "step": 360
    },
    {
      "epoch": 6.898089171974522,
      "grad_norm": 0.3021167233528452,
      "learning_rate": 4.153236761488266e-09,
      "loss": 0.0523,
      "step": 361
    },
    {
      "epoch": 6.917197452229299,
      "grad_norm": 0.2634021956191592,
      "learning_rate": 1.8459539971804608e-09,
      "loss": 0.0352,
      "step": 362
    },
    {
      "epoch": 6.936305732484076,
      "grad_norm": 0.30858275497051213,
      "learning_rate": 4.614991483686826e-10,
      "loss": 0.0874,
      "step": 363
    },
    {
      "epoch": 6.955414012738854,
      "grad_norm": 0.2820445593615768,
      "learning_rate": 0.0,
      "loss": 0.0458,
      "step": 364
    },
    {
      "epoch": 6.955414012738854,
      "step": 364,
      "total_flos": 8.44150893179945e+16,
      "train_loss": 0.3398333612092576,
      "train_runtime": 2635.8839,
      "train_samples_per_second": 13.278,
      "train_steps_per_second": 0.138
    }
  ],
  "logging_steps": 1,
  "max_steps": 364,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.44150893179945e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}