{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.07545835052761894,
  "eval_steps": 160,
  "global_step": 480,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00015720489693253945,
      "grad_norm": 1.3751904964447021,
      "learning_rate": 0.0,
      "loss": 3.5741,
      "step": 1
    },
    {
      "epoch": 0.00015720489693253945,
      "eval_loss": 3.4173049926757812,
      "eval_runtime": 2315.7248,
      "eval_samples_per_second": 3.998,
      "eval_steps_per_second": 1.999,
      "step": 1
    },
    {
      "epoch": 0.0003144097938650789,
      "grad_norm": 1.231239676475525,
      "learning_rate": 5e-06,
      "loss": 3.3021,
      "step": 2
    },
    {
      "epoch": 0.00047161469079761836,
      "grad_norm": 1.3657807111740112,
      "learning_rate": 1e-05,
      "loss": 3.6333,
      "step": 3
    },
    {
      "epoch": 0.0006288195877301578,
      "grad_norm": 1.3117496967315674,
      "learning_rate": 1.5e-05,
      "loss": 3.3731,
      "step": 4
    },
    {
      "epoch": 0.0007860244846626972,
      "grad_norm": 1.4118576049804688,
      "learning_rate": 2e-05,
      "loss": 3.612,
      "step": 5
    },
    {
      "epoch": 0.0009432293815952367,
      "grad_norm": 1.3155895471572876,
      "learning_rate": 2.5e-05,
      "loss": 3.3296,
      "step": 6
    },
    {
      "epoch": 0.001100434278527776,
      "grad_norm": 1.2847192287445068,
      "learning_rate": 3e-05,
      "loss": 3.2168,
      "step": 7
    },
    {
      "epoch": 0.0012576391754603156,
      "grad_norm": 1.1421078443527222,
      "learning_rate": 3.5e-05,
      "loss": 3.085,
      "step": 8
    },
    {
      "epoch": 0.0014148440723928551,
      "grad_norm": 0.9923035502433777,
      "learning_rate": 4e-05,
      "loss": 3.0472,
      "step": 9
    },
    {
      "epoch": 0.0015720489693253944,
      "grad_norm": 0.795043408870697,
      "learning_rate": 4.5e-05,
      "loss": 2.6666,
      "step": 10
    },
    {
      "epoch": 0.001729253866257934,
      "grad_norm": 0.5987974405288696,
      "learning_rate": 5e-05,
      "loss": 2.473,
      "step": 11
    },
    {
      "epoch": 0.0018864587631904734,
      "grad_norm": 0.4488905668258667,
      "learning_rate": 4.9999999236547564e-05,
      "loss": 2.3731,
      "step": 12
    },
    {
      "epoch": 0.002043663660123013,
      "grad_norm": 0.3517301380634308,
      "learning_rate": 4.999999694619029e-05,
      "loss": 2.2158,
      "step": 13
    },
    {
      "epoch": 0.002200868557055552,
      "grad_norm": 0.3045121431350708,
      "learning_rate": 4.999999312892831e-05,
      "loss": 2.3351,
      "step": 14
    },
    {
      "epoch": 0.002358073453988092,
      "grad_norm": 0.24488244950771332,
      "learning_rate": 4.9999987784761884e-05,
      "loss": 2.2693,
      "step": 15
    },
    {
      "epoch": 0.0025152783509206312,
      "grad_norm": 0.22892728447914124,
      "learning_rate": 4.999998091369132e-05,
      "loss": 2.1006,
      "step": 16
    },
    {
      "epoch": 0.0026724832478531705,
      "grad_norm": 0.23219206929206848,
      "learning_rate": 4.999997251571704e-05,
      "loss": 2.215,
      "step": 17
    },
    {
      "epoch": 0.0028296881447857102,
      "grad_norm": 0.24427154660224915,
      "learning_rate": 4.999996259083956e-05,
      "loss": 2.1708,
      "step": 18
    },
    {
      "epoch": 0.0029868930417182495,
      "grad_norm": 0.2640205919742584,
      "learning_rate": 4.999995113905947e-05,
      "loss": 2.1709,
      "step": 19
    },
    {
      "epoch": 0.003144097938650789,
      "grad_norm": 0.26644033193588257,
      "learning_rate": 4.999993816037749e-05,
      "loss": 2.1733,
      "step": 20
    },
    {
      "epoch": 0.0033013028355833285,
      "grad_norm": 0.2621535062789917,
      "learning_rate": 4.9999923654794414e-05,
      "loss": 2.0059,
      "step": 21
    },
    {
      "epoch": 0.003458507732515868,
      "grad_norm": 0.2586187422275543,
      "learning_rate": 4.999990762231111e-05,
      "loss": 2.0336,
      "step": 22
    },
    {
      "epoch": 0.003615712629448407,
      "grad_norm": 0.26732271909713745,
      "learning_rate": 4.9999890062928566e-05,
      "loss": 2.0566,
      "step": 23
    },
    {
      "epoch": 0.003772917526380947,
      "grad_norm": 0.2357867807149887,
      "learning_rate": 4.999987097664787e-05,
      "loss": 1.9529,
      "step": 24
    },
    {
      "epoch": 0.003930122423313486,
      "grad_norm": 0.2297009825706482,
      "learning_rate": 4.999985036347016e-05,
      "loss": 2.0369,
      "step": 25
    },
    {
      "epoch": 0.004087327320246026,
      "grad_norm": 0.20529747009277344,
      "learning_rate": 4.9999828223396705e-05,
      "loss": 1.9781,
      "step": 26
    },
    {
      "epoch": 0.004244532217178565,
      "grad_norm": 0.18342873454093933,
      "learning_rate": 4.999980455642887e-05,
      "loss": 1.9986,
      "step": 27
    },
    {
      "epoch": 0.004401737114111104,
      "grad_norm": 0.16487397253513336,
      "learning_rate": 4.999977936256809e-05,
      "loss": 1.9063,
      "step": 28
    },
    {
      "epoch": 0.004558942011043644,
      "grad_norm": 0.1762266606092453,
      "learning_rate": 4.99997526418159e-05,
      "loss": 1.9517,
      "step": 29
    },
    {
      "epoch": 0.004716146907976184,
      "grad_norm": 0.16371938586235046,
      "learning_rate": 4.999972439417394e-05,
      "loss": 1.7734,
      "step": 30
    },
    {
      "epoch": 0.004873351804908723,
      "grad_norm": 0.17309769988059998,
      "learning_rate": 4.999969461964392e-05,
      "loss": 1.8732,
      "step": 31
    },
    {
      "epoch": 0.0050305567018412625,
      "grad_norm": 0.15772338211536407,
      "learning_rate": 4.9999663318227683e-05,
      "loss": 1.7537,
      "step": 32
    },
    {
      "epoch": 0.005187761598773802,
      "grad_norm": 0.17521986365318298,
      "learning_rate": 4.9999630489927126e-05,
      "loss": 2.0077,
      "step": 33
    },
    {
      "epoch": 0.005344966495706341,
      "grad_norm": 0.15462292730808258,
      "learning_rate": 4.999959613474425e-05,
      "loss": 1.8576,
      "step": 34
    },
    {
      "epoch": 0.005502171392638881,
      "grad_norm": 0.15280336141586304,
      "learning_rate": 4.999956025268117e-05,
      "loss": 1.862,
      "step": 35
    },
    {
      "epoch": 0.0056593762895714205,
      "grad_norm": 0.14518432319164276,
      "learning_rate": 4.999952284374006e-05,
      "loss": 1.8893,
      "step": 36
    },
    {
      "epoch": 0.005816581186503959,
      "grad_norm": 0.16087624430656433,
      "learning_rate": 4.999948390792321e-05,
      "loss": 1.8658,
      "step": 37
    },
    {
      "epoch": 0.005973786083436499,
      "grad_norm": 0.17504698038101196,
      "learning_rate": 4.999944344523301e-05,
      "loss": 1.7647,
      "step": 38
    },
    {
      "epoch": 0.006130990980369039,
      "grad_norm": 0.17786233127117157,
      "learning_rate": 4.999940145567191e-05,
      "loss": 1.8133,
      "step": 39
    },
    {
      "epoch": 0.006288195877301578,
      "grad_norm": 0.1628972887992859,
      "learning_rate": 4.999935793924249e-05,
      "loss": 1.7731,
      "step": 40
    },
    {
      "epoch": 0.006445400774234117,
      "grad_norm": 0.13461466133594513,
      "learning_rate": 4.9999312895947406e-05,
      "loss": 1.7558,
      "step": 41
    },
    {
      "epoch": 0.006602605671166657,
      "grad_norm": 0.12960125505924225,
      "learning_rate": 4.99992663257894e-05,
      "loss": 1.7639,
      "step": 42
    },
    {
      "epoch": 0.006759810568099196,
      "grad_norm": 0.10991287231445312,
      "learning_rate": 4.9999218228771324e-05,
      "loss": 1.7538,
      "step": 43
    },
    {
      "epoch": 0.006917015465031736,
      "grad_norm": 0.11583230644464493,
      "learning_rate": 4.999916860489612e-05,
      "loss": 1.715,
      "step": 44
    },
    {
      "epoch": 0.007074220361964275,
      "grad_norm": 0.10344280302524567,
      "learning_rate": 4.999911745416681e-05,
      "loss": 1.6907,
      "step": 45
    },
    {
      "epoch": 0.007231425258896814,
      "grad_norm": 0.10546118766069412,
      "learning_rate": 4.999906477658651e-05,
      "loss": 1.7294,
      "step": 46
    },
    {
      "epoch": 0.007388630155829354,
      "grad_norm": 0.11775675415992737,
      "learning_rate": 4.9999010572158465e-05,
      "loss": 1.7146,
      "step": 47
    },
    {
      "epoch": 0.007545835052761894,
      "grad_norm": 0.11109112203121185,
      "learning_rate": 4.999895484088596e-05,
      "loss": 1.6939,
      "step": 48
    },
    {
      "epoch": 0.007703039949694433,
      "grad_norm": 0.1116517186164856,
      "learning_rate": 4.999889758277242e-05,
      "loss": 1.7271,
      "step": 49
    },
    {
      "epoch": 0.007860244846626972,
      "grad_norm": 0.11245547980070114,
      "learning_rate": 4.999883879782132e-05,
      "loss": 1.7333,
      "step": 50
    },
    {
      "epoch": 0.008017449743559512,
      "grad_norm": 0.1150551363825798,
      "learning_rate": 4.999877848603626e-05,
      "loss": 1.7036,
      "step": 51
    },
    {
      "epoch": 0.008174654640492052,
      "grad_norm": 0.10856381803750992,
      "learning_rate": 4.999871664742093e-05,
      "loss": 1.7493,
      "step": 52
    },
    {
      "epoch": 0.008331859537424591,
      "grad_norm": 0.10760089010000229,
      "learning_rate": 4.9998653281979095e-05,
      "loss": 1.6292,
      "step": 53
    },
    {
      "epoch": 0.00848906443435713,
      "grad_norm": 0.0932115837931633,
      "learning_rate": 4.9998588389714634e-05,
      "loss": 1.6608,
      "step": 54
    },
    {
      "epoch": 0.00864626933128967,
      "grad_norm": 0.09837482124567032,
      "learning_rate": 4.9998521970631504e-05,
      "loss": 1.7834,
      "step": 55
    },
    {
      "epoch": 0.008803474228222209,
      "grad_norm": 0.08872833847999573,
      "learning_rate": 4.9998454024733775e-05,
      "loss": 1.6484,
      "step": 56
    },
    {
      "epoch": 0.008960679125154749,
      "grad_norm": 0.08829163759946823,
      "learning_rate": 4.9998384552025577e-05,
      "loss": 1.5913,
      "step": 57
    },
    {
      "epoch": 0.009117884022087288,
      "grad_norm": 0.09087682515382767,
      "learning_rate": 4.999831355251117e-05,
      "loss": 1.6809,
      "step": 58
    },
    {
      "epoch": 0.009275088919019828,
      "grad_norm": 0.08675853163003922,
      "learning_rate": 4.9998241026194884e-05,
      "loss": 1.6519,
      "step": 59
    },
    {
      "epoch": 0.009432293815952368,
      "grad_norm": 0.08463481813669205,
      "learning_rate": 4.999816697308114e-05,
