{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.37142120196027856,
  "eval_steps": 500,
  "global_step": 900,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00041269022440030954,
      "grad_norm": 5.357641696929932,
      "learning_rate": 0.0,
      "loss": 6.1189,
      "step": 1
    },
    {
      "epoch": 0.0008253804488006191,
      "grad_norm": 5.467103004455566,
      "learning_rate": 8.19672131147541e-07,
      "loss": 6.1624,
      "step": 2
    },
    {
      "epoch": 0.0016507608976012382,
      "grad_norm": 5.270254611968994,
      "learning_rate": 2.459016393442623e-06,
      "loss": 6.1301,
      "step": 4
    },
    {
      "epoch": 0.002476141346401857,
      "grad_norm": 4.018209457397461,
      "learning_rate": 4.098360655737704e-06,
      "loss": 6.1305,
      "step": 6
    },
    {
      "epoch": 0.0033015217952024763,
      "grad_norm": 3.973541259765625,
      "learning_rate": 5.737704918032787e-06,
      "loss": 6.0095,
      "step": 8
    },
    {
      "epoch": 0.0041269022440030955,
      "grad_norm": 3.5741689205169678,
      "learning_rate": 7.3770491803278695e-06,
      "loss": 5.9256,
      "step": 10
    },
    {
      "epoch": 0.004952282692803714,
      "grad_norm": 4.449726581573486,
      "learning_rate": 9.016393442622952e-06,
      "loss": 5.8059,
      "step": 12
    },
    {
      "epoch": 0.005777663141604333,
      "grad_norm": 4.23346471786499,
      "learning_rate": 1.0655737704918032e-05,
      "loss": 5.6952,
      "step": 14
    },
    {
      "epoch": 0.006603043590404953,
      "grad_norm": 2.9237682819366455,
      "learning_rate": 1.2295081967213116e-05,
      "loss": 5.5323,
      "step": 16
    },
    {
      "epoch": 0.007428424039205571,
      "grad_norm": 2.1714608669281006,
      "learning_rate": 1.3934426229508196e-05,
      "loss": 5.3692,
      "step": 18
    },
    {
      "epoch": 0.008253804488006191,
      "grad_norm": 1.9534616470336914,
      "learning_rate": 1.557377049180328e-05,
      "loss": 5.1891,
      "step": 20
    },
    {
      "epoch": 0.00907918493680681,
      "grad_norm": 1.7847157716751099,
      "learning_rate": 1.721311475409836e-05,
      "loss": 5.0308,
      "step": 22
    },
    {
      "epoch": 0.009904565385607429,
      "grad_norm": 1.329651117324829,
      "learning_rate": 1.8852459016393442e-05,
      "loss": 4.8984,
      "step": 24
    },
    {
      "epoch": 0.010729945834408047,
      "grad_norm": 1.3483397960662842,
      "learning_rate": 2.0491803278688525e-05,
      "loss": 4.7523,
      "step": 26
    },
    {
      "epoch": 0.011555326283208666,
      "grad_norm": 1.213932991027832,
      "learning_rate": 2.2131147540983607e-05,
      "loss": 4.6048,
      "step": 28
    },
    {
      "epoch": 0.012380706732009285,
      "grad_norm": 1.2783870697021484,
      "learning_rate": 2.377049180327869e-05,
      "loss": 4.5203,
      "step": 30
    },
    {
      "epoch": 0.013206087180809905,
      "grad_norm": 1.4986013174057007,
      "learning_rate": 2.540983606557377e-05,
      "loss": 4.3679,
      "step": 32
    },
    {
      "epoch": 0.014031467629610524,
      "grad_norm": 0.9324406981468201,
      "learning_rate": 2.7049180327868856e-05,
      "loss": 4.2537,
      "step": 34
    },
    {
      "epoch": 0.014856848078411143,
      "grad_norm": 2.258493185043335,
      "learning_rate": 2.8688524590163935e-05,
      "loss": 4.1994,
      "step": 36
    },
    {
      "epoch": 0.015682228527211763,
      "grad_norm": 2.1433424949645996,
      "learning_rate": 3.0327868852459017e-05,
      "loss": 4.1349,
      "step": 38
    },
    {
      "epoch": 0.016507608976012382,
      "grad_norm": 2.0129520893096924,
      "learning_rate": 3.19672131147541e-05,
      "loss": 4.0318,
      "step": 40
    },
    {
      "epoch": 0.017332989424813,
      "grad_norm": 1.2179840803146362,
      "learning_rate": 3.360655737704918e-05,
      "loss": 3.9541,
      "step": 42
    },
    {
      "epoch": 0.01815836987361362,
      "grad_norm": 1.133366346359253,
      "learning_rate": 3.524590163934427e-05,
      "loss": 3.8931,
      "step": 44
    },
    {
      "epoch": 0.01898375032241424,
      "grad_norm": 1.4476598501205444,
      "learning_rate": 3.6885245901639346e-05,
      "loss": 3.8386,
      "step": 46
    },
    {
      "epoch": 0.019809130771214857,
      "grad_norm": 1.8356342315673828,
      "learning_rate": 3.8524590163934424e-05,
      "loss": 3.7593,
      "step": 48
    },
    {
      "epoch": 0.020634511220015476,
      "grad_norm": 1.1986051797866821,
      "learning_rate": 4.016393442622951e-05,
      "loss": 3.7367,
      "step": 50
    },
    {
      "epoch": 0.021459891668816095,
      "grad_norm": 1.4108922481536865,
      "learning_rate": 4.1803278688524595e-05,
      "loss": 3.6536,
      "step": 52
    },
    {
      "epoch": 0.022285272117616713,
      "grad_norm": 1.194887638092041,
      "learning_rate": 4.3442622950819674e-05,
      "loss": 3.6377,
      "step": 54
    },
    {
      "epoch": 0.023110652566417332,
      "grad_norm": 1.5970392227172852,
      "learning_rate": 4.508196721311476e-05,
      "loss": 3.592,
      "step": 56
    },
    {
      "epoch": 0.02393603301521795,
      "grad_norm": 1.7871198654174805,
      "learning_rate": 4.672131147540984e-05,
      "loss": 3.5467,
      "step": 58
    },
    {
      "epoch": 0.02476141346401857,
      "grad_norm": 2.0405406951904297,
      "learning_rate": 4.836065573770492e-05,
      "loss": 3.5071,
      "step": 60
    },
    {
      "epoch": 0.025586793912819192,
      "grad_norm": 1.6245758533477783,
      "learning_rate": 5e-05,
      "loss": 3.4754,
      "step": 62
    },
    {
      "epoch": 0.02641217436161981,
      "grad_norm": 1.3766052722930908,
      "learning_rate": 5.163934426229509e-05,
      "loss": 3.4631,
      "step": 64
    },
    {
      "epoch": 0.02723755481042043,
      "grad_norm": 1.3058711290359497,
      "learning_rate": 5.327868852459017e-05,
      "loss": 3.4197,
      "step": 66
    },
    {
      "epoch": 0.028062935259221048,
      "grad_norm": 1.545015573501587,
      "learning_rate": 5.491803278688525e-05,
      "loss": 3.4313,
      "step": 68
    },
    {
      "epoch": 0.028888315708021667,
      "grad_norm": 1.439721703529358,
      "learning_rate": 5.6557377049180324e-05,
      "loss": 3.3894,
      "step": 70
    },
    {
      "epoch": 0.029713696156822286,
      "grad_norm": 1.6934937238693237,
      "learning_rate": 5.819672131147541e-05,
      "loss": 3.3193,
      "step": 72
    },
    {
      "epoch": 0.030539076605622904,
      "grad_norm": 1.3454101085662842,
      "learning_rate": 5.9836065573770495e-05,
      "loss": 3.3252,
      "step": 74
    },
    {
      "epoch": 0.03136445705442353,
      "grad_norm": 1.610787272453308,
      "learning_rate": 6.147540983606557e-05,
      "loss": 3.2966,
      "step": 76
    },
    {
      "epoch": 0.03218983750322414,
      "grad_norm": 2.0271148681640625,
      "learning_rate": 6.311475409836067e-05,
      "loss": 3.2892,
      "step": 78
    },
    {
      "epoch": 0.033015217952024764,
      "grad_norm": 2.165980100631714,
      "learning_rate": 6.475409836065574e-05,
      "loss": 3.2792,
      "step": 80
    },
    {
      "epoch": 0.03384059840082538,
      "grad_norm": 1.7957913875579834,
      "learning_rate": 6.639344262295082e-05,
      "loss": 3.2543,
      "step": 82
    },
    {
      "epoch": 0.034665978849626,
      "grad_norm": 1.3472362756729126,
      "learning_rate": 6.80327868852459e-05,
      "loss": 3.251,
      "step": 84
    },
    {
      "epoch": 0.03549135929842662,
      "grad_norm": 1.3264447450637817,
      "learning_rate": 6.967213114754098e-05,
      "loss": 3.2185,
      "step": 86
    },
    {
      "epoch": 0.03631673974722724,
      "grad_norm": 1.5266629457473755,
      "learning_rate": 7.131147540983607e-05,
      "loss": 3.2058,
      "step": 88
    },
    {
      "epoch": 0.037142120196027854,
      "grad_norm": 1.36456298828125,
      "learning_rate": 7.295081967213115e-05,
      "loss": 3.1753,
      "step": 90
    },
    {
      "epoch": 0.03796750064482848,
      "grad_norm": 1.3469734191894531,
      "learning_rate": 7.459016393442624e-05,
      "loss": 3.1905,
      "step": 92
    },
    {
      "epoch": 0.03879288109362909,
      "grad_norm": 1.2221981287002563,
      "learning_rate": 7.622950819672131e-05,
      "loss": 3.1526,
      "step": 94
    },
    {
      "epoch": 0.039618261542429714,
      "grad_norm": 1.3852319717407227,
      "learning_rate": 7.78688524590164e-05,
      "loss": 3.156,
      "step": 96
    },
    {
      "epoch": 0.04044364199123033,
      "grad_norm": 1.3925163745880127,
      "learning_rate": 7.950819672131148e-05,
      "loss": 3.149,
      "step": 98
    },
    {
      "epoch": 0.04126902244003095,
      "grad_norm": 1.4574236869812012,
      "learning_rate": 8.114754098360656e-05,
      "loss": 3.1252,
      "step": 100
    },
    {
      "epoch": 0.042094402888831574,
      "grad_norm": 1.1478595733642578,
      "learning_rate": 8.278688524590165e-05,
      "loss": 3.1093,
      "step": 102
    },
    {
      "epoch": 0.04291978333763219,
      "grad_norm": 1.4830694198608398,
      "learning_rate": 8.442622950819673e-05,
      "loss": 3.0921,
      "step": 104
    },
    {
      "epoch": 0.04374516378643281,
      "grad_norm": 1.4275153875350952,
      "learning_rate": 8.606557377049181e-05,
      "loss": 3.0861,
      "step": 106
    },
    {
      "epoch": 0.04457054423523343,
      "grad_norm": 2.0077579021453857,
      "learning_rate": 8.770491803278689e-05,
      "loss": 3.0694,
      "step": 108
    },
    {
      "epoch": 0.04539592468403405,
      "grad_norm": 1.533471703529358,
      "learning_rate": 8.934426229508197e-05,
      "loss": 3.0526,
      "step": 110
    },
    {
      "epoch": 0.046221305132834664,
      "grad_norm": 1.394168496131897,
      "learning_rate": 9.098360655737706e-05,
      "loss": 3.0655,
      "step": 112
    },
    {
      "epoch": 0.047046685581635286,
      "grad_norm": 1.6954408884048462,
      "learning_rate": 9.262295081967214e-05,
      "loss": 3.0338,
      "step": 114
    },
    {
      "epoch": 0.0478720660304359,
      "grad_norm": 1.4712835550308228,
      "learning_rate": 9.426229508196722e-05,
      "loss": 3.0511,
      "step": 116
    },
    {
      "epoch": 0.048697446479236524,
      "grad_norm": 1.473305344581604,
      "learning_rate": 9.59016393442623e-05,
      "loss": 3.031,
      "step": 118
    },
    {
      "epoch": 0.04952282692803714,
      "grad_norm": 1.5957138538360596,
      "learning_rate": 9.754098360655737e-05,
      "loss": 3.003,
      "step": 120
    },
    {
      "epoch": 0.05034820737683776,
      "grad_norm": 1.7283776998519897,
      "learning_rate": 9.918032786885247e-05,
      "loss": 3.0025,
      "step": 122
    },
    {
      "epoch": 0.051173587825638384,
      "grad_norm": 1.286211609840393,
      "learning_rate": 9.999995343827644e-05,
      "loss": 3.0046,
      "step": 124
    },
    {
      "epoch": 0.051998968274439,
      "grad_norm": 1.612631916999817,
      "learning_rate": 9.99995809450083e-05,
      "loss": 2.9935,
      "step": 126
    },
    {
      "epoch": 0.05282434872323962,
      "grad_norm": 1.7912741899490356,
      "learning_rate": 9.9998835961247e-05,
      "loss": 3.0016,
      "step": 128
    },
    {
      "epoch": 0.053649729172040236,
      "grad_norm": 1.7926831245422363,
      "learning_rate": 9.999771849254263e-05,
      "loss": 2.9718,
      "step": 130
    },
    {
      "epoch": 0.05447510962084086,
      "grad_norm": 1.4038861989974976,
      "learning_rate": 9.999622854722017e-05,
      "loss": 2.9792,
      "step": 132
    },
    {
      "epoch": 0.055300490069641474,
      "grad_norm": 1.7067828178405762,
      "learning_rate": 9.99943661363795e-05,
      "loss": 2.968,
      "step": 134
    },
    {
      "epoch": 0.056125870518442096,
      "grad_norm": 1.5349268913269043,
      "learning_rate": 9.999213127389536e-05,
      "loss": 2.9373,
      "step": 136
    },
    {
      "epoch": 0.05695125096724271,
      "grad_norm": 1.2763527631759644,
      "learning_rate": 9.99895239764172e-05,
      "loss": 2.9384,
      "step": 138
    },
    {
      "epoch": 0.057776631416043334,
      "grad_norm": 1.3789610862731934,
      "learning_rate": 9.998654426336905e-05,
      "loss": 2.9416,
      "step": 140
    },
    {
      "epoch": 0.05860201186484395,
      "grad_norm": 1.6175661087036133,
      "learning_rate": 9.998319215694936e-05,
      "loss": 2.9323,
      "step": 142
    },
    {
      "epoch": 0.05942739231364457,
      "grad_norm": 1.5398344993591309,
      "learning_rate": 9.997946768213095e-05,
      "loss": 2.9334,
      "step": 144
    },
    {
      "epoch": 0.060252772762445186,
      "grad_norm": 1.6788642406463623,
      "learning_rate": 9.997537086666063e-05,
      "loss": 2.9218,
      "step": 146
    },
    {
      "epoch": 0.06107815321124581,
      "grad_norm": 1.4843031167984009,
      "learning_rate": 9.997090174105919e-05,
      "loss": 2.9366,
      "step": 148
    },
    {
      "epoch": 0.06190353366004643,
      "grad_norm": 1.3358060121536255,
      "learning_rate": 9.996606033862102e-05,
      "loss": 2.9279,
      "step": 150
    },
    {
      "epoch": 0.06272891410884705,
      "grad_norm": 1.9375636577606201,
      "learning_rate": 9.996084669541397e-05,
      "loss": 2.9202,
      "step": 152
    },
    {
      "epoch": 0.06355429455764766,
      "grad_norm": 1.3849859237670898,
      "learning_rate": 9.9955260850279e-05,
      "loss": 2.895,
      "step": 154
    },
    {
      "epoch": 0.06437967500644828,
      "grad_norm": 1.3628286123275757,
      "learning_rate": 9.994930284482993e-05,
      "loss": 2.8983,
      "step": 156
    },
    {
      "epoch": 0.0652050554552489,
      "grad_norm": 1.2561815977096558,
      "learning_rate": 9.994297272345319e-05,
      "loss": 2.9089,
      "step": 158
    },
    {
      "epoch": 0.06603043590404953,
      "grad_norm": 1.3705800771713257,
      "learning_rate": 9.993627053330732e-05,
      "loss": 2.878,
      "step": 160
    },
    {
      "epoch": 0.06685581635285014,
      "grad_norm": 1.2955900430679321,
      "learning_rate": 9.99291963243228e-05,
      "loss": 2.8591,
      "step": 162
    },
    {
      "epoch": 0.06768119680165076,
      "grad_norm": 1.231101155281067,
      "learning_rate": 9.992175014920161e-05,
      "loss": 2.8616,
      "step": 164
    },
    {
      "epoch": 0.06850657725045138,
      "grad_norm": 1.1412620544433594,
      "learning_rate": 9.991393206341677e-05,
      "loss": 2.8353,
      "step": 166
    },
    {
      "epoch": 0.069331957699252,
      "grad_norm": 1.102623701095581,
      "learning_rate": 9.990574212521205e-05,
      "loss": 2.8262,
      "step": 168
    },
    {
      "epoch": 0.07015733814805261,
      "grad_norm": 1.0235016345977783,
      "learning_rate": 9.98971803956014e-05,
      "loss": 2.8581,
      "step": 170
    },
    {
      "epoch": 0.07098271859685323,
      "grad_norm": 1.475123643875122,
      "learning_rate": 9.988824693836864e-05,
      "loss": 2.8709,
      "step": 172
    },
    {
      "epoch": 0.07180809904565386,
      "grad_norm": 0.8155277371406555,
      "learning_rate": 9.98789418200669e-05,
      "loss": 2.8426,
      "step": 174
    },
    {
      "epoch": 0.07263347949445448,
      "grad_norm": 1.3113749027252197,
      "learning_rate": 9.98692651100181e-05,
      "loss": 2.8017,
      "step": 176
    },
    {
      "epoch": 0.0734588599432551,
      "grad_norm": 1.2450861930847168,
      "learning_rate": 9.985921688031252e-05,
      "loss": 2.8317,
      "step": 178
    },
    {
      "epoch": 0.07428424039205571,
      "grad_norm": 1.304402470588684,
      "learning_rate": 9.984879720580816e-05,
      "loss": 2.8157,
      "step": 180
    },
    {
      "epoch": 0.07510962084085633,
      "grad_norm": 1.1851410865783691,
      "learning_rate": 9.983800616413026e-05,
      "loss": 2.8245,
      "step": 182
    },
    {
      "epoch": 0.07593500128965695,
      "grad_norm": 1.2967396974563599,
      "learning_rate": 9.982684383567071e-05,
      "loss": 2.8363,
      "step": 184
    },
    {
      "epoch": 0.07676038173845758,
      "grad_norm": 1.2011407613754272,
      "learning_rate": 9.981531030358746e-05,
      "loss": 2.8142,
      "step": 186
    },
    {
      "epoch": 0.07758576218725818,
      "grad_norm": 1.0165106058120728,
      "learning_rate": 9.980340565380382e-05,
      "loss": 2.7913,
      "step": 188
    },
    {
      "epoch": 0.0784111426360588,
      "grad_norm": 1.3044579029083252,
      "learning_rate": 9.979112997500792e-05,
      "loss": 2.7805,
      "step": 190
    },
    {
      "epoch": 0.07923652308485943,
      "grad_norm": 1.1849685907363892,
      "learning_rate": 9.9778483358652e-05,
      "loss": 2.7707,
      "step": 192
    },
    {
      "epoch": 0.08006190353366005,
      "grad_norm": 0.9122027158737183,
      "learning_rate": 9.976546589895175e-05,
      "loss": 2.7777,
      "step": 194
    },
    {
      "epoch": 0.08088728398246066,
      "grad_norm": 1.0830117464065552,
      "learning_rate": 9.975207769288556e-05,
      "loss": 2.8048,
      "step": 196
    },
    {
      "epoch": 0.08171266443126128,
      "grad_norm": 1.1544275283813477,
      "learning_rate": 9.973831884019387e-05,
      "loss": 2.7761,
      "step": 198
    },
    {
      "epoch": 0.0825380448800619,
      "grad_norm": 0.8355935215950012,
      "learning_rate": 9.972418944337835e-05,
      "loss": 2.7593,
      "step": 200
    },
    {
      "epoch": 0.08336342532886253,
      "grad_norm": 1.203262209892273,
      "learning_rate": 9.970968960770124e-05,
      "loss": 2.7695,
      "step": 202
    },
    {
      "epoch": 0.08418880577766315,
      "grad_norm": 1.23800790309906,
      "learning_rate": 9.969481944118443e-05,
      "loss": 2.7576,
      "step": 204
    },
    {
      "epoch": 0.08501418622646376,
      "grad_norm": 0.8839966058731079,
      "learning_rate": 9.96795790546088e-05,
      "loss": 2.7442,
      "step": 206
    },
    {
      "epoch": 0.08583956667526438,
      "grad_norm": 0.9399611949920654,
      "learning_rate": 9.966396856151326e-05,
      "loss": 2.7402,
      "step": 208
    },
    {
      "epoch": 0.086664947124065,
      "grad_norm": 1.1721992492675781,
      "learning_rate": 9.964798807819397e-05,
      "loss": 2.7378,
      "step": 210
    },
    {
      "epoch": 0.08749032757286562,
      "grad_norm": 0.9647835493087769,
      "learning_rate": 9.963163772370352e-05,
      "loss": 2.7256,
      "step": 212
    },
    {
      "epoch": 0.08831570802166623,
      "grad_norm": 0.9155466556549072,
      "learning_rate": 9.961491761984996e-05,
      "loss": 2.7255,
      "step": 214
    },
    {
      "epoch": 0.08914108847046685,
      "grad_norm": 0.9373721480369568,
      "learning_rate": 9.959782789119592e-05,
      "loss": 2.7544,
      "step": 216
    },
    {
      "epoch": 0.08996646891926748,
      "grad_norm": 0.9547314643859863,
      "learning_rate": 9.958036866505772e-05,
      "loss": 2.7333,
      "step": 218
    },
    {
      "epoch": 0.0907918493680681,
      "grad_norm": 1.0028138160705566,
      "learning_rate": 9.956254007150432e-05,
      "loss": 2.7232,
      "step": 220
    },
    {
      "epoch": 0.0916172298168687,
      "grad_norm": 1.2652791738510132,
      "learning_rate": 9.954434224335649e-05,
      "loss": 2.7268,
      "step": 222
    },
    {
      "epoch": 0.09244261026566933,
      "grad_norm": 1.1313235759735107,