      "loss": 1.6234,
      "step": 60
    },
    {
      "epoch": 0.009589498712884906,
      "grad_norm": 0.08403950184583664,
      "learning_rate": 4.999809139317448e-05,
      "loss": 1.6533,
      "step": 61
    },
    {
      "epoch": 0.009746703609817445,
      "grad_norm": 0.08155622333288193,
      "learning_rate": 4.99980142864795e-05,
      "loss": 1.6726,
      "step": 62
    },
    {
      "epoch": 0.009903908506749985,
      "grad_norm": 0.08056480437517166,
      "learning_rate": 4.999793565300093e-05,
      "loss": 1.5881,
      "step": 63
    },
    {
      "epoch": 0.010061113403682525,
      "grad_norm": 0.07879023998975754,
      "learning_rate": 4.999785549274355e-05,
      "loss": 1.5568,
      "step": 64
    },
    {
      "epoch": 0.010218318300615065,
      "grad_norm": 0.07828455418348312,
      "learning_rate": 4.9997773805712265e-05,
      "loss": 1.6464,
      "step": 65
    },
    {
      "epoch": 0.010375523197547604,
      "grad_norm": 0.08054805546998978,
      "learning_rate": 4.9997690591912075e-05,
      "loss": 1.6213,
      "step": 66
    },
    {
      "epoch": 0.010532728094480142,
      "grad_norm": 0.07610727101564407,
      "learning_rate": 4.999760585134805e-05,
      "loss": 1.5729,
      "step": 67
    },
    {
      "epoch": 0.010689932991412682,
      "grad_norm": 0.07693428546190262,
      "learning_rate": 4.999751958402537e-05,
      "loss": 1.5444,
      "step": 68
    },
    {
      "epoch": 0.010847137888345222,
      "grad_norm": 0.0810319185256958,
      "learning_rate": 4.99974317899493e-05,
      "loss": 1.7045,
      "step": 69
    },
    {
      "epoch": 0.011004342785277762,
      "grad_norm": 0.07729896157979965,
      "learning_rate": 4.9997342469125205e-05,
      "loss": 1.6268,
      "step": 70
    },
    {
      "epoch": 0.011161547682210301,
      "grad_norm": 0.07730107754468918,
      "learning_rate": 4.999725162155855e-05,
      "loss": 1.658,
      "step": 71
    },
    {
      "epoch": 0.011318752579142841,
      "grad_norm": 0.08072328567504883,
      "learning_rate": 4.9997159247254864e-05,
      "loss": 1.5045,
      "step": 72
    },
    {
      "epoch": 0.011475957476075379,
      "grad_norm": 0.08120577782392502,
      "learning_rate": 4.9997065346219805e-05,
      "loss": 1.568,
      "step": 73
    },
    {
      "epoch": 0.011633162373007919,
      "grad_norm": 0.08131498098373413,
      "learning_rate": 4.99969699184591e-05,
      "loss": 1.6035,
      "step": 74
    },
    {
      "epoch": 0.011790367269940458,
      "grad_norm": 0.08395873010158539,
      "learning_rate": 4.9996872963978584e-05,
      "loss": 1.5844,
      "step": 75
    },
    {
      "epoch": 0.011947572166872998,
      "grad_norm": 0.08502068370580673,
      "learning_rate": 4.999677448278417e-05,
      "loss": 1.6661,
      "step": 76
    },
    {
      "epoch": 0.012104777063805538,
      "grad_norm": 0.08467952907085419,
      "learning_rate": 4.999667447488188e-05,
      "loss": 1.5537,
      "step": 77
    },
    {
      "epoch": 0.012261981960738078,
      "grad_norm": 0.19682182371616364,
      "learning_rate": 4.999657294027782e-05,
      "loss": 1.5051,
      "step": 78
    },
    {
      "epoch": 0.012419186857670617,
      "grad_norm": 0.08586428314447403,
      "learning_rate": 4.999646987897818e-05,
      "loss": 1.565,
      "step": 79
    },
    {
      "epoch": 0.012576391754603155,
      "grad_norm": 0.08156823366880417,
      "learning_rate": 4.999636529098928e-05,
      "loss": 1.6627,
      "step": 80
    },
    {
      "epoch": 0.012733596651535695,
      "grad_norm": 0.08715341240167618,
      "learning_rate": 4.9996259176317486e-05,
      "loss": 1.5862,
      "step": 81
    },
    {
      "epoch": 0.012890801548468235,
      "grad_norm": 0.09664586186408997,
      "learning_rate": 4.999615153496928e-05,
      "loss": 1.5741,
      "step": 82
    },
    {
      "epoch": 0.013048006445400774,
      "grad_norm": 0.08438891172409058,
      "learning_rate": 4.999604236695125e-05,
      "loss": 1.5933,
      "step": 83
    },
    {
      "epoch": 0.013205211342333314,
      "grad_norm": 0.08333732932806015,
      "learning_rate": 4.999593167227006e-05,
      "loss": 1.5904,
      "step": 84
    },
    {
      "epoch": 0.013362416239265854,
      "grad_norm": 0.07945791631937027,
      "learning_rate": 4.9995819450932455e-05,
      "loss": 1.5763,
      "step": 85
    },
    {
      "epoch": 0.013519621136198392,
      "grad_norm": 0.07682961225509644,
      "learning_rate": 4.9995705702945304e-05,
      "loss": 1.5197,
      "step": 86
    },
    {
      "epoch": 0.013676826033130932,
      "grad_norm": 0.07547677308320999,
      "learning_rate": 4.999559042831555e-05,
      "loss": 1.6825,
      "step": 87
    },
    {
      "epoch": 0.013834030930063471,
      "grad_norm": 0.07293456047773361,
      "learning_rate": 4.999547362705025e-05,
      "loss": 1.5466,
      "step": 88
    },
    {
      "epoch": 0.013991235826996011,
      "grad_norm": 0.07730914652347565,
      "learning_rate": 4.999535529915651e-05,
      "loss": 1.5775,
      "step": 89
    },
    {
      "epoch": 0.01414844072392855,
      "grad_norm": 0.07689664512872696,
      "learning_rate": 4.9995235444641565e-05,
      "loss": 1.5881,
      "step": 90
    },
    {
      "epoch": 0.01430564562086109,
      "grad_norm": 0.07754997909069061,
      "learning_rate": 4.999511406351275e-05,
      "loss": 1.5037,
      "step": 91
    },
    {
      "epoch": 0.014462850517793628,
      "grad_norm": 0.07229866087436676,
      "learning_rate": 4.999499115577746e-05,
      "loss": 1.5077,
      "step": 92
    },
    {
      "epoch": 0.014620055414726168,
      "grad_norm": 0.07491567730903625,
      "learning_rate": 4.9994866721443215e-05,
      "loss": 1.5461,
      "step": 93
    },
    {
      "epoch": 0.014777260311658708,
      "grad_norm": 0.07258685678243637,
      "learning_rate": 4.9994740760517605e-05,
      "loss": 1.5516,
      "step": 94
    },
    {
      "epoch": 0.014934465208591248,
      "grad_norm": 0.07643327116966248,
      "learning_rate": 4.9994613273008334e-05,
      "loss": 1.6223,
      "step": 95
    },
    {
      "epoch": 0.015091670105523787,
      "grad_norm": 0.0740588903427124,
      "learning_rate": 4.999448425892318e-05,
      "loss": 1.5322,
      "step": 96
    },
    {
      "epoch": 0.015248875002456327,
      "grad_norm": 0.44172239303588867,
      "learning_rate": 4.999435371827003e-05,
      "loss": 1.5498,
      "step": 97
    },
    {
      "epoch": 0.015406079899388867,
      "grad_norm": 0.0756363570690155,
      "learning_rate": 4.999422165105684e-05,
      "loss": 1.559,
      "step": 98
    },
    {
      "epoch": 0.015563284796321405,
      "grad_norm": 0.07251248508691788,
      "learning_rate": 4.99940880572917e-05,
      "loss": 1.5903,
      "step": 99
    },
    {
      "epoch": 0.015720489693253945,
      "grad_norm": 0.06931837648153305,
      "learning_rate": 4.999395293698275e-05,
      "loss": 1.4849,
      "step": 100
    },
    {
      "epoch": 0.015877694590186484,
      "grad_norm": 0.07403590530157089,
      "learning_rate": 4.9993816290138254e-05,
      "loss": 1.5191,
      "step": 101
    },
    {
      "epoch": 0.016034899487119024,
      "grad_norm": 0.07027724385261536,
      "learning_rate": 4.999367811676655e-05,
      "loss": 1.5655,
      "step": 102
    },
    {
      "epoch": 0.016192104384051564,
      "grad_norm": 0.07320379465818405,
      "learning_rate": 4.9993538416876093e-05,
      "loss": 1.4869,
      "step": 103
    },
    {
      "epoch": 0.016349309280984103,
      "grad_norm": 0.0726180374622345,
      "learning_rate": 4.9993397190475396e-05,
      "loss": 1.4629,
      "step": 104
    },
    {
      "epoch": 0.016506514177916643,
      "grad_norm": 0.07542011886835098,
      "learning_rate": 4.999325443757309e-05,
      "loss": 1.5976,
      "step": 105
    },
    {
      "epoch": 0.016663719074849183,
      "grad_norm": 0.07440067082643509,
      "learning_rate": 4.9993110158177895e-05,
      "loss": 1.5469,
      "step": 106
    },
    {
      "epoch": 0.016820923971781723,
      "grad_norm": 0.07547372579574585,
      "learning_rate": 4.999296435229863e-05,
      "loss": 1.5328,
      "step": 107
    },
    {
      "epoch": 0.01697812886871426,
      "grad_norm": 0.07532137632369995,
      "learning_rate": 4.999281701994419e-05,
      "loss": 1.6742,
      "step": 108
    },
    {
      "epoch": 0.0171353337656468,
      "grad_norm": 0.07249438762664795,
      "learning_rate": 4.999266816112358e-05,
      "loss": 1.4799,
      "step": 109
    },
    {
      "epoch": 0.01729253866257934,
      "grad_norm": 0.07399806380271912,
      "learning_rate": 4.999251777584589e-05,
      "loss": 1.5438,
      "step": 110
    },
    {
      "epoch": 0.017449743559511878,
      "grad_norm": 0.08135057240724564,
      "learning_rate": 4.99923658641203e-05,
      "loss": 1.5608,
      "step": 111
    },
    {
      "epoch": 0.017606948456444418,
      "grad_norm": 0.07508935779333115,
      "learning_rate": 4.99922124259561e-05,
      "loss": 1.5894,
      "step": 112
    },
    {
      "epoch": 0.017764153353376957,
      "grad_norm": 0.07432372123003006,
      "learning_rate": 4.999205746136265e-05,
      "loss": 1.4818,
      "step": 113
    },
    {
      "epoch": 0.017921358250309497,
      "grad_norm": 0.07694194465875626,
      "learning_rate": 4.999190097034942e-05,
      "loss": 1.5629,
      "step": 114
    },
    {
      "epoch": 0.018078563147242037,
      "grad_norm": 0.07384433597326279,
      "learning_rate": 4.999174295292597e-05,
      "loss": 1.4829,
      "step": 115
    },
    {
      "epoch": 0.018235768044174577,
      "grad_norm": 0.07152919471263885,
      "learning_rate": 4.999158340910195e-05,
      "loss": 1.4748,
      "step": 116
    },
    {
      "epoch": 0.018392972941107116,
      "grad_norm": 0.07719701528549194,
      "learning_rate": 4.999142233888709e-05,
      "loss": 1.5524,
      "step": 117
    },
    {
      "epoch": 0.018550177838039656,
      "grad_norm": 0.07540587335824966,
      "learning_rate": 4.999125974229125e-05,
      "loss": 1.4661,
      "step": 118
    },
    {
      "epoch": 0.018707382734972196,
      "grad_norm": 0.0787581130862236,
      "learning_rate": 4.9991095619324344e-05,
      "loss": 1.6455,
      "step": 119
    },
    {
      "epoch": 0.018864587631904736,
      "grad_norm": 0.07454577833414078,
      "learning_rate": 4.999092996999641e-05,
      "loss": 1.5083,
      "step": 120
    },
    {
      "epoch": 0.019021792528837272,
      "grad_norm": 0.0751076266169548,