      "learning_rate": 9.952577531618574e-05,
      "loss": 2.7417,
      "step": 224
    },
    {
      "epoch": 0.09326799071446995,
      "grad_norm": 0.7514833211898804,
      "learning_rate": 9.950683942831328e-05,
      "loss": 2.6898,
      "step": 226
    },
    {
      "epoch": 0.09409337116327057,
      "grad_norm": 0.9731917381286621,
      "learning_rate": 9.948753472080907e-05,
      "loss": 2.686,
      "step": 228
    },
    {
      "epoch": 0.0949187516120712,
      "grad_norm": 0.8640966415405273,
      "learning_rate": 9.946786133749071e-05,
      "loss": 2.7168,
      "step": 230
    },
    {
      "epoch": 0.0957441320608718,
      "grad_norm": 0.9116567969322205,
      "learning_rate": 9.944781942492242e-05,
      "loss": 2.7123,
      "step": 232
    },
    {
      "epoch": 0.09656951250967243,
      "grad_norm": 1.0034291744232178,
      "learning_rate": 9.942740913241386e-05,
      "loss": 2.7146,
      "step": 234
    },
    {
      "epoch": 0.09739489295847305,
      "grad_norm": 0.8208848237991333,
      "learning_rate": 9.94066306120191e-05,
      "loss": 2.6773,
      "step": 236
    },
    {
      "epoch": 0.09822027340727367,
      "grad_norm": 0.8781367540359497,
      "learning_rate": 9.938548401853547e-05,
      "loss": 2.719,
      "step": 238
    },
    {
      "epoch": 0.09904565385607428,
      "grad_norm": 0.7302896976470947,
      "learning_rate": 9.93639695095024e-05,
      "loss": 2.7011,
      "step": 240
    },
    {
      "epoch": 0.0998710343048749,
      "grad_norm": 0.705086350440979,
      "learning_rate": 9.934208724520024e-05,
      "loss": 2.6648,
      "step": 242
    },
    {
      "epoch": 0.10069641475367552,
      "grad_norm": 0.8350553512573242,
      "learning_rate": 9.931983738864904e-05,
      "loss": 2.687,
      "step": 244
    },
    {
      "epoch": 0.10152179520247614,
      "grad_norm": 0.6524394154548645,
      "learning_rate": 9.92972201056074e-05,
      "loss": 2.7015,
      "step": 246
    },
    {
      "epoch": 0.10234717565127677,
      "grad_norm": 0.6503209471702576,
      "learning_rate": 9.927423556457121e-05,
      "loss": 2.6148,
      "step": 248
    },
    {
      "epoch": 0.10317255610007738,
      "grad_norm": 0.7506954073905945,
      "learning_rate": 9.925088393677236e-05,
      "loss": 2.6914,
      "step": 250
    },
    {
      "epoch": 0.103997936548878,
      "grad_norm": 1.1561987400054932,
      "learning_rate": 9.922716539617746e-05,
      "loss": 2.6659,
      "step": 252
    },
    {
      "epoch": 0.10482331699767862,
      "grad_norm": 1.0000964403152466,
      "learning_rate": 9.920308011948665e-05,
      "loss": 2.6626,
      "step": 254
    },
    {
      "epoch": 0.10564869744647924,
      "grad_norm": 0.8899397850036621,
      "learning_rate": 9.917862828613214e-05,
      "loss": 2.6666,
      "step": 256
    },
    {
      "epoch": 0.10647407789527985,
      "grad_norm": 1.1503660678863525,
      "learning_rate": 9.915381007827698e-05,
      "loss": 2.6395,
      "step": 258
    },
    {
      "epoch": 0.10729945834408047,
      "grad_norm": 0.8070819973945618,
      "learning_rate": 9.912862568081364e-05,
      "loss": 2.6531,
      "step": 260
    },
    {
      "epoch": 0.1081248387928811,
      "grad_norm": 0.8623407483100891,
      "learning_rate": 9.910307528136266e-05,
      "loss": 2.6588,
      "step": 262
    },
    {
      "epoch": 0.10895021924168172,
      "grad_norm": 0.9573660492897034,
      "learning_rate": 9.907715907027129e-05,
      "loss": 2.6823,
      "step": 264
    },
    {
      "epoch": 0.10977559969048233,
      "grad_norm": 1.0500940084457397,
      "learning_rate": 9.905087724061195e-05,
      "loss": 2.6545,
      "step": 266
    },
    {
      "epoch": 0.11060098013928295,
      "grad_norm": 1.0520515441894531,
      "learning_rate": 9.902422998818094e-05,
      "loss": 2.6371,
      "step": 268
    },
    {
      "epoch": 0.11142636058808357,
      "grad_norm": 0.9879215955734253,
      "learning_rate": 9.899721751149688e-05,
      "loss": 2.6474,
      "step": 270
    },
    {
      "epoch": 0.11225174103688419,
      "grad_norm": 0.8972532749176025,
      "learning_rate": 9.896984001179925e-05,
      "loss": 2.6271,
      "step": 272
    },
    {
      "epoch": 0.11307712148568481,
      "grad_norm": 0.6369883418083191,
      "learning_rate": 9.894209769304696e-05,
      "loss": 2.6054,
      "step": 274
    },
    {
      "epoch": 0.11390250193448542,
      "grad_norm": 0.6478956937789917,
      "learning_rate": 9.891399076191674e-05,
      "loss": 2.6168,
      "step": 276
    },
    {
      "epoch": 0.11472788238328605,
      "grad_norm": 0.8620642423629761,
      "learning_rate": 9.888551942780162e-05,
      "loss": 2.6313,
      "step": 278
    },
    {
      "epoch": 0.11555326283208667,
      "grad_norm": 0.740717887878418,
      "learning_rate": 9.885668390280941e-05,
      "loss": 2.6307,
      "step": 280
    },
    {
      "epoch": 0.11637864328088729,
      "grad_norm": 0.7513862252235413,
      "learning_rate": 9.882748440176109e-05,
      "loss": 2.625,
      "step": 282
    },
    {
      "epoch": 0.1172040237296879,
      "grad_norm": 0.8409993052482605,
      "learning_rate": 9.879792114218921e-05,
      "loss": 2.6034,
      "step": 284
    },
    {
      "epoch": 0.11802940417848852,
      "grad_norm": 0.8200739622116089,
      "learning_rate": 9.876799434433628e-05,
      "loss": 2.599,
      "step": 286
    },
    {
      "epoch": 0.11885478462728914,
      "grad_norm": 0.9191763401031494,
      "learning_rate": 9.873770423115314e-05,
      "loss": 2.6168,
      "step": 288
    },
    {
      "epoch": 0.11968016507608976,
      "grad_norm": 0.7739763855934143,
      "learning_rate": 9.870705102829723e-05,
      "loss": 2.6279,
      "step": 290
    },
    {
      "epoch": 0.12050554552489037,
      "grad_norm": 0.6580247282981873,
      "learning_rate": 9.867603496413103e-05,
      "loss": 2.599,
      "step": 292
    },
    {
      "epoch": 0.121330925973691,
      "grad_norm": 0.7197789549827576,
      "learning_rate": 9.864465626972023e-05,
      "loss": 2.5948,
      "step": 294
    },
    {
      "epoch": 0.12215630642249162,
      "grad_norm": 0.9027787446975708,
      "learning_rate": 9.861291517883213e-05,
      "loss": 2.6058,
      "step": 296
    },
    {
      "epoch": 0.12298168687129224,
      "grad_norm": 1.048640489578247,
      "learning_rate": 9.858081192793378e-05,
      "loss": 2.6128,
      "step": 298
    },
    {
      "epoch": 0.12380706732009286,
      "grad_norm": 0.827551543712616,
      "learning_rate": 9.85483467561903e-05,
      "loss": 2.6058,
      "step": 300
    },
    {
      "epoch": 0.12463244776889347,
      "grad_norm": 0.9751214981079102,
      "learning_rate": 9.851551990546306e-05,
      "loss": 2.57,
      "step": 302
    },
    {
      "epoch": 0.1254578282176941,
      "grad_norm": 1.0783475637435913,
      "learning_rate": 9.848233162030794e-05,
      "loss": 2.6116,
      "step": 304
    },
    {
      "epoch": 0.1262832086664947,
      "grad_norm": 0.8441977500915527,
      "learning_rate": 9.84487821479734e-05,
      "loss": 2.59,
      "step": 306
    },
    {
      "epoch": 0.12710858911529532,
      "grad_norm": 1.0184507369995117,
      "learning_rate": 9.841487173839873e-05,
      "loss": 2.579,
      "step": 308
    },
    {
      "epoch": 0.12793396956409595,
      "grad_norm": 0.9782202243804932,
      "learning_rate": 9.838060064421217e-05,
      "loss": 2.5996,
      "step": 310
    },
    {
      "epoch": 0.12875935001289657,
      "grad_norm": 0.8052064180374146,
      "learning_rate": 9.834596912072897e-05,
      "loss": 2.594,
      "step": 312
    },
    {
      "epoch": 0.1295847304616972,
      "grad_norm": 0.765544056892395,
      "learning_rate": 9.831097742594958e-05,
      "loss": 2.581,
      "step": 314
    },
    {
      "epoch": 0.1304101109104978,
      "grad_norm": 0.8481123447418213,
      "learning_rate": 9.827562582055765e-05,
      "loss": 2.6006,
      "step": 316
    },
    {
      "epoch": 0.13123549135929843,
      "grad_norm": 0.8604638576507568,
      "learning_rate": 9.823991456791811e-05,
      "loss": 2.5875,
      "step": 318
    },
    {
      "epoch": 0.13206087180809906,
      "grad_norm": 0.6848525404930115,
      "learning_rate": 9.820384393407525e-05,
      "loss": 2.5788,
      "step": 320
    },
    {
      "epoch": 0.13288625225689968,
      "grad_norm": 0.7860177159309387,
      "learning_rate": 9.816741418775066e-05,
      "loss": 2.5961,
      "step": 322
    },
    {
      "epoch": 0.13371163270570027,
      "grad_norm": 0.7415710091590881,
      "learning_rate": 9.813062560034134e-05,
      "loss": 2.5805,
      "step": 324
    },
    {
      "epoch": 0.1345370131545009,
      "grad_norm": 0.8323041200637817,
      "learning_rate": 9.809347844591753e-05,
      "loss": 2.5799,
      "step": 326
    },
    {
      "epoch": 0.13536239360330152,
      "grad_norm": 0.9540163278579712,
      "learning_rate": 9.805597300122081e-05,
      "loss": 2.5704,
      "step": 328
    },
    {
      "epoch": 0.13618777405210214,
      "grad_norm": 0.7344382405281067,
      "learning_rate": 9.801810954566195e-05,
      "loss": 2.5649,
      "step": 330
    },
    {
      "epoch": 0.13701315450090276,
      "grad_norm": 0.7706190347671509,
      "learning_rate": 9.797988836131884e-05,
      "loss": 2.5583,
      "step": 332
    },
    {
      "epoch": 0.13783853494970338,
      "grad_norm": 0.7073199152946472,
      "learning_rate": 9.794130973293445e-05,
      "loss": 2.5523,
      "step": 334
    },
    {
      "epoch": 0.138663915398504,
      "grad_norm": 0.6726153492927551,
      "learning_rate": 9.790237394791461e-05,
      "loss": 2.5673,
      "step": 336
    },
    {