      "learning_rate": 4.9990762794317545e-05,
      "loss": 1.4874,
      "step": 121
    },
    {
      "epoch": 0.01917899742576981,
      "grad_norm": 0.07733119279146194,
      "learning_rate": 4.999059409229798e-05,
      "loss": 1.6308,
      "step": 122
    },
    {
      "epoch": 0.01933620232270235,
      "grad_norm": 0.07897089421749115,
      "learning_rate": 4.999042386394802e-05,
      "loss": 1.5906,
      "step": 123
    },
    {
      "epoch": 0.01949340721963489,
      "grad_norm": 0.07758141309022903,
      "learning_rate": 4.999025210927804e-05,
      "loss": 1.5604,
      "step": 124
    },
    {
      "epoch": 0.01965061211656743,
      "grad_norm": 0.07845707982778549,
      "learning_rate": 4.9990078828298544e-05,
      "loss": 1.5901,
      "step": 125
    },
    {
      "epoch": 0.01980781701349997,
      "grad_norm": 0.0772818773984909,
      "learning_rate": 4.998990402102012e-05,
      "loss": 1.4516,
      "step": 126
    },
    {
      "epoch": 0.01996502191043251,
      "grad_norm": 0.07795504480600357,
      "learning_rate": 4.998972768745344e-05,
      "loss": 1.4642,
      "step": 127
    },
    {
      "epoch": 0.02012222680736505,
      "grad_norm": 0.0784008800983429,
      "learning_rate": 4.998954982760926e-05,
      "loss": 1.5936,
      "step": 128
    },
    {
      "epoch": 0.02027943170429759,
      "grad_norm": 0.07791212201118469,
      "learning_rate": 4.9989370441498465e-05,
      "loss": 1.4705,
      "step": 129
    },
    {
      "epoch": 0.02043663660123013,
      "grad_norm": 0.07785367220640182,
      "learning_rate": 4.9989189529132004e-05,
      "loss": 1.5085,
      "step": 130
    },
    {
      "epoch": 0.02059384149816267,
      "grad_norm": 0.07916689664125443,
      "learning_rate": 4.9989007090520925e-05,
      "loss": 1.5365,
      "step": 131
    },
    {
      "epoch": 0.02075104639509521,
      "grad_norm": 0.0775083601474762,
      "learning_rate": 4.9988823125676367e-05,
      "loss": 1.5286,
      "step": 132
    },
    {
      "epoch": 0.020908251292027745,
      "grad_norm": 0.08110442757606506,
      "learning_rate": 4.998863763460956e-05,
      "loss": 1.5779,
      "step": 133
    },
    {
      "epoch": 0.021065456188960285,
      "grad_norm": 0.0814640000462532,
      "learning_rate": 4.998845061733185e-05,
      "loss": 1.4778,
      "step": 134
    },
    {
      "epoch": 0.021222661085892824,
      "grad_norm": 0.08069492131471634,
      "learning_rate": 4.998826207385465e-05,
      "loss": 1.5317,
      "step": 135
    },
    {
      "epoch": 0.021379865982825364,
      "grad_norm": 0.07377774268388748,
      "learning_rate": 4.998807200418948e-05,
      "loss": 1.5258,
      "step": 136
    },
    {
      "epoch": 0.021537070879757904,
      "grad_norm": 0.0787922590970993,
      "learning_rate": 4.9987880408347945e-05,
      "loss": 1.5185,
      "step": 137
    },
    {
      "epoch": 0.021694275776690444,
      "grad_norm": 0.07662995159626007,
      "learning_rate": 4.9987687286341745e-05,
      "loss": 1.4637,
      "step": 138
    },
    {
      "epoch": 0.021851480673622983,
      "grad_norm": 0.08528955280780792,
      "learning_rate": 4.9987492638182676e-05,
      "loss": 1.4776,
      "step": 139
    },
    {
      "epoch": 0.022008685570555523,
      "grad_norm": 0.08089053630828857,
      "learning_rate": 4.9987296463882626e-05,
      "loss": 1.5885,
      "step": 140
    },
    {
      "epoch": 0.022165890467488063,
      "grad_norm": 0.08029694855213165,
      "learning_rate": 4.998709876345358e-05,
      "loss": 1.4557,
      "step": 141
    },
    {
      "epoch": 0.022323095364420602,
      "grad_norm": 0.07918502390384674,
      "learning_rate": 4.9986899536907614e-05,
      "loss": 1.4285,
      "step": 142
    },
    {
      "epoch": 0.022480300261353142,
      "grad_norm": 0.0813126415014267,
      "learning_rate": 4.998669878425689e-05,
      "loss": 1.5958,
      "step": 143
    },
    {
      "epoch": 0.022637505158285682,
      "grad_norm": 0.07935188710689545,
      "learning_rate": 4.998649650551368e-05,
      "loss": 1.5249,
      "step": 144
    },
    {
      "epoch": 0.02279471005521822,
      "grad_norm": 0.08163304626941681,
      "learning_rate": 4.9986292700690324e-05,
      "loss": 1.483,
      "step": 145
    },
    {
      "epoch": 0.022951914952150758,
      "grad_norm": 0.08277447521686554,
      "learning_rate": 4.998608736979928e-05,
      "loss": 1.6212,
      "step": 146
    },
    {
      "epoch": 0.023109119849083298,
      "grad_norm": 0.08285827934741974,
      "learning_rate": 4.9985880512853076e-05,
      "loss": 1.4495,
      "step": 147
    },
    {
      "epoch": 0.023266324746015837,
      "grad_norm": 0.082750603556633,
      "learning_rate": 4.998567212986437e-05,
      "loss": 1.4335,
      "step": 148
    },
    {
      "epoch": 0.023423529642948377,
      "grad_norm": 0.07986058294773102,
      "learning_rate": 4.998546222084587e-05,
      "loss": 1.4704,
      "step": 149
    },
    {
      "epoch": 0.023580734539880917,
      "grad_norm": 0.08105576783418655,
      "learning_rate": 4.9985250785810396e-05,
      "loss": 1.5183,
      "step": 150
    },
    {
      "epoch": 0.023737939436813457,
      "grad_norm": 0.08202917128801346,
      "learning_rate": 4.9985037824770866e-05,
      "loss": 1.5423,
      "step": 151
    },
    {
      "epoch": 0.023895144333745996,
      "grad_norm": 0.08937894552946091,
      "learning_rate": 4.998482333774029e-05,
      "loss": 1.5731,
      "step": 152
    },
    {
      "epoch": 0.024052349230678536,
      "grad_norm": 0.08333728462457657,
      "learning_rate": 4.9984607324731766e-05,
      "loss": 1.5133,
      "step": 153
    },
    {
      "epoch": 0.024209554127611076,
      "grad_norm": 0.08529175072908401,
      "learning_rate": 4.998438978575849e-05,
      "loss": 1.516,
      "step": 154
    },
    {
      "epoch": 0.024366759024543615,
      "grad_norm": 0.08508963882923126,
      "learning_rate": 4.998417072083374e-05,
      "loss": 1.5646,
      "step": 155
    },
    {
      "epoch": 0.024523963921476155,
      "grad_norm": 0.08971578627824783,
      "learning_rate": 4.99839501299709e-05,
      "loss": 1.4714,
      "step": 156
    },
    {
      "epoch": 0.024681168818408695,
      "grad_norm": 0.08380109816789627,
      "learning_rate": 4.998372801318345e-05,
      "loss": 1.4476,
      "step": 157
    },
    {
      "epoch": 0.024838373715341235,
      "grad_norm": 0.08533143252134323,
      "learning_rate": 4.9983504370484945e-05,
      "loss": 1.4866,
      "step": 158
    },
    {
      "epoch": 0.02499557861227377,
      "grad_norm": 0.08318709582090378,
      "learning_rate": 4.998327920188905e-05,
      "loss": 1.5274,
      "step": 159
    },
    {
      "epoch": 0.02515278350920631,
      "grad_norm": 0.08486370742321014,
      "learning_rate": 4.9983052507409525e-05,
      "loss": 1.4713,
      "step": 160
    },
    {
      "epoch": 0.02515278350920631,
      "eval_loss": 1.5136528015136719,
      "eval_runtime": 2318.8971,
      "eval_samples_per_second": 3.992,
      "eval_steps_per_second": 1.996,
      "step": 160
    },
    {
      "epoch": 0.02530998840613885,
      "grad_norm": 0.08242359757423401,
      "learning_rate": 4.9982824287060195e-05,
      "loss": 1.5069,
      "step": 161
    },
    {
      "epoch": 0.02546719330307139,
      "grad_norm": 0.08547423779964447,
      "learning_rate": 4.9982594540855014e-05,
      "loss": 1.4973,
      "step": 162
    },
    {
      "epoch": 0.02562439820000393,
      "grad_norm": 0.08345580101013184,
      "learning_rate": 4.9982363268808016e-05,
      "loss": 1.5078,
      "step": 163
    },
    {
      "epoch": 0.02578160309693647,
      "grad_norm": 0.0830339640378952,
      "learning_rate": 4.9982130470933316e-05,
      "loss": 1.4098,
      "step": 164
    },
    {
      "epoch": 0.02593880799386901,
      "grad_norm": 0.08568515628576279,
      "learning_rate": 4.998189614724514e-05,
      "loss": 1.4628,
      "step": 165
    },
    {
      "epoch": 0.02609601289080155,
      "grad_norm": 0.08261829614639282,
      "learning_rate": 4.998166029775779e-05,
      "loss": 1.4492,
      "step": 166
    },
    {
      "epoch": 0.02625321778773409,
      "grad_norm": 0.08944887667894363,
      "learning_rate": 4.998142292248569e-05,
      "loss": 1.5633,
      "step": 167
    },
    {
      "epoch": 0.02641042268466663,
      "grad_norm": 0.08632911741733551,
      "learning_rate": 4.998118402144332e-05,
      "loss": 1.5106,
      "step": 168
    },
    {
      "epoch": 0.026567627581599168,
      "grad_norm": 0.08733859658241272,
      "learning_rate": 4.998094359464528e-05,
      "loss": 1.5607,
      "step": 169
    },
    {
      "epoch": 0.026724832478531708,
      "grad_norm": 0.08667927235364914,
      "learning_rate": 4.9980701642106245e-05,
      "loss": 1.4544,
      "step": 170
    },
    {
      "epoch": 0.026882037375464244,
      "grad_norm": 0.08655022084712982,
      "learning_rate": 4.9980458163841006e-05,
      "loss": 1.5264,
      "step": 171
    },
    {
      "epoch": 0.027039242272396784,
      "grad_norm": 0.08899988234043121,
      "learning_rate": 4.9980213159864426e-05,
      "loss": 1.4778,
      "step": 172
    },
    {
      "epoch": 0.027196447169329323,
      "grad_norm": 0.09411856532096863,
      "learning_rate": 4.997996663019147e-05,
      "loss": 1.5269,
      "step": 173
    },
    {
      "epoch": 0.027353652066261863,
      "grad_norm": 0.087191641330719,
      "learning_rate": 4.997971857483719e-05,
      "loss": 1.5166,
      "step": 174
    },
    {
      "epoch": 0.027510856963194403,
      "grad_norm": 0.08959636092185974,
      "learning_rate": 4.997946899381675e-05,
      "loss": 1.5503,
      "step": 175
    },
    {
      "epoch": 0.027668061860126943,
      "grad_norm": 0.0951187014579773,
      "learning_rate": 4.997921788714537e-05,
      "loss": 1.4879,
      "step": 176
    },
    {
      "epoch": 0.027825266757059482,
      "grad_norm": 0.09324768930673599,
      "learning_rate": 4.997896525483841e-05,
      "loss": 1.5714,
      "step": 177
    },
    {
      "epoch": 0.027982471653992022,
      "grad_norm": 0.08633986115455627,
      "learning_rate": 4.997871109691129e-05,
      "loss": 1.4198,
      "step": 178
    },
    {
      "epoch": 0.028139676550924562,
      "grad_norm": 0.08947525173425674,
      "learning_rate": 4.9978455413379535e-05,
      "loss": 1.4702,
      "step": 179
    },
    {
      "epoch": 0.0282968814478571,
      "grad_norm": 0.09275490790605545,
      "learning_rate": 4.9978198204258766e-05,