      "epoch": 0.13948929584730463,
      "grad_norm": 0.6806092262268066,
      "learning_rate": 9.786308129632598e-05,
      "loss": 2.5748,
      "step": 338
    },
    {
      "epoch": 0.14031467629610522,
      "grad_norm": 0.7212201952934265,
      "learning_rate": 9.782343207089377e-05,
      "loss": 2.5615,
      "step": 340
    },
    {
      "epoch": 0.14114005674490585,
      "grad_norm": 0.7233206629753113,
      "learning_rate": 9.778342656699964e-05,
      "loss": 2.5509,
      "step": 342
    },
    {
      "epoch": 0.14196543719370647,
      "grad_norm": 0.6308603882789612,
      "learning_rate": 9.77430650826795e-05,
      "loss": 2.5133,
      "step": 344
    },
    {
      "epoch": 0.1427908176425071,
      "grad_norm": 0.676368772983551,
      "learning_rate": 9.770234791862125e-05,
      "loss": 2.5293,
      "step": 346
    },
    {
      "epoch": 0.1436161980913077,
      "grad_norm": 0.687326192855835,
      "learning_rate": 9.766127537816256e-05,
      "loss": 2.531,
      "step": 348
    },
    {
      "epoch": 0.14444157854010833,
      "grad_norm": 0.7347912788391113,
      "learning_rate": 9.761984776728864e-05,
      "loss": 2.5468,
      "step": 350
    },
    {
      "epoch": 0.14526695898890896,
      "grad_norm": 0.5843812227249146,
      "learning_rate": 9.757806539462985e-05,
      "loss": 2.539,
      "step": 352
    },
    {
      "epoch": 0.14609233943770958,
      "grad_norm": 0.7662450671195984,
      "learning_rate": 9.753592857145957e-05,
      "loss": 2.5297,
      "step": 354
    },
    {
      "epoch": 0.1469177198865102,
      "grad_norm": 0.6813721656799316,
      "learning_rate": 9.749343761169171e-05,
      "loss": 2.5519,
      "step": 356
    },
    {
      "epoch": 0.1477431003353108,
      "grad_norm": 0.7090803384780884,
      "learning_rate": 9.745059283187857e-05,
      "loss": 2.515,
      "step": 358
    },
    {
      "epoch": 0.14856848078411142,
      "grad_norm": 0.5888863801956177,
      "learning_rate": 9.74073945512082e-05,
      "loss": 2.5622,
      "step": 360
    },
    {
      "epoch": 0.14939386123291204,
      "grad_norm": 0.7369230389595032,
      "learning_rate": 9.736384309150233e-05,
      "loss": 2.5482,
      "step": 362
    },
    {
      "epoch": 0.15021924168171266,
      "grad_norm": 0.6921555995941162,
      "learning_rate": 9.731993877721377e-05,
      "loss": 2.5304,
      "step": 364
    },
    {
      "epoch": 0.15104462213051328,
      "grad_norm": 0.755409300327301,
      "learning_rate": 9.727568193542403e-05,
      "loss": 2.5214,
      "step": 366
    },
    {
      "epoch": 0.1518700025793139,
      "grad_norm": 0.9340344071388245,
      "learning_rate": 9.723107289584095e-05,
      "loss": 2.5248,
      "step": 368
    },
    {
      "epoch": 0.15269538302811453,
      "grad_norm": 0.9866952300071716,
      "learning_rate": 9.718611199079617e-05,
      "loss": 2.5281,
      "step": 370
    },
    {
      "epoch": 0.15352076347691515,
      "grad_norm": 0.8538560271263123,
      "learning_rate": 9.714079955524269e-05,
      "loss": 2.5436,
      "step": 372
    },
    {
      "epoch": 0.15434614392571577,
      "grad_norm": 0.6923696398735046,
      "learning_rate": 9.709513592675236e-05,
      "loss": 2.5431,
      "step": 374
    },
    {
      "epoch": 0.15517152437451637,
      "grad_norm": 0.6311334371566772,
      "learning_rate": 9.704912144551341e-05,
      "loss": 2.5473,
      "step": 376
    },
    {
      "epoch": 0.155996904823317,
      "grad_norm": 0.7558380961418152,
      "learning_rate": 9.700275645432784e-05,
      "loss": 2.4998,
      "step": 378
    },
    {
      "epoch": 0.1568222852721176,
      "grad_norm": 0.8375957608222961,
      "learning_rate": 9.695604129860889e-05,
      "loss": 2.524,
      "step": 380
    },
    {
      "epoch": 0.15764766572091823,
      "grad_norm": 0.9554900527000427,
      "learning_rate": 9.690897632637852e-05,
      "loss": 2.5184,
      "step": 382
    },
    {
      "epoch": 0.15847304616971886,
      "grad_norm": 0.7582331299781799,
      "learning_rate": 9.686156188826478e-05,
      "loss": 2.5177,
      "step": 384
    },
    {
      "epoch": 0.15929842661851948,
      "grad_norm": 0.8506788611412048,
      "learning_rate": 9.681379833749915e-05,
      "loss": 2.4866,
      "step": 386
    },
    {
      "epoch": 0.1601238070673201,
      "grad_norm": 0.738979697227478,
      "learning_rate": 9.676568602991399e-05,
      "loss": 2.5286,
      "step": 388
    },
    {
      "epoch": 0.16094918751612072,
      "grad_norm": 0.6935485005378723,
      "learning_rate": 9.671722532393985e-05,
      "loss": 2.532,
      "step": 390
    },
    {
      "epoch": 0.16177456796492132,
      "grad_norm": 0.7109572887420654,
      "learning_rate": 9.666841658060282e-05,
      "loss": 2.5126,
      "step": 392
    },
    {
      "epoch": 0.16259994841372194,
      "grad_norm": 0.8224465847015381,
      "learning_rate": 9.661926016352178e-05,
      "loss": 2.5062,
      "step": 394
    },
    {
      "epoch": 0.16342532886252256,
      "grad_norm": 0.7705041766166687,
      "learning_rate": 9.656975643890578e-05,
      "loss": 2.5173,
      "step": 396
    },
    {
      "epoch": 0.16425070931132318,
      "grad_norm": 0.6699206829071045,
      "learning_rate": 9.651990577555122e-05,
      "loss": 2.5047,
      "step": 398
    },
    {
      "epoch": 0.1650760897601238,
      "grad_norm": 0.6773229241371155,
      "learning_rate": 9.64697085448392e-05,
      "loss": 2.5437,
      "step": 400
    },
    {
      "epoch": 0.16590147020892443,
      "grad_norm": 0.6482560634613037,
      "learning_rate": 9.641916512073268e-05,
      "loss": 2.4866,
      "step": 402
    },
    {
      "epoch": 0.16672685065772505,
      "grad_norm": 0.5687413811683655,
      "learning_rate": 9.636827587977368e-05,
      "loss": 2.5004,
      "step": 404
    },
    {
      "epoch": 0.16755223110652567,
      "grad_norm": 0.591502845287323,
      "learning_rate": 9.63170412010806e-05,
      "loss": 2.4925,
      "step": 406
    },
    {
      "epoch": 0.1683776115553263,
      "grad_norm": 0.7202515602111816,
      "learning_rate": 9.626546146634523e-05,
      "loss": 2.4834,
      "step": 408
    },
    {
      "epoch": 0.1692029920041269,
      "grad_norm": 0.5923997163772583,
      "learning_rate": 9.621353705982998e-05,
      "loss": 2.4832,
      "step": 410
    },
    {
      "epoch": 0.1700283724529275,
      "grad_norm": 0.519095242023468,
      "learning_rate": 9.616126836836508e-05,
      "loss": 2.4909,
      "step": 412
    },
    {
      "epoch": 0.17085375290172813,
      "grad_norm": 0.6338533759117126,
      "learning_rate": 9.61086557813456e-05,
      "loss": 2.5027,
      "step": 414
    },
    {
      "epoch": 0.17167913335052876,
      "grad_norm": 0.5483947992324829,
      "learning_rate": 9.60556996907286e-05,
      "loss": 2.4864,
      "step": 416
    },
    {
      "epoch": 0.17250451379932938,
      "grad_norm": 0.6210249662399292,
      "learning_rate": 9.600240049103017e-05,
      "loss": 2.4987,
      "step": 418
    },
    {
      "epoch": 0.17332989424813,
      "grad_norm": 0.6927972435951233,
      "learning_rate": 9.594875857932258e-05,
      "loss": 2.4987,
      "step": 420
    },
    {
      "epoch": 0.17415527469693062,
      "grad_norm": 0.6685944199562073,
      "learning_rate": 9.589477435523118e-05,
      "loss": 2.4794,
      "step": 422
    },
    {
      "epoch": 0.17498065514573125,
      "grad_norm": 0.89150470495224,
      "learning_rate": 9.584044822093157e-05,
      "loss": 2.467,
      "step": 424
    },
    {
      "epoch": 0.17580603559453187,
      "grad_norm": 0.8726872205734253,
      "learning_rate": 9.57857805811465e-05,
      "loss": 2.4917,
      "step": 426
    },
    {
      "epoch": 0.17663141604333246,
      "grad_norm": 0.8355669379234314,
      "learning_rate": 9.573077184314294e-05,
      "loss": 2.5095,
      "step": 428
    },
    {
      "epoch": 0.17745679649213308,
      "grad_norm": 0.7523061037063599,
      "learning_rate": 9.567542241672891e-05,
      "loss": 2.4695,
      "step": 430
    },
    {
      "epoch": 0.1782821769409337,
      "grad_norm": 0.5890743732452393,
      "learning_rate": 9.561973271425061e-05,
      "loss": 2.4748,
      "step": 432
    },
    {
      "epoch": 0.17910755738973433,
      "grad_norm": 0.6020349860191345,
      "learning_rate": 9.55637031505892e-05,
      "loss": 2.4746,
      "step": 434
    },
    {
      "epoch": 0.17993293783853495,
      "grad_norm": 0.6291228532791138,
      "learning_rate": 9.550733414315776e-05,
      "loss": 2.4537,
      "step": 436
    },
    {
      "epoch": 0.18075831828733557,
      "grad_norm": 0.6267942786216736,
      "learning_rate": 9.545062611189821e-05,
      "loss": 2.4663,
      "step": 438
    },
    {
      "epoch": 0.1815836987361362,
      "grad_norm": 0.7870015501976013,
      "learning_rate": 9.539357947927815e-05,
      "loss": 2.4794,
      "step": 440
    },
    {
      "epoch": 0.18240907918493682,
      "grad_norm": 0.818481981754303,
      "learning_rate": 9.53361946702877e-05,
      "loss": 2.4538,
      "step": 442
    },
    {
      "epoch": 0.1832344596337374,
      "grad_norm": 0.6101433634757996,
      "learning_rate": 9.527847211243635e-05,
      "loss": 2.5041,
      "step": 444
    },
    {
      "epoch": 0.18405984008253803,
      "grad_norm": 0.7772427201271057,
      "learning_rate": 9.52204122357498e-05,
      "loss": 2.4497,
      "step": 446
    },
    {
      "epoch": 0.18488522053133866,
      "grad_norm": 0.6459339261054993,
      "learning_rate": 9.516201547276668e-05,
      "loss": 2.4636,
      "step": 448
    },
    {
      "epoch": 0.18571060098013928,
      "grad_norm": 0.6417956948280334,
      "learning_rate": 9.510328225853549e-05,
      "loss": 2.4489,
      "step": 450
    },
    {
      "epoch": 0.1865359814289399,
      "grad_norm": 0.5950794219970703,
      "learning_rate": 9.50442130306111e-05,
      "loss": 2.4418,
      "step": 452
    },
    {
      "epoch": 0.18736136187774052,
      "grad_norm": 0.4874417185783386,
      "learning_rate": 9.498480822905176e-05,
      "loss": 2.4458,
      "step": 454
    },
    {
      "epoch": 0.18818674232654115,
      "grad_norm": 0.6025642156600952,
      "learning_rate": 9.492506829641566e-05,
      "loss": 2.4582,
      "step": 456
    },
    {
      "epoch": 0.18901212277534177,
      "grad_norm": 0.5527782440185547,
      "learning_rate": 9.486499367775764e-05,
      "loss": 2.4732,
      "step": 458
    },
    {
      "epoch": 0.1898375032241424,
      "grad_norm": 0.5585253238677979,
      "learning_rate": 9.480458482062594e-05,
      "loss": 2.4494,
      "step": 460
    },
    {
      "epoch": 0.19066288367294298,
      "grad_norm": 0.7675066590309143,
      "learning_rate": 9.474384217505883e-05,
      "loss": 2.4404,
      "step": 462
    },
    {
      "epoch": 0.1914882641217436,
      "grad_norm": 0.6983161568641663,
      "learning_rate": 9.468276619358129e-05,
      "loss": 2.4409,
      "step": 464
    },
    {
      "epoch": 0.19231364457054423,
      "grad_norm": 0.6638076305389404,
      "learning_rate": 9.462135733120156e-05,
      "loss": 2.4746,
      "step": 466
    },
    {
      "epoch": 0.19313902501934485,
      "grad_norm": 0.7150386571884155,
      "learning_rate": 9.455961604540784e-05,
      "loss": 2.4841,
      "step": 468
    },
    {
      "epoch": 0.19396440546814547,
      "grad_norm": 0.6076739430427551,
      "learning_rate": 9.449754279616481e-05,
      "loss": 2.4154,
      "step": 470
    },
    {
      "epoch": 0.1947897859169461,
      "grad_norm": 0.6565660238265991,
      "learning_rate": 9.443513804591026e-05,
      "loss": 2.4033,
      "step": 472
    },
    {
      "epoch": 0.19561516636574672,
      "grad_norm": 0.5531965494155884,
      "learning_rate": 9.43724022595516e-05,
      "loss": 2.4344,
      "step": 474
    },
    {
      "epoch": 0.19644054681454734,
      "grad_norm": 0.6353370547294617,
      "learning_rate": 9.430933590446244e-05,
      "loss": 2.4631,
      "step": 476
    },
    {
      "epoch": 0.19726592726334796,
      "grad_norm": 0.6833090782165527,
      "learning_rate": 9.424593945047906e-05,
      "loss": 2.4402,
      "step": 478
    },
    {
      "epoch": 0.19809130771214856,
      "grad_norm": 0.5862318277359009,
      "learning_rate": 9.418221336989695e-05,
      "loss": 2.4308,
      "step": 480
    },
    {
      "epoch": 0.19891668816094918,
      "grad_norm": 0.6512525081634521,
      "learning_rate": 9.411815813746726e-05,
      "loss": 2.4239,
      "step": 482
    },
    {
      "epoch": 0.1997420686097498,
      "grad_norm": 0.6480604410171509,
      "learning_rate": 9.405377423039331e-05,
      "loss": 2.4211,
      "step": 484
    },
    {
      "epoch": 0.20056744905855042,
      "grad_norm": 0.5605840682983398,
      "learning_rate": 9.398906212832699e-05,
      "loss": 2.4346,
      "step": 486
    },
    {
      "epoch": 0.20139282950735105,
      "grad_norm": 0.6704816818237305,
      "learning_rate": 9.392402231336518e-05,
      "loss": 2.4309,
      "step": 488
    },
    {
      "epoch": 0.20221820995615167,
      "grad_norm": 1.019185185432434,
      "learning_rate": 9.38586552700462e-05,
      "loss": 2.4341,
      "step": 490
    },
    {
      "epoch": 0.2030435904049523,
      "grad_norm": 0.816562294960022,
      "learning_rate": 9.379296148534619e-05,
      "loss": 2.4389,
      "step": 492
    },
    {
      "epoch": 0.2038689708537529,
      "grad_norm": 0.8160488605499268,
      "learning_rate": 9.372694144867544e-05,
      "loss": 2.4259,
      "step": 494
    },
    {
      "epoch": 0.20469435130255353,
      "grad_norm": 0.6071799993515015,
      "learning_rate": 9.36605956518748e-05,
      "loss": 2.4479,
      "step": 496
    },
    {
      "epoch": 0.20551973175135413,
      "grad_norm": 0.5928800106048584,
      "learning_rate": 9.359392458921198e-05,
      "loss": 2.4467,
      "step": 498
    },
    {
      "epoch": 0.20634511220015475,
      "grad_norm": 0.6091005802154541,
      "learning_rate": 9.352692875737787e-05,
      "loss": 2.4204,
      "step": 500
    },
    {
      "epoch": 0.20717049264895537,
      "grad_norm": 0.6261982917785645,
      "learning_rate": 9.34596086554829e-05,
      "loss": 2.3813,
      "step": 502
    },
    {
      "epoch": 0.207995873097756,
      "grad_norm": 0.597626805305481,
      "learning_rate": 9.339196478505321e-05,
      "loss": 2.4288,
      "step": 504
    },
    {
      "epoch": 0.20882125354655662,
      "grad_norm": 0.6542629599571228,
      "learning_rate": 9.332399765002698e-05,
      "loss": 2.4192,
      "step": 506
    },
    {
      "epoch": 0.20964663399535724,
      "grad_norm": 0.5791048407554626,
      "learning_rate": 9.32557077567507e-05,
      "loss": 2.3969,
      "step": 508
    },
    {
      "epoch": 0.21047201444415786,
      "grad_norm": 0.7887916564941406,
      "learning_rate": 9.318709561397537e-05,
      "loss": 2.4048,
      "step": 510
    },
    {
      "epoch": 0.21129739489295848,
      "grad_norm": 0.5982603430747986,
      "learning_rate": 9.311816173285268e-05,
      "loss": 2.4113,
      "step": 512
    },
    {
      "epoch": 0.21212277534175908,
      "grad_norm": 0.6093983054161072,
      "learning_rate": 9.304890662693123e-05,
      "loss": 2.4111,
      "step": 514
    },
    {
      "epoch": 0.2129481557905597,
      "grad_norm": 0.6429126858711243,
      "learning_rate": 9.297933081215273e-05,
      "loss": 2.4279,
      "step": 516
    },
    {
      "epoch": 0.21377353623936032,
      "grad_norm": 0.5441975593566895,
      "learning_rate": 9.29094348068481e-05,
      "loss": 2.4287,
      "step": 518
    },
    {
      "epoch": 0.21459891668816095,
      "grad_norm": 0.635387659072876,
      "learning_rate": 9.283921913173368e-05,
      "loss": 2.43,
      "step": 520
    },
    {
      "epoch": 0.21542429713696157,
      "grad_norm": 0.5462561845779419,
      "learning_rate": 9.276868430990726e-05,
      "loss": 2.4366,
      "step": 522
    },
    {
      "epoch": 0.2162496775857622,
      "grad_norm": 0.53558748960495,
      "learning_rate": 9.269783086684428e-05,
      "loss": 2.4123,
      "step": 524
    },
    {
      "epoch": 0.2170750580345628,
      "grad_norm": 0.6640864014625549,
      "learning_rate": 9.262665933039381e-05,
      "loss": 2.4034,
      "step": 526
    },
    {
      "epoch": 0.21790043848336343,
      "grad_norm": 0.6774006485939026,
      "learning_rate": 9.255517023077472e-05,
      "loss": 2.4487,
      "step": 528
    },
    {
      "epoch": 0.21872581893216406,
      "grad_norm": 0.7681392431259155,
      "learning_rate": 9.248336410057168e-05,
      "loss": 2.4377,
      "step": 530
    },
    {
      "epoch": 0.21955119938096465,
      "grad_norm": 0.6729565262794495,
      "learning_rate": 9.24112414747312e-05,
      "loss": 2.4305,
      "step": 532
    },
    {
      "epoch": 0.22037657982976527,
      "grad_norm": 0.5404065847396851,
      "learning_rate": 9.233880289055761e-05,
      "loss": 2.3878,
      "step": 534
    },
    {
      "epoch": 0.2212019602785659,
      "grad_norm": 0.5672969818115234,
      "learning_rate": 9.226604888770911e-05,
      "loss": 2.4138,
      "step": 536
    },
    {
      "epoch": 0.22202734072736652,
      "grad_norm": 0.5919613838195801,
      "learning_rate": 9.219298000819376e-05,
      "loss": 2.4219,
      "step": 538
    },
    {
      "epoch": 0.22285272117616714,
      "grad_norm": 0.6129051446914673,
      "learning_rate": 9.211959679636535e-05,
      "loss": 2.3922,
      "step": 540
    },
    {
      "epoch": 0.22367810162496776,
      "grad_norm": 0.5665661096572876,
      "learning_rate": 9.204589979891946e-05,
      "loss": 2.4229,
      "step": 542
    },
    {
      "epoch": 0.22450348207376838,
      "grad_norm": 0.6748535633087158,
      "learning_rate": 9.197188956488931e-05,
      "loss": 2.3801,
      "step": 544
    },
    {
      "epoch": 0.225328862522569,
      "grad_norm": 0.666181743144989,
      "learning_rate": 9.189756664564167e-05,
      "loss": 2.3888,
      "step": 546
    },
    {
      "epoch": 0.22615424297136963,
      "grad_norm": 0.6297335028648376,
      "learning_rate": 9.182293159487281e-05,
      "loss": 2.3812,
      "step": 548
    },
    {
      "epoch": 0.22697962342017022,
      "grad_norm": 0.6825816631317139,
      "learning_rate": 9.174798496860433e-05,
      "loss": 2.3837,
      "step": 550
    },
    {
      "epoch": 0.22780500386897085,
      "grad_norm": 0.7759227156639099,
      "learning_rate": 9.167272732517903e-05,
      "loss": 2.3769,
      "step": 552
    },
    {
      "epoch": 0.22863038431777147,
      "grad_norm": 0.6477057933807373,
      "learning_rate": 9.159715922525673e-05,
      "loss": 2.3852,
      "step": 554
    },
    {
      "epoch": 0.2294557647665721,
      "grad_norm": 0.8262442946434021,
      "learning_rate": 9.152128123181013e-05,
      "loss": 2.3868,
      "step": 556
    },
    {
      "epoch": 0.2302811452153727,
      "grad_norm": 0.6992378830909729,
      "learning_rate": 9.14450939101206e-05,
      "loss": 2.3896,
      "step": 558
    },
    {
      "epoch": 0.23110652566417333,
      "grad_norm": 0.6080542206764221,
      "learning_rate": 9.136859782777394e-05,
      "loss": 2.3806,
      "step": 560
    },
    {
      "epoch": 0.23193190611297396,
      "grad_norm": 0.7257338166236877,
      "learning_rate": 9.129179355465621e-05,
      "loss": 2.4114,
      "step": 562
    },
    {
      "epoch": 0.23275728656177458,