      "loss": 1.5252,
      "step": 180
    },
    {
      "epoch": 0.02845408634478964,
      "grad_norm": 0.08761609345674515,
      "learning_rate": 4.9977939469564676e-05,
      "loss": 1.505,
      "step": 181
    },
    {
      "epoch": 0.02861129124172218,
      "grad_norm": 0.08683087676763535,
      "learning_rate": 4.997767920931308e-05,
      "loss": 1.5059,
      "step": 182
    },
    {
      "epoch": 0.02876849613865472,
      "grad_norm": 0.08931361883878708,
      "learning_rate": 4.997741742351988e-05,
      "loss": 1.5003,
      "step": 183
    },
    {
      "epoch": 0.028925701035587257,
      "grad_norm": 0.08820109069347382,
      "learning_rate": 4.997715411220105e-05,
      "loss": 1.5132,
      "step": 184
    },
    {
      "epoch": 0.029082905932519797,
      "grad_norm": 0.09284964948892593,
      "learning_rate": 4.997688927537268e-05,
      "loss": 1.4561,
      "step": 185
    },
    {
      "epoch": 0.029240110829452336,
      "grad_norm": 0.09472864121198654,
      "learning_rate": 4.997662291305094e-05,
      "loss": 1.4729,
      "step": 186
    },
    {
      "epoch": 0.029397315726384876,
      "grad_norm": 0.08725330233573914,
      "learning_rate": 4.997635502525211e-05,
      "loss": 1.3994,
      "step": 187
    },
    {
      "epoch": 0.029554520623317416,
      "grad_norm": 0.09085626900196075,
      "learning_rate": 4.9976085611992536e-05,
      "loss": 1.4695,
      "step": 188
    },
    {
      "epoch": 0.029711725520249956,
      "grad_norm": 0.09322400391101837,
      "learning_rate": 4.9975814673288684e-05,
      "loss": 1.4753,
      "step": 189
    },
    {
      "epoch": 0.029868930417182495,
      "grad_norm": 0.08927160501480103,
      "learning_rate": 4.99755422091571e-05,
      "loss": 1.4465,
      "step": 190
    },
    {
      "epoch": 0.030026135314115035,
      "grad_norm": 0.09317070990800858,
      "learning_rate": 4.997526821961442e-05,
      "loss": 1.5124,
      "step": 191
    },
    {
      "epoch": 0.030183340211047575,
      "grad_norm": 0.08911167085170746,
      "learning_rate": 4.9974992704677385e-05,
      "loss": 1.4515,
      "step": 192
    },
    {
      "epoch": 0.030340545107980114,
      "grad_norm": 0.09432853013277054,
      "learning_rate": 4.997471566436282e-05,
      "loss": 1.4623,
      "step": 193
    },
    {
      "epoch": 0.030497750004912654,
      "grad_norm": 0.09417332708835602,
      "learning_rate": 4.997443709868764e-05,
      "loss": 1.5103,
      "step": 194
    },
    {
      "epoch": 0.030654954901845194,
      "grad_norm": 0.09564542025327682,
      "learning_rate": 4.997415700766887e-05,
      "loss": 1.4929,
      "step": 195
    },
    {
      "epoch": 0.030812159798777734,
      "grad_norm": 0.09101004898548126,
      "learning_rate": 4.997387539132361e-05,
      "loss": 1.4225,
      "step": 196
    },
    {
      "epoch": 0.03096936469571027,
      "grad_norm": 0.09196274727582932,
      "learning_rate": 4.997359224966906e-05,
      "loss": 1.4701,
      "step": 197
    },
    {
      "epoch": 0.03112656959264281,
      "grad_norm": 0.09573279321193695,
      "learning_rate": 4.997330758272251e-05,
      "loss": 1.4425,
      "step": 198
    },
    {
      "epoch": 0.03128377448957535,
      "grad_norm": 0.09180758893489838,
      "learning_rate": 4.9973021390501354e-05,
      "loss": 1.4426,
      "step": 199
    },
    {
      "epoch": 0.03144097938650789,
      "grad_norm": 0.09583238512277603,
      "learning_rate": 4.997273367302306e-05,
      "loss": 1.5158,
      "step": 200
    },
    {
      "epoch": 0.03159818428344043,
      "grad_norm": 0.09394747018814087,
      "learning_rate": 4.997244443030521e-05,
      "loss": 1.4306,
      "step": 201
    },
    {
      "epoch": 0.03175538918037297,
      "grad_norm": 0.09470199793577194,
      "learning_rate": 4.9972153662365474e-05,
      "loss": 1.5286,
      "step": 202
    },
    {
      "epoch": 0.031912594077305505,
      "grad_norm": 0.09274959564208984,
      "learning_rate": 4.997186136922161e-05,
      "loss": 1.4803,
      "step": 203
    },
    {
      "epoch": 0.03206979897423805,
      "grad_norm": 0.09344369918107986,
      "learning_rate": 4.997156755089145e-05,
      "loss": 1.5449,
      "step": 204
    },
    {
      "epoch": 0.032227003871170584,
      "grad_norm": 0.09794919937849045,
      "learning_rate": 4.997127220739296e-05,
      "loss": 1.4383,
      "step": 205
    },
    {
      "epoch": 0.03238420876810313,
      "grad_norm": 0.09698093682527542,
      "learning_rate": 4.997097533874418e-05,
      "loss": 1.4462,
      "step": 206
    },
    {
      "epoch": 0.032541413665035664,
      "grad_norm": 0.09690559655427933,
      "learning_rate": 4.997067694496323e-05,
      "loss": 1.4735,
      "step": 207
    },
    {
      "epoch": 0.03269861856196821,
      "grad_norm": 0.09657544642686844,
      "learning_rate": 4.9970377026068336e-05,
      "loss": 1.5672,
      "step": 208
    },
    {
      "epoch": 0.03285582345890074,
      "grad_norm": 0.09483659267425537,
      "learning_rate": 4.9970075582077825e-05,
      "loss": 1.4931,
      "step": 209
    },
    {
      "epoch": 0.033013028355833286,
      "grad_norm": 0.09744243323802948,
      "learning_rate": 4.9969772613010104e-05,
      "loss": 1.4638,
      "step": 210
    },
    {
      "epoch": 0.03317023325276582,
      "grad_norm": 0.09521006047725677,
      "learning_rate": 4.9969468118883665e-05,
      "loss": 1.4127,
      "step": 211
    },
    {
      "epoch": 0.033327438149698366,
      "grad_norm": 0.09646004438400269,
      "learning_rate": 4.996916209971713e-05,
      "loss": 1.5139,
      "step": 212
    },
    {
      "epoch": 0.0334846430466309,
      "grad_norm": 0.09292810410261154,
      "learning_rate": 4.996885455552916e-05,
      "loss": 1.4399,
      "step": 213
    },
    {
      "epoch": 0.033641847943563445,
      "grad_norm": 0.09986516088247299,
      "learning_rate": 4.996854548633857e-05,
      "loss": 1.4637,
      "step": 214
    },
    {
      "epoch": 0.03379905284049598,
      "grad_norm": 0.09723702073097229,
      "learning_rate": 4.996823489216421e-05,
      "loss": 1.5673,
      "step": 215
    },
    {
      "epoch": 0.03395625773742852,
      "grad_norm": 0.09608977288007736,
      "learning_rate": 4.996792277302507e-05,
      "loss": 1.4428,
      "step": 216
    },
    {
      "epoch": 0.03411346263436106,
      "grad_norm": 0.09329380095005035,
      "learning_rate": 4.99676091289402e-05,
      "loss": 1.3892,
      "step": 217
    },
    {
      "epoch": 0.0342706675312936,
      "grad_norm": 0.0959913358092308,
      "learning_rate": 4.996729395992875e-05,
      "loss": 1.5219,
      "step": 218
    },
    {
      "epoch": 0.03442787242822614,
      "grad_norm": 0.09832671284675598,
      "learning_rate": 4.996697726600999e-05,
      "loss": 1.5259,
      "step": 219
    },
    {
      "epoch": 0.03458507732515868,
      "grad_norm": 0.10061636567115784,
      "learning_rate": 4.996665904720325e-05,
      "loss": 1.5216,
      "step": 220
    },
    {
      "epoch": 0.03474228222209122,
      "grad_norm": 0.09742400050163269,
      "learning_rate": 4.9966339303527965e-05,
      "loss": 1.3819,
      "step": 221
    },
    {
      "epoch": 0.034899487119023756,
      "grad_norm": 0.09629969298839569,
      "learning_rate": 4.996601803500367e-05,
      "loss": 1.5341,
      "step": 222
    },
    {
      "epoch": 0.0350566920159563,
      "grad_norm": 0.09776200354099274,
      "learning_rate": 4.996569524164998e-05,
      "loss": 1.5054,
      "step": 223
    },
    {
      "epoch": 0.035213896912888835,
      "grad_norm": 0.1008530780673027,
      "learning_rate": 4.996537092348661e-05,
      "loss": 1.5333,
      "step": 224
    },
    {
      "epoch": 0.03537110180982138,
      "grad_norm": 0.09749735891819,
      "learning_rate": 4.996504508053338e-05,
      "loss": 1.3899,
      "step": 225
    },
    {
      "epoch": 0.035528306706753915,
      "grad_norm": 0.10522401332855225,
      "learning_rate": 4.9964717712810175e-05,
      "loss": 1.5413,
      "step": 226
    },
    {
      "epoch": 0.03568551160368646,
      "grad_norm": 0.09566272795200348,
      "learning_rate": 4.9964388820336996e-05,
      "loss": 1.435,
      "step": 227
    },
    {
      "epoch": 0.035842716500618994,
      "grad_norm": 0.10133984684944153,
      "learning_rate": 4.996405840313393e-05,
      "loss": 1.445,
      "step": 228
    },
    {
      "epoch": 0.03599992139755153,
      "grad_norm": 0.09702739119529724,
      "learning_rate": 4.996372646122116e-05,
      "loss": 1.4287,
      "step": 229
    },
    {
      "epoch": 0.036157126294484074,
      "grad_norm": 0.1012992411851883,
      "learning_rate": 4.996339299461896e-05,
      "loss": 1.382,
      "step": 230
    },
    {
      "epoch": 0.03631433119141661,
      "grad_norm": 0.09877166152000427,
      "learning_rate": 4.99630580033477e-05,
      "loss": 1.5729,
      "step": 231
    },
    {
      "epoch": 0.03647153608834915,
      "grad_norm": 0.1033129170536995,
      "learning_rate": 4.996272148742783e-05,
      "loss": 1.4754,
      "step": 232
    },
    {
      "epoch": 0.03662874098528169,
      "grad_norm": 0.09901215881109238,
      "learning_rate": 4.9962383446879914e-05,
      "loss": 1.5153,
      "step": 233
    },
    {
      "epoch": 0.03678594588221423,
      "grad_norm": 0.10241983830928802,
      "learning_rate": 4.996204388172458e-05,
      "loss": 1.5131,
      "step": 234
    },
    {
      "epoch": 0.03694315077914677,
      "grad_norm": 0.09574593603610992,
      "learning_rate": 4.9961702791982594e-05,
      "loss": 1.5285,
      "step": 235
    },
    {
      "epoch": 0.03710035567607931,
      "grad_norm": 0.10309838503599167,
      "learning_rate": 4.996136017767477e-05,
      "loss": 1.5751,
      "step": 236
    },
    {
      "epoch": 0.03725756057301185,
      "grad_norm": 0.09928470849990845,
      "learning_rate": 4.996101603882204e-05,
      "loss": 1.5108,
      "step": 237
    },
    {
      "epoch": 0.03741476546994439,
      "grad_norm": 0.10514767467975616,
      "learning_rate": 4.996067037544542e-05,
      "loss": 1.4206,
      "step": 238
    },
    {
      "epoch": 0.03757197036687693,
      "grad_norm": 0.10411518812179565,
      "learning_rate": 4.996032318756601e-05,
      "loss": 1.5628,
      "step": 239
    },
    {
      "epoch": 0.03772917526380947,
      "grad_norm": 0.0989808738231659,
      "learning_rate": 4.9959974475205045e-05,
      "loss": 1.4444,
      "step": 240
    },
    {
      "epoch": 0.03788638016074201,
      "grad_norm": 0.10069911926984787,