      "grad_norm": 0.6741151213645935,
      "learning_rate": 9.121468166294945e-05,
      "loss": 2.3753,
      "step": 564
    },
    {
      "epoch": 0.23358266701057517,
      "grad_norm": 0.6519246101379395,
      "learning_rate": 9.113726272712734e-05,
      "loss": 2.3937,
      "step": 566
    },
    {
      "epoch": 0.2344080474593758,
      "grad_norm": 0.7125470042228699,
      "learning_rate": 9.105953732395116e-05,
      "loss": 2.4047,
      "step": 568
    },
    {
      "epoch": 0.23523342790817642,
      "grad_norm": 0.5559272170066833,
      "learning_rate": 9.098150603246517e-05,
      "loss": 2.3927,
      "step": 570
    },
    {
      "epoch": 0.23605880835697704,
      "grad_norm": 0.5721775889396667,
      "learning_rate": 9.090316943399255e-05,
      "loss": 2.3992,
      "step": 572
    },
    {
      "epoch": 0.23688418880577766,
      "grad_norm": 0.5262630581855774,
      "learning_rate": 9.082452811213095e-05,
      "loss": 2.3898,
      "step": 574
    },
    {
      "epoch": 0.23770956925457828,
      "grad_norm": 0.5832549333572388,
      "learning_rate": 9.074558265274819e-05,
      "loss": 2.3943,
      "step": 576
    },
    {
      "epoch": 0.2385349497033789,
      "grad_norm": 0.5760109424591064,
      "learning_rate": 9.066633364397786e-05,
      "loss": 2.4223,
      "step": 578
    },
    {
      "epoch": 0.23936033015217953,
      "grad_norm": 0.5974195003509521,
      "learning_rate": 9.058678167621493e-05,
      "loss": 2.402,
      "step": 580
    },
    {
      "epoch": 0.24018571060098015,
      "grad_norm": 0.5613354444503784,
      "learning_rate": 9.050692734211142e-05,
      "loss": 2.3602,
      "step": 582
    },
    {
      "epoch": 0.24101109104978075,
      "grad_norm": 0.6307066082954407,
      "learning_rate": 9.042677123657191e-05,
      "loss": 2.4034,
      "step": 584
    },
    {
      "epoch": 0.24183647149858137,
      "grad_norm": 0.7120583057403564,
      "learning_rate": 9.034631395674917e-05,
      "loss": 2.3909,
      "step": 586
    },
    {
      "epoch": 0.242661851947382,
      "grad_norm": 0.7483564615249634,
      "learning_rate": 9.026555610203964e-05,
      "loss": 2.4026,
      "step": 588
    },
    {
      "epoch": 0.2434872323961826,
      "grad_norm": 0.5825770497322083,
      "learning_rate": 9.018449827407905e-05,
      "loss": 2.3736,
      "step": 590
    },
    {
      "epoch": 0.24431261284498323,
      "grad_norm": 0.5669400095939636,
      "learning_rate": 9.010314107673783e-05,
      "loss": 2.376,
      "step": 592
    },
    {
      "epoch": 0.24513799329378386,
      "grad_norm": 0.6108945608139038,
      "learning_rate": 9.002148511611675e-05,
      "loss": 2.3765,
      "step": 594
    },
    {
      "epoch": 0.24596337374258448,
      "grad_norm": 0.499024361371994,
      "learning_rate": 8.993953100054224e-05,
      "loss": 2.4033,
      "step": 596
    },
    {
      "epoch": 0.2467887541913851,
      "grad_norm": 0.4951154291629791,
      "learning_rate": 8.985727934056207e-05,
      "loss": 2.3808,
      "step": 598
    },
    {
      "epoch": 0.24761413464018572,
      "grad_norm": 0.4967551529407501,
      "learning_rate": 8.977473074894052e-05,
      "loss": 2.3749,
      "step": 600
    },
    {
      "epoch": 0.24843951508898632,
      "grad_norm": 0.6024683117866516,
      "learning_rate": 8.969188584065412e-05,
      "loss": 2.3745,
      "step": 602
    },
    {
      "epoch": 0.24926489553778694,
      "grad_norm": 0.5306781530380249,
      "learning_rate": 8.960874523288683e-05,
      "loss": 2.3824,
      "step": 604
    },
    {
      "epoch": 0.25009027598658756,
      "grad_norm": 0.5470788478851318,
      "learning_rate": 8.952530954502557e-05,
      "loss": 2.3828,
      "step": 606
    },
    {
      "epoch": 0.2509156564353882,
      "grad_norm": 0.5227565765380859,
      "learning_rate": 8.944157939865556e-05,
      "loss": 2.3512,
      "step": 608
    },
    {
      "epoch": 0.2517410368841888,
      "grad_norm": 0.5495042204856873,
      "learning_rate": 8.935755541755569e-05,
      "loss": 2.3553,
      "step": 610
    },
    {
      "epoch": 0.2525664173329894,
      "grad_norm": 0.554063081741333,
      "learning_rate": 8.927323822769386e-05,
      "loss": 2.3492,
      "step": 612
    },
    {
      "epoch": 0.25339179778179005,
      "grad_norm": 0.596449077129364,
      "learning_rate": 8.918862845722243e-05,
      "loss": 2.3708,
      "step": 614
    },
    {
      "epoch": 0.25421717823059065,
      "grad_norm": 0.5647444128990173,
      "learning_rate": 8.910372673647336e-05,
      "loss": 2.3716,
      "step": 616
    },
    {
      "epoch": 0.2550425586793913,
      "grad_norm": 0.6036911606788635,
      "learning_rate": 8.901853369795361e-05,
      "loss": 2.3728,
      "step": 618
    },
    {
      "epoch": 0.2558679391281919,
      "grad_norm": 0.6191892027854919,
      "learning_rate": 8.893304997634045e-05,
      "loss": 2.36,
      "step": 620
    },
    {
      "epoch": 0.25669331957699254,
      "grad_norm": 0.5799134373664856,
      "learning_rate": 8.884727620847669e-05,
      "loss": 2.3686,
      "step": 622
    },
    {
      "epoch": 0.25751870002579313,
      "grad_norm": 0.6006870865821838,
      "learning_rate": 8.876121303336596e-05,
      "loss": 2.3705,
      "step": 624
    },
    {
      "epoch": 0.2583440804745938,
      "grad_norm": 0.6628080606460571,
      "learning_rate": 8.867486109216789e-05,
      "loss": 2.3723,
      "step": 626
    },
    {
      "epoch": 0.2591694609233944,
      "grad_norm": 0.6496407389640808,
      "learning_rate": 8.858822102819347e-05,
      "loss": 2.3438,
      "step": 628
    },
    {
      "epoch": 0.259994841372195,
      "grad_norm": 0.5622187256813049,
      "learning_rate": 8.850129348690004e-05,
      "loss": 2.3729,
      "step": 630
    },
    {
      "epoch": 0.2608202218209956,
      "grad_norm": 0.6995213031768799,
      "learning_rate": 8.841407911588675e-05,
      "loss": 2.3889,
      "step": 632
    },
    {
      "epoch": 0.2616456022697962,
      "grad_norm": 0.6655398607254028,
      "learning_rate": 8.832657856488949e-05,
      "loss": 2.3803,
      "step": 634
    },
    {
      "epoch": 0.26247098271859687,
      "grad_norm": 0.7044237852096558,
      "learning_rate": 8.82387924857762e-05,
      "loss": 2.3709,
      "step": 636
    },
    {
      "epoch": 0.26329636316739746,
      "grad_norm": 0.6213721036911011,
      "learning_rate": 8.815072153254195e-05,
      "loss": 2.3698,
      "step": 638
    },
    {
      "epoch": 0.2641217436161981,
      "grad_norm": 0.6705930233001709,
      "learning_rate": 8.806236636130411e-05,
      "loss": 2.329,
      "step": 640
    },
    {
      "epoch": 0.2649471240649987,
      "grad_norm": 0.4926806688308716,
      "learning_rate": 8.797372763029742e-05,
      "loss": 2.3431,
      "step": 642
    },
    {
      "epoch": 0.26577250451379936,
      "grad_norm": 0.5596938729286194,
      "learning_rate": 8.78848059998691e-05,
      "loss": 2.3888,
      "step": 644
    },
    {
      "epoch": 0.26659788496259995,
      "grad_norm": 0.4732595980167389,
      "learning_rate": 8.779560213247395e-05,
      "loss": 2.3267,
      "step": 646
    },
    {
      "epoch": 0.26742326541140055,
      "grad_norm": 0.5391475558280945,
      "learning_rate": 8.770611669266938e-05,
      "loss": 2.3869,
      "step": 648
    },
    {
      "epoch": 0.2682486458602012,
      "grad_norm": 0.5030598640441895,
      "learning_rate": 8.76163503471105e-05,
      "loss": 2.3535,
      "step": 650
    },
    {
      "epoch": 0.2690740263090018,
      "grad_norm": 0.6067845225334167,
      "learning_rate": 8.752630376454511e-05,
      "loss": 2.3711,
      "step": 652
    },
    {
      "epoch": 0.26989940675780244,
      "grad_norm": 0.5053196549415588,
      "learning_rate": 8.743597761580877e-05,
      "loss": 2.3643,
      "step": 654
    },
    {
      "epoch": 0.27072478720660303,
      "grad_norm": 0.5403187274932861,
      "learning_rate": 8.734537257381973e-05,
      "loss": 2.3531,
      "step": 656
    },
    {
      "epoch": 0.2715501676554037,
      "grad_norm": 0.498923659324646,
      "learning_rate": 8.7254489313574e-05,
      "loss": 2.3423,
      "step": 658
    },
    {
      "epoch": 0.2723755481042043,
      "grad_norm": 0.5593265295028687,
      "learning_rate": 8.716332851214024e-05,
      "loss": 2.3548,
      "step": 660
    },
    {
      "epoch": 0.2732009285530049,
      "grad_norm": 0.5866305232048035,
      "learning_rate": 8.707189084865481e-05,
      "loss": 2.3677,
      "step": 662
    },
    {
      "epoch": 0.2740263090018055,
      "grad_norm": 0.5718086957931519,
      "learning_rate": 8.698017700431662e-05,
      "loss": 2.3416,
      "step": 664
    },
    {
      "epoch": 0.2748516894506061,
      "grad_norm": 0.6683026552200317,
      "learning_rate": 8.688818766238208e-05,
      "loss": 2.3658,
      "step": 666
    },
    {
      "epoch": 0.27567706989940677,
      "grad_norm": 0.682115912437439,
      "learning_rate": 8.679592350816007e-05,
      "loss": 2.3287,
      "step": 668
    },
    {
      "epoch": 0.27650245034820736,
      "grad_norm": 0.6516245007514954,
      "learning_rate": 8.670338522900672e-05,
      "loss": 2.3095,
      "step": 670
    },
    {
      "epoch": 0.277327830797008,
      "grad_norm": 0.5818968415260315,
      "learning_rate": 8.66105735143204e-05,
      "loss": 2.3354,
      "step": 672
    },
    {
      "epoch": 0.2781532112458086,
      "grad_norm": 0.5822069644927979,
      "learning_rate": 8.651748905553656e-05,
      "loss": 2.3402,
      "step": 674
    },
    {
      "epoch": 0.27897859169460926,
      "grad_norm": 0.5988590717315674,
      "learning_rate": 8.642413254612251e-05,
      "loss": 2.3488,
      "step": 676
    },
    {
      "epoch": 0.27980397214340985,
      "grad_norm": 0.5264620184898376,
      "learning_rate": 8.633050468157234e-05,
      "loss": 2.3336,
      "step": 678
    },
    {
      "epoch": 0.28062935259221045,
      "grad_norm": 0.5173693895339966,
      "learning_rate": 8.623660615940165e-05,
      "loss": 2.3201,
      "step": 680
    },
    {
      "epoch": 0.2814547330410111,
      "grad_norm": 0.5370919108390808,
      "learning_rate": 8.61424376791425e-05,
      "loss": 2.353,
      "step": 682
    },
    {
      "epoch": 0.2822801134898117,
      "grad_norm": 0.5648570656776428,
      "learning_rate": 8.604799994233798e-05,
      "loss": 2.3517,
      "step": 684
    },
    {
      "epoch": 0.28310549393861234,
      "grad_norm": 0.5215715169906616,
      "learning_rate": 8.595329365253719e-05,
      "loss": 2.3333,
      "step": 686
    },
    {
      "epoch": 0.28393087438741293,
      "grad_norm": 0.6153488755226135,
      "learning_rate": 8.585831951528991e-05,
      "loss": 2.3617,
      "step": 688
    },
    {
      "epoch": 0.2847562548362136,
      "grad_norm": 0.6054225564002991,
      "learning_rate": 8.576307823814132e-05,
      "loss": 2.3149,
      "step": 690
    },
    {
      "epoch": 0.2855816352850142,
      "grad_norm": 0.5253807902336121,
      "learning_rate": 8.566757053062678e-05,
      "loss": 2.3114,
      "step": 692
    },
    {
      "epoch": 0.28640701573381483,
      "grad_norm": 0.5196321606636047,
      "learning_rate": 8.55717971042665e-05,
      "loss": 2.3275,
      "step": 694
    },
    {
      "epoch": 0.2872323961826154,
      "grad_norm": 0.5359232425689697,
      "learning_rate": 8.54757586725603e-05,
      "loss": 2.3108,
      "step": 696
    },
    {
      "epoch": 0.288057776631416,
      "grad_norm": 0.5835343599319458,
      "learning_rate": 8.537945595098222e-05,
      "loss": 2.3423,
      "step": 698
    },
    {
      "epoch": 0.28888315708021667,
      "grad_norm": 0.49723029136657715,
      "learning_rate": 8.528288965697523e-05,
      "loss": 2.2839,
      "step": 700
    },
    {
      "epoch": 0.28970853752901726,
      "grad_norm": 0.5888681411743164,
      "learning_rate": 8.518606050994591e-05,
      "loss": 2.3485,
      "step": 702
    },
    {
      "epoch": 0.2905339179778179,
      "grad_norm": 0.5658313632011414,
      "learning_rate": 8.508896923125901e-05,
      "loss": 2.359,
      "step": 704
    },
    {
      "epoch": 0.2913592984266185,
      "grad_norm": 0.5612741112709045,
      "learning_rate": 8.499161654423219e-05,
      "loss": 2.3482,
      "step": 706
    },
    {
      "epoch": 0.29218467887541916,
      "grad_norm": 0.51905757188797,
      "learning_rate": 8.48940031741305e-05,
      "loss": 2.3349,
      "step": 708
    },
    {
      "epoch": 0.29301005932421975,
      "grad_norm": 0.5525624752044678,
      "learning_rate": 8.479612984816112e-05,
      "loss": 2.3318,
      "step": 710
    },
    {
      "epoch": 0.2938354397730204,
      "grad_norm": 0.6043487191200256,
      "learning_rate": 8.469799729546781e-05,
      "loss": 2.3399,
      "step": 712
    },
    {
      "epoch": 0.294660820221821,
      "grad_norm": 0.5291466116905212,
      "learning_rate": 8.459960624712556e-05,
      "loss": 2.3175,
      "step": 714
    },
    {
      "epoch": 0.2954862006706216,
      "grad_norm": 0.5733122229576111,
      "learning_rate": 8.450095743613512e-05,
      "loss": 2.3366,
      "step": 716
    },
    {
      "epoch": 0.29631158111942224,
      "grad_norm": 0.5417082905769348,
      "learning_rate": 8.440205159741752e-05,
      "loss": 2.3335,
      "step": 718
    },
    {
      "epoch": 0.29713696156822283,
      "grad_norm": 0.565000593662262,
      "learning_rate": 8.430288946780865e-05,
      "loss": 2.3384,
      "step": 720
    },
    {
      "epoch": 0.2979623420170235,
      "grad_norm": 0.6104756593704224,
      "learning_rate": 8.420347178605367e-05,
      "loss": 2.3587,
      "step": 722
    },
    {
      "epoch": 0.2987877224658241,
      "grad_norm": 0.5492483973503113,
      "learning_rate": 8.410379929280168e-05,
      "loss": 2.3169,
      "step": 724
    },
    {
      "epoch": 0.29961310291462473,
      "grad_norm": 0.5351945161819458,
      "learning_rate": 8.400387273059998e-05,
      "loss": 2.3195,
      "step": 726
    },
    {
      "epoch": 0.3004384833634253,
      "grad_norm": 0.5342651605606079,
      "learning_rate": 8.39036928438887e-05,
      "loss": 2.2912,
      "step": 728
    },
    {
      "epoch": 0.301263863812226,
      "grad_norm": 0.5106682181358337,
      "learning_rate": 8.380326037899522e-05,
      "loss": 2.3283,
      "step": 730
    },
    {
      "epoch": 0.30208924426102657,
      "grad_norm": 0.5767691731452942,
      "learning_rate": 8.370257608412857e-05,
      "loss": 2.3286,
      "step": 732
    },
    {
      "epoch": 0.30291462470982716,
      "grad_norm": 0.54031902551651,
      "learning_rate": 8.360164070937389e-05,
      "loss": 2.3243,
      "step": 734
    },
    {
      "epoch": 0.3037400051586278,
      "grad_norm": 0.6625233888626099,
      "learning_rate": 8.350045500668681e-05,
      "loss": 2.3232,
      "step": 736
    },
    {
      "epoch": 0.3045653856074284,
      "grad_norm": 0.6196743845939636,
      "learning_rate": 8.339901972988795e-05,
      "loss": 2.3216,
      "step": 738
    },
    {
      "epoch": 0.30539076605622906,
      "grad_norm": 0.5990370512008667,
      "learning_rate": 8.329733563465711e-05,
      "loss": 2.3401,
      "step": 740
    },
    {
      "epoch": 0.30621614650502965,
      "grad_norm": 0.5713345408439636,
      "learning_rate": 8.319540347852786e-05,
      "loss": 2.3416,
      "step": 742
    },
    {
      "epoch": 0.3070415269538303,
      "grad_norm": 0.5813875198364258,
      "learning_rate": 8.30932240208817e-05,
      "loss": 2.2979,
      "step": 744
    },
    {
      "epoch": 0.3078669074026309,
      "grad_norm": 0.6877428889274597,
      "learning_rate": 8.299079802294258e-05,
      "loss": 2.317,
      "step": 746
    },
    {
      "epoch": 0.30869228785143155,
      "grad_norm": 0.5318434834480286,
      "learning_rate": 8.288812624777109e-05,
      "loss": 2.2947,
      "step": 748
    },
    {
      "epoch": 0.30951766830023214,
      "grad_norm": 0.5925495624542236,
      "learning_rate": 8.278520946025884e-05,
      "loss": 2.3114,
      "step": 750
    },
    {
      "epoch": 0.31034304874903273,
      "grad_norm": 0.5941009521484375,
      "learning_rate": 8.268204842712278e-05,
      "loss": 2.3463,
      "step": 752
    },
    {
      "epoch": 0.3111684291978334,
      "grad_norm": 0.5976232886314392,
      "learning_rate": 8.25786439168994e-05,
      "loss": 2.2996,
      "step": 754
    },
    {
      "epoch": 0.311993809646634,
      "grad_norm": 0.5421935319900513,
      "learning_rate": 8.24749966999391e-05,
      "loss": 2.3143,
      "step": 756
    },
    {
      "epoch": 0.31281919009543463,
      "grad_norm": 0.5300918817520142,
      "learning_rate": 8.237110754840043e-05,
      "loss": 2.2946,
      "step": 758
    },
    {
      "epoch": 0.3136445705442352,
      "grad_norm": 0.5926389098167419,
      "learning_rate": 8.22669772362443e-05,
      "loss": 2.3312,
      "step": 760
    },
    {
      "epoch": 0.3144699509930359,
      "grad_norm": 0.5365331768989563,
      "learning_rate": 8.216260653922823e-05,
      "loss": 2.3165,
      "step": 762
    },
    {
      "epoch": 0.31529533144183647,
      "grad_norm": 0.5278505086898804,
      "learning_rate": 8.205799623490055e-05,
      "loss": 2.2811,
      "step": 764
    },
    {
      "epoch": 0.31612071189063706,
      "grad_norm": 0.5568265318870544,
      "learning_rate": 8.195314710259475e-05,
      "loss": 2.2899,
      "step": 766
    },
    {
      "epoch": 0.3169460923394377,
      "grad_norm": 0.6386959552764893,
      "learning_rate": 8.184805992342342e-05,
      "loss": 2.3193,
      "step": 768
    },
    {
      "epoch": 0.3177714727882383,
      "grad_norm": 0.5363957285881042,
      "learning_rate": 8.174273548027262e-05,
      "loss": 2.3255,
      "step": 770
    },
    {
      "epoch": 0.31859685323703896,
      "grad_norm": 0.6581987142562866,
      "learning_rate": 8.163717455779602e-05,
      "loss": 2.3159,
      "step": 772
    },
    {
      "epoch": 0.31942223368583955,
      "grad_norm": 0.7844798564910889,
      "learning_rate": 8.153137794240903e-05,
      "loss": 2.3241,
      "step": 774
    },
    {
      "epoch": 0.3202476141346402,
      "grad_norm": 0.5486951470375061,
      "learning_rate": 8.142534642228288e-05,
      "loss": 2.2834,
      "step": 776
    },
    {
      "epoch": 0.3210729945834408,
      "grad_norm": 0.6187033653259277,
      "learning_rate": 8.13190807873389e-05,
      "loss": 2.31,
      "step": 778
    },
    {
      "epoch": 0.32189837503224145,
      "grad_norm": 0.6599840521812439,
      "learning_rate": 8.121258182924247e-05,
      "loss": 2.3075,
      "step": 780
    },
    {
      "epoch": 0.32272375548104204,
      "grad_norm": 0.6156419515609741,
      "learning_rate": 8.110585034139723e-05,
      "loss": 2.3171,
      "step": 782
    },
    {
      "epoch": 0.32354913592984264,
      "grad_norm": 0.652897834777832,
      "learning_rate": 8.099888711893917e-05,
      "loss": 2.2963,
      "step": 784
    },
    {
      "epoch": 0.3243745163786433,
      "grad_norm": 0.5405826568603516,
      "learning_rate": 8.089169295873058e-05,
      "loss": 2.299,
      "step": 786
    },
    {
      "epoch": 0.3251998968274439,
      "grad_norm": 0.5489581823348999,
      "learning_rate": 8.078426865935432e-05,
      "loss": 2.3051,
      "step": 788
    },
    {
      "epoch": 0.32602527727624453,
      "grad_norm": 0.5330743789672852,
      "learning_rate": 8.067661502110768e-05,