      "learning_rate": 4.9959624238383804e-05,
      "loss": 1.4805,
      "step": 241
    },
    {
      "epoch": 0.038043585057674544,
      "grad_norm": 0.10637518763542175,
      "learning_rate": 4.995927247712367e-05,
      "loss": 1.5289,
      "step": 242
    },
    {
      "epoch": 0.03820078995460709,
      "grad_norm": 0.10085684061050415,
      "learning_rate": 4.995891919144614e-05,
      "loss": 1.5288,
      "step": 243
    },
    {
      "epoch": 0.03835799485153962,
      "grad_norm": 0.09989017248153687,
      "learning_rate": 4.995856438137279e-05,
      "loss": 1.5444,
      "step": 244
    },
    {
      "epoch": 0.038515199748472166,
      "grad_norm": 0.10382463037967682,
      "learning_rate": 4.9958208046925294e-05,
      "loss": 1.4621,
      "step": 245
    },
    {
      "epoch": 0.0386724046454047,
      "grad_norm": 0.10208063572645187,
      "learning_rate": 4.99578501881254e-05,
      "loss": 1.5003,
      "step": 246
    },
    {
      "epoch": 0.038829609542337246,
      "grad_norm": 0.1028011366724968,
      "learning_rate": 4.9957490804994977e-05,
      "loss": 1.516,
      "step": 247
    },
    {
      "epoch": 0.03898681443926978,
      "grad_norm": 0.10475701838731766,
      "learning_rate": 4.995712989755598e-05,
      "loss": 1.5333,
      "step": 248
    },
    {
      "epoch": 0.039144019336202325,
      "grad_norm": 0.1038154736161232,
      "learning_rate": 4.995676746583044e-05,
      "loss": 1.4779,
      "step": 249
    },
    {
      "epoch": 0.03930122423313486,
      "grad_norm": 0.10413440316915512,
      "learning_rate": 4.99564035098405e-05,
      "loss": 1.5241,
      "step": 250
    },
    {
      "epoch": 0.039458429130067404,
      "grad_norm": 0.09869382530450821,
      "learning_rate": 4.995603802960838e-05,
      "loss": 1.442,
      "step": 251
    },
    {
      "epoch": 0.03961563402699994,
      "grad_norm": 0.10138234496116638,
      "learning_rate": 4.995567102515641e-05,
      "loss": 1.5393,
      "step": 252
    },
    {
      "epoch": 0.039772838923932484,
      "grad_norm": 0.10225867480039597,
      "learning_rate": 4.995530249650701e-05,
      "loss": 1.4516,
      "step": 253
    },
    {
      "epoch": 0.03993004382086502,
      "grad_norm": 0.09942895174026489,
      "learning_rate": 4.995493244368268e-05,
      "loss": 1.4543,
      "step": 254
    },
    {
      "epoch": 0.040087248717797556,
      "grad_norm": 0.11218860745429993,
      "learning_rate": 4.995456086670602e-05,
      "loss": 1.4985,
      "step": 255
    },
    {
      "epoch": 0.0402444536147301,
      "grad_norm": 0.10839337855577469,
      "learning_rate": 4.9954187765599736e-05,
      "loss": 1.4805,
      "step": 256
    },
    {
      "epoch": 0.040401658511662636,
      "grad_norm": 0.10317599028348923,
      "learning_rate": 4.9953813140386595e-05,
      "loss": 1.4412,
      "step": 257
    },
    {
      "epoch": 0.04055886340859518,
      "grad_norm": 0.10285656154155731,
      "learning_rate": 4.99534369910895e-05,
      "loss": 1.476,
      "step": 258
    },
    {
      "epoch": 0.040716068305527715,
      "grad_norm": 0.10330680012702942,
      "learning_rate": 4.995305931773141e-05,
      "loss": 1.5157,
      "step": 259
    },
    {
      "epoch": 0.04087327320246026,
      "grad_norm": 0.1086694598197937,
      "learning_rate": 4.99526801203354e-05,
      "loss": 1.4999,
      "step": 260
    },
    {
      "epoch": 0.041030478099392795,
      "grad_norm": 0.10800144821405411,
      "learning_rate": 4.995229939892464e-05,
      "loss": 1.4764,
      "step": 261
    },
    {
      "epoch": 0.04118768299632534,
      "grad_norm": 0.10645303875207901,
      "learning_rate": 4.9951917153522355e-05,
      "loss": 1.4404,
      "step": 262
    },
    {
      "epoch": 0.041344887893257874,
      "grad_norm": 0.10440964996814728,
      "learning_rate": 4.9951533384151906e-05,
      "loss": 1.3678,
      "step": 263
    },
    {
      "epoch": 0.04150209279019042,
      "grad_norm": 0.10993078351020813,
      "learning_rate": 4.995114809083673e-05,
      "loss": 1.5064,
      "step": 264
    },
    {
      "epoch": 0.041659297687122954,
      "grad_norm": 0.10710245370864868,
      "learning_rate": 4.9950761273600366e-05,
      "loss": 1.4134,
      "step": 265
    },
    {
      "epoch": 0.04181650258405549,
      "grad_norm": 0.11030582338571548,
      "learning_rate": 4.995037293246644e-05,
      "loss": 1.5299,
      "step": 266
    },
    {
      "epoch": 0.04197370748098803,
      "grad_norm": 0.1058267131447792,
      "learning_rate": 4.994998306745866e-05,
      "loss": 1.3654,
      "step": 267
    },
    {
      "epoch": 0.04213091237792057,
      "grad_norm": 0.10541702806949615,
      "learning_rate": 4.994959167860084e-05,
      "loss": 1.4297,
      "step": 268
    },
    {
      "epoch": 0.04228811727485311,
      "grad_norm": 0.11085420846939087,
      "learning_rate": 4.994919876591689e-05,
      "loss": 1.4876,
      "step": 269
    },
    {
      "epoch": 0.04244532217178565,
      "grad_norm": 0.11054470390081406,
      "learning_rate": 4.994880432943081e-05,
      "loss": 1.574,
      "step": 270
    },
    {
      "epoch": 0.04260252706871819,
      "grad_norm": 0.11234510689973831,
      "learning_rate": 4.994840836916668e-05,
      "loss": 1.5079,
      "step": 271
    },
    {
      "epoch": 0.04275973196565073,
      "grad_norm": 0.11040106415748596,
      "learning_rate": 4.994801088514869e-05,
      "loss": 1.5091,
      "step": 272
    },
    {
      "epoch": 0.04291693686258327,
      "grad_norm": 0.10639887303113937,
      "learning_rate": 4.994761187740111e-05,
      "loss": 1.4495,
      "step": 273
    },
    {
      "epoch": 0.04307414175951581,
      "grad_norm": 0.11268071085214615,
      "learning_rate": 4.994721134594833e-05,
      "loss": 1.5057,
      "step": 274
    },
    {
      "epoch": 0.04323134665644835,
      "grad_norm": 0.10079260170459747,
      "learning_rate": 4.994680929081479e-05,
      "loss": 1.4145,
      "step": 275
    },
    {
      "epoch": 0.04338855155338089,
      "grad_norm": 0.11474710702896118,
      "learning_rate": 4.994640571202506e-05,
      "loss": 1.5061,
      "step": 276
    },
    {
      "epoch": 0.04354575645031343,
      "grad_norm": 0.10946876555681229,
      "learning_rate": 4.994600060960377e-05,
      "loss": 1.5306,
      "step": 277
    },
    {
      "epoch": 0.04370296134724597,
      "grad_norm": 0.11192137002944946,
      "learning_rate": 4.994559398357569e-05,
      "loss": 1.5347,
      "step": 278
    },
    {
      "epoch": 0.0438601662441785,
      "grad_norm": 0.10744784027338028,
      "learning_rate": 4.994518583396564e-05,
      "loss": 1.4686,
      "step": 279
    },
    {
      "epoch": 0.044017371141111046,
      "grad_norm": 0.11113352328538895,
      "learning_rate": 4.9944776160798544e-05,
      "loss": 1.4101,
      "step": 280
    },
    {
      "epoch": 0.04417457603804358,
      "grad_norm": 0.11456230282783508,
      "learning_rate": 4.994436496409943e-05,
      "loss": 1.4036,
      "step": 281
    },
    {
      "epoch": 0.044331780934976125,
      "grad_norm": 0.11608672887086868,
      "learning_rate": 4.994395224389342e-05,
      "loss": 1.4949,
      "step": 282
    },
    {
      "epoch": 0.04448898583190866,
      "grad_norm": 0.1232326403260231,
      "learning_rate": 4.9943538000205705e-05,
      "loss": 1.5501,
      "step": 283
    },
    {
      "epoch": 0.044646190728841205,
      "grad_norm": 0.11791515350341797,
      "learning_rate": 4.994312223306159e-05,
      "loss": 1.4542,
      "step": 284
    },
    {
      "epoch": 0.04480339562577374,
      "grad_norm": 0.11657550930976868,
      "learning_rate": 4.9942704942486476e-05,
      "loss": 1.4724,
      "step": 285
    },
    {
      "epoch": 0.044960600522706284,
      "grad_norm": 0.11560262739658356,
      "learning_rate": 4.994228612850584e-05,
      "loss": 1.4036,
      "step": 286
    },
    {
      "epoch": 0.04511780541963882,
      "grad_norm": 0.10999175906181335,
      "learning_rate": 4.994186579114527e-05,
      "loss": 1.4489,
      "step": 287
    },
    {
      "epoch": 0.045275010316571364,
      "grad_norm": 0.11586826294660568,
      "learning_rate": 4.9941443930430436e-05,
      "loss": 1.5486,
      "step": 288
    },
    {
      "epoch": 0.0454322152135039,
      "grad_norm": 0.11349951475858688,
      "learning_rate": 4.994102054638711e-05,
      "loss": 1.5698,
      "step": 289
    },
    {
      "epoch": 0.04558942011043644,
      "grad_norm": 0.11978698521852493,
      "learning_rate": 4.9940595639041134e-05,
      "loss": 1.3933,
      "step": 290
    },
    {
      "epoch": 0.04574662500736898,
      "grad_norm": 0.11438622325658798,
      "learning_rate": 4.994016920841846e-05,
      "loss": 1.5005,
      "step": 291
    },
    {
      "epoch": 0.045903829904301516,
      "grad_norm": 0.11395915597677231,
      "learning_rate": 4.9939741254545155e-05,
      "loss": 1.4521,
      "step": 292
    },
    {
      "epoch": 0.04606103480123406,
      "grad_norm": 0.11659599095582962,
      "learning_rate": 4.993931177744734e-05,
      "loss": 1.5166,
      "step": 293
    },
    {
      "epoch": 0.046218239698166595,
      "grad_norm": 0.11053171753883362,
      "learning_rate": 4.9938880777151254e-05,
      "loss": 1.4459,
      "step": 294
    },
    {
      "epoch": 0.04637544459509914,
      "grad_norm": 0.11428084224462509,
      "learning_rate": 4.993844825368321e-05,
      "loss": 1.4448,
      "step": 295
    },
    {
      "epoch": 0.046532649492031675,
      "grad_norm": 0.10734150558710098,
      "learning_rate": 4.993801420706964e-05,
      "loss": 1.3388,
      "step": 296
    },
    {
      "epoch": 0.04668985438896422,
      "grad_norm": 0.11137369275093079,
      "learning_rate": 4.993757863733703e-05,
      "loss": 1.4155,
      "step": 297
    },
    {
      "epoch": 0.046847059285896754,
      "grad_norm": 0.1221408098936081,
      "learning_rate": 4.993714154451202e-05,
      "loss": 1.4884,
      "step": 298
    },
    {
      "epoch": 0.0470042641828293,
      "grad_norm": 0.11707969009876251,
      "learning_rate": 4.993670292862127e-05,
      "loss": 1.4605,
      "step": 299
    },
    {
      "epoch": 0.047161469079761834,
      "grad_norm": 0.11751751601696014,
      "learning_rate": 4.993626278969158e-05,
      "loss": 1.5538,
      "step": 300
    },
    {
      "epoch": 0.04731867397669438,
      "grad_norm": 0.11617731302976608,
      "learning_rate": 4.993582112774984e-05,
      "loss": 1.438,
      "step": 301
    },
    {
      "epoch": 0.04747587887362691,