      "loss": 2.306,
      "step": 790
    },
    {
      "epoch": 0.3268506577250451,
      "grad_norm": 0.5833327174186707,
      "learning_rate": 8.056873284599648e-05,
      "loss": 2.3197,
      "step": 792
    },
    {
      "epoch": 0.3276760381738458,
      "grad_norm": 0.638880729675293,
      "learning_rate": 8.046062293772922e-05,
      "loss": 2.309,
      "step": 794
    },
    {
      "epoch": 0.32850141862264637,
      "grad_norm": 0.6887226104736328,
      "learning_rate": 8.035228610171085e-05,
      "loss": 2.3192,
      "step": 796
    },
    {
      "epoch": 0.329326799071447,
      "grad_norm": 0.5981518030166626,
      "learning_rate": 8.024372314503701e-05,
      "loss": 2.2958,
      "step": 798
    },
    {
      "epoch": 0.3301521795202476,
      "grad_norm": 0.5953544974327087,
      "learning_rate": 8.013493487648782e-05,
      "loss": 2.3161,
      "step": 800
    },
    {
      "epoch": 0.3309775599690482,
      "grad_norm": 0.576503574848175,
      "learning_rate": 8.002592210652202e-05,
      "loss": 2.2901,
      "step": 802
    },
    {
      "epoch": 0.33180294041784886,
      "grad_norm": 0.6154365539550781,
      "learning_rate": 7.991668564727082e-05,
      "loss": 2.3093,
      "step": 804
    },
    {
      "epoch": 0.33262832086664945,
      "grad_norm": 0.5416200757026672,
      "learning_rate": 7.98072263125319e-05,
      "loss": 2.3004,
      "step": 806
    },
    {
      "epoch": 0.3334537013154501,
      "grad_norm": 0.5973731875419617,
      "learning_rate": 7.969754491776329e-05,
      "loss": 2.3224,
      "step": 808
    },
    {
      "epoch": 0.3342790817642507,
      "grad_norm": 0.5144022107124329,
      "learning_rate": 7.958764228007741e-05,
      "loss": 2.2899,
      "step": 810
    },
    {
      "epoch": 0.33510446221305135,
      "grad_norm": 0.627142608165741,
      "learning_rate": 7.947751921823488e-05,
      "loss": 2.3196,
      "step": 812
    },
    {
      "epoch": 0.33592984266185194,
      "grad_norm": 0.5556638240814209,
      "learning_rate": 7.936717655263841e-05,
      "loss": 2.2991,
      "step": 814
    },
    {
      "epoch": 0.3367552231106526,
      "grad_norm": 0.5637221336364746,
      "learning_rate": 7.925661510532681e-05,
      "loss": 2.3093,
      "step": 816
    },
    {
      "epoch": 0.3375806035594532,
      "grad_norm": 0.6272327899932861,
      "learning_rate": 7.91458356999687e-05,
      "loss": 2.3286,
      "step": 818
    },
    {
      "epoch": 0.3384059840082538,
      "grad_norm": 0.5124315023422241,
      "learning_rate": 7.903483916185654e-05,
      "loss": 2.2879,
      "step": 820
    },
    {
      "epoch": 0.33923136445705443,
      "grad_norm": 0.5467550158500671,
      "learning_rate": 7.892362631790035e-05,
      "loss": 2.3107,
      "step": 822
    },
    {
      "epoch": 0.340056744905855,
      "grad_norm": 0.5180369019508362,
      "learning_rate": 7.881219799662164e-05,
      "loss": 2.2784,
      "step": 824
    },
    {
      "epoch": 0.3408821253546557,
      "grad_norm": 0.5402101874351501,
      "learning_rate": 7.870055502814714e-05,
      "loss": 2.2704,
      "step": 826
    },
    {
      "epoch": 0.34170750580345627,
      "grad_norm": 0.4879278242588043,
      "learning_rate": 7.858869824420272e-05,
      "loss": 2.2913,
      "step": 828
    },
    {
      "epoch": 0.3425328862522569,
      "grad_norm": 0.5525892972946167,
      "learning_rate": 7.847662847810713e-05,
      "loss": 2.3071,
      "step": 830
    },
    {
      "epoch": 0.3433582667010575,
      "grad_norm": 0.517867922782898,
      "learning_rate": 7.836434656476583e-05,
      "loss": 2.2748,
      "step": 832
    },
    {
      "epoch": 0.34418364714985816,
      "grad_norm": 0.5947239398956299,
      "learning_rate": 7.825185334066475e-05,
      "loss": 2.2687,
      "step": 834
    },
    {
      "epoch": 0.34500902759865876,
      "grad_norm": 0.6157646775245667,
      "learning_rate": 7.813914964386401e-05,
      "loss": 2.2769,
      "step": 836
    },
    {
      "epoch": 0.34583440804745935,
      "grad_norm": 0.5485296845436096,
      "learning_rate": 7.802623631399176e-05,
      "loss": 2.2663,
      "step": 838
    },
    {
      "epoch": 0.34665978849626,
      "grad_norm": 0.4869697093963623,
      "learning_rate": 7.791311419223791e-05,
      "loss": 2.2908,
      "step": 840
    },
    {
      "epoch": 0.3474851689450606,
      "grad_norm": 0.5534776449203491,
      "learning_rate": 7.779978412134783e-05,
      "loss": 2.2998,
      "step": 842
    },
    {
      "epoch": 0.34831054939386125,
      "grad_norm": 0.5947521328926086,
      "learning_rate": 7.768624694561604e-05,
      "loss": 2.2907,
      "step": 844
    },
    {
      "epoch": 0.34913592984266184,
      "grad_norm": 0.5616730451583862,
      "learning_rate": 7.757250351088004e-05,
      "loss": 2.3003,
      "step": 846
    },
    {
      "epoch": 0.3499613102914625,
      "grad_norm": 0.5840707421302795,
      "learning_rate": 7.745855466451385e-05,
      "loss": 2.2683,
      "step": 848
    },
    {
      "epoch": 0.3507866907402631,
      "grad_norm": 0.5235263705253601,
      "learning_rate": 7.734440125542186e-05,
      "loss": 2.2976,
      "step": 850
    },
    {
      "epoch": 0.35161207118906374,
      "grad_norm": 0.5673782229423523,
      "learning_rate": 7.723004413403238e-05,
      "loss": 2.2936,
      "step": 852
    },
    {
      "epoch": 0.35243745163786433,
      "grad_norm": 0.5472978353500366,
      "learning_rate": 7.711548415229135e-05,
      "loss": 2.2896,
      "step": 854
    },
    {
      "epoch": 0.3532628320866649,
      "grad_norm": 0.5697288513183594,
      "learning_rate": 7.700072216365602e-05,
      "loss": 2.3239,
      "step": 856
    },
    {
      "epoch": 0.3540882125354656,
      "grad_norm": 0.6162300705909729,
      "learning_rate": 7.688575902308854e-05,
      "loss": 2.2973,
      "step": 858
    },
    {
      "epoch": 0.35491359298426617,
      "grad_norm": 0.4970763027667999,
      "learning_rate": 7.677059558704965e-05,
      "loss": 2.3018,
      "step": 860
    },
    {
      "epoch": 0.3557389734330668,
      "grad_norm": 0.486848920583725,
      "learning_rate": 7.665523271349221e-05,
      "loss": 2.3116,
      "step": 862
    },
    {
      "epoch": 0.3565643538818674,
      "grad_norm": 0.5172179341316223,
      "learning_rate": 7.65396712618549e-05,
      "loss": 2.2898,
      "step": 864
    },
    {
      "epoch": 0.35738973433066806,
      "grad_norm": 0.5573694705963135,
      "learning_rate": 7.642391209305581e-05,
      "loss": 2.2866,
      "step": 866
    },
    {
      "epoch": 0.35821511477946866,
      "grad_norm": 0.540338397026062,
      "learning_rate": 7.630795606948592e-05,
      "loss": 2.3066,
      "step": 868
    },
    {
      "epoch": 0.3590404952282693,
      "grad_norm": 0.5216573476791382,
      "learning_rate": 7.619180405500284e-05,
      "loss": 2.2489,
      "step": 870
    },
    {
      "epoch": 0.3598658756770699,
      "grad_norm": 0.5518911480903625,
      "learning_rate": 7.607545691492421e-05,
      "loss": 2.2621,
      "step": 872
    },
    {
      "epoch": 0.3606912561258705,
      "grad_norm": 0.5395997166633606,
      "learning_rate": 7.595891551602139e-05,
      "loss": 2.2828,
      "step": 874
    },
    {
      "epoch": 0.36151663657467115,
      "grad_norm": 0.5570027828216553,
      "learning_rate": 7.584218072651291e-05,
      "loss": 2.2721,
      "step": 876
    },
    {
      "epoch": 0.36234201702347174,
      "grad_norm": 0.530053436756134,
      "learning_rate": 7.572525341605805e-05,
      "loss": 2.3146,
      "step": 878
    },
    {
      "epoch": 0.3631673974722724,
      "grad_norm": 0.5206712484359741,
      "learning_rate": 7.560813445575032e-05,
      "loss": 2.2633,
      "step": 880
    },
    {
      "epoch": 0.363992777921073,
      "grad_norm": 0.5709933638572693,
      "learning_rate": 7.549082471811105e-05,
      "loss": 2.2524,
      "step": 882
    },
    {
      "epoch": 0.36481815836987364,
      "grad_norm": 0.5382503271102905,
      "learning_rate": 7.53733250770828e-05,
      "loss": 2.2871,
      "step": 884
    },
    {
      "epoch": 0.36564353881867423,
      "grad_norm": 0.49123620986938477,
      "learning_rate": 7.525563640802286e-05,
      "loss": 2.2906,
      "step": 886
    },
    {
      "epoch": 0.3664689192674748,
      "grad_norm": 0.48953843116760254,
      "learning_rate": 7.513775958769683e-05,
      "loss": 2.2912,
      "step": 888
    },
    {
      "epoch": 0.3672942997162755,
      "grad_norm": 0.5467056632041931,
      "learning_rate": 7.501969549427195e-05,
      "loss": 2.2961,
      "step": 890
    },
    {
      "epoch": 0.36811968016507607,
      "grad_norm": 0.5301052331924438,
      "learning_rate": 7.49014450073106e-05,
      "loss": 2.2588,
      "step": 892
    },
    {
      "epoch": 0.3689450606138767,
      "grad_norm": 0.5778504014015198,
      "learning_rate": 7.478300900776387e-05,
      "loss": 2.2627,
      "step": 894
    },
    {
      "epoch": 0.3697704410626773,
      "grad_norm": 0.5600552558898926,
      "learning_rate": 7.46643883779648e-05,
      "loss": 2.2874,
      "step": 896
    },
    {
      "epoch": 0.37059582151147796,
      "grad_norm": 0.5173321962356567,
      "learning_rate": 7.454558400162195e-05,
      "loss": 2.2787,
      "step": 898
    },
    {
      "epoch": 0.37142120196027856,
      "grad_norm": 0.534444272518158,
      "learning_rate": 7.442659676381275e-05,
      "loss": 2.2883,
      "step": 900
    }
  ],
  "logging_steps": 2,
  "max_steps": 2424,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 300,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.0132651008589824e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}