      "grad_norm": 0.15164637565612793,
      "learning_rate": 4.993537794282302e-05,
      "loss": 1.4607,
      "step": 302
    },
    {
      "epoch": 0.047633083770559456,
      "grad_norm": 0.12434446811676025,
      "learning_rate": 4.9934933234938193e-05,
      "loss": 1.4167,
      "step": 303
    },
    {
      "epoch": 0.04779028866749199,
      "grad_norm": 0.12518739700317383,
      "learning_rate": 4.993448700412251e-05,
      "loss": 1.4003,
      "step": 304
    },
    {
      "epoch": 0.04794749356442453,
      "grad_norm": 0.11146944761276245,
      "learning_rate": 4.993403925040323e-05,
      "loss": 1.3913,
      "step": 305
    },
    {
      "epoch": 0.04810469846135707,
      "grad_norm": 0.11682326346635818,
      "learning_rate": 4.993358997380771e-05,
      "loss": 1.3415,
      "step": 306
    },
    {
      "epoch": 0.04826190335828961,
      "grad_norm": 0.1197504773736,
      "learning_rate": 4.993313917436336e-05,
      "loss": 1.515,
      "step": 307
    },
    {
      "epoch": 0.04841910825522215,
      "grad_norm": 0.14647473394870758,
      "learning_rate": 4.993268685209775e-05,
      "loss": 1.4529,
      "step": 308
    },
    {
      "epoch": 0.04857631315215469,
      "grad_norm": 0.12431525439023972,
      "learning_rate": 4.9932233007038484e-05,
      "loss": 1.5426,
      "step": 309
    },
    {
      "epoch": 0.04873351804908723,
      "grad_norm": 0.11715538799762726,
      "learning_rate": 4.9931777639213284e-05,
      "loss": 1.4615,
      "step": 310
    },
    {
      "epoch": 0.04889072294601977,
      "grad_norm": 0.12391876429319382,
      "learning_rate": 4.993132074864997e-05,
      "loss": 1.4138,
      "step": 311
    },
    {
      "epoch": 0.04904792784295231,
      "grad_norm": 0.11894181370735168,
      "learning_rate": 4.9930862335376444e-05,
      "loss": 1.4383,
      "step": 312
    },
    {
      "epoch": 0.049205132739884846,
      "grad_norm": 0.1225295439362526,
      "learning_rate": 4.9930402399420695e-05,
      "loss": 1.3847,
      "step": 313
    },
    {
      "epoch": 0.04936233763681739,
      "grad_norm": 0.11435995995998383,
      "learning_rate": 4.9929940940810825e-05,
      "loss": 1.4254,
      "step": 314
    },
    {
      "epoch": 0.049519542533749926,
      "grad_norm": 0.11988761276006699,
      "learning_rate": 4.9929477959575024e-05,
      "loss": 1.4787,
      "step": 315
    },
    {
      "epoch": 0.04967674743068247,
      "grad_norm": 0.11983373016119003,
      "learning_rate": 4.992901345574155e-05,
      "loss": 1.4341,
      "step": 316
    },
    {
      "epoch": 0.049833952327615005,
      "grad_norm": 0.13395054638385773,
      "learning_rate": 4.992854742933878e-05,
      "loss": 1.4315,
      "step": 317
    },
    {
      "epoch": 0.04999115722454754,
      "grad_norm": 0.12578143179416656,
      "learning_rate": 4.9928079880395186e-05,
      "loss": 1.4143,
      "step": 318
    },
    {
      "epoch": 0.050148362121480085,
      "grad_norm": 0.1401878446340561,
      "learning_rate": 4.992761080893932e-05,
      "loss": 1.4665,
      "step": 319
    },
    {
      "epoch": 0.05030556701841262,
      "grad_norm": 0.13048145174980164,
      "learning_rate": 4.9927140214999826e-05,
      "loss": 1.4266,
      "step": 320
    },
    {
      "epoch": 0.05030556701841262,
      "eval_loss": 1.450086236000061,
      "eval_runtime": 2316.1877,
      "eval_samples_per_second": 3.997,
      "eval_steps_per_second": 1.999,
      "step": 320
    },
    {
      "epoch": 0.050462771915345164,
      "grad_norm": 0.13121232390403748,
      "learning_rate": 4.992666809860545e-05,
      "loss": 1.4946,
      "step": 321
    },
    {
      "epoch": 0.0506199768122777,
      "grad_norm": 0.13547195494174957,
      "learning_rate": 4.9926194459785015e-05,
      "loss": 1.5532,
      "step": 322
    },
    {
      "epoch": 0.050777181709210244,
      "grad_norm": 0.11797169595956802,
      "learning_rate": 4.992571929856747e-05,
      "loss": 1.4118,
      "step": 323
    },
    {
      "epoch": 0.05093438660614278,
      "grad_norm": 0.12734922766685486,
      "learning_rate": 4.992524261498183e-05,
      "loss": 1.4427,
      "step": 324
    },
    {
      "epoch": 0.05109159150307532,
      "grad_norm": 0.12444902211427689,
      "learning_rate": 4.99247644090572e-05,
      "loss": 1.4369,
      "step": 325
    },
    {
      "epoch": 0.05124879640000786,
      "grad_norm": 0.12244518846273422,
      "learning_rate": 4.99242846808228e-05,
      "loss": 1.4587,
      "step": 326
    },
    {
      "epoch": 0.0514060012969404,
      "grad_norm": 0.12424397468566895,
      "learning_rate": 4.9923803430307916e-05,
      "loss": 1.3949,
      "step": 327
    },
    {
      "epoch": 0.05156320619387294,
      "grad_norm": 0.1352718621492386,
      "learning_rate": 4.9923320657541944e-05,
      "loss": 1.504,
      "step": 328
    },
    {
      "epoch": 0.05172041109080548,
      "grad_norm": 0.12855666875839233,
      "learning_rate": 4.992283636255438e-05,
      "loss": 1.4271,
      "step": 329
    },
    {
      "epoch": 0.05187761598773802,
      "grad_norm": 0.129829540848732,
      "learning_rate": 4.99223505453748e-05,
      "loss": 1.455,
      "step": 330
    },
    {
      "epoch": 0.052034820884670555,
      "grad_norm": 0.12780050933361053,
      "learning_rate": 4.992186320603286e-05,
      "loss": 1.4045,
      "step": 331
    },
    {
      "epoch": 0.0521920257816031,
      "grad_norm": 0.13515712320804596,
      "learning_rate": 4.992137434455834e-05,
      "loss": 1.4335,
      "step": 332
    },
    {
      "epoch": 0.052349230678535634,
      "grad_norm": 0.15026766061782837,
      "learning_rate": 4.99208839609811e-05,
      "loss": 1.5386,
      "step": 333
    },
    {
      "epoch": 0.05250643557546818,
      "grad_norm": 0.13422101736068726,
      "learning_rate": 4.992039205533108e-05,
      "loss": 1.454,
      "step": 334
    },
    {
      "epoch": 0.05266364047240071,
      "grad_norm": 0.13735777139663696,
      "learning_rate": 4.991989862763833e-05,
      "loss": 1.4415,
      "step": 335
    },
    {
      "epoch": 0.05282084536933326,
      "grad_norm": 0.12985137104988098,
      "learning_rate": 4.9919403677932994e-05,
      "loss": 1.385,
      "step": 336
    },
    {
      "epoch": 0.05297805026626579,
      "grad_norm": 0.1301167607307434,
      "learning_rate": 4.9918907206245285e-05,
      "loss": 1.4364,
      "step": 337
    },
    {
      "epoch": 0.053135255163198336,
      "grad_norm": 0.1407599002122879,
      "learning_rate": 4.991840921260553e-05,
      "loss": 1.4454,
      "step": 338
    },
    {
      "epoch": 0.05329246006013087,
      "grad_norm": 0.12763133645057678,
      "learning_rate": 4.9917909697044164e-05,
      "loss": 1.4008,
      "step": 339
    },
    {
      "epoch": 0.053449664957063416,
      "grad_norm": 0.1443052589893341,
      "learning_rate": 4.991740865959167e-05,
      "loss": 1.5184,
      "step": 340
    },
    {
      "epoch": 0.05360686985399595,
      "grad_norm": 0.13496418297290802,
      "learning_rate": 4.991690610027866e-05,
      "loss": 1.3888,
      "step": 341
    },
    {
      "epoch": 0.05376407475092849,
      "grad_norm": 0.12681293487548828,
      "learning_rate": 4.991640201913583e-05,
      "loss": 1.42,
      "step": 342
    },
    {
      "epoch": 0.05392127964786103,
      "grad_norm": 0.13178062438964844,
      "learning_rate": 4.9915896416193965e-05,
      "loss": 1.4178,
      "step": 343
    },
    {
      "epoch": 0.05407848454479357,
      "grad_norm": 0.14452503621578217,
      "learning_rate": 4.991538929148394e-05,
      "loss": 1.4248,
      "step": 344
    },
    {
      "epoch": 0.05423568944172611,
      "grad_norm": 0.1352955400943756,
      "learning_rate": 4.991488064503674e-05,
      "loss": 1.4304,
      "step": 345
    },
    {
      "epoch": 0.05439289433865865,
      "grad_norm": 0.14846469461917877,
      "learning_rate": 4.991437047688343e-05,
      "loss": 1.4784,
      "step": 346
    },
    {
      "epoch": 0.05455009923559119,
      "grad_norm": 0.12475849688053131,
      "learning_rate": 4.9913858787055156e-05,
      "loss": 1.4131,
      "step": 347
    },
    {
      "epoch": 0.054707304132523726,
      "grad_norm": 0.13835409283638,
      "learning_rate": 4.991334557558318e-05,
      "loss": 1.4913,
      "step": 348
    },
    {
      "epoch": 0.05486450902945627,
      "grad_norm": 0.13921529054641724,
      "learning_rate": 4.991283084249885e-05,
      "loss": 1.3713,
      "step": 349
    },
    {
      "epoch": 0.055021713926388806,
      "grad_norm": 0.13188250362873077,
      "learning_rate": 4.9912314587833586e-05,
      "loss": 1.3608,
      "step": 350
    },
    {
      "epoch": 0.05517891882332135,
      "grad_norm": 0.12457428872585297,
      "learning_rate": 4.991179681161895e-05,
      "loss": 1.4427,
      "step": 351
    },
    {
      "epoch": 0.055336123720253885,
      "grad_norm": 0.12452542781829834,
      "learning_rate": 4.9911277513886535e-05,
      "loss": 1.4179,
      "step": 352
    },
    {
      "epoch": 0.05549332861718643,
      "grad_norm": 0.14799195528030396,
      "learning_rate": 4.9910756694668074e-05,
      "loss": 1.4532,
      "step": 353
    },
    {
      "epoch": 0.055650533514118965,
      "grad_norm": 0.13485541939735413,
      "learning_rate": 4.991023435399538e-05,
      "loss": 1.4114,
      "step": 354
    },
    {
      "epoch": 0.0558077384110515,
      "grad_norm": 0.1422443389892578,
      "learning_rate": 4.990971049190034e-05,
      "loss": 1.377,
      "step": 355
    },
    {
      "epoch": 0.055964943307984044,
      "grad_norm": 0.12994804978370667,
      "learning_rate": 4.990918510841496e-05,
      "loss": 1.4474,
      "step": 356
    },
    {
      "epoch": 0.05612214820491658,
      "grad_norm": 0.1429785192012787,
      "learning_rate": 4.990865820357133e-05,
      "loss": 1.4435,
      "step": 357
    },
    {
      "epoch": 0.056279353101849124,
      "grad_norm": 0.12979790568351746,
      "learning_rate": 4.9908129777401625e-05,
      "loss": 1.4039,
      "step": 358
    },
    {
      "epoch": 0.05643655799878166,
      "grad_norm": 0.1332644671201706,
      "learning_rate": 4.990759982993812e-05,
      "loss": 1.4377,
      "step": 359
    },
    {
      "epoch": 0.0565937628957142,
      "grad_norm": 0.13796579837799072,
      "learning_rate": 4.99070683612132e-05,
      "loss": 1.3951,
      "step": 360
    },
    {
      "epoch": 0.05675096779264674,
      "grad_norm": 0.14315246045589447,
      "learning_rate": 4.9906535371259294e-05,
      "loss": 1.4042,
      "step": 361
    },
    {
      "epoch": 0.05690817268957928,
      "grad_norm": 0.1463768631219864,
      "learning_rate": 4.9906000860108974e-05,
      "loss": 1.461,
      "step": 362
    },
    {
      "epoch": 0.05706537758651182,
      "grad_norm": 0.14041170477867126,
      "learning_rate": 4.9905464827794884e-05,
      "loss": 1.4147,
      "step": 363
    },
    {
      "epoch": 0.05722258248344436,
      "grad_norm": 0.19242697954177856,
      "learning_rate": 4.990492727434976e-05,
      "loss": 1.3435,
      "step": 364
    },
    {
      "epoch": 0.0573797873803769,
      "grad_norm": 0.1556611955165863,
      "learning_rate": 4.990438819980644e-05,
      "loss": 1.4075,
      "step": 365
    },
    {
      "epoch": 0.05753699227730944,
      "grad_norm": 0.13157570362091064,
      "learning_rate": 4.990384760419784e-05,
      "loss": 1.3334,
      "step": 366
    },
    {
      "epoch": 0.05769419717424198,
      "grad_norm": 0.17953743040561676,
      "learning_rate": 4.990330548755698e-05,
      "loss": 1.4609,
      "step": 367
    },
    {
      "epoch": 0.057851402071174514,
      "grad_norm": 0.14179491996765137,
      "learning_rate": 4.990276184991697e-05,
      "loss": 1.4344,
      "step": 368
    },
    {
      "epoch": 0.05800860696810706,
      "grad_norm": 0.16522593796253204,
      "learning_rate": 4.9902216691311024e-05,
      "loss": 1.3794,
      "step": 369
    },
    {
      "epoch": 0.05816581186503959,
      "grad_norm": 0.12736016511917114,
      "learning_rate": 4.9901670011772425e-05,
      "loss": 1.4167,
      "step": 370
    },
    {
      "epoch": 0.058323016761972137,
      "grad_norm": 0.15869787335395813,
      "learning_rate": 4.990112181133456e-05,
      "loss": 1.4293,
      "step": 371
    },
    {
      "epoch": 0.05848022165890467,
      "grad_norm": 0.14410504698753357,
      "learning_rate": 4.990057209003093e-05,
      "loss": 1.4357,
      "step": 372
    },
    {
      "epoch": 0.058637426555837216,
      "grad_norm": 0.1567080020904541,
      "learning_rate": 4.9900020847895086e-05,
      "loss": 1.4146,
      "step": 373
    },
    {
      "epoch": 0.05879463145276975,
      "grad_norm": 0.1430107057094574,
      "learning_rate": 4.989946808496071e-05,
      "loss": 1.3415,
      "step": 374
    },
    {
      "epoch": 0.058951836349702295,
      "grad_norm": 0.146332785487175,
      "learning_rate": 4.989891380126156e-05,
      "loss": 1.496,
      "step": 375
    },
    {
      "epoch": 0.05910904124663483,
      "grad_norm": 0.13674487173557281,
      "learning_rate": 4.989835799683149e-05,
      "loss": 1.3611,
      "step": 376
    },
    {
      "epoch": 0.059266246143567375,
      "grad_norm": 0.1321984827518463,
      "learning_rate": 4.989780067170444e-05,
      "loss": 1.4695,
      "step": 377
    },
    {
      "epoch": 0.05942345104049991,
      "grad_norm": 0.1535942554473877,
      "learning_rate": 4.9897241825914464e-05,
      "loss": 1.3564,
      "step": 378
    },
    {
      "epoch": 0.059580655937432454,
      "grad_norm": 0.1538037806749344,
      "learning_rate": 4.989668145949568e-05,
      "loss": 1.3502,
      "step": 379
    },
    {
      "epoch": 0.05973786083436499,
      "grad_norm": 0.15744829177856445,
      "learning_rate": 4.989611957248232e-05,
      "loss": 1.4318,
      "step": 380
    },
    {
      "epoch": 0.05989506573129753,
      "grad_norm": 0.17178332805633545,
      "learning_rate": 4.98955561649087e-05,
      "loss": 1.4643,
      "step": 381
    },
    {
      "epoch": 0.06005227062823007,
      "grad_norm": 0.15913072228431702,
      "learning_rate": 4.989499123680923e-05,
      "loss": 1.487,
      "step": 382
    },
    {
      "epoch": 0.060209475525162606,
      "grad_norm": 0.15134060382843018,
      "learning_rate": 4.9894424788218415e-05,
      "loss": 1.4705,
      "step": 383
    },
    {
      "epoch": 0.06036668042209515,
      "grad_norm": 0.13704389333724976,
      "learning_rate": 4.989385681917085e-05,
      "loss": 1.4756,
      "step": 384
    },
    {
      "epoch": 0.060523885319027686,
      "grad_norm": 0.14025503396987915,
      "learning_rate": 4.989328732970122e-05,
      "loss": 1.443,
      "step": 385
    },
    {
      "epoch": 0.06068109021596023,
      "grad_norm": 0.1822325438261032,
      "learning_rate": 4.9892716319844325e-05,
      "loss": 1.3996,
      "step": 386
    },
    {
      "epoch": 0.060838295112892765,
      "grad_norm": 0.15639656782150269,
      "learning_rate": 4.989214378963502e-05,
      "loss": 1.3656,
      "step": 387
    },
    {
      "epoch": 0.06099550000982531,
      "grad_norm": 0.15097728371620178,
      "learning_rate": 4.989156973910828e-05,
      "loss": 1.4055,
      "step": 388
    },
    {
      "epoch": 0.061152704906757845,
      "grad_norm": 0.18977142870426178,
      "learning_rate": 4.989099416829917e-05,
      "loss": 1.4472,
      "step": 389
    },
    {
      "epoch": 0.06130990980369039,
      "grad_norm": 0.1596304178237915,
      "learning_rate": 4.989041707724284e-05,
      "loss": 1.4373,
      "step": 390
    },
    {
      "epoch": 0.061467114700622924,
      "grad_norm": 0.171820729970932,
      "learning_rate": 4.988983846597454e-05,
      "loss": 1.468,
      "step": 391
    },
    {
      "epoch": 0.06162431959755547,
      "grad_norm": 0.14266176521778107,
      "learning_rate": 4.98892583345296e-05,
      "loss": 1.4037,
      "step": 392
    },
    {
      "epoch": 0.061781524494488003,
      "grad_norm": 0.13375528156757355,
      "learning_rate": 4.988867668294346e-05,
      "loss": 1.437,
      "step": 393
    },
    {
      "epoch": 0.06193872939142054,
      "grad_norm": 0.13332228362560272,
      "learning_rate": 4.988809351125165e-05,
      "loss": 1.3892,
      "step": 394
    },
    {
      "epoch": 0.06209593428835308,
      "grad_norm": 0.17180980741977692,
      "learning_rate": 4.988750881948977e-05,
      "loss": 1.3494,
      "step": 395
    },
    {
      "epoch": 0.06225313918528562,
      "grad_norm": 0.1419111043214798,
      "learning_rate": 4.988692260769355e-05,
      "loss": 1.3748,
      "step": 396
    },
    {
      "epoch": 0.06241034408221816,
      "grad_norm": 0.17256620526313782,
      "learning_rate": 4.9886334875898776e-05,
      "loss": 1.3549,
      "step": 397
    },
    {
      "epoch": 0.0625675489791507,
      "grad_norm": 0.2243422418832779,
      "learning_rate": 4.988574562414137e-05,
      "loss": 1.4465,
      "step": 398
    },
    {
      "epoch": 0.06272475387608324,
      "grad_norm": 0.15700723230838776,
      "learning_rate": 4.9885154852457294e-05,
      "loss": 1.4477,
      "step": 399
    },
    {
      "epoch": 0.06288195877301578,
      "grad_norm": 0.14497259259223938,
      "learning_rate": 4.988456256088264e-05,
      "loss": 1.3861,
      "step": 400
    },
    {
      "epoch": 0.06303916366994831,
      "grad_norm": 0.14747034013271332,
      "learning_rate": 4.988396874945359e-05,
      "loss": 1.4206,
      "step": 401
    },
    {
      "epoch": 0.06319636856688086,
      "grad_norm": 0.17671054601669312,
      "learning_rate": 4.98833734182064e-05,
      "loss": 1.2475,
      "step": 402
    },
    {
      "epoch": 0.0633535734638134,
      "grad_norm": 0.16974316537380219,
      "learning_rate": 4.9882776567177446e-05,
      "loss": 1.4955,
      "step": 403
    },
    {
      "epoch": 0.06351077836074594,
      "grad_norm": 0.15419775247573853,
      "learning_rate": 4.988217819640317e-05,
      "loss": 1.4209,
      "step": 404
    },
    {
      "epoch": 0.06366798325767847,
      "grad_norm": 0.13987664878368378,
      "learning_rate": 4.988157830592012e-05,
      "loss": 1.456,
      "step": 405
    },
    {
      "epoch": 0.06382518815461101,
      "grad_norm": 0.24560455977916718,
      "learning_rate": 4.988097689576493e-05,
      "loss": 1.3567,
      "step": 406
    },
    {
      "epoch": 0.06398239305154356,
      "grad_norm": 0.13870076835155487,
      "learning_rate": 4.9880373965974334e-05,
      "loss": 1.3752,
      "step": 407
    },
    {
      "epoch": 0.0641395979484761,
      "grad_norm": 0.16167718172073364,
      "learning_rate": 4.987976951658517e-05,
      "loss": 1.4766,
      "step": 408
    },
    {
      "epoch": 0.06429680284540863,
      "grad_norm": 0.1700398474931717,
      "learning_rate": 4.9879163547634346e-05,
      "loss": 1.427,
      "step": 409
    },
    {
      "epoch": 0.06445400774234117,
      "grad_norm": 0.15502458810806274,
      "learning_rate": 4.987855605915887e-05,
      "loss": 1.3965,
      "step": 410
    },
    {
      "epoch": 0.06461121263927372,
      "grad_norm": 0.14834032952785492,
      "learning_rate": 4.987794705119584e-05,
      "loss": 1.4399,
      "step": 411
    },
    {
      "epoch": 0.06476841753620625,
      "grad_norm": 0.22443649172782898,
      "learning_rate": 4.987733652378246e-05,
      "loss": 1.3736,
      "step": 412
    },
    {
      "epoch": 0.06492562243313879,
      "grad_norm": 0.14396560192108154,
      "learning_rate": 4.9876724476956015e-05,
      "loss": 1.4648,
      "step": 413
    },
    {
      "epoch": 0.06508282733007133,
      "grad_norm": 0.15352006256580353,
      "learning_rate": 4.987611091075389e-05,
      "loss": 1.4988,
      "step": 414
    },
    {
      "epoch": 0.06524003222700388,
      "grad_norm": 0.13210074603557587,
      "learning_rate": 4.987549582521356e-05,
      "loss": 1.3705,
      "step": 415
    },
    {
      "epoch": 0.06539723712393641,
      "grad_norm": 0.16056782007217407,
      "learning_rate": 4.98748792203726e-05,
      "loss": 1.3388,
      "step": 416
    },
    {
      "epoch": 0.06555444202086895,
      "grad_norm": 0.18992343544960022,
      "learning_rate": 4.9874261096268647e-05,
      "loss": 1.3842,
      "step": 417
    },
    {
      "epoch": 0.06571164691780149,
      "grad_norm": 0.1789916455745697,
      "learning_rate": 4.9873641452939466e-05,
      "loss": 1.3622,
      "step": 418
    },
    {
      "epoch": 0.06586885181473402,
      "grad_norm": 0.21043789386749268,
      "learning_rate": 4.9873020290422915e-05,
      "loss": 1.3477,
      "step": 419
    },
    {
      "epoch": 0.06602605671166657,
      "grad_norm": 0.15355254709720612,
      "learning_rate": 4.987239760875691e-05,
      "loss": 1.3643,
      "step": 420
    },
    {
      "epoch": 0.06618326160859911,
      "grad_norm": 0.1433190107345581,
      "learning_rate": 4.9871773407979496e-05,
      "loss": 1.3753,
      "step": 421
    },
    {
      "epoch": 0.06634046650553165,
      "grad_norm": 0.17479249835014343,
      "learning_rate": 4.987114768812879e-05,
      "loss": 1.3809,
      "step": 422
    },
    {
      "epoch": 0.06649767140246418,
      "grad_norm": 0.186944842338562,
      "learning_rate": 4.987052044924302e-05,
      "loss": 1.3616,
      "step": 423
    },
    {
      "epoch": 0.06665487629939673,
      "grad_norm": 0.15202952921390533,
      "learning_rate": 4.986989169136048e-05,
      "loss": 1.4479,
      "step": 424
    },
    {
      "epoch": 0.06681208119632927,
      "grad_norm": 0.16295532882213593,
      "learning_rate": 4.9869261414519575e-05,
      "loss": 1.3713,
      "step": 425
    },
    {
      "epoch": 0.0669692860932618,
      "grad_norm": 0.19577625393867493,
      "learning_rate": 4.986862961875881e-05,
      "loss": 1.4199,
      "step": 426
    },
    {
      "epoch": 0.06712649099019434,
      "grad_norm": 0.22768542170524597,
      "learning_rate": 4.986799630411677e-05,
      "loss": 1.3529,
      "step": 427
    },
    {
      "epoch": 0.06728369588712689,
      "grad_norm": 0.25184011459350586,
      "learning_rate": 4.986736147063212e-05,
      "loss": 1.3944,
      "step": 428
    },
    {
      "epoch": 0.06744090078405943,
      "grad_norm": 0.15565118193626404,
      "learning_rate": 4.986672511834366e-05,
      "loss": 1.4505,
      "step": 429
    },
    {
      "epoch": 0.06759810568099196,
      "grad_norm": 0.16559922695159912,
      "learning_rate": 4.986608724729024e-05,
      "loss": 1.3742,
      "step": 430
    },
    {
      "epoch": 0.0677553105779245,
      "grad_norm": 0.14826242625713348,
      "learning_rate": 4.986544785751081e-05,
      "loss": 1.4008,
      "step": 431
    },
    {
      "epoch": 0.06791251547485704,
      "grad_norm": 0.16543184220790863,
      "learning_rate": 4.986480694904444e-05,
      "loss": 1.3433,
      "step": 432
    },
    {
      "epoch": 0.06806972037178959,
      "grad_norm": 0.15332931280136108,
      "learning_rate": 4.986416452193027e-05,
      "loss": 1.4459,
      "step": 433
    },
    {
      "epoch": 0.06822692526872212,
      "grad_norm": 0.18880733847618103,
      "learning_rate": 4.986352057620752e-05,
      "loss": 1.3902,
      "step": 434
    },
    {
      "epoch": 0.06838413016565466,
      "grad_norm": 0.1513829231262207,
      "learning_rate": 4.986287511191554e-05,
      "loss": 1.3485,
      "step": 435
    },
    {
      "epoch": 0.0685413350625872,
      "grad_norm": 0.15241704881191254,
      "learning_rate": 4.9862228129093745e-05,
      "loss": 1.3051,
      "step": 436
    },
    {
      "epoch": 0.06869853995951974,
      "grad_norm": 0.1956702321767807,
      "learning_rate": 4.986157962778165e-05,
      "loss": 1.4647,
      "step": 437
    },
    {
      "epoch": 0.06885574485645228,
      "grad_norm": 0.2027936428785324,
      "learning_rate": 4.9860929608018866e-05,
      "loss": 1.3602,
      "step": 438
    },
    {
      "epoch": 0.06901294975338482,
      "grad_norm": 0.1623186320066452,
      "learning_rate": 4.986027806984509e-05,
      "loss": 1.4154,
      "step": 439
    },
    {
      "epoch": 0.06917015465031735,
      "grad_norm": 0.16111283004283905,
      "learning_rate": 4.985962501330011e-05,
      "loss": 1.4311,
      "step": 440
    },
    {
      "epoch": 0.0693273595472499,
      "grad_norm": 0.16754299402236938,
      "learning_rate": 4.985897043842382e-05,
      "loss": 1.349,
      "step": 441
    },
    {
      "epoch": 0.06948456444418244,
      "grad_norm": 0.1766330897808075,
      "learning_rate": 4.985831434525621e-05,
      "loss": 1.3714,
      "step": 442
    },
    {
      "epoch": 0.06964176934111498,
      "grad_norm": 0.1742810308933258,
      "learning_rate": 4.985765673383733e-05,
      "loss": 1.4161,
      "step": 443
    },
    {
      "epoch": 0.06979897423804751,
      "grad_norm": 0.17025281488895416,
      "learning_rate": 4.985699760420736e-05,
      "loss": 1.3925,
      "step": 444
    },
    {
      "epoch": 0.06995617913498005,
      "grad_norm": 0.19201375544071198,
      "learning_rate": 4.985633695640655e-05,
      "loss": 1.4158,
      "step": 445
    },
    {
      "epoch": 0.0701133840319126,
      "grad_norm": 0.1636267751455307,
      "learning_rate": 4.985567479047524e-05,
      "loss": 1.4071,
      "step": 446
    },
    {
      "epoch": 0.07027058892884513,
      "grad_norm": 0.19676333665847778,
      "learning_rate": 4.9855011106453894e-05,
      "loss": 1.3449,
      "step": 447
    },
    {
      "epoch": 0.07042779382577767,
      "grad_norm": 0.17712907493114471,
      "learning_rate": 4.985434590438303e-05,
      "loss": 1.3421,
      "step": 448
    },
    {
      "epoch": 0.07058499872271021,
      "grad_norm": 0.18515101075172424,
      "learning_rate": 4.985367918430329e-05,
      "loss": 1.4051,
      "step": 449
    },
    {
      "epoch": 0.07074220361964276,
      "grad_norm": 0.17168915271759033,
      "learning_rate": 4.985301094625538e-05,
      "loss": 1.3093,
      "step": 450
    },
    {
      "epoch": 0.0708994085165753,
      "grad_norm": 0.1891397386789322,
      "learning_rate": 4.9852341190280127e-05,
      "loss": 1.3075,
      "step": 451
    },
    {
      "epoch": 0.07105661341350783,
      "grad_norm": 0.17731457948684692,
      "learning_rate": 4.985166991641843e-05,
      "loss": 1.3986,
      "step": 452
    },
    {
      "epoch": 0.07121381831044037,
      "grad_norm": 0.18817296624183655,
      "learning_rate": 4.985099712471129e-05,
      "loss": 1.3531,
      "step": 453
    },
    {
      "epoch": 0.07137102320737292,
      "grad_norm": 0.1782791018486023,
      "learning_rate": 4.9850322815199795e-05,
      "loss": 1.4064,
      "step": 454
    },
    {
      "epoch": 0.07152822810430545,
      "grad_norm": 0.18053874373435974,
      "learning_rate": 4.984964698792514e-05,
      "loss": 1.4607,
      "step": 455
    },
    {
      "epoch": 0.07168543300123799,
      "grad_norm": 0.286338746547699,
      "learning_rate": 4.984896964292858e-05,
      "loss": 1.3036,
      "step": 456
    },
    {
      "epoch": 0.07184263789817052,
      "grad_norm": 0.2560707926750183,
      "learning_rate": 4.98482907802515e-05,
      "loss": 1.3428,
      "step": 457
    },
    {
      "epoch": 0.07199984279510306,
      "grad_norm": 0.19296897947788239,
      "learning_rate": 4.984761039993537e-05,
      "loss": 1.3502,
      "step": 458
    },
    {
      "epoch": 0.07215704769203561,
      "grad_norm": 0.19685949385166168,
      "learning_rate": 4.9846928502021725e-05,
      "loss": 1.4015,
      "step": 459
    },
    {
      "epoch": 0.07231425258896815,
      "grad_norm": 0.1548481583595276,
      "learning_rate": 4.984624508655223e-05,
      "loss": 1.3698,
      "step": 460
    },
    {
      "epoch": 0.07247145748590068,
      "grad_norm": 0.16076034307479858,
      "learning_rate": 4.984556015356862e-05,
      "loss": 1.3627,
      "step": 461
    },
    {
      "epoch": 0.07262866238283322,
      "grad_norm": 0.18571603298187256,
      "learning_rate": 4.9844873703112726e-05,
      "loss": 1.3506,
      "step": 462
    },
    {
      "epoch": 0.07278586727976577,
      "grad_norm": 0.1540035605430603,
      "learning_rate": 4.984418573522648e-05,
      "loss": 1.4483,
      "step": 463
    },
    {
      "epoch": 0.0729430721766983,
      "grad_norm": 0.1730145364999771,
      "learning_rate": 4.984349624995188e-05,
      "loss": 1.3678,
      "step": 464
    },
    {
      "epoch": 0.07310027707363084,
      "grad_norm": 0.26254212856292725,
      "learning_rate": 4.984280524733107e-05,
      "loss": 1.401,
      "step": 465
    },
    {
      "epoch": 0.07325748197056338,
      "grad_norm": 0.2079063057899475,
      "learning_rate": 4.984211272740623e-05,
      "loss": 1.3655,
      "step": 466
    },
    {
      "epoch": 0.07341468686749593,
      "grad_norm": 0.21711499989032745,
      "learning_rate": 4.9841418690219653e-05,
      "loss": 1.4011,
      "step": 467
    },
    {
      "epoch": 0.07357189176442847,
      "grad_norm": 0.18226252496242523,
      "learning_rate": 4.984072313581375e-05,
      "loss": 1.4213,
      "step": 468
    },
    {
      "epoch": 0.073729096661361,
      "grad_norm": 0.1463780552148819,
      "learning_rate": 4.9840026064230984e-05,
      "loss": 1.4519,
      "step": 469
    },
    {
      "epoch": 0.07388630155829354,
      "grad_norm": 0.18232892453670502,
      "learning_rate": 4.983932747551394e-05,
      "loss": 1.3657,
      "step": 470
    },
    {
      "epoch": 0.07404350645522607,
      "grad_norm": 0.19644559919834137,
      "learning_rate": 4.9838627369705285e-05,
      "loss": 1.3988,
      "step": 471
    },
    {
      "epoch": 0.07420071135215862,
      "grad_norm": 0.16292576491832733,
      "learning_rate": 4.983792574684776e-05,
      "loss": 1.4369,
      "step": 472
    },
    {
      "epoch": 0.07435791624909116,
      "grad_norm": 0.2244543433189392,
      "learning_rate": 4.983722260698425e-05,
      "loss": 1.4269,
      "step": 473
    },
    {
      "epoch": 0.0745151211460237,
      "grad_norm": 0.2582489848136902,
      "learning_rate": 4.9836517950157666e-05,
      "loss": 1.3986,
      "step": 474
    },
    {
      "epoch": 0.07467232604295623,
      "grad_norm": 0.15564194321632385,
      "learning_rate": 4.983581177641108e-05,
      "loss": 1.3871,
      "step": 475
    },
    {
      "epoch": 0.07482953093988878,
      "grad_norm": 0.2301008552312851,
      "learning_rate": 4.9835104085787596e-05,
      "loss": 1.3572,
      "step": 476
    },
    {
      "epoch": 0.07498673583682132,
      "grad_norm": 0.21603424847126007,
      "learning_rate": 4.9834394878330444e-05,
      "loss": 1.3803,
      "step": 477
    },
    {
      "epoch": 0.07514394073375386,
      "grad_norm": 0.16744717955589294,
      "learning_rate": 4.9833684154082937e-05,
      "loss": 1.4233,
      "step": 478
    },
    {
      "epoch": 0.07530114563068639,
      "grad_norm": 0.23016415536403656,
      "learning_rate": 4.98329719130885e-05,
      "loss": 1.3962,
      "step": 479
    },
    {
      "epoch": 0.07545835052761894,
      "grad_norm": 0.19687114655971527,
      "learning_rate": 4.983225815539061e-05,
      "loss": 1.3667,
      "step": 480
    },
    {
      "epoch": 0.07545835052761894,
      "eval_loss": 1.3748993873596191,
      "eval_runtime": 2315.5952,
      "eval_samples_per_second": 3.998,
      "eval_steps_per_second": 1.999,
      "step": 480
    }
  ],
  "logging_steps": 1,
  "max_steps": 12722,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 160,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.4587428485188813e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